E QUINE L OCOMOTION

Commissioning Editor: Robert Edwards

Development Editor: Louisa Talbott/Catherine Jackson
Project Manager: Julie Taylor
Designer/Design Direction: Stewart Larking
Illustration Manager: Jennifer Rose
Illustrator: Antbits Ltd
EQUINE S ECOND E DITION

LOCOMOTION
Edited by

Willem Back
DVM, Cert. KNMvD (CKRD), Cert. Pract. KNMvD (Equine Practice), PhD, Spec. KNMvD
(Equine Surgery), Dipl. ECVS, Prof. (U Ghent)
Department of Equine Sciences, Faculty of Veterinary Medicine
Utrecht University
Utrecht, The Netherlands
and
Department of Surgery and Anaesthesiology of Domestic Animals
Faculty of Veterinary Medicine
Ghent University
Merelbeke, Belgium

Hilary M. Clayton
BVMS, PhD, Dipl. ACVSMR, MRCVS
McPhail Dressage Chair in Equine Sports Medicine
College of Veterinary Medicine, Michigan State University
East Lansing, MI, USA

Foreword by
Peter D. Rossdale
MA, PhD, Dr.h.c. (Berne, Edinburgh, Sydney), DESM, FRCVS

Edinburgh  London  New York  Oxford  Philadelphia  St Louis  Sydney  Toronto  2013
© 2013 Elsevier Ltd. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopying, recording, or any information storage and retrieval system, without permission in writing from
the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our
arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be
found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as
may be noted herein).

First edition 2000

Second edition 2013

ISBN 978-0-7020-2950-9

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data

A catalog record for this book is available from the Library of Congress

Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understand-
ing, changes in research methods, professional practices, or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any
information, methods, compounds, or experiments described herein. In using such information or methods they should
be mindful of their own safety and the safety of others, including parties for whom they have a professional
responsibility.
With respect to any drug or pharmaceutical products identified, readers are advised to check the most current informa-
tion provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the rec-
ommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of
practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages
and the best treatment for each individual patient, and to take all appropriate safety precautions.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any
injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or
operation of any methods, products, instructions, or ideas contained in the material herein.

The
Working together to grow publisher’s
policy is to use
libraries in developing countries paper manufactured
from sustainable forests
www.elsevier.com | www.bookaid.org | www.sabre.org

Printed in China
 Contents

Foreword vii
Foreword to the first edition ix
Preface xi
Preface to the first edition xiii
Acknowledgments xv
Contributors xvii
Glossary xxi
Nancy R. Deuel, Hilary M. Clayton

1. History 1
P. René van Weeren
2. Measurement techniques for gait analysis 31
Hilary M. Clayton, Henk C. Schamhardt (deceased)
3. Signals from materials 61
Christian Peham
4. Locomotor neurobiology and development 73
Albert Gramsbergen
5. Gaits and interlimb coordination 85
Eric Barrey
6. Forelimb function 99
Hilary M. Clayton, Henry Chateau, Willem Back
7. Hind limb function 127
Hilary M. Clayton, Willem Back
8. The role of the hoof and shoeing 147
Willem Back, Frederik Pille
9. Gait adaptation in lameness 175
H.H. Florian Buchner
10. The neck and back 199
Claudia Wolschrijn, Fabrice Audigié, Inge D. Wijnberg, Christopher Johnston,
Jean-Marie Denoix, Willem Back
11. The effects of conformation 229
Mikael Holmström, Willem Back
12. Genetic contributions to exercise and athletic performance 245
Emmeline W. Hill, Bart J. Ducro, P. René van Weeren, Albert Barneveld, Willem Back
13. The response of musculoskeletal tissues to exercise 267
Helen L. Birch, Pieter A.J. Brama, Elwyn C. Firth, Allen E. Goodship, José Luis L. Rivero,
P. René van Weeren
14. Performance in equestrian sports 305
Hilary M. Clayton, P. René van Weeren
15. Horse–rider interaction 341
Agneta Egenvall, Anna Byström, Michael Weishaupt, Lars Roepstorff
16. Ethology and welfare aspects 369
Natalie K. Waran, Machteld C. Van Dierendonck
17. Rehabilitation of the locomotor apparatus 381
Narelle Stubbs, Eveline Menke, Willem Back, Hilary M. Clayton

v
 Contents

18. Metabolic energetics 419

Mathew P. Gerard, E. de Graaf-Roelfsema, David R. Hodgson, J.H. Van der Kolk
19. Mechanical analysis and scaling 443
Antonie J. Van den Bogert, Liduin S. Meershoek, At L. Hof
20. Modeling, simulation and animation 467
Adam Arabian, Jonathan Merritt, Franca Jonquiere, Antonie J. Van den Bogert,
Hilary M. Clayton

Index 489

vi
 Foreword

Scientists strive to explain the relationship between cause and effect; two decades, science has developed the technology on which our
in the context of locomotion, structure and function. Those in understanding has expanded enormously, not only of the mechani-
practice, such as clinicians and horse trainers, are concerned with cal attributes of the horse’s musculoskeletal system, but also of the
the application of scientific findings, but this requires monitoring biological and pathological processes involved in the structures on
in the spirit of evidence-based medicine (EBM). which equine ‘athleticism’ is based. This is a never-ending progres-
In the foreword to the first edition of this book, Professor McNeill sion in keeping with scientific method and is likely to continue
Alexander FRS drew attention to early studies initiated by the pho- indefinitely in the future. The content of the present edition of
tography of Muybridge (1899) and of digital ligaments by Camp Equine Locomotion represents a significant step in this pathway.
and Smith (1942). James Gray (1936) proposed the paradox as to Consultation with two colleagues active in practice led to the
how dolphins obtain high speeds and acceleration with such small following personal communications:
muscle mass against the drag forces of water. He hypothesized that
their skin must have special anti-drag properties. The study of locomotion should be applied to the young
This was subsequently discredited as summarized by Wei (2008), racehorse to help circumvent injury or lameness developing as,
but serves to illustrate the biological need to elucidate the features for example, in the specific condition of plantar osteochondral
of structure and function. From the viewpoint of a veterinary clini- disease (POD) or more simply a stress-related bone injury to
cian, this is the essential contribution of locomotory science with the lower cannon. Science might guide us towards different
the objective of improving performance and reducing the likeli- training methods and programmes, shoeing, surfaces, tack or
hood of injury, or improving the chances of successful treatment if other unknowns to improve the soundness and welfare of the
injury occurs; thereby offering overall benefits in the welfare of horse and reduce wastage and days lost from training.
horses.
M.C. Shepherd BVSc Massey, MRCVS
The introduction of techniques applied to elucidate the mysteries
of movement illustrates the advances reported in the present book.
There is, for example, very substantial potential for telemetry involv-
ing inertial motion sensors which, especially when combined with Breeders of modern sport horses are currently selecting for
global positioning system (GPS) technology, make possible the traits that adapt the individual to the needs for a chosen
collection of a great deal of data about movement and limb kine- discipline. With this, biomechanically specific pathologies
matics, such as reported in the study of Pfau et al. (2006). typical of a discipline, its training and competition conditions
Viewpoints differ according to the circumstances of those involved are challenging and taxing the mental and physical limits of
in exercise physiology and locomotion: the scientist pursues the the animal in its artificial environment. The understanding
essential elements of knowledge. Veterinary clinicians require the of the changing demands each discipline exerts on the
means to the end of diagnosis, prophylaxis, therapy and prognosis; phenotypical development of the more slowly changing genotype
the owner uses knowledge to improve the return on investment; is essential to the welfare of the horses and sustained sporting
and the trainer of horses, be they for racing, competition or riding, and economic success of a discipline. Learning about ground
needs to improve on methods to achieve the purpose for which the and force interaction, evolving equitation and training ideas
horse is kept. on repetitive biomechanical stresses and their effects on
The questions of practice include, why do individuals differ as to skeletal, articular and muscular sustainable long term athletic
their performance or susceptibility to injury on differing track sur- health is essential for the sporting success of the current and
faces, and what in the structure or body function of the individual future generations of show jumpers, eventers, dressage as well
determines sprinting or stamina as a feature? as endurance, driving, reining horses and polo ponies.
The modern horse is, somewhat paradoxically, the product of F.E. Barrelet DrVetMed Berne, MRCVS
man’s interference with the evolutionary process due to our selec-
tion for various roles, ranging from the drawing and carrying of The history of scientific progress is described eloquently in the
loads to the participation in war and war-like activities, including first chapter by René van Weeren. There are then nineteen chapters
chariot racing, and, more recently, on the racecourse, in endurance devoted to the application of scientific methods and measurement
and in eventing competitions. But the relationship between the techniques regarding limb and bodily function in motion. The list
biological composition of the individual and the stresses and strains of 45 contributors from twelve different countries and three conti-
of functional challenge imposed must, in large measure, be consid- nents is evidence of the substantial weight of scientific input into
ered unnatural as far as outcome of performance and susceptibility the subject of equine locomotion; and lends emphasis to the impor-
to injury are concerned. tance of the book and the successful endeavours of the two co-
Up until about the 1970s, the disciplines of locomotion and editors, Wim Back and Hilary Clayton, who deserve congratulations
exercise physiology were largely in the hands of clinicians and from all concerned, contributors and readers alike.
horsemen, who based their diagnostic and therapeutic measures
upon experience and trial and error associated with outcome. The
Peter D. Rossdale
introduction of scientific methods has not only solved problems,
but also challenged those responsible for the care and well-being Emeritus Editor Equine Veterinary Journal
of horses for whatever purpose they are used. Particularly in the past Rossdale and Partners, Newmarket, UK

vii
 Foreword

References
Camp, C.L., Smith, N., 1942. Phylogeny and Muybridge, E., 1899. Animals in motion. Wei, T., 2008. Gray’s paradox solved:
function of the digital ligaments of the Chapman & Hall, London. researchers discover secret of speedy
horse. University of California Memoirs Pfau, T., Witte, T.H., Wilson, A.M., 2006. dolphins.
Zoology 13, 69–124. Centre of mass movement and mechanical Science Daily. Available at: http://www.
Gray, J., 1936. Studies in animal locomotion energy fluctuation during gallop sciencedaily.com/releases/2008/11/
V1: The propulsive powers of the dolphin. locomotion in the Thoroughbred racehorse. 081124131334.htm. Accessed 11 November
J. Expt. Biol. 13, 192–199. J. Exp. Biol. 209, 3742–3757. 2009.

viii
 Foreword
to the first edition

Horse locomotion has lost most of the practical importance it had valuable insight into the stresses that bones have to withstand when
in past centuries, but a great many people remain passionately surgically implanted strain gauges were used to record strains in the
interested in horses for leisure, racing or other purposes. This, and lower leg bones of running and jumping horses (Biewener et al.,
the high value of horses, sustain veterinary interest in the species. 1983). It was in a study of horses that Rome et al., (1990) showed
Horses have also attracted the interest of many research scientists. us the remarkable range of properties that can be found within a
Muybridge’s (1899) sequences of photographs of horses in motion, single muscle. Some fibers in the soleus muscle of horses are capable
taken with multiple cameras before the cine camera had been of contracting ten times faster than others. And observations on
invented, have an honored place in the history of motion pictures. horses were the first to show that tendons can be damaged by
Camp and Smith’s (1942) study of the digital ligaments of horses overheating, by heat liberated by the repeated stretching and recoil,
was an early recognition of the importance of tendon and ligament that occurs in running (Wilson & Goodship, 1994).
elasticity in running; horses, people, dogs, kangaroos and many Besides these studies that have wide significance, there have been
other animals save energy by bouncing along, using their tendons others, equally fascinating, that concern horses alone. The recipro-
like the spring of a child’s pogo stick. Horses were particularly good cal apparatus in the hind limb makes movements of the stifle joint
material for this study because their adaptations for elastic energy drive those of the hock (van Weeren et al., 1992). The hooves of
savings are surpassed only by the camel (Dimery et al., 1986). horses have been shown to be designed in a remarkably sophisti-
Recent research provides many examples of research on horses cated way, to withstand impact on the ground (Thomason et al.,
that has wide significance for the understanding of animal and 1992). The hock joint of horses has bistable properties that make
human movement. I will cite just a few of my favourites. it click like an electric switch from one extreme position to the
Most quadrupedal mammals walk, trot and gallop, but the sig- other, a property which, curiously, is much more marked in domes-
nificance of changing gaits became apparent only when Hoyt and tic horses than in zebra or Przewalsky’s horse (Alexander & Trestik,
Taylor (1981) trained ponies to change gait on command, so that 1989).
they could be made (for example) to trot at speeds at which they These examples make it clear that horse locomotion has inspired
would have preferred to walk, and vice versa. By measurements of a great deal of excellent and interesting science, much of which
oxygen consumption they showed that walking is the gait that needs throws light on the biology of other animals, as well as of horses.
least energy at low speeds, trotting at intermediate speeds, and gal- That is one reason why this book will be so welcome. Another is
loping at high speeds, and that each gait is used in the range of that it has been written by a carefully chosen team of scientists
speeds in which it is most economical. Bramble and Carrier (1983) whose research has added substantially to our knowledge of horse
depended largely on observations of horses for their demonstration locomotion.
that galloping mammals take one breath per stride, their breathing Sadly, the team has been depleted by the death of Henk
apparently driven by the movements of locomotion. They specu- Schamhardt, co-author of one of our chapters, who was pre-
lated that this might depend on the viscera functioning as an iner- eminent among researchers in the field of equine biomechanics.
tial piston, shifting forward and back in the trunk as the animal
accelerated and decelerated in the course of each stride. Disappoint- R. McNeill Alexander FRS
ingly, this seems not to be the case; the breathing of a galloping
horse seems to be driven by the bending and extension of the back, Emeritus Professor of Zoology
functioning like a bellows (Young et al., 1992). Horses gave us University of Leeds, UK

References
Alexander, R.McN., Trestik, C.L., 1989. Bistable the locomotion of horses (Equus caballus). van Weeren, P.R., Jansen, M.O., van den
properties of the hock joint of horses. J. J. Zool. 210, 415–425. Bogert, A.J., Barneveld, A., 1992. A
Zool. 218, 383–391. Hoyt, D.F., Taylor, C.R., 1981. Gait and the kinematic and strain gauge study of the
Biewener, A.A., Thomason, J., Lanyon, L.E., energetics of locomotion in horses. reciprocal apparatus in the equine hind
1983. Mechanics of locomotion and Nature 292, 239–240. limb. J. Biomech. 25, 1291–1301.
jumping in the forelimb of the horse Muybridge, E., 1899. Animals in Motion. Wilson, A.M., Goodship, A.E., 1994.
(Equus): In vivo stress developed in the Chapman and Hall, London. Exercise-induced hyperthermia as a
radius and metacarpus. J. Zool. 201, 67–82. Rome, L.C., Sosnicki, A.A., Goble, D.O., 1990. possible mechanism for tendon
Bramble, D.M., Carrier, D.R., 1983. Running Maximum velocity of shortening of three degeneration. J. Biomech. 27, 899–905.
and breathing in mammals. Science 219, fibre types from horse soleus muscle: Young, I.S., Alexander, R.McN., Woakes, A.J.,
251–256. implications for scaling with body size. J. Butler, P.J., Anderson, L., 1992. The
Camp, C.L., Smith, N., 1942. Phylogeny and Physiol. 431, 173–185. synchronisation of ventilation and
function of the digital ligaments of the Thomason, J.J., Biewener, A.A., Bertram, J.E.A., locomotion in horses (Equus caballus).
horse. University of California Memoirs 1992. Surface strain on the equine hoof J. Exp. Biol. 166, 19–31.
in Zoology 13, 69–124. wall in vivo: implications for the material
Dimery, N.J., Alexander, R.McN., Ker, R.F., design and functional morphology of the
1986. Elastic extensions of leg tendons in wall. J. Exp. Biol. 166, 145–168.

ix
This page intentionally left blank
 Preface

The repercussions of societal revolutions are felt far and wide, and and gamers. The new chapter on modeling, simulation and anima-
the equine industry is certainly not immune. The industrial revolu- tion draws on the expertise of multiple authors to adequately re­
tion brought major changes in agriculture and transportation that present this exciting area.
were the impetus for a change in the predominant use of horses Considerable new information has been added to the chapters
from work to sport and recreation. The baby boomer generation on signals from materials, genetic contributions to exercise and
adopted a fitness craze that ultimately resulted in an increase in performance, the horse–rider interaction, locomotion ethology and
sport-related injuries and a growing need for sports medicine prac- welfare, and rehabilitation. These chapters present state-of-the-art
titioners. Equestrian sports were also growing in popularity and information from the researchers who are at the forefront in these
there was a demand for scientific information about the structure, areas.
function and diseases of sport horses. Production of this text was timed so that the chapter authors
Gait analysis has always been recognized as a key element in could include new information presented at the International Con-
understanding equine athletic performance but it was not until the ference on Equine Exercise Physiology in November 2010. At this
technological revolution that the tools became available to make conference, biomechanics was the strongest subject area; there were
rapid progress in this area. Biomechanists now have at their disposal numerous excellent presentations made at the conference and
a dazzling array of equipment and techniques to analyze equine published in the proceedings book. This second edition of Equine
locomotion. Consequently the past decade has seen tremendous Locomotion truly represents the current state of knowledge about
growth in the body of literature in this area. Since the first edition equine locomotion and the multitude of factors that can influence
of Equine Locomotion was published in 2001, the knowledge base locomotor performance in the horse.
has expanded in the traditional areas and many new foci of inves- As the editors of this edition we thank all our contributors for
tigation have emerged. Therefore, there was no shortage of new their expertise and dedication in making the information so readily
information for the second edition. available to all those who have an interest in this fascinating subject.
In this edition, almost every chapter has undergone substantial
revision to include new information. Notable among these is the Willem Back (Utrecht, The Netherlands)
chapter on modeling of locomotion, an area that was in its infancy
a decade ago but has now become much more mainstream and Hilary M. Clayton (East Lansing, Michigan, USA)
accessible to researchers rather than being the domain of animators 2013

xi
This page intentionally left blank
 Preface
to the first edition

As a result of their diverse athletic abilities, horses have been used reader a complete picture of the horse in motion, as we now know
as beasts of burden, as vehicles of war and as partners in sports and it. The book begins with a history of man’s association with the
recreation. The past century has seen an explosion in the popularity horse both in sports and in veterinary medicine, which sets the stage
of equestrian sports with a concomitant increase in competition for a comprehensive description of the present state of knowledge
standards. The expectations for performance in today’s competition beginning with the initiation of gait and ending with the more
horse require a high level of care and training, which can be futuristic area of computer modeling. In areas where studies of
achieved only through a comprehensive understanding of the horses are lacking, ideas from other species have been introduced.
anatomy and physiology of the elite equine athlete. The scientific The list of authors comprises individuals who are acknowledged
community responded to this need for information by embracing experts in their subject areas. We thank them for the enormous time
the discipline of equine sports science, which has developed in and effort they have invested in producing this book.
parallel with the growth of equestrian sports. Gait analysis, which Unfortunately, one of our most esteemed authors, Henk C.
is the study of locomotion, is an area of equine sports science that Schamhardt, PhD, died in an accident on June 26th 1999 during
has made great strides (quite literally) in the last 30 years. his sabattical leave in Australia. As a leading member of the Utrecht
By 1991, gait analysis was sufficiently established as a scientific Equine Biomechanics Research Group, Henk made an enormous
discipline to warrant the establishment of an International Work- contribution to the development of equine biomechanics across the
shop on Animal Locomotion (IWAL). The idea of IWAL was con- world. Those who did not know Henk personally are familiar with
ceived and brought to fruition by Henk Schamhardt and Ton van his work through his excellent conference presentations, numerous
den Bogert at Utrecht University. Subsequent IWALs have been publications, and his role as co-editor of the proceedings of the first
organized by Hilary Clayton in California in 1993 and by Eric two IWALs. As a result of his generosity in sharing his skills and
Barrey in Saumur, France in 1996. By the time this book is pub- knowledge, Henk acted as a mentor to a new generation of research-
lished, Florian Buchner will have organized IWAL4 in Vienna and ers in equine locomotion and biomechanics. It is a fitting tribute
IWAL5 will be in the planning stages. to dedicate this book to him.
Each IWAL proceedings contains a collection of manuscripts that It has been an honour and a pleasure to edit this millennium
reflects the recent and on-going research projects in locomotion book about the locomotion of our mutual friend, the horse. As our
laboratories around the world. Although these proceedings have Swedish colleagues would say while standing with one boot on a
proven to be a valuable resource, the information they contain is chair and the other one on the table ‘To the horse!’
not intended to cover the various aspects of the discipline com-
pletely. The need for a more comprehensive source of information Willem Back (Utrecht, The Netherlands)
on equine locomotion was recognized by Wim Back. W.B. Saunders
supported the concept and Hilary Clayton was enlisted to assist in Hilary Clayton (East Lansing, Michigan, USA)
the editorial duties. Our goal in producing this book is to give the 2000

xiii
This page intentionally left blank
 Acknowledgments

We thank the staff at Elsevier, especially Alison Ashmore, Rita husband Richard. We appreciate their support and tolerance of the
Demetriou-Swanwick, Robert Edwards, Catherine Jackson, Jennifer time we spent working on the book, which was usually during off
Jones, Bhargavi Natarajan, Joyce Rodenhuis, Julie Taylor, and Louisa duty hours.
Talbott for their help in completing this book.
We must also recognize the important role of our families, Wim’s
wife Tia and their children Niels, Floris and Milou, and Hilary’s

xv
This page intentionally left blank
 Contributors

R. McNeill Alexander CBE, MA, PhD, DSc, Hon DSc Antonie J. Van den Bogert PhD
(Aberdeen), Dr.h.c. (Wageningen), FSB, FRS, EAS Parker-Hannifin Endowed Chair in Human Motion and
Emeritus Professor of Zoology Control
Faculty of Biological Sciences Department of Mechanical Engineering
School of Biology Cleveland State University
University of Leeds Cleveland, OH, USA
Leeds, UK
Pieter A.J. Brama DVM, PhD, MBA, Spec. KNMvD (Equine
Fabrice Audigié DVM, PhD, LA-ECVDI Surgery), Dipl. ECVS
Professor in Equine Diagnostic Imaging Professor of Veterinary Surgery
CIRALE-École Nationale Vétérinaire d’Alfort Head Veterinary Clinical Sciences
Maisons-Alfort, France School of Veterinary Medicine
University College Dublin
Adam Arabian PhD, PE
Belfield, Dublin, Ireland
Department of Engineering
Seattle Pacific University H.H. Florian Buchner DVM, PhD
Seattle, WA, USA Associate Professor of Orthopaedics in Large Animals
Equine Surgery Section
Willem Back DVM, Cert. KNMvD (CKRD), Cert. Pract.
Clinic for Horses
KNMvD (Equine Practice), PhD, Spec. KNMvD (Equine
Department for Small Animals and Horses
Surgery), Dipl. ECVS, Prof. (U Ghent)
University of Veterinary Medicine Vienna
Assistant Professor in Equine Surgery
Vienna, Austria
Department of Equine Sciences, Faculty of Veterinary Medicine
Utrecht University Anna Byström DVM
Utrecht, The Netherlands; Department of Anatomy, Physiology and Biochemistry
Visiting Professor in Equine Orthopaedics Swedish University of Agricultural Science
Department of Surgery and Anaesthesiology of Domestic Uppsala, Sweden
Animals
Henry Chateau DVM, PhD
Faculty of Veterinary Medicine
Associate Professor in Anatomy
Ghent University
INRA-ENVA Biomechanics and Locomotor Pathology of the
Merelbeke, Belgium
Horse
Albert Barneveld DVM, PhD, Spec. KNMvD (Equine Ecole Nationale Vétérinaire d’Alfort
Surgery) Maisons-Alfort, France
Professor of Equine Surgery
Hilary M. Clayton BVMS, PhD, Dipl. ACVSMR, MRCVS
Department of Equine Sciences
McPhail Dressage Chair in Equine Sports Medicine
Faculty of Veterinary Medicine
Professor of Large Animal Clinical Sciences,
Utrecht University
College of Veterinary Medicine, Michigan State University
Utrecht, The Netherlands
East Lansing, MI, USA
Eric Barrey DVM, PhD
Jean-Marie Denoix DVM, PhD, Agrégation in Anatomy,
Research Director
Assoc. LA-ECVDI
INRA
Professor in Equine Anatomy and Locomotor System
Animal Genetics and Integrative Biology
Pathology
Jouy-en-Josas,
CIRALE-École Nationale Veterinaire d’Alfort
Inserm, Integrative Biology of Exercise Adaptations,
Maison-Alfort, France
Evry University,
Genopole, Evry, France Nancy R. (Deuel) Toby MSc, PhD
Freelance Editor
Helen L. Birch BSc, PhD, FHEA
St. Michaels, MA, USA
Senior Lecturer in Musculoskeletal Pathobiology
Institute of Orthopaedics and Musculoskeletal Science
University College London,
Stanmore Campus
London, UK

xvii
 Contributors

Machteld C. Van Dierendonck PhD, Prof (U Ghent) Emmeline W. Hill PhD

Equine Behaviour Specialist Lecturer in Equine Science and Chairman Equinome Ltd
Equus Research and Therapy College of Agriculture, Food Science and Veterinary Medicine
Stroe, The Netherlands; University College Dublin
Associate Honorary Scientist Belfield, Dublin, Ireland
Utrecht University
David R. Hodgson BVSc, PhD, FACSM, Dipl. ACVIM
Utrecht, The Netherlands
Professor and Head of Department
Honorary Professor in Equine Behaviour and Animal Welfare
Virginia-Maryland Regional College of Veterinary Medicine
Faculty of Veterinary Medicine
Blacksburg, Virginia, USA
Ghent University
Merelbeke, Belgium At L. Hof PhD
Associate Professor in Biomechanics
Bart J. Ducro PhD
Center for Human Movement Sciences
Assistant Professor in Animal Breeding and Genetics
University Medical Center Groningen
Animal Breeding and Genomics Centre
University of Groningen
Wageningen University
Groningen, The Netherlands
Wageningen, The Netherlands
Mikael Holmström DVM, PhD
Agneta Egenvall DVM, PhD
Holmstroem Thoroughbred Management
Professor of Veterinary Epidemiology
Torreby, Sweden
Swedish University of Agricultural Science
Department of Clinical Sciences Christopher Johnston DVM, PhD
Uppsala, Sweden Mälaren Equine Hospital
Sigtuna, Sweden;
Elwyn C. Firth BVSc MS, PhD, Dipl. ACVS, DSc
Swedish University of Agricultural Science
Professor
Uppsala, Sweden
Department of Exercise Science
Liggins Institute, National Research Centre for Growth and Franca Jonquiere DVM
Development Department of Media and Culture Studies
University of Auckland Faculty of Humanities,
Auckland, New Zealand Department of Farm Animal Health
Faculty of Veterinary Medicine
Mathew P. Gerard BVSc, PhD, Dipl. ACVS
Utrecht University
Department of Molecular Biomedical Sciences
Utrecht, The Netherlands
College of Veterinary Medicine
North Carolina State University J.H. (Han) van der Kolk DVM, PhD, Spec. KNMvD (Equine
Raleigh, North Carolina, USA Internal Medicine), Dipl. ECEIM
Head, section Equine Metabolic and Genetic Diseases
Allen E. Goodship BVSc, PhD, MRCVS
Euregio Laboratory Services
Director and Head of Centre
Maastricht, The Netherlands
Centre for Comparative and Clinical Anatomy
Vesalius Clinical Training Centre Eveline Menke MSc(Psych), PT, APT
University of Bristol Department of Equine Sciences
Bristol, UK; Faculty of Veterinary Medicine
Professor of Orthopaedic Sciences Utrecht University
Institute of Orthopaedics and Musculoskeletal Science and Utrecht, The Netherlands
Royal National Orthopaedic Hospital
Stanmore Campus Liduin S. Meershoek MSc, PhD
UCL Division of Surgery & Interventional Science Faculty of Veterinary Medicine
University College London, Utrecht University
London, UK; Utrecht, The Netherlands
Professor of Orthopaedic Sciences Jonathan S. Merritt BE, BSc
Royal Veterinary College Department of Veterinary Clinic and Hospital
Hatfield, Herts, UK Faculty of Veterinary Science Equine Centre
E. de Graaf-Roelfsema DVM, PhD, Spec. KNMvD (Equine The University of Melbourne
Internal Medicine), Dipl. ECEIM Victoria, Australia
Assistant Professor in Equine Medicine Christian Peham PhD
Department of Equine Sciences Professor
Faculty of Veterinary Medicine Movement Science Group
Utrecht University Clinic for Horses,
Utrecht, The Netherlands University of Veterinary Medicine Vienna
Albert Gramsbergen MD, PhD Vienna, Austria
Professor of Developmental Neurosciences
Department of Medical Physiology,
University of Groningen
Groningen, The Netherlands

xviii
 Contributors

Frederik Pille DVM, PhD, Dipl. ECVS Natalie K. Waran BSc (Hons), PhD
Professor of Orthopaedics in Large Animals Jeanne Marchig Professor of Animal Welfare Education
Department of Surgery and Anaesthesiology of Domestic Director of the Jeanne Marchig International Centre for
Animals Animal Welfare Education Royal (Dick) School of Veterinary
Faculty of Veterinary Medicine Studies
Ghent University The University of Edinburgh
Merelbeke, Belgium Easter Bush, Roslin Midlothian, UK
Lars Roepstorff DVM, PhD P. René van Weeren DVM, PhD, Spec. KNMvD (Equine
Professor of Equine Functional Anatomy Surgery), Dipl. ECVS, Dipl. RNVA
Unit of Equine Studies Professor of Equine Musculoskeletal Biology
Swedish University of Agricultural Science Department of Equine Sciences
Uppsala, Sweden Faculty of Veterinary Medicine
Utrecht University
José Luis L. Rivero DVM, PhD
Utrecht, The Netherlands
Professor of Veterinary Anatomy
Head of the Laboratory of Muscular Biopathology Michael A. Weishaupt PD, Dr.med.vet., PhD
Faculty of Veterinary Medicine, Sports Medicine Section
University of Cordoba, Equine Department
Cordoba, Spain Vetsuisse Faculty
University of Zürich
Peter D. Rossdale MA, PhD, Dr.h.c. (Berne, Edinburgh,
Zürich, Switzerland
Sydney), DESM, FRCVS
Founder Rossdale and Partners Inge D. Wijnberg DVM, PhD, Spec. KNMvD (Equine
Emeritus Editor Equine Veterinary Journal Internal Medicine), Dipl. ECEIM
Director Romney Publications Ltd Assistant Professor in Equine Medicine
Newmarket, UK Department of Equine Sciences
Faculty of Veterinary Medicine
Henk C. Schamhardt PhD (deceased)
Utrecht University
Faculty of Veterinary Medicine
Utrecht, The Netherlands
Utrecht University
Utrecht, The Netherlands Claudia Wolschrijn DVM, PhD
Associate Professor in Veterinary Anatomy
Narelle Stubbs BAppSc (PT), MAnimSt (Animal
Department of Pathobiology
Physiotherapy), PhD
Faculty of Veterinary Medicine
Department of Large Animal Clinical Sciences
Utrecht University
College of Veterinary Medicine
Utrecht, The Netherlands
Michigan State University
East Lansing, MI, USA

xix
This page intentionally left blank
 Glossary

Abduct, abductor, abduction:  Pertaining to Biphasic:  A movement or measurement Diagonal width:  The lateral distance,
the movement of a body part away from with two distinct major amplitudes   measured perpendicular to the direction
the midsagittal plane. per cycle. of motion, between the placements of
Acceleration:  The time rate change of Canter, rotary:  One of two footfall the diagonal fore and hind limbs during
velocity. sequences of the canter gait in which a diagonal stance.
Acceleration due to gravity:  The lateral pair of limbs move synchronously Digitize:  Transform analog data, e.g. a
acceleration of a body freely falling in a and the leading limbs are on opposite photographic image, to digital form,  
vacuum, the magnitude of which varies sides of the body for the fore and hind e.g. an electronic file.
slightly with location. Standardized limb pairs. Displacement:  Length measured along a
value at sea level is 9.807 m/s2. Canter, transverse:  One of two footfall straight line from starting point to
Accelerometer, accelerometry:  Pertaining sequences of the canter gait in   finishing point. A vector quantity.
to the measurement of accelerations. which a diagonal pair of limbs move Distal:  Away from the main mass of the
Adduct, adductors, adduction:  Pertaining synchronously and the leading limbs are body; opposite of proximal.
to the movement of a body part toward on the same side of the body for the Distance:  Length measured along the path
the midsagittal plane. fore and hind limb pairs. taken by the body from starting point to
Advanced completion/lift-off:  Time Caudal, caudad:  Toward the tail. finishing point. A scalar quantity.
elapsing between lift-off of two specified Center of mass:  The point about which the Dorsal:  Toward the dorsum (back) of the
limbs. total mass of a body is evenly balanced. body, opposite of ventral.
Advanced placement:  Time elapsing Used to represent the mass of an object Drag:  Resistance to motion within a fluid.
between ground contact of two specified in physical calculations. Duration:  The period of time during which
limbs. Central pattern generators:  Neural a given state lasts.
Air resistance:  Drag on a body produced networks located in the spinal cord   Duty factor:  The duration of the stance
by the frictional effects of moving that produce rhythmic limb movements. phase of a specified limb as a proportion
through air. Cervical:  The anatomic name for the neck; of the total limb cycle duration or stride
Allometry:  The study of factors that change horses have seven cervical vertebrae. duration.
the shape and functionality of an animal Contralateral:  Located on opposite (left or Dynamics:  The branch of mechanics that
with increasing size. right) sides of the body. deals with motion and the way in which
Angular acceleration:  The time rate change Cost of transport:  The energy expended by forces produce motion.
of angular velocity. an animal in moving a given distance. Efficiency:  The ratio of useful energy
Angular displacement:  The change in Couplet:  When the alternating footfalls of delivered by a dynamic system to the
orientation of a line segment, as the a pair of limbs (diagonal, lateral, fore or energy supplied to it.
plane angle between initial and final hind) are separated by unequal intervals Elastic energy:  The potential energy stored
orientation, regardless of the rotational of time, the two closely-spaced footfalls by a body as a result of deformation.
path taken. occur as a couplet. Electrogoniometry, electrogoniometer,
Angular velocity:  The time rate change of Cranial, craniad:  Toward the head. electrogoniometric:  Pertaining to
angular displacement. A vector quantity. Damping:  A gradual reduction in electrical signals generated by devices
Arthromere:  One of the body segments of the amplitude of an oscillating which monitor the change in angle of  
a jointed animal, somite. movement. a joint.
Arthrometry:  The study of joint movement Density:  The concentration of matter, Electromyography, electromyographer,
range and mobility. expressed as mass per unit volume. electromyographic, EMG:  Pertaining to
Articular, articulate, articulated, Diagonal dissociation:  temporal electrical potentials generated in muscle
articulation:  Pertaining to a structure dissociation at contact or lift-off of a cells during contractions.
constructed of segments united by joints. diagonal pair of limbs that appear to Energetics:  The study of energy, force, and
Axis:  A straight line about which rotation move in synchrony. efficiency.
occurs. Diagonal limb (pair):  A forelimb of one Extend, extension, extensor (joint): 
Biokinematics:  Kinematics applied to side and the hind limb of the opposite Pertaining to factors that cause an
biological systems or entities. side. Customarily, the ‘right diagonal’   increase in joint angle.
Biokinetics:  The study of the forces is right fore and left hind; the ‘left Fatigue:  A reduction in power output,
responsible for the movements of   diagonal’ is left fore and right hind. comfort, and/or efficiency associated
living organisms. Diagonal stance, left/right:  A stride phase with prolonged or excessive exertion.
Biomechanics:  The application of in which the body is supported by a Flex, flexion, flexor (joint):  Pertaining to
mechanical laws to living structures. diagonal pair of limbs. factors that cause a reduction in joint
Biped, bipedal:  Having, utilizing, or Diagonal step length, left/right:  The angle.
supported by two feet. distance along the direction of motion Flexion/extension:  The movement of a
Bipedal:  A portion of the stride in which between the diagonal fore and hind joints around its transverse axis. The
two limbs support the body. hooves during diagonal stance. flexor and extensor sides of the joints are

xxi
 Glossary

defined according to anatomical Lateral limb (pair):  A forelimb and hind Normal:  Perpendicular to a plane, usually
convention. limb on the same side of the body. the ground.
Force:  The mechanical action or effect of Lateral stance, left/right:  A stride phase in Normalize:  Mathematically convert
one body on another, which causes the which the body is supported by a lateral measurements to a common frame of
bodies to accelerate relative to an inertial pair of limbs. reference to facilitate comparisons
reference frame. Lateral step length, left/right:  The distance between individuals or groups.
Force plate:  A device that measures ground along the direction of motion between Ontogeny:  The sequential development of
reaction forces. the placements of the ipsilateral hind the individual organism.
Frequency:  The number of repetitions of a and forelimbs. Outliers:  Values of data that are in excess of
periodic event that occur within a given Laterality:  Asymmetry between the a certain cut-off point.
time interval, usually expressed in Hz left and right sides of the body in Overlap:  Stride phase in which two or
(cycles per second). motion or limb usage that occurs more specified limbs are simultaneously
Gait:  Cyclic pattern of limb movements. naturally and is not accounted for by in stance.
Each complete cycle is one stride. pathology or injury, as in handedness in Pedobarograph:  Device for measuring the
Gallop, rotary:  One of two footfall humans. pressure distribution beneath the foot.
sequences of the gallop gait in which   Levator, elevate, elevation:  Pertaining to Phylogeny:  The complete developmental
the footfall of the leading hind limb is lifting a body part vertically. history through evolution of a group of
followed by the footfall of the ipsilateral Lift-off:  The moment when the hoof leaves animals.
forelimb and the leading forelimb and the ground. Marks the transition from Pitch:  Body rotation about the transverse
leading hind limb are on opposite sides stance to swing phase. (mediolateral) axis.
of the body. Limb, lead/leading:  The second of the two Placement interval:  Elapsed time between
Gallop, transverse:  One of two footfall hind limbs or two forelimbs to contact the contacts of two specified limbs
sequences of the gallop gait in which   the ground in each stride of an within a stride.
the footfall of the leading hind limb is asymmetrical gait.
Potential energy:  Energy of a body
followed by the footfall of the diagonal Limb, trail/trailing:  The first of the two associated with position or
forelimb and the leading forelimb and hind limbs or two forelimbs to contact configuration.
leading hind limb are on the same side the ground in each stride of an
Power:  The rate at which work is done or
of the body. asymmetrical gait.
energy is expended.
Ground reaction force:  force exerted by the Longitudinal axis:  A line in the median
Pressure:  The force applied per unit area.
ground against a limb that is in contact sagittal plane extending from head to
with the ground. Acts in opposition to Pronate, pronator, pronation:  Pertaining
tail; craniocaudal axis.
the force exerted by the limb against   to axial rotation of a limb in which the
Lumbar:  The anatomic name for the loin
the ground. lateral aspect rotates in a cranial
of the horse; there are usually 6 lumbar
direction.
Horsepower:  A quantity of power equal to vertebrae.
735.5 watts (or joules per second) Protract, protraction, protractor: 
Mass:  The quantity of matter contained
(metric unit), or 745.7 watts (imperial Pertaining to moving a body part
by a body, with standardized
unit). forward (cranially).
international measurement units  
Hyperextension:  Excessive or extreme Proximal:  Toward the main mass of the
of kilograms.
extension of a joint; in a range that may body; opposite of distal.
Mass moment of inertia:  The measure of a
cause injury. Quadruped, quadrupedal:  Having,
body’s resistance to accelerated angular
Hyperflexion:  Excessive or extreme flexion utilizing, or supported by four feet.
motion about an axis.
of a joint; in a range that may cause Quadruple or quadrupedal support:  A
Mechanical energy:  The capacity to do
injury. portion of the stride in which four limbs
work, equal to the sum of potential
Ipsilateral:  Located on the same (left or support the body.
energy and kinetic energy.
right) side of the body. Retract, retraction, retractor:  Pertaining to
Mechanics:  The branch of physics
Jump suspension:  The phase during moving a body part backward
concerned with the behavior of  
jumping when the horse has no   (caudally).
physical bodies when subjected to  
contact with the ground. forces or displacements, and the Retrograde motion:  Movement opposite to
Kinematics:  The branch of mechanics that subsequent effects of the bodies   the accustomed direction, i.e. in the
is concerned with the description of on their environment. caudal direction.
movements. Modeling:  A theoretical, simplified Roll:  Body rotation about the longitudinal
Kinetic energy:  Energy of a body associated mathematical construct of a physical (craniocaudal) axis.
with translational and rotational phenomenon. ROM:  Range of motion is the total angular
motion. Moment of a couple:  The resultant change during one stride.
Kinesiology:  The study of the mechanics of moment of two equal but oppositely Sagittal plane:  Plane parallel to the
motion. Usually described with reference directed, non-collinear parallel forces median plane.
to human anatomy. (the couple). Scalar quantity:  A quantity that is
Kinesthesia:  The perception or sensing of Moment of a force:  The turning effect completely expressed in terms of  
the motion, weight, or position of the (torque) of a force about a point. its magnitude.
body. Morphometry, morphometric:  The study Single support:  A portion of the stride in
Kinetics:  The study of internal and external of the form and dimensions of a body. which only one limb is in the stance
forces, energy, power, and efficiency Newton:  The standard unit of force, phase.
involved in the movement of a body. equivalent to that which will cause a Somite:  Body segment.
Lateral:  Pertaining to the sides, left and mass of one kilogram to accelerate   Speed:  The rate of change of distance;
right. 1 m/s2. a scalar quantity.

xxii
 Glossary

Stance phase of a limb:  The period of Support, bipedal or double:  Ground Time:  Measurable quantity in the temporal
ground contact of an individual limb contact and/or weight bearing by two domain, with standardized international
within a stride beginning at ground limbs. units of seconds.
contact and ending at lift-off. Support phase:  phase of the stride when Toe-off:  The instant at the end of the
Stance phase of the stride:  Summation of one or more limbs are in contact with stance phase at which the toe is no
time periods during a stride when one the ground. longer contacting the ground. Marks  
or more limbs are in contact with the Support, quadrupedal or quadruple:  the transition between stance and swing
ground. Ground contact and/or weight bearing phases.
Statics:  The branch of mechanics by four limbs. Torque:  A turning or twisting force; also
concerned with the analysis of loads Support, single:  Ground contact and/or referred to as moment of force.
(force, torque/moment) on physical weight bearing by one limb. Transverse axis:  A mediolateral line
systems in static equilibrium. Support, tripedal or triple:  Ground contact that is mutually perpendicular to the
Step:  The movement pattern in normal and/or weight bearing by three limbs. longitudinal and vertical axes;
locomotion in transitioning from Suspension:  An aerial or airborne phase mediolateral axis.
weight-bearing by one limb to another. of the stride in which all four limbs are Triped, tripedal:  Having, utilizing, or
Step duration:  The time between ground simultaneously in the swing phase and supported by three feet.
contacts of two specified limbs. free from weight bearing. Tripedal support:  A portion of the
Step height:  The height to which a Suspension, extended:  Aerial phase of the stride in which three limbs support  
specified location on a limb is raised stride that occurs with between lift-off of the body.
from the ground during locomotion. the leading hind limb and contact of the Ungulate:  Hoofed mammal.
Step length:  The horizontal distance in trailing forelimb when the vertebral Uniped, unipedal:  Having, utilizing, or
the direction of motion between the column is extended. supported by one foot.
footfalls of two specified limbs. Suspension, gathered:  Aerial phase of the Unipedal support:  A portion of the
Strain:  Deformation resulting from the stride that occurs between lift-off of the stride in which one limb supports the
application of stress. leading forelimb and contact of the body.
Strain rate:  The rate of strain generation. trailing hind limb when the vertebral Units, SI:  Standard international units (or
Strength:  The maximal force produced by column is flexed. their standard multiples) accepted by the
muscular contraction. Swing phase:  The portion of the stride International Society of Biomechanics
Stress:  An external force that acts on a cycle or limb motion cycle in which   for scientific measurement, including
body. a designated limb is free from contact length in meters, mass in kilograms,
Stride:  A complete cycle of the repetitive with the ground. time in seconds and plane angle in
series of limb movements that Symmetrical:  A movement or morphology radians.
characterize a particular gait. that is substantially similar in being a Vector quantity:  A quantity that has both
Stride duration:  The time required to mirror image on left and right sides of magnitude and direction.
complete one stride. the body. Velocity:  The time rate change of
Stride frequency:  The number of Symmetrical gait:  A gait in which the displacement. A vector quantity.
repetitions of the stride per unit   limb coordination pattern of one side Ventral:  Toward the ventrum (belly,
time. repeats that of the other side, half a underside) of the body; opposite  
stride later. of dorsal.
Stride length:  The horizontal distance
traveled in the direction of motion Temporal:  Concerning time, duration. Weight:  The force of gravity acting
during a single stride, or between Tension:  The application of forces acting to on a body, equal to the product  
consecutive hoof prints of the same   stretch an object. of mass and the acceleration due  
hoof. Tetrapod, tetrapodal:  Having, utilizing, or to gravity.
Supinate, supinator, supination:  Pertaining supported by four feet. Work:  The result of a force acting to
to axial rotation of a limb in which the Thoracic:  The anatomic name for the chest displace a body in a given direction.
lateral aspect rotates in a caudal of the horse; there are usually 18 Yaw:  Body rotation about the vertical
direction. thoracic vertebrae. (dorsoventral) axis.

xxiii
This page intentionally left blank
 C H A PTER 1 

History
P. René van Weeren

Introduction warfare. In fact, it looked as if the species would become entirely

marginalized.
However, interest in the species was revived at the end of the
Man has always been fascinated by the creatures that surrounded sixties and in the early seventies of the 20th century when eques-
him. Rock paintings and engravings from prehistoric times trian sports enjoyed an immense popularity that continues to
show that, apart from man himself, it was the large mammalian increase today. This popularity has again made the horse into an
species that were most often depicted. Among these, the horse important economic factor, worthy of serious investment. At the
played a prominent role in those latitudes where this species was same time, interest in locomotion analysis has revived. This time
abundant. first in Sweden, but soon followed by other countries and regions
After domestication interest in the species naturally deepened as where the horse has gained importance as a sports and leisure
the role changed from simple animals of prey to that of an impor- animal like in North America and North-western Europe. This
tant economic entity. The horse, unlike almost all other species, was renewed interest in equine locomotion coincided with the elec-
domesticated for its locomotor capacities rather than as a supplier tronic revolution, which made computer-aided analysis a reality,
of food or clothing materials. It was this tremendous capacity to thus creating the possibility for much more advanced and profound
move that for millennia gave the horse its pivotal role in transport studies of equine locomotion than ever before.
in many of the major civilizations on this planet and that also made The science of equine locomotion is thriving. This book aims to
it a most feared weapon in warfare from ancient times until very present the state of the art in this branch of science. In the first
recently. chapter an attempt is made to give an overview of how this science
As a result of this important role in transportation combined with developed over time against the background of evolving veterinary
its proximity to man, the horse became the primary focus when science, but more so against the background of the evolving rela-
veterinary science developed in the ancient societies and later as it tionship of mankind with what has been called our closest ally; the
became a flourishing branch of science during the heyday of the horse.
Greek and Roman civilizations. It is therefore not surprising that it
was during Antiquity that the first scientific comments were made
on gait.
The decline of the Antique culture and the subsequent fall of the Prehistoric times
Roman Empire brought science to a virtual standstill in most of
Europe during the dark Middle Ages and it is not until the Renais- The oldest known art to be produced by man is the rock art found
sance that we see a renewed scientific interest that also extends to in various caves in the Franco-Cantabrian region, covering what is
veterinary medicine. First directed at the legacy of Antiquity, science nowadays South-western France and North-western Spain. Here,
took a step forward in the 18th century when the modern approach about 30 000 years ago the Cro-Magnon race of people began
of making observations and drawing conclusions, later followed depicting their environment by means of large and impressive
by the conjunction of hypotheses with subsequent experimental paintings on the walls of rock caves. At first still somewhat crude,
testing, was adopted. Then we also see the founding of the first artistic heights were reached about 15 000 years ago in the Magda-
veterinary colleges. These focused almost exclusively on the horse, lenian period, so called after the rock shelter of La Madeleine, near
which, throughout this entire period, had maintained its primary present-day Montauban. In those days of the last Ice Age, South-
role in transport and warfare. It was France that took the lead and western Europe must have known abundant wildlife. In the paint-
it was also French scientists that published the first scientific study ings two classes of animals prevail: ruminants such as cattle, bison,
completely dedicated to the locomotion of the horse. deer and ibex, and horses. The way horses were represented does
France retained the lead in veterinary medicine, and in equine not reveal a profound knowledge of equine locomotion. In most
gait analysis, for almost a century until the end of the 19th century. cases the animals were painted standing with all four legs on the
Then, with one notable exception in the United States, German ground, or in an unnatural jump-like action with the forelimbs
scientists took over and explored with their characteristic thorough- extended forward and the hind limbs backward, in much the same
ness the possibilities of novel techniques like cine film. way as horses were still erroneously depicted in many 18th century
The outbreak of World War II brought this thriving research to a and early 19th century paintings (Fig. 1.1). Species like rhinoceros,
halt. There was no recovery after the end of the war because the mammoth, bear and the felidae are present, but to a much lesser
mechanical revolution, which had already started during World War extent. Perhaps the plains, which covered that part of Europe in this
I, made horse power redundant and brought to a definitive end the period, looked much like the great plains of East Africa, such as
traditional role the horse had played for millennia in transport and the Serengeti, nowadays. Here too, ruminants like buffalo and

1
 1 History

Fig 1.1  Przewalski-type horse as depicted in the

cave at Lascaux (about 15 000 BC).
Reproduced from Dunlop, R.H., Williams, D.J. (Eds.), 1996. Vet-
erinary Medicine. An illustrated history. Mosby, St. Louis, with
permission from Elsevier.

wildebeest are abundant together with equids (zebras), while other These changes in human society were to a large extent possible
species such as rhinoceros and the large cats occur in significantly thanks to a new phenomenon: the domestication of animal
smaller numbers. species. It is widely believed that the dog was the first animal to be
Man was still a hunter-gatherer in those days and, for this reason, domesticated about 12 000 years ago. Like most early domestica-
the wild animals comprised an essential part of his diet. Remains tions, this event took place in Western Asia’s Fertile Crescent (the
of large mammals eaten by man, including horses, have been found area of fertile land from the Mediterranean coast around the
at many sites. It is interesting to note that the vast majority of rock Syrian Desert to Iraq), which was the cradle of human civilization.
paintings concerned animals, most of them large mammals, whereas There also the next domestication took place: small ruminants
man himself was depicted rarely and other parts of the environment were domesticated approximately 10 000 years ago, sheep and
such as the vegetation or topographical peculiarities were never goats in about the same period. Cattle were domesticated 2000
shown. Also non-mammalian species such as birds, reptiles, fish or years later in Anatolia (Western Turkey). Cats were domesticated
insects were virtually unrepresented. The rock art found in various (or adopted man as some people state) as early as 9000 years ago.
parts of Zimbabwe and other parts of Southern Africa was some- The first camelids to be domesticated were llamas in South
what different. These paintings were made by the Bushmen from America, perhaps as early as 7500 years ago. The horse arrived
13 000–2000 years ago. Here again, the large mammalian species rather late on the scene. There is evidence that the first horses were
prevailed, with the zebra representing the equids, but man was domesticated in what is now Southern Russia approximately 5000
depicted more often and there were some paintings of fish and years ago. However, the domestication of the horse dramatically
reptiles (Adams & Handiseni, 1991). The Bushmen culture has influenced the history of mankind, mainly because of its enor-
survived until the present day, though in a much diminished and mous potency in warfare.
nowadays heavily endangered form, and it is known that these The horse was definitely the most revolutionary innovation in
people, who were hunter-gatherers, lived in a very close relationship warfare before the invention of gunpowder. First, the animals were
with their environment, forming an integral part of the entire eco- used to draw heavy war chariots which, used against traditional
system. It is easy to imagine that under such circumstances the large infantry, could provoke enormous massacres while being them-
mammals, which were the most impressive fellow-creatures giving selves rather invulnerable. Later, with the development of the skill
rise to mixed feelings of awe, admiration and a certain form of soli- of horse riding and increasing horsemanship, a real cavalry of
darity, inspired the creation of works of art. mounted soldiers was developed which, with their greater agility,
The world changed dramatically when, at the beginning of the replaced the chariots. This development enabled rapid conquests of
Neolithic period about 10 000–12 000 years ago, man changed from vast territories. The Hittites conquered Asia Minor (present-day
being a hunter-gatherer to primitive forms of agriculture and pas- Turkey) in 2000 BC with their horse drawn chariots. A thousand
toralism. The capacity of most natural savanna habitats to support years later the Scythians, originally a Eurasian nomadic tribe settled
fixed human nutritional requirements is estimated at only one or in the area north of the Black Sea. The Scythians were excellent
two persons per square mile (Dunlop & Williams, 1996). The horse riders and they became masters in the tactics of cavalry-based
advent of agriculture and pastoralism meant that the nutritional steppe warfare, enslaving agricultural peoples and plundering what
constraints on population growth were lifted and an unprecedented came in their way. Later, other tribes that were mostly of Eastern
population growth followed. It also meant that man definitively origin, succeeded. The Huns overran the Roman Empire from the
and irreversibly placed himself apart from his fellow-creatures 4th to the 6th century AD, and in the 13th century AD, Genghis Khan
and outside the existing ecosystems where the numbers of species reached the gates of Western Europe. For those peoples the horse
were determined by the unmanipulated carrying capacity of the was more than just a domesticated animal; it was central in their
environment. culture, as a weapon, food, drink, a friend and a god. The warriors

2
 The Antique world

were capable of staying in the saddle for an entire day. They ate alluded to earlier (Simpson, 1951), the oldest reports of domesti-
horsemeat, drank mare’s milk and intoxicated themselves during cated horses are from the Shung dynasty (1766–1027 BC). Like
their feasts with the fermented form of it. It is even said that sol- everywhere else, the horse was used first to draw chariots, then for
diers, traveling without rations, opened the veins of their horses, a mounted cavalry. From the latter period dates the famous ‘army’
drank the blood, closed the wounds, and remounted (Simpson, of terra cotta figures (including large numbers of horses) that was
1951). excavated at the burial site of Shih Hunagdi (259–210 BC). He is
The changes in attitude towards animals by man, including also called the first emperor as the formerly divided China was
domestication and changes in the use of domestic animals that were united by then. The horse gained great importance in China during
partly dictated by changing environmental conditions are magnifi- the Han dynasty when emperor Wu sent out a military expedition
cently demonstrated by the North African rock art found, for to capture 3000 horses of a heavier and sturdier breed, which he
instance, in the Hoggar and Air mountain ranges in what is now called the ‘Horses of Heaven’. They were probably related to the
the central Sahara Desert. Several thousands of years ago, North Tarpan breed that still roamed the steppes of Southern Russia.
Africa was not covered to such a large extent by the extremely arid Only 50 of them survived the 2000-mile journey home. By the
and inhospitable Sahara Desert as it is now. The oldest art dates to middle of the 7th century AD, during the Tang dynasty, horse
about 7000 years ago and depicts wild animals such as buffalo, breeding in China reached unprecedented heights when numbers
giraffe, elephant, ostriches, etc., suggesting that the area must have increased from 5000 to over 700 000. The Chinese were excellent
looked like large parts of Eastern Africa do now. About 4000 BC in designing saddlery and harnesses. They invented the trace
cattle and fat-tailed sheep appeared as first representations of harness, in which the power of the horse is transmitted by a belt
domestic species. Horses appeared around 1200 BC, first drawing around the chest, long before it was used in Europe where a collar-
chariots. These chariots are believed to have belonged to Cretan type of harness was common. This latter type of harness com-
invaders because they are similar to pictures of chariots from this presses the trachea and jugular veins when force is applied and
island (Lhote, 1988). In those days there was a Trans-saharan route therefore permits the exertion of a force only one sixth that of
running from Tripoli and probably also Egypt to Gao on the Niger using a trace harness. Also stirrups are a Chinese invention, dating
River, thus connecting the Mediterranean, Egyptian and Nubian from the 3rd century AD.
cultures to the Bantu cultures of the Niger River valley. More recent A few reports on equine veterinary medicine have survived from
rock art shows riders instead of chariots. With the increasing aridity these cultures, some of which are quite extensive and methodical,
of the Sahara, the horse became unsuitable for traveling large dis- like some Egyptian works. However, no specific studies on equine
tances with ever diminishing water resources, and was supple- locomotion are known. Horses in general were depicted as they had
mented by camels around 100 BC. been in prehistoric times and would remain until quite recently:
either in a rather natural pose at a slow gait or in the characteristic
unnatural pose that was used to indicate the gallop: forelimbs
extended forward and hind limbs backward.
The ancient cultures

The first human civilizations, characterized by urbanization and the

invention of a script, developed in Mesopotamia, the area around The Antique world
the Euphrates and Tigris Rivers, about 3000 years BC. The first report
on hippiatry dates from the 14th century BC from the Assyrian There is perhaps no revolution that has changed the course of
culture. In those days the city of Hethiter was famous for the pro- human history more than the revolutionary change in thinking
curement and training of horses. Also, donkeys were already cross- that originated in the Greek port towns of Asia Minor (Western
bred with horses to produce the sturdier, but infertile mules. There Turkey) and nearby islands around 600 BC. Leaving ordained pre-
is even evidence that horses were crossbred with the then abundant conceptions and supernatural speculations or dogmas behind as
wild onagers. explanations for natural phenomena, nature was now studied in
After the decline of the Assyrian empire in the 8th century BC the a rigorous, rational way. The proponents of this new way of
Medean and Persian cultures took over. These were, to a large extent, thinking were called natural scientists or philosopher scientists
horse-based societies where horsemanship was developed to great (Dunlop & Williams, 1996). Here, the foundations were laid
heights. In fact, the word ‘Persia’ is derived from the word for for the great Greek schools of philosophy, like the Athenian
horseman. It was the forceful Persian cavalry that created the Academy. This institution became the intellectual center of the
largest empire the world had ever seen until then under Darius I. world a couple of centuries later, producing the great philoso-
An empire that was only to be conquered by another horse-based phers Socrates, Plato and Aristotle. It can be stated without any
army, that of Alexander the Great in 322 BC. hesitation that with the onset of the great Greek philosophical era
One of the other great ancient cultures, that of Egypt, lived its science was born. Nature in all its aspects was studied for the first
first periods of glory (the Old Kingdom from 2620–2170 BC and time by theoretical speculation (proposing hypotheses), followed
the Middle Kingdom from 2080–1760 BC) before the arrival of by critical reappraisal and revision. What was still missing was the
(horse-borne) invaders. During this period the horse was an experimental testing of hypotheses, which is essential to modern
unknown animal to the Egyptians who were surrounded by many science. The Greek philosophical schools tried to resolve all prob-
domesticated species like dogs, cattle and cats, the last of which lems by logical reasoning, so the balance was far to the intellec-
gradually obtained a divine status during the later periods of Egyp- tual side. Nowadays it is not uncommon to see a reversed
tian culture. The horse was introduced in the New Kingdom when tendency with strong emphasis on strictly controlled experimental
invaders from the Palestinian region, who used chariots drawn by testing, but with sometimes hardly any evidence of critical
swift Arabian horses, began to challenge Egyptian sovereignty. The thinking.
only way to refute them was by using the same weapon: the horse. Aristotle was a teacher of Alexander the Great who, seated on his
In the New Kingdom (1539–1078 BC) the Egyptians became masters black stallion Bucephalus, conquered most of the then known
in horsemanship and horse breeding, producing the finest Arabian world. Alexander was a skillful rider who is said to have been the
horses. The horse even allowed them to expand their empire as far only one able to ride the horse. Bucephalus had been given to him
as the Euphrates River. by his father when he was 12 years old and served Alexander for 17
In China, where some think that a separate domestication of the years (Fig. 1.2). Alexander greatly favored the development of
horse had taken place, independent of the site in Southern Russia science and created a center of learning in the city that was named

3
 1 History

Fig 1.2  Alexander the Great attacking Persian

horsemen on Bucephalus. From the sarcophage of
Alexander, Syria.
Reproduced from Dunlop, R.H., Williams, D.J. (Eds.), 1996. Vet-
erinary Medicine. An illustrated history. Mosby, St. Louis, with
permission from Elsevier.

after him: Alexandria. This city on the mouth of the Nile would Greece, Asia Minor and, of course, Alexandria. However, the Romans
remain the intellectual center of the world from 300 BC to 500 AD. were excellent in organizing and implementing the scientific and
In the vast library 700 000 scrolls were housed compiling all knowl- technical advances of others. Consequently, they created one of the
edge that had been gained in the preceding millennium. The burning vastest empires the world has ever known and which still influences
of the library on the orders of Caliph Omar in AD 642 was an act of many aspects of daily life.
barbarism, narrow-mindedness and, in its deepest meaning, of fear Horses played a pivotal role in the Roman army, which employed
for the unknown. It resembles the burning of books that took place large numbers of veterinarians to care for them. These were first
in more recent history and is still taking place on the instigation of called ‘mulomedici’, but after an overhaul of the military regula-
totalitarian regimes and intolerant sectarian cults. tions under Commodus (180–192 AD) the term ‘veterinarii’ appears.
The first extensive work on equine conformation was performed Thanks to the enormous popularity of horse racing (chariots drawn
by Xenophon (430–354 BC). Apparently a man of great experience, by 2 or 4 horses), there was also employment for a category of
he described in full detail the desirable and undesirable traits of veterinary specialists not unknown today: the racetrack veterinarian
horses. Many of his criteria were equal to those used today. Though of which Pelagonius in the 4th century was a famous example.
his work was more of a hippiatric caliber than a scientific work, he However, these veterinarians were mainly engaged in the treatment
already recognized the role of the hindquarters as the motor of of diseases and healing of the many wounds. Empiricists, they relied
locomotion. heavily on their Greek and Hellenistic counterparts for some theo-
It is not surprising that the first documented study on animal retical basis. A noteworthy exception was the physician Galenus.
locomotion originated from one of the great Greek philosophers, Galenus was born in Pergamon, Asia Minor, in 130 AD, but he
Aristotle (384–322 BC). In his youth, Aristotle was intrigued by worked for decades in Rome. He conducted large numbers of experi-
natural history, and he wrote various volumes on biological and ments on animals to advance medical knowledge and can be seen
medical matters. In his works De motu animalium and De incessu as the founder of the experimental basis of comparative medicine.
animalium (On the movements of animals and On the progression of He produced vast numbers of treatises of which about 20% have
animals) he accurately described quadrupedal locomotion, at least survived the ages. None of those is dedicated to the study of animal
in the slower gaits. In De incessu animalium (Aristotle, 1961) he or human locomotion.
states that: Emperor Diocletianus (284–306 AD) had divided the Roman
Empire into an eastern and a western half. Constantine the Great
reunited the empire in 324, but it was divided again in 395. The
western part fell with the abdication of Romulus Augustus in 476;
the eastern part was to survive for an additional 1000 years as the
Byzantine Empire with Constantinople (Istanbul) as capital. The
veterinary profession was at a high level as can be judged from
(The bendings, then, of the limbs take place in this manner the compilation of all that was known in this field under the
and for the reasons stated. But the hind limbs move diagonally name Corpus Hippiatricorum Graecorum or Hippiatrika. Though
in relation to the forelimbs; for after the right forelimb published in the 9th or 10th century, most of the contents date
animals move the left hind limb, then the left forelimb, and back to the 4th century. The contributions of Apsyrtos (300–360),
after it the right hind limb.) the chief military veterinarian in the army of Constantine the
Great, are of outstanding quality. Though the care and treatment
The Romans were more doers than thinkers and lacked the intel- of the locomotor system have a prominent position in this
lectual drive that characterized the Greeks. Throughout the whole work, no specific comments on locomotion itself or gait analysis
period of the Roman Empire, the intellectual center remained in are made.

4
 From the Renaissance to the 18th century

Through the Dark Ages to the Renaissance mystical powers towards a more rational, naturalistic approach.
Perhaps no one was as closely associated with this revolutionary
process as the genial artist and scientist Leonardo da Vinci (1452–
After the fall of the Western Roman Empire, the existing administra- 1519). Leonardo himself is known to have been interested in the
tive structures collapsed and much of the knowledge that had been movements of animals and he even projected to:
gained over the centuries was lost. For centuries most of Europe
became an incoherent assembly of tribes and mini-states where … write a separate treatise describing the movement of
insecurity and ignorance reigned. During this period the impressive animals with four feet, among which is man, who likewise in
Arab conquest started from Mecca in the Arabian Peninsula where its infancy crawls on all fours.
the prophet Muhammad had died in 632. Within a century, the Clayton, 1996
Arabs conquered millions of square miles of land from Northern
India to Spain. This could be accomplished thanks to their aggres-
Da Vinci was intrigued by the flexibility of the equine spine and
sive light cavalry, which was based on the swift and enduring Arab
produced a series of fine drawings, now in the British Royal Collec-
horses, and their great horsemanship. They were halted by the
tion at Windsor Castle, with horses in a number of exceptional, but
troops of the Frankish king Charles Martel at Poitiers in 752. The
not impossible poses.
Franks were only able to withstand them because they employed a
It has been stated that the renaissance in veterinary medicine
heavily armored cavalry, which was rather invulnerable to the light
started with the publication of the first great textbook on veterinary
cavalry of the Arabs, not unlike the use of the first tanks in World
anatomy Dell Anatomia et dell’Infirmita del Cavallo (On the Anatomy
War I.
and Diseases of the Horse) by Carlo Ruini in 1598. The anatomical
While Europe was in cultural decline, the Arab culture flourished.
part presented the first real new work since Antiquity. However,
It is thanks to many Arab scientists that at least a part of what had
the part on diseases did not pass the standards of Jordanus Ruffus,
been written in Antiquity has survived to the present day. They
De Medicina Equorum that had been published 350 years earlier.
translated the works into Arabic and in the later Middle Ages these
It may not be surprising that this ambiance of emerging science
Arabic versions were translated again into Latin to lay the founda-
fostered the first contribution to the science of equine locomotion
tion for the scientific revival in the Renaissance. The Arabs also
since Aristotle. Giovanni Alphonso Borelli (1608–1679) was a pro-
contributed to veterinary medicine with original works. Akhi Hizam
fessor of mathematics at Pisa University and applied physical theory
al-Furusiyah wa al-Khayl wrote the first book on the characteristics,
to the study of animal locomotion. He calculated the force of
behavior and diseases of horses in 860. Abu Bakr ibn el-bedr al
muscle action and recognized that the muscles were under nervous
Baytar (1309–1340) wrote an excellent work on veterinary medi-
control (Fig. 1.3). In his book De motu Animalium (On the movement
cine, the Kamil as Sina’atayn. This book features aspects of equine
of animals), he describes the center of gravity and also makes obser-
management and care including the tricks of horse-dealers(!),
vations about limb placement in the various gaits (Borelli, 1681).
together with remarks on appearance, conformation and gait
(Dunlop & Williams, 1996). The horse had a very high standing in
the Arab world. Abu Bakr held the opinion that the horse was so
important to an Arab man that it would be reunited with him in
paradise, together with his wives. There is also an Arab maxim
stating that:
Every grain of barley given to a horse is entered by God in the
register of Good Works.
Simpson, 1951

In the first part of the Middle Ages the medical and veterinary
professions stood at a low level in most of Europe. The link with
Antiquity had been broken and the Christian church, which saw
diseases as a divine punishment to be cured with the help of
supernatural power, had a hostile attitude towards the few rational
natural scientists. In medieval matters mystics and superstition
played an important role. It was not until the late Middle Ages that,
mainly through the translation of Arab texts (originals and transla-
tions of classical works), the tide changed. The emperor Frederick
II was a man ahead of his time. He formed a bridge between the
Christian Western world and the Islamic East. He was a great pro-
ponent of science and had a special interest in animals. It was his
chief marshal, Jordanus Ruffus, who, supported by the emperor,
published the first new work on equine medicine De Medicina
Equorum in 1250. Fredrick II was a great, but ruthless, innovator. At
first supported by Pope Innocent III, he fell into disgrace with his
successors who deprived him of his kingdoms in 1245. This enlight-
ened man can be seen as a very early protagonist of the wave of
renewal that was to blow over Europe and which would mean an
end to the Middle Ages: the Renaissance.

From the Renaissance to the 18th century Fig 1.3  Page from De motu animalium by Giovanni Borelli comparing equine
and human locomotion.
In Italy a change in scientific attitude developed; this involved a Reproduced from Dunlop, R.H., Williams, D.J. (Eds.), 1996. Veterinary Medicine. An illustrated
change from the concept of life as the product of supernatural and history. Mosby, St. Louis, with permission from Elsevier.

5
 1 History

He was obviously ahead of his time; this line of investigation would interesting for everybody dedicated to the art of horse-riding. The
not be further pursued until the end of the 18th century. work is of extreme importance in the history of equine gait analysis.
The 17th century was the age of the great horse marshals. One of Though not entirely correct with respect to limb placement in the
these was William Cavendysh, the first Duke of Newcastle (1592– faster gaits, the study is well done and enters into great detail. Gaits
1676). He was one of the most famous horse trainers of his days of horses are represented by a ‘piste’ (a graphical representation of
but was, as a royalist, forced to leave Britain when Charles I’s army the footfall pattern), a kind of schematic stick diagram, an elaborate
was defeated by Cromwell’s troops. In exile in Antwerp he wrote an table, and by what we now call a gait diagram (Fig. 1.4). This latter
extensive work on all aspects of the horse, which first appeared in representation of equine gait was invented by Goiffon and Vincent
French. It includes a chapter on the gaits, which he studied with and has proved so useful that it is used in many present-day
help of the sounds that the hooves make when they strike the publications in an essentially unaltered form. Goiffon and Vincent
ground (Cavendysh, 1674). Another great marshal of this era was called it an ‘échelle odochronométrique’. Regarding the origin of the
Jacques de Solleysel from France. His great work Le Parfait Maréschal word, they state:
qui enseigne a connoistre la beauté, la bonté, et les défauts des chevaux
Cette denomination est composée de trois mots grecs, dont l’un
(The perfect marshal who teaches how to know the beauty, the virtue and
signifie chemin, l’autre temps & le troisieme mesure. C’est la
the defects of horses) consists of two volumes. The first one is dedi-
définition exacte de notre échelle; elle est la mesure du temps
cated to horse management, the second to equine diseases. De
& du chemin fait pendant ce temps.
Solleysel makes some remarks on limb placement in various gaits,
(This name is composed of three Greek words: one means
but appears not to really have made a study of the subject (De
distance, one means time and the third means measurement.
Solleysel, 1733).
This is the exact definition of our scale, which in fact measures
time and the distance covered in that time.)
The start of veterinary education Goiffon & Vincent, 1779

Notwithstanding some progress in the preceding centuries, the vet-

erinary profession was still far from any scientific status in the
middle of the 18th century. George Leclerc, Comte de Buffon
The 19th century
(1707–1778), who was the dominant personality in zoology during
the second half of the 18th century, wrote at the end of his zoologi- By the end of the first half of the 19th century veterinary schools
cal description of the horse: had been founded in practically all countries that belonged to the
then developed world, which did not yet include the United States.
I cannot end the story of the horse without writing with regret There, the first (private) veterinary school was the Veterinary College
that the health of this useful and precious animal has been up of Philadelphia, founded in 1852. However, progress was relatively
to now surrendered to the care and practice, often blind, of slow and some of these institutions, an example being the Royal
people without knowledge and without qualification. Veterinary College in London under Coleman, even had a question-
Dunlop & Williams, 1996 able academic level. The profession still had a low status and the
interest in the courses sometimes was marginal. In Holland, where
However, the chances for the veterinary profession were to change veterinary education had started in 1821 with 24 first-year students,
for the better, mainly because of two reasons. On the one hand the only 8 first-year students entered during the 7(!) years from 1848
need for better veterinary care became dramatically evident in this until 1855 (Kroon et al., 1921). In the second half of the century,
period by the huge losses of horses on the many battlefields of those things would change dramatically thanks to decisive breakthroughs
days, and by the incredible losses of livestock caused by various in microbiology, especially in bacteriology. Scientists such as Pasteur
waves of cattle plagues, mostly rinderpest, that swept over Europe. and Koch provided the clues for many diseases that so far had been
It has been estimated that more than 200 000 000 cattle died in of mysterious origin. The brilliant pathologist Virchow laid the basis
Europe in the years between 1711 and 1780 because of this disease for a cell-based pathology that broke with the old humoral theories.
which had a profound demoralizing effect on the rural areas. On These discoveries had enormous implications, not only for human
the other hand, a new intellectual movement, which came to be medicine, but also for the veterinary sciences. It was the time that
known as the Enlightenment, was spreading through Europe from the great cattle plagues rinderpest and contagious pleuropneumo-
its origins in France. Philosophers such as Montesquieu, Rousseau nia came under control. In the horse the causative agent of glanders,
and Voltaire rejected the idea of mere authority as the source of then the greatest plague of this species and, as a zoonosis, a poten-
truth and emphasized the role of reason. These two concurrent tial threat to man, was isolated by Schütz and Löffler in 1886. All
circumstances created an optimal starting point for veterinary edu- these developments boosted the interest in veterinary medicine and
cation. It was Claude Bourgelat, director of the Academy of Equita- stressed the importance of the profession.
tion in Lyon and, himself, one of the authors of the great work, The Throughout the 19th century the horse retained its primary posi-
Enlightenment, a highly controversial and unorthodox encyclopedia tion in society. On the battlefields the cavalry remained as decisive
that trampled the toes of many established authorities, who finally as ever though heavier losses of horses were inflicted because of the
obtained royal permission to transform the riding school into the increasing artillery firepower. Napoleon lost more than 30 000
first veterinary school in 1761. A few years later he would establish horses (and over 300 000 men) during his Russian expedition. In
a second one in the center of the French empire; Paris. Other Euro- the Boer war of 1899–1902, 326 073 horses perished. Over the
pean countries followed the French example with the Austro- entire century, millions more must have lost their lives in battle.
Hungarian Empire as the first (1766). The horse remained equally important in the transport sector
The newly established veterinary center at Alfort near Paris did though in the second half of the century they were increasingly
not fail to produce results. In 1779 the first modern work that replaced by the rapidly expanding rail network for long-distance
focuses entirely on equine gait was published. The authors were the transport. In Britain, travel by mail coach peaked in the 1830s and
late Mr Goiffon and his deputy Vincent, who was employed by the then declined because of the increasing rail services. The extension
school and later became one of the first pupils of the Alfort School. of the railroad network in Great Britain tripled between the
The book of Goiffon and Vincent was primarily intended to help years 1850–1875 from 5000–14 500 miles. However, horse-drawn
artists depict their horses in a more natural way, which had been transport remained important at a local level and in the rural areas
a problem throughout the ages, but it was considered equally until well into the 20th century.

6
 The 19th century

In the field of locomotion analysis some progress was made in cycle of a limb into six periods. This was a simplification of the
the first three quarters of the 19th century. In Switzerland, Conrad system used by Goiffon and Vincent who had used 12 time inter-
von Hochstätter published, from 1821–1824, his Theoretisch- vals. Raabe’s division was also used by Lenoble du Teil (1873). In
praktisches Handbuch der äussern Pferdekenntniß, und der Wartung und his 1893 publication, when he had a leading position at the famous
Pflege der Pferde (Theoretical and practical handbook of the conformation national stud Haras du Pin, Lenoble du Teil used his studies and
of the horse and of horse grooming and care), which includes the first similar works of others to take a strong stand against the classical
considerations of the mechanisms underlying equine gait, based on Italian school of riding.
his own observations. He also discusses the consequences of a In the Anglo-Saxon world there was not much research on the
number of faulty conformations for performance. Unfortunately, topic. However, the problem of gait analysis was a point of discus-
this work remained largely unnoticed by the veterinary profession sion from time to time as exemplified by a scientific quarrel between
(Schauder, 1923a). In Germany, there was an increasing interest in Joseph Gamgee (Edinburgh) and Neville Goodman (Cambridge)
the explanation of locomotion by specific muscle action. This cul- in the Journal of Anatomy and Physiology in the early 1870s. Discuss-
minated in the classical work Die topographische Myologie des Pferdes ing the canter, Gamgee stated that:
mit besonderer Berücksichtigung der lokomotorischen Wirkung der
The horse in the fast paces, as in the slowest movement, has
Muskeln (Topographical myology of the horse with special attention to the
never less than two of his feet acting on the ground.
locomotor effect of muscles) by Karl Günther in 1866. In Austria, Bayer
(1882) did some experimental work on the hoof mechanism using
an electrical device (Fig. 1.5) while in Germany Peters (1879) also This statement was (correctly) attacked by Goodman. However,
dedicated himself to the hoof. there was, as yet, no means to prove this (Gamgee, 1869, 1870;
In the meantime, in France, attention remained focused on gait Goodman, 1870, 1871). In 1873, Pettigrew published a book on
analysis as initiated by Goiffon and Vincent. In his book on the animal locomotion in which he put forward some ideas that were
conformation of the horse, Lecoq (1843) introduced a different gait later taken by Marey from France, who was to become much more
diagram from that of Goiffon and Vincent (Fig. 1.6). It did not find famous.
general acceptance because, although it was unequivocal regarding
limb placement, it did not give temporal information.
In his book Locomotion du cheval (Locomotion of the horse, 1883), Muybridge and Marey: revolution
Captain Raabe presented an ingenious system consisting of two
discs, a fixed one and a rotating one, with which the sequence of
in gait analysis
limb placement in all symmetrical gaits can be determined (Fig. About a century after the French Revolution, which meant the end
1.7). Raabe, who first published his work in 1857, divided the stride of an era and changed world politics forever, a revolution took place

   

 

 
  
 
  
  

Fig 1.4  Forms in which equine gait was represented by Goiffon and Vincent in the first work that was entirely dedicated to the subject: (A) stick diagram;
(B) table; (C) footfall pattern; (D) ‘échelle odochronométrique’ or the precursor of the present day gait diagram.
Reproduced from Goiffon & Vincent, 1779. Mémoire artificielle des principes relatifs à la fidelle répresentation des animaux tant en peinture, qu’en sculpture. Ecole Royale Vétérinaire, Alfort.

7
 1 History

Fig 1.4  Continued

8
 The 19th century

in the field of equine gait analysis. So far, treatises on gait analysis It has been argued that ‘the invention of motion pictures can be
had largely consisted of theoretical considerations while conclu- traced to an argument among the ancient Egyptians whether a trot-
sions based on experimental data were scarce. This was mainly due ting horse ever had all four feet off the ground at once’ (Simpson,
to the limitations of the human eye when observing the faster gaits. 1951). Though this certainly is a bit of an overstatement, it was this
In the middle of the 19th century it was still contentious whether still unresolved question that led to Muybridge’s first photographic
the faster gaits had moments when all limbs were in suspension or
not, as illustrated by the dispute between Gamgee and Goodman
alluded to above.
It is thanks to the efforts of two men that decisive breakthroughs
were made. The English-born American photographer Eadweard
Muybridge and the French physiologist Etienne Jules Marey used
the technology of their time to study equine gait.

Fig 1.6  Gait diagram as introduced by Lecoq (trot). Though later used by
celebrities such as Muybridge, it did not find general acceptance because it
provided information about the spatial position of the limbs only, but not
about temporal aspects. This in contrast to the original ‘échelle odochronomé-
Fig 1.5  Electrical device as developed by Bayer to measure the movement trique’ invented by Goiffon and Vincent which, after modification by Marey,
of the lateral hoof wall. St, needle; B, wire to battery; SA, wire to electric bell. became the world standard.
Reproduced from Bayer, J., 1882. Experimentelles über Hufmechanismus. Oesterr. Monatsschr. Reproduced from Lecoq, F., 1843. Traité de l’extérieur du cheval, et des principaux animaux
Thierheilk. 7, 72–74. domestiques. Savy, Lyon.

Fig 1.7  System consisting of a fixed and a rotating disc

to determine the sequence of limb placement in the
symmetrical gaits.
Reproduced from Raabe, 1883. Locomotion du cheval. Cadran hip-
pique des allures marchées. T. Symonds, Paris.

9
 1 History

Fig 1.8  The set-up at Palo Alto in California as

used by Muybridge during his research on the
locomotion of the horse for Leland Stanford with
24 cameras in a linear array.
Reproduced from Stillman, J.D.B., 1882. The horse in motion
as shown by instantaneous photography with a study on
animal mechanics founded on anatomy and the revelations
of the camera in which is demonstrated the theory of quadru-
pedal locomotion. James R. Osgood & Co., Boston.

Fig 1.9  Position of cameras and track as used in the studies conducted by Muybridge during his period at the University of Pennsylvania. B, lateral background;
C, transverse backgrounds; F, R, batteries of 12 automatic photo-electric cameras; L, lateral battery of 24 automatic photo-electric cameras; O, position of opera-
tor; T, track.
Reproduced from Muybridge, E., 1957. Animals in motion. Original edition 1899. Republished: Brown, L.S. (Ed.), Dover Publications, New York.

experiments. The railroad magnate Leland Stanford, the founder of The thrust against each thread completed an electric circuit and
Stanford University, was intrigued by this question with respect to effectuated a photographic exposure. He managed to get pictures
his trotter ‘Occident’ and it was at his farm in Palo Alto, California, of an excellent quality as, through an ingenious combination of
that Muybridge commenced his experiments in 1872 (Fig. 1.8). clockworks and electro-magnetic circuits, he had finally succeeded
His first efforts were unsuccessful because his camera lacked a in bringing exposure time down to one six-thousandth second (in
fast shutter. Then the project was interrupted because Muybridge a time when an exposure of half a second was considered
was being tried for the murder of his wife’s lover. Though acquit- instantaneous!).
ted, he found it expedient to travel for a number of years in Muybridge did not only study equine locomotion, though it
Mexico and Central America taking publicity photographs for the formed the major part of his work. He also focused on other domes-
Union Pacific railroad, owned by Stanford. In 1877, he returned ticated species, wild animals and man. His book Animal Locomotion,
to California and later pursued his work at the University of first published in 1887, has been republished several times. Muy-
Pennsylvania (Fig. 1.9). Muybridge placed 24 single-lens cameras bridge also invented the ‘zoöpraxiscope’, a device that consisted of
on a row. The cameras were triggered in sequence by a series of a large glass disc on which successive pictures were printed. By
thin threads that were stretched across the path of the animal. projecting these in rapid succession on a screen, it gave an

10
 The 19th century

A B C

Fig 1.10  Devices developed by Marey to quantify equine locomotion. (A) ‘Chaussure exploratrice’ or exploratory shoe; (B) the recorder indicating limb placement
of the horse; (C) air-filled bracelet for the discrimination between stance and swing phase on hard surfaces where the ‘chaussure exploratrice’ could not be used.
Reproduced from Marey, E.J., 1882. La machine animale. Locomotion terestre et aérienne. Germer Baillière et Cie, Paris.

impression of a moving picture. In fact, this was a forerunner of have to for ever envy the birds and insects for their wings or
present-day cine film, the invention of which is usually credited to whether it would be possible some day for him to travel
Thomas Edison, though it is known that Edison derived some of through the skies as he travels the oceans? At various times
his basic ideas from Muybridge. Early in 1888 Muybridge even scientific authorities have declared, after having made
discussed with Edison the possibility of producing talking pictures elaborate calculations, that this was an idle dream. But how
by synchronizing a zoöpraxiscope with a phonograph. As the pho- many inventions did we not see that had been declared equally
nograph at the time was not loud enough to be heard by an audi- impossible beforehand?)
ence, the idea was abandoned (Muybridge, 1957). It would take
another 40 years before the talking picture would conquer the Twenty years later the Wright brothers would make their first flight
world. and 80 years later long-distance maritime passenger transport
Though originally a photographer, Muybridge was also some- would have been almost totally replaced by air travel.
thing of a scientist. His book Animals in motion has a better scientific In his book, Marey studied both terrestrial and aerial locomotion.
base than the book The horse in motion as shown by instantaneous The studies on terrestrial gait focused on the horse. Three ingenious
photography with a study on animal mechanisms founded on anatomy devices were used to study the equine gaits in a relatively accurate
and the revelation of the camera in which is demonstrated the theory of way. To discriminate between stance and swing phase Marey used
quadrupedal locomotion by the physician J.D.B. Stillman (1882). a ‘chaussure exploratrice’ or ‘exploratory shoe’ (Fig. 1.10A). This was
Stanford provided many of Muybridge’s photographs to Stillman essentially an India rubber ball filled with horsehair that was
without giving credit to the original photographer. Muybridge also attached to the horse’s foot. At hoof placement the ball was com-
suggested in a letter to Nature in 1883 that the photographic tech- pressed. The increase in pressure, transmitted by airtight rubber
nique could be used to identify the winner of horse races when the tubing, was registered by a recorder in the rider’s hand (Fig. 1.10B).
finish was very close (Leach & Dagg, 1983). Indeed, in 1888, the The recorder consisted of a charcoal-blackened rotating cylinder on
world’s first photo finish was made in New Jersey. which traces were made by a needle that reacted to changes in air
In the meantime, in France, the physiologist and university pro- pressure. As this device wore rapidly on hard surfaces, a second
fessor E.J. Marey investigated equine gait with equally inventive, but instrument was made (Fig. 1.10C). It consisted of a kind of bracelet
somewhat different techniques. Marey was intrigued by the similar- that was fastened to the distal limb just above the fetlock joint and
ity of natural mechanisms and mechanical machinery and was that functioned according to the same principle. A third device
convinced that a more profound study of the former, especially in consisted of two collapsible drums that were fastened to the withers
the area of locomotion, would lead to substantial progress in and the croup, with levers attached to record vertical movements in
mechanical engineering. In the preface of his book La machine the gaits.
animale, locomotion terrestre et aérienne (Marey, 1882a) he writes: Marey discussed various notations of gaits and concluded that
the notation by Goiffon and Vincent was by far superior. He adapted
Quant à la locomotion aérienne, elle a toujours eu le privilège this method somewhat and his notation, depicting limb placement
d’exciter vivement la curiosité chez l’homme. Que de fois ne by sequential open and filled bars, is still in common use today.
s’est-il pas demandé s’il devrait toujours envier à l’oiseau et à Marey worked out the exact sequence of foot contacts, but his cal-
l’insecte leurs ailes, et s’il ne pourrait aussi voyager à travers culation of how long each foot remained on the ground was too
les airs, comme il voyage à travers les océans? A différentes short. Like Muybridge, he demonstrated the short suspension phase
époques, des hommes qui faisaient autorité dans la science ont of the trot and he also correctly deduced that the hindquarters gave
proclamé, à la suite de longs calculs, que c’était là un rêve the main propulsion whereas the forequarters provided support
chimérique. Mais que d’inventions n’avons-nous pas vu réaliser (Leach & Dagg, 1983).
qui avaient été pareillement déclarées impossibles? Though the techniques used in La machine animale are mainly of
(Aerial locomotion has always provoked a vivid curiosity in a mechanical nature and not photographic, Marey in fact is also
man. How many times did man wonder whether he would one of the pioneers in photography (Marey, 1882b, 1883). At first

11
 1 History

he used multiple exposures on the same photographic plate, later photographer Otoman Anschütz for his largely theoretical treatise
he made a rotating plate not unlike Muybridge’s ‘zoöpraxiscope’. He on the jumping horse (1912). In the proceedings of the Kaiserlich-
also produced flight arcs of several segments of the body by repeated Königliche Botanische Gesellschaft in Vienna in early 1917 professor
exposure of black objects with reflecting markers at anatomically Keller showed the results of own kinematic experiments. Keller had
defined points moving against a black background. Most of these constructed a turntable, which was moved by the horse itself, in the
techniques were applied to study human locomotion, but photo- center of which was a camera. This system ensured a strictly lateral
graphs of horses were also made. In the latter the superposition of view although the horse was not walking along a straight line. Keller
hind limb markers over forelimb markers using the repeated expo- filmed at a rate of 32–50 fps (frames per second), and showed the
sure technique made interpretation of the data a difficult job. He film at the standard rate of 16 fps, thus creating a slow motion
and his coworkers, Pagès and Le Hello, published a fine series of effect. Schmaltz (1922a,b) used essentially the same technique, but
articles on the subject in the Comptes Rendus Hebdomodaires des used the film mainly to produce a photographic series that shows
Séances de l’Academie des Sciences (Le Hello, 1896, 1897, 1899; the characteristics of each gait. It is Walter (1925) who, under the
Marey & Pagès, 1886, 1887; Pagès, 1885, 1889). guidance of Schmaltz, extensively used the turntable in his study of
Coincidentally, these two great men of equine gait analysis, Muy- limb placement sequence and changes in joint angles during walk,
bridge and Marey, were born 1 month apart in 1830 and died trot and gallop. Walter admits the disadvantage of the circular
within 1 week from each other in 1904. They met in Marey’s labora- movement and indicates that the clinical department of the Berlin
tory, in the presence of a large number of scientists from all over Veterinary High School made use of a linear rail over which a
the world who attended the Electrical Congress in Paris in 1881. camera could be moved by mechanical power in order to keep up
This was when Muybridge gave the first demonstration of his with a horse moving on a parallel path.
‘zoöpraxiscope’ in Europe (Muybridge, 1957). In a publication from 1923, A.O. Stoss from Munich used pho-
tographic methods to study the anatomy and kinematics of the
equine limbs. In the section on the shoulder he remarked that it
German supremacy until World War II was a pity he could not use Muybridge’s pictures as for the exact
location of the skeletal parts that make up the shoulder, because
A nice synopsis of late 19th century state of the art in the field of only skinny horses could be used for this purpose. Apparently,
equine gait analysis is given in the book by Goubaux and Barrier Muybridge’s horses were too fat!
De l’extérieur du cheval (On the conformation of the horse, 1884). They Also, studies using techniques other than photography were
describe both Marey’s techniques and Muybridge’s work. They cite being performed, again mainly by anatomists. Dörrer (1911),
Marey as saying that, if he had to do his experiments again, he working at the Königliche Tierärztliche Hochschule in Dresden,
would use an electric circuit instead of a pneumatic system. In fact, wrote a thesis on the tension in the flexor tendons and the suspen-
it appears that Barrier repeated his measurements in 1899 using an sory ligament during various phases of the stride cycle. For his in
electrical device to improve accuracy (Schauder, 1923b). Goubaux vitro work he used a device that had originally been designed by
and Barrier also described some other means to represent equine Moser (Fig. 1.12). Strubelt (1928), who worked in Hanover, found
gait such as a kind of adjustable rattle that would reproduce the that transecting either the lacertus fibrosus or the peroneus tertius
sounds made by the hoof beats in the various gaits, and a wooden muscle did not affect locomotion in the living animal, or the ana-
table that was over 8 feet high through which limb placement in tomical relations in a specimen of the hind limb that was brought
various gaits could be visualized (Fig. 1.11). under tension.
Following Goubaux and Barrier, the period of French supremacy Bethcke (1930) focused on the relationship between morpho-
in gait analysis ended. The French were very ingenious inventors, metric data and performance in the trotter. Earlier studies on the
but there is little work by French authors on gait analysis dating subject had been performed by Bantoiu (1922) in Berlin and Birger
from after 1900 and it was essentially the Germans who chose to Rösiö (1927) who performed measurements on Standardbreds in
follow the path indicated by Marey and Muybridge. It was only Sweden, Germany and the United States. Bantoiu was one of a series
through sequential photography, soon followed by cine film, which of Rumanian vets who, under the guidance of professor Schöttler
is essentially the same, that the faster equine gaits could be fully in Berlin studied the relationship between conformation and per-
explored. formance in various breeds. His colleagues Stratul (1922), Nico-
After laying the foundations in the late 19th century, German lescu (1923) and Radescu (1923) studied this relationship in
veterinary science had its golden age in the first half of the 20th Thoroughbreds and Hanoverian horses. Though Bethcke is able to
century. Many disciplines flourished, but perhaps none so abun- give some data on anatomical differences between various breeds
dantly as the discipline of veterinary anatomy. Wilhelm Ellenberger of horses, he has to conclude that he could not predict performance,
(1848–1928) and Hermann Baum (1864–1932) published many stating that:
editions of their Handbuch der vergleichenden Anatomie der Haustiere
(Handbook of comparative anatomy of domestic animals). This work was Wenn uns so auch die Maße für die Beurteilung eines
so complete that a facsimile version of the 1943 edition was still in Trabers gewisse Anhaltspunkte geben, so sind für seine
print in the 1970s. Another excellent anatomist was Paul Martin Leistungsfähigkeit letzten Endes doch noch andere Faktoren
who moved from Zurich to Giessen in the first years of the 20th wie Training, Temperament, Abstammung, Ausdauer,
century. Two of his pupils were to become famous anatomists them- Beschaffenheit der inneren Organe usw. mit ausschlaggebend.
selves: Wilhelm Schauder at Giessen University and Reinhold (Though the measurements may give us some clues for the
Schmaltz in Berlin. In Germany, the study of muscle function and judgement of a trotter, performance is in the end more
locomotion was in the hands of the anatomists. It is therefore not determined by other factors such as training, character,
surprising that many of these individuals became involved in descent, endurance and condition of the internal organs.)
equine gait analysis and/or biomechanical studies. It has been
reported that Ellenberger studied the gallop by attaching 4 different- This conclusion has not changed in the past 80 years.
sounding bells to the feet of the horses (Schauder, 1923b). Schmaltz The relationship of conformation and locomotion was heavily
was among the first to extensively use cine film for equine gait studied in pre-war Germany. Wiechert (1927) studied East-Prussian
analysis, and Schauder published on equine gait and related topics cavalry horses to find morphometric criteria for performance poten-
throughout a long career that extended until after World War II. tial. Though he finds some biometric differences between horses
Chief veterinary officer in the first Dragoon Regiment at Berlin, selected for specific purposes, the study lacks any statistical elabora-
Werner Borcherdt, used the pictures by Muybridge and the German tion of the data. Buchmann (1929) focused mainly on stride length

12
 The 19th century

Fig 1.11  Wooden table (A) invented by Goubaux and Barrier

to represent the footfall pattern in various gaits. Drawers A
and B contain nickel-plated shoes for fore and hind limbs.
Drawer C contains a series of cards (B) on which the correct
position of the feet in a certain gait are described.
Reproduced from Goubaux, A., Barrier, G., 1884. De l’extérieur du cheval.
Asselin et Cie, Paris.

in various breeds. Kronacher and Ogrizek (1931) published a com- conformational indications could be found. Therefore, it was better
prehensive study using 60 Brandenburger mares. A follow-up to this to accurately measure the actual stride length when buying cavalry
study was performed by Horst Franke (1935) who studied 186 horses! In 1934 Wagener published a comprehensive study on
mares from the famous stud in Trakehnen (East-Prussia). A positive jumping horses and in 1944 Wehner wrote on the relation of bone
relationship between the length of some limb segments and stride axes and joint angles and stride length in the German Coldblood.
length was found, but results on the influence of joint angulation In Munich research was more biomechanically oriented. Moskov-
were not consistent. While Kronacher and Ogrizek report a clear its (1930) studied statical-mechanical aspects of the equine meta-
positive relationship between stride length and shoulder and elbow carpus. Max Kadletz studied the biomechanical behavior of the
angles, Franke is more cautious stating that joint angles are much small tarsal joints in detail in relation to the pathogenesis of bone
less important in determining stride length than the dimensions of spavin (Kadletz, 1937). He also paid attention to the movements
limb segments. The study of Schmidt (1939) is even less conclusive. of fore and hind limbs, and addressed the age-old problem of how
He studied 100 cavalry horses with the aim of determining confor- artists should depict horses, inventing some makeshifts for this
mational characteristics that were indicative of a long stride, which purpose (Kadletz, 1926, 1932, 1933).
was a desirable trait of cavalry horses that often had to cover more Possibly the most prolific of the pre-war German scientists was
than 30 miles a day, and concluded that no unequivocal Wilhelm Krüger from Berlin. He published a very elaborate and

13
 1 History

Fig 1.13  The so-called ‘Balance of Borelli’ as used by Krüger to determine the
center of gravity of the horse.
Reproduced from Krüger, W., 1939. Die Fortbewegung des Pferdes. Paul Parey, Berlin.

Fig 1.12  Device to (semi-quantitatively) determine the tension in the flexor on the gaits of the horse, Magne de la Croix from Argentina started
tendons and the suspensory ligament in various positions of the limb as to study photographic material, including the series made by Muy-
designed by Moser and used by Dörrer. bridge, and became so interested in the subject that he wrote an
Reproduced from Dörrer, H., 1911. Über die Anspannung der Beugesehnen des Pferdefußes article on the evolution of the gallop in which he stated that, in
während der verschiedenen Bewegungsstadien derselben. Inauguraldissertation, Dresden. evolutionary terms, the rotatory gallop was more advanced than the
transverse gallop (Magne de la Croix, 1928). He can be seen as a
forerunner of the great American zoologist Milton Hildebrand, who
authoritative study on the kinematics of both fore- and hind limbs four decades later stated that the various gaits are not that distinct
(Krüger, 1937, 1938). Krüger did not use a turntable like his prede- as hitherto presumed but in fact form a continuum that may gradu-
cessors Schmaltz and Walter, but made use of a vehicle that moved ally change into each other, when he writes:
alongside and at the same speed as the horse, probably the same
installation that had already been described by Walter. He was well Un hecho en el cual nunca me había fijado, resaltó de repente
aware of the artifact introduced by the use of skin markers, which a mis ojos, y es que el galope de carrera de los varios animales,
had already been noted by Fick (1910) who had stated: en vez de presentarse sólo bajo dos formas: transverse gallop y
rotatory gallop, como los llama Muybridge, ofrece, en realidad,
Vor allem ist es schwierig, an den bewegten Gliedern una infinidad de variantes que constituyen una cadena
bestimmte Punkte sicher zu markieren die ihre Lage auf der insensible y continua.
Körperoberfläche während der Bewegung nicht ändern, da sich (A fact that so far had never occurred to me became
ja auch die Haut in der Nähe der Gelenke beträchtlich suddenly clear. This is that the racing gallop of the various
verschiebt. species, instead of presenting in only two forms: transverse and
(In the first place it is difficult on the moving limbs to rotatory gallop as Muybridge calls them, in fact presents as an
mark sites that do not alter their positions on the body surface infinite number of variations that form a continuous,
during movement, as the skin near the joints shifts imperceptibly changing chain.)
considerably.)
He developed the idea further and included all gaits and many
Krüger avoided the problem by oiling the skin and using oblique different animals in his extensive paper on the phylogeny of qua-
lighting, thus visualizing the position of the bones directly. Though drupedal and bipedal locomotion in vertebrates (Magne de la
the skin displacement artifact is taken into account in most pre-war Croix, 1929) (Fig. 1.14). He continued publishing on the subject
research (in contrast to many much more recent studies), this was until the mid-1930s (Magne de la Croix, 1932, 1936). Armando
not always the case and Krüger does not hesitate to blame this Chieffi in Brazil started his investigations of locomotion with a large
artifact for the discrepancies of his findings with those by Aepli study of the position of the center of gravity (Chieffi & de Mello,
(1937) who, in his cinematographic measurements of joint angles 1939). He continued with papers on the ‘marcha’ (an artificial gait)
during locomotion, failed to correct for the artifact (Krüger, 1938). in the Mangalarga horse (1943) and with studies comparing the
Krüger was a dedicated scientist and a prolific writer. Apart from his stance phases of different gaits and the change of gallop (Chieffi,
cinematographic work on gait analysis, which he summarized in a 1945, 1946). Finally he wrote a thesis (1949) on the subject of the
treatise on the movement of the horse (1939a) (Fig. 1.13), he wrote transition of gaits, a subject on which little scientific work had been
on the oscillations of the vertebral column (1939b), specific limb done as he concluded from his review of the literature:
placement during the gallop and while jumping (1939c,d), the
position of the center of gravity during locomotion (1941a), and O exame de literatura a respeito…… revelou que pouco ou
the effect of hauling heavy loads on the tendons in the forelimb nade existe sôbre transição de andamentos.
(1941b). (The literature search showed that there was very little if
Though German scientists were dominant in the era between the anything on the subject of changes of gait.)
turn of the century and World War II, this was not the only site of
activity. The work of Aepli (1937), which has been referred to It still is a relatively unexplored area.
already, was performed in Zurich. Before World War II, there were In Belgium, Zwaenepoel published the same study on the impulse
also scientists in South America who focused on equine gait analy- of gait in the horse twice in different journals (Zwaenepoel, 1910a,b,
sis. When asked to write a series of articles for a sports magazine 1911a,b). In Holland, Kroon (1922, 1929) and van der Plank

14
 The horse in decline

Fig 1.14  Depiction of various gaits by Magne de la Croix. (A) Trotting horse; (B) pacing camel; (C) 3-beat gallop; (D) horse featuring 4-beat racing gallop.
Reproduced from Magne de la Croix, P., 1929. Filogenia de las locomociones cuadripedal y bipedal en los vertebrados y evolución de la forma consecutiva de la evolución de la locomoción. Anal.
Soc. Argent. 108, 383–406.

In the United States work in this area of research was scarce. In

1912, Gregory published a largely theoretical treatise on quadrupe-
dal locomotion and Chubb of the American Museum of Natural
History used photographs of horses as aids in the preparation of
accurate specimens for the Museum (Berger, 1923). In 1934, Harry
H. Laughlin, from the Carnegie Institution’s Department of Genet-
ics, published a study in which he developed a formula based on
age, weight carried, distance run and time taken that quantified
Thoroughbred performance in a given race. These figures were
meant to serve as a ‘mathematical yardstick’ to be used in the search
to find the natural laws governing the transmission of inborn racing
capacity (Laughlin, 1934a,b). Less specific to the horse was the work
of John T. Manter of South Dakota who wrote on the dynamics of
quadrupedal walking (1938). The same applies to the book on
Fig 1.15  Electric device to measure deformation of the hoof wall, adapted speed of animals by A. Brazier Howell (1944). Also in 1944, J. Gray
by Kroon (1922) from the original by Bayer (1882). from Cambridge (UK) published a comprehensive study on the
Reproduced from Kroon, H.M., 1922. Een bijdrage tot de studie van het hoefmechanisme. mechanics of the tetrapod skeleton. Even in Bolshevik Russia,
Tijdschr. Dierg. 49 (11), 399–406. research was performed. Ivanov and Borissov (1935) continued the
line set by Forssell (1915) and Strubelt (1928); studying the impor-
tance of the Lacertus fibrosus in the standing and moving horse.
(1929) performed some detailed studies on the hoof mechanism Schtserbakow (1935) tried to use kinematic data for the determina-
using electric devices, not photography or cine film (Fig. 1.15). tion of the optimal amount of work a horse should be given during
Captain Carnus from the military training school at Saumur was training. Afanasieff (1930) claimed that he found correlations
one of the few French authors of the era who wrote on the subject. between some morphometric variables and maximal speed. He did
His was a largely theoretical study on the role of the muscles of the not fail to mention that, whereas before the Russian Revolution
neck in forelimb motion (Carnus, 1935). From an earlier date is there was a great interest in the American trotter and many crossings
the work by another cavalry officer, who wrote a book on equine were made with the Russian Orlow trotter, the present breeding goal
motion that contains experimental cinematographic parts on show was to develop the pure Orlow trotter breed and to cut out any
jumping and the racing gallop (De Sévy, 1918). In Sweden, the American influence.
well-known veterinarian Forssell investigated the effect of the tran-
section of various tendons and ligaments (Forssell, 1915) and
Palmgren (1929) studied the angulation pattern of the elbow joint The horse in decline
in relation to muscle use and muscle conformation (‘feathering’ of
certain muscle groups) in Swedish Standardbreds using high-speed World War II left large parts of Europe in ruins. Apart from the tragic
cine film. In Czechoslovakia, Kolda (1937) published on the func- loss of life for millions of people, the war had a devastating effect
tional anatomy of the equine shoulder joint. on many aspects of pre-war society. In Germany, where research on

15
 1 History

equine gait analysis and biomechanics had flourished as nowhere Giessen University, continued publishing. He was principally inter-
else, Bernhard Grzimek was still performing research on handed- ested in the functional development of the musculoskeletal system.
ness in horses, parrots and monkeys in Frankfurt Zoo in the summer In his pre-war publications he had focused on the development of
of 1944 when the Allied Forces already were fighting their way to various parts of the equine musculoskeletal system (Schauder,
Paris through the heavily defended French province of Normandy. 1924a, 1924b, 1932). After the war he continued with this subject
He submitted the paper on October 5th 1944, about at the time when quoting Goethe to express his basic presumption: ‘Gestalten-
Montgomery’s troops were defeated in the Battle of Arnhem, which lehre ist Verwandlungslehre’ (Morphology is the science of change,
would prolong the war for about half a year. However, after the Schauder, 1949). In the early 1950s, he concentrated on shock-
capitulation of the Third Reich in May 1945, science in Germany, absorbing structures in the equine limbs and rump (Schauder, 1951,
like practically all aspects of public life, came to a grinding halt. 1952, 1954), which he discussed on a theoretical basis.
Grzimek’s paper finally was published in 1949. The tradition at Giessen was continued by L. Krüger (not to be
Although public life regained its vitality earlier in the formerly confused with Wilhelm Krüger from Berlin), who published on the
occupied and now liberated parts of Europe, the situation was not hauling capacity of horses (and cattle) in order to determine per-
essentially different. Attention was focused on repairing the enor- formance capacity (Krüger, 1957). At the Institute of Animal Physi-
mous damage caused by the war, rather than on developing new ology of the University of Bonn (which features no veterinary
scientific approaches. faculty), Kaemmerer (1960) used photography to study the flight
Apart from the direct consequences of the war, there were other arcs of various parts of the equine limbs. He concluded that these
developments that failed to stimulate new research in the field of flight arcs cannot be seen as parts of a circle, as had been stated
equine locomotion. In World War I horses had been used exten- before, but in fact are complicated cycloids which change their form
sively. The British expeditionary force in France in 1914 began with when the horse is more heavily loaded. He stated, not without
53 000 horses, but it is estimated that in 1917 the army had more surprise, that the work by Walter (1925) had not received the rec-
than 1 000 000 horses in active service over all fronts (Dunlop & ognition it merited. He had come across Walter’s work only after
Williams, 1996). On the German side the number of horses was finishing his own experiments, and concluded that both investiga-
reported to be 1 236 000 in the same year, not counting those tions, performed using different techniques, generally confirmed
belonging to the army contingents employed at home or in the each other.
occupied territories (von den Driesch, 1989). In German East Africa Equine gait analysis still formed a topic for a few of the many
(Tanzania), the commander of the relatively weak forces, First Lieu- veterinary doctorate theses at the German universities. Richter
tenant von Lettow-Vorbeck, developed a ‘veterinary strategy’ to (1953) in Berlin, continued the tradition of the Rumanian veteri-
combat an overwhelmingly more numerous army, consisting of narians from the early 1920s when he studied the American-bred
British, South African, Belgian and Portuguese forces. Using his trotter in order to correlate morphometric data with performance.
superior knowledge of the local situation of tropical diseases, espe- In Giessen, Maennicke (1961) and Genieser (1962) were among the
cially trypanosomiasis, and giving better prophylactic care to his first to work with (Shetland) ponies and not with horses. The
own animals, he consistently retreated through the tsetse infested Turkish veterinarian, Ihsan Aysan (1964), analyzed the gait of lame
areas, inflicting heavy losses in animals on the Imperial Forces. On horses during his years in Giessen in which he prepared his thesis.
Armistice Day in 1918, he still was at large with his last 1323 troops, He worked with a 3-m long track and with a fixed camera using a
pursued by an army of 120 000. film speed of 132 fps. As the distance of the horse to the camera
At the start of World War II, in September 1939, Polish lancers was not constant, complicated mathematical procedures were nec-
tried to stop the invading German tanks. Not surprisingly, they suf- essary to calculate the exact locations of certain anatomical sites.
fered heavy losses and were not able to slacken the advance of the Though the principle of this mathematical data processing is the
enemy. Though the German army still used horses extensively for same in the modern video-based systems, one should be aware of
transportation (a mean population of about 1 350 000 has been the fact that this study was undertaken before the advent of the
estimated of which an average of 59% was lost), it became clear computer. As Aysan could not statistically process the data, his work
that the role of the horse in warfare finally had come to an end after basically consists of extended case reports of various types of
5000 years. There is no modern army in which the horse plays a lameness.
prominent role, except perhaps for some ceremonial duties. Though the interest in equine locomotion certainly was at a low
The increasing mechanization not only influenced the military in the period between the end of World War II and the early 1970s,
role of the horse. Numbers of horses in the United States reached some seeds were sown of what in later decades would become rich
their peak in 1918 at 21 000 000. After that year, which is also the fruit-bearing trees. In Sweden, Björck (1958) was the first to use a
year that automobile production first passed the 1 million mark, force-shoe to analyze the ground reaction forces exerted by the
the population more than halved to an estimated 8 million horses horse. In Vienna, the opening of the reconstructed lecture hall of
and 2 million mules in 1947 (Simpson, 1951). In Britain, the agri- the anatomy department in 1950 was celebrated with a lecture by
cultural horse population in 1913 was 1 324 400; in 1956 it was Schreiber on the old theme of the anatomically (in)correct depic-
only 233 500 (Brayley Reynolds, 1957). It was to be expected that tion of horses in art.
the day would not be far away that the last commercially used horse The then young assistant Peter Knezevic used strain gauges and
could be turned out to pasture after retirement. Of course, this trend cinematography to study the hoof mechanism. With a 4-channel
affected the numbers of patients presented for treatment to the recorder he was able to make synchronous ungulographic record-
veterinary schools. In the pre-war period, the Utrecht Clinic of Large ings at walk, trot and gallop (Knezevic, 1962). In Holland, Slijper
Animal Surgery received more equine than bovine patients. This had already published an extensive study on the vertebral column
ratio changed to 1 : 1 in the early post-war period and remained so and spinal musculature of mammals in 1946. However, the tradi-
through the early 1960s, followed by an increase in equine patients tion of biomechanical research and gait analysis at Utrecht Univer-
again from 1964 onwards, now in the form of sports and leisure sity, which is maintained to the present day, can be said to have
animals (Offringa, 1981). started with the publication by Dick Badoux in Nature on the fric-
In view of the declining role of the horse in society, it is not tion between feet and ground (1964). Many publications were to
surprising that research on equine locomotion received less priority follow, all focusing on the biomechanics of specific parts of the
than in the years before the war. Nevertheless, some activity musculoskeletal system (Badoux, 1966, 1970a, 1970b). There was
remained. In Germany the old tradition had not completely been also a start of the input of biomechanics in essentially clinical work,
broken by the war, though the relative number of publications as shown by the thesis of Rathor (1968) on disorders of the equine
decreased considerably. Of the pre-war scientists, Schauder, at and bovine femoropatellar articulation.

16
 Equine locomotion research centers and activities

Elsewhere, activity was very limited in this field of research. In competition. The official history of horse racing tells us that this
France, a thesis was published by Marcel André (1949) on static, sport started in 648 BC in Olympia in Greece (Simpson, 1951), but
dynamic and cinematic aspects of equine locomotion. In Switzer- it is highly improbable that horse races were not common in the
land some work was performed on the biomechanics of the equine few thousand years between domestication of the horse and that
elbow joint in the late 1960s (Mosimann & Micheluzzi, 1969). In date. Apart from its role in competition, the horse, unlike any other
Eastern Europe, there was some interest in the relation between animal with the possible exception of the dog, has always had man’s
conformation and performance. Fehér focused principally on bio- affection. This special bond between man and the horse is already
metric data concerning the horse with a normal configuration evident in the earliest human writings on the species in Antiquity
(Fehér, 1957, 1958). From the early 1960s, Dušek from Czechoslo- and remains so, through the great horse marshals of the 17th
vakia started a series of publications on the relation of a number of century and many others in the course of time, to the present day.
conformational parameters and performance like jumping ability With the booming economy in the 1960s, popular interest in the
and how to correctly evaluate these parameters. Unfortunately, horse could be materialized. Equestrian sports had existed for thou-
some of this work was published in the Czech language, which is sands of years, but had always been restricted to the lucky few. Now
not accessible to many scientists (Dušek & Dušek, 1963, Dušek they became within reach of the general public. From the end of
et al., 1970). In the German Democratic Republic, some work was the 1960s and beginning of the 1970s equestrian sports flourished
done on the conformation of trotters by von Lengerken and Werner as never before. This was evident in very old and well-known
(1969). However, the breed was not very important in socialist days branches of the equestrian sports such as flat racing and harness
as there was no more than one racetrack in the whole country. In racing all over the world, the Western style activities in the United
the Soviet Union, Sukhanov (1963) focused in a more general sense States, and dressage and jumping in Europe. Popular interest also
on the evolution of gait. In Japan Nomura published a series of increased in other, less-known, areas such as three-day-eventing,
reports on the mechanics of a number of equine joints in the early vaulting, four-in-hand driving and endurance competitions. New
1950s (Nomura, 1953a,b,c). competitions were created in several of these disciplines and the
In the Anglo-Saxon world, the interest in equine gait analysis and time was ripe for the organization of large events like the World
biomechanics was still limited. In 1951 Grogan published a general Equestrian Games, the first of which was held in Stockholm
descriptive article about gaits in horses which ominously opens in 1990.
with the words: ‘With the drop in the number of working horses, The increased interest in the horse was the trigger for the revival
the study of the horse has held less interest for most veterinarians in equine locomotion research that started in the early 1970s.
and received less attention in the schools’. In England, H.W. Dawes Though the economic role of the horse had disappeared (except for
(1957) published a paper on the relationship between conforma- the increasing economic significance of the horse industry itself),
tion and soundness, giving detailed descriptions of some more or the need for research into equine locomotion was now even greater
less desirable traits. The same topic is discussed by Pritchard in the than before. While in earlier times the horse had to be able to do
United States nearly a decade later (1965). Though interesting, both its job properly which required a functional, but not necessarily
papers are based on clinical impressions rather than scientific superior, locomotor system, the horse now had become an athlete
research. This is different with the work of the great American zoolo- upon which high demands were made. This prompted the need for
gist Milton Hildebrand who studied, using film among other tech- a highly accurate analysis of normal and abnormal gait and of the
niques, the gaits of tetrapods. Though he used the horse frequently ways in which equine locomotion could be influenced or improved.
as a study object, his interests were broader. The analysis and inter- Rapid developments in computer technology enabled the produc-
pretation of the gaits of tetrapods, including the energetics of oscil- tion of both hardware and software that facilitated capture and
lating legs in relation to conformation and gait were his main analysis of the faster movements of the horse in intricate detail.
targets, not the horse itself (Hildebrand, 1959, 1960, 1965, 1966). These were the main factors that determined the explosion in
However, in the late 1960s some horse-specific research was pub- equine locomotion research that started in the late 1970s and con-
lished in the United States (Taylor et al., 1966; Rooney, 1968, 1969; tinues today.
Solá, 1969; Cheney et al., 1970), indicating that we are in the wake Thus far, an attempt has been made in this chapter to provide a
of what may be called revolutionary changes in the science of comprehensive and, to the author’s knowledge, as complete as pos-
equine biomechanical research and gait analysis. sible overview of the literature on equine gait analysis because many
of the older works can not be found with help of the currently used
electronic databases. However, from this point on this is no longer
The revival in equine locomotion research the case. It is far beyond the scope of this chapter to provide an
exhaustive bibliography of the vast amount of work that has been
In most countries in the world the era directly after World War II done in recent decades. These references can easily be found in
had been a period of hardship in which the damage caused by the modern databases and most of them will be mentioned in the other
war, either materially or economically, had to be repaired. In large chapters of this book, which are dedicated to virtually all aspects of
parts of the world it was a period of lack of resources in which equine locomotion. In the following section a brief overview will
people were forced to work hard while leading a life deprived of be given of the development of the main centers of equine locomo-
any luxury. Gradually this picture changed since, from the late tion research in order to provide a link from the rich history of this
1950s and early 1960s, economies began to boom. The mid 1960s area of research to the practice of today. References will be restricted
onwards marked the start of a period of unprecedented economic to some key publications, but make no pretense of completeness.
growth and increasing prosperity in the developed world. While old
sources of richness, such as the former colonial empires, disap-
peared new technological developments enabled manifold increases Equine locomotion research centers
in human production capacity, leading to a large increase in cheap and activities
consumer goods. This development led to a period of wealth and
prosperity in the industrial world.
The horse had lost its role in the military completely and in Europe
agriculture and transport to a large extent. However, it had, after
5000 years of close alliance, not lost its appeal to man. The horse Stockholm/Uppsala
had always been a very useful instrument for the satisfaction of one Ingvar Fredricson and coworkers may be credited for the initiation
of the most fundamental drives of mankind, the need for of the revival of research on equine locomotion. In 1970, they

17
 1 History

published a report about the quantitative analysis of hoof motion Gunnar Hjertén, focused on kinematic analysis of the Swedish Stan-
patterns of harness horses using high-speed film in the proceedings dardbred. In Sweden, harness racing had always been a popular
of a congress on high-speed photography (Fredricson et al., 1970). sport, but the industry boomed in the 1970s and 1980s with
Shortly after that, the new method of investigating equine locomo- Sweden and France becoming the most important countries for
tion was published in the then recently founded Equine Veterinary harness racing in Europe. This increase in popularity of the sport
Journal (Fredricson & Drevemo, 1971). Fredricson used high-speed was partly due to the very generous fiscal legislation for horse
film (with frame rates up to 500 fps) and analysis methods derived owners. The interest in harness racing was also evident in the exten-
from the aviation industry to process his data. This approach sive research this group has performed on the design of racetracks,
enabled him to analyze in 3 dimensions the very fast movements signaling deleterious effects of poor racetrack design and giving
of the distal limbs of Standardbreds trotting at high speed (Fig. possible solutions for improvement (Fredricson et al., 1975a, b).
1.16). These investigations resulted in his thesis, which may be seen Indeed, their work resulted in considerable improvements with
as the starting point of the modern era of equine locomotion respect to banking and curve geometry of many racetracks. The
research (Fredricson, 1972). Swedish group may also be credited for being the first to use a
The Swedish group, the nucleus of which was formed by Ingvar treadmill for equine locomotion analysis (Fredricson et al., 1983),
Fredricson and Stig Drevemo, later joined by Gøran Dalín and an example that was soon to be followed by many research centers
all over the world.
After Fredricson left the group to head the national stud at
Flyinge, Stig Drevemo took the helm. He became a professor of
Anatomy at the Faculty of Veterinary Medicine of the Uppsala Uni-
versity of Agricultural Sciences into which the formerly independent
Royal Veterinary College had been converted. His series of papers
on equine locomotion that appeared in the Equine Veterinary
Journal in 1980 can be considered as classic (Drevemo et al.,
1980a–c). Drevemo was a strong protagonist of international coop-
eration in the field of equine locomotion research. Together with
Doug Leach from Saskatoon, Canada, he published the first (and
last) edition of a Bibliography of research in equine locomotion and
biomechanics in 1988.
The almost exclusive emphasis on kinematics of the Standard-
bred was broken by the arrival in Uppsala of the English-born
globetrotter Leo Jeffcott. His broad interest in the field of equine
orthopedics included problems related to the back. With the excep-
tion of the 1939 paper by Wilhem Krüger, little attention had been
paid so far to the equine back. There is little doubt that this appar-
ent lack in interest was partly caused by the inaccessibility of this
structure. Jeffcott worked with Dalín and other members of the
Swedish team on normal biomechanics of the back and on various
back-related disorders (Jeffcott & Dalín, 1980; Jeffcott et al., 1982,
1985).
After Jeffcott left for Australia, the Swedish group continued to
perform equine locomotion research using high-speed cinematog-
raphy. Though very reliable and accurate, the technique has a major
drawback in that data analysis is extremely labor-intensive. The
Swedes tried to overcome this problem by automating as far as pos-
sible this analysis using advanced and expensive techniques such as
the Trackeye® system (Drevemo et al., 1993). In the 1990s they
finally opted for the video-based Proreflex® system. Though older
members of the group left or became absorbed in administrative
functions, the Swedish group succeeded in maintaining momentum
by attracting young researchers such as Holmström, Johnston and
Roepstorff. Of these, Holmström focused on the kinematic analysis
of top-level dressage horses (Holmström, et al. 1995), while Roep-
storff and Johnston were engaged in the development of a force
shoe suitable for use on the treadmill, in fact continuing the tradi-
tion begun by Björck in the late 1950s (Roepstorff & Drevemo,
1993).
In 1990, Uppsala housed the International Conference on Equine
Exercise Physiology (ICEEP), a four-yearly conference that had
started in 1982. Although the major part of this conference is dedi-
cated to exercise physiology in a more strict sense, and thus is more
cardiovascular and respiratory system oriented, the conference has
always featured a part on biomechanical musculoskeletal system
Fig 1.16  Diagrammatic representation of the data capture and data analysis oriented research.
procedure as used by the Swedish group in the 1970s when they worked with Energetically directed by Chris Johnston and Lars Roepstorff
high-speed film. (with Stig Drevemo still in the background as Head of Department)
Reproduced Fredricson, I., Drevemo, S., Dalín, G., Hjertén, G., Björne, K., 1980. The application of locomotion research in Uppsala has remained strong. Around the
high-speed cinematography for the quantitative analysis of equine locomotion. Equine vet. J. turn of the century collaboration with the group from Utrecht was
12 (2), 54–59. intensified in a project focusing at the kinematics of the equine

18
 Equine locomotion research centers and activities

back. With the results of that study it became possible to develop, 2004; Peham et al., 2004). However, more fundamental work on
together with the manufacturer of the Proreflex® system Qualisys modeling was performed as well (Peham & Schobesberger, 2004).
AB, a reliable and accurate program (Backin®) for the analysis of
non-invasively captured kinematic data from the equine thoraco-
lumbar spine. Back motion had always been a very difficult area Utrecht
because of the difficult accessibility of the back, the relatively small In Holland, the foundation for the line on biomechanical research
ranges of motion of the components of the back and the fact that, had been laid in the 1960s by Dick Badoux from the Department
unlike the limbs, movements out of the sagittal plane contribute of Anatomy at the University of Utrecht. One of the products of this
significantly to overall motion. Work from Uppsala has made a line was the still authoritative study by Wentink on the biomechan-
substantial contribution to the knowledge about equine back ics of the hind limb of horse and dog (Wentink, 1978a,b). However,
motion in normal and clinically affected horses (Johnston et al., it was the strategic alliance between the Departments of Anatomy
2004; Roethlisberger-Holm et al., 2006; Rhodin et al., 2005; and of General and Large Animal Surgery that really boosted this
Wennerstrand et al., 2004). Other lines pursued in Uppsala included kind of research from the end of the 1970s onwards. For this stra-
work on shoeing (Roepstorff et al., 2001; Johnston & Back, 2006) tegic alliance, the vision of the department heads, Professors W.
and on modeling and the more technical analytical aspects of Hartman and A.W. Kersjes respectively, should be given full credit.
equine motion (Halvorsen et al., 2008). The Uppsala group was, First, Schamhardt and Merkens focused on the analysis of the
together with researchers from Zurich and Utrecht, one of the ground reaction forces in sound and lame horses using a force plate
founders of the so-called SDS (Swiss-Dutch-Swedish) consortium (Merkens et al., 1986; Schamhardt & Merkens, 1987). Later, kine-
that initiated a groundbreaking research project on the influence of matic analysis was added. The Dutch group did not opt for high-
head and neck position on dressage horses that managed to bring speed film or the then recently introduced first video-based systems
science closer to the equestrian sports than it ever had been (see the such as Selspot® or Vicon®, but chose a new invention from England:
paragraph about the Zurich group) and later on started an FEI- the CODA-3® system. This optoelectronic system used a concept
backed project on the influence of the surface on the orthopedic that was basically different from any video-based system and thus
health of the distal limb (Hernlund et al., 2010). avoided a number of disadvantages inherent to these systems.
However, when the prototype with serial number 007 was delivered,
it proved to be not exactly ready-to-use. In fact, it took about 4 years
Vienna and a considerable amount of manpower before the machine was
Another center where research on equine locomotion was initiated working well under the conditions encountered when performing
in the 1970s was Vienna. Peter Knezevic had written a thesis on the gait analysis in the horse. It proved to have a high spatial and tem-
biomechanics of the hoof using strain gauges when he was an poral (300 fps) resolution. Van Weeren, who was intended to be
assistant at the Veterinary High School. After he became head of the first user of the newly purchased system, was forced to redirect
the Department of Orthopaedics of Even and Uneven-hoofed his line of research. He started an investigation into a topic in
Animals of the Faculty of Veterinary Medicine of the University of equine kinematic gait analysis that had been correctly identified by
Vienna, he strongly promoted further research in this field. From most pre-war researchers, but which had been largely ignored so far
the late 1970s, the Vienna group started to publish regularly on in more recent research: the problem of skin displacement. Together
both kinematic (using high-speed cinematography and later also with van den Bogert, Schamhardt and Barneveld, he developed
the Selspot® system) and kinetic studies (using force plates) in the techniques using intra-osseous LEDs, which shone through the
horse (Knezevic et al., 1978; Knezevic & Floss, 1984). The histori- skin, and transcutaneous pins to produce correction factors for skin
cal interest of Knezevic is shown by his 1985 paper in the Wiener displacement (Van Weeren & Barneveld, 1986; Van Weeren et al.,
Tierärztliche Monatsschrift (where many of the papers from the first 1992). In his study on the coupling between stifle and hock joint
Golden Age of equine locomotion research had been published). by the reciprocal apparatus, he was finally able to use the CODA-3®
However, this paper is principally a German copy of the Equine system for the first time (van Weeren et al., 1990). The first person
Veterinary Journal paper published on the same topic by Leach and using the CODA-3® system extensively was Willem Back who, in a
Dagg in 1983. Knezevic’s coworker Girtler wrote an excellent and long-term project, studied the longitudinal development of gait
elaborate thesis on the temporal stride characteristics of lame from 4-month-old foal to adult horse (Back et al., 1994a), conclud-
horses at the walk and trot which, again, appeared in the Wiener ing that the gait pattern of the individual horse does not essentially
Tierärztliche Monatsschrift (Girtler, 1988). Also after the retirement alter during this development. He also made comparisons between
of Knezevic locomotion research remained a priority topic in hard data from kinematic gait analysis and the subjective evaluation
Vienna. The research facilities greatly improved when the Faculty of horses as is done at horseshows and during sire selection proce-
was moved to the vast new premises at the end of the 1990s. The dures, identifying kinematic parameters that determine the judg-
group was substantially reinforced by the arrival of Florian Buchner ment of quality of equine gait (Back et al., 1994b). The German
in 1996, who had spent 4 years with the Utrecht Equine Biome- veterinarian Florian Buchner used his time in Utrecht to study the
chanics Research group working on his PhD thesis. The growing kinematics of lame horses, and the symmetry of gait (Buchner et al.,
importance of the Austrian group was recognized and in 2000 the 1996). Besides, he performed some important work on determining
4th International Workshop on Animal Locomotion (IWAL-4) was the centers of gravity of various body segments (Buchner et al.,
organized in Vienna. 1997). Van den Bogert mathematically modeled equine gait (Van
The new facilities, which included a treadmill and a Motion den Bogert et al., 1989). This excellent scientist included a floppy
Analysis® kinematic gait analysis system, were not left unused. The disc in his thesis with an animation of equine gait, which was a
research effort was boosted by the arrival of the vet Theresia Licka, novelty at the end of the 1980s.
who had graduated in Vienna but had spent a couple of years after- Apart from the kinematic and kinetic studies, other lines of
wards in Scotland, and the engineer Christian Peham, who became research were followed as well. Some work on bone strain in the
the technical figure of the group. Here too, equine back motion was tibia was performed (Hartman et al., 1984; Schamhardt et al.,
studied extensively using a variety of techniques including electro- 1984). Riemersma started a line on tendon research using many
myography (Licka et al., 2001). The Vienna group focused on a techniques, including kinetic and kinematic analyses (Riemersma
number of practical applications of clinical research, such as studies et al., 1988a,b). This line was continued by Jansen who made use
into saddle pressure under various circumstances (Fruehwirth et al., of mercury in-silastic strain gauges (Jansen et al., 1993) and by
2004), the influence of other tack (Kicker et al., 2004), and the Becker who focused on the inferior check ligament (Becker et al.,
effect of skill of the rider on the way a horse moves (Licka et al., 1994).

19
 1 History

Fig 1.17  Investigations on the kinematics of the

equine spine at the Utrecht Biomechanics Research
Lab at the end of the 1990s by a combined Dutch
and Swedish team. A total of 7 Proreflex® cameras
are used.
Photograph courtesy of Utrecht Equine Biomechanics
Research Group.

In 1992, a specially constructed equine biomechanics research foals tend to continually protract the same limb when they spread
lab was opened in Utrecht, featuring a treadmill and a force plate their front limbs in order to reach the ground for grazing, which
track. This building greatly facilitated research. In the late 1990s, leads to the development of uneven feet with possible clinical con-
the CODA-3® system was replaced by a ProReflex® system (Fig. sequences in the long term (van Heel et al., 2006b). Maarten
1.17). By the early 1990s, the Utrecht Equine Biomechanics research Bobbert of the Department of Human Movement Sciences of the
group had become one of the world’s leading centers for equine Free University of Amsterdam, filled the vacuum caused by the
locomotion research. This had been possible in somewhat more death of Henk Schamhardt, providing invaluable scientific input.
than a decade because of the vision of the heads of department The extensive reconstruction of the entire Equine Clinic, which had
(after the retirement of Kersjes as head of the Department of General been concluded in 2008, enabled the incorporation of the formerly
and Large Animal Surgery, he had been succeeded by Ab Barneveld separately housed biomechanics lab in the clinic, opening possibili-
who strongly supported this line of research and, in fact, was the ties for a better integration of the lab with clinical work. A new
driving force behind many of the research projects), and the willing- player in the field in the Netherlands is Wageningen University,
ness to cooperate. This involved collaboration between depart- where the Department of Experimental Zoology has expanded into
ments, and between people with different scientific backgrounds. equine biomechanical research, focusing on the interaction of horse
Henk Schamhardt, a driving force behind a large number of the and rider (De Cocq et al., 2009).
research projects who unfortunately died as a consequence of a
tragic accident in 1999, was a physicist not a veterinarian; as is Ton
van den Bogert, who later left for Canada and the United States, Alfort
changing horses for humans. In France, the cradle of modern equine locomotion research, work
Just before the turn of the century, the Dutch group initiated a on the topic had been extremely limited after the turn of the
research project on the quantification of back kinematics that century. However, here too interest in this kind of research regained
would, to a large extent, be carried out together with the group from momentum at the end of the 1970s, which led to the re-emergence
Uppsala. An invasive experiment in which Steinmann pins were of France as one of the leading nations in equine locomotion
implanted in the spinous processes of various thoracic and lumbar research from the mid-1980s onwards. The new work originated in
vertebrae was performed that enabled the accurate quantification of part from the veterinary faculties of the French universities, but also
back motion (Faber et al., 1999, 2001, 2002), but also served as the from institutions such as the Institut National de Recherches
basis from which the skin marker-based, hence non-invasive, Agronomiques (INRA, National Institute for Agronomic Research)
Backin® program was developed. Once established and validated, and the national Stud Services (Haras Nationaux) as is the case
this technique was extensively used, among other things to study with the work by Langlois and co-workers (1978) relating confor-
the effects of induced lameness on back motion (Gómez Álvarez mation with ability in trotting, galloping and jumping. In these
et al., 2007), the effect of manual manipulation or chiropractic institutions research work of a more applied rather than fundamen-
treatment of the equine spine (Faber et al., 2003; Gómez Álvarez tal character has been carried out in various fields of equestrian
et al. 2008), and the influence of various head and neck positions sports, including jumping and harness racing, which is an impor-
on equine kinematics in the SDS collaborative project alluded to tant discipline in France.
earlier (Gómez Álvarez et al., 2006). Other topics of research Alfort took the lead among the veterinary schools. Eric Barrey
included a five-year study into the effect of early training on jumping studied the biomechanics of the equine foot extensively (Barrey,
capacity and the predictive value of foal jumping technique for later 1987), while Jean-Marie Denoix started his very active career by
athletic performance (Santamaría et al., 2005; Bobbert et al., 2005), studying contact areas in the fetlock joint (Denoix, 1987a) and the
saddle – related studies (De Cocq et al., 2004) and various studies kinematics of the thoracolumbar spine (Denoix, 1987b). Originally
on the effects of shoeing techniques and shoe types on equine an anatomist, Denoix’s interest in clinical orthopedic disorders
locomotion (van Heel et al., 2006a). An interesting observation in would later broaden considerably. His special interest in the bio-
the last project was that many of today’s long-legged Warmblood mechanics of the back, however, remained.

20
 Equine locomotion research centers and activities

A B

Fig 1.18  Trotter performing at high speed, equipped with force shoe and device for the non-invasive measurement of tendon force (A). The metal case on
the sulky contains equipment for data storage. (B): non-invasive tendon force measuring device.
Photographs courtesy of Unité Mixte de Recherche Biomécanique et Pathologie Locomotrice du Cheval.

The Alfort crew aligned with INRA to form the Unité Mixte de ground reaction force) with a Mustang® treadmill, thus avoiding the
Recherche Biomécanique et Pathologie Locomotrice du Cheval, great nuisance of conventional force plates that require many runs
which has been proven to be extremely productive and innovative to be made, especially at faster gaits, to get sufficient good hits
in the past decade, led by Nathalie Crevier-Denoix. Henry Chateau (Weishaupt et al., 2002, 2004). The instrument proved crucial in
made a very detailed and well-performed study into the three- the collaborative study that was performed by the SDS consortium
dimensional motion of the equine distal limb using an invasive in February 2005, which looked into the effect of various head and
technique and sound-based system to capture motion (Chateau neck positions on equine biomechanics (Fig. 1.19). In that study,
et al., 2004). Philippe Pourcelot, an engineer, was instrumental to hardware from all three research groups was combined, and mea-
the development of a device that could non-invasively quantify surements were performed using the instrumented treadmill, a
tendon force in the living and moving animal through the measure- 12-camera ProReflex® system and a Novel® saddle pressure mat,
ment of (changes in) speed of transmission of ultrasound signals simultaneously. This research project turned out to have been very
through the tissue (Pourcelot et al., 2005) (Fig. 1.18). The group timely, as shortly afterwards a fierce debate started in the equestrian
introduced more innovative technology such as a new lightweight lay press about the use of the so-called Rollkur (Meyer, 1992) or
force shoe and miniaturized accelerometers, both to be used at high hyperflexion (an extremely low and flexed position of head and
speed in the performance horse (Chateau et al., 2009, Robin et al., neck that used to be practiced by some (obviously successful) train-
2009). Another example of the innovative technology produced by ers and riders). The debate was highly emotional and the Zurich
the group from Alfort is the video-based analysis system for the project provided some of the very little hard data of the effects of
quantification of conformation that can easily be used at various head and neck position. This made the outcome of the project
horse events, such as stallion selection shows, as it requires only a (Gómez Álvarez et al., 2006; Weishaupt et al., 2006) play an impor-
single run of a horse over a defined track without any need for tant role in the eventual decision of the FEI (Fédération Equestre
markers (Pourcelot et al., 2002; Crevier-Denoix et al., 2004). In Internatonale) not to ban the practice after round table meetings
2008, the French group hosted the 6th International Conference on held in Lausanne in 2006 and 2010, although admittedly the evi-
Equine Locomotion (ICEL-6). The French group focuses on funda- dence was not conclusive, as the experiment concerned biome-
mental (technological) research, as well as on more applied inves- chanical effects only and did not look into possible behavioral
tigations (Vergari et al., 2012). consequences of the method.

Zurich United Kingdom

In Zurich, some work was done in the 1970s and 1980s using a In the UK, where research in equine biomechanics and gait analysis
(Swiss-built) Kistler force plate (Koch, 1973; Hugelshofer, 1982), had never been a strong area, a group interested in equine biome-
but it was not until the late 1990s that the vet school in Zurich chanics emerged around Lance Lanyon and Allen Goodship.
seriously took up locomotor research by the development of a very While Lanyon focused on biomechanics of bone (Lanyon, 1971;
special tool: a treadmill with inbuilt force plate. Through a very Lanyon & Rubin, 1980), Goodship dedicated himself and his group
ingenious construction, Mike Weishaupt and colleagues, among to the study of the equine tendon using force plate techniques
which technician Thomas Wiestner played an important role, suc- (Goodship et al., 1983). The latter technique was also used in the
ceeded in combining a force plate (that can measure the vertical extensive evaluation of various treatments of tendon injury as

21
 1 History

Other places in Europe

In the rest of Europe there were few developments. In Germany
some isolated investigations were performed (Wilsdorf, 1971; Bayer,
1973; Preuschoft & Fritz, 1977), but no new centers of research
emerged. In Spain, some kinematic work has been done on Anda-
lusian horses (Cano et al., 2001) and in Eastern Europe there are
other priorities for the time being. However, some of the states in
Eastern Europe have a strong tradition in horse breeding and other
horse-related activities and a start with some activities in the field
of equine locomotion has been made in Hungary and the Czech
Republic.

North America
In North America work on equine biomechanics and gait analysis
had been very limited after the great advances brought to this area
of science by Eadweard Muybridge. Publications from the period
that in Europe can be designated as the first Golden Age of equine
locomotion research, the period between World Wars I and II, are
virtually non-existent. After World War II this did not really change
with only a few isolated publications on rather heterogeneous
Fig 1.19  Horse and rider on the instrumented treadmill at the Zurich lab topics. However, the scene changed dramatically after the revolution
during the investigations into the effect of head and neck position on equine started by the Swedes. In just a few years a number of research
biomechanics. Both rider and horse are equipped with reflective markers; two groups were founded which would become very prolific.
ProReflex® cameras can be seen in the background. The small threads running
to the hooves of the horses serve to determine the exact position of each
limb, which is necessary input for the analysis of the force plate data. Saskatoon
Photograph courtesy of Equine Hospital, University of Zurich.
At the Western College of Veterinary Medicine in Saskatoon, Canada,
Doug Leach founded a school of equine biomechanical research
documented in the first Supplement of the Equine Veterinary that was to foster a number of excellent scientists. Leach himself
Journal, the famous Silver report (Silver et al., 1983). Also from published on a variety of topics including the effects of fatigue
Britain, R. McNeill Alexander should be mentioned as the author (Leach & Sprigings, 1979) and temporal stride characteristics (Leach
of several studies on quadrupedal gait. Like the American zoologist et al., 1987). He was also interested in the history of equine loco-
Milton Hildebrand, Alexander regarded the horse as an example motion research (Leach & Dagg, 1983) and was heavily involved in
rather than a goal in itself (Alexander, 1980; Alexander & Jayes, the definition of correct terminology (Leach et al., 1984). Another
1983). At the Royal Veterinary College, Alan Wilson followed in his researcher from Saskatoon was Hugh Townsend who dedicated his
foot steps. Whereas his earlier work concerned specific equine items thesis to thoracolumbar kinematics and the relationships between
(Wilson et al., 1998), he later focused more on general mechanisms vertebral morphology and pathologic changes, which led to some
in equine locomotion (Wilson et al., 2001) and developed a more important papers on this difficult subject (Townsend et al., 1983,
comparative approach (Wilson et al., 2003; Williams et al., 2007), 1986; Townsend & Leach, 1984).
a strategy that reduced clinical applicability, but certainly was better
appreciated by the scientific community at large. Thilo Pfau has
taken the lead in pursuing the clinical applications of gait analysis. Michigan
His work has focused on the use of inertial measurement units to Leach’s coworker and later chair of the Department of Veterinary
characterize movement asymmetries of the axial skeleton associated Anatomy in Saskatoon, Hilary Clayton, was to become an even
with lameness when horses move on the straight and on circles greater authority in the field. At first she focused on the kinematic
(Pfau et al., 2012). Whereas locomotion research in the UK does analysis of various causes of lameness using high-speed film
not have a strong historical basis, as stated before, the increase in (Clayton, 1986, 1987). Later her research developed a more funda-
this type of research in the past decade is remarkable, as is the mental character when it was directed towards the kinematic analy-
diversity of institutions involved. Equine locomotion research is sis of elite sport horses (Clayton, 1993, 1994, 1997). Collaboration
performed at vet schools, with the Royal Veterinary College and between Clayton and Bob Colborne, a human kinesiologist, and
Bristol (Burn 2006; Parsons et al., 2008; Pfau et al., 2007; Thorpe Joel Lanovaz, a mechanical engineer, broadened the scope of the
et al., 2009) as the best examples, but certainly not limited to them. work in Saskatoon to the calculation of net joint moments and joint
Work is done at engineering departments of universities not featur- powers at all joints of the equine limbs during both the stance and
ing a vet school, such as the University of Central Lancashire swing phases of the stride (Colborne et al., 1998; Clayton et al.,
(Hobbs et al., 2006) and Newcastle University (Lawson et al., 1998). In 1997, Clayton left Canada to become the first incumbent
2007), but is also commonplace in polytechnic type institutions of the Mary Anne McPhail Dressage Chair in Equine Sports Medi-
such as Hartpury College (Lovett et al., 2005) and Myerscough cine at Michigan State University in East Lansing (Fig. 1.20). The
College (Forsyth et al., 2006). large research endowment that founded this unique chair and the
construction of the new Mary Anne McPhail Equine Performance
Center made equine locomotion research at Michigan State Univer-
Belgium sity flourish over the past decade. The excellent facilities have
In Belgium two main centers have been developed: in Liege the enabled research into a great many areas of fundamental, clinical
exercise physiology group (Art & Lekeux, 1988) and in Ghent and sport-related research (Clayton & Sha, 2006; Clayton et al.,
(Oosterlinck et al., 2010) the focus was put on the use of pressure 2008; Geutjens et al., 2008), offering the possibility to many young
plate in the orthopedic examination of the horse. researchers to become acquainted with equine locomotion research.

22
 Equine locomotion research centers and activities

Fig 1.20  Kinematic research conducted at the

Equine Sports Medicine Center of Michigan State
University, East Lansing.
Photograph courtesy of Mary Anne McPhail, Dressage Chair
in Equine Sports Medicine.

In 2004 the Center hosted the 5th International Conference on et al., 1979). The gallop was the gait most intensively studied by
Equine Locomotion (succeeding the 4th International Workshop this group (Ratzlaff et al., 1995). The start of the work by George
on Animal Locomotion, the name change indicating that from then Pratt Jr. dates from about the same period. He used the force plate
on it was a single-species conference). A selection of the papers and concentrated on the Thoroughbred and the interaction with the
describing the research presented at the conference was published racetrack (Pratt & O’Connor, 1976; Pratt, 1984).
in a special issue of Equine Veterinary Journal in December 2004. Nancy Deuel, who spent part of her career at the University of
In 2009 the facilities were improved by the installation of an array Maryland in College Park, wrote her thesis on the kinematic analy-
of six force plates designed to provide data simultaneously from all sis of the gallop in Quarter Horses (Deuel, 1985). She later became
supporting limbs during locomotion on the straight and on circles interested in top-level performance horses and obtained permission
and to facilitate postural sway analysis with each hoof positioned to make recordings during the World Equestrian Games in 1990
on a separate force plate (Clayton & Nauwelaerts, 2012). and the Olympic Games in 1988, 1992 and 1996 (Deuel & Park,
1990, 1991). Calvin Kobluk performed equine kinematic research,
first using cinematography at Guelph University, then the video-
Missouri based Motion Analysis® system at the University of Minnesota
At Missouri State University, an endowed chair was installed for (Kobluk et al., 1989).
locomotor research related to lameness. It was Kevin Keegan who Apart from the fundamental research, more applied investiga-
took the lead there and produced many lameness-related studies tions were also being carried out. At Texas A&M University,
(Keegan, 2007; Keegan et al., 2008). His mathematical background Gingerich et al. (1979) used the force plate to study the effect of
drove him into new approaches, such as the fuzzy clustering tech- pharmaceuticals on joint function. In the same university, Swiss-
nique to describe the motion of horses suspected of ataxia (Keegan born Jörg Auer experimented with the so-called Kaegi-Straße, a
et al., 2004). This research led to the development of a commer- 5-m long track of rubber matting consisting of a large number of
cially marketed lameness detection system, the Lameness Locator™ tiny liquid-filled chambers, which in fact were electrical circuits.
(Equinosis® LLC, Columbia, Missouri, USA). The pressure of the hoof changed the electric resistance and thus
resulted in a change in the shape of the signal (Auer & Butler,
1985). This ‘diagnostic street’ was intended to aid in diagnosing
Other places in North America specific lameness causes. After the first enthusiastic reports, it died
There have been several places in the United States where work on a silent death as things turned out not to be as simple as that. At
equine locomotion has been or is performed, but thus far, apart Tufts University, Howard Seeherman constructed his Equine Per-
from the groups in Michigan and Missouri mentioned above, no formance Lab, which, among many other tests for equine perfor-
lasting research lines with a strong tradition have developed. mance, featured gait analysis techniques. Apart from the detection
As an early researcher in equine biomechanics Jim Rooney, a of gait irregularities, these were mainly used to correctly balance
pathologist in the Mecca of the American Thoroughbred industry, the hooves of the patients (Seeherman et al., 1987; Seeherman
Kentucky, should be mentioned. In 1969 he published a book on et al., 1996).
the Biomechanics of Lameness in horses. After that, a long series of At Cornell University, Kevin Haussler, a vet and qualified chiro-
papers on a wide variety of biomechanical subjects followed. practor, conducted research on the effects of chiropractic treatment
Rooney can to a certain extent be compared with the old scientists on spinal kinematics (Haussler et al., 2001, 2007), but later moved
of the pre-World War II generation. The vast majority of his papers to the orthopedic research center of Colorado State University in
are single-authored and it sometimes is hard to tell which of his Fort Collins, where the research climate was more welcoming for
statements are personal opinions based on deductions from obser- this type of investigations. For several years the team of Don Hoyt,
vations, and which are real facts based on hard, statistically sound, Steve Wickler, Ed Cogger and Darren Dutto had an active equine
scientific data. biomechanics and exercise physiology lab at the California State
From the mid 1970s, Marc Ratzlaff and his group at Washington Polytechnic University Pomona. This group performed a number of
State University in Pullman were one of the few who extensively fundamental studies on the relationships between gait and energet-
used electrogoniometry for gait analysis in the horse, in addition to ics and the effects of moving at different speeds and on an incline
other techniques such as cinematography (Ratzlaff, 1974; Ratzlaff or decline (Hoyt et al., 2006).

23
 1 History

Elsewhere in the world was the latter two aspects that ensured the horse’s survival when
mechanization relieved the horse of its former strongholds of mili-
Elsewhere in the world, research remained limited. In Australia, Leo tary and economic importance and, at the end of the 1940s, a total
Jeffcott began investigations into the non-invasive measurement of eclipse of the species threatened. Despite its retreat from many areas
bone quality and the influence of exercise thereon (Jeffcott et al., in society, the horse has remained within the public domain and is
1987), while Wilson et al. (1988a,b) did some work on the kine- now more popular than ever.
matics of trotters. However, the only long-lasting research line was The interest in the species from the veterinary perspective has kept
set up by Helen Davies at the University of Melbourne, where she pace with public appreciation of the horse. The same applies to
focused on bone strain in Thoroughbred racehorses, both under in equine locomotion research, though another factor should be men-
vivo and ex vivo conditions (Davies, 2006; Davies et al., 1993). In tioned too: the state of technology. The development of technology
New Zealand, Chris Rogers introduced kinematic research at Massey in general, and photography in particular, made possible the great
University. He did a lot of work in collaboration with the group in breakthroughs accomplished by Muybridge and Marey at the end
Utrecht (Rogers & Back, 2007) and aims, as is the case with the SDS of the 19th century, which were the prelude to the first Golden Age
consortium mentioned earlier, to bridge the gap between the scien- of equine locomotion research. It was also the combination of the
tist and the end-user in the equestrian sports (Rogers et al., 1999). renewed interest in the horse and the rapidly developing computer
In Japan, Tokuriki and coworkers made advances in electromyogra- technology, which led to the second Golden Age of this branch of
phy in the horse (Aoki et al., 1984), while Niki and colleagues used research.
the force plate for the study of equine biomechanics (Niki et al., What about the future? The bond between man and the horse
1982). has proved to be strong enough to survive the disappearance of
In the second decade of the 21st century, the second Golden Age what seemingly was the raison d’être of the alliance between the two
of equine biomechanical and locomotion research has certainly species. There is no reason to suppose that interest in the horse will
surpassed the first in researchers involved and in what had been diminish as long as economic conditions do not become too harsh.
accomplished. This type of research is now well established with As for technology, developments in this area seem to happen at ever
various centers where high-quality research is being performed in increasing speed. It can be anticipated that the increasing availabil-
North Western Europe and North America. Besides these major ity of high-performance micronized equipment that uses telemetry
research groups, smaller scale projects are undertaken in many other for data transfer, or is equipped with sophisticated lightweight data
places, the number of which is growing and starts to include places logging tools, will be increasingly used in real-life training and
where no real tradition existed. competition conditions. The day may even be not too far off that
equipment of this kind becomes a compulsory feature of competi-
tion in some kinds of equestrian activities (think of rein strain
Concluding remarks gauges for dressage or heart rate equipment for endurance horses).
Another branch of equine biomechanical research that is likely to
After its domestication 5000 years ago, the horse has had a close flourish in the next decade is advanced modeling of various parts
relationship with man. There is no doubt that, from a historical of the equine locomotor system with the aim to assess the effect of
viewpoint, the horse’s most important role has been that of a certain equestrian activities or environmental conditions (such as
machine of war, followed by its economic significance as a draft surface type) on lesion-prone structures of the horse. Research of
animal. However, from the very early days after domestication, the this type will be driven, as always, by scientific curiosity, but also
horse has had other roles also: as a sports and competition animal may become badly needed because of considerations of animal
and as an animal with which man formed a bond of affection. It welfare that become of increasing concern to the general public.

References
Adams, L., Handiseni, T., 1991. A tourist guide studies on supraspinatus and infraspinatus Back, W., Barneveld, A., Schamhardt, H.C.,
to rock art sites in Northern Zimbabwe, muscles of the horse with or without a Bruin, G., Hartman, W., 1994a.
Queen Victoria Museum, Harare. rider in walk, trot and canter. Bull. Equine Longitudinal development of the
Aepli, A., 1937. Betrachtungen über die Res. Inst. 21, 100–104. kinematics of 4-,10-,18- and 26-month-old
messung der oberen extremitätenwinkel Aristotle, 1961. (384–322 BC) Parts of animals; Dutch Warmblood horse. Equine Vet. J.
sowie neue meß- und movements of animals; progression of Suppl. 17, 3–6.
beobachtungsversuche durch die animals. In: Peck, A.l., Forster, E.S. (Eds.), Badoux, D.M., 1964. Friction between feet and
filmaufnahme beim pferd, The Loeb Classical Library, Heinemann, ground. Nature 202, 266–267.
inauguraldissertation, Zürich. London. Badoux, D.M., 1966. Mechanics of
Afanasieff, S., 1930. Die untersuchung des Art, T., Lekeux, P., 1988. Pulmonary mechanics the acropodium of the horse: a
exterieurs, der wachstumsintensität und der during treadmill exercise in race ponies. practical application of photoelastic
korrelation zwischen renngeschwindigkeit Vet. Res. Commun. 12, 245–258. research. Tijdschr. Dierg. 91 (19),
und exterieur beim traber. Z. Tierz. Zücht. Auer, J.A., Butler, K.D., 1985. An introduction 1207–1232.
Biol. 18, 171–209. to the Kaegi equine gait analysis system in Badoux, D.M., 1970a. Some biomechanical
Alexander McN., R., 1980. Optimum walking the horse. Proc. 31st Annu. Conv. Am. aspects of the elbow joint in the horse
techniques for quadrupeds and bipeds. J. Assoc. Equine Pract. 209–226. during normal gait. Proc. K. Ned. Akad.
Zool. Lond. 192, 97–117. Aysan, I., 1964. Beitrag zur analyse der Wetensch. 73 (1), 35–47.
Alexander McN., R., Jayes, A.S., 1983. A bewegungsanomalien beim pferd mit hilfe Badoux, D.M., 1970b. The statical function of
dynamic similarity hypothesis for the gaits der kinematographischen methode, some crural muscles in the horse. Acta
of quadrupedal mammals. J. Zool. Lond. inauguraldissertation, Gießen. Anat. 75, 396–407.
201, 135–152. Back, W., Barneveld, A., Bruin, G., Schamhardt, Bantoiu, C., 1922. Messungen an
André, M., 1949. Mécanique equestre, Thèse, H.C., Hartman, W., 1994b. Kinematic trabern und die beurteilung der
Lyon. detection of superior gait quality in young leistungsfähigkeit auf grund der
Aoki, O., Tokuriki, M., Kurakawa, Y., Hataya, trotting warmbloods. Vet. Quart. 166 (S2), mechanischen verhältnisse,
M., Kita, T., 1984. Electromyographic 91–96. inauguraldissertation, Berlin.

24
 References

Barrey, E., 1987. Biomécanique du pied du miembre antérieur. Rev. Vét. Milit. 19, phase of the trot. Equine Vet. J. 30 (5),
cheval: étude experimental, Thèse, École 563–570. 384–389.
Nationale Vétérinaire d’Alfort. Cavendysh, G., 1674. Méthode nouvelle et Clayton, H.M., Sha, D.H., 2006. Head and
Bayer, A., 1973. Bewegungsanalysen an invention extraordinaire de dresser les body centre of mass movement in horses
trabrennpferden mit hilfe der chevaux et les travailler selon la nature, trotting on a circular path. Equine Vet. J.
ungulographie. Zentbl. Vet. Med. Reihe A Tho. Milbourne, Londres, pp. 119–126. Suppl. 36, 462–467.
20, 209–221. Chateau, H., Degeuerce, C., Denoix, J.M., Clayton, H.M., White, A.D., Kaiser, L.J.,
Bayer, J., 1882. Experimentelles über 2004. Evaluation of three-dimensional Nauwelaerts, S., Lavignino, M., Stubbs,
hufmechanismus. Oesterr. Monatsschr. kinematics of the distal portion of the N.C., 2008. Short-term habituation of
Thierheilk. 7, 72–74. forelimb in horses walking in a straight equine limb kinematics to tactile
Becker, C.K., Savelberg, H.H.C.M., Barneveld, line. Am. J. Vet. Res. 65, 447–455. stimulation of the coronet. Vet. Comp.
A., 1994. In vitro mechanical properties of Chateau, H., Robin, D., Veniard, M., Falala, S., Orthop. Traumatol. 21, 211–214.
the accessory ligament of the deep digital Poucelot, P., Valette, J.-P., et al., 2009. Clayton, M., 1996. Leonardo da Vinci – a
flexor tendon in horses in relation to age. Effects of a synthetic all-weather waxed curious vision, Merrell Holberton, London,
Equine Vet. J. 26, 454–459. track versus a crushed sand track on 3D pp. 158–159.
Berger, A.K., 1923. Mounting horse skeletons acceleration of the foot in horses trotting at Clayton, H.M., Nauwelaerts, S., 2012. Is a
to exemplify different gaits and actions. A high speed. Equine Vet. J. 41, 247–251. single force plate adequate for
glimpse behind the scenes at the American Cheney, J.A., Henwood, K., Chen, C.K., 1970. stabilographic analysis in the horse? Equine
Museum, Natural History Magazine (Nov/ Thoroughbred lameness, a report on Vet. J. 44, 550–553.
Dec). (Reprinted in: Natural History 89 research into its relationship to the Colborne, G.R., Lanovaz, J.L., Sprigings, E.J.,
(4), 100–102, 1980.) composition of California racing Schamhardt, H.C., Clayton, H.M., 1998.
Bethcke, H., 1930. Ist es möglich auf Grund surfaces. The Thoroughbred of Calif. June, Forelimb joint moments and power during
der mechanischen verhältnisse die 906–924. the walking stance phase of horses. Am. J.
leistungsfähigkeit eines trabers zu Chieffi, A., 1943. A marcha no cavalo Vet. Res. 59 (5), 609–614.
bestimmen? Z. Für Veterinärkunde 42 (5), mangalarga. Rev. Fac. Med. Vet. S. Paulo 2, Crevier-Denoix, N., Erlinger, D., Concordet,
161–170. 177–192. D., Tavernier, L., Lagache, C., Ricard, A., et
Björck, G., 1958. Studies on the draught forces Chieffi, A., 1945. Estudo comparativo de fases al., 2004. Comparison of the conformation
of horses. Development of a method using de apôio semelhantes, nos diversos of 20 international and 20 low level
strain gauges for measuring forces between andamentos do cavalo. Rev. Fac. Med. Vet. jumping horses using a 3-D video
hoof and ground. Acta Agricult. Scand. 8, S. Paulo 3, 109–130. morphometric measurement method. Proc.
(Suppl. 4). Chieffi, A., 1946. Contribuição para o estudo Eur. Assoc. Vet. Anat. p75.
Bobbert, M.F., Santamaría, S., van Weeren, da mudança do galope do cavalo pela Davies, H.M., 2006. Estimating peak strains
P.R., Back, W., Barneveld, A., 2005. Can cinematografia. Rev. Fac. Med. Vet. S. Paulo associated with fast exercise in
jumping of adult show jumping horses be 3 (3), 61–69. thoroughbred racehorses. Equine Vet. J.
predicted on the basis of submaximal free Chieffi, A., 1949. Contribuição para o estudo Suppl. 36, 383–386.
jumps at foal age? A longitudinal study. da transição dos andamantos, no cavalo, Davies, H.M., McCarthy, R.N., Jeffcott, L.B.,
Vet. J. 170, 212–221. pela interpretação de filmes, em camara 1993. Surface strain on the dorsal
Borcherdt, W., 1912. Studien über die lenta, Thesis, University of Sao Paulo, 74 metacarpus of thoroughbreds at different
sprungbewegung des pferdes, pp.. speeds and gaits. Acta Anat. 146,
inauguraldissertation, Bern. Chieffi, A., de Mello, L.H., 1939. Contribuição 148–153.
Borelli, G.A., 1681. De motu animalium, para o estudo da localização do centro de Dawes, H.W., 1957. The relationship of
Roma. gravidade no corpo dos animais soundness and conformation in the horse.
Brayley Reynolds, E., 1957. In: Dawes, H.W. domésticos e dos fatores que produzem seu Vet. Rec. 69, 1367–1374.
(Eds.), The relationship of soundness and deslocamento temporário ou permanente. De Cocq, P., Prinsen, H., Springer, N.C.N., van
conformation in the horse. Vet. Rec. 69, Rev. Fac. Med. Vet. S. Paulo 1, 97–154. Weeren, P.R., Schreuder, M., Muller, M., et
1367–1374. Clayton, H.M., 1986. Cinematographic al., 2009. The effect of rising and sitting
Buchmann, J., 1929. Untersuchungen über analysis of the gait of lame horses. I: trot on back movements and head-neck
schritt- und trabgang des pferdes, fractured supraglenoid tubercle. J. Equine position of the horse. Equine Vet. J. 41,
dissertation, Breslau. Vet. Sci. 6 (2), 70–78. 423–427.
Buchner, H.H.F., Savelberg, H.H.C.M., Clayton, H.M., 1987. Cinematographic De Cocq, P., van Weeren, P.R., Back, W., 2004.
Schamhardt, H.C., Barneveld, A., 1997. analysis of the gait of lame horse. IV: The effect of a girth a saddle and weight on
Inertial properties of Dutch Warmblood fracture of the third carpal bone. J. Equine the movements of the horse. Equine Vet. J.
horses. J. Biomech. 30, 653–658. Vet. Sci. 7 (3), 130–135. 36, 758–763.
Buchner, H.H.F., Schamhardt, H.C., Barneveld, Clayton, H.M., 1993. The extended canter: a De Sévy, L., 1918. Saut d’obstacles et galop de
A., 1996. Limb movement adaptation in comparison of some kinematic variables in course, Librairie Le Goupy et Librairie
horses with experimentally induced fore- or horses trained for dressage and racing. Acta Chapelot, Paris.
hind limb lameness. Equine Vet. J. 28, Anat. 146, 183–187. De Solleysel, J., 1733. Le parfait maréschal,
63–70. Clayton, H.M., 1994. Comparison of the stride que enseigne a connoistre la beauté, la
Burn, J.F., 2006. Time domain characteristics kinematics of the collected, working, bonté et les défauts des chevaux, 2ième
of hoof-ground interaction at the onset of medium and extended trot in horses. partie, p66. Pierre-Jean Mariette, Paris.
stance phase. Equine Vet. J. 38, 657–663. Equine Vet. J. 26 (3), 230–234. Denoix, J.M., 1987a. Étude biomécanique de
Cano, M.R., Vivo, J., Miró, F., Morales, J.L., Clayton, H.M., 1997. Classification of la main du cheval: extensometrie des
Galisteo, A.M., 2001. Kinematic collected trot, passage and piaffe based on rayons metacarpo-phalangiens et surfaces
characteristics of Andalusian, Arabian and temporal variables. Equine Vet. J. Suppl. articulaires de contact (sur membre isolé
Anglo-Arabian horses: a comparative study. 23, 54–57. soumis a compression), Thèse, Université
Res. Vet. Sci. 71, 147–153. Clayton, H.M., Lanovaz, J.L., Schamhardt, Claude-Bernard, Lyon.
Carnus, O.A.C., 1935. Contribution à l’étude H.C., Willemen, M.A., Colborne, G.R., Denoix, J.M., 1987b. Kinematics of the
de la cinématique du cheval. L’action des 1998. Net joint moments and joint powers thoracolumbar spine of the horse during
muscles de l’encolure dans le jeu du in the equine forelimb during the stance dorsoventral movements: a preliminary

25
 1 History

report. Proc. 2nd Int. Conf. Equine Exerc. Fehér, G., 1958. Beiträge zur Statik und function in equine arthritis. J. Vet.
Physiol. 607–614. Dynamik der vorderen extremitäten des Pharmacol. Therap. 2, 291–298.
Deuel, N.R., 1985. A kinematic analysis of the pferdes. Acta Vet. Acad. Sci. Hung. 8, Girtler, D., 1988. Untersuchungen über die
gallop of the horse, Thesis, University of 187–198, 1958. dauer des bewegungszyklus –
Illinois. Fick, R., 1910. Allgemeine Gelenk- und stützbeinphase, hangbeinphase,
Deuel, N.R., Park, J., 1991. Kinematic analysis Muskelmechanik. In: von Bardeleben, K. phasenverschiebung – bei lahmen und
of jumping sequences of Olympic show (Ed.), Handbuch der anatomie des bewegungsgestörten pferden im schritt und
jumping horses. Proc. 3rd Conf. Equine menschen. Bd. 2 Abt. 1, Fischer, Jena. trab sowie kinematische beurteilungen zu
Exerc. Physiol. 158–166. Forssell, G., 1915. Untersuchungen über die deren bewegungsmuster. Wien. Tierärztl.
Deuel, N.R., Park, J.J., 1990. The gait patterns Wirkungsweise der Beugesehnen am Mon. Schr. 75 (5), 185–196; 75 (6),
of Olympic dressage horses. Int. J. Sport Vorderfuß des Pferdes. Z. Tiermed. 18, 217–231; 75 (7), 255–270; 75 (8), 310–324.
Biomech. 6, 198–226. 184–188. Goiffon, G.C., Vincent, A.F., 1779. Mémoire
Dörrer, H., 1911. Über die Anspannung der Forsyth, R.K., Brigden, C.V., Northrop, A.J., artificielle des principes relatifs à la fidelle
Beugesehnen des pferdefußes während der 2006. Double blind investigation of the répresentation des animaux tant en
verschiedenen bewegungsstadien derselben, effects of oral supplementation of peinture, qu’en sculpture. I. Partie
inauguraldissertation, Dresden. combined glucosamine hydrochloride concernant le cheval, Ecole Royale
(GHCl) and chondroitin sulphate (CS) on Vétérinaire, Alfort.
Drevemo, S., Dalín, G., Fredricson, I., Björne,
K., 1980c. Equine locomotion 3. The stride characteristics of veteran horses. Gómez Álvarez, C.B., l’Ami, J.J., Moffat, D.,
reproducibility of gait in Standardbred Equine Vet. J. Suppl. 36, 622–625. Back, W., van Weeren, P.R., 2008. Effect of
trotters. Equine Vet. J. 12 (2), 71–73. Franke, H., 1935. Untersuchungen über den chiropractic manipulations on the
einfluß des körperbaus auf die schrittlänge kinematics of back and limbs in horses
Drevemo, S., Dalín, G., Fredricson, I., Hjertén,
des pferdes, Dissertation with clinically diagnosed back problems.
G., 1980a. Equine locomotion 1. The
Landwirtschaftliche Hochschule, Berlin. Equine Vet. J. 40, 153–159.
analysis of linear and temporal stride
characteristics of trotting Standardbreds. Fredricson, I., 1972. Equine joint kinematics Gómez Álvarez, C.B., Rhodin, M., Bobbert,
Equine Vet. J. 12 (2), 60–65. and co-ordination. Thesis, Royal Veterinary M.F., Meyer, H., Weishaupt, M.A., Johnston,
College, Stockholm. Published in: Acta Vet. C., et al., 2006. The effect of different head
Drevemo, S., Fredricson, I., Dalín, G., Björne,
Scand. Suppl. 37, 1–136. and neck positions on the thoracolumbar
K., 1980b. Equine locomotion 2. The
Fredricson, I., Andersson, S., Dandanell, R., kinematics in the unridden horse. Equine
analysis of coordination between limbs of
Moen, K., Andersson, B., 1970. Vet. J. Suppl. 36, 445–451.
trotting Standardbreds. Equine Vet. J. 12
(2), 66–70. Quantitative analysis of hoof motion Gómez Álvarez, C.B., Wennerstand, J.,
patterns using high-speed films of harness Bobbert, M.F., Lamers, L., Johnston, C.,
Drevemo, S., Roepstorff, L., Kallings, P.,
horses. Proc. 9th Int. Congr. On High-Speed Back, W., et al., 2007. The effect of induced
Johnston, C., 1993. Application of TrackEye
Photogr. 347–350. forelimb lameness on thoracolumbar
in equine locomotion research. Acta Anat.
Fredricson, I., Dalín, G., Drevemo, S., Hjertén, kinematics during treadmill locomotion.
146, 137–140.
G., 1975b. A biotechnical approach to the Equine Vet. J. 39, 197–201.
Dunlop, R.H., Williams, D.J. (Eds.), 1996.
geometric design of racetracks. Equine Vet. Goodman, N., 1870. The action of the horse.
Veterinary Medicine. An illustrated history,
J. 7 (2), 91–96. J. Anat. Phys. 4, 8–11.
Mosby, St. Louis.
Fredricson, I., Dalín, G., Hjertén, G., Nilsson, Goodman, N., 1871. The action of the horse. J.
Dušek, J., Dušek, J., 1963. Vyznam telesne
G., Alm, L.O., 1975a. Ergonomic aspects of Anat. Phys. 5, 89–91.
stavby koni z hlediska pracovniho vykonu
v zemedelstvi. Sb. Zivoc. Výroba 8, poor racetrack design. Equine Vet. J. 7 (2), Goodship, A.E., Brown, P.N., MacFie, H.J.H.,
303–310. 63–65. Lanyon, L.E., Silver, I.A., 1983. A
Fredricson, I., Drevemo, S., 1971. A new quantitative force plate assessment of
Dušek, J., Ehrlein, H.J., von Engelhardt, W.,
method of investigating equine equine locomotor performance. Proc. 1st
Hörnicke, H., 1970. Beziehungen zwischen
locomotion. Equine Vet. J. 3 (4), Int. Conf. Equine Exerc. Physiol. 263–270.
trittlänge, trittfrequenz und
137–140. Goubaux, A., Barrier, G., 1884. De l’extérieur
Geschwindigkeit bei Pferden. Z. Tierärztl.
Zücht. Biol. 87, 177–188. Fredricson, I., Drevemo, S., Dalín, G., Hjertén, du cheval, Asselin et Cie, Paris, pp.
G., Björne, K., Rynde, R., et al., 1983. 547–582.
Faber, M., Johnston, C., van Weeren, R.,
Treadmill for equine locomotion analysis. Gray, J., 1944. Studies in the mechanics of the
Barneveld, A., 2002. Repeatability of back
Equine Vet. J. 15 (2), 111–115. tetrapod skeleton. J. Exp. Biol. 20, 88–116.
kinematics of horses during treadmill
locomotion. Equine Vet. J. 34, 235–241. Fruehwirth, B., Peham, C., Scheidl, M., Gregory, W.K., 1912. Notes on the principles
Schobesberger, H., 2004. Evaluation of of quadrupedal locomotion and on the
Faber, M., Schamhardt, H., van Weeren, R.,
pressure distribution under an English mechanism of the limbs in hoofed
1999. Determination of 3D spinal
saddle at walk, trot and canter. Equine Vet. animals. Annals N.Y. Acad. Science 22,
kinematics without defining a local
J. 36, 754–757. 267–294.
vertebral co-ordinate system. J. Biomech.
32, 1355–1358. Gamgee, J., 1869. On the action of the horse. Grogan, J.W., 1951. The gaits of horses. J. Am.
J. Anat. Physiol. 3, 370–376. Vet. Med. Assoc. 119, 112–117.
Faber, M.J., Schamhardt, H.C., van Weeren,
P.R., Barneveld, A., 2001. Methodology and Gamgee, J., 1870. The action of the horse. J. Grzimek, B., 1949. Rechts- und linkshändigkeit
validity of assessing kinematics of the Anat. Phys. 4, 235–236. bei pferden, papagaien und affen. Z.
thoracolumbar vertebral column in horses Genieser, D., 1962. Analyse der Tierpsychol. 6, 406–432.
on the basis of skin-fixated markers. Am. J. bewegungswechsel beim shetlandpony, Günther, K., 1866. Die topographische
Vet. Res. 62, 301–306. dissertation, Gießen. myologie des pferdes. mit besonderer
Faber, M.J., van Weeren, P.R., Schepers, M., Geutjens, C.A., Clayton, H.M., Kaiser, L.J., berücksichtigung der lokomotorischen
Barneveld, A., 2003. Long-term follow-up 2008. Forces and pressures beneath the wirkung der muskeln, Carl Rümpler,
of manipulative treatment in a horse with saddle during mounting from the ground Hannover.
back problems. J. Vet. Med. A 50, 241–245. and from a raised mounting platform. Vet. Halvorsen, K., Johnston, C., Back, W., Stokes,
Fehér, G., 1957. Beiträge zu den funktionellen J. 175, 332–337. V., Lanshammar, H., 2008. Tracking the
Veränderungen der form und struktur des Gingerich, D.A., Auer, J.A., Fackelman, G.E., motion of hidden segments using
schulterblatts beim pferd. Acta Vet. Acad. 1979. Force plate studies on the effect of kinematic constraints and Kalman filtering.
Sci. Hung. 7, 19–48. exogenous hyaluronic acid on joint J. Biomech. Eng. 130, 011012.

26
 References

Hartman, W., Schamhardt, H.C., Lammertink, Jeffcott, L.B., Dalín, G., Drevemo, S., Knezevic, P.F., Floss, F.N., 1984. Klinische
J.L.M.A., Badoux, D.M., 1984. Bone strain Fredricson, I., Björne, K., Bergqvist, A., ergebnisse der rechnergestützten
in the equine tibia: An in vivo strain gauge 1982. Effect of induced back pain on gait bewegungsanalyse beim pferd. Arch.
analysis. Am. J. Vet. Res. 45 (5), 880–884. and performance of trotting horses. Equine Tierärztl. Fortbild. 8, 140–148.
Haussler, K.K., Bertram, J.E.A., Gellman, K., Vet. J. 14 (2), 129–133. Knezevic, P.F., Seebacher, M., Fischerleitner, F.,
Hermanson, J.W., 2001. Segmental in vivo Jeffcott, L.B., Dalín, G., Ekman, S., Olsson, Tauffkirchen, W., Benedikter, G., 1978.
vertebral kinematics at the walk, trot and S.E., 1985. Sacroiliac lesions as a cause of Ermittlung des Zusammenhanges zwischen
canter: a preliminary study. Equine Vet. J. chronic poor performance in competitive Bewegung und Belastung des Pferdehufes
Suppl. 33, 160–164. horses. Equine Vet. J. 17 (2), 111–118. mit Hilfe von Hochfrequenzphotographie
Haussler, K.K., Hill, A.E., Puttlitz, C.M., Johnston, C., Back, W., 2006. Hoof ground und Mehrkomponentenkraftmeßplatte.
McIlwraith, C.W., 2007. Effects of vertebral interaction: when biomechanical stimuli Biomed. Tech. 23, 154–155.
mobilization and manipulation on challenge the tissues of the distal limb. Kobluk, C.N., Schnurr, D., Horney, F.D.,
kinematics of the thoracolumbar region. Equine Vet. J. 38, 634–641. Sumner-Smith, G., Willoughby, R.A.,
Am. J. Vet. Res. 68, 508–516. Johnston, C., Holm, K.R., Erichsen, C., Eksell, DeKleer, V., 1989. Use of high-speed
Hernlund, E., Egenvall, A., Roepstorff, L., 2010. P., Drevemo, S., 2004. Kinematic cinematography and computer generated
Kinematic characteristics of hoof landing in evaluation of the back in fully functioning gait diagrams for the study of equine hind
jumping horses at elite level. Equine Vet. J. riding horses. Equine Vet. J. 36, 495–498. limb kinematics. Equine Vet. J. 21, 48–58.
Suppl. 38, 462–467. Kadletz, M., 1926. Die formenwechsel der Koch, A., 1973. Zeit- und
Hildebrand, M., 1959. Motions of the running hinterhandmuskulatur des pferdes während belastungsmessungen an den vorderhufen
cheetah and horse. J. Mammal. 40 (4), der bewegung. Wiener Tierärztl. des pferdes mittels der mehrkomponenten-
481–495. Monatsschr. 13, 185–198. kraftmeßplatte ‘Kistler’, dissertation, Zürich.
Hildebrand, M., 1960. How animals run. Kolda, J., 1937. Srovnávací anatomie a
Kadletz, M., 1932. Ueber die physiologischen
Scient. Am. 202 (5), 148–157. mechanica pažního kloubo.
kreiselungen und über die statik der
Zverolek Rozpr. 11, 265–276.
Hildebrand, M., 1965. Symmetrical gaits of schultergliedmaße des pferdes; gleichzeitig
ein erklärungsversuch der sog. struppigkeit Kronacher, C., Ogrizek, A., 1931. Exterieur und
horses. Science 150, 701–708.
und des sog. Fuchtelns. Münchn. Tierärztl. leistungsfähigkeit des pferdes mit
Hildebrand, M., 1966. Analysis of the besonderen berücksichtigung der
symmetrical gaits of tetrapods. Folia Wschr. 83 (37), 433–436.
gliedmaßenwinkelung und
Biotheor. VI, 9–22. Kadletz, M., 1933. Ueber einige behelfe zur
schrittlängenverhältnisse. Z. Zücht. Reihe B
Hobbs, S.J., Richards, J., Matuszewski, B., erkennung der bewegungsarten des pferdes.
Tierz. Züchtungsbiol. 32 (2), 183–228.
Brigden, C., 2006. Development and Dtsch. Tierärztl. Wschr. 41 (28), 433–438.
Kroon, H.M., 1922. Een bijdrage tot de studie
evaluation of a noninvasive marker cluster Kadletz, M., 1937. Über die Bewegungsweise
van het hoefmechanisme. Tijdschr. Dierg.
technique to assess three-dimensional des tarsalgelenkes des pferdes. 3. beitrag
49 (11), 399–406.
kinematics of the distal portion of the zur spatgenese. Arch. Wissensch. Prakt.
Kroon, H.M., 1929. Studie van het
forelimb in horses. Am. J. Vet. Res. 67, Tierheilk. 71 (4), 279–296.
hoefmechanisme. Tijdschr. Dierg. 55 (19),
1511–1518. Kaemmerer, K., 1960. Deskriptives über einige 961–967.
Holmström, M., Fredricson, I., Drevemo, S., bahnkurven bei der gliedmaßenbewegung
Kroon, H.M., Paimans, W.J., Ihle, J.E.W., 1921.
1995. Biokinematic effects of collection on des pferdes. Arch. Exp. Vet. Med. 14, 97–110.
Een eeuw veeartsenijkundig onderwijs. ‘S
the trotting gaits in the elite dressage horse. Keegan, K.G., 2007. Evidence-based lameness rijksveeartsenijschool – veeartsenijkundige
Equine Vet. J. 27 (4), 281–287. detection and quantification. Vet. Clin. hoogeschool 1821–1921, Senaat der
Hoyt, D.F., Wickler, S.J., Dutto. D.J., Catterfeld, North Am. Equine Pract. 23, 403–423. Veeartsenijkundige Hoogeschool, Utrecht.
G.E., Johnsen, D., 2006. What are the Keegan, K.G., Arafat, S., Skubic, M., Wilson, Krüger, L., 1957. Die bestimmung der
relations between mechanics, gait D.A., Kramer, J., Messer, N., et al., 2004. arbeitsfähigkeit bei pferd und rind durch
parameters, and energetics in terrestrial Detection of spinal ataxia in horses using leistungsprüfungen, physiologische und
locomotion? J. Exp. Zool. A Comp. Exp. fuzzy clustering of body position psychologische meßwerte und durch die
Biol. 305, 912–922. uncertainty. Equine Vet. J. 36, 712–717. exterieurbeurteilung. Z. Tierz. Zücht. Biol.
Howell, A.B., 1944. Speed in animals, Hafner Keegan, K.G., Messer, N.T., Reed, S.K., Wilson, 69, 289–320.
Publishing Co, New York. D.A., Kramer, J., 2008. Effectiveness of Krüger, W., 1937. Ueber den bewegungsablauf
Hugelshofer, J., 1982. Vergleichende kraft- und administration of phenylbutazone alone or an dem oberen teil der vordergliedmaße
belastungszeitmessungen an den concurrent administration of des pferdes im schritt, trab und galopp.
vorderhufen von gesunden und an phenylbutazone and flunixin meglumine to Tierärztl. Rundsch. 43 (49/50), 809–816;
podotrochleose erkrankten pferden, alleviate lameness in horses. Am. J. Vet. 825–827.
dissertation, Zürich. Res. 69, 167–173. Krüger, W., 1938. Ueber den bewegungsablauf
Ivanov, S., Borissov, V.M., 1935. Importance of Keller, K., 1917. Kinematographische analyse an dem oberen teil der hintergliedmaße
the lacertus fibrosus in standing and der bewegungen des pferdes. Verhandl. des pferdes im schritt, trab und galopp.
motion of the horse. Archiv Anatomii, Zool. Botan. Ges. Wien 67, 93–94. Tierärztl. Rundsch. 44 (34), 549–557.
Gistologii I Embriologii 14, 51–55. Kicker, C.J., Peham, C., Girtler, D., Licka, T., Krüger, W., 1939a. Die fortbewegung des
Jansen, M.O., van den Bogert, A.J., Riemersma, 2004. Influence of support boots on fetlock pferdes, Paul Parey, Berlin.
D.J., Schamhardt, H.C., 1993. In vivo joint angle of the forelimb of the horse at Krüger, W., 1939b. Ueber die schwingungen
tendon forces in the forelimb of ponies at walk and trot. Equine Vet. J. 36, 769–771. der wirbelsäule -insbesondere der
the walk, validated by ground reaction force Knezevic, P., 1962. Klinik des wirbelbrücke- des pferdes während der
measurements. Acta Anat. 146, 162–167. trachtenzwanghufes und Grundlagen der bewegung. Berl. Münchn. Tierärztl. Wschr.
Jeffcott, L.B., Buckinhgham, S.H.W., ungulographie mit dehnungsmeßstreifen 13, 197–203.
McCartney, R.N., 1987. Noninvasive beim pferde. Wien. Tierärztl. Mon. Schr. 49 Krüger, W., 1939c. Ueber die arbeit der vier
measurement of bone quality in horses and (10), 777–824. gliedmaßen des pferdes beim galopp.
changes associated with exercise. Proc. 2nd Knezevic, P.F., 1985. Zur geschichte der Tierärztl. Rundsch. 45 (13), 250–265.
Int. Conf. Equine Exerc. Physiol. 615–630. bewegungslehre des pferdes – eine Krüger, W., 1939d. Bemerkungen zur
Jeffcott, L.B., Dalín, G., 1980. Natural rigidity historische betrachtung der gliedmaßenfolge beim renngalopp und
of the horse’s backbone. Equine Vet. J. 12, untersuchungstechniken. Wien. Tierärztl. beim sprung des pferdes. Tierärztl.
101–108. Mon. Schr. 72 (12), 399–405. Rundsch. 45 (15), 287–293.

27
 1 History

Krüger, W., 1941a. Ueber das verhalten des Lenoble du Teil, J., 1893. Les allures de cheval gedämpfte, erzwungene schwingung.
schwerpunktes bei der normalen dévoilées par la méthode expérimentelle, Zentbl. Vet. Med. Reihe A 16, 180–184.
fortbewegung des pferdes. Tierärztl. Berger-Levrault et Cie, Paris. Moskovits, S., 1930. Die statisch-mechanische
Rundsch. 47 (12/13), 147–152; Lhote, H., 1988. La route des chars. In: Vaes, beurteilung der arbeitstiere durchgeführt
162–166. B., del Marmol, G., d’Otreppe, A., (Eds.), am metacarpus der pferde. Wissensch.
Krüger, W., 1941b. Die schwere zugarbeit und Guide du Sahara., Hachette, Paris, pp. Arch. Landwirtsch. Abt. B Tiernähr. Tierz.
ihre auswirkungen auf die sehnen der 464–466. 2 (3), 372–421.
vordergliedmaße des pferdes. Dtsch. Licka, T., Kapaun, M., Peham, C., 2004. Muybridge, E., 1957. Animals in motion.
Tierärztl. Wschr. 49 (17), 203–208. Influence of the rider on lameness in the Original edition 1899. Republished Brown,
Langlois, P., Froidevaux, J., Lamarche, L., trotting horse. Equine Vet. J. 36, 734–736. L.S., (Ed.). Dover Publications, New York.
Legault, C., Legault, P., Tassencourt, L., et Licka, T., Peham, C., Zohmann, E., 2001. Nicolescu, J., 1923. Messungen über die
al., 1978. Analyse des liaisons entre la Treadmill study of the range of back mechanik des hannoverschen pferdes in
morphologie et l’aptitude au galop, au trot movement at the walk in horses without vergleich zum vollblut und traber,
et au saut d’obstacles chez le cheval. Ann. back pain. Am. J. Vet. Res. 62, 1173–1179. dissertation, Berlin.
Génét. Sél. Anim. 10, 443–474. Lovett, T., Hodson-Tale, E., Nankervis, K., Niki, Y., Ueda, Y., Yoshida, K., Masumitsu, H.,
Lanyon, L.E., 1971. Use of an accelerometer to 2005. A preliminary investigation of rider 1982. A force plate study in equine
determine support and swing phases of a position during walk, trot and canter. biomechanics. 2. The vertical and fore-aft
limb during locomotion. Am. J. Vet. Res. Equine Comp. Exerc. Physiol. 2, 71–76. components of floor reaction forces and
32 (7), 1099–1101. Maennicke, E.M., 1961. Untersuchungen über motion of equine limbs at walk and trot.
Lanyon, L.E., Rubin, C.T., 1980. Loading of die Bewegungsvorgänge des Shetlandponys, Bull. Equine Res. Inst. 19, 1–17.
mammalian long bones during Dissertation, Gießen. Nomura, S., 1953a. Mechanical studies on the
locomotion. J. Physiol. 303, 72P. Magne de la Croix, P., 1928. Sobre la evolución elbow joint of the horse. Jap. J. Vet. Sci. 15,
Laughlin, H.H., 1934a. Racing capacity in the del galope de carrera y la consecutiva de 21–36.
Thoroughbred horse. Part 1. The measure la forma. Causa de la evolución en Nomura, S., 1953b. On the so-called
of racing capacity. The Sci. Monthly 38, perisodáctiles y artiodadáctiles. Anal. Soc. ‘Fäderungsphänomen’ of elbow joint of
210–222. Cient. Argent. 106 (6), 317–331. horse. Jap. J. Vet. Sci. 15, 175–182.
Laughlin, H.H., 1934b. Racing capacity in the Magne de la Croix, P., 1929. Filogenia de las Nomura, S., 1953c. A mechanical study on the
Thoroughbred horse. Part 2. The locomociones cuadripedal y bipedal en los tarsal joint of the horse. Jap. J. Vet. Sci. 15,
inheritance of racing capacity. The Sci. vertebrados y evolución de la forma 281–293.
Monthly 38, 310–321. consecutiva de la evolución de la Offringa, C., 1981. Van gildestein naar uithof.
Lawson, S.E., Chateau, H., Pourcelot, P., locomoción. Anal. Soc. Cient. Argent. 108, dl. 2, Faculteit der Diergeneeskunde,
Denoix, J.M., Crevier-Denoix, N., 2007. 383–406. Utrecht.
Sensitivity of an equine distal limb model Magne de la Croix, P., 1932. Evolución del Oosterlinck, M., Pille, F., Huppes, T., Gasthuys,
to perturbations in tendon paths, origins galope transverso. Anal. Soc. Cient. Argent. F., Back, W., 2010. Comparison of pressure
and insertions. J. Biomech. 40, 2510–2516. 113, 38–41. plate and force plate gait kinetics in sound
Le Hello, P., 1896. Du rôle des membres Magne de la Croix, P., 1936. The evolution of Warmbloods at walk and trot. Vet. J. 186,
postérieurs dans la locomotion du cheval. locomotion in mammals. J. Mamm. 17, 347–351.
Compt. Rend. Hebd. Séances Acad. Sci. 51–54. Pagès, C., 1885. Analyse cinématique de la
122, 1356–1360. Manter, J.T., 1938. The dynamics of locomotion du cheval. Compt. Rend.
Le Hello, P., 1897. Sur l’action locomotrice des quadrupedal walking. J. Exp. Biol. 15, Hebd. Séances Acad. Sci. 101, 702–705.
membres antérieurs du cheval. Compt. 522–540. Pagès, C., 1889. De la marche chez les
Rend. Hebd. Séances Acad. Sci. 124, Marey, E.J., 1882a. La machine animale. animaux quadrupèdes. Compt. Rend.
913–914. Locomotion terrestre et aérienne, 3ième ed. Hebd. Séances Acad. Sci. 108, 194–196.
Le Hello, P., 1899. Du rôle des organes Germer Baillière et Cie, Paris, pp. 144–186. Palmgren, A., 1929. Zur kenntnis der
locomoteurs du cheval. Compt. Rend. Marey, E.J., 1882b. Emploi de la photographie sogenannten schnappgelenke. ii die
Hebd. Séances Acad. Sci. 129, 179–181. instantanée pour l’analyse des mouvements physiologische bedeutung des
Leach, D.H., Dagg, A.I., 1983. Evolution of chez les animaux. Compt. Rend. Hebd. fäderungsphänomens. Z. Ges. Anat. Abt. I
equine locomotion research. Equine Vet. J. Séances Acad. Sci. 94, 1013–1020. Z. Anat Entwickl. Ges. 89, 194–200.
15 (2), 87–92. Marey, E.J., 1883. Emploi des photographies Parsons, K.J., Pfau, T., Wilson, A.M., 2008.
Leach, D.H., Ormrod, K., Clayton, H.M., 1984. partielles pour étudier la locomotion de High-speed gallop locomotion in the
Standardised terminology for the l’homme et des animaux. Compt. Rend. thoroughbred racehorse. I. The effect of
description and analysis of equine Hebd. Séances Acad. Sci. 94, 1827–1831. incline on stride parameters. J. Exp. Biol.
locomotion. Equine Vet. J. 16 (6), Marey, E.J., Pagès, C., 1886. Analyse 211, 935–944.
522–528. cinématique de la locomotion du cheval. Peham, C., Licka, T., Schobesberger, H.,
Leach, D.H., Sprigings, E., 1979. Gait fatigue Compt. Rend. Hebd. Séances Acad. Sci. Meschan, E., 2004. Influence of the rider
in the racing Thoroughbred. J. Equine Med. 103, 538–547. on the variability of the equine gait. Hum.
Surg. 3, 436–443. Marey, E.J., Pagès, C., 1887. Locomotion Movement Sci. 23, 663–671.
Leach, D.H., Sprigings, E.J., Laverty, W.H., comparée. Mouvement du membre pelvien Peham, C., Schobesberger, H., 2004. Influence
1987. A multivariate statistical analysis of chez l’homme, l ‘éléphant et le cheval. of the load of a rider on a region of
stride timing measurements of nonfatigued Compt. Rend. Hebd. Séances Acad. Sci. increased stiffness of the equine back: a
racing Thoroughbreds. Am. J. Vet. Res. 48 105, 149–156. modelling study. Equine Vet. J. 36,
(5), 880–888. Merkens, H.W., Schamhardt, H.C., Hartman, 703–705.
Lecoq, F., 1843. Traité de l’extérieur du cheval, W., Kersjes, A.W., 1986. Ground reaction Peters, F., 1879. Mechanische untersuchungen
et des principaux animaux domestiques, force patterns of Dutch Warmblood horses an den gelenken und hufe des pferdes,
Savy, Lyon. at normal walk. Equine Vet. J. 18 (3), August Hirschwald, Berlin.
Lenoble du Teil, J., 1873. Étude sur la 207–214. Pettigrew, J.B., 1873. Animal locomotion or
locomotion du cheval et des quadrupèdes Meyer, H., 1992. Rollkur. St. Georg 11, 70–73. walking, swimming and flying with a
en général, Librairie Militaire de J. Mosimann, W., Micheluzzi, P., 1969. Die dissertation on aeronautics, Henry S. King
Demaine, Paris. bewegung im cubitus des pferdes als & Co, London.

28
 References

Pfau, T., Robilliard, J.J., Weller, R., Jespers, K., Riemersma, D.J., Schamhardt, H.C., Hartman, Schauder, W., 1924a. Anatomische und
Eliashar, E., Wilson, A.M., 2007. W., Lammertink, J.L.M.A., 1988a. Kinetics metrische Untersuchungen über die
Assessment of mild hind limb lameness and kinematics of the equine hindleg. In Muskeln der Schultergliedmaße des
during over ground locomotion using vivo tendon loads and force plate Pferdes. Z. F. D. Ges. Anat. I. Abt. 71,
linear discriminant analysis of inertial measurements in ponies. Am. J. Vet. Res. 559–637.
sensor data. Equine Vet. J. 39, 407–413. 49 (8), 1344–1352. Schauder, W., 1924b. Die fetale entwicklung
Pfau, T., Stubbs, N.C., Kaiser, L.J., Brown, L.E., Riemersma, D.J., Van den Bogert, A.J., der ‘sehnenmuskeln’ des pferdes. Arch.
Clayton, H.M., 2012. Effect of trotting Schamhardt, H.C., Hartman, W., 1988b. Mikrosk. Anat. Entwickl. Mechanik 102,
speed and circle radius on movement Kinetics and kinematics of the equine 211–262.
symmetry in horses during lunging on a hindleg. In vivo tendon strain and joint Schauder, W., 1932. Ueber korrelative
soft surface. Am. J. Vet. Res. 73, kinematics. Am. J. Vet. Res. 49 (8), funktionelle gestaltungen an den
1890–1899. 1353–1359. gliedmaßen von ungulaten besonders
Pourcelot, P., Audigié, F., Lacroix, V., Denoix, Robin, D., Chateau, H., Pacquet, L., Falala, S., equiden. Münchn. Tierärztl. Wschr. 83
J.-M., Crevier-Denoix, N., 2002. A method Valette, J.-P., Pourcelot, P., et al., 2009. Use (21), 241–244; (22), 257–261.
free of markers to measure 3-D of a 3D dynamometric horseshoe to assess Schauder, W., 1949. ‘Nur im werden erfaßt,
morphometrical data using the DLT the effects of an all-weather waxed track wird das gewordene verständlich’. Gießener
technique. Arch. Physiol. Biochem. 110, 66. and a crushed sand track at high speed trot. Hochschulnachrichten 56–75.
Pourcelot, P., Defontaine, M., Ravary, B., Equine Vet. J. 41, 253–256. Schauder, W., 1951. allgemeine stoßbrechende
Lemâtre, M., Crevier-Denoix, N., 2005. A Roepstorff, L., Drevemo, S., 1993. Concept of einrichtungen an den gliedmaßen des
non-invasive method of tendon force a force-measuring horseshoe. Acta Anat. pferdes in funktionell-anatomische
measurement. J. Biomech. 38, 2124–2129. 146, 114–119. hinsicht. Dtsch. Tierärztl. Wschr. 58,
Pratt Jr., G.W., 1984. Racing surfaces – A survey Roepstorff, L., Johnston, C., Drevemo, S., 350–352.
of mechanical behavior. Proc. 30th Annu. 2001. In vivo and in vitro expansion in Schauder, W., 1952. Die besonderen
Conv. Am. Assoc. Equine Pract. 321–331. relation to shoeing and frog pressure. stoßbrechenden einrichtungen an den
Pratt Jr., G.W., O’Connor Jr., J.T., 1976. Force Equine Vet. J. Suppl. 33, 54–57. gliedmaßen des pferdes, in angewandt-
plate studies of equine biomechanics. Am. Roethlisberger-Holm, K., Wennerstrand, J., anatomische hinsicht. Dtsch. Tierärztl.
J. Vet. Res. 37, 1251–1255. Lagerquist, U., Eksell, P., Johnston, C., Wschr. 59, 35–38.
Preuschoft, H., Fritz, M., 1977. Mechanische 2006. Effect of local analgesia on Schauder, W., 1954. Stoßabfangende
beanspruchungen im bewegungsapparat movement of the equine back. Equine Vet. einrichtungen in rumpfe des pferdes.
von springpferde. Fortschr. Zool. 24, J. 38, 65–69. Dtsch. Tierärztl. Wschr. 61, 7–9.
75–98. Rogers, C.W., Back, W., 2007. The effect of Schmaltz, R., 1922a. Analyse der gangarten des
Pritchard, C.C., 1965. Relationship between plain, eggbar and 6 degrees-wedge shoes pferdes durch den film. Berl. Tierärztl.
conformation and lameness in the foot. on the distribution of pressure under the Wschr. 38 (46), 523–527.
Auburn Vet. 22, 11–26, 29. hoof of horses at the walk. N. Z. Vet. J. 55, Schmaltz, R., 1922b. Die analyse der gangarten
Raabe, C.H., 1883. Locomotion du cheval. 120–124. des pferdes durch den film. bemerkung
Cadran hippique des allures marchées, T. Rogers, C.W., Davies, A.S., Pfeiffer, D.U., dazu. Berl. Tierärztl. Wochenschr. 38, 551.
Symonds, Paris. Davie, P.S., 1999. Temporal changes in the Schmidt, H., 1939. Beziehungen zwischen
Radescu, T., 1923. Biometrische trot between the first and third horse schrittlänge und bau der gliedmaßen des
untersuchungen an vollblutpferden in inspection at a CCI*** 3-day event. Equine pferdes. Dtsch. Tierärztl. Wschr. 47 (46),
vergleich mit rennleistung, dissertation, Vet. J. Suppl. 30, 254–257. 689–692.
Berlin. Rooney, J.R., 1968. Biomechanics of equine Schreiber, J., 1950. Gestalt und bewegung.
Rathor, S.S., 1968. Clinical aspects of the lameness. Cornell Vet. 58 (Suppl), 49–58. Eine anatomische betrachtung des pferdes
functional disorders of the equine and Rooney, J.R., 1969. Biomechanics of lameness in der darstellenden kunst. Wien. Tierärztl.
bovine femoro-patellar articulation with in horses, Williams & Wilkins, Baltimore. Monatsschr. 37 (2), 73–105.
some remarks on its biomechanics, Thesis, Rösiö, B., 1927. Die bedeutung des exterieurs Schtserbakow, N.M., 1935. The determination
Utrecht. und der konstitution des pferdes für seine of the work for which a horse is suitable by
Ratzlaff, M., 1974. Locomotion of the horse. leistungsfähigkeit, Almqvist & Wiksells, means of measurement of its gait. Bull.
Anat. Histol. Embryol. 3, 376–377. Uppsala. Moscow Vet. Inst. 2, 113–146.
Ratzlaff, M.H., Grant, B.D., Adrian, M., Santamaría, S., Bobbert, M.F., Back, W., Seeherman, H.J., Morris, E., O’Callaghan,
Feeney-Dixon, C., 1979. Evaluation of Barneveld, A., van Weeren, P.R., 2005. M.W., 1996. Comprehensive clinical
equine locomotion using electrogoniometry Effect of early training on the jumping evaluation of performance. In: Auer, J.A.
and cinematography: research and clinical technique of horses. Am. J. Vet. Res. 66, (Ed.), Equine surgery, Saunders,
applications. Proc. 25th Annu. Conv. Am. 418–424. Philadelphia, 1133–1173.
Assoc. Equine Pract. 381. Schamhardt, H.C., Hartman, W., Lammertink, Seeherman, H.J., Morris, E.A., Fackelman, G.,
Ratzlaff, M.H., Shindell, R.M., White, K.K., J.L.M.A., Badoux, D.M., 1984. Bone strain 1987. Computerized force plate
1995. The interrelationships of stride in the equine tibia: inertia as a cause of the determination of equine weight-bearing
length and stride times to velocities of presupport peak. Am. J. Vet. Res. 45, profiles. Proc. 2nd Int. Conf. Equine Exerc.
galloping horses. J. Equine Vet. Sci. 15, 885–887. Physiol. 536–552.
279–283. Schamhardt, H.C., Merkens, H.W., 1987. Silver, I.A., Brown, P.N., Goodship, A.E.,
Rhodin, M., Johnston, C., Holm, K.R., Quantification of equine ground reaction Lanyon, L.E., McCullagh, K.G., Perry, S.C.,
Wennerstrand, J., Drevemo, S., 2005. The force patterns. J. Biomech. 20 (4), 443–446. et al., 1983. A clinical and experimental
influence of head and neck position on Schauder, W., 1923a. Historisch-kritische study of tendon injury, healing and
kinematics of the back in riding horses at studie über die bewegungslehre des pferdes. treatment in the horse. Equine Vet. J.
the walk and trot. Equine Vet. J. 37, 7–11. (1. teil) Berl. Tierärztl. Wschr. 39 (12), Suppl. 1.
Richter, O., 1953. Korrelaten zwischen 123–126. Simpson, G.G., 1951. Horses, Oxford
körpermaßen und leistungen bei dem auf Schauder, W., 1923b. Historisch-kritische University Press, New York.
amerikanischer grundlage gezogenen studie über die bewegungslehre des pferdes. Slijper, E.J., 1946. Comparative biologic-
deutschen traber, inauguraldissertation, (2. teil) Berl. Tierärztl. Wschr. 39 (13), anatomical investigations on the vertebral
Berlin. 135–137. column and spinal musculature of

29
 1 History

mammals. Proc. K. Ned. Acad. Wetensch. with a rolled toe optimises hoof- S., 2004. Kinematic evaluation of the back
42 (5), 1–128. unrollment and lowers peak loading during in the sport horse with back pain. Equine
Solá, G.A., 1969. Environmental factors breakover. Equine Vet. J. 38, 258–262. Vet. J. Suppl. 36, 707–711.
affecting the speed of pacing horses, MS Van Weeren, P.R., Barneveld, A., 1986. A Wentink, G.H., 1978a. An experimental study
Thesis, Ohio State University. technique to quantify skin displacement in on the role of the reciprocal tendinous
Stillman, J.D.B., 1882. The horse in motion as the walking horse. J. Biomech. 19 (10), apparatus of the horse at walk. Anat.
shown by instantaneous photography with 879–883. Embryol. 154, 143–151.
a study on animal mechanics founded on Van Weeren, P.R., Van den Bogert, A.J., Wentink, G.H., 1978b. Biokinetical analysis of
anatomy and the revelations of the camera Barneveld, A., 1992. Correction models for the movements of the pelvic limb of the
in which is demonstrated the theory of skin displacement in equine kinematic gait horse and the role of the muscles in the
quadrupedal locomotion. Boston: James R. analysis. J. Equine Vet. Sci. 12 (3), walk and trot. Anat. Embryol. 154,
Osgood & Co. 178–192. 261–272.
Stoss, A.O., 1923. Anatomie und kinematik Van Weeren, P.R., Van den Bogert, A.J., Wiechert, F., 1927. Messungen an
der gelenke der pferdeextremitäten. Z. Anat. Barneveld, A., Hartman, W., Kersjes, A.W., ostpreußischen kavalleriepferden und
Entwickl. Ges. 69, 5–31. 1990. The role of the reciprocal apparatus solchen mit besonderen leistungen und die
Stratul, J., 1922. Biometrische untersuchungen in the hind limb of the horse investigated beurteilung der leistungsfähigkeit auf grund
an vollblutpferden mit rückschlüssen auf by a modified CODA-3 opto-electronic gait der mechanischen verhältnisse. DGfZ 34,
rennleistung, inauguraldissertation, Berlin. analysis system. Equine Vet. J. Suppl. 9, 1–67.
95-100. Williams, S.B., Payne, R.C., Wilson, A.M.,
Strubelt, H., 1928. Über die bedeutung des
Vergari, C., Pourcelot, P., Ravary-Plumioën, B., 2007. Functional specialisation of the
lacertus fibrosus und des tendo
Dupays, A.G., Denoix, J.M., Mitton. D., pelvic limb of the hare (Lepus europeus). J.
femorotarseus für das stehen und die
Laugier, P., Crevier-Denoix, M., 2012. First Anat. 210, 472–490.
bewegung des pferdes. Arch. Tierheilk. 57,
575–585. application of axial speed of sound to Wilsdorf, G., 1971. Beitrag zur biomechanik
follow up injured equine tendons. der vorderextremität des galopprennpferdes
Sukhanov, V.B., 1963. Forms of movement
Ultrasound Med. Biol. 38, 162–167. und deren beziehung zur sogenannten
(gaits) of land vertebrates (a theory of
Von den Driesch, A., 1989. Geschichte der schienbeinerkrankung. Mon. Hefte Vet.
locomotion and of the evolution of its
Tiermedizin. 5000 Jahre Tierheilkunde, Med. 26 (24), 939–944.
forms). Byull. Mosk. Obshch. Ispyt. Prirody
Otd. Biol. 67 (5), 136–137. Callwey, München. Wilson, A.M., McGuigan, M.P., Su, A., van den
Bogert, A.J., 2001. Horses damp the spring
Taylor, B.M., Tipton, C.M., Adrian, M., Von Hochstätter, C., 1821–24. Theoretisch-
in their step. Nature 4141 (6866),
Karpovich, P.V., 1966. Action of certain praktisches handbuch der äussern
895–899.
joints in the legs of the horse recorded pferdekenntniß, und der wartung und
pflege der pferde, Haller, Bern. Wilson, A.M., Seelig, T.J., Shield, R.A.,
electrogoniometrically. Am. J. Vet. Res. 38
Silverman, B.W., 1998. The effect of
(10), 1675–1677. Von Lengerken, G., Werner, K., 1969. Das
foot imbalance on point of force
Thorpe, C.T., Marlin, D.J., Franklin, S.H., exterieur der zucht- und renntraber in
application in the horse. Equine Vet. J.
Colborne, G.R., 2009. Transverse and der DDR. Wiss. Z. Univ. Halle 18 (5),
30, 540–545.
dorso-ventral changes in thoracic 505–518.
Wilson, A.M., Watson, J.C., Lichtwark, G.A.,
dimension during equine locomotion. Vet. Wagener, H., 1934. Untersuchungen an 2003. Biomechanics: A catapult action for
J. 179, 370–377. spitzenpferden des spring- und schulstalles rapid limb protraction. Nature 421 (6918),
Townsend, H.G.G., Leach, D.H., 1984. The der kavallerie-schule hannover. DGfZ 65, 35–36.
relationship between intervertebral jont 1–117.
Wilson, B.D., Neal, R.J., Howard, A.,
morphology and mobility in the equine Walter, K., 1925. Der bewegungsablauf an den Groenendyk, S., 1988a. The gait of pacers.
thoracolumbar spine. Equine Vet. J. 16 (5), freien gliedmaßen des pferdes im schritt, I. Kinematics. Equine Vet. J. 20 (5),
461–465. trab und galopp. Arch. Wissensch. Prakt. 341–346.
Townsend, H.G.G., Leach, D.H., Doige, C.E., Tierheilk. 53, 316–352.
Wilson, B.D., Neal, R.J., Howard, A.,
Kirkaldy-Willis, W.H., 1986. The Wehner, R., 1944. Lassen sich beziehungen der Groenendyk, S., 1988b. The gait of pacers.
relationship between spinal biomechanics knochenachsen und gliedmaßenwinkel und II. Factors influencing racing speed. Equine
and pathological changes in the equine schrittlänge beim rheinisch-deutschen Vet. J. 20 (5), 347–351.
thoracolumbar spine. Equine Vet. J. 18 (2), kaltblut nachweisen? Z. Tierz. Zücht. Biol. Xenophon (430–354 BC) Über die reitkunst.
107–112. 56, 321–353. (In: Widdra K (1965) Xenophon. Reitkunst.
Townsend, H.G.G., Leach, D.H., Fretz, P.B., Weishaupt, M.A., Hogg, H.P., Wiestner, T., Schriften und quellen der alten welt 16:
1983. Kinematics of the equine Denoth, J., Stussi, E., Auer, J.A., 2002. Berlin).
thoracolumbar spine. Equine Vet. J. 15 (2), Instrumented treadmill for measuring Zwaenepoel, M., 1910a. Démonstration
117–122. vertical ground reaction forces in horses. expérimentale du mécanisme de
Van den Bogert, A.J., Schamhardt, H.C., Am J. Vet. Res. 63, 520–527. l’impulsion et du recul chez le cheval. Ann.
Crowe, A., 1989. Simulation of Weishaupt, M.A., Wiestner, T., Hogg, H.P., Méd. Vétérin. 59, 322–334.
quadrupedal locomotion using a dynamic Jordan, P., Auer, J.A., 2004. Vertical ground Zwaenepoel, M., 1910b. Démonstration
rigid body model. J. Biom. 22 (1), 33–41. reaction force-time histories of sound expérimentale du mécanisme de
Van der Plank, G.M., 1929. De verdeeling der warmblood horses trotting a treadmill. Vet. l’impulsion et du recul chez le cheval. Ann.
lichaamslast over de draagvlakte van den J. 168, 304–311. Bull. Soc. Roy. Sciences Méd. Natur.
voorhoef, bij verhooging der verzenen. Weishaupt, M.A., Wiestner, T., Von Peinen, K., Bruxelles 68, 83–91.
Tijdschr. Dierg. 55 (10), 498–500. Waldern, N., Roepstorff, L., van Weeren, R., Zwaenepoel, M., 1911a. Démonstration
Van Heel, M.C.V., Kroekenstoel, A.M., Van et al., 2006. Effect of head and neck expérimentale du mécanisme de
Dierendonck, M.C., van Weeren, P.R., Back, position on vertical ground reaction forces l’impulsion chez le cheval. Ann. Méd.
W., 2006b. Uneven feet in a foal may and interlimb coordination in the dressage Vétérin. 60, 461–478.
develop as a consequence of lateral grazing horse ridden at walk and trot on a Zwaenepoel, M., 1911b. Démonstration
behaviour induced by conformation. treadmill. Equine Vet. J. Suppl. 36, expérimentale du mécanisme de
Equine Vet. J. 7, 646–651. 387–392. l’impulsion chez le cheval. Bull. De l’Acad.
Van Heel, M.C.V., van Weeren, P.R., Back, W., Wennerstrand, J., Johnston, C., Roethlisberger- Roy. De Méd. IVe Série Tôme XXV,
2006a. Shoeing sound Warmblood horses Holm, K., Erichsen, C., Eksell, P., Drevemo, 527–558.

30
 C H A PTER 2 

Measurement techniques for

gait analysis
Hilary M. Clayton, Henk C. Schamhardt (deceased)

Introduction of correct hits usually has a lower SEM. Are data from that limb
more ‘correct’ than those from the other limb? Obviously not, but
it is not easy to define a universal, statistically correct recipe to deal
In a qualitative evaluation of a horse’s gaits or movements, the with this problem. In practice, most laboratories collect data until
human eye captures the image and the brain processes the informa- a certain minimum number of correct hits have been recorded from
tion to form an opinion based on evaluating the observed motion each limb. In sound horses, both the kinematic and force variables
in the context of previous experience. Even though some judges of are quite stable, and analysis of three to five strides is sufficient to
horses and equestrian sports are very astute observers, subjectivity give a representative value for kinematic (Drevemo et al., 1980a) or
is inherent in this type of judgment. In assigning a lameness grade, GRF (Schamhardt, 1996) analysis. The mean value is then used in
clinicians draw on their powers of observation and previous experi- further stages of the analysis as being representative of that variable
ence to assign a lameness score in a semi-quantitative analysis. for a particular limb in one horse. Most of the stride variables show
Experienced clinicians may be consistent in their scores (Back et al., good repeatability over the short and long term (Drevemo et al.,
1993), but there is considerable variation between clinicians 1980b; van Weeren et al., 1993; Ishihara et al., 2005; Lynch et al.,
(Keegan et al., 1998). 2005), and the stride kinematics of a young horse have already
Scientific analysis requires accurate quantitative data describing assumed the characteristics that they will have at maturity by the
the movements and the associated forces. Kinematic analysis mea- time the foal is 4 months of age (Back et al., 1994).
sures the geometry of movement without considering the forces that When planning an experiment, knowledge of the reliability of a
cause the movement. At the present time, the majority of kinematic variable should be combined with estimates of the sample size
evaluations are performed using optoelectronic systems consisting required to detect significant differences. Thus, knowledge of vari-
of integrated hardware and software components. Kinetics is the ability within and between recording sessions is helpful in selecting
study of the internal and external forces that are associated with the variables for analysis and in calculating the number of trials and
movements. A variety of transducers, including strain gauges, piezo- subjects needed to obtain meaningful results in research studies.
electric and piezoresistive transducers, accelerometers, gyroscopes Some horses may, inherently, show more variability than others
and magnetometers are used in kinetic studies. Several transducers (Lynch et al., 2005) and variability may be affected by many of the
can be combined into force plates and force shoes for measuring conditions that are being studied, such as lameness, the presence
ground reaction forces (GRFs) or into mats for mapping pressure of a rider and the fit of the saddle. Horses with more severe lame-
distribution. Electromyography (EMG) detects the electrical activity ness tend to have larger coefficients of variation for kinetic variables,
associated with muscular contraction as a means of determining both within and between horses (Ishihara et al., 2005). On the
muscular activation patterns during different activities. Variables other hand, variability in stride length increased significantly when
that are not or cannot be measured directly may be calculated from lameness was reduced by intra-articular or perineural anesthesia
the data by computer modeling. (Peham et al., 2001). It was suggested that variations from the
optimal motion pattern are associated with pain, so lame horses
maintain a highly consistent kinematic pattern, whereas sound
Interpreting the effects of biological variability horses are not constrained by the association between kinematic
variation and pain, so other influences on motion, such as external
Variability is inherent to data obtained from repeated measure- stimuli become more influential. In horses moving on a treadmill,
ments on biological material. For example, when the horse is variability in forward velocity and acceleration decreased when the
guided over the force plate, the chance for a correct hit by one fore horses were ridden but the same variables had higher variability
hoof is about 50%, with the number of hits being approximately when the rider used a poorly fitting saddle compared with a well
evenly distributed between the right and left limbs. After several fitting saddle (Peham et al., 2004).
runs, it is likely that a different number of correct hits will have Variability between individual horses affects the response to
been recorded from the right and left limbs. When calculating the certain interferences, such as drug treatment and shoeing, which
mean and the standard error of the mean (SEM) for a force variable differ qualitatively and quantitatively in different animals. Impres-
(e.g. peak vertical force) by averaging the data of these runs, the sive libraries of statistical routines have been developed to extract
mean of the data obtained from the limb with the higher number trends in the data, to detect differences between groups, or to

31
 2 Measurement techniques for gait analysis

identify a ‘statistically significant’ response to a certain treatment. 20

Unfortunately, the prerequisites for these statistical tests, such as
normal distribution of the data or lack of correlation between vari- 15
ables, may invalidate their use in a particular study. Before conduct-
ing an experiment, an hypothesis is formulated, and an appropriate

Percent stride
10
experimental design and statistical model are determined to test
that hypothesis. After collection of data, it is not unusual to find
dependency between variables within series of data, which disquali- 5
fies a particular statistical test. A detailed discussion of the pitfalls
and problems associated with statistical testing is beyond the scope 0
this text. Readers are advised to consult a suitable statistical text or
a statistician to avoid perpetuating the use of incorrect analyses, -5
which have appeared frequently in the literature. 012 5 012 5 0 5 0 5 0 5 0 5
A statistical test determines the likelihood that a certain hypoth- I II III V VII IX
esis can be accepted, or has to be rejected. However, the answer is
not absolute: for example, having selected an uncertainty level of I–IX sessions 0–5 minutes training
p < 0.05, the correct decision to accept or reject the hypothesis will
be made in 95% of cases but has a 5% chance of being wrong, and
one observation out of 20 will differ significantly due to chance. 0.05
Therefore, statistical tests are not proof that a certain hypothesis is
true or false. The majority of equine locomotion studies are based 0.00
on a rather small number of subjects, which may be insufficient to
-0.05
give the required power for a statistical analysis. In these cases,

Seconds
trends in the data may suggest a biologically significant effect that
-0.10
cannot be proven statistically but is, nevertheless, important.
-0.15

Treadmill evaluation -0.20

-0.25
The treadmill is extremely useful for equine gait analysis due to the 012 5 012 5 0 5 0 5 0 5 0 5
ability to control the speed of movement and the environment I II III V VII IX
around the horse. It must be recognized, however, that stride kine-
I–IX sessions 0–5 minutes training
matics on the treadmill differ in some respects from over ground
locomotion (Fredricson et al., 1983; Barrey et al., 1993; Buchner Fig 2.1  Habituation to treadmill locomotion in 10 horses determined by
et al., 1994b). Horses trotting at the same speed under both condi- changes in hind limb stance duration, expressed as a percentage of stride
tions use a higher stride frequency and a longer stride length on the duration (above) and in seconds (below). The horizontal axis shows the
treadmill (Barrey et al., 1993). The treadmill is also associated with number of training sessions, each of 5 min duration. The vertical axis shows
longer stance durations, earlier placement of the forelimbs, greater the relative stance duration. Reductions in relative stance duration are
retraction of both fore and hind limbs and reduced vertical excur- regarded as a sign of habituation. The horizontal line indicates the
sions of the hooves and the withers (Buchner et al., 1994b). ‘habituation limit’ based on data of the final recording session. Vertical bars
A period of habituation is required before horses move consis- indicate standard deviations within 10 horses.
tently on the treadmill, with habituation occurring more rapidly at Reprinted from Buchner, H.H.F., Savelberg, H.H.C.M., Scharmhardt, H.C., et al., 1994.
faster gaits. Rapid adaptation is seen during the first few training Kinematics of treadmill versus overground locomotion in horses, Veterinary Quarterly, Vol 16,
sessions, and by the end of the third 5-min session, the kinematics supp2, S87–S90, with permission of Taylor & Francis Ltd, http://www.tandf.co.uk/journals.
of the trot have stabilized (Fig. 2.1), whereas walk kinematics are
not fully adapted even at the tenth session (Buchner et al., 1994a).
During the first session and, to a lesser extent at the start of subse-
quent sessions, the horse takes short, quick steps, with the withers diminish its value for clinical and research studies involving com-
and hindquarters lowered, and the feet splayed to the side to widen parisons between locomotion on the treadmill under different con-
the base of support. Even experienced horses take at least one ditions, for example in evaluating hoof balance and the flight arc
minute for their gait pattern to stabilize each time the treadmill belt of the hoof.
starts moving (Buchner et al., 1994a). When measurements are Kinematic analysis of horses moving on a treadmill has been used
made during treadmill locomotion, it is recommended that horses to study many aspects of equine locomotion, including movements
habituate for one minute after a change of gait or speed before of the limbs (Back et al., 1995a, 1995b), ontogeny of the trot (Back
making steady state measurements (Buchner et al., 1994a). et al., 1994), response to training (van Weeren et al., 1993; Corley
Horses moving on a treadmill use less energy than horses moving & Goodship, 1994), development of gait asymmetries (Drevemo
over ground at the same speed (Sloet & Barneveld, 1995), which et al., 1987) and kinematic adaptations used by the horse to manage
may be partly due to a power transfer from the treadmill to lameness (Peloso et al., 1993; Buchner et al., 1995, 1996a, 1996b).
the horse. Although the speed of the treadmill belt is assumed The ability to incline the treadmill belt allows studies of stride
to be constant, in fact it is decreased by about 9% in early stance kinematics when horses move on an incline or decline (Sloet et al.,
due to the frictional effect of the vertical force component and the 1997; Hoyt et al., 2000; Dutto et al., 2004b; Hodson-Tole 2006).
decelerating effect of the longitudinal force component exerted by Mounted studies have also been performed on the treadmill,
the horse’s hoof. Towards the end of the stance phase the frictional which provides a consistent speed and environment for assessing
effect of the vertical force declines while the propulsive longitudinal the influence of a rider (Barrey et al., 1993; Peham et al., 2004), the
force tends to accelerate the belt (Schamhardt et al., 1994). effects of different training techniques or riding styles (Gomez
Although the kinematics and energetics of treadmill locomotion Alvarez et al., 2006; Weishaupt et al., 2006b) and the role of saddle
are not exactly equivalent to over ground locomotion, this does not fit (Peham et al. 2004; Meschan et al., 2007).

32
 Kinematic analysis

1–3 cm in diameter, and are available from hobby stores. If neces-

sary, the balls can be cut in half before covering them with strips of
retro-reflective tape or reflective paint. The optimal diameter of the
spheres depends on the resolution of the cameras. Smaller markers
are lighter in weight, which facilitates marker retention, and allows
a larger number of markers to be differentiated within a small
volume.
Fig 2.2  Stick figure of the forelimb of a horse trotting over ground. The Adhesion of skin markers presents some problems, especially
figures are shown at intervals of 10% of stance duration. when the horse travels at high speed or sweats profusely. Double-
sided tape or high performance glue may be effective for securing
the markers, but some types of glue, such as cyanoacrylate, are dif-
ficult to remove completely at the end of the study, which may pose
Kinematic analysis a problem in client-owned horses. A little experimentation may be
needed to find a range of appropriate products that are suitable for
attaching markers under a variety of circumstances.
Kinematic analysis quantifies the features of gait that are assessed When markers are fixed to the skin, they are usually attached
qualitatively by visual evaluation in terms of temporal (timing), over specific bony landmarks. Accurate palpation skills are a pre-
linear (distance) and angular measurements. Visualization of kine- requisite. Inaccuracy in identification of anatomical landmarks is
matic data is facilitated by constructing animated stick figures (Fig. a major source of error in kinematic analysis (Weller et al., 2006).
2.2) and by plotting the results graphically. Statistical analysis can In order to minimize the inaccuracies, care should be taken that
be used to compare the entire curves (e.g. Deluzio & Astephen, the horse is standing squarely with weight on all four limbs. Any
2007) or to compare discrete values that represent important events. change in position or loading of a limb alters the cutaneous rela-
At the present time, the most common method of motion analy- tionship to the underlying bone. It is useful to mark the attach-
sis involves optical motion capture, which may be accomplished ment sites on the skin or hoof wall to facilitate reattachment in
using passive markers, active markers or marker-less methods. the same place if a marker is lost during data collection or if
Passive markers are coated with a reflective material that reflects markers are removed and replaced during sequential recording
back along the same path as the incident light. Active markers are sessions.
usually LEDs that are distinguished by flashing in a temporal When markers fixed to the skin are used to represent motion of
pattern. Marker-less systems are based on pattern recognition soft- the bones, sliding of the skin over the bones introduces a significant
ware that locates and tracks the area of interest. source of error. In the proximal limb, the skin moves by as much
Optical motion capture is generally adequate for animation and as 12 cm relative to the bone in the sagittal plane (van Weeren et al.,
gait analysis. Markers are tracked automatically based on the emis- 1990b), which is sufficient to change the entire shape of the angle–
sion and detection of infrared or visible light, though other methods, time diagrams at the proximal joints (Back et al., 1995a). In the
such as ultrasonic tracking and videographic analysis are also used. proximal limbs, uncorrected data cannot be used for absolute
Optical technology is limited by its use of cameras to record activity; angular computations or for measuring muscle or tendon lengths
multiple cameras surrounding the action are needed to generate based on limb kinematics. Correction for skin displacement may
three-dimensional models, which generally restricts optical motion not be of primary importance in clinical applications, especially
capture to a laboratory setting. In addition to the accurate (and when comparing data in a repeated measures design, but it is essen-
expensive) systems used for precision analysis, less sophisticated tial in biomechanical applications when absolute values are impor-
video-based analysis systems are available that provide less accurate tant (van Weeren et al., 1992).
measurements but at a more affordable price. Bone-fixed markers are usually attached via a Steinmann pin
The sequence of events for motion analysis involves attaching inserted percutaneously into the bone. A cluster of at least three
markers to the subject, setting up and calibrating the recording tracking markers is rigidly attached to the pin immediately before
space, recording the subject in motion, tracking the markers to data collection (Chateau et al., 2004, 2006; Clayton et al., 2004,
obtain digital coordinates, and data analysis that might include 2007a, 2007b; Khumsap et al., 2004).
smoothing, normalization, transformation and interpretation of Marker locations are chosen in accordance with the purposes of
the results. Not all of these steps are required for every motion the analysis. Calculation of the angle between two limb segments
analysis project. in two dimensions requires a minimum of three markers: one
over the center of joint rotation and one on each segment, prefer-
Markers ably as far as possible from the marker representing the joint
center. Figure 2.3 shows the approximate centers of joint rotation
Most motion analysis systems are based on tracking markers fixed in the sagittal plane in the fore and hind limbs (Leach & Dyson,
to the skin or the bones. For two-dimensional analysis, circular 1988). These locations are used with software that requires
markers, 1–3 cm in diameter are used, with bigger markers giving markers to be placed over the joint centers. An alternative tech-
better accuracy when the resolution of the system is poor (Scham- nique uses two markers aligned along the long axis to represent
hardt et al., 1993a). Retroreflective material, which reflects light the orientation of the segment with the difference between seg-
back along the same path as the incident light, can be purchased in mental angles being the joint angle. Ideally, marker locations
sheets (Scotchlite, 3M Corp., St Paul, MN) and used to make should be easily identified by palpation and should be in sites
markers of a suitable size or precut circles can be purchased at a where skin displacement relative to the underlying bones is
higher price. Retroreflective paint (Scotchlite 7210 Silver, 3M Corp., minimal or correction algorithms are available for extracting the
St Paul, MN) is also available, but clumping of the reflective beads effects of skin displacement.
makes it difficult to use. Markers are placed on the dorsal midline of the back to evaluate
For three-dimensional studies, the views from multiple cameras trunk motion (Licka & Peham, 1998; Faber et al., 2002; Johnston
are integrated so that movements of the markers can be extrapolated et al., 2004). These trunk markers can also be used to check that
into three dimensions. Spherical or hemispherical markers are used the horse is moving straight and is aligned with the axes of the
because they retain their circular shape when viewed from different Global Coordinate System (GCS) during data collection (Fig. 2.4).
angles. Spherical markers can be purchased or they can be fabricated Markers on the poll, withers and croup are useful for evaluating
from polystyrene or balsa wood balls. These markers are typically vertical excursions of these reference points and for detecting

33
 2 Measurement techniques for gait analysis

Fig 2.3  The black dots represent locations that are commonly used for skin marker placement for two-dimensional kinematic analysis. Limb markers are
placed over the centers of joint rotation.

asymmetries associated with lameness (Buchner et al. 1996a; to the cameras during locomotion and at locations for which the
Keegan et al., 2004). skin displacement is known. A stationary file is recorded with
Three-dimensional analysis requires a minimum of three non- both the virtual and tracking markers in place, after which the
colinear markers per segment. Ideally, the markers should be widely virtual markers are removed. Trials are recorded with only the track-
distributed over the segment and, as for two-dimensional analysis, ing markers in place.
placed in locations that show minimal skin displacement or for In the sagittal plane, skin displacement relative to the underlying
which correction algorithms for skin displacement are available. bones has been quantified and correction algorithms have been
Each marker must be visible to at least two cameras throughout the developed to calculate skin motion relative to specific bony land-
movement and accuracy improves with an increase in the number marks from the scapula to the metacarpus and from the pelvis to
of cameras tracking a marker. When markers are required in loca- the metatarsus of walking and trotting Dutch Warmblood horses
tions where they are difficult to track, it is possible to use a virtual (van Weeren et al., 1990a, 1990b, 1992). However, these algorithms
targeting system (Nicodemus et al., 1999). This method relies on are only valid for horses of similar conformation, moving at the
the fact that, for rigid body motion, the location of any point on a same gaits and at similar speeds. For three-dimensional analysis,
body does not change with respect to that body. Therefore, if the correction algorithms are available for sites on the crural and meta-
location of a point on a segment is known with respect to the posi- tarsal segments (Lanovaz et al., 2004) and on the forearm segment
tion of the markers on that segment and the orientation of (Sha et al., 2004) of trotting horses.
the segment is known in a GCS, then the location of that point
on that segment can be calculated in the GCS. The virtual targeting
method employs two sets of markers: tracking markers and virtual
Calibration
markers. Three (virtual or tracking) markers attached to each The recording area or volume must be calibrated in order to scale
segment are used to define the segmental coordinate system: two the linear measurements. The accuracy of the calibration directly
are oriented along the long axis of the segment and the third is determines the accuracy of the final three-dimensional data
perpendicular to that axis. Three non-collinear tracking markers are (DeLuzio et al., 1993), which emphasizes the importance of invest-
placed on the segment in appropriate positions to track the motion ing the necessary effort into calibration of the volume space in
of the segment in the GCS, i.e. at locations that are readily visible which the measurements are made.

34
 Kinematic analysis

movement angle is less than about 15°, this error is smaller than
5%. If the analysis involves measuring a transverse distance, such
as step width between the left and right limbs, the error is larger
and increases in proportion to the sine of the oblique movement
angle. If the horse moves at an angle of 15°, the error in the trans-
verse direction can be as large as 26%.
For three-dimensional studies a calibration frame with non-
coplanar control points can be used. A larger frame with more
numerous control points gives a more accurate reconstruction. The
accuracy of the data is markedly reduced outside the calibrated
volume, so a large, custom-designed frame is required for equine
studies.
Modern motion analysis systems are calibrated using a dynamic
linearization technique. First, the x, y and z axes of the GCS are
defined, then a wand of appropriate length for the size of the
capture volume and with markers at known locations is waved in
three planes throughout the capture volume to establish camera
linearization parameters. Software locates the cameras, calibrates
the volume, corrects for camera lens distortions and calculates the
error in linear measurements.

Sampling frequency
In order to reconstruct (interpolate) a signal from a sequence of
samples, sufficient samples must be recorded to capture the peaks
and troughs of the original waveform. If a waveform is sampled at
less than twice its frequency the reconstructed waveform will effec-
tively contribute only noise. This phenomenon is called ‘aliasing’.
For most equine kinematic studies, a sampling frequency of 120 Hz
is adequate, unless the objective is to study short duration events
such as impact, which require a much higher sampling frequency.

Digitization
Digitization determines the coordinates of the markers in two-
dimensional or three-dimensional space within the GCS. Markers
may be tracked manually, for example when there is too much
ambient light for automated tracking or when markers cannot be
used as in a competition. In addition to the time required, this
tedious process creates more digitizing noise than automated
marker tracking. With automated tracking, the system locates each
marker and then calculates its centroid. Many automated systems
can track complex marker sets in real time. However, the operator
should check each digitized field during post-processing to make
adjustments for digitizing errors, such as marker misidentification,
Fig 2.4  Dorsal view of horse trotting through data collection volume before accepting the data for further analysis.
recorded using Motion Analysis System. The grid pattern shows the
alignment of the horizontal axes of the global coordinate system (GCS). The
position of the markers distributed along the horse’s dorsal midline and on Smoothing
the left and right sides of the head and neck indicates that the craniocaudal
axis of the horse’s body is aligned with the longitudinal axis of the GCS. The During digitization small errors are introduced that constitute
shoulders are slightly turned to the right due to the retraction of the right ‘noise’ in the signal. The effect of noise is not too great in the dis-
forelimb. placement data, but it becomes increasingly apparent in the time
derivatives, i.e. the velocity and acceleration data (Fioretti & Jetto,
1989) as shown in Figure 2.5. Smoothing removes high-frequency
noise introduced during the digitization process using one of two
general approaches: a digital filter followed by finite difference tech-
For two-dimensional analysis, a rectangular frame or a linear ruler nique or a curve-fitting technique (e.g. polynomial or spline curve
is recorded in the plane of movement in different areas of the move- fitting). Selection of an appropriate smoothing algorithm and
ment space. If the horse deviates from the plane of calibration smoothing parameter for a specific purpose requires some expertise
during data collection, errors are introduced in the linear data, and is discussed further in Chapter 3. As a guideline, a low-pass
though the timing data are not affected. Correction algorithms can digital filter with a cut-off frequency of 10–15 Hz is adequate for
be used to adjust the linear data if the horse moves along a line most kinematic studies of equine gait. However, if the movement
parallel to the intended plane of motion, but if the horse travels at of a marker has an oscillatory component, as occurs when loose
an oblique angle to the camera, it introduces image distortion. connective tissue is interposed between the skin and the underlying
Linear measurements in the longitudinal direction (e.g. stride bones, these oscillations are essentially tied to the movements
length) are proportional to the cosine of the angle at which the themselves and cannot be removed by smoothing (Schamhardt,
horse moves relative to the desired direction. As long as the oblique 1996).

35
 2 Measurement techniques for gait analysis

4 stride duration expresses the values of the variables as a percentage

3
of stride duration. This facilitates comparisons between strides
that have slightly different durations, and allows the construction
2 of mean curves from a number of strides. Depending on the objec-
1 tives of the study, normalization to stance duration or swing dura-
tion may be more appropriate than using total stride duration.
0
The most common method of performing time normalization is
-1 to use cubic spline interpolation to resample the curve at a set
-2 number of intervals (usually 101), which gives values at intervals
from 0–100%.
-3
Many kinematic variables are velocity-dependent. In designing
-4 research studies, consideration should be given to standardization
of velocity. There are several ways in which this might be achieved.
4 One method is to set a target velocity that all horses must adhere
3 to within a predetermined, narrow range. This may be adequate
when all the subjects are of similar size but is less satisfactory for
2 horses that differ in size. On the treadmill, an energetically optimal
1 speed can be determined at which the variation between motion
0
cycles is minimized (Peham et al., 1998). This is a good method
when performing repeated analyses of the same horse. However, if
-1 the protocol involves lameness, it is likely that the horse will have
-2 a slower preferred velocity when lame. An alternative approach to
controlling the effects of velocity involves measuring velocity
-3
dependent changes in the variables, then developing a regression
-4 equation to predict the values of these variables at different veloci-
ties (Khumsap et al., 2001).
1000 Joint angles in the sagittal plane may be reported in terms of the
800 absolute joint angle, which is usually measured on the anatomical
600 flexor aspect of the joint. Alternatively, flexion (positive) and exten-
sion (negative) angles can be expressed relative to the position at
400
which the proximal and distal segments are aligned. In some
200 studies, joint angular measurements have been standardized to
0 those obtained in the square standing horse (Back et al., 1994), so
-200 that the angles are reported as deviations from this square standing
-400 position (flexion positive, extension negative). Although the pat-
terns are the same regardless of the method of measurement, the
-600
values differ considerably, which impairs comparisons between
-800 data from different studies.
Fig 2.5  Effect of noise on derived functions. The graphs show a smooth A drawback to reporting absolute values for kinematic variables
curve (top) to which a small amount of noise has been added (center) and is that conformational differences increase the variability between
the second derivative from the smooth (thick red line) and noisy (thin red horses, which decreases statistical power and may make it difficult
line) data (bottom). to detect real differences between experimental conditions. In a
comparison of different methods of standardization for removing
the effects of conformational variation, it was shown that subtrac-
tion of either the standing angle or the impact angle reduced vari-
Transformation ability without changing the data. Subtraction of the average joint
angle during the trial or multiplicative scatter correction reduced
The transformation process integrates the calibration information variability even further but some of the variables changed signifi-
with the digitized coordinates to scale the data. For calculating the cantly (Mullineaux et al., 2004)
x, y and z coordinates of the markers within the GCS, direct linear Horses vary greatly in size and to compensate for the effect of
transformation (DLT) is the standard procedure for combining two size, data can be scaled such that the kinetic and gravitational forces
or more two-dimensional views into a single three-dimensional are proportional to the horse’s height and mass (Alexander, 1977).
view. This method has the advantage of not needing any informa- Temporal variables are adjusted for height at the withers using
tion about camera locations; the transformation is based on knowl- the equations for acceleration and velocity (Alexander, 1977).
edge of the coordinates of the control points on the calibration
frame, which are determined for each camera view.
For studies in which three-dimensional angles are calculated, data distance distance height
acceleration = then time = =
are transformed from the GCS onto a segmental coordinate system time 2 acceleration g
for each segment. This process involves recording a stationary file
with markers in locations that can be used to establish the axes for distance distance
velocity = = = distance ∗ acceleration
each segmental coordinate system. These markers may be the same time distance
as the tracking markers or they may be virtual markers that are acceleration
removed after recording the temporary file.
= height ∗ g

where height is subject height at the withers (m) and g is gravita-

Normalization tional acceleration (9.8 m/s2).
Normalization of data facilitates comparisons between different Stride, stance and swing duration may be standardized by divid-
horses by standardizing certain parameters. Time normalization to ing measured stride, stance and swing duration by the adjusted time

36
 Kinematic analysis

and then expressed as stride duration in dimensionless units (strid- Multi-planar analysis
eDU), stance time in dimensionless units (stanceDU) and swing
duration in dimensionless units (swingDU).12 An approach that has been used in some equine studies is to project
the three-dimensional coordinate data onto three orthogonal
stride duration planes that are tied to the GCS (Fredricson & Drevemo, 1972).
strideDU (s/s) = Provided the horse moves in a direction parallel to a global coor-
height/g
dinate axis, the analytic planes become the sagittal (side view),
stance duration frontal (front or rear view) and dorsal (dorsal or ventral view)
stanceDU (s/s) =
height/g planes. In effect, this method degrades the three-dimensional analy-
sis into a series of quasi-two-dimensional analyses. Joint motion
swing duration that is not parallel to one of the projection planes cannot be accu-
swingDU (s/s) =
height/g rately measured and, since the segments are defined as simple lines
between landmarks, rotations along the long axis of a segment are
Similarly, subject velocities are standardized by dividing measured impossible to measure.
velocity by adjusted velocity and expressed as velocity in dimension-
less units (VDU).
Three-dimensional analysis
velocity
VDU (ms/ms) = A true three-dimensional analysis requires multiple cameras that
height ∗ g
are synchronized precisely and with each marker visible to at
An objective method of predetermining an equivalent velocity for least two cameras at all times. Each length measurement has
horses of different sizes is to adjust the velocity on an individual three components in space and a segment requires three angle
basis by taking into account height at the withers and the effects of measurements to define its orientation. A three-dimensional joint
gravitational acceleration as described above (Alexander, 1977). coordinate system (JCS) is established for the joint based on embed-
After setting the VDU, the target velocity for each horse is ding an anatomically meaningful coordinate system within each
calculated. limb segment comprising the joint. Flexion/extension, adduction/
abduction, and internal/external rotation are usually expressed as
target velocity = VDU (ms/ms) ∗ ( height ∗ g ) motions of the distal segment relative to the fixed proximal segment
(Fig. 2.6). The angles can be described independently of each other,
GRFs are standardized to the subject’s body mass and expressed as which allows for examination of complex coupled motion in a
force per kg body mass (N/kg). Impulses are affected by both mass joint. Angles measured with the JCS are independent of the joint
and height at the withers. The measured impulse (force*stance centers of rotation.
duration) is divided by the force due to body mass (weight*g) One method of expressing three-dimensional joint motions is
multiplied by the adjusted time, and expressed as the impulse in based on a strictly ordered sequence of three rotations, known as
dimensionless units (IDU). The equation to obtain IDU is: Euler angles. In aeronautical terminology these are referred to as
force ∗ stance time pitch, yaw and roll. The drawback to the use of Euler angles is the
IDU (Ns/Ns) = need to specify the order of rotations. In an effort to express joint
weight ∗ g ∗ height/g
motion in a context that is more meaningful to clinicians, Grood
and Suntay (1983) proposed the use of a JCS for the human knee
joint in which the rotations are consistent with the clinical defini-
Kinematic data tions of flexion/extension, abduction/adduction and internal/
Kinematic data include temporal, linear and angular variables. Tem- external rotation. The first axis of rotation (flexion/extension) was
poral data, which describe the stride duration and the limb coordi- attached to the segment proximal to the joint, the third (internal/
nation patterns, are calculated from the frame numbers and the external rotation) axis was attached to the distal segment and the
sampling frequency. Linear variables, which describe stride length, second (adduction/abduction) axis was a floating axis that was
the distances between limb placements, and the flight paths of the mutually perpendicular to the other two axes and was not aligned
body parts, are calculated from the coordinates of the markers with the planes of either segment (Fig. 2.6). For each axis, there is
combined with the calibration information. Angular variables a rotation around the axis and a translation along the axis. All three
describing the rotational motion of the body segments and joints rotations take place at the same time, as in a mechanical linkage,
are calculated from the coordinate data. thus eliminating the need to specify the order of rotations. The
Analyses may be performed in two or three dimensions. Since the method of Grood and Suntay (1983) brought a wider acceptance
horse’s limbs have evolved to move primarily in a sagittal plane, of the value of the JCS in a clinical setting, though it is now under-
most of the useful information is captured by the two-dimensional stood that their method is actually identical to the Euler angle
lateral view, and in many situations the extra effort involved in method.
extracting three-dimensional data is not warranted. However, there When using the JCS method, kinematic analysis is typically per-
are times when knowledge of abduction/adduction or internal/ formed in three steps. The first step is to define a coordinate system
external rotations would be useful, especially during sporting activi- on each bone. Second, the rotation matrix and translation vector
ties and in relation to lameness. relating the GCS with the JCS is obtained from three-dimensional
marker coordinates during motion using a singular value decom-
position method (Soderkvist & Wedin, 1993). This algorithm pro-
duces a rotation matrix and a translation vector that describe the
Sagittal plane analysis rotation and translation of the segmental coordinate system from
Data are collected in a global (or laboratory) coordinate system its neutral position in the stationary file to the orientation in each
(GCS). A common convention is for the z-axis to be vertical and to frame of tracked data. In the final step, matrix equations are used
align the two horizontal axes (y: longitudinal; x: transverse) with to extract the three rotation angles and three translations (Grood &
the force plate axes or with the longitudinal and transverse axes of Suntay, 1983). Relative angular motions (helical angle changes)
the runway. If the horse moves straight along the length of the between the segments may be calculated using a spatial attitude
runway, the y-z plane of the GCS corresponds with the horse’s method (Spoor & Veldpaus, 1980; Woltring, 1994) or Euler angles
sagittal plane. may be calculated (Ramakrishnan & Kadaba, 1991). Software is

37
 2 Measurement techniques for gait analysis

20

10

Add (+)/abd (-) (deg)

0
20 40 60 80 100

-10

-20

Fig 2.7  Adduction and abduction of the metacarpophalangeal joint during

trotting for the segmental coordinate system that was defined by a
standing pose (black line), after +10° (inward) rotation (red line) and after
−10° (outward) rotation (blue line) of the segmental coordinate system of
the proximal phalanx around the longitudinal axis of the bone.

Adduction/abduction
mathematically valid, differences in output affect interpretation of
the results. Clinical conclusions can be upheld or refuted, based on
the same data set, subject to coordinate system definitions (Beard-
sley et al., 2007). Choice of the coordinate system is, therefore,
critical to the outcome of a study.
In equine studies, limb segment axes have been based on ana-
tomical landmarks identified by palpation and/or fluoroscopy
(Khumsap et al., 2004; Clayton et al., 2004, 2007a, 2007b) or using
a template attached to the bones (Chateau et al., 2004, 2006). A
Flexion/extension right-handed coordinate system is used to establish the three axes
around which flexion/extension, abduction/adduction and internal/
external rotation occur. Errors are reduced if the axis aligned with
the longest dimension of the bone is established first.
Kinematic crosstalk is a common problem in using the JCS
method. Flexion/extension angles are robust due to their large
signal to noise ratio and misalignments between the horse or the
plane of motion and the GCS tend to have little effect on flexion/
extension measurements where this is the predominant type of
motion (Ramakrishnan & Kadaba, 1991). However, these misalign-
ments will cause some of the flexion movement to be misinter-
preted as abduction. Figure 2.7 shows the effect of a small (±10°)
change in alignment of the flexion/extension axis on the amount
of motion ascribed to abduction/adduction at the fetlock joint
(Clayton et al., 2007a). Another problem with the JCS methodol-
ogy is gimbal lock, which causes the rotation angles to become
increasingly sensitive to measuring errors when the second rotation
Internal/external approaches ±90° (Woltring, 1994).

Fig 2.6  Three-dimensional joint coordinate system for kinematic analysis.

Flexion/extension and external/internal rotation are expressed in terms of Methods of kinematic analysis
the distal segment rotating relative to the proximal segment around axes
embedded in the proximal and distal segments, respectively. Abduction and
This section will give a brief review of older techniques used in
adduction are relative to the floating axis, which is mutually perpendicular
kinematic data collection, and will then describe the principles of
to the other two axes.
modern motion analysis systems in greater detail.

available to perform these steps or computer code can be accessed

Electrogoniometry
in the public domain. Markers on the model are defined and, from An electrogoniometer or elgon is a device for measuring joint angle
the marker trajectories, joint rotations and translations are solved changes. It consists of a potentiometer attached to two rigid, rotat-
iteratively by global optimization. ing arms that are fixed to the limb with tape or straps, so that the
When using a segment-based or bone-based local coordinate center of the elgon lies over the center of rotation of the joint (Fig.
system, decisions regarding the establishment and orientation of 2.8). A change in joint angle alters the electrical resistance of the
the axes have important effects on the results of the analysis and potentiometer, which is calibrated with a protractor. Permanent
must, therefore, be stated clearly (Capozzo et al., 2005). Both joint records, or goniograms, can be recorded on an oscilloscope and the
kinematics (Beardsley et al., 2007) and net joint moments (Schache data stored for later analysis.
& Baker, 2007) vary significantly according to the coordinate system In horses, electrogoniometry has been used to record joint move-
used and, even though the alternative frames of reference are ments at different gaits in normal and lame horses (Adrian et al.,

38
 Methods of kinematic analysis

A standard camcorder records 30 frames/s in the NTSC format

and 25 frames/s in the PAL format. Each frame consists of two
interlaced video fields recorded 1/60 s (NTSC) or 1/50 s (PAL)
apart. Software displays successive fields sequentially giving an
effective sampling rate of 60 fields/s (NTSC) or 50 fields/s (PAL).
High-speed video cameras are useful for studies of short-duration
events or rapid movements but the lighting conditions become
critical at faster recording speeds. Comparisons of joint displace-
ment data from horses cantering on a treadmill showed minimal
loss of information in terms of angular data when the sampling
rate was reduced from 200 Hz to 50 Hz (Lanovaz, unpublished).
Linford (1994) compared the temporal stride variables in horses
trotting on a treadmill by analyzing the same ten strides with
cameras sampling at 60 Hz and 1000 Hz, respectively. Mean
values for stride duration, stance duration, swing duration, and
breakover did not differ by more than 3.3 ms. Thus, 60 Hz is an
adequate sampling rate for kinematic analysis of many aspects of
equine locomotion, but a large number of strides must be ana-
lyzed to produce a representative mean value for temporal events
of short duration. At gaits faster than a walk, a higher sampling
rate is preferable, especially if the displacement data will be pro-
cessed further.
High-speed video cameras capable of recording at very high frame
rates (up to several thousand frames per second) usually offer only
a few seconds of recording time unless the video is recorded to a
Fig 2.8  Electrogoniometer (elgon) placed over the equine fetlock joint. hard drive in a separate recording unit or computer. These cameras
require very good lighting; the shorter the exposure time, the more
illumination is needed to maintain acceptable image quality. High
speed, high-resolution camcorders are available that are easy and
convenient to use and record for a reasonable duration to a digital
1977), to diagnose obscure lameness, and to evaluate changes in tape or DVD.
joint motion after medical or surgical treatment (Taylor et al., 1966; A number of software packages with varying levels of sophistica-
Ratzlaff et al., 1979). The availability of image-based motion analy- tion are marketed for video-based kinematic analysis. These
sis systems has superseded the use of electrogoniometry. packages may offer marker-less tracking, simultaneous multiple
two-dimensional video capture and stroboscopic imaging. Marker-
less tracking digitizes unique user specified patterns on a frame-
Photographic systems by-frame basis without markers or manual intervention. If the
In the past high-speed cinematography was the most commonly image contains a distinguishable pattern that is visible across several
used optical technique for kinematic analysis, but it was an expen- frames, the system can recognize a point within the pattern and
sive and tedious method that has declined in popularity since the track it automatically for the entire sequence (Fig. 2.9). By simulta-
advent of high-resolution video cameras and optoelectronic motion neously collecting multiple two-dimensional camera views, it is
analysis systems. TrackEye (Image Systems AB, Linköping, Sweden) possible to perform a multi-planar analysis. Images that represent
is an image digitizing system that processes high-speed film to distinct events, such as ground contact, midstance and lift-off, can
provide accurate two-dimensional or three-dimensional data. This be viewed to create stroboscopic files.
system has been used for equine kinematic analysis at Uppsala Some packages, such as Dartfish (Dartfish, Fribourg, Switzer-
University. land), that are designed for use by athletic coaches and trainers are
Videography is now the standard photographic recording and relatively inexpensive and easy to use. Software targeted to the
analysis technique. The availability of video replay is especially equine market includes Equinalysis (Equinalysis, Gwehelog Usk,
useful for sports applications or in a clinical setting. The level of UK), Equine Gait Trax Digital Motion Analysis System (Motion
sophistication ranges from simple visual appraisal through the use Imaging Corp., Simi Valley, CA) and Ontrack (Lameness Solutions,
of interactive software to turnkey systems that autodigitize reflective LLC., Altrincham, UK). Software offering a more complete biome-
markers. Some video systems also offer a manual digitizing option chanical profile that can be used with video data includes the Ariel
that can be used when there are no markers on the subject, for Performance Analysis System (Ariel Dynamics Inc., Trabuco Canyon,
example during competition. Manual digitizing is also a useful CA), SIMI Motion (Simi Reality Motion Systems GmbH, Unter-
option when the autodigitizing system has difficulty differentiating schleissheim, Germany), KinTrak/Orthotrak (Motion Analysis
between markers placed close together, markers that cross each Corp., Santa Rosa, CA) and Vicon MOTUS (Vicon, Los Angeles, CA).
other as the horse moves or markers that are temporarily obscured. Appropriate software should be chosen depending on the level of
It is important to verify marker identifications before accepting the sophistication required and the available budget.
data for further analysis to avoid errors in the raw data being propa-
gated throughout the subsequent analysis.
For single camera, two-dimensional studies, the video camera is
precisely oriented with its axis perpendicular to the plane of inter-
Optoelectronic systems
est. To overcome the limitations of a small field of view, a panning The majority of equine gait labs use digital optical motion capture
camera can be used. For three-dimensional studies, precise camera systems that offer automatic marker identification with real-time
positioning is less important, so long as each marker is visible to at processing of the three-dimensional coordinates to display graphs
least two cameras at all times. If the angle between cameras is small, and stick figures of the motion and other types of calculated
however, it may reduce the accuracy along the axis running toward data during or immediately after the movement occurs. Other
the cameras. sources of analog or digital data from force platforms, EMG systems,

39
 2 Measurement techniques for gait analysis

Fig 2.9  Black lines track the trajectories of a

marker on the horse and a marker on the rider
during a video clip showing a transition from
extended trot to collected trot. Produced using
Dartfish TeamPro software (Dartfish, Fribourg,
Switzerland).

Fig 2.10  Data collection using a Motion Analysis

System in the McPhail Equine Performance
Center at Michigan State University, USA.

dynamometers, foot switches, event recorders, etc., can be synchro- an electronic shutter synchronized with the strobed lights will
nized with the system. Kinematic data may be recorded using active remove motion blur.
markers (markers that emit a signal) or passive markers (markers Software packages for kinematic analysis offer a simple and pow-
that detect or reflect a signal). Most of the equine gait laboratories erful interface that facilitates set up, calibration, motion capture
currently use optoelectronic systems based on passive markers, such (real-time and for post processing), editing and saving data in a
as the Motion Analysis System (Motion Analysis Corp., Santa Rosa, chosen format. The software delivers six-degrees-of-freedom data
CA), Vicon MX (Vicon, Oxford, UK) or ProReflex MCU (Qualysis for the full range of targets using reflective markers and high speed,
Inc., Glastonbury, CT). Figure 2.10 shows data collection in progress high-resolution cameras. In this way, the system can provide the
using a Motion Analysis System in the Mary Anne McPhail Equine highest order positional and angular accuracy. The interface syn-
Performance Center at Michigan State University. chronizes with other analog and digital devices, such as video
In choosing cameras for an optoelectronic system, there is a cameras, force plate, pressure mat and EMG. The on-screen display
balance between range, speed, size, performance and cost. The may allow simultaneous display of several panels showing video, a
current industry standard is to use digital camera systems that have three-dimensional stick figure viewed from any perspective, together
a ring of strobed lights (visible red, infra-red) surrounding the lens; with graphical outputs of the data. A considerable volume of litera-
the strobe frequency corresponds with the sampling rate. Desirable ture in the field of equine locomotion has been based on the use
features include a combination of high resolution (high pixel of optoelectronic systems with passive markers.
count) and a high frame rate, though at higher sampling rates the Motion analysis systems based on active strobed markers offer
resolution may be reduced or the image may be displayed in split- automated marker identification by sequencing the temporal output
screen mode. Cameras with more pixels allow the use of larger of different markers, which avoids marker confusion or swapped
capture volumes, smaller markers and more complex marker sets, trajectories. These systems include Optotrak Certus (NDI, Waterloo,
all of which are advantageous when working with horses. The ON) and Codamotion (Charnwood Dynamics Ltd., Rothley, UK).
advantages of digital technology over its analog predecessors are An advantage to Codamotion is that its precalibrated sensor units
that there is no degradation of the signal over distance, less noise, are portable and can be set up in any location to measure three-
and no re-sampling of data on another piece of electronics. The dimensional coordinates without the need for a calibration process.
signal processing is embedded in the camera and the signal goes
directly to the tracking computer via an ethernet connection. By
streamlining the system of motion capture from camera to com-
puter, the amount of hardware is reduced and there is less potential
Electromagnetic systems
for equipment problems. Regardless of the type of camera used, a Electromagnetic systems, such as Flock of Birds (Ascension Tech-
high-quality lens will improve the accuracy of the entire system and nologies, Burlington, VT) are used as three-dimensional tracking

40
 Kinetic analysis

A B

Fig 2.11  Zebris system in use at the National Veterinary School of Alfort, France. Triads of ultrasound microphones fixed rigidly to the limb segments (left)
are located relative to a fixed system of three ultrasound transmitters as data are collected from a horse moving over ground (right).
Photograph courtesy of Dr. H. Chateau.

devices for applications that include real-time visualization, and walking and trotting over ground and in analysis of the stance and
target acquisition. The magnetic tracking has accuracy comparable swing phases of horses walking and trotting on a treadmill (Chateau
with an optical motion system (Hassan et al., 2007). Drawbacks to et al., 2004, 2006).
the use of an electromagnetic system include the problem of metal
interference, the relative heaviness of the receivers, the presence of
wires and the small data capture volume. Kinetic analysis

Kinetic analysis measures locomotor forces, both external and inter-

Ultrasonographic system nal to the body. Forces developed by muscles are transformed into
An ultrasonographic system for real-time, three-dimensional gait rotations of the limb segments that ultimately produce movement.
analysis (Zebris CMS-HS, Medizintechnik Gmbh, Isny, Germany) is The forces between the hoof and the ground during locomotion can
based on temporal delay of ultrasound signals. The system uses be recorded using a force plate (e.g. Pratt & O’Connor, 1976; See-
marker triads consisting of three small ultrasound microphones herman et al., 1987; Merkens et al., 1988, 1993a, 1993b; Dutto
(Fig. 2.11) that operate sequentially and are mounted at a predeter- et al., 2004b) or a force shoe (Björck 1958; Frederick & Henderson
mined distance from each other (Fig. 2.11). The position of the 1970; Hjertén & Drevemo, 1987; Barrey, 1990; Ratzlaff et al., 1990,
markers is determined relative to a fixed system of three ultrasound 1994; Keg et al., 1992; Roepstorff & Drevemo, 1993; Kai et al. 2000;
transmitters and the x, y, z coordinates of the markers are deter- Rollot et al., 2004; Roland et al., 2005). Transmission of forces and
mined by triangulation, based on the delay in the ultrasound accelerations through the body are recorded by strain transducers,
pulses. ultrasonic transducers, accelerometers, gyroscopes and magnetom-
In standard laboratory conditions, the precision of the Zebris eters attached directly to the segment.
system was measured as ±0.14 mm for linear measurements and
±0.16° for angular measurements, but the precision decreased if the
distance between the transmitters and the microphones exceeded
Ground reaction force
1.5 m. This restricts the size of the data collection volume and does During the stance phase of the stride, the hoof exerts a force against
not permit recording of data for a complete stride during over the ground and, according to Newton’s third law of motion, the
ground locomotion (Chateau et al., 2003). Although the maximum ground exerts a reaction force against the hoof that is equal in
sampling frequency is 100 Hz, the relationship between sampling magnitude and acts in the opposite direction. This ground reaction
rate, measuring distance and time between ultrasonic pulses is such force (GRF) is fully described by its magnitude, direction and point
that a sampling frequency of 50–60 Hz is practical for equine of application. To aid in understanding the effects of the GRF, the
studies (Chateau et al., 2003). Ultrasound reflections can induce three-dimensional force vector may be resolved into components
systematic errors (about 0.2°) and impair the repeatability of the acting in the vertical, longitudinal (craniocaudal), and transverse
measurements. Wind decreases the precision of the measurements (mediolateral) directions. When the frame of reference changes, for
in proportion with its speed and temperature also has an effect, so example if the horse moves on an incline or decline, the compo-
it is necessary to configure temperature in the software. This system nents may be aligned with the ground rather than the global refer-
has been used in precision analysis of the stance phase of horses ence frame (Dutto et al., 2004b). Measurement devices may provide

41
 2 Measurement techniques for gait analysis

12 12

10 10

8 8
Force (N/kg)

Force (N/kg)
6 6

4 4

2 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Percent of stance Percent of stance

1.5 1.5
1.0 1.0
0.5 0.5
0.0 0.0
Force (N/kg)

Force (N/kg)
-0.5 -0.5
-1.0 -1.0
-1.5 -1.5
-2.0 -2.0
-2.5 -2.5
0 20 40 60 80 100 0 20 40 60 80 100
Percent of stance Percent of stance

0.6 0.6
0.4 0.4
0.2 0.2
Force (N/kg)

Force (N/kg)

0.0 0.0
-0.2 -0.2
-0.4 -0.4
-0.6 -0.6
-0.8 -0.8
0 20 40 60 80 100 0 20 40 60 80 100
Percent of stance Percent of stance

Fig 2.12  Vertical (above), longitudinal (center) and transverse (below) components of the ground reaction force for a horse at the walk (left panel) and at
the trot (right panel). For the longitudinal force the cranial direction is positive and for the transverse force the medial direction is positive. Forelimb (red line);
hind limb (green line).

complete three-dimensional data or a more limited range of GRF result of friction that prevents the hoof slipping forward. Later in
components, so the method of measurement should be chosen in the stance phase, it changes to a propulsive (accelerating) force. The
accordance with the goals of the study. direction of the horse’s motion across the force plate determines
Standard GRF patterns have been developed for Dutch Warm- whether acceleration or deceleration is recorded as positive. Soft-
bloods at a walk (Merkens et al., 1988), trot (Merkens et al., 1993a) ware correction is applied to standardize the sign convention. The
and canter (Merkens et al., 1993b). Since there are significant dif- transverse force, which increases during sideways or turning move-
ferences between breeds (Back et al., 2007), the Dutch Warmblood ments, is small in magnitude when moving in a straight line and is
data must be adapted to other breeds by incorporating appropriate directed medially. The left to right values recorded by the force plate
parameters and weighting factors into the formulae used to develop can be converted during post-processing to represent medial and
the standard patterns. For example, when forces are normalized to lateral values for each limb. During turning the transverse force acts
body weight, Quarter Horses have lower vertical GRFs than Dutch toward the center of the turn or toward the direction of lateral
Warmbloods trotting at a similar velocity (Back et al., 2007). motion.
Figure 2.12 shows the three force components (vertical, longitu- For a walking horse, the peak vertical GRF has a magnitude of the
dinal, transverse) during the stance phase of a forelimb at a walk order of 60% body weight at walk and 90% body weight at a moder-
and trot. The vertical force, which represents the anti-gravity support ate speed trot (Fig. 2.12). At slower walking velocities the vertical
function of the limb, is always positive. The longitudinal force, force trace tends to be biphasic with the second peak being higher
which provides acceleration and deceleration, has negative and in the forelimbs and the first peak being higher in the hind limbs.
positive phases. In the early part of the stance phase, the longitudi- During walking, peak values of the vertical and longitudinal forces
nal force brakes (decelerates) the horse’s forward movement as a in the forelimb increase with velocity, but the corresponding

42
 Kinetic analysis

impulses decrease as a consequence of a reduced stance duration 14

(Khumsap et al., 2002). In the hind limb, however, an increase in
walking velocity affects only the first peak in vertical force (Khumsap 12
et al., 2001). In the trot, sharp spikes usually occur immediately
10
after initial ground contact during the period of impulsive loading

Force (N/kg)
(impact phase). The trace then rises smoothly to peak around the 8
time the limb is at its midstance position, which is marked by the
cannon segment being vertical, after which it decreases to lift-off 6
(Fig. 2.12). The peak value of the longitudinal force is 10–15% of
the horse’s body weight at the walk and trot, with marked spiking 4
occurring during the impact phase at trot. The transverse force is 2
much smaller in magnitude, around 2% body weight at the trot
(Fig. 2.12). The center of pressure is located under the middle of 0
the hoof during most of the stance phase, moving rapidly toward 0 20 40 60 80 100
the toe at the start of breakover. Time (% stance)
Values representing the peak forces and their times of occurrence
are extracted from the force history. Impulses are determined by 1.0
time integration of the force curves. A procedure that combined over
90 numbers describing the peak amplitudes, their times of occur- 0.5
rence, and the impulses has been described as the H(orse) INDEX
(Schamhardt & Merkens, 1987). This method is valid but has some 0.0
20 40 60 80 100

Force (N/kg)
drawbacks in that the variables used to calculate the index are
selected by the user and are essentially dependent on the shape of -0.5
the signal. Moreover, it does not take account of the real pattern of
the curve, which can be accomplished by different techniques that -1.0
are more suitable for comparison of curve patterns.
The stance durations, GRF amplitudes and impulses are sym- -1.5
metrical in sound horses at the walk (Merkens et al., 1986, 1988)
and trot (Seeherman et al., 1987; Merkens et al., 1993a; Linford -2.0
1994; Weishaupt et al., 2004a). Analysis of ground reaction forces
Time (% stance)
is a reliable way to quantify lameness, with peak vertical force and
vertical impulse being the variables with highest limb specificity Fig 2.13  Vertical (above) and longitudinal (below) ground reaction forces
and sensitivity in grading lameness (Weishaupt, 2004b, 2006a). An of forelimbs in sound condition (mean value of left and right forelimbs in
example of the changes in ground reaction forces in superficial blue) and after induction of superficial digital flexor tendonitis in one
digital flexor tendonitis is shown in Figure 2.13. Lameness models forelimb (lame, forelimb in green and compensating forelimbs in red shown
include pressure on the hoof sole (Merkens & Schamhardt, 1988); separately).
collagenase-induced tendinitis in the flexor tendons (Clayton et al., From Clayton et al. (2000) Am. J. Vet. Res. 61, 191–196 with permission of American
2000a) or desmitis of the suspensory ligament (Keg et al., 1992); Veterinary Medical Association, http://avmajournals.avma.org/loi/ajvr.
and surgical creation of a cartilage defect (Morris & Seeherman,
1987; McIlwraith et al., 2011). In addition to its value for detecting
lameness, the force plate is a sensitive tool for measuring the a non-slip material (Fig. 2.15). When a horse steps on the plate, the
response of lame horses to diagnostic anesthesia (Keg et al., 1992) force is detected by transducers at its corners, and is converted to
or to therapeutic intervention (Gingerich et al., 1979). an electrical signal that is amplified and recorded. Two types of force
Ground reaction force has also been used for postural sway analy- transducers are used: strain gauge and piezoelectric. Strain gauges
sis, which projects movements of the center of pressure onto a have a linear response, are relatively insensitive to temperature, and
horizontal plane. A stabilogram (Fig. 2.14) is a graphic representa- can be miniaturized. The disadvantage is their low sensitivity. By
tion of the motion of the center of mass in craniocaudal and medio- comparison, piezoelectric transducers have high responsiveness and
lateral directions over a period of time. Techniques for postural sway can be miniaturized but they are sensitive to temperature and
analysis has been described in horses (Clayton et al., 2003a; 2012) humidity and are not amenable to static calibration. Variables
and disturbances in balance after administration of a sedative have measured by the force plate include stance duration, magnitude of
been quantified (Bialski et al., 2004). Postural sway analysis also the vertical, longitudinal (horizontal craniocaudal) and transverse
shows promise as a technique for detection of neurological diseases (horizontal mediolateral) forces, time when the peak forces occur,
(Clayton et al., 1999). the impulses (areas under the force–time curves), and point of
application of the force (center of pressure or point of zero moment).
In selecting a force plate for equine use, it is important to choose
Normalization one that has a linear response over an appropriate range of forces,
taking into account the weight of the horses and the gaits and
GRFs vary with body mass (Barr et al., 1995). Comparisons between
speeds to be studied. The dimensions of the plate should maximize
horses are facilitated by normalizing the force traces to the horse’s
the chance of getting a good strike from one fore hoof followed by
weight, so they are expressed in Newtons/kilogram body weight (N/
the hind hoof on the same side at the selected gait. If two hooves
kg). GRFs may also be standardized to the duration of the stance
strike the plate simultaneously, it is not possible to separate their
phase, which allows comparison of the forces occurring at a percent-
effects. Shorter force plates (60–90 cm) are ideal for collecting data
age of stance duration in strides with different durations.
at the walk, but a length of 90–120 cm is preferable for use at the
faster gaits. Width of the plate is not generally a limiting factor:
50–60 cm is adequate. Depending on the dimensions of the force
Force plate plate, it is typical for a good strike to be recorded for every two to six
A force plate is a steel plate that can be mounted in a runway or passes at the walk, trot and canter (Niki et al., 1982; Merkens et al.,
recessed into the ground. For equine use, it is usually covered with 1986, 1993a, 1993b). During jumping, the obstacle is moved

43
 2 Measurement techniques for gait analysis

relative to the stationary force plate, to increase the likelihood of

6
getting a good strike with a particular limb at lift-off or landing
(Schamhardt et al., 1993b). The horse should move parallel with
the long axis of the force plate to avoid cross talk between the
4 horizontal transducers. Companies that manufacture force plates
suitable for equine use include AMTI (Watertown, MA), Bertec
Corporation (Columbus, OH) and Kistler Instruments Corp.
(Amherst, NY).
2
Installation and calibration of the force plate are critically impor-
Longitudinal sway (mm)

tant to the quality of the data collected. The plate should be embed-
ded in a concrete pit to isolate it from surrounding vibrations and
the supporting surface must be absolutely level to avoid cross talk
-3 -2 -1 0 1 2 3 between the vertical and horizontal channels. Before recording data,
the calibration should be checked by placing a known weight,
similar in magnitude to the loads that will be applied during
-2 normal use, at different locations on the force plate. The same verti-
cal force should be recorded independent of location, and the
position of the center of pressure should match the actual location
-4 of the load.
Simple procedures for checking force plate calibration on a daily
basis have been described (e.g. Lewis et al., 2007). Since GRFs vary
with velocity (McLaughlin et al., 1996; McGuigan & Wilson, 2003),
-6 measurement and control of velocity are important in studies of
GRFs. Provided the horse’s velocity over the force plate is main-
Transverse sway (mm)
tained within a narrow range, the GRFs are consistent and repeat-
Fig 2.14  The stabilogram is a graphic representation of the motion of the able between strides, with analysis of five strides being sufficient to
horse’s center of mass projected onto a horizontal plane. In the graph provide representative data (Merkens et al., 1986). Methods of mea-
shown here, the line traces the motion of the center of mass during a 20 s suring the horse’s average velocity over the plate include the use of
recording period at 1000 Hz. The 0,0 location, which is the centroid of all timing lights that record the time taken to cover a known distance
the points, is located at the intersection of the transverse and longitudinal or dedicated kinematic software. Sensors for a simple infrared
axes. The head is toward the top of the figure and the horse’s left is on the timing device that measures the time elapsing as the horse passes
left. Note the different scales on the vertical and horizontal axes. through the data collection volume can be purchased inexpensively.
Alternatively, software can be written to track a position marker (as
described above) on the horse’s torso and calculate average speed
and speed variation during the trial. Depending on the objectives
of the study, data from runs that fall outside a predetermined range
of velocities may be discarded.
Bad trials occur when the horse fails to strike the force plate, when
the hoof contacts the edge of the plate, when more than one hoof
is on the plate simultaneously, and when the horse’s body axis or
direction of movement are not aligned with the axes of the force
plate. These problems may be recognized by the force traces having
an unusual shape or magnitude or failing to return to the baseline
between individual limb contacts. It is helpful to use kinematic data
to verify which hoof is on the force plate and its precise location,
and also to check the alignment of the horse and the direction of
movement.
Variables with low variability require fewer animals, trials and
days to obtain accurate data while maintaining adequate statistical
power. Inter-day reliability of force plate data over three consecutive
days is high for stance duration, peak vertical force and vertical
impulse (<10% variability of the mean), while longitudinal force
peaks and impulses have larger variation (>20% variability of the
mean) (Lynch et al., 2005). In lame horses, peak vertical force and
vertical impulse have the lowest between- and within-horse coeffi-
cients of variation and the highest correlations with subjective lame-
ness grade. Peak vertical force also has high sensitivity and specificity
for lameness classification (Ishihara et al., 2005).
Although force plates for equine use are usually covered with
rubber matting, it is possible to place different surface materials on
top of the force plate. For example, a steel-reinforced concrete top-
plate has been used to simulate the surface of a road in a study of
energy dissipation during hoof slippage at the start of stance
(Wilson & Pardoe, 2001).
At the University of Zurich, a force plate suitable for equine use
has been embedded in a treadmill (Fig. 2.16) to measure the vertical
forces of all four limbs over an unlimited number of steps at any
Fig 2.15  Horse stepping on force plate during data collection. gait (Weishaupt et al., 2002). The force measuring system consists

44
 Kinetic analysis

Angular encoders
X Rubber string

Y Z

Force sensor Treadmill platform

R1 R2 R3 R4 R5 R6 R7 R8 R9

R1 R2 R3 R4 R5 R6 R7 R8 R9

No sensor Treadmill frame

0.5m

Traction belt Shock absorber Treadmill platform Treadmill frame

Fig 2.17  Horse being ridden on a treadmill instrumented for measurement
R1 R2 R3 R4 R5 R6 R7 R8 R9 of vertical ground reaction forces at University of Zurich, Switzerland. The
hooves are attached by rubber bands to goniometers that triangulate the
hoof positions.
Fig 2.16  Schematic illustrations representing the overhead (above) and Photograph courtesy of Dr. Michael Weishaupt.
cross-sectional lateral (below) views of the instrumented treadmill for use in
determination of vertical ground reaction forces in horses at the University
of Zurich. There are 18 force sensors and four triangulation units for Both strain gauge and piezoelectric transducers have been used
determining coordinates of each hoof. in the construction of equine force shoes, with the same advantages
Diagram courtesy of Dr. Michael Weishaupt. and disadvantages as when they are used in force plates. An addi-
tional practical consideration in the force shoe application is that
the output of piezoelectric transducers can change in response to
movement of the lead wires.
An early force shoe (Björck, 1958) used strain gauges to measure
of 16 piezoelectric force transducers mounted between the treadmill vertical and horizontal forces in draft horses. The force patterns
frame and the supporting steel plate over which the belt moves. resembled those obtained by other methods but the size of the shoe
Each transducer measures the vertical force at the corresponding (height: 24 mm; weight: 1800 g) limited its use. Several years later,
bearing of the supporting plate. Transfer coefficients have been a similar shoe weighing 2300 g was used to measure vertical and
determined and tabulated for each of the 16 transducers for each horizontal components of the GRF at trot (Hjertén & Drevemo,
square centimeter of the treadmill surface by the application of a 1987). A different device had three force sensors sandwiched
test force. The coordinates of a hoof on the treadmill surface are between a base plate nailed to the hoof wall and a ground plate
calculated by triangulation based on angle values derived from two that was attached to the base plate, thus preloading the force sensors
electrogoniometers. For each sampling instant, a set of 16 linear (Frederick & Henderson, 1970). It was used to measure vertical
equations are formulated containing the four unknown hoof forces, forces at walk, trot and gallop.
the four x–y coordinates of the hoof force application, their corre- A much lighter (200–300 g) force shoe based on three strain
sponding transfer coefficients and the 16 forces from the sensors, gauge measuring units, one at the toe and one at each quarter, was
from which the individual hoof forces are extracted. This treadmill developed and used to measure vertical, longitudinal and transverse
has been used to determine GRFs and interlimb temporal coordina- forces (Roepstorff & Drevemo, 1993). The output correlated well
tion data of clinically sound Warmblood horses at the trot with that of a force plate when all three measuring units were in
(Weishaupt et al., 2004a), the characteristics of lameness in the contact with the ground, but when one or more units lost contact
hind limbs (Weishaupt et al., 2004b) and in the forelimbs with the ground, for example during breakover, the correlation
(Weishaupt et al., 2006a) and the effect of head and neck position between force shoe and force plate signals deteriorated. This shoe
on vertical ground reaction forces in dressage horses ridden at walk was used to compare GRFs on treadmill belts with different com-
and trot (e.g. Weishaupt et al., 2006b) (Fig. 2.17). positions (Roepstorff et al., 1994).
Kai et al. (2000) developed a force shoe based on four strain
gauges mounted on an aluminum plate and supported by a stain-
Force shoes less steel plate in such a way that vertical forces were transmitted by
A force shoe is a portable device that measures one or more com- the load cells, whereas shear forces were borne, not by the load cells,
ponents of the ground reaction force during a large number of but by bolts that prevented horizontal movement of the steel plate
successive stance phases, thus overcoming a limitation of the force relative to the underlying aluminum plate. Three accelerometers
plate. Force shoes also have the advantages of being amenable to were integrated into the shoe to measure three-dimensional impact
use on different surface types, and being able to collect data from accelerations. The total weight of the assembly was 480 g. The shape
more than one limb at a time. The force-measuring device may be of the vertical GRF curves during walking, trotting and cantering on
attached directly to the hoof, sandwiched between a base plate and a treadmill resembled force plate recordings.
a shoe, or placed inside a boot attached to the hoof. Force shoes A force shoe developed specifically for characterizing the GRF and
are technically difficult to construct and the weight of the shoe moment vectors during galloping (Fig. 2.18) (Roland et al., 2005)
may affect limb kinematics and forces. Several researchers have had two parallel plates with three rectangular strain-gauged sensing
used force shoes experimentally, but none is currently marketed posts sandwiched between them: two at the widest part of the shoe
commercially. and one at the toe. The size and contour of the solar surface of the

45
 2 Measurement techniques for gait analysis

Ridge around
Hoof Plate z-axis the hoof wall

Sensor 2

y-axis
A Ground Plate

0.5 mm radius

Strain gage rosette for

measuring normal strain
created by force
applied in the z direction Strain gage rosettes for
measuring shear strains
created by forces applied
in the x and y directions

Fig 2.19  Force shoe attached to hoof and viewed from the solar aspect.
HL, lateral heel; QL, lateral quarter; QM, medial quarter; HM, medial heel; and
B Flange for mating with ground plate global anatomical frame (Xhoof, Yhoof).
Reprinted from Rollot, Y., Lecuyer, E., Chateau, H., Crevier-Denoix, N., 2004. Development of a
Fig 2.18  Above: dynamometric horseshoe viewed from the toe with
coordinate axes superimposed. The x-axis is perpendicular to the plane of 3D model of the equine distal forelimb and of a GRF shoe for noninvasive determination of
the photograph. The elevation of the x–y plane coincides with the bottom in vivo tendon and ligament loads and strains. Equine Vet. J. 36, 677–682, with permission
plane of the ground plate. The location of the y–z plane is palmar from the from the Equine Veterinary Journal.
center of sensor 2, two-thirds of the distance from sensor 2 to a line
between the other two sensors (not visible). Below: diagram indicating 15 kN. The total weight of the assembly was around 700 g. When
strain gage placement on the sensors as viewed from below. Opposite compared with data from a force plate, the force shoe represented
sensor faces have the same gage pattern. the shape of the GRF curve but the amplitude was reduced, which
Reprinted from Roland, E.S., Hull, M.L., Stover, S.M., 2005. Design and demonstration of a was explained by the fact that the screws used to sandwich the
dynamometric horseshoe for measuring ground reaction loads of horses during racing sensors were supporting some of the load. Another issue was that
conditions. J. Biomech. 38 (10), 2102–2112, with permission from Elsevier. rotation of the hoof and shoe relative to the ground, especially
during contact and breakover, resulted in crosstalk between the
longitudinal and vertical forces. On sand, weight bearing over the
base plate matched the size and shape of the solar surface of a Thor- frog and sole reduced the force recorded by the instrumented shoe
oughbred hoof. The data were measured in a local coordinate system relative to the actual GRF.
aligned with the hoof, whereas force plate measurements are in a A force shoe with four transducers located on the medial and
global coordinate system fixed to the ground. When the hoof was lateral sides of the toe and quarters, was integrated into the bottom
flat on the ground, the three-dimensional GRF patterns and ampli- of an easy boot and used to measure the vertical GRF component
tudes agreed well with previously reported force plate data but there at walk and trot (Barrey, 1990). This shoe was used to compare
were differences at the start and end of stance when the hoof was different track surfaces (Barrey et al., 1991).
not flat on the ground. The entire shoe weighed 860 g, which is Although a force shoe would be an ideal method of measuring
likely to affect swing phase kinematics due to the added inertia. GRFs, there are significant technical difficulties in constructing one
The first equine force shoe based on piezoelectric sensors was that is accurate and reliable throughout the equine stance phase.
developed by researchers at Washington State University. The first Working within these limitations, some useful data have been
prototype had a single piezoelectric transducer in a housing over generated.
the frog. A later version that had three piezoelectric transducers
located at the medial heel, the lateral heel and the toe gave a better Indirect methods of GRF measurement
correlation with simultaneous force plate recordings. It was used
primarily in studies of galloping Thoroughbreds (e.g. Ratzlaff et al., An alternative method of determining GRFs is by calculation rather
1990), and provided unique data when a horse wearing the shoe than direct measurement. There are different methods of GRF esti-
sustained a rupture of the distal sesamoidean ligaments while gal- mation, each of which has advantages and disadvantages. One
loping on a training track (Ratzlaff et al., 1994). method of estimating vertical GRF is based on the spring-like prop-
Another design used four three-axis piezoelectric load cells erties of the distal forelimb, which implies that vertical loading of
located medially and laterally at the quarters and heels (Fig. 2.19) the limb is directly related to the distance from elbow to hoof and
(Rollot et al., 2004). In this case, the load cells were sandwiched also to fetlock angle, which would allow the vertical GRF to be
between a base plate that was nailed to the hoof and a shoe that calculated from limb kinematics (McGuigan & Wilson, 2003). This
was attached to the base plate using four screws. These screws were method likens the limb to a strain gauge; the relationship between
tightened to preload the load cells with a compressive force of about limb force and elbow-hoof distance or fetlock angle is calibrated

46
 Tissue strain

using kinematic and force data that are collected synchronously. Inverse dynamic analysis has been applied in horses to evaluate
Peak vertical force can then be estimated from kinematics. A further forelimb net joint moments and powers at walk (Clayton et al.,
step involves calibrating the forces indirectly (and without the need 2000b, 2001) and trot (Clayton et al., 1998; Lanovaz et al., 1999)
for force plate analysis) by estimating limb forces from duty factor, and during jumping (Dutto et al., 2004a; Bobbert & Santamaría,
which is the percentage of stride that a limb spends in the stance 2005). The effects of lameness on net joint moments and powers
phase. A prerequisite of this method is a knowledge of the distribu- have also been evaluated (e.g. Clayton et al., 2000c; Khumsap et al.,
tion of the total ground reaction force between the individual limbs, 2003).
which is gained through direct force measurements. The results of
a study that used hoof-mounted accelerometers to determine duty
factor and then estimate peak vertical forces yielded an overestima- Tissue strain
tion of 13% at walk, an underestimation of 3% at trot, an under-
estimation of 16% for the trailing limb at canter and an Body tissues deform or strain with respect to their preferred shape
overestimation of 19% for the leading limb at canter (Witte et al., as a consequence of applied external forces competing with the
2004). The substantial errors at canter were attributed to the differ- microscopic internal forces that hold the tissue together. When a
ent functions performed by the trailing and leading limbs in this tensile force is applied to a solid material, it causes the length to
asymmetrical gait. increase, whereas a compressive force causes the length to decrease.
Bobbert et al. (2007) used a slightly different approach to calcu- A bending force causes both increases and decreases in length in
late force-time histories for the individual limbs based on the different parts of the tissue. Stress is the force per unit area; strain
assumption that they operated as linear springs. Kinematic data is the change in length expressed as a percentage of the original
were used to calculate the acceleration of the horse’s center of mass length.
and the magnitude and direction of the total GRF vector. Angular In ideally linearly elastic materials, deformation is proportional
momentum of the horse, also calculated from the kinematic data, to the applied force and the material restores its original shape as
was combined with the GRF vector to determine the moment arms soon as the deforming force is removed. Deformation of the mate-
of the total GRF relative to the two supporting hooves during rial, usually expressed in terms of strain (ε), is defined as:
bipedal support (Bobbert & Santamaria, 2005). The moment arm
ratio between fore and hind hooves indicates how the total GRF is l1 − l0
ε=
distributed between the supporting limbs, so the individual limb l0
forces can be calculated. The forces can be calibrated against distal
where l0 is the original (or resting) length, l1 is the length after
limb length or fetlock angle, allowing force–time data for the limb
deformation, and l1−l0 represents the change in length. Usually, the
can be estimated based on the time history of distal limb length or
resting length is defined as the length at zero loading force. Because
fetlock angle. Compared with values measured by a force plate, peak
strain is a relative measure, it has no units. It is expressed as a frac-
vertical GRFs calculated by this method were quite accurate (Bobbert
tion or as percentage strain.
et al., 2007).

Strain transducers
Inverse dynamic analysis Strain transducers measure the amount of strain based on electrical
Inverse dynamic analysis (IDA) uses kinematic data, GRF data and resistance (liquid metal strain gauge, buckle transducer, implant-
knowledge of segmental morphology and inertial properties to cal- able force transducer and pressure transducer), variation of mag-
culate internal forces within the limb that cannot be measured netic field (Hall effect transducer), or variation of light flow (optic
directly (Winter, 1990). The net joint moment describes the net fiber). Ideally, transducers should be reliable, easy to implant,
torque across a joint produced by the soft tissues (muscle, tendon, induce minimal distortion or length change in the tissue and not
ligament and joint capsule). Net joint power is a measure of the net cause a significant inflammatory reaction. Implantation requires
mechanical work done across a joint; it is the product of the net surgery using local or general anesthesia. A drawback is that the
joint moment and joint angular velocity. If the net joint moment time required to recover from anesthesia delays the start of data
has the same polarity as the joint angular velocity, then the power collection, which is a relevant concern when using transducers that
is positive (power generation) indicating that the muscles are per- have a limited lifespan or that induce the development of fibrous
forming positive (concentric) work, in which the muscle fibers healing reactions (Ravary et al., 2004). Since strain gauges are small
shorten as they generate tension. If the net joint moment has the devices, they collect data from a limited area that is not necessarily
opposite polarity to the joint angular velocity, the power is negative representative of the total strain or load of the entire structure.
(power absorption) and the muscles perform negative (eccentric) The most commonly used type of strain gauge changes its electri-
work, in which the muscle fibers lengthen as they generate tension. cal resistance in response to deformation in a certain direction; the
Power absorption occurs when muscles act to restrain the rate of change in resistance is converted to a voltage output that is propor-
joint motion in opposition to gravity or some other force. Net work tional to the strain. Voltage outputs are stored for later processing
performed over a period of time is calculated by mathematical by a computer. A combination of three strain gauges stacked at 45°
integration of the power curve with time. angles to each other forms a rosette gauge capable of measuring
Input into the inverse dynamic model comes from synchronized three-dimensional strains.
kinematic and GRF data, which are combined with morphometric
information (segment mass and center of mass) and segment iner-
tial parameters. A link segment model of the limb is constructed,
Measuring strains in hard tissues
with each limb segment represented as a solid bar with its center Hard tissues, such as bone, deform slightly in multiple directions
of mass located relative to the marker coordinates that define the as a result of the combined effects of weight bearing, tension in the
segment. An inverse dynamic solution described by Winter (1990) muscles and tendons, and inertial effects due to acceleration of the
and adapted for use in horses (Colborne et al., 1997) is then used limb. Rosette (three-dimensional) gauges, bonded to a bone surface
to calculate the sagittal plane net joint moments at each joint. Net using a thin layer of cyanoacrylate glue, deform with the surface to
joint moment and net joint power can be normalized to horse mass provide information about the compressive and tensile forces
and expressed per kg body weight (Nm/kg, W/kg, respectively). For (Lanyon, 1976). The best sites for attaching strain gauges to bones
further details see Chapter 19. are in areas where the bone lies subcutaneously, so soft tissue

47
 2 Measurement techniques for gait analysis

Fig 2.20  Strain gage attached to equine third metacarpal with wires
attached. Fig 2.21  Strain gage attached to the hoof wall.
Photograph courtesy of Dr. H. Davies. Photograph courtesy of Dr. J. Thomason.

trauma during surgery is minimized. The bone surface is prepared of loading conditions (e.g. Thomason et al., 1992, 2001) and after
by removing the periosteum and drying the underlying bone before application of therapeutic shoes (Hansen et al. 2005).
bonding the strain gauge to the bone using cyanoacrylate adhesive
(Fig. 2.20). The wires exit the skin through a separate incision. It is
important to shield the wires from movement and trauma, since
Measuring strains in tendons and ligaments
damage to or loosening of the wires is the most frequent reason for The use of strain and force transducers in biomechanical studies of
failure of the gauges. During data collection, the gauges deform as tendon and ligament has been reviewed by Ravary et al. (2004).
if they were part of the bone surface. The resulting electrical signal Forces can be estimated (e.g. van den Bogert, 1989) or measured
is amplified and transmitted to a data recorder or computer for using implanted transducers. Regardless of the type of transducer
storage. used, implantation within a tendon or ligament causes tissue
A practical problem in quantifying bone strain is that the resting damage, which may affect the signal output and limit the number
length of bone is difficult to determine. When the horse is standing of measurements that can be made.
quietly with the limb lifted from the ground, the loading may be The long tendons in the lower limb of the horse can be consid-
assumed to be small. However, the effects of muscular contraction ered as more or less elastic, homogeneous cables. When loaded,
cannot be excluded completely, and the influence of gravity may their length increases. However, strain in tendons is not as well
also affect the zero-strain determination. Software has been devel- defined as in bones. An unloaded tendon shrinks in length and the
oped to calculate a ‘zero-strain compensation’ for in vivo strain tendon fibers become wrinkled or crimped. When elongated, the
gauge data of horses at the walk, using the assumption that crimp in the fibers is first straightened out, after which the fibers
strain is minimal in the middle of the swing phase when the limb are stretched elastically up to a point, beyond which permanent
is moving forward with an almost constant velocity (Schamhardt & elongation occurs.
Hartman, 1982). The load–elongation curve for a tendon has a ‘toe’ region, which
Surface strain is a consequence of the forces loading the bone. is characterized by having a large elongation for a small load. This
However, the relationship between surface strain and load is com- region represents straightening of the crimp in the collagen fibers.
plicated, especially in non-homogeneous, non-linear, viscoelastic As loading increases beyond the ‘toe’ region, there is a linear rela-
structures such as bones (Rybicki et al., 1974). Roszek et al. (1993) tionship between load and elongation in the elastic region until the
presented an elegant technique to quantify the loading forces from yield point, which occurs around 10–12% strain. Beyond the yield
a post-mortem calibration using multiple strain gauges and known point permanent elongation results as the tendon fibers begin to
bending and torsional loading forces. Without this kind of calibra- rupture. A problem in measuring tendon strain lies in defining the
tion, however, bone strain recordings are a valuable, but qualitative, initial length of the tendon and the position of zero load, which
estimate of bone loading. affects the magnitude of the strains recorded throughout the physi-
By using three or four strain rosettes around the perimeter of a ological range. It appears that the resting length, or the length at
long bone shaft, and combining their output with knowledge of the zero force, can only be approximated. Studies that rely on different
bone’s geometry, the distribution of principal strains can be deter- criteria for defining zero load give very different strain values during
mined. It has been shown that the loading pattern of each bone is similar activities and at the yield point. An objective method of
fairly consistent in different activities, though peak strain and strain determining the transition between the toe and the elastic region
rate vary with gait and speed (Rybicki et al., 1974). This informa- has been described (Riemersma & van den Bogert, 1993).
tion has been applied in locating the tension surface of various long In tendons and ligaments, unidirectional strain gauges are ade-
bones, which is the surface of choice for the application of bone quate to record tensile strains during loading. The force or strain
plates. Strain gauges have been bonded to equine long bones to developed within these structures can be measured directly using
investigate bone loading under various conditions during exercise an implanted transducer that should be sufficiently compliant to
and training (e.g. Hartman et al., 1984; Schamhardt et al., 1985; measure tissue tension without interfering with its normal use
Nunamaker et al., 1990; Davies et al., 1993, 2004; Davies 2005). during locomotion (Ravary et al., 2004).
Strain gauges are easily bonded to the hoof wall (Fig. 2.21) to Regardless of the type of transducer used, the electrical signal
study the functional anatomy of the hoof capsule under a variety must be converted to strain (relative elongation of the tissue) or

48
 Tissue strain

load (internal force within the tissue) by a calibration process in across the primary loading axis of the tendon then threading a
which the sensor itself or the tissue containing the implanted sensor sterile fiberoptic sensor through the needle. The needle is then
are stretched. Methods of calibration include traction in a materials removed leaving the sensor embedded in the tendon. One end of
testing machine, tensile loading with known weights, stretching the the fiber is connected to the light transmitter unit and the other end
tendons in situ in the limb by manipulation of the isolated limb or is connected to the light receiver unit. As the tendon stretches lon-
by manually moving the joint, or electrical stimulation of the gitudinally, its transverse dimension decreases and squeezes the
attached muscle. In animals, direct calibration is often performed fiberoptic sensor, causing it to lose light, which is quantified by the
post-mortem after anatomical dissection of the tissue. In live receiver. The amount of light lost varies linearly with the applied
animals, calibration must use an indirect method based on math- tension in the tendon (Komi & Ishakawa, 2007). Calibration of the
ematical equations of static equilibrium. sensor is similar to the buckle transducer. The fiberoptic transducer
Several types of strain gauges have been used to study tendon is relatively unobtrusive (diameter 0.25–0.5 mm), easy to implant
strains in horses. In a buckle transducer, the tendon is wound over and appears to be well tolerated.
a buckle and preloaded as it passes over the middle support An ultrasound probe has been used to measure tendon strain
bar. Tendon damage is reduced by adapting the transducer dimen- non-invasively (Pourcelot et al., 2005) based on the fact that the
sions and the shape of the crossbar to the width of the tendon. speed at which acoustic waves travel through a tendon varies with
Tensile loading straightens the tendon and loads the buckle, and the mechanical properties of the tissue; when the tendon is loaded,
this loading is detected by strain gauges bonded on the transducer the ultrasound wave travels faster. An ultrasound probe consisting
surface to measure deformation of the buckle. Calibration of a of an emitter and multiple receivers is placed on the skin with the
buckle strain gauge is achieved by ex vivo transection of the tendon receivers spaced at regular distances along the skin overlying the
and application of known force or in vivo evaluation of tendon force superficial digital flexor tendon. The ultrasound wave speed is cal-
from joint moment data. Buckle transducers appear to be poorly culated from knowledge of the distance between the receivers. The
tolerated and are associated with more postoperative pain than probe is fixed to the skin of the palmar metacarpal area using an
other types of strain gauges. Another problem with the buckle adapted boot (Fig. 2.22). It has been used to evaluate the effects of
transducer is that, by forcing the tendon to follow a curved course, farriery modifications on loading of the superficial digital flexor
its initial strain and tension are altered. The possibility of preload- tendon at walk and slow trot (Crevier-Denoix et al., 2004).
ing of the tissues is a particular concern when making measure-
ments in short ligaments.
Liquid metal strain gauges consist of compliant tubes filled, Sonomicrometry
under slight pressure, with a metal that is liquid at room tempera-
ture (e.g. mercury in silastic). Strain-induced changes in the resis- Sonomicrometry measures distances using transducers, commonly
tance of the liquid metal are measured, which depend on the length called ‘crystals’, made from piezoelectric ceramic material that trans-
and cross-sectional area of the column of metal. When the gauge is mit and receive sound energy. The crystals are implanted in soft
stretched, the liquid column becomes longer and narrower, which tissue structures where they emit an ultrasonic pulse and detect
is associated with an increase in electrical resistance. Since the gauge similar pulses from other crystals. The sonomicrometer measures
is part of a circuit with constant current, a change in electrical resis- the time between transmission of the sound wave from one crystal
tance induces a change in voltage. These gauges are implanted with to its reception by the other crystals. The time differences are con-
some prestrain to ensure detection of any changes in length. Liquid verted to distance measurements. Specialized digital circuitry makes
metal strain gauges have the advantage of being calibrated in vivo, continuous calculations of the distances between transducer crystals
but they have to be custom made, which is a tricky process. Micro- implanted in or attached to soft tissue structures to quantify the
damage in the area of implantation alters the tensile properties of active and passive functional properties of moving tissues.
the tendon within a few days, so readings must be taken as soon as Sonomicrometry is useful in studies of muscle function to deter-
mine whether muscles are working concentrically, isometrically or
possible after surgery (Jansen et al., 1998). Longevity is also
eccentrically and to relate the type of contraction with the amount
adversely affected when oxygen diffuses through the porous silastic
of force produced. This technique has been used in horses to inves-
tubing and oxidizes the mercury. There is also a possibility of break-
tigate the relationship between muscle strain and length change of
ing the connecting wires or loss of contact between the wire and
the musculotendinous unit (Hoyt et al., 2005)
the liquid metal column. To reduce the risk of problems, readings
should be taken as soon as possible after implantation and it may
be necessary to limit the number of in vivo measurements that are
collected. Liquid metal strain gauges have been used to investigate
Calculation of strains
load distribution between the flexor tendons and suspensory liga- Numerical modeling of the limb allows simultaneous evaluation of
ment (Jansen et al., 1993), to characterize the behavior of the limb the strain contributions of different ligaments and tendons, even
(Riemersma et al., 1988a, 1988b), to assess the effect of support those that are deeply placed. The effect of interventions, such as a
bandages (Keegan et al., 1992) and to detect changes in the loading change in hoof angle can be studied (e.g. Lawson et al., 2007a).
pattern in response to changes in surface type or shoeing adapta- Bone-fixed markers have been used to calculate three-dimensional
tions (Riemersma et al., 1996a, 1996b). bone segment trajectories, from which geometric displacement of
A transducer based on the Hall effect, in which the voltage output tendon and ligament insertions were calculated to determine strain
of a semiconductor is proportional to the strength of a magnetic of the digital flexor tendons in vitro (Rollot et al., 2004). The estima-
field, was used to measure strain in the superficial digital flexor tions agreed fairly well with simultaneous measurements of tendon
tendon. Although the strains recorded were higher than would be length using needles pinned perpendicularly through the superficial
expected, this may have been due to the definition of initial length digital flexor tendon in the mid-metacarpal and digital regions.
as the length at heel strike (Stephens et al., 1989). Differences between calculated and measured strains, which were
A force transducer that detects strain from a very small part of the in the range of 12–15%, were due to integration of the length of
tendon has been applied in horses (Platt et al., 1994). A drawback the proximal scutum (approximately 4 cm), which was not taken
is that the small area sampled is not necessarily representative of into account in the direct measurement of tendon length.
strain in the entire tendon. Meershoek et al. (2001a) quantified the parameters for a pulley
A fiberoptic transducer consists of a plastic optical fiber that loses model to describe the relationships between force and strain of the
transmitted light under compression. It is implanted into a tendon flexor tendons and between fetlock joint angle and suspensory liga-
under local anesthesia by passing a 19-gauge needle transversely ment strain. Evaluation of the sensitivity of the technique indicated

49
 2 Measurement techniques for gait analysis

Fig 2.23  Telemetered system for measuring rein tension showing strain
gages inserted between rein and bit with battery operated power source
and transmitter-attached to the noseband underneath the horse’s chin.

the dorsopalmar path of the tendons at the level of the proximal

sesamoid bones have more influence on calculated tendon strain
than changes in their attachments (Lawson et al., 2007b).

Measurement of rein tension

Tension in the reins is one of the mechanisms used by the rider to
control the horse’s speed and direction. Rein tension can be mea-
sured using strain gauge transducers (Preuschoft et al., 1999;
Clayton et al., 2003b, 2005). Sensors used for this purpose must be
small in size and light in weight (Fig. 2.23) to avoid interfering with
the normal movements or functions of the reins and the coupling
between bit, transducer and rein must be inextensible. Ideally, the
power source and transmitter should be small enough to be worn
B on the bridle or breastplate (Fig. 2.23) or attached around the
horse’s neck. Synchronized video images are useful to relate the
Fig 2.22  Above: ultrasound probe for dynamic measurement of tendon timing of kinematic events, such as footfalls, with patterns in
strain. Below: probe shown beside the special boot used to hold it in secure the rein tension graphs (Fig. 2.24). Useful variables include
contact with the shaved skin overlying the digital flexor tendons. maximal, minimal and average tension values, rate of increase in
Photograph courtesy of Dr. Nathalie Crevier-Denoix. tension and width of the tension spikes. Rein tension measure-
ments are applied in evaluations of the effects of tack and rider
technique.
that modeling errors in the mechanical properties of the suspensory
ligament and measurement errors in the point of application of the Accelerometers, gyroscopes and
GRF and position of the marker on the distal interphalangeal joint
had a substantial effect on accuracy. After correction of these errors,
magnetometers
the method was used to calculate mean tendon forces and to evalu- Accelerometers, gyroscopes and magnetometers provide different
ate the influence of factors such as surface properties, type of shoe, pieces of information that can be combined to provide a more
speed, and fatigue on tendon forces during trotting (Meershoek & complete picture of the movement pattern. They are used in various
Lanovaz, 2001) and jumping (Meershoek et al., 2001b). combinations in gait analysis. Miniaturization of the components
When using musculoskeletal models to calculate soft tissue allows their attachment to the extremities without affecting kine-
strains in the superficial and deep digital flexor tendons, changes in matics or kinetics.

50
 Tissue strain

40
Left rein
Right rein

30
Rein tension (N)

20

10

0
0 1 2 3 4
Time(s)

Fig 2.24  Rein tension between the bit and leather side reins for a horse Fig 2.25  Electronics, accelerometers and power source shown with
trotting in hand in a straight line. polyurethane housing for mounting on the hoof wall.
Reprinted from Ryan, C.T., Dallap Schaer, B.L., Nunamaker, D.L., 2006. A novel wireless data
acquisition system for the measurement of hoof accelerations in the exercising horse. Equine
Accelerometers measure acceleration of the surface to which they Vet. J. 38, 671–674, with permission from the Equine Veterinary Journal.
are attached. Sub-miniature accelerometers (mass <1 g) are avail-
able. They contain silicon beams that deform during acceleration
causing changes in capacitance that are output as a voltage propor- mechanical energy fluctuation in galloping Thoroughbreds (Pfau
tional to the applied acceleration. Tilt relative to the earth’s gravi- et al., 2006) and in horses moving on an incline (Parsons et al.,
tational field can also be measured using these accelerometers. 2008), and to compare movements of the left and right tuber coxae
Gravity deforms the silicon beam, changing the capacitance and in horses with mild hind limb lameness (Pfau et al., 2007). Inertial
associated voltage of the sensor. motion tracking has inherently poor resolution and, over time, the
Since each accelerometer measures acceleration in one direction readings drift.
only, two or three accelerometers are often stacked with their axes A magnetometer measures the strength and/or direction of the
perpendicular to each other to obtain two-dimensional or three- magnetic field around it, which can be used to provide an absolute
dimensional accelerations. Interpretation of the output is subject to measure of orientation using the earth’s magnetic field, as in a
errors due to misalignment of the axes of the accelerometers with compass. A triaxial magnetometer with a tilt sensor can be used to
the body axes or to rotations of the instrumented body segments sense absolute three-dimensional orientation.
during locomotion. Orientation of the accelerations relative to the In equine studies, accelerometers have most often been applied
body’s frame of reference can be achieved using optical motion to the hoof wall, where they are used to detect initial ground contact
capture or by integration of the angular velocity output of three and to measure the associated accelerations. In one design (Fig.
orthogonal gyroscopes. 2.25), three accelerometers, a power source and the associated elec-
Gyroscopic sensors measure orientation based on the principle tronics were potted in a shock-attenuating silicon shell inside a
of conservation of angular momentum. The essence of the instru- semi-rigid polyurethane housing that was contoured to fit the
ment is a spinning wheel on an axle that tends to resist changes to dorsal hoof wall. The total mass of the assembly was only 148 g
its orientation due to the angular momentum of the wheel. If two (Ryan et al., 2006).
gyroscopes with their axles perpendicular to each other are mounted By mounting accelerometers to the hoof wall and to the bones
on a platform inside a set of gimbals, the platform will remain of the digit, Lanovaz et al. (1998) studied the attenuation of impact
completely rigid as the gimbals rotate. Sensors on the gimbals’ axes shock in the distal digit in vitro. Willemen et al. (1997) performed
detect platform rotation. When used in combination with acceler- similar measurements both in vitro and in vivo. A hoof-mounted
ometers, the direction of motion and changes in motion can be accelerometer is an effective tool for measuring certain characteris-
measured in all three directions. Miniature gyroscopes (mass <1 g) tics of the footing (Barrey et al., 1991) and the efficiency of shock
measure angular velocity through the phenomenon of Coriolis absorbing shoes and pads (Benoit et al., 1991). Hoof mounted
force, which is generated when a rotational angular velocity is accelerometers have also shown that impact accelerations increase
applied to a vibrating element. It is a by-product of measuring with speed and are affected by the dynamic properties of the surface
coordinates relative to a rotating coordinate system. the horse moves over (Ratzlaff et al., 2005; Ryan et al., 2006).
Small, lightweight inertial-based sensors are generally robust and A different application used two accelerometers secured beneath
have some useful applications for recording data from a large the horse’s sternum to measure longitudinal and dorsoventral accel-
number of consecutive strides, when the use of an optical motion erations of the trunk segment (Barrey et al., 1994) with data trans-
analysis system may not be practical. In addition to providing accu- mitted telemetrically to a receiver connected to a portable computer.
rate orientation data, inertially based systems can be used to calcu- Analysis of the symmetry of the trunk acceleration patterns during
late displacements (Pfau et al., 2005). The large volumes of data trotting revealed subtle asymmetries in lame horses, though in some
collected by systems based on inertial sensors necessitate automatic of these horses the side of hind limb lameness was identified incor-
processing, especially if the system is to be used in a clinical setting, rectly (Barrey & Desbrosse, 1996). The same device has been used
where results are needed rapidly for timely decision-making. One to study accelerations of the trunk during jumping (Barrey &
such technique involves the use of a hidden Markov model-based Galloux, 1997). Accelerometers attached to the saddle have been
stride segmentation technique based on stochastic pattern recogni- used to measure the acceleration at different gaits and the findings
tion. It has been used to evaluate data from trunk-mounted, six have been applied in the development of a mechanical horse that
degrees of freedom, inertial sensors in horses (Pfau et al., 2008). simulates the motions during walking, trotting, cantering and
Other studies have evaluated center of mass movement and jumping (Galloux et al., 1994).

51
 2 Measurement techniques for gait analysis

The output from vertically oriented, uni-axial accelerometers, one

attached to the halter close to the poll and the other glued to the
top of the croup, have been used to compute vertical displacements
of the head and pelvis as the second integral of the accelerations.
Movement asymmetries of the head or croup in lame horses are
detected. When used in combination with gyroscopic transducers
attached to one hoof to indicate the timing and duration of
its stance phase, lameness can be partitioned between fore and
hind limbs, then assigned to the left or right side (Keegan et al.,
2001). Based on head movement data, the lameness is further
assigned to the loading or push off phase. Comparison between this
accelerometric-gyroscopic system and a video-based motion analy-
sis system showed a high linear correlation for quantification of the
degree of lameness in both fore and hind limbs (Keegan et al.,
2004).
A three-directional gyroscopic transducer attached to the hoof has
been used on a standalone basis to evaluate the effects of trimming
and shoeing modifications on hoof motion and the direction of
breakover (Keegan et al., 2005).
Combinations of accelerometers, gyroscopes and magnetometers
have many applications, for example in virtual reality systems. With
regard to motion analysis, combination of orientation data from
gyroscopes with three-dimensional accelerations allows motion Fig 2.26  Pressure profile beneath the hoof collected with electronic
pressure mat (RSFootScan, RSScan International, Olen, Belgium).
tracking with six degrees of freedom. As an example, an inertial
Photograph courtesy of Dr. W. Back.
orientation sensor that has been modified for equine use to provide
a full set of movement parameters (displacement, velocity, accelera-
tion) in three directions. The sensor was attached over the spine of pressure, calibration procedure, duration of pressure application
horses exercising on a treadmill to measure orientation and relative and length of time for which the insole has been used (Hsiao et al.,
position at the canter (Pfau et al., 2005). After high pass filtering to 2002). Some types of pressure-measurement systems are particu-
separate non-cyclical and cyclical components of the movement, larly sensitive to calibration, with measurement errors being reduced
the data were comparable with those from an optical motion when the sensors are calibrated within the range of pressures that
capture system. This combination of sensors has been used to study they will be used to measure. An equine study suggested that sensors
movements of the center of mass and mechanical energy fluctuation be used for a maximum of three trials before recalibration (Carter
during galloping (Pfau et al., 2006) and the effects of an incline on et al., 2001). Data recorded simultaneously from a force plate and
center of mass movement and mechanical energy fluctuation a pressure mat have confirmed the value of the pressure mat for
(Parsons et al., 2008). This technology has also been marketed as evaluating temporal variables and kinetic symmetry ratios but peak
a clinical lameness evaluation tool (Keegan et al., 2004). vertical force and vertical impulse values are lower than those
recorded by the force plate (Oosterlinck et al., 2010). This problem
can be overcome by calibrating the pressure mat dynamically using
Pressure mats reference values from a force plate. In this method, the sum of the
vertical forces applied to all the sensors in the pressure mat is
Pressure mats contain an array of sensors used to map the pattern adjusted to match the vertical component of the GRF (van Heel
of pressure across a surface. The main applications in horses are to et al., 2004; Oosterlinck et al., 2012).
map pressure distribution beneath the hoof or beneath the saddle. In horses, practical applications of the use of sensor mats beneath
In human gait analysis, ultra-thin, in-shoe sensors have been used the hooves have focused on lameness detection (Carter et al., 2001;
to map pressure beneath the foot and force between the foot and Perino et al., 2007) and evaluation of hoof dynamics and farriery
the ground as a means of evaluating podiatric and orthopedic (van Heel et al., 2004).
problems. Sensor mats trimmed to the shape of the hoof have been attached
using adhesive tape (Judy et al., 2001) or placed inside a hoof boot
(Perino et al., 2007) to measure simple temporal variables, pressure
distribution across the solar surface of the hoof and total force
Hoof pressure mats summed over all sensors. The total force recorded simultaneously
Pressures beneath the hoof are used to evaluate and compare far- by the pressure mat and a force plate (Perino et al., 2007) did not
riery practices and in the detection of mild lameness. To perform differ statistically, but there was an overall lack of agreement
these functions, the mat and the sensors must be sufficiently robust, between the two sets of measurements (up to 1000 N or 20% of the
accurate and reliable to measure relatively high forces and pressures. total measurement) with higher coefficients of variation for the
Larger mats (e.g. Matscan®, Tekscan Inc., Boston, MA) are placed on pressure mat (10.2%) than the force plate (6.6%). The authors sug-
the floor or on top of a force plate. Some smaller mats can be cut gested that shear forces may have interfered with the attachment of
to the appropriate shape and attached beneath the hoof or used sensors around the periphery of the trimmed mat leading to greater
inside a boot. The Hoof TM System (Tekscan Inc., Boston, MA) dis- variability in forces recorded from the trimmable versus untrim-
plays contact area and color-coded pressure profiles (Fig. 2.26), mable pressure mats.
average and peak pressures, center of pressure and its trajectory. The An RSFootScan (RSscan International, Olen, Belgium) was
associated software displays force-time and peak pressure-time embedded in an aluminum plate on top of a force plate. This com-
graphs, and allows the user to isolate and analyze data from specific bination allowed dynamic calibration of the force mat and provided
regions of the hoof or to make side-by-side comparisons of pre- and force and pressure data that gave a more detailed analysis of the
post-treatment conditions. areas of contact between the hoof and the surface than was possible
In human studies, accuracy and precision of in-sole pressure mats with the force plate alone, especially during loading and unloading
have been shown to vary with manufacturer, levels of applied (van Heel et al., 2004).

52
 Electromyography

Saddle pressure mats and comparison: left and right sides; front and back halves; quad-
rants, etc. Small changes in position of the horse or saddle relative
Knowledge of pressure beneath the saddle (Fig. 2.27) is potentially to the position of the mat affect the force distribution and the
useful to evaluate saddle fit, rider position and riding technique. assignment of summed forces to different areas, so it is important
Pressure mats marketed specifically for this purpose have an array to standardize the position of the mat relative to both the horse’s
of sensors to measure forces acting normal (perpendicular) to their back and to the saddle and to standardize the horse’s position for
surface. Early models were hard-wired to a laptop computer via a standing measurements.
long tether, which imposed limitations on the horse’s mobility. The As with all quantitative measurement devices, validity and reli-
latest generation of pressure mats use wireless technology, allowing ability are important considerations. For a saddle pressure mat,
data to be collected on-board using a PDA or transmitted telemetri- there should be a high correlation between the sum of the measured
cally to a computer, while the horse moves around freely. forces and the mass of the riders and saddle. Pressure bench tests
Each sensor detects the force applied perpendicular to its surface (Hochmann et al., 2002) indicate that the Pliance system (Novel
(normal force) and the output on the computer screen is color- GmbH, Munich, Germany) might be more reliable than the FSA
coded to represent bands of force. The operator sets the force range (Vista Medical Ltd., Winnipeg, Canada), ClinSeat (Tekscan Inc.,
within each color band depending on the objective of the evalua- Boston, MA) and Xsensor (XSENSOR Technology Corporation,
tion. If saddle fit is being evaluated without a rider, a smaller range Calgary, Canada) systems. The Pliance mat has also been shown to
of forces is required, whereas if the horse will be ridden, the added be more reliable than the FSA device in equine practice (Jeffcott
weight of the rider requires a larger range of forces. The output can et al., 1999; de Cocq et al., 2006, 2009). Subsequently, the Pliance
be displayed as a sensor plot, a two-dimensional contour plot or a mat was proven valid for this purpose, but reliability measures
three-dimensional surface plot (Fig. 2.28). Total force summed over indicate that the mat should remain on the horse’s back when
all sensors and maximal force in an individual sensor are also mea- comparing different conditions (De Cocq et al., 2009). Pressure
sured. The data can be partitioned into smaller areas for analysis mats have been used in saddle fitting to compare different saddle
brands (Werner et al., 2002) and tree width on saddle fit (Meschan
et al., 2007). In ridden horses, they have been applied in studies of
the horse’s gait and movements and rider technique (e.g. Freuh-
wirth et al., 2004; Peham et al., 2004; Geutjens et al., 2008).

Electromyography

Muscle contraction is preceded by electrical activation, which can be

detected and recorded as the electromyogram, to provide informa-
tion on the state of activity of the motor neurons at rest, during reflex
contraction and during voluntary contraction. Surface electromyog-
raphy (EMG) is non-invasive, and can be performed on the con-
scious horse and used during locomotion (Korsgaard, 1982; Tokuriki
et al., 1989; Jansen et al., 1992; Tokuriki & Aoki, 1995; Robert et al.,
1999, 2000, 2001a, 2001b, 2002; Tokuriki et al., 1999; Colborne
et al., 2001; Licka et al., 2004, 2009; Wickler et al., 2005; Wakeling
et al., 2007; Zsoldos et al., 2010). Intramuscular EMG, in which the
Fig 2.27  Pressure mat (Pliance, Novel GmbH, Munich, Germany) placed electrodes are inserted into the muscle tissue, is used to study normal
beneath saddle. Power source and transmitter are in the pack carried at the (Schuurman et al., 2003) and pathological (Wijnberg et al., 2004;
front of the saddle. The wires shown in loops on the horse’s shoulder fit into Macgregor & Graf von Schweinitz, 2006) muscular activity. The
the green bag during data collection. ability to investigate conditions, such as ataxia, myotonia, myositis,

Fig 2.28  Output from Pliance force mat (Novel GmbH, Munich, Germany) shown as individual sensor map (left), two-dimensional planar view (center) and
three-dimensional view (right). Pommel is toward top and left side is on the left for sensor map and two-dimensional planar view. The three-dimensional
view has been rotated slightly counterclockwise. Pressures are color-coded according to the scale on the right.

53
 2 Measurement techniques for gait analysis

shivers and stringhalt, and to diagnose neurologic disorders, such as

equine motor neuron disease (EMND), hyperkalemic periodic paral-
ysis, myotonia congenita and equine myotonic dystrophy, has
resulted in EMG becoming a useful diagnostic tool. 500 mV
A
100 ms
EMG equipment
Myoelectric signals are collected by electrodes connected to a pre-
amplifier located close to the muscle. Signal conditioning, consist-
ing of data sampling, analog-to-digital conversion, multiplexing 20 mV
and coding are performed before transmission to the recording unit. B
10 ms
Data are transferred from the receiving unit to the data recorder by
cables, optical fibers or radio-telemetry. Further amplification of the
signal, filtering and, in some cases, digital-to-analog conversion are
performed by the receiver.
The electrodes function as the antenna to pick up the electrical
signal. They may be placed on the skin surface, inserted percutane-
C 5000 mV
ously into the muscle or implanted surgically. Surface electrodes
10 ms
have the advantage of being non-invasive but they provide only a
gross estimation of muscle activity in the large superficial muscle Fig 2.29  Normal electromyographic findings: (A) insertional activity; (B)
groups. Electrodes with higher spatial resolution are now available endplate noise; (C) motor unit action potential (MUAP).
(Huppertz et al., 1997). Jansen et al. (1992) found surface EMG to Reprinted from van Wessum, R., Sloet van Oldruitenborgh-Oosterbaan, M.M., Clayton, H.M.,
be a reliable and reproducible technique. However, many locomo- 1999. Electromyography in the horse in veterinary medicine and in veterinary research: a
tor muscles of the horse are deeply placed or lie beneath the thick review, with permission of Taylor & Francis Ltd, http://www.tandf.co.uk/journals.
cutaneous muscles, making them unsuitable for study by surface
electrodes. Cross talk may be a significant problem when using
surface electrodes (Schuurman et al., 2003). excitation the resting potential is temporarily reversed to 40 mV,
The percutaneous technique involves using a hypodermic needle with the outside being negative. As the action potential travels along
to introduce fine wires into the muscle belly to measure localized the muscle fiber, the small electrical potential generated across the
electrical activity in a small number of fibers close to the electrodes. surface membrane is dissipated in the surrounding interstitial fluid,
Several electrode configurations are used. A unipolar electrode which is a good electrical conductor. The summation of the electri-
usually has a single strand of wire coated with an insulating mate- cal changes in the interstitial fluid is recorded as the EMG.
rial except for a small length at its distal end. The wire is inserted Surface EMG is a non-invasive technique that gives a general
percutaneously through a small hypodermic needle and the end of picture of muscle activation in subcutaneous muscles, whereas
the wire is bent back on itself to act as a barb. The potential is intramuscular EMG measures localized electrical activity in a small
measured between the uninsulated tip of the wire and a reference number of fibers located around the electrodes. With the intramus-
electrode on the skin. Unipolar electrodes detect the time of activa- cular technique, there are three phases of electrical activity: inser-
tion of the muscle, rather than the amplitude of the contraction. tional activity, resting activity and activity during muscular
Bipolar electrodes measure the voltage difference between two elec- contraction. The mechanical stimulation associated with insertion
trical contacts. A simple bipolar design has two hooked wires or movement of the EMG provokes spontaneous activity with posi-
inserted through a hypodermic needle. Another type is the concen- tive or negative high frequency spikes in a cluster that may last for
tric bipolar electrode, which measures the potential difference around 500 ms and are followed by electrical silence (Fig. 2.29)
between a point-like recording contact and an average of all the (Wijnberg et al., 2003). Abnormally shaped spikes or abnormal
potentials in a ring surrounding it at some fixed distance. The duration of electrical activity are signs of muscle pathology. Inser-
advantages of this configuration are that the recording electrode is tional activity tends to be short in duration when muscle fibers have
shielded by the outer needle from electromagnetic noise and the been lost due to fibrosis or in ischemia, whereas neurogenic or
reference is kept as close to the electrode as possible. Concentric myogenic problems induce abnormal muscular discharges and
bipolar electrodes are often used in clinical applications (Wijnberg potentials (Wijnberg et al., 2003). Hyperirritability and instability
et al., 2003). of the muscle fiber membrane is usually a sign of early denervation
To ensure accurate placement in deeper muscles, the electrodes atrophy but is also seen in myotonic disorders and myositis (Kimura,
may be inserted under ultrasonic guidance. Surgical implantation 1984) and in equine motor neurone disease (Podell et al., 1995).
of EMG electrodes is a more complex, time-consuming and poten- Following the burst of insertional activity, relaxed skeletal muscle
tially damaging procedure, but it provides the best results in terms is electrically silent (Naylor et al., 1992), resulting in a fairly flat
of electrical and mechanical reliability, since the operator has direct trace in the baseline electromyogram. When the electrode is posi-
visual control of the position of the electrodes. tioned near an endplate it gives rise to endplate noise (Fig. 2.29)
Muscle potentials may be displayed on a computer with a data and endplate spikes, due to spontaneous release of acetylcholine.
acquisition system using the convention that a positive potential is Repositioning the needle eliminates these signals. Irritation of nerve
indicated by a downward deflection. Special software is used to terminals in the vicinity of the electrodes may also induce electrical
process the information, to measure variables such as amplitude discharges (Wijnberg et al., 2003).
and frequency of the spikes, to rectify the output, to construct an Motor unit action potentials (MUPs) are the summation of
envelope that touches the peaks of the spikes, and to measure the muscle action potentials from the voluntary or reflex contraction of
areas under the resulting curves. myofibers in a motor unit. Approximately 50 muscle fibers around
a needle electrode contribute to the observed potential. The wave-
form of the MUP is described in terms of its amplitude and duration
The electromyogram and the number of phases and turns in the graph. Dedicated soft-
ware is used to measure the amplitude and frequency of the MUPs,
A resting muscle fiber has a 90 mV potential difference across its rectify the curves, construct envelopes, and calculate the area under
surface membrane, with the outer side being positive. During the curves.

54
 References

The duration, from initial elevation until return to baseline, is (Kimura, 1989). In human subjects, it may be possible to estimate
affected by the synchrony in firing of the individual muscle fibers the number of active motor units at a given muscle contraction
and the conduction velocity. Discharges generated further from the (Kimura, 1989) but in horses, it is difficult to judge the level of force
electrode form the low amplitude areas at the start and end of the and, therefore, the value of evaluating the recruitment pattern is
MUP. The amplitude of the MUP, defined from peak to peak, limited (Wijnberg et al., 2002a, 2002b). This would require a more
depends on the distance from the electrode tip to the center of sophisticated muscle model that incorporates the muscle architec-
maximal potential of the discharging unit. For example, fewer than ture, the force–length and force–velocity relationships of the muscle
20 muscle fibers within a 1 mm radius of a concentric needle elec- fibers, and the activation of the muscle, possibly from EMG signal
trode contribute to the spike of the motor unit. Both amplitude and analysis.
duration of the MUP may be affected by the horse’s age and muscle Electromyographic studies of athletic horses have described the
temperature. The shape of the MUP is described in terms of the activation pattern of various muscles during normal locomotion
number of phases, with a phase being defined as the part of (Wentink, 1978; Korsgaard, 1982; Tokuriki et al., 1989; Tokuriki &
the curve between departure from and return to baseline, and turns, Aoki, 1995; Robert et al., 1999) and have investigated the effects of
which are the changes in direction of the curve. Normal equine speed and incline (Robert et al., 2000, 2001a, 2001b, 2002; Wickler
MUPs have been described as potentials that have two or three et al., 2005) on muscle activation patterns. Over ground locomo-
phases, a duration of 3–10 ms and an amplitude of about 1500 µV tion has been compared with swimming and walking in an under-
(Mayhew et al., 1978). More recently, MUP analyses have been water treadmill (Tokuriki et al., 1999) and the effects of fatigue on
reported in the horse for the subclavian muscle (Wijnberg et al., the EMG median frequency have been investigated (Colborne et al.,
2002a) and for the triceps and lateral vastus muscles (Wijnberg 2001).
et al., 2002b).
The amplitude of the EMG signal depends on the dimensions of
the electrodes, their electrical contact with the muscle, and the kind New technologies
of electrodes: signals from indwelling wire electrodes usually are
much lower than those obtained from surface electrodes. However, The 21st century is an age in which technology is changing rapidly.
the major influence on the EMG signal amplitude is caused by the Although the contents of this chapter are current at the time of
degree of activation of the muscle. When the muscle is completely writing, by the time the text is printed some of the information may
activated, the EMG signal will reach a maximum. This relationship be outdated and new methods and systems for locomotion analysis
allows the EMG signal amplitude to be used as a measure of the will undoubtedly have become available. It is recommended
degree of activation, and thus indirectly, of muscle force develop- that the reader treat the information in this chapter as a framework
ment (Hof, 1984). Recruitment is defined as successive activation from which to evaluate future developments in gait analysis
of motor units with increasing strength of voluntary contraction methodologies.

References
Adrian, M., Grant, B., Ratzlaff, M., Ray, J., representations of equine hind limb translation measures. J. Biomech. 40,
Boulton, C., 1977. Electrogoniometric kinematics. Equine Vet. J. 27, 39–45. 1417–1422.
analysis of equine metacarpophalangeal Barr, A.R.S., Dow, S.M., Goodship, A.E., 1995. Benoit, P., Barrey, E., Regnault, J.C., Brochet,
joint lameness. Am. J. Vet. Res. 38, 431. Parameters of forelimb ground reaction J.L., 1991. Comparison of the damping
Alexander, R.McN., 1977. Terrestrial forces in 48 normal ponies. Vet. Rec. 136, effect of different shoeing by the
locomotion. In: Mechanics and energetics 283–286. measurement of hoof acceleration. Acta
of animal locomotion. John Wiley and Barrey, E., 1990. Investigation of the vertical Anat. 146, 109–113.
Sons, New York, pp. 168–203. hoof force distribution in the equine Bialski, D., Lanovaz, J.L., Bohart, G.V.,
Back, W., Barneveld, A., Schamhardt, H.C., forelimb with an instrumented horseboot. Mullineaux, D.R., Clayton, H.M., 2004.
Bruin, G., Hartman, W., 1994. Longitudinal Equine Vet. J. 9 (Suppl.), 35–38. Effect of detomidine on postural sway in
development of the kinematics of 4-, 10-, Barrey, E., Desbrosse, F., 1996. Lameness horses. Equine Compar. Exerc. Physiol. 1,
18- and 26-month-old dutch warmblood detection using an accelerometric device. 45–50.
horses. Equine Vet. J. 17 (Suppl.), 3–6. Pferdeheilkunde 12, 617–622. Björck, G., 1958. Studies on the draught forces
Back, W., Barneveld, A., van Weeren, P.R., van Barrey, E., Galloux, P., 1997. Analysis of the of horses: development of a method using
den Bogert, A.J., 1993. Kinematic gait equine jumping technique by strain gauges for measuring forces between
analysis in equine carpal lameness. Acta accelerometry. Equine Vet. J. 23 (Suppl.), hoof and ground. Acta Agric. Scand. Suppl.
Anat. 146, 86–89. 45–49. 4, 1–109.
Back, W., MacAllister, C.G., van Heel, M.C.V., Barrey, E., Galloux, P., Valette, J.P., Auvinet, B., Bobbert, M.F., Gómez Álvirez, C.B., van
Pollmeier, M., Hanson, P.D., 2007. Vertical Wolter, R., 1993. Stride characteristics of Weeren, P.R., Weishaupt, M.A., Roepstorff,
frontlimb ground reaction forces of sound over ground versus treadmill locomotion. L., 2007. Validation of vertical ground
and lame warmbloods differ from those in Acta Anat. 146, 90–94. reaction forces on individual limbs
quarter horses. J. Equine Vet. Sci. 27, Barrey, E., Hermelin, M., Vaudelin, J.L., Poirel, calculated from kinematics of horse
123–129. D., Valette, P., 1994. Utilisation of an locomotion. J. Exp. Biol. 210, 1885–1896.
Back, W., Schamhardt, H.C., Savelberg, accelerometric device in equine gait Bobbert, M.F., Santamaría, S., 2005.
H.C.C.M., van den Bogert, A.J., Bruin, G., analysis. Equine Vet. J. 17 (Suppl.), 7–12. Contribution of the forelimbs and hind
Hartman, W., et al., 1995a. How the horse Barrey, E., Landjerit, B., Wolter, R., 1991. Shock limbs of the horse to mechanical energy
moves. 1. Significance of graphical and vibration during hoof impact on changes in jumping. J. Exp. Biol. 208,
representations of equine forelimb different track surfaces. Equine Exerc. 249–260.
kinematics. Equine Vet. J. 27, 31–38. Physiol. 3, 97–106. Buchner, H.H.F., Savelberg, H.H.C.M.,
Back, W., Schamhardt, H.C., Savelberg, Beardsley, C.L., Paller, D.J., Peura, G.D., Schamhardt, H.C., et al., 1994a.
H.C.C.M., van den Bogert, A.J., Bruin, G., Brattbakk, B., Beynnon, B.D., 2007. The Habituation of horses to treadmill
Hartman, W., et al., 1995b. How the horse effect of coordinate system choice and locomotion. Equine Vet. J. 17 (Suppl.),
moves. 2. Significance of graphical segment reference on RSA-based knee 13–15.

55
 2 Measurement techniques for gait analysis

Buchner, H.H.F., Savelberg, H.H.C.M., 2000c. Net joint moments and joint during locomotion. Acta Anat. 146,
Schamhardt, H.C., et al., 1994b. Kinematics powers in horses with superficial digital 148–153.
of treadmill versus over ground locomotion flexor tendinitis. Am. J. Vet. Res. 61L, Davies, H.M.S., Merritt, J.S., 2004. Surface
in horses. Vet. Quart. 16 (Suppl. 2), 197–201. strains around the midshaft of the third
S87–S90. Clayton, H.M., Schott, H.C., Littlefield, L., metacarpal bone during turning. Equine
Buchner, H.H.F., Savelberg, H.H.C.M., Lanovaz, J.L., 1999. Center of pressure Vet. J. 36, 689–692.
Schamhardt, H.C., et al., 1995. Temporal analysis in horses. Am. Coll. Vet. Int. Med. De Cocq, P., Clayton, H.M., Terada, K., Muller,
stride patterns in horses with 13, 241. M., van Leeuwen, J.L., 2009. Usability of
experimentally induced fore or hind limb Clayton, H.M., Sha, D.H., Stick, J.A., Elvin, normal force distribution measurements to
lameness. Equine Vet. J. 18 (Suppl.), N.G., 2007a. 3D kinematics of the evaluate asymmetrical loading of the back
161–165. metacarpophalangeal joint at walk and of the horse and different rider positions.
Buchner, H.H.F., Savelberg, H.H.C.M., trot. Vet. Compar. Orthop. Traum. 20, Vet. J. 181, 266–273.
Schamhardt, H.C., et al., 1996a. Limb 86–91. De Cocq, P., van Weeren, P.R., Back, W., 2006.
movement adaptations in horses with Clayton, H.M., Sha, D.H., Stick, J.A., Saddle pressure measuring: validity,
experimentally induced fore or hind limb Mullineaux, D.R., 2004. Three dimensional reliability and power to discriminate
lameness. Equine Vet. J. 28, 63–70. kinematics of the equine carpus at trot. between saddle-fits. Vet. J. 17:, 265–273.
Buchner, H.H.F., Savelberg, H.H.C.M., Equine vet. J. 36, 671–676. Deluzio, K.J., Astephen, J.L., 2007.
Schamhardt, H.C., et al., 1996b. Head and Clayton, H.M., Sha, D.H., Stick, J.A., Biomechanical features of gait waveform
trunk movement adaptations in horses Robinson, P., 2007b. 3D kinematics of the data associated with knee osteoarthritis an
with experimentally induced fore or hind proximal and distal interphalangeal joints application of principal component
limb lameness. Equine Vet. J. 28, 71–76. of walking and trotting horses. Vet. analysis. Gait Posture 25, 86–93.
Cappozzo, A., Croce, U.D., Leardini, A., Compar. Orthop. Traum. 2, 1–7. DeLuzio, K.J., Wyss, U.P., Li, J., Costigan, P.A.,
Chiari, L., 2005. Human movement Clayton, H.M., Singleton, W.H., Lanovaz, J.L., 1993. A procedure to validate three-
analysis using sterophotogrammetry, part 1: Cloud, G.L., 2003b. Measurement of rein dimensional motion assessment systems. J.
theoretical background. Gait and Posture tension during horseback riding using Biomech. 26, 753–759.
21, 186–196. strain gage transducers. Exp. Tech. 27, Drevemo, S., Dalin, G., Fredricson, I., Björne,
Carter, J.E., Galuppo, L.D., Snyder, J.R., 34–36. K., 1980b. Equine locomotion 3: the
Willitis, N.H., 2001. Evaluation of an Clayton, H.M., Singleton, W.H., Lanovaz, J.L., reproducibility of gait in standardbred
in-shoe pressure measurement system in Cloud, G.L., 2005. Strain gauge trotters. Equine Vet. J. 12, 71–73.
horses. Am. J. Vet. Res. 62, 23–27. measurement of rein tension during riding: Drevemo, S., Dalin, G., Fredricson, I., Hjertén,
Chateau, H., Degueurce, C., Denoix, J.-M., a pilot study. Equine Compar. Exerc. G., 1980a. Equine locomotion 1: the
2004. Evaluation of three-dimensional Physiol. 2, 203–205. analysis of linear and temporal stride
kinematics of the distal portion of the Clayton, H.M., Willemen, M.A., Schamhardt, characteristics of trotting standardbreds.
forelimb in horses walking in a straight H.C., Lanovaz, J.L., Colborne, G.R., 2000a. Equine Vet. J. 12, 60–65.
line. Am. J. Vet. Res. 65, 447–455. Kinematics and ground reaction forces in Drevemo, S., Fredricson, I., Hjertén, G., 1987.
Chateau, H., Degueurce, C., Denoix, J.-M., horses with superficial digital flexor Early development of gait asymmetries in
2006. Three-dimensional of the distal tendinitis. Am. J. Vet. Res. 61, 191–196. trotting standardbred colts. Equine Vet. J.
forelimb in horses trotting on a treadmill Colborne, G.R., Lanovaz, J.L., Sprigings, E.J., 19, 189–191.
and effects of elevation of heel and toe. Schamhardt, H.C., Clayton, H.M., 1997. Dutto, D.J., Hoyt, D.F., Clayton, H.M., Cogger,
Equine Vet. J. 38, 164–169. Joint moments and power in equine gait: a E.A., Wickler, S.J., 2004a. Moments and
Chateau, H., Girard, D., Degueurce, C., preliminary study. Equine Vet. J. Suppl 2:, powers generated by the horse (equus
Denoix, J.-M., 2003. Methodological 33–36. caballus) hind limb during jumping. J. Exp.
considerations for using a kinematic Colborne, R.R., Birtles, D.-M., Cacchione, I.-C., Biol. 207, 667–674.
analysis system based on ultrasonic 2001. Electromyographic and kinematic Dutto, D.J., Hoyt, D.F., Cogger, E.A., Wickler,
triangulation. Innovation Technol. Biol. indicators of fatigue in horses: a S.J., 2004b. Ground reaction forces in
Méd. 24, 69–78. pilot study. Equine Vet. J. 33 (suppl), horses trotting up an incline and on the
Clayton, H.M., Bialski, D., Lanovaz, J.L., 89–93. level over a range of speeds. J. Exp. Biol.
Mullineaux, D.L., 2003a. Reliability of a Corley, J.M., Goodship, A.E., 1994. Treadmill 207, 3507–3514.
technique to measure postural sway in training induced changes to some Faber, M., Johnston, C., van Weeren, P.R.,
horses. Am. J. Vet. Res. 64, 1354–1359. kinematic variables measured at the canter Barneveld, A., 2002. Repeatability of back
Clayton, H.M., Hodson, E.F., Lanovaz, J.L., in thoroughbred fillies. Equine Vet. J. 17 kinematics in horses during treadmill
2000b. The forelimb in walking horses: 2. (Suppl.), 20–24. locomotion. Equine Vet. J. 34, 235–241.
Net joint moments and joint powers. Crevier-Denoix, N., Ravary, B., Defontaine, M., Fioretti, S., Jetto, L., 1989. Accurate derivative
Equine Vet. J. 32, 295–299. Lematre, M., Denoix, J.-M., Pourcelot, P., estimation from noisy data: a state-space
Clayton, H.M., Hodson, E.F., Lanovaz, J.L., 2004. Use and application of a non- approach. Int. J. Systems Sci. 20, 33–53.
Colborne, R.G., 2001. The hind limb in invasive procedure for measuring loads Frederick, F.H. Jr, Henderson, J.M., 1970.
walking horses: 2. Net joint moments and supported by the superficial flexor tendon, Impact force measurement using
joint powers. Equine Vet. J. 33, 44–48. in the moving horse: interest for the preloaded transducers. Am. J. Vet. Res. 31,
Clayton, H.M., Lanovaz, J.L., Schamhardt, evaluation of corrective shoes and the soils. 2279–2283.
H.C., Willemen, M.A., Colborne, G.L., Proceed. 30ème Journée de la Recherche Fredricson, I., Drevemo, S., 1972.
1998. Net joint moments and powers in Équine, Paris, pp. 1–12. Methodological aspects of kinematics of
the equine forelimb in the stance phase of Davies, H.M., 2005. The timing and the joints in the forelimbs of fast moving
the trot. Equine Vet. J. 30, 384–389. distribution of strains around the surface of horses. Acta Vet. Scand. 37, 93–136.
Clayton, H.M., Nauwelaerts, S., 2012. Is a the midshaft of the third metacarpal bone Fredricson, I., Drevemo, S., Dalin, G., et al.,
single force plate adequate for during treadmill exercise in one 1983. Treadmill for equine locomotion
stabilographic analysis in horses? Equine thoroughbred racehorse. Austr. Vet. J. 83, analysis. Equine Vet. J. 15, 111–115.
Vet. J. 44, 550–553. 157–162. Freuhwirth, B., Peham, C., Scheidl, M.,
Clayton, H.M., Schamhardt, H.C., Lanovaz, Davies, H.M.S., McCarthy, R.N., Jeffcott, L.B., Schobeberger, H., 2004. Evaluation of
J.L., Colborne, G.R., Willemen, M.A., 1993. Strain in the equine metacarpus pressure distribution under an english

56
 References

saddle at walk, trot and canter. Equine Vet. in relation to length change of the Keegan, K.G., Satterley, J.M., Skubic, M.,
J. 36, 754–757. muscle-tendon unit. J. Exp. Biol. 208, Yonezawa, Y., Cooley, J.M., Wilson, D.A., et
Galloux, P., Richard, N., Dronka, T., et al., 1175–1190. al., 2005. Use of gyroscopic sensors for
1994. Analysis of equine gait using Hoyt, L., Wickler, S.J., Cogger, E.A., 2000. Time objective evaluation of trimming and
three-dimensional accelerometers fixed on of contact and step length: the effect of shoeing to alter time between heel and toe
the saddle. Equine Vet. J. 17 (Suppl.), limb length, running speed, load carrying lift-off at end of the stance phase in horses
44–47. and incline. J. Exp. Biol. 203, 221–227. walking and trotting on a treadmill. Am. J.
Geutjens, C.A., Clayton, H.M., Kaiser, L.J., Huppertz, H.J., Disselhorst-Klug, C., Silny, J., Vet. Res. 66, 2046–2054.
2008. Forces and pressure beneath the Rau, G., Heimann, G., 1997. Diagnostic Keegan, K.G., Wilson, D.A., Wilson, D.J., et al.,
saddle during mounting from the ground yield of non invasive high spatial 1998. Evaluation of mild lameness in horse
and from a raised mounting platform. Vet. resolution EMG in neuromuscular diseases. trotting on a treadmill by clinicians and
J. 175, 332–337. Muscle Nerve 20, 1360–1370. interns or residents and correlation of their
Gingerich, D.A., Auer, J.A., Fackelman, G.E., Ishihara, A., Bertone, A.L., Rajala-Schultz, P.J., assessments with kinematic gait analysis.
1979. Force plate studies on the effect of 2005. Association between subjective Am. J. Vet. Res. 59, 1370–1377.
exogenous hyaluronic acid on joint lameness grade and kinetic gait parameters Keegan, K.G., Yonezawa, Y., Pai, P.F., Wilson,
function in equine arthritis. J. Vet. in horses with experimentally induced D.A., Kramer, J., 2004. Evaluation of a
Pharmacol. Therap. 2, 291–298. forelimb lameness. Am. J. Vet. Res. 66, sensor-based system of motion analysis for
Gomez Alvarez, C.B., Rhodin, M., Bobbert, 1805–1815. detection and quantification of forelimb
M.F., Meyer, H., Weishaupt, M.A., van Jansen, M.O., Schamhardt, H.C., van den and hind limb lameness in horses. Am. J.
Johnston, C., et al., 2006. The effect of Bogert, A.J., Hartman, W., 1998. Vet. Res. 65, 665–670.
head and neck position on the Mechanical properties of the tendinous Keg, P.R., van den Belt, A.J.M., Merkens, H.W.,
thoracolumbar kinematics in the unridden equine interosseus muscle affected by in Barneveld, A., Dik, K.J., 1992. The effect of
horse. Equine Vet. J. Suppl. 36, 445–451. vivo transducer implantation. J. Biomech. regional nerve blocks on the lameness
Grood, E.S., Suntay, W.J., 1983. A joint 31, 485–490. caused by collagenase induced tendonitis
coordinate system for the clinical Jansen, M.O., van den Bogert, A.J., in the midmetacarpal region of the horse: a
description of three-dimensional motions: Schamhardt, H.C., 1993. In vivo tendon study using gait analysis, and
applications to the knee. J. Biomech. Eng. forces in the forelimb of ponies at ultrasonography to determine tendon
105, 136–144. the walk, validated by ground reaction healing. J. Vet. Med. A39, 349–364.
Hansen, M.S., Buchner, H.H.F., Haller, J., force measurements. Acta Anat. 146, Khumsap, S., Clayton, H.M., Lanovaz, J.L.,
Windischbauer, G., 2005. Evaluation using 162–167. 2001. Effect of walking velocity ground
hoof wall strain gauges of a therapeutic Jansen, M.O., van Raaij, J.A.G.M., van den reaction force variables in the hind limb of
shoe and a hoof cast with a heel wedge as Bogert, A.J., Schamhardt, H.C., Hartman, clinically normal horses. Am. J. Vet. Res.
potential supportive therapy for horses W., 1992. Quantitative analysis of 62, 901–906.
with laminitis. Vet. Surg. 34, 630–636. computer-averaged electromyographic Khumsap, S., Clayton, H.M., Lanovaz, J.L.,
Hartman, W., Schamhardt, H.C., Lammertink, profiles of intrinsic limb muscles in ponies Bouchey, M., 2002. Effect of walking
J.L.M.A., Badoux, D.M., 1984. Bone strain at the walk. Am. J. Vet. Res. 53, velocity on forelimb kinematics and
in the equine tibia: an in vivo strain gauge 2343–2349. kinetics. Equine Vet. J. Suppl. 34, 325–329.
analysis. Am. J. Vet. Res. 45, 880–884. Jeffcott, L.B., Holmes, M.A., Townsend, Khumsap, S., Lanovaz, J.L., Clayton, H.M.,
Hassan, E.A., Jenkyn, T.R., Dunning, C.E., H.G.G., 1999. Validity of saddle pressure 2004. Three-dimensional kinematic
2007. Direct comparison of kinematic data measurements using force-sensing array analysis of horses with induced tarsal
collected using an electromagnetic tracking technology – preliminary studies. Vet. J. synovitis. Equine Vet. J. 36, 659–664.
system versus a digital optical system. J. 158, 113–119. Khumsap, S., Lanovaz, J.L., Rosenstein, D.L.,
Biomech. 40, 930–935. Johnston, C., Holm, K.R., Erichsen, C., Eksell, Byron, C., Clayton, H.M., 2003. Effect of
Hisao, H., Guan, J., Weatherly, M., 2002. P., Drevemo, S., 2004. Kinematic unilateral synovitis of distal intertarsal and
Accuracy and precision of two in-shoe evaluation of the back in fully functioning tarsometatarsal joints on sagittal plane
pressure measurement systems. Ergonomics riding horses. Equine Vet. J. 36, 495–498. kinematics and kinetics of trotting horses.
45, 537–555. Judy, C.E., Galuppo, L.D., Snyder, J.R., Willits, Am. J. Vet. Res. 64, 1491–1495.
Hjertén, G., Drevemo, S., 1987. A method to N.H., 2001. Evaluation of an in-shoe Kimura, J., 1984. Electrodiagnosis in disease of
analyze forces and moments in the pressure measurement system in horses. nerve and muscle: principles and practice.
extremities of the horse during the stance Am. J. Vet. Res. 62, 23–28. F.A. Davis, Philadelphia.
phase at the trot. In: Gillespie, J.R., Kai, M., Aoki, O., Hiraga, A., Oki, H., Tokuriki, Kimura, J., 1989. Techniques and normal
Robinson, N.E. (Eds.), Equine exercise M., 2000. Use of an instrument findings. In: Electrodiagnosis in diseases of
physiology 2. Davis, Calif: ICEEP sandwiched between the hoof and shoe to nerve and muscle: principles and practice,
Publications, pp. 587–598. measure vertical ground reaction forces and second. ed. F.A.Davis and Co.,
Hochmann, D., Diesing, P., Boenick, U., 2002. three-dimensional acceleration at the walk, Philadelphia. pp. 221–224.
Evaluierung der messmethoden zur trot, and canter in horses. Am. J. Vet. Res. Komi, P.V., Ishikawa, M., 2007. In vivo
bewertung des therapeutischen nutzens von 61, 979–985,. function of human achilles and patella. In:
antidekubitus-systemen. Biomedizinische Keegan, K.G., Baker, G.J., Boero, M.J., Woo, S., Renstrom, P., Arnoczky, S. (Eds.),
Technik 4, 816–819. Pijanowski, G.J., Phillips, J.W., 1992. Tendinopathy in athletes. Blackwell
Hodson-Tole, E., 2006. Effects of treadmill Evaluation of support bandaging during Publishing, pp. 101–122.
inclination and speed on forelimb muscle measurement of proximal sesamoidean Korsgaard, E., 1982. Muscle function in the
activity and kinematics in the horse. strain in horses by use of a mercury strain forelimb of the horse. An
Equine Compar. Exerc. Physiol. 3, 61–72. gauge. Am. J. Vet. Res. 53, 1203–1208. electromyographical and kinesiological
Hof, A.L., 1984. EMG and muscle force: an Keegan, K.G., Pai, P.F., Wilson, D.A., et al., study. PhD thesis, Kopenhagen.
introduction. Human Movement Sci. 3, 2001. Signal decomposition method of Lanovaz, J.L., Clayton, H.M., Colborne, G.R.,
119–153. evaluating head movement to measure Schamhardt, H.C., 1999. Forelimb
Hoyt, D.F., Wickler, S.J., Biewener, A.A., induced forelimb lameness in horses kinematics and joint moments during the
Cogger, E.A., De La Paz, K., 2005. In vivo trotting on a treadmill. Equine Vet. J. 33, swing phase of the trot. Equine Vet. J. 30
muscle function vs speed I. Muscle strain 446–451. (Suppl.), 235–239.

57
 2 Measurement techniques for gait analysis

Lanovaz, J.L., Clayton, H.M., Watson, L.G., McIlwraith, C.W., Fortier, L.A., Frisbie, D.D., Naylor, J.M., Robinson, J.A., Crichlow, E.C.,
1998. In vitro attenuation of impact shock Nixon, A.J., 2011. Equine models of Steiss, J.E., 1992. Inheritance of myotonic
in equine digits. Equine Vet. J. 26 (Suppl.), articular cartilage repair. Cartilage 2, discharges in american quarter horses and
96–102. 317–326. the relationship to hyperkalemic periodic
Lanovaz, J.L., Khumsap, S., Clayton, H.M., McLaughlin, R.M., Gaughan, E.M., Roush, K., paralysis. Can. J. Vet. Res. 56, 62–66.
2004. Quantification of three-dimensional Skaggs, C., 1996. Effects of subject velocity Nicodemus, M.C., Lanovaz, J.L., Corn, C.,
skin displacement artefacts on the equine on ground reaction force measurements Clayton, H.M., 1999. The application of
tibia and third metatarsus. Equine Compar. and stance times in clinically normal virtual markers to a joint coordinate system
Exerc. Physiol. 1, 141–150. horses at the walk and trot. Am. J. Vet. Res. for equine three-dimensional motions.
Lanyon, L.E., 1976. The measurement of bone 57, 7–11. Proc. Equine Nutr. Physiol. Soc. 16, 24–25.
strain in vivo. Acta Orthop. Belg. 42 (Suppl. Meershoek, L.S., Lanovaz, J.L., 2001c. Niki, Y., Ueda, Y., Yoshida, K., Masumitsu, H.,
1), 98–108. Sensitivity analysis and application to 1982. A force plate study in equine
Lawson, S.E.M., Chateau, H., Pourcelot, P., trotting of a noninvasive method to biomechanics. 2. The vertical and fore-aft
Denoix, J.-M., Crevier-Denoix, N., 2007a. calculate flexor tendon forces in the components of floor reaction forces and
Effect of toe and heel elevation on equine forelimb. Am J. Vet. Res. 62, motion of equine limbs at walk and trot.
calculated tendon strains in the horse and 1594–1598. Bull. Equine Res. Inst. 19, 1–17.
the influence of the proximal Meershoek, L.S., Schamhardt, H.C., Roepstorff, Nunamaker, D.M., Butterweck, D.M., Provost,
interphalangeal joint. J. Anat. 210, L., Johnston, C., 2001b. Forelimb tendon M.T., 1990. Fatigue fractures in
583–591. loading during jump landings and the thoroughbred racehorses: relationships
Lawson, S.E.M., Chateau, H., Pourcelot, P., influence of fence height. Equine Vet. J. with age, peak bone strain, and training. J.
Denoix, J.-M., Crevier-Denoix, N., 2007b. Suppl. 33, 6–10. Orthop. Res. 8, 604–611.
Sensitivity of an equine distal limb model Meershoek, L.S., van cen Bogert, A.J., Oosterlinck, M., Pille, F., Huppes, T., Gasthuys,
to perturbations in tendon paths, origins Schamhardt, H.C., 2001a. Model F., Back, W., 2010. Comparison of pressure
and insertions. J. Biomech. 40, 2510–2516. formulation and determination of in vitro plate and force plate gait kinetics in sound
Leach, D.H., Dyson, S., 1988. Instant centres parameters of a noninvasive method to Warmbloods at walk and trot. Vet. J. 186,
of rotation of equine limb joints and their calculate flexor tendon forces in the 347–351.
relationship to standard skin marker equine forelimb. Am. J. Vet. Res. 62, Oosterlinck, M., Pille, F., Sonneveld, D.C.,
locations. Equine Vet. J. 6 (Suppl.), 1585–1593. Oomen, A.M., Gasthuys, F., Back, W., 2012.
113–119. Merkens, H.W., Schamhardt, H.C., 1988. Contribution of dynamic calibration to the
Lewis, A., Stewart, C., Postans, N., Trevelyan, Evaluation of equine locomotion during measurement accuracy of a pressure plate
J., 2007. Development of an instrumented different degrees of experimentally induced system throughout the stance phase in
pole test for use as a gait laboratory quality lameness 1: lameness model and sound horses. Vet. J. 193, 471–474.
check. Gait and Posture 26, 317–322. quantification of ground reaction force Parsons, K., Pfau, T., Wilson, A.M., 2008. High
Licka, T., Frey, A., Peham, C., 2009. patterns of the limbs. Equine Vet. J. 6 speed gallop locomotion in the
Electromyographic activity of the (Suppl.), 99–106. thoroughbred racehorse. II The effect of
longissimus dorsi muscles in horses when Merkens, H.W., Schamhardt, H.C., Hartman, incline on centre of mass movement and
walking on a treadmill. Vet. J. 180, 71–76. W., Kersjes, A.W., 1986. Ground reaction mechanical energy fluctuation. J. Exp. Biol.
Licka, T., Peham, C., 1998. An objective force patterns of dutch warmblood horses 211, 945–956.
method for evaluating the flexibility of the at normal walk. Equine Vet. J. 18, 207–214. Peham, C., Licka, T., Girtler, D., Scheidl, M.,
back of standing horses. Equine Vet. J. 30, Merkens, H.W., Schamhardt, H.C., Hartman, 2001. The influence of lameness on equine
412–415. W., Kersjes, A.W., 1988. The use of stride length consistency. Vet. J. 162,
Licka, T., Peham, C., Frey, A., 2004. horse(INDEX), a method of analysing the 153–157.
electromyographic activity of the ground reaction force patterns of lame and Peham, C., Licka, T., Mayr, A., Scheidl, M.,
longissimus dorsi muscles in horses during normal gaited horses at the walk. Equine Girtler, D., 1998. Speed dependency of
trotting on a treadmill. Am. J. Vet. Res. 65, Vet. J. 20, 29–36. motion pattern consistency. J. Biomech. 31,
155–158. Merkens, H.W., Schamhardt, H.C., Osch, 769–772.
Linford, R.L., 1994. Camera speeds for G.J.V.M., Bogert, A.J., 1993a. Ground Peham, C., Licka, T., Schobesberger, H.,
optoelectronic assessment of stride-timing reaction force patterns of dutch warmblood Meschan, E., 2004. Influence of the rider
characteristics in horses at the trot. Am. J. horses at normal trot. Equine Vet. J. 25, on the variability of the equine gait. Hum.
Vet. Res. 55, 1189–1195. 134–137. Move. Sci. 23, 663–671.
Lynch, J.A., Clayton, H.M., Mullineaux, D.R., Merkens, H.W., Schamhardt, H.C., van Osch, Peloso, J.G., Stick, J.A., Caron, J.P., Peloso,
2005. The reliability of force platform data G.J.V.M., Hartman, W., 1993b. Ground P.M., Soutas-Little, R.W., 1993. Computer-
from trotting horses. Equine Compar. reaction force patterns of dutch assisted three-dimensional gait analysis of
Exerc. Physiol. 2, 129–132. warmbloods at canter. Am. J. Vet. Res. 54, amphotericin-induced carpal lameness in
Macgregor, J., Graf von Schweinitz, D., 2006. 670–674. horses. Am. J. Vet. Res. 54, 1535–1543.
Needle electromyographic activity of Meschan, E.M., Peham, C., Schobesberger, H., Perino, V.V., Kawcak, C.E., Frisbie, D.D., Reiser,
myofascial trigger points and control sites Licka, T.F., 2007. The influence of the width R.F., McIlwraith, C.W., 2007. The accuracy
in equing=e cleidobrachialis muscle – an of the saddle tree on the forces and the and precision of an equine in-shoe
observational study. Acupunct. Med. 24, pressure distribution under the saddle. Vet. pressure measurement system as a tool for
61–70. J. 173, 578–584. gait analysis. J. Equine Vet. Sci. 27,
Mayhew, I.G., Whitlock, R.H., de la Hunta, A., Morris, E.A., Seeherman, H.J., 1987. 161–166.
1978. Spinal cord disease of the horse, Redistribution of ground reaction forces in Pfau, T., Ferrari, M., Parsons, K., Wilson, A.J.,
electromyograpic and radiographic studies. experimentally induced carpal lameness. 2008. A hidden markov model-based stride
Cornell Vet. 68 (suppl), 44–62. Equine Exerc. Physiol. 2, 553–563. segmentation technique applied to equine
McGuigan, M.P., Wilson, A.M., 2003. The Mullineaux, D.R., Clayton, H.M., Gnagey, inertial sensor trunk movement data. J.
effect of gait and digital flexor muscle L.M., 2004. Effects of offset-normalizing Biomech. 41, 216–220.
activation on limb compliance in the techniques on variability in motion Pfau, T., Robilliard, J.J., Weller, R., Jespers, K.,
forelimb of the horse equus caballus. J. analysis data. J. Appl. Biomech. 20, Eliasher, E., Wilson, A.J., 2007. Assessment
Exp. Biol. 206, 1325–1336. 177–184. of mild hind limb lameness during over

58
 References

ground locomotion using linear kinematics of the equine hindleg, I: In vivo Roszek, B., van Loon, P., Weinans, H., Huiskes,
discriminant analysis of inertial sensor tendon load and force plate measurements. R., 1993. In vivo measurements of the
data. Equine Vet. J. 39, 407–413. Am. J. Vet. Res. 49, 1344–1352. loading conditions on the tibia of the goat.
Pfau, T., Witte, T.H., Wilson, A.J., 2005. Riemersma, D.J., van den Bogert, A.J., 1993. A Acta Anat. 146, 188–192.
A method for deriving displacement method to estimate the initial length of Ryan, C.T., Dallap Schaer, B.L., Nunamaker,
data during cyclical movement using an equine tendons. Acta Anat. 146, 120–122. D.L., 2006. A novel wireless data
inertial sensor. J. Exp. Biol. 208, Riemersma, D.J., van den Bogert, A.J., Jansen, acquisition system for the measurement of
2503–2514. M.O., Schamhardt, H.C., 1996a. Tendon hoof accelerations in the exercising horse.
Pfau, T., Witte, T.H., Wilson, A.J., 2006. Centre strain in the forelimbs as a function of gait Equine Vet. J. 38, 671–674.
of mass movement and mechanical energy and ground characteristics, and in vitro Rybicki, E.F., Simonen, F.A., Mills, E.J., et al.,
fluctuation during gallop locomotion in limb loading in ponies. Equine Vet. J. 28, 1974. Mathematical and experimental
the thoroughbred racehorse. J. Exp. Biol. 133–138. studies on the mechanics of plated
209, 3742–3757. Riemersma, D.J., van den Bogert, A.J., Jansen, transverse fractures. J. Biomech. 7,
Platt, D., Wilson, A.M., Timbs, A., Wright, M.O., Schamhardt, H.C., 1996b. Influence 377–384.
I.M., Goodship, A.E., 1994. Novel of shoeing on ground reaction forces and Schache, A.G., Baker, R., 2007. On the
force transducer for the measurement of tendon strains in the forelimbs of ponies. expression of joint moments during gait.
tendon forces in vivo. J. Biomech. 27, Equine Vet. J. 28, 126–132. Gait Posture 25, 440–452.
1489–1493. Riemersma, D.J., van den Bogert, A.J., Schamhardt, H.C., 1996. Quantitative analyses
Podell, M., Valentine, B.A., Cummings, J.F., et Schamhardt, H.C., Hartman, W., 1988b. of equine locomotion. In: Ossenkopp,
al., 1995. Electromyography in acquired Kinetics and kinematics of the equine K.-P., Kavakiers, M., Sanberg, P.R. (Eds.),
equine motor neuron disease. Prog. Vet. hindleg, II: In vivo tendon strain and Measuring movement and locomotion:
Neurol. 6, 128–134. joint kinematics. Am. J. Vet. Res. 49, from invertebrates to humans. Springer-
Pourcelot, P., Defontaine, M., Ravary, B., 1344–1352. Verlag, Heidlberg, pp. 189–211.
Lemâtre, M., Crevier-Denoix, N., 2005. A Robert, D., Audigie, F., Valette, J.-P., Pourcelot, Schamhardt, H.C., Hartman, W., 1982.
non-invasive method of tendon force P., Denoix, J.-M., 2001a. The effects of Automatic zero strain compensation of in
measurement. J. Biomechanics 38, treadmill inclination and speed on the vivo bone strain recordings. J. Biomech. 15,
2124–2129. activity of three trunk muscles in the 621–624.
Pratt, G.W. Jr, O’Connor, J.T. Jr, 1976. Force trotting horse. Equine Vet. J. 33, 466–472. Schamhardt, H.C., Hartman, W., Lammertink,
plate studies of equine biomechanics. Am. Robert, D., Valette, J.-P., Degueurce, C., J.L.M.A., 1985. In vivo bone strain in the
J. Vet. Res. 37, 1251–1255. Denoix, J.-M., 1999. Correlation between equine tibia before and after transection of
Preuschoft, H., Witte, H., Recknagel, St, Baer, surface electromyograph and kinematics of the peroneus tertius muscle. Res. Vet. Sci.
H., Lesch, C., Wuthrich, M., 1999. “Ueber the hind limb of horses at trot on a 39, 139–144.
die wirkung gebrauchlicher zaeumungen treadmill. Cell Tissues Organs 165, Schamhardt, H.C., Merkens, H.W., 1987.
auf das pferdes,” Deutsche Tierartz. 113–122. Quantification of equine ground reaction
Wochensch. 106, 169–175. Robert, D., Valette, J.-P., Denoix, J.-M., 2000. force patterns. J. Biomech. 20, 443–446.
Ramakrishnan, H.K., Kadaba, M.P., 1991. On The effects of treadmill inclination and Schamhardt, H.C., van den Bogert, A.J.,
the estimation of joint kinematics during speed on the activity of two hind limb Hartman, W., 1993a. Measurement
gait. J. Biomech. 24, 969–997. muscles in the trotting horse. Equine Vet. J. techniques in animal locomotion analysis.
Ratzlaff, M.H., Grant, B.D., Adrian, M., 32, 312–317. Acta Anat. 146, 123–129.
Feeney-Dixon, C., 1979. Evaluation of Robert, D., Valette, J.-P., Denoix, J.-M., 2001b. Schamhardt, H.C.C., van den Bogert, A.J.,
equine locomotion using Effects of treadmill inclination and speed Lammertink, J.L.M.A., 1994. Power transfer
electrogoniometry and cinematography: on the mechanics of the back in the from treadmill engine to athlete. Proc. Can.
research and clinical applications. Proc. trotting saddlehorse. Equine Vet. J. 33 Soc. Biomech. 306–307.
Am. Assoc. Equine Practnrs. 25, 381. (suppl), 154–159. Schamhardt, H.C., Merkens, H.W., Vogel, V.,
Ratzlaff, M.H., Grant, B.D., Hyde, M.L., Balch, Robert, D., Valette, J.-P., Pourcelot, P., Audigie, Willekens, C., 1993b. External loads on the
O.K., 1994. Rupture of the distal R., Denoix, J.-M., 2002. Effects of trotting limbs of the jumping horse at take-off and
sesamoidean ligaments of a horse: vertical speed on muscle activity and kinematics in landing. Am. J. Vet. Res. 54, 675–680.
forces and temporal components of the saddlehorses. Equine Vet. J. 34 (suppl), Schuurman, S.O., Kersten, W., Weijs, W.A.,
strides before, during and after injury. J. 295–301. 2003. The equine hind limb is actively
Equine Vet. Sci. 14, 45–52. Roepstorff, L., Drevemo, S., 1993. Concept of stabilized during standing. J. Anat. 202,
Ratzlaff, M.H., Hyde, M.L., Grant, B.D., Balch, a force-measuring horseshoe. Acta Anat. 355–362.
O., Wilson, P.D., 1990. Measurement of 146, 114–119. Seeherman, H.J., Morris, E.A., Fackelman, G.E.,
vertical forces and temporal components of Roepstorff, L., Johnston, C., Drevemo, S., 1987. Computerized force plate
the strides of galloping horses using 1994. The infleunces of different treadmill determination of equine weightbearing
instrumented shoes. J. Equine Vet. Sci. 10, constructions on ground reaction forces as profiles. Equine Exerc. Physiol. 2, 537–552.
23–25. determined by the use of a force-measuring Sha, D.H., Mullineaux, D.R., Clayton, H.M.,
Ratzlaff, M.H., Wilson, P.D., Hutton, D.V., shoe. Equine Vet. J. 17 (Suppl.), 71–74. 2004. Three-dimensional analysis of
Slinker, B.K., 2005. Relationships between Roland, E.S., Hull, M.S., Stover, S., 2005. patterns of skin displacement over the
hoof-acceleration patterns of galloping Design and demonstration of a equine radius. Equine Vet. J. 36, 665–670.
horses and dynamic properties of the track. dynamometric horseshoe for measuring Sloet van Oldruitenborgh-Oosterbaan, M.M.,
Am. J. Vet. Res. 66, 589–595. ground reaction loads of horses during Barneveld, A., 1995. Comparison of the
Ravary, B., Pourcelot, P., Bortolussi, C., racing conditions. J. Biomech. 38, workload of dutch warmblood horses
Konieczka, S., Crevier-Denoix, N., 2004. 2102–2112. ridden normally and on a treadmill. Vet.
Strain and force transducers used in human Rollot, Y., Lecuyer, E., Chateau, H., Crevier- Rec. 137, 136–139.
and veterinary tendon and ligament Denoix, N., 2004. Development of a 3D Sloet van Oldruitenborgh-Oosterbaan, M.M.,
biomechanical studies. Clin. Biomech. 19, model of the equine distal forelimb and of Barneveld, A., Schamhardt, H.C., 1997.
433–447. a GRF shoe for noninvasive determination Effects of treadmill inclination on
Riemersma, D.J., Schamhardt, H.C., Hartman, of in vivo tendon and ligament loads and kinematics of the trot in dutch warmblood
W., Lammertink, J.L., 1988a. Kinetics and strains. Equine Vet. J. 36, 677–682. horses. Equine Vet. J. Suppl. 23, 71–75.

59
 2 Measurement techniques for gait analysis

Soderkvist, I., Wedin, P.A., 1993. Determining van Weeren, P.R., van den Bogert, A.J., veterinary medicine and in veterinary
the movements of the skeleton using Barneveld, A., 1990b. Quantification of research – a review. Vet. Quart. 21, 3–7.
well-configured markers. J. Biomech. 26, skin displacement in the proximal parts of Wickler, S.J., Hoyt, D.F., Biewener, A.A.,
1473–1477. the limbs of the walking horse. Equine Vet. Cogger, E.A., De La Paz, K.L., 2005. In vivo
Spoor, C.W., Veldpaus, F.E., 1980. Rigid body J. 9 (Suppl.), 110–118. muscle fnuction vs speed. II. Muscle
motion calculated from spatial co-ordinates van Weeren, P.R., van den Bogert, A.J., function trotting up an incline. J. Exp. Biol.
of markers. J. Biomech. 13, 391–393. Barneveld, A., 1992. Correction models for 208, 1191–1200.
Stephens, P.R., Nunamaker, D.M., Butterweck, skin displacement in equine kinematic gait Wijnberg, I.D., Back, W., de Jong, M., Zuidhof,
D.M., 1989. Application of a hall-effect analysis. J. Equine Vet. Sci. 12, 178–192. M.C., van den Belt, A.J.M., van der Kolk,
transducer for measurement of tendon Wakeling, J.M., Ritruechaia, P., Dalton, S., J.H., 2004. The role of electromyography in
strain in horses. Am. J. Vet. Res:. 50, Nankervis, K., 2007. Segmental variation in clinical diagnosis of neuromuscular
1089–1095. the activity and function of the equine locomotor problems in the horse. Equine
Taylor, B.M., Tipton, C.M., Adrian, M., longissimus dorsi muscle during walk and Vet. J. 36, 718–722.
Karpovich, P.V., 1966. Action of certain trot. Comp. Exerc. Physiol. 4, 95–103. Wijnberg, I.D., Franssen, H., van der Kolk,
joints in the legs of the horse recorded Weishaupt, M.A., Hogg, H.P., Wiestner, T., J.H., Breukink, H.J., 2002a. Quantitative
electrogoniometrically. Am. J. Vet. Res. 27, Denoth, J., Stüssi, E., Auer, J.A., 2002. motor unit action potential analysis of the
85–89. Instrumented treadmill for measuring normal subclavian muscle in the horse.
Thomason, J.J., Biewener, A.A., Bertram, J.E.A., vertical ground reaction forces in horses. Am. J. Vet. Res. 63, 198–203.
1992. Surface strains on the equine hoof Am. J. Vet. Res. 63, 520–527. Wijnberg, I.D., Franssen, H., van der Kolk,
wall in vivo: implications for the material Weishaupt, M.A., Wiestner, T., Hogg, H.P., J.H., Breukink, H.J., 2002b. Quantitative
design and functional morphology of the Jordan, P., Auer, J.A., 2004a. Vertical motor unit action potential analysis of
wall. J. Exp. Biol. 166, 145–168. ground reaction force-time histories of skeletal muscles in the warmblood horse.
Thomason, J.J., Bignell, W.W., Sears, W., 2001. sound warmblood horses trotting on a Equine Vet. J. 34, 556–562.
Components of variation of surface hoof treadmill. Vet. J. 168, 304–311. Wijnberg, I.D., van der Kolk, J.H., Franssen,
strain with time. Equine Vet. J. Suppl. 33, Weishaupt, M.A., Wiestner, T., Hogg, H.P., H., Breukink, H.J., 2003. Needle
63–66. Jordan, P., Auer, J.A., 2004b. Compensatory electromyography in the horse compared
Tokuriki, M., Aoki, O., 1995. load redistribution of horses with induced with its principles in man: a review. Equine
Electromyographic activity of the hind limb weightbearing hind limb lameness trotting Vet. J. 35, 9–17.
muscles during the walk, trot and canter. on a treadmill. Equine Vet. J. 36, 727–733. Willemen, M.A., Jacobs, M.W.H., Schamhardt,
Equine Vet. J. 18 (Suppl.), 152–155. Weishaupt, M.A., Wiestner, T., Hogg, H.P., H.C., 1997. In vitro transmission and
Tokuriki, M., Aoki, O., Niki, Y., Kurakawa, Y., Jordan, P., Auer, J.A., 2006a. Compensatory attenuation of impact vibrations in the
Hataya, M., Kita, T., 1989. load redistribution of horses with induced lower forelimb of the horse. In: Willemen,
Electromyographic activity of cubital joint weight-bearing forelimb lameness trotting M.A. (Ed.), Horseshoeing: a biomechanical
muscles in horses during locomotion. Am. on a treadmill. Vet. J. 171, 135–146. analysis. PhD thesis, Utrecht University.
J. Vet. Res. 50, 950–957. Weishaupt, M.A., Wiestner, T., Peinen, K., von, Wilson, A.M., Pardoe, C.H., 2001.
Tokuriki, M., Ohtsuki, R., Kai, M., Hiraga, A., Waldern, N., Roepstorff, L., van Weeren, R., Modification of a force plate system for
Oki, H., Miyahara, Y., et al., 1999. EMG et al., 2006b. Effect of head and neck equine gait analysis on hard road surfaces:
activity of the muscles of the neck and position on vertical ground reaction forces a technical note. Equien Vet. J. Suppl. 33,
forelimbs during different forms of and interlimb coordination in the dressage 67–69.
locomotion. Equine Vet. J. 30 (Suppl), horse ridden at walk and trot on a Winter, D.A., 1990. Kinematics: forces and
231–234. treadmill. Equine Vet. J. Suppl. 36, moments of force. In: Biomechanics and
van den Bogert, A.J., 1989. Computer 387–392. motor control of human movement,
simulation of locomotion in the horse. Weller, R., Pfau, T., Babbage, D., Brittin, E., second ed. John Wiley & Sons, New York,
PhD Thesis, University of Utrecht, Utrecht. May, S.A., Wilson, A.M., 2006. Reliability of pp. 75–102.
van Heel, M.C.V., Barneveld, A., van Weeren, conformational measurements in the horse Witte, T.H., Knill, K., Wilson, A.M., 2004.
P.R., Back, W., 2004. Dynamic pressure using a three-dimensional motion analysis Determination of peak ground reaction
measurements for the detailed study of system. Equine Vet. J. (2006) 38 (7), force from duty factor in the horse
hoof balance: the effect of trimming. 610–615. (equus caballus). J. Exp. Biol. 207,
Equine Vet. J. 36, 778–782. Wentink, G.H., 1978. Biokinetical analysis of 3639–3648.
van Weeren, P.R., van den Bogert, A.J., Back, the movements of the pelvic limb of the Woltring, H.J., 1994. 3-D attitude
W., Bruin, G., Barneveld, A., 1993. horse and the role of the muscles in the representation of human joints: a
Kinematics of the standardbred trotter walk and trot. Anat. Embryol. 152, standardization proposal. J. Biomech. 27,
measured at 6, 7, 8 and 9 m/s on a 261–272. 1388–1414.
treadmill, before and after 5 months Werner, D., Nyikos, S., Kalpen, A., Geuden, M., Zsoldos, R.R., Kotschwar, A., Kotschwar, A.B.,
prerace training. Acta Anat. 146, 154–161. Haas, C., Vontobel, H.D., et al., 2002. Rodriguez, C.P., Peham, C., Licka, T., 2010.
van Weeren, P.R., van den Bogert, A.J., Druckmessungen unter dem sattel: einem Activity of the equine rectus abdominis and
Barneveld, A., 1990a. A quantitative elektronischen sattel-messsystem (Novel oblique external abdominal muscles
analysis of skin displacement in the GmbH). Pferdeheilkunde, 18, 125–140. measured by surface EMG during walk and
trotting horse. Equine Vet. J. 9 (Suppl.), Wessum, R. van, Sloet, M.M., Clayton, H.M., trot on the treadmill. Equine Vet. J. 42
101–109. 1999. Electromyography in the horse in (Suppl. 38), 523–529.

60
 C H A PTER 3 

Signals from materials

Christian Peham

The concept of this chapter follows the signal beginning with mea- Motion of a wheel
suring, then analog to digital conversion, signal processing and,
finally, decision-making. A typical example of digital measurement equipment is a film
camera. Let us consider an example when the camera takes 25 time-
equidistant samples (pictures) of a motion every second. Depend-
Introduction and definition of signals ing on the speed of the moving object this sampling rate may or
may not be sufficient.
In equine motion analysis a signal is a measured physical property In movies there is often a strange effect in moving stagecoaches.
of a movement presented as time-depended parameter or variable. It is clearly visible that the stagecoach moves from left to right. But
It is acquired by different measurement methods, such as kine- the wheels of the stagecoach appear to be revolving in the opposite
matic measurement (Fig. 3.1). This physical parameter can be a direction (Figs 3.4 and 3.5).
coordinate (distance from origin), a force, acceleration, an angle This so-called ‘alias effect’ is the result of under sampling. The
and so on. These signals are so-called bio signals. A bio signal pres- conclusion is that the sampling rate is too low.
ents a time series of a physical parameter of a living being (Shiavi, Figure 3.6 shows another example of a turning wheel with more
1999), and appears naturally in time-continuous form. Figure 3.2 spokes. The conclusion is that in fast movements more than 16
presents a typical signal in time-continuous form. pictures/turn (because there are eight segments) are necessary.
Due to the measuring technique (analog to digital converter The conclusion of the wheel experiment is the Nyquist – Shannon
(ADC)) today nearly every signal is detected as digital signal sampling theorem, which states that the sampling frequency must
(Semmlow, 2004). The ADCs quantify the captured values. The be at least twice the signal frequency to avoid the alias effect
effective range is always a power of 2. So a 12-bit ADC means that (Werner, 2006):
the effective range is divided into 212 (4096) steps. A digital signal Fsample > 2*Fsignal where F is the frequency
consists of values measured at given time points, in most cases Normally oversampling is used during the measurement. For
time-equidistance. Figure 3.3 shows the sampling (measuring) instance, a sampling frequency that is five times higher than the
process of a signal. theoretical minimum sample frequency is used.
The sampling signal represents a camera, which takes a picture of
the motion every 0.008 s (120 Hz sample rate). The result is a
sampled signal with measured values on given time points. Resampling and normalization
Therefore in (equine) motion analysis, signal processing is digital
Often two different measurement systems are synchronized, e.g. a
signal processing (DSP). This knowledge is important for measuring
kinematic system with a force plate or EMG equipment. Both
(sampling frequency) and signal processing.
systems use different sampling frequencies. Since EMG measure-
Normally noise and/or any other disturbances interfere with the
ments need a much higher sampling frequency than the kinematic
desired signal (Semmlow, 2004). Sources of these disturbances may
measurements, the result is two different time scales, i.e. one for
be errors of the measurement equipment, ADC (quantization
kinematic and one for EMG. To solve this problem, resampling is
error), influence of an electrical field in case of EMG or ECG, etc.
needed. If the sample frequency from EMG is reduced (Peham et al.,
DSP is necessary to get the most out of the measured data. The first
2001a,b; Licka et al., 2004), it is comparable to smoothing or low-
step is to set up the measurement equipment in a way that the
pass filtering. If there is a whole-numbered relation between the
digital signal represents the measured signal sufficiently.
two sample frequencies, it is very easy to reduce or add the samples.
The simplest method to reduce the samples is for instance to take
Choosing the sampling frequency only every second or third sample and to add samples to calculate
the values of the new samples by a linear relation between two
(Nyquist – Shannon sample theorem) neighboring samples. Usually the procedure of resampling is done
in two steps. The first step is to fit the curve (e.g. cubic spline, Fourier
Certain preconditions have to be fulfilled to represent a time series, etc.) in the second step; the new samples will be extracted
continuous signal by a digital signal (time-equidistance sampled from the fitted curve. If the sample frequency is to be reduced, it is
values). This characteristic is described by the Nyquist – Shannon necessary to limit the bandwith by a low-pass filter.
sample theorem (Oppenheim & Willsky, 1996; Semmlow, 2004). The effect of reducing the samples (resampling) can be demon-
The last graph of Figure 3.3 shows a sampled signal. Now the strated by the moving average. The moving average is one of the
question arises, is the sample frequency sufficient? oldest and most popular technical analysis tools in motion analysis.

61
 3 Signals from materials

Fig 3.1  Horse on the treadmill during kinematic

data acquisition.

Horizontal coordinate of hoof motion of a horse Analog signal of vertical coordinate of hoof motion
walking on treadmill of a horse walking on treadmill
1200 Vertical coordinate (mm) 100
Horizontal coordinate (mm)

80
1000 60
800 40
20
600 0
400 0 0.25 0.5 0.75 1 1.25
200 Time (s)
0
0 1 2 3 4 5 6 7 8 9 10 Sampling signal (Camera)
Time (s) 1
ADC

Vertical coordinate of hoof motion of a horse 0

walking on treadmill
120
Analog signal of vertical coordinate of hoof motion
Vertical coordinate (mm)

100 of a horse walking on treadmill

80 100
60 90
Vertical coordinate (mm)

80
40 70
20 60
50
0 40
0 1 2 3 4 5 6 7 8 9 10 30
Time (s) 20
10
Fig 3.2  Typical example of bio signals; hoof motion of a horse in walk. 0
0 0.25 0.5 0.75 1
Time (s)
A moving average uses a fixed number of samples. These samples Fig 3.3  The sampling process.
will be averaged to give a new sample. Then the working window
is shifted by one sample.
In case of a moving average of three, the mean is calculated from evident when the duration of a motion cycle is normalized to 100%.
the first three samples ( (s1 + s2 + s3)/3) to give the first new sample. Normalization or a relative time scale is used as it makes the com-
The next sample will be calculated by the mean of the samples parison of different motion cycles easier, and allows averaging of
shifted by one ( (s2 + s3 + s4)/3) and so on (Fig. 3.7). This is a kind multiple movement cycles into a single curve (a so-called ‘ensemble
of resampling with the reduction of the sampling frequency. The average’). The disadvantage is that the absolute time scale is lost.
effect is a shift of the curve by one sample interval. If a moving Sometimes the information of the variation of the duration of
average of five samples is used, the delay will be two sample inter- motion cycle is needed, which is then done before normalization.
vals and so on. This combination of more samples is very similar Figure 3.7 shows a signal and the smoothed signal with a moving
to a reduction of the sample frequency. So if it is used, it must be average of three samples. It is obvious that the smoothing will shift
realized that the signal is now low-pass filtered. A resampling is also the signal.

62
 Signal processing

Motion
Moving average of the motion
1.615 MA9
MA8
MA7 S9
1.61 S8
MA6
S7
MA5
1.605
S6

Amplitude (m)
MA4
S5
1.6
MA3
S4
1.595 MA2
S3
MA1 = (S1+S2+S3)/3
1.59 MA1
MA2 = (S2+S3+S4)/3
S2
MAn = (Sn+Sn+1+Sn+2)/3
S1
1.585
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Time (s)
Fig 3.4  An example of a stagecoach moving from left to right. The sequence
consists of two consecutive pictures. If the wheel turned clockwise for 30° (to Fig 3.7  Moving average.
the right) then to the viewer it may appear to have moved anticlockwise for
15° (to the left), as the two consecutive spokes are 45° apart. So we interpret
(interpolate) a backward motion of the wheel. The actual turn of the wheel is
shown by the blue-colored segment of the wheel. In the second picture the
starting position is shown by red bars.
B
Distance between A and B

Sample Interval

Fig 3.5  Turning wheel of a stagecoach. Fig 3.8  Derivation of a sampled signal.

angles, EMG, ECG and many others. Processing of such signals

? ? includes storage and reconstruction, separation of information
from disturbances and noise. We separate the signal processing in
two steps, i.e. analysis of the time curve and analysis of the fre-
quency domain.

Time curve analysis

A B C
Differentiation
Fig 3.6  (A) An example of a wheel with only one spoke. (B) The spoke has Differentiation is often needed in motion analysis and biomechan-
moved 180°, but it is not clear in which direction the wheel has turned. (C) ics (e.g. to calculate velocity and acceleration from the motion).
More than two pictures are necessary during a whole turn to show the direc- Furthermore, acceleration is needed to calculate the acting forces
tion of the rotation.
(force = mass × acceleration).

Aliasing – anti-aliasing filter Definition of differentiation

The definition of the differentiation can be approached in two dif-
To avoid under sampling it is possible to use an anti-aliasing filter, ferent ways. One is geometrical (as a slope of a curve), which was
which is a kind of low-pass filter. A low-pass filter limits the band- used by Leibniz, the other is physical (as a rate of change), which
width (cuts off high frequencies) of the signal. The cut-off frequency was the method of Newton. Historically, it was a struggle between
is smaller than the half of the sample frequency. Leibniz and Newton for being the pioneer of calculus. The approach
Fsample > 2*Fcut-off = Fsignal of Newton was to calculate the velocity; therefore this method will
So the sample theorem is fulfilled. be shown.

The physical concept of Newton

Signal processing This approach was used by Newton in developing his ‘Classical
Mechanics’. The main idea is the calculation of velocity.
Signal processing is the analysis, interpretation and manipulation Figure 3.8 shows one sample interval of a coordinate motion. The
of signals. Signals of interest are biological signals such as motion, sample rate gives us the time scale, whereas the coordinates give the

63
 3 Signals from materials

distance from A to B. The velocity for each sample interval is calcu- When differentiation is applied to data that contains noise, results
lated by dividing the distance between two samples by the duration can be inaccurate. This will be discussed later.
of a sample interval. The mean velocity for each interval is the Positive velocity indicates that the distance is increasing, whereas
quotient of the distance and the elapsed time. negative velocity stands for the decreasing distance.
distance between two samples In Figure 3.9, it can be seen that when the head goes up the
Vmean = . velocity is positive. When the head goes down the velocity is
duration of a sample interval
negative.
The motion is represented by the slope of a straight line between At zero velocity, the motion reaches an extreme value, i.e.
two points. If the sample rate is infinitive, the straight line is the maximum or minimum. When the velocity changes from positive
tangent to this curve at a given time point. In the real world the to a negative value, this indicates a maximum. Whereas, when it
sample rate is always finite. So in DSP it is possible to calculate the changes from negative to a positive value it indicates a minimum.
mean velocity between two consecutive measures. Time series of
velocity can be calculated by repeating this step for all intervals.
Integration
Phase-plane analysis (practical use of differentiation) Integration is very important in motion analysis. Additionally,
Phase-plane analyses are used to show the stability of a system or many powerful mathematical tools are based on integration, e.g.
a motion. Examples are stability of equine gait on a treadmill differential equations are the direct consequence of the develop-
(Peham et al., 1998), harmony of horse and rider (Peham et al., ment of integration. Calculation of impulse from the time curve of
2001c) and stability of coupling via saddle of horse and rider force (Osterlinck et al., 2009), integrated EMG (Wijnberg et al.,
(Peham et al., 2004). A phase-plane is a graph of a signal (motion) 2009), and computation of speed from acceleration (Galloux et al.,
versus its derivative (speed). 1994) are a few examples of integration in motion analysis.
Figure 3.9 shows the motion (first graph) and the velocity (second The integration of a function or data involves computing the area
graph) of vertical motion of the head of the horse. beneath the time curve. In most cases the time interval is constant,
because of constant sample rate of measurement equipment, e.g.
camera, force plate, EMG, and accelerometer, etc. It makes it very
Increasing, decreasing and finding a local maximum or
easy to determine the area under the expected curve between two
the minimum (extreme values) sample values.
The derivative of data or a function can provide information about Figure 3.10 shows how the area under a curve can be calculated.
any increase or decrease in the function and position of extreme The first step is computation of area of the square (side length =
values. first sampled acceleration a1, Δt1 = sample interval) A1 = a1 × Δt1.
Since the differentiation is the local linearization, the curve can The rest of the area is a right-angled triangle (side length a = differ-
be replaced by tangents in every time point. We presented here only ence between the first and the second sampled acceleration a2 − a1,
the simple linear method of differentiation (Fig. 3.8). More point side length b, Δt1 = sample interval) A2 = (a2 − a1) × Δt1/2.
methods (e.g. 5-point) based on the Taylor series are often used.
A s,1 = A  + A ∆ = a1*∆t1 + (a 2 − a1 )*∆t1/2 = (a1 + a 2 )*∆t1/2
∆t
As = ∑ i =1(ai + ai +1 ) ⋅
n
+ abegin (= Sum area under the curve)
2
ai … samples, ∆t … time interval
1.66
The equation above shows the calculation of the area for one inter-
1.64
val and then for n sampled values (n-1 intervals).
1.62
Vertical motion of the
horses' head (m)

1.6
1.58
1.56
1.54 ∆t1=∆t2=∆t
1.52
1.5 Acceleration
0 0.5 1 1.5 2 2.5
Time (s)

1
0.8 a2
0.6
Velocity of vertical motion
of the horses' head (m/s)

0.4
0.2
0
-0.2 (a2-a1) *∆t1/2
a1
-0.4
-0.6 a1 *∆t1
-0.8
-1
0 0.5 1 1.5 2 2.5 ∆t1 ∆t2
Time
Time (s)
Fig 3.9  Differentiation of a signal. Fig 3.10  Calculation of the integral.

64
 Signal processing

Horizontal motion of hoof of a horse

1
walking on concrete surface
3000 0.5
2500
-2
2000 -0.5
Motion (mm)

1500 -1
1000 Stride length
500 Fig 3.12  A rotating pointer giving a sine curve.
0
0.5 1 1.5 2 2.5
-500
Time (s)

Horizontal motion
4500
4000 Horizontal hoof motion
3500 Motion of the treadmill
SL SR SL SR SL SR SL SR
Motion (mm)

3000
2500
A B
2000
1500 Fig 3.13  (A) Vertical head motion of a sound trotting horse. (B) Vertical head
1000 motion of a trotting horse with a supporting forelimb lameness of the left
500 extremity. SL, stance left; SR, stance right.
0
0 0.5 1 1.5 2 2.5
Time (s) chewing motion without the movement of the whole head (Niederl
et al., 2006).
Added motion of hoof and treadmil
10000 Fourier analysis (time domain –
9000 frequency domain)
8000 The Fourier transform is named in honor of the French mathemati-
Motion (mm)

Stride length cian Jean Baptiste Joseph Fourier. The Fourier analysis is the most
7000
important tool in signal processing. It is used in digital image pro-
6000 cessing, medical image processing (CT, MRI, ultrasound, etc.) as
well as in engineering (communication, electrical and mechanical
5000
engineering) and physics (optics, mechanics, etc.). The Fourier
4000 transform separates a signal into sine waves with different frequen-
0 0.5 1 1.5 2 2.5 cies, whereas the inverse Fourier transform composes the time curve
Time (s) from its frequency components. In other words the Fourier trans-
form takes a signal from the time domain to the frequency domain
Fig 3.11  Difference between firm ground and a treadmill motion (coordinate
transform).
and the inverse Fourier transform takes a signal from the frequency
domain to the time domain (Nambu et al., 2000).
Gait is a periodic process, so it can be compared to the motion
of a wheel. The motion of a wheel will be described by a sine func-
tion (Fig. 3.12).
A useful analogy is to interpret the Fourier analysis as a curve
Transformation of the coordinate system fitting tool, which uses only sinusoidal curves (sine and cosine). A
A coordinate transformation is a conversion from one system to very simple example is the vertical motion of the head, withers or
another to describe the same space. For example, in case of a com- sacral bone in walk and trot. Walk and trot are symmetrical gaits
parison of data from a moving horse on firm ground to data from and therefore used for lameness or symmetry evaluation. Figure
a horse moving on a treadmill a coordinate transformation is neces- 3.13A shows the vertical motion of the head of a sound horse. The
sary. In this simple case, the horizontal excursions of the horse and head is lowered during both stance phases, i.e. twice in a motion
the treadmill have to be added. Figure 3.11 shows the difference cycle. This curve can be represented by a sine wave twice the fre-
between the motion signal on firm ground (first graph) and on a quency of the motion cycle.
treadmill (second graph). The third graph shows the transformed If the horse is severely lame on one of its front limbs, the head
motion (from treadmill to firm ground). motion changes. The head is then lowered only during the stance
The description of the skeletal-system movement involves the phase of the sound limb and is less lowered during the stance phase
definition of specific sets of axes or frames that are either global or of the painful limb (Fig. 3.13B). It is now evident that the vertical
local (Cappozzo et al., 2005). The origin of the coordinate system motion of the head of the horse can be represented by a sine wave
has to be fixed on the belt where only the backward-movement is with the frequency of the motion cycle.
taken into account. It makes the trajectories of the fore and back Normally both frequency parts are included in the vertical head
movements comparable. This is a translation of the coordinate motion. So it is easy to calculate the symmetry index from this
system. A universal translation includes translations in the direction sinusoidal wave. The Fourier analysis decomposes the curve to these
of all the three coordinates (x, y, z) and two rotations, e.g. a coor- frequency parts (Peham et al., 1996, 1999, 2001d; Audigie et al.,
dinate system is fixed on the head of the horse to observe the 2002). Sha et al. (2004) used Fourier series to determine the skin

65
 3 Signals from materials

Mag. (dB)

Level 3 Level 3 Level 2 Level 1

0 Apass
fn/8 fn/4 fn/2 fn
Frequency Astop

Fig 3.14  Frequency domain of a discrete wavelet transformation as shown

in Figure 3.15.
0 Fpass Fstop Fs/2 f (Hz)

Fig 3.16  Shows the necessary filter parameters. Apass, ripple range of the
Signal passband; Astop, attenuation of the stopband; Fs, sample frequency; 0-Fpass,
frequency range of the passband; Fstop-Fs/2, frequency range of the
stopband.

HP LP
Filtering
DS/2 DS/2
Filtering of the collected data is crucial in signal processing. The
ideal filter will separate noise from the wanted signal, based on their
Level 1 Coeff. different frequency content. In reality there is never a perfect separa-
tion between noise and signal. Filtering the collected data is, there-
fore, always a compromise between removing too much of the noise
and too much of the signal (Strampp & Vorozhtsov, 2004; Meffert
HP LP
& Hochmuth, 2004; Werner, 2006).
As we will see in the following example, differentiation needs a
DS/2 DS/2 low-pass filter, because of its characteristic to gain higher frequency
parts (Giakas & Baltzopoulos, 1997; Bisseling & Hof, 2006). The
frequency range of equine motion lies between 0 and the half of
Level 2 Coeff. the sampling frequency. In walking humans with a stride frequency
of 1 Hz, the highest harmonics were found to be in the toe and heel
trajectories, and it was found that 99.7% of the signal power was
HP LP contained in the lower seven harmonics (below 6 Hz) (Winter,
1990). In most cases sampling frequencies of 100 Hz are used.
Clayton (1996) stated as a guideline that a digital low-pass filter
DS/2 DS/2 with a cut-off frequency of 10–15 Hz is adequate for most video-
graphic studies on equine gait.
Level 3 Coeff. It is generally assumed that raw surface EMG should be high-pass
filtered with cut-offs of 10–30 Hz to remove motion artifact before
Fig 3.15  The wavelet decomposition of a signal via filter cascades. DS/2, subsequent processing to estimate muscle force (Potvin & Brown,
down-sampling by a factor of two; HP, high-pass; LP, low-pass. 2004).
If peak values are detected (e.g. EMG signal processing determines
the peak envelope), this will lead to an overestimate of the values.
displacement over the radius. Back et al. (2006) showed changes in This can be avoided by the use of low-pass filtering. On the other
the frequency spectrum at impact between differently shod horses. hand the cut-off frequency of the low-pass filter influenced the peak
It is necessary to know the concept of the Fourier transform to values obtained with the largest changes occurring between 15 and
understand the following. 40 Hz. Typically, researchers using lower sample rates have to filter
The Fourier transform is limited when the frequency is changed around 10 Hz and consequently are likely to underestimate peak
during the measurements, i.e. when a horse accelerates or changes angular velocities (Digby et al., 2005).
the gait. In this case when the signals are non-stationary, Fourier It is nearly impossible to determine the time occurrence in unfil-
transform is not sufficient. The first attempt to fix this problem was tered EMG signals (Durkin & Callaghan, 2005; Ives & Wigglesworth,
to use very short time windows, but it has the disadvantage that 2003).
resolution of the frequency scale is very wide (frequency resolution The filter characteristic will be determined in the frequency
= 1/duration of time window). The wavelet transform solves this domain. In motion analysis, filters with a special transfer function
dilemma of resolution to a certain extent (Fig. 3.14). Barrey and are used. Most of this transfer functions are named to honour the
Galloux (1997) used the wavelet transform to analyze acceleration mathematicians who developed the transfer function or shape
of jumping horses. Burn et al. (1997) decomposed the impact accel- (polynomial) of the filter. The most popular are Butterworth (see
eration (instationary signal) of jumping horses. Fig. 3.17), Chebyshev and Cauer.
In short, the wavelet transform uses variable time windows (Fig. The filters are defined by their type or polynomial (Butterworth,
3.14) and therefore results in a time to frequency relation, a variable Chebyshev or Cauer), order of the polynomial (e.g. third order,
frequency scale and an efficient resolution in frequency domain fourth order, etc.) and cut-off frequency used.
(Meffert & Hochmuth, 2004; Strampp & Vorozhtsov, 2004; Polikar, The first step of filter design is to define the cut-off frequency. The
1995). Figure 3.15 shows the analysis process of the signal via cut-off frequency is the frequency where signal is attenuated by 3 dB
wavelets. The signal will be low-pass and high-pass filtered and then (square root of two, i.e. 1.41). It is possible to construct low-pass,
the sample frequency will be reduced by a factor of two. This will high-pass, band-pass and band-stop filters. See Figures 3.16 and
be done iteratively (Fig. 3.15). Keegan et al. (2003) used the wavelet 3.17. In motion analysis the most common type is the Butterworth
transform to detect lameness in horses. filter.

66
 Signal processing

1.2 1.66
N=1 N=3 N=10
1.64
1 N=2 N=4
1.62
0.8 1.6

Motion (m)
0.6 1.58
1.56
0.4
1.54
0.2 1.52
1.5
0 0 0.5 1 1.5 2 2.5
0 2 4 5 6 10
A Time (s)
Fig 3.17  Shows the influence of the order N to the filter characteristic.
2
1.5
2 1 1
H(ω ) = 2N
ω 0.5
1+  
 ωc 

Velocity (m)
0
where N is the order of the polynom, ω standardized frequency, ωc -0.5
standardized cut-off frequency.
-1
In comparison to the other filter types the advantage of a But-
terworth filter is, that the signal distortion is low (linear phase). -1.5
Disadvantages are the relative wide transition range and the high -2
order compared to Chebyshev and Cauer filters.
The term ‘spline’ stems from shipbuilding. A spline is a curve -2.5
0 0.5 1 1.5 2 2.5
which connects given data points (samples) smoothly. So this
method is very useful in motion analysis. It can be used to smooth B Time (s)
data, then it is called smoothing spline, or as a curve fitting tool to
300
fill gaps in the trajectories from incomplete measures (missing
data), or to resample data (curve fitting, calculation of the new
samples). In more detail, a function in most cases, cubic splines 200
(polynom of the third order) will be fitted in the data by optimizing
squared residuals (Vint & Hinrichs, 1996).
The following section will show the necessity of low-pass filtering 100
Acceleration (m/s2)

when the differentiation is calculated.

0
Differentiation in time and frequency domain
How to calculate a derivative is explained earlier. Figure 3.18 shows -100
real data of vertical motion, velocity and acceleration of the head
of a horse. Noise increases with every derivation. The wanted signal
has disappeared in the acceleration. What is the reason for this -200
increase in noise?
In the time domain, the derivative is divided by time/interval.
How are time and frequency related? -300
If the duration of the period of a cosine wave is one second, the Time (s)
frequency of this signal is one Hertz. If the duration is half a C
second, the frequency of this signal is two Hertz. The frequency is Fig 3.18  Effect of derivation of an unsmoothed signal. Noise is increasing.
related to the reciprocal value of the time. If signal is divided In the graphs of velocity and acceleration and the actual signal disappears.
by time in the time domain, the same operation in the frequency
domain will be a multiplication by the frequency (see Fig.
3.19). A signal part, two times derivated at 1 Hz, is multiplied by
1 (12); a signal part, two times derivates at 2 Hz, is multiplied Integration in time and frequency domain
by 4 (22); and a signal part, two times derivates at 3 Hz is multi- The method of calculating an integration has been explained earlier.
plied by 9 (32). Figure 3.20 gives an example of an integrated EMG. See also resa-
This behavior is similar to a high-pass filter. High frequency parts mpling. It is obvious that the noise is reduced by calculating the
are amplified in comparison to low frequency parts. Noise is pro- integral of the EMG. What is the reason for this smoothing?
portional to the bandwidth (frequency range). Now it is clear why In the time domain, the integral was calculated by determining
noise is amplified. So we have to limit the bandwidth by a low-pass the area under the curve. This was done by a multiplication of time
filter to get representative results (Uhlir et al., 1997). Figure 3.9 (amplitude × time). Now we know that time is related to the fre-
shows the same data as Figure 3.18 but band-limited by low-pass quency by the reciprocal value. In the frequency domain, the
filtering. frequency parts are divided by their entire frequency.

67
 3 Signals from materials

Amplitude Amplitude

D=2*π*f 1=1/2*π*f

Frequency Frequency
1 Hz 2 Hz 3 Hz 4 Hz 0.5Hz 1Hz 2Hz 3Hz 4Hz

Amplitude
Amplitude
Differentiated signal =
signal*frequency
Integrated signal =
signal/frequency
Frequency

1 Hz 2 Hz 3 Hz 4 Hz Frequency
Fig 3.19  Effect of the derivation to the signal. Differentiation is a multiplica- 0.5Hz 1Hz 2Hz 3Hz 4Hz
tion by the frequency. If the bandwidth is not limited, the noise of the higher
frequencies will be amplified and the signal to noise ratio will be very low. Fig 3.21  Effect of the integration. Lower frequency parts are amplified and
higher frequency parts are dampened.

400
350 is much easier in the frequency domain, because the convolution
300 integral in time domain corresponds to a multiplication in the
frequency domain. Therefore, in most cases the filters are developed
250
IEMG (mVs)

in the frequency domain. Data are transformed in the frequency

200 domain and then multiplied by the transfer function. See Figures
3.19 and 3.21.
150
This property can also be used to study the system behavior. See
100 equations below.
Time domain:
50
+∞
0 output(t ) = ∫ transfer function(t − τ ) ⋅ input(t )dt
−∞
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Frequency domain:
Time (s)
output(ω ) = transfer function(ω ) ⋅ input(ω )
Fig 3.20  Integrated EMG signal of the long back muscle. In case of filtering, the transfer function and measured data (input)
are given and filtered data (output) are calculated. In case of system
analysis, input and output are available.

Figure 3.21 shows a signal in the frequency domain. The inte- output(ω )
transfer function(ω ) =
grated signal part at 1 Hz is divided by 1; the integrated signal part input(ω )
at 2 Hz is divided by 2; and the integrated signal part at 3 Hz is Peham and Schobesberger (2006) used the transfer function to
divided by 3. This procedure gives us smoothed data and is com- calculate the elasticity of the equine back by relating the EMG of
parable to the moving average using two neighboring data points. the long back muscle as input to the motion of the horses back as
This is similar to a low-pass filter, because low frequency parts output. The transfer function was developed in a differential equa-
(below zero) are amplified and high frequency parts are dampened. tion of second order and compared to the motion equation.
In other words, a moving average reduces the samples by the The convolution is similar in nature to the cross correlation (CC).
number of data points that are included. The convolution calculates the CC of two signals whereas the
reverse function of the second signal is used. The CC can be used
System analysis in time and frequency domain to compare two different signals or find the content of one signal
in the other signal. Strobach et al. (2006) used CC to show the limb
(convolution, special filters) coordination of ataxic, sedated and sound horses (by using the CC
If special characteristics of the data are known, it is possible to to compare the motion of different limbs). Peham et al. (1996)
develop a filter characteristic adapted to the wanted signal. Peham used CC to determine the wanted signal parts of vertical head
et al. (1996) used this to filter the vertical head motion. Peham et al. motion, which were necessary for lameness evaluation.
(2001b) filtered electromyographic signals of the left and right lon- The auto-correlation is the CC of a signal with itself. This can be
gissimus dorsi muscles, middle gluteal muscles and triceps brachii used to find periodic elements in the signal or to determine, e.g. the
muscle of horses walking on a treadmill with a special signal stride frequency of a motion signal. Strobach et al. (2006) used the
adapted filter characteristic. In both cases, the transfer function was auto-correlation to show the differences in the gait pattern in ataxic,
developed from the motion of the limb. sedated and sound horses. If the auto-correlation was high, a stable
In the time domain, the convolution integral of the signal and motion pattern was present. Peham et al. (2001b) used the auto-
the transfer function has to be calculated to get the filter signal. This correlation to calculate the signal to noise ratio.

68
 Signal processing

Are the dependent variables

categorical or continuous?

Categorical Continuous

Are the data independent Two or more groups?

Independent Dependent Two groups More than two groups

Chi-Square test Mc Nemar test Independent Dependent Independent Dependent

Are the dependent variables

normally distributed?

Yes No

Independent Paired ANOVA ANOVA Mann-Whitney Wilcoxon- Kruskal-Wallis- Friedmann-

T-test T-test for repeated U-test Test Test Test
measures

Fig 3.22  Basic guidelines for choosing a statistical test.

Kalman filters are used to estimate the dynamical behavior of Artificial neural network
linear systems. Halvorsen et al. (2008) used Kalman filtering to
exclude skin-movement of the distal part of the limb of horse. Artificial neural network (ANN) is a method of highly sophisti-
cated decision-making. ANN, often just called a ‘neural network’
(NN), is a mathematical model or computational model inspired
Decision-making from biological neural networks. It consists of a network of inter-
Signal processing is necessary to interpret the signal or characteris- connected group of artificial neurons. The artificial neurons are
tics of the signal. To get reliable information from the processed in most cases switches that will be activated if a certain input
signal, statistics are needed to reflect trends, if any. Figure 3.22 threshold is reached. Combination, activation and weight of
shows a simple flow chart regarding selection of an appropriate neurons are adapted to the problem. So an ANN is an adaptive
statistical test. A more detailed discussion of statistical methods is system that changes its structure based on external or internal
given by Hopkins (2007). information that flows through the network during the learning
phase.
An overview of ANN in motion analysis is given by Schöllhorn
Fuzzy logic (2004).
The classical (binary) logic distinguishes between two values ‘black’ In more practical terms neural networks are non-linear, non­
and ‘white’. This binary logic implies the law of the excluded parametric statistical data modeling tools. They can be used to
middle, which states that for all propositions, either black or white model a complex relationship between inputs and outputs or
(not black) must be true, there being no middle true proposition to identify patterns in data (Schöllhorn, 2004; Jang et al., 1997).
(gray) between them. Savelberg et al. (1997a) used ANN as a tool for complex calcula-
It was Jan Łukasiewicz who first proposed a systematic alternative tion to determine ground reaction forces in horses with hoof wall
to the bi-valued logic of Aristotle and described the 3-valued logic deformation. Further use of ANN was seen for computing tendon
(black-gray-white), with the third value being possible (Jang et al., forces from EMG data in cats (Savelberg et al., 1997b) and for cal-
1997). culation of fore and back force from pressure data (Savelberg et al.,
Fuzzy sets have been introduced by Lotfi A. Zadeh (1965) as an 1999).
extension of the classical notion of set. In classical binary set theory, Use of ANN as a decisive statistical tool was shown by Schobes-
two states (black and white) are possible. Other states are not ele- berger and Peham (2002). They used a multilayer feed forward ANN
ments of the set. By contrast, fuzzy set theory permits the gradual to quantify lameness in horses. The classification via ANN was
assessment of the membership of elements in a set (e.g. 50% of correct in 78.6% of cases. They concluded that after proper training,
black and 50% of white gives us gray). Let us say, ‘fuzzy logic is ANNs were potentially capable of making a non-human diagnosis
more adapted to the (bio-) logic of the real world, because in the of equine lameness.
real world (life) there are always more than two colors or a fixed Schöllhorn et al. (2006) showed that a Kohonen map (a special
number of solutions’. type of ANN; it compares the input data with standard data by
Fuzzy clustering is used in clinical motion analysis to distinguish vector products) combined with a cluster analysis is sensitive
between sound and lame gait patterns. Keegan et al. (2004) used enough to identify a specific motion pattern in the horse-rider
fuzzy clustering to detect spinal ataxia in horses. interaction.

69
 3 Signals from materials

References
Audigie, F., Pourcelot, P., Degueurce, C., series. Prentice Hall, Upper Saddle River, Peham, C., Scheidl, M., Licka, T., 1996. A
Geiger, D., Denoix, J.M., 2002. Fourier NJ, o7458. method of signal processing in motion
analysis of trunk displacements: a method Keegan, K.G., Arafat, S., Skubic, M., Wilson, analysis of the trotting horse. J. Biomech.
to identify the lame limb in trotting D.A., Kramer, J., 2003. Detection of 29 (8), 1111–1114.
horses. J. Biomech. 35 (9), 1173–1182. lameness and determination of the affected Peham, C., Schobesberger, H., 2006. A novel
Back, W., Van Schie, M.H.M., Pol, J.N., 2006. forelimb in horses by use of continuous method to estimate the stiffness of the
Synthetic shoes attenuate hoof impact in wavelet transformation and neural network equine back. J. Biomech. 39 (15),
the trotting warmblood horse. Equine classification of kinematic data. Am. J. Vet. 2845–2849.
Comp. Exerc. Physiol. 3 (3), 143–151. Res. 64 (11), 1376–1381. Polikar, R., 1995. The engineer’s ultimate
Barrey, E., Galloux, P., 1997. Analysis of the Keegan, K.G., Arafat, S., Skubic, M., Wilson, guide to wavelet analysis. The Wavelet
equine jumping technique by D.A., Kramer, J., Messer, N.M., et al., 2004. Tutorial.
accelerometry. Equine Vet. J. Suppl. (23), Detection of spinal ataxia in horses using Potvin, J.R., Brown, S.H.M., 2004. Less is
45–49. fuzzy clustering of body position more: high pass filtering, to remove up to
Bisseling, R.W., Hof, A.L., 2006. Handling of uncertainty. Equine Vet. J. 36 (8), 99% of the surface EMG signal power,
impact forces in inverse dynamics. J. 712–717. improves EMG-based biceps brachii muscle
Biomech. 39, 2438–2444. Meffert, B., Hochmuth, O., 2004. Werkzeuge force estimates. J. Electromyo. Kinesio.
Burn, J.F., Wilson, A., Nason, G.P., 1997. der signalverarbeitung. Pearson Education 14, 389–399.
Impact during equine locomotion: Deutschland GmbH. Savelberg, H.H., de Lange, A.L., 1999.
techniques for measurement and analysis. Nambu, Y., Morimoto, M., Lemieux, J., 2000. Assessment of the horizontal, fore-aft
Equine Vet. J. Suppl. (23), 9–12. Who is fourier? a mathematical adventure. component of the ground reaction force
Cappozzo, A., Croce, U.D., Leardini, A., Transnational College of LEX, USA. from insole pressure patterns by using
Chiari, L., 2005. Human movement Niederl, M., Anen, C., Simhofer, H., Peham, artificial neural networks. Clin. Biomech.
analysis using stereophotogrammetry – part C., 2006. Kinematic analysis of the chewing 14 (8), 585–592.
1: theoretical background. Gait Posture 21, cycle of horses before and after dental Savelberg, H.H., Herzog, W., 1997b. Prediction
186–196. treatment. J. Biomech. 39 (Suppl. 1), 567. of dynamic tendon forces from
Clayton, H.M., 1996. Instrumentation and Oosterlinck, M., Pille, F., Back, W., Dewulf, J., electromyographic signals: an artificial
techniques in locomotion and lameness. Gasthuys, F., 2009. Use of a stand-alone neural network approach. J. Neurosci.
Vet. Clin. North Am. Equine Pract. 12 (2), pressure plate for the objective evaluation Methods. 78 (1–2), 65–74.
337–350. of forelimb symmetry in sound ponies at Savelberg, H.H., Van Loon, T., Schamhardt,
Digby, C.J., Lake, M.J., Lees, A., 2005. walk and trot. Vet. J. Epub ahead of print. H.C., 1997a. Ground reaction forces in
High-speed non-invasive measurement of Oppenheim, A.V., Willsky, A.S., 1996. Signals horses, assessed from hoof wall
tibial rotation during the impact phase of and systems. In: Hamid, S., Hamid Nawab, deformation using artificial neural
running. Ergonomics 48 (11–14), S. (Eds.), Signal processing series, second networks. Equine Vet. J. Suppl. (23),
1623–1637. ed. Prentice Hall Inc., Englewood Cliffs, NJ, 6–8.
Durkin, J.L., Callaghan, J.P., 2005. Effects of USA. Schobesberger, H., Peham, C., 2002.
minimum sampling rate and signal Peham, C., Frey, A., Licka, T., Scheidl, M., Computerized detection of supporting
reconstruction on surface 2001a. Evaluation of the EMG activity of forelimb lameness in the horse using an
electromyographic signals. J. Electromyo. the long back muscle during induced back artificial neural network. Vet J. 163 (1),
Kinesio. (15), 474–481. movements at stance. Equine Vet. J. Suppl. 77–84.
Galloux, P., Richard, N., Dronka, T., Leard, A., (33), 165–168. Schöllhorn, W.I., 2004. Applications of
Perrot, A., Jouffroy, J.L., et al., 1994. Peham, C., Licka, T., Girtler, D., Scheidl, M., artificial neural nets in clinical
Analysis of equine gait using three- 1999. Supporting forelimb lameness: biomechanics. Clin. Biomech. 19 (9),
dimensional accelerometers fixed on the clinical judgement vs. computerised 876–898.
saddle. Equine Vet. J. Suppl. (17), 44–47. symmetry measurement. Equine Vet. J. 31 Schöllhorn, W.I., Peham, C., Licka, T., Scheidl,
Giakas, G., Baltzopoulos, V., 1997. Optimal (5), 417–421. M., 2006. A pattern recognition approach
digital filtering requires a different cut-off Peham, C., Licka, T., Girtler, D., Scheidl, M., for the quantification of horse and rider
frequency strategy for the determination of 2001d. Hind limb lameness: clinical interactions. Equine Vet. J. Suppl. (36),
the higher derivatives. J. Biomech. 30 (8), judgement versus computerised symmetry 400–405.
351–355. measurement. Vet Rec. 148 (24), 750–752. Semmlow, J.L., 2004. Biosignal and
Halvorsen, K., Johnston, C., Back, W., Peham, C., Licka, T., Kapaun, M., Scheidl, M., biomedical image processing MATLAB-
Stokes, V., Lanshammar, H., 2008. 2001c. A new method to quantify harmony based applications. signal processing and
Tracking the motion of hidden of the horse-rider system in dressage. Sports communication series. Marcel Dekker Inc.,
segments using kinematic constraints Engineering 4, 95–101. New York Basel.
and kalman filtering. J. Biomech. Eng. Peham, C., Licka, T., Mayr, A., Scheidl, M., Sha, D.H., Mullineaux, D.R., Clayton, H.M.,
130 (1), 011012. Girtler, D., 1998. Speed dependency of 2004. Three-dimensional analysis of
Hopkins, W.G., 2007. A new view of statistics. motion pattern consistency. J. Biomech. 31 patterns of skin displacement over the
<http://sportsci.org/resource/stats/index. (9), 769–772. equine radius. Equine Vet. J. 36 (8),
html>. Peham, C., Licka, T., Scheidl, M., 2001b. 665–670.
Ives, J.C., Wigglesworth, J.K., 2003. Sampling Evaluation of a signal-adapted filter for Shiavi, R., 1999. Introduction to applied
rate effects on surface EMG timing and processing of periodic electromyography statistical signal analysis. Academic Press,
amplitude measures. Clin. Biomech. 18, signals in horses walking on a treadmill. San Diego San Francisco New York Boston
543–552. Am. J. Vet. Res. 62 (11), 1687–1689. London Sydney Tokyo.
Jang, J.S.R., Sun, C.T., Mizutani, E., 1997. Peham, C., Licka, T., Schobesberger, H., Strampp, W., Vorozhtsov, E.V., 2004.
Neuro-fuzzy and soft computing – a Meschan, E., 2004. Influence of the rider Mathematische methoden der
computational approach to learning and on the variability of the equine gait. Hum. signalverarbeitung. Oldenburg Verlag,
machine intelligence. matlab curriculum Mov. Sci. 23 (5), 663–671. München Wien.

70
 References

Strobach, A., Kotschwar, A., Mayhew, I.G., trotting on a treadmill. Am. J. Vet. Res. 65 signals to monitor the effect of
Peham, C., Licka, T., 2006. Gait pattern of (2), 155–158. a botox treatment in six healthy
the ataxic horse compared to sedated and Vint, P.F., Hinrichs, R.N., 1996. Endpoint error ponies and two horses with stringhalt:
nonsedated horses. Equine Vet. J. Suppl. in smoothing and differentiating raw preliminary study. Equine Vet. J. 41 (3),
(36), 423–426. – kinematic data: an evaluation of four 313–318.
Uhlir, C., Licka, T., Kübber, P., Peham, C., popular methods. J. Biomech. 29 (12), Winter, D.A., 1990. Processing of raw
Scheidl, M., Girtler, D., 1997. 1637–1642. kinematic data. In: Biomechanics and
Compensatory movements of horses with a Werner, M., 2006. Digitale signalverarbeitung motor control of human movement,
stance phase lameness. Equine Vet. J. mit MATLAB. Friedrich Vieweg & Sohn second ed, John Wiley & Sons, Inc.
Suppl. (23), 102–105. Verlag. Zadeh, L.A., 1965. Fuzzy sets. Information and
Uhlir, C., Licka, T.F., Peham, C., Frey, A., 2004. Wijnberg, I.D., Schrama, S.E., Elgersma, A.E., Control 8, 338–353.
Electromyographic activity of the Maree, J.T., de Cocq, P., Back, W., et al.,
longissimus dorsi muscles in horses during 2009. Quantification of surface EMG

71
This page intentionally left blank
 C H A PTER 4 

Locomotor neurobiology
and development
Albert Gramsbergen

such as teleosts, reptiles, amphibians and birds. Larger mammals

Introduction are rarely included in these comparisons (e.g. Ariens Kappers et al.,
1967). Also, data on neural connections in the horse’s brain are
The last decennia have seen an explosion of knowledge on the scarce. Verhaart and Sopers-Jurgens (1957) described Bagley’s
neurobiology of locomotion. Much research has been devoted to bundle by dissection of the horse’s brain. This projection runs from
elucidating the basis of normal and disturbed locomotion in man. the cerebral cortex to the mesencephalic and pontine tegmentum
Investigations mainly in cats and rats and also in man have focused and is considered to be the phylogenetic predecessor of the corti-
upon the steering of locomotion by central brain areas, upon central cospinal tract (CST) which is present in other mammals.
pattern generators (CPG), upon the recovery after brain damage or In the last four or three decades of the 20th century a variety of
peripheral nerve lesions and a variety of other topics. It has been neuroanatomical tracing techniques have been developed and these
known for almost a century, that the neuronal assemblies called opened a multitude of new insights into brain circuitry. By the
CPGs play an important role in rhythmic leg movements (and also injection of retrogradely or anterogradely transported vital stains,
in breathing and chewing movements). The research in lampreys, terminals of neural projections or their site of origin can be studied.
cats and developing rats into the neural organization of these CPGs, Such investigations, however, require surgical interference as well
their wiring patterns and the transmitters involved has given an as, after a short survival period, post-mortem processing of the brain
impressive momentum to studies into the neurobiology of locomo- tissue. These studies have therefore not been performed in horses.
tion in bipeds and in quadrupeds. Similarly, the cerebellum and For similar reasons, studies involving the implantation of depth
motor learning, the basal ganglia and the planning of motor acts electrodes for recording or stimulating central nervous structures
as well as the sensory-motor cortex and skilled movements in rela- have not been carried out in horses. Modern imaging techniques of
tion to posture and locomotion have seen a vivid research the brain (as fMRI) have been widely applied in human research
interest. and results may indicate the brain areas involved in specific tasks.
However, detailed knowledge is still lacking on the neuroana- These studies require immobility of the skull in the scanner and
tomical and neurophysiological basis of locomotion in horses. that is one of the reasons why these techniques do not lend them-
Investigations into such neurobiological aspects generally require selves to investigations in horses. Possibly, the newest advance in
invasive procedures, and ethical considerations, the economical brain-imaging, MRI-diffusion tractography, which allows non-
value and the size of this animal are the main reasons for a invasive identification of tracts in the CNS, might ultimately be
paucity of data. Non-invasive techniques such as transcranial mag- applied to study neural projections in animals and possibly in larger
netic stimulation with the recording of responses in the EMG or animals (Behrens et al., 2007).
fMRI-techniques have not yet been adapted to equine research for In horses, the recording of simultaneous EMG signals in collec-
a variety of reasons. On the other hand, research into the biome- tions of muscles have been performed with kinematic recordings
chanical aspects of walking, galloping and trotting in horses, and to analyze coordination patterns during a variety of locomotion
also into gait analysis with high-speed filming, have a long tradi- patterns. These studies are also important for diagnosing abnor-
tion. Results from this research have set standards which, in turn, malities in movement patterns and diseases of muscles and
inspired investigations in man and smaller laboratory animals. motoneurons.
The ‘golden age’ of research into equine locomotion in the 1960s, This chapter bases itself upon research in laboratory animals and
(Van Weeren, 2001) developed a promising sequel by research in notions on neurobiological aspects of trotting and galloping in
several laboratories in the USA, in Canada as well as in several horses should be derived by the extrapolation of this knowledge.
European countries, most notably in The Netherlands and in When extrapolating, it should be kept in mind that not only the
France. The first edition of Equine Locomotion marked this interest size but also the behavioral competences of horses on the one hand,
and the results obtained so far were comprehensively summarized and laboratory animals such as rats, cats or monkeys on the other,
there. differ widely. These differences are related to the ecological biotopes
Concepts on the neurobiological basis of locomotor movements from which these animals stem. Behavioral competence is mutually
in horses, therefore, necessarily mainly rely upon knowledge interconnected with the neural organization of the nervous system
obtained in laboratory animals. Older textbooks on comparative and interspecies extrapolations therefore have to be interpreted with
neuroanatomy include only a few data on horses. These textbooks the greatest caution.
are mainly limited to comparing the gross anatomy of the brains The relationships between the terminal projection patterns of the
in smaller mammals and man, with that of primitive vertebrates CST in a variety of animals in relation to their manipulative skills,

73
 4 Locomotor neurobiology and development

is a case in point (Armand, 1982). In higher primates, the pyramidal Summary

neurons in the primary motor cortex are monosynaptically con-
nected to the motoneurons of the hand and wrist muscles, located Actin and myosin molecules shorten in the sarcomere by connect-
in the cervical spinal cord. By virtue of these direct connections, ing troponin. This reaction is initiated by an action potential arriv-
apes and man are able to independently move their fingers and toes, ing at the muscle cell membrane which, in turn, releases the
to manipulate objects and to use tools, while standing upright intermediary Ca++ ion.
(Lawrence & Hopkins, 1976). In cats, however, fibers of the CST
mainly end in the dorsal and medial laminae of the spinal cord and Muscle fiber-types
they connect to the so-called premotor neurons at cervical and
lumbar levels. Indeed, cats cannot move their toes independently. Some muscles may appear dark-red on macroscopical inspection
Rats take an intermediate position between primates and cats as while other muscles are much paler. Examples of such dark-red
these animals do have a limited number of direct connections muscles are the long back muscles and also the soleus muscles in
between cortical and spinal motoneurons and these do allow them the hind-legs. A large majority of the fibers in these muscles are of
to manipulate their food by independent toe movements. In rabbits the so-called type I, and these fibers derive their energy (to regener-
and goats the CST only descends as far as the upper cervical seg- ate ATP from ADP) from the aerobic oxidation of glucose (which
ments and its fibers do not even reach the premotor neurones of takes place in mitochondria). Glucose enters the myocytes from
the cervical and lumbar spinal cord. Horses most probably are capillaries and the extracellular space. These muscles appear dark-
similar to these animal species with respect to the descent of the red because the type I myocytes contain high levels of the deep red
CST or its analogon (Armand, 1982; Verhaart & Sopers-Jurgens, myoglobin molecules, which bind oxygen, and also because of the
1957). many capillaries around the muscle fibers. Glucose is readily replen-
In this chapter, current knowledge on locomotion and its neuro- ished from the blood vessels during muscle contractions and there-
biological substrate will be briefly reviewed. The order which will fore these muscles are able to produce contractions for long periods
be followed is from the peripheral elements to central structures, of time.
or, from the muscles via the spinal cord towards central motor Other muscles are much paler. These muscles mainly contain type
systems. As developmental research into locomotor functions and II muscle fibers (and a minority of type I fibers). In type II muscle
structural changes in motor systems has made important contribu- fibers, the ATP is regenerated by the anaerobic catabolism of glyco-
tions to our understanding of the neurobiology of walking and gen which is stored in the muscle. This reaction is much faster but
postural maintenance at adult age, these aspects will be discussed energetically less efficient than the aerobic metabolism of glucose.
where appropriate. Moreover, the supplies of glycogen are limited. As the recovery of
these supplies takes a considerable period of time, the contractions
of the fibers in these myocytes can only be brief. Other fibers within
the same muscle will take over but, obviously, this has its limits.
Muscles, motors of movements Two subtypes of type II muscle fibers can be distinguished. Type IIb
fibers derive their energy exclusively from the anaerobic metabolism
Leg and trunk movements are produced by skeletal muscles. Histo- of glycogen and type IIa fibers have a mixed metabolism; these
logical and neurophysiological aspects of muscles at adult age have make use both of the anaerobic as well as the aerobic route to gener-
been reviewed several times (Burke, 1981; Partridge & Benton, ate energy.
1981; Kernell, 1998) and a few of these will be repeated only briefly The mechanical force of single muscle fibers has been recorded
here. after briefly stimulating the motor nerve. Such stimulation induces
Muscles consist of hundreds or thousands of multi-nuclear a so-called muscle-twitch which is a single contraction of the fiber.
muscle cells or muscle fibers and each of these contain many myo- It has been demonstrated that the force produced by a twitch of
fibrils. In addition, satellite cells occur which at later stages may type I muscle fibers increases relatively slowly, has a low peak value
transform into functional muscle cells. The myofibrils in the muscle and also decreases slowly. When stimulating for sustained periods
cell consist of a chain of thousands of ‘sarcomeres’, the basic and with such frequencies that the separate twitches fuse into a
element of a striated muscle fiber. A sarcomere consists of longitu- tetanic contraction, it appears that the tension in type I muscle
dinal actin molecules with short troponin molecules attached and fibers can be maintained for longer periods. In other words, these
myosin molecules. The contraction of a myofibril is produced by ‘slow-twitch’ muscle fibers are ‘resistant to fatigue’ and this is related
the contraction of the actin and myosin-complexes in each of the to their oxidative metabolism (see above). On the other hand, the
sarcomeres. This contraction is produced by the formation of cross type II muscle fibers, after brief electrical stimuli, show a much
bridges between the actin and myosin molecules. This active process faster increase in their tension and also the peak value of tension is
requires energy which is derived from ATP (which then degrades much higher than that in the type I fibers. This increased tension
into ADP). The relaxation of the muscle by lengthening of the sar- is explained by the formation of alternative cross bridges between
comeres basically is a passive process. the actin and mysosin molecules. The tension after a twitch declines
The contraction of a muscle fiber is induced by an action poten- rapidly in these fibers. When stimulated with high, constant
tial in the motor nerve which arrives at the motor endplate. The frequency leading to a tetanic contraction, the tension in these
release of acetyl-choline (ACh) in the motor endplate and the ‘fast-twitch, fast-fatigable’ muscle fibers diminishes after a relatively
binding of ACh to nicotinic ACh-receptors, leads to the opening of short period of time. After such contractions, these fibers need a
ion channels in the muscle cell membrane. The net influx of Na+ considerable time to replenish their glycogen stores and to wash
ions leads to a local depolarization of the cell membrane which out the lactic acid, which is produced by the metabolism of glyco-
spreads along the membrane. This depolarization, in turn, leads to gen. Type IIb fibers produce the highest tension, and these fibers
the release of Ca++ via the sarcoplasmic reticulum. Ca++ binds to show signs of fatigue most rapidly, while type IIa muscle fibers take
troponin which then induces the formation of the cross bridges an intermediate position between the type I and type IIb fibers. After
between actin and mysin, producing the contraction of the muscle stimulation their force rises and falls rapidly (though, less than that
fibers. Both the spread of the depolarization along the membrane of the type IIb fibers) and because of a mixed aerobic and anaerobic
and the transport of Ca++ through the muscle cell is a relatively slow metabolism, they are more resistant to fatigue than the type IIb
process and the delay between the arrival of the action potential fibers.
and the onset of the contraction, the so-called ‘action–contraction The different types of muscle fibers can be visualized histo­
coupling’, may last 200 ms. logically. Post-mortem studies allow the investigation of the

74
 Muscles, motors of movements

distribution of fiber types in sections taken perpendicular to the storage of large quantities of glycogen for rapid use in type II
longitudinal axis of the muscle at several levels. Multiple biopsies muscle fibers.
only allow an approximation of such data. In a classical method
of processing (Dubowitz & Brooke, 1973), the muscle material is Summary
preincubated at different pHs and the ATPase staining technique
then enables differentiation of the muscle fiber types. More recent Muscles are composed of type I muscle fibers (slow-twitch, fatigue-
methods apply myosin-related antibodies to identify several sub- resistant) type IIA fibers (fast-twitch, moderately fatiguable) and
types of muscle fibers (Schiaffino et al., 1986). Investigations with type IIB (fast-twitch, fatiguable); intermediate types also exist.
these methods have demonstrated that all muscles consist of a
mixture of type I, type IIa and type IIb muscle fibers, but the rela-
tive proportions of these fibers differ greatly. Extremity muscles
Muscular development
and particularly the flexor muscles may contain higher propor- Muscle cells or myocytes develop by the fusion of a longitudinal
tions of type IIa and IIb fibers and fewer type I muscle fibers. The array of mesenchymal cells. The first generation of these multinu-
antigravity muscles in the extremities, in contrast, contain clear myoblasts, extending from tendon to tendon, are the primary
increased proportions of types I and IIa muscle fibers. This is myotubes. Each of these is soon innervated by a multitude of moto-
understandable when realizing that these muscles have to carry neuronal axons. At a later stage, a second generation of secondary
the weight of the animal for long periods of time. The propor- myotubes develops in the vicinity of the middle region of the pri-
tions of type I muscle fibers are particularly pronounced in the maries. The secondary myotubes also are initially innervated by
soleus muscle in the hind legs, and portions of the flexor carpi multiple motoneuronal axons (Ontell & Dunn, 1978; Goldspink &
ulnaris muscle in the forelegs. In these muscles, the type I Ward, 1979; Bennett, 1983). During further development, the con-
muscles fibers may be as high as 100% (e.g. in rats). It is note- tractile proteins actin and myosin increase enormously in the myo-
worthy that the type I fibers both in the extensors and the tubes of both types. The motor endplate develops its folded
flexor-muscles are mostly concentrated in the deeper portions appearance and from a certain stage of development these cells have
of the muscles, close to the skeletal elements (Wang & Kernell, acquired the characteristics of the mature myocyte. It is generally
2001). Similar data have been obtained in horses (Grotmol accepted that the primary myotubes are the precursors of type I
et al., 2002). muscle fibers and the secondary myotubes of the type II muscle
The trunk muscles also contain elevated percentages of type I fibers (see Wilson et al., 1988).
and IIa muscle fibers. This holds particularly for the multifidus and Wilson and coworkers studied in rats the effects of severe under-
the longissimus muscles in the back (Gramsbergen et al., 1996). nutrition from early gestation and through the lactation period on
Similar results have been obtained in horses (Gellman et al., muscular development. The number of primary myotubes remained
2002). The high percentage of type I and IIa muscle fibers make unaffected but they detected a dramatic decrease in the number of
these particular muscles fatigue resistant and therefore these secondary myotubes. These and other results has led them to
muscles are ideally suited for postural functions, such as keeping hypothesize that the development of primary myotubes is geneti-
the trunk straight and carrying the body weight (for a review on cally determined while that of the secondary myotubes is suscep-
regional differences in the distribution of type I and type II muscle tible to environmental factors, such as under-nutrition, hormonal
fibers, see Kernell (1998)). factors and also training (Wilson et al., 1988). In addition, non-
Horse breeds differ in the size and composition of their muscles. differentiated satellite cells develop and these cells may differentiate
Horses of some breeds are heavily built with bulky muscles that at a later stage into myocytes after damage to the muscle or after
have a large cross-sectional area and these horses are kept in order training.
to sustain enduring agricultural or transportation tasks. Their The muscle-fiber type distribution within muscles is, to some
muscles contain large proportions of type I and type IIa muscle extent, related to their topological origin. The muscles in the trunk
fibers which are able to produce forces for long-lasting periods (see arise from the segmentally arranged myotomes. The intercostal
Miyata et al., 1999). Horses of other breeds are characterized by a muscles (together with the rib and the spinal cord segment inner-
slender and athletic build that is appropriate for speed racing. Their vating these muscles) have clearly retained this segmental origin in
muscles have increased proportions of type II muscle fibers that all mammals. But, the long muscles in the neck and in the back,
allow them to produce immense forces but only for relatively short the abdominal muscles and the diaphragm are also of axial origin.
periods of time. These muscle are particularly important for the stabilization of the
Investigations in man with EMG recordings and, in addition, trunk and the head and, therefore, for postural control. At adult age,
power spectral analysis of the EMG and the calculation of EMG- the trunk muscles contain high proportions of slow-twitch, fatigue-
fatigue indices, indicated that loading for long periods induces the resistant muscle fibers.
recruitment of satellite cells in the muscle (see below), their devel- The muscles in the four legs arise from limb buds. The limb buds
opment into muscle cells and a relative increase in type II muscle develop at the ventrolateral aspect of the body axis and these origi-
fibers (Mileva et al., 2009). Investigations, e.g. in rats, have shown nate from groups of mesenchymal cells which migrated from the
that per individual muscle the number of type I muscle fibers trunk. The extensor muscles in the limbs (e.g. the gastrocnemius
remains unchanged (or, is genetically determined; see Wilson et al., and the quadriceps muscles in the hind leg) develop from the
1988; see also, Mileva et al., 2009) and Mileva’s results, therefore, ventral plates in the buds and they are involved in ‘anti-gravity’ tasks
point towards an absolute increase in type II fibers. Other investiga- such as standing, and also in the heel strike phase of walking and
tors suggest, in contrast, that training may influence the fiber type running. These muscles later contain considerable percentages of
distribution of muscles by shifts in the numbers of both type I and type I muscle fibers. The flexor muscles develop from the dorsal
type II fibers (e.g. Duclay et al., 2009). plate in the limb bud, they are particularly active during the swing
Other training effects are hypertrophy of muscle fibers as indi- phase of the leg and these muscles contain increased proportions
cated by an increase in their cross sectional area, increases of type II muscle fibers.
in mitochondrial content (important for aerobic glycolysis)
(Grünheid et al., 2009), changes in muscular architecture (which
might be secondary to the muscle fiber hypertrophy) (Duclay
Summary
et al., 2009), as well as vascular changes by angiogenesis and an Muscles develop, cytologically, from primary and secondary myo-
increase in arteriolar density (for review, Laughlin and Roseguini tubes. Trunk and extremity muscles originate from myotomes and
(2008)). Short-term effects of training aim at increasing the limb buds, respectively.

75
 4 Locomotor neurobiology and development

The motor unit and motoneurones (Wang et al., 2002). Therefore, the polyneural innervation at first
and its regression thereafter should rather be regarded as a fine-
tuning in the matching-process.
Groups of muscle fibers in a muscle are innervated by one moto-
neuron in the spinal cord (or in the brain stem) and the complex
of an α-motoneuron, together with the muscle fibers, is called a Summary
‘motor unit’. All muscle fibers of a motor unit tend to have identical Muscle fibers initially are polyneurally innervated. Mononeural
physiological properties (such as twitch speed and relaxation time) innervation develops by a regressive process before the muscle is
and these properties match the physiological and morphological involved in its adult functioning.
properties of the motoneurons.
Research mainly in cats, has indicated that the small-sized moto-
neurons innervate the type I muscles fibers and in neurophysio­ Size principle
logical research, these units have been coined ‘S-type units’ The motoneurons within a motoneuronal pool differ in their sizes
(slow-twitch, fatigue-resistant). Motoneurons of intermediate sizes (see above). Henneman & Mendell (1981) have demonstrated that
innervate type IIa muscle fibers and physiologists have distinguished within a motoneuronal pool, the motoneurons are recruited in an
‘FR-type units’, fast-twitch, fatigue-resistant units and ‘Fint-type orderly fashion from small to large, with increasing intensities of
units’, with fast-twitch and intermediate fatigue-resistant units. The neuronal drive (e.g. from higher brain centers). During motor tasks
largest motoneurons are part of the ‘FF-type units’, innervating type which require only low muscle forces only the smallest motoneu-
IIb muscle fibers; these units are fast-twitch and fatigable (Burke, rons are activated (these motorneurons innervate groups of type
1981). I-muscle fibers). With increasing forces, additional motoneurones
The motorunits systematically differ in the numbers of muscle with larger sizes are recruited, first those with FR properties, later
fibers which are innervated by the motoneuron. The smaller moto- those with Fint properties and finally the FF units are activated with
neurons innervate relatively small numbers of type I muscle fibers. their large numbers of type IIb muscle fibers. This orderly recruit-
Both the low tension produced by each of these fibers and the rela- ment of motoneurons is related to differences in the membrane
tively small number of muscle fibers in each of these units, explain properties of the smaller and larger motoneurons. Interestingly, the
why these units produce relatively low forces. The larger motoneu- relative force-increase by recruiting additional motor units at adult
rons innervate larger numbers of (type II) muscle fibers. Each of age is more or less the same over a wide range of forces (Milner-
these fibers produce higher tensions, and consequently these units Brown et al., 1973) and this leads to a smooth increase of force
produce relatively large forces (for a review see Kernell, 2006). with increasing intensities of neuronal drive.

Summary Summary
Small motoneurons innervate a group of slow-twitch, fatigue- Motoneurons are recruited in an orderly fashion from small to
resistant muscle fibers, producing low forces, motoneurons of inter- large, with increasing intensities of neuronal drive.
mediate sizes innervate groups of fast-twitch, intermediate-fatigable,
and the largest neurons produce strong forces by their fast-twitch,
fatigable muscle fibers. Dendrite bundles
Apart from their size, motoneurons also differ in the morphology
of their dendritic tree. The motoneurons in particular pools have
Development of muscular innervation their dendrites running in bundles both in a longitudinal direction
At early stages of development, muscle cells are innervated by more as well as in transverse directions, while in other motoneuronal
than one axon (see above) and this is known as polyneural innerva- pools such organization is absent. Motoneurons with dendrite
tion. Later, and already in early development, this multiple inner­ bundles have been described in cats (Scheibel & Scheibel, 1970), in
vation of muscle cells is replaced by mononeural innervation (for rats (Gramsbergen et al., 1996) and in the human spinal cord
reviews, Bennett, 1983; Jansen & Fladby, 1990). Research from (Schoenen, 1982) and because of their wide occurrence it seems
many groups has indicated that this regressive process is near com- highly likely that dendrite bundles also occur in the horse’s spinal
pletion when the behavioral function in which the muscle is cord. Dendrite bundles consist of up to 15 dendrites in close vicin-
involved has developed (Gramsbergen et al., 1997). Rats are a pre- ity for several hundreds of millimeters (Fig. 4.1), and connected via
cocious species and they are born at an early stage of brain develop- long-stretched gap-junctions (Matthews et al., 1971; van der Want
ment. Still, mononeural innervation of respiratory muscles in this et al., 1998). Dendrite bundles specifically occur in pools of axial
species has already been reached at birth when the animals start muscles in the trunk and neck as well as in pools of anti-
breathing continuously. On the other hand, in the psoas muscle in gravity muscles in the extremities (notably, the soleus and the
the hind leg, this regression of polyneural innervation is only com- gastrocnemius muscles in the hind limb and the flexor carpi ulnaris
pleted around the 16th postnatal day (IJkema-Paasen & Gramsber- muscle in the forelimb). Interestingly, these are the very muscles
gen, 1998). This is precisely the point at which they start the that contain high percentages of type I muscle fibers.
adult-like pattern of walking. Horses are a precocious species, born The physiological significance of dendrite bundles is not known
with a nervous system at an advanced stage of development. They but one possibility is that they serve to electrotonically couple
are able to stand and walk shortly after their birth and, therefore, it motoneurons in pools stretching over several spinal cord segments
is safe to postulate that most of their muscles are mononeurally which innervate long muscles, e.g. in the back or abdomen. Our
innervated from birth. study into the development of the dendritic tree of motoneurons
It has been suggested that the great excess of axons innervating innervating the soleus muscle (mainly consisting of type I muscle
the primitive muscle fibers at early stages and the selective regres- fibers) revealed that, until the 14th postnatal day in rats, the den-
sion of supernumerary nerve fiber-endings thereafter plays a role in drites of these neurons run in a seemingly disorganized fashion.
matching the properties of muscle fibers to those of the motoneu- From that age, however, the dendrites reorganize into prominent
rons (O’Brien et al., 1978; Greensmith & Vrbova, 1991). More bundles. This fast development coincides with the occurrence of the
recent research has indicated, however, that even at the earliest adult-like walking pattern in rats. On the other hand, the dendrites
stages ingrowing axons follow a trajectory towards a specific mus- of the motoneurons innervating the tibialis anterior muscle (a
cular region with a preponderance of particular muscle fiber types flexor muscle, and the antagonist to the soleus muscle) do not

76
 The motor unit and motoneurones

Fig 4.1  Motorneuronal pools of the longus capitis muscle: (A) microphotograph and (B) reconstruction of another pool. Bars: 100 µm.
Reprinted from Gramsbergen, A., Ijkema-Paassen, J., Westerga, J., Geisler, H.C., 1996. Dendrite bundles in motoneural pools of trunk and extremity muscles in the rat. Experi. Neuro. 137 (1), 34–42,
with permission from Elsevier.

develop dendrite bundles at any age (Westerga & Gramsbergen, on a treadmill. Forssberg et al. (1980) demonstrated after spinal
1992). The organization of dendrites into bundles depends on the cord transection that the coordination patterns in the limbs follow
ingrowth of descending projections from higher levels as (in rats) the increasing speed of the treadmill, provided that the limb affer-
a spinal cord transection at early age prevents these bundles from ents are left intact. At higher speeds they even observed a gallop-like
developing (Gramsbergen et al., 1995). In cats, such interference at coordination. Shik and Orlovsky (1976) studied cats after spinal
adult ages leads to their hypotrophy or even atrophy (Reback et al., cord transection when walking on a treadmill with two belts, driven
1982). at different speeds. When the speed differences became too large,
the frequencies of the CPG stepwise shifted to a 1 : 2 relationship
and on this basis and other evidence it is considered that each limb
Summary has its own CPG (Fig. 4.2).
The neural principles involved in CPGs have been elucidated in
Dendrite bundles are typical for motoneuronal pools of muscles
a series of investigations in the lamprey, a primitive fish which has
with an important postural function.
a less complex spinal cord than mammals and no extremities. Here,
each spinal cord segment along the trunk has a CPG which subse-
quently and with a short phase lag become active. This leads to the
Central pattern generators rhythmic curving of the trunk which produces the swimming move-
Walking, trotting or galloping is produced by rhythmic and alter- ments. Grillner and coworkers demonstrated that each spinal cord
nating leg movements in varying patterns. Graham Brown (1914) segment has two networks on the left and right side, each consisting
was the first to demonstrate that such rhythmic and alternating of excitatory and inhibitory interneurons and with motoneurons as
extremity movements can occur even after several dorsal roots of the output elements (Grillner et al., 1991). The networks on both
the lumbar spinal cord have been severed. This demonstrates that sides are connected via inhibitory connections functioning as
proprioceptive feedback is not essential for these movements to coupled oscillators with an output of alternating activation on the
occur. A few years earlier, Sherrington (1910) had demonstrated right and the left side. Networks in adjacent spinal cord segments
that rhythmic limb movements also remain after severing descend- are connected via propriospinal interneurons.
ing projections in the thoracic spinal cord. These results together Based on neuroanatomical and neurochemical data derived from
indicate that local neuronal networks in the spinal cord are able studies in the lamprey, the network was modeled and computer
to autonomously generate rhythmic movement patterns. Brown simulations indicated that increases in the frequency of the most
has coined these assemblies ‘half-centers’ and later they became rostral segment are spread along the spinal cord and this leads to
know as ‘central pattern generators’ (CPGs). Von Holst in 1935 higher swimming speeds (Grillner et al., 1991). Such increases in
similarly described that rhythmic swimming movements in tele- speed in natural life are induced by supraspinal influences.
osts remain after transection of the spinal cord and Weiss reported The CPGs for extremity movements in tetrapods obviously are
such results in salamanders after partial deafferentation of the more complex and probably consist of a series of coupled oscilla-
limbs (Weiss, 1936). These results provide strong counterevidence tors for the different muscle groups. In experiments performed in
against the concept of rhythmic limb movements being a chain of the isolated spinal cord of newborn rats by Cazalets et al. (1995)
reflexes. fictive locomotion was recorded from electrical activity in the
Most of the studies on CPGs for extremity movements have con- ventral roots (see also Kjaerulff & Kiehn, 1996). Cazalets and
centrated on the hind limb. Grillner and Zangger (1975) demon- co-workers showed that the CPG for hind limb movements in the
strated that adult cats with a low thoracic spinal cord transection rat is located in the first lumbar segment. They also identified the
maintained a delicate pattern of orderly starting and stopping of neural transmitters which are involved such as the excitatory amino
EMG activities in the hind limb flexors and extensors during walking acids L-glutamate and aspartate as well as 5-HT, dopamine and

77
 4 Locomotor neurobiology and development

An intriguing problem with respect to locomotion in horses,

other animals and man is how the fluent postural adjustments in
Descending command
(sets general level of activity)
the trunk muscles are effected during locomotion at various speeds.
It is well known from kinematic studies that the activation and
coordination of trunk muscles differ importantly between walking,
trotting and galloping. Theoretical possibilities are, that the trunk
muscles are rhythmically activated via the CPG for limb move-
LF RF ments, that the rhythmic activity in the trunk muscles is produced
by separate CPGs (perhaps similar to the chain of CPGs in fish) and
that the activity in these CPGs is coupled to the CPG for limb move-
Coordinating neurons ments, and, thirdly, that the trunk muscle activity is adjusted sepa-
rately via supraspinal influences.
Our own experiments involving EMG recordings of limb muscles
and dorsal trunk muscles in adult rats indicated that at low walking
LH RH speeds tonic activation without modulation prevails in the back
A
muscles. At higher speeds, and particularly during accelerations,
stronger and more uniform activation patterns occurs in the long
back muscles and shortly preceding the stance phases of the extrem-
ities at the ipsilateral side (see below; Gramsbergen et al., 1999). As
vestibular deprivation from the 5th postnatal day in rats leads to
long-lasting abnormalities in the coordination between extremity
movements and trunk control (Geisler & Gramsbergen, 1998) we
proposed that supraspinal influences play a decisive role in the
control of the back muscles (Gramsbergen, 1998).

Summary
CPGs in the spinal cord are able to autonomously generate rhyth-
mical locomotor patterns in the trunk muscles in fish and amphib-
B iae and in the extremities in quadrupeds and bipeds. Afferent
feedback and CPGs mutually influence each other. In rats, the CPG
Fig 4.2  Schematic representation of the neural control of locomotion. for hind limb movements is localized in the first lumbar segment.
(A) Dorsal view. Interlimb coordination is achieved by interaction of the central Postural adjustments during walking might be governed largely by
pattern generators (CPGs) of each of the four limbs. The general level of activity supraspinal influences.
in the different generators that are capable of rhythmic activity is set by a
supraspinal descending command, whereas actual coordination is due to inter-
action between the four different generators via coordinating neurons. (B)
Lateral view. The CPG of each limb is influenced by peripheral input from a
Descending projections
variety of somatosensory receptors as well as descending influence from supra-
spinal centers. Through peripheral output pathways, appropriate muscles are In vertebrates, the topography of the descending projections in the
contracted to perform their task and result in intralimb coordination. spinal cord to the motoneurones innervating the trunk and extrem-
(A) Reprinted from Grillner, S., 1975. Locomotion in vertebrates: central mechanisms and reflex ity muscles reflect to some extent their phylogenetical descent. Axial
interaction, 55 (2), 247–304, with permission from The American Physiological Society. muscles in the trunk (the ‘oldest’ muscles from an evolutionary
(B) Reprinted by permission of the publisher from FUNCTIONAL VERTEBRATE MORPHOLOGY, perspective) are innervated by medially descending motor projec-
edited by Milton Hildebrand, Dennis M. Bramble, Karel F. Liem, and David W. Wake, p. 339, tions as the reticulospinal and vestibulospinal tracts. On the other
Cambridge, Mass.: The Belknap Press of Harvard University Press, Copyright © 1985 by the hand, the motoneurones innervating the ‘newer’ extremity muscles
President and Fellows of Harvard College. are located laterally in the ventral horn and these are innervated by
laterally descending tracts.
GABA. However, the neural elements and the circuitry of these CPGs The medially descending projections, such as the vestibulospinal,
for rhythmic alternating limb movements in higher vertebrates are reticulospinal and tectospinal projections, are particularly impor-
still unknown. tant for the innervation of the muscles in the neck and trunk and
Obviously, locomotion in intact animals is adapted to environ- also muscles in the proximal segments of the extremities. Anatom­
mental constraints. Avoiding an object in the trajectory, walking on ical investigations in rats demonstrated that these projections
a slope or anticipating an unevenness in the path lead to adjust- descend early during neuro-ontogeny.
ments of the limb excursions or their rhythmicity. Exteroceptive and In quadrupeds, the rubrospinal tract (RST) is particularly impor-
proprioceptive information influences the extension and flexion tant in the innervation of the motoneurones of the extremity
phases of the step cycle via segmentally arranged circuitry. Stum- muscles. This projection has not been studied yet in horses but it
bling over unexpected objects on the trajectory leads to adjustments seems safe to speculate that in this species this will be the main
of the step cycle (Rossignol et al., 1988) and other experiments have tract involved in adjusting the excitation of motoneurons and in
demonstrated that increasing the load on the extended limb and steering non-rhythmic extremity movements (such as jumping,
input from Golgi-tendon organs via group II afferents also leads to etc.). The RST arises from the red nucleus and descends contra­
adjustments of the step cycle (Duysens & Pearson, 1980). laterally. In rats and cats, this tract arises mainly from the most
The CPG itself, in turn, influences the input from proprioceptive caudally located magnocellular part and to a lesser extent from the
afferents. This probably is affected by activating the inhibitory inter- rostral parvocellular part. It crosses in the ventral tegmental decus-
neurons mediating the reciprocal Ia inhibition (Hultborn, 1972). sation and descends via the lateral funiculus. The axons terminate
Input from the visual and the vestibular system is relayed along upon excitatory and inhibitory interneurons in the spinal cord.
descending projections and cerebellar circuitry probably plays an The red nucleus receives its major input from the deep cerebellar
important role in adjustments (see below). nuclei and possibly also from the motor cortex, both via direct

78
 Central brain mechanisms and locomotion

connections (and in horses possibly via collaterals from Bagley’s the transitions into other coordination patterns probably are medi-
bundle). The cerebellum and the vestibular complex play an impor- ated by indirect activation, via noradrenergic projections arising
tant role in the regulation of locomotor movements and the regula- from the locus coeruleus and adjacent cell groups. These nuclei are
tion of postural maintenance. located close to the MLR and the fiber-projections are part of the
The CST will be less well expressed in horses (Verhaart & Sopers- LMS (see above).
Jurgens, 1957) and horses may, in this respect, resemble the expres-
sion of the CST in rabbits. In rabbits, the unmyelinated fibers of the
CST terminate at the 2nd cervical segment (Hobbelen et al., 1992; The cerebellum
for review see Nudo & Masterton, 1988). Verhaart and Sopers-
Jurgens demonstrated Bagley’s bundle by dissection of the horse’s The cerebellum plays important roles in adjusting motor com-
brain and concluded that this tract only descends as far as upper mands during locomotion. Cerebellar lesions in cats and rats (e.g.
spinal cord levels. In horses, transcranial magnetic stimulation on Brooks, 1975; Gramsbergen, 1982) or cooling of the cerebellum
the skull and the recording of muscle activation in extremity muscles (Udo et al., 1979) lead to atactic gait and irregularities in foot
might help in future research to elucidate the nature of the descend- placing, and, for this reason, it is generally considered that the
ing projections (see Nollet et al., 2003). cerebellum plays a key role in finely adjusting the limb movements
Parallel to the projections of the (three) so-called somatic motor during the step cycle.
systems referred to above (the vestibule- and tectospinal tracts, the The cerebellum receives massive information from spinal
RST and the CST), another system of diffusely projecting monoami- sources both from a direct and indirect nature. The projections
nergic fibers, the ‘fourth motor system’ plays an important role in convey information on the ‘ongoing movements’ (by input from
modulating motor activities. This was termed the limbic motor premotor neurons, interneurons adjacent to the motoneurons) as
system (LMS) by Kuypers (1982) and the emotional motor system well as information on the ‘position of limbs and the tension in
(EMS) by Holstege (1991, 1995). The LMS originates in the muscles’ (by input from muscle spindles and tendon organs). The
medial portions of the hypothalamus and the mesencephalon. information from both these sources reaches the cerebellum along
Its medial components modulate the excitatory state of interneu- the ventral and dorsal spinocerebellar tracts respectively. Indeed,
rons and motoneurons via diffusely projecting noradrenergic and Arshavsky and coworkers (1972, 1986) have recorded rhythmic
serotonergic fibers. An increased excitatory state induces motoneu- activity in the ventral spinocerebellar in phase with rhythmic
rons being more readily activated by a neuronal drive (e.g. from activity in the limbs and this is in agreement with the anatomical
suprasegmental levels). The LMS is the earliest of the descending data. The projections terminate on the parallel fibers in the cere-
projections to develop. At birth in the rat, its fibers have descended bellar cortex (with collaterals, e.g. on neurons in the deep cerebel-
already to lumbar levels where they have established 5-HT contain- lar nuclei). The parallel fibers contact the Purkinje cells. The axons
ing synapses (Rajaofetra et al., 1989); in comparison, the CST in of the Purkinje cells are the sole output of the cerebellar cortex.
the rat has descended to lumbar levels only around the 10th post- Another important input which is highly relevant for steering
natal day. As the terminals are initially widespread, but at later locomotor activity stems from the vestibular nuclei and these
stages restricted to the dorsal and ventral horns, Rajofetra hypoth- give information on the position as well as the movements of
esized that an initial role during early stages of these 5-HT contain- the head.
ing fibers might be to stabilize the innervation patterns of other An important indirect input to the cerebellum is via the inferior
projections to the spinal cord. olivary nucleus. The inferior olivary nucleus receives indirect
(multisynaptic) information from the spinal cord and several
areas in the CNS. Axons from these neurons reach the Purkinje
Summary cells in the cerebellar cortex as climbing fibers, each of which has
thousands of synaptic contacts with one Purkinje cell (e.g.
The medially descending vestibulospinal and reticulospinal tracts, Ruigrok & Cella, 1995). The physiological significance of the
project upon motoneurons of axial muscles and these phylogeneti- input from the inferior olivary nucleus to the cerebellum is still
cally old systems are functional from early stages. The laterally unclear. In one theory, the climbing fiber system, together with
descending projections project upon motoneurons of extremity the parallel fiber system (see above), is considered to be decisive
muscles. The cerebellum, via the crossed rubrospinal tract starts its in motor learning (Marr, 1969). A more recent theory suggests
adult-like functioning at a late stage. that the ION plays an important role in regulating a distributed
processing of afferent information in the cerebellar cortex (Llinas
& Muhletaler, 1988).
Central brain mechanisms and locomotion The outflow of cerebellar processing is transported via the deep
cerebellar nuclei to the red nucleus and thalamic nuclei and from
there along descending tracts to pre-motoneurons of extremity
The mesencephalic locomotor region muscles. This input is crucial for postural adjustments and postural
Strong evidence has been collected from research in cats that the control during walking and running.
rhythmic leg movements in locomotion are ‘initiated’ by activity in Our own experiments in rats indicated that the cerebellum plays
a group of cells in the mesencephalic brain stem. Electrical stimula- an important role in adjusting postural control to limb movements
tion of this area in intact cats induced the leg movements on a during walking. This postulate is based upon the finding that ves-
moving treadmill (Shik et al., 1966). Since then this area has been tibular deprivation in rats retards the development of postural
termed the mesencephalic locomotor region (MLR). Anatomically, control and this, in turn, leads to a delay in the development of the
its neurones are localized around the pedunculopontine tegmental adult-like walking pattern (see, above; Gramsbergen, 1998, 2001,
nucleus (Spann & Grofova, 1989). Descending fibers probably 2005a, b).
course via the reticulospinal tract and impinge upon the CPGs
where they start the rhythmical activity.
The experiments indicated also that by increasing the stimulus
Summary
intensity the walking speed increased (by an increase in the CPG- Locomotion is initiated by activity in the MLR, which in turn might
frequency) and eventually the animals made a transition to trot or be activated by higher centers. The LMS plays a role in the accelera-
gallop (indicating an alternative phasing-pattern within the CPGs) tion or deceleration of walking speed. Cerebellar circuitry plays a
(Shik et al., 1966; Shik & Orlovsky, 1976). The speed ‘increases’ and role in adjustments of postural control during walking.

79
 4 Locomotor neurobiology and development

Walking in adult animals µV

200
During walking, limbs are in contact with the ground during the
stance phase and are moved forward during the swing phase. The GC, right
foot placement sequence during walking at low speeds typically is
LH–LF–RH–RF, with the body always being supported by three
limbs. The extension phases of the two hind limbs are phase shifted 0
* * * * * *
by 0.5, which implies that when one limb is starting its stepcycle 100
the other limb is halfway through its cycle. When walking along
irregular trajectories, these phase shifts change. This alternating gait
TA, right
pattern (Grillner, 1975, 1981) applies not only to rats but also to
many quadrupeds, including horses, during walking at lower speeds.
When walking speed increases and when the footfall pattern 0
indicated above is maintained, the stride length increases and the
stance phase decreases, but the duration of the swing phase remains 30
more or less constant (Westerga & Gramsbergen, 1990). During
further increases in speed, e.g. in horses, walking changes into a trot LL, right
and then to a gallop, which are characterized by different footfall caudal
patterns. Transitions between gaits during acceleration or decelera-
tion occur at variable speeds (Grillner, 1975, 1981). 0
The kinematic and neurophysiological aspects of these patterns 60
in rats have not been studied in detail but abundant data are avail-
able on horses (see other chapters in this book). The few studies LL, right
that have been performed aimed at investigating the energy expen- intermediate
diture and force production by muscles during such patterns (e.g.
in mice: James et al., 1995; in rats: Sullivan & Armstrong, 1978;
0
Taylor et al., 1982; Perry et al., 1988).
An important difference between smaller and larger animals 40
during fast locomotion is that in species such as rats, cats and dogs
the hind limbs are used for acceleration and maintenance of speed LL, right
and the forelimbs for braking, while in larger animals such as rostral
horses, the hind and forelimbs promote similar functions during
acceleration and braking (Heglund et al., 1982). These differences 0
most probably are reflected in neural circuitries as well as in mus- 0 1
cular and skeletal specializations of the hind and forelimbs. Time (s)
Fig 4.3  Averaged EMG records of a rat aged 32 days. Recordings from the
Summary gastrocnemius (GC) muscle at the right side, the tibialis (TA) muscles at the
right side, and the longissimus (LL) muscles at caudal, intermediate and rostral
Walking, trotting and galloping are characterized by different foot-
levels. Asterisks indicate the onsets of the stance phase in the hind limb at the
fall patterns and transitions between these patterns occur at variable
right side.
speeds. In horses, the hind and forelimbs are both used for accel-
Reprinted from Gramsbergen, A., Geisler, H.C., Taekema, H., van Eyrken, L.A., 1999, The activa-
erating and braking.
tion of back muscles during locomotion in the developing rat. Developmental Brain Research
112 (2), 217–228, with permission from Elsevier.
EMG recordings during locomotion
in adult animals shortly before the onset of the swing phase. Bursts in these muscles
are characterized by a short ‘attack’ phase and the bursts generally
Most of the investigations into electromyographic activity of leg only last during the first part of the swing phase (the F phase).
muscles during walking in rats and cats have been performed in the Between the bursts EMG activity generally is absent. Remarkably,
hind limb (for reviews, Grillner, 1975, 1981; Grillner et al., 1991). the onsets of the EMG bursts in these muscles coincide, despite the
Globally, the extensors of the hip, knee, ankle and digits are acti- fact that they move different limb segments and that slight phase
vated during the stance phase and the flexors are active during the shifts in joint angle trajectories occur (Westerga & Gramsbergen,
swing phase (Fig. 4.3). EMG patterns in the forelimb have been 1990). During increases in speed, the burst durations in these flexor
studied less frequently. In the newt, extensors and flexors of the muscles remain unchanged, but the EMG amplitudes increase
forelimb are activated rhythmically and reciprocally but the EMG (Westerga & Gramsbergen, 1993; Gramsbergen et al., 1999). In cats,
patterns are more complex and co-contractions are often observed EMG recordings in the iliopsoas and the extensor digitorum longus
(Szekely et al., 1969). muscles as well as in the tibialis anterior muscle showed similar
In the literature on kinematic and electromyographic aspects of results (e.g. Engberg & Lundberg, 1962, 1969).
the step cycle in cats, the ‘extension phase’ generally is subdivided The gastrocnemius, the soleus (ankle extensors) and the quadri-
into an E1 phase indicating the onset of limb extension before ceps femoris (knee extensor) muscles are activated shortly before
contacting the ground, the E2 phase during the midphase of ground the onset of the stance phase (during the E2 phase) and EMG activ-
contact and the E3 phase with the foot still in contact with the ity lasts until shortly before the onset of the next swing phase
ground but shortly before toe-off. The swing phase thus consists of (Westerga & Gramsbergen, 1993, 1994; Gramsbergen et al., 1999).
the F (flexion) and E1 phase, and the stance phase consists of the The bursts in these extensor muscles start simultaneously (as occurs
E2 and E3 phases. in the flexors) and the profiles of the EMG bursts are more or less
In the hind limbs of freely moving rats, the tibialis anterior (ankle identical. During increases in speed, the burst amplitudes in
flexor) and the semitendinosus muscles (knee flexor) are activated these muscles remain the same but in the extensors the burst

80
 Development of postural control

F E1 E2 E3 Summary
Flexors in the hind limb are activated simultaneously, shortly before
the swing phase of the hind limb and extensors are activated shortly
QF Knee before the stance phase. The longissimus muscles in the back are
GC Ankle
activated during extension of the hind limbs and EMG activity in
this muscle is strongest during the stance phase of the ipsilateral
SOL Ankle hind limb.

ST Knee Development of postural control

TA Ankle
Adequate postural control is an important factor in the develop-
Swing Stance ment of standing and locomotion. Posture may be defined as the
relative positions of the head, trunk and the extremities as well as
Flexion Extension the orientation of the body in space. In adult animals, both feed-
back (‘static’; e.g. during standing) and feed-forward (‘dynamic’;
during locomotion and head movements) control mechanisms play
H a role (Massion, 1992). Although it is impossible to distinguish the
K
walking movements from postural adjustments from a behavioral
A
point of view, it should be remembered that the neural systems
Fig 4.4  Schematic representation of EMG activity in a few hind limb muscles governing posture and movement are organized differently and
in relation to extension and flexion phases and swing and stance phases. QF, have a different evolutionary descent (see above). On the one hand,
quadriceps femoris muscle; GC, gastrocnemius muscle; SOL, soleus muscle; specific central motor areas, specialized sensory systems and medi-
ST, semitendinosus muscle; TA, tibialis anterior muscles (extensors and flexors ally descending spinal projections are involved in postural control.
of the knee and ankle, respectively). Blocks indicate burst activity: E1, extension On the other hand, neural systems with laterally descending fiber
of hind limb before stance phase; E2, midstance phase; E3, extension shortly projections are involved in governing extremity movements. Obvi-
before toe-off; F, flexion. Stick diagrams are derived from video recordings: ously, intimate connections exist between these systems at adult age
H, hip; K, knee; A, ankle. (and the cerebellum is crucial in that coupling).
Horses are born at an advanced stage of brain development and
within hours of birth they are able to stand and to walk. This indi-
durations decrease (Fig. 4.4), which agrees with data obtained in cates that postural mechanisms and vestibular reflexes must have
cats (Grillner, 1981). matured before birth in this species. Experiments in guinea pigs
EMG activity in the postural muscles in the back is related in a (guinea pigs are (as horses) a precocial species and neural develop-
complex way to the activity in the hind limb extensors and flexors. ment for an important part has occurred intra-uterinely) have shed
In adult rats, EMG patterns were recorded in the medially located light on the development of the vestibulum. The structural develop-
multifidus muscles and the laterally located longissimus muscles ment of the vestibulum in guinea pigs is completed before birth but
(Geisler et al., 1996b; Gramsbergen et al., 1999). The multifidus its functioning is inhibited prenatally possibly due to low levels of
muscle is tonically active during all movements that require stabi- oxygen content in the placental blood (Peter Schwartze, University
lization of the trunk, and during locomotion the activity in this of Leipzig, Germany, pers. commun.). This active inhibition is most
muscle modulates slightly with the frequency of the left and right useful as it prevents the occurrence of vestibular reflexes of the
hind limb. The EMG in the longissimus muscle, on the other fetuses before birth (e.g. during vigorous movements of the mother,
hand, is more phasic in character. EMG bursts in these muscles are such as during walking). In horses this might be similar. Immedi-
phase-linked to bursts in the gastrocnemius muscles during the ately after birth, the inhibition is relieved because of a fast rise in
stance phase (Gramsbergen et al., 1999). This is remarkable the oxygen content in the arterial vascular bed. This, in turn, leads
because intuitively, it seems that a swing phase in one of the hind to sudden development of vestibular reflexes (which enables an
limbs induces a relative instability of the trunk which should be upright stance).
counteracted by activity in the trunk muscles. During one step Experiments in rats by Geisler et al. (1996a) showed that
cycle of both hind limbs two bursts in the longissimus muscles the development of postural control is the limiting factor for
occur. The burst with the ‘higher’ amplitude accompanies exten- walking development. She demonstrated that vestibular deprivation
sion of the ‘ipsilateral’ hind limb and that with a ‘lower’ activity by plugging the semicircular canals from the 5th postnatal day
the extension of the ‘contralateral’ leg. EMG activity at more selectively induces a retardation in postural development. Motor
rostral levels in the longissimus muscle is variable in amplitude, patterns which specifically require high levels of trunk control such
duration and phasing of the bursts (with reference to the stance as grooming (self-cleaning behavior) were retarded by 1–2 days.
phases in the hind legs). In horses, similar results were obtained Rearing (standing on the hind limbs without support from the wall)
by recordings of the multifidus muscles during trotting on a was even delayed by as much as 5 days. Movements of the extremi-
moving belt (Licka et al., 2004). ties were essentially normal, but the adult-like type of smooth
During rapid acceleration and fast walking, simultaneous con- walking (with the body free from the floor – indicating control of
tractions may occur over the entire length of the muscle but at slow the trunk muscles) was retarded by 3–4 days. These experiments
speeds the rostral portions of the longissimus muscles only show demonstrate that postural development is the limiting factor for the
tonic activity without modulations in amplitude (Fig. 4.4). development of walking (Gramsbergen, 1998, 2001, 2005a, b).
These data on adult rats are in agreement with results from adult
cats. Carlson et al. (1979) reported a similar coupling in the activa- Summary
tion of back muscles with the vastus lateralis muscle in the hind
limb during treadmill walking. Their data, however, indicate that In rats, around the end of the 2nd week after birth, the immature
the two bursts-per-step cycle in both hind limbs and in the longis- walking pattern is replaced by a smooth adult-like walking pattern.
simus muscles are more or less identical in amplitude, which might The development of postural control probably is the limiting factor
point to a peculiarity of treadmill walking. for this transition.

81
 4 Locomotor neurobiology and development

Synthesis MLR (initiation)

LMS (pace)
Control of the postural muscles in the trunk and the neck on the
one hand, and that of the extremity muscles on the other are gov- VSCT VSCT
erned by different neural motor systems. Neural control of axial SRCP SRCP
muscles in the trunk and neck is provided by old structures from
an evolutionary point of view. On the other hand, the neural LVST LVST
control of the extremities is from a more recent evolutionary stage. RST RST
From this perspective it is remarkable that postural control, by the
older system, is the limiting factor for the development of the adult
pattern of fluent and swift locomotion. This has been demonstrated
in rats and also data on the development of standing and free
walking in human infants are congruent with this conclusion. As
horses are born at an advanced stage of brain development, this
maturational order is less obvious. The cerebellum plays important
Fl CPG Fl
functions, not only in finely tuning muscle forces during move-

Muscles

Muscles
ments and acquisition and performance of skilled movement pat-
terns but also in making postural adjustments during extremity Ext Ext
movements. As to this latter function, the data available suggest that
the cerebellum is the key structure in coupling and adjusting the Feedback Feedback
activities of the evolutionary newer structures, i.e. the muscles in
the legs and their control with those of the older structures, i.e. the Fig 4.5  Diagram representing CPG in relation to motorneurons (Fl, flexors;
axial muscles along the trunk and their neural control. Trunk Ext, extensors) and ascending and descending tracts. CPG, central pattern
muscles and particular extremity muscles, the so-called anti-gravity generator; Feedback, proprioceptive feedback from muscles, tendons and
muscles are specialized to subserve postural tasks. They contain joint receptors; RST, rubrospinal tract; LVST, lateral vestibulospinal tract; VSCT,
ventral spinocerebellar tract; SRCP, spinoreticulocerebellar tract; MLR, mesen-
large proportions of type I muscle fibers which are specifically able
cephalic locomotor region; LMS, limbic motor system.
to sustain force-loads for long periods. These muscles are innervated
by motoneurons which, in part, are connected by dendrite bundles
(possibly an evolutionary older type of interneuronal connectivity;
Gramsbergen et al., 1996; Gramsbergen, 2005b). These muscles,
containing large proportions of type I muscle fibers are innervated rats) a locomotor region in the mesencephalon exists. This cell
by motoneuronal pools with conspicuous dendrite bundles (as in group, in turn, is activated by other brain areas, and most probably
the soleus muscle and the trunk muscles), or in other cases by by the sensory-motor cortex (or it’s analogue in horses). The speed
motoneuronal pools with a few regions of dendrite bundles (as the of walking is adjusted by the LMS. In situations of fear or strong
gastrocnemius and vastus medius muscles). Other extremity muscles activation during races, the LMS may induce the CPG to switch to
and particularly the flexor muscles are composed of large propor- increased rhythms and incidentally to other coordination
tions of type IIa and IIb muscles. These muscles are particularly patterns.
involved in heavy exercises which they can, however, can only Training procedures in horses aim at optimizing muscular sizes
sustain for relatively short periods. and muscular properties. EMG-recordings and advanced automatic
The basic rhythmicity in alternating limb movements (even at analysis techniques such as the computation of power density
varying footfall patterns) is produced by neuronal circuits, localized spectra and the recording of shifts in these spectra during exercises
at spinal levels, CPGs (Fig. 4.5). and training or fatigue in conjunction with kinematic recordings
Walking starts by activation from the MLR in the mesencephalic and analysis will prove to be important tools in devising the most
brain stem. It is supposed that also in horses (like in cats and in effective training schemes.

References
Ariens Kappers, C.U., Huber, G.C., Crosby, Behrens, T.E., Berg, H.J., Jbadi, S., Rushworth, Burke, R.E., 1981. Motor units: anatomy,
E.C., 1967. The comparative anatomy of M.F., Woolrich, M.W., 2007. Probabilistc physiology, and functional organization.
the nervous system of vertebrates including diffusion tractography with multiple fibre In: Brooks, V.B. (Ed.), Handbook of
man. Hafner, New York. orientations: what can we gain. physiology. American Physiological Society,
Armand, J., 1982. The origin, course and Neuroimage 34, 144–155. Bethesda, Md., pp. 345–423.
terminations of corticospinal fibers in Bennett, M.R., 1983. Development of Carlson, H., Halbertsma, J., Zomlefer, M.,
various mammals. In: Kuypers, H.G.J.M., neuromuscular synapses. Physiol. Rev. 63, 1979. Control of the trunk during walking
Martin, G.F. (Eds.), Descending pathways 915–1048. in the cat. Acta. Physiol. Scand. 105,
to the spinal cord. Progress in Brain Brooks, V.B., 1975. Roles of the cerebellum 251–253.
Research, vol. 57. Elsevier, Amsterdam, pp. and basal ganglia in initiation and control Cazalets, J.R., Borde, M., Clarac, F., 1995.
327–360. of movements. Can. J. Neurol. Sci. 2, Localization and organization of the
Arshavsky Yu, I., Berkinblit, M.B., Fukson, O.I., 265–277. central pattern generator for hind limb
Gelfand, I.M., Orlovsky, G.N., 1972. Origin Brown, T.G., 1914. On the nature of the locomotion in newborn rat. J. Neurosci.
of modulation in neurones of the ventral fundamental activity of the nervous centers; 15, 4943–4951.
spinocerebellar tract during locomotion. together with an analysis of the Dubowitz, V., Brooke, M.H., 1973. Muscle
Brain Res. 43, 276–279. conditioning of rhythmic activity in biopsy: a modern approach. In:
Arshavsky Yu, I., Gelfand, I.M., Orlovsky, G.N., progression, and a theory of the evolution Walton, J.N. (Ed.), Major problems in
1986. Cerebellum and rhythmical of function in the nervous system. J. neurology, vol. 2. W.B. Saunders, London,
movements. Springer, Berlin, Heidelberg. Physiol. (Lond.) 48, 18–46. pp. 10–100.

82
 References

Duclay, J., Martin, A., Duclay, A., Cometti, G., of dendrite bundles in the rat spinal cord. Holstege, G., 1991. Descending motor
Pousson, M., 2009. Behavior of fascicles Soc. Neurosci. Abstr. 21, 1790. pathways and the spinal motor system:
and the myotendinous junction of human Gramsbergen, A., IJkema-Paassen, J., Nikkels, limbic and non-limbic components. Prog.
medial gastrocnemus following eccentrc P.G.J., Hadders-Algra, M., 1997. Regression Brain Res. 87, 307–421.
strength training. Muscle & Nerve 39, of polyneural innervation in the human Holstege, G., 1995. The basic, somatic, and
819–827. psoas muscle. Early Human Develop 49, emotional components of the motor
Duysens, J., Pearson, K.G., 1980. Inhibition of 49–61. system in mammals. In: Paxinos, G. (Ed.),
flexor burst generation by loading ankle Gramsbergen, A., IJkema-Paassen, J., Westerga, The rat nervous system, 2nd edn. Academic
extensor muscles in walking cats. Brain. J., Geisler, H.C., 1996. Dendrite bundles in Press, San Diego, pp. 137–154.
Res. 187, 321–332. motoneuronal pools of trunk and extremity Hultborn, H., 1972. Convergence on
Engberg, I., Lundberg, A., 1962. An muscles in the rat. Exp. Neurol. 137, interneurons in the reciprocal Ia inhibitory
electromyographic analysis of stepping in 34–42. pathway to motoneurones. Acta. Physiol.
the cat. Experientia 18, 174. Gramsbergen, A., van Eykern, L.A., Taekema, Scand. 375 (Suppl.), 1–42.
Engberg, I., Lundberg, A., 1969. An H., Geisler, H.C., 1999. The activation of IJkema-Paassen, J., Gramsbergen, A., 1998.
electromyographic analysis of muscular back muscles during locomotion in the Regression of polyneural innervation in the
activity in the hind limb of the cat during developing rat. Dev. Brain Res. 112, psoas muscle of the developing rat. Muscle
unrestrained locomotion. Acta. Physiol. 217–228. and Nerve 21, 1058–1063.
Scand. 75, 614–630. Greensmith, L., Vrbova, G., 1991. James, R.S., Altringham, J.D., Goldspink, D.F.,
Forssberg, H., Grillner, S., Halbertsma, J., Neuromuscular contacts in the developing 1995. The mechanical properties of fast
Rossignol, S., 1980. The locomotion of the rat soleus depend on muscle activity. Dev. and slow skeletal muscles of the mouse in
low spinal cat. 2: interlimb coordination. Brain. Res. 62, 121–129. relation to their locomotory function. J.
Acta. Physiol. Scand. 108, 283–295. Grillner, S., 1975. Locomotion in vertebrates: Exp. Biol. 198, 491–502.
Geisler, H.C., Gramsbergen, A., 1998. The central mechanisms and reflex interaction. Jansen, J.K.S., Fladby, T., 1990. The perinatal
EMG development of the longissimus and Physiol. Rev. 55, 247–304. reorganization of the innervation of
multifidus muscles after plugging the Grillner, S., 1981. Control of locomotion in skeletal muscle in mammals. Prog.
horizontal semicircular canals. J. Vestib. bipeds, tetrapods, and fish. In: Brooks, V.B. Neurobiol. 34, 39–90.
Res. 8, 1–11. (Ed.), Handbook of physiology. American Kernell, D., 1998. Muscle regionalization. Can.
Geisler, H.C., van der Fits, I.B.M., Physiological Society, Bethesda, Md., pp. J. Appl. Physiol. 23, 1–22.
Gramsbergen, A., 1996a. The effect of 1179–1236. Kernell, D., 2006. The motoneurone and its
vestibular deprivation on motor Grillner, S., Wallén, P., Brodin, L., Lansner, A., muscle fibres. Oxford University Press,
development in the rat. Behav. Brain Res. 1991. Neuronal network generating Oxford.
86, 89–96. locomotor behaviour in lamprey: circuitry, Kjaerulff, O., Kiehn, O., 1996. Distribution of
Geisler, H.C., Westerga, J., Gramsbergen, A., transmitters, membrane properties, and networks generating and coordinating
1996b. The function of the long back simulation. Ann. Rev. Neurosci. 14, locomotor activity in the neonatal spinal
muscles; and EMG study in the rat. Behav. 169–199. cord in vitro: a lesion study. J. Neurosci. 16,
Brain Res. 80, 211–215. Grillner, S., Zangger, P., 1975. How detailed 5777–5794.
Gellman, K.S., Bertram, J.E., Hermans, J.W., is the central pattern generator for Kuypers, H.G.J.M., 1982. A new look at the
2002. Morphology, histochemistry, and locomotion? Brain. Res. 88, 367–371. organization of the motor system. In:
function of epaxial cervical musculature in Grotmol, S., Totland, G.K., Kryvi, H., Bristol, Kuypers, H.G.J.M., Martin, G.F. (Eds.),
the horse (equus caballus). J. Morphol. A., Essén-Gustavsson, B., Lindholm, A., Descending pathways to the spinal cord.
251, 182–194. 2002. Spatial distribution of fibre types Progress in brain research, vol. 57. Elsevier,
Goldspink, G., Ward, P.S., 1979. Changes in within skeletal muscle fascicles from Amsterdam, pp. 381–403.
rodent muscle fiber types during postnatal standardbred horses. Anat Rec. 268, Laughlin, M.H., Roseguini, B., 2008.
growth, undernutrition and exercise. J. 131–136. Mechanisms for exercise training-induced
Physiol. (Lond.) 296, 453–469. Grünheid, T., Langenbach, G.E., Korfage, J.A., increases in skeletal muscle blood flow
Gramsbergen, A., 1982. The effects of Zentner, A., van Eijden, T.M., 2009. The capacity: differences with interval sprint
cerebellar hemispherectomy in the young adaptive response of jaw muscles to varying training versus aerobic endurance
rat. I behavioural sequelae. Behav. Brain functional demands. Eur J. Orthod. (E.publ training. J. Physiol. Pharmacol. 59
Res. 6, 85–92. ahead of print). (Suppl. 7), 71–88.
Gramsbergen, A., 1998. Posture and Heglund, N.C., Cavagna, G.A., Taylor, C.R., Lawrence, D.G., Hopkins, D.A., 1976. The
locomotion in the rat: independent or 1982. Energetics and mechanics of development of motor control in the
interdependent development. Neurosci. terrestrial locomotion. III energy changes of rhesus monkey: evidence concerning the
BioBehav. Rev. 22, 547–554. the center of mass as a function of speed role of corticomotoneuronal connections.
Gramsbergen, A., 2001. Neuro-ontogeny of and body size in birds and mammals. J. Brain 99, 235–254.
motor behaviour in the rat. In: Kalverboer, Exp. Biol. 97, 41–56. Licka, T.F., Peham, C., Frey, A., 2004.
A.F., Gramsbergen, A. (Eds.), Handbook of Henneman, E., Mendell, L.M., 1981. Electromyographic activity of the
brain and behaviour in human Functional organization of motoneuronal longissimus dorsi muscles in horses during
development. Kluwer Ac Publ, Dordrecht, pool and its inputs. In: Brooks, V.B. (Ed.), trotting on a treadmill. Am. J. Vet. Res. 65,
Boston, London, pp. 467–512. Handbook of physiology. American 155–158.
Gramsbergen, A., 2005a. Postural control in Physiological Society, Bethesda, Md., pp. Llinas, R., Muhletaler, M., 1988.
man, the phylogenetic perspective. Neural 43–106. Electrophysiology of guinea pig cerebellar
Plast. 12, 77–88. Hobbelen, J.F., Gramsbergen, A., van Hof, nuclear cells in the in vitro brainstem-
Gramsbergen, A., 2005b. Brain and behavioral M.W., 1992. Descending motor cerebellar preparation. J. Physiol. (Lond.)
development (I): sub-cortical. In: Hopkins, pathways and the hopping response 404, 241–258.
B. (Ed.), The cambridge encyclopedia of in the rabbit. Brain. Behav. Res. 51, Marr, D., 1969. A theory of cerebellar cortex. J.
child development. Cambridge University 217–221. Physiol. (Lond.) 202, 437–470.
Press, Cambridge, pp. 287–295. Holst, E. von, 1935. Über den prozess der Massion, J., 1992. Movements, posture and
Gramsbergen, A., IJkema-Paassen, J., Klok, F., zentralnervösen koordination. Pflugers equilibrium: interaction and coordination.
1995. Factors influencing the development Arch. 236, 149–158. Prog. Neurobiol. 38, 35–56.

83
 4 Locomotor neurobiology and development

Matthews, M.A., Willis, W.D., Williams, V., Rossignol, S., Lund, J.P., Drew, T., 1988. The Udo, M., Matsukawa, K., Kamei, H., 1979.
1971. Dendrite bundles in lamina IX of cat role of sensory inputs in regulating patterns Effects of partial cooling of cerebellar
spinal cord. A possible source for electrical of rhythmical movements in higher cortex at lobules V and IV of the
interaction between motoneurons? Anat. vertebrates. In: Cohen, A.H., Rossignol, S., intermediate part in decerebrate walking
Rec. 171, 313–328. Grillner, S. (Eds.), Neural control of cats under monitoring vertical floor
Mileva, K.N., Morgan, J., Bowtell, J., 2009. rhythmic movements in vertebrates. Wiley, reaction forces. Brain Res. 160, 559–564.
Differentiation of power and endurance New York, pp. 201–283. van der Want, J.J.L., Gramsbergen, A.,
athletes based on their muscle fatigability Ruigrok, T.J.H., Cella, F., 1995. Precerebellar IJkema-Paassen, J., Weerd, H., de Liem,
assessed by new spectral electromyographic nuclei and red nucleus. In: Paxinos, G. R.S.B., 1998. Dendro-dendritic connections
indices. J. Sports Sci. 27, 611–623. (Ed.), The rat nervous system. 2nd edn. between motoneurons in the rat spinal
Milner-Brown, H.S., Stein, R.B., Yemm, R., Academic Press, San Diego, pp. 277–300. cord. An electronmicroscopic investigation.
1973. The orderly recruitment of human Scheibel, M.E., Scheibel, A.B., 1970. Brain Res. 779, 342–345.
motor units during voluntary isometric Developmental relationship between spinal Van Weeren, P.R., 2001. History of locomotor
contractions. J. Physiol. (Lond.) 230, motoneuron dendrite bundles and research, In: Back, W., Clinton, H. (Eds.),
359–370. patterned activity in the hind limb of cats. Equine locomotion. W.B. Saunders
Miyata, H., Sugiura, T., Kai, M., Hiraga, A., Exp. Neurol. 29, 328–335. London, pp. 1–36.
Tokuriki, M., 1999. Muscle adaptation of Schiaffino, S., Saggin, L., Viel, A., Ausoni, S., Verhaart, W.J.C., Sopers-Jurgens, M.R., 1957.
thoroughbred racehorses trained on a flat Sartore, S., Gorza, L., 1986. Muscle fiber Aspects of comparative anatomy of the
or sloped track. Am. J. Vet. Res. 60, types identified by monoclonal antibodies mammalian brain stem. Acta. Morph.
1536–1539. to myosin heavy chains. In: Benzi, G., Neerl-Scand. 1, 246–255.
Nollet, H., Van Ham, L., Dewulf, J., Packer, L., Siliprandi, N. (Eds.), Wang, L.C., Copray, S., Brouwer, N., Meek,
Vanderstraeten, G., Deprez, P., 2003. Biochemical aspects of physical exercise. M.F., Kernell, D., 2002. Regional
Standardization of transcranial magnetic Elsevier, Amsterdam, pp. 27–34. distribution of slow-twitch muscle fibres
stimulation in the horse. Vet. J. 166, Schoenen, J., 1982. Dendritic organization of after reinnervation in adult rat hind limb
244–250. the human spinal cord: the motoneurons. muscles. Muscle and Nerve 25, 85–815.
Nudo, R.J., Masterton, R.B., 1988. Descending J. Comp. Neurol. 21, 226–247. Wang, L.C., Kernell, D., 2001. Fibre type
pathways to the spinal cord: a comparative Sherrington, C.S., 1910. Flexion reflex of the regionalisation in lower hind limb muscles
study of 22 mammals. J. Comp. Neurol. limb, crossed extension reflex and reflex of rabbit, rat and mouse: a comparative
277, 53–79. stepping and standing. J. Physiol. (Lond.) study. J. Anat. 199, 631–643.
O’Brien, R.A.D., Ostberg, A.J.C., Vrbova, G., 40, 28–121. Weiss, P.A., 1936. A study of motor
1978. Observations on the elimination of Shik, M.L., Orlovsky, G.N., 1976. coordination and tonus in deafferented
polyneural innervation in developing Neurophysiology of locomotor limbs of amphibia. Am. J. Physiol. 115,
mammalian skeletal muscle. J. Physiol. automatism. Physiol. Rev. 56, 465–501. 461–475.
(Lond.) 282, 571–582. Shik, M.L., Severin, F.V., Orlovsky, G.N., 1966. Westerga, J., Gramsbergen, A., 1990. The
Ontell, M., Dunn, R.F., 1978. Neonatal muscle Control of walking and running by means development of locomotion in the rat.
growth: a quantitative study. Am. J. Anat. of electrical stimulation of the midbrain. Develop. Brain Res. 57, 163–174.
152, 539–556. Biophysics 11, 756–765. Westerga, J., Gramsbergen, A., 1992. Structural
Partridge, L.D., Benton, L.A., 1981. Muscle, the Spann, B.M., Grofova, I., 1989. Origin of changes of the soleus and the tibialis
motor. In: Brooks, V.B. (Ed.), Handbook of ascending and spinal pathways from the anterior motoneuron pool during
physiology. American Physiological Society, nucleus tegmenti pedunculopontinus in development in the rat. J. Comp. Neurol.
Bethesda, Md., pp. 43–106. the rat. J. Comp. Neurol. 283, 13–27. 319, 406–416.
Perry, A.K., Blickhan, R., Biewener, A.A., Sullivan, T.E., Armstrong, R.B., 1978. Rat Westerga, J., Gramsbergen, A., 1993.
Heglund, N.C., Taylor, C.R., 1988. Preferred locomotory muscle fiber activity during Changes in the EMG of two major
speeds in terrestrial vertebrates: are they trotting and galloping. J. Appl. Physiol. 44, hind limb muscles during locomotor
equivalent? J. Exp. Biol. 137, 207–219. 358–363. development in the rat. Exp. Brain Res.
Rajaofetra, N., Sandillon, F., Geffard, M., Szekely, G., Czeh, G., Vörös, G., 1969. The 92, 479–488.
Orivat, A., 1989. Pre- and post-natal activity pattern of limb muscles in freely Westerga, J., Gramsbergen, A., 1994.
ontogeny of serotonergic projections to the moving and deafferented newts. Exp. Brain Development of the EMG of the soleus
rat spinal cord. J. Neurosc. Res. 22, Res. 9, 53–62. muscle in the rat. Dev. Brain Res. 80,
305–321. Taylor, C.R., Heglund, N.C., Maloiy, G.M., 233–243.
Reback, P.A., Scheibel, A.B., Smith, J.L., 1982. 1982. Energetics and mechanics of Wilson, S.J., Ross, J.J., Harris, A.J., 1988. A
Development and maintenance of dendrite terrestrial locomotion. I. metabolic energy critical period for formation of secondary
bundles after cordotomy in exercised and consumption as a function of speed and myotubes defined by prenatal
nonexercised cats. Exp. Neurol. 76, body size in birds and mammals. J. Exp. undernourishment in rats. Development
428–440. Biol. 97, 1–21. 102, 815–821.

84
 C H A PTER 5 

Gaits and interlimb coordination

Eric Barrey

Introduction Another type of gait classification distinguishes between stepping

or walking gaits that have no period of suspension (there is always
a contact with the ground) and running gaits that have one or more
The locomotor apparatus is a complex set of systems including suspension phases in each stride (no foot in contact with the
muscle, bone segments and joints, which are controlled by the ground). The main characteristics of equine gaits are described in
central nervous system to produce well-coordinated locomotion. Table 5.1. Within each gait there exist continuous variations, from
Biomechanically, locomotion involves moving all the body and a collected type of gait with a slow speed, to an extended type of
limb segments in rhythmic and automatic patterns, which define gait with a higher speed.
the various gaits. A great diversity exists in equine gait patterns Many types of illustrations have been proposed to describe the
because quadrupedal locomotion allows many combinations of limb movements more precisely in time and space: drawings, chro-
inter-limb coordination. Furthermore, horse breeds have been nophotographs, bar diagrams, phase diagrams and pie diagrams
genetically selected for different occupations: draft, riding, driving (Fig. 5.1). Some of the methods are only descriptive and show either
and meat production. Sport horses compete in a variety of sports, temporal or linear characteristics of gaits:
including pacing, trotting and galloping races, eventing, show
jumping, dressage, endurance and western disciplines. Conse- • Drawings of the footfall sequence (Barroil, 1887) (Fig. 5.1A).
quently, a large range of gaits and gait variations can be observed • Gait or bar diagrams and hoof imprints which shows tracks
in horses, including the walk and its many variations, trot, pace, and linear distances (Lenoble du Teil, 1893) (Fig. 5.1B). Gait
canter and gallop. These gaits can be analyzed and classified accord- and bar diagrams can also describe temporal footfall sequence
ing to their linear, temporal and dynamic characteristics using the (Marey, 1873) (see Fig. 5.4B).
measuring techniques that have been described in Chapter 3. The • Pie gait diagram showing the footfall sequence and relative
variety and complexity of gaits have always created difficulties in durations of the suspension phase, stance phases and overlaps
defining a terminology that is, at the same time, broad enough and (Deuel & Lawrence, 1984) (Fig. 5.1D).
also sufficiently specific to describe the locomotor phenomenon. Two other methods represent and classify the gaits in a more func-
Some efforts have been made to define a standard terminology for tional way using phase lag or advance between the footfalls:
describing equine locomotion (Leach et al., 1984a; Clayton, 1989;
Leach, 1993) The gait terminology used in this book is defined in • The continuum of symmetric gaits was described by a diagram
the glossary (p. xix). proposed by Hildebrand (1965) (Fig. 5.2). The stance duration
of the hind limb was plotted against the lateral advanced
placement. On the x-axis, the stance duration of the hind limb
Classification and description of gaits indicates if the gait is classified as walking (no suspension
phase) or running (two suspension phases per stride in the
symmetric gaits). On the y-axis, the lateral advanced placement
A gait can be defined as a complex and strictly coordinated, rhyth-
quantifies the phase lag of the lateral fore and hind limbs. The
mic and automatic movement of the limbs and the entire body of
two-beat gaits are at the top and bottom of the diagram, with
the animal, which results in the production of progressive move-
the four-beat gaits between them. A similar diagram has been
ments. In horses, a two-, three- or four-beat gait corresponds to the
proposed for illustrating and comparing the diagonal gaits by
number of footfalls that can be heard during each stride of trot,
plotting the hind stance phase duration against the diagonal
canter and gallop, respectively. The sounds are related to the footfall
advanced placement (Clayton, 1997).
pattern of the gait but, when the interval between footfalls is very
short, the human ear is not able to perceive the separation. • A group of methods describe the type of coupling between
the four limbs. All the running gaits can be modeled using
One method of classifying gaits depends on the symmetry
the relative phases of the limb cycle (Alexander, 1984) (Fig.
between the left and right sides. In a symmetric gait, the left and
5.1C). However, this diagram describes only the footfalls
right footfalls of the fore and hind limbs are evenly spaced in time.
sequence but does not give any linear or temporal
In an asymmetric gait, the footfalls of the fore and/or hind limbs
characteristics of the stride. A more sophisticated method
occur as couplets. The first limb of a couplet to contact the ground
based on a series of coupled oscillators has been proposed to
is the trailing limb; the second is the leading limb:
describe and simulate both symmetric and asymmetric gaits
• symmetric gaits: walk, trot, running walk, rack, toelt, fox trot, (Collins & Stewart, 1993). This model has the advantage of
paso, stepping pace being able to describe gait transitions and abnormal gaits like
• asymmetric gaits: canter, transverse and rotary gallop, half the aubin, in which the hind limbs trot while the forelimbs
bound. gallop, and traquenard, in which the hind limbs gallop while

85
86
Table 5.1  Classification and main characteristics of equine gaits
5

Classification Gait Gait variations Footfall Rhythm Type of Speed Stride Stride Limb Suspension
sequence (beat/stride) symmetry (m/s) length frequency stance phase (s or %
(m) (stride/s) phase (s or stride)
% stride)
Walking gaits Walk Collected, medium, RH, RF, LH, LF 4 Right/left 1.2–1.8 1.5–1.9 0.8–1.1 65–75% 0
extended, free bipedal
Toelt, paso, RH, RF, LH, LF 4 Right/left 3.4–5.3 1.7–2.3 2.23–2.36 40–55% 0
rack, foxtrot lateral

Running gaits Trot Piaffe, passage, RH-LF, susp, 2 Right/left 2.8–14.2 1.8–5.9 0.9–2.52 26–53% 0–9%
collected, working, LH-RF, susp diagonal
Gaits and interlimb coordination

medium, extended,
flying-trot
Pace RH-RF, susp, 2 Right/left 9.1–16.0 4.5–6.3 1.8–2.4 0.130–0.138 s 0.081–0.094 s
LH-LF, susp lateral
Canter Collected, medium, Trail.H, 3 Asymmetry 2.9–9 1.9–4.6 1.6–2.0 0.28–0.30 s 0–0.013 s
extended, disunited lead.H-trail.F, with a phase
lead.F, susp lag between
limb pair
Gallop Transverse, rotary Transverse: 4 Asymmetry 9–20 4.5–7.2 2.27–2.92 0.085–0.09 s 0.063–0.114 s
trail.H, lead.H, with a phase 16–28%
trail.F, lead.F, lag between
susp limb pair
 Classification and description of gaits

Attitude 1

0.35
0.35 1st time
1st time

Attitude 2 1.00 0 0.5 0 0.5 0 0.5 0 0.8

Attitude 3 0.75 0.25 0.5 0 0 0.5 0.8 0.5

0.80 Walk Trot Pace (Rack) Cantor

0 0.1 0 0.1 0 0 0 0

Attitude 4

0.5 0.6 0.6 0.5 0.5 0.5 0 0

0.35 Transverse Rotary Bound Pronk
0.35
gallop gallop
2nd time 2nd time
A C

Stride duration
Contralateral
advance placement

Stance duration Swing duration

RF LF RF

LH RH LH

Diagonal advance placement Suspension

B

Fig 5.1  Methods for representing the footfall sequences and temporal characteristics of the gaits. (A) Drawings of the footfall sequence observed at the
flying trot. (B) Gait and bar diagram of the trot. The bars represent the stance phase duration of each limb. (C) Drawings of the relative phases between the
limbs in various quadrupedal gaits.

the forelimbs trot. The model consists of four coupled (Barbeau & Rossignol, 1987). There are four distinct
oscillators that simulate the cyclical patterns of the rhythm generators (CPG) for the two hind limbs and for the
four limb movements. It is possible to generate all types two forelimbs (Forssberg et al., 1980a, b). Such a generator
of equine gaits using five ways of coupling the oscillators. has been identified in the lumbar spinal cord of the newborn
This type of functional model can be useful for rat (Cazalets et al., 1996). In horses, the characteristics of
understanding locomotor control by the central these rhythm generators should determine the stride
nervous system. Experimental results in neurophysiology frequency and its variability. Great stability is required in
demonstrated that the rhythmic activity of the skeletal dressage while rapid changes are nessary for jumping or
muscles of each limb comes from the central nervous system racing horses.

87
 5 Gaits and interlimb coordination

Fig 5.1  Continued (D) Pie gait diagram of the gallop in Quarter Horses showing the footfall sequence and the durations of the suspension phase, stance
phases and overlaps expressed in seconds and as a percentage of stride duration. The sequence rotates counterclockwise.
(A) Reproduced from Barroil (1887). (C) From Alexander, R. McN. The gait of bipedal and quadrupedal animals. Int. J. Robotics Res. 3, 49–59, Copyright ©1984, reprinted by permission of SAGE.
(D) Reproduced from Deuel & Lawrence (1987).

Walk lateral or diagonal couplets (Fig. 5.3) (Nicodemus & Clayton,

2003). Gaits with these characteristics are called toelt, paso, running
The walk is a four-beat gait with large overlap times between stance walk, rack, stepping pace or slow gait. These gaits are comfortable
phases of the limbs and no period of suspension. It is the slowest for the rider because the amplitude of the dorsoventral displace-
equine gait but probably one of the more complex gaits because of ment is lower than at the trot, which is a consequence of not having
the variability in the overlap and lag time between limbs. In a lame- a period of suspension. Furthermore vertical movements of the fore
ness examination, the variability in the regularity and symmetry of and hind limbs are out of phase, allowing a smoother gait (Bikn-
the stride measured at the walk was higher than at the trot (Barrey evicius et al., 2006). The speed ranges between 1.7–2.3 m/s for the
& Desbrosse, 1996). In dressage horses, the speed of the walk toelt and the natural gait transition sequence is walk-toelt-canter
increased from the collected walk (1.37 m/s) to the extended walk (Grasselli et al., 1991).
(1.82 m/s) with only a small increase in stride frequency (Clayton,
1995). The speed change was mainly the result of lengthening the
stride by increasing the over-tracking distance. Even in highly Trot
trained dressage horses, a regular four-beat rhythm of the footfalls
The trot is a two-beat, symmetric, diagonal gait (Fig. 5.4). The varia-
was observed in only one of the six horses.
tions of the trot of saddle horses are the collected, working, medium
For a breeding purposes, the calculation of heritabilities of the
and extended trots, with the speed of the gait increasing from col-
gait parameters are interesting in order to know if the measure-
lected to extended trot. A positive, hind first diagonal advanced
ments of these traits could be useful for genetic selection. Herita-
placement has been measured at the collected trot in elite dressage
bility (h2) of a quantitative trait, like a gait parameter, estimates
horses (Holmström et al., 1994; Clayton, 1997), with the hind limb
the genetic component vs. non-genetic components (breeding
contacting the ground about 20–30 ms before the diagonal fore-
management, nutrition, environment, training), which may influ-
limb. Passage and piaffe are diagonal exercises derived from col-
ence this quantitative trait. Heritability is expressed as a percent-
lected trot. From trot in hand to passage, the speed (−2.18 m/s) and
age of the genetic variance against the total variance including all
stride length (−1.18 m) are reduced while the stride duration
the other effects. Table 5.2 shows the heritabilities obtained in
(0.279 s) and diagonal advanced placement (9.7 ms) increase
French saddle horses for the walk in hand. Walk heritabilities
(Holmström et al., 1995). Dressage finalists at the Olympic Games
were rather low (mean heritability, h2 = 0.15) except for vertical
in Barcelona showed differences between the temporal variables of
activity, longitudinal propulsion and percentage of four-beat walk
the collected trot, passage and piaffe (Clayton, 1997). The stride
which were highly heritable (Barrey, 2004). The low heritabilities
duration is longer for piaffe (1.08 s) and passage (1.09 s) than for
could be explained because the walk is a complex four-beat gait
collected trot (0.84 s), which means that passage and piaffe have a
which can be slightly modified by many factors: rider actions,
lower stride frequency. For most of the other temporal variables,
environment.
collected trot and passage were similar to each other except that the
suspension phase was short in passage.
Heritabilities of the trot parameters have been calculated in
Other walking gaits French saddle horses using the same method as for the walk
Icelandic horses, Paso Finos, and certain other gaited breeds exhibit (Table 5.2). Most of the trot variables had a moderate to high
a four-beat symmetric gait in which the footfalls are coordinated as heritability (mean heritability, h2 = 0.24) (Barrey, 2004). The trot

88
 Classification and description of gaits

Hind limb stancephase (% of stride duration)

70 60 50 40

1. (31-0)

10
4. (36-6)

2. (57-9) 3. (43-10)
5. (29-12)

20

10. (32-22)
Lateral advanced placement (% of stride duration)

7. (66-22) 9. (46-22)
8. (54-24)

30 6. (73-25)

14. (30-29)
12. (54-31) 13. (40-31)

11. (66-33)

40

15. (59-38)

17. (46-44)
50

18. (36-50) 19. (23-??)

16. (54-50)

60
20. (33-57)
Fig 5.2  Classification of symmetrical gaits according to the temporal characteristics of the gait.
From Hildebrand, M., 1965. Symmetrical gaits of horses. Science 150 (3657), with permission from AAAS.

characteristics are very important for dressage ability and should France, the French saddle horses were mainly selected for jumping.
be used for early selection at 2 or 3 years old. Slow stride fre- However, a new breeding program for dressage was set up at the
quency, high dorsoventral activity and high propulsion accelera- end of the 20th century. It has been shown that specific conforma-
tion vector and longitudinal activity are the trot characteristics tion and gait characteristics in German breeds can explain their
required for performing in dressage. The heritabilities were higher higher ability for dressage competition (Barrey et al., 2002). For
at trot than at walk. dressage performance, the trot characteristics and variations are very
Because of the heritabilities of the gait parameters, it was assumed important because they form the basis of passage and piaffe which
that some of these traits could be genetically selected and contrib- should be performed at the top level. According to the FEI rules,
uted to dressage performance. Several breeds have been specifically the trot should be a two-beat gait with free, active and regular steps;
selected for dressage. In Spain, Andalusian horses have been bred the regularity and elasticity of the steps, and engaged hind quarters,
for dressage since the 15th century. The Andalusian horses were should be the main qualities of the trot. The same cadence and
used for military work at several royal riding academies in Spain, rhythm should be maintained during trot variations. The trot char-
Italy, France and Austria from the 15th to 18th centuries. According acteristics of the German horses showed good similarities with the
to the international competition results, several German breeds or FEI rules: a slow stride frequency, high regularity, large dorsoventral
crossed breeds (Hannoverian, Oldenburger, Westphalian and Dutch displacement and activity, which means elasticity and good propul-
Warmblood) are today the best performers. In these studbooks, the sion (see Table 5.3). Spanish horses have a shorter stride length, a
dressage ability was the main objective of genetic selection. In higher stride frequency and a lower dorsoventral displacement and

89
 5 Gaits and interlimb coordination

HL Table 5.2  Genetic components (heritabilities estimates h2 <1) of

FL the gait parameters measured in French saddle horses using an
HR acclerometric gait analysis device (Barrey, 2004)
FR
Gait variables h2 (SE) h2 (SE) h2 (SE)
A 0 50 100
for walk for trot for gallop
Stride characteristics
HL
Speed n/a 0.30 (0.14) n/a
FL
HR Stride length n/a 0.29 (0.13) n/a

FR Stride frequency n/a 0.20 (0.15) 0.32 (0.19)

B 0 50 100
Dorsoventral motion
Fig 5.3  Bar diagrams illustrating two types of footfall sequences of the toelt Symmetry 0.10 (0.08) 0.12 (0.08) n/a
in Icelandic horses: (A) diagonal couplets and (B) lateral couplets. HL, left
Regularity 0.10 (0.13) 0.12 (0.09) n/a
hind limb; HR, right hind limb; FL, left forelimb; FR, right forelimb.
(B) Grasselli, A., Grasselli, R., Iotti, P., 1991. Quantitative analysis of 4-beat leral gaits. In Proc. Displacement 0.16 (0.12) 0.14 (0.08) n/a
42nd Annual meeting of the EAAP, 8–12 September, Berlin, H4-2, pp. 558–559, reprinted with
permission. Dorsoventral activity 0.41 (0.12) 0.22 (0.10) 0.50 (0.13)
Gait tempo 0.29 (0.10) 0.05 (0.07) n/a

Longitudinal motion
Mean propulsion vector 0.19 (0.07) 0.20 (0.11) n/a
Propulsion duration 0.69 (0.13) 0.38 (0.15) n/a
Longitudinal activity n/a 0.44 (0.14) 0.46 (0.20)

h2, heritability; n/a, data not available.

Table 5.3  Changes in trot variables with the stage of training

Trot 4 years 5 years 6 years 7 years

parameters and older
Stride frequency 1.34a 1.32a 1.26b 1.27b
A (stride/s)
Stride regularity 186a 185a 185a 179b
Withers (/200)
Sacrum Dorsoventral 7.8 8.3 8.7 10.4
activity (W/kg)
Longitudinal 1.14 1.62 1.55 2.1
activity (W/kg)
Propulsion 2.3 2.2 1.4 3
LF
acceleration
RH vector (g)
RF
B LH Dorsoventral 0.11a 0.13b 0.135b 0.13b
Fig 5.4  Footfalls sequence recorded at the trot using a pneumatic gait displacement
recorder. (A) Horse equipped with pneumatic accelerometers attached to (m)
the limbs, saddle and tuber sacrale for measuring temporal gait parameters.
Values followed by different superscript letters are significantly different (p < 0.05).
The white spot indicates the suspension phase as shown on the figure.
(B) Time-related changes in the pressure obtained from the pneumatic
accelerometers at the trot (LF, left forelimb; RF, right forelimb; RH, right hind
limb; LH, left hind limb). The shaded bars indicate the stance phase
durations of each limb. The gaps between the shaded bars are the
suspension phases between the diagonal supports.
Reproduced from Marey (1873).

90
 Classification and description of gaits

Fig 5.5  Flying trot of a Standardbred.

Drevemo, S., Dalin, G., Fredricson, I., Hjerten, G., 1980. Equine locomotion 3: the reproducibility of gait in Standardbred trotters. Equine Vet. J. 12, 71–73, reused with permission from the Equine
Veterinary Journal.

activity than German horses. Spanish horses exhibited elevated flying trot. At racing speed, the pace, like the trot, becomes a four-
movements (large flexion of carpus and tarsus) rather than extended beat gait, with dissociation of the lateral limb pairs at impact and
movements of the limbs. Spanish and German groups have a high lift-off. The hind limb contacts the ground about 26–30 ms before
propulsion and longitudinal activity, which should be an advantage the lateral forelimb (Wilson et al., 1988a). In comparison with the
for collecting the trot to passage and piaffe. The German horses had flying trot, there is less problem of limb interference at the pace
gait characteristics more adapted for dressage competition and that because the lateral sequence avoids any contact between the ipsilat-
could already be measured in 3-year-old horses. eral limbs. Consequently, there are fewer coordination problems
In harness trotters, the trot is so extended that it can reach a and it is easier for the horse to increase stride length. These differ-
maximum speed of 14.2 m/s with a maximum stride frequency of ences may explain the higher speed records obtained by pacers
2.52 strides/s and a maximum stride length of 5.92 m (Barrey et al., 9.4–16.0 m/s (Wilson et al., 1988b) than by trotters 11.8–14.2 m/s
1995). The diagonal sequence usually changes to a four-beat rhythm (Barrey et al., 1995).
due to asynchrony of the impact (and lift-off) of the diagonal limb
pairs (Drevemo et al., 1980). This particular gait is named the flying
trot (Fig. 5.5). The hind limb touches the ground first (positive Canter and gallop
advanced placement), and the dissociation at lift-off is greater than Canter and gallop refer to the same gait performed at different
at impact. speeds: the canter is a slow speed, three-beat gait and the gallop
Various irregularities in the rhythm of the trot can occur during is a four-beat gait performed at a higher speed. At the canter, the
a harness race, which may result in the horse being disqualified by stance phases of the diagonal limb pair (leading hind and trailing
the gait judges. Irregular gait patterns that occur relatively frequently fore) are synchronized while at the gallop the footfalls of the
are called the aubin and the traquenard in French. At the ‘aubin’, diagonal are dissociated, with the leading hind limb contacting
the forelimbs gallop and the hind limbs trot, while, at the ‘traque- the ground before the trailing forelimb. The gallop is the fastest
nard’, the forelimbs trot and the hind limbs gallop. A trotter is also equine gait, and is the racing gait of Thoroughbreds and Quarter
disqualified for pacing or galloping. With increasing speed, the Horses. Canter parameters are highly heritable (mean heritability,
stride length increases linearly but interference between the hind h2 = 0.43) and could also be used for genetic selection (Table 5.2)
limb and the lateral forelimb becomes a limiting factor. A large (Barrey, 2004).
amount of overreaching by the hind limbs is possible only if the The canter and gallop show asymmetric movements of both the
hind limbs move outside (lateral to) the forelimbs during the swing hind and forelimbs. There are two possible footfall sequences: right
phase. Heritabilities of the trot parameters measured in harness lead canter or gallop and left lead canter or gallop. Horses at liberty
trotters (French totters) at high speed are moderate to high (stride prefer to canter or gallop through a turn with the inside limbs
frequency, 0.40 stride/s; stride length, 0.36 m; longitudinal activity, leading.
0.23 W/kg (Leleu, 2004) ). The heritability of the trot symmetry The ‘lead change’ is the transition between the footfall sequences
index is zero and the regularity of the stride in time is very low of the right and left leads. Racehorses usually change the forelimb
(0.09), which demonstrate that these coordination parameters are lead before the hind limb lead. However, in dressage the rider can
more trained than inherited. elicit the canter lead change during the suspension phase so the
The foxtrot is a four-beat symmetric gait in which the footfalls of change is initiated in the hind limbs (Clayton, 1994). In racing,
the diagonal limbs occur as couplets. The interval between footfalls horses change leads eight or more times per mile to avoid excessive
of the fore hoof and the diagonal hind hoof is 15% of stride, com- muscular fatigue due to the asymmetric work of the limbs and also
pared with 35% of stride between footfalls of the hind hoof and the to minimize the centrifugal forces as they accommodate to the curve
lateral fore hoof. During a complete stride, the overlap periods were (Leach et al., 1987).
tripedal with two hind limbs and one forelimb (8.9%), diagonal At the gallop, there are two ways of coordinating the hind or fore
bipedal (60.6%), tripedal with one hind limb and two forelimbs footfalls. These are called the ‘transverse gallop’ and the ‘rotary
(8.9%) and lateral bipedal (21.7%) (Clayton & Bradbury, 1995). gallop’ (Fig. 5.6). The transverse gallop is used more frequently by
horses than the rotary gallop but the rotary sequence is observed
temporarily during a lead change initiated by the forelimbs or when
Pace muscular fatigue occurs during racing. The disunited canter has the
This lateral symmetric gait is used in harness racing mainly in North same footfall sequence as the rotary gallop except that stance phase
America and Australia. The maximum speed is higher than at the of the lateral limbs is synchronized. It can be observed for one or

91
 5 Gaits and interlimb coordination

Rotary gallop

LH LF

RH RF

Transverse gallop

LH LF

RH RF
Fig 5.6  Differences of the footfall sequences of transverse and rotary gallops.
Reprinted from Leach, 1984. Stride characteristics of horses competing in Grand Prix jumping, with permission from the American Journal of Veterinary Research http://avmajournals.avma.
org/loi/ajvr.

more strides after a bad lead change in dressage or landing after The footfall sequence of various gait transitions has been
jumping. described by Marey (1873), Barroil (1887) and Lenoble du Teil
(1893) (Fig. 5.7). Kinematic studies have described alternative foot-
fall sequences observed in dressage horses during transitions
Jump between the walk and trot transition (Argue & Clayton, 1993a) and
during the transitions between trot and canter (Argue & Clayton,
The jump is a gallop stride in which the airborne phase is a long
1993b).
dissociation of the diagonal. The footfalls of the jump stride
The combination of gait analysis by accelerometry and wavelet
are: trailing hind and leading hind at lift-off, jump suspension, then
analysis allows quantitative description of some temporal and
trailing fore and leading fore at landing. At lift-off, the hind limb
kinetic characteristics of gait transitions (Biau et al., 2002). Transi-
stance phases are more synchronized than in a normal gallop stride
tion duration, dorsoventral activity and frequency were specific for
to produce a powerful push-off. The footfalls of the forelimbs at
each transition. Training improved smoothness of braking decelera-
landing are not synchronized (Leach et al., 1984b). A lead change
tion and frequency changes with a long transition duration. The
can take place during the airborne phase and, in this case, the
walk-halt transition was characterized by the largest transition dura-
change of forelimb placement order occurs before that of the hind
tion, the lowest dorsoventral activity, and a low braking decelera-
limbs. A disunited canter can be observed after the jump if the lead
tion. The change of the vertical activity during this transition was
change of the hind limbs does not occur immediately after the
smooth but the decrease of stride frequency was sudden. The trot-
landing phase.
walk transition was characterized by a short duration with a large
braking deceleration and a great change of vertical activity. An
increase of transition duration allowed a smooth transition with
Gait transitions lower braking deceleration. Canter-halt, canter-walk and canter-trot
transitions were characterized by a great dorsoventral activity and a
In order to increase its velocity, the horse can switch gaits from walk middle braking deceleration. It was explained by progressive
to trot, from trot to canter and then extend the canter into a gallop. decrease of dorsoventral activity before the gait change, which mini-
Each gait can be extended by changing the spatial and temporal mized the braking deceleration at the end of the canter, especially
characteristics of its strides. Ponies were shown to have a preferred for experienced horses. The transition duration increased signifi-
speed for the trot to gallop transition and this particular speed was cantly with training for trot-walk, canter-halt and canter-trot transi-
related to an optimal metabolic cost of running (Hoyt & Taylor, tions. The lengthening of the transition duration allowed a slow
1981). However, another experiment demonstrated that the trot- decrease of stride frequency (canter-trot transition) and a smooth
gallop transition was triggered when the peak of ground reaction decrease of vertical activity (canter-halt and trot-halt transition).
force reached a critical level of about 1–1.25 times the body weight The rider adapted his technique to the locomotion and level of the
(Farley & Taylor, 1991). Carrying additional weight reduced the horse. By lengthening transition duration, experienced horses could
speed of the trot-gallop transition. perform a smooth deceleration. In contrast, young horses could not

92
 Velocity-related changes in stride variables

Position 1 4

0.30
0.90

Stride length (m)

0.60

3
0.15
Position 2

2
0.30 0.30

0
1 2 3 4 5 6 7 8 9 10 11 12
Position 3
Velocity (m/s)
Fig 5.8  Linear relationship between the stride length and velocity of the
gaits. Data are from 6-month-old Quarter Horse foals.
Reprinted from Leach, 1984. Stride characteristics of horses competing in Grand Prix
jumping, with permission from the American Journal of Veterinary Research http://
0.20 avmajournals.avma.org/loi/ajvr.

Position 4
3

0.90
0.30

2
Frequency (strides /s)

Position 5

1st time
1
C–G
1.00
Position 6 T
W
1.50
0
1.50 1.50 0 1 2 3 4 5 6 7 8 9 10 11 12
2nd time 2nd time Velocity (m/s)
Fig 5.7  Example of limb placement sequence during a transition from walk Fig 5.9  Non-linear relationship between the stride frequency of the gaits.
to canter.
Reprinted from Leach, 1984. Stride characteristics of horses competing in Grand Prix
Reproduced from Barroil (1887).
jumping, with permission from the American Journal of Veterinary Research http://
avmajournals.avma.org/loi/ajvr.

prepare their gait transition and would suddenly brake which pro- order to repeat the limb movements more frequently. The stride
duced a high peak of deceleration. The amplitude of frequency frequency (SF) and stride length (SL) are the two main components
change during the gait transition decreased with training for all of speed. The mean speed can be estimated by the product of the
transitions, especially for canter-halt and trot-walk transition. stride frequency and stride length: speed = SF × SL. The speed-
related changes in stride parameters have been studied in many
horse breeds and disciplines. Stride length increases linearly with
the speed of the gait (Fig. 5.8).
Velocity-related changes in stride variables Stride frequency increases non linearly and more slowly (Dusek
et al., 1970; Leach & Cymbaluk, 1986; Ishii et al., 1989) (Fig. 5.9).
To increase speed at a particular gait, the amplitude of the steps During rapid acceleration, such as that occurring at the start of a
becomes larger and the duration of the limb cycle is reduced in gallop race, the stride frequency reaches its maximum value very

93
 5 Gaits and interlimb coordination

Speed Stride length Stride frequency there are some objective indicators of good trot. A slow stride fre-
20 quency with a long swing phase, a large amplitude of scapular
rotation, maximal forelimb retraction and maximal hind limb pro-
traction lead to a long stride length with a good trot. The vertical
8.0

Stride length (m)

elasticity of the trot was associated with maximal fetlock extension,
15
and maximal stifle and tarsal flexion (Back et al., 1994b).

Stride frequency (f/s)

7.0
In racehorses, the influence of training has been investigated in
6.0
Velocity(m/s)

Standardbreds and Thoroughbreds. After 3 years of training, Stan-

10 5.0 dardbreds showed increases in stride length, stride duration and
swing phase duration (Drevemo et al., 1980). In Thoroughbreds,
4.0 stride duration and stride length increased (Leach & Springings,
3.0
3.0 1979). After 8 weeks of high intensity training on a treadmill, the
5 2.5 stance phase duration of the Thoroughbred gallop stride was
reduced by 8–20% (Corley & Goodship, 1994).
2.0 In young saddle horses (2.5 years old), some kinematic changes
were observed on a treadmill at trot after a 70-day training period
0 for dressage and jumping (Back et al., 1995). The protraction and
0 10 20 30 retraction range of the forelimbs decreased and the stance duration
Strides and the flexion of the hind limbs decreased. The engagement
increased with a maximal protraction of the hind limbs that took
Fig 5.10  Changes in velocity, stride frequency and stride length at the start
place earlier in the stride cycle. The stride frequency stayed constant.
dash in a galloping horse.
During the same 70-day period, the stride frequency in the control
Reprinted from Higara, A., Yamanobe, A., Kubo, K., 1994. Relationships between stride length,
group (in pasture day and night) was reduced with a longer swing
stride frequency, step length and velocity at the start dash in a racehorse, J. Equine Sci. 5,
phase. The trained horses were trotting with impulsion ‘on the bit’
127–130, with permission of the Japanese Society of Equine Science.
with the same stride frequency while the pastured horses were trot-
ting in a more relaxed way with a lower stride frequency.
Dressage requires a high level of locomotor control by the rider,
rapidly to produce the initial acceleration, while the maximum which is obtained progressively through exercise and collecting the
stride length increases more slowly to its maximum value (Hiraga gaits. A horse’s ability for collection seems to be one of the main
et al., 1994) (Fig. 5.10). limiting factors for dressage because it is impossible to execute the
In Thoroughbred racehorses, the effects of fatigue include more complex exercises correctly without having attained a good
increases in the overlap time between the leading hind limb and collected gait at trot and canter. Some locomotor parameters were
the trailing forelimb, the stride duration and the suspension phase identified as favoring collection ability, extended gaits and the
duration (Leach & Springings, 1979). The compliance of the track expressiveness of the gait (Holmström et al., 1994b; Back et al.,
surface can also influence the stride parameters when the horse is 1994b). A slow stride frequency including a long swing phase is
trotting or galloping at high speed. At the gallop, the stride duration required for good trot quality. The elapsed time between the hind
tends to be reduced on a harder track surface (Fredricson et al., limb contact and the diagonal forelimb contact defines the diagonal
1983). There is a slight increase in the stride duration on a wood- advanced placement and should be positive and high at the trot.
fiber track in comparison with a turf track at the same speed. The diagonal hind limb should touch the ground about 30 ms
When the rider stimulated the horse with a whip, there was a before the diagonal forelimb. In space, a good engagement is
reduction in the stride length and an increase in the stride frequency required; the hind limb footfall should be placed as far as possible
corresponding to a reduction of the forelimb stance phase duration. under the body for a good propulsion activity.
However, the velocity was not significantly influenced (Deuel & Dressage horses should be trained to improve their coordination,
Lawrence, 1987b). their suppleness and their gait collection. The collection of the gait
means that the forward movements become more upward move-
ments and the stride frequency decreases. A group of horses compet-
Gait development and training effects ing in dressage has been tested from 4 to 7 years old to determine
the training-related changes in locomotion variables (Barrey & Biau,
Gait patterns are influenced by the age of the horse, but little is 2002). The changes of the trot (Table 5.3) were more pronounced
known about gait development. Studies in different breeds of foals than the changes in the walk and canter. At the trot, the stride fre-
have analyzed the relationship between conformation and stride quency decreases between 5 and 6 years old. In the same time the
variables in foals aged 6–8 months. In Quarter Horse foals, speed walk and gallop stride frequencies stay the same. At the trot, the
increases were obtained by a longer stride length in heavier foals dorsoventral displacement increases after the 1st year of training.
and a higher stride frequency in taller foals (Leach & Cymbaluk, The increase of dorsoventral activity with age corresponded to an
1986). In Dutch Warmblood foals, the elbow, carpal and fetlock increase of muscular power and an increase of the dorsoventral
joint angle flexions were the most significant differences between displacement in the collected gaits. The stride regularity and sym-
the stride kinematics of individual foals (Back et al., 1993). The metry increase initially, but then decrease after 6 years old.
stride and stance durations increased with age as a consequence of
the increase in height but the swing duration and the protraction
and retraction angles were consistent over time. The joint angle Influence of the treadmill on
patterns recorded at 4 months and 26 months of age were very
similar. The duration of the trot swing phase, the maximal range of gait characteristics
protaction-retraction of the limbs and the maximal flexion of hock
joint were well correlated between 4 and 26 months of age (Back Under laboratory conditions it is possible to study the locomotion
et al., 1994a). The good correlations between the kinematic param- of horses running on an experimental track or on a treadmill. The
eters measured in the foals and adults make it possible to assess latter provides an excellent means of controlling the regularity of
these parameters in young horses in order to predict the gait quality the gaits because the speed and slope of the treadmill belt are
of adult horses. According to the correlations with the judged scores, determined by the operator. In order to analyze the gaits without

94
 Locomotion and respiratory coupling

(1) S F = 2.11 (1-0.75v) r = 0.95** Inspiration Expiration

2.0

Stride frequency (St./5) 1.9 (3) S F = 1.93 (1-0.69v) r = 0.93**

1.8 (2) S F = 1.91 (1-0.70v) r = 0.93** Left hind

Left fore and

1.7 right fore
400 450 500 550 600 650 ** P < 0.01
A Velocity (m/min) Right fore

(2) S L = 0.44V + 0.85 r = 0.99**

5.5 (3) S L = 0.44V + 0.96 r = 0.98**

5.0
Stride length (m)

0.1 seconds
4.5 A Time
(1) S L = 0.38V + 1.15 r = 0.99**
4.0

3.5 Inspiration Expiration

B 400 450 500 550 600 650 ** P < 0.01
Velocity (m/min)

Track (1) Right

Treadmill 0% (2) hind
Treadmill 3.5% (3)
Left
Fig 5.11  Relationship of velocity with (A) stride frequency and (B) stride hind
length for locomotion on a track, a flat treadmill and a treadmill with a 3.5%
inclination. Right
fore
Reprinted from Barrey, E., Galloux, P., Valette J.P., Auvinet, B., Wolter, R., 1993a. Stride
characteristics of overground versus treadmill locomotion in the saddle horse. Acta Left
Anatomica 146, 90–94, with permission from S Karger AG. fore

stress, some pre-experimental exercise sessions are required to

accustom the horse to this unusual exercise condition (Buchner 0.1 seconds
et al., 1994). The horse adapts rapidly at trot, and stride measure- B Time
ments can be undertaken at the beginning of the third session. For
the walk, however, adaptation occurs more slowly and many stride Fig 5.12  Schematic diagrams showing the relationship between the limb
parameters are not stable even after the nineth training session. cycle and the respiratory cycle at (A) the canter and (B) gallop.
Within a session, a minimum of 5 min of walking or trotting is Reprinted from Attenburrow, D.P., 1982. Time relationship between the respiratory cycle and
required to reach a steady state of locomotion. limb cycle in the horse. Equine Vet. J. 14, 69–72, with permission from the Equine Veterinary
Many locomotion studies have been performed on high-speed Journal.
treadmills, since the development of the first installation of this type
of machine at the Swedish University of Agricultural Science in
Uppsala (Fredricson et al., 1983). However, it was demonstrated reduced in the forelimbs and increased in the hind limbs on
experimentally that the stride length was longer on the treadmill at the incline, which indicated that the hind limbs generate more
the trot and canter than at the same speed on a track (Barrey et al., propulsion on the inclined than on the flat treadmill. The inclina-
1993a; Couroucé et al., 1998) (Fig. 5.11). tion of the treadmill did not change stride length nor did it change
The mechanical reasons for these differences are not entirely the stance, swing or stride duration in a cantering Thoroughbred
known, but some explanations have been suggested by the experi- (Kai et al., 1997).
mental and theoretical results. The speed of the treadmill belt fluc-
tuates in relationship to the hoof impact on the belt (Savelberg
et al., 1994). The energy transfer between the hooves and the tread-
mill belt are not exactly the same as in overground locomotion Locomotion and respiratory coupling
because the belt and the hooves are driven backwards by the engine
of the treadmill. During level treadmill exercise, the horse receives Some relationships have been established between stride parame-
some mechanical energy from the treadmill. This assumption is ters and other physiological variables. At the canter and gallop, the
based on the fact that the heart rate response and blood lactate respiratory and limb cycle are synchronized. Inspiration starts at the
concentration, which reflect the horse’s workload, were lower on a beginning of the suspension phase and ends when the trailing
treadmill than overground during a standardized exercise test forelimb contacts the ground. Expiration then occurs during the
(Valette et al., 1992; Barrey et al., 1993b). At a slow trot, a treadmill forelimb stance phases (Attenburrow, 1982) (Fig. 5.12). Expiration
inclination of 6% tended to increase the stride duration and signifi- is facilitated by compression of the rib cage between the weight-
cantly increased the stance duration, more so in the hind limbs than bearing forelimbs. This functional coupling of respiration to the
the forelimbs (Sloet et al., 1997). Maximal fetlock extension was stride cycle might be a limiting factor for ventilation at maximal

95
 5 Gaits and interlimb coordination

exercise intensity. At the gallop racing speed, less than 50% of the muscular fatigue onset may occur quickly at high speed because of
stride duration is available for the expiration. At 16 m/s, the expira- this decrease in expiration duration, which is mechanically linked
tion during the trailing and leading forelimb support represents to the stride cycle both at the trot and gallop.
45% of the stride duration (Barrey et al., 2000). At high speed on At the walk, trot and pace there is no consistent coupling between
a treadmill (11 m/s), the respiratory and stride cycles are exactly the locomotor and respiratory cycles. At trot, the ratio between
synchronized with one inspiration/expiration per stride in Stan- locomotor and respiratory frequency ranged between 1–3 depend-
dardbreds. The inspiration takes place during the support phase of ing on speed, duration of exercise and breed of horse (Hörnicke
one diagonal and most of the preceding and following suspension et al., 1987; Art et al., 1990). A similarly variable coupling mecha-
phases. About 57% of the stride duration was available for inspira- nism was observed at the pace where the ratio between the stride
tion and only 43% for expiration. Hypercapnia, acidosis and and respiratory frequencies ranged from 1–1.5 (Evans et al., 1994).

References
Alexander, R.McN., 1984. The gait of bipedal three-day-event horse, conference on Clayton, H.M., 1994. Comparison of
and quadrupedal animals. Int. J. Robotics equine sports medecine and science. Arno the collected, working, medium and
Res. 3, 49–59. Lindner Ed, pp 17–32. extended canters. Equine Vet. J. Suppl.
Argue, C.K., Clayton, H.M., 1993a. A Barrey, E., Couroucé, A., Orsetti, H., Evans, D., 17, 16–19.
preliminary study of transitions between Robert, C., Rose, J.R., 2000. Couplage de la Clayton, H.M., 1995. Comparison of the stride
the walk and trot in dressage horses. Acta ventilation respiratoire avec la locomotion kinematics of the collected, medium and
Anat. 146, 179–182. du cheval de course. Entretiens de Laval, extended walks in horses. Am. J. Vet. Res.
Argue, C.K., Clayton, H.M., 1993b. A study of Equathlon 30, 32–35. 56, 849–852.
transitions between the trot and canter in Barrey, E., Desbrosse, F., 1996. Lameness Clayton, H.M., Bradbury, J.W., 1995. Temporal
dressage horses. J. Equine Vet. Sci. 13, detection using an accelerometric device. characteristics of the foxtrot: a symmetrical
171–174. Pferdeheilkunde 12, 617–622. equine gait. Appl. Anim. Behav. Sci. 42,
Art, T., Desmecht, D., Amory, H., Lekeux, P., Barrey, E., Desliens, F., Blouin, C., Langlouis, 153–159.
1990. Synchronization of locomotion and B., 2002. Heritabilities of gait Collins, J.J., Stewart, I.N., 1993. Coupled
repiration in trotting ponies. J. Vet. Med. A characteristics: application for breeding in nonlinear oscillators and the symmetries
37, 95–103. dressage. The Elite dressage and three-day- of animal gaits. J. Nonlinear Sci. 3,
Attenburrow, D.P., 1982. Time relationship event horse, conference on equine sports 349–392.
between the respiratory cycle and limb medecine and science. Arno Lindner Ed, pp Corley, J.M., Goodship, A.E., 1994. Treadmill
cycle in the horse. Equine Vet. J. 14, 69–72. 91–95. training induced changes to some
Back, W., Barneveld, A., Bruin, G., Schamhardt, Barrey, E., Desliens, F., Poirel, D., Biau, S., kinematic variables measured at the canter
H.C., Hartman, W., 1994b. Kinematic LeMaire, S., Rivero, J.L.L., et al., 2002. Early in thoroughbred fillies. Equine Vet. J.
detection of superior gait quality in young evaluation of dressage ability in different Suppl. 17, 20–24.
trotting warmbloods. Vet. Q. 16, S91–S96. breeds. Equine Vet. J. Suppl. 34, 319–324. Couroucé, A., Geoffroy, O., Barrey, E., Rose,
Back, W., Barneveld, A., Schamhardt, H.C., Barrey, E., Galloux, P., Valette, J.P., Auvinet, B., R.J., 1998. Comparison of exercise tests on
Bruin, G., Hartman, W., 1994a. Wolter, R., 1993a. Stride characteristics of different tracks and on an uninclined
Longitudinal development of the overground versus treadmill locomotion in treadmill in French trotters. Equine Vet. J.
kinematics of 4-, 10-, 18- and 26-month- the saddle horse. Acta Anatomica 146, Suppl. 29, (in press).
old Dutch warmblood horses. Equine Vet. 90–94. Deuel, N.R., Lawrence, L.M., 1984. Computer-
J. Suppl. 17, 3–6. Barroil, E., 1887. L’art équestre de haute école drawn gait diagrams. J. Equine Vet. Sci. 4,
Back, W., Hartman, W., Schamhardt, H.C., d’équitation: traité pratique. J. Rothschild 228–229.
Bruin, G., Barneveld, A., 1995. Kinematic Ed., fig. 31. Deuel, N.R., Lawrence, L.M., 1987. Effect of
response to a 70 day training period in Biau, S., Lemaire, S., Barrey, E., 2002. Analysis urging by the rider on equine gallop stride
trotting Dutch warmbloods. Equine Vet. J. of gait transitions in dressage horses using limb contacts. Proc. 10th Equine Nutrition
Suppl. 18, 127–131. wavelet analysis of dorsoventral and Physiology Symposium, pp 487–492.
Back, W., van den Bogert, A.J., van Weeren, acceleration. Pferdeheilkunde 4, 343–350. Drevemo, S., Dalin, G., Fredricson, I., Hjerten,
P.R., Bruin, G., Barneveld, A., 1993. Biknevicius, A.R., Mullineaux, D.R., Clayton, G., 1980. Equine locomotion 3: the
Quantification of the locomotion of Dutch H.M., 2006. Locomotor mechanics of the reproducibility of gait in Standardbred
warmblood foals. Acta Anatomica 146, tölt in Icelandic horses: taking the walk for trotters. Equine Vet. J. 12, 71–73.
141–147. a run. Am. J. Vet. Res. 67, 1505–1510. Dusek, J., Ehrlein, H.J., von Engelhardt, W.,
Barebeau, H., Rossignol, S., 1987. Buchner, H.H.F., Savelberg, H.C.M., Hornicke, H., 1970. Beziehungen zwischen
Recovery of locomotion after chronic Schamhardt, H.C., Merkens, H.W., trittlange, trittfrequenz und geschwindigkeit
spinalisation in the adult cat. Brain Res. Barneveld, A., 1994. Habituation of horses bei pferden. Zeitschrift fur Tierzuechtung
412, 84–95. to treadmill locomotion. Equine Vet. J. und Zuechtungsbiologie 87, 177–188.
Barrey, E., 2004. Influence of genetics in Suppl. 17, 13–15. Evans, D.L., Silverman, E.B., Hodgson, D.R.,
equine exercise physiology, 9th annual Cazalets, J.R., Borde, M., Clarac, F., 1996. The Eaton, M.D., Rose, R.J., 1994. Gait and
congress of european college of sport synaptic drive from the spinal locomotor respiration in Standardbred horses when
science. Clermont-Ferrand, France, July network to motoneurons in the newborn pacing and galloping. Res. Vet. Sci. 57,
3–6, Abstracts book, p. 119. rat. J. Neuroscience 16, 298–306. 233–239.
Barrey, E., Auvinet, B., Couroucé, A., 1995. Clayton, H., 1997. Classification of collected Farley, C.T., Taylor, C.R., 1991. A mechanical
Gait evaluation of race trotters using an trot, passage and piaffe based on temporal trigger for the trot-gallop transition in
accelerometric device. Equine Vet. J. Suppl. variables. Equine Vet. J. Suppl. 23, 54–57. horses. Sciences 25, 306–308.
18, 156–160. Clayton, H.M., 1989. Terminology for the Forssberg, H., Grillner, S., Halbertsma, J.,
Barrey, E., Biau, S., 2002. Locomotion of description of equine jumping kinematics. Rossignol, D., 1980a. The locomotion of
dressage horses. The Elite dressage and J. Equine Vet. Sci. 9, 341–348. low spinal cat. 1. Coordination within a

96
 References

hind limb. Acta Physiol. Scand. 108, Equine exercise physiology. Granta Leleu, C., 2004. Analyse des facteurs
269–281. Editions, Cambridge, pp. 7–16. locomoteurs de performance chez le cheval
Forssberg, H., Grillner, S., Halbertsma, J., Hoyt, D.F., Taylor, C.R., 1981. Gait and trotteur. PhD thesis, Rennes University, pp
Rossignol, D., 1980b. The locomotion energetics of locomotion in horses. Nature 188.
of low spinal cat. 2. Interlimb 292, 239. Lenoble du, Teil, 1893. Les allures du cheval
coordination. Acta Physiol. Scand. Ishii, K., Amano, K., Sakuraoka, H., 1989. dévoilées par la méthode expérimentale.
108, 282–295. Kinetics analysis of horse gait. Bull. Equine Berger Levrault, Paris and Nancy, pp.
Fredricson, I., Drevemo, S., Dalin, G., Hjerten, Res. Ins. 26, 1–9. 192–211.
G., Björne, K., Rynde, R., et al., 1983. Kai, M., Hiraga, A., Kubo, K., Tokuriki, M., Marey, E.J., 1873. La machine animale:
Treadmill for equine locomotion analysis. 1997. Comparison of stride characteristics locomotion terrestre et aérienne. In: Marey,
Equine Vet. J. 15, 111–115. in a cantering horse on a flat and inclined E.J. (Ed.), Coll. bibliothèque science
Fredricson, I., Hellander, J., Hjertén, J., treadmill. Equine Vet. J. Suppl. 23, 76–79. internationale, second ed. Librairie Gerner
Drevemo, S., Björne, K., Dalin, G., et al., Leach, D.H., 1987. Locomotion of the athletic Baillere et Cie, Paris, pp. 145–186.
1983. Galloppaktion II – Basala horse. In:Gillespie, J.R., Robinson, N.E. Nicodemus, M.C., Clayton, H.M., 2003.
gangartsvariabler i relation till banunderlag. (Eds.), Equine exercise physiology 2. ICEEP Temporal variables of four-beat, stepping
Svensk Veterinärtidning 35 (Suppl. 3), publications, Davis, pp. 516–535. gaits of gaited horses. Appl. Ani. Behav. Sci.
83–88. Leach, D.H., 1993. Recomended terminology 80, 133–142.
Grasselli, A., Grasselli, R., Iotti, P., 1991. for researchers in locomotion and Savelberg, H.C.M., Vostenbosch, M.A.T.M.,
Quantitative analysis of 4-beat leral gaits. biomechanics of quadrupedal animals. Acta Kamman, E.H., van de Weijer, J.,
In Proc. 42nd annual meeting of the EAAP, Anatomica 146, 130–136. Schamhardt, H.C., 1994. The effect of
8–12 September, Berlin, H4–2, pp Leach, D.H., Cymbaluk, N.F., 1986. intra-stride speed variation on treadmill
558–559. Relationship between stride length, stride locomotion. Proc. 2nd world congress of
Hildebrand, M., 1965. Symmetrical gaits of frequency, velocity and morphometrics of biomechanics, amsterdam.
horses. Science 191, 701–708. foals. Am. J. Vet. Res. 47, 2090–2097. Sloet van Oldruitenborgh-Oosterbaan, M.M.,
Hiraga, A., Yamanobe, A., Kubo, K., 1994. Leach, D.H., Ormrod, K., Clayton, H.M., Barneveld, A., Schamhardt, H.C., 1997.
Relationships between stride length, stride 1984a. Standardised terminology for Effect of treadmill inclination on
frequency, step length and velocity at the the description and analysis of equine kinematics of the trot in Dutch warmblood
start dash in a racehorse. J. Equine Sci. 5, locomotion. Equine Vet. J. 16, horses. Equine Vet. J. Suppl. 23, 71–75.
127–130. 522–528. Valette, J.P., Barrey, E., Auvinet, B., Galloux, P.,
Holmström, M., Fredricson, I., Drevemo, S., Leach, D.H., Ormrod, K., Clayton, H.M., Wolter, R., 1992. Comparison of track and
1994. Biokinematic differences between 1984b. Stride characteristics of horses treadmill exercise tests in saddle horses: a
riding horses judged as good and porr at competing in grand prix jumping. Am. J. preliminary report. Annale de Zootechnie
the trot. Equine Vet. J. Suppl. 17, 51–56. Vet. Res. 45, 888–892. 41, 129–135.
Holmström, M., Fredricson, I., Drevemo, S., Leach, D.H., Springings, E.J., 1979. Gait Wilson, B.D., Neal, R.J., Howard, A.,
1995. Variation in angular pattern fatigue in the racing thoroughbred. J. Groenendyk, S., 1988a. The gait of pacers:
adaptation from trot in hand to passage Equine Med. Surg. 3, 436–443. I. Kinematics of the racing stride. Equine
and piaffe in the grand prix dressage horse. Leach, D.H., Springings, E.J., Laverty, W.H., Vet. J. 20, 341–346.
Equine Vet. J. Suppl. 17, 51–56. 1987. Multivariate statistical analysis of Wilson, B.D., Neal, R.J., Howard, A.,
Hörnicke, H., Meixner, R., Pollmam, U., 1987. stride-timing measurements of nonfatigued Groenendyk, S., 1988b. The gait of pacers:
Respiration in exercising horses. In: Snow, racing thoroughbreds. Am. J. Vet. Res. 48, II. Kinematics of the racing stride. Equine
D.H., Persson, S.G.B., Rose, R.J. (Eds.), 880–888. Vet. J. 20, 347–351.

97
This page intentionally left blank
 C H A PTER 6 

Forelimb function
Hilary M. Clayton, Henry Chateau, Willem Back

Terminology of the joint moment and that joint’s angular velocity, measures the
rate of mechanical energy generation and absorption across a joint.
Discrete bursts of positive and negative work can be quantified as
This chapter reviews the structure and functions of the equine fore- the areas under the positive and negative phases, respectively, of the
limbs in relation to locomotor activity, including kinematics (move- power curve. Positive work is done when the net joint moment acts
ments) and kinetics (forces) during the stride. A stride is regarded in the same direction as the angular velocity of the joint, indicating
as the unit of measurement. The stride starts and ends with consecu- that the muscle shortens as it generates tension (concentric contrac-
tive occurrences of the same event, which is often a specific footfall. tion). Negative work is done when the net joint moment acts in the
Within a stride, each limb has a stance phase when the hoof is in opposite direction to the angular velocity of the joint, so the muscle
contact with the ground, and a swing phase when the hoof is swing- lengthens as it generates tension (eccentric contraction) and acts to
ing through the air. The stance phase starts at the moment of initial restrain joint movement in opposition to gravity or some other
ground contact, after which the hoof is decelerated during the external force.
impact phase. In the middle part of the stance phase the limb is This chapter describes the structure of the forelimb musculature,
loaded by the horse’s body weight then unloaded. Breakover is the the movements of the forelimb and the role of specific muscle
terminal part of stance when the heels leave the ground and rotate groups in causing and controlling those movements. These concepts
around the toe, which is still in ground contact. The various gaits are important in understanding how the muscles of the forelimb
are defined by the sequence and timing of the limb movements work to cause or control segmental and joint motion. The role of
during the stride. the forelimb joints and musculature as determined by inverse
Forelimb kinematics are described in terms of temporal (timing), dynamics analysis will be described later in this chapter.
linear (distance) and angular variables. In order to interpret data
describing segment and joint angles, it is important to know where
the angles were measured. For example, joint angles may be mea-
sured between the proximal and distal segments on the anatomical Musculotendinous architecture
flexor aspect or as the angle by which the distal segment deviates
from alignment with the proximal segment, or some combination In small, non-cursorial mammals, the forelimb is attached to the
of these methods (Fig. 6.1). Furthermore, the angle may be expressed trunk via a shoulder girdle, in which the clavicle articulates with
in absolute terms or it may be normalized to the standing angle, the sternum and scapula, imposing some constraints on forelimb
the angle at ground contact or the average angle during the stride motion. Horses do not have a clavicle or shoulder girdle. Instead,
(Mullineaux et al., 2004). the articulation between the forelimb and the trunk is a synsarco-
The ground reaction force (GRF) vector is usually resolved into sis, consisting of a substantial group of extrinsic muscles and their
vertical, longitudinal and transverse components to facilitate inter- associated soft tissues. The absence of a clavicle allows the scapula
pretation of its effects. Due to their proximity to the horse’s center more freedom to rotate and translate relative to the ribcage,
of mass, the forelimbs carry more weight (57–58%) than the hind which may contribute to an increase in stride length. As a conse-
limbs (42–43%) and have proportionately higher vertical forces quence of this translational motion, the instantaneous center of
and impulses. This supportive function is reflected in the pillar-like rotation of the scapulothoracic joint changes throughout the
alignment of the antebrachial and metacarpal segments. stride.
GRF data can be combined with kinematic data using a link Locomotor muscles account for about 42% of the horse’s body
segment model to calculate internal forces within the limb that mass (Gunn, 1978) with the large, powerful muscles concentrated
cannot be measured directly (see Chapter 19 for details). Briefly, in in the proximal limb, while the distal forelimb makes use of long,
a two-dimensional link segment model, each segment is repre- elastic tendons to reduce the metabolic cost of locomotion. The
sented as a solid bar and the location of its center of mass is known bulk of the musculature is in the proximal limb, which reduces the
relative to the coordinates that define the segment. The input for moment of inertia of the limb as a whole. The functions of
the model comprises kinematic and force data that are synchro- the musculotendinous system of the equine forelimb include con-
nized in time and space, together with segment morphometric data necting the forelimb to the trunk; supporting the body mass; stabi-
(Fig. 6.2, Table 6.1). An inverse dynamic solution is used to compute lizing the joints in opposition to the force of gravity during the
net joint moments and net joint powers (Colborne et al., 1997a,b). stance phase; generating forces that are used for propulsion, braking
The net joint moment represents the net torque acting around a and turning; and flexing the joints to lift the hoof clear of the
joint, which is produced primarily by the soft tissues (muscle, ground during the swing phase. At first glance, there appears to be
tendon and ligament). Net joint power, calculated as the product considerable redundancy in muscular function but muscles with

99
 6 Forelimb function

4
5

6
7
A B

Fig 6.1  Two methods of measuring joint angles of the forelimbs with the Fig 6.2  Skin markers used to locate the centers of mass of the forelimb
measured angles being represented by black arcs. Left: measurement of the segments in Table 6.1, which are separated according to the incision lines
angle between the proximal and distal segments on the anatomical flexor shown in red.
aspect. Right: measurement of the angle by which the distal segment differs Reprinted from Buchner, H.H.F., Savelberg, H.H.C.M., Scharmhardt, H.C., Barneveld, A., 1997,
from alignment with the proximal segment; deviation toward the flexor Inertial properties of Dutch Warmblood horses, Journal of Biomechanics, 30 (6), 653–658,
aspect is negative (−), deviation toward the extensor aspect is positive (+). with permission from Elsevier.

Table 6.1  Forelimb segmental masses, densities, reference lines for division of segments (see Fig. 6.2 for key to marker locations), and
position of segmental centers of mass in the sagittal plane*

Segment Mass (kg) Density Proximal and Distance as % segment length

(g/cm3) distal markers
x-axis y-axis
Scapula 11.5 ± 0.9 1.04 ± 0.01 1–2 27 ± 4 −12 ± 5
Brachium 8.6 ± 1.5 1.05 ± 0.01 2–3 51 ± 8 −5 ± 3
Antebrachium 6.7 ± 0.6 1.12 ± 0.03 3–4 35 ± 3 −2 ± 2
Metacarpus 1.6 ± 0.1 1.29 ± 0.04 4–5 44 ± 4 −1.0 ± 0.3
Fore pastern 0.73 ± 0.03 1.25 ± 0.03 5–6 46 ± 4 −11 ± 4
Fore hoof 1.08 ± 0.01 1.18 ± 0.01 6–7 29 ± 6 −20 ± 4.0

*Position of center of mass is expressed along the x-axis (longitudinal, positive distally from the more proximal marker), then along the y-axis (perpendicular to x-axis, positive
cranially). Distances are expressed as percentage segment length between the two reference markers. Location along the x-axis is measured first from the proximal reference
marker toward the distal reference marker, then shifted along the y-axis. Data are mean ± SD for 12 forelimbs of 6 Warmblood horses.
Reprinted from Willemen, M.A., Savelberg, H.H.C.M., Barneveld, A., 1997, The improvement of the gait quality of sound trotting warmblood horses by normal shoeing and its
effect on the load on the lower forelimb, Livestock Production Science, 52 (2), 145–153, with permission from Elsevier.

similar attachments may, in fact, have quite different functional volume, has been determined to be 1.075 g/cm3 over a range of
responsibilities. Evaluation of the geometry and architecture of the muscles, with different muscles varying by only a small amount
musculotendinous units is helpful in understanding whether their (Brown et al., 2003). Fiber length and pennation angle affect the
function is to produce rapid movements or generate large forces to range of motion through which the muscle contracts and its ability
stabilize the joints. to generate force. Long fibers arranged in parallel with the long axis
Muscle size is expressed in terms of its mass and volume, which of the muscle belly have the greatest capacity to shorten the muscle.
are closely correlated. Muscle density, calculated as mass divided by Thus parallel fibers impart the largest range of motion and the most

100
 Musculotendinous architecture

Table 6.2  Architectural properties of the extrinsic muscles of the equine forelimb

Muscle Mass (kg) Volume (cm3) MFL (mm) PCSA (cm2) Force (kN) Power (W)
Pectoralis transversus 1.54 1434 200 77 2310 231
Pectoralis descendens 2.84 2649 461 60 1800 415
Pectoralis ascendens 21.0 1954 292 72 2160 315
Serratus ventralis cervicis 29.9 2781 49 577 17310 424
Serratus ventralis thoracis 24.3 2349 693 62 1860 644
Brachiocephalicus/Omotransversarius 13.0 1217 519 23 690 179
Subclavius 0.68 631 191 42 1260 120
Trapezius 1.83 1705 378 53 1590 301
Rhomboideus cervicis 0.55 503 311 15 450 70
Rhomboideus thoracis 0.43 409 139 24 720 50
Latissimus dorsi 1.83 1705 378 53 1590 301

MFL, mean fiber length; PCSA, physiological cross-sectional area; force, maximal isometric force generation capacity estimated by multiplying muscle PCSA by the maximal
isometric stress of skeletal muscle, taken as 0.3 MPa; power, maximal power output calculated as one-tenth of the product of force and maximal contraction velocity, which
was estimated based on published values of equine muscle fiber-typing.
Reprinted from Payne, R.C., Veenman, P., Wilson, A.M., 2004, The role of the extrinsic thoracic limb muscles in equine locomotion, Journal of Anatomy, with permission from
John Wiley and Sons.

rapid shortening velocities. These characteristics are associated with Extrinsic muscles of the forelimb
muscles that act as prime movers. When fibers insert into the tendon
at an angle (pennation angle), the amount of force transmitted to The extrinsic muscles of the forelimb, which have an attachment to
the tendon is determined as force developed in the fiber multiplied the bones of the limb and an attachment to the trunk, are respon-
by the cosine of the pennation angle. Thus, fibers with a pennation sible for suspending the trunk between the forelimbs and for
angle of 40° transmit only 77% of the force generated by the muscle moving the forelimbs relative to the trunk. In general, these are large
fiber to the tendon. In the equine antebrachial muscles, a close muscles with long fibers arranged parallel with the long axis of the
inverse relationship between muscle fiber length and pennation muscle belly (Table 6.2) that insert on the bones via short tendons
angle has been demonstrated (Brown et al., 2003), and this rela- or aponeurotic sheets (Payne et al., 2004). This internal architec-
tionship may be true for other anatomical areas. tural style favors the ability to contract rapidly through a wide range
Physiological cross-sectional area (PCSA), which is a determinant of motion at the expense of the ability to generate high forces. The
of the maximal isometric force that can be generated by a muscle, exception is serratus ventralis thoracis, which has short, pennate fibers
is calculated from muscle volume divided by fiber length. For and is encased in a strong aponeurotic sheath. These features suggest
muscles with equal volumes, PCSA is larger in muscles with short, that serratus ventralis thoracis bears primary responsibility for anti-
pennated fibers because a larger number of fibers can be contained gravitational support of the trunk, whereas the other extrinsic
within the volume. These characteristics, which confer an increase muscles move the forelimb relative to the trunk during the swing
in passive stiffness to the muscle, are typical of the anti-gravity phase or advance the trunk over the grounded limb during the
muscles that support the body during the stance phase. Short, stance phase.
highly pennated fibers in combination with a long, elastic tendon Serratus ventralis thoracis (Fig. 6.3, Table 6.2) is the largest extrinsic
are characteristic of muscles involved in elastic energy storage and muscle of the forelimb both in mass and volume and has the short-
release. est mean fiber length. It is a broad flat muscle covered medially and
The force of contraction of a muscle can be estimated in a Hill- laterally by broad aponeurotic sheets. The shortness of the muscle
type model based on four parameters: fiber length, maximal fiber fibers (49 mm) relative to the length of the aponeurosis (500 mm)
shortening velocity, pennation angle, and peak isometric muscle offers limited ability for muscular contraction to compensate for
force (Zajac, 1989). In general, muscle size (volume and mass) and stretching of the aponeurosis, which implies that this muscle would
fiber length decrease in a proximal to distal direction within the not be effective in moving or positioning the limb. The elasticity of
equine limbs. Muscles in the proximal forelimb tend to be large the aponeurosis may contribute to the overall elastic properties of
and powerful with long fibers arranged parallel to the muscle belly the limb by acting in series with the spring-like tendons in the more
to move the joints through a large range of motion. By comparison, distal part of the limb and may support protraction of the scapula
muscles in the distal limb are smaller and less powerful with short, (Smythe et al., 1993). Pennation of the fibers and the consequent
pennate fibers that are not capable of a large amount of shortening increase in cross-sectional area gives the muscle considerable passive
but are well suited to contract isometrically. When muscle length is stiffness and allows it to generate sufficiently high forces to with-
controlled by an isometric contraction, the long tendons are loaded stand gravitational loading of the limbs during galloping. Serratus
elastically as the limb accepts weight in early stance, then recoil to ventralis thoracis is the primary muscular component of the thoracic
release the stored elastic energy at the end of stance. Large, cursorial sling, which suspends the trunk between the forelimbs and controls
animals use this musculotendinous arrangement to move at high the position of the thorax and withers relative to the scapulae when
speeds with a relatively economical metabolic cost (Alexander, the forelimbs are weight-bearing. During standing, the trunk is sup-
2002). ported passively by elongation of the muscle fibers and the series

101
 6 Forelimb function

Splenius

Rhomboideus

Trapezius

Latissimus dorsi

Serratus ventralis cervicis

Brachiocephalicus

Omotransversarius

Descending pectoral

Serratus ventralis
Ascending pectoral thoracis

Fig 6.3  Extrinsic muscles of the forelimb.

From Stubbs and Clayton (2008) with permission of Sport Horse Publications.

elastic elements. Much higher forces must be resisted during loco- descendens and pectoralis transversus) are smaller muscles with
motion when gravitational and inertial loads on the limbs produce medium length fibers (Payne et al., 2004).
peak vertical forces of at least 8 kN per limb for a 500 kg horse Subclavius (Table 6.2) has long fibers (519 mm) that allow genera-
(McGuigan & Wilson, 2003). Loading during locomotion is easily tion of large forces to assist in adduction and retraction of the
resisted by the maximal isometric force-generating capacity of ser- forelimb or stabilization of the scapula (Payne et al., 2004).
ratus ventralis thoracis, which is estimated to exceed 17 kN. This Trapezius (Fig. 6.3) and rhomboideus are the smallest extrinsic
muscle is more variable in its mass than the other extrinsic muscles, muscles of the forelimb and have medium-length fibers (Table 6.2)
which may reflect adaptation in response to the amount and type (Payne et al., 2004). Both muscles insert on the dorsal scapula and
of training. have fiber orientations ranging from vertical in the middle of the
Serratus ventralis cervicis (Fig. 6.3, Table 6.2) is somewhat smaller muscle to almost longitudinal towards the extremities. The orienta-
than the thoracic part of the muscle and differs in having relatively tion of the vertical fibers suggests that their function is to hold the
long fibers and a smaller cross-sectional area (Payne et al., 2004), proximal scapula against the trunk (preventing winging of the
that confer the ability to support the base of the neck or to retract scapula) by opposing the action of the pectoral muscles that adduct
the limb by rotating the proximal scapula cranially. In dogs, it has the distal scapula and humerus. The longitudinal fibers likely con-
been suggested that the primary function of serratus ventralis cervicis tribute to forelimb protraction, retraction and stabilization.
is to stabilize the position of the fulcrum about which the forelimb Brachiocephalicus (Fig. 6.3) and omotransversarius are considered
rotates in a craniocaudal direction during active retraction of the together since their fibers cannot be separated close to their
forelimb, thus ensuring that the GRF vector passes close to the origins in the shoulder region. They form a large, powerful muscle,
center of scapular rotation (Carrier et al., 2006). with long fibers oriented parallel to the muscle belly (Payne et al.,
The pectoral muscles (Fig. 6.3, Table 6.2) are large with relatively 2004). Fiber lengths in Table 6.2 represent a mean of the two
long fibers that suggest a primary role in adducting the forelimb muscles, though brachiocephalicus has longer fibers than omotrans-
(Payne et al., 2004). The deep pectoral (pectoralis ascendens) is a versarius. The fiber direction suggests a primary role in forelimb
large, powerful muscle with long fibers. It is active through most of protraction.
the stance phase when it may assist in moving the trunk forward Latissimus dorsi (Fig. 6.3, Table 6.2) is a moderately large and
over the grounded limb. The superficial pectorals (pectoralis powerful muscle with fairly long fibers (Payne et al., 2004). It is

102
 Musculotendinous architecture

enveloped by an aponeurotic sheath that is part of the thoracolum- movement is required. In a galloping horse, the biceps tendon has
bar fascia. The fiber orientation suggests a primary role in forelimb been estimated to release 243 J in 0.11 s, which would require the
retraction. Compared with the primary limb protractors, brachioce- power output of 50 kg of non-elastic muscle (Wilson & Watson,
phalicus and omotransversarius, latissimus dorsi develops similar 2003). Thus the biceps catapult is an effective and efficient mech-
amounts of force but less than half as much power. Electromyo- anism for protracting the forelimb in galloping horses. A further
graphic studies indicate that this muscle is active in late swing and benefit of using tendon elasticity rather than muscular force is
at hoof contact (Preedy, 1998), when it may act to retract the limb that the tendon has low energy dissipation, returning an impres-
in preparation for ground contact and stabilize the trunk during the sive 93% of the work done stretching it and with only 7% dissipa-
impact phase. tion as heat.
Part of the internal tendon of biceps brachii emerges from the
muscle and continues distally as the lacertus fibrosus, a tendinous
band that blends with the epimysium of extensor carpi radialis. Lac-
Intrinsic muscles of the forelimb ertus fibrosus has a much smaller mass than the internal tendon of
The intrinsic muscles of the forelimb are characterized by being biceps brachii and is capable of storing much less energy (10–28 J)
smaller in volume than the extrinsic musculature with short, highly (Watson & Wilson, 2007). These properties are consistent with its
pennate fibers and long tendons relative to muscle length. Total role in stabilizing the forelimb as part of the stay apparatus.
mass of the intrinsic musculature is around 17 kg in each forelimb. Biceps brachii has tonic activity during standing (Tokuriki et al.,
Pennation of the muscle fibers results in a larger PCSA than for 1989), which supports the suggestion that the lateral part of the
equal-sized muscles with parallel fibers. This confers the ability to muscle, with its high proportion of type I muscle fibers, acts in series
resist elongation of the muscle (isometric contraction) as the limb with lacertus fibrosus and extensor carpi radialis to stabilize the shoul-
is loaded, so elongation of the musculotendinous unit is due to der as part of the passive stay apparatus (Hermanson, 1997). During
stretching of the tendon that acts in series with the muscle. As the locomotion at walk, trot and canter, biceps brachii is active in early
tendons stretch, they store elastic energy, which is released later in and midstance (Tokuriki et al., 1989). This is in contrast to brachia-
the stance phase when the limb is unloaded. This mechanism lis, which acts as an elbow flexor in early swing (Tokuriki et al.,
reduces muscular work and increases the economy of locomotion 1989) and contributes to the flexor moment at the elbow at this
(Alexander, 2002). time in walk (Clayton et al., 2000a, b) and trot (Lanovaz et al.,
Supraspinatus, with a mass of 793–1546 g and fiber length of 1999).
5–12 cm, has limited force generating capacity and a small moment Triceps brachii is a biarticular muscle crossing the flexor aspect of
arm at the shoulder, which makes it more suitable for stabilization the shoulder (long head) and extensor aspect of the elbow. It is by
than dynamic movement of the joints (Watson & Wilson, 2007). far the largest of the intrinsic forelimb muscles. As the name sug-
Supraspinatus and infraspinatus are active during early and mid- gests, it has three heads. The biarticular long head (mass, 3200–
stance in walk, trot and canter (Aoki et al., 1984; Robert et al., 6663 g; fiber length, 19–26 mm) and the monoarticular lateral
1998), when the primary action of supraspinatus appears to be sta- head (mass, 514–1240 g; fiber length, 17–24 mm) comprise 81%
bilization of the shoulder joint. The instability seen with paralysis and 15%, respectively, of the extensor muscle mass at the elbow
of the supraspinous nerve (Sweeney) supports this presumptive (Ryan et al., 1992). They contain predominantly fast-twitch fibers
function. suggesting they are important in locomotion. By contrast, the small
The medial and lateral heads of the biarticular biceps brachii span medial head of triceps (mass, 85–271 g; fiber length, 9–17 cm) and
the extensor aspect of the shoulder and the flexor aspect of the the anconeus muscle each account for only 2% of elbow extensor
elbow. The lateral head, with a mass of 171–343 g, is composed of muscle mass and both are composed almost entirely of slow-twitch
short (5–8 mm), pennate fibers (Watson & Wilson, 2007), a large fibers, suggesting their role is to support the elbow in extension
percentage of which are slow-twitch and well suited to postural during stance (Ryan et al., 1992).
control. The medial head, which has longer fibers (15–40 mm) and The actions of triceps are to extend the elbow, to retract and
fewer slow-twitch fibers, may be more important in locomotion extend the distal forelimb and, perhaps, to extend the limb when
(Hermanson & Hurley, 1990). Compared with supraspinatus, biceps it is being used to raise the forehand. It is active in late swing and
brachii has a larger force generating capacity and a larger moment early stance (Tokuriki et al., 1989; Preedy, 1998; Robert et al.,
arm at the shoulder joint, which suggests that it may be a more 1998), with activity in the long head preceding activity in the lateral
effective extensor of the shoulder (Watson & Wilson, 2007). head. Since biceps brachii and the long head of triceps brachii are
A strong internal tendon (mass, 122–260 g; fiber length, biarticular, their interaction affects motion and stability of both
9–17 cm) runs through the muscle belly of biceps brachii uniting shoulder and elbow joints. The nature of this interaction differs
the tendons of origin and insertion. This tendon plays an impor- between the inverted pendulum behavior of the limb during
tant role in elastic energy storage, having the potential to store walking and the mass-spring behavior during trotting. In the walk­
277–591 J. It is estimated to withstand forces of 3.2 × 104 − 5.4 × ing horse, the shoulder acts primarily as an energy damper, with a
104 N when stretched as the forelimb is retracted in late stance large burst of energy absorption on its extensor (cranial) aspect in
with the shoulder and elbow in extension (Watson & Wilson, midstance (Clayton et al., 2000a), which is likely due to eccentric
2007). Recoil of the stretched biceps tendon in the galloping action of biceps brachii controlling extension of the shoulder. At the
horse has been described as a catapult mechanism that provides elbow there are bursts of energy generation on the extensor aspect
rapid acceleration of the distal forelimb (Wilson & Watson, in early stance, which is thought to be due to concentric action of
2003). In a catapult, a large force is applied to store energy, which triceps brachii, and on the flexor aspect in late stance, which coin-
is then released rapidly to accelerate a small mass. In the horse’s cides with electrical activity in biceps brachii (Tokuriki et al., 1989).
forelimb, the biceps tendon is stretched by forward movement of During trotting, the elbow has well-defined bursts of energy absorp-
the trunk and the changing orientation of the ground reaction tion on the extensor side of the joint in early stance followed by
force vector relative to the shoulder and elbow joints during the energy generation on the extensor aspect in midstance (Clayton
stance phase. Stretching continues as long as the carpus is locked et al., 1998). Energy absorption and generation on the extensor
in extension. When the carpus buckles, the forearm is released (cranial) aspect of the shoulder joint are similar in magnitude to
allowing the biceps tendon to recoil. The effect is rapid extension those at the elbow joint but occur slightly later in the stance phase.
of the shoulder, flexion of the elbow and forward acceleration of There is an additional burst of energy generation on the extensor
the distal limb. Tendons can recoil elastically much faster than aspect of the shoulder in late stance (Clayton et al., 1998) corre-
muscles shorten, which is beneficial in situations where rapid sponding with activity in biceps brachii (Tokuriki et al., 1989). Thus

103
 6 Forelimb function

Table 6.3  Architectural properties of the muscles of the equine antebrachium

Muscle Mass (kg) Volume (cm3) MFL (mm) PCSA (cm2) MPA (°)
Flexor carpi radialis 1.80 166 89.7 18 6.7
Extensor carpi radialis 8.15 754 76.0 99 16.0
Flexor carpi ulnaris 2.62 245 18.3 134 31.6
Ulnaris lateralis 3.64 337 17.4 194 34.3
Superficial digital flexor 2.45 227 7.5 303 41.6
Deep digital flexor – humeral head 5.61 524 43.8 120 21.6
Deep digital flexor – radial head 0.45 42 10.6 39 32.3
Deep digital flexor – ulnar head 0.96 89 37.2 24 20.3
Common digital extensor 3.16 295 81.4 36 13.3

MFL, mean fiber length; PCSA, physiological cross sectional area; MPA, mean pennation angle.
Reprinted from Brown, N.A.T., Kawcak, C.E., McIlwraith, W., Pandy, M.G., 2003, Architectural properties of distal forelimb muscles in horses, Equus caballus, Journal of Morphology,
with permission from John Wiley and Sons.

biceps brachii appears to be responsible for a net extensor moment the inertially driven carpal flexion (Colborne et al., 1997a, b).
at the shoulder through most of stance, while triceps brachii is gen- Rupture of the extensor carpi radialis tendon allows the carpus to
erating an extensor moment at the elbow. In the swing phase, hyperflex during the swing phase at walk and may cause the horse
forelimb protraction is driven by the elbow flexors, which generate to stumble or fall at trot because it cannot protract the forelimb
a flexor moment in early swing, and retraction is driven by the rapidly enough. Later in the swing phase, the net joint moment
elbow extensors, which generate an extensor moment in late swing moves to the caudal side of the carpus (Lanovaz et al., 1999), with
(Lanovaz et al., 1999). flexor carpi radialis controlling carpal extension and retracting the
The muscles of the forearm move and stabilize the carpal and distal limb in preparation for ground contact.
digital joints. The architectural properties of these muscles have The digital flexor and extensor muscles (Table 6.3) are character-
been described (Hermanson, 1997; Hagen et al., 2002; Brown et al., ized by having long tendons relative to their muscle length. However,
2003; Zarucco et al., 2004) and are summarized in Table 6.3. The there are marked differences in muscle architecture and tendon
carpal flexors and extensors have longer muscle bellies with shorter properties of the deep and superficial digital flexor muscles that are
tendons than the digital flexors and extensors. The carpal net joint indicative of the different roles played by these musculotendinous
moment acts on the palmar aspect through most of stance, suggest- units in locomotion.
ing that tension in the carpal flexors assists the passive tendinous The deep digital flexor (DDF) has three distinct muscle bellies,
structures in maintaining joint stability (Clayton et al. 1998, humeral, ulnar and radial, each of which is innervated by a separate
2000a, b). Flexor carpi ulnaris and ulnaris lateralis have short, highly branch of the median nerve suggestive of neuromuscular compart-
pennated fibers (pennation angle close to 30o), which results in a mentalization (Zarucco et al., 2004). The DDF is also compartmen-
large PCSA (Brown et al., 2003). Both of these muscles insert on talized morphologically into regions with different lengths of fibers
the accessory carpal bone, which increases their moment arm and in the range 5–117 mm (Hagen et al., 2002; Brown et al., 2003;
facilitates their ability to stabilize the carpus during stance. These Zarucco et al., 2004). The humeral head, which comprises about
muscles show electromyographic activity in late swing and early 74% of total muscle volume, has the longest fibers of any muscle
stance (Jansen et al., 1992) indicating a possible role in stabilizing in the antebrachium, but also has some short and medium length
the carpus through the impact phase. (Note that ulnaris lateralis fibers. Its tendon is long, but not as long as that of the ulnar head,
represents the extensor carpi ulnaris but, since its distal attachment which also has long fibers but relatively small mass and PCSA
is to the accessory carpal bone in horses, it acts as a carpal flexor compared with the humeral head. In contrast, the fibers in the
rather than a carpal extensor.) radial head are short and highly pennated (pennation angle close
Flexor carpi radialis and extensor carpi radialis (Table 6.3) have long to 30°). The functions of the DDF are to flex the digital joints during
muscle fibers with small pennation angles (less than 20°), a small the swing phase and to generate a propulsive force during the
PCSA, and short tendons (Brown et al., 2003). These qualities second half of stance. Since the DDF muscle has a relatively high
suggest a role in initiating and controlling carpal flexion/extension percentage of fast-twitch fibers, it is susceptible to fatigue during
during the swing phase. Extensor carpi radialis also shows evidence exercise (Hermanson & Cobb, 1992; Butcher et al., 2007). The DDF
of a degree of compartmentalization; the proximal portion has a tendon, which functions as a positional tendon, has a higher
predominance of fast-twitch fibers that can contract powerfully, modulus of elasticity than the SDF tendon, which is used for elastic
while the distal portion has more slow-twitch fibers (Hermanson, energy storage and release (Birch, 2007). At the walk, the deep head
1997). The fatigue resistant slow-twitch fibers arranged in series of DDF is active throughout stance with peaks of activity in early
with the tendons of biceps brachii and lacertus fibrosus are part of the stance and late stance (Jansen et al., 1992). Horses with palmar foot
passive stay apparatus of the forelimb. Electromyographic activity pain may prolong the activation of DDF in early stance to move the
in extensor carpi radialis is concentrated at the beginning of swing center of pressure beneath the hoof in a dorsal direction, thereby
(Jansen et al., 1992) when the elbow and carpus are flexing. At this relieving weight-bearing in the palmar part of the hoof (Wilson
time, the net joint moment is on the cranial side of the joints et al., 2001).
(Lanovaz et al., 1999) indicating that the biarticular extensor carpi The superficial digital flexor (SDF) muscle has a much smaller
radialis is flexing the elbow to protract the limb and is controlling volume than the DDF muscle. However, the fibers are very short

104
 Sagittal plane analysis of forelimb kinematics and kinetics

(2–10 mm) and highly pennate (pennation angle up to 60°) In the equine distal limb, the suspensory ligament and SDF
(Grandage, 1981; Dimery et al., 1986; Hermanson & Cobb, 1992; tendon are primarily responsible for storing and releasing elastic
Biewener, 1998; Wilson et al., 2001; Brown et al., 2003; Zarucco energy. They are subject to high tendon strains as the limb is loaded
et al., 2004), resulting in the largest cross-sectional area of the during stance, especially at the trot. In the canter, overall limb
antebrachial muscles and endowing a large force-generating capac- loading decreases with less elastic energy being stored in the SDF
ity (Hagen et al., 2002; Brown et al., 2003). As a consequence of tendon, and the DDF tendon being more loaded (Butcher et al.,
the large pennation angle, the SDF is capable of minimal shortening 2007). With an increase in galloping speed, the DDF muscle and
and only about 71% of the total muscle force is transmitted to the its associated tendon assume a greater role in support of the MCP
tendon. The SDF muscle has a high percentage of slow-twitch joint, thus relieving some stress on the SDF tendon. The DDF
muscle fibers (Butcher et al., 2007) that are resistant to fatigue. This muscle, however, has fast-twitch fibers that contract concentrically
fiber composition is well suited for its support role as part of the and are susceptible to fatigue. When the DDF muscle becomes
stay apparatus (Swanstrom et al., 2005), and for attenuating high- fatigued, the SDF tendon is over-loaded and predisposed to strain
frequency forces associated with impact (Wilson et al., 2001). The injury (Butcher et al., 2007), which occurs frequently in equine
fatigue-resistant, slow-twitch fibers of the SDF tendon act eccentri- athletes, especially in Thoroughbred racehorses (Peloso et al.,
cally or isometrically during stance with changes in length of the 1994). Even though the elastic tendons are highly energy efficient
musculotendinous unit being due almost entirely to stretching of and dissipate only 7% of the stored elastic energy as heat, tempera-
the elastic tendon (Butcher et al., 2007). tures around 45°C have been found in the SDF tendon of race-
The third interosseous muscle (suspensory ligament) acts as an horses after galloping (Birch et al., 1997). Repeated loading of the
energy-storing tendon and has an even lower modulus of elasticity musculotendinous unit during training and racing may lead to the
than the SDF tendon (Birch, 2007). It supports the metacarpopha- accumulation of subclinical microdamage if the processes of repair
langeal (MCP) joint during the stance phase, which is critical to the and adaptation are unable to keep up with the rate of tissue damage
function of the equine limb. During standing, the suspensory liga- (Hill et al., 2001).
ment is fully capable of supporting the horse’s weight passively Hoof angle affects strain distribution between the tendoligamen-
(Dyce et al., 1996). It has been speculated that reduction in the tous structures in the distal limb. Elevation of the heels reduces peak
muscular function of the equine interosseus began about 15 million strain in the DDF tendon and increases peak strain in the SDF
years ago, when ancestral horses were increasing in size and moving tendon and suspensory ligament, whereas raising the toe has the
to the grasslands where efficient overground locomotion was opposite effects (Lawson et al., 2007).
required (Camp & Smith, 1942). Today, muscle fibers comprise
about 10% of the suspensory ligament. These muscle fibers are short
(~0.88 mm) and highly pennate (pennation angles >45°), which Sagittal plane analysis of forelimb kinematics
increases the PCSA and contributes to the ability to generate forces
but produces little work. The vast majority (95%) of the muscle
and kinetics
fibers are type I, and presumably slow-twitch (Wilson et al., 2001;
Soffler & Hermanson, 2006). Thus, the muscular content of the The joints of the horse’s forelimb from the elbow distally are more
suspensory ligament is small but significant and the architecture of or less constrained to move in a sagittal plane with relatively small
the fibers suggests that it contributes to forelimb stability and elastic amounts of abduction/adduction and internal/external rotation
energy storage during locomotion. (Thompson et al., 1992; Degueurce et al., 1996). Studies of three-
The common digital extensor (CDE) (Table 6.3) and lateral dimensional forelimb kinematics, which will be described at the
digital extensor muscles have long fibers, small PCSA and long end of this chapter, have confirmed that flexion/extension is the
tendons (Brown et al., 2003). The CDE muscle is active in terminal dominant rotation at the carpal, metacarpophalangeal, proximal
swing at the walk, when it extends the digit in preparation for interphalangeal (PIP) and distal interphalangeal (DIP) joints
ground contact, (Jansen et al., 1992). The CDE tendon has a stiffer (Chateau et al., 2004, 2006; Clayton et al., 2004, 2007a, b; Hobbs
matrix than the SDF tendon, which may be due to the smaller fibril et al., 2006). Inverse dynamic analysis has been used to calculate
diameters (Birch, 2007). net moments of force, net joint powers and net joint energies for
As horses bounce over the ground in the trot, canter and gallop, the joints of the equine forelimb at walk and trot (Hjertén &
the forelimbs have been estimated to contribute one-third of the Drevemo, 1993; Clayton et al., 1998, 2000a, b; Colborne et al.,
energy storage compared with two-thirds in the hind limbs 1998; Lanovaz et al., 1999; Hodson et al., 2000; Khumsap et al.,
(Biewener, 1998). During trotting, changes in potential and kinetic 2002, Dutto et al., 2006).
energy of the horse’s center of mass are in phase, which allows the When using skin markers to represent two-dimensional motion
distal limb to make substantial contributions to elastic energy of the limb segments, a minimum of two markers per segment is
storage (Biewener, 1998). In canter and gallop, the relationship required. Typical marker configurations involve either placing a
between kinetic and potential energy varies during the stride marker over the center of rotation of each joint or aligning two
(Minetti et al., 1999), which limits the ability to store elastic markers along the long axis of each segment (Fig. 6.4). Intra-limb
energy in the distal limb. At faster speeds, vertical excursions of coordination patterns can be visualized using stick figures or joint angle–
the center of mass are reduced and the limb sweeps through a time graphs (Fredricson & Drevemo, 1972; Fleiss et al., 1984;
larger angle during its stance phase causing the horse to bounce Martinez-del Campo et al., 1991; Holmström et al., 1994; Back
off the ground more quickly (Farley et al., 1993). The concept et al., 1994; Degueurce et al., 1997; Nicodemus & Holt, 2006;
of the forelimb acting as a spring implies that changes in joint Martuzzi et al., 2007; Nicodemus & Booker, 2007).
angles as the limb accepts weight result in shortening of the bony Skin displacement relative to the underlying bones is always a
column and stretching of the musculotendinous units. During gal- concern when kinematic studies are based on skin-fixed markers.
loping, the proximal limb from scapula to elbow shortens by In the equine forelimb, artifacts due to skin displacement may be
about 12 mm, whereas the limb distal to the elbow shortens small enough to be negligible on the antebrachial and metacarpal
by around 127 mm. One of the most obvious articular changes is segments, but are large enough to alter sagittal plane kinematics
extension of the MCP joint, the magnitude of which can be pre- significantly on the scapular, brachial, and pastern segments. Since
dicted from vertical ground reaction force (McGuigan & Wilson, skin displacement has a cyclic pattern, it has been possible to
2003). In general, the limb spring is stiffer in larger animals develop mathematical correction algorithms for many of the ana-
but, within an individual, stiffness of the limb spring is nearly tomical locations that are commonly used for marker placement
independent of speed. (van Weeren et al., 1990a, b; Sha et al., 2004).

105
 6 Forelimb function

13 retraction of the limb. Maximal protraction precedes initial ground

13 contact, and maximal retraction occurs just after lift-off. Angle–time
curves before and after correction for skin displacement are similar,
except that maximal retraction and the magnitude of a dip at mid-
swing are underestimated without correction for skin displacement
(Fig. 6.5).

12
Shoulder joint
11 The shoulder joint is extended at initial ground contact, then flexes
as the limb is loaded. There is so much skin movement over the
11 scapula and humerus that it is not possible to describe shoulder
joint kinematics in late stance or during the swing phase based on
10 skin-fixed markers unless correction procedures are applied (Fig.
6.5) (van Weeren et al., 1990a; Back et al., 1995a). The net joint
moment at the shoulder (Fig. 6.6) moves from the cranial to the
9 caudal aspect and back to the cranial (extensor) aspect in early
10 stance, then sustains a cranial (extensor) moment until lift-off due
to the action of biceps brachii assisted by other extensors of the
8 shoulder (Clayton et al., 1998). The power profile suggests a com-
7 ponent of elastic energy storage and release as the joint is loaded
6 in early stance, followed by energy generation on the extensor aspect
5 in terminal stance (Fig. 6.6). Early in the swing phase, there is an
extensor moment at the shoulder joint as the extensor musculature
4
acts eccentrically to control flexion, then the net joint moment
4 5 moves to the flexor side, where the muscles work eccentrically to
14 slow extension and initiate flexion (Lanovaz et al., 1999; Clayton
1 3 et al., 1998; Singleton et al., 2003).
2
A B

Fig 6.4  Bony landmarks underlying skin-fixed markers in two marker- Elbow joint
placement schemes. Left: markers placed over centers of joint rotation with The elbow joint extends through most of stance reaching maximal
limb segments being represented by lines joining the markers. Right: two extension near the end of stance (Fig. 6.7). During the swing phase
markers placed along the long axis of each segment are joined to represent it undergoes a flexion cycle. The amplitudes of flexion and extension
the segment with adjacent segments intersecting at the joints. The hoof do not appear to be influenced much by skin displacement, but
may be represented by different combinations of markers with radiographic without correction the swing phase flexion peak occurs somewhat
identification of the center of rotation of the DIP joint relative to the hoof later (Fig. 6.4) (van Weeren et al., 1990a; Back et al., 1995a).
markers. 1, hoof at toe; 2, hoof at mid-lateral distal wall; 3, hoof at heel; The net joint moment at the elbow acts predominantly on the
4, hoof at coronary band; 5, distal condyle of metacarpus; 6, proximal
caudal (extensor) side during most of the stance phase to resist
metacarpus; 7, ulnar carpal bone; 8, distal radius at lateral styloid process; 9,
collapse of the joint under the influence of gravity (Fig. 6.7). This
proximal radius at collateral ligament of elbow; 10, distal humerus at lateral
epicondyle; 11, proximal humerus at caudal greater tubercle; 12, distal
is probably due to the action of triceps brachii, which shows electro-
scapular spine; 13, proximal scapular spine; 14, proximal first phalanx. myographic activity at this time (Korsgaard, 1982; Tokuriki et al.,
1989; Robert et al., 1998). The net joint moment shifts to the
cranial (flexor) side of the elbow in latestance. The power profile
indicates a phase of energy generation and absorption in early
stance, followed by a larger phase of energy generation in mid-
stance (Fig. 6.7). There is a final period of energy absorption on
The trot the flexor aspect as elbow extension slows at the end of stance
The trot is the most important gait for evaluation of the quality of (Singleton et al., 2003). During the first half of swing, the elbow
a horse’s movement and for detection of lameness. It is a two-beat flexes with a flexor moment (Lanovaz et al., 1999) indicating active
gait with the limbs coordinated by diagonal pairs. The diagonal flexion driven by the internal tendon of biceps brachii and flexor
support phases are usually separated by aerial or suspension phases muscles, such as brachialis (Tokuriki et al., 1989). There is a short
in which all feet are off the ground (Alexander & Jayes, 1978). Back period at midswing when the elbow continues to flex after the net
et al. (1995a) used standardized procedures to describe sagittal joint moment has moved to the extensor side coinciding with
plane kinematics of the trot in a large group of Warmblood horses. the activation of triceps brachii (Korsgaard, 1982; Tokuriki et al.,
The joint angle–time diagrams were analyzed simultaneously with 1989) and, later, flexor carpi ulnaris (Korsgaard, 1982; Jansen et al.,
corresponding stick figures and marker diagrams to create a com- 1992). The elbow extensor muscles slow flexion (eccentric action)
plete picture of equine forelimb motion at the trot that could be and then actively extend the elbow joint (concentric action) in
related to limb function (Back et al., 1995a). Variability between preparation for ground contact (Lanovaz et al., 1999; Singleton
horses was evaluated, and the effect of correcting for skin displace- et al., 2003).
ment on joint angle–time diagrams was quantified and illustrated
graphically (Fig. 6.5) (van Weeren et al., 1990a; Back et al., 1995a).
Carpal joint
After initial ground contact the carpus quickly snaps into its over-
Scapula extended (close-packed) position and remains in this stable posi-
The scapula rotates around a point close to its proximal end tion during most of stance. At faster trotting speeds, the carpus
with a sinusoidal pattern that corresponds with protraction and may snap into its stable position just before initial ground contact

106
 Sagittal plane analysis of forelimb kinematics and kinetics

Fig 6.5  Mean joint angle–time diagrams of the forelimb of an individual horse trotting on a treadmill at a speed of 3 m/s before (continuous line) and after
(dotted line) correction for skin displacement.
Reprinted from Back, W., Schamhardt, H.C., Savelberg, H.H.C.M,. et al. (1995a) How the horse moves: significance of graphical representations of equine forelimb kinematics. Equine Vet. J. 27,
31–38, with permission from the Equine Veterinary Journal.

(Back et al., 1993, 1995a; Johnston et al., 1997). A functional simi- where it controls the whiplash effect in the distal limb that results
larity between the equine carpus and the human knee has been in carpal extension (Fig. 6.8) (Lanovaz et al., 1999). The flexor
identified (Colborne et al., 1997a,b). After a second extension peak moment increases during the last one-third of swing as flexor carpi
at midstance, the carpus starts to flex and, before the end of stance, ulnaris is activated (Korsgaard, 1982). In preparation for initial
initiates rapid flexion that peaks in midswing (Fig. 6.8). Correction ground contact, elbow extension is accompanied by synchronous
for skin displacement does not significantly alter carpal kinematics extension of the carpus and MCP joint with a proximal to distal
(Fig. 6.4) (van Weeren et al., 1990a; Back et al., 1995a). The amount transfer of angular momentum.
of hyperextension during stance increases with speed and is greater If the carpal joint of an anesthetized horse is flexed, it snaps back
when moving on an uphill gradient than when traveling on flat into its stable extended position. The mechanism responsible is
ground at the same speed (Burn et al., 2006). known as the clicking phenomenon (Alexander & Trestik, 1989). It
The net carpal moment is on the palmar aspect through most of is due to the eccentric attachments of the collateral ligaments rela-
stance and peaks around midstance indicating that the deep palmar tive to their rotation axis, which makes the joint bistable: it springs
carpal ligament and other soft tissues on the palmar aspect limit either into full extension or strong flexion. Functions attributed to
hyperextension (Fig. 6.8) (Clayton et al., 1998). There are small this clicking phenomenon include storage of elastic energy (Rooney,
bursts of positive and negative work but, compared with the other 1990) and damping of oscillations (Mosimann & Micheluzzi,
forelimb joints, the carpus plays only a small role in energy genera- 1969). The moments required to produce these movements in vitro
tion and absorption in the stance phase. However, it does play an are so small compared to those occurring in vivo that if the clicking
important role in supporting the forelimb as a propulsive strut phenomenon plays any role in locomotion, it must be during the
(Smythe et al., 1993). swing phase when the limb is not loaded. It is speculative whether
During early swing, carpal flexion is controlled by an extensor the inflection in the carpal flexion curve in early swing that coin-
moment (Lanovaz et al., 1999) that is most likely due to the action cides with the first metacarpophalangeal flexion peak is due
of passive structures, such as lacertus fibrosus, together with activa- to a clicking phenomenon and thus enables rapid movement
tion of extensor carpi radialis (Korsgaard, 1982; Jansen et al., 1992). (Palmgren, 1929) contributing to the elegance of gait (Alexander &
In midswing, the moment at the carpus moves to the flexor side Trestik, 1989).

107
 6 Forelimb function

140 150

140
135
Angle (deg)

Angle (deg)
130

130
120

125 110
0 20 40 60 80 100 0 20 40 60 80 100

1.5 0.3

1.0 0.2

0.1

Moment (Nm/kg)
Moment (Nm/kg)

0.5
0.0
20 40 60 80 100
0.0 -0.1
20 40 60 80 100
-0.2
-0.5
-0.3
-1.0 -0.4
Time (% stance)
3.0 3.0

2.0
Power (W/kg)

Power (W/kg)

1.0 0.0
20 40 60 80 100

0.0
20 40 60 80 100

-1.0 -1.0
Stance (%) Swing (%)
A B

Fig 6.6  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the shoulder joint during the stance phase (left) and swing phase (right) at the trot.

joint and its maximal palmar moment. Both the SDF and DDF
Metacarpophalangeal (MCP) joint muscles contract actively in the first half of stance (Korsgaard,
The angle-time graph (Fig. 6.9) shows rapid fetlock extension in 1982) to generate a palmar moment at the distal joints (Lanovaz
early stance peaking at midstance, after which the joint flexes et al., 1999). In addition, the proximal and distal accessory liga-
through late stance, sometimes with an inflexion preceding hoof ments provide mechanisms for passive support and the generation
lift. During the swing phase, two flexion peaks are separated by a of a palmar moment later in stance. The net joint power profile of
slight extension, and then, in late swing, the fetlock extends in the MCP joint during stance shows almost equal amounts of nega-
preparation for ground contact. Correction for skin displacement tive and positive work (Clayton et al., 1998), which is typical of
does not change the MCP joint pattern when the proximal (P1) elastic energy storage and release. Energy is absorbed in the first
and middle (P2) phalanges are treated as a single, rigid segment half of stance by the SDF tendon and suspensory ligament then
(Back et al., 1995a). The net joint moment (Fig. 6.9) acts on the released later as a result of elastic recoil.
palmar (flexor) side of the MCP joint during the entire stance In the swing phase, the MCP joint shows two flexion peaks (Fig.
phase peaking around midstance (Clayton et al., 1998). This 6.9) (Back et al., 1995a; Lanovaz et al., 1999), the first of which
reflects the elastic support of the palmar soft tissues as the MCP occurs soon after toe-off and is an elastic phenomenon due to
joint extends during weight acceptance then flexes for push-off. loading of the SDF tendon and suspensory ligament during the
The SDF tendon, DDF tendon and suspensory ligament experience preceding stance phase. The amount of flexion is controlled by a
peak strains around midstance (Riemersma et al., 1988a, b), net dorsal moment provided by the digital extensor tendons
which corresponds with the time of peak extension of the MCP (Lanovaz et al., 1999). After the elastic rebound, MCP joint flexion

108
 Sagittal plane analysis of forelimb kinematics and kinetics

140 140

Angle (deg)
Angle (deg)

130
135
120

130 110
0 20 40 60 80 100 0 20 40 60 80 100

1.3 0.20

0.9 0.10

Moment (Nm/kg)
Force (N/kg)

0.5 0.00
20 40 60 80 100

0.1 -0.10
20 40 60 80 100
-0.3 -0.20

5.0 0.8

0.6
3.0
Power (W/kg)

Power (W/kg)

0.4
1.0
0.2
20 40 60 80 100
-1.0
0
20 40 60 80 100
-3.0 -0.2
A Stance (%) B Swing (%)

Fig 6.7  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the elbow joint during the stance phase (left) and swing phase (right) at the trot.

can be fully explained by inertial forces: the dip in the joint flexion tendon acting through its distal accessory ligament combined with
curve is simply a relative extension of the MCP joint due to rapid tension in the navicular ligaments. The extensor branches of the
flexion of the carpus. When carpal flexion slows, MCP flexion suspensory ligament are taut to control hoof placement and prevent
resumes. After midswing, the proximal forelimb is decelerated and buckling of the interphalangeal joints in early stance (Jansen et al.,
a distal transfer of angular momentum results in a whiplash effect 1992). Later in stance, the DIP joint extends against a palmar
in the digit, causing the MCP joint to extend under the control of moment provided by the DDF tendon acting through its distal
a net palmar moment. Many of the movements of the MCP joint accessory ligament (Riemersma et al., 1988a, 1998b). In early
are driven inertially in the swing phase, with input from the flexor swing, a small extensor moment controls joint flexion after lift-off
and extensor muscles in late swing that prepares the hoof for (Lanovaz et al., 1999). Since there is little evidence of activity in the
ground contact. extensor muscles at this time (Korsgaard, 1982; Jansen et al., 1992),
this can be interpreted as a passive effect of tendinous and ligamen-
tous attachments. As the limb extends in the later part of the swing
Distal interphalangeal (DIP) joint phase, there is a whiplash effect with angular momentum being
The DIP joint flexes in early stance reaching maximal flexion before transferred distally causing the DIP joint to extend under the control
midstance (Fig. 6.10), then extends as the forelimb is retracted. of a flexor moment provided by the DDF muscle (Korsgaard, 1982;
Maximal extension occurs just after heel-off, after which the DIP Jansen et al., 1992) that slows joint extension prior to initial ground
joint flexes rapidly through early swing (Back et al., 1995a). The net contact.
DIP joint moment (Fig. 6.10) acts on the palmar (flexor) side The power curve (Fig. 6.10) has a burst of energy absorption
throughout the stance phase, peaking at 63% stance (Clayton et al., beginning one-third of the way through stance, peaking around
1998). The palmar moment is provided by tension in the DDF 75% stance when the longitudinal propulsive force is maximal, and

109
 6 Forelimb function

190

185
Angle (deg)

180

180
175

Angle (deg)
160
170
0 20 40 60 80 100 140

120
0.0
20 40 60 80 100
100
0 20 40 60 80 100
Moment (Nm/kg)

0.1
-1.0

Moment (Nm/kg)
0.0
20 40 60 80 100

-2.0

2.0 -0.1

1.0 0.05
Power (W/kg)

0.0 -0.05 20 40 60 80 100

20 40 60 80 100
Power (W/kg)

-1.0 -0.15

-2.0 -0.25

-3.0 -0.35
A Stance (%) B Swing (%)

Fig 6.8  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the carpal joint during the stance phase (left) and swing phase (right) at the trot.

continuing into breakover as the DIP joint extends against a palmar distal limb. In midswing, the forward velocity of the proximal limb
moment and acts as an energy damper (Clayton et al. 1998). is reduced and it may even stop rotating temporarily, which pro-
duces a whiplash effect that advances the distal limb rapidly with
a proximal to distal transfer of angular momentum that helps to
Functional interpretation conserve energy during swing (Hildebrand, 1987).
In the swing phase, forelimb motion has been likened to a pendu- During the stance phase, the net joint moment is on the caudal/
lum rotating around the proximal scapula (Krüger, 1938). All the palmar side of all joints except the shoulder through most of the
forelimb joints have their net joint moment on the cranial/dorsal stance phase at the trot (Clayton et al., 1998). During impact, the
side at the start of swing (Lanovaz et al., 1999), which is the exten- hoof is decelerated and impact shock is absorbed by mechanisms
sor side for all joints except the elbow. The elbow joint drives the that include shortening of the limb at the distal radius and proximal
forelimb movements during the swing phase at the trot; at the other metacarpus (Hjertén & Drevemo, 1993), with a corresponding
joints the muscles act eccentrically to control the motion of the deviation in the vertical GRF (Merkens & Schamhardt, 1994). At the
joints (Lanovaz et al., 1999). This is associated with generally low same time, rapid flexion of the DIP, MTP, and elbow joints assist in
EMG signals during the majority of swing (Korsgaard, 1982; damping the rapid build-up of force at impact (Back et al., 1995c;
Tokuriki et al., 1989; Jansen et al., 1992). The cranial/dorsal Johnston et al., 1995). Forelimb mechanics in the stance phase at
moments gradually decrease in magnitude and move to the caudal/ trot are described in terms of a spring-mass system in which limb
palmar side between 35% and 52% of swing. Peak values of the loading by the body mass stretches elastic springs that subsequently
swing phase net joint moments decrease in a proximal to distal recoil as the limb is unloaded thus increasing the efficiency of gait
direction, with peak moments at the shoulder and elbow joints (Dimery et al., 1986). The shoulder, elbow, carpal and MCP joints
being several times larger than those at the distal joints. At the show elastic behavior (Clayton et al., 1998) with the MCP joint
elbow, the net joint moment actively flexes and extends the joint being the main site of elastic energy storage and release primarily
during swing, which has the effect of protracting and retracting the due to the actions of the SDF tendon and the suspensory ligament.

110
 Sagittal plane analysis of forelimb kinematics and kinetics

240

230
200
Angle (deg)

220
180

Angle (deg)
210
160
200
0 20 40 60 80 100
140
0 20 40 60 80 100
0.0
20 40 60 80 100
-0.2 0.01
-0.4
Moment (Nm/kg)

-0.6

Moment (Nm/kg)
-0.8
-1.0 0.0
-1.2 20 40 60 80 100
-1.4
-1.6
-1.8 -0.01

6.0
0.04
4.0
-0.00
20 40 60 80 100
2.0
Power (W/kg)

Power (W/kg)

-0.04
0.0
20 40 60 80 100 -0.08
-2.0
-0.12
-4.0

-6.0 -0.16

A Stance (%) B Swing (%)

Fig 6.9  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the metacarpophalangeal joint during the stance phase (left) and swing phase (right) at the trot. The pastern segment (proximal and middle
phalanges) was treated as a rigid segment.

In addition to their role in energy efficiency, the limb springs act as similar with regard to the joint motion pattern and swing duration
shock absorbers (Schauder, 1952). (Table 6.3, Fig. 6.11), but the faster speed of the trot requires the
limb to perform the same movements in a shorter time. Similarities
in stance distance and swing time at different speeds and in differ-
The walk ent gaits (Table 6.4) have also been noted in cats and dogs (Grillner,
Like the trot, the walk is a symmetrical gait, but inter-limb coordina- 1975). The limbs cover a shorter distance over the ground during
tion is quite different in walking and trotting. The walk is a four-beat the swing phase at walk and there is less forelimb protraction. Also,
gait with a lateral sequence of limb placements and support the elbow and carpal joints are less flexed in the swing phase at
sequences that always include at least one fore and one hind foot walk, resulting in a longer pendulum and a slower forward motion
on the ground. The walk does not have a suspension phase and with less protraction. Apparently, the limbs have to be more loaded
walking mechanics are described in terms of an inverted pendulum during stance and more flexed during swing to enable faster
mechanism that implies an exchange between kinetic and potential locomotion.
energy as the body vaults over the supporting limb (see Chapter 19). The forelimb MCP joint extension pattern is quite different at
In the literature, forelimb stick figures and joint angle–time dia- walk and trot (Fig. 6.11). In walking, there are two extension peaks
grams for the walk have been reported for individual horses (Walter, that ‘melt’ together at the trot, which mirrors the vertical ground
1925; Krüger, 1937, 1938; Fleiss et al., 1984) and for groups of reaction force traces at walk and trot (Niki et al., 1982; Back et al.,
horses of various breeds (Back et al., 1996; Galisteo et al., 1996; 1996; Khumsap et al., 2002). Schryver et al. (1978) explained this
Hodson et al., 2000; Nicodemus & Holt, 2006; Martuzzi et al., phenomenon as a result of the transition from a double to a single
2007). In a comparison of the kinematics of walk (1.6 m/s) and limb support. Alexander and Jayes (1978) proposed that this phe-
trot (4.0 m/s), Back et al. (1996) showed that the two gaits were nomenon was related to mechanical properties of the distal limb

111
 6 Forelimb function

210 210

200
190
Angle (deg)

Angle (deg)
190

170
180

150 170
0 20 40 60 80 100 0 20 40 60 80 100

0.0 0.02
20 40 60 80 100
0.01
Moment (Nm/kg)

Moment (Nm/kg)
0.00
-0.2 20 40 60 80 100
-0.01

-0.02

-0.4 -0.03

1.0 0.04

0.0 0.02
20 40 60 80 100
Power (W/kg)

Power (W/kg)

-1.0 0.00
20 40 60 80 100

-2.0 -0.02

-3.0 -0.04

A Stance (%) B Swing (%)

Fig 6.10  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the distal interphalangeal joint during the stance phase (left) and swing phase (right) at the trot. The pastern segment (proximal and middle
phalanges) was treated as a rigid segment.

and presented a mathematical model that is capable of describing neck, which undergo larger excursions as velocity increases. The
both possibilities and the gradation between them. Computer sim- timing of these oscillations is such that the head is lowered as each
ulation experiments have shown that oscillations in the vertical GRF forelimb provides a braking force and raised as each forelimb pro-
have roughly (not exactly, since muscle stiffness depends on muscle vides propulsion. The resulting torque around the cervicothoracic
activation) the same frequency in walk and trot (van den Bogert, junction tends to rotate the trunk in the opposite direction, which
1989). Thus, in the trot only one peak occurs since the limb is lifted increases the braking effect as the head is lowered and increases the
before the second peak. propulsive effect as the head is raised (Khumsap et al., 2002). There-
Peak vertical ground reaction force is smaller at walk than trot fore, the larger head and neck excursions at faster walking velocities
(Ueda et al., 1981; Merkens & Schamhardt, 1994; Schamhardt & are a mechanism to increase braking in early stance and propulsion
Merkens, 1994), which is not surprising since a suspension phase in late stance, with the balance between braking and propulsion
is lacking in the walk. The vertical force has two peaks separated by being maintained regardless of velocity.
a slight dip, with the second peak being higher than the first in the During the walking stance phase, retraction of the forelimb is
forelimbs (Merkens & Schamhardt, 1988; Hodson et al., 2000). As associated with the predominant joint moment being on the
walking speed increases, the heights of both peaks increase. The caudal/palmar aspect of all forelimb joints except the shoulder, at
longitudinal ground reaction force initially has a braking effect, which the peak moment is considerably higher than at any other
followed by a propulsive effect, with the braking impulse being joint (Clayton et al., 2000b). The net moment moves to the cranial/
larger than the propulsive impulse in the forelimbs. Peak braking dorsal side of all the joints during protraction in early swing, then
and propulsive forces increase with walking velocity but the corre- moves to the caudal/palmar side in late swing as protraction slows
sponding impulses decrease (Khumsap et al., 2002) as a conse- and retraction begins.
quence of a large reduction in stance duration. The entire forelimb shows a net absorption of energy in both
The velocity-dependent increases in peak ground reaction forces stance and swing phases (Clayton et al., 2000a). The elbow pro-
are thought to be associated with cyclic movements of the head and vides most of the positive work and is the only joint that shows net

112
 Sagittal plane analysis of forelimb kinematics and kinetics

80 100 Fig 6.11  Mean joint angle–time diagrams of

Flex Shoulder Flex Elbow the forelimb joints of a group of horses walking
75 (1.6 m/s, red line) and trotting (4.0 m/s, green
80 line) on a treadmill. The joint angles are defined
as zero when the adjacent bone segments are
70
Angle (degrees)

Angle (degrees)
60 aligned: extension is negative; flexion is positive.
The red bar represents the duration of the
65 forelimb stance phase at walk. The green bar
40 represents the duration of the forelimb stance
60 phase at trot.
Reprinted from Buchner, H.H.F., Savelberg, H.H.C.M.,
20
55 Scharmhardt, 1994, Kinematics of treadmill versus
Walk Walk
Ext Trot Ext Trot
overground locomotion in horses, Veterinary Quarterly, 16,
50 0 sup 2, with permission from Taylor & Francis Ltd, http://
0 20 40 60 80 100 0 20 40 60 80 100 www.informaworld.com.
Time (% of stride) Time (% of stride)

100 40
Flex Carpus Flex Fore fetlock
80 20

60 0
Angle (degrees)

Angle (degrees)

40 -20

20 -40

0 -60
Walk Walk
Ext Trot Ext Trot
-20 -80
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

generation of energy over the entire stride. Bursts of energy genera- is the leading limb. Horses use a transverse sequence in which the
tion at the elbow occur on the extensor (caudal) aspect in early trailing and leading limbs are on the same side for the fore and
stance to maintain forward motion of the trunk as the braking hind limb pairs. The sequence of footfalls is trailing hind limb,
longitudinal force increases and on the flexor (cranial) aspect leading hind and trailing forelimbs together, then the leading fore-
during breakover to initiate swing phase protraction. Extension of limb. An aerial (suspension) phase usually follows lift-off of the
the carpus aligns the limb allowing it to act as a supportive strut leading forelimb.
during stance, but does not play an important role in either energy In the literature, the canter has been described at speeds ranging
absorption or generation. A small burst of positive work on the from 3.0 to 11.0 m/s under various circumstances overground and
flexor aspect of the carpus at the start of breakover is indicative of on a treadmill (Clayton, 1994, 1995; Corley & Goodship, 1994;
an active role in initiating breakover at the walk. The fetlock func- Deuel & Park, 1990; Back et al., 1997). Forelimb kinematic patterns
tions elastically to store and release strain energy during stance, during cantering have been illustrated using stick figures (Walter,
while the coffin joint acts as an energy damper through most of 1925; Krüger, 1937, 1938), and have been described in terms of
stance with a small burst of energy generation on the flexor aspect inter-limb timing variables (Deuel & Park, 1990; Clayton, 1993,
during breakover. An increase in walking velocity is not accompa- 1994; Nicodemus & Booker, 2007). Changes in temporal and linear
nied by a comparable increase in energy generation across the joints kinematics induced by training have been documented (Corley &
of the forelimb (Khumsap et al., 2002). Goodship, 1994).
In the swing phase, peak magnitudes of the net joint powers Back et al. (1997) compared the kinematics of the leading and
decrease in a proximal to distal sequence (Clayton et al., 2000a). trailing forelimbs of Dutch Warmbloods, cantering on a treadmill
The elbow is the only joint that has a positive power profile, which at a speed of 7.0 m/s (Fig. 6.12). The leading forelimb is more
is indicative of its role in driving limb protraction and retraction. protracted due to greater elbow flexion, whereas the trailing fore-
Power is generated by the elbow flexors in early swing to protract limb is retracted further due to greater caudal rotation of the
the distal limb and by the elbow extensors later in swing to scapula. Also, the MCP joint is more extended in the trailing fore-
reverse the direction of limb movement in preparation for ground limb suggesting that this limb is relatively more loaded than the
contact. The shoulder joint absorbs energy during swing. At the leading forelimb (Sloet et al., 1995), which is in accordance with
carpal and digital joints, net joint powers act to control joint force plate studies showing higher vertical forces in the trailing
motions that are driven by inertial forces. forelimb of horses ridden overground at the canter (Niki et al.,
1984; Merkens et al., 1993) (Table 6.4). In order to achieve the
larger range of protraction and retraction in the trailing forelimb,
The canter the carpal joint is more flexed in the swing phase to allow the distal
The canter is an asymmetrical gait in which the footfalls of the limb to rotate further in the same swing time. Greater carpal flexion
forelimb pair and hind limb pair occur as couplets. The first limb has been reported as an effect of training in young Thoroughbreds
of the couplet to contact the ground is the trailing limb, the second (Corley & Goodship, 1994) and Quarter Horses (Deuel, 1994).

113
 6 Forelimb function

Table 6.4  Forelimb kinematic variables for horses walking at 1.6 m/s, trotting at 4.0 m/s (Back et al., 1996) and cantering at 7.0 m/s (Back
et al., 1997)

Variable Walk Trot Trailing forelimb Leading forelimb

canter canter
Stride duration (s) 1.09 ± 0.06* 0.67 ± 0.03* 0.54 ± 0.02 0.56 ± 0.04

Stance duration (s) 0.69 ± 0.04* 0.27 ± 0.01* 0.18 ± 0.01 0.19 ± 0.02

Stance duration (%) 63.2 ± 1.7* 40.3 ± 1.7* 33.6 ± 1.4 34.5 ± 2.4

Swing duration (s) 0.40 ± 0.03 0.40 ± 0.03 0.36 ± 0.02 0.36 ± 0.03

Scapula
Angle of max protraction (°) 20.2 ± 1.6* 21.5 ± 1.5* 24.9 ± 2.1# 33.4 ± 3.2#
Angle of max retraction (°) −22.5 ± 1.7 −22.8 ± 1.5 −31.6 ± 1.8# −24.7 ± 1.7#
Max pro-/retraction range (°) 42.8 ± 2.5* 44.4 ± 1.7* 56.5 ± 1.6 58.1 ± 3.2

Shoulder joint
Angle of max extension (°) 53.8 ± 5.9 54.6 ± 4.2 83.7 ± 3.9 80.2 ± 4.4
Angle of max flexion (°) 66.9 ± 4.2* 69.9 ± 5.1* −0.5 ± 0.8 −2.0 ± 2.2

Range of motion (°) 13.1 ± 1.9 15.3 ± 1.8 25.8 ± 1.6# 23.9 ± 2.3#
Elbow joint
Angle of max extension (°) 23.4 ± 3.1* 24.5 ± 3.1* 20.1 ± 2.1# 23.5 ± 3.8#
Angle of max flexion (°) 75.2 ± 3.6* 84.7 ± 4.3* 81.9 ± 4.6# 91.2 ± 6.7#
Range of motion (°) 51.8 ± 3.6* 60.2 ± 4.0* 61.8 ± 4.8# 67.7 ± 5.6#

Carpal joint
Angle of max extension (°) 1.3 ± 2.3* −2.9 ± 2.6* −6.7 ± 3.6 −4.4 ± 2.8

Angle of max flexion (°) 69.9 ± 6.2* 87.9 ± 7.4* 83.0 ± 7.9 #
91.2 ± 8.3#
Range of motion (°) 68.7 ± 5.7* 90.8 ± 7.1* 89.7 ± 7.0 95.6 ± 8.1

Metacarpophalangeal joint
Angle of max extension (°) −41.1 ± 6.1* −55.9 ± 6.2* −62.4 ± 4.7 −59.0 ± 8.1

Angle of max flexion (°) 16.5 ± 6.4* 21.3 ± 6.8* 22.8 ± 7.3# 29.1 ± 6.1#
Range of motion (°) 60.2 ± 1.4* 80.6 ± 7.1* 85.2 ± 8.0 88.1 ± 7.5

Joint angles are defined zero when the adjacent bone segments are aligned: extension is negative; flexion is positive. Scapular angle is measured relative to the vertical;
protraction angles are positive, retraction angles are negative. Values are mean ± SD.
*Variables that differ significantly between walk and trot (p <0.05).
#
Variables that differ significantly between trailing and leading limbs at canter (p <0.05).
Reproduced from Back et al. (1996, 1997).

The gallop 1987). Stance duration has been reported to be longer in the leading
forelimb than the trailing forelimb (Leach et al., 1987), but Witte
The gallop is an asymmetrical gait in which the footfalls of both et al. (2006) did not find any differences between kinematic vari-
the fore and hind limbs occur as couplets, with an aerial (suspen- ables for trailing and leading forelimbs. Mean stance durations of
sion) phase following lift-off of the leading forelimb. The transition the forelimbs have been reported to be 131 ms and 77 ms, at speeds
from a slow speed canter to the higher speed gallop involves a dis- of 9.0 m/s and 17.0 m/s, respectively. Although the duration of the
sociation of the footfalls of the diagonal limb pair, such that the suspension phase is independent of speed, overlaps between the
leading hind contacts the ground before the trailing forelimb in the limbs decreases with speed and approaches zero at maximal speed
gallop. (Witte et al., 2006). The intervention of a suspension phase between
The function of the forelimbs during galloping is to support the lift-off of the trailing forelimb and contact of the leading forelimb
forehand as the body mass moves forward and to generate the forces has been described at very high galloping speeds in some horses
needed for braking and turning. As gallop speed increases, the trail- (Seder & Vickery, 2003).
ing and leading fore and hind limb pairs act more synchronously, Fatigue is associated with changes in limb coordination: the per-
so the gait becomes more like a bound or bunny hop (Leach et al., centage of stride between contacts of the leading hind and trailing

114
 Three-dimensional kinematics

90 100 Fig 6.12  Mean joint angle–time diagrams of

Re Scapula Flex Elbow the leading and trailing forelimbs for a group of
85 horses cantering on a treadmill at 7.0 m/s. Joint
80 angles are defined as zero when the adjacent
80
bone segments are aligned: extension is
Angle (degrees)

Angle (degrees)
75 60 negative; flexion is positive. The red bar
represents the duration of the stance phase in
70 the trailing forelimb, the green bar represents
40 the duration of the stance phase of the leading
65
forelimb. The red bar represents the stance
60 duration of the trailing forelimb, the green bar
20
represents the stance duration of the leading
55 Trailing Trailing forelimb.
Pro Leading Ext Leading
50 0 Reprinted from Back, W., Hartman, W., Schamhardt, H.C.,
0 20 40 60 80 100 0 20 40 60 80 100 Bruin, G. and Barneveld, A. (1995b) Kinematic response to
Time (% of stride) Time (% of stride) a 70-day training period in trotting Dutch Warmbloods.
Equine Vet. J. 18 (suppl.), 127–131, with permission from
100 40 the Equine Veterinary Journal.
Flex Carpus Flex Fore fetlock
80 20

60 0
Angle (degrees)

Angle (degrees)

40 -20

20 -40

0 -60
Trailing Trailing
Ext Leading Ext Leading
-20 -80
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

forelimbs decreases, the percentage of stride between contacts of the higher for horses with good forelimb motion and these horses were
two hind limbs and the two forelimbs increases, and the percentage 58% more likely to win a graded stakes race than horses with bad
of stride occupied by the suspension phase increases (Leach & forelimb motion.
Sprigings, 1979). It has been suggested that at the end of a race, the
fatigued Thoroughbred is at serious risk of developing chip fractures
of the carpal bones (Johnston et al., 1999) due to carpal hyperex-
Three-dimensional kinematics
tension, and Burn et al. (2006) have confirmed that carpal hyper-
extension increases with speed. Furthermore, fatigue of the DDF In two-dimensional kinematic studies, angular data are reported as
muscle may overload the SDF tendon and suspensory ligament flexion and extension in the sagittal plane with the estimated
increasing the risk of injury to the MCP joint and its supporting soft centers of joint rotation being used as landmarks for placement of
tissues (Butcher et al., 2007). skin markers. Measurements are limited to one degree of freedom
In Quarter Horses galloping at 13.1 m/s, it was observed that in rotation. This type of analysis ignores abduction–adduction and
temporal patterns of forelimb motion originated proximal to the axial rotation of the joints and the role of movements occurring
elbow and there was a high degree of reciprocity in shoulder joint outside the sagittal plane in the pathogenesis of injuries cannot be
action in the trailing and leading forelimbs. In the stance phase, explored.
maximal carpal extension preceded maximal elbow extension, An alternative method is to establish a three-dimensional joint
while in the swing phase, maximal flexions occurred sequentially coordinate system, based on the axes of the bone segments, which
at the carpus, shoulder and elbow joints (Deuel, 1994). are independent of the joint centers of rotation. Establishment of a
In the Thoroughbred, forelimb motion in the swing phase has joint coordinate system requires the definition of orthogonal frames
been correlated with racetrack performance and earnings by mea- that are rigid with the bones and numerically described with an
suring metacarpal angulation in the sagittal and frontal planes orientation matrix by the use of a minimum of three, non-colinear
(Seder & Vickery, 2005). In the sagittal plane, peak angulation of kinematic markers per segment (for details, see Chapters 2 and 19).
the metacarpus to the ground during the swing phase retraction was With this method the three angles of rotation of the joints can be
graded on a scale of 1–5. Grade 1 was forelimbs minimally lifted described.
with little rotation of the metacarpus, which is commonly referred In the equine forelimb, in vivo three-dimensional kinematics have
to as a ‘daisy cutter action’. Grade 5 was excessive metacarpal rota- been reported on a hard surface at walk (Chateau et al., 2004;
tion with the hoof hitting the elbow, which is described as ‘high Hobbs et al., 2006) and trot (Clayton et al., 2004, 2007a, b) and
knee action’. In the frontal plane, abduction (‘winging’) and adduc- also at trot on a treadmill (Chateau et al., 2006).
tion (‘paddling’) of the distal limb was graded on a scale of 1–5 In Chateau et al. (2004) and Hobbs et al. (2006), joint rotations
with 1 being very little lateral motion in the forelimbs (i.e. almost were calculated by use of a cardan sequence x, y, z, also known as
vertical) and 5 being excessive lateral motion in the forelimbs. the joint coordinate system (JCS) introduced by Grood and Suntay
Horses with good forelimb motion (low scores) earned more purse (1983). With this method, flexion-extension is calculated around
money and had greater stakes level success than horses with bad the transverse axis of the proximal segment of the joint, axial
forelimb motion (high scores). Median earnings per start was 83% rotation around the vertical axis of the distal segment, and

115
 6 Forelimb function

abduction–adduction around a floating axis perpendicular to the In the digital joints, abduction and adduction occur passively
two other axes (Fig. 6.13). This technique, which follows the recom- rather than being generated by abductor or adductor muscles.
mendations of the International Society of Biomechanics (Wu & Denoix (1999) introduced the term collateromotion to describe these
Cavanagh, 1995), clearly defines the axes around which rotations passive movements: lateromotion is passive abduction and medio-
are expressed and preserves the link with clinical and physiological motion is passive adduction.
terminology. Clayton et al. (2004, 2007a, 2007b) used a slightly The majority of equine kinematic studies have been based on
different method in which relative angular motions (helical angle tracking markers attached to the skin, which provide a simple and
changes) between the segments were calculated using a spatial atti- non-invasive method of visualizing joint motion (Clayton, 2007b),
tude method (Woltring, 1994). This method, which is also well but large errors may arise due to soft tissue movement over the
known and recognized in human biomechanics, calculates the underlying bone (Reinschmidt et al., 1997; van Weeren et al., 1986;
angles around slightly different axes and this can explain small dif- Crevier-Denoix et al., 2001). In three-dimensional kinematics,
ferences between the results of different studies. those errors are crucial because they counteract the rigid-body
theory (Fuller et al., 1997). Besides, skin markers are difficult to use
in the distal forelimb because of the small size of P1 and P2. Indeed,
those phalanges have usually been modeled as a single, rigid
segment, and any motion at the PIP joint has been incorporated
into the kinematics of the MCP and DIP joints. To overcome the
problems inherent in using surface markers, bone-fixed markers can
be used for direct measurement of skeletal motion. This technique
provides the most accurate means for determining bone movements
(Ramsey & Wretenberg, 1999). Several kinematic studies (Ramsey
Lateromotion et al., 2001; Lafortune et al., 1992; Reinschmidt et al., 1997) have
Floating axis (XI) (+)
implanted bone pins in human subjects, none of whom reported
pain or substantial discomfort during the experiments. Bone pins
have also been used in horses (van Weeren et al., 1986; Lanovaz
et al., 2002; Dyhre-Poulsen et al., 1994). Steinmann pins implanted
Mediomotion (-) in a 6 mm canal drilled through the bone cortex, do not have a
significant effect on locomotion of horses (van Weeren et al., 1986).
More recently, Chateau et al. (2004, 2005, 2006) and Clayton et al.
(2004, 2007a, 2007b) have used markers rigidly fixed to the radius,
third metacarpal bone, P1, P2 and the hoof wall to characterize
three-dimensional rotations of the carpal, MCP, PIP and DIP joints.
Development of non-invasive methods for quantifying three-
dimensional motion would be advantageous in a clinical setting.
Extension Hobbs et al. (2006) used plaster casts to attach marker clusters to
the skin and were able to measure flexion–extension motions,
including the PIP joint, with an acceptable degree of accuracy, but
not internal–external rotation, which highlights the difficulties in
quantifying extrasagittal movements non-invasively (Hobbs et al.,
Yp 2006).

Medial rotation (-) Flexion Three-dimensional forelimb kinematics for

locomotion at walk and trot in a straight line
Lateral rotation (+)
Distal interphalangeal joint
Using three-dimensional kinematics, the pattern of flexion and
extension of the DIP joint at the walk and trot (Figs 6.14 and 6.15)
has been shown to be very similar to two-dimensional kinematic
data (Chateau et al., 2004, 2006; Clayton et al., 2007b). At both
gaits, the DIP joint undergoes rapid flexion during impact. After
hoof stabilization, the DIP joint continues to flex but at a slower
rate. Maximal flexion of the DIP joint precedes maximal extension
of the MCP joint. The DIP joint then extends in the second half of
stance.
When the horse moves in a straight line, small amounts of
Zd extrasagittal motion of the DIP joint have been measured (Table
Fig 6.13  Definition of a joint coordinate system (example of the left 6.6), mostly during landing and breakover, when the transverse
metacarpophalangeal joint). The flexion–extension angle is calculated orientation of the foot is changing. At the beginning of the stance
around the transverse axis of the proximal segment, the axial rotation angle phase, the limb is globally adducted, resulting in the lateral aspect
around the longitudinal axis of the distal segment, and the of the hoof being closer to the ground and making first contact with
abduction–adduction angle (collateromotion) around a floating axis the ground. Changes in the mediolateral orientation of the hoof
perpendicular to the two other axes. during landing unequivocally affect extrasagittal motions of the
Reprinted from Degueurce, C., Chateau, H., Pasqui-Boutard, V., Pourcelot, P., Audigié, F., digital joints. As the hoof rocks medially, the DIP joint undergoes
Crevier-Denoix, N., Jerbi, H., Geiger, D. and Denoix, J.-M., 2000, Concrete use of the joint medial rotation associated with lateromotion that narrows the
coordinate system for the quantification of articular rotations in the digital joints of the articular space laterally on the side of the hoof that hits the ground
horse, Veterinary Rsearch, 31, 297–311, with permission from EDP Sciences. first (Chateau et al., 2004). These findings highlight a major

116
 Three-dimensional kinematics

Pitch Roll Yaw

IMP HS MS CH CH TO IMP HS MS CH CH TO IMP HS MS CH CH TO
50 4 5

Lateral
Backward

Medial
45 2 0
0
Hoof

40 -5
-2
35 -4 -10

Forward

Lateral
Medial
30 -6 -15
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

IMP HS MS CH CH TO IMP HS MS CH CH TO
30 10

Backward

Adduction
20 8
10 6 WALK
4 CIRCLE mean se
Mc3

0
2
-10 0 STRAIGHT LINE mean se

Abduction
Forward

-20 -2
-30 -4
0 20 40 60 80 100 0 20 40 60 80 100

Flexion/Extension Collateromotion Axial rotation

IMP HS MS CH CH TO IMP HS MS CH CH TO IMP HS MS CH CH TO
170 2

Lateral
Flexion

Lateromotion
165 0 0
160 -2
155 -2
-4
MCPJ

150 -4
145 -6

Mediomotion
-6
-8
Extension

140
-8

Medial
135 -10
130 -10
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

IMP HS MS CH CH TO IMP HS MS CH CH TO IMP HS MS CH CH TO

200

Lateral
Extension

Lateromotion

4 2
198
2 0
196
0 -2
PIPJ

194
-2 -4
Mediomotion

192
-4 -6
Flexion

190

Medial
-6 -8
188
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

IMP HS MS CH CH TO IMP HS MS CH CH TO IMP HS MS CH CH TO

210 2 Lateral
Extension

Lateromotion

8
200 0
6
190 -2
4
DIPJ

180 -4
2
Mediomotion

170 -6
0
-8
Flexion

160
Medial

-2
150 -10
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

Time (% stance)
Fig 6.14  Mean attitude angle and joint angle–time diagrams of the left distal forelimb of a group of four horses walking in a straight line and turning
sharply to the left. Angles (°) are plotted against time (expressed as % of the stance phase). Mc3, third metacarpal bone; MCPJ, metacarpophalangeal joint;
PIPJ, proximal interphalangeal joint; DIPJ, distal interphalangeal joint; IMP, impact; HS, hoof stabilization; MS, midstance; CH, heel-off; TO, toe-off. Note that for
some curves the standard error bars are superimposed on the mean values.
Reprinted from Chateau, H., Degueurce, C. and Denoix, J.-M. (2005) Three-dimensional kinematics of the equine distal forelimb : effects of a sharp turn at the walk. Equine vet. J. 37, 12–18,
with permission from the Equine Veterinary Journal.

117
 6 Forelimb function

Pitch Roll Yaw

IMP HS MS HO TO IMP HS MS HO TO IMP HS MS HO TO
10 4

Backward

Lateral

Medial
40 8 2
6 0
30 4
Hoof

-2
20 2
0 -4

Forward
-6

Lateral
Medial
10 -2
-4 -8
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

IMP HS MS HO TO IMP HS MS HO TO
30 10

Backward

Adduction
20
10 5
0 TROT (4m/s on a treadmill) mean se
Mc3

-10
-20 0

Abduction
Forward

-30
-40 -5
0 20 40 60 80 100 0 20 40 60 80 100

Flexion/Extension Collateromotion Axial rotation

IMP HS MS HO TO IMP HS MS HO TO IMP HS MS HO TO
170
Flexion

Lateromotion

Lateral
0
160 -2
-2
150 -4
MCPJ

-4
140 -6
Mediomotion
-6
Extension

130 -8

Medial
120 -8
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

IMP HS MS HO TO IMP HS MS HO TO IMP HS MS HO TO

205
Extension

Lateromotion

Lateral
2 0
200 0 -2
PIPJ

195 -2 -4
Mediomotion

190 -4 -6
Flexion

Medial
185 -6 -8
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

IMP HS MS HO TO IMP HS MS HO TO IMP HS MS HO TO

200
Extension

Lateromotion

190 4 0 Lateral
180
2 -2
DIPJ

170
0 -4
Mediomotion

160
-6
Flexion

150 -2
Medial

140
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

Time (% stance)
Fig 6.15  Mean attitude angle and joint angle–time diagrams of the left distal forelimb in a group of three horses trotting on a treadmill (4 m/s). Angles (°)
are plotted against time (expressed as % of the stance phase). Mc3, third metacarpal bone; MCPJ, metacarpophalangeal joint; PIPJ, proximal interphalangeal
joint; DIPJ, distal interphalangeal joint; IMP, impact; HS, hoof stabilization; MS, midstance; CH, heel-off; TO, toe-off). Note that for some curves the standard
error bars are superimposed on the mean values.
Reprinted from Chateau H, Degueurce C, Denoix J-M. (2006) Three-dimensional of the distal forelimb in horses trotting on a treadmill and effects of elevation of heel and toe. Equine vet. J. 38,
164–169, with permission from the Equine Veterinary Journal.

118
 Three-dimensional kinematics

Table 6.5  Mean value ± SD for range of motion in flexion and Table 6.6  Three-dimensional motion of the hoof, third
extension of the digital joints of the left forelimb of four horses metacarpal bone and digital joints in three horses trotting on
walking in a straight line on a hard track at 1.3 m/s (Chateau a treadmill at 4.0 m/s (Chateau et al., 2006). Values are mean
et al., 2004) and three horses trotting on a treadmill at 4.0 m/s and (SD)
(Chateau et al., 2006)
Landing Weight Breakover
Walk Trot bearing
MCPJ ext −27.6 ± 3.3 −45.7 ± 7.0 Hoof Pitch 1.7 (4.8) 0.1 (0.9) −20.3 (6.7)
MCPJ flex 24.9 ± 3.1 41.2 ± 8.2 Roll −8.9 (4.4)* −1.2 (1.0) 3.4 (1.6)*
PIPJ flex −0.9 ± 0.9 −1.7 ± 2.6
Yaw 0.9 (2.7) −0.4 (1.2) −0.7 (3.0)
PIPJ ext 10.0 ± 2.5 14.3 ± 2.8
Mc3 Pitch −6.0 (1.8) −38.1 (3.8) −13.5 (4.3)
DIPJ flex −16.4 ± 3.1 −21.7 ± 3.8
MCPJ Flex-Ext −25.6 (3.7)* 2.8 (5.5) 18.3 (6.7)
DIPJ ext 42.4 ± 4.4 45.6 ± 5.5
Collat 0.8 (1.7)* 1.0 (1.5)* 0.5 (2.5)
MCPJ, metacarpophalangeal joint; PIPJ, proximal interphalangeal joint; DIPJ, distal −0.4 (1.0)
Axial rot 1.2 (1.6)* −1.1 (1.2)*
interphalangeal joint.
Reproduced from Chateau et al. (2004, 2006). PIPJ Flex-Ext 1.3 (2.6)* 11.5 (2.7)* −0.9 (1.4)*
Collat −0.4 (2.0) −0.2 (1.4) 0.2 (1.4)
function of the digital joints in compensating for asymmetrical hoof
Axial rot −1.5 (2.1)* 1.7 (1.6)* 0.2 (0.5)
placement. Interestingly, medial rotation and lateromotion of the
DIP joint were coupled, which has been observed during asymmet- DIPJ Flex-Ext −19.3 (5.0)* 31.3 (7.1)* 11.9 (5.8)*
ric loading of isolated forelimbs in other studies (Denoix, 1999;
Chateau et al., 2002). When a heel wedge was applied unilaterally, Collat 1.4 (3.3)* 0.5 (2.3) −0.5 (1.6)
the DIP joint showed collateromotion with narrowing of the joint Axial rot −1.7 (3.0) −2.3 (2.9)* 1.3 (1.5)*
space on the side of the wedge, together with axial rotation away
from the raised side. The same association of movement has also *Values differ significantly (p <0.05) between the beginning and the end of the
been reported at the start of breakover with a combination of lat- period.
eromotion (abduction) and medial (internal) rotation (Chateau Mc3, third metacarpal bone; MCPJ, metacarpophalangeal joint; PIPJ, proximal
et al., 2004; Clayton et al., 2007b). interphalangeal joint; DIPJ, distal interphalangeal joint.
Landing is from impact to hoof stabilization; Weight bearing is from hoof
stabilization to heel-off; Breakover is from heel-off to toe-off.
Proximal interphalangeal joint
For segment angles, pitch angle around the transverse axis is positive during
The PIP joint (Figs 6.14 and 6.15, Table 6.5) flexes in early stance backward (palmar or plantar) rotation; roll angle around the longitudinal axis is
reaching maximal flexion at about 14% of stance (Chateau et al., positive during lateral inclination; and yaw angle around the vertical axis is positive
2004; Clayton et al., 2007b). This movement is brief and slight in during medial rotation of the segment. For joint angles, lateromotion (passive
walking horses (less than 1°) and almost comparable at the trot abduction) and lateral rotation are designated positive.
(about 1.7°). Then, the PIP joint extends through most of stance Reprinted from Chateau H, Degueurce C, Denoix J-M. (2006) Three-dimensional of
until the start of breakover. Chateau et al. (2004, 2006) reported the distal forelimb in horses trotting on a treadmill and effects of elevation of heel
the mean amplitude of this extension to be 10° ± 2.5° at walk and and toe. Equine vet. J. 38, 164–169, with permission from the Equine Veterinary
14.3° ± 2.8° at trot, while Clayton et al. (2007b) reported 9.4° ± Journal.
1.7° at walk and 8.5° ± 0.6° at trot. The results confirm that flexion
and extension of the PIP joint cannot be neglected in biomechani-
cal models of the distal portion of the forelimbs. In horses walking in a straight line. The three-dimensional motions of the PIP joint,
in a straight line, the PIP joint accounts for approximately 5% of especially axial rotation, coupled with asymmetric landing of the
the entire interphalangeal flexion (0.9° of PIP joint flexion vs. 16.4° hoof at impact, should be taken into account, even when a horse
of DIP joint flexion) and 19% of the entire interphalangeal exten- is walking in a straight line, to explain concussion of the joints and
sion (10° of PIP joint extension vs. 42.4° of DIP joint extension). pain that may result from mediolateral imbalances.
The substantial involvement of the PIP joint in digital extension
could explain why arthrodesis is sometimes followed by a worsen-
ing of navicular conditions (Martin et al., 1984), since loss of exten- Metacarpophalangeal joint
sion following PIP joint arthrodesis is likely to be compensated During the stance phase, the MCP joint (Figs 6.14 and 6.15, Table
primarily by the DIP joint. 6.5) undergoes extension and then flexion. As already observed in
For locomotion in a straight line, extrasagittal motions of the PIP two-dimensional kinematic studies at the walk, extension is pro-
joint are minimal (Figs 6.14 and 6.15, Table 6.6). Chateau et al. longed with a tendency toward two peaks separated by a slight dip,
(2004) did not detect a significant pattern in collateromotion for whereas in trot there is a single cycle of extension. No particular
the PIP joint. Internal rotation of P2 relative to P1 occurs during pattern of collateromotion was detected for the MCP joint during
landing, concurrent with the medial rocking motion of the hoof stance (Chateau et al., 2004) but P1 twisted laterally relative to the
(Chateau et al., 2004; Clayton et al., 2007b). During breakover, third metacarpal bone during extension and medially during
lateral rotation of the PIP joint occurs in association with lateral flexion. This is in contrast to the study of Clayton et al. (2007a), in
rocking motion and lateral rotation of the hoof (Chateau et al., which all horses showed lateromotion during stance. The difference
2004). between studies is a consequence of the alignment of the segmental
Extrasagittal movements of the PIP joint, like those of the DIP coordinate systems. If the coordinate system of P1 used by Clayton
joint, occur mainly during landing and breakover when horses walk et al. (2007a) is rotated externally around its longitudinal axis by

119
 6 Forelimb function

80 swing; and axial rotation 4° ± 0° in stance and 11° ± 7° in swing

(Clayton et al., 2004).
60
Flexion(+)/extension(-) (deg)

Three-dimensional forelimb kinematics when

40 turning sharply at walk
Three-dimensional kinematics of the distal forelimb reveal small
20 amounts of collateromotion and axial rotation when the horse
moves in a straight line (Table 6.6). During turning, especially on
a hard surface, frontal plane alignment between the hoof and the
0
20 40 60 80 100 third metacarpus is disrupted and the digital joints must compen-
sate for this misalignment. Pathogenesis of several digital injuries
-20 is believed to be connected with these extrasagittal movements and
examination of the horse on a circle at the walk and trot is useful
8 to further characterize a lameness because horses affected with spe-
cific digital injuries are often more lame with the involved limb on
the inside of the circle (Swanson, 1988).
Adduction(+)/abduction(-) (deg)

6 Motions outside the sagittal plane in the digital joints have been
studied ex vivo during asymmetric loadings of isolated forelimbs.
4 Intra-articular injection of a colored solution was used to evaluate
variations in the contact areas between articular surfaces as a
consequence of asymmetrical hoof placement (Denoix, 1999).
2
Radiographic assessment has also been proposed (Caudron et al.,
1998), but this method was limited to the measurement of com-
0 bined movements between the PIP and DIP joints and carries a risk
20 40 60 80 100 of misinterpretation due to the geometrical projection on a single
-2 plane. Changes in the transverse orientation of the foot quantified
ex vivo using three-dimensional kinematics (Chateau et al., 2001,
2002) have shown that asymmetric placement of the foot induced
10 a collateromotion (narrowing of the articular space) in the direction
of the elevated part of the foot and an axial rotation in the opposite
8
Int(+)/ext(-) rotation (deg)

direction in the digital joints. However, in vitro studies or quasi-

6 static radiographic studies cannot merely be extrapolated to the
moving horse.
4 More recently, three-dimensional motion of the digital joints has
been measured in vivo as horses turned sharply at the walk (Fig.
2 6.14, Table 6.7) (Chateau et al., 2005). In this study, four horses
equipped with bone-fixed markers on the left forelimb were led in
0
20 40 60 80 100 hand around a left turn approximately 1.5 m in diameter with the
-2 instrumented left forelimb on the inside of the turn. The three-
dimensional kinematics, calculated using the principle of the JCS,
Time (% stride) were compared with data from the same horses walking in a straight
Fig 6.16  Mean joint angle–time graphs for three-dimensional carpal joint
line (Table 6.7, Fig. 6.14).
motion of three horses trotting overground at 3.1 m/s.
Reprinted from Clayton H.M., Sha D.H;, Stick J.A., Mullineaux D.R. (2004) Three dimensional Landing (from first impact to hoof stabilization)
kinematics of the equine carpus at trot. Equine vet. J. 36: 671–676, with permission from the
For motion in a straight line, the four horses studied by Chateau
Equine Veterinary Journal.
et al. (2005) showed global adduction of the distal limb at impact,
which lowered the lateral aspect of the hoof and resulted in initial
10°, there is close agreement in collateromotion between the two contact on the lateral side, followed by a medial rocking motion.
studies. When moving through a turn, the inclination of the third metacar-
pal bone in the transverse plane (roll angle) shows that the distal
part of the limb on the inside of the turn is substantially more
Carpal joint abducted at impact and landing is flat-footed, with little subsequent
In horses trotting at 3.1 m/s, three-dimensional carpal kinematics rocking motion of the hoof.
show low variability between strides within a horse but greater vari- When the hoof lands during locomotion in a straight line, the
ability between horses, especially in the amount of carpal flexion DIP joint undergoes a lateromotion and a medial rotation associ-
at the instant of lift-off which influences the amount of flexion– ated with the medial rocking motion of the hoof. During flat-footed
extension assigned to the stance and swing phases (Clayton et al., landing in a turn, however, the DIP joint underwent no significant
2004). In early stance, the carpus extends and the metacarpus movement of collateromotion or axial rotation. At the PIP joint,
rotates internally relative to the radius as the joint assumes its close- internal rotation occurs during landing when moving a straight line
packed position (Fig. 6.16). This position is maintained with only but not when turning.
minor oscillations until just before lift-off, when the carpus begins
a cycle of flexion combined with a little internal rotation that con- Bearing phase (from hoof stabilization
tinues until midswing (Fig. 6.16). The direction of abduction–
adduction varies between horses. Ranges of carpal motion during
to heel-off)
trotting are: flexion–extension 15° ± 6° in stance and 76° ± 13° in In a turn, the inside forelimb is abducted at the beginning of the
swing; abduction–adduction 5° ± 1° in stance and 13° ± 6° in stance phase because the horse places this limb toward the center

120
 Three-dimensional kinematics

Table 6.7  Three-dimensional motion of the hoof, third metacarpal bone and digital joints in four horses walking in a straight line and
around a small circle (Chateau et al., 2005). Values are mean and (SD)

Landing (IMP–HS) Weight bearing (HS–HO) Breakover (HO–TO)

Straight line Circle Straight line Circle Straight line Circle

Hoof Pitch −4.1 (5.4)* 0.4 (6.0) #
0.0 (1.0) 0.4 (1.1) −9.8 (2.9)* −8.2 (4.1)*

Roll −6.2 (2.7)* 0.7 (3.8)# −0.6 (0.9) 0.2 (0.8) 1.9 (1.3)* 4.6 (2.8)*#

Yaw 0.9 (3.7) −2.4 (4.2) −0.1 (0.9) −0.1 (2.6) −1.3 (4.5) −11.3 (8.6)*#
Mc3 Pitch −3.7 (1.7)* −3.5 (1.0)* −41.0 (4.4)* −33.6 (3.7)*# −5.8 (1.4)* −7.7 (3.0)*#
Roll 1.2 (1.2)* −0.6 (1.1)# −7.5 (5.5)* 7.3 (5.2)*# −1.5 (1.0)* −2.7 (1.8)*#
MCPJ Flex-Ext −10.9 (4.6)* −12.2 (5.1)* −0.1 (3.4) −7.4 (5.9)*# 9.0 (1.9)* 13.4 (3.4)*#
Collat 0.2 (1.7) −0.1 (0.7) 0.3 (0.8) 0.0 (0.7) −0.3 (0.9) 0.0 (1.0)
Axial rot 0.2 (0.9) 1.3 (0.7)* #
0.2 (1.3) −1.1 (0.5)* #
−0.3 (0.8) 0.2 (0.6)

PIPJ Flex-Ext 0.9 (1.8) 0.9 (4.2) 9.2 (2.4)* 6.5 (2.1)* #
−0.2 (0.6) 1.3 (0.9)*#
Collat −0.6 (1.5) −0.2 (0.9) 0.4 (1.1) −0.3 (0.8) 0.5 (0.7)* 0.8 (0.6)*
Axial rot −2.2 (1.1)* −0.5 (1.3)# 1.1 (1.7)* −4.0 (1.7)*# 0.6 (0.8)* 2.9 (0.8)*#
DIPJ Flex-Ext −12.1 (2.2)* −9.3 (4.8)*# 33.1 (3.4)* 18.4 (5.9)*# 5.1 (2.5)* 12.3 (4.4)*#
Collat 0.9 (2.4)* −0.4 (2.2)# −1.9 (1.6)* 2.0 (1.8)*# −0.3 (0.8) 0.2 (1.1)

Axial rot −1.7 (1.6)* 1.0 (2.0)# −2.2 (2.5)* −10.2 (3.9)*# 1.5 (1.6)* 3.6 (1.3)*#

*Values differ significantly (p <0.05) between the beginning and the end of the period.
#
ROM differs significantly (p <0.05) from the straight line.
Mc3, third metacarpal bone; MCPJ, metacarpophalangeal joint; PIPJ, proximal interphalangeal joint; DIPJ, distal interphalangeal joint; IMP, impact; HS, hoof stabilization; HO,
heel-off; TO, toe-off.
Landing is from impact to hoof stabilization; Weight bearing is from hoof stabilization to heel-off; Breakover is from heel-off to toe-off.
For segment angles, pitch angle is positive during backward rotation, roll angle is positive during lateral inclination and yaw angle is positive during medial rotation of the
segment. For joint angles, lateromotion (passive abduction) and lateral rotation are designated positive.
Reprinted from Chateau, H., Degueurce, C. and Denoix, J.-M. (2005) Three-dimensional kinematics of the equine distal forelimb : effects of a sharp turn at the walk. Equine vet. J.
37: 12–18, with permission from the Equine Veterinary Journal.

of the circle. During the rest of stance, the inside forefoot is immo- the stance phase do not have the same biomechanical consequences
bile on the ground and the horse’s body mass moves over the limb on the three-dimensional motion of the interphalangeal joints and
in the direction of the turn. The third metacarpal bone undergoes thus on the stress imposed on the collateral ligaments.
large movements relative to the hoof in the transverse plane Alterations induced by limb adduction also induce a substantial
(Chateau et al., 2005) showing an adduction that is maximal when involvement of the PIP joint (Chateau et al., 2005), which under-
the heels leave the ground (Fig. 6.14). goes medial rotation during the bearing phase in a turn. At heel-off,
Because the third metacarpal bone is more abducted at the begin- the PIP joint contributes 33% of the total amount of axial rotation
ning of stance in the limb on the inside of a turn, the DIP joint in the digital joints during a sharp turn compared with 10% for the
initially shows a combination of mediomotion (adduction) and MCP joint and 57% for the DIP joint. The amount of axial rotation
lateral rotation. Adduction of the limb during the stance phase with was greater than the amount of flexion at the DIP joint during sharp
the hoof flat on the ground, reverses the direction of these move- turns, which demonstrates the strong involvement of this joint in
ments and the DIP joint undergoes a combination of lateromotion movements that occur outside the sagittal plane. These results cor-
(2.0 ± 1.8) and medial rotation (10.2 ± 3.9) (Chateau et al., 2005). roborate clinical observations related to pathological conditions of
During this movement, the articular space becomes pinched later- this joint showing that degenerative diseases of the PIP joint occur
ally and P3 rotates medially relative to P2. In other words, collat- most frequently in horses that make tight turns and rapid twisting
eromotion occurs in the direction of movement, while axial rotation movements (Stashak, 1987; Schaer et al., 2001). Injuries include
occurs in the opposite direction. Interestingly this combination of pulling or tearing of the attachments of the joint capsule and col-
articular rotations is the same as those observed in vitro during lateral ligaments (McIlwraith & Goodman, 1989). Thus decreased
asymmetric loadings of the limb (Caudron et al., 1998; Denoix, range of motion following PIP arthrodesis may exacerbate stresses
1999; Chateau et al., 2002; Viitanen et al., 2003). on the DIP joint and navicular apparatus and explain why long-
The changes in direction of DIP joint angulation are easily seen term radiographic evaluation after PIP joint arthrodesis often
on the angle-time diagrams when the curves for collateromotion reveals degenerative disease involving the DIP joint or even navicu-
and axial rotation intersect (Fig. 6.14). The effects of these angular lar disease (Martin et al., 1984; MacLellan et al., 2001). The biome-
changes should be considered during examination of a lame horse chanical importance of the PIP joint should be kept in mind when
on a tight circle at the walk because the cranial and caudal parts of arthrodesis is considered as a treatment option and particular

121
 6 Forelimb function

attention should be paid to mediolateral hoof balance for manage- ground during breakover. Average amplitude of this movement
ment of the treated horses (Clayton et al., 2007b). when walking around a sharp turn on a hard surface is about 11°.
This movement of the hoof contributes to the sudden realignment
of the interphalangeal joints that were strongly medially rotated just
Breakover (from heel-off to toe-off) before the start of breakover (Fig. 6.14).
The results of Chateau et al. (2005) show that, during turning, Delaying breakover is likely to exacerbate extrasagittal stresses in
extrasagittal stress in the interphalangeal joints is maximal just the interphalangeal joints, so prevention and treatment strategies
before the start of breakover because adduction of the limb is for traumatic or degenerative diseases of the interphalangeal
maximal at heel-off. Unweighting of the palmar part of the foot joints should take into account the biomechanical influence on
allows the heels to lift-off and the hoof to rotate laterally on the breakover.

References
Alexander, R.M., 2002. Tendon elasticity and requirements. Int. J. Exp. Path. 88, Audigié, F., et al., 2001. Normal three-
muscle function. Comp. Biochem. Physiol. 241–248. dimensional behaviour of the
Part A-Molecular And Integrative Birch, H.L., Wilson, A.M., Goodship, A.E., metacarpophalangeal joint and the effect of
Physiology 133, 1001–1011. 1997. The effect of exercise-induced uneven foot bearing. Equine Vet. J. 33
Alexander, R.McN., Jayes, A.S., 1978. Vertical localized hyperthermia on tendon cell (Suppl.), 84–88.
movements in walking and running. J. survival. J. Exp. Biol. 200, 1703–1708. Chateau, H., Degueurce, C., Jerbi, H.,
Zool. 185, 27–40. Brown, N.A.T., Kawcak, C.E., McIlwraith, C.W., Crevier-Denoix, N., Pourcelot, P., Audigié,
Alexander, R.McN., Trestik, C.L., 1989. Bistable Pandy, M.G., 2003. Architectural properties F., et al., 2002. Three-dimensional
properties of the hock joint of horses of distal forelimb muscles in horses, equus kinematics of the equine interphalangeal
(Equus spp.). J. Zool. 218, 383–391. caballus. J. Morph. 258, 106–114. joints: articular impact of asymmetric
Aoki, O., Tokuriki, M., Kurakawa, Y., Hataya, Buchner, H.H.F., Savelberg, H.H.C.M., bearing. Vet. Res. 33, 371–382.
M., Kita, T., 1984. Electromyographic Schamhardt, H.C., Barneveld, A., 1997. Clayton, H.M., 1993. The extended canter: a
studies on supraspinatus and infraspinatus Inertial properties of Dutch warmblood comparison of the stride kinematics of
muscles of the horse with or without a horses. J. Biomech. 30, 653–658. horses trained for dressage and for racing.
rider in walk, trot and canter. Bull. Equine Burn, J.F., Portus, B., Brockington, C., 2006. Acta Anat. 146, 183–187.
Res. Inst. 22, 100–104. The effect of speed and gradient on Clayton, H.M., 1994. Comparison of the stride
Back, W., Barneveld, A., Schamhardt, H.C., hyperextension of the equine carpus. Vet. J. kinematics of the collected, working,
Bruin, G., Hartman, W., 1994. Longitudinal 171, 169–171. medium and extended canter in horses.
development of the kinematics in 4-, 10-, Butcher, M.T., Hermanson, J.W., Ducharme, Equine Vet. J. 17 (Suppl.), 16–19.
18-, and 26-month-old Dutch warmblood N.G., Mitchell, L.M., Soderholm, L.V., Clayton, H.M., 1995. Comparison of the stride
horses. Equine Vet. J. 17 (Suppl.), 3–6. Bertram, J.E.A., 2007. Superficialdigital kinematics of the collected, medium, and
Back, W., Hartman, W., Schamhardt, H.C., flexor tendon lesions in racehorses as a extended walks in horses. Am. J. Vet. Res.
Bruin, G., Barneveld, A., 1995b. Kinematic sequela to muscle fatigue: a preliminary 56, 849–852.
response to a 70-day training period in study. Equine Vet. J. 39, 540–545. Clayton, H.M., Hodson, E.F., Lanovaz, J.L.,
trotting Dutch warmbloods. Equine Vet. J. Camp, C.L., Smith, N., 1942. Phylogeny and 2000a. The forelimb in walking horses: 2.
18 (suppl.), 127–131. functions of the digital ligaments of the Net joint moments and joint powers.
Back, W., Schamhardt, H.C., Barneveld, A., horse. Mem Univ Calif Press, 13, pp. Equine vet. J. 32, 295–299.
1996. Are kinematics of the walk related to 69–124. Clayton, H.M., Lanovaz, J.L., Schamhardt,
the locomotion of a warmblood horse at Carrier, D.R., Deban, S.M., Fischbein, T., 2006. H.C., Willemen, M.A., Colborne, G.R.,
the trot? Vet. Q. 18, S79–84. Locomotor function of the pectoral girdle 1998. Net joint moments and powers in
Back, W., Schamhardt, H.C., Barneveld, A., ‘muscular sling’ in trotting dogs. J. Exp. the equine forelimb during the stance
1997. Kinematic comparison of the leading Biol. 209, 2224–2237. phase at the trot. Equine Vet. J. 30,
and trailing limb at the canter. Equine Vet. Caudron, I., Grulke, S., Farnir, F., Aupaix, R., 384–389.
J. 23 (Suppl.), 80–83. Serteyn, D., 1998. Radiographic assessment Clayton, H.M., Schamhardt, H.C., Lanovaz,
Back, W., Schamhardt, H.C., Hartman, W., of equine interphalangeal joints J.L., Colborne, G.R., Willemen, M.A.,
Barneveld, A., 1995c. Repetitive loading asymmetry: articular impact of phalangeal 2000b. Net joint moments and joint
and oscillations in the distal fore and hind rotations (Part 1). J. Vet. Med. A 45, powers in horses with superficial digital
limb as predisposing factors for equine 319–325. flexor tendinitis. Am. J. Vet. Res. 61,
lameness. Am. J. Vet. Res. 56, 1522–1528. Chateau, H., Degueurce, C., Denoix, J.-M., 197–201.
Back, W., Schamhardt, H.C., Savelberg, 2004. Evaluation of three-dimensional Clayton, H.M., Sha, D.H., Stick, J.A., Elvin,
H.H.C.M., et al., 1995a. How the horse kinematics of the distal portion of the N.G., 2007a. 3D kinematics of the
moves: significance of graphical forelimb in horses walking in a straight metacarpophalangeal joint at walk and
representations of equine forelimb line. Am. J. Vet. Res. 65, 447–455. trot. Vet. Comp. Orthop. Traum. 20, 86–91.
kinematics. Equine Vet. J. 27, 31–38. Chateau, H., Degueurce, C., Denoix, J.-M., Clayton, H.M., Sha, D.H., Stick, J.A.,
Back, W., van den Bogert, A.J., van Weeren, 2005. Three-dimensional kinematics of the Mullineaux, D.R., 2004. Three dimensional
P.R., Bruin, G., Barneveld, A., 1993. equine distal forelimb : effects of a sharp kinematics of the equine carpus at trot.
Quantification of the locomotion of Dutch turn at the walk. Equine Vet. J. 37, 12–18. Equine Vet. J. 36, 671–676.
warmblood foals. Acta Anat. 146, 141–147. Chateau, H., Degueurce, C., Denoix, J.-M., Clayton, H.M., Sha, D.H., Stick, J.A.,
Biewener, A.A., 1998. Muscle-tendon stresses 2006. Three-dimensional of the distal Robinson, P., 2007b. 3D kinematics of the
and elastic energy storage during forelimb in horses trotting on a treadmill proximal and distal interphalangeal joints
locomotion in the horse. Comp. Biochem. and effects of elevation of heel and toe. of walking and trotting horses. Vet. Comp.
Physiol. Biochem. Mol. Biol. 120, 73–87. Equine Vet. J. 38, 164–169. Orthop. Traum. 20, 1–7.
Birch, H.L., 2007. Tendon matrix composition Chateau, H., Degueurce, C., Jerbi, H., Colborne, G.R., Lanovaz, J.L., Sprigings, E.J.,
and turnover in relation to functional Crevier-Denoix, N., Pourcelot, P., Schamhardt, H.C., Clayton, H.M., 1997a.

122
 References

Joint moments and power flow in equine Farley, C.T., Glasheen, J., McMahon, T.A., evaluation of a noninvasive marker cluster
gait: a preliminary study. Equine Vet. J. 23 1993. Running speed: springs and animal technique to assess three-dimensional
(Suppl.), 33–36. size. J. Exp. Biol. 185, 71–86. kinematics of the distal portion of the
Colborne, G.R., Lanovaz, J.L., Sprigings, E.J., Fleiss, O., Edl, M., Schubert, P., 1984. Intra forelimb in horses. Am. J. Vet. Res. 67,
Schamhardt, H.C., Clayton, H.M., 1997b. und interlimbkoordination bei 1511–1518.
Power flow in the equine forelimb. Equine verschiedenen ganggeschwindigkeiten des Hodson, E.F., Clayton, H.M., Lanovaz, J.L.,
Vet. J. 23 (Suppl.), 37–40. pferdes. Arch. Tierärztl. Fortbild. 8, 2000. The forelimb in walking horses: 1.
Colborne, G.R., Lanovaz, J.L., Sprigings, E.J., 116–128. Kinematics and ground reaction forces.
Schamhardt, H.C., Clayton, H.M., 1998. Fredricson, I., Drevemo, S., 1972. Equine Vet. J. 32, 287–294.
Forelimb joint moments and power during Methodological aspects of kinematics of Holmström, M., Fredricson, I., Drevemo, S.,
the walking stance phase of horses. Am. J. the joints in the forelimbs of fast-moving 1994. Biokinematic analysis of the Swedish
Vet. Res. 59, 609–614. horses. In: Fredricson, I. (ed) A warmblood riding horse at trot. Equine Vet.
Corley, J.M., Goodship, A.E., 1994. Treadmill photogrammetic study applying high speed J. 26, 235–240.
training induced changes to some cinematography. Acta Vet. Scand. 37 Jansen, M.O., Raaij, J.A.G.M., van Bogert, A.J.,
kinematic variables measured at the canter (Suppl.), 1–136. van den Schamhardt, H.C., Hartman, W.,
in thoroughbred fillies. Equine Vet. J. 17 Fuller, J., Liu, L.-J., Murphy, M.C., et al., 1997. 1992. Quantitative analysis of computer-
(Suppl.), 20–24. A comparison of lower-extremity skeletal averaged electromyographic profiles of
Crevier-Denoix, N., Roosen, C., Dardillat, C., kinematics measured using skin- and intrinsic limb muscles in ponies at the
Pourcelot, P., Jerbi, H., Sanaa, M., et al., pin-mounted markers. Human Movement walk. Am. J. Vet. Res. 53, 2343–2349.
2001. Effects of heel and toe elevation Sci. 16, 219–242. Johnston, C., Drevemo, S., Roepstorff, L.,
upon the digital joint angles in the Galisteo, A.M., Cano, M.R., Miro, F., Vivo, J., 1997. Kinematics and kinetics of the
standing horse. Equine Vet. J. Suppl. 33, Morales, J.L., Aguera, E., 1996. Angular carpus. Equine Vet. J. 23 (Suppl.), 84–88.
74–78. joint parameters in the Andalusian horse at Johnston, C., Gottlieb-Vedi, M., Drevemo, S.,
Degueurce, C., Chateau, H., Pasqui-Boutard, walk, obtained by normal videography. J. Roepstorff, L., 1999. The kinematics of
V., Pourcelot, P., Audigié, F., Crevier- Equine. Vet. Sci. 16, 73–77. loading and fatigue in the standardbred
Denoix, N., et al., 2000. Concrete use of Grandage, J., 1981. Penniform muscles of the trotter. Equine Vet. J. 30 (Suppl.),
the joint coordinate system for the equine distal forelimb. J. Anat. Abstr. 132, 249–253.
quantification of articular rotations in the 318. Johnston, C., Roepstorff, L., Drevemo, S.,
digital joints of the horse. Vet. Res. 31, Grillner, S., 1975. Locomotion in vertrebrates: 1995. Kinematics of the distal forelimb
297–311. central mechanisms and reflex interaction. during the stance phase in the fast trotting
Degueurce, C., Dietrich, G., Pourcelot, P., Physiol. Rev. 55, 247–304. standardbred. Equine Vet. J. 18 (Suppl.),
Denoix, J.M., Geiger, D., 1996. Three Grood, E.S. Suntay, W.J., 1983. A joint 170–174.
dimensional kinematic technique for coordinate system for the clinical Khumsap, S., Clayton, H.M., Lanovaz, J.L.,
evaluation of horse locomotion in outdoor description of three-dimensional motions: Bouchey, M., 2002. Effect of walking
conditions. Med. Biol. Engineering Comp. application to the knee. J. Biomech. Eng. velocity on forelimb kinematics and
5, 249–252. 105, 136–144. kinetics. Equine Vet, J. Suppl. 34, 325–329.
Degueurce, C., Pourcelot, P., Audigie, F., Gunn, H.M., 1978. Differences in the Korsgaard, E., 1982. Muscle function in the
Denoix, J.M., Geiger, D., 1997. Variability histochemical properties of skeletal- forelimb of the horse. An
of the limb joint patterns of sound horses muscles of different breeds of horses and electromyographical and kinesiological
at the trot. Equine Vet. J. 23 (Suppl.), dogs. J. Anat. 127, 615–634. study. Thesis, Copenhagen.
89–92. Hagen, R., McGuigan, M.P., Wilson, A.M., Krüger, W., 1937. Ueber den bewegungsablauf
Denoix, J.M., 1999. Functional anatomy 2002. The architecture of the digital flexor an dem oberen teil der vordergliedmaβe
of the equine interphalangeal joints. Proc. muscles of the equine distal forelimb. des pferdes im schritt, trab and galopp.
Amer. Assoc. Equine Practnr. 45, 174–177. Comp. Biochem. Physiol. A 132A, 77. Tierärztl. Rundsh. 43, 1–24.
Deuel, N.R., 1994. Coordination of equine Hermanson, J., 1997. Architecture and the Krüger, W., 1938. Ueber den bewegungsablauf
forelimb motion during the gallop. Equine division of labor in the extensor carpi an dem oberen teil der hintergliedmaβe
Vet. J. 17 (Suppl.), 29–34. radialis muscle of hotses. Acta Anat. (Basel) des pferdes im schritt, trab und galopp.
Deuel, N.R., Park, J.J., 1990. Canter leads 159, 127–135. Tierärztl. Rundsh. 44, 549–557.
change kinematics of superior dressage Hermanson, J.W., Cobb, M.A., 1992. Four Lafortune, M.A., Cavanagh, P.R., Sommer, H.J.,
horses. J. Equine Vet. Sci. 10, 287–298. forearm flexor muscles of the horse, et al., 1992. Three-dimensional kinematics
Dimery, N.J., Alexander, R.McN., Ker, R.F., equus-caballus – anatomy and of the human knee during walking. J.
1986. Elastic extension of leg tendons in histochemistry. J. Morph. 212, 269–280. Biomech. 25, 347–357.
the locomotion of horses (equus caballus). Hermanson, J.W., Hurley, K.J., 1990. Lanovaz, J.L., Clayton, H.M., Colborne, G.R.,
J. Zool. 210, 415–425. Architectural and histochemical analysis of Schamhardt, H.C., 1999. Forelimb
Dutto, D.J., Hoyt, D.F., Clayton, H.M., Cogger, the biceps brachii muscle of the horse. Acta kinematics and net joint moments during
E.A., Wickler, S.J., 2006. Joint work and Anat. 137, 146–156. the swing phase of the trot. Equine Vet. J.
power for both the forelimb and hind limb Hildebrand, M., 1987. The mechanics of the 30 (Suppl.), 235–239.
during trotting in the horse. J. Exp. Bio. horse leg. American Scientist 75, 594–601. Lanovaz, J.L., Khumsap, S., Clayton, H.M.,
209, 3990–3999. Hill, A.E., Stover, S.M., Gardner, I.A., Kane, Stick, J.A., Brown, J., 2002. Three-
Dyce, K.M., Sack, W.O., Wensing, C.J.G., 1996. A.J., Whitcomb, M.B., Emerson, A.G., 2001. dimensional kinematics of the tarsal joint
The forelimb of the horse. In: Textbook of Risk factors for and outcomes of at the trot. Equine Vet. J. Suppl. 34,
veterinary anatomy, second ed. W.B. noncatastrophic suspensory apparatus 308–313.
Saunders, Philadelphia, pp. 573–609. injury in thoroughbred racehorses. J. Am. Lawson, S.E.M., Chateau, H., Pourcelot, P.,
Dyhre-Poulsen, P., Smedegaard, H.H., Roed, J., Vet. Med. Assoc. 218, 1136–1144. Denoix, J.M., Crevier-Denoix, N., 2007.
Korsgaard. E., 1994. Equine hoof function Hjertén, G., Drevemo, S., 1993. Shortening of Effect of toe and heel elevation on
investigated by pressure transducers inside the forelimb in the horse during the stance calculated tendon strains in the horse and
the hoof and accelerometers mounted on phase. Acta Anat. 146, 193–195. the influence of the proximal
the first phalanx. Equine Vet. J. 26, Hobbs, S.J., Richards, J., Matuszewski, B., interphalangeal joint. J. Anat. 210,
362–366. Brigden, C., 2006. Development and 583–591.

123
 6 Forelimb function

Leach, D.H., Sprigings, E., 1979. Gait fatigue horse equine compar. Exerc. Physiol. 4, Equine Sports Med. Sci. Wageningen Press,
in the racing thoroughbred. J. Equine Med. 59–70. Wageningen, The Netherlands. pp. 80–85.
Surg. 3, 436–443. Nicodemus, M.C., Holt, H.M., 2006. Two- Rooney, J.R., 1990. The jump behaviour of the
Leach, D.H., Sprigings, E.J., Laverty, W.H., dimensional kinematics of the flat-walking humeroradial and tarsocrural joints of the
1987. Multivariate statistical analysis of tennessee walking horse yearling. Horse. horse. J. Equine Vet. Sci. 4, 311–314.
stride-timing measurements of nonfatigued Equine Compar. Exerc. Physiol. 3, 101–108. Ryan, J. M., Cobb, M. A., Hermanson, J. W.,
racing thoroughbreds. Am. J. Vet. Res. 48, Niki, Y., Ueda, Y., Masumitsu, H., 1984. A 1992. Elbow extensor muscles of the horse
880–888. forceplate study in equine biomechanics 3. – postural and dynamic implications. Acta
MacLellan, K.N., Crawford, W.H., Macdonald, The vertical and fore aft components of Anat. 144, 71–79.
D.G., 2001. Proximal interphalangeal joint floor reaction forces and motion of equine Schaer, T.P., Bramlage, L.R., Embertson, R. M.
arthrodesis in 34 horses using two parallel limbs at canter. Bull. Equine Res. Inst. 21, Hance, S., 2001. Proximal interphalangeal
5.5-mm cortical bone screws. Vet. Surg. 30, 8–18. arthrodesis in 22 horses. Equine Vet. J. 33,
454–459. Niki, Y., Ueda, Y., Yoshida, K., Masumitsu, H., 360–365.
Martin, G.S., McIlwraith, C.W., Turner, A.S., 1982. Force plate study of equine Schamhardt, H.C. Merkens, H.W., 1994.
Nixon, A.J., Stashak, T.S., 1984. Long-term biomechanics. 2. The vertical and fore-aft Objective determination of ground contact
results and complications of proximal components of floor reaction forces and of the limbs of the horse at the walk and
interphalangeal arthrodesis in horses. J. motion of equine limbs. Bull. Equine Res. trot: comparison between ground reaction
Am. Vet. Med. Assoc. 184, 1136–1140. Inst. 19, 1–17. forces, accelerometer data and kinematics.
Martinez-del Campo, L.J., Kobluk, C.N., Greer, Palmgren, A., 1929. Zur kenntnis der Equine Vet. J. 17 (Suppl.), 75–79.
N., et al., 1991. The use of high-speed sogenannten schnappgelenke. II Die Schauder, W., 1952. Die besonderen
videography to generate angle-time and physiologische bedeuting des stoβbrechenden einrichtungen an den
angle-angle diagrams for the study of federungsphänomens. Zeitschr. anat. gliedmaβen des pferdes. Dtsch. tierärztl.
equine locomotion. Vet. Comp. Orthop. entwickl. Gesch. 89, 194–200. Wschr. 59, 35–38.
Traum. 4, 120–131. Payne, R.C., Veenman, P., Wilson, A.M., 2004. Schryver, H.F., Bartel, D.L., Langrana, N.,
Martuzzi, F., Vaccari, S.F., Gosi, S., Catalano, The role of the extrinsic thoracic limb Lowe, J.E., 1978. Locomotion in the horse:
A.L., 2007. Kinematic characteristics of muscles in equine locomotion. J. Anat. kinematics and external and internal forces
walk in the bardigiano horse breed. Ann. 205, 479–490. in the normal equine digit in the walk and
Fac. Medic. Di Parma 27, 229–234. Peloso, J.G., Mundy, G.D., Cohen, N.D., 1994. trot. Am. J. Vet. Res. 39, 1728–1733.
McGuigan, M.P. Wilson, A.M., 2003. The effect Prevalence of, and factors associated with, Seder, J.A., Vickery, C.E., 2003. Double and
of gait and digital flexor muscle activation musculoskeletal racing injuries of triple fully airborne phases in the gaits of
on limb compliance in the forelimb of the thoroughbreds. J. Am. Vet. Med. Assoc. racing speed Thoroughbreds. J. Equine Vet.
horse equus caballus. J. Exp. Biol. 206, 204, 620–626. Sci. 23 (Suppl. 1), S73–S81.
1325–1336. Preedy, D., 1998. Analysis of horse thoracic Seder, J.A., Vickery, C.E., 2005. The
McIlwraith, C.W., Goodman, N.L., 1989. limb muscle activity (EMG) at walk and relationship of subsequent racing
Conditions of the interphalangeal joints. trot. Honours Dissertation, Bristol performance to foreleg flight patterns
Vet. Clin. N. Amer. Equine Pract. 5, University. during racing speed workouts of unraced
161–178. Ramsey, D.K., Lamontagne, M., Wretenberg, 2-year-old Thoroughbred racehorses at
Merkens, H.W., Schamhardt, H.C., 1988. P.F., et al., 2001. Assessment of functional auctions. J. Equine Vet. Sci. 25, 505–522.
Distribution of the ground reaction forces knee bracing: an in vivo three-dimensional Sha, D.H., Mullineaux, D.R., Clayton, H.M.,
of concurrently loaded limbs of the Dutch kinematic analysis of the anterior cruciate 2004. Three-dimensional analysis of
warmblood horse at the normal walk. deficient knee. Clin. Biomech. (Bristol, patterns of skin displacement over the
Equine Vet. J. 20, 209–213. Avon) 16, 61–70. equine radius. Equine Vet. J. 36,
Merkens, H.W., Schamhardt, H.C., 1994. Ramsey, D.K., Wretenberg, P.F., 1999. 665–670.
Relationship between ground reaction force Biomechanics of the knee: methodological Singleton, W.H., Clayton, H.M., Lanovaz, J.L.,
patterns and kinematics in the walking and considerations in the in vivo kinematic Prades, M., 2003. Effects of shoeing on
trotting horses. Equine Vet. J. 17 (Suppl.), analysis of the tibiofemoral and forelimb swing phase kinetics of trotting
67–70. patellofemoral joint. Clin. Biomech. horses. Vet. Comp. Orthop. Traum. 16,
Merkens, H.W., Schamhardt, H.C., van Osch, (Bristol, Avon) 14, 595–611. 16–20.
G.J.V.M., van den Bogert, A.J., 1993. Reinschmidt, C., van den Bogert, A.J., Sloet van Oldruitenborgh-Oosterbaan, M.M.,
Ground reaction force patterns of Dutch Lundberg, A., et al., 1997. Tibiofemoral van Schamhardt, H.C., Barneveld, A., 1995.
warmbloods at the canter. Am. J. Vet. Res. and tibiocalcaneal motion during walking: Effects of weight and riding on workload
54, 670–674. external vs. skeletal markers. Gait Posture and locomotion during treadmill exercise.
Minetti, A.E., Ardigo, L.P., Reinach, E., Saibene, 6, 98–109. Equine Exerc. Physiol. 4, 413–417.
F., 1999. The relationship between Riemersma, D.J., Bogert, A.J., van den Smythe, R.H., Goody, P.C., Gray, P., 1993.
mechanical work and energy expenditure Schamhardt, H.C., Hartman, W., 1988b. Horse structure and movement. Allen &
of locomotion in horses. J. Exp. Biol. 202, Kinetics and kinematics of the equine Co, London.
2329–2338. hind limb: in vivo tendon strain and Soffler, C., Hermanson, J.W., 2006. Muscular
Mosimann, W., Micheluzzi, P., 1969. Die joint kinematics Am. J. Vet. Res. 49, design in the equine interossues muscle. J.
bewegung im cubitus des pferdes als 1353–1359. Morph. 267, 696–704.
gedampfte, erzwungene schwingung. Riemersma, D.J., Schamhardt, H.C., Hartman, Stashak, T.S., 1987. Adams’ lameness in
Zentbl. Vet. Med. Reihe A 16, 180–184. W., Lammertink, J.L.M.A., 1988a. Kinetics horses, fourth ed. Philadelphia: Lea &
Mullineaux, D.R., Clayton, H.M., Gnagey, and kinematics of the equine hind limb: in Febiger, pp. 76, 88, 102 and 487–785.
L.M., 2004. Effects of offset-normalizing vivo tendon loads and force plate Stubbs, N.C., Clayton, H.M., 2008. Activate
techniques on variability in motion measurements in ponies. Am. J. Vet. Res. your horse’s core. Mason: Sport Horse
analysis data. J. Appl. Biomech. 20, 49, 1344–1352. Publications.
177–184. Robert, C., Valette, J.P., Denoix, J.M., 1998. Swanson, T.D., 1988. Degenerative disease of
Nicodemus, M.C., Booker, J.E., 2007. Surface electromyographic analysis of the the proximal interphalangeal (pastern)
Two-dimensional kinematics of the jog and normal horse locomotion: a preliminary joint in performance horses. Proc. Am.
lope of the stock breed western pleasure report. In: Lindner, A. (Ed.), Proc. Conf. Assoc. Equine Practnr. 34, 393–397.

124
 References

Swanstrom, M.D., Zarucco, L., Stover, S.M., Van Weeren, P.R., van den Bogert, A.J., Wilson, A.M., Watson, J.C., 2003. A catapult
Hubbard, M., Hawkins, D.A., Driessen, B., Barneveld, A., 1990a. A quantitative action for rapid limb protraction. Nature
et al., 2005. Passive and active mechanical analysis of skin displacement in the 421, 35–36.
properties of the superficial and deep trotting horse. Equine Vet. J. 9 (Suppl.), Witte, T.H., Hirst, C.V., Wilson, A.M.,
digital flexor muscles in the forelimbs 101–109. 2006. Effect of speed on stride
of anesthetized thoroughbred horses. Van Weeren, P.R., van den Bogert, A.J., parameters in racehorses at gallop
J. Biomech. 38, 579–586. Barneveld, A., 1990b. Quantification of in field conditions. J. Exp. Biol. 209,
Thompson, K.N., Herring, L., Shapiro, R., skin displacement in the proximal parts of 4389–4397.
1992. A three dimensional kinematic study the limbs of the walking horse. Equine Vet. Woltring, H.J., 1994. 3-D attitude
of the metacarpophalangeal joint in horses. J. 9 (Suppl.), 110–118. representation of human joints: a
J. Equine Vet. Sci. 12, 172–176. Viitanen, M.J., Wilson, A.M., McGuigan, H.R., standardization proposal. J. Biomech. 27,
Tokuriki, M., Aoki, O., Niki, Y., Kurakawa, Y., Rogers, K.D., May, S.A., 2003. Effect of foot 1399–1414.
Hataya, M., Kita, T., 1989. balance on the intra-articular pressure in Wu, G., Cavanagh, P.R., 1995. ISB
Electromyographic activity of cubital joint the distal interphalangeal joint in vitro. recommendations for standardization in
muscles in horses during locomotion. Am. Equine Vet. J. 35, 184–189. the reporting of kinematic data. J.
J. Vet. Res. 50, 950–957. Walter, K., 1925. Der bewegungsablauf an den Biomech. 28, 1257–1261.
Ueda, Y., Niki, Y., Yoshida, K., Masumitsu, H., freien gliedmassen des pferdes im schritt, Zajac, F. E., 1989. Muscle and tendon
1981. Force plate study of equine trab und galop. Arch. Wiss. Pract. Tierheilk. – properties, models, scaling, and
biomechanics: floor reaction forces of 53, 316–352. application to biomechanics and motor
normal walking and trotting horses. Bull. Watson, J.C., Wilson, A.M., 2007. Muscle control. Crit. Rev. Biomed. Eng. 17,
Equine Res. Inst. 18, 28–41. architecture of bideps brachii, triceps 359–411.
van den Bogert, A.J., 1989. Computer brachii and supraspinatus I the horse. Zarucco, L., Taylor, K. T., Stover, S. M., 2004.
simulation of locomotion in the horse. J. Anat. 210, 32–40. Determination of muscle architecture and
Thesis, Utrecht. Wilson, A.M., McGuigan, M.P., Su, A., van fiber characteristics of the superficial and
Van Weeren, P.R., Barneveld, A., 1986. A den Bogert, A.J., 2001. Horses damp deep digital flexor muscles in the forelimbs
technique to quantify skin displacement in the spring in their step. Nature 414, of adult horses. Am. J. Vet. Res. 65,
the walking horse. J. Biomech. 19, 879–883. 895–899. 819–828.

125
This page intentionally left blank
 C H A PTER 7 

Hind limb function

Hilary M. Clayton, Willem Back

Introduction that are well-suited for generating force economically. Muscle size
and fiber length are not correlated, however, so the larger muscles
do not necessarily have longer fibers and smaller muscles do not
This chapter reviews the structure and functions of the equine hind always have shorter fibers (Table 7.1). Fiber length variability is low
limbs in relation to stride kinematics (movements) and kinetics in the digital flexor muscles both within an individual horse and
(forces), and highlights differences between the fore and hind between different horses. The amount of force transmitted to the
limbs. The hind limbs carry only about 43% of body mass, so the tendon is calculated as the product of PCSA and the cosine of the
vertical ground reaction force and vertical impulse are lower in the pennation angle. When pennation angle exceeds about 20° it sig-
hind limbs than in the forelimbs. In the longitudinal direction, nificantly reduces the component of force transmitted to the tendon.
the propulsive component of the ground reaction force tends to be The benefit of larger pennation angles is the increase in force-
larger and the braking component smaller in the hind limbs com- generating capacity as a result of the larger PCSA for a given volume.
pared with the forelimbs. The angulations of the hip, stifle and These muscular characteristics are typical of musculotendinous
tarsal joints are well suited to play an active role in generating pro- units used for elastic energy storage and recoil, which reduces the
pulsion, which is in contrast to the supporting function of the more need for energetically expensive changes in muscle length, while
strut-like forelimbs. minimizing distal limb mass. The suspensory ligament represents
the end point of such modification, having minimal muscular tissue
in the adult horse (Klomkleaw et al., 2002). Some muscles, notably
Musculotendinous architecture fibularis tertius and superficial digital flexor, have so little muscular
tissue that they function as strong tendinous bands to synchronize
Horses are cursorial animals and within the genus Equus different flexion–extension movements of stifle and tarsus in an arrangement
breeds have been developed for different occupations. Breeds spe- known as the reciprocal apparatus (Wentink, 1978b).
cialized for racing tend to have long, gracile limbs, whereas draft During trotting, changes in potential and kinetic energy of the
breeds have shorter, stouter limbs in proportion to body size (Gunn, horse’s center of mass are in phase, which allows the distal limb to
1983). Fat and muscle distribution also differ between breeds but, make substantial contributions to elastic energy storage (Biewener,
regardless of breed, successful athletes tend to have body fat per- 1998). In canter and gallop, the relationship between kinetic and
centages in the range of 7.8–8.8% (Lawrence et al., 1992; Kearns potential energy varies during the stride (Minetti et al., 1999),
et al., 2002). In athletic horses, muscles comprise about 53% of which limits the ability to store elastic energy in the distal limb. It
body mass compared with 42% in non-athletes, and the propulsive has been estimated that the hind limbs are responsible for two-
muscles in the hind limb account for a larger portion of that muscle thirds of the elastic energy savings in horses (Biewener, 1998).
mass (Gunn, 1987). These muscles are responsible for generating Payne et al. (2005) described the morphology and architecture
external work that is required for acceleration or to raise the center of the hind limb musculature in horses of different sizes (body mass
of mass when moving uphill or jumping (Clayton et al., 2002; 430–600 kg) and ages (10–30 years) (Table 7.1). Data for muscle
Dutto et al., 2004a, b). The muscles in the proximal hind limb are mass were scaled to body mass and measurements of fiber length
characterized by having large physiological cross sectional areas were scaled to (body mass (kg) )1/3, which resulted in similar values
(PCSA), associated with large mass and volume and relatively long among the seven subjects for most muscles, with the exceptions of
fibers that move the joints through a large range of motion. Many gluteus medius and the hamstrings, which may have adapted in
of these muscles have multiple bellies: the more proximal bellies response to different types of training. After scaling for geometric
have both their origin and insertion located more proximally than similarity, fiber lengths varied more than muscle masses, particu-
the distal bellies allowing the entire muscle to exert force over a larly in the vertebral head of semitendinosus and in semimembranosus.
wide range of joint positions without needing long muscle fibers. In general, the hamstring group is capable of generating large forces
The muscle fibers are arranged in series with elastic tissue and they and high powers. All the hind limb muscles have pennation angles
attach to the bones directly or via short tendons. These characteris- greater than 20° and many have considerably larger angles. Most
tics are typical of muscles that are specialized for doing work. Many of the proximal muscles have little, if any, tendon and, when tendon
of the hind limb muscles are multi-articular and have complex tissue is present, it is light in weight (5.3–34.2 g) compared with
fascial attachments and connections that make it difficult to sepa- tendons in the distal limb (44.8–208.7 g).
rate extrinsic and intrinsic functions. Hind limb muscle architecture has been compared in Arabians,
Muscle volume and fiber length decrease, and pennation angles a breed that excels in endurance races, with Quarter Horses, a
increase in a proximal to distal direction, so the muscles in the distal breed noted for its sprinting speed (Crook et al., 2008). Overall
hind limb are smaller and less powerful with short, pennate fibers height and mass of the horses did not differ but the Quarter Horses

127
 7 Hind limb function

Table 7.1  Architectural properties of the muscles of the equine hind limb

Muscle Mass Volume MFL PCSA Angle Force Power

(g) (cm3) (mm) (cm2) (°) (kN) (W)
Psoas minor 432 408 67 (50–205) 61 37 (35–38) 1833 65
Psoas major 1182 1116 198 (95–280) 56 29 (20–35) 1694 177
Iliacus 1395 1316 244 (170–305) 54 26 (15–40) 1617 209
Gluteus superficialis 646 609 102 (55–190) 60 36 (28–50) 1793 97
Gluteus medius 8577 8091 203 (135–300) 398 28 (15–45) 11942 1287
Gluteus profundus 1351 1275 118 (42–220) 108 30 (15–40) 3249 203
Tensor fascia lata 1448 1366 97 (60–150) 140 34 (20–43) 4210 217
Biceps femoris
Intermediate head 870 820 235 (190–260) 27 27 (20–40) 1048 130
Vertebral head 6112 5766 258 (130–330) 37 37 (28–50) 6705 917
Caudal head 946 892 245 (170–300) 39 39 (30–45) 1092 142
Semitendinosus
Vertebral head 26841 2532 274 (105–289) 92 28 (20–35) 2770 403
Pelvic head 727 1630 312 (180–355) 52 35 (25–45) 1567 259
Semimembranosus 3834 3617 342 (80–760) 106 35 (20–45) 3171 575
Sartorius 484 456 376 (250–460) 12 21 (20–25) 364 73
Gracilis 1760 1661 123 (80–175) 135 31 (22–35) 4037 264
Pectineus 447 422 78 (49–320) 54 29 (13–50) 1614 67
Adductor 3924 3702 176 (80–390) 211 35 (25–40) 6322 589
Quadriceps femoris
Rectus femoris 2291 2161 98 (40–152) 220 40 (25–53) 6610 344
Vastus medialis 1878 1772 119 (90–145) 148 33 (25–41) 4448 282
Vastus intermedius 501 473 105 (60–220) 45 41 (30–55) 1355 75
Vastus lateralis 1734 1636 155 (92–220) 105 36 (30–40) 3163 260
Popliteus 280 264 38 (21–170) 70 42 (35–50) 2107 42
Gastrocnemius
Medialis 817 771 48 (25–99) 161 36 (20–45) 4836 123
Lateralis 808 762 56 (36–70) 137 34 (30–45) 4098 121
Soleus 6 6 121 (110–151) 0 22 (22–22) 15 1
Flexor digitorum superficialis 111 105 3 (1–6) 417 52 (40–60) 12 514 17
Flexor digitorum profundus
Flexor digitorum medialis 161 152 70 (4–100) 22 27 (12–40) 652 24
Flexor digitorum lateralis 660 622 10 (3–55) 644 44 (30–60) 19 324 99
Tibialis caudalis 224 211 57 (40–117) 37 31 (20–60) 1106 34
Tibialis cranialis 309 291 40 (17–218) 73 41 (25–60) 2199 46
Extensor digitorum longus 462 435 81 (60–110) 54 29 (25–45) 1620 69
Extensor digitorum lateralis 192 181 70 (35–110) 26 28 (20–45) 776 29

MFL, mean fascicle length; PCSA, mean physiological cross sectional area; Angle, mean (range) of pennation angles of fibers; Force, maximal isometric force generation capacity
estimated as the product of PCSA and maximal isometric stress of skeletal muscle, taken as 0.3 MPa; Power, maximal power output calculated as one-tenth of the product of
force and maximal contraction velocity, which was estimated based on published values of equine muscle fiber-typing. Unless indicated, different heads of the muscles were
combined.
Data from Payne et al. (2005).

128
 Musculotendinous architecture

Superficial gluteal

Middle gluteal

Biceps femoris
Semitendinosus
Semimembranosus

Gracilis

Superficial digital flexor

Tensor fasciae latae Lateral gastrocnemius

Long digital extensor

Medial gastrocnemius
Lateral digital extensor
Deep digital flexor

Fig 7.1  Musculature in the proximal hind limb viewed from the side (left) and from behind (right).

had greater muscle mass and volume in both proximal and distal and early stance (Robert et al., 1998). It contributes to a burst of
muscles and larger PCSAs allowing the development of larger power generation across the hip joint that pushes the trunk forward
forces (Crook et al., 2008). Thoroughbreds trained for sprinting over the grounded hind limb (Clayton et al., 2001). The muscle
have a higher hind limb muscle to body mass ratio than those fibers attach onto broad aponeurotic sheets that are somewhat com-
trained for hurdling (Gunn, 1987). Fiber lengths and pennation pliant and may provide some elastic energy storage and return.
angles are similar in different breeds of horses (Crook et al., 2008). Gluteus medius represents 2.0% of body mass in Quarter Horses and
This is in contrast to human sprinters who tend to have longer 1.7% in Arabians (Crook et al., 2008), which is indicative of the
fibers and smaller pennation angles in their calf muscles compared specialization of Quarter Horses for acceleration and speed. Inter-
with distance runners (Abe et al., 2000), though this may be an nally, gluteus medius shows functional compartmentalization; the
effect of a specific type of training. PCSA and calculated maximal superficial parts of the muscle are composed primarily of type IIb
isometric force decrease as follows: biceps femoris; semitendinosus; fibers, which suggests a propulsive function, whereas the deeper
gastrocnemius; vastus lateralis; extensor digitorum longus; tibialis parts have more type I fibers that are typical of a postural function
cranialis. (Lopez-Rivero et al., 1992; Serrano et al., 1996). Muscle fiber
Gluteus superficialis (Fig. 7.1, Table 7.1), the smallest of the gluteal lengths do not appear to have a superficial to deep gradation and
muscles, is a hip flexor. It is one of the few muscles in the proximal range from 135 mm to 300 mm throughout the muscle.
hind limb that has a distinct tendon. This tendon is short in length Gluteus profundus (Table 7.1) is a smaller muscle than gluteus
(980 mm), small in mass (14.0 g) and volume (12.5 cm3) but has medius in terms of size and PCSA, with shorter fibers but similar
a relatively large cross-sectional area (1.39 cm2). It may provide pennation angles.
some elastic energy storage and return. Calculated stress in the Biceps femoris (Fig. 7.1, Table 7.1) is a multi-articular muscle that
tendon is quite low (13.0 MPa). can act as a hip extensor, stifle flexor and tarsal extensor. It plays an
Gluteus medius (Fig. 7.1, Table 7.1) is a monoarticular hip exten- important role in stabilization of the hip and stifle joints. The three
sor and is the largest muscle of the hind limb in terms of mass heads (intermediate, vertebral and caudal) have a combined mass
(8577 g) and PCSA (398 cm2). The fibers are long with pennation of 7928 g making it one of the largest muscles in the hind quarters.
angles in the range 15–45° (Payne et al., 2005). These characteris- It has a large PCSA, especially the biarticular vertebral head, due to
tics are indicative of a primary role in force and power generation. its large volume (Payne et al., 2005). This study reported pennation
In fact, gluteus medius has been estimated to have by far the largest angles within the range of 20–55° in all three heads, whereas Crook
force-generating capacity of all the hind limb muscles and plays an et al. (2008) described a parallel fiber orientation (pennation angle,
important role in propulsion. Electromyographic studies at trot ≤5°) in the intermediate and caudal heads and 55° pennation
indicate that it is active during hind limb retraction in late swing angles in the vertebral head. Biceps femoris is the only muscle in the

129
 7 Hind limb function

proximal hind limb that has a large tendon (mass, 106 g; CSA, action of the resultant load on the tibia thus reducing the strain due
9.54 cm2) though this tendon is short in length (100 mm). During to bending (Wentink, 1978a).
trotting, it is active during hind limb retraction in late swing and Fibularis (peroneus) tertius forms the cranial arm of the reciprocal
early stance (Robert et al., 1998). apparatus and is entirely tendinous (mass, 64.3 g; volume, 57.4 cm3;
Semitendinosus (Fig. 7.1) has vertebral and pelvic heads. Accord- CSA, 1.59 cm2). Rupture of fibularis tertius as a consequence of
ing to Payne et al. (2005), both heads have pennation angles in the hyperextension of the tarsus, does not have much effect on the
range of 20–45° (Table 7.1) whereas Crook et al. (2008) identified horse’s stance or walking ability (Strubelt, 1928; Schamhardt et al.,
distinct differences between the two head with the pelvic head 1985), although Wentink (1978b) recorded slight hyperextension
having parallel fibers (pennation angle, ≤5°) and fibers in the ver- at the end of stance and less flexion of the tarsal joint during swing.
tebral head having a mean pennation angle of 23°, which supports At trot, the loss of elastic rebound of fibularis tertius is responsible
its presumptive role in propulsion This muscle is thought to be for delayed protraction of the distal limb in early swing.
important in generating large forces during hind limb retraction. Flexor digitorum superficialis (SDF) (Fig. 7.1, Table 7.1) in the hind
PCSA of semitendinosus in Quarter Horses is double that of Arabians limb is almost entirely tendinous, forming the caudal arm of the
(Crook et al., 2008). Semitendinosus is active during retraction of reciprocal apparatus that synchronizes stifle and tarsal motion in
both the ipsilateral and the contralateral hind limb in late swing the sagittal plane. The muscle belly has a small volume of only
and early stance (Robert et al., 1998). 105 cm3 with uniformly short (1–6 mm) and highly pennate (40–
Semimembranosus (Fig. 7.1) has vertebral and pelvic heads that 60°) fibers, resulting in a surprisingly large PCSA (417 cm2) (Table
cannot easily be separated and are considered together in Table 7.1. 7.1). Estimates suggest that SDF has a capacity to generate high force
A notable feature of this muscle is its large range of fiber lengths (12500 N) but low power (17 W). The SDF tendon is long
from 80–760 mm. Some fibers run the entire length of the belly (748 mm) with a large mass (2.25 cm2) (Payne et al., 2005). As a
from origin to insertion; others are staggered along the muscle belly. consequence of having short muscle fibers and a long tendon, the
Similar to the other hamstring muscles, semimembranosus is mul- hind limb SDF is particularly well suited to elastic energy storage
tiarticular and is estimated to have the capacity to develop large and release and is more effective in this regard than the forelimb
force and high power (Payne et al., 2005). SDF (Brown et al., 2003).
Adductor has two parts (magnus and brevis) that are united, and Flexor digitorum profundus (DDF) has three parts. Flexor digitorum
so are considered together in Table 7.1. Like semimembranosus, the medialis and tibialis caudalis are characterized by small volumes
adductor is a large muscle characterized by a wide range of fascicle and short to medium length fibers (Fig. 7.1, Table 7.1), whereas
lengths (80–390 mm) and it is estimated to produce large force and flexor digitorum lateralis has a larger volume and short fibers (3–
high power. 55 mm) embedded within large amounts of aponeurotic tendi-
Quadriceps femoris has four parts, of which the biarticular head, nous tissue that give the muscle belly a striated appearance on
rectus femoris, is notable for its large PCSA (220 cm2) as a conse- gross morphological examination (Payne et al., 2005). This muscle
quence of its large volume (2161 cm3). The fiber lengths are short is estimated to have the capacity to develop high force (21100 N)
(<179 mm). Quadriceps assists in support and stabilization of the but relatively low power (157 W) (see Table 7.3). Although the
stifle joint, which is the key joint in the hind limb stay apparatus. short muscle fibers and large PCSA are not compatible with active
Details of this mechanism are given later. Rectus femoris may also shortening of the musculotendinous unit, this architecture is well
play a role in hip flexion. suited to work isometrically during elastic energy storage and
Tensor fascia latae is estimated to be capable of producing moder- release.
ate force and power (Fig. 7.1, Table 7.1). The tendon is relatively Flexor digitorum lateralis is particularly effective in this regard, and
long (227 mm) but with a small CSA (1.35 cm2). Electromyo- more so than the humeral, radial or ulnar head of DDF in the
graphic studies indicate that this muscle is active during late swing forelimb (Brown et al., 2003; Payne et al., 2005). The DDF tendon
and early stance (Robert et al., 1998). It is thought to play a role in has the largest cross-sectional area of the distal limb tendons
stabilizing the stifle during the stance phase (Tokuriki & Aoki, (3.64 cm2). Stress in the flexor digitorum lateralis tendon has been
1995). estimated to be as high as 105 MPa (Ker et al., 1988) or as low as
Tibialis cranialis is a small muscle with a wide range of fiber 40–50 MPa (Biewener et al., 1998). Tendons fail at stresses of
lengths (17–218 mm) and highly pennate fibers (41°) capable of approximately 120 MPa (Zajac, 1989), indicating that flexor digito-
generating fairly high force but low power (Table 7.1). It is notable rum lateralis may be operating close to, and may even exceed, its
for having a short (92 mm), thick (CSA, 2.61 cm2) tendon of attach- limit during isometric muscular contractions that generate high
ment (Payne et al., 2005). Fibers in the proximal part of this muscle forces. In spite of this prediction, digital flexor tendon injuries do
are longer than those in the distal part (Crook et al., 2008) and it not occur frequently in the equine hind limbs.
is the only muscle in the distal hind limb that has long fibers. Extensor digitorum longus is a small muscle that generates a moder-
Gastrocnemius (Fig. 7.1, Table 7.1) is the largest muscle of the ate force but little power (Fig. 7.1, Table 7.1). Its long (472 mm)
distal limb. The medial and lateral heads combined have a volume tendon has a small cross-sectional area (1.13 cm2) and experiences
of 1543 cm3 and mass 1625 g (Payne et al., 2005), with the lateral low stress. Estimates of peak stress range from 14.4 MPa (Payne
head being larger than the medial head but having shorter fiber et al., 2005) to 36 MPa (Ker et al., 1988).
lengths (Crook et al., 2008). The muscle belly contains extensive
tendinous bands organized in series and in parallel and united by
short (48–56 mm), pennate muscle fibers (see Table 7.3) (Payne The stay apparatus
et al., 2005). This architecture is consistent with a capacity for eco-
nomical force generation via tendinous stretch and recoil within the In the standing horse, the body weight acts through the hip joint
muscle belly. As a consequence of the fiber pennation, the lateral and is counteracted by the ground reaction force acting through the
head has the largest PCSA of all the hind limb muscles (644 cm2). hoof. The net effect is to exert a compressive force on the hind limb.
The estimated force-generating capacity of gastrocnemius is quite The joints are maintained in extension by the action of the extensor
high (8930 N) and is combined with a relatively large power- musculature. In general, the magnitude of extensor muscle force
generating ability (244 W). The two heads have a thick (3.32 cm2) required to maintain the standing posture increases with body mass
common (calcanean) tendon. Stress in the calcanean tendon has and with joint angulation. Thus, smaller animals tend to adopt a
been estimated as 47 MPa (Ker et al., 1988), 30 MPa (Biewener crouched posture with more flexion of the joints, whereas larger
et al., 1998) and 27 MPa (Payne et al., 2005). The gastrocnemius and animals stand with more upright limb angulations to reduce
cranial-tibial muscles may also play a role in centering the line of loading of the extensor musculature (Biewener, 1989). Horses

130
 Sagittal plane analysis of hind limb kinematics and kinetics

spend a considerable proportion of their time standing, which is

facilitated by the presence of the passive stay apparatus that uses
tendoligamentous structures to reduce the muscular activity
required to maintain the standing posture (Dyce et al., 1996).
The reciprocal apparatus is responsible for passively stabilizing
the hind limb during standing. Extension of the stifle joint, which
is the key to hind limb stabilization, is achieved by hooking
the patella over the medial femoral trochlea (Sack, 1989). When
the trochlea protrudes between the middle and medial parts of the
patellar ligament, it holds the stifle in extension by mimicking the
action of quadriceps femoris. Tension in the largely tendinous SDF
prevents tarsal flexion and the more distal joints are stabilized pas-
sively by a system of ligaments and tendons. The equine patella is
enveloped by the parapatellar fibrocartilage, which gives attachment
to all parts of quadriceps as well as to tensor fascia latae, biceps femoris,
gracilis and sartorius. During quiet standing with the patellar locking
mechanism engaged, the only muscular force needed to stabilize
the stifle is provided by tonic, low-level activity in vastus medialis,
which inserts on the medial aspect of the parapatellar fibrocartilage.
Tension in this muscle is estimated to be only 2% of the force that
would be needed without the patellar locking mechanism (Schuur-
man et al., 2003).
Upward fixation of the patella is an inability to disengage the
patellar locking mechanism. It may be mild causing a slight hesita-
tion in hind limb protraction or it can be severe enough to fix the
entire hind limb in extension, effectively precluding locomotion.
A B
Contributing factors include shrinking of the patellar fat pad in
underweight animals, abnormal coordination of muscles, such as
vastus lateralis, that actively disengage the patella (Wentink, 1978a) Fig 7.2  Two methods of measuring joint angles of the hind limbs with the
or spastic activity in vastus medialis that precludes disengagement measured angles being represented by black arcs. Left: measurement of the
(Schuurman et al., 2003). angle between the proximal and distal segments comprising the joint on
the anatomical flexor aspect. Right: measurement of the angle by which
the distal segment differs from alignment with the proximal segment,
deviation toward the flexor aspect is negative (−), deviation toward the
Sagittal plane analysis of hind limb extensor aspect is positive (+).
kinematics and kinetics

Although kinematic gait analysis of the equine hind limb has been
performed since the early part of this century (Walter, 1925; Krüger,
1938), the development of the computer gave new impetus to this are required. Typical marker configurations involve either placing a
field of research, both in the area of clinical applications (Fleiss marker over the center of rotation of each joint or aligning two
et al., 1984; Kobluk et al., 1989; Martinez-del Campo et al., 1991; markers along the long axis of each segment (Fig. 7.3). Intra-limb
Back et al., 1995a, 1995b) and in computer simulation studies (van coordination patterns can be visualized using stick figures or by
den Bogert & Schamhardt, 1993). The reciprocal apparatus, which joint angle–time graphs (Fredricson & Drevemo, 1972; Martinez-
couples stifle and tarsal joint motion, has been studied extensively del Campo et al., 1991; Holmström et al., 1994; Back et al., 1994;
(Strubelt, 1928; Molenaar, 1983; Wentink, 1978b; van Weeren Hodson et al., 2001, Dutto et al., 2006).
et al., 1990). Back et al. (1995b) standardized the graphical presen- Errors introduced into kinematic data due to skin movement
tation of kinematic data of the equine hind limb using joint angle– relative to the underlying skeletal landmarks (van Weeren et al.,
time diagrams in a large group of horses. 1992) may be small enough to be neglected on the crural and
The hip joint is the pivot point for rotation of the hind limb, and metatarsal segments, but are large enough to cause obvious changes
the ball and socket construction allows some motion outside of the in sagittal plane kinematics on the pelvic, thigh and pastern seg-
sagittal plane. The more distal joints are constrained to move pri- ments (Fig. 7.4). Errors as large as 15° in stifle angle and 30% in
marily in a sagittal plane with only small amounts of abduction– moment arm of gastrocnemius have been attributed to the effects of
adduction and internal–external rotation (Lanovaz et al., 2002). skin displacement in a walking pony (van den Bogert et al., 1990).
Thus kinematic analysis in the two-dimensional sagittal plane cap- Correction algorithms have been developed for many of the ana-
tures most of the kinematic information describing hind limb tomical locations that are commonly used for marker placement for
movement patterns. Hind limb joint angles may be measured in two-dimensional analysis of hind limb kinematics (van Weeren
several ways (Fig. 7.2): between the proximal and distal segments et al., 1992).
on the anatomical flexor aspect; as the angle by which the distal Kinematic information can be combined with ground reaction
segment deviates from alignment with the proximal segment; or as forces and morphometric data for the equine hind limb segments
some variation of these methods. The angle may be expressed in (Buchner et al., 1997; Nauwelaerts et al., 2011) (Fig. 7.5, Table 7.2)
absolute terms or it may be normalized to the standing angle, the to calculate net joint moments and net joint powers using inverse
angle at ground contact or the average angle during the stride (Mul- dynamic analysis (see Chapter 19 for details). This method offers a
lineaux et al., 2004). Three-dimensional kinematics of the tarsal more complete description of the functional responsibilities of the
joint have been measured using bone-fixed markers (Lanovaz et al., joints. In the equine hind limb, inverse dynamic analysis has been
2002) and will be described at the end of this chapter. used to improve understanding of joint function during walking
When skin markers are used to represent sagittal plane motion (Colborne et al., 1997a, 1997b; Clayton et al., 2001; Dutto et al.,
of the hind limb segments, a minimum of two markers per segment 2006) and trotting (Clayton et al., 2002; Dutto et al., 2006).

131
 7 Hind limb function

1 Fig 7.3  Locations of skin-fixed markers overlying bony landmarks in two

2 2 marker-placement schemes. Left: markers are placed over the centers of
joint rotation and the limb segments are represented by lines joining the
markers. Right: two markers aligned along the long axis of each segment
1 are joined to represent the segment and adjacent segments intersect at the
joints. The hoof may be represented by different combinations of markers
with or without radiographic identification of the center of rotation of the
DIP joint. Anatomical locations of the markers on the hind limb: 1, tuber
coxae; 2, proximal femur at cranial greater trochanter; 3, distal femur at
lateral epicondyle; 4, proximal tibia at fibular head; 5, distal tibia at lateral
malleolus; 6, talus; 7, proximal metatarsus; 8, distal metatarsus; 9, hoof at
coronary band; 10, toe of hoof; 11, mid-lateral hoof; 12, heel of hoof;
3 13, proximal pastern.

3
4

5
6
7

8 8
9
13
9
A 11 B 10 12

30 50
Flex Hip Flex Stifle

20
30
Joint angle (degrees)
Angle (degrees)

10
10
0

-10
-10

Ext Ext
-20 -30
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

60 80
Flex Tarsus Flex Hind
60 Fetlock
40
Joint angle (degrees)

Joint angle (degrees)

40

20 20

0
0
-20
Ext Ext
-20 -40
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)
Fig 7.4  Mean angle–time diagrams for the joints of the hind limb in a group of horses trotting on a treadmill at 3 m/s before (red line) and after (green line)
correction for skin displacement. The horizontal zero line indicates the joint angle of the horse standing square. The vertical dashed line marks the transition
from stance to swing phase. At the fetlock there is no difference after correction for skin displacement.
Reprinted from Back, W., Schamhardt, H.C., Savelberg, H.H.C.M., et al., 1995b. How the horse moves: significance of graphical representations of equine hind limb kinematics. Equine Vet. J. 27,
39–45, with permission from the Equine Veterinary Journal.

132
 Sagittal plane analysis of hind limb kinematics and kinetics

The trot breeds of horses (Clayton, 1994a; Holmström et al., 1994; Back
et al., 1995b; Galisteo et al., 1997; Galisteo et al., 1998; Cano et al.,
The trot is a symmetrical, two-beat gait with the limbs moving by 1999; Cano et al., 2001; Nicodemus & Holt, 2006; Nicodemus &
diagonal pairs. The diagonal stance phases are usually separated by Booker, 2007). The following paragraphs describe sagittal plane
aerial or suspension phases, so the trot is classified as a running or kinematics and kinetics of the joints of the hind limb during trot-
leaping gait in which the body mass is modeled as a spring mass ting and indicate the effects of skin displacement on the amplitude
system. Since the trot is the most important gait for lameness detec- and pattern of joint angular displacement (Fig. 7.4).
tion, it has been the focus of many kinematic studies in various In Warmblood horses trotting at 4.0 m/s, stride duration is
0.67 s, and hind stance duration occupies about 40% of stride dura-
tion (Table 7.3) (Back et al., 1995b). Stride and stance durations
are negatively correlated with velocity (Robert et al., 2002).
Although hind limb swing duration is the same at walk and trot,
the limbs rotate through a larger range of motion at trot, requiring
greater force and impulse generation, and with more flexion of the
stifle and tarsal joints during the swing phase, which shortens the
limb pendulum and hastens protraction. As trotting speed increases,
range of motion increases in the hip and tarsal joints (Hoyt et al.,
2002).
As in the forelimb, gait efficiency in the hind limb is increased
by using the tendons as energy-conserving elastic springs (Dimery
et al., 1986). Energy is stored by stretching the tendons as the limb
is loaded and it is released during the unloading phase when it
contributes to propulsion and joint flexions during the swing phase.
1 In early swing, the release of elastic energy supports tarsal and stifle
flexion (Wentink, 1978b), which occur synchronously, since these
joints are coupled by the reciprocal apparatus. In general, electro-
2 myographic activity in the hind limb musculature is low during the
swing phase at trot (Wentink, 1978a; Jansen et al., 1992; Tokuriki
& Aoki, 1995), which supports the idea that movements in the
3 swing phase are heavily reliant on passive energy from elastic recoil
of the tendons. In the absence of ground reaction forces, resistance
to limb movements in the swing phase is due mainly to segmental
inertia. Flexion of all the hind limb joints as the limb is protracted
4 reduces the effective limb length and moment of inertia about the
hip joint.
5

Hip joint
6
The hip joint shows a more or less sinusoidal flexion and extension
Fig 7.5  Locations of skin markers used to locate the centers of mass of the pattern. In early stance, hip angle is constant during the impact
hind limb segments in Table 7.2. The segments are separated along the phase, after which the joint extends to reach maximal extension in
incision lines shown in red. terminal stance. The swing phase flexion cycle peaks just after mid-
Reprinted from Buchner, H.H.F., Savelberg, H.H.C.M., Scharmhardt, H.C., Barneveld, A., 1997. swing, which represents the position of maximal hind limb protrac-
Inertial properties of Dutch Warmblood horses. J. Biomech. 30 (6), 653–658, with permission tion. The hip then extends and retracts the limb prior to hoof
from Elsevier.. contact with the ground (Fig. 7.6). In the absence of correction for

Table 7.2  Hind limb segmental masses, densities, reference lines for division of segments (see Fig. 7.2), and position of segment centers of
mass in sagittal plane

Segment Mass (kg) Density Proximal Distance as % segment length

(g/cm3) and distal
x-axis y-axis
markers
Thigh 18.6 ± 2.3 1.05 ± 0.007 1–2 59 ± 8 −12 ± 4
Crus 8.3 ± 0.8 1.11 ± 0.019 2–3 38 ± 2 −8 ± 2
Metatarus 2.84 ± 0.22 1.28 ± 0.03 3–4 32 ± 4 −7 ± 2
Hind pastern 0.89 ± 0.04 1.23 ± 0.03 4–5 43 ± 5 −13 ± 4
Hind hoof 0.99 ± 0.17 1.18 ± 0.01 5–6 31 ± 5 −22 ± 4

Position of center of mass is determined first along the segmental x-axis (longitudinal, positive distally from the proximal reference marker toward the distal reference marker),
then shifted along the y-axis (perpendicular to x-axis, positive cranially). Distances are measured as a percentage of the segment length between the two reference markers.
Data are mean ± SD for 12 hind limbs of six Warmblood horses.
Reprinted from Muir, W.W., Hubbell, J.A.E., 2008. Equine Anaesthesia, monitoring and emergency therapy, 2e, Mosby, with permission from Elsevier.

133
 7 Hind limb function

Table 7.3  Hind limb kinematic variables for horses walking at 1.6 m/s, trotting at 4.0 m/s (Back et al., 1996b), and cantering at 7.0 m/s
(Back et al., 1997)

Walk Trot Trailing hind limb Leading hind limb

canter canter
Stride duration (s) 1.09 ± 0.06* 0.67 ± 0.03* 0.553 ± 0.02 0.547 ± 0.02
Stance duration (s) 0.69 ± 0.04* 0.27 ± 0.01* 0.177 ± 0.01# 0.185 ± 0.01#
Stance duration (%) 63.4 ± 1.4* 40.5 ± 1.9* 32.1 ± 1.3 #
33.9 ± 1.3#
Swing duration (s) 0.40 ± 0.03 0.40 ± 0.03 0.376 ± 0.02 0.362 ± 0.02

Pro-/retraction
Angle of max protraction (°) 23.1 ± 1.8* 21.6 ± 1.3* 22.7 ± 1.4# 35.8 ± 1.7#
Angle of max retraction (°) −23.6 ± 1.6* −26.6 ± 1.5* −37.6 ± 1.2# −29.7 ± 1.0#
Max pro/retraction range (°) 46.7 ± 2.9* 48.1 ± 1.6* 60.3 ± 1.7# 65.5 ± 1.6#

Hip joint
Angle of max extension (°) 68.9 ± 3.8* 67.8 ± 3.9* 66.7 ± 4.4 73.8 ± 3.2
Angle of max flexion (°) 93.7 ± 3.2* 91.1 ± 3.6* 91.0 ± 3.6 96.3 ± 3.2
Range of motion (°) 24.8 ± 1.9* 23.3 ± 1.8* 24.3 ± 1.9 22.5 ± 1.8

Stifle
Angle of max extension (°) 7.0 ± 4.9* 11.0 ± 4.0* 13.5 ± 4.6# 2.3 ± 5.1#
Angle of max flexion (°) 46.1 ± 5.9* 58.3 ± 5.1* 55.5 ± 3.0 57.9 ± 4.3
Range of motion (°) 39.1 ± 3.7* 47.3 ± 3.8* 42.0 ± 3.9 #
55.5 ± 4.7#

Tarsus
Angle of max extension (°) 10.4 ± 3.5 10.3 ± 2.9 7.0 ± 2.6 8.1 ± 3.0
Angle of max flexion (°) 46.0 ± 0.1* 65.7 ± 5.1* 56.0 ± 8.0 #
64.5 ± 8.6#
Range of motion (°) 35.6 ± 4.5* 55.4 ± 5.3* 49.0 ± 8.3# 56.5 ± 7.6#

Metatarsophalangeal
Angle of max extension (°) −37.0 ± 4.2* −53.0 ± 4.0* −60.6 ± 3.4# −53.9 ± 3.8#
Angle of max flexion (°) 34.8 ± 9.1* 32.0 ± 7.5* 35.6 ± 8.9# 43.8 ± 7.3#
Range of motion (°) 71.8 ± 9.0* 85.0 ± 7.7* 96.1 ± 9.1 97.7 ± 7.0

*Indicates variables that differ significantly between walk and trot (p <0.05).
#
Indicates variables that differ significantly between trailing and leading limbs at canter (p <0.05).
Angles are expressed relative to 180° alignment of the proximal and distal segments; flexion angles are positive, extension angles are negative. Protraction and retraction angles
are measured relative to the vertical: protraction positive, retraction negative. Values are mean ± SD.
Reproduced from Back et al. (1996b, 1997).

skin displacement, the pattern of hip joint motion hardly changes range of motion of around 11° during trotting (Gomez Alvarez
but the range of motion is reduced by about half resulting in an et al., 2008).
underestimation of the range of flexion–extension by as much as There is an extensor moment at the hip joint through most of
30° (Fig. 7.4) (Back et al., 1995b). stance with two bursts of energy generation in the extensor muscu-
The hip joint–time curve and the cranio-caudal oscillations of a lature that push the trunk forward over the grounded hind hoof
marker on the distal metatarsus have a similar pattern of movement (Fig. 7.6). The net joint moment moves to the flexor aspect during
(Back et al., 1995b) indicating that the entire hind limb can be breakover and remains there through the first one third of swing.
considered to move like a pendulum with a rotation point in the Positive work is done as hind limb protraction is initiated by con-
acetabulum. Maximal protraction occurs almost 10% later in the centric action of the hip flexors followed by passive recoil of elastic
stride for the hind limb compared with the forelimb. Pelvic rotation structures, such as the tendon of tensor fascia latae (Tokuriki & Aoki,
contributes little to hind limb motion in the sagittal plane as the 1995) and fibularis tertius (Wentink, 1978b). Energy generation in
pelvis maintains a fairly constant angle relative to the horizon. the hip flexors in early swing increases linearly with trotting velocity
However, pelvic rotation in the horizontal plane is larger with a (Clayton et al., 2002). The net joint moment is low during

134
 Sagittal plane analysis of hind limb kinematics and kinetics

120 130

110 110
Angle (deg)

Angle (deg)
100 90

90
0 20 40 60 80 100 70
0 20 40 60 80 100
1.0
0.8 0.6
0.6
Moment (Nm/kg)

0.4 0.3

Moment (Nm/kg)
0.2
0.0 0.0
20 40 60 80 100 20 40 60 80 100
-0.2
-0.4 -0.3
-0.6
-0.60
2.5

2 1.4
1.2
1.5
Power (W/kg)

1.0
Power (W/kg)

1
0.8
0.5 0.6
0 0.4
20 40 60 80 100 0.2
-0.5
Time (% stance) 0.0
A 0 20 40 60 80 100

B Time (% swing)

Fig 7.6  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the hip joint during the stance phase (left) and swing phase (right) at the trot.

midswing, then in late swing the extensor moment increases as the 2002; Dutto et al., 2006). In early swing, an extensor moment cor-
hip extensor muscles do positive work to initiate hind-limb retrac- responds with a period of energy absorption as the stifle extensors
tion in preparation for ground contact (Khumsap, 2002). Gluteus control the amount of flexion. In late swing, a flexor moment is
medius, biceps femoris and semitendinosus are active at this time present as energy is absorbed by the stifle flexors, which are acting
(Robert et al., 1998). The magnitude of this late swing phase burst eccentrically to control large bursts of joint extension (Fig. 7.7)
increases with velocity squared (Clayton et al., 2002), which may (Khumsap, 2002). The early swing phase burst of energy absorption
reflect the need to overcome the forward kinetic energy increases linearly with velocity, whereas the later burst increases
( 1 2 mv2) of the limb. with velocity cubed (Clayton et al., 2002).

Stifle joint Tarsal joint

After initial ground contact the stifle joint flexes rapidly during The tarsal joint flexes rapidly in early stance, after which the flexion–
loading then shows only minor angular changes through the extension pattern in midstance varies between horses; some have a
remainder of the stance phase (Fig. 7.7). Without correction for skin second flexion phase, some show a fairly constant angle through
displacement there are errors in the amount of stifle flexion in late midstance, and in others the joint extends after the initial phase of
stance and in the time of maximal stance phase flexion (Fig. 7.4). flexion. Maximal tarsal extension occurs around the end of stance
In the swing phase, peak flexion of the stifle in midswing tends to and is followed by a swing phase flexion cycle peaking in midswing
be underestimated by skin markers without correction for skin via an inflection in early swing that occurs just after passing its angle
displacement (Back et al., 1995b). at square stance (Fig. 7.8) (Kobluk et al., 1989; Holmström et al.,
The net joint moment is on the flexor (caudal) side of the stifle 1994; Back et al., 1995b; Khumsap et al., 2003). Correction for skin
through most of the stance with energy generation on the flexor displacement (Fig. 7.4) produces only minor differences in the
side in early stance and midstance due to the action of the ham- tarsal flexion–extension curves (van Weeren et al., 1990; Back et al.,
string muscles that provide support and propulsion (Khumsap, 1995b). Even after correction for skin displacement, the tarsus

135
 7 Hind limb function

170 170

165 150
Angle (deg)

Angle (deg)
160 130

155 110
0 20 40 60 80 100 0 20 40 60 80 100

0.4 0.15
0.2 0.10
Moment (Nm/kg)

0.05

Moment (Nm/kg)
0.0
20 40 60 80 100 0.00
-0.2 20 40 60 80 100
-0.05
-0.4
-0.10
-0.6 -0.15
-0.20
3.0

2.0 0.1
0.0
Power (W/kg)

1.0 -0.1 20 40 60 80 100

-0.2
Power (W/kg)

0.0 -0.3
20 40 60 80 100 -0.4
-1.0 -0.5
-0.6
-2.0
-0.7
A Time (% stance) -0.8
Time (% swing)
B

Fig 7.7  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the stifle joint during the stance phase (left) and swing phase (right) at the trot.

appears to extend more than the stifle in late stance due to stretch- The passive spring action and coupling mechanism in the tarsus
ing of fibularis tertius, especially at faster speeds. In attempting to can be reproduced in a horse under general anesthesia in lateral
explain these discrepancies, Molenaar (1983) assigned functional recumbency. The joint moves rapidly into maximal flexion or
significance to the axial and abaxial insertions of fibularis tertius. extension as it passes through a critical mid-point angle (Alexan-
The net joint moment is on the extensor aspect of the tarsus der & Trestik, 1989). This snapping action is a consequence of the
throughout stance, with the value peaking in midstance. The power eccentric attachments of the tarsal collateral ligaments relative to
profile varies somewhat according to the kinematic pattern in mid- the center of rotation of the tarsocrural joint (Updike, 1984;
stance but consistent features include a burst of energy absorption Badoux, 1987). Inflections in the joint angle–time curves of the
in the first half of stance and a burst of energy generation later in tarsal joint in vitro indicate the positions at which a labile equilib-
stance that may represent elastic recoil (Fig. 7.8) (Dutto et al., rium is passed and the joint rapidly flexes or extends. Evidence of
2006). The net joint moment is on the flexor aspect in the first half this phenomenon is also seen in vivo; three-dimensional analysis
of swing. It corresponds with a phase of elastic energy release in of tarsal kinematics shows that craniocaudal translation becomes
fibularis tertius, which was stretched in late stance. There may also decoupled from tarsal flexion–extension as the joint passes
be active energy generation in the tarsal flexors and/or digital exten- through −50° of flexion and becomes coupled again later in swing
sors, such as the long digital extensor, which shows electromyo- as the joint passes back through that same angle (Lanovaz et al.,
graphic activity at this time (Jansen et al., 1992). The net joint 2002).
moment moves to the extensor (caudal) side of the tarsus in the
second half of swing, when energy generation in the tarsal extensors
actively extends the joint and lowers the hoof in preparation for Metatarsophalangeal (MTP) joint
ground contact. This is in contrast to the forelimb in which carpal After initial ground contact, the MTP joint extends and reaches
extension in late swing is due to inertia and is controlled by the maximal extension at midstance. The joint then flexes through the
carpal extensor muscles. The early swing phase burst of energy remainder of stance to reach a flexion peak in early swing through
generation increases linearly with trotting velocity, whereas the the passive action of the SDF tendon (Molenaar, 1983). After a
burst of energy generation in late swing increases with velocity slight extension in midswing there is a second flexion peak that
cubed (Clayton et al., 2002). occurs close to the time of maximal flexion of the stifle and tarsal

136
 Sagittal plane analysis of hind limb kinematics and kinetics

180 170

160
150
Angle (deg)

Angle (deg)
140

120 130

100
0 20 40 60 80 100 110
0 20 40 60 80 100
1.2
0.15
1.0
0.10
Moment (Nm/kg)

0.8

Moment (Nm/kg)
0.6 0.05
0.4
0.00
0.2 20 40 60 80 100
-0.05
0.0
20 40 60 80 100 -0.10
-0.2
-0.15
4.0
0.3
Power (W/kg)

2.0
Power (W/kg)

0.2

0.0
20 40 60 80 100 0.1

-2.0
Time (% stance) 0.0
A 0 20 40 60 80 100

B Time (% swing)

Fig 7.8  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the tarsal joint during the stance phase (left) and swing phase (right) at the trot.

joints. At the end of swing, when the stifle and tarsal joints are extension is during breakover in the terminal part of stance. In the
extending, the MTP joint extends under the control of the digital swing phase, the DIP joint flexes rapidly via a small flexion peak to
flexor muscles (Fig. 7.9). No differences are found between the reach peak flexion at midswing. It then extends to initial ground
curves before and after correction for skin displacement when contact (Fig. 7.10) (Back et al., 1995b).
the pastern is treated as a rigid segment (Fig. 7.4) (Back et al., The net joint moment is on the flexor (plantar) aspect throughout
1995b). stance. Some horses show a little negative work in early stance, the
The net joint moment is on the flexor (plantar) aspect of the MTP presence of which may be related to hoof angulation, then in late
joint throughout the stance phase, where it acts to support and stance there is a burst of energy absorption on the flexor aspect of
control the amount of extension. The power profile is typical of the DIP joint (Fig. 7.10). The net joint moment is on the extensor
elastic energy storage and release with a phase of energy absorption aspect of the DIP joint in early swing where it acts to control joint
followed immediately by an almost equal phase of energy genera- flexion during protraction (Khumsap, 2002).
tion (Fig. 7.9) (Khumsap, 2002; Dutto et al., 2006). At the begin-
ning of the swing phase, MTP flexion is actively supported by
activity of the DDF muscle (Jansen et al., 1992), but the net joint Hoof interaction with the ground
moment is on the extensor aspect at this time indicating that the Although the fore and hind distal limbs are anatomically similar,
extensors are controlling MTP flexion. The net moment moves to they have different angles, velocities and accelerations during the
the flexor aspect in the second half of swing to control MTP exten- impact phase (Back et al., 1995d). The hind hooves have a more
sion by the long digital extensor tendon in preparation for ground exaggerated heel first contact and so take longer to become flat on
contact (Khumsap, 2002; Dutto et al., 2006). the ground (Merkens & Schamhardt, 1993; Schamhardt & Merkens,
1994; Back et al., 1995d). Furthermore, the angle of the metatarsus
at impact is smaller than that of the metacarpus (Clayton, 1994a;
Distal interphalangeal (DIP) joint Back et al., 1995d).
The DIP joint flexes during the rapid loading phase at the begin- Within approximately 3% of stride duration, hoof angle, vertical
ning of stance reaching peak flexion before midstance. Maximal velocity and vertical acceleration of the hoof are zero, but it is not

137
 7 Hind limb function

240 200

230
Angle (deg)

220 180

Angle (deg)
210

200 160

190
0 20 40 60 80 100
140
0.2 0 20 40 60 80 100
0.0
20 40 60 80 100 0.008
Moment (Nm/kg)

-0.2
-0.4 0.006
-0.6
0.004
-0.8

Moment (Nm/kg)
-1.0 0.002

-1.2 0.000
20 40 60 80 100
-0.002
5
4 -0.004
3
-0.006
Power (W/kg)

2
1
0 0.05
-1 20 40 60 80 100
-2 0.00
-3 20 40 60 80 100
-0.05
-4
Power (W/kg)

Time (% stance) -0.10

A
-0.15
-0.20
-0.25
-0.30

B Time (% swing)

Fig 7.9  Joint angle (above, extension positive), net joint moment (center, extensor moment positive) and net joint power (below, power generation
positive) for the metatarsophalangeal joint during the stance phase (left) and swing phase (right) at the trot.

until approximately 6% of stride duration that the horizontal veloc-

ity and acceleration decline to zero.
The walk
The fore hoof contacts the ground with a higher vertical velocity The walk and trot are both symmetrical gaits, but they differ in
and bounces more at impact, whereas the hind hoof has a higher inter-limb coordination. The walk is a four-beat gait with a lateral
horizontal velocity and shows more sliding (Back et al., 1995d; sequence of limb placements in which there is always at least one
Johnston et al., 1995, 1996). Impact is associated with high fre- fore and one hind foot on the ground. Since there is no suspension
quency oscillations that are still visible in the MTP joint acceleration– or aerial phase, the walk is classified as a stepping gait, which affects
time diagram until 14% of stride duration (Fig. 7.11) (Schamhardt the patterns of limb loading and force distribution.
& Merkens, 1994; Back et al., 1995d). Impact oscillations are damp- In the literature stick figures and joint angle–time diagrams for
ened by the hoof and other structures as they are transmitted proxi- the hind limbs during walking have been reported for individual
mally. The net effect of changes in hind limb joint angles during horses (Walter, 1925; Krüger, 1937, 1938; Fleiss et al., 1984) and
early stance is to shorten the distance from hip to hoof (Hjertén groups of horses of various breeds (Back et al., 1996, Galisteo
et al., 1994), which also contributes to the shock-absorbing capac- et al., 1996; Hodson et al., 2001; Nicodemus & Holt, 2006; Mar-
ity of the hind limb. tuzzi et al., 2007). Walk and trot kinematics are similar with
Kinematic differences between the distal fore and hind limb regard to the intra-limb coordination pattern (Fig. 7.12), stance
might explain the lower incidence of chronic lameness in the hind distance and swing duration but since the trot is a faster gait than
limbs compared with the forelimbs (Stashak, 1987). Repetitive walk, the limb shows the same movements in a shorter time (Back
impulsive loading in combination with rapid oscillations in the et al., 1996). The hind limb is less loaded during stance and has
joint, even within physiological limits, plays a mechanical role in smaller vertical ground reaction forces at walk compared to trot
the development of osteoarthrosis (Radin et al., 1991). (Ueda et al., 1981; Merkens & Schamhardt, 1994; Schamhardt &

138
 Sagittal plane analysis of hind limb kinematics and kinetics

210 200

190
190
Angle (deg)

Angle (deg)
180

170
170
160

150 150
0 20 40 60 80 100 0 20 40 60 80 100

0.05 0.004
0.00
20 40 60 80 100
Moment (Nm/kg)

-0.05 0.002

Moment (Nm/kg)
-0.10
0.000
-0.15 20 40 60 80 100
-0.20
0.002
-0.25
0.004
0.5
0.0
-0.5 20 40 60 80 100 0.08
Power (W/kg)

-1.0
-1.5 0.04
Power (W/kg)

-2.0
-2.5 0.00
20 40 60 80 100
-3.0
-0.04
A Time (% stance)

-0.08
B Time (% swing)

Fig 7.10  Joint angle (above, extension positive), net joint moment (center, extension moment positive) and net joint power (below, power generation
positive) for the distal interphalangeal joint during the stance phase (left) and swing phase (right) at the trot.

Merkens, 1994), which is associated with less MTP extension and change with velocity (Khumsap et al., 2001a). The left and right
less tarsal flexion during stance. Push-off is less powerful in walk hind stance phases overlap by 12% of stride duration and, during
than trot. this time, the vertical and braking longitudinal forces increase in
At a walking speed of 1.6 m/s, hind stance duration occupies the hind limb that is accepting weight. The first vertical force peak
63.4% of stride duration (Table 7.3) (Back et al., 1996). Stance coincides with lift-off of the contralateral hind limb. The dip in
duration is negatively correlated with walking velocity (McLaughlin vertical force occurs during tripedal support (one hind, two fore),
et al., 1996; Khumsap et al., 2001a). Although swing duration is then the second peak is during the support by a lateral pair of limbs.
the same for walk and trot, the limbs swing through a smaller The midstance position, marked by the hind hoof being vertically
range of motion at walk, requiring less force and impulse genera- beneath the hip joint, occurs at 28% stride, which is close to the
tion. The stifle and tarsal joints are less flexed during the swing time that the longitudinal force changes from braking to propulsive
phase at walk, which results in a longer pendulum and a slower at 30% stride (Hodson et al., 2001).
forward swing. Flexion and extension of the hip joint are responsible for protrac-
During walking, the vertical ground reaction force of the hind tion and retraction of the entire hind limb during walking. Maximal
limb shows two peaks separated by a dip. As walking velocity protraction occurs shortly before the end of swing, and maximal
increases, the first vertical force peak increases, while the dip retraction is during breakover. The thigh, crural and metatarsal seg-
decreases and becomes more distinct. This indicates that, as the ments rotate caudally with fairly constant angular velocities through
horse walks faster, the limb is loaded more in early stance and then the middle part of stance. As the limb is loaded in early stance, the
unloaded to a greater extent in midstance, which is partly due to stifle, tarsal and DIP joints flex and the MTP joint hyperextends.
rebound of the more heavily loaded limb spring. At very low veloci- After an initial flexion, the tarsus extends through the remainder of
ties, the vertical dip almost disappears (Khumsap et al., 2001a). The stance, while the stifle flexes a little after midstance. The MTP joint
hind limb longitudinal ground reaction force has a braking phase extension pattern during the stance phase at walk has two extension
followed by a propulsive phase that peaks just before the start of peaks coinciding with those in the vertical ground reaction force.
breakover. At walk, only one limb provides braking and only one These peaks are more pronounced at a faster walk and ‘melt’ into
limb provides propulsion at any time (Merkens et al., 1986). The one extension peak at the trot, similar to vertical ground reaction
timing of the peaks in the longitudinal force do not appear to be force recordings at walk and trot (Back et al., 1996; Niki et al.,
related to kinematic events and the magnitudes of the peaks do not 1982).

139
 7 Hind limb function

60 start of the swing phase, the hip, stifle and tarsal joints flex to raise
Flex Fetlock angle and protract the distal limb. The crural, pastern and hoof segments
start to swing cranially soon after the hind hoof lifts off (around
40
70% stride), which coincides with the first swing phase flexion peak
Joint angle (degrees) at the MTP joint. Peak flexions of the stifle, tarsal and DIP joints
20 occur around 80% stride, followed by peak hip flexion at 85%
stride. The thigh and cannon segments reverse their direction of
rotation and begin to swing caudally at 90% stride followed by the
0
crural segment at 95% stride. Maximal hind-limb protraction occurs
at 97% stride.
-20 The net joint moment is on the caudal/plantar side of all hind
limb joints at the start of stance when the limb is being actively
Ext
retracted. It moves to the cranial/dorsal side around 24% stride at
-40 the hip and stifle and in terminal stance at the more distal joints.
0 20 40 60 80 100
A It remains on the cranial/dorsal side of all joints during the first
Time (% of total stride)
half of swing when the proximal joints provide active limb pro-
traction, then moves to the caudal/plantar aspect to initiate retrac-
tion prior to ground contact. The hip joint is the main source of
1500 energy generation during walking. It is assisted by the tarsal joint,
Flex Fetlock velocity
in both stance and swing phases, and by the MTP joint during
1000 stance. The DIP joint acts as an energy damper during stance,
Joint angle velocity (degrees/s)

whereas the stifle joint absorbs almost equal amounts of energy

500 in stance and swing phases. The DIP and MTP joints absorb
energy as the limb is protracted and retracted during the swing
0 phase suggesting that their movements are driven by inertial
forces and are controlled by muscles acting eccentrically (Clayton
-500 et al., 2001).
An increase in walking velocity is associated with increases in
-1000 magnitudes of the net joint moments and peak powers at the hip,
Ext stifle, tarsal and MTP joints in early stance. Energy generation at the
-1500 hip makes the largest contribution to the increase in velocity assisted
0 20 40 60 80 100 by the tarsus, which contributes to the forward and upward accelera-
B Time (% of total stride) tion of the limb into the swing phase (Khumsap et al., 2001b). Peak
powers and energy bursts are considerably lower at walk than at trot
(Clayton et al., 2001), which is similar to the situation in humans
(Chapman & Caldwell, 1983).
150000
Flex Fetlock acceleration
Joint angle acceleration (degrees/s2)

100000 The canter

The canter is an asymmetrical gait in which the footfalls of both the
50000 fore and hind limbs occur as couplets. The first limb of the couplet
to contact the ground is defined as the trailing limb, the second is
0 the leading limb. The sequence of footfalls is trailing hind limb,
leading hind and trailing forelimbs together, then the leading fore-
-50000 limb. An aerial phase usually follows lift-off of the leading forelimb.
The transition from a slow speed canter to the higher speed gallop
-100000 involves dissociation of the footfalls of the diagonal limb pair
Ext (leading hind and trailing fore).
-150000 In the literature, canter has been described at speeds ranging
0 20 40 60 80 100 from 3–11 m/s. The recordings were made under various circum-
C Time (% of total stride) stances from a collected canter of a mounted horse overground
to an extended canter of an unmounted horse on a treadmill
Fig 7.11  Metatarsophalangeal joint angle–time (above), angular velocity– (Clayton, 1994b; Corley & Goodship, 1994; Deuel & Park, 1990).
time (middle) and angular acceleration-time (below) graphs of the fore- and Canter kinematics have been illustrated using stick diagrams
hind-limb metatarsophalangeal joints for a group of horses trotting on a (Walter, 1925; Krüger, 1937, 1938) and diagrams illustrating the
treadmill. At bottom of graph, blue bar indicates stance phase, white bar coordination of inter-limb timing variables (Deuel & Park, 1990;
indicates swing phase.
Clayton, 1993, 1994b). Changes in temporal and linear kinematics
Reprinted from Back, W., 1995. Repetitive loading and oscillations in the distal fore and hind
induced by training have also been described (Corley & Goodship,
limb as predisposing factors for equine lameness, American Journal of Veterinary Research,
1994).
http://avmajournals.avma.org/loi/ajvr.
Back et al. (1997) compared the kinematics of the leading and
trailing hind limbs of Warmblood horses, cantering on a treadmill
at a speed of 7 m/s (Fig. 7.13), and found greater protraction in the
After the heels leave the ground at 55% stride, the thigh and leading hind limb, due to more hip flexion, and greater retraction
metatarsal segments reverse their direction of rotation and begin to in the trailing hind limb (Table 7.3). As the leading hind limb has
swing cranially during breakover, while the crus continues to rotate a somewhat larger total range of maximal protraction and retrac-
caudally with almost constant angular velocity. Maximal hind limb tion, the tarsal joint has to be more flexed to allow the distal limb
retraction occurs just before lift-off at 63% stride and coincides with to reach further in the same swing time. As a consequence of the
the reversal in the direction of rotation of the thigh segment. At the asymmetry in inter-limb timing between trailing and leading limbs

140
 Sagittal plane analysis of hind limb kinematics and kinetics

100 60 Fig 7.12  Mean joint angle–time diagrams of

Flex Hip Flex Stifle the hind limb joints of a group of horses
walking at 1.6 m/s (red line) and trotting at
4 m/s (green line) on a treadmill. The joint
90
angles are defined as zero when the adjacent
40
Angle (degrees)

Angle (degrees)
bone segments are aligned. Bars at bottom of
graph indicate stance and swing phases at walk
80 (red bars) and at trot (green bars).
Reprinted from Keg, P.R., Scharmhardt, H.C., van Weeran,
20 P.R., et al., 1996. The effect of the high palmar nerve block
70 and the ulnar nerve block on lameness provoked by a
collagenese-induced tendonitis of the lateral branch of
Walk Walk
Ext Trot Ext Trot the suspensory ligament, Veterinary Quarterly, 18, sup 2,
60 0 with permission from Taylor & Francis Ltd, http://www.
0 20 40 60 80 100 0 20 40 60 80 100 informaworld.com.
Time (% of stride) Time (% of stride)

70 40
Flex Tarsus Flex Hind fetlock
60 20
50
0
Angle (degrees)

Angle (degrees)

40
-20
30
-40
20

10 -60
Walk Walk
Ext Trot Ext Trot
0 -80
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

100 70 Fig 7.13  Mean joint angle–time diagrams of

Flex Hip Flex Stifle the trailing (red line) and leading (green line)
60 hind limbs of a group of horses cantering on a
90 treadmill at 7 m/s. Joint angles are defined zero
50
when the adjacent bone segments are aligned.
Angle (degrees)

Angle (degrees)

80 40 The bars at the bottom of the graphs show the

stance phase of trailing limb in red and stance
30 phase of leading limb in green.
70 20 Back, W., Hartman, W., Schamhardt, H.C., Bruin, G.,
Barneveld, A., 1995c. Kinematic response to a 70-day
10 training period in trotting Dutch Warmbloods. Equine Vet.
60
J. 18 (Suppl.), 127–131, with permission from the Equine
Trailing 0 Trailing
Ext Leading Ext Leading Veterinary Journal.
50 -10
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

70 60
Flex Tarsus Flex Hind fetlock
60 40

50 20
Angle (degrees)

Angle (degrees)

40 0

30 -20

20 -40

10 Trailing -60 Trailing

Ext Leading Ext Leading
0 -80
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

141
 7 Hind limb function

Medial Lateral

Abduction Adduction

Flexion

Tibial
BCS
Extension
Metatarsal
BCS

External rotation Internal rotation

A B

Fig 7.14  Caudolateral view of the right tarsus showing the local coordinate systems of the crural and metatarsal segments (left) and the three rotations
measured in three-dimensional analysis of the tarsus (right).
Reprinted from Lanovaz, J.L., Khumsap, S., Clayton, H.M., Stick, J.A., Brown, J., 2002. Three-dimensional kinematics of the tarsal joint at trot, Equine Vet. J., 34 (Suppl.), 308–313, with permission
from the Equine Veterinary Journal.

at canter, the hind footfalls occur as couplets. The trailing limb is to be coupled with flexion–extension; there is a small cycle of
the supporting limb and carries more load. abduction in stance and a larger cycle in swing with approximately
1° of abduction for every 3° flexion. In stance, there is a small
cycle of internal rotation followed by a small cycle of external rota-
Three-dimensional kinematics of the tarsal tion in swing (Fig. 7.15). Translational movements in the stance
phase consist of cranial then caudal translation, lateral then medial
joint at trot translation and distal followed by proximal translation. The same
patterns occur in the swing phase but the magnitudes are larger in
Three-dimensional tarsal kinematics have been measured using accordance with the greater angular excursions.
marker triads fixed rigidly to the tibia and third metatarsus (Fig. The tarsocrural joint is assumed to be the source of most of the
7.14) to describe three-dimensional motion in terms of three rota- motion of the tarsus. The trochlea of the talus and the tibial cochlea
tions (flexion–extension, abduction–adduction, internal–external act like threads on a bolt with the tibia following an almost circular
rotation) and three translations (mediolateral, craniocaudal and path along the talar ridges (Badoux, 1987). Experimental measure-
proximodistal) (Lanovaz et al., 2002). At a velocity of 2.8 m/s, ments are in general agreement with these predictions (Lanovaz
stance duration occupied 38% of the stride. Flexion–extension pat- et al., 2002). If the tarsocrural joint behaves as a pure screw joint and
terns and magnitudes were very similar to those recorded during is the only source of motion within the tarsus, all movements should
sagittal plane analysis, even though the flexion–extension axis is be highly coupled to the flexion–extension angle and any non-
defined differently in two-dimensional and three-dimensional coupled movement would indicate motion at other sites within the
analyses (Fig. 7.14). This is not surprising, however, since even joint complex. Abduction–adduction and the three translations are
large misalignments of the flexion–extension axis have little effect strongly coupled with flexion–extension but internal–external rota-
on the measured values when this is the dominant motion tion is uncoupled during weight-bearing. In the swing phase, around
(Ramakrishnan & Kadaba, 1991). The tarsus flexes during stance, 55% of stride duration, the cranial–caudal displacement becomes
reaches peak extension around the time of lift-off, then undergoes uncoupled from flexion–extension indicating that the third metatar-
a larger flexion cycle during swing. Abduction–adduction appears sus is displaced slightly cranially relative to the talus and tibia.

142
 References

10 15
Flexion(-)/Extension(+)
0

Cranial(+)/Caudal(-)
Displacement (mm)
10
-10
Angle (deg)

-20 5
-30
0
-40
-50 -5
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

4 5
Adduction(+)/Abduction(-)

Displacement (mm)
Medial(+)/Lateral(-)
0
0
Angle (deg)

-4
-5
-8

-12 -10
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)

2 5
0
Internal(+)/External(-)
Rotation Angle (deg)

Proximal(+)/Distal(-)
Displacement (mm)

0
-5
-2 -10
-15
-4
-20
-6 -25
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) Time (% of stride)
Fig 7.15  Three rotations and three translations measured at the tarsal joint during trotting overground at 2.8 m/s. Left panel shows rotations:
flexion–extension (above), abduction–adduction (middle) and external–internal rotation (below). Right panel shows translations: craniocaudal (above),
mediolateral (middle) and proximodistal (below). Vertical dashed lines show transition from stance to swing phase.
Reprinted from Lanovaz, J.L., Khumsap, S., Clayton, H.M., Stick, J.A., Brown, J., 2002. Three-dimensional kinematics of the tarsal joint at trot. Equine Vet. J. 34 (Suppl.), 308–313, with permission
from the Equine Veterinary Journal.

References
Abe, T., Kumagai, K., Brechue, W.F., 2000. leading and trailing limb at the canter. Biewener, A.A., 1998. Muscle-tendon stresses
Fascicle length of leg muscles is greater in Equine Vet. J. 23 (Suppl.), 80–83. and elastic energy storage during
sprinters than distance runners. Med. Sci. Back, W., Schamhardt, H.C., Hartman, W., locomotion in the horse. Comp.
Sports Exerc. 32, 1125–1129. Barneveld, A., 1995d. Repetitive loading Biochem. Physiol. B-Biochem. Mol. Biol.
Alexander, R.McN., Trestik, C.L., 1989. Bistable and oscillations in the distal fore and hind 120, 73–87.
properties of the hock joint of horses limb as predisposing factors for equine Brown, N.A.T., Kawcak, C.E., McIlwraith, C.W.,
(equus spp.). J. Zool. 218, 383–391. lameness. Am. J. Vet. Res. 56, 1522–1528. Pandy, M.G., 2003. Architectural properties
Back, W., Barneveld, A., Schamhardt, H.C., Back, W., Schamhardt, H.C., Savelberg, of distal forelimb muscles in horses, equus
Bruin, G., Hartman, W., 1994. Kinematic H.H.C.M., et al., 1995a. How the horse caballus. J. Morphol. 258, 106–114
detection of superior gait quality in moves: significance of graphical Buchner, H.H.F., Savelberg, H.H.C.M.,
young trotting warmbloods. Vet. Q. 16, representations of equine forelimb Schamhardt, H.C., Barneveld, A., 1997.
S91–96. kinematics. Equine Vet. J. 27, 31–38. Inertial properties of Dutch warmblood
Back, W., Hartman, W., Schamhardt, H.C., Back, W., Schamhardt, H.C., Savelberg, horses. J. Biomech. 30, 653–658.
Bruin, G., Barneveld, A., 1995c. Kinematic H.H.C.M., et al., 1995b. How the horse Cano, M.R., Miro, F., Vivo, J., Galisteo, A.M.,
response to a 70-day training period in moves: significance of graphical 1999. Comparative biokinematic study of
trotting Dutch warmbloods. Equine Vet. J. representations of equine hind limb young and adult Andalusian horses at the
18 (Suppl.), 127–131. kinematics. Equine Vet. J. 27, 39–45. trot. J. Vet. Med. Ser. A-Physiol. Pathol.
Back, W., Schamhardt, H.C., Barneveld, Badoux, D.M., 1987. Some biomechanical Clin. Med. 46, 91–101.
A.1996. Are kinematics of the walk related aspects of the structure of the equine Cano, M.R., Vivo, J., Miro, F., Morales, J.L.,
to the locomotion of a warmblood horse at tarsus. Anatomischer Anzeiger 164, 53–61. Galisteo, A.M., 2001. Kinematic
the trot? Vet. Quart. 18, S79–84. Biewener, A.A., 1989. Scaling body support in characteristics of Andalusian, Arabian and
Back, W., Schamhardt, H.C., Barneveld, mammals – limb posture and muscle Anglo-Arabian horses: a comparative study.
A.1997. Kinematic comparison of the mechanics. Science 245, 45–48. Res. Vet. Sci. 71, 147–153.

143
 7 Hind limb function

Chapman, A.E., Caldwell, G.E., 1983. Kinetic a range of speeds. J. Exp. Biol. 207, Johnston, C., Roepstorff, L., Drevemo, S.,
limitations of maximal sprinting speed. 3507–3514. 1995. Kinematics of the distal forelimb
J. Biomech. 16, 79–83. Dyce, K.M., Sack W.O., Wensing C.J.G.1996. during the stance phase in the fast trotting
Clayton, H.M., 1993. The extended canter: a Textbook of veterinary anatomy, second ed. Standardbred. Equine Vet. J. 18 (Suppl.),
comparison of the stride kinematics of Philedelphia: Saunders. 170–174.
horses trained for dressage and for racing. Fleiss, O., Edl, M., Schubert, P., 1984. Intra Johnston, C., Roepstorff, L., Drevemo, S.,
Acta Anat. 146, 183–187. und interlimbkoordination bei 1996. Kinematics of the distal hind limb
Clayton, H.M., 1994a. Comparison of the verschiedenen ganggeschwindigkeiten des during the stance phase in the fast trotting
stride kinematics of the collected, working, pferdes. Arch. Tierärztl. Fortbild. 8, Standardbred. Equine Vet. J. 28, 263–268.
medium, and extended trot in horses. 116–128. Kearns, C.F., McKeever, K.H., Kumagai, K., Abe,
Equine Vet. J. 26, 230–234. Fredricson, I., Drevemo, S., 1972. T., 2002. Fat-free mass is related to
Clayton, H.M., 1994b. Comparison of the Methodological aspects of kinematics of one-mile race performance in elite
stride kinematics of the collected, working, the joints in the forelimbs of fast-moving Standardbred horses. Vet. J. 163, 260–266.
medium and extended canter in horses. horses. In: Fredricson, I. (Eds.), A Ker, R.F., Alexander, R.M., Bennett, M.B., 1988.
Equine Vet. J. 17 (Suppl.), 16–19. photogrammetic study applying high speed Why are mammalian tendons so thick?
Clayton, H.M., Hodson, E., Lanovaz, J.L., cinematography. Acta Vet. Scand. 37 J. Zool. Lond. 216, 309–324.
Colborne, G.R., 2001. The hind limb in (Suppl.), 1–136. Khumsap, S., 2002. Biomechanics of the
walking horses: 2. Net joint moments and Galisteo, A.M., Cano, M.R., Miro, F., Vivo, J., equine tarsal joint. PhD Dissertation.
joint powers. Equine Vet. J. 33, 44–48. Morales, J.L., Aguera, E., 1996. Angular Michigan State University.
Clayton, H.M., Hoyt, D.F., Wickler, S.J., joint parameters in the Andalusian horse at Khumsap, S., Clayton, H.M., Lanovaz, J.L.,
Cogger, E.A., Lanovaz, J.L., 2002. Hind walk, obtained by normal videography. J. 2001a. Effect of walking velocity on ground
limb net joint energies during swing phase Equine. Vet. Sci. 16, 73–77. reaction force variables in the hind limb of
as a function of trotting velocity. Equine Galisteo, A.M., Cano, M.R., Morales, J.L., Vivo, clinically normal horses. Am. J. Vet. Res.
Vet. J. 34 (Suppl.), 363–367. J., Miro, F., 1998. The influence of speed 62, 901–906.
Clayton, H.M., Lanovaz, J.L., Schamhardt, and height on the kinematics of sound Khumsap, S., Clayton, H.M., Lanovaz, J.L.,
H.C., Willemen, M.A., Colborne, G.R., horses at the hand trot. J. Equine. Vet. Sci. 2001b. Effect of walking velocity on hind
1998. Net joint moments and powers in 16, 73–77. limb kinetics during stance in normal
the equine forelimb during the stance Galisteo, A.M., Vivo, J., Cano, M.R., Morales, horses. Equine Vet. J. Suppl. 33, 21–26.
phase at the trot. Equine Vet. J. 30, J.L., Miro, F., Aguera, E., 1997. Differences Khumsap, S., Lanovaz, J.L., Rosenstein, D.L.,
384–389. between breeds (Dutch warmblood vs Byron, C., Clayton, H.M., 2003. Effect of
Colborne, G.R., Lanovaz, J.L., Sprigings, E.J., Andalusian purebred) in forelimb unilateral synovitis of distal intertarsal and
Schamhardt, H.C., Clayton, H.M., 1997a. kinematics. J. Equine Sci. 8, 43–47. tarsometatarsal joints on sagittal plane
Joint moments and power flow in equine Gomez Alvarez, C.B., L’ami, J.J., Moffat, D., kinematics and kinetics of trotting horses.
gait: a preliminary study. Equine Vet. J. 23 Back, W., van Weeren, P. R., 2008. Effect of Am. J. Vet. Res. 64, 1491–1495.
(Suppl.), 33–36. chiropractic manipulations on the Klomkleaw, W., Kasashima, Y., Fuller, G.A.,
Colborne, G.R., Lanovaz, J.L., Sprigings, E.J., kinematics of back and limbs in horses Kobayashi, A., Yoshihara, T., Oikawa, M.A.
Schamhardt H.C., Clayton, H.M., 1997b. with clinically diagnosed back problems. et al., 2002. Horse lumbrical muscle:
Power flow in the equine forelimb. Equine Equine Vet. J. 40, 153–159. possible structural and functional
Vet. J. 23 (Suppl.), 37–40. Gunn, H.M., 1983. Morphological attributes reorganization in regressive muscle. J. Vet.
Corley, J.M., Goodship, A.E., 1994. Treadmill associated with speed of running in horses. Med. Ser. C- Anat. Histol. Embryol. 31,
training induced changes to some Equine exercise physiology. Proceedings of 85–98.
kinematic variables measured at the canter the first international conference 271–274. Kobluk, C.N., Schnurr, D., Horney, F.D.,
in Thoroughbred fillies. Equine Vet. J. 17 Gunn, H.M., 1987. Muscle, bone and fat Sumnersmith, G., Willoughby, R.A.,
(Suppl.), 20–24. proportions and muscle distribution of Dekleer, V. et al., 1989. Use of high speed
Crook, T.C., Cruickshank, S.E., McGowan, Thoroughbreds and other horses. Equine cinematography and computer generated
C.M., Stubbs, N.., Wakeling, J.M., Wilson, Exerc. Physiol. 2, 253–264. gait diagrams for the study of equine hind
A.M. et al., 2008. Comparative anatomy Hjertén, G., Drevemo, S., Eriksson, L.E., 1994. limb kinematics. Equine Vet. J. 21, 48–58.
and muscle architecture of selected hind Shortening of the hind limb in the horse Krüger, W., 1937. Ueber den bewegungsablauf
limb muscles in the quarter horse and arab. during the stance phase. Equine Vet. J. 17 an dem oberen teil der vordergliedmaβe
J. Anat. 212, 144–152. (Suppl.), 48–50. des pferdes im schritt, trab and galopp.
Deuel, N.R., Park, J.J., 1990. Canter leads Hodson, E., Clayton, H.M., Lanovaz, J.L., Tierärztl. Rundsh. 43, 1–24.
change kinematics of superior dressage 2001. The hind limb in walking horses: 1. Krüger, W., 1938. Ueber den bewegungsablauf
horses. J. Equine Vet. Sci. 10, 287–298. Kinematics and ground reaction forces. an dem oberen teil der hintergliedmaβe
Dimery, N.J., Alexander, R.M.c.N., Ker, R.F., Equine Vet. J 33, 38–43. des pferdes im schritt, trab und galopp.
1986. Elastic extension of leg tendons in Holmström, M., Fredricson, I., Drevemo, S., Tierärztl. Rundsh. 44, 549–557.
the locomotion of horses (equus caballus). 1994. Biokinematic analysis of the Swedish Lanovaz, J.L., Khumsap, S., Clayton, H.M.,
J. Zool. 210, 415–425. warmblood riding horse at trot. Equine Vet. Stick, J.A., Brown, J., 2002. Three-
Dutto, D., Hoyt, D., Clayton, H. M., Cogger, J. 26, 235–240. dimensional kinematics of the tarsal joint
E.A., 2006. Joint work and power for both Hoyt, D.F., Molinari, M., Wickler, S.J., Cogger, at the trot. Equine Vet. J. 34 (Suppl.),
the forelimb and hind limb during trotting E.A., 2002. Effect of trotting speed, load 308–313.
in the horse. J. Exp. Biol. 209, 3990–3999. and incline on hind limb stance-phase Lawrence, L., Jackson, S., Kline, K., Moser L.,
Dutto, D.J., Hoyt, D.F., Clayton, H.M., Cogger, kinematics. Equine Vet. J. 34 (Suppl.), Powell, D., Biel, M., 1992. Observations on
E.A., Wickler, S.J., 2004a. Moments and 330–336. body-weight and condition of horses in a
powers generated by the horse (equus Jansen, M.O., Raaij, J.A.G.M., van Bogert, A.J., 150-mile endurance ride. J. Equine Vet. Sci.
caballus) hind limb during jumping. J. Exp. van den Schamhardt, H.C., Hartman, W., 12, 320–324.
Biol 667–674. 1992. Quantitative analysis of computer- Lopez-Rivero, J.L., Serrano, A.L., Diz, A.M.,
Dutto, D.J., Hoyt, D.F., Cogger, E.A., Wickler, averaged electromyographic profiles of Galisteo, A.M.1992. Variability of
S.J., 2004. Ground reaction forces in horses intrinsic limb muscles in ponies at the muscle-fiber composition and fiber size
trotting up an incline and on the level over walk. Am. J. Vet. Res. 53, 2343–2349. in the horse gluteus medius – an

144
 References

enzyme-histochemical and morphometric tennessee walking horse yearling. Horse. Strubelt, H., 1928. Uber die bedeutung des
study. J. Anat. 181, 1–10. Equine Compar. Exerc. Physiol. 3, 101–108. lacertus fibrosus und des tendo
Martinez-del Campo, L.J., Kobluk, C.N., Greer, Niki, Y., Ueda, Y., Yoshida, K., Masumitsu, H., femorotarsicus fur das stehen und die
N., et al, 1991. The use of high-speed 1982. Force plate study of equine bewegung des pferdes. Arch. Tierheilk. 57,
videography to generate angle-time and biomechanics. 2. The vertical and fore-aft 575–585.
angle-angle diagrams for the study of components of floor reaction forces and Tokuriki, M., Aoki, O., 1995.
equine locomotion. Vet. Comp. Orthop. motion of equine limbs. Bull. Equine Res. Electromyographic activity of hind
Traum. 4, 120–131. Inst. 19, 1–17. limb joint muscles during walk, trot
Martuzzi, F., Vaccari, S.F., Gosi, S., Catalano, Payne, R.C., Hutchinson, J.R., Robilliard, J.J., and canter. Equine Vet. J. 18 (Suppl.),
A.L., 2007. Kinematic characteristics of Smith, N.C., Wilson, A.M., 2005. 152–155.
walk in the Bardigiano horse breed. Ann. Functional specialisation of pelvic limb Ueda, Y., Niki, Y., Yoshida, K., Masumitsu, H.,
Fac. Medic. Di Parma 27, 229–234. anatomy in horses (equus caballus). 1981. Force plate study of equine
McLaughlin, R.M., Gaughan, E.M., Roush, J.K., J. Anat. 206, 557–574. biomechanics: floor reaction forces of
Skaggs, C.L., 1996. Effects of subject Radin, E.L., Yang, K.H., Riegger, C., Kish, V.L., normal walking and trotting horses. Bull.
velocity on ground reaction force O’Connor, J.J., 1991. Relationship between Equine Res. Inst. 18, 28–41.
measurements and stance times in lower limb dynamics and knee joint pain. Updike, S.J., 1984. Functional-anatomy of the
clinically normal horses at the walk and J. Orthop. Res. 9, 398–405. equine tarsocrural collateral ligaments. Am.
trot. Am. J. Vet. Res. 57, 7–11. Ramakrishnan, H.K., Kadaba, M.P., 1991. On J. Vet. Res. 45, 867–874
Merkens, H.W., Schamhardt, H.C., 1994. the estimation of joint kinematics during van den Bogert, A.J., Schamhardt, H.C., 1993.
Relationship between ground reaction force gait. J. Biomech. 24, 969–977. Multi-body modelling and simulation of
patterns and kinematics in the walking and Robert, C., Valette, J.P., Denoix, J.M., 1998. animal locomotion. Acta Anat. 146,
trotting horses. Equine Vet. J. 17 (Suppl.), Surface electromyographic analysis of the 95–102.
67–70. normal horse locomotion: a preliminary van den Bogert, A.J., van Weeren, P.R.,
Merkens, H.W., Schamhardt, H.C., Hartman, report. In: Lindner, A. (Ed.), Proc. Conf. Schamhardt, H.C., 1990. Correction for
W., Kersjes, A.W., 1986. Ground reaction Equine Sports Med. Sci. Wageningen Press, skin displacement errors in movement
force patterns of Dutch warmblood Wageningen, The Netherlands, pp. 80–85. analysis of the horse. J. Biomech. 23,
horses at normal walk. Equine Vet. J. 18, Robert, C., Valette, J.P., Pourcelot, P., Audigie, 97–101.
207–214. F., Denoix, J.M., 2002. Effects of trotting Walter, K., 1925. Der Bewegungsablauf an den
Merkens, H.W., Schamhardt, H.C., van Osch, speed on muscle activity and kinematics in freien gliedmassen des pferdes im schritt,
G.J.V.M., van den Bogert, A.J., 1993. saddlehorses. Equine Vet. J. 34 (Suppl.), trab und galop. Arch. Wiss. Pract. Tierheilk.
Ground reaction force patterns of Dutch 295–301. 53, 316–352.
warmbloods at the canter. Am. J. Vet. Res. Sack, W.O., 1989. The stay-apparatus of the Van Weeren, P.R., van den Bogert, A.J.,
54, 670–674. horse’s hind limb – explained. Equine Prac. Barneveld, A., 1992. Correction models
Minetti, A.E., Ardigo, L.P., Reinach, E., Saibene, 11, 31–35. for skin displacement in equine kinematic
F., 1999. The relationship between Schamhardt, H.C., Hartman, W., Lammertink, gait analysis. J. Equine Vet. Sci. 12,
mechanical work and energy expenditure J.M.L.A., 1985. In vivo bone strain in the 178–192.
of locomotion in horses. J. Exp. Biol. 20, equine tibia before and after transection of Van Weeren, P.R., van den Bogert, A.J.,
2329–2338. the fibularis tertius muscle. Res. Vet. Sci. 39, Barneveld, A., Hartman, W., Kersjes, A.W.,
Molenaar, G.J., 1983. Kinematics of the 139–144. 1990. The role of the reciprocal apparatus
reciprocal apparatus in the horse. J. Vet. Schamhardt, H.C., Merkens, H.W., 1994. in the hind limb of the horse investigated
Med. Ser. C- Anat. Histol. Embryol. 12, Objective determination of ground contact by a modified CODA-3 opto-electronic
278–287. of the limbs of the horse at the walk and kinematic analysis system. Equine Vet. J. 9
Mullineaux, D.R., Clayton, H.M., Gnagey, trot: comparison between ground reaction (Suppl.), 95–100.
L.M., 2004. Effects of offset-normalizing forces, accelerometer data and kinematics. Wentink, G.H., 1978a. Biokinetical analysis of
techniques on variability in motion Equine Vet. J. 17 (Suppl.), 75–79. the movement of the pelvic limb of the
analysis data. J. Appl. Biomech. 20, Schuurman, S.O., Kersten, W., Weijs, W.A., horse and the role of the muscles in the
177–184. 2003. The equine hind limb is actively walk and trot. Anat. Embryol. 152,
Nauwelaerts, S., Allen, W., Lane, J., Clayton, stabilized during standing. J. Anat. 202, 261–272.
H.M., 2011. Inertial properties of equine 355–362. Wentink, G.H., 1978b. An experimental study
limb segments. J. Anat. 218, 500–509. Serrano, A.L., Petrie, J.L., Rivero, J.L. L., on the role of the reciprocal tendinous
Nicodemus, M.C., Booker, J.E., 2007. Hermanson, J.W., 1996. Myosin isoforms apparatus of the horse at walk. Anat.
Two-dimensional kinematics of the jog and and muscle fiber characteristics in equine Embryol. 154, 143–151.
lope of the stock breed western pleasure gluteus medius muscle. Anat. Rec. 244, Zajac, F.E., 1989. Muscle and tendon
horse equine compar. Exerc. Physiol. 4, 444–451. – properties, models, scaling, and
59–70. Stashak, T.S., 1987. Adams’ lameness in application to biomechanics and motor
Nicodemus, M.C., Holt, H.M., 2006. Two- horses. fourth ed. Lea & Febiger, control. Critical Reviews Biomed. Eng. 17,
dimensional kinematics of the flat-walking Philadelphia, pp. 76, 88, 102. 359–411.

145
This page intentionally left blank
 C H A PTER 8 

The role of the hoof and shoeing

Willem Back, Frederik Pille

Introduction connects the skin above from the hoof wall below. Deep to the
coronet lies the dermis that produces the horn tubules of the hoof
wall via the horn papillae, which project distally into the horn of
To deliver maximal performance it is essential that a horse has good the hoof wall. The horn tubules are responsible for the striated
balance. A part of the required balance comes from factors con- appearance of the hoof wall. The nonpigmented innermost part of
nected to the hoof (Balch et al., 1991a; Curtis, 1999, Aoki, 1999; the tubular hoof wall is less stiff than the outer part and serves as
Arabian et al., 2001; Eliashar, 2007). ‘No foot, no horse’ is a well- a mechanical buffer zone assisting the feet of horses in transmitting
known saying that emphasizes the important role of the hoof load through the tissues without inflicting damage (Wagner et al.,
(Davies, 2002). 2001, 2002).
After domestication of the horse 5000–6000 years ago, man Projecting from the inner surface of the hoof wall are the primary
assumed much of the responsibility for the balance between hoof epidermal lamellae (600/hoof) and secondary epidermal lamellae
growth and hoof wear. The Greek horse people (hippiaters) were in (100–200/primary lamella). The insensitive horn of these epider-
favor of breeding horses with good hoof quality that did not need mal lamellae interdigitates with the sensitive dermal lamellae to
shoes. The Romans invented the hipposandal that was attached to form the functional connection between the hoof wall and P3. The
the hoof with straps and was used to protect the feet of the horse horn that forms the epidermal lamellae is unstructured (non-
en route to the battlefield, where they were removed. The first iron tubular). As the hoof wall grows, it moves distally by a mechanism
horseshoes with nails were made by the Celts 2000 years ago and that allows the primary epidermal lamellae to slide past the station-
were similar to the ones we use today (Fig. 8.1). ary secondary epidermal lamellae. Primary epidermal laminae
During locomotion every stride involves forces at the hoof– morphology (spacing, orientation and curvature) is responsive to
ground interface that load the locomotor apparatus. Repeated appli- mechanical stress at the laminar junction (Thomason et al., 2005,
cation of forces that have a high magnitude and/or an abnormal 2008). On the solar surface, the white line demarcates the junction
direction of action, overload the limb and may lead to the develop- between sensitive and insensitive tissues and is in fact a distal pro-
ment of pathological processes (Johnston & Back, 2006). When we jection of dermal and epidermal laminae (cf. terminal papillae).
have an insight how these forces work, we can modulate them by When shoes are nailed in place, the nails should enter peripheral
corrective hoof trimming and optimal shoeing as rational measures to the white line.
to treat and prevent lameness, e.g. at purchase (Anderson, 1992). The bony skeleton of the hoof consists of the distal phalanx (P3)
That is the basic theme of this chapter. or coffin bone, the navicular bone and part of the middle phalanx
(P2) or short pastern bone. All three of these bones take part in the
formation of the distal interphalangeal or coffin joint (Fig. 8.3).
Functional anatomy of the foot During flexion of the coffin joint, the navicular bone generally
follows the coffin bone although small but consistent motions exist
General anatomy between the navicular bone and the coffin bone (Bowker et al.,
2001; Van Dixhoorn et al., 2002).
The equine foot has evolved from the third digit, which has been The medial and lateral hoof cartilages (ungual cartilages) are
greatly elongated and strengthened. The hoof wall has developed square-shaped structures located on either side of P3 and are con-
from the nail of the third digit. The hoof of the first digit is still nected to P1 and P2 by connective tissue. They are large and flexible.
present, though rudimentary, in the form of the chestnut on the Between them lie the digital cushion and a venous plexus. Compres-
medial side of the radius (forelimb) or metatarsus (hind limb), sion of these structures during locomotion assists the return of
while that of the fifth digit, especially in Coldbloods, persists as the venous blood to the heart (Hoffman et al., 2001; Pietra et al.,
ergot on the palmar/plantar side of the metacarpo-/metatarsopha- 2004). The navicular bone is connected to P3 by the distal navicular
langeal joint (Dyce & Wensing 1980; Pollitt 1995). or ‘impar’ ligament, to P2 by connective tissue and the synovial
The hoof capsule consists of several parts: the coronet, the wall, membranes of the coffin joint, navicular bursa and tendon sheath,
the sole, the frog and the heels (Fig. 8.2). The horn of each part and to P1 by the proximal navicular ligaments. The navicular bone
of the hoof is produced by a corresponding area of dermis (corium). is covered on one side by hyaline cartilage and on the opposite side
The estimated horn regeneration time based on a growth rate of by fibrocartilage (Viitanen et al., 2003b). Its hyaline cartilage side
8–10 mm/month is 12 months for the toe, 6–8 months for the is in contact with the palmar/plantar aspect of P2 and P3. The deep
quarters, and 4–5 months for the heels. The coronet separates and digital flexor tendon (DDFT) curves around the palmar/plantar

147
 8 The role of the hoof and shoeing

contents; low palmar nerve blocks just distal to the fetlock joint
anesthetize the entire hoof.

Proprioception
Bowker et al. (1993, 1995) described the concentration of nocicep-
tors in the palmar/plantar part of the frog and in the proximal
navicular area. It is hypothesized that these so-called lamellar
A B bodies supply the central nervous system and the brain with pro-
prioceptive information describing the location of the body in space
(Van Wulfen & Bowker, 2002). This information is needed to
control the central pattern generator (CPG) and thus intra- and
inter-limb coordination (Fig. 8.6).
Studies of the effects of local anesthesia in sound horses have
yielded conflicting results. Keg et al. (1996) demonstrated no effect
of a low palmar nerve block on gait symmetry as evaluated kineti-
cally, whereas Kübber et al. (1994) did detect some kinematic
effects. It is not known whether the differences between these
studies were due to the sensitivity of the analytic equipment or to
D
the (un-) soundness of the horses.
C

Hoof mechanics in the standing horse

Hoof–pastern axis
Ideally, hoof trimming optimizes the interaction between the hoof
and the ground during locomotion. Since the hoof is a three-
E dimensional structure, it should be balanced in both the craniocau-
dal and mediolateral planes. Forces at the toe, medial and lateral
Fig 8.1  The development of the horseshoe. (A) Ancient grass sandal. (B) heels collectively are optimally distributed when the hoof and
Roman iron hipposandal fastened without nails. (C) Celtic horseshoe with pastern angles are aligned (Balch et al., 1997; Davies, 2002).
oval nailholes. (D) Medieval shoe with square nail holes. (E) Renaissance
shoe with fullering.
Reprinted with kind permission of Butler, Doug, The Principles of Horseshoeing, 1995, Doug Hoof balance
Butler Enterprises, Inc. www.dougbutler.com.
Craniocaudal balance
Craniocaudal balance evaluates the hoof in a lateral view (Fig.
8.7). It is assessed with the horse standing square on a level
aspect of the navicular bone before attaching to the flexor surface surface. Alignment of the dorsal hoof wall with the pastern axis is
of P3. The navicular bursa is interposed between the DDFT and the achieved by adjusting the absolute and relative lengths of the
fibrocartilage of the navicular bone. heels, the quarters and the toe. A hoof that is balanced in this
manner usually contacts the ground flat-footed or slightly heel
first. When the hoof has a more acute angle than the pastern, the
Vascular supply hoof–pastern axis is said to be broken backwards. Conversely,
when the angle of the hoof is more upright than that of the
The blood supply enters the hoof via the palmar/plantar digital pastern, the hoof–pastern axis is said to be broken forward. Radio-
arteries, which run through the terminal arch within P3 (Fig. 8.4). graphic studies have shown that P1 is always a little more upright
A network of arteries perforates the dorsal surface of P3 and ramifies (vertical) than P2 and P3 with the three phalanges being most
in the lamellar dermis. Some branches from this plexus are directed closely aligned when the hoof is trimmed with the dorsal hoof
proximally to supply the coronet, where they anastomose with wall parallel to the pastern axis (Bushe et al., 1987). Under these
branches of the circumflex artery of the coronet. The dorsal branches circumstances the angle between the solar surface of P3 and the
leave the palmar/plantar digital arteries just before they enter the ground is approximately 5°, whereas the dorsal face of P3 is paral-
terminal arch of P3. The palmar/plantar branches run to the heel lel to the dorsal hoof wall at a distance of about 19 ± 0.5 mm in
region where they form the venous plexus and anastomose with the Warmbloods (Back, 2001).
circumflex artery of the sole. There is a difference between the shape of the fore and hind
hooves: the fore hoof is wider and usually has a smaller (more
acute) hoof angle, while the hind hoof is narrower and has a larger
hoof angle. In farriery books (Hermans, 1984; Hertsch et al., 1996;
Nerve supply Ruthe et al., 1997), angles ranging from 45–50° for the forelimbs
The hoof is innervated by the palmar/plantar digital nerves, which and 50–55°for the hind limbs have been reported, but more
are located caudal to the vein and artery in the pastern region (Fig. recently the average angles have been reported to be over 50° in the
8.5). The nerves can be blocked by injection of a local anesthetic at forelimb and over 55° in the hind limbs (Clayton, 1988; Kobluk
the proximal pastern level where the nerves are crossed by the liga- et al., 1990; Balch et al., 1991a; Butler, 1995; Hickman & Hum-
ment of the ergot (low palmar block) or more distally as the nerve phrey, 1997; Oosterlinck et al., 2010c). In racehorses, a significant
passes deep to the hoof cartilage (palmar digital block). Palmar decrease in hoof angle was observed in association with starting
digital nerve blocks anesthetize the caudal third of the hoof and its fast exercise work (Peel et al., 2006). According to the same study,

148
 Hoof mechanics in the standing horse

Primary lamellae

1
2

5
Secondary
lamellae

5
4
5
3

3
4 4

5
6 6

7 7 Fig 8.2  The hoof capsule and its parts. (A) Lateral view. 1, Coronary band
epidermal layer; 2, coronary band germinative layer; 3, horn papillae and
tubulae; 4, unstructured horn; 5, primary and secondary epidermal lamellae. (B)
B Ventral view. 1, Pigmented outer hoofwall; 2, unpigmented inner hoofwall; 3,
white line; 4, sole; 5, frog, 6, sulcus; 7, heels. (The interrupted lines indicate the
projected location of the navicular bone and the third phalanx.) (C) Schematic
drawing of the correct location of a hoof nail.
Reprinted from Dyce, K.M. and Wensing, C.J.G. (1980) De Anatomie Van Het Paard. Utrecht:
C Uitgeverij Bohn, Scheltema, Holkema, with permission of K.M. Dyce.

wet pasture conditions may also be associated with a reduced back-foot axis that becomes more upright with increased bone
hoof angle. growth and tendon strength.
There exist also differences in hoof shape between contralateral The second component of craniocaudal balance is the location
limbs. A study on radiographic measurements of the front feet of of the bearing surface of the hoof relative to the weight-bearing axis
normal Warmblood horses found that in 70% of the horses the left through the cannon bone. The bulbs of the heel should lie vertically
hoof capsule and coffin bone were significantly larger than the right below the central axis of the cannon bone in the sagittal plane. In
(Kummer et al., 2006; White et al., 2008). According to Wilson some horses, although the hoof–pastern axis is aligned, the whole
et al. (2009), contralateral differences in hoof spread are related to hoof capsule is located too far forward so that the bulbs of the heels
generalized asymmetry in forelimb left–right morphometry. Con- are ahead of the central axis of the cannon bone. The resulting
tralateral differences in hoof shape may suggest unequal loading of caudal concentration of the weight bends the hoof tubules at the
limbs, which in turn may contribute to injuries and reduced perfor- heels, which reduces their ability to withstand compression, and
mance. Kroekenstoel et al. (2006) determined that foals with leads to underrun heels (Stashak, 1987; Balch et al., 1997). Nor-
uneven feet show asymmetrical loading of the interphalangeal mally, the dorsal toe wall and caudal heel wall should run parallel
joints (Lejeune et al., 2006), whereas foals are born with a broken to each other and the toe wall length relative to heel wall length

149
 8 The role of the hoof and shoeing

1i

11
17

13
1

2i

7
2
17

3
6
5
8 14
17
2
12
9
17 3 3ii

15
10
4 3i

2
B

A 4 16 6 5
Fig 8.3  (A) Sagittal section of the equine distal fore limb. 1, Metacarpal bone (McIII); 2, first phalanx (P1); 3, second phalanx (P2); 4, third phalanx (P3);
5, proximal sesamoid bone; 6, navicular bone; 7, fetlock joint; 8, pastern joint; 9, coffin joint; 10, navicular bursa; 11, suspensory ligament; 12, straight
sesamoidean ligament; 13, deep digital flexor tendon; 14, superficial digital flexor tendon; 15, synovial membrane of coffin joint, navicular bursa and tendon
sheath with connective tissue; 16, distal navicular ‘impar’ ligament; 17, attachments of common digital extensor tendon to first, second and third phalanges.
(B) Schematic drawing of tendons and ligaments of the equine lower limb. 1, Superficial digital flexor tendon; 1I, proximal accessory ligament; 2, deep digital
flexor tendon; 2I, distal accessory ligament; 3, suspensory ligament; 3I, straight sesamoideum ligament; 3II, rami extensori; 4, proximal navicular ligament; 5,
distal navicular ‘impar’ ligament; 6, common digital extensor tendon; 7, extensor carpi radialis tendon.
Reprinted from Dyce, K.M. and Wensing, C.J.G. (1980) De Anatomie Van Het Paard. Utrecht: Uitgeverij Bohn, Scheltema, Holkema, with permission of K.M. Dyce.

should be 2 : 1.0 in the fore hooves and 2 : 1.5 in the hind hooves Caudron et al. (1997a,b, 1998) used radiographs to evaluate and
(Hermans, 1984). Underrun heels have been defined as having a correct this balance. In the dynamic method, the hoof is trimmed
difference between the angles of the toe and heels that is more than so that the medial and lateral sides contact the ground simultane-
5° (Balch et al., 1997). ously, which adds a new dimension to the equation and makes the
solution even more elusive. In a horse with ideal conformation,
static and dynamic balancing will show a rather similar result, but
when conformational defects are present the two methods produce
Mediolateral balance different results.
Mediolateral balance evaluates the hoof in a frontal plane and When a horse is standing quietly, the force due to gravity is com-
attempts to optimize hoof balance by either a static or a dynamic pensated by the ground reaction force (GRF) acting near the geo-
evaluation (Seeherman, 1991; Balch et al., 1997). Static balance metric center of pressure (CP) of the foot, which lies around the
seeks to achieve symmetry in the square standing horse so that a apex of the frog (Barrey, 1990; Balch et al., 1997; Ovnicek, 1997).
line that bisects the limb longitudinally is intersected at 90° by a The hoof rotates around the instantaneous center of rotation of the
transverse line drawn across the heels (geometric limb axis). coffin joint, which is located in the distal part of P2. Equilibrium

150
 Hoof mechanics in the standing horse

Proximal phalanx

Arterial circle of proximal phalanx

Proximal interphalangeal joint

Middle phalanx
Arterial circle of middle phalanx
Palmar digital arteries
Coronary circumflex artery

Coronary band Branches to, dorsal middle phalanx, coronary band

Branches to digital cushion, frog, lameallae
Distal interphalangeal joint
of heels and bars and palmar coronet
Hoof wall

Lamellar arteries

Circumflex artery of sole

Terminal arch Distal phalanx Notch in palmar process
A Dorsal branch

Proximal navicular artery

Dorsal branch of
middle phalanx

Proximal navicular
plexus
Distal navicular
plexus
Dorsal branch of
distal phalanx

Distal navicular
artery

Palmar branch of Circumflex artery of sole Fig 8.4  The main arteries of the foot: (A) lateral view, (B) caudal view.
distal phalanx Distal interosseus ligament Reprinted from Pollit, C. (1995) Color Atlas of the Horse’s Foot. Mosby, with permission from
B Elsevier and C. Pollit.

151
 8 The role of the hoof and shoeing

on a firm surface. This leaves a raised area projecting downwards at

the heel buttress. There is a gradual arc between the impression
mark at the lateral side of the toe and the heel impression mark on
each side of the foot. The hoof wall has four loading points: one
at each side of the toe and one at each heel. Little to no sole, frog
1 and bars are ever removed. Within a few months this method of
trimming should result in a stronger hoof structure with a cupped
sole, spreading heels, and a well-developed frog. Dejardin et al.
2 (2001) demonstrated that four point trimming results in strain
concentration above the hoof contact points with the strain magni-
4 tude being dependent on contact area.
Hood et al. (2001a) found that on a deformable surface, load
distribution is principally solar and located transversely across the
central region of the foot. Therefore, it is hypothesized that the
typical hoof conformation observed in feral horses is the result of
increased erosion of the central solar and quarter regions of the
bearing wall, allowing the heels and the lateral and medial toe to
remain relatively long. The rapid disappearance of this initially
noneroded areas when horses are maintained on a flat nondeform-
able surface further supports that they are caused by lack of contact
when the horse is on a more natural deformable terrain. Hood et al.
3
(2001a) concluded that so called four-point loading is the product
of ground surface abrasion and, as such, is the mirror image of the
loading pattern seen in horses kept on a natural terrain. Hence, the
B hoof ground interaction after an artificial four-point trim may only
partially mimic the natural hoof ground interaction observed in
feral horses.

Hoof mechanics during locomotion

3’ 5 Several techniques have been used to document the kinematics and

kinetics of the hoof. One of the requirements of the recording
equipment is that it should have a high enough sampling frequency
to distinguish events such as initial ground contact, heel-off and
A toe-off with sufficient accuracy (Linford, 1994). This can be espe-
cially difficult on outdoor tracks where reflective markers disappear
in the sand (Merkens & Schamhardt, 1994; Scheffer & Back, 2001).

Initial ground contact

When the horse is observed in the lateral view, initial ground
contact is heel first, flat-footed or toe first. The hind limbs show a
greater tendency to heel first contacts than the forelimbs, and heel
Fig 8.5  Sensory innervation of the foot, medial view. 1, Medial palmar
nerve; 2, ramus communicans; 3, digital medial palmar nerve; 3′, ramus
first contacts occur more frequently during high-speed locomotion
dorsalis; 4, medial palmar artery and vein; 5, medial digital palmar artery and (Back et al., 1995). However, in some movements, such as piaffe,
vein; (A) site of digital nerve block; (B) site of low palmar (abaxial) nerve toe first contacts are normal. The frequency of toe first contacts
block. increases when the hoof is trimmed with an acute angle (long toes
Reprinted from Dyce, K.M. and Wensing, C.J.G. (1980) De Anatomie Van Het Paard. Utrecht: and/or low heels). Conversely, when horses are trimmed with a
Uitgeverij Bohn, Scheltema, Holkema, with permission of K.M. Dyce. steep hoof angle (short toe and/or long heels), heel first contacts
are more numerous (Clayton, 1990a). Therefore, the manner of
initial ground contact depends on speed, gait and farriery. Neverthe-
exists when the moment (torque) of the GRF about the coffin joint less, Van Heel et al. (2004) found that most horses tend to land on
(FGRF* dGRF) is equal to the moment of the DDFT (FDDFT* dDDFT). Ideally, the lateral side of the foot, especially in the hind limbs, which is
the central reference point (CRP) should be located midway only minimally influenced by trimming.
between the toe and the heels (Wright & Douglas, 1993), although
in wild horses the distance from the toe to the CRP is only about Impact
one-third of the toe-to-heel distance (Ovnicek, 1997; Fig. 8.8).
Immediately after initial ground contact, the hoof is rapidly deceler-
ated by the vertical landing forces and horizontal braking forces that
Four point trimming reduce its speed to zero during the impact phase that follows initial
Ovnicek (1997) studied the hooves of 65 wild horses. From his ground contact (Back et al. 1995; Burn 2006; Parsons & Wilson,
observations he developed the natural trimming technique for 2006). The forces and decelerations associated with impact have
unshod horses. The principles of this technique are that the heels been measured using a force plate (Merkens & Schamhardt, 1994;
are trimmed back to the widest point of the frog, along the sole Merkens et al., 1994; Wilson & Pardoe, 2001; Weishaupt et al.,
plane. The toe is beveled to a 15–20° angle in a manner akin to 2002; Oosterlinck et al., 2010b), force shoe (Barrey, 1990; Frederick
what would be done in preparation for a rocker-toe shoe. The quar- & Henderson, 1970; Ratzlaff et al., 1985, 1993; Kai et al., 2000;
ters are hollowed slightly by rasping so they are not weight-bearing Rollot et al., 2004; Roland et al., 2005; Chateau et al., 2009a; Robin

152
 Hoof mechanics during locomotion

Fig 8.6  (A) Sensory receptors in the foot. In the sagittal view the four zones
of the sole (S1–S4) represent locations of the lamellated corpuscles with the
relative density being indicated by the intensity of the green shading in zones
S1–S4. S4 shows the highest density of lamellated corpuscles. In the ventral
view the lamellated corpuscles were obtained from the areas shaded in green.
(B) Sensory receptors in the navicular region. Inset shows the three levels from
which drawings a–c were obtained. Note that the lamellated corpuscles
(dots) are located primarily abaxially and proximal to the collateral
sesamoidean ligament (CSL) often in association with the palmar digital nerve
(n), as well as with the artery (a) and vein (v). d is a parasagittal drawing from
the abaxial region of CSL. Sensory receptors are shown as red dots. PII: pastern
bone; PIII: coffin bone; DDF: deep digital flexor tendon; Nav: navicular bone.
(A) Reprinted from Bowker, 1993, Sensory receptors in the equine foot, American Journal of
Digital Veterinary Research, with permission from American Journal of Veterinary Research http://
cushion avmajournals.avma.org/loi/ajvr. (B) Reprinted from Bowker, R.M., Linder, K.E., Sonea, I.M. and
Guida, L.A. (1995) Sensory nerve fibers and receptors in the equine distal forelimbs and their
potential role in locomotion. Equine Vet. J. 18, 141–146, with permission of the Equine
Hoof dermis Sole S1 S2 S3 S4 Veterinary Journal.

Central

Medial

Lateral

Frog

Lamellated
corpuscles

v,a,n

A PII
DDF

B PII
C
A

PIII

PII
DDF

PII
B
Nav
PIII

PII
DDF

B C

153
 8 The role of the hoof and shoeing

Toe in Straight Toe out

A Broken out Straight Broken in

Sloping Ideal Stumpy

Fig 8.7  Diagrams of straight, ideal and correct hoof–pastern axes and common
Broken back Correct Broken forward deviations viewed (A) from the front and (B) from the side.
Reprinted with kind permission of Butler, Doug, The Principles of Horseshoeing, 1995, Doug Butler
B Enterprises, Inc. www.dougbutler.com.

et al., 2009) and accelerometers (Benoit et al., 1993; Burn et al., The impulsive loading that occurs during impact has been impli-
1997; Back et al., 2006). The hoof and interphalangeal joints atten- cated as a causative factor in arthritis in animals (Radin et al.,
uate the shock wave associated with impact (Dyhre-Poulsen et al., 1981; Pratt, 1997; Radin 1999) and in humans (Folman et al.,
1994; Lanovaz et al., 1998; Willemen et al., 1998). Approximately 1986; Ker et al., 1989). The time during which these forces have to
67% of the damping of impact vibrations takes place at the interface be absorbed is reduced as speed increases. For example, the time
between the hoof wall and the distal phalanx. The attenuation of taken to absorb impact shock at the canter (7 m/s) is 50% of that
impact vibrations at the interphalangeal joints and the metacarpo- at the walk (1.6 m/s), which is a consequence of the shorter time
phalangeal joint seems considerably less (12% and 9% respectively) the foot is on the ground. Back et al. (1995) found that during
leaving approximately 12% of the impact vibrations detectable trotting on a treadmill the hoof was flat on the ground, and the
at the level of the metacarpus (Willemen et al., 1999a). Neverthe- vertical speed and acceleration were zero within 3% (20 ms) of the
less, the forces acting in different directions have the potential to total stride duration. Surprisingly, it took 6% (40 ms) of total
damage the body (Fig. 8.9). Friction between the hoof and the stride duration for the horizontal speed to reach zero, but as dem-
ground and hardness of the ground affect the forces applied to the onstrated by Gustås et al. (2001) the time lapse of this horizontal
limb during the impact phase (Hjertén & Drevemo, 1993, 1994). retardation of the hoof is an important factor in the attenuation of
The amplitude of impact vibrations at the level of the hoof wall is the impact.
15% higher in shod versus unshod hooves independent from the The forelimbs appeared to land with a higher vertical velocity and
type of shoe that is used (shoes with or without pad). At the level the hind limbs with a higher horizontal velocity (Back et al., 1995).
of the first phalanx and the metacarpus the difference between shod Thus, the forelimbs ‘bounce’, whereas the hind limbs ‘slide’. On the
and unshod vanishes (Willemen et al., 1999a). other hand, Gustås et al. (2004) found no significant difference in

154
 Hoof mechanics during locomotion

CA

Acceleration (m/s2)
Time (s)
A

8
VGRF
Front contact CP Rear contact
points CRP LGRF
points 6

e
4 d
A Force (kN) f
b c g
2
HGRF

0
TGRF
-2
0 100 200 300
Time (ms)
B
Fig 8.9  (A) The impact phase is the short period immediately following
initial ground contact, in which the decelerating hoof is oscillating relative
to the ground until hoof velocity has been reduced to zero. (B) During
impact the vertical landing (VGRF) and horizontal braking (HGRF) forces
acting on the hoof are transferred into longitudinal (LGRF) and transverse
(TGRF) forces acting on the equine lower limb.
Reprinted from Hjertén, G. and Drevemo, S. (1994) Semi-quantitative analysis of hoof strike in
the horse. J. Biomech. 17, 997–1004, with permission from Elsevier.

hoof deceleration between fore and hind limb using accelerometers

C mounted on the hooves. The sliding of the hoof in the sagittal plane
P1 A has been measured for horses trotting on a treadmill; steel shoes
slide for longer than rubber shoes and the hoof stops less abruptly
P2 on a coir mat than on a rubber belt (Roepstorff et al., 1994). It has
been reported that the sliding of the hoof took 16 ms on a dirt track,
whereas on a hard surface (asphalt or concrete) it took 21 ms for
the unshod hoof and 32 ms for steel shoes (Back, 2001). Pardoe
et al. (2001) could not demonstrate a significant difference in either
slip time or distance between steel, plastic and rubber shoes,
whereas Back et al. (2006) noted less gliding in freshly steel shod
Abaxial contact Abaxial contact horses with protruding nails than in the same horses shod with
point point plastic shoes.
R S
The active neurophysiological response time is the time required
B
for the muscles to respond to a stimulus. It has been estimated to
Fig 8.8  Schematic representation of foot balance showing the ideal be around 30 ms in humans (Nigg et al., 1981) and horses (Bowker
equilibrium of the foot relative to the center of articulation in the distal part et al., 1993; Hjertén & Drevemo, 1993). This suggests that the forces
of the second phalanx. (A) Craniocaudal balance: Lateral and solar view and acting on the equine lower limb during the impact phase are deter-
(B) mediolateral balance: heel view. CA, center of articulation; CP, center of mined mainly by passive factors like coordination and ground
pressure; CRP, central reference point. properties (Hjertén & Drevemo, 1993).

155
 8 The role of the hoof and shoeing

expansion in a consistent manner (Colles, 1989a). The latter is in

A B contrast to more recent work of Roepstorff et al. (2001) demonstrat-
ing that increased pressure on the frog and sole does increase heel
expansion, although some heel expansion still occurs when the frog
Dorsum and sole are unsupported. Surprisingly, the digital cushion has no
Quarters active role in this heel expansion (Taylor et al., 2005). Shoes with
a toe clip and two side clips placed behind the 3rd nail (type of
Motion of dorsal wall shoe used in the treatment of a fractured P3) minimize the hoof
and third phalanx mechanism (Kersjes et al., 1985; Hinterhofer et al., 2001).
Tension in laminar junction Strains on the hoof wall are considerably higher at the trot than
Motion of third phalanx at the walk (Thomason, 1998) and are higher in the trailing limb
relative to sides of wall than the leading limb at a gallop (Summerley et al., 1998). Surpris-
ingly, riding decreases the strain in the quarters by 30%, while hoof
C Dorsal hoof wall D wall strains are 20% higher on the medial side when the rider sits
Hoof in the saddle and 20% higher on the lateral side for the forward
wall seat (Summerley et al., 1998). Turning increases the hoof strain in
Third Third the quarter that is on the inside of the turn by 40% (Summerley
phalanx phalanx et al., 1998). A larger hoof angle and a longer toe length increased
the strain at the toe, even when the length of the toe was propor-
tional to the body size of the animals (Thomason, 1998). In the
medial and lateral wall this relation was reversed: more upright feet
Force due to were stiffer (Thomason, 1998). Elasticity of the hoof structures,
mass of horse however, is also affected by the moisture content: the wall consists
GRF GRF of 82% keratin and 16% water, while the frog is 56% keratin and
42% water (Hertsch et al., 1996). Tubules are unlikely to be involved
Fig 8.10  Schematic drawing of the hoof mechanism phenomenon. The in the hydration status of the foot, but appear to have a more
solid line represents the unloaded hoof wall, the dashed line shows the mechanical function in the hoof wall to redirect and resist cracks
change in shape that occurs during weight-bearing. Under load the dorsal (Kasapi & Gosline, 1997, 1998).
wall flattens and moves palmarly, while the heels move laterally and
caudally. GRF, ground reaction force.
Reprinted from Douglas, J.E., Biddick, T.L., Thomasson, J.J. and Jofriet, J.C. (1998) Stress/strain
behaviour of the equine laminar junction. J. Exp. Biol. 201, 2287–2297, with permission from Center of pressure
the Journal of Experimental Biology, http://jeb.biologists.org/content/201/15/2287.full. By tracking the path of the center of pressure in a craniocaudal plane
pdf+html it can be seen that immediately after initial ground contact it moves
from the point of hoof contact to a position close to the apex of
the frog (Barrey, 1990; Balch et al., 1997; Ovnicek, 1997; Van Heel
Hoof mechanism et al., 2004). It remains there for most of the stance phase. At the
Since the hoof shows more elasticity in the heel area than at the end of the stance phase the center of pressure starts moving toward
toes or quarters, its geometry changes when it is loaded: the proxi- the toe. At toe-off it lies beneath the dorsal hoof wall at its breakover
mal dorsal wall moves back, the quarters flare to the side and sole point (Barrey, 1990). Van Heel et al. (2005) studied the changes in
and frog perform a downward movement (Hinterhofer et al., 2000, location of center of pressure and hoof-unrollment pattern in rela-
2001; Burn & Brockington, 2001; Hobbs et al., 2004, 2009); this is tion to an 8-week shoeing interval and determined that horses can
called the hoof mechanism (Fig. 8.10). compensate to a certain extent for changes in hoof conformation
At impact the distance between the heels is wider than at rest, that develop during 8 weeks on shoes. However, the capacity to
while at breakover, because of increased pressure in the toe area, compensate is less in the forelimbs and thus, the relative increase
the distance between the heels is narrower than at rest. Thus, the in loading of these limbs during a shoeing interval is larger com-
foot has three functional areas: the heels for damping; the wall, frog pared to the hind limbs.
and sole for support; and the toe for propulsion (Barrey, 1990). The location of the center of pressure in a mediolateral plane
Deformation of the foot during the stance phase optimizes these shows that when the horse is standing squarely the direction is
functions, and also enables the hoof to act as a pump for the blood concentrated on the medial quarters (Colahan et al., 1991). During
circulation (Ratzlaff et al., 1985). locomotion more force is recorded on the medial hoof with force
The hoof mechanism has been studied using strain gauges (Kne- shoes (Ratzlaff et al., 1993; Balch et al., 1997; Kai et al., 2000;
zevic, 1966; Bayer, 1973; Dyhre-Poulsen et al., 1994; Summerley Rollot et al., 2004; Roland et al., 2003; 2005, Chateau et al., 2009a),
et al., 1998; Thomason, 1998; Thomason et al., 2001, 2002; Roep- and with force and pressure plates in horses and ponies (Van Heel
storff et al., 2001), photoelastic material (Davies, 1997; Dejardin et al., 2004; Oosterlinck et al., 2010a,b,c). Transverse forces as
et al., 1997, 1999), optical systems (Roepstorff et al., 2001; Burn & recorded with force plates are not consistent throughout the gaits
Brockington, 2006), doppler (Hoffman et al., 2001; Pietra et al., in the fore and hind limbs (Merkens & Schamhardt, 1994; Merkens
2004), and special horse boots (Barrey, 1990; Preuschoft, 1989). As et al., 1994; Weishaupt et al., 2002). Furthermore, extrasagittal joint
hoof strain and limb loading are related, the GRF can be estimated motions in the forelimb will show bilateral variation since horizon-
from hoof strain data using artificial neural networks in experiments tal moments around the hoof center of pressure are not symmetric
where force plates cannot be used (Savelberg et al., 1997). Photo- (Colborne et al., 2009).
elastic materials did not show any concentrations of colored fringes Changes in fetlock and coffin joint angulation affect strain in the
in the foot of the standing horse following trimming (Davies, palmar soft tissues that support the limb during the stance phase
1996). Shoeing stabilized (Thomason, 1998) or even decreased (Leach 1983; Bushe et al., 1987; Thompson et al., 1993; Riemersma
hoof movements (Colles, 1989b; Dyhre-Poulsen et al., 1994). et al., 1996a). The suspensory ligament (SL) and superficial digital
Shoes with 5° raised heels reduce hoof deformation compared to flexor tendon (SDFT) are affected by the fetlock joint angle; strain
flat horseshoes (Hinterhofer et al., 2000) whereas the extra frog increases as the fetlock extends. The distal accessory ligament (DAL)
pressure provided by heart bar shoes does not influence hoof and deep digital flexer tendor (DDFT) are affected by the coffin

156
 Hoof mechanics during locomotion

Fig 8.11  Shoes used to compare breakover. Left, a standard flat shoe (Mustad 22/10 LB); right and a shoe with a rolled toe (Mustad 22/10 equilibrium).
Reprinted from Van Heel, M.C., van Weeren, P.R., Back, W. (2006b) Shoeing sound warmblood horses with a rolled toe optimises hoof-unrollment and lowers peak loading. Equine Vet. Jour.
38 (3), 258–62, with permission of the Equine Veterinary Journal.

joint angle: strain increases when this joint extends. The load dis- 12
tribution between the tendinous structures changes with alterations
in hoof balance: strain reduction in one tendon may result in an 10
increased strain in another structure. Also the horse may compen-
sate, to a certain extent, for hoof imbalances by adjusting the length
8
of the muscle bellies of the SDFT and DDFT or by changing the

Shift CoP (mm)

configuration of the joint angles in the proximal limb, e.g. when it
experiences navicular bone pain in the distal limb (Wilson et al., 6
2001).
4

Breakover 2
Breakover is the terminal part of the stance phase from heel-off to
toe-off. Rotation of the hoof is brought about as a result of tension 0
in the DDFT and DAL, and in the navicular ligaments (Schamhardt 0 50 100 150 200
et al., 1991; Riemersma et al., 1996b). Farriery modifications that Time (msecs)
facilitate breakover may reduce tension in the DAL and in the
navicular ligaments and also reduce pressure of the DDFT against Fig 8.12  Craniocaudal displacement of the center of pressure (COP) from
the navicular bone. The onset and duration of breakover are sensitive midstance to toe off in the same horse with a flat shoe (red line and
to changes in hoof balance, especially hoof angle and toe length. squares) and a rolled toe (blue line and squares). The peak indicates heel lift.
The higher peak in the flat shoe condition represents a more abrupt
Hooves trimmed with a low heel and long toe have a significantly
breakover process.
longer breakover duration, although other stride variables like stride
Reprinted from Van Heel, M.C., van Weeren, P.R., Back, W. (2006b) Shoeing sound
duration do not change significantly (Clayton, 1988, 1990a,b). On
warmblood horses with a rolled toe optimises hoof-unrollment and lowers peak loading.
the other hand, rocker, rolled and square toe shoes (Fig. 8.11) did
Equine Vet. Jour. 38 (3), 258–62, with permission of the Equine Veterinary Journal.
not significantly alter breakover duration of horses trotting on a
hard surface (Clayton et al., 1991; van Heel et al. 2006b) or on a
rubber floor (Willemen et al., 1996). However, breakover tends to After the hoof leaves the ground, the limb swings forward to reach
occur earlier (Eliashar et al., 2002) and is smoother with a lower its position of maximal protraction, and is then retracted prior to
peak loading (van Heel et al., 2006b; Fig 8.12). In contrast to the contact with the ground. The final retraction is important for reduc-
effect of heel wedges, shoes that facilitate breakover do not reduce ing the forward velocity of the hoof relative to the ground and so
the peak distal interphalangeal joint moment nor the absolute force decreasing hoof deceleration at ground contact and preventing the
exerted on the navicular bone (Eliashar et al., 2002). horse from stumbling. Protraction of the limb is driven by muscles
Off-center breakover has been associated with swing abnormali- in the proximal limb, with the distal limb following passively (Back
ties of the foot that interfere with the opposite limb and contouring et al., 1995). In the forelimb the joints from the carpus proximally
the hoof or shoe on one side are common techniques to correct for are driven by muscular action, while the fetlock, pastern and
this. Recent research by Keegan et al. (2005) using gyroscopic coffin joints move in response to inertial effects (Lanovaz et al.,
sensors mounted on the hoof confirmed that shoes with a con- 1999). As maximal protraction is approached, the motion of the
toured lateral branch induce greater lateral roll during breakover. proximal limb is slowed and reversed by muscular action, while the
However, this effect was only observed at a trot and was limited to distal limb continues moving forward until resisted by the passive
the first half of breakover. structures (bones, ligaments, tendons). Swinging the limbs back and
forth uses considerable energy, and a number of energy-saving mech-
anisms have evolved. One of the most important is the use of elastic
Flight arc structures as springs; energy is stored when elastic tissues are
The flight arc of the hoof represents the summation of all the joint stretched as the limb is loaded during the stance phase, then released
movements in the limb (Fig. 8.13). The highest point in the flight during unloading to bounce the limb off the ground and assist in
arc occurs soon after lift-off with a second, smaller elevation, which flexing the joints. At the trot the SDFT, DDFT and SL are maximally
may coincide with an upward flip of the toe, at the time of maximal stretched at midstance which corresponds with the time of maximal
protraction. This gives a slightly biphasic flight arc (Clayton, 1990a; weight-bearing. Thus, the elbow, carpal and fetlock joints behave
Balch et al., 1991a, 1997; Back et al., 1995). elastically during the stance phase at the trot (Clayton et al., 1998).

157
 8 The role of the hoof and shoeing

Fig 8.13  (A) Special shoes with glass-fiber molds that were mounted on the
hooves of trotter and recorded using high-speed cine cameras. (B) Three-
dimensional rotations (pitch, red; yaw,blue; roll, green) of the hoof during
4 trotting. could be illustrated in (B) a three-dimensional hoof coordination
ROLL diagram.
Reprinted from Fredricson, I. and Drevemo, S. (1971) A new method of investigating equine
3 locomotion. Equine Vet. J. 3, 137–140, with permission of the Equine Veterinary Journal.

PITCH

YAW
A

Degrees Degrees
50
Support Swing Support Swing Support Swing 40

Yaw angle
20
0 0
Yaw –20
Pitch angle

–40
–50
Pitch 40
20

Roll angle
–100
0
Roll –20
–150 –40
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900
B Time (msec)

Below the elbow and stifle joints the horse’s limbs are designed pastern becomes more upright creating a broken-back hoof pastern
to move in a sagittal plane, which is another energy-saving strategy. axis (Bushe et al., 1987). Conversely, when horses are trimmed with
The distal limbs sometimes deviate from this ideal pattern by being a short toe and/or long heels, the hoof angle becomes more upright
abducted (winging) or adducted (plaiting) during protraction as a and the pastern angle becomes more sloping. Raising the heels
result of slight asymmetries in the articular surfaces, which also have decreases strain in the DDFT and DAL (Leach, 1983; Bushe et al.,
a tendency to cause breakover to occur on the medial or lateral side 1987; Thompson et al., 1993; Riemersma et al. 1996a), but
of the toe. If horses that naturally show these deviations are shod increases strain in the SDFT and SL (Willemen et al., 1999b). One
in a manner that forces them to breakover the center of the toe, it of the goals of hoof trimming is to achieve a flat landing with the
creates torsional forces before and after breakover. As the hoof objective of disseminating the forces on the foot as much as pos-
leaves the ground, these torsional forces cause it to deviate medially sible. When the hoof angle is more upright, the hoof has a more
or laterally, depending on the type of asymmetry. Careful observa- exaggerated heel first landing (Clayton, 1988; Back et al., 1995).
tion of the horse in motion, together with an examination of the Barrey (1990) looked at the relation between hoof angle and
wear pattern on the ground surface of the shoe or hoof, will reveal force distribution: 75% of the weight was borne by the heels when
the preferred side of breakover. If the horse is shod to facilitate the hoof angle was 39°, and this was reduced to 57% when the
breakover at the preferred location, there will often be a marked hoof angle was increased to 55° (Fig. 8.14). This agrees with the
reduction in winging or plaiting and this, in turn, affects the hoof’s finding that a larger toe angle results in more strain of the hoof wall
angle of approach to, and contact with, the ground. at the toe (Thomason, 1998). Nevertheless, the quarters experience
Light-weight tactile stimulation devices attached around the higher forces during stance than the toe (Barrey, 1990).
pastern may increase the height of the flight arc and thus may rep- Many racehorses are trimmed with an acute hoof angle because
resent a useful adjunct in physical therapy aiming at mobilization it is believed that the long toe low heel conformation enhances
of joints and reinforcement of (atrophied) muscles (Clayton et al., performance by increasing stride length. Comparison of the trot
2011). However, short-term habituation does occur, especially in the stride for a normal hoof angle versus an acute hoof angle showed
forelimbs (Clayton et al., 2008). Application of the stimulators for no significant changes in stride length or suspension, and the
short periods is recommended, although it is unknown how rapidly flight arc of the hoof was almost identical with the two angula-
horses rehabituate to the effect of tactile stimulation of the pastern tions (Clayton, 1990a; Balch et al., 1997; Girtler et al., 1995).
over different sessions. However, the acute hoof angle was associated with an increased
frequency of toe-first contacts, which was thought to be a conse-
quence of the proprioceptive reflexes ensuring a fairly flat place-
Effects of hoof manipulations ment of P3 regardless of the shape of the hoof capsule (Clayton,
1990a). Since toe-first contacts are associated with a tendency to
trip or stumble, this may be an undesirable effect. Decurnex et al.
Hoof angle (2009) investigated the effect of training on proximal hoof cir-
When a horse is trimmed with relatively long toes and/or short cumference in young Thoroughbred racehorses being prepared for
heels the hoof angle becomes more acute or sloping, while the racing. Front hoof circumference immediately below the coronary

158
 Effects of hoof manipulations

Normal

47°

Force ratio 75% 61% 57%

on quarters

Hoof force
pattern at
Quarter
the walk Toe Quarter
Toe

A t t t t

Quarter

16.9 ±6.5
Toe
.1

±2.0
.8 ± 242
Force (daN/kg)

Force (daN/kg)

16.9 ±5.4 4.5±2.0

10.0 ±6.2
pe = 453

Walk
.7
31
Impulse =
±2
Initial slo

6.7± 3.5
6.5

3.4±1.5
18

Date
19.0 40.4 54.4 100 55.8
±2.5 ±5.0 ±7.6 ±3.8

23.7±6.1
Force (daN/kg)

Force (daN/kg)

Trot
.2

10.7±4.6
± 382

.1
25

4.4 ± 1.3
1072.9

±2

1.5±0.8
6.9
44

27.8 400
B ±7.9 ±6.1
Fig 8.14  (A) The influence of foot axis on the individual vertical hoof force distribution at the toe and at the quarters. (B) Forces experienced at the medial
toe and quarters at walk (above) and trot (below) in a group of 10 horses. Values on the graphs indicate mean +/− SD for force peaks, slope of initial slope,
impulse and times of occurrence of the force peaks and troughs.
Reprinted from Barrey, E. (1990) Investigation of the vertical hoof force distribution in the equine forelimb with an instrumented horseboot. Equine Vet. J. 9 (Suppl.), 35–38, with permission
from the Equine Veterinary Journal.

band was measured weekly with a measuring tape in all horses The duration of breakover was prolonged with the acute hoof
present at the stable. Most horses showed a similar pattern of angulation and the orientation of the limb segments at the start of
change. The proximal hoof circumference decreased during the breakover suggested an increased tension in the DAL and navicular
training periods and increased when the horse was rested. Appar- ligaments (Fig. 8.15). The effects of an acute hoof angle on break-
ently, horses showed a decrease in circumference during race train- over may be mitigated on a softer surface that allows penetration
ing that reversed when they were rested, thus contributing to a of the toe during the terminal part of the stance phase, since flexion
possible ‘environmental’ explanation for the long toe low heel of the coffin joint reduces tension in the DAL and navicular
conformation often found in racehorses. Furthermore, measure- ligaments.
ment of front hoof circumference is a simple method to assess The hoof wall grows approximately 1 cm every 6 weeks, with
change in hoof shape. It provides an opportunity to investigate the the wall at the toe growing faster than at the heels (Hertsch et al.,
relationships between specific training, hoof shape and soundness 1996). When a horse is shod the hoof wears mainly in the heel
(Peel et al., 2006). region. Consequently, the hoof angle changes between farriery

159
 8 The role of the hoof and shoeing

treatments and this is associated with alterations in the pressure injury. Therefore, horses should receive regular farrier treatment
on the navicular area (Hermans, 1984). Willemen et al. (1999b) every 6–8 weeks or earlier when considerable changes in hoof
calculated a decrease in force on the navicular bone of 24% when angle can be expected in a particular horse. Recent studies also
the angle was 6° more upright. The mean toe angle of the fore- indicate that, by 2 weeks of age, foals already show a limb prefer-
limbs of the 12 horses used in that study was 55°. At the end of ence when grazing at pasture (Van Heel et al., 2006a; Fig. 8.16),
an 8-week shoeing interval, the decrease in hoof angle results in a and thus their hooves can become uneven (Kroekenstoel et al.,
broken-back hoof-pastern axis thereby increasing extension in the 2006; Fig. 8.17).
distal interphalangeal joint and decreasing extension of the fetlock
joint (Tacchio et al., 2002; Moleman et al., 2006; van Heel et al.,
2006c). The deep digital flexor tendon has to compensate for this
Left foot low hoof angle
conformational change which puts the tendon at increased risk for
10.0
8.0
6.0
4.0

Right preference

Left preference
2.0
0.0
-3.0 -2.0 -1.0 1.0 2.0 3.0
-2.0
-4.0
-6.0
-8.0
-10.0
Right foot low hoof angle

Fig 8.17  Relationship (r = 0.89) between the strength of preference during

field observations and the degree of unevenness. On the x-axis the results
of preference in the field, expressed as mean Z-value; on the y-axis
unevenness in hoof angle at 27 weeks. Note that the Z-value is below the
level of significance (1.96) in the majority of foals, which results from the
inclusion of all observations of the first 31 weeks. In this period laterality
Toed-out Toed-in Base-wide Base-narrow developed and was not yet present in the first weeks. Inclusion of these
observations decreased the mean Z-value while leaving the relationship of
Cow-hocked Bow-legged Base-wide Base-narrow
preference and hoof angle unaltered.
Fig 8.15  The influence of foot axis and conformation of the lower limb on Reprinted from Van Heel, M.C.V., Kroekenstoel, A.M., Van Dierendonck, M., Van Weeren, P.R.,
the flight pattern of the hoof from above. Back, W. (2006a) Uneven feet in a foal may develop as a consequence of lateral grazing
Reprinted from Stashak, T.S. (1987) Adam’s Lameness in Horses, with permission from John behaviour induced by conformation . Equine Vet. J. 38, 646–651, with permission from the
Wiley and Sons. Equine Veterinary Journal.

Fig 8.16  A foal grazing at pasture in the typical

posture it has to adopt to reach the ground.
Reprinted from Van Heel, M.C.V., Kroekenstoel, A.M., Van
Dierendonck, M., Van Weeren, P.R., Back, W. (2006a) Uneven
feet in a foal may develop as a consequence of lateral
grazing behaviour induced by conformation . Equine Vet. J.
38, 646–651, with permission from the Equine Veterinary
Journal; picture courtesy of P.J.H.M. Meeus, DVM.

160
 Effects of shoe manipulations

Hoof length
Long hooves, often augmented by pads and weights, are a feature
of some gaited breeds in which they are used to give a showy, exag-
gerated elevation of the distal limbs during the swing phase.
In a study designed to investigate the effects of overall hoof length
on the flight arc of the hoof, pads were applied to increase hoof
length by 5 cm without changing the total weight of the shoe-pad A
combination (Balch et al., 1994). Compared with a normal hoof
length, the long hooves were associated with a prolongation of
stride duration, swing duration and breakover, but overall stride
length and stance duration did not change. The flight arc of the hoof
peaked earlier and higher with the longer hooves, but the normal
movement pattern was re-established in the later part of the swing
phase. Although stride length did not change, the prolongation of
the swing phase may be esthetically pleasing. However, longer toes
also lead to more strain on the dorsal hoof wall, which predisposes
to hoof wall pathology (Thomason, 1998). Under natural social
and environmental conditions, hoof length shows seasonal varia-
tions with ‘self-trimming’ attributable to periods of increased wear
being an important phenomenon in the self-maintenance of the B
equine hoof (Florence & McDonnell, 2006).

Effects of shoe manipulations

b
Shoes are applied to protect against too much wear of the hoof wall,
to improve performance and to provide additional support for the aI
horse on slippery surfaces. On the other hand, shoes restrict the
hoof mechanism, increase the weight of the distal limb and increase
the impact shock (Hermans, 1984). When good-quality shoes are c
C a
put on correctly, with the least number of nails and not too far
backwards, and when adequate hoof care by a farrier takes place on Fig 8.18  Basic principles to increase the weight of the shoe–foot
a regular basis, then the aforementioned objections can be more or combination. (A) Double shoe; (B) toe (left), side (middle) and heel (right)
less compensated (Moyer, 1975). Nevertheless, quantitative evalu- weighted shoes; (C) extra toe or heel weight attached to hoof wall.
ation of trimming and shoeing procedures identified significant
differences between farriers in balancing a hoof (Kummer et al.,
2009).
The increase in maximal vertical ground reaction force when shod
was also found by Roepstorff et al. (1999).
Shoe weight Balch et al. (1996) doubled the weight of the shoe from 348–
The distal limb of the horse has evolved to become very light in 417 g to 724–869 g, and did not find any changes in stride length,
weight. One of the goals when shoeing a horse is to keep the weight stride duration or breakover time, but again there were increases in
of the shoe as low as possible (Balch et al., 1997). This decreases the maximal heights of the hoof, fetlock and carpus during the
its inertia, which reduces energy expenditure in protracting and swing phase, with the peak height of the flight arc tending to occur
retracting the limbs. When weight is added to the horse’s limb, for later in the swing phase. The hoof and pastern segments had a more
example by the application of shoes or protective boots, the effect acute angle at initial ground contact, probably as a result of the
depends on the location of the weight (Fig. 8.18) and increases with increased momentum of the distal limb during the swing phase.
a more distal placement. The weight of a shoe is likely to affect both The location of the weight on the hoof may also have an influence
the energetics and the kinematics of locomotion. Therefore, in race- on its effect.
horses, steel training shoes are replaced by aluminum plates for Butler (1995) stated that toe weight would improve fold and heel
racing (Curtis, 1999). The increase in energy expenditure has been weight would improve reach. Willemen et al. (1994) applied 88-g
estimated to be the same for 1 oz (28.4 g) at the foot as for 30 oz toe weights to the fore hooves of Standardbreds. During trotting on
(852 g) at the withers; the maximal weight that should added be to the treadmill the flight arc was elevated with the extra weight. Inter-
the hoof is 5–10 oz (142–284 g) (Butler, 1995). estingly, the effects of toe weights on stride kinematics varied with
Willemen et al. (1997) investigated the influence of shoeing on the individual locomotion pattern: horses that initially showed too
stride kinematics of young horses that were shod the first time little carpal flexion showed more flexion, whereas those with ade-
(Table 8.1). The average weight of the horses was 516 kg and the quate carpal flexion did not show a significant change. The effect
average weight of the shoes was 478 g. With shoes there was an on stride length was doubtful and there were no significant changes
increase in the maximal height of the flight arc of the hoof and in stride duration or relative stance duration.
greater flexion of the coffin, fetlock and carpal joints during the When egg bar shoes were put on young horses, the increased
swing phase, which improved the quality or ‘animation’ of the trot. ‘animation’ effect was less than that found in horses wearing flat
Stride duration and the relative swing phase duration were longer shoes. It has been speculated that it might be an effect of the
in the shod horses, but stride length was not significantly different. increased weight distribution over the heels of this bar shoe (Wil-
During the stance phase the load on the navicular bone increased lemen et al., 1997).
by 14% when shod, probably as a consequence of these swing phase The Seattle shoe was designed to increase the energetic efficiency
adaptations which were achieved with a less protracted forelimb. of locomotion but its main effect appeared to be due to its weight

161
 8 The role of the hoof and shoeing

Table 8.1  Kinetic comparison of unshod versus shod conditions Table 8.2  Kinetic comparison of three shoe manipulations of a
in a group of 12 horses trotting on a treadmill at 4 m/s group of 12 horses trotting on a treadmill at 4 m/s

Variable Unshod Shod Variable Unshod Flat shoe Egg bar

Stride length (cm) 2.78 ± 0.12 2.82 ± 0.12 Maximum GRF (N) 6364 ± 600 6421 ± 647 6477 ± 699
Stride duration (ms) 694 ± 31 706 ± 28* Vertical displacement at 7.2 ± 1.8 8.0 ± 1.5 7.9 ± 1.5
scapular spine (cm)
Stance phase duration (%) 37.0 ± 1.3 36.0 ± 1.2*
Maximal total moment at 158 ± 30 175 ± 48 175 ± 46
Range of pro/retraction (°) 43.5 ± 1.6 42.7 ± 1.3* coffin joint (Nm)
Maximum protraction (°) 112.0 ± 1.3 111.2 ± 0.8* Maximal total moment at 862 ± 124 835 ± 148 829 ± 100
Maximum retraction (°) 68.5 ± 1.8 68.5 ± 1.5 fetlock joint (Nm)

Swing phase retraction (%) 9.8 ± 3.7 6.9 ± 3.4* Maximal moment at 211 ± 40 231 ± 67 231 ± 63
fetlock joint of DDFT (Nm)
Range of carpal motion (°) 86.7 ± 7.1 98.2 ± 6.1*
Maximal moment at 726 ± 115* 676 ± 110 674 ± 97
Maximum carpal flexion (°) 80.6 ± 6.5 92.5 ± 6.0* fetlock joint of SDFT and
SL (Nm)
Range of fetlock motion (°) 81.2 ± 8.3 91.5 ± 8.8*
Force on navicular bone 3060 ± 438* 3546 ± 526 3504 ± 459
Maximum fetlock extension (°) 25.5 ± 2.7 24.3 ± 3.5
by DDFT (N)
Maximum fetlock flexion (°) 55.6 ± 8.3 67.2 ± 8.5*
*Indicates a statistically significant difference compared with the value of the flat
Range of vertical displacement 10.0 ± 3.6 17.4 ± 4.5* shoe (p < 0.05). DDFT, deep digital flexor tendon; GRF, ground reaction force; Nm,
hoof (cm) Newton meters; SDFT, superficial digital flexor tendon; SL, suspensory ligament.
Values are mean ± SD.
Range of vertical displacement 18.0 ± 3.9 20.0 ± 3.7*
Reprinted from Willemen, M.A., Savelberg, H.C.C.M. and Barneveld, A. (1997) The
fetlock (cm)
improvements of the gait quality of sound trotting warmblood horses by normal
Range of vertical displacement 11.0 ± 3.4 13.9 ± 3.0* shoeing and its effect on the load on the lower forelimb. Livestock. Prod. Sci. 52,
carpus (cm) 145–153, with permission from Elsevier.

Maximum total moment coffin 158 ± 30 175 ± 48

joint (Nm)
Maximum total moment fetlock 862 ± 124 835 ± 148
joint (Nm)

*Indicates a statistically significant difference between unshod and shod (p < 0.05).
Nm, Newton meters. Values are mean ± SD.
Reprinted from Willemen, M.A., Savelberg, H.C.C.M. and Barneveld, A. (1997) The
improvements of the gait quality of sound trotting warmblood horses by normal
shoeing and its effect on the load on the lower forelimb. Livestock. Prod. Sci. 52,
145–153, with permission from Elsevier. A B

of 310–325 g compared with control shoes weighing 76–85 g

(Wilson et al., 1992).
Weight attached to the feet increases the height to which the feet
are lifted from the ground through greater limb flexion. This is a
well-known effect used to improve the performance of gaited, car-
riage horses. Effects related to the location of the weight on the foot
(heel/toe), or direct beneficial influence on stride length, remain
speculative. Empirically, young trotters can find their balance more
easily using weights, and interference can be prevented.

C D
Shoe length
Fig 8.19  Shoes to support the caudal heel at stance. (A) Egg bar;
The hoof is more elastic in the heel area and its geometry changes (B) straight bar; (C) (adjustable) heart bar; (D) trailer with caulk at the
when loaded: the heels expand and sink caudally, while the toe lateral heel.
retracts (Douglas et al., 1998). Lengthening the heels of the shoe
depending on foot conformation in routine shoeing automatically
allows for this change in shape (Balch et al., 1997). (Auer & Butler, 1986; Willemen et al. 1999b) (Table 8.2). The net
The extended heel of an egg bar shoe does not change the posi- effect is that egg bar shoes have a negligible effect on tendon strain
tion of the center of pressure within the hoof during the stance patterns in sound ponies at a walk (Riemersma et al., 1996a).
phase. Compared with flat shoes, however, it reduces the torque at Bouley shoes, which have a lengthened heel, may be effective,
the fetlock joint by changing the orientation of the GRF vector so however, in cases of severed tendons, especially if the DDFT or DAL
that its line of action is closer to the fetlock joint on its dorsal side is involved (Figs 8.19 & 8.20).

162
 Effects of shoe manipulations

D
Fig 8.20  The longer heels of a Bouley shoe support a dropped fetlock joint
caused by severe flexor tendon lacerations.
B
In the absence of tension in the DDFT, the hoof rotates to such
an extent that the point of application of the GRF vector is directly
below the center of rotation of the coffin joint. The tip of the toe
lifts up and the hoof rolls back onto the heels, which puts the hoof
in an unstable position. In these cases, egg bar shoes stabilize the
hoof, and provide pain relief in horses with injury of the DDFT or
DAL, laminitis or navicular disease (Auer & Butler, 1986). Bar shoes C E
may also be useful to prevent lifting of the tip of the toe on soft
ground (Wright & Douglas, 1993). Fig 8.21  Toe manipulations of shoes to reduce breakover time. (A) Rolled
In horses trotting on a rubber floor there was no significant dif- shoe; (B) rocker toed shoe; (C) full roller motion shoe; (D) square toe shoe;
ference in limb loading between egg bars and flat shoes (Willemen (E) reversed shoe.
et al., 1999b), although the ‘animation’ effect was more pronounced
with flat shoes. Egg bar shoes, as a result of having more ground
support and stability, should prevent the heels from sinking into unshod hooves the impact shock had a velocity of 450 m/s2 and
the ground and allow reestablishment of heel growth leading to an this increased to 800 m/s2 when steel shoes were applied, which is
increased toe angle (Scheffer & Back, 2001, Chateau et al., 2006b). in accordance with Hertsch et al. (1996) and Back et al. (2006).
Furthermore, it has been proven indirectly that the dorsal laminar Various shock-absorbing pads have been developed to decrease the
blood flow is enhanced by egg bar or heart bar shoes in laminitic amount of shock transmitted to the limb (Marks et al., 1971). Good
horses, as a result of distributing the weight more caudally and thus attenuation of the impact shock was achieved using an aluminum
protecting the toe region of the foot (Ritmeester et al., 1998). It shoe combined with a polyurethane pad, which had a value of
should be remembered that egg bars on the fore hooves are more 200 m/s2. Benoit et al. (1993) recommended using a shoe with a
likely to be stepped on and torn off by overreaching. Bell boots can pad of rubber or polyurethane and working the horse on soft
be applied to the forelimbs to prevent this. footing to reduce the risk of lameness. A polyurethane/elastomer/
Bar shoes in general provide more ground support and thus viscoelastic hoof pad is a much better shock absorber than rubber
might reduce limb rotation and pain in horses that are lame due to (Marks et al., 1971; Vasko & Farr, 1984; Rooser et al., 1988; Back
bone spavin. This effect can also be achieved using a unilateral et al., 2006). For an optimal effect, hoof pads or silicone should be
trailer (heel extension) at an angle of 45° (Stashak, 1987; Balch put under pressure to prevent it moving and to show increased
et al., 1997; Martinelli & Ferrie, 1997). Trailers are also supposed shock damping (Jörgensen & Ekstrand, 1988). This is also ideal for
to give more medial or lateral support and to change the medio­ increasing the surface area to disseminate high-frequency oscilla-
lateral landing pattern depending on whether they are applied on tions at impact.
the inside or outside. In foals with contracted flexor tendons toe An additional benefit is that increased deformation of the hoof
extensions are used to stretch the DAL and DDFT at the end of the also leads to an increase in the hoof mechanism and thus
stance phase. circulation.
It should be noted that most of the studies of the effects of shoes
and pads on impact shock have evaluated the effects at the level of
Shoe width the hoof wall. Little information is available regarding attenuation
As mentioned previously, the geometry of the hoof changes when of the amplitude or frequency of the vibrations within the hoof. In
the hoof is loaded: the heels expand and sink caudally, and the toe vitro studies have suggested that the hoof acts as a filter to protect
retracts (Douglas et al., 1998). Increasing the width of the shoe the more proximal bones and joints from the potentially damaging
accommodates for these movements (Balch et al., 1997). Another effects of impact shock (Dyhre-Poulsen et al., 1994; Lanovaz et al.,
application for widening the shoes is in foals with angular limb 1998; Willemen, 1998).
deformities. The goal is to bring the weight-bearing surface of the
limb, and thus the point of application of the GRF, under the fetlock
joint as opposed to under the hoof capsule. Toe of the shoe
Breakover is defined as the time from the moment of heel-off to
Hoof pads toe-off. In theory, the speed of hoof rotation during breakover can
be enhanced by shoes that facilitate the movement of the hoof in
Shoes increase pressure on the navicular region (Willemen et al., rolling over the toe or by raising the heel (Fig. 8.21). For horses
1997, 1999b). Compared with horses that are bare-footed, shod trotting even on a hard surface, however, the duration of breakover
horses have less damping of the impact forces and both the median was no different for a rolled toe, a rocker toe or a square toe, com-
and maximal amplitudes of the frequencies are higher (Balch et al., pared with a flat shoe (Clayton et al., 1991; Willemen et al., 1996).
1991b; Benoit et al., 1993; Dyhre-Poulsen et al., 1994). A possible disadvantage, particularly with the square-toed shoes, is
Benoit et al. (1993) explored the shock absorbing damping effect that they direct the breakover point to the middle of the toe, which
on the hoof wall of different combinations of shoes and pads. For is not the natural position in all horses. This might enhance the

163
 8 The role of the hoof and shoeing

likelihood of lameness, especially at high speeds when rapid move- not taken into account, which may have affected the accuracy of the
ments of the hoof occur (Clayton, 1990a; Wilson et al., 1992). measurements.
Therefore, Caudron et al. (1997b) successfully used a full roller With a toe wedge, strain in the DDFT and DAL increases, whereas
motion shoe to provide easy breakover in all directions for horses strain in the SDFT and SL decreases both at a walk and a trot
with poorly balanced hooves, whereas the principle behind these (Thompson et al., 1993; Lawson et al., 2007). Since the DAL
shoes was proven by Van Heel et al. (2006b). has no muscular component that can actively change its length,
strain in this structure is totally dependent on limb configuration,
especially the angle of the coffin joint. The DAL is normally maxi-
Wedges mally strained at the start of breakover, which is when heel wedges
have their greatest effect on the GRF (Fig. 8.22). This emphasizes
Side wedges the importance of the DAL in influencing limb forces and move-
Wedges are applied to one or both heels for specific reasons. In ments in the final part of the stance phase. Raising the heels seems
horses with bone spavin, a lateral wedge is used to relieve pressure appropriate in DAL injury, though this may slightly increase SDFT
from the medial side of the hock. In horses with upward fixation loading. During recuperation from DAL injury, however, the limita-
of the patella, lateral wedges should enhance inward rotation of the tions on exercise make it unlikely that the safety margin of the SDFT
hoof and promote rolling over the medial toe. In the light of will be exceeded even with heel wedges in place. Toe wedges are
evidence-based medicine, Back et al. (2003) studied the effect of used clinically as a passive flexion test in lameness examination.
lateral heel wedges on sagittal and transverse plane hind limb kine- Nevertheless, one should realize that a balanced foot is the ideal.
matics. It was found that lateral heel wedges cause significant Altered foot orientation as a result of the use of wedges influences
changes in the transversal plane angles of all joints in the hind limb. intra-articular pressure and articular contact area at the level of the
Just before the end of the stance phase, the stifle joint becomes distal interphalangeal joint. A study by Viitanen et al. (2003a) has
more adducted whereas the tarsal and fetlock joint become more demonstrated a significant increase in distal interphalangeal joint
abducted, in fact a more ‘cow hocked’ position of the tarsal joints. pressure and pressure on the dorsal side of the joint by elevating
Conceivably, lateral heel wedges relieve tension from the medial the heels by 5°.
patellar ligament and decrease pressure on the medial side of the
tarsal joint. However, the fetlock joint experiences considerably
more out of plane stress (Back et al., 2003) while digital imbalance Toe grabs and heel caulks
puts the horse at risk for interphalangeal joint injury (Chateau On a soft, slippery surface, caulks are screwed into the horse’s shoes,
et al., 2002). where they act in a similar manner to spikes in a runner’s shoes
Colahan et al. (1991) and Wilson et al. (1998) found that lateral (Thompson & Herring, 1994). Toe grabs are used in racing Thor-
wedges shifted the center of pressure to the lateral side of the hoof. oughbreds with the objective of increasing the propulsion at the
Firth et al. (1988) put lateral wedges under the feet of foals and toe. There have not been any studies to prove these beneficial effects.
found a rapid compensation of bone strain to this alteration. Lateral However, results of a preliminary study indicate that landing and
corrective trimming restored the mediolateral balance of 15 horses breakover in horses with toe grab shoes occurs much more abruptly
with chronic digital lameness using a radiological protocol to assess which could represent an increased risk for injury (Ryan et al., 2006;
the imbalance of the foot (Caudron et al., 1998). Schaer et al., 2006). A correlation has been shown between the
presence of toe grabs and the incidence of breakdowns involving
Heel/toe wedges the suspensory-sesamoidean apparatus (Kane et al., 1996). The
higher the toe grab, the greater the risk of injury. Thompson and
Heel wedges facilitate rolling over the toe and are used to relieve Herring (1994) found also a decrease in the dorsal metacarpopha-
pressure from the heels, as in navicular disease. A wedge of 6° langeal, and an increase in the dorsal inter phalangeal joint excur-
resulted in a 24% load reduction on the navicular bone in trotting sion in the sagittal plane during the stance phase, while in the
horses (Willemen et al., 1999b). Heel wedges caused only slight transverse plane the distal limbs appeared to take a limb conforma-
changes in strain of the SDFT, DDFT and SL during walking tion similar to a varus limb.
(Riemersma et al., 1996a,b), though a larger increase in SDFT strain
has been recorded at the trot (Stephens et al., 1989). Using in vivo
data and an accurate subject-specific model, Lawson et al. (2007) Rims, clips and nails
calculated that at walk and trot, heel wedges decrease DDFT peak
A minimum number of nails should be used, usually 6–8, to mini-
strain and increase SL peak strain. Calculated SDFT peak strain,
mize damage to the hoof wall and to reduce blocking the hoof
however, increases at a walk but not a trot. With heel wedges there
mechanics to a minimum (Balch et al., 1997). Therefore nailing
is an earlier shift of the center of pressure from the mid-hoof to the
should be done from the toe to the widest part of the hoof, with
toe and the unloading of the heels is enhanced, whereas toe wedges
the clinches being formed one-third of the way up in a line parallel
delay the forward shift of the center of pressure and the unloading
to the ground. Clips and rims reduce the shearing stress upon the
of the heels is delayed (Riemersma et al., 1996a,b; Wilson et al.,
nails especially when more traction is needed, for example in barrel
1998; Crevier-Denoix et al., 2001). Heel wedges affect the sagittal
racing and polo (Stashak, 1987; Balch et al., 1997; Martinelli &
plane kinematics of the digital joints both in vitro (Deguerce et al.,
Ferrie, 1997).
2001) and in vivo. A wedge of 6° significantly increases maximal
flexion of the proximal (PIPJ) and distal (DIPJ) interphalangeal
joints while decreasing maximal extension of these joints at heel-off
both at a walk and at a trot (Degueurce et al., 2001; Chateau et al., Effect of footing
2004, 2006a). Inverse effects were observed with the use of toe
wedges (except for PIPJ maximal extension at a trot). In the same Properties of the ground
studies, it was found that heel wedges significantly increase maximal
extension of the metacarpophalangeal joint at a walk (Chateau The hoof–ground interaction has been found to be the major deter-
et al., 2004) but not at a trot (Chateau et al., 2006a). The latter minant in studies relating track properties to limb pathology, both
finding is in contrast with the study of Scheffer and Back (2001) in horses (Kane et al., 1996) and in humans (Folman et al., 1986).
that found a decrease in maximal fetlock extension when horses Two important phenomena play a role: the banking and the
were shod with heel wedges. However, in that study the PIPJ was surface type.

164
 Effect of footing

10 10
Tendon strain at the walk Tendon strain at the trot

Strain (%) 5 5

Strain (%)
0 0
ICL ICL
SL SL
-5 SDFT -5 SDFT
DDFT DDFT

-10 -10
0 20 40 60 80 100 0 20 40 60 80 100

A Time (% of stride) Time (% of stride)

Fig 8.22  (A) Strain in the palmar soft tissues at walk (left) and trot (right)
(Schamhardt et al., 1991); SL, suspensory ligament; ICL, inferior check
ligament; SDFT, superficial digital flexor tendon; DDFT, deep digital flexor
tendon. (B) Some examples of wedged shoes used to change strain
distribution among the palmar soft tissues. Left, shoe with wedged heel;
center, shoe with heel calk; right, shoe with wedge pad.
(A) Reprinted from Schamhardt, H.C., Hartman, W., Jansen, M.O. and Back, W. (1991)
Biomechanics of the thoracic limb of the horse. Swiss Vet. 8, 7–10, with permission from Swiss
vet. (B) Reprinted from Wright, I.M. and Douglas, J., Biomechanical considerations in the
treatment of navicular disease. Vet. Rec. 7, 109–114, Copyright© 1993 with permission from
B BMJ Publishing Group Limited.

Fredricson et al. (1971, 1975 a,b) made three recommendations the softer track due to the lower impact shock combined with more
for banking to improve racetracks for trotters: increase the banking elastic rebound (Buchner et al., 1994; Riemersma et al., 1996b).
of the curve, incorporate a transition curve and eliminate slopes in Newer track surfaces like synthetic all-weather waxed tracks further
the straight part of the track. Davies (1997) also stressed the impor- reduce the impact shock by 50% compared to sand tracks (Robin
tance of the track shape to prevent shin soreness. She found a et al., 2009; Chateau et al., 2009a,b). Furthermore, all-weather
considerable increase in dorsal bone strain when horses gallop waxed tracks smooth horizontal braking as a result of decreased
through turns. friction. As a drawback, they reduce stride length and acceleration
Most tracks and arenas are constructed with two layers: a looser of the hoof at breakover. Hence, the all-weather waxed track
cushion on top of a firmer base. Variations in the depth and seems an ideal training track but is most likely less efficient for
quality of the base and the cushion affect both performance and performing at racing speed.
soundness of the horses (Thomason & Peterson, 2008). Running Barrey et al. (1991) categorized the damping properties into three
on a rough instead of smooth track surface changed the vertical types: no damping (asphalt), friction damping (sand) and struc-
hoof force and balance of the resultant hoof forces (Kai et al., tural damping (wood). Sand tracks can have different properties
1999). Hard surfaces absorb little energy, which leads to fast race depending on the moisture content, dry density and depth of the
times, but they are associated with a high incidence of lameness layers (Ratzlaff et al., 1997). Apart from damping, sand tracks allow
(Cheney et al., 1973; Pratt, 1997). The hardness of the ground is a forward rotation of the hoof at midstance and thus relief of pres-
related to the impact time: the harder the surface, the shorter the sure in the navicular area (Scheffer & Back, 2001). Surprisingly, the
impact time (Drevemo & Hjertén, 1991). Ratzlaff et al. (2005) same study observed a concurrent decrease in maximal fetlock
found a significant correlation between track rebound rate and extension and thus unloading of the fetlock joint.
negative vertical hoof acceleration peaks. Any factor that reduces In wild horses, hoof growth appeared to be related to hoof wear.
phase of the hooves will increase stride efficiency by allowing a By natural selection only horses with good hoof quality survive.
smoother transition from braking to propulsion. Hence, track Ovnicek et al. (1995, 1997) found differences between the hooves
dynamic properties are important in determining the safety of of the wild horses that had lived for a prolonged period on one of
racing surfaces. three surfaces: sand, gravel or firm soil (Fig. 8.23). The form of the
A loose top layer of 5 cm or more can give dynamic response hoof adapts to the surface (Hampson et al., 2011). On sand, less
values that reduce impact by 40–60% (Cheney et al., 1973). Reduc- shock damping was needed so the hoof wall carried more weight.
tion of impact forces and enhancement of energy absorption can The hooves did not show more wear as one would have expected,
be achieved using a layer of wood shavings (Barrey et al., 1991; but on both sides of the wall spikes were formed at the quarters to
Drevemo & Hjertén, 1991). Wood products are effective shock give more grip in the sand (‘natural caulks’). The hoof angle was
absorbers but are sufficiently resilient that little impulse is lost. about 55°. On a firmer surface of gravel, the hoof angle was more
Wood products (shavings, mulch, chips) mixed with sand give good upright at the toe but the spikes at the quarters were lower, so the
shock damping with lower vibration frequencies (Barrey et al., angle was again 55°. On firm soil the hoof wall was worn
1991). Rubber chips from recycled tires can also be mixed with sand flat without a heel spike, which would not be useful in this type
to reduce impact shock without losing much impulse. Differences of ground. Again the hoof angle was 55°. Regardless of the type of
were found for horses trotting on a concrete track compared with a surface the horse moved over, the toe was always rolled (Ovnicek
rubber or a sand track: stride and swing duration were longer on et al., 2003).

165
 8 The role of the hoof and shoeing

55˚ 55˚ 55˚

11˚ 6˚
A B C

Fig 8.23  Hoof conformation of wild horses living on different ground surfaces. (A) Soft, (B) medium and (C) hard surfaces.
Reprinted from Ovnicek, G., 1997, New Hope for Soundmess, reprinted with kind permission of EDSS Publishers, Equine Digit Support System, Inc.

Practical applications to performance Practical application to lameness

Interference Lameness is caused by pain, a mechanical deficit, a shortage of

oxygen or a neurological problem. Therapeutic trimming and
Horses trimmed with normal angles in their fore hooves and acute shoeing can be used to improve lameness by reducing pain or
angles in their hind hooves showed a prolonged breakover time and influencing limb mechanics (Stashak, 1987; Martinell & Ferrie,
delayed lift-off in the hind hooves (Hermans, 1984; Clayton, 1997; Aoki, 1999; Eliashar, 2007). This section highlights some
1990b). However, the hind limbs were protracted more rapidly to practical applications to treat specific lameness conditions, thereby
re-establish the normal sequence of limb placements by the time illustrating the aforementioned principles.
of ground contact (Clayton, 1990a,b). Therefore, delaying break-
over in the hind hooves seems unlikely to have a beneficial effect
in horses that interfere (Curtis, 1999). Hoof cracks
A more effective solution to interference problems may be to
Hoof wall defects can be very painful in the acute stages, which
hasten breakover and lift-off in the fore hooves. Furthermore,
makes the horse reluctant to bear weight on the affected foot. Cracks
another solution is to put on square toe shoes more towards
should be relieved from hoof wall pressure and should be repaired
the heels of the hind hooves so that the hoof wall instead of
with the objective of minimizing movement and propagation of the
the shoe hits the forelimbs (toe preventer shoe). Also, weights are
crack using mechanical or composite material (Wilson & Pardoe,
used to balance horses moving at high speed and thus prevent
1998). Cracks are stress raisers and can be detected using photo-
interference.
elastic material (Davies, 1997). The hoof wall tubules seem to play
a major mechanical role in resisting and redirecting these hoof-
Action cracks (Kasapi & Gosline, 1998). Full hoof support, including frog
and sole using polyurethane hoofpads, is one of the treatment
It is recognized among horse trainers that heavier shoes give horses options, as by using this technique the loading pressure on the hoof
more action. Double shoes are applied in trotters just before the is significantly relieved from the nails and the hoofwall, and
race to encourage them to lift the limbs higher, in spite of the fact disseminated over the whole supported foot surface (Van Heel &
that heavier shoes require greater energy expenditure to overcome Back, 2006).
inertia at the start of the swing phase and to overcome momentum
at the end of the swing phase. Young trotters are shod with heavier
shoes in the forelimb than in the hind limb to improve their P3 fracture
balance (Hermans, 1984; Butler, 1995; Ruthe et al., 1997). Too little
folding of the forelimb allows interference by the hind limbs; too Fractures of P3 can be stabilized successfully by reducing hoof
much folding results in elbow hitting. Mediolateral imbalances movements during weight-bearing. Heel expansion can be stabi-
leads to knee hitting by the contralateral limb, which can also be lized by using a full bar shoe with quarter and heel clips (Hermans,
compensated for by lateral weights on the hind limbs so that their 1984; Colles, 1989a,b; Thomason, 1998), and the effect can be
forward swing is outside the forelimbs. enhanced by applying a rigid cast over the clip shoe (Kersjes et al.,
Willemen et al. (1994) found that the effect of toe weights on the 1985). However, frequent trimming does not help to prevent P3
flight arc dependent on the individual locomotion pattern of the fractures from developing in foals (Kaneps et al., 1998).
horse. Horses that normally trotted with too little carpal flexion
showed more action with toe weights. Stride duration and relative Laminitis
stance duration did not change at high speeds (10–11 m/s), and it
is doubtful whether there was any effect on stride length. Laminitis, an aseptic pododermatitis mainly caused by endotoxins
In young horses shod for the first time the weight of the shoes and gastrointestinal disorders, affects the laminae primarily on the
increased the maximal heights of the hoof, fetlock and carpus dorsal aspect of the foot. The avascular damage to those laminae
during the swing phase, which thus resulted in a longer, more ‘ani- and thus to the connection between P3 and the hoof wall can lead
mated’ swing phase at the trot (4 m/s). This is the desired effect of to rotation and sinking of P3 within the hoof capsule. Quantifica-
double shoes applied in harness horses just before the race (Balch tion of limb load distribution using force platforms seems a promis-
et al., 1996; Wilson et al., 1992; Willemen et al., 1997). ing screening method for detecting acute laminitis, grading severity

166
 Practical application to lameness

of lameness and monitoring rehabilitation of horses with chronic

Table 8.3  Growth plate closure times and periods of rapid
laminitis (Hood et al., 2001b). An important parameter is the load
growth for equine long bones
distribution profile (LDP), which is an index of the frequency of
redistribution of load over the four limbs of a horse during a 5-min
measuring period. In acute laminitis, a decreased LDP is observed Growth plate Age at closure Rapid growth period
indicating unwillingness to move and even normal weight shifting
Distal radius 24–30 months <8 months
is avoided as much as possible. In chronic laminitis, LDP increase
with the severity of disease mostly as a result of cycling of load Distal tibia 24 months <6 months
between forelimbs. Except for brief periods during motion, horses
seem physically incapable of shifting substantial weight to the Distal McIII/MtIII 12 months <3 months
hind limbs. Proximal Ph I 6 months <3 months
Basic principles of shoeing treatment in laminitic horses are to
protect the foot and damp impact shock, to facilitate breakover and Mc, metacarpal; Mt, metatarsal; Ph, phalanx
thus relieve tension on the dorsal laminae, and to prevent sole pres- Reprinted from Curtis, S.J. (1999) From Foal to Racehorse, with kind permission of
sure in the toe region by posterior support of the frog and heels Newmarket Farrier Consultancy.
(Goetz, 1987; Olivier et al., 2001; Oosterlaan-Mayer et al., 2002;
Hansen et al., 2005). One of the ensuing factors in P3 rotation is
deep digital flexor tendon tension. However, once rotation of P3 higher risk of developing navicular disease (Balch et al., 1997) and
did occur, deep digital flexor tendon force is much reduced until navicular bone forces are well correlated to the ratio between heel
late stance where it nears normal values again (McGuigan et al., and toe height and to the angle of the distal phalanx with the
2005). From this observation, farriery in laminitic horses with P3 ground (Willemen et al., 1999; Eliashar et al., 2004). Bowker et al.
rotation should focus on regimens that reduce the force in the deep (2001) found a significant increases in contact area and contact
digital flexor tendon at the end of stance. load at dorsiflexion in the joints between the distal phalanx and
In chronic founders it is important to restore normal hoof con- navicular bone and between the middle phalanx and navicular
formation by corrective trimming. A reversed shoe laid back from bone, when the digit was subjected to a load. This may explain
the toe facilitates breakover, relieves pressure from the dorsal the biomechanical background of morphological changes in the
laminae and gives caudal heel support. The use of a pad protects navicular bone.
the sole and damps oscillations (White & Bagget, 1983; O’Grady, Shoeing principles for horses with navicular syndrome aim to
1997). Substantial clinical improvement following therapeutic reduce concussion and facilitate breakover by elevating the heels
shoeing of horses with chronic laminitis should not be expected and rolling the toe. When the horse is bare-foot, pressure of the
during the first 7 days (Taylor et al., 2002). Dorsal hoof wall resec- DDFT on the navicular bone is 14% lower than when flat shoes are
tion and the use of adjustable heart bar shoes, however, still remain applied (Willemen et al., 1999b). Therefore, it might be beneficial
somewhat controversial (Eustace & Caldwell, 1989a,b; Peremans for shod horses suffering for navicular pain to have their shoes
et al., 1991; Jurga, 1997). removed; this will also decrease concussion. Elevation of the heels
using heel wedges may be beneficial (Wintzer, 1971; Leach, 1983;
Turner, 1986, 1988; Bushe et al., 1987; Ratzlaff & White, 1989;
Wright & Douglas, 1993; Keegan et al., 1998) due to the reduced
Navicular syndrome net joint moment at the coffin joint (Willemen et al., 1999b) (Table
Navicular syndrome comprises pathological changes to the navicu- 8.3). Wedge shoes promote breakover but the increased pressure at
lar bone, its ligaments and the DDFT caused by mechanical over- the heels may have a long-term negative effect on the growth and
loading and/or vascular obstruction (Wright et al., 1998; Trotter, health of the heels (Rogers & Back, 2007). Empirically, the use of a
2001). Recent histological examination confirms that lesions in the rolled toe/rocker toe shoe is promoted, although this type of
navicular bone occur predominantly on the fibrocartilage side and shoeing did not reduce breakover time in sound horses (Clayton
are comparable to osteoarthritis (Wright et al., 1998; Viitanen et al., et al., 1991; Willemen et al., 1996), but breakover tends to occur
2003b). Pain is due to pressure of the DDFT on the navicular bursa earlier (Eliashar et al., 2002) and more smoothly with a lower peak
and bone, tension in the navicular ligaments, and elevated intra- loading (van Heel et al., 2006b). Egg bar shoes may prevent hyper-
medullary bone pressure in the navicular bone (Pleasant et al., extension of the coffin joint on a soft surface and rotation of the
1993). According to Wilson et al. (2001), horses with navicular foot, thus avoiding high stresses on the navicular region (Østblom
disease experience higher compressive forces on the navicular bone et al., 1984; Auer & Butler, 1986; Wright & Douglas, 1993). By
compared to normal horses. This finding is attributable to a contrac- preventing the heels from sinking into the ground, they increase
tion of the deep digital flexor muscle in early stance and the both distal and proximal interphalangeal joint flexion by several
increased stress on the flexor surface of the navicular bone puts the degrees at the beginning of stance while decreasing distal interpha-
horse at risk for progression of the disease. Regional analgesia of langeal joint extension by almost 4° at heel-off (Chateau et al.,
the palmar digital nerves lowers the compressive forces on the 2006b). Further, egg bar shoes promote a more gradual loading of
navicular bone indicating that horses with navicular disease attempt the hoof (flatter loading curve) with lower peak pressures, particu-
to unload their heels in response to pain (McGuigan & Wilson, larly at the heel region (Rogers & Back, 2007). Egg bars in combina-
2001). Hence, the whole mechanism identifies navicular disease as tion with mixed sand tracks allowed a forward rotation of the hoof
a possible end point for a variety of heel related painful conditions. at midstance and thus relief of pressure in the navicular area (Schef-
According to Williams (2001), however, horses with navicular fer & Back, 2001).
disease have abnormal limb-loading force patterns that are not
altered by loss of sensation in the palmar digital region and thus
some horses seem predisposed to develop the disease as a result of Arthrosis and arthritis
an inherently abnormal gait pattern.
The navicular bone in the forelimb is significantly larger than in The high-frequency oscillations together with rapid loading of the
the hind limb, probably to compensate for the larger forces in the joints experienced at impact promote the development of osteoar-
forelimb (Gabriel et al., 1997; Bentley et al., 2007), while the shape throsis (Radin et al., 1981, 1999). Prevention and treatment aim to
of the navicular bone has been related to the biomechanical risk of reduce these impact oscillations with a shock-absorbing pad (Marks
developing lameness (Dik et al., 2001). Smaller feet seem to have a et al., 1971; Benoit et al., 1993). In vitro tests, however, revealed that

167
 8 The role of the hoof and shoeing

the hoof capsule (Willemen, 1998) and the distal limb were able
to attenuate impact shock quite considerably (Dyhre-Poulsen et al.,
1994; Lanovaz et al., 1998). A polyurethane/elastomer/viscoelastic
pad is a much better shock absorber than rubber (Marks et al., 1971;
Vasko & Farr, 1984; Rööser et al., 1988; Back et al. 2006). A leather
pad is more susceptible to variations in moisture content of the
ground (Stashak, 1987). For arthritis of the distal interphalangeal
joint, Caudron et al. (1997a,b, 1998) used a special full roller
motion shoe to facilitate breakover and to reduce stress and thus
pain from the ligaments and the joint capsule while adapting to the
breakover path of the particular horse, using a similar principle as
Van Heel et al. (2006b).

A
Spavin
Bone spavin is an aseptic osteoarthritis on the dorsomedial aspect
of the distal intertarsal joint or the tarsometatarsal joint (Stashak,
1987). The rationale behind treating bone spavin by lowering the
medial side of the hoof or elevating the lateral heel is that it may
relieve pressure from the medial side (Firth et al., 1988; Colahan
et al., 1991). Indeed, Back et al. (2003) proved that when using
lateral heel wedges in the hind limb, the tarsal joint becomes more
abducted just before the end of the stance phase, in fact a more ‘cow
hocked’ position of the tarsal joints. Conceivably, lateral heel
wedges decrease pressure on the medial side of the tarsal joint. It
has also been suggested that heel wedges and rolled toes prevent
the hind limb from hyperextending the distal intertarsal and/or
tarsometatarsal joints. Trailers have been used in selected cases to
reduce rotation of the distal limb, but this can also exacerbate the
symptoms (Stashak, 1987; Balch et al., 1997; Martinelli & Ferrie, B
1997).

Patellar fixation
In horses with patellar fixation, lateral heel wedges and rolled toes
prevent hyperextension and outward rotation of the stifle and thus
avoid locking the patella (Stashak, 1987). Most affected horses are
aged 3–5 years. When corrective trimming and shoeing are com-
bined with training of the appropriate muscles, symptoms often
disappear. Back et al. (2003) confirmed that practical experience in
his study on the effect of lateral heel wedges on sagittal and trans-
verse plane hind limb kinematics. It was found that by using lateral
heel wedges the stifle joint becomes more adducted, and thus these
lateral heel wedges would relieve tension from the medial patellar C
ligament. Empirically, some horses, especially trotters, respond
Fig 8.24  Shoeing principles for flexural deformities in foals. (A) Toe
better to a medial wedge than a lateral wedge, as this would also
extension for contracted deep digital flexor tendon (left) treated with a
prevent them from interfering.
raised heel (right). (B) Contracted superficial digital flexor tendon. (C) Flaccid
flexor tendons (left) treated with a heel extension and appropriate hoof
Flexural limb deformities trimming (right).

The greatest threat to the young foal’s locomotor system is the foal
itself. In the first 4–5 months of its life a foal grows as much as can be supported by shoes with longer heels (Curtis, 1992a; Ellis,
during the rest of its life. Just recently, it has been hypothesized that 1998).
foals from older mares, that are poorly developed at birth because
of the smaller placenta but then receive a surplus of milk, are prone
to develop pathology in their locomotor system. Asynchronous
Angular limb deformities
bone–tendon growth results in flexural deformities. In the rapidly growing young foal asymmetric bone growth, caused
Contraction of the DDFT is responsible for a hyperflexed coffin by congenital or acquired factors can lead to angular limb deformi-
joint found at 1–6 months of age, while at 1 year of age SDFT ties. They should be corrected at the fetlock joint before 3 months
contraction leads to a hyperflexed fetlock joint (Fig. 8.24). Treat- of age, while in those at the carpus/tarsus region this should be
ment options depend on the severity of the contraction. Possible corrected before 6–8 months of age (Table 8.3). At that time the
treatments for DDFT contraction include lowering the heels, growth plates start to close. Treatment options of first choice are
extending the toe of the shoe using a cast, and desmotomy of the corrective wall/toe rasping, a medial/lateral extension shoe (Dallric,
distal accessory (check) ligament. For SDFT contraction a wedged Dallmer Salzhausen-Putensen, Germany; Baby Glu, Mustad Hoof-
pad would be an option. So-called ‘weak’ flexor tendons are often care SA, Bulle, Switzerland) and a limited exercise regime (Curtis,
seen in premature foals and lead to hyperextension of the distal 1992b, 1999). The objective is to center the hoof-bearing surface
limb. If this does not have a traumatic origin, the initial treatment under the fetlock joint (Fig. 8.25).

168
 References

Conclusion

The major effects of farriery are similar in most horses of a group,

but there are likely to be minor differences in the response of indi-
vidual horses. Basic farriery considerations that influence locomo-
tion are toe angle and weight of the shoe, which interact with the
type of footing. The mechanisms by which trimming and shoeing
affect locomotion and lameness involve relieving pain, altering the
mechanics of the stance phase and influencing inertia during the
swing phase. In experimental studies the mean effect of a particular
modification on the kinetics and kinematics is often smaller than
the differences between individuals. Therefore, individual responses
may have to be taken into account when evaluating farriery effects.
In young foals, regular trimming is necessary from 1 month of
age (every 4–6 weeks). If pathological deformities or hoof uneven-
ness start to develop, therapeutic trimming should be performed at
least every 2 weeks. Prevention of these deformities is the best
option, of course, because forced correction of conformation by
Fig 8.25  Diagram of a foal with a fetlock valgus deviation (left) treated with trimming and corrective shoeing at an older age will lead to lame-
a medial extension glue-on shoe (right). ness sooner or later.

References
Anderson, G.F., 1992. Evaluation of the hoof hooves. J. Am. Vet. Med. Ass. 198, microscopic changes in articular cartilage
and foot relevant to purchase. Vet. Clin. 1980–1989. and ligaments. Am. J. Vet. Res. 62,
North Am. Equine Pract. 8, 303–318. Balch, O.K., Butler, D., Collins, M.A., 1997. 414–424.
Aoki, O., 1999. Biomechanical analysis of Balancing the normal foot: hoof Bowker, R.M., Breuer, A.M., Vex, K.B., et al.,
horse shoeing. Equine Vet. J. 30 (Suppl.), preparation, shoe fit and shoe modification 1993. Sensory receptors in the equine foot.
629–630. in the performance horse. Equine Vet. Am. J. Vet. Res. 54, 1840–1844.
Arabian, A.K., Lanovaz, J.L., Clayton, H.M., Educ. 9, 143–154. Bowker, R.M., Linder, K.E., Sonea, I.M., Guida,
2001. Determination of hoof mass and Balch, O.K., Clayton, H.M., Lanovaz, J.L., L.A., 1995. Sensory nerve fibers and
centre of mass from morphological 1994. Effects of increasing hoof length on receptors in the equine distal forelimbs and
measurements. Equine Vet. J. 33 (Suppl.), limb kinematics of trotting horses. Proc. their potential role in locomotion. Equine
46–49. Ann. Conv. Am. Assoc. Equine Pract. 40, Vet. J. 18, 141–146.
Auer, J., Butler, K.D., 1986. An introduction 40–43. Buchner, H.H.F., Savelberg, H.H.C.M.,
into the Kaegi equine gait analysis system Balch, O.K., Clayton, H.M., Lanovaz, J.L., Schamhardt, H.C., Merkens, H.W., 1994.
in the horse. Proc. Ann. Conv. Am. Assoc. 1996. Weight-and length-induced changes Kinematics of treadmill versus overground
Equine Pract. 31, 209–226. in limb kinematics in trotting horses. Proc. locomotion in horses. Vet. Quart. 16,
Back, W., 2001. The role of hoof and shoeing. Ann. Conv. Am. Assoc. Equine Pract. 42, S87–S90.
In: Back, W., Clayton, H.M. (Eds.), Equine 218–219. Burn, J.F., 2006. Time domain characteristics
locomotion, first ed. W.B. Saunders, Barrey, E., 1990. Investigation of the vertical of hoof-ground interaction at the onset
London, pp. 135–166. hoof force distribution in the equine of stance phase. Equine Vet. J. 38,
Back, W., Remmen, J.L., Knaap, J., de Koning, forelimb with an instrumented horseboot. 657–663.
J.J., 2003. Effect of lateral heel wedges on Equine Vet. J. 9 (Suppl.), 35–38. Burn, J.F., Brockington, C., 2001.
sagittal and transverse plane kinematics of Barrey, E., Landjerit, B., Wolter, R., 1991. Shock Quantification of hoof deformation using
trotting Shetland ponies and the influence and vibration during the hoof impact on optical motion capture. Equine Vet. J. 33
of feeding and training regimes. Equine different track surfaces. Equine Exerc. (Suppl.), 50–53.
Vet. J. 35, 606–612. Physiol. 3, 97–106. Burn, J.F., Wilson, A., Nason, G.P., 1997.
Back, W., Schamhardt, H.C., Hartman, W., Bayer, A., 1973. Bewegungsanalysen an Impact during equine locomotion:
Barneveld, A., 1995. Kinematic differences trabrenpferden mit hilfe der ungulographie. techniques for measuring and analysis.
between the distal portions of the Zbl. Vet. Med. A 20, 209–221. Equine Vet. J. 23 (Suppl.), 9–12.
forelimbs and hind limbs of horses at the Benoit, P., Barrey, E., Regnault, J.C., Brochet, Bushe, T., Turner, T.A., Poulos, P.W., Harwell,
trot. Am. J. Vet. Res. 56, 1522–1528. J.L., 1993. Comparison of the damping N.M., 1987. The effect of hoof angle on
Back, W., Van Schie, M.H.M., Pol, J.N., 2006. effect of different shoeing by the coffin, pastern and fetlock joint angles.
Synthetic shoes attenuate hoof-impact in measurement of hoof acceleration. Acta Proc. Am. Assoc. Equine Pract. 33,
the trotting warmblood horse. Equine Anat. 146, 109–113. 729–738.
Comp. Exerc. Phys. 3, 143–151. Bentley, V.A., Sample, S.J., Livesey, M.A., Butler, D., 1995. The principles of
Balch, O., Ratzlaff, M., Hyde, M., 1991b. The Scollay, M.C., Radtke, C.L., Frank, J.D., et horseshoeing doug butler enterprises, Inc
effects of different pads on locomotion al., 2007. Morphologic changes associated www.dougbutler.com.
forces exerted by a horse exercising on a with functional adaptation of the navicular Caudron, I., Grulke, S., Farnir, F.,
high speed treadmill. Proc. Ann. Meet. bone of horses. J. Anat. 211, 662–672. Vanschepdael, P., Serteyn, D., 1998.
Assoc. Equine Sports Med. 10, 3–7. Bowker, R.M., Atkinson, P.J., Atkinson, T.S., Clinical and radiological assessment of
Balch, O., White, K., Butler, D., 1991a. Factors Haut, R.C., 2001. Effect of contact stress in corrective trimming in lame horses. Vet. Q.
involved in the balancing of equine bones of the distal interphalangeal joint on 20, 131–135.

169
 8 The role of the hoof and shoeing

Caudron, I., Miesen, M., Grulke, S., the equine forelimb in the stance phase of Dejardin, L.M., Arnoczky, S.P., Cloud, G.L.,
Vanschepdael, P., Serteyn, D., 1997a. the trot. Equine Vet. J. 30, 384–389. Stick, J.A., 2001. Photoelastic stress analysis
Clinical and radiological assessment of Clayton, H.M., Sigafoos, R., Curle, R.D., 1991. of strain patterns in equine hooves after
corrective trimming in lame horses. J. Effect of three shoe types on the duration four-point trimming. Am .J. Vet. Res. 62,
Equine. Vet. Sc. 17, 375–379. of breakover in sound trotting horses. J. 467–473.
Caudron, I., Miesen, M., Grulke, S., Equine Vet. Sci. 11, 129–132. Dejardin, L.M., Arnoczky, S.P., Fox, D.B.,
Vanschepdael, P., Serteyn, D., 1997b. Clayton, H.M., White, A.D., Kaiser, L.J., Ciavelli, M.J., Makidon, P.E., 1997. The use
Radiological assessment of the effect of a Nauwelaerts, S., Lavagnino, M., Stubbs, of photoelastic technology in the analysis
full roller motion shoe during N.C., 2008. Short-term habituation of and localization of strain on the loaded
asymmetrical weightbearing. Equine Vet. J. equine limb kinematics to tactile equine hoof – an in vitro pilot study. Vet.
23 (Suppl.), 20–22. stimulation of the coronet. Vet. Comp. Surg. 26, 413.
Chateau, H., Degueurce, C., Denoix, J.M., Orthop. Traumatol. 21, 211–214. Dik, K.J., van den Belt, A.J., van den Broek, J.,
2004. Effects of 6 degree elevation of the Colahan, P., Leach, D., Muir, G., 1991. Center 2001. Relationships of age and shape of the
heels on 3D kinematics of the distal of pressure location of the hoof with and navicular bone to the development of
portion of the forelimb in the walking without hoof wedges. Equine Exerc. navicular disease: a radiological study.
horse. Equine Vet. J. 36, 649–654. Physiol. 3, 113–119. Equine Vet J. 33, 172–175.
Chateau, H., Degueurce, C., Denoix, J.M., Colborne, G.R., Heaps, L.A., Franklin, S.H., Douglas, J.E., Biddick, T.L., Thomasson, J.J.,
2006a. Three-dimensional kinematics of 2009. Horizontal moment around the Jofriet, J.C., 1998. Stress/strain behaviour
the distal forelimb in horses trotting on a hoof’s centre of pressure during walking in of the equine laminar junction. J. Exp. Biol.
treadmill and effects of elevation of heel a straight line. Equine Vet. J. 41, 242–246. 201, 2287–2297.
and toe. Equine Vet. J. 38, 164–169. Colles, C.M., 1989a. The relationship of frog Drevemo, S., Hjertén, G., 1991. Evaluation of
Chateau, H., Degueurce, C., Denoix, J.M., pressure to heel expansion. Equine Vet. J. a shock absorbing woodchip layer on a
2006b. Effects of egg-bar shoes on the 21, 13–16. harness race-track. Equine Exerc. Physiol. 3,
3-dimensional kinematics of the distal Colles, C.M., 1989b. A technique for assessing 107–112.
forelimb in horses walking on a sand track. hoof function in the horse. Equine Vet. J. Dyce, K.M., Wensing, C.J.G., 1980. De
Equine Vet. J. 36 (Suppl.), 377–382. 21, 17–22. anatomie van het paard. Uitgeverij Bohn,
Chateau, H., Degueurce, C., Jerbi, H., Crevier-Denoix, N., Roosen, C., Dardillat, C., Scheltema, Holkema, Utrecht.
Crevier-Denoix, N., Pourcelot, P., Audigié, Pourcelot, P., Jerbi, H., Sanaa, M., et al., Dyhre-Poulsen, P., Smedegaard, H.H., Roed, J.,
F., et al., 2002. Three-dimensional 2001. Effects of heel and toe elevation Korsgaard, E., 1994. Equine hoof function
kinematics of the equine interphalangeal upon the digital joint angles in the investigated by pressure transducers inside
joints: articular impact of asymmetric standing horse. Equine Vet. J. 33 (Suppl.), the hoof and accelerometers mounted on
bearing. Vet. Res. 33, 371–382. 74–78. the first phalanx. Equine Vet. J. 26,
Chateau, H., Robin, D., Falala, S., Pourcelot, Curtis, S.J., 1992a. Farriery in the treatment of 326–366.
P., Valette, J.P., Ravary, B., et al., 2009b. flexural deformities and a discussion on Eliashar, E., 2007. An evidence-based
Effects of a synthetic all-weather waxed applying shoes to young horses. Equine assessment of the biomechanical effects of
track versus a crushed sand track on 3D Vet. Educ. 4, 193–197. the common shoeing and farriery
acceleration of the front hoof in three Curtis, S.J., 1992b. The assessment and farriery techniques. Vet. Clin. North Am. Equine
horses trotting at high speed. Equine Vet. J. treatment of mediolateral deformities in Pract. 23, 425–442.
41, 247–251. foals. Equine Athelete 9, 15–19. Eliashar, E., McGuigan, M.P., Rogers, K.A.,
Chateau, H., Robin, D., Simonelli, T., Pacquet, Curtis, S.J., 1999. From foal to racehorse, first Wilson, A.M., 2002. A comparison of three
L., Pourcelot, P., Falala, S., et al., 2009a. ed. Newmarket: R and W Publications. horseshoeing styles on the kinetics of
Design and validation of a dynamometric Davies, H.M., 2002. No hoof, no horse! breakover in sound horses. Equine Vet. J.
horseshoe for the measurement of The clinical implications of modelling 34, 184–190.
three-dimensional ground reaction force on the hoof capsule. Equine Vet. J. 34, Eliashar, E., McGuigan, M.P., Wilson, A.M.,
a moving horse. J. Biomech. 9, 336–340. 646–647. 2004. Relationship of foot conformation
Cheney, J.A., Shen, C.K., Wheat, J.D., 1973. Davies, H.M.S., 1996. A technical report on and force applied to the navicular bone of
Relationship of track surface to lameness in the distribution of strain in the hoof wall sound horses at the trot. Equine Vet. J. 36,
the thoroughbred racehorse. Am. J. Vet. of a standing horse before and after 431–435.
Res. 34, 1258–1289. trimming. Pferdeheilkunde 12, 679–680. Ellis, D.R., 1998. Conditions of the hoof wall
Clayton, H.M., 1988. Comparison of the stride Davies, H.M.S., 1997. Non invasive in young horses and corrective farriery with
of trotting horses trimmed with a normal photoelastic method to show distribution regard to limb deformities. Equine Vet.
and a broken-back hoof axis. Proc. Ann. of strain on the hoof wall of a living horse. Educ. 10, 155–160.
Conv. Am. Assoc. Equine Pract. 33, Equine Vet. J. 23, 13–15. Eustace, R., Caldwell, M.N., 1989a. The
289–298. construction of the heart bar shoe and the
Decurnex, V., Anderson, G.A., Davies, H.M.,
Clayton, H.M., 1990a. The effect of an acute 2009. Influence of different exercise technique of dorsal wall resection. Equine
hoof angulation on the stride kinematics of regimes on the proximal hoof Vet. J. 21, 367–369.
trotting horses. Equine Vet. J. 9 (Suppl.), circumference in young thoroughbred Eustace, R.J., Caldwell, M.N., 1989b.
86–90. horses. Equine Vet. J. 41, 233–236. Treatment of solar prolapse using the
Clayton, H.M., 1990b. The effect of an acute Degueurce, C., Chateau, H., Jerbi, H., heart bar shoe and dorsal hoof wall
angulation of the hind hooves on diagonal Crevier-Denoix, N., Pourcelot, P., Audigié, resection technique. Equine Vet. J. 21,
synchrony of the trotting horse. Equine Vet. F., et al., 2001. Three-dimensional 370–372.
J. 9 (suppl.), 91–94. kinematics of the proximal interphalangeal Firth, E.C., Schamhardt, H.C., Hartman, W.,
Clayton, H.M., Kaiser, L.J., Stubbs, N.C., 2011. joint: effects of raising the heels or the toe. 1988. Measurements of bone strain in foals
Hind limb flexion response to different Equine Vet. J. 33 (Suppl.), 79–83. with altered foot balance. Am. J. Vet. Res.
types of tactile devices. Am. J. Vet. Res. 72, Dejardin, L.M., Arnoczky, S.P., Cloud, G.L., 49, 261–265.
1489–1495. 1999. A method for determination of Florence, L., McDonnell, S.M., 2006. Hoof
Clayton, H.M., Lanovaz, J.L., Schamhardt, equine hoof strain patterns using growth and wear of semi-feral ponies
H.C., Willemen, M.A., Colborne, G.L., photoelasticity. Equine Vet. J. 31, during an annual summer ‘self-trimming’
1998. Net joint moments and powers in 232–237. period. Equine Vet. J. 38, 642–645.

170
 References

Folman, Y., Wosk, J., Voloshin, A., Liberty, S., and lowered heels on the biomechanics of effect of hoof trimming on the occurrence
1986. Cyclic impact on heel strike: a the equine hoof assessed by finite element of distal phalangeal palmar process
possible biomechanical factor in the analysis (FEA). J. Vet. Med. A Physiol. fractures in foals. Equine Vet. J. 26
etiology of degenerative disease of the Pathol. Clin. Med. 47, 73–82. (Suppl.), 36–45.
human locomotor system. Acta. Orthop. Hinterhofer, C., Stanek, C., Haider, H., 2001. Kasapi, M.A., Gosline, J.M., 1997. Design
Trauma. Surg. 104, 363–365. Finite element analysis (FEA) as a model to complexity and fracture control in the
Frederick, F.H., Henderson, J.M., 1970. Impact predict effects of farriery on the equine equine hoof wall. J. Exp. Biol. 200,
force measurement using preloaded hoof. Equine Vet. J. 33 (Suppl.), 58–62. 1639–1659.
transducers. Am. J. Vet. Res. 31, 2279–2283. Hjertén, G., Drevemo, S., 1993. Shortening of Kasapi, M.A., Gosline, J.M., 1998. Exploring
Fredricson, I., Dalin, G., Drevemo, S., Hjertén, the forelimb in the horse during the stance the possible functions of equine hoof
G., 1975a. Ergonomics of poor racetrack phase. Acta Anat. 146, 193–195. wall tubulus. Equine Vet. J. 26 (Suppl.),
design. Equine Vet. J. 7, 63–65. Hjertén, G., Drevemo, S., 1994. Semi- 10–14.
Fredricson, I., Dalin, G., Drevemo, S., Hjertén, quantitative analysis of hoof strike in the Keegan, K.G., Satterley, J.M., Skubic, M.,
G., 1975b. A biotechnical approach to the horse. J. Biomech. 17, 997–1004. Yonezawa, Y., Cooley, J.M., Wilson, D.A., et
design of racetracks. Equine Vet. J. 7, Hobbs, S.J., Mather, J., Rolph, C., Bower, J.A., al., 2005. Use of gyroscopic sensors for
91–96. Matuszewski, B., 2004. In vitro objective evaluation of trimming and
Fredricson, I., Drevemo, S., 1971. A new measurement of internal hoof strain. shoeing to alter time between heel and toe
method of investigating equine Equine Vet. J. 36, 683–688. lift-off at end of the stance phase in horses
locomotion. Equine Vet. J. 3, 137–140. Hobbs, S.J., Mather, J., Rolph, C., Richards, J., walking and trotting on a treadmill. Am. J.
Gabriel, A., Youshi, S., Detilleux, J., Dessy- 2009. The effects of limb posture on Vet. Res. 66, 2046–2054.
Doizé, C., Bernard, C., 1997. Bone relationships between in vitro radial hoof Keegan, K.G., Wilson, D.J., Wilson, D.A.,
morphometric study of the equine strain, load and joint angles. Equine Vet. J. Barnett, C.D., Smith, B., 1998. Effects of
navicular bone: comparison between fore 41, 229–232. balancing and shoeing of the forelimb feet
and rear limbs. J. Vet. Med. A. 44, Hoffmann, K.L., Wood, A.K., Griffiths, K.A., on kinematic gait analysis in five horses
579–594. Evans, D.L., Gill, R.W., Kirby, A.C., 2001. with navicular disease. J. Equine Vet. Sci.
Girtler, D., Kübber, P., Kastner, C.H., Scheidl, Doppler sonographic measurements of 18, 522–528.
M., 1995. Kinematische untersuchungen arterial blood flow and their repeatability Keg, P.R., Schamhardt, H.C., van Weeren, P.R.,
des vorführsbogen bei pferden mit in the equine foot during weight bearing Barneveld, A., 1996. The effect of
unterschiedlichen dorsalwandwinkel des and non-weight bearing. Res. Vet. Sci. 70, diagnostic nerve blocks in the forelimb on
hufes. Wien Tier. Arztl. Mschr. 82, 199–203. the locomotion of clinically sound horses.
145–151. Hood, D.M., Taylor, D., Wagner, I.P., 2001a. Vet. Q. 18, S106–S109.
Goetz, T.H., 1987. Anatomic hoof and shoeing Effects of ground surface deformability, Ker, R.F., Bennett, M.B., Alexander, R.McN.,
considerations for the treatment of trimming, and shoeing on quasistatic hoof Kester, R.C., 1989. Foot strike and the
laminitis in horses. J. Am. Vet. Med. Assoc. loading patterns in horses. Am. J. Vet. Res. properties of the human heel pad. Proc.
190, 1323–1332. 62, 895–900. Inst. Mech. Engrs. 203, 191–196.
Gustås, P., Johnston, C., Roepstorff, L., Hood, D.M., Wagner, I.P., Taylor, D.D., Kersjes, A.W., Németh, F., Rutgers, L.J.E., 1985.
Drevemo, S., 2001. In vivo transmission of Brumbaugh, G.W., Chaffin, M.K., 2001b. Atlas of large animal surgery. Uitgeverij
impact shock waves in the distal forelimb Voluntary limb-load distribution in horses Bunge, Utrecht, p. 133.
of the horse. Equine Vet. J. 33 (Suppl.), with acute and chronic laminitis. Am. J. Knezevic, P., 1966. Measuring of strain in the
11–15. Vet. Res. 62, 1393–1398. hoof capsule of the horse. Proc. Ann. Conv.
Gustås, P., Johnston, C., Roepstorff, L., Johnston, C., Back, W., 2006. Hoof ground Am. Assoc. Equine Pract. 12, 293–304.
Drevemo, S., Lanshammar, H., 2004. interaction: when biomechanical stimuli Kobluk, C., Robinson, R., Gordon, B., Clanton,
Relationships between fore- and hind limb challenge the tissues of the distal limb. C., Trent, A., Ames, T., 1990. The effect of
ground reaction force and hoof Equine Vet. J. 38, 634–641. conformation and shoeing: a cohort study
deceleration patterns in trotting horses. Jörgensen, U., Ekstrand, J., 1988. Significance of 95 thoroughbred racehorses. Proc. Ann.
Equine Vet. J. 36, 737–742. of heel pad confinement for the shock Conv. Am. Assoc. Equine Pract. 35,
Hampson, B.A., Connelley, A.D., de Laat, absorption at heel strike. Int. J. Sports 9, 259–274.
M.A., Mills, P.C., Pollitt, C.C., 2011. Sole 468–471. Kroekenstoel, A.M., van Heel, M.C., van
depth and weight-bearing characteristics of Jurga, F., 1997. Progress in laminitis therapy. Weeren, P.R., Back, W., 2006.
the palmar surface of the feet of feral Hoofcare and Lameness 68, 16–23. Developmental aspects of distal limb
horses and domestic thoroughbreds. Am J Kai, M., Aoki, O., Hiraga, A., Oki, H., Tokuriki, conformation in the horse: the potential
Vet Res. 72, 727–735. M., 2000. Use of an instrument consequences of uneven feet in foals.
Hansen, N., Buchner, H.H., Haller, J., sandwiched between the hoof and shoe to Equine Vet. J. 38, 652–656.
Windischbauer, G., 2005. Evaluation using measure vertical ground reaction forces and Kübber, P., Kastner, J., Girtler, D., Knezevic,
hoof wall strain gauges of a therapeutic three-dimensional acceleration at the walk, P.F., 1994. Erkenntnisse über den einfluss
shoe and a hoof cast with a heel wedge as trot, and canter in horses. Am. J. Vet. Res. der tiefen palmarnervanästhesie auf das
potential supportive therapy for horses 61, 979–985. gangbild des lahmheitsfreien pferdes mit
with laminitis. Vet. Surg. 34, 630–636. Kai, M., Takahashi, T., Aoki, O., Oki, H., 1999. hilfe einer kinematischen Messmethode.
Hermans, W.A., 1984. Hoefverzorging en Influence of rough track surfaces on Pferdeheilkunde 10, 11–21.
hoefbeslag. groene reeks. Uitgeverij Terra components of vertical forces in cantering Kummer, M., Geyer, H., Imboden, I., Auer, J.,
Zutphen, Zutphen. thoroughbred horses. Equine Vet. J. 30 Lischer, C., 2006. The effect of hoof
Hertsch, B., Höppner, S., Dallmer, H., 1996. (Suppl.), 214–217. trimming on radiographic measurements
The hoof and how to protect it without Kane, A.J., Stover, S., Gardner, I.A., et al., 1996. of the front feet of normal warmblood
nails, first ed. Salzhauzen-Putensen: Horseshoe characteristics as possible risk horses. Vet. J. 172, 58–66.
Dallmer Publications. factor in fatal musculoskeletal injury of Kummer, M., Gygax, D., Lischer, C., Auer, J.,
Hickman, J., Humphrey, M., 1997. Hickman’s thoroughbred racehorses. Am J. Vet. Res. 2009. Comparison of the trimming
farriery, second ed. J.A. Allen, London. 37, 1147–1152. procedure of six different farriers by
Hinterhofer, C., Stanek, C., Haider, H., 2000. Kaneps, A.J., O’Brien, T.R., Willits, N.H., quantitative evaluation of hoof
The effect of flat horseshoes, raised heels Dykes, J.E., Stover, S.M., 1998. The radiographs. Vet. J. 179, 401–406.

171
 8 The role of the hoof and shoeing

Lanovaz, J.L., Clayton, H.M., Colborne, G.R., Nigg, B.M., Denoth, J., Neukomm, P.A., 1981. Pietra, M., Guglielmini, C., Nardi, S., Gandini,
Schamhardt, H.C., 1999. Forelimb Quantifying the load on the human body: G., Cipone, M., 2004. Influence of weight
kinematics and joint moments during the problems and some possible solutions. In: bearing and hoof position on Doppler
swing phase of the trot. Equine Vet. J. 30 Morecki, A., Fideliu, K., Kedzior, K., Wit, A. evaluation of lateral palmar digital arteries
(Suppl.), 235–239. (Eds.), Biomechanics VII. University Park in healthy horses. Am. J. Vet. Res. 65,
Lanovaz, J.L., Clayton, H.M., Watson, L.G., Press, Baltimore, pp. 88–99. 211–215.
1998. In vitro attenuation of impact shock O’Grady, S., 1997. Shoeing the laminitic horse. Pleasant, R.S., Baker, G.J., Foreman, J.H.,
in equine digits. Equine Vet. J. (Suppl. 26), Am. Farriers J. 3, 33–38. Eurell, J.C., Losonsky, J.M., 1993.
96–102. Olivier, A., Wannenburg, J., Gottschalk, R.D., Intraosseous pressure and pathologic
Lawson, S.E., Chateau, H., Pourcelot, P., van der Linde, M.J., Groeneveld, H.T., 2001. changes in horses with navicular disease.
Denoix, J.M., Crevier-Denoix, N., 2007. The effect of frog pressure and downward Am. J. Vet. Res. 54, 7–12.
Effect of toe and heel elevation on vertical load on hoof wall weight-bearing Pollitt, C., 1995. Color atlas of the horse’s
calculated tendon strains in the horse and and third phalanx displacement in the foot. Mosby–Wolfe, London.
the influence of the proximal horse–an in vitro study. J. S. Afr. Vet. Assoc. Pratt, G.W., 1997. Model for injury to the
interphalangeal joint. J. Anat. 210, 72, 217–227. foreleg of the thoroughbred racehorse.
583–591. Oosterlaan-Mayer, B., Back, W., Sloet van Equine Vet. J. 23 (Suppl.), 30–32.
Leach, D.H., 1983. Biomechanical Oldruitenborgh-Oosterbaan, M.M., 2002. Preuschoft, H., 1989. The external force and
considerations in raising and lowering the The effect of a treatment protocol on the internal stresses in the feet of dressage and
heel. Proc. Ann. Conv. Am. Assoc. Equine prognosis of equine laminitis. Tijdschr jumping horses. Z. Säugetierkunde 54,
Pract. 28, 333–342. Diergeneeskd. 127, 644–649. 172–190.
Lejeune, J.P., Schneider, N., Caudron, I., Oosterlinck, M., Pille, F., Back, W., Dewulf, J., Radin, E.L., 1999. Subchondral bone changes
Duvivier, D.H., Serteyn, D., 2006. Gasthuys, F., 2010a. Use of a stand-alone and cartilage damage. Equine Vet. J. 31,
Radiographic evolution of the forelimb pressure plate for the objective evaluation 94–95.
digit in Ardenner horses from weaning to of forelimb symmetry in sound ponies at Radin, E.L., Parker, H.G., Pugh, J.W., Steinberg,
28 months of age and its clinical walk and trot. Vet. J. 183, 305–309. R.S., Paul, I.L., Rose, R.M., 1981. Response
significance. J. Vet. Med. A Physiol. Pathol. Oosterlinck, M., Pille, F., Back, W., Dewulf, J., of joints to impact loading III. Relationship
Clin. Med. 53, 364–370. Gasthuys, F., 2010c. A pressure plate study between trabecular microfractures and
Linford, R.L., 1994. Camera speeds for on fore and hind limb loading and the cartilage degeneration. J. Biomech. 6,
optoelectronic assessment of stride timing association with hoof contact area in 51–57.
characteristics in horses at the trot. Am. J. sound ponies at walk and trot. Vet J. 2010 Ratzlaff, M.H., Hyde, M.L., Hutum, D.V.,
Vet. Res. 55, 1189–1195. Sep 25. [Epub ahead of print]. Rathgeber, R.A., Balch, O.K., 1997.
Marks, D., Mackay-Smith, M.P., Cushing, L.S., Oosterlinck, M., Pille, F., Huppes, T., Gasthuys, Interrelationships between moisture
Leslie, J.A., 1971. Use of an elastomer to F., Back, W., 2010b. Comparison of pressure content of the track, dynamic properties of
reduce concussion to horses’ feet. J. Am. plate and force plate gait kinetics in sound the track and the locomotor forces exerted
Vet. Med. Assoc. 158, 1361–1365. warmbloods at walk and trot. Vet. J. 186, by galloping horses. J. Equine Vet. Sci. 17,
Martinelli, M.J., Ferrie, J.W., 1997. Farrier 347–351. 35–42.
primer for the student and practising Østblom, L.C., Lund, C., Meslen, F., 1984. Ratzlaff, M.H., Shindell, R.M., Debowes, R.M.,
veterinarian. Equine Vet. Educ. 9, 89–97. Navicular bone disease: results of treatment 1985. Changes in digital venous pressures
McGuigan, M.P., Walsh, T.C., Pardoe, C.H., using egg bar shoeing principle. Equine Vet. of horses moving at the walk and trot. Am.
Day, P.S., Wilson, A.M., 2005. Deep digital J. 16, 203–206. J. Vet. Res. 46, 1545–1549.
flexor tendon force and digital mechanics Ovnicek, G., 1997. New hope for soundness. Ratzlaff, M.H., White, K.K., 1989. Some
in normal ponies and ponies with rotation EDSS Publishers, Columbia Falls. biomechanical considerations concerning
of the distal phalanx as a sequel to Ovnicek, G., Erfle, J.B., Peters, D.F., 1995. Wild navicular disease. J. Equine Vet. Sci. 9,
laminitis. Equine Vet. J. 37, 161–165. horse hoof patterns offer a formula for 149–153.
McGuigan, M.P., Wilson, A.M., 2001. The preventing and treating lameness. Proc. Ratzlaff, M.H., Wilson, P.D., Hutton, D.V.,
effect of bilateral palmar digital nerve Ann. Conv. Am. Assoc. Equine Pract. 41, Slinker, B.K., 2005. Relationships between
analgesia on the compressive force 258–260. hoof-acceleration patterns of galloping
experienced by the navicular bone in Ovnicek, G.D., Page, B.T., Trotter, G.W., 2003. horses and dynamic properties of the track.
horses with navicular disease. Equine Vet. J. Natural balance trimming and shoeing: its Am. J. Vet. Res. 66, 589–595.
33, 166–171. theory and application. Vet. Clin. North Ratzlaff, M.H., Wilson, P.D., Hyde, M.L., Balch,
Merkens, H.W., Schamhardt, H.C., 1994. Am. Equine Pract. 19, 353–377. O.K., Grant, B.D., 1993. Relationships
Relationship between ground reaction force Pardoe, C.H., McGuigan, M.P., Rogers, K.M., between locomotor forces, hoof position
patterns and kinematics in the walking and Rowe, L.L., Wilson, A.M., 2001. The effect and joint motion during the support phase
trotting horse. Equine Vet. J. 17 (Suppl.), of shoe material on the kinetics and of the stride of galloping horses. Acta Anat.
67–70. kinematics of foot slip at impact on 146, 200–204.
Merkens, H.W., Schamhardt, H.C., van Osch, concrete. Equine Vet J. 33 (Suppl.), 70–73. Riemersma, D.J., van den Bogert, A.J., Jansen,
G.J.V.M., van den Bogert, A.J., 1994. Parsons, K.J., Wilson, A.M., 2006. The use of M.O., Schamhardt, H.C., 1996a. Influence
Ground reaction force patterns of Dutch MP3 recorders to log data from equine of shoeing on ground reaction forces and
warmblood horses at normal trot. Equine hoof mounted accelerometers. Equine Vet. tendon strains in the forelimbs of ponies.
Vet. J. 25, 134–137. J. 38, 675–680. Equine Vet. J. 28, 126–132.
Moleman, M., van Heel, M.C., van Weeren, Peel, J.A., Peel, M.B., Davies, H.M., 2006. The Riemersma, D.J., van den Bogert, A.J., Jansen,
P.R., Back, W., 2006. Hoof growth between effect of gallop training on hoof angle in M.O., Schamhardt, H.C., 1996b. Tendon
two shoeing sessions leads to a substantial thoroughbred racehorses. Equine Vet. J. 36 strain in the forelimbs as a function of gait
increase of the moment about the distal, (Suppl.), 431–434. and ground characteristics and in vitro
but not the proximal, interphalangeal joint. Peremans, K., Verschooten, F., de Moor, A., limb loading in ponies. Equine Vet. J. 28,
Equine Vet. J. 38, 170–174. Desmet, P., 1991. Laminitis in the pony: 133–138.
Moyer, W., 1975. Lameness caused by conservative treatment versus dorsal Ritmeester, A.J., Blevins, W.E., Ferguson, D.W.,
improper shoeing. J. Am. Vet. Med. Assoc. hoof wall resection. Equine Vet. J. 23, Adams, S.B., 1998. Digital perfusion,
1, 47–52. 243–246. evaluated scintigraphy and hoof wall

172
 References

growth in horses with chronic laminitis Schamhardt, H.C., Hartman, W., Jansen, M.O., racing horses. Vet. Clin. North Am. Equine
treated with egg bar heart bar shoeing and Back, W., 1991. Biomechanics of the Pract. 24, 53–77.
coronary grooving. Equine Vet. J. 26 thoracic limb of the horse. Swiss Vet. 8, Thompson, K.H., Cheung, T.K., Silverman, M.,
(Suppl.), 111–118. 7–10. 1993. The effect of toe angle on tendon,
Robin, D., Chateau, H., Pacquet, L., Falala, S., Scheffer, C.J., Back, W., 2001. Effects of ligament, band and hoofwall strains. J.
Valette, J.P., Pourcelot, P., et al., 2009. Use ‘navicular’ shoeing on equine distal Equine Vet. Sci. 13, 651–654.
of a 3D dynamometric horseshoe to assess forelimb kinematics on different track Thompson, K.N., Herring, L.S., 1994.
the effects of an all-weather waxed track surface. Vet. Q. 23, 191–195. Metacarpophalangeal and phalangeal joint
and a crushed sand track at high speed trot: Seeherman, H.J., 1991. The use of high speed kinematics in horses shod with hoof
preliminary study. Equine Vet. J. 41, treadmills for lameness and hoof balance caulks. J. Equine Vet. Sci. 14, 319.
253–256. evaluation in the horse. Vet. Clin. North Trotter, G., 2001. The biomechanics of what
Roepstorff, L., Johnston, C., Drevemo, S., Am. Equine Pract. 7, 271–309. really causes navicular disease. Equine Vet.
1994. The influence of different treadmill Stashak, T.S., 1987. Adam’s lameness in J. 33, 334–336.
constructions on ground reaction forces as horses. WB Saunders, Philadelphia, pp. Turner, T.A., 1986. Shoeing principles for the
determinant by the use of a force 786–833. management of navicular disease. J. Am.
measuring horseshoe. Equine Vet. J. 17 Stephens, P.R., Nunamaker, D.M., Butterweck, Vet. Med. Assoc. 189, 298–301.
(Suppl.), 71–74. D.M., 1989. Application of hall-effect Turner, T.A., 1988. Proper shoeing and
Roepstorff, L., Johnston, C., Drevemo, S., transducer for measurement of tendon shoeing principles for the management
1999. The effect of shoeing on kinetics and strains in horses. Am. J. Vet. Res. 50, of navicular syndrome. Proc. Ann.
kinematics during the stance phase. Equine 1089–1095. Conv. Am. Assoc. Equine Pract. 33,
Vet. J. 30 (Suppl.), 279–285. Summerley, H.L., Thomason, J.J., Bignell, 299–305.
Roepstorff, L., Johnston, C., Drevemo, S., W.W., 1998. Effect of rider style on Van Dixhoorn, I.D., Meershoek, L.S., Huiskes,
2001. In vivo and in vitro heel expansion deformation of the front hoof wall in R., Schamhardt, H.C., 2002. A description
in relation to shoeing and frog pressure. warmblood horses. Equine Vet. J. 26 of the motion of the navicular bone during
Equine Vet. J. 33 (Suppl.), 54–57. (Suppl.), 81–85. in vitro vertical loading of the equine
Rogers, C.W., Back, W., 2007. The effect of Tacchio, G., Davies, H.M., Morgante, M., forelimb. Equine Vet. J. 34, 594–5497.
plain, eggbar and 6 degrees-wedge shoes Bernardini, D., 2002. A radiographic Van Heel, M.C., Barneveld, A., van Weeren,
on the distribution of pressure under the technique to assess the longitudinal P.R., Back, W., 2004. Dynamic pressure
hoof of horses at the walk. N. Z. Vet J. 55, balance in front hooves. Equine Vet. J. 34 measurements for the detailed study of
120–124. (Suppl.), 368–372. hoof balance: the effect of trimming.
Roland, E., Stover, S.M., Hull, M.L., Dorsch, K., Taylor, D., Hood, D.M., Wagner, I.P., 2002. Equine Vet. J. 36, 778–782.
2003. Geometric symmetry of the solar Short-term effect of therapeutic shoeing Van Heel, M.C., Moleman, M., Barneveld, A.,
surface of hooves of thoroughbred on severity of lameness in horses with Van Weeren, P.R., Back, W., 2005. Changes
racehorses. Am. J. Vet. Res. 64, 1030–1039. chronic laminitis. Am. J. Vet. Res. 63, in location of centre of pressure and
Roland, E.S., Hull, M.L., Stover, S.M., 2005. 1629–1633. hoof-unrollment pattern in relation to an
Design and demonstration of a Taylor, D.D., Hood, D.M., Potter, G.D., Hogan, 8-week shoeing interval in the horse.
dynamometric horseshoe for measuring H.A., Honnas, C.M., 2005. Evaluation of Equine Vet. J. 37, 536–540.
ground reaction loads of horses during displacement of the digital cushion in Van Heel, M.C., van Weeren, P.R., Back, W.,
racing conditions. J. Biomech. 38, response to vertical loading in equine 2006b. Shoeing sound warmblood horses
2102–2112. forelimbs. Am J. Vet. Res. 66, 623–629. with a rolled toe optimises hoof-
Rollot, Y., Lecuyer, E., Chateau, H., Crevier- Thomason, J.J., 1998. Variation in surface train unrollment and lowers peak loading
Denoix, N., 2004. Development of a on the equine hoof wall at the midstep during breakover. Equine Vet. J. 38,
3D model of the equine distal forelimb with shoeing, gait, substrate, direction of 258–262.
and of a GRF shoe for noninvasive travel, and hoof angle. Equine Vet. J. 26 Van Heel, M.C., van Weeren, P.R., Back, W.,
determination of in vivo tendon and (Suppl.), 86–95. 2006c. Compensation for changes in hoof
ligament loads and strains. Equine Vet. J. Thomason, J.J., Bignell, W.W., Sears, W., 2001. conformation between shoeing sessions
36, 677–682. Components of variation of surface hoof through the adaptation of angular
Rööser, R., Ekbladh, R., Lidgren, L., 1988. The strain with time. Equine Vet. J. 33 (Suppl.), kinematics of the distal segments of the
shock absorbing effect of soles and insoles. 63–66. limbs of horses. Am. J. Vet. Res. 67,
Int. Orthop. 12, 335–338. Thomason, J.J., Faramarzi, B., Revill, A., Sears, 1199–1203.
Ruthe, H., Müller, H., Reinhard, F., 1997. Der W., 2008. Quantitative morphology of the Van Heel, M.C.V., Back, W., 2006. Ground
huf. Ferdinand Enke Verlag, Stuttgart. equine laminar junction in relation to surface and poly-urethane (PU) filling alter
Ryan, C.T., Schaer, B.L., Nunamaker, D.M., capsule shape in the forehoof of the pressure distribution pattern in square
2006. A novel wireless data acquisition standardbreds and thoroughbreds. Equine standing horses. Pferdeheilkunde 22,
system for the measurement of hoof Vet. J. 40, 473–480. 592–596.
accelerations in the exercising horse. Thomason, J.J., McClinchey, H.L., Faramarzi, Van Heel, M.C.V., Kroekenstoel, A.M., Van
Equine Vet. J. 38, 671–674. B., Jofriet, J.C., 2005. Mechanical behavior Dierendonck, M., Van Weeren, P.R., Back,
Savelberg, H.H.C.M., van Loon, T., and quantitative morphology of the equine W., 2006a. Uneven feet in a foal may
Schamhardt, H.C., 1997. Ground reaction laminar junction. Anat. Rec. A Discov. Mol. develop as a consequence of lateral grazing
forces in horses assessed from hoof Cell Evol. Biol. 283, 366–379. behaviour induced by conformation.
wall deformation using artificial Thomason, J.J., McClinchey, H.L., Jofriet, J.C., Equine Vet. J. 38, 646–651.
neural networks. Equine Vet. J. 23 2002. Analysis of strain and stress in the Van Wulfen, K.K., Bowker, R.M., 2002.
(Suppl.), 6–8. equine hoof capsule using finite element Evaluation of tachykinins and their
Schaer, B.L., Ryan, C.T., Boston, R.C., methods: comparison with principal strains receptors to determine sensory innervation
Nunamaker, D.M., 2006. The horse- recorded in vivo. Equine Vet. J. 34, in the dorsal hoof wall and insertion of the
racetrack interface: a preliminary study on 719–725. distal sesamoidean impar ligament and
the effect of shoeing on impact trauma Thomason, J.J., Peterson, M.L., 2008. deep digital flexor tendon on the distal
using a novel wireless data acquisition Biomechanical and mechanical phalanx in healthy feet of horses. Am. J.
system. Equine Vet. J. 38, 664–670. investigations of the hoof-track interface in Vet. Res. 63, 222–228.

173
 8 The role of the hoof and shoeing

Vasko, K.A., Farr, D., 1984. A viscoelastic Willemen, M.A., 1998. Een onderzoek naar de Wilson, A.M., Pardoe, C.H., 1998. Equine hoof
polymer as an aid in injury management biomechanische effecten van hoefbeslag. cracks: mechanical considerations and
and prevention in equine athletes. J. Tijdschr. Diergeneesk 123, 408–410. repair techniques. Equine. Vet. Educ.
Equine. Vet. Sci. 4, 278–280. Willemen, M.A., Jacobs, M.W., Schamhardt, Manual 4, 52–56.
Viitanen, M., Bird, J., Smith, R., Tulamo, R.M., H.C., 1999a. In vitro transmission and Wilson, A.M., Pardoe, C.H., 2001.
May, S.A., 2003. Biochemical attenuation of impact vibrations in the Modification of a force plate system for
characterisation of navicular hyaline distal forelimb. Equine Vet. J. 30 (Suppl.), equine gait analysis on hard road surfaces:
cartilage, navicular fibrocartilage and the 245–248. a technical note. Equine Vet. J. Suppl. 33,
deep digital flexor tendon in horses with Willemen, M.A., Savelberg, H.C.C.M., 67–69.
navicular disease. Res. Vet. Sci. 75, Barneveld, A., 1997. The improvements of Wilson, A.M., Seelig, T.J., Shield, R.A.,
113–120. the gait quality of sound trotting Silverman, B.W., 1998. The effect of
Viitanen, M.J., Wilson, A.M., McGuigan, H.R., warmblood horses by normal shoeing and foot imbalance on point of force
Rogers, K.D., May, S.A., 2003a. Effect of its effect on the load on the lower forelimb. application in the horse. Equine.
foot balance on the intra-articular pressure Livestock. Prod. Sci. 52, 145–153. Vet. J. 30, 540–545.
in the distal interphalangeal joint in vitro. Willemen, M.A., Savelberg, H.C.C.M., Wilson, G.H., McDonald, K., O’Connell, M.J.,
Equine Vet. J. 35, 184–189. Barneveld, A., 1999b. The effect of 2009. Skeletal forelimb measurements and
Wagner, I.P., Hood, D.M., 2002. Effect of orthopaedic shoeing on the force exerted hoof spread in relation to asymmetry in
prolonged water immersion on equine by the deep digital flexor tendon on the the bilateral forelimb of horses. Equine Vet.
hoof epidermis in vitro. Am. J. Vet. Res. 63, navicular bone in horses. Equine Vet. J. 31, J. 41, 238–241.
1140–1144. 25–30. Wilson, P.D., Ratzlaff, M.H., Grant, B.D., Hyde,
Wagner, I.P., Hood, D.M., Hogan, H.A., 2001. Willemen, M.A., Savelberg, H.C.C.M., Bruin, M.L., Balch, O.K., 1992. The effects of a
Comparison of bending modulus and yield G., Barneveld, A., 1994. The effect of toe compressible plastic – the Seattle-shoe on
strength between outer stratum medium weights on linear and temporal stride the kinematics of the strides of galloping
and stratum medium zona alba in equine characteristics of Standardbred trotters. Vet. thoroughbred horses. J. Equine Vet. Sci. 12,
hooves. Am. J. Vet. Res. 62, 745–751. Q. 16, S97–100. 374–381.
Weishaupt, M.A., Hogg, H.P., Wiestner, T., Willemen, M.A., Savelberg, H.C.C.M., Wintzer, H.J., 1971. Besitzen der hufbeslag
Denoth, J., Stüssi, E., Auer, J.A., 2002. Jacobs, M.W.H., Barneveld, A., 1996. und die hufplege eine bedeutung bei der
Instrumented treadmill for measuring Biomechanical effects of rocker-toed shoes behandlung und verhutung der
vertical ground reaction forces in horses. in sound horses. Vet. Q. 18, S75–S78. podotrochleose. Wien Tierärtzl. Mschr. 58,
Am. J. Vet. Res. 63, 520–527. Williams, G.E., 2001. Locomotor characteristics 148–151.
White, J.M., Mellor, D.J., Duz, M., Lischer, C.J., of horses with navicular disease. Am. J. Vet. Wright, I.M., Douglas, J., 1993. Biomechanical
Voute, L.C., 2008. Diagnostic accuracy of Res. 62, 206–210. considerations in the treatment of navicular
digital photography and image analysis for Wilson, A.M., McGuigan, M.P., Fouracre, L., disease. Vet. Rec. 7, 109–114.
the measurement of foot conformation in MacMahon, L., 2001. The force and contact Wright, I.M., Kidd, L., Thorp, B.H., 1998.
the horse. Equine Vet. J. 40, 623–628. stress on the navicular bone during trot Gross, histological and histomorphometric
White, N.A., Baggett, N., 1983. A method of locomotion in sound horses and horses features of the navicular bone and related
corrective shoeing for laminitis in horses. with navicular disease. Equine Vet. J. 33, structures in the horse. Equine Vet. J. 30,
Vet. Med. Small Anim. Clin. 5, 775–778. 159–165. 220–234.

174
 C H A PTER 9 

Gait adaptation in lameness

H. H. Florian Buchner

Introduction stride when the disturbance is caused. A supporting limb lameness

is caused by pain during the stance phase, while swinging limb
lameness is caused by problems during the swing phase. Pure
Horses are kept as domestic animals due to their outstanding swinging limb lameness without pain during the stance phase is
locomotor skills, which have been employed for military use, extremely rare and is most often a component of mixed lameness
transportation and sports. Perfect athletic performance, of course, with features of both supporting and swinging limb lameness. The
needs a sound locomotor system. Maintenance of soundness and third classification describes the site of the lameness as a fore or
detection of lameness are of prime importance for horse owners. hind limb lameness, or as a bilateral lameness when both fore limbs
The ability to study the gait of horses and assess the small devia- or both hind limbs are affected.
tions associated with locomotor problems is limited by the physi- If a specific structure can be proven as the origin of lameness and
ological ability of the human eye. Therefore, locomotion research a specific diagnosis is established, locomotion analysis techniques
in horses started very early with studies of the gait of lame horses. can be used to define the associated locomotor patterns. A special
Already in 1899, Muybridge published series of photographic situation is represented by induced lameness, which has been used
plates of both sound and lame horses in his fascinating book, in several clinical studies for evaluations of diagnostic or therapeu-
Animals in Motion. These nice studies were the start of modern tic regimes. Lameness models have been reported to induce hoof
kinematic research about lameness in horses, which is and, will lameness (Merkens & Schamhardt, 1988a; Foreman & Lawrence,
continue to be, a central theme of equine locomotion analysis 1991; Keegan et al., 2000), arthritis of the carpal joint (Auer et al.,
(Leach & Crawford, 1983). 1980; Firth et al., 1987; Peloso et al., 1993) and tendonitis (Silver
et al., 1983; Williams et al., 1984). The ability to induce a more or
less transient lameness in groups of sound horses offers an impor-
Definitions tant and reliable method to study the locomotor pattern in horses
Lameness is not a disease but a symptom of a locomotor distur- with specific, well-defined lameness in a controlled manner that
bance. Generally, lameness can be defined as an alteration of the minimizes individual variations.
normal gait due to a functional or structural disorder in the Locomotion research can provide information regarding several
locomotor system (Wittman, 1931; Knezevic, 1982; Stashak, 1987; aspects of equine locomotion. This chapter will describe the results
Wyn-Jones, 1988; Speirs, 1994; Wilson & Keegan, 1995). A detailed of locomotion research in the area of lameness during the past
description of the locomotion pattern is a central part of each lame- century from different viewpoints and with different objectives:
ness examination. The goal is to localize the cause of the lameness 1. Firstly, the kinematics that are specific to different types of
and to make a diagnosis that is as specific as possible as a basis for lameness will be described. This will provide a more complete
the veterinary therapy. insight into the gait adaptations and how horses deal with
and compensate for pain in a limb. The complete picture
of movement changes of all body parts due to fore or hind
Classification of lameness limb lameness makes it possible to distinguish the original
In a clinical setting various classifications of lameness are used to and compensatory lameness. The findings are of fundamental
differentiate various pathological gait patterns. Such classifications importance for the orthopedic specialist.
provide only a rough framework and a real lameness cannot always 2. The effectiveness of the locomotor adaptations in terms of
be fitted into a single category. Nevertheless, placing a lameness in load reduction in the painful limb and load redistribution
one of the categories is a first step for the clinician in reaching a will be evaluated using kinetic methods. The connection of
diagnosis. Locomotion analysis, on the other hand, should provide kinetic and kinematic findings allows an analysis of the
specific details of the various lamenesses as well as fundamental mechanisms by which a horse manages a lameness.
relations and principles. These can be used to understand the dif- 3. The patterns of some well-defined lameness will be presented
ferent types of lameness and to provide the scientific basis for and analyzed to determine their specific signs, which will
interpreting the observations made during lameness examinations. serve as a diagnostic database.
Figure 9.1 shows the traditional classifications for lameness. The 4. The possible use of locomotion analysis techniques in clinical
cause of the lameness is most often pain in one or more limbs. settings, such as the documentation of diagnostic or
Sometimes diseases of peripheral nerves, blood vessels or muscles therapeutic regimes or the diagnosis of insidious lameness,
cause specific lameness and occasionally purely mechanical restric- will then be described and discussed including the pros and
tions can be found. The type of lameness describes the phase of the cons of these methods.

175
 9 Gait adaptation in lameness

often originates in ailments of the hoof or the distal limb. These

Lameness
ailments cause pain during loading and the horses try to minimize
the pain by changes in various aspects of the locomotor pattern,
which can be described within four categories: the temporal stride
pattern, the movement of the hoof, the angular pattern of the limb
Cause Type Localization joints and head and trunk movements.

Temporal stride pattern

Pain Supporting limb Forelimb
Many studies (Hugelshofer, 1982; Clayton, 1986b; Girtler,
1988a,b,c; Tietje, 1992; Buchner et al., 1995a; Keegan et al., 1997;
Mechanical Swinging limb Hindlimb
Weishaupt et al., 2006) have tried to find relationships between
types of lameness and the timing of the limb placements, described
as stride variables. However, these studies did not find typical
changes or asymmetry patterns. This is in contradiction to subjective
Neurogenic, Mixed Bilateral impressions, where some observers describe that a shortening of the
muscular, vascular loading phase of the lame limb seems to be a typical sign for a
supporting limb lameness (Wittmann, 1931; Ratzlaff et al., 1982;
Fig 9.1  Classification of lameness. Hajer et al., 1988). Nevertheless, some general temporal patterns
can be seen if the lameness is moderate or severe. Slight or ‘subclini-
cal’ lameness, on the other hand, does not show significant tempo-
5. Finally, the effects of several frequently used diagnostic or ral deviations from the sound stride pattern (Buchner et al., 1995a;
therapeutic aids will be evaluated. In these studies locomotion Weishaupt et al., 2006).
analysis technology is used to evaluate the proposed effects Generally, lame horses tend to move with a slower velocity. This
and to support or to reject their use based on scientific is of utmost importance for the interpretation of the stride pattern
fundamentals. in lame horses, because all stride variables are heavily dependent
on velocity (Dusek et al., 1970; Leach & Cymbaluk, 1986; Leach &
Drevemo, 1991). Furthermore, the lameness degree is influenced by
speed, as moderate lameness increases with speed (Peham et al.,
Kinematics of lameness 2000). Comparisons of different assessments, for example before
and after diagnostic anaesthesias, are valid only if recorded at the
The judgment of an observer as to whether a horse is lame or not same velocity or using statistical correction methods (Kübber et al.,
is based on a comparison of the gait or details of the locomotion 1994). Horses reduce their velocity by decreasing stride duration,
pattern with a sound reference. Such a reference for an objective stride length and all the dependent variables, such as stance dura-
determination of a lameness can be found using three different tion or contralateral advanced placement. On the other hand, when
approaches: horses are forced to maintain a constant velocity on the treadmill,
1. A comparison with the ‘normal’ locomotion pattern in the variable results regarding stride duration and stride length in lame
population, which is based on the experience of the clinician. horses have been reported (Buchner et al., 1995a; Pollhammer-
This requires a sufficient number of observations in sound Zeilinger, 1996; Keegan et al., 1997). When showing different
horses to establish a standard and to eliminate the problem degrees of hoof lameness, provoked by the screw model of
of individuality. This ‘standard horse’, however, may not be Merkens & Schamhardt (1988a), horses maintained the velocity by
the best method for kinematic assessments due to the small taking shorter, but quicker, strides than in the sound condition
differences caused by lameness compared to the wide (Table 9.1) (Buchner et al., 1995a; Galisteo et al., 1997; Weishaupt
interindividual variation in sound horses. et al., 2006). However, in horses suffering from navicular disease
2. A comparison using the horse’s own pattern as an individual no consistent changes in stride duration and stride length were
control. Since the sound pattern of the individual is usually found between the lame pattern and the sound pattern after
unknown during lameness examinations, diagnostic nerve diagnostic anesthesia (Pollhammer-Zeilinger, 1996; Keegan et al.,
blocks serve as a tool to restore kinematics that are close to 1997).
the sound pattern during a clinical evaluation. In a research Generally, a consistent feature in all studies is a shortening of the
setting similar comparisons are possible using lameness swing duration in lame horses compared to the same horses
models, where the sound locomotion pattern of each horse without lameness (Hugelshofer, 1982; Tietje, 1992; Buchner et al.,
before lameness induction serves as an individual control and 1995a; Keegan et al., 1997). In other words, horses lengthen the
minute deviations due to lameness can be easily differentiated stance phase duration rather than shortening it to diminish the
from the small intraindividual variations. pain. This is in contradiction to the opinions that shortened stance
3. An assessment of the movements of the left and right sides durations are indicative of supporting limb lameness (Ratzlaff
of the body in terms of locomotor symmetry or asymmetry et al., 1982; Clayton, 1986a). Furthermore, there is no difference in
within an individual horse. This approach allows an stance duration between the lame and the sound limb (Girtler,
immediate comparison of affected and unaffected limbs 1988c; Tietje, 1992; Buchner et al., 1995a; Weishaupt et al., 2006).
independent of individual characteristics. The horses maintain their symmetry; both limbs are kept on the
ground longer with increasing lameness, while both swing dura-
Using one of these three methods several general characteristics of tions decrease. This does not contradict the findings of differences
supporting limb lameness and swinging limb lameness have been in stance durations between left and right limbs in some particular
described. horses. Absolute symmetry is nearly impossible in nature and sig-
nificant left/right differences as sign of handedness or leggedness
are found in various breeds (Drevemo et al., 1987; Deuel & Law-
Supporting limb lameness rence, 1987; Meij & Meij, 1980). Asymmetry in stance duration is
Lameness that results from the efforts of the horse to avoid pain therefore not a sign of a supporting limb lameness, but a typical
during the stance phase is defined as supporting limb lameness and individual characteristic.

176
 Kinematics of lameness

0.50
Table 9.1  Mean and standard deviation of stride variables in 11
horses evaluated when sound and with two degrees of induced
forelimb lameness in the hoof at the trot (3.5 m/s)

Vertical movement (m)

0.30
Variable Lameness degree

0 1 2
0.10 *
Stride duration (ms) *
715 (36) 707 (42) 693 (53)c
-0.10
Stance duration (ms) -25 -15 -5 5 15 25
Lame forelimb 312 (17) 313 (15) 318 (16) R/protraction (degrees)
Sound forelimb 307 (16) 306 (15) 312 (15)c Fig 9.2  Hoof trajectory of a lame forelimb during different degrees of
forelimb lameness. (blue line) Lameness degree 0 (sound); (red line)
Ipsilateral hind limb 291 (20) 287 (25) 290 (22)
lameness degree 1; (green line) lameness degree 2.
a
Diagonal hind limb 291 (17) 284 (20) 285 (19)c Reprinted with permission from: Buchner, H.H.F., Savelberg, H.H.C.M., Schamhardt, H.C.,
Barneveld, A., 1996. Limb movement adaptations in horses with experimentally induced fore
Relative stance duration (%) or hind limb lameness. Equine Vet J 28, 63–70.
Lame forelimb 43.7 (2.4) 44.3 (2.4) 46.0 (2.8)bc
Sound forelimb 43.0 (2.4) 43.4 (2.8) 45.3 (3.4)bc
step duration without lameness (Table 9.1) as a result of the clearly
Ipsilateral hind limb 40.7 (3.1) 40.7 (3.7) 42.0 (3.8)bc shortened suspension between the lame and sound diagonals.
Diagonal hind limb 40.7 (2.9) 40.3 (3.3) 41.3 (3.2) An interesting pattern is seen in the diagonal advanced place-
ment. As already mentioned, during forelimb lameness, the stance
Advanced placement (ms) phases of the forelimbs tend to increase, with the fore hooves being
placed earlier and lifted later. This results in an earlier placement of
Lame diagonal 0 (18) −9 (19)a −16 (26)c the forelimbs in relation to the diagonal hind limbs. Sound horses
Sound diagonal 3 (22) −2 (24) −10 (24)bc place their diagonal limb pairs almost synchronously or even place
the hind limbs earlier than the forelimbs (positive diagonal
Lame to sound forelimb 360 (20) 358 (26) 341 (28)bc advanced placement). The latter sequence may be indicative of
Sound to lame forelimb 355 (18) 350 (19) 352 (28) superior gait quality in dressage horses (Holmström et al., 1994).
Lameness reverses this pattern. Similarly, horses on a treadmill
Ipsilateral to diagonal 358 (19) 357 (20) 359 (21) show this earlier forelimb placement, perhaps due to a need for
hind limb longer ground contact, possibly as a sign of a remaining insecurity
even after a long period of habituation to the treadmill (Buchner
Diagonal to ipsilateral 357 (18) 351 (24) 335 (36)bc
et al., 1994).
hind limb
During hind limb lameness, generally the same temporal pattern
Suspension (ms) is found in most stride variables, however, the amount of changes
in limb timing is much less in hind limb lameness. Stride frequency
Lame diagonal 36 (24) 35 (26) 14 (29)bc increases as in forelimb lameness, but the increase in relative stance
Sound diagonal 39 (20) 38 (20) 32 (27) duration is small and both diagonal advanced placement and sus-
pension phases do not change at all (Buchner et al., 1995a). Horses
Significant differences between the values of a variable for different lameness with hind limb lameness keep this variable constant and perfectly
degrees are indicated by superscripts: a, degree 0 versus 1; b, degree 1 versus 2; symmetrical.
c
, degree 0 versus 2. Diagonal advanced placement is positive when the hind limb There are two possible explanations for this temporal stability.
precedes the forelimb and vice versa. Firstly, load damping may be more effective in the hind limbs as a
Reprinted with permission from: Buchner et al. (1995). result of greater tarsal flexion during loading of the lame limb. And
secondly, a lower percentage of the body weight is carried by the
hind limbs (46.8%, trot) compared with the forelimbs (53.2%)
(Merkens et al., 1993), which may facilitate the management of a
In forelimb lameness, a real asymmetry due to lameness can be supporting limb lameness.
found in the suspension phase at the trot (Table 9.1) (Clayton,
1986a; Buchner et al., 1995a; Weishaupt et al., 2006). The suspen- Summary
sion phase following the lame diagonal stance phase, which means Horses with supporting limb lameness tend to have longer stance
the time when none of the limbs is on the ground after the stance durations in both, lame and contralateral sound limb, but provide
phase of the lame forelimb and diagonal hind limb, is significantly less propulsion during the stance phase of the lame limb, so that
shortened in lame horses. This is a sign of reduced propulsion the following suspension phase is reduced. This is an important
during the stance phase of the lame limb. The suspension following method of reducing the peak loads on the limbs, which will be
the sound diagonal, on the other hand, is nearly unchanged. This discussed in detail in a later paragraph.
asymmetry can only be seen in forelimb lameness, not in hind limb
lameness. The asymmetric suspension phase in forelimb lameness
has some implications for the co-ordination of the placements of
Hoof movement
the different limbs. The duration of the step from the lame to the Hoof movement can be visualized as the hoof trajectory during the
sound forelimb is shorter than the contralateral step or than the stride. Figure 9.2 shows the trajectory of a fore hoof on a treadmill

177
 9 Gait adaptation in lameness

seen from the right side. The stance phase is characterized by a by Knezevic et al. (1982) for a horse with carpal lameness. Changes
constant vertical position on the treadmill belt during the horizon- in limb movement during the swing phase may also cause changes
tal movement from right to left in the figure. After lift-off, the flight in the hoof-landing pattern. However, a lack of quantitative data for
arc of the hoof shows the elevation during the swing phase and ends different lameness causes still precludes making precise conclusions
with the hoof landing at the start of the next stance phase. at this time. To differentiate between individual landing patterns
Different features can be seen in this figure showing typical and specific lameness patterns, it will be necessary to perform
changes in the hoof trajectory due to various degrees of forelimb studies with a number of horses that have a similar diagnosis or the
lameness. The maximal height of the hoof during protraction of the same induced lameness.
limb is said to be lower during both supporting limb lameness and
swinging limb lameness (Stashak, 1987). In supporting limb lame- Summary
ness, a lower flight arc might reduce pain on impact. In swinging The hoof flight is slightly lower in the lame limb compared to the
limb lameness, difficulties in flexing the joints may cause a lowering sound limb. Forelimb lameness causes reduced retraction and
of the flight arc. However, this could not be confirmed in recordings sometimes increased protraction of the lame limb; hind limb lame-
of horses suffering from navicular disease. Several authors investi- ness shows a reversed pattern. The reduced swing duration also
gated this lameness for characteristic patterns, but did not find leads to a shorter step length from lame to sound limb but slightly
consistent changes in maximal hoof height (Ratzlaff & Grant, 1986; higher flight arc in the sound limb.
Pollhammer-Zeilinger, 1996; Keegan et al., 1997). In experimen-
tally induced forehoof lameness (Buchner et al., 1996a) as well as
in patients with forelimb lameness (Girtler et al., 1987) a higher
flight arc was found in the sound forelimb and an unchanged height
Limb movement and joint angle patterns
in the lame limb (Fig. 9.2). During hind limb lameness, a lower The joint movement patterns of the equine limbs are important
flight arc was found in the lame limb, while the sound limb had an indicators of both physiologic locomotor capacity (Holmström
unchanged flight arc (Buchner et al., 1996a). Both patterns give the et al., 1994; Back et al., 1994) and disturbances of the gait due to
same impression, the hoof of the lame side is lower than the con- lameness (Adrian et al., 1977; Ratzlaff & Grant, 1986; Back et al.,
tralateral hoof. In induced toe pain lameness, a shift of the instant 1993; Peloso et al., 1993; Buchner et al., 1996a; Keegan et al.,
of maximal hoof height nearer to midswing has been found (Keegan 1997). The amount of hyperextension of the fetlock joint and
et al., 2000). flexion of the carpal and tarsal joints correlated very well with sub-
Changes in the pro- and retraction of the limbs of lame horses jective judgments of gait quality in the areas of suppleness and
are visible at the walk. Retraction of the forelimbs is slightly reduced strength (Back et al., 1994). In lame horses, the proximal (shoulder,
during forelimb lameness. For the interpretation of this pattern a carpus, stifle, tarsus) and distal (fetlock, coffin) joints reflect differ-
comparison with the changes in hind limb movement during hind ent aspects of limb motion and show different changes due to
limb lameness is interesting. Lame hind limbs show a reduction in lameness.
protraction rather than retraction. Perhaps the position of the limb During supporting limb lameness the horse tries to reduce the
relative to the body center of mass influences this feature. During load on the painful limb. The amount of loading can be measured
walking vertical ground reaction forces of the forelimbs reach peak directly using a force plate, and it can be assessed indirectly from
values in the second half of the stance phase, when the limb is the distal joint patterns during the stance phase. At the trot, after
retracted, which brings it closer to the body center of mass. On the landing the fetlock joint shows increasing hyperextension until the
other hand, the position of the hind limbs relative to the body moment of maximal loading in the middle of the stance phase at
center of mass causes peak loading in the first half of the stance (Fig. 9.3A). The hyperextension then decreases gradually until the
phase (Merkens et al., 1986). The changes in pro- or retraction, end of the stance phase. The fetlock joint angle during the stance
which are more obvious during walking than trotting, might reduce phase is determined by and resembles the pattern of the vertical
the total load on the lame limb by shortening the period of ground reaction force as measured in sound horses with a force
high load. plate (Riemersma et al., 1988) or a force shoe (Ratzlaff et al., 1993).
The changes in the temporal stride pattern also have implications This relationship is valid also in lame horses, when changes in the
for the linear stride variables. The distances between hoof place- fetlock and coffin joint patterns correspond to a decrease in the
ments of sound and lame limbs and vice versa, which are called the vertical ground reaction force in the lame limb and a compensatory
step lengths, might give information about the cause of the lame- increase in the contralateral sound limb (Merkens & Schamhardt,
ness. A shortening of the step length from the lame to the sound 1988b). The correlation of the ground reaction forces with the distal
limb is said to be typical for a supporting limb lameness (Witt- limb joint angle pattern is even used to calculate limb forces based
mann, 1931; Stashak, 1987). Horses with sesamoiditis (Clayton, on kinematic data of distal limb length (Bobbert et al., 2007).
1986b) and hoof lameness (Buchner et al., 1996a) showed this During supporting limb lameness, both the fetlock and coffin
feature. This shortening corresponds with the temporal variables in joint patterns change distinctly with increasing lameness (Table 9.2)
terms of the shorter advanced placement between lame and sound (Buchner et al., 1996a). In the lame limb, fetlock hyperextension at
forelimbs as well as the clearly shortened suspension phase follow- the middle of the stance phase is reduced with each degree of lame-
ing the lame diagonal. The linear stride variables offer further proof ness (Fig. 9.3A). In contrast, fetlock hyperextension in the contra-
of reduced propulsion during the stance phase of the painful limb. lateral sound limb shows increased maximal values (Fig. 9.3B). This
Few studies have described specific changes in the hoof-landing asymmetry indicates a compensation of the reduced loading of the
pattern due to lameness. Toe first or heel first might give informa- lame limb by the contralateral sound limb. Similarly, coffin joint
tion about the localization of the pain, in the heel or toe region flexion is reduced, but the effects are a little less obvious and occur
(Ratzlaff & Grant, 1986; Stashak, 1987; Clayton, 1988; Tietje, 1992; earlier in the stride cycle (Fig. 9.3C,D). Based on this strong correla-
Wilson & Keegan, 1995). Measurements of induced hoof lameness, tion, the fetlock joint pattern can be used as indicator of a support-
caused by pressure-inducing screws on the sole, did not influence ing limb lameness or the supporting limb component of a mixed
the hoof-landing angle (Buchner et al., 1996a), which may have lameness (Back et al., 1993). However, the range of fetlock joint
been due to the position of the screws in the middle between toe motion proved to be less sensitive in detecting slight lameness
and heel. The study of Keegan et al. (2000) inducing the hoof pain (Peloso et al., 1993), probably due to the higher variability in swing
at the toe showed an increased protraction of the lame limb, but phase flexion. Therefore, maximal fetlock hyperextension during
the hoof-landing pattern could not be distinguished. Differences in stance, which resembles the clinical assessment of fetlock sinking,
the hoof position and motion just before landing were described is a sensitive measure of a supporting limb lameness. The

178
 Kinematics of lameness

70 70

50 50
Flexion angle (degrees)

Flexion angle (degrees)

30 30

10 10

-10 -10

-30 -30
0 20 40 60 80 100 0 20 40 60 80 100
A Time (% of stride) B Time (% of stride)

30 70

20 20

Flexion angle (degrees)

Flexion angle (degrees)

10 10

0 0

-10 -10

-20 -20

-30 -30
0 20 40 60 80 100 0 20 40 60 80 100
Time (% of stride) D Time (% of stride)
C

Fig 9.3  Joint angle pattern of the fetlock (A) and coffin (C) joints of a lame forelimb, and the fetlock (B) and coffin (D) joints of the contralateral sound
forelimb during different degrees of forelimb lameness. (blue line) Lameness degree 0 (sound); (red line) lameness degree 1; (green line) lameness degree 2.
Arrows indicate hyperextension (fetlock joint) or flexion (coffin joint) during the middle of the stance phase.
Reprinted with permission from: Buchner, H.H.F., Savelberg, H.H.C.M., Schamhardt, H.C., Barneveld, A., 1996. Limb movement adaptations in horses with experimentally induced fore or hind
limb lameness. Equine Vet J 28, 63–70.

where passive support by the interosseous (suspensory) ligament is

Table 9.2  Joint angle variables of the forelimbs of 11 horses at
the most significant factor. In contrast to the distal joints, flexion
the trot when sound and with two degrees of induced forelimb
of the proximal joints during loading of the limbs is increased
lameness (3.5 m/s)
rather than reduced in the lame limb (Buchner et al., 1996a). In
the shoulder joint of the lame forelimb this increase is rather
Variable Limb Lameness degree small, but tarsal flexion changes more distinctly, which may indi-
cate an increased functioning of a shock absorbing mechanism
0 1 2
(Fig 9.4A–D) (Hjertén et al., 1994; Back et al., 1995b). These
Maximal carpal Lame 80.3 (7.7) 81.1 (6.3) 79.4 (5.2) increases in shoulder and tarsal joint flexion during the stance
flexion phase of the lame limb are not due to increased loading. They are
Sound 79.8 (10.4) 80.1 (10.0) 78.3 (11.2) the result of a more gentle braking of the flexion by the extensor
muscles. Consequently, loading of the lame limb with the body
Maximal fetlock Lame −19.1 (4.0) −17.0 (4.1)a
−11.5 (5.2)bc
weight occurs more gradually, and this reduces the peak forces in
hyperextension
Sound −20.0 (3.7) −21.2 (3.4)a −21.9 (3.9)c the hoof. This finding corresponds with the changes in the ground
reaction force pattern in lame horses: there is a marked decrease in
Maximal coffin Lame 23.6 (3.3) 22.8 (3.5) 20.4 (3.4)bc the first peak of the vertical ground reaction force in a lame limb at
flexion the walk, while the forces at midstance are slightly higher (Merkens
Sound 23.5 (4.0) 24.6 (4.1) 27.3 (3.5)bc
& Schamhardt, 1988a). This increased damping is expressed more
Data are presented as mean (SD), and are expressed in degrees. Significant in the hind limb than the forelimb, which is thought to be related
differences (p < 0.05) between different lameness degrees are indicated by to the considerably greater range of motion in the tarsal/stifle joint
superscripts: a, degree 0 versus 1; b, degree 1 versus 2; c, degree 0 versus 2. complex compared with the shoulder/elbow joint complex (Back
Reprinted with permission from: Buchner et al. (1996). et al., 1995a, b).

asymmetric pattern of lame and contralateral sound limbs can be Summary

used as a symmetry variable for lameness quantifications. The proximal joint pattern shows the efforts of the horse to smooth
In contrast to the distal joints, the proximal joints play a more limb loading, while the distal joint pattern reflects an absolute
active role in lameness management. The shoulder and tarsus nor- decrease in loading of the lame limb. This decrease in limb loading
mally flex as the limbs are loaded, but movements of these joints cannot be achieved by changing the limb movement pattern. It is
are more dependent on muscular control than the fetlock joint, the result of adaptations in head and trunk movements.

179
 9 Gait adaptation in lameness

70 70
Flexion angle (degrees)

Flexion angle (degrees)

50 50

30 30

10 10

-10 -10
0 20 40 60 80 100 0 20 40 60 80 100
A Time (% of stride) B Time (% of stride)

70 70

50 50
Flexion angle (degrees)

Flexion angle (degrees)

30 30

10 10

-10 -10

-30 -30
0 20 40 60 80 100 0 20 40 60 80 100

C Time (% of stride) D Time (% of stride)

Fig 9.4  Joint angle pattern of the tarsal (A) and fetlock (C) joints of a lame hind limb and tarsal (B) and fetlock (D) joints of the contralateral sound hind limb
during different degrees of hind limb lameness. (blue line) Lameness degree 0 (sound); (red line) lameness degree 1; (green line) lameness degree 2. Arrows
indicate flexion (tarsal joint) or hyperextension (fetlock joint) during the middle of the stance phase.
Reprinted with permission from: Buchner, H.H.F., Savelberg, H.H.C.M., Schamhardt, H.C., Barneveld, A., 1996. Limb movement adaptations in horses with experimentally induced fore or hind
limb lameness. Equine Vet J 28, 63–70.

head movement pattern (Girtler & Floss, 1984; Peloso et al., 1993;
Head and trunk movement Buchner et al., 1996b; Keegan et al., 1997). The lowering and lifting
The simplest, most sensitive, and most frequently used indicator for of the head during the stance phase of the lame limb decreases, with
the clinical diagnosis of lameness is the characteristic vertical move- a compensatory increase in both movements during the stance
ment of the horse’s head and trunk (Stashak, 1987; Wyn-Jones, phase of the contralateral sound forelimb (Fig. 9.5). These changes
1988). A more or less asymmetric pattern of head movement is the are proportional to the degree of lameness and in severe lameness
starting point for each student to diagnose a forelimb lameness and, the first wave may not be visible, there is no lifting of the head
similarly, croup or hip movement is used to diagnose a hind limb anymore. The sinusoidal pattern with two cycles per stride then
lameness. Sound horses at a trot show a perfect sinusoidal pattern changes to show a single cycle during each stride (Girtler & Floss,
for all midline body locations including the head, withers and 1984). Vertical velocity of the head changes in accordance with the
croup (Girtler & Floss, 1984; Buchner et al., 1996b). During one vertical movement. Both minimal and maximal vertical velocity
stride, two symmetric waves can be seen (Fig. 9.5A), which occur during the lame stance phase decrease. During the sound stance
almost synchronously in all three body parts. The height of these phase both values increase, which also results in a more positive
structures falls from the beginning of the diagonal stance phase velocity of the head at the start of the subsequent lame stance phase.
reaching the lowest position at midstance, then rising to their This means that, at impact, the head shows less downward move-
highest level at or shortly after the end of the stance phase. During ment with increasing lameness. During severe forelimb lameness,
the suspension phase, the body starts to fall again into the next the head may even be lifted slightly at the instant of impact of the
diagonal limb stance phase. These sinusoidal cycles are repeated lame limb. Finally, vertical acceleration of the head also changes
twice in each stride. The derivatives of this vertical movement, verti- from a symmetrical pattern to an asymmetrical pattern (Fig. 9.5C).
cal velocity and vertical acceleration of the head or withers, show For a quantitative analysis of the lameness, the acceleration values
similar sinusoidal patterns, but shifted by 12.5% per derivative to are even more useful than the head movement pattern, since vertical
the left (Fig. 9.5B, C). Vertical velocity reaches minimal values acceleration is less sensitive to changes in absolute head height.
shortly after the beginning of the stance phase and maximal values Furthermore, changes in the acceleration peaks quite accurately
shortly before the end of stance. Acceleration is maximal in the represent changes in the vertical forces acting on the limbs, since
middle of the stance phase and minimal during the suspension forces (F) are determined by the mass (m) of a body and its vertical
phase when it falls to its minimum value which is equal to gravita- acceleration (a): F = ma. Therefore, reduced vertical acceleration of
tional acceleration. All midline body locations, head, withers and the head and trunk during the stance phase of the lame limb results
croup raise and sink simultaneously, they are in phase. in a lower vertical force, or less loading of the lame limb.
During lameness, characteristic changes in the patterns of all During forelimb lameness the withers and croup show the same
these body segments occur. The most obvious of these is the vertical vertical movement pattern and the same timing as the head, but the

180
 Kinematics of lameness

1.75 An interesting hypothesis is stated by Keegan (2005), who

describes four different patterns of head movement during forelimb
lameness characterized by different relative heights of the head
1.70 during the various instants of the stride. He proposed that different
Displacement (m)

instants of maximal pain, e.g. after landing and the first half of the
stance or before breakover, may lead to these four different patterns
1.65 of head movement. This specific head movement pattern would
lead to a more specific local lameness diagnosis than only limb and
lameness degree, but has still to be proven by the analysis of a
1.60
Lame limb Sound limb number of horses with known lameness cause.
The locomotion patterns of the head, withers and croup during
1.55 hind limb lameness are similar to those of a forelimb lameness, but
0 20 40 60 80 100 show some distinctive characteristics. The os sacrum, which lies on
the midline, shows a perfectly sinusoidal up and down motion in
A Time (% of stride)
sound horses. During lameness, it shows less lowering and lifting
1.00 during the stance phase of the lame limb, which is exactly the same
as the motion of the withers during forelimb lameness (Buchner
et al., 1996b). The tuber coxae, however, which is more laterally
0.50 placed, has an asymmetric locomotion pattern even in sound horses
(Buchner et al., 1993). The amplitude of motion, which is mea-
Velocity (m/s)

0.00 sured as the distance between lowest and highest positions of the
tuber coxae, is smaller during the stance phase of the ipsilateral hind
0.50 limb, than during the contralateral stance phase (Fig. 9.6A). Rota-
tion of the croup around the sagittal axis through the hip joint
-1.00 Lame limb Sound limb causes different displacements in the ipsilateral and contralateral
tuber coxae. The sum of the rotational movement and the vertical
-1.50 translational movement of the whole trunk causes this typical asym-
0 20 40 60 80 100 metric vertical movement pattern of the tuber coxae even in sound
Time (% of stride) horses. In hind limb lameness, the asymmetry on the lame side of
B
the body increases. This means that vertical motion of the left
20 tuber coxae during left hind limb lameness is diminished or
absent during the left hind limb stance phase (Fig. 9.6B) and
increased during the right hind limb stance phase (May & Wyn-
10 Jones, 1987; Buchner et al., 1996b). These large motion amplitudes
Acceleration (m/s2)

of the tuber coxae are more easily detected by many people com-
pared with midline movements of the os sacrum. However, the
0 pattern of the os sacrum is similar to head and withers movement
and is, therefore, easier to describe.
Movement of the withers during hind limb lameness is similar
-10 Lame limb Sound limb to, but less pronounced, than os sacrum movement. Head move-
ment, on the other hand, shows a different pattern and different
timing than withers or sacrum movements. While the head move-
-20
0 20 40 60 80 100 ments are unchanged or even increased during the stance phase of
the lame hind limb, the displacement amplitude of the head
C Time (% of stride)
decreases during the stance phase of the sound hind limb. This
means that head and trunk motion is out of phase, the small verti-
Fig 9.5  Head movement pattern during different degrees of forelimb
cal amplitude of the croup during the stance of the lame hind
lameness: (A) vertical displacement, (B) vertical velocity, (C) vertical
limb is associated with a large head amplitude and vice versa. In
acceleration. The horizontal bars indicate the stance phases of the limbs.
(blue line) Lameness degree 0 (sound); (red line) lameness degree 1;
other words, in moderate-to-severe hind limb lameness the head
(green line) lameness degree 2. drops on the diagonal forelimb. This is to allow the horse to
Reprinted with permission from: Buchner, H.H.F., Savelberg, H.H.C.M., Schamhardt, H.C.,
reduce some of the load on the lame hind limb due to the
Barneveld, A., 1996. Head and trunk movement adaptations in horses with experimentally
changed moment of head and neck. This load redistribution
induced fore or hind limb lameness. Equine Vet J 28, 71–76.
mechanisms will be discussed in more detail in a subsequent
section.

Compensatory lameness
oscillations are less pronounced (Buchner et al., 1996b). This Dropping of the head as a compensating mechanism for a hind
means less lowering of the croup simultaneously to the withers and limb lameness mimics the head movements in a supporting limb
head during the lame diagonal stance and more lowering during lameness of the ipsilateral forelimb and, consequently, is described
the sound stance phase. Looking from behind the horse with mod- as sagittal compensatory forelimb lameness (Uhlir et al., 1997). A
erate forelimb lameness this resembles a slight lameness of the compensatory or false lameness is an apparent lameness in the
diagonal hind limb, a diagonal ‘compensatory’ lameness (see later opposite end of the body in a quadrupedal animal due to the kine-
in this chapter). Nevertheless, there is a large decrease in the vertical matic changes caused by pain in one limb, the primary lameness
acceleration of the trunk depending on the degree of lameness. Due (Uhlir et al., 1997; Kelmer et al., 2005). During a moderate hind
to the mass of the trunk, which accounts for about 65% of total limb lameness, the dropping of the head during the lame diagonal
body mass (Buchner et al., 1997), this causes a highly significant stance phase resembles an ipsilateral (or sagittal) forelimb lame-
decrease in limb loading. ness. A different situation can be seen during moderate forelimb

181
 9 Gait adaptation in lameness

1.65 1.50
Vertical displacement (m)

Vertical displacement (m)

1.60 1.45

1.55 1.40

1.50 1.35

1.45 1.30

1.40 1.25
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Time (s) Time (s)

20 20
Vertical acceleration (m/s2)

Vertical acceleration (m/s2)

10 10

0 0

–10 –10

–20 –20
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Time (s) Time (s)

Stance Stance Left hindlimb Stance Stance

Stance Stance Right hindlimb Stance Stance

Fig 9.6  Vertical displacement and vertical acceleration of the right tuber coxae in a sound horse (A) and in a horse with a moderate lameness of the right
hind limb (B) at the trot. The horizontal bars indicate the stance phases of the hind limbs.
Reprinted with permission from: Buchner, F., Kastner, J., Girtler D., Knezevic, P.F., 1993. Quantification of hind limb lameness in the horse. Acta Anat 146, 196–199, S Karger AG.

lameness. In that case the croup movements change simultaneously in this limb did not cause pain, but the limb showed hardly any
with the head movements, dropping of both head and croup, and flexion or extension during the entire stride. The author has observed
resemble a diagonal hind limb lameness. A compensatory lameness a similar locomotion pattern in horses with ankylosis of the stifle
can be differentiated from a real or true lameness by diagnostic or talocrural joint. In all cases, the affected limb could support the
anesthesias. For example, a positive diagnostic anesthesia in the body without pain, but also without any flexion or extension during
lame hind limb reduces the sagittal compensatory forelimb lame- the stride.
ness too. A true forelimb lameness, on the other hand, will not be Much more common and of clinical significance is the mixed
significantly changed by the hind limb anesthesia. lameness, where pain or pain reactions are obvious during both
stance and swing phases. Typical examples are shoulder and carpal
Summary lameness or stifle and tarsal lameness. The most extensive informa-
tion is available for carpal lameness. Several clinical studies in
The pattern of head and croup movements in lame horses allows patients (Ratzlaff et al., 1982; Ratzlaff & Grant, 1986; Clayton,
localization of the lame limb and assessment of the degree of lame- 1986a, 1987a, b) as well as studies using lameness models (Morris
ness. The asymmetric movements of head and trunk enable signifi- & Seeherman, 1987; Back et al., 1993; Peloso et al., 1993) investi-
cant unloading of the painful limb. Based on a knowledge of these gated the locomotion pattern of horses with carpal joint problems.
principles and the interactions between trunk and head movement In these studies the various symptoms can be differentiated into
patterns, an accurate quantification of the lameness using several signs that are typical of a supporting limb lameness and other
mathematical methods are possible which will be discussed later in signs that are typical of swing phase problems. The temporal vari-
the chapter. ables do not show a consistent pattern in the different studies.
Clayton (1986a, 1987a, b) studied three horses, suffering from a
shoulder, carpal and tarsal problems. She found in all three
Swinging limb and mixed lameness cases shorter stance phase durations in the lame limb compared
Pure swinging limb lameness, which is caused by pain during the to the sound limb. Ratzlaff et al. (1982), focusing on carpal lame-
swing phase without pain during the stance phase, is extremely rare. ness, found variable results regarding stance duration and suggested
The two major reasons for a pure swinging limb lameness are a that a shortened stance duration occurred with a predominance of
mechanical problem, such as an advanced ankylosis of a limb joint, the supporting limb lameness component, while a lengthened
or a neural problem such as paresis of the radial nerve. Adrian et al. stance duration indicated a predominance of the swinging limb
(1977) published goniograms of a 9-year-old stallion with an lameness. Of course, the variability of the patient histories, even in
osseous ankylosis of the metacarpophalangeal joint. Flexion testing horses suffering from carpal lameness, precludes making general

182
 Kinetics of lameness

conclusions. Three studies provoked carpal lameness using toxins phase is indicative of the swinging limb lameness component,
(Back et al., 1993), antibiotics (Peloso et al., 1993) or surgical while reduced fetlock joint hyperextension during the stance is
manipulation (Morris & Seeherman, 1987). If trotting velocity was indicative for the supporting limb lameness component.
checked to be constant, no significant changes were found in the
stance phase duration between the sound and lame limbs, or
between the same limb before and after lameness induction (Morris Bilateral lameness
& Seeherman, 1987; Peloso et al., 1993). Obviously, individual left/ A special situation and a challenging task for a veterinarian is the
right differences play the major role compared with the influence presence of a bilateral lameness. A number of orthopedic diseases,
of a mixed lameness. In most cases, a high degree of symmetry is such as navicular disease, distal limb arthrosis and bone spavin,
maintained and the right/left differences in temporal variables do may be present in both limbs to a more or less similar extent. The
not help to diagnose a specific lameness. presence and degree of bilateral problems are often difficult to
The characteristics of the caudal and cranial phases of the stride assess, because of the lack of the typical asymmetric pattern associ-
might be a useful criterion to distinguish between swinging and ated with unilateral lameness (Seeherman, 1991). The locomotion
supporting limb lameness (Wittmann, 1931; Hajer et al., 1988). In of horses suffering from bilateral lameness has been described as
swinging limb lameness the cranial phase is said to be shorter in stiff, short, or shuffling (Stashak, 1987). Such characteristics are
the lame limb due to pain during the swing phase and its effect on difficult to distinguish from the normal locomotor pattern of the
protraction. Indeed, Clayton (1986a, 1987b) found shorter cranial individual horse, which is usually unknown to the veterinarian. In
phases in shoulder and tarsal lameness, but in a horse with carpal this situation, local diagnostic anesthesia is used to detect locomo-
lameness (Clayton, 1987a) both phases were equal. Therefore, tor asymmetries by eliminating pain in one limb and thus facilitate
reduction of the cranial phase is indicative of some swing phase the detection of pain in the contralateral limb. However, this diag-
problems, but lack of such a reduction does not exclude a swinging nostic method is complicated by gait changes seen also in sound
limb lameness. horses with desensitized limbs. In a study on the effects of local
Equivocal results were found regarding the joint angle pattern in anesthesia in sound horses, they showed changes in the locomotion
horses with mixed lameness. Looking at the different limb joints, a pattern too, perhaps due to the changed proprioception in the
very clear differentiation between swing phase and stance phase anesthetized limb (Kübber et al., 1994). Therefore, in mild cases it
problems can be made. During the swing phase, flexion of the is difficult to decide whether a change in locomotion pattern is due
affected joint (carpal, tarsal) was decreased in all studies. The restric- to bilateral lameness, or just to the lost sensitivity in one limb.
tion can be seen in both the maximal flexion and the total range of Quantitative locomotion analysis in bilaterally lame horses
motion. The limb gives the impression of moving stiffly with a suffers from the same problems as the clinical assessment. If the
lower flight arc of the hoof. This is most obvious during carpal horse does not show left-to-right asymmetries and if there are no
lameness (Back et al., 1993), but is also found in tarsal (Clayton, individual control data for the horse when it was sound, it is impos-
1987b; Khumsap et al., 2004) and shoulder lameness (Clayton, sible to make a diagnosis of lameness (Hugelshofer, 1982). The use
1986a). This feature is clearly different from a supporting limb of a bilateral hoof lameness model showed some gait adaptations
lameness, where the carpal joint angle during the swing phase was that allowed the horses to reduce their discomfort, even when the
unchanged, and the tarsal joint angle was even increased (Buchner usual transfer of weight to the contralateral limb was impossible
et al., 1996a). Therefore, the presence and degree of left/right dif- (Buchner et al., 1995b). The adaptations were similar to those seen
ferences in flexion of a carpal or tarsal joint seem to be useful in unilateral supporting limb lameness: stride duration was reduced,
indicators for the swinging limb component of a mixed lameness, while relative stance duration increased. The most distinct change
as Back et al. (1993) proposed. The decreased flexion of a painful in temporal co-ordination, however, was in the diagonal advanced
joint results from the horse’s efforts to avoid painful positions. This placement. During slight bilateral lameness, horses placed their
effort can be detected even earlier in the stride cycle at the start of forelimbs earlier than the hind limbs compared with the sound
the swing phase; when flexion of the carpus begins, the angular situation. Again this variable proved to be a reliable indicator for
velocity of this joint is reduced (Ratzlaff et al., 1982). unilateral or bilateral locomotor problems.
The characteristics of the supporting limb lameness component A second variable that is indicative of bilateral pain is fetlock
of the mixed lameness have already been described. In general, all hyperextension, which was reduced equally in both forelimbs
variables typical for a supporting limb lameness can also be seen during an induced bilateral lameness. By changing the temporal
in mixed lameness (Clayton, 1986a; Peloso et al., 1993). The most stride pattern and slightly reducing the vertical displacement of the
significant features are that fetlock hyperextension in the lame limb trunk, the horse achieved the same amount of load reduction (and
is reduced as sign of less loading, which is achieved by reduced head pain relief) as seen in unilateral lameness.
and trunk displacement. The contributions of the supporting and
swinging limb lameness components to the individual lameness Summary
pattern are quite variable depending on the individual condition.
The combined angle-angle graph of both joints illustrates the Horses are able to relieve the pain in bilateral lameness even when
changes during various lameness degrees very clearly. Fetlock hyper- the typical asymmetric, contralateral compensation is not possible.
extension and carpal flexion give useful information about both Adjustments in temporal stride patterns and vertical trunk move-
aspects of the lameness, but not necessarily about its location. ment reduce the peak forces of all limbs. Diagnosis of bilateral
Though the proximal limb joints are more at risk for swinging limb lameness remains difficult, especially in mild cases. Repeated mea-
lameness, distal limb problems more often show supporting limb surements using diagnostic anesthesia or by screening the horses
lameness. The coupling of different joints, such as the tarsal and over longer periods can be used to monitor fetlock joint angle and
stifle joints, impedes the ability to localize the disease. Furthermore, diagonal advanced placement to detect mild gait deficits.
slow protraction due to joint pain causes similar decreases in the
range of joint motion in the coupled joint regardless of whether the
pain is in this joint or the neighboring joint. Kinetics of lameness

If horses feel pain in a limb when moving, they adjust their locomo-
Summary tion pattern to diminish the pain. The various kinematic changes
Carpal and tarsal or stifle lameness are typical examples of mixed described in the previous sections are components of the overall
lameness. Reduced flexion of a proximal joint during the swing effort to achieve pain reduction. In a supporting limb lameness,

183
 9 Gait adaptation in lameness

pain reduction means reducing the load on the affected limb. The 1988a) as well as at the trot (Morris & Seeherman, 1987; Clayton
success of the efforts of the load-reducing strategy can be assessed et al., 2000a; Weishaupt et al., 2004, 2006). At the walk the peak
by measuring the ground reaction forces (GRFs) using a force plate vertical force decreases, while the GRF pattern becomes smoother:
(Gingerich et al., 1979; Morris & Seeherman, 1987; Merkens & the dip between the two vertical force peaks diminishes in mild
Schamhardt, 1988a, b; Weishaupt et al., 2004, 2006), force shoes lameness and disappears in moderate lameness at the walk (Fig.
(Hugelshofer, 1982) or the EGA-system (Tietje, 1992). A unique 9.7) (Merkens & Schamhardt, 1988a). At the trot, the shapes of the
technology to record continuously the vertical ground reaction vertical and longitudinal GRF force curves in the lame limb resem-
forces of horses is the instrumented treadmill developed in Zürich/ ble those of sound horses, however with characteristic decreases in
Switzerland, which allows to measure continuously vertical forces the force amplitudes. Generally, with increasing lameness, horses
of all four limbs independently during treadmill locomotion show a reduced impact force, a reduced rate of loading during the
(Weishaupt et al., 2002). A synchronous view of kinetic and kine- first half os stance and a reduced peak vertical force (Fzmax).
matic findings provides a more complete picture of how horses (Clayton et al., 2000a; Weishaupt et al., 2006). Already in only a
manage a lameness and gives a better understanding of the mecha- subtle lameness, Fzmax is significantly reduced by 4%. This reduc-
nisms that redistribute the load away from the lame limb without tion reached 9% in mild lameness and 24% in moderate lameness
serious overloading of the sound limbs. (Weishaupt et al., 2006). Similar results are reported by Morris and
Seeherman (1987) which measured 11.5% decrease of Fzmax in a
easy observable carpal lameness, or Clayton et al. (2000a) with
Load reduction in the lame limb 27% less Fzmax, in horses with an induced tendinitis. However,
As lameness increases, horses show progressive reductions in their these horses, showing moderate to severe lameness also reduced
vertical and longitudinal deceleration forces. These differences can their running velocity by 10%, which contributes to the overall
be seen in the GRF tracings at the walk (Merkens & Schamhardt, load reduction.
Mediolateral force (N/kg)

Mediolateral force (N/kg)

100 100
Forelimb Hindlimb
0.5 0.5

0 0
-0.5 -0.5

-1.0 -1.0
0 20 40 60 80 100 0 20 40 60 80 100
Stance time (%) Stance time (%)

1.5 1.5
Longitudinal force (N/kg)

Longitudinal force (N/kg)

1.0 1.0

0 0

0 0

0.5 0.5
-1.0 -1.0

-1.5 -1.5
0 20 40 60 80 100 0 20 40 60 80 100
Stance time (%) Stance time (%)

8 8
Vertical force (N/kg)

Vertical force (N/kg)

6 6

4 4

2 2

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Stance time (%) Stance time (%)

Fig 9.7  Ground reaction force patterns of left fore limb (left panel) and left hind limb (right panel) at the walk during different degrees of left forelimb
lameness. The forces are expressed in N/kg body mass during a standardized stance time. Control session (blue line); only lame at the trot (red line); mild
lameness at the walk (green line); moderate lameness at the walk (dark green line); digital nerve block (orange line).
Reprinted with permission from: Merkens, H.W., Schamhardt, H.C., 1988a. Evaluation of equine locomotion during different degrees of experimentally induced lameness. I. Lameness model
and quantification of ground reaction force patterns of the limbs. Equine Vet. J. (Suppl.) 6, 99–106.

184
 Kinetics of lameness

Similar reduction of loading can be found in the braking longi- differences in the vertical force of nearly 500 N and differences of
tudinal force, where several local peaks of the longitudinal force are about 230 Nm in the sagittal torque acting on the trunk. Therefore,
reduced in the lame limb. The third parameter used to describe limb the head is an effective tool for redistributing the load from the
loading is the vertical impulse (Iz) of the limb, which represents the painful limb to the other limbs. These dynamic influences of the
time integrated force during one stance phase. While the vertical head and neck movements should not be confused with the rela-
force decreases continuously with increasing lameness, the vertical tively small static influences on the position of the body center of
impulse is less affected. The increase in relative stance duration is mass (BCM) due to different head positions. If the head and neck
responsible for this smaller increase in vertical impulse. Compared move caudally by 10 cm, the BCM shifts caudally by only 1 cm. For
to the 24% decrease of Fzmax in moderate lameness, Iz decreased only a 500-kg horse, assuming a distance of 100 cm between fore and
by 20% (Weishaupt et al., 2006). This difference between Fzmax and hind limbs, this means a transfer of only 50 N from fore to hind
Iz is an important feature for the lameness management of horses limb. This assumption is substantiated by Buchner et al. (2001)
who try to maintain their normal locomotion pattern as much as finding a caudal translation of the BCM during midstance of the
possible together with maximal pain reduction, which will be dis- lame limb in a grade 2 lamenesss of only 9 mm. Therefore, the
cussed in the next paragraph. If lameness is more severe, the load dynamic influences of the asymmetric head and neck movement in
is reduced progressively to zero load in a non-weightbearing lame horses have about 10 times more effect than the static
lameness. component.
In lame horses, the fetlock joint angle time diagram (Back et al.,
1993; Buchner et al., 1996a) resembles the vertical force pattern
shown by Morris and Seeherman (1987) or Clayton et al. (2000a). Summary
The reduction of the fetlock hyperextension of about 8° in a grade Summing up all details, the body receives less vertical impulse
2 lameness might be quite comparable to the reduction of fetlock during the stance phase of the lame limb despite the small prolon-
extension by about 10% that was reported in horses with an easily gation of the stance duration. This smaller vertical impulse leads to
observable lameness by Morris and Seeherman (1987). Similarly, a distinct reduction in the duration of the subsequent suspension
Clayton et al. (2000a) found a decrease in fetlock hyperextension phase and the contralateral limb has to cope with the need for
of 11° corresponding to the decrease in the Fzmax of 27%. a higher vertical impulse associated with the descending trunk
and head.
Summary
Maximal vertical force, assumed to be responsible for the pain
during loading, is effectively reduced in the lame limb. Vertical Redistribution of the load
impulse is reduced too, but due to an increased stance duration less
than maximal vertical force. The reduction of the peak load in the lame limb is not accompanied
by equivalent increases of peak loads in the sound limbs. This
remarkable fact has been observed both in kinematic studies that
evaluated the fetlock joint pattern of all limbs and in kinetic studies
Mechanisms of load reduction of the vertical loads on all four limbs. Using the maximal hyperex-
Based on the kinematic analyses, the reduced loading in a lame tension of the fetlock joint as a load indicator, at the trot a decrease
forelimb can be explained by two major mechanisms: 1) a smoother of 7.6° in the lame limb was accompanied by an increase of only
loading of the lame limb and 2) decreased vertical movements of 1.9° in the contralateral forelimb, no change in the diagonal hind
the head and trunk. limb and a 2.6° decrease in the ipsilateral hind limb (Buchner et al.,
The loading is smoothed by small adjustments in several aspects 1996a). Simultaneous kinetic recordings of the vertical forces on
of limb timing. The swing duration of the forelimbs is reduced, the instrumented treadmill allowed for the quantification of the
while the relative stance duration is increased. The longer ground effective load changes in all four limbs with increasing fore or hind
contact distributes the effort required to lift the body over a longer limb lameness (Weishaupt et al., 2004, 2006). As already seen in
period. Furthermore, the earlier placement of the lame forelimb, the kinematic analyses, the contralateral limb does not get a higher
seen in a changed diagonal advanced placement of the lame limb peak load at the trot. Fzmax of the contralateral limb was nearly
(Buchner et al., 1995a; Weishaupt et al., 2006), allows it to accept unchanged during subtle or moderate lameness. Similarly, peak
the body load at a time when the trunk is at a higher point in its load in the diagonal hind limb was nearly constant, and it was only
sinusoidal motion cycle and has a relatively low downward velocity. during moderate lameness a slight increase of 2% could be found.
This reduces the effort required to lift the body as discussed in the The ipsilateral limb shows more changes, but it is loaded less rather
second mechanism. Additionally, braking of the descending body than more (6% less in moderate lameness). In fact, hardly any
is smoothed by slightly more flexion of the shoulder joint. compensatory increase of peak load can be found in the nonlame
The second mechanism to reduce peak load involves the vertical limbs. This is true also for hind limb lameness, which is even more
displacement of the body. During the stance phase of the lame limb stable: while peak load in the lame limb gradually decreases with
the head and neck and, to a lesser extent, the trunk, are not lowered increasing lameness, no change in Fzmax could be found in the three
as normal. Keeping the body at a more constant height needs less other limbs (Weishaupt et al., 2004).
maximal vertical acceleration and results in a decreased peak verti- The same relationships were found by Morris and Seeherman
cal force in the lame limb. Additionally horses adjust their head and (1987) and Tietje (1992).
neck movement to control the load distribution in all limbs. Since Analysis of the characteristic head and trunk movements com-
the head and neck are heavy, representing about 10% of total body bined with model calculations (Vorstenbosch et al., 1997) offer an
mass (Buchner et al., 1997), and they have a long lever arm relative explanation for the way horses redistribute the body load. The
to the body center of mass, their position has a relatively large effect reduction of maximal head acceleration during the lame diagonal
in loading or unloading the forelimbs. This influence of head and stance phase reduces the load on the lame forelimb, but increases
neck movement has been evaluated using an inverse dynamics the load on the diagonal hind limb due to a reduced torque at the
model (Vorstenbosch et al., 1997). The dynamic forces acting on neck-trunk connection. At the same time, reduced acceleration of
the trunk and then on the limbs were calculated from the kinemat- the whole trunk adds to the reduction of loading of the lame fore-
ics of head and neck, their inertial properties and the geometric limb, but counteracts and equalizes the torque effects of the head
properties. Differences of only 10 cm in vertical amplitude of the on the diagonal hind limb. During the sound diagonal stance
head during the stance phases of lame and sound limbs caused phase, the trunk acceleration hardly changes. Increased head

185
 9 Gait adaptation in lameness

movements, however, slightly increase the loading of the contralat- These results show that a lameness in one limb will not neces-
eral forelimb and decrease that of the ipsilateral hind limb. sarily increase the risk of damaging the contralateral limb or the
The fact that the reduction of the peak load in the lame limb is other sound limbs. If high peak forces are responsible for compen-
not compensated by higher peak loads in the other limbs raises the satory injuries, the mechanisms described above enable a maximal
question as to where the load is going. Two phenomena are involved decrease of peak force in the lame limb together with a minimal
in answering this question: increase of risk in the sound limbs.
1. A redistribution of the load within the limbs
2. A decrease of total load of all four limbs during the whole Summary
stride cycle. There is no redistribution of peak load from the lame to the sound
limbs and only a little redistribution in vertical impulse. A lower
The redistribution within the limb is based on an earlier placement
total stride impulse is compensated by a higher stride frequency.
of the limb, the prolonged stance duration and the increased
damping in the proximal limb joints. These movement adaptations
smooth the loading of the lame limb and enable the horse to reduce Specific lameness
the peak load significantly more than the vertical impulse. As
already mentioned for a moderate forelimb lameness, peak vertical
load reduction was 24% compared to a reduction of vertical impulse The most ambitious aim in the use of locomotion analysis for the
of only 20%. Similar results were found for a carpal lameness by diagnosis of lameness in horses is the exact localization of the
Morris and Seeherman (1987) whose results showed 11.5% reduc- ailment within a limb thus making the diagnosis of a specific
tion in peak force, but only 8.4% reduction in impulse in the lame disease (Leach & Crawford, 1983; Clayton, 1986a). A precondition
limb. In severe induced tendonitis, a decrease of 27% in Fzmax cor- for such a specific diagnosis is a database that includes kinematic
responded to 15% reduction in the vertical impulse (Clayton et al., patterns of various specific diseases, based on sufficient measure-
2000a) (Fig. 9.8). The same principle works in the contralateral ments to eliminate the variation due to individual locomotion
limb. More or less unchanged peak loads during a prolonged stance patterns. Recordings have been made in small groups of horses with
phase led to increases in vertical impulse in the contralateral limb specific ailments, and studies have been performed in groups of
of 7% (Weishaupt et al., 2006) or 4.7% (Morris & Seeherman, horses with equal, induced lameness. Analysis of the results of these
1987). studies showed two major problems that impede formulation of a
The remaining difference in impulses between lame and sound definition of characteristic lameness patterns in specific lameness:
limbs can be explained by the second phenomenon, the reduction • The individual locomotion pattern: Several studies found a
of the sum of the load of all four limbs, the total load or total stride high level of reproducible individuality in the locomotion
impulse. This has already been suggested by Morris and Seeherman pattern of horses (Drevemo et al., 1980; van Weeren et al.,
(1987) and is later confirmed by results of kinematic and kinetic 1993) seen as a low intraindividual variation compared to
studies. When forced to keep a constant velocity on the treadmill, the high interindividual variation. The low intraindividual
horses increase their stride frequency by reducing the duration of variation allows for repeated, reliable assessments of
the swing phase. In this way the horses reduce the load within one the locomotion pattern after diagnostic or therapeutic
stride by distributing it over a larger number of strides. manipulations. The accurate measurement techniques enable
the detection of really small, but significant, differences
due to the various kinds of lameness. However, these
differences are often smaller than differences between horses
due to their individual locomotion pattern.
180 • The wide variety of orthopedic diseases: Many horses suffer
from more than one disease. Even horses suffering from very
Sound
similar diseases or syndromes, such as navicular disease or
160 Lameness left forelimb carpal lameness, show a variety of orthopedic and radiological
findings. In two studies analyzing seven horses with carpal
Vertical impulse (G msec)

lameness syndrome seven different specific diagnoses were

140 reported (Ratzlaff et al., 1982; Clayton, 1986a). Navicular
disease seems to be more uniform, but additional ailments
can occur. In the thorough study of Girtler (1988a,b,c) six
120 out of seven horses with navicular disease had additional
problems, such as arthrosis, tendinitis, sesamoiditis or bone
spavin. This problem might have existed, though it was not
100 described, in other studies of horses with navicular disease
by Pollhammer-Zeilinger (1996), Hütter (1997) and Keegan
et al. (1997). Therefore, it is difficult to use horses with
80
LF RF LH RH naturally occurring lameness to extract specific patterns for a
specific ailment. General conclusions can be drawn only by
Limb
analysis of a large number of horses suffering from more or
Fig 9.8  Redistribution of vertical impulse from the lame forelimb to the less the same disease. Then general symptoms have to be
other limbs during induced superficial digital flexor tendinitis in one separated from specific symptoms. Large numbers of horses
forelimb. have been studied for only a few diseases and in these
Data from Clayton, H.M., Schamhardt, H.C., Willemen, M.A., Lanovaz, J.L., Colborne, G.R., diseases the general symptoms dominate over specific details.
2000a. Kinematics and ground reaction forces in horses with superficial digital flexor
tendinitis. Am J Vet Res 61, 191–196. ©American Veterinary Medical Association.
Reprinted with permission from: Merkens H.W., Schamhardt H.C., 1988a. Evaluation of
Navicular disease
equine locomotion during different degrees of experimentally induced lameness. I. Lameness There are five studies reporting kinematic data from horses with
model and quantification of ground reaction force patterns of the limbs. Equine Vet J navicular disease (Ratzlaff & Grant, 1986; Girtler, 1988a,b,c;
(Suppl.) 6, 99–106. Pollhammer-Zeilinger, 1996; Keegan et al., 1997; Hütter, 1997).

186
 Specific lameness

Comparing their results with those from other lameness studies, the SDF tendonitis could be found especially at the beginning of the
same general characteristics as described for supporting limb lame- stance phase. These very promising results for a specific lameness
ness can be found. There is a tendency to longer stance durations, diagnosis, however, are based on a quite sophisticated analytic
but this feature is not shown by all horses (Pollhammer-Zeilinger, method that has not been pursued.
1996). Fetlock joint hyperextension is reduced and vertical head
movement shows some of its characteristic asymmetric pattern,
which depends on the degree of asymmetry of the pathologic Laminitis
process in both limbs. Even hoof landing, as described by Ratzlaff A similar inverse dynamic study as for navicular disease was per-
and Grant (1986), does not show a consistent pattern in all horses formed in ponies with chronic laminitis (McGuigan et al., 2005).
(Hütter, 1997). Keegan et al. (1997) found also a reduced maximal These ponies all had rotations of the distal phalanx of 6–13° and
flexion in the carpal joint during the swing phase, which he explained a stilted gait at the trot. Ground reaction forces (GRF) showed
as a general symptom of a reduced loading followed by a less ener- similar typical signs of bilateral lameness, with lower Fzmax and
getic protraction of the limb. Therefore, no really specific kinematic a prolonged stance duration. The most interesting detail was
patterns for navicular disease could be found to differentiate navicu- the change in the moment arm of the GRF at the coffin joint. The
lar disease from other types of supporting limb lameness. moment arm was below zero until about 40% of the stance phase,
However, locomotion analysis studies using the advanced meth- and in the later part of stance was still significantly shorter than the
odology of inverse dynamic analysis, which combines kinetic and moment arm in sound ponies. As a consequence of this situation
kinematic measurements, offered additional insights into the spe- there was no tension in the DDFT until nearly midstance. In the
cific locomotion pattern of horses suffering from navicular disease second half of the stance phase, DDFT force gradually increased,
(or low palmar foot pain) (Wilson et al., 2001; McGuigan & Wilson, but reached only 64% of the peak values in sound ponies at heel-off
2001). These studies calculated the moment arm of the coffin joint (Fig. 9.10). This shows nicely how displacement of the distal
and the compressive force of the deep digital flexor tendon (DDFT) phalanx reduces tension in the DDFT, which means that rotation
on the navicular bone. The most surprising and significant finding of the distal phalanx during the laminitis is a self-limiting process.
was an increase in compressive force shortly after landing and When the rotation reaches a critical amount around 15–20°, there
during the whole stance phase in horses suffering from navicular is no tension in the DDFT and therefore no rotational force on the
pain. Sound horses start with a low DDFT force and a low navicular distal phalanx.
compressive force, which increases during the stance phase reaching
a peak shortly before heel-off. The lame horses with navicular pain
show a paradoxical increase in this force, which is expected to
increase their pain too (Fig. 9.9). Obviously, while they try to
Carpal and tarsal lameness
decrease heel loading by tension in the DDFT, they increase pressure Carpal lameness is the most intensively studied orthopedic syn-
on the navicular bone and thereby even contribute to a progression drome using locomotion analysis techniques. Both patient studies
of the disease. This obvious inability of the horse to unload the (Ratzlaff et al., 1982; Clayton, 1987a) as well as studies of induced,
diseased structure might explain the bad prognosis of navicular uniform carpal lameness (Auer et al., 1980; Morris & Seeherman,
disease and also the positive effects of mechanical support offered 1987; Back et al., 1993; Peloso et al., 1993) serve as a database for
by orthopedic shoeing, such as raised heels, facilitation of break a detailed analysis. As in navicular disease, there are reliable general
over, or protection from overextension of the coffin joint by signs of lameness, including reduced fetlock hyperextension and
extended heels (egg bar shoes). a head movement pattern that are indicative of the supporting
Another interesting approach was described by Williams et al. limb lameness component. Reduced carpal flexion is a consistent
(1999) using the pattern of the GRF. When they performed a prin- sign of the swinging limb lameness component. Ratzlaff et al.
cipal component analysis of the vertical and longitudinal forces (1982) also described changes in the angular velocity as a good
specific differences between sound horses and horses suffering from indicator to differentiate swinging and supporting limb lameness.

7 12

6 10
Compressive stress on the

5
navicular bone (MPa)

DDFT force (N/kg)

8
4
6
3
4
2
2
1
0
0 0 20 40 60 80 100
0 20 40 60 80 100 Time (% of stance)
Percentage of stance
Fig 9.10  Mean force in the deep digital flexor tendon (N/kg body mass)
Fig 9.9  Compressive stress on the navicular bone by the DDFT during trot during the stance phase in normal horses (blue lines) and horses with
in normal horses (red lines) and horses with navicular disease (blue lines). navicular disease (red lines). Values are mean (solid lines) and SD (broken
Values are mean (solid lines) and SD (broken lines). lines).
Reprinted with permission from: Wilson A.M., McGuigan M.P., Fouracre L., et al., 2001. The Reprinted with permission from: McGuigan M.P., Walsh T.C., Pardoe C.H., et al., 2005. Deep
force and contact stress on the navicular bone during trot locomotion in sound horses and digital flexor tendon force and digital mechanics in normal ponies and ponies with rotation
horses with navicular disease. Equine Vet J 33, 159–165. of the distal phalanx as a sequel to laminitis. Equine Vet J 37 (2), 161–165.

187
 9 Gait adaptation in lameness

Clayton (1987a) described a specific detail in the vertical movement the population. Clayton et al. (2000a) also found a significant dif-
pattern of the carpal region in a horse with a fracture of the third ference in the slope of the vertical ground reaction force that was
carpal bone. This involved lowering of the lame limb at lift-off and indicative of a greatly reduced loading rate in the lame limb and a
during the early swing phase. This pattern is visible also in the horse more rapid loading rate in the compensating limb. It coincided with
recorded by Ratzlaff and Grant (1986) as well as in one horse of a divergence between the angles of the coffin and fetlock in the two
Girtler (1988b). Another horse in the study of Girtler (1988b), that limbs as a consequence of the differences in loading. However, the
was suffering from tendinitis, showed the same carpal pattern, large variations in the force curves between and within horses are
which might point to a common phenomenon of pain during likely to preclude their use as a diagnostic tool.
maximal extension of the carpal joint and the wish to flex and
unload the lame limb as early as possible.
Several studies analyzed the gait in induced arthritis of the distal Ataxia
tarsal joints (Clayton, 1987b; Kramer et al., 2000; Khumsap et al., Several attempts were made during recent years to find an objective
2003; Khumsap et al., 2004). Osteoarthritis of the distal tarsal joints documentation for ataxic movements, which are symptoms of neu-
is known as bone spavin, a very common degenerative joint disease rological disturbances in horses. The diagnosis of this gait pattern
in the horse. Like carpal lameness, characteristics of both supporting is based primarily on subjective gait changes, like in coordination,
limb lameness as well as swinging limb lameness were found. weakness or dysmetria. Clayton et al. (2003) and Bialski et al.
Typical load reduction, seen as reduced vertical force and impulse (2004) introduced the force plate as a tool to assess postural sway
or reduced fetlock joint hyperextension was measured. Tarsal flexion as an indicator of neurological disease. The horses stood with both
was reduced during midstance as well as midswing. Furthermore, a fore or both hind feet on the force plate, without and with a blind-
net power analysis revealed reduced negative joint work during early fold and movements of the center of pressure were tracked over a
stance and reduced positive work at push-off at the tarsal joint period of time. Variables that were measured include the radius and
(Khumsap et al., 2003). Another study by Khumsap et al. (2004) velocity of the movements of the center of pressure, and the cranio-
offered even more insight reporting from three-dimensional joint caudal and mediolateral range of motion. The findings are plotted
movement pattern in tarsal synovitis. These results pointed to move- graphically as a stabilogram (Fig. 9.11). Increases in the measured
ments like proximodistal or craniocaudal translation of the meta- variables have been detected in horses with various neurological
tarsus, which, while coupled to tarsal flexion and extension to a diseases including cervical vertebral stenosis (wobblers), equine
major part, show some additional movements. protozoal myelitis and vestibular disease. The agreement boundary
of the center of pressure movement was described to be a sensitive
parameter to distinguish between sound horses or those having
Tendinitis balance deficiencies. Keegan et al. (2004) used the kinematics of 21
markers on the trunk and limbs together with a sophisticated math-
An integrated study of the kinematics and GRFs before and after the ematical procedure using a fuzzy clustering to distinguish between
induction of superficial digital flexor (SDF) tendinitis in one fore- sound and spinal ataxic horses. They found the movement
limb (Clayton et al., 2000a) showed typical signs of a supporting pattern of one body marker on the back together with one marker
limb lameness, e.g. a lower peak vertical GRF in the lame limb each at on a fore and hind limb sufficient to achieve a correct clas-
together with less flexion of the coffin joint and less hyperextension sification of the horses. Another kinematic method was used by
of the fetlock joint in the lame limb during midstance compared Strobach et al. (2006), which used the autocorrelation and cross
with the compensating limb. Carpal joint kinematics did not correlation of hoof movement successfully to evaluate the consis-
change. At the lame evaluation the compensating limb had a more tency of the gait.
protracted orientation throughout its stance phase, though its total
range of limb rotation from ground contact to lift-off did not
change. This facilitated a smooth transfer of body weight from the Peculiar lameness
lame to the compensating limb without the need to raise the body
mass into a suspension. In association with its more protracted There are some rare orthopedic diseases in which the locomotion
orientation, the compensating limb had a higher braking longitu- pattern is so unique that it will be recognized by everybody who
dinal force and impulse than the lame limb, while the propulsive
components of the longitudinal ground reaction force did not differ
between limbs. The center of pressure began to move rapidly toward
the toe relatively early in the stance phase in the lame limb, which 0.4
was interpreted as a consequence of the lower GRFs in the lame 0.3 Neurological
Craniocaudal movement (cm)

limb. The lame limb also showed significant reductions in the peak Normal
values for the net joint moments on the palmar aspect of the fetlock, 0.2
carpal and elbow joints, which are the joints crossed by the SDF
0.1
musculotendinous unit. The total mechanical energy absorbed was
significantly lower at every joint in the lame limb compared with 0.0
the compensating limb (Clayton et al., 2000b).
-0.1
In the early stance phase, oscillations in the longitudinal force
peaks correspond approximately with changes in the vertical -0.2
loading pattern of the limb, and it has been suggested (Dow et al.,
1991) that changes in the slope of the vertical ground reaction force -0.3
during phase periods corresponds to a reduced rate of loading of -0.4
the fetlock as this joint reaches its full extension. Unfortunately, -8 -6 -4 -2 0 2 4 6 8
findings from the lame and compensating limbs were not differenti-
Time (% of stride)
ated by Dow et al. (1991), but all horses whose values for the verti-
cal force slope for one or both limbs were outside the 95% Fig 9.11  Stabilogram showing craniocaudal and mediolateral movements
confidence limits of the population under study had an SDF injury. of the center of pressure of the two forelimbs in a normal horse (red line)
Clinical lameness was only apparent when the deviation of the and a horse with vestibular disease (blue line). Measurements were made
vertical force slope was outside the 99% confidence limits of over a period of 10 s at a sampling frequency of 2000 Hz.

188
 Clinical use of locomotion analysis

Fig 9.12  Typical high elevation of the hind limb during the swing phase in Fig 9.13  Typical movement pattern of a hind limb in a horse at the walk
a horse at the walk suffering from stringhalt. suffering from fibrotic myopathy of the semitendinosus muscle. The foot
jerks backward and downward just before landing.
Reprinted from Stashak, T.S., 1987. Adam’s Lameness in Horses, with permission from John
Wiley and Sons, Inc.
has seen it previously. Two of these specific lamenesses have been
described kinematically in detail.
84% of the swing phase and the hoof contacted the ground toe first
Stringhalt after being lowered almost vertically rather than being retracted
prior to contact. Consequently, the diagonal distance was shorter
This extraordinary locomotion pattern can be seen in horses often
for the affected diagonal limb pair. These differences were found
without a known underlying cause (idiopathic stringhalt), probably
even at the trot, in which the characteristics are less obvious than
due to a neurologic impairment with disturbed spinal motorneuron
at the walk, because at the walk more hind limb protraction is seen
regulation. This pattern is characterized by the exaggerated lifting
than at the slow trot (Buchner et al., 1996a).
of one or both hind limbs during the swing phase (Fig. 9.12).
Girtler (1988c) presented the flight arc and temporal data of such
a patient. The hoof was lifted to a maximum height of nearly 60 cm Summary
instead of the normal height of 15–20 cm, with the peak of the Kinematic or kinetic analyses of lameness in different ailments show
flight arc being reached quite late during the swing phase and patterns that allow for a reliable classification into supporting or
the hoof then being lowered nearly vertically to the ground. In the swinging limb lameness. More specific patterns could not be detected
affected limb, the stance phase was shortened and the swing phase using these techniques alone, probably as a consequence of the quite
was prolonged, whereas the sound limb showed the opposite uniform way horses react to pain in a limb, due to the limited degree
changes. Step duration from the sound to the lame limb was longer of freedom in their locomotion patterns. Inverse dynamic analyses
than the contralateral step. This strange locomotion pattern is much allow more specific characteristics of a lameness to be identified,
more obvious at the walk than at the trot, even at very slow veloci- including changes in functional units like the deep digital flexor
ties. Recently, both kinematic as well as surface EMG techniques tendon (including its check ligament and the navicular bone). There-
were used to document these typical gait characteristics of horses fore, specific lameness diagnosis using gait analysis techniques is still
with stringhalt and the effects of an intramuscular botox therapy a hot topic and field for further research in the future.
(Wijnberg et al., 2009).
Clinical use of locomotion analysis
Fibrotic myopathy
Like stringhalt, the typical locomotion pattern of fibrotic myopathy Accurate and objective assessments of locomotor disturbances using
is more obvious at the walk than at the trot. Fibrosis of torn muscle locomotion analysis techniques provide the veterinarian with valu-
fibers of the semitendinosus muscle impedes protraction of the able tools for advanced diagnostics in a clinical setting. Both quali-
hind limb at its most cranial position, when the muscle is maxi- tative and quantitative methods can be used to assess lameness
mally stretched. At the walk, the hind hoof jerks backward and objectively, to augment diagnostic procedures or to verify therapeu-
downward just before landing instead of a smooth forward move- tic effects. Qualitative assessments use slow motion video-recordings
ment to impact (Fig. 9.13). Clayton (1988) analyzed a typical to detect changes in hoof motion and landing as well as locomotor
example of a horse with this ailment and described some kinematic asymmetries (Seeherman, 1991, 1992). The big advantage of quali-
details for the trot. The flight arc stopped its forward movement at tative gait analysis is the low technical effort and the simplicity of

189
 9 Gait adaptation in lameness

Table 9.3  Symmetry indices using kinematic or kinetic variables

Author Index Variables Method Sound value

Morris & Seeherman (1987) L/R ratio Ground reaction forces, impulses, timing Quotient 1
Merkens et al. (1988) H(orse)INDEX Ground reaction forces, impulses, timing Quotient, factors 1
Kastner (1989) HAAS Head vertical acceleration Quotient 0
Buchner et al. (1993) HAQ Hip vertical acceleration Quotient 1.28
Kübber et al. (1994) WAAS Withers vertical acceleration Quotient 0
Peham et al. (1996) Symmetry Vertical head movement Fourier analysis 100%
Buchner et al. (1996b) Various indices Head, withers, sacrum, tuber coxae, vertical Quotient 1
displacement, maximal acceleration,
vertical acceleration amplitude
Uhlir et al. (1997) SAAS Sacral vertical acceleration Quotient 0
Pourcelot et al. (1997) Symmetry indices, forelimb, Joint displacement, joint angles Inter-correlation 1
hind limb index
Weishaupt et al. (2001) Various indices Ground reaction forces, impulses Quotient 0
Audigie et al. (2002) Energy ratio Trunk vertical displacement Fourier analysis 100%
Keegan et al. (2004) Accelerometric ratios Poll and pelvis acceleration Fourier analysis 1

interpretation of the slow motion video images by veterinarian and The HAAS index is zero in sound horses with symmetric head move-
horse owner. The same criteria as during a standard orthopedic ments during the left and right forelimb stance phases. The index
examination are assessed, but with a higher temporal resolution tends to −1 if the left limb is lame and it tends to +1 if the right
and with as many replays as necessary. limb is lame. This simple equation has also been applied to loca-
Quantitative assessments need more sophisticated methods and tions other than the head, giving rise to WAAS (Withers Acceleration
can measure all kinematic or kinetic details described in the previ- ASymmetry) (Kübber et al., 1994) or SAAS (sacral acceleration
ous sections. For clinical applications, however, the need for a rapid asymmetry) (Uhlir et al., 1997). Similarly, using slightly different
turnaround time usually restricts the analysis to localization of the calculation methods, the hip acceleration can be recorded and ana-
lame limb and quantification of the lameness degree. Several lyzed (Buchner et al., 1993) or the vertical displacements in both
methods and variables are available to define symmetry indices for hips can be compared (May & Wyn-Jones, 1987). These symmetry
lameness quantification (Table 9.3). These indices have to meet indices perform very well, but still need considerable effort during
several criteria to be of practical, diagnostic value. Firstly, they must recording and analyzing.
be sufficiently sensitive to lameness that even small disturbances in More sophisticated calculation methods were used by Peham
the locomotion pattern are reflected by distinct changes in the sym- et al. (1996) to reduce the influence of unsteady head movements.
metry index. Secondly, an index value should be indicative of a After processing the data using a system-matched filter, the sym-
certain degree of lameness, and the interindividual variation in this metry of the vertical head movement pattern was analyzed by deter-
variable should be smaller than the differences between lameness mination of the Fourier series. The symmetry of the horse’s
degrees. Thirdly, the variable should be easy to measure and inter- movements can then be calculated by comparing the values of the
pret. These criteria can be fulfilled to a variable extent by variables Fourier coefficients and presented as symmetry percentage (Table
derived from both kinetic and kinematic analyses. However, the 9.3) (Peham et al., 1995). Similar methodology was used by Keegan
third criterion is still especially difficult to achieve and is the subject et al. (2001) and Audigie et al. (2002), where Audigie et al. (2002)
of further technical development. only used a trunk marker to determine fore or hind limb lameness
as well as lameness degree.
A different kinematic method was reported by Pourcelot et al.
Kinematic lameness indicators (1997). They applied an intercorrelation method to analyze the
The typical asymmetric pattern of head, withers or croup move- contralateral symmetry of both the vertical joint motion and the
ments in lame horses enable their use for the calculation of sym- joint angle changes during one stride. The results of these calcula-
metry indices that are similar to the traditional, subjective tions are presented as kinematic indices for the comparison of each
assessment by the veterinarian. Kastner (1989) first used the vertical pair of markers or limb joints, or as an averaged fore or hind limb
accelerations of the head to calculate a symmetry index named index. In contrast to the HAAS, these symmetry indices have the
HAAS: head acceleration asymmetry. He used the following value 1 in perfectly sound and symmetric horses and tend to zero
equation: in severe lameness.
LAA − RAA
HAAS =
BAA Accelerometer
HAAS:  Head acceleration asymmetry Acceleration can be calculated by double differentiation of the dis-
LAA:  Left vertical head acceleration amplitude placement of a point, or it can be measured directly using an accel-
RAA:  Right vertical head acceleration amplitude erometer. Conversely, acceleration data can be integrated to calculate
BAA:  Bigger vertical head acceleration amplitude. velocity or displacement. Barrey et al. (1994) used accelerometers

190
 Clinical use of locomotion analysis

fixed to the sternum of horses to record the horizontal and vertical 8

movements of the trunk. In addition to providing performance H0
indices, this method can be used to assess both symmetry between
successive half strides and regularity between successive strides by 6 Left forelimb

Vertical force (N/kg)

calculating the autocorrelation function of the acceleration signal.
This allows the detection of lame horses as those showing symmetry Right forelimb
values lower than 95%, with 100% as the result in perfectly sym- 4
Left hindlimb
metrically moving horses (Barrey & Debrosse, 1996).
An accelerometer mounted to the horse’s head has been used to Right hindlimb
2
detect head motion asymmetries by Weishaupt et al. (1993). They
recorded the vertical head acceleration directly and quantified the
improvement of the gait in a lame horse due to analgesics and local 0
anesthetics, in a similar manner to Kastner (1989). Church et al., –1.5 –1.0 –0.5 0 0.5 1.0 1.5
2009 again proved the significance of the vertical movement of the Longitudinal force (N/kg)
pelvis to quantify hind limb lameness by using inertial sensors on
both tubera coxae and tuber sacrale.
All accelerometric methods have the advantage of needing little 8 8
F2 H2
instrumentation and of offering results in a short time using auto-
mated data analysis software.
6 6

Vertical force (N/kg)

Vertical force (N/kg)

Force plates, force shoes, EGA system
4 4
Pratt and O’Connor (1978) introduced the force plate in equine
locomotion research. The vertical and horizontal force tracings
allow a clinically appropriate analysis of limb loading and have 2 2
often been used to satisfy the urgent need for objective assessments
of orthopedic therapies (see next paragraph), for example in carpal
lameness or tendinitis (Gingerich et al., 1979; Goodship et al., 0 0
1983; Morris & Seeherman, 1987). Merkens et al. (1988) and –1.5 –1.0 –0.5 0 0.5 1.0 1.5 –1.5 –1.0 –0.5 0 0.5 1.0 1.5
Merkens and Schamhardt (1988a, b) tried to establish a clinically Longitudinal force (N/kg) Longitudinal force (N/kg)
oriented assessment system based on the bilateral symmetry of fore
and hind limbs. Both amplitude and temporal variables were used
to define various symmetry indices that were combined into a single 8 8
F3 H3
H(orse)INDEX (Table 9.3). Furthermore, graphical displays of the
vertical and horizontal forces are presented that allow a quick quali-
6 6
Vertical force (N/kg)

Vertical force (N/kg)

tative assessment of the presence of asymmetries by the veterinarian
(Fig. 9.14) (Merkens et al., 1988). The major advantage of force
plate analysis is that it does not require any instrumentation on 4 4
the horse, so there are minimal problems for the horse and its
owner. Furthermore, the results are very accurate and reliable. A
disadvantage is the large number of trials needed to get enough data 2 2
when using a small force plate, which requires some time and
patience for all concerned. This disadvantage was overcome by the
instrumented treadmill in Zürich, which allows for continuous 0 0
assessment of vertical GRFs, which, however, is still an unique –1.5 –1.0 –0.5 0 0.5 1.0 1.5 –1.5 –1.0 –0.5 0 0.5 1.0 1.5
measurement device (Weishaupt et al., 2002). Longitudinal force (N/kg) Longitudinal force (N/kg)
A special application of the force plate as a diagnostic instrument
has been described by Aviad (1988), who used the stability of the Fig 9.14  Vertical-longitudinal force curves (vectordynamograms) of all
limbs of one horse during sound control session (H0), moderate left
force versus time signal of horses standing with the lame limb on
forelimb lameness (F2), severe left forelimb lameness (F3), mild left hind
a force plate to assess the weight bearing stability and improvement
limb lameness (H2) and moderate left hind limb lameness (H3).
due to therapy. Similarly, Clayton et al. (2003) have used the force
Reprinted with permission from: Merkens H.W., Schamhardt H.C., 1988b. Evaluation of
plate to detect postural sway in standing horses as an indicator of
equine locomotion during different degrees of experimentally induced lameness. II.
neurological disease.
Distribution of ground reaction force patterns or the concurrently loaded limbs. Equine Vet J
Besides force plate studies, two other kinetic devices have been
(Suppl.) 6, 107–112.
used for the analysis of equine lameness: force shoes and the EGA
system. Force shoes measure GRFs in special measuring devices
integrated into a horseshoe. Different types of force shoes using
piezoelectric transducers (Ratzlaff et al., 1990) or strain gauges
(Roepstorff & Drevemo, 1993; Kai et al., 2000) have been designed. are integrated into a special floor. This system needs no instrumen-
Problems in the design of force shoes include the delicate nature of tation on the horse and offers data describing the loading of each
the instrumentation and the considerable weight of the shoes limb as well as temporal stride variables. Despite the simplicity of
(Hugelshofer, 1982), which impedes their use in a clinical situation. use, the high cost of the EGA system prevented the widespread
However, the technology is improving, and a clinically useful force adoption of its use.
shoe may eventually be developed.
Auer and Butler (1980) first described the Kaegi system as a tool
for assessing limb loading patterns in lame horses. The system, later
Artificial neural networks
improved and described as the EGA system (Huskamp et al., 1990; A nice computational method to assess the results of locomotion
Tietje, 1992), measures vertical forces in a number of sensors, which analysis, and to detect asymmetry or lameness, is the use of artificial

191
 9 Gait adaptation in lameness

neural networks (ANN). ANNs are computer programs that use

similar methods to the human brain to recognize characteristic pat-
terns in the kinematic or kinetic variables. Similar to human experi-
ence, neural networks need to be trained with input data and the
corresponding output message. For example, the ANN might be
given input data describing the joint angle values or symmetry
indices of the head and sacrum, together with the corresponding
output message that there is a grade 2 lameness in the left forelimb.
After several training cycles, the networks can use different, new
input data to arrive at a conclusion describing the corresponding,
unknown output information.
There have been three studies in which kinematic data from
groups of lame horses were used as input data for an ANN: Schobes- + -
berger and Peham (2002) used data of 175 horses with a variety of
different orthopedic diseases, Van Loon et al. (1995) used the sym-
metry indices of the lameness study of Buchner et al. (1995a,
1996a, b), and Keegan et al. (2003) based the ANN on the wavelet
transformation of head movement in 12 horses. All authors were
able to detect the lame limb and to assess the lameness degree with Fig 9.15  Flex-o-meter. (A) Polyvinylchloride plate with rubber tube as inner
probabilities between 75 and 85%. However, they concluded that lining; (B) electronic manometer; (C) charge amplifier.
the quality and the reliability depended on using a sufficiently high
number of training data. Schobesberger and Peham (2002) esti-
mated the necessary amount of input data for a correct diagnosis
in 90% of the cases to be about 6600 input values, and to reach for
95% probability of a correct diagnosis required 13 200 input values. difference in mean force: female examiners used a mean force of
The fact the ANNs require such a large amount of kinematic input 114 N compared with 144 N in male examiners. The intraindivid-
data to be adequately trained is likely to restrict their clinical use ual variation, which is the variation within repeated flexion tests by
for lameness diagnosis in the near future. the same person, was smaller, only about 12%. Therefore, to achieve
a high repeatability, it is necessary to be consistent in performing
the flexion tests. The same person should perform flexion tests in
Biomechanical studies of diagnostic or both limbs of the same horse to allow for a correct interpretation,
and it is not possible to equate the results found by different
therapeutic aids in equine orthopedics examiners.
The second variable, the duration of the forced joint flexion, is
A number of kinetic or kinematic techniques have been used to also very important. Two studies analyzed the influence of changing
evaluate the effects of diagnostic or therapeutic procedures in flexion force and duration on the outcome of the test in sound
equine orthopedics. Some of these procedures are frequently used, horses (Keg et al., 1997b; Verschooten & Verbeeck, 1997). Both
but their effects are often small and difficult to quantify. Knowledge studies found dramatic increases in the number of positive flexion
regarding the effects of these procedures has been based primarily tests by increasing the time of flexion to 3 or 5 min. This factor was
on experience or logical assumptions, but not on objectively more important than the force, in which an increase of 25% resulted
determined results. Especially due to the need for an objective vali- in only one additional horse out of eight becoming lame (Keg et al.,
dation of therapies for an evidence based medicine approach, the 1997b). Increasing the force from 100–150 N caused only about
quantification of equine locomotion became an important topic. 6% more positive results in the flexion tests (Verschooten & Ver-
Detailed biomechanical studies are available for diagnostic proce- beeck, 1997). Both studies stressed the importance of standardizing
dures like the effects of flexion tests and diagnostic nerve blocks. the flexion tests in regard to force and duration, and both concluded
Similarly several therapeutic aids have been tested: casts and ban- that 60 s is an adequate duration for the flexion test. For the applied
dages, medical therapies, extracorporal shock wave therapy and the force, Keg et al. (1997b) suggested 150 N, while Verschooten and
effects of desmotomy of the accessory ligament of the deep digital Verbeeck (1997) preferred only 100 N. More important, however,
flexor tendon. An important field for biomechanical evaluations is the principle of using the same person to apply all the flexion
are studies about the value of various orthopedic shoeing tech- tests within one horse to achieve the necessary consistency in the
niques; these studies are described elsewhere and will not be con- applied force.
sidered here.
Diagnostic anesthesia
Diagnostic aids The effects of diagnostic nerve blocks on the gait of lame horses has
often been described (Girtler et al., 1987; Merkens & Schamhardt,
Flexion tests 1988a, b; Keg et al., 1992, 1994, 1996a; Keegan et al., 1997) and
A flexion test is a standard procedure during equine lameness exam- can easily be quantified. More ambiguous results are reported
inations. The objective is to test if pain in a certain region and a concerning the effect of diagnostic anesthesias on the gait
worsening of a lameness can be provoked by prolonged flexion of of sound horses. This information is important for differentiating
the regional joints. Typical examples of flexion tests are digital sound horses from those with slight bilateral lameness. Bilaterally
flexion, carpal flexion and hock flexion (spavin) tests. Two factors lame horses do not show any asymmetries in their gait patterns, but
have a significant influence on the outcome of a flexion test: the diagnostic nerve blocks of one affected limb cause a visible lame-
force applied and the duration of force application. Keg et al. ness of the contralateral limb. This is the only way to prove a bilat-
(1997a) analyzed the forces used by a group of clinicians using a eral orthopedic problem. This diagnostic procedure, however, is
force transducer, the ‘flex-o-meter’ (Fig. 9.15). They found consider- based on the assumption that diagnostic blocks in sound horses do
able interindividual variation between different clinicians of about not affect their locomotor pattern, or at least, have a different effect
20% of the mean force. Furthermore, there was a significant sex than in bilaterally lame horses. Several investigators tested this

192
 Biomechanical studies of diagnostic or therapeutic aids in equine orthopedics

assumption using different methods. All the studies of diagnostic mat. They measured local peak pressures up to 14.4 N/cm2 at the
anesthesia in sound horses found slight gait changes, but they dif- gallop and maximal total forces up to 235 N under the bandage.
fered in quality and quantity. Kübber et al. (1994) and Drevemo They concluded that even when no kinematic effect can be seen,
et al. (1999) found some kinematic changes that were indicative of the pressure under the bandage might impede blood flow in the
a slight increase in weight bearing on the anaesthetized limb. distal limb, which could be counterproductive in terms of protec-
Kübber et al. (1994) reported increased asymmetry in the symmetry tion and performance of the horse. Furthermore, if bandages were
variables of the head and withers after the nerve block in 9 out of applied tightly enough to support the tendons during locomotion,
12 sound horses, while Drevemo et al. (1999) found an increased the forces applied to the limb at different locations under the
range in the joint angle pattern of the fetlock joint. The changes bandage, for example the metacarpal skin, are likely to cause prob-
were very small and some were very close to the border of statistical lems that would outweigh any positive effects at the tendons or
significance. Looking at the loading of the limbs using a force plate, fetlock joint.
however, Keg et al. (1996b) could not find differences in the vertical
load before and after local nerve blocks. Only one variable, the time
of change from a decelerative to an accelerative horizontal force
Analgesia, NSAIDs, ESWT
changed significantly. This indicates a change in the proprioceptive The effectiveness of various pharmaceuticals for the therapy of lame-
information and a consequent slight change in the locomotion ness has been quantified in terms of changes in the lameness degree
pattern of the horses. This change in proprioception was also pro- or loading of the limbs. Mostly the peak vertical force was used to
posed by Kübber et al. (1994) and Drevemo et al. (1999) and leads follow the effects of an anti-inflammatory therapy, like phenylbuta-
to the conclusion that sound horses do indeed show small changes zone, flunixin, etodolac, cox-2 inhibitors (firocoxib) or combina-
in their gait pattern after a local nerve block. However, these tions with intrasynovial corticocoids and orthopedic shoeings (Back
changes are very small and there might be a gray zone in which it et al., 2009; Hu et al., 2005; Erkert et al., 2005; Schoonover et al.,
is impossible to differentiate between the responses to desensitiza- 2005; Symonds et al., 2006). Similarly, the immediate analgesic
tion of the limbs in horses with a slight or subclinical, bilateral effect of extracorpral shock wave therapy (ESWT) was tested and
lameness from the reactions of sound horses. found to be negligible in horses suffering from navicular disease
using the simple parameter peak vertical force (Brown et al., 2005).

Therapeutic aids
Desmotomy
Bandages A study of the therapeutic effects of desmotomy of the accessory
Bandages on the limbs of horses are a widely used tool and serve ligament of the deep digital flexor tendon (also known as the distal
several different functions. They have important medical uses in check ligament, DCL) took a step further and used inverse dynamic
wound management or immobilization, and therapeutic uses for analysis to assess the consequences of this procedure for the digital
applying pressure, cold, heat, water or various medical agents joints and tendons (Buchner et al., 1996c; Becker et al., 1998a).
(Stashak, 1987). Another important use of bandages is to protect Desmotomy of the DCL is usually performed in young, growing
against injuries of the distal limbs due to interfering, overreaching, horses suffering from flexural deformity of the distal interphalan-
hitting a fence or any other cause of skin trauma (Dyson, 1994). geal joint (McIlwraith & Fessler, 1978). Recently, desmotomy of the
The most controversial use of bandages is to prevent tendon strains DCL has also been proposed as a possible therapy for DCL desmi-
or joint injuries associated with loading during locomotion. The tis, similar to desmotomy of the accessory ligament of the superfi-
distal limbs are often bandaged to prevent or support tendinitis or cial digital flexor tendon to treat SDF tendinitis (Becker et al.,
arthritis of the fetlock joint. There are some studies, however, which 1998a). In a long-term study all clinical, ultrasonographical, histo-
disprove this protective function or even indicate possible adverse logical and biomechanical aspects of the desmotomy were assessed
effects of bandages for this purpose. The effects of bandages on to evaluate the advantages or disadvantages of the desmotomy for
tendon strain can be evaluated directly using invasive methods, treating chronic desmitis of the DCL. A combination of kinetic,
such as strain gauge implants into the tendons, or indirectly, using kinematic and radiologic techniques was used to study the function
non-invasive, kinematic methods. Keegan et al. (1992) implanted of the DCL for normal locomotion and to follow the changes in
strain gauges into the suspensory ligament of 9 horses. They com- joint moments, tendon forces and limb kinematics at intervals of
pared ligamentous strain while standing and walking with two types 10 days and 6 months after DCL desmotomy in sound horses.
of casts and four types of supportive bandaging materials as well as Ten days after desmotomy the horses had no visible lameness or
different bandaging techniques. The results clearly showed that sig- changes in limb loading, but alterations in the locomotion pattern
nificant support can only be expected for the full cast and the dorsal were clearly indicative of the loss of biomechanical function of the
fetlock splint. There was no effect of any bandage on suspensory DCL (Buchner et al., 1996c). During the whole stance phase a
ligament strain. caudad shift of the point of force application at the hoof (center of
Kobluk et al. (1990) applied the close relationship between pressure) reduced the moment arm of the GRF and, consequently,
fetlock joint angle and suspensory ligament strain to evaluate the the net joint moment at the coffin joint. Despite the loss of DCL
supporting capacities of bandages. They measured the kinematics function, the joint motion pattern at the beginning and in the
of galloping horses that had been wrapped with different types of middle of the stance phase was not changed. An increase in function
support bandages. Some horses showed slightly decreased fetlock of the SDFT compensated for the loss of the DCL, so the fetlock
hyperextension with bandages, but others showed no decrease or joint angle was unchanged during maximal vertical loading at mid-
even had increased fetlock joint angles. In general, there was no stance. At the end of the stance phase, it was primarily the deep
proof for a protective effect of supporting bandages on fetlock joint digital flexor tendon (DDFT), rather than the SDFT, that took over
angle or tendon strain during loading of the limbs. There might be the function of the deficient DCL (Fig. 9.16). However, the DDFT
a physical restriction of the amount of flexion during the swing could not fully compensate the DCL function. Consequently, some
phase, but this is unlikely to prevent injuries associated with limb kinematic changes occurred in the later part of the stance phase: the
loading. fetlock joint remained hyperextended for a longer time before
On the other hand, bandages cause pressure on the limb, with flexing rapidly at the end of the stance phase, whereas the carpal
the amount of pressure being dependent on the type of bandage joint started to flex earlier in the stride cycle. The increased loading
and the speed of locomotion. Morlock et al. (1994) quantified the of SDFT and DDFT was found to be within the normal range of
pressures on the metacarpal skin using a small pressure-sensitive loading at the walk and the trot. Therefore, it was concluded that

193
 9 Gait adaptation in lameness

0.60 (1989). By using a computer model to simulate the loss of the DCL,
SL + SDFT
the increase in the DDFT force, as well as the change in the position
0.50
of the point of force application, were predicted correctly, but the
Moments (Nm/kg)

0.40 increase in the force of the suspensory ligament (SL) was overesti-
DDFT + DCL mated. Obviously, the SDFT could compensate for the DCL des-
0.30
motomy by maintaining the normal midstance angle of the fetlock
0.20 joint so that SL loading was unchanged.
0.10 Six months after the desmotomy, the function of the DCL was
partly restored by scar tissue formation, which restored 80% of the
0.00 original tensile strength of the DCL, while its length had increased
0 10 20 30 40 50 60 70
by 1 cm (Becker et al., 1998b). This healing process restored its
Time (% of stride) biomechanical function to a certain extent and reduced the typical
Fig 9.16  Tendon moments at the fetlock joint of suspensory ligament and
locomotor changes seen 10 days after the desmotomy. However,
superficial digital flexor tendon (SL + SDFT), deep digital flexor tendon and most of the locomotor changes persisted, especially the caudad shift
its distal check ligament (DDFT + DCL) averaged over six horses at the walk of the point of force application, and it was assumed that healing
before and after desmotomy of the DCL. Solid lines, before desmotomy; would continue for a longer period of time.
dashed lines, after desmotomy. In conclusion this study documented the biomechanical
Reprinted from Buchner, H.H.F., Savelberg, H.H.C.M., Becker, C.K., 1996. Load redistribution changes due to DCL desmotomy and quantified the consequences
after desmotomy of the accessory ligament of the deep digital flexor tendon in adult horses. for the net joint moments and the loading of the tendons. Even
Veterinary Quarterly 18, S2, S70–74, with permission from Taylor & Francis Ltd, http://www. 6 months after desmotomy, healing was still in progress and DCL
informaworld.com. function was not fully restored. These experimental and model
studies of DCL function show the potency of non-invasive
methods for assessing the biomechanical effects of therapeutic
there is no risk of damage to the compensating tendons after DCL procedures on the internal forces in limb joints and tendons.
desmotomy, provided locomotion is controlled during the recovery Similar studies have been performed to assess the effects of
period. various types of orthopedic shoes and to enable an objective
These experimental findings nicely confirmed most of the details assessment of the benefit of these therapeutic measures for the
that had been predicted by a model analysis of van den Bogert et al. health of the horse.

References
Adrian, M., Grant, B., Ratzlaff, M., et al., 1977. representations of equine forelimb kinematics of horse locomotion. J. Exper.
Electrogoniometric analysis of equine kinematics. Equine Vet. J. 27, 31–38. Biol. 210, 1885–1896.
metacarpophalangeal joint lameness. Am. Back, W., Schamhardt, H.C., Savelberg, Brown, K.E., Nickels, F.A., Caron, J.P., et al.,
J. Vet. Res. 38, 431–435. H.H.C.M., et al., 1995b. How the horses 2005. Investigation of the immediate
Audigie, F., Pourcelot, P., Degueurce, C., 2002. move: 2. Significance of graphical analgesic effects of extracorporal shock
Fourier analysis of trunk displacements: a representations of equine hind limb wave therapy for treatment of navicular
method to identify the lame limb in kinematics. Equine Vet. J. 27, 39–45. disease in horses. Vet. Surg. 34, 554–558.
trotting horses. J. Biomech. 35, 1173–1182. Barrey, E., Debrosse, F., 1996. Lameness Buchner, H.H.F., Kastner, J., Girtler, D., et al.,
Auer, J., Butler, K.D., 1980. An introduction to detection using an accelerometric device. 1993. Quantification of hind limb
the kaegi equine gait analysis system in the Pferdeheilkunde 12, 617–622. lameness in the horse. Acta Anat. 146,
horse. J. Am. Vet. Med. Assoc. 209–226. Barrey, E., Hermelin, M., Vaudelin, J.L., et al., 196–199.
Auer, J.A., Fackelman, G.E., Gingerich, D.A., 1994. Utilisation of an accelerometric Buchner, H.H.F., Obermüller, S., Scheidl, M.,
Fetter, A.W., 1980. Effect of hyaluronic acid device in equine gait analysis. Equine Vet. J. 2001. Body centre of mass movement in
in naturally occurring and experimentally Suppl. 17, 7–12. the lame horse. Equine Vet. J. Suppl. 33,
induced osteoarthritis. Am. J. Vet. Res. 41, Becker, C.K., Savelberg, H.H.C.M., Buchner, 122–127.
568–574. H.H.F., et al., 1998a. Long-term Buchner, H.H.F., Savelberg, H.H.C.M., Becker,
Aviad, A.D., 1988. The use of the standing consequences of experimental desmotomy C.K., 1996c. Load redistribution after
force plate as a quantitative measure of of the accessory ligament of the deep desmotomy of the accessory ligament of
equine lameness. J. Equine Vet. Sci. 8, digital flexor tendon in adult horses. Am. J. the deep digital flexor tendon in adult
460–462. Vet. Res. 59 (3), 347–351. horses. Vet. Q. 18, S70–S74.
Back, W., Barneveld, A., Bruin, G., et al., 1994. Becker, C.K., Savelberg, H.H.C.M., Buchner, Buchner, H.H.F., Savelberg, H.H.C.M.,
Kinematic detection of superior gait quality H.H.F., et al., 1998b. Effects of Schamhardt, H.C., et al., 1994. Kinematics
in young trotting warmbloods. Vet. Q. 16, experimental desmotomy on material of treadmill versus overground locomotion
91–96. properties, histomorphology and in horses. Vet. Q. 16 (Suppl. 2), 87–90.
Back, W., Barneveld, A., van Weeren, P.R., et ultrasonographic features of the accessory Buchner, H.H.F., Savelberg, H.H.C.M.,
al., 1993. Kinematic gait analysis in equine ligament of the deep digital flexor tendon Schamhardt, H.C., et al., 1995a. Temporal
carpal lameness. Acta Anat. Suppl. 146, in clinically normal horses. Am. J. Vet. Res. stride patterns in horses with
86–89. 59 (3), 352–358. experimentally induced fore or hind limb
Back, W., MacAllister, C.G., van Heel, M.C.V., Bialski, D., Lanovaz, J.L., Bohart, G.V., et al., lameness. Equine Vet. J. Suppl. 18,
et al., 2009. The use of force plate 2004. Effect of detomidine on postural 161–165.
measurements to titrate the dosage of a sway in horses. Equine and Comparative Buchner, H.H.F., Savelberg, H.H.C.M.,
new COX-2 inhibitor in lame horses. Ex. Physiol. 1, 45–50 Schamhardt, H.C., et al., 1995b. Bilateral
Equine Vet. J. 41, 309–312. Bobbert, M.F., Gómez Álvarez, C.B., van lameness in horses – a kinematic study.
Back, W., Schamhardt, H.C., Savelberg, Weeren, P.R., et al., 2007. Validation of Vet. Q. 17, 103–105.
H.H.C.M., et al., 1995a. How the horses vertical ground reaction forces on Buchner, H.H.F., Savelberg, H.H.C.M.,
move: 1. Significance of graphical individual limbs calculated from Schamhardt, H.C., et al., 1996a. Limb

194
 References

movement adaptations in horses with anaesthesia of the forelimb in the sound Goodship, A.E., Brown, P.N., Mac Fie, H.J.H.,
experimentally induced fore or hind limb horse. Equine Vet. J. Suppl. 30, 266–269 et al, 1983. A quantitative force plate
lameness. Equine Vet. J. 28, 63–70. Dusek, J., Ehrlein, H.J., Engelhardt, W.V., et al., assessment of equine locomotor
Buchner, H.H.F., Savelberg, H.H.C.M., 1970. Beziehungen zwischen trittlänge, performance. In: Snow, D.H., Persson,
Schamhardt, H.C., et al., 1996b. Head and trittfrequenz und geschwindigkeit bei S.G.B., Rose, R.J. (Eds.), Equine exercise
trunk movement adaptations in horses pferden. Zeitschrift für Tierzüchtung und physiology. Granta, Cambridge, pp.
with experimentally induced fore or hind Züchtungsbiologie 87, 177–188. 263–270.
limb lameness. Equine Vet. J. 28, 71–76. Dyson, S.J., 1994. Training the event horse. In: Hajer, R., Hendrikse, J., Rutgers, J.E., et al.,
Buchner, H.H.F., Savelberg, H.H.C.M., Hodgson, D.R., Rose, R.J. (Eds.), The 1988. Het klinisch onderzoek bij grote
Schamhardt, H.C., et al., 1997. Inertial athletic horse. Saunders, Philadelphia, pp. huisdieren. Wetenschappelijke Uitgeverij
properties in Dutch warmblood horses. J. 419–428. Bunge, Utrecht, pp. 178–214
Biomech. 30, 653–658. Erkert, R.S., MacAllister, C.G., Payton, M.E., et Hjertén, G., Drevemo, S., Eriksson, L.E., 1994.
Church, E.E., Walker, A.M., Wilson, A.M., et al., 2005. Use of force plate analysis to Shortening of the hind limb in the horse
al., 2009. Evaluation of discriminant compare the analgesic effects of during the stance phase. Equine Vet. J.
analysis based on dorsoventral symmetry intravenous administration of Suppl. 17, 48–50.
indices to quantify hind limb lameness phenylbutazone and flunixin meglumine in Holmström, M., Fredricson, I., Drevemo, S.,
during overground locomotion in the horses with navicular syndrome. Am. J. Vet. 1994. Biokinematic differences between
horse. Equine Vet. J. 41, 304–308. Res. 66, 284–288. riding horses judged as good and poor at
Clayton, H.M., 1986a. Cinematographic Firth, E.C., Wensing, T., Seuren, F., 1987. An the trot. Equine Vet. J. Suppl. 17, 51–56.
analysis of the gait of lame horses. J. induced synovitis disease model in ponies. Hu, H.H., MacAllister, C.G., Payton, M.E., et
Equine Vet. Sci. 6, 70–78. Cornell Vet. 77, 107–118. al., 2005. Evaluation of the analgesic effects
Clayton, H.M., 1986b. Cinematographic Foreman, J.H., Lawrence, L.M., 1991. Lameness of phenylbutazone administered at a high
analysis of the gait of lame horses II: and heart rate elevation in the exercising or low dosage in horses with chronic
chronic sesamoiditis. J. Equine Vet. Sci. 6, horse. J. Vet. Sci. 11, 353–356. lameness. J. Am. Vet. Med. Assoc. 226,
310–320. 414–417.
Galisteo, A.M., Cano, M.R., Morales, J.L., et al.,
Clayton, H.M., 1987a. Cinematographic 1997. Kinematics in horses at the trot Hugelshofer, J., 1982. Vergleichende kraft und
analysis of the gait of lame horses III: before and after an induced forelimb belastungszeit messung an den vorderhufen
fracture of the third carpal bone. J. Equine supporting lameness. Equine Vet. J. Suppl. von gesunden und an podotrochlose
Vet. Sci. 7, 130–135. 23, 97–101. erkrankten pferden. Vet. Med. Diss.
Clayton, H.M., 1987b. Cinematographic Gingerich, D.A., Auer, J.A., Fackelman, G.E., Giessen.
analysis of the gait of lame horses IV: 1979. Force plate studies on the effect of Huskamp, B., Tietje, S., Novak, M., et al.,
degenerative joint disease of the distal exogenous hyaluronic acid on joint 1990. Fussungs- und bewegungsmuster
intertarsal joint. J. Equine Vet. Sci. 7, function in equine arthritis. J. Vet. gesunder und strahlbeinlahmer pferde
274–278. Pharmacol. Therap. 2, 291–298. – gemessen mit dem equine gait analysis
Clayton, H.M., 1988. Cinematographic Girtler, D., 1988a. Untersuchungen über die system (EGA-system). Pferdeheilkunde 6,
analysis of the gait of lame horses V: dauer des bewegungszyklus – 231–236.
fibrotic myopathy. J. Equine Vet. Sci. 8, stützbeinphase, hangbeinphase, Hütter, H., 1997. Aussagekraft der
297–301. phasenverschiebung – bei lahmen und Winkeländerung der Zehengelenke von
Clayton, H.M., Bialski, D.E., Lanovaz, J.L., et bewegungsgestörten pferden im schritt und podotrochlosekranken Pferden.
al., 2003. Assessment of the reliability of a trab sowie kinetische beurteilungen zu Dissertation Veterinärmedizinischen
technique to measure postural sway in deren bewegungsmuster. Wiener Universität Wien.
horses. Am. J. Vet. Res. 64, 1354–1359 Tierärztliche Monatszeitschrift 5, 217–231. Kai, M., Aoki, O., Hiraga, A., 2000. Use
Clayton, H.M., Schamhardt, H.C., Willemen, Girtler, D., 1988b. Untersuchungen über die of an instrument sandwiched between
M.A., et al., 2000a. Kinematics and ground dauer des bewegungszyklus – the hoof and shoe to measure vertical
reaction forces in horses with superficial stützbeinphase, hangbeinphase, ground reaction forces and three-
digital flexor tendinitis. Am. J. Vet. Res. 61 phasenverschiebung – bei lahmen und dimensional acceleration at the walk, trot,
(2), 191–196. bewegungsgestörten pferden im schritt und and canter in horses. Am. J. Vet. Res. 61,
Clayton, H.M., Schamhardt, H.C., Willemen, trab sowie kinetische beurteilungen zu 979–985.
M.A., et al., 2000b. Net joint moments and deren bewegungsmuster. Wiener Kastner, J., 1989. Bewegungsmessung auf dem
joint powers in horses with superficial Tierärztliche Monatszeitschrift 5, 225–270. weg zur klinischen methode. Österreich
digital flexor tendinitis. Am. J. Vet. Res. 61 Girtler, D., 1988c. Untersuchungen über die Hochschulzeitung 9, 15–16.
(2), 197–201. dauer des bewegungszyklus – Keegan, K.G., 2005. Head movement pattern
Deuel, N.R., Lawrence, L.M., 1987. Laterality stützbeinphase, hangbeinphase, in horses with forelimb and hind limb
in the gallop gait of horses. J. Biomech. 20, phasenverschiebung – bei lahmen und lameness. Proceedings of the.annual
645–649. bewegungsgestörten pferden im schritt und Convention of the American Association of
Dow, S.M., Leendertz, J.A., Silver, C.A., et al., trab sowie kinetische beurteilungen zu Equine Practitioners 51, 114–120.
1991. Identification of subclinical tendon deren bewegungsmuster. Wiener Keegan, K.G., Arafat, S., Skubic, M., et al.,
injury from ground reaction force analysis. Tierärztliche Monatszeitschrift 5, 310–324. 2003. Detection of lameness and
Equine Vet. J. 23, 266–272. Girtler, D., Floss, F.N., 1984. Zur bewegung determionation of the affected forelimb in
Drevemo, S., Dalin, G., Fredricson, I., et al., gesunder und bewegungsgestörter pferde. horses by use of continuous wavelet
1980. Equine locomotion: 3. The In: Knezevic, P.F. (Ed.), Orthopädie bei transformation and neural network
reproducibility of gait in Standardbred huf- und klauentieren. Schlütersche, classification of kinematic data. Am. J. Vet.
trotters. Equine Vet. J. 12, 71–73. Hannover, pp. 132–139. Res. 64, 1376–1381.
Drevemo, S., Fredricson, I., Hjertén, G., et al., Girtler, D., Kastner, J., Holzreiter, S., 1987. Die Keegan, K.G., Baker, G.J., Boero, M.J., et al.,
1987. Early development of gait bewegung eines stützbeinlahmen pferdes 1992. Evaluation of support bandaging
asymmetries in trotting Standardbred colts. vor und nach diagnostischen during measurements of proximal
Equine Vet. J. 19, 189–191. leitungsanästhesien dargestellt in weg-zeit- sesamoidean ligament strain in horses by
Drevemo, S., Johnston, C., Roepstorff, L., et diagrammen. Wiener Tierärztliche use of a mercury strain gauge. Am. J. Vet.
al., 1999. Nerve block and intra-articular Monatszeitschrift 4, 135–142. Res. 53, 1203–1208.

195
 9 Gait adaptation in lameness

Keegan, K.G., Pai, P.F., Wilson, D.A., et al., tarsometatarsal joints on sagittal plane Merkens, H.W., Schamhardt, H.C., 1988a.
2001. Signal decomposition method of kinematics and kinetics of trotting horses. Evaluation of equine locomotion during
evaluating head movement to measure Am. J. Vet. Res. 64 (12), 1491–1495. different degrees of experimentally induced
induced forelimb lameness in horses Knezevic, P.F., 1982. Untersuchung auf lameness. I. Lameness model and
trotting on a treadmill. Equine Vet. J. 33 bewegungsstörung und /oder lahmheit. In: quantification of ground reaction force
(5), 446–451. Dietz, O., Wiesner, E. (Eds.), Handbuch der patterns of the limbs. Equine Vet. J. Suppl.
Keegan, K.G., Wilson, D.A., Smith, B.K., et al., pferdekrankheiten für wissenschaft und 6, 99–106.
2000. Changes in kinematic variables praxis. S.Karger, Basel, pp. 40–53. Merkens, H.W., Schamhardt, H.C., 1988b.
observed during pressure-induced Knezevic, P.F., Floss, F.N., Girtler, D., 1982. Evaluation of equine locomotion during
forelimb lameness in adult horses Erste klinische ergebnisse rechnergestützter different degrees of experimentally induced
trotting on a treadmill. Am. J. Vet. Res. kinematographischer bewegungsanalyse lameness. II. Distribution of ground
61, 612–619. beim pferd. Proceedings of the 14. reaction force patterns on the concurrently
Keegan, K.G., Wilson, D.J., Wilson, D.A., et al., Conference of the European Society of loaded limbs. Equine Vet. J. Suppl. 6,
1997. Effects of anesthesia of the palmar Veterinary Surgery, Istanbul, pp. 89–94. 107–112.
digital nerves on kinematic gait analysis in Kobluk, C.N., Martinez del Campo, L., Merkens, H.W., Schamhardt, H.C., Hartman,
horses with and without navicular disease. Harvey-Fulton, K.A., et al., 1990. A W., et al., 1986. Ground reaction force
Am. J. Vet. Res. 58, 218–223. kinematic investigation of the effect of a patterns of Dutch warmblood horses at
Keegan, K.G., Yonezawa, Y., Pai, P.F., et al., cohesive elastic bandage on the gait of the normal walk. Equine Vet. J. 18, 207–214.
2004. Evaluation of a sensor based system exercising Thoroughbred racehorse. Merkens, H.W., Schamhardt, H.C., Hartman,
of motion analysis for detection and Proceedings of the.annual Convention of W., et al., 1988. The use of H(orse) INDEX:
quantification of forelimb and hind limb the American Association of Equine a method of analysing the ground reaction
lameness in horses. Am. J. Vet. Res. 65, Practitioners Lexington, pp. 135–148. force patterns of lame and normal gaited
665–670. Kramer, J., Keegan, K.G., Wilson, D.A., et al., horses at the walk. Equine Vet. J. 20,
Keg, P.R., Barneveld, A., Schamhardt, H.C., et 2000. Kinematics of the hind limb in 29–36.
al., 1994. Clinical and force plate trotting horses after induced lameness of Merkens, H.W., Schamhardt, H.C., van Osch,
evaluation of the effect of a high plantar the distal tarsal and tarsometatarsal joints G.J.V.M., et al., 1993. Ground reaction
nerve block in lameness caused by induced and intraarticular administration of force patterns of Dutch warmblood horses
mid-metatarsal tendinitis. Vet. Q. 16, anesthetic. Am. J. Vet. Res. 61 (9), at normal trot. Equine Vet. J. 25, 134–137.
S70–S75. 1031–1036 Morlock, M.M., Kobluk, C.N., Jones, J.H., et
Keg, P.R., Schamhardt, H.C., van Weeren, P.R., Kübber, P., Kastner, J., Girtler, D., et al., 1994. al., 1994. Influence of bandage material on
1996b. The effect of diagnostic regional Erkenntnisse über den einfluss der tiefen pressure distribution under the bandage on
nerve blocks in the forelimb on the palmaranästhesie auf das gangbild des the distal forelimb of the galloping horse.
locomotion of clinically sound horses. Vet. lahmheitsfreien pferdes mit hilfe einer Gait Posture 2, 253–260.
Q. 18, S106–S109. kinematischen meßmethode. Morris, E.A., Seeherman, H.J., 1987.
Keg, P.R., Schamhardt, H.C., van Weeren, P.R., Pferdeheilkunde 1, 11–21. Redistribution of ground reaction
et al., 1996a. The effect of high palmar Leach, D.H., Crawford, W.H., 1983. forces in experimentally induced equine
nerve block on lameness provoked by Guidelines for the future of equine carpal lameness. Equine Ex. Physiol. 2,
collagenase induced desmitis of the lateral locomotion research. Equine Vet. J. 15, 553–563.
branch of the suspensory ligament. Vet. Q. 103–110. Muybridge, E., 1899. Animals in motion.
18, S103–S105.
Leach, D.H., Cymbaluk, N.F., 1986. Republished (1957). In: Brown, L.S. (Ed.).
Keg, P.R., van den Belt, A.J.M. , Merkens, H.W.,
Relationship between stride length, stride Dover publications, New York.
et al., 1992. The effect of regional nerve
frequency, velocity and morphometrics of Peham, C., Scheidl, M., Girtler, D., 1995.
blocks on the lameness caused by
foals. Am. J. Vet. Res. 47, 2090–2097. Definition eines lahmheitsmaßes mit
collagenase-induced tendonitis in the
midmetacarpal region of the horse: a study Leach, D.H., Drevemo, S., 1991. Velocity- hilfe der spektralanalyse. In: Knezevic, P.F.
using gait analysis, and ultrasonography to dependent changes in stride frequency and (Ed.), Orthopädie bei huf- und
determine tendon healing. J. Vet. Med. length of trotters on a treadmill. Equine Ex. klauentieren. Schattauer, Stuttgart, pp.
Assoc. 39, 349–364. Physiol. 3, 136–140. 412–423.
Keg, P.R., van Weeren, P.R., Back, W., et al., May, S.A., Wyn-Jones, G., 1987. Identification Peham, C., Scheidl, M., Licka, T., 1996. A
1997b. Influence of the force applied and of hind limb lameness. Equine Vet. J. 19, method of signal processing in motion
its period of application on the outcome of 185–188. analysis of the trotting horse. J. Biomech.
the flexion test of the distal forelimb of the McGuigan, M.P., Walsh, T.C., Pardoe, C.H., et 29, 1111–1114.
horse. Vet. Rec. 141, 463–466. al., 2005. Deep digital flexor tendon force Peham, C.h., Licka, T., Mayr, A., et al., 2000.
Keg, P.R., van Weeren, P.R., Schamhardt, H.C., and digital mechanics in normal ponies Individual speed dependency of forelimb
et al., 1997a. Variations in the force and ponies with rotation of the distal lameness in trotting horses. Vet. J. 160,
applied to flexion tests of the distal limb of phalanx as a sequel to laminitis. Equine 135–138.
horses. Vet. Rec. 141, 435–438. Vet. J. 37 (2), 161–165. Peloso, J.G., Stick, J.A., Soutas-Little, R.W., et
Kelmer, G., Keegan, K.G., Kramer, J., et al., McGuigan, M.P., Wilson, A.M., 2001. The al., 1993. Computer-assisted three
2005. Computer-assisted kinematic effect of bilateral palmar nerve analgesia on dimensional gait analysis of amphotericin-
evaluation of induced compensatory the compressive force experienced by the induced carpal lameness in horses. Am. J.
movements resembling lameness in horses navicular bone in horses with navicular Vet. Res. 54, 1535–1543.
trotting on a treadmill. Am. J. Vet. Res. 66 disease. Equine Vet. J. 33 (2), 166–171. Pollhammer-Zeilinger, S., 1996. Messung des
(4), 646–655. McIlwraith, C.W., Fessler, J., 1978. Evaluation bewegungsmusters podotrochlosekranker
Khumsap, S., Lanovaz, J.L., Clayton, H., 2004. of inferior check ligament desmotomy for pferde auf dem laufband mit hilfe eines
Three-dimensional kinematic analysis of treatment of aquired flexor tendon high speed videosystems. Dissertation,
horses with induced tarsal synovitis. Equine contracture in the horse. J. Am. Vet. Med. Wien.
Vet. J. 36 (8), 659–663. Assoc. 294, 293. Pourcelot, P., Degueurce, C., Audigié, F., et al.,
Khumsap, S., Lanovaz, J.L., Rosenstein, D.S., Meij, H.S., Meij, J.C.P., 1980. Functional 1997. Kinematic analysis of the locomotion
et al., 2003. Effect of induced unilateral asymmetry in the motor system of the symmetry of sound horses at a slow trot.
synovitis of distal intertarsal and horse. South African J. Sci. 76, 552–556. Equine Vet. J. Suppl. 23, 93–96.

196
 References

Pratt, G.W., O’Conner, J.T., 1978. A Stashak, T.S., 1987. Diagnosis of lameness. In: Weishaupt, M.A., Schatzmann, U., Straub, R.,
relationship between gait and breakdown Stashak, T.S. (Eds.), Adams’ lameness in 1993. Quantifizierung der
in the horse. Am. J. Vet. Res. 39, 249–253. horses, fourth ed. Lea & Febiger, stützbeinlahmheit mit hilfe
Ratzlaff, M., Hyde, M.L., Grant, B., et al., 1990. Philadelphia, pp. 100–56, 840–877. akzelerometrischer messungen am kopf des
Measurements of vertical forces and Strobach, A., Kotschwar, A., Mayhew, I.G., et pferdes. Pferdeheilkunde 9, 375–377.
temporal components of the strides of al., 2006. Gait pattern of the ataxic horse Weishaupt, M.A., Wiestner, T., Hogg, H.P., et
horses using instrumented shoes. J. Equine compared to sedated and nonsedated al., 2001. Assessment of gait irregularities
Vet. Sci. 10, 23–35. horses. Equine Vet. J. Suppl. 36, 423–426. in thehorse: eye vs. gait analysis. Equine
Ratzlaff, M.H., Grant, B.D., 1986. The use of Symonds, K.D., MacAllister, C.G., Erkert, R.S., Vet. J. 33, 135–140.
electrogoniometry and cinematography in et al., 2006. Use of force plate analysis to Weishaupt, M.A., Wiestner, T., Hogg, H.P., et
the diagnosis and evaluation of forelimb assess the analgesic effects of etodolac in al., 2004. Compensatory load
lameness. Proceedings of the Annual horses with navicular syndrome. Am. J. Vet. redistribution of horses with induced
Convention of the American Association of Res. 67, 557–561. weight-bearing hind limb lameness trotting
Equine Practitioners 31, 183–198. Tietje, S., 1992. Das EGA-system (equine gait on a treadmill. Equine Vet. J. 36 (8),
Ratzlaff, M.H., Grant, B.D., Adrian, M., 1982. analysis) – eine möglichkeit zur 727–733.
Quantitative evaluation of equine carpal bewegungsanalyse und Weishaupt, M.A., Wiestner, T., Hogg, H.P., et
lameness. J. Equine Vet. Sci. 2, 78–88. lahmheitsuntersuchung beim pferd. al., 2006. Compensatory load
Ratzlaff, M.H., Wilson, P.D., Hyde, M.L., et al., Dissertation München. redistribution of horses with induced
1993. Relationships between locomotor Uhlir, C., Licka, T., Kübber, P., et al., 1997. weight-bearing forelimb lameness trotting
forces, hoof position and joint motion Compensatory movements of horses with a on a treadmill. Vet. J. 171 (1), 135–146.
during the support phase of the stride stance phase lameness. Equine Vet. J. Wijnberg, I.D., Schrama, S.E.A., Elgersma, A.E.,
of galloping horses. Acta Anat. 146, Suppl. 23, 102–105. et al., 2009. Quantification of surface EMG
200–204. Van den Bogert, A.J., Schamhardt, H.C., signals to monitor the effect of a botox
Sauren, A.A.H.J., et al., 1989. Computer treatment in six healthy ponies and two
Riemersma, D.J., Schamhardt, H.C., Hartman,
simulation of equine locomotion and its horses with stringhalt: preliminary study.
W., et al., 1988. Kinetics and kinematics of
application to the study of force Equine Vet. J. 41, 313–318.
the equine hind limb: in vivo tendon loads
and force plate measurements in ponies. distribution in the limbs. In: van den Williams, G.E., Silverman, B.W., Wilson, A.M.,
Am. J. Vet. Res. 49, 1344–1352. Bogert, A.J. (Ed.), Computer simulation of et al., 1999. Disease-specific changes in
locomotion in the horse. PhD Thesis, equine ground reaction force data
Roepstorff, L., Drevemo, S., 1993. Concept of
Utrecht University, pp. 111–136. documented by use of principal
a force-measuring horse shoe. Acta Anat.
Van Loon, J.P.A.M., Savelberg, H.H.C.M., component analysis. Am. J. Vet. Res. 60,
146, 114–119.
Buchner, H.H.F., et al., 1995. Lameness 549–555.
Schobesberger, H., Peham, C., 2002.
diagnosis in horses using artificial neural Williams, I.F., McGullagh, K.G., Goodship,
Computerized detection of supporting limb
networks. ESMAC Conference, Enschede, p. A.E., et al., 1984. Studies on the
lameness in the horse using an artificial
43 pathogenesis of equine tendonitis
neural network. Vet. J. 163, 77–84.
Van Weeren, P.R., van den Bogert, A.J., Back, following collagenase injury. Res. Vet. Sci.
Schoonover, M.J., Jann, H.W., Blaik, M.A., W., et al., 1993. Kinematics of the 36, 326–338.
2005. Quantitative comparison of three Standardbred trotter measured at 6, 7, 8 Wilson, A.M., McGuigan, M.P., Fouracre, L., et
commonly used treatments for navicular and 9 m/s on a treadmill, before and after al., 2001. The force and contact stress on
syndrome in horses. Am. J. Vet. Res. 66, 5 months of prerace training. Acta Anat. the navicular bone during trot locomotion
1247–1251. 146, 154–161. in sound horses and horses with navicular
Seeherman, H.J., 1991. The use of high-speed Verschooten, F., Verbeeck, J., 1997. disease. Equine Vet. J. 33, 159–165.
treadmills for lameness and hoof balance Flexion test of the metacarpophalangeal Wilson, D.A., Keegan, K.G., 1995.
evaluations in the horse. Vet. Clin. N. Am. and interphalangeal joints and flexion Pathophysiology and diagnosis of
Equine Prac. 7, 271–309. angle of the metacrpophalangeal musculoskeletal disease. In: Kobluk, C.N.,
Seeherman, H.J., 1992. Gait analysis. In: joint in sound horses. Equine Vet. J. 29, Ames, T.R., Geor, R.J. (Eds.), The horse,
Robinson, N.E. (Ed.), Current therapy in 50–54. diseases & clinical management. W.B.
equine medicine 3. Saunders, Philadelphia, Vorstenbosch, M.A.T.M., Buchner, H.H.F., Saunders, Philadelphia, pp. 607–658.
pp. 786–790. Savelberg, H.H.C.M., et al., 1997. Wittmann, F., 1931. Die untersuchung auf
Silver, C.A., Brown, P.N., Goodship, A.E., Modelling study of compensatory head lahmheit. In: Wittman, F. (Ed.),
1983. Biomechanical assessment of movements in lame horses. Am. J. Vet. Res. Chirurgische diagnostik des pferdes.
locomotor performance in the horse. 58, 713–718. Ferdinand Enke Verlag, Stuttgart, pp.
Equine Vet. J. Suppl. 1, 23–35. Weishaupt, M.A., Hogg, H.P., Wiestner, T., 140–149.
Speirs, V.C., 1994. Lameness: approaches to et al., 2002. Instrumented treadmill Wyn-Jones, G., 1988. The diagnosis of the
therapy and rehabilitation. In: Hodgson, for measuring vertical ground reaction causes of lameness. In: Equine lameness.
D.R., Rose, R.J. (Eds.), The athletic horse. forces in horses. Am. J. Vet. Res. 63, Blackwell Scientific Publications, Oxford,
Saunders, Philadelphia, pp. 343–370. 520–527. pp. 1–22.

197
This page intentionally left blank
 CHAPTER 10 

The neck and back

Claudia Wolschrijn, Fabrice Audigié, Inge D. Wijnberg,
Christopher Johnston, Jean-Marie Denoix, Willem Back

Function and neck changes the center of gravity, which provides means for
changing the speed of the gaits and even movement on the same
position. The thoracolumbar part of the vertebral column is rela-
The goal of this chapter is to present a synthesis of the current tively stiff; it has to oppose the forces exerted by the relatively heavy
knowledge on in vivo neck and back function and dysfunction at abdominal organs and it has to transmit the propulsive forces
walk, trot, gallop. Descriptions of equine vertebral column move- brought about by the hind limbs.
ments are often limited to the relative displacements of different
axial regions of the horse (mainly neck versus back). These displace-
ments induced and controlled by long muscle chains have been Bones
measured using kinematics and accelerometrics, of which in general The first cervical vertebra, the atlas, consists of two arches (dorsal
surface EMG measures muscular activity and needle EMG muscular and ventral), two lateral wings (enlarged transverse processes) and
functionality. An anatomical overview of the vertebral column pre- no vertebral body. On its cranial end it carries a dorsal tubercle
cedes the locomotor data review on its function and dysfunction (remnant of the spinous process) on its caudal end a ventral one.
(Clayton et al., 2005). The wings are an important palpable landmark, to which the head
and neck muscles responsible for rotary movements of the head,
attach. During development of the vertebrae in early embryonic life
Anatomy (a part of) the body of the first vertebra fuses with the body of the
The spinal column of the horse extends from the occipital condyles second, providing the latter with a cranial protrusion, the dens. The
to the tail and consists of approximately 50 separate short bones, dens, resting on the floor of the vertebral foramen of the atlas,
the vertebrae. The spinal column supports the body axis and thus provides the axis of rotation for the atlas and the head. The body
the maintenance of posture. The vertebrae are connected by short of the axis is long and carries a ventral crest that ends on the caudal
and long ligaments, and two types of articulations, and therefore, ventral tubercle. The high spinous process of the axis gives rise to
the spinal column also plays an important role in locomotion. The two articular facets at its caudal extremity; these meet the cranial
typical vertebral formula for the horse is seven cervical, 18 thoracic, facets on the third cervical vertebra. The cervical vertebrae become
six lumbar, five sacral, and 15–21 caudal vertebrae (Fig. 10.1), each progressively shorter towards caudal. The third, fourth and fifth
with its own specific characteristics. The basic form of vertebrae (this cervical vertebrae are long and strong and carry a sharp robust
does not account for the first cervical vertebra) consists of a ventrally ventral crest on their body that terminates in the ventral tubercle.
placed vertebral body and a dorsally placed vertebral arch, which The spinous processes are weakly developed. The transverse pro-
encloses the vertebral foramen. The body is composed of an outer cesses are split into a cranioventral and a dorsocaudal tubercle,
cortex of compact bone, which surrounds trabecular bone. The connected by a crest. The sixth cervical vertebrae is shorter, but
cranial aspect of the body is usually convex, the caudal aspect carries an impressive ventral crest. The ventral tubercle of the trans-
concave. Both are covered with hyaline cartilage and can as such be verse process is replaced by a ventral plate (lamina ventralis). The
regarded as the non-ossified part of the epiphysis of the vertebral seventh cervical vertebra has the shortest body of the cervical verte-
body. The basal part of the arch is notched on both the cranial and brae, still a weakly developed spinous process and no ventral plate.
caudal margin; the two notches of successive vertebrae form one On its caudal aspect the body carries not only the articular process
intervertebral foramen that transmits the spinal nerves and vessels for the cervicothoracic joint but also the fovea costalis for articula-
(Fig. 10.2). The spinous process, which is located, on top of the tion with the first rib.
arch, varies in its shape, length and inclination. On the left and right The 18 (range 17–19) thoracic vertebrae have short bodies, low
junctions of the body and the arch the transverse processes can be articular processes, closely fitting vertebral arches and long spinous
found. Mamillary processes, if present, are located between the processes. The latter increase rapidly in size with T2–T9 forming the
cranial articular process and the transverse process; accessory pro- basis of the withers. Normally, the tip of T5 or T6 forms the highest
cesses are located between the caudal articular process and the point of the withers. Thereafter a gradual decline in height of the
transverse process. The head and neck unit plays an important role spinous processes can be seen. The spinous processes of the first 14
in balancing and in locomotion. Changing the position of the head thoracic vertebrae have a backward inclination, the 15th one has an

199
 10 The neck and back

C1 Cervical C7

T1 Thoracic T18

L1 Lumbar L6 S1 Sacral S5

Ca1 Caudal

Fig 10.1  The vertebral formula for the horse is seven cervical, 18 thoracic, six lumbar, five sacral, and 15–21 caudal vertebrae.
Denoix, J.M., Pailloux, J.P., 1996. Anatomy and basic biomechanical concepts. In: Physical Therapy and Massage for the Horse, with permission from Manson Publishing Ltd.

Nerve roots upright spinous process, and the processes of the remaining tho-
racic vertebrae and of the loin or lumbar vertebrae have a forward
Spinal cord
inclination. Serial features are an appearance of an additional
Vertebral facet mamillary process as a projection from the transverse process and
its gradual migration to join the cranial articular process and a
change toward the end of the number of the thoracic vertebrae in
the character of the articular facets from the cervical to the lumbar
pattern (Fig. 10.1). There are sternal (10) and asternal (8) ribs, the
neck of the ribs become gradually shorter caudally. Costal facets are
present on both extremities of the vertebral body for articulation
with the rib heads and on the stubby transverse processes for the
rib tubercles. The articular facets of the heads and tubercle become
confluent in the caudal thoracic vertebrae which increases the
mobility of the last few pairs of ribs. The head of a rib has a cranial
A and a caudal facet, separated by a groove for the attachment of the
intra-articular ligament of the head of the rib.
Dorsal branch Intervertebral foramen The average number of lumbar vertebrae is six (range five to seven).
Dorsomedial branch
Their vertebral bodies are longer than those of the thoracic verte-
brae, the spinous processes shorter, and the transverse processes are
long, flattened and positioned in a cranioventral direction. The
transverse processes of the last two lumbar vertebrae and those of
the last lumbar and first sacral vertebrae articulate with each other.
Transverse processes cranial to L5 can also form articulations.
Fusion of the transverse processes can occur without clinical
implications.
The sacrum is a completely rigid structure, constructed of five
sacral vertebrae and their ossified intervertebral discs. The fusion is
complete at the age of 4–5 years. The fusion increases the effective-
ness of the forward thrust in locomotion from the hind limb to the
vertebral column. The fused transverse processes of the sacral ver-
tebrae are enlarged by the sacral wings, which project laterally and
originate from the first sacral vertebra. The sacral wings form a rigid
Dorsolateral branch joint with the ilial wings through the oval area (facies auricularis),
Ramus communicans
which is covered with cartilage. The dorsal margin of the sacral wing
B (tuberositas sacralis) is roughened for the attachment of the sacro-
Fig 10.2  The basal part of the arch is notched on both the cranial and iliac ligament.
caudal margin; the two notches of successive vertebrae form one The number of tail (coccygeal) vertebrae is very variable (15–21).
intervertebral foramen that transmits the spinal nerves and vessels: (A) a Unlike in other species such as the cat, the biomechanical role of
transverse view, (B) a sagittal view. the tail in horses is limited.

200
 Function

Vertebral joints supraspinous bursa) is present. Additional bursae can be found

between the nuchal ligament and the atlas or axis and are called bursa
The atlanto-occipital joint consists of two separate ellipsoid joints
subligamentosa nuchalis cranialis and caudalis respectively. The
between the occipital condyles and the concave articular surfaces of
fibrous supraspinous ligament is composed of two parallel parts like
the atlas. In old horses there can be a ventral communication
the funiculus of the nuchal ligament. In the lumbar region, the two
between the two joint cavities. The shape of the articular surface
parts can slope down from the top of the spinous processes. Its inner
restricts movement between the atlas and the skull to mainly flexion
structure is built up of successive fascicles that span 3–4 vertebrae. In
and extension; though some lateral bending is also allowed by this
the thoracolumbar region the supraspinous ligament is partly fused
type of joint. The atlanto-axial joint is a pivot joint between the
with the thoracolumbar fascia to which the large extensor muscles of
dens of the axis and the ventral part of the vertebral foramen of the
the back are attached. In the sacral region it unites with the gluteal
atlas. Furthermore, there are caudal articular facets on the body of
fascia and the sacroiliac ligament. The supraspinous ligament plays
the atlas and cranial ones on the axis. The joint construction allows
an important role in biomechans of the back. It stretches with flexion
rotational movements along the longitudinal axis of the dens. The
of the vertebral column and relaxes upon extension; it thus opposes
subsequent cervical vertebrae articulate with two different joint
the spreading of the spinous processes (Fig. 10.3).
types. First, there is the joint between the extremities of the bodies,
The dorsal longitudinal ligament can be found along the floor of
formed by the intervertebral discs and second there are joints
the vertebral canal, from the dens of the axis to the sacrum and is
between the articular facets on the arches of the vertebrae. The
attached to each intervertebral disc. The considerably shorter ventral
intervertebral discs allow movement in every plane, the facet joints
longitudinal ligament is located on the ventral aspect of the verte-
restrict the movement to one main plane, depending on their posi-
bral bodies from the 8th thoracic vertebra to the sacrum, and is also
tion. In the cervical vertebrae, the facet joints are more or less hori-
attached to each intervertebral disc.
zontally oriented allowing mainly lateroflexion, in the lumbar
The short ligaments comprise the ligamenta flava, the intertrans-
area the facet joints are sagittally orientated, allowing mainly
verse ligaments and the interspinous ligaments. The ligamenta flava
extension–flexion movements. In the thoracic vertebrae there is a
are elastic (yellow) sheets that fill the interarcuate spaces. The inter-
gradual change in position of the facet joints.
transverse ligaments stretch out between the transverse processes of
the lumbar vertebrae and are tensed during lateroflexion and rota-
Intervertebral discs
tion. The interspinous ligaments between the spinous processes are
The intervertebral discs between the vertebral bodies are composed elastic in the cranial part of the spine. They prevent the vertebrae
of a pulpy nucleus and a fibrous ring (anulus fibrosus), although from sliding dorsally, and limit ventral flexion of the spine.
especially in the adult horse the distinction between these two parts
is difficult (Yovich et al., 1985). The thickness of the discs decreases
throughout the thoracic and lumbar region to reach their minimum Muscles and nerves
thickness in the lumbar region. The pulpier nucleus is maintained The forelimb is attached to the trunk through muscles, the so-called
under pressure and spreads the compressive forces over a wider synsarcosis. The involved extrinsic muscles are the trapezius, rhom-
surface. As a result the surrounding fibrous ring and the ventral and boid, pectoral (superficial and deep), and the ventral serrate muscle
dorsal longitudinal ligaments are also brought under tension. Flexion (Payne et al., 2004). The latter consists of a cervical part (longer
of the vertebral arch brings about compression of the ventral side of fibers, less tendinous) (Fig. 10.4) and a thoracic part (short fibers,
the intervertebral disc and an upward (dorsal) movement of the disc. much tendinous tissue). The ventral serratus muscle is innervated
by a brachial plexus nerve, the long thoracic nerve. The following
Ligaments four muscles are also extrinsic muscles and therefore might be
The spine is passively stabilized by ligamentous structures. Liga- considered to be part of the synsarcosis as well: the subclavius, the
ments prevent too much excursion of the vertebrae. The ligaments latissimus dorsi, the omotransverse and the brachiocephalic muscle.
can be divided into the nuchal ligament, long and, short ligaments. The synsarcosis muscles are all involved in stabilizing the position
The specific ligaments of the head and neck region stabilize the of the scapula relative to the trunk, but during movement they also
joints between the skull and C1 and between C1 and C2. The lateral assist in protraction or retraction of forelimb. The synsarcosis pro-
ligaments are situated between the left and right paracondylar vides flexibility to absorb the impact forces upon landing on the
process of the skull and the left and right wings of the atlas. The ground. The trapezius muscle, consisting of a cervical and thoracic
dorsal and ventral atlanto–occipital membranes reinforce the joint part, assists both in pro- and retraction (cervical, pulls the scapula
capsule and arise dorsally and ventrally from the foramen magnum. to dorsocranial and thoracic part respectively). Both parts are motor
The well-developed elastic dorsal atlanto-axial ligament is located innervated by the accessory nerve, the sensory innervation is trans-
between the dorsal tubercle of the atlas and the spinous process of duced via the dorsal rami of the cervical and thoracic nerves. The
the axis. The less developed ventral atlanto–axial ligament can be rhomboid muscle lifts the scapula and turns the scapula to cranial.
found between the ventral tubercle of the atlas and the ventral part Dorsal branches of the cervical nerves and a special branch from
of the body of the axis. the brachial plexus (dorsal nerve of the scapula) innervate the
The structure and function of the nuchal ligament has been muscle (Fig. 10.5). The muscle covers the dorsoscapular ligament,
described by (Gellman & Bertram, 2002a, b). It supports the weight which is unique in the horse. A second head of the rhomboid
of the head, when the head is held high. The nuchal ligament is muscle (m. rhomboideus cervicis) (Fig. 10.6 (3)) runs cranially to
composed of two paired parts, the funiculus and the laminae. The the second cervical vertebra. The latissimus dorsi, which is a flexor
funiculus originating from the external occipital protuberance of the shoulder joint, becomes a powerful retractor with aid of the
receives the nuchal laminae at the level of the third cervical vertebra, ground reaction forces. The brachiocephalic muscle (cleidobrachia-
and inserts on the spinous process of the fourth thoracic vertebra. The lis, cleidocephalicus) is a protractor, but can also bend the neck and
two parallel parts of the funiculus broaden in the region of the head or move it to one side upon action. It is closely related to the
withers from where it continues as the supraspinous ligament to omotransversarius from which it is hard to separate. The cleidoce-
the caudal spinous processes of the lumbar vertebrae, and possibly phalicus and omotransversarius muscle are innervated by the acces-
the sacrum, until the first coccygeal vertebrae. The laminae (two sory nerve, the cleidobrachial part by the axillary nerve. Sensory
sheets of tendinous tissue) originate from the spinous processes of information is relayed by the ventral rami of the cervical nerves
the axis, the successive cervical vertebrae and insert on the first tho- (Dyce et al., 2010) (Fig. 10.7).
racic vertebrae. Between the nuchal ligament and the second or third The following muscles coordinate the movements of the head,
thoracic vertebra a bursa (bursa subligamentosa supraspinalis or especially by the atlanto-occipital and atlantoaxial joints, like

201
 10 The neck and back

Fig 10.3  The nuchal ligament supports the weight of the head, when the head is held high. It is composed of two paired parts, the funiculus and
the laminae.
Denoix, J.M., Pailloux, J.P., 1996. Anatomy and basic biomechanical concepts. In: Physical Therapy and Massage for the Horse, with permission from Manson Publishing Ltd.

shaking, tilting, flexing and turning the head. The m. rectus capitis muscle. It originates from the cranial border of the thoracolumbar
dorsalis major is located between the spine of the axis and the fascia, the spinal processes of the first three thoracic vertebrae and
occiput–nuchal crest, lateral to the nuchal ligament; it contains a the median neck raphe. The cephalic parts inserts the nuchal crest
deep and superficial part. The muscle extends the atlanto–occipital and mastoid process with the longissimus capitis. The cervical part
joint and is innervated by the dorsal branch of the first cervical inserts on the caudal part of the alar wing and the transverse pro-
nerve. Beneath it lies the minor rectus capitus dorsalis muscle, cesses of third and fifth cervical vertebrae. The muscle is a neck
which covers the dorsal atlanto–occipital membrane and shares the extensor and upon unilateral action it moves the neck to the side.
function and innervation with the major muscle. The lateral rectus It is thought to play an important role in maintaining balance
capitis muscle is a small muscular band, which extends from the during galloping. The muscle is innervated by the dorsal branches
ventral arch of the atlas to the paracondylar process of the skull. It of the cervical nerves. The long muscle of the head (m. longus
flexes the atlanto–occipital joint and tilts the head and is innervated capitis) (Fig. 10.6 (9)) is a strong muscle, which can be regarded as
by the ventral branch of the first cervical nerve. The ventral rectus a cranial continuation of the long muscle of the neck. It has long
capitis muscle between the ventral arch of the atlas and the base of muscles fibers, and is located on both sides next to the ventral part
the skull near the tympanic bulla also flexes the atlanto–occipital of the second to fourth cervical vertebra. It unites with the contra-
joint and has the same innervation as the lateral rectus muscle. The lateral muscle and ends between the guttural pouches on the base
m. obliquus capitis cranialis extends obliquely craniolaterally over of the skull. It flexes the atlanto–occipital joint, draws the head
the atlanto–occipital joint, covered by the splenius and a part of sideways, or the neck downward and is innervated by the ventral
the brachiocephalic muscle. It extends the atlanto–occipital joint branches of the first four cervical nerves. The segmented long muscle
and bends the head to the contracting side. It is innervated by the of the neck (m. longus cervicis) extends from the atlas to T6. It runs
dorsal branch of the first cervical nerve. The m. obliquus capitis craniomedially to insert on the bodies of the more cranial cervical
caudalis (Fig. 10.6 (12)) is positioned more caudally between vertebrae near the midline. The muscle flexes the neck and is inner-
the spinous process of the axis and the wing of the atlas; it receives vated by ventral branches of both the intercostal as the cervical
its innervation through the dorsal branch of the second cervical nerves dependent on the part of the muscle. Of the scalenus muscle,
vertebra. Unilateral action provokes rotation of the atlas and thus only the ventral (Fig. 10.6 (11)) and middle part are present. It takes
the head on the dens of the axis. Bilaterally it acts as a fixator of its origin from the first rib, and is divided by the brachial plexus. It
the head. inserts on the transverse processes of the third to seventh cervical
A neck muscle, not belonging to the large extensor systems, is the vertebrae (middle part) and the seventh cervical vertebra (ventral
splenius muscle, which is a large muscle, located at the dorsolateral part). The ventral branches of the fifth to eighth cervical nerve and
part of the neck, under the cranial (cervical) parts of the rhomboid the first two thoracic nerves innervate both parts. It flexes the neck,

202
 Function

basis of their topographical position and local presence. Unilateral

contraction evokes lateroflexion of the spine. The lumbar part of
the iliocostalis and longissimus system share the same origin at the
iliac crest, and are in this area strongly fused together, and covered
with a shiny aponeurosis. The thoracolumbar part of the longissi-
mus is composed of series of overlapping fascicles, that are cranio-
ventrally orientated and span several vertebral segments. The
iliocostalis muscle runs from the intermuscular fascia near the iliac
crest (hence the name), but also fascicles arise from the transverse
1 processes of the lumbar vertebrae and the ribs to attach to more
2I cranial segments with glistening small tendons on the caudal edges
2II of ribs 1–15 and the transverse process of the last cervical vertebra.
3 In the neck this muscle is weakly developed (Fig. 10.6 (5)). The
longissimus system is topographically divided into a lumbar, tho-
4
racic, cervical, atlantic (Fig. 10.6 (7)) and capital part (Fig. 10.6 (6)).
Its segmental origin is still reflected in the numerous individual
20 attachments of its muscle bundles. The origin is at the ilium and
sacrum, which have their insertion on the spinous and transverse
6 processes of the lumbar and thoracic vertebrae. The thoracodorsal
part is covered with a fascia. The cervical longissimus muscle is the
5 cranial continuation of the thoracic part. It has a triangular shape
and consists of four separate bundles that insert on the transverse
7 processes of the caudal cervical vertebrae. The atlantal longissimus
muscle (Fig. 10.6 (7)) extends from the last four cervical and first
two thoracic vertebrae to the wing of the atlas. It partly fuses with
the cranial part of the omotransversarius muscle. The capital longis-
10 simus muscle (Fig. 10.6 (6)) is a separate muscle, located medial to
9 the cervical part and the splenius muscle, originating at the first
11 thoracic vertebrae and extending to the mastoid process of the skull.
8 At the level of the wings of the atlas, it is strongly fused with the
12 splenius muscle. The corresponding dorsal branches of the spinal
19 17 nerves innervate these muscle parts. The muscle extends the back,
is active during the swing phase, and is also responsible for raising
18
16 the body during rearing and kicking with both the hind limbs.
When unilaterally contracted, it flexes the vertebral column laterally
13 or is responsible for rotation of the head.
14
15 The transversospinalis system is the most medial and deepest of
the neck and back extensors. Separate bundles connect the spinous
and transverse processes with subsequent vertebrae in the lumbar,
thoracic and cervical regions. Within the transversospinalis system
Fig 10.4  Transverse section of the neck at the level of the fourth vertebra. two separate groups of muscles can be distinguished. The first one
1, m. rhomboideus, cervical part; 2, nuchal ligament; 2′, funicular part; 2′′, is located directly over the skeleton occupying the space between
laminar part; 3, m. splenius; 4, m. semispinalis capitis; 5, m. longissimus the spinous processes, the vertebral arches and the transverse pro-
capitis; 6, m. spinalis cervicis; 7, m. intertransversarius; 8, m. cesses, and is named the spinal system. If the muscle bundles extend
brachiocephalicus; 9, m. longissimus cervicis; 10, m. longus capitis; 11, m.
between the spinal and the transverse processes of adjacent verte-
longus colli; 12, m. omohyoideus; 13, m. sternocephalicus; 14, m.
brae they belong to the transversospinal system. It is common to
sternothyroideus; 15, m. sternohyoideus; 16, trachea; 17, oesophagus; 18,
external jugular vein; 19, common carotid artery; 20. m. multifidi.
name both systems together as the transversospinalis system. The
fiber orientation is caudoventrolateral to craniodorsomedial. The
spinalis thoracis and cervicis muscle passes between the spinous
processes of adjacent vertebrae; it forms a common muscle belly,
draws it downwards and bends it sideways; it also supports inspira- which bridges several segments. The muscle is present between the
tion. The sternocephalic muscle (Fig. 10.6 (1)) has a fusiform body last lumbar vertebra and the third cervical vertebra; its fibers are
and a long cranial insertion tendon. Dependent on uni- or bilateral craniodorsally orientated. The muscle is innervated by the dorsal
action it flexes the head or provides lateroflexion. The muscle has branches of the topographically corresponding nerves. The semispi-
motor innervation from a large branch of the accessory nerve, the nalis capitis muscle is a strong muscle that is located in the triangle
sensory information is relayed by the ventral branch of the second between the skull, the transverse processes of the cervical vertebrae
cervical spinal nerve. and the nuchal ligament, and is covered by the longissimus and
The epaxial musculature of the neck and back can, apart from the splenius muscles. It is composed of two parts, dorsomedially the
above mentioned muscles, be divided into three main systems, the biventer cervicis (Fig. 10.6 (4)) and ventrolaterally the complexus
iliocostal, the longissimus and the transversospinal system, that are (Fig. 10.6 (8)) muscle. The biventer cervicis has its origin from the
located more or less next to each other in the thoracolumbar region, cranial thoracic vertebrae and its insertion on the skull. The muscle
where extensive fusion takes place. The most median muscle (trans- can be easily recognized because of the connective tissue inscrip-
versospinal system) group is deep to the other two groups, the most tions on the muscle belly, which show the multi-segmented origin
lateral one (iliocostal system) is very small in the cervical region. of the muscle. The semispinalis capitis muscle raises the head when
They lie parallel on both sides of the spinous process, and if they contracted bilaterally and flexes the head and neck to the lateral
are well-developed the supraspinous ligament can be regarded as side when acting unilaterally and is innervated by the dorsal
the bottom of the groove between the muscles. All three systems are branches of the cervical nerves. The superficial cervical epaxial
divided into a lumbar, thoracic, cervical and capital part on the muscles have been described by Gellman et al. (2002c).

203
 10 The neck and back

Cerebrum Cerebellum
C1 to C6
C6 to T2
T3 to L3 L4 to S2
S3 to S5

Brainstem

Head and face

Neck

Thoracic limb

Abdomen

Pelvic limb

Perineum and tail

Fig 10.5  A sagittal view with the different functional parts of the neurological system within the vertebral column: face and head, neck (C1–C5), thoracic
limb (C6–T2), abdominen (T3–L3), pelvic limb (L4–S2), tail and perineum (S3–S5).

12 8
3
6
4
7
9
2

11
5
1 10

Fig 10.6  A sagittal view to the cervical part of the muscles around the vertebral column: 1, m. sternocephalicus; 2, m. serratus ventralis cervicis; 3, m.
rhomboideus cervicis; 4, m. semispinalis capitis, biventer cervicis head; 5, m. iliocostali; 6, m. longissimus, capital part; 7, m. longissimus, atlantal part; 8, m.
semispinalis capitis, complexus head; 9, m. longus capitis; 10, m. scalenus, ventral part; 11, m. scalenus, middle part; 12, m. obliquus capitis caudalis.

Recently, the multifidous musculature has been the subject of the thoracic and cervical vertebrae to the spinous processes of the
study in the horse, because it atrophies in people with lower back preceding vertebrae. The bundles can span up to five segments in
pain (Stubbs et al., 2011). The muscle is composed of numerous the thoracic region. Cranially it unites with the cranial oblique
short bundles, which extend from the articular or mamillary pro- muscle of the head, caudally with the musculature of the tail. In
cesses of the lumbar vertebrae, and from the transverse processes of the neck five bundles are present at each vertebral level that are

204
 Function

nerve. The psoas minor muscle, which is located between the last
two or three thoracic vertebrae, and the first four or five lumbar
vertebrae, lies ventrally on the psoas major muscle. Due to its inser-
tion on the ilium it stabilizes the vertebral column and rotates the
iliosacral joint. The muscle can be easily recognized from its shiny
5
insertion tendon. It shares its innervation with the quadratus lum-
2 borum muscle.

4 6
1
Biomechanics
3
Several in vitro studies have been performed on the biomechanics
of the neck and back (Jeffcott & Dalin, 1980; Townsend et al., 1983;
Townsend & Leach 1984; Denoix, 1987; Clayton & Townsend,
A 1989a,b; Sleutjens et al., 2010), but our knowledge of these areas is
still incomplete. In vitro studies were mainly performed to evaluate
the nature and maximal amount of regional or intervertebral move-
ments of flexion and extension, axial rotation and lateral bending
(Gellman & Bertram, 2002a,b). With the different protocols used,
wide displacements were studied (Denoix, 1999a), but mobility of
3 the vertebral column rarely reaches its maximal limits during loco-
motion and sporting performance, although back pain due to pres-
2 sure, tension and/or shearing on vertebral structures (Denoix,
1999b) can alter locomotion of sport and racehorses (Jeffcott et al.,
1 1982; Denoix, 1998; Jeffcott, 1980). These studies on cadavers
4 found that the lumbar part of the spinal column was very rigid,
especially with respect to lateroflexion, with increasing mobility in
forward direction. The same applied to axial rotation. It should be
5
emphasized, however, that, although this work is of great value as
it gives insight into the potential for movement of various parts of
6 the equine spine, it does not represent reality, as there was obviously
B no influence of any of the active structures.
Of special interest for the biomechanical function of the vertebral
Fig 10.7  A transverse section of (A) the thoracic (1, m. latissimus dorsi; 2, m.
column is head and neck movement influence the movement of the
rhomboideus thoracis; 3, m. iliocostalis; 4, m. longissimus dorsi; 5, m. spinalis;
back. If the head is lowered, the nuchal ligament will exert a pull
6, m. multifidi) and (B) the lumbar part (1, m. iliocostalis; 2, m. longissimus
dorsi; 3, m. multifidus; 4, m. psoas major; 5, m. psoas minor; 6, crus of on the withers and flex the spinal column. Vice versa, lifting of the
diaphragm) of the muscles around the vertebral column. head will extend the back (Fig. 10.3). Understanding of this concept
is of great importance in the athletic training of sport horses. The
question of how the function of the back can best be described dates
innervated by the dorsal branches of the corresponding cervical back to Antiquity. Galenus (130–200 AD) developed the concept of
nerves. The attachment of the hind limb is much more stable, com- the ‘vaulted roof’ in which the back and the upper part of the rib
pared to the front limb, to optimize the transmission of the thrust cage form a roof over the abdominal and thoracic cavities. A col-
generated by the hind limbs to the back. The hind limb is connected lapse of this roof would be prevented by the spinous processes.
to the pelvis through the hip joint, the pelvis itself is connected with However, the fact that the spinous processes do not make contact
the rigid iliosacral joint to the spine. Extrinsic muscles are respon- in the normal situation makes this representation improbable. The
sible for pro- and retraction of the limb: the gluteal muscles (super- next concept was developed by Bergmann in 1847 and further
ficial, medial, deep and piriform), the sartorius, the rectus femoris, elaborated by Zschokke (1892). This concept implies the represen-
the biceps femoris, the semitendinosus and the semimembranosus tation of the back by a bridge that is resting on four piers (the
muscles. limbs). The upper ledger represents the supraspinal ligament and
The flexed arch of the spine is, for a large part, maintained by the withstands tensional forces, the lower ledger, the vertebral bodies,
abdominal musculature, especially the rectus abdominis muscle. is loaded under compression. The smaller girders between both
The hypaxial musculature also stabilizes the vertebral column, the ledgers represent the spinous processes and the ligaments in
quadratus lumborum muscle, the iliopsoas muscle and the psoas between these. Although this concept was generally adhered to until
minor muscle. World War II, it contains a basic error in that such a bridge will not
The quadratus lumborum muscle is a segmented muscle posi- be loaded by tension dorsally and by compression ventrally, but
tioned directly ventral to the bodies, the transverse processes and just the other way round. It was the zoologist Slijper who, in 1946,
the intervertebral discs from the last thoracic vertebrae to the sacral after a meticulous study of the anatomical form of the vertebrae
wings. It serves to stabilize the vertebral column: the ventral and especially of the inclination of the spinous processes in various
branches of the lumbar nerves and the last intercostal nerves inner- species, came up with the model that is still holding today. His
vate it. The psoas major muscle is located lateral to quadratus so-called bow-and-string concept not only takes into account the
lumborum, between the bodies and transverse processes of the last vertebral column and the limbs, but also the sternum and muscu-
two thoracic vertebrae and ribs and the lumbar vertebrae. The lature of the ventral abdomen. In this concept the vertebral column
iliacus muscle fuses at its insertion with the major psoas muscle to is a bow that is held under intrinsic tension by the abdominal wall
form the iliopsoas muscle. It originates from the medial side of the (Fig. 10.3).
ilium and is divided in a larger lateral part and a smaller medial Now the bony vertebral column is loaded under compression and
part. The common insertion is on the lesser trochanter and can flex the supraspinal ligament under tension (which is the only force it
the pelvic joint and externally rotate the femur. The muscle is inner- can resist). Various factors determine the ultimate loading of
vated by the ventral branches of the lumbar nerves and the femoral the system. Contraction of the abdominal musculature, especially

205
 10 The neck and back

of the rectus abdominis muscle, will tense the bow (i.e. flex the at the beginning of the stance phase and maximal at about mid-
back). The same effect will be achieved indirectly by retraction of stance of both thoracic and pelvic limbs (Buchner et al., 1996). The
the forelimbs or protraction of the hind limbs. The fusion of the acceleration signal recorded over the sternum showed also two
fibrous sheet of the gluteus medius muscle and the longissimus similar dorsoventral deviations corresponding to each half-stride
dorsi muscle as alluded to earlier may be of importance for the latter (Barrey et al., 1994).
mechanism. The string will be tensed (i.e. the back extended) by A complementary analysis of the walk on a treadmill using three-
protraction of the forelimbs and retraction of the hind limbs, but dimensional accelerometers fixed to the front of a saddle (Galloux
also by the considerable weight of the abdominal organs. The latter et al., 1994) showed that, at the walk, the amplitude of movement
effect is nicely illustrated by the appearance of many old brood was higher in the vertical axis than in the transverse or longitudinal
mares with their often very hollow backs. Although often misun- axes; rotation around the transverse axis (pitching motion) was
derstood, activity of the epaxial musculature will have an extending higher than rotation around the longitudinal axis (rolling) and
effect on the back as well, that makes the back hollower. Given the vertical axis (twisting or yaw). Furthermore, the twisting movement
fact that the vast majority of the musculature is located on top of was greater in the walk than in the other gaits.
the thoracolumbar vertebral column, contraction of this muscula- Head movement adaptations associated with a supporting fore-
ture will automatically lead to extension of the back. limb lameness induced by pressure on the hoof sole resembled
those of the trot: the amplitude of the dorsoventral oscillation
decreased during the stance phase of the lame limb and increased
Coordination and locomotion during the contralateral limb stance phase (Buchner et al., 1996).
Similarly, the maximum acceleration amplitude over the sternum
Considering the complete equine vertebral column, the movements was reduced during the stance phase of the lame forelimb (Barrey
of the neck and trunk and structural and functional description of et al., 1994). For induced hind limb lameness, elevation of the
the behavior of the axial muscles during different gaits and sport tuber sacrale was slightly reduced during the lame stance phase
exercises has been published previously (Denoix, 1987; Denoix & while head and withers movements were not significantly altered
Pailloux, 1996; Denoix, 1999a,b). At present, few in vivo studies (Buchner et al., 1996).
have been published. A kinematic method to evaluate back flexibil- Surface electromyographic (EMG) activity of neck and trunk
ity on standing horses was presented (Licka & Peham, 1998). Neck muscles was performed while horses walked on a treadmill or on
and back motion and coordination are very different according to hard surfaces, using skin-mounted surface electrodes (Fig. 10.8).
the gait and movement that are being performed and there is a need It was found that the left and right splenius muscles act simulta-
for better knowledge of the nature and amplitude of in vivo equine neously before the landing of each forelimb (or during the second
vertebral mobility. Neck and back flexibility and active mobility are half of the opposite forelimb stance phase) to elevate the head and
essential to the accomplishment of sport exercises and are the neck and facilitate forelimb protraction by the elongated brachio-
subject of constant observation and concern for trainers and riders cephalicus muscle. The sternocephalicus muscles had reciprocal
(Colborne et al., 2001; Cassiat et al., 2004). activity to splenius muscles and act before and during the first half
The behavior of the osteoarticular components of the equine of each forelimb stance phase, the longissimus dorsi muscles act
thoracolumbar vertebral column mainly based on in vitro investiga- during the intermediate part of each hind limb stance phase to
tions has been performed (Denoix, 1999a). Three major move- facilitate propulsion (Peham et al., 2001a,b; Licka et al., 2009), and
ments take place in the equine intervertebral joints (Jeffcott, 1980; the rectus abdominis muscles do not show any significant EMG
Townsend et al., 1983; Clayton & Townsend, 1989a; Denoix, activity during the walk (Tokuriki et al., 1997; Robert et al., 1998),
1999a): and this is correlated to the limited vertical acceleration of the
• Flexion (ventral bending inducing a dorsal convexity) and abdominal visceral mass due to the lack of a suspension phase in
extension (dorsal bending inducing a ventral convexity) this gait.
movements, occurring in the median plane, around a A previous study performed with intramuscular fine-wire elec-
transverse axis trodes on four horses with and without a rider reported that the
• Lateral bending (lateroflexion) to the left and right sides,
developed in the horizontal plane, around a dorsoventral axis
• Left or right axial rotation (left or right deviation of a Stance phases
vertebral body with respect to the following one) around a LH
longitudinal axis. LF
RH
RF
Translational movements of minor amplitude take place in a verti-
cal transverse plane:
• Vertical shearing (dorsoventral translation) is associated with Sp
flexion and extension movements SC
• Transverse shearing (left to right displacement) is associated
with lateroflexion and rotation.
LD
Longitudinal compression and tension occur in some parts of the RA No electromyographic activity
intervertebral joints for every major movement of the vertebral
column. 80 0 20 40 60 80 0 20
Percentage of stride activity

Walk Period of electromyographic activity;

right and left side muscles are active simultaneously
Kinematic and accelerometric studies of head and trunk movement
at walk and trot on treadmills have been performed (Barrey et al., Fig 10.8  Electromyographic activity of neck and trunk muscles at walk
1994; Buchner et al., 1996). At the walk, the vertical displacement, (1.8 m/s) on a hard surface. Limbs: LF, left forelimb; RF, right forelimb; LH, left
velocity and acceleration of the head, withers and tuber sacrale hind limb; RH, right hind limb. Muscles: Sp, splenius; SC, sternocephalicus;
showed a sinusoidal pattern with two similar oscillations during LD, longissimus dorsi; RA, rectus abdominis.
each stride. The height of the withers and tuber sacrale was minimal Courtesy of Céline Robert, Maison-Alfort, France.

206
 Function

multifidus lumborum muscle was active in the intermediate part of stance phase, the neck and trunk rotate downward to become closer
the stance phase of the ipsilateral hind limb and the obliquus to the horizontal. During the suspension phase, extension between
externus abdominis muscle had intermittent activity (Tokuriki the trunk and neck occurs (Fig. 10.9).
et al., 1991). On the standing horse, in the neck region, the splenius A three-dimensional in vivo kinematic study of flexion and exten-
muscle showed some activity, while the sternocephalicus and bra- sion movements of the thoracolumbar spine was performed at the
chiocephalicus muscles were silent (Tokuriki & Aoki, 1991). Clinical trot in 13 sound horses using five skin markers placed on the median
observation shows that it is probably during walking that the mobil- plane of the back over the 6th and 13th thoracic spinal processes
ity of the thoracolumbar vertebral column is most diversified, with as well as at the thoracolumbar, lumbosacral and sacrocaudal junc-
association of rotation and lateroflexion combined with limited tions (Pourcelot et al., 1998; Audigié et al., 1999). This study showed
movements of flexion and extension. that maximal thoracic extension occurs near midstance (Fig. 10.10).
This passive movement is produced by the visceral mass inertia. The
Trot maximal thoracolumbar extension occurs in the second half of each
stance phase, followed by the maximal lumbosacral extension at the
Kinematic and electromyographic data of neck and trunk motion end of the stance phase (Audigié et al., 1999).
and coordination have been established on sound horses with a The maximal thoracic flexion takes place during the swing phase
three-dimensional kinematic analysis system allowing simultane- and is concomitant with an elevation of the neck (Figs 10.10, 10.11).
ous recording of the left and right sides of the horse (Audigié et al., It is followed by the maximal thoracolumbar flexion. Finally, the
1999; Robert et al., 1998; Pourcelot, 1999). Lame horses have also maximal flexion of the lumbosacral junction occurs at the end of
been investigated using kinematic analysis (Buchner et al., 1996; the swing phase at the time of maximal protraction of the hind limb
Audigié et al., 1999). (Audigié et al., 1999).
The vertical displacement versus time of the head, withers and The influence of the neck orientation on the thoracic spine dor-
tuber sacrale at trot shows a sinusoidal pattern (Buchner et al., soventral mobility has been investigated in cadaver specimens
1996) with two symmetrical oscillations during a stride. The general (Denoix, 1987). The results showed that elevation of the neck facili-
orientation of the neck and trunk and their alignment vary during tates thoracic flexion (Fig. 10.11).
the stride (Fig. 10.9). With respect to the tuber sacrale position, the An accelerometric device fixed over the sternum area by an elastic
withers elevate during the first and intermediate parts of each diago- girth (Barrey et al., 1994) showed that, for horses trotting on a
nal stance phase (upward rotation of the trunk) and descend during treadmill, the dorsoventral acceleration curves had a deviation cor-
the last part of the stance phase as well as during the suspension responding to each half-stride. The height of the dorsoventral accel-
phase (downward rotation of the trunk). Additionally, the croup eration signal was linearly correlated with speed. Longitudinal
presents wider dorsoventral movements (lowering and elevation) accelerations were less repetitive than dorsoventral ones; their mag-
than the withers. nitude increased with the increasing speed of the gait. Analysis of
In sound trotting horses, the head position is highest in the first the trot on a treadmill using three-dimensional accelerometers fixed
half of each diagonal stance phase (Vorstenbosch et al., 1997). to the saddle (Galloux et al., 1994) showed that the linear move-
During the major part of the stance phase (Fig. 10.9), the neck ments along the three axes (longitudinal, transverse and vertical)
rotates downward and becomes closer to the horizontal, with the were similar in amplitude. Rotation around the transverse axis
head reaching its lowest point before the end of this phase. During (pitching motion) was smaller in comparison with the other two
the suspension and beginning of the following stance phase the gaits (walk and canter).
neck becomes more oblique (Fig. 10.9), and the head is elevated The surface electromyographic activity of neck and trunk muscles
again (upward rotation of the neck). at trot has been investigated on hard track surfaces, with or without
In early stance both the neck and trunk undergo an upward rota- riders, as well as on a treadmill (Tokuriki & Aoki, 1991; Tokuriki
tion; during the middle part of the stance phase the angle between et al., 1991, 1997; Robert et al., 1998, 1999, 2001, 2002; Licka et al.,
the neck and trunk flexes (Fig. 10.9) and during the last part of the 2004).

Left diagonal Right diagonal

stance phase stance phase
8 5

+
6 2
Trunk angle (o)
Neck angle (o)

4 0

2 -2
0 20 40 60 80 100
Time (% of stride)
Fig 10.9  Neck and trunk orientation in sound horses guided by hand on hard track at trot (3.1 m/s). Markers are placed over the zygomatic arch, withers
(6th thoracic spinal process) and tuber sacrale. The straight line between the first two markers represents the neck orientation; the line between the last two
markers represents the trunk orientation. The figure shows the angle variations occurring between these lines and the horizontal plane during one stride. A
positive angle indicates an upward angulation of the cranial marker relative to the caudal one as shown in the image on the right. Blue line: neck angle (°)
with respect to the horizontal plane; green line: trunk angle (°) with respect to the horizontal plane; diagonal stance phase: full fore hoof–ground contact.

207
 10 The neck and back

Left diagonal Right diagonal

stance phase stance phase
8 167 T18 T13 T6

Thoracic angle (o)

6 165 +
Neck angle (o)

4 163

2 161
0 20 40 60 80 100
Time (% of stride)

Fig 10.10  Coordination between the neck orientation and thoracic spine dorsoventral movements in sound horses guided by hand on a hard track at trot
(3.1 m/s). Markers are placed over the zygomatic arch, withers (6th thoracic spinal process), 13th and 18th thoracic spinal processes. The straight line
between the first two markers represents the neck orientation. The thoracic angle is drawn between the last three markers. The figure shows the angle
variations occurring between the neck orientation and the horizontal plane as well as the thoracic angle variations during one stride. For the neck, an
increasing angle indicates an upward rotation; the thoracic angle increases during flexion of the back and decreases during extension. Blue line: neck angle
(°) with respect to the horizontal plane; green line: thoracic angle (°); diagonal stance phase: full fore hoof–ground contact.

Accelerometer (LF) LF landing LF landing

12 2E+00
6 3
5 1E+00
4 5E+01
2 3 -0E+00
1
5 6 -5E+01
4 -1E+00 LF RH LF RH
Splenius stance phase stance phase
2E+04
1E+04
0E+00
2 -1E+04
1 2 5 1 6 5
6
3 4 -2E+04
3 4
Sternpcephalicus
1E+04
5E+05
Fig 10.11  Schematic representation of the trunk movements and neck 0E+00
positions during half a stride with reference to the vertebral column -5E+05
dorsoventral peaks of mobility. 1, Maximal thoracic flexion; 2, maximal -1E+04
thoracolumbar flexion; 3, maximal lumbosacral flexion; 4, maximal thoracic
extension; 5, maximal thoracolumbar extension; 6, maximal lumbosacral Longissimus dorsi
1E+04
extension. The same numbers indicate the respective positions of the neck
5E+05
and limbs (left diagonal).
0E+00
-5E+05
-1E+04
Neck muscle activity and neck orientation Rectus abdominis
3E+04
Comparison of neck muscle activity and neck orientation demon-
1E+04
strates that the splenius muscle acts before and during the first part
of the stance phase of each forelimb (Figs 10.12, 10.13) to limit -1E+04
lowering of the neck (antigravitational activity). As mentioned by -3E+04
Tokuriki and Aoki (1991), muscle activity tends to be higher during
the stance phase of the contralateral forelimb than for the ipsilateral Fig 10.12  Electromyographic activity of two neck and two trunk muscles
one (Fig. 10.12). The sternocephalicus muscles have a reciprocal (left side) at trot on a treadmill (4 m/s) during three consecutive strides. An
activity during the suspension phase to control neck elevation (Figs accelerometer was placed on the left forelimb. LF, left forelimb; RF, right
10.12, 10.13). The brachiocephalicus muscle activity is high during forelimb; LH, left hind limb; RH, right hind limb.
the later part of the ipsilateral forelimb stance phase and during the Courtesy of Céline Robert, Maisons-Alfort, France.
suspension phase (Fig. 10.13) to achieve protraction of the ipsilat-
eral forelimb; this muscle is inactive during most of the stance phase
of the ipsilateral forelimb in order to avoid limiting forelimb
propulsion.

208
 Function

RF-LH LF-RH RF-LH Accelerometer

1E+00
Landing Landing
5E-01
Sp
BC 0E-00

NAtural position
SC -5E-01 LF RH LF RH
Splenius stance phase stance phase
3E-04
LD
RA 1E-04

-1E-04
80 0 20 40 60 80 0 20
Percentage of stride activity -3E-04

Muscular activity: Splenius

3E-04
High Left side
Moderate Right side 1E-04
Low Both sides
-1E-04
None

Neck lowered
-3E-04
Fig 10.13  Electromyographic activity of neck and trunk muscles at trot
Accelerometer
(4 m/s) on a treadmill (mean periods for five horses). Limbs: LF, left forelimb; 1E-00
RF, right forelimb; LH, left hind limb; RH, right hind limb. Muscles: Sp,
splenius; BC, brachiocephalicus; SC, sternocephalicus; LD, longissimus dorsi; 5E-01
RA, rectus abdominis.
Courtesy of Céline Robert, Maisons-Alfort, France. 0E-00
-5E-01
Effect of neck orientation on neck muscle activity
Lowering of the neck using the reins induces a reduction of splenius Fig 10.14  Influence of the neck position on the electromyographic activity
muscle activity during the first half of the stance phase of each of the left splenius at trot (4 m/s) with reference to the left front limb.
Courtesy of Céline Robert, Maisons-Alfort, France.
diagonal (Fig. 10.14). These data can be explained by the lower
amplitude of the neck displacement as well as by increased tension
in passive anatomical structures that support the head and neck,
such as the nuchal ligament.
the back, markers indicative of neck and trunk movements were
Trunk muscle activity and thoracolumbar placed over the zygomatic arch, shoulder joint and tuber coxae (Fig.
movements 10.15). The head and neck were positioned with side reins and the
horses were filmed under competitive conditions at a left lead canter
Comparison of trunk muscle activity and thoracolumbar move- from outside of the circle. The neck and trunk angle curves showed
ments shows that the rectus abdominis muscle acts during the that the cranial part of the neck and trunk becomes lower during
stance phase (Figs 10.12, 10.13) to limit the passive thoracolumbar the intermediate part of the support phase (downward rotation)
extension induced by the visceral mass acceleration. The longissi- and elevates (upward rotation) during the leading forelimb stance
mus dorsi muscles act at the end of each stance phase and during phase and the suspension phase (Fig. 10.15). The orientation
the suspension phase (Figs 10.12, 10.13) to induce lumbosacral changes of the neck angle occur before those of the trunk. An analy-
extension and facilitate hind limb propulsion as well as to stabilize sis of the canter using three-dimensional accelerometers fixed to the
the thoracolumbar spine as it flexes. saddle (Galloux et al., 1994) showed higher amplitudes of motion
A previous study (Tokuriki et al., 1991) showed that the multifi- than for the other gaits (walk and trot), especially along the longi-
dus lumborum and longissimus lumborum were active before and tudinal and vertical axes. Rotation around the transverse axis (pitch-
after lift-off of each hind limb, with the activity being higher for the ing motion) and around the longitudinal axis (rolling) were greater
ipsilateral hind limb. According to this study the obliquus externus than in the other gaits, while twisting around the vertical axis was
abdominis and rectus abdominis had roughly reciprocal activity to lower. EMG activity of trunk and neck muscles was recorded with
these epiaxial muscles. In another study (Tokuriki et al., 1997), the surface electrodes at the canter (Figs 10.16, 10.17). Recording of
longissimus lumborum was thought to play a role in limiting lateral neck muscle activity showed that the splenius muscles were
bending of the trunk during symmetrical gaits. active once in the stride cycle, during the trailing diagonal stance
Dorsoventral movements of flexion and extension have been phase (Fig. 10.16). These muscles limit neck lowering and cause
quantified in sound horses (Audigié et al., 1999; Pourcelot et al., the neck to extend during the leading forelimb stance phase and
1998). The maximal range of vertical displacements occurred near the suspension phase. The sternocephalicus muscles had a reci­
the 13th thoracic vertebra and, with respect to the tuber sacrale, procal activity and were active from the end of the leading forelimb
reached an average value of 1.5 ± 0.2 cm. In clinical cases with stance phase to the first part of the trailing hind limb stance
intervertebral osteoarthrosis a significant reduction of this range of phase (Fig. 10.16). The brachiocephalicus muscle moving the trail-
motion was observed (Audigié et al., 1999). ing forelimb was mainly active during the stance phase of the
leading forelimb. The brachiocephalicus muscle on the side of
the leading forelimb was mainly active during the suspension
Canter phase. This left to right dissociation is correlated to the asymmetry
Kinematic analysis was performed in nine high-level vaulting horses of the gait and to the respective chronology of each forelimb
cantering at 4–5 m/s on a circle. As no markers could be placed on protraction.

209
 10 The neck and back

Right hindlimb
stance phase

Right forelimb
stance phase
45 -5

35 -15 +

Trunk angle (o)

Neck angle (o)

25 -25

15 -35
0 20 40 60 80 100
Time (% of stride)

Fig 10.15  Neck and trunk orientation in a sound horse being lunged with side reins at left lead canter (4.5 m/s) on soft ground (sand) in competitive
conditions (vaulting gymnastics). Markers are placed over the zygomatic arch, point of shoulder and tuber coxae. The straight line between the first two
markers represents the neck orientation; the line between the last two markers represents the trunk orientation. The figure shows the angle variations
occurring between these lines and the horizontal plane during one stride. An increasing angle indicates an upward rotation; a decreasing angle indicates a
downward rotation (blue line). Neck angle (°) with respect to the horizontal plane; (green line) trunk angle (°) with respect to the horizontal plane.

LF Left frontlimb
RF LH
RH RF LH
RH
Sp
LF LF
Left hindlimb RF-LH RF-LH
BC RH RH
SC

LD Left longissimus dorsi

RA

OI
Right longissimus dorsi
80 0 20 40 60 80 0 20
Percentage of stride duration

Muscular activity:
Left rectus abdominis
High Left side
Moderate Right side
Low Both sides
None
Right rectus abdominis
Fig 10.16  Electromyographic activity of neck and trunk muscles at ridden
left lead canter (6 m/s) on hard ground (mean values on three horses).
Limbs: LF, left forelimb; RF, right forelimb; LH, left hind limb; RH, right hind
limb. Muscles: Sp, splenius; BC, brachiocephalicus; SC, sternocephalicus; LD,
longissimus dorsi; RA, rectus abdominis; OI, obliquus internus abdominis. Left obliquus internus abdominis
Courtesy of Céline Robert, Maisons-Alfort, France.

Right obliquus internus abdominis

Fig 10.17  Hoof accelerometer signals and electromyographic activity of

trunk muscles at ridden left lead canter (6 m/s) for three consecutive strides.
LF, left forelimb; RF, right forelimb; LH, left hind limb; RH, right hind limb.

210
 Neck dysfunction

The longissimus dorsi muscles were active once per stride at the Subsequently, there is transition from the lumbar back to the tho-
end of the leading forelimb stance phase, during the suspension racic back that leads to a progressive (smoother) rotation of the
phase and during the trailing hind limb stance phase (Figs 10.16, chest vertebrae. During diagonal support phases (e.g. left forelimb,
10.17). The function of these muscles was to prepare the landing of right hind limb) in the trot and canter, the forelimb pushes the chest
both hind limbs and to extend the trunk before forelimb landing. vertebrae in one direction while the hind limb rotates the pelvis
The rectus abdominis muscles had reciprocal activity during the and lumbar back in the opposite direction. This twisting occurs at
support phase of the diagonal limbs; they act to support the visceral about the same time as the maximal extension of the back. The
mass and to initiate thoracolumbar flexion during the leading fore- strings of the back (abdominal muscles) restrain these movements.
limb stance phase (Fig. 10.18) and the suspension phase. Bursts of The above-mentioned movements of the back were found to be very
activity were observed simultaneously in the longissimus and rectus symmetrical. The relationship between the movement of the limbs
abdominis muscles on the side opposite to the leading forelimb to the movement of the back is paramount to the symmetrical
during the stance phase of the trailing hind limb. bending of the back. The shape of the movement partly depends
The obliquus internus abdominis muscle corresponding to the on the gait. Flexion–extension movement has a double sinusoidal
leading forelimb was active during most of the intermediate part of motion pattern at the walk and trot, but a single sinusoidal pattern
the support phase (Figs 10.16, 10.17), contributing to visceral at the canter. Lateroflexion has the form of a single peak and trough
support and trunk lateroflexion. The opposite obliquus internus at all gaits, as has axial rotation. This has to do with the symmetry
abdominis muscle had a burst of activity at the end of the support of the gait and the effect of hind limb placement on spinal
phase and beginning of the swing phase. kinematics.
During the canter, the use of intramuscular wire electrodes has Variability within the same horse is limited for flexion–extension
shown that the multifidus lumborum and longissimus lumborum and axial rotation (6–8%), but considerably more for lateroflexion
are active during the swing phase of the trailing hind limb. The (8–18%). The variability between horses is larger, as could be
rectus abdominis and obliquus externus abdominis have reciprocal expected, and the same applies here: lateroflexion may vary as much
activity (Tokuriki et al., 1991). as 16–25% between individual horses, which is considerably more
than the variation in the rotation around the other two axes
(10–16%).
Walk vs trot vs canter
The horses worked at a walk, trot and canter on a treadmill and Neck dysfunction
movement of the back and limbs was measured with a complex
series of high-speed cameras. The studies have demonstrated far less
movement of the back at trot than in the other two gaits. At the Since the publication of Ricardi and Dyson (1993) there has been
walk, the range of motion for flexion–extension is fairly constant a growing interest in the head and neck region as a potential source
for vertebrae behind T10 (approximately 7°), lateral bending is of pathological locomotor disturbances and the availability of EMG
greatest in the more cranial thoracic vertebrae and in the pelvic (Giovagnoli et al., 1998; Tokuriki et al., 1999; Van Wessum et al.,
segments (values >5.6°), but less in the lumbar region between T17 1999; Wijnberg et al., 2004) and MEP techniques (Nollet et al.,
and L5 (<4°). Axial rotation increases gradually from 4° at T6 to 2003a,b, 2004) has proven to be a great asset in investigating this
13° at the tuber coxae. At the trot the range of flexion–extension region. These studies proved that a neuropathy in combination with
for all vertebrae does not exceed 2.8–4.9°, lateral bending is even myelopathy and facet joint arthrosis can, like in humans, also result
less (1.9–3.6°). Axial rotation at this gait is in the order of 3°. At in neck pain, clinical signs of incoordination or front leg lameness,
the canter flexion–extension movement is substantially larger stringhalt and disability in horses (Schnebel et al., 1989; Moore
(maximal range 15.8 ± 1.3°). Lateral bending is maximally 5.2 ± et al., 1992; Dunbar et al., 2008; Levine et al., 2008). Moreover,
0.7° and axial rotation 7.8 ± 1.2°. An important relationship was different head and neck positions (HNP) have become under
demonstrated on the movement of the lumbar back and pelvis. debate in the equine community, especially in the disciplines of
Functionally, this part of the back essentially rotates (twists) as a dressage and show jumping, as they might have an effect on normal,
single unit is readily observed when the walk is viewed from behind. physiological function too (Jeffcott et al., 2006; McGreevy et al.,
2010; Sleutjens et al., 2012; Wijnberg et al., 2010).
The general aim of training the equine athlete is to achieve a
well-balanced horse in harmony with its rider that is able to show
its individual gait qualities (Weishaupt et al., 2006; Heuschmann,
2007). The head and neck position is believed to be an important
aid in achieving this goal, as it influences fore and hind limb kinet-
ics and kinematics, as well as thoracolumbar movement (Denoix &
Audigie, 2001; Rhodin et al., 2005; Gómez-Alvarez et al., 2006;
Weishaupt et al., 2006; Rhodin et al., 2009; Waldern et al., 2009).
The use of specific head and neck positions in training is not
uncontested. The recent discussion in the international dressage
world focuses on the extremely flexed head and neck position ‘hyper-
flexion’ or ‘Rollkur’, or ‘low, deep and round’ (Jeffcott et al., 2006;
Heuschmann, 2007; Van Bostel et al., 2009; Wijnberg et al., 2010),
which is believed by some trainers to be a useful tool to improve
the gymnastic ability that is asked from today’s high performance
dressage horses and is rejected by others on the basis of presumed
negative effects on equine welfare (McGreevy et al., 2010).
Some recent studies on locomotion demonstrated that, although
a flexed head and neck position induced an increase in range of
motion in the lumbar back in the unridden horse and would imply
Fig 10.18  Leading forelimb stance phase during left lead canter while a more animated use of the hind limbs, a longer stride length at
lunging on a circle. See the rectus abdominis contraction (red arrowheads) walk and a more equal weight distribution between fore and hind
inducing lumbosacral flexion (white arrow). limbs (Gómez-Alvarez et al., 2006), this effect could not be

211
 10 The neck and back

reproduced while horses were ridden (Rhodin et al., 2009). An model, developed from measured cervical vertebral dimensions,
extremely elevated neck, however, caused an increase in extension calculated centers of rotations, determined intervertebral angle limi-
of the thoracic and lumbar back in the unridden as well as in the tations, and earlier reported nuchal ligament properties, and subse-
ridden horse (Gómez-Alvarez et al., 2006; Rhodin et al., 2009). quently whether the model could be used to estimate the relative
Furthermore, an extremely elevated neck was shown to affect the differences in loading between the objectively quantified HNPs
functionality of the locomotor apparatus much more than an (Sleutjens et al., 2010).
extremely low neck by an increase in peak vertical force in the Goals were to determine the effect of dorsoventral flexion and
forelimbs, which is, among other factors, a potential risk factor for extension on intervertebral foramina dimensions in the equine cer-
the development of injuries (Biau et al., 2002; Weishaupt et al., vical spine (Sleutjens et al., 2010, 2012), to evaluate the effect of
2006; Waldern et al., 2009). different HNPs on single fiber (SF) and quantitative electromyo-
Dysfunction of the vertebral column can originate from pathology graphic (EMG) examination directly after moderate exercise (Wijn-
of bone, cartilage, joints, ligaments, tendons, muscles or nerves and berg et al., 2010), to evaluate the effect of different head and neck
therefore the clinical signs can be rather diverse. Several clinical signs positions on intrathoracic airway resistance and corresponding arte-
can occur and vary from lameness, to less specific signs such as rial blood gas values during a regular training exercise (Sleutjens
paresis, muscle atrophy, decreased range of motion, expressions of et al., 2009) and to investigate the effect of head and neck position
pain such as resistance to perform or to position (Sleutjens et al., (HNP) on behavior and cortisol levels in blood and salivar in
2012). Pathology in the nervous system can also result in clinical Warmblood riding horses (Dierendonck et al., 2010). This study
signs of upper motor neuron disease such as dysmetria, hypermetria, used 7 healthy, base-level trained Dutch Warmblood riding horses
hypometria and deficits in proprioception. A thorough and detailed (sex: five mares, two geldings, age: 10.3 ± 3.6 years (mean ± SD),
lameness exam and neurological exam are necessary to localize the height at the withers: 161.2 ± 1.4 cm, weight: 531 ± 47.3 kg), with
area of interest that can be subjected to additional diagnostic no history of respiratory disease, cardiovascular disease, musculo-
methods such as local blocks, radiology, ultrasound, scintigraphy skeletal or neuromuscular disorders. Radiographic and ultrasound
and last, but not least, EMG needle examination. A study using examination of their spinal column showed no abnormalities.
fone-wire EMG to localize a problem (Wijnberg et al., 2004) showed Videos were taken at the walk on a straight line in five different
that lesions in the segment L3–S3 were overdiagnosed and lesions head and neck positions, which were accomplished using side rains
in the C1–T2 region were underdiagnosed without (semi) quantita- (Fig. 10.19). The five head and neck positions, of which HNP1,
tive EMG examination and Motor Unit Action Potential (MUP) HNP2, HNP4, and HNP5 were identical to those used in an earlier
Analysis. In addition, this study showed that if generalized muscle experiment (Weishaupt et al., 2006; Gomez Álvarez et al., 2006;
atrophy was present, it was more often associated with a generalized Rhodin et al., 2009; Waldern et al., 2009), were defined as follows
neuropathy than a myopathy. If the problem was defined as lame- (Fig. 10.20):
ness of unknown origin, most often the diagnosis was a neurogenic
problem in the segment of C1-T2. Remarkably signs of hypermetria
• HNP1: Free, unrestrained, neutral position.

or the hind limb or hind limbs appeared to be the result of a cervical

• HNP2: Neck raised; bridge of the nose around the vertical.

lesion rather than a pelvic lesion.

Quantitative analysis of motor units recorded with needle EMG
allows detection of a disturbance of the functionality of the neuro-
muscular system and appeared to be able to diagnose myopathy or
neuropathy in an early stage of the disease (Wijnberg et al., 2000,
2003a,b). The absence of raised activity of muscle enzymes appeared
not to exclude myopathies, and is explained by the fact that not all
myopathies will coincide with leakage of muscle enzymes, at that
time point or indeed at any time. This makes this technique interest-
ing since in contrast to diagnostic imaging techniques it provides
information about the functionality of the region examined (Wijn-
berg et al., 2004). Since normative values have recently become
available on two paraspinal neck muscles, this technique can con-
tribute further insights into neck pathology and its relevance (Wijn-
berg et al., 2002, 2011).

Effect of different head and neck positions on

physiology and performance during exercise
A range of positions of the head and neck (HNP) have been evalu-
ated and reported during exercise (Gómez-Alvarez et al., 2006; Sloet
et al., 2006; Van Breda, 2006; Weishaupt et al., 2006; Rhodin et al.,
2009; Waldern et al., 2009; Wijnberg et al., 2010). However, these
studies contain some subjective elements and neck positions that
have not yet been fully objectively quantified. Furthermore, little is
known about the biomechanical consequences of these head and
neck positions on the loading of the cervical vertebrae.
Therefore, a study designed in order to test whether five in vivo
head and neck positions commonly used during training and com- Fig 10.19  From the captured video-frame at LF midstance, the coordinates
petition in sport horses, could be objectively quantified in the sagit- of the reflective markers were defined, from which the five head and neck
tal plane from straightforward angles and distances using a positions (HNPs) were objectively defined using the angles between
home-video camera and standard anatomical landmarks (Elgersma C1–T6-Hor (‘angle no. 1’), CF-C1–T6 (‘angle no. 2’), C1–CF-Vert (‘angle no. 3’),
et al., 2010). The second objective was to test whether the loading BN′–BN-Vert (‘angle no. 4’) and the horizontal distance between CF and TS
on the cervical vertebrae could be calculated from an anatomical (‘distance A’) and the vertical distance between CF and PS (‘distance B’).

212
 Neck dysfunction

Fig 10.20  Different HNPs evaluated in this study shown at the moment of LF midstance in an anatomical simulation model.

213
 10 The neck and back

Fig 10.20  Continued

• HNP4: Neck lowered and considerably flexed; nose pointing accustom the horses to the experimental set-up, they were trained
toward the chest. in the different head and neck positions on the lunge for at least 3
• HNP5: Neck raised and considerably extended; bridge of the weeks.
nose in front of the vertical. Spherical reflective markers were glued to the skin over the dorsal
• HNP7: Neck lowered and flexed; nose pointing towards the spinous process of Th6, the wing of the atlas (C1), the rostral part
carpus. of the facial crest (CF), the suprascapular tubercle (TS) and the
lateral styloid process of the radius (PS). Two additional markers
In addition, they were evaluated by an international dressage team were placed on the horse’s bridle, while the bridge of the nose (BN,
to check for realistic and correct interpretation, and thus considered BN′) could be identified in every video frame. The known distance
as being currently representative and commonly used in training between these two bridle markers was used as a calibration refer-
and competition. HNP7 was included because two interpretations ence for the video-frame analysis. A home-video camera was set on
of the hyperflexion, ‘Rollkur’ or ‘low, deep and round’ training posi- the line A-C perpendicular to the long side of the riding arena,
tion were found to exist among riders (HNP4 and HNP7). To where the horses were evaluated.

214
 Neck dysfunction

Video analysis was done using custom software program to deter- The horses were acclimatized to the experimental neck positions
mine the marker coordinates. Head and neck positions were during a training period of at least 3 weeks, until they accepted the
described using four angles and two distances. For each HNP the induced neck positions. Acceptance was defined as walking in the
following angles and distances were measured (Fig. 10.19): required position with no pressure on the bit. A standardized exercise
test was performed namely: trot1 for 10 min at a heart rate (HR) of
• (C1–T6-Hor): the angle between the wing of the atlas (C1),
101 ± 8 bpm, canter for 4 min at a HR of 128 ±10 bpm, trot2 for 5 min
the dorsal spinous process (T6) with the horizontal: angle no.
at a HR of 104 ± 7 bpm and walk for 5 min at a HR of 73 ± 8 bpm.
1. This angle has a negative value when the atlas is lower than
Mean consecutive difference (MCD) of single fiber potentials and
T6.
motor unit action potential (MUP) variables (amplitude, duration,
• (CF-C1–T6): the angle between the rostral part of the facial
area, turns and phases) were recorded in each position directly after
crest (CF), the wing of the atlas (C1) and the dorsal spinous
exercise at rest in a fixed HNP using commercial EMG equipment.
process of (T6): angle no. 2.
Muscle enzyme activity was measured before and 4, 6 and 24 h after
• (C1–CF-Vert): the angle between the wing of atlas (C1), the
exercise (Wijnberg et al., 2010).
rostral part of the facial crest (CF) with the vertical: angle no.
Intrathoracic pressure was indirectly measured during exercise
3. This angle is negative when the facial crest is caudal to the
using a portable intra-oesophageal pressure (Pes) measuring device.
atlas.
Arterial blood samples were taken immediately after every test inter-
• (BN′–BN-Vert): bridge of nose with the vertical: angle no. 4.
val from a catheter placed into the transverse facial artery. Arterial
This angle is negative value when the bridge of the nose is
oxygen pressure (PartO2), oxygen saturation (SartO2) and carbon
behind the vertical.
dioxide pressure (PartCO2) values were corrected for core tempera-
• (CF–TS): the horizontal distance between the rostral part of the
ture (Sleutjens et al., 2009). Continuous telemetric ECG recordings
facial crest (CF) and the supraglenoid tubercle (TS): distance A.
were made using the Televet system in order to determine or exclude
• (CF–PS): the vertical distance between the rostral part of the
cardiac arrhythmias that potentially could occur as a result of
facial crest (CF) and the lateral styloid process (PS): distance
hypoxia.
B. For each HNP, 5 video captures were taken per horse on
The horses performed an exercise test in each (HNP), which
the left and right sides with the camera perpendicular to the
consisted of warming-up at trot (15 min), canter (4 min), trot2
line of motion. Measurements were made with the left front
(10 min), walk (5 min) and cooling down. External influences
limb (LF) at midstance with the left metacarpal bone vertical
were prevented and the horses were videoed. Afterwards, predefined
(Figs 10.19, 10.20). The four angles (angle no. 1, 2, 3 and 4)
behavioral elements were scored by the same person in The
and the two distances (A and B) were measured and the mean
Observer® program. Blood and salivary samples were collected at
and the SEM of the group of seven horses were calculated.
8.00 am, prior to and directly at the end of the exercise test, 30 and
For every HNP digital L/M radiographs were made of the cervical 60 min after the exercise test and at 8.00 pm, in order to determine
vertebrae C2–C7 of one horse (horse no. 1), while it was standing basal cortisol levels and the effect of exercise in the different HNPs.
square with the head and neck in the sagittal plane to determine Several behavioral expressions were recorded such as:
the in vivo longitudinal angles between the cervical vertebrae (C2–
C7). Cervical angles were given a negative value when the body of • Head toss: The horse attempts to move or moves the head in
a quick forward-upward motion.
the cranial vertebra sloped in regards to the body of its successive
caudal vertebra. • Head shake: The horse attempts to shake or shakes its head in
a quick left to right motion.
From an additional Warmblood horse (horse no. 8), which
was euthanized for other reasons than musculoskeletal or neuro- • Head pull: The horse lets its head hang in the reins and pulls
the reins forward.
muscular disorders, the ex vivo longitudinal angles between the
cervical vertebrae (C2–C7) were measured in five different positions • Shying: The horse shies away from an object or side of the arena.
(neutral, extension 20° and 40°, flexion 20° and 40°), which • Blowing: Non-pulsated sound produced by forceful expulsion
of air through the nostrils.
according to Sleutjens et al. (2009) represents respectively the in
vivo HNP angles: HNP 1, HNP 2 and 5, and HNP 4 and 7. • Snorting: Sound of forceful exhalation through the nostrils
with an audible flutter pulsation, while the horse attempts to
The (ex vivo) CT images (horse no. 8) were overlaid on the (in
lower the head (Dierendonck et al., 2010).
vivo) digital radiographical images (horse no. 1) using imaging
processing software to facilitate accurate and representative cervical Elgersma et al. (2010) showed that mean angles were significantly
intervertebral angle measurements to feed the model. All data were different between the 5 HNPs for C1–T6 with the horizontal ((C1–
transferred into commercially available modeling software. The Th6-Hor): angle no. 1) and for CF–C1 ((C1–CF-Vert): angle no. 3)
model was based on the inertial properties data of Buchner et al. and bridge nose ((BN′–BN-Vert): angle 4) with the vertical (p <
(1997), Gelmann et al. (2002) and Gelmann and Bertram 0.05). For angle 2 ((CF-C1–Th6)) all HNP, except for HNP2 and 4,
(2002a,b). were significantly different (p < 0.05). The horizontal distance from
The first step in building the model was to measure the cervical CF to TS (distance A) was different only between HNP1 and all
vertebral dimensions from the ex vivo specimen (horse no. 8). For others, while the vertical distance from CF to PS (distance B) was
this, digital images were produced of the vertebrae C1–Th2 using significantly different between all 5 HNPs (p < 0.05).
a CT scan (Sleutjens et al., 2009) and a rectangle was projected Sleutjens et al. (2010) showed that an extension of 20° causes a
over the vertebral body with the cranial head and the caudal tail decrease in intervertebral foramina dimensions at segment C5/
as longitudinal, sagittal reference length. Thus, using free available C6–C7/T1. A decrease in intervertebral foramina dimensions is
software the dimensions of the vertebrae were determined. The fol- caused by 40° extension at segment C4/C5–C7/T1. Flexion causes
lowing values were found: C1, 4.4 cm; C2, 13.4 cm; C3, 11.0 cm; no significant change in foramina dimensions, except for the length
C4, 11.1 cm; C5, 10.6 cm; C6, 9.8 cm; C7, 8.3 cm; T1, 5.9 cm; T2, at segment C6/C7 at 20° flexion.
5.9 cm; T3, 5.6 cm; T4, 5.3 cm; T5, 5.1 cm; T6, 5.1 cm. Cervical CT Wijnberg et al (2010) showed that mean MCD in all HNPs were
images of six cadaver equine cervical spines were taken in the neutral significantly higher than in HNP1 (p < 0.05) of which HNP4 was
position, extension (20° and 40°) and flexion (20° and 40°). highest (p < 0.05) with 39 compared to 30 in HNP2. HNP5 and 7
Images were reconstructed in the oblique plane perpendicular to with 25 MCD were lower than the MCD in HNP 2 and 4 (p < 0.05).
the long axis of each intervertebral foramen from the C2/C3 to C7/ Odds ratio for potential pathological MCD and conduction block
T1 level. Foramina height and length were measured in each posi- was 13 : 6 in HNP4 compared to HNP1 (p < 0.05), but there are no
tion and compared to the neutral position (Sleutjens et al., 2010). reference values for horse patients yet. Number of turns and

215
 10 The neck and back

duration followed the same pattern as MCD indicating temporal MUP variables indicated that head positions affect neuromuscular
dispersion of the MUP and a non-synchroneous arrival of the indi- functionality. The meaning of this elevation of MCD, in a range that
vidual muscle fiber action potentials that contribute to the MUP. in humans is considered to be pathological, remains unclear as long
Only LDH activity increased at all time points in HNP4 at time as there is no information from patients with known nerve damage.
points 4 (p < 0.05), 6 (p < 0.05) and 24 (p < 0.05) and in HNP5 Therefore, it would be interesting to measure patients with cervical
and HNP7 at 4 (p < 0.05) (Wijnberg et al., 2010). vertebral pathology (Wijnberg et al., 2010). Especially HNP4 leads
Sleutjens et al. (2009, 2012) proved that using HNP1 as reference, to a higher elevation in muscle enzyme activity that caused the
inspiratory intrathoraric pressure (IP) became more negative during horses to develop a dynamic obstruction of the respiratory tract,
trot1 (each HNP) (p < 0.05), canter and trot2 (HNP4) (p < 0.05) and evidenced by an increase in intrathoracic pressure difference and a
walk (HNP4, HNP5) (p < 0.05). Intrathoracic pressure difference more negative pressure during inspiration. However, the authors did
(IPΔ) increased during trot1 and walk (each HNP) (p < 0.05) and not find a change in arterial oxygen or carbon dioxide pressure,
canter and trot2 (HNP4) (p < 0.05). PartO2 increased during canter respectively. In contrast, they found a significant increase in arterial
(HNP4) (p < 0.05). PartCO2 (mmHg) and SartO2 (%) did not oxygen pressure in HNP4 during canter. This may be explained by
change significantly. the fact that the horses were by no means working at maximal
Van Dierendonck et al. (2010) demonstrated that head tossing capacity, leading to a lot of reserve capacity that prohibits a measur-
and head shaking were increased during exercise in HNP2 (p < 0.05) able decline or increase in arterial oxygen or carbon dioxide pressure,
and head pulling in HNP2 (p < 0.05). During detaching of the side respectively. Further research is needed to elucidate the origin of
rains, head shaking was increased in all HNPs (p < 0.05), but most this effect on arterial blood gases and to evaluate the effect of an
in HNP2, holding the head low was increased in HNP5 (p < 0.05). extremely flexed head and neck position (HNP4) in patients with
During cooling down, holding the head low was increased in HNP5 upper/lower airway disease (Sleutjens et al., 2009, 2012). HNP2 and
(p < 0.05), HNP2 (p < 0.05) and HNP4 (p < 0.05). During exercise, HNP5 cause more resistance behaviors during exercise and attach-
helper encouragement was increased in HNP2, HNP4 and HNP7 ment of the side rains. After detaching the side reins (HNP5) and
(p < 0.05), helper slow down was increased in HNP2 and HNP5 during cooling down after HNP5, 2 and 4, the horses displayed a
(p < 0.05), shying was decreased in HNP7 (p < 0.05), blowing was lower head and neck position, possibly to compensate pain in the
increased in HNP2 and HNP4 (p < 0.05), snorting was increased in exhausted neck muscles. Cortisol levels are needed to further com-
HNP7 (p < 0.05) and swishing of the tail was increased in HNP2 plete the knowledge of the effect of HNP on the mental state of the
(p < 0.05). During attaching and detaching of the side rains, tail equine athlete (Van Dierendonck et al., 2010).
swishing was increased in respectively HNP2 (p < 0.05) and HNP2,
7 (p < 0.05, p < 0.05).
Back dysfunction

Discussion of measurements from the neck Introduction

This study objectively quantified and measured differences between Dysfunction of the back is an important reason for poor perfor-
five in vivo head and neck positions commonly used during training mance in the horse. Clinical history and examination are important
and competition in sport horses. Mean angles of C1–T6 with the features in the diagnosis of back problems. However, the interpreta-
horizontal (C1–T6-Hor) and of CF–C1 with the vertical (C1–CF- tion of these findings is subjective and varies due to experience,
Vert) in combination with the horizontal distance from CF and TS tradition and personal bias. Therefore, ancillary aids such as radi-
(CF–TS) and the vertical distance from CF to PS (CF–PS) were ography, scintigraphy, ultrasonography, thermography and local
shown to differ between HNPs. anesthesia are commonly used in the diagnosis of back problems.
The in vivo ‘hyperextended’ HNP5 (ex vivo, 40° extension) had The relationship between these ancillary aids and pain or dys-
the largest, most positive (C1–T6-Hor), (C1–CF-Vert) and (BN′– function is debatable as many fully functioning riding horses can
BN-Vert) angles and the longest vertical (CF–PS) distance; in HNP2 display pathological changes that apparently do not diminish their
(ex vivo, 20° extension) the (C1–CF-Vert) angle was close to zero, athletic performance. There is a clear need to document the func-
the in vivo ‘neutral’ HNP1 had the largest (CF-C1–T6) angle and the tional and dysfunctional back objectively to assist our clinical exam-
longest horizontal (CF–TS) distance; HNP4 (ex vivo, 20° flexion) ination and to determin of the health status of the equine back
the smallest (C1–CF-Vert) and (BN′–BN-Vert) angle; the in vivo (Haussler et al., 2001, 2006a,b; De Heus et al., 2010).
‘hyperflexed’ HNP7 (ex vivo, 40° flexion) was the HNP with the Back pathology, with pain and reduced motion, has been identi-
smallest, most negative (C1–T6-Hor) angle and the shortest vertical fied as an important cause of poor performance (Jeffcott, 1975,
(CF–PS) distance (p < 0.05). 1980; Denoix, 1998). A general geometric approach to the biome-
In conclusion from the angles of C1–T6 with the horizontal and chanics of the back in relation to vertebral pathology has been
of CF–C1 with the vertical in combination with the vertical distance presented (Rooney, 1982). Kissing spines, vertebral spondylosis
from CF to PS, we were able to objectively quantify the head and (Jeffcott, 1980; Haussler et al., 1999) and osteoarthrosis of the
neck position in a particular horse during training and competition dorsal synovial intervertebral joints between the articular processes
from home-video images. This enables routine (standard video) (Denoix, 1998; Girodroux et al., 2009) are the most significant
monitoring, including timing, of the applied HNPs during training vertebral lesions responsible for back discomfort and clinical mani-
and competition in the practical field situation. The ex vivo findings festations in horses.
suggest that an increase in extension of the equine cervical spine Thus, to comprehensively understand the different, complex
could cause compression of the cervical nerve root at the level C4/ movements of the back horses in daily use, horses in regular train-
C5–C7/T1. This information contributes to a more complete under- ing, horses with complaints of poor performance with the back as
standing of different training techniques with respect to the head the cause, and of horses with obvious back pain were objectively
and neck position (Sleutjens et al., 2010) but it also raises questions evaluated for their function and dysfunction.
with regard to the findings in the study on neuromuscular function
measured using stimulated single fiber EMG. Based on these findings
it cannot be ruled out that especially HNP4, leads to a rather large
diminished MCD compared to the control position, and that factors
Kinematic examination of normal horses
other than compression might play a role in this induced delay. For In the study of a group of normally functioning riding horses (John-
example, traction on the nerve in this flexed position. MCD and ston et al., 2004), the clinical examination of the back consisted of

216
 Back dysfunction

visual inspection, palpation and observation of movement when in markers, a camera system consisting of six cameras was used. The
hand, on a lunge, and ridden. Individual limbs were flexed to reg- system is based on passive markers and infrared cameras. Data were
ister reaction and lameness. Only horses that were clinically sound captured during steady state locomotion during walking and
and did not demonstrate abnormal reaction to palpation, flexion trotting. Movement of the back was described in terms of
tests nor abnormal movement (behavior) during the riding or extension–flexion, lateral bending, lateral excursion and axial rota-
lunging tests, were used. The clinical examination in the asymptom- tion. The range and symmetry of movement were calculated. Addi-
atic, normal horse illustrated the importance of the normal appear- tionally, characteristics of the horses’ back, height at withers, and
ance of well functioning horses. Still, in the majority of the horses weight were determined. The characteristics that were derived from
some reaction to palpation was apparent. Furthermore, the location marker data were length of the thoracic back (horizontal distance
of reaction to palpation seemed to be dependent on the main use T6–T17), length of the lumbar back (horizontal distance L1–L5),
of the horses. This could be explained by the differences in training length of whole back (horizontal distance T6–S3), width (i.e. the
these horses were subjected to, in their respective discipline. Dres- distance between the tubera coxae), and curvature of the mid-
sage horses and show jumpers in this study were also ridden with thoracic back.
different types of saddles, which might influence how the back is Walk and trot are symmetric gaits and the movements of the back
loaded. The dressage rider normally sits ‘deeper’ on the horse, while in these trained animals were also highly symmetric. Normal func-
the rider of a show jumper will sit more in a forward seat. However, tion and symmetry are apparently highly related both in gait and
this should not lead to a pathological reaction upon palpation, as movement in the back. The only difference in conformation that
then there might be something wrong. Horses that were reactive to dressage horses had longer lumbar backs than jumpers. Dressage
palpation of the back, did not always show impaired function, horses demonstrated more lateral movement than jumpers. Further-
dysfunction or lack of performance. more, dressage horses demonstrated less symmetry in the mid-
The horses were trained four times on a treadmill (Johnston thoracic region than did the jumpers at the walk. Suppleness is
et al., 2004). The dorsal spinous processes were identified by palpa- thought to be associated with increased back length and may there-
tion and spherical, reflective markers were placed over the identified fore explain a conformational preference toward the dressage horse.
landmarks (Fig. 10.21). For collecting the position data of the Because the musculature of the back is passive during the walk,

T6
L5 S3
T10 T13 T17 L1 L3
1

2 4

Fig 10.21  Marker placement on back and limbs. Lumbar vertebrae 1, 3 and 5 (L1, L3, L5); sacral vertebra 3 (S3). 1, Proximal spina scapula; 2, lateral collateral
ligament of the metacarpophalangeal joint; 3, cranial part of the trochanter major of the femur; 4, lateral collateral ligament of the metatarsophalangeal joint.
Reprinted from De Cocq, P., van Weeren, P.R., Back, W., 2004. Effects of girth, saddle and weight on movements of the horse. Equine Vet. J. 36, 758–763, with permission of the Equine Veterinary
Journal.

217
 10 The neck and back

lateral movement may be greater and less symmetric in the dressage reactions like bolting or rearing, tail swishing, unruliness, rapid
horse. Gender was found to be an important factor in the lateral caudal movement of the ears or stiff, jerky movements.
movement of some of the segments of the back while working at Horses with back pain show an aberrant pattern of the movement
the trot. Mares demonstrated greater lateral excursion (displace- of the back, which, with the aid of objective measurement tech-
ment of the back to each side) in the cranial back than did geldings, niques, makes it possible to classify a horse with back dysfunction.
and greater movement symmetry. The functional significance is not It is reasonable to believe that a horse with a sore back will try to
known, however mares show increased occurrence of bony changes move in a way that, if possible, may alleviate the pain. Apparently,
on the underside of the vertebrae (spondylosis) when compared to this will best be accomplished by a stiffer dorsoventral (flexion-
geldings and stallions. Perhaps excessive lateral movement is a con- extension) movement in the caudal thoracic back and at the thora-
tributing factor in the degenerative processes between the vertebral columbar junction, at both the walk and trot. The abnormal lateral
structures of this part of the mid-thoracic back that is most suscep- movement seen at the withers and decreased rotation (AR) of the
tible to spondylosis. Age was also a determinant in the movement pelvis, results in a side-to-side swaying walk as seen from behind.
of the back. Older horses had decreased flexion–extension at the Presumably, the symptomatic horse has altered the neuromuscular
transition between thoracic and lumbar back at the trot. Perhaps control of the walk and trot to cope with back pain. Acquired
wear and tear due to age in this area of high shearing is an important pathological limitations could also be an initial source of the
factor in the pathogenesis of certain disorders. These trained riding problem and therefore crucial factors to the decreased range of
horses demonstrated highly symmetrical movement of the back, movement (ROM) of the back.
which is in contrast to research horses used in an earlier study Apparently, flexion of the back is reduced to limit the relative
(Faber et al., 2000, 2001a,b,c, 2002). Training and perhaps quality displacement of the individual segments of the thoracolumbar
of horse may influence movement symmetry and function of the back, perhaps due to excessive muscular activity as aggravated by
back. Factors such as gender and age, thus are important in the nociception. The normal movement of the back is more controlled
function of the back. The effect of gender is intriguing and requires by muscle activity at the trot than at the walk, where the movement
further investigation, while age is more intuitive as it is likely related is more passive with greater amplitude for the lateral and twisting
to the wear and tear of use. Perhaps the most influential factor (not movements. This may be the reason why the ROM for the AR of the
related to use) for the function of the back is the perceived confor- pelvis is decreased in a horse with back pain at the walk, but not at
mation for the respective disciplines (Johnston et al., 2002). Further the trot.
investigation on the relationship of back conformation and func- The reduction in the ROM for the flexion and extension (FE) and
tion to orthopedic health is warranted as to determine the most AR movements has earlier been observed in another study – a case
appropriate conformation and function for the different sporting study of one horse with increased responsiveness to palpation of
disciplines. the lumbar and sacral back (Faber et al., 2003). The shorter stride
There were no statistically significant associations between the length observed in the horses with back pain at the walk, coinciding
clinical, kinematic, radiographic and scintigraphic results (Erichsen with the decreased FE movement of the back, is well in accordance
et al., 2003a,b, 2004). The age, gender, use, weight and height of with the findings of Jeffcott (1980) and Faber et al. (2003). It is also
the horses did not influence, the presence or absence of clinical or in agreement with the positive relationship between the pro- and
radiographic and scintigraphic changes nor the classification of pos- retraction of the hind limbs and the FE movement of the back that
sible outliers in the range of movement and the symmetry of move- has been established in clinically sound horses at the walk (Faber
ment. However, there was a statistically significant positive et al., 2000) and trot (Faber et al., 2001). In the present study, the
association between the total number of times a horse was graded expected reduction in stride length in the patients was not seen at
as a possible outlier in the range of movement and the symmetry the trot though. The explanation for this is not obvious, but it is
of movement and increased isotope uptake (IRU) during a scinti- possible that the muscle activity in the hind limbs was altered and
graphic examination (Erichsen et al., 2003a,b, 2004). The number may have influenced the stride length. The slight difference in stride
of times a horse was graded as a possible outlier in the range of velocity at the trot is not likely to have caused the decrease in the
movement was significantly associated with the presence of coincid- ROM (Robert et al., 2001).
ing radiographic changes, IRU and lunging abnormalities respec- Since all horses showed muscle soreness on palpation of the back,
tively. Deviation from the normal function of the back is related to and more than half of them had pathological skeletal reactions, it
clinical, radiological and scintigraphic abnormalities. Consequently, is reasonable to believe that this was the main reason for the
the objective determination of function may indeed suggest sub- decreased dorsoventral FE movement at both gaits and the changed
clinical structural and functional abnormalities. Kinematic studies lateral movement at the walk.
on the treadmill (Licka et al., 2001a,b; Johnston et al., 2004; This study supports the use of objective measurements of the
Johnson & Moore-Colyer, 2009) and overground (Audigie et al., back kinematics as a valuable tool to help identify horses with back
1999; Gomez Alvarez et al., 2007, 2009) provide a viable means of dysfunction. Before using it more extensively it is necessary to
determining difficult performance problems that are related to the further evaluate the method, including measurements on patients
back. whose diagnoses can be confirmed, and long-term follow-up of
back patients after treatment.

Kinematic examination of horses with

back pain Differences between horses with and without
To study back dysfunction a group of riding horses with impaired
back pain
performance, altered movements during work and clinical back The studies of Wennerstrand et al. (2004) and Johnston et al.
pain were used (Wennerstrand et al., 2004). Horses were Warm- (2004) nicely show significant and objective differences between
blood riding horses between 5 and 15 years old. Only horses that horses with back pain and asymptomatic controls at both the walk
had not been treated for back pain, with the exception of rest and/ and trot. The differences are most striking at the thoracolumbar
or convalescence training, during the last 3 months were allowed junction, where the patients, in addition to the changes in the ROM,
to participate. Back pain was considered present if the horse showed also were less symmetrical for the FE movement at the walk. These
clear signs of pain/discomfort on palpation of the back and the findings indicate that the thoracolumbar junction is one part of the
reaction did not decrease with repeated palpation. Commonly, back especially predisposed to impairment (Denoix, 1998; Holm
horses demonstrating back pain reacted to palpation by adverse et al., 2006; Peham & Schobesberger, 2006). The most apparent

218
 Back dysfunction

criterion between horses with back pain and horses without prob- diagnosed back problem was much higher than in horses without
lems is their function. Despite many changes in the radiographs and such a problem (Landman et al., 2004).
scintigraphy, the horses in the healthy group managed to maintain The diagnosis of back pain is elusive and complex. Single factors
their normal work. Imaging techniques are important in the isola- such as conformation, lameness, equipment, and way of riding are
tion of horses with back problems, but do not seem to be predictive, likely to contribute to the clinical manifestation. Consequently,
at least at this point in time, in determining if horses will become function or dysfunction is ultimately the determining factor in the
dysfunctional. Encouraging, however, is a clear separation of diagnosis. Indiscriminate uses of imaging or palpation do not seem
patients from normal, as horses with back pain have more outliers to be accurate determinants of back pain. A complete clinical evalu-
than the fully functional horses. In other words, horses with back ation is paramount and in this study an objective measure of func-
pain move significantly differently from the fully functional horse tion may certainly contribute to an accurate diagnosis. Techniques
as determined from kinematic data. Thus, kinematic studies are in the diagnosis of back pain, such as local anesthesia, need to be
valuable in the definition of clinical and perhaps subclinical clas- evaluated to appreciate their correct usage and interpretation. It can
sification of horses with dysfunction of the back. be anticipated that, notwithstanding the limitations all kinematic
The results from the clinical, kinematic, radiographic and scinti- gait analysis systems suffer (Van Weeren, 2002), computerized anal-
graphic examinations are likely to overlap between horses with and ysis of spinal kinematics will become more popular as an aid in
those without back pain or back dysfunction. The sensitivity and diagnosis and to monitor recovery and/or the success of chosen
specificity of these, as well as other tests or examinations, depend therapies. Understanding back pain requires a fundamental under-
on the cut-off points set in order to determine the presence or standing of the normal function of the back and how normal
absence of a change. Cut-off levels of changes set so low, that intervention can affect its function. The biomechanical concept of
changes would be present in a majority of asymptomatic horses, the action of the back is crucial too for a good understanding of
would not be helpful in the evaluation of symptomatic horses. various riding techniques in dressage horses and how certain train-
Since mild changes were common in these studies (Erichsen et al., ing methods could affect the musculoskeletal system (Van Weeren,
2003a,b, 2004), they are not likely to be good indicators of clini- 2005). It should be emphasized, however, that this matter is
cally significant pathology in symptomatic horses. Therefore, mild extremely complex, as there are many factors that may influence
changes were grouped together with no changes to allow some performance. Of these, use of the horse, the quality of the rider and
variation in the lower range of the classification system and to make the tack are among the most important.
the criteria less sensitive. The level from which the changes should
be considered significant, is also important if changes are early signs
of later clinical problems. This can only be determined by perform- The effect of head and neck position on the
ing a longitudinal study. movement of the back
It may be that the examinations used in this study are not the
optimal or final combination for the detection or description of Head and neck positions have been standardized by the use of side
symptomatic or asymptomatic horses in terms of back pain or reins (Fig. 10.22). Flexion–extension movement of the back at walk
dysfunction. Data on the ultrasonographic and thermographic or with the head and neck in the high position was significantly lower
quantitative electromyographic appearance of the back of the as compared to the movement with the head and neck in a free or
asymptomatic horse have not been published extensively and, few, in a low position at the caudal part of the back (Biau et al., 2002;
if any, other practically useful tests would contribute significantly Weishaupt et al., 2006; Gómez-Alvarez et al., 2006; Rhodin et al.,
in a general evaluation of the equine back. 2009; Waldern et al., 2009). The same tendency was seen with the
A major advantage of the protocol used in these studies is the head and neck in the low position as compared to the free position
broad and complementary approach of the examinations. The but probably because of the low number of observations, no sig-
detailed clinical examination is quintessential to identify and clas- nificant differences were found. At trot only flexion–extension
sify the signs of reduced or impaired performance. The kinematic movement of segment T17 was lower at the high position.
examination is functional, objective, and relatively easy to perform. At walk the stride length was shortest with the head in the high
The radiographic and scintigraphic examinations are indicative of position but this was not seen at trot where the stride length was
specific anatomic lesions within the relative context. In the study of constant. It is, however, known that the movement of the back is
asymptomatic horses, none of the examinations by themselves is related to stride length. Horses with longer strides extend and flex
sufficiently specific to describe these horses as asymptomatic. Kine- their backs in the caudal saddle region to a greater extent at walk
matics and radiographic and scintigraphic results support the clini- (Johnston et al., 2002). Faber et al. (2000, 2001) found that an
cal examination suggesting that these horses are asymptomatic. increasing stride length of the hind limb was correlated with an
The examination of the equine back remains a complicated and increasing flexion–extension range of movement for most of the
interesting challenge but with information gained in this study used vertebrae both in the walk and in the trot. This is in accordance
for comparison, the combined results from clinical, kinematic and with our findings but the stride length was also associated with the
radiographic and scintigraphic examinations of the back in symp- head and neck position.
tomatic horses can be used to decrease the risk of excessive false The different results at walk and trot may be due to the different
positive results. In addition, muscle pathology of the back is not characteristics of the gaits (Roepstorff et al., 2002; Bystrom et al.,
reflected by an increase in muscle enzyme activity and therefore 2006). At walk the horse moves the head and neck to a greater
needle EMG can often be helpful in detecting underlying muscle extent as compared to trot, where the head and neck position is
fiber loss. more constant. When the head and neck are fixed, as with side reins,
It can be concluded that the equine back is a very complex struc- this impedes the natural movement at walk more than at trot. At
ture that takes a central position in the entire equine musculoskel- trot, speed influences movement of the back (Robert et al., 2001),
etal system and hence can be decisive for performance. A good but in this study the speed at trot was constant.
knowledge of how the back works is therefore essential. The close It seems probable that the decreased flexion–extension move-
interrelationship between the limbs and the back (and neck) is ment of the back in the high head position is directly caused by the
often underestimated. In a recent survey a group of orthopedic shorter stride length, which in turn is caused by the head and neck
patients and a control group of animals presented for a prepurchase position. The lateral bending movement of the lumbar back with
exam were subjected to a full lameness examination as well as a full the head and neck in the high position was at walk significantly
back examination, irrespective of their eventual complaints. In that lower as compared to the movement with the head and neck in a
study it appeared that the prevalence of lameness in horses with a free position, which was similar for flexion–extension. On the other

219
 10 The neck and back

on the movement of the asymptomatic equine back, and the useful-

ness of kinematic studies as an objective and quantitative tool in
evaluating local analgesia in clinical practice.
The kinematics of the back in ten clinically sound horses were
measured on two occasions at the walk and trot before and after
infiltration with mepivacaine (a local anaesthetic) and infiltration
with sodium chloride (saline or water and salt solution that is
similar to water in the body) around the spaces between T16 and
L2. The kinematics were compared between the two occasions
before infiltrations and before and after each infiltration. The range
of motion for dorsoventral flexion–extension of the back was sig-
nificantly increased in all measured segments other than T10 at the
walk, as was the lateral bending (LB) at T10, L3 and L5 after infiltra-
tion of mepivacaine. For lateral excursion (side to side movement
of the back) the total movement increased at all measured seg-
ments. At the trot, the only affected segment was L3, where the
infiltration with mepivacaine decreased the ROM for FE. After infil-
tration of sodium chloride the ROM for FE increased at T13 and
T17 at the walk. Lateral bending and lateral excursion were not
affected, while AR increased at the walk. At the trot, LB increased at
L3 and L5. The symmetry of movement was not affected by infiltra-
tion of either mepivacaine or sodium chloride at the walk and only
at L5 for LB at the trot after infiltration of sodium chloride. Local
infiltration of aesthetic solution affects the movement of the back
in asymptomatic horses, which is important to consider when using
it in clinical practice. Thus, the effect of local analgesia of the back
could be qualitatively and quantitatively evaluated from kinematics.
Diagnostic infiltration of local anesthetic solution affects the func-
tion of the back also in the asymptomatic horse, which must be
considered when interpreting the use of this clinical aid in assessing
clinical cases of back dysfunction.

General conclusions
Fig 10.22  The six different head positions measured in the study. From
right to left and top to bottom, the positions are designated free, FEI like Research on the function and dysfunction of the neck and back has
position, FEI like position but head slightly behind the vertical, deep been of great academic and public interest. The popularity of the
overrolled or ‘Rollkur’ position, the extremely high and stressed position, the research is due to the fundamental understanding of the central roll
relaxed low position. of the neck and back in the riding horse, but as well as the lack of
in-depth insight into the function of the neck and back. Restricting
and restraining the position and movement of the head and neck
hand the thoracic region of the back had a significantly larger lateral alter the movement of the back. The flexion–extension movement
bending movement. The lower movement in the lumbar back may of the caudal back, with the head and neck in the high position
have been due to the shorter stride length and the larger movement at the walk, has the most pronounced reduction; most likely
in the thoracic region of the back can be compensation or the direct depending on the significantly shorter stride length. There are only
cause of the head and neck position. small difference between long reins and side reins with regard to
The axial rotation was significantly lower with the head in a high the movement of the back at the walk and trot. Consequently, long
position as compared to the free and low position at walk. For axial reining and free movement (with or without side reins) on a tread-
rotation at the walk, there is a high degree of synchronization mill are similar. Considering the intention and normal use of long
between the vertebrae T13 through the pelvis; only T10 moved reining, proper technique is thus essential as the positioning of the
independently and there is no significant correlation with the stride head (use of hand) is central in the movement of the horse.
length. For pro- and retraction of the hind limb there is a positive Movement and forces are not necessarily similar in all cases as
correlation with the stride length. In this experimental set-up we demonstrated by the use of draw reins. Similarly, more extreme
did not have the possibility to encourage the horses forward and movement does not induce similar and more extreme differences
thereby affect the hind limbs. If a constant stride length could be in forces. To induce a shift in weight (though only slight) from the
achieved, for example by long reining, together with variable head forelimb to the hind limb draw reins with normal reins are required.
and neck positions the causal relationship between the head Extreme bending of the head and neck by use of the draw rein did
and neck position and the movement of the back could be proven not lead to the same result. As in long reining proper equipment
(Fig. 10.22). and technique when using for example draw reins is required to
achieve the intended result.
The head and neck positions (HNP) in sport horses, however, are
The effect of local analgesia in the movement under debate in the equine community, as they would interfere
with equine welfare (Jeffcott et al., 2006). These HNPs have, until
of the back recently, never been objectively quantified and no information used
Diagnostic infiltration of local aesthetic solution is commonly used to be available on their head and neck loading.
in cases of equine back pain. The evaluation is subjective and it is To quantify in vivo HNPs in horses and develop a model to esti-
not known how local analgesia affects an asymptomatic back. Holm mate loading on the cervical vertebrae in these positions, videos
et al. (2006) evaluated the effect of infiltration of local anesthetics were taken of seven Warmbloods at walk on a straight line in five

220
 General conclusions

Table 10.1  Mean ± SEM in vivo measured angles and distances defining different head and neck positions (HNP) commonly used in
training and competition evaluated in a group of horses (n = 7)

Measured values HNP1 HNP2 HNP4 HNP5 HNP7

Angle 1 −16.2 ± 2.1* 3.6 ± 2.0* −33.3 ± 1.2* 34.8 ± 2.0* −41.5 ± 1.7*
Angle 2 131.9 ± 2.0* 87.2 ± 1.9a
84.4 ± 1.3a 119.2 ± 3.8* 103.6 ± 2.6*
Angle 3 25.7 ± 1.7* 0.9 ± 2.5* −38.9 ± 1.1 a
63.9 ± 3.6* −27.8 ± 1.5*
Angle 4 11.9 ± 1.3* −12.8 ± 2.5* −51.7 ± 1.8* 47.2 ± 3.5* −41.0 ± 2.2*
Distance A 73.5 ± 3.4* 43.9 ± 1.7b 34.5 ± 1.3c 45.4 ± 2.1b 37.2 ± 1.7c
Distance B 60.2 ± 4.9* 89.8 ± 3.0* 46.7 ± 2.1* 144.8 ± 3.9* 18.6 ± 3.3*

*p < 0.05; pairs a, b and c are not significantly different.

Reprinted with permission from Elgersma et al. 2010.

positions, representing all HNPs during Warmblood training and neck position. The results on the cortisol test are needed to define
competition. Markers were glued at five anatomical landmarks. whether the horse does indeed feel more relaxed in these positions,
Two-dimensional angles and distances were determined from video or just does not have the ability to behave in an adverse manner.
frames for the five HNPs (p < 0.05). This is probably also influenced by the character of the indi-
Thus, Elgersma et al. (2010) was able to objectively quantify the vidual and its tendency to submissive behavior. A temporary
different head and neck positions by angles and distances, which decreased conduction velocity indicating a decreased neuromuscu-
were significantly different from each other (Table 10.1). Therefore, lar transmission is an unwanted, negative effect, although not
it was possible to induce repeatable and reproducible the desired per se a pathological effect, but the question remains how these
head and neck position in different experimental circumstances by values will be when some degree of facet joint arthrosis is present,
using side rains. Sleutjens et al. (2010) found from ex vivo data that thereby pressing to some extent on peripheral nerve roots, like
extension of the cervical spine could have an adverse effect on is experienced in humans (Yoo et al., 1992; Farmer & Wisneski,
neuromuscular functionality with the most profound effect on the 1994; Humphreys et al., 1998; Lu et al., 2000; Muhre et al., 2001;
segment C6–C7. The concommittant in vivo study of Wijnberg et al. Nuckley et al., 2002; Kitagawa et al., 2004; Ebraheim et al., 2006;
(2010) on neuromuscular transmission showed that compared to Morishita et al., 2006; Tanaka et al., 2006; Abbed & Coumans,
HNP1, each head and neck position had a potentially adverse effect 2007; Hubbard et al., 2008; Down et al., 2009). Nevertheless,
on neuromuscular functionality. The greatest adverse affect was seen trainers/riders insisting on using HNP4 in the base-level trained
in HNP4, though based upon reference values for humans. This horse for a longer period of time should at least realize the possible
effect could be even more profound in patients with cervical pathol- consequences. There are still several remaining questions, which
ogy. HNP 7, another interpretation of a flexed HNP from the need to be answered in order to complete our understanding of the
induced neck positions, had the least affect on neuromuscular func- effect of head and neck position on the well-being of the equine
tionality. This effect might be caused by another factor than a athlete. It would for example be interesting to repeat the study in
decrease in intervertebral foramina dimensions. Furthermore, espe- elite dressage horses, to see if there is a difference between horses
cially HNP4 leads to an increased intrathoraric pressure difference trained at different levels. Additionally, the training circumstances
and a more negative inspiratory pressure. These effects may be could be expanded. Although it could possibly pose a greater chal-
caused by a dynamic upper airway obstruction and are possibly lenge on standardization and safety, it would be interesting to
even more profound in patients with upper airway disease. Last but measure in the ridden horse or the horse working on a treadmill.
not least, HNP4 caused a shift to sympathetic dominance during Furthermore, besides the healthy horse, there are many patients
walk, based on heart rate variability, possibly indicating (exercise) with some sort of upper/lower airway disease or a more or less
stress. HNP2 and HNP5 caused more active resistance behavior advanced stadium of cervical arthrosis, which are still used in
during exercise and applying of the side rains. Cortisol levels are training/competition on various levels.
under examination to further complete the knowledge of the effect The following logical question is which head and neck position
of HNP on the mental state of the equine athlete, and the type of would be advisable. HNP1 is the natural position of the horse
stress detected using heart rate variability. during exercise. However, it is unreasonable to believe that in the
These results justify to conclude that training with an extreme opinion of trainers/riders this position would be sufficient to
flexed head and neck position (HNP4) has an effect on the healthy, prepare the horses for competition at any level. As stated above
base-level trained Warmblood riding horse during moderate exer- HNP4 is questionable as is HNP5 because of its effect on the back,
cise for a longer period (± 30 min) of time and during lunging locomotor apparatus and behavior (Gómez-Alvarez et al., 2006;
exercise. However, it remains a challenge to judge for another Rhodin et al., 2009; Weishaupt et al., 2006; Waldern et al., 2009).
subject than yourself to which extent this effect is necessarily nega- HNP2 is still the position desired during competition by the
tive. For example, we could argue that an organism needs to have national (KNHS) and international (FEI) organizations and pre-
some degree of stress in order to perform. However, the resistance ferred above HNP 4 for gymnastic training of a horse on the lunge
behavior we observed in HNP2 and HNP5 is not desirable during and for shorter times, although we demonstrated that it affected
training/competition. This behavior was not seen in HNP4 and neuromuscular transmission and increased resistance behavior
HNP7, our impression is that the horses showed more submissive during exercise. After all, HNP7 looks at a position which could be
behavior during these positions which is in agreement with the used during training, since round and deep posture is trained with
argument that riders feel that their horse is more under control and less effect measured so far. In the end, variation is probably the key
more sensitive to the riders input while trained in a flexed head and word while adapting the training strategy to the level of the

221
 10 The neck and back

individual horse, rider and training goal while being cautious with appreciate these pathological changes thus limits our ability to
extreme flexed head and neck positions in the base-level trained identify them as factors in the orthopedic function of the horse. The
Warmblood riding horse. All this information will be of potential back offers thus a considerable challenge for the clinician. Presently,
help in better understanding the occurrence, and eventually in the diagnosis of back pain is often through exclusion of sources of
improving treatment and prevention of locomotor disturbances in reduced performance. Back pain is thus a functional diagnosis and
sport horses, induced by suboptimal head and neck positions requires considerable experience to understand and properly
during training and competition. In this way, this approach may manage. The interaction between lameness and back dysfunction is
contribute to equine welfare and indirectly to promotion of the apparent in the clinics, though the interrelationship to the success-
equine industry (McGreevy et al., 2010). ful treatment of lameness and back dysfunction is not as apparent.
The back is a complex entity that provides stability and flexibility. In many cases horses need only to be treated for the lameness and
Normal movement is highly symmetrical and specific. The thoracic any symptom from the back is resolved. Not uncommon are lame-
back is an extension of the flexible neck to allow for rotation and nesses that are readily isolated but do not respond to therapy until
lateral bending (Stubbs et al., 2006). As the forelimb is loaded, the combined therapies for the lameness and back pain are initiated.
back will need to accommodate for the strut like system. The scapula Finally, some clinical cases are overtly lame and the lameness is first
will assumedly impose bending of the forward segments of the resolved (isolated) to the back. Subsequently, back pain is likely a
thoracic back. On the other hand, the stable lumbar back-pelvis symptom and a source of other injuries and performance problems,
complex transfers the propulsive energy of the hind limb to lift and as well as being a primary cause of locomotor disturbance. The
propel the rest of the body. Intersegmental movement of the lumbar information gained to date reveals that clinical and potentially
back is slight to provide the needed stability. Consequently, kinematic examinations are fundamental to the diagnosis of back
the lumbar back, with the exception of the joint between the last dysfunction. Aids such as radiology, scintigraphy, and diagnostic
lumbar vertebra and sacrum, acts as a functional unit transferring anesthesia are only helpful when the functional (clinical) diagnosis
the movement through the joints of the back and supported by is established. Knowing the characteristic way of walking can
the muscles and ligaments that surround these bony structures. The provide a strong indication for back dysfunction as indicated from
muscles of the lumbar back (longissimus dorsi) limit flexion while the kinematic studies. A protocol might become a basis for a clinical
the muscles of the abdomen limit extension of the lumbar and later examination of horses with performance problems that may be
thoracic back. A more complex system of muscles function about related to the back, and thus documentation for further studies. The
the thoracic back to aid in lateral movement and limit extension, correct evaluation of therapies and rehabilitation programs can also
while the large ligament of the neck causes extension or limits be established through this protocol. Potentially, a preferably lon-
flexion of the part of the back. In equine sports, engagement of the gitudinal study using a protocol can offer a golden standard of
hind limbs and manipulation of the head and neck are thus impor- diagnosis determination. However, many details on interrelation-
tant features in training and competition. Our use of the horse has ship of lameness and back pain, involvement of difficult areas of
a central role on the function of the back. Perception of conforma- clinical evaluation such as the lumbar back, and interaction of
tion, riding technique, equipment, and training will impinge on the riding/training/equipment/tradition affect the orthopedic health of
normal function of the back and consequently function of the horse the back are needed to successfully apply the information gained
as a whole. The manipulation of the head and neck alters the move- in this study. Because of their limited mobility and heavy muscles,
ment specific regions of the back. The use of reins is frequently used the axial regions of the horse are probably the most difficult to
to aid in this manipulation and thus education of the horse. Raising investigate with biomechanical techniques and diagnostic proce-
the head and neck extends and limits the amount of movement of dures. Nevertheless, in the recent years a lot of information has been
the back, while lowering the head and neck tends to flex the back. gained on the biomechanics of the neck and back, especially with
In the ridden condition, the back becomes extended due to the kinematic analysis and quantitative (needle) electromyographic
weight of the rider (De Cocq et al., 2004, 2006) and stride length studies (Wijnberg et al., 2004, 2008, 2010). However, detailed
decreases. Consequently, raising or lowering of the head and neck knowledge of the functional behavior of the different structures of
would exacerbate or mitigate this condition. Training with weights the axial regions is still incomplete. Further research is required to
may help to mobilize and strengthen the muscles of the back and understand the pathogenesis of the different pathological condi-
limbs to help prepare the horse for loading as engendered by riding. tions of the back as well as to identify the sources of pain in back
Intriguingly interesting and important to recognize is that move- problems. These data are essential for application of the different
ment of the limbs (as induced by increased weight) has also a sig- methods of physical therapy on equine athletes. Further investiga-
nificant effect on the back. The innate movement pattern, lameness, tions are needed to establish a rational basis for the training of sport
and shoeing practices can potentially induce changes that may affect and racehorses in order to develop their neck and back athletic
the normal function of the back. A direct relationship between capabilities as well as to identify physical exercises helpful for the
shoeing and movement of the back is difficult to appreciate. We management of back problems.
know that horses respond to similar farrier techniques similarly in Studies on the effect of lameness from limbs on the movement
the hoof capsule and bony segments of the distal limb (under the of the back found a changed posture and a reduced mobility of the
carpus and tarsus), but quite differently when taking into consider- caudal back (Gomez Alvarez et al., 2007, 2008). Perhaps the most
ation of the movement of the whole body. difficult anatomical region of the horse to understand, from a vet-
Draw reins used with normal reins resulted in lowering the neck erinary perspective, is indeed the caudal back. The caudal back is
and placing the head more vertical or behind the vertical and the lumbar back and the connection between the back and the
increased propulsion from the hind limbs. Interestingly, exaggera- pelvis. The caudal back is constructed to support and transmit the
tion of this condition with draw reins only did not exaggerate hind limb’s energy to the forelimb (Van den Bogert et al., 1994;
engagement of the hind limb; on the contrary the hind limbs Spadavecchia et al., 2002, 2004). Indeed, considerable pathological
reduced their driving function. Function of the horse (and back) changes are found in this part of the back at pathological examina-
is thus the complex interaction of rider-horse that is influenced tion (Jeffcott, 1980; Haussler et al., 1999) that would assumingly
by training, equipment, and tradition; this interaction having affect the function of the horse as a whole. We therefore need to
the most potential influence on the orthopedic health of the horse address this area of the horse in a methodical manner to better
(and back). The close proximity of these bony structures as well understand the changes to these structures that may be important
as other structures of the back is subsequently site for many to the normal function of the athletic horse.
pathological changes that may or may not limit performance Perhaps the single most important factor to the orthopedic health
through dysfunction/pain. The bulky muscles limit our ability to of the riding horse is the interaction with the rider. The weight of

222
 General conclusions

the rider is known to induce extension of the lumbar back and alter confused and indeed in some cases horses are being treated with
the way of moving. While we do not exactly know what these altera- methods that are not efficacious and perhaps injurious. A systematic
tions may result in the health of the horse, it is plausible to suggest approach in the evaluation of therapies and rehabilitations pro-
that if the horse is not properly prepared for this task than injury grams is required to assure the proper care of horses with back-
will assuredly ensue. In the studies with high level dressage horses related problems.
walking and trotting on a force measuring treadmill with and
without their normal rider we could measure the three-dimensional
movement of the rider and horse as well as the pressure from
the rider and saddle on the back of the horse, and the forces at all
four hooves simultaneously (Fig. 10.23). Follow-up studies in this
area would be aimed at the effect of rider position and education
on the movement and forces in the high-level dressage horse. The
effect of the saddle on the movement of the horse is very poorly
understood and is subject to considerable speculation. Recent
studies using modeling (Schlacher et al., 2004; Peham & Shobes-
berger, 2004, 2006) have also contributed considerably to our
present understanding, Considering the relationship of the rider
and horse, the saddle is supposed to be important in the orthopedic
health of the horse (De Cocq et al., 2004, 2006; Back, 2013)
(Fig. 10.24).
Having these measurements techniques available there is a
great need on the objective evaluation of different therapies
and rehabilitation programs on horse with dysfunction and/or
pain of the back. Currently, most of our therapies are based
on experience and conjecture (Marks et al., 1999; Xie et al., 2005;
Haussler et al., 2007; Van de Weerd et al., 2007; Gomez
Alvarez et al., 2008; Sullivan et al., 2008). Animal owners are often Fig 10.23  Horse ridden on the force measuring treadmill.

A B

C D

Fig 10.24  Pictures of all four situations: (A) unloaded; (B) with lunging girth; (C) with saddle; (D) with saddle and 75 kg lead.
Reprinted from De Cocq, P., van Weeren, P.R., Back, W., 2004. Effects of girth, saddle and weight on movements of the horse. Equine Vet. J. 36, 758–763, with permission of the Equine Veterinary
Journal.

223
 10 The neck and back

References
Abbed, K.M., Coumans, J.V., 2007. Cervical objectively monitor the effect of diagnostic method. Vet. Radiol. Ultrasound 44,
radiculopathy: pathophysiology, palpation by a physiotherapist in 699–706.
presentation, and clinical evaluation. warmblood horses. J. Equine Vet. Sci. 30, Erichsen, C., Eksell, P., Widstrom, C.,
Neurosurg. 60, S28–S34. 310–321. Roethlisberger Holm, K., Johnston, C.,
Audigié, F., Pourcelot, P., Degeurce, C., Denoix, Denoix, J.M., 1987. Kinematics of the Lord, P., 2003a. Scintigraphic evaluation of
J.M., Geiger, D., 1999. Kinematics of the thoracolumbar spine of the horse during the thoracic spine in the asymptomatic
equine back: flexion-extension movements dorsoventral movements: a preliminary riding horse. Vet. Radiol. Ultrasound 44,
in sound trotting horses. Equine Vet. J. report. Proceedings of the 2nd 330–338.
Suppl. 30, 210–213. International Conference on Equine Faber, M., Johnston, C., Schamhardt, H.C., van
Back, W., 2013. Back in the driver’s seat and Exercise and Physiology, San Diego, pp. Weeren, P.R., Roepstorff, L., Barneveld, A.,
the need for an objective evaluation of 607–614. 2001a. Basic three-dimensional kinematics
saddle fit. Vet. J. 195, 12–13. Denoix, J.M., 1998. Diagnosis of the cause of of the vertebral column of horses trotting
Barrey, E., Hermelin, M., Vaudelin, J.L., et al., back pain in horses. Proc. Conf. Equine on a treadmill. Am. J. Vet. Res. 62,
1994. Utilisation of an accelerometric Sports Med. Sci., Cordoba, pp. 97–110. 757–764.
device in equine gait analysis. Equine Vet. J. Denoix, J.M., 1999a. Spinal biomechanics and Faber, M., Johnston, C., Schamhardt, H.C., van
17 (Suppl.), 7–12. functional anatomy. Vet. Clin. North Am. Weeren, P.R., Roepstorff, L., Barneveld, A.,
Bergmann, C., 1847. Über die verhältnisse der Equine Pract. 15, 27–60. 2001b. Three-dimensional kinematics of
wärme-ökonomie der thiere zu ihrer grösse. Denoix, J.M., 1999b. Ultrasonographic the equine spine during canter. Equine Vet.
Göttinger Studien Abt. 1. evaluation of back lesions. Vet. Clin. North J. 33 (Suppl.), 145–149.
Biau, S., Couve, O., Lemaire, S., Barrey, E., Am. Equine Pract. 15, 131–159. Faber, M., Johnston, C., van Weeren, P.R.,
2002. The effect of reins on kinetic Denoix, J.M., Audigie, F., 2001. The neck and Barneveld, A., 2002. Repeatability of back
variables of locomotion. Equine Vet. J. back. In: Back, W., Clayton, H.M. (Eds.), kinematics in horses during treadmill
Suppl. 34, 359–362. Equine locomotion, 1st edn. W.B. locomotion. Equine Vet. J. 34, 235–241.
Buchner, H.H.F., Savelberg, H.H.C.M., Saunders, London, pp. 167–192. Faber, M.J., Schamhardt, H.C., van Weeren,
Schamhardt, H.C., Barneveld, A., 1997. Denoix, J.M., Pailloux, J.P., 1996. Anatomy P.R., Barneveld, A., 2001c. Methodology
Inertial properties of dutch warmblood and basic biomechanical concepts. In: and validity of assessing kinematics of the
horses. J. Biomech. 30, 653–658. Physical therapy and massage for the horse. thoracolumbar vertebral column in horses
Manson Publishing Ltd, London, pp. based on skin-fixated markers. Am. J. Vet.
Buchner, H.H.F., Savelberg, H.H.H.C.,
21–63. Res. 62, 301–306.
Schamhardt, H.C., et al., 1996. Head and
trunk movement adaptations in horses Down, S.S., Henson, F.M.D., 2009. Faber, M.J., Schhamhardt, H.C., van Weeren,
with experimentally induced fore- and hind Radiographic retrospective study of the P.R., Johnston, C., Roepstorff, L., Barneveld,
limb lameness. Equine Vet. J. 28, 71–76. caudal cervical articular process joints in A., 2000. Basic three-dimensional
the horse. Equine Vet. J. 41, 518–524. kinematics of the vertebral column of
Byström, A., Roepstorff, L., Johnston, C., 2006.
Dunbar, D.C., Macpherson, J.M., Simmons, horses walking on a treadmill. Am. J. Vet.
The effect of draw reins on limb
R.W., Zarcades, A., 2008. Stabilization and Res. 61, 399–406.
kinematics. Equine Vet. J. 36, 452–456.
mobility of the head, neck and trunk in Faber, M.J., van Weeren, P.R., Schepers, M.,
Cassiat, G., Pourcelot, P., Tavernier, L., Geiger, Barneveld, A., 2003. Long-term follow-up
horses during overground locomotion:
D., Denoix, J.M., Degueurce, D., 2004. of manipulative treatment in a horse with
comparisons with humans and other
Influence of individual competition level back problems. J. Vet. Med. A Physiol.
primates. J. Exp. Biol. 211, 3889–3907.
on back kinematics of horses jumping a Pathol. Clin. Med. 50, 241–245.
vertical fence. Equine Vet. J. 36, 748–753. Dyce, K.M., Sack, W.O., Wensing, C.J.G., 2010.
Textbook of veterinary anatomy, fourth ed. Farmer, J.C., Wisneski, R.J., 1994. Cervical
Clayton, H.M., Flood, P.F., Rosenstein, D.S., spine nerve root compression. An analysis
Elsevier, London, p. 84.
2005. Clinical anatomy of the horse. of neuroforaminal pressures with varying
Mosby/Elsevier Publishers. Ebraheim, N.A., Liu, J., Shafiq, Q., Lu, J.,
Pataparla, S., Yeasting, R.A., et al., 2006. head and arm positions. Spine 19,
Clayton, H.M., Townsend, H.G.G., 1989a. 1850–1855.
Quantitative analysis of changes in cervical
Kinematic of the cervical spine of the adult Galloux, P., Richard, N., Dronka, T., et al.,
intervertebral foramen size with vertebral
horse. Equine Vet. J. 21, 189–192. 1994. Analysis of equine gait using
translation. Spine 31, E62–65.
Clayton, H.M., Townsend, H.G.G., 1989b. Elgersma, A.E., Wijnberg, I.D., Sleutjens, J., three-dimensional accelerometers fixed on
Cervical spinal kinematics: a comparison Van der Kolk, J.H., Van Weeren, P.R., Back, the saddle. Equine Vet. J. 17 (Suppl.),
between foals and adult horses. Equine Vet. W., 2010. A pilot study on objective 44–47.
J. 21, 193–195. quantification and anatomical modelling Gellman, K.S., Bertram, J.E., Hermanson, J.W.,
Colborne, G.R., Birtles, D.M., Cacchione, I.C., of in vivo head and neck positions 2002. Morphology, histochemistry, and
2001. Electromyographic and kinematic commonly applied in training and function of epaxial cervical musculature in
indicators of fatigue in horses: a pilot competition of sport horses. Equine Vet. J. the horse (Equus caballus). J. Morphol.
study. Equine Vet. J. 33 (Suppl.), 89–93. 42 (Suppl. 38), 436–443. 251, 182–194.
De Cocq, P., van Weeren, P.R., Back, W., 2004. Erichsen, C., Eksell, P., Holm, K.R., Lord, P., Gellman, K.S., Bertram, J.E.A., 2002a. The
Effects of girth, saddle and weight on Johnston, C., 2004. Relationship between equine nuchal ligament 1: structural and
movements of the horse. Equine Vet. J. 36, scintigraphic and radiographic evaluations material properties. Vet. Comp. Orthop.
758–763. of spinous processes in the thoracolumbar Traumatol. 15, 1–6.
De Cocq, P., van Weeren, P.R., Back, W., 2006. spine in riding horses without clinical signs Gellman, K.S., Bertram, J.E.A., 2002b. The
Saddle pressure measuring: validity, of back problems. Equine Vet. J. 36, equine nuchal ligament 2: passive dynamic
reliability and power to discriminate 458–465. energy exchange in locomotion. Vet. Comp.
between different saddle-fits. Vet. J. 172, Erichsen, C., Eksell, P., Widstrom, C., Berger, Orthop. Traumatol. 15, 7–14.
265–273. M., Holm, K.R., Johnston, C., 2003b. Giovagnoli, C., Pieramati, G., Castellano, G.,
De Heus, P., Van Oossanen, G., Van Scintigraphy of the sacroiliac joint region et al., 1998. Analysis of neck muscle
Dierendonck, M.C., Back, W., 2010. A in asymptomatic riding horses: (splenius) activity during jumping by
pressure algometer is a useful tool to scintigraphic appearance and evaluation of surface video-electromyography technique.

224
 References

Proc. Conf. Equine Sports Med. Sci., myelinated axonal degeneration near the Licka, T.F., Peham, C., Frey, A., 2004.
Cordoba, pp. 57–60. biomechanical thresholds for mechanical Electromyographic activity of the
Girodroux, M., Dyson, S., Murray, R., 2009. behavioural hypersensitivity. Exp. Neurol. longissimus dorsi muscles in horses during
Osteoarthritis of the thoracolumbar 212, 482–489. trotting on a treadmill. Am. J. Vet. Res. 65,
synovial intervertebral articulations: clinical Humphreys, S.C., Chase, J., Patwardhan, A., 155–158.
and radiographic features in 77 horses with Shuster, J., Lomasney, L., Hodges, S.D., Licka, T.F., Peham, C., Zohmann, E., 2001a.
poor performance and back pain. Equine 1998. Flexion and traction effect on C5-C6 Treadmill study of the range of back
Vet. J. 41, 130–138. foraminal space. Arch. Phys. Med. Rehabil. movement at the walk in horses without
Gomez Alvarez, C.B., Bobbert, M.F., Lamers, L., 79, 1105–1109. back pain. Am. J. Vet. Res. 62, 1173–1179.
Johnston, C., Back, W., van Weeren, P.R., Jeffcott, L.B., 1975. The diagnosis of diseases Lu, J., Ebraheim, N.A., Huntoon, M., Haman,
2008. The effect of induced hind limb of the horse’s back. Equine Vet. J. 7, 69–78. S.P., 2000. Cervical intervertebral disc space
lameness on thoracolumbar kinematics Jeffcott, L.B., 1980. Disorders of the narrowing and size of intervertebral
during treadmill locomotion. Equine Vet. J. thoracolumbar spine of the horse. A foramina. Clin. Orthop. Relat. Res. (370),
40, 147–152. survey of 443 cases. Equine Vet. J. 12, 259–264.
Gomez Alvarez, C.B., L’Ami, J.J., Moffat, D., 197–209. Marks, D., 1999. Cervical nerve root
Back, W., van Weeren, P.R., 2008. Effect of Jeffcott, L.B., Attock, A., Higgins, A., 2006. The impingement in a horse, treated by
chiropractic manipulations on the use of over bending (“Rollkur”) in FEI epidural injection of corticosteroids. J.
kinematics of back and limbs in horses competition. Report of the FEI Veterinary Equine Vet. Sci. 19, 399–401.
with clinically diagnosed back problems. and Dressage Committees’ Workshop. McGreevy, P.D., Harman, A., McLean, A.,
Equine Vet. J. 40, 153–159. Fédération Equestre International (FEI), Hawson, L., 2010. Over-flexing the horse’s
Gómez Alvarez, C.B., Rhodin, M., Byström, A., Lausanne. neck: A modern equestrian obsession? J.
Back, W., van Weeren, P.R., 2009. Back Jeffcott, L.B., Dalin, G., 1980. Natural rigidity Vet. Behav. Clin. Appl. Res. 5, 180–186.
kinematics of healthy trotting horses of the horse’s backbone. Equine Vet. J. 12, Moore, B.R., Holbrook, T.C., Stefanacci, J.D.,
during treadmill versus over ground 101–108. Reed, S.M., Tate, L.P., Menard, M.C., 1992.
locomotion. Equine Vet. J. 41, 297–300. Jeffcott, L.B., Dalin, G., Drevemo, S., et al., Contrast-enhanced computed tomography
Gómez Alvarez, C.B., Wennerstrand, J., 1982. Effect of induced back pain on gait and myelography in six horses with cervical
Bobbert, M.F., Lamers, L., Johnston, C., and performance of trotting horses. Equine stenotic myelopathy. Equine Vet. J. 24,
Back, W., et al., 2007. The effect of induced Vet. J. 14, 129–133. 197–202.
forelimb lameness on thoracolumbar Morishita, Y., Hida, S., Naito, M., Arimizu, J.,
Johnson, J.L., Moore-Colyer, M., 2009. The
kinematics during treadmill locomotion. Matsushima, U., Nakamura, A., 2006.
relationship between range of motion of
Equine Vet. J. 39, 197–201. Measurement of the local pressure of the
lumbosacral flexion-extension and canter
Gomez-Alvarez, C.B., Rhodin, M., Bobbert, intervertebral foramen and the
velocity of horses on a treadmill. Equine
M.F., Meyer, H., Weishaupt, M.A., Johnston, electrophysiologic values of the spinal
Vet. J. 41, 301–303.
C., et al., 2006. The effect of head and neck nerve roots in the vertebral foramen. Spine
Johnston, C., Holm, K., Faber, M., Erichsen, 31, 3076–3080.
position on the thoracolumbar kinematics C., Eksell, P., Drevemo, S., 2002. Effect of
in the unridden horse. Equine Vet. J. Suppl. Muhle, C., Resnick, D., Ahn, J.M., Sudmeyer,
conformational aspects on the movement
36, 445–451. M., Heller, M., 2001. In vivo changes in the
of the equine back. Equine Vet. J. 34
neuroforaminal size at flexion-extension
Haussler, K.K., Bertram, J.E., Gellman, K., (Suppl.), 314–318.
and axial rotation of the cervical spine in
Hermanson, J.W., 2001. Segmental in vivo Johnston, C., Holm, K.R., Erichsen, C., Eksell, healthy persons examined using kinematic
vertebral kinematics at the walk, trot and P., Drevemo, S., 2004. Kinematic magnetic resonance imaging. Spine 26,
canter: a preliminary study. Equine Vet. J. evaluation of the back in fully functioning E287–293.
33 (Suppl.), 160–164. riding horses. Equine Vet. J. 36, 495–498.
Nollet, H., Deprez, P., van Ham, L., Dewulf, J.,
Haussler, K.K., Erb, H.N., 2006a. Mechanical Kitagawa, T., Fujiwara, A., Kobayashi, N., Saiki, Decleir, A., Vanderstraeten, G., 2004.
nociceptive thresholds in the axial skeleton K., Tamai, K., Saotome, K., 2004. Transcranial magnetic stimulation: normal
of horses. Equine Vet J. 38, 70–75. Morphologic changes in the cervical neural values of magnetic motor evoked potentials
Haussler, K.K., Erb, H.N., 2006b. Pressure foramen due to flexion and extension: in in 84 normal horses and influence of
algometry for the detection of induced vivo imaging study. Spine 29, 2821–2825. height, weight, age and sex. Equine Vet. J.
back pain in horses: a preliminary study. Landman, M.A.A.M., de Blaauw, J.A., van 36, 51–57.
Equine Vet. J. 38, 76–81. Weeren, P.R., Hofland, L.J., 2004. Lameness Nollet, H., Van Ham, L., Deprez, P.,
Haussler, K.K., Hill, A.E., Puttlitz, C.M., prevalence in horses with back problems Vanderstraeten, G., 2003a. Transcranial
McIlwraith, C.W., 2007. Effects of vertebral compared to a control population: a field magnetic stimulation: review of the
mobilization and manipulation on study. Vet. Rec. 155, 165–168. technique, basic principles and
kinematics of the thoracolumbar region. Levine, J.M., Ngheim, P.P., Levine, G.J., Cohen, applications. Vet. J. 166, 28–42.
Am. J. Vet. Res. 68, 508–516. N.D., 2008. Associations of sex, breed, Nollet, H., Van Ham, L., Dewulf, J.,
Haussler, K.K., Stover, S.M., Willits, N.H., andage with cervical vertebral compressive Vanderstraeten, G., Deprez, P., 2003b.
1999. Pathologic changes in the myelopathy in horses: 811 cases (1974- Standardization of transcranial magnetic
lumbosacral vertebrae and pelvis in 2007). Vet. Radiol. Ultrasound 49, 1–6. stimulation in the horse. Vet. J. 166,
Thoroughbred racehorses. Am. J. Vet. Res. Licka, T., Frey, A., Peham, C., 2009. 244–250.
60, 143–153. Electromyographic activity of the Nuckley, D.J., Konodi, M.A., Raynak, G.C.,
Heuschmann, G., 2007. Tug of war: Classical longissimus dorsi muscles in horses when Ching, R.P., Mirza, S.K., 2002. Neural space
versus “modern” dressage (Translated from walking on a treadmill. Vet. J. 180, 71–76. integrity of the lower cervical spine: effect
Finger in der Wunde). Trafalgar Square Licka, T., Peham, C., 1998. An objective of normal range of motion. Spine 27,
Books, Vermont. method for evaluating the flexibility of the 587–595.
Holm, K.R., Wennerstrand, J., Lagerquist, U., back of standing horses. Equine Vet. J. 30, Payne, R.C., Veenman, P., Wilson, A.M., 2004.
Eksell, R., Johnston, C., 2006. Effect of 412–415. The role of the extrinsic throacic limb
local analgesia on movement of the equine Licka, T., Peham, C., Zohmann, E., 2001b. muscles in equine locomotion. J. Anat.
back. Equine Vet. J. 38, 65–69. Range of back movement at trot in horses 205, 479–490.
Hubbard, R.D., Winkelstein, B.A., 2008. without back pain. Equine Vet. J. 33 Peham, C., Frey, A., Licka, T., Scheidl, M.,
Dorsal root compression produces (Suppl.), 150–153. 2001. Evaluation of the EMG activity of the

225
 10 The neck and back

long back muscle during induced back Schlacher, C., Peham, C., Licka, T., Stubbs, N.C., Kaiser, L.J., Hauptman, J.,
movements at stance. Equine Vet. J. 33 Schobesberger, H., 2004. Determination of Clayton, H.M., 2011. Dynamic mobilisation
(Suppl.), 165–168. the stiffness of the equine spine. Equine exercises increase cross sectional area of
Peham, C., Licka, T.F., Scheidl, M., 2001. Vet. J. 36, 699–702. musculus multifidus. Equine Vet. J. 43,
Evaluation of a signal-adapted filter for Schnebel, B.E., Watkins, R.G., Dillin, W., 1989. 522–529.
processing of periodic electromyography The role of spinal flexion and extension in Sullivan, K.A., Hill, A.E., Haussler, K.K., 2008.
signals in horses walking on a treadmill. changing nerve root compression in disc The effects of chiropractic, massage and
Am. J. Vet. Res. 62, 1687–1689. herniations. Spine 14, 835–837. phenylbutazone on spinal mechanical
Peham, C., Schobesberger, H., 2004. Influence Sleutjens, J., Smiet, E., van Weeren, P.R., van nociceptive thresholds in horses without
of the load of a rider or of a region with der Kolk, J.H., Back, W., Wijnberg, I.D., clinical signs. Equine Vet. J. 40, 14–20.
increased stiffness on the equine back: a 2009. Effect of head and neck position on Tanaka, Y., Kokubun, S., Sato, T., Ozawa, H.,
modelling study. Equine Vet. J. 36, intrathoracic airway resistance and arterial 2006. Cervical roots as origin of pain in
703–705. blood gas values. Proceedings of the 4th the neck or scapular regions. Spine 31,
Peham, C., Schobesberger, H., 2006. A novel World Equine Airways Symposium E568–573.
method to estimate the stiffness of the Congress, p. 244. Tokuriki, M., Aoki, O., 1991. Neck muscles
equine back. J. Biomech. 39, 2845–2849. Sleutjens, J., Smiet, E., Van Weeren, P.R., Van activity in horses during locomotion with
Pourcelot, P., 1999. Développement d’un der Kolk, J.H., Back, W., Wijnberg, I.D., and without a rider. Equine Exerc. Physiol.
système d’analyse cinématique 3-D. 2012. Effect of head and neck position on 3, 146–150.
Application à l’étude de la symétrie intrathoracic pressure and arterial blood Tokuriki, M., Nakada, A., Aoki, O., 1991.
locomotrice du cheval au trot. PhD thesis, gas values in Dutch Warmblood riding Electromyographic activity of trunk muscles
Université Paris XII -Val de Marne. horses during moderate exercise. Am. J. Vet. in the horse during locomotion and
Res. 73, 522–528. synaptic connections of neurons of trunk
Pourcelot, P., Audigié, F., Degueurce, C.,
Denoix, J.M., Geiger, D., 1998. Kinematics Sleutjens, J., Voorhout, G., Van der Kolk, J.H., muscles in the cat. Proceeding of the First
of the equine back: a method to study the Wijnberg, I.D., Back, W., 2010. The effect of (ESB) Workshop on Animal Locomotion,
thoracolumber flexion-extension ex vivo flexion and extension on Utrecht.
movements at the trot. Vet. Res. 29, intervertebral foramina dimensions in the Tokuriki, M., Ohtsuki, R., Kai, M., Hiraga, A.,
519–525. equine cervical spine. Equine Vet. J. 42 Oki, H., Miyahara, Y., et al., 1999. EMG
(Suppl. 38), 425–430. activity of the muscles of the neck and
Rhodin, M., Gomez Alvarez, C.B., Bystrom, A.,
Johnston, C., van Weeren, P.R., Roepstorff, Slijper, E.J., 1946. Comparative biologic- forelimbs during different forms of
L., et al., 2009. The effect of different head anatomical investigations on the vertebral locomotion. Equine Vet. J. Suppl. 30,
and neck positions on the caudal back and column and spinal musculature of 231–234.
hind limb kinematics in the elite dressage mammals. Proc. K. Ned. Acad. Wetensch. Tokuriki, M., Ohtsuki, R., Kai, M., Hiraga, A.,
horse at trot. Equine Vet. J. 41, 274–279. 42, 1–128. Oki, H., Miyahara, Y., et al., 1999. EMG
Rhodin, M., Johnston, C., Holm, K.R., Sloet van Oldruitenborgh-Oosterbaan, M.M., activity of the muscles of the neck and
Wennerstrand, J., Drevemo, S., 2005. The Blok, M.B., Begeman, L., Kamphuis, forelimbs during different forms of
influence of head and neck position on M.C.D., Lameris, M.C., Spierenburg, A.J., locomotion. Equine Vet. J. 30 (Suppl.),
kinematics of the back in riding horses at et al., 2006. Workload and stress in 231–234.
the walk and trot. Equine Vet. J. 37, 7–11. horses: comparison in horses ridden Tokuriki, M., Otsuki, R., Kai, M., et al., 1997.
deep and round (‘Rolkür’) with a draw Electromyographic activity of trunk muscles
Ricardi, G., Dyson, S.J., 1993. Forelimb rein and horses ridden in a natural
lameness associated with radiographic during locomotion on a treadmill (5th
frame with only light rein contact. WEVA Congr. Abstracts: Padova). J. Equine
abnormalities of the cervical vertebrae. Tijdschrift voor Diergeneeskunde 131,
Equine Vet. J. 25, 422–426. Vet. Sci. 17, 488.
114–119.
Robert, C., Audigié, F., Valette, J.P., Pourcelot, Townsend, H.G., Leach, D., 1984. Relationship
Spadavecchia, C., Andersen, O.K., Arendt-
P., Denoix, J.M., 2001. Effects of treadmill between facit joint morphology and
Nielsen, L., Spadavecchia, L., Doherr, M.,
speed on the mechanics of the back in the mobility in the equine thoracolumbar
Schatzmann, U., 2004. Investigation of the
trotting saddlehorse. Equine Vet. J. 33 spine. Equine Vet. J. 16, 461–465.
facilitation of the nociceptive withdrawal
(Suppl.), 154–159. reflex evoked by repeated transcutaneous Townsend, H.G.G., Leach, D.H., Fretz, P.B.,
Robert, C., Valette, J.P., Degueurce, C., Denoix, electrical stimulations as a measure of 1983. Kinematics of the equine
J.M., 1999. Correlation between surface temporal summation in conscious horses. thoracolumbar spine. Equine Vet. J. 15,
electromyography and kinematics of the Am. J. Vet. Res. 65, 901–908. 117–122.
hind limb of horses at trot on a treadmill. Spadavecchia, C., Spadavecchia, L., Andersen, Van Breda, E., 2006. A nonnatural head-neck
Cells Tissues Organs. 165, 113–122. O.K., Arendt-Nielsen, L., Leandri, M., position (Rolkür) during training results in
Robert, C., Valette, J.P., Denoix, J.M., 1998. Schatzmann, U., 2002. Quantitative less acute stress in elite, trained, dressage
Surface electromyographic analysis of the assessment of nociception in horses by use horses. J. Appl. Anim. Welfare Sc. 9, 59–64.
normal horse locomotion: A preliminary of the nociceptive withdrawal reflex evoked Van den Bogert, A.J., Jansen, M.O., Deuel,
report. Proc. Conf. Equine Sports Med. by transcutaneous electrical stimulation. N.R., 1994. Kinematics of the hind limb
Science, Cordoba, pp. 80–85. Am. J. Vet. Res. 63, 1551–1556. push-off in elite show horses. Equine Vet. J.
Robert, C., Valette, J.P., Pourcelot, P., Audigié, Stubbs, N.C., Clayton, H.M., Hodges, P.W., Suppl. 17, 80–86.
F., Denoix, J.M., 2002. Effects of trotting Jeffcott, L.B., McGowan, C.M., 2010. Van Dierendonck, M.C., Sleutjens, J.,
speed on muscle activity and kinematics in Osseous spinal pathology and epaxial Tiggelman, S.L., Van Denderen, J.G.,
saddlehorses. Equine Vet. J. 34 (Suppl.), muscle ultrasonography in Thoroughbred Wijnberg, I.D. Back, W., et al., 2010. The
295–301. racehorses. Equine Vet. J. 42 (Suppl.), s38, effect of head- and neck position on
Roepstorff, L., Johnston, C., Drevemo, S., 654–661. behaviour and cortisol levels in blood and
Gustas, P., 2002. Influence of draw reins on Stubbs, N.C., Hodges, P.W., Jeffcott, L.B., salivary. Proceedings ISES, in press.
ground reaction forces at the trot. Equine Cowin, G., Hodgson, D.R., McGowan, Van Weeren, P.R., 2002. The clinical
Vet. J. Suppl. 34, 349–352. C.M., 2006. Functional anatomy of the applicability of automated gait analysis
Rooney, J.M., 1982. The horse’s back: caudal thoracolumbar and lumbosacral systems. Equine Vet. J. 34, 218–219.
Biomechanics of lameness. Equine Pract. 4, spine in the horse. Equine Vet. J. 36 Van Weeren, P.R., 2005. Equine ergonomics: a
17–27. (Suppl.), 393–399. new era? Equine Vet. J. 37, 4–6.

226
 References

Van Wessum, R., Sloet van Oldruitenborgh- Wennerstrand, J., Johnston, C., Roethlisberger- M.M.E., Barneveld, A., et al., 2008.
Oosterbaan, M.M., Clayton, H.M., 1999. Holm, K., Erichsen, C., Eksell, P., Drevemo, (Over)training effects on quantitative
Electromyography in the horse in veterinary S., 2004. Kinematic evaluation of the back electromyography and muscle enzyme
medicine and in veterinary research–a in the sport horse with back pain. Equine activities in standardbred horses. J. Applied
review. Vet. Q. 21, 3–7. Vet. J. 36, 707–711. Physiol. 105, 1746–1753.
Vandeweerd, J.M., Desbrosse, F., Clegg, P., Wijnberg, I.D., Back, W., De Jong, M., Zuidhof, Wijnberg, I.D., van der Kolk, J.H., Franssen,
Hougardy, V., Brock, L., Welch, A., et al., M.C., Van den Belt, A.J.M., Van der Kolk, H., Breukink, H.J., 2003a. Needle
2007. Innervation and nerve injections of J.H., 2004. The role of electromyography in electromyography in the horse compared
the lumbar spine of the horse: a cadaveric clinical diagnosis of neuromuscular with its principles in man: a review. Equine
study. Equine Vet. J. 39, 59–63. locomotor problems in the horse. Equine Vet. J. 35, 9–17.
Von Borstel, U.U., Duncan, I.J.H., Shoveller, Vet. J. 36, 718–722. Wijnberg, I.D., Van der Kolk, J.H., Sleutjens, J.,
A.K., Merkies, K., Keeling, L.J., Millman, Wijnberg, I.D., Back, W., van der Kolk, J.H., Back, W., 2010. Effect of head and neck
S.T., 2009. Impact or riding in a coercively 2000. The use of electromyographic position on outcome of quantitative
obtained Rollkur posture on welfare and examination as a diagnostic tool and neuromuscular diagnostic techniques in
fear of performance horses. Appl. Anim. phenytoin sodium as treatment in a case Warmblood riding horses directly following
Behav. Sc. 116, 228–236. of classic springhalt in a Dutch moderate exercise. Equine Vet. J. 42 (Suppl.
Vorstenbosch, M.A.T., Buchner, H.H.F., warmblood horse. Tijdschr. Diergeneeskd. 38), 261–267.
Savelberg, H.H.C.M., et al., 1997. Modeling 15, 743–747. Xie, H., Colahan, P., Ott, E.A., 2005.
study of compensatory head movements in Wijnberg, I.D., Back, W., Van der Kolk, J.H., Evaluation of electroacupuncture treatment
lame horses. Am. J. Vet. Res. 58, 713–718. 2002. Use of EMG examination in 7 horses of horses with signs of chronic
Waldern, N.M., Wiestner, T., von Peinen, K., to localize cervical lesions. Congress British thoracolumbar pain. J. Am. Vet. Med.
Gomez Alvarez, C.G., Roepstorff, L., Equine Veterinary Association, Glasgow, Assoc. 15, 281–286.
Johnston, C., et al., 2009. Influence of UK. September 2002. Proceedings, p. 205. Yoo, J.U., Zou, D., Edwards, W.T., Bayley, J.,
different head-neck positions on vertical Wijnberg, I.D., Franssen, H., Jansen, G.H., Yuan, H.A., 1992. Effect of cervical spine
ground reaction forces, linear and time Back, W., van der Kolk, J.H., 2003b. motion on the neuroforaminal dimensions
parameters in the unridden horse walking Quantitative electromyographic of human cervical spine. Spine 17,
and trotting on a treadmill. Equine Vet. J. examination in myogenic disorders of 6 1131–1136.
41, 268–273. horses. J. Vet. Intern. Med. 17, 185–193. Yovich, J.V., Powers, B.E., Stashak, T.S., 1985.
Weishaupt, M.A., Wiestner, T., von Peinen, K., Wijnberg, I.D., Graubner, C., Auriemma, E., Morphologic features of the cervical
Waldern, N., Roepstorff, L., van Weeren, R., Van den Belt, A.J.M., Gerber, V., 2011. intervertebral disks and adjacent vertebral
et al., 2006. Effect of head and neck Quantitative motor unit action potential bodies of horses. Am. J. Vet. Res. 46,
position on vertical ground reaction forces analysis in 2 paraspinal neck muscles in 2372–2377.
and interlimb coordination in the dressage adult royal dutch sport horses. J. Vet. Zschokke, E., 1892. Untersuchungen über das
horse ridden at walk and trot on a Intern. Med. 25, 592–597. verhältnis der knochenbildung zur statik
treadmill. Equine Vet. J. Suppl. 36, Wijnberg, I.D., Van Dam, K.G., De Graaf- und mechanik des vertebraten-skelettes.
387–392. Roelfsema, E., Keizer, H.A., Van Ginneken, Thesis, Zürich.

227
This page intentionally left blank
 CHAPTER 11 

The effects of conformation

Mikael Holmström, Willem Back

Introduction 1940; Bengtsson, 1983). Much of these horsemen’s knowledge is

still in practice but parts of it seem to have been forgotten in recent
years.
For as long as the horse has been used by man, conformation has In the practical situation, it often appears that the results from
been regarded as an important indicator of performance and research based on both subjective evaluation and quantitative anal-
soundness. In accordance with the variety of uses, many different ysis has confirmed some and rejected other aspects of these old
types and breeds of horses have evolved, ranging from heavy relationships. Some of the ‘forgotten’ relationships have also been
draught horses to light, refined racehorses. The result is that a rediscovered. Even though objective methods for conformation
huge number of breeds exist today, each with its own specific evaluation will probably play a more important role in the future,
conformational characteristics. Nevertheless, conformational traits traditional subjective evaluation will always be important. It must
are not always a guarantee for performance and soundness. The be kept in mind that there are aspects of conformation that cannot
effects of conformation on lameness and athletic potential have be measured by objective methods.
mostly been evaluated subjectively and based on anecdotal evi-
dence and experience of the observer (Van Weeren & Crevier-
Denoix, 2006). Some people would consider this an ‘art form’ in
which some people have a natural talent for assessing the influ-
Subjective evaluation
ence of different conformational characteristics on injury and Traditional subjective evaluation of conformation is currently per-
talent. This chapter will focus on conformational characteristics formed in many different ways. Almost every country has its own
that, subjectively or objectively evaluated, are considered impor- protocol. Furthermore, many breeds have their own specific regula-
tant for the performance and soundness of the sport horse and tions that instruct the judges how to judge according to the stan-
the racehorse. dards of the breed. A problem with traditional evaluation is that
subjective evaluations of conformation vary greatly between judges,
although some morphological characteristics are assessed more
The study of equine conformation consistently than others (van Vleck & Albrechtsen, 1965; Grundler,
1980; Magnusson, 1985a; Van der Veen, 1996). It is obvious that
the reliability of the evaluation is dependent on the skill and experi-
History and tradition ence of the individual judge.
Much has been written about the conformation of the horse during The importance of the traditional conformation evaluation also
the past 200 years, but very little is based on research. This does not varies between breeds. In racehorses and Standardbred trotters,
mean that the work is of less interest. It is thought provoking that racing performance is the main parameter when selecting individu-
many of the relationships between conformation and performance als for breeding. In riding horses intended for dressage and jumping,
described by Bourgelat (1750) and Magne (1866) correspond well the selection is mainly based on conformation and performance
with the results from recent research. They stressed the importance tests carried out at 3–4 years of age. Since riding horses perform in
of the hindquarter conformation. Horses with hind limbs placed competition much later in life than racehorses, it is desirable to have
well underneath themselves were found suitable for dressage work, a method for prediction of breeding value in young stallions and
whereas those with hind limbs camped out behind were likely to mares that do not have any performance record. Because the selec-
show good speed. Another early author, Hörman (1837), wrote that tion of stallions in most countries is stepwise, and performance tests
a long and forwardly sloping femur facilitates lifting of the hind are only applied to those that reach minimum conformation stan-
limb and the horse’s ability to step under itself. Ehrengranat (1818) dards, it is very important that conformation evaluation is based on
maintained that a sloping shoulder, long radius, short fore cannon correct criteria.
and a flat croup are desirable for good movement. According to A step towards a more objective interpretation and analysis of
Schmidt (1928), a small hock angle means that, although the horse subjectively obtained conformation scores is the so-called linear
will be able to step under itself easily, it will not be able to carry scoring system, in which trained members of the jury score the
weight on the hind quarters due to decreased strength in the hock. conformation of the horse relative to the mean of the population
Thus, he regarded a normal hock angle as advantageous for the (Van der Veen, 1996). Thus, judging still remains subjective, but it
function of the hind limbs. Most authors of handbooks on confor- is the interpretation and the analysis of the judges’ scores that
mation evaluation believe that a large hock angle leads to rigid and becomes more objective. In the Netherlands this is done for the
incorrect hind limb movements as well as an increased strain on Warmblood (KWPN) as well as the Friesian (KFPS) horses at stud-
the hind limb joints (Wrangel, 1911–1913; Forsell, 1927; Anon, book admission (Fig. 11.1).

229
 11 The effects of conformation

gender: name:
height at withers: reg. no.: cat. no.:
markings: sire: dam:
markings not allowed: reg. no.: reg. no.:
judging location: date: jury member:
element characteristic groep 5 10 15 20 25 30 35 40 45 characteristic
head plain racial noble
head-neck connection heavy frame light
neck short rac/frame long
horizontal racial vertical
shoulder steep frame sloping
back weak frame tight
loins weak frame tight
croup straight frame sloping
short frame long
body downhill frame upstanding
length forearm short frame long
frontlegs straight feet&legs standing under
hindlegs sickled feet&legs straight
pasterns short feet&legs long
hooves small feet&legs large
quality of legs course feet&legs hard
hair little racial much
color faded black racial jet black
frontlegs toeing-in feet&legs toeing-out
walk short walk long
weak walk powerful
trot short trot long
weak trot powerful
unbalanced trot balanced
not supple trot supple
racialtype frame feet & legs walk trot

Fig 11.1  Typical example of a linear scoring form to judge a Friesian sport horse.
Data from http://www.kfps.nl.

Quantitative analysis further developed and refined for use in the Thoroughbred race-
horse industry in the USA (M Holmstrom and D Lambert, unpub-
Quantitative methods for measuring conformation can be used for lished work, 2004). The method was originally based on reference
an objective evaluation (Rossdale & Butterfield, 2006). Several studies points marked on the horse with small paper dots glued to the skin
have been carried out, especially on riding horses. The results do often (Fig. 11.2), and when first used all measurements were registered
show a great deal of conformity, but in most cases it is impossible to ‘by hand’ from a picture projected on a wall using a simple measur-
directly compare such parameters as angle measurements from dif- ing band and a protractor. Later the whole procedure was computer-
ferent studies, due to differences in methods of measurement. ized and the measurements were obtained from a digital photo on
Quantitative conformational analysis, as a complement to the a laptop computer. The measurements were registered by clicking
traditional evaluation, has been proven to increase the accuracy in with the mouse on the white markers. The computer then calculates
the prediction of performance potential in young riding horses all length and angle measurements (Fig. 11.3). In the most recent
(Holmström & Philipsson, 1993). However, applying this knowl- version joint angles and length and proportional measurements can
edge in practice, i.e. in breeding evaluation programs, talent scout- be obtained from a picture with just one paper dot on the hip.
ing, etc., has been very difficult. Much of the resistance to quantitative
analysis is due to the deeply rooted tradition of subjective evalua-
tions in the horse industry, combined with a lack of familiarity with
the new methods. In addition, there are some shortcomings in the
Conformation and growth
method itself. Obtaining quantitative measurements has, until Conformation evaluation in young growing horses calls for an
recently, been a slow procedure – too slow to be incorporated into understanding of the effect of age, i.e. the effect of growth on the
stallion tests and other similar events. New computerized methods conformation. The accuracy of quantitative methods applied to
have now been developed that will speed up the process and make foals and yearlings, with the aim of predicting future performance
it possible to measure many horses in a short time. and soundness, are dependent on age adjustments according to
In recent years an objective method, developed in Sweden by reliable growth curves for each individual measurement. Growth
Magnusson (1985a) for a study of Standardbred trotters, has been curves for some body measurement such as height at the withers

230
 The study of equine conformation

2 9

10

4 11

5 12

13
6
7 14

8 15

Fig 11.2  Standard position and reference points used in conformation analysis of Thoroughbred racehorses. Head and forelimbs: 1, the cranial end of the
wing of the atlas; 2, the proximal end of the spine of the scapula; 3, the posterior part of the greater tubercle of the humerus; 4, the transition between the
proximal and middle thirds of the lateral collateral ligament of the elbow joint; 5, the lateral tuberosity of the distal end of the radius; 6, the space between
the fourth carpal and the third and fourth metacarpal bones; 7, the proximal attachment of the lateral collateral ligament of the fetlock joint to the distal end
of the third metacarpal bone; 8, the proximal attachment of the lateral collateral ligament of the pastern joint to the distal end of the first phalanx. Hind
limbs: 9, the proximal end of the lateral angle of the ilium; 10, the center of the anterior part of the greater trochanter of the femur; 11, the proximal
attachment of the lateral collateral ligament of the stifle joint to the femur; 12, the attachment of the long lateral ligament of the hock joint to the plantar
border of the calcaneus bone; 13, the space between the fourth tarsal and the third and fourth metatarsal bones; 14, the proximal attachment of the lateral
collateral ligament of the fetlock joint to the distal end of the third metatarsal bone; 15, the proximal attachment of the lateral collateral ligament of the
pastern joint to the distal end of the first phalanx.

b 5

1 6
2 g
c
7
3

d h

8
e i
4
9
f

Fig 11.3  Length measurements: a, neck; b, shoulder; c, humerus; d, radius; e, fore cannon; f, fore pastern; g, femur; h, tibia; i, hind cannon; j, hind pastern.
Angle measurements: 1, shoulder inclination; 2, shoulder joint; 3, elbow joint; 4, fore fetlock joint; 5, pelvis inclination; 6, femur inclination; 7, stifle joint; 8,
hock joint; 9, hind fetlock joint.

231
 11 The effects of conformation

Correct Outwardly rotated Bench knee Base wide Narrow at the knees Toe-in

Buck knee Calf knee Correct Outwardly rotated Narrow at the knees Wide at the hocks

Fig 11.4  Deviations of limb and toe axes in the fore and hind limbs.

are mainly breed specific while many other growth curves are spe- not sufficient. Many important variables are almost impossible to
cific for the method used for measuring a trait. Thus, growth graphs judge correctly without objective measurements.
obtained by one quantitative method cannot be used together with This chapter will focus on objective measurements and subjective
a different quantitative method. conformational characteristics that are important for the function
of sport horses and racehorses. For those who have a specific interest
in the characteristics of a specific breed, that information can be
Conformation, performance and soundness obtained from the respective breeding organization.
Searching for talented young sport horses involves evaluation of
many different qualities. Temperament, movement and jumping
ability are of course the most important in riding horses, but the
significance of conformation must not be neglected. Furthermore,
Deviations of limb and toe axis
dressage riders seek ‘good looking’ horses, with conformation that One important part of the subjective evaluation is to describe devia-
facilitates good movements, soundness and, above all, the ability tions of limb and toe axes, if any. The most common deviations are
to show a high level of collection. The competition results of Grand described in Figure 11.4. Deviations from the straight (normal) limb
Prix horses are certainly dependent on the skill of their riders and and toe axes have traditionally been considered as a considerable
trainers, but their conformation and movement must have the basic weakness. However, not all deviations from what has been the
qualities that create the necessary conditions for successful training desired conformation should be judged as abnormal. About 80%
of the horse. Resistance from the horse is often interpreted as poor of all Warmblood riding horses and Standardbred trotters had out-
temperament but might just as well be due to pain or lack of ability wardly rotated hind limbs (Magnusson, 1985a; Holmström et al.,
to carry weight on the hind limbs caused by inappropriate confor- 1990). The frequency of this ‘faulty conformation’ is so high that it
mation and/or movement. Potential world class Grand Prix dres- must be regarded as normal. Boldt (1978) maintains in his dressage
sage horses are difficult to find because many promising young handbook that outwardly rotated hind limbs facilitate exercises
horses with excellent gaits fail to learn passage, piaffe and other such as half-pass and shoulder-in. In this context, it is important to
collected movements, resulting in years of wasted training. The distinguish between hind limb rotation and toe-out conformation
ability to collect and work in balance is also very important in a as well as between rotated hind limbs and hind limbs that are
jumping horse. narrow at the hocks (cow hocked). Confusion between these char-
Conformation is also important in the selection of most other acteristics should be avoided, especially as the rotated hind limb
performance horses including Thoroughbred racehorses and Stan- conformation should be considered normal. To judge this aspect of
dardbreds. However, a successful use of conformation as an indica- conformation, the observer should stand behind the point of the
tor of performance requires good knowledge of the relationships hock rather than behind the tail of the horse. The latter position
between conformation and performance, as well as a reliable makes it impossible to see if the rotated hind limb is also narrow
method for its evaluation. In most cases a subjective evaluation is at the hocks.

232
 The study of equine conformation

Table 11.1  Frequencies (percent) of subjectively scored deviations of limb and toe axes in different groups of riding horses

Characteristic Mild Moderate Severe Total

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Forelimbs
Front view
Outwardly rotated 3.0 7.1 3.0 3.7 – 3.6 – 0.6 – – – – 3.0 10.7 3.0 4.3
Bench knee 48.4 32.5 53.0 48.4 6.1 21.4 11.0 11.8 – – – – 54.5 53.6 64.0 60.2
Base wide – – – – – – 1.0 0.6 – – – – – – 1.0 0.6
Narrow at knees 9.1 21.4 10.0 11.8 – – – – – – – – 9.1 21.4 10.0 11.8
Toe-in 39.4 25.7 37.0 37.2 9.1 10.7 7.0 8.1 – – 3.0 1.9 48.5 46.4 47.0 47.2

Side view
Buck knee 18.2 35.7 6.0 13.8 3.0 10.7 2.0 3.7 6.1 – – 1.2 27.3 46.4 8.0 18.7
Calf knee 9.1 7.1 21.0 16.2 3.0 – 3.0 2.5 – – – – 12.1 7.1 24.0 18.7

Hind limbs
Rear view
Outwardly rotated 84.9 82.1 67.0 73.3 3.0 3.6 7.0 5.6 – – – – 87.9 85.7 74.0 78.9
Narrow at the hocks – 7.1 10.0 7.4 – – – – – – – – – 7.1 10.0 7.4
Wide at the hocks – 14.3 3.0 4.4 – – 1.0 0.6 – – – – – 14.3 4.0 5.0
Toe-out – 0 1.0 0.6 – – – – – – – – – – 1.0 0.6

Side view
Camped under – 7.1 4.0 3.7 – – 1.0 0.6 – – – – – 7.0 5.0 4.3

1, elite dressage horses (n = 33); 2, elite show jumpers (n = 28); 3, riding school horses (n = 100); 4, total (n = 161).
Data from Holmström et al. (1990). ©EVJ Ltd.

Bench kneed conformation in the forelimbs together with a The frequency of toe-in and/or bench knee conformation, as well
toe-in conformation is a very common type of deviation in riding as of most other deviations, was the same among the elite dressage
horses, while outwardly rotated forelimbs are seldom seen in adult horses and show jumpers as in a group of riding school horses
riding horses (Holmström et al., 1990). According to Magnusson (Table 11.1). This indicates that mild-to-moderate deviations from
and Thafvelin (1985a) the converse is the case in Standardbred the ‘normal’ limb conformation do not impair either soundness or
trotters. This difference may partly be explained by the fact that performance in riding horses. However, it is important to empha-
trotters are narrower through the chest than the riding horses, even size that, even though mild and moderate deviations can be
after adjustment for different height at the withers. Several other accepted, severe deviations of any type should be considered as a
authors have also found that bench knee and toe-in conformation major weakness. Several authors have claimed that there is an
seems to be related to weight and width of the breast (Santschi increased stress to the distal parts of the limbs in horses with toe-out
et al., 2006; Firth et al., 1998). Santschi et al. (2006) studied the and toe-in conformation (Churchill, 1962; Rooney, 1968; Beeman,
development of front limb deviations in foals and concluded that 1973; Magnusson, 1985c). This can often be noticed as synovial
calf knee conformation in foals tends to self correct and eventually distensions of the fetlock and the coffin joints, as well as swelling
and it was also found that the incidence of bench knee increased of the distal metacarpal growth plate.
between the age of 1 week and 18 months. Anderson et al. (2004) According to Adams (1974), calf knee (back at the knee) confor-
compared conformation to injury in Thoroughbreds and found that mation may predispose to lameness, whereas buck knee (over at the
calf knees were somewhat protective agains injuries. Today most knee) conformation is less serious. This is probably true for Stan-
clinicians feel that mild carpal valgus is protective and is actually dardbreds as well as Thoroughbred racehorses that overextend the
normal during growth. carpal joint considerably during the midstance phase. Steel et al.
As in the hind limbs, there is often confusion between outwardly (2006) reported that carpal lameness occurred in 28% of a group
rotated forelimbs and a toe-out conformation. Even though bench of Standardbred horses and was present in 56% with forelimb lame-
knee conformation in a riding horse does not have any documented ness. Of the variables studied, poor forelimb conformation and
negative effects on the long-term performance, bench kneed horses more intense speed training were predisposing factors. Back et al.
have been reported to have a higher frequency of splints on the (1996) proved in Warmbloods that there is a relationship between
medial side of the third metacarpal bone (Adams, 1974; Davidson, a calf knee conformation at square stance and carpal hyperexten-
1970; Nordin, 1980). sion at trot.

233
 11 The effects of conformation

Head, neck and body inclusion of objective measurements can minimize these discrepan-
cies. Figure 11.5 shows two horses, one with a low set neck and the
Many of the criteria that are used to describe the head, neck and other with a well set neck.
body are very difficult or impossible to analyze objectively. Such In handbooks on evaluation of the conformation of sport horses,
criteria are type, the setting of the neck, the shape of the withers, some characteristics are described more consistently. A long and
the ‘harmony’ of the horse, etc. They have to be evaluated subjec- well-set neck are considered important for most kinds of perfor-
tively. Other characteristics, such as length of the neck and body or mances (Wrangel, 1911–1913; Forsell, 1927; Anon, 1940; Boldt,
height at the withers, can be objectively measured or subjectively 1978; Bengtsson, 1983) and a long well-developed withers is said
scored (Van der Veen, 1996). to be important for horses that are working under saddle (Van der
It is obvious that there are significant relationships between head, Veen, 1996). In a large, cohort study of a Warmblood population,
neck and body characteristics and performance, which are reflected in which linear scoring results collected at studbook admission were
in the difference between, for example, a slow, heavy draught horse related to later performance in dressage and show jumping, it
and a fast, elegant racehorse. Thus, differences between breeds have, appeared that when horses were genetically selected on height at
to a large extent, originated from different demands upon the the withers, neck length was shorter. Longer necks were genetically
horses. However, within a certain breed it is much more difficult to related to a conformation more suited for dressage performance but
correctly describe and evaluate the relatively subtle differences, and had less impact on jumping performance (Ducro et al., 2009a).
this difficulty has been associated with a rather large variation In contrast, results from objective studies show that elite show
between judges (van Vleck & Albrechtsen, 1965; Grundler, 1980; jumpers have significantly longer necks than elite dressage horses
Magnusson, 1985a). Long experience, a deep understanding of the and ‘normal’ horses (Holmström, unpublished data) (Table 11.2).
influence of conformation on performance and, when appropriate, Other studies showed significantly shorter necks in elite dressage
horses than in other riding horses but there was no significant cor-
relation between the length of the neck and gaits under saddle in
4-year-old riding horses (Holmström et al., 1990; Holmström &
Philipsson, 1993). Available objectively obtained results indicate
that a long neck might be an advantage for jumping horses, prob-
ably because it makes it easier for the horse to maintain balance
over the fence. In dressage horses, the setting of the neck is probably
more important than its length (Van der Veen, 1996). It is generally
agreed among riders and trainers of riding horses that a low set neck
makes it very difficult to work the horse in a proper frame. However,
there are no data supporting this statement, mainly because of the
difficulties of objectively measuring the setting of the neck.
In Thoroughbred racehorses the length of the neck is part of the
overall subjective conformation evaluation of young stock. The
general opinion is that sprinters should have shorter necks than
stayers but objective analysis of 780 grade 1 and group 1 winners
in the USA and Europe showed that on both dirt and turf the sprint-
ers, i.e. distances up to 7 furlongs, had slightly longer necks com-
pared to the rest of the grade 1 and group 1 winners (M Holmstrom,
unpublished work, 2009). The differences were small but statisti-
cally significant.
A The setting of the head to the neck is also of importance (Van der
Veen, 1996). A wide throat latch has always been considered impor-
tant in racehorses, Standardbred trotters and Quarter Horses
because it is said to facilitate breathing. In dressage horses, a wide
distance between the wing of the atlas (first cervical vertebra) and
the posterior ridge of the mandible has been considered important
by riders. In a study of differences between elite dressage horses,
elite jumping horses and ‘normal’ horses it was found that both the
dressage horses and the jumpers had significantly greater width in
this area than other horses (Holmström, unpublished data). A pos-
sible explanation is that a small distance might cause problems,
most likely a mechanical resistance, at higher levels of collection
when the horse is required to perform a maximal flexion at the poll.
Height at the withers has been linked to jumping performance in
several studies (Neisser, 1976; Langlois et al., 1978). Müller and
Schwark (1979) found that show jumpers were taller at the withers
than dressage horses. Even though there is a positive correlation
between height at the withers and jumping performance up to a
certain limit (around 172 cm), it must be remembered that there
is a great variation among elite show jumpers. The range of the
height at the withers was between 158 cm and 178 cm in a study
of the World Cup finalists in 1997 (Holmström, unpublished data).
According to Ducro et al. (2009a) height at the withers was geneti-
cally related to dressage ranking but not to jumping ranking.
B It is generally agreed that height at the withers is not related to
stride length in different gaits (von Wagener, 1934; Krüger, 1957;
Fig 11.5  Horses with (A) low set neck and (B) well set neck. Dušek et al., 1970; Dušek, 1974). However, Holmström and

234
 The study of equine conformation

Table 11.2  Comparison of adjusted means for length and angle Table 11.3  Significant conformational differences between
measurements* horses treated for recurrent lameness and other insured horses

Variable 4-year Dressage Jumping Injured Variable Lame horses Sound horses
olds mean (SD) mean (SD)
Length measurement (cm) Height at the withers 165.4 (4.2) 163.5 (3.60)*
a ab c bc
Neck 70.51 71.55 74.02 72.87 Femur inclination 86.7 (1.7) 85.4 (2.3)**
a c a ab
Scapula 40.07 42.00 40.51 41.00 Hind fetlock joint angle 156.6 (6.2) 153.8 (6.4)*
a b a a
Humerus 32.10 33.07 32.14 32.01 *p < 0.05; **p < 0.01.
a b a
Radius 37.15 37.64 37.69 37.58ab
Fore 20.92a 21.56b 21.37b 21.14ab
cannon
In Standardbred trotters many authors have found a positive cor-
Fore 9.12a 9.52b 9.51b 9.16a relation between height at the withers and performance (Bantoiu,
pastern 1922; Richter, 1953; Magnusson, 1985d), while a negative correla-
tion between height at the withers and soundness was found by
Femur 40.24a 41.20b 40.36a 40.09a
Magnusson (1985c). In Thoroughbred racehorses no significant
Tibia 48.52 a
49.14 ab
49.74 b
49.40ab effect of the height at the withers on performance has been found
by Holmström (unpublished work, 2004) but the height at the
Hind 26.83a 27.36b 26.68a 26.55a
withers is likely to have a similar negative correlation to soundness
cannon
as in Standardbred trotters. In a study of relationships between
Hind 8.74a 9.01ab 9.34b 9.07ab conformation and soundness in insured riding horses in Sweden,
pastern horses with recurrent lameness problems were significantly taller at
the withers than the sound horses (Table 11.3) (Holmström, unpub-
Angle measurements (°) lished data). This was confirmed by a (UK) study among owners of
Shoulder 64.5a 66.3b 67.0b 66.5b registered dressage horses, in which it was reported that 33% of the
inclination horses had been lame at some time during their career, 24% of these
within the previous 2 years (Murray et al., 2010). A number of
Shoulder 126.3a 124.5b 126.2a 126.0a factors were associated with the occurrence of lameness including
joint increased risk for older and bigger horses.
Swedish Warmblood riding horses (Holmström et al., 1990)
Elbow joint 152.4a 148.5b 151.1a 150.7a
and Hannovarian stallions (Dušek, 1974) seemed to have a rectan-
Fore fetlock 148.7a 149.9ab 151.3b 148.2a gular body shape, i.e. the length of the body is greater than the
joint height at the withers. Oldenburg and East Friesian breeds have also
been shown to have this body form (Degen, 1953; Weferling,
Pelvis 31.0a 27.5b 28.2b 31.0a 1964). At the Royal Dutch Warmblood selections, stallions are
inclination selected for ‘rectangular’ body proportions in contrast to a ‘square’
Femur 85.4a 84.7b 84.5b 87.8c body (Van der Veen, 1996). Müller and Schwark (1979), on the
inclination other hand, measured 687 horses competing in dressage, show
jumping and 3-day events and found them to have a rather short
Stifle joint 154.1a 155.6b 154.0a 153.4a body form. The horses had about the same height at the withers as
Hock joint 159.4a 160.4a 159.2a 157.0b Swedish Warmbloods but it was not clear how the body length was
measured. Neisser (1976) and Schwark et al. (1977) found nega-
Hind 154.6a 153.4a 155.7a 156.1a tive correlations between body length and performance in show
feltlock jumping.
joint The average body length of 780 grade 1 and group 1 winners in
the USA and Europe analyzed by Holmström (unpublished work,
*Between elite dressage horses (n = 40), elite show jumpers (n = 51), 4-year-olds
2009) was 108% of the height at the withers, i.e. the Thoroughbred
tested at quality events (n = 217) and horses with back problems or recurrent
racehorses had, on average, a long body form. No significant cor-
lameness problems (n = 52). Differences in sex, and, for the stifle angle, differences
in femur inclination, have been taken into consideration. relation between body length and performance has been found but
a short body form is very rare among top class racehorses. No sig-
Values with different superscripts differ significantly from each other.
nificant differences in body length between grade 1 and group 1
winners on dirt and turf were found. The influence of body length
on performance is still somewhat unclear but a majority of the
available studies indicate that a somewhat long body form is
Philipsson (1993) found positive correlations between height at the preferable.
withers and subjective scores for the canter in 4-year-old riding There is not much objective data available on the effect of the
horses. However, elite dressage horses have not been found to be length of the body on soundness but Magnusson (1985c) found
larger than other riding horses (Barrey et al., 2002). Ducro et al. that Standardbred trotters with a short back had fewer problems
(2009a) found that height at the withers was positively genetically with back pain then those with a long back. On the other hand,
correlated to conformation grades for sports performance. The horses with short backs showed more scalping problems. This con-
results above indicate that judges might be positively influenced by firms the statement that a short back is a strong back but predis-
the size of the horse, while in reality it has just a marginal impact poses to interference, such as over-reaching, forging and scalping
on elite performance in dressage and show jumping. (Pritchard, 1965; Nordin, 1980; Van der Veen, 1996).

235
 11 The effects of conformation

Scope, as defined by Holmström and Lambert, i.e. the relationship Subjectively, it is very difficult to correctly estimate the real slope
between the length of the back and the distance between the fore- of the shoulder. The problem is that in some horses there is a con-
limbs and hind limbs, has shown to have a significant impact on siderable discrepancy between the external outline of the shoulder
the performance of Thoroughbred racehorses on dirt tracks in the and the real inclination of the scapula (Fig. 11.7). To be able to
USA (Holmström, unpublished work, 2004). The study, including judge the slope of the shoulder correctly, it is necessary to palpate
2600 racehorses in the USA and Europe, showed that the scope, as the position of the scapula. On the other hand there is one impor-
with most other conformational traits, had a non-linear correlation tant aspect of the shoulder that might have been underestimated.
to performance on a mile and longer, while it was of less importance A seemingly long and sloping shoulder in combination with a long
for sprinter performance. It was also less important for the perfor- and well-developed withers will place the rider more to the rear on
mance on European tracks. The optimal scope increased with race the horse, resulting in better balance. As a result of the better
distance, i.e. the longer the race distance, the longer the horse’s back balance, the horse will be able to move its forelimbs more freely,
relative to the distance between the forelimbs and hind limbs. and reach higher and more forward. Thus, the effect of a ‘subjec-
tively’ sloping shoulder on the forelimb movements might be more
important than the real or ‘objective’ slope.
Forelimbs Of the forelimb conformational details, the length of the humerus
A long and sloping shoulder has always been considered as advanta- showed the strongest correlation to good gaits in 4-year-old riding
geous for the movements of the forelimbs (Ehrengranat, 1818; Van horses (Table 11.4) (Holmström et al., 1993; Holmström, unpub-
der Veen, 1996). Sellet et al. (1981) suggested that a long sloping lished data). Elite dressage horses have been shown to have a sig-
scapula was associated with ergonomic efficiency in 2-year-old nificantly longer humerus than both show jumpers and ‘normal’
pacing fillies. In a study of conformational characteristics of elite horses (Table 11.2). In the older literature, the importance of a long
dressage horses and show jumpers, both groups had significantly
more sloping shoulders (Holmström et al., 1990). There was also
a significant correlation between high gait scores and a sloping
shoulder in 4-year-old riding horses (Holmström et al., 1993). Nev-
ertheless, in a recent study (Table 11.2), the opposite result was
found (Holmström, unpublished data). Apparently, the slope of the
shoulder, objectively measured, does not have to be directly con-
nected to elite performance in riding horses. Sport horses can
perform at top level with a less sloping shoulder because other
qualities are more important. The Royal Dutch Warmblood stud-
book uses the criterion that horses should have a functional, but
nonetheless ‘eye-catching’ conformation, that would allow and
preferably facilitates sports performance at the highest level (Van
der Veen, 1996). Henninges (1933), however, found a positive cor-
relation between a sloping shoulder and stride length in walk.
Theoretically, a more sloping shoulder might facilitate the forward
and upward movement of the forelimbs during the last part of the
swing phase in trot (Fig. 11.6). Back et al. (1996) correlated the
joint angles at square stance to their kinematics at trot in a group
of young adult Warmbloods (n = 24) and proved that a sloping
scapula was correlated to a more protracted forelimb.

Shoulder
inclination = +5°

B
Fig 11.6  Possible effect of different shoulder conformation on forelimb
orientation at the beginning of the swing phase retraction: 5° difference in Fig 11.7  (A,B) Horses with different slopes of the shoulders and scapula.
shoulder inclination and all other angles unchanged. Note the discrepancy between the real slope and the outer contour in (B).

236
 The study of equine conformation

Table 11.4  Results from multiple regression analyses of the effects of conformation measurements and scores on gaits under saddle and
jumping abilitya

Gaits under saddle Jumping ability

Variable 1984 1996 1984 1996

Objective measurements
Humerus length 0.16*** 0.12** 0.12** 0.10** −0.12 −0.13
Pelvis inclination −0.04* −0.02 −0.005 −0.005 0.05 0.03
Femur inclination −0.21*** −0.15*** −0.13*** −0.08** −0.06 0
Stifle angle 0.04* 0.01 −0.006 −0.01 0.09 −0.04
Hock angle – – – – 0.01 0.11*

Subjective scores
Type – 0.11 – −0.04 0.40 0.02
Head, neck and body – 0.12 – 0.11 −0.13 0.18
Extremities – −0.08 – 0.02 0.18 −0.28
Walk – 0.98*** – 0.25*** 0.08 0.05
Trot – 0.19** – 0.45*** 0.16 0.11

Coefficient of determination, R2 (%)

Conformation score – 26.12 – 25.13 3.86 0
Objective measurements 24.76 24.76 13.32 13.32 1.85 1.31
Score + measurements – – 42.24 42.61 3.84 0.42

*p < 0.05; **p < 0.01; ***p < 0.001.

a
Partial regression coefficients for variables that showed either significant correlation to the performance traits or significant differences between elite and ‘normal’ horses were
calculated. Effects of sex and event (site) were included in the model.

radius and a short fore cannon is often stressed (Ehrengranat, 1818;

Schmidt, 1928; Wrangel, 1911–1913), but surprisingly the length of
the humerus is never mentioned.
In the subjective judging procedure good forelimb movements
are often described in terms of ‘freedom of the shoulders’. The
analyses of high-speed films have shown that ‘freedom of the
shoulders’ is not a completely adequate expression. The most sig-
nificant differences between good and poor forelimb movements
were found in the elbow and carpal joints, with the elbow being
more important (Clayton et al., 2004; Holmström et al., 1994). At
the most forward position of the forelimb, the elbow joint was
flexed ~30° more in the horses with good forelimb movements
compared to those with poor movement. Consequently, the fore
hoof was lifted higher above the ground and reached more forward.
The effect of the greater flexion in the elbow and carpal joints is
shown in Figure 11.8. The significance of elbow joint flexion for the Good trot First
forelimb movements explains why a correlation was found between hoof-ground
the length of the humerus and gait score in the conformational contact
studies: a long humerus gives a long triceps muscle that may facili-
tate a larger range of elbow movements. The difference in forelimb
movements between good and poor horses can also to some extent Poor trot
be explained by differences in the rotation of the scapula in the
sagittal plane parallel to the thorax. Back et al. (1994) found that Fig 11.8  Forelimb angles at the most forward position of the limb.
high subjective gait scores were significantly correlated to more
scapular rotation. The humerus is significantly shorter in horses
where less knee action in the forelimbs is preferred, such as Thor-
oughbred racehorses, or in horses where the forelimb movements
have no influence on performance (Holmström, unpublished

237
 11 The effects of conformation

data). According to Sellet et al. (1981) a long forearm is important

in pacing Standardbreds.
In Thoroughbred racehorses the slope of the humerus has shown
to be of significant importance for the efficiency of the stride (Holm-
ström, unpublished work). A vertical humerus has a negative impact
on the performance. The most likely explanation is that a vertical
humerus limits the forward reach of the forelimb and the stance
duration of the forelimbs, which in turn has a negative effect on the
horse’s ability to run at high speed and breathe normally. A horse
cantering or galloping can only exhale during the stance phase of the
front limbs. If the stance duration of the front limbs at full speed
becomes too short the horse will not be able to exhale the same
volume as it inhales, which will cause it to stop breathing normally.
The fore pastern has been shown to be significantly longer in elite
dressage and jumping horses than in other riding horses, while the
inclination to the horizontal plane showed no differences (Holm-
ström, 1990). Horses with short, upright fore pasterns are considered
to be prone to injuries, as are those with long sloping pasterns
(Adams, 1974). Magnusson (1985c) found more swellings of the
superficial flexor tendon in horses with upright pasterns. The short
upright pastern probably also has a negative effect on the movements A
due to less elasticity. When joint angles at square stance were corre-
lated to their kinematics at trot in a group of young adult Warm-
bloods (n = 24), it was proved that a more extended fore fetlock joint
(larger palmar angle) was correlated to more extension and more
protraction at the trot (Back et al., 1996). This would generate more
suppleness in the forelimb with generally a passive ‘strut’ function,
which is correlated to higher scores for the trot and ‘good gait’ (Back
et al., 1994). Ergonomic efficiency in pacing horses has been shown
to be correlated to an elastic fetlock joint (Sellet et al., 1981).

Hind limbs
The importance of the hind limbs for sport horse performance is
obvious (Back et al., 1995). The hind quarters constitute the ‘engine’
of the horse that, depending on the type of performance, should
lift the horse over a fence, push it forward at a high speed or over
long distances or, as in dressage horses, carry a lot of weight.
Both horsemanship handbooks and many scientific studies indi-
cate that good conformation of the hind limbs is essential for good
performance and soundness (Van der Veen, 1996).
A small angle of the pelvis to the horizontal plane has been B
reported to have a positive effect on performance in dressage and
jumping horses. In a recent study, elite dressage and jumping horses Fig 11.9  Horse with large discrepancy between (A) the slopes of the pelvis
had significantly flatter pelvises than ‘normal’ riding horses (Table and the croup, compared to a horse (B) with rather good correspondence
11.2). The same results were found in a similar study in 1988 (Hol- between pelvis and croup.
mström et al., 1990). Holmström et al. (1993) also found a positive
correlation between the slope of the pelvis and scores for the walk
at 4-year-old tests for riding horses. Ehrengranat (1818) states that position of the pelvis and a more horizontal position of the femur
a flat croup is desirable for good movements. However, there is not at square stance. Furthermore, the more vertical position of the
always a good correspondence between the slope of the croup and pelvis was correlated to a less retracted hind limb providing a more
pelvis. Many horses have a flat croup combined with a steep pelvis collected appearance. This does not, however, mean that a more
(Fig. 11.9). This must be taken into consideration when evaluating sloping pelvis has a positive correlation to real gait quality or dres-
conformation subjectively. In analyses of high-speed films it has sage performance. The majority of studies on dressage horses and
been found that dressage horses classified as good movers had a show jumpers lead to the conclusion that these horses’ perfor-
larger rotation of the pelvis during the stride than those classified mances are facilitated by relatively flat pelvises. Thus, it is important
as poor movers (Holmström et al., 1994). Pelvic rotation is one of to evaluate pelvis slope correctly in riding horses.
the biomechanical parameters that contribute to elastic gaits in In Thoroughbred racehorses Holmström (unpublished work,
good dressage horses, while a non-rotating pelvis results in short 2004) found strong correlations between a smaller angle between
and inelastic gaits. A flat pelvic conformation facilitates pelvic rota- the pelvis and horizontal plane (flatter pelvis) and top performance.
tion. Rotation of the pelvis during the stride is more pronounced As mentioned earlier a flat pelvis facilitates pelvic rotation, which
in passage than in the other paces, and the degree of change in the very important part of an efficient stride in the racehorse.
pelvic movement pattern from trot in hand to passage is approxi- In Standardbred trotters a steeply sloping pelvis has been reported
mately the same in all horses, irrespective of the initial pattern to be related to synovial distention in the femoropatellar joint and
(Holmström et al., 1995a,b). Pelvic rotation in the trot might be a the medial synovial sac of the femorotibial joint, and also to have
determinant of passage performance. Back et al. (1996) found that a negative effect on the soundness of the hock joints (Magnusson,
a flexed hip joint at movement was correlated to a more vertical 1985c). On the other hand, a flat pelvis is significantly correlated

238
 The study of equine conformation

with pain on palpation of the croup muscles. The length of the (Bourgelat, 1750; Magne, 1866), which is very much the result of a
pelvis has been reported to have a positive correlation to jumping forward sloping femur.
ability (Langlois et al., 1978) and to stride length in the walk (Kro- The femur slope is also important in Thoroughbred racehorses
nacher & Ogrizek, 1931). and shows a significant non-linear correlation to top performance
Probably the most important individual conformational detail (Holmström, unpublished data). There are different optimal femur
for most sport horses is the femur. A long and forwardly sloping slopes depending on distance and type of track. The optimal femur
femur places the hind limb more under the horse, which allows the in a sprinter is a bit less forward sloping than in a stayer. Further-
horse to keep its balance more easily and carry more weight on the more, horses with a femur slope close to the vertical have a better
hind limbs, since the hind limb position is closer to the center of chance for top performance on turf. A combination of a very sloping
gravity. Van der Veen (1996) has suggested that optimally the stifle pelvis and a vertical femur has been shown to disqualify racehorses
joint at square stance should be situated below the tuber coxae. In from top performance. In a study of approximately 17 000 Thor-
a study of Brandenburg horses, Kronacher and Ogrizek (1931) oughbred racehorses with quantitative conformation data (Holm-
showed a positive correlation between stride length in walk and the ström, unpublished data) there were no grade 1 or group 1 winners
length of the pelvis and femur. The slope of the femur has been with this type of proximal hind limb conformation. The vast major-
reported by many authors to be related to performance. Langlois ity of the horses with a combination of very sloping pelvis and verti-
et al. (1978) showed that good jumping performance was signifi- cal femur had below average performance records. The reason for the
cantly correlated to a forwardly sloping femur, i.e. a small angle of large negative effect on performance is probably that the very sloping
the femur to the horizontal plane (Fig. 11.10). In 4-year-old riding pelvis limits the rotation of the pelvis during gallop and also together
horses the forward sloping femur has been shown to have the with the vertical femur limits the range of motion of the femur,
strongest correlation to gait quality of all studied variables (Holm- especially retraction, resulting in a short and inefficient stride.
ström et al., 1993). In a recent study both elite dressage and A forwardly sloping femur has also been shown to have a positive
jumping horses had significantly more forward sloping femurs than effect on soundness. In a study of riding horses with recurrent lame-
‘normal’ horses (Table 11.2). More than 150 years ago this correla- ness and back problems attending one of the major horse clinics in
tion was stated by Hörman (1837). Other authors claimed that Sweden, the femur was significantly more vertical than in ‘normal’
horses intended for dressage work should be well camped under horses and elite horses (Table 11.2). Going through insurance
company records of riding horses that were measured at quality
events as 4 year olds, horses with recorded recurrent lameness and
back problems had a significantly more vertical femur than the
sound horses (Table 11.3) (Holmström, unpublished data).
In riding horses, the stifle angle should not be too small. Elite
dressage horses have been shown to have a significantly straighter
stifle angle than elite show jumpers and ‘normal’ horses (Table
11.2). A small stifle angle results in a lot of strain to the quadriceps
femoris muscle, i.e. the muscle that extends the stifle. The quadri-
ceps femoris probably is the most strained group of muscles when
a horse works in collected gaits. If the muscles cannot ‘lock’ the stifle
in an extended position when maximum load is put on the hind
limbs, the horse must transfer weight to its forelimbs and it is then
no longer working in balance. In Standardbred trotters a positive
correlation has been found between stifle angle and performance
(Magnusson, 1985d). The same has been found in Thoroughbred
racehorses but again, there is a non-linear correlation to perfor-
mance and an optimal stifle angle, which vary depending on the
conformation of the rest of the hind limb (Holmström, unpub-
A lished data). Generally a very small stifle angle leads to a weak hind
limb and poor performance. The hind limb has to be able withstand
the huge load at high speed and at the same time store elastic
energy, which not possible with a very small stifle angle.
Comparing the hock angles of elite dressage and jumping horses
with ‘normal’ horses (Fig. 11.11), it has been found that the dressage
horses in general had larger hock angles or, more correctly, there were
no sickle hocked dressage horses (Holmström et al., 1990). However,
it could not be proved that the gaits improved with larger hock angles
(Holmström et al., 1993). A more recent study failed to show any
differences in the hock angle between elite horses and others, mainly
because the mean hock angle in the ‘normal’ horses had increased
from 155.4–159.4°, which was almost the same as in the elite horses
(Table 11.2). Horses with lameness and back problems had signifi-
cantly smaller hock angles than sound horses (Table 11.2).
Magnusson (1985c) found that small hock angles (sickle hocks)
were related to more synovial distentions in the stifle and hock joints
as well as to more curbs. This has also been reported by several other
authors (Smythe, 1963; Pritchard, 1965; Davidson, 1970; Beeman,
1973; Adams, 1974). Rooney (1968) and Hickman (1977) were of
B the opinion that sickle hocked horses more frequently experienced
bone spavin. Icelandic Toelter horses show a significant correlation
Fig 11.10  Comparison of horses with (A) a forward sloping femur and (B) a between small hock angles and bone spavin according to Eksell
vertical femur. Note the effect on the overall replacement of the hind limb. et al. (1998). Axelsson et al. (2001) found that Icelandic horses with

239
 11 The effects of conformation

B B

Fig 11.11  Horses with (A) small hock joint angle and (B) large hock joint Fig 11.12  Horses with (A) good and (B) poor hind limb conformation.
angle.

a larger tarsal angle had a lower prevalence of radiographic signs of It is concluded that sickle hocked horses, or horses with small hock
degenerative joint disease (DJD) in the distal tarsus confirming that joint angles, should be avoided.
the tarsal conformation is associated with OA in the distal tarsus, The slope of the hind pastern is, to some extent, influenced by
probably by altering the biomechanics of the distal tarsal joints. In the rest of the hind limb conformation. A straight hock is signifi-
addition, Björnsdóttir et al. (2004) examined Icelandic horses at cantly correlated to a more sloping pastern (Magnusson, 1985b).
young age by high detail radiography and histology and concluded Riding horses with soundness problems had a significantly steeper
that the development of OA in the centrodistal tarsal joint of young pastern. This might be an effect of a smaller hock angle in these
Icelandic horses seems to be due to poor conformation or joint horses; however, the difference in hock angle was not statistically
architecture rather than trauma or overloading. significant (Table 11.3) (Holmström, unpublished data). In contrast
A small hock angle generally has a negative effect on performance to the forelimb, a straighter hind fetlock joint has been shown to
in Thoroughbred racehorses (Holmström, unpublished work). The be correlated to a longer stride and swing duration at the trot (Back
negative correlation between hock angle and performance is stron- et al.1996). A straighter hind fetlock is more efficient in storing
ger on dirt tracks than on turf. All these findings confirm what was elastic energy, which contributes to more power from the hind limb.
said by Schmidt (1928), that a horse with a small hock angle will Generally, the hind limb conformation should be regarded as one
be able to step underneath itself but will not be able to carry weight unit when evaluated. There are many different combinations of
on the hind limbs due to decreased resistance and strength in angles that result in a ‘good hind limb conformation’, but it is
the hock. generally characterized by a somewhat flat pelvis, a forwardly
Studies using high-speed films have shown that in good dressage sloping femur, a normal to straight stifle and a normal to straight
horses there is considerable compression of the hock joint (~15°) hock (Fig. 11.12). If the horse is weak in one part this can be com-
during the midstance phase in trot and even greater compression pensated by strength somewhere else. The overall result is a strong
in passage and piaffe (Holmström et al., 1995a). In jumping horses hind limb that can endure stress, carry weight and store elastic strain
and racehorses, the compression is larger than in dressage horses. energy. Thus, looking at one characteristic at a time is not sufficient.

240
 Predicting performance and soundness by conformation analysis

Fig 11.13  Laterality and uneven feet: demo of a clinical provocation test.

All aspects of hind limb conformation must be taken into account. Thoroughbreds differed from other breeds, not only with regard to
This makes it more difficult to do an accurate evaluation subjec- segment lengths but also with regard to joint angles and deviations.
tively, and that is why inclusion of objective measurements improves The results from these studies illustrate the difficulties scientists face
the accuracy of the evaluation (Holmström et al., 1993). when designing conformational studies in horses. First of all subjec-
Preliminary studies on riding ponies and Icelandic Toelter horses tive evaluation of the conformation has limitations mainly because
indicate that the most favorable conformation for each type of of the difficulties to accurately and consistently register the subtle
performance is almost the same as that described above in sport differences between good and poor conformation within a breed
horses. The differences found are variations on the same theme with relatively homogeneous conformation and secondly quantita-
(Holmström, unpublished data). These variations are nevertheless tive analysis of the conformation has great potential as a predictive
important to investigate, and further studies will be carried out in tool only if used with the right approach. Quantitative studies
the future. of smaller groups of horses or random samples without inclusion
Extensive studies of the Thoroughbred racehorse conformation of a sufficient number of top performers might give the wrong
have shown that there are surprisingly small differences in confor- answers to the questions asked. Holmström (unpublished data)
mation between elite sport horses and top class racehorses. The has found that there are very small and subtle conformational
optimal hind limb shape for top performance is very similar, while differences between top class Thoroughbred racehorses and Warm-
there are some subtle differences in the forelimb conformation. blood sport horses when comparing grade 1 winners and Grand
Prix dressage horses and show jumpers, while when comparing
horses of lesser quality significant differences could be found.
Predicting performance and soundness by It is also very important to understand that conformation analysis
as a selection tool can only work as a negative predictor, i.e. sorting
conformation analysis out horses with conformation that gives them a significantly lower
chance of being successful in the sport or a significantly higher
There is no doubt that there are significant correlations between chance of injuries. The latter information however, would give
conformation and performance and soundness. The efficiency of a insurance companies an opportunity to improve their risk
selection procedure based on conformation evaluation is, however, evaluation.
highly dependent on the method used. Love et al. (2006) analyzed Applying conformation analysis as a negative predictor to wean-
a limited number of conformational traits assessed by a single vet- lings, yearlings and 2 year olds within the Thoroughbred racehorse
erinary observer against pedigree and racing records of 3916 Thor- business in the USA has proven that quantitative conformation
oughbred yearlings sold at public auctions during a 7-year period. analysis is a very useful tool in the selection of young horses
Only a weak association was found between performance and con- intended for top class performance. Between 75 and 80 % of the
formation. In the study by Santschi et al. (2006) the validation horses analyzed as weanlings or yearlings, and that subsequently
portion of their study showed that the coefficient of variation for won grade 1 races, at 8 furlongs or longer, were categorized as horses
subjective assessment of conformation was greater than 10%, which with ‘grade 1 conformation’ (Holmström, unpublished data).
is well below a clinically useful tool. Weller et al. (2006a,b,c) objec- There are many studies on the relationship between conforma-
tively evaluated the conformation of a group of 108 National Hunt tion and soundness in different breeds showing the complexity of
racehorses using a digital motion analysis system. It appeared that the problem. For example, Gnagey et al. (2006) evaluated the effect

241
 11 The effects of conformation

of standing tarsal angle on joint kinematics and kinetics and com- 1

pared tarsal kinematics and kinetics in horses with large, intermedi- 0.9
ate and small tarsal angles. They found that in horses with large 0.8
tarsal angles, less concussion was absorbed during the impact phase,
0.7
which may be a factor in the development of degenerative joint
disease. At the same time, the smaller net joint moment may reduce 0.6

Survival
the risk of plantar ligament desmitis, illustrating the delicate balance 0.5
between conformation, kinematics and kinetics of the tarsal joint 0.4
in relation to soundness (Baird & Pilsworth, 2001). 0.3
Recently, the ‘grassfoot’ phenomenon has been unraveled, as 0.2
foals with relatively longer limbs and shorter neck and head length
0.1
seem to have a preference in forelimb position when grazing leading
to an asymmetric development of their front feet, the so-called 0
0 1 2 3 4 5 6 7 8 9
uneven feet (Van Heel et al., 2006; Kroekenstoel et al., 2006; Ducro A Years in sport
et al., 2009a). The steep foot is habitually positioned behind and
the sloping foot is positioned more forward (Fig. 11.13). Studbooks
tend not to select horses that have this asymmetry. The effect of this 1
phenomenon has been evaluated at a population level comparing 0.9
studbook admission data, scoring the existence of uneven feet, with 0.8
what horses have delivered in sports (Ducro et al., 2009b). Length 0.7
of competitive life was shorter for jumping than for dressage. A 0.6

Survival
different set of risk factors was found for each level and discipline,
0.5
e.g. height at withers was a risk factor at basic level in dressage and
jumping, while pastern angle was a risk factor at the elite level of 0.4
jumping and dressage. The trait ‘uneven feet’ tended to shorten the 0.3
competitive life in dressage, but was a significant risk factor at the 0.2
elite level of jumping (Fig. 11.14). Davies and Watson (2005) 0.1
proved that there is also such a laterality asymmetry in the third 0
metacarpal bone and midshaft dimensions in Thoroughbred race- 0 1 2 3 4 5 6 7 8 9
horses (n = 40). They proposed that racehorses with longer right B Years in sport
MC3 bones were more able to control the loading of the right MC3
than the left during fast exercise. In Warmbloods, this phenomenon Fig 11.14  Survival curves for (A) dressage: basic dressage (n = 12 776):
has led to a locomotor asymmetry at a young adult age, possibly an even feet (n = 11 940; blue); uneven feet (n = 836; green); elite dressage
explanation for the so-called ‘natural asymmetry’ in the perfor- (n = 684): even feet (n = 657; orange), uneven feet (n = 27; red); and
mance of a horse at the beginning of their career that is a challenge (B) jumping performance: basic jumping (n = 8738): even feet (n = 8221;
blue), uneven feet (n = 517; green); elite jumping (n = 756): even feet
for every rider (Van Heel et al., 2010).
(n = 725; orange), uneven feet (n = 31; red); and the number of years
Further studies of conformational effects on soundness is neces-
horses were registered as being in competition.
sary but at the moment there are no data available that contradict
Reprinted from Ducro, B.J., Gorissen, B., van Eldik, P., Back, W., 2009b. Influence of foot
that belief that a conformation that facilitates top class performance
conformation on duration of competitive life in a Dutch Warmblood horse population.
and longevity of the horses’ careers in general also is positive for
Equine Vet. J. 41: 144–148, with permission from the Equine Veterinary Journal
the horses’ soundness.

References
Adams, O.R., 1974. Lameness in Horses, third detection of superior gait quality in grund der mechanischen verhältnisse.
ed. Lea and Febiger, Philadelphia, pp. young trotting warmbloods. Vet. Q. 16, Inauguraldissertation, Berlin.
1–32. S91–S96. Barr, A.R., 1994. Carpal conformation in
Anderson, T.M., McIlwraith, C.W., 2004. Back, W., Hartman, W., Schamhardt, H.C., relation to carpal chip fracture. Vet. Rec.
Longitudinal development of equine Bruin, G., Barneveld, A., 1995. Kinematic 134, 646–650.
conformation from weanling to age 3 years response to a 70-day training period in Barrey, E., Desliens, F., Poirel, D., Biau, S.,
in the Thoroughbred. Equine Vet. J. 36, trotting Dutch Warmbloods. Equine Vet. J. Lemaire, S., Rivero, J.L., et al., 2002. Early
563–570. 18 (Suppl.), 127–131. evaluation of dressage ability in different
Anderson, T.M., McIlwraith, C.W., Douay, P., Back, W., Schamhardt, H.C., Barneveld, A., breeds. Equine Vet. J. Suppl. 34, 319–324.
2004. The role of conformation in 1996. The influence of conformation on Beeman, G.M., 1973. Correlations of defects
musculoskeletal problems in the racing fore and hind limb kinematics of the in conformation to pathology in the horse.
Thoroughbred. Equine Vet. J. 36, 571–575. trotting Dutch Warmblood horse. Proc. 19th Ann. Meetg. Am. Assoc. Equine
Anon 1940. Riding Instructions for the Pferdeheilkunde 12, 647–650. Pract., Atlanta, Georgia, pp. 177–197.
Swedish Army. Baird, D.H., Pilsworth, R.C., 2001. Bengtsson, 1983. ABC om hästar, seventh ed.
Axelsson, M., Björnsdottir, S., Eksell, P., Wedge-shaped conformation of the LT:s Förlag, [In Swedish] Stockholm.
Häggström, J., Sigurdsson, H., Carlsten, J., dorsolateral aspect of the third tarsal bone Björnsdóttir, S., Ekman, S., Eksell, P., Lord, P.,
2001. Risk factors associated with hind in the Thoroughbred racehorse is 2004. High detail radiography and
limb lameness and degenerative joint associated with development of slab histology of the centrodistal tarsal joint of
disease in the distal tarsus of Icelandic fractures in this site. Equine Vet. J. 33, Icelandic horses age 6 months to 6 years.
horses. Equine Vet J. 33, 84–90. 617–620. Equine Vet. J. 36, 5–11.
Back, W., Barneveld, A., Bruin, G., Schamhardt, Bantoui, C., 1922. Messungen an trabern und Boldt, H., 1978. In: Haberbeck, H., Haberbeck,
H.C., Hartman, W., 1994. Kinematic die beurteiling der leistungsfähigkeit auf H.-R. (Eds.), Das dressurpherd.

242
 References

Verlagsgesellschaft GmbH, Lage-Lippe, pp. Grundler, C., 1980. Aussagewert verschiedener utvisa hästars brukbarhet till olika behov.
60–71. hilfsmerkmale zur beurteilung des zücht [In Swedish]. Stockholm, pp. 38–93.
Bourgelat, C., 1750. Elémens d’hippiatique au – und gebrauchswertes von Magnusson, L.-E., 1985a. Studies on the
nouveaux principes sur la connaissance et warmblutpferden. Dissertation, München. conformation and related traits of
sur la medicine de chevaux. T. 1, Lyon. Henninges, H.E., 1933. Untersuchungen über Standardbred trotters in Sweden. I. An
Churchill, E.A., 1962. Predisposing factors to den einfluss der körpermasse und objective method for measuring the equine
lameness in Standardbreds. Proceedings of gliedmassenvinkel auf die conformation. Thesis, SLU, Skara.
8th Annual Meeting of the American leistungsschrittlänge bei pommerschen Magnusson, L.-E., 1985b. Studies on the
Association of Equine Practitioners, warmblutstuten. Dissertation, Berlin. conformation and related traits of
Chicago, Illinois, pp. 191–194. Hickman, J., 1977. Farriery. J.A. Allen & Co. Standardbred trotters in Sweden. II. The
Clayton, H.M., Sha, D., Stick, J.A., Mullineaux, Ltd, London. variation in conformation of the
D.R., 2004. Three-dimensional carpal Holmström, M., Fredricson, I., Drevemo, S., Standardbred trotter. Thesis, SLU, Skara.
kinematics of trotting horses. Equine Vet. J. 1994. Biokinematic differences between Magnusson, L.-E., 1985c. Studies on the
36, 671–676. riding horses judged as good and poor at conformation and related traits of
Davidson, A.H., 1970. Some relationships of the trot. Equine Vet. J. 17 (Suppl.), 51–56. Standardbred trotters in Sweden. IV.
conformation to lameness and the Holmström, M., Fredricson, I., Drevemo, S., Relationship between conformation and
evaluation of the potential. Proceedings of 1995a. Biokinematic effects of collection in soundness in 4-year old Standardbred
16th Annual Meeting of the American the elite dressage trot. Equine Vet. J. 27, trotters. Thesis, SLU, Skara.
Association of Equine Practitioners, 281–287. Magnusson, L.-E., 1985d. Studies on the
Montreal, Quebec, pp. 399–404. Holmström, M., Fredricson, I., Drevemo, S., conformation and related traits of
Davies, H.M., Watson, K.M., 2005. Third 1995b. Variation in angular pattern Standardbred trotters in Sweden. V.
metacarpal bone laterality asymmetry and adaptation from trot in hand to passage Relationship between conformation and
midshaft dimensions in Thoroughbred and piaffe in the Grand Prix dressage horse. performance in 4-year old Standardbred
racehorses. Aust. Vet. J. 83, 224–226. Equine Vet. J. 18 (Suppl.), 132–137. trotters. Thesis, SLU, Skara.
Degen, K., 1953. Variationsstatistische Holmström, M., Magnusson, L.-E., Philipsson, Müller, J., Schwark, H.J., 1979.
körpermassuntersuchungen an 500 J., 1990. Variation in conformation of Merkmalsvarians und genetische
oldenburger zuchtstuten. Dissertation, Swedish Warmblood horses and bedingtkeit von im turniersport erfassten
Hannover. conformational characteristics of elite sport leistungsmerkmalen. Züchterische
horses. Equine Vet. J. 22, 186–193. weiterentwicklung der sportpferderassen.
Ducro, B.J., Bovenhuis. H., Back, W., 2009a.
Holmström, M., Philipsson, J., 1993. Vorträge de III Int. Wissenschaftliches
Heritability of foot conformation and its
Relationship between conformation, Symp., Leipzig, pp. 45–52.
relationship to sports performance in a
performance and health in 4-year old Murray, R.C., Walters, J.M., Snart, H., Dyson,
Dutch Warmblood horse population.
Swedish Warmblood Riding Horses. S.J., Parkin, T.D., 2010. Identification of risk
Equine Vet. J. 41, 139–143.
Livestock Prod. Sci. 33, 293–312. factors for lameness in dressage horses. Vet
Ducro, B.J., Gorissen, B., van Eldik, P., Back,
Hörman, J., 1837. Hästens exteriör. Hästen J. 184, 27–36.
W., 2009b. Influence of foot conformation
både naturhistoriskt och tekniskt betraktad. Neisser, E., 1976. Korrelative zusammenhänge
on duration of competitive life in a Dutch
[In Swedish] Skara, pp. 39–122. zwischen phänotyp und leistung beim
Warmblood horse population. Equine Vet.
J. 41, 144–148 Kroekenstoel, A.M., Van Heel, M.C.V., Van sportpferd. Arch. Tierzücht. 19, 51–60.
Weeren, P.R., Back, W., 2006. Nordin, S., 1980. Sören nordins travskola. Trav
Dušek, J., 1974. Beitrag zum problem der
Developmental aspects of distal limb och galoppronden 49, [In Swedish].
beziehung zwischen der höhe des pherdes
conformation in the horse: differences with
und der bewegungsmechanik bei den Pritchard, C.C., 1965. Relationship between
mature horses and the potential
hannoveranerhengsten. Bayer. Landw. Jb. conformation and lamemess in the equine
consequences of uneven feet in foals.
52, 209–213. foot. Auburn Vet. 22, 11–26; 29.
Equine Vet. J. 38, 652–656.
Dušek, J., Ehrlein, H.-J., v Engelhardt, W., Kronacher, C., Ogrizek, A., 1931. Exteriur und Richter, O., 1953. Korrelaten zwischen
Hörnicke, H., 1970. Beziehungen zwischen leistungsfähigkeit des pferdes und körpermaßen und leistungen bei dem auf
trittlänge, trittfrekvenz und geschwindigheit besondere berücksichtigung der amerikanischer grundlage gezogenen
bei pferden. Z. Tierzücht. ZüchtBiol. 87, gliedmassenvinklung und deutschen traber. Inauguraldissertation,
177–188. schrittlängenverhältnisse. Z. Tierzücht. Berlin.
Ehrengranat, A., 1818. Om hästens rörelser i ZüchtBiol. 23, 183–228. Rooney, J.R., 1968. Biomechanics of the
deras samband med ridkonsten. [In Krüger, L., 1957. Die estimmung der equine lameness. Cornell Vet. 58 (Suppl.),
Swedish] Lund. arbeitsfähigkeit bei pferd und rind durch 49–58.
Eksell, P., Axelsson, M., Broström, H., Ronéus, leistungsprüfungen, physiologische und Rossdale, P.R., Butterfield, R.M., 2006.
B., Häggström, J., Carlsten, J., 1998. psychologische. Messweret und durch die Conformation: a step along the pathway to
Prevalence and risk factors of bone spavin exteriurbeurteilung. Z. Tierzüct. ZüchtBiol. objective assessment. Equine Vet. J. 38,
in icelandic horses in Sweden: a 69, 289–320; 70: 1–20. 588–590.
radiographic field study. Acta Vet. Scand. Langlois, B., Froideveaux, J., Lamarche, L., Santschi, E.M., Leibsle, S.R., Morehead, J.P,
39, 339–348. Legault, P., Tassencourt, L., Theret, M., Prichard, M.A., Clayton, M.K., Keuler, N.S.,
Firth, E.C., Schamhardt, H.C., Hartman, W., 1978. Analyse de liaisons centre la 2006. Carpal and fetlock conformation of
1988. Measurements of bone strain in foals morphologie et l’aptitude au galop, au trot the juvenile Thoroughbred from birth to
with altered foot balance. Am. J. Vet. Res. et au saut d’obstacle chez le cheval. Ann. yearling auction age. Equine Vet. J. 38,
49, 261–265. Génét. Sél. Anim. 10, 443–474. 604–609.
Forsell, G., 1927. Om hästen, dess anatomiska Love, S., Wyse, C.A., Stirk, A.J., Stear, M.J., Schmidt, P., 1928. Handbok i hästbedömning.
byggnad, ytterlära, vård och sjukdomar. Calver, P.,. Voute, L.C., Mellor, D.J., 2006. Almquist and Wiksell, [In Swedish].
Hugo Gerbers Förlag, [In Swedish] Prevalence, heritability and significance of Uppsala, Stockholm.
Stockholm, pp. 34–51. musculoskeletal conformational traits in Schwark, H.J., Schreiber, K., Sasse, L., 1977.
Gnagey, L., Clayton, H.M., Lanovaz, J.L., 2006. Thoroughbred yearlings. Equine Vet. J. 38, Merkmalskorrelationen bei sportpferden.
Effect of standing tarsal angle on joint 597–603. Paper presented at: 28th Annual Meeting of
kinematics and kinetics. Equine Vet. J. 38, Magne, J.L., 1866. Valet af hästar eller the European Association of Animimal
628–633. beskrivning av alla de kännetecken som Production Brussels, H1.05, pp. 1–7.

243
 11 The effects of conformation

Sellet, L.C., Albert, W.W., Groppel, J.L., 1981. Van Heel, M.C.V., Kroekenstoel, A.M., Van Weller, R., Pfau, T., Babbage, D., Brittin, E.,
Forelimb kinematics of the Standardbred Dierendonck, M., Van Weeren, P.R., Back, May, S.A., Wilson, A.M., 2006a. Reliability
pacing gait. Proc. Equine Nutr. Physiol., pp. W., 2006. Uneven feet in a foal may of conformational measurements in
210–215. develop as a consequence of lateral grazing the horse using a three-dimensional
Smythe, R.H., 1963. Horses in action. Country behaviour induced by conformation. motion analysis system. Equine Vet. J. 38,
Life Ltd, London. Equine Vet. J. 38, 646–651. 610–615.
Steel CM, Hopper BJ, Richardson JL, Alexander van Vleck, L.D., Albrechtsen, R., 1965. Weller, R., Pfau, T., May, SA, Wilson, A.M.,
GR, Robertson ID., 2006. Clinical findings, Differences between appraisers in the New 2006b. Variation in conformation in a
diagnosis, prevalence and predisposing York type appraisal program. J. Dairy Sci. cohort of National Hunt racehorses. Equine
factors for lameness localised to the middle 48, 61–64. Vet. J. 38, 616–621.
carpal joint in young Standardbred Van Weeren, P.R., Crevier-Denoix, N., Weller, R., Pfau, T., Verheyen, K., May S.A.,
racehorses. Equine Vet. J. 38, 152–157. 2006. Equine conformation: clues to Wilson A.M., 2006c. The effect of
Van der Veen, G., 1996. Het paard in partjes. performance and soundness. Equine Vet. J. conformation on orthopaedic health and
Tirion uitgevers, The Netherlands. 38, 591–596. performance in a cohort of National Hunt
Van Heel, M.C., Van Dierendonck, M.C., Von Wagener, H., 1934. Untersuchungen an racehorses: preliminary results. Equine Vet.
Kroekenstoel, A.M., Back, W., 2010. spitzenpferden des spring – und J. 38, 622–627.
Lateralised motor behaviour leads to schulstalles der kavallerischule hannover. Wrangel, C.G., 1911–1913. Handbok för
increased unevenness in front feet and Scharper Verlag, Hannover. hästvänner. [In Swedish] Facsimile-uppl.
asymmetry in athletic performance in Weferling, C.G., 1964. Körpermessungen bei Bröd. Ekstrand Tryckeri AB, 1981, Lund,
young mature Warmblood horses. Equine oldenburgischen und ostfriesischen pp. 336–455.
Vet. J. 42, 444–450. warmblutstuten. Dissertation, Hannover.

244
 CHAPTER 12 

Genetic contributions to exercise and

athletic performance
Emmeline W. Hill, Bart J. Ducro, P. René van Weeren, Albert Barneveld,
Willem Back

Introduction recent developments in equine exercise genomics and the implica-

tions for equine athletic performance are reviewed.
The genomes of agricultural livestock species and other domestic
animals have been shaped by artificial selection for favorable traits Historical background and the
for thousands of years. The domestication of animal species led to
the manipulation of gene pools by selectively breeding on the basis pre-Mendelian era
that various traits were observed to be passed on from generation
to generation. In the past millennia, humans have produced many The relationship between conformation and gait and the observa-
unique phenotypic characteristics in domesticates that clearly tion that variation in many conformational traits is, to a large
demarcate breeds and differentiate livestock from their wild ances- degree, inherited, has formed the basis for the selection of perfor-
tors and related species. Recent rapid advances in genomics tech- mance horses for millennia. Still, this selection forms the basis for
nologies have provided platforms from which livestock scientists many breeds of horse today. As early as the 4th century BC the Greek
may understand the genetic variation responsible for such pheno- hippologist Xenophon gave detailed descriptions of the conforma-
typic adaptation and variation. The availability of the draft genome tional characteristics of a good performance horse. He made com-
sequence for the horse (Equus caballus) and associated genomics ments on the quality of the horn, the slope of the pastern, the
platforms is set to revolutionize the understanding of the genetic robustness of the metacarpal and metatarsal bones and the desired
variation underlying phenotypic variation. The variation in pheno- joint angles in fore and hind limbs. According to Xenophon a well-
type is attributed to structural variation that produces variation in developed hind quarter was the best sign that a horse would move
gene expression, which has clear phenotypic consequences. Dedi- fast and smoothly; this observation demonstrates a knowledge that
cated functional genomics tools for the horse have emerged since propulsion stems from the hind limbs. This insight is also evi-
around 2010. Equine genomics scientists can now utilize platform- denced by the observation that a short, strong lumbar region is
based tools such as the Illumina® EquineSNP BeadChip and gene preferred to a long back (cited by Schauder, 1923). After Xenophon,
expression microarrays as well as next-generation sequencing tech- all great hippologists stressed the same conformational characteris-
nologies to deepen our understanding of the molecular control of tics as desirable traits. This applies to the great Arab horsemen from
cellular function. In particular these tools may be used to identify the early Middle Ages, such as Akhi Hizam al-Furusiyah wa al-Khayl
key genetic determinants of exercise adaptation in the horse. who published on the conformation of horses in AD 860 (Dunlop
In relation to the equine locomotor system, genetics is important & Williams, 1996), and also to the famous European ecuyers from
at a number of levels. For example, conformation characteristics the 17th and 18th centuries such as William Cavendysh (1674) and
have been observed to have considerable heritable components and Jacques de Solleysel (1733).
have a strong relationship with athletic capacity. Also, there is a All modern horse breeds descend from common ancestors that
strong relationship between conformation and other performance- roamed the central Eurasian steppe some 6000 years ago (Levine,
determining characteristics, such as the capacity of the circulatory 1999). Natural selection in different biotopes and selection by man
and respiratory system as well as behavioral characteristics of the for different purposes has resulted in a wide variety of horse breeds,
horse. Additionally, genetics may have an important influence ranging from the small and sturdy Shetland pony, that stands on
on locomotor capacity by determining susceptibility to musculosk- average no more than 11 hands, to the Shire, ‘the great horse of the
eletal injury. Furthermore, there are a number of genetically Middle Ages,’ that carried heavily armored knights. Since horses
determined disorders that may directly influence locomotor were domesticated they have been selected for strength, speed and
performance. endurance-exercise traits. More recently the development of specific
In this chapter the historical role of horse breeding with respect breeds has resulted in selection for athletic phenotypes that enable
to locomotor performance is outlined. Developments in breeding in the use of the horse for riding, recreation, sport and racing. Some
the pre-Mendelian era and during the 20th century, when it became breeds are particularly suited to racing (e.g. Thoroughbred, Stan-
possible to introduce quantitative genetic methods to breeding, will dardbred, French trotter, Swedish trotter and Quarter Horse) while
be discussed. In particular, attention is paid to a number of disorders others have been strongly selected for jumping and other athletic
of the locomotor system that have a genetic background. Finally, phenotypes (e.g. Warmbloods).

245
 12 Genetic contributions to exercise and athletic performance

This is different when traits related to moving and jumping are

Selection for locomotor performance considered. Most studbooks also score gaits and elements of (free)
jumping during studbook admission, field tests and station perfor-
In horse breeding, selection has traditionally been based on con- mance tests. From a review of Thorén et al. (2006) it appeared that
formational characteristics and the evaluation of gait. These traits traits related to gaits showed moderate-to-good heritabilities as well
may be scored in a semi-quantitative way, but some degree of sub- as good genetic correlations to dressage performance in competi-
jectivity remains. This is in contrast to many other economically tion. Compared to gaits, jumping traits showed even higher herita-
important livestock species where traits such as milk production, bilities as well as higher genetic correlations to show jumping in
growth rate, fat thickness and food conversion can be easily mea- competition. Genetic parameter estimates were somewhat lower
sured and used as a basis for improved breeding programs. It is when records were collected at field tests rather than the more
generally considered that in the horse athletic performance and uniform station performance tests (Ducro et al., 2007). These
locomotor characteristics, such as gait, are polygenic traits that may results clearly indicate that information from studbook admissions,
also be largely influenced by environmental factors. Heritability is field tests and stationary performance tests are valuable additions
the term given to the proportion of variation in a particular trait to an efficient breeding program. Most studbooks have therefore
that is attributed to genetics. Estimates of heritability (h2) provide adopted such a breeding system.
a means to determine the relative importance of genetics and the
environment to a given characteristic.
Performance in the horse may be broadly divided into three New genomics technologies for trait mapping
categories: working, riding and racing (Hintz, 1980). With respect
to working, pulling has been considered by various authors and
and gene expression
estimates of heritability range from 0.12 ± 0.14 (Lonka, 1946) to For some time, equine genomics resources lagged behind those of
0.58 ± 0.20 (Varo, 1947) with a mean estimated heritability from other agricultural species, as the horse was not formally recognized
several reports of 0.25 (Hintz, 1980). Cutting performance or ‘cow as an economically important production species, or a model for
sense’ has received little attention so far, although Ellersieck et al. human disease. However, more recently, genomics tools for the
(1985) estimated a heritability of 0.19 ± 0.05 for this trait. With horse have surpassed those available for other livestock species, and
respect to riding performance in competition, most of the estimates the stage is now set to employ state-of-the-art genomics technolo-
are in the range of 0.10–0.30 (Ricard et al., 1998). gies to understand the genetic basis for athletic performance char-
Although racing performance may seem relatively straight- acteristics, including the complex phenotypes of conformation
forward to quantify using parameters such as number of races won, and gait.
best time and (total) earnings, many factors complicate this proce- With just a few markers mapped on the equine genome in the
dure. For instance, many methods use the ranking in a certain early millennium, there have been unexpected advances and a rapid
athletic event. However, although high ranking in a certain event explosion of information, perhaps up to 100-fold, during the past
may be synonymous with good performance for a given individual 10 years. A coordinated interest in horse genome mapping began
horse, this does not necessarily provide information about the in 1995, when an international workshop on genetics and disease
genetic value of this individual for the population, as much depends in the horse was held in Interlaken, Switzerland. Various general
on the level of the event in which the individual took part. Herita- genetics topics and a large number of inherited disorders were
bility estimates for transformed earnings or best time per kilometer discussed (Rossdale, 1995). The horse genome mapping project
range from 0.20 to 0.40 (Hintz, 1980). began in earnest later that year when ~70 scientists from over 20
Nowadays, conformation is recorded in a semi-quantitative way countries met in Kentucky, USA, to discuss the initial plan to con-
by most studbooks, using scoring systems. Traditionally, the scale struct a genetic map for the horse. It was agreed that resources,
of scoring has an appreciative character, e.g. the scale runs from knowledge and findings would be shared on a regular basis. There-
‘bad’ to ‘good’. At the end of the 1980s, the Royal Dutch Warm- after, a biannual International Horse Genome Mapping Workshop
blood Studbook introduced a system for the linear scoring of funded by The Dorothy Russell Havemeyer Foundation was estab-
conformational traits and gait. This system, together with the lished. During the past 15 years, hundreds of equine scientists,
assessment of some health-related items, is one of the pillars of the specialists and breeders have attended workshops in the USA, Aus-
Dutch selection system for stallions and their progeny (Barneveld, tralia, South Africa, Sweden, Ireland and UK.
1996). In this system a large number of conformation traits are Several genetic maps, including the linkage map (Guérin et al.,
scored on a 0–40 scale in which a population mean of 20 is 1999; Lindgren et al., 1998; Penedo et al., 2005; Swinburne et al.,
assumed. The rather fine scale and the use of more specific charac- 2000, 2006; Tozaki et al., 2007), radiation hybrid (RH) map (Chow-
teristics (such as slope of the shoulder, length of the back, slope of dhary et al., 1992), mapping of specific genes and assigning micro-
the pastern, etc.), makes a better estimation of heritability possible. satellite markers by RH panel (Beck et al., 2005; Böneker et al.,
In general, heritability was found to be low to moderate for confor- 2005a,b, 2006; Brenig et al., 2004; Bricker et al., 2005; Dempsey &
mation traits, ranging from 0.09 (slope of the pastern) to 0.26 Wagner, 1999; Dranchak et al., 2006; Klukowska-Rötzler et al.,
(length of the neck and position of the croup) (Koenen et al., 2006a, b; Leeb et al., 2005; Mickelson et al., 2004; Momozawa
1995). Also, in other studbooks linear scoring systems have been et al., 2005, 2007a,b; Müller et al., 2005a,b,c; Perrocheau et al.,
successfully introduced (Van Bergen & van Arendonk, 1993; Samoré 2006; Takahashi et al., 2004; Ward et al., 2003; Wittwer et al.,
et al., 1997). 2005), cytogenetic map by fluorescent in situ hybridization (FISH)
Some of these conformational characteristics are correlated with (Brinkmeyer-Langford et al., 2005; Chowdhary & Gustavsson,
performance. A long, sloping shoulder is positively correlated 1992; Goh et al., 2007; Gustafson-Seabury et al., 2005; Milenkovic
with dressage results (Holmström et al., 1990, 1994). A long and et al., 2002; Perrocheau et al., 2005; Raudsepp et al., 2004b), com-
strongly muscled croup is positively related to show jumping per- parative map (Brinkmeyer-Langford et al., 2005; Caetano et al.,
formance (Langlois et al., 1978). However, the vast majority of 1999; Chaudhary et al., 1998; Chowdhary & Gustavsson, 1992;
conformational traits that are scored using the linear system do not Milenkovic et al., 2002; Perrocheau et al., 2005; Raudsepp et al.,
directly correlate with performance. This fact, together with the low- 1996; Yang et al., 2003, 2004), BAC contig map (Brinkmeyer-
to-moderate heritability estimates for conformation (Preisinger Langford et al., 2008; Gustafson et al., 2003; Raudsepp et al.,
et al., 1991; von Butler-Wemken et al., 1992), indicate that confor- 2004b; Tallmadge et al., 2005) and other high-resolution chromo-
mation results should be considered of minor importance in a some maps including a comprehensive Y chromosome
direct selection for performance (Koenen et al., 1995). map (Brinkmeyer-Langford et al., 2008; Dempsey & Wagner, 1999;

246
 Selection for locomotor performance

Dierks et al., 2006; Goh et al., 2007; Gustafson-Seabury et al., 2005; (Gu et al., 2009). This work involved a genome scan of genetic
Lee et al., 2004; Perrocheau et al., 2006; Raudsepp et al., 2004a,b), variation at 394 autosomal and X chromosome microsatellite loci
have been contributed to by collaborations among international in four geographically diverse horse populations (Connemara,
laboratories. Akhal-Teke, Tuva and Thoroughbred). Positively selected loci were
A major breakthrough in equine genomics was the selection of identified in the extreme tail-ends of the distributions for popula-
the horse for genome sequencing by the US National Human tion genetic parameters and test statistics that identified departures
Genome Research Institute (NHGRI). The full length sequencing of from patterns of genetic variation expected under neutral genetic
the horse genome began in February 2006, and was performed at drift (Gu et al., 2009). Within the outlier loci there was a statistically
The Broad Institute, Massachusetts Institute of Technology and significant enrichment for genes involved in phosphatidylinositol
Harvard University in collaboration with the Equine Genome 3-kinase (PI3K) mediated signaling, insulin receptor signaling and
Sequencing Consortium. A single female Thoroughbred horse was lipid transport – biochemical pathways with well-characterized
chosen for sequencing using the whole genome shotgun approach. roles in adaptation to exercise. Furthermore, the importance of
The 6.8× coverage of the initial assembly (EquCab1.0) was released muscle function in the recent evolution of the Thoroughbred was
in January 2007, and the current assembly (EquCab2.0) was highlighted by a significant overrepresentation of sarcoglycan
released in September 2007 (Wade et al., 2009). complex and focal adhesion pathway genes located within the
In addition to whole genome sequencing, a single nucleotide selected regions. In summary, these data indicate that recent selec-
polymorphism (SNP) database (EquCab 2.0) containing >1 million tion in the ancestors of the present-day Thoroughbred population
SNPs (~1 per 1500 bp) was constructed by comparing sequences principally targeted genes associated with fatty acid oxidation,
from horses of disparate geographic origin, including Akhal-Teke, increased insulin sensitivity and muscle strength, highlighting the
Andalusian, Arabian, Icelandic, Quarter Horse, Standardbred and central role for muscle function and integrity in the Thoroughbred
Thoroughbred. Stemming from the horse genome sequencing athletic phenotype. More recently, using the Illumina® EquineSNP50
project new technologies such as the Illumina® EquineSNP Bead- BeadChip platform, the Equine Genetic Diversity Consortium, led
Chip (McCue et al., 2012) and various gene expression microarrays by researchers at the University of Minnesota, investigated signa-
(Bright et al., 2009) have been designed and may be used to under- tures of selection in over 30 horse breeds that had been selected for
stand trait association and gene function. different phenotypic traits (Petersen et al., 2013). In this study, the
The recent completion of the draft sequence of the horse genome most significant signature of selection in the Thoroughbred was
(Wade et al., 2009), combined with technological and method- found on ECA17, which was previously implicated as having selec-
ological advances in the analysis of complex genetic traits in humans tive importance in the Thoroughbred (Gu et al., 2009). Among all
(Steemers & Gunderson, 2007) has provided veterinary genetic epi- breeds, the strongest signature of selection was identified in the
demiologists with an array of tools for the study of diseases with a Paint and Quarter Horse; a 5.5-Mb region of ECA18, which centered
heritable component. Specifically, the advent of high-density single on the myostatin (MSTN) gene.
nucleotide polymorphism (SNP) based genotyping arrays and the To date, three genes with molecular functions relevant to physi-
concomitant growth in knowledge of the haplotypic structure of ological processes important for exercise have been reported to be
mammalian genomes has led to the adoption of genome-wide associated with racing performance, including the myostatin gene
association studies (GWAS) for disease gene mapping (Frazer et al., [MSTN] (Hill et al., 2010b,; Tozaki et al., 2010a,b), the cytochrome
2007; Orr & Chanock, 2008). These studies, which have recently c oxidase, subunit 4, isoform 2 gene [COX4I2] (Gu et al., 2010) and
been successful in the identification of large numbers of genetic loci the pyruvate dehydrogenase kinase isozyme 4, mitochondrial gene
contributing to disease in humans, exploit the high degree of cor- [PDK4] (Hill et al., 2010c). A variant in the genomic sequence for
relation between genetic variants at any particular region of a chro- PDK4 is the first example of a statistically significant association of
mosome to efficiently interrogate the entire genomes of large a SNP with elite race winning performance (Hill et al., 2010). In a
numbers of samples using a minimally redundant set of so-called set of n = 278 Thoroughbred samples, segregated by retrospective
‘tagging-SNPs’ (Howie et al., 2006). The GWAS paradigm in human racetrack performance into elite performers and non-winners, a SNP
studies has been to genotype hundreds of thousands of SNPs in at the PDK4 gene locus (g. 38973231A>G) was significantly associ-
many thousands of affected and unaffected individuals. These large ated with elite racing ability (p = 0.0004, odds ratio = 1.97, C.I. (95)
numbers are a result of the hypothesis that multiple loci with a = 1.35–2.87), with the A : A and A : G genotypes more common
range of effect sizes contribute to the etiology of complex diseases; among elite (70%) than non-elite (47%) racehorses. On average,
large numbers of samples are thus required to attain the requisite the A : A and A : G genotypes had a 16.2–16.6 lb handicap advantage
statistical power for unambiguous detection of such loci. Charlier over G : G horses. As the PDK4 gene locus was identified among a
et al. (2008) and Karlsson et al. (2007) have demonstrated that set of positively selected loci in Thoroughbreds (Gu et al., 2009),
association based mapping can be effective in the detection of single has functions known to be relevant to exercise (Wende et al., 2005;
locus recessive traits in animals using much smaller numbers of Scarpulla, 2008; Pilegaard & Neufer, 2004), and in horse skeletal
individuals. muscle is significantly differentially regulated post-exercise (Eivers
The advances in equine genomics have also rapidly enabled the et al., 2010; Hill et al., 2010) this gene variant represents a promis-
identification of genomic sequence variation associated with ath- ing opportunity to integrate molecular genetic information in selec-
letic performance phenotypes in Thoroughbreds (Gu et al., 2009; tion and decision-making processes in the Thoroughbred.
Hill et al., 2010b,d; Tozaki et al., 2010a,b). Thoroughbred horses The locus that has been most extensively studied in regard to an
represent a unique opportunity to identify genomic contributions athletic performance trait contains the gene encoding myostatin
to exercise adaptation and athletic performance as Thoroughbreds (MSTN). Three studies have identified variation at this locus associ-
have been selected for exceptional racing performance for 300 years ated with optimum race distance in European, North American and
and stem from a small number of founder animals. Intense selec- Australasian Thoroughbreds (Hill et al., 2010; Binns et al., 2010)
tion for elite racing performance in the Thoroughbred has resulted and with performance rank and lifetime earnings among Japanese
in a number of adaptive physiological phenotypes relevant to exer- Thoroughbreds (Tozaki et al., 2010a,b). Myostatin is a growth and
cise; however, the underlying molecular mechanisms responsible differentiation factor that functions as a negative regulator of skel-
for these characteristics are not yet well understood. etal muscle mass development. In several mammalian species,
The first report of genes and functional groups of genes contribut- including cattle, sheep, dogs and mice, muscle hypertrophy pheno-
ing to the athletic phenotype in the Thoroughbred employed a types are associated with sequence variants in the MSTN gene
population genetics-based approach to identify regions of the Thor- (Grobet et al., 1997; McPherron et al., 1997; McPherron & Lee,
oughbred genome that have been selected for exercise-relevant traits 1997; Schuelke et al., 2004; Mosher et al., 2007).

247
 12 Genetic contributions to exercise and athletic performance

Employing a candidate gene approach, Hill and colleagues identi- microarrays and through the use of next-generation sequencing
fied a sequence polymorphism in intron 1 of the MSTN gene technologies. Using equine-specific cDNA microarrays McGivney
(g.66493737C>T) that may be used to predict sprinting ability and and colleagues (2009) identified novel genes and key regulatory
racing stamina in Thoroughbred horses (Hill et al., 2010). In a set of pathways responsible for exercise adaptation in skeletal muscle.
n = 197 elite race winning Thoroughbreds, a very strong association Skeletal muscle biopsy samples were collected from a cohort of
with best race distance has been observed (p = 3.28 × 10−13). Two Thoroughbred horses at rest (T0) and at two time points following
alleles were observed at this biallelic SNP: a ‘C’ allele and a ‘T’ allele, a single bout of treadmill exercise (immediately post-exercise (T1),
with the ‘C’ allele more than twice as frequent in the short distance and 4 h post-exercise (T2) (Fig. 12.1)). While only two genes had
(≤7 furlongs) cohort of animals compared to the long distance (>8 increased expression at T(1) (p < 0.05), by T(2) 932 genes had
furlongs) group (0.75 and 0.34 respectively), corresponding to an increased (p < 0.05) and 562 genes had decreased expression
odds ratio of 5.81 (Hill et al., 2010d). It was observed that C : C horses (p < 0.05). Functional analysis of genes differentially expressed
are suited to fast, short-distance races; C : T horses compete favorably during the recovery phase (T2) revealed an over-representation of
in middle-distance races; and T : T horses have greater stamina. Evalu- genes localized to the actin cytoskeleton and with functions in the
ation of retrospective racecourse performance (n = 142) and stallion MAPK signaling, focal adhesion, insulin signaling, mTOR signaling,
progeny performance predict that C : C and C : T horses are more p53 signaling and type II diabetes mellitus pathways (Table 12.1).
likely to be successful 2-year-old racehorses than T : T animals. His- At T(1), using a less stringent statistical approach, an over-
tological analysis of gluteal muscle biopsies in Quarter horses representation of genes involved in the stress response, metabolism
showed that the C-allele was associated with higher Type 2B and and intracellular signaling was observed (Table 12.2). These find-
lower Type 1 muscle fiber proportions, demonstrating a functional ings suggested that protein synthesis, mechanosensation and muscle
consequence for selection this locus. Each copy of the C-allele remodeling contribute to skeletal muscle adaptation towards
increased Type 2B muscle fibres by 4.79% (Petersen et al., 2013). improved integrity and hypertrophy.
The importance of this locus in determining the type of racing To evaluate changes in gene expression following multiple exer-
an individual is most suited to, was confirmed in a genome wide cise bouts (i.e. conditioning or training) McGivney et al. (2010)
SNP association study (GWAS) using the Illumina® EquineSNP50 used digital gene expression (DGE) mRNA tag profiling, which is
BeadChip. A study using a set of n = 118 elite Thoroughbreds identi- based on the high-throughput sequencing by synthesis technology
fied the genomic region on chromosome 18 containing the MSTN available from Illumina Inc. (Mardis, 2008; Fox et al., 2009; Moro-
gene as the highest ranked region for optimum race distance and a zova et al., 2009), to characterize the assembly of genes expressed
set of seven SNPs within a 1.7 Mb region that reached genome-wide in equine skeletal muscle and to identify the subset of genes that
significance (Hill et al., 2010d). A comparison of trait association were differentially expressed following a 10-month period of exer-
in the same set of samples demonstrated the superior power of the cise training. Skeletal muscle biopsies were collected from the
g.66493737C>T SNP (Punadj. = 1.02 × 10−10) for the prediction of best gluteus medius of seven Thoroughbred horses at rest at two time
race distance when compared with the best marker on the array: points: T(1)-untrained (9 ± 0.5 months old) and T(2)-trained (20
BIEC2-417495 (Punadj. = 1.61 × 10−9). A genetic test is now available ± 0.7 months old). This study found that the most abundant mRNA
to horse breeders and trainers for this polymorphism. The Equi- transcripts in the muscle transcriptome were those involved in
nome Speed Gene Test may be used to make a prediction about the muscle contraction, aerobic respiration and mitochondrial func-
optimum race distance for an individual and may be used to tion. A previously unreported over-representation of genes related
improve decision-making in selection, breeding and training. to RNA processing, the stress response and proteolysis was observed.
Functional genomics studies provide a mechanism to further Following training 92 tags were differentially expressed of which 74
understand the importance of genes and gene functions in exercise were annotated. Sixteen genes showed increased expression, and
adaptation. Gene expression studies in an exercise context in the among the 58 genes with decreased expression, tags representing
horse are growing (Eizema et al., 2005; Jose-Cunilleras et al., 2005; the gene encoding myostatin (MSTN), had the greatest decrease
Barrey et al., 2006; Mucher et al., 2006; McGivney et al., 2009; (−4.2-fold, p = 0.0043). This demonstrates the key role of the myo-
Eivers et al., 2010; Hill et al., 2010a; Martin et al., 2010; McGivney statin protein in the adaptive response and its function as a negative
et al., 2010). The products of many genes are likely to influence regulator of muscle development. Functional groups displaying
system-wide physiological responses. However, the protein products highly significant increased expression included mitochondrion,
of two genes have been identified as key regulators of the adaptive oxidative phosphorylation and fatty acid metabolism, while func-
response to exercise in humans and model species. These are the tional groups with decreased expression were mainly associated
hypoxia inducible factor 1, alpha subunit (basic helix-loop-helix with structural genes and included the sarcoplasm, laminin complex
transcription factor) gene (HIF1A), which encodes HIF-1α and the and cytoskeleton. Recently, the Illumina® EquineSNP50 BeadChip
peroxisome proliferator-activated receptor gamma, coactivator 1 platform was used in a GWAS to identify a gene having a major
alpha gene (PPARGC1A), which encodes PGC-1α (for reviews effect on gaitedness in horses. Andersson and colleagues (Anders-
Bonen, 2009; Gibala, 2009; Lundby et al., 2009; Yan, 2009; Lira son et al., 2012) identified a premature stop codon in the doublesex
et al., 2010; Olesen et al., 2010). Eivers and colleagues (2010) have related mab transcription factor gene 3 (DMRT3 gene), which influ-
reported the gene expression responses of a panel of HIF-responsive enced the pattern of locomotion in Icelandic breeds. The presence
genes in skeletal muscle biopsies collected from the gluteus medius of the mutation allows for the performance of alternate gaits, i.e.
before and after a standardized incremental-step treadmill exercise instead of the two-beat contralateral gait of the trot, some horses
test in untrained Thoroughbred horses. Analyses of mRNA profiles perform the pace, a two-beat ipsilateral gait. Other natural varia-
revealed significant transcriptomic differences 4 h post-exercise (T2) tions in movement include four-beat ambling gaits characteristic of
for the CKM, COX4I1, COX4I2, PDK4, PPARGC1A, and SLC2A4 the Tennessee Walking Horse, Peruvian Paso, Paso Fino, and others,
genes relative to pre-exercise levels. The observed relationships with with unique variations in rhythm between breeds. This finding
measured physiological variables (VHRmax and [La]peak) indicated was validated in an investigation of genomic regions under selec-
that local transcriptional microadaptations influence the overall tion in gaited breeds (Petersen et al., 2013). A 186kb region on
athletic phenotype. Also, the data highlighted the roles of genes ECA23 spanning the DMRT3 gene locus was found to have been
responsible for the regulation of oxygen-dependent metabolism, targeted by selection in breeds with the ability to perform alterna-
glucose metabolism, and fatty acid utilization in equine skeletal tive gaits. Together these studies provide strong evidence that
muscle adaptation to exercise. DMRT3 is a major effect gene for locomotion/gait phenotypes;
Employing global approaches to gene expression studies is now however specific variations in gait among breeds may be modified
possible with the availability of equine-specific gene expression by other loci.

248
 Selection for locomotor performance

1000
800
600

No. D.E. Probes

400
200
0
-200
-400

T0 T1 T2
~10 min exercise <10 min post-exercise 4 hrs post-exercise

Stimulus Over-represented functional groups Recovery/adaptive response

T1 T2
Ribosome
Intramolecular Return to homeostasis,
Metabolic due to Oxidative phosphorylation
oxidoreductase activity enhanced response to
endurance exercise Proton-transporting ATP
Insulin signaling pathway future bouts of exercise
synthase complex

Heat shock protein binding

Metabolic or mechanical Recovery from oxidative
Protein folding
due to endurance or Response to stress and mechanical stress
MAPK signaling pathway
resistance exercise associated with exercise
P53 signaling pathway

Actin cytoskeleton
Stress fiber
Mechanical due to Muscular repair,
Cell cycle
resistance exercise muscle hypertrophy
Focal adhesion
mTOR signaling pathway

Fig 12.1  Skeletal muscle biopsy samples collected from a cohort of Thoroughbred horses at rest (T0) and at two time points following a single bout of
treadmill exercise (immediately post-exercise (T1), and 4 h post-exercise (T2)).
Reprinted from McGivney, B.A., Eivers, S.S., MacHugh, D.E., MacLeod, J.N., O’Gorman, G.M., Park, S.D., Katz, L.M., Hill, E.W., 2009. Transcriptional adaptations following exercise in thoroughbred
horse skeletal muscle highlights molecular mechanisms that lead to muscle hypertrophy. BMC Genomics 10, 638, http://www.biomedcentral.com/1471-2164/10/638 ©McGivney et al; licensee
BioMed Central Ltd.

Table 12.1  Over-represented functional groups of differentially expressed genes relative to all genes expressed in muscle were obtained
using the online tool DAVID. Subsequent clustering of functional groups was aided by the functional group clustering tool in DAVID

Stimulus Clusters of over-represented Number of Number Adaptive response

functional groups groups of genes
Metabolic due to Oxidative phosphorylation 34 92 Enhanced aerobic capacity
endurance exercise Increased Mt volume
Mitochondria 18 162 Enhanced fatty acid metabolism
Fatty acid metabolism 14 25
Mechanical due to Muscle development 16 27 Increase muscle mass
resistance exercise Muscular remodeling
Actin cytoskeleton 28 81
Increased protein turnover
Development/apoptosis 17 185 Enhanced contractile properties

Amino acid metabolism 14 60

Muscle contraction 32 42

Mt, mitochondrial.

249
250
Table 12.2  Differential gene expression between pre and post-exercise time-points in a panel of exercise-relevant genes located in positively selected (Dh/SD and FST) genomic regions
12

in Thoroughbred horses
T0 Vs T1 T0 Vs T2

Molecular function Gene symbol FC P value FC P value Chr Locus Dh/sd P FST P
Angiogenesis ANGPT2 1.29 0.051 4.35 0.001 27 VHL150 −2.863 0.017 0.309 0.038
Carbohydrate/Glucose metabolism PDK2 1.24 0.234 1.51 0.033 11 TKY033 −2.916 0.002 0.392 0.013
PDK3 1.07 0.583 −1.16 0.486 X Lex026 −3.890 0.006 0.020 NS
PDK4 1.80 0.010 2.19 0.001 4 TKY222 −6.117 0.000 0.450 0.005
Fatty acid metabolism/Gluconeogenesis/Glycolysis ADHFE1 1.54 0.020 1.51 0.060 9 COR008 −9.436 0.000 0.258 NS
Fatty acid biosynthesis/Insulin signaling pathway ACACA −1.1 0.379 −1.18 0.342 11 TKY033 −2.916 0.002 0.392 0.013
ACACB −1.05 0.821 −1.02 0.929 8 AHT025 −1.360 NS 0.319 0.023
Insulin signaling pathway FOXO1A 2.17 0.008 3.69 0.002 17 NVHEQ079 −3.602 0.007 0.380 0.018
GRB2 −1 0.985 −1.06 0.517 11 NVHEQ040 −3.848 0.007 0.170 NS
IRS1 – – – – 6 UMNe197 −2.571 0.014 0.315 0.028
PRKAR2B −1.05 0.714 −2.04 0.040 4 NVHEQ029 −0.272 NS 0.305 0.041
PTPN1 −1.01 0.949 −1.1 0.707 22 HMS047 −4.444 0.004 0.135 NS
Genetic contributions to exercise and athletic performance

SOCS3 4.09 0.011 4.57 0.010 11 NVHEQ040 −3.848 0.007 0.170 NS

Oxidative phosphorylation ATPSH 1.14 0.221 1.13 0.26 11 NVHEQ040 −3.848 0.007 0.170 NS
ATP6V1A 1.07 0.417 1.58 0.008 19 ASB011 −.3646 0.007 0.039 NS
ATP6V1G1 −1.27 0.079 −1.42 0.071 25 TKY316 −6.775 0.000 0.352 0.020
COX4I1 1.09 0.424 1.08 0.450 3 LEX057 −4.752 0.002 0.078 NS
NDUFA13 −1.8 0.001 −1.09 0.667 21 AHT059 −3.956 0.005 0.160 NS
NDUFA8 1.17 0.142 1.2 0.139 25 TKY316 −6.775 0.000 0.352 0.020
Phosphatidylinositol signaling system/VEGF PIK3R1 1.19 0.085 1.42 0.053 21 AHT059 −3.956 0.005 0.160 NS
signaling pathway
PPAR signaling PPARA 1.17 0.327 1.02 0.914 28 UCDEQ425 −3.956 0.005 0.160 NS
Insulin signaling pathway /PPAR signaling /Pyruvate PCK1 – – – – 22 HMS047 −4.444 0.004 0.135 NS
metabolism
Skeletal development BMP7 – – – – 22 HMS047 −4.444 0.004 0.135 NS
MEF28 1.07 0.639 −1.20 0.323 21 AHTO59 −3.956 0.005 0.160 NS
 Selection for locomotor soundness

Selection for locomotor soundness comprising 23 116 records of horses for which their conformation
scores and duration of their sports career were available. Survival
analysis was used to determine which of the conformation traits
A discrimination should be made between inherited conforma- had a significant effect on duration of sports career in dressage and
tional traits that may predispose to lameness or other aberrations jumping at basic and elite level. Duration of competitive life was
of the locomotor system and genetically determined disorders of shorter for jumping than for dressage. A different set of risk factors
the musculoskeletal system that will affect performance. was found for each level and discipline. The trait ‘uneven feet’
tended to shorten the competitive life in dressage, but was a signifi-
Conformational lameness-causing disorders cant risk factor at the elite level of jumping. Thus, limb conforma-
tion and, in particular, the conformation of the distal limb, are
Stashak (1987a) describes conformation of the horse as ‘the key to important for duration of competitive life. From the prevalence of
its method of progression’. He then elaborates on a number of faulty uneven feet in sports disciplines, it may be concluded that this is
conformations that may predispose to various pathologic condi- an undesirable trait, particularly at the elite level of jumping, since
tions of the musculoskeletal system: for instance a toe-out confor- uneven feet have a detrimental effect on the duration of competitive
mation results in a greater likelihood of limb interference and life in a sport horse population. This study provided evidence that
plaiting; a palmar deviation of the carpal joints (‘calf knees’) may the conformation trait uneven feet has a negative effect on Warm-
predispose to slab fractures of the carpal bones; and the commonly blood jumping performance and, therefore, breeders should be
seen cow hocked conformation may lead to bone spavin. Apart encouraged to avoid this phenomenon at foal age. Moreover, Ducro
from these conformation-related disorders, there are direct geneti- et al. (2009b) assessed the prevalence and heritability of uneven
cally determined aberrations of the musculoskeletal system, which feet and its genetic relationship to other conformation traits as well
are much rarer than conformational imperfections. Some of these as to sporting performance later in life in Warmblood riding horses.
have been known for a long time to have a genetic basis, but have Warmblood horse studbooks aim to breed horses with a conforma-
only recently been elucidated using modern molecular genetic tech- tion that will enable elite future performance, but reduce the risk
niques. Ducro et al. (2009a) investigated the significance of foot of injuries and lameness. Negative conformational traits, such as
conformation at young age to duration of the career of sport horses. asymmetrical or ‘uneven’ forefeet would possibly diminish perfor-
Warmblood horse studbooks aim to breed horses with a conforma- mance. The databases of the Royal Dutch Warmblood Studbook
tion that will enable elite future sports performance, but reduce the (KWPN, n = 44 840 horses) and Royal Dutch Equestrian Sports
risk of early retirement due to lameness. Negative conformational Federation (KNHS, n = 33 459 horses in dressage and n = 30 474
traits, such as asymmetrical or ‘uneven’ forefeet may possibly horses in show jumping) were linked through the unique number
shorten the career of sport horses. Databases of the Royal Dutch of each registered horse (Table 12.3). Therefore, heritabilities and
Warmblood Studbook (KWPN) and of the Royal Dutch Equestrian genetic and phenotypic correlations could be estimated from the
Sports Federation (KNHS) were matched and resulted in a dataset scores of the jury at studbook admission and the sports

Table 12.3  Means and description of lower and upper value of traits scored at studbook entry (n = 44 480) and graded in sports
(n = 33 459 in dressage, n = 30 474 in jumping)

Lower value Upper value Mean

Value Description Value Description
Score at studbook entry
Neck length −40 Short −0 Long −19.0
Forelimb −40 Calf kneed −0 Buck kneed −18.5
conformation
Pastern angle −40 Upright −0 Weak −20.5
Hoof shape −40 Narrow −0 Broad −19.9
Heel height −40 Low −0 High −19.1
Limb quality −40 Blurred −0 Dry −19.8
Bone circumference −40 Light −0 Heavy −19.4
Conformation grade 40 Bad 100 Good 67.5
Uneven feet (no/yes) 0 Absent 1 Present 0.05
Height at withers (cm) 158.3 178.2 164.8

Grade in sports
Dressage ranking* 0 Bad 200 Good 52.5
Jumping ranking 0 Bad 200 Good 43.0

*Original marks, before analysis square root transformed.

Reprinted from Ducro, B.J., Bovenhuis, H., Back, W., 2009. Heritability of foot conformation and its relationship to sports performance in a Dutch Warmblood population. Equine
Vet. J. 41, 139–143, with permission from the Equine Veterinary Journal.

251
 12 Genetic contributions to exercise and athletic performance

Table 12.4  Heritabilities (diagonal, bold and underlined) and genetic (below diagonal) and phenotypic (above diagonal) correlations
between traits scored at studbook entry (n = 44 480) and graded in sports (n = 33 459 dressage (Dre), n = 30 474 jumping (Jum))

HW NL UF FC PA HS HH LQ BC CG Jum Dre
Height at withers 0.67 −0.12 0.01 0.04 −0.05 0.19 0.03 −0.17 0.12 0.19 −0.02 0.14
Neck length −0.44 0.23 −0.01 −0.12 0.08 −0.03 0.04 0.10 −0.01 −0.30 0.02 0.04
Uneven feet −0.03 0.10 0.12 −0.01 −0.06 −0.17 0.05 −0.03 0.04 −0.06 −0.01 −0.01
Forelimb conformation −0.06 0.16 −0.05 0.16 0.00 −0.04 0.01 0.10 0.00 0.08 −0.01 −0.08
Pastern angle −0.03 0.01 −0.30 0.21 0.17 0.13 −0.16 0.10 −0.04 −0.04 −0.12 −0.04
Hoof shape 0.28 0.01 −0.49 −0.12 0.27 0.27 0.07 0.00 0.25 0.03 0.09 −0.07
Heel height −0.01 0.15 0.47 −0.19 −0.42 −0.41 0.16 0.08 0.06 0.15 0.20 0.04
Limb quality −0.29 0.22 −0.12 0.22 0.22 0.04 −0.01 0.19 −0.25 0.33 0.04 0.05
Bone circumference 0.35 0.02 0.20 −0.17 −0.16 0.50 0.18 −0.58 0.24 −0.13 0.09 0.03
Conformation grade 0.35 −0.59 −0.04 0.25 0.15 0.08 0.28 0.67 −0.23 0.30 0.04 0.19
Jumping ranking 0.01 0.14 −0.12 0.07 0.01 0.11 0.19 0.20 −0.09 0.29 0.14 ne
Dressage ranking 0.33 0.32 −0.09 −0.01 0.07 −0.01 0.15 0.36 −0.13 0.67 ne 0.14

BC, bone circumference; CG, conformation grade; FC, forelimb conformation; HH, heel height; HS, hoof shape; HW, height at withers; LQ, limb quality; NL, neck length; PA,
pastern angle; UF, uneven feet (%).
Standard errors of estimates were below 0.03; ne, not estimated.
Reprinted from Ducro, B.J., Bovenhuis, H., Back, W., 2009, Heritability of foot conformation and its relationship to sports performance in a Dutch Warmblood population. Equine
Vet. J. 41, 139–143, with permission from the Equine Veterinary Journal.

Table 12.5  Least square means of traits scored at studbook entry for level (basic and elite) and for sports discipline (n = 33 459 dressage,
n = 30 474 jumping)

Dressage Jumping
Basic Elite Prob Basic Elite Prob
Uneven feet (%) 6.20 5.78 0.78 6.43 5.32 0.32
Height at withers 165.6 166.1 <0.0001 165.6 165.6 0.4
Neck length −18.36 −17.90 <0.0001 −18.55 −18.00 <0.0001
Forelimb conformation −20.02 −19.97 0.14 −19.91 −19.98 0.43
Pastern angle −20.71 −20.42 0.006 −20.61 −20.71 0.05
Hoof shape −18.99 −18.91 0.62 −18.97 −18.85 0.02
Heel height −19.38 −19.28 0.05 −19.47 −19.14 0.005
Limb quality −19.66 −19.10 <0.0001 −19.62 −19.28 <0.0001
Conformation grade 67.62 69.27 <0.0001 67.53 68.33 <0.0001

Prob, probability of equal means of basic and elite level in each discipline.
From Ducro et al., 2009b.

performance of that population in dressage and jumping over the mean prevalence in the offspring of registered breeding stallions
period 1990–2002. The prevalence of uneven feet was 53% on (Fig. 12.2). Because of weak genetic correlations, the increased
average, and increased from under 4.5% during the first 3 years of prevalence is not directly associated with selection for better sports
recording to over 8% in the years from 2000 onwards. Heritability performance or higher conformation grade. If the trait ‘uneven feet’
estimates of foot conformation traits were moderate and ranged arises from a disproportionate relationship between height at the
from 0.16 for heel height to 0.27 for hoof shape (Table 12.4). The withers and neck length, then selection on conformation grade
genetic correlation between the trait of uneven feet and perfor- might result in development of uneven feet (Fig. 12.3). In general,
mance in competition was negative but weak: −0.09 with dressage limb conformation has a moderate genetic relationship to confor-
and −0.12 with show jumping (Table 12.5). Predisposition to mation grade and foot conformation traits have a genetic relation-
uneven feet can be reduced by selection, as there is a difference in ship to sporting performance. Reducing occurrence of uneven feet

252
 Selection for locomotor soundness

by selection is possible, without limiting progress in sport breeding, illnesses, and concurrent myopathies. Valberg et al.
performance. (1996) gave evidence for a familial basis for polysaccharide storage
myopathy and associated exertional rhabdomyolysis in Quarter
Horse-related breeds, the pattern of inheritance of which resembled
Neuromuscular lameness-causing disorders an autosomal recessive disorder.
Many monogenic anomalies in the horse have been described, a Some disorders are inherited as dominant traits. An example that
minority of which are related to the locomotor system (Galizzi Vec- has been known for centuries in both animals and man is hereditary
chiotti Antaldi, 1980a,b). Most of these disorders are inherited as multiple exostoses (Li et al., 1989; Leone et al., 1987). In the horse
autosomal recessive genes. Björck et al. (1973) described progres- the condition is generally inherited as a single autosomal dominant
sive congenital cerebellar ataxia in the Gotland pony breed as being gene (Gardner et al., 1975; Shupe et al., 1979), though some report
inherited as an autosomal recessive gene with full penetrance. In that, as in man, three genes are involved, two autosomal and one
Arabs a clinically similar condition described by Gerber et al. X-linked (Monteiro & Barata, 1980). The American Association of
(1995) caused almost total hypoplasia or atrophy of the Purkinje Equine Practitioners (AAEP) has listed genetic tests for nine single
cell layer. A lethal form of arthrogryposis (‘muscle contracture’) gene diseases (Nollet & Deprez, 2005; Finno et al., 2009) and
associated with polydactylia in the Norwegian Fjord horse was this number is likely to increase in the coming years as the new
described by Nes et al. (1982) as an autosomal recessive mutation, genomics tools are utilized to identify the genetic variants underly-
whereas Buoen et al. (1997) suggested a possible relation to auto- ing key equine disorders. ‘Autosomal dominant’ disorders include
somal trisomy. Collinder et al. (1997) found for the equine hyperkalemic periodic paralysis (HYPP) in the Quarter Horse, type
rhabdomyolysis syndrome (RHA) in Standardbreds that gene fre- 1 polysaccharide storage myopathy (PSSM) in different breeds, and
quencies for several markers in the RHA groups differed signifi- malignant hyperthermia in Quarter Horse-related breeds (Aleman
cantly from those estimated for the total population. A et al., 2009; Magdesian, 2009). ‘Autosomal dominant’ disorders
rhabdomyolysis risk group could be characterized using 4 or 5 include overo-lethal-white syndrome in the Paint Horse (McCabe
genetic marker loci. Beech and Haskins (1987) described a neuroax- et al., 1990; Blendinger et al., 1994), combined immunodeficiency
onal dystrophy in the Morgan. Aleman et al. (2009) proved that in Arabian Horses (McGuire et al., 1974; Mottironi et al., 1981;
Malignant Hyperthermia (MH) is a potentially fatal disease of Bernoco & Bailey, 1998), glycogen branching enzyme deficiency
Quarter Horses that could be triggered by halogenated anesthetics (GBED) in Quarter Horse-related breeds (Stockham et al., 1994,
and other nonanesthetic factors that may include exercise, stress, Wagner et al., 2006), junctional epidermolysis bullosa (JEB) in Bel-
gians (Johnson et al., 1988; Frame et al., 1988; Milenkovic et al.,
2003; Mömke & Distl, 2007) and in Saddlebred horses (Lieto et al.,
300 2002; Lieto & Cothran, 2003; Graves et al., 2009), and hereditary
equine regional dermal asthenia (HERDA) in Quarter Horse-
250 related breeds (White et al., 2004; Tryon et al., 2005; Graves et al.,
2009).
200 Jamison et al. (1987) described a congenital form of myotonia
Number of sires

with dystrophic changes in a Quarter Horse, which later became

150 known as hyperkalemic periodic paralysis (HYPP). HYPP in Quarter
Horses is an autosomal dominant disorder with a variable pene-
100 trance that has been extensively studied in recent years, not in the
least because the disease is homologous to adynamica episodica
50 hereditaria or Gamstorp’s disease in man (Pickar et al., 1991; Naylor
et al., 1992; Valberg et al., 1992; Rudolph et al., 1992a,b; Spier
0 et al., 1993; Meyer et al., 1999; Naylor et al., 1999). It has been
0 5 10 15 20 25 30 35
shown that the HYPP mutation in the Quarter Horse population
Uneven feet (%)
can be traced back to a single stallion (Bowling et al., 1996; Valberg
Fig 12.2  Mean prevalence of uneven feet in offspring related to number of et al., 1996). Indications exist that the inherited predisposition for
sires (n = 44 480 horses from n = 630 sires). the disease has a positive genetic correlation with muscular appear-
Reprinted from Ducro, B.J., Bovenhuis, H., Back, W., 2009. Heritability of foot conformation ance. This could mean that the HYPP rate has increased as a result
and its relationship to sports performance in a Dutch Warmblood population. Equine Vet. J. of selection for a more muscular appearance (Naylor, 1993, 1994a).
41, 139–143, with permission from the Equine Veterinary Journal. In addition, Quarter Horses affected with HYPP performed

167 -16 Fig 12.3  Least square means by year of height at

16.5 withers (red, in cm) of neck length (green, score) and
166.6
prevalence of uneven feet (bars, with percentage
166 -17 prevalence on top) (n = 44 480).
165.5 -17.5 Reprinted from Ducro, B.J., Bovenhuis, H., Back, W., 2009.
-18 Heritability of foot conformation and its relationship to sports
165 performance in a Dutch Warmblood population. Equine Vet. J. 41,
-18.5
Score

164.5 139–143, with permission from the Equine Veterinary Journal.

cm

-19
164
9.4 -19.5
163.5 9.0
7.3 7.5 8.1 -20
7.2
163 5.5 5.9
4.5 5.0 4.7 -20.5
3.8 3.7
162.5 -21
162 -21.5
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

253
 12 Genetic contributions to exercise and athletic performance

significantly better in halter classes than did unaffected horses (2009) fine mapped a quantitative trait locus on horse chromosome
(Naylor, 1994b). The mean total performance points were not sig- 2 associated with radiological signs of navicular disease in Hanove-
nificantly different. He stated that selection of Quarter Horses rian Warmblood horses. Németh (1974) demonstrated a pheno-
affected with HYPP was enhanced by show judges. It has been typic correlation between navicular disease and sesamoiditis. For
demonstrated that the disease manifests as a result of a defect in sesamoiditis a relatively low heritability of 0.11–0.17 has been esti-
the sodium channel subunit gene (Rudolph et al., 1992; Zhou mated (Barneveld, 1996). For osteoarthritis or degenerative joint
et al., 1994; Cannon et al., 1995). A mutation in glycogen synthase disease of the fetlock joint, which can also be graded using a radio-
1(GYS1) has been shown to be associated with equine PSSM and logical scale, heritability was found to be comparable (0.13–0.26;
exertional rhabdomyolysis (Valberg et al., 2001; Ward et al., 2003, Barneveld, 1996). Degenerative joint disease (DJD) of the distal
2004; McCue et al., 2008a,b, 2009a; Herszberg et al., 2009). The tarsal joints, often called bone spavin, is commonly associated with
GYS1 mutation is an important cause of exertional rhabdomyolysis poor conformation in which sickle hocks and cow hocks are pre-
but does not account for all forms of PSSM. However, a significant disposing factors (Stashak, 1987b). It is therefore not surprising that
proportion of horses with histopathological evidence of PSSM and/ the disease to a certain extent is thought to have a genetic basis.
or exertional rhabdomyolysis have different diseases (Valentine Barneveld (1983) studied 168 3- and 4-year-old Warmblood off-
et al., 1997; Valberg et al., 1998; Stanley et al., 2009). Tryon et al. spring of 11 stallions and concluded that bone spavin indeed had
(2009) estimated the allele frequencies of HYPP, lethal white foal a hereditary basis and was related to the conformation of the hind
syndrome (LWFS), glycogen branching enzyme deficiency (GBED), limb. In the study by Van der Veen et al. (1994) bone spavin was
HERDA, and PSSM genes in elite performance subgroups of Ameri- found to have a heritability of 0.20–0.35. Radiographic examina-
can Quarter Horses (AQHs). Accurate estimates of disease-causing tion is considered essential for diagnosis of bone spavin (Butler
alleles in AQHs and APHs may guide the use of diagnostic genetic et al., 1993) but palpation of the distal tarsus and hind limb motion
testing, aid management of genetic diseases, and help minimize evaluation before and after the flexion test of the tarsus have also
production of affected foals. been shown to be of importance. In German riding horses, the
heritability of bone spavin, based on radiographic diagnosis, has
been estimated to be low, 0.02–0.04 on an observed scale (Winter
Degenerative lameness-causing disorders et al., 1996). In Icelandic horses the heritability for bone spavin in
Frequently encountered lameness-causing disorders of the locomo- the distal tarsal joint was 0.1 (Björnsdóttir et al., 2000) whereas
tor system that are not developmental in origin, but are more of a heritability for hind limb lameness after the flexion test was 0.47.
degenerative nature, include navicular disease, sesamoiditis, osteo- The flexion test is regarded as a less specific method for the diagnosis
arthritis of the fetlock joint and bone spavin. of bone spavin. They suggest to select against bone spavin based on
Navicular disease is a degenerative disorder of the podotrochlea flexion test of the tarsus followed by radiographic diagnosis, since
including the navicular bone, the navicular bursa, the distal sesa- the combined trait has a considerable higher heritability than only
moid impar ligament, the collateral sesamoid ligament and the considering radiography. This is an interesting example of combin-
distal part of the deep digital flexor tendon, has been known as an ing two methods of diagnosis to develop a more accurate measure
important cause of lameness for ages (Youatt, 1836). An inherited of the phenotype, resulting in a higher heritability of the disorder.
basis for the disease has been suggested (Numans & van de Water-
ing, 1973).
Using a radiological classification of the navicular bone as a
Developmental lameness-causing disorders
measure for the disease, Bos et al. (1986) found variation between Hermans et al. (1987) investigated the genetic background of con-
daughter groups consisting of 3-year-old mares from different sires, genital luxation of the patella in Shetland ponies. Hermans (1970)
supporting the theory that navicular disease has, to a certain extent, described ulnar and tibial malformation (persistence) in the Shet-
a genetic basis. However, they also concluded that the results of the land pony with associated locomotor problems. Vertebral compres-
radiological evaluation of the sires did not have predictive value for sive myelopathy or ‘wobbler disease’ is a pathological disorder of
the progeny. the spinal cord rather than of the locomotor system. However, as it
In a large study that included 590 female offspring from 30 sires, produces ataxia the condition may severely affect locomotor perfor-
the heritability of the radiographic classification of the navicular mance. An early report on Wobbler disease stated that the syndrome
bone was estimated as 0.26–0.34 (Van der Veen et al., 1994) and was familial (Dimock, 1950). However, this could not be confirmed
were in agreement with heritability estimates ranging from in a later large-scale retrospective study (Falco et al., 1976), nor in
0.20 to 0.31 found by Willms et al. (1999) on 492 Holstein horses. a prospective study in which clinically and radiographically con-
Winter et al. (1996) analyzed 3566 German Warmblood horses firmed ‘wobbler’ mares and stallions were mated (Wagner et al.,
selected for auction sale without clinically manifest navicular 1985). In the latter study the high incidence of a number of devel-
disease. The heritability estimate was 0.06 on the observed scale and opmental orthopedic diseases (OC, physitis, contracted tendons)
no transformation to liability scale was performed. The study of was remarkable.
Stock and Distl (2006) gave estimates varying from 0.10 to 0.34.
From these studies it can be concluded that radiological changes of
the navicular bone are genetically influenced. An explanation might
Complex chondrodysplasia: osteochondrosis
be the shape of the proximal articular border of the navicular bone Osteochondrosis (OC) can be defined as a disturbance of the
has been found to be inherited and to predispose the pathogenesis process of endochondral ossification as this occurs in the growing
of navicular disease; the more concave or undulating, the higher the individual. Irregular ossification leads to the formation of thick
risk for the disease (Dik et al., 2001). Selection against navicular cartilage plugs and areas of focal necrosis, eventually resulting in
disease was successfully performed by the Dutch Warmblood stud- flattened bone contours and loose fragments, commonly referred
book. In 1997 selection was started by excluding from breeding the to as osteochondrosis dissecans (OCD) (Jeffcott, 1997; van de Lest
stallions with severest grade (grade 4) of the disease has resulted in et al., 1999). Of the group of so-called developmental orthopedic
a considerable decrease of the two worst grades (3 and 4). From diseases it is by far the most common (McIlwraith, 1986). Epide-
1997 to 2002 the prevalence of the two most severe grades (3 and miological data suggest that the disorder is present in many breeds
4) decreased from 11 to 2% (Van den Belt et al., 2003). Diesterbeck of horses in 10–25% of the population (Jeffcott, 1997). The exact
and Distl (2007) reviewed the genetic aspects of radiological altera- pathogenesis of the disease is still unclear and subject to investiga-
tions in the navicular bone of the horse and found Genome-wide tion (Jeffcott, 1991; Jeffcott & Henson, 1998). However, there is
significant QTL were on ECA2 and on ECA10, whereas Lopes et al. universal agreement that the disease is multifactorial (Hurtig &

254
 Selection for locomotor soundness

Pool, 1996) and that genetics play a role. Therefore, like many often are converted afterwards to the underlying scale, using appro-
diseases and performance characteristics, the phenotypic expression priate formulas (e.g. Dempster & Lerner, 1950).
of OC is influenced by the environment plus a genetic component The analysis is commonly performed using a sire or an animal
consisting of many genes in the genome (Gerber & Bailey, 1995). model. The animal model better accounts for the state of selection
Van Weeren and Barneveld (1999) showed that osteochondrosis in females, whereas the assumption underlying a sire model is that
is a very dynamic process in which lesions develop, but may also females are unselected. In horse breeding in which assortative
regress spontaneously. They also demonstrated the presence of mating is common. As a result of poor data quality and treatment,
osteochondrotic lesions in many joints other than the commonly genetic differences between joints and breeds might be
affected hock and stifle. These findings shed a different light on this compromised.
condition and ask the question of whether there is a genetic differ- OC/OCD in the stifle seems to have low heritability with esti-
ence between animals that show the condition at 3 years of age and mates around 0.08 and not significantly different from zero (Tables
animals that have shown the condition since foals, but in which the 12.6 and 12.7). Measuring OC in the stifle is probably more difficult
lesions subsequently regressed spontaneously. In addition to this, than in the hock, which introduces random noise. The heritability
when considering the stifle joint, they found significant difference estimate for OC in the stifle might therefore be lower than for the
in the prevalence of the condition in foals that were the offspring hock, although the prevalence is comparable. This is supported by
of OC-free parents compared with foals whose parents were suffer- the significant effect of sedation during radiography (Van Grevenhof
ing from the condition in the same joint. This was not the case with et al., 2009a). OC/OCD in the FP more often leads to lameness
respect to oseteochondrotic lesions in the hock joint and might be than OC/OCD in the TC (Auer & Stick, 2006), but OC/OCD in the
an indication of a different genetic background in these two joints hock joint is more frequently investigated. When only considering
(Van Weeren et al., 1999). OC, heritability lies mostly in the area of 0.15, whereas OCD is
Osteochondrotic lesions are found in different joints and joints somewhat higher at 0.27 (Stock & Distl, 2006; Van Grevenhof et al.,
primarily affected are the tarsocrural (TC), femeropatellar (FP), gle- 2009b). OC in trotters might be somewhat higher as indicated by
nohumeral (GH), metacarpophalangeal, (MCP), metatarsophalan- studies of Grondahl and Dolvik (1993) and Philipsson et al.
geal (MTP) and the cervical intervertebral (CI) joints (Radiostits (1993). Grondahl and Dolvik suspected an overestimation in their
et al., 2007). Most of the genetic studies are considering only one result, because of a very high incidence of OC in one group of
or a few of the joints, of which the stifle, hock and fetlocks are the offspring. OC/OCD in the fetlock has a heritability of 0.10 (Schober
most frequent ones (Table 12.6). The large range in prevalence and et al., 2003; Van Grevenhof et al., 2009b); whereas heritability of
heritability reported may be attributed to: use of field data, defini- OCD was somewhat higher at 0.17–0.18 (Stock et al., 2005; Phil-
tion of OC, and statistical model of analysis in relation to the scale ipsson et al., 1993), which was accompanied by a higher prevalence
of OC-scoring. As radiographic examinations are expensive and of 25%. Van Grevenhof et al. (2009b) found only a value of 0.09
labor intensive, datasets collected at other occasions (auctions and when clearly separating fragments from flattenings. They argued
presale purchases) are often subject to genetic studies. There is a that only osseous fragments seen at the proximodorsal parts of the
serious risk that these datasets are too small or preselected, leading sagittal ridge of the third metacarpal and metatarsal bone can be
to inaccurate and/or biased estimates. The heritability estimates considered as caused by the same etiology as other OC and not by
may also vary as a result of different definitions of OC, in particular mechanical injuries (Hurtig & Pool, 1996).
with respect to bony irregularities at different predilection sites in Genetic correlations reflect to what extent traits are affected by
the joints. Depending on the predilection site, lesions are assumed the same genetic complex and therefore contribute to unraveling
to have either a traumatic origin or an osteochondrotic origin (Jef- the etiology of traits. Sandgren et al. (1993) found genetic correla-
fcott, 1991). For example, Philipsson et al. (1993) and Grøndahl tions between hock and fetlock close to zero and this corresponds
and Dolvik (1993) only regarded bony fragments at the intermedi- to the findings of Grondahl and Dolvik (1993) and Van Grevenhof
ate ridge of the distal tibia and at lateral/medial trochlear ridge of et al. (2009b). Stock et al (2005) estimated a moderate positive
the talus as OC, whereas McIlwraith (1993a,b) additionally regarded genetic correlation between hock and fetlock OCD. Genetic correla-
the lateral/medial malleolus of the tibia as a predilection site. Preva- tion of stifle with hock was moderately positive, and of stifle with
lence and heritability are likely affected by the number of predilec- fetlock was low. Although standard errors in most of the studies
tion sites considered. Similarly, fragments and flattening might have were too high to make significant statements, the genetic correla-
a different genetic basis and it is questionable whether these two tions between joints did not indicate that OC as manifested in the
forms should be treated alike in a genetic study. Scoring of OC is various joints are the same traits (or solely regulated by the same
mostly on binary scales (yes/no OC), which does not give the genes). This is supporting the hypothesis of Van Weeren and Van
opportunity to record further details with respect to form and Barneveld (1999) of a different genetic background in the different
degree of severity of OC. Instead Van Grevenhof et al. (2009b) used joints (Van Weeren et al., 1999). Additionally, genetic correlations
a 5-point scale in their study on which form as well as severity of between the joints are not high enough to rely on measuring in
disorder could be recorded. In most joints heritability-estimate only a part of the joints to set up a selection program to eradicate
improved when using the 5-point scale, compared to a binary scale OC in all joints. Whether the two forms of OC are genetically
(Table 12.7). In contrast, Ricard et al. (2002) propagated a binary related was investigated by van Grevenhof et al. (2009b). Genetic
scale, because heritabilities based on multipoint scales were lower, correlation between flattenings and fragments over joints was high
due to introduction of random noise in the data when using a (0.80) and not significantly differing from one. This finding sup-
multipoint scale. This illustrates that the extra categories should be ports the hypothesis of a common genetic background of both
useful differentiators. forms of OC.
After data collection, further differences in heritability estimates Genetic correlations of OC with conformation and performance
could arise from different models of analysis. The binary trait could could aid to understanding the etiology as well, and reveals how
be treated accordingly using a threshold model, although often an selection against OC can be implemented in a horse breeding
ordinary linear model is used. A linear model reveals the heritability program. There are only a few studies on the relation with perfor-
on the observed scale, which is an underestimation of the true mance and expectations are based on phenotypic observations.
frequency because of its dependence on the prevalence in the popu- Affected horses suffering from pain and distress are likely to show
lation. A threshold model estimates heritability on the underlying reduced performance in sports. However, given the high preva-
scale, which can be interpreted as the sensitivity to or risk of devel- lence of radiographic findings in horses without signs of lameness
oping the disorder (irrespective of having OC), and better reflects or stiff joints, the relevance of particular radiographic findings,
the genetic potency. Therefore, estimates on the observable scale especially with respect to sport performance, is being questioned.

255
 12 Genetic contributions to exercise and athletic performance

Table 12.6  Prevalences and heritabilities (h2) of osteochondrosis (OC) and fragments (OCD), by joint, for different populations of
warmblood (WB) horses

Population n Radiographic Prevalence h2 SE Method of Reference*

finding author
Femoropatellar (FP) OC/OCD
Dutch WB stallions (n = 1965) OC 11.5 0.09 ATM (REML, DL) I
French WB 103 sires (n = 733) OC 1–7 0.00–0.17 LSM II
Italian WB 75 sires (n = 350) OCD 16.6 0.09 0.24
ATM (REML) III

Tarsocrural (TC) OC/OCD

Dutch WB stallions (n = 1965) OC 16.0 0.11 ATM (REML, DL) I
Dutch WB mares 30 sires (n = 590) OC 13.7 0.01 0.06
LSM (REML) IV
Dutch WB mares 30 sires (n = 590) OC 13.7 0.14 0.17
LAM (REML) IV
French WB 103 sires (n = 733) OC 11–13 0.00–0.02 LSM II
Hanoverian WB 165 sires (n = 624) OC 10.5 0.06 0.06
LAM (REML) V
SB trotters 39 sires (n = 644) OC 14.3 0.52 STM (REML) VI
SB trotters 24 sires (n = 793) OC 10.5 0.27 0.08
LSM VII
0.06
Hanoverian WB 3725 OCD 9.6 0.37 LAM (REML, DL) VIII
Hanoverian WB 569 sires (n = 5231) OCD 9.2 0.28 0.04
LAM (REML, DL) IX
Hanoverian WB 569 sires (n = 5231) OCD 9.2 0.27 0.04
LSM (REML, DL) IX
Hanoverian WB 569 sires (n = 5231) OCD 9.2 0.17 0.07
STM (GS) IX
Danish Trotters 9 sires (n = 325) OCD 12.0 0.260.14 STM X

Metacarpophalangeal/metatarsophalangeal (MCP/MTP) OC/OCD

French WB 103 sires (n = 733) OC 8–11 0.04–0.21 LSM II
Hanoverian WB 165 sires (n = 624) OC 18.3 0.12 0.09
LAM (REML) V
Hanoverian WB (n = 3725) OCD 20.8 0.19 0.03
LAM (REML, DL) VIII
Hanoverian WB 569 sires (n = 5231) OCD 23.5 0.17 0.03
LAM (REML, DL) IX
Hanoverian WB 569 sires (n = 5231) OCD 23.5 0.170.03 LSM (REML, DL) IX
Hanoverian WB 569 sires (n = 5231) OCD 23.5 0.120.05 STM (GS) IX
SB trotters 39 sires (n = 644) OCD 11.8 0.21 STM (REML) VI
SB trotters 24 sires (n = 793) OCD 21.5 0.170.06 LSM VII

All joints (FP+TC+MCP/MTP)

Italian WB 75 sires (n = 350) OCD 16.6 0.140.23 LAM (REML, DL) III

ATM, animal threshold model; DL, Dempster-Lerner transformation; GS, Gibbs sampling; LAM, linear animal model; LSM, linear sire model; REML, restricted maximum likelihood;
SB, Standardbred; STM, sire threshold model; WB, Warmblood.
*I, der Kinderen (2005); II, Ricard et al. (2002); III, Pieramati et al. (2003); IV, KWPN (1994); V, Schober et al. (2003); VI, Grøndahl and Dolvik (1993); VII, Philipsson et al. (1993); VIII,
Stock et al. (2005); IX, Stock and Distl (2006); X, Schougaard et al. (1990).
Adapted from Van Grevenhof, E. M., B. J. Ducro, P. R. van Weeren, J. M. F. M., van Tartwijk, A. J. van der Belt, and P. Bijma, 2009. Prevalence of various radiographic manifesta-
tions of osteochondrosis and their correlations between and within joints in Dutch Warmblood horses (KWPN). Equine Vet. J. 41, 11–16, with permission from the Equine
Veterinary Journal.

Reduced sports performance was found in trotters (Grøndahl & make inferences on whether sires that inherit potential for a suc-
Engeland, 1995) and osteoarticular findings were directly respon- cessful sport career also inherit high or low susceptibility to OC.
sible for failure to qualify in 31% of the horses participating In the study of Stock and Distl (2006) hock-OCD showed high
(Robert et al., 2006). In contrast, no significant differences were negative genetic correlation to sports performance in both disci-
found in racing performance of Standardbred trotters with radio- plines whereas fetlock-OCD was only negatively related to show
graphic signs of OCD, relative to their contemporaries (Brehm & jumping. Negative genetic correlations are favorable because they
Straecker, 2000; Storgaard Jørgensen et al., 1997). These results are point out that breeding for lower prevalence of OC coincides with
indicative for phenotypic correlations, but do not allow us to breeding for higher sport performance. The authors suspected

256
 Selection for locomotor soundness

circumference of the cannon bone and the carpus (Pagan &

Table 12.7  Heritabilities estimated (h2) using continuous, binary
Jackson, 1996; Van Weeren et al., 1999). In contrast to hock OC,
and ‘1 to 5’ scales
foals with lower body weight were found to have a significantly
higher prevalence of OC in fetlock joints (Schober et al., 2003).
Level2 OC3 σp h2 h2 h2 In conclusion, the genetic relation of OC to growth is not
continuous binary 1-to−5 straightforward and depends on the joint considered. Develop-
scale1 scale1 scale1 ment of OC is probably related to how the growth-phase of the
body and its components interfere with the (disproportionate)
Animal ALL 4.05 0.23 (0.09) 0.15 (0.08) 0.23 (0.09) weight bearing of each individual joint, which itself is in a
FLAT 3.16 0.08 (0.06)
process of growing.
Until now selection against OC has not been very successful,
FRAG 2.45 0.22 (0.09) despite the heritable character of OC. Most of the selection pro-
grams have been based on the phenotype of the candidate-breeding
FP ALL 2.09 0.05 (0.05) 0.05 (0.05) 0.06 (0.05) stallion. Observable heritability is low which has the consequence
FLAT 1.17 0.07 (0.06) that stallions’ own phenotype is not a reliable predictor of its geno-
type. A more reliable breeding value can be realized from progeny
FRAG 1.22 0.02 (0.04) testing. The limited use of the stallions’ phenotype to predict its
TC ALL 2.01 0.36 (0.11) 0.25 (0.10) 0.36 (0.11) breeding values is illustrated by the studies of Schougaard et al.
(1990) and Van Grevenhof et al. (2009a), where the recorded
FLAT 1.47 0.15 (0.08) animals all are descending from breeding stallions which have been
checked for OC. Still there was large variation in OC prevalence
FRAG 1.39 0.26 (0.09)
among the progeny groups.
MCP/MTP ALL 2.24 0.14 (0.08) 0.10 (0.07) 0.11 (0.07) To date diagnosis of OC has always been on radiographic findings
or clinical state and more advanced techniques like magnetic reso-
FLAT 1.86 0.08 (0.10)
nance imaging or biomarkers will play a more prominent role in
FRAG 1.36 0.06 (0.07) the future (Vervuert et al., 2007). In particular, biomarkers, sup-
ported by mRNA expressions factors (Gläser et al., 2009; Austbø
1
Heritability values are expressed as h2 (SE). The continuous scale refers to et al., 2010; Serteyn et al., 2010), may serve as an early and more
transforming osteochondrosis (OC) scores on a continuous liability scale; the objective measure.
bivariate scale refers to OC scored as 0 or 1; and the scale of 1 to 5 refers to Austbø et al. (2010) used a PCR (RAP-PCR) and a real-time
transforming A to E scores into corresponding values ranging from 1 to 5.
RT-PCR and tissue samples of the distal intermediate ridge of the
2
FP = femoropatellar joint; TC = tarsocrural joint; MCP/MTP = metacarpophalangeal tibia of foals considered predisposed to OC when the parents had
and metatarsophalangeal joints. OC, and identified two genes not previously correlated with OC as
3
Refers to the joints level. ALL = both flattened bone contours and fragments; FLAT differentially expressed, and which were identical to TLK2 and an
= flattened bone contours; FRAG = fragments. equine EST. Serteyn et al. (2010) compared leucocytes of OC-affected
Taken from Van Grevenhof, E.M., Ducro, B.J., van Weeren, P.R., van Tartwijk, J.M.F.M., and non-OC-affected horses using digital gene expression analysis
van der Belt, A.J., Bijma, P., 2009a. Prevalence of various radiographic manifestations (DGE) and real-time PCR. They showed an obvious dysregulation
of osteochondrosis and their correlations between and within joints in Dutch of several signaling pathways related to cartilage formation or car-
Warmblood horses (KWPN). Equine Vet. J. 41, 11–16. tilage repair, including Wnt, Indian hedgehog, and TGF-β signaling.
Other genes, including ISG, APOB, MGAT4, and TBC1D9, showed
a significantly different expression between groups. They proposed
that these genes might play a role in high carbohydrate diet, abnor-
some overestimation of the correlations, but nevertheless a nega- mal insulin metabolism, or inflammation, mechanisms suspected
tive correlation was expected since hock joint lesions are most to be involved in OC.
likely to interfere with load bearing of the hind limbs in dressage Also, these methods of diagnosis might help to break through
sport. The authors therefore concluded that orthopedic status of the binary measure of the trait currently used and might reveal a
the horse could genetically be improved, without compromising more quantitative measure of bone constitution, making it possible
selection for sports performance. However, Winter et al. (1996) to allow horses that are not OC-affected to be distinguished from
estimated a positive genetic correlation between show jumping the never-affected. For the detection of animals at risk genetic
and occurrence of OCD. markers have the potential to form a basis for reducing OC positive
As intensive sport performance will affect joint biomechanics, animals in a population (Dierks et al., 2007; Lampe et al., 2009a,b;
so does conformation. Deviating conformation might lead to dis- Wittwer et al., 2007a,b, 2008, 2009). Lykkjen et al. (2010) were the
proportionate weight bearing and subsequently OC. Many confor- first to perform a genome-wide association study for osteochondro-
mational traits are considerably heritable and a genetic sis of the tibiotarsal joint in Norwegian Standardbred trotters using
relationship to OC might be expected. Taller horses have a predis- the recently released Illumina Equine SNP50 BeadChip. On chro-
position for osseous fragments in the fetlock and hock joints mosomes (Equus callabus chromosome (ECA)) 5, 10, 27 and 28 they
(Stock et al., 2005), which agree with findings in German and found moderate evidence of association representing interesting
Dutch Warmbloods and Swedish trotters (Van der Veen et al., areas for future research, validation studies and fine mapping of
1994; Pagan & Jackson, 1996; Sandgren et al., 1993; Van Weeren candidate regions. Dierks et al. (2010) refined the quantitative trait
et al., 1999; Winter et al., 1996). This could indicate positive locus for equine osteochondrosis (OC) on horse chromosome
genetic correlations, but the literature shows contradictory results. (ECA) 2 to a genome-wide significant interval at 20.08–30.94 Mb.
Negative genetic correlations were found between height at the
withers and the incidence of hock OC (Philipsson et al., 1993;
Willms et al., 1999). Large carpal circumference was genetically
associated with a high incidence of hock OC but a low incidence
Simple chondrodysplasia: dwarfism
of fetlock OCD (Philipson et al., 1993). Foals affected with hock Within the Friesian horse breed, congenital dwarfism has been
OC had a higher average daily weight gain (Sandgren et al., recognized for many years and occurs at a frequency of 0.25%
1993), they had a larger frame and a markedly larger (Osinga, 2000; Back et al., 2008). A full phenotypic

257
 12 Genetic contributions to exercise and athletic performance

4
-log10 P-Value

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 3031
Chromosome
Fig 12.4  Manhattan plot of P-value in the Friesian horse dwarfism GWAS. Association of 34 429 SNPs with dwarfism represented by –log10 P-values from a
two degree of freedom x2 test plotted by chromosome and sorted by chromosomal position. No SNP in the GWAS remained statistically significant after
correction for multiple testing.
Reprinted from Orr, N., Back, W., Gu, P., et al., 2010. Genome-wide SNP association based localization of a dwarfism gene in Fresian dwarf horses. Animal Genetrics 41 (Suppl. 2), 2–7, with
permission from John Wiley and Sons.

characterization for Friesian dwarfs has been reported (Back et al., 5

2008). Briefly, the Friesian dwarf phenotype results from physeal
growth retardation in both limbs and ribs, reflected in a character- BIEC2-239376
istic disproportional growth disturbance. The potential for post- P=4.54x10-5
natal growth in these animals, albeit at a reduced rate, is responsible 4
for mature dwarfs having a head of the same size as unaffected
animals, a broader chest with narrowing at the costochondral junc-
tion, a disproportionally long back and abnormally short limbs. 3
-log10 P-Value

Furthermore, radiographs reveal a dysplastic metaphysis of the

distal metacarpus and metatarsus. Light microscopy of growth
plates at the costochondral junction demonstrate an irregular tran-
sition from cartilage to bone, and thickening and disturbed forma- 2
tion of chondrocyte columns, which is similar to findings in
osteochondrodysplasia (Back et al., 2008). The molecular mecha-
nism underlying the growth disturbance in Friesian foals has yet to
be determined, although the trait is heritable and appears to follow 1
an autosomal recessive pattern of transmission (Osinga, 2000; Back
et al., 2008). Orr et al. (2010) have reported the first genetic
mapping of the genetic determinants of dwarfism in the Friesian 0
horse using GWAS methodologies in a small number of Friesian
dwarfs and normal controls. Ten Friesian dwarf horses were selected 16200 47300 78400
by the investigators over an 8 year period (2001–2008) based on Chromosome 14 position (kb)
their unique phenotype (Back et al., 2008) and availability of DNA.
Ten normal Friesian horses (controls) were selected in 2008 based Fig 12.5  Association plot of chromosome 14 SNPs with dwarfism, with
chromosomal position on the x-axis and –log10 P-value on the y-axis. The
on a normal phenotype.
most significant chromosome 14 SNP, BIEC2-239376, is indicated by a red
A peak of association was observed on chromosome 14, the best
diamond. Each square on the plot represents a single SNP; the color of each
SNP being BIEC2-239376 (p = 4.54 × 10−5) (Fig. 12.4) and the square represents the strength of correlation between that SNP and
significance of association at BIEC2-239376 was greatest under a BIEC2-239376, with dark red indicating an r2 of 1.
recessive model (p = 7.74 × 10−6). All 10 of the dwarfs in the GWAS Reproduced from Orr et al. (2010).
were TT homozygotes, while the controls were comprised of 4 CC
homozygotes and 6 heterozygotes. Fine-mapping in the region was
performed by genotyping 374 SNPs, of which 319 were not on the singletons; none was associated with the dwarfism phenotype. Thus,
GWAS array, in the region of association on chromosome 14. The resequencing analysis of the protein coding components of PROP1
BIEC2-250663 (p = 4.94 × 10−5) SNP was more significantly associ- failed to detect a putative functional variant.
ated with dwarfism than BIEC2-239376 (p = 6.12 × 10−5) (Fig. 12.5); Perturbations in homologous gene families often result in the
dwarfs were fully homozygous for both markers and the homozy- manifestation of phenotypically similar traits between species. As
gous genotype observed in dwarfs was not present in controls. An such, Orr and colleagues screened the Online Mendelian Inheri-
additional SNP, BIEC2-249929 was in perfect linkage disequilib- tance in Man (OMIM) database (http://www.ncbi.nlm.nih.gov/
rium with BIEC2-239376. BIEC2-239376 is located 34 Kb from the omim/) to determine whether any of the genes in the chromosome
gene encoding the equine homolog of the prophet of PIT1, paired- 14-association region were implicated in dwarfism in humans. The
like homeodomain transcription factor protein, PROP1. Inactivat- equine homolog of PROP1 was identified as a strong positional
ing mutations in PROP1 are known to cause dwarfism in humans candidate gene for dwarfism in Friesians. Inactivating mutations in
via combined pituitary hormone deficiency. Therefore re-sequencing PROP1 cause disruption of the growth hormone (GH) axis and are
in the PROP1 gene was performed resulting in the detection of nine manifested as a dwarfism phenotype. Simple, central pituitary
SNPs, all of which were novel based on their absence from the underproduction of GH leads typically to proportional dwarfism,
EquCab 2.0 SNP database. Each novel SNP was rare, and many were as seen in the German Shepherd dog (Andresen & Willeberg, 1976;

258
 References

Hanson et al., 2006). In the Friesian dwarf horses, however, a dis- Finally, both FGFR1 and FGFR2 are included within the critical
proportional growth disturbance is seen, which would imply a region of chromosome 14 flanked by BIEC2-249929 and BIEC2-
local defect or disturbance in one of the regulatory systems for 250663 based on the equine build from UCSC Broad Institute
growth plate development. Growth plate development is under the (UCSC Genome Browser: http://genome.ucsc.edu). Both fibroblast
control of many autocrine and paracrine factors (Kronenberg, growth factor receptors play key roles in skeletal development and
2003). Recently, de Graaf and colleagues investigated the function- mutations have been related to skeletal dysplasia and dwarfing syn-
ing of the hypothalamic-pituitary growth axis in three Friesian dromes (White et al., 2005; Eswarakumar et al., 2002). Profound
dwarfs. No evidence of hypothalamic-pituitary dysfunction or effects on bone elongation have been shown through supposed sup-
failure of IGF-1 production was found, suggesting that the cause of pression of chondrocyte and osteoblast function. However, in con-
the congenital growth abnormality was located distal or peripheral trast to the relative normal head proportions seen in the Friesian
to the level of the GH receptor in the liver and may have been a dwarf syndrome, FGFR1 and FGFR2 mutations seem to have a sig-
defect in a peripheral IGF-1 or GH receptor, or may not involve the nificant effect on flat bone growth and skull formation.
GH-IGF-1 axis at all (De Graaf-Roelfsema et al., 2009). Concomi- In conclusion, a putative region that may harbor a gene for dwarf-
tant with this observation, no coding polymorphisms in PROP1 ism in Friesian horses has been identified on ECA14. Validation of
were detected that were associated with dwarfism in the Friesian this finding in a larger group of animals and segregation analysis in
horse samples. When the normal bone remodeling process is dis- known pedigrees is warranted before further localization of the
turbed in horses, abnormal defects in the growth plate may result. causative mutation is conducted. This may prove extremely chal-
A local defect or disturbance in one of the regulatory systems for lenging given the strong linkage disequilibrium in the region. The
growth plate development potentially can result in a dispropor- study suggests that with the advent of new genomic tools, studies
tional growth disturbance typically seen in Friesian dwarf horses of equine diseases may yield important new insight into pathogen-
(Vaughan, 1976; Jeffcott & Henson, 1998; Gee et al., 2005). esis and may be translatable to orthologous human traits.
Further screening of the OMIM database for genes in the chromo-
some 14 region that could be linked to a disturbed bone remodel-
ing process in abnormal growth plate development identified
ZNF346, COL23A1 and B4GALT7 as candidate genes. ZNF346 is Conclusions
proposed to play a role in apoptosis (cell death), a process crucial
for the normal transition of cartilage into bone seen during normal The recent completion of the horse genome sequence and the com-
physeal growth (Gibson, 1998; Ballock & O’Keefe, 2003). It could mercial availability of an equine SNP genotyping array will facilitate
be speculated that disturbed apoptosis plays a role in the physeal an acceleration in the mapping of disease genes in the horse during
growth retardation typically seen in Friesian dwarfism. Both the coming years. While environmental factors such as management
COL23A1 and B4GALT7 are proposed to play a role in collagen and training play an important role in shaping the equine athletic
network formation. Although a disturbed collagen network is phenotype, it is likely that genetic testing for conformation, disease
known to effect calcification and subsequent transformation of car- and performance traits will become mainstream and that this infor-
tilage into bone (Wassen et al., 2000), the specific role of both mation will enhance decision-making processes in the selection and
genes in collagen formation is highly speculative and largely breeding of horses. A crucial step in the success of developing gene
unclear. B4GALT7 is also proposed to play a role in connective tests is the phenotyping. Poor phenotyping will severely compro-
tissue disorders and has been related to disturbed fibril organiza- mise the success of genotyping. For a set of diseases this is less
tion and proteoglycan synthesis. Both processes could play a role relevant as the phenotype is clear, but for complex multi-factorial
in abnormal development of bone and subsequent retardation of diseases the phenotype is less clear. Defining accurate phenotypes
growth in the growth plate seen in Friesian dwarfism (Kvist et al., for performance and conformation, in which subjective measure-
2006; Burdan et al., 2009). ments are involved will be the challenge in the near future.

References
Aleman, M., Nieto, J.E., Magdesian, K.G., Back, W., van der Lugt, J.J., Nikkels, P.G., van rearranged by inversion with a breakpoint
2009. Malignant hyperthermia associated den Belt, A.J., van der Kolk, J.H., Stout, in the distal class I region.
with ryanodine receptor 1 (C7360G) T.A.E., 2008. Phenotypic diagnosis of Immunogenetics. 56,702–709.
mutation in Quarter Horses. J. Vet. Intern. dwarfism in six Friesian horses. Equine Vet. Beech, J., Haskins, M., 1987. Genetic studies
Med. 23, 329–334. J. 40, 282–287. of neuraxonal dystrophy in the Morgan.
Andersson, L.S., Larhammar, M., Memic, F., Ballock, R. T., O’Keefe, R. J., 2003. The Biology Am. J. Vet. Res. 48, 109–113.
Wootz, H., Schwochow, D., Rubin, C.J., of the growth plate. J. Bone Joint Sur. Am. Bernoco, D., Bailey, E., 1998. Frequency of the
et al, 2012. Mutations in DMRT3 affect 85, 715–726. SCID gene among Arabian horses in the
locomotion in horses and spinal circuit Barneveld, A., 1983. Spat bij het paard. PhD USA. Anim. Genet. 29, 41–42.
function in mice. Nature 488, 642–646. thesis, Utrecht University. Binns, M.M., Boehler, D.A., Lambert, D.H.,
Andresen, E., Willeberg, P., 1976. Pituitary Barneveld, A., 1996. The role of breeding in 2010. Identification of the myostatin locus
dwarfism in German shepherd dogs sports performance and health with special (MSTN) as having a major effect on
– additional evidence of simple, autosomal emphasis on lameness prevention. optimum racing distance in the
recessive inheritance. Nordisk Veterinaer Pferdeheilk. 12, 689–692. Thoroughbred horse in the USA. Anim
Medicin 28, 481–486. Barrey, E., Mucher, E., Robert, C., Amiot, F., Genet. 41 (Suppl. 2), 154–158.
Auer, J.A., Stick, J.A., 2006. Equine Surgery, Gidrol, X., 2006. Gene expression profiling Björck, G., Everz, K.E., Hansen, H.-J.,
third ed. Elsevier Publishers, London. in blood cells of endurance horses Henricson, B., 1973. Congenital cerebellar
Austbø, L., Røed, K.H., Dolvik, N.I., Skretting, completing competition or disqualified due ataxia in the Gotland pony breed. Zbl. Vet.
G., 2010. Identification of differentially to metabolic disorder. Equine Vet. J., Suppl. Med. A. 20, 341–354.
expressed genes associated with 36, 43–49 Björnsdóttir, S., Axelsson, M., Eksell, P.,
osteochondrosis in Standardbred horses Beck, T.W., Menninger, J., Murphy, W.J., Nash, Sigurdsson, H., Carlsten, J., 2000.
using RNA arbitrarily primed PCR. Anim. W.G., O’Brien, S.J., Yuhki, N., 2005. The Radiographic and clinical survey of
Biotechnol. 21, 135–359. feline major histocompatibility complex is degenerative joint disease in the distal

259
 12 Genetic contributions to exercise and athletic performance

tarsal joints in Icelandic horses. Equine Vet. mammals. Mamm. Genome. 16 (8), Dierks, C., Mömke, S., Drögemüller, C., Leeb,
J. 32 (3), 268–272. 631–649. T., Chowdhary, B.P., Distl, O., 2006. A
Blendinger, C., Müller, G., Bostedt, H., 1994. Buoen, L.C., Zhang, T.Q., Weber, A.F., Turner, high-resolution comparative radiation
The “lethal white foal” syndrome]. T., Bellamy, J., Ruth, G.R., 1997. hybrid map of equine chromosome
Tierarztl. Prax. 22, 252–255. Arthrogryposis in the foal and its possible 4q12-q22. Anim. Genet. 37 (5), 513–517.
Boneker, C, Kuiper, H, Drögemüller, C, relation to autosomal trisomy. Equine Vet. Diesterbeck, U., Distl, O., 2007. Review of
Chowdhary, B.P., Distl, O., 2006. Molecular J. 29, 60–62. genetic aspects of radiological alterations in
characterization of the equine collagen, Burdan, F., Szumiło, J., Korobowicz, A., the navicular bone of the horse. Dtsch.
type IX, alpha 2 (COL9A2) gene on horse Farooquee, R., Patel, S., Patel, A., et al., Tierarztl. Wochenschr. 114, 404–411.
chromosome 2p16–>p15. Cytogenet. 2009. Morphology and physiology of the Dik, K.J., van den Belt, A.J., van den Broek, J.,
Genome Res. 115 (2), 107–114. epiphyseal growth plate. Folia Histochem. 2001 Relationships of age and shape of the
Böneker, C., Kuiper, H., Wöhlke, A., Cytobiol. 47 (1), 5–16. navicular bone to the development of
Drögemüller, C., Chowdhary, B.P., Distl, O., Butler, D., 1993. WHO plans guidelines for navicular disease: a radiological study.
2005a. Assignment of the COL16A1 gene genetics research. Nature. 366 (6455), 500. Equine Vet. J. 33 (2), 172–175.
to equine chromosome 2p15.1-p15.3 by Caetano, A.R., Shiue, Y.L., Lyons, L.A., O’Brien, Dimock, W.W., 1950 “Wobbles” an hereditary
FISH and confirmation by RH mapping. S.J., Laughlin, T.F., Bowling, A.T., Murray, disease in horses. J Hered. 41 (12),
Anim. Genet. 36 (3), 262–263. J.D., 1999. A comparative gene map of the 319–323.
Böneker, C., Müller, D., Kuiper, H., horse (Equus caballus). Genome Res. 9 Dranchak, P.K., Valberg, S.J., Onan, G.W.,
Drögemüller, C., Chowdhary, B.P., Distl, O., (12), 1239–1249. Gallant, E.M., Binns, M.M., Swinburne, J.E.,
2005b. Assignment of the COL8A2 gene to Cannon, S.C., Hayward, L.J., Beech, J., Brown et al., 2006. Exclusion of linkage of the
equine chromosome 2p15-p16 by FISH Jr, R.H., 1995. Sodium channel inactivation RYR1, CACNA1S, and ATP2A1 genes to
and confirmation by RH mapping. Anim. is impaired in equine hyperkalemic recurrent exertional rhabdomyolysis in
Genet. 36 (5), 444–445. periodic paralysis. J. Neurophysiol. 73, Thoroughbreds. Am. J. Vet. Res. 67,
Bonen, A., 2009. PGC-1alpha-induced 1892–1899. 1395–1400.
improvements in skeletal muscle Cavendysh, G., 1674. Méthode nouvelle et Ducro, B.J., Bovenhuis, H., Back, W., 2009a.
metabolism and insulin sensitivity. Appl. invention extraordinaire de dresser les Heritability of foot conformation and its
Physiol Nutr Metab. 34 (3), 307–314. chevaux et les travailler selon la nature. relationship to sports performance in a
Bos, H., van der Meij, G.J.W., Dik, K.J., 1986. Tho. Milbourne, Londres. Dutch Warmblood horse population.
Heredity of navicular disease. Vet. Q. 8, Charlier, C., Coppieters, W., Rollin, F., Equine Vet. J. 41, 139–143.
68–72. Desmecht, D., Agerholm, J.S., Cambisano, Ducro, B.J., Gorissen, B., van Eldik, P., Back,
Bowling, A.T., Byrns, G., Spiers, S., 1996. N., et al., 2008. Highly effective SNP-based W., 2009b. Influence of foot conformation
Evidence for a single pedigree source of the association mapping and management of on duration of competitive life in a Dutch
hyperkalemic periodic paralysis recessive defects in livestock. Nat. Genet. Warmblood horse population. Equine Vet.
susceptibility gene in Quarter Horses. 40, 449–454. J. 41, 144–148.
Anim. Gen. 27, 279–281. Chaudhary, R., Raudsepp, T., Guan, X.Y., Ducro, B.J., Koenen, E.P.C., van Tartwijk,
Brehm, W., Staecker, W., 2000. Zhang, H., Chowdhary, B.P., 1998. J.M.F.M., Bovenhuis, H., 2007. Genetic
Osteochondrosis in the tarsocrural joint of Zoo-FISH with microdissected arm specific relations of movement and free-jumping
Standardbred trotters. Correlation between paints for HSA2, 5, 6, 16, and 19 refines traits with dressage and showjumping
radiographic findings and racing known homology with pig and horse performance in competition of Dutch
performance. Pferdeheilkunde 16 (6), chromosomes. Mamm. Genome. 9 (1), warmblood horses. Livest. Prod. Sci. 107,
590–593. 44–49. 227–234.
Brenig B., Beck J., Hall A.J., Broad T.E., Collinder, E., Lindholm, A., Rasmuson, M., Dunlop, R.H., Williams, D.J. (Eds.), 1996.
Chowdhary B.P., Piumi F., 2004. 1997. Genetic markers in Standardbred Veterinary Medicine. An illustrated history.
Assignment of the equine solute carrier trotters susceptible to the rhabdomyolysis Mosby, St. Louis.
26A2 gene (SLC26A2) to equine syndrome. Equine Vet. J. 29, 117–120. Eivers, S.S., McGivney, B.A., Fonseca, R.G.,
chromosome 14q15–>q21 (ECA14q15– De Graaf-Roelfsema, E., Back, W., Keizer, H.A., MacHugh, D.E., Menson, K., Park, S.D.,
>q21) by in situ hybridization and Stout, T.A.E., van der Kolk, J.H., 2009. et al., 2010. Alterations in oxidative gene
radiation hybrid panel mapping. Normal function of the hypothalamic- expression in equine skeletal muscle
Cytogenet. Genome Res. 107 (1–2), 139. pituitary growth axis in three dwarf Friesian following exercise and training. Physiol.
Bricker, S.J., Brault, L.S., DelValle, A., Millon, foals. Vet. Rec. 165, 373–376. Genomics 40, 83–93.
L.V., Murray, J.D., Penedo, M.C., 2005. De Solleysel, J., 1733. Le parfait maréschal, Eizema, K., van den Burg, M.M., de Jonge,
Radiation hybrid and linkage mapping of qui enseigne a connoistre la beauté, la H.W., Dingboom, E.G., Weijs, W.A., Everts,
six new type I markers in the horse. Anim. bonté et les défauts des chevaux, 2ième M.E., 2005. Myosin heavy chain isoforms
Genet. 36 (2), 182–184. partie. Pierre-Jean Mariette, Paris. in equine gluteus medius muscle:
Bright, L.A., Burgess, S.C., Chowdhary, B., Dempsey, J.A., Wagner, P.D., 1999 Exercise- comparison of mRNA and protein
Swiderski, C.E., McCarthy, F.M., 2009. induced arterial hypoxemia. J. Appl. expression profiles. J. Histochem.
Structural and functional-annotation of an Physiol. 87 (6), 1997–2006. Cytochem. 53 (11), 1383–1390. Epub 2005
equine whole genome oligoarray. BMC Dempster, E.R., Lerner, I.M., 1950. Heritability Jun 27.
Bioinformatics 10 (Suppl. 11), S8. of threshold characters. Genetics 35 (2), Ellersieck, M.R., Lock, W.E., Vogt, D.W.,
Brinkmeyer-Langford, C., Raudsepp, T., 212–236. Aippersbach, R., 1985. Genetic evaluation
Gustafon-Seabury, A., Chowdhary, B.P., Dierks, C., Komm, K., Lampe, V., Distl, O., of cutting scores in horses. J. Equine Vet.
2008. A BAC contig map over the proximal 2010. Fine mapping of a quantitative trait Sci. 5, 287–289.
approximately 3.3 Mb region of horse locus for osteochondrosis on horse Eswarakumar, V.P., Monsonego-Ornan, E.,
chromosome 21. Cytogenet. Genome Res. chromosome 2. Anim. Genet. 41 (Suppl. Pines, M., Antonopoulou, I., Morriss-Kay,
120 (1–2), 164–172. Epub 2008 Apr 30. 2), 87–90. G.M., Lonai, P., 2002. The IIIc alternative
Brinkmeyer-Langford, C., Raudsepp, T., Lee, Dierks, C., Löhring, K., Lampe, V., Wittwer, C., of Fgfr2 is a positive regulator of bone
E.J., Goh, G., Schäffer, A.A., Agarwala, R., Drögemüller, C., Distl, O., 2007. Genome- formation. Development 129 (16),
et al., 2005. A high-resolution physical wide search for markers associated with 3783–3793.
map of equine homologs of HSA19 shows osteochondrosis in Hanoverian warmblood Falco, M.J., Whitwell, K., Palmer, A.C., 1976.
divergent evolution compared with other horses. Mamm. Genome 18, 739–747. An investigation into the genetics of

260
 References

‘wobbler disease’ in Thoroughbred deletion in the bovine myostatin gene racecourse performance. Anim. Genet. 41
horses in Britain. Equine Vet. J. 8, causes the double-muscled phenotype in (Suppl. 2), 56–63.
165–169. cattle. Nat. Genet. 17 (1), 71–74. Hill, E.W., McGivney, B.A., Gu, J., Whiston, R.,
Finno, C.J., Spier, S.J., Valberg, S.J., 2009. Grøndahl, A.M., Dolvik, N.I., 1993. Machugh, D.E., 2010c. A genome-wide
Equine diseases caused by known genetic Heritability estimation of osteochondrosis SNP-association study confirms a sequence
mutations. Vet. J. 179, 336–347. in the tibiotarsal joint and of bony variant (g.66493737C>T) in the equine
Fox, S., Filichkin, S., Mockler, T.C., 2009. fragments in the palmar/plantar portion of myostatin (MSTN) gene as the most
Applications of ultra-high-throughput the metacarpo-and metatarsophalangeal powerful predictor of optimum racing
sequencing. Methods Mol. Biol. 553, joints of horses. J. Am. Vet. Med. Assoc. distance for Thoroughbred racehorses. BMC
79–108. 203, 101–104. Genomics 11, 552.
Frame, S.R., Harrington, D.D., Fessler, J., Grøndahl, A.M., Engeland, A., 1995. Influence Hintz, R.L., 1980. Genetics of performance in
Frame, P.F., 1988. Hereditary junctional of radiographically detectable orthopedic the horse. J. Anim. Sci. 51, 582–594.
mechanobullous disease in a foal. J. Am. changes on racing performance in Holmström, M., Fredricson, I., Drevemo, S.,
Vet. Med. Assoc. 193, 1420–1424. Standardbred trotters. J. Am. Vet. Med. 1994. Biokinematic differences between
Frazer, K.A., Eskin, E., Kang, H.M., Bogue, Assoc. 206 (7), 1013–1017. riding horses with good and poor trot.
M.A., Hinds, D.A., Beilharz, E.J., et al., Gu, J., Machugh, D. E., Mcgivney, B. A., Park, Equine Vet. J. 17 (Suppl.), 51–56.
2007. A sequence-based variation map of S. D. E., Katz, L. M., Hill, E. W., 2010. Holmström, M., Magnusson, L.E., Philipsson,
8.27 million SNPs in inbred mouse strains. Association of sequence variants in CKM J., 1990. Variation in conformation of
Nature 448 (7157), 1050–1053. (creatine kinase, muscle) and COX4I2 Swedish warmblood horses and
(cytochrome c oxidase, subunit 4, isoform conformational characteristics of élite sport
Galizzi Vecchiotti Antaldi, G., 1980a. Ereditá
2) genes with racing performance in horses. Equine Vet. J. 22, 186–193.
patologica nel cavallo. I. Riv. Zootec. Vet. 9,
Thoroughbred horses. Equine Vet. J. 42, Howie, B.N., Carlson, C.S., Rieder, M.J.,
53–61.
569–575. Nickerson, D.A., 2006. Efficient selection of
Galizzi Vecchiotti Antaldi, G., 1980b. Ereditá
Gu, J., Orr, N., Park, S.D., Katz, L.M., tagging single-nucleotide polymorphisms
patologica nel cavallo. II. Riv. Zootec. Vet.
Sulimova, G., MacHugh, D.E., et al., 2009. in multiple populations. Hum. Genet. 120
9, 123–132.
A genome scan for positive selection in (1), 58–68.
Gardner, E.J., Shupe, J.C., Leone, N.C., Olson, thoroughbred horses. PLoS One 4, e5767. Hurtig, M.B., Pool, P.R., 1996. Pathogenesis of
A.E., 1975. Hereditary multiple exostosis. A
Gustafson, A.L., Tallmadge, R.L., Ramlachan, equine osteochondrosis. In: McIlwraith,
comparative genetic evaluation in man and
N., Miller, D., Bird, H., Antczak, D.F., et al., C.W., Trotter, G.W. (Eds.), Joint disease in
horses. J. Hered. 66, 318–322.
2003. An ordered BAC contig map of the the horse. Saunders, Philadelphia, pp.
Gee, E.K., Firth, E.C., Morel, P.C., Fennessy, equine major histocompatibility complex. 335–358.
P.F., Grace, N.D., Mogg, T.D., 2005 Articular Cytogenet. Genome Res. 102 (1–4), Jamison, J.M., Baird, J.D., Smith-Maxie, L.L.,
/ epiphyseal osteochondrosis in 189–195. Hulland, T.J., 1987. A congenital form of
Thoroughbred foals at 5 months of age:
Gustafson-Seabury, A., Raudsepp, T., Goh, G., myotonia with dystrophic changes in a
influences of growth of the foal and
Kata, S.R., Wagner, M.L., Tozaki, T., et al., quarterhorse. Equine Vet. J. 19, 353–358.
prenatal copper supplementation of the
2005. High-resolution RH map of horse Jeffcott, L.B., 1991. Osteochondrosis in the
dam. N. Z Vet. J. 53 (6), 448–456.
chromosome 22 reveals a putative ancestral horse – searching for the key to
Gerber, H., Gaillard, C.L., Fatzer, R. et al., vertebrate chromosome. Genomics 85 (2), pathogenesis. Equine Vet. J. 13, 331–338.
1995. Cerebellär abiotrophie bei 188–200. Jeffcott, L.B., 1997. Osteochondrosis in horses.
vollblutaraber-fohlen. Pferdeheilk. 11, Hanson, J.M., Mol, J.A., Leegwater, P.A., In Practice 19, 64–71.
423–431. Kooistra, H.S., Meij, B.P., 2006. The Jeffcott, L.B., Henson, F.M., 1998. Studies on
Gerber, V, Bailey, E., 1995. Genetics and leukemia inhibitory factor receptor gene is growth cartilage in the horse and their
disease in the horse. Equine Vet. J. 27, not involved in the etiology of pituitary application to aetiopathogenesis of
400–401. dwarfism in German shepherd dogs. Res. dyschondroplasia (osteochondrosis). Vet. J.
Gibala, M., 2009 Molecular responses to Vet. Sci. 81 (3), 316–320. 156, 177–192.
high-intensity interval exercise. Appl. Hermans, W.A., 1970. A hereditary anomaly in Johnson, G.C., Kohn, C.W., Johnson, C.W.,
Physiol. Nutr. Metab. 34 (3), 428–432. Shetland ponies. Neth. J. Vet. Sci. 3, 55–63. Garry, F., Scott, D., Martin, S., 1988.
Gibson, G., 1998 Active role of chondrocyte Hermans, W.A., Kersjes, A.W., van der Mey, Ultrastructure of junctional epidermolysis
apoptosis in endochondral ossification. G.J.W., Dik, K.J., 1987. Investigations into bullosa in Belgian foals. J. Comp. Pathol.
Microsc. Res. Tech. 43 (2), 191–204. the heredity of congenital lateral patellar 99, 329–336
Gläser, K.E., Sun, Q., Wells, M.T., Nixon, A.J., (sub)luxation in the Shetland pony. Vet. Q. Jose-Cunilleras, E., Hayes, K.A., Toribio, R.E.,
2009. Development of a novel equine 9: 1–8. Mathes, L.E., Hinchcliff, K.W., 2005.
whole transcript oligonucleotide GeneChip Herszberg, B., McCue, M.E., Larcher, T., Mata, Expression of equine glucose transporter
microarray and its use in gene expression X., Vaiman, A., Chaffaux, S., et al., 2009. A type 4 in skeletal muscle after glycogen-
profiling of normal articular-epiphyseal GYS1 gene mutation is highly associated depleting exercise. Am. J. Vet. Res. 66 (3),
cartilage. Equine Vet. J. 41, 663–670. with polysaccharide storage myopathy in 379–385.
Goh, G., Raudsepp, T., Durkin, K., Wagner, Cob Normand draught horses. Anim. Karlsson, E.K., Baranowska, I., Wade, C.M.,
M.L., Schäffer, A.A., Agarwala, R., et al., Genet. 40, 94–96. Salmon Hillbertz, N.H., Zody, M.C.,
2007 High-resolution gene maps of horse Hill, E.W., 2010b. Livestock issue. Brief Funct Anderson, N., et al., 2007. Efficient
chromosomes 14 and 21, additional Genomics 9 (3), 191–192. mapping of Mendelian traits in dogs
insights into evolution and rearrangements Hill, E.W., Gu, J., Eivers, S.S., Fonseca, R.G., through genome-wide association. Nat.
of HSA5 homologs in mammals. McGivney, B.A., Govindarajan, P., et al., Genet. 39 (11), 1321–1328.
Genomics 89 (1), 89–112. 2010a. A sequence polymorphism in MSTN Kinderen L. der., 2005. Heritability of
Graves, K.T., Henney, P.J., Ennis, R.B., 2009. predicts sprinting ability and racing osteochondrosis in Dutch warmblood
Partial deletion of the LAMA3 gene is stamina in thoroughbred horses. PLoS One stallions from the second stallion
responsible for hereditary junctional 5, e8645. inspection. MSc. thesis. Wageningen
epidermolysis bullosa in the American Hill, E.W., Gu, J., McGivney, B.A., MacHugh, University, The Netherlands.
Saddlebred Horse. Anim. Genet. 40, 35–41. D.E., 2010d. Targets of selection in the Klukowska-Rötzler, J., Bugno, M., Sander, P.,
Grobet, L., Martin, L.J., Poncelet, D., Pirottin, Thoroughbred genome contain exercise- Slota, E., Dolf, G., Chowdhary, B.P., et al.,
D., Brouwers, B., Riquet, J., et al., 1997 A relevant gene SNPs associated with elite 2006b. Chromosomal assignment of the

261
 12 Genetic contributions to exercise and athletic performance

two candidate genes (EGFR, CLCA1) for Li, J.K., Moloney, B.K., Shupe, J.L., Gardner, myopathy in horses. Vet. Pathol. 46,
equine recurrent airway obstruction (RAO) E.J., Leone, N.C., Elsner, Y., 1989. DNA 1281–1291.
by FISH and RH mapping. Hereditas. polymorphism analysis of hereditary McCue M.E., Bannasch D.L., Petersen J.L.,
143:138–141. multiple exostoses in horses. Am. J. Vet. Gurr J., Bailey E., Binns M.M., et al.,
Klukowska-Rötzler, J., Jost, U., Schelling, C., Res. 50, 978–983. 2012. A high density SNP array for the
Dolf, G., Chowdhary, B.P., Leeb, T., et al., Lieto, L.D., Cothran, E.G., 2003. The domestic horse and extant Perissodactyla:
2006a. Characterization and RH mapping epitheliogenesis imperfecta locus maps to utility for association mapping, genetic
of six gene-associated equine microsatellite equine chromosome 8 in American diversity, and phylogeny studies. PLoS
markers. Anim. Genet. 37 (3), 305–306. Saddlebred horses. Cytogenet. Genome Res. Genet. 8 (1), e1002451. doi: 10.1371/
Koenen, E.P.C., van Veldhuizen, A.E., 102, 207–210. journal.pgen.1002451. Epub 2012
Brascamp, E.W., 1995. Genetic parameters Lieto, L.D., Swerczek, T.W., Cothran, E.G., Jan 12.
of linear scored conformation traits and 2002. Equine epitheliogenesis imperfecta McCue, M.E., Ribeiro, W.P., Valberg, S.J., 2006.
their relation to dressage and show- in two american saddlebred foals is a Prevalence of polysaccharide storage
jumping performance in the Dutch lamina lucida defect. Vet. Pathol. 39, myopathy in horses with neuromuscular
Warmblood riding horse population. 576–580. disorders. Equine Vet. J. 36 (Suppl.),
Livest. Prod. Sci. 43, 85–94. Lindgren, G., Sandberg, K., Persson, H., et al., 340–344.
Kronenberg, H.M., 2003. Developmental 1998. A primary male autosomal linkage McCue, M.E., Valberg, S.J., 2007. Estimated
regulation of the growth plate. Nature 423, map of the horse genome. Genome Res. 8, prevalence of polysaccharide storage
332–336. 951–966. myopathy among overtly healthy Quarter
Kvist, A.J., Johnson, A.E., Mörgelin, M., Lira, F.S., Yamashita, A., Carnevali Jr, L.C., Horses in the United States. J. Am. Vet.
Gustafsson, E., Bengtsson, E., Lindblom, K., Gonçalves, D.C., Lima, W.P., Rosa, J.C., Med. Assoc. 231, 746–750.
et al., 2006. Chondroitin sulfate perlecan et al., 2010. Exercise training reduces PGE2 McCue, M.E., Valberg, S.J., Jackson, M., Borgia,
enhances collagen fibril formation. levels and induces recovery from steatosis L., Lucio, M., Mickelson, J.R., 2009a.
Implications for perlecan in tumor-bearing rats. Horm. Metab. Res. Polysaccharide storage myopathy
chondrodysplasias. J. Biol. Chem. 281 (44), 42 (13), 944–949. phenotype in Quarter Horse-related breeds
33127–33139. Lonka, T., 1946. The evaluation of the pulling is modified by the presence of an
KWPN, 1994. The frequency and heredity of ability of horses. Valt. Maatalousk. Julk. dominant RYR1 mutation. Neuromuscul.
navicular disease, sesamoidosis, fetlock joint 126, 45–50. Disord. 19, 37–43.
arthrosis, bone spavin and osteochondrosis Lopes, M.S., Diesterbeck, U., da Câmara McCue, M.E., Valberg, S.J., Lucio, M.,
of the hock. A radiographic progeny study. Machado, A., Distl, O., 2009. Fine mapping Mickelson, J.R., 2008b. Glycogen synthase
Koninklijke Vereniging Warmbloed a quantitative trait locus on horse 1 (GYS1) mutation in diverse breeds with
Paardenstamboek Nederland, Zeist. chromosome 2 associated with radiological polysaccharide storage myopathy. J. Vet.
Lampe, V., Dierks, C., Distl, O., 2009a. signs of navicular disease in Hanoverian Intern. Med. 22, 1228–1233.
Refinement of a quantitative trait locus on warmblood horses. Anim. Genet. 40, McCue, M.E., Valberg, S.J., Miller, M.B., Wade,
equine chromosome 5 responsible for 955–957. C., DiMauro, S., Akman, H.O., et al.,
fetlock osteochondrosis in Hanoverian Lundby, C., Calbet, J.A., Robach, P., 2009. The 2008a. Glycogen synthase (GYS1) mutation
warmblood horses. Anim. Genet. 40, response of human skeletal muscle tissue causes a novel skeletal muscle glycogenosis.
553–555. to hypoxia. Cell Mol. Life Sci. 66 (22), Genomics 91, 458–466.
Lampe, V., Dierks, C., Komm, K., Distl, O., 3615–3623. McGivney, B.A., Eivers, S.S., MacHugh, D.E.,
2009b. Identification of a new quantitative Lykkjen, S., Dolvik, N.I., McCue, M.E., MacLeod, J.N., O’Gorman, G.M., Park,
trait locus on equine chromosome 18 Rendahl, A.K., Mickelson, J.R., Roed, K.H., S.D., et al., 2009. Transcriptional
responsible for osteochondrosis in 2010. Genome-wide association analysis of adaptations following exercise in
Hanoverian warmblood horses. J. Anim. osteochondrosis of the tibiotarsal joint in thoroughbred horse skeletal muscle
Sci. 87:3477–3481. Norwegian Standardbred trotters. Anim. highlights molecular mechanisms that lead
Langlois, B., Froidevaux, J., Lamarche, L. et al., Genet. 41 (Suppl. 2), 111–120. to muscle hypertrophy. BMC Genomics 10,
1978. Analyse des liaisons entre la Magdesian, K.G., 2009. Malignant 638.
morphologie et l’aptitude au galop, au trot hyperthermia associated with ryanodine McGivney, B.A., McGettigan, P.A., Browne, J.A.,
et au saut d’obstacles chez le cheval. Ann. receptor 1 (C7360G) mutation in Evans, A.C., Fonseca, R.G., Loftus, B.J.,
Génét. Sél. Anim. 10, 443–474. Quarter Horses. J. Vet. Intern. Med. 23, et al., 2010. Characterization of the equine
Lee, E.J., Raudsepp, T., Kata, S.R., Adelson, D., 329–334. skeletal muscle transcriptome identifies
Womack, J.E., et al., 2004. A 1.4-Mb Mardis, E.R., 2008. Next-generation DNA novel functional responses to exercise
interval RH map of horse chromosome 17 sequencing methods. Ann. Rev. Genomics training. BMC Genomics 11, 398.
provides detailed comparison with human Hum. Genet. 9, 387–402. McGuire, T.C., Poppie, M.J., Banks, K.L., 1974.
and mouse homologues. Genomics 83, Martin, A.M., Elliott, J.A., Duffy, P., Blake, Combined (B- and T-lymphocyte)
203–215 C.M., Ben Attia, S., Katz, L.M., et al., 2010. immunodeficiency: a fatal genetic disease
Leeb, T., Bruhn, O., Philipp, U., Kuiper, H., Circadian regulation of locomotor activity in Arabian foals. J. Am. Vet. Med. Assoc.
Regenhard, P., Paul, S., et al., 2005. and skeletal muscle gene expression in the 164, 70–76.
Assignment of the equine S100A7 gene horse. J. Appl. Physiol. 109 (5), McIlwraith, C.W., 1986. Incidence of
(psoriasin 1) to chromosome 5p12–>p13 1328–1336. developmental joint problems. In:
by fluorescence in situ hybridization and McCabe, L., Griffin, L.D., Kinzer, A., Chandler, McIlwraith, C.W. (Ed.), AQHA
radiation hybrid mapping. Cytogenet M., Beckwith,J.B., McCabe, E.R., 1990. developmental orthopedic disease
Genome Res. 109 (4), 533. Overo lethal white foal syndrome: equine symposium. AQHA, Amarillo, pp. 15–20.
Leone, N.C., Shupe, J.L., Gardner, E.J., Millar, model of aganglionic megacolon McIlwraith, C.W., 1993a. Inferences from
E.A., Olson, A.E., et al., 1987. Hereditary (Hirschsprung disease). Am. J. Med. Genet. referred clinical cases of osteochondritis
multiple exostosis. A comparative 36, 336–340. dissecans. Equine Vet. J. 16:27–30.
human-equine-epidemiologic study. J. McCue, M.E., Armién, A.G., Lucio, M., McIlwraith, C.W., 1993b. What is
Hered. 78, 171–177. Mickelson, J.R., Valberg, S.J., 2009b. developmental orthopedic disease,
Levine, M.A., 1999. Investigating the origins of Comparative skeletal muscle osteochondrosis, osteochondritis,
horse domestication. Equine Vet. J. Suppl. histopathologic and ultrastructural features metabolic bone disease? Proc. Am. Assoc.
(28), 6–14. in two forms of polysaccharide storage Equine Practnr. 39:35–44.

262
 References

McPherron, A.C., Lawler, A.M., Lee, S.J., 1997. Arabian horse. Transplantation 31, Orr, N., Back, W., Gu, J., Leegwater, P.,
Regulation of skeletal muscle mass in mice 290–294. Govindarajan, P., Conroy, J., et al., 2010.
by a new TGF-beta superfamily member. Mucher, E., Jayr, L., Rossignol, F., Amiot, F., Genome-wide SNP association based
Nature 387 (6628), 83–90. Gidrol, X., Barrey, E., 2006. Gene localization of a dwarfism gene in Friesian
McPherron, A.C., Lee, S.J., 1997. Double expression profiling in equine muscle dwarf horses. Anim. Gen. 41 (Suppl. 2),
muscling in cattle due to mutations in the tissues using mouse cDNA microarrays. 2–7.
myostatin gene. Proc. Natl. Acad. Sci. U S Equine Vet. J. Suppl. (36), 359–364. Orr, N., Chanock, S., 2008. Common genetic
A. 94 (23), 12457–12461. Müller, D., Kuiper, H., Böneker, C., Mömke, S., variation and human disease. Adv. Genet.
Meyer, T.S., Fedde, M.R., Cox, J.H., Erickson, Drögemüller, C., Chowdhary, B.P., et al., 62, 1–32.
H.H., 1999. Hyperkalaemic periodic 2005b. Physical mapping of the PTHR1 Osinga, A., 2000. Het Fokken van het Friese
paralysis in horses: a review. Equine Vet. J. gene to equine chromosome 16q21.2. Paard. Schaafsma & Brouwer, Dokkum.
31, 362–367. Anim. Genet. 36 (3), 282–284. Pagan, J.D., S.G. Jackson, 1996. The incidence
Mickelson, J.R., Wagner, M.L., Goh, G., Wu, Müller, D., Kuiper, H., Böneker, C., Mömke, S., of developmental orthopedic disease on a
J.T., Morrison, L.Y., Alexander, L.J., et al., Drögemüller, C., Chowdhary, B.P., et al., Kentucky Thoroughbred farm.
2004. Thirty-five new equine microsatellite 2005c. Assignment of BGLAP, BMP2, Pferdeheilunde 12:351–354.
loci assigned to genetic linkage and CHST4, SLC1A3, SLC4A1, SLC9A5 and Penedo, M.C., Millon, L.V., Bernoco, D.,
radiation hybrid maps. Anim. Genet. 35 SLC20A1 to equine chromosomes by FISH Bailey, E., Binns, M., Cholewinski, G., et
(6), 481–484. and confirmation by RH mapping. Anim. al., 2005. International Equine Gene
Milenkovic, D., Chaffaux, S., Taourit, S., Genet. 36 (5), 457–461. Mapping Workshop Report: a
Guérin, G., 2003. A mutation in the Müller, D., Kuiper, H., Mömke, S., Böneker, C., comprehensive linkage map constructed
LAMC2 gene causes the Herlitz junctional Drögemüller, C., Chowdhary, B.P., et al., with data from new markers and by
epidermolysis bullosa (H-JEB) in two 2005a. Assignment of the COMP gene to merging four mapping resources.
French draft horse breeds. Genet. Sel. Evol. equine chromosome 21q12-q14 by FISH Cytogenet. Genome Res. 111 (1), 5–15.
35, 249–256. and confirmation by RH mapping. Anim. Perrocheau, M., Boutreux, V., Chadi, S., Mata,
Milenkovic, D., Oustry-Vaiman, A., Lear, T.L., Genet. 36 (3), 277–279. X., Decaunes, P., Raudsepp, T., et al., 2006.
Billault, A., Mariat, D., Piumi, F., et al., Naylor, J.M., 1994a. Equine hyperkalemic Construction of a medium-density horse
2002. Cytogenetic localization of 136 genes periodic paralysis: review and implications. gene map. Anim. Genet. 37 (2), 145–155.
in the horse: comparative mapping with Can. Vet. J. 35, 279–285. Perrocheau, M., Boutreux, V., Chadi-Taourit,
the human genome. Mamm. Genome. 13 Naylor, J.M., 1994b. Selection of Quarter S., Di Meo, G.P., Perucatti, A., Incarnato,
(9), 524–534. Horses affected with hyperkalemic periodic D., et al., 2005. Equine FISH mapping of
Mömke, S., Distl, O., 2007. Molecular genetic paralysis by show judges. J. Am. Vet. Med. 36 genes known to locate on human
analysis of the ATP2A2 gene as candidate Assoc. 204, 926–928. chromosome ends. Cytogenet. Genome
for chronic pastern dermatitis in German Naylor, J.M., Jones, V., Berry, S.L., 1993. Res. 111 (1), 46–50.
draft horses. J. Hered. 98, 267–271. Clinical syndrome and diagnosis of Petersen, J.L., Mickelson, J.R., Rendahl, A.K.,
Momozawa, Y., Kusunose, R., Kikusui, T., hyperkalaemic periodic paralysis in Quarter Valberg, S.J., Andersson, L.S., Axelsson, J.,
Takeuchi, Y., Mori, Y., 2005. Assessment of Horses. Equine Vet. J. 25, 227–232. et al., 2013. Genome-wide analysis reveals
equine temperament questionnaire by Naylor, J.M., Nickel, D.D., Trimino, G., Card, selection for important traits in domestic
comparing factor structure between two C., Lightfoot, K., Adams, G., 1999. horse breeds. PLoS Genet. 9, e1003211.
separate surveys. App. Anim. Behav. Hyperkalaemic periodic paralysis in Philipsson, J., Andréasson, E., Sandgren, B.,
Science. 92, 77–84. homozygous and heterozygous horses: a Dalin, G., Carlsten, J., 1993.
Momozawa, Y., Takeuchi, Y., Tozaki, T., co-dominant genetic condition. Equine Vet. Osteochondrosis in the tarsocrural joint
Kikusui, T., Hasegawa, T., Raudsepp, T., et J. 31, 153–159. and osteochondral fragments in the fetlock
al., 2007a. SNP detection and radiation Naylor, J.M., Robinson, J.A., Bertone, J., 1992. joints in Standardbred trotters. II.
hybrid mapping in horses of nine Familial incidence of hyperkalemic periodic Heritability. Equine Vet. J. 16 (Suppl.),
candidate genes for temperament. Anim. paralysis in Quarter Horses. J. Am. Vet. 38–41.
Genet. 38 (1), 81–83. Med. Assoc. 200, 340–343. Pickar, J.G., Spier, S.J., Snyder, R.J., Carlsen,
Momozawa, Y., Terada, M., Sato, F., Kikusui, Naylor, J.M., Robinson, J.A., Crichlow, E.C., A.C., 1991. Altered ion permeability in
T., Takeuchi, Y., Kusunose, R., et al., 2007b. Steiss, J.E., 1992. Inheritance of myotonic skeletal muscle from horses with
Assessing equine anxiety-related parameters discharges in American Quarter Horses and hyperkalemic periodic paralysis. Am. J.
using an isolation test in combination with the relationship to hyperkalemic periodic Physiol. 260, C926–C933.
a questionnaire survey. J. Vet. Med. Sci. 69 paralysis. Can. J. Vet. Res. 56, 62–66. Pieramati, C., Pepe, M., Silvestrelli, M., Bolla,
(9), 945–950. Németh, F., 1974. De sesambeenskreupelheid A., 2003. Heritability estimation of
Monteiro, J., Barata, G., 1980. Exostoses no bij het paard. PhD thesis, Utrecht osteochondrosis dissecans in Maremmano
cavalo. Estudo da sua hereditariedade. Rev. University. horses. Livest. Prod. Sci. 79, 249–255.
Port. Ciênc. Vet. 75, 31–39. Nes, N., Løno, O.M., Bjerkås, I., 1982. Pilegaard H., Neufer P.D., 2004.
Morozova, O., Hirst, M., Marra, M.A., 2009. Hereditary lethal arthrogryposis (‘muscle Transcriptional regulation of pyruvate
Applications of new sequencing contracture’) in horses. Nord. Vet. Med. 34, dehydrogenase kinase 4 in skeletal muscle
technologies for transcriptome analysis. 425–430. during and after exercise Proc. Nutr. Soc.
Ann. Rev. Genom. Hum. Genet. Nollet, H., Deprez, P., 2005. Hereditary 63, 221–226.
10:135–151. skeletal muscle diseases in the horse. A Preisinger, R., Wilkens, J., Kalm, E., 1991.
Mosher, D.S., Quignon, P., Bustamante, C.D., review. Vet. Q. 27, 65–75. Estimation for conformation traits for foals
Sutter, N.B., Mellersh, C.S., Parker, H.G., Numans, S.R., van de Watering, C.C., 1973. and their practical implications. Livest.
et al., 2007. A mutation in the myostatin Navicular disease: podotrochleitis chronica Prod. Sci. 29, 77–86.
gene increases muscle mass and enhances aseptica, podotrochleosis. Equine Vet. J. 5, Radiostits, O.M., 2007. Veterinary medicine; A
racing performance in heterozygote dogs. 1–7. textbook of the diseases of cattle, horses,
PLoS Genet. 3 (5), e79. Olesen, J., Kiilerich, K., Pilegaard, H., 2010. sheep, pigs and goats.
Mottironi, V.D., Perryman, L.E., Pollara, B., PGC-1alpha-mediated adaptations in Raudsepp, T., Frönicke, L., Scherthan, H.,
Mickey, M.R., Swift, R., McGrath, P., 1981. skeletal muscle. Pflugers Arch. 460 (1), Gustavsson, I., Chowdhary, B.P., 1996.
Major histocompatibility locus in the 153–162. Zoo-FISH delineates conserved

263
 12 Genetic contributions to exercise and athletic performance

chromosomal segments in horse and man. osteochondrosis in a selected population of Storgaard Jorgensen, H., Proschowsky, H.,
Chromosome Res. 4 (3), 218–225. trotting horses in Denmark and its possible Falk-Ronne, J., Willeberg, P., Hesselholt, M.,
Raudsepp, T., Lee, E.J., Kata, S.R., Brinkmeyer, genetic significance. Equine Vet. J. 22, 1997. The significance of routine
C., Mickelson, J.R., Skow, L.C., et al., 288–289. radiographic findings with respect to
2004a. Exceptional conservation of Schuelke, M., Wagner, K.R., Stolz, L.E., subsequent racing performance and
horse-human gene order on X chromosome Hübner, C., Riebel, T., Kömen, W., et al., longevity in Standardbred trotters. Equine
revealed by high-resolution radiation 2004. Myostatin mutation associated with Vet. J. 29 (1), 55–59.
hybrid mapping. Proc. Natl. Acad. Sci. gross muscle hypertrophy in a child. N. Swinburne, J.E., Boursnell, M., Hill, G., Pettitt,
U S A. 101 (8), 2386–2391. Engl. J. Med. 350 (26), 2682–2688. L., Allen, T., Chowdhary, B., et al., 2006.
Raudsepp, T., Santani, A., Wallner, B., Kata, Serteyn, D., Piquemal, D., Vanderheyden, L., Single linkage group per chromosome
S.R., Ren, C., Zhang, H.B., et al., 2004b. A Lejeune, J.P., Verwilghen, D., Sandersen, C., genetic linkage map for the horse, based on
detailed physical map of the horse Y 2010. Gene expression profiling from two three-generation, full-sibling, crossbred
chromosome. Proc. Natl. Acad. Sci. U S A. leukocytes of horses affected by horse reference families. Genomics 87 (1),
101 (25), 9321–9326. osteochondrosis. J. Orthop. Res. 28, 1–29.
Ricard, A., 1998. Developments in the genetic 965–970. Swinburne, J.E., Lockhart, L., Aldridge, V.,
evaluation of performance traits in horses. Shupe, L., Leone, N.C., Olson, A.E., Gardner, Marti, E., Breen, M., Binns, M.M., 2000.
Proceeding of the World Congress on E.J., 1979. Hereditary multiple exostoses: Characterisation of 25 new physically
Genetics Applied to Livestock Production clinicopathologic features of a comparative mapped horse microsatellite loci:
6, 388–395. study in horses and man. Am. J. Vet. Res. AHT24++-48. Anim. Genet. 31 (3),
Ricard, A., 2002. Genetic background of 40, 751–757. 237–238.
osteochondrosis. Proceeding of the 55th Spier, S.J., Carlson, G.P., Harrold, D., Bowling, Takahashi, T., Yawata, M., Raudsepp, T., Lear,
Annual Meeting of the European A., Byrns, G., Bernoco, D., 1993. Genetic T.L., Chowdhary, B.P., Antczak, D.F., et al.,
Association of Animal Producers, Bled, study of hyperkalemic periodic paralysis in 2004. Natural killer cell receptors in the
Slovenia, Sep. 5–8, p. 304. horses. J. Am.Vet. Med. Assoc. 202, horse: evidence for the existence of
933–937. multiple transcribed LY49 genes. Eur. J.
Robert, C., Valette, J.P., Denoix, J.M., 2006
Correlation between routine radiographic Stanley, R.L., McCue, M.E., Valberg, S.J., Immunol. 34 (3), 773–784.
findings and early racing career in French Mickelson, J.R., Mayhew, I.G., McGowan, Tallmadge, R.L., Lear, T.L., Antczak, D.F., 2005.
trotters. Equine Vet. J. Suppl. (36), C., et al., 2009. A glycogen synthase 1 Genomic characterization of MHC class I
473–478. mutation associated with equine genes of the horse. Immunogenetics 57
polysaccharide storage myopathy and (10), 763–774.
Rossdale, P.D., 1995. Genetics and disease in
exertional rhabdomyolysis occurs in a
the horse: Heinz Gerber International Thorén Hellsten, E., Viklund, A., Koenen,
variety of UK breeds. Equine Vet. J. 41,
Workshop. Equine Vet. J. 27, 411–415. E.P.C., Ricard, A., Bruns, E., Philipsson, J.,
597–601.
Rudolph, J.A., Spier, S.J., Byrns, G., Hoffman, 2006. Review of genetic parameters
Stashak, T.S., 1987a. The relationship between
E.P., 1992a. Linkage of hyperkalaemic estimated at stallion and young horse
conformation and lameness. In: Stashak,
periodic paralysis in Quarter Horses to the performance tests and their correlations
T.S. (Ed.), Adams’ lameness in horses.
horse adult skeletal muscle sodium channel with later results in dressage and show-
fourth ed. Philadelphia, Lea and Febiger,
gene. Anim. Gen. 23, 241–250. jumping competition. Livest. Sci. 103,
pp. 71–99.
Rudolph, J.A., Spier, S.J., Byrns, G., Rojas, C.V., 1–12.
Stashak, T.S., 1987b. Bone spavin
Bernoco, D., Hoffman, E.P., 1992b. Tozaki, T., Hirota, K., Hasegawa, T., Ishida, N.,
(osteoarthritis or degenerative joint disease
Periodic paralysis in Quarter Horses: a Tobe, T., 2007. Whole-genome linkage
of the distal tarsal joints). In: Stashak, T.S.
sodium channel mutation disseminated by disequilibrium screening for complex traits
(Ed.), Adams’ lameness in horses. fourth
selective breeding. Nat. Genet. 2, 144–147. in horses. Mol. Genet. Genomics 277 (6),
edn. Philadelphia, Lea and Febiger, p. 695.
Samoré, A.B., Pagnacco, G., Miglior, F., 1997. 663–672.
Steemers, F.J., Gunderson, K.L., 2007. Whole
Genetic parameters and breeding values for genome genotyping technologies on the Tozaki, T., Hirota, K., Sugita, S., Ishida, N.,
linear type traits in the Haflinger horse. BeadArray platform. Biotechnol. J. 2 (1), Miyake, T., Oki, H., et al., 2010b. A
Livest. Prod. Sci. 52, 105–111. 41–49. genome-wide scan for tying-up syndrome
Sandgren, B., Dalin, G., Carlsten, J., 1993. in Japanese Thoroughbreds. Anim. Genet.
Stock, K.F., Hamann, H., Distl, O., 2005.
Osteochondrosis in tarsocrural joint and 41 (Suppl. 2), 80–86.
Estimation of genetic variables for the
osteochondral fragments in the fetlock prevalence of osseous fragments in limb Tozaki, T., Miyake, T., Kakoi, H., Gawahara, H.,
joints in Standardbred trotters. I. joints of Hanoverian Warmblood horses. Sugita, S., Hasegawa, T., et al., 2010a. A
Epidemiol. Equine Vet. J. (Suppl. 16), J. Anim. Breed. Genet. 122, 271–280. genome-wide association study for racing
31–37. Stock, K.F., Distl, O., 2005. Evaluation of performances in Thoroughbreds clarifies a
Scarpulla, R.C., 2008. Nuclear control of expected response to selection for candidate region near the MSTN gene.
respiratory chain expression by nuclear orthopedic health and performance traits Anim. Genet. 41 (Suppl. 2), 28–35.
respiratory factors and PGC-1-related in Hanoverian Warmblood horses. Am. J. Tryon, R.C., Penedo, M.C., McCue, M.E.,
coactivator. Ann. N. Y. Academy of Vet. Res. 66, 1371–1379. Valberg, S.J., Mickelson, J.R., Famula, T.R.,
Sciences. 1147, 321–334. Stock, K.F., Distl, O., 2006. Genetic analyses of et al., 2009. Evaluation of allele frequencies
Schauder, W., 1923. Historisch-kritische Studie the radiographic appearance of the distal of inherited disease genes in subgroups of
über die Bewegungslehre des Pferdes. (1. sesamoid bones in Hanoverian Warmblood American Quarter Horses. J. Am.Vet. Med.
Teil) Berl. Tierärztl. Wschr. 39, 123–126. horses. Am. J. Vet. Res. 67, 1013–1019. Assoc. 234, 120–125.
Schober, M., Coenen, M., Distl, O., Hertsch, Stock, K.F., Distl, O., 2006. Genetic Valberg, S.J., Cardinet 3rd, G.H., Carlson, G.P.,
B., Christmann, L., Bruns, E., 2003. correlations between conformation traits DiMauro, S., 1992. Polysaccharide storage
Estimation of genetic parameters of and radiographic findings in the limbs of myopathy associated with recurrent
osteochondrosis (OC) in Hanoverian German Warmblood riding horses. Genet. exertional rhabdomyolysis in horses.
Warmblood foals. In: Proceeding of 54th Sel. Evol. 38, 657–671. Neuromuscul. Disord. 2, 351–359.
Annual Meeting of Europe Association for Stockham, S.L., Harvey, J.W., Kinden, D.A., Valberg, S.J., Geyer, C., Sorum, S.A., Cardinet
Animal Production Rome, Italy. 1994. Equine glucose-6-phosphate III, G.H., 1996. Familial basis of exertional
Schougaard, H., Falk Rønne, J., Philipsson, J., dehydrogenase deficiency. Vet. Pathol. 31 rhabdomyolysis in Quarter Horse related
1990. A radiographic study of tibiotarsal (5), 518–527. breeds. Am. J. Vet. Res. 57, 286–290.

264
 References

Valberg, S.J., Townsend, D., Mickelson, J.R., Varo, M., 1947. The development of the immunohistological and ultrastructural
1998. Skeletal muscle glycolytic capacity Finnish horses’ pulling power. Maataloust. findings. Vet. Dermatol. 15, 207–217.
and phosphofructokinase regulation in Aikakausk. 19, 69–82. Willms, F.; Röhe, R.; Kalm, E., 1999,
horses with polysaccharide storage Vaughan, LC., 1976. Growth plate defects in Genetische Analyse von
myopathy. Am. J. Vet. Res. 59, 782–785 dogs. Vet Rec. 98 (10), 185–189. Merkmalskomplexen in der Reitpferdezucht
Valberg, S.J., Ward, T.L., Rush, B., Kinde, H., Vervuert, I., Winkelsett, S., Christmann, L., unter Berücksichtigung von
Hiraragi,H., Nahey, D., et al., 2001. Bruns, E., Hoppen, H.O., Distl, O., et al., Gliedmaßenveränderungen. 1. Mitteilung:
Glycogen branching enzyme deficiency in 2007. Evaluation of the influences of Züchterische Bedeutung von
Quarter horse foals. J. Vet. Intern. Med. 15, exercise, birth date, and osteochondrosis Gliedmaßenveränderungen.
572–580. on plasma bone marker concentrations in Züchtungskunde 71, 330–345.
Valentine, B.A., Credille, K.M., Lavoie, J.P., Hanoverian Warmblood foals. Am. J. Vet. Winter, D., Bruns, E., Glodek, P., Hertsch, B.,
Fatone, S., Guard, C., Cummings, J.F., et al., Res. 68, 1319–1323. 1996. Genetische Disposition von
1997. Severe polysaccharide storage Von Butler-Wemken, I., Duda, J., Kaiser, M., Gliedmaβenerkrankungen beim Reitpferd.
myopathy in Belgian and Percheron 1992. Genetische Analysen für Züchtungsk. 68, 92–108.
draught horses. Equine Vet. J. 29, 220–225. Exterieurgesamtbeurteilungen und Wittwer, C., Chowdhary, B.P., Distl, O., 2005.
Van Bergen, H.M.J.M., van Arendonk, J.A.M., Beziehungen zu Körpermaβen bei Radiation hybrid mapping of equine
1993. Genetic parameters for linear type Trakehner Stuten. Züchtungsk. 64, 92–100. CDK2, DGKA, DNAJC14, MMP19,
traits in Shetland ponies. Livest. Prod. Sci. CTSL and GAS1. Anim. Genet. 36 (6),
Wade, C.M., Giulotto, E., Sigurdsson, S., Zoli,
36, 273–284. 536–537.
M., Gnerre, S., Imsland, F., et al., 2009.
Van de Lest, C.H., van den Hoogen, B.M., van Genome sequence, comparative analysis, Wittwer, C., Dierks, C., Hamann, H., Distl, O.,
Weeren, P.R., Brouwers, J.F., van Golde, and population genetics of the domestic 2008. Associations between candidate gene
L.M., Barneveld, A., 1999. Changes in bone horse. Science 326, 865–867. markers at a quantitative trait locus on
morphogenic enzymes and lipid equine chromosome 4 responsible for
Wagner, M.L., Valberg, S.J., Ames, E.G., Bauer,
composition of equine osteochondrotic osteochondrosis dissecans in fetlock joints
M.M., Wiseman, J.A., Penedo, M.C., et al.,
subchondral bone. Equine Vet. J. Suppl. of South German Coldblood horses.
2006. Allele frequency and likely impact of
(31), 31–37. J. Hered. 99, 125–129.
the glycogen branching enzyme deficiency
Van Den Belt, A. J., Dik K. J., Van Den Broek gene in Quarter Horse and Paint Horse Wittwer, C., Hamann, H., Distl, O., 2009. The
J., 2003. The use of radiography as a tool populations. J. Vet. Intern. Med. 20, candidate gene XIRP2 at a quantitative
for the efficacy of selective breeding of a 1207–1211. gene locus on equine chromosome 18
poor radiographic navicular bone associated with osteochondrosis in fetlock
Wagner, P.C., Grant, B.D., Watrous, B.J., and hock joints of South German
condition (grade3 and 4). In: Proceedings
Appell, L.H., Blythe, L.L., 1985. A study of Coldblood horses. J. Hered. 100,
of the 13th International Veterinary
the heritability of cervical malformation in 481–486.
Radiological Society, Midrand, South
horses. Proc. Am. Assoc. Equine Pract. 31,
Africa. Wittwer, C., Hamann, H., Rosenberger, E.,
43–50.
Van der Veen, G., Kingmans, J., van Distl, O., 2007b. Genetic parameters for the
Ward, T.L., Valberg, S.J., Adelson, D.L., Abbey, prevalence of osteochondrosis in the limb
Veldhuizen, A.E., et al., 1994. The
C.A., Binns, M.M., Mickelson, J.R., 2004. joints of South German Coldblood horses.
frequency and heritability of navicular
Glycogen branching enzyme (GBE1) J. Anim. Breed Genet. 124, 302–307.
disease, sesamoidosis, fetlock joint
mutation causing equine glycogen storage
arthrosis, bone spavin, osteochondrosis of Wittwer, C., Lohring, K., Drogemuller, C.,
disease IV. Mamm. Genome 15, 570–577.
the hock: a radiographic progeny study. Hamann, H., Rosenberger, E., Distl, O.,
Koninklijk Warmbloed Paardenstamboek Ward, T.L., Valberg, S.J., Lear, T.L., Guérin, G., 2007a. Mapping quantitative trait loci for
Nederland, Zeist, p. 47. Milenkovic, D., Swinburne, J.E., et al., osteochondrosis in fetlock and hock joints
Van Grevenhof, E. M., Ducro, B. J., van 2003. Genetic mapping of GBE1 and its and palmar/plantar osseous fragments in
Weeren, P. R.,van Tartwijk, J. M. F. M., van association with glycogen storage disease IV fetlock joints of south German Coldblood
der Belt, A. J., Bijma, P., 2009a. Prevalence in American Quarter Horses. Cytogenet. horses. Anim. Genet. 38, 350–357.
of various radiographic manifestations of Genome Res. 102, 201–206. Yan, Z., 2009. Exercise, PGC-1alpha, and
osteochondrosis and their correlations Wassen, M.H., Lammens, J., Tekoppele, J.M., metabolic adaptation in skeletal muscle.
between and within joints in Dutch Sakkers, R.J., Liu, Z., Verbout, A.J., et al., Appl. Physiol. Nutr. Metab. 34 (3),
Warmblood horses (KWPN). Equine Vet. J. 2000. Collagen structure regulates fibril 424–427.
41, 11–16. mineralization in osteogenesis as revealed Yang, D.E., Zhang, C.L., Zhang, D.S., Jin, D.M.,
Van Grevenhof, E. M., Schurink, A., Ducro, B. by cross-link patterns in calcifying callus. Weng, M.L., Chen, S.J., et al., 2004.
J., van Weeren, P. R., van Tartwijk, J. M. F. J. Bone Miner Res. 15 (9), 1776–1785. Genetic analysis and molecular mapping
M., et al., 2009b. Genetic variables of Wende, A.R., Huss, J.M., Schaeffer, P.J., of maize (Zea mays L.) stalk rot resistant
various manifestations of osteochondrosis Giguère, V., Kelly, D.P., 2005. PGC-1alpha gene Rfg1. Theor. Appl. Genet. 108 (4),
and their correlations between and within coactivates PDK4 gene expression via the 706–711.
joints in Dutch warmblood horses. orphan nuclear receptor ERRalpha: a Yang, F., Fu, B., O’Brien, P.C., Robinson, T.J.,
J. Anim. Sci. 87:1906–1912. mechanism for transcriptional control of Ryder, O.A., Ferguson-Smith, M.A., 2003.
Van Weeren, P.R., Barneveld, A., 1999. The muscle glucose metabolism. Mol. Cell Biol. Karyotypic relationships of horses and
effect of exercise on the distribution and 25 (24), 10684–10694. zebras: results of cross-species chromosome
manifestation of osteochondrotic lesions in White, K.E., Cabral, J.M., Davis, S.I., Fishburn, painting. Cytogenet. Genome Res. 102
the Warmblood foal. Equine Vet. J. 31 T., Evans, W.E., Ichikawa, S., et al., 2005. (1–4), 235–243.
(Suppl.), 16–25. Mutations that cause osteoglophonic Youatt, W., 1836. Animal pathology. The
Van Weeren, P.R., Sloet van Oldruitenborgh- dysplasia define novel roles for FGFR1 in Veterinarian 9, 362–368.
Oosterbaan, M.M., Barneveld, A., 1999. bone elongation. Am. J. Hum. Genet. 76 Zhou, J., Spier, S.J., Beech, J., Hoffman, E.P.,
The influence of birth weight, rate of (2), 361–367. 1994. Pathophysiology of sodium
weight gain and final achieved height and White, S.D., Affolter, V.K., Bannasch, D.L., channelopathies: correlation of normal/
sex on the development of osteochondrosis Schultheiss, P.C., Hamar, D.W., Chapman, mutant mRNA ratios with clinical
in a population of genetically predisposed P.L., et al., 2004. Hereditary equine phenotype in dominantly inherited
warmblood foals. Equine Vet. J. 31 regional dermal asthenia (“hyperelastosis periodic paralysis. Hum. Mol. Genet. 3,
(Suppl.), 26–30. cutis”) in 50 horses: clinical, histological, 1599–1603.

265
This page intentionally left blank
 CHAPTER 13 

The response of musculoskeletal

tissues to exercise
Helen L. Birch, Pieter A.J. Brama, Elwyn C. Firth, Allen E. Goodship,
José Luis L. Rivero, P. René van Weeren

1996). Unlike most other species, horses have always been used to
General introduction increase the mobility and transport capacity of mankind, with more
tangible products such as meat, milk and leather figuring as
There is a constant interaction between any biological entity and its by-products only. Although the role of the horse in society has
environment, which may consist of other biological systems (such changed dramatically in the past 50–60 years in comparison to the
as individuals from the same species, predators or invading micro- previous five millennia, it is still its athletic capacity that gives the
organisms), or physical influences such as weather conditions. In horse its role in present-day society.
fact, the (in)ability to respond actively to these environmental chal- On a much shorter timescale, functional adaptation towards
lenges can be seen as a criterion for the discrimination between life optimal locomotion capacity is also of crucial importance for the
and death. The environmental influences can be best accommo- individual animal. The demands on the equine musculoskeletal
dated when the individual is optimally suited to cope with these system of many modern performance horses seem likely to be sub-
challenges. In other words, when optimal functional adaptation stantially more than those borne previously by animals principally
occurs. This is of particular interest when the environmental stimuli used in transport, agriculture and the military, and are most prob-
are repetitive in nature, such as those generated by the locomotor ably much heavier than under most circumstances in the wild. The
system. This concept of functional adaptation applies to widely heavy challenges placed on the musculoskeletal system by present-
different timescales. On an evolutionary scale the capacity to adapt day equestrian activities, together with the inevitable loss in robust-
functionally is decisive for the survival of a species; for a given ness and sturdiness that goes with the process of domestication,
individual the ability to adapt throughout life determines whether may well lie at the heart of the high incidence of musculoskeletal
or not vitality is retained and for how long. injury and elevated wastage for orthopedic reasons in the equine
The horse has evolved from a forest-dwelling browser called species (Rossdale et al., 1985; Williams et al., 2001).
Hyracotherium or Eohippus, which had four hoofed toes in the Any musculoskeletal injury can be seen as a failure of the tissues
front feet and three in the hind feet and was not more than 8–9 to respond to the (mostly biomechanical) challenges placed upon
inches (approximately 2 hands) at the withers, to the contemporary them. Such failure may be the consequence of a high-impact single
large ungulate measuring up to 18 hands. During the evolutionary traumatic event, but more often follows the accumulation of repeti-
process, which started in the Eocene and took some 60 million tive micro-trauma. In the latter, training may be helpful to enhance
years, the horse changed from a typical forest inhabitant, which the resistance of the musculoskeletal tissues to biomechanical
relied on hiding rather than flight for survival into an animal that loading. In fact, training is the artificial and purposeful enhance-
was optimally adapted for living on the open steppe (Simpson, ment of the degree of functional adaptation of tissues. Classically,
1951). Through this process the locomotor system underwent pro- there has been much interest in the effect of training on cardiovas-
found change, as it became the most important factor in escaping cular performance and on muscle strength, but other tissues respond
from predators. The morphological changes that have imparted to training too, though it should be clear that the extent of response
advantage for high-speed locomotion include the development of of different tissues to the same training protocol may differ vastly.
long and slender limbs with the muscles located proximally and Another complicating factor is the age of the individual. Tradition-
close to the center of mass, reduced degrees of freedom of joint ally, research has been done in young horses, mostly aged 2–4 years,
motion in the distal limbs, which grossly limits motion to the sagit- because of the early onset of the athletic career in the racing breeds
tal plane, and the use of collagenous components of muscle and and resulting commercial pressure. These horses cannot be classi-
tendon to reduce energy requirements in posture and locomotion. fied as juvenile, nor have they yet reached full skeletal maturity
A very important athletic asset of the present-day horse is the storage (Gabel et al., 1977; Strand et al., 2007). There has been little inter-
of kinetic energy in the form of elastic energy in the tendinous est in the effects of exercise on older horses, but some large studies
structures of (especially) the limb flexors. This spring-like mecha- have been conducted in the past decade on the effects of exercise
nism prevents energy loss and limits the role of the flexor muscles in very young animals, e.g. in newborn foals. The latter research has
to dampening of these springs rather than using them to generate made it clear that age indeed does matter. Whereas bone and muscle
propulsion, thus greatly reducing the muscle volume needed are known to respond to exercise throughout life, articular cartilage,
(Wilson et al., 2001). tendons and ligaments are tissues that are much less, if at all,
The potential of the locomotor system led man to begin domes- responsive in mature individuals. However, there is mounting evi-
ticating the horse in approximately 3000 BC (Dunlop & Williams, dence from recent research that these latter three tissues, which rank

267
 13 The response of musculoskeletal tissues to exercise

among the first with respect to injury prevalence and are notorious Articular surface
for their poor healing tendency, might be much more responsive to
exercise in juvenile animals. This might open up new avenues for Proteoglycan
better adapting the equine musculoskeletal system to athletic chal- aggregates
lenge through programs of conditioning exercise at young age. Collagen fibrils
However, such programs could be developed only if the exact effects
Tidemark
of exercise on the diverse musculoskeletal tissues are known, if the
exercise regimen to effect adaptation is known, and the appropriate Calcified layer
time window in which to apply it. In the following chapter the Subchondral
major constituting tissues of the musculoskeletal system: bone, boneplate
articular cartilage, tendons and muscle, are discussed with respect Fig 13.1  Schematic drawing of articular cartilage showing the arcade-like
to their responsiveness to exercise. Discrimination is made between architecture of the collagen fibrils as first described by Benninghoff (1925).
the effect of exercise in mature individuals and in the growing and Typically, the collagen fibrils emerge from the calcified layer perpendicular
still developing juvenile animal. to the tidemark and start to arch in the intermediate zone, to become
aligned with the articular surface in the superficial zone.

Articular cartilage

Introduction the collagen network (lysylpyridinoline or LP-cross-links and

hydroxylysylpyridinoline or HP-cross-links respectively) as the
Joint disorders have a major impact on equine athletic performance, most common covalent cross-links. Cross-links have a major influ-
as they rank either first or second as causes for wastage in perfor- ence on structural and consequently biomechanical characteristics
mance horses, depending on the equestrian discipline involved of the collagen network (Eyre & Wu, 1995) and the formation of
(Rossdale et al., 1985; Todhunter & Lust, 1990; Williams et al., cross-links is mainly an irreversible process (Ray et al., 1996). A
2001). A joint is a complex entity that is made up of subchondral special category of cross-links is formed through the process of
bone, articular cartilage, synovial membrane, and fibrous joint non-enzymatic glycation. As collagen molecules have an excep-
capsule and ligaments. Sometimes other structures such as a physis, tionally long lifetime once incorporated into the ECM of cartilage
tendon insertions, menisci or ligaments are present intra-articularly (see later), they are susceptible to the accumulation of advanced
as well. Articular cartilage plays a pivotal role in the physiology of non-enzymatic glycation end products (AGEs) via the so-called
the joint because of its triple function: transmission of the forces Maillard reaction (Monnier, 1989). The process results in increased
generated by locomotion, acting as a shock absorber and providing cross-linking, such as pentosidine formation from lysine, sugar
a smooth gliding surface to enable supple joint motion. Articular and arginine moieties. Pentosidine is one of the few Maillard
cartilage has a crucial role in joint pathology, as in most joint dis- cross-links of which the structure has been elucidated, and is used
orders it is the principally affected joint tissue and has, at least in as a sensitive marker for the entire process (Vlassara et al., 1995).
mature individuals, a very limited capacity for repair. In the follow- The accumulation of AGEs depends on the turnover rate of a
ing section first a brief synopsis will be given of some essential protein or tissue and is thus a measure for the metabolic rate of
principles of articular cartilage biology, followed by a paragraph on that structure (Verzijl et al., 2000).
the developmental aspects of articular cartilage, which are essential Interspersed between the collagen fibrils and connected to them
to appreciate the effects of exercise on young, growing individuals either directly or via hyaluronan or hyaluronic acid molecules are
and the possible implications thereof. The section ends with a proteoglycan aggregates, of which aggrecan is the most important.
paragraph on the effects of exercise (or lack thereof) in more mature Aggrecan consists of a core protein with large numbers of sulphated
individuals. glycosaminoglycans (chondroitin sulphate and keratan sulphate) as
side chains. Because of their sulphate groups, these side chains are
negatively charged and attract water. The combination of the swell-
Basic principles of articular cartilage biology ing pressure of the sulphated glycosaminoglycans and the confine-
Articular cartilage plays an important role in both the transmis- ment due to the stiffness of the collagen network creates an intrinsic
sion and attenuation of forces generated by locomotion, which pressure within the ECM of articular cartilage that gives it the
leads to seemingly contradicting requirements of both elasticity unique combination of resilience under compression and great
and great strength. Nature provides a solution to this problem by resistance against tensional and shear forces (Todhunter, 1996).
conceiving an extracellular matrix (ECM) that consists of two prin- Different areas of articular cartilage are subject to different types
cipal components: collagen fibrils and proteoglycan aggregates. of loading, such as low-level constant loading, intermittent loading,
The collagen fibrils are made up of collagen type II molecules (3 and very high impact loading (Fig. 13.2). These variable loading
α1-chains in triple helix configuration) and form a three- conditions can only adequately be met, without being damaged in
dimensional arcade network, as was first described by Benninghoff the long term, by cartilage possessing different mechanical proper-
(1925). The fibrils start from the layer of calcified cartilage just ties (Herzog & Federico, 2006; Palmer & Bertone, 1996) through
above the subchondral bone, then course perpendicular to this appropriately different biochemical and ultrastructural characteris-
layer towards the joint surface and bend back, running tangen- tics over the cartilage surface (Brama et al., 2000b,c; Brommer et al.,
tially to the joint surface and directly under it, to close the arch by 2005; Murray et al., 2001a; Palmer et al., 1995a,b) (Fig. 13.3). Two
returning perpendicularly again to the calcified cartilage layer (Fig. sites of the proximal first phalangeal bone, known to be subjected
13.1). Cross-links that interconnect α1-chains within the same col- to strongly differing loading conditions and exhibiting a difference
lagen molecule and interconnect collagen molecules and different in predisposition to cartilage disease, have received considerable
collagen fibrils provide structural coherence and extra strength. attention with respect to topographical heterogeneity in cartilage
Cross-link formation is one of the so-called post-translational composition and loading characteristics (Brama et al., 1999b, 2001,
modifications of collagen and is the last chemical modification 2002) (Fig. 13.4).
that occurs during the formation of the primary collagen structure. Cartilage also displays zonal heterogeneity along its depth. The
There are various types of cross-links with the pyridinoline superficial layer contains flattened chondrocytes, has densely packed
cross-links that form between lysyl and hydroxylysyl residues in type II collagen fibrils oriented parallel to the surface, a small

268
 Articular cartilage

A D

B E

Fig 13.2  Topographical distribution patterns of pressures (MPa) on the

proximal articular surface of the proximal phalanx as measured under ex vivo
conditions. Loads correspond with those measured in vivo using a force
plate technique (1800 N: stance; 3600 N: walk; 5400 N: trot; 10 500 N: canter;
12 000 N: jumping an 80 cm fence). Dors., dorsal; lat., lateral; med., medial;
palm, palmar; N, Newton; MPa, megapascal.
C
Reproduced from Brama et al, 2001. Equine Veterinary Journal 33, 26–32., with permission.

amount of proteoglycans (PG) and high water content (Aydelotte percentage of water of any zone, and chondrocytes arranged in
et al., 1988; Aydelotte & Kuettner, 1988). The middle (transitional) columns perpendicular to the subchondral bone (Todhunter,
zone has lower water content, a higher concentration of PG, and a 1996).
lower concentration of collagen fibrils that are less organized, and Articular cartilage is, like any living tissue, not a static entity but
rounded chondrocytes dispersed irregularly in the extracellular undergoes constant remodeling through anabolic and catabolic
matrix. The deep zone has collagen fibrils oriented perpendicular processes. The chondrocyte, the only cell type found in normal
to the articular surface, the highest concentration of PG, lowest articular cartilage, produces all ECM components. Collagen

269
 13 The response of musculoskeletal tissues to exercise

Site 1 Site 1

A Proximal phalanx

Normal weight-bearing Hyperextension due

to peak loading

Site 1
Site 2
Site 1

Site 2

B Fig 13.4  Differently loaded sites at the proximal articular surface of the
proximal phalanx. Site 1 is located at the medial dorsal margin of the joint
surface. This site is not loaded when standing still or in a slowly moving
animal, but it is subject to high intermittent peak loading in overextended
joints at high speeds and during jumping. Site 2 is located in the central
fovea of the articular surface and is continually loaded when the limb is
weight bearing, but the site experiences lower absolute forces than site 1.

and proteoglycans, have very different turnover rates. Normal turn-

over time of proteoglycans is variable and ranges in mature indi-
viduals from about 300 days for overall turnover in canine articular
cartilage to around 20 years for the long-lived globular hyaluronic
acid binding domain of aggrecan in human cartilage (Maroudas,
1980; Maroudas et al., 1998; Verzijl et al., 2001). However, collagen
turnover times have been estimated to be up to 350 years in mature
C human cartilage (Maroudas et al., 1992). Comparable data for the
horse do not exist, but are assumed to be in the same order of
Fig 13.3  Patterns of distribution of glycosaminoglycan content (GAG), magnitude. It is this very low turnover rate and thus extremely
collagen content, and hydroxylysylpyridinoline (HP) cross-linking over the limited repair capacity that lies at the basis of the famous observa-
proximal articular surface of the proximal phalanx and the distal articular tion by William Hunter (1743) that ‘ulcerated cartilage when
surface of the third metacarpal bone. Dors., dorsal; GAG, glycosaminoglycan;
destroyed, will never repair’, and that makes OA in both humans
HP, hydroxylysylpyridinoline; lat., lateral; MC, distal third metacarpal bone;
and horses into such a debilitating disease with a very poor long-
med., medial; palm., palmar; P1, proximal first phalanx.
term prognosis.
Reproduced from Brama et al, 2000. Equine Veterinary Journal 32, 19–26, with permission.

Developmental aspects of articular cartilage

molecules are produced intracellularly as pro-collagen and are Theoretically, there are two possible origins of topographical het-
extracellularly modified by enzymatic cleavage of the C- and erogeneity of articular cartilage biochemistry: genetic predetermina-
N-terminal propeptides before they are integrated in the collagen tion or formation during early life when cartilage metabolism is still
network. Degradation of ECM is effectuated by various proteinases, high during active growth. The first indication that the latter could
of which the matrix metalloproteinases (MMPs) and aggrecanases be true came from a study, which showed that proteoglycan content
(members of the disintegrin and metalloproteinase with thrombo- was identical all over the joint in neonatal lambs, unlike the situa-
spondin motifs (ADAMTS) gene family) are the most important. tion in adult sheep where there is topographical heterogeneity
However, the principal constituents of the cartilage ECM, collagens (Little & Ghosh, 1997). The same was true for the horse, in which

270
 Articular cartilage

proteoglycan content and collagen, including post-translational 43 1-week-old foals

modifications of collagen such as cross-links, were homogeneously
distributed, at least when looking at average values of full thickness
samples, across the joint of the newborn in what has been called a 14 1 4 15
‘blank joint’ (Brama et al., 2002). Most of the topographical het-
erogeneity appeared before the age of 5 months, but HP cross-links 24 h box-rest Box-rest with exercise 24 h pasture
(the most abundant pyridinoline cross-link in articular cartilage)
did not reach their mature values until after 12 months (Brama
et al., 1999a, 2000b, 2002). Recently, this maturation process of Weaning at 5 months,
articular cartilage in early life was confirmed by an investigation euthanasia of 3x8 foals and
using equine-specific cDNA micro-arrays. The expression of genes post-mortem analysis of tissues
encoding matrix proteins and matrix modifying enzymes was 1 9
clearly different between neonatal and mature individuals with
the former expressing more collagens, matrix-modifying enzymes Identical exercise protocol until
and provisional matrix non-collagenous proteins and the latter 11 months, then euthanasia and
showing a transition from growth to homeostasis and tissue post-mortem analysis of tissues
function related to coping with shear and compressive forces
(Mienaltowski et al., 2008). Fig 13.5  Schematic diagram of the experimental set-up of the so-called
The changes in biochemical composition of the extracellular EXOC study. Forty-three Warmblood foals were divided into three different
matrix of articular cartilage during early life are accompanied by exercise groups at age 1 week. Exercise regimens consisted of 24 h/day box
structural changes. The arcade structure of the collagen network as rest (n = 14), a similar sedentary protocol but with additional high-intensity
described by Benninghoff (1925) is typical for mature cartilage, but sprinting exercise for 30 min/day (n = 14), and free pasture exercise (n = 15).
not yet present in the newborn. Polarized light microscopy has After weaning at age 5 months, 8 foals from each group were euthanized
shown that at birth collagen fibril orientation is predominantly and their tissues harvested. The remaining 19 foals were housed together
and given a moderate exercise regimen. These animals were euthanized at
parallel to the joint surface and subchondral bone and then gradu-
age 11 months for tissue harvest.
ally changes to the more classic Benninghoff arcade structure
(Rieppo et al., 2008). Preliminary data from the horse suggest that
here too the mature configuration is basically formed during the
first 5–6 months of life (van Turnhout et al., 2008).

1.25
The effect of exercise on articular cartilage
Collagen ratio site I/II

during growth and development

1.0
If the topographical heterogeneity of the extracellular matrix of
articular cartilage enables the joint to meet the biomechanical chal- Training
lenges to which it is subjected and if this topographical heterogene- Pasture
ity is formed from scratch in the young individual in the so-called 0.75 Exercise
Boxrest
process of functional adaptation (Brama et al., 2002), then biome- Start regime End
chanical loading is the most likely driving force of this adaptive
process. The first evidence that this is true came from a study that Neonatal 5 mo 11 mo
was originally designed to study the influence of exercise on the n=12 n=8 n=7
developmental orthopaedic disease osteochondrosis (van Weeren &
Fig 13.6.  The influence of different exercise levels during the first 5 months
Barneveld, 1999a). In that study, named EXOC, a group of 43 after birth on the collagen ratio of site 1/site 2 (see Fig. 13.4). Values are
Warmblood foals was divided into three groups from the age of given as mean ± standard error of the mean. Pasture: pasture exercise;
1 week. One group (boxed; n = 14) was kept in box stalls for training: box rested/sprinted animals; box rest: box rested animals (see also
24 h/day; the second group (boxed/sprinted; n = 14) was kept in Fig. 13.5). n, number of animals; *, p < 0.01. Note that both groups that were
similar box stalls, but given exercise in the form of a number of exercised develop a site 1:site 2 ratio that is significantly different from one
short sprints during a restricted period, i.e. in this group high- in the first 5 months of life, indicates that they developed topographical
intensity exercise was alternated with a very sedentary lifestyle; the heterogeneity. The box rested animals do not develop this topographical
third group (pastured; n = 15) had free pasture exercise for 24 h/ heterogeneity and fail to compensate for this in the following 6 months
day. The exercise regimens were maintained until weaning at 5 when undergoing a moderate exercise regimen.
months of age, at which age 8 foals from each group were eutha- Reproduced from Brama et al., 2002. Equine Veterinary Journal 34, 265–269., with
nized and their tissues harvested for biochemical and other analy- permission.
ses. The remaining 19 foals were then kept together in a single group
and allowed a moderate exercise regimen. These animals were euth-
anized at age 11 months and had their tissues analyzed to see if any
difference provoked by the different exercise regimens at 5 months mineral density (Cornelissen et al., 1999; Firth et al., 1999b)),
would remain or was reversible once a similar exercise regimen was levels of collagen or the collagen-related hydroxylysine levels
given to all foals (Fig. 13.5). remained abnormal and failed to develop the topographical hetero-
The study showed that topographical heterogeneity, expressed as geneity that is characteristic of mature individuals (Fig. 13.6)
a significant difference from zero of the ratio of two very differently (Brama et al., 2002). Therefore, whereas this study unequivocally
loaded sites within the joint, developed in the pastured group and demonstrated the importance of biomechanical loading for the
in the boxed/sprinted group alike, but failed to develop in the development of the topographical heterogeneity of the extracellular
boxed foals. The most interesting observation, however, was that, matrix, it showed also that, at least with respect to the collagen
where proteoglycans became normal after the common training component, there is a limited window in time when this process of
program from 5 to 11 months (as did exercise-induced differences functional adaptation can take place. This stresses the importance
in other tissues from the same experimental animals such as bone of sufficient exercise in the early juvenile phase.

271
 13 The response of musculoskeletal tissues to exercise

Another interesting observation was made in the boxed/sprinted metacarpal cartilage (Kawcak et al., 2010). However, the exercised
animals that had been subjected to a combination of short bouts horses had more viable chondrocytes in the more heavily loaded
of high-intensity exercise with a sedentary lifestyle. The degree of sites than in the less loaded sites of the same metacarpal cartilage
development of topographical heterogeneity in this group was com- (Dykgraaf et al., 2008). Also, a detailed study of contiguous approx-
parable to the pastured animals, which were regarded as controls. imately 100-µm thick slices taken from the proximal articular
However, when culture experiments were carried out with chondro- surface of the proximal phalanx down to the tidemark at differently
cytes harvested at age 11 months from animals from the former loaded sites (see above) showed not only obvious site differences,
exercise groups, it appeared that chondrocytes from the former box/ but also exercise-related changes. There was no exercise effect on
sprinted group could not be stimulated to increase metabolic activ- proteoglycan content, indicating that the exercise level had not been
ity, in contrast to samples from both other groups. This was inter- strenuous and confirming the work by Nugent et al. (2004) and
preted as a deleterious long-term effect of the combination of Kawcak et al. (2010); but at 18 months old the normal physiological
sedentarism with bouts of high intensity exercise, possibly repre- increase in collagen at the site located at the joint margin was less
senting a kind of cellular exhaustion that might be due to over- in CONDEX animals (Brama et al., 2009a). This was interpreted as
stimulation during the first 5 months of life (van den Hoogen et al., a precocious cessation of collagen remodeling at this site due to
1999). Similar phenomena were observed in other tissues from advancement in time of the normal maturation process. The bio-
these animals (Barneveld & van Weeren, 1999). Also, forced exercise chemical analysis of similar slices from the same site in the meta-
(box/sprinted) had a negative effect on collagen turnover, as mea- tarsophalangeal joint pointed in the same direction: in the CONDEX
sured by serum markers CPII and Ctx1, when compared to pastured animals, hydroxylysine, HP cross-links and pentosidine cross-links
foals (Billinghurst et al., 2003). were all higher, all indicative of advancement of the normal process
Another large-scale exercise study concerning foals was conducted of functional adaptation (van Weeren et al., 2008). The increased
by the Global Equine Research Alliance (GERA, a consortium of the pentosidine levels in the CONDEX animals are of particular inter-
equine orthopedic research groups from Massey University, New est, because they indicate a lower metabolic activity in this group,
Zealand, Colorado State University, USA, Royal Veterinary College, confirming more rapid progression of the physiological age-related
UK and Utrecht University, the Netherlands). This study became decrease in matrix turnover and thus indicative of an advanced
known under the acronym GEXA. In the GEXA study, 33 Thorough- degree of maturation compared to the PASTEX animals (Fig. 13.8).
bred foals were raised from 0–18 months under different exercise Other evidence for the difference in maturation rate came from
conditions. One group (PASTEX, n = 15) had free pasture exercise ultrastructural and biomechanical studies. Polarized light micros-
all year round, in agreement with common New Zealand practice; copy techniques were used to investigate the spatial arrangement of
the second group (CONDEX, n = 18) was raised under similar the fibrils of the collagen network throughout the depth of the
conditions, but subjected to an additional exercise program as well. cartilage, measured as parallelism index (PI, a measure of the degree
The exercise program consisted of cantering and sprinting on a track to which the collagen fibrils are aligned with each other) and ori-
which increased overall workload by a moderate, but significant, entation index (OI, a measure of the average angle of collagen fibrils
30% (Rogers et al., 2008a). At age 18 months, 12 animals (six from with respect to the articular surface). Parallelism index was higher
each group) were sacrificed and their tissues harvested for detailed in CONDEX animals, again indicating advanced maturation (Brama
analyses. The remaining animals were broken in and trained and et al., 2009b) (Fig. 13.9). An interesting observation was that ori-
raced as 2- and 3-year-olds (Rogers et al., 2008b) (Fig. 13.7). entation of collagen fibrils in the deep zone of the cartilage was less
The articular cartilage of the metacarpophalangeal and metatar- perpendicular than expected, which is possibly an adaptation of the
sophalangeal joints from the GEXA animals harvested at 18 months direction of collagen fibers to high shear forces in the equine meta-
was meticulously researched using a variety of techniques. Bio- carpophalangeal joint, which is a heavily loaded joint with an
chemical and biomechanical analyses of full-thickness samples exceptionally high range of motion where an oblique insertion of
from the distal metacarpal bone showed site-related differences, but
no exercise effect (Nugent et al., 2004). The same applied to the
metabolic rate of chondrocytes (measured by 35S-uptake) from third

2
Pastex
Months

33 foals Condex
Pentosidine (mmol/triple helix)

18 15 Pasture 18 Pasture
without training without training 1

20 racing
35
6† 6†

0
Fig 13.7  Schematic diagram of the experimental set-up of the so-called 1 2 3 4 5 6 7 8 9
GEXA study. Thirty-three Thoroughbred foals were divided into 2 different Slice from surface
exercise groups from average age 3 weeks until age 18 months. Exercise
regimens consisted of 24 h/day pasture exercise (PASTEX, n = 15), or pasture Fig 13.8  Mean values for pentosidine cross-link content (expressed as
exercise with additional track training that increased workload by 30% millimole per mole triple helix) for the slices produced at site 1 and 2 for
(CONDEX, n = 18; for details of training protocol see Rogers et al. 2008a). At both PASTEX and CONDEX animals. The most superficial layer (slice 1) is
age 18 months, six animals from each group were euthanized and their located at the joint surface; the deepest layer (slice 9) is adjacent to the
tissues harvested. One PASTEX animal was lost due to an accident and the calcified cartilage.
remaining 20 animals were trained for racing as 2- and 3-year-olds. Cross Reproduced from van Weeren et al, Equine Veterinary Journal 40, 128–135, 2008, with
symbol indicates euthanasia. permission.

272
 Articular cartilage

90% discussed in the preceding paragraph. In the so-called short-term

80% Bristol study 2-year-old Thoroughbreds were subjected to a 19-week
exercise regimen that consisted of three times/week galloping on a
Degree of parallelism (%)

70%
treadmill, trotting in a mechanical horse walker three times weekly
60%
and 40 min walking exercise six times weekly. The control group
50% was only walked for 40 min six times per week during the same
40% Trained period (Murray et al., 1999b). In the MUGES study (Massey Uni-
Pasture versity Grass Exercise Study), a group of 2-year-old Thoroughbreds
30%
20% was trained for 13 weeks on grass and sand tracks, after which their
musculoskeletal tissues were analyzed (Firth et al., 2004a).
10%
Cartilage taken from the proximal articular surface of the radial,
0% intermediate and 3rd carpal bone from exercised animals that had
0 90 180 270 360 451 541 631 721 811
participated in the Bristol study had a thicker calcified, but not
Thickness (micrometers)
hyaline cartilage layer than from non-exercised animals (Murray
Fig 13.9  The average degree of collagen parallelism (PI; mean ± SD) over et al., 1999a). The cartilage of the strenuously trained animals was
the entire depth of the cartilage, measured with quantitative polarized light biomechanically less stiff, showed more fibrillation and more chon-
microscopy at points 9 µm apart from each other from the cartilage surface drocyte clusters than the gentler trained animals, suggesting that
to the osteochondral junction. A PI of 0% indicates random course of fibrils strenuous exercise may lead to deterioration of cartilage (Murray
and 100% complete parallelism. Pastured animals (PASTEX) compared to et al., 1999b). Also fibronectin (Murray et al., 2000) and cartilage
exercised animals (CONDEX). oligomeric matrix protein (COMP) distribution (Murray et al.,
2001b) appeared to be influenced by exercise. When cartilage from
animals of the 19-week Bristol study was cultured to measure chon-
drocyte metabolic activity, proteoglycan synthesis rates were higher
the collagen fibers in the calcified cartilage layer may yield better in the trained animals than in the controls, confirming the overall
resistance to repeated shear forces than would a perpendicular con- effect of strenuous exercise on the proteoglycan component of
figuration. Based on this observation it was hypothesized that the articular cartilage ECM (Bird et al., 2000). There were no overall
Benninghoff arcade model may be more flexible than commonly effects on collagen levels, but sites predisposed to clinical signs
thought and that local adaptation includes a predominantly non- contained significantly less collagen in horses undergoing the exer-
perpendicular direction of the collagen fibrils in the deep zone. cise program; also, strenuously exercised animals had higher glycos-
Biomechanical analysis of samples from different sites showed that aminoglycan content than horses from the more gently trained
in CONDEX animals site-related differences known to develop with group, which was most marked in the cartilage from the dorsal
age (Brommer et al., 2005) were less marked than in PASTEX radial and dorsal intermediate carpal surfaces, sites known to be
animals with respect to Young’s modulus (ratio of axial stress and heavily loaded during exercise (Murray et al., 2001a). Mechanisti-
strain in unconfined compression) and dynamic modulus (propor- cally, the hypothesis that the degrading enzymes Cathepsin B and
tional to the elastic and viscous energy dissipated in the loading D might be involved in the exercise-induced alterations could not
process and essentially a measure of dynamic stress and strain be confirmed, but it was demonstrated that the two enzymes were
(Palmer & Bertone, 1996)). This may have been caused by earlier regulated differently by mechanical loading (Bowe et al. 2007). In
down regulation of collagen metabolism in trained animals, as contrast to the Bristol study, the MUGES study found a significant
signaled earlier (van Weeren et al., unpublished results). The increase in thickness of (hyaline) cartilage after 13 weeks of training
quicker maturation of the collagen matrix and associated other (Firth & Rogers, 2005), compared to pasture-raised controls, in
pattern of topographical heterogeneity need not necessarily be del- particular in sites within the intercarpal joint. At these sites there
eterious, as the same study showed that the cartilage degeneration was no patho-anatomical damage, leading to the conclusion that
index (CDI) as described by Brommer et al. (2003) was significantly the response of the tissue was adaptive and not degenerative in
less in the CONDEX animals, indicating better surface integrity in nature. However, there was a difference between joints, because in
this group (van Weeren et al., unpublished results). the metacarpophalangeal (MCP) joints of the same animals wear
There is also evidence from other species that the exercise regimen lines and fibrillation were present and water content was increased
at early age may affect cartilage properties and that this effect need while HP cross-links had fallen compared to the controls, which
not always be reversible. Kiviranta et al. (1987) showed in young was interpreted as signs of loosening of the collagen network due
Beagles in which one of the hind limbs was immobilized for 11 to micro-damage (Brama et al., 2000a). The metacarpophalangeal
weeks that GAG content may be reduced by almost 50% in the joint is known to be the equine joint most susceptible to damage
non-weight-bearing limb, whereas cartilage thickness and GAG- (Pool, 1996) and predilection for patho-anatomical damage in the
content was increased in the contralateral limb. When the study was MCP joint has been shown in both treadmill-exercised horses
repeated, but now followed by a 15-week rehabilitation program, (Kawcak et al., 2000) and in wild horses (Cantley et al., 1999). Also
the normal situation was not fully restored; the authors concluded the study by Little et al. (1997) showed that there is a level at which
that immobilization of skeletally immature joints may affect carti- exercise may become deleterious. In an ex vivo study in which mate-
lage development such that recovery is very slow or alterations are rial from the dorsal radial facet of the third carpal bone, an area
permanent (Kiviranta et al., 1994). well-known to be subjected to high contact stresses in galloping
horses, was cultured that came from horses having undergone a
strenuous exercise regimen, a significant reduction in aggrecan syn-
The effect of exercise on articular cartilage in thesis and a concomitant increase in decorin synthesis could be
determined compared to less vigorously trained controls. The sug-
young adult animals gestion was made that this alteration in articular cartilage metabo-
No large studies have been conducted on the effect of exercise on lism could be a predisposing factor for cartilage degeneration and
cartilage from completely mature or elderly horses, but work has OA at a later stage.
been performed on 2-year-old Thoroughbreds. These animals can An effect of exercise on the biomechanical properties of cartilage
be classified as ‘young adult’, although they cannot be considered was shown in a relatively short (6 weeks) treadmill study (Palmer
as fully skeletally mature yet because they are still growing (Green, et al., 1995a), in which the permeability constant increased at all
1976), although at a much lower rate than the foals that were sites investigated, but the Poisson’s ratio increased only at specific

273
 13 The response of musculoskeletal tissues to exercise

sites, stressing the topographical variation in response to biome- mean for the horse that it might be better to start athletic training
chanical stimuli, which is most probably related to local differences rather than later, as has been suggested (Smith et al., 1999).
in joint architecture and geometry. However, long-term epidemiological studies are necessary to estab-
The overall metabolism-enhancing effect of exercise on articular lish whether differences in the development of the biochemistry
cartilage has been studied indirectly in several studies that focused and ultrastructure of articular cartilage that are now known to be
on a variety of potential biomarkers in either serum or synovial modulated by early exercise have significant implications for the
fluid. Plasma GAG levels, and serum keratan sulphate and COMP orthopedic health and/or performance of the equine athlete during
all increased after moderate to strenuous exercise (Calatroni et al., its career.
2008; Okumura et al., 2002; Helal et al., 2007). Brown et al. (2007)
looked in more detail into changes in glycosaminoglycan metabo-
lism due to exercise and found higher chondroitin sulphate peak Bone
chain lengths, but shorter hyaluronic acid chain lengths in exercised
versus rested horses. Investigating both serum and synovial fluid, Introduction
Frisbie et al. (2008) confirmed the increase in proteoglycan markers
after exercise, and showed that the same was true for a number of Bone is the tissue that confers rigidity to the musculoskeletal system
collagen markers. However, moderate exercise did not affect MMP-1 and its intimate association with articular cartilage allows both
activity in synovial fluid (Brama et al., 2004). There is a reciprocal tissues to sustain the large forces generated by normal and athletic
interaction between synovial fluid and articular cartilage and the locomotor activity. Bone derives its resistance to deformation from
former is not just a reflection of the status of the latter. The effects the hydroxyapatite crystals that are laid down on a collagen matrix,
of joint loading on cartilage are, at least in part, mediated by altera- providing the tissue with great strength, but also reducing elasticity
tions in the synovial fluid, as cartilage explants cultured in post- substantially. The responsiveness of bone to biomechanical influ-
exercise synovial fluid showed enhanced GAG synthesis and ences has been recognized universally since the late 19th century
diminished release when compared to cultures using pre-exercise when Julius Wolff proposed his law, stating that bone will adapt
synovial fluid (van den Hoogen et al., 1998). both architecturally and with respect to composition to changes in
Also in other species there is evidence for an influence of exercise mechanical loading (Wolff, 1892). Failure of bone results in (micro)
on articular cartilage, at least in young individuals. In young Beagle fractures. The incidence of bone failure in equine orthopedics varies
dogs, moderate loading levels tended to increase GAG production, strongly with the discipline involved, but can be substantial, par-
while strenuous exercise led to GAG depletion in high-load areas ticularly in racing (Rossdale et al., 1985; Williams et al., 2001).
(Arokoski et al., 1993; Kiviranta et al., 1988, 1992, 1994; Säämä- Clinically, bone failure in the horse may vary from relatively mild,
men et al., 1994). In man it has been shown that children not yet economically important, micro-fractures and fissures as seen in
undergoing programs of vigorous activity had approximately 25% sore shins (Boston & Nunamaker, 2000), to catastrophic fractures
less cartilage than children that had been mildly active (Jones et al., of long bones (Parkin et al., 2006). In the following section an
2003). introduction on the morphology and physiology of bone is given
first, followed by a discussion of the influences of exercise on bone
in general and more specifically for the horse. Most of the earlier
work has been done in young adult animals; recently the effect of
Conclusion exercise on foals has become an area of interest.
There is no doubt that articular cartilage is responsive to biome-
chanical loading and that responsiveness is highest in the youngest
animals. In fact, there is now substantial evidence that the charac-
Morphology and physiology of bones
teristic topographical heterogeneity in composition and ultrastruc- The diaphysis of long bones has a marrow cavity, which allows a
ture of the extracellular matrix of mature cartilage is formed under better strength/weight ratio to resist bending and torsional forces,
the influence of loading, i.e. through physical exercise. It has been while positively affecting gait velocity by minimizing the weight of
suggested, therefore, that the collagen network of juvenile animals the distal limb. Most features of a long bone and cuboidal bone are
that are in a phase of rapid growth might respond to loading in a illustrated in Figure 13.10. There is a gradual transition from the
fashion similar to that in which trabecular bone is known to dense bone of the diaphyseal cortex to the metaphysis, in which the
respond according to Wolff’s law (van Weeren & Brama, 2003). The junction of the outer compact bone with the cancellous bone is
difference with bone is that, whereas bone will continue to respond indistinct. The epiphysis is contained within a dense shell of
to exercise throughout life, the collagen network can most probably compact bone, which in articulating regions is referred to as the
do so only in the young, growing animal before the collagen remod- subchondral bone (SCB), is regionally thickened at various sites,
eling rate falls to the extremely low level related to the very long and has a gradual transition into the bulk epiphyseal trabecular
collagen turnover times in mature animals mentioned earlier. bone (BETB). The orientation and architecture of the trabecular
Indeed, although some studies have shown a response of the pro- bone of the metaphysis and epiphysis is central to how load is
teoglycan component of the ECM of cartilage to exercise in (young) transferred from the diaphyseal cortex to the joint surface and
adult horses, no adaptive response of collagen has ever been dem- thence across the joint. The metaphyseal cortex is thin, and this is
onstrated in this age group. This makes the early juvenile period the most obvious close to the physis, and especially in the young pre-
only window of opportunity to manipulate the collagen network sumably because of the rapid growth rates in early life. The cross
through modification of loading, e.g. differences in exercise. Recent section of the whole bone is greatest at the level of the physis, and
studies show that such manipulation is possible and may irrevers- the contact area of the articulating surface of the bone with its
ibly delay the normal maturation process as the result of a decreased opposing counterpart in the close-packed position is also large. This
exercise level or lead to accelerated maturation resulting in a differ- maximization of the force-bearing surfaces allows articular and
ent topographical pattern in animals subjected to more exercise physeal cartilage and immature bone to sustain lower stress,
than free pasture exercise. There is thus reason to believe that exer- although the force transmitted across the joint and through the
cise at an early age may indeed have long-lasting effects on the physis and metaphysis equals that borne by the diaphysis, which is
collagen matrix of articular cartilage, lending credibility to the much more dense and much smaller in cross-section than elsewhere
hypothesis that variations in exercise regimens of juvenile persons in the bone.
or animals may influence the risk for the development of degenera- The extracellular matrix of bone is made up of a framework of
tive joint diseases later in life (Helminen et al., 2000). This may osteoid, consisting of about 90% collagen (type I), around and in

274
 Bone

General response of bone to exercise

It is possible to measure if bone responds to exercise in various
ways including biochemical, gene expression, biomarkers, morpho-
logical and compositional change, functional change, and results of
that change (such as athletic performance) the latter two of which
might not be specific to change in bone. The relationship between,
for instance, serum biomarker elevation or altered gene expression
on phenotypic effect may not be known, and thus careful whole
animal study is required before the significance of such changes can
be known. It is paradoxical that now it has become easier to
produce data from sophisticated laboratories, it appears to be more
difficult to obtain performance data to test the validity of the altera-
tions detected.
Most methods used to assess the response of bone to exercise are
based on quantifying the amount and disposition of mineral within
the bone. The mechanical properties of bone are closely related to
mineral in the matrix, and the modulus of elasticity is directly
related to volumetric bone mineral density (BMDv) of the bone
tissue (Currey 1988b; Ferretti et al 1996), at least within the physi-
ological range of BMDv. Collagen orientation and cross-linking also
affect variance in mechanical properties of bone (Riggs, 1993;
Oxlund, 1995; Viguet-Carrin et al., 2006), but the extent to which
collagen parameters affect bone strength, and when this is rate
limiting in the growing and training horse if injury is to be pre-
Fig 13.10  Radiograph of a 1.5-mm thick sagittal slice of the radius and vented, have not been established.
intermediate carpal of a 19-month-old pasture-raised Thoroughbred filly. Bone responds to exercise in one or more possible ways. These
The radial cortex blends gradually with the trabecular bone of the include increase in mineralized mass (Goodship et al., 1979; Woo
metaphysis, and the cortical shell of the epiphysis with the bulk epiphyseal et al., 1981; Gordon et al., 1989) , increase in bone material density
trabecular bone. At the joint surfaces, the subchondral bone is evident as a (Boyde & Jones, 1996), increase in bone mineral density of the
radiodense line of uniform thickness. whole bone (Cornelissen et al., 1999; Firth et al., 1999b, 2000),
change in trabecular bone density and orientation (Young et al.,
1991b; Firth et al., 1999a,c) alteration in remodeling, and increase
which hydroxyapatite crystals are precipitated. After suitable stimu- in maturation of secondary osteons (Reid & Boyde, 1987), change
lation, the osteoid is 70% calcified within a period of days, but in geometry through size increase as a result of changed endosteal
complete mineralization takes several months (Hayes, 1991). About and subperiosteal apposition rates (Gordon et al., 1989; Kontulai-
50% by volume and 75% by weight of bone is mineral, consisting nen et al., 2003), change in collagen cross-link concentrations
mostly of calcium phosphate and calcium carbonate in the form of (Oxlund et al., 1995; Köwitz et al., 1997), change in predominant
hydroxyapatite. The importance of collagen in determining bone collagen fiber orientation (Martin & Ishida, 1989; Boyde & Riggs,
strength has been rarely included in analyses of variance of 1990; Martin & Boardman, 1993; Riggs et al., 1993) and change in
bone strength, most likely because bone mineral density parameters bone collagen synthesis and remodeling (Tidswell et al., 2008).
became dominant, being so easy to measure. Conversely, quantifica- Changes in bone geometry and density resulting from exercise
tion of collagen in bone was, until recently, much more difficult are highly significant in altering the bone’s resistance to bending.
and required invasive sampling. But the central importance of col- The extent of the changes and the specific stimuli under which they
lagen and its orientation is clear since bone mineral is deposited in occur provide an understanding of the extent and rate of bone
and on collagen; and the directionality of the collagen fibrils deter- strength changes at the time of early conditioning exercise, training,
mines the orientation of the mineral plates or spindles (Wassen athletic competition, or lack of any of these stimuli.
et al., 2000; Viguet-Carrin et al., 2006). The diaphysis of immature individuals consists of primary osteo-
Compact cortical bone contains spaces within it, including osteo- nal bone embedded between more highly mineralized concentric
cyte lacunae, blood vessels in Haversian and Volkmann canals, and plates of woven bone. Bone formed post-natally at the periphery is
resorption spaces, the volume of which varies with age, recent more porous, contains wider, less mineralized lamellae, and has
loading regimen, metabolic status, injury, endocrine status, and osteons of greater diameter than does cortical bone formed pre-
lactation. Bone matrix is the same material in diaphyseal cortex and natally. The bone shaft can respond to exercise by enlargement as
a single trabecular plate. However cancellous bone tissue has a bone a result of the formation of woven-fibred bone in the periosteum,
volume fraction (proportion of total volume occupied by bone) of which then forms lamellar bone and becomes connected by radially
0.47–0.82 in the normal third carpal bone (Young et al., 1991a), oriented bony plates, with apposition of bone leading to new
which is significantly less than the range 0.92–0.98 for the mid- primary osteons (Stover et al., 1992). This allows prompt and rapid
radius cortex (Riggs et al., 1993), because the inter-trabecular spaces periosteal new bone formation, with an increase in resistance of the
are far larger than the spaces in the cortex. Within (any given) whole bone to bending and torsion. Compared with this periosteal
bone, the amount of mineral will differ between its regions, due to response in the diaphysis and metaphysis, the capability for epiphy-
differences in the relative amount and placement of cancellous and seal enlargement in response to the biomechanical stimuli is appar-
cortical tissue at different sites. The word ‘bone’ is applied to three ently limited, in both extent and rate, and occurs largely in early
different entities: bone material, bone tissue consisting of bone life.
material plus the spaces and tissues interposed around the material,
and bone as an ‘organ’ or ‘whole’ bone. The three entities have dif-
ferent mineral densities, which have been referred to as material,
The need for exercise
compartment, and total bone mineral densities respectively (Rauch The development and maintenance of tissues requires the delivery
& Schoenau, 2001). of a certain number of cycles per day at certain strain rates, in order

275
 13 The response of musculoskeletal tissues to exercise

to stimulate various tissues: for instance, in young rats taught to osteochondrosis in Warmblood foals. Three groups of foals were
jump, five cycles per day of this unusual activity produced as great kept from birth for 5 months in one of three exercise conditions:
a response in terms of resistance to fracture than did 40 jumps in a box stall with the mare (boxed group), boxed but exposed to
(Umemura et al., 1997). The effect on other tissues and species may sprint exercise 5 days per week in a 50-m long walled enclosure
differ, and each may have its own stimulus requirements if it is to from a young age (boxed/sprinted), or kept at pasture continually
adapt. Unfortunately, the number of other activities that deliver (pastured). After euthanasia of some foals at 5 months of age, the
these novel higher forces to tissues is not documented in either remaining foals until 11 months of age were out of the weather in
domesticated or wild horses. It is assumed to include the playing, a large stall, which was always open to a small enclosure. The
fighting, fast locomotion and other exuberant activities that young change in exercise regimen meant that the previously boxed group
mammals engage in habitually and adults less often. The duration had greater opportunity for exercise, the previously boxed/sprinted
of these high-intensity exercises need not be long. The natural and pastured groups had less (van Weeren & Barneveld, 1999a).
grazing and social slow movement behavior of wild horses is known At 5 months, pQCT showed that the cross sectional area (CSA)
to occupy ~17 h/day, whereas Dutch Warmblood foals at pasture of the third metacarpal bone (Mc3) was significantly less in the
did not canter more than approximately 3 min per day (Kurvers boxed than pastured group, but the BMDv was not different between
et al., 2006). groups, although BMDv in the dorsal cortex was significantly higher
The animal management systems of agricultural and competition in the boxed group, possibly due to a difference in the remodeling
animals have been in place for some hundreds of years, so it is rate. The between-group differences had disappeared in the foals
perhaps understandable that some profess to know the normal aged 11 months, and the pooled values at 11 months for area were
exercise requirements at all stages of a horse’s life. It is less reassur- higher and for density were lower than at 5 months (Cornelissen
ing that the literature contains hardly any data on how horses are et al., 1999), although there was no decrease in other foals pastured
managed, and much less data on how much horses move, whether from birth, and scanned at 6 and 9 months of age (Grace et al.,
they be domesticated, wild or feral. It is only recently that modern 2003), possibly due to different experimental design and technique
tracking technologies such as GPS have been applied in the equine (Firth, 2006).
field, and it is assuring that their price is falling to an affordable In the proximal sesamoid bone, the trabecular BMDv was highest
level. The upshot remains that we have no data to know how much in the boxed/sprinted group, and lowest in the boxed group, and
exercise a horse requires, in terms of duration, frequency, number the value in the latter increased greatly after the exercise regimen
of cycles, and when repetitions of different types of activity are changed at 6 months if age. The BMDv of the previously pastured
spread over time. In turn this implies firstly, that we may under- or group had increased by 11 months, but that of the previously
over-estimate how much exercise animals require physiologically, boxed/sprinted group had not. Confinement apparently had
secondly, the ideal amount of imposed exercise or imposed rest may resulted in a retardation of normal development which was com-
be doubly difficult to estimate, and thirdly, the control group of pensated for, at least partly, when the restriction on exercise was
some experiments may not have been ‘normal’ but rather abnor- lifted (Cornelissen et al., 1999).
mally inactive or sedentary, which may mean the effect of an The BMDv of the medial aspect of the third carpal bone was
imposed exercise regimen appears greater than if it had been significantly less in boxed foals at 5 months old, with the effect of
imposed on a normal control regimen (whatever that is). sprint training being similar to that of pasture exercise. By 11
This lack of knowledge is one reason for the current interest in months old, the reduced exercise possibility in previously pastured
the area of exercise in the young foal. But there is also little research and boxed/sprinted groups and the greater exercise in previously
in the area of training for particular athletic objectives. There is close boxed foals were associated with a significantly greater BMDv than
similarity of equine structure and function to that in people, and at 5 months old, but no difference between groups. The average
there is increasing interest from the human health community in dorso-palmar depth of the third carpal bone was not significantly
the responses of the tissues of children to exercise, or the lack of it. different between foals of 5 and 11 months old (Firth et al., 1999b).
The outcome parameters that have been used to quantify a bone BMDv was determined in the SCB of the proximal articular surface
response include radiography, radiogrammetry, radiographic bone of the proximal phalangeal bone of these six groups (three exercise
aluminium equivalence, point projection radiography, dual X-ray regimens, two ages) at 1-, 2-, 3-, 4- and 5-mm depth from the
absorptiometry (DXA), computed tomography (CT), peripheral articular surface at two differently loaded anatomical sites (site 1 at
quantitative computed tomography (pQCT), magnetic resonance the proximo-dorsal eminence and site 2 at the middle of the medial
imaging (MRI) (Firth, 2004), and clinical and pathological features articular surface; site 2 is continuously loaded, site 1 intermittently).
associated with, for instance, cartilage and subchondral bone At 5 months BMDv was significantly higher at site 2 than 1 with
lesions related to imposition of exercise. The latter are not consid- highest values and most apparent difference at depths to 3 mm. By
ered in this chapter. 11 months, site differences were less, due to a greater increase in
BMDv at site 1 in the previous 6 months, and BMDv was significantly
higher in site 2 than 1 in previously boxed/sprinted and trained
groups but not previously pastured foals. The significantly higher
The response of bone to exercise in the horse BMDv in pasture and boxed/sprinted than boxed foals at 5 months
The following is a summary of evidence so far available on bone was no longer present at 11 months at site 1, and at site 2 the previ-
tissue responses of horses, grouped in order of age when the exercise ously boxed/sprinted foals had significantly higher BMD than previ-
began, because there is difference in the response of the immature ously boxed and pastured groups. Continuous loading in the central
as opposed to the older skeleton, possibly because of the turnover area of the metacarpophalangeal joint surface, which is present even
time of collagen and its maturation in older subjects. The responses during standing and slow gaits, apparently stimulated distinct and
presented are mostly in either BETB or SCB of the epiphysis or of a significant changes in the SCB to a greater extent than intermittent
particular cuboidal bone, or diaphyseal cortex. Few studies mention loading at the proximal eminence of the proximal phalangeal bone
responses in the metaphysis, perhaps because this area is considered (Brama et al., 2009).
to be of less interest after a young age and because of the difficulty In the GEXA study 18 of 33 Thoroughbred foals born and raised
in using sensitive imaging methods to examine it, partly caused by at pasture began gentle exercise at 3 weeks of age and continued
the inability to separate trabecular from cortical metaphyseal bone until ~17 months of age. The remaining 15 foals had spontaneous
in projected radiographic images. pasture exercise only. The bone parameters of the foals having only
Several aspects of bone development became evident in an inves- spontaneous pasture exercise (PASTEX group) were compared with
tigation (EXOC study) of the relationship between exercise and those of the foals subjected to additional exercise (1030 m per day,

276
 Bone

5 days/week; CONDEX group) (Rogers et al., 2008a), by pQCT

undertaken serially at average ages of 4 days, and 2, 4, 12 and
17 months (Firth et al., 2010). The mid-diaphysis of the proximal
phalangeal and third metacarpal bone increased its resistance to
deformation by increasing bone mass, circumference, and mean
cortical thickness. The imposed exercise resulted in highly signifi-
cant differences in these parameters when serial values over
time were pooled, but were significant at specific ages (12 and
17 months) only in the phalangeal bone. There was no significant
difference between PASTEX and CONDEX groups with regard to
cortical bone density, indicating that the increase in bone strength
is related to a far greater extent to increase in bone size and mineral
content or to other factors such as collagen characteristics (see
above) than to change in BMDv. In the proximal metaphysis, the
trabecular BMDv (inner 45% of the scan slice) was lower in
the CONDEX than PASTEX group at 2 months of age and over the
whole 17 months when values were pooled, most likely associated
with the bone area of the CONDEX group being greater, although
not significantly. In the outer ring of the metaphysis, consisting of
both trabecular and cortical bone, bone mineral content was sig-
nificantly greater over time in CONDEX than PASTEX; bone area
was also greater but not significantly. The interpretation is that the
increase in size and mineral content of the outer part of the metaph-
ysis made the bone stiffer, resulting in lower strain and less mineral
accretion response in the central part of the metaphysis of the
CONDEX group. The conditioning exercise from 3 weeks to Fig 13.11  Radiograph of an identical preparation to that in Figure 13.10
17 months of age appeared to have no negative effects on bone, from a 19-month-old pasture-raised Thoroughbred filly that underwent 13
since no clinical radiographic abnormalities were detected. Also, the weeks of early training for racing. The orientation of the metaphyseal and
reduced bone density in the sagittal groove detected by CT (Kawcak epiphyseal trabecular bone is different from that in the untrained horse in
Fig. 13.10, there is increased radiodensity in the dorsal load path in the
et al., 2009) and previously described (Doube et al., 2007) were
intermediate carpal and radial epiphysis, and the subchondral bone line is
present in both PASTEX and CONDEX foals. Point projection radi-
regionally thickened. The dorsal cortical shell of the radial epiphysis and
ography (Firth et al., 2009) of the same joints of both conditioned
intermediate carpal bone is thicker. The changes are typical of the general
and non-conditioned horses showed osteochondrosis lesions at the changes present after galloping exercise.
disto-dorsal aspect of the sagittal ridge, which apparently heal spon-
taneously without negative effect on subsequent athletic activity
(Kane et al., 2003). This supported previous indications (Barneveld
& van Weeren, 1999; van Weeren & Barneveld, 1999b) that the for 19 months in an earlier experiment (Firth et al., 1999c). Some
presence and severity of osteochondrosis is not affected by exercise of these changes were discernible after 13 weeks of early training
in the young horse. (Fig. 13.11). Difference in bone density and in radionuclide uptake
In nineteen of the GEXA trial horses the bone response to race in carpal bones was not detected between 2-year-old horses exer-
training and withdrawal from training was determined (Firth et al., cised on a treadmill 5 days per week for 6 months and control
2007; Firth et al., 2012) by pQCT scanning immediately before horses walked for 7 min daily; that there were differences in the
training began, after training stopped, and after the horses had been fetlock joints shows not only the specific adaptive response within
at pasture for some months. In the diaphysis, the density of the bones, but also a regional difference in the response of different
third metacarpal bone was approximately 0.8% higher at the end bones to the exercise imposed upon the animal (Kawcak et al.,
of 2-year-old training than before it started, but it had returned to 2000). Recently, study of carpal bone tissue from Thoroughbreds
the previous value after being at pasture for several months. During aged 3–22 years showed that increased bone formation was accom-
training, diaphyseal bone strength index (the resistance of the bone panied by increased bone collagen synthesis and remodeling, par-
to bending and torsion) increased, due to both the slight density ticularly within the trabecular regions of the bone of horses in
increase but much more to the increase in size of the bone; the racing compared to those with no such history. The increase in bone
strength index continued to increase after pasturing, most likely density would lead to greater stiffness, particularly in the ‘stiffer’
because the horses were still growing, although the independent cortical bone, the failure of which was suggested to be associated
effect of the training exercise itself on diaphyseal size increase could with reduced support from the weaker underlying trabecular bone
not be tested. In the epiphysis, there was no increase in cross- because of lower mature cross-link content (Tidswell et al., 2008).
sectional area, but substantial and significant increase in mineral These findings and their suggested significance, in terms of end-
content and BMDv. In the inner zone of the epiphysis, trabecular stage disease that supervenes in racehorses (Pool, 1996), remain
bone density increased, as did density of the outer zone consisting unclear at this time.
of trabecular and more compact bone fractions; in neither did Horses aged 22 months trained conventionally on grass and sand
withdrawal from exercise for several months at pasture result in racetrack surfaces were compared with horses kept in large grass
significant decrease in BMDv. yards for 13 weeks. Four horses completed the training program,
In 19-month-old Thoroughbred fillies, which had been treadmill and three completed only cantering exercise and did almost no
exercised for 18 weeks, bone density of the distal radius, radial galloping (Firth et al., 2004). The latter had Mc3 diaphyseal bone
carpal bone, third carpal bone and proximal third metacarpal bone CSA, mineral content, periosteal circumference, and bone strength
was significantly greater than in age-matched horses that had similar to untrained horses, but less than those horses that had
been given only walker exercise (Firth et al., 1999b). The adaptive galloped. However, cantering exercise had not caused growth of the
changes in the trabecular bone was very obviously localized to the bone shaft, and cortical BMDv of horses that had cantered only was
load path of the higher forces resulting from the higher speed exer- similar to that of galloping horses, both being significantly higher
cise, and resembled those found in horses that had been exercised than that in untrained horses, despite numbers being small (Firth

277
 13 The response of musculoskeletal tissues to exercise

et al., 2005). The results were consistent with the previous studies Interestingly, when loading is discontinued, e.g. when horses are
in which metacarpal cortical thickness increased in treadmill- withdrawn from training, diaphyseal circumference will not dimin-
trained yearlings that had galloped (McCarthy & Jeffcott, 1992). ish, but the (relatively minor) increase in bone density will be
Significant relationships existed between months in training and reversed. In the epiphyses, which are for the larger part made up of
Mc3 bone geometry and moment of inertia in 2–4-year-old race- trabecular bone, increased loading through the intensification of
horses (Sherman et al., 1995). It was further shown that lunging exercise programs will lead to a substantial increase in density, which
yearling Quarter Horses for 20 min/day for 8 weeks had no signifi- is apparently retained after withdrawal from training exercise.
cant effect on osteocalcin concentrations (Fenton et al., 1999), and When assessing the effect of exercise on bone it should be realized
that velocities >12 m/s were required to affect the Mc3 dorsal cortex that the area is complex, with the nature of the exercise program
dimensions (Davies et al., 1999). itself being only one of the factors involved. Possible confounding
In the Mc3 epiphysis, early training resulted in highly localized factors include the base exercise in the subjects of study, and the
BMDv increases, with different patterns to that in Mt3 (Firth et al., effects of the early environment in which they were raised. Also, the
2005), which was described in detail (Boyde & Firth, 2005). The background of the control group matters, as no exercise effect can
changes in the galloped horses resembled those found in the proxi- be assessed against a nil option. To obtain better insight into the
mal sesamoid bone of 2-year-old racehorses, in which 5 months of effects of exercise in the horse, it is imperative that exercise levels
training on dirt tracks was associated with significantly higher densi- of individual horses be recorded on a large scale, including various
ties and greater trabecular width compared with horses at pasture age classes over prolonged, preferably career-spanning, periods of
(Young et al., 1991b). Similar regional adaptive changes were time. Relating workload patterns to data on performance, orthope-
present in the medial facet of the third carpal bone, in which the dic health and wastage may then provide better clues regarding the
remodeling response of the SCB was not uniform between horses, effects of specific exercise programs under field conditions for a
with higher densities being present in those that had galloped least. variety of equestrian disciplines.
The bulk trabecular bone apparently responded by formation of
increased bone mass in an arrangement resembling pillar structures
deep to the SCB (Firth & Rogers, 2005). Such changes are consistent Muscles
with indentation studies in the dorsal aspect of third carpal bones
of 2- and 3-year-old Thoroughbred horses that had raced or trained,
which had higher density and stiffness of the third carpal bone than
Overview
did untrained horses. Area fraction and stiffness were closely related, The horse’s skeletal musculature is highly developed, particularly in
and were higher in horses with than without pathological change athletic breeds. In contrast to most mammals, in which 30–40% of
(Young et al., 1991a). bodyweight consists of muscle, and to non-athletic horse breeds
Lastly, in this cohort of 2-year-old horses, changes in subchondral (~42%), more than half (~55%) of a mature Thoroughbred’s body
bone were detected in subchondral bone and articular calcified weight comprises skeletal muscle (Gunn, 1987). Low body fat and
cartilage of the distal Mc3. The lesions were present in both trained a large amount of muscle likely reflect adaptation and selective
and untrained horses, and it was suggested that the forces associated breeding of animals used for elite endurance and sprint racing
with athletic training might be less important than factors related (Kearns et al., 2002). Total muscle blood flow in horses that are
to bone development (Boyde & Firth, 2008). Although it is impos- exercising at a level when O2 consumption is at its maximum
sible to quantify the etiological significance of either, the findings (VO2max, 134 ± 2 mL/min/kg) has been estimated at 226 L/min,
are mentioned here to reinforce the point already alluded to above which represents approximately 78% of total cardiac output (Arm-
under cartilage, namely that optimal development requires a certain strong et al., 1992). During exercise and locomotion metabolites
biomechanical environment to proceed optimally, and inappropri- and oxygen reach skeletal muscle fibers via the respiratory, cardio-
ate exercise imposed early or later in juvenile life, can negatively vascular, and hematologic systems; in turn the muscle fibers produce
affect bone development. energy in the form of adenosine triphosphate (ATP), which, via the
The age at which horses are considered adult remains less than contractile machinery, is converted into mechanical work. The struc-
definitive, but if adulthood is accepted as being around the end of tural arrangement of the musculoskeletal system permits the use of
the horse’s fourth year, there are almost no studies of the effect of this energy to move the horse’s limbs in a characteristic rhythmic
exercise on bone of mature horses, besides studies of cumulative or pattern that is well established for each gait.
excessive exercise loads resulting in pathology and/or clinical signs Equine skeletal muscle is relatively heterogeneous and has con-
which are not considered in this text. There remains difficulty in siderable potential to adapt during training, largely mediated by the
assessing the effect of exercise since much of the adult response structural and functional plasticity of muscle fibers. These long-term
most likely is influenced by adaptive changes in earlier years, the (weeks to months) adaptations occur independently from the
nature of which will vary between individuals. This explains why immediate or short-term muscular metabolic responses to either a
the little information on adaptive responses to exercise in adult single bout of sub-maximal or near-maximal exercise. They are
horses is confined to cross-sectional information, such as that con- associated with altered rates and regulation of transcription of spe-
cerning carpal change (see above) and radial diaphyseal BMDv cific genes and consequently a change in the amount of specific
(Riggs, 1993b). isoforms of proteins expressed within individual fibers (Williams &
Neufer, 1996). Depending on the basal muscle status (e.g. breed,
age, sex, level of fitness and training history of the horse), and
characteristics of the stimulus (e.g. nature, intensity, duration, and
Conclusion frequency of exercise bouts, and total length of the conditioning
This section dealt principally with the young horse and the effects program), the adaptive response to training can take two different
of early exercise and its possible relationship to development and forms (Table 13.1). First, the quantitative responses, when myofi-
readiness for later athletic endeavour. It does not deal with excessive bers increase in size (hypertrophy) but otherwise retain their basal
or long-duration exercise that, through supra-maximal impact or structural, physiological, and biochemical properties. And secondly,
wear-and-tear, may lead to (pre-) lesional change. the qualitative responses or remodeling without hypertrophy, where
It can be concluded that, in compliance with Wolff’s law, the myofibers do not change in size but acquire markedly different
changes elicited by the forces that are applied to bone are regionally enzymatic and structural characteristics (e.g. fiber type transitions).
specific and consistent with the load path. Increase in strength of In practice, the most common adaptive responses of equine skeletal
the diaphysis is more dependent on increase in size than in density. muscle to training is a mixed form of remodeling with discrete

278
 Muscles

standards for training protocols and updated methods, have been

Table 13.1  Summary of the two basic responses of skeletal
published in this field. The readers of this book can be referred to
muscle to training; possible stimuli and the nature of the
recent book chapters (Rivero & Piercy, 2004, 2008), and reviews
responses and main physiological implications are indicated
(Rivero, 2007; Rogers et al., 2007; Votion et al., 2007; Rivero et al.,
2008) by the author covering these aspects. Although the majority
Quantitative responses: Qualitative responses: of these studies have been conducted on racehorses (Thorough-
hypertrophy remodeling breds, Standardbreds, and endurance horses), there are a small
number of specific studies about muscular adaptations to training
Stimulus
in horses from other disciplines (show jumpers, dressage, and
Short isometric actions Tonic and prolonged actions driving). Altogether, these recent studies provide valuable informa-
Light and prolonged Repeated activities (endurance tion about two particular aspects: 1) what can be modified in horse
overextension training) muscle by training? and 2) how can this be obtained. It is certain
Compensatory hypertrophy Interval and high intensity actions that answers to these two capital questions are of great interest to
people involved in equine sport practice. But to achieve this goal,
Plyometric contractions
some preliminary considerations about muscle exercise physiology
Responses are mandatory in order to provide an appropriate scenario for sub-
sequent considerations.
Muscle fiber hypertrophy Intracellular reorganization
Increased synthesis of muscle Changes in the expression of
proteins metabolic proteins A synopsis of equine muscle exercise
Increase in thickness and length of Increase in mitochondria and physiology
muscle mass capillaries
Changes in the expression of Since the pioneering study by Lindholm and Piehl (1974), analysis
contractile proteins of equine muscle physiology has centered on the use of the percu-
taneous needle biopsy technique (Fig. 13.13), which has proved to
Implications be an invaluable tool in defining the morphological, histochemical,
and biochemical properties of equine skeletal muscle. A detailed
Improvement of muscle strength Increase of muscle resistance
review of the technique for performing percutaneous needle muscle
generation Decrease of muscle shortening
biopsy and sample preparation can be obtained elsewhere (Rivero,
Improvement of explosive actions velocity
1996; Rivero & Piercy, 2004).
Increase in power generation Higher efficiency and economy of Biochemical analysis of homogenized equine muscle samples has
metabolism enabled the study of broad metabolic responses to exercise, and
metabolic adaptation to training through analysis of selected
muscle enzyme activities, their substrates, and intermediary metab-
or substantive hypertrophy, that is often accompanied by increases olites (see Snow & Valberg, 1994 for a review). Histochemical evalu-
in the number of capillaries (see Rivero, 2007; Rivero & Piercy, ation of muscle, combined with image analysis (Henckel et al.,
2008). 1998), has also provided invaluable information about contractile
As already mentioned, the modality and amplitude of the mus- and metabolic properties of equine muscles, specifically regarding
cular responses to training also depend significantly on the muscle fiber types, oxidative and glycolytic capacities, fiber sizes, and capil-
profile before training (Pette & Staron, 1997) because the increased lary density. Cellular and molecular diversity within equine muscle
contractile activity that is associated with training always induces a has also been addressed via study of myofibrillar and non-contractile
change toward slow and oxidative muscle profiles. Hence fast-twitch proteins by immunohistochemistry (Rivero et al., 1996a), gel elec-
fibers (and therefore muscles that contain a higher proportion of trophoresis (Rivero et al., 1997), and enzyme-linked immunosor-
fast-glycolytic fibers) can show a relatively greater training adapta- bent assay (Rivero et al., 1999). Molecular biology and genomic
tion than slow-twitch fibers. As a general rule, this response is techniques have also recently been applied to the field of equine
particularly prominent in young, inactive (Yamano et al, 2002) and muscle physiology (Eizema et al., 2003, 2005; Chikuni et al., 2004).
aged, sedentary (Kim et al, 2005) horses, which have a high percent- The fact that during the early phase of transformation between fiber
age of glycolytic (low oxidative) type IIX fibers (Fig. 13.12A,B), in types, isoforms-specific mRNA should be detectable some time
contrast to active mature horses, which have muscle fiber type pro- before a change in the specific protein, proves invaluable for exam-
files that are more oxidative (Fig. 13.12C,D). Although it can be ining exercise and training effects (Eizema et al., 2005).
hard to differentiate muscular adaptations caused by growth and Locomotor muscles are mainly located proximally on the
training, specific training effects have been delineated in growing appendicular skeleton, which has the effect of reducing the weight
foals (Dingboom et al, 2002; Eto et al, 2003; Rietbroek et al, of the lower limb and decreasing the energy necessary to over-
2007a,b), and young horses (Essén-Gustavsson et al, 1983; Ronéus come inertia when the limb swings back and forth (Barrey, 2008).
et al. 1992, 1994; Rivero & Serrano, 1999; Serrano & Rivero, 2000; Movements of the distal limb are mainly passive and result from
Serrano et al, 2000; Yamano et al, 2002). the release of elastic energy of the digital flexor tendons and sus-
From the extensive literature available, there is now sufficient pensor ligament when the limb is unloaded (Wilson et al., 2001).
evidence that equine skeletal muscle has great potential to adapt In contrast, movements of the proximal limb result from active
during exercise and, overall, these adaptations have important phys- muscular contraction. In general, most locomotor muscles are
iological implications for locomotion that include effects on power active during the propulsion stage of weight bearing in each limb
generation (strength), resistance to fatigue (stamina), and maximum (Robert et al., 1998). Muscle activity, as measured by electromyog-
velocity of shortening (speed) (Table 13.1). Much of this work is raphy, is positively correlated with running speed, but shows that
summarized in the excellent review chapter by Snow and Valberg muscles within the same synergic group have significant differ-
(1994). But most of these earlier studies on muscular responses to ences in their activities when a horse exercises at constant speed
training failed to describe accurately the training-induced muscular (Hyyppä & Hänninen, 1998).
changes. As a consequence, results from many of these studies are The total force exerted across a muscle is the sum of an active
of little, if any, application in practice. In recent years, however, a force generated by the contractile proteins and a passive force
considerable number of experimental studies, with enhanced provided by structures that are arranged in parallel with the

279
 13 The response of musculoskeletal tissues to exercise

A C

B D

Fig 13.12  Transverse serial sections of gluteus medius muscle biopsies stained with myofibrillar ATPase after acid preincubation at pH 4.45 (A and C) and
succinate dehydrogenase (B and D) in a young adult (3-year-old) and untrained Andalusian horse (A and B), and in an adult and regularly trained (10-year-
old) Andalusian horse (C and D). Note that the young animal has a lower proportion of type I fibers (black fibers in A and C) and a higher percentage of
low-oxidative type IIX fibers (gray fibers in A and C) than the adult horse; furthermore, differences in fiber size are much more pronounced in the young
animal than in the adult. This therefore may explain the broader range of adaptive responses to training in young untrained horses than in adults or regularly
trained animals. Bar = 50 µm.

contractile elements. Most of the natural equine gaits typically The unit of muscular contraction is the sarcomere, which is
consist of stretch-shortening exercise cycle (plyometric muscular formed by thick and thin filaments of contractile proteins. Muscle
contractions), characterized by multi-joint actions, rapid eccentric contraction occurs when, within each sarcomere, thin myofilaments
phases and explosive concentric muscular contractions potentiated slide over thick myofilaments, bringing the adjacent Z disks closer
by the stretch reflex. More than 90% of a muscle is made up of together. The myosin head of thick filaments is the motor domain
myofibers, with the rest consisting of nerves, blood vessels, and the that contains the ATP-blinding site, and the myofibrillar ATPase
fat and connective tissue that separate the individual fibers (endo- enzyme. The force that is generated in sarcomeres is transmitted via
mysium), the fascicles (perimysium), and the whole muscle (epi- the contractile apparatus to intermediate filament proteins that
mysium). The connective tissue merges with both the origin and provide a structural link first to the sarcolemma and then to the
the insertion tendons of the muscle, as well as with internal tendons extracellular matrix and the connective tissues of tendons, and ulti-
in compartmentalized muscles. Regional functional architecture mately to the bones of the skeleton.
and specialization of muscles of the pelvic and thoracic limbs have Functionally, muscle fibers are organized as motor units. A motor
been reported in horses (Payne et al., 2005a,b). In general, the unit consists of an α-motoneuron and the set of skeletal muscle
proximal limb is characterized by muscles with large volumes, long fibers that it innervates. The ability of equine muscle tissue to
and parallel fascicles, and extremely short tendons/aponeuroses, perform efficiently in spite of very different types of exercise is
whereas the distal lib is characterized by muscles with small enhanced by the muscle heterogeneity. Equine skeletal muscle is
volumes, short, pinnate fascicles and long tendons. Hence, in composed of three main pure fiber types termed I, IIA, and IIX, and
general, proximal limb musculature is specialized for work output, a fourth hybrid phenotype termed IIAX, which express myosin
while the distal limb musculature is specialized for generating force heavy chain (MHC) isoforms (Fig. 13.14). There are important dif-
economically. ferences in the physiological, biochemical, and morphological

280
 Muscles

Fig 13.13  (A) Instruments used in percutaneous

muscle biopsy. Scalpel (left) and biopsy needles
(right). The top needle, which has an outer
diameter of 6 mm, was first introduced by
Bergström (1962) and was further modified with
finger and thumb rings by Henckel (1983). (B) Site
for the collection of biopsy specimens from the
right gluteus medius muscle according to
Lindholm and Piehl (1974); this fixed site is
located one-third of the distance along a line
running from the tuber coxae to the root of the
tail. (C–E) An illustration of the various steps for
the needle biopsy technique; (C) the needle,
A B together with the internal cutting cylinder, is
inserted into the muscle; (D) within the muscle,
the cutting cylinder is partially withdrawn so
that the window is opened, allowing muscle
to enter the slot, and a piece is then cut by
pushing down the internal cylinder; (E) between
50 and 150 mg of muscle tissue is usually
acquired.

C D E

properties of these muscle fiber types. Hybrid IIAX fibers exist in non-myogenic (neural input, neuromuscular activity, and extracel-
equine locomotor muscles in significant numbers (Dingboom lular substances) in nature, regulate this percentage. Much more
et al., 1999; Linnane et al., 1999), although the co-expression of IIA important is, however, the significance for performance of this
and IIX MHCs at the protein level seems not to be reflected by muscle fiber type diversity. For example, endurance ability is cor-
co-transcription of the corresponding genes (Eizema et al., 2005). related with high percentages of types I and IIA fibers (Rivero et al.,
This suggest that hybrid fibers may represent fibers that are undergo- 1993), whereas speed ability is correlated with high percentages of
ing committed fiber type switching, towards the type corresponding fast-twitch type IIA and IIX fibers (Barrey et al., 1999), and power
to the mRNA currently being expressed, and that this conversion generation (strength) is proportional to fiber size (Rivero et al.,
occurs spontaneously in equine muscle not subjected to any specific 2001). Interestingly, some myofibers characteristics are positively
training stimulus. correlated with both the percentages of IIA fibers (Persson et al.,
All fibers within a single motor unit are of the same histochemical 1991) and fiber size (Rivero & Clayton, 1996), and the stance time
type. Motor units are selectively recruited in a specific pattern that of the stride is inversely correlated with the percentage of IIX fibers
changes according to the gait and the intensity and duration of (Ronéus et al., 1995) and fiber diameter (Rivero & Clayton, 1996).
exercise (Eto et al., 2006). For the maintenance of posture, only type Furthermore, certain primary muscular adaptations to training
I motor units are recruited. As the intensity and duration of exercise occur with concomitant modifications in the temporal characteris-
increase, further motor units are recruited, in the rank order: tics of the stride (Rivero et al., 2001).
I→IIA→IIAX→IIX. Type IIX motor units are only recruited at near- Muscle cannot contract without a source of energy provided by
maximal exercise intensity (sprint and jumping), and during the hydrolysis of ATP that occurs at the head of the myosin mole-
extremely prolonged submaximal exercise (Valberg, 1996; Yamano cule. As muscle (ATP) is limited and it is metabolized within a few
et al., 2006). seconds of the exercise, replacement of ATP is mandatory. There are
The percentage of each muscle fiber type present in a muscle (e.g. different metabolic pathways for this goal within the muscle
fiber type composition) varies from horse to horse (Fig. 13.15), and during exercise (see Votion et al., 2007). Within mitochondria,
multiple factors, both myogenic (lineage, breed, age, and sex), and β-oxidation of free fatty acids, the tricarboxylic cycle and oxidative

281
 13 The response of musculoskeletal tissues to exercise

A B C

D E F

G H I

J K L

Fig 13.14  Serial cryosections of adult horse M. gluteus medius stained by immunohistochemistry, enzyme histochemistry and histology. (A–D) Sections were
stained with a number of monoclonal antibodies against specific myosin heavy chain (MHC) isoforms: BA-D5 (A, anti MHC-β/slow), SC-71 (B, anti MHC-IIA),
BF-35 (C, anti MHCs β/slow and IIA), and S5–8H2 (D, anti MHCs β/slow and IID/X). (E–G) Sections were assayed for myofibrillar actomyosin adenosine
triphosphatase activity after acid (pH 4.45, E) and alkaline (pH 10.45, F) preincubations, and by Blanco and Sieck’s quantitative histochemical procedure. (H–J)
Sections assayed for succinate dehydrogenase and (H), glycerol-3-phosphate dehydrogenase activities (I) and periodic acid-Schiff (PAS) for selective staining
of glycogen (J). (K–L) PAS with α-amylase digestion, for visualizing capillaries (K) and hematoxylin and eosin for visualizing myonuclei (L). The fibers labeled 1,
2, 3, 4 and 5 are types I, IIA, IIAx (e.g. a hybrid fiber with more MHC-IIA content than MHC-IIX), IIaX (e.g. a hybrid fiber with less MHC-IIA content than
MHC-IIX), and IIX, respectively. Bar = 50 µm.
(G) Reproduced from Blanco & Sieck (1992).

282
 Muscles

Fig 13.15  Transverse sections of M. gluteus medius specimens (equivalent depth) stained with myofibrillar adenosine triphosphatase after acid preincubation
(pH 4.4), removed at the same depth, from four different breeds of athletic horses. Note how the percentage of type I (oxidative) fibers (black fibers) and the
fiber size increases from the fastest breed (Thoroughbred) towards the slower, more endurance-suited breed (Arabian). Bar = 150 µm.

phosphorylation (via the electron transport chain) combine to point when the increased rate of lactate production can be detected
produce ATP aerobically (aerobic pathways). Two additional anaer- in the plasma is known as the anaerobic threshold. This threshold
obic mechanisms (anaerobic pathways) exist: the high-energy phos- varies and depends on several factors, including the muscle’s
phate pathway and glycolysis. Aerobic production of ATP is a fiber type composition and the level of fitness. In practice (and
relatively slow but efficient process, while anaerobic pathways henceforth in this chapter), aerobic capacity is defined as the
produce energy rapidly but relatively inefficiently. Although both maximum amount of energy provided by aerobic pathways, whereas
pathways are active during exercise, the relative contribution anaerobic capacity is the maximum amount of energy provided
depends on the nature, intensity, duration and frequency of the by anaerobic pathways (Hinchcliff et al., 2002).
exercise, the muscle’s fiber type composition, the availability of At the beginning of the sub-maximal exercise (<85% of VO2max),
oxygen and substrates, and the presence of intermediary metabo- muscle glycogenosis is the main mechanism for providing acetyl-
lites that may potentially activate or inhibit selected enzymes. CoA from pyruvate. Either intramuscular glycogen or blood glucose
In exercise of low or moderate intensities, the main energy provi- are the substrates for this pathway. But as energy demands increase,
sion for muscular functions comes from aerobic pathways. As exer- higher rates of pyruvate oxidation tend to cause a shift towards free
cise intensity increases, a greater proportion of the energy is supplied fatty acids β-oxidation. The overall effect is that muscle glycogenesis
by the anaerobic pathway. At the point where the availability of declines over time during sub-maximal exercise, whereas free fatty
oxygen becomes a limiting factor in mitochondria, pyruvate from acids β-oxidation increases. Although lipids are the predominant
the anaerobic glycolysis cannot be converted into acetyl-CoA (the fuel utilized during prolonged sub-maximal exercise, fatigue occurs
substrate for the Krebs’ cycle), but it is converted to lactate. The by intramuscular glycogen depletion, as the main cause.

283
 13 The response of musculoskeletal tissues to exercise

During maximal exercise (>85% of VO2max), the high functional et al., 1996). When considered together, the only explanation for
demands require recruitment of most motor units. At this time, an increase in muscle mass, despite either no change or a reduction
intramuscular glycogen and blood glucose act as the predominant in fiber size, is a parallel increase in the number of muscle fibers
fuels to replenish ATP through anaerobic glycolysis. Limitations (hyperplasia), as previously demonstrated in humans (Sjöström
imposed by this high-intensity exercise result in greater amounts of et al., 1991).
pyruvate being converted to lactate rather than acetyl-CoA. As a
consequence, muscle (lactate) increases in a rate proportional to the
percentage of type II fibers. Initially, intracellular lactate accumula- Muscle fiber type transitions
tion is removed from the cell by active transport into the blood, but Muscle fiber type distribution and MHC composition are strongly
saturation of this mechanism results in a sudden exponential rise influenced by training (Figs 13.12, 13.16, 13.17). Studies on endur-
in intracellular lactate accumulation (anaerobic threshold), that ance training in horses have demonstrated (by myofibrillar ATPase
generally occurs when the plasma (lactate) reaches about 4 mmol/L. histochemistry, immunohistochemistry and electrophoresis),
The rise in intracellular lactate results in a significant reduction in increases in the fraction of type IIA fibers and a concomitant
cytoplasmic pH, the main cause of fatigue during maximal exercise. decrease in the proportion of IIX fibers (Essén-Gustavsson & Lind-
Local acidosis leads to impairment of both muscle structure and holm, 1985; Essén-Gustavsson et al., 1989; López-Rivero et al.,
function. A fall in cytoplasmic pH is partially overcome by a buffer- 1991; Sinha et al., 1993; Rivero et al., 1995a, 2001, 2007; Rivero,
ing system within myofibers. 1996; Tyler et al., 1998; Miyata et al., 1999; Rivero & Letelier, 2000;
From the biochemical perspective, the main goal of muscle con- Serrano & Rivero, 2000; Serrano et al., 2000; Yamano et al., 2002;
ditioning is to increase performance by: (1) increasing aerobic and/ D’Angelis et al., 2005; Gondim et al., 2005; Kim et al., 2005; Riet-
or anaerobic capacities, (2) reducing the major causes of fatigue broek et al., 2007b). In addition, several endurance training studies
during sub-maximal (e.g. heavy intramuscular glycogen depletion) in horses have reported fiber transitions beyond type IIA fibers, e.g.
or near-maximal (e.g. intramuscular acidosis) exercise, or (3) both an increase in hybrid I+IIA fibers and pure type I fibers (Henckel,
situations. 1983; Ronéus et al., 1987; Rivero et al., 1995a; Serrano et al.,
2000). Fiber type transitions during resistance training appear to
Muscular adaptations to training resemble qualitatively, those observed in endurance training. Hence,
strength training in horses has been shown to result in both
Muscle fiber size increased IIA : IIX fiber ratio (Gottlieb et al., 1989; Rivero & Serrano,
1999) and, when training is long enough, the I : II fiber ratio
The effects of training on equine muscle fiber size are still contro- (Serrano et al., 2000). Similarly, sprint training in horses results in
versial. In general, the adaptive response of equine skeletal muscle increased numbers of type IIA and decreased numbers of type IIX
to early and long-term exercise training takes the form of remodel- fibers (Lovell & Rose, 1991), with a corresponding change in their
ing with minimal, if any muscle fiber hypertrophy (see Snow & respective MHC content (Miyata et al., 1999; Rivero et al., 2002).
Valberg, 1994; and Rivero, 1997 for reviews). However, specific In contrast to endurance and strength training, a specific decrease
muscle fiber hypertrophy can be stimulated with bursts of high of type I fibers has been reported as an early and probably transitory
resistance muscle activity (Heck et al, 1996; Serrano & Rivero, response to high-intensity training (Lovell & Rose, 1991; Rivero
2000) and by prolonged stretch beyond normal resting length et al., 2002).
(Tyler et al, 1998; Serrano et al, 2000). Weight bearing, as a form When these various training studies are considered in combina-
of progressive resistance exercise training, has been investigated in tion, it is reasonable to assume that fiber type transitions occur in
ponies for improving strength, and resulted in increased muscle a graded and orderly sequential manner and typically change from
power, muscle size, and increased (though not statistically signifi- faster, more glycolytic fibers to slower and more oxidative fiber
cant) type II fiber cross sectional area, without parallel changes in types, e.g. IIX → IIAX → IIA → IIA+I → I (Pette & Staron, 1997). A
MHC composition (Heck et al, 1996). Six months of conventional dose–response relationship between the duration (in total) of the
jump training in competitive show jumpers also induced a selective training program and the magnitude of induced changes has
hypertrophic growth of type II fibers, with minimal switching recently been demonstrated at the molecular level (Serrano &
between myofiber phenotypes (Rivero & Letelier, 2000). Other lon- Rivero, 2000; Serrano et al., 2000). This relationship can be explained
gitudinal studies have also reported significant and early (less than more readily in terms of a threshold for the type IIX → IIA transition
3 months) increases in the mean cross-sectional areas of type I and/ during the early phase of training, and then a further threshold for
or IIA fibers after training (Gottlieb et al, 1989; López-Rivero et al, the type IIA → type I transition. Thus, a single fiber is capable of a
1992; Rivero et al, 1995a, 2001, 2007; Tyler et al, 1998; Miyata et al, complete fast-to-slow transformation in response to a sufficiently
1999; Serrano et al, 2000; Yamano et al, 2002; D’Angelis et al, long physiological training stimulus. Finally, although many reports
2005), a result also noted in horses less than a year of age (Eto et al, describe the training response shown by muscle fibers in terms of
2003; Rietbroek et al, 2007b). A marked and generalized fiber the MHC component, it is important to remember that many other
hypertrophy (~50%) has recently been documented in endurance- protein isoforms of the sarcomere (such as the regulatory proteins
trained Arabians performing moderate intensity exercise (~80% of of the thin filaments) and the calcium regulatory proteins of the
VO2max) of long duration (50–80 min/day), 4 days/week for 3 sarcoplasmic reticulum, change in parallel (Pette, 1998).
months (D’Angelis et al, 2005). Even though a fiber transition
occurs towards more oxidative fibers (IIX → IIA → I) with normally
smaller cross sectional area, the hypertrophy may represent either Metabolic changes and increased
the new fiber’s retaining size characteristics of its previous type, or
generalized increased protein production (or both). In contrast,
capillary density
other studies of Standardbreds and Thoroughbreds reveal minimal Perhaps the most commonly detected and earliest muscular adapta-
changes (Henckel, 1983; Lindholm et al, 1983; Foreman et al, tion to training is an increase in the activity of enzymes of aerobic
1990; Rivero et al, 1996b) or even a reduction in fiber cross- metabolism, such as selected enzymes of the TCA cycle, the
sectional area (Essén-Gustavsson & Lindholm, 1985; Essén-Gustavs- electron transport chain and fat oxidation (Lovell & Rose, 1991;
son et al., 1989; Ronéus et al, 1992, 1993; Rivero et al, 2002). These Ronéus et al., 1991; Rivero et al., 1995b, 2007; Eaton et al., 1999;
observations are hard to reconcile with the prominent increase in Miyata et al., 1999; Rivero & Letelier, 2000; Serrano et al.,
muscle mass, especially in the hindquarters, generally observed in 2000; Yamano et al., 2002; McGowan et al., 2002; Eto et al., 2003;
horses after most training programs (Snow & Valberg, 1994; Heck Kim et al., 2005; Rietbroek et al., 2007b). These changes are

284
 Muscles

A C

B D

Fig 13.16  Transverse serial sections of M. gluteus medius biopsies (depth, 6 cm) of the same horse before (A and B) and after (C and D) a long-term
endurance training program (9 months in total). (A and C) Sections are stained by immunohistochemistry with a monoclonal antibody to types I or IIA
myosin heavy chain isoforms; note that almost all muscle fibers express either or both of these isoforms after training. (B and D) The same sections stained
by immunohistochemistry with a monoclonal antibody specific to type IIA myosin heavy chain isoform; note the significant increase in the number of fibers
expressing this isoform after 9 months of training. Bar = 50 µm. Details of the training program are described in Serrano et al. (2000).

associated with increased mitochondrial and capillary densities high-intensity, short duration exercise (~100–140% of the speed
(Fig. 13.18) (Sinha et al., 1993; Misumi et al., 1995; Rivero et al., that results in a blood lactate of 4 mmol/L, (or V4) for 15 min, or
1995a, 2007; Ronéus et al., 1995; Tyler et al., 1998), the latter moderate intensity, long duration exercise (~65% of V4, 60–90 min)
response promoting improved oxygen diffusion and more expedi- exercise for a total of 5 weeks (Rivero et al., 2002). Similar results
tious removal of waste products (such as CO2). A recent study in have recently been reported in adolescent Thoroughbreds after 3
growing foals has reported, however, that exercise training had a weeks of high-intensity training (Rivero et al., 2007). Although
negative effect on both the capillary supply and oxidative capacity training also causes increased AMP deaminase activity along with
of skeletal muscle in horses (Rietbroek et al., 2007a). other purine nucleotide cycle enzymes, such as creatine kinase (dis-
The activities of key anaerobic enzymes, such as phosphofructo- cussed in Snow & Valberg, 1994), the concentration of total nucleo-
kinase and lactate dehydrogenase, either do not change or decrease tide stores is usually not affected by training (Lovell & Rose, 1991).
following training (Snow & Valberg, 1994; Rivero et al., 1995b; Nevertheless, moderate intensity (~55% of VO2max) and long dura-
Geor et al., 1999; Miyata et al., 1999; Serrano et al., 2000). However, tion (60 min, over ~13–14 km) exercise for 10 consecutive days,
high-intensity training (~80–100% of VO2max, 5 min × 2, 5 days/ increases muscle phosphocreatine concentration and reduces
week for 12–16 weeks) for both young (Yamano et al., 2002) and muscle creatine content at the point of fatigue (Geor et al., 1999).
adult (Eto et al., 2004). Thoroughbreds did result in increased gly- Training has also been shown to result in a modest increase in
colytic phosphofructokinase enzyme activity. Histochemical assess- muscle glycogen storage (Foreman et al., 1990; Gansen et al., 1999;
ment of glycerol-3-phosphate dehydrogenase enzyme activity, an Serrano et al., 2000; McGowan et al., 2002; Rivero et al., 2002) that
enzyme well correlated with other glycolytic enzymes, was also may well be associated with reduced activities of glycolytic enzymes,
increased by exercise consisting of alternating days of either since the capacity to mobilize endogenous glycogen is partially

285
 13 The response of musculoskeletal tissues to exercise

increase in muscle fiber size, recruitment of larger (type II) motor

units, and increased synchronization of the muscle fiber action
potentials.
Training has significant effects on the electrical and ionic mem-
brane properties of equine skeletal muscle (McCutcheon et al.,
1999; Suwannachot et al., 1999, 2001; Rietbroek et al., 2007a).
Short-term exercise training of moderate intensity results in an
increase in the density of Na+/K+-ATPase pumps measured by
ouabain binding assays, together with an attenuation in K+ efflux
from working muscles during high-intensity exercise. Although the
physiologic implication of these training-induced adaptations is
unclear, enhanced ionic control at the sarcolemma may result in
myofibers that are better able to respond to the rate of motor
neuron discharge (Bottinelli & Reggiani, 2000). Additional
responses following physical conditioning include increased SR
calcium uptake (Ca2+-ATPase activity) and an attenuation of the
exercise-induced decline of calcium uptake (Wilson et al., 1998).
Several studies have reported an increase in buffering capacity of
equine skeletal muscle after a few weeks of sprint training (Fox
et al., 1987; McCutcheon et al., 1987; Sinha et al., 1991, 1993), and
after prolonged (34 weeks) training (McGowan et al., 2002). This
increase may be due to 1) increased incorporation of myofiber
protein, 2) increased creatine phosphate concentration or 3) higher
muscle carnosine concentrations (Hyyppä & Pösö, 1998), although
with regard to the latter, no differences in carnosine concentration
were found in muscle from either trained or untrained Thorough-
breds (Marlin et al., 1989).

Other training consequences

Training increases the expression of heat shock proteins (Pösö et al.,
2002) and splice variants of neuronal nitric oxide synthase (Gondim
et al., 2005), which protect muscle cells from exercise-induced oxi-
dative stress. Furthermore, the reported apoptosis that occurs in
Fig 13.17  Eight percent sodium dodecyl polyacrilamide gel electrophoresis trained equine skeletal muscle (Boffi et al., 2002) probably pro-
of muscle samples from the M. gluteus medius of a control horse (upper row) motes the work/recovery/rebound/supercompensation cycle, when
and a trained horse (bottom row) throughout an experiment to investigate unconditioned muscle cells undergo programmed cell death, to be
the effects of prolonged endurance exercise training and detraining replaced by new, stronger cells. In fact, training attenuates exercise-
program. 0, pre-training; 3 and 8, after 3 and 8 months of training; 11, after a
induced ultrastructural damage in equine skeletal muscle (Kim
further 3 months of detraining. More details of the experiment in Serrano
et al., 2005), probably via a parallel fiber type transition, myofiber
and Rivero (2000). The three myosin heavy chain isoforms are identified as I,
IIA and IIX. Note the effect of training on the relative densities of the I and
hypertrophy or increased satellite cell activation (or perhaps
IIX bands, indicating a fiber type transition in the order IIX → IIA → I. all three).

Overtraining
influenced by the absolute activity of anaerobic enzymes expressed Overtraining is a syndrome that has been recognized as a cause of
within muscle fibers (Booth & Baldwin, 1996). In experimental poor performance in racehorses over the course of a high intensity
animals training is known to increase the sensitivity of glucose and prolonged (more than 4 months) training (Bruin et al., 1994;
uptake across the sarcolemma via increased GLUT-4 expression Tyler et al., 1998; McGowan et al., 2002; Rivero et al., 2008).
(Rodnick et al., 1992). In horses, moderate-intensity exercise train- However clinical signs of overtraining are accompanied by relatively
ing increases middle gluteal muscle GLUT-4 protein content, but few characteristic changes in skeletal muscle. In the study by Tyler
this change is not reflected by increased sarcolemmal glucose trans- et al. (1998), overtraining caused significant type IIA fiber atrophy
port in post exercise muscle samples (McCutcheon et al., 2002). (~8%), a decrease in type IIA : IIX mATPase fiber ratio, and increased
Finally, the transfer of FFAs from the vascular to the intracellular mitochondrial volume in type I (~16%) and type II (~39%) fibers
compartment is also enhanced with endurance training, and is accompanied by an increase in VO2max (~8%). As a consequence,
promoted by increased extracellular (interstitial) albumin concen- muscle fibers of overtrained horses are generally highly oxidative
tration (Heilig & Pette, 1988). (Karlstrom & Essén-Gustavsson, 2002). Additional metabolic adap-
tation of overtrained muscle includes ATP depletion (Bruin et al.,
1994), and reduced resting glycogen concentration (McGowan
Physiological adaptations and buffering capacity et al., 2002), though this latter finding may relate to delayed glyco-
Training results in increased motor unit action potential amplitude gen replenishment seen in normal horses, particularly given that
indicative of recruitment of larger motor units and synchronization glycogen utilization during exercise itself, is not influenced by over-
when analyzed by quantitative electromyography (Wijnberg et al., training (McGowan et al., 2002).
2008). An increase of muscle force as a result of training can be,
among others, the result of increased muscle fiber hypertrophy and/
or motor unit synchronization (Gabriel et al., 2006). Thus, the
Detraining
significant increase in motor unit potential amplitude seen after Maintenance of the trained muscle signature during inactivity is
training (Winjberg et al., 2008) is most likely the result of an more prolonged in horses than other athletic species, lasting

286
 Muscles

A C

B D

Fig 13.18  Transverse serial sections of muscle biopsy samples from the M. gluteus medius from the same horse taken before (A, B) and after (C, D) 9 months
of prolonged endurance training. (A, C) Sections are stained with succinate dehydrogenase to demonstrate the oxidative capacity of individual muscle fibers;
note the increase in the number of fibers with dark staining after training. (B, D) Sections are stained with the α-amylase PAS to visualize capillaries; note the
increased capillary density (e.g. number of capillaries per mm2) after the training program. Bar = 75 µm.

throughout 5–6 weeks of inactivity (Essén-Gustavsson et al., 1989; phenotype parallel a reversion of the muscle’s size and return of the
Foreman et al., 1990; Tyler et al., 1998) although not beyond 12 muscle’s metabolic and capillary characteristics to pre-training
weeks (Rivero et al., 1995a; Serrano & Rivero, 2000; Serrano et al., levels (Serrano et al., 2000). In summary, fiber sizes decrease,
2000). Researchers have suggested that expression of the MHC-IIX together with a decline in mitochondrial density, aerobic enzyme
gene constitutes a default setting that may be altered (decreased) by activities and glycogen content and normalization of anaerobic
chronic increased contractile activity (e.g. training), and compen- enzyme activities.
sated for, by increased expression of MHC-IIA (Goldspink et al.,
1992). In line with this hypothesis is the observation that a return Possible mechanisms underlying muscular
to sedentary activity levels following a prolonged endurance train- adaptations to training
ing period normalizes the expression of MHC-IIX, via a slow-to-fast
fiber type transformation in the order I → IIA → IIX (Fig. 13.17) Skeletal muscle responds to altered functional demands by specific
(Serrano et al., 2000). These detraining-induced changes in MHC quantitative and/or qualitative alterations in gene expression

287
 13 The response of musculoskeletal tissues to exercise

Muscle fiber
Box 13.1  Steps in cascade of events by which exercise and
increased neuromuscular activity lead to physiologically
Assembly relevant changes in the characteristics of skeletal muscle
Protein Sarcomere
• Exercise
• Contractile activity
• Messengers and signal transduction pathways
Synthesis Degradation
Amino acids • Regulatory genes
Nucleotides • Structural genes
Translation • Muscle fiber characteristics
Degradation • Muscle tissue characteristics
mRNA
Stability

Transport storage
control
requirement and energy supply is possibly the most important
DNA signal that triggers appropriate adjustment of contractile and meta-
Processing bolic protein expression (Green et al., 1992). Finally, recent years
Transcription
mRNA have seen significant advances in our understanding of the signaling
Nucleus mechanism by which the information contained in specific action
potential patterns alters transcription within muscle fiber nuclei
(Yan et al., 2001). For example, Ras-mitogen-activated protein
Fig 13.19  Schematic diagram showing different steps in the regulation of kinase (Murgia et al., 2000) and calcineurin (Serrano et al., 2001)
gene expression in skeletal muscle associated with increased contractile signaling are implicated in the α-motor neuron induction of slow
activity. muscle fiber phenotype, but not muscle growth. Conversely, a
protein kinase β-dependent and rapamycin-sensitive pathway con-
trols myofiber growth but not fiber type specification (Pallafacchina
et al., 2002).
(Fig. 13.19) provided the stimuli are of sufficient magnitude and
Specific genes known to undergo altered expression following
duration (Pette, 1998). Repeated or persistent elevation of neuro-
contractile activity include those encoding sarcomeric and cytosolic
muscular activity (e.g. during exercise and training) induces a series
proteins and enzymes of the glycolytic pathway, TCA cycle, the
of concerted changes in gene expression, evoking either myofiber
electron transport chain and fat oxidation (Williams & Neufer,
hypertrophy or myofiber remodeling, or both (Williams & Neufer,
1996). Additionally, signals inducing the expression of proteins
1996). Myofiber hypertrophy is generally characterized by a coor-
derived from mitochondrial genes are coordinated with the activa-
dinated increase in abundance (per fiber) of most protein constitu-
tion of the various nuclear genes that encode mitochondrial
ents. To some extent this process includes the selective and transient
proteins.
activation of specific genes immediately following the onset of work
The factors that promote angiogenesis in skeletal muscle in
overload. The major events, however, underlying muscle hypertro-
response to training have not been clarified although they may be
phy involve a general and non-specific augmentation of protein
related to a chronic increase in muscle capillary blood flow and the
synthesis within cells. Remodeling of myofiber phenotype, with
corresponding endothelial shear stress, as well as increased capillary
minimal or no hypertrophy, is the typical muscular response to
wall tension (Hudlicka et al., 1992). Hudlicka and colleagues spec-
prolonged training in the horse (Serrano et al., 2000). During this
ulated that endothelial stress might disturb the luminal surface,
type of adaptation, myofibers undergo striking reorganization, with
resulting in the release of bound proteases that damage the base-
selective activation and repression of many genes, and switching
ment membrane and contribute to an increase in basic fibroblast
between different myofibrillar isoproteins occurring in a graded and
growth factor release (Hudlicka et al., 1992). Subsequently, growth
orderly sequential manner (Pette, 1998). These changes occur in
factors may enhance vascular growth and satellite cell proliferation
parallel, but not simultaneously over time, and correspond to the
(Morrow et al., 1990). Still unclear, however, is the influence of
changes observed in enzymatic profiles, cytosolic proteins and
training intensity and duration on neovascularization and the
membrane receptors and transporters.
mechanisms that underlie the increase seen in intramuscular sub-
The complexity and pleiotropic nature of the physical and meta-
strates in response to long-term endurance training. These latter
bolic stimuli presented to myofibers during contraction that ulti-
adaptations may be related to either 1) increased glucose and FFA
mately results in altered gene regulation were reviewed by Williams
availability (via GLUT-4 and albumin respectively), 2) a lower uti-
and Neufer (1996). Acetylcholine released from motor neurons and
lization of these substrates for energy production or 3) possible
other signaling molecules of neural origin, bind to cell surface
artifacts imposed by experimental design (e.g. they may be a reflec-
receptors on myofibers, triggering intracellular events that may be
tion of increased dietary intake of soluble starches and fat from a
linked to altered gene expression (Box 13.1). Additional signals are
parallel change in diet in horses in training (Snow & Valberg,
probably derived from contraction-induced mechanical stress per-
1994)).
turbing the sarcolemma and extracellular matrix, thus exerting
tension via intermediate filaments on the cytoskeleton, organelles
and the nucleus (Milner et al., 1999). Changes in the intracellular Implications of training-induced changes to
concentrations of ions and metabolites during repeated muscle
contraction are also implicated in the activation of signaling path-
the physiologic response to exercise
ways (Booth & Baldwin, 1996). These signals include intracellular The main physiologic consequence of increased muscle mass in
calcium, hydrogen ions produced during anaerobic exercise, the response to training is to produce a muscle with a greater peak force
marked reduction in phosphorylation potential of the adenylate capacity, because force output is proportional to total cross-sectional
system (ATP/ADPfree), a depletion of the redox state (NADH/NAD) area of the fiber mass recruited (Bottinelli & Reggiani, 2000). At
and hypoxia. Among all these factors, an imbalance between energy slow speeds, this adaptation has an impact on gait, causing a

288
 Muscles

marked reduction of both stance time and stride duration (Rivero

Table 13.2  Summary of main muscular adaptations to training
et al., 2001). This adaptation significantly affects the performance
reported in horses and their functional significance for
of show jumpers via enhanced power output from the hindquarters
endurance, strength and speed
(Rivero & Letelier, 2000). Furthermore, because increased power
output results in a greater ability to accelerate and may increase
stride length, these training adaptations (strength rather than Muscular adaptations Physiological traits
endurance) may be important for racehorses competing over short
distances (Snow & Valberg, 1994). However enhanced power Stamina Strength Speed
through training comes with the cost of a corresponding decline in
Morphological adaptations
aerobic potential, because the increased mass of recruited fibers and
concomitant rise in ATP utilization occur simultaneously with a Muscle fiber hypertrophy − +++ +
relative inability of oxygen to diffuse into the larger fibers (Essén-
Gustavsson et al., 1989). Muscle fiber atrophy + −−− −
From a physiologic standpoint, remodeling of myofiber pheno- Increased number of capillaries +++ nc nc
type with minimal or with no hypertrophy in response to training,
produces a muscle that is much more resistant to fatigue but with Increased mitochondrial volume +++ nc nc
decreased maximal shortening velocity: the former corresponding Increased myonuclear density nc +++ nc
to each myofiber’s increased oxidative capacity, and the latter, to the
increased expression of slow MHC and other contractile protein Metabolic adaptations
isoforms (Rome et al., 1990). In a similar but reciprocal fashion to
that described for strength training, some conventional training Increased aerobic muscle +++ nc nc
programs of young racehorses produce a decrease in the size of type enzyme activities
II fibers (Ronéus et al., 1993) and a corresponding decline of both Increased glucose and fatty acid +++ nc
speed and force of contraction (Valberg, 1996). Clearly a balance transport
must be acquired at a level deemed most appropriate for the
intended use of the horse (see below). Increased muscle (glycogen) +++ nc nc
Following endurance training, exercise at submaximal intensity and its sparing during exercise
is facilitated by optimal delivery of oxygen and blood borne sub- Increased muscle (triglycerides) +++ nc nc
strates, early activation of oxidative metabolism with low utilization
of endogenous carbohydrates and increased reliance on fat oxida- Decreased post-exercise muscle +++ nc nc
tion as an energy source. The increased oxidative capacity observed (lactate)
in skeletal muscle after training occurs concurrently with increased
Increased anaerobic muscle nc nc +++
maximum oxygen uptake (Tyler et al., 1998) and a significant
enzymes
reduction in the net rate of muscle glycogenolysis and anaerobic
metabolism (Geor et al., 1999). As a consequence, in the trained Decreased anaerobic muscle nc nc −−−
state the speed at which a horse begins to accumulate lactate enzymes
increases gradually (e.g. there is a delay in the onset of lactate accu-
mulation and ATP depletion; Ronéus et al., 1994; Valberg, 1996; Increased muscle (high-energy nc nc +++
Rivero, 1997). This is accompanied by enhanced muscle buffering phosphate)
capacity and more efficient excitation–contraction coupling. Hence Increased muscle buffering + nc +++
collectively, endurance may be enhanced by a wide variety of related capacity
factors that delay the onset of fatigue during anaerobic exercise and,
although relative sparing of muscle glycogen underlies the delay in Contractile adaptations
fatigue onset during this type of exercise in the trained state, it is
Unidirectional IIX→IIA→I fiber + nc −−−
likely that all the metabolic adaptations summarized above combine
type (MyHC isoform) transition
to increase endurance.
Although controversial then, it is clear from extensive scientific Bidirectional IIX→IIA←I fiber + nc −−−
literature that equine skeletal muscle has considerable potential to type (MyHC isoform) transition
adapt during training, and overall, that these adaptations have
important functional implications affecting stamina, strength and Increase of IIA : IIX fiber type + nc −−−
speed (Table 13.2). Ideally therefore, conditioning programs of ath- (MyHC isoform) ratio
letic horses should promote characteristics that optimize equilib- Increase of I : IIA fiber type + nc −−−
rium between these physiological traits (Snow & Valberg, 1994; (MyHC isoform) ratio
Rivero & Piercy, 2008). While interpreting Table 13.2, first note that
certain changes occur together, whereas others occur independently. Increase of IIA : I fiber type − nc +++
For example reduced intrinsic muscle shortening velocity occurs (MyHC isoform) ratio
together with increased endurance capacity following IIX → IIA
The symbols +++ and −−− indicate primary implication (positive and negative,
fiber transition, whereas increased aerobic muscle enzyme activities respectively) towards the particular characteristic; + and −, secondary implication
are never associated with decreased mitochondrial volume. Second, (positive and negative, respectively); nc: no contribution to the particular
note that certain changes precede others: for instance, aerobic meta- characteristic.
bolic adaptation occurs prior to any structural adaptation. Thus
significant increases in muscle glycogen concentration occur after
only 10 consecutive days of training (Geor et al., 1999), whilst a then, occur in the first 3–4 months: prolonging training beyond this
significant degree of type II–type I fiber type conversion needs much period, although improving aerobic capacity, reduces anaerobic
longer (8 months; Serrano et al., 2000). Third, adaptations are capacity and has no effect on strength. Simultaneously, the risk of
cumulative and dose-dependent, but there is an upper limit above overtraining increases. Finally, certain adaptations to training are
which further adaptations do not occur, even when the stimulus reversible, tending to return to the pre-training situation when the
continues (Tyler et al., 1998). Most relevant training adaptations stimulus is stopped.

289
 13 The response of musculoskeletal tissues to exercise

As stated earlier, the response of equine muscle to training very high (~165% of VO2max) intensity and short duration or dis-
depends on two sets of factors: 1) the basal status of the muscle, tance (1.6–5.3 min; 1 600–3 600 m) 5 days/week for 16 weeks
(determined by the breed, age, sex and level of fitness of the (Yamano et al., 2002).
horse) and 2) the stimulus applied (e.g. type, intensity, duration, In general though, it seems clear that improved stamina through
frequency and volume of the training exercise). Unfortunately, enhanced aerobic capacity is the most common response of equine
little is known about the relative influence of most of these factors. skeletal muscle to training, regardless of either the basal status of
For example, only a few studies compare muscular adaptations the muscle or the training-exercise program. Given that many
that occur with different training programs. Of these, two studies equestrian disciplines, including Thoroughbred racing, are largely
(with contradictory results), examine the influence of different aerobic in nature, this underlies the likely benefit of training to the
exercise intensity (Sinha et al., 1993; Eaton et al., 1999), and only equine athlete.
one study has examined the combined influence of intensity and
duration (Gansen et al., 1999). This latter study concluded that
low intensity exercise (~50% of V4) for long duration (45 min) Tendons and ligaments
was, after 6 weeks, more effective in improving aerobic capacity
exercise than high-intensity exercise (~100% of V4) of moderate
duration (25 min). Recent years, however, have seen the publica-
Introduction
tion of several experimental studies with well-documented exer- A great proportion of locomotor injuries in equine athletes involve
cise protocols (Table 13.3). From these it seems that moderate to the flexor tendons, accessory ligaments and the suspensory liga-
high intensity (~80–100% of VO2max) exercise of short duration ment. The injuries occur both in training and during athletic com-
(5–10 min) improves both stamina and strength in racehorses fol- petition, resulting in significant wastage, with consequent financial
lowing 12–16 weeks of training (Tyler et al., 1998; Miyata et al., and welfare implications (Rossdale et al., 1985; Williams et al.,
1999), whereas anaerobic capacity, can only be increased in the 2001).
short to mid term (up to 16 weeks), by introducing supramaximal The extent of clinical pathological change can range from a minor
intensity exercise (~100–150% of either VO2max or V4) of short disruption of a small number of fibrils to complete catastrophic
(2 min) to moderate (15 min) duration (Rivero et al., 2002; rupture of the tendon. However, the last event is unlikely to occur
Yamano et al., 2002; Eto et al., 2004). Interestingly, adaptation without preceding preclinical injury and changes to the matrix at a
occurs more readily in younger (~2 year old) than in mature race- molecular and ultrastructural level occur before clinical injury is
horses, e.g. muscular adaptations compatible with a combination evident, making tendon damage from frequent cumulative micro-
of improved stamina, strength and speed were reported in young trauma into the most frequent cause for tendon injury (Kannus &
Thoroughbreds following exercise of high (~100% of VO2max) to Josza, 1991; Selvanetti et al., 1997).

Table 13.3  Physiological implications of muscular adaptations in various training programs scientifically evaluated in horses

Horses Conditioning parameters Functional significance Reference

Breed Age Intensity Duration Frequency Volume Stamina Strength Speed
(distance)
Thoroughbred 4–8 years ~55% VO2max 60 min Daily 10 days + Ni − Geor et al.,
(~13–14 km) 1999
Thoroughbred 2–9 years ~80% VO2max 3 min 6 days/week 6 weeks + Ni − Sinha et al.,
(1500 m) × 2 1993
Thoroughbred 5–7 years ~80–100% 5 min (× 2) 5 days/week 12 weeks Ne Ni ++ Eto et al., 2004
VO2max
Thoroughbred 2–3 years ~100–165% 1.6–5.3 min 5 days/week 16 weeks ++ + − Miyata et al.,
VO2max (1600–3200 m) 1999
Thoroughbred 2 years ~100–165% 1.6–5.3 min 5 days/week 16 weeks ++ + + Yamano et al.,
VO2max (1600–3200 m) 2002
Standardbred 3–5 years ~60–100% 6–12.5 min 5 days/week 16 weeks ++ ++ Ne Tyler et al.,
VO2max (~3–9 km) 1998
32 weeks ++ + −
Standardbred 2 years ~100–140% V4 15 min nd
2 day 5 weeks + − + Rivero et al.,
nd 2002
~65% V4 60–90 min 2 day
Arab 8.6 years ~80% V4 50–80 min 3 days/week 12 weeks + +++ − D’Angelis
(~10–20 km) et al., 2005
Andalusian ~4 years ~25–30% V4 45–60 min 5 days/week 12 weeks + + − Serrano et al.,
2000
~50–60% V4 75–120 min 32 weeks +++ + −

Intensity is expressed as a fraction of either VO2max (velocity at maximal aerobic capacity) or V4 (velocity inducing a blood lactate concentration of 4 mmol/L). The symbols + and
− indicate that either the muscular adaptations to training had a positive or negative effect respectively towards the particular characteristic; the number of symbols is
proportional to the magnitude of the adaptation; Ne, no effect; Ni, not investigated.

290
 Tendons and ligaments

Lesions in the superficial digital flexor tendon (SDFT) of the collagen and minimal concentrations of several other types can be
horse are very similar to injuries seen in the human Achilles tendon, found (Riley, 2004). The basic structural unit of collagen is tropo-
in the quadriceps tendons of jumping athletes and also the degen- collagen, a triple-helix of polypeptide chains. Five tropocollagen
erative changes occurring in the rotator cuff of the shoulder. Thus, molecules pack together forming collagen microfibrils, which are
the horse may provide a natural model of tendon degeneration organized into the larger structural units of the tendon: fibrils,
from which the underlying biological mechanisms and pathologi- fibers, sub-fascicles (primary fiber bundle), fascicles (secondary
cal processes of a number of tendon and ligament lesions, resulting fiber bundles), tertiary fiber bundles and the tendon itself (Kastelic
from both exercise and aging, may be elucidated. et al., 1978; Wang, 2006) (Fig. 13.20).
Recovery from tendon injury is protracted and incomplete; once In the relaxed state the fascicles show a planar wave form or
injured, the tendon is permanently compromised to a greater or ‘crimp’, which can be well demonstrated using plane polarized light
lesser extent (Watkins, 1999). Treatments are numerous and, microscopy (Gathercole & Keller, 1991). It is the crimp that, by
although recent cell-based therapies have shown encouraging flattening out, causes the initial, nonlinear load/deformation
results (Smith, 2008), not one approach thus far has resulted in behavior of the tendon. Intra- and interfibril collagen cross-links
completely restoring the functional capacity of an injured tendon. can be formed after enzymatic transformation, such as hydroxyly-
Rehabilitation is thought to be one of the most important compo- sylpyridinoline (HP) and lysylpyridinoline (LP) cross-links or via
nents of the treatment regimen. Consequently, an appreciation of non-enzymatic glycation (NEG), of which pentosidine cross-links
the mechanisms involved in the response of normal tendon to are the best-known examples. Cross-linking is essential for the sta-
physical training could have implications for developing rehabilita- bility of collagen fibrils and HP cross-links contribute significantly
tion strategies in the management of tendon injuries. to the biomechanical properties of a tendon (Parry, 1988; Tsuzaki
In most competition- and racehorses there is a high predisposi- et al., 1993). The level of NEG products is closely related to the
tion to injury in the SDFT and, to some extent, the suspensory liga- remodeling rate of tendon tissue. NEG products are formed in a
ment of the forelimb. In dressage horses many injuries in the time-dependent fashion via the so-called Maillard reaction (Sell
suspensory ligament of the hind limb are seen. In addition to bio- et al., 1991; Vlassara et al., 1994). In metabolically active tissues
mechanical predisposition resulting from individual conformation, they are broken down in the process of tissue turnover and thus
external mechanical influences such as track surface contribute to remain at a constant level. However, in tissues with a very low
injury risk. In some racetracks in Japan a change in track surface has turnover rate they will accumulate and can therefore be used as an
led to a great increase in tendon injury, while reducing the influence estimate of tissue age (Verzijl et al., 2000).
of bone fractures. The major non-collagenous constituents of tendons are water
The methods used in training to condition tendons are for the (~65% of wet weight) and proteoglycans (PGs, ~1% of dry weight).
most part based on a traditional empirical approach. Trainers often PGs are composed of a core protein to which one or more sulphated
describe the effect of training as ‘hardening’ the tendons, yet changes glycosaminoglycans (GAGs) are covalently attached. The strongly
do not appear to be related to biological events within the tendon negatively charged and therefore hydrophilic GAGs attract water
structure. The response of tendons to load and thus to training may molecules and the collagen network prevents the tissue from
also differ between specific tendons that have different biological expanding beyond its limits, thus giving the tissue an intrinsic
functions (Birch et al., 1997a; Goodman et al., 2004). The inci- tension and its characteristic viscoelastic properties (Scott, 1995).
dence of injury being related to particular tendons suggests that the Besides this direct effect PGs are also known to have an effect on
predisposition is associated with a specific function. The extensor fibrillogenesis and matrix architecture (Iozzo, 1998), thus indirectly
tendons, together with the deep digital flexor tendons (DDFT) have, influencing tensile strength (Garg et al., 1989). The content and
in general, a low incidence of failure in equine athletes (apart from types of proteoglycans vary between biomechanically differently
the navicular area of the DDFT that has received much attention in loaded areas within the SDFT (Micklethwaite et al., 1999).
recent years because of the advent of sophisticated diagnostic Fibroblast-like cells account for 90–95% of all cells in tendons.
methods such as high-field magnetic resonance imaging (MRI) and They have been histologically classified as tenoblasts and tenocytes
contrast-enhanced computed tomography (CECT) (Dyson & and are important for the production, organization and mainte-
Murray, 2007; Puchalski et al., 2007), whereas the SDFT and the nance of the extracellular matrix (Kannus, 2000). Tenoblasts are
suspensory ligament have a very high incidence of injury. immature, ovoid to spindle-shaped tendon cells, with a high meta-
With exercise, and particularly elite athletic performance, the dif- bolic activity. As they mature, tenoblasts transform into tenocytes
ferent functional requirements will result in the need for changes with a more flattened, slender appearance. Tenocytes show a lower
in the composition and structure of tendons. It is this relationship metabolic activity, due to their lower nucleus-to-cytoplasm ratio,
and the mechanisms controlling it that are important in condition- than tenoblasts (Kannus, 2000). Tenocytes are able to adjust their
ing horses for peak performance while minimizing injury risk. In shape and metabolic activity with respect to the production of
the following part first a brief overview of structure and function of extracellular matrix components (Robertson, 1994). Different types
tendons will be given with special attention to the reaction of loading (compression, tension) elicit different types of metabolic
of tendons to mechanical challenge. After that, sections will follow response (Arnoczky et al., 2007; Gillard et al., 1979). The whole
on the response of tendon tissue in juvenile and older horses to tendon is surrounded by the paratenon, a layer of loose connective
imposed exercise. tissue that allows movement of the tendon in relation to the sur-
rounding tissues. The tendon itself is encapsulated in a fine connec-
tive tissue sheath called epitenon that contains vascular, lymphatic
and nerve supplies to the tendon. This layer penetrates deep into
Tendon composition and structure the tendon as the endotenon covering the tertiary fiber bundles
Tendons consist of a relatively small cellular fraction and a corre- (Kastelic et al., 1978). Tendons can be covered by a synovial sheath
spondingly large extracellular matrix (ECM), which is composed of in areas where the mechanical loading is not parallel to the longi-
densely packed and hierarchically arranged collagen filaments, tudinal axis of the tendon (Jozsa & Kannus, 1997).
embedded in a hydrophilic, proteoglycan rich matrix that provides Blood vessels enter tendons at three main sites: the so-called
tendons with characteristic viscoelastic properties (Goodship et al., intrinsic system at both the myotendinous junction and the osteo-
1994). Collagen is the main component of the ECM (60–85% of tendinous insertion and the extrinsic system via the paratenon or
the dry mass of a tendon), of which type I is predominantly present, the vincula within synovial sheaths (Carr & Norris, 1989). The ratio
accounting for 95% of the total collagen content (von der Mark, of blood supply from these sources differs between different
1981; Williams et al., 1980). Collagen type III represents 3% of total tendons, but the extrinsic route is usually the most important,

291
 13 The response of musculoskeletal tissues to exercise

Primary fiber Secondary

Collagen Collagen Collagen bundle fiber bundle Tertiary
molecule fibril fiber (Subfascicle) fascicle) fiber bundle Tendon

Fibroblast

Endolenon Paratenon

Crinp wave

Fig 13.20  Schematic diagram representing the hierarchical structure of tendon.

Modified from Wang et al. (2006).

especially in areas distal from the myotendinous junction and close depicted in a so-called stress–strain curve (Fig. 13.22) (Jozsa &
to the insertion on the bone (Carr & Norris, 1989). The blood Kannus, 1997). This curve has four distinct regions. The initial
supply of tendons, which is already scarce, decreases further with concave portion of the curve (toe region) (I) represents the straight-
increasing age and mechanical loading (Astrom, 2000). ening of the crimp wave, where minor loading results in a consider-
able elongation until about 3% strain (Evans & Barbenel, 1975).
Beyond this point, tendons deform in a linear fashion as a result of
Tendon biomechanics sliding of collagen molecules and fibrils and the fibers become
During the stance phase of the stride cycle flexor tendons are loaded arranged in a more parallel fashion. If the strains do not exceed
under tensile stress with loads that may attain easily two times body 4–5% the tendon behaves in an elastic fashion and changes are
weight, depending on the gait (Schryver et al., 1978). The tendons still completely reversible. The gradient of this almost linearly
are able to store much of the kinetic energy when they are loaded elastic part of the curve (II) represents the tensile stiffness or elastic
during the first half of the stance phase as elastic energy and release modulus (EM) of the tendon (EM = Δstress/Δstrain). In the third
it during the second half of the stance phase where it aids in propul- region (III) higher strains, up to 8%, will result in dissociation of
sion. In fact, flexor tendons act basically as energy saving elastic fibrils due to interfibrillar bond failure, which causes a plastic defor-
springs, while the main function of the flexor muscles is to damp mation and microscopic failure (Evans & Barbenel, 1975; Jozsa &
this spring-like action (Wilson et al., 2001). This energy saving role Kannus, 1997). With further increasing strains larger tendon
of tendons is also utilized in other species and is perhaps most bundles start to fail and macroscopic failure occurs (region IV)
evident in the kangaroo, which is able to increase speed without an (Jozsa & Kannus, 1997). The strain at which each phase occurs
associated increased oxygen uptake (Morgan et al., 1978), by utiliz- differs in functionally distinct tendons; tendons with a spring-like
ing the large Achilles tendons as springs. function experience higher strains before plastic deformation
Optimization for this role results in very low failure safety occurs.
margins compared with other less mass-critical structures of the Cyclical loading of a tendon, to a level not reaching yield, in a
musculoskeletal system; consequently, at peak performance levels, materials testing machine over a period of time shows two phenom-
functional strains are close to failure strains (Fig. 13.21). Functional ena that are relevant to the effects of exercise in the short term. First,
tensile strains in equine flexor tendons have been recorded in vivo. as shown in Figure 13.23, there is a shift of the load–deformation
Tensile strains measured were approximately 3% at the walk, 6–8% curve to the right with each successive loading cycle until eventually
at the trot and 12–16% at the gallop (Stephens et al., 1989). Labo- a steady state is reached. This is known as ‘preconditioning’ and,
ratory testing showed that the SDFT of horses will rupture at a level although it is a feature seen in laboratory testing of tendon; it may
of 12–16% in the Thoroughbred (Wilson, 1991) and at approxi- have in vivo relevance in relation to ‘warming up’ prior to peak
mately 12.5% in the Warmblood (Riemersma & Schamhardt, athletic performance. Secondly, as illustrated, the plot of loading
1985). followed by unloading of the tendon forms a loop, starting and
The relationship between load (stress) and deformation (strain) finishing at the same point. The loop, termed hysteresis, is formed
of elastic structures, such as flexor tendons can be graphically as a result of the elastic recoverability of the tendon being less than

292
 Tendons and ligaments

Fig 13.21  Flexor tendons are heavily loaded

during high-level equestrian activities.

Toe Linear Microscopic Macroscopic 12

Cyclic conditioning of tendon
region region failure failure Rupture
Equine SDFT
10
Force (kN) (thousands)

6
Stress (N/mm2)

0
0 2 4 6 8 10 12 14 16 18 20
% Strain

0 3 6 9 12 Fig 13.23  Cyclical loading of a tendon within the elastic range. After the
Strain (%) first cycle, the tendon will be longer at zero force than originally. In the
following cycles this phenomenon repeats, but to a much smaller extent
Fig 13.22  Stress–strain curve; the relationship between stress generated by until a steady state is reached. This is called preconditioning. The loops are
a certain force applied to a tendon and the resulting elongation (strain). open because of hysteresis: more energy is required to stretch the tendon
(left arm of the loop) than is released when relaxing (right arm). The deficit
is dissipated as heat.
100%. The area of the loop represents loss of energy, largely in the
form of heat, which occurs during stretch and release of the tendon.
It has been shown that the internal temperature of the SDFT may
rise to 45–47°C during 7–10 min of peak performance (Wilson & occur. Furthermore tendon injury will lead to significant local
Goodship, 1994). Although this level and duration of hyperthermia changes in tissue properties and thus in non-homogeneous bio­
would normally result in the death of most cell types, tenocytes mechanical behavior.
within the inner core of the SDFT are able to withstand much higher
temperatures than most other fibroblast-like cells (Birch et al.,
1997b). It is possible, however, that hyperthermia, although being
Effects of exercise
not lethal for these cells, still plays a role in the initiation of degen- Tendon tissue is responsive to loading, giving the biomechanical
erative processes as often seen in the core region of the SDFT. environment an important role in the determination of tendon
It is clear that the biomechanical characteristics of the tendon are composition and structure. This is best shown by the regional dif-
determined by tendon composition and ultra-structure, as outlined ferences in molecular composition of the SDFT. The mid-metacarpal
above. As long as the tissue is homogeneous, deformation (strain) area, which is exclusively subjected to tensional loads, has a high
will occur evenly. Riemersma and Schamhardt (1985) showed that collagen content and typically has a resident population of slender,
the strain at different levels of the SDFT is uniform, despite differ- elongated tenocytes, whereas the area that wraps around the fetlock
ences in cross-sectional area. As a function of both age and exercise and sustains compressive loads in addition to tensile forces, con-
changes in matrix composition and associated functional properties tains relatively less collagen, more glycosaminoglycans and rounded

293
 13 The response of musculoskeletal tissues to exercise

tenocytes, taking a more chondrocyte-like appearance (Lin et al., Comparison of the stress–strain curves
2005a). Also, the content of cartilage oligomeric matrix protein of central and peripheral fibers
(COMP) is higher in this region. Cells in tensional regions of
tendons thus synthesize predominantly collagen, whereas cells in C P
regions of tendons that wrap around bone produce collagen type II
and other matrix components typical of the fibrocartilaginous F/CSA
matrix seen in these areas of compression. Interestingly, this appears

Stress
to be a dynamic and reversible process. In a study where the com-
pression region of the deep digital flexor tendon was released from
compression and subjected to tensile forces, the nature changed F/CSA
from fibrocartilaginous to fibrous type I collagen (Gillard et al.,
1979).
The horse is a precocial animal that loads its musculoskeletal
system within a few hours after birth and will remain doing so
throughout its life. This, together with the responsiveness of tendon
tissue to loading, makes it difficult, if not impossible, to discrimi-
nate between the effects of growth, maturation and eventually
ageing, and exercise, especially with respect to the energy-storing DL% Strain
tendons and ligaments that are the subject of most studies, as they
Fig 13.24  A load/deformation plot of collagen fibers from central and
sustain by far most injuries and in fact are the only ones that are
peripheral regions of the superficial; digital flexor tendon in an old horse,
clinically relevant. However, there have been a number of studies
illustrating the effect of a lower toe limit strain on the greater increase in
focusing specifically at ageing or the effect of well-controlled exer- stress with extension experienced by the central fibers.
cise protocols that have increased our insight in the complex inter- Reproduced from Wilmink et al. (1992).
action of the equine tendon with its biomechanical environment.

The effects of ageing on tendon tissue

During embryonic development the collagen fibrils of the tendon foal with a fractured limb, COMP levels were shown to have
that are formed are small in size (diameter approximately 40 nm) increased dramatically in the sound limb that was loaded more
and all have a more or less similar appearance. This is called a than normal, but not in the unloaded fractured limb that was
unimodal distribution. During maturation this distribution changes immobilized in a cast (Smith et al., 1997).
into a bimodal one through the development of larger size fibrils There are also changes in crimp angle with age and these depend
(diameter up to 250 nm). Although there may be differences in the on the location within the tendon. In older horses, particularly with
timing of this process between precocial species, such as the horse, some history of exercise, the crimp characteristics were significantly
and altricial species, such as the rat and man, this appears to be a different between peripheral and central core regions of the tendon.
universal mechanism that has been demonstrated in all mammals In older horses, a reduction in crimp angle in the core region can
that have been examined (Parry et al., 1978) and is thought to be be observed that makes stress increase at a greater rate than in
an adaptation to biomechanical challenges in the postnatal envi- peripheral fibers (Wilmink et al., 1992) (Fig. 13.24). This seems to
ronment. The fibril distribution pattern can be quantified in terms be a naturally occurring phenomenon, as similar age-related
of collagen fibril index and mass average diameter (Flint et al., changes have been observed in wild horses subjected to ‘natural’
1984) and has been shown to reach mature levels in the Thorough- levels of exercise (Patterson-Kane et al., 1997a) and might be a
bred by 2 years of age (Patterson-Kane et al., 1997b). The changes (partial) explanation for the higher incidence of central core lesions
in fibril profile may be related to changing GAG or COMP levels. than of peripheral lesions in the equine SDFT. In an in vitro study
Levels of these molecules change during development (Cherd- using tendon explants, Dudhia et al. (2007) showed that tissue
chutham et al., 1999) and both have been implicated in the regula- from older individuals (10–30 years) was more susceptible to weak-
tion of fibrillogenesis (Parry et al., 1982; Smith et al., 2002). In fact, ening by cyclical loading than specimens from younger individuals.
all biochemical components, including collagen and DNA as a This effect was possibly due to increased MMP activity and may be
measure for cellularity, have been shown to change significantly related to the age-related changes alluded to above.
during early life, also in the absence of specific exercise regimens
(Cherdchutham et al., 1999). A study using material from horses
aged 0, 5, 12 and 36 months showed that changes in matrix com-
The effects of exercise on tendon tissue
position occur fastest in the early juvenile period. Interestingly, the There seem to be three major factors that determine the effect of
study also showed that, in contrast to articular cartilage, tendon exercise on tendons. These are the type of tendon, the age at which
tissue is not biochemically ‘blank’ (e.g. homogeneous) with respect the exercise is given and the level of background exercise.
to region (tension or compression) at birth, possibly due to intra- An increase in biomechanical loading in the form of the imposi-
uterine movements (Lin et al., 2005b). In another study it was tion of an exercise regimen will elicit an adaptive response in most
shown that the levels of the gap junction proteins Connexin 43 and musculoskeletal tissues, resulting in increased tissue mass and
32 (Cx43; Cx32) were significantly higher in fetal tendons than in related strength. This general principle seems to apply to ligaments
tendons from other age classes (Stanley et al., 2007). Gap junctions and extensor tendons (Woo et al., 1982, 1987; Newton et al.,
play a role in intracellular communication, which may correlate 1995), but not or to a much less extent to flexor tendons (Birch
with synthetic responsiveness that is known to be highest in very et al., 1999), probably because a substantial increase in the mass
young animals. Cartilage oligomeric matrix protein (COMP) seems of energy-storing tendons will increase the stiffness of the structure
to be of specific interest with respect to its sensitivity regarding and hence hamper its energy conservation function (Smith & Good-
biomechanical loading. In the tensile regions of tendons levels rise ship, 2008). If substantial changes in cross-sectional area (CSA) of
from birth to peak at around 2 years of age, after which they decline. the SDFT are seen, this is the case in (very) young animals, which
However, in the compression region of the tendon COMP levels seem to be more adaptive than mature individuals. Some degree of
plateau at about 5–7 years of age after their initial increase and exercise seems to be necessary for the proper functioning of tendons,
remain stable after that (Smith et al., 1999). In a single case of a as there is growing evidence that stress deprivation may be as

294
 Tendons and ligaments

Fig 13.25  (A) Electron micrograph of biopsy from a foal in the pasture group at age 2 months. Note the predominance of thick fibrils. (B) Electron
micrograph of peripheral sample from the same foal as in Fig. 13.25B after euthanasia at age 5 months. There is an increase in small-diameter fibrils.
(C) Electron micrograph of biopsy from a foal in the pasture group after euthanasia at age 11 months. The number of small fibrils has further increased.
This appeared to be not the case in both other groups, indicating that additional exercise did not lead to full restoration of the normal situation.
Reproduced from Cherdchutham et al. (2001b).

deleterious for tendon function as overloading (Arnoczky et al., (CDET) were increased in the trained horses compared to the con-
2007). In the case of young animals, the amount of this background trols (by 8 and 16% respectively). In the SDFT (but not the CDET)
exercise may be an important factor determining the effect of this increase in size was accompanied by a decrease in tissue density.
the exercise regimen. Because no evidence of pathology could be detected, this response
The largest exercise effect on SDFT CSA has been measured in the was seen as adaptive. There were no changes in mechanical proper-
EXOC study that has been referred to earlier (van Weeren & ties of the tendon (maximal load and stress at failure) (Firth et al.,
Barneveld, 1999a). In this study pasture exercise was compared to 2004b).
box rest and box rest with additional sprint exercise in 0–5 month- In the long-term Bristol study young Thoroughbreds were exer-
old foals. The CSA of the SDFT of foals kept at pasture was signifi- cised for 18 months on a treadmill and compared to controls
cantly greater (by a mean of almost 50%) than of the confined foals undergoing a much less intensive exercise regimen. There were no
and of the foals that were confined and sprinted for a few minutes differences in collagen content or CSA of the SDFT, but there was a
per day (Cherdchutham et al., 2001a). This means that the back- distinct difference in collagen fibril diameter parameters with col-
ground loading as provided by free pasture exercise is an important lagen fibril mass average diameter in the central core region of the
factor in tendon development in the juvenile animal. This conclu- trained animals lower than in the control horses (Patterson-Kane
sion was supported by a study on collagen fibril characteristics in et al., 1997c). By determining the level of glycosylation it was
the same foals. The change from a unimodal to a bimodal fibril evident that the collagen was not newly formed, but a degradation
diameter distribution was most advanced at age 5 months in the of larger fibrils. Further, mean collagen crimp angle and crimp
pastured foals and least in the box-rested foals. Additional exercise length were less in the exercised group (Patterson-Kane et al.,
from 5–11 months led to only partial recovery (Cherdchutham 1997d, 1998). In contrast to the findings in the SDFT, no exercise-
et al., 2001b) (Fig. 13.25). related changes in collagen mass average diameter or collagen fibril
Further support came from a study in which Thoroughbred foals index were found in the CDET of these same horses. This observa-
were exercised on a treadmill from age 2 to 15 months where there tion supported the concept of the functionally distinct nature of the
was a greater rate of increase in SDFT CSA in the trained animals CDET and SDFT, resulting in fundamentally different responses to
compared to the controls, but no significant difference at the end high-speed exercise (Edwards et al., 2005). It was initially suggested
of the study due to high variance (Kasashima et al., 2002). Bio- that the changes in the SDFT might not be adaptive in nature, but
chemically, COMP was significantly increased in the positional more indicative of micro-trauma and thus detrimental to tendon
CDET, but not in the energy-storing SDFT. Other matrix compo- function (Birch et al., 1998). In later work the same research group
nents and mechanical properties were not significantly changed performed detailed biomechanical and biochemical analyses of
(Kasashima et al., 2008). In that study foals did not have continu- both CDET and SDFT tissue specimens from animals pertaining to
ous access to pasture and the exercise effect was against a back- the long-term Bristol study in which no signs of degeneration were
ground of ‘partial’ confinement. found. They concluded that high-intensity long-term exercise on
In the GERA study that has been mentioned earlier (Rogers et al., skeletally mature individuals results in changes that suggest acceler-
2008a,b) there was no difference in SDFT CSA between the trained ated aging in the SDFT and adaptation in the CDET (Birch et al.,
CONDEX and the untrained (but pasture exercised) PASTEX 2008).
animals after an experimental period lasting from age 3 weeks to
age 18 months, although there was a trend (p = 0.058) towards a
larger CSA in the CONDEX group (Moffat et al., 2008).
Conclusion
Whereas there are no data on the effects of exercise on tendons The adaptive response of tendon tissue to exercise is limited. It is
of old horses, various experiments have investigated the exercise clear that the possibility to exercise is crucial to young foals, as lack
effect on tendon characteristics of young-mature horses, often of exercise in juvenile animals may lead to the development of
2-year-old Thoroughbreds. biomechanically inferior tendons, but little seems to be gained by
In the MUGES study alluded to earlier (Firth et al., 2004a) CSA imposing additional training above a baseline of free pasture exer-
of both the SDFT and the Common Digital Extensor Tendon cise. The time window in early life in which the constitution of

295
 13 The response of musculoskeletal tissues to exercise

tendons can be influenced by exercise is short. Already at an age of musculo­skeletal system (in the mature individual) as a fact, the
11 months the juvenile tendon has biomechanical properties that question becomes, what would the optimal exercise load be? It has
are similar to those of mature animals (Cherdchutham et al., been shown that the amount of voluntary exercise in foals raised
2001). This supports the opinion that early competition at young with 24 h/day access to pasture is surprisingly similar to what has
age is not detrimental to horses (Smith et al., 1999). In young- been observed in foals from entirely free living populations (Kurvers
mature animals the adaptive response of flexor tendons is minimal. et al., 2006). This, together with the fact that in the EXOC study
In the MUGES study a small increase in CSA was observed, but mentioned earlier the pasture exercised foals developed much better
together with a decrease in tissue density and without improve- than the foals in the other exercise groups, which were deprived of
ment of biomechanical characteristics. Further, the division line exercise or were subjected to a combination of box rest with bouts
between adaptive response, if any, and degenerative effects is of high-intensity exercise, makes it safe to proclaim that, based on
extremely thin. This can be explained by the facts that flexor present-day knowledge, free exercise for 24 h/day is the baseline or
tendons are known to operate very close to their physiological limit gold standard for growing foals. This does not mean, however, that
during strenuous athletic activity and that, as energy-storing struc- this exercise level is by definition optimal for raising all foals, irre-
tures, flexor tendons cannot adapt by a simple increase in size (and spective of future use. There is now evidence that the development
thus strength) as other tendons and ligaments and most other of the musculoskeletal system can be manipulated beyond the
tissues of the musculoskeletal system would do. The accurate deter- effects produced by ‘natural’ exercise through the imposition of
mination of the delicate balance between the training load that is additional exercise in the young animal. In the case of articular
needed for an optimal preparation of the horse for competition cartilage such intervention led to advancement (and earlier
and the load that can be sustained by the SDFT and suspensory cessation) of the natural process of functional adaptation (van
ligament, as perhaps the most vulnerable structures of the equine Weeren et al., 2008). No immediate clinical effects were seen
musculoskeletal system, will remain the best prevention against in that study, which led to the conclusion that the imposed exercise
tendon injury and the greatest challenge for the trainer. This level was safe (Rogers et al., 2008a). However, the long-term
endeavor would be greatly facilitated if suitable biomarkers or effect on morbidity in the later athletic career could not be estab-
hyper-sensitive quantitative ultrasonographic techniques would lished due to low numbers (Rogers et al., 2008b). Later work
become available that would permit the timely detection of immi- strongly suggests an overall positive influence of early exercise: epi-
nent tendon injury. demiological data from Thoroughbreds have shown that horses in
training or racing as 2-year-olds had better musculoskeletal health
throughout life than horses starting their career later (Tanner et al.,
2013).
General conclusion In the mature horse attention has focused on the adaptive
response provoked by exercise and the thin borderline between
The musculoskeletal system is the organ system that generates loco- these adaptive responses and early damage. No work has been done
motion and is, therefore, together with the cardiovascular and respi- on the establishment of a baseline workload needed for optimal
ratory systems, likely to be influenced by the exercise regimen an maintenance of the musculoskeletal system. It is the question
individual is subjected to. The actual response of the system is whether this lack of interest is justified given the increasing insight
complex, however, and depends on the type of tissue and in some that lack of exercise may have an adverse effect on some structures
cases even on the specific structure involved. Age is a crucial factor, in mature individuals (Arnoczky et al., 2007) and the deleterious
as exercise may have a different effect on the young, growing animal effect of a combination of exercise restriction and bouts of heavy,
than in the mature individual. There is enough evidence now to high-intensity loading, an exercise regimen that is all too common
state that physical exercise in young individuals is pivotal for correct for many sports and leisure horses.
development of the musculoskeletal system, which is of great It can be concluded that exercise is a must for both the correct
importance for those tissues known not to remodel or to show development and maintenance of the musculoskeletal system.
minimal remodeling at adult age (and related to this have a very Restriction of space and of time spent on them means that many
limited healing capacity after injury), such as articular cartilage and horses kept under modern management conditions most probably
possibly also tendons (van Weeren et al., 2010; Rogers et al., 2012). exercise too little rather than too much. Apart from a severe impact
However, there is evidence that also in bone, a tissue that is known on well-being, this situation is likely to also affect musculoskeletal
to respond to exercise and to retain its healing ability throughout morbidity rates. Meticulous monitoring and administration of
life, exercise-induced structural changes may be retained even long workload and wastage due to illness or injury in the relevant eques-
after the cessation of exercise (Karlsson, 2007), possibly because of trian disciplines in large cohorts of horses over prolonged periods
an effect on the underlying collagen skeleton. of time and adequate epidemiological processing of these data is
Taking the need for exercise for either correct development the way forward to establish optimal training and conditioning
(in the young animal) or for adequate maintenance of the protocols for sport horses.

References
Armstrong, R.B., Essén-Gustavsson, B., Arokoski, J., Kiviranta, I., Jurvelin, J., Tammi, II. Proteoglycan metabolism. Connect.
Hoppeler, H., et al., 1992. O2 delivery at M., Helminen, H.J., 1993. Long-distance Tissue Res. 18, 223–234.
VO2max and oxidative capacity in muscles running causes site-dependent decrease of Aydelotte, M.B., Kuettner, K.E., 1988.
of standardbred horses. J. Appl. Physiol. cartilage glycosaminoglycan content in the Differences between sub-populations of
73, 2274–2282. knee joints of beagle dogs. Arthritis cultured bovine articular chondrocytes. I.
Arnoczky S.P., Lavagnino, M., Egerbacher, M., Rheum. 36, 1451–1459. Morphology and cartilage matrix
Caballero, O., Gardner, K., 2007. Matrix Astrom M., 2000. Laser doppler flowmetry in production. Connect. Tissue Res. 18,
metalloproteinase inhibitors prevent a the assessment of tendon blood flow. 205–222.
decrease in the mechanical properties of Scand. J. Med. Sci. Sports 10, 365–367. Barneveld, A., van Weeren, P.R., 1999.
stress-deprived tendons: an in vitro Aydelotte, M.B., Greenhill, R.R., Kuettner, K.E., Conclusions regarding the influence of
experimental study. Am. J. Sports Med. 35, 1988. Differences between sub-populations exercise on the development of the equine
763–769. of cultured bovine articular chondrocytes. musculoskeletal system with special

296
 References

reference to osteochondrosis. Equine Vet. J. Booth, F.W., Baldwin, K.M., 1996. Muscle Brama, P.A., Tekoppele, J.M., Bank, R.A., van
Suppl. 31, 112–119. plasticity: energy demand and supply Weeren, P.R., Barneveld, A., 1999a.
Barrey, B., Valette, J.P., Jouglin, M., et al., 1999. processes. In: Rowel, L.B., Shepherd, J.T. Influence of different exercise levels and
Heritability of percentage of fast myosin (Eds.), Handbook of physiology. American age on the biochemical characteristics of
heavy chains in skeletal muscles and Physiological Society, Bethesda, MD, pp. immature equine articular cartilage. Equine
relationship with performance. Equine Vet. 1075–1123. Vet. J. Suppl. 31, 55–61.
J. Suppl. 30, 289–292. Boston, R.C., Nunamaker, D.M., 2000. Gait Brama, P.A., Tekoppele, J.M., Bank, R.A., van
Barrey, E., 2008. Biomechanics of locomotion and speed as exercise components of risk Weeren, P.R., Barneveld, A., 1999b.
in the athletic horse. In: Hinchcliff, K.W., factors associated with onset of fatigue Influence of site and age on biochemical
Geor, R.J., Kaneps A.J. (Eds.), Equine injury of the third metacarpal bone in characteristics of the collagen network of
exercise physiology. The science of exercise 2-year-old Thoroughbred racehorses. Am. J. equine articular cartilage. Am. J. Vet. Res.
in the athletic horse. Saunders Elsevier, Vet. Res. 61, 602–608. 60, 341–345.
Edinburgh, UK, pp. 143–168. Bottinelli, R., Reggiani, C., 2000. Human Brama, P.A.J., Firth, E.C., van Weeren, P.R.,
Benninghoff, A., 1925. Form und Bau der skeletal muscle fibres: molecular and Tuukkanen, J., Holopainen, J., Helminen,
Gelenkknorpel in ihren Beziehungen zur functional diversity. Prog. Biophys. Mol. H.J., et al., 2009c. Influence of intensity
Funktion. Zweiter Teil: Der Aufbau des Biol. 73, 195–262. and changes of physical activity on bone
Gelenkknorpels in seinen Beziehungen zur Bowe, E.A., Murray, R.C., Jeffcott, L.B., Davies, mineral density of immature equine
Funktion. Zsch. Zelforsch. Mikroskop. M.E., 2007. Do the matrix degrading subchondral bone. Equine Vet. J. 41,
Anat. 2, 783–862. cathepsins B and D increase following a 564–571.
Bergström, J., 1962. Muscle electrolytes in man high intensity exercise regime? Brama, P.A.J., Holopainen, J., van Weeren, P.R.,
determinated by neutron activation analysis Osteoarthritis Cartilage 15, 343–349. Firth, E.C., Helminen, H.J., Hyttinen, M.M.,
on needle biopsy specimens: a study in Boyde, A., Firth, E.C., 2005. Musculoskeletal 2009a. Influence of exercise and joint
normal subjects, kidney patients and responses of 2-year-old thoroughbred topography on depth-related spatial
patients with chronic diarrhoea. Scand. J. horses to early training. 8. Quantitative distribution of proteoglycan and collagen
Clin. Invest. Suppl. 14, 1–110. back-scattered electron scanning electron content in immature equine cartilage.
Billinghurst, R.C., Brama, P.A.J., van Weeren, microscopy and confocal fluorescence Equine Vet. J. 41, 557–563.
P.R., Knowlton, M.S., McIlwraith, C.W., microscopy of the epiphysis of the third Brama, P.A.J., Holopainen, J., van Weeren, P.R.,
2003. Significant exercise-related changes in metacarpal bone. N. Z. Vet. J. 53, 123–132. Helminen, H.J., Hyttinen, M.M., 2009b.
the serum levels of two biomarkers of Effect of loading on the organization
Boyde, A., Firth, E.C., 2008. High resolution
collagen metabolism in young horses. of the collagen fibril network in juvenile
microscopic survey of third metacarpal
Osteoarthritis Cartilage 1111, 760–769. equine cartilage. J. Orth. Res. 27,
articular calcified cartilage and subchondral
Birch, H.L., Bailey, A.J., Goodship, A.E., 1998. 1226–1234.
bone in the juvenile horse: possible
Macroscopic “degeneration” of equine implications in chondro-osseous disease. Brama, P.A.J., van den Boom, R., DeGroot, J.,
superficial digital flexor tendon is Microsc. Res. Tech. 71, 477–488. Kiers, G.H., van Weeren, P.R., 2004.
accompanied by a change in extracellular Collagenase-1 (MMP-1) activity in equine
Boyde, A., Jones, S.J., 1996. Scanning electron
matrix composition. Equine Vet. J. 30, synovial fluid: influence of age, joint
microscopy of bone: instrument, specimen,
534–539. pathology, exercise and repeated
and issues. Microsc. Res. Tech. 33, 92–120.
Birch, H.L., McLaughlin, L., Smith, R.K., arthrocentesis. Equine Vet. J. 36, 34–40.
Goodship, A.E., 1999. Treadmill exercise- Boyde, A., Riggs, C.M., 1990. The quantitative Brommer, H., Brama, P.A.J., Laasanen, M.S.,
induced tendon hypertrophy: assessment of study of the orientation of collagen in Helminen, H.J., Van Weeren, P.R., Jurvelin,
tendons with different mechanical compact bone slices. Bone 11, 35–39. J.S., 2005. Functional adaptation of
functions. Equine Vet. J. Suppl. 30, Brama, P.A., Karssenberg, D., Barneveld, A., articular cartilage from birth to maturity
222–226. van Weeren, P.R., 2001. Contact areas and under the influence of loading: A
Birch, H.L., Wilson, A.M., Goodship, A.E., pressure distribution on the proximal biomechanical analysis. Equine Vet. J. 37,
1997a. Physical training induces alterations articular surface of the proximal phalanx 148–154.
in tendon matrix composition which are under sagittal plane loading. Equine Vet. J. Brommer, H., van Weeren, P.R., Brama, P.A.J.,
structure specific. Proc. Brit. Orthopaed. 33, 26–32. 2003. New approach for quantitative
Res. Soc., Cardiff. Brama, P.A., Tekoppele, J.M., Bank, R.A., assessment of articular cartilage
Birch, H.L, Wilson, A.M., Goodship, A.E., Barneveld, A., Firth, E.C., van Weeren, P.R., degeneration in horses with osteoarthritis.
1997b. The effect of exercise-induced 2000a. The influence of strenuous exercise Am. J. Vet. Res. 64, 83–87.
localised hyperthermia on tendon cell on collagen characteristics of articular Brown, M.P., Trumble, T.N., Plaas, A.H., Sandy,
survival. J. Exp. Biol. 200, 1703–1708. cartilage in thoroughbreds age 2 years. J.D., Romano, M., Hernandez, J., et al.,
Birch, H.L., Wilson, A.M., Goodship, A.E., Equine Vet. J. 32, 551–554. 2007. Exercise and injury increase
2008. Physical activity: does long-term, Brama, P.A., Tekoppele, J.M., Bank, R.A., chondroitin sulfate chain length and
high-intensity exercise in horses result in Barneveld, A., van Weeren, P.R., 2000b. decrease hyaluronan chain length in
tendon degeneration? J. Appl. Physiol. 105, Functional adaptation of equine articular synovial fluid. Osteoarthritis Cartilage 15,
1927–1933. cartilage: the formation of regional 1318–1325.
Bird, J.L., Platt, D., Wells, T., May, S.A., Bayliss, biochemical characteristics up to age one Bruin, G., Kuipers,H., Keizer, H.A., et al., 1994.
M.T., 2000. Exercise-induced changes in year. Equine Vet. J. 32, 217–221. Adaptation and overtraining in horses
proteoglycan metabolism of equine Brama, P.A., Tekoppele, J.M., Bank, R.A., subjected to increasing training loads. J.
articular cartilage. Equine Vet. J. 32, Barneveld, A., van Weeren, P.R., 2002. Appl. Physiol. 76, 1908–1913.
161–163. Development of biochemical heterogeneity Calatroni, A., Avenoso, A., Feriazzo, A.M.,
Blanco, C.E., Sieck, G.C., 1992. Quantitative of articular cartilage: influences of age and Lindner, A., Campo, G.M., 2008. Transient
determination of calcium-activated myosin exercise. Equine Vet. J. 34, 265–269. increase with strenuous exercise of plasma
adenosine triphosphatase activity in rat Brama, P.A., Tekoppele, J.M., Bank, R.A., levels of glycosaminoglycans in humans
skeletal muscle fibres. Histochem. J. 24, Karssenberg, D., Barneveld, A., van Weeren, and horses. Connect. Tissue Res. 49,
431–444. P.R., 2000c. Topographical mapping of 416–425.
Boffi, F.M., Citar, J., Balskus, G., et al., 2002. biochemical properties of articular cartilage Cantley, C.E.L., Firth, E.C., Delahunt, J.W.,
Training-induced apoptosis in skeletal in the equine fetlock joint. Equine Vet. J. Pfeiffer, D.U., Thompson, K.G., 1999.
muscle. Equine Vet. J. Suppl. 34, 275–278. 32, 19–26. Naturally occurring osteoarthritis in the

297
 13 The response of musculoskeletal tissues to exercise

metacarpophalangeal joints of wild horses. Dunlop, R.H., Williams, D.J (Eds.), 1996. Evans, J.H., Barbenel, J.C., 1975. Structural
Equine Vet. J. 31, 73–81. Veterinary medicine. An illustrated history. and mechanical properties of tendon
Carr A.J., Norris, S.H., 1989. The blood supply St. Louis, Mosby. related to function. Equine Vet. J. 7, 1–8.
of the calcaneal tendon. J. Bone Joint Surg. Dykgraaf, S., Firth, E.C., Rogers, C.W., Kawcak, Eyre, D.R., Wu, J.J., 1995. Collagen structure
Br. 71, 100–101. C.E., 2008. Effects of exercise on and cartilage matrix integrity. J. Rheumatol.
Cherdchutham, W., Becker, C., Smith, R.K.W., chondrocyte viability and subchondral 22, 82–85.
Barneveld, A., van Weeren, P.R., 1999. bone sclerosis in the distal third metacarpal Fenton, J.I., Orth, M.W., Chlebek-Brown, K.A.,
Age-related changes and effect of exercise and metatarsal bones of young horses. Vet. Nielsen, B.D., Corn, C.D., Waite, K.S., et
on the molecular composition of immature J. 178, 53–61. al., 1999. Effect of longeing and
equine superficial digital flexor tendon in Dyson, S., Murray, R.C., 2007. Magnetic glucosamine supplementation on serum
foals. Equine Vet. J. Suppl. 31, 86–94. resonance imaging evaluation of 264 markers of bone and joint metabolism in
Cherdchutham, W., Becker, C.K., Spek, E.R., horses with foot pain: the podotrochlear yearling quarter horses. Can. J. Vet. Res. 63,
Voorhout, W.F., van Weeren, P.R., 2001b. apparatus, deep digital flexor tendon and 288–291.
Effects of exercise on the diameter of collateral ligaments of the distal Ferretti, J.L., Capozza, R.F., Zanchetta, J.R.,
collagen fibrils in the central core and interphalangeal joint. Equine Vet. J. 39, 1996. Mechanical validation of a
periphery of the superficial digital flexor 340–343. tomographic (pQct) index for noninvasive
tendon in foals. Am. J. Vet. Res. 62, Eaton, M.D., Hodgson, D.R., Evans, D.L., et estimation of rat femur bending strength.
1563–1570. al., 1999. Effects of low- and moderate- Bone 18, 97–102.
Cherdchutham, W., Meershoek, L.S., van intensity training on metabolic responses Firth, E.C., 2004. Problems in quantifying
Weeren, P.R., Barneveld, A., 2001a. Effects to exercise in Thoroughbreds. Equine Vet. J. bone response to exercise in horses: a
of exercise on biomechanical properties of Suppl. 30, 521–527. review.
the superficial digital flexor tendon in foals. Edwards, L.J., Goodship, A.E., Birch, H.L., Firth, E.C., 2006. The response of bone,
Am. J. Vet. Res. 62, 1859–1864. Patterson-Kane, J.C., 2005. Effect of exercise articular cartilage and tendon to exercise in
Chikuni, K., Muroya, S., Nakajima, I., 2004. on age-related changes in collagen fibril the horse. J. Anat. 208, 513–526.
Absence of the functional myosin heavy diameter distributions in the common Firth, E.C., Delahunt, J., Wichtel, J.W., Birch,
chain 2b isoform in equine skeletal digital extensor tendons of young horses. H.L., Goodship, A.E., 1999c. Galloping
muscles. Zoolog Sci 21, 589−596. Am. J. Vet. Res. 66, 564–568. exercise induces regional changes in bone
Cornelissen, B.P.M., van Weeren, P.R., Eizema, K., van den Burg, M., Kiri, A., et al., density within the third and radial carpal
Ederveen, A.G.H., Barneveld, A., 1999. 2003. Differential expression of equine bones of thoroughbred horses. Equine Vet.
Influence of exercise on bone mineral myosin heavy-chain mRNA and protein J. 31, 111–115.
density of immature cortical and trabecular isoforms in a limb muscle. J. Histochem. Firth, E.C., Goodship, A.E., Delahunt, J.,
bone of the equine metacarpus and Cytochem. 51, 1207−1216. Smith, T., 1999a. Osteoinductive response
proximal sesamoid bone. Equine Vet. J. in the dorsal aspect of the carpus of young
Eizema, K., van den Burg, M.M.M., de Jonge,
Suppl. 31, 79–85. thoroughbreds in training occurs within
H.W., et al., 2005. Myosin heavy chain
Currey, J.D., 1988. The effect of porosity and isoforms in equine gluteus medius muscle: months. Equine Vet. J. Suppl. 30, 552–554.
mineral content on the Young’s modulus of comparison of mRNA and protein Firth, E.C., Rogers, C.W., 2005.
elasticity in compact bone. J. Biomech. 21, expression profiles. J. Histochem. Musculoskeletal responses of 2-year-old
131–139. Cytochem. 53, 1383–1390. thoroughbred horses to early training. 7.
D’Angelis, F.H., Ferraz, G.C., Boleli, I.C., et al., Bone and cartilage response in the carpus.
Essén-Gustavsson, B., Lindholm, A., 1985.
2005. Aerobic training, but not creatine N. Z. Vet. J. 53, 113–122.
Muscle fiber characteristics of active and
supplementation, alters the gluteus medius Firth, E.C., Rogers, C.W., Anderson, B.H.,
inactive standardbred horses. Equine Vet. J.
muscle. J. Anim. Sci. 83, 579−585. 2004b. Musculoskeletal responses of
17, 434–438.
Davies, H.M., Gale, S.M., Baker, I.D., 1999. 2-year-old thoroughbred horses to early
Radiographic measures of bone shape in Essén-Gustavsson, B., Lindholm, A., McMiken, training. 4. Morphometric, microscopic and
young thoroughbreds during training for D., et al., 1983. Skeletal muscle biomechanical properties of the digital
racing. Equine Vet. J. Suppl. 30, 262–265. characteristics of young standardbreds in tendons of the forelimb. N. Z. Vet. J. 52,
relation to growth and early training. In: 285–292.
Dingboom, E.G., Dijkstra, G., Enzerink, E., et
Snow, D.H., Persson, S.G.B., Rose, R.J.
al., 1999. Postnatal muscle fibre Firth, E.C., Rogers, C.W., Doube, M., Jopson,
(Eds.), Equine exercise physiology.
composition of the gluteus medius muscle N.B., 2005. Musculoskeletal responses of
Burlington Press, Foxton, Cambridge, pp.
of Dutch warmblood foals: maturation and 2-year-old thoroughbred horses to early
200–210.
the influence of exercise. Equine Vet. J. training. 6. Bone parameters in the third
Suppl. 31, 95–100. Essén-Gustavsson, B., McMiken, D., Karlström, metacarpal and third metatarsal bones. N.
K., et al., 1989. Muscular adaptation of Z. Vet. J. 53, 101–112.
Dingboom, E.G., Van, O.H., Eizema, K., et al.,
horses during intensive training and Firth, E.C., Rogers, C.W., Jopson, N.B., 2000.
2002. Changes in fibre type composition of
detraining. Equine Vet. J. 21, 27–33. Effects of racetrack exercise on third
gluteus medius and semitendinosus
muscles of Dutch Warmblood foals and the Eto, D., Yamano, S., Hiraga, A. et al., 2006. metacarpal and carpal bone of New
effect of exercise during the first year Recruitment pattern of muscle fibre type Zealand thoroughbred horses. J.
postpartum. Equine Vet. J. 34, 177–183. during flat and sloped treadmill running in Musculoskel. Neuronal. Interat. 1,
Doube, M., Firth, E.C., Boyde, A., 2007. thoroughbred horses. Equine Vet. J. Suppl. 145–147.
Variations in articular calcified cartilage by 36, 349–353. Firth, E.C., Rogers, C.W., Perkins, N.R.,
site and exercise in the 18-month-old Eto, D., Yamano, S., Kasashima, Y., et al., Anderson, B.H., Grace, N.D., 2004a.
equine distal metacarpal condyle. 2003. Effect of controlled exercise on Musculoskeletal responses of 2-year-old
Osteoarthritis Cartilage 15, 1283–1292. middle gluteal muscle fibre composition in thoroughbred horses to early training. 1.
Dudhia, J., Scott, C.M., Draper, E.R., thoroughbred foals. Equine Vet. J. 35, Study design and clinical, nutritional,
Heinegård, D., Pitsillides, A.A., Smith, R.K., 676−680. radiological and histological observations.
2007. Aging enhances a mechanically- Eto, D., Yamano, S., Mukai, K., et al., 2004. N. Z. Vet. J. 52, 261–271.
induced reduction in tendon strength by an Effect of high intensity training on Firth, E.C., Rogers, C.W., van Weeren, P.R.,
active process involving matrix anaerobic capacity of middle gluteal Barneveld, A., Kawcak, C.E., McIlwraith,
metalloproteinase activity. Aging Cell 6, muscle in thoroughbred horses. Res. Vet. C.W., et al., 2007. Changes in diaphyseal
547–556. Sci. 76, 139–144. and epiphyseal bone parameters in

298
 References

thoroughbred horses after withdrawal from collagen fibre formation. Biomaterials 10, Hayes, W.C., 1991. Biomechanics of cortical
training. J. Musculoskelet. Neuronal. 413–419. and trabecular bone: implications for
Interact. 7, 74–76. Gathercole, L.J., Keller, A., 1991. Crimp assessment of fracture risk. In: Hayes, W.C.,
Firth, E.C., Rogers, C.W., van Weeren, P.R., morphology in the fibre-forming collqaens. Mow, V.C. (Eds.), Basic Orthopedic
Barneveld, A., Kawcak, C.E., McIlwraith, Matrix 11, 214–234. Biomechanics. Raven Press, New York, pp.
C.W., et al., 2010. Mild exercise in Geor, R.J., McCutcheon, L.J., Shen, H., 1999. 93–140.
life produces changes in bone size and Muscular and metabolic responses to Heck, R.W., McKeever, K.H., Always, S.E., et al.,
strength but not density in proximal moderate-intensity short-term training. 1996. Resistance training-induced increases
phalangeal, third metacarpal and Equine Vet. J. Suppl. 30, 311–317. in muscle mass and performance in ponies.
third carpal bones of foals. Vet. J. 190, Gillard, G.C., Reilly, H.C., Bell-Booth, P.G., Med. Sci. Sports. Exerc. 28:877–883.
383–390. Flint, M.H., 1979. The influence of Heilig, A., Pette, D., 1988. Albumin in rabbit
Firth, E.C., Rogers, C.W., van Weeren, P.R., mechanical forces on the skeletal muscle. Origin, distribution and
Barneveld, A., McIlwraith, C.W., Kawcak, glycosaminoglycan content of the rabbit regulation by contractile activity. Eur. J.
C.E., et al., 2009. Responses in proximal flexor digitorum profundus tendon. Biochem. 171, 503–508.
phalangeal, third metacarpal and third Connect. Tissue Res. 7, 37–46. Helal, I.E., Misumi, K., Tateno, O., Kodama, T.,
carpal bone of foals exposed to exercise Goldspink, G., Scutt, A., Loughna, P.T., et al., Ishimaru, M., Yamamoto, J., et al., 2007.
early in life. J. Orthop. Res., submitted. 1992. Gene expression in skeletal muscle Effect of exercise on serum concentration
Firth, E.C., Rogers, C.W., van Weeren, P.R., in response to stretch and force generation. of cartilage oligomeric matrix protein in
Barneveld, A., McIlwraith, C.W., Kawcak, Am. J. Physiol. 262: R356–R363. Thoroughbreds. Am. J. Vet. Res. 68,
C.E., et al., 2012. The effect of previous 134–140.
Gondim, F.J., Modolo, L.V., Campos, G.E., et
conditioning exercise on diaphyseal and Helminen, H.J., Hyttinen, M.M., Lammi, M.J.,
al., 2005. Neuronal nitric oxide synthase is
metaphyseal bone to imposition and Arokoski, J.P.A., Lapveteläinen, T., Jurvelin,
heterogeneously distributed in equine
withdrawal of training in young J., et al., 2000. Regular joint loading in
myofibers and highly expressed in
thoroughbred horses. Vet. J. 192, youth assists in the establishment and
endurance trained horses. Can. J. Vet. Res.
34–40. strengthening of the collagen network of
69, 46−52.
Firth, E.C., van Weeren, P.R., Pfeiffer, D.U., articular cartilage and contributes to the
Goodman, S.A., May, S.A., Heinegård, D., prevention of osteoarthrosis later in life: a
Delahunt, J., Barneveld, A., 1999b. Effect of Smith, R.K., 2004. Tenocyte response to
age, exercise and growth rate on bone hypothesis. J. Bone Min. Metab. 18,
cyclical strain and transforming growth 245–257.
mineral density (BMD) in third carpal factor beta is dependent upon age and site
bone and distal radius of Dutch Henckel P., 1983. Training and growth
of origin. Biorheology 41, 613–628.
Warmblood foals with osteochondrosis. induced changes in the middle gluteal
Goodship A.E., Birch, H.L., Wilson, A.M., muscle of young standardbred trotters.
Equine Vet. J. Suppl. 31, 74–78. 1994. The pathobiology and repair of Equine Vet. J. 15, 134–140.
Flint, M.H., Craig, A.S., Reilly, H.C., Gillard, tendon and ligament injury. Vet. Clin.
G.C., parry, D.A.D., 1984. Collagen fibril Henckel, P., Ducro, B., Oksbjerg, N., et al.,
North Am. Equine Pract. 10, 323–349.
diameters and glycosaminoglycan content 1998. Objectivity of two methods of
Goodship, A.E., Lanyon, L.E., McFie, H., 1979. differentiating fibre types and repeatability
of skins – indices of tissue maturity and Functional adaptation of bone to increased
function. Conn. Tissue Res. 13, 69–81. of measurements by application of the
stress. An experimental study. J. Bone Joint TEM image analysis system. Eur. J.
Foreman, J.H., Bayly, W.M., Allen, H., et al., Surg. Am. 61, 539–546. Histochem. 42, 49–62.
1990. Muscle responses of thoroughbreds Gordon, K.R., Perl, M., Levy, C., 1989. Herzog, W., Federico, S., 2006. Considerations
to conventional race training and Structural alterations and breaking strength on joint and articular cartilage mechanics.
detraining. Am. J. Vet. Res. 51, 909–913. of mouse femora exposed to three activity Biomech. Mod. Mechanobiol. 5, 64–81.
Fox, G., Henckel, P., Juel, C., et al., 1987. regimens. Bone 10, 303–312. Hinchcliff, K.W., Lauderdale, M.A., Dutson, J.,
Skeletal muscle buffer capacity changes in Gottlieb, M., Essén-Gustavsson, B., Lindholm, et al., 2002. High intensity exercise
standardbred horses: effects of growth and A., et al., 1989. Effects of a draft-loaded conditioning increases accumulated oxygen
training. In: Gillespie, J.R., Robinson, N.E. interval-training program on skeletal deficit of horses. Equine Vet. J. 34, 9–16.
(Eds.), Equine exercise physiology 2. ICEEP muscle in the horse. J. Appl. Physiol. 67, Hudlicka, O., Brown, M., Egginton, S., 1992.
Publications, Davis, CA, pp. 341–347. 570–577. Angiogenesis in skeletal and cardiac
Frisbie, D.D., Al-Sobayil, F., Billinghurst, R.C., Grace, N.D., Rogers, C.W., Firth, E.C., Faram, muscle. Physiol. Rev. 72, 369–417.
Kawcak, C.E., McIlwraith, C.W., 2008. T.L., Shaw, H.L., 2003. Digestible energy Hunter, W., 1743. On the structure and
Changes in synovial fluid and serum and effect of increased calcium intake on diseases of articulating cartilage. Philos.
biomarkers with exercise and early bone parameters of grazing thoroughbred Trans. R. Soc. Lond. 42, 514–521.
osteoarthritis in horses. Osteoarthritis weanlings in New Zealand. N. Z. Vet. J. 51, Hyyppä, S., Hänninen, O., 1998. Application
Cartilage 16, 1196–1204. 165–173. of surface electromyography in horses
Gabel, A.A., Spencer, C.P., Pipers, F.S., 1977. A Green, D.A., 1976. Growth rate in during physical exercise. In: Lindner, A.
study of correlation of closure of the distal thoroughbred yearlings and two-year-olds. (Ed.), Proceedings of the conference on
radial physis with performance and injury Equine Vet. J. 8, 133–134. equine sport medicine and sciences.
in the standardbred. J. Am. Vet. Med. Wageningen Pers, Wageningen, The
Green, H.J., Düsterhöft, S., Dux, L., et al.,
Assoc. 170, 188–194. Netherlands, pp. 156–162.
1992. Metabolite patterns related to
Gabriel, D.A., Kamen, G., Frost, G., 2006. exhaustion, recovery, and transformation of Hyyppä, S., Pösö, A.R., 1998. Fluid, electrolyte,
Neural adaptations to resistive exercise: chronically stimulated rabbit fast-twitch and acid–base responses to exercise in
mechanisms and recommendations for muscle. Pflüg Arch. Eur. J. Physiol. 420, racehorses. Vet. Clin. North Am. Equine
training practices. Sports Med. 36, 359–366. Pract. 14, 121–136.
133–149. Iozzo R.V., 1998. Matrix proteoglycans: from
Gunn, H.M., 1987. Muscle, bone and fat
Gansen, S., Lindner, A., Marx, S., et al., 1999. proportions and muscle distribution of molecular design to cellular function.
Effects of conditioning horses with thoroughbreds and quarter horses. In: Annu. Rev. Biochem. 67, 609–652.
lactate-guided exercise on muscle glycogen Gillespie, J.R., Robinson, N.E. (Eds.), Jones, G., Bennell, K., Cicuttini, F.M., 2003.
content. Equine Vet. J. Suppl. 30, 329–331. Equine exercise physiology 2. Effect of physical activity on cartilage
Garg A.K., Berg, R.A., Silver, F.H., Garg, H.G., ICEEP Publications, Davis, CA, pp. development in healthy kids. Br. J. Sports
1989. Effect of proteoglycans on type I 253–264. Med. 37, 382–383.

299
 13 The response of musculoskeletal tissues to exercise

Jozsa L., Kannus P., 1997. Functional and Articular cartilage thickness and between antigenic and ATPase
mechanical behaviour of tendons. In: glycosaminoglycan distribution in the determinants. J. Anat. 181, 363–372.
Jozsa, L.G., Kannus, P. (Eds.), Human canine knee joint after strenuous running Little, C.B., Ghosh, P., 1997. Variation in
tendons: anatomy, physiology, and exercise. Clin. Orthop. Rel. Res. 283, proteoglycan metabolism by articular
pathology. Human Kinetics Ltd., Illinois, 302–308. chondrocytes in different joint regions is
pp. 98–113. Kiviranta, I., Tammi, M., Jurvelin, J., Arokoski, determined by post-natal mechanical
Kane, A.J., McIlwraith, C.W., Park, R.D., J., Säämänen, A.M., Helminen, H.J., 1994. loading. Osteoarthritis Cartilage 5, 49–62.
Rantanen, N.W., Morehead, J.P., Bramlage, Articular cartilage thickness and Little, C.B., Ghosh, P., Rose, R., 1997. The
L.R., 2003. Radiographic changes in glycosaminoglycan distribution in the effect of strenuous versus moderate exercise
thoroughbred yearlings. Part 2: Associations young canine knee joint after on the metabolism of proteoglycans in
with racing performance. Equine Vet. J. 35, remobilization of the immobilized limb. J. articular cartilage from different weight-
366–374. Orthop. Res. 12, 161–167. bearing regions of the equine third carpal
Kannus, P., 2000. Structure of the tendon Kiviranta, I., Tammi, M., Jurvelin, J., bone. Osteoarthritis Cartilage 5, 161–172.
connective tissue. Scand. J. Med. Sci. Sports. Säämänen, A.M., Helminen, H.J., 1988. López-Rivero, J.L., Agüera, E., Monterde, J.G.,
10, 312–320. Moderate running exercise augments et al., 1992. Skeletal muscle fiber size in
Kannus P., Jozsa, L., 1991. Histopathological glycosaminoglycans and thickness of untrained and endurance-trained horses.
changes preceding spontaneous rupture of articular cartilage in the knee joint of Am. J. Vet. Res. 53, 847–850.
a tendon. A controlled study of 891 young Beagle dogs. J. Orthop. Res. 6, López-Rivero, J.L., Morales-López, J.L.,
patients. J. Bone Joint Surg. Am. 73, 188–195. Galisteo, A.M., et al., 1991. Muscle fibre
1507–1525. Kontulainen, S., Sievänen, H., Kannus, P., type composition in untrained and
Karlsson, M.K., 2007. Does exercise during Pasanen, M., Vuori, I., 2003. Effect of endurance-trained Andalusian and Arab
growth prevent fractures in later life? Med. long-term impact-loading on mass, size horses. Equine Vet. J. 23, 91–93.
Sport Sci. 51, 121–136. and estimated strength of humerus and Lovell, D.K., Rose, R.J., 1991. Changes in
radius of female racquet-sports players: a skeletal muscle composition in response to
Karlstrom, K., Essén-Gustavsson, B., 2002.
peripheral quantitative computed interval and high intensity training. In:
Myosin heavy chain-based fibre types in
tomography study between young and old Persson, S.G.B., Lindholm, A., Jeffcott, L.B.
red cell hyper- and normovolaemic
starters and controls. J. Bone Miner. Res. (Eds.), Equine exercise physiology 3. ICEEP
standardbred trotters. Equine Vet. J. Suppl.
18, 352–359. Publications, Davis, CA, pp. 215–222.
34, 279–282.
Köwitz, J., Knippel, M., Schuhr, T., Mach, J., Marlin, D.J., Harris, R.C., Gash, S.P., et al.,
Kasashima, Y., Smith, R.K., Birch, H.L.,
1997. Alteration in the extent of collagen I 1989. Carnosine content of the middle
Takahashi, T., Kusano, K., Goodship, A.E.,
hydroxylation, isolated from femoral heads gluteal muscle in thoroughbred horses with
2002. Exercise-induced tendon
of women with a femoral neck fracture relation to age, sex and training. Comp.
hypertrophy: cross-sectional area changes
caused by osteoporosis. Calcif. Tissue Int. Biochem. Physiol. 93:629–635.
during growth are influenced by exercise.
60, 501–505. Maroudas, A., 1980. Metabolism of
Equine Vet. J. Suppl. 34, 264–268.
Kurvers, C.M.H.C., van Weeren, P.R., Rogers, cartilaginous tissues: A quantitative
Kasashima, Y., Takahashi, T., Birch, H.L.,
C.W., van Dierendonck, M.C., 2006. approach. In: Holborow, E.J., Maroudas, A.
Smith, R.K., Goodship, A.E., 2008. Can
Quantification of spontaneous locomotion (Eds.), Studies in joint disease, Vol. 1.
exercise modulate the maturation of
activity in foals kept in pastures under Tunbridge Wells, Pitman Medical, pp.
functionally different immature tendons in
various management conditions. Am. J. Vet. 59–86.
the horse? J. Appl. Physiol. 104, 416–422.
Res. 67, 1212–1217. Maroudas, A., Bayliss, M.T., Uchitel-
Kastelic J., Galeski, A., Baer, E., 1978. The
Lin, Y.L., Brama, P.A., Kiers, G.H., DeGroot, J., Kraushansky, N., Schneiderman, R., Gilav,
multicomposite structure of tendon.
van Weeren, P.R., 2005b. Functional E., 1998. Aggrecan turnover in human
Connect. Tissue Res. 6, 11–23.
adaptation through changes in regional articular cartilage: use of aspartic acid
Kawcak, C.E., McIlwraith, C.W., Firth, E.C., biochemical characteristics during racemisation as a marker of molecular age.
2010. Effects of early exercise on maturation of equine superficial digital Arch. Biochem. Biophys. 350, 61–71.
metacarpophalangeal joints in horses. Am. flexor tendon. Am. J. Vet. Res. 66, Maroudas, A., Palla, G., Gilav, E., 1992.
J. Vet. Res. 71, 405–411. 1623–1629. Racemization of aspartic acid in human
Kawcak, C.E., McIlwraith, C.W., Norrdin, R.W., Lin, Y.L., Brama, P.A.J., Kiers, G.H., van articular cartilage. Connect. Tissue Res. 28,
Park, R.D., Steyn, P.S., 2000. Clinical effects Weeren, P.R., DeGroot, J., 2005a. 161–169.
of exercise on subchondral bone and Extracellular matrix composition of the Martin, R.B., Boardman, D.L., 1993. The
metacarpophalangeal joints in horses. Am. equine superficial digital flexor tendon: effects of collagen fiber orientation,
J. Vet. Res. 61, 1252–1258. relationship with age and anatomical site. porosity, density, and mineralization on
Kearns, C.F., McKeever, K.H., Abe, T., 2002. J. Vet. Med. A. 52, 333–338. bovine cortical bone bending properties. J.
Overview of horse composition and muscle Lindholm, A., Essén-Gustavsson, B., McMiken, Biomech. 26, 1047–1054.
architecture: implications for performance. D., et al., 1983. Muscle histochemistry Martin, R.B., Ishida, J., 1989. The relative
Vet. J. 164, 224−234. and biochemistry of thoroughbred effects of fiber orientation, porosity, density
Kim, J.S., Hinchcliff, K.W., Yamaguchi, M., et horses during growth and training. and mineralization on bone strength. J.
al., 2005. Exercise training increases In: Snow, D.H., Persson, S.G., Rose, R.J. Biomech. 22, 419–426.
oxidative capacity and attenuates exercise- (Eds.), Equine exercise physiology. McCarthy, R.N., Jeffcott, L.B., 1992. Effects of
induced ultrastructural damage in skeletal Burlington Press, Foxton, Cambridge, treadmill exercise on cortical bone in the
muscle of aged horses. J. Appl. Physiol. 98, pp. 211–217. third metacarpus of young horses. Res. Vet.
334−342. Lindholm, A., Piehl, K., 1974. Fibre Sci. 52, 28–37.
Kiviranta, I., Jurvelin, J., Tammi, M., composition, enzyme activity and McCutcheon, L.J., Geor, R.J., Hinchcliff, K.J.,
Säämänen, A.M., Helminen, H.J., 1987. concentrations of metabolites and 2002. Changes in skeletal muscle GLUT 4
Weight bearing controls glycosaminoglycan electrolytes in muscles of standardbred content and muscle membrane glucose
concentration and articular cartilage horses. Acta Vet. Scand. 15, 280–309. transport following 6 weeks of exercise
thickness in the knee joints of young Linnane, L., Serrano, A.L., Rivero, J.L.L., 1999. training. Equine Vet. J. Suppl. 34, 199–204.
beagle dogs. Arthr. Rheum. 30, 801–809. Distribution of fast myosin heavy McCutcheon, L.J., Geor, R.J., Shen, H., 1999.
Kiviranta, I., Tammi, M., Jurvelin, J., Arokoski, chain-based muscle fibres in the gluteus Skeletal muscle Na+-K+-ATPase and K+
J., Saamanen, A.M., Helminen, H.J., 1992. medius of untrained horses: mismatch homeostasis during exercise: effects of

300
 References

short-term training. Equine Vet. J Suppl. articular cartilage fibronectin distribution in UK racehorses- fracture descriptions and
30, 303–310. associated with training, joint location and pre-existing pathology. Vet. J. 171,
McCutcheon, L.J., Kelso, T., Bertocci, L.A., et cartilage deterioration. Equine Vet. J. 32, 157–165.
al., 1987. Buffering and aerobic capacity in 47–51. Parry D.A., 1988. The molecular and fibrillar
equine muscle: variation and effect of Murray, R.C., Smith, R.K.W., Henson, F.M.D., structure of collagen and its relationship to
training. In: Gillespie, J.R., Robinson, N.E. Goodship, A.E., 2001b. The distribution of the mechanical properties of connective
(Eds.), Equine exercise physiology 2. ICEEP cartilage oligomeric matrix protein tissue. Biophys. Chem. 29, 195–209.
Publications, Davis, CA, pp. 348–358. (COMP) in equine carpal articular cartilage Parry, D.A., Flint, M.H., Gillard, G.C., Craig,
McGowan, C.M., Golland, L.C., Evans, D.L., et and its variation with exercise and cartilage A.S., 1982. A role for glycosaminoglycans
al., 2002. Effects of prolonged training, deterioration. Vet. J. 162, 121–128. in the development of collagen fibrils.
overtraining and detraining on skeletal Murray, R.C., Whitton, R.C., Vedi, S., Goodhip, FEBS lett. 149, 1–7.
metabolites and enzymes. Equine Vet. J. A.E., Lekeux, P., 1999a. The effect of Parry, D.A.D., Craig, A.S., Barnes, G.R.G.,
Suppl. 34, 257–263. training on the calcified zone of equine 1978. Tendons and ligaments from the
Micklethwaite L., Ghosh, P., Little, C.B., middle carpal articular cartilage. Equine horse: an ultrastructural study of collagen
Hodgson, D.R., Rose, R.J., 1999. Vet. J. Suppl. 30, 274–278. fibrils and elastic fibres as a function of
Proteoglycan composition and biosynthesis Murray, R.C., Zhu, C.F., Goodship, A.E., age. Proc. R. Soc. Lond. B. 203, 293–302.
in superficial and deep digital flexor Lakhani, K.H., Agrawal, C.M., Athanasiou,
Patterson-Kane, J.C., Firth, E.C., Goodship,
tendons of standardbred horses. Vet. Comp K.A., 1999b. Exercise affects the mechanical
A.E., Parry, D.A., 1997a. Age-related
Orthop. Traumatol. 12, 166–172. properties and histological appearance of
differences in collagen crimp patterns in
Mienaltowski, M.J., Huang, L., Stromberg, A.J., equine articular cartilage. J. Orthop. Res.
the superficial digital flexor tendon core
MacLeod, J.N., 2008. Differential gene 17, 725–731.
region of untrained horses. Aust. Vet. J. 75,
expression associated with postnatal equine Newton, P.O., Woo, S.L.Y., Mackenna, D.A., 39–44.
articular cartilage maturation. BMC Akeson, W.H., 1995. Immobilisation of the
Patterson-Kane, J.C., Parry, D.A.D., Birch, H.L.,
Muskuloskelet. Disord. 9, 149. knee alters the mechanical and
Goodship, A.E., Firth, E.C., 1997b. An
Milner, D.J., Taffet, G.E., Wang, X., et al., 1999. ultrastructural properties of the rabbit
age-related study of morphology and
The absence of desmin leads to anterior cruciate ligament. J Orthop. Res.
cross-link composition of collagen fibrils in
cardiomyocyte hypertrophy and cardiac 13, 191–200.
the digital flexor tendons of young
dilation with compromised systolic Nugent, G.E., Law, A.W., Wong, E.G., Temple, thoroughbred horses. Connect. Tissue Res.
function. Mol. Cell Cardiol. 31, M.M., Bae, W.C., Chen, A.C., et al., 2004. 36, 253–260.
2063–2076. Site- and exercise-related variation in
Patterson-Kane, J.C., Parry, D.A.D., Goodship,
Misumi, K., Skamoto, H., Shimizu, R., 1995. structure and function of cartilage from
A.E., Firth, E.C., 1997d. Exercise modifies
Changes in skeletal muscle composition in equine distal metacarpal condyle.
the age-related change in crimp pattern in
response to swimming training for young Osteoarthritis Cartilage 12, 826–833.
the core region of the equine superficial
horses. J. Vet. Med. Sci. 57, 959–961. Okumura, M., Kim, G.H., Tagami. M.,
digital flexor tendon. N. Z. Vet. J. 45,
Miyata,H., Sugiura, T., Kai, M., et al., 1999. Haramaki, S., Fujinaga, T., 2002. Serum
135–139.
Muscle adaptation of thoroughbred keratan sulphate as a cartilage metabolic
racehorses trained on a flat or sloped track. marker in horses: the effect of exercise. J. Patterson-Kane, J.C., Wilson, A.M., Firth, E.C.,
Am. J. Vet. Res. 60, 1536–1539. Vet. Med. A Physiol. Pathol. Clin. Med. 49, Parry, D.A., Goodship, A.E., 1997c.
195–197. Comparison of collagen fibril populations
Moffat, P.A., Firth, E.C., Rogers, C.W., Smith,
in the superficial digital flexor tendons of
R.K.W., Barneveld, A., Goodship, A.E., et Oxlund, H., Barckman, M., Ortoft, G.,
exercised and non-exercised thoroughbreds.
al., 2008. The influence of exercise during Andreassen, T.T., 1995. Reduced
Equine Vet. J. 29, 121–125.
growth on ultrasonographic parameters of concentrations of collagen cross-links are
the superficial digital flexor tendon of associated with reduced strength of bone. Patterson-Kane, J.C., Wilson, A.M., Firth, E.C.,
young thoroughbred horses. Equine Vet. J. Bone 17 (Suppl.), 365S-S371. Parry, D.A.D., Goodship, A.E., 1998.
40, 136–140. Exercise-related alterations in crimp
Pallafacchina, G., Calabria, E., Serrano, A.L., et
morphology in the central regions of
Monnier, V.M., 1989. Toward a maillard al., 2002. A protein kinase B-dependent
superficial digital flexor tendons from
reaction theory of ageing. Prog. Clin. Biol. and rapamycin-sensitive pathway controls
young thoroughbreds: a controlled study.
Res. 304, 1–22. skeletal muscle growth but not fiber type
Equine Vet. J. 30, 61–64.
Morgan, D.L., Proske, U., Warren, D., 1978. specification. Proc. Nat. Acad. Sci. USA 99,
Measurements of muscle stiffness and the 9213–9218. Payne, R.C., Hutchinson, J.R., Robilliard, J.J.,
mechanisms of elastic storage of energy in Palmer, J., Bertone, A., 1996. Joint et al., 2005b. Functional specialization of
hopping kangaroos. J. Physiol. (Lond.) biomechanics in the pathogenesis of pelvic limb anatomy in horses (Equus
282, 253–261. traumatic arthritis. In: McIlwraith, C.W., caballus). J. Anat. 205, 557−574.
Morrow, N.G., Kraus, W.E., Moore, J.W., et al., Trotter, G.W. (Eds.), Joint disease in the Payne, R.C., Veenman, P., Wilson, A.M., 2005a.
1990. Increased expression of fibroblast horse. Saunders, Philadelphia, pp. 104–119. The role of the extrinsic thoracic limb
growth factors in a rabbit muscle model of Palmer, J.L., Bertone, A.L., Malemud, C.J., muscles in equine locomotion. J. Anat.
exercise conditioning. J. Clin. Invest. 85, Carter, B.G., Papay, R.S., Mansour, J., 206, 193−204.
1816–1820. 1995a. Site-specific proteoglycan Persson, S.G.B., Essén-Gustavsson, B.,
Murgia, M., Serrano, A.L., Calabria, E., et al., characteristics of third carpal articular Lindholm, A., 1991. Energy profile and the
2000. Ras is involved in nerve-activity- cartilage in exercised and nonexercised locomotory pattern of trotting on an
dependent regulation of muscle genes. horses. Am. J. Vet. Res. 56, 1570–1576. inclined treadmill. In: Persson, S.G.B.,
Nature Cell Biol. 2, 142–147. Palmer, J.L., Bertone, A.L., Mansour, J., Carter, Lindholm, A., Jeffcott, L.B. (Eds.), Equine
Murray, R.C., Birch, H.L., Lakhani, K., B.G., Malemud, C.J., 1995b. Biomechanical exercise physiology 3. ICEEP Publications,
Goodship, A.E., 2001a. Biochemical properties of third carpal articular cartilage Davis, CA, pp. 231–238.
composition of equine carpal articular in exercised and nonexercised horses. J. Pette, D., 1998. Training effects on the
cartilage is influenced by short-term Orthop. Res. 13, 854–860. contractile apparatus. Acta. Physiol. Scand.
exercise in a site-specific manner. Parkin, T.D., Clegg, P.D., French, N.P., 162, 367–376.
Osteoarthritis Cartilage 9, 625–632. Proudman, C.J., Riggs, C.M., Singer, E.R., et Pette, D., Staron, S., 1997. Mammalian
Murray, R.C., Janicke, H.C., Henson, F.M.D., al., 2006. Catastrophic failure of the lateral skeletal muscle fiber type transitions. Int.
Goodship, A.E., 2000. Equine carpal condyle of the third metacarpus/metatarsus Rev. Cytol. 170, 143–221.

301
 13 The response of musculoskeletal tissues to exercise

Pool, R.R., 1996. Pathologic manifestations of Rivero, J.L.L., 2007. A scientific background for immunohistochemical and electrophoretic
joint disease in the athletic horse. In: skeletal muscle conditioning in equine study. Anat. Rec. 246, 185–194.
McIlwraith, C.W., Trotter, G.W. (Eds.), Joint practice. J. Vet. Med. A 54, 321–332. Rivero, J.L.L., Talmadge, R.J., Edgerton, V.R.,
disease in the horse. Saunders, Rivero, J.L.L., Clayton, H., 1996. The potential 1996b. Correlation between myofibrillar
Philadelphia, pp. 87–104. role of the muscle in kinematic ATPase activity and myosin heavy chain
Pösö, A.R., Eklund-Uusitalo, S., Hyypä, S., characteristics. Pferdeheilkunde 12, composition in equine skeletal muscle and
Pirilä, E., 2002. Induction of heat shock 635–640. the influence of training. Anat. Rec. 246,
protein 72 mRNA in skeletal muscle byb Rivero, J.L.L., Letelier, A.I., 2000. Skeletal 195–207.
exercise and training. Equine Vet. J. Suppl. muscle profile of show jumpers: Rivero, J.L.L., Talmadge, R.J., Edgerton, V.R.,
34, 214–218. physiological and pathological 1997. A sensitive electrophoretic method
Puchalski, S.M., Galuppo, L.D., Hornof, W.J., considerations. In: Lindner, A. (Ed.), The for the quantification of myosin heavy
Wisner, E.R., 2007. Intraarterial contrast- elite show jumper, conference on equine chain isoforms in horse skeletal muscle:
enhanced computed tomography of the sports medicine and science 2000. Lensing histochemical and immunocytochemical
equine distal extermity. Vet. Radiol. Druck, Dortmund, Germany, pp. 57–76. verifications. Electrophoresis 18,
Ultrasound 48, 21–29. Rivero, J.L.L., Piercy, R.J., 2004. Muscle 1967–1972.
Rauch, F., Schoenau, E., 2001. Changes in physiology: responses to exercise and Rivero, J.L.L., van Breda, E., Rogers, C.W., et
bone density during childhood and training. In: Hinchcliff, K.W., Geor, R.J., al., 2008. Unexplained underperformance
adolescence: an approach based on bone’s Kaneps A.J. (Eds.), Equine sport medicine syndrome in sport horses: classification,
biological organisation. J. Bone Miner. Res. and surgery. chapter 5, Saunders Elsevier potential causes and recognition. Equine
16, 597–604. Ltd, Edinburgh, UK, pp. 45–76. Vet. J. 40, 611–618.
Ray, C.S., Poole, A.R., McIlwraith, C.W., 1996. Rivero, J.L.L., Piercy, R.J., 2008. Muscle Robert, C., Valette, J.P., Denoix, J.M., 1998.
Use of synovial fluid and serum markers in physiology: responses to exercise and Surface electromyographic analysis of the
articular disease. In: McIlwraith, C.W., training. In: Hinchcliff, K.W., Geor, R.J., normal horse locomotion: a preliminary
Trotter, G.W. (Eds.), Joint disease in the Kaneps, A.J. (Eds.), Equine exercise report. In: Lindner, A. (Ed.), Proceedings of
horse. Saunders, Philadelphia, pp. 203–216. physiology. The science of exercise in the the conference on equine sport medicine
Reid, S.A., Boyde, A., 1987. Changes in the athletic horse. Saunders Elsevier, and sciences. Wageningen Pers,
mineral density distribution in human Edinburgh, pp. 30–80. Wageningen, The Netherlands, pp. 80–85.
bone with age: image analysis using Rivero, J.L.L., Ruz, A., Martí-Korff, S., et al., Robertson M., 1994. The extracellular matrix
backscattered electrons in the SEM. J. Bone 2007. Effects of intensity and duration of of animals. In: Robertson, M. (Ed.),
Miner. Res. 2, 13–22. exercise on muscular responses to training Molecular biology of the cell, 2nd edn.
Riemersma D.J., Schamhardt, H.C., 1985. In of thoroughbred racehorses. J. Appl. Garland Publishing, New York, pp.
vitro mechanical properties of equine Physiol 102, 1871–1882. 971–994.
tendons in relation to cross-sectional area Rivero, J.L.L., Ruz, M.C., Serrano, A., et al., Rodnick, K.J., Henriksen, E.J., James, D.E.,
and collagen content. Res. Vet. Sci. 39, 1995a. Effects of a 3 month endurance et al., 1992. Exercise training, glucose
263–270. training programme on skeletal muscle transporters, and glucose transport in rat
Rieppo, J., Hyttinen, M.M., Halmesmaki, E., histochemistry in Andalusian, Arabian and skeletal muscles. Am. J. Physiol. 262,
Ruotselainen, H., Vasara, M., Kiviranta, I., Anglo-Arabian horses. Equine Vet. J. 27, C9–C14.
et al., 2008. Changes in spatial collagen 51–59. Rogers, C.W., Bolwell, C.F., Tanner, C.J., van
content and collagen network architecture Rivero, J.L.L., Serrano, A.L., 1999. Skeletal Weeren, P.R., 2012. Early exercise in the
in porcine articular cartilage during growth myosin heavy chain and carriage training. horse. J. Vet. Behav. Clin. Appl. Res. 7,
and maturation. Osteoarthritis Cartilage, in Equine Vet. J. Suppl. 30, 318–323. 375–379.
press. Rivero, J.L.L., Serrano, A.L., Barrey, E., et al., Rogers, C.W., Firth, E.C., McIlwraith, C.W.,
Rietbroek, N.J., Dingboom, E.G., Joosten, 1999. Analysis of myosin heavy chains at Barneveld, A., Goodship, A.E., Kawcak,
B.J.E., et al., 2007b. effect of show jumping the protein level in horse skeletal muscle. C.E., et al., 2008a. Evaluation of a new
training on the development of locomotory J. Muscle Res. Cell Motil. 20, 211–221. strategy to modulate skeletal development
muscle in young horses. Am. J. Vet. Res. 68, Rivero, J.L.L., Serrano, A.L., Henckel, P., in thoroughbred performance horses by
1232–1238. 1995b. Activities of selected aerobic and imposing track-based exercise during
Rietbroek, N.J., Dingboom, E.G., Schuurman, anaerobic enzymes in the gluteus medius growth. Equine Vet. J. 40, 111–118.
S.O., et al., 2007a. Effect of exercise on muscle of endurance horses with different Rogers, C.W., Firth, E.C., McIlwraith, C.W.,
development of capillary supply and performance records. Vet. Rec. 137, Barneveld, A., Goodship, A.E., Kawcak,
oxidative capacity in skeletal muscle of 187–192. C.E., et al., 2008b. Evaluation of a new
horses. Am. J. Vet. Res. 68, 126–1231. Rivero, J.L.L., Serrano, A.L., Henckel, P., et al., strategy to modulate skeletal development
Riggs, C.M., Lanyon, L.E., Boyde, A., 1993. 1993. Muscle fiber type composition and in racehorses by imposing track-based
Functional associations between collagen fiber size in successfully and unsuccessfully exercise during growth: the effects on
fibre orientation and locomotor strain endurance-raced horses. J. Appl. Physiol. 2- and 3-year-old racing careers. Equine
direction in cortical bone of the equine 75, 1758–1766. Vet. J. 40, 119–127.
radius. Anat. Embryol. (Berl.) 187, Rivero, J.L.L., Serrano, A.L., Quiroz-Rothe, E., Rogers, C.W., Rivero, J.L.L., van Breda, E.,
231–238. et al., 2001. Co-ordinated changes of et al., 2007. Describing workload and
Riley G., 2004. The pathogenesis of kinematics and muscle fibre properties scientific information on conditioning
tendinopathy. A molecular perspective. with prolonged endurance training. Equine horses. Equine Comp. Exerc. Physio. 4,
Rheumatology (Oxford) 43, 131–142. Vet. J. Suppl. 33, 104–108. 1–6.
Rivero, J.L.L., 1996. Muscle biopsy as a tool Rivero, J.L.L., Sporleder, H.P., Quiroz-Rothe, E., Rome, L.C., Sosnicki, A.A., Gobble, D.O.,
for assessing muscular adaptation to et al., 2002. Oral L-carnitine combined 1990. Maximum velocity of shortening of
training in horses. Am. J. Vet. Res. 57, with training promotes changes in three fibre types from horse soleus muscle:
1412–1416. skeletal muscle. Equine Vet. J. Suppl. 34, implications for scaling with body size.
Rivero, J.L.L., 1997. Muscle variables. In: 269–274. J. Physiol. (Lond) 431, 173–185.
Lindner, A. (Ed.), Performance diagnosis of Rivero J.L.L., Talmadge, R.J., Edgerton, V.R., Ronéus, M., Essén-Gustavsson, B., Arnason, T.,
horses. Wageningen Pers, Wageningen, The 1996a. Myosin heavy chain isoforms in 1993. Longitudinal changes of muscle
Netherlands, pp. 44–71. adult equine skeletal muscle: an characteristics and racing performance in

302
 References

standardbred trotters. J. Equine Vet. Sci. 13, Serrano, A.L., Rivero, J.L.L., 2000. Myosin immature and adult equine digital tendons.
355–359. heavy chain profile of equine gluteus J. Anat. 211, 325–334.
Ronéus, M., Essén-Gustavsson, B., Lindholm, medius muscle following draught-exercise Stephens, P.R., Nunamaker, D.M., Buitterweck,
A., et al., 1987. A field study of circulatory training and detraining. J. Muscle Res. Cell D.M., 1989. Application of a Hall-effect
response and muscle characteristics in Motil. 21, 235–245. transducer for measurement of tendon
young standardbreds. In: Gillespie, J.R., Sherman, K.M., Miller, G.J., Wronski, T.J., strains in horses. Am. J. Vet. Res. 50,
Robinson, N.E. (Eds.), Equine exercise Colahan, P.T., Brown, M., Wilson, W., 1089–1095.
physiology 2. ICEEP Publications, Davis, 1995. The effect of training on equine Stover, S.M., Pool, R.R., Martin, R.B., Morgan,
CA, pp. 376–383. metacarpal bone breaking strength. Equine J.P., 1992. Histological features of the
Ronéus, M., Essén-Gustavsson, B., Lindholm, Vet. J. 27, 135–139. dorsal cortex of the third metacarpal bone
A., et al., 1992. Skeletal muscle Simpson, G.G., 1951. Horses. Oxford mid-diaphysis during postnatal growth in
characteristics in young trained and University Press, New York. thoroughbred horses. J. Anat. 181,
untrained standardbred trotters. Equine Sinha, A.K., Ray, S.P., Rose, R.J., 1991. Effect of 455–469.
Vet. J. 24, 292–294. training intensity and detraining on Strand, E., Braathen, L.C., Hellsten, M.C.,
Ronéus, M., Lindholm, A., Asheim, A., 1991. adaptations in different skeletal muscles. Huse-Olsen, L., Bjornsdottir, S., 2007.
Muscle characteristics in thoroughbreds of In: Persson, S.G.B, Lindholm, A., Jeffcott, Radiographic closure time of appendicular
different ages and sexes. Equine Vet. J. 23, L.B. (Eds.), Equine exercise physiology 3. growth plates in the Icelandic horse. Acta
207–210. ICEEP Publications, Davis, CA, pp. Vet. Scand. 49, 19.
Ronéus, N., Essén-Gustavsson, B., Johnston, 223–230. Suwannachot, P., Verkleij, C.B., Kocsis, S., et
C., et al., 1995. Lactate response to Sinha, A.K., Ray, S.P., Rose, R.J., 1993. Effect of al., 2001. Specificity and reversibility of the
maximal exercise on the track: relation to constant load training on skeletal muscle training effects on the concentration of
muscle characteristics and kinematic histochemistry of thoroughbred horses. Na+,K+-ATPase in foal skeletal muscle.
variable. Equine Vet. J. Suppl. 18, 191–194. Res. Vet. Sci. 54, 147–159. Equine Vet. J. 33, 250–255.
Ronéus, N., Essén-Gustavsson, B., Lindholm, Sjöström, M., Lexell, J., Eriksson, A., et al., Suwannachot, P., Verkleij, C.B., Weijs, W.A., et
A., et al., 1994. Plasma lactate response to 1991. Evidence of fibre hyperplasia in al., 1999. Effects of training on the
submaximal and maximal exercise tests human skeletal muscles from healthy concentration of Na+,K+-ATPase in foal
with training, and its relationship to young men? A left-right comparison in the muscle. Equine Vet. J. Suppl. 31, 101–105.
performance and muscle characteristics in fibre number in whole anterior tibialis
Tanner, J.C., Rogers, C.W., Firth, E.C., 2013.
standardbred trotters. Equine Vet. J. 26, muscles. Eur. J. Appl. Physiol. 62, 301–304.
The association of two-year-old training
117–121. Smith, R.K., Birch, H., Patterson-Kane, J., Firth, milestones with career length and racing
Rossdale, P.D., Hopes, R., Wingfield Digby, E.C., Williams, L., Cherdchutham, W., et success in a sample of thoroughbred horses
N.J., Offord, K., 1985. Epidemiological al., 1999. Should equine athletes in New Zealand. Equine Vet. J. 45, 20–24.
study of wastage among racehorses 1982 commence training during skeletal
Tidswell, H.K., Innes, J.F., Avery, N.C., Clegg,
and 1983. Vet. Rec. 116, 66–69. development?: changes in tendon matrix
P.D., Barr, A.R., Vaughan-Thomas, A., et al.,
associated with development, ageing,
Säämämen, A.M., Kiviranta, I., Jurvelin, J., 2008. High-intensity exercise induces
function and exercise. Equine Vet. J. Suppl.
Helminen, H.J., Tammi, M., 1994. structural, compositional and metabolic
30, 201–209.
Proteoglycan and collagen alterations in changes in cuboidal bones – findings from
Smith, R.K.W., 2008. Mesenchymal stem cell an equine athlete model. Bone 43,
canine knee articular cartilage following
therapy for equine tendinopathy. Disabil. 724–733.
20 km daily running exercise for 15 weeks.
Rehabil. 30 (20–22), 1752–1758.
Connect. Tissue Res. 30, 191–201. Todhunter, R.J., 1996. Anatomy and
Smith, R.K.W., Gerard, M., Dowling, B., Dart,
Schryver, H.F., Bartel, D.L., Langrana, N., Lowe, physiology of synovial joints. In:
A.J., Birch, H.L., Goodship, A.E., 2002.
J.E., 1978. Locomotion in the horse; McIlwraith, C.W., Trotter, G.W. (Eds.), Joint
Correlation of cartilage oligomeric matrix
kinematics and external and internal forces disease in the horse. Philadelphia,
protein (COMP) levels in equine tendons
in the normal equine digit in the walk and Saunders, pp. 1–28.
with mechanical properties: a proposed
trot. Am. J. Vet. Res. 39, 1278–1733. Todhunter, R.J., Lust, G., 1990.
role for COMP in determining function-
Scott, J.E., 1995. Extracellular matrix, specific mechanical characteristics of Pathophysiology of synovitis: clinical signs
supramolecular organisation and shape. J. locomotor tendons. Equine Vet. J. Suppl. and examination in horses. Comp. Cont.
Anat. 187, 259–269. 34, 241–244. Educ. Pract. Vet. 12, 980–992.
Sell, D.R., Nagaraj, R.H., Grandhee, S.K., Smith, R.K.W., Goodship, A.E., 2008. The Tsuzaki M., Yamauchi, M., Banes, A. J., 1993.
Odetti, P., Lapolla, A., Fogarty, I., et al., effect of early training and the adaptation Tendon collagens: extracellular matrix
1991. Pentosidine: a molecular marker for and conditioning of skeletal tissues. Vet. composition in shear stress and tensile
the cumulative damage to proteins in Clin. North Am. Equine Pract. 24, 37–51. components of flexor tendons. Connect.
diabetes, aging, and uremia. Diabetes Tissue Res. 29, 141–152.
Smith, R.K.W., Zunino, L., Webbon, P.M.,
Metab. Rev. 7, 239–251. Heinegård, D., 1997. The distribution of Tyler, C.M., Golland, L.C., Evans, D.L., et al.,
Selvanetti, A., Cipolla, M., Puddu, G., 1997. Cartilage Oligomeric Matrix Protein 1998. Skeletal muscle adaptations to
Overuse tendon injuries: basic science and (COMP) in tendon and its variation with prolonged training, overtraining and
classification. Oper. Tech. Sports Med. 5, site, age, and load. Matrix Biol. 16, detraining in horses. Pflüg Arch. Eur. J.
110–117. 255–271. Physiol. 436, 391–397.
Serrano, A.L., Murgia, M., Pallafacchina, G., et Snow, D.H., Valberg, S.J., 1994. Muscle Umemura, Y., Ishiko, T., Yamauchi, T., Kurono,
al., 2001. Calcineurin controls nerve anatomy, physiology and adaptations to M., Mashiko, S., 1997. Five jumps per day
activity-dependent specification of slow exercise and training. In: Hodgson, D.R., increase bone mass and breaking force in
skeletal muscle fibers but not muscle Rose, R.J. (Eds.), The athletic horse: rats. J. Bone Miner. Res. 12, 1480–1485.
growth. Proc. Nat. Acad. Sci. USA 98, principles and practice of equine sports Valberg, S.J., 1996. Muscular causes of exercise
13108–13113. medicine. Saunders, Philadelphia, pp. intolerance in horses. Vet. Clin. North. Am.
Serrano, A.L., Quiroz-Rothe, E., Rivero, J.L.L., 145–179. Equine Pract. 12, 495–515.
2000. Early and long-term changes of Stanley, R.L., Fleck, R.A., Becker, D.L., Van den Hoogen, B.M., van de Lest, C.H.A.,
equine skeletal muscle in response to Goodship, A.E., Ralphs, J.R., Patterson- van Weeren, P.R., Lafeber, F.P., Lopes-
endurance training and detraining. Pflüg. Kane, J.C., 2007. Gap junction protein Cardozo, M., van Golde, L.M., et al., 1998.
Arch. Eur. J. Physiol. 441, 263–274. expression and cellularity: comparison of Loading-induced changes in synovial fluid

303
 13 The response of musculoskeletal tissues to exercise

affect cartilage metabolism. Br. J. clinical implications for diabetes and aging. and mechanical properties of tendon. PhD
Rheumatol. 37, 671–676. Lab. Invest. 70, 138–151. thesis, University of Bristol, UK.
Van den Hoogen, B.M., van de Lest, C.H.A., Vlassara, H., Li, Y.M., Imani, F., Wojciechowicz, Wilson, A.M., Goodship, A.E., 1994. Exercise-
van Weeren, P.R., van Golde, L.M.G., D., Yang, Z., Liu, F.T., et al., 1995. induced hyperthermia as a possible
Barneveld, A., 1999. Effect of exercise on Identification of galectin-3 as a high- mechanism for tendon degeneration. J.
proteoglycan metabolism of articular affinity binding protein for advanced Biomech. 27, 899–905
cartilage in growing foals. Equine Vet. J. glycation end products (EGE): a new Wilson, A.M., McGuigan, M.P., Su, A., van den
Suppl. 31, 62–66. member of the AGE-receptor complex. Mol. Bogert, A.J., 2001. Horses damp the spring
Van Turnhout, M.C., Haazelager, M.B., Gijsen, Med. 1, 634–646. in their step. Nature 414, 895–899.
M.A., Schipper, H., Kranenbarg, S., van Von der Mark K., 1981. Localisation of Wilson, J.A., Kronfeld, D.S., Gay, L.S., et al.,
Leeuwen, J.L., 2008. Quantitative collagen types in tissues. Int. Rev. Connect. 1998. Sarcoplasmic reticulum responses to
description of collagen structure in the Tissue Res. 9, 265–324. repeated sprints are affected by
articular cartilage of the young and adult Votion, D.M., Navet, R., Lacombe, V.A., et al., conditioning of horses. J. Anim. Sci. 76,
equine distal metacarpus. Anim. Biol. 58, 2007. Muscle energetic in exercising horses. 3065–3071.
353–370. Equine Comp. Exerc. Physiol. 4, 105–118. Wolff, J., 1892. Das Gesetz der Knochen.
Van Weeren, P.R., Barneveld, A., 1999a. Study Wang J.H., 2006. Mechanobiology of tendon. Hirschwald, Berlin.
design to evaluate the influence of exercise J. Biomech. 39, 1563–1582. Woo, S.L., Gomez, M.A., Woo, Y.K., Akeson,
on the development of the musculoskeletal Wassen, M.H., Lammens, J., TeKoppele, J.M., W.H., 1982. Mechanical properties of
system of foals up to age 11 months. Sakkers, R.J., Liu, Z., Verbout, A.J., et al., tendons and ligaments. II. The
Equine Vet. J. Suppl. 31, 4–8. 2000. Collagen structure regulates fibril relationships of immobilization and
Van Weeren, P. R., Barneveld, A., 1999b. The mineralization in osteogenesis as revealed exercise on tissue remodeling. Biorheology
effect of exercise on the distribution and by cross-link patterns in calcifying callus. J. 19, 397–408.
manifestation of osteochondrotic lesions in Bone Miner. Res. 15, 1776–1785. Woo, S.L., Kuei, S.C., Amiel, D., Gomez, M.A.,
the warmblood foal. Equine Vet. J. Suppl. Watkins, J.P., 1999. Tendon and ligament Hayes, W.C., White, F.C., et al., 1981.
31, 16–25. biology. In: Auer, J.A., Stick, J.A. (Eds.), The effect of prolonged physical training
Van Weeren, P.R., Brama, P.A.J., 2003. Equine Equine surgery, 2nd edn. W.B. Saunders, on the properties of long bone: a study of
joint disease in the light of new Philadelphia, pp. 704–711. Wolff’s Law. J. Bone Joint Surg. Am. 63,
developments in articular cartilage research. Wijnberg, I.D., van Dam, K.G., de Graaf- 780–787.
Pferdeheilk. 19, 336–344. Roelfsema, E., et al., 2008. (Over)training Woo, S.L.Y., Gomez, M.A., Sites, T.J., Newton,
Van Weeren, P.R., Firth, E.C., Brama, P.A.J., effects on quantitative electromyography P.O., Orlando, C.A., Akeson, W.H., 1987.
2010. To move or to perish: the importance and muscle enzyme activities in The biomechanical and morphological
of exercise during musculoskeletal standardbred horses. J. Appl. Physiol. 105, changes in the medial collateral ligament
development in the horse. Pferdeheilk 26, 1746–1753. of the rabbit after immobilization and
581–587. Williams I.F., Heaton, A., McCullagh, K. G., remobilization. J. Bone Joint Surg. 69A,
Van Weeren, P.R., Firth, E.C., Brommer, B., 1980. Cell morphology and collagen types 1200–1211.
Hyttinen, M.M., Helminen, A.E., Rogers, in equine tendon scar. Res. Vet. Sci. 28, Yamano, S., Eto, D., Hiraga, A., Miyata, H.,
C.W., et al., 2008. Early exercise advances 302–310. 2006. Recruitment pattern of muscle
the maturation of glycosaminoglycans and Williams, R.B., Harkins, L.S., Hammond, C.J., fibre type during high intensity
collagen in the extracellular matrix of Wood, J.L., 2001. Racehorse injuries, exercise (60–100% VO2max) in
articular cartilage in the horse. Equine Vet. clinical problems and fatalities recorded on thoroughbred horses. Res. Vet. Sci.
J. 40, 128–135. British racecourses from flat racing and 80, 109–115.
Verzijl, N., DeGroot, J., Bank, R.A., Bayliss, National Hunt racing during 1996, Yamano, S., Eto, D., Sugiura, T., et al., 2002.
M.T., Bijlsma, J.W.J., Lafeber, F.P.J.G., et al., 1997 and 1998. Equine Vet. J. 33, Effect of growth and training on muscle
2001. Age-related accumulation of the 478–486. adaptation in thoroughbred horses. Am. J.
advanced glycation end product Williams, R.S., Neufer, P.D., 1996. Regulation Vet. Res. 63, 1408−1412.
pentosidine in human articular cartilage of gene expression in skeletal muscle by Yan, Z., Serrano, A.L., Schiaffino, S., et al.,
aggrecan: The use of pentosidine levels as a contractile activity. In: Rowell, L.B., 2001. Regulatory elements governing
quantitative measure of protein turnover. Shepherd, J.T. (Eds.), Handbook of transcription in specialized myofiber
Matrix Biol. 20, 409–417. physiology: integration of motor, subtypes. J. Biol. Chem. 276,
Verzijl, N., DeGroot, J., Thorpe, S.R., Bank, circulatory, respiratory and metabolic 17361–17366.
R.A., Shaw, J.N., Lyons, T.J., et al., 2000. control during exercise. American Young, D.R., Nunamaker, D.M., Markeld,
Effect of collagen turnover on the Physiological Society, Bethesda, MD, pp. M.D., 1991b. Quantitative evaluation of the
accumulation of advanced glycation end 1124–1150. remodeling response of the proximal
products. J. Biol. Chem. 275, Wilmink, J., Wilson, A.M., Goodship, A.E., sesamoid bones to training-related stimuli
39027–39031. 1992. Functional significance of the in thoroughbreds. Am. J. Vet. Res. 52,
Viguet-Carrin, S., Garnero, P., Delmas, P. D., morphology and micro-mechanism of 1350–1356.
2006. The role of collagen in bone collagen fibres in relation to partial rupture Young, D.R., Richardson, D.W., Markel, M.D.,
strength. Osteoporos. Int. 17, 319–336. of the superficial digital flexor tendon in Nunamaker, D.M., 1991a. Mechanical and
Vlassara, H., Bucala, R., Striker, L., 1994. racehorses. Res. Vet. Sci. 53, 354–359. morphometric analysis of the third carpal
Pathogenic effects of advanced Wilson, A.M., 1991. The effect of exercise bone of thoroughbreds. Am. J. Vet. Res. 52,
glycosylation: biochemical, biological, and intensity on the biochemistry, morphology 402–409.

304
 CHAPTER 14 

Performance in equestrian sports

Hilary M. Clayton, P. René van Weeren

Introduction Within any gait, speed is the product of stride length and stride
frequency. In sprinting races, a high stride frequency is the prime
requirement; the ability to take longer strides becomes increasingly
The use of horses for different occupations resulted in selective important as the distance of the race increases (Deuel & Park, 1990).
breeding to produce different morphological types in accordance Stride length has been reported to show a rapid and fairly linear
with specific intended uses. Massive draught horses could haul increase with speed, whereas stride frequency increased more slowly
heavy guns and carry medieval knights with their cumbersome and and in a non-linear manner (Dušek et al., 1970; Leach & Cymbaluk,
bulky armor, whereas slender, more agile Arabians and Thorough- 1986). In Thoroughbreds galloping at 13.7–19.8 m/s which is
breds displayed greater speed and endurance. Today, horses compete approaching maximal speed, Seder and Vickery (2003) reported a
is a diverse range of sports each requiring specific athletic talents. linear increases in stride length and stride frequency.
For sprinting sports the prime requirements are rapid acceleration
and the ability to generate a high maximum speed. Middle distance
racing and endurance racing call for stamina to maintain submaxi- Stride length
mal speed over a longer distance. Other sports depend on different Stride length increases with effective limb length. The forelimbs are
degrees of technical skills, visual acuity, fast reflexes or esthetics of generally considered to rotate around a point close to the tuber
movement. spinae scapulae and the attachment of serratus ventralis, though the
Over the past half-century, breeders have produced performance precise point of rotation may vary between individuals. The hind
horses with highly specialized athletic abilities, such as elegant gaits limbs rotate around the hip joint in the symmetrical gaits (walk,
and efficient jumping technique. Consequently, today’s sport horses trot, pace) and around the lumbosacral joint in the asymmetrical
can truly be described as ‘equine athletes’, which emphasizes the gaits (canter, gallop). The advantage of moving the rotation point
importance of their locomotor apparatus. This chapter will consider to the lumbosacral joint is that it increases the effective limb length
the mechanical factors involved in different types of equine athletic (Fig. 14.1). Interestingly, a necropsy study showed that 32% of
performance under the general headings of racing sports, sport Thoroughbreds had maximal range of motion in flexion–extension
horses and Western sports. at the joint between the fifth and sixth lumbar vertebrae rather than
at the lumbosacral joint, and 8% had evidence of sacralization of
the sixth lumbar vertebra (Stubbs et al., 2006). Moving the effective
Racing sports lumbosacral articulation more cranially might confer an advantage
in terms of further lengthening the hind limb. In the symmetrical
Racing encompasses a broad spectrum of sports that occur over gaits, particularly the walk, lateral bending of the vertebral column
distances from as short as 200 m to greater than 160 km. The may enhance stride length, whereas in the asymmetrical gaits dor-
common denominator in all these sports is that the winner main- soventral flexion and extension make a larger contribution to stride
tains the highest average speed over the distance of the race. In length (Hildebrand, 1962). In horses, the thoracolumbar spine is
sprinters, generation of high stride rates is of prime importance. fairly rigid to provide support for the large body mass and facilitate
Stride length assumes greater significance for horses racing over transmission of propulsive forces from the hind limbs. The need
middle distances. Economy of movement and energy efficiency are for spinal stability is a priority in gaits that have a suspension
prime considerations in endurance racing competitors. phase, which include the racing trot, pace and gallop. Thus, the
need for stabilization predominates and intervertebral motion
makes a relatively small contribution to stride length in racehorses
Speed with the exception of the lumbosacral joint. This is in contrast
Speed is the determining factor in racing, and is also a desirable to smaller animals, such as cats, that make use of a more flexible
characteristic in many other sports. Horses change speed by altering thoracolumbar spine to increase stride length. For more informa-
the spatial and temporal relationships between the limbs to produce tion on vertebral kinematics, the reader is referred to the chapter on
different gaits and to vary the extension within a gait. Each horse neck and back movements (Chapter 10).
has an optimal speed within a gait at which the metabolic cost
is minimized; slower and faster speeds both result in a higher meta-
bolic cost (Hoyt & Taylor, 1981). It has been suggested that transi- Stride frequency
tions between gaits may be triggered by energetic cost (Hoyt & Stride frequency describes the rate at which the limbs are protracted
Taylor, 1981) or musculoskeletal forces (Farley & Taylor, 1991). and retracted. The ability to cycle the limbs rapidly is favored by
Whatever the reason, horses naturally select particular speeds at having a high percentage of fast-twitch fibers in the extrinsic
which to make transitions between gaits. muscles, which influences both the rapidity with which the limbs

305
 14 Performance in equestrian sports

Fig 14.1  Rotation points of the fore and hind limbs are represented by asterisks for the trot (left) and the gallop (right). The forelimb rotates around a point
in the upper part of the scapula. The hind limb rotates around the hip joint in trot (left) and around the lumbosacral joint in asymmetrical gaits (right).
Reprinted from The Dynamic Horse with permission of Sport Horse Publications, Mason, MI, USA.

are moved and the ability to generate large forces and impulses
during the stance phase. Stance duration is negatively correlated 0.12
with the diameter of the muscle fibers (Rivero & Clayton, 1996),
whereas stride length is positively correlated with the percentages
of type I and type IIa muscle fibers, and negatively correlated with
the percentage of type IIb muscle fibers in Standardbreds. Stride
frequency is positively correlated with the percentage of type IIa
muscle fibers (Persson et al., 1991). Stance duration has a negative 0.08

Height (m)
correlation with the percentage of type IIb muscle fibers (Roneus
et al., 1995) that is indicative of an ability to generate high ground
reaction forces (GRFs) and so create the necessary impulse over a
shorter period of time.

Efficiency of movement 0.04

During locomotion, energy is used to move the horse’s center of

mass and to cycle the limbs back and forth. For sprint races, energy
expenditure is relatively unimportant, but as race distance increases 2 3 4 5 6
beyond the capacity of the anaerobic energy production systems, Speed (m/s)
energetic efficiency has a greater effect on the ability to maintain
submaximal speed over a distance. Factors that contribute to loco- Fig 14.2  Peak height of the hind hoof during the swing phase of the stride
motor economy include the masses and inertial properties of the in five horses trotting at a range of velocities. Each horse is represented by a
limb segments, kinematics, storage and release of elastic strain different type of data point.
energy and inter-segmental transfers of kinetic energy (Cavagna From Clayton et al. (2002).
et al., 1977).
Segment mass, measured as a percentage of body mass, and its
distribution within the segment determine its inertial properties. In elastic energy storage and release as the spine flexes and extends
horses, heavy muscular tissue is confined to the proximal limb, (Minetti et al., 1999).
while the distal limb is composed of less dense tissue including
skin, connective tissue, ligament, tendon and bone. This proximal
concentration of mass reduces the moment of inertia of the limb
Quarter Horse racing
and thus facilitates protraction and retraction and reduces the ener- The Quarter Horse is the elite equine sprinter and was named for
getic expenditure. its ability to sprint over a distance of a quarter mile. The most
Protraction and retraction of the limbs uses energy and several important qualities of a sprinter are rapid acceleration and a high
mechanisms are employed to improve energetic efficiency. The limb maximal speed. In sprinters speed is more strongly influenced by
is folded during protraction, which brings the distal limb closer to stride frequency than stride length (Deuel & Lawrence, 1986). This
the pivot point, thereby reducing its inertia. The amount of joint is in contrast to middle distance Thoroughbreds in which speed is
flexion and limb folding increase with speed with a consequently altered primarily by changes in stride length (Ratzlaff et al., 1985).
higher hoof flight arc (Clayton et al., 2002) (Fig. 14.2). Storage and Comparing Quarter Horses and Thoroughbreds galloping at 15 m/s,
release of elastic strain energy greatly reduce the metabolic energy the stride length of the Quarter Horses is shorter by 1.0–2.0 m
requirement (Alexander, 2002). In the first half of the stance phase, (Deuel & Lawrence, 1986), with a correspondingly higher stride
the elastic elements are stretched, thereby storing elastic energy. frequency. The stride variables of 2-year-old Quarter Horse fillies
As the body mass rolls over the limb, the elastic tissues recoil, releas- galloping at a mean speed of 13.3 m/s are shown in Table 14.1.
ing the stored elastic energy, which assists in protracting (e.g. biceps
brachii) and raising (e.g. superficial digital flexor tendon) the limb.
During trotting the fetlock, carpal, elbow and shoulder joints show Stride variables
elastic behavior in the soft tissues that cross these joints (Clayton There is no significant covariance between stride frequency and
et al., 1998). The elastic contributions are relatively small at the stride length in galloping Quarter Horses, so these variables are
walk (Clayton et al., 2000) but are much larger during trotting and regarded as independent factors within the time domain and space
in cantering and galloping. In the asymmetrical gaits, the elastic domain (Deuel & Lawrence, 1986). As speed increases, stride fre-
lumbodorsal fascia may also contribute to gait efficiency through quency is increased by shortening the single support periods of the

306
 Racing sports

trailing hind limb and leading forelimb, and the duration of the stride. Use of the whip did not change the speed, but the horses
suspension. Linear regression analysis of the data predicts that at maintained their speed with a reduced stride length and a higher
14.9 m/s, overlap between the leading hind limb and trailing fore- stride frequency during urging. Also, the stance duration of the
limb approaches zero. Interestingly, at approximately this speed, forelimbs was reduced (Deuel & Lawrence, 1987b).
two of four horses in the study showed strides with an extended
suspension (i.e. a period of suspension between lift-off of the
leading hind limb and ground contact of the trailing forelimb) of Thoroughbred racing
4.0–8.0 ms duration (Deuel & Lawrence, 1986). At a constant stride Thoroughbreds race over short to middle distances, mostly in the
frequency, stride length is adjusted by increasing the distances range of one to two miles, though shorter and longer races exist.
between footfalls of the trailing and leading forelimbs and between Over the shorter distances, the requirements for acceleration and
the leading forelimb and the trailing hind limb. In Quarter Horse speed are similar to those for Quarter Horse racing, but as the dis-
foals there is a significant correlation between stride length and tance increases stamina assumes more importance. In this breed
mass (Leach & Cymbaluk, 1986). Small foals prefer to change speed high maximal speeds are achieved primarily as a result of having
by adjusting stride length, whereas foals with larger dimensions a long stride length (Leach et al., 1987). Top class Thoroughbreds
prefer to adjust their stride frequency. galloping at racing speed have had their stride lengths measured at
7.38 m for Secretariat and 6.66 m for Riva Ridge (Pratt & O’Connor,
Sidedness 1976). However, individual horses vary in terms of their prefer­
ence to increase stride length or stride frequency as they approach
Two-year-old Quarter Horse fillies show a leading limb preference; top speed. In general, for Thoroughbreds galloping at submaxi­
approximately twice as many strides were recorded on the right lead mal speeds, stride length tends to show a decreased rate of increase
versus the left lead in spite of the horses being cued for the left and whereas stride frequency tends to show an increased rate of
right leads an equal number of times. The horses either picked up increase (Yamanobe et al., 1992). Stride lengths and stride frequen-
the right lead during the transition or changed from the left to the cies for Thoroughbreds galloping at a range of speeds are shown in
right lead (Deuel & Lawrence, 1987a). Both speed and stride length Table 14.1 (Ishii et al., 1989).
were significantly longer on the left lead than the right lead, but
stride frequency did not differ between the two leads. The trailing
forelimb had a longer stance duration and a shorter period of Stride variables
overlap with the leading forelimb on the left lead, which was inter- A study of part-Thoroughbred horses galloping on a variety of
preted as a sign of an increased reliance on the right forelimb for surfaces (sand, turf, woodchips, dirt) showed a linear relationship
support. The hind limbs, however, did not show any linear or tem- between stride length and speed, with stride length increasing from
poral asymmetries between the left and right leads. about 3.5 m at 6 m/s to around 7.5 m at 19 m/s. At the same time,
stride duration decreased from 580 ms at 6 m/s to 440 ms at
Effect of urging 19 m/s. The reduction in stride frequency was achieved through a
large reduction in the swing phase from 200 to 100 ms, and a
The effect of urging by the rider was studied by having the rider use smaller reduction in the stance phase from 400 to 350 ms (Hel-
a whip on the shoulder of the leading forelimb in rhythm with the lander et al., 1983).
Stride length at the gallop comprises the sum of four inter-limb
distances: the hind step, the diagonal step, the fore step and the
Table 14.1  Mean values for stride variables of galloping Quarter suspension distance (Fig. 14.3). Alterations in stride length are
Horses (QH) and Thoroughbreds (TB) associated with greater changes in the diagonal distance and the
suspension distance than in the hind and fore steps (Ishii et al.,
Speed (m/s) Stride Stride frequency Breed 1989). The suspension distance shows the greatest increase at mod-
length (m) (strides/s) erate speeds but has a tendency to level off as maximal speed is
approached, whereas the diagonal step shows an increasing rate of
13.3 5.06 2.62 QH increase at higher speeds. Lengthening of the diagonal step is a
consequence of an increased propulsive force from the hind limbs
6.6 3.63 1.82 TB
but there may be an associated increase in forelimb braking force,
9.9 4.89 2.02 TB which, in turn, limits the suspension distance (Yamanobe et al.,
1992).
13.8 6.18 2.23 TB The temporal variables for the gallop stride include periods of
15.8 6.66 2.37 TB single support when one limb only is in the stance phase, periods
of overlap when two or more limbs are in the stance phase simul-
19.8 7.77 2.55 TB taneously and one or more suspension phases when the horse is
airborne. Duration of the suspension phase is highly correlated with
Data for Quarter Horses from Deuel and Lawrence (1986) and data for Thorough-
stride duration. For Thoroughbreds galloping at 15 m/s, the suspen-
breds are from Ischii et al. (1989) and Seder and Vickery (2003).
sion phase occupies 28% of stride duration (Leach et al., 1987).

Hind Diagonal Fore Suspension

step distance step distance
Fig 14.3  Interlimb distances in the gallop. The sum of the interlimb distances is the stride length.

307
 14 Performance in equestrian sports

The stance phase durations of the four limbs differ significan­ (the stride in which the forelimb lead is changed) the sequence
tly from each other, with the leading hind limb and trailing fore- of limb placements follows a rotary sequence. Sometimes the hind
limb having shorter stance durations than the contralateral limbs. limb lead is not changed until several strides later, in which case
Overlap is longest between the two hind limbs, shortest between the rotary sequence of limb placements is maintained during the
the two forelimbs, and of intermediate duration between the intervening strides; this is also known as being disunited or on a
leading hind-trailing forelimb pair. As galloping speed increases crossed lead.
there is a marked reduction in all the overlap durations, with
overlap between the leading hind and trailing forelimbs decreasing
linearly down to about 50 ms (Hellander et al., 1983). Comparison
Acceleration
between the extended canter of horses trained for dressage with During acceleration, the initial increase in speed is due to a
horses trained for racing reveals that, at the same speed, there is rapid increase in stride frequency accompanied by a relatively slow
no difference in stride length or stride frequency, but the race­ increase in stride length (Hiraga et al., 1994). Stride frequency
horses had significantly shorter stance durations and overlap times peaks within a few strides after leaving the starting gate, whereas
(Clayton, 1993), which may be a consequence of race training. Fol- stride length requires 25–30 strides to reach its maximal value. The
lowing 8 weeks of high-intensity training on a treadmill, the stance diagonal distance increases relatively rapidly to reach maximal
duration of galloping Thoroughbreds was reduced by 8–20% length within the first few strides. The airborne distance increases
(Corley & Goodship, 1994). These changes can be ascribed to linearly after the first few strides until the 20th stride. Forelimb step
greater muscular strength; the ability to generate higher ground length and the hind limb step length tend to level off after the 20th
reaction forces allows the necessary impulses to be created during stride. Hind step length is very short in the initial strides during
a shorter stance phase, and the greater forces project the horse into which the horses use a half bound before establishing a leading
a prolonged suspension phase. Race training also results in increases limb for the hind limb pair. The majority of horses use a rotary
in stride duration and stride length (Leach et al., 1987). sequence of limb placements (crossed lead) during the initial accel-
The advanced placements (time between footfalls) change in eration before establishing the normal transverse sequence (Kai &
such a manner that, at faster galloping speeds, the hind limbs act Kubo, 1993; Hiraga et al., 1994). Horses tend to hold their breath
more in unison with each other while becoming more dissociated for 3–4 s immediately after leaving the starting gate after which they
from the forelimbs. Thus the rhythm becomes more like a rabbit settle into a breathing rhythm that is synchronized with the stride
hop. As the advanced placement between the leading hind and the cycle (Kai & Kubo, 1993).
trailing forelimbs increases, the distance between placements of
these limbs also increases. Dissociation of the actions of the hind
limbs and forelimbs causes a greater reliance on spinal flexion and
Ground reaction forces
extension. Thoroughbreds traveling at speeds greater than 13.5 m/s Vertical ground reaction forces have been studied at racing speed in
may show a short extended suspension between lift-off of the horses galloping round a 0.8-km track with well-banked turns and
leading hind limb and contact of the trailing forelimb. A third wearing instrumented shoes on all four feet (Ratzlaff et al., 1987).
suspension, between lift-off of the trailing forelimb and contact of On the straightaway the greatest vertical force was exerted by the
the leading forelimb, has also been recorded occasionally. The per- leading forelimb, followed by the leading hind limb, trailing hind
centage of Thoroughbreds showing multiple suspension phases limb and trailing forelimb. On the banked turns, the greatest verti-
increases with speed from 27% of horses at 15.5 m/s, to 50% at cal force was exerted on the leading forelimb, followed by the trail-
17.0 m/s, and 70% at 19.5 m/s. The duration of these suspension ing forelimb, leading hind limb and trailing hind limb. It is not
phases also tends to increase with speed (Seder & Vickery, 2003). surprising, therefore, that the majority of racing injuries occur in
the leading limbs, especially the leading forelimb.
Head and neck motion
The head and neck, which comprise approximately 10% of body Limb kinetics
weight, move in synchrony with the limbs. The neck is actively An early modeling study indicated that each forelimb would
extended and lowered during the hind limb stance phases with the support 170% body weight when galloping at 14 m/s (Kingsbury
head reaching its lowest point during the propulsive phase of the et al., 1978). Subsequent modeling studies, however, have esti-
leading hind limb (Bramble, 1984). This movement may counteract mated widely different forces in the superficial and deep digital
the tendency of the forequarters to rise as a result of hind limb flexor tendons (Brown et al., 2003; Swanstrom et al., 2005b). A
propulsion (Pratt, 1983). Elevation of the head and neck begins detailed musculoskeletal model for dynamic simulation of the
as the trailing forelimb contacts the ground, continues through Thoroughbred forelimb during the stance phase when galloping at
the forelimb stance phases and peaks at lift-off of the leading 18 m/s (Swanstrom et al., 2005a) indicated peak vertical ground
forelimb. reaction force of 210% body weight combined with longitudinal
forces of 30% body weight (braking) and 63% body weight (pro-
pulsion). This model also predicted palmar flexion angles of 148°
Lead changes for the distal interphalangeal joint and 169° for the proximal inter-
Horses normally use a transverse gallop in which the leading limb phalangeal joint during the first one third of stance and 262° for
is on the same side of the body for the fore and hind limb pairs. A the metacarpophlangeal joint just before midstance. Angular ranges
lead change involves a reversal in the order of placement of the of motion for these joints were 50° (distal interphalangeal joint),
contralateral limb pairs, and Thoroughbreds normally change leads 16° (proximal interphalangeal joint) and 54° (metacarpophalan-
several times during a race. The most common strategy is to use the geal joint). Peak strains in the superficial digital flexor tendon
inside lead through the turns, and the opposite lead in the straight- (4.9%) and suspensory ligament (8.5%) occurred prior to mid-
aways. Racehorses usually initiate the lead change with the fore- stance, whereas strains in the deep digital flexor tendon (2.8%) and
limbs. During the forelimb swing phase the leading forelimb the accessory ligament of the superficial digital flexor tendon (8.8%)
abbreviates the cranial part of its swing phase and is placed early peaked around midstance. The accessory ligament of the deep
to become the new trailing forelimb. The limb that was the trailing digital flexor tendon had peak strain of 7.1% around the start of
forelimb increases its swing phase duration substantially to become breakover. Transection of the accessory ligament of the superficial
the new leading forelimb. The hind limbs then change their lead in digital flexor tendon, which is used to treat superficial digital flexor
a similar manner during their next swing phase. For one stride tendonitis, resulted in increased force and strain in other soft

308
 Racing sports

tissues, increases in metacarpophalangeal joint extension and distal hoof, which increases locomotor efficiency by allowing a smoother
interphalangeal joint flexion, and a reduction in ground reaction transition from braking to propulsion, is achieved by having a
force (Swanstrom et al., 2005a). deeper cushion or by reducing the dry density of the cushion (Rat-
zlaff et al., 2005). Hoof-mounted accelerometers (Ratzlaff et al.,
2005) indicate wide variations in acceleration from stride to stride
Fatigue on the same surface.
Fatigue implies a decrement in performance shown by reductions Track surface composition and characteristics have a profound
in stride frequency and running speed. Stride length may increase effect on the loading conditions of the horse’s limb, with the shear
or decrease depending on the horse. The absolute duration of both strength of the granular material being particularly influential.
the stance and suspension phases of the stride increases, with sus- Therefore, selection of soil for structure may be more important
pension occupying a greater percentage of stride duration. Footfall than the addition of soil amendments that alter damping charac-
of the leading hind is followed more closely by that of the trailing teristics (Reiser et al., 2001). Many racing surfaces are a composite
forelimb, and overlap between these two limbs increases during of sand, rubber granules and fibers coated with a wax. Horses
fatigue (Leach & Sprigings, 1979). Since the stride and respiratory trained and raced more slowly when the surface reached a tem-
cycles are synchronized in galloping horses, it has been suggested perature of 43-46 degrees at which the lower molecular weight
that the changes in limb coordination associated with fatigue may components of the wax melt (Peterson et al., 2010). In one study,
be related to the respiratory demands. Prolongation of the suspen- a synthetic racing surface was shown to have lower peak accelera-
sion phase increases the duration of the inspiratory phase of the tions, mean vibrations and peak ground reaction forces than
respiratory cycle, which may be advantageous as the horse becomes turf or dirt surface (Setterbo et al., 2009). Track maintenance
fatigued (Leach & Sprigings, 1979). also affects the mechanical properties of the track (Peterson &
McIlwraith, 2008).

Injury
Standardbred racing
Analysis of racetrack patrol videos (Ueda et al., 1993) has shown
that breakdown injuries often occur immediately after a lead change Racing Standardbreds are divided into two groups according to the
(47% cases), use of the whip (38% cases) or an oblique movement gait they perform during racing. The tendency to trot, in which the
of the horse across the track (21% cases). diagonally opposite limb pairs move together, or to pace, in which
the lateral pair of limbs move together, has been shown to be geneti-
cally determined (Andersson et al., 2012). The offspring of pacers
Training almost always pace, whereas about 20% of the offspring of trotters
An important consideration in Thoroughbred training is the need are pacers (Cothran et al., 1987).
to stimulate skeletal adaptation without causing injury to the bones
and joints. It has been suggested that the training regime should Trotters
incorporate short sprints on a relatively frequent basis (every 3–6
days) to stimulate bone adaptation, especially with regard to the Stride variables
prevention of bucked shins (Moyer & Fisher, 1991). However, the A good trotter should have a high maximal speed with a stride
distance of the sprints must be limited because horses that accumu- frequency in excess of 2.4 strides/s and a stride length greater
late extensive distances at high speeds are more likely to suffer a than 5.45 m. It has been suggested that trotters achieve a high
fatal skeletal injury (Estberg et al., 1995). Many complete fractures racing speed by selecting an optimal stride length then accelerat-
of the long bones are preceded by incomplete fractures (Stover ing to the wire by increasing their stride frequency (Barrey et al.,
et al., 1992) indicating that they are a consequence of repeated limb 1995).
trauma during training and racing. Different studies vary in their findings regarding the relationship
between speed, stride length and stride frequency. Barrey et al.
(1995) found that both stride length and stride frequency increase
Track surface linearly with speed, whereas Drevemo et al. (1980a) found speed
Material properties of the track surface affect both performance and to be moderately correlated with stride length but not correlated
lameness. On a harder surface there is a reduction in stride duration, with stride duration. The latter study also found a close correlation
whereas on a more compliant surface stride duration increases between stride duration and swing duration. Swing duration,
(Fredricson et al., 1983). This offers at least a partial explanation which occupied about 75% stride, had much more effect on overall
for the fact that faster race times are recorded on firmer track sur- stride duration than stance duration, which occupied only about
faces. However, there is a price to pay in terms of soundness. The 25% of stride. The treadmill study of Weeren et al. (1993) also
incidence of lameness in Thoroughbreds that are in race training found that stride length contributes more than stride frequency to
increases with the hardness of the track surface (Cheney et al., an increase in speed, with the swing duration remaining constant
1973). One of the factors that affect hardness is the composition at all speeds.
and depth of the cushion. Hardness is reduced as the cushion gets The temporal and linear stride variables of 30 Standardbreds
deeper, but if the cushion depth is too deep (greater than 10 cm), trotting on a racetrack were studied using a camera car driven paral-
the footing becomes insecure. Surface moisture content tends to lel to the horse as a pacemaker (Drevemo et al., 1980a). With the
vary, whereas the moisture content of the compacted cushion is pacemaker car moving at 12 m/s, the actual speed of the individual
more constant (Clanton et al., 1991). There also tended to be more strides varied from 11.3 to 12.4 m/s. The values of the stride vari-
compaction around the starting chutes than in other areas of the ables are shown in Table 14.2. There was very little intra-individual
track. Next to the rail, track surface was softer than the surface variation (Drevemo et al., 1980b; Kobluk et al., 1989), indicating
toward the middle of the track. that each horse has a stable locomotion pattern that is repeated
Dynamic properties of the track, such as hardness, rebound, regularly with only minor deviations. The variation between indi-
deceleration rate, rebound rate and penetration, can be assessed viduals was two to three times greater than that within horses.
with a track-testing instrument. Accelerometers have been attached Depending on the variable being studied, the coefficients of varia-
to the hoof wall for dynamic testing of the track surface (Barrey tion were of the order of 8–12%, though the variability within
et al., 1991; Ratzlaff et al., 2005). Lower peak acceleration of the horses was about 60% less than that between horses.

309
 14 Performance in equestrian sports

For Standardbreds trotting on a treadmill, the swing phase trajec-

Table 14.2  Mean values for stride variables of Standardbred
tories of the fore hooves are quite different from those of the hind
trotters and pacers
hooves. The fore hooves are lifted higher but show less lateral devia-
tion than the hind hooves in which the limbs swing wide to avoid
Speed (m/s) Stride Stride Suspension interference. The hoof movements as seen in the frontal plane seem
– gait length (m) duration (ms) (ms) to be characteristic of the individual animal (Weeren et al., 1993).
Drevemo et al. (1987) analyzed temporal kinematic variables in
12.0 – trot 5.45 455.0 99.0
ten Standardbreds trotting at 4 m/s on a treadmill at the ages of 8,
14.2 – trot 5.65 396.8 n/a 12 and 18 months. After scaling the data to height at the withers,
a decrease in relative stride length was found.
11.5 – pace 5.57 485.0 148.6
12.0 – pace 5.57 465.9 141.2 Reproducibility of gait
The kinematic stride variables of trotting Standardbreds show good
13.1 – pace 6.04 460.0 135.3
reproducibility in the short term during both overground (Drevemo
14.0 – pace 6.28 448.4* 133.4 et al., 1980c) and treadmill (Weeren et al., 1993) locomotion. A
5-month training period did not result in any significant changes
*In the original paper the stride duration is given as 499.4 ms, and the stride in temporal parameters, but other aspects of the stride kinematics
frequency is 2.23 strides/s. Apparently, the value for the stride duration is a did change. In the forelimb there was more flexion of the elbow,
typographical error.
more overextension of the carpus and a decrease in fetlock flexion
Data for trotters moving at 12.0 m/s from Drevemo et al. (1980a and 1980b) and at after training. In the hind limb there was an increase in overexten-
14.2 m/s from Barrey et al. (1995). Data for pacers from Wilson et al. (1988b). sion of the fetlock (Weeren et al., 1993). A longer training period
of 3 years was associated with significant increases in stride length,
stride duration and swing duration. The increase in stride dura­
Good trotters show a short stance phase duration, especially in tion resulted almost exclusively from the longer swing duration
the forelimbs (Bayer, 1973). The stance phase has been subdivided (Drevemo et al., 1980c).
at the instant when the metacarpus is vertical in the forelimbs or
when the hoof is vertically beneath the hip joint in the hind limbs Ground reaction forces
(Drevemo et al., 1980a). The early part was named the restraint
phase and the later part was named the propulsion phase. Although The peak vertical ground reaction force in the forelimb of a Stan-
the names suggest that these terms define functional phases of the dardbred trotting at a relatively slow speed of 8 m/s was 99% body
stride, it should be noted that they are defined kinematically and weight and occurred around 50% stance. The braking longitudinal
so the implied functional relationship to the longitudinal ground force had a maximal value of 6.4% body weight at 15% stance,
reaction forces is an approximation only. Using this definition, the while the propulsive component peaked at 75% stance with a value
restraint phase accounts for the initial 40–45% stance, and is of 9.2% body weight. The forelimb longitudinal force changed from
slightly longer in the hind limbs than in the forelimbs. The propul- braking to propulsive around 40–50% stance (Quddus et al., 1978;
sive phase is significantly longer in the right hind than in the other Chateau et al., 2009).
limbs, which has been interpreted as an indication of sidedness.
The duration of the forelimb propulsive phase is negatively corre- Sidedness
lated with speed. By definition the trot is a symmetrical gait, which implies that the
Diagonal dissociation is a term that encompasses the dissociation movements of the left and right limbs are out of phase by 50%,
of the diagonal pair at contact with the ground (diagonal advanced with the left and right steps being equally spaced in time. In the 30
placement) and at lift-off (diagonal advanced completion). Disso- horses studied by Drevemo et al. (1980b), the kinematic variables
ciation of the movements of the diagonal pair of limbs results in a for the entire group were symmetrical on the left and right sides.
short period of single support between the periods of bipedal However, many individuals showed left–right asymmetries when
support and periods of suspension. Duration of the diagonal dis- trotting at high speeds, especially in the hind limbs. Six of 30 horses
sociation is significantly longer at lift-off than at contact, with the had significant asymmetries between the fore steps, while 12 out of
forelimb acting in advance of the hind limb both at contact and 30 horses had significant asymmetries between the hind steps
lift-off in the majority of horses. For a given diagonal, the duration (Drevemo et al., 1980b). Some individuals had such marked asym-
of the dissociation at contact and lift-off are highly correlated. metries that, rather than trotting, they showed a transitional type of
When the forelimb precedes the hind limb, diagonal advanced gait with a galloping motion of the hind limbs, which is known as
placement is inversely correlated with the distance between the the traquenard. The asymmetries detected at racing speed by slow
diagonal limb pair during bipedal support (diagonal length). When motion analysis are often not apparent during a clinical examina-
the hind limb precedes the forelimb, the diagonal advanced place- tion when the horse trots at a slower speed. Detection is important,
ment is directly correlated with the diagonal distance (Drevemo though, because an asymmetrical gait is thought to be associated
et al., 1980b). About 25% of trotters showed highly significant dif- with poor performance (Dalin et al., 1985) and the development
ferences between the left and right diagonal dissociations. of locomotor pathology (Rooney, 1969).
The mean duration of diagonal limb support at a speed of 12 m/s Even 8-month-old Standardbreds that have not yet been trained
was 99 ms, with highly significant differences between the left and show some differences between the kinematics of the left and right
right diagonals in four out of 30 horses. Diagonal length, which is limbs (Drevemo et al., 1987). The fact that asymmetries are present
the distance between the diagonal limb pair during their stance at such a young age is strong evidence for congenital laterality or
phase, had a high coefficient of variation on the right diagonal, and sidedness. Furthermore, when the same horses were re-evaluated at
the values for the left and right diagonals differed significantly in 18 months of age the differences between the left and right sides
five out of 30 horses. The mean duration of the suspension phase were even more pronounced in horses that had been trained, but
was 99 ms, with 66% of the horses showing no significant differ- not in those that were not trained in the interim.
ence between the left and right sides. The left and right suspensions An asymmetric appearance of the hindquarters has been associ-
were strongly correlated with the length and duration of the left and ated with poor racing performance (Dalin et al., 1985). Five
right hind steps, respectively. There was also a moderate correlation hundred Standardbreds in race training were evaluated for sym-
between suspension and stride duration (Drevemo et al., 1980b). metry in the height of their tubera sacrale. The left tuber sacrale

310
 Racing sports

was lower than the right one in nine horses, and the right tuber hoof. At racing speed the dissociation is about 30 ms, which repre-
sacrale was lower than the left one in 30 horses. The asymmetric sents 7% of stride duration. With this amount of dissociation, the
horses had significantly larger body size than the 461 horses that pace can be considered a four-beat gait (Wilson et al., 1988a).
did not show asymmetry. The asymmetric horses had lower total The stride lengths and stride durations of Standardbreds pacing
earnings, a lower number of races per horse and fewer good racing at a range of speeds between 11.5 and 14.0 m/s are shown in Table
records. It was concluded that horses in which the height of the 14.2. Pacing speed increases primarily as a result of an increase in
tubera sacrale is asymmetrical are less likely to become successful stride length with minimal change in stride frequency (Wilson
racehorses. et al., 1988b). Lengthening of the stride is a result of covering a
greater distance during the suspension phases, without much
Effect of track design change in the distance between the lateral pair of limbs during
Lameness is a common problem in racing horses, and gait asym- stance. Although the overall stride rate does not show much change
metries that arise due to defects in the geometrical design of the with speed, overlap time is reduced and single support time is
racetrack were identified as a predisposing factor. Dalin et al. (1973) increased as a consequence of having longer diagonal advanced
showed that when horses trot on a track with under-banked semi- placements and diagonal advanced completions. These changes in
circular turns the gait becomes asymmetrical at high speeds with timing create a distinctly four-beat gait. The best discriminators
the horses changing to a galloping motion in the hind limbs. To between speeds were stride length and overlap time, while the vari-
overcome the tendency to slip sideways on the turns, the horse leans ables that limited pacing speed were stride rate and suspension time
into the turns and places the feet at an angle, using friction against (Wilson et al., 1988b).
the track surface to overcome the centrifugal force. The left limbs At the end of a race, a pacer is moving at a speed close to 15 m/s.
are adducted and placed on the ground closer to the horse’s midline, When horses with different levels of ability were classified according
which is associated with a lateral landing of the left hooves and to their finishing placing, low-order finishers were found to have
medial deviation of the fetlock throughout the stance phase. Ther- shorter distances between the lateral pair of limbs during stance,
mographic evaluation revealed that, after the horses performed at whereas high-order finishers had a greater range of motion in the
speed around inadequately banked turns, the left fore fetlock fore and hind limbs. Comparing the last lap of a race with the
became warmer than the right one. preceding lap, speed and stride length were reduced during the last
The traditional racetrack design of two straights joined by under- lap for low-order finishers but increased for high-order finishers. At
banked semi-circular turns is obviously not optimal for either per- the end of a race, an increase in the amount of overlap expressed
formance or soundness of the horses. Fredricson et al. (1975) made as a percentage of the stride was a sign of fatigue.
suggestions for improving the geometric design of racetracks. They
recommended the use of wide sweeping turns to reduce the horse’s Sidedness
tendency to slide outwards and to facilitate adequate banking; on Although the pace is generally described as a symmetrical gait, many
tighter turns the amount of banking required is so great that the pacers show temporal asymmetries between the left and right sides.
surface material cannot be stabilized. Only the inner half of the Gait variables that show left–right asymmetries include stance dura-
track needs to be banked sufficiently for maximal speed; the middle tion, suspension, overlap, diagonal dissociation and single support
and outer lanes can be less steeply banked to accommodate moder- (Crawford & Leach, 1984).
ate and slow speed training. It has also been observed that gait
asymmetries are most prevalent as the horse enters and leaves the Conformation
turns. Therefore, the intercalation of a transitional curve between Preliminary kinematic data of 2-year-old pacing fillies suggest that
the true curve and the straight is highly recommended (Fredricson a long sloping scapula, substantial development of the brachioce-
et al., 1975). phalicus, a long forearm and an elastic fetlock joint are associated
with ergonomic efficiency (Sellett et al., 1981).
Conformation of trotters
A study relating conformation to performance in Standardbred
trotters (Magnusson, 1985) showed that traits associated with supe- Endurance racing
rior racing performance included height at the withers and large The primary requirement for an endurance horse is economy of
girth circumference around the point of the withers, which was movement, but the stride variables that contribute to energetic
indicative of prominent withers and a large area for attachment economy have not yet been evaluated. During a race an endurance
of the locomotor muscles. In contrast, girth circumference at the horse takes many thousands of strides. Horses that maximize trans-
lowest point of the back had a negative correlation with perfor- fers of angular momentum between limb segments reduce their
mance. Outward rotation of the limb axes, which is typically mechanical energy expenditure and so use less metabolic energy to
regarded as a conformational fault, had a positive effect on racing cover a given distance. Prolonged endurance training of 8 months’
ability. Conformational features that had a negative influence on duration resulted in significant reductions in stride and stance dura-
performance included tied in at the knees and hocks, narrow hooves tions in association with an increase in the percentage of Type IIA
and greater width/circumference of the forelimbs. Overall it was muscle fibers and a decrease in the percentage of Type IIX muscle
concluded that the better performing horse was a lightweight, fibers, which was interpreted as an indication of more efficient
gracile type with tall withers, with open angles at the shoulder and propulsion associated with an increase in muscle fatigue resistance
stifle joints and normal-sized hooves. and strength (Rivero et al., 2001).
In one study, stride length increased during the course of an
Pacers 82 km endurance ride. Compared with pre-race values, stride length
at the walk increased to 115% at 41 km and 123% after 82 km. The
Stride variables length of the trot stride increased to 151% at 41 km and 149% after
The pace is a symmetrical gait in which the lateral pair of limbs 82 km compared with pre-race values (Lewczuk & Pfeffer, 1998). It
move more or less synchronously. The ability to pace is genetically was suggested that the increase in stride length was due to warming
determined (Andersson et al., 2012). Slow motion studies have up during the course of the race. A different study that evaluated
shown that ground contact of the hind hoof precedes that of the the effects of fatigue found that speed decreased from 4.55 to
front hoof (Crawford & Leach, 1984; Wilson et al., 1988a), unlike 4.03 m/s during an 80 km race due to a reduction in stride length
the racing trot in which the front hoof usually precedes the hind without any change in temporal variables (Wickler et al., 2006).

311
 14 Performance in equestrian sports

This is in contrast to horses exercised to exhaustion at a similar limbs is called diagonal dissociation, with the more specific terms
speed (4.5 m/s) on a treadmill with 6% incline, in which the con- diagonal advanced placement (time between fore and hind contact)
stant speed was maintained using longer stride lengths and shorter and diagonal advanced completion (time between fore and hind
stride frequencies (Wickler et al., 2006). lift-off) also being used. Positive values are assigned if the hind limb
acts in advance of the forelimb, and negative values are assigned if
the forelimb acts in advance of the hind limb. A positive dissocia-
Sport horses tion at contact is associated with good movement and a well-
balanced horse (Holmström et al., 1994b; Morales et al., 1998a,
1998b). The magnitude of the positive dissociation at contact tends
The modern sport horse is elegant in appearance, usually a descen-
to increase with training due to greater elevation of the forehand as
dant of draft breeds crossed with hot-blooded Thoroughbreds and
the degree of collection and self-carriage improve (Tans et al.,
Arabians, that has been refined by many generations of selective
submitted).
breeding to produce the Warmblood breeds.

Quality of movement Forelimb kinematics

Sport horses are selected by horsemen, often at a young age, with The forelimb has been likened to a pendulum rotating around the
one of the important selection criteria being visual appraisal of proximal scapula, with better moving horses showing more retrac-
their gait patterns. The definition of ‘good gait’ or ‘good action’ tion in the forelimbs and more protraction of the hind limbs (Fig.
is, however, rather subjective (Smythe, 1957), and always leads 14.4). To increase stride length, an increased scapular movement is
to discussion between experts (Oliver & Langrish, 1992). Observa- obligatory (Spooner, 1977). Therefore Smythe (1957) and Spooner
tional gait analysis seems to be a convenient technique to quantify (1977) both stressed the importance of having enough space
gait (Krebs et al., 1985), but it should be standardized by using a between the elbow and the ribs to facilitate the sliding movement
uniform scoring system to improve reliability between judges (East- of the proximal forelimb over the rib cage.
lack et al., 1991). Many of the Warmblood registries use such a The spring-like action of the fetlock joint is valued as a method
system to score gait quality (e.g. van Veldhuizen, 1991). to reduce impact shock (Smythe, 1957; Alexander, 1988; Gray,
The most important gait for selection of sport horses is the 1993), to give smoothness of action (Gray, 1993) and to store
trot (Podhajsky, 1981; Clayton, 1994a). Podhajsky (1981) suggested elastic energy (Clayton et al., 1998). The action of the fetlock con-
that the ideal gait and speed to train a horse overground under tributes to the appearance that the horse moves in a graceful, float-
saddle is the trot at 3.75 m/s, which is between the speeds recorded ing manner (Oliver & Langrish, 1992).
for the working and medium trot (Clayton, 1994a; Sloet & Clayton, Much of the expression in the trot is judged from the forelimb
1999). At a speed of 4.0 m/s horses of all ages move on a treadmill movements during the swing phase (Fig. 14.5). At the position
at a comfortable though demanding trot that can be used to evalu- of maximal forelimb protraction, expressive movers have greater
ate stride characteristics (Back et al., 1994a). The same speed has flexion of the elbow and more elevation of the carpus (Fig.
been used to analyze the gait of Standardbreds and Thorough­ 14.5) (Holmström et al., 1994b). This type of movement is indica-
breds (Drevemo et al., 1987; Herring et al., 1992). Standardization tive of the horse controlling the motion of the distal joints
of speed is important because many kinematic variables are speed during protraction rather than relying on inertial forces and ana-
dependent, with the relationship being closer for linear than tem- tomical constraints to determine the joint position in maximal
poral variables (Galisteo et al., 1998). protraction.
Modern kinematic analysis equipment has also been used to With regard to conformation, a sloping shoulder facilitates
assess the quality of movement in elite dressage and show- forward and upward movement of the limb during the swing phase,
jumping horses during competition (Leach et al., 1984; Deuel & so conformation of the shoulder may play a role in gait quality
Park, 1990, 1991; Clayton et al., 1995, 1996; Colborne et al., 1995; (Holmström et al., 1990). After the limb reaches its maximally
Burns & Clayton, 1997; Clayton, 1997a; Hodson et al., 1999) and protracted position, it is retracted before making ground contact
to compare the quality of the overground trot scored by a judge (Fig. 14.6). This period, which is known as the swing phase retrac-
with kinematic variables measured during treadmill locomotion tion, has a longer duration in the forelimbs of horses that receive
(Back et al., 1994b). In the latter study the judge used a linear high trot scores (22% of stride duration) than for those with low
scoring system, in which the trot was rated on three criteria: length, trot scores (10% of stride duration). The fore hoof trajectory reaches
strength and suppleness. The measured kinematic variables included a height of 20–25 cm during midswing in horses with a high gait
forelimb and hind limb stride and swing durations, scapular rota- score, whereas horses with lower gait scores do not raise the hoof
tion, forelimb maximal fetlock extension, forelimb maximal retrac- as high and the flight arc peaks earlier in the swing phase (Holm-
tion, hind limb maximal protraction, maximal stifle flexion and ström et al., 1994b).
maximal tarsal flexion. The kinematic variables correlated very well
with the judged score and allowed the subjective terms used by
horsemen to be quantified. Hind limb kinematics
Better moving sport horses have more protraction of the hind limbs,
shown as more forward placement of the hind hoof relative to the
Stride and swing duration tuber coxae as a result of the combined actions of the joints of the
Sport horses with good gaits move with longer stride lengths and a hind limb (Holmström et al., 1994b). When horses show a combi-
lower stride frequency (Knopfhart, 1966). The lower stride rate and, nation of large ranges of hind limb protraction and simultaneous
hence, longer stride duration in horses with a better trot were associ- forelimb retraction, the trot has the appearance of greater ‘supple-
ated with longer durations of both stance and swing phases in the ness’ in lateral bending of the body, which is one of the visual
forelimb (Back et al., 1994b; Morales et al., 1998a, 1998b). indicators in horses with a good gait (Back et al., 1994b). Stifle and
Although the trot is described as a two-beat gait in which the tarsal joint flexion, which are closely related through the reciprocal
diagonal limb pairs make contact and lift-off simultaneously, slow apparatus (Weeren, 1990), support protraction with greater flexion
motion analysis has shown that there is usually a period of dissocia- being indicative of superior action in the hind limb (Spooner, 1977;
tion with either the fore or hind hoof making ground contact and Alexander & Trestik, 1989; Oliver & Langrish, 1992; Back et al.,
lift-off earlier. The time between contact or lift-off of the diagonal 1994b; Morales et al. 1998a, 1998b).

312
 Sport horses

Fig 14.4  A good quality trot is recognized by a long swing phase duration and a long suspension phase with more retraction of the forelimb and more
protraction of the hind limb.
Courtesy of Ellen van Leeuwen, The Netherlands.

Fig 14.5  Photograph of ‘Mythilus’ ridden by Courtney King-Dye performing the trot with good flexion of the elbow joint raising the carpus and giving
expression to the forelimb movement.
Courtesy of Jacob Melissen, The Netherlands.

During the stance phase, horses with high gait scores show greater
inclination of the pelvis, flexion of the hock and extension of the
Breed differences
fetlock joints as the limb accepts weight. This is associated with a Although European Warmblood breeds currently predominate in
significantly larger angular velocity during tarsal flexion between international dressage competitions, other breeds are competing
20 and 55% stance and during tarsal extension between 75 and 85% more frequently. Warmbloods typically have lofty gaits with a long
stance. In the later part of the stance phase, the limb is retracted suspension phase, which requires the generation of large vertical
faster in good movers (Holmström & Drevemo, 1997). Maximal ground reaction forces and impulses. Compared with Quarter
hoof height occurs soon after lift-off in the hind limb and does not Horses, Warmbloods have been shown to have higher peak vertical
differ with gait quality (Holmström et al., 1994b). forces normalized to body weight (Back et al., 2007). Comparison

313
 14 Performance in equestrian sports

100

80

Carpal joint angle (degrees)

60

Stance phase Swing phase 40

120
110 Retraction Protraction Swing 20
phase
100
retraction
90 0
Degrees

n
io
Fle
80 ens xio
Ext n
70 -20
60 -60 -40 -20 0 20 40
50 A Fetlock joint angle (degrees)
40
0 10 20 30 40 50 60 70 80 90 100
Percent of one stride 100
Fig 14.6  Illustration of the relationship between scapular angle to the
80
horizontal and protraction and retraction of the forelimb during one stride.

Carpal joint angle (degrees)

From Holmström et al. (1994a).
60

of the gaits of Andalusian, Arabian and Angle-Arabian horses 40

showed that Andalusians have more carpal flexion resulting in a
larger range of carpal joint motion and also have less forelimb 20
retraction than the other breeds at walk (Galisteo et al., 2001). Cano
et al. (2001) found that Andalusians had more flexion in the joints 0
of the forelimb at trot than Arabians or Anglo-Arabians, which
accounted for the elevated movements that are characteristic of the -20
breed. Anglo-Arabians had a smaller range of protraction–retraction -60 -40 -20 0 20 40
in both the fore and hind limbs. Tracking length was positive (over- B Fetlock joint angle (degrees)
tracking) in Arabians and Anglo-Arabians, but negative (not track-
ing up) in Andalusians. 100

Ontogeny of gait 80
Carpal joint angle (degrees)

Horse breeders are interested in identifying gait characteristics of 60

young animals that are predictive for future performance, so it is
important to know at what age the adult gait characteristics are 40
acquired. Using modern automated kinematic analysis equipment
Back et al. (1994a, 1995a) compared the gaits of foals with those
20
of adults using angle–angle diagrams that are indicative of intra-
limb coordination (Kobluk et al., 1989; Martinez-del Campo et al.,
0
1991). Plots of carpal angle against fetlock angle (Fig. 14.7) showed
that some horses accelerate their carpal joint flexion during the
-20
swing phase (‘spike’ pattern) and some maintain a constant carpal
-60 -40 -20 0 20 40
angle during the same period (‘straight’ pattern), and others dem-
C Fetlock joint angle (degrees)
onstrate a smoother rate of carpal flexion (‘convex’ pattern).
A longitudinal study of 24 Warmblood horses was performed Fig 14.7  Angle–angle diagrams of the carpal and fetlock joints of three
with evaluations at 4, 10, 18 and 26 months of age at walk, trot and horses (A, B, C) as foals (green line) and adults (red line).
canter on the treadmill (Back et al., 1993, 1994a). A treadmill was From Back et al. (1994a).
used to standardize speed, because Dušek (1974) was not able to
find any correlations between height and temporal gait variables
during overground locomotion when the individuals selected their
own, preferred walking or trotting speed. Intra-limb coordination kinematics were consistent throughout this age range (Back et al.,
features could be recognized as early as 4 months of age. Variability 1994b). Within an individual animal, angle–time and angle–angle
of kinematic variables was similar throughout the growth period graphs for the joints of the fore and hind limbs were remarkably
indicating that kinematic patterns are already established at a similar both visually and numerically from 4 to 26 months (Fig.
young age. Remarkable resemblance between angle–angle diagrams 14.7). Each animal appeared to have an inherent and characteristic
recorded in the foal compared with the adult can be observed espe- intra-limb pattern that was retained from foal to adulthood, the
cially in individuals showing the more typical ‘spike’ versus ‘convex’ ‘gait fingerprint’.
pattern in the carpal–fetlock kinematics (Fig. 14.7). Prediction aims to select horses with a good gait quality at an
There were age-related increases in stride and stance durations early age and Back et al. (1995b) proved that locomotor variables,
(Back et al., 1994a) as a result of increases in limb length and height including swing duration, maximal protraction–retraction angles,
at the withers (Beck et al., 1981; Vilensky et al., 1990a, 1990b). and maximal tarsal flexion, showed a good correlation between
Swing duration, protraction and retraction angles, and joint angular 4-month-old foals and adults. Since these kinematic variables are

314
 Sport horses

indicative of gait quality (Back et al., 1994b) this explains why 0.32
astute horse breeders are said to be able to predict future gait per-
formance at a young age (Clayton, 1989a; Grant, 1989). 0.30 4
The inherent patterns of the angle–angle diagrams that are recog-
3 3 4 4
nized at foal age (Back et al., 1994a) might be related to the future 0.28 44 444 44

Stance duration (s)

3 3 44 44
ability to perform gait transitions or piaffe and passage in dressage 0.26 3 4 4 43
horses, while in jumping horses it might be related to jumping 2 32 43 43243 2333 4
33 3
technique, which has become a more substantial part of selection 22 2 332 3
0.24 1 1 2 21 2 32 3 333
procedures. Gait analysis can be used to quantify performance, 21 22222
while neural networks might be developed to automate pattern 0.22 1 1 22 22
1
recognition in the data (Holzreiter & Köhle, 1993; Dalin, 1994). In 1 111 11 12 2
1 111
1 1
this way, more specific selection criteria for dressage or jumping 0.20 11
could be defined and objectively measured. Finally, it should be
0.18
remembered that horses can be trained to perform well, but for élite
120 130 140 150 160 170 180
performance in competition the mental capacities of rider and
horse might be the crucial factor. A Height at the withers (cm)

Normalization of temporal variables 0.48

4 3
Temporal gait variables in differently sized animals of the same 0.46
3
breed can be made comparable after a transformation based on 4 3
linear and dynamic similarity principles (Günther, 1975; Alexander, 0.44 1 2 3 4 4
11 2 2 3

Swing duration (s)

1984). However, models found in the literature that would 1 1 3
2 3 3
0.42 2 21 233 33 34
compensate for the influence of height on temporal kinematic 1 1
2 22 3 4
1 1 1 2 4
variables, like trot–gallop transitions (elastic similarity) and Froude 2 22 23 3 3 3343 4 44
0.40 1 11 11
numbers (dynamic similarity), showed inconclusive results 1 2 23 43 4 4 4
4 4
(Vilensky & Gankiewicz, 1986; Drevemo et al., 1987). The principle 0.38 1
1 1 2 22 3
of Günther (1975), in which frequencies are related to linear 1 2 2 22 4 34
dimensions, showed a decrease in relative stride duration with age, 0.36 11 4
which is in accordance with Back et al. (1994a) that younger
0.34
animals take relatively longer strides (Fig. 14.8). Back et al. (1995b)
120 130 140 150 160 170 180
hypothesized that, after normalization for differences in conforma-
tion, foal and adult kinematics are similar within a certain range B Height at the withers (cm)
(Tables 14.3–14.5).
Fig 14.8  Correlation between height at withers and stance duration (A)
Differences in height at the withers influence temporal kinemat-
and height at withers and swing duration (B) recorded at trot in a group
ics. Models that compare the kinematics of differently sized animals
of horses at ages 4 months (1), 10 months (2), 18 months (3), and 26
are those of linear or dynamic similarity. months (4).
From Back et al., (1994a).
Linear similarity Due to the increased limb length in adult horses, the distance the
Consider a small and a large horse, in which the dimensions of the body moves forward over the grounded hoof during the stance
large one are a simple multiplication of the dimensions of the phase is longer in adult horses trotting at the same velocity because
smaller one. If the forelimb is regarded as a pendulum that moves the body moves forward over a greater distance during the stance
the body forward at constant horizontal velocity with the same phase (Back et al., 1994a). Compensation for the difference in
pro-/retraction angles and stance durations, the distance moved stance duration is achieved by multiplying stance duration by the
during the stance phase will increase almost linearly with height. ratio of the height at the withers, assuming linear similarity. Stride
During the swing phase the limb must move over a greater distance duration of foals became similar to adults after dynamic correction.
in the same time. If the maximal protraction and retraction angles When using linear similarity principles to correct stride duration, it
of the limbs are the same (Back et al., 1994a) and the speed of was found that adult horses took shorter strides than foals (Back
progression is also the same, the larger animal moves with a longer et al., 1994a, 1995b; Drevemo et al., 1987).
stride length and a slower stride frequency. Normalization can be From the latter study it became clear that as the swing duration
achieved by multiplying the quotient of the height at the withers of at foal age was already similar to that at adult age, but the limbs
the larger and smaller horse with, for example, the stride duration were less flexed in foals than adults because a shorter distance had
of the smaller horse (Günther, 1975). to be covered in the same time. This is nicely illustrated in the
forelimb, in which the carpal joint showed greater flexion in adult
horses. The extension dip in the fetlock joint at the end of the swing
Dynamic similarity phase is more pronounced in adults compared with foals, because
Forces occurring in a pendular locomotor system are considered mass and speed of the hoof are higher. In the hind limb, the pen-
dynamically similar when the ratio of kinetic energy ( 1 2 mv 2 ) and dular response of tarsal flexion to the increase in limb length is
potential energy (mgh, where g = 9.81 m/s2) is the same regardless minimal. This fact, together with the remarkably high correlation
of horse size. The ratio between these energy variables is known coefficient, illustrates the ‘adult-like’ capacities of this joint at
as the Froude number (v2/gh). When the substitution of v = h/t is foal age.
used, the Froude number equals h/gt2. This relationship leads to the
following scaling procedure if dynamic similarity conditions are Coordination
fulfilled: the stride duration of the smaller horse is multiplied by
the square root of the quotient of the height at the withers of the Limb timing and coordination patterns affect many aspects of per-
larger and the smaller horse (Alexander, 1984). formance including locomotor efficiency, quickness and esthetics of

315
 14 Performance in equestrian sports

Table 14.3  Recorded and predicted temporal kinematics of the forelimb and the hind limb of a group of horses trotting on a treadmill at
foal and adult age

Variable Foal Adult

Mean Not corrected predicted Linearly predicted Dynamically predicted Recorded
mean
Range Correlation Value Range Correlation Value Range Correlation
Forelimb
Stance 34.7 33.4–36.0 0.10 42.9 41.0–44.9 0.82† 38.6 37.1–40.1 0.95† 40.3*
duration (%)
Stance 0.21 0.20–0.22 0.43† 0.26 0.25–0.28 0.89† 0.24 0.22–0.25 0.97† 0.27*
duration (s)
Swing 0.40 0.38–0.42 0.65† 0.50 0.47–0.53 0.90† 0.45 0.42–0.47 0.97† 0.40
duration (s)
Stride 0.62 0.59–0.65 0.83† 0.76 0.72–0.80 0.86† 0.68 0.65–0.72 0.96† 0.67*
duration (s)

Hind limb
Stance 34.3 32.1–36.5 0.63† 42.3 39.7–44.9 0.92† 38.1 35.8–40.4 0.98† 40.5*
duration (%)
Stance 0.21 0.20–0.22 0.65† 0.26 0.25–0.27 0.61† 0.23 0.22–0.25 0.66† 0.27*
duration (s)
Swing 0.41 0.38–0.44 0.77† 0.50 0.46–0.54 0.68† 0.45 0.42–0.49 0.73† 0.40
duration (s)
Stride 0.62 0.59–0.65 0.86† 0.76 0.72–0.80 0.71† 0.68 0.65–0.72 0.80† 0.67*
duration (s)

recorded adult variable outside the predicted range of foal age.

*, †p < 0.05: significant difference and correlation between foal and adult.
Predicted range is based on foal kinematics ± SD.
Data from Back et al. (1995b).

movement. The basic patterns of limb coordination are governed Effect of training
by central pattern generators in the spinal cord that regulate flexion
and extension of the joints, which are the basis for stride rate and Prediction of good gait at a young age seems to be possible, but
inter-limb coordination patterns that give rise to the recognizable the question remains as to whether good gait also results in
gaits. Individual horses vary slightly in the synchronization of their good performance when the horse is being ridden. Swedish
limb movements, which can have a marked effect on the esthetics researchers found that, during ridden exercise, good movers had
and energetics of sporting performance. a longer stride duration compared to poor movers (Holmström
When a new motor skill is learned, the initial attempts are et al., 1994b). Furthermore, they calculated that the subjective eval-
clumsy and inefficient. After a period of practice, efficiency in per- uation of conformation and locomotion accounted for 43% of the
forming motor tasks improves. In human athletes, muscular activa- variation of their gaits under saddle (Holmström & Philipsson,
tion patterns are modified by practice; champion athletes have a 1993).
specific temporal and sequential pattern of muscle activation in Training is defined as a program of exercise to improve the horse’s
the execution of their specific skill (Normand et  al., 1982). Sus- physical performance in a particular task (Blood & Studdert, 1990).
tained practice establishes an advanced level of muscular control Most of the literature on training involves the effect on physiologi-
and economy of effort. Codification that develops as a result of cal parameters and on blood chemistry. This section mainly focuses
practice is attributed to reorganization of the motor program on the influence of training on kinematics of riding horses. Clayton
required to execute the specific task within the central nervous (1993) found differences in stride variables of the extended canter
system. These modifications are manifest as inhibition of unde- under a rider between horses trained for dressage and horses trained
sired cocontractions in the antagonistic muscles, contributing to for racing. These included longer stance durations and longer over-
lower overall tension and greater economy of effort. This is an laps between stance phases of different limbs in the dressage horses.
important training consideration, especially in high-intensity The level of training of a dressage horse affects the ability to make
sports, such as polo, cutting and eventing, that combine the need a transition directly between walk and trot (Argue & Clayton,
for speed and endurance with highly coordinated technical skills. 1993a) and between halt and trot (Tans et al., 2009), but not
For more information, the reader is referred to the chapters on between trot and canter (Argue & Clayton, 1993b). Horses compet-
initiation and coordination of gait. ing at a higher level performed transitions between walk and trot

316
 Sport horses

Table 14.4  Recorded and predicted kinematics of the forelimb of a group of horses trotting on a treadmill at foal and adult age

Forelimb variable Foal Adult Foal–adult correlation

Recorded Predicted range Recorded
Pro- and retraction (degrees)
Max protraction 23.0 21.1–24.9 21.5* 0.44†
Max retraction −21.7 −23.0–20.4 −22.8* 0.03
Range of max pro- and retraction 44.7 42.8–46.6 44.4 0.50†

Scapula (degrees)
Range of max rotation 15.5 13.8–17.2 17.8* 0.47†

Elbow (degrees)
At IGC 53.6 50.1–57.1 50.1* 0.12
Max extension relative to IGC −25.7 −28.7–22.7 −25.6 0.36
Max flexion relative to IGC 30.9 26.6–35.2 34.6* 0.48†
Range of movement 56.6 52.1–61.1 60.2* 0.44†

Carpus (degrees)
At IGC 4.4 1.8–7.0 6.0 0.03
Max extension relative to IGC −7.6 −9.9–5.3 −9.0 0.36
Max flexion relative to IGC 70.0 64.9–75.1 81.9* 0.42†
Range of movement 77.5 73.2–81.8 90.8* 0.47†

Fetlock (degrees)
At IGC −15.3 −18.4–12.2 −16.8 0.01
Max extension relative to IGC −36.9 −40.4–33.4 −39.1* 0.37
Max flexion relative to IGC 42.2 37.8–46.6 41.5 0.47†
Range of movement 79.1 73.6–84.6 80.6 0.36

recorded adult variable outside the predicted range at foal age.

*, †p < 0.05: significant difference and correlation between foal and adult.
Predicted range is based on foal kinematics ± SD. IGC, initial ground contact.
Data from Back et al. (1995b).

more ‘cleanly’, that is with fewer support sequences that were not duration increased after 1 year of training. Similarly, in Andalusian
part of the normal pattern for either gait. horses, 10 months of training resulted in an increase in swing dura-
Many Warmblood registries have a stallion performance test as tion and a decrease in hind stance percentage (Cano et al., 2000).
part of the process for determines which stallions are accepted for Corley and Goodship (1994) also reported a decrease in stance
breeding. In Dutch Warmbloods a training period of 70 days has duration in cantering Thoroughbreds trained on a treadmill, as
been shown to correlate well with the future performance (Huiz- did Drevemo et al. (1987) in four young Standardbreds that
inga, 1991). Back et al. (1995a) compared the effects of a 70-day were trained over a 3-year period. Training changes the relation­
training period with turn out on pasture by evaluating the temporal, ship between stance and swing durations (Cano et al., 2000) and
angular and segmental kinematics of the fore and hind limbs. between stride duration and stride frequency (Muñoz et al., 1997),
In the hind limbs of the trained horses, the same stride duration but the definition and effects of ‘training’ may differ between
was achieved with a significantly shorter stance duration after the studies.
70-day period, illustrating the development of ‘impulsion’ through In trained horses the load seems to shift from the forelimbs
improvement of muscular strength (Fig. 14.9). Horses pastured for towards the hind limbs, in which maximal fetlock extension
70 days also decreased their stance percentage, but this was associ- increases, and the horse is said to ‘carry itself ’ (Crossley, 1993). Back
ated with increases in both swing and stride duration (Back et al., et al. (1995a) hypothesized that increased impulsion in trained
1995a). horses might be visible as increased extension of the tarsal and hind
Schwarz (1971) analyzed some temporal kinematic variables fetlock joints, in contrast to the increased extension of the carpal
of Hannoverian stallions walking and trotting overground and and fore fetlock joints seen in pastured horses. Interestingly, young
found that swing duration expressed as a percentage of total stride Standardbreds that were trained for 5 months showed both

317
 14 Performance in equestrian sports

Table 14.5  Recorded and predicted kinematics of the hind limb of a group of horses trotting on a treadmill at foal and adult age

Hind limb variable Foal Adult Foal–adult correlation

Recorded Predicted range Recorded
Pro- and retraction (degrees)
Max protraction 18.1 17.1–19.1 21.6* 0.28
Max retraction −29.7 −30.9–28.5 −26.6* 0.48†
Range of max pro- and retraction 47.8 46.5–49.1 48.1 0.58†

Pelvis (degrees)
Range of max rotation 7.6 5.9–9.3 9.1* 0.77†

Hip (degrees)
At IGC 91.7 87.5–95.9 88.4* 0.07
Max extension relative to IGC −20.9 −23.4–18.4 −20.6 0.70†
Max flexion relative to IGC 2.0 0.9–3.1 2.7* 0.55†
Range of movement 22.9 20.9–24.9 23.3 0.59†

Stifle (degrees)
At IGC 15.9 11.7–20.1 12.0* 0.08
Max extension relative to IGC −0.8 −1.6–0.0 −1.0 0.19
Max flexion relative to IGC 48.4 43.8–53.0 46.3* 0.48†
Range of movement 49.2 44.4–54.0 47.3* 0.55†

Tarsus (degrees)
At IGC 21.4 18.0–24.8 16.1* 0.42†
Stance flexion relative to IGC 9.6 6.8–12.4 10.6* 0.52†
Max extension relative to IGC −6.1 −8.2–4.0 −5.8 0.17
Max flexion relative to IGC 47.3 42.3–52.3 49.7* 0.65†
Range of movement 53.4 48.3–58.5 55.4* 0.70†

Fetlock
At IGC −8.8 −12.9–4.7 −13.5* 0.27
Max extension relative to IGC −37.8 −42.0–33.6 −39.5* 0.53†
Max flexion relative to IGC 44.8 38.9–50.7 45.5* 0.61†
Range of movement 82.6 75.6–89.6 85.0 0.56†

recorded adult variable outside the predicted range at foal age.

*, †p < 0.05: significant difference and correlation between foal and adult.
Predicted range is based on foal kinematics ± SD. IGC, initial ground contact.
Data from Back et al. (1995b).

phenomena: increased carpal and fetlock extension in the forelimb similar to those of pastured horses (Table 14.6) (Back et al., 1995a).
and increased tarsal and fetlock extension in the hind limbs (Weeren Overall, horses kept at pasture show increases in swing and stride
et al., 1993). durations, they have a larger range of protraction and retraction and
At the beginning of a horse’s career under saddle the movement increases in maximal carpal and fetlock extension of the forelimb.
often appears to become ‘shorter in front.’ For example, Andalusian They appear to be more on the forehand with a more relaxed type
horses had a shortened stride length after 10 months in training of movement.
(Cano et al., 2000). After a further period of training, young horses A more prolonged period of dressage training teaches horses
find their balance, and are able to carry a rider ‘on the bit.’ These to move with greater collection and self-carriage. The hind limbs
changes are accompanied by re-establishment of movement of the provide more of the forward propulsive force as self-carriage devel-
proximal limb segments, which is reflected by gradual increases in ops. At the same time, the forelimbs show a reduced propulsive
cranio-caudal movement of the distal segments of the forelimb force and develop an increased braking force to control balance and

318
 Sport horses

Fig 14.9  ‘Ravel’ ridden by Steffen Peters moves with impulsion and shows a large amount of forelimb retraction and hind limb protraction.
Courtesy of Jacob Melissen, The Netherlands.

Table 14.6  Maximal cranial and caudal movement in centimeters of the forelimb segments relative to the proximal scapula before and
after 70 days in training or 70 days at pasture

Forelimb segment Training (n = 12) Pasture (n = 12) ANOVA (post-hoc)

Before After Before After

Scapula
Distal 5.1 5.6* 4.7 5.3*

Humerus
Proximal 11.4 12.3* 11.1 12.2*
Distal 22.7 23.6* 22.4 24.0*

Radius
Proximal 27.9 29.2* 27.2 29.1*
†
Distal 62.1 62.4 61.2 62.7*

Metacarpus
† #
Proximal 75.6 75.1 74.3 76.3* ()
† #
Distal 93.9 92.9 92.2 93.9* ()

Hoof
† #
Coronet 112.2 111.2 110.3 112.6* ()
† #
Heel 113.5 112.7 11.3 114.2* ()
† #
Toe 116.2 115.8 113.9 116.5* ()

*Indicates that the values before and after 70 days are significantly different (p < 0.05) within the training or pasture groups (Student’s t-test).
†
Indicates that the changes in the values over the period of 70 days are significantly different (p < 0.05) between the training and pasture groups (ANOVA).
#
Indicates the changes in the values over the period of 70 days are also significantly different between the training and pasture groups after a Bonferroni post-hoc test
(p ≤ 0.05).
There was a statistically significant interaction between ‘time’ (before/after) and ‘group’ (training/pasture), and between ‘segment’ (marker no. 1–10), ‘time’ and ‘group’ (p < 0.05;
ANOVA).
Data from Back et al. (1995a).

319
 14 Performance in equestrian sports

Table 14.7  Time relative to initial ground contact (t = 0) of the maximal protraction of the different segments of the hind limb before and
after 70 days in training or 70 days at pasture. Time is expressed as a percentage of total stride duration

Hind limb segment Training (n = 12) Pasture (n = 12) ANOVA (post-hoc)

Before After Before After

Femur
Distal −1.9 −3.2* −2.7 −2.9

Tibia
Proximal −2.4 −3.8* −3.0 −3.5
Distal −1.4 −2.4* −1.6 −1.8 †

Metatarsus
Proximal −1.6 −2.5* −1.7 −1.9 †

Distal −2.1 −2.8* −2.1 −2.0 †

Hoof
Coronet −2.0 −2.7* −1.8 −1.9
Heel −1.9 −2.6* −1.8 −1.8
Toe −1.8 −2.5* −1.7 −1.8

*Indicates that the values before and after 70 days are significantly different (p < 0.05) within the training or pasture groups (Student’s t-test).
†
Indicates that the changes in the values over the period of 70 days are significantly different (p < 0.05) between the training and pasture groups (ANOVA).
There was a statistically significant interaction between ‘time’ (before/after) and ‘group’ (training/pasture), and between ‘segment’ (marker no. 1–10), ‘time’ and ‘group’ (p < 0.05;
ANOVA).
Data from Back et al. (1995d).

prevent the horse from falling onto the forehand as a result of the
Table 14.8  Response to the 70 days in training or 70 days at
increased hind limb propulsion. Consequently, the trained horse
pasture assessed according to the change in maximal fetlock
does not roll over the forelimbs; instead the forehand is vaulted
extension of the forelimb and the hind limb
upwards over the forelimb strut. ‘Engagement of the hind quarters’,
which is one of the primary goals of training the young sport horse
(Crossley, 1993), might be visible as an earlier maximal protraction Forelimb
of the hind limb with respect to the retracting ipsilateral forelimb
Training (n = 12) Pasture (n = 12)
as a consequence of the ability to generate the impulse needed from
the hind limbs during a shorter stance time (Fig. 14.9; Table 14.7). Decrease Increase Decrease Increase
At the same time, increased fetlock extension illustrates more weight
carrying by the hind limbs (Back et al., 1995a) (Table 14.8). After Hind limb
an initial 10 months of dressage training Andalusian horses had Decrease 3 2 1 6
more flexion of the hip and stifle joints at the beginning of stance
and at the beginning of swing, which may indicate the development Increase 5 2 2 3
of engagement (Cano et al., 2000).
Total 8 4 3 9†

recorded adult variable outside the predicted range at foal age.

Effect of a rider †
Indicates that the ratio of the number of horses that showed an increased and
The presence of a rider affects both the kinematics and the ground decreased value after 70 days is significantly different (p < 0.05) between the training
reaction forces, with the changes being more pronounced in the and pasture groups (Chi square test).
forelimbs (Schamhardt et al., 1991; Sloet and Barneveld, 1995; Data from Back et al. (1995d).
Clayton et al., 1999). When the rider performs a posting trot there
are left–right asymmetries in the ground reaction forces, and expe-
rienced riders shift part of the weight towards the hind limbs stance. During walking, the effect of an experienced rider was almost
(Schamhardt et al., 1991). Comparing horses trotting at the same identical to that of a sandbag of equal mass.
speed in hand and with a rider, the peak vertical ground reaction
force normalized to the mass of the system (horse or horse plus
rider) is lower in both the forelimbs and the hind limbs when Dressage
horses are ridden. In the forelimbs the mass normalized propulsive
force is higher in the second half of stance in ridden horses (Fig. Dressage is a judged sport in which the quality of the horse’s
14.10), and the fetlock joint is more extended between 50 and 70% gaits (kinematics) and the horse’s innate athleticism (balance,

320
 Dressage

Vertical GRF Vertical GRF

Hindlimb Forelimb
8000 8000
7000 7000
6000 6000
5000 5000
Force (N)

Force (N)
4000 4000
3000 3000
2000 2000
1000 1000
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Stance (%) Stance (%)

Longitudinal GRF Longitudinal GRF

Hindlimb Forelimb
1000 800
600
500
400
0 200
20 40 60 80 100
Force (N)

Force (N)
0
–500 20 40 60 80 100
-200
–1000 -400
-600
–1500
-800
–2000 -1000
Stance (%)
Fig 14.10  Vertical (above) and longitudinal (below) ground reaction forces (GRFs) of the hind limbs (left) and the forelimbs (right) for horses trotting at the
same speed in hand (blue line) and with a rider (red line).

coordination) are determinants of competitive success. Since dres- cycle of hind limb movement is well forward under the horse’s
sage competitions are judged subjectively, the esthetic quality of the body, and the vertebral column is rounded while the forehand is
gaits and the expressiveness of the movements are important factors relatively elevated. Horses that are able to carry themselves in this
in determining competition scores. Estimated heritabilities for dres- manner are said to move in self-carriage. As the horse becomes more
sage in the Dutch Warmblood breed were 0.17 ± 0.05 and genetic collected, the range of pendular motion of the hind limb is reduced
correlations between conformation and performance were low to (Holmström & Drevemo, 1997).
moderate (Koenen et al., 1995). The length of the neck, length and
position of the shoulders, shape and length of the croup and mus-
cularity of the haunches had a significant moderate genetic correla- Rider effects
tion with dressage performance. Due to the low genetic correlations
with performance traits, it was concluded that indirect selection for The dressage rider strives follow the movements of the horse har-
performance based on conformation is of limited value. In Swedish moniously during all the gaits. This is a learned skill that improves
Warmbloods, similar heritabilities for gait traits were found (range with practice. For example, novice riders stabilize their position by
0.09–0.27) and, when combined with estimated heritabilities for gripping with the adductor muscles but these muscles become more
competitive dressage performance, they were interpreted as being relaxed as coordination between rectus abdominis and erector spinae
useful for early genetic evaluation and selection of both mares and muscles improves (Terada, 2000). Horses ridden by professional
stallions for sport performance traits (Wallin et al., 2003). Compari- riders perform more harmoniously and with a more consistent
son of gait variables of walk and trot with skeletal conformation motion pattern than horses ridden by recreational riders (Peham
measurements in 142 3-year-old horses of three breeds (German, et al., 2003). Experienced riders maintain an almost vertical trunk
French and Spanish saddle horses) were interpreted as indicating position in sitting trot (Schils et al., 1993) and are able to maintain
that the gaits of the German horses were more adapted for dressage dynamic equilibrium by anticipating the translational movements
competition, whereas purebred Spanish horses could be considered of the horse, leaning forward slightly by about 5° in late stance in
as a reference for collected gaits as used in academic dressage preparation for the push-off into the suspension, then leaning back-
(Barrey et al., 2003). ward by about 5° in preparation for the deceleration of impact
The qualities of rhythm and relaxation are important at all levels (Terada et al., 2006). Movements of the rider’s shoulder and elbow
of dressage training and as the horse progresses to the more joints are synchronized with these forward-backward trunk oscilla-
advanced levels, the amount of collection and self carriage must tions, which allow the rider’s hand to maintain a consistent posi-
be developed to a higher degree. Collection describes a manner of tion relative to the bit (Terada et al., 2006). The upper and middle
moving in which the strides become shorter and more elevated, the trapezius and middle deltoid muscles are active in early stance to

321
 14 Performance in equestrian sports

Regular Table 14.9  Mean values for stride kinematics for the collected,
rhythm: RH RF LH LF RH
medium and extended walks in dressage horses

Lateral Collected Medium Extended

couplets: RH RF LH LF RH
walk walk walk
Diagonal Speed (m/s) 1.4a 1.7b 1.8b
couplets: RH RF LH LF RH
a b
Stride length (m) 1.57 1.87 1.93b
Left walk
pirouette: RH RF LH LF RH Stride frequency 52a 55a,b 56b
(strides/min)
Time
Lateral distance (cm) 158a 167a 166a
Fig 14.11  Temporal characteristics of a regular walk, a walk with lateral Tracking distance (cm) −7a 19b 27b
couplets, a walk with diagonal couplets and a half pirouette in walk.
Different superscripts indicate values that differ significantly (p < 0.05).
Data are from Clayton (1995).

stabilize the rider’s neck and shoulder through the impact phase
(Terada et al., 2004). At the same time, biceps brachii stabilizes the
elbow and prevents the forearm descending under the effect of These values are similar to the speed of 1.82 m/s achieved with a
gravity, while the carpal flexors and extensors stabilize the wrist. stride length of 1.93 m and a stride duration of 1.06 s reported by
It is not only the effect of the horse on the rider that must be Clayton (1995) in horses of slightly lower caliber. On average,
taken into consideration; the position and movements of the rider bipedal contact accounted for 61% of stride duration and tripedal
have a dynamic effect on the horse’s ground reaction forces (Clayton contact for 39% of stride duration in the Olympic competitors
et al., 1999). For example, in the sitting trot the rider’s descent (Deuel & Park, 1990).
during the stance phase increases the vertical ground reaction force Stride length in the walk can be considered as the sum of the
on the horse’s forelimbs. lateral distance (distance between the hind hoof and the next place-
ment of the ipsilateral fore hoof) plus the tracking distance (dis-
The gaits tance between the fore hoof and the next placement of the ipsilateral
hind hoof). Changes in stride length at the walk are almost entirely
The gaits performed in competition are the walk, trot, canter, rein due to adjustments in tracking distance (Table 14.9). Lengthening
back, passage and piaffe. The walk, trot and canter each have several of the stride is accompanied by a wider arc of limb rotation during
variations that differ in speed of progression. From slowest to fastest the stance phase. The angle between the metapodial bone and the
these are the collected, working, medium and extended gaits. Horses ground is more acute at ground contact and more obtuse at lift-off
are supposed to maintain the same stride frequency (tempo) during in the extended walk than in the collected walk, without any change
the transitions between these gait types. In other words, they should in the carpal or tarsal angle at impact or at lift-off (Clayton, 1995).
change stride length independently of stride frequency. Although dressage horses are required to maintain a regular, four-
beat rhythm in the walk, only a minority of horses achieve this
(Clayton, 1995). When the rhythm becomes irregular, the horses
The walk show either lateral couplets (lateral or pacing rhythm), in which
The walk is a four-beat gait with a lateral footfall sequence in which there is a shorter time between the lateral footfalls, or they show
contact of a hind limb is followed by contact of the lateral forelimb: diagonal couplets (diagonal rhythm), in which there is a shorter
RH, RF, LH, LF. The footfalls should be evenly spaced in time, giving time between the diagonal footfalls (Fig. 14.11). Clayton (1995)
a regular, four-beat rhythm. The limb support sequences alternate found that a majority of national level dressage horses showed
between bipedal and tripedal supports (Fig. 14.11). The bipedal lateral couplets, with the same footfall pattern being present in all
supports always consist of a forelimb and a hind limb, which may types of walk. The average step durations measured for dressage
be a diagonal or a lateral pair. The tripedal supports may be two finalists competing in the Seoul Olympics indicate that they moved
hind limbs and one forelimb or two forelimbs and one hind limb. with diagonal couplets (Deuel & Park, 1990).
The Fédération Equestre Internationale (FEI) recognizes four In an investigation of the desirable characteristics of the walk,
types of walk: collected, medium, extended and free. (There is no Biau and Barrey (2004) attached an accelerometer beneath the
working walk.) The free walk, in which the horse is allowed freedom sternum of young horses (4–6 years old) and experienced horses
to stretch the neck, is only performed in lower levels of competition. (7–13 years old). Comparison between the accelerometric data and
The speed, tempo and stride length of the collected, medium and the scores awarded by dressage judges showed that the horses were
extended walks measured in a group of national level FEI horses are rewarded for having a slow, regular and symmetrical pattern of
compared in Table 14.9 (Clayton, 1995). Speed of the medium and movement with a large amount of dorsoventral displacement.
extended walks was significantly faster than that of the collected In highly trained dressage horses, both the collected and extended
walk. Compared with the collected walk, stride length was 23% walk strides have a longer stance duration in the hind limbs than
longer and stride frequency was 8% longer in the extended walk, in the forelimbs (Clayton, 1995; Hodson et al., 1999). This may be
which agrees with Dušek et al. (1970) who found that increases in related to the requirement that highly trained horses move in self-
speed at all gaits up to a moderate gallop were caused mainly by carriage, which implies lightness of the forehand and a greater
increasing stride length. However, dressage horses did not fulfill the reliance on the hind limbs for propulsion.
FEI requirement that the same stride frequency be maintained at all In Andalusian horses, medium to high heritabilities were found
types of walk. for many kinematic parameters of the walk, including stride length
For dressage horses competing in the Seoul Olympics, the mean and duration, maximal height of the hind hoof, range of forelimb
speed of the extended walk was 1.88 m/s, the stride length was protraction–retraction, hind limb stance duration, and fore and
1.95 m and the stride duration was 1.03 s (Deuel & Park, 1990). hind limb swing durations (Molina et al., 2008).

322
 Dressage

Half pirouette at the walk frequency is significantly slower in collected than extended trot
(Clayton, 1994a), which does not meet the requirements to main-
The half pirouette in collected walk is a half circle in which the tain a constant stride frequency. Deuel and Park (1990) have shown
forehand moves around the hindquarters. The inside hind limb acts a positive relationship between speed and stride length and a nega-
as a pivot point for the movement, but it continues to step in the tive relationship between speed and stride duration in a group of
rhythm of the walk strides. In a study at the Atlanta Olympics top-level competitors. Interestingly, dressage horses that qualified
(Hodson et al., 1999) the majority (8/11) of the horses completed for the individual medal finals in the Seoul Olympics tended to
the half pirouette at walk in three strides; the remaining horses used have faster speeds, longer stride lengths and higher stride frequen-
four strides. This is consistent with the FEI rules, which stipulate cies in the extended trot than horses that failed to qualify (Deuel &
three to four strides in a half pirouette. None of the horses main- Park, 1990).
tained a regular four-beat rhythm in the walk pirouette. Instead, the Stride length at the trot depends on the diagonal distance (dis-
footfall of the inside hind limb occurred relatively early in the tance between the diagonal pair of limbs during their stance phase)
stride. Consequently, the time between footfalls of the outside fore and the tracking distance (distance between the fore hoof and the
and inside hind hooves was short, while the time between footfalls next contact of the ipsilateral hind hoof). Diagonal distance shows
of the inside hind and inside fore hooves was long (Fig. 14.11). This a non-significant increase of 4.0–5.0 cm between working and
indicates that, to compensate for the lack of forward movement, the medium trot, which is probably a consequence of the lengthening
horses become more reliant on the inside hind limb to maintain of the horse’s frame. However, most of the lengthening of the stride
their balance. is a result of greater over-tracking (46 cm increase from collected to
extended trot), which represents the distance covered during the
suspension (Clayton, 1994a). The best way to increase over-tracking
The trot and, therefore, stride length is to prolong the suspension by
The trot is a two-beat gait in which the diagonal pairs of limbs move pushing-off with a higher vertical velocity. Suspension in the
more or less synchronously, and the footfalls of the diagonal limb medium and extended trots is twice as long as that of the collected
pairs are evenly spaced in time. The diagonal support phases are and working trots (Clayton, 1994a).
usually separated by periods of suspension, except in a very slow In the highest scoring dressage horses in the Seoul Olympics
(jog) trot. Therefore, each stride has two diagonal support phases the speed of the extended trot was strongly influenced by stride
and two suspensions. In young horses, an accelerometric study length but not closely related to stride duration (Deuel & Park,
showed that judges awarded higher marks to horses showing a slow 1990). This indicates greater reliance on changes in stride length
trot with a large amount of dorsoventral displacement of the trunk rather than stride frequency in elite dressage horses. The speed of
(Biau & Barrey, 2004). the extended trot (4.93 m/s) in the national level competitors
Slow motion analysis has shown a slight dissociation between studied by Clayton (1994a) is similar to the speed recorded in
ground contact and lift-off of the diagonal fore and hind limbs. Olympic competitors (4.98 m/s) by Deuel and Park (1990).
The interval between the fore and hind contacts is known as diago- However the Olympic competitors achieved this speed using a
nal dissociation. The value is positive if the hind limb contacts longer stride length (3.79 m versus 3.55 m) and a longer stride
the ground before the forelimb, zero if the diagonal pair contact the duration (0.763 s versus 0.722 s). Therefore, horses of a slightly
ground simultaneously, and negative if the hind limb contacts lower caliber achieve the speed required in the extended trot using
the ground after the forelimb. Positive diagonal dissociation at shorter, faster strides.
contact (Fig. 14.9) is considered a desirable characteristic that is The stance durations of the forelimbs and hind limbs do not
indicative of good balance (Holmström et al., 1995). It occurs in differ from each other in any type of trot (Clayton, 1994a), but both
horses that travel with an elevated forehand, which is a character- the fore and hind stance durations are significantly shorter in
istic of collection. However, a negative diagonal dissociation does extended trot than in collected trot. The angles of the cannon
not preclude a horse from being successful in dressage. In the Seoul segment to the horizontal are significantly more acute at hoof
Olympics, 15% of the extended trot strides that were analyzed had contact and more obtuse at lift-off in the extended trot than in the
a negative diagonal dissociation at contact (Deuel & Park, 1990). collected trot (Clayton, 1994a). The hind cannon consistently has
Four types of trot are performed in competition: collected, a more acute angle to the ground on its plantar side than the fore
working, medium and extended. Table 14.10 shows that speed and cannon throughout the stance phase. The difference ranges from
stride length differ significantly between each type of trot, and stride about 10° at hoof contact to 20° at lift-off.

Table 14.10  Mean values for stride kinematics of the collected, working, medium and extended trots in FEI level dressage horses

Collected trot Working trot Medium trot Extended trot

a b c
Speed (m/s) 3.20 3.61 4.47 4.93d
Stride length (m) 2.50a 2.73b 3.26c 3.55d
Stride frequency (strides/min) 77a 79a,b 82a,b 83b
Diagonal distance (cm) 132a 132a,b 136a,b 137b
Tracking distance (cm) −7a 4b 27c 39d
Suspension (ms) 16a 17a 32b 37b

Different superscripts indicate values that differ significantly (p < 0.05).

Data from Clayton (1994a).

323
 14 Performance in equestrian sports

Table 14.11  Mean values for stride kinematics of the collected

trot, passage and piaffe in dressage horses competing in the
individual medal finals at the Barcelona Olympics

Collected trot Passage Piaffe

a b
Speed (m/s) 3.3 1.6 0.2c
Stride length (m) 2.50a 1.75b 0.20c
Stride frequency 71a 55b 55b
(strides/min)

Different superscripts indicate values that differ significantly (p < 0.05).

Data from Clayton (1997a).

The longitudinal GRF is almost entirely retarding in the forelimbs,

and almost entirely propulsive in the hind limbs. The forelimbs
have the effect of elevating the forehand, while the hind limbs
provide forward and upward propulsion (Fig. 14.13). On a tread-
mill, passage shows higher vertical impulses than collected trot in
the fore and hind limbs without an increase in peak vertical forces.
The load shifts from the fore to the hind limb of the diagonal pairs
(Weishaupt et al., 2009).

The piaffe
Piaffe (or piaffer) is a highly collected, cadenced, elevated diagonal
movement giving the impression of remaining in place. The horse’s
back is supple and elastic. The hindquarters are lowered; the
haunches with active hocks are well engaged, giving great freedom,
lightness and mobility to the shoulders and forehand. Each diago-
Fig 14.12  ‘Salinero’ ridden by Anky van Grunsven performing passage and nal pair of legs is raised and returned to the ground alternately, with
showing elevation of the hooves in midswing. spring and an even cadence (Anon, 2009). As in passage, the limbs
Courtesy of Jacob Melissen, The Netherlands. pause momentarily at their position of maximal elevation.
Stride duration is longer in piaffe than in collected trot but is
similar in piaffe and passage (Clayton, 1997a; Holmström et al.,
The passage 1995). The rules require piaffe to be performed in place in the
According the FEI Rules for Dressage, passage is a measured, very Grand Prix test, whereas in the Intermediate II test the horse is
collected, elevated and cadenced trot. It is characterized by a pro- allowed to move forward with a stride length of 20 cm (Table
nounced engagement of the hindquarters, a more accentuated 14.11), which is in accordance with the amount of forward progres-
flexion of the knees and hocks, and the graceful elasticity of the sion recorded during Olympic competition (Argue, 1994). Since
movement (Anon, 2009). The limbs pause momentarily at their piaffe has little, if any, forward momentum, horses maintain their
position of maximal elevation (Fig. 14.12), when the toe of the balance by increasing the stance durations and overlaps between
raised forefoot should be level with the middle of the cannon bone limbs. Piaffe has longer fore, hind and diagonal stance durations
of the other supporting forelimb. The toe of the raised hind foot than passage or trot and these are associated with longer overlaps
should be slightly above the fetlock joint of the other supporting between limbs (Clayton, 1997a). There is always at least one hoof
hind limb (Anon, 2009). None of the horses competing in the in contact with the ground, so piaffe has no suspension phase
individual medal finals at the Barcelona Olympics achieved (Holmström et al., 1995; Clayton, 1997a). Therefore, piaffe is a
the required elevation in the forelimbs in passage (Argue, 1994). stepping gait, rather than a leaping gait, with a gradual transfer of
The diagonal limb pairs move more or less in synchrony in passage body weight from one diagonal to the other. The amount of overlap
but with a pronounced positive diagonal dissociation (Clayton, between successive diagonal stance phases is, however, shorter in
1997a; Holmström et al., 1995; Weishaupt et al., 2009), which is the better-quality piaffe (Clayton, 1997a).
longer in the more successful competitors (Clayton, 1997a). Each horse performs the piaffe with its own, highly individual-
Stride length and speed are significantly reduced from collected ized coordination pattern and, although the mean value of the
trot to passage and from passage to piaffe (Table 14.11). Passage diagonal dissociation at contact for piaffe in a group of horses was
and piaffe have the same stride frequency, which is significantly negative (Argue, 1994; Holmström et al., 1995), the best competi-
slower than that of collected trot (Clayton, 1997a). tive horses had a positive diagonal dissociation (Clayton, 1997a).
Kinematic analysis indicated that the stifle and tarsal joints were Piaffe is a highly collected gait and the collection is associated with
more flexed at ground contact and the tarsus was also more flexed more pelvic tilting throughout the stride cycle, greater stifle and
in midstance in passage compared with collected trot (Holmström tarsal flexion at the start of stance and greater tarsal flexion at mid-
et al., 1995). For horses moving on a treadmill, there was more stance compared with passage and collected trot. Horses do not,
lumbosacral flexion throughout the stride in passage than in col- however, step further underneath themselves in piaffe than in
lected trot (Weishaupt et al., 2009). passage or collected trot (Holmström et al., 1995).
The ground reaction forces in passage resemble those of the col- A unique feature of piaffe is the GRF profiles (Clayton, unpub-
lected trot in horses moving overground (Clayton, unpublished). lished data). The vertical GRF has a smaller amplitude than passage
The forelimbs have a higher peak vertical force than the hind limbs. or collected trot and the trace has a flattened profile during the long

324
 Dressage

Vertical GRF Vertical GRF Vertical GRF

Collected trot Passage Piaffe
6000 6000 6000
5000 5000 5000
4000 4000 4000
Force (N)

Force (N)

Force (N)
3000 3000 3000
2000 2000 2000
1000 1000 1000
0 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Stance (%) Stance (%) Stance (%)

Longitudinal GRF Longitudinal GRF Longitudinal GRF

Collected trot Passage Piaffe
400 400 400
200 200 200
0 0 0
Force (N)

Force (N)

Force (N)
–200 –200 –200
–400 –400 –400
–600 –600 –600
–800 –800 –800
–1000 –1000 –1000
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
Stance (%) Stance (%) Stance (%)
Fig 14.13  Vertical (above) and longitudinal (below) ground reaction forces for collected trot (left), passage (center) and piaffe (right) for the forelimbs (blue
line) and hind limbs (red line).

Table 14.12  Stride kinematics of the collected, working, medium and extended canters in FEI level dressage horses

Collected canter Working canter Medium canter Extended canter

a b c
Speed (m/s) 3.27 3.91 4.90 5.97d
Stride length (m) 2.00a,b,c 2.35a,d,e 2.94b,d,f 3.47c,e,f
Stride frequency (strides/min) 99a 99a 101a 105a
Suspension (ms) 0a,b 5c,d 54a,c 87b,d

Different superscripts indicate values that differ significantly (p < 0.05).

Data from Clayton (1994b).

stance phase (Fig. 14.13). The longitudinal force is almost entirely to stay airborne longer and to cover a greater distance during the
propulsive in the forelimbs and almost entirely braking in the hind suspension.
limbs (Fig. 14.13), which is the opposite of passage and collected For horses competing in the individual medal final at the Seoul
trot. Olympic Games, the speed of the extended canter (7.03 m/s) was
considerably faster than that recorded in slightly lower-caliber
national level competitors (5.97 m/s). The difference was primarily
The canter a result of a longer stride length (4.15 m versus 3.47 m), which was
The canter is the only asymmetrical gait of dressage horses. It has a combined with a slightly slower stride frequency (101 strides/min
transverse sequence of limb placements, so the leading fore and in the national level horses versus 105 strides/min in the Olympic
hind limbs are on the same side of the body. Four types of canter competitors). Higher overall competition scores have been recorded
are performed in dressage competitions: collected, working medium, for horses with faster speeds and longer stride lengths in the
and extended, which have significantly different speeds (Clayton, extended canter. No upper limit was detected for optimal stride
1994b). The stride frequency is the same for the different types of length (Deuel & Park, 1990). Higher scoring horses also showed
canter, so changes in speed are accomplished by alterations in stride shorter periods of ground contact of the limbs, while increasing
length (Table 14.12). Stride length increases as a result of a small both the duration and distance covered during the suspension.
increase in the distance between the two hind limbs, a small increase The rhythm of the stride differs between the collected and
in the distance between the two forelimbs, and a large increase in extended canters (Clayton, 1994b). In collected canter, the three
the distance covered during the suspension. The ability to generate footfalls are separated by relatively long intervals but the suspension
a high vertical velocity at the start of the suspension allows the horse is relatively short. In extended canter the three footfalls are further

325
 14 Performance in equestrian sports

Leading Fore Trailing Fore

Table 14.13  Stride variables for the one tempi and two tempi
Leading Hind Trailing Hind
lead changes
14

12 One tempi Two tempi Two tempi

stride strides before strides after
10 change change
Force (N/kg)

8 Speed (m/s) 3.36 3.65 3.95

6 Stride length (m) 2.08 2.21 2.44
4 Hind step 0.94 0.87 0.88
TrH–LdH (m)
2
Diagonal step 0.97 0.97 0.93
0 LdH–TrF (m)
0 20 40 60 80 100
Fore step 0.80 0.79 0.89
2.0 TrF–LdF (m)

1.5
Suspension −0.53 −0.47 −0.33
LdF–TrH (m)
1.0
Stride frequency 97 99 97
0.5 (strides/min)
Force (N/kg)

0.0 LdF, leading forelimb; LdH, leading hind limb; TrF, trailing forelimb; TrH, trailing
20 40 60 80 100 hind limb.
-0.5 Data from Deuel and Park (1990).
-1.0

-1.5
characteristics of the canter strides during the two tempi (alternate
-2.0
stride) and one tempi (every stride) lead changes are shown in Table
Time (% stance)
14.13 (Deuel & Park, 1990). During the lead changes, all four limbs
Fig 14.14  Vertical (above) and longitudinal (below) ground reaction forces have long stance durations and short swing durations, and there is
of the four limbs in the canter. a diagonal dissociation with contact of the leading hind hoof pre-
ceding that of the trailing fore hoof by 14.0–24.0 ms.
In the two tempi changes (Deuel & Park, 1990), the stride imme-
diately preceding the change (pre-change stride) has a slower speed,
apart spatially but are more closely grouped temporally and the shorter stride length and higher stride frequency than the stride
suspension is longer. The diagonal limb pair (leading hind and following the change (post-change stride). The shorter stride length
trailing fore) do not always make contact with the ground synchro- of the pre-change stride is a result of taking a shorter step between
nously, though the dissociation is very small and usually only the two forelimbs and covering less distance during the suspension
detectable with the aid of slow motion analysis. In the extended (Table 14.13). In pre-change strides ground contact of the trailing
canter either the fore or hind limb may contact the ground earlier forelimb precedes that of the leading hind limb, whereas in the
(Deuel & Park, 1990; Clayton, 1994b). post-change strides the sequence is reversed.
The ground reaction forces at the canter (Merkens et al., 1993) In the one tempi lead changes, the strides share characteristics of
show marked differences between the trailing and leading limbs both the pre- and post-change strides in the two tempi changes
(Fig. 14.14). In collected canter the peak vertical ground reaction (Table 14.13). The support sequence varies between horses, reflect-
force is smallest in the trailing hind limb (approximately equal to ing individual differences in technique, but the diagonal dissocia-
body weight) and largest in the trailing forelimb (1.5 times body tion almost always involves placing the leading hind limb before
weight). The peak vertical force in the leading hind and leading the trailing forelimb. There is a pronounced reliance on overlap
forelimbs was approximately 1.2 times body weight. With regard to between the two hind limbs in the early part of the stride, which is
the longitudinal ground reaction force, the trailing hind limb is similar to the post-change stride for the two tempi changes. Later
primarily propulsive; it acts to change the direction of movement in the stride there is a reliance on forelimb overlap. Higher competi-
of the center of mass from forward and downward to forward and tion scores are associated with a longer time in hind limb support
horizontal. The leading hind limb and trailing forelimb (diagonal and a shorter time in forelimb support (Deuel & Park, 1990).
pair) are principally responsible for supporting the body weight and
supplying forward propulsion, with the trailing forelimb having
a particularly large propulsive component. The leading forelimb The canter pirouette
raises the center of mass as the horse moves into the suspension by In the canter pirouette, horses are supposed to maintain the tempo
exerting large vertical and braking forces. and rhythm of the collected canter strides. However, a study of
horses competing in the individual medal finals at the Barcelona
Olympics showed that neither the tempo nor the rhythm of the
Lead changes at the canter collected canter strides was maintained in the canter pirouette. The
When a dressage horse performs a flying lead change, the leading tempo was significantly slower in the pirouette strides (68 strides/
hind limb and leading forelimb change during the suspension min) than in the collected canter strides (95 strides/min). The
phase. In high-level competitions, lead changes are performed footfalls of the diagonal limb pair were dissociated in the
in series at intervals of four, three, two or one strides. The pirouette strides, giving a distinct four-beat rhythm, in contrast

326
 Jumping

ability to perform type 1 transitions from walk to trot is significantly

Collected Canter
related to level of training; more highly trained horses perform a
TrH TrH larger number of type 1 walk–trot transitions (Argue & Clayton,
LdH LdH 1993a).
TrF TrF Trot to walk transitions are initiated by placement of the second
LdF LdF forelimb during a diagonal support phase at the trot. In type 1
transitions, the horse then proceeds in the walk. In type 2 transi-
tions, the tripedal support phase is followed by lifting of the diago-
Canter Pirouette nal pair leaving a forelimb in single support. The diagonal hind
TrH TrH limb is then placed, and the horse continues in a normal walk
LdH LdH sequence. The frequency of type 1 transitions from trot to walk is
TrF significantly related to level of training, with the more highly
LdF trained horses showing a higher percentage of type 1 transitions
(Argue & Clayton, 1993a).
Transitions between trot and canter are differentiated according
Time to whether the horse initiates the canter from a forelimb or a
hind limb (Argue & Clayton, 1993b). Neither type of transition
Fig 14.15  Temporal characteristics of the collected canter strides and the
involves limb support sequences that are not typical of the trot or
canter pirouette strides. The shaded bars represent the stance phases of the
limbs. LdF, leading forelimb; LdH, leading hind limb; TrF, trailing forelimb; canter. A transition initiated by the forelimb starts with a diagonal
TrH, trailing hind limb. support phase in trot. The opposite forelimb is then placed on the
ground and it becomes the leading limb as the horse continues in
a canter sequence. For the transitions initiated by a hind limb, the
forelimb is lifted during the diagonal support phase of the trot. The
to the three-beat rhythm of the canter strides. Also, there was no hind limb that remains on the ground becomes the trailing hind
suspension between successive pirouette strides (Burns & Clayton, limb. The diagonal pair (leading hind–trailing forelimbs) are then
1997). placed on the ground and the horse continues in a canter sequence.
The lack of forward movement in the pirouette makes it difficult The type of trot to canter transition is not related to the horse’s level
for the horse to maintain its balance. The horse compensates by of training.
increasing the limb stance durations and the overlaps between Two types of transitions were observed between canter and trot.
limbs (Fig. 14.15). This was particularly obvious for the inside hind In one type, the horse initiates the change by springing from the
limb, in which the stance duration was greatly prolonged as a means diagonal support phase of the canter to the opposite diagonal and
of maintaining the horse’s balance in the absence of forward move- then continues in the trot. The other type of transition is initiated
ment (Burns & Clayton, 1997). from the leading forelimb single support. Instead of proceeding
into a suspension, the trailing hind limb is placed on the ground
to give a diagonal support phase, from which the horse springs into
The capriole a suspension and continues in a trot sequence. The type of transi-
Knezevic et al. (1987) used high-speed cinematography to describe tion from canter to trot is not related to level of training (Argue &
the capriole in a Lipizzan stallion at the Spanish Riding School in Clayton, 1993b).
Vienna. This is a movement in which the horse elevates the fore- A different study (Biau et al., 2002) used an accelerometer
hand, then leaps into the air and, at the culmination of the airborne attached beneath the sternum to measure dorsoventral and longi-
phase, kicks backwards with the hind limbs. At the culmination of tudinal acceleration during downward transitions from canter to
the leap, the ventral rump contour was 158 cm above the ground. trot, canter to walk, canter to halt, trot to walk, trot to halt and walk
The hip, stifle and tarsal joints showed extreme extension when the to halt. Wavelet analysis indicated that duration of the transition,
hind limbs were fully stretched behind the horse. Maximal exten- and the energy and frequency of the dorsoventral acceleration dis-
sion of the fetlock and coffin joints occurred about 20 ms later. tinguished between the types of transitions. Dressage training was
associated with an increase in the duration of the transition and
minimization of the variation in energy and frequency.
Transitions
A study of transitions between the walk and trot classified the transi-
tions into two types (Argue & Clayton, 1993a). Type 1 transitions
Conformation of dressage horses
are those in which all the limb support sequences are typical of the Elite dressage horses tend to have short necks, which perhaps
walk or trot. Type 2 transitions include intermediate steps between reduces the leverage of the head and neck and so facilitates collec-
the two gaits in which the support sequences are not typical of tion and self-carriage. In the forelimb, the scapula is sloping and
either gait. Type 1 transitions are thought to be more desirable and the elbow joint has a large angle. A long humerus, a small hip angle
should be rewarded by judges. The likelihood of performing a type and a long sloping femur are associated with high gait scores. The
1 transition may be influenced by the skill of the rider in giving the fore and hind phalanges tend to be long and upright (Holmström
aids at the appropriate time in the stride. et al., 1990).
For the transitions from walk to trot, in a type 1 transition the
horse springs from a diagonal support phase in the walk directly
into the trot. Type 2 transitions are initiated from either a diagonal Jumping
or a lateral support phase in the walk. From a diagonal support
phase, either the fore or hind hoof is raised, then the opposite Jumping sports require the horse to raise the center of mass (CM)
diagonal pair is placed for a three-limb support. The original fore high enough for all body parts to clear the height and width of an
or hind limb is lifted, and the horse springs from the remaining obstacle. The phase of the stride in which the horse jumps is an
diagonal pair into the trot. In the transitions initiated from a lateral extended suspension. The pattern of lift-off from the hind limbs
support phase, one limb is raised leaving a single supporting limb. and landing on the forelimbs is most easily incorporated into the
The diagonal fore or hind limb is then placed to give a diagonal canter or gallop stride, though horses can also jump from the trot,
support phase from which the horse springs into the trot. The walk or even the halt.

327
 14 Performance in equestrian sports

Terminology 1989). Fewer knock-downs occurred when there was a larger dis-
tance between the trailing forelimb in approach stride 1 and the
Terminology for the strides during the approach to and departure base of the fence (Deuel & Park, 1991).
from a fence for horses jumping at a canter has been described The timing and coordination of the limb movements during the
(Clayton, 1989a). The stride in which the jump occurs is the jump approach, jump stride, and departure are shown in Table 14.14.
stride; its components are the lift-off, jump suspension and landing. Approach stride 2 and the strides that precede it are fairly typical
The lift-off comprises the stance phases of the two hind limbs canter strides performed with an elevated head and neck position.
immediately preceding the jump. The jump suspension is the air- The ability to maintain a high stride frequency during the approach
borne phase starting when the last hind limb leaves the ground at strides is a characteristic of good jumpers (Barrey & Galloux, 1997).
lift-off and ending at ground contact of the trailing forelimb during Approach stride 1 is a short, quick stride; both the stride length
landing. The landing comprises the stance phases of the two fore- and stride duration are significantly shorter than in approach stride
limbs after the jump suspension. The strides preceding the jump 2. This stride has a distinctly four-beat rhythm with the leading hind
stride are the approach strides and those following the jump stride limb contacting the ground before the trailing forelimb. The neck
are the departure strides. Both the approach and departure strides is stretched forward and downward in preparation for lift-off. This
are numbered from the jump outwards (Fig. 14.16). action is similar to the ‘gather’ shown by human jumpers during
the transition between the approach and lift-off. The objective of
the gather is to lower the CM prior to the lift-off foot contacting the
Jumping mechanics ground. This minimizes the need to overcome a downward move-
ment of the CM before driving the body into the air. It has even
The approach and lift-off been suggested that the amount of lowering during the gather bears
The path of the horse’s CM and the angular momentum of its body a direct relationship to the height jumped. The leading hind limb
during the jump suspension are determined during the lift-off. After has a very short stance duration in approach stride 1 and the horse’s
the jump suspension begins, these properties cannot be changed body does not roll forward over this limb as in a normal stride.
until the horse makes contact with another object or with the The forelimbs are stretched forward at ground contact and, conse-
ground. Therefore, the approach and lift-off are extremely impor- quently, have a relatively small angle between the palmar aspect
tant in determining the outcome of the jump. of the metacarpus and the ground (Clayton & Barlow, 1991). The
The positions of the limb placements on the lift-off side do not forelimbs initiate the upward movement of the forehand, convert-
differ between a vertical fence and an oxer, or between fences of ing forward movement into vertical movement. This involves reduc-
different heights in the range of 1.10–1.40 m. One study showed ing the horizontal speed, elevating the forehand and rotating the
that, in 92 of 96 trials, the limb placed closest to the fence on the trunk segment into an appropriate position for lift-off. Because the
lift-off side was the leading forelimb in approach stride 1. In the forehand is already starting to move upwards, the leading forelimb
remaining four trials, all of which were over fences, the limb placed is pulled off the ground relatively early, when it is more or less in
closest to the fence was the leading hind limb (Clayton & Barlow, a vertical position.

Approach stride 2 Approach stride 1 Take–off Jump suspension Landing Departure stride 1
Jump stride
Fig 14.16  Terminology for the strides preceding and following the jump.

Table 14.14  Stride characteristics of horses jumping a vertical fence 1.55 m high

Approach stride 1 Approach stride 2 Jump stride Departure stride 1

Horizontal speed (m/s) 7.3 6.3 5.9 6.5
Stride length (m) 4.07 2.39 4.87 3.35
Stride frequency (strides/min) 108 157 73 116
Stance TrH (ms) 157 145 197 147
Stance LdH (ms) 179 129 195 171
Stance TrF (ms) 149 188 140 174
Stance LdF (ms) 134 159 176 168

LdF, leading forelimb; LdH, leading hind limb; TrF, trailing forelimb; TrH, trailing hind limb.
Data from Clayton and Barlow (1991).

328
 Jumping

When a horse jumps a fence that is less than 1 m high, it requires the hind limbs and land on the forelimbs. Both hind limbs have
little, if any, elevation of the CM to clear the fence and, conse- relatively long stance durations, which allows the generation of
quently, the forces required to jump a small fence are not much large impulses. The two hind limbs show almost identical ground
greater than those in a canter stride. For example, in jumping a fence reaction force profiles. The vertical force rises to a plateau, with a
0.8 m high, the combined vertical impulses of the fore and hind peak amplitude around 130% body weight for a horse jumping
limbs increase by only about 8% on the lift-off side and 3% on the a vertical fence 1.3 m high. The longitudinal forces are predomi-
landing side. For fences higher than about 1.0 m, the need to elevate nantly propulsive (Fig. 14.17) (Schamhardt et al., 1993).
the CM increases progressively with the height of the fence, with a Studies using an accelerometer attached to the thorax beneath the
consequent increase in the vertical forces at lift-off and landing. A sternum have shown that the action of the hind limbs produces a
considerable increase in vertical forces has been recorded when the lower acceleration peak on the trunk segment than the push-off by
height of the fence was raised from 1.3 to 1.5 m (Fig. 14.17) the forelimbs (Fig. 14.18). However, the inclination of the trunk
(Schamhardt et al., 1993), and a vertical force of 3.85 times body axis during the hind limb push-off reduces the amplitude registered
weight has been measured in a horse jumping a vertical fence by the direction-sensitive accelerometer, so the acceleration due to
1.53 m high (Preuschoft, 1989). Both forelimbs exert a large the hind limb action may be underestimated. The hind limb accel-
braking impulse that decelerates the forward movement. The trail- eration peak at lift-off is significantly greater for fences with width
ing forelimb provides a little propulsion in the terminal part of its (oxer, 1.48 g; water jump, 1.74 g) than for those with height only
stance phase, but the leading forelimb provides only a braking force (vertical, 1.27 g) (Barrey & Galloux, 1997). This is because the flight
as a result of the early cessation of its stance phase (Schamhardt path must be higher and longer when jumping a wider obstacle.
et al., 1993). The braking action of the forelimbs causes a reduction For elite show jumping horses jumping a vertical fence 1.50 m
in horizontal velocity (Clayton & Barlow, 1991) and an upward high during an international competition (Bogert et al., 1994), a
acceleration of the trunk segment (Barrey & Galloux, 1997). After similar movement pattern was observed in all horses during lift-off.
the forelimbs leave the ground, the head and neck are raised, which The duration of the lift-off phase averaged 221 ms. At the start of
helps to establish an advantageous position for optimal power lift-off the trunk was rotating backwards with an angular velocity of
production during the push-off by the hind limbs. 150°/s. The direction of rotation changed almost linearly to 50°/s
A short suspension intervenes between approach stride 1 and the forward rotation at the start of the jump suspension. During this
jump stride. The jump stride is distinguished from the preceding time the horizontal velocity initially decreased from 4.5 m/s at the
and following strides by the inclusion of the jump suspension, start of lift-off to 3.6 m/s, then increased to 6.5 m/s at lift-off. The
which results in significant increases in stride length and stride vertical velocity, which was zero at the start of lift-off, reached a
duration. At lift-off the hind limbs often contact the ground syn- negative value of 1.0 m/s before increasing to 4.0 m/s at lift-off. The
chronously and at almost equal distances from the fence. Their initial downward movement of the trunk was a result of the total
functions are to provide upward and forward propulsion, and to vertical GRF being less than body weight. The neck rotated down-
reverse the direction of rotation of the trunk segment. Forward rota- ward relative to the trunk throughout lift-off with an almost con-
tion of the trunk is necessary for the horse to leave the ground from stant angular velocity.

3.0 3.0 3.0 3.0

Force (N/kg)

0 0 0 0

–3.0 –3.0 –3.0 –3.0

20 20 20 20
Force (N/kg)

0 0 0 0
0 100 0 100 0 100 0 100

Time (% stance)
Fig 14.17  Longitudinal (above) and vertical (below) ground reaction forces during the approach and lift-off for a horse jumping a vertical fence 1.1 m high.
The traces from left to right represent the trailing forelimb in approach stride 1, the leading forelimb in approach stride 1, the trailing hind limb at lift-off
and the leading hind limb at lift-off. Stance durations of the limbs are: trailing forelimb, 0.20 s; leading forelimb 0.18 s; trailing hind limb, 0.21 s; leading
hind limb, 0.22 s).

329
 14 Performance in equestrian sports

4
Stride frequency and FFT Acceleration
module of the approach strides peaks
3
Landing
Forelimbs

2
Dorsoventral acceleration (g)

Hindlimbs
1

-1

-2

Landing
Take off
Stride-2

Stride-1

Jump
Durations
-3

Time (s)
Fig 14.18  Dorsoventral accelerometer recording of the approach, lift-off, jump suspension, landing and departure. FFT, Fast Fourier transformation.
Reprinted from Barrey & Galloux (1997) © EVJ Ltd.

The hip joint extended during the first 152 ms at lift-off, then jumper over a fence 1.3 m high. It might be expected from these
maintained the same angle until leaving the ground. The stifle and results that horses with poor technique would show earlier signs
hock joints initially flexed then extended with the transition from of wear and tear injuries than horses with good technique. Poor
flexion to extension occurring earlier in the hock than the stifle. jumpers, even when they clear a fence, tend to have a higher ratio
The fetlock showed two phases of extension with a relatively con- between the acceleration provided during lift-off by the forelimbs
stant angle in the period between. Maximal fetlock extension and the hind limbs. In other words, they increase the braking action
occurred late in the stance phase after which there was rapid of the forelimbs in approach stride 1 to compensate for a weak
flexion. The total power production by the hind limbs was esti- acceleration impulse of the hind limbs at lift-off (Barrey & Galloux,
mated to be 13 000 J in this study, which represents an average 1997).
power production of 59 000 W (or about 100 W/kg) during the
lift-off phase. These findings indicate that most of the energy
required to clear the fence is provided by the hind limbs during The jump suspension
lift-off (Bogert et al., 1994). Power production, however, appears The jump stride has a much longer stride length than the approach
to be very dependent on the height of the fence and the speed with or departure strides due to the distance covered during the jump
which it is approached. In the study by Bogert et al. (1994), fence suspension. In one study the total distance jumped from lift-off to
height was 1.50 m. Dutto et al. (2004) studied horses jumping a landing did not differ between a vertical fence and an oxer, but
much lower fence (0.63 m) which they approached at trot not increased significantly with fence height (Clayton & Barlow, 1989).
canter; these authors calculated a power output during hind limb In a different study, the jump distance was longer over an oxer than
push of only 30 W/kg. When studying horses jumping a 1.15 m a vertical fence, with the increase being approximately equal to the
obstacle, Bobbert and Santamaría (2005) found a value of 75 W/ spread of the oxer (Deuel & Park, 1991).
kg. In the latter study the authors also investigated the contribu- After lift-off, the motions of the body and limb segments are
tion of the forelimbs. They calculated that during forelimb push- coordinated so that the angular velocity of the trunk remains almost
off total energy first decreased by 3.2 J/kg and then increased again constant throughout the jump suspension. All the segments act
by 4.2 J/ kg. Energy stored in the elastic structures of the limb was synergistically with the approximate contributions of the different
calculated to be only 0.4 J/kg in the trailing forelimb and 0.23 J/kg segments to the angular momentum being 50% from the trunk,
in the leading forelimb. From this observation it was concluded 25% from the hind limbs, 2% from the head–neck and 5% from
that the forelimbs are not merely used as passive springs, but a the forelimbs (Galloux & Barrey, 1997). The presence of the rider
considerable amount of energy is first dissipated and then regener- has virtually no effect on rotation of the horse’s body.
ated by muscular power. As the horse passes over the top of the fence, the forelimbs gener-
Each horse has its own individual jumping technique that is ally have less clearance than the hind limbs, and clearance by all
repeatable from jump to jump and produces a characteristic accel- limbs decreases with fence height (Jelen, 1976). Furthermore, a
eration profile (Barrey & Galloux, 1997). Over a small fence, 0.8 m good horse paired with a good rider shows a smaller discrepancy
high, horses with good technique that folded their lower limbs between the height of the fence and the height of the limbs; in other
during the jump suspension, and had vertical and longitudinal words horses show less tendency to over-jump with a good rider.
force profiles that were similar in shape and magnitude to those of Most jumping errors are a result of inappropriate aids from the rider
a canter stride. However, a horse with poor jumping technique (Lauk et al., 1991). A study of trunk accelerations showed that 87%
registered considerably higher forces both at lift-off and landing of jumping faults could be blamed on a forelimb error (Barrey &
(Schamhardt et al. 1993). The forces used by a poor jumper over a Galloux, 1997). More faults occur at vertical fences than at oxers or
fence 0.8 m high were similar in magnitude to those used by a good water jumps.

330
 Jumping

sides. The high peak vertical forces at landing in the trailing fore-
The landing and departure limb also translate to higher peak flexor joint moments and hence
During landing, the two forelimbs are separated by only a short higher loading of the flexor tendons than in the leading forelimb
distance. The metacarpal segment of the trailing forelimb (the first (Meershoek et al., 2001a). In a study using an inverse dynamics
limb to land) is almost vertical when the hoof makes ground approach Meershoek et al. (2001b) analyzed horses jumping fences
contact while the leading forelimb contacts the ground with a more of 0.8 , 1.0 m and 1.2 m and calculated peak forces in the suspen-
acute angulation (Clayton & Barlow, 1991). The trailing hind limb sory ligament, the deep digital flexor tendon and its check ligament,
is usually placed between the previous placements of the two fore- and the superficial digital flexure tendon, which reached maximum
limbs, and this characteristic is associated with fewer jumping pen- values of 15 500 kN, 13 900 kN and 11 700 kN, respectively. When
alties (Deuel & Park, 1991). The leading hind limb is placed further comparing the recorded values to the maximal load at rupture of
away from the fence than the forelimbs. these structures, it became clear that the superficial digital flexor
The placements of the leading forelimb and the leading hind tendon is loaded dangerously close to its limit during show jumping,
limb on the landing side are closer to the last element of an oxer possibly explaining the high incidence of lesions in this structure
than a vertical at all fence heights, and there is a trend toward in jumping horses.
landing further from the vertical than the oxer for the other limbs. The forelimbs absorb the initial force of landing and this is
The limb displacements from the fence on the landing side increase reflected in the high peak amplitudes of the vertical forces, espe-
with fence height but this effect becomes less apparent in the later cially the trailing forelimb in which a vertical force peak of twice
strides as the horse moves away from the fence (Clayton & Barlow, body weight has been recorded in a horse jumping a vertical fence
1991). Therefore, the fewer the strides between fences in a combina- 1.3  m high (Schamhardt et  al., 1993). This limb makes ground
tion, the more important it is to take into account the height and contact with an almost vertical orientation, which is not conducive
type of the in-going fence in determining the appropriate distance to generating a braking force. Therefore, the longitudinal force of
between fences. the trailing forelimb at landing is entirely propulsive, whereas the
During landing the trailing forelimb has a very short stance dura- leading forelimb exerts predominantly a braking force with a
tion as the horse rapidly rolls forward onto the leading forelimb smaller propulsive component in its terminal part (Merkens et  al.,
(Clayton & Barlow, 1991; Deuel & Park, 1991). Horses that have a 1991; Schamhardt et  al., 1993). The actions of the trunk, hind
longer time interval between contact of the two forelimbs are less limbs and head and neck segments play a large role in reversing
likely to knock the fence down (Deuel & Park, 1991). the direction of rotation of the horse’s trunk which is necessary
The trailing forelimb has the highest peak vertical forces both at to allow the hind limbs to contact the ground underneath the
lift-off and at landing (Fig. 14.19) (Schamhardt et al., 1993), which body mass.
may explain the fact that some horses have a preferred lead for lift- In departure stride 1 the horse regains its balance and the hind
off and landing and habitually switch to this lead one or two strides limbs re-establish forward movement by generating large propulsive
before the jump. These horses probably have either a subclinical longitudinal forces and impulses, especially in the trailing hind
lameness or a marked strength asymmetry between the left and right limb (Schamhardt et al., 1993). This stride has a four-beat rhythm

3.0 3.0 3.0 3.0

Force (N/kg)

0 0 0 0

–3.0 –3.0 –3.0 –3.0

20 20 20 20
Force (N/kg)

0 0 0 0
0 100 0 100 0 100 0 100

Time (% stance)
Fig 14.19  Longitudinal (above) and vertical (below) ground reaction forces during the landing and departure for a horse jumping a vertical fence 1.3 m
high. The traces from left to right represent the trailing forelimb at landing, the leading forelimb at landing, the trailing hind limb in departure stride 1 and
the leading hind limb in departure stride 1. Stance durations of the limbs are: trailing forelimb, 0.19 s; leading forelimb, 0.21 s; trailing hind limb, 0.17 s;
leading hind limb, 0.22 s.

331
 14 Performance in equestrian sports

with the leading hind limb contacting the ground in advance of the correlates to good jumping performance. The latter feature has the
trailing forelimb. The stride length is relatively short (Clayton & strongest correlation to overall gait quality of all conformational
Barlow, 1991), and the distance between the two hind limb place- variables (Holmström & Philipsson, 1993). Show jumpers have
ments is particularly variable. Sometimes they are placed almost shorter backs than dressage horses, possibly related to the supple-
equidistant from the fence, other times they are widely separated. ness required in the dressage horse (Johnston et al., 2004).
Fewer knock-downs were recorded when the trailing hind limb Kinematics can, to a certain extent, be seen as a dynamic measure
landed closer to the fence (Deuel & Park, 1991). A study of a two- of conformation. In a study relating back kinematics to jumping
stride double in which the two fences were separated by a distance performance using a relatively simple 4-marker set-up. Cassiat et al.
of 10.96 m (Hole et al., 2002). The two intermediate strides had (2004) found significant differences in kinematic patterns between
similar velocities, but the second intermediate stride had signifi- good and poor jumpers when jumping a 1.0 m fence, from which
cantly shorter stride length and stride duration. The stride lengths they concluded that criteria based on back kinematics might be
of the two intermediate strides showed low inter-horse variability, developed that could help in selecting talented show jumpers.
but the step lengths were much more variable indicating that dif-
ferent horses achieved similar stride lengths using different combi-
nations of step lengths. Compared with published descriptions of The effect of early training on jumping ability
horses jumping a single fence of similar size, the first intermediate and the predictability of jumping performance
stride resembled the first departure stride and the second intermedi-
ate stride resembled the final approach stride. Unlike racehorses, peak performance in show jumpers and dressage
horses is achieved at a later age (approximately 10–16 years). There-
fore, the time lapse between birth and the final evaluation of the
Water jump technique athletic capacity of the animal is very long and requires a substan­
A study performed during the Barcelona Olympics identified the tial investment of time and money before it is known whether
factors that influence success in clearing a water jump. The vertical any financial return can be expected. As a consequence the inter-
velocity and the angle of projection at lift-off were significantly generational interval is very long and genetic progress through
greater, and the horizontal distance between the leading hind limb selective breeding is slow. In a long-term study, Santamaría and
and the CM at lift-off was significantly shorter, in horses that were co-workers investigated two possible ways in which show jumpers
successful in clearing the width of the water jump compared with might be raised more cost-effectively. Their first research question
those that failed to clear the entire width (Clayton et al., 1995, was whether good predictors for future jumping ability could pos-
1996; Colborne et al., 1995). The horizontal velocity at lift-off did sibly be found through the analysis of jumping technique at foal
not influence success, nor did the height of the CM at the start and age (Santamaría et al., 2005). This would make earlier selection
end of the jump suspension. It was concluded that, to be a success- possible. An indication that jumping capacity might be more innate
ful water jumper, a horse must generate a large vertical velocity than acquired through specific training came from the work of
during lift-off. Vertical velocity at the start of the jump suspension Fabiani (1973), who reported that jumping ability showed little
is highly correlated with trunk angle at lift-off and horizontal dis- change after 6 months of jump training. The second question was
tance between the leading hind hoof and the horse’s CM at both whether jumping performance could be improved by starting train-
the start and end of lift-off (Colborne et al., 1995). ing much earlier than the usual age of 3 years (Santamaría et al.,
2006). This was based on the empirical observation that many elite
human athletes start their careers as children, which suggests a
Puissance jumping technique lasting effect of such early athletic activity.
To answer these questions a cohort of 40 Warmblood foals
When jumping very high fences body position at lift-off is particu- was selected after weaning at age 6 months and divided into two
larly important. In a study of horses during a puissance competition groups. One group was raised traditionally (free paddock exercise
that started at a height of 1.80 m and ended in the last round at or group housing in open front stalls, depending on the season)
2.27 m, success was significantly positively correlated with the fol- until they were broken at age 3; the other group was subjected to a
lowing variables at lift-off: vertical velocity of CM, height of CM, specific jumping training program twice weekly combined with a
distance of CM from the fence; and was significantly negatively more general exercise regimen to ensure a sufficient level of physical
correlated with the distance of the leading hind limb to the CM fitness from 6 months to 3 years of age. At that age all animals were
(Powers, 2005). joined in a single group to be broken and to receive a common
exercise program to prepare for work under saddle. After a rest
Conformation of jumping horses period of 6 months, 30 of the now 4-year-olds were trained in a
single group as show jumpers for an entire year. This group of 30
The conformational trait that has been linked most consistently to horses consisted of 15 animals from each training group; the five
successful jumping performance is height at the withers (Fabiani, worst animals from each exercise group had been culled because of
1973; Langlois et al., 1978), which is more important in this economical constraints.
equestrian discipline than in dressage. The shoulder is preferably Kinematic examinations were performed while free jumping at
sloping and the fore pastern has been shown to be significantly ages 6 months, 4 years and 5 years and also at 5 years with a rider.
larger in elite jumping (and dressage) horses compared to other The outcome parameter for jumping performance was a puissance
riding horses (Holmström et al., 1990). Also, elite show jumpers competition held at the end of the experiment during which horses
have larger tarsal angles than other horses (Holmström et al., had to jump fences up to 1.50 m, whereas in the preceding year
1990). The ratio between the chest girth and wither height tends to they had been trained using obstacles with a maximum height of
be relatively low in good show jumpers, as does the ratio between 1.20 m. During the puissance competition performance was char-
cannon circumference and wither height (Fabiani, 1973). A long acterized as good (faultless), moderate (able to finish the contest
neck is thought to be an advantage for jumping horses, possibly but with faults and/or refusals), or bad (unable to finish the
because this feature makes it easier for the horse to maintain competition).
balance over the fence (Holmström, 2001). When the horse is Jumping technique was quite consistent from foal to adult horse
viewed from behind, width through the hips and shoulders is desir- (Santamaría et al., 2002, 2004). When the kinematic variables that
able (Langlois et al., 1978). The length of the pelvis has a positive were consistent over time were related to the outcome of the
correlation with jumping ability and a forward sloping femur also puissance competition as the main parameter for performance, it

332
 Eventing

appeared that elbow flexion, retraction angle of the hind limbs and would-be breeding stock at ages of 3 or 4 years (Santamaría et al.,
inclination of the trunk with respect to the horizontal were among 2006). It should be emphasized that the lack of effect of specific
the variables that were related to performance at 5 years of age and training for jumping does not take away the well-documented benefi-
could already be identified in the 6-month-old foals (Bobbert et al., cial effect of early exercise on the general development of the equine
2005; Santamaría et al., 2002, 2004). It was concluded, therefore, musculoskeletal system (Weeren et al., 2000; Weeren, 2007).
that it is indeed possible to use some of the characteristics of the
jumping technique exhibited by foals when they are free jumping
as predictors for future athletic performance. However, it should be The effect of a rider
stressed that many other factors influence eventual success of a In a kinematic study of the effect of a rider on jumping technique,
show jumper beyond jumping technique alone. The kinematic char- Powers and Harrison (2002) compared the same group of horses
acteristics that were identified are therefore probably of more help jumping a 1.0 m obstacle freely or with a rider. The rider signifi-
in developing an effective culling strategy than in picking out a cantly influenced the following variables: vertical velocity at lift-off,
future top show jumper. height of the CM at lift-off, distance of the CM to the fence at lift-
Early training resulted in a less variable and clearly more efficient off, maximal height of CM during the suspension phase, position
jumping technique in the 4-year-olds (Fig. 14.20) (Santamaría et al., of the CM when passing over the fence, and height of the CM at
2004, 2005). However, when reassessing the effects after a year of landing. They concluded that the effect of the rider on the horse
common and similar training, no significant difference could be was more due to behavioral influences than to the inertial effect of
detected anymore between the two former training groups. Also, the mass of the rider.
there was no relation of early training with performance, as measured
by the outcome of the puissance competition. Therefore, specific
training for jumping at foal age, at least when carried out according Eventing
to the relatively mild protocol as used in this study, has no measur-
able long-term effect. It was noted, however, that there was a tempo- The sport of eventing was originally devised as a test of athleticism
rary effect of early training on naïve horses that made them jump for cavalry horses, and the name ‘the military’ is still retained in
more efficiently and thus seemingly better than untrained horses. some languages. Event horses perform three tests: dressage, speed
Early jump training might therefore bias selection events of stud- and endurance, and show jumping. Since dressage and show
books, many of which nowadays include free jumping sessions for jumping are covered elsewhere in this chapter, this section will
focus on studies that are specific to the sport of eventing, especially
Trained horses Untrained horses those describing performance in the speed and endurance phases.
The multi-faceted nature of the sport dictates that event horses
must be versatile athletes. They should show relaxation and supple-
ness in the dressage phase; speed, strength, stamina and athleticism
in the endurance phase; and suppleness and coordination in the
show jumping phase. A long and energetically efficient galloping
stride is an important quality in a potential eventer. More data are
needed describing the techniques used to jump different types of
cross-country fences.

Dressage performance
In a study at the Barcelona Olympics, performance in the dressage
phase was correlated with finishing place. Stride length and speed
Forelimbs clearing fence in the extended canter were positively related to points awarded by
the judges for the canter. Horses that failed to complete the entire
3-day event had longer stride lengths and faster speeds in the
extended canter during the dressage phase than horses that finished
the competition (Deuel, 1995). The author suggested that charac-
teristics of the extended canter that are highly rewarded by the
judges have little relation to subsequent galloping and jumping
performance, and qualities favored by dressage judges may even
predict failure to finish the event.

Speed and endurance performance

Success in the speed and endurance phase is largely dependent on
having a ground-covering stride. Recordings made during the steeple­
chase phase of the 3-day event at the Seoul Olympics (Deuel & Park,
1993) showed an average speed of 12.1 m/s with a stride length of
Hindlimbs clearing fence 6.0 m and a stride frequency of 2.0 strides/s. Superior performance
Fig 14.20  Representative schematic drawings of horses from the two in the entire competition was significantly associated with a faster
training groups clearing a fence under saddle at 4 years of age (i.e. before speed and a longer stride length in the steeplechase phase. The score
the common 1-year specific jump training). (A) Clearing the fence with the was optimized with a gallop stride frequency between 1.85 and
forelimbs. Left: trained animals; right: untrained animals. The trained animals 2.05 strides/s and a speed between 13.0 and 14.3 m/s. Up to 7 m
have their center of gravity closer to the fence and a more flexed elbow there appeared to be no upper limit for optimal stride length.
joint. (B) Clearing the fence with the hind limbs. Left: untrained animals; In a study of the temporal and linear kinematics of horses
right: trained animals. The trained animals exhibit more retroflexion of the jumping a steeplechase fence (1.4 m high, 2.6 m wide at the base)
hind limbs. during a young riders’ competition (Leach & Ormrod, 1984), the

333
 14 Performance in equestrian sports

average distance jumped was 5.2 m. The distances of the limbs from
Table 14.15  Temporal and linear kinematic variables for horses
the base of the fence on the lift-off side were: trailing forelimb
jumping a table fence with a 61-kg rider (rider weight) and with a
2.6 m, leading forelimb 2.1 m, trailing hind limb 2.0 m and leading
61-kg rider plus an 18-kg weighted saddle pad (added weight)
hind limb 2.0 m. Unlike show jumpers, in which the trailing hind
limb distance from the fence was between those of the two fore-
limbs at lift-off, during steeplechasing both hind limbs were placed Rider weight Added weight
closer to the fence than the leading forelimb. The forelimb stance
Stance duration, TrF at 0.18 0.18
durations were approximately 0.13 s, whereas the hind limbs had
landing (s)
longer stance durations of around 0.17 s. In a regression analysis
the only temporal variable that was a significant predictor of the Stance duration, LdF at 0.21 0.21
distance jumped was the overlap between the forelimb stance landing (s)
phases. The longer the forelimbs overlapped, the shorter the hori-
zontal distance jumped. For the linear variables, the distance of the Stance duration, TrH in stride 0.19* 0.20*
trailing forelimb from the base of the fence was positively related D1 (s)
to the distance jumped. When this limb was further from the jump, Stance duration, LdH in stride 0.21* 0.22*
the horse used a flatter trajectory, which carried it farther across the D1 (s)
fence.
Distance TrF to fence at 173.5 159.3
landing (cm)
Effect of added weight on jumping Distance LdF to fence at 240.4* 222.7*
performance landing (cm)
In the past the competition rules stipulated that event horses had Maximum carpal angle, TrF at 191.0 190.0
to carry a minimum weight of 75 kg (rider and tack), which repre- landing (degrees)
sented anything from 10 to 17% of body weight depending on the
weight of the horse. It was suggested that this imposed an unsafe Maximum fetlock angle, TrF at 252.9 251.9
burden on small horses. A study was performed (Clayton, 1997b) landing (degrees)
to compare the landing kinematics of small (maximum height Maximum carpal angle, LdF at 190.7* 193.4*
164 cm) event horses when carrying a 61 kg rider versus carrying landing (degrees)
the same rider plus an 18 kg weighted saddle pad. The test fence
was a table fence with a sloping face that measured 1.1 m high, Maximum fetlock angle, LdF at 247.4* 250.7*
1.9 m wide at the base and 1.3 m wide at the top. The results (Table landing (degrees)
14.15) showed that with the extra weight the leading forelimb
landed closer to the fence, which confirmed the riders’ subjective *Values that differ significantly (p < 0.05).
impression that the horses were ‘cutting down’ during the landing. D1, first departure stride; LdF, leading forelimb; LdH, leading hind limb; TrF, trailing
This was probably a result of failing to generate sufficient impulse forelimb; TrH, trailing hind limb.
during the lift-off to compensate for the extra weight. Data from Clayton (1997b).
During landing both the fetlock and carpal joints of the leading
forelimb were significantly more extended with the extra weight as
a direct result of the greater force at landing. Since strain of the Cutting demonstrates the need for maneuverability during rapid
suspensory ligament and the superficial digital flexor tendon increase acceleration, deceleration and turning. Many sports call for a com-
with extension of the fetlock, the extra weight may have put the bination of stability and maneuverability.
horses at greater risk of suspensory desmitis or superficial digital
flexor tendonitis. Stance durations of both hind limbs increased in
departure stride 1, which may have been indicative of the use of Reaction time
these limbs to restore the horse’s balance by elevating the forehand, Reaction time is defined as the time that elapses between an external
which was burdened by the extra weight. As a result of these findings stimulus and the initial response to that stimulus. It is extremely
the minimum weight rule for event horses was abolished. important in sports that require fast movements and quick reflexes,
such as cutting, in which there is an offensive–defensive component
Western sports and in which the horse must perceive a visual stimulus and respond
appropriately within a short space of time.
Reaction time in response to an external stimulus is determined by
Western sports are traditionally associated with skills required of the sum of the pre-motor time and the motor time. Pre-motor time
horses and riders in ranching, and are based on breeds such as the is dependent on the speed of processing in the central nervous system
Quarter Horse or Australian Stock Horse. The sports usually require and speed of conduction along the nerve to the motor end plate.
a short period of intense exercise, and may require speed and accel- Motor time reflects the speed of muscle contraction, which is depen-
eration, strength, stability, or explosive power. dent on muscle fiber type. The reaction times of human athletes are
substantially faster than those of non-athletes (Kroll & Clarkson,
1977). Because nerve conduction velocities do not appear to differ
Stability and maneuverability between individuals (Bodine-Rees & Bone, 1976), it has been sug-
The qualities of stability and maneuverability are required to differ- gested that the faster reaction time of athletes may be attributed to
ent degrees in different sports. Stability is enhanced by lowering the the superior functioning of the central nervous system. Furthermore,
COM and positioning it centrally within the base of support. the total reaction time is significantly correlated with pre-motor time
Maneuverability, on the other hand, is enhanced when the center but not with motor time (Viitsalo & Komi, 1981). In equine athletes,
of mass is high and lies close to the perimeter of the base of support, Clayton (1989b) found that the reaction time of cutting horses in
so that it is easily displaced outside the base of support. An example response to a visual stimulus ranged from 110 to 370 ms, and that
of a sport based on stability is roping, in which the horse must horses with shorter reaction times had higher competition earnings
remain balanced while resisting the forces applied by a roped calf. than those that were less successful in competition.

334
 Western sports

Western pleasure gaits Cutting is a high-intensity activity in which the horse must have
quick reaction times and the ability to turn and accelerate in either
Stock types breeds, such as the American Quarter Horse and the direction in response to the unpredictable movements of the calf.
American Paint Horse, comprise the majority of horses in North Some horses are naturally more talented than others, and ‘cow
America. One of the popular competitions for these horses is sense’ is a highly prized trait. Genetic evaluation of over 3000 horses
western pleasure classes, in which the horses are required to walk, competing at the World Championship Futurities over a 9-year
jog and lope in both directions of the arena. The jog is a type of period yielded a heritability estimate of cutting ability of 19 ± 5%
trot defined as a two-beat diagonal gait that is smooth and ground- (Ellersieck et al., 1985).
covering. The lope is a type of canter defined as a three-beat gait The tactics used to keep the calf separate from the herd are similar
performed with rhythm, forward motion and ease. to defensive play that blocks the progress of the offensive player in
Analysis of the temporal gait kinematics of the jog and lope sports such as American football. A study was performed to deter-
indicated that both gaits were actually performed as four-beat step- mine factors that differed between horses with different levels
ping gaits with a lateral sequence of footfalls (Nicodemus & Clayton, of ability (Clayton, 1989b). A group of cutting horses that were
2001). Thus, both gaits had the same footfall sequence as the walk. of similar ages, training histories and competitive opportunities,
Furthermore, the lope was performed without suspension and the were divided into two groups according to their competitive win-
limb support sequence included a period of quadrupedal support. nings. The five horses in the ‘average’ group had won less than
In 2003, the executive committee of the American Quarter Horse $35 000. The seven horses in the ‘elite’ group had won more
Association made changes to the rules with the intention of encour- than $35 000. The horses worked a mechanical flag rather than a
aging more forward-moving gaits. However, subsequent kinematic live steer, which allowed the use of a standard test to ensure that all
analysis (Nicodemus & Booker, 2007) performed during a national horses performed a similar series of turns and runs in each direc-
caliber show indicated that the speed of the jog was 1.13 ± 0.14 m/s tion. The mechanical flag is a piece of heavy cloth measuring about
and the speed of the lope was 1.77 ± 0.13 m/s, both of which are 30 cm2. All the horses had worked with the mechanical flag regu-
slower than the typical speed for the walk to trot transition. It is not larly and frequently during their careers.
surprising that both gaits were again performed with a four-beat Horses in the two groups differed significantly in several ways
rhythm and lateral sequence of limb placements, as in the walk (Table 14.17). Elite horses had faster reaction times than average
(Table 14.16). horses, which allowed them to respond more quickly when the flag
In the jog, the rhythm showed diagonal couplets and stance dura- began moving and to stop sooner after the flag stopped moving.
tions were around 68% of stride duration for the forelimbs and The faster reaction times of the elite horses may have some value
60% of stride duration for the hind limbs. During each half of the as a predictor of performance. As a result of their faster reaction
stride, the limb support sequences, starting at contact of a hind limb times the elite horses were significantly closer to the flag throughout
were as follows: quadrupedal (8.5%); tripedal with two hind limbs the run and were less likely to overrun the flag when it stopped
(2%); bipedal diagonal (30%); and tripedal with two forelimbs moving than the average horses. Other features of the performance
(9.5%). were that elite horses leaned their shoulders in the direction of the
In the lope, the footfalls occurred as diagonal couplets and the turn before pushing off against the ground, which is a more effective
typical footfall sequence was: trailing hind; trailing fore; leading method of turning. The average horses were more likely to move
hind; leading fore. There was no period of suspension. The limb their hooves sideways as the first indication of turning, rather than
support phases, starting with contact of the trailing hind limb were leaning into the turn and pushing sideways. The elite horses also
as follows: bipedal diagonal (leading front, trailing hind); unipedal tended to turn using fewer strides than the average horses though
trailing hind; bipeal lateral (trailing hind and trailing fore); tripedal this difference did not reach statistical significance.
(trailing hind, trailing fore, leading fore); bipedal diagonal; tripedal In summary, cutting is a high-intensity sport that requires great
(leading hind, trailing fore, leading fore); bipedal lateral (leading agility and an inherent ‘cow sense’. Successful cutting horses have
hind, leading fore); and unipedal leading fore. Thus, the limb quick reaction times and an economical turning technique that
support sequence is very different from that of a canter. allows them to follow the calf closely at all times.

Cutting
Cutting horses perform almost independently of the rider; during Table 14.17  Comparison of elite and average cutting horses
the judging points are lost if the rider cues the horse. The objective working a mechanical flag
is to sort one calf from a herd and then to prevent it returning to
the other calves. The sport evolved from the practical aspects of Elite horses Average horses
ranch work, with the first competition taking place around the turn
of the century. Since then, there has been increasing interest in Reaction time after flag starts 200 ± 66* 282 ± 74*
cutting as a sport and it is currently one of the most rapidly growing moving (ms)
equestrian sports. In competition the horse is judged for 2.5 min
Time to stop after flag stops 386 ± 108* 492 ± 94*
during which it usually cuts two or three calves.
moving (ms)

Table 14.16  Speed and stride duration jog and lope of the Distance from flag at start of 52 ± 23* 81 ± 32*
western pleasure horse run (cm)
Maximal distance from flag 148 ± 28* 221 ± 23*
Jog Lope during run (cm)

Speed (m/s) 1.13 ± 0.14 1.77 ± 0.13 Distance from flag at end of 55 ± 26* 78 ± 32*
run (cm)
Stride length (m) 1.09 ± 0.07 1.27 ± 0.04
*Values that differ significantly (p < 0.05).
Stride duration (s) 0.92 ± 0.03 0.72 ± 0.04
Values are mean ± SD.
Data from Nicodemus and Clayton (2001). Data from Clayton (1989b).

335
 14 Performance in equestrian sports

References
Alexander, R.M., 1984. Stride length and speed accelerometric device. Equine Vet. J. 18 Cano, M.R., Miro, F., Diz, A.M., Aguera, E.,
for adults, children, and fossil hominids. (Suppl.), 156–160. Galisteo, A.M., 2000. Influence of training
Am. J. Phys. Anthropol. 63, 23–27. Barrey, E., Desliens, F., Poirel, D., Biau, S., on the biokinematics in trotting
Alexander, R.M., 2002. Tendon elasticity and Lemaire, S., Rivero, J.L., et al., 2003. Early Andalusian horses. Vet. Res. Commun. 24,
muscle function. Comp. Biochem. Physiol. evaluation of dressage ability in different 477–489.
Part A-Mol. Integr. Physiol. 133, 1001–1011. breeds. Equine Vet. J. 34 (Suppl.), 319–324. Cano, M.R., Vivo, J., Miro, F., Morales, J.L.,
Alexander, R.M., Trestik, C.L., 1989. Bistable Barrey, E., Galloux, P., 1997. Analysis of the Galisteo, A.M., 2001. Kinematic
properties of the hock joint of horses equine jumping technique by characteristics of Andalusian, Arabian and
(Equus spp.). J. Zool. 218, 383–391. accelerometry. Equine Vet. J. 23 (Suppl.), Anglo-Arabian horses: a comparative study.
Alexander, R.McN., 1988. Elastic mechanisms 45–49. Res. Vet. Sci. 71, 147–153.
in animal movement. Cambridge Barrey, E., Landjerit, B., Wolter, R., 1991. Shock Cassiat, G., Pourcelot, P., Tavernier, L., Geiger,
University Press, Cambridge. and vibration during the hoof impact on D., Denoix, J.M., Degueurce, C., 2004.
Andersson, L.S., Larhammar, M., Memic, F., different track surfaces. Equine Exerc. Influence of individual competition level
Wootz, H., Schwochow, D., Rubin, C.-J., Physiol. 3, 97–106. on back kinematics of horses jumping a
et al., 2012. Mutations in DMRT3 affect Bayer, A., 1973. Bewegungsanalysen an vertical fence. Equine Vet. J. 36, 748–753.
locomotion in horses and spinal circuit trabrennpferden mit hilfe der Cavagna, G.A., Heglund, N.C., Taylor, C.R.,
function in mice. Nature 488, 642–646. ungulographie. Zbl. Vet. Med. A 20 1977. Mechanical work in terrestrial
Anon, 2011. Rules for dressage events, twenty (Suppl.), 209–221. locomotion: two basic mechanisms for
fourth ed. Fédération Equestre Beck, R.J., Andriacchi, T.P., Kuo, K.N., Fermier, minimizing energy expenditure. Am. J.
Internationale, Switzerland. R.W., Galante, J.O., 1981. Changes in the Physiol. 233: R243–R261.
Argue, C.K., 1994. The kinematics of piaffe, gait patterns of growing children. J. Bone Chateau, H., Robin, D., Simonelli, T., Pacquet,
passage and collected trot of dressage Joint Surg. Am 63, 1452–1457. L., Pourcelot, P., Falala, S., et al, 2009.
horses. PhD thesis, University of Biau, S., Barrey, E., 2004. Relationship between Design and validation of a dynamometric
Saskatchewan, Saskatoon. stride characteristics and scores in dressage horseshoe for the measurement of
tests. Pferdeheilkunde 20, 140–144. three-dimensional ground reaction force
Argue, C.K., Clayton, H.M., 1993a. A
on a moving horse. J. Biomech. 42,
preliminary study of transitions between Biau, S., Lemaire, S., Barrey, E., 2002. Analysis
336–340.
the walk and trot in dressage horses. Acta of gait transitions in dressage horses
Anat. 146, 179–182. using wavelet analysis of dorsoventral Cheney, J.A., Shen, C.K., Wheat, J.D., 1973.
acceleration. Pferdeheilkunde 18, Relationship of racetrack surface to
Argue, C.K., Clayton, H.M., 1993b. A study of
343–350. lameness in the Thoroughbred racehorse.
transitions between the trot and canter in
Am. J. Vet. Res. 34, 1285–1289.
dressage horses. J. Equine Vet. Sci. 13, Blood, D.C., Studdert, V.P., 1990. Ballière’s
171–174. comprehensive veterinary dictionary, Clanton, C., Kobluk, C., Robinson, R.A.,
sceond ed. Ballière Tindall, London, Gordon, B., 1991. Monitoring surface
Back, W., Barneveld, A., Bruin, G., Schamhardt,
p. 927. conditions of a thoroughbred racetrack. J.
H.C., Hartman, W., 1994a. Kinematic
Am. Vet. Med. Assoc. 198, 613–620.
detection of superior gait quality in young Bobbert, M.F., Santamaría, S., 2005.
trotting warmbloods. Vet. Quart. 16 Contribution of the forelimbs and hind Clayton, H.M., 1989a. Terminology for the
(Suppl 2), S91–S96. limbs of the horse to mechanical energy description of equine jumping kinematics.
changes in jumping. J. Exp. Biol. 208, J. Equine Vet. Sci. 9, 341–348.
Back, W., Barneveld, H.C., Schamhardt, H.C.,
Bruin, G., Hartman, W., 1994b. 249–260. Clayton, H.M., 1989b. Kinematic analysis of
Longitudinal development of the Bobbert, M.F., Santamaría, S., van Weeren, cutting horses working a mechanical flag.
kinematics of 4-, 10-, 18- and 26-month- P.R., Back, W., Barneveld, A., 2005. Can Am. J. Vet. Res. 50, 1418–1422.
old Dutch Warmblood horses. Equine Vet. jumping of adult show jumping horses be Clayton, H.M., 1993. The extended canter: a
J. 17 (Suppl.), 3–6. predicted on the basis of submaximal free comparison of some kinematic variables in
Back, W., Bogert, A.J., van den Weeren, P.R., jumps at foal age? A longitudinal study. horses trained for dressage and racing. Acta
Bruin, G., Barneveld, A., 1993. Vet. J. 170, 212–221. Anat. 146, 183–187.
Quantification of the locomotion of Bodine-Rees, P., Bone, J.P., 1976. Premotor Clayton, H.M., 1994a. Comparison of the
dutch warmblood foals. Acta Anat. 146, reaction time component of biceps brachii, stride kinematics of the collected, working,
141–147. brachialis and brachioradialis muscles in medium and extended trot in horses.
Back, W., Hartman, W., Schamhardt, H.C., varsity versus non-varsity women. Equine Vet. J. 26, 230–234.
Bruin, G., Barneveld, A., 1995a. Kinematic Biomechanics V-A. University Park Press, Clayton, H.M., 1994b. Comparison of
response to a 70 day training period in Baltimore, pp. 96–101. the collected, working, medium and
trotting Warmbloods. Equine Vet. J. 18 Bogert, A.J. van den Jansen, M.O., Deuel, N.R., extended canters. Equine Vet. J. 17
(Suppl.), 127–135. 1994. Kinematics of the hind limb push-off (Suppl.), 16–19.
Back, W., MacAllister, C.G., van Heel, M.C.V., in elite jumping horses. Equine Vet. J. 23 Clayton, H.M., 1995. Comparison of the stride
Pollmeier, M., Hanson, P.D., 2007. Vertical (Suppl.), 80–86. kinematics of the collected, medium, and
frontlimb ground reaction forces of sound Bramble, D.M., 1984. Locomotor-respiratory extended walks in horses. Am. J. Vet. Res.
and lame warmbloods differ from those in integration in running mammals. Proc. 56, 849–852.
Quarter Horses. J. Equine Vet. Sci. 27, Assoc. Equine Sports Med. 4, 42–53. Clayton, H.M., 1997a. Classification of
123–129. Brown, N.A.T., Kawcak, C.E., McIlwraith, C.W., collected trot, passage and piaffe using
Back, W., Schamhardt, H.C., Hartman, W., Pandy, M.G., 2003. Architectural properties stance phase temporal variables. Equine
Bruin, G., Barneveld, A., 1995b. Predictive of distal forelimb muscles in horses, Equus Vet. J. 2 (Suppl.), 54–57.
value of foal kinematics for adult caballus. J. Morphol. 258, 106–114. Clayton, H.M., 1997b. Effect of added weight
locomotor performance. Res. Vet. Sci. 59, Burns, T.E., Clayton, H.M., 1997. Comparison on landing kinematics of jumping horses.
64–69. of the temporal kinematics of the canter Equine Vet. J. 23 (Suppl.), 50–53.
Barrey, E., Auvinet, B., Courouce, A., 1995. pirouette and collected canter. Equine Vet. Clayton, H.M., Barlow, D.A., 1989. The effect
Gait evaluation of race trotters using an J. 23 (Suppl.), 58–61. of fence height and width on the limb

336
 References

placements of show jumping horses. Deuel, N.R., Lawrence, L.M., 1986. Gallop Fabiani, M., 1973. Proba wczesnej oceny
J. Equine. Vet. Sci. 9, 179–185. velocity and limb contact variables of zdolnosci koni do skokow. II Konie z
Clayton, H.M., Barlow, D.A., 1991. Stride Quarter Horses. J. Equine Vet. Sci. 6, Zakladow Treningowych w Kwidzynie i
characteristics of four Grand Prix jumping 143–147. Bialym Borze. Prace i Materialy
horses. Equine Exerc. Physiol. 3, 51–157. Deuel, N.R., Lawrence, L.M., 1987a. Laterality Zootechniczne 4, 39–54.
Clayton, H.M., Colborne, G.R., Burns, T.E., in the gallop gait of horses. J. Biomech. 20, Farley, C., Taylor, R.C., 1991. A mechanical
1995. Kinematics analysis of successful and 645–649. trigger for the trot-gallop transition in
unsuccessful attempts to clear a water Deuel, N.R., Lawrence, L.M., 1987b. Effects of horses. Science 253, 306–308.
jump. Equine. Vet. J. 18 (Suppl.), 166–169. urging by the rider on equine gallop stride Fredricson, I., Dalin, G., Drevemo, S., Hjertén,
Clayton, H.M., Colborne, G.R., Lanovaz, J., limb contacts. Proc. Equine Nutr. Physiol. G., 1975. A biotechnical approach to the
Burns, T.E., 1996. Linear kinematics of 10, 487–494. geometric design of racetracks. Equine Vet.
water jumping in Olympic show jumpers. Deuel, N.R., Park, J., 1990. The gait patterns of J. 7, 91–96.
Pferdeheilkunde 12, 657–660. Olympic dressage horses. Int. J. Sport Fredricson, I., Hellander, J., Hjertén, G.,
Clayton, H.M., Hodson, E., Lanovaz, J.L., Biomech. 6, 198–226. Drevemo, S., Bjoerne, K., Dalin, G., et al.,
2000. The forelimb in walking horses. 2. Deuel, N.R., Park, J., 1991. Kinematic analysis 1983. Gait-pattern stability in
Net joint moments and powers. Equine of jumping sequences of Olympic show Thoroughbreds galloping on different track
Vet. J. 32, 295–300. jumping horses. Equine Exerc. Physiol. 3, surfaces. Svensk Veterinartidning 35 (Suppl.
Clayton, H.M., Hoyt, D.F., Wickler, S.J., 158–166. 3), 83–88.
Cogger, E.A., Lanovaz, J.L., 2002. Hind Deuel, N.R., Park, J., 1993. Gallop kinematics Galisteo, A.M., Cano, M.R., Morales, J.L., Vivo,
limb net joint energies during swing phase of Olympic three-day event horses. Acta J., Miro, F., 1998. The influence of speed
as a function of trotting velocity. Equine Anat. 146, 168–174. and height at the withers on the kinematics
Vet. J. 34 (Suppl.), 363–367. Drevemo, S., Dalin, G., Fredricson, I., Bjorne, of sound horses at the hand-led trot. Vet.
Clayton, H.M., Lanovaz, J.L., Schamhardt, K., 1980c. Equine locomotion. 3. The Res. Commun. 22, 415–423.
H.C., van Wessum, R., 1999. Rider effects reproductibility of gait in Standardbred Galisteo, A.M., Morales, J.L., Cano, M.R., Miro,
on ground reaction forces and fetlock trotters. Equine Vet. J. 12, 71–73. F., Aguera, E., Vivo, J., 2001. Inter-breed
kinematics at the trot. Equine Vet. J. 30 Drevemo, S., Dalin, G., Fredricson, I., Hjerten, differences in equine forelimb kinematics
(Suppl.), 235–239. G., 1980a. Equine locomotion. 1. The at the walk. J. Vet. Med. Ser. A-Physiol.
Clayton, H.M., Lanovaz, J.L., Schamhardt, analysis of linear and temporal stride Pathol. Clin. Med. 48, 277–285.
H.C., Willemen, M.A., Colborne, G.R., characteristics of trotting Standardbreds. Galloux, P., Barrey, E., 1997. Components of
1998. Net joint moments and joint powers Equine Vet. J. 12, 60–65. the total kinetic moment in jumping
in the equine forelimb during the stance Drevemo, S., Fredricson, I., Dalin, G., Bjorne, horses. Equine Vet. J. 23 (Suppl.), 41–44.
phase of the trot. Equine. Vet. J. 30, K., 1980b. Equine locomotion. 2. The Grant, B.D., 1989. Performance prediction.
384–389. analysis of coordination between limbs of The Equine Athlete 2, 1–2.
Colborne, G.R., Clayton, H.M., Lanovaz, J.L., trotting Standardbreds. Equine Vet. J. 12, Gray, P., 1993. Smythe and Goody’s Horse
1995. Factors that influence vertical velocity 66–70. structure and movement. Allen, London,
during take off over a water jump. Equine Drevemo, S., Fredricson, I., Hjerten, G., pp. 152–202.
Vet. J. 18 (Suppl.), 138–140. McMiken, D., 1987. Early development of Günther, B., 1975. Dimensional analysis and
Corley, J.M., Goodship, A.E., 1994. Treadmill gait asymmetries in trotting Standardbred theory of biological similarity. Physiol. Rev.
training induced changes to some colts. Equine Vet. J. 19, 189–191. 55, 659–699.
kinematic variables measured at the canter Dušek, J., 1974. The relationship between Hellander, J., Fredricson, I., Hjertén, G.,
of Thoroughbred fillies. Equine Vet. J. 17 height and gait in Hanover stallions. Drevemo, S., Dalin, G., 1983. The
(Suppl.), 20–24. Bayerisches Landwirtschaftliches Jahrbuch galloping gait. I. Variable features at
Cothran, E.G., MacCluer, J.W., Weitkamp, L.R., 51, 209–213. different speeds. Svensk Veterinartidning
Bailey, E., 1987. Genetic differentiation Dušek, J., Ehrlein, H.-J., Engelhardt, W.V., 35, 75–82.
associated with gait within American Hornicke, H., 1970. Beziehungen zwischen Herring, L., Thompson, K.N., Jarret, S., 1992.
Standardbred horses. Anim. Genetics 18, tritlange, trittfrequenz und geschwindigkeit Defining normal three-dimensional
285–296. bei pferden. A Tierzucht ZuchtBiol. 87, kinematics of the lower forelimb in the
Crawford, W.H., Leach, D.H., 1984. The effect 177–188. horse. J. Equine Vet. Sci. 12, 172–176.
of racetrack design on gait symmetry of the Dutto, D.J., Hoyt, D.F., Clayton, H.M., Cogger, Hildebrand, M., 1962. Walking, running and
pacer. Can. J. Comp. Med. 48, 374–380. E.A., Wickler, S.J., 2004. Moments and jumping. Int. Zoologist 2, 151–155.
Crossley, A., 1993. Training the young horse: power generated by the horse (Equus Hiraga, A., Yamanobe, A., Katsuyoshi, K.,
the first two years, 10th edn. Stanley Paul & caballus) hind limb during jumpig. J. Exp. 1994. Relationships between stride length,
Co. Ltd, London, Biol. 207, 667–674. stride frequency, step length and velocity at
Dalin, G., 1994. Pattern-recognition in equine Eastlack, M.E., Arvidson, J., Snydermackler, L., the start dash in a racehorse. J. Equine Sci.
locomotion. Equine Vet. J. 26, 173–174. Danoff, J.V., Mcgarvey, C.L., 1991. Interrater 5, 127–130.
Dalin, G., Drevemo, S., Fredricson, I., Jonsson, reliability of videotaped observational Hodson, E., Clayton, H.M., Lanovaz, J.L.,
K., Nilsson, G., 1973. Ergonomic aspects of gait-analysis assessments. Phys. Ther. 71, 1999. Temporal analysis of walk
locomotor asymmetry in Standardbred 465–472. movements in the Grand Prix dressage test
horses trotting through turns. Acta Vet. Ellersieck, M.R., Lock, W.E., Vogt, G.W., at the 1996 Olympic Games. Appl. Anim.
Scand. 44 (Suppl.), 111–139. Aipperspach, R., 1985. Genetic evaluation Behav. Sci. 62, 89–97.
Dalin, G., Magnusson, L.-E., Thafvelin, B.C., of cutting scores in horses. J. Equine Vet. Hole, S.M., Clayton, H.M., Lanovaz, J.L., 2002.
1985. Retrospective study of hindquarter Sci. 5, 287–289. A note on the linear and temporal stride
asymmetry in Standardbred trotters and its Estberg, L., Stover, S., Gardner, I.A., Drake, kinematics of Olympic show jumping
correlation with performance. Equine. Vet. C.M., Johnson, B., Case, J.T., et al., 1995. horses between two fences. Appl. Anim.
J. 17, 292–296. Several racing-career intensity characteristics Behav. Sci. 75, 317–323.
Deuel, N.R., 1995. Dressage canter kinematics are associated with racing-related fatal Holmström, M., 2001. The effects of
and performances in an Olympic three-day skeletal injury in California Thoroughbred conformation. In: Back, W., Clayton, H.M.
event. Proc. Eur. Assoc. Anim. Prod. 46, racehorses. Proc. Am. Assoc. Equine (Eds.), Equine Locomotion. W.B. Saunders,
341–344. Practnr. 41, 82–83. London, p. 281.

337
 14 Performance in equestrian sports

Holmström, M., Drevemo, S., 1997. Effects of gait diagrams for the study of equine hind Meershoek, L.S., Schamhardt, H.C., Roesptorff,
trot quality and collection on the angular limb kinematics. Equine Vet. J. 21, 48–58. L., Johnston, C., 2001b. Forelimb tendon
velocity in the hind limbs of riding horses. Koenen, E.P.C., van Veldhuizenb, A.E., loading during jump landings and the
Equine Vet. J. 2 (Suppl.), 62–65. Brascamp, E.W., 1995. Genetic parameters influence of fence height. Equine Vet. J. 33
Holmström, M., Fredricson, I., Drevemo, S., of linear scored conformation traits and (Suppl.), 6–10.
1994a. Biokinematic analysis of the their relation to dressage and show- Merkens, H.W., Schamhardt, H.C., van Osch,
Swedish Warmblood riding horse at trot. jumping performance in the Dutch G.J.V.M., Hartman, W., 1993. Ground
Equine Vet. J. 26, 235–240. Warmblood Riding Horse population. reaction force patterns of Dutch
Holmström, M., Fredricson, I., Drevemo, S., Livest. Prod. Sci. 43, 85–94. Warmbloods at the canter. Am. J. Vet. Res.
1994b. Biokinematic differences between Krebs, D.E., Edelstein, J.E., Fishman, S., 1985. 54, 670–674.
riding horses judged as good and poor Reliability of observational kinematic gait Merkens, H.W., Schamhardt, H.C., van Osch,
at the trot. Equine Vet. J. 17 (Suppl.), analysis. Phys. Ther. 65, 1027–1033. G.J.V.M., van den Bogert, A.J., 1991.
51–56. Kroll, W., Clarkson, P.M., 1977. Fractionated Ground reaction force analysis of Dutch
Holmström, M., Fredricson, I., Drevemo, S., reflex time, resisted and unresisted Warmblood horses at canter and jumping.
1995. Biokinematic effects of collection on fractionated reflex time under normal and Equine Exerc. Physiol. 3, 128–135.
the trotting gaits in the elite dressage horse. fatigued conditions. In: Landers, D.M., Minetti, A.E., Ardigò, L.P., Reinach, E., Saibene,
Equine Vet. J. 27, 281–287. Christiana, R.W. (Eds.), Psychology F., 1999. The relationship between
Holmström, M., Magnusson, L.E., Philipsson, of Motor Behavior and Sport. Human mechanical work and energy expenditure
J., 1990. Variation in conformation of Kinetic Publishers Inc., Champaign, of locomotion in horses. J. Exp. Biol. 202,
Swedish Warmblood horses and pp. 106–129. 2329–2338.
conformational characteristics of elite sport Langlois, B., Froidevaux, J., Lamarche, L., et al., Molina, A., Valera, M., Galisteo, A.M., Vivo, J.,
horses. Equine Vet. J. 22, 186–193. 1978. Analyse des liasons entre la Gomez, M.D., Rodero, A., Aguera, E., 2008.
Holmström, M., Philipsson, J., 1993. morphologie et l’aptitude au galop au trot Genetic parameters of biokinematic
Relationship between conformation, et au saut d’obstacles chez le cheval. Ann. variables at walk in the Spanish Purebred
performance and health in 4-year old Genet. Sel. Anim. 10, 443–474. (Andalusian) horse using experimental
Swedish Warmblood Riding Horses. Lauk, H.D., Auer, J., von Plocki, K.A., 1991. treadmill records. Livest. Sci. 116, 137–145.
Livestock Prod. Sci. 33, 293–312. Zum Problem “Barren” Uberblick, Morales, J.L. Manchado, M., Cano, M.R., Miro,
Holzreiter, S.H., Köhle, M.E., 1993. Assessment biomechanische Berechnungen und F., Galisteo, A.M., 1998a. Temporal and
of gait patterns using neural networks. J. Verhaltungsbeobachtungen. linear kinematics in elite and riding horses
Biomech. 26, 645–651. Pferdeheilkunde 7, 225–235. at the trot. J. Equine Vet. Sci. 18, 835–839.
Hoyt, D.F., Taylor, C.R., 1981. Gait and Leach, D.H., Cymbaluk, N., 1986. Morales, J.L., Manchado, M., Vivo, J., Galisteo,
energetics of locomotion in horses. Nature Relationships between stride length, stride A.M., Aguera, E., Miro, F., 1998b. Angular
292, 239–240. frequency, velocity and morphometrics of kinematic patterns of limbs in elite and
Huizinga, H.A., 1991. Genetic studies on the foals. Am. J. Vet. Res. 47, 2090–2097. riding horses at the trot. Equine Vet. J. 30,
performance of Dutch Warmblood riding Leach, D.H., Ormrod, K., 1984. The technique 528–533.
horses. Thesis, Utrecht University, Utrecht, of jumping a steeplechase fence by Moyer, W., Fisher, J.R.S., 1991. Bucked shins:
The Netherlands. competing event-horses. Appl. Anim. effects of differing track surfaces and
Ischii, K., Amano, K., Sakuraoka, H., 1989. Behav. Sci. 12, 15–24. proposed training regimens. Proc. Am.
Kinematic analysis of horse gaits. Bull. Leach, D.H., Sprigings, E., 1979. Gait fatigue Assoc. Equine. Practnr. 37, 541–547.
Equine Res. Inst. 26, 1–9. in the racing Thoroughbred. J. Equine Med. Muñoz, A., Santisteban, R., Rubio, M.D., Vivo,
Jelen, B., 1976. Faktyczna I umowna wyokosc Surg. 3, 436–443. R., Aguera, E.I., Escribano, B.M., et al.,
skokow koni przez przeszkody I niektore Leach, D.H., Sprigings, E.J., Laverty, W.H., 1997. Training as an influential factor on
warunkujace ja czynniki. Zootechniczna 1987. Multivariate statistical analysis of the locomotor pattern in Andalusian
97, 79–91. stride-timing measurements of nonfatigued horses. J. Vet. Med. Ser. A-Physiol. Pathol.
Johnston, C., Roethlisberger-Holm, K., racing Thoroughbreds. Am. J. Vet. Res. 48, Clin. Med. 44, 473–480.
Erichsen, C., Eksell, P., Drevemo, S., 2004. 880–888. Nicodemus, M.C., Booker, J.E., 2007.
Kinematic evaluation of the back in fully Lewczuk, D., Pfeffer, M., 1998. Investigation Two-dimensional kinematics of the jog and
functioning riding horses. Equine Vet. J. 36, on the horses’ stride length during lope of the stock breed western pleasure
495–498. endurance race competition using video horse. Equine Compar. Exerc. Physiol. 4,
Kai, M., Kubo, K., 1993. Respiration pause and image analysis. In: Conference on Equine 59–70.
gait after the start dash in a horse. Bull. Sports Medicine and Science. Wageningen Nicodemus, M.C., Clayton, H.M., 2001.
Equine Res. Inst. 30, 26–29. Pers, The Netherlands, pp. 249–250. Temporal variables of the four-beat
Kingsbury, H.B., Quddus, M.A., Rooney, J.R., Magnusson, L.-E., 1985. Studies on the stepping jog and lope. Proc. Equine Nutr.
Geary, J.E., 1978. Laboratory system for conformation and related traits of Physiol. Symp. 17, 247–252.
production of flexion rates and forces in Standardbred trotters in Sweden. PhD Normand, M.C., Lagasse, P.P., Rouillard, C.A.,
forelimb of horse. Am. J. Vet. Res. 39, thesis, Skara, Swedish University of Tremblay, L.E., 1982. Modifications
365–369. Agricultural Sciences. occurring in motor programs during
Knezevic, P.F., Kastner, J., Girtler, D., Martinez-del Campo, L.J., Kobluk, C.N., Greer, learning of a complex task in man. Brain
Holzreiter, S., 1987. High-frequency N., Trent, A.M., Stoner, L.J., Wickstrom, L., Res. 241, 87–93.
cinematographic measurement of the horse et al., 1991. The use of high-speed Oliver, R., Langrish, B., 1992. A photographic
during capriole. Deutsche Tierarztliche videography to generate angle-time and guide to conformation, second ed. Allen,
Wochenschrift 94, 141 angle-angle diagrams for the study of London, pp. 114–124.
Knopfhart, A., 1966. Het rijpaard: beoordeling equine locomotion. Vet. Comp. Orthop. Peham, C., Licka, T. Kapaun M., Scheidl, M.,
en keuze. Veen ‘s Uitgeversmaatschappij, Traum. 4, 120–131. 2003. A new method to quantify harmony
Amsterdam, pp. 70–77. Meershoek, L.S., Roepstorff, L., Schamhardt, of the horse–rider system in dressage.
Kobluk, C.N., Schnurr, D., Horney, F.D., H.C., Johnston, C., Bobbert, M.F., 2001a. Sports Engin. 4, 95–101.
Sumner-Smith, G., Willoughby, R.A., Joint moments in the distal forelimbs of Persson, S.G.B., Essen-Gustavsson, B.,
DeKleer, V., 1989. Use of high-speed jumping horses during landing. Equine Vet. Lindholm, A., 1991. Energy profile and the
cinematography and computer generated J. 33, 410–415. locomotor pattern of trotting on an

338
 References

inclined treadmill. Equine Exerc. Physiol. 3, variables. Equine Vet. J. 18 (Suppl.), Smythe, R.H., 1957. What makes a good
231–238. 191–194. horse: its structure and performance.
Peterson, M., McIlwraith, C.W., Reiser, R.F., Rooney, J.R., 1969. Lameness of the forelimb. Country Life Ltd, London, pp. 86–102.
2008. System development for in-situ In: Biomechanics of Lameness in Horses. Spooner, G., 1977. The Handbook of
characterization of horse racing track Williams and Wilkins, Baltimore, 114–196. Showing. Allen, London, pp. 50–58.
surfaces. Biosyst. Eng. 101, 260–269. Santamaría, S., Back, W., van Weeren, P.R., Stover, S.M., Johnson, B.J., Daft, B.M., Read,
Peterson, M.L., McIlwraith, C.W., 2008. Effect Knaap, J., Barneveld, A.(2002) Jumping D.H., Anderson, M., Barr, B.C., et al., 1992.
of track maintenance on mechanical characteristics of naive foals: Lead changes An association between complete and
properties of a dirt racetrack: a preliminary and description of temporal and linear incomplete stress fractures of the humerus
study. Equine Vet. J. 40, 602–605. parameters. Equine Vet. J. 34 (Suppl.), in racehorses. Equine Vet. J. 24, 260–263.
Peterson, M.L., Reiser, R.F., Kuo, P.H., Radford, 302–307. Stubbs, N.C., Hodges, P.W., Jeffcott, L.B.,
D.W., McIlwraith, C.W., 2010. Effect of Santamaría, S., Bobbert, M., Back, W., Cowin, G., Hodgson, D.R., McGowan,
temperature on race times on a synthetic Barneveld, A., van Weeren, P.R., 2004. C.M., 2006. Functional anatomy of the
surface. Equine Vet. J. 42, 351–357. Variation in free jumping technique within caudal thoracolumbar and lumbosacral
Podhajsky, A., 1981. The Complete Training of and among horses with little experience in spine in the horse. Equine Vet. J. 36
Horse and Rider, tenth ed. Wilshire Book show jumping. Am. J. Vet. Res. 65, (Suppl), 393–399.
Company, London, p. 34. 938–944. Swanstrom, M.D., Zarucco, L., Hubbard, M.,
Powers, P., 2005. Linear kinematics at take-off Santamaría, S., Bobbert, M.F., Back, W., Stover, S.M., Hawkins, D.A., 2005a.
in horses jumping the wall in an Barneveld, A., van Weeren, P.R., 2005. Musculoskeletal modeling and dynamic
international Puissance competition. Sports Effect of early training on the jumping simulation of the thoroughbred equine
Biomech. 4, 149–162. technique of horses. Am. J. Vet. Res. 66, forelimb during stance phase of the gallop.
Powers, P., Harrison, A., 2002. Effects of the 418–424. J. Biomech. 127, 318–328.
rider on the linear kinematics of jumping Santamaría S., Bobbert, M.F., Back, W., Swanstrom, M.D., Zarucco, L., Stover, S.M.,
horses. Sports Biomech. 1 (2), 135–146. Barneveld, A., van Weeren, P.R., 2006. Can Hubbard, M., Hawkins, D.A., Driessen, B.,
Pratt, G.W., 1983. Remarks on gait analysis. early training of show jumpers bias et al., 2005b. Passive and active mechanical
Equine Exerc. Physiol. 2, 245–262. outcome of selection events? Livest. Sci. properties of the superficial and deep
Pratt, G.W., O’Connor Jr, J.T., 1976. Force 102, 163–170. digital flexor muscles in the forelimbs of
plate studies of equine biomechanics. Am. Schamhardt, H.C., Merkens, H.W., van Osch, anesthetized thoroughbred horses. J.
J. Vet. Res. 37, 1251–1255. G.J.V.M., 1991. Ground reaction force Biomech. 38, 579–586.
Preuschoft, H., 1989. The external forces and analysis of horse ridden at walk and trot. Tans, E., Nauwelaerts, S., Clayton, H.M. 2009.
internal stresses in the feet of dressage and Equine Exerc. Physiol. 3, 120–127. Dressage training affects temporal variables
jumping horses. Z. Säugetierkunde 54, Schamhardt, H.C., Merkens, H.W., Vogel, V., in transitions between trot and halt. Comp.
172–190. Willekens, C., 1993. External loads on the Exerc. Physiol. 6, 89–97.
Quddus, M.A., Kingsbury, H.B., Rooney, J.R., limbs of jumping horses at take-off and Terada, K., 2000. Comparison of head
1978. A force and motion study of the landing. Am. J. Vet. Res. 54, 675–680. movement and EMG activity of muscles
foreleg of a Standardbred trotter. J. Equine Schils, S.J., Greer, N.L., Stoner, L.J., Kobluk, between advanced and novice horseback
Med. Surg. 2, 233–242. C.N., 1993. Kinematic analysis of the riders at different gaits. J. Equine Sci. 11,
Ratzlaff, M.H., Grant, B.D., Frame, J.M., Hyde, equestrian – walk, posting trot and sitting 83–90.
M.L., 1987. Locomotor forces of galloping trot. Hum. Movement Sci. 12, 693–712. Terada, K., Mullineaux, D.R., Clayton, H.M.,
horses. Equine Exerc. Physiol. 2, 574–586. Schwarz, H.G., 1971. Bewegunganalyse beim Kato, K., 2006. Stabilization of wrist
Ratzlaff, M.H., Shindell, R.M., White, K.K., pferd und untersuchung der position during horseback riding. Equine
1985. The interrelationships of stride trainingseinwirkung auf den Comp. Exerc. Physiol. 3, 179–184.
length and stride times to velocities of bewegungsablauf im schritt und trab. PhD Terada, K., Mullineaux, D.R., Lanovaz, J.L.,
galloping horses. J. Equine Vet. Sci. 5, thesis, University of Giessen, Germany. Kato, K., Clayton, H.M., 2004.
279–283. Seder, J.A., Vickery, C.E., 2003. Double and Electromyographic analysis of the rider’s
Ratzlaff, M.H., Wilson, P.D., Hutton, D.V., triple fully airborne phases in the gaits of muscles at trot. Equine Comp. Exerc.
Slinker, B.K., 2005. Relationships between racing speed Thoroughbreds. J. Equine Vet. Physiol. 1, 193–198.
hoof-acceleration patterns of galloping Sci. 23 (Suppl. 1), S73–S81. Ueda, Y., Yoshida, K., Oikawa, M., 1993.
horses and dynamic properties of the track. Sellett, L.C., Albert, W.W., Groppel, J.L., 1981. Analyses of race accident conditions
Am. J. Vet. Res. 66, 589–595. Forelimb kinematics of the Standardbred through use of patrol video. J. Equine Vet.
Reiser, R.F., Peterson, M.L. McIlwraith, C.W., pacing gait. Proc. Equine Nutr. Physiol. Sci. 13, 707–710.
Woodward, B., 2001. Simulated effects on 210–215. van Veldhuizen, A.E., 1991. Nogmaals uitleg
racetrack material properties on the vertical Setterbo, J.J., Garcia, T.C., Campbell, I.P., over het lineair scoren. In de Strengen 58,
loading of the equine forelimb, Sports Reese, J.L., Morgan, J.M., Kim, S.Y., 34–35.
Engineering 3 (1), 1–11. et al., 2009. Hoof accelerations and van Weeren, P.R., 2007. La conformation et
Rivero, J.L., Serrano, A.L., Quiroz-Rothe, E., ground reaction forces of Thoroughbred l’exercice du poulain déterminent–ils ses
Aguilera-Tejero, E., 2001. Coordinated racehorses measured on dirt, synthetic, performances futures? Prat Vét. Equine 39
changes of kinematics and muscle fibre and turf track surfaces. Am. J. Vet. Res. 70, (numéro spécial), 119–126.
properties with prolonged endurance 1220–1229. van Weeren, P.R., Brama, P.A.J., Barneveld, A.,
training. Equine Vet. J. 33 (Suppl.), Sloet van Oldruitenborgh–Oosterbaan, M.M., 2000. Exercise at young age may influence
104–108. Barneveld, A., 1995. Comparison of the the final quality of the equine
Rivero, J.-LL., Clayton, H.M., 1996. The workload of Dutch warmblood horses musculoskeletal system. Proc. Am. Assoc.
potential role of the muscle in kinematic ridden normally and on a treadmill. Vet. Equine Pract. 46, 29–35.
characteristics. Pferdeheilkunde 12, Rec. 137, 136–139. van Weeren, P.R., van den Bogert, A.J., Back,
635–640. Sloet van Oldruitenborgh–Oosterbaan, W., Bruin, G., Barneveld, A., 1993.
Roneus, N., Essen-Gustavsson, B., Johnston, M.M., Clayton, H. M., 1999. Advantages Kinematics of the Standardbred trotter
C., Persson, S., 1995. Lactate response to and disadvantages of track vs. treadmill measured at 6, 7, 8 and 9 m/s on a
maximal exercise on the track: relation to tests. Equine Vet. J. 30 (Suppl.), treadmill, before and after 5 months of
muscle characteristics and kinematic 645–647 prerace training. Acta Anat. 2–3, 81–204.

339
 14 Performance in equestrian sports

van Weeren, P.R., van den Bogert, A.J., longitudinal approach. Am. J. Phys. parameters and hind limb length in horses
Barneveld, A., Hartman, W., Kersjes, A.W., Anthropol. 81, 429–439. fatigued on a treadmill and at an
1990. The role of the reciprocal apparatus Vilensky, J.A., Gankiewicz, E., Townsend, D.W., endurance ride. Equine Vet. J. 36 (Suppl.),
in the hind limb of the horse investigated 1990b. Effect of growth and speed on hind 60–64.
by a modified CODA–3 opto–electronic limb joint angular displacement patterns in Wilson, B.D., Neal, R.J., Howard, A.,
kinematic analysis system. Equine Vet. J. 9 vervet monkeys (Cercopithecus aethiops). Groenendyk, S., 1988a. The gait of pacers
(Suppl.), 95–100. Am. J. Phys. Anthrop. 81, 441–449. 1: kinematics of the racing stride. Equine
Viitsalo, J.T., Komi, P.V., 1981. Wallin, L., Strandberg, E., Philipsson, J., 2003. Vet. J. 20, 341–346.
Interrelationships between Genetic correlations between field test Wilson, B.D., Neal, R.J., Howard, A.,
electromyographic, mechanical, muscle results of Swedish Warmblood Riding Groenendyk, S., 1988b. The gait of pacers
structure and reflex time measurements. Horses as 4-year-olds and lifetime 2: factors influencing pacing speed. Equine
Acta Physiol. Scand. 111, 97–103. performance results in dressage and show Vet. J. 20, 347–351.
Vilensky, J.A., Gankiewicz, E., 1986. Effects of jumping. Livestock Prod. Sci. 82, 61–71. Yamanobe, A., Hiraga, A., Kubo, K.,
size on vervet (Cercopithecus–aethiops) Weishaupt, M.A., Bystrom, A., von Peinen, K., 1992. Relationships between stride
gait parameters– A preliminary analysis. Wiestner, T., Meyer, H., Waldern, N., et al., frequency, stride length, step length
Folia Primatol. 46, 104–117. 2009. Kinetics and kinematics of the and velocity with asymmetric gaits in
Vilensky, J.A., Gankiewicz, E., Townsend, D.W., passage. Equine Vet. J. 41, 263–267. the Thoroughbred horse. Jap. J. Equine
1990a. Effects of size on vervet Wickler, S.J., Greene, H.M., Egan, K., Astudillo, Sci. 3, 143–148.
(Cercopithecus aethiops) gait parameters: a A., Dutto, D.J., Hoyt, D.F., 2006. Stride

340
 C H A P T ER 15 

Horse–rider interaction
Agneta Egenvall, Anna Byström, Michael Weishaupt, Lars Roepstorff

Introduction standards. In order to construct well-fitting saddles it is essential to

know how the rider, saddle and horse interact, and to explain the
source of saddle pressure peaks.
In many countries the role of the horse in society has changed
dramatically during the last half century from being used in agri-
culture and warfare to a companion in sports and leisure. Today The electronic saddle mat
considerably fewer people grow up with horses as a daily part of
their lives while the general interest in horses grows. The lack of The most obvious way to study the seat in an objective and quan-
natural training and ‘education’ of working with horses on a daily titative way is to use a saddle pressure mat. There are today several
basis has led to an increased demand for educational programs in brands on the market that can be used for measuring pressure
equine care and management from very basic principles to between the horse’s back and saddle dynamically. Though this
university-level education. equipment has been available for many years, the first peer-reviewed
More formalized education increases the demands for knowledge paper was published in 1994 (Harman, 1994). One reason that
and a natural consequence has been an increase in research. Studies relatively few papers have been published might be the difficulties
of the horse–rider interaction have included a biomechanical or in obtaining good quality, quantitative data using this technology.
ethological approach, together with research on health and perfor- There are several problems including relatively poor performance
mance of both horse and rider. Other studies have evaluated train- in terms of accuracy and drift. The published papers mainly deal
ing techniques or rehabilitation of rider or horse. In this chapter we with maxima, local peaks or mean forces, which are a very limited
focus on the biomechanical approaches to horse–rider performance part of the total information registered by the pressure sensitive mats
and, to a limited extent, on horse health. which usually consist of between 256 and 2000 sensors, each sam-
The biomechanical methods used so far to assess horse–rider pling at up to several hundred samples per second. The variability
interaction comprise kinetic measurements with force plates or a between strides makes it useful for the interpretation to calculate a
force-measuring treadmill; kinematic measurements using motion mean stride for each experimental situation, which has not been
analysis or accelerometers; pressure measurements between the generally done so far. Mat slip can be an issue and the correct posi-
saddle and horse’s back using electronic pressure mats; rein tension tion has to be checked before and after measurement. Placement of
studies; and electromyography. This chapter will begin by describ- the mat relative to both the horse’s back and the saddle needs to be
ing studies of saddles or other research using saddle mats, then defined in a standardized way. In most of the studies to date the mat
studies of reins and bits. This will lead to a description of the rider’s data has been used and presented in a qualitative or semi-quantitative
seat and other rider characteristics, including various rider-body way. The mats may be difficult to calibrate and the software that
factors related to skill, and rider weight. The next part of the chapter handles the data needs further development. There is a need for
devotes itself to concepts used in riding methodology and specifi- standardized methods of presenting, analyzing and comparing
cally to the central concept of collection and its definitions. saddle mat measurements (see also Chapter 2).
The following paragraphs describe studies in which the saddle
mat has been used scientifically to study the interaction between
Saddle the rider, saddle and horse’s back. Saddle pressure measurements
have also been used to confirm the association between clinical
The saddle is an important part of the rider’s equipment, making it evidence of back pain and pressure peaks acting on the horse’s back
comfortable to sit on the horse as well as distributing the weight of (Werner et al., 2002; Nyikos et al., 2005; von Peinen et al., 2010)
the rider over a larger area of the horse’s back, thus making it more but these studies will not be further elucidated here.
comfortable also for the horse. The idea of distributing weight over
a larger area thereby lowering pressure (force over surface area) with
the purpose of avoiding pressure injuries to the horse’s back is a
Saddle pressure in the standing horse
logical and obvious solution at first sight. However, under closer De Cocq et al. (2006) calculated a correlation coefficient between
scrutiny there are some potential problems in this reasoning. Is it total measured force and the weight of 28 different riders, using
always better to have lower pressure, especially if it is more static? paired measurements with defined air-pressure inside the panels of
Are all tissues equally susceptible to pressure? Does pressure on a the saddle in standing horses. Total force correlated well with the
certain part of the back, for example behind the 16th thoracic ver- weight of the rider and the results within trials agreed well, but for
tebra, limit movement of the back? Many of the questions debated trials in which the saddle mat was removed and replaced the agree-
by saddle manufacturers and saddle fitters demand objective ments varied from poor to excellent. The results indicated that

341
 15 Horse–rider interaction

Fig 15.1  Graph showing total force during a typical trial for mounting from the ground. Still photographs and the corresponding maps of pressure
distribution under the saddle are shown for the points marked 1–5 on the graph. This sequence shows the most common pattern with a single force peak as
the right leg swings upwards. In the pressure maps, the cranial part of the mat is up and the left side is to the left.
Reprinted from Geutjens, C.A., Clayton, H.M., Kaiser, L.J., 2008. Forces and pressures beneath the saddle during mounting from the ground and from a raised mounting platform. The Veterinary
Journal 175 (3), 332–337, with permission from Elsevier.

highly standardized conditions were necessary to obtain useful and is not possible, it is recommended to use a mounting block and
valid results and that further technical development is necessary. mount alternately from the left and right sides.
Better results, relative to agreement between different measurement
conditions basically measuring the same things, was found in
another study using a different type of saddle mat (de Cocq et al.,
The effect of the saddle on motion of the back
2009a). Different rider positions were evaluated and the mat regis- With the use of pressure sensitive saddle mats, the influence of the
tered increased force beneath the saddle in the direction towards saddle on the horse’s back can be studied while moving at different
which the rider was leaning (de Cocq et al., 2009a). Therefore, the gaits. However, there is only a weak association between the saddle
mat could discriminate between rider positions and asymmetrical pressures acting on the back in a standing horse and those exerted
weighting of the saddle. during movement (Jeffcott et al., 1999). Total saddle force curves
A study of forces on the horse’s back during mounting showed that show a characteristic, gait dependent pattern (Pullin, 1996; Jeffcott
the highest forces coincided with the rider’s free leg swinging up and et al., 1999; Fruehwirth et al., 2004; von Peinen et al., 2009).
over the horse’s back. When the rider mounted from the conventional Fruehwirth et al. (2004) registered force data from quarters of the
(left) side forces were highest on the right side of the horse’s back saddle mat in horses ridden at walk, trot and right lead canter (12
adjacent to the withers. High forces were also found lower on the left horses and 12 riders) and found that force patterns were reproduc-
shoulder. Heavier riders yielded higher forces (Geutjens et al., 2008). ible from stride to stride in each gait and force magnitude increased
The use of a mounting block produced a smoother force application with speed.
and lower peak force (Fig. 15.1). In light of these findings it is recom- On the contrary, de Cocq et al. (2004) evaluated nine Dutch
mended that, if possible, the rider should mount without using the Warmbloods walking, trotting and cantering on a treadmill in an
stirrup to avoid the asymmetrical loading of the horse’s back. If this unloaded condition, wearing a surcingle, a saddle or a saddle plus

342
 Bits and reins

75 kg weight. The most marked changes occurred with the addition Von Peinen et al. (2010) demonstrated that the mean and
of extra weight; the horse’s back moved through the same range of maximal saddle pressure in walk, trot and canter were larger in
motion but with more extension and less flexion. Fewer changes 23 horses with clinical evidence of ill-fitting saddles compared to
were seen when comparing an unweighted saddle with the unsad- 16 control horses. Pressure was measured at the areas of dry spots
dled situation. Forelimb retraction increased in walk and trot when (dry non-sweating areas that sustained local ischemia due to high
the horse had a weighted saddle compared with a saddle, a surcingle saddle pressure) or acute soreness in the withers area in the horses
or unloaded. Increases in both forelimb retraction and hind limb with the problems and compared to the same areas in the control
protraction are mechanisms that counteract back-extension, so the horses. The results indicated that mean pressure was more repeat-
increase in forelimb retraction is likely to counteract over-extension able than maximal pressure and was, therefore, a more reliable
of the back (Jeffcott, 1979). variable for assessing saddle fit. Similar to the decubitus ulcer
probem in people, critical values could be established for different
Saddle types levels of saddle sores in the withers area. Such critical values would
help fitting saddles in the future.
Early studies using a pressure mat evaluated various saddles quali- In her pioneering work Harman (1994) looked at saddle fit of
tatively (Harman, 1994). Werner et al. (2002) studied two types of equine patients and evaluated the effects of 14 different pads on one
dressage saddles in 25 horses ridden by the same rider in various rider–horse combination, concluding that many pads make the pres-
gaits. They found differences between saddles and horses and sure distribution worse compared to using a simple cotton pad.
smaller differences with gait. It was suggested that mean pressure Pullin et al. (1996) evaluated two saddles, two horses and 14 pad-
>15 kPa and maximal pressure >35 kPa at sitting trot correlated liner combinations. Both static and dynamic measurements and
with back pain. Further, patient horses (n = 25) with various degrees results were evaluated qualitatively. Repeatable differences were seen
of and reasons for back pain were evaluated with an interval of at between the tested pads leading the authors to recommend that
least 12 weeks, during which time saddlery problems were corrected dynamic evaluations were more useful, and that static scans without
based on saddle mat data (Mönkemöller, 2005). Qualitatively, riders should not be used for optimal saddle pad evaluation.
improvement was judged to be significant based on distribution A study of how the fit of a saddle with a tree that was too wide
and magnitude of pressure. was affected by using a pad made of gel, leather, foam or reindeer
In a study comparing an English saddle with a side saddle, horses fur was undertaken in one rider who rode 18 horses in trot and
(n = 13) without clinical signs of back-pain were evaluated at walk, walk. All pad-saddle combinations were compared to the saddle
trot and canter using a saddle mat and kinematic measurements of without pad. In many pad-saddle-horse combinations, maximal
back movement both in front of the withers and behind the saddle overall force increased when using a pad and the pressure differ-
(4th lumbar vertebra, L4). The results showed asymmetrical loading ences between adjacent areas of the saddle often did not decrease.
under the side saddle with the center of pressure located to the right The potential benefits of using a pad need careful consideration in
of midline and slightly more caudally than the center of pressure horses with back problems (Kotschwar et al., 2010a,b).
under the English saddle in all gaits. The sidesaddle also influenced
back movement. The authors suggested that the two saddle types Panels and girths
could be used to change the load distribution on the back of horses
with potential back pain and as a training variation (Winkelmayr De Cocq et al. (2006) documented that the force beneath the rear
et al., 2006). part of the panels could be measured more reliably than the force
In racehorses (n = 8), no difference was found between saddles beneath the more forward part of the panels because of the inclina-
with a normal tree, a flexible tree or treeless relative to the stipulated tion of the horse’s back. The sensors measure force applied perpen-
critical pressure on the lower thoracic back at canter and gallop dicular to their surface but shear forces give cross-talk to the
(Latif et al., 2010), which was the main concern. However, loading measurements, i.e. the forces will not be correctly measured. This
and peak pressures were relatively high at the withers during canter- means that a larger part of the force is captured in areas of the back
ing and galloping but shifted toward the mid and rear regions while that have a more horizontal inclination. As a consequence of the
trotting. The authors concluded that the type of tree had no influ- shape of the back on either side of the withers, forces are likely to
ence on the pressure profile at the rear of the saddle and that the be underestimated by a larger amount in this area.
high peak pressures at the rear of the saddle in trot with all saddles Another study examined saddle pressure patterns during riding
may limit back motion. The pressure distribution in trot and gallop in trot and canter, comparing traditional and v-system girth strap
was mainly due to jockey position. The high pressure under the rear placement, and wool and synthetic foam panel flocking material
of the treeless saddle, as well as the saddle with the flexible tree, was (six horses, three riders) (Byström et al., 2010a). Controlling for
directly beneath the rider. With respect to the treed saddle, it was speed, stride maximum pressures below the hind part of the saddle
due to an improperly-fitting tree or cushioning. increased by 7–12% and the area below the saddle with a stride
mean pressure >11 kPa increased by 114 cm2 at trot and 127 cm2 at
canter with foam-filled panels compared with wool-stuffed panels.
Saddle fit and saddle pads With the v-system girthing the area with pressure >11 kPa increased
The kinematics of horses (n = 21) ridden by one rider on a treadmill by 53 cm2 and 38 cm2. It was concluded that both flocking material
at trot with either a well-fitting or ill-fitting saddle (Peham et al., and girthing are relevant considerations in saddle fitting. In this
2004), were documented using three saddles, and two kinematic study, wool performed better than foam as panel flocking material
markers (L4 and right fore hoof). Variability of the motion cycles and traditional placement of the girth was better or at least as good
was calculated as well as the derivates (velocities and accelerations) compared to the v-system.
in the x, y, and z directions. With a rider the variability of velocity
and acceleration in the forward direction and acceleration in the
transverse direction decreased with a well-fitting saddle compared Bits and reins
to an ill-fitting saddle. Meschan (2007) tested differently fitting
saddles with a pressure mat and concluded that the load under
poorly fitting saddles is distributed over a smaller area than under
Bits
properly fitting saddles, leading to potentially harmful pressure Clayton and Lee (1984) described the use of a fluoroscopic technique
peaks. Typical pressure profiles for saddles with correct, wide and to evaluate bit position, the interactions of the bit with the intra-oral
narrow trees were shown. tissues and the effects of rein tension on bit position. This technique

343
 15 Horse–rider interaction

was applied in a further study of a variety of bits that varied in shape

and mechanical action (Clayton, 1985). It was shown that bilateral
rein tension caused the bits to indent the tongue. Horses accepted
the bit quietly with minimal intra-oral motion when the bit was
adjusted so that there were one or two small wrinkles at the commis-
sures of the lips. When the cheek pieces were adjusted to allow the
bit to lie lower on the tongue or when the bit was considerably wider
than the distance between the oral commissures the horse could
move the bit around more freely within the mouth and tended to
show more movements of the tongue that could be interpreted as
resistances (Clayton & Lee, 1984). With some bits it was possible to
restrict the movements of the mouthpiece, for example, using keepers
on a cheek snaffle. For some bits the severity of their action depended
on their orientation within the oral cavity, such as the reversible Dr.
Bristol bit. In the equestrian literature, when using the double bridle,
the bridoon (or bradoon) and the curb are said to have quite different
actions, the bradoon lifts and extends the neck while the curb is
described as having a leverage effect and powerful action on the bars
(Karl, 2008). However, in this study the bradoon of the double bridle
moved with the curb, suggesting that the effect of the bradoon and
the curb would be somewhat hard to separate within the oral cavity
though the horse’s sensation likely differs between the actions of the
bits (Clayton, 1985).
A further study using lateral-view radiographs (Manfredi et al.,
2005b) described the positions of six bits relative to the horse’s
palate and second premolar teeth with and without 25-N rein
tension which is equivalent to peak tension recorded in horses trot-
ting with side reins adjusted to the resting length of the horse’s neck
(Fig. 15.2) (Manfredi et al., 2005b). Both with and without rein
tension, the bits varied in their proximity to the premolars, and
angulation of the cannons. Rein tension moved the mouthpiece
away from the palate by indenting the horse’s tongue. The amount
of time spent with the mouth quiet, gently mouthing the bit,
opening the mouth, using the tongue to raise the bit between the
cheek teeth and displacing the dorsum of the tongue over the bit
were measured. The horses’ intra-oral behaviors with these six bits
were evaluated with a loose rein and with 25  N tension. When rein
tension was applied, the time spent with the mouth quiet or gently
mouthing the bit decreased and the time spent in the other behav-
iors increased. The behaviors differed significantly between horses
but not between bit types (Manfredi et al., 2010).
It has been suggested that horses are unable to swallow when
wearing a bit (Cook, 1999). However, Manfredi et al. (2005a) per-
formed an endoscopic study of horses cantering on a treadmill with
tight side reins to maintain poll flexion while wearing a halter, a
bitless bridle, a loose ring jointed snaffle and a Myler bit. Swallow-
ing occurred frequently under all four conditions.
Materials, types and sizes of bits are often changed by riders when
they think they experience a ‘mouth problem’ in the horse. Usually
the trainer’s perception guides the choice and the conclusions about
how these act, but scientific evidence to support these ideas is
minimal. For example, relative to dressage the difference between
the action of a single and double bridle warrants scientific investiga-
tion in a larger population of horses. One reason why this is difficult
to study is that tension sensors in the reins measure the combined
effects of horse and rider and cannot partition the effects between
the two. Thus it is not possible to separate variation in a horse’s
response to a bit from the force applied by the rider or differences
in rider technique.

Fig 15.2  Lateral radiographic views of bit positions without rein tension
(left panel) and with 25 ± 5 N bilateral rein tension (right). Bits represented
from top to bottom are jointed snaffle, KK Ultra, Boucher, Myler low port
comfort snaffle, Myler ported barrel, Myler correctional-ported barrel.
Reprinted from Manfredi, J., Clayton, H.M., Rosenstein, D., 2005b. Radiographic study of bit
position within the horse’s oral cavity. Equine Comp. Exerc. Physiol. 2, 195–201, with
permission from Cambridge University Press.

344
 Bits and reins

(Biau et al., 2002). The reins were: ‘rubber bands’ going from the
lowest point of the girth to the bit; a ‘Chambon’ going from lowest
point of the girth through a pulley at the poll area to the bit; and
‘back lifts’ that were attached to the sides of the girth by a double
mechanism, and to the poll. Rubber bands increased propulsion (i.e.
the mean acceleration vector used for propulsive work) of the fore-
limbs at trot. Back lifts increased forelimb propulsion and dorsoven-
tral activity at walk and trot (i.e. the FFT-integrated power spectrum
signal of the dorsoventral acceleration signal). Chambons increased
hind limb propulsion and forelimb propulsion and hind limb dor-
soventral activity at trot. The authors’ interpretation was that a
Chambon was likely the most effective if the aim was to increase hind
limb activity in a long and low position.
Draw reins used alone or in combination with normal reins were
A compared to normal reins only in horses ridden at trot on sand-
covered asphalt over a forceplate (Roepstorff et al., 2002). The body
weight was shifted caudally only when draw reins were used with
normal reins (Byström et al., 2006). Kinematic analysis of the head
angle, measured on the front of the nose relative to the horizontal
plane (angles >90° equals nose in front of vertical), was different
in all situations. With draw reins alone it was ~97°, combining draw
reins with normal reins it was ~109°, while with normal reins it
was ~116°. With draw reins forelimb retraction was increased at
hoof lift. The hip joint extended more quickly and the hock joint
flexed more with both reins compared to normal reins only. The
pelvis was more horizontally inclined when draw reins were com-
bined with a normal rein in comparison with draw reins alone, but
the hip joint angular pattern was not significantly different between
B
these conditions. It was concluded that the riders were, to a certain
extent, able to shift weight towards the hind quarters with the com-
bination of reins. An important finding, though, was that when
using the draw rein alone to simulate an overuse of the draw rein,
the horse’s weight shifted to the fore quarters. So, when used incor-
rectly draw reins had the opposite effect from that which is intended
and the caution to use draw reins with the utmost care (Podhajsky,
1965) seems to be scientifically validated.

C Rein tension
From a technical point of view it is quite simple to measure rein forces
Fig 15.3  The three studied reins. (A) Rubber bands, (B) chambons and (C) acting between the hand of the rider and the bit. However, it is impor-
back lift. tant to use appropriate force transducers for dynamic measurements;
Reprinted from Biau, S., Couve, O., Lemaire, S., Barrey, E., 2002. The effect of reins on kinetic
transducers based on a spring will not provide accurate force data and
variables of locomotion. Equine Vet. J. 34 (Suppl. ), 359–362, with permission from the Equine
will underestimate forces when used in a dynamic system. Interpreta-
Veterinary Journal.
tion of the data can also be problematical due to the inherent vari-
ability associated with a pattern of spikes that are characteristic of each
gait (Clayton et al., 2005). The mean tension ignores this inherent
Reins variability. A more detailed and informative approach measures the
Auxiliary reins, defined as reins other than simple inelastic reins forces as minima, maxima and means with standard deviations. The
passing directly from the rider’s hand to the bit, differ widely in their force distribution can be mapped over time and compared between
popularity among riders. Many see them as coercive, because they left and right reins. However, the force cannot be partitioned between
most often force the horse’s head in a downward direction, with some horse and rider using this method. Some studies have presented forces
types being perceived as having a more severe effect. Other trainers in weight units (see below), but for the sake of comparison we have
use them for a short time as temporary aids to provide better long- here assumed that 1 kg (weight) equals 9.8 N (force).
term control of the horse’s head and neck position (see more on this In pioneering work, it is stated that unexpectedly large rein forces
under collection below). For example, with draw reins the rider can were found, 5–75 N, up to maximal 150 N, on each rein. Most
pull the horse’s head downwards and backwards to its anatomical riders, using double bridles, used forces of around 5 N at halt
limit, whereas upwards movements are restricted by a running mar- increasing to 20 N at canter (Preuschoft et al., 1995). In further
tingale in which the reins slide freely through loose rings that are work, different types of riders were compared (Preuschoft et al.,
held in place by a neck strap attached to the girth. The rings control 1999). Very low rein tensions were found for western riders, for
the line of action of rein tension and prevent the head of the horse riders riding only on the curb and at a few dressage stables where
coming above the base of the skull. A few auxiliary reins have been forces registered below or just above 20 N, when performing
compared (Biau et al., 2002) or evaluated (Roepstorff et al., 2002; demanding exercises. In other dressage stables, mean forces ranged
Byström et al., 2006; Heleski et al., 2009) but more information is from 59 to 147 N, without any difference between single and
needed to provide a comprehensive picture. In a study of eight saddle double bridles. Peak force during regular work was below 49 N in
horses, normal reins were compared with three types of auxiliary driving horses, 98–176 N during dressage training rising to 245 N
reins (Fig. 15.3), using accelerometers attached at the sternum and when slowing down from canter. In trotters moving at high speed,
sacrum as the horses walked and trotted unridden in a horse walker forces up to 392 N were found (Preuschoft et al., 1999).

345
 15 Horse–rider interaction

Cartier d’Yves and Ödberg (2005) studied right rein tension during 1000
a dressage test performed by three riders and 30 horses, including Left
riding school horses, leisure horse and professional competition Right
800

No. measurements
horses. Mean rein tension was approximately 13 N in riding school
horses and 9.8 N in leisure horses, with intermediate values in the 600
professional competition horses. Mean rein tension values were not
statistically different between groups. Additionally, judge’s or rider’s 400
subjective evaluation of lightness did not correlate with actual mea-
surements of mean rein tension. Clayton et al. (2003) also report that 200
the rider’s perceptions of rein tension differed from the registered
forces in one horse. In further work by the same author, one experi- 0
enced rider rode one horse, chosen because it was known to lean

0–4

5–8

16

20

24

28

32

33–
9–1
against the bit. Although the rider perceived a constant rein tension,

13–

17–

21–

25–

29–
in walk there were regularly occurring spikes with a frequency of 108 Tension (kp)
per min, in trot 168 spikes per min and in canter 90 spikes per min
(Clayton et al., 2005). The maximum values of the rein tension Fig 15.4  Distribution of number of left and right rein force recordings (y)
ranged from 4 to 43 N at walk, 19 to 51 N at trot and 21 to 104 N at by amount of tension (x) in one horse and rider in sitting trot on a treadmill
canter. In this rider, synchronized video analysis showed that peak during a period of 15 s with a sampling frequency of 140 Hz.
rein tension occurred during diagonal stance in both trot and canter, From Roepstorff (unpublished).
which coincided with downward motion of the horse’s head.
Rein tension and the steadiness of the hand were evaluated in horse in walk, trot and canter. There were differences between the
four horses and nine novice riders (Heleski et al., 2009). Three sides for the two horses. For example, there was more slackening of
experimental conditions were studied: running martingales that the reins of the horse’s non-preferred side and taken together the
stabilized the angle of the rein to the horses mouth during riding, left rein, especially in the left-lateralized horse, had generally higher
elastic rein inserts that yielded if tension exceeded 20 N, which rein tension. The authors speculated that right-lateralized riders
happened very rarely in this study (in 0.6% of the spikes evaluated), might benefit from riding right-lateralized horses because this
and controls without either rein aid. Conflict behaviours of the seemed to enhance symmetry in the riding.
horse, evaluated using an ethogram, did not differ between experi- Warren-Smith et al. (2007) studied one professional, one interme-
mental conditions. Mean rein tension was highest for the condition diate and one novice rider on 22 horses performing a simple dres-
with running martingales (4.1 N with a martingale versus 3.5 N for sage test under saddle and the same horses being long-reined by a
both rein inserts and controls). The authors suggested that carefully professional handler. Mean rein tension during riding was 7.4 N and
fitted running martingales may be advantageous for novice riders during long-reining 10.7 N. Rein tension was highest when halting
on the basis that the head position of the horse was less often and higher when turning compared to riding on straight lines. When
judged as unacceptably high. It is suggested by many riders and horses were ridden, tension in the left rein when turning left was
veterinarians that a somewhat lower head position actually spares higher than the tensions in either rein when turning right, which
the locomotor health of the horse. could be interpreted as evidence of motor laterality in the rider or
The rein tension of three horses, ridden by their own expert rider the horse. The novice rider used more tension for turning. Tension
on a treadmill, was studied at walk, trot and passage (Roepstorff, did not differ between bridles and halters. Tensions were higher in
unpublished). In the vertical head-neck position, data from both ex-racehorses compared to other horses.
reins averaged over four trials, showed that the three horses had In an effort to separate the effects of horse and rider on rein
different mean low and high values of 11 N and 16 N, 1.5 N and tension, Clayton et al. (2011) studied the tension in side reins made
3.5 N, and 2.7 N and 4.8 N. Mean rein tension increased with of three materials with different amounts of elasticity and adjusted
increasing collection (as determined by the head and neck posi- to three lengths. The materials were inelastic, stiff elastic and, com-
tion). This is contrary to most riding texts and to the concept that pliant elastic. The rein lengths were neutral (equal to the length of
rein contact should be able to be dropped in collection (descente the horse’s neck when standing in a relaxed position), long (10 cm
de main) with the horse remaining in the collected movement/ longer than neutral) and short (10 cm shorter than neutral). Horses
position when rein contact is released (Decarpentry, 1949). In walk, (n = 8) trotted in hand at consistent speed in a straight line wearing
trot and passage the mean rein tension was higher in eight of nine a bit and bridle with the three types of side reins being evaluated
comparisons when the head and neck were in the vertical (competi- in random order. Strain gauge transducers (240 Hz) measured rein
tion) position compared with positions slightly behind the vertical, tension dynamically and the authors reported minimal, maximal
with a low neck or with a neck elevated and the bridge of the nose and mean rein tension, rate of loading and impulse. Rein tension
in front of the vertical. On the contrary, all three horses had a much oscillated in a regular pattern with a peak during each diagonal
lower mean rein tension when walking with the neck low and stance phase. Within each rein type minimal, maximal and mean
forward compared with the competition position. The maximal tension were higher with shorter reins. At neutral or short lengths,
peaks in walk were found at contralateral hind limb stance and in minimal tension increased and maximal tension decreased with a
trot there was one peak in each rein at stance, which agrees with more elastic rein. Short, inelastic reins had the highest maximal
the results found by Clayton et al. (2003, 2011). tension and highest rate of loading. Since the tension variables
In walk and sitting trot, rein tension was generally lower in one responded differently to rein elasticity at different lengths, it was
of the reins in two of three horse–rider combinations. A typical recommended that studies of rein tension should report a set of
example from the vertical head–neck position in one of the horses variables representing different aspects of rein tension.
is shown (Fig. 15.4), where the right rein more often has lower In conclusion, some studies only measured tension when horses
tension, while the opposite pattern is seen for the two next bars. were moving straight, others have looked at rein tension during
This is further evidence of motor laterality, but the data cannot dressage tests and correlated rein tension to various activities or to
detect whether it comes from the horse, the rider or both. In rising different head and neck positions. Only a few studies have correlated
trot the tendency was that the tensions in both reins were higher rein tension to the footfall sequence. It is likely that the situations
for the diagonal on which the rider was sitting, compared to the measured so far are relatively standardized experimental situations
diagonal on which the rider rose. and, in real life, much higher forces will be applied in emergency
Kuhnke et al. (2010) studied 11 right-lateralized (right-handed) situations or by novice riders. Since some studies presented their
riders that rode one left-lateralized horse and one right-lateralized results as mean tension there is a need for more information about

346
 Basic rider movements

the tension patterns, minimal and maximal tensions, the evenness with diagonal hoof placement in which each diagonal stance
of the contact and the rein forces applied during half halts, halts and encompasses a braking phase when the horse’s limbs are com-
jumping. It will also be necessary to study horse and rider laterality pressed and a propulsive phase when the horse’s limbs extend
with respect to rein forces. If the rider/trainer could better analyze (Farley et al., 1993). The horse keeps its back rigid through strong
the distribution of forces on the horse’s mouth, this may lead to a muscle activity to resist wasteful lateral and rotational movements
more efficient and humane riding technique. One of the challenges induced by the diagonal stance (Robert et al., 1998; Faber et al.,
facing researchers is to distinguish between correctly trained sport 2001). Accordingly it can be expected that the movement pattern of
horses that accept the bridle with a light soft contact and horses that the rider will be dominated by the effects of and reactions to the
achieve a light contact by dropping the bit and working in a pseudo- alternating impacts and push-offs. Even in high-level dressage riders
collected frame (with the nose behind the vertical, flexion not at the riding at collected trot, i.e. rather slowly (~3 m/s), the rider’s move-
poll but further down the spine, see also further down on collection) ments can largely be explained from the vertical and horizontal
(German National Equestrian Federation, p 143). decelerations and acceleration of the horse’s trunk that take place
during each diagonal stance (Byström et al., 2009).

Basic rider movements

Rider movements at sitting trot
The following description of the rider’s movement starts at ground
Trot contact of one diagonal and ends after the following suspension
Of the three basic gaits, trot is the gait at which rider movements phase. The description is based on kinematic measurements of
have been most extensively studied, particularly the sitting trot, seven high-level dressage riders riding at collected trot on a force-
although rising trot has also been evaluated and compared with measuring treadmill (Byström et al., 2009) unless a different refer-
sitting trot. A number of studies describe aspects of the basic rider ence is given.
movements (Schils et al., 1993; Terada, 2000; Terada et al., 2004, Both the trunk of the horse and the rider’s seat reach maximal
2006; Peham et al., 2001; Matsuura et al., 2003, 2005; Lovett et al., upwards displacement at the end of the suspension phase. Following
2005; Lagarde et al., 2005; de Cocq et al., 2010a; Byström et al., diagonal ground contact both horse and rider move downwards.
2009, 2010b; Symes & Ellis 2009; von Peinen et al., 2009) mostly Because the horse descends more quickly than the rider, the vertical
at trot and, to a lesser extent, at walk and canter. distance from the rider’s seat to the saddle increases during the first
Starting with rider movements at trot, one of the fundamental 5% of the stride (Fig. 15.5). The rider is thus little influenced by move-
questions addresses the characteristics of the trot that the rider must ments of the horse during the landing phase and the rider therefore
react and adapt to. The trot is a regular symmetrical gait continues the movements induced by the preceding push-off.

Longitudinal distance rider's neck -

Pitch angle rider's pelvis Flexion–extension rider's back horse's L3
10 10 60
40
5 5
20
Distance (mm)
Angle (º)

Angle (º)

0 0 0
-20
-5 -5
-40
-10 -10 -60

0 50 100 0 50 100 0 50 100

% stride % stride % stride

Vertical distance rider's seat -

Vertical position horse's L5 Vertical position rider's seat saddle front
60 60 60
40 40 40
20 20 20
Distance (mm)
Position (mm)

Position (mm)

0 0 0
-20 -20 -20
-40 -40 -40
-60 -60 -60

0 50 100 0 50 100 0 50 100

% stride % stride % stride
Fig 15.5  Upper row, left to right: rider’s pelvic pitch, flexion–extension of the riders back, longitudinal distance from the rider’s neck at the C7 vertebra to
the horse’s L3; lower row, left to right: vertical position of the horse’s L5, vertical position of the rider’s seat, and vertical distance from the rider’s seat to the
front of the saddle in seven horses ridden in sitting trot on a treadmill (Byström unpublished). Solid line is mean, dashed lines indicate one standard
deviation. On the y axes, zero represents the mean value for the stride with each horse/rider combination curve being adjusted so individual means are zero.
For the rider’s pelvis, positive pitch represents anterior movement of the upper part of the pelvis relative to the lower part. For the rider’s back, flexion is
anterior pitch of the upper body relative to the pelvis, i.e. flexion of the lumbar spine. The lower bars show the stance of the left fore, right fore, left hind and
right hind (top to bottom).

347
 15 Horse–rider interaction

Shortly after first contact the rider’s trunk (hip to shoulder) is tilted of maximal propulsive ground reaction force (Merkens et al., 1993),
maximally backwards as a consequence of both forward motion of the rider’s seat also starts to move forwards in relation to the horse.
the rider’s hip and a backward motion of the rider’s shoulders, which As the rider is pushed out of the saddle the rider’s trunk tilts for-
was interpreted as a feed forward mechanism by which experienced wards due to a combination of relative forwards movement of the
riders anticipate and compensate for the longitudinal deceleration of shoulders and backwards movement of the hips (Terada et al.,
the horse (Terada et al., 2006). As the rider’s seat moves forward in 2006). At this time the rider’s pelvis rotates posteriorly and the
relation to the horse, the rider’s neck and feet move backwards. At the rider’s back and neck straighten. The rider’s legs extend and the feet
same time the rider’s pelvis is maximally posteriorly rotated (around move medially and backwards in relation to the hips. These move-
the transverse axis), the rider’s back and neck are stretched in exten- ments then continue into the following suspension phase.
sion and the rider’s legs are also maximally extended (Fig. 15.6).
From 5% stance onwards both vertical and braking ground reac- The seat at rising trot
tion forces increase rapidly in unridden horses (Barr et al., 1995)
and the forwards movement of the horse’s center of mass is rapidly Rising trot is a rather recently invented technique that makes trotting
decelerated (Buchner et al., 2000). At the same time the distance more comfortable for the rider and reduces peak forces on the
between rider and horse decreases causing the rider’s seat to be horse’s back (de Cocq et al., 2010). Using this technique, the rider
pressed against the saddle. Following this the pressure of the saddle alternately sits on one diagonal pair of limbs and rises on the other.
against the back of the horse also increases rapidly (Fig. 15.7). The
impact forces are now transmitted from the horse to the rider and 2.5
this has a clear effect on the rider’s movements. The rider’s pelvis
starts to rotate anteriorly, the rider’s back extends (hollows) and the
rider’s neck flexes. The rider’s trunk assumes a vertical position 2
(Terada et al., 2006) and the rider’s neck moves forwards in relation

Normalized force
to the horse. The rider’s leg joints flex, the heels are lowered and
1.5
the feet move laterally and forwards in relation to the rider’s hips.
As the rider sinks deeper into the saddle the forward movement of
the rider’s seat eventually ceases at about 16% of the stride after 1
diagonal contact. The rider’s seat then starts to move backwards in
relation of the horse as midstance is approached.
At midstance both horse and rider reach maximal downwards 0.5
displacement, the rider sits deepest into the saddle (Byström et al.,
2009) (Fig. 15.8) and the pressure of the saddle against the back of
the horse is maximal. 0
During the propulsive phase, the upwards–forwards directed 0 50 100
push-off by the horse is transmitted to the rider’s seat through the % stride
saddle. The rider’s seat starts to move out of the saddle as soon as Fig 15.7  The saddle force normalized to rider weight (N/kg) at trot with
the horse’s trunk starts to move upwards (Fig. 15.5) and the pressure stance bars in seven horses ridden in sitting trot on a treadmill (Byström
of the saddle against the back of the horse then starts to decrease unpublished). Solid line is mean, dashed lines indicate one standard
(Fig. 15.7). A little later, at 36% of the stride following ground deviation. The lower bars show the stance of the left fore, right fore, left
contact of the diagonal and coinciding approximately with the time hind and right hind (top to bottom).

Fig 15.6  Rider in sitting trot landing on the right diagonal Fig 15.8  Rider in sitting trot at midstance on the left diagonal at midstance.
Photograph courtesy of the Equine Biomechanics Research Team from the Universities of Courtesy of the Equine Biomechanics Research Team from the Universities of Uppsala,
Uppsala, Utrecht and Zürich. Utrecht and Zürich.

348
 Basic rider movements

The full stride sequence for rider movements during rising trot has treadmill (Byström et al., 2010b) unless a different reference is given.
not yet been documented. A preliminary kinematic study of five As in the trot, it is paramount to know the characteristics of the gait
riders performing rising trot on one horse showed that at first contact in order to understand rider movement. The walk is a symmetrical
of the sitting diagonal the rider’s trunk and thigh were more vertical gait without a suspension phase. The horse alternates between
while the shank was angled more backwards compared with the tripedal and bipedal support, with the limbs functioning as rigid
equivalent time on the rising diagonal when the trunk leaned more struts to raise the withers or croup from first contact to midstance.
forward and the thigh was also more angled forward while the shank This is in contrast to trot, where both withers and croup move
was closer to the vertical (Lovett et al., 2005). These findings suggest downwards during the first half of stance and then upwards in the
that the rider first leans the trunk forward by flexing the hip joints following propulsion and suspension phase. As the forelimbs of the
and simultaneously moves the feet more forward relative to the hip walking horse move approximately 25% of the stride out of phase
before rising with a forward–upward movement of the seat through in relation to the hind limbs, the withers are raised when the croup
extension of the hip joints such that both trunk and thigh position is lowered and vice versa. All together this causes the vertical dis-
approach the vertical. This interpretation must be, however, consid- placement of the saddle area of the horse’s back at walk to be about
ered preliminary and perhaps incomplete as it is based on measure- half that of the trot (Matsuura et al., 2003) while rotation of the
ments at diagonal first contacts. It is also possible that there are horse’s trunk around the transverse axis (pitch) is almost doubled
relatively large variations in rider technique in rising trot. (Table 15.1) (Byström et al., 2010b). The amplitudes of lateral and
Saddle pressure measurements of one high-level dressage rider longitudinal excursions of the horse’s body center of mass, which
riding at sitting and at rising trot on several different horses showed are necessary for the horse to maintain its balance during the alter-
a tendency towards smaller longitudinal excursion of the center of nating tripedal and bipedal support phases, are approximately equal
pressure during rising trot compared to sitting trot (Peham et al., to the vertical excursion at walk (Buchner et al., 2000). Such dis-
2009). This means that the rider maintains a more constant position placement can be accomplished through lateral bending and axial
relative to the horse even though the longitudinal displacement of rotation of the back and the axial rotation range of motion is almost
the rider’s body is larger at rising trot compared to sitting trot (Mat- doubled during walk compared to trot (Faber et al., 2000; Faber
suura et al., 2005). A preliminary study of 30 riders with a strain et al., 2001). Therefore, extra-sagittal movements constitute a pro-
gauge between the stirrup and the stirrup leather concluded that the portionally larger part of the assembled motion pattern of the walk
force was higher at rising than sitting trot (van Beek et al., 2008). compared to trot, and it is to be expected that the rider’s movement
One study examined saddle and leg forces using three saddle mat pattern will be equally more complex. To gain full understanding
systems between the saddle and the horse in 11 horse–rider combi- of the horse–rider interaction at walk, the rotations of the rider’s
nations of high but not international level, moving in straight body segments around the vertical and longitudinal axes must also
line, shoulder-in (when the horse is bent away from the line of be considered. However, focusing on sagittal plane movements is a
travel) and in travers (when the horse is bent towards the line good place to start describing the rider’s movements.
of travel) (de Cocq et al., 2010b). Mean total saddle force was sig-
nificantly lower when riding straight ahead (671 N) than when
riding shoulder-in (707 N) or travers (726 N). Mean inside saddle Rider movements in the sagittal plane at walk
force was significantly higher when riding travers (468 N) than
At walk both the range of vertical displacement of the rider’s body
when riding straight on (425 N) or shoulder-in (413 N). Maximum
(Matsuura et al., 2003) and the rider’s seat in relation to the front
outside leg force was significantly higher when riding travers
part of the saddle are smaller than at trot (Table 15.1). Further, the
(47.2 N) than when riding straight (31.6 N) or shoulder-in
rider’s seat moves upwards in relation to the front part of the saddle
(34.2 N). This information enhances the understanding of contem-
while moving downwards in relation to L3 of the horse and vice
porary rider movement but should be complemented with data on
versa (Fig. 15.9). This indicates that the vertical displacement of the
more riders and various dressage riding styles.
rider’s seat at walk results mainly from the croup being lowered and
raised in relation to the withers rather than true vertical displace-
Walk ment of the rider in relation to horse and saddle. The fact that the
The description is based on kinematic measurements of seven high- rider shows a decreased range of motion for pelvic pitch, back
level dressage riders riding at collected walk on a force-measuring flexion–extension and all head–neck rotations at walk compared

Vertical distance rider's seat - Vertical distance rider's seat -

Pitch angle horse's trunk horse's L3 saddle front
30 30
4
20 20
2
10 10
Distance (mm)

Distance (mm)
Angle (º)

0 0 0

-2 -10 -10

-20 -20
-4
-30 -30
-6
0 50 100 0 50 100 0 50 100
% stride % stride % stride
Fig 15.9  Left to right: The horse’s trunk pitch angle, the vertical distance between the rider’s seat and the horse’s third lumbar vertebra (L3) and the vertical
distance from the rider’s seat to the front of the saddle in seven horses ridden in walk on a treadmill. Solid line is mean, dashed lines indicate one standard
deviation. Zero value on the y axis represents mean value for the stride with each horse/rider combination curve being adjusted so individual means are
zero. The lower bars show the stance of the left fore, right fore, left hind and right hind (top to bottom).
Data from Byström (unpublished).

349
 15 Horse–rider interaction

Table 15.1  Stride range of motion in degrees or mm for variables measured in high-level dressage horses and riders (n = 7) during
collected walk and trot on a treadmill#. Pitch is rotation around a transverse axis, roll is rotation around a longitudinal axis and yaw is
rotation around a vertical axis.

Walk ROM (°/mm) Difference trot–walk (%) Trot ROM (°/mm)

Saddle Pitch 6.1 ± 0.9 9 ± 13 5.6 ± 0.6

Roll 8.5 ± 2.7 53 ± 69 7.3 ± 5.2

Yaw 8.3 ± 1.5 49 ± 33* 5.7 ± 1.0

Rider pelvis Pitch 9.7 ± 2.0 −28 ± 20* 13.9 ± 2.2
Roll 5.6 ± 0.6 18 ± 31 5.1 ± 1.1
Yaw 8.2 ± 1.9 9 ± 31 7.9 ± 2.1
Rider upper body Pitch 6.0 ± 2.0 −42 ± 14* 10.7 ± 3.4
Roll 5.0 ± 1.8 12 ± 42 4.9 ± 1.8
Yaw 5.0 ± 1.3 −6 ± 21 5.5 ± 1.1
Rider head–neck Pitch 11.0 ± 4.8 −30 ± 19* 15.7 ± 4.5
Roll 4.0 ± 0.9 −28 ± 22* 5.9 ± 1.1
Yaw 4.1 ± 1.3 −25 ± 22* 5.7 ± 2.4
Rider seat–saddle front Vertical distance 25 ± 6 −55 ± 14* 58 ± 10
Rider seat–horse (L3) Vertical distance 38 ± 9 −5 ± 39 45 ± 13
Longitudinal distance 41 ± 4 −14 ± 24 50 ± 24
Rider neck–horse (L3) Longitudinal distance 53 ± 14 21 ± 33 45 ± 6
Horse neck Pitch 8.0 ± 1.6 43 ± 50 6.0 ± 1.4
Horse trunk Pitch 6.0 ± 0.7 53 ± 22* 4.0 ± 0.7
Horse (L5) Vertical displacement 62 ± 7 −42 ± 8* 106 ± 8
#
Table values represent group mean ± SD.
*Significant difference (p < 0.05) between gaits in a paired non-parametric test (Wilcoxon).
ROM, range of motion.
Reprinted from Byström, A., Rhodin, M., von Peinen, K., et al., 2009. Basic kinematics of the saddle and rider in high-level dressage horses trotting on a treadmill. Equine Vet. J. 41,
280–284, with permission from the Equine Veterinary Journal.

with trot (Table 15.1) may reflect the absence of impact effects on
the rider at walk. This is supported by the fact that the range of
Lateral and rotational rider movements at walk
motion for vertical displacement of the rider’s body is only about For the rider’s body movements at walk the lateral amplitude is
1 cm at walk (Matsuura et al., 2003) and the amplitude of the approximately equal to the longitudinal amplitude and signifi-
rider’s body movements is significantly lower in the vertical than in cantly greater than the vertical amplitude (Matsuura et al., 2005).
the lateral and longitudinal directions, which is contrary to the Consequently, the extra-sagittal movements constitute a significant
findings at trot (Matsuura et al., 2005). Riders are therefore more part of the assembled motion pattern of the walk, both for the
likely to be adjusting their movements at walk to the alternating horse and the rider and these have to be taken into account to
difference in height between the horse’s croup and withers rather form a comprehensive picture. This is clearly reflected in the
than to vertical and horizontal deceleration and acceleration of the saddle pressure pattern of walk (Fig. 15.13) and the fact that the
horse’s trunk as in trot (Byström et al., 2009). vertical movements of the horse and rider were insufficient to
To appreciate the rider’s positional adjustments at walk, we note explain the various force peaks observed (von Peinen et al., 2009).
that from the tripedal support of one forelimb and both hind limbs To explain this pattern the rotation of the rider’s pelvis around the
to the tripedal support of both forelimbs and one hind limb, i.e. longitudinal axis (roll rotation) as well as the lateral bending and
from forelimb midstance to hind limb midstance, the horse’s croup axial rotation of the horse’s back must be considered in addition
is raised while the withers are lowered in relation to the croup (Fig. to sagittal plane movements. Rotations of the rider’s pelvis around
15.9). When the horse’s trunk starts to rotate cranially this transmits the vertical axis (yaw rotation) generally follow the movements of
a forward push to the rider’s seat from behind that causes the seat the saddle and seem to have little influence on the saddle pressure
to slide forward (Byström et al., 2010b). To maintain balance the pattern.
rider’s pelvis rotates posteriorly (around the transverse axis), the The motion pattern has been described in high-level dressage
neck moves backwards relative to the horse and the feet are retracted riders riding at collected walk on a treadmill (von Peinen et al.,
through extension of the leg joints, particularly the hip joints (Fig. 2009; Byström et al., 2010b). During the tripedal dual hind limb
15.10). These movements are reversed after hind limb midstance support phase, the horse’s croup is rapidly lowered and, at the same
when the croup descends relative to the withers. time, the horse’s back is bent laterally towards the left hind limb

350
 Basic rider movements

Pitch angle rider's pelvis Flexion–extension rider's back Flexion–extension rider's hip joint
5 10
4
5
0 2
Angle (º)

Angle (º)

Angle (º)
0 0
-5 -2 -5
-4
-10 -10
-6
0 50 100 0 50 100 0 50 100
% stride % stride % stride

Longitudinal distance rider's neck - Longitudinal distance rider's seat -

horse's L3 horse's L3
20
40

20 0
Distance (mm)

Distance (mm)

-20 -20

-40
-40
-60
0 50 100 0 50 100
% stride % stride
Fig 15.10  Top row, left to right: Rider pelvic pitch, flexion–extension of the rider’s back and flexion/extension of the rider’s hip joints in seven horses ridden
in walk on a treadmill. Bottom row, left to right: longitudinal distances from the rider’s neck and from the rider’s seat to the horse’s third lumbar vertebra (L3)
in seven horses ridden in walk on a treadmill. Solid line is mean, dashed lines indicate one standard deviation. Zero value on the y axis represents mean value
for the stride with each horse/rider combination curve being adjusted so individual means are zero. For the rider’s pelvis, positive pitch represents anterior
movement of the upper part of the pelvis relative to the lower part. For the rider’s back, flexion is cranial pitch of the upper body relative to the pelvis, i.e.
flexion of the lumbar spine. The lower bars show the stance of the left fore, right fore, left hind and right hind (top to bottom).
Data from Byström (unpublished).

FGFL FGHL
FGFR FGHR
GRF (%hnwt)

60
30
0

120 P3
P1
M1 P2
110
FS (%nwt)

M2 M3
100

90

0 14 27 41 50 63 77 92 100
% stride duration
Fig 15.11  Mean stride normalized ground reaction forces of the four limb (FGFL, left forelimb; FGFR, right forelimb; FGHL, left hind limb; FGHR, right hind
limb) (top) and total saddle force (bottom) of seven horses at walk. Ground reactions forces of individual limbs (FG limb) are expressed as percentage of the
total weight of horse and rider (% hrwt). Total saddle force (FS) is expressed as percentage of the rider’s weight (% rwt). The thick line is mean value and thin
lines indicate one standard deviation. The local extremes (minima M1 to M3, maxima P1 to P3) are indicated.
Reprinted from Peinen, K., von, Wiestner, T., Bogisch, S., Roepstorff, L., van Weeren, P.R., Weishaupt, M.A., 2009. Relationship between the forces acting on the horse’s back and the movements
of rider and horse while walking on a treadmill. Equine Vet. J. 41, 285–291, with permission from the Equine Veterinary Journal.

such that the back is convex to the right, and rotated around the movements causes a marked reduction in pressure under both the
longitudinal axis such that the left tuber coxae is higher than left and right front thirds of the saddle while pressure increases
the right. Simultaneously the rider’s seat moves rapidly backwards mainly on the left side under the rear part of the saddle (Fig. 15.13)
in relation to the horse and the rider’s left hip is being lowered in (von Peinen et al., 2009).
relation to the right hip, i.e. the rider’s pelvis undergoes roll rotation When the right hind limb starts to break-over and the croup rises
to the left (Fig. 15.12). This combination of horse and rider relative to the withers, the pressure under the rear left third of the

351
 15 Horse–rider interaction

Roll angle rider's pelvis Yaw angle rider's pelvis

5
4
2 0
Angle (º)

Angle (º)
0
-2 -5
-4
-10
-6
0 50 100 0 50 100
% stride % stride

Longitudinal distance rider's seat - horse L3 Pitch angle horse's trunk

20
4
Distance (mm)

0 2
Angle (º) 0
-20 -2
-4
-40
-6
0 50 100 0 50 100
% stride % stride
Fig 15.12  Rider kinematics in seven horses ridden at walk on a treadmill. Top row, left to right: rider’s pelvic roll angle (positive is rotation to the right when
viewed from behind) and rider’s pelvic yaw angle (positive yaw is rotation to the right when viewed from below) during the walk. Bottom row, left to right:
longitudinal distance from the rider’s seat to the horse’s third lumbar vertebra (L3) and pitch angle of the horse’s trunk (positive pitch is lowering of the
withers relative to the croup). Solid line is mean, dashed lines indicate one standard deviation. The lower bars show the stance of the left fore, right fore, left
hind and right hind (top to bottom).
Data from Byström (unpublished).

Left half of saddle mat Right half of saddle mat

20 20

10 10
%rwt

FSL FSR
0 0

-10 -10

10 10
5 5
%rwt

FSLf FSRf
0 0
-5 -5

10 10
5 5
%rwt

FSLc FSRc
0 0
-5 -5

10 10
5 5
%rwt

FSLh FSRh
0 0
-5 -5
0 14 27 41 50 83 77 92 100 0 14 27 41 50 83 77 92 100
% stride % stride
Fig 15.13  Mean (thick line) ± SD (dashed lines) for relative changes of stride standardized saddle forces of the left (FSL, left panel) and right (FSR, right panel)
mat half and partial forces of six sectors (n = 7 horses). Each half of the mat was divided into a front (f), central (c) and hind (h) sector; partial forces are
labeled accordingly. Forces are presented as deviation from the stride mean value of the respective sector. Bold curve sections mark significant deviations
from the stride mean (p < 0.05). For scaling of the abscissa see Fig. 15.11. Solid line is mean, dashed lines indicate one standard deviation.
Reprinted from Peinen, K., von, Wiestner, T., Bogisch, S., Roepstorff, L., van Weeren, P.R., Weishaupt, M.A., 2009. Relationship between the forces acting on the horse’s back and the movements
of rider and horse while walking on a treadmill. Equine Vet. J. 41, 285–291, with permission from the Equine Veterinary Journal.

352
 Rider skill

saddle increases even further (Fig. 15.12). This induces not only Rider skill
forwards movement of the rider’s seat but also roll rotation of the
rider’s pelvis to the right.
During the following diagonal (right front, left hind) support General
phase the rider’s seat moves rapidly forwards in relation to the
The following section presents the results of studies that evaluate
horse. At the same time the horse’s dorsal back muscles contract
the effect of rider skill level with most of the results being derived
strongly on the left side in the cranial saddle area during retraction
from traditional riders. A large kinematic overground study by
of the right forelimb (studied without rider) (Licka et al., 2009).
Schils et al. (1993) included 63 riders divided into three groups of
This coincides with a marked increase in pressure under the left
beginner, intermediate and advanced based on video evaluation of
front third of the saddle while the rider’s pelvis continues to roll to
their skill level by judges. Midstance and midswing rider angles were
the right.
determined. Other studies have strived to evaluate rider skill (Lovett
After the transition from bipedal to tripedal support at ground
et al., 2005) and some of these results will be presented in the fol-
contact of the left forelimb the dorsal back muscles in the left
lowing paragraphs. Speed and temporal characteristics of the horse
cranial saddle area start to relax (Licka et al., 2009) and the forwards
as well as riding style and the rider’s degree of control of the
rotation of the horse’s trunk slows as the left hind limb approaches
horse’s movements are likely to be important factors affecting
midstance. Following this the pressure distribution between the
the rider’s movement pattern. The influence of these factors on the
front thirds of the saddle starts to even out as the pressure increases
rider’s movements has however not been evaluated.
under the right front third (Fig. 15.13). Just after this, roll rotation
of the rider’s pelvis to the right also ceases (Fig. 15.12).
When the left hind limb has passed its midstance position the
horse’s back becomes progressively more axially rotated to the left
Trot
(lower on the left side) and bent to the right (convex to the left). The rider needs to have excellent postural control through well-
This decreases the support under the left rear third of the saddle, timed muscle activity to constantly maintain the same posture in
which is reflected in the very low pressure measured under this part every trotting stride. A study using surface EMG measurements in
of the saddle at the end of tripedal support (right front, left hind, experienced riders showed that the upper part of m. trapezius has
left front) and during the following (left hind, left front) bipedal strong and distinct activity during the first half of stance to stabilize
ipsilateral support phase (Fig. 15.13). Decreased support of the left the rider’s head and neck and that m. rectus abdominis is active
rear third of the saddle, perhaps combined with the slight increase during the middle and later parts of the stance phase (Terada et al.,
in pressure under the right front third of the saddle following lift-off 2004) to maintain pressurization of the abdomen and stabilize the
of the right forelimb, explains why the rider’s pelvis starts to roll to trunk. The sequence and timing of activation of the other abdomi-
the left at lift-off of the right forelimb. Most riders also retract their nal muscles of the rider have not been evaluated but are likely to
left hip in relation to the right hip (left yaw rotation), this being the be equally, if not more important, than the action of rectus abdomi-
only period in the stride (approximately forelimb lift-off to ground nis, which is more easily accessible to surface EMG. Postural control
contact of the next hind limb) when yaw rotation of the rider’s pelvis is clearly something the rider must learn through training.
does not follow yaw rotation of the saddle. However, at the same Comparing riders of different skill levels, three studies found that
time the horse’s dorsal back muscles in the middle saddle area con- the cranio-caudal movements of the rider’s head or shoulders were
tract strongly on the left side as the left hind limb is retracted (Licka more regular between strides in advanced compared to less experi-
et al., 2009). It is therefore not surprising that left roll and yaw rota- enced riders (Terada, 2000; Peham et al., 2001; Lagarde et al.,
tion of the rider’s pelvis are followed by an increase in pressure under 2005). More efficient postural control in more skilled riders allowed
the left central third of the saddle, peaking at ground contact of the them to maintain their shoulders closer to the vertical at diagonal
right hind limb. Following this peak, roll rotation of the rider’s pelvis midstance whereas the novice riders tilted more forwards (Schils
is abruptly redirected to a right yaw rotation as the rider’s pelvis starts et al., 1993). Comparisons of EMG recordings of two important
to follow the saddle again and right roll rotation continues during postural muscles, m. erector spinae and m. rectus abdominis, and the
the early part of the right hind limb stance. great adductor of the thigh, m. adductor magnus, also confirm differ-
ences in activity pattern between advanced and novice riders
(Terada, 2000). Advanced riders had forceful contractions in both
Canter m. erector spinae and m. rectus abdominis while m. adductor magnus
The canter shows mixed gait mechanics where the trailing hind and was largely inactive. Novice riders, however, had proportionally
leading forelimbs function essentially as rigid struts while the diag- more forceful contractions of m. rectus abdominis compared to m.
onal stance of the leading hind and trailing fore is best likened to erector spinae and m. adductor magnus was markedly more active
a bounce. The trunk of the horse rotates quite extensively around compared to the advanced riders (Terada, 2000). Unbalanced activ-
its transverse axis and the back, particularly the lumbosacral joint, ity between abdominal and epaxial muscles could reflect a less
has the largest range of motion in flexion–extension of all gaits. The efficient postural control in novice riders and relatively greater activ-
rolling motion pattern of the canter is sometimes perceived by the ity of the abdominal muscles could contribute to a more protracted
rider as being easier to sit than the trot. This perception is supported (rounded) shoulder position. Increased activity of the thigh adduc-
by the fact that measurements of the accelerations of the rider’s tors is likely an attempt to compensate for the unstable upper body
helmet showed no differences between advanced and novice level position (Terada, 2000). Although the movements of the horse
riders at canter, which is in contrast to the differences found at walk seem to dictate the basic pattern of the rider’s movements, postural
and sitting trot (Terada, 2000). control is clearly influenced by the rider’s active responses to the
The full stride sequence for rider movements during canter movements of the horse and these responses vary with skill level.
has not yet been documented. However, a preliminary study com- A different study (Pantall et al., 2009) used surface EMG to evalu-
pared rider position at first contact of the trailing hind limb, the ate activation of m. iliocostalis lumborum and m. rectus abdominis in
diagonal limb pair and the leading forelimb. It showed that rising trot. All riders showed coactivation of the right and left
the rider’s trunk was tilted forwards during trailing hind limb muscles. In the novice riders m. rectus abdominis and m. iliocostalis
stance then rotated back towards the vertical during leading fore- lumborum were coactivated whereas in experienced riders there was
limb stance and suspension (Lovett et al., 2005). These trunk a phase shift between them. M. rectus abdominis behaved as an
movements of the rider to follow the rotation of the horse’s trunk agonist in the experienced rider, contracting as the rider made
at canter. contact with the saddle on the sitting diagonal.

353
 15 Horse–rider interaction

The rising trot is often perceived by the novice rider as being less moved the neck caudally in relation to the horse when the horse’s
tiring than sitting trot (German National Equestrian Federation, croup was raised, but the relative contribution of the pelvis and
2002, p. 50). As in sitting trot the skill level of the rider seems to upper body to these movements varied between the riders (Byström
have a significant influence on the rider’s movements. Advanced et al., 2010b). One rider compensated by flexion–extension of the
riders have their shoulders further backwards relative to their hips lumbar back while keeping the neck more stationary, while others
at diagonal midstance both at the sitting and rising phases and they tended to use their entire upper body to balance the movements of
also extend their hip joints more when rising (Schils et al., 1993). the horse (Fig. 15.14). Skill level dependency of upper body move-
ments at walk is supported by the finding that advanced riders had
a more regular pattern in their vertical head movements whereas
novice riders had greater between-stride variability (Terada, 2000).
Walk This was associated with proportionally more concentric activity of
At walk there is more variation, compared to the trot, in the upper the postural muscles m. erector spinae and m. rectus abdominis in the
body movements between individual riders both regarding extent advanced riders while the novices showed a more static activity
and regularity and likely also timing in relation to the horse. It is pattern (Terada, 2000). More skilled riders may exert a more active
suggested that variation is caused by the individual style of the rider postural control leading to a more regular movement pattern and
and/or skill level. In a group of high-level dressage riders all riders smaller phase shift in relation to the horse.

Horse 1 Horse 2
Pitch angle rider's pelvis
10 10

5 5
Angle (º)

0 0

-5 -5

-10 -10
0 50 100 0 50 100
% stride % stride
Flexion–extension rider's back
10 10

5 5
Angle (º)

0 0

-5
-5
0 50 100 0 50 100
% stride % stride

Longitudinal distance rider's neck - horse's L3

40 40

20 20

0 0
Angle (º)

-20 -20

-40 -40

0 50 100 0 50 100
Distance (mm) Distance (mm)
Fig 15.14  Rider’s upper body kinematics in two horses (Horse 1, left panel, Horse 2, right panel) ridden at walk on a treadmill. Data are for one stride starting
at left hind contact. Top row, pitch angle of the rider’s pelvis, middle row, flexion/extension of the rider’s back; bottom row, longitudinal distance from the
rider’s neck to the horse’s third lumbar vertebra (L3). Zero represents mean value for the stride. Solid line is mean, dashed lines indicate one standard
deviation. The lower bars show the stance of the left fore, right fore, left hind and right hind (top to bottom).
From Byström, unpublished.

354
 Rider skill

The rider’s stability can be assessed by tracking movements of the indicating that the rider has a significant influence on the horse–
center of pressure (COP) using a pressure mat beneath the saddle. rider system.
Able-bodied riders have been shown to have smaller amplitudes of Peham et al. (2001) compared a professional with a recreational
COP motion in both the anteroposterior and mediolateral direc- rider riding 20 horses at sitting trot over a sand arena with high-
tions compared with riders with cerebral palsy. COP velocity in the speed cameras recording marker positions. Angular velocities and
mediolateral direction was also significantly smaller in able-bodied accelerations were derived, and resulting vectors in the phase space
riders (Clayton et al., 2010). Although this study was directed were computed. The professional horse–rider system had the most
toward the value of COP measurements for monitoring progress in consistent motion pattern and higher dressage scores (evaluated
a therapeutic riding program, the technique may also be useful for during the experiment), compared to the recreational horse–rider
assessing stability of riders in general. system. Using the same material, motion asymmetries of the head
Rider skill level also seems to influence the rider’s leg movements and sacrum were compared using markers tracked at 120 Hz (Licka
at walk. At forelimb midstance (i.e. approximately at hind limb et al., 2004). The lameness of the unridden horse (assessed visually
contact) advanced riders show less hip flexion but similar shoulder using a dichotomous scale) was not reproduced under either rider.
position compared to novices (Schils et al., 1993). This could be For the dressage rider there was a significant increase in hind-limb
due to a phase-shift between riders at different skill levels or it could asymmetry, compared to the unridden situation. Using partly the
reflect the fact that advanced riders are more exact and efficient in same trials (n = 14 horses), Schöllhorn et al. (2006) used artificial
their movements. Compared to novice riders, advanced riders have neural networks to analyze kinematic parameters over time. The
greater concentric activity in m. adductor magnus at walk (Terada, professional rider was able to control 13 of the 14 horses with
2000), which could indicate a more active control of leg movements respect to the head, while the recreational rider controlled only
as well as upper body movements. However, walking speed was not three horses. On the contrary, hind fetlock and hock movements
presented in these studies (Schils et al., 1993; Terada, 2000) and it were not influenced much by the riders, but were more affected by
is possible that differences in the horse’s stride length and/or stride the inherent motion pattern of the horse.
frequency between rider skill levels contributed to the observed Terada (2000) studied two experienced and two novice riders
differences in rider movement patterns. riding two horses over an 80-m sand arena in the three gaits, with
instructions to use as few aids as possible. A camera recorded the
The hand stride durations, the riders’ head movements were monitored with
an accelerometer, and activity in three of the riders’ muscles were
The equestrian literature often states that the hand should be inde- recorded electromyographically. Differences were found in the walk
pendent from the seat, meaning that the movements of the hand and trot, where it was concluded that novice riders had more dif-
should follow the movements of the horse’s head and mouth, ficulty in stabilizing the body. At sitting trot, novice riders used their
rather than the movements of the rider’s torso. Accordingly, well- m. adductor magnus muscle to maintain their posture because of lack
timed flexion–extension movements of the rider’s shoulder and of coordination between the m. rectus abdominis and m. erector spinae
elbow joints are necessary to compensate for movements of the (Terada, 2000; Terada et al., 2004).
rider’s upper body in relation to the horse. High-level dressage In earlier work, Schamhardt et al. (1991) studied walk and rising
riders indeed do perform such movements (Terada et al., 2006; trot using ground reaction force analysis (x, y and z) in 13 Dutch
Byström et al., 2009) and EMG measurements in experienced riders Warmbloods comparing one experienced rider, one novice rider
indicate that these movements are actively induced by the rider and an equal amount of dead weight (two sandbags). They con-
(Fig. 15.15). cluded that, compared with sandbags, the riders managed to shift
During trotting, in the first half of stance the rider’s neck moves part of the weight towards the hind limbs. However skill of rider
forwards in relation to the horse, the rider’s trunk (hip to shoulder) did not, in general, influence the patterns of the ground reaction
tilts forwards and the distance from the rider’s shoulders to the bit forces. It should be cautioned that velocity was not standardized in
in the horse’s mouth decreases. Simultaneously the rider’s shoulder that study. Another study in which horses moved at the same trot-
and elbow joints flex while the m. biceps brachii and m. deltoideus ting speed with and without a rider confirmed that there was a
muscles are active and m. flexor carpi radialis is also active to flex or reduction in mass normalized peak vertical force with an experi-
stabilize the wrist. During the second half of stance the rider’s neck enced rider (Clayton et al., 1999).
moves backwards in relation to the horse, the rider’s trunk tilts Kinematics of 20 horses were evaluated using high-speed video
backwards and the distance from the rider’s shoulders to the bit to compare three conditions: unridden at trot, ridden at sitting trot
increases. At the same time the rider’s shoulder and elbow joints by a professional rider and ridden at sitting trot by a recreational
extend and m. triceps brachii becomes active while the activity of the rider (Kapaun et al., 1998). Significant differences between the
flexor muscles ceases. These compensatory movements stabilize the unridden situation and the professional rider were found for impul-
riders’ hands at a more constant position in relation to the horse’s sion, pelvic inclination, fore and hind limb vertical movement, and
moving head (Terada et al., 2006) (Fig. 15.16). Vertical displace- fore and hind stance durations. After speed-normalization, only the
ment of the rider’s wrist was significantly smaller than that of the difference in vertical movement of the fore and hind limbs remained
rider’s hips and shoulders and the distance from the rider’s wrist to significant. Comparison between the professional and recreational
the bit varied significantly less than the distance from the rider’s hip rider showed differences in speed, stride length, head angle, hind
to the bit (Terada et al., 2006). A preliminary study comparing one limb protraction, pelvic inclination, fore and hind limb vertical
novice and one expert rider also found that the expert rider had a movement, and fore stance duration. After speed-normalization
significantly lower phase shift between the vertical movements of only impulsion remained significantly different.
the head and shoulder and the wrist and elbow (Lagarde et al., With regard to jumping kinematics, 50 Hz video recordings of 10
2005), further supporting the idea that these arm movements are horses jumping a 1.05-m fence with an experienced and a novice
under the rider’s active control and are learned through training, rider revealed no significant differences in velocity or stride length
rather than induced by the movements of the horse. during the approach, take-off or landing (Powers & Harrison,
2002). It was concluded that horses jumped in the same way,
Horse movement in relation to rider skill regardless of what the riders were doing. However, the authors
cautioned that the similar results could have arisen because the
Compared with an average rider, an expert rider’s movements are horses were riding school horses and not competition horses, the
more consistent and less phase-shifted in relation to the horse’s fence was not very high, the weight of the riders differed between
movements (Terada, 2000; Peham et al., 2001; Lagarde et al., 2005) 12 and 15% of the horse’s weight, the novice rider was possibly too

355
 15 Horse–rider interaction

(%) Rectus Abdominis (%) Flexor Carpi Radialls

100 100
80 80
60 60
40 40
20 20
0 0

Upper Trapezius Extensor Carpi Ulnaris

100 100
80 80
60 60
40 40
20 20
0 0

Middle Trapezius Biceps Brachii

100 100
80 80
60 60
40 40
20 20
0 0

Lower Trapezius Triceps Brachii

100 100
80 80
60 60
40 40
20 20
0 0

Serratus Anterior Middle Deltoid

100 100
80 80
60 60
40 40
20 20
0 0

Teres Major Pectoralis Major

100 100
80 80
60 60
40 40
20 20
0 0
ht1

rly ht1

ds 1

sta 1
rly 1
La ght2

rly ht2

ds 2

sta 2
e2

ht1

rly ht1

ds 1

sta 1
rly 1
La ght2

rly ht2

ds 2

sta 2
e2
Mi nce

e
Ea nce

Mi nce

Mi nce

e
Ea ce

Mi nce

e
La anc

La anc

nc

La anc

La anc

nc
flig

Ea flig

Ea flig

flig

Ea flig

Ea flig
n
fli

fli
sta

sta

sta

sta
t

t
r ly

te

te

r ly

te

te
te

te

te

te
La

La
Ea

Ea

Fig 15.15  Muscular activity for 12 muscles of six subjects over 10 phases of the stride. Each datum point represents mean activity during a phase of the
stride expressed as a percentage of maximal activity for that muscle. Dashed lines indicate one standard deviation. Vertical dotted lines separate the four
consecutive phases of flight 1, stance 1, flight 2 and stance 2.
Reprinted from Terada, K., Mullineaux, D.R., Lanovaz, J., Kato, K., Clayton, H.M., 2004. Electromyographic analysis of the rider’s muscles at trot. Equine Comp. Exerc. Physiol. 1, 193–198, with
permission of Cambridge University Press.

experienced and the riders may not be representative of experienced for turning and the authors concluded that experienced riders can
and novice riders in general. use finer aids.
Warren-Smith et al. (2007) studied rein tension in one profes- Most studies that have compared experienced and novice riders
sional, one intermediate and one novice rider on 22 horses per- relative to the motion of the horse have used only a few riders. The
forming a simple dressage test. The novice rider used more tension experienced riders often have a long riding experience and have

356
 Riding methodology

Bit to wrist Bit to hip extension in both fore and hind limbs increases with a rider (Sloet
Bit to shoulder van Oldruitenborgh-Oosterbaan et al., 1995). Furthermore, the
40 horse’s head is carried in a higher position with a rider, so the neck
is also more extended (Sloet van Oldruitenborgh-Oosterbaan et al.,
1996).
20
Distance (mm)

Load carrying at trot

0
10 20 30 40 50 When a rider is mounted, peak vertical force and impulse increase
in both fore and hind limbs (Schamhardt et al., 1991; Clayton
-20 et al., 1999). However, the relative increase is 50 to 100% higher in
the forelimbs compared with the hind limbs (Schamhardt et al.,
1991; Clayton et al., 1999). Overall, the relative weight bearing on
-40 the forelimbs increases by approximately 1–2% (Weishaupt et al.,
2006; Waldern et al., 2009). The time of maximal vertical force,
Elbow Trunk peak propulsion and the vertical force curve show a shift to the right
Shoulder Forearm in the second half of stance such that the vertical force remains close
10 to maximum for an extended length of time before it starts to
decline (Schamhardt et al., 1991; Clayton et al., 1999). Stride dura-
tion is not influenced by the presence of a rider at trot, but relative
5 stance duration tends to increase (Schamhardt et al., 1991; Sloet
van Oldruitenborgh-Oosterbaan et al., 1995). In the forelimb, the
Angle (º)

fetlock joint is more extended in the second half of stance compared

0 to the unloaded condition, with the fetlock angle–time curve reflect-
10 20 30 40 50 ing the shape of the vertical ground reaction force curve (Sloet van
Oldruitenborgh-Oosterbaan et al., 1995; Clayton et al., 1999).
-5

-10
Rider weight
Time (% stride) Peham and Schobesberger (2004) performed a biomechanical sim-
ulation using a combination of electromyographical data from m.
Fig 15.16  Mean values of linear angular variables in six experienced riders longissimus dorsi, saddle mat data and kinematic data from two back
during half a stride at trot. Distances are measured from the bit to the riders’ markers of 15 horses ridden overground in trot with their customary
wrist, shoulder and hip (top panel) and angulations of the riders’ elbow riders and saddles. They concluded that the forces were higher when
joint, shoulder joint, trunk to vertical (tilting the upper body back is positive) the back was stiffer, compared to the simple rider weight. It is of
and forearm to vertical (moving the elbow forward relative to the shoulder course natural that heavier riders in general yield higher forces and
is positive) (bottom). Ground contact time of the right forelimb is indicated pressures (Geutjens et al., 2008). The effects of rider weight should
by the vertical line at 11% stride. The data have been normalized so the be studied in the dynamic situation to gain a better understanding
mean value during the entire trial is represented as zero for each variable.
of what constitutes unacceptably high pressures from heavyweight
Reprinted from Terada, K., Clayton, H.M., Kato, K., 2006. Stabilization of wrist position during
riders.
horseback riding at trot. Equine Comp. Exerc. Physiol. 3, 179–184, with permission from
Cambridge University Press.
Riding methodology
competed up to high levels, while the novice riders have less riding Numerous texts on riding and equestrianism have been written
experience. However, the experience and ability of the novice riders during the past 500 years but the terminology is often poorly and
likely varies between studies. A general conclusion is that profes- inconsistently defined from both a scientific and practical perspec-
sional riders are more stable than novice riders. tive and terms are sometimes circularly referenced (Roepstorff,
unpublished). Even though riding will always have an artistic
element, especially in dressage, there are also parts of the perfor-
The influence of the rider’s weight mance that should be quantifiable. Measurements of equestrian
technique require definitions that relate equestrian terminology
Learning how to ride involves developing the ability to follow the with measured biomechanical variables.
movements of the horse and to influence the horse’s movements. The training scale, as described in the Official Instruction Handbook
Since the rider’s weight has a marked influence on the horse’s move- of the German National Equestrian Federation (German National
ments, the effects of the rider’s weight will be discussed first. Equestrian Federation, 2002) consists of a series of steps, which are
translated as rhythm, suppleness/relaxation, contact, impulsion,
straightness and collection. These terms will be used as a basis for
Load carrying at walk discussion of biomechanical research.
When the horse carries a rider, the vertical and longitudinal ground
reaction force peaks and impulses increase and the peaks occur Rhythm
significantly later during stance in the forelimb (Schamhardt et al.,
1991). In the hind limb all force peaks increase except peak braking Rhythm, the first requirement in the training scale, is described
force. biomechanically by temporal variables. The stride rate is the number
Stride duration is not influenced by the presence of a rider at of strides per stride interval, also denoted beat, which should not
trot, but relative stance duration tends to increase in both fore be confused with rhythm. The rhythm of the footfalls within each
and hind limbs (Schamhardt et al., 1991). Fetlock joint maximal stride is described by the time intervals between first-contacts of the

357
 15 Horse–rider interaction

individual limbs, such as time of advanced placements and step In a later study, reflective markers were placed bilaterally on the
durations. The same rhythm should be maintained through the facial crests and the lateral aspects of C1 and C3 in six horses, to
variations in gait types (collected, working, medium, extended, calculate the head and neck angles (Zsoldos et al., 2010a). Activity
free). of m. splenius was measured with surface EMG. In trot, functional
The rhythm of the different gaits is described in Chapter 14. The stabilization against flexion of the head and neck was found with
walk should have a regular four-beat rhythm. More irregularities maximal activity of this muscle at the beginning of the forelimb
were found in the medium and the extended walk, compared to the stance phases. Unilateral activity of m. splenius to stabilize against
collected walk (Clayton, 1995), which is in accordance with general lateral movement was not found.
dressage rider perceptions. At trot the regular, two-beat rhythm Surface EMG activities of the m. rectus abdominis and the external
was maintained from collected through working, medium and abdominal oblique muscles and kinematics of the hooves, withers
extended trot though stride duration was significantly shorter in and sacrum were measured at walk and trot in six horses (Zsoldos
the extended trot compared with collected trot (Clayton, 1994b). et al., 2010b). EMG values (minimum, maximum and mean) were
In canter stride duration did not differ significantly between col- significantly higher at trot than at walk in all horses for the external
lected, working, medium and extended canters but the timing abdominal oblique muscle and in five out of six horses for m. rectus
between footfalls changed such that in extended canter the three abdominis. The activities of both muscles differed between left
beats were more closely associated while the time elapsing between and right sides for all horses in walk and for four out of six horses
contact of the leading forelimb and the next contact of the trailing at trot. The ratio of muscle coactivation between the external
hind limb increased (Clayton, 1994a). abdominal oblique muscles and m. rectus abdominis, which was
Elite dressage horses show more irregularities in walk pirouette interpreted as providing trunk stability, was lower at walk than trot.
strides than in canter pirouette strides (Burns & Clayton, 1997; The external abdominal oblique muscles were more clearly acti-
Hodson et al., 1999). Another aspect of rhythm is that the expert vated in an alternating left and right sequence while m. rectus abdom-
rider’s movements have been found to be more consistent and less inis was activated more simultaneously on the left and right sides.
phase-shifted in relation to the horse compared to novice riders In both walk and trot, the activity of m. rectus abdominis is out of
(Terada, 2000; Peham et al., 2001; Lagarde et al., 2005). These find- phase with that of m. longissimus dorsi (Licka et al., 2004; Licka et al.,
ings may explain the greater smoothness in the horse’s gaits when 2009).
ridden by expert riders, often in a more collected form. Using surface electrodes the quotient of the mean of the maximal
and minimal surface EMG activity of m. longissimus dorsi in unrid-
Suppleness/relaxation den horses (n = 15) at walk was judged to be higher at the twelfth
thoracic vertebra (T12) than at the sixteenth thoracic vertebra (T16)
Suppleness and relaxation are prerequisites for all further training or the third lumbar vertebra (L3) (Licka et al., 2009), with some
and, along with rhythm, are an essential aim of the preliminary tension being maintained at the minima. Activity was maximal
training phase. The horse should be free from physical and mental during the ipsilateral hind limb support; this occurred at the begin-
tension with ‘a rhythmically swinging back’ (German National ning of the support phase for L3, followed by T12 and then T16. At
Equestrian Federation, 2002). Even if the rhythm is maintained, the these instances low activity was found in the contralateral muscle.
movement cannot be considered correct unless the muscles are free The authors’ interpretation was that the large activity at T12 com-
from unnecessary tension. pared to T16 and L3 suggested that m. longissimus dorsi is mainly
Suppleness is difficult to measure objectively with biomechanical responsible for stabilising the back during locomotion, which
methods. An approach might be to use electromyography (EMG) agrees with other authors (Robert et al., 2001; Robert et al., 2002;
in relation to kinetic and/or kinematic studies. Biologically, muscles Groesel et al., 2010). In another study it was shown that the func-
must generate tension to support the body and to produce locomo- tion of m. longissimus dorsi may vary at different spinal levels; uni-
tion. In addition, a certain amount of contraction of antagonistic lateral activity predominates in the more cranial segments at the
muscles is needed to stabilize the joints. Thus, it is difficult to evalu- level of the fourteenth thoracic vertebra (T14) during walking,
ate EMG results in relation to ‘necessary’ and ‘unnecessary’ muscular which was interpreted as indicating that the muscle developed
activity, because there are no common and widely accepted defini- lateral bending moments, whereas there were more cocontractions
tions of which muscles should be active at specific periods of the in the caudal segments (Wakeling et al., 2007).
stride cycle in different gaits. During lateral bending, EMG activity in m. longissimus dorsi at
Robert et al. (1998) have published several EMG studies of T12, T16 and L3 was present only on the concave side of the back.
horses. Three horses were ridden in walk, trot and canter and activ- When the back position of standing horses was manually induced
ity of the m. longissimus dorsi was recorded in relation to the stage to extend or bend laterally, EMG amplitudes in m. longissimus dorsi
of the stride cycle determined by accelerometers. The longissimus were highest at T16. During extension there was a temporal correla-
dorsi muscle showed two bursts of activity during each stride at walk tion at the three locations, but less so during lateral bending (Peham
and trot but only one burst during each stride at canter. Muscles on et al., 2001; Wakeling et al., 2007).
the right and left sides acted simultaneously. There was considerable
variation with speed and between-horse variation was also substan-
tial (Robert et al., 1998). In the unridden horse, activity of the
Contact
longissimus dorsi muscle increased both with speed and uphill slope Contact is defined as a soft, steady, connection between the hand
when trotting on a treadmill (n = 4). At faster speed, muscle activity of the rider and the mouth of the horse. The horse should go
began and ended earlier in the stride cycle, while the opposite was forward rhythmically according to the driving aids of the rider and
seen with increasing slope (Robert et al., 2001). seek a contact with the hand of the rider, thus going into the contact
In a study of a larger number of muscles, the basic pattern of (German National Equestrian Federation, 2002).
trunk muscle activation at trot showed two bursts of EMG activity There are no studies directly aiming at defining or validating what
per stride. Activity of m. splenius started just before contact of each contact is. It has been documented that in some rider–horse com-
forelimb and ended in the middle of the corresponding stance binations the contact varies between the left and right sides with
phase. M. longissimus dorsi contracted bilaterally before each hind one hand consistently having a larger proportion of low contacts
limb lifted off. M. rectus abdominis was consistently active as each (Roepstorff, unpublished).
hind foot contacted the ground (Robert et al., 2002). Similar to Studies of rein tension have shown that the rhythmic head and
previous results, increasing speed yielded earlier EMG activity and neck movements of the horse produce a pattern of regular spikes in
increasing treadmill slope postponed it. rein tension when the horses are ridden by experienced riders

358
 Riding methodology

(Clayton et al., 2005) or when side reins are used (Clayton et al., with riding the horse in a vertical head and neck position (HNP2
2011) but not when ridden by novice riders (Heleski et al., 2009). see definition in Fig. 15.19) there was an increase in asymmetry in
It has been interpreted that the baseline tension comes from the a number of kinematic parameters, such as the lumbar spine and
rider or side rein and the spikes are from the head motion of the sacral angle (Rhodin et al., 2009). The roll rotation of the rider’s
horse. A prerequisite to establishing a correct contact is that the rider upper body was markedly asymmetrical between diagonals for
can stabilize the hand position thus encouraging the horse to seek some rider–horse combinations, which is likely to be an indication
a contact through controlled relaxation of the cervical musculature. of rider-laterality (Byström et al., 2009). A study of 17 dressage
Within this baseline contact, inertially driven movements of the riders with varying levels of experience showed that all had a shorter
head and neck result in the rein tension spikes (Clayton, right limb, moved their left shoulders more than the right, and that
unpublished). they were differently asymmetric between right and left canter with
the most inharmonious pattern in the right canter (Symes & Ellis,
2009).
Impulsion
The horse is considered to have impulsion, when the energy created Asymmetry in rising trot
by the hind limbs is transmitted into the gait and into every aspect
of the forward movement (German National Equestrian Federation, At rising trot the load on the horse’s back is lower on the rising
2002). This is another area that has not been specifically addressed diagonal than on the sitting diagonal and, furthermore, both of
with biomechanical methodology. Accelerometers have been used these loads are lower than when the rider sits the trot (de Cocq
to study forward motion of the body parts in relation to fore and et al., 2010a). The same effect is seen in the ground reaction forces
hind quarter impulsion during the use of different types of auxiliary (Schamhardt et al., 1991). Parameters for rising trot, mainly vertical
reins (Biau et al., 2002) (see previous section). peak force and impulse, were higher for the diagonal on which the
riders were sitting. Similarly for riders performing rising trot on a
treadmill force loading was generally increased in the limbs of the
Straightness sitting diagonal (Roepstorff et al., 2009). The following asymme-
tries were found: the lumbar spine was lower between midstances
A horse is considered straight when the hind quarters follow the of the sitting and rising stance phases; pelvic roll was limited
same track as the forehand and the longitudinal axis of the body is and heights of the tubera coxarum were lower on the sitting side
aligned with the straight or curved track that the horse moves along. (Fig. 15.17); maximal hind limb protraction was decreased and
The weight should be evenly distributed between the left and right forelimb retraction was increased; and T6 height decreased. The
sides of the body (German National Equestrian Federation, 2002). conclusion was that the rising and sitting motion of the rider
However, every horse and rider can be assumed to be born with induced an uneven biphasic load that affected the horse’s back,
some kind of sidedness. Sidedness has been studied quite exten- pelvis and limb kinematics and the vertical ground reaction force.
sively in humans but to a much lesser extent in horses. The interest- It should be noted that phase-shifting in the rising trot differed
ing question is how to straighten the horse and correct asymmetrical between the left and right diagonals, which provides further evi-
locomotor patterns associated with sidedness. A further important dence that inherent or rider-induced asymmetry was present in
consideration is that asymmetry in the movement pattern is the these top-level dressage horses.
basic concept of the lameness work up. From the diagnostic point Both genetic and environmental factors have been implicated
of view there is a gray area between normal physiological responses in equine inter-stride asymmetry (motor laterality, sidedness)
(e.g. sidedness inherent in the horse or induced by the rider’s sided- (Murphy & Arkins, 2008). There are likely two aspects of laterality,
ness), and asymmetry in response to painful processes (lameness). one defined by ethologists (Murphy & Arkins, 2008) and another,
Asymmetry directly related to rider influence has been reported in possibly a different phenomenon, studied by biomechanists. From
several papers and has been elucidated in the section on rider skills. the rider’s perspective, the biomechanical asymmetry is commonly
Alignment of the forehand with the hind quarters has not been described as the neck muscles being more contracted on the ‘stiff’
scientifically addressed. A number of kinematic studies have side, compared to the other ‘hollow’ side. The ‘hollow’ side has
reported on different aspects of asymmetry. In the following we the haunches positioned ipsilaterally relative to the forehand
discuss the most relevant from a horse–rider interaction perspective. (German National Equestrian Federation, 2002). Any intra-stride
Studies targeting lameness are addressed in Chapter 9. asymmetry might encompass motor laterality and/or training-
induced asymmetry.
In a study of seven horses, all were asymmetric to the same side
Asymmetry basics (Roepstorff et al., 2009). Regardless of whether the riders were
In trot ridden overground, Clayton determined that three of six rising on the left or right diagonal, the left fore/right hind diagonal
highly trained dressage horses in collected, working, medium and was relatively more loaded than the opposite diagonal. The more
extended trot were asymmetrical (Clayton, 1994). This included a loaded left fore and right hind limbs were likely placed further
longer suspension phase (on the right one horse), longer diagonal under the horse. The forward placement of the right hind limb
distance (on the right one horse), and longer diagonal advanced might indicate unloading of the left hind limb (German National
completion times (right and left, one horse each). The fact that riders Equestrian Federation, 2002). This is generally accepted as the horse
often perceive one side as easier for a canter transition, or that most being stiff to the left, which is perceived as the more common side
horses have a preferred canter lead when unridden (unpublished (German National Equestrian Federation, 2002).
observations) points to the presence of asymmetries between the two In conclusion, straightness and related expected symmetry in
leads at canter, though the effect of the rider on this has not yet locomotion and force patterns are extremely important to study
been evaluated. In a study of the kinematics of high level dressage further and in more depth. This is definitely an important key to
horses and expert riders on a treadmill, some movements of the successful performance and in keeping the horse sound. However,
riders and saddles were more asymmetric than expected. The study these two concepts are also heavily interrelated by the fact that it is
design could not distinguish between asymmetries caused by lateral- hard to compete or perform with an unsound horse. In mounted
ity of the horse or rider (Byström et al., 2009). studies it is important to realize and take into account the fact that
The asymmetry during riding with various head and neck posi- the rider’s asymmetry affects the laterality of the horse.
tions (see below) was larger when the rider tried to affect the horse It may be asked whether there is correlation between forces
more. For example, when trotting on a loose rein was compared exerted on horseback and kinematic evaluation of rising trot

359
 15 Horse–rider interaction

Left tuber coxae rising diagonal. Another study with 12 horses ridden overground
by one intermediate rider found an increased range of flexion–
extension in the back in rising trot compared to sitting trot
80
(de Cocq et al., 2009). Taken together, there does not seem to be a
60 correlation between increased range of forces and increased range
of motion in back flexion–extension. Whether it is good for the
40 horse to have an increased range of motion of the back remains
Relative height (mm)

20 to be seen.

0
Collection
-20

-40 General
-60 The aim of all gymnastic training is to create a horse that is athletic
and ready and willing to perform. For the horse to meet these cri-
-80 teria, its weight, plus that of its rider, must be distributed as evenly
as possible over all four limbs. This means reducing the weight on
0 10 20 30 40 50 60 70 80 90 100
the forelimbs (which naturally carry more of the load than the hind
% stride
limbs), and increasing the weight on the hind limbs, which were
originally intended mainly for creating the forward movement.
Equestrian texts indicate that in collection the hind limbs (the hock
Right tuber coxae
and stifle joints) flex more, stepping further underneath the horse
in the direction of the center of gravity, thereby taking a greater
60 share of the load. This in turn lightens the forehand, allowing the
forelimbs to move more freely. The horse looks more ‘uphill’. The
40 strides become shorter, but without losing the energy or activity.
The impulsion is maintained in full at the trot and canter, and as a
20
result the strides become more expressive and ‘stately’ (German
Relative height (mm)

0 National Equestrian Federation, 2002).

The role of the forelimbs is to support rather than to push and
-20 they can only be strengthened to a very limited degree by training.
It is therefore more sensible, and indeed necessary, to transfer some
-40 of the weight to the hind quarters. The increased flexion of the hind
limbs results in raising of the neck. When the carrying capacity of
-60 the hind quarters is sufficiently developed, the horse is able to move
in balance and self-carriage in all three gaits (German National
-80
Equestrian Federation, 2002).
-100 Collection can be defined by several terms, which can be studied
0 10 20 30 40 50 60 70 80 90 100 biomechanically. These definitions will be described, relevant find-
% stride ings from research so far elucidated and suggestions provided for
what can be done to further study the subject of collection.
Fig 15.17  Stride-cycle parameters for kinematic variables in left (solid line)
As already stated many of the terms are poorly defined. For
and right (dashed line) rising trot (in left rising trot the rider sits when the
left fore and right hind limb are grounded). The upper bar shows when at example, ‘increased weight bearing by the hind limbs’ can be deter-
least five consecutive points were statistically significant at p < 0.05. The mined using appropriate force measuring methodology, while ‘low-
lower bars show the stance of the left fore, right fore, left hind and right ering of the back’ is a more subtle poorly defined concept, therefore
hind (top to bottom). also more difficult to set up a measuring strategy accepted by riders,
Reprinted from Roepstorff, L., Egenvall, A., Rhodin, M., Byström, A., Johnston, C., van Weeren, trainers or peers.
P.R., Weishaupt, M.A., 2009. Kinetics and kinematics of the horse comparing left and right It is important to remember that collected gaits are performed at
rising trot. Equine Vet. J. 41 (Suppl.), 292–296, with permission of the Equine Veterinary slower speeds than the same gaits with less collection. Another
Journal.. confounding issue relates to differences in the kinematic patterns
between overground and treadmill locomotion.
In this section the concepts and the scientific background will be
described for kinematic variables that are frequently used to describe
collection by riders and trainers; forward stepping of the hind limbs
compared to sitting trot. A treadmill study using 10 horses ridden relative to the hind quarters (hind limb protraction–retraction and
by one experienced rider with correctly fitted saddles, found that overreach distance), lowering of the hind quarters and increased
the range of overall saddle mat force was lower in rising trot com- joint flexions, diagonal displacement and relative weight bearing.
pared to sitting trot (Peham et al., 2009). This means that the force First, the biomechanical effect of using various head and neck posi-
at rising trot varies less than at sitting trot, i.e. the rising trot evens tions (HNP) will be described. A correct HNP is thought by many
out the forces acting between the saddle and the horse’s back that to be a prerequisite for collection. Accordingly, large efforts are put
are caused by the vertical oscillation of the horse back. This was into attaining a desired HNP during riding. Both novice and expe-
confirmed by de Cocq et al. (2010a) who used rider kinematics to rienced riders frequently fixate on HNP, perhaps as a consequence
model saddle reaction forces in sitting and rising trot. The results of the visual ease of determining the HNP, which is also relatively
showed that mean force was always equal to rider weight but straightforward for judges to assess. In the following description,
peak forces, which occurred in diagonal midstance, were symmetri- HNPs in relation to the frame of the horse will be discussed using
cal in sitting trot but differed between the two diagonals in rising terms defined by the German National Equestrian Federation
trot with significantly higher values on the sitting diagonal than the (2002).

360
 Collection

• HNP3: Neck raised, poll high and bridge of the nose slightly
behind the vertical (behind).
• HNP4: Neck lowered and flexed, bridge of the nose
considerably behind the vertical (low: to imitate Rollkür,
neck low).
• HNP5: Neck extremely elevated and bridge of the nose
considerably in front of the vertical (above; absolute
elevation).
• HNP6: Neck and head extended forward and downward (low
and open; ridden, only performed at walk, not shown in
HNP1 HNP2 Fig. 15.19).

Walk
Kinetic evaluation in the different HNPs at walk showed that, in
general, being ‘above’ the bit or ‘behind’ the vertical differed from
the vertical position in the same direction, while having the neck
‘free’ or ‘low and open’ differed in the opposite direction (Weishaupt
et al., 2006) (Table 15.2).
Comparing the ‘competition’ HNP with the other HNPs several
kinematic parameters were found to be different. The sixth thoracic
HNP3 HNP4 vertebra (T6) height was affected during three-limb support by one
forelimb and two hind limbs, except for in the low and open HNP
when it was increased throughout the stride. In the free and above
positions there was a tendency for L5 to be higher during three-limb
support by one forelimb and two hind limbs. In the behind and
low and open HNPs the height of L5 increased during the second
half of hind limb stance. The lumbar back angle (the ‘small’ angle
between a horizontal plane and a line through L5 to L3) increased
in the ‘behind’, ‘neck low’ and ‘above’ positions and decreased in
the free position with the peak value coinciding with three-limb
support with one hind and two forelimbs.
In the free and above positions, pelvis pitch increased (tilted
HNP5 HNP6 forward in the free position) and decreased (tilted backward in the
above position) respectively, from forelimb three-limb support to
Fig 15.18  The head and neck positions used in the Zurich study. See text hind limb three-limb support. In the behind position a decrease in
for description of positions. pelvic pitch was found throughout the stride and in the low and
Reprinted from Rhodin, M. et al., 2009. The effect of different head and neck positions on the open pelvic position pitch decreased at three-limb support with one
caudal back and hindlimb kinematics in the elite dressage horse at trot. Equine Vet. J. 41 (3), forelimb.
274–279, with permission from the Equine Veterinary Journal. In the free and above HNPs the femur angle was increased in the
later part of stance in the free position and decreased in early swing
in the above position, relative to the competition HNP. In the
Head and neck positions ‘behind’ HNP it was smaller at the end of stance.
In the free position the stifle joint was more flexed during the
In 2005 an experiment was conducted with seven advanced-level
later part of stance. In the behind position the stifle was more flexed
horses ridden by their own riders on a treadmill (Gomez Alvarez
during diagonal stance and following three-limb support with two
et al., 2006; Weishaupt et al., 2006; Rhodin, 2008; Rhodin et al.,
hind limbs. With a low neck it was more flexed during midswing.
2009). Vertical ground reaction forces were measured by the instru-
In the above position the stifle joint was more flexed during diago-
mented treadmill and kinematic analysis was performed with a
nal support, less flexed during ipsilateral support and more flexed
high-speed video system tracking 80 markers distributed over ana-
during the first half of swing, indicating a greater ROM (Rhodin,
tomical landmarks on the horse, rider and saddle. Forces between
2008). Figure 15.19 compares the hind limb position for the ‘com-
the saddle and the horse’s back were measured with a pressure sensi-
petition’ position to the free and above positions. Generally, ROM
tive saddle mat and, in three horses, rein forces were measured.
increased in the low positions implying that these are good for
The horses were ridden at walk and trot with six different HNPs
gymnastic effects on the horse, i.e. increasing suppleness.
(Fig. 15.18). A correct HNP was assumed to be one result, or a
prerequisite of, collection. The vertical ‘competition’ HNP was used
for comparison and therefore a series of recordings at a range of Trot
speeds were made in this position to enable speed-matched com-
The horses shifted weight to the forequarters in the low position and
parisons with the other positions. All of these trials were also per-
a number of kinetic parameters were affected in the ‘above’ position
formed without the rider achieving the head neck positions with
(Table 15.3) (Weishaupt et al., 2006). Kinematically, in the ‘free’
different kinds of auxiliary reins. Finally also passage trials were
position sacral flexion (pelvic pitch) decreased during suspension
performed. All in all approximately 240 trials were recorded during
and early stance and in the ‘above’ position the sacrum was signifi-
a 2-week period:
cantly more angled during breakover and suspension (Table 15.4).
• HNP1: Free or natural; voluntarily acquired position, In the free position the femur angle was larger during breakover and
unrestrained with loose reins (free). early swing whereas in the above position it was smaller in early and
• HNP2: Neck raised, poll high and bridge of the nose slightly late swing. In the free position the stifle flexed more slowly in early
in front of the vertical; reference position (competition, the swing and in the above position the stifle showed increased flexion
Ramener). during stance and the first two-thirds of swing (Rhodin et al., 2009).

361
 15 Horse–rider interaction

Table 15.2  The directions of the statistically significant differences found for kinetic parameters comparing the competition position to
various head and neck positions in ridden walk. One arrow, percentage changes <5%; two arrows, percentage changes >5%

Parameters Head and neck position

Free Behind Neck, low Above Low, open
Stride duration ↑ ↓ ↓↓
Stride impulse ↑ ↓ ↓↓
Stride length ↑ ↓ ↓↓
First forelimb force peak ↑↑ ↓
Second forelimb force peak ↑ ↑
First hind limb force peak ↑↑ ↑
Second hind limb force peak ↓ ↓
Percentage of stride impulse carried by the forehand ↑ ↓ ↑
Forelimb duty factor ↓ ↑
Hind limb duty factor ↓ ↑ ↑ ↓
Ipsilateral step duration relative to stride duration ↑↑
Diagonal step duration relative to stride duration ↓↓
Overreach distance ↑↑ ↓↓ ↓↓

Adapted and reprinted from Weishaupt, M.A., Wiestner, T., von Peinen, K., et al., 2006. Effect of head and neck position on vertical ground reaction forces and interlimb
coordination in the dressage horse ridden at walk and trot on a treadmill. Equine Vet. J. 36 (Suppl.), 387–392, with permission from the Equine Veterinary Journal.

HNP2 versus HNP1 HNP2 versus HNP5

1400 1400

1200 1200

1000 1000
mm

mm

800 800

600 600

400 400

200 200
-1000 -800 -600 -400 -200 0 -800 -600 -400 -200 0 200
mm mm
Fig 15.19  Stick figures illustrating the hind limb position during walking at first contact, midstance and toe-off for HNP2 (competition position) (blue line),
HNP1 (free position) (red line on left) and HNP5 (above position) (red line on right). Markers from proximal to distal; greater trochanter of the hip joint, stifle
joint, hock joint, fetlock joint and hind hoof.
Reprinted from Rhodin, M., 2008. A biomechanical analysis of relationship between the head and neck position, vertebral column and limbs in the horse at walk and trot. Dissertation,
Uppsala., with kind permission of Marie Rhodin.

Data from the same horses were studied without a rider in walk In this study the highest neck position had the largest weight
and trot to study back kinematics that are impossible to interpret shift to the hind quarters (Weishaupt et  al., 2006). This might be
fully with a rider. The comparison base was the free HNP. In the interpreted as collection (being one definition of collection).
‘competition’, behind and above HNPs, the back was more extended However, the front limb locomotor pattern changed with stance
in the anterior thoracic region and the ROM in flexion–extension being shortened to such an extent that, even though weight was
was reduced in the lumbar region (Gomez Alvarez et al., 2006). For shifted to the hind quarters, the maximal peak force in the fore-
the low neck HNP the pattern was opposite. In the above HNP the limb was increased. The forelimb stride simply became very
intravertebral pattern symmetry was negatively affected and hind short and sharp, which is not desired from an equestrian
limb protraction was reduced. perspective.

362
 Collection

Table 15.3  The directions of the statistically significant differences found for kinetic parameters comparing the competition position to
various head and neck positions in ridden trot as well as passage (using the competition position). One arrow, percentage changes <5%;
two arrows, percentage changes >5%

Parameters Head and neck position

Free Behind Neck, low Above Passage
Stride duration ↑ ↑ ↑↑
Stride impulse ↑ ↑ ↑↑
Stride length ↑ ↓↓
Forelimb force peak ↓ ↑↑
Hind limb force peak ↑
Proportion of impulse carried by forehand ↓ ↓
Suspension duration relative to stride duration ↓↓ ↑↑
Forelimb duty factor ↑ ↓↓
Hind limb duty factor ↓
Diagonal dissociation 1
↑↑2 ↑↑2
Diagonal advanced completion relative to stride duration ↑↑ ↓↓ ↓↓ ↑↑
Over-track distance ↓↓ 3
n.a.4
1
Also called ‘diagonal advanced placement’.
2
Mean value positive.
3
Negative, but less negative compared to competition position.
4
Not applicable.
Adapted and reprinted from Weishaupt, M.A., Wiestner, T., von Peinen, K., et al., 2006. Effect of head and neck position on vertical ground reaction forces and interlimb
coordination in the dressage horse ridden at walk and trot on a treadmill. Equine Vet. J. 36 (Suppl.), 387–392 and Weishaupt, M.A., Byström, A., von Peinen, K., et al., 2009.
Kinetics and kinematics of the passage. Equine Vet. J. 41, 263–267, with permission from the Equine Veterinary Journal.

The collected body orientation achieved in this way at the considered more demanding. Classically, it is said that the horse
passage redistributed the diagonal vertical impulse to the hindquar- must not drop the bit and work behind the vertical. Approximating
ters. The shift amounted to approximately 3-fold of that which was the behind position to such a state, rather few changes were found
observed in the same horses trotting in an extremely high elevated compared to the competition position and only in walk. Interest-
head-neck position (Weishaupt et al., 2006). However, because of ingly taking the head ‘back (behind)’ or ‘down’ (rollkur) had almost
the longer absolute stance durations, the higher limb impulses at the same main effects, i.e. the measurements of back height
the passage did not result in higher peak forces, in either the fore- or increased and pelvic pitch and yaw decreased. The above the bit
hind limbs. In this respect the passage seems to be less of a strain position had the largest effects on the horses and from the perspec-
for the distal limbs compared to horses moving with the head-neck tive of collection, the horse actually shifted the vertical stride
in a high elevated position in which, despite the redistribution of impulse to the hind quarters in both walk and trot. However, force
the vertical impulse to the rear, peak vertical forces were increased peaks increased in the forelimbs in walk and in both fore and hind
in the forelimbs due to shortened stance durations (Weishaupt limbs in trot. Such force peaks may be detrimental to, for example,
et al., 2006). riding school horses, which often trot with their heads in a high
This observation leads us to the conclusion that correct collection position. In walk, the total forces are lower than at trot and likely
of the horse’s movement is characterized by relative elevation in to be less harmful. Rather few riders voluntarily strive for ‘above’
the forehand matched to the degree of ‘Hankenbeugung’ which HNP, with the exception of piaffe, where pictures of old masters are
includes increased flexion at the lumbosacral junction and of the often shown with the nose in front of the vertical (Karl, 2008). A
proximal limb joints (hip, stifle, hock) during weight-bearing. few elite horses took part in the highly controlled treadmill experi-
Whereas an absolute elevation of the head and neck with the back ment and whether the results can be generalized extensively to more
held in extension (Rhodin et al., 2009) and with shortened over- horses of various levels ridden overground remains to be studied.
track distance and consequently insufficient engagement of the In another study of seven horses lunged in various HNPs it was
hind legs forward under the body (Waldern et al., 2009; Weishaupt found that the below position (in this text translated to rollkur) had
et al., 2006) results in an overload situation with increased peak the highest rise in serum lactate dehydrogenase and most increases
vertical forces (Weishaupt et al., 2006). in variables measuring neuromuscular functionality, which was
In conclusion, as anticipated by riders the HNPs were shown to measured by single muscle fibre potentials and motor unit action
have effects on the horses. Among dressage riders one of these posi- potentials (Wijnberg et al., 2010). Artificial positions in general
tions, the competition position, could be considered as an ideal, affected neuromuscular functionality more than the free position.
i.e. as a way to collection. However, most riders consciously vary Further, the interaction between the rider and the horse when
the neck and head positions and thereby increase the range of placing the head of the horse needs elucidation also from the per-
motion of several body parts, to exert a gymnastic effect on the spective of how much self-carriage the horse is in (whether leaning
horse, for example during warm-up, compared to the work on the reins) to non-biomechanical aspects, including horse

363
 15 Horse–rider interaction

Table 15.4  The directions of the statistically significant differences found for kinematic parameters comparing the competition position to
various head and neck positions in ridden trot as well as passage (with the competition position)

Parameters Head and neck position

Free Behind Neck, low Above Passage
Maximum T6 height ↓ ↑ ↑
Minimum T6 height ↓
Stride ROM for T6 height ↓ ↑ ↑
Maximum L5 height ↑
Minimum L5 height ↓ ↓
Stride ROM for L5 height ↑ ↑
Lumbar back angle at midstance (horizontal L5 to L3) ↓ ↑ n.a.3
Sacral flexion ↓ ↑ ↑4
Femur angle (see text) ↑ ↓ n.a.
Stifle joint angle (see text) ↑ ↓ ↓
Stride retraction forelimb1 ↑ ↓ ↓ ↓
Stride protraction forelimb ↓ ↑
Stance retraction forelimb ↓ n.a.
Stance protraction forelimb ↓ ↓ n.a.
Stride retraction hind limb 2
↑ ↓
Stride protraction hind limb ↓
1
Forelimb retraction/protraction measured as the horizontal angle through the elbow joint, to the fore hoof, except for in passage when it is evaluated from T6.
2
Hind limb retraction/protraction measured as the horizontal angle through L5, to the hind hoof.
3
Not applicable.
4
Increased flexion of the lumbosacral junction throughout the stride.
Adapted and reprinted from Rhodin, M., Gomez Álvarez, C.B., Byström, A., et al., 2009. The effect of different head and neck positions on the caudal back and hindlimb
kinematics in the elite dressage horse at trot. Equine Vet. J. 41, 274–279 and Weishaupt, M.A., Byström, A., von Peinen, K., et al., 2009. Kinetics and kinematics of the passage.
Equine Vet. J. 41, 263–267, with permission from the Equine Veterinary Journal.

behavior (in which studies have been done), human perception and when trotting unridden on a treadmill, hind limb protraction was
horse longevity. reduced in the above HNP (Gomez Alvarez et al., 2006).

Forward stepping of the hind limbs relative to the Over-track distance

hind quarters Over-track distance (also defined as over-reach) is a well-defined
biomechanical concept, i.e. the distance that a hind limb steps in
High-level dressage horses ridden overground do not step further front of the ipsilateral forelimb. It varies with speed of the gaits
under themselves as collection increases at trot, or even in passage within an individual and is affected by conformation.
and piaffe (Holmström et al., 1995). However, it can be argued that A related but slightly different variable is the position of the fore
an increased level of collection simply implies a lower speed which and hind limbs compared to each other during late swing phase,
means shorter strides and that this speed effect obscures a relative which can be measured by determining the minimum distance
and perceived forward stepping of the hind limbs. Since it is difficult between the contralateral fore/hind limbs in late swing.
to study collection independent of speed, alternative solutions that The mean over-track distance in overground walk decreased with
correct for the speed effect may be necessary. The following sections collection and was negative (mean, −7 cm) in collected walk, 19 cm
compare different studies specifically related to hind-limb at medium walk and 27 cm at extended walk (Clayton, 1995). In
protraction–retraction or over-track distance. walk on a treadmill the over-tracking distance relative to stride
length was significantly larger by 34% in the free HNPs, compared
Hind limb protraction–retraction? to the competition HNP (Weishaupt et al., 2006).
The hind limb protraction–retraction angle is usually measured as The mean over-track distance in overground trot decreased with
the angle between a line connecting the hoof to L5 and the hori- collection and was negative (mean -7 cm) in collected trot (Clayton,
zontal plane in front of the limb. In horses walking overground 1994b), 4 cm in working trot and 39 cm in extended trot. In trot on
fore- and hind-limb retraction decreased in the competition HNP a treadmill the over-tracking mean distances both for free and col-
compared to the free HNP (Rhodin et al., 2009). In trot on a tread- lected trot were negative (i.e. within a stride the hind hooves did not
mill, hind-limb retraction decreased after toe-off in the competition reach the imprints of the ipsilateral fore hooves (Weishaupt et al.,
HNP compared with the free HNP (Rhodin et al., 2009) and was 2006). The difference in results has been attributed to differences
further reduced in passage (Weishaupt et al., 2009). Additionally, between overground and treadmill locomotion.

364
 Collection

Lowering of the hind quarters, and increased 1.2

hind joint angulation
1.0
A common term is that the horse lowers its haunches. The anatomi-
cal manifestations of this term could theoretically translate to
several biomechanically measurable features, including increased 0.8
flexion of the lumbar spine, increased inclination of the pelvis
(pitch), and increased angulations of the hind-limb joints. 0.6

Force
At walk and trot, the metacarpus and metatarsus had a more verti-
cal orientation both at contact and at lift-off in the more collected 0.4
gaits. In other words, the limb rotated through a smaller angle
during stance with less forward progression of the body mass in the
0.2
more collected gaits. However, metacarpal/metatarsal angulations
were generally more acute in the hind limbs than the forelimbs as
a consequence of conformational differences (Clayton, 1994b, 0.0
1995). In walk when comparing the competition HNP to the free -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6
HNP, pelvic pitch decreased in the competition HNP, the height of Time
L5 during three-limb support with two hind limbs increased, the
Fig 15.20  A schematic drawing of vertical ground reaction force over time,
lumbar back angle increased, T6 height during three-limb support
with the half-sine shape that is observed in running gaits. The highest point
with two forelimbs decreased, and the femoral angle in late stance
is the peak force and the area under the curve is the impulse. Note that the
and early swing decreased (Rhodin, 2008). The stifle joint was less three traces have equal impulses, but peak vertical forces are inversely
flexed during late stance. When trotting on a treadmill sacral flexion related to the stance durations according to Fpeak ~ 1/D if the shape of the
(pelvic pitch) was increased in the competition HNP compared force curve remains similar.
with the free HNP (Rhodin et al., 2009).
Piaffe has more pelvic inclination during the entire stride cycle
compared to collected trot and passage (Holmström et al., 1995). Commonly the total impulse of the fore/hind limbs is determined,
Furthermore, the stifle and hock joints were more flexed at the start which can be done with a force-measuring horseshoe, force plate
of stance and the hock joint was also more flexed in midstance in or an instrumented treadmill. However, it is important to distin-
passage and piaffe compared to collected trot (Holmström et al., guish between impulse (total weight bearing of one limb during
1995). When comparing the passage with the collected trot, the one stance phase) and peak force (Fig. 15.20).
horses moved at a slower forward velocity (−43.2%) and with a
lower stride frequency (cadence; −23.6%), with a more elevated
HNP and an overall pronounced upwards excursion of the trunk Absolute versus relative weight
during the suspension phase. The strides were more elevated result- The weight distribution of an unridden horse trotting freely over-
ing from reduced limb pro- and retraction and increased limb ground is 55/45% on the fore/hind limbs for (Schamhardt et al.,
flexion at the carpus, stifle and tarsus during the swing phase. 1991), 56/44% (Weishaupt et al., 2004), and 58/42–44% (Waldern
However, the prolonged suspension duration could not be con- et al., 2009) when trotting on a treadmill, and 59/41% at walk on
firmed. Also Clayton (1997) observed a gradual shortening of the a treadmill (Waldern et al., 2009). The presence of a rider sitting
airborne phase from collected trot to passage. passively, leads to a quotient of 52/48%.
The greater reliance for hind limb support, indicating a higher From the free HNP to the HNPs in which most riders need to use
degree of self-carriage and balance, was reflected in prolonged their reins, in walk there is a small increase in weight bearing of the
stance duration and positive diagonal dissociation. Similar changes hind quarters that is statistically significant for the ‘above’ HNP
were described by Clayton (1994, 1997) and Holmström et al. (vertical impulse of the forelimbs decreased by 1.6% in walk and
(1995) for the over-ground situation in which the passage stride 1.8% in trot). However, this position was judged to be uncomfort-
duration and fore- and hind limb stance duration increased while able for the horse (Weishaupt et al., 2006). In the high HNPs,
speed and stride length were reduced. 2.5–3% less vertical impulse was carried by the forehand compared
It seems difficult to prove biomechanically that lowering of the with the free HNP (Weishaupt et al., 2006). In passage the vertical
hind quarters is a consistent finding in collection. One likely expla- impulse carried by the forehand was decreased by 4.8% compared
nation is that the human eye confuses movement during the swing to collected trot (Weishaupt et al., 2009).
phase with temporal characteristics of limb placement. Weishaupt et al. (2006) confirmed that, indeed, relative shifts
of the center of mass can be induced by altering the HNP of the
horse. The HNPs with the neck stretched out (HNP1 and HNP6)
Diagonal dissociation shifted the weight towards the forelimbs; more restrained HNPs and
In speed-matched trots on a treadmill, significant decreases in diag- especially the elevated neck position (HNP5) shifted the weight
onal dissociation relative to stride duration were found in the com- towards the hind quarters in both gaits. However, compared to the
petition HNP compared to the free HNP (Weishaupt et al., 2006), reference position HNP2, the changes were surprisingly small; the
and the mean value was negative, meaning that the fore hooves cranio-caudal load distribution varied only between +0.7% and
made contact before the diagonal hind hooves. Diagonal dissocia- −1.0% at walk and +0.5% and −1.0% at the trot
tion was further significantly longer on the treadmill (Weishaupt A further essential finding was to quantitatively demonstrate that
et al., 2009) in passage compared to collected trot. In this study a load shift towards certain limbs would not necessarily increase
speed-matching was not done. the peak forces in those limbs and vice versa. In the free HNP and
the low and open HNP peak vertical forces (Fzpeak) in the forelimbs
decreased although the center of mass was shifted towards the
Relative weight bearing forehand. In a single peak, sinusoidal curve such as the vertical
ground reaction force curve of the trot, maximal values can be
Increased weight bearing by the hind quarters is another common reduced by distributing the impulse (area under the curve) over a
term used to describe collection. Although it is easy to define, it is longer stance duration, which was the case for the free HNP and
not so easy to measure comprehensively in the practical situation. the low and open HNP. The opposite was observed in the above

365
 15 Horse–rider interaction

HNP where the shift of load from the forehand to the hind quarters association with collection, i.e. the impulse may be shifted to earlier
was not associated with a reduction of Fzpeak in the forelimbs. On within the stride.
the contrary, because of the pronounced vertical stride pattern of Trainers use terms such as ‘letting the aids through’ and ‘working
the forelimbs coupled with unproportionally short stance dura- through the back’ that are related to the feeling of the horse being
tions, Fzpeak actually increased compared to the speed-matched refer- ‘between the aids’ when the rider perceives that it is possible to do
ence. In movement patterns that have a short stance duration, whatever he or she wishes without any resistance from the horse.
higher peak forces are to be expected! Most riders with a certain amount of experience can identify a
Conflicting results between different studies may have arisen specific feeling when the horse and rider seem to become one unit,
because the changes being measured are so small or they may be and the feeling is that every aid is accepted with ease by the horse.
confounded by the use of a treadmill. Changes in timing of the It is likely that this is related to timing between fore- and hind-
force applications by the fore and hind limbs may also be important quarter locomotion and temporal coordination of back muscle
in achieving collection. As an example the vertical force in the hind activity (Robert et al., 1998; Licka et al., 2009). However, this field
limb might increase faster during the first part of stance in is yet to be explored.

References
Barr, A.R., Dow, S.M., Goodship, A.E., 1995. medium and extended walks in horses. evaluate asymmetrical loading of the
Parameters of forelimb ground reaction Am. J. Vet. Res. 56, 849–852. back of the horse and different riders
force in 48 normal ponies. Vet. Rec. 136, Clayton, H.M., 1997. Classification of positions on a standing horse. Vet. J. 181,
283–286. collected trot, passage and piaffe based on 266–273.
Beek, F., van, de Cocq, P., Muller, J.A., van temporal variables. Equine Vet. J. 23 De Cocq, P., Dunker, A.A., Clayton, H.M.,
Leeuwen, J.L., 2008. Stirrup force (Suppl.), 54–57. Bobbert, F.M., Muller, M., van Leeuwen,
measurement during equine locomotion. Clayton, H.M., Kaiser, L.J., de Pue, B., Kaiser, J.L., 2010a. Vertical forces on the horse’s
Abstracts/Comparative Biochemistry and L., 2010. Center of pressure movements back in sitting and rising trot. J. Biomech.
Physiology Part A 150, S74–S92. during horseback riding: A comparison 43, 627–631.
Biau, S., Couve, O., Lemaire, S., Barrey, E., between able-bodied and disabled riders. De Cocq, P., Mooren, M., Dortmans, A., van
2002. The effect of reins on kinetic Am. J. Occup. Ther. 65, 211–216. Weeren, P.R., Timmermans, M., Muller, M.,
variables of locomotion. Equine Exerc. Clayton, H.M., Lanovaz, J.L., Schamhardt, et al., 2010b. Saddle and leg forces during
Physiol. 34 (Suppl.), 359–362. H.C., van Wessum, R., 1999. The effects of lateral movements in dressage. Equine Vet.
Buchner, H.H., Obermuller, S., Scheidl, M., a rider’s mass effects on ground reaction J. 42 (Suppl.), 644–649.
2000. Body centre of mass movement in forces and fetlock kinematics at the trot. De Cocq, P., Prinsen, H., Springer, N.C.N., van
the sound horse. Vet. J. 160, 225–234. Equine Vet. J. 30 (Suppl.), 218–221. Weeren, P.R., Schreuder, M., Muller, M., et
Burns, T.E., Clayton, H.M., 1997. Comparison Clayton, H.M., Lanovaz, J.L., Schamhardt, al., 2009b. The effect of rising and sitting
of the temporal kinematics of the canter H.C., van Wessum, R., 1999. Rider mass trot on back movements and head-neck
pirouette and collected canter. Equine Vet. effects on ground reaction forces and position of the horse. Equine Vet. J. 41,
J. 23 (Suppl.), 58–61. fetlock kinematics at the trot. Equine Vet. J. 423–427.
Byström, A., Rhodin, M., von Peinen, K., 30 (Suppl.), 218–221. De Cocq, P., van Weeren, P.R., Back, W., 2004.
Weishaupt, M.A., Roepstorff, L., 2009. Basic Clayton, H.M., Larson, B., Kaiser, L.J., Effects of girth, saddle and weight on
kinematics of the saddle and rider in Lavagnino, M., 2011. Length and elasticity movements of the horse. Equine Vet. J. 36,
high-level dressage horses trotting on a of side reins affect rein tension at trot. Vet. 758–763.
treadmill. Equine Vet. J. 41, 280–284. J. 188, 291–294. De Cocq, P., van Weeren, P.R., Back, W., 2006.
Byström, A., Rhodin, M., von Peinen, K., Clayton, H.M., Lee, R., 1984. A flouroscopic Saddle pressure measuring: Validity,
Weishaupt, M.A., Roepstorff, L., 2010b. study of the position and action of the reliability and power to discriminate
Kinematics of saddle and rider in high-level jointed snaffle bit in the horse’s mouth. J. between different saddle-fits. Vet. J. 172,
dressage horses performing collected walk Equine Vet. Sci. 4, 193–195. 265–273.
on a treadmill. Equine Vet. J. 42, 340–345. Clayton, H.M., Singleton, W.H., Lanovaz, J., Decarpentry, A., 1949. Academic equitation. A
Byström, A., Roepstorff, L., Johnston, C., 2006. Cloud, G.L., 2003. Measurement of rein system based on the methods of D’aure,
Effects of draw reins on limb kinematics. tension during horseback riding using baucher and L’hotte. Allens Classic Series 4,
Equine Vet. J. 36 (Suppl.), 452–456. strain gage transducers. Exp. Tech. 27, 79–82.
Byström, A., Stalfelt, A., Egenvall, A., von 34–36. Faber, M., Johnston, C., Schamhardt, H., van
Peinen, K., Morgan, K., Roepstorff, L., Clayton, H.M., Singleton, W.H., Lanovaz, J., Weeren, R., Roepstorff, L., Barneveld, A.,
2010a. Influence of girth strap placement Cloud, G.L., 2005. Strain gauge 2001. Basic three-dimensional kinematics
and panel flocking material. Equine Vet. J. measurement of rein tension during riding: of the vertebral column of horses trotting
38 (Suppl.), 502–509. a pilot study. Equine Comp. Exerc. Physiol. on a treadmill. Am. J. Vet. Res. 62,
Clayton, H.M., 1985. A fluoroscopic study of 2, 203–205. 757–764.
the position and action of different bits in Cook, W.R., 1999. Pathophysiology of bit Faber, M., Schamhardt, H., van Weeren, R.,
the horses mouth. J. Equine Vet Sci. 5, control in the horse. J. Equine Vet. Sci. 19, Johnston, C., Roepstorff, L., Barneveld, A.,
68–77. 196–203. 2000. Basic three-dimensional kinematics
Clayton, H.M., 1994a. Comparison of the De Cartier D’Yves, A., Ödberg, F.O., 2005. A of the vertebral column of horses walking
collected, working, medium and extended preliminary study on the relation between on a treadmill. Am. J. Vet. Res. 61,
canters. Equine Vet. J. 17 (Suppl.), 16–19. subjectively assessing dressage performance 399–406.
Clayton, H.M., 1994b. Comparison of the and objective welfare parameters. 1st Int. Farley, C.T., Glasheen, J., McMahon, T.A.,
stride kinematics of the collected, working, Equitation Science Symposium, Melbourne, 1993. Running springs: speed and animal
medium and extended trot in horses. Australia, pp. 89–110. size. J. Exp. Biol. 185, 71–86.
Equine Vet. J. 26, 230–234. De Cocq, P., Clayton, H.M., Terada, K., Muller, Fruehwirth, B., Peham, C., Scheidl, M.,
Clayton, H.M., 1995. Comparison of the stride J.A., van Leeuwen, J.L., 2009a. Usability of Schobesberger, H., 2004. Evaluation of
kinematics of the collected, working, normal force distribution measurements to pressure distribution under an English

366
 References

saddle at walk, trot and canter. Equine Vet. forces and pressure distribution beneath a Mönkemöller, S., Keel, R., Hambsch, D.,
J. 36, 754–757. fitting saddle. Equine Vet. J. 42, 114–118. Müller, J., Kalpen, A., Geuder, M., et al.,
German National Equestrian Federation, 2002. Kotschwar, A.B., Baltacis, A., Peham, C., 2010b. 2005. PLIANCE MOBILE - 16HE: A study
The principles of riding. The Kenilworth The influence of different saddle pads on about measurements under the saddle after
Press Ltd., Addington, Bucks, pp. 50, force and pressure changes beneath saddles the adjustment of saddle fit. Pferdeheilk.
136–139. with excessively wide trees. Vet. J. 184, 21, 102–114.
Geutjens, C.A., Clayton, H.M., Kaiser, L.J., 322–325. Murphy, J., Arkins, S., 2008. Facial hair whorls
2008. Forces and pressures beneath the Kuhnke, S., Dumbell, L., Gauly, M., Johnson, (trichoglyphs) and the incidence of motor
saddle during mounting from the ground J.-L., McDonald, K., von Borstel, U.-K., laterality in the horse. Behav. Process. 79,
and from a raised mounting platform. Vet. 2010. A comparison of rein tension of the 7–12.
J. 175, 332–337. rider’s dominant and non-dominant hand Nyikos, S., Werner, D., Muller, J.A.,
Gomez Alvarez, C.B., Rhodin, M., Bobbert, and the influence of the horse’s laterality. Buess, C., Keel, R., Kalpen, A., et al.,
M.F., Meyer, H., Weishaupt, M.A., Johnston, Comp. Exerc. Physiol. doi: 10.1017/ 2005. Measurements of saddle
C., et al., 2006. The effect of head and neck S1755254010000243. pressure in conjunction with back
position on the thoracolumbar kinematics Lagarde, J., Kelso, J.A., Peham, C., Licka, T., problems in horses. Pferdeheilk. 21,
in the unridden horse. Equine Vet. J. 36, 2005. Coordination dynamics of the 187–198.
445–451. horse-rider system. J. Mot. Behav. 37, Pantall, A., Barton, S., Collins, P., 2009.
Groesel, M., Zsoldos, R.R., Kotschwar, A., 418–424. Surface electromyography of abdominal
Gfoehler, M., Peham, C., 2010. A Latif, S.N., von Peinen, K., Wiestner, T., and spinal muscles in adult horse riders
preliminary model study of the equine Bitschnau, C., Renk, B., Weishaupt, M.A., during rising trot. Abstracts of the XXVII
back including activity of longissimus dorsi 2010. Saddle pressure patterns of three Int. Conf. Biomech. Sport. In: Proceedings
muscle. Equine Vet. J. 42 (Suppl. 38), different training saddles (normal tree, of the First International Equitation Science
401–406. flexible tree, treeless) in Thoroughbred Symposium. Melbourne, Australia, pp.
Harman, J.C., 1994. Practical use of a racehorses at trot and gallop. Equine Vet. J. 89–110.
computerized saddle pressure measuring 42 (Suppl.), 630–636. Peham, C., Frey, A., Licka, T., Scheidl, M.,
device to determine the effects of saddle Licka, T., Frey, A., Peham, C., 2009. 2001a. Evaluation of the EMG activity of
pads on the horse’s back. J. Equine Vet. Sci. Electromyographic activity of the the long back muscle during induced back
14, 606–611. longissimus dorsi muscles in horses when movements at the stance. Equine Vet. J. 33
Heleski, C.R., McGreevy, P.D., Kaiser, L.J., walking on a treadmill. Vet. J. 180, 71–76. (Suppl.), 165–168.
Lavagnino, M., Tans, E., Bello, N., et al., Licka, T., Kapaun, M., Peham, C., 2004. Peham, C., Kotschwar, A.B., Borkenhagen, B.,
2009. Effects on behaviour and rein Influence of rider on lameness in trotting Kuhnke, S., Molsner, J., Baltacis, A., 2009. A
tension on horses ridden with and without horses. Equine Vet. J. 36, 734–736. comparison of forces acting on the horse’s
martingales and rein inserts. Vet. J. 181, Lovett, T., Hodson-Tole, E., Nankervis, K., back and the stability of the rider’s seat in
56–62. 2005. A preliminary investigation of rider different positions at the trot. Vet J. 184,
Hodson, E.F., Clayton, H.M., Lanovaz, J.L., position during walk, trot and canter. 56–59.
1999. Temporal analysis of walk Equine Comp. Exerc. Physiol. 2, 71–76. Peham, C., Licka, T., Kapaun, M., Scheidl, M.,
movements in the Grand Prix dressage test Manfredi, J., Clayton, H.M., Derksen, F., 2001b. A new method to quantify harmony
at the 1996 Olympic Games. Appl. Animal 2005a. Swallowing frequency in cantering of the horse-rider system in dressage. Sports
Behav. Sci. 62, 89–97. horses: Effects of different bits and bridles. Eng. 4, 95–101.
Holmström, M., Fredricson, I., Drevemo, S., Equine Comp. Exerc. Physiol. 2, 241–244. Peham, C., Licka, T., Schobesberger, H.,
1995a. Biokinematic effects of Manfredi, J., Clayton, H.M., Rosenstein, D., Meschan, E., 2004. Influence of the rider
collection on the trotting gaits in 2005b. Radiographic study of bit position on the variability of the equine gait. Hum.
the elite dressage horse. Equine Vet. J. 27, within the horse’s oral cavity. Equine Mov. Sci. 23, 663–671.
281–287. Comp. Exerc. Physiol. 2, 195–201. Peham, C., Schobesberger, H., 2004. Influence
Holmström, M., Fredricson, I., Drevemo, S., Manfredi, J.M., Rosenstein, D., Lanovaz, J.L., of the load of a rider or of a region with
1995b. Variation in angular pattern Nauwelaerts, S., Clayton, H.M., 2010. increased stiffness on the equine back: a
adaptation from trot in hand to passage Flouroscopic Study of Oral Behaviours in modelling study. Equine Vet. J. 36,
and piaffe in the Grand Prix dressage horse. Response to the Presence of a Bit and the 703–705.
Equine Vet. J. 18 (Suppl.), 132–137. Effects of Rein Tension. Equine Comp. Peinen, K., von, Wiestner, T., Bogisch, S.,
Jeffcott, L.B., 1979. Back problems in the horse Exerc. Physiol. 6, 143–148. Roepstorff, L., van Weeren, P.R., Weishaupt,
– a look at past present and future Matsuura, A., Matsumura, K., Chiba, Y., M.A., 2009. Relationship between the
progress. Equine Vet. J. 11, 129–136. Nakatsuji, H., Kondo, S., 2005. Analysis of forces acting on the horse’s back and the
Jeffcott, L.B., Holmes, M.A., Townsend, rhythmical movements of both horse and movements of rider and horse while
H.G.G., 1999. Validity of saddle pressure rider using accelerometer. Anim. Behav. walking on a treadmill. Equine Vet. J. 41,
measurements using force-sensing array Manag. 41, 5–11. 285–291.
technology – preliminary studies. Vet. J. Matsuura, A., Takita, N., Shingu, Y., Kondo, S., Peinen, K., von, Wiestner, T., von Rechenberg,
158, 113–119. Matsui, A., Hiraga, A., et al., 2003. Rhythm B., Weishaupt, M.A., 2010. Relationship
Kapaun, M., Peham, C., Licka, T., Scheidl, M., analysis for movements of horse and rider between saddle pressure measurements and
1998. The influence of the rider on the on a treadmill by sequential still VTR clinical signs of saddle soreness at the
trotting motion of the horse. In: Riehle, pictures. J. Equine Sci. 14, 125–131. withers. Equine Vet. J. 42 (Suppl.),
H.J., Vieten, M.M. (Eds.), Proceedings of Merkens, H.W., Schamhardt, H.C., Van Osch, 650–653.
the XVI International Symposium on G.J., Van den Bogert, A.J., 1993. Ground Podhajsky, A., 1965. The complete training of
Biomechanics in Sports. Konstanz: reaction force patterns of Dutch horse and rider in the principles of classical
UVK-Universitätsverlag Konstanz. warmblood horses at normal trot. Equine horsemanship. Doubleday & Company,
Karl, P., 2008. Twisted truths about modern Vet. J. 25, 134–137. Inc, pp. 246–247.
dressage. A search for a classical alternative. Meschan, E., 2007. The influence of the width Powers, P.N.R., Harrison, A.J., 2002. Effects of
Cadmos verlag GmbH, Brunsbek, Austria. of the saddle tree on the forces and the rider on the linear kinematics of
Kotschwar, A.B., Baltacis, A., Peham, C., 2010a. pressure distribution under the saddle. Vet. jumping horses. Sports Biomech. 1,
The effects of different saddle pads on J. 173, 578–584. 135–146.

367
 15 Horse–rider interaction

Preuschoft, H., Falaturi, P., Lesch, C., 1995. Physiology. ICEEP Publications, Davis, horse and handler during specific
Grenzen der Einwirkungen des Reiters auf California, pp. 120–127. equitation movements Appl. Animal Behav.
das Pferd. Tierarztl. Umsch. 50, 511–521. Schils, S.J., Greer, N.L., Stoner, L.J., Kobluk, Sci. 108, 157–169.
Preuschoft, H., Witte, H., Recknagel, S., Bar, C.N., 1993. Kinematic analysis of the Weishaupt, M.A., Byström, A., von Peinen, K.,
H., Lesch, C., Wuthrich, M., 1999. Effect on equestrian – walk, posting trot and sitting Wiestner, T., Meyer, H., Waldern, N., et al.,
horses of commonly used bridles. Deut. trot. Hum. Mov. Sci. 12, 693–712. 2009. Kinetics and kinematics of the
Tierarzt. Wochenschr. 106, 169–175. Schöllhorn, W.I., Peham, C., Licka, T., Scheidl, passage. Equine Vet. J. 41, 263–267.
Pullin, 1996. Use of force sensing array M., 2006. A pattern recognition approach Weishaupt, M.A., Wiestner, T., Hogg, H.-P.,
technology in the development of a new for the quantification of horse and rider Jordan, P., Auer, J.A., 2004. Vertical ground
equine saddle pad: static and dynamic interactions. Equine Vet. J. 36 (Suppl.), reaction force-time histories of sound
evaluations and technical considerations. J. 400–405. Warmblood horses trotting on a treadmill.
Equine Vet. Sci. 16, 207–216. Sloet van Oldruitenborgh-Oosterbaan, M.M., Vet. J. 168, 304–311.
Rhodin, M., 2008. A biomechanical analysis of Barneveld, A., Schamhardt, H.C., 1995. Weishaupt, M.A., Wiestner, T., von Peinen, K.,
relationship between the head and neck Effects of weight and riding on workload Waldern, N., Roepstorff, L., van Weeren, R.,
position, vertebral column and limbs in and locomotion during treadmill exercise. et al., 2006. Effect of head and neck
the horse at walk and trot. Dissertation, Equine Vet. J 18, 413–417. position on vertical ground reaction forces
Uppsala. Sloet van Oldruitenborgh-Oosterbaan, M.M., and interlimb coordination in the dressage
Rhodin, M., Gomez Álvarez, C.B., Byström, A., Barneveld, A., Schamhardt, H.C., 1996. horse ridden at walk and trot on a
Johnston, C., van Weeren, P.R., Roepstorff, Kinematics of unmounted and mounted treadmill. Equine Vet. J. 36 (Suppl.),
L., et al., 2009. The effect of different head horses at walk before and after treadmill 387–392.
and neck positions on the caudal back and exercise. Pferdeheilk. 12, 651–655. Werner, D., Nyikos, S., Kalpen, A., Geuder, M.,
hind limb kinematics in the elite dressage Symes, D., Ellis, R., 2009. A preliminary study Haas, C., Vontobel, H.D., et al., 2002.
horse at trot. Equine Vet. J. 41, 274–279. into rider asymmetry within equitation. Pressure measurements under the saddle:
Robert, C., Valette, J.P., Denoix, J.M., 1998. Vet. J. 181, 34–37. a study using an electronic saddle mat
Surface electromyographic analysis of the Terada, K., 2000. Comparison of head system (Novel GmbH). Pferdeheilk. 18,
normal horse locomotion: a preliminary movements and EMG activity of muscles 125–140.
report. In: Proceedings: Conference on between advanced and novice horseback Wijnberg, D., Sleutjens, J., van der Kolk, J.H.,
Equine Sports Medicine and Science 24-26 riders at different gaits. J. Equine Sci. 11, Back, W., 2010. Effect of head and neck
April 1998, Cordoba, Spain, pp. 80–85. 83–90. position on outcome of quantitative
Robert, C., Valette, J.P., Denoix, J.M., 2001. The Terada, K., Clayton, H.M., Kato, K., 2006. neuromuscular diagnostic techniques in
effects of treadmill inclination and speed Stabilization of wrist position during Warmblood riding horses directly following
on the activity of three trunk muscles in horseback riding at trot. Equine Comp. moderate exercise. Equine Vet. J. 42
the trotting horse. Equine Vet. J. 33, Exerc. Physiol. 3, 179–184. (Suppl.), 261–267.
466–472. Terada, K., Mullineaux, D.R., Lanovaz, J., Kato, Winkelmayr, B., Peham, C., Fruhwirth, B.,
Robert, C., Valette, J.P., Pourcelot, P., Audigié, K., Clayton, H.M., 2004. Electromyographic Licka, T., Scheidl, M., 2006. Evaluation of
F., Denoix, J.M., 2002. Effects of trotting analysis of the rider’s muscles at trot. the force acting on the back of the horse
speed on muscle activity and kinematics in Equine Comp. Exerc. Physiol. 1, 193–198. with an English saddle and a side saddle at
saddlehorses. Equine Vet. J. 34 (Suppl.), Wakeling, J.M., Ritruechai, P., Dalton, S., walk, trot and canter. Equine Exerc.
295–301. Nankervis, K., 2007. Segmental variation in Physiol. 36 (Suppl.), 406–410.
Roepstorff, L., Egenvall, A., Rhodin, M., the activity and function of the equine Zsoldos, R.R., Kotchwar, A., Kotchwar, A.B.,
Byström, A., Johnston, C., van Weeren, P.R., longissimus dorsi muscle during walk and Rodriquez, C.P., Peham, C., Licka, T.,
et al., 2009. Kinetics and kinematics of the trot. Equine Comp. Exerc. Physiol. 4, 2010b. Activity of the equine rectus
horse comparing left and right rising trot. 95–103. abdominis and oblique external abdominal
Equine Vet. J. 41 (Suppl.), 292–296. Waldern, N.M., Wiestner, T., von Peinen, K., muscles measured by surface EMG during
Roepstorff, L., Johnston, C., Drevemo, S., Gomez-Alvarez, C.G., Roepstorff, L., walk and trot on the treadmill. Equine Vet.
Gustas, P., 2002. Influence of draw reins on Johnston, C., et al., 2009. Influence of J. 38 (Suppl.) 523–529.
ground reaction forces at the trot. Equine different head-neck positions on vertical Zsoldos, R.R., Kotchwar, A.B., Kotchwar, A.,
Vet. J. 34 (Suppl.), 349–352. ground reaction forces, linear and time Groesel, M., Licka, T., Peham, C.,
Schamhardt, H.C., Merkens, H.W., van Osch, parameters in the unridden horse walking 2010a. Electromyography activity of the
G.J.V.M., 1991. Ground reaction force and trotting on a treadmill. Equine Vet. J. equine splenius muscle and neck
analysis of horses ridden at the walk and 41, 268–273. kinematics during walk and trot on the
trot. In: Persson, S.G.B., Lindholm, A., Warren-Smith, A.K., Curtis, R.A., Greetham, L., treadmill. Equine Vet. J. 42 (Suppl.),
Jeffcottt, L.B. (Eds.), Equine Exercise McGreevy, P.D., 2007. Rein contact between 455–461.

368
 C H A P T ER 16 

Ethology and welfare aspects

Natalie K. Waran, Machteld C. Van Dierendonck

Introduction from one maternal group to another or to a bachelor band, as

occurs in many feral herds. For example, Berger (1986) reported
that fillies moved on average 4 km and males moved between
Through the process of domestication the horse has gained survival 10 and 15 km away from their natal herd in the Navada region of
advantages such as food, shelter, veterinary care and protection, but the USA, and in Japan, the Misaki horses will emigrate over much
this has been at the cost of restriction of movement, limited breed- greater distances. Having evolved from animals adapted to live in
ing opportunities and a requirement to expend energy for the the unpredictable and open environment of the Steppe region of
benefit of another species. In fact it is likely that many of the Asia, it is not surprising that they exhibit anatomical specializations
demands made on the domestic horse conflict with the evolution- for running. These adaptations include anatomical changes in the
ary process that shaped the behavior of its predecessors (Goodwin, 3rd phalanx (the caput ridge on the distal metacarpi and metatarsi)
1999). The evolution of the horse began some 60 million years ago as well as the development of the ‘spring mechanism’.
(Fig. 16.1). The ancestor of the domestic horse was Eohippus or the Due to the spring mechanism, the energy the horse needs for
‘Dawn Horse’ (Hyracotherium). Eohippus was a small creature locomotion is halved through storing and returning elastic strain
about the size of a fox, with adaptations for life in the swampy energy in spring-like muscle–tendon units, and is optimized by the
regions of North America, with four toes on its front feet and three anatomical construction of the tendons and the number of distal
on the hind feet. The ‘Dawn horse’ died out, and by the Oligocene joints affected by the tendons (Wilson et al., 2001). These features
period its successors had grown to the size of a large dog with three allowed the horse to respond quickly through early detection of
toes on each foot. Pliohippus was the first equid to walk on one danger due to their well-developed senses (as social behavior devel-
digit, the two ‘other’ toes being reduced to long thin vestiges, more oped – it is likely that their sensory development may have become
commonly referred to as the splint bones in the domestic horse. enhanced), followed by flight as their first defense mechanism
The various breeds of Equus caballus are derived from a 13–14 hand against predators. Therefore sociality and flight became neurologi-
(130–140 cm) at the withers, social, Steppe-dwelling animal that cally ‘blue printed’, and are considered a biological need even for
thrived on low-quality forage through the development of domesticated and intensively managed horses. Equus caballus (the
hypsodontic molars as well as by evolutionary adaptations to the modern ancestor of the little dawn horse) exhibits similar motiva-
digestive tract. tion in relation to much of their locomotion and behavior.
The domestic horse is a non-territorial member of the family
Equidae (Goodwin, 2002), and is often described as a follower
species, because the precocial young stand and move to feed within
a few hours of birth, after which they will follow the mother as she Development of locomotion
moves off with the herd. Thus the horse performs locomotory
behavior from only a few hours of age. This makes perfect evolu-
tionary sense, since as a prey animal the horse relies upon early
Early kinetic activity
predator detection and flight as its primary defense mechanisms. Simple movements initiated by the developing fetus can be detected
In terms of locomotion, in its natural state, the horse has been from only a few months old and include limb stretching, becoming
observed to range over considerable distances. In ethological terms, more complex and coordinated as it matures. Once born, the foal
the horse’s home range contains resources important for survival becomes mobile through a series of stages described by Fraser
such as watering holes and suitable grazing areas. Home ranges vary (1992) that result in teat seeking and sucking. Foals differ in their
in size between and within areas where feral horses have been ability to achieve each stage of mobility, but usually they are stable
studied depending on the availability resources, and the available on their limbs within hours of birth (Fig. 16.2). Its interesting that
area ranges have been reported from 0.9 to 48 km2 and more in although most foals are able to walk, canter and suckle very soon
different study sites (Berger, 1986; Boyd & Keiper, 2005; Tyler, after birth, they need a few days to develop the skills needed for
1972). In addition, the members of a horse band (the structured lying down and standing up. Many foals will literally fall asleep on
social unit within a herd) follow similar movement patterns within their feet, during their first days of life, illustrating that even in
a common home range. Changes in band composition may occur precoccial species, complex locomotory movements need to
through natal dispersal, which is the movement of young horses be learned and practiced. It is worth considering how much the

369
 16 Ethology and welfare aspects

South America North America Eurasia

PLESIOCENE RECENT

Modern horse (equus)

0.01 mya

Sindohipparion
Hippidion group Equus
1.05 mya
PLIOCENE

Nannippus
Neohipparion Hipparion
Pilohippus
Calippus Sinohippus
Hipparion Megahippus
3.3 mya

Pilohippus Archaeohippus
Anchtherium
MIOCENE

Hypohippus

Merychippus

Merychippus
Parahippus Anchitherlines
23.7 mya
Miohippus
OLIGOCENE

Miohippus Mesohippus

36.6 mya Epihippus Palaeotheres

EOCENE

Hypracotherium

Hypracotherium
Fig 16.1  Evolutionary tree of the horse. Note the transition from browsing forest dweller to hypsodontic grass eating Steppe roamer in the Miocene, as well
as the extinction in the America’s in the last ice age.

intensive housing of foals at this early stage may hamper the devel- is an instinctive tendency in a foal, refined through learning during
opment of many behaviors, including locomotion. the first week of life. Over time, the mare and foal will spend more
One of the primary activities foals engage in is stretching, which time apart, until by the age of 1 year the youngster may be grazing
seems to be of greatest importance during the first few weeks after for approximately 44 min/h alongside the rest of the herd. Grazing
birth. Foals have been observed to perform this activity 40–60 times involves moving whilst foraging, an activity that horses in their
each day (more than sucking) (Fraser, 1992). It is suggested that natural state will perform for around 16 h each day depending on
both recumbent and upright stretching are important for athletic their physiological and reproductive state, the time of year, available
development, growth and joint formation. Especially in the first 10 forage and weather conditions. Thus horses have evolved to spend
days of their life, foals are engaged in many short bouts of solitary much of their time moving slowly with their head and necks
locomotory play around their dams, allowing their muscles to lowered, whilst selecting suitable forage. In the natural environ-
develop. ment, horses may travel up to 80 km during grazing and for water,
and even when kept on restricted pastures, horses are estimated to
move around 20 km each day (McGreevy, 2004). Increased locomo-
Grazing tory activity has been found in horses kept socially isolated from
During the first week of life the mare and foal spend approximately others in which case they will walk and trot three times more fre-
90% of their time within 5 m of each other (Waring, 2003). During quently than horses kept in groups or in visual contact with another
these early weeks, the mare repeatedly leaves the foal and moves a horse (Houpt & Houpt, 1989), apparently to try to get into more
short distance away, and the foal follows. This following behavior close contact with conspecifics.

370
 Development of locomotion

Fig 16.2  A foal just after birth quickly learns to balance and to move Fig 16.4  Fifty percent of foals develop a side preference while equally
around so it can follow its dam, even in canter. However, complex behaviors distributed left and right.
like lying down take much longer to learn. © Arnd Bronkhorst/www.arnd.nl.
© Arnd Bronkhorst/www.arnd.nl.

engage in compensatory locomotion activity. However this was

insufficient to reach the level of locomotion of foals kept continu-
ally outside (Kurvers et al., 2006).
There have been relatively few studies of play in non-domestic
equids and where recorded this has generally been anecdotal, but
social and solitary-locomotor play have been recorded in Hart-
mann’s zebra foals (Joubert, 1972) and also in captive juvenile
Przewalski horses (Zharkikh, 1999) as well as in free-ranging and
feral horse populations in the UK and USA (Tyler, 1972; Berger,
1986). In these groups, the majority was social play behavior with
activity patterns including: play fighting, neck wrestling and chasing
and with solitary-locomotor play including gamboling, high speed
turns and sudden stops (see Goodwin & Hughes, 2005). During
play, absolute force values as exerted on the developing tissues are
probably not substantially different from those created by canter-
ing, but play activities result in other loading directions, including
eccentric loading and loading in the sagittal plane, which might be
important for the development of musculoskeletal structures
(Kurvers et al., 2006) (see Chapter 13).
Fig 16.3  Play in young horses is essential for the development of their The findings of various studies suggest that (24 h) pasture-
physical as well as their social skills. Playmates of similar age provide a based management systems for foals may be superior to more inten-
strong stimulus to the young horse to move. sive management systems due not only to the recognized advantages
Courtesy of Machteld van Dierendonck. of social interaction with other mares and young foals but also the
physical gains due to increased and varied locomotion in early
development. One of the impacts of the pressures of producing
horses as early in the season as possible, as is the case in the Thor-
Play oughbred racing industry and many sports horse registration
Apart from movement associated with following and grazing with schemes, is that in some climates, it is difficult to persuade breeders
the herd, the foal engages in various locomotory activities described of the social, mental, physical and thus welfare advantages of the
as play (Fig. 16.3). There appears to be no universally agreed func- pasture-based system, when their foals are being born during the
tion or homogeneous character to define play, but it is apparent coldest months of the year.
that play has survived the pressures of natural selection and is an
almost universal trait in mammals. There are a number of proposed
functions which include development of social skills (Fraser, 1992);
Development of laterality (sidedness)
motor strength and co-ordination; behavioral flexibility (Spinka Equine laterality (or sidedness in horses) has been investigated in
et al., 2001) and cognitive training. In horses, play occurs most relation to equine performance and training as well as trying to
frequently in younger animals, with colts performing more than understand the underlying factors that lead one horse to favor a left
fillies (Kurvers et al., 2006). When the spontaneous locomotion lead over a right leading leg. In a series of studies, (van Heel et al.,
activity of foals kept under various management conditions was 2006; Kroekenstoel et al., 2006; van Heel et al., 2010) healthy
compared (Kurvers et al., 2006), it was found when foals were kept jumping bred KWPN foals were born and raised outside at the same
for 24 h/day at pasture, they exercised at a level comparable with location, and were then followed weekly for a year to observe the
feral foals. Interestingly the total observed daytime workload in the development of laterality (Fig. 16.4).
first month of life was approximately twice that of the following They were assessed at 3, 15, 27 and 55 weeks of age for pressure
months, while locomotion activity decreased with increasing age. distribution under their feet (to determine the Centre of Pressure),
Foals that were stabled for some portion of the day appeared to external and internal development of the foot including phalangeal

371
 16 Ethology and welfare aspects

80 that information regarding the natural activity patterns of the horse

Absolute canter score should be used to plan the training schedules and competition
management of athletic horses. However a recent study of the
60 impact of biorhythm changes associated with long distance travel
for competition purposes suggests that horses are far more flexible
40 than humans in relation to coping with change (Murphy, 2008). In
humans a change in the light/dark cycle has a number of effects
y = 20.257x -1.1402 including disturbed sleep patterns and lethargy. Given that race-
20 R2 = 0.7932 horses and other high level sport horses are frequently transported
long distances between hemispheres there are questions about how
quickly they recover from any jetlag in order to perform well. The
0 findings of the study by Murphy (2008) showed that melatonin
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
levels in the two test horses that experienced an abrupt change in
Absolute z-value leg preference
light/dark cycles, were reset within 24 h, whereas regulation of body
Fig 16.5  Relationship between canter (% correct lead when horse is asked temperature took approximately 3 days. These results suggest that
to canter; 15 repeats per horse) and side preference (15 repeats) of 3-year horse performance is less compromised by the time change than
old horses (n = 17). human athletes and that horses possess the ability to rapidly adjust
Reprinted with permission from van Heel et al. (2010). to any changes in photoperiod.

The impact of intensive management

development and ‘handedness’ (or side preference). In addition on locomotion
several conformation measures were made such as relative leg In the natural environment, horses move continuously as they graze
length and head size, heel height and toe angle. It was shown that between food patches with short periods of rest, social activity and
about 50% of the foals developed a significant preference for pro- locomotion. Apart from the movement associated with grazing,
tracting the same limb systematically while grazing, which resulted horses spend approximately 9% of their daily time budget engaged
in changes to the balance of their feet and subsequently uneven in other forms of locomotion (which could be play, travelling, or
loading patterns. There was no overall preference for either left or due to flight responses). This daily time budget changes dramati-
right sidedness, with an even distribution between the two sides. cally once the horse is managed more intensively. Horses housed
However it appears that foals with relatively long limbs and small in stables spend a great deal of their time standing and very little
heads were predisposed to develop laterality and, consequently time moving whilst eating.
unevenness (van Heel et al. 2006, Kroekenstoel et al. 2006). For example, various studies of feral horses have shown that
Seventeen of these horses were tested again when 3 years old, horses will spend approximately 60–80% a day on feeding, during
after they had only been handled for normal management proce- which they walk around 5–10 km, and even up to 20 km a day
dures like trimming and de-worming (van Heel et al., 2010). The (VanDierendonck et al., 1996; Feist & McCullough, 1976; McGreevy,
assessments revealed that the laterality as well as the unevenness 2004). By contrast in stabled horses both feeding and moving time
still existed, but the relationship between the body conformation are significantly reduced due to the restricted space and the fact that
and the unevenness had weakened. In addition when these horses there is no need to move in order to obtain food (Mills & Clarke,
were also assessed in free trot to canter transitions and free jumping, 2002; Sweeting et al., 1985; Vervuert & Coenen, 2002; Waring,
laterality and unevenness were strongly related to incorrect transi- 2003). When there is a high motivation for a stabled horse to
tions (either disunited canter or counter-canter) (Fig. 16.5). perform certain behaviors but expression is restricted due to envi-
Since in ridden horses there is a preference for left sidedness ronmental factors this can lead to frustration and hence to the
(McGreevy & Rogers, 2005; McGreevy & Thomson, 2006), the results development of abnormal behavior (Mills & Clarke, 2002; Rushen
of studying 3-year-olds suggests that left sidedness can be induced et al., 1993) including behavioral pathologies that are extremely
by the human interaction with the horse, which, in most cases, is difficult to manage. Foraging in itself is an appetitive behavior and
predominantly from the left side of the horse. Apart from the is a biological need which a horse must be able to perform other-
concern that horses with strong lateral preferences might be more wise its welfare could be hampered (VanDierendonck, 2006). One
prone to overloading injuries of the distal limb it is also interesting of the problems with the intensive management of horses kept in
to note that this could relate to performance problems later in life. stables or stalls is their restricted movement associated with the
For example: horses may be less balanced when running around acquisition of food. These horses are provided with all their daily
racetracks that run to the left or right, or have problems with picking requirements in ‘meals’ and no foraging is required. Providing
up the correct lead in canter when under saddle. There is a need for stabled/stalled horses with multiple forage types in different loca-
more applied work in this area to determine how motor laterality tions within the indoor housing environment can increase move-
can be influenced through breeding, environment and training. ment and has also been shown to reduce abnormal activities such
as weaving (see Thorne et al., 2005)
Natural rhythms of activity
Circadian rhythms reflect extensive programming of biological
Feeding
activity that meets and exploits the challenges and opportunities The effects of diet on equine locomotory responses has received
offered by the periodic nature of the environment. Bertolucci et al. surprisingly little scientific attention, despite the fact that many
(2008) found that athletic horses monitored for activity during four sports horses are fed high-energy diets in order to boost perfor-
different times of the year (vernal equinox, summer solstice, autumn mance. A study by Nicol et al. (2005) examined the effect of differ-
equinox and winter solstice) experienced seasonal variations in ent feeds on the behavior of foals aged from 2 to 40 weeks. Each
their daily activity rhythms. Athletic Thoroughbred horses kept in foal received either a starch and sugar (SS) diet, or a fat and fibre
individual boxes (stables) with access to a paddock showed the (FF) diet, and their behavior was monitored immediately after
highest daily amount of activity during the vernal equinox (spring) weaning. The results showed that horses receiving the FF diet can-
and the lowest during the winter solstice but in all cases horses were tered less frequently and for a shorter duration and the authors
most active in the middle of the photoperiod. The authors suggest described the foals receiving the fat/fiber diet as more settled. This

372
 Effect of management regimes on locomotion and welfare

suggests that at least during a stressful time such as occurs at is not necessary, and they do not require as careful monitoring of
weaning, high-energy diets appear to be associated with increased exercise and feeding as competition horses do. This system works
locomotion. However, inappropriate or over-feeding high-energy best where group composition can be kept stable, and where there
foods to horses is also known to be associated with unwanted is space available for horses to escape any unwanted attention from
locomotory responses, such as explosive reactions to signals (aids), others in the group. Its advantages are that horses can interact freely,
and this is particularly the case for horses housed in stables (Riveria move fairly freely and live a more natural life. Group sizes vary from
et al., 2002). Hyperactivity responses, including shying, pulling, two or three to 60 or more and space allowances of the same dimen-
bolting, and rushing could all be considered to be the result of a sions as a loose-box per individual are recommended.
restricted environment, lack of exercise and too much sugar/starch Relatively recently technology borrowed from the automatic
in the diet. It is likely that the behavioral problems often experi- milking systems for dairy cows has been adapted and used in group
enced by horse owners and that lead to poor owner–horse relation- housing for horses. Unpublished studies by one of the authors
ships, and thus lower equine welfare, could be prevented through (MvD) have shown that despite these systems needing more techni-
improved understanding of the effect of food and feeding on horse cal (software) development, the horses housed in this way more
behavior and by providing the horse with a more natural time- closely resemble the natural (semi-feral) feeding, locomotion and
budget including access to an area to enable it to perform a normal resting time budgets than stabled/stalled horses. Moreover, in these
locomotion pattern. systems the animals can perform biologically essential behavior
when they want to: so they can be social, move freely and perform
Impact of housing/restricted grazing locomotory behaviors more similar to those seen in horses in
paddocks/at grass. This ability to perform more natural responses
on locomotion including locomotion is considered to lead to a reduction in stress
(Weiss, 1971) and therefore enhanced welfare as compared with
As far as locomotion behavior under more natural conditions is other intensive housing systems.
concerned, horses have been shown to travel distances of up to
65–80 km/day. Where horses are managed at pasture, grazing is the Effect of management regimes on locomotion
main initiator of locomotion and the distance they cover is there-
fore usually much less. There are a number of different methods for
and welfare
managing sports horses, however the demands of competition and
training are often associated with controlled exercise, restricted The method of management obviously impacts greatly on horse
feeding regimes and restricted housing for at least part of the day. locomotion time and type. The confined conditions in which most
There are three basic types of housing for intensively managed performance and recreation horses are kept at least for part of a
horses, which include the stall, the stable or loose box and group 24-h period, clearly restricts normal behavior. The stable appears to
or loose housing. conflict with many of the horse’s survival instincts, and the associ-
The stall (known as the straight stall in the USA), in which the ated effects of restricted feeding and exercise are thought to lead to
horse is tethered in a space usually not much greater than the width a number of problems for the horse (VanDierendonck & Goodwin,
and length of the animal is therefore the most restrictive for the 2005). Changes in kinetic behavior associated with housing are
horse. Generally there is very limited access to other horses, and no quite extreme – horses in stables will stand for up to 40% of their
space to turn or move around voluntarily. The horse is limited to daily time budget, and move infrequently, whereas horses free-
some forward and backward movement, and lying down on its ranging in Camargue district of France will only stand for 20% of
brisket and sometimes its side if the tether allows. Although stalls the day, and will spend 60% grazing and 10% in other activities
are less popular now than during the era of the working draught including social grooming, rolling and play (Kiley-Worthington,
horse, they are still used where space is limited or where traditional 1987). These figures are confirmed in studies of feral horses in other
methods of housing are valued. In some countries (e.g. Germany area’s around the world such as New Zealand, USA, Iceland and
‘Bundesländer’, Sweden and Denmark) they are forbidden by law. Mongolia (Linklater et al., 2000; VanDierendonck et al., 1996; Cran
The most common type of housing in Europe is the stable or loose- et al., 1997; Sigurjónsdóttir et al., 2003).
box (also called the box stall in the USA), where the horse has However there is a question regarding whether exercise is actually
limited freedom of movement, some external stimulation, and dif- important to the horse. That is, although there are undoubted dif-
fering degrees of access to conspecifics depending on the internal ferences in the amount of time, the type and the space available for
partitions. Most Western European countries have a recommended movement in the different housing systems, is this really a problem
minimal stable size for adult horses and ponies (e.g. 2 × height-at- for a horse? This issue has recently arisen due to public concern
the-withers). This allows a pony/horse to perform its normal about the welfare of horses used in the Pregnant Mare Urine (PMU)
feeding, resting and maintenance behaviors without space limita- industry in the USA, and in particular the fact that mares are con-
tions. Despite this minimum requirement, in many countries (e.g. fined for long periods in straight stalls in which they can not turn
UK, The Netherlands, Belgium, Germany and Spain) most stables around and do not have the chance to move or exercise. To answer
are standardized at 3 × 3 m or less and therefore often very restric- this, Houpt et al. (2000) developed a device based on the phenom-
tive for the most common breeds. In the UK 4 × 4 m is recom- enon of operant conditioning. In their study, horses had to learn to
mended for horses ‘kept inside for long periods’, while the European press a panel a number of times in succession (the switch) in order
Union law uses 5 × 4 m for research horses. to release themselves from the stall. Houpt and co-workers mea-
Horses in the restricted stalls perform no ‘normal’ locomotion, sured the strength of a horse’s motivation for each of three com-
possibly stepping back and forward being the only movement being modities: food, exercise (release into a paddock for 3 min) or
possible, apart from the hour or so during which they may be exer- companionship (access to another mare for 3 min). They found
cised. Horses in stables or loose-boxes can move, but can only do that the chance to leave the stall (i.e. to experience free movement)
so within the confines of the enclosure and generally are restricted was less important than leaving to access food and the opportunity
to only walking, although some horses manage to perform other to exercise and have access to another mare was equally chosen. In
forms of locomotion such as bucking and rearing. a further trial the authors looked at the preference of the horse for
The third type of housing used for the performance horses is engaging in exercise on a treadmill or alternatively, returning to the
group or loose-housing in barns or yards. Group housing is mainly home stall. Interestingly most horses chose their home stall and
used for keeping breeding mares or young stock together, since actively avoided the treadmill, suggesting that exercise per se, may
these are rarely handled, and so access to each individual every day not be as important to the horse as often considered. However when

373
 16 Ethology and welfare aspects

Houpt and Ogilvie-Graham (2002) compared the behavior of

horses allowed daily turn-out with those only given access to a
paddock every 2 weeks they found that there was much more move-
ment, in fact more running, in horses that had been restricted for
longer periods. This suggests that horses need to move but do not
necessarily choose to exercise. Despite whether they choose to or
not, exercise is important for the physical development of young
horses (Barneveld et al., 1999). Group housed young horses show
more locomotion than animals kept alone (Christensen et al.,
2002), and it is suggested that due to the performance of move-
ments such as bucking, rearing and chasing, they are likely to
develop more coordinated movements, which is highly desirable in
the sports horse. However along with such activities there is a risk
of injury, especially if the enclosure is not large enough or the
surfaces/internal environment is unsafe. Since it seems increased
play, aggression and locomotion are associated with the rebound
effect of solitary housing (Christensen et al., 2002), there appear to
be more risks in relation to future injuries and behavioral issues
associated with housing horses in the conventional way than raising
them in groups outside where they have no need to compensate for Fig 16.6  Abnormal behaviours like weaving and box walking can be
the lack of opportunity to perform natural responses. reduced by offering visual (and preferably physical) access to a conspecific;
but in cases where this is not possible, a well placed mirror can be a
valuable substitute.
Impact of exercise on welfare Photo courtesy of Professor D. Mills.

The inability of the stabled horse to escape from the situation it

finds itself in, or to control any aspect of its physical environment,
often leads to the development of somatic or body-based forms of leads to greater muscle development. However, many horses who
abnormal behavior such as stereotyped pacing or box walking, do show this rebound effect show uncontrolled and dangerous
weaving, head nodding and pawing. All of these activities are associ- locomotion patterns and tend to be hyper reactive to stimuli from
ated with the stable environment, and some are associated with a their environment, leading to a perceived need for greater rider
reduction in health and therefore their appearance must be viewed controllability of the horse through the use of stronger bits and
as welfare concern. bridles often leading to greater problems as the horse becomes
Weaving appears to be a form of frustrated locomotion, in which more stressed and the development of inappropriate responses and
the horse that is prevented from moving forward by a stable door musculature.
or field gate, will move its head, neck and body from side to side,
or head nod. This is often accompanied by the shifting of weight
from one foreleg to the other. Box-walking (or stall-walking) is a Overtraining
form of stereotypic pacing that follows the walls of the stable, and
pacing is often seen along field fences, again apparently following In order to improve performance in sports, the training load has to
the line of the barrier. be gradually increased over time. This improvement is only achieved
The motivation for such abnormal locomotory activities has been if the training load is properly balanced with recovery. However, in
suggested as being related to lack of exercise, boredom, social isola- both human and equine elite athletes, the ratio between training
tion or general frustration due to confinement and management load and recovery is often in a precarious equilibrium. When this
procedures. These abnormal behaviors can also be induced when imbalance occurs over a longer period, it may eventually lead to a
horses have to undergo sudden box rest, due to some kind of injury severe mal-adaptation to the increased workload defined as over-
or illness and this sudden change is not anticipated for by adapting reaching or even, in more severe cases, overtraining. In addition, in
the management around the affected horse. Reduction in these both horses and humans, several gradual stages of maladaptation
abnormal behaviors has been found through supplying the horse to training can be observed: mechanical overload, metabolic over-
with increased social contact, stable mirrors, increased visual hori- load (or overreaching) and overtraining (or staleness). Since a full
zons (through increasing the number of openings to the outside blown overtraining syndrome can lead to a long absence from
environment), increased exercise, and changes in food presentation competition and is a severe welfare issue for the athlete involved, it
(Cooper et al., 2000; Cooper & McGreevy, 2002) (Fig. 16.6). is important to prevent this (VanDierendonck et al., 2007). In a
Structured exercise such as occurs in many racing yards may lead study to systematically assess the horse’s attitude, behavior
to a reduction in the frequency of locomotor stereotypies but this and response to intensified training for identifying factors related
is not necessarily the case with abnormal oral behavior such as to early overtraining or overreaching (VanDierendonck et al., 2007),
crib-biting, demonstrating that the underlying motivational causes early overtraining was induced in a treadmill study (Fig. 16.7).
for these behaviors are quite different (McBride & Long, 2001). All horses started with an 18-month training protocol on a tread-
Increased and abnormal kinetic activity can therefore be used as mill. Next, half of the horses (controls) continued on the same
an indicator of stress in managed horses. Such measures include a schedule (4 training days, 2 days endurance training, 2 days high-
reduction in normal recreational or functional locomotion such as intensity training), the other half went to a more intensive schedule,
seen in play or grazing as well as an increase in flight responses in ending with 7 days a week unpredictable training. The study fin-
response to the environment. ished with 4 weeks of detraining (de Graaf-Roelfsema, 2007). All
Some trainers make inadvertent use of the rebound effect in train- experimental horses showed significant changes in their behavior
ing due to the housing and management regime in race yards. It is (VanDierendonck et al., 2007) and in their growth hormone release
suggested that when horses are housed for up to 23 h each day, and (De Graaf-Roelfsema et al., 2009). After 4 weeks of detraining, most
then are given the opportunity to be exercised in short time frames, horses were not back to their pre early overtraining levels in relation
they are likely to be more motivated to locomote which in turn to performance and behavior. The significant factor with the greatest

374
 Overtraining

potentially life-threatening situation. As a prey animal, the horse

has a particularly well-developed sense of self-preservation and
there are a variety of actions it may take including bolting, bucking,
rearing, kicking, shying and freezing. These behaviors are often the
result of a conflict between the horse’s desire to escape from pain
or pressure and their inability to do so due to the environment,
situation or impact of the devices that control them (see McGreevy
(2007) for a discussion of equitation science).
There are many trainers, advocators of different methods and
training devices that advertise the various ways for dealing with
problem behavior in horses; however, if a horse has learned to buck
and run away from pain due to inappropriate riding or equipment
causing pain, it is clear that understanding the behavioral mecha-
nisms that underpin the response must be the most effective way
to deal with it and prevent such responses in future.
There are a variety of ways in which humans can impact nega-
tively on horse welfare in relation to locomotion. Performance
horses are used for jumping high fences, running at high speeds for
prolonged periods, performing highly elevated and exaggerated
movements in dressage, swerving and cutting or sliding to a halt
from a gallop. Although horses naturally perform these behaviors,
they are trained so that they are performed more frequently, more
exuberantly and to a greater extent than the animal has evolved to
do. In many cases the horse has to be trained through rigorous
gymnastic work to be physically strong enough, developed enough
and mentally able to perform such locomotory activities for com-
Fig 16.7  Treadmill used to induce early overtraining experimentally by de petition, and of course there are injuries, stresses and wastage (see
Graaf Roelfsema and co-workers. Waran, 2002).
Courtesy of Wim Back.

Locomotion and the impact of different

impact on the horses seemed to be the absence of ‘resting’ days in training methods
the training protocol. Since there were no alterations at the muscu-
lar level, it was concluded that the applied stress involved altera-
Head and neck positions
tions in the central neuro-endocrine system preceding the peripheral The aim of all training is to produce a horse that is useful and
systems, indicating that physical stress started with alterations in the responsive to signals of the rider or trainer. For the horse to meet
neuro-endocrine system: that is, the overtrained horses endured these conditions, its weight, plus that of its rider, must be correctly
more mental stress than physical stress (de Graaf-Roelfsema, 2007). distributed over all four limbs. Horses naturally carry more of their
Therefore, it seems that locomotion may be important to the horse weight over the forelimbs than over their hind limbs. In collection,
but there are serious welfare implications when early signals of the hind limbs flex more, stepping further underneath the horse in
overtraining (i.e. too much (forced) locomotion) are not recognized the direction of the center of gravity, thereby taking a greater share
and dealt with. of the load. This in turn lightens the forehand, allowing the fore-
limbs to move more freely (Gomez-Alvarez et al., 2006; Rhodin,
2008).
Problems with locomotory behavior during In attempting to achieve competition goals, trainers and riders
may employ training methods that have the potential to cause pain
training or performance and distress to the horse. For example in trying to achieve greater
Not all forms of locomotion are desirable in the horse. Problem collection, some training methods advocate prolonged extreme
behaviors often result from the unwanted performance of the head and neck positions (either very high carriage or very low posi-
horse’s normal flight response during the horse’s training. ‘Train- tions, also called hyperflexion of the neck). Hyperflexion of the
ing’ can be defined as the intentional modification of the frequency horse’s neck is currently employed as a training method to reach
and /or intensity of specific behavioral responses. Such modifica- maximal collection by a number of high-level competition dressage
tion is achieved through different means: the shaping of a response riders, and is increasingly used by eventers and show jumpers (also
where the specific behavior is positively reinforced in a step-wise referred to as Rollkur or LDR (low, deep, round)) (Fig. 16.8). The
fashion towards a specific end point and the suppression of unde- FEI defines hyperflexion as ‘Hyperflexion of the neck is a technique
sirable responses. Successful and humane training relies upon the of working/training to provide a degree of longitudinal flexion of
trainer having a good understanding of the application of learning the mid-region of the neck that cannot be self-maintained by the
theory, horse ethology and a clear understanding of the goals and horse for a prolonged time without welfare implications’. There is
the limitations of the training approach being used. Humans influ- considerable debate regarding the use of such techniques, both in
ence the behavior of horses, in-hand and under saddle, with relation to efficacy and on welfare grounds.
stimuli from their hands on the reins and their legs on the sides of Several scientific study results as well as the personal opinions of
the horse and more discreetly with the use of their seat, weight leading riders, trainers and veterinarians were summarized during a
position and movement. Horse riding at its most humane relies on special meeting held in Lausanne (2006). Most of the presented
a clear understanding of the way in which the horse makes these studies related to positive and negative influences on different
associations. bony and soft tissue anatomical structures (Lausanne 2006), and
When problems occur during training, these usually manifest few addressed concerns regarding welfare implications. Since the
as locomotory behaviors designed to remove the horse from a Lausanne meeting some of these studies have been published

375
 16 Ethology and welfare aspects

(McGreevy & McLean, 2007). All these effects could possibly have
psychological and welfare implications.

Breeding or training certain types of

locomotory responses and welfare implications
There are almost as many breeds of horses as there are different
reasons why horses are kept. However the majority have directly or
indirectly to do with the horse’s kinetic possibilities; some per-
formed in competition, others for recreational use (see Chapter
14). Those involved with competition have the additional risk of
reducing the horses welfare when human prestige and/or prize
money is involved. There are many different possible welfare issues
related to training methods, so only a selection of issues is men-
tioned below.
In Europe and America, those disciplines in which the horse com-
petes in age classes (or where early maturity is considered a competi-
tive advantage) produce foals as near to January first as possible
(Gibbs & Cohen, 2001). In many countries this coincides with the
middle of the winter, preventing most dams and foals from being
outside and exercising sufficiently thus increasing the chances of
future problems in relation to bone and muscle development as dis-
cussed earlier in this chapter. These horses also often start to train
and compete before they are fully grown and physically matured,
possibly making them more prone to injuries or forcing early
retirement.
Apart from the obvious risks in relation to the locomotory
demands involved in jumping high and fixed obstacles, horses
training or performing for long durations (endurance, eventing,
long marathons in four-in-hand driving) or in hot climates have
Fig 16.8  Hyperflexion according to the Fédération Equestre International extra challenges on their locomotion apparatus with welfare con-
(Lausanne, 2006). sequences. There are also concerns regarding the impact of fre-
© Jacques Toffi – www.arnd.nl. quent and often long journeys overland and by air, as horses are
transported internationally for competition and breeding pur-
poses. Research has clearly indicated that even land journeys of a
short duration involve restriction on locomotion in relation to
(Gomez-Alvarez et al., 2006; Van Breda, 2006) and there are now a space and activity, and the accelerative forces due to the transport
number of new studies to add to the body of knowledge (Sloet van process and imposed on the horse, cause it to adopt a ‘bracing’
Oldruitenborgh-Oosterbaan et al., 2006; Rhodin, 2008; Von Borstel posture to allow it to balance in a moving vehicle (see Waran
et al., 2009). et al., 2002). Apart from the increased heart rate, cortisol levels
A recent analysis (Visser & VanDierendonck, 2009) of these and oxygen consumption during transport, such postures as seen
studies revealed that there are some methodological issues with during transport are associated with elevated muscle enzymes
many of the current studies, and there is still insufficient scientific indicating muscle damage. Transport is a physical challenge for
evidence to confirm unequivocally whether or not there are welfare the performance horse, and as such needs to be considered as a
issues involved in training techniques using hyperflexion or others welfare risk.
involving a very high carriage of the head and neck, as compared There are also the welfare considerations related to the breeding
with training methods advocating a relaxed or ‘on-the-bit’ head and and training of horses used in competitions where primarily the
neck position (see Chapter 15, Fig. 15.19). subjective ‘beauty’ of the horses’ movements is judged (e.g. Arabi-
Additionally some aspects did not receive much attention in the ans, Hackneys, American Saddlebreds, Warmbloods bred for dres-
latest research, even though they may have an important bearing sage, Icelandics, and Friesians). There is concern regarding the use
on the perception of the horse. For example, hyperflexion most of inhumane training methods to encourage the horses to lift their
likely severely restricts the horse’s vision in the direction of travel limbs higher or wider or differently as required by the breed stan-
(McGreevy, 2004), could cause airway obstruction (changes in arte- dard and competition. Some of these methods include keeping the
rial O2, and CO2 concentrations and increased heart rates), could animals in the dark until seconds before performing as seen in Arab
cause physiological exhaustion and may disturb the horse’s biome- showing (making use of slow light adaptation of the horses pupil)
chanical equilibrium (King et al., 1994; Brown et al., 2004). The to all kinds of additions to or underneath the shoes and gadgets
possibility that these two impediments can lead to a state called designed to attach to the horse’s limbs to train a new way of moving
‘learned helplessness’ is still open to debate, but has not been seri- (e.g. see Fig. 16.9). Interestingly most of the unethical methods are
ously addressed by any study to date. either frowned upon or disallowed, yet they are difficult to police
In terms of training locomotory responses, the concern is due to and as such continue to be used. It is clear that the welfare concerns
the confusion induced in the horse by applying one cue/signal regarding the training of horses to perform unnatural gaits is an
(pulling the reins with both hands) for two responses (slowing/ area for more research and public awareness.
stopping versus over-flexing the neck). Such confusion is considered Within jumping there are many different unwanted training
to lead to a dangerous detraining of the deceleration (stop) response. methods that have been the subject of media attention and are still
Furthermore, the subsequent confusion can induce conflict resolu- mentioned regularly. These methods are used to get the horse to
tion behaviors in horses. This can prompt horses to escalate their jump higher and to tuck the front or back limbs up higher to avoid
active coping strategies such as causing them to trial hyper-reactive knocking an obstacle. Such methods include the use of caustic
predator removal behaviors (such as bucking, rearing and shying) substances underneath the protective leg bandages, lifting of poles

376
 References

Concluding remarks – new approaches

including equitation science

Several welfare issues have been the subject of the previous para-
graphs, challenging scientists to find acceptable solutions to train-
ing and management issues. As van Weeren (2008) points out, the
equestrian industry is a conservative one slow to adopt the available
technological and scientific advances to help progress training
and better safeguard equine welfare. This has been recognized
as an emerging area combining many fields of equine research
called equitation science (McGreevy, 2007; Goodwin et al., 2008;
McGreevy & McLean, 2007).
Ever since horses were domesticated people have managed their
behavior in a restrictive manner. This has had and still has serious
consequences for the horse. As discussed, research has shown that
foals post partum need enough ‘Steppe like’ space and social stimu-
lation to exercise and develop sufficiently (Kurvers et al., 2006;
Brama et al., 2001, 2009) (see Chapter 13) and it is clear that
modern demands on the horse differ greatly from the requirements
of horses in previous times (see Chapter 1).
Technologies which can be attached to the horse during training
and use telemetry to transfer data regarding the horse and the
horse–rider interaction (e.g. tension meters in the reins, and pres-
sure sensors in the stirrups (Warren-Smith et al., 2006), pressure
meters in saddle pads (de Cocq et al., 2004; Geutjes et al., 2008),
on the limbs or head (Warren-Smith, pers. comm.) and heart (Von
Borell et al., 2007) all have the potential to provide us with more
objective empirical information to aid with training and manage-
ment. In addition detailed experiments using force places, kine-
Fig 16.9  The use of so called ‘shoes’ in training Hackneys in which the matic analyses with motion cameras (see Chapter 2) (van Weeren,
strap between the two front legs is too short, forcing the horse to lift its 2002) and other such advances will lead toward better ergonomics
opposite leg earlier and higher than natural, as well as that the horse cannot (van Weeren, 2005), prevent overtraining (de Graaf-Roelfsema,
stand on both feet at the same time. 2007) and stimulate better skeletal development in the horse
© Arnd Bronkhorst/www.arnd.nl. (Rogers et al., 2008a,b). Add to this advanced practice involved in
the application of ethologically sound training practices using
knowledge of learning processes and the horse’s natural behavioral
(rapping) as a horse jumps an obstacle, and other such painful responses (McGreevy & McLean, 2007), we will develop a greater
stimuli to cause the horse to avoid touching the jumping poles. understanding of what is acceptable from the horse’s point of view.
Given that this is a welfare issue and banned by the professional There are many and varied demands made upon the horse by
body, the extent to which this occurs and the effect on the (short humans. All forms of performance and recreational use of the horse
and long term) welfare of the horse is largely unknown. A recent relate to its locomotory behavior – and in all cases humans have
issue concerns the use of different performance enhancing boots on modified the horse’s normal behavior to accommodate our needs.
distal limbs in jumping horses. A study by Murphy (2008) revealed The breeding, management and training of horses to enable humans
that weight and extra pressure does indeed alter the flexion of the to enjoy interacting with the horse often compromise the horse’s
hind legs above the jump and according to Murphy: ‘In the absence welfare. Recognition of this fact, and questioning the way in which
of scientific appraisal, it is unclear if such boots are acceptable and we apply such pressure on horses is an important step forward in
innovative training aids within equitation’. improving our relationship with them.

References
Barneveld, A., van Weeren, R., Knapp, J., 1999. evolution, development and management Effect of laryngoplasty on respiratory noise
Influence of early exercise on the of its behavior. Cambridge University Press, reduction in horses with laryngeal
locomotion system. Paper presented at the pp. 55–82. hemiplegia. Equine Vet. J. 36 (5), 420–425.
50th Meeting of the EAAP (Zurich). Brama, P.A., Bank, R.A., Tekoppele, J.M., Van Christensen, J.W., Ladewig, J., Sondergaard, E.,
Berger, J., 1986. Wild horses of the Great Weeren, P.R., 2001. Training affects the Malmkvist, J., 2002. Effects of individual
Basin: social competition and population collagen framework of subchondral bone versus group stabling on social behavior in
size. The University of Chicago Press, in foals. Vet. J. 162 (1), 24–32. domestic stallions. Appl. Anim. Behav. Sci.
Chicago and London, pp. 181–195. Brama, P.A., Holopainen, J., van Weeren, P.R., 75 (3), 233–248.
Bertolucci, C., Giannetto, C., Fazio, F., Firth, E.C., Helminen, H.J., Hyttinen, M.M., Cooper, J.J., Mcdonald, L., Mills, D.S., 2000.
Piccione, G., 2008. Seasonal variations in 2009. Effect of loading on the organization The effect of increasing visual horizons on
daily rhythms of activity in athletic horses. of the collagen fibril network in juvenile stereotypic weaving: implications for the
Animal 2, 1055–1060. equine articular cartilage. J. Ortho. Res. 27 social housing of stabled horses. Appl.
Boyd, L., Keiper, R., 2005. Behavioral ecology (9), 1226–1234. Anim. Behav. Sci. 69 (1), 67–83.
of feral horses. In: Mills, D., McDonnell, Brown, J.A., Derksen, F.J., Stick, J.A., Cooper, J.J., McGreevy, P., 2002. Stereotypic
S.M. (Ed.), The domestic horse: the Hartmann, W.M., Robinson, N.E., 2004. behavior in the stabled horse: causes effects

377
 16 Ethology and welfare aspects

and prevention without compromising environmental preferences. Proceedings of (Ed.), Welfare of horses. Kluwer Academic
horse welfare. In: Waran, N. (Ed.), The the 5th ISAE North American Regional Publishers, pp. 77–97.
welfare of horses. Kluwer Academic Meeting, University of Guelph, Ontario Murphy, B.A., 2008. Equine Chronobiology
Publishers, pp. 99–124. Canada (June 10–11th). – An emerging scientific field with
Cran, K.K., Smith, M.A., Reynolds, D., 1997. Houpt, K.A., Ogilvie-Graham, T.S., 2002. implications for health and performance in
Habitat selection patterns of feral horses in Comfortable quarters for horses at research the horse. In: Proceedings of the IVth
Southcentral Wyoming. J. Range Manage. institutions. In: Reinhardt, V., Reinhardt, A. conference of the International Society of
50 (4), 374–380. (Eds.), Comfortable quarters for laboratory Equitation Science, Dublin July 1–3.
De Cocq, P., van Weeren, P.R., Back, W., 2004. animals. Washington Animal Welfare Murphy, J., 2008. Boots on horses: limb
The effects of a girth, a saddle and weight Institute. protection or hyperflexion training aids in
on the movement of the horse. Equine Vet. Joubert, E., 1972. Activity patterns shown by the Showjumping horse. J. Appl. Anim.
J. 36, 758–763. mountain zebra Equus zebra hartmannae Welfare Sci. 11 (3), 223–228.
De Graaf-Roelfsema, E., 2007. in South West Africa with reference to Nicol, C., Badnell-Waters, A., Bice, B., Kelland,
Endocrinological and behavioral climatic factors. Zoologica Africana 7 (1), A., Wilson, A., Harris, P., 2005. The effects
adaptations to experimentally induced 309–331. of diet and weaning method on the
physical stress in horses. PhD thesis Kiley-Worthington, M., 1987. The behavior of behavior of young horses. Appl. Anim.
Veterinary Faculty Utrecht. horses. J.A. Allen, London. Behav. Sci. 95 (3–4), 205–221.
De Graaf-Roelfsema, E., Veldhuis, P.P., Keizer, King, C.M., Evans, D.L., Rose, R.J., 1994. Rhodin, M., 2008. A biomechanical analysis of
H.A., van Ginneken, M.M.E., van Dam, Cardiorespiratory and metabolic responses relationship between the head and neck
K.G., Johnson, M.L., et al., 2009. to exercise in horses with various position, vertebral column and limbs in
Overtrained horses alter their resting abnormalities of the upper respiratory tract. the horse at walk and trot. PhD thesis.
pulsatile growth hormone secretion. Am. J. Equine Vet. J. 26 (3), 220–225. Swedish University of Agricultural Sciences,
Physiol.-Regul. Int. Compara. Physiol. 297 Kroekenstoel, A.M., van Heel, M.C.V., van Uppsala.
(2), 403–411. Weeren, P.R., Back, W., 2006. Rivera, E., Benjimen, S., Nielsen, B., 2002.
Feist, J.D., McCullough, D.R., 1976. Behavior Developmental aspects of distal limb Behavioral and physiological responses of
patterns and communication in feral conformation in the horse: the potential horses to initial training: comparison
horses. Zeitschrift fur Tierpsychologie 41, consequences of uneven feet in foals. between pastured versus stalled horses.
337–371. Equine Vet. J. 38, 652–656. Appl. Anim. Behav. Sci. 78, 235–252.
Fraser, A.F., 1992. The behavior of the horse. Kurvers, C.M.H., van Weeren, P.R., Rogers, Rogers, C.W., Firth, E.C., McIlwraith, C.W.,
CAB International, Oxon, UK. C.W., VanDierendonck, M.C., 2006. Barneveld, A., Goodship, A.E., Kawcak,
Geutjes, C.A., Clayton, H.M., Kaiser, L.J., 2008. Quantification of spontaneous locomotion C.E., et al., 2008a. Evaluation of a
Forces and pressures beneath the saddle activity in foals kept in pastures under new strategy to modulate skeletal
during mounting from the ground and various management conditions. Am. J. Vet. development in Thoroughbred
from a raised mounting platform. Vet. J. 3, Res. 67, 1212–1217. performance horses by imposing track-
137–140. Lausanne, 2006. The use of over bending based exercise during growth. Equine
Gibbs, P.G., Cohen, N.D., 2001. Early (“Rollkur”) in: FEI Competition Report of Vet. J. 40 (2), 111–118.
management of race-bred weanlings and the FEI Veterinary and Dressage Rogers, C.W., Firth, E.C., McIlwraith, C.W.,
yearlings on farms. J. Equine Vet. Sci. 21 Committees’ Workshop. 31 January 2006 Barneveld, A., Goodship, A.E., Kawcak,
(6), 279–283. during the FEI Veterinary Committee C.E., et al., 2008b. Evaluation of a new
Gomez-Alvarez, C.B., Rhodin, M., Bobber, meeting at the Olympic Museum, strategy to modulate skeletal development
M.F., Meyer, H., Weishaupt, M.A., Johnston, Lausanne. in racehorses by imposing track-based
C., et al., 2006. The effect of head and neck Linklater, W.L., Cameron, E.Z., Stafford, K.J., exercise during growth: the effects on
position on the thoracolumbar kinematics Veltman, C.J., 2000. Social and spatial 2- and 3-year-old racing careers. Equine
in the unridden horse. Equine Vet. J. Suppl. structure and range use by Kaimanawa wild Vet. J. 40 (2), 119–127.
36, 445–451. horses (Equus Caballus: Equidae). N. Z. J. Rushen, J., Lawrence, A., Terlouw, E.M.C.,
Goodwin, D., 1999. The importance of Ecol. 24 (2), 139–152. 1993. The motivational basis of
ethology in understanding the behavior of McBride, S., Long, L., 2001. Management of stereotypes. In: Lawrence, A. B., Rushen, J.
the horse. WALTHAM Symposium. Equine horses showing stereotypic behavior: owner (Eds.), Stereotypic anmial behavior. CAB
Vet. J. Suppl. 28, 15–19. perception and implications for welfare. International, Wallingford, UK, pp. 41–64.
Goodwin, D., 2002. Horse behavior: Vet. Rec. 148 (26), 799–802. Sigurjónsdóttir, H., Van Dierendonck, M.C.,
evolution, domestication and feralisation. McGreevy, P., 2004. Equine behavior, a guide Snorrason, S., Thorhallsdottir, A.G., 2003.
In: Waran, N. (Ed.), Welfare of horses. for veterinarians and equine scientists. Social relationships in a group of horses
Kluwer Academic Publishers, pp. 1–18. Saunders, Edinburgh, UK. without a mature stallion. Behav. 140 (6),
Goodwin, D., Hughes, C.F., 2005. Equine Play McGreevy, P.D., 2007. The advent of 783–804.
Behavior. In: Mills, D., McDonnell, S.M. equitation science. Vet. J. 174, 492–500. Sloet van Oldruitenborgh-Oosterbaan, M.M.,
(Eds.), In the domestic horse: the origins, McGreevy, P.D., McLean, A.N., 2007. The Blok, M.B., Begeman, L., Kamphuis,
development and management of its roles of learning theory and ethology in M.C.D., Lameris, M.C., Spierenburg, A.J., et
behavior. Cambridge University Press, pp. equitation. J. Vet. Behav. Clin. Appl. Res. al., 2006. Workload and stress in horses:
150–160. 2, 108–118. comparison in horses ridden deep and
Goodwin, D., McGreevy, P., Heleski, C., McGreevy, P.D., Thomson, P.C., 2006. round (‘rollkur’) with a draw rein and
Randle, H., Waran, N., 2008. Equitation Differences in motor laterality between horses ridden in a natural frame with only
science: the application of science in breeds of performance horse. Appl. Anim. light rein contact. Tijschrift Voor
equitation. J. Appl. Anim. Welfare Sci. 11 Behav. Sci. 99 (1-2), 183–190. Diergeneeskunde 131 (5), 152–157.
(3), 185–191. McGreevy, P.D., Rogers, L.J., 2005. Motor Spinka, M., Newberry, R.C., Bekoff, M., 2001.
Houpt, K.A., Houpt, T.R., 1989. Social and and sensory laterality in thoroughbred Mammalian play: training for the
Illumination preference of mares. J. Anim. horses. Appl. Anim. Behav. Sci. 92 (4), unexpected. Quart. Rev. Biol. 76 (2),
Sci. 66, 2159–2164. 337–352. 141–168.
Houpt, K.A., Lee, J., Houpt, T.R., 2000. Mills, D.S., Clarke, A., 2002. Housing, Sweeting, M.P., Houpt, C.E., Houpt, K.A.,
Consumer demand theory of equine management and welfare. In: Waran, N. 1985. Social facilitation of feeding and

378
 References

time budgets instabled ponies. J. Anim. Sci. Takhi or Przewalski Horses (Equus ferus activity for assessing stress and welfare in
60 (2), 369–374. przewalskii) in Mongolia. Appl. Anim. farm animals – a Review. (2007). Physiol.
Thorne, J.B., Goodwin, D., Kennedy, M.J., Behav. Sci. 50 (2), 95–114. Behav. 92 (3), 293–316.
Davidson, H.P.B., Harris, P., 2005. Foraging VanDierendonck, M.C., de Graaf-Roelfsema, von Borstel, U.U., Duncan, I.J.H., Shoveller,
enrichment for individually housed horses: E., van Breda, E., Wijnberg, I.D., Keizer, A.K., Merkies, K., Keeling, L.J., Millman,
Practicality and effects on behavior. Appl. H.A., van der Kolk, J.H., et al., 2007. S.T., 2009. Impact of riding in a coercively
Anim. Behav. Sci. 94 (1-2), 149–164. Intensified training induces ethological obtained Rollkur posture on welfare and
Tyler, S., 1972. The behavior and social effects in Standardbreds. In: fear of performance horses. Appl. Anim.
organisation of the New Forest ponies. Endocrinological and behavioral Behav. Sci. 116 (2-4), 228–236.
Anim. Behav. Mono. 2, 86–194. adaptations to experimentally induced Waran, N.K., 2002. (Ed.), The welfare of
van Breda, E., 2006. A nonnatural head-neck physical stress in horses. E de Graaf- horses. Springer, The Netherlands.
position (Rollkur) during training results in Roelfsema, Utrecht: PhD thesis, pp. Waran, N.K., Leadon, D., Friend, T., 2002. The
less acute stress in elite, trained, dressage 168–188. effects of transportation on the welfare of
horses. J. Appl. Anim. Welfare Sci. 9 (1), VanDierendonck, M.C., Goodwin, D., 2005. horses. In: Waran, N. (Ed.), Welfare of
59–64. Social contact in horses: implications for horses. Kluwer Academic Publishers, pp.
van Heel, M.C.V., Kroekenstoel, A.M., human-horse interactions. The Human- 125–150.
VanDierendonck, M.C., van Weeren, P.R., animal relationship: forever and a day. Waring, G.H., 2003. Horse Behavior, second
Back, W., 2006. Uneven feet in a foal may Royal van Gorcum, Assen, pp. 65–82. ed. Noyes Publications/William Andrew
develop as a consequence of lateral grazing Vervuert, I., Coenen, M., 2002. Feeding and Publishing, Norwich, NY.
behavior induced by conformational traits. housing management in horses. Warren-Smith, A.K., Curtis, R.A., Greetham, L.,
Equine Vet. J. 38, 646–651. Pfedeheilkunde 18 (6), 629–632. McGreevy, P.D., 2006. Rein contact between
van Heel, M.C.V., VanDierendonck, M.C., van Weeren, P.R., 2008. How long will horse and handler during specific
Kroekenstoel, A.M., Back, W., 2010. equestrian traditionalism resist science? equitation movements. Appl. Anim. Behav.
Lateralised behavior leads to increased (Guest editorial). Vet. J. 175, 289–290. Sci. 108 (1-2), 157–169.
unevenness in front feet and asymmetry van Weeren, R., 2005. Equine ergonomics: a Weiss, J.M., 1971. Effects of coping behavior in
in performance in young adult new era? (Editorial). Equine Vet. J. 37 (1), different warning signal conditions on
Warmblood horses. Equine Vet. J. 42 (5), 4–6. stress pathology in rats. J. Compara.
444–450. Visser, K., VanDierendonck, M.C., 2009. A Physiol. Psychol. 77, 1–13.
van Weeren, P.R., 2002. The clinical note on the welfare implications of Wilson, A.M., McGuigan, M.P., Su, A., van
applicability of automated gait analysis hyperflexion in the training of the den Bogert, A.J., 2001. Horses damp the
systems. Equine Vet. J. 34 (3), 218–219. ridden horse. Commissioned by Dutch spring in their step. Nature 414 (6866),
VanDierendonck, M.C., 2006. The importance Ministry of Agriculture, Nature and Food 895–899.
of social relationships in horses. PhD Quality. Zharkikh, T.L., 1999. Development of the
thesis, Utrecht University. Von Borell, E., Langbein, J., Despres, G., Przheval’skii horse, Equus przewalskii
VanDierendonck, M.C., Bandi, N., Batdorj, D., Hansen, S., Leterrier, C., Marchant-Forde, J., (Perissodactyla) play behavior at the
Dugerlham, S., Munkhtsog, B., 1996. et al., 2007. Heart rate variability as a “Askania-Nova” reserve. Zoologicheskii
Behavioral Observations of Reintroduced measure of autonomic regulation of cardiac Zhurnal. 78 (7), 878–884.

379
This page intentionally left blank
 C H A P T ER 17 

Rehabilitation of the locomotor

apparatus
Narelle Stubbs, Eveline Menke, Willem Back, Hilary M. Clayton

General introduction subjective clinical outcome measures of performance and function

are also vital in the rehabilitation setting which may include force
plates to measure and monitor neurological signs and or lameness
Equine sport sciences are evolving along similar lines to human during the rehabilitation process (Clayton et al., 2003; Ishihara
sports sciences. There has been considerable research on normal et al., 2009) and motion analysis to measure intervertebral and
and pathological gaits, and the high prevalence of locomotor inju- limb motion as described elsewhere in the text. For example, Faber
ries in athletic horses is well established (Ely et al., 2009; Scott, and colleagues presented a single case study in a dressage horse
2008). Physical therapy and rehabilitation play an important role with scoliosis at T10–L1 that caused severe loss of performance. After
in performance enhancement, injury prevention and restoration of two sessions of thoracolumbar (TL) manipulation the effects were
full function during recovery from injury. This is particularly rele- assessed using a validated kinematic analysis protocol (Faber et al.,
vant in relation to back pain and lameness prevention strategies in 1999, 2001) to measure TL motion and spatiotemporal parameters
horses participating in exercise or performance-related activities, with the horse walking and trotting at standardized speeds on a
including racing, and in horses stabled in smaller operations for treadmill. Spinal motion was quantified by range of motion in the
pleasure activities, as there is increased risk of lameness for active three planes of motion and intervertebral pattern symmetry (Faber
horses (n = 3925 horses in 138 randomly selected yards; Ross et al., et al., 2000). Data were collected pre-treatment, 48 h after treat-
1998). As described in previous chapters, numerous studies have ment and at weeks 3, 7 and 36. Treatment involved ‘direct short-
described factors that may predispose the horse to injury, lameness lever thrust towards the neutral position of the areas where an
and lack of or loss of performance (Burns et al., 2006). These impaired mobility had been noted during the clinical examination’,
include: morphology; conformation; training; environmental con- labeled as ‘orthomanual realignment technique’. Progressive mobi-
ditions; type of competition; age and sex; performance limitations; lization exercises were also given (walking along serpentines).
exercise effects on the neuromuscular and skeletal systems; and The aim of this review is to summarize equine rehabilitation from
interaction between rider and horse. Rehabilitation encompasses an evidence-based perspective utilizing references from databases,
broad-based concepts that relate to all the aforementioned factors such as Pubmed. There is a long history of physical therapy in the
with a focus on tissue healing, biomechanics and neuromotor treatment of animals. A book on physiotherapy for horses, written
control, which all ultimately affect equine locomotion. Veterinary by the physiotherapist Sir Charles Strong, was published in 1967 in
sports medicine and rehabilitation follows in the footsteps of the the UK. Two other early publications on physiotherapy for horses
human counterparts encompassing many professional fields and appeared in the 1970s (Hopes, 1970; Downer, 1978) focusing on
areas of research, including: physiotherapy (PT), osteopathy, chiro- back pain.
practic, and other complementary medicines. In this chapter, back pain will be used as an example to illustrate
Objective measures of locomotor kinematics and kinetics (see the principles and practice of rehabilitation. Fundamental concepts
Chapter 2) are useful in monitoring progress and obtaining outcome of neuromotor control and rehabilitation will be addressed in the
measures in rehabilitation, both clinically and for research pur- context of equine sports medicine.
poses. Conventional diagnostic equipment can also be used in
rehabilitation including electromyography (EMG) (Cottriall et al.,
2008) and ultrasonography (McGowan et al., 2007b; Stubbs et al.,
2010, 2011), along with relatively inexpensive functional measure-
Historical background: rehabilitation defined
ment tools such as pressure algometry (Haussler& Erb, 2003, The word rehabilitation comes from the Latin ‘rehabilitare’ meaning
2006a,b; Haussler, 2006; Varcoe-Cocks et al., 2006; Sullivan et al., to make fit again. Traditionally this encompasses treatments
2008; De Heus et al., 2010), goniometry (Liljebrink & Bergh, 2010), designed to facilitate the process of recovery and restoration to a
digital photography, videography and the tape measure. However, former capacity following an illness or injury. These are the found-
as shown in human studies many of the measurement tools such ing principles of the physiotherapy profession as described by the
as goniometry, used to document limb passive motion at the World Confederation for Physical Therapy. Complementary medi-
fetlock, carpus and hock are only reliable when used by the same cine including PT in all species is concerned with identifying and
investigator, with further studies needed to validate this tool in the maximizing quality of movement potential within the spheres of
horse (Liljebrink & Bergh, 2010). Complementary objective and promotion of health, prevention of injuries, treatment/intervention,

381
 17 Rehabilitation of the locomotor apparatus

rehabilitation and, more recently, performance enhancement and pain and dysfunction techniques (Simons et al., 1999), electro-
sports medicine. Physiotherapy has been defined as ‘A holistic therapy (Watson, 2008) and integrative therapeutic approaches
approach to the prevention, diagnosis and therapeutic management relating to mechanical passive constraints of locomotion and neu-
of pain, disorders of movement or optimization of function to romotor control including facilitation and strengthening techniques
enhance the health and welfare of the community from an indi- (McGill, 2007; Mooney, 2007; Lee & Vleeming, 2007).
vidual or population perspective’. McGowan et al. (2007c) empha-
sized that one of the fundamental differences between PT and the
medical or veterinary profession is that physiotherapists are trained
to focus on the assessment and management of a patient’s function,
Longevity and musculoskeletal disorders
rather than focusing purely on the specific patho-anatomical diag- The clinical need for rehabilitation interventions and justification
nosis. The aim of PT and rehabilitation is to restore function and for research into this field are verified by the amount of wastage in
promote tissue healing by assisting normal physiological processes equine sports due to musculoskeletal injuries (Peloso et al., 1994;
through the application of manual therapy techniques, electromo- Valentine, 2008; Barr et al., 2009; Ely et al., 2009). Statistics related
dalities and exercise-based regimens. Buchner and Schildboeck to longevity and life spans are often collated according to insurance
(2006) and McGowan et al. (2007c) recognized these specific areas: company records. These indicate that diseases of the musculoskel-
manual/manipulative therapies (PT, massage, chiropractic, osteopa- etal system are the predominant cause of death in sport horses
thy) and complementary alternative medicine (CAM) also known (Clausen et al., 1990; AGRIA, 1995 cited in Wallin et al., 2000;
as physical or technical/modality based therapies (physiotechnical Heisele, 1995), whereas diseases of the digestive system are the most
therapies): acupuncture, electrotherapy, exercise, hydrotherapy, laser frequently reported cause of death in companion horses (Baker &
therapy, magnetic field therapy, thermotherapy and therapeutic Ellis, 1981). In a follow-up 5-year survival study of 2495 horses with
ultrasound. at least one costly veterinary-care event and 15 576 horses with no
Rehabilitation in veterinary medicine involves the veterinarian costly veterinary care event evaluated for 1 year, the risk of death
as primary patho-anatomical diagnostician, a thorough objective increased linearly with age and with increasing life-insurance value
functional assessment of the patient and consultation with other (Egenvall et al., 2006). Horses with previous lameness had the
health professionals (McGowan et al., 2007a). Using knowledge lowest survival. Horses with previous locomotor problems contin-
and skills unique to these professions the patient’s movement ued to have considerably more veterinary-care events and higher
potential is assessed, with all information and confounding factors costs for locomotor problems during the follow-up period (Egen-
being incorporated to establish an accurate functional diagnosis, vall et al., 2008), highlighting the potential need for further and
problem list, management plan and goals. Clinically, and particu- improved rehabilitation strategies. The most common disease in
larly in relation to research, it is imperative that during the reha- 107 310 horses of varying age, gender, breed and use requiring vet-
bilitation process valid and reliable objective measurements are erinary care covered by complete insurance in Sweden from 1997
taken to determine accurate outcome measures. The integration of to 2000 was arthritis, which most frequently affected the fetlock
research into evidence-based practice continues to be a challenge in (28%) or multiple joints (16%) (Penell et al., 2005).
veterinary medicine and rehabilitation is no exception. Currently In a 5-year study of 5140 horses from 136 riding schools the
there is limited research in equine locomotion in relation to reha- overall yearly incidence rate was 1584 events of veterinary care per
bilitation, though many animal models have been used for human 10 000 horse-years at risk (Egenvall et al., 2009). The total and
rehabilitation research with the findings being integrated into diagnostic mortalities were 790 and 763 deaths per 10 000 horse-
current concepts of equine rehabilitation. years at risk. Rates varied substantially among riding schools. For
Textbooks and review articles in the area of veterinary PT and locomotor problems the rates were 1116 events of veterinary care
rehabilitation include Animal physiotherapy: assessment, treatment and and 524 deaths per 10 000 horse-years at risk. For the outcome
rehabilitation of animals (McGowan et al., 2007a). This is an evidence- veterinary care for locomotor problems, the hazard ratio (HR)
based textbook that reviews the literature in relation to animal PT increased with increasing life-insurance value, was higher in horses
and is a useful reference with respect to outlining the foundations than ponies, and was higher in Warmbloods than other horses. The
of rehabilitation and equine locomotion. A recent review Equine HR increased by 33% for each year of age at entry. Age at entry
Physiotherapy: the science behind the profession (McGowan et al., ≥8 years was associated with decreased HR due to locomotor
2007c) also highlights the potential input of PT in equine rehabili- problems.
tation. Another useful text is Physical Therapy and Massage for the Wallin et al. (2000) investigated the longevity, causes of death
Horse which complements both evidence based and clinical practice and culling of Swedish Warmblooded and Coldblooded horses via
(Denoix & Pailloux, 2005). In addition there are clinical and evi- a retrospective owner questionnaire. Data were retrieved from
dence based articles, chapters and textbooks (Porter, 2005; Bromi- horses born 1968–1986 participating in Riding Horse Quality Tests
ley, 2007; Henson et al., 2009; Pusey et al., 2010). The profile of as 4-year-olds, with information about the horses available until
rehabilitation and performance enhancement has also been height- 1990. Of the 1847 Warmbloods, 503 were dead, 85% had com-
ened in recent years by the official use of physiotherapists during peted in different sporting disciplines and many were used as
international equestrian competitions. In this context, equestrian leisure horses after retirement from competition. Data also included
sports are rapidly catching up to other international competitive horses of the Swedish Cavalry Horse Foundation born between
sports, such as football and athletics, in which individuals and 1970 and 1975 with information available until 1989: 208/344
teams utilize the professional service of physiotherapists not only Warmbloods and 97/204 Coldbloods were dead. The most
in treatment of injuries, but also in maintenance and enhancement common causes of death were musculoskeletal diseases (56–57%),
of performance. respiratory diseases (8–9%), diseases of the digestive system (5–
In addition, there is a growing body of literature describing 6%), accidents (3–9%), and causes unknown in 13.0% of Warm-
research studies in equine rehabilitation (e.g. Wakeling et al., 2006; blooded horses. In the Coldblooded horses death was attributable
Haussler et al., 2007, 2010; Ramon et al., 2004; Wennerstrand et al., to temperamental disorders (23%), diseases of the musculoskeletal
2006; Xie, 2005; Clayton et al., 2008, 2010a,b, 2011a,b; De Heus system (14%) and hoof diseases (8%). A written questionnaire
et al., 2010; Stubbs et al., 2010, 2011). Many forms of PT, manual among (Swiss) horse owners (n = 2912 horses and ponies) indi-
therapies and alternative medicine can play a role in the horse’s cated that a veterinarian examined 718 horses (24.7% of the sample
return to optimal performance. Human therapeutic texts should be population) within the 12 months prior to the survey (Knubben et
consulted to further the reader’s understanding of manual therapies, al., 2008). Orthopedic and traumatic disorders (41.5%) had the
soft tissue mobilization/massage (Boyling et al., 2004), myofascial largest proportion, followed by gastrointestinal (27.1%) and

382
 General introduction

respiratory (14.0%) diseases. Half of the lameness cases occurred Gibson et al. (2002) reported the frequency of soft tissue injury
as a direct consequence of an injury. In 25.6% of all cases diag- in 70 elite sport horses competing at the 2000 Sydney Olympics in
nosed by a veterinarian, alternative therapeutic methods were used dressage (8), show jumping (15), and eventing (47). As reported
either in addition to traditional medicine or exclusively (Knubben previously (Reef, 1998), suspensory desmitis (n = 4) was most
et al., 2008). common in dressage horses affecting both fore and hind limbs.
A UK study surveyed owners of registered dressage horses (n = Some were able to compete despite having suspensory desmitis,
2554) reporting that 33% of horses had been lame at some time others had to withdraw. The most common lesion in show jumpers
during their career, with 24% of these within the previous 2 years (n = 10) was suspensory desmitis including desmitis of the extensor
(Murray et al., 2010). A number of factors were associated with the branches, with four medalists amongst this group. The authors
occurrence of lameness in the last 2 years, with increased risk for suggest that show jumpers can continue to compete successfully
older and bigger horses; use of arenas that were indoors, not pri- with good management despite chronic low-grade injury. Tendon
vately owned, and that became deeper in wet conditions or were and ligament injuries occurred more frequently in eventers due to
sand-based; use of and longer time spent in horse-walkers (cause the requirement for speed and jumping. Soft tissue lesions were
or effect); not lunging; shorter turnout time; and back pain. present in 45/47 horses that were presented to the veterinary clinic,
It is well established in the literature that flexor tendon and sus- 10 horses had multiple affected sites, most commonly superficial
pensory ligament injuries are the most frequently reported injuries digital flexor tendonitis and/or suspensory desmitis (proximal or
in sports requiring the horse to perform at high speed and over branches), tenosynovitis, desmitis, core lesions and diffuse loss of
jumps (Rooney & Genovese, 1981; Jeffcott et al., 1982; Rossdale echogenicity. Many did not complete due to retiring on course (n
et al., 1985; Mohammed et al., 1991; Marr et al., 1993; Colbourn = 10), elimination on course (n = 3) or withdrawal before the third
& Yovich, 1994; Reef, 1998). In the Thoroughbred (TB) racing inspection (n = 6). However, 20 horses passed the veterinary inspec-
industry musculoskeletal injuries account for three times more tions and completed the event, whilst adhering to the veterinary
wastage than all other medical problems (Rossdale et al., 1985; rules of the Fédération Equestre Internationale (FEI) (2011) with
Robinson & Gordon, 1988). More than half of the 2–4-year-old respect to non-pharmaceutical treatments (www.fei.org). Fourteen
racehorses become lame and 20% of all racehorses eventually suffer horses were presented for veterinary evaluation before the start of
a career ending musculoskeletal injury or disease (Bourke, 1995). the competition, only one of these completed all three phases
Ely et al. (2009) found that fractures, tendon and suspensory liga- (proximal suspensory desmitis), five did not complete because of
ment injuries were important causes of morbidity and mortality in the existing injury at the time of arrival in Sydney or during training
1223 National Hunt racehorses in training in the UK. Ramzan and leading up to the Games. The remaining eight horses fell or were
Palmer (2010) recently investigated musculoskeletal injuries in eliminated during competition for reasons other than a lesion.
three training yards (616 horses) in Newmarket (UK) finding a total A study in event horses training to compete in Concours Com-
of 248 injuries that occurred in 217 horses with fractures of the tibia plète Internationale (CCI) showed that 21% did not start due to
(20.7%) and proximal phalanx (14.5%) being the most common. injury: 43% had soft tissue injuries, 33.3% involving the superficial
A post mortem study of the third metacarpal and metatarsal bones digital flexor tendon and 30.6% involving the suspensory ligament
of 64 TB racehorses reported a 67% prevalence of palmar/plantar (Singer et al., 2008). Of those competing in the CCI, the most
osteochondral disease (Barr et al., 2009), which is considered to be common injuries included lacerations and abrasions to the carpus
a consequence of repetitive cyclical high intensive overload result- and stifle, superficial digital flexor tendonitis and exertional rhab-
ing in arthrosis (Pool, 1996). This painful condition is recognized domyolysis. These injuries were significantly higher in CCI competi-
in most racing breeds including TB (Arthur et al., 2003; Pilsworth, tions than at one-day events, most likely due to the increase in
2003), Standardbreds (Mitchell et al., 2003; Torre, 2003), Quarter demands of athletic performance.
Horses (Lewis, 2003) and Scandinavian Coldblooded trotters Scott (2008) performed an extensive literature review summariz-
(Ertola & Houttue, 2003). Overall, the incidence of catastrophic ing the musculoskeletal injuries in non-racing Quarter Horses
musculoskeletal injury of Thoroughbreds in the USA has been which included palmar foot pain, osteoarthritis of the proximal
reported as 1.2/1000 race starts (Hernandez et al., 2001). Incidence interphalangeal joints, pastern fractures, suspensory ligament des-
of injury was significantly higher for turf races (2.3/1000 starts) than mitis, osteoarthritis of the distal tarsal joints, stifle injuries and back
for dirt races (0.9/1000 starts). The number of days since the last pain. The review presents the current literature on the diagnosis,
race (≥33 days) was associated with a higher risk of injury. treatment and rehabilitation strategies including concepts of injury
A study of 265 Danish Standardbreds evaluated over a 5-month predisposition and prevention. A retrospective study by Dabareiner
period showed that a change of trainer affected the risk of lameness et al. (2005a) identified types of musculoskeletal problems associ-
(Vigre et al., 2002). Compared to the period in which horses had ated with lameness or poor performance in 118 horses used for
been with the same trainer for >3 months, horses that entered a barrel racing. The forelimbs were more frequently affected than the
different training regime within the past 1.5–2.5 months had a hind limbs, with forelimb foot pain and osteoarthritis of the distal
higher risk of lameness. Participation in races increased the risk of tarsal joints being the most common abnormalities. Reasons for
lameness significantly in the 5 days following a race. presentation were lameness (61%) or poor performance (39%). The
In a (USA) study of 357 lameness cases, 78.6% were reported to most common performance change (41%) was refusal or failure to
have recovered after a median duration of 18 days (Ross et al., turn properly around the first barrel. Dabareiner et al. (2005b) also
1999). Some type of treatment was administered in 82.9% of lame- reported on 118 team roping horses presenting with lameness or
ness incidents. Of 619 total treatments used, 53.2% were adminis- poor performance. A significantly greater proportion (74/118) of
tered or applied by a veterinarian. Horses experiencing hoof horses used for ‘heading’ (roping the steer around the horns) pre-
lameness were more likely to recover than those with other types sented for examination compared to (44/118) horses used for
of lameness. Horses that had participated in exercise-related activi- ‘heeling’ (roping around the hind limbs of the steer). Most horses
ties during the study period and prior to the development of lame- examined for poor performance were lame, however, the biome-
ness were more likely to recover. Treatment of the lameness was chanical loading patterns influenced the type of lameness: ‘headers’
associated with an increased likelihood of recovery. Cases with a had more right forelimb lameness (35%) compared with ‘heelers’
veterinarian involved in the diagnosis were associated with a (16%); ‘headers’ had a significantly greater proportion of bilateral
decreased likelihood of recovery and a longer duration of lameness, forelimb lameness (24%) compared with ‘heelers’ (9%), but
which might indicate that these cases were more complex or severe. ‘heelers’ had more bilateral hind limb lameness (7%) compared to
Although cases treated for lameness were more likely to recover, ‘headers’ (0%). The most common musculoskeletal problems for
treatment did not affect lameness duration (Ross et al., 1999). ‘headers’ were pain limited to the distal sesamoid (navicular) area,

383
 17 Rehabilitation of the locomotor apparatus

with or without osteoarthritis of the distal tarsal joints and soft These percentages were significantly higher than those recorded in
tissue injury in the forelimb proximal phalangeal (pastern) region. the same study for a control population of 399 horses, in which
The most common signs of pain in ‘heelers’ were in the navicular 20% were lame and 12% had back problems. Therefore there is a
area, osteoarthritis of the metatarsophalangeal joints and osteoar- strong association between lameness and back problems. Even
thritis of the distal tarsal joints. subtle hind limb lameness can cause changes in spinal kinematics
Veterinarians are more often consulted for the more complex at trot involving increased range of motion and hyperextension in
lameness cases, with 25% of the cases diagnosed by a veterinarian the TL spine and reduced range of motion in the lumbosacral region
being reported to receive alternative therapeutic methods (Knubben (Gomez Alvarez et al., 2007a).
et al., 2008). Physiotherapists who work in racehorse training yards It is widely reported that the TL spine is predisposed to damage
routinely treat horses’ backs and hindquarters. It has been reported and/or pain (Wennerstrand et al., 2004). A retrospective study by
that racehorses presented for PT showing pelvic bony asymmetry, Jeffcott (1980) reviewed 443 cases of equine back pain. The primary
muscle atrophy of the hindquarters and/or spasm or tenderness on pathological lesions associated with TL pain were vertebral lesions
palpation of the gluteal muscles should alert the physiotherapist to (38.6%), soft tissue injuries (25%), sacroiliac strain (13%) and
the possibility of impending pelvic or hind limb fracture (Hesse & non-TL lesions (13%). Vertebral lesions were predominantly crowd-
Verheyen, 2010). ing and over-riding dorsal spinous processes (DSP), which were
most common beneath the saddle from T12–T17 and were most
prevalent in young adult to middle aged horses used for jumping
Prevalence of neck and back pain or dressage and in Thoroughbreds (TB) with short backs. Soft tissue
lesions were predominantly in the longissimus dorsi muscles and
Equine back pain is a condition where PT and rehabilitation strate- supraspinous ligament in the caudal withers and cranial lumbar
gies are clinically useful and is an area of expanding research inter- regions. Specific causes of back pain that have now been identified
est. As discussed in Chapter 10, back problems are associated with include muscle strain (Jeffcott & Dalin, 1980; Piercy & Weller,
alterations of gait and performance (cause and/or effect). Improved 2009); ligamentous lesions (Jeffcott, 1980; Henson et al., 2007;
diagnostic capabilities combined with increased clinical and Tomlinson et al., 2003); fractures of the TL and/or lumbo-pelvic
research interest has raised awareness of the importance of equine complex (Sumner, 1948; Mason, 1971: Jeffcott & Whitwell 1976;
back pain. Historically, Jeffcott (1980) reported the prevalence of Vaughan & Mason 1976; Haussler & Stover, 1998; Driver & Pils-
equine back pain to be 0.9% in general veterinary practice, 2% in worth, 2009); vertebral body osteophytes and spondylosis (Geres,
TB racehorse practice, 5% in a veterinary school referral practice, 1978; Jeffcott, 1980; Haussler et al., 1999b; Meehan et al., 2009);
13% in a mixed equine practice (dressage, show jumpers, eventing), osteoarthritis and ankylosis of the inter-transverse and or lateral
and 47% in a spinal research clinic. These figures can be compared inter-transverse joints (Mitchell, 1933; Stecher & Goss, 1961;
with a reported rate of 94% at an equine chiropractic clinic Smythe, 1962; Haussler et al., 1999b); impingement of the DSPs
(Haussler, 1999a, 2000). Racehorse trainers in Sydney, Australia (Roberts, 1968; Jeffcott & Hickman, 1975; von Salis & Huskamp,
reported back problems to be in the top quartile ranked conditions 1978; Walmsley et al., 2002; Cousty et al., 2010); sacroiliac disease
(Bailey et al., 1997), and one of the most common injuries prevent- (Rooney, 1977; Jeffcott, 1980; Jeffcott et al., 1985; Haussler et al.,
ing training and racing. In a survey of registered dressage horse 1999b; Goff et al., 2008); degenerative intervertebral disc disease
owners in the UK, 25% (644/2554) of the respondents indicated (Hansen, 1959; Rooney, 1970; Taylor et al., 1977; Townsend et al.,
that their horse had a ‘back problem’ (Murray et al., 2010), though 1986; Denoix, 2007); and osteoarthritis of the synovial interverte-
the majority (80%) had not been diagnosed by a veterinary surgeon. bral articulations (facet joints) (Jeffcott, 1980; Haussler et al.,
Of those that sought treatment (447/644) complementary therapy 1999b; Girodroux et al., 2009; Cousty et al., 2010; Stubbs et al.,
was the most common (63%) followed by saddle fit (24%); veteri- 2010). Historically, the most widely documented osseous lesion is
nary involvement (20%); rest (20%); a change in training (13%); impingement or over-riding of the DSPs (Jeffcott, 1980; Walmsley
and other (6%) (Murray et al., 2010). From the 25% of horses et al., 2002; Erichsen et al., 2004; Cousty et al., 2010). More recently,
thought to have back pain, 2.5% received veterinary care alone and osteoarthritis of the facet joints has been identified as a source of
3% received a combination of veterinary care and complimentary pain and dysfunction (Denoix & Dyson, 2003; Girodroux et al.,
therapy, highlighting the need for a team approach to treatment and 2009; Stubbs et al., 2010). Nuclear scintigraphy showed moderate
owner education. Interestingly horses reported to have a ‘previous to intense increased radiopharmaceutical uptake in the facet joints
back problem’ that was resolved by complimentary therapy or rest of horses with pain in the region from T13–L1 compared with clini-
were more likely to have been lame in the last year. Historically cally normal horses (Gillen et al., 2009). However, only 61.5% of
Jeffcott (1979) reported that in 190 horses treated for chronic back 67 horses with back pain associated with radiographic evidence of
problems 57% recovered completely, 17% showed no improvement osteoarthritis had increased radiopharmaceutical uptake in one or
and 38% had a recurrence or continuation of signs of low-grade more facet joints, highlighting the inherent difficulties, even with
back pain. This highlights the need for research into rehabilitation advanced technology, in localizing the cause of back pain and loss
strategies and long-term management of chronic and recurrent back of function.
pain in horses. Haussler et al. (1999b) highlighted the potential under-diagnosis
Equine back pain often presents with more than one lesion or of TL vertebral or pelvic lesions in a post mortem study of 36 TB
problem area including limb lameness. Osseous lesions of the tho- racehorses that were euthanized for reasons unrelated to back pain.
racolumbar (TL) spine and the lumbopelvic complex are widely An alarming rate of osseous lesions was reported in the caudal
recognized as significant causes of equine back pain, poor perfor- thoracic and lumbar regions. Degenerative changes were observed
mance, loss of performance, and altered back and limb kinematics at lumbar intertransverse joints and sacroiliac articulations in all
(Jeffcott, 1975, 1980; Townsend et al., 1986; Denoix, 2005, 2007; specimens, with variable degrees of degenerative changes of the TL
Wennerstrand et al., 2004; Cousty et al., 2010; van Weeren et al., articular processes in 97% of specimens. Impingement of the DSPs
2010). Such lesions may arise as primary injuries to the vertebral (92%) and transverse processes (97%) was very prevalent, with
column and related structures, or secondary to other musculoskel- many specimens having widespread and severe osseous changes
etal injuries (Landman et al., 2004; Meehan et al., 2009; Girodroux including stress fractures of the facet joints. The relationship
et al., 2009). In a population of 805 horses (70% dressage, 20% between the changes observed at necropsy examination and the
show jumpers, 10% trotters) with orthopedic problems, 74% that presence of pain or loss of function could not be established. Stubbs
were presented with a back problem were lame and 32% of those et al. (2010) reported osseous pathologies in a group of 22 TB
presented for lameness had a back problem (Landman et al., 2004). racehorses euthanized at the Hong Kong Jockey Club for reasons

384
 General introduction

unrelated to back pain. Osseous lesions were graded as mild, mod- ligament injuries, muscle pain, wither injuries, polysaccharide
erate or severe at vertebral levels from T9–Ca1 in a non-cumulative storage myopathy (PSSM) and saddle fit problems. Sixteen horses
fashion. All horses exhibited lesions of moderate severity and 77% also had limb lameness. The treatment protocol included adminis-
had evidence of severe (grade 3) osseous pathology at various sites tration of systemic anti-inflammatories and/or a local anti-
in the TL spine and pelvis (Fig. 17.1). inflammatory injection and a variable combination of acupuncture,
Turner (2003) evaluated 5352 medical records from 1997 to 2002 chiropractic techniques, massage therapy, electro-stimulation, mag-
at a veterinary teaching hospital and found the occurrence of true netic therapy, therapeutic ultrasound, extracorporeal shockwave
back problems to be 2.2% in the lameness caseload of mixed breeds and training (exercise) management. Follow-up information was
and sporting disciplines. One hundred and twenty-four horses pre- available on 112 horses with 90% returning to work; however, 15
sented with the complaint of back pain with the diagnosis con- of those horses did not return to their previous level and were
firmed in 102 (82%) cases. An additional 16 horses that were retired or used for other activities. Of the 86 horses that returned
presented for other problems were diagnosed with back problems. to their previous level of work, 60% did not need further therapy
The most prevalent lesions involved the sacroiliac area (66). Other but the remaining 40% continued to receive some form of therapy.
problems included kissing spinous processes, dorsal spinous The need for therapy was based on the owner’s and trainer’s

A B

Fig 17.1  Severe osseous lesions and asymmetry in the lumbosacral/pelvic region in a Thoroughbred racehorse. (A) Sacral transitional that is ankylosed to
the pelvis on the right side of the specimen (left side of photo). The biomechanical effects of unilateral ankylosis on the lumbosacral and sacroiliac
biomechanics likely contributed to (B) a fracture with some displacement on the opposing sacral facet, lamina and articular pillar. (C) Severe sacroiliac
degenerative joint disease, including marked new bone formation on the ventral ilial surface and bilateral fracture lines, and (D) an intra-articular facet joint
fracture.

385
 17 Rehabilitation of the locomotor apparatus

impression of the horse’s behavior. The rate of successful return to pathology in horses with back pain but may also be due to pathol-
previous level of work is much higher than reported by Jeffcott ogy of the muscles themselves or to a generalized muscular disorder
(1979), which may reflect advancements in veterinary diagnostics, (Valberg, 1999; Quiroz-Rothe et al., 2002). It is suggested that this
medical management and the implementation of complementary is due to altered motor control as a result of the underlying lesion
therapies as part of the rehabilitation strategy. in the spine and/or due to peripheral joint disease with pain and
There is a smaller body of literature describing the equine cervical inflammation which causes reflex inhibition of motor neurons
spine, neck pain and pathology. Studies of cervical osseous patholo- resulting in weakness and atrophy of associated muscles (Young,
gies have focused mainly on cervical vertebral compressive myelop- 1993). Local muscle damage attributed to a poorly fitting saddle,
athy (van Biervliet et al., 2006; Levine et al., 2007). Arthropathy of for example, can also cause atrophy of the epaxial muscles (Gellman,
the cervical facet joints has been cited in the aetiology of reduced 1998; Harman, 1999). Figure 17.2 visually shows three examples
performance in the horse and has been reported to cause forelimb of asymmetrical hindquarter muscle development, or relative. The
lameness (Ricardi & Dyson, 1993), stiffness in movement, neck authors suggest objectively measuring muscle size using ultrasonog-
muscle atrophy, and neck pain (Beck et al., 2002; Dyson, 2003). A raphy, which may be a reliable tool for comparative measures across
recent examination of six cadaveric necks by Claridge et al. (2010) time and during the rehabilitation process. Research is currently
developed a three-dimensional model of the cervical facet joints underway using this modality in the horse. Jeffcott (1980) surveyed
based on radiographs and computed tomography (CT) images and 443 cases referred with TL disorders finding that 23.37% had evi-
used the model to determine that effusion within the articular facet dence of epaxial muscle pain. Stubbs et al. (2010) reported that
joints of the cervical spine is unlikely to cause compression of the there was a relationship between muscle function and pathology
spinal cord, which is known to be associated with neurological via ultrasonographic and necropsy analysis where significant
manifestations (Ricardi & Dyson, 1993). atrophy of the multifidus muscles was evident at the level and site/
In vitro kinematic studies of the cervical spine have demonstrated side of TL lesions in 22 racehorses.
that dorsoventral flexion and extension, axial rotation and lateral Valentine (2008) investigated the pathological findings in equine
bending take place at each of the intervertebral joints with the muscle (excluding PSSM) in 229 equids (217 horses and eight
largest ranges of motion in the upper (occiput to C2) and lower ponies of multiple breeds, three donkeys and one mule) over a 2.5-
(C5–T1) joints (Clayton & Townsend, 1989a). There is a general year period through necropsy and histopathology. Muscle lesions
reduction in overall cervical spinal mobility from foal to adulthood were present in 65% with the most common findings being chronic
(Clayton & Townsend, 1989b). In vivo kinematics of the cervical myopathic changes (15%), generalized muscle atrophy (13%),
spine have confirmed that most of the flexion–extension occurs at denervation atrophy (6%), and other lesions and pathologies that
the poll and the base of the neck when the horse voluntarily per- occurred less frequently (<5%) including myonecrosis, bone frac-
forms dynamic mobilization exercises to end range of motion in tures, bacterial infections, muscle rupture, selenium deficiency, exer-
cervical flexion (Clayton et al., 2010a) or in cervical lateral bending tional rhabdomyolysis, intramuscular protozoa, neoplasia, injection
(Clayton, et al., 2012). site reactions, lymphocytic infiltrates, ring fibers, fiber splitting, and
There have been few published anatomical descriptions of the fat infiltrations. The aetiology was undetermined in 4% of cases.
facet joints in the equine cervical spine, particularly with reference Hunt et al. (2008) performed an epidemiological study of myopa-
to the prevalence, clinical signs and pathological features of degen- thies in Warmblood horses in which the most common was PSSM
erative changes. Ultrasonographic imaging of the equine cervical (72/132 horses), followed by RER (7/132), neurogenic or myogenic
region provided a reference for normal appearance of cervical ver- atrophy (7/132) and non-specific myopathic changes (14/132).
tebrae, facet joints, and paravertebral structures in eight horses of Thirty-two biopsies were normal. Recently there have been signifi-
unspecified breed, between the ages of 2 and 14 years (Berg et al., cant advances in understanding the etiopathogenesis underlying
2003). A retrospective study of radiographic images of 122 horses these disorders including recognizing the similarities between
concluded that the size of the caudal cervical facet joint at the level human and equine muscle diseases (Piercy & Rivero, 2004). Myopa-
of C5–C6 increases with age but it is not known whether there is thies often present as gait abnormalities or overt RER, with slow
an association between such changes and clinical symptoms or improvements in clinical signs through dietary management and a
performance (Down & Henson, 2009). Claridge et al. (2010) regulated exercise routine as part of their rehabilitation (Hunt et al.,
described normal anatomical shape, spatial orientation and joint 2008).
volume for the cervical facet joints in six subjects.
Little is known about cervical intervertebral disc disease in horses.
The only study of the gross anatomy of the discs (Bullwein &
Hannichen, 1989) involved evaluations of midline sections the
Patho-anatomical diagnosis
cervical vertebrae without preparing histological sections. The
authors saw no evidence of an annulus fibrosus or a nucleus pulpo- Accurate patho-anatomical diagnosis of back pain is clinically chal-
sus and some discs showed disintegration of the tissue fibers. lenging. It requires a lengthy clinical examination and multiple
Cervical disc disease has been reported in horses, but is likely under- diagnostic procedures including radiography, ultrasonography,
reported since it poses a diagnostic challenge that may be solved by scintigraphy, local analgesia, kinematics and EMG imaging in
modern imaging equipment. Information on disc structure and the attempt to reach a diagnosis and increase the knowledge about
common pathological changes will facilitate the interpretation of the possible causes (Jeffcott, 1980; Steckel, 1992; Weaver et al.,
images of the discs. The joints in the mid neck (C3–C4, C4–C5 and 1999; Denoix & Dyson, 2003; Peham et al., 2001; Peham &
C5–C6) undergo considerably less motion in vivo even when the Schobesberger, 2006; Roethlisberger Holm et al., 2006; Gillen
neck is fully flexed or extended (Clayton et al., 2010a). It seems et al., 2009; Girodroux et al., 2009; Meehan et al., 2009; Fuglbjerg
likely, therefore, that the equine cervical intervertebral discs will et al., 2010). Further, as previously stated, many horses have multiple
vary in thickness. Future CT imaging studies may provide informa- osseous lesions that can be detected along the length of the verte-
tion about the size and structure of the discs that cannot be obtained bral column (Meehan et al., 2009; Girodroux et al., 2009; Gillen
from standard radiographic or ultrasonographic imaging. et al., 2009; Cousty et al., 2010; Stubbs et al., 2010). Post mortem
As discussed in Chapter 13, disorders and lesions affecting skel- studies have confirmed the presence of multiple types of osseous
etal muscle and peripheral nerves are common in horses and lesions at multiple sites; unfortunately these horses were unable to
adversely affect athletic ability (Cardinet & Holliday, 1979; Free- be examined ante-mortem to determine the relationship between
stone & Carlson, 1991; Martin et al., 2000; Gregory, 2004). Fre- the grades of lesions and pain and function (Townsend et al., 1986;
quently, muscular dysfunctions are secondary to underlying bone Haussler & Stover, 1998; Haussler et al., 1999b). Thus it is

386
 Patho-anatomical diagnosis

A B C

Fig 17.2  Three horses with asymmetrical hindquarter muscle development or relative atrophy on one side. (A) Right hemipelvis has markedly reduced
muscle mass compared with the left, particularly biceps femoris and gluteus medius, secondary to degenerative joint disease of the right tarsus. (B) Right
hemipelvis muscle mass moderately reduced compared to left in a successful international eventer with a history of limb lesions and degenerative
pathology. (C) Gross generalized disuse and secondary muscle atrophy affecting primarily the left hind limb following luxation of the superficial digital flexor
tendon off the calcaneus. The lines on the photographs facilitate a more objective assessment of asymmetry.

still difficult to determine the relationship between clinical signs, summary, the clinical signs often are unclear, so it is important to
diagnostic findings (the grade/severity of the lesions) and level of exclude other problems in order to arrive at a diagnosis of back
pain, as this is very variable between individual performances and pain. Microtrauma from chronic overuse because of poor saddle
horses (Haussler et al., 1999b; Meehan et al., 2009). The clinical and tack fit (Harman, 1999), sub-optimal riding technique or an
signs associated with a variety of osseous lesions of the vertebral inappropriate training schedule can all predispose horses to back
column and pelvis are poorly described, often non-specific and pain (Haussler, 2000).
difficult to validate objectively, with a decrease in performance Objective assessment of musculoskeletal pain in horses is chal-
being the main complaint of the rider/owner. Clinical diagnosis is lenging (Casey, 2002). Local tenderness is the major manifestation
further complicated by the fact that horses vary in their response to of most, if not all, musculoskeletal pain (Vanderweeën et al., 1996).
pain and it is believed that temperament also plays a role (Jeffcott, Tenderness in the axial skeleton has traditionally been assessed by
1999). manual palpation, although the interpretation of its outcome is
highly subjective (Haussler et al., 2007). Nonetheless, palpation
plays an important role in the clinical examination of cases with
suspected neck and back muscle sensitivity. Therefore, both equine
Functional assessment veterinarians and physiotherapists have included palpation in their
Clinical signs of back pain in horses include both behavioral and physical examination protocol (Ranner et al., 2002; Denoix & Pail-
physical signs including altered sensitivity to palpation, which may loux, 2005; Cauvin, 1997). This is obviously particularly important
be described as decreased mechanical nociceptive threshold (MNT); for clinical decision-making and evaluation of therapeutic interven-
sinking when placing the saddle on the horse, when securing the tion (Ashley et al., 2005). Pathological findings of the vertebral
girth (‘girthiness’) or when the rider mounts (‘cold back’); resistance column identified on radiographs often do not correlate with the
in work, for example, not wanting to trot, canter or rein back, refus- clinical findings (Gundel et al., 1997; Ranner et al., 2002). Thus,
ing to jump or tail swishing; and/or lameness of one or more limbs Ranner et al. (2002) concluded that palpation remains one of the
without a cause or possible diagnosis (Martin & Klide, 1997; most important clinical examination methods to determine whether
Haussler et al., 2006; Mills et al., 2007). Horses may express pain or not a horse is suffering from neck and back muscle sensitivity or
by fleeing or evasion; adopting an abnormal stance, gait or speed; pain. Pain in the TL region of a horse is very complex as it can
vocalizing or showing signs of aggression during movement or originate from various pain receptor structures. Afferent nerve
manipulation; restlessness; swishing the tail; and muscle tension endings are found in various types of connective tissue including
and tremors (Gregory, 2004; Mills et al., 2007). When back pain fascia around muscle bundles, muscle spindles, joint capsules,
becomes chronic, muscular atrophy of the back can become visible tendons and ligaments, blood vessel walls and bone. Some of these
(Fig. 17.2). Chronic pain, unlike acute pain, can be present without receptors are nociceptors (pain receptors) and respond to noxious
inflammation or noticeable tissue damage. Horses with chronic stimuli, including mechanical, thermal or chemical stimulation.
pain can be aggressive or evasive when they cannot react in their Muscle pain may range from very intense sharp localized pain in
natural way by fleeing. This is why owners and/or riders sometimes the acute setting (for example, a muscle tear), to chronic muscle
believe their horse has a behavioral problem, as they are not able pain, which may be dull and diffuse. Referred pain from abdominal
to identify the underlying painful process (Ridgway et al., 2005). In or thoracic organs can project onto different parts of the body

387
 17 Rehabilitation of the locomotor apparatus

including the back (Gregory, 2004). Although very difficult to quan- to assess tissue irritability, gross and intersegmental movement
tify in the horse neurogenic and or referred pain may also be present evaluation, and specific provocation tests of the vertebral column,
which is well documented in the human literature. Macgregor and osseous structures and pelvic soft tissue. Additional techniques
von Schweinitz (2006) identified equine myofascial trigger points include radiography, ultrasonography, CT, magnetic resonance
with similar objective signs and electrophysiological properties to imaging (MRI), EMG, scintigraphy and kinematic analysis.
those documented in human and rabbit skeletal muscle tissue.
Unfortunately, referred pain patterns and reproduction of the pain
profile cannot be determined in animals. Comparative pain charac- Kinematics
teristics between muscular and ligamentous saline injections have
Kinematic measurements describe stride and step timing, stride and
been investigated, as these methods are commonly used to induce
step lengths, and intervertebral joint motion and the symmetry in
human and animal back pain (Tsao et al., 2010). Interspinous liga-
different gaits. It is a useful tool in evaluating the effectiveness of
ment injections produced pain of greater intensity and duration
therapeutic interventions like chiropractic manipulation (Faber
compared to injecting the paraspinal muscles in the normal human
et al., 2003, Gomez Alvarez et al., 2008; Haussler et al., 2010).
lumbar spine. Interestingly muscle pain was reduced with contract-
Johnston et al. (2004) performed a kinematic evaluation of the
ing and stretching the injected muscle, but this did not affect the
back in horses without a history of back problems in order to
pain produced from the ligament injection. The authors were also
develop a database on kinematics. Older horses were shown to have
surprised that many of the subjects (n = 10) pointed to a region of
decreased flexion and extension of the TL junction in trot. Kine-
pain 1–2 segments away from the injection site (Tsao et al., 2010).
matic analysis can be a useful tool to identify horses with back pain
Pain works as a protective mechanism to prevent further tissue
because the movement pattern of the back changes so that dorso-
damage and to allow healing of wounds and damaged tissue. Pain
ventral flexion and extension at T13 and T17 at walk and at T17
cannot be objectively measured, because it has no units. Changes
and L1 at trot are significantly reduced compared to horses without
in heart rate, blood pressure, plasma cortisol and behavior can be
back pain but there is no difference in lateral bending between
helpful to identify and study pain (Ashley et al., 2005; Robertson,
horses with and without back pain. During walk, horses with back
2006). Examples of ways to subdivide pain are acute versus chronic,
pain have a decreased axial rotation of the pelvis. Decreased
somatic versus visceral and physiological versus pathological, which
back motion can also be present due to pathological conditions
occurs when the tissue is damaged and responds with inflammation
(Wennerstrand et al., 2004; Gomez Alvarez et al., 2007a, b), with
(Robertson, 2002). In mammals, inflammation is the greatest
the most significant differences being visible in walk rather than
source of pain (Gregory, 2004). During inflammation, the nerve
trot. This is likely due to the fact that the trot inherently has less
endings that respond to noxious stimuli such as heat, pressure and
intervertebral motion than the walk. However, these studies were
chemical stimuli become sensitized. When normal, non-painful
performed only on a treadmill and further investigation is war-
stimuli are applied to sensitized tissue, the patient experiences a
ranted on a circle, on varying gradients and under saddle in a larger
painful sensation. In the human this is commonly referred to as
number of subjects with different types of back pain/pathology.
‘allodynia’ (Woolf & Mannion, 1999). Bussières et al. (2008) estab-
lished the value of behavioral and physiological criteria by develop-
ing and validating a composite multifactorial pain scale (CPS) in EMG
an experimental equine model of acute orthopedic pain. Eighteen
horses were allocated to control and experimental groups that The ‘normal’ movement of the horse’s back has been investigated
received Amphotericin-B injection to induce pain in the tarsocrural via EMG in conjunction with comparisons of locomotor kinematics
joint and various forms of analgesia. Inter- and intra-observer repro- and kinetics. Many factors contribute to locomotion including the
ducibility was good (0.8 < K < 1), with the key specific and sensitive muscular system (longissimus dorsi, iliocostalis, semitendinosus, biceps
behavioral indices being palpation of the painful area and the femoris, and gluteus medius), acceleration from the hind legs, gravity
horse’s posture. Other less valuable signs were pawing the floor, and the influence of the rider (Licka et al., 2004, 2009; Peham &
kicking the abdomen and head movement. There was a statistical Schobesberger 2006; Wakeling et al., 2007; Zaneb et al., 2009).
correlation between the CPS and both non-invasive blood pressure Gluteus medius is considered an epaxial muscle as it covers the
(p < 0.0001) and blood cortisol (p < 0.002). Van Loon et al. (2010) majority of the dorsal lumbar spine, with attachments onto the
recently investigated the reliability and clinical applicability of thoracolumbar fascia and iliocostalis ventrally. The anatomical
applying a CPS to objectively monitor somatic and visceral pain epaxial musculature of the lumbar spine is displayed in Figure 17.3.
under hospital conditions in 94 horses (control, acute and chronic EMG is becoming a widely accepted tool for the examination of
surgical and non-surgical groups). CPS showed low baseline values muscle function in horses (Robert et al., 2000; Wakeling et al.,
in healthy animals with non-painful conditions and these were not 2007; Zsoldos et al., 2010; Groesel et al., 2010). The longissimus dorsi
affected when general anesthesia was the only intervention. Inter- is medially located directly overlying multifidus (Stubbs et al., 2006).
observer reliability was very high (n = 23 horses; weighted kappa It runs from the transverse processes of the thoracic and lumbar
correlation coefficient, κ = 0.81). Horses with painful conditions vertebrae and the dorsal extremity of the ribs to the tuber sacralae
responding well to analgesic treatment could be discriminated from and the ventro-medial aspect of the iliac crest (Haussler, 1999a;
horses that had to be euthanized on humane grounds because of Stubbs et al., 2006). The longissimus dorsi muscle contributes to sup-
painful non-responsive conditions. It was concluded that the CPS porting the weight of the rider and saddle against dynamic forces
is a promising tool to potentially provide a basis for direct day-to- (Licka et al., 2004). Bilateral contraction of the epaxial muscles
day assessment of pain status in equine patients with various produces extension of the spine and unilateral contraction produces
painful conditions in the future. The CPS may be a useful subjective lateral flexion and rotation (Haussler, 1999a; Denoix & Pailloux,
tool as an adjunct to other subjective/objective measurements 2005; Peham et al., 2001). At the walk, the back moves passively
during the rehabilitation process. but at the trot, the back muscles are active to control back move-
ments (Robert et al., 2001). Walk and trot are symmetrical gaits, and
the movement of the back in these gaits is highly symmetrical (Jef-
fcott, 1979). Wakeling et al. (2007) found segmental variation in
Objective measurements the activity and function of longissimus during walk and trot. Zaneb
For evaluation of back function there are various diagnostic possi- et al. (2009) investigated the quantitative differences of back
bilities. Firstly, a thorough clinical evaluation is important, includ- and pelvic muscles during walk and trot between chronically lame
ing a complete lameness examination, general and specific palpation and non-lame horses using surface EMG (SEMG). For each muscle

388
 Patho-anatomical diagnosis

B C

Fig 17.3  Lateral view of the dorsal thoracolumbar spine and pelvis. (A) Dissection of the caudal thoracic, lumbar and sacral outline of the middle gluteal
muscle, which has been removed. Note that this muscle covers a large area of the lumbar region and attaches cranially at the level of the last rib which can
be seen ventrally. Underneath the middle gluteal lies the thoracolumbar fascia (TLF), longissimus dorsi, and the more lateral iliocostalis. The dorsal and lateral
parts of the dorsal sacroiliac ligament (DSIL) are labeled. Underneath the lateral DSIL are sacrocaudalis dorsalis medialis and lateralis, the latter having
attachments to the caudal lumbar dorsal spinous processes. (B) Part of longissimus dorsi and iliocostalis have been removed to reveal the underlying
multifidus, which has multiple fascicles originating from each spinous process that run caudally spanning 1–5 more caudal vertebrae. (C) Cross-sectional
image at the level of the 13th thoracic vertebra showing the structure and location of multifidus (M) and longissimus dorsi (L) muscles. Note the fibrous
content of multifidus.

(longissimus dorsi, semitendinosus, biceps femoris, gluteus medius, exten- trotting on a treadmill is at T12 (Peham et al., 2001; Licka et al.,
sor digitorum longus) kinematics and SEMG data were recorded syn- 2004).
chronously at walk and trot on a treadmill; mean, maximum, and It is only through needle/fine wire EMG that the electrical activity
minimum muscle activities and maximum-to-mean and minimum- of the motor unit can be evaluated. In this way a possible pathologi-
to-mean activity ratios were calculated. For each horse (lame or cal process can be detected and the difference between neurogenic
non-lame) pelvic limb data were averaged; in lame horses, data were and myogenic disorders can be made (Wijnberg et al., 2002). The
also examined separately (NL–L and L–L values, respectively). Com- advantages of SEMG compared to needle EMG are that it is less
parisons were made between gaits. At walk the NL–L maximum-to- invasive and that measurements during movement are possible
mean ratios for biceps femoris and gluteus medius muscles were (Franssen, 1995). However, the clinical relevance of SEMG is still
significantly greater, and in lame horses, L–L and NL–L minimum- limited because of the great variation in measurement outcomes
to-mean ratios for semitendinosus, biceps femoris, gluteus medius, and (Licka et al., 2004). In pathological conditions, muscle tension
longissimus dorsi muscles were significantly less in NL horses. At trot changes the relation of the minimal and maximal EMG activity
minimum-to-mean ratios for semitendinosus, gluteus medius and lon- independent of muscle mass, the conduction of the skin, the exact
gissimus dorsi muscles in lame horses were significantly lower than location of the electrodes, the EMG measuring device or its configu-
those for NL horses. Activity of extensor digitorum longus muscle was ration (Licka et al., 2004), as well as the presence of the cutaneous
not affected by lameness. In chronically lame horses, back and trunci muscle directly attaching to the skin in the shoulder and tho-
pelvic limb muscle activities were affected differently during walking racolumbar and abdominal regions. In people, back pain can be
and trotting. SEMG has confirmed that motor control is altered in induced, by injecting hypertonic saline into the back muscles with
the lame horse, whereby the compensation strategy is an increase recordings showing an increase in mean EMG signal compared to a
in muscle activity in the NL limb which may have major implica- control group. However there was no correlation between degree of
tions in the assessment and treatment of horses, especially those pain and changes in the EMG signal (Arendt-Nielsen et al., 1995).
with chronic lameness and adaptive motor control patterns and/or In the horse manual therapy has been proven to significantly
secondary muscle atrophy. alter/decrease muscle tone and total muscle activity measured by
By combining kinematics and EMG, the relationship between the EMG (Wakeling et al., 2006), though the direct mechanism has yet
muscle activity and back movement can be determined (Peham & to be established. In contrast, EMG measurements following acu-
Schobesberger, 2006). The maximal EMG activity of the back pre- puncture in healthy humans, showed no difference in muscle activ-
cedes the maximal vertical movement. It is theorized that at the end ity (Tough et al., 2006), although whether motor control may be
of stance and during the suspension phases of every stride cycle, the altered by therapeutic intervention is debatable. A decrease in left-
longissimus dorsi muscle provides lumbosacral extension and pro- right asymmetry of the EMG signal was found after applying acu-
vides the propulsion force for the hind leg (Denoix & Pailloux, puncture to the paraspinal muscles in the human back (Tanaka
2005). The highest muscle activity of longissimus dorsi both during et al., 1998). Further research, particularly using fine wire EMG with
passive flexion and extension of the back in stance and during respect to rehabilitation techniques is definitely warranted.

389
 17 Rehabilitation of the locomotor apparatus

Pressure algometry There is still much debate amongst human and equine researchers
with respect to whether this technique requires further validation
Pressure algometry (PA) is a subjective and potentially objective utilizing other objective measures such as EMG. The human litera-
diagnostic tool used in both human and equine research to assist ture still classes PA as a subjective tool that should be combined
in quantifying musculoskeletal pain relative to the tissues response with other pain scale measurements, such as the visual analogue
(irritability) to mechanical pressure, whereby the mechanical noci- scale, for clinical and experimental purposes (Ylinen, 2007; Ylinen
ceptive threshold is determined (MNT). It is used in research and et al., 2007).
clinical practice to evaluate treatment results in humans and horses Pressure algometry is a relatively new tool in equine medicine.
(Fischer, 1998; Pöntinen, 1998; Haussler & Erb, 2003). Pressure is Benefits include the non-invasive character of the method and the
applied via the algometer probe with the investigator subjectively ease of use in horses (Haussler & Erb, 2003). Regional median MNT
noting the tissue response (e.g. muscle fasciculation or spasms and/ values were 9 kg/cm2 in the cervical area, 12 kg/cm2 in the thoracic
or behavioral avoidance reactions). The MNT is expressed in kg/cm2 area, 13 kg/cm2 in the lumbar area and 16 kg/cm2 in the pelvic area
(Fig. 17.4). Pressure algometry has proved to be repeatable in (Haussler & Erb, 2006a). MNT values were higher in non-TB horses,
humans and horses (Pöntinen, 1998; Ylinen 2007; Ohrbach & Gale, thus emphasizing the need to collect MNT values for horses of dif-
1989a; Fischer, 1987; Varcoe-Cocks et al., 2006; Love et al., 2011), ferent breeds and performance levels.
but its reliability depends on the device itself as well as the experi- When using PA, the pressure should be increased gradually at a
ence and technique of the examiner (Pöntinen, 1998; Nussbaum & constant rate and should stop when the horse shows a behavioral
Downes, 1998; Ohrbach & Gale, 1989a; Antonaci et al., 1998). and/or tissue reaction that may reflect pain (Varcoe-Cocks et al.,

Fig 17.4  Two types of pressure algometers used in research. (A) FPK 60
pressure algometer. Note the position of the device in the hand of the
operator and the rubber tip. The dial indicates pressure applied to the skin
in kg/cm2. (B) COMMANDER™ algometer. Pressure applied to the tip of the
algometer is registered digitally on the unit worn on the operator’s arm.
With this unit, the digital readings can be stored and downloaded without
revealing the values to the examiner.
(A) Reprinted from De Heus, P., Van Oossanen, G., Machteld, C. et al., 2010, A pressure
algometer is a useful tool to objectively monitor the effect of diagnostic palpation by a
physiotherapist in warmblood horses, Journal of Equine Veterinary Science 30 (6), 310–321,
with permission from Elsevier (B) Courtesy of JTECH Medical, Salt Lake City, Utah, USA.

390
 Patho-anatomical diagnosis

2006; De Heus et al., 2010; Love et al., 2011). This can be a twitching physiotherapists. This study tested inter-examiner repeatability of
of the skin, ear movement, sinking, stepping aside, kicking or biting. animal physiotherapists and tried to correlate subjective clinical
The maximal applied pressure until a pain reaction occurs is scores of neck and back function in horses. Owners frequently
recorded as the MNT for that site. When performed by an experi- believe that their horse has improved after the physiotherapist made
enced investigator, PA may be a reliable and repeatable tool that a diagnosis, even though the actual treatment had not yet started.
correlates significantly with palpation score in horses. It may con- Therefore, it was hypothesized that PT diagnostic intervention in
tribute to a clinical examination of the back and may be a useful itself could have a therapeutic, objectively measurable effect that
adjunct subjective tool for evaluation of the effectiveness of thera- would increase the MNT of horses evaluated physiotherapeutically
peutic or analgesic interventions (Varcoe-Cocks et al., 2006; compared to the control group. Therapeutic effects were especially
Haussler, 2006; Sullivan et al., 2008; De Heus et al., 2010). However, expected from stiffness testing, since this is also a spinal manipula-
horses have been shown to develop a learned response in which tive technique used as therapy by physiotherapists and there is some
they respond to a stimulus in anticipation of it becoming aversive. evidence that manipulative therapy is effective in horses (Haussler
Regular sham testing may avoid this problem (Kamerling et al., et al., 2007; Haussler et al., 2010). The effects of three consecutive
1985). diagnostic palpation examinations within 1 day on MNTs were
In humans, bruising was observed in some subjects after a 3-day measured, with three physiotherapists examining and grading the
PA study with a total of 18 measurements on one location (3 days, same horses, which made it possible to evaluate agreement between
three trials/day by two examiners) (Nussbaum & Downes, 1998). physiotherapists.
In horses, bruising after repeated measurements has not yet been De Heus et al. (2010) studied six clinically sound Dutch Warm-
reported (Haussler & Erb, 2006a), but local bruising can be difficult bloods. Since they were riding school horses, they were expected to
to differentiate since dark skin coloration masks the presence of a have some neck or back pain that could lower the MNT and alter
hematoma. Different pain scales are commonly used in humans other palpation criteria, even though they did not have a current
including visual analog scales to assess pain intensity, which can history of lameness or therapy. Horses were randomly assigned to
also be transferred to horses: the numeric rating scale (NRS) and an experimental (n = 3) or control (n = 3) group. The experimental
the simple descriptive scale (SDS) (Holton et al., 1998; Williamson group received diagnostic PT intervention between morning and
& Hoggart, 2005). For the NRS, the patient picks the number that evening PA measurements, the control group received only the two
best represents his or her pain. The SDS is usually a four-point scale PA measurement sessions.
with numbers attached to descriptive terms, such as 0 (no pain) to A non-electrical PA (FPK 60 Wagner Instruments Inc., Post Office
3 (severe pain) (Holton et al., 1998; Williamson & Hoggart, 2005). Box 12.17, Greenwich, Connecticut, USA) was used (Fig. 17.4A),
The pain intensity scales were developed for self-reporting of pain which has a force gauge with a 1 cm2 rubber tip and a maximal
in humans, something that is obviously impossible in veterinary reading of 30 kg. According to the certificate of calibration it has an
medicine, thus an observer has to score the pain on behalf of the accuracy of ± 2 graduations below 2.5 kg and an accuracy of ± 1
animal, relying on behavioral observations and sometimes physi- graduation above 2.5 kg. Since there is a learning curve associated
ological variables (i.e. heart rate, respiratory rate). The physiological with the use of a PA (Nussbaum & Downes, 1998; Ohrbach & Gale,
variables are, however, not exclusively influenced by pain and are 1989a,b; Antonaci et al., 1998), operators learned to increase pres-
more useful in combination with behavioral responses (Pritchett sure at a constant rate (force/time) and to recognize pain behavioral
et al., 2003). The pain behavior construct proposes that pain affects reactions of horses in response to the pressure algometer. The
behavior in ways that are accessible to observers and that the mag- authors suggested practicing exerting a constant force over time
nitude of change correlates with the severity of the pain (Hansen using a pressure mat (RsScan Flexible Solutions System, RSscan
2003). As previously stated, scoring systems used in equine research INTERNATIONAL, Lammerdries 27 B-2250 Olen, Belgium), which
to assess pain severity were designed to evaluate the effectiveness of is designed for measurement of intra-articular pressure in the
an epidural block or to quantify postoperative pain in orthopedic equine fetlock joint. Secondly, the operators practiced PA on horses
surgery or colic (Pritchett et al., 2003). Ashley et al. (2005) con- under the supervision of an equine behaviorist. MNT measurements
cluded in a review that there is strong evidence that aggressive should be made in a quiet room, with both the room and the
behavior emerges from pain. technique being familiar to the horses to avoid a conditioned fear
Dyson and Murray (2003) reported clear behavioral responses to response that could interfere with the reaction to a painful stimulus
palpatory actions in horses with pain in the sacroiliac joint. Vertical through opiate release (Fanselow et al., 1989).
forces or painful stimuli applied near the lumbosacral junction Pressure algometer measurements were performed with the neck
often induced maximal extension at the lumbosacral joint (Haussler straight holding the PA in one hand and stabilizing the neck on the
et al., 2001). Wennerstrand et al. (2004) used palpation as an inclu- contralateral side with the other hand. The plunger tip of the PA
sion criterion for back pain in her research on back kinematics in was held between thumb and forefinger (Fig. 17.4A), to avoid acci-
horses with back pain. She reported the following adverse reactions dental shifting of the tip (Haussler & Erb, 2006a). Pressure was
to palpation commonly seen in horses with back pain: bolting or applied perpendicular to the surface and increased at a constant
rearing, tail swishing, unruliness, rapid caudal movement of the value of 10 kg/cm2, which was similar to that used by Haussler &
ears, or stiff, jerky movements. The palpation scoring system of Erb, (2006a). The pressure stimulus was applied until a behavioral
Varcoe-Cocks’s et al. (2006) included many aspects of palpation avoidance reaction was evoked (muscle fasciculations, cutaneous
(muscle tone, pain and stiffness) in a single scale. Wolf (2002) trunci reflex, active vertebral movement or stepping away (Haussler
related muscle tone and pain through chronic irritation of muscles & Erb, 2003)). The presence of the ‘pain face’, which is recognized
leading to permanent elevation in muscle tone, which continuously as a fixed stare, with the eyes puckered slightly, the ears held back
feeds back to reinforce the condition. This is the mechanism by slightly, and the nostrils dilated (Fraser, 1969), was observed and
which segmental dysfunction leads to muscle spasm. Wennerstrand reported. When an avoidance reaction was observed, the pressure
et al. (2004) found that the clinical manifestation of back pain was immediately stopped and the value recorded was considered to
resulted in diminished flexion–extension movement at or near the be the MNT value. To increase reliability, measurements were
TL junction. These examples again indicate that musculoskeletal repeated in three successive trials at one site (Nussbaum & Downes,
pain, increased muscle tone and stiffness often concur, and seem to 1998). Measurements were performed blinded: the PA was passed
be related. to a third party to record the result.
Research conducted by De Heus et al. (2010) is an example The 35 anatomical locations were adopted from Haussler & Erb
of subjective/objective evidence of the use of PA to evaluate the (2006a) and Varcoe-Cocks et al. (2006) (Table 17.1). The anatomi-
use and possible effect of (diagnostic) palpation applied by cal measurement sites were marked because identification of

391
 17 Rehabilitation of the locomotor apparatus

Table 17.1  Anatomical locations for pressure algometry measurements

Abbreviation Anatomical location

Bony landmark (four bilateral sites)
1, 19 Caudoventral aspect of the wing of the atlas
3, 21 Dorsolateral aspect of the caudal portion of the 3rd cervical transverse
process
4, 22 Dorsolateral aspect of the caudal portion of the 5th cervical transverse
process
1 19
16, 33 Dorsal aspect of the sacral tuber

Supraspinal ligament landmark (five sites) 2 20

3 21
10 Between 15th and 16th thoracic spinous process
28 Between 17th and 18th thoracic spinous process 4 22
5 23
11 Between 18th thoracic and 1st lumbar spinous process 6 24
31 Between 1st and 2nd lumbar spinous process
14 Between 2nd and 3rd lumbar spinous process

Muscle landmark (11 bilateral sites)

2, 20 Middle of the splenius muscle at the 3rd cervical vertebral level 7 25
5, 23 Dorsal portion of the cervical serratus ventralis muscle at the 6th cervical
vertebral level 8 26
9 27
6, 24 Brachiocephalicus muscle at the base of the neck, at the level of C7
7, 25 Dorsolateral aspect of the thoracic spinalis muscle at the 9th thoracic spinous 10
28
process 11 29
12 30
9, 27 Thoracic longissimus muscle at the 13th thoracic vertebral level, 2 cm lateral 13 31
to the dorsal midline 14 32
15
8, 26 Mid-portion of the thoracic longissimus muscle at the 13th thoracic vertebral 16 33
level, 10 cm lateral to the dorsal midline 17 34
12, 30 Thoracic longissimus muscle at the 18th thoracic vertebral level, 2 cm lateral 18 35
to the dorsal midline
13, 29 Mid-portion of the thoracic longissimus muscle at the 18th thoracic vertebral
level, 10 cm lateral to the dorsal midline
15, 32 Mid portion of the middle gluteal muscle at the 3rd lumber vertebral level,
10 cm lateral to the dorsal midline
17, 34 Mid point between the cranial aspect of the tuber sacrale and tuber coxae
(middle gluteal muscle)
18, 35 20 cm dorsal to the greater trochanter, over the vertebral head of the biceps
femoris muscle

anatomical landmarks is subject to wide variation (Weller et al., 4, hypertonicity. Mobility was scored as: 0, hypermobility; 1, normal
2006). Measurements started on the left of the withers moving mobility; 2, slight decrease in expected range of motion; 3, moder-
caudally, then continued at the poll and neck. They were repeated ate decrease in expected range of motion; 4, stiff and rigid, no range
in the same order on the right side. All three physiotherapists/ of motion. For further description of the scoring system see De Heus
examiners graded four elements of neck and back dysfunction et al. (2010). A mean physioscore was calculated from the scores of
(‘temperature’, ‘pain’, ‘muscle tone’, and ‘mobility’) in the experi- the three individual physiotherapists.
mental group (n = 3) between the MNT measurements using a Local assessment of back or neck dysfunction at and adjacent to
scoring system specifically developed for this study. The elements the MNT measurement was recorded, since stiffness and increased
were scored according to a grading system. Temperature was scored: muscle tone can be manifestations of back pain and they are
0, cold; 1, normal temperature; 2, mildly increased temperature; 3, included in the scales modeled as an SDS (Holton et al., 1998;
moderately increased temperature; 4, severely increased tempera- Williamson & Hoggart, 2005).
ture. Pain was scored as: 0, pain free; 1, mildly painful; 2, moder- Both the MNT measurements and the physiotherapeutic diag­
ately painful; 3, moderately increased pain; 4, severely painful. nostic intervention were well accepted by the horses. The three
Muscle tone was scored as: 0, hypotonicity; 1, normal tone; 2, successive pressure algometer measurements did not differ signifi-
mildly increased muscle tone; 3, moderately increased muscle tone; cantly from each other, and there were no significant differences

392
 Patho-anatomical diagnosis

between left and right sides, which were pooled for analysis, or at algometry measurements. Sullivan et al. (2008) reported 20%
the same site over time. There were significant differences in MNT sequential increase, 13% sequential decrease and 67% no change
values at different sites and between morning and evening values or consistent pattern, similar to De Heus et al. (2010). Pain thresh-
at the same site (Tables 17.2 and 17.3). Median MNT values old shows diurnal variation in humans (Göbel & Cordes, 1990;
increased from cranial to caudal in both the morning and the Procacci, 1993) but different pathological conditions vary in the
evening (Tables 17.2 and 17.3). MNT values of muscle were signifi- time of day when pain is more severe. Circadian rhythms of pain
cantly higher than those for the supraspinous ligament and bone. could be related to the time-dependent variation in endogenous
There were significant differences (p < 0.05), in scores between the opiod peptides, which are higher in the morning and lower in the
three physiotherapists for ‘temperature’, ‘muscle tone’ and ‘mobil- evening in man (Labracque & Vanier, 2003). Hamra et al. (1993)
ity’ but not for ‘pain’. found a strong positive correlation between the equine plasma
This study showed that the palpation findings of individual phys- immunoreactive β-endorphin and nocioceptive threshold, both
iotherapists were correlated with the objective MNT measurements peaking in the morning at 09.00 hours and again less high at 15.00
Horses that received scores indicative of being more sensitive hours, which corresponds with the significantly higher MNTs in the
(higher scores for pain, temperature, and muscle tone) had lower morning measurements.
MNT values. On the contrary, there was a positive correlation In humans as well as in horses the MNT between individuals
between the subjective ‘mobility’ score and the objective PA mea- varied more than that within one individual (Haussler & Erb,
surements, illustrating that the more rigid vertebral columns will 2006a; Ylinen et al., 2007; Pöntinen, 1998; Vanderweeën et al.
show higher MNT values. Inter-horse differences in MNT reflected (1996), which was confirmed in this study. Haussler and Erb
differences in individual sensitivity. The lower MNT values of horses (2003), noticed that the individual MNT measurements varied
in the evening may have been due to sensitization of the measure- greatly between horses, but were very consistent within horses.
ment location, a learning effect of the horses or a diurnal fluctua-
tion. Haussler & Erb, (2003) found that the majority of horses did
not demonstrate accommodation or sensitization to serial pressure Table 17.3  MNT recorded in the evening (mean ± SD) from a
group of riding school horses (Heus et al., 2010)

Table 17.2  MNT recorded in the morning (mean ± SD) from a Site Left Middle Right Site
group of riding school horses (Heus et al., 2010)
Cervical
Site Left Middle Right Site 1 4.5 ± 0.9 3.8 ± 1.2 19
Cervical 2 5.1 ± 1.3 4.6 ± 0.5 20
1 4.9 ± 1.6 3.7 ± 1.7 19 3 4.5 ± 2.5 4.7 ± 1.7 21
2 6.0 ± 2.6 4.8 ± 1.9 20 4 4.7 ± 1.7 5.7 ± 1.4 22
3 5.2 ± 1.9 4.8 ± 1.9 21 5 7.7 ± 3.3 6.4 ± 0.8 23
4 5.5 ± 1.6 5.5 ± 0.9 22 6 4.5 ± 1.4 5.7 ± 1.3 24
5 7.2 ± 1.2 6.8 ± 1.2 23 Thoracic
6 5.8 ± 2.6 5.8 ± 1.8 24 7 6.3 ± 1.4 6.6 ± 0.9 25
Thoracic 8 6.1 ± 1.6 6.7 ± 1.4 26
7 7.3 ± 1.8 6.6 ± 1.3 25 9 6.4 ± 0.8 6.5 ± 1.3 27
8 6.7 ± 1.7 7.5 ± 0.9 26 10 5.8 ± 1.7 10
9 6.8 ± 1.2 6.6 ± 1.1 27 28 5.9 ± 0.9 28
10 8.1 ± 2.4 10 11 6.6 ± 1.6 11
28 7.3 ± 2.2 28 12 7.7 ± 1.8 7.9 ± 1.5 30
11 6.7 ± 1.0 11 13 7.4 ± 1.7 7.5 ± 0.9 29
12 6.9 ± 1.7 8.8 ± 1.5 30 Lumbar
13 6.6 ± 2.1 8.0 ± 1.7 29 31 6.7 ± 2.4 31
Lumbar 14 7.4 ± 1.9 14
31 7.3 ± 1.4 31 15 8.5 ± 2.5 8.8 ± 1.4 32
14 7.6 ± 1.6 14 Pelvis
15 8.2 ± 1.6 9.3 ± 1.8 32 16 9.9 ± 2.5 9.7 ± 3.3 33
Pelvis 17 8.9 ± 2.9 9.3 ± 1.5 34
16 9.7 ± 2.5 11.0 ± 2.4 33 18 11.7 ± 2.7 11.4 ± 2.9 35
17 9.7 ± 2.7 10.2 ± 2.5 34 Reprinted from Clayton, H.M., Lavagnino, M., Kaiser, L.J. and Stubbs, N.C. (2011b) Hind
18 13.3 ± 2.8 12.7 ± 1.4 35 limb flexion response to different types of tactile devices, with permission from the
American Journal of Veterinary Research, http://avmajournals.avma.org/loi/ajvr

393
 17 Rehabilitation of the locomotor apparatus

Pöntinen, (1998) argued therefore that it would be impossible to Ultrasonography provides an immediate image of structures
determine absolute reference values, since the variation between beneath the probe enabling the practitioner to determine soft tissue
individuals is too large. Therefore, relative values are considered to sizes, particularly ligamentous and muscle CSA and linear measure-
be more specific and more sensitive. Interpretation of the measure- ments (Stokes & Young, 1986; Pickersgill et al., 2001). It also allows
ments at the cervical points were disputable in the study of De Heus real-time acquisition of muscle activity, whereby the practitioner
et al. (2010) because some horses started head shaking before pres- can visually observe and measure changes in muscle shape during
sure was applied, possibly in anticipation of experiencing pressure both concentric and eccentric volitional activation in comparison
that they found aversive. Since right-left MNTs are not significantly to the relaxed state (Kidd et al., 2002; Hodges et al., 2003b;
different (Haussler & Erb, 2006a; De Heus et al., 2010), the contra- McMeeken et al., 2004). In people, changes in muscle thickness as
lateral side can be used to compare MNT values in case of unilateral measured by ultrasound have been shown to be highly correlated
pathology. to EMG activity of the lumbar multifidus in asymptomatic subjects
In conclusion, a PA can be a useful tool to objectively monitor (Kiesel et al., 2006). McMeeken et al. (2004) also found that
the palpation of individual physiotherapists. It has proven difficult changes in thickness of transversus abdominus are indicative of elec-
to maintain a constant rate of pressure leading to repeatable results. trical activity of the muscle. The use of US in this manner has been
Introduction and implementation of a more detailed protocol extensively described in the human literature, including reliability
describing the procedure for diagnostic examination by palpation and normative data, particularly with respect to back pain and
of the neck and back area, describing for example hand movement rehabilitation (Hides et al., 1998; 2001: 2008; Ferreira et al., 2004;
and aimed-for hand pressure, could aid in achieving a more uniform Hodges, 2005; Koppenhaver et al., 2009).
scoring outcome. This is not standardized yet in physiotherapeutic Despite US being an essential imaging modality in the assessment
and veterinary diagnostic palpation protocols. The effects of phys- and rehabilitation of the equine musculoskeletal system, data on
iotherapeutic diagnostic palpation in relation to MNT must be its accuracy and reproducibility in veterinary medicine are extremely
taken into account when applying the PA to assess the effects of limited. Research in equine biomechanics and PT interventions is
physiotherapeutic intervention in rehabilitation programs. Pressure using both functional objective measures including US and the
algometry may be useful for quantifying clinical neck and back human motor control model to investigate various equine prob-
muscle sensitivity in horses possibly leading to dysfunction, as well lems including TL-pelvic pain including sacroiliac joint biomechan-
as for evaluating treatment outcomes. Repeated measurements on ics (Degueurce et al., 2004, Goff et al. 2006) and equine back pain
the same day and at the same location along the vertebral column (Denoix, 1999, Stubbs et al., 2006; Stubbs et al., 2010a; 2011;
may influence absolute MNT values. It is important that the opera- Clayton et al., 2010). Data describing repeatability and validity of
tor is trained in use of the PA. US in the horse are limited to the tendon and epaxial musculature
(Pickersgill et al., 2001; van Schie et al., 1999, 2000; McGowan
Ultrasonography et al., 2007c; Stubbs et al., 2010). Pickersgill et al. (2001) performed
a quantitative investigation of the reliability of US analysis of the
In the field of human physiotherapeutic rehabilitation, ultrasonog- morphometric properties of the equine SDF tendon. The effects of
raphy (US) has emerged as an invaluable tool for objective func- three variables on tendon CSA measurements were determined by
tional assessment and management of neck, low back and pelvic comparing two skilled operators acquiring US images and measur-
girdle pain and dysfunction in relation to muscle function and ing CSA values in 16 National Hunt TB racehorses. Variables ana-
motor control (see section XVIII.3.2: Motor control) (Jull & Rich- lyzed included inter-operator reliability during image acquisition
ardson, 2000; Pietrek et al., 2000; Richardson et al., 2002; Whit- and CSA measurement, and intra-operator reliability when using
taker et al., 2007; Hodges & Cholewicki, 2007). Ultrasonographic different analytical equipment from previously stored images. There
measurement of multifidus CSA and function has proven to be a was no statistically significant difference in the inter-operator image
reliable, objective guide in the assessment, management and pre- acquisition (p > 0.05) but inter-operator image analysis showed a
vention of recurrence of back pain in man (Hides et al., 2008). significant difference (p < 0.01) with one operator consistently
Comparison of measurements of muscle size (cross sectional area, returning larger measurements. The use of different equipment
CSA) from the left and right sides at a given vertebral level provide within one operator was not reliable in the distal metacarpal region.
a method of direct assessment of muscle function by indicating It was concluded that consistency of image acquisition across skilled
whether the muscle is of normal size, atrophied, or hypertrophied. ultrasonographers was good but the images should be evaluated by
This is a functional assessment that can be used to monitor the a single-skilled operator.
effectiveness of treatment over time in rehabilitation (Rantanen Van Schie et al. (1999) used first-order gray-level statistics of US
et al., 1993; Stokes et al., 2005; Whittaker et al., 2007). Ultrasonog- images in an in vitro study to quantify the effects of some instrumen-
raphy is used extensively in equine veterinary practice and many tal variables on the evaluation of equine SDF tendons to compare
practitioners already have an ultrasound machine, thus the use of normal tendon tissue, an acute lesion, or a chronic scar. Results
ultrasound measurements has practical value in equine rehabilita- showed considerable effects on the gray levels of the US image with
tion (Stubbs et al., 2010). slight variations in scanner settings and transducer handling. Tilting
Historically in veterinary medicine, US has been widely used to the transducer had substantial effects on mean gray levels, especially
diagnose and monitor tendon and ligament lesions in the distal in acute lesions (40%) and, to a lesser extent, in normal tendon
limbs, with more recent clinical applications being recognized in tissue (18%) and chronic scar tissue (12%). Displacement of the
the vertebral column and pelvis (Tomlinson et al., 2001; 2003; transducer causes relatively small changes in normal tendon tissue,
Kersten & Edinger, 2004; Mattoon et al., 2004; Denoix & Dyson, but changes the mean gray-level by 7% in chronic scar tissue and
2003; McGowan et al., 2007c; Stubbs et al., 2010; 2011). In com- 20% in an acute lesion. The US image is also substantially influenced
parison to other modern diagnostic and biomechanical technology, by the total amplifier gain output: a low gain setting in an acute
US is relatively inexpensive, non-invasive, easy to use, and allows lesion results in an almost completely black image and, conversely,
rapid analysis by an experienced practitioner. For clinical and a marked ‘filling in’ effect on the lesion occurs with higher gain set-
research purposes both objective and subjective scales of measure- tings. Therefore, the value of exclusively using a quantitative evalu-
ment are used (Reef et al., 1998; Genovese et al., 1997; Pickersgill ation method on the first-order gray-level statistics may not
et al., 2001). These involve comparison of contralateral tissues and discriminate accurately enough to assess the integrity of tendon
determination of echogenicity using gray-scale analysis and prede- during rehabilitation. A further study that quantified the transverse
termined ordinal scales (Martinoli et al., 1993; Gillis et al., 1993; US image by use of first-order gray-level statistics concluded that
Tsukiyama et al., 2005). the method was not sufficiently sensitive to accurately and

394
 Patho-anatomical diagnosis

unequivocally determine the type of tendon tissue (van Schie et al., 1998a,1998b; Pressler et al., 2006; Wallwork et al., 2007; Lee et al.,
2000). Therefore, quantitative analysis should incorporate both 2007; Delaney et al., 2010).
transverse and longitudinal information. These studies highlight the In the horse, McGowan et al. (2007c) determined the reliability
need for standardized techniques and instrumentation for image of US measurements of CSA of multifidus mm, longissimus dorsi mm,
acquisitions over the period of rehabilitation. and sacrocaudalis dorsalis mm complex. These studies determine
In human medicine and rehabilitation, reliability data and clini- intra-operator reliability of repeated measurements of muscle CSA,
cal rehabilitative applications are more accessible, and US is widely using magnetic resonance imaging (MRI) as the gold standard for
used for functional muscle diagnosis, real time analysis during treat- comparison. MRI images were taken in 1 cm axial slices from T11
ment, and as an objective measure of motor control enabling the to the tail. Measurements of CSA were repeated three times using
practitioner to monitor rehabilitation and performance (Hides the standard Bruker analysis package on each image of the multifidus
et al., 1992; 1994, 1995; Kidd et al., 2002; Hodges et al., 2003b; muscle at four vertebral regions (T13/14, T18/L1, L3/4, L5/6) and
McMeeken et al., 2004; Whittaker, 2007). The methodology and the sacrocaudalis dorsalis muscle complex at the S3 (Figs 17.3 &
intra-operator reliability for the use of ultrasound imaging to 17.5). Digital photography was used to record the CSA in the same
measure skeletal muscle has been described for the following spinal segments that were cut in cross-section after freezing. Ana-
muscles: anterior tibial; quadriceps; transversus abdominis; multifidus; tomical dissection verified the anatomical relationship of multifidus,
rectus abdominis; internal and external obliques; erector spinae; splenius longissimus dorsi and sacrocaudalis dorsalis muscles and comparisons
capitis, semispinalis capitis; masseter; iliopsoas; muscles of the pelvic were made with MRI (Fig. 17.5). Median coefficient of variation
floor; and the diaphragm (Hides et al., 1992; Martinson & Stokes, (CV) for all individual MRI readings was 1.7%, showing good intra-
1991; Stokes & Young, 1986; Kelly & Stokes, 1993; Blaney et al., operator reliability of measurement of multifidus CSA using MRI.
1999; Norasteh et al., 2007; Ferreira et al., 2004; Watanabe et al., The pooled SD for all readings was calculated to be 0.22 cm2 with
2004; Raadsheer et al., 1994; Barker et al 2004; Rezasoltani et al., 95% confidence interval ± 0.43 cm.

Fig 17.5  Ultrasonographic images of multifidus (M), sacrocaudalis dorsalis medialis (SM) and sacrocaudalis dorsalis lateralis (SL) muscles at the levels of the
13th thoracic vertebra (T13) (top left), the 1st lumbar vertebra (L1) (top right), the 5th lumbar vertebra (L5) (bottom left), and the 3rd sacral vertebra (S3)
(bottom right). The vertebral spinous process (SP) forms the medial border of these muscles. Longissimus dorsi (L) lies on the lateral side in the
thoracolumbar region and biceps femoris (B) is on the lateral side at S3. The ventral border at T13 is the costovertebral joint (CVJ), at L5 is the transverse
process (TP) and at S3 is the lateral sacral crest (LS). The gluteus medius (MG) and iliocostalis (I) muscles are also shown.
Reprinted from Stubbs, N.C., Clayton H.M., Hodges P.W., Jeffcott, L.B. and McGowan C.M. (2010) Osseous spinal pathology and epaxial muscle ultrasonography in Thoroughbred racehorses.
Equine Vet. J. 42 (Suppl. 38), 654–661, with permission from the Equine Veterinary Journal.

395
 17 Rehabilitation of the locomotor apparatus

McGowan et al. (2007c) further investigated the reliability of back pain, showed that 91% had asymmetrical development of
measuring multifidus CSA using a 4–7 MHz curved linear probe at biceps femoris muscle and 82% had asymmetrical gluteal muscle
a depth of 15 cm. Two skilled examiners independently collected mass (McGowan et al., 2007c). Ultrasonographic images were
three images at each site. CSA measurements were analyzed blindly acquired on the left and right sides at 5 TL levels to measure CSA
on three occasions by both examiners using Image J software. Data of multifidus or sacrocaudalis dorsalis. At necropsy, osseous pathologi-
were analyzed for overall intra- and inter-operator reliability at each cal lesions of the TL spine and pelvis were recorded according
level, and measurements of left and right sides in the same horse to spinal level and anatomical site, and graded (0–3) according to
were compared. Images taken over the intervertebral space at all severity. The mean typical measurement error in estimating
levels were most repeatable, with greater variability over the facet multifidus/sacrocaudalis dorsalis CSA was used to determine the sig-
joints possibly due to differences in facet size and/or shape. Reli- nificance of left/right asymmetries. An association between multifi-
ability for multifidus CSA was good: Interclass coefficient (ICC) was dus CSA asymmetry and asymmetrical grading of pathological
0.83; standard error of the measurement (SEM) was 0.78 cm; and lesions was sought using Pearson’s chi-squared analysis. All horses
the smallest detectable difference was 1.5 cm across all levels. The had significant left/right asymmetry of multifidus CSA at >2 spinal
relative variance components of measurement of CSA were 56% levels, most commonly at L5 with a total of 74 sites affected in the
inter-operator variance, 11% intra-operator variance and 33% image 22 horses. Seventeen horses had severe (Grade 3) pathology and 16
variance. Hence the variance was minimized by having just one of these had ipsilateral atrophy of multifidus/sacrocaudalis dorsalis.
operator and taking several images of each site. Comparing repeat- There was a significant association between pathological grade and
ability of US measurement of CSA with the gold standard of MRI, degree of multifidus asymmetry. Severe osseous pathological changes
the pooled SD was 0.4 cm2, almost double, but still within accept- were common in this population of TB racehorses and were associ-
able limits. Longissimus dorsi CSA could be measured in the lumbo- ated with measurable left/right asymmetry in multifidus at or close
sacral region (L5), just cranial to the tuber sacrale, where it narrows to the level of pathology (Fig. 17.6). Ultrasonography of multifidus
down sufficiently to be entirely visible in the US image. muscles is a useful and reliable clinical tool in the functional diag-
In normal horses without back pain the epaxial muscles are sym- nosis and rehabilitation of back problems in horses, potentially
metrical on the left and right sides. Ultrasonography of thoracolum- together with other US functional muscular measures (linear, and
bar and sacral epaxial musculature revealed significant individuality contraction measurements).
and regional variation in the shape and size of multifidus, which is B-mode ultrasonography, which has the ability to produce images
largest in the lumbosacral region. From a functional biomechanical of moving objects almost instantaneously, has been used to measure
perspective this would be expected since maximal dorsoventral CSA of longissimus dorsi (D’Angelis et al., 2007). Images were
flexion and extension occurs at the lumbosacral junction, and the acquired before, during and after a 90-day aerobic training program
primary role of multifidus mm is stability and proprioception on a treadmill. With the horse stationary, images were recorded at
(Stubbs et al., 2006). the level of the last rib with longissimus dorsi relaxed. CSA of longis-
Ultrasonography is a repeatable and reliable tool for measuring simus dorsi increased significantly during the exercise period. Unfor-
epaxial muscle size in clinical cases of equine back pathology tunately, reliability data were not published for the US methodology
(McGowan et al., 2007c). Stubbs et al. (2010) reported that in 22 with respect to both image acquisition and calculation of CSA.
TB racehorses there was a significant reduction in multifidus size In the human field advancements have also been made in the
(CSA) at the level of significant injury or pathology seen on post measurement of muscle contraction over time using M-mode ultra-
mortem examination. This is comparable with research in people sonography. This technique is reliable within a single operator
with back pain (Hides et al., 1994, 1996) in which US imaging has when the probe is fixed to the antero-lateral abdominal wall to
shown a significant reduction in CSA of multifidus on the symptom- measure transversus abdominis in supine, standing and walking
atic side of the spine, indicating a relationship between pain and human subjects with and without back pain (Bunce et al., 2002;
muscle atrophy (Hides et al., 1994). Pre-clinical examination of 22 Kidd et al., 2002). To date there are no reports in the literature using
racehorses presented for euthanasia for primary reasons other than M-mode ultrasonographic muscle assessment in the horse.

Fig 17.6  Left and right ultrasonographic images of multifidus (M) at the 5th lumbar vertebra. Osseous pathology on the right side (facet and neural arch
fractures with incomplete vertebral body spondylosis) matches the reduction in multifidus CSA on the right. (Left mean CSA 15.07 cm2 (SD ± 0.21); right
mean CSA 11.21 cm2 (SD ± 0.24); absolute difference: 3.86 cm2.)
Reprinted from Stubbs, N.C., Clayton H.M., Hodges P.W., Jeffcott, L.B. and McGowan C.M. (2010) Osseous spinal pathology and epaxial muscle ultrasonography in Thoroughbred racehorses.
Equine Vet. J. 42 (Suppl. 38), 654–661, with permission from the Equine Veterinary Journal.

396
 Rehabilitation techniques: manual therapy

In conclusion, US is a valuable functional, objective, diagnostic

tool that can be used to monitor the rehabilitation process, espe-
cially in cases where objective measures are difficult to derive,
such as back pain (Stubbs et al., 2011) which will be discussed later
in this chapter.

Rehabilitation techniques: manual therapy

The modern sports medicine literature indicates that rehabilitation

of an athlete should begin as soon as possible after an injury occurs,
in combination with therapeutic measures such as anti-inflammatory
and other pain killing agents (Frontera, 2007). Complementary and
alternative medicine (CAM) has applications in performance
enhancement, injury prevention and therapy, especially with regard
to back pain in the equine athlete. Examples of CAM treatment are
PT, chiropracty, osteopathy and acupuncture (Chan et al., 2001,
Haussler, 1999b), exercise-based therapies and swimming (Roeth-
lisberger Holm et al., 2006). To educate and advise the owner/rider
of horses with back pain, riding instructions are also vitally impor-
tant in successful rehabilitation (Marks, 1999).
Manual therapy techniques are recommended for many condi-
tions following a veterinary patho-anatomical diagnosis. Within the
PT profession, manual therapy is defined as clinical methods using
skilled, specific hands-on techniques, including but not limited to Fig 17.7  A therapist performs myofascial release, a deep soft tissue
manipulation/mobilization to diagnose and treat soft tissues and mobilization technique, in the cervical spine.
joint structures for the purpose of modulating pain; increasing
range of motion (ROM); reducing or eliminating soft tissue inflam-
mation; inducing relaxation; improving contractile and non-
contractile tissue repair, extensibility, and/or stability; facilitating ultimately improve motor control. Addressing these impairments is
movement; and improving function (American Academy of Ortho- a key feature of the rehabilitation process in an attempt to regain
pedic Manual Physical Therapists). A chiropractic definition of function and optimize performance.
manual therapy is ‘procedures by which the hands directly contact It has been reported in the clinical literature that following soft
the body to treat the articulations and/or soft tissues’ (Gatterman tissue trauma or pathology, dynamic and or passive stretching can
& Hansen, 1994). be implemented to regain and maintain full range of motion during
Degenerative joint disease (osteoarthritis) is an example of a the healing process. As a preventative measure this also applies to
common cause of lameness and loss of performance in the horse. any horse that is on restricted exercise or confined to the stable for
It has been reported clinically that manual therapy can be success- other reasons, to assist with avoiding the detrimental effects of
fully applied, as in human cases. Many manual therapy techniques immobilization (Paulekas & Haussler, 2009).
such as passive mobilizations and manipulation can be applied to Passive stretching techniques are difficult to apply because horses
the articular system including the vertebral column and peripheral are usually standing during treatment and this requires persistent
joints, with complementary soft tissue techniques applied to the muscle tonicity. Therefore the stretching procedure is always
associated neuromuscular and fascial tissues. These repeated move- dynamic and never completely passive, even when intermittently
ment techniques have been reported to have effects on the intra- ‘stretching’ the horse’s limbs. This may also be the case in the
articular, peri-articular (capsular, ligamentous) and extra-articular unsedated human according to a newer sensory theory suggesting
(muscular, fascial, neural) structures, thereby affecting the passive modulation of muscle activity, rather than an increase in extensibil-
and active constraints of the joint complex and assisting in pain ity of the myofascial and neural systems (Weppler & Magnusson,
modulation (Goff & Jull, 2007; Haussler, 2009; Goff, 2009). The 2010). Until recently the proposed physiological mechanisms to
human literature supports manual therapy techniques and the explain the immediate increased extensibility of muscle as a result
underlying biomechanical and neurophysiological mechanisms, of intermittent stretching have been predominantly passive and
and many of the studies have used animal models (Wright, 1995; mechanical in nature including: viscoelastic deformation (Chan
Vicenzino et al., 1998, 2007; Sluka et al., 2006). Evidence has et al., 2001; Willy et al., 2001), plastic or permanent deformation
shown that manual therapy produces an initial local treatment- (Warren et al., 1971; Sapega et al., 1981; Wessling et al., 1987),
specific hypoalgesic and sympathetic-excitatory effect beyond increase in sarcomeres in series with sustained prolonged stretching
placebo or control conditions, followed by a non-opioid-mediated (Gajdosik, 2001) and neuromuscular relaxation as a result of a
systemic hypoalgesia (Sluka et al., 2006). stretch reflex, many of which are not supported experimentally
Numerous myofascial and neuromuscular mobilization tech- (Sharman et al., 2006). In the 1990s, many researchers tested the
niques can be used, including massage (friction, effleurage, petris- mechanical theories of stretching and found that the only change
sage, tapôtement, vibrations and shaking), trigger point therapy, in passive torque/angle curves was an increased joint angle at end
direct and indirect myofascial release, positional release, reflex inhi- range with the same applied torque. It was concluded that the end
bition techniques, craniosacral therapy, adverse neural tension tech- point was determined by a subjective sensation (Magnusson et al.,
niques and stretching (mobilization exercises) (McGowan et al., 1996a, b; Weppler & Magnussun, 2010). The mechanism of altered
2007a; Paulekas & Haussler, 2009) (Fig. 17.7). The aim of many myofascial extensibility is still under debate in the human stretching
manual therapy techniques directed at the myofascia, tendons and literature, however the current theory indicates that stretching alters
ligaments is to normalize tissue irritability, muscle ‘tone’, which has pain perception and thus there is more tolerance immediately after
also been demonstrated in the horse (Wakeling et al., 2006), exten- stretching and in the short term (3 to 8 weeks) (Folpp et al., 2006;
sibility, length, contractility, strength and coordination, and Weppler & Magnussun, 2010).

397
 17 Rehabilitation of the locomotor apparatus

Rose et al. (2009) studied the effects of two different 8-week walking program along with other forms of therapy as soon as the
stretching regimes on stride length and range of motion (ROM) in lesion is determined to be stabilized. It is the authors’ suggestion
the equine trot. There was no significant effect on stride length but to also include unmounted dynamic mobilization exercises to
a number of significant differences between treatments were found increase the dynamic ‘core’ muscle strength as described above
in ROM of the shoulder, stifle and hock joints, which suggested (Stubbs & Clayton, 2008; Clayton et al., 2010a, 2012; Stubbs et al.,
negative biomechanical effects of the stretching. However, the meth- 2011).
odology was flawed and the results need further clarification; mul-
tiple variables including individual horse variation may have played
a role and the specific stretching techniques and forces applied need Evaluation of information on manipulative/
to be quantified in a larger number of subjects. mobilization intervention
An osteopathic technique that is reported in the clinical literature
suggests that a form of gentle passive stretching (‘positional release’) Specific techniques in PT, chiropracty and osteopathy are passive
in the treatment of chronic myofascial restrictions can be performed accessory and physiological joint mobilization and manipulation
under sedation or anesthesia. Clinically this technique has been techniques that have been clinically reported to be successfully
reported to have some success (Pusey et al., 2010). The technique adapted to treat the horse (Goff, 2009; Haussler, 2009; Paulekas &
is often utilized after all other avenues have been exhausted due to Haussler, 2009), with some evidence indicating positive effects in
the anesthetic risk. These techniques are performed rhythmically or reducing muscle tone and increasing range of motion (Wakeling
in a sustained manner dependent on the desired response and effect et al., 2006; Gomez-Alvarez et al., 2008). These techniques encom-
on the tissues. Therefore tissue and/or joint ‘mobilization with pass the application of very specific passive and/or active assisted
movement’ exercises may be a more apt title for many forms of movements by the therapist to the horse, to manage and/or alter
stretching techniques in the horse, an example being those exercises pain and dysfunction of the articular, neural and muscular systems.
described in Stubbs et al., (2011), which have been coined in the Manual therapy techniques are based on a wide range of methods
clinical and lay literature as ‘neck and back stretches’, which is and theories related to the intervertebral and peripheral joint com-
discussed later in this chapter. plexes, myofascia and the neuromuscular system.
If there is a primary lesion of a soft-tissue structure such as Passive mobilizations and manipulations are applied at different
muscle, ligament or tendon, it is the authors’ suggestion that these amplitudes, velocities and directions determined by the assessment
active mobilizations with movement exercises of the affected struc- procedure and biomechanics of the joint complex (Goff & Stubbs,
tures and/or the whole horse are appropriate, in conjunction with 2007). Figure 17.8 shows an active assisted direct modified ‘Mait-
other therapies. The mobilizations may be appropriate to avoid land’ mobilization technique into end of range of motion (Mait-
excessive scar formation and disorientation of the fiber alignment land et al., 2000). In this technique, the therapist facilitates lateral
in the subacute and chronic phases of healing, along with primary bending and dorsoventral flexion to attain end range of motion of
or secondary disuse atrophy. With respect to rehabilitation of distal the thoracolumbar spine by applying pressure to one side of the
tendon and ligament lesions, it has been widely reported in the hindquarters (pressure from the middle of the medial gluteal to the
human literature that dynamic motor control-based exercises base of the sacrum). The position is maintained while applying a
including eccentric loading exercises accelerate healing and tensile lateral flexion and rotation oscillating mobilization to the lateral
strength, especially in relation to Achilles tendon lesions (Kingma dorsal spinous process at the desired amplitude and frequency. This
et al., 2007). This type of rehabilitation strategy, sometimes referred technique, if modified appropriately, can become what is termed in
to as active rest, is being implemented in horses by commencing a the human literature a ‘Mulligan mobilization-with-movement

Fig 17.8  Visual demonstration of an active assisted direct modified ‘Maitland’ mobilization technique into end of range of motion (Maitland et al., 2000). The
therapist facilitates lateral bending and dorsoventral flexion to attain end range of motion of the thoracolumbar spine by exerting pressure to one side of the
hindquarter (pressure from the middle part of gluteus medius to the base of the sacrum). While maintaining this position, a lateral flexion and rotation
oscillating mobilization technique with lateral flexion and rotation is applied laterally to the dorsal spinous process at the desired amplitude and frequency.

398
 Rehabilitation techniques: manual therapy

technique’. These techniques have been positively reviewed in the (n = 12) and control (n = 12) groups. Outcomes were measured
scientific literature with regard to producing immediate and sub- each week pre- and post-intervention and showed significantly
stantial pain reduction and improved function. Vicenzino et al. higher amplitudes and applied forces post intervention in the treat-
(2007) reviewed the clinical efficacy, effects and putative mecha- ment group compared to the controls, with a trend for increased
nisms of action of the approach, including the mechanisms and passive spinal stiffness. Further studies with improved objective
action in both biomechanical and pain science paradigms. This measures of musculoskeletal dysfunction are needed to assess the
technique (Fig. 17.8) requires the oscillatory mobilization and effectiveness of manual therapies (Haussler, 2009).
response techniques to be applied simultaneously and repeated
during multiple oscillations, hence both active lateral bend and
dorsoventral flexion occur whilst the mobilization technique is Evaluation of information on exercise and
applied. Manual passive assessment and treatment techniques
directed at a joint, soft tissue or neural structure are well tolerated facilitation techniques
by the horse due to the rhythmical motion being applied at a com- Exercise-based treatment techniques:
fortable speed as has been widely reported in people (Hurwitz et al.,
1996). This may include high-velocity, low-amplitude thrust motor control
(HVLA) techniques. In the human literature it is hypothesized that In the human rehabilitation literature much emphasis is placed on
the velocity may be more important than the amplitude, as dem- functional assessment and objective measurements in relation to
onstrated and standardized by using a mechanical ‘activator’ to neuromotor control. As described in Chapter 4, locomotion can be
perform the technique (Fuhr & Meke, 2005). Although the HVLA viewed as an emerging pattern within a complex system initiated
involves high speed, horses also appear to tolerate these techniques by spinal central pattern generators. Locomotion involves central
well, as they take place in a very short period of time, in conjunction neural structures, peripheral organs and interactions with the envi-
with what is theorized as a rebound parasympathetic reaction and ronment (Latash, 2008). In equestrian sports, the rider must be
thus relaxation (Haussler, 2000; Pickar, 2002; Triano, 2005; Haussler included (Peham et al., 2004; Valentin et al., 2010). Dynamic
et al., 2010). control involves a spectrum of control strategies during locomotion,
The effect of manual therapy in horses in relation to equine back ranging from muscular co-contraction causing stiffening, to more
pain has been evaluated (Haussler et al., 1999a, 2007; Faber et al., dynamic control strategies that involve carefully timed muscle activ-
2003; Colborne et al., 2004; Wakeling et al., 2006; Gomez-Alvarez ity and movement. Multiple factors influence selection of the
et al., 2008). These studies highlight the difficulties in conducting appropriate dynamic control strategy: load, movement, predictabil-
manual therapy research, especially across time. There is certainly a ity, proprioceptive function and error-tolerance or robustness
lack of randomized clinical trials. Methodological limitations are (Hodges & Cholewicki, 2007). An essential element of motor
primarily due to limited subject numbers, variations within and control is accurate internal and external receptor activity, especially
between horses and practitioners, and the difficulty in the differen- regarding proprioceptive feedback on joint position during locomo-
tial diagnosis and monitoring of back pain. Even with modern tion, which can be altered due to pathology. In the last two decades,
motion analysis systems there are inherent difficulties in measuring much of the research in human PT and sports medicine in relation
intervertebral motion accurately and non-invasively (see Chapters to musculoskeletal disorders has focused on the neurosciences
2 and 10), and in measuring pain objectively. In spite of these limi- including neuromotor control and pain associated with musculo-
tations, joint mobilization and chiropractic manipulations have skeletal conditions, as well as the inter-relationship between pain
been shown to affect the kinematics of the back. For example, kine- and function. This complements work in biomechanics and kine-
matic measurements following treatment showed less extension in matics. In this research, functional impairments and interventions
the thoracic region and a greater symmetry of pelvic motion (Gomez have been studied via randomized clinical trials investigating and
Alvarez et al., 2008). Haussler, (2009) recently reviewed the scien- comparing one or more approaches.
tific literature supporting manipulative therapy in the horse, where One of the major and current areas of interest in human muscu-
the focus of chiropractic has been on the clinical effects. In a pilot loskeletal PT is the field of neuromotor control and dynamic stabil-
study, Haussler et al. (1999a) showed that manipulative techniques ity. Research in this area has allowed major advancements in the
produce substantial spinal motion. In one study manipulative prevention and treatment of important problems in people includ-
therapy modulated MNT, and a clinical trial using PA indicated that ing pelvic pain (Richardson et al., 2002) and low back pain (Hodges
manual and mechanical spinal manipulation increase MNTs (Sul- et al., 2003a; Hodges, 2005), as well as prevention and treatment
livan et al., 2008; Haussler & Erb, 2003). Further preliminary studies of peripheral joint injuries (Reimann & Lephart, 2002a, b) and
have investigated the effects of manipulative techniques on verte- performance enhancement in athletes (Saunders et al., 2005).
bral mobility and muscle tone (Haussler et al., 2007; Wakeling These are also key issues in equine performance and wastage (Jeff-
et al., 2006). Wakeling et al. (2006) investigated spinal (McTim- cott et al., 1982; Stubbs et al., 2010) and research in these areas for
oney) manipulations (Colloca et al., 2009) and reflex inhibition horses is clearly warranted and currently underway in a number of
therapy versus a control group in 26 randomly assigned horses. research facilities around the world.
Compared with controls, both treatment groups had significantly For example, research in human back pain, spinal stability and
less muscle tone as measured by EMG after treatment. Collectively, control of movement is highly dependent on the contribution of the
these studies suggest that manipulative therapy elicits changes in TL muscular system. In the past, attention focused on muscle strength
and pelvic kinematics, and muscle tone, which are likely to be alone. However, it is now recognized that the central nervous sys-
beneficial (Faber et al., 2003; Gomez Alvarez et al., 2008). However, tem’s control of the muscular system (when and how the muscles
further investigation is warranted to improve the objectivity of the work) is probably of greater importance to the muscle system’s
methodology, decrease the number of variables, and increase ability to satisfy the needs of spinal movement and stability. The
sample size. central nervous system must plan suitable strategies of muscle
Recently Haussler et al. (2010) reported on the efficacy of dorso- recruitment, co-ordination, and levels of activity to meet the
ventral spinal manipulation and mobilization on trunk flexibility demands of internal and external forces and initiate appropriate
and stiffness in 24 actively ridden horses. Passive spinal mobility responses to unexpected disturbances of movements and function.
was assessed once per week for 3 weeks by measuring peak vertical Research has shown that the deep, local muscles of the TL region
displacement, loading and unloading velocities, applied force and (for example, the transversus abdominis and the segmental lumbar
frequency of truncal oscillations induced during dorsoventral spinal multifidi) play key roles in modulating the stiffness of the lumbar
mobilization at 5 TL sites. This was compared between treatment spinal segments and pelvic joints during limb and lumbo-pelvic

399
 17 Rehabilitation of the locomotor apparatus

movements (Indahl et al., 1997; Richardson et al., 2002; Hodges trunk muscles are then re-educated to retrain painless and con-
et al., 2006; Kalichman et al., 2009; Mok et al., 2011). It has also trolled functional activities. Progressively the stability system is
been shown that the central nervous system pre-programs activity functionally challenged with load (static and dynamic exercises) as
in certain trunk muscles in preparation for limb movement. For control improves. Most importantly there is growing evidence that
instance, the transversus abdominis and multifidi activate prior to limb this exercise approach can reduce low back pain and possibly reduce
movement, regardless of direction. This serves to increase segmental its recurrence rate (O’Sullivan et al., 1997; Hides et al., 2006; Mac-
stiffness for spinal segmental support prior to loading (Hodges & Donald et al., 2006). Specific physiotherapeutic intervention in
Richardson, 1997, 1999; Hodges et al., 2003b; Holm et al., 2002; people with multifidus dysfunction following an episode of acute
MacDonald et al., 2010). Back pain patients display delayed activa- back pain reduced the rate of recurrence of injury to 30% in the PT
tion of the transversus abdominis and multifidi, depriving the painful intervention group compared with a recurrence rate of 84% in
and injured spinal segments of timely support (MacDonald et al., controls (Hides et al., 2001).
2009). Findings to date indicate that the horse follows the same The human motor control model has been applied in investiga-
sequence of events (Stubbs et al., 2010). Further research into the tions of the equine multifidus muscle (Stubbs et al., 2006, 2011)
activation patterns in normal horses and those with back pain is revealing striking similarities in structure and function across
necessary and currently underway. species. The equine multifidus has similar morphological orienta-
Until recently relatively little was known about other epaxial and tion; therefore, its biomechanical activity is comparable to man,
hypaxial muscles in normal horses or those with back pain. In the with the primary function being to stabilize intervertebral motion.
human literature there is a vast body of evidence (>80 peer reviewed Sacrocaudalis dorsalis lateralis forms the caudal continuation of mul-
research papers: Pubmed), with in vivo and in vitro observations of tifidus in the equine sacro-caudal spine (Stubbs et al., 2006) (Figure
the role of transversus abdominis and multifidus muscles in controling 17.5). Multifidus CSA is largest in the lumbosacral region, which is
intervertebral motion. The multifidi provide intersegmental stabili- the area with the greatest amount of motion (Townsend et al.,
zation and stiffen the spine, contributing two-thirds of the total 1983; Stubbs et al., 2006).
increase in spinal stiffness imparted by muscular action (Wilke Anatomical variations in the equine lumbosacral region occur in
et al., 1995). In vivo studies in pigs confirmed that the multifidi are at least a third of horses (Haussler et al., 1999b). The predominant
also a major stabilizer of lumbar intersegmental motion in quad- variation is divergence between the dorsal spinous processes of L5
rupeds (Kaigle, 1995). Morphological changes occur in the multifi- and L6, with L6 being a transitional vertebra and functioning as
dus muscle in association with low back pain. In both acute and part of the lumbosacral joint (Stubbs et al., 2006). These variations
chronic low back pain, reduced CSA of multifidus was observed may affect range of motion and performance.
within 24 h of injury at the same intervertebral level and on the In people with back pain, the deep spinal stabilizer multifidus is
same side as the spinal pathology (Hides et al., 1994, 1996; Hodges inhibited ipsilaterally leading to muscle atrophy, asymmetry in CSA
et al., 2006; 2009). In a recent study, a decrease in density of and loss of its function leading to intervertebral instability. Specific
multifidus/erector spinae was found in association with facet joint physiotherapeutic exercises are required to reactivate multifidus. This
osteoarthritis, spondylolisthesis and disc narrowing (Kalichman information has influenced the management of back pain in humans
et al., 2009). This also occurs within 3 days after a unilateral experi- (MacDonald et al., 2006, 2009). In horses, Stubbs et al. (2011)
mental lesion to an intervertebral disc in pigs (Hodges et al., 2006). applied these physiotherapeutic and motor control principles to the
Another interesting finding is that multifidus does not automati- horse in a study that assessed the effects of dynamic mobilization
cally resume its normal function following recovery from or resolu- exercises on size and symmetry of multifidus in the equine caudal
tion of an episode of acute back pain (Hides et al., 1996). Specific thoracic and lumbar spine. Eight horses performed dynamic mobi-
physiotherapeutic interventions (exercises) are required to restore lization exercises (three cervical flexions, one cervical extension, and
the size and function of multifidus after an episode of acute back three lateral bending exercises to the left and right sides) with five
pain in people and these interventions reduce the rate of recurrence repetitions/exercise/day on 4–5 days/week for 3 months during
of injury from 84% in untreated controls to 30% (Hides et al., which time they were not ridden (Fig. 17.9). These exercises are
2001) in patients who received specific interventions. In horses, widely used in clinical practice with the aim of increasing ‘core’
generalized secondary atrophy of the epaxial muscles, especially muscle activation (Stubbs & Clayton, 2008). Left and right multifidus
longissimus dorsi and gluteus medius have been reported in horses CSA was measured ultrasonographically at 6 levels from T10–L5 at
with back pain (Jeffcott et al., 1982; Quiroz-Rothe et al., 2002). the start (initial evaluation) and end (final evaluation) of the
Changes in multifidus CSA associated with ipsilateral osseous 3-month study. Changes in CSA of the left and right multifidi and
pathology in the horse has been reported (Stubbs et al., 2010). symmetry of multifidus CSA on the right and left sides between the
Furthermore, the muscles hypertrophy in response to regular per- two evaluations were sought using analysis of variance (p < 0.05).
formance of dynamic mobilization exercises (Stubbs et al., 2011). Between the initial evaluation and final evaluations multifidus CSA
Knowledge gained from ultrasonographic research in relation to increased significantly at all six spinal levels on both right and left
changes in neuromotor control associated with back pain has trans- sides. Asymmetries in multifidus CSA between the right and left sides
lated to the development of new rehabilitation strategies for the decreased significantly between the initial and final evaluations at
lumbo-pelvic muscles in human back pain patients. This is an all six spinal levels indicating the dynamic mobilization exercises are
example of anatomical and biomechanical research creating a effective in activating the equine multifidus (Stubbs et al., 2011).
framework for future neuromotor control research in the horse Clayton et al. (2010a, 2012) reported the intervertebral angula-
along the same lines as the human research model. MacDonald and tions in the end position of the dynamic mobilization exercises
colleagues (2006) recently reviewed the evidence of these treatment performed in flexion and lateral bending, respectively. Most of the
strategies in clinical practice. Rehabilitation in this context using movements occurred at the cranial and caudal cervical joints with
ultrasonography-guided feedback places emphasis on motor less motion in the mid neck. The results were interpreted in the
relearning to optimize motor control for spinal dynamic stability. context that articulations at the cervicothoracic junction are primar-
The rehabilitation first uses the end organs of the neuromotor ily responsible for raising, lowering and turning the neck while the
system, the muscles, with the aim that cognitive, repeated contrac- cranial cervical joints position the head relative to the neck, which
tions of the muscles and correct movement patterns will result in a is important with regard to proprioceptive input.
transition to automated use (i.e. skill training) (O’Sullivan et al., Regular performance of dynamic mobilization exercises may also
1998). Initially, the deep muscles such as transversus abdominis and have neuromuscular training effects as a consequence of the activa-
lumbar multifidus are repeatedly activated in the relearning process tion, recruitment, and strengthening of the deep dynamic segmental
during rehabilitation. Movement patterns and strategies for all stability muscles in the neck and back (longus colli, intertransversarii,

400
 Rehabilitation techniques: manual therapy

A B C D

E F G

Fig 17.9  Dynamic mobilization exercises (baited stretches) performed in flexion–extension (top row): (A) chin to chest, (B) chin between carpi, (C) chin
between fetlocks, and (D) cervical extension; and in lateral bending (bottom row): (E) chin to girth/ribcage, (F) chin towards stifle, and (G) chin towards hock.
Reprinted from Stubbs, N.C. and Clayton, H.M. (2008) Activate Your Horse’s Core: Unmounted Exercises for Dynamic Mobility, Strength, and Balance., with permission from Sport Horse
Publications.

multifidi); the thoracic sling (primarily serratus ventralis thoracis); the Jeffcott, 1986), which is conisitent with the finding that the direc-
hypaxial musculature (abdominals, iliopsoas complex); and the tion of greatest movement is in the lateral and oblique planes (Goff
pelvic dynamic stabilizers (superficial and medial gluteals, tensor et al., 2006). In conclusion, research to date supports the use of
facia latae, biceps femoris). It is believed that the horse develops more manual provocation tests for SID in horses.
appropriate dynamic control strategies as a consequence of per-
forming these exercises. It is further suggested that, as the horse’s
dynamic stability and strength improve, the exercises should be
Facilitation-based exercise therapy
made more challenging, for example, by deliberately having the In conjunction with direct manual therapy techniques, facilitation
horse stand in a slightly unbalanced position or by teaching techniques may be utilized to cause the horse to move through a
the horse to perform the exercises with one leg raised off the ground. desired range of motion. These techniques and exercises encompass
It is important for the therapist to assess each horse’s compensation many forms of direct and indirect mobilizations with/through
strategies and to always encourage horses to use the appropriate movement techniques/exercises, which clinically have been seen to
dynamic control. be very effective, especially in relation to functional motor control.
Research into the biomechanics and neuromotor control of the Altered motor control (neuromuscular function) may be a result of
human sacroiliac joint (SIJ) has also contributed to clinicians’ an underlying lesion in the spine and/or peripheral joint disease
ability to diagnose sacroiliac disease (SID) in humans. It has been with pain and inflammation causing reflex inhibition of motor
shown that non-invasive, manual SIJ provocation tests are as predic- neurons, resulting in weakness and atrophy of associated muscles
tive for SIJ being the source of pain as diagnostic joint blocks (van (Young, 1993). Many of these mobilization techniques/exercises are
der Wurff et al., 2006). These pain provocation tests compress the reported to use neuromuscular reflexive responses along with mus-
SIJ articular surfaces and/or stress the extra-articular structures of cular facilitation and inhibition. In horses, these techniques are
the joint. Manual tests based on the examiner’s impression of the based on similar theoretical principles to those used in human
amount of motion (hypo/hypermobility) and quality of motion manual therapy, including muscle energy techniques to gain motion
(the status of the neuromotor system) at the SIJ are also used clini- that is limited by restrictions of neuromuscular structures and
cally to assess the functional status of the SIJ. This includes assess- restore or normalize motor control. The human patient is asked to
ment of the amount of movement of the SIJ during application of contract and relax specific muscles in a given range of motion and
manual force and, specifically, analysis of sacral movement relative often resistance is applied to the body part. In theory these tech-
to the pelvis in weight bearing through specific motion tests, such niques promote muscle relaxation by activating the Golgi tendon
as the one leg flexion test in standing or the ‘stork test’ (Gillet test) reflex (DiGiovanna et al., 2005) via the principles of reciprocal
(Lee, 2004). Diagnosis of equine sacroiliac disease using manual inhibition and post-isometric relaxation. Reciprocal inhibition uses
tests similar to those used in the human literature has been eluded the body’s antagonist–inhibition reflex to induce relaxation of a
to (Haussler, 1999b, 2003a, b; Varcoe-Cocks et al., 2006; Goff & muscle. Conversely, other techniques use the antagonist–inhibition
Crook, 2007). Motion between the sacrum and ilium (dorsal sac- reflex to incrementally restore range of motion, whereby a muscle
roiliac ligament provocation test) during the application of manual is stretched immediately following an isometric contraction due to
force has been reported and the direction and amount of movement the neuromuscular apparatus becoming briefly refractory or unable
of the ilium relative to a fixed sacrum in response to manual pres- to respond to further excitation.
sure has been measured in an in vitro model (Goff et al., 2006). The Although these techniques have not yet been investigated scien-
sacroiliac joint lies at an angle of 30o to the horizontal (Dalin & tifically, it is the author’s opinion that many of these applied

401
 17 Rehabilitation of the locomotor apparatus

techniques/exercises are very useful in the majority of rehabilitation the chest or between carpi/fetlocks stimulate flexion, whereas move-
cases to maintain and improve mobility, strength and dynamic ment of the chin toward the girth, flank or hock is achieved using
stability, especially in regions where the affected joint complexes coupled lateral bending and axial rotation. The aim of these exer-
are inaccessible due to the horse’s morphology such as the horse’s cises is not only to mobilize both the axial and appendicular skel-
cervicothoracic region and back. The intention is not only to eton, but also to facilitate core muscle activity (thoracic sling,
increase joint mobility but also to activate and strengthen muscles hypaxial/epaxial and pelvic musculature), improve neuromuscular
that move the horse’s body to the desired position, promote inter- control and strengthen the active musculature, which is a key
segmental joint stability, and restore functional locomotion and concept in rehabilitation and sports medicine.
sport specific motion. Indirect mobilization techniques can be uti- The human scientific literature indicates that core stability exer-
lized clinically to mobilize the cervicothoracic junction, whereby a cises improve neuromotor control in back pain patients to aid
muscle reflex/response occurs in response to applying constant recovery and prevent further reccurrence (Hides et al., 2001, 2006).
pressure between the index finger and the thumb while they are It seems likely that core training exercises as described in horses by
wrapped around the distal third of the brachiocephalicus muscle, Stubbs & Clayton (2008) will have similar therapeutic applications
producing flexion in this region. The brachiocephalic response can in horses that are rehabilitating from neck and back pain. Further-
be coupled with ventral pressure to the sternum to mobilize the more, regular use of the exercises may help to prevent the occur-
joint complexes by stimulating flexion in the cranial thoracic region. rence or recurrence of equine neck and back pain. Further studies
This may be combined with lifting of one forelimb to progress the in equine neck and back pain patients are needed to test this
technique (Fig. 17.10). hypothesis.
Many combinations of mobilization, core strengthening and A vital aspect of musculoskeletal and neurological rehabilitation
stability/balancing exercises have been described in the clinical lit- and performance enhancement is related to motor control and
erature (Stubbs & Clayton, 2008; Goff, 2009; Paulekas & Haussler, motor-skill retraining, not only with respect to pain management
2009; Stubbs et al., 2011). These often use food as an incentive or but directly related to and mediated by proprioceptive and mecha-
bait (such as a carrot) to encourage the horse to move into the noreceptive afferent feedback from joints, tendons, ligaments, fascia
desired posture, with the end range of motion posture being main- and skin, which modulate efferent neuromuscular control. As stated
tained for a period of 3–10 s. The exercises that move the chin to previously, the overall aim of rehabilitation is to restore the appro-
priate neuromuscular pathways, and strengthen the musculature to
allow the horse to return to perform to its athletic potential.
Sensory integration is a form of facilitation-based exercise therapy
that involves tactile stimulation during exercise. These techniques
have clinically (Goff & Stubbs, 2007; Stubbs & Clayton, 2008; Goff,
2009; Paulekas & Haussler, 2009) and scientifically (Ramon et al.,
2004; Clayton et al., 2008, 2010b, 2011a,b) been reported to be
useful tools in the treatment of the horse. The aim of these tech-
niques is to alter the horse’s kinesthesia (proprioception and joint
position sense), which ultimately influences static and dynamic
posture. Static and dynamic position senses are dependent on feed-
back from peripheral receptors that determine joint angulations in
all planes and their rate of change. Peripheral receptors include
cutaneous mechanoreceptors, nociceptors, Golgi tendon organs,
muscle spindles and capsular receptors, which relay information to
the central nervous system via type Ia and type II afferent fibers, to
modulate and coordinate locomotion (Loeb, 2005). Sensory inte-
gration is very effective because the horse has a heightened cutane-
ous mechanoreceptive system, in conjunction with the underlying
myofascial attachments of cutaneous trunci to the skin. Theoretically,
tactile stimulants including adhesive tape, exercise bands, loose-
fitting tactile limb bracelets (55 g), or tactile cues can be applied to
the skin over targeted regions including the limbs or specific
muscles, with the aim of altering mechanoreceptive and proprio-
ceptive feedback, which alters motor control. There is a vast body
of literature to support the application of these techniques in people
(Seki et al., 2003; Rose & Scott, 2003).
A series of studies has explored the effects of sensory integration
techniques in horses (Clayton et al., 2008, 2010b, 2011a,b). The first
study determined the short-term habituation to the effects of light-
weight (55 g) stimulators consisting of a strap with seven double
stranded oval brass links that were 7 cm in length. The straps were
loosely attached around the pastern. Trials were conducted under
three conditions: no stimulators, stimulators attached to the pas-
terns of both forelimbs and stimulators attached to the pasterns of
both hind limbs, with 10 consecutive trails collected for each condi-
Fig 17.10  Combined facilitation technique used to assess and treat tion (Clayton et al., 2008). Eight sound Arabians (450 ± 56.1 kg)
dynamic stability in the cervicothoracic and thoracic sling regions. The first were trained by an experienced handler to trot in a straight line
step is to perform a ‘sternal lift’ or ‘rounding response’ by pressing upward at a consistent velocity. Kinematic data were collected using an
on the sternum. The horse responds to the noxious provocation by flexing automated motion analysis system (eight cameras; data collection
the cervicothoracic spine with the flexed position being maintained for volume 5 × 3 × 2 m) with reflective markers attached to the head,
3–5 s. A progression of the exercise is to lift one forelimb off the ground trunk (T10 and L6), tuber spinae scapulae, tuber coxae, and each
while performing the sternal lift. hoof. Velocity did not vary significantly across all trials (± 0.4 m/s)

402
 Rehabilitation techniques: manual therapy

and hoof height in the control condition did not differ significantly Control Lightweight chain Leg weight +
across the 10 trials. As determined by the slope of the regression Strap Leg weight lightweight chain
lines, the effect on peak hoof height was largest immediately after
400
the application of the fore and hind stimulators and decreased over
time. Peak hoof height was significantly greater than the control 350
across the first six trials for the forelimbs and for all 10 trials for the
hind limbs with extrapolation of the regression line estimating that 300
hind hoof elevation would no longer be significant after the 14th

Hoof height (mm)

trial. There was, however, considerable individual variation in the 250
rate of habituation of different horses. It was suggested that habitu-
200
ation may be more rapid in the forelimbs due to the inherently
higher extensor tone of the forelimbs and the influence of ascending 150
neurological inhibition from the hind limbs to the forelimbs (De
Lahunta, 1983). Crossover effects between the limbs were observed 100
(Clayton, unpublished). The effects of longer-term habituation and
the response to repeated applications of the stimulators have not 50
yet been investigated. It is theorized that there may be a potential
0
effect on training due to altering kinesthesia and joint position 0 20 40 60 80 100
sense, thus modulating the central pattern generators via the extra-
Time (% swing)
pyramidal system and alteration of the horse’s gait over time, which
may be an ultimate goal during the rehabilitation process. Fig 17.11  Flight arc of right hind hoof without stimulators and with four
In a different study, swing phase kinematic and kinetic effects of types of stimulators attached around both hind pasterns. Each curve
the same tactile stimulators (55 g) loosely attached to the hind represents one trial with all trials being from the same horse.
pasterns were investigated in nine sound horses during four trials
at trot compared with a control condition, with the two conditions
applied in random order (Clayton et al., 2010b). Peak height of the lowest for the control (5.42 ± 1.38 cm) and loose straps (6.72 ±
flight arc and peak flexion angles were measured, and net positive 2.19 cm), significantly higher for lightweight tactile stimulators
and negative work performed across each joint were calculated (14.13 ± 7.33 cm) and leg weights (16.86 ± 15.93 cm) and signifi-
using inverse dynamics analysis and compared pairwise (tactile cantly higher again for leg weights plus lightweight tactile stimula-
stimulant versus control). Speed and stride duration did not change tors (24.35 ± 13.06 cm). The shape of the flight arc of the hind
but the lightweight tactile stimulators were associated with a reduc- hooves changed with loss of the biphasic shape as hoof height
tion in hind stance duration. The flight arc was higher as a result of increased (Fig. 17.11).
a significant increase in flexion at the stifle, tarsus, metartarsopha- Compared with no stimulators, peak joint flexions increased sig-
langeal and distal interphalangeal joints. Interestingly, the pattern nificantly at the stifle, tarsal and fetlock joints when using the
of the flight arc changed shape from biphasic to monophasic as a combination leg weights plus tactile stimulator and tarsal flexion
result of increased muscular work across the limb joints. Negative increased with the tactile stimulator (Table 17.4). There were large
work (eccentric contraction) increased across the stifle, MTP and differences in the response of individual horses that affected the
DIP, and positive work (concentric contraction) in the tarsal mus- ability to detect significant differences even with large changes in
culature but not the hip musculature. No changes in hip flexion or mean peak flexion values across conditions. Net energy generation
in the range of limb protraction/retraction were found. increased at the tarsal joint and net energy absorption increased at
Another study evaluated the response to weighting the hind pas- the stifle joint for lightweight chains, leg weights and leg weights
terns, which has previously been investigated by Wennerstrand plus lightweight chains but only the two weighted conditions were
et al. (2006). Swing phase kinematics and kinetics were measured associated with increased net energy generation across the hip joint
at trot (Clayton et al., 2011a) in six horses to compare control/no (Table 17.4).
weight condition with firmly attached leg weights (700 g). The two In conclusion, this series of studies has determined that both the
conditions were applied in random order. Peak height of the flight type and weight of tactile stimulators can alter the magnitude of the
arc was significantly higher with leg weights due to increased flexion kinematic and kinetic variables in the swing phase. The results
at the stifle, tarsus and metatarsophalangeal joints. Positive (con- suggest that the effects of tactile stimulation of the hind pasterns
centric) work increased in the tarsal and hip flexors in early swing on joint motion and muscle activation may be used in PT and
to lift the hoof and protract the limb and in the tarsal and hip rehabilitation to restore or increase flexion of the hind limb joints
extensors in late swing to lower the hoof and retract the limb. There with the exception of the hip joint. The authors suggest that ‘reha-
was also increased negative (eccentric) work across the stifle and bilitation after an orthopedic injury should progress from light-
metatarsophalangeal joints. The increased work was needed to over- weight tactile stimulators that restore range of motion by facilitating
come the additional inertia of the leg weights. In a rehabilitation stifle and tarsal musculature, to weighted stimulators that also
setting, a weight firmly attached to the pastern may be a useful strengthen the hip musculature’. The ability to stimulate concentric
method of facilitating and strengthening the musculature of the hip, activity of the tarsal musculature may have therapeutic applications
stifle and hock. The authors concluded that ‘increased muscular in conditions such as toe dragging (Clayton et al., 2010b). Lame-
activity occurs without an increase in hip flexion suggesting that it ness is also often associated with reduced joint ranges of motion,
will not improve hind limb engagement though the concurrent even after the lameness resolves. Clinically this is often associated
increase in tarsal flexion may give the impression of greater hind with degenerative joint disease or as a result of a period of immo-
limb activity’ (Clayton et al., 2011a). bilization. Van Harreveld et al. (2002) reported long-term reduced
Another in this series of studies made direct comparisons of the joint range of motion in sound horses after being cast for 7 weeks,
biomechanical effects of different types of stimulation devices even following a progressive 8 week exercise program that included
applied to the hind pasterns of trotting horses (Clayton et al., trotting on a treadmill. This could be a result of altered mechanical
2011b). Eight sound horses were tested under the following condi- properties of the articular and periarticular structures (Akeson
tions applied in random order: no stimulators, loose straps (10 g), et al., 1987; Michlovitz et al., 2004; Katalinic et al., 2009) or altera-
loose lightweight tactile stimulators (55 g), leg weights (700 g) and tions in neuromotor control (Jerosch & Prymka, 1996; Cowan
leg weights with tactile stimulators (700 g). Peak hoof height was et al., 2001). In the human literature there is evidence of

403
 17 Rehabilitation of the locomotor apparatus

Table 17.4  Effects of four types of hind limb stimulators on swing phase kinematics at trot

Type of stimulator None Strap Tactile stimulator Leg weight Leg weight + tactile stimulator
Peak hoof height (cm) 5.4 ± 1.4 a,b,c
6.7 ± 2.2 d,e,f
14.1 ± 7.3 a,d,g
16.9 ± 15.9 b,e,h
24.4 ± 13.1c,f,g,h

Peak flexion angle (degrees)

Hip 73.2 ± 9.6 74.6 ± 6.4 72.2 ± 9.6 71.1 ± 11.0 71.3 ± 9.9
Stifle 118.3 ± 10.2a,b,c 117.7 ± 6.1d 104.5 ± 15.8a 95.9 ± 21.2b 95.2 ± 13.8c,d
Tarsus 116.5 ± 5.0 a,b
114.6 ± 2.5 c
96.0 ± 15.1 a
87.7 ± 21.7 84.8 ± 13.5b,c
Fetlock 133.8 ± 8.7a 131.2 ± 9.0b 116.7 ± 18.1c 107.8 ± 20.8 101.5 ± 15.4a,b,c

Net joint energy (J/kg)

Hip 0.160 ± 0.099a,b 0.259 ± 0.026c 0.246 ± 0.046 0.367 ± 0.029a,c 0.320 ± 0.016b
Stifle −0.141 ± 0.057a,b,c −0.184 ± 0.018d −0.211 ± 0.030a,e −0.320 ± 0.036b,d,e −0.312 ± 0.043c
Tarsus 0.052 ± 0.024a,b,c 0.066 ± 0.006d,e 0.088 ± 0.16a,f 0.145 ± 0.051b,d 0.146 ± 0.038c,e,f

Values in the same row with the same superscript are significantly different from each other (p < 0.05).
Reprinted with permission from Clayton et al., 2011b.

underlying neuromotor control problems in human patients with This type of taping is commonly used in human athletes and is
back and knee pain, including abnormal knee joint position sense, known as kinesiology taping. In theory, the effects are achieved
altered joint proprioception (Baker et al., 2002), altered electro- through tension that increases the kinesthetic awareness along the
myographic (EMG) firing patterns of vastus medialis obliquus and fibers of a muscle or muscles when a facilitatory effect is desired.
vastus lateralis that contribute to maltracking of the patella (Cowan The horse performs specific exercises with the tape attached, includ-
et al., 2001; Owings et al., 2002), reduced knee flexion in early ing sport-specific activities under saddle, which leads to alteration
stance due to the delay in onset of eccentric activity in vastus media- of neuromotor function over time. The author has used taping
lis obliquus (Crossley et al., 2004) and significantly decreased techniques with clinical success to facilitate activity of the core
strength in the more proximal musculature including hip external dynamic stability musculature, more specifically biceps femoris and
rotators and abductors (Ireland et al., 2003) in patients with patel- the abdominal complex (Fig. 17.13). Further investigation into
lofemoral pain syndrome. Similar neuromotor control problems these sensory facilitation techniques is definitely warranted given
may occur in the horse and confound many gait abnormalities and the clinical improvements observed.
disease processes, as shown by subtle, and often related, changes in These sensory integration techniques may be combined with
lameness patterns and back kinematics (Gomez-Alvarez et al. many forms of in-hand and ridden exercises that have been
2007a,b, 2008) due to altered neuromotor control. The use of described in the clinical literature (Paulekas & Haussler, 2009;
tactile stimulation as a form of sensory integration may be a valu- Steiner & Bryant, 2003; Tellington-Jones & Burns, 1988). The exer-
able tool to facilitate restoration of normal locomotor function by cise protocol should be individually tailored and monitored by the
activating and strengthening specific muscle groups, with a training therapist, as there are vast individual variations in responses from
effect over time. the horse to these therapeutic strategies. A skilled handler/trainer is
Clinically it has been reported that other forms of sensory integra- required to perform the exercises on a daily basis. The exercise
tion are useful during training but further investigation is needed protocol should follow the principles of conditioning (Clayton,
to confirm their value. These include the use of different materials 1991), with ongoing re-assessment of the horse’s motion to ensure
such as bandages, such as Vetrap or ‘body wrap’, and exercise-bands that compensatory strategies are minimal and the desired effects are
wrapped around the horse’s hindquarter, abdomen and/or chest evident. Considerations should include exercise time, gaits, transi-
(Goff & Stubbs, 2007). It is the authors’ opinion that exercise bands tions, direction, surfaces and gradients/slopes, with constant moni-
are the most useful of these techniques in a clinical setting as they toring for signs of fatigue.
appear to increase the horse’s body awareness (kinesthesia) and the
ability to use the core muscles when exercised in hand, on the longe
or under saddle (Fig. 17.12). Hydrotherapy
Another effective technique for sensory stimulation in a clinical There is limited evidence in the literature regarding the musculo­
setting is the ‘functional proprioceptive taping technique’, which is skeletal benefits of equine hydrotherapy. Irwin & Howell (1980)
widely used in human rehabilitation, sports medicine and athletic suggested several indications for and some observations of swim-
performance. In the veterinary literature, Ramon et al. (2004) ming horses. Kato et al. (2003) determined the effects of immersion
reported that ‘rigid mechanical athletic taping’ of the fetlock did not in warm spring-water (38–40°C) on autonomic nervous system
alter the kinematics of the forelimb during stance, but did limit activity in horses. Variations in heart rate (HR) were evaluated from
fetlock flexion in the swing phase. There was also a decrease in peak the power spectrum in terms of low-frequency (LF) and high-
vertical ground reaction force, which may have been due to an frequency (HF) power as indices of autonomic nervous activity.
increased proprioceptive effect. The authors concluded that reduced They found that an increase in HF power, indicative of parasympa-
vertical force may be beneficial in preventing or reducing the risk thetic nervous activity, occurred during immersion in warm spring-
of injury and might be useful for tendinous or ligamentous reha- water, which may thus provide a means of relaxation for horses.
bilitation in equine patients. Kohn et al. (1999) proved that active cooling by washing with cold
Different effects may be seen when using a functional, rather than (15.6 ± 0.6°C) water is a safe, effective means of facilitating heat
mechanical, taping technique to facilitate increased muscle activity. dissipation of horses after exercise in a hot, humid environment.

404
 Rehabilitation techniques: manual therapy

Fig 17.12  Three applications of exercise bands to the horse. (A) Exercise band around the
hindquarters for exercise in hand or on the longe. (B) Exercise band around the hindquarters
for exercise under saddle. (C) Exercise band in the thoracic sling region attached ventrally to
C the girth, held in place at the manubrium via a ring, and then attached to the front of the
saddle on each side.

Evaluation of information on mechanical and temperature is reduced to therapeutic threshold of 10–19°C at

depths of 1–4 cm (Michlovitz, 1996). In the horse there has been
electro-physical intervention considerable research on using cryotherapy in the distal limbs,
It is beyond the scope of this chapter to define and describe all the where it has been shown to markedly reduce the severity of acute
electromodalities and biophysical agents that are used in PT and laminitis due to the reduction of enzymatic mediators (Van Eps &
rehabilitation to alter and/or accelerate tissue healing or to mini- Pollitt, 2010). The most effective form of cryotherapy in the equine
mize the effects of disuse atrophy, immobilization and denervation. limb is ice water immersion (Kaneps et al., 2000), which is a
Readers are encouraged to search the human literature for further common practice in many forms of equestrian sports post-exercise,
information because studies in the human field are more numerous such as eventing (Fig. 17.14). Horses are able to stand immersed in
and extensive than those in the veterinary field where many ques- ice below the carpi or tarsi for extended periods of time, but on
tions need to be addressed. In horses, the clinical efficacy of thera- other parts of the horse’s body it is suggested that the human recom-
peutic ultrasound and optimal methods of application are uncertain, mendations for cryotherapy should be adhered to. Wet ice applica-
with no peer reviewed studies to date. Factors to consider that can tions (wet towel with crushed ice) are best applied for up to 20 min,
alter penetration and absorption are the horse’s coat (which needs every 2–4 h, to avoid tissue damage and cold-induced vasodilation
to be shaved to apply therapeutic ultrasound) and thickness of the (‘hunting response’). Tissue damage has been shown to occur in
skin, fascia and subcutaneous fat. humans if the tissue temperature is reduced by more than 10°C.
It is well established in the human literature that cryotherapy is Superficial heat therapy (hot packs and hot water hosing) may be
an effective method of modulating pain and inflammation in the less useful in the horse. A study by Kaneps et al. (2000) showed that
acute phase of healing following injury (first 48 h), if the tissue subcutaneous and deep tissue temperatures never reached the

405
 17 Rehabilitation of the locomotor apparatus

A B

Fig 17.13  Examples of functional proprioceptive taping techniques using kinesiotape. (A) The abdominal complex is taped to facilitate activity of transversus
abdominis, internal oblique and rectus abdominis. (B) Taping technique to facilitate activity of the pelvic stabilizer muscles including: biceps femoris, gluteus
superficialis and medius, and tensor fascia latae.

Fig 17.14  Event horse after returning to its stable following the cross-country phase at an international competition. The horse is standing in boots filled
with a slush of ice and water and has a wet ice towel placed along its back.

therapeutic threshold of 41°C in the metacarpal region. It is not Electrotherapy entails the introduction of electrical currents to the
known whether other anatomical regions might be different. body to gain therapeutic effects. These modalities utilize various
Therapeutic ultrasound is another modality commonly used in types and frequencies of current, dependent on the desired effect.
people to heat tissue (3–5 cm depth) and increase fluid motion In the human, transcutaneous electrical nerve stimulation (TENS)
around cell membranes, but evidence of a therapeutic effect is still has been well documented to provide temporary pain relief by
lacking (Paulekas & Haussler, 2009). central release of endogenous endorphins and opioids, and/or by

406
 Rehabilitation techniques: manual therapy

stimulation of inhibitory inter-neurons at the spinal level (Watson, body, but mostly they appear on stable, anatomically described
2008). Even though there is no evidence that TENS is effective in points (Janssens, 1992). There is a correlation between the location
horses, it is the authors’ experience that it is a well-tolerated and clini- of trigger points and the location of acupuncture points, which were
cally useful modality to complement other treatment strategies. discovered independently in Eastern and Western medicine, respec-
Neuromuscular electrical stimulation (NMES) is a useful electro- tively. Pressure applied to a trigger point can produce pain in the
modality in cases of muscle atrophy due to disuse, immobility and/ trigger point itself, or referred pain in myofascial or visceral
or nerve damage in people. When a patient is unable to voluntarily structures.
contract the muscle(s), NMES can be applied to generate 80–90% It is generally perceived that by applying pressure to a painful
of the maximal voluntary contraction to assist in maintaining neu- point on the body, the pain decreases. Physical therapists use this
romuscular control and muscle development (Paulekas & Haussler, information during therapy (Simons et al., 1999). Instead of apply-
2009). In denervated muscle the findings are inconsistent though ing pressure, acupuncturists insert needles into these points
there is some evidence of a delay in denervation atrophy (Watson, (Melzack et al., 1977). Acupoints are located at specific locations of
2008). To date there are no equine studies, but it is the authors’ the body where the skin has a localized decrease in electrical resis-
opinion that, if the horse will tolerate it, NMES is a useful adjunct tance. Needle insertion influences the energy in the meridian and
to assess and treat horses with muscle dysfunction and/or atrophy. in the corresponding organ. The theory of acupuncture therapy is
Cases in which NEMS may be useful include suprascapular and to restore the balance of energy (Qi) in the body (van der Molen,
radial nerve lesions. 1999; Xie et al., 2005). Besides the specific acupoints described in
Anecdotal reports suggest beneficial effects from a variety of treat- manuals and atlases, every point on the body can be used as an
ment modalities (e.g. Bromiley, 1999; Porter, 2005). However, acupoint. Points that are painful on palpation, so called Ah Shi
objective evidence to support their use is not available. Extrapola- points, can be stimulated by any form of acupuncture. Insertion of
tion from the human field is complicated by the fact that there are an acupuncture needle causes micro trauma, which produces a local
differences in skin resistance, pigmentation, energy transmission inflammatory reaction (Ridgway, 2005) and always has a degree of
and absorption between equine tissue and the tissues of people or afferent sensory stimulation (Skarda et al., 2002). The use of a
research animals that the modalities have been tested on. Conse- twitch in horses is based on the same principle as acupuncture;
quently, therapeutic regimes are recommended by the manufacturer stimulation of the receptors in the skin of the nose produces anal-
rather than being based on scientific evidence. gesia (Lagerweij et al. 1984; Macgregor and von Schweinitz, 2006).
The effect of the twitch and of acupuncture can be blocked by
naloxon, a specific opioid antagonist (Xie et al., 2005), which indi-
Acupuncture cates that opioids play a role in the pain transmission (Veeneklaas,
Acupuncture is a part of the traditional Chinese medicine (TCM) 1999; Han, 2004). A significant increase of β-endorphins in plasma
dating back to 1500 BC. In addition to acupuncture, TCM also uses and cerebrospinal fluid has been found following acupuncture
herbal medicine, massage, breathing exercises and others. The word treatment in horses (Xie et al., 2001). Electro-acupuncture produces
acupuncture is formed by the combination of the Latin words acus, a greater β-endorphin release in the cerebrospinal fluid than
‘needle’, and pungere, ‘prick’. Typically, acupuncture involves insert- needle acupuncture (Skarda et al., 2002). Besides the β-endorphin
ing needles at specific acupuncture points (acupoints), although the release, acupuncture is also thought to act via the gate control
points may be stimulated by other methods including electro- theory in which pain in the central nervous system is inhibited by
acupuncture, laser acupuncture and acupressure (massaging of acu- stimulation of sensory receptors (by needle insertion) in the same
puncture points). Since the formation of the International Veterinary innervation area (Veeneklaas, 1999; Ammendolia et al., 2008).
Acupuncture Society (IVAS) in 1974, acupuncture has become more In human research there is a correlation between activation and
popular as a therapy in animals. There are some indications of a deactivation of specific areas in the brain by stimulation of acu-
positive clinical effect (Xie et al., 2001; Martin & Klide, 1987), but puncture points compared to stimulation of sham points (points
literature reviews indicate insufficient evidence to accept or reject that are not described as being acupuncture points, but which are
the effectiveness of acupuncture for certain problems in both veteri- located nearby the used acupuncture points and are innervated by
nary and human medicine (Cherkin et al., 2003; Habacher et al., the same spinal segment). Some brain areas are activated and deac-
2006; Johnston et al., 2008). Prospective, randomized, double tivated by both acupuncture points and sham points (Yan et al.,
blind, clinical trials are necessary to determine whether acupuncture 2005; Hsieh et al., 2001). Stimulation by acupuncture needles at
has more than a placebo effect (Sluijs, 2000). This section will the vertebral level of the pathological process gives better analgesia
highlight objective evidence for the use of acupuncture in the treat- in the short term than stimulation distal to the pathological process
ment of horses suffering from chronic back pain (Ridgway, 1999; (Skarda et al., 2002). In research on the effect of acupuncture
Martin et al., 1987; Xie et al., 2005). in people with back pain, a positive effect has been described.
The philosophy is based on the Yin-Yang principle, which explains The largest positive effect was found when acupuncture was com-
that everything in nature is constantly subject to change. Yin and pared to a negative control group or when acupuncture was used as
Yang are each other’s opposites but on the other hand they comple- a complementary therapy. There are conflicting results from studies
ment each other. Diseases can also be divided into Yin or Yang. Yin that compare acupuncture with needling of sham points with
diseases are chronic conditions in which the pain decreases by insufficient evidence to prove that the effect of acupuncture is
applying warmth and movement. Yang diseases are inflammations, greater than the effect of needling sham points (Ammendolia
which usually show redness, swelling, heat and a sharp pain. Yang et al., 2008).
pain decreases by applying rest and cold. Yin and Yang are parts of Acupuncture points in horses originate from two sources. One
the life energy, called Qi or Chi, which is present everywhere, but system used human acupuncture points and ‘translated’ them to the
researchers have not been able to quantify or measure Qi. Qi circu- equine model, using corresponding anatomical landmarks. Western
lates through 12 channels, called meridians, each of which is con- acupuncturists generally use the transpositional atlas in equine acu-
nected to a certain organ or organ system. Most of the acupuncture puncture (Fleming, 2001). Two articles are found with relevant data
points are located on the meridians (Van der Molen, 1999). on the effect of acupuncture in horses with chronic back pain. Xie
Trigger points are used in Western medicine for making a diag- et al. (2005) used fifteen horses with TL pain that were randomly
nosis and for treatment of pathological pain (Simons et al., 1999). assigned to one of the three treatment groups. The pain-relieving
They can be recognized as painful, hard nodular structures within effect of acupuncture in horses with back pain was compared with
a muscle or fascia. Sometimes they can be localized subcutaneously phenylbutazone treatment and with a control group in which the
or in the periosteum. Trigger points can occur anywhere on the horses received 0.9 % NaCl per os. Group 1 (four horses) received

407
 17 Rehabilitation of the locomotor apparatus

electroacupuncture every 3 days for five treatments. Group 2 (seven 3) analysis of weight distribution using a force plate. Though the
horses) received phenylbutazone 2.2 mg/kg, PO, q 12 h for five majority of horses improved (60% with chronic laminitis, 70% with
days. Group 3 (four horses) was the control group which received navicular disease) there were no statistically significant differences
0.9% NaCl 20 ml, PO, q 12h for 5 days. The horse owners and between treatment and control groups. Skarda et al. (2002) found
trainers were not informed as to what treatment their horse had that electroacupuncture was more effective than acupuncture in
received. TL pain score (TPS) was evaluated before, on days 1, 4, 7, stimulating the spinal cord to release β-endorphins into the CSF of
10 and 13 during treatment and on days 7 and 14 after the last horses. Both methods provided cutaneous analgesia in horses, mea-
electroacupuncture treatment. The mean score of 2 TPS investiga- sured by use of skin twitch reflex latency without adverse cardiovas-
tions was used because of the subjective nature of behavior with cular and respiratory effects.
one TPS score being evaluated from videotape of the examinations. A considerable body of literature on acupuncture in human and
After a washout period of 4 weeks, the measurements were repeated, veterinary alternative medicine is available, but there is a lack of
but for unexplained reasons group 3 (control) was not included good-quality study designs. Patient groups are often relatively small,
and horses in this group were reallocated to one of the other groups. negative controls are not included, and blinding is difficult. Articles
The results showed that after two treatments there was no significant have described some of the effects of acupuncture, like endorphin
difference in pain scores between the three groups. After treatment release or improvement of performance. Controversial outcomes
3 until 14 days after the final treatment, the group treated with are given in articles comparing acupuncture with sham treatment.
acupuncture had a significantly decreased pain score compared to It is unclear if specific acupuncture points are of greater value
the other groups. Between the NaCl and phenylbutazone group than a specific needling of the painful area. Further studies with
there was no significant difference. The article concluded that the larger sample sizes and relevant control groups are necessary to
decrease in TPS scores for the acupuncture group occurred after obtain scientific evidence to accept or reject the effectiveness of
three treatments and could last at least 14 days after the fifth treat- acupuncture.
ment. Deficiencies in the study include the lack of a negative control
group during the second phase and the fact that treatment duration
in the three groups was different. Phenylbutazone and NaCl were
administered for 5 days, but acupuncture therapy was continued for Conclusions
15 days (once every 3 days for five treatments). Furthermore, it is
not clear if the person investigating the TPS scores was blinded to The use of rehabilitation strategies and PT in equine veterinary
the given treatment. medicine is rapidly expanding; to date there is a relatively small
Martin and Klide (1987) used 15 horses that were referred to a body of scientific evidence of their applications and benefits.
university clinic by other veterinarians. All horses showed signs of However, the existing data are extremely promising, highlighting
chronic back pain, which existed for 2–24 months and had not the beneficial effects of complementary and alternative medicine,
improved with previous therapies. Hind limb lameness, poor saddle particularly from a manual therapy and motor control approach as
fit and poor riding technique were ruled out as causes of the back an adjunct to traditional medical management in the prevention,
pain. All horses were treated by injecting 1 mL of saline at the same treatment, and rehabilitation of sport horses. Equine researchers
nine acupoints once a week for 7–12 weeks with an average of nine should evaluate the human literature to develop research strategies,
treatments. Signs of back pain (clinical signs, behavioral signs, per- paying particular attention to therapeutic research that has used
formance changes) were evaluated by the investigators, rider/trainer quadrupeds as an animal model for human interventions. In clini-
and referring veterinarian on three occasions: before, during and cal practices that are widely used including acupuncture, electromo-
after the treatment period. Response to treatment was classified dalities and biophysical agents including therapeutic ultrasound,
as alleviation of signs or no change. The horses were classified as magnetic field therapy, laser therapy, hydrotherapy (swimming
having their signs alleviated only if all three evaluators agreed that pools, spas, underwater treadmills) and thermotherapy. These
the horse’s performance was normal, there were not clinical signs modalities are reported in the lay literature to have beneficial effects
of back pain and the owner/rider thought the horse could perform (Bromiley, 1999; Porter, 2005) but objective evidence and scientific
normally. Signs were alleviated in 13/15 horses. Six to twelve indications for their use are not yet clear in horses. Moreover, many
months after the final treatment, 11/13 horses that returned to of these treatment techniques are still the subject of scientific
normal performance were still competing without signs of back dispute in the human literature. TENS and NMES, although not
pain and the other two were no longer competing for reasons unre- proven effective in the horse, are supported by many studies in
lated to back pain. The nomenclature of the acupoints used in this relation to pain modulation and muscle tissue stimulation. Increas-
study is different from the traditional and transpositional points, ing evidence is appearing in the equine scientific literature to
which are commonly used worldwide. No explanation is given on support the implementation of manual therapy and exercise-based
why different points were used. No control group was included in therapies derived from scientific motor control principles. These
the study, which makes it impossible to rule out a placebo effect. include intervertebral joint mobilizations/manipulations, dynamic
The exact moment of evaluation of back pain signs is not made mobilization exercises and proprioceptive sensory integration
clear in the report and the method of classification of the horses techniques.
is not described. Because the classification of the horses was not Although research and documented clinical experiences are still
performed objectively, the outcome has a high risk of being necessary to complement the adaptation of scientific-based princi-
subjective. ples presented in this chapter, it is the authors’ opinion and clinical
Other studies have shown further systemic effects. Cheng et al. experience that rehabilitation/PT strategies can be successfully
(1980) found that electro-acupuncture elevated blood cortisol incorporated in conjunction with traditional veterinary medicine.
levels in naive horses whereas sham treatment did not. Bossut et al. A team approach with the necessary clinical reasoning skills is vital
(1984) noted production of cutaneous analgesia by electroacupunc- to choose the appropriate rehabilitation strategy, PT technique and
ture in horses, with variations in response due to sex of subject and objective monitoring of the patient’s progress and protocol. Con-
locus of stimulation. Steiss et al. (1989) used electroacupuncture to stant reassessment and, where possible, the use of objective mea-
treat lameness in horses and ponies with chronic laminitis (n = 10) sures to validate the outcome of the interventions are indicated for
or navicular disease (n = 10) and assessed the degree of lameness successful treatment strategies involving PT and rehabilitative
by: 1) a grading scheme, 2) measurement of stride lengths and techniques.

408
 References

References
AGRIA, 1995. 90–94-Femårsskrift AGRIA, Beck, C., Middleton, D., Maclean, A., Lavelle, Chan, S.P., Hong, Y., Robinson, P.D., 2001.
Stockholm (Insurance Statistics 1990– R., 2002. Osteochondrosis of the second Flexibility and passive resistance of the
1994). Cited in Wallin, L., Strandberg. E., cervical vertebra of a horse. Equine Vet. J. hamstrings of young adults using two
Philipson, J. and Dalin, G. (2000) 34 (2), 210–212. different static stretching protocols. Scand.
Estimates of longevity and causes of culling Berg, L.C., Nielsen, J.V., Thoefner, M.B., J. Med. Sci. Sports. 11, 81–86.
and death in Swedish warmblood and Thomsen, P.D., 2003. Ultrasonography of Chan, W.-W., Chen, K.-Y., Liu, H., Wu, L.-S.,
coldblood horses. Livest. Prod. Sci. 63 (3), the equine cervical region: a descriptive Lin, J.-H., 2001. Acupuncture for general
275–289. study in eight horses. Equine Vet. J. 35, veterinary practice. J. Vet. Med. Sci. 63,
American Academy of Orthopaedic Manual 647–655. 1057–1062.
Physical Therapists. http://www.aaompt. Blaney, F., English, C.A., Sawyer, T., 1999. Cheng, R., McKibbin, L., Roy, B., Pomeranz, B.,
org/. Sonographic measurement of 1980. Electroacupuncture elevates blood
Akeson, W., Arniet, D., Abel, M., Garfin, S., diaphragmatic displacement during tidal cortisol levels in naive horses; sham
Woo, L.-Y., 1987. Effects of immobilization breathing manoeuvres – a reliability study. treatment has no effect. Int. J. Neurosc. 10,
on joints. Clin. Orthop. 219, 28–37. Aust. J. Physiother. 45 (1), 1–80. 95–97.
Ammendolia, C., Furlan, A.D., Imamura, M., Bossut, D.F., Page, E.H., Stromberg, M.W., Cherkin, D.C., Sherman, K.J., Deyo, R.A.,
Irvin, E., van Tulder, M., 2008. Evidence- 1984. Production of cutaneous analgesia Shekelle, P.G., 2003. A review of the
informed management of chronic low back by electroacupuncture in horses: variations evidence for the effectiveness, safety, and
pain with needle acupuncture. Spine J. 8, dependent on sex of subject and locus of cost of acupuncture, massage therapy, and
160–172. stimulation. Am. J. Vet. Res. 45, 620–625. spinal manipulation for back pain. Ann.
Antonaci, F., Sand, T., Lucas, G.A., 1998. Bourke, J.M., 1995. Wastage in Thoroughbreds. Intern. Med. 138, 898–906.
Pressure algometry in healthy subjects: Animal Seminar, Equine Branch, New Claridge, H.A.H., Piercy, R.J., Parry, A., Weller,
inter-examiner variability. Scand. J. Rehab. Zealand Veterinary Association. Foundation R., 2010. The 3-D anatomy of the cervical
Med. 30, 3–8. for Continuing Education 167, 17–119. articular process joints in the horse and
Arendt-Nielsen, L., Graven-Nielsen, T., Citing American Association of Equine their topographical relationship to the
Svarrer, H., Svensson, P., 1995. The Practitioners Report (1994). spinal cord. Equine Vet. J., doi:
influence of low back pain on muscle Boyling, J.D., Grieve, G.P., Jull., G.A., 2004. 10.1111/j.2042–3306.2010.00114.x.
activity and coordination during gait: a Grieve’s modern manual therapy: the Clausen, M., Preisinger, R., Kalm, E., 1990.
clinical and experimental study. Pain 64, vertebral column, 3rd edn. Elsevier Health Analyse von krankenheitsdaten in der
231–240. Sciences, Edinburgh. deutschen warmblutzucht. Züchtungskunde
Arthur, R.M., Ross, M.W., Moloney, P.J., 62, 167–178.
Bromiley, M., 2007. Equine injury, therapy and
Cheney, W.W., 2003. North American rehabilitation. third ed. Blackwell Clayton, H.M., 1991. Conditioning Sport
Thoroughbred. In: Ross, M.W., Dyson, S.J. Publishing. Chichester, West Sussex, UK. Horses. Sport Horse Publications, Mason,
(Eds.), Diagnosis and management of MI.
Bromiley, M.W., 1999. Physical therapy for the
lameness in the horse. W.B. Saunders, equine back. Vet. Clin. N. Am.: Equine Clayton, H.M., Bialski, D.E., Lanovaz, J.L.,
Philadelphia, pp. 868–879. Pract. 15, 223–246. Mullineaux, D.R., 2003. Assessment of the
Ashley, F.H., Waterman-Pearson, A.E., Whay, reliability of a technique to measure
Buchner, H.H., Schildboeck, U., 2006.
H.R., 2005. Behavioural assessment of pain postural sway in horses. Am. J. Vet. Res. 64
Physiotherapy applied to the horse: a
in horses and donkeys: application to (11), 1354–1359.
review. Equine Vet. J. 38, 574–580.
clinical practice and future studies. Equine Clayton, H.M., Kaiser, L.J., Lavagnino, M.,
Vet. J. 37, 565–575. Bullwein, A., Hanichen, T., 1989. Age-related Stubbs, N.C., 2010a. Dynamic
changes in the intervertebral discs of the mobilizations in cervical flexion: effects on
Bailey, C.J., Reid, S.W.J., Hodgson, D.R.,
cervical vertebral column in the horse. intervertebral angulations. Equine Vet. J.
Suann, C.J., Rose, R.J., 1997. Risk factors
Tierartzliche Praxis 17, 73–76. (suppl) 38, 688–694.
associated with musculoskeletal injuries in
Australian Thoroughbred racehorses. Prev. Bunce, S.M., Moore, A.P., Hough, A.D., 2002. Clayton, H.M., Kaiser, L.J., Lavagnino, M.,
Vet. Med 32 (1–2), 47–55. M-mode ultrasound: a reliable measure of Stubbs, N.C., 2012. Evaluation of
transversus abdominis thickness? Clin. intersegmental vertebral motion during
Baker, J.R., Ellis, C.E., 1981. A survey of post
Biomech. 17, 315–317. performance of dynamic mobilization
mortem findings in 480 horses 1958 to
1980: (1) Causes of death. Equine Vet. J. Burns, E.M., Enns, R.M., Garrick, D.J., 2006. exercises in cervical lateral bending in
13, 43–46. The effect of simulated censored data on horses. Am. J. Vet. Res. 73, 1153–1159.
estimates of heritability of longevity in the Clayton, H.M., Lavagnino, M., Kaiser, L.J.,
Baker, V., Bennell, K., Stillman, B., Cowan, S.,
Thoroughbred racing industry. Genet. Mol. Stubbs, N.C., 2011a. Swing phase kinematic
Crossley, K., 2002. Abnormal knee joint
Res. 5 (1), 7–15. and kinetic response to weighting the hind
position sense in individuals with
patellofemoral pain syndrome. J. Orthop. Bussières, G., Jacques, C., Lainay, O., pasterns. Equine Vet. J. 43, 210–215.
Res. 20, 208–214. Beauchamp, G., Leblond, A., Cadoré, J.L., doi:10.1111/j.2042–3306.2010.00139.x.
et al., 2008. Development of a composite Clayton, H.M., Lavagnino, M., Kaiser, L.J.,
Barker, K.L., Shamley, D.R., Jackson, D., 2004.
orthopaedic pain scale in horses. Res. Vet. Stubbs, N.C., 2011b. Hind limb flexion
Changes in the cross-sectional area of
Sci. 85 (2), 294–306. response to different types of tactile
multifidus and psoas in patients with
unilateral back pain: the relationship to Cardinet, G.H., Holliday, T.A., 1979. devices. Am. J. Vet. Res. In press.
pain and disability. Spine 29 (22), Neuromuscular diseases of domestic Clayton, H.M., Townsend, H.G.G., 1989a.
e515–e519. animals: a summary of muscle biopsies Kinematics of the cervical spine of the
Barr, E.D., Pinchbeck, P.D., Clegg, P.D., Boyde, from 159 cases. Ann. NY Acad. Sci. 317, adult horse. Equine Vet. J. 21, 189–192.
A., Riggs, C.M., 2009. Post mortem 290–313. Clayton, H.M., Townsend, H.G.G., 1989b.
evaluation of palmar osteochondral disease Casey, R.A., 2002. Pain. In: Waran, N. (Ed.), Cervical spinal kinematics: a comparison
(traumatic osteochondrosis) of the The Welfare of Horses. Kluwer Academic between foals and adult horses. Equine Vet.
metacarpo/metatarsophalangeal joint in Publishers, Dordrecht, pp. 25–30. J. 21 (3), 193–195.
Thoroughbred racehorses. Equine Vet. J. 41, Cauvin, E., 1997. Assessment of back pain in Clayton, H.M., White, A.D., Kaiser, L.J.,
366–371. horses. In Practice 19, 522–533. Nauwelaerts, S., Lavagnino, M., Stubbs,

409
 17 Rehabilitation of the locomotor apparatus

N.C., 2008. Short term habituation of warmblood horses. J. Equine Vet. Sci. 30, between costly veterinary-care events and
equine limb kinematics to tactile 310–321. 5-year survival of Swedish insured
stimulation of the coronet. Vet. Comp. De Lahunta, A., 1983. The upper motor warmblooded riding horses. Prev. Vet. Med.
Orthop. Traumatol. 21, 211–214. neuron system In: Veterinary neuroanatomy 77, 122–136.
Clayton, H.M., White, A.D., Kaiser, L.J., and clinical neurology, 2nd edn. W.B. Egenvall, A., Lönnell, C., Roepstorff, L., 2009.
Nauwelaeerts, S., Lavagnino, M., Stubbs, Saunders Company, Philadelphia, pp. Analysis of morbidity and mortality data in
N.C., 2010b. Hind limb response to tactile 130–155. riding school horses, with special regard to
stimulation of the pastern and coronet. Degueurce, C., Chateau, H., Denoix, J.M., locomotor problems. Prev. Vet. Med. 88,
Equine Vet. J. 42 (suppl. 38), 227–233. 2004. In vitro assessment of movements of 193–204.
Colborne, G.R., McDonald, C., Barrett, A., the sacroiliac joint in the horse. Equine Vet. Ely, E.R., Avella, C.S., Price, J.S., Smith, R.K.W.,
Targett, S., Fuller, C.J., 2004. Effects of J. 36 (8), 694–698. Wood, J.L.N., Verheyen, K.L.P., 2009.
reflex inhibition technique for muscular Delaney, S., Worsley, P., Warner, M., Taylor, M., Descriptive epidemiology of fracture,
back pain. Equine Comp. Exerc. Physiol. 1 Stokes., M., 2010. Assessing contractile tendon and suspensory ligament injury in
(2), A8. ability of the quadriceps muscle using National Hunt racehorses in training.
Colbourn, C.M., Yovich, J.V., 1994. Suspensory ultrasound imaging. Muscle Nerve. 42 (4), Equine Vet. J. 41, 372–378.
ligament injuries in racing horses: 530–538. Erichsen, C., Eksell, R., Holm, K.R., Lord, R.,
ultrasonographic diagnosis and long term Denoix, J.M., 1999. Ultrasonographic Johnston, C., 2004. Relationship between
follow up. Aust. Equine Vet. J. 12, 119–128. evaluation of back lesions. Vet. Clin. N. scintigraphic and radiographic evaluations
Colloca, C.J., Cunliffe, C., Pinnock, M.H., Kim, Am.: Equine Pract. 15 (1), 131–159. of spinous processes in the thoracolumbar
Y.K., Hinrichs, R.N., 2009. Force-time Review. spine in riding horses without clinical signs
characteristisation of the mctimoney Denoix, J.M., 2005. Review of diagnosis and of back problems. Equine Vet. J. 36,
toggle-torque-recoil technique. J. Manip. treatment of lumbosacral pain in sport and 458–465.
Physiol. Th. 32 (5), 372–376. racehorses. Proc. Am. Ass. Equine Practnrs. Ertola, K., Houttu, J., 2003. The Finnish horse
51, 366–373. and other Scandinavian cold-blooded
Cottriall, S., Ritruechai, P., Wakeling, J.M.,
trotters. In: Ross, M.W., Dyson, S.J. (Eds.),
2008. The effects of training aids on the Denoix, J.M., 2007. Discovertebral pathology
Diagnosis and management of lameness in
longissimus dorsi in the equine back. in horses. Equine Vet. Educ. 19, 72–73.
the horse. W.B. Saunders, Philadelphia,
Comp. Exerc. Physiol. 5, 111–114. Denoix, J.M., Dyson, S., 2003. Thoracolumbar pp. 946–950.
Cousty, M., Retureau, C., Tricaud, C., Geffroy, spine. In: Ross, M.W., Dyson, S.J. (Eds.),
Faber, M., Schamhardt, H., van Weeren, R.,
O., Caure, S., 2010. Location of radiological Diagnosis and management of lameness in
Barneveld, A., 2001. Methodology and
lesions of the thoracolumbar column in the horse. W.B. Saunders, Philadelphia, pp.
validity of assessing kinematics of the
French trotters with and without signs of 509–521.
thoracolumbar vertebral column in horses
back pain. Vet. Rec. 166, 41–45. Denoix, J.M., Pailloux, J.P., 2005. Physical on the basis of skin-fixated markers. Am. J.
Cowan, S.M., Bennell, K.L., Hodges, P.W., therapy and massage for the horse, second Vet. Res. 62, 301–306.
Crossley, K.M., McConnell, J., 2001. ed. Manson Publishing Ltd, London.
Faber, M., Schamhardt, H., van Weeren, R.,
Delayed onset of electromyographic activity DiGiovanna, E.L., Schiowitz, S., Dowling, D.J., Johnston, C., Roepstorff, L., Barneveld, A.,
of vastus medialis obliquus relative to 2005. An osteopathic approach to 2000. Basic three-dimensional kinematics
vastus lateralis in subjects with diagnosis and treatment, 3rd edn. of the vertebral column of horses walking
patellofemoral pain syndrome. Arch. Phys. Lippincott Williams and Wilkins, on a treadmill. Am. J. Vet. Res. 61,
Med. Rehabil. 82, 183–189. Philadelphia. 399–406.
Crossley, K.M., Cowan, S.M., Bennell, K.L., Down, S.S., Henson, F.M.D., 2009. Faber, M.J., Schamhardt, H.C., van Weeren,
McConnell, J., 2004. Knee flexion during Radiographic retrospective study of the P.R., 1999. Determination of 3D spinal
stair ambulation is altered in individuals caudal cervical facet articulations in the kinematics without defining a local
with patellofemoral pain. J. Orthop. Res. horse. Equine Vet. J. 41 (1), 01–07. vertebral co-ordinate system. J. Biomech.
22, 267–274. Downer, A.H., 1978. Physical therapy for 32, 1355–1358.
D’Angelis, F., Mota, M., Freitas, E., Ferraz, G., animals: selected techniques. Charles C. Faber, M.J., van Weeren, P.R., Schepers, M.,
Abrahão, A., Lacerda-Neto, J., et al., 2007. Thomas Publishers, Springfield, Il. Barneveld, A., 2003. Long-term follow-up
Aerobic training, but not creatine, modifies Driver, A., Pilsworth, R., 2009. Traumatic of manipulative treatment in a horse with
longissimus dorsi muscle composition. J. damage to the back and pelvis. In: Henson, back problems. J. Vet. Med. 50, 241–245.
Equine Vet. Sci. 27 (3), 118–122. F. (Ed.), Back pathology: diagnosis and Fanselow, M.S., Calcagnetti, D.J., Helmstetter,
Dabareiner, R.M., Cohen, N.D., Carter, G.K., treatment. Blackwell Publishing Ltd., F.J., 1989. Role of mu and kappa opioid
Nunn, S., Moyer, M., 2005a. Oxford, pp. 135–146. receptors in conditional fear-induced
Musculoskeletal problems associated with Dyson, S., Murray, R., 2003. Pain associated analgesia: the antagonistic actions of nor
lameness and poor performance among with the sacroiliac joint region: a clinical binaltorphimine and the cyclic
horses used for barrel racing: 118 cases study of 74 horses. Equine Vet. J. 35, somatostatin octapeptide, Cys2Tyr3Pen7-
(2000–2003). J. Am. Vet. Med. Assoc. 227 240–245. amide. J. Pharmac. Exp. Therap. 250,
(10), 1646–1650. Dyson, S.J., 2003. The cervical spine and soft 825–830.
Dabareiner, R.M., Cohen, N.D., Carter, G.K., tissues of the neck. In: Ross, M.W. and Fédération Equestre Internationale, 2011.
Nunn, S., Moyer, M., 2005b. Lameness and Dyson, S.J. (Eds.), diagnosis and Veterinary regulations, twelfth ed. http://
poor performance in horses used for team management of lameness in the horse. www.fei.org/veterinary/doping-and-
roping: 118 cases (2000–2003). J. Am. Vet. Saunders, Philadelphia, pp. 522–531. controlled-medication.Switzerland
Med. Assoc. 226, 1694–1699. Egenvall, A., Bonnett, B., Wattle, O., Copyright © 2010 Fédération Equestre
Dalin, G., Jeffcott, L.B., 1986. Sacroiliac joint Emanuelson, U., 2008. Veterinary-care Internationale
of the horse. 2. Morphometric features. events and costs over a 5-year follow-up Ferreira, P.H., Ferreira, M.L., Hodges, P.W.,
Anat. Histol. Embryol.15 (2), 97–107. period for warmblooded riding horses with 2004. Changes in recruitment of the
De Heus, P., Van Oossanen, G., Van or without previously recorded locomotor abdominal muscles in people with low
Dierendonck, M.C., Back, W., 2010. A problems in Sweden. Prev. Vet. Med. 83, back pain: ultrasound measurement of
pressure algometer is a useful tool to 130–143. muscle activity. Spine 29 (22), 2560–2566.
objectively monitor the effect of diagnostic Egenvall, A., Bonnett, B.N., Olson, P., Penell, Fischer, A.A., 1987. Pressure algometry over
palpation by a physiotherapist in J., Emanuelson, U., 2006. Association normal muscles. Standard values, validity

410
 References

and reproducibility of pressure threshold. Gibson, K.T., Snyder, J.R., Spier, S.J., 2002. Back., W., van Weeren, P.R., 2007b. The
Pain 30, 115–126. Ultrasonographic diagnosis of soft tissue effect of induced forelimb lameness on
Fischer, A.A., 1998. Algometry in diagnosis of injuries in horses competing at the Sydney thoracolumbar kinematics during treadmill
musculoskeletal pain and evaluation of 2000 Olympic Games. Equine Vet. Ed. 14 locomotion. Equine Vet. J. 39 (3),
treatment outcome: an update. J. (3), 149–156. 197–201.
Musculoskeletal Pain 6, 5–32. Gillen, A., Dyson, S., Murray, R., 2009. Nuclear Gregory, N.G., 2004. Pain. In: Physiology and
Fleming, P., 2001. The location of equine back scintigraphic assessment of the behaviour of animal suffering. Blackwell
Shu points: Traditional Chinese versus thoracolumbar synovial intervertebral Science, Oxford, pp. 94–130.
transpositional. In: Schoen, A.M. (Ed.), articulations. Equine Vet. J. 41, 534–540. Groesel, M., Zsoldos, R.R., Kotschwar, A.,
Veterinary acupuncture, ancient art to Gillis, C.L., Meagher, D.M., Pool, R.R., Stover, Gfoehler, M.C., Peham, C., 2010. A
modern medicine. CV Mosby, St Louis, pp. S.M., Craychee, T.J., Willits, N., 1993. preliminary model study of the equine
393–431. Ultrasonographically detected changes in back including activity of longissimus dorsi
Folpp, H., Deall, S., Harvey, L.A., Gwinn, T., equine superficial digital flexor tendons muscle. Equine Vet. J. 42 (Suppl. 38),
2006. Can apparent increases in muscle during the first months of race training. 401–406.
extensibility with regular stretch be Am. J. Vet. Res. 54, 1797–1802. Gundel, M., Schatzmann, U., 1997.
explained by changes in tolerance to Girodroux, M., Dyson, S., Murray, R., 2009. Rückenprobleme beim Pferd: 3. Vorschlag
stretch? Aust. J. Physiother. 52, 45–50. Osteoarthritis of the thoracolumbar eines klinischen Untersuchungsprotokolls
Franssen, J.L.M., 1995. Registratie en synovial intervertebral articulations: zur Abklärung einer Rückenproblematik
bewerking van de oppervlakte-EMG. Clinical and radiographic features in 77 beim Reitpferd. Pferdeheilkunde 13,
In: Handboek oppervlakte- horses with poor performance and back 213–221.
elektromyografie. De Tijdstroom, pain. Equine Vet. J. 41, 130–138. Gundel, M., Schatzmann, U., Ueltschi, G.,
Utrecht, pp. 117–142. Göbel, H., Cordes, P., 1990. Circadian 1997. Rückenprobleme beim Pferd: Ein
Fraser, J.A., 1969. Some observations on the variation of pain sensitivity in pericranial Vergleich der bildgebenden Verfahren mit
behaviour of the horse in pain. British Vet. Musculature. Headache 30, 418–422. der klinischen Untersuchung.
J. 125, 150–151. Goff, L., Crook, T., 2007. Physiotherapy Pferdeheilkunde 13, 659–666.
Freestone, F.J., Carlson, G.R., 1991. Muscle assessment for animals. In: McGowan, C., Habacher, G., Pittler, M.H., Ernst, E., 2006.
disorders in the horse: a retrospective Goff, L., Stubbs, N. (Eds.), Animal Effectiveness of acupuncture in veterinary
study. Equine Vet. J. 23 (2), 86–90. physiotherapy: assessment, treatment and medicine: systematic review. J. Vet. Intern.
Frontera, W.R., 2007. Chapter 1. An overview rehabilitation of animals. Blackwell Med. 20, 480–488.
of sports medicine. Clinical sports Publishing, Ames IA, pp. 136–163. Hamra, J.G., Kamerling, S.G., Wolfsheimer,
medicine: medical management and Goff, L., Stubbs, N., 2007. Equine treatment K.J., Bagwell, C.A., 1993. Diurnal variation
rehabilitation. Elsevier, London, pp. 3–6. and rehabilitation. In: McGowan, C., Goff, in plasma ir-beta-endorphin levels and
Fruehwirth, B., Peham, C., Scheidl, M., L., Stubbs, N. (Eds.), Animal physiotherapy: experimental pain thresholds in the horse.
Schobesberger, H., 2004. Evaluation of assessment, treatment and rehabilitation of Life Sci. 53, 121–129.
pressure distribution under an English animals. Blackwell Publishing, Ames IA, Han, J.-S., 2004. Acupuncture and endorphins.
saddle at walk, trot and canter. Equine Vet. pp. 238–251. Neurosci. Lett. 361 (1-3), 258–261.
J. 36 (8), 754–757. Goff, L.M., 2009. Manual therapy for the Hansen, B.D., 2003. Assessment of pain in
Fuglbjerg, V., Nielson, J.V., Thomsen, P.D., Horse–a contemporary perspective. J. dogs: veterinary clinical studies. Inst. Lab.
Berg, L.C., 2010. Accuracy of ultrasound- Equine Vet. Sci. 29, 799–808. Anim. Res. J. 44, 197–205.
guided injections of thoracolumbar Goff, L.M., Jasiewicz, J., Jeffcott, L.B., Condie, Hansen, H.J., 1959. Comparative views of the
articular process joints in horses: a P., McGowan, T.W., McGowan, C.M., 2006. pathology of disc degeneration in animals.
cadaveric study. Equine Vet. J. 42, 18–22. Movement between the equine ilium and Lab. Invest. 8, 1242–1257.
Fuhr, A.W., Menke, J.M., 2005. Status of sacrum: in vivo and in vitro studies. Equine Harman, J., 1999. Tack and saddle fit. Vet.
activator methods chiropractic technique, Vet. J. 38 (Suppl. 36), 457–461. Clin. N. Am.: Equine Pract. 15, 247–261.
theory, and practice. J. Manipulative Goff, L.M., Jeffcott, L.B., Jasiewicz, J., Haussler, K.K., 1999a. Anatomy of the TL
Physiol. Ther. 28, e1–e20. McGowan, C.M., 2008. Structural and vertebral region. Vet. Clin. N. Am.: Equine
Gajdosik, R.L., 2001. Passive extensibility of biomechanical aspects of equine sacroiliac Pract. 15, 13–26.
skeletal muscle: review of the literature joint function and their relationship to Haussler, K.K., 1999b. Chiropractic evaluation
with clinical implications. Clin. Biomech. clinical disease. Vet. J. 176 (3), 281–293. and management. Vet. Clin. N. Am.:
(Bristol, Avon) 16, 87–101. Goff, L.M., Jull, G., 2007. Manual therapy In: Equine Pract. 15, 195–209.
Gatterman, M.I., Hansen, D.T., 1994. McGowan, C., Goff, L.,Stubbs, N. (Eds.), Haussler, K.K., 2000. Equine chiropractic:
‘Development of chiropractic nomenclature Animal physiotherapy: assessment, general principles and clinical applications.
through consensus’. J. Manipulative Physiol. treatment and rehabilitation of animals. Am. Assoc. Equine Pract. Proceedings 46,
Ther. 17 (5), 302–309. Blackwell Publishing, Ames IA, pp. 84–93.
Gellman, K., 1998. An integrated approach to 164–176. Haussler, K.K., 2003a. Diagnosis and
diagnosing and treating back pain in Gomez Alvarez, C.B., Bobbert, M.F., Lamers, L., management of sacroiliac joint injuries. In:
horses. Proc. Conf. Equine Sports Med. Sci. Johnston, C., Back, W., van Weeren, P.R., Ross, M.W., Dyson, S.J. (Eds.), Diagnosis
119. 2007a. The effect of induced hind limb and management of lameness in the horse.
Genovese, R., Longo, K., Berthold, B., lameness on the TL kinematics during Saunders, Philadelphia, pp. 803–811.
Jorgenson, J., 1997. Quantitative treadmill locomotion. Equine Vet. J. 40, Haussler, K.K., 2003b. Chiropractic evaluation
sonographic assessment in the clinical 147–152. and management of musculoskeletal
management of superficial digital flexor Gomez Alvarez, C.B., L’Ami, J.J., Moffatt, D., disorders. In: Ross, M.W., Dyson, S.J. (Eds.),
injuries in thoroughbred racehorses. Proc. Back, W., van Weeren, P.R., 2008. Effect of Diagnosis and management of lameness in
Am. Assoc. Equine Pract. 43, 187–206. chiropractic manipulations on the the horse. Saunders, Philadelphia, pp.
Geres, V., 1978. Radiological findings in kinematics of back and limbs in horses 501–508.
diseases of skeletal parts of vertebral with clinically diagnosed back problems. Haussler, K.K., 2006. Pressure algometry for
column, discs and spinal meninges in dog Equine Vet. J. 40, 153–159. the detection of induced back pain in
cat and horse. Zbl. Vet. Med., A 25, Gomez-Alvarez, C.B.G., Wennerstrand, J., horses: a preliminary study. Equine Vet. J.
363–382. Bobber, M.F., Lamers., L., Johnston., C., 37, 76–81.

411
 17 Rehabilitation of the locomotor apparatus

Haussler, K.K., 2009. Review of manual subsequent diagnosis of pelvic or hind muscle contraction with ultrasound
therapy techniques in equine practice. limb fracture in racing Thoroughbreds. imaging. Muscle Nerve 27 (6), 682–692.
J. Equine Vet. Sci. 29, 849–869. Equine Vet. J. 42, 234–239. Hodges, P.W., Richardson, C., 1999.
Haussler, K.K., Bertram, J.E.A., Gellman, K., Hides, J., Cooper, D., Stokes, M., 1992. Transversus abdominis and the superficial
1999a. In-vivo segmental kinematics of the Diagnostic ultrasound imaging for abdominal muscles are controlled
thoracolumbar spinal region in horses and measurement of the lumbar multifidus independently in postural tasks. Neurosci.
effects of chiropractic manipulations. Proc. muscle in normal young adults. Physiother. Lett. 265, 91–94.
Am. Assoc. Equine Pract. 45, 327–329. Theory Pract. 8, 19–26. Holm, S., Indahl, A., Solomonow, M., 2002.
Haussler, K.K., Bertram, J.E.A., Gellman, K., Hides, J., Gilmore, C., Stanton, W., Sensorimotor control of the spine. J.
Hermanson, J.W., 2001. Segmental in vivo Bohlscheida, E., 2008. Multifidus size and Electromyogr. Kinesiol. 12 (3), 219–234.
vertebral kinematics at the walk, trot and symmetry among chronic LBP and healthy Holton, L.L., Scott, E.M., Nolan, A.M., Reid, J.,
canter: a preliminary study. Equine Vet. J. asymptomatic subjects. Manual Ther. 13, Welsh, E., Flaherty, D., 1998. Comparison
33 (Suppl.), 160–164. 43–49. of three methods used for assessment of
Haussler, K.K., Erb, H.N., 2003. Pressure Hides, J., Richardson, C., Jull, G., 1995. pain in dogs. J. Am. Vet. Med. Assoc. 212,
algometry: objective assessment of back Magnetic resonance imaging and 61–66.
pain and effects of chiropractic treatment. ultrasonography of the lumbar multifidus Hopes, R., 1970. Physiotherapy in the horse.
In: 49th Annual Convention of the muscle. Spine 20 (1), 54–58. Physiotherapy 56, 56–59.
American Association of Equine Hides, J., Richardson, C., Jull, G., 1996.
Practitioners, 2003, New Orleans, Hsieh, J.-C., Tu, C.-H., Chen, F.-P., Chen,
Multifidus muscle recovery is not M.-C., Yey, T.-C., Cheng, H.-C., et al., 2001.
Louisiana. Internet Publisher, International automatic after resolution of acute,
Veterinary Information Service (www.ivis. Activation of the hypothalamus
first-episode low back pain. Spine 21 (23), characterizes the acupuncture stimulation
org), Ithaca, New York, USA. 2763–2769.
Haussler, K.K., Erb, H.N., 2006a. Mechanical at the analgesic point in human: a positron
Hides, J., Stokes, M., Saide, M., Jull, G., emission tomography study. Neuroscience
nociceptive thresholds in the axial skeleton
Cooper, D., 1994. Evidence of lumbar Letters 307, 105–108.
of horses. Equine Vet. J. 38, 70–75.
multifidus wasting ipsilateral to symptoms Hunt, L.M., Valberg, S.J., Steffenhagen, K.,
Haussler, K.K., Erb, H.N., 2006b. Pressure in patients with acute/subacute low back
algometry for the detection of induced McCue, M.E., 2008. An epidemiological
pain. Spine 19 (2), 165–172. study of myopathies in Warmblood horses.
back pain in horses: a preliminary study.
Hides, J., Wilson, S., Stanton, W., McMahon, Equine Vet. J. 40 (2), 171–177.
Equine Vet. J. 37, 76–81.
S., Keto, H., McMahon, K., et al., 2006. An Hurwitz, E., Aker, P., Adams, A., Meeker, W.,
Haussler, K.K., Hill, A.E., Puttlitz, C.M., MRI investigation into the function of the
McIlwraith, C.W., 2007. Effects of vertebral Shekelle, P., 1996. Manipulation and
transversus abdominis muscle during mobilization of the cervical spine: a
mobilisation and manipulation on “drawing-in” of the abdominal wall. Spine
kinematics of the TL region. Am. J. Vet. Res. systematic review of the literature. Spine 21,
31 (6), e175–e178. 1746–1759.
68, 508–516.
Hides, J.A., Jull, G.A., Richardson, C.A., 2001. Indahl, A., Kaigle, A.M., Reikeräs, O., Holm,
Haussler, K.K., Martin, C.E., Hill, A.E., 2010.
Long-term effects of specific stabilizing S.H., 1997. Interaction between the porcine
Efficacy of spinal manipulation and
exercises for first-episode low back pain. lumbar intervertebral disc, zygapophysial
mobilisation on trunk flexibility and
Spine 26 (11), e243–e248. joints, and paraspinal muscles. Spine 22
stiffness in horses: a randomised clinical
trial. Equine Vet. J. 42 (Suppl. 38), Hides, J.A., Richardson, C.A., Jull, G.A., 1998. (24), 2834–2840.
695–702. Use of real-time ultrasound imaging for
Ireland, M.L., Willson, J.D., Ballantyne, B.T.,
feedback in rehabilitation. Manual Ther. 3
Haussler, K.K., Stover, S.M., 1998. Stress Davis, I.M., 2003. Hip strength in females
(3), 125–131.
fractures of the vertebral lamina and pelvis with and without patellofemoral pain.
in Thoroughbred racehorses. Equine Vet. J. Hodges, P., Holm, A.K., Hansson., T., Holm, J. Orthop. Sports Phys. Ther. 33, 671–676.
30 (5), 374–381. S., 2006. Rapid atrophy of the lumbar
Irwin, D.H., Howell, D.W., 1980. Some
Haussler, K.K., Stover, S.M., Willits, N.H., multifidus follows experimental disc or
thoughts on swimming horses in a pool.
1999b. Pathologic changes in the nerve root injury. Spine 31 (25),
J. S. Afr. Vet. Assoc. 51, 189–191.
lumbosacral vertebrae and pelvis in 2926–2933.
Ishihara, A., Reed, S.M., Rajala-Schultz, P.J.,
Thoroughbred racehorses. Am. J. Vet. Res. Hodges, P., Kaigle Holm, A., Holm, S.,
Robertson, J.T., Bertone, A.L., 2009. Use of
60, 143–153. Ekström, L., Cresswell, A., Hansson, T.,
kinetic gait analysis for detection,
Heisele, C., 1995. Schadenfälle in der et al., 2003a. Intervertebral stiffness of the
quantification, and differentiation of hind
langfristigen pferdelebensversicherung spine is increased by evoked contraction of
limb lameness and spinal ataxia in horses.
unter berucksichtigung der tierärtlichen transversus abdominis and the diaphragm:
J. Am. Vet. Med. Assoc. 234 (5), 644–651.
mitwirkung in der in vivo porcine studies. Spine 28 (23),
2594–2601. Janssens, L.A.A., 1992. Trigger point therapy.
pferdelebensversicherung. Thesis, University Problems Vet. Med. 4, 117–124.
of München. München, Germany. Hodges, P.W., 2005. Ultrasound imaging in
rehabilitation: just a fad? J. Orthop. Sports Jeffcott, L.B., 1975. The diagnosis of diseases
Henson, F., 2009. Equine back pain: diagnosis
Phys. Ther. 35 (6), 333–337. of the horse’s back. Equine Vet. J. 7, 69–78.
and treatment. Wiley-Blackwell, Ames, IA.
Hodges, P.W., Cholewicki, J., 2007. Functional Jeffcott, L.B., 1979. Back problems in the
Henson, F., Lamas, L., Knezevic, S., Jeffcott,
control of the spine. In: Vleeming, A., horse- a look at past, present and future
L.B., 2007. Ultrasonographic evaluation of
Mooney, V., Stoeckart, R. (Eds.), Movement, progress. Equine Vet. J. 11, 129–136.
the supraspinous ligament in a series of
ridden and unridden horses and horses stability and lumbopelvic pain, 2nd edn. Jeffcott, L.B., 1980. Disorders of the
with unrelated back pathology. BMC Vet. Churchill Livingstone, Elsevier, Edinburgh, thoracolumbar spine of the horse-a survey
Res. 3 (3), 1746–6148. pp. 489–512. of 443 cases. Equine Vet. J. 12 (4),
Hernandez, J., Hawkins, D.L., Scollay, M.C., Hodges, P.W., Galea, M.P., Holm, S., Holm, 197–210.
2001. Race-start characteristics and risk of A.K., 2009. Corticomotor excitability of Jeffcott, L.B., 1999. Back problems. Historical
catastrophic musculoskeletal injury in back muscles is affected by intervertebral perspective and clinical indications. Vet.
Thoroughbred racehorses. J. Am. Vet. Med. disc lesion in pigs. Eur. J. Neurosci. 29 (7), Clin. N. Am.: Equine Pract. 15, 1–12.
Assoc. 218, 83–86. 1490–1500. Jeffcott, L.B., Dalin, G., 1980. Natural rigidity
Hesse, K.L., Verheyen, K.L., 2010. Associations Hodges, P.W., Pengel, L.H.M., Herbert, R.D., of the horse’s backbone. Equine Vet. J. 12,
between physiotherapy findings and Gandevia, S.C., 2003b. Measurement of 101–108.

412
 References

Jeffcott, L.B., Dalin, G., Ekman, S., Olsson, Kersten, A.A.M., Edinger, J., 2004. Lewis, R.D., 2003. The Racing Quarter Horse
S.E., 1985. Sacroiliac lesions as a cause of Ultrasonographic examination of the In: Ross, M.W., Dyson, S.J. (Eds.),
chronic poor performance in competitive equine sacroiliac region. Equine Vet. J. 36 Diagnosis and management of lameness in
horses. Equine Vet. J. 17 (2), 111–118. (7), 602–608. the horse. W.B. Saunders, Philadelphia, pp.
Jeffcott, L.B., Hickman, J., 1975. The treatment Kidd, A.W., Magee, S., Richardson, C.A., 2002. 927–930.
of horses with chronic back pain by Reliability of real-time ultrasound for the Licka, T.F., Peham, C., Frey, A., 2004.
resecting the summits of the impinging assessment of transversus abdominis Electromyographic activity of the
dorsal spinous processes. Equine Vet. J. 7, function. J. Gravit. Physiol. 9 (1), 131–132. longissimus dorsi muscles in horses during
115–119. Kiesel, K., Uhl, T., Underwood, F., Rodd, D., trotting on a treadmill. Am. J. Vet. Res. 65,
Jeffcott, L.B., Rossdale, P.D., Freestone, J., Nitz, A., 2006. Measurement of lumbar 155–158.
Frank, C.J., Towers-Clark, P.F., 1982. An multifidus muscle contraction with Liljebrink, Y., Bergh, A., 2010. Goniometry: is
assessment of wastage in Thoroughbred rehabilitative ultrasound imaging. Man. it a reliable tool to monitor passive joint
racing from conception to 4 years of age. Ther. 12 (2), 161–166. range of motion in horses? Equine Vet. J.
Equine Vet. J. 14, 185–198. Kingma, J.J., de Knikker, R., Wittink, H.M., 42 (Suppl.), 676–682.
Jeffcott, L.B., Whitwell, K.E., 1976. Fractures of Takken, T., 2007. Eccentric overload Loeb, G.E., 2005. Biomimetic posture sensing
the thora-columbar spine in the horse. training in patients with chronic Achilles and feedback for proprioception.
Proc. 23rd Ann. Conv. Am. Ass. Equine tendinopathy: a systematic review. Br. J. Proceedings 27th Annual International
Pract. 91–102. Sports Med. 41, e3. Conference on Engineering in Medicine
Jerosch, J., Prymka, M., 1996. Proprioception Knubben, J.M., Gygax, L., Auer, J., Fürst, A., and Biology 77389–77392.
and joint stability. Knee Surg., Sports Stauffacher, M., 2008. [Frequency of Love, E.J., Murrel, J., Whay, H.R., 2011.
Traumatol. Arthroscopy 4, 171–179. diseases and injuries in the Swiss horse Thermal and mechanical nociceptive
Johnston, B.C., da Costa, B.R., Devereaux, P.J., population]. Schweiz Arch Tierheilkd 150, threshold testing in horses: a review. Vet.
Akl, E.A., Busse, J.W., 2008. The use of 399–408. Anaesth. Analg. 38, 3–14.
expertise-based randomized controlled Kohn, C.W., Hinchcliff, K.W., McKeever, K.H., MacDonald, D., Moseley, G.L., Hodges, P.W.,
trials to assess spinal manipulation and 1999. Evaluation of washing with cold 2009. Why do some patients keep hurting
acupuncture for low back pain: a systematic water to facilitate heat dissipation in horses their back? Evidence of ongoing back
review. Spine 33, 914–918. exercised in hot, humid conditions. Am. J. muscle dysfunction during remission from
Johnston, C., Roethlisberger Holm, K., Vet. Res. 60, 299–305. recurrent back pain. Pain 142 (3),
Erichsen, C., Eksell, P., Drevemo, S., 2004. Koppenhaver, S., Hebert, J., Fritz, J., Parent, E., 183–188.
Kinematic evaluation of the back in fully Teyhen, D., Magel, J., 2009. Reliability of MacDonald, D., Moseley, G.L., Hodges, P.W.,
functioning riding horses. Equine Vet. J. 36, rehabilitative ultrasound imaging of the 2010. People with recurrent low back pain
495–498. transversus abdominis and lumbar respond differently to trunk loading despite
Jull, G.A., Richardson, C.A., 2000. Motor multifidus muscles. Arch. Phys. Med. remission from symptoms. Spine 35 (7),
control problems in patients with spinal Rehab. 90 (1), 87–94. 818–824.
pain: A new direction for therapeutic Labracque, G., Vanier, M.C., 2003. Rhythms, MacDonald, D.A., Moseley, G.L., Hodges, P.W.,
exercises. J. Manip. Physiol. Ther. 23 (2), pain and pain management. In: Redfern, P. 2006. The lumbar multifidus: Does the
115–117. (Ed.), Chronotherapies. Pharmaceutical evidence support clinical beliefs? Man.
Kaigle, A.M., Sten, M.S., Holm, H., Press, London, pp. 212–233. Ther. 11, 254–263.
Hansson, T.H., 1995. Experimental Lagerweij, E., Nelis, P.C., Wiegant, V.M., van Macgregor, J., von Schweinitz, D.G., 2006.
instability in the lumbar spine. Spine 20 Ree, J.M., 1984. The twitch in horses: a Needle electromyographic activity of
(4), 421–430. variant of acupuncture. Science 225, myofascial trigger points and control sites
Kalichman, L., Hodges, P., Li, L., Guermazi, A., 1172–1174. in equine cleidobrachialis muscle – an
Hunter, D.J., 2009. Changes in paraspinal Landman, M.A.A.M., de Blaauw, J.A., van observational study. Acupunct. Med. 24
muscles and their association with low Weeren, P.R., Hofland, L.J., 2004. Field (2), 61–70.
back pain and spinal degeneration: CT study of the prevalence of lameness in Magnusson, S.P., Simonsen, E.B., Aagaard, P.,
study. Eur. Spine J. 19 (7), 1136–1144. horses with back problems. Vet. Rec. 155, et al., 1996a. Mechanical and physical
Kamerling, S.G., Weckman, T.J., Dequick, D.J., 165–168. responses to stretching with and without
Tobin, T., 1985. A method for studying Latash, M.L., 2008. Neurophysiological basis preisometric contraction in human skeletal
cutaneous pain perception and analgesia in of movement, second ed. Human Kinetics, muscle. Arch. Phys. Med. Rehabil. 77,
horses. J. Pharmacol. Methods 13, Champaign, IL, USA, p. 17. 373–378.
267–274. Lee, D., 2004. The pelvic girdle, 3rd edn. Magnusson, S.P., Simonsen, E.B., Aagaard, P.,
Kaneps, A.J., 2000. Tissue temperature Elsevier Science, Edinburgh. et al., 1996b. A mechanism for altered
response to hot and cold therapy in the Lee, D., Vleeming, A., 2007. An intergrated flexibility in human skeletal muscle. J.
metacarpal region of a horse. Proc. Am. therapeutic approach to the treatment of Physiol. 497, 291–298.
Assoc. Equine Practnrs. 46, 208–213. pelvic girdle pain. In: Vleeming, A., Maitland, G., Hengeveld, E., Banks, K.,
Katalinic, O.M., Harvey, L.A., Herbert, R.D., Mooney, V., Stoeckart, R. (Eds.), Movement, English, K., 2000. Maitland’s vertebral
Moseley, A.M., Lannin, N.A., Schurr, K., stability and lumbopelvic pain, second ed. manipulation. 6th edn. Butterworth-
2009. Stretch interventions for contractures, Churchill Livingstone, Elsevier Science, Heinemann, Oxford, UK.
the cochrane collaboration. John Wiley and Edinburgh, pp. 622–638. Marks, D., 1999. Medical management of back
Sons, Oxford. Lee, J., Tseng, W., Shau, Y., Wang, C., Wang, pain. Vet. Clin. N. Am.: Equine Pract. 15,
Kato, T., Ohmura, H., Hiraga, A., Wada, S., H., Wang, S., 2007. Measurement of 179–194.
Kuwahara, M., Tsubone, H., 2003. Changes segmental cervical multifidus contraction Marr, C.M., Love, S., Boyd, J.S., McKellar, Q.,
in heart rate variability in horses during by ultrasonography in asymptomatic 1993. Factors affecting the clinical outcome
immersion in warm springwater. Am. J. Vet. adults. Man. Ther. 12 (3), 286–294. of injuries to the superficial digital flexor
Res. 64, 1482–1485. Levine, J.M., Adam, E., Mackay, J., et al., 2007. tendon in National Hunt and point-to-
Kelly, S.J., Stokes, M.J., 1993. Symmetry of Confirmed and presumptive cervical point racehorses. Vet. Rec. 132, 476–479.
anterior tibial muscle size measured by vertebral compressive myelopathy in older Martin, B.B., Klide, A., 1987. Use of
real- time ultrasound imaging in young horses: a retrospective study (1994–2004). acupuncture for the treatment of chronic
females. Clin. Rehabil. 7, 222–228. J. Vet. Intern. Med. 21, 812–819. back pain in horses: stimulation of

413
 17 Rehabilitation of the locomotor apparatus

acupuncture points with saline solution Mills, D., Millman, S., Levine, E., 2007. vastus lateralis muscles is disrupted during
injections. J. Am. Vet. Med. Assoc. 190, Applied animal behaviour: assessment, eccentric contractions in subjects with
1177–1180. pain and aggression. In: McGowan, C., patellofemoral pain. J. Sports Med. 30,
Martin, B.B., Klide, A., 1997. Diagnosis and Goff, L., Stubbs, N. (Eds.), Animal 483–487.
treatment of chronic back pain in horses. physiotherapy: assessment, treatment and Paulekas, R., Haussler, K.K., 2009. Principles
AAEP Proceedings 43, 310–311. rehabilitation of animals. Blackwell and practice of therapeutic exercise for
Martin Jr, B.B., Reef, V.B., Parente, E.J., Sage, Publishing, Oxford, pp. 3–14. horses. J. Equine Vet. Sci. 29, 870–893.
A.D., 2000. Causes of poor performance of Mitchell, J.B., Mitchell, J.S., Nolan, P.M., Ross, Peham, C., Frey, A., Licka, T.F., Scheidl, M.,
horses during training, racing, or showing: M.W., 2003. The North American 2001. Evaluation of the EMG activity of the
348 cases (1992–1996). J. Am. Vet. Med. Standardbred. In: Ross, M.W., Dyson, S.J. long back muscle during induced back
Assoc. 216, 554–558. (Eds.), Diagnosis and management of movements at stance. Equine Vet. J. 33
Martinoli, C., Derchi, L.E., Pastorin, C., lameness in the horse. W.B. Saunders, (Suppl.), 165–168.
Bertolotto, M., Silvestri, E., 1993. Analysis Philadelphia, pp. 895–912.
Peham, C., Licka, T., Schobesberger, H.,
of echotexture of tendons with US. Mitchell, W.M., 1933. Some aspects of the Meschan, E., 2004. Influence of the rider
Radiology 186, 839–843. general osteoarthritic problem in the horse. on the variability of the equine gait. Hum.
Vet. Rec. 13, 918–924. Mov. Sci. 23 (5), 663–671.
Martinson, H., Stokes, M., 1991. Measurement
of anterior tibial muscle size using Mohammed, H.O., Hill, T., Lowe, J., 1991. Peham, C., Schobesberger, H., 2006.
real- time ultrasound imaging. Eur. J. Appl. Risk factors associated with injuries in A novel method to estimate the stiffness
Physiol. 63, 250–254. Thoroughbred horses. Equine Vet. J. 23, of the equine back. J. Biomech. 39,
445–448. 2845–2849.
Mason, B.J.E., 1971. A case of spinal cord
compression causing paraplegia of a foal. Mok, N.W., Brauer, S.G., Hodges, P.W., 2011.
Peloso, J.M., Mundy, G.D., Cohen, N.D., 1994.
Equine Vet. J. 3, 155–157. Changes in lumbar movement in people
Prevalence of and factors associated with
with low back pain are related to
Mattoon, J.S., Drost, W.T., Grguric, M.R., Auld, musculoskeletal racing injuries of
compromised balance. Spine 36 (1),
D.M., Reed, S.M., 2004. Technique for Thoroughbreds. J. Am. Vet. Med. Assoc.
e45–e52.
equine cervical articular process joint 204, 620–626.
Mooney, V., 2007. Effective rehabilitation of
injection. Vet. Radiol. Ultrasound 45 (3), Penell, J.C., Egenvall, A., Bonnett, B.N., Olson,
lumbar and pelvic girdle pain. In:
238–240. P., Pringle, J., 2005. Specific causes of
Vleeming, A., Mooney, V., Stoeckart, R.
McGill, S.M., 2007. The painful and unstable morbidity among Swedish horses insured
(Eds.), Movement, stability and
lumbar spine: a foundation and approach for veterinary care between 1997 and 2000.
lumbopelvic pain, 2nd edn. Churchill
for restabilization. In: Vleeming, A., Vet. Rec. 157, 470–477.
Livingstone, Elsevier Science, Edinburgh,
Mooney, V., Stoeckart, R. (Eds.), Movement, pp. 573–588. Pickar, J.G., 2002. Neurophysiological effects
stability and lumbopelvic pain, second ed. of spinal manipulation. Spine J. 2,
Murray, R.C., Walters, J.M., Snart, H., Dyson,
Churchill Livingstone, Elsevier Science, 357–371.
S.J., Parkin, T.D., 2010. Identification of risk
Edinburgh, pp. 529–545. Pickersgill, C.H., Marr, C.M., Reid, S.W., 2001.
factors for lameness in dressage horses. Vet.
McGowan, C., Goff, L., Stubbs, N., 2007a. J. 184, 27–36. Repeatability of diagnostic ultrasonography
Animal physiotherapy: Assessment, in the assessment of the equine superficial
Norasteh, A., Ebrahimi, E., Salavati, M., Rafiei,
treatment and rehabilitation of animals. digital flexor tendon. Equine Vet. J. 33 (1),
J., Abbasnejad, E., 2007. Reliability of
Blackwell Publishing, Ames, IA. 33–37.
B-mode ultrasonography for abdominal
McGowan, C.M., Hodges, P.W., Jeffcott, L.B., muscles in asymptomatic and patients with Piercy, R.J., Rivero, J.L.L., 2004. Chapter 2.1
2007b. Epaxial musculature and its acute low back pain. J. Bodywork Muscle disorders of equine athletes. In:
relationship with back pain in the horse. A Movement Ther. 11 (1), 17–20. Hinchcliff, K.W., Kaneps, A. J., Geor, R. J.
report for the Rural Industries Research and Nussbaum, E.L., Downes, L., 1998. Reliability (Eds.), Equine sports medicine and surgery:
Development Corporation. (www.rirdc.gov. of clinical pressure-pain algometric basic and clinical sciences of the equine
au) measurements obtained on consecutive athlete. W.B. Saunders, London, UK, pp.
McGowan, C.M., Stubbs, N.C., Jull, G.A., days. Phys. Ther. 78, 160–169. 77–110.
2007c. Equine physiotherapy: a Ohrbach, R., Gale, E.N., 1989a. Pressure pain Piercy, R.J., Weller, R., 2009. Chapter 16:
comparative view of the science thresholds, clinical assessment, and Muscular disorders of the equine back. In:
underlying the profession. Equine Vet. J. differential diagnosis: reliability and Henson, F. (Ed.), Back pathology: diagnosis
39, 90–94. validity in patients with myogenic pain. and treatment. Blackwell Publishing Ltd,
McMeeken, J.M., Beith, I.D., Newham, D.J., Pain 39, 157–169. Oxford, pp. 168–180.
Milligan, P., Critchley, D.J., 2004. The Ohrbach, R., Gale, E.N., 1989b. Pressure pain Pietrek, M., Sheikhzadeh, A., Hagins, M.,
relationship between EMG and change in thresholds in normal muscles: reliability, Nordin, M., 2000. Evaluation of abdominal
thickness of transversus abdominis. Clin. measurement effects, and topographic muscles by ultrasound imaging: reliability,
Biomech. 19 (4), 337–342. differences. Pain 37, 157–169. and comparison to electromyography. Eur.
Meehan, L., Dyson, S., Murray, R., 2009. O’Sullivan, P.B., Phyty, G.D., Twomey, L.T., Spine J. 9, 309.
Radiographic and scintigraphic evaluation Allison, G.T., 1997. Evaluation of specific Pilsworth, R.C., 2003. The European
of spondylosis in the equine stabilizing exercise in the treatment of thoroughbred. In: Ross, M.W., Dyson, S.J.
thoracolumbar spine: a retrospective study. chronic low back pain with radiologic (Eds.), Diagnosis and management of
Equine Vet. J. 41, 800–807. diagnosis of spondylolysis or lameness in the horse. W.B. Saunders,
Melzack, R., Stillwell, D.M., Fox, E.J., 1977. spondylolisthesis. Spine 22 (24), Philadelphia, pp. 879–894.
Trigger points and acupuncture points for 2959–2967. Pöntinen, P.J., 1998. Reliability, validity,
pain: correlations and implications. Pain 3, O’Sullivan, P.B., Twomey, L., Allison, G.T., reproducibility of algometry in diagnosis of
3–23. 1998. Altered abdominal muscle active and latent tender spots and trigger
Michlovitz, S., 1996. Thermal agents in recruitment in patients with chronic back points. J. Musculoskeletal Pain 6, 61–71.
rehabilitation. FA, Davis, Philadelphia, pain following a specific exercise Pool, R.R., 1996. Pathologic Manifestations of
Michlovitz, S.A.L., Harris, B.A., Watkins, M.P., intervention. J. Orthop. Sports Phys. Ther. joint disease in the athletic horse. In:
2004. Therapy interventions for improving 27 (2), 114–124. McIlwraith, C.W., Trotter, G.W. (Eds.), Joint
joint range of motion: a systematic review. Owings, T.M., Grabiner, M.D., 2002. Motor disease in the horse. W.B. Saunders,
J. Hand Ther. 17, 118–131. control of the vastus medialis oblique and Philadelphia, pp. 87–104.

414
 References

Porter, M., 2005. Equine rehabilitation therapy measured by real-time ultrasonography. Rose, K.P., Scott, S.H., 2003. Sensory-motor
for joint disease. Vet. Clin. N. Am.: Equine Clin. Rehab. 12 (1), 36–44. control: a long-awaited behavioural
Pract. 21, 599–607. Rezasoltani, A., Ylinen, J., Kauhanen, H., correlate of presynaptic inhibition. Nat.
Pressler, J.F., Heiss, G.D., Buford, J.V., 1998b. Real-time ultrasonography of the Neurosci. 6, 1243–1245.
Chidley, J.A., 2006. Between day neck extensor muscles during isometric Rose, N.S., Northrop, A.J., Brigden, C.V.,
repeatability and symmetry of multifidus neck extension. In: Häkkinen, K. (Ed), Martin, J.H., 2009. Effects of a stretching
cross-sectional area measured using Conference book. international conference regime on stride length and range of
ultrasound imaging. J. Orthop. Sports on weightlifting and strength training. motion in equine trot. Vet. J. 181, 53–55.
Phys. Ther. 36 (1), 10–18. Gummerus Printing, Jyväskylä, Finland, p. Ross, W.A., Kaneene, J.B., Caron, J.P.,
Pritchett, L.C., Ulibarri, C., Roberts, M.C., 205. Gallagher, K.F., Gardiner, J.C., 1999. Factors
Schneider, R.K., Sellon, D.C., 2003. Ricardi, G., Dyson, S.J., 1993. Forelimb influencing recovery from and duration of
Identification of potential physiological lameness associated with radiographic lameness in Michigan (USA) horses. Prev.
and behavioural indicators of postoperative abnormalities of the cervical vertebrae. Vet. Med. 40, 127–138.
pain in horses after exploratory celiotomy Equine Vet. J. 25, 422–426. Ross, W.A., Kaneene, J.B., Gardiner, J.C., 1998.
for colic. Appl. Anim. Behav. Sc. 80, 31–43. Richardson, C., Snijders, C.J., Hides, J.A., Survival analysis of risk factors associated
Procacci, P., 1993. Chronobiological studies Damen, L., Pas, M.S., Storm, J., 2002. with the occurrence of lameness in a
on pain threshold. Pain 55, 277. The relation between the transversus Michigan horse population. Am. J. Vet. Res.
Pusey, A., Brooks, J., Jenks, A., 2010. abdominis muscles, sacroiliac joint 59, 23–29.
Osteopathy and the treatment of horses. mechanics and low back pain. Spine 27, Rossdale, P.D., Hopes, R., Wingfield Digby,
Wiley-Blackwell, Ames, IA. 399–405. N.J., Offord, K., 1985. Epidemiological
Quiroz-Rothe, M., Novales, M., Aguilare- Ridgway, K., 1999. Acupuncture as a treatment study of wastage among racehorses 1982
Tejero, E., Rivero, J., 2002. Polysaccharide modality for back problems. Vet. Clin. N. and 1983. Vet. Rec. 116, 66–69.
storage myopathy in the M. longissimus Am.: Equine Pract. 15, 211–221. Sapega, A.A., Quedenfeld, T.C., Moyer, R.A.,
lumborum of show jumpers and dressage Ridgway, K., 2005. Diagnosis and treatment of Butler, R.A., 1981. Biophysical factors in
horses with back pain. Equine Vet. J. 34 equine musculoskeletal pain. The role of range-of-motion exercise. Phys. Sportsmed.
(2), 171–176. the complementary modalities: 9, 57–65.
Raadsheer, M.C., van Eijden, T.M., van acupuncture and chiropractic. AAEP Saunders, S.W., Schache, A., Rath, D., Hodges,
Spronsen, P.H., van Ginkel, F.C., Kiliaridis, Proceedings 51, 403–408. P.W., 2005. Changes in three dimensional
S., Prahl-Andersen, B., 1994. A comparison Robert, C., Valette, J.P., Denoix, J.M., 2000. lumbo-pelvic kinematics and trunk muscle
of human masseter muscle thickness The effects of treadmill inclination and activity with speed and mode of
measured by ultrasonography and magnetic speed on the activity of two hind limb locomotion. Clin. Biomech. 20 (8),
resonance imaging. Arch. Oral. Biol. 39 muscles in the trotting horse. Equine Vet J. 784–793.
(12), 1079–1084. 32, 312–317. Scott, M., 2008. Musculoskeletal injuries in
Ramon, T., Prades, M., Armengou, L., Lanovaz, Robert, C., Valette, J.P., Denoix, J.M., 2001. The Nonracing Quarter Horses. Vet. Clin.
J.L., Mullineaux, D.R., Clayton, H.M., 2004. effects of treadmill inclination and speed Equine Prac. 24, 133–152.
Effects of athletic taping of the fetlock on on the activity of three trunk muscles in Seki, K., Perlmuter, S.I., Fetz, E., 2003.
distal limb mechanics. Equine Vet. J. 36 the trotting horse. Equine Vet. J. 33, Sensory input to primate spinal cord
(8), 764–768. 466–472. is presynaptically inhibited during
Ramzan, P.H.L., Palmer, L., 2010. Roberts, E.J., 1968. Resection of the thoracic or voluntary movement. Nat. Neurosci. 6,
Musculoskeletal injuries in Thoroughbred lumbar spinous processes for the relief of 1309–1316.
racehorses: A study of three large training pain responsible for lameness and some Sharman, M.J., Cresswell, A.G., Riek, S.,
yards in Newmarket, UK (2005–2007). Vet. other locomotor disorders of horses. Proc. 2006. Proprioceptive neuromuscular
J. 87, 325–329. 14th Ann. Conv. Am. Ass. Equine Practnrs. facilitation stretching: mechanisms and
Ranner, W., Gerhards, H., Klee, W., 2002. 13–30. clinical implications. Sports Med. 36,
Diagnostische validität der palpation bei Robertson, S., 2006. The importance of 929–939.
pferden mit rückenproblemen. Berl. assessing pain in horses and donkeys. Simons, D.G., Travel, J.G., Simons, L.S., 1999.
Münch. tierärztl. Wochenschr. 115, Equine Vet J. 38 (1), 5–6. Travell & Simons’ myofascial pain and
420–424. Robertson, S.A., 2002. What is pain? J. Am. dysfunction: The trigger point manual.
Rantanen, J., Hurme, M., Falck, B., Alaranta, Vet. Med. Assoc. 221, 202–205. Vol:1. Upper Half of Body. Vol:2. The
H., Nykvist, F., Lehto, M., et al., 1993. Robinson, R.A., Gordon, B., 1988. American Lower Extremities. second ed. Lippincott
The lumbar multifidus muscle five years association of equine practitioners track Williams & Wilkins, Baltimore.
after surgery for a lumbar intervertebral breakdown studies – horse results. 7th Singer, E.R., Barnes, J., Saxby, F., Murray, J.K.,
disc herniation. Spine 18 (5), 568–574. International Conference of Racing Analysts 2008. Injuries in the event horse: training
Reef, V.B., 1998. Musculoskeletal and Veterinarians, Lexington, KY, USA, versus competition. Vet. J. 75, 76–81.
ultrasonography. In: Reef, V.B. (Ed.), pp. 385–394. Skarda, R.T., Tejwani, G.A., Muir, W.W., 2002.
Equine diagnostic ultrasound. W.B. Roethlisberger Holm, K., Wennerstrand, J., Cutaneous analgesia, hemodynamic and
Saunders, Philadelphia, pp. 39–186. Lagerquist, U., Eksell, P., Johnston, C., respiratory effects, and β-endorphin
Reimann, B.L., Lephart, S.M., 2002a. The 2006. Effect of local analgesia on concentration in spinal fluid and plasma of
sensorimotor system, Part I: The movement of the equine back. Equine Vet. horses after acupuncture and
physiologic basis of functional joint J. 38, 65–69. electroacupuncture. Am. J. Vet. Res. 64,
stability. J. Athletic Training 37, 71–79. Rooney, J.R., 1970. Autopsy of the Horse. 1435–1442.
Reimann, B.L., Lephart, S.M., 2002b. The Williams & Wilkins Co, Baltimore. Sluka, K.A., Skyba, D.A., Radhakrishnan, R.,
sensorimotor system. Part II: The role of Rooney, J.R., 1977. Sacroiliac arthosis and Leeper, B.J., Wright, A., 2006. Joint
proprioception in motor control and ‘stifle lameness’. Mod. Vet. Pract. 28, mobilization reduces hyperalgesia
functional joint stability. J Athletic Training 138–139. associated with chronic muscle and joint
37, 80–84. Rooney, J.R., Genovese, R.L., 1981. A survey inflammation in rats. J. Pain 7, 602–607.
Rezasoltani, A., Kallinen, M., Mälkiä, E., and analysis of bowed tendon in Smythe, R.H., 1962. Ankylosis of the equine
Vihko, V., 1998a. Neck semispinalis capitis Thoroughbred racehorses. Equine Vet. Sci. spine: pathologic or biologic? Mod. Vet.
muscle size in sitting and prone positions 5, 49–53. Pract. 43, 50–51.

415
 17 Rehabilitation of the locomotor apparatus

Stecher, R.M., Goss, L.F., 1961. Ankylosing Tomlinson, J.E., Sage, A.M., Turner, T.A., 2003. minimally invasive sacroiliac joint
lesions of the spine. J. Am. Vet. Med. Ass. Ultrasonographic abnormalities detected in procedures. Arch. Phys. Med. Rehabil. 87
138, 248–255. the sacroiliac area in twenty cases of upper (1), 10–14.
Steckel, R., 1992. Scintigraphic diagnosis of hind limb lameness. Equine Vet. J. 35, Van Eps, A.W., Pollitt, C.C., 2010. Equine
thoracolumbar spinal disease in horses: a 48–54. laminitis: cryotherapy reduces the severity
review of 50 cases. Proc. Am. Ass. Equine Tomlinson, J.E., Sage, A.M., Turner, T.A., of the acute lesion. Equine Vet. J. 36,
Practnrs. 37, 301–314. Feeney, D.A., 2001. Detailed 255–260.
Steiner, B., Bryant, J., 2003. A gymnastic riding ultrasonographic mapping of the pelvis in Van Harreveld, P.D., Lillich, J.D., Kawcak, C.E.,
system using mind body and spirit. clinically normal horses and ponies. Am. J. Gaughan, E.M., McLaughlin, R.M.,
Progressive training for the horse and rider. Vet. Res. 62 (11), 1768–1775. DeBowes, R.M., 2002. Clinical evaluation
Trafalgar Sqaure Publishing, Hong Kong. Torre, F., 2003. The European and Australasian of the effects of immobilization followed
Steiss, J.E., White, N.A., Bowen, J.M., 1989. Standardbreds. In: Ross, M.W., Dyson, S.J. by remobilization and exercise on the
Electroacupuncture in the treatment of (Eds.), Diagnosis and management of metacarpophalangeal joint in horses. Am.
chronic lameness in horses and ponies: a lameness in the horse. W.B. Saunders, J. Vet. Res. 63, 282–288.
controlled clinical trial. Can. J. Vet. Res. 53, Philadelphia, pp. 913–927. Van Loon, J.P.A.M., Back, W., Hellebrekers, L.J.,
239–243. Tough, L., 2006. Lack of effect of acupuncture Van Weeren, P.R., 2010. Application
Stokes, M., Rankin, G., Newham, D.J., 2005. on electromyographic (EMG) activity- a of a composite pain scale to objectively
Ultrasound imaging of lumbar multifidus randomized controlled trial in healthy monitor horses with somatic and
muscle: normal reference ranges for volunteers. Acupuncture Med. 24, 55–60. visceral pain under university hospital
measurements and practical guidance on Townsend, H.G.G., Leach, D.H., Doige, C.E., conditions. J. Equine Vet. Sci. 30 (11),
the technique. Man. Ther. 10 (2), 116–126. Kirkaldy-Willis, W.H., 1986. Relationship 641–649.
Stokes, M., Young, A., 1986. measurement of between spinal biomechanics and van Schie, H.T., Bakker, E.M., Jonker, A.M., van
quadriceps cross-sectional area by pathological changes in the equine Weeren, P.R., 2000. Ultrasonographic tissue
ultrasonography: A description of the thoracolumbar spine. Equine Vet. J. 18, characterization of equine superficial
technique and its applications to 107–112. digital flexor tendons by means of gray
physiotherapy. Physiother. Pract. 2, 31–36. Townsend, H.G.G., Leach, D.H., Fretz, B., level statistics. Am. J. Vet. Res. 16 (2),
Stubbs, N.C., Clayton, H.M., 2008. Activate 1983. Kinematics of the equine 210–219.
your horse’s core: Unmounted exercises for thoracolumbar spine. Equine Vet. J. van Schie, J.T., Bakker, E.M., van Weeren, P.R.,
dynamic mobility, strength, and balance. 15 (2), 117–122. 1999. Ultrasonographic evaluation of
Sport Horse Publications, Mason, MI. Triano, J., 2005. The theoretical basis for equine tendons: a quantitative in vitro
Stubbs, N.C., Clayton, H.M., Hodges, P.W., spinal manipulation. In: Haldeman, S. study of the effects of amplifier gain level,
Jeffcott, L.B., McGowan, C.M., 2010. (Ed.), Principles and practice of transducer-tilt, and transducer-
Osseous spinal pathology and epaxial chiropractic. third ed. McGraw-Hill, New displacement. Vet. Radiol. Ultrasound 40
muscle ultrasonography in Thoroughbred York, NY, pp. 361–381. (2), 151–160. Review.
racehorses. Equine Vet. J. 42 (Suppl. 38), Tsao, H., Tucker, K.J., Coppieters, M.W., van Sluijs, F.J., 2000. De pretenties van de
654–661. Hodges, P.W., 2010. Experimentally induced (dieren)arts. Tijdschr. Diergeneeskd. 125,
Stubbs, N.C., Hodges, P.W., Jeffcott, L.B., low back pain from hypertonic saline 153–154.
Cowin, G., Hodgson, D.R., McGowan, injections inot lumbar ligament and erector Van Weeren, P.R., McGowan, C.M., Haussler,
C.M., 2006. Functional anatomy of the spinae muscle. Pain 150 (1), 167–172. K.K., 2010. Science overview: development
caudal thoracolumbar and lumbosacral Tsukiyama, K., Acorda, J.A., Yamada, H., 2005. of a structural and functional
spine in the horse. Equine Vet. J. 36 Evaluation of superficial digital flexor understanding of the equine back. Equine
(Suppl.), 393–399. tendinitis in racing horses through gray Vet. J. 42 (Suppl.), 393–400.
Stubbs, N.C., Kaiser, L.J., Hauptman, J., scale histogram analysis of tendon Vanderweeën, L., Oostendorp, R.A.B., Vaes, P.,
Clayton, H.M., 2011. Dynamic mobilization ultrasonograms. Vet. Radiol. Ultrasound 37 Duquet, W., 1996. Pressure algometry in
exercises increase cross sectional area of (1), 46–50. manual therapy. Man. Ther. 1, 258–265.
multifidus. Equine Vet. J. doi: Turner, T.A., 2003. Back problems in Horses. Vanderweeën, L., Oostendorp, R.A.B., Vaes, P.,
10.1111/j.2042–3306.2010.00322.x. Proc. 49th Annu. Conv. Am. Assoc. Equine Duquet, W., 1996. Pressure algometry in
Sullivan, K.A., Hill, A.E., Haussler, K.K., 2008. Practitioners, New Orleans, LA. USA, pp. manual therapy. Man. Ther. 1 (5),
The effects of chiropractic, massage and 21–25. 258–265.
phenylbutazone on spinal nociceptive Valberg, S., 1999. Spinal muscle pathology. Varcoe-Cocks, K., Sagar, K.N., Jeffcott, L.B.,
thresholds in horses without clinical signs. Vet. Clin. N. Am.: Equine Pract. 15 (1), McGowan, C.M., 2006. Pressure algometry
Equine Vet. J. 40, 14–20. 87–96. to quantify muscle pain in racehorses with
Sumner, H., 1948. Equine accidents in the Valentin, S., Davis, R., Wilson, A., Pfau, T., suspected sacroiliac dysfunction. Equine
field of sport. Vet. Rec. 60, 579–581. 2010. The effect of loading on the equine Vet. J. 38, 558–562.
Tanaka, T.H., Leisman, G., Nishijo, K., 1998. spine – a preliminary study. Vet. Med. Vaughan, L.C., Mason, J.E., 1976. A clinic-
Dynamic electromyographic response Austria 97, 74–80. doi: 10.4112–10–921. pathological study of racing accidents in
following acupuncture: possible influence Valentine, B.A., 2008. Pathologic findings in horses. A report of a study on equine fatal
on synergistic coordination. Int. J. equine muscle (excluding polysaccharide accidents on race courses. Financed by the
Neurosci. 95, 51–61. storage): a necropsy study. J. Vet. Diagn. Horserace Betting Levy Board.
Taylor, H.W., Vandevelde, M., Firth, E.C., 1977. Invest. 20 (5), 572–579. Veeneklaas, R.J., 1999. Neurofysiologische
Ischemic myelopathy caused by Van Biervliet, J., Mayhew, J., de Lanhunta, A., basis van de acupunctuur.
fibrocartilaginous emboli in a horse. Vet. 2006. Cervical vertebral compressive Diergeneeskundig Memorandum 46,
Pathol. 14 (5), 479–481. myelopathy: Diagnosis. Clin. Tech. Equine 26–36.
Tellington-Jones, L., Bruns., U., Isenbugel, Pract. 5, 54–59. Vicenzino, B., Collins, D., Benson, H., Wright,
V.M.D.E., 1988. An introduction to the van der Molen, C., 1999. Acupunctuur, 4th A., 1998. An investigation of the
tellington-jones equine awareness method: edn. Elsevier, De Tijdstroom, Maarssen. interrelationship between manipulative
The T.E.A.M. approach to problem-free van der Wurff, P., Buijs, E.J., Groen, G.J., 2006. therapy-induced hypoalgesia and
training. Breakthrough Publications, New A multitest regimen of pain provocation sympathoexcitation. J. Manip. Physiol.
York. tests as an aid to reduce unnecessary Ther. 21, 448–453.

416
 References

Vicenzino, B., Paungmali, A., Teys, P., 2007. Watson, T., 2008. Electrotherapy: evidence- Willy, R.W., Kyle, B.A., Moore, S.A., Chleboun,
Mulligan’s mobilisation-with –movement, based practice (Physiotherapy Essentials), G.S., 2001. Effect of cessation and
positional faults and pain relief: Current 12th edn. Churchill Livingstone, Elsevier, resumption of static hamstring muscle
concepts forom a critical review of the Edinburgh. stretching on joint range of motion.
literature. Man. Ther. 12 (2), 98–108. WCPT (World Confederation for Physical J. Orthop. Sports Phys. Ther. 31,
Vigre, H., Chriél, M., Hesselholt, M., Falk- Therapy) http://www.wcpt.org/. 138–144.
Rønne, J., Kjaer Ersbøll, A., 2002. Risk Weaver, M., Jeffcott, L., Nowak, M., 1999. Back Wolf, L., 2002. The role of complementary
factors for the hazard of lameness in problems radiology and scintigraphy. Vet. techniques in managing musculoskeletal
Danish Standardbred trotters. Prev. Vet. Clin. N. Am.: Equine Pract. 15, 113–129. pain in performance horses. Vet. Clin. N.
Med. 56, 105–117. Weller, R., Pfau, T., Babbage, D., Brittin, E., Am.: Equine Pract. 18, 107–115.
von Salis, B., Huskamp, B., 1978. Vorlaufige May, S.A., Wilson, A.M., 2006. Reliability of Woolf, C.J., Mannion, R.J., 1999. Neuropathic
erfahrungen mit der konservativen and conformational measurements in the horse pain: aetiology, symptoms, mechanisms,
chirurgischen behandlung der using a three-dimensional motion analysis and management. Lancet 353,
wirbersaulener-krakrung der pferd. Der system. Equine Vet. J. 38, 610–615. 1959–1964.
praktische Tierarzt. 59, 281–283. Wennerstrand, J., Johnston, C., Roethlisberger Wright, A., 1995. Hypoalgesia post-
Wakeling, J.M., Barnett, K., Price, S., Nankervis, Holm, K., Erichsen, C., Eksell, P., Drevemo, manipulative therapy: a review of a
K., 2006. Effects of manipulative therapy S., 2004. Kinematic evaluation of the back potential neurophysiological mechanism.
on the longissimus dorsi in the equine in the sport horse with back pain. Equine Man. Ther. 1, 11–16.
back. Equine Comp. Exerc. Phys. 3, Vet. J. 36, 707–711. Xie, H., Colahan, P., Ott, E.A., 2005.
153–160. Wennerstrand, J., Johnston, J., Rhodin, M., Evaluation of electroacupuncture treatment
Wakeling, J.M., Ritruechia, P., Dalton, S., Roethlisberger-Holm, K., Drevemo, S., of horses with signs of chronic TL pain. J.
Namkervis, K., 2007. Segmental variation 2006. The effect of weighted boots on the Am. Vet. Med. Assoc. 227, 281–286.
in the activity and function of the equine movement of the back in the asymptomatic Xie, H., Ott, E.A., Colahan, P., 2001.
longissimus dorsi muscle during walk and riding horse. Equine Comp. Exerc. Physiol. Influence of acupuncture on experimental
trot. Equine Comp. Exer. Physiol. 4, 3 (1), 13–18. lameness in horses. AAEP Proceedings 47,
95–103. Weppler, C.H., Magnusson, S., 2010. Increasing 347–357.
Wallin, L., Strandberg, E., Philipson, J., Dalin, muscle extensibility: A matter of increasing Yan, B., Xu, J., Li, K., Liu, H., Shan, B., Tang,
G., 2000. Estimates of longevity and causes length of modifying sensation? Phys. Ther. X., 2005. Acupoint-specific fMRI patterns in
of culling and death in Swedish 90 (3), 438–449. human brain. Neurosci. Lett. 383,
warmblood and coldblood horses. Livest. Wessling, K.C., DeVane, D.A., Hylton, C.R., 236–240.
Prod. Sci. 63 (3), 275–289. 1987. Effects of static stretch versus static Ylinen, J., 2007. Pressure algometry. Aust. J.
Wallwork, T.L., Hides, J.A., Stanton, W.R., stretch and ultrasound combined on triceps Physiother. 53, 207.
2007. Intrarater and interrater reliability of surae muscle extensibility in healthy Ylinen, J., Nykänen, M., Kautiainen, H.,
assessment of lumbar multifidus muscle women. Phys. Ther. 67, 674–679. Häkkinen, A., 2007. Evaluation of
thickness using rehabilitative ultrasound Whittaker, J.L., Elliott, J.M., Cook, K., repeatability of pressure algometry on the
imaging. J. Orthop. Sports Phys. Ther. 37 Langevin, H.M., Dahl, H.H., Stokes, M., neck muscles for clinical use. Man. Ther.
(10), 608–612. et al., 2007. Rehabilitative ultrasound 12, 192–197.
Walmsley, J.P., Pettersson, H., Winberg, F., imaging: understanding the technology and Young, A., 1993. Current issues in
McEvoy, F., 2002. Impingment of the dorsal its applications. J. Orthop. Sports Phys. 37 arthrogenous inhibition. Ann. Rheum. Dis.
spinous processes in two hundred and (8), 434–449. 52, 829–834.
fifteen horses: case selection, surgical Wijnberg, I.D., Franssen, H., van der Kolk, Zaneb, H., Kauffman, V., Stanek, C., Peham,
technique and results. Equine Vet. J. 34, J.H., Breukink, H.J., 2002. Quantative C., Licka, T., 2009. Quantitative differences
23–28. motor unit action potential analysis of in activities of back and limb muscles
Warren, C.G., Lehmann, J.F., Koblanski, J.N., skeletal muscles in the warmblood horse. during walking and trotting between
1971. Elongation of rat tail tendon: effect Equine Vet. J. 34, 556–561. chronically lame and non lame horses. Am.
of load and temperature. Arch. Phys. Med. Wilke, H.J., Wolf, S., Claes, L.E., Arand, M., J. Vet. Res. 70 (9), 1129–1134.
Rehabil. 52, 465–474. Wiesend, A., 1995. Stability increase of the Zsoldos, R.R., Kotschwar, A.B., Kotschwar, A.,
Watanabe, K., Miyamoto, K., Masuda, T., lumbar spine with different muscle groups. Groesel, M., Licka, T., Peham, C., 2010.
Shimizu, K., 2004. Use of ultrasonography A biomechanical in vitro study. Spine 20 Electromyography activity of the
to evaluate thickness of the erector spinae (2), 192–198. equine splenius muscle and neck
muscle in maximum flexion and extension Williamson, A., Hoggart, B., 2005. Pain: A kinematics during walk and trot on the
of the lumbar spine. Spine 29 (13), review of three commonly used pain rating treadmill. Equine Vet. J. 42 (suppl. 38),
1472–1477. scales. J. Clin. Nurs. 14, 798–804. 455–461.

417
This page intentionally left blank
 CHAPTER 18 

Metabolic energetics
Mathew P. Gerard, E. de Graaf-Roelfsema, David R. Hodgson, J.H. van der Kolk

Introduction critical as the flux of glucose from the circulation during exercise is
limited and is likely to provide no more than ~10% of the energy
used during intense exercise. Mitochondria produce the energy for
The horse is a superb athlete, the result of the evolutionary adapta- muscle contraction. It is not surprising therefore that skeletal muscle
tions required to live as a species on open landscapes. These of horses contains approximately twice the concentration of mito-
adaptations resulted in speed, to escape predators, and endurance, chondria found in humans, sheep and cattle. These factors com-
necessary in the search for nutrition. Selective breeding by humans bined with the profound cardiac output and oxygen carrying
has further modified these evolutionary traits. capacity of horses permits this species to have a higher whole
Domesticated horses have been selected for certain characteristics animal maximal aerobic capacity when compared to less athletic
depending on the intended use. Heavy breeds were selected for draft domestic species (Hinchcliff et al., 2004).
or military work whereas lighter horses were chosen because of their The circulation is responsible for transport of oxygen from the
speed and endurance. This has resulted in myriad breeds and capac- lungs to exercising muscle and other tissues. Cardiac output and
ities of each. For example, Thoroughbreds can achieve speeds of oxygen-carrying capacity of blood are key determinants in this
>65 km/h when racing over 800–7000 m, Standardbreds trot or process. As stated earlier horses achieve rapid increases in the
pace at up to 55 km/h for up to 4000 m, Quarter Horses sprint over oxygen-carrying capacity of blood via exercise-induced splenic
400 m (1/4 mile) at speeds approaching 90 km/h, yet endurance contraction. This epinephrine-induced contraction (autoinfusion)
horses (of Arabian breed) can cover 160 km at average speeds of results in substantive increases (~50%) in blood hemoglobin con-
~25 km/h. In contrast, a pair of draft horses pull a sled weighing centrations without concomitant increases in plasma volume.
>2000 kg over specific distances from 3.7 to 8.5 m, whereas breeds The limbs of horses are long and their associated muscles have
such as Warmbloods may participate in eventing, show jumping relatively low power outputs when compared to smaller athletic
and dressage competitions. species. However, despite this, horses have adapted such that their
The athletic capacity of horses results from physiologic properties. gait is energetically efficient. This results from the fact that up to
These may be innate and as such are not affected by training, for 50% of the muscular work occurring during galloping is stored as
example respiratory capacity. In contrast other variables do respond elastic energy in muscle and tendon units. This mechanism is of
to training, including skeletal muscle respiratory capacity and blood such magnitude that it has been estimated that in the forelimb of
volume. Overall, particularly when compared to humans, the note- horses, energy storage and subsequent release means key muscles
worthy athletic ability of horses can be attributed to: of the forelimb are substantially smaller than if this process were
not available (Hinchcliff et al., 2004).
• High maximal aerobic capacity.
Vertebrate locomotion requires the controlled integration of
• Large intramuscular stores of energy, particularly glycogen.
numerous physiological and metabolic pathways that have an
• High respiratory capacity of skeletal muscle.
impact on the musculoskeletal system to provide the organism with
• Splenic contraction, which results in the oxygen-carrying
mobility. Perhaps the most important pathways are those concerned
capacity of blood increasing by up to 50% soon after the
onset of exercise. with the production of energy, for without energy muscles cannot
contract and mobility is not achieved. One of the major survival
• Highly efficient and adaptable gait(s).
mechanisms for the wild equid is swift retreat from threatening
• Well-developed capacity for effective thermoregulation
circumstances. Consequently, the horse has an efficient locomotor
(Hinchcliff et al., 2004).
system designed to move a relatively large mass with great speed
The maximal aerobic capacity of horses is over 2-fold greater and endurance.
than that of similarly sized cattle. This is due to many of the adapta- Muscular movement requires the transformation of chemical
tions outlined in the previous paragraph. Substrate is required energy stored in metabolic fuels to the kinetic energy of muscular
to support these high metabolic rates during exercise. Substrate to contraction. All pathways integral to energy supply are concerned
support exercise is generally either carbohydrate or fatty acids. In with the ultimate production of adenosine triphosphate (ATP), the
most species oxidation of fatty acids, although highly efficient in final carrier of energy ‘packages’ utilized by muscle for contraction.
terms of energy yield, is limited, and is likely to reach capacity when Muscular contraction involves coupling of thin actin and thick
work rate requires ~50% of the maximal oxygen consumption. myosin filaments to form cross-bridges and then these filaments
Above this exercise intensity substrate is provided by the oxidation slide relative to each other by a change in orientation of the cross-
of carbohydrates, predominantly glycogen. Athletic species such as bridges (Guyton, 1986). Energy is necessary for the change in ori-
horses and dogs have high intramuscular concentrations of glyco- entation of the cross-bridges to occur. Cleavage of a high-energy
gen relative to humans. This locally available source of substrate is phosphate bond from ATP results in adenosine diphosphate (ADP),

419
 18 Metabolic energetics

High-energy bonds The availability of O2 to the exercising muscle is the rate-

limiting step for oxidative phosphorylation. Oxygen immediately
available to the muscle at the onset of exercise, from myoglobin
within the muscle (MbO2), hemoglobin within the circulatory
Adenosine P P P + H2O system (HbO2) or O2 dissolved in the body fluids, is in sufficient
quantities for only a few seconds of exercise. Therefore the deliv-
ery of O2 to the exercising muscles via the cardiorespiratory system
Myosin
is crucial for the capacity to continuously produce energy via
ATPase
aerobic means.
The two major electron donor substrates for aerobic phosphory-
Adenosine P P + energy + Pi + H+ lation are carbohydrates (CHO) and fatty acids. Glucose is the main
CHO and if it is not used for immediate energy production, it is
Fig 18.1  Hydrolysis of ATP to ADP by the enzyme myosin ATPase, with stored as glycogen, mostly in skeletal muscle and to a lesser extent
the release of energy for use by working muscle. P, phosphate; Pi: in the liver (Lindholm, 1979; Hodgson et al., 1983, 1984). Adipose
orthophosphate. tissue constitutes the largest store of fatty acids (FA) (Robb et al.,
Modified and reprinted from Clayton, H.M. (1991) Energy production. In: Clayton, H.M. (ed.) 1972; Stryer, 1988). Adipocytes store fat within their cytoplasm as
Conditioning Sport Horses. With permission from Sport Horse Publications. triglycerides. Triglyceride storage also occurs to a much lesser extent
in muscle (Lindholm, 1979).

free phosphate, a proton and the release of energy. This hydrolysis

reaction of ATP occurs at the head of each myosin filament and is Aerobic glycolysis
catalyzed by the enzyme myosin ATPase (Fig. 18.1). The energy The importance of CHO as a substrate for energy production
released is utilized by the working muscle (Cain & Davies, 1962). increases as exercise intensity increases (Hodgson, 1985; Lawrence,
In addition, ATP is the source of energy required to restore the 1990). Glucose diffuses into the muscle cell cytoplasm from the
contracted muscle to a relaxed or resting state via the distribution circulation facilitated by the glucose transporter 4. Glucose trans-
of calcium ions (Åstrand & Rodahl, 1986). Under normal condi- port in mammalian skeletal muscle is almost exclusively mediated
tions there is a finite store of ATP within muscle, sufficient to by the glucose transporter 4 (GLUT-4). Insulin and muscle contrac-
maintain muscular activity for only a few seconds (Lindholm, tion are two potent stimuli for GLUT-4 translocation and both
1979; Åstrand & Rodahl, 1986). Therefore to perform continuous recruit GLUT-4 from intracellular storage sites. In equine muscle
muscular exertion it is necessary to resynthesize ATP and this is GLUT-4 is predominantly seen in the cytosol of fast type 2B fibers
performed by the pathways of aerobic (oxidative) and anaerobic (Lacombe et al., 2003; van Dam, 2004). Following uptake, CHO is
phosphorylation. phosphorylated to glucose-6-phosphate (G-6-P) in a reaction cata-
lyzed by the enzyme hexokinase (HK) and requiring one ATP mol-
Production of energy ecule. Glucose-6-phosphate is then transferred into the glycolytic
pathway for immediate energy production or reversibly converted
to glucose-1-phosphate (G-1-P) and then glycogen for storage.
Aerobic phosphorylation Glycogen stores provide most of the glucose required for energy
production during exercise. In the glycolytic pathway, G-6-P
Production of ATP via aerobic pathways occurs within the inner is phosphorylated to fructose-6-phosphate (F-6-P). Fructose-6-
membrane of mitochondria in a series of single oxidation reactions phosphate is then phosphorylated to fructose-1,6-bisphosphate
known as the electron transport or respiratory chain. Oxidation is the (F-1,6-BP) in a reaction catalyzed by phosphofructokinase (PFK) and
donation or loss of electrons (often in the form of hydrogen) from at the expense of one ATP molecule. Fructose-1,6-bisphosphate is
an atom or molecule, while reduction is the acceptance of electrons subsequently split into the triose phosphate isomers, glyceraldehyde-
(hydrogen) by an atom or molecule. When electrons are donated, 3-phosphate (Gl-3-P) and dihydroxyacetone-phosphate (DiH-P).
considerable chemical energy is liberated and a portion of this Dihydroxyacetone-phosphate is readily converted into Gl-3-P by
energy is captured for the rephosphorylation of ADP to ATP, with triose phosphate isomerase (TPI). So one molecule of glucose or gly-
the remainder being lost as heat energy (Guyton, 1986). Nicotin- cogen gives 2 molecules of Gl-3-P. Glyceraldehyde-3-phosphate
amide adenine dinucleotide (NAD) and flavin adenine dinucleo- then proceeds through a series of reactions with the end
tide (FAD) act as hydrogen carriers (acceptors) during glycolysis, result being the production of pyruvate, ATP, NADH, water and
β-oxidation and the tricarboxylic acid (TCA) cycle and therefore are hydrogen ions. The net reaction for the glycolytic pathway utilizing
reduced to NADH and FADH2. These coenzymes are essential for glucose is:
aerobic and anaerobic phosphorylation but their concentrations
within the muscle are low. Therefore, NADH and FADH2 must be Glucose + 2Pi + 2 ADP + 2NAD+
reoxidized to NAD+ and FAD via the electron transport chain. Spe- → 2 Pyruvate + 2 ATP + 2 NADH + 2H+ + 2H2O
cific mitochondrial enzymes incorporated in the electron transport
chain catalyze the oxidation through a process of dehydrogenation. Under aerobic conditions the hydrogen atoms are transferred to
The donated hydrogen atoms provide the electrons that are trans- the electron transport chain and pyruvate is transported into the
ported from one enzyme complex to another by electron carriers, mitochondrial matrix as a substrate for acetyl coenzyme A (acetyl
e.g. cytochrome c. The importance of oxygen (O2) in this whole CoA). Acetyl CoA then enters the TCA cycle by combining with
process is that it acts as the final hydrogen acceptor to form water. oxaloacetate to form citrate. The normal function of the TCA cycle
The energy released by the step-by-step transfer of electrons from requires 3 NAD+ and 1 FAD to accept hydrogen atoms during the
NADH or FADH2 to O2 via the electron carriers is used to pump oxidative conversion of citrate back to oxaloacetate. When O2 is
protons from the inner matrix of the mitochondrion into the outer available, the NADH and FADH2 are then reoxidized back to NAD+
chamber between the inner and outer mitochondrial membranes. and FAD in the electron transport chain, thus replenishing the
This creates a strong transmembrane electric potential. Energy for adenine dinucleotide stores and producing ATP (Fig. 18.2). The
the phosphorylation of ADP to ATP is obtained as the protons flow complete aerobic utilization of one mole of glucose generates
back through an inner membrane enzyme complex called ATP syn- 36–38 ATP. When O2 is not available, pyruvate is converted to
thetase (Guyton, 1986; Stryer, 1988). lactate as described later.

420
 Production of energy

CARBOHYDRATES FATS

Glycolysis Triglycerides

Glucose Glycogen
ATP
HK Glycerol Fatty acids (Cn)
ADP Glucose-1-P
ATP NAD+ CoA ATP
Glucose-6-P ADP NADH AMP+2Pi

Dihydroxyacetone Acyl CoA

phosphate
Fructose-6-P

ATP
PFK
ADP
Anaerobic

TPI
Fructose-1,6-BP

Dihydroxyacetone-P Glyceraldehyde-3-P
TPI
ADP NAD+ Cytosol
ATP NADH
Mitochondria
Phosphoenolpyruvate
Acyl CoA
ADP NAD+
PK FAD
ATP
Pyruvate Beta oxidation

C(n-2)
NAD+ NADH
NADH
Lactate Acetyl CoA (C2) FADH2
Cytosol LDH PDH

Mitochondria e-

Citrate (C6) Oxaloacetate (C4)

TCA cycle H2O ETC O2

3NAD+ ATP
Aerobic

FAD
3NADH 2CO2
FADH2

2e- Electron
ATP
transport chain
ATP 2H+
ATP
H2O O2

Fig 18.2  Schematic representation of the principal components of glycolysis, fatty acid oxidation, the tricarboxcylic acid (TCA) cycle and the electron
transport chain (ETC) in a muscle cell. ATP, adenosine triphosphate; ADP, adenosine diphosphate; BP, bisphosphate; C, carbon; CS, citrate synthase; FAD,
FADH2, flavin adenine dinucleotide; HK, hexokinase; LDH, lactate dehydrogenase; NAD+, NADH, nicotinamide adenine dinucleotide; P, phosphate; PDH,
pyruvate dehydrogenase; PFK, phosphofructokinase; PK, pyruvate kinase; TPI, triose phosphate isomerase.

421
 18 Metabolic energetics

Fatty acid utilization In the gluteus medius muscle of Standardbreds, the size of the CP
pool is estimated to be 15–20 mmol/kg wet muscle (Lindholm &
Following lipolysis, non-esterified fatty acids (NEFAs) are released Piehl, 1974; Lindholm, 1979). This source of ATP replenishment
into the circulation and are subsequently available as hydrogen would support maximum intensity exercise for no more than a few
donors for energy production in skeletal muscle. NEFAs likely seconds (Åstrand & Rodahl, 1986; Clayton, 1991).
diffuse into muscle cells down a concentration gradient (Hodgson,
1985) as well as being actively transported across the cell mem-
brane. The latter is facilitated by a membrane-associated protein Myokinase reaction
called membrane fatty acid translocase (FAT/CD36). This mem-
The myokinase enzyme catalyzes the synthesis of ATP and adenosine
brane fatty acid translocase resides in intracellular membrane frac-
monophosphate (AMP) from two ADP molecules:
tions and translocates to the plasma membrane of skeletal muscle
cells also after insulin or contraction stimuli and is seen intracel- ADP + ADP → ATP + AMP
lularly in equine muscle exclusively in type 1 fibers (van Dam, At rest, this reaction proceeds at an approximately equal rate in both
2004). At the cytoplasmic surface of the outer mitochondrial directions with little net ATP being produced. In working muscle,
membrane, the NEFAs are esterified (activated) enzymatically AMP-deaminase reduces AMP concentration by converting it to
forming long-chain acyl CoA molecules. The acyl CoA molecules ionosine monophosphate (IMP) and ammonia. This provides the
are then linked to carnitine and shuttled across to the matrix side driving force for the myokinase reaction towards the production of
of the inner mitochondrial membrane. In the mitochondria the ATP (McMiken, 1983). Again this pathway only has the capabilities
acyl CoA molecules undergo a series of four reactions known as of providing small amounts of ATP.
β-oxidation. With each cycle of β-oxidation two-carbon (C2) units
are sequentially removed from the acyl CoA molecule and acetyl
CoA, NADH and FADH2 are produced (Fig. 18.2). NADH and Anaerobic glycolysis
FADH2 subsequently donate their electrons in the electron trans- The anaerobic production of two molecules of pyruvate from one
port chain generating ATP and being reoxidized to NAD+ and FAD molecule of glucose or glycogen is identical to that described for
in the process. Acetyl CoA is utilized in the TCA cycle as previously aerobic glycolysis. In the absence of available O2, pyruvate accepts
described. The splitting of C2 units from the parent acyl CoA mol- hydrogen atoms from NADH and is converted to lactate, rather than
ecule is repeated until the whole chain has been cleaved into the being converted to acetyl CoA and entering the TCA cycle. The reac-
acetyl CoA molecules. The number of carbon atoms in the parent tion is catalyzed by lactate dehydrogenase and the regeneration of
FA chain will determine the net energy yield from β-oxidation. NAD+ during the reduction of pyruvate to lactate sustains glycolysis
Complete oxidation of one palmitic acid molecule produces 129 under anaerobic conditions (Fig. 18.2). The net result of anaerobic
molecules of ATP. glycolysis is the production of three molecules of ATP from one
molecule of glycogen or two molecules of ATP from one molecule
of glucose. This form of energy production is relatively rapid com-
Anaerobic phosphorylation pared to aerobic glycolysis but yields a significantly lower amount
The pathways of anaerobic phosphorylation occur solely in the of ATP and substrates are limited.
muscle cell cytoplasm, with no reactions in the mitochondria as
there are for aerobic phosphorylation. However, in general mito-
chondria are no longer the once-thought uniform ATP-producing Regulation of aerobic and anaerobic pathways
organelles, identical in all organisms, but are rather diverse and
adapted to distinct conditions. Next to the classical aerobic mito- At all exercise levels both systems of energy supply are active;
chondria, which contain a respiratory chain and use oxygen as a however, one will predominate, depending in particular on the
final electron acceptor, anaerobically functioning mitochondria intensity and duration of the activity. A complex method of meta-
were identified that also contain a respiratory chain and perform bolic regulation controls the input of each pathway. Substrate and
oxidative phosphorylation, but do not use oxygen as a terminal enzyme availability, end product concentrations and various feed-
electron acceptor. The energy metabolism in anaerobic mitochon- back mechanisms contribute to pathway dynamics.
dria differs from that in aerobic mitochondria, as no external final Oxygen supply to muscle and the ratio of ATP:ADP are the most
electron acceptors are used (Tielens & van Hellemond, 2007). significant regulators of the energy producing pathways. When
With the initiation of exercise there is a lag period before oxida- adequate O2 is available aerobic phosphorylation persists, provid-
tive energy production becomes an important source of ATP. During ing a driving force for substrate to enter the TCA cycle and produce
this time rapid supplies of ATP must still be available if muscular high concentrations of citrate. Citrate retards the activity of PFK, the
contraction is to continue. Stores of ATP in skeletal muscle are enzyme responsible for the irreversible conversion of F-6-P to
limited (4–6 mmol/kg wet muscle) and contribute little to the total F-1,6-BP in the glycolytic pathway. This has the effect of inhibiting
energy supply (Lindholm & Piehl, 1974; McMiken, 1983). Until glucose and glycogen metabolism. Accumulation of glycolytic inter-
aerobic phosphorylation makes a substantial contribution to energy mediates before the PFK step, including G-6-P, inhibits the activities
supply, rapid regeneration of ATP must occur in the absence of O2. of HK and phosphorylase, thus also dampening the glycolytic
The anaerobic phosphorylation of ADP is achieved by three path- pathway. High cellular concentrations of ATP (i.e. a high ATP : ADP
ways: the phosphocreatine reaction, the myokinase reaction and ratio) also inhibits PFK activity. Pyruvate kinase, the third control
anaerobic glycolysis. The former two pathways may be described site in glycolysis, is inhibited by ATP and activated by F-1,6-BP.
as anaerobic alactic reactions because no lactate is produced as it is When the relatively slow production of energy by oxidative phos-
in the latter process (Clayton, 1991). phorylation is unable to meet the demands for ATP, the ATP : ADP
ratio swings in favor of ADP and this in turn stimulates PFK and
enhances glycolysis and the utilization of glucose and glycogen
Phosphocreatine reaction stores. The pyruvate produced is metabolized to lactate and this
allows the reoxidization of NADH to NAD+ for continued electron
In this pathway the enzyme creatine kinase catalyzes a reversible acceptance in the glycolytic pathway.
reaction where creatine phosphate (CP or phosphocreatine) donates Control of the TCA cycle starts with regulation of the irreversible
its high-energy phosphate to ADP producing ATP: oxidative decarboxylation of pyruvate to acetyl CoA by pyruvate
CP + ADP → Creatine + ATP dehydrogenase (PDH). Acetyl CoA, NADH and ATP inhibit PDH

422
 Energy substrates

activity. Once acetyl CoA is formed, its combination with oxaloac-

Table 18.1  Rates of glycogen utilization in horses performing
etate to produce citrate is inhibited by high concentrations of ATP.
various types of athletic activities
At least two other sites in the TCA cycle have regulatory mecha-
nisms that respond to the cellular concentrations of ATP. Overall
the rate of the TCA cycle is reduced when abundant ATP is present Type of exercise Average speed Glycogen
but it should also be remembered the TCA cycle operates only (m/min) utilization*
under aerobic conditions because of the constant need for NAD+
Endurance ride
and FAD.
The rate of oxidative phosphorylation is regulated by the cellular 40 km 187 1.47
ADP concentration. This is known as respiratory control. For electrons
to flow through the electron transport chain to O2 the simultaneous 110 km 173 0.41
phosphorylation of ADP to ATP is usually required. When ATP is 160 km 180–135 0.69
utilized ADP concentrations increase and therefore the rate of oxi-
dative phosphorylation increases, assuming an adequate O2 supply 3-day event
is available. So oxidative phosphorylation is coupled to ATP utiliza-
Cross-country phase 293 (range 220–690) 4.08
tion via the relative ADP concentration.
(23.1 km)
Enzymes in the β-oxidation pathway are inhibited by NADH and
acetyl CoA; however, the rate of FA oxidation is determined greatly Trotting
by substrate availability.
4 h 300 0.50† (2.0)
1 h 500 0.175† (7.0)
Energy pathway contributions in the
5 min 750 15.0† (6.0)
exercising horse
Racing
The duration and intensity of exercise determine the metabolic
requirements of muscle. At any given time the most effective com- 506 m 870 149.4
bination of the various energy-producing pathways will occur and,
again, it should be emphasized that no form of exercise by the horse 800 m 920 191.9
is entirely aerobic or anaerobic in nature. 1025 m 846 129.3
At the initiation of exercise, the immediate source of energy is
locally available ATP (McMiken, 1983). As previously noted, this 1200 m 960 126.5
supply of ATP is limited and rapidly depleted, so for exercise to 1600 m 756 66.5
continue ATP must be replenished by other processes. Creatine
phosphate and the phosphocreatine pathway provide the next 3620 m 684 18.8
rapidly available ATP source; however, this energy supply is also of
limited capacity. The myokinase reaction provides a further means Note: At speeds greater than 500–600 m/min, intensity of exercise is such that
aerobic phosphorylation is unable to meet all the energy needs of the working
of regenerating ATP but it is restricted to certain muscle fiber types
muscle; therefore anaerobic glycolysis plays a major role in the supply of energy.
and is used in anaerobic exercise only when these fibers are recruited As a result, glycogen utilization increases dramatically at these and faster speeds.
(Meyer et al., 1980). The myokinase reaction (and associated NH3
*Millimoles glucose units per kg muscle per min (dry weight).
production) is believed to have a minor role in energy production
†
overall. Millimoles glucose units per kg muscle per min (wet weight). Approximate figures
for dry weight values are presented in parentheses.
The glycolytic pathway with the production of pyruvate and
lactate provides the main ongoing ATP source and reaches peak Adapted and reprinted from Hodgson, D.R. (1985) Energy considerations during
metabolism within about 30 s of the onset of exercise (McMiken, exercise. Vet. Clin. N. Am.: Equine Pract. 1 (3), 447–460, with permission from
Elsevier.
1983). The delay in maximal glycolytic output is possibly due to
the multiple and complex reactions required (McMiken, 1983). The
large stores of glycogen in equine muscle (Lindholm & Piehl, 1974;
Lindholm et al., 1974; Lindholm, 1979; Nimmo & Snow, 1983; during endurance rides where it is well recorded that blood concen-
Hodgson et al., 1984) allow this pathway to provide an early con- trations of NEFA increase and the glycogen utilization rate is low
sistent source of energy, but still there is a finite limit to this sub- (Lucke & Hall, 1980; Snow et al., 1982; Hodgson et al., 1983;
strate. On the other hand, this large store of glycogen in equine Hodgson, 1985) (Tables 18.1 and 18.2).
muscle might explain the fact that ketogenesis is a very limited As exercise intensity increases, ADP accumulates and this stimu-
pathway of energy production in the equine species (van der Kolk, lates anaerobic glycolytic energy production with a dramatic
1995). increase in the use of CHO substrates (see Table 18.1). Galloping
Aerobic mechanisms for ATP replenishment represent the most and bursts of intense exercise, such as during polo and jumping,
efficient means of substrate production. However, it is the slowest rely heavily on anaerobic energy supply. The self-limiting nature of
pathway to respond to exercise demands, owing to the cardiovascu- anaerobic power output (substrate exhaustion) means the horse can
lar lag in supplying O2 to the cells and the intricacy of the reactions. only maintain maximal speed for about 600–800 m. After this
Oxidative processes are in full production approximately 1 min distance, energy supply falls back to the slower aerobic pathways,
after the onset of exercise and then muscular energy is more likely necessitating a reduction in speed of exercise (McMiken, 1983;
to be dependent on the rate of oxygen transport to the cells rather Hodgson, 1985).
than substrate availability (McMiken, 1983).
At rest and during low-intensity exercise (walking and trotting),
aerobic pathways provide most energy requirements after the initial
Energy substrates
lag period (Åstrand & Rodahl, 1986). At this exercise intensity the
ratio of ATP to ADP will be high; PFK will therefore be inhibited Scientific data on the relationship between nutrition and equine
and β-oxidation of fatty acids will provide the main method for ATP performance has occupied the attention of many researchers;
regeneration (Hodgson, 1985; Lawrence, 1990). Such is the case however, it is difficult to design controlled experiments that isolate

423
 18 Metabolic energetics

equine muscle. The primary dietary sources of energy stores for the
Table 18.2  Mean plasma non-esterified fatty acid (NEFA)
horse are soluble and fiber CHOs and fats. Protein is considered to
concentration in horses performing various athletic activities
play a minor role as an energy source.

Type of exercise Mean plasma NEFA

concentration (µmol/L) Carbohydrates
80 km endurance ride Absorbed CHO is immediately available as an energy source in the
form of blood glucose. Muscle and liver provide the reservoirs for
Pre-exercise 47
excess CHO where it is stored as glycogen. Numerous studies have
Post-exercise 1254 documented the depletion of muscle glycogen stores that occurs
with exercise in the horse (Lindholm et al., 1974; Lindholm, 1979;
3-day event cross-country with phases A, B, C and D Hodgson et al., 1983; Nimmo & Snow, 1983). The rate of, and
Pre-exercise 156 percentage, depletion that results is a function of the intensity and
duration of exercise. Muscle glycogen utilization/min is greatest at
After steeplechase and roads and 586 the faster speeds over shorter distances (Nimmo & Snow, 1983;
tracks section Hodgson et al., 1984) (Table 18.1) but the total percentage of gly-
cogen depletion increases with increasing duration of exercise
After cross-country section 324
(Snow et al., 1981, 1982; Hodgson et al., 1983). Liver glycogen
Galloping (1.2 km) stores are also depleted significantly during exercise (Lindholm
et al., 1974; Lindholm, 1979). Restoration of depleted muscle gly-
Pre-exercise 246 cogen stores takes up to 48–72 h in the horse (Davie et al., 1995;
Post-exercise 279 Lacombe et al., 2004), which is 2–3 times slower compared to man
(up to 24 h). Recent research suggests that the low rate of glycogen
Adapted and reprinted from Hodgson, D.R. (1985) Energy considerations during synthesis in horses may be due to a reduced rate of absorption of
exercise. Vet. Clin. N. Am.: Equine Pract. 1 (3), 447–460, with permission from CHO from the gastrointestinal tract, combined with reduced skel-
Elsevier. etal muscle glucose uptake compared to man (Waller & Lindinger,
2010). The rate of skeletal muscle glycogen uptake is determined by
several factors among them the amount of initial muscle glycogen,
carbohydrate availability, glucose transport into muscle and activity
Table 18.3  Estimation of energy stores for a 500-kg horse with a
of glycogen synthase (GS).
muscle mass of 206 kg (approximately 55 kg being locomotor
In man, glycogen synthesis in muscle post exercise occurs in two
muscles), adipose tissue of 25 kg and a liver of 6.5 kg
phases. During the first hour of recovery, there is a rapid insulin
independent phase stimulated by glycogen depletion and contrac-
Fuel Energy tion stimulated GLUT-4 translocation, followed by a slower insulin
dependent phase with marked increased muscle insulin sensitivity
kJ kcal
up to 48 h (Price et al., 1994). Post exercise increases in insulin
ATP 38 9 sensitivity facilitate rapid muscle glycogen synthesis after glycogen
depleting exercise in man. However, in the horse no increase in
Creatine phosphate 188 45 insulin sensitivity could be found 24 h after a single bout of exercise
Glycogen 75 300 17 988 in trained horses (De Graaf-Roelfsema et al., 2006; Pratt et al.,
2007). In addition, no increase in crude muscle membrane GLUT-4
Fat 640 000 152 889 content was found after exercise inducing 40–50% glycogen deple-
tion (Pratt et al., 2007; Nout et al., 2003).
Adapted and reprinted from McMiken, D.F. (1983) An energetic basis of equine
Glycogen synthase (GS) is activated by insulin binding to its cell
performance. Equine Vet. J. 15 (2), 123–133, with permission from the Equine
Veterinary Journal.
surface receptor and the rate of GS synthesis is inversely proportional
to muscle glycogen in contraction- and insulin-stimulated muscle in
horses (Pratt et al., 2007) although less marked than in man.
Although single bouts of exercise does not seem to influence
nutritional influences on performance. Subtle, yet important effects
glycogen replenishment rate, training and overtraining have been
of nutritional alterations may go undetected partly because the
shown to influence peripheral insulin sensitivity. Several studies
power of statistical studies is limited by the small numbers of horses
describe an increase in insulin sensitivity after training horses for a
often used in experiments (Hintz, 1994). The relationship between
short period (Pratt et al., 2006, Stewart-Hunt et al., 2006). Also
energy and exercise is complex and inseparable. The amount
overtrained horses have been shown to increase insulin sensitivity
of energy required depends on the type and duration of activity
(De Graaf-Roelfsema et al., 2012). The underlying mechanisms are
and the horse’s body weight. Maintenance digestible energy (DE)
not completely understood yet.
requirements are linearly related to body weight (Pagan & Hintz,
1986a). During submaximal exercise energy expenditure is expo-
nentially related to speed and proportional to the body weight of
the riderless horse or the combined weight of the horse plus rider
Fat
(Pagan & Hintz, 1986b). The method used by Pagan and Hintz Assimilated fats are stored as triglycerides (uncharged esters of glyc-
(1986b) for calculating energy requirements was based only on the erol) in adipose tissue and muscle. Quantitatively, adipose tissue
amount of work performed and may not account for any follow-on constitutes the largest energy store in the body (see Table 18.3).
demands for energy in recovery that the work bout stimulates (Law- Triglyceride concentrations in muscle are considerably less than in
rence, 1990). fat (Lindholm, 1979). The triglycerides are highly concentrated
The stores of major fuels in the horse for muscular contraction stores of energy because they are reduced and anhydrous (Stryer,
are outlined in Table 18.3 as calculated by McMiken (1983). It is 1988). The initial event in the utilization of triglycerides as
clear that ‘fast’ energy stores (i.e. ATP, creatine phosphate and gly- an energy substrate is their hydrolysis by lipases to glycerol and
cogen) are limited despite the high capacity for glycogen storage in free fatty acids (FFA). Lipolysis is stimulated by epinephrine,

424
 Energy substrates

Starch/glucose insulin catecholamines lipolysis glycogenolysis endurance and speed?

Fats citrate phosphofructokinase glycogenolysis

Glucose during exercise glycemia, extends endurance time?

Fig 18.3  Effect of consumption of carbohydrate or fat immediately before or during exercise on energy metabolism.
Reprinted from Frape, D.L., (1988) Dietary requirements and athletic performanceof horses, Equine Vet J, 20 (3), 163–172, with permission from the Equine Veterinary Journal.

norepinephrine, glucagon and adrenocorticotrophic hormone and and therefore earlier onset of fatigue. Free FA mobilization is also
inhibited by insulin via hormone sensitive lipase. Glycerol is con- inhibited by insulin. Frape (1988) summarized the effects of con-
verted in a number of steps to Gl-3-P, which is an intermediate in sumption of CHO or fat before or during exercise on metabolism,
both the glycolytic and gluconeogenic pathways. The FFAs undergo as depicted in Figure 18.3.
β-oxidation and enter the TCA cycle as previously described. Oleic, On the other hand, a lack of available CHO during submaximal
palmitic and linoleic acids (Robb et al., 1972; Westermann, 2008a) exercise can also limit performance and there is strong evidence
as well as 3-OH-butyric and 3-OH-iso-butyric acid (Westermann, supporting the use of high CHO diets by humans for endurance
2008a) represent the major FAs in the equine species. exercise (Lawrence, 1990). In humans, muscle glycogen loading was
Fat has been shown to be the major energy substrate during low- achieved by performing intense exercise and then consuming a
intensity exercise. This is best evidenced by a decrease in the respira- CHO-rich diet (Lindholm, 1979). Current practice is to combine a
tory exchange ratio (R) (McMiken, 1983; Pagan et al., 1987; Rose program of decreased activity with increased CHO consumption a
et al., 1991) and an increase in plasma NEFA concentrations (Lind- few days before competition to achieve a glycogen load. Glycogen
holm, 1979; Rose et al., 1980; Essén-Gustavsson et al., 1991) that loading in horses has been accomplished but no obvious improve-
occurs with prolonged submaximal exercise. R is calculated by divid- ment in work performance has been demonstrated (Topliff et al.,
ing the volume of carbon dioxide (CO2) expired by the volume of 1983, 1985; Frape, 1988; Lawrence, 1990; Snow, 1994). Intrave-
O2 consumed during exercise. R values around 0.7 indicate fat utili- nous, but not oral, glucose supplementation has increased glycogen
zation whereas for CHO utilization the value is 1.0. Values within repletion rates after exercise (Snow et al., 1987; Davie et al., 1994,
this range reflect various mixtures of FA and CHO metabolism. 1995; Snow, 1994). Although glycogen loading is not recom-
When anaerobic metabolism predominates R values will exceed 1.0 mended in the horse adequate CHO intake must still be ensured
because lactate production is high, thereby adding to the CO2 load (Hintz, 1994). A low CHO diet and regular exercise leads to glyco-
to be eliminated. In addition, in the equine species carnitine forms gen depletion and decreased performance in horses (Topliff et al.,
esters predominantly with short- and medium-chain fatty acids in 1983, 1985). In a study where fit Standardbreds were exercised
plasma. Submaximal exercise in Standardbreds mainly affected con- strenuously for 3 consecutive days to achieve a 55% depletion of
centrations of short-chain acylcarnitines (Westermann, 2008a). the muscle glycogen store, anaerobic, but not aerobic, capacity was
impaired (Lacombe et al., 1999). However the association between
glycogen depletion and impaired anaerobic metabolism is not con-
Protein clusive as confounding effects of other exercise-induced changes on
performance could not be eliminated (Lacombe et al., 1999). When
Digested protein is absorbed from the small intestine as amino muscle glycogen was depleted by 22% there was no significant effect
acids and small peptides. When amino acids are available in excess on performance of Thoroughbreds exercising at high intensities
of the animal’s requirements they may be broken down to provide (Davie et al., 1996).
energy. Degradation by deamination or transamination reactions A CHO supplement taken an hour or two before exercise does
occurs mostly in the liver, with the final product being acetyl CoA not seem to benefit endurance performance but intake of glucose
for utilization in the TCA cycle. Leucine, a branched-chain amino during exercise may supplement waning plasma concentrations and
acid, may undergo oxidation directly in muscle (Lawrence, 1990). delay onset of fatigue (Lawrence, 1990) (see Fig. 18.3).
The contribution of amino acids to energy production during exer- The beneficial effects of feeding high-fat diets to horses remains
cise is minor compared to that of CHO and FA (Åstrand & Rodahl, shrouded in controversy. Differences in the condition of the horses,
1986), perhaps in the range of 1–15% (Lawrence, 1990). High type of exercise, the length of the adaptation period to the diets, the
protein diets (up to 16%) were once thought necessary to sustain type of fat used as the supplement, and the level of fat supple-
the performance of mature equine athletes but now it is considered mented, particularly in relation to CHO, make comparing the
that approximately 10% protein in the diet is adequate (Snow, published results difficult (Lawrence, 1990; Hintz, 1994). Many
1994). The main amino acids in equine plasma are serine, gluta- variations in study designs influence results obtained. Feeding an
mine, proline, glycine, alanine, and valine. Training of Standard- increased level of fat is suggested to cause metabolic adaptations
breds reduced plasma concentrations of aspartic acid (Westermann, like increase in equine heparin-released lipoprotein lipase activity
2010). (Geelen et al., 2001) that permits horses to preferentially utilize fat
and spare glycogen during exercise but the evidence to support such
a proposal is inconclusive (Lawrence, 1990; Hintz, 1994; Snow,
Effects of dietary alterations on energy 1994; Geelen et al., 2000). Furthermore, it should be realized that
dietary soyabean oil depresses the apparent digestibility of fiber
substrate utilization when substituted for an iso-energetic amount of cornstarch or
Many published reports have described the effect that altering com- glucose (Jansen et al., 2002).
ponents of the normal diet has on substrate utilization and perfor- It has been suggested that improvements in oxidative capacity can
mance in the horse. The consumption of large amounts of digestible be brought about by certain amino acid supplements (Lawrence,
CHO within a few hours of strenuous activity may depress the 1990; Hintz, 1994). Higher than necessary protein diets are often
performance of that exercise (Åstrand & Rodahl, 1986). This is fed to performance horses but studies to indicate that this practice
possible because insulin-stimulated uptake of blood glucose results enhances exercise capabilities are lacking (Lawrence, 1990). On
in hypoglycemia and a greater dependence on muscle glycogen CHO-rich and fat-rich diets, plasma concentrations of glucose,

425
 18 Metabolic energetics

ammonia, lactate, alanine and the muscle concentrations of G-6-P Atmospheric O2 Altitude
and lactate were higher at the end of exercise compared to normal
diets (Essén-Gustavsson et al., 1991). Higher pre-exercise muscle
glycogen concentrations and FFA concentrations were present in the Airway resistance
O2 transport via
horses fed a CHO-rich diet when compared to the fat-rich and Ventilation rate
airways to alveoli
normal-diet fed periods. No significant difference in performance Tidal volume
during trotting at submaximal intensity on a horizontal treadmill
was detected between the three diets (Essén-Gustavsson et al., 1991) P(A-a)O2 difference
with the average time to fatigue being 51–56 min. Whether or not O2 diffusion across
Ventilation/
alveolar/capillary interface
the diets would alter performance in shorter or longer exercise perfusion ratio
periods remains unanswered. The effects on protein metabolism
need to be further investigated as both the CHO-rich and the fat-
rich diets were associated with significant increases in branched- Hematocrit O2 binding to
chain amino acids in the plasma during and at the end of exercise Hb concentration hemoglobin
compared to the normal diet (Essén-Gustavsson et al., 1991). The
resting plasma concentration of the branched-chain amino acids Heart rate
was increased 26% on the fat-rich diet but only 8% on the CHO- O2 distribution
Stroke volume
rich diet. via the circulation
Blood volume
A 9% (control) or 18.5% (high) crude protein diet had no effect
on hepatic or muscular glycogen utilization and did not affect
exercise performance in Quarter Horses exercising at submaximal O2 uptake
Capillarization
intensities (Miller-Graber et al., 1991). Performance of Arabian by muscle
endurance horses was not augmented by excessive protein in
their diet (Hintz, 1983). In contrast, Standardbreds fed a high-
Oxidative enzymes
protein diet (20%) or high-fat diet (15% soybean oil) showed O2 utilization
Mitochondrial
greater muscle and liver glycogen utilization during prolonged exer- by mitochondria
density
cise compared to when fed a control diet of 12% crude protein
(Pagan et al., 1987). During higher-intensity, shorter-duration exer-
Fig 18.4  The oxygen transport chain, indicating the steps in the transfer of
cise, glycogen utilization was less when horses were fed the high-
oxygen from the inspired air to the final utilization by the mitochondria.
protein or high-fat diets. Of interest, there was no difference in
Physiological variables that will influence the capacity of the oxygen
concentrations of various plasma amino acids in overtrained Stan- transport chain are listed beside each step they affect. Hb, hemoglobin;
dardbreds and age-matched controls (Westermann, 2010). P(A–a)O2, alveolar – arterial oxygen pressure.
The timing of feeding and what to feed before exercise has con-
siderable influence on the metabolic and physiological responses
to exercise (Lawrence et al., 1995; Harris & Graham-Thiers, 1999).
In one study, it was concluded that feeding only hay shortly before
exercise would not adversely affect performance but feeding grain
would, and that therefore grain should be withheld (Pagan & Oxygen uptake at rest and during submaximal exercise
Harris, 1999). At rest, V.O2 is in the order of 3–5 mL/kg/min or 1.5–2.5 L/min for
Of course, many other nutritional components not discussed in a 500-kg horse (Thornton et al., 1983; Eaton, 1994). It can be dif-
this chapter may play roles in equine performance (Marlin & Nan- ficult to accurately obtain a basal V.O2 prior to exercise as often
kervis, 2002; Hinchcliff et al., 2004). These include water, electro- horses are excited in anticipation of impending activity. Therefore
lytes, acid–base balance, minerals and vitamins. a resting V.O2 level of 2 mL/kg/min may be more realistic (Eaton,
1994).
During submaximal exercise a number of factors will influence
Energy expenditure the level of V.O2, including speed of exercise, load being carried,
degree of incline on which exercise is being performed, duration of
exercise, thermoregulation and track surface.
Aerobic power There is a well-established linear relationship between V.O2 and
the speed of exercise at submaximal intensities in horses (Hoyt &
Oxygen uptake Taylor, 1981; Hörnicke et al., 1983; Evans & Rose, 1987, 1988a;
The oxygen consumed by the body at a given time is a measure of Rose et al., 1990a; Eaton, 1994) and humans (Åstrand & Rodahl,
the body’s total aerobic metabolic rate and is termed the oxygen 1986). When speed increases such that V.O2max is approached, this
uptake (V.O2). Units of measurement are usually milliliters of linear relationship is lost as V.O2 plateaus and anaerobic sources of
oxygen per kilogram of body weight per minute (mL/kg/min) or energy production become significant. In addition, if horses exercise
liters per minute (L/min), therefore representing a rate of consump- at unnatural (extended or restricted) gaits the linear relationship
tion and not a finite capacity. The maximum rate of oxygen uptake will be lost due to a loss in economy of locomotion (Hoyt & Taylor,
is called the V.O2max. Oxygen consumption by the body is princi- 1981; Eaton, 1994; Preedy & Colborne, 2001; Wickler et al., 2001;
pally a function of the cardiorespiratory system to supply oxygen Griffin et al., 2004).
and the capacity of end organs to utilize oxygen. The sequence of Few equine sports are performed without the horse carrying an
events is described as the oxygen transport chain (Fig. 18.4). It is extra load in the form of a rider or driver. Oxygen consumption (or
influenced by the O2 concentration in the air, ventilation of the energy expenditure) increases in proportion to the load carried
lungs, diffusion of O2 through the alveolar wall, circulatory perfu- (Taylor et al., 1980; Pagan & Hintz, 1986b; Thornton et al., 1987;
sion of the lungs and affinity of hemoglobin (Hb) for O2, distribu- Gottlieb-Vedi, et al., 1991).
tion of O2 to the periphery by the circulation, extraction by the end Taylor and colleagues (1980) reported that when a 10% load was
organ (muscle) and, finally, O2 utilization by the mitochondria. A added to the horse when trotting, V.O2 increased approximately
large number of physiological variables contribute to the capacity 10% and this direct proportionality was consistent for loads between
of the oxygen transport chain. 7% and 27% of the horse’s body mass. A direct proportionality

426
 Energy expenditure

between load and V.O2 was also demonstrated for trotting rats, trot- and electrolyte losses in the sweat, contributing to thermoregulatory
ting and galloping dogs and running humans (Taylor et al., 1980). and circulatory problems, would be key factors in the horse’s ability
As a consequence, small animals use more oxygen and expend more to perform endurance activity (Rose & Evans, 1986). Naturally, the
energy to carry each gram of a load a given distance than do large intensity of exercise will be a determinant of the duration of any
animals, be it their own body mass or an additional load carried. activity.
Pagan and Hintz (1986b) demonstrated that a 450-kg horse with a The effect of temperature on V.O2 will be a consequence of any
50-kg rider would expend the same amount of energy as a 500-kg impedance that altered thermoregulation may have on energy
horse. Thornton and colleagues (1987) found no significant differ- demands. Redistribution of cardiac output to skin for heat dissipa-
ence in the oxygen cost per kilogram per meter traveled between tion, fluid shifts and metabolic disturbances may all contribute to
loaded and unloaded horses. The increase in V.O2 due to load is a less-efficient oxygen transport chain and therefore diminished
achieved largely by an increase in ventilation until maximum tidal performance. The optimum temperature range for oxygen utiliza-
volume is approached (Thornton et al., 1987) and this is readily tion has yet to be established.
explained by the close and linear relationship between V.O2 Track surfaces may affect the economy of locomotion due to
and pulmonary ventilation (Hörnicke et al., 1983; Gottlieb-Vedi altered stride patterns (change in frequency and length of stride) in
et al., 1991). slippery, uneven or ‘heavy’ conditions. Quantifying track effects
The implications for racing performance should be considered. on energy expenditure is difficult but Thoroughbreds and endur-
A horse of less mass will expend proportionally less total energy to ance horses have longer race times in heavy conditions (Eaton,
move the same distance when compared to a heavier horse. 1994). Different treadmills also influence the energetic cost of
The degree of incline on which exercise is being performed has a locomotion: it costs less energy for horses to walk, trot or canter
significant impact on V.O2. For Standardbreds, trotting on a 6.25% on a stiffer treadmill than on a more compliant treadmill (Jones
inclined treadmill at an average speed of 5.2 m/s, V.O2 increased et al., 2006).
from a mean of 17.7 L/min on the flat to 31.1 L/min on the slope It has been reported that major horse race times and records
(mean change of 13.4 L/min; 76%, p <0.001) (Thornton et al., improved by 5–7% around 1900 when jockeys adopted a crouched
1987). The addition of a load when doing the inclined exercise did posture. When animals carry loads, there is a proportionate increase
not significantly add to the oxygen cost of the exercise. Thorough- in metabolic cost, and in humans this increase in cost is reduced
breds exercising on a treadmill at speeds of 1–13 m/s also showed when the load is elastically coupled to the load bearer. Pfau et al.,
a substantial increase in V.O2 when the treadmill slope was elevated (2009) showed that jockeys move to isolate themselves from the
from 0–5% and 10% (Eaton et al., 1995a). Exercising on a 10% movement of their mount, which would be difficult or impossible
slope can double the energy expenditure at some speeds. with a seated or upright, straight-legged posture. This isolation
When trotting on an inclined treadmill over a range of speeds, means that the horse supports the jockey’s body weight but does
V.O2max is higher during inclined than level running (McDonough not have to move the jockey through each cyclical stride path. This
et al., 2002) and a greater volume of muscle would have to be posture requires substantial work by jockeys, who have near-
recruited to generate an equivalent force for body support, which maximum heart rates during racing.
is reflected in significant increases in the EMG intensity (IEMG) of
muscles (Wickler et al., 2005). In addition, normal training in Stan-
dardbreds resulted in a significant adaptation of quantitative needle Maximum aerobic power
electromyography (QEMG) parameters. Compared with normal When V.O2 no longer increases despite an increase in workload, the
trained controls, intensively trained Standardbreds showed a stron- horse is defined as having reached V.O2max. This value represents the
ger adaptation (e.g. higher amplitude, shorter duration and fewer ‘gold standard’ measure for maximum or peak aerobic power. Thor-
turns) in QEMG variables resembling potentially synchronization oughbred horses have mean V.O2max values around 150–170 mL/kg/
of individual motor unit fiber action potentials (Wijnberg et al., min (Evans & Rose, 1987; Rose, et al., 1990a), easily twice the values
2008). for elite human athletes (69–85 mL/kg/min), 1.5 times those of
Wickler et al. (2004) found that the costs of swinging the limbs greyhounds (100 mL/kg/min) and 3 times those of racing camels
in the horse are considerable and the addition of weights to the (51 mL/kg/min) (Derman & Noakes, 1994). Oxygen consumption
distal limb can have a profound effect on not only the energetics can be calculated by the following equation:
of locomotion but also the kinematics, at least in the hind limb. V.O2 = CO i (a-v )O2
Thus, they proved that the use of weighted shoes, intended to
increase animation of the gait, increases the metabolic effort of where CO = cardiac output (heart rate • stroke volume) and (a-v)
performance horses a disproportionate amount. The additional O2 is the arterio-mixed venous oxygen difference.
mass also increases the joint range of motion and, potentially, the The horse’s tremendous ability to achieve a higher V.O2max than
likelihood of injury. other athletic species is related to its massive heart rate response
The terrain of endurance rides and cross-country tracks in three- and ability to substantially augment its circulating red blood cell
day events ensure that much work up and down gradients will be mass, and therefore oxygen-carrying capacity, during exercise
performed and this will play a large role in determining energy (Thomas & Fregin, 1981). Trends indicate that top human athletes
expenditure. Little investigation has been done regarding the effect (e.g. runners, cyclists and cross-country skiers) will generally have
of a downhill gradient on energy expenditure in horses but in higher V.O2max values; however, there is considerable variation
humans the energy cost of moving down a slope decreases up to a between athletes of similar ability (Derman & Noakes, 1994). In
certain steepness and then becomes more expensive compared to horses, a positive correlation between running speed and V.O2max
level exercise (Åstrand & Rodahl, 1986). has been described and this correlation became stronger as the
The effect that duration of exercise has on V.O2 has not been distance ran increased with an increase in distance run (Harkins
frequently investigated. Rose and Evans (1986) monitored cardio- et al., 1993). It was suggested that faster horses utilize more oxygen
respiratory and metabolic alterations during 90 min of submaximal during maximal exercise intensity.
exercise in Standardbreds. The horses trotted on a slope of 2% at
3 m/s. Many of the respiratory variables measured, including V.O2,
reached a steady state within 5 min of the start of exercise and
Anaerobic power
remained stable for the duration of the exercise period. Oxygen Anaerobic energy supply becomes significant when exercise inten-
consumption from 5 min onwards did not alter significantly until sity is at a level beyond that which aerobic pathways can accom-
a slight decrease was identified at 90 min. It was proposed that fluid modate alone. The faster glycolytic pathways may be recruited under

427
 18 Metabolic energetics

250
Muscle glycogen
concentration VO2 demand
Percentage of Muscle
type II fibers buffering capacity 200
VO2 max
Anaerobic

VO2 (ml/kg/min)
capacity 150

Muscle Rate of
fiber area Muscle concentration glycogenolysis 100
of high-energy phosphates

Fig 18.5  Major factors contributing to anaerobic capacity. 50

0
0 2 4 6 8 10 12 14 16 18 20
A Speed (m/s)
two conditions: when energy demand increases so rapidly that the
slower aerobic systems cannot match the supply rate required; or
when the total energy demand exceeds what the aerobic pathways
are capable of supplying at peak capacity. Workloads at intensities 250
beyond that provided for by V.O2max have been referred to as supra-
maximal intensities. This level of energy utilization is experienced
by racing Thoroughbreds, Standardbreds and Quarter Horses during 200 Anaerobic capacity
competition. Shorter duration races, e.g. Quarter Horse 400-m
sprints, rely predominately on the rapid supply of energy by anaero-

VO2 (ml/kg/min)
bic means (Eaton, 1994). Anaerobic power is considered a finite 150
capacity and not a rate because the supply of substrates for anaero-
bic phosphorylation is limited. Factors that influence anaerobic 100 Aerobic contribution
capacity are depicted in Figure 18.5.
Theorizing that anaerobic capacity is a function of the area of
type II fibers in the locomotor muscles, McMiken (1983) stated that 50
to measure maximal anaerobic capacity one should calculate the
type II fiber area and the activities of anaerobic pathway enzymes
in the muscle. Maximum accumulated oxygen deficit (MAOD) has 0
0 30 60 90 120 150
been investigated as a measure of anaerobic capacity in horses B Time (s)
(Eaton et al., 1992, 1995b; Eaton, 1994) following preliminary
studies indicating its usefulness in humans (Mebø et al., 1988; Scott Fig 18.6  Determination of maximum accumulated oxygen deficit (MAOD).
et al., 1991). (A) Initially, the V.O2 versus speed plot is generated by performing a
Oxygen deficit refers to the deficiency in V.O2 that occurs at the standardized incremental exercise test on an inclined treadmill. The V.O2max
commencement of exercise until the responding cardiorespiratory is 175 mL/kg/min, and for this horse to exercise at an intensity of 125% of
system meets the oxygen demand of the tissues (Åstrand & Rodahl, V.O2max, by extrapolation it can be seen that the V.O2 demand would be
1986). The total oxygen deficit that accumulates during exercise at 219 mL/kg/min. To exercise at this supramaximal intensity the horse would
supramaximal intensities is the MAOD and this is the difference need to run at 15 m/s. (B) This figure demonstrates the relationship
between the oxygen demand and the actual V.O2 achieved. The O2 between V.O2 and time for the horse exercising at 125% of V.O2max. The
demand is calculated by extrapolating from the linear relationship previously calculated O2 demand is drawn in as the dotted line at 219 mL/
between V.O2 and speed at submaximal intensities (Fig. 18.6). To kg/min. At the onset of the exercise there is a lag in V.O2 but it quickly
determine MAOD, horses on a treadmill are rapidly accelerated to reaches V.O2max. The exercise ceases when the horse can no longer keep
pace with the treadmill. The difference in the O2 demand and the actual
speeds equivalent to supramaximal intensities (defined as a percent-
oxygen uptake is defined as the MAOD and is a measure of the anaerobic
age of V.O2max measured in a previous exercise test and extrapolated
capacity.
from the V.O2 versus speed plot) (Rose et al., 1988; Eaton et al.,
Adapted from Eaton (1994).
1995b). The V.O2 is measured at frequent intervals until the horse
fatigues. The area between the O2 demand and the V.O2 curve is the
MAOD (Eaton et al., 1995b) (Fig. 18.6).
In humans, the power vs time-to-fatigue (P:TTF) relationship has
For exercise intensities requiring 105–125% V.O2max the MAOD
been used as an accepted method for assessing anaerobic work
was similar at 31 mL O2 equivalents per kg of bodyweight but the
capacity and this relationship has now been investigated in horses
proportion of energy supplied by anaerobic processes increased
(Lauderdale & Hinchcliff, 1999). In humans, the relationship is best
from 14 to 30% (Eaton et al., 1995b). V.O2max was not correlated to
described by the hyperbolic equation:
MAOD, suggesting that anaerobic capacity is unlikely to be depen-
dent on the rate of oxygen uptake. Eaton and colleagues (1995b) t = W9/(P 2ØPA )
proposed from their results that anaerobic energy supply would where t is the time to fatigue (s); P is power (watts); ØPA is power
contribute less than 30% of the total energy input in Thoroughbred asymptote, or critical power, which represents the maximum sus-
and Standardbred races, which is considerably lower than previ- tainable power output or anaerobic threshold, and W9 is a constant
ously suggested (Bayly, 1985). Using peak blood or plasma lactate representing anaerobic capacity or the finite amount of work that
concentrations as an indicator of anaerobic capacity appears limited can be performed above ØPA (Fig. 18.7). Similarly to humans, the
because of the many variables like function of monocarboxylate P:TTF relationship in Standardbreds is best represented by a hyper-
transporters (MCT) as lactate carriers (Koho et al., 2006) that affect bolic function; however, the technique needs to be validated against
lactate concentrations including rates of flux between fluid the more traditional MAOD measure of anaerobic capacity before
compartments. its usefulness in horses as a predictor of fitness and anaerobic

428
 Energy expenditure

or oxygen debt. The EPOC may only account for a small fraction of
the net total oxygen cost (NTOC) of exercise. In humans exercising
at 30–70% of V.O2max for up to 80 min the EPOC was only 1.0–
8.9% of the NTOC of the exercise (Gore & Withers, 1990). In the
only comprehensive study to date in horses, Rose and colleagues
(1988) measured oxygen debt as the area under the O2 recovery
Time-to-fatigue (s)

curve following a bout of exhaustive exercise at an intensity equiva-

lent to 120% of V.O2max. Oxygen debt represented nearly 52% of
the NTOC, which is dramatically higher than that in the previously
quoted study of humans. This can be attributed in part to the very
different exercise intensities performed in the two studies and the
relative fitness of the subjects.
EPOC is considered to have an initial fast phase and then a slower
phase. In the horse these phases were complete by 1.4 and 18.3 min,
respectively, after supramaximal exercise (Rose et al., 1988). The fast
ØPA
phase is associated with the resaturation of myoglobin and hemo-
globin and the replenishment of the high-energy phosphagen pool
Power (W) (CP and ATP). Perhaps less than 1.5% of the EPOC was required to
restore the muscle CP pool in the horse and this occurred at a slower
Fig 18.7  Schematic of the hyperbolic relationship between power (P) and pace than may have been expected (Rose et al., 1988). Post-exercise
time-to-fatigue (t) represented by the equation t = W′/(P2ØPA). W′, anaerobic tachycardia and tachypnea would also contribute small compo-
capacity; ØPA, critical power; power units = watts (W). nents to the EPOC because of increased consumption of O2 by the
myocardium and respiratory muscles until resting levels are reached.
The slow phase is associated with the oxidation of lactate that accu-
mulates during exercise. However, not all the lactate that is metabo-
capacity can be investigated (Lauderdale & Hinchcliff, 1999). The lized is accounted for by the EPOC and some is utilized in
P:TTF does not require collection of respiratory gases or blood for gluconeogensis and amino acid synthesis. A poor relationship
its calculation and this may be an advantage over the more intensive existed between the restoration of muscle metabolites to pre-
effort required to measure blood lactate concentrations or deter- exercise levels and the recovery of V.O2. Muscle and plasma lactate
mine MAOD. However, currently a high-speed treadmill and mul- concentrations remained elevated after 60 min of recovery whereas
tiple high-intensity exercise tests are still a prerequisite, as the ability V.O2 had returned to near pre-exercise levels (Rose et al., 1988).
to calculate the P:TTF relationship in field trials is yet to be A number of other factors considered associated with EPOC,
determined. including exercise-induced hyperthermia, have not been quantified
For any measure of anaerobic capacity in the horse, it remains to in the horse.
be examined what relationship exists between performance and
anaerobic capacity.
Economy of locomotion
The economy of locomotion refers to the net energy cost in mL O2 per
Anaerobic threshold kg of body weight per meter traveled (mL O2/kg/m). It is indepen-
Anaerobic threshold is defined as the level of work or V.O2 consump- dent of speed and load (or body weight) (Eaton, 1994). There are
tion just below that at which metabolic acidosis and the associated conflicting reports on the values for economies of locomotion but
changes in pulmonary gas exchange occur (Wasserman et al., 1973). the importance of gait in these studies needs to be recognized
It represents the transition when anaerobic means of energy supply (Eaton, 1994). Horses on treadmills may be forced to work at
becomes important during exercise. In humans, this level of work defined speeds, and in doing so, utilize extended or restricted gaits
has been correlated with a blood lactate concentration of 4 mmol/L that might alter the true cost of locomotion they would otherwise
(Åstrand & Rodahl, 1986). As exercise intensity increases, lactate naturally incur if allowed to control their own pace. Thornton and
accumulation in the circulation rises in an exponential manner. colleagues (1987) have reported values of 0.122 mL O2/kg/m for
Hence the anaerobic threshold has also been described as the onset speeds of 4.5–6.25 m/s and 0.124 mL O2/kg/m for speeds of 6.5–
of blood lactate accumulation (OBLA). The velocity or intensity of 8.14 m/s obtained on a horizontal plane. These results are similar
work at which a blood lactate concentration of 4 mmol/L (VLA4) is to the 0.133 mL O2/kg/m that can be derived from the results of
reached has been used to assess the relative fitness of horses and Taylor and coworkers (1980) for a 119-kg pony working at 3.11 m/s.
humans and their response to training (Persson, 1983; Thornton Eaton and colleagues (1995a) recorded a range of 0.10–0.16 mL
et al., 1983; Rose et al., 1990a; Auvinet, 1996). The VLA4 increases O2/kg/m at speeds of 5–13 m/s on a horizontal treadmill. Exercis-
with training (Thornton et al., 1983; Eaton et al., 1999) and in ing on a positively inclined treadmill will increase the oxygen cost
general, the higher the VLA4 the fitter the horse (Rose & Hodgson, of work about 2–2.5 times that measured on a flat plane, depending
1994a). on the steepness of the slope (Thornton et al., 1987; Eaton et al.,
Anaerobic threshold and OBLA are determined during an incre- 1995a). When averaging a range of speeds for each gait, Hörnicke
mental exercise test. Anaerobic threshold is identified by the point and colleagues (1983) reported higher values for the walk (0.21 mL
of non-linear increase in respiratory variables such as minute ven- O2/kg/m) and for the trot and gallop (0.19 mL O2/kg/m) for horses
tilation and carbon dioxide production (Wasserman et al., 1973). exercising on the track. By averaging the economies of a number of
It is assumed that this point is highly correlated with the OBLA but speeds, a higher oxygen cost can be expected as values will include
this may not be the case (Åstrand & Rodahl, 1986). less-efficient gait velocities than that at which the horses may be
required to exercise (Hörnicke et al., 1983).
It is well documented that horses will choose a gait at any speed
Postexercise oxygen consumption that results in the least possible expenditure of energy (Hoyt &
At the cessation of exercise, O2 continues to be consumed above Taylor, 1981; Eaton, 1994; Preedy & Colborne, 2001; Wickler et al.,
basal rates as it declines in an exponential manner to resting levels. 2001; Griffin et al., 2004). Ponies (110–170 kg) were trained to
This is referred to as excess postexercise oxygen consumption (EPOC) walk, trot and gallop, and to extend their gait on command on a

429
 18 Metabolic energetics

28 (2006) found it is confusing to distinguish between walking and

running by the duty factor (DF). When changing from pendulum
Walk
Trot
to spring mechanics, there is a change in the slope of metabolic rate
24 (MR) vs. speed. At the trot–gallop transition, where quadrupeds are
Gallop
hypothesized to change from spring mechanics to some combina-
20 tion of spring and pendulum mechanics, there is a change in slope
ml O2 consumed to move 1 m

of MR vs. speed in horses but not in other species. Stride frequency

(SF) is a logarithmic function of walking speed in all species, a
16 linear function of trotting/running speed, and nearly independent
15
of speed in galloping. In humans and horses there is a discontinuity
in SF at the walk-trot (run) transition. The slope of time of contact
12 vs. speed does not change with mechanics in most species, but it

No. of observations
10 does in humans. In horses and humans, there is a discontinuity at
8 the walk-trot (run) transition and data for other species do not
permit generalization. Duty factor (DF) in humans is >0.5 in
5 walking (pendulum mechanics) and <0.5 when running (spring
4 mechanics). However, this is not true in many species that have DF
>0.5 at the lowest speeds where they use spring mechanics, includ-
ing horses performing the tölt (Biknevicius et al., 2006). Appar-
0 0
0 1 2 3 4 5 6 7 ently, different energy-conserving mechanics (i.e. pendulum and
spring) used in different gaits reflected in differences in energetics
and/or stride parameters.
The effects of training and athletic activities on the economy of
Walk Gallop equine locomotion need to be considered. Endurance horses forced
Trot to use extended gaits for prolonged periods may fatigue more
rapidly than if they were to move at their natural speed for each
Running speed (m/s)
gait. Clearly, it becomes a matter of achieving the right balance of
Fig 18.8  Economy of locomotion. The oxygen cost to move a unit distance speed and energy consumption to complete the distances and be
(rate of oxygen consumption divided by speed) declined to a minimum and successful in such events. Standardbred trotters and pacers may be
then increased with increasing speed in a walk and trot. It also declined to a able to work at a wide range of speeds using a single gait without
minimum in a gallop but the treadmill did not go fast enough to observe any loss of economy (Thornton, et al., 1987; Eaton, 1994).
any increase at higher galloping speeds. The minimum oxygen cost to
move a unit distance was almost the same in all three gaits. The histogram
shows gaits when a horse was allowed to select its own speed while Fatigue
running over ground. The three speeds chosen coincided with the Fatigue is a complex and intricate physiological response to exercise,
energetically optimal speed for each gait. leading to the inability to sustain further activity at the current
From Hoyt, D.F., Taylor, C.R., Gait and the energetic of locomotion in horses. Reprinted by
intensity. Fatigue can be categorized as structural, acute or chronic.
permission from Macmillan Publishers Ltd, Nature Publishing Group, ©1981.
Structural fatigue refers to biomechanical failure of tissues, for
example tendons, ligaments and bone, which inadequately adapt
to the stresses placed upon them. Chronic fatigue is a function of
prolonged conditions such as chronic anemia and starvation. Acute
treadmill. Rates of oxygen consumption increased curvilinearly with fatigue is directly related to energy production in the muscle and
speed for walking and trotting. The maximum speed of the tread- occurs in events requiring maximal work effort for short periods,
mill prevented sufficient data from being obtained for galloping e.g. Thoroughbred or Standardbred racing. It has been labeled
velocities. Gait transitions occurred at speeds when the oxygen con- anaerobic fatigue (McMiken, 1983) and has different causal factors
sumption was similar for the two gaits, but when the ponies were to those that limit aerobic performance in endurance type events.
forced to exercise at an extended gait beyond the normal range of Fatigue appears to involve central (psychologic/neurologic) and
speeds, oxygen consumption was higher (Hoyt & Taylor, 1981). peripheral (muscular) contributions (Hodgson & Rose, 1994a).
Thus, there was a speed for each gait where the energy cost of loco- Overtrained horses can become listless and ‘sour’ with a decline in
motion was minimal and this cost was similar for the walk, trot and performance that may be partly a manifestation of psychological
gallop (Hoyt & Taylor, 1981) (Fig. 18.8). So, at the optimal speed fatigue. Rivero et al., (2008) differentiate overtraining syndrome
of each gait, the amount of energy consumed to move a given dis- (OTS) from over-reaching, a term used for horses that, after suffer-
tance is much the same. When a horse was allowed to move at its ing a loss of performance without an obvious clinical reason,
natural pace, it did so by selecting speeds within each gait around recover their performance within 1 or 2 weeks. When inadequate
the most energy efficient speed (Hoyt & Taylor, 1981). The optimal training stress is applied and recovery time is insufficient, perfor-
value for economy was similar to the 0.122–0.133 mL O2/kg/m mance reduction and chronic maladaptation occurs. Overtraining
values derived elsewhere for flat treadmill exercise (Taylor et al., syndrome (OTS) is known as a complex condition, which often
1980; Thornton et al., 1987). afflicts horses in top training. Peripheral causes of fatigue have been
Griffin et al. (2004) found that the absolute walk–trot transition studied more widely as they are easier to define. Recently, De Graaf-
speed increased with size, but it occurred at nearly the same Froude Roelfsema et al., (2009) induced overtraining (performance
number. In addition, horses spontaneously switched between gaits decreased nearly 20% compared with controls and did not improve
in a narrow range of speeds that corresponded to the metabolically after 4 weeks of detraining) by intensified training and found that
optimal transition speed. These results support the hypotheses that endocrinological and behavioral alterations occurred before periph-
the walk–trot transition is triggered by inverted-pendulum dynam- eral adaptations for example in the muscles (De Graaf-Roelfsema
ics and occurs at the speed that maximizes metabolic economy. Of et al., 2009).
interest, some overtrained Standardbreds made a transition to Fatigue in response to high-intensity exercise is likely due to a
canter rather than continue trotting during a standardized exercise combination of factors including depletion of the phosphagen
test (De Graaf-Roelfsema et al., 2009). Nevertheless, Hoyt et al., pool (ATP and CP), decreased intracellular pH, and possibly

430
 Training programs

accumulation of lactate (McMiken, 1983; Hodgson, 1985; Hodgson 1. A foundation phase

& Rose, 1994a; Essén-Gustavsson et al., 1999). The main event 2. A cardiovascular or aerobic phase
appears to be a reduction in the concentration of ATP. Acidity can 3. An anaerobic interval phase.
impair the respiratory capacity of muscle and have a direct effect on
The foundation phase is considered continuous or endurance-type
the contractile apparatus. In addition, acidosis and increased muscle
training and is vital for the cardiovascular and musculoskeletal
temperature could be associated with impaired sarcoplasmic reticu-
systems to make early adaptations to the stresses placed upon them.
lum function. Finally, altered electrolyte gradients (potassium and
Subsequent phases include more intense, shorter runs.
calcium) will add to the overall deleterious effects on muscle
Improvement in V.O2max is independent of submaximal training
metabolism.
intensity when horses are exercised for the same distance (Knight
During prolonged, submaximal exercise, hyperthermia, altered
et al., 1991; Eaton et al., 1999) and as little as 10–14 days of train-
fluid and electrolyte balance, and fuel depletion have all been con-
ing is all that is required to achieve an increase in aerobic capacity
sidered as contributors to fatigue during this type of exercise (Lucke
(Knight et al., 1991; Geor et al., 1999). Geor and colleagues (1999)
& Hall, 1980; Snow et al., 1982; Hodgson, 1985; Hodgson & Rose,
reported an 8.9% increase in V.O2max after 10 days of moderate-
1994a). Performance capacity or onset of fatigue in horses and
intensity training at 55% V.O2max for 60 min each day. Concurrent
humans has been correlated with a depletion in glycogen stores in
metabolic studies revealed a decrease in muscle glycogenolysis and
working muscle (Snow et al., 1981, 1982; McMiken, 1983; Åstrand
anaerobic metabolism but no concomitant increase in muscle oxi-
& Rodahl, 1986). A decline in extracellular glucose concentrations
dative enzyme activities. So, although there was improvement in
may also be important. However, fatigue or reduced muscle power
the aerobic capacity of the horses the mechanisms underlying it
occurs before the complete depletion of any substrate and it is the
were not clear (Geor et al., 1999). In contrast, normal training in
rate of ATP production that appears crucial. Neural fatigue in short
Standardbreds resulted in decreased muscle hexokinase activity
events is considered very unlikely but may be important in endur-
(Wijnberg et al., 2008). Guy and Snow (1977) reported an increase
ance events (McMiken, 1983). However, in supramaximal exercise
in muscle glycogen content and an increase in aerobic and anaero-
lasting only a few seconds, fatigue may be related to an inability of
bic enzymes in response to training.
the neuromuscular junction to maintain the propagation of excit-
Various conditioning programs studied by Gansen and colleagues
atory action potentials into the muscle fibers, possibly leading to
(1999) revealed significant increases in muscle glycogen storage in
early musculoskeletal damage (Leach & Sprigings 1979). Racehorses
Haflinger stallions after low-intensity, long-duration exercise but
that were being whipped as they were progressing through the field
not after higher-intensity, shorter-duration exercise. This adaptation
were at much greater risk of falling compared to horses that were
in the middle gluteal muscle fibers was only noted at 6-cm depth
not being whipped and that had no change in position or lost posi-
and not at 2-cm depth, demonstrating the variable response of dif-
tion through the field (Pinchbeck et al., 2004).
ferent regions of muscle fibers to training.
Few scientific studies comparing conventional and interval train-
ing techniques for Thoroughbreds have been reported. In one study
Training programs it was concluded that interval training led to higher lactate produc-
tion and increased plasma lactate clearance during a 1000-m sprint
compared to conventionally trained horses and this was equated
The implications of the various energy supply pathways are far-
with greater anaerobic capacity (Harkins et al., 1990). An earlier
reaching in terms of specific training programs for horses competing
study failed to identify significant differences in the response to
in different athletic events. An understanding of the patterns of
conventional and interval training techniques but the total slow-
energy substrate supply and utilization in different athletic events
and fast-work distances undertaken were the same in both training
allows tailored training strategies to be adopted to maximize the
schedules (Gabel et al., 1983).
adaptations in various body systems. Although no activity is exclu-
The various physiologic adaptations that occur in response to
sively aerobic or anaerobic in nature, aerobic energy production is
training in the horse have been extensively reviewed elsewhere
vital in all cases. Hence an emphasis on establishing a good ‘aerobic
(Hodgson & Rose, 1994b; Rivero et al., 2001, 2007; Marlin & Nan-
foundation’ is considered paramount in any training program
kervis, 2002; Hinchcliff et al., 2004; Rivero et al., 2007).
(Rivero et al., 2008).
Even today, there is probably more desire to use training methods
that have ‘stood the test of time’ over many years and have been Measuring energy expenditure
passed on from generation to generation, than to develop a training
program based on current scientific knowledge. Naturally, the suc- Until the 1980s, some veterinarians and horse trainers relied on the
cessful training of any athlete relies on far more than simple physi- complete blood count and the plasma or serum biochemistry panel
ologic adaptations to exercise. The maintenance of motivation, the in an attempt to assess clinical and subclinical disease in perfor-
skill of predicting a horse’s tolerance levels and the fostering of that mance horses. Evaluation of resting and/or post-exercise hematol-
psychogenic factor, ‘the will to win,’ are vitally important and ogy and biochemistry to assess fitness capacity or response to
unquantifiable. The ability to train a horse to its limits without training is of limited use.
causing it to ‘break down’ is necessary. These ‘human factor’ com- A range of equipment to determine various performance indices
ponents of any training program separate the elite trainers from has now been available and extensively used for the last decade and
their average peers. a half, including treadmills, heart rate meters and rapid lactate
Training may be of a continuous or intermittent nature. Continu- analyzers.
ous training, also known as endurance training, refers to exercise
performed in a single bout that can vary in intensity and duration.
Intermittent or interval training implies a series of intense work Treadmills
sessions interspersed with rest periods of varying duration. The The first use of a high-speed equine treadmill for experimental
intensity and duration of training are two variables that must be purposes was in 1960 in Stockholm, Sweden (Fredricson et al.,
considered in any training program. It seems that the shorter the 1983; Sloet van Oldruitenborgh-Oosterbaan & Clayton, 1999).
race the more consideration should be given to the intensity of Early studies focused on metabolic variables with locomotor studies
training and the longer the race the more important endurance becoming more common in later years (Barrey et al., 1993; Gottlieb-
training becomes (Bayly, 1985). A basic model of training incorpo- Vedi & Lindholm, 1997; Couroucé et al., 1999). Equine treadmills
rates three phases (Derman & Noakes, 1994; Evans, 1994): have been installed in many university veterinary colleges and

431
 18 Metabolic energetics

equine research centers for ongoing studies of exercise physiology to determine V.O2max is a standardized rapid incremental test on a
and for the diagnosis of poor performance conditions, in particular 10% treadmill incline (Rose et al., 1990b) and the repeatability of
dynamic upper airway obstructions. In addition, treadmills are now results is good (Evans & Rose, 1988a; Seeherman & Morris, 1990).
commonly used in training stables as a complementary training Oxygen uptake can improve quickly with the onset of training but
tool and by stud farms to walk and trot yearlings for conditioning relative training intensity, when kept constant and submaximal,
prior to sale. does not appear to affect the rate of change of V.O2max (Knight et al.,
Most current knowledge concerning the physiologic response of 1991).
the horse to exercise has come from numerous treadmill-based
studies. Cardiovascular, respiratory, metabolic, hematologic, ther-
moregulatory, hormonal, musculoskeletal and locomotory changes
Lactate analysis
in the horse, exercising over various intensities and durations, have Response to training and the relative intensity of an exercise session
been thoroughly examined. can be assessed by the simple measurement of blood or plasma
Treadmill exercise is not equivalent to track exercise (Sloet van lactate, during or after exercise (Milne et al., 1977). Lactate increases
Oldruitenborgh-Oosterbaan & Clayton, 1999). The effects of air at an exponential rate with increasing workload, and fitter horses
movement, track surface and rider impact are not duplicated on the show a slower accumulation during submaximal exercise. The VLA4
treadmill and horses have no forward momentum on the treadmill is often calculated to compare fitness between horses and response
because the moving belt provides the driving force (Rose & Hodgson, to training. Classically, the VLA4 has been considered to approximate
1994b). So the amount of work performed by a horse on the tread- the anaerobic threshold, mirroring the metabolic transition from
mill is quantitatively different from work on the track. For a track predominantly aerobic to anaerobic energy sources, and this calcu-
exercise test, horses require only a short habituation period and can lated value increases with improved fitness (Hodgson & Rose,
be worked in their standard manner, often with the usual rider or 1994b). Plasma lactate values are 30–50% higher than whole blood
driver (Sloet van Oldruitenborgh-Oosterbaan & Clayton, 1999). lactate (Rose & Hodgson, 1994b). However, because of great inter-
Nevertheless, there are clear advantages to studying responses to individual variation in lactate distribution between plasma and red
exercise on the treadmill. A consistent exercise surface, controlled blood cells (RBCs) after exercise and in the rate of lactate influx into
environmental conditions, precise control over intensity of exercise RBCs, there is no consistent relationship between the two lactate
and ease of measuring physiologic variables to monitor fitness are reservoirs (Pösö et al., 1995; Väihkönen & Pösö, 1998). Recent
all strong indications to pursue treadmill-based studies (Rose & evidence suggests that whole blood lactate concentrations should
Hodgson, 1994b; Sloet van Oldruitenborgh-Oosterbaan & Clayton, be measured when estimating the accumulation of lactate from
1999). By positively inclining the treadmill, a horse can be exercised exercising muscle, to minimize variation due to factors that influ-
at its maximum power output at a relatively slower speed than if it ence transport of lactate from plasma into RBCs (Väihkönen et al.,
were on a flat plane (Sexton & Erickson, 1990). This potentially 1999). If whole blood is to be used, the sample should be imme-
reduces the risk of musculoskeletal injury because speeds above diately deproteinized to halt post-collection production of lactate
12–13 m/s are unnecessary, but the steeper the slope the greater the within the RBC (Ferrante, 1995); however, storage at 0°C for up to
effects on gait and mechanics. It is suggested that muscles may be an hour before deproteinization does not affect the lactate concen-
recruited differently when the horse is exercised on a slope versus tration (Ferrante & Kronfeld, 1994).
the flat (Sloet van Oldruitenborgh-Oosterbaan & Barneveld, 1995). Whether plasma or whole blood lactate is assessed, one method
A slope of 10% (5.71°) is recommended for treadmill testing as should be adhered to by the laboratory or investigator, to reduce
most horses will reach their maximum oxygen uptake at speeds of variability in measurements. Post-exercise blood and plasma lactate
10–12 m/s compared with 14–15 m/s on the flat (Rose & Hodgson, concentrations were significantly correlated with race performance
1994b). It is desirable to standardize the incline that exercise for Thoroughbreds undergoing a submaximal treadmill exercise test
tests are performed on, to allow better comparison between studies (Evans et al., 1993). Harkins and colleagues (1993) found the VLA4
from different institutions (Sloet van Oldruitenborgh-Oosterbaan to be one of the best correlates of running speed for Thoroughbreds.
& Clayton, 1999). This was a negative correlation, indicating that faster horses attained
Significant differences in locomotor and metabolic variables have a plasma lactate of 4 mmol/L at a lower velocity than did slower
been reported in studies comparing track versus treadmill exercise horses. The faster horses also had the highest peak lactate concentra-
(Sloet van Oldruitenborgh-Oosterbaan & Clayton, 1999) and tions, implying that plasma lactate concentrations of faster horses
future research may continue to elucidate the etiology of these dif- rise more rapidly and to higher levels than do those of slower
ferences. Currently, treadmill tests are preferable for most research horses. However, no correlation was found between performance
purposes but track tests may be of greater importance when examin- and post-exercise blood or plasma lactate concentrations taken after
ing locomotor variables and fitness of sport horses (Sloet van maximal activity during a field trial (Evans et al., 1993).
Oldruitenborgh-Oosterbaan & Clayton, 1999). Recent research (Lindner et al., 2010) defined the maximum
lactate at steady state (maxLASS) in horses based on the definition
by Heck et al. (1985) and showed that not V4 but V2 most closely
Oxygen consumption resembles the maxLASS in horses. They defined maxLASS as the
The measurement of V.O2 by the exercising body is the single most maximal speed at which the [LA] does not change by more than
important step in the evaluation of exercise capacity. In the horse, 1 mmol/L between the 5th and the 25th min of exercise at a con-
various mask systems have been investigated and an open-flow stant pace. More research is needed to confirm this conclusion,
mask without valves is the currently accepted apparatus. Bayly and though it has already been shown that training of endurance horses
colleagues (1987) compared flow-through mask systems with at V2 for 4 weeks was able to significantly improve the V4 during
valved masks and found a funnel-shaped, valveless flow-through standardized exercise tests (Trilk et al., 2002).
system to have the least impedance on airway function at flow rates
of 6300 L/min. Evans and Rose (1988b) investigated a valved mask
system and showed a negative influence on arterial blood gases,
Heart rate
namely an exacerbation of exercise-induced hypercapnea and Heart rate is measured to monitor exercise intensity on the basis
hypoxemia. Respiratory frequency was lower when the mask was that there is a linear relationship between heart rate and work per-
worn but arterial acid–base tensions and heart rate were minimally formed in the range of 120–210 beats per min (bpm) (Persson,
affected. The respiratory effects were attributed to alveolar hypoven- 1983). The velocity at a heart rate of 200 bpm (V200) has been used
tilation (Evans & Rose, 1988b). The most appropriate exercise test to assess fitness and response to training but care should be taken

432
 Training programs

when evaluating this variable (Rose & Hodgson, 1994a). Conflict- gluteal muscle was used as the preferred site of sampling and con-
ing reports on the correlation between V200 and V.O2max have been tinues to be favored today although other muscles, e.g. the semi-
published (Evans & Rose, 1987; Rose et al., 1990a) but this is attrib- tendinosis, biceps femoris, and lateral vastus have been examined.
uted to the dissimilar numbers of horses tested in each study and The latter muscle has the advantage to be able to compare muscle
it is accepted that the two variables are significantly positively cor- histochemistry and histopathology with quantitative needle elec-
related. The usefulness of V200 to assess response to training in the tromyography (QEMG) parameters in the conscious horse
field using Thoroughbreds has been confirmed (Kobayashi et al., (Wijnberg et al., 2008) as for instance the gluteal muscle has very
1999). A number of heart rate meters are available with generally poor basal motor unit action potentials (MUPs). Examination of
good accuracy (Evans & Rose, 1986; Rose & Hodgson, 1994b; muscle tissue has allowed fiber typing and evaluation of responses
Holopherne et al., 1999). Telemetered electrocardiography is the to exercise and training, particularly alterations in substrate utiliza-
favored technique in treadmill laboratories. To achieve accurate tion and the oxidative capacity of muscle. One concern with the
heart rate values good electrode contact with the skin is required, use of muscle biopsies is the degree of variation in samples from
and this can be difficult to maintain in the galloping horse. Electrode very similar sites. A standardized approach to the muscle biopsy
gels can enhance contact and glue adhesives can hold electrode procedure in French trotters has been described, incorporating
casings firmly on the skin (Hill et al., 1977; Rose & Hodgson, 1994b). anatomical landmarks, age, sex and hip width of the horses (Valette
et al., 1999). Interestingly, muscle fiber composition has been cor-
related with locomotor patterns in horses (stride frequency and
Blood gases stride length) (Rivero & Clayton, 1996, Rivero et al., 2006) and
Respiratory disorders that may interfere with the transport of oxygen therefore may indirectly influence the economy of locomotion.
from the atmosphere to the pulmonary vasculature can be assessed Furthermore, epaxial muscle biopsy characterized histopathologi-
for their significance by measuring arterial blood gas tensions. Arte- cally and by electron microscopy is a good option in diagnosing
rial blood samples are normally collected from a catheterized trans- back problems in horses when clinical examination and imaging
verse facial artery during a treadmill exercise test. Values should be techniques do not provide a precise diagnosis (Quiroz-Rothe et al.,
corrected for central venous blood temperature. Hypoxemia and 2002) (Fig. 18.9). In addition, urinary excretion of organic acids,
hypercapnea are recognized responses to high-intensity exercise glycine conjugates and acylcarnitines (Westermann et al., 2008b),
(Bayly et al., 1983, 1987) and the severity of hypoxemia increases needle electromyography (EMG) (Wijnberg et al., 2002, 2003,
with training (Christley et al., 1997). There is a strong negative cor- 2008) and proteomics (Bouwman et al., 2010) are attractive
relation between minimal arterial oxygen content and VO2max in additional tools to evaluate myopathy besides increased plasma
trained horses, indicating the importance of assessing both vari- muscle enzyme activities (like CK) and histological evaluation
ables before interpreting blood gas data (Christley et al., 1997). of muscle biopsies.
Horses with functional airway obstructions may have a greater Using modern cDNA microarrays, Barrey et al., (2006) showed
degree of hypoxemia or hypercapnea when approaching, and at, that genes are modulated in leucocytes in relationship with perfor-
maximal exercise intensities compared to normal horses (Rose & mance and clinical status of the horses. It appeared that the gene
Hodgson, 1994a). The degree of hypoxemia in normal Thorough- ontology classification showed that more genes were up-regulated
breds performing a standardized treadmill exercise test was not in successful than in disqualified endurance horses. More genes
correlated with their running speed on an 800 m track (Harkins
et al., 1993).

Blood volume
The Evans blue dye technique of measuring the plasma volume
(PV) in horses was first described by Persson in 1967. It is a simple
and highly reproducible technique with a coefficient of variation
of 3–4%. To ensure an accurate assessment of the plasma volume,
the splenic erythrocyte pool must be mobilized either by intense
exercise or an epinephrine injection. A post-exercise hematocrit is
preferred and is typically measured for blood volume calculations.
A major determinant of the oxygen-carrying capacity of the horse
is the red cell volume (RCV), which can be calculated from the
hematocrit and PV. Evaluation of the hematocrit alone as an indi-
cator of RCV can be misleading due to variations in PV. Plasma
volume increases in response to training in all species studied
(Oscai et al., 1968; Persson, 1968; McKeever et al., 1985, 1987).
Total blood volume (TBV) has been positively correlated with
fitness level and may be a useful measure of such as long as varia-
tions due to body size, age, sex and breed of horse are taken into
account (Persson, 1968). A significant positive correlation has also
been found between TBV and racing performance in Standardbred
trotters (Persson, 1968). During a single high-intensity exercise
bout TBV increases substantially in the horse mainly due to the
release of red cells from the splenic reservoir. It remains to be elu-
cidated whether PV or RCV can be used to predict performance in Fig 18.9  Electron micrograph of lateral vastus muscle transverse section
horses. from a 12-year-old Warmblood mare illustrating normal mitochondria.
Mitochondria are no longer the once-thought uniform ATP-producing
organelles, identical in all organisms, but are rather diverse and adapted to
Muscle biopsy distinct conditions. Next to the classical aerobic mitochondria, which
Needle biopsy of skeletal muscle was first described in the horse contain a respiratory chain and use oxygen as a final electron acceptor,
more than 35 years ago by Lindholm and Piehl (1974). The middle anaerobically functioning mitochondria were identified.

433
 18 Metabolic energetics

were down-regulated in the disqualified horses. Some genes were disorder, such as long-term increased CRH secretion raised by
expressed in relationship with the clinical phenotype observed in chronic stress or indirectly by increased secretion of an ACTH inhib-
the disqualified horses: rhabdomyolysis and hemolysis. Recently, itory factor that counteracts the action of CRH raised by chronic
Eivers et al., (2010) investigated the adaptive changes in mRNA social stress as described by Alexander et al. (1996) in chronic
expression to training in equine skeletal muscle biopsies by real- socially stressed horses.
time qRT-PCR for a panel of candidate exercise-response genes Pituitary growth hormone (GH) is usually secreted in an episodic
following a standardized incremental-step treadmill exercise test in manner as a result of a delicate interaction between the two hypo-
a group of untrained Thoroughbred horses. Significant differences thalamic peptides, GH-releasing hormone (GHRH) and somatosta-
in gene expression were detected for several genes 4 h after exercise. tin (SRIF). GH stimulates the liver to produce insulin-like growth
Investigation of relationships between mRNA and velocity at factor (IGF-1), which mediates most of the effects of GH via the
maximum heart rate and peak postexercise plasma lactate concen- IGF receptor family. Defects in the function of the GH-IGF-1-axis
tration revealed significant associations with postexercise gene could, in theory, be located at any level from the hypothalamus
expression, and between plasma lactate concentration and basal down to the target receptors in skeletal tissues. Diagnostic tests for
gene expression. These findings highlight the roles of genes respon- assessing the function of the GH-IGF-1 axis include measurements
sible for the regulation of oxygen-dependent metabolism, glucose of multiple serum GH concentrations, single serum IGF-1 concen-
metabolism, and fatty acid utilization in equine skeletal muscle tration, and various endocrine challenge tests, for example, with
adaptation to exercise. GHRH. GH pulsatility characteristics can be determined from the
GH data series by visual identification of presumptive pulses.
However, this method is not very objective. Therefore, computerized
Hormone profile algorithms have been developed in order to analyze GH hormone
In general, stress occurs when body homeostatic balance is dis- pulsatility and regularity (Veldhuis et al., 2008). Because monitor-
turbed (Sapolsky et al., 2000; Selye 1951). When the body experi- ing the pattern of spontaneous GH secretion is labor-sensitive and
ences environmental or internal stressors it responds by secreting a difficult, challenge tests to measure GH ‘secretory reserves’ are often
whole array of hormones to reestablish homeostatic balance. Regen- used instead. GH deficiency and GH insensitivity also influence the
erative processes continue after restoration of homeostatic balance serum concentrations of IGF-1 and IGF-2 and their binding pro-
such that, if the same stressor were imposed again, the homeostatic teins, which makes the latter good indices of GH status. Further-
mechanisms would not be displaced to the same extent, resulting more, the serum concentrations of IGF-1 and IGF-2 are relatively
in overcompensation (Selye, 1951). The overcompensation is seen constant during the day, so that stimulation tests or multiple sam-
as positive stress. plings are not necessary (De Graaf-Roelfsema et al., 2011).
The hormonal events during reestablishment of homeostasis due
to (exercise) stress can be divided into two phases. Initially, a cata-
bolic phase can be distinguished, with decreased tolerance of effort, Effects of exogenous GH
characterized by reversible biochemical, hormonal and immuno- It is well known that GH has anabolic effects, and its consequent
logical changes. The two main hormonal axes activated in this phase abuse is a concern in many sports, including horse racing. The avail-
are the sympathetic-adrenal medullary (SAM) axis and the ability of reGH in Australia has attracted the attention of horse
hypothalamic-pituitary-adrenocortical (HPA) axis (Armstrong & racing authorities worldwide and urged the development of
van Heest, 2002). An anabolic phase follows with a higher adaptive methods for the detection of its abuse (De Kock et al., 2001). The
capacity and enhanced performance capacity, in which both the availability of reGH also attracted the attention of researchers to
GH-IGF-I axis as well as the gonadal-axis are activated (Urhausen determine the responses to administration of reGH in horses.
et al., 1995). When homeostatic balance is not restored, the body The biological responses to reGH administration in adult horses
experiences chronic stress which induces chronic activation of the are similar to those common in other species: hyperglycemia,
endocrine system and possibly ending in a neuroendocrine disorder hyperinsulinemia, insulin resistance, decreased plasma urea nitro-
like the overtraining syndrome. gen concentrations, increased plasma IGF-I and IGFBP-3 concentra-
So far, no single diagnostic parameter has been identified for tions (De Kock et al., 2001, Julen Day et al., 1998, Malinowski
overreaching and the overtraining syndrome in humans or equines et al., 1997, Popot et al., 2001, Smith et al., 1999, Thatcher and
and so far a combination of different (hormonal) parameters Thompson 2002).
appear to be the best indicators of overreaching/overtraining (Hug Beneficial effects of reGH administration reported to date in adult
et al., 2003, Rivero et al., 2008). Standardized exercise tests are sug- horses include increased nitrogen retention, muscularity and granu-
gested to provide a way to detect subtle changes in hormonal locyte numbers in aged mares (Malinowski et al., 1997).
responses in the individual, which may make an important contri- No effects of reGH administration were found in age-related
bution to the detection of early overtraining (Rivero et al., 2008). declines in various immune parameters in adult horses (Guirnalda
Alterations in functioning of the HPA axis has been described in et al., 2001), on modulation of the in vitro biomechanical properties
trained, overreached and overtrained horses. Several mechanisms of superficial digital flexor tendon (SDFT) (Dowling et al., 2002a)
underlying the alterations in functioning have been suggested, or on second intention wound healing (Dart et al., 2002). Also no
among them diminished adrenocortical sensitivity to adrenocorti- effects were found on aerobic capacity or exercise performance in
cotropic hormone (ACTH) and a decreased negative feedback sen- geriatric mares (McKeever et al., 1998) or on exercise capacity or
sitivity (Golland et al., 1999; Marc et al., 2000; Hamlin et al., 2002; indices of fitness in young Standard bred horses in training (Gerard
Persson et al., 1980). However, so far only ACTH challenges were et al., 2002).
used to study the alterations in the HPA axis in horses during (over) ReGH administration had a negative effect on the biomechanical
training. Recently, the influences of overtraining on the function of properties of healing SDFT (Dowling et al., 2002b) and peripheral
the HPA axis using a corticotropin releasing hormone (CRH)- insulin sensitivity (De Graaf-Roelfsema et al., 2005). Long-term
stimulation test were investigated (De Graaf-Roelfsema et al., administration of reGH in foals did not influence body weight, long
2008). The main finding was a reduction in the ACTH response to bone growth and other body sizes. Generally, basal glucose concen-
administration of CRH in the overtrained group. This finding was tration and insulin response to glucose infusion were higher in
not accompanied by significant differences between the control and reGH treated foals. Endogenous GH secretion in response to GH
overtrained group in plasma cortisol concentrations after CRH secretagogue (EP51389) was significantly reduced in treated foals.
administration. It was hypothesized that the decrease in pituitary The prolactin and thyroid-stimulating hormone (TSH) responses to
sensitivity must have been caused by a more centrally located TRH were not altered by reGH treatment. Mean serum IGF-1

434
 Thermoregulatory consequences of exercise

concentrations were not significantly higher in the reGH treated to be the underlying mechanism for the observed changes in GH
group. However, there was a significant increase of IGF-1 in the first pulsatility pattern. A 4-week recovery period did not normalize the
5 weeks of treatment in foals receiving reGH compared to controls. pulsatile growth hormone secretion, thereby excluding over-
At necropsy, many internal organ weights were increased, but little reaching as its cause.
effect on histopathologic characteristics was found in the same foals
(Capshaw et al., 2001; Kulinski et al., 2002).
Thermoregulatory consequences of exercise
Endogenous GH responses to exercise
During the production of energy for locomotion the mechanical
The GH-IGF-I response to exercise is well described in humans efficiency of the processes described above is ~20%. That is, one-
(Wideman et al., 2002; Birzniece et al., 2011), but not in horses. fifth of the energy produced goes into exercise with the remaining
Nevertheless, some studies provide information about the acute 80% given off as heat. Given that the mass-specific maximal
changes in GH-IGF-I axis due to a bout of exercise (Cartmill et al., oxygen uptake of horses is at least 2-fold higher than in man, at a
2003; Thompson et al., 1992, 1994). To induce a significant plasma given workload, the metabolic heat load in horses is considerably
GH concentration elevation post-exercise, horses should perform higher. Further, relative to body mass, the surface area available for
exercise of at least 10 min duration with intensity above the lactate dissipation of heat in horses is approximately 50% of that in man
threshold (50–70% VO2max) to overcome autonegative feedback. (Hodgson et al., 1993; Hodgson et al., 1994). On the other hand,
There is no indication that exercise modifies IGF-1 concentrations the horse’s efficient thermoregulatory mechanisms provide for
in plasma in trained adult horses (Popot et al., 2001; Noble et al., effective transfer of heat from contracting skeletal muscle to the
2007). environment.
De Graaf-Roelfsema et al., (2009) successfully used the resting Also, the horse can selectively cool the brain during exercise or
nocturnal pulsatile growth hormone secretion pattern to detect the heat exposure by cooling the venous blood within the cavernous
(over)training status of Standardbreds in an experimental setting sinuses during respiration (McConaghy et al., 1995). The primary
(Fig. 18.10). The overtrained horses altered their resting pulsatile physiological mechanisms driving heat loss during exercise are an
growth hormone secretion with an increase in the number of con- increased proportion of cardiac output directed toward the cutane-
centration peaks, a smaller peak secretion pattern with a prolonged ous circulation and an increased rate of sweat secretion. Sweating
growth hormone half-life, and an increased approximate entropy and cutaneous evaporation are the most important heat dissipatory
(ApEn). The increased irregularity of nocturnal GH pulsatility mechanisms in horses, accounting for 65–70% of heat loss during
pattern is indicative of a loss of coordinated control of GH regula- prolonged exercise (Hodgson et al., 1993; Hodgson et al., 1994).
tion. Longer phases of somatostatin withdrawal were hypothesized Sweating rates of 20–55 g/m2/min have been measured on the
necks and backs of exercising horses; for a 500-kg horse these rates
correspond to sweat fluid losses of 6–15 L/h. When expressed in
5 terms of sweating rate per unit area of skin, these rates are 2- to
3-fold greater than those reported for human subjects (McCutcheon
4 & Geor, 2000; McCutcheon & Geor, 2004).
Unsurprisingly, the thermal responses to exercise are affected by
3 ambient conditions. Under conditions of high heat (>30°C) and
GH (µg/l)

humidity evaporative heat loss is severely limited with resultant

2 increases in the rate of heat storage and degree of hyperthermia. In
horses, the rate of heat storage when exercising in hot, humid condi-
1 tions can be more than twice the rate occurring during exercise at
the same intensity in cool, dry conditions (Geor et al., 1995; Kohn
0 et al.1999). Increased demands for respiratory heat loss are reflected
0 100 200 300 400 500 by an increase in respiratory rate during and after exercise (Geor
A Time (minutes) et al., 1995; McConaghy et al., 2002; McCutcheon & Geor, 2004).
Further, dehydration associated with profuse sweat fluid losses
5 can further compromise heat transfer and exacerbate hyperthermia
(McCutcheon & Geor, 2004). An important consequence of this
4 impairment of heat dissipation during exercise in hot weather is a
decrease in the time to attain a critical upper limit in core body
3
GH (µg/l)

temperature that results in development of fatigue (Gonzalez-

Alonso et al., 1999; Nybo & Nielsen 2001). Moreover, exercise in
2 such conditions increases the risk of developing heat-related ill-
nesses in horses (McCutcheon & Geor, 2004).
1 Several factors may contribute to this decrease in performance
when exercise is undertaken in hot versus cool conditions, includ-
0 ing the effects of hyperthermia on brain and muscle function, com-
0 100 200 300 400 500
promise of muscle blood flow (McConaghy et al., 2002) and
B Time (minutes)
reduced aerobic power (Art & Lekeux, 1995). In trained human
subjects, exhaustion during exercise in the heat corresponds to a
Fig 18.10  Mean nocturnal GH profiles for intensified trained (red line)
horses and control horses (blue line) after the training period (A) and
core temperature of about 40°C (Gonzalez-Alonso et al., 1999;
intensified training period (continued normal training for the control Nybo & Nielsen, 2001). The onset of fatigue at this critical upper
horses) (B). limit may represent a mechanism to avoid heat stroke.
Reprinted from De Graaf-Roelfsema, E., Veldhuis, P.P., Keizer, H.A., et al., 2009, Overtrained Measurements of central blood (pulmonary artery) temperature
horses alter their resting pulsatile growth hormone secretion, Am J Physiol-Regulatory, in horses during heavy exercise have demonstrated that fatigue
Integrative and Comparative Physiology, 297 (2), R403–411, with permission from the occurs as blood temperature approaches 42.5–43°C. As such,
American Physiological Society. horses may have greater thermal tolerance than man, perhaps in

435
 18 Metabolic energetics

12 180.0

10 160.0
Unfit Fit
140.0
Lactate (mmol/liter)

80
120.0

V02max (ml/kg/min)
60 VLA4 VLA4
100.0
Anaerobic threshold
40 80.0

20 60.0

40.0
0
100 200 300 400 500 20.0
Velocity (m/min)
0.0
Fig 18.11  Training increases the Vla4 and moves the lactate curve to the Before After
right. Good performers have significantly higher Vla4 than poor performers,
which makes it a valid parameter to assess fitness in Standardbred horses. Fig 18.12  Training usually increases VO2max between 10 and 25%.
From Hinchcliff et al. (2004) and Lindner (2010). Percentage improvement depends on the initial fitness of the horse and
type of the horse. The figure shows improvement of VO2max after 4 weeks of
aerobic training in untrained small (blue bars) and midsize (red bars) ponies,
and in detrained TB. (Green bars, mean ± SEM.)
part because the selective brain cooling mechanism maintains Adapted from Katz et al. (2000).
hypothalamic temperature approximately 1°C lower than central
blood temperature during exercise (McConaghy et al., 1995).
Clinical experience has indicated that poor physical conditioning, on performance has been heavily investigated but continues to be
prolonged exercise (e.g. endurance races; speed and endurance test an area of considerable controversy; comparison between studies is
of a 3-day event) in hot environments, lack of heat acclimatization, difficult and at this time meaningful conclusions cannot be drawn.
and dehydration are factors that may increase the risk of exertional Well-thought-out training programs tailored to the horse’s particu-
heat illnesses in horses. Horses with a history of anhydrosis are lar activity are important for preparing the horse for athletic
obviously at higher risk for development of exercise-associated heat endeavors (Fig. 18.11). Every program should have an initial low-
illnesses. intensity foundation phase to allow body tissues to adapt to the
stresses placed upon them. The economy of locomotion refers to
the optimal gait at a given speed of exercise at which the energy cost
is least and this gait is naturally chosen by freely moving horses.
Conclusions The measurement of V.O2max remains the single most important
assessor of a horse’s relative fitness (Fig. 18.12). Fundamental dif-
It is imperative that anyone with an interest in the horse as an ferences between track and treadmill exercise tests make both
athlete should understand equine energy production and utiliza- methods of performance evaluation attractive; however, because of
tion. Both aerobic and anaerobic pathways of energy supply are more controllable conditions treadmill exercise tests continue to be
necessary for all forms of exercise. Evidence suggests that the aerobic favored. The emphasis is on further development and evaluation of
system has a much greater role to play than previously thought, in procedures that can be readily and reliably applied to the field
short-duration, supramaximal exercise bouts. The effect of nutrition situation.

References
Alexander, S.L., Irvine, C.H.G., Donald, R.A., Auvinet, B., 1996. Performance testing and tension in Thoroughbred horses. In: Snow,
1996. Dynamics of the regulation of the improvement in human athletes. D.H., Persson, S.G.B., Rose, R.J. (Eds.),
Hypothalamo-pituitary-adrenal (HPA) axis Pferdeheilkunde 12 (4), 455–456. Equine exercise physiology. Granta
determined using a nonsurgical method for Barrey, E., Galloux, P., Valette, J.P., Auvinet, B., Editions, Cambridge UK, pp. 400–407.
collecting pituitary venous blood from Wolter, R., 1993. Determination of the Bayly, W.M., Schulz, D.A., Hodgson, D.R.,
horses. Frontiers Neuroendocrinol. 17, optimal treadmill slope for reproducing the Gollnick, P.D., 1987. Ventilatory responses
1–50. same cardiac response in saddle horses as of the horse to exercise: effect of gas
Armstrong, L.E., Van Heest, J.L., 2002. The overground exercise conditions. Vet. Rec. collection systems. J. Appl. Physiol. 63 (3),
unknown mechanism of the overtraining 133, 183–185. 1210–1217.
syndrome: clues from depression and Barrey, E., Mucher, E., Robert, C., Amiot, F., Biknevicius, A.R., Mullineaux, D.R., Clayton,
psychoneuroimmunology. Sports Med. 32, Gidrol, X., 2006. Gene expression profiling H.M., 2006. Locomotor mechanics of the
185–209. in blood cells of endurance horses tölt in Icelandic horses: Taking the walk for
Art, T., Lekeux, P., 1995. Respiratory completing competition or disqualified due a run. Am. J. Vet. Res. 67, 1505–1510.
adjustments in unacclimatised to metabolic disorder. Equine Vet. J. 36 Birzniece, V., Nelson, A.E., Ho, K.K., 2011.
horses exercised under hot, humid (Suppl.), 43–49. Growth hormone and physical
conditions. Equine Vet. J. 18, Bayly, W.M., 1985. Training programs. Vet. performance. Trends Endocrinol. Metab. 22
289–293. Clin. N. Amer.: Equine Pract. 1 (3), (5), 171–178.
Åstrand, P.-O., Rodahl, K., 1986. Textbook of 597–610. Bouwman, F.G., Van Ginneken, M.M., Van der
work physiology: Physiological bases of Bayly, W.M., Grant, B.D., Breeze, R.G., Kramer, Kolk, J.H., Van Breda, E., Mariman, E.C.,
exercise. third ed. McGraw-Hill Book J.W., 1983. The effects of maximal exercise 2010. Novel markers for tying-up in horses
Company, New York. on acid-base balance and arterial blood gas by proteomics analysis of equine muscle

436
 References

biopsies. Comp. Biochem. Physiol. Part D Wijnberg, I.D., et al., 2005. Effects of Eaton, M.D., Rose, R.J., Evans, D.L., Hodgson,
Genomics Proteomics 5, 178–183. short- and long-term recombinant equine D.R., 1992. The assessment of anaerobic
Cain, D.F., Davies, R.E., 1962. Breakdown of growth hormone and short-term capacity of Thoroughbred horses using
adenosine triphosphate during a single hydrocortisone administration on tissue maximal accumulated oxygen deficit. Aust.
contraction of working muscle. Biochem. sensitivity to insulin in horses. Am. J. Vet. Equine Vet. 10 (2), 86.
Biophys. Res. Comm. 8 (5), 361–366. Res. 66, 1907–1913. Eivers, S.S., McGivney, B.A., Fonseca, R.G.,
Capshaw, E.L., Thompson Jr., D.L., Kulinski, De Graaf-Roelfsema, E., van Ginneken, M.E., MacHugh, D.E., Menson, K., Park, S.D.,
K.M., Johnson, C.A., French, D.D., 2001. van Breda, E., Wijnberg, I.D., Keizer, H.A., et al., 2010. Alterations in oxidative gene
Daily treatment of horses with equine van der Kolk, J.H., 2006. The effect of expression in equine skeletal muscle
somatotropin from 4 to 16 months of age. long-term exercise on glucose metabolism following exercise and training. Physiol.
J. Anim. Sci. 79, 3137–3147. and peripheral insulin sensitivity in Genomics 8, 83–93.
Standardbred horses. Equine Vet. J. 38 Essén-Gustavsson, B., Blomstrand, E.,
Cartmill, J.A., Thompson Jr., D.L., Storer, W.A.,
(Suppl.), 221–225. Karlstrom, K., Lindholm, A., Persson,
Gentry, L.R., Huff, N.K., 2003. Endocrine
responses in mares and geldings with high De Graaf-Roelfsema, E., Veldhuis, P.P., Keizer, S.G.B., 1991. Influence of diet on substrate
body condition scores grouped by high vs H.A., van Ginneken, M.M., van Dam, K.G., metabolism during exercise. In: Persson,
low resting leptin concentrations. J. Anim. Johnson,M.L., et al., 2009. Overtrained S.G.B., Lindholm, A., Jeffcott, L.B. (Eds.),
Sci. 81, 2311–2321. horses alter their resting pulsatile growth Equine exercise physiology 3. ICEEP
hormone secretion. Am. J. Physiol. Regul. Publications, Davis, California,
Christley, R.M., Hodgson, D.R., Evans, D.L.,
Integr. Comp. Physiol. 297, R403–411. pp. 288–298.
Rose, R.J., 1997. Effects of training on the
De Graaf-Roelfsema, E., Veldhuis, P.P., van Essén-Gustavsson, B., Gottlieb-Vedi, M.,
development of exercise-induced arterial
Dam, K.G., Menheere, P., Keizer, H.A., Lindholm, A., 1999. Muscle adenine
hypoxemia in horses. Am. J. Vet. Res. 58
Johnson, M.L., et al., 2011. Assessment of nucleotide degradation during submaximal
(6), 653–657.
endogenous growth hormone pulsatility treadmill exercise to fatigue. Equine Vet. J.
Clayton, H.M., 1991. Energy production. In: in gelded yearling horses using
Clayton, H.M. (Eds.), Conditioning Sport 30 (Suppl.), 298–302.
deconvolution analysis. Vet. Q. doi:
Horses. Sport Horse Publications, Evans, D.L., 1994. Training Thoroughbred
10.1080/01652176.2011.571381
Saskatoon, pp. 31–43. racehorses. In: Hodgson, D.R., Rose, R.J.
De Kock, S.S., Rodgers, J.P., Swanepoel, B.C., (Eds.), The Athletic Horse: Principles and
Couroucé, A., Geffroy, O., Barrey, E., Auvinet, Guthrie, A.J., 2001. Administration of
B., Rose, R.J., 1999. Comparison of exercise practice of equine sports medicine. WB
bovine, porcine and equine growth Saunders, Philadelphia, pp. 393–397.
tests in French trotters under training track, hormone to the horse: effect on insulin-
racetrack and treadmill conditions. Equine Evans, D.L., Harris, R.C., Snow, D.H., 1993.
like growth factor-1 and selected IGF
Vet. J. 30 (Suppl.), 528–532. Correlation of racing performance with
binding proteins. J. Endocrinol. 171,
Dart, A.J., Cries, L., Jeffcott, L.B., Hodgson, blood lactate and heart rate after exercise in
163–171.
D.R., Rose, R.J., 2002. The effect of equine Thoroughbred horses. Equine Vet. J. 25 (5),
Derman, K.D., Noakes, T.D., 1994. 441–445.
recombinant growth hormone on second Comparative aspects of exercise physiology.
intention wound healing in horses. Vet. In: Hodgson, D.R., Rose, R.J. (Eds.), The Evans, D.L., Rose, R.J., 1986. Method of
Surg. 31, 314–319. Athletic Horse: Principles and practice of investigation of the accuracy of four
Davie, A.J., Evans, D.L., Hodgson, D.R., Rose, equine sports medicine. WB Saunders, digitally-displaying heart rate meters
R.J., 1994. The effects of an oral glucose Philadelphia, pp. 13–25. suitable for use in the exercising horse.
polymer on muscle glycogen resynthesis in Equine Vet. J. 18 (2), 129–132.
Dowling, B.A., Dart, A.J., Hodgson, D.R., Rose,
Standardbred horses. J. Nutr. 124, R.J., Walsh, W.R., 2002a. Recombinant Evans, D.L., Rose, R.J., 1987. Maximum
2740S–2741S. equine growth hormone does not affect the oxygen uptake in racehorses: changes with
Davie, A.J., Evans, D.L., Hodgson, D.R., Rose, in vitro biomechanical properties of equine training state and prediction from
R.J., 1995. Effects of intravenous dextrose superficial digital flexor tendon. Vet. Surg. submaximal cardiorespiratory
infusion on muscle glycogen resynthesis 31, 325–330. measurements. In: Gillespie, J.R., Robinson,
after intense exercise. Equine Vet. J. 18 N.E., (Eds.), Equine exercise physiology 2.
Dowling, B.A., Dart, A.J., Hodgson. D.R., Rose,
(Suppl.), 195–198. ICEEP Publications, Davis, California,
R.J., Walsh, W.R., 2002b. The effect of
pp. 52–67.
Davie, A.J., Evans, D.L., Hodgson, D.R., Rose, recombinant equine Growth Hormone on
R.J., 1996. Effects of glycogen depletion on the biomechanical properties of healing Evans, D.L., Rose, R.J., 1988a. Determination
high intensity exercise performance and superficial digital flexor tendons in horses. and repeatability of maximum oxygen
glycogen utilization rates. Pferdeheilkunde Vet. Surg. 31, 320–324. uptake and other cardiorespiratory
12 (4), 482–484. Eaton, M.D., 1994. Energetics and measurements in the exercising horse.
performance. In: Hodgson, D.R., Rose, R.J. Equine Vet. J. 20 (2), 94–98.
De Graaf-Roelfsema, E., Keizer, H.A., van
Breda, E., Wijnberg, I.D., van der Kolk, J.H., (Eds.), The Athletic Horse: Principles and Evans, D.L., Rose, R.J., 1988b. Effect of a
2008. The influences of overtraining on the practice of equine sports medicine. WB respiratory gas collection mask on some
hypothalamo-pituitaryadrenocortical axis Saunders, Philadelphia, pp. 49–61. measurements of cardiovascular and
in Standardbred horses. In: Lindner, A. Eaton, M.D., Evans, D.L., Hodgson, D.R., Rose, respiratory function in horses exercising on
(Eds.), The acute poorly performing sport R.J., 1995a. Effect of treadmill incline and a treadmill. Res. Vet. Sci. 44 (2), 220–225.
horse, Proc. CESMAS, Utrecht. Wageningen speed on metabolic rate during exercise in Ferrante, P.L., 1995. Lactate measurement and
Academic Publishers, The Netherlands, pp. Thoroughbred horses. J. Appl. Physiol. 79 interpretation: blood vs. plasma. Equine
159–162. (3), 951–957. Vet. J. 18 (Suppl.), 478–479.
De Graaf-Roelfsema, E., Keizer, H.A., van Eaton, M.D., Evans, D.L., Hodgson, D.R., Rose, Ferrante, P.L., Kronfeld, D.S., 1994. Effect of
Breda, E., Wijnberg, I.D., van der Kolk, J.H., R.J., 1995b. Maximum accumulated oxygen sample handling on measurement of
2012. Effects of intensified training and deficit in Thoroughbred horses. J. Appl. plasma glucose and blood lactate
subsequent reduced training on glucose Physiol. 78 (4), 1564–1568. concentrations in horses before and after
metabolism rate and peripheral insulin Eaton, M.D., Hodgson, D.R., Evans, D.L., Rose, exercise. Am. J. Vet. Res. 55 (11),
sensitivity in Standardbreds. Am. J. Vet. Res. R.J., 1999. Effects of low- and moderate- 1497–1500.
73, 1386–1393. intensity training on metabolic responses Frape, D.L., 1988. Dietary requirements and
De Graaf-Roelfsema, E., Tharasanit, T., van to exercise in Thoroughbreds. Equine Vet. J. athletic performance of horses. Equine Vet.
Dam, K.G., Keizer, H.A., van Breda, E., 30 (Suppl.), 521–527. J. 20 (3), 163–172.

437
 18 Metabolic energetics

Fredricson, I., Drevemo, S., Dalin, G., Hjerten, with identical draught resistances. Vet. Rec. Hodgson, D.R., Rose, R.J., 1994a. Evaluation
G., Bjorne, K., Rynde, R., Franzen, G., 1983. 140, 525–528. of performance potential. In: Hodgson,
Treadmill for equine locomotion analysis. Griffin, T.M., Kram, R., Wickler, S.J., Hoyt, D.R., Rose, R.J. (Eds.), The Athletic Horse:
Equine Vet. J. 15, 111–115. D.F., 2004. Biomechanical and energetic Principles and Practice of Equine Sports
Gabel, A.A., Milne, D.W., Muir, W.W., Skarda, determinants of the walk-trot transition in Medicine. WB Saunders, Philadelphia,
R.T., Weingold, M.F., 1983. Some horses. J. Exp. Biol. 207, 4215–4223. pp. 231–243.
physiological responses of Standardbred Guirnalda, P.D., Malinowski, K., Roegner, V., Hodgson, D.R., Rose, R.J., 1994b. Training
horses to a submaximal exercise test Horohov, D.W., 2001. Effects of age and regimens: physiologic adaptations to
following conventional and interval recombinant equine somatotropin (eST) training. In: Hodgson, D.R., Rose, R.J.
training. In: Snow, D.H., Persson, S.G.B., administration on immune function in (Eds.), The Athletic Horse: Principles and
Rose, R.J., (Eds.), Equine exercise female horses. J. Anim. Sci. 79, 2651–2658. practice of equine sports medicine. WB
physiology. Granta Editions, Cambridge Guy, P.S., Snow, D.H., 1977. The effect of Saunders, Philadelphia, pp. 379–385.
UK, pp. 497–504. training and detraining on muscle Hodgson, D.R., Rose, R.J., Allen, J.R., 1983.
Gansen, S., Lindner, A., Marx, S., Mosen, H., composition in the horse. J. Physiol. 269, Muscle glycogen depletion and repletion
Sallmann, H.-P., 1999. Effects of 33–51. patterns in horses performing various
conditioning horses with lactate-guided Guyton, A.C., 1986. Textbook of medical distances of endurance exercise. In: Snow,
exercise on muscle glycogen content. physiology, seventh ed. WB Saunders, D.H., Persson, S.G.B., Rose, R.J. (Eds.),
Equine Vet. J. 30 (Suppl.), 329–331. Philadelphia. Equine exercise physiology. Granta
Geelen, S.N., Jansen, W.L., Geelen, M.J., Sloet Hamlin, M.J., Shearman, J.P., Hopkins, W.G., Editions, Cambridge, UK, pp. 229–236.
van Oldruitenborgh-Oosterbaan, M.M., 2002. Changes in physiological parameters Hodgson, D.R., Rose, R.J., Allen, J.R.,
Beynen, A.C., 2000. Lipid metabolism in in overtrained Standardbred racehorses. Dimauro, J., 1984. Glycogen depletion
equines fed a fat-rich diet. Int. J. Vitam. Equine Vet. J. 34, 383–388. patterns in horses performing maximal
Nutr. Res. 70, 148–152. Harkins, J.D., Beadle, R.E., Kamerling, S.G., exercise. Res. Vet. Sci. 36, 169–173.
Geelen S.N., Jansen W.L., Sloet van 1993. The correlation of running ability Holopherne, D.L.M., Hodgson, D.R., Rose,
Oldruitenborgh-Oosterbaan M.M., and physiological variables in R.J., 1999. Investigation of the accuracy of
Breukink H.J., Beynen, A.C., 2001. Fat Thoroughbred racehorses. Equine Vet. J. 25 a new heart rate meter for use in exercising
feeding increases equine heparin-released (1), 53–60. horses. J. Equine Vet. Sci. 19 (9), 552.
lipoprotein lipase activity. J. Vet. Intern. Harkins, J.D., Kamerling, S.G., Bagwell, C.A., Hörnicke, H., Meixner, R., Pollmann, U., 1983.
Med. 15, 478–481. Karns, P.A., 1990. A comparative study of Respiration in exercising horses. In: Snow,
Geor, R.J., McCutcheon, L.J., Ecker, G.L., interval and conventional training in D.H., Persson, S.G.B., Rose, R.J. (Eds.),
Lindinger, M.I., 1995. Thermal and Thoroughbred racehorses. Equine Vet. J. 9 Equine exercise physiology. Granta
cardiorespiratory responses of horses to (Suppl.), 14–19. Editions, Cambridge, UK, pp. 7–16.
submaximal exercise under hot and humid Harris, P., Graham-Thiers, P.M., 1999. To Hoyt, D.F., Taylor, C.R., 1981. Gait and the
conditions. Equine Vet. J. 20, 125–132. evaluate the influence that ‘feeding state’ energetics of locomotion in horses. Nature
Geor, R.J., McCutcheon, L.J., Shen, H., 1999. may exert on metabolic and physiological 292, 239–240.
Muscular and metabolic responses to responses to exercise. Equine Vet. J. 30 Hoyt, D.F., Wickler, S.J., Dutto, D.J., Catterfeld,
moderate-intensity short-term training. (Suppl.), 633–635. G.E., Johnsen, D., 2006. What are the
Equine Vet. J. 30 (Suppl.), 311–317. Heck, H.A., Mader, G., Hess, S., Mücke, R. relations between mechanics, gait
Gerard, M.P., Hodgson, D.R., Lambeth, R.R., M.üller, Hollmann, W., 1985. Justification parameters, and energetics incterrestrial
Ray, S.P., Rose, R.J., 2002. Effects of of the 4-mmol/l lactate threshold. Int. J. locomotion? J. Exp. Zool. A Comp. Exp.
somatotropin and training on indices of Sports Med. 6, 117–130. Biol. 305, 912–922.
exercise capacity in Standardbreds. Equine Hill, G., Atkins, R., Littlejohn, A., Kruger, J.M., Hug, E., 2003. Training modalities: over-
Vet. J. 34 (Suppl.), 496–501. Bowles, F., 1977. Exercise studies in horses: reaching and over-training in athletes,
Golland, L.C., Evans, D.L., Stone, G.M., 1. A simple telemetry system for recording including a study of the role of hormones.
Tyler-McGowan, C.M., Hodgson, D.R., exercise ECGs in horses. Equine Vet. J. 9 Best Pract. Res. Clin. Endocrinol. Metab.
Rose, R.J., 1999. Plasma cortisol and (2), 72–74. 17, 199–209.
ß-endorphin concentrations in trained and Hinchcliff, K.W., Kaneps, A.J., Geor, R.J., 2004. Jansen, W.L., Geelen, S.N., Van der Kuilen, J.,
overtrained Standardbred racehorses. Equine sports medicine and surgery: basic Beynen, A.C., 2002. Dietary soyabean oil
Europ. J. Physiol. 439, 11–17. and clinical sciences of the equine athlete. depresses the apparent digestibility of fibre
Gonzalez-Alonso, J., Teller, C., Andersen, S.L., WB Saunders, Edinburgh. in trotters when substituted for an
Jensen, F.B., Hyldig, T., Nielsen, B., 1999. Hintz, H.F., 1983. Nutritional requirements of iso-energetic amount of corn starch or
Influence of body temperature on the the exercising horse – a review. In: Snow, glucose. Equine Vet. J. 34, 302–305.
development of fatigue during prolonged D.H., Persson, S.G.B., Rose, R.J. (Eds.), Jones, J.H., Ohmura, H., Stanley, S.D., Hiraga,
exercise in the heat. J. Appl. Physiol. 86, Equine exercise physiology. Granta A., 2006. Energetic cost of locomotion on
1032–1039. Editions, Cambridge, UK, pp. 275–290. different equine treadmills. Equine Vet. J.
Gore, C.J., Withers, R.T., 1990. Effect of Hintz, H.F., 1994. Nutrition and equine 36 (Suppl.), 365–369.
exercise intensity and duration on performance. J. Nutr. 124, 2723S–2729S. Julen Day, T.R., Potter, G.D., Morris, E.L.,
postexercise metabolism. J. Appl. Physiol. Hodgson, D.R., 1985. Energy considerations Greene, L.W., Simmons, J.B., 1998.
68 (6), 2362–2368. during exercise. Vet. Clin. N. Am.: Equine Physiologic and skeletal response to
Gottlieb-Vedi, M., Essén-Gustavsson, B., Pract. 1 (3), 447–460. exogenous equine somatotropin (eST) in
Persson, S.G.B., 1991. Draught load and Hodgson, D.R., Davis, R.E., McConaghy, F.F., two-year-old Quarter Horses in race
speed compared by submaximal tests on a 1994. Thermoregulation in the horse in training. J. Equine Vet. Sci. 18, 321–328.
treadmill. In: Persson, S.G.B., Lindholm, A., response to exercise. Br. Vet. J. 150, Katz, L.M., Bayly, W.M., Roeder, M.J., Kingston,
Jeffcott, L.B. (Eds.), Equine exercise 219–235. J.K., Hines, M.T., 2000. Effects of training
physiology 3. ICEEP Publications, Davis, Hodgson, D.R., McCutcheon, L.J., Byrd, S.K., on maximum oxygen consumption of
California, pp. 92–96. Brown,W.S., Bayly, W.M., Brengelmann, ponies. Am. J. Vet. Res. 61, 986–991.
Gottlieb-Vedi, M., Lindholm, A., 1997. G.L., et al., 1993. Dissipation of metabolic Knight, P.K., Sinha, A.K., Rose, R.J., 1991.
Comparison of Standardbred trotters heat in the horse during exercise. J. Appl. Effects of training intensity on maximum
exercising on a treadmill and a race track Physiol. 74, 1161–1170. oxygen uptake. In: Persson, S.G.B.,

438
 References

Lindholm, A., Jeffcott, L.B. (Eds.), Equine electrolytes in muscles of Standardbred McMiken, D.F., 1983. An energetic basis of
exercise physiology 3. ICEEP Publications, horses. Acta Vet. Scand. 15, 287–309. equine performance. Equine Vet. J. 15 (2),
Davis, California, pp. 77–82. Lindner, A.E., 2010. Maximal lactate steady 123–133.
Kobayashi, M., Kuribara, K., Amada, A., 1999. state during exercise in blood of horses. Mebø, J.I., Mohn, A.-C., Tabata, I., Bahr, R.,
Application of V200 values for evaluation J. Anim. Sci. 88, 2038–2044. Vaage, O., Sejersted, O.M., 1988. Anaerobic
of training effects in the young Lucke, J.N., Hall, G.N., 1980. Further studies capacity Determined by maximum
Thoroughbred under field conditions. on the metabolic effects of long distance accumulated O2 deficit. J. Appl. Physiol. 64
Equine Vet. J. 30 (Suppl.), 159–162. riding: Golden Horseshoe Ride 1979. (1), 50–60.
Kohn, C.W., Hinchcliff, K.W., McKeever, K.H., Equine Vet. J. 12 (4), 189–192. Meyer, R.A., Dudley, G.A., Terjung, R.L., 1980.
1999. Total body washing with cool water Malinowski, K., Christensen, R.A., Konopka, Ammonia and IMP in different skeletal
facilitates heat dissipation in horses A., Scanes, C.G., Hafs, H.D., 1997. Feed muscle fibers after exercise in rats. J. Appl.
exercised in hot, humid conditions. Am. J. intake, body weight, body condition score, Physiol. (Respiration, Environment and
Vet. Res. 60, 299–305. musculation, and immunocompetence in Exercise Physiology) 49 (6), 1037–1041.
Koho N.M., Hyyppä S., Pösö A.R., 2006. aged mares given equine somatotropin. Miller-Graber, P.A., Lawrence, L.M., Foreman,
Monocarboxylate transporters (MCT) as J. Anim. Sci. 75, 755–760. J.H., Bump, K.D., Fisher, M.G., Kurcz, E.V.,
lactate carriers in equine muscle and red Marc, M., Parvizi, N., Ellendorff, F., Kallweit, 1991. Dietary protein level and energy
blood cells. Equine Vet. J. 36 (Suppl.), E., Elsaesser, F., 2000. Plasma cortisol and metabolism during treadmill exercise in
354–358. ACTH concentrations in the warmblood horses. J. Nutr. 121 (9), 1462–1469.
Kulinski, K.M., Thompson Jr., D.L., Capshaw, horse in response to a standardized Milne, D.W., Gabel, A.A., Muir, W.W., Skarda,
E.L., French, D.D., Oliver, J.L., 2002. Daily treadmill exercise test as physiological R.T., 1977. Effects of training on heart rate,
treatment of growing foals with equine markers for evaluation of training status. cardiac output, and lactic acid in
somatotropin: Pathologic and J. Anim. Sci. 78, 1936–1946. Standardbred horses, using a standardized
endocrinologic assessments at necropsy Marlin, D., Nankervis, K.J., 2002. Equine exercise test. J. Equine Med. Surg. 1 (4),
and residual effects in live animals. exercise physiology. Blackwell Science, UK. 131–135.
J. Anim. Sci. 80, 392–400. Nimmo, M.A., Snow, D.H., 1983. Changes in
McConaghy, F.F., Hales, J.R.S., Rose, R.J.,
Lacombe, V., Hinchcliff, K.W., Geor, R.J., Hodgson, D.R., 1995. Selective brain muscle glycogen, lactate and pyruvate
Lauderdale, M.A., 1999. Exercise that cooling in the horse during exercise and concentrations in the thoroughbred horse
induces substantial muscle glycogen environmental heat stress. J. Appl. Physiol. following maximal exercise. In: Snow, D.H.,
depletion impairs subsequent anaerobic 79, 1849–1854. Persson, S.G.B., Rose, R.J. (Eds.), Equine
capacity. Equine Vet. J. 30 (Suppl.), exercise physiology. Granta Editions,
McConaghy, F.F., Hales, J.R.S., Rose, R.J.,
293–297. Cambridge, UK, pp. 237–244.
Hodgson, D.R., 2002. Thermoregulatory-
Lacombe, V.A., Hinchcliff, K.W., Devor, S.T., induced compromise of muscle blood flow Noble, G.K., Houghton, E., Roberts, C.J.,
2003. Effects of exercise and glucose in ponies during intense exercise in the Faustino-Kemp, J., de Kock, S.S.,
administration on content of insulin- heat: a contributor to the onset of fatigue. Swanepoel, B.C., Sillence, M.N., 2007.
sensitive glucose transporter in equine Equine Vet. J. 34, 491–495. Effect of exercise, training, circadian
skeletal muscle. Am. J. Vet. Res. 64, rhythm, age, and sex on insulin-like growth
McCutcheon, L.J., Geor, R.J., 2000. Influence
1500–1506. factor-1 in the horse. J. Anim. Sci. 85,
of training on sweating responses during
Lacombe, V.A., Hinchcliff, K.W., Kohn, C.W., submaximal exercise in horses. J. Appl. 163–171.
Devor, S.T., Taylor, L.E., 2004. Effects of Physiol. 89, 2463–2471. Nout, Y.S., Hinchcliff, K.W., Jose-Cunilleras, E.,
feeding meals with various soluble Dearth, L.R., Sivko, G.S., DeWille, J.W.,
McCutcheon, L.J., Geor, R.J., 2004.
carbohydrate content on muscle glycogen 2003. Effect of moderate exercise
Thermoregulation and exercise-associated
synthesis after exercise in horses. Am. J. Vet. immediately followed by induced
heat illnesses. In: Hinchcliff, K.W., Kaneps,
Res. 65, 916–923. hyperglycemia on gene expression and
A.J., Geor, R.J. (Eds.), Equine sports
Lauderdale, M.A., Hinchcliff, K.W., 1999. medicine and surgery. WB Saunders, content of the glucose transporter-4 protein
Hyperbolic relationship between time-to- Philadelphia, pp. 919–936. in skeletal muscles of horses. Am. J. Vet.
fatigue and workload. Equine Vet. J. 30 Res. 11, 1401–1408.
McDonough, P., Kindig, C.A., Ramsel, C.,
(Suppl.), 586–590. Nybo, L., Nielsen, B., 2001. Hyperthermia and
Poole, D.C., Erickson, H.H., 2002. The
Lawrence, L.M., 1990. Nutrition and fuel effect of treadmill incline on maximal central fatigue during prolonged exercise in
utilization in the athletic horse. Vet. Clin. oxygen uptake, gas exchange and the humans. J. Appl. Physiol. 91, 1055–1060.
N. Am.: Equine Pract. 6 (2), 393–418. metabolic response to exercise in the horse. Oscai, L.B., Williams, B.T., Hertig, B.A., 1968.
Lawrence, L.M., Hintz, H.F., Soderholm, L.V., Exp. Physiol. 87, 499–506. Effect of exercise on blood volume. J. Appl.
Williams, J., Roberts, A.M., 1995. Effect of McKeever, K.H., Malinowski, K., Christensen, Physiol. 24 (5), 622–624.
time of feeding on metabolic response to R.A., Hafs, H.D., 1998. Chronic Pagan, J.D., Essén-Gustavsson, B., Lindholm,
exercise. Equine Vet. J. 18 (Suppl.), recombinant equine somatotropin (eST) A., Thornton, J., 1987. The effect of dietary
392–395. administration does not affect aerobic energy source on exercise performance in
Leach, D.H., Sprigings, E., 1979. Gait fatigue capacity or exercise performance in geriatric Standardbred horses. In: Gillespie, J.R.,
in the racing Thoroughbred. J. Equine Med. mares. Vet. J. 155, 19–25. Robinson, N.E. (Eds.), Equine exercise
Surg. 3, 436–443. McKeever, K.H., Schurg, W.A., Convertino, physiology 2. ICEEP Publications, Davis,
Lindholm, A., 1979. Substrate utilization and V.A., 1985. Exercise training-induced California, pp. 686–700.
muscle fiber types in Standardbred trotters hypervolemia in greyhounds: role of water Pagan, J.D., Harris, P.A., 1999. The effects of
during exercise. Proc. Am. Assoc. Equine intake and renal mechanisms. Am. J. timing and amount of forage and grain on
Pract. 25, 329–336. Physiol. (Regulatory, Integrative and exercise response in Thoroughbred horses.
Lindholm, A., Bjerneld, H., Saltin, B., 1974. Comparative Physiology) 248 (17), Equine Vet. J. 30 (Suppl.), 451–457.
Glycogen depletion pattern in muscle fibres R422–R425. Pagan, J.D., Hintz, H.F., 1986a. Equine
of trotting horses. Acta Physiol. Scand. 90, McKeever, K.H., Schurg, W.A., Jarrett, energetics. 1. Relationship between body
475–484. S.H., Convertino, V.A., 1987. Exercise weight and energy requirements in horses.
Lindholm, A., Piehl, K., 1974. Fibre training-induced hypervolemia in the J. Anim. Sci. 63 (3), 815–821.
composition, enzyme activity and horse. Med. Sci. Sports Exerc. 19 (1), Pagan, J.D., Hintz, H.F., 1986b. Equine
concentrations of metabolites and 21–27. energetics. 2. Energy expenditure in horses

439
 18 Metabolic energetics

during submaximal exercise. J. Anim. Sci. and duration to training-linked myosin Rose, R.J., Knight, P.K., Bryden, W.L., 1991.
63 (3), 822–830. transitions in thoroughbreds. Equine Vet. J. Energy use and cardiorespiratory responses
Persson, S.G.B., 1967. On blood volume and 36 (Suppl.), 311–315. to prolonged submaximal exercise. In:
working capacity. Acta. Vet. Scand. 19 Rivero, J.L., van Breda, E., Rogers, C.W., Persson, S.G.B., Lindholm, A., Jeffcott, L.B.
(Suppl.), 1–189. Lindner, A., van Oldruitenborgh- (Eds.), Equine exercise physiology 3. ICEEP
Persson, S.G.B., 1968. Blood volume, state of Oosterbaan, M.M., 2008. Unexplained Publications, Davis, California,
training and working capacity of underperformance syndrome in sport pp. 281–287.
racehorses. Equine Vet. J. 1, 52–58. horses: classification, potential causes and Sapolsky, R.M., Romero, M.L., Munck, A.U.,
Persson, S.G.B., 1983. Evaluation of exercise recognition. Equine Vet. J. 40, 611–618. 2000. How do glucocorticoids influence
tolerance and fitness in the performance Rivero, J.-L.L., Clayton, H.M., 1996. The stress responses? Integrating permissive,
horse. In: Snow, D.H., Persson, S.G.B., potential role of the muscle in kinematic suppressive, stimulatory, and preparative
Rose, R.J. (Eds.), Equine exercise characteristics. Pferdeheilkunde 12 (4), actions. Endocr. Rev. 21, 55–89.
physiology. Granta Editions, Cambridge, 635–640. Scott, C.B., Roby, F.B., Lohman, T.G., Bunt,
UK, pp. 441–457. Rivero, J.L.L., Ruz, A., Martí-Korff, S., Estepa, J.C., 1991. The maximally accumulated
Persson, S.G.B., Larsson, M., Lindholm, A., J.C., Aguilera-Tejero, E., Werkman, J., oxygen deficit as an indicator of anaerobic
1980. Effects of training on adreno-corticol Sobotta, M., Lindner, A., 2007. Effects of capacity. Med. Sci. Sports Exerc. 23 (5),
function and red-cell volume in trotters. intensity and duration of exercise on 618–624.
Zbl. Vet. Med. 27, 261–268. muscular responses to training of Seeherman, H.J., Morris, E.A., 1990.
Pfau, T., Spence, A., Starke, S., Ferrari, M., Thoroughbred racehorses. J. Appl. Physiol. Methodology and repeatability of a
Wilson, A., 2009. modern riding style 102, 1871–1882. standardised treadmill exercise test for
improves horse racing times. Science 325, Rivero, J.L.L., Serrano, A.L., Quiroz-Rothe, E., clinical evaluation of fitness in horses.
289. Aguilera-Tejero, E., 2001. Co-ordinated Equine Vet. J. 9 (Suppl.), 20–25.
Pinchbeck, G.L., Clegg, P.D., Proudman, C.J., changes of kinematics and muscle fibre Self, Z.T., Spence, A.J., Wilson, A.M., 2012.
Morgan, K.L., French, N.R., 2004. Whip use properties with prolonged endurance Speed and incline during Thoroughbred
and race progress are associated with horse training. Equine Vet. J. 33 (Suppl.), horse racing: racehorse speed supports a
falls in hurdle and steeplechase racing in 104–108. metabolic power constraint to incline
the UK. Equine Vet. J. 36, 384–389. Robb, J., Harper, R.B., Hintz, H.F., Reid, J.T., running but not to decline running.
Popot, M.A., Bobin, S., Bonnaire, Y., Delahaut, Lowe, J.E., Schryver, H.F., Rhee, M.S.S., J. Appl. Physiol. 13, 602–607.
P.H., Closset, J., 2001. IGF-I plasma 1972. Chemical composition and energy Selye, H., 1951. The general-adaptation-
concentrations in non-treated horses and value of the body, fatty acid composition syndrome. Ann. Rev. Med. 2, 327–342.
horses administered with methionyl equine of adipose tissue, and liver and kidney size Sexton, W.L., Erickson, H.H., 1990. Effects of
somatotropin. Res. Vet. Sci. 71, 167–173. in the horse. Anim. Prod. 14, 25–34. treadmill elevation on heart rate, blood
Pösö, A.R., Lampinen, K.J., Räsänen, L.A., Rose, R.J., Evans, D.L., 1986. Metabolic and lactate concentration and packed cell
1995. Distribution of lactate between red respiratory responses to prolonged volume during graded submaximal exercise
blood cells and plasma after exercise. submaximal exercise in the horse. In: in ponies. Equine Vet. J. 9 (Suppl.), 57–60.
Equine Vet. J. 18 (Suppl.), 231–234. Saltin, B., (Eds.), Biochemistry of exercise Sloet van Oldruitenborgh-Oosterbaan, M.M.,
Pratt, S.E., Geor, R.J., McCutcheon, L.J., 2006. VI. Human Kinetics Publishers, Barneveld, A., 1995. Comparison of the
Effects of dietary energy source and Champaign, Illinois, pp. 459–466. workload of Dutch warmblood horses
physical conditioning on insulin sensitivity Rose, R.J., Hendrickson, D.K., Knight, P.K., ridden normally and on a treadmill. Vet.
and glucose tolerance in Standardbred 1990a. Clinical exercise testing in the Rec. 137, 136–139.
horses. Equine Vet. J. 38 (Suppl 36), normal Thoroughbred racehorse. Aust. Vet. Sloet van Oldruitenborgh-Oosterbaan, M.M.,
579–584. J. 67 (10), 345–348. Clayton, H.M., 1999. Advantages and
Pratt, S.E., Geor, R.J., Spriet, L.L., McCutcheon, Rose, R.J., Hodgson, D.R., 1994a. Clinical disadvantages of track vs. treadmill tests.
L.J., 2007. Time course of insulin sensitivity exercise testing. In: Hodgson, D.R., Rose, Equine Vet. J. 30 (Suppl.), 645–647.
and skeletal muscle glycogen synthase R.J. (Eds.), The Athletic Horse: Principles Smith, L.A., Thompson Jr., D.L., French, D.D.,
activity after a single bout of exercise in and Practice of Equine Sports Medicine. Leise, B.S., 1999. Effects of recombinant
horses. J. Appl. Physiol. 103, 1063–1069. WB Saunders, Philadelphia, pp. 245–257. equine somatotropin on wound healing,
Preedy, D.F., Colborne, G.R., 2001. A method Rose, R.J., Hodgson, D.R., 1994b. An overview carbohydrate and lipid metabolism, and
to determine mechanical energy of performance and sports medicine. In: endogenous somatotropin responses to
conservation and efficiency in equine gait: Hodgson, D.R., Rose, R.J. (Eds.), The secretagogues in geldings. J. Anim. Sci. 77,
a preliminary study. Equine Vet. J. 33 Athletic Horse: Principles and practice of 1815–1822.
(Suppl.), 94–98. equine sports medicine. WB Saunders, Snow, D.H., 1994. Ergogenic aids to
Price, T.B., Rothman, D.L., Taylor, R., Avison, Philadelphia, pp. 3–11. performance in the racehorse: nutrients or
M.J., Shulman, G.I., Shulman, R.G., 1994. Rose, R.J., Hodgson, D.R., Bayly, W.M., drugs. J. Nutr. 124, 2730S–2735S.
Human muscle glycogen resynthesis after Gollnick, P.D., 1990b. Kinetics of VO2 and Snow, D.H., Baxter, P., Rose, R.J., 1981. Muscle
exercise: insulin-dependent and VCO2 in the horse and comparison of five fibre composition and glycogen depletion
-independent phases. J. Appl. Physiol. 76 methods for determination of maximum in horses competing in an endurance ride.
(1), 104–111. oxygen uptake. Equine Vet. J. 9 (Suppl.), Vet. Rec. 108, 374–378.
Quiroz-Rothe, E., Novales, M., Aguilera-Tejero, 39–42. Snow, D.H., Harris, R.C., Harman, J.C., Marlin,
E., Rivero, J.L., 2002. Polysaccharide storage Rose, R.J., Hodgson, D.R., Kelso, T.B., et al., D.J., 1987. Glycogen repletion following
myopathy in the M. longissimus lumborum 1988. Maximum O2 uptake, O2 debt and different diets. In: Gillespie, J.R., Robinson,
of showjumpers and dressage horses with deficit, and muscle metabolites in N.E. (Eds.), Equine exercise physiology 2.
back pain. Equine Vet. J. 34, 171–176. Thoroughbred horses. J. Appl. Physiol. ICEEP Publications, Davis, California,
Rivero, J.L., 2007. A scientific background for 64(2), 781–788. pp. 701–710.
skeletal muscle conditioning in equine Rose, R.J., Ilkiw, J.E., Arnold, K.S., Backhouse, Snow, D.H., Kerr, M.G., Nimmo, M.A., Abbott,
practice. J. Vet. Med. A Physiol. Pathol. J.W., Sampson, D., 1980. Plasma E.M., 1982. Alterations in blood, sweat,
Clin. Med. 54, 321–332. biochemistry in the horse during 3-day urine and muscle composition during
Rivero, J.L., Ruz, A., Marti-Korfft, S., Lindner, event competition. Equine Vet. J. 12 (3), prolonged exercise in the horse. Vet. Rec.
A., 2006. Contribution of exercise intensity 132–136. 110, 377–384.

440
 References

Stewart-Hunt, L., Geor, R.J., McCutcheon, L.J., hydrogenosomes. Springer-Verlag, Berlin, Wasserman, K., Whipp, B.J., Koyal, S.N.,
2006. Effects of short-term training on pp. 85–103. Beaver, W.L., 1973. Anaerobic threshold
insulin sensitivity and skeletal muscle Topliff, D.R., Potter, G.D., Dutson, T.R., and respiratory gasexchange during
glucose metabolism in Standardbred Kreider, J.L., Jessup, G.T., 1983. Diet exercise. J. Appl. Physiol. 35, 236–243.
horses. Equine Vet. J. 38 (Suppl 36), manipulation and muscle glycogen in the Westermann, C.M., Dorland, B., De Sain-van
226–232. equine. In: Proc. 8th Equine Nutr. Physiol. der Velden, M.G., Wijnberg, I.D., Van
Stryer, L., 1988. Biochemistry, third ed. WH Symp, Lexington, Kentucky, pp. 119–124. Breda, E., De Graaf-Roelfsema, E., et al.,
Freeman and Company, New York. Topliff, D.R., Potter, G.D., Kreider, J.L., 2008a. Plasma acylcarnitine and fatty acid
Taylor, C.R., Heglund, N.C., McMahon, T.A., Dutson, T.R., Jessup, G.T., 1985. Diet profiles during exercise and training in
Looney, T.R., 1980. Energetic cost of manipulation, muscle glycogen metabolism Standardbreds. Am. J. Vet. Res. 69,
generating muscular force during running: and anaerobic work performance in the 1469–1475.
a comparison of large and small animals. equine. In: Proc. 9th Equine Nutr. Physiol. Westermann, C.M., Dorland, L., Votion, D.M.,
J. Exp. Biol. 86, 9–18. Symp. East Lansing, Michigan, De Sain-van der Velden, M.G., Wijnberg,
Thatcher, C.E., Thompson Jr., D.L., 2002. pp. 224–229. I.D., Wanders, R.J., et al., 2008b. Acquired
Effects of frequency of treatment with Trilk, J.L., Lindner, A.J., Greene, H.M., multiple Acyl-CoA dehydrogenase
recombinant equine somatotropin on Alberghina, D., Wickler, S.J., 2002. A deficiency in 10 horses with atypical
selected biological responses in geldings. lactate-guided conditioning programme to myopathy. Neuromuscul. Disord. 18,
Domest. Anim. Endocrinol. 22, 127–143. improve endurance performance. Equine 355–364.
Thomas, D.P., Fregin, G.F., 1981. Vet. J. 34 (Suppl.), 122–125. Westermann, C.M., Dorland, L., Wijnberg,
Cardiorespiratory and metabolic responses Urhausen, A., Gabriel, H., Kindermann, W., I.D., De Sain-van der Velden, M.G., Van
to treadmill exercise in the horse. J. Appl. 1995. Blood hormones as markers of Breda, E., Barneveld, A., et al., 2010. Amino
Physiol. (Respiratory, Environmental and training stress and overtraining. Sports acid profile during exercise and training in
Exercise Physiology) 50 (4), 864–868. Med. 20, 251–276 Standardbreds. Res Vet. Sci. (in press).
Thompson Jr., D.L., DePew, C.L., Ortiz, A., Väihkönen, L.K., Hyyppä, S., Pösö, A.R., 1999. Wickler, S.J., Hoyt, D.F., Biewener, A.A.,
Sticker, L.S., Rahmanian, M.S., 1994. Factors affecting accumulation of lactate in Cogger, E.A., De La Paz, K.L., 2005. In vivo
Growth hormone and prolactin red blood cells. Equine Vet. J. 30 (Suppl.), muscle function vs speed. II. Muscle
concentrations in plasma of horses: sex 443–447. function trotting up an incline. J. Exp. Biol.
differences and the effects of acute exercise Väihkönen, L.K., Pösö, A.R., 1998. 208, 1191–1200.
and administration of growth hormone- Interindividual variation in total and Wickler, S.J., Hoyt, D.F., Clayton, H.M.,
releasing hormone. J. Anim. Sci. 72, carrier-mediated lactate influx into red Mullineaux, D.R., Cogger, E.A., Sandoval,
2911–2918. blood cells. Am. J. Physiol. (Regulatory, E., et al., 2004. Energetic and kinematic
Thompson Jr., D.L., Rahmanian. M.S., DePew, Intergrative and Comparative Physiology) consequences of weighting the distal limb.
C.L., Burleigh, D.W., DeSouza, C.J., 274 (43), R1025–R1030. Equine Vet. J. 36, 772–777.
Colborn, D.R., 1992. Growth hormone in Valette, J.P., Barrey, E., Jouglin, M., Courouce, Wickler, S.J., Hoyt, D.F., Cogger, E.A., Hall,
mares and stallions: pulsatile secretion, A., Auvinet, B., Flaux, B., 1999. K.M., 2001. Effect of load on preferred
response to growth hormone-releasing Standardisation of muscular biopsy of speed and cost of transport. J. Appl.
hormone, and effects of exercise, sexual gluteus medius in French trotters. Equine Physiol. 90, 1548–1551.
stimulation, and pharmacological agents. Vet. J. 30 (Suppl.), 342–344. Wideman, L., Weltman, J.Y., Hartman, M.L.,
J. Anim. Sci. 70, 1201–1207. Van Dam, K.G., Van Breda, E., Schaart, G., Van Veldhuis, J.D., Weltman, A., 2002. Growth
Thornton, J., Essén-Gustavsson, B., Lindholm, Ginneken, M.M., Wijnberg, I.D., De hormone release during acute and chronic
A., McMiken, D., Persson, S., 1983. Effects Graaf-Roelfsema, E., et al., 2004. aerobic and resistance exercise: recent
of training and detraining an oxygen Investigation of the expression and findings. Sports Med. 32, 978–1004.
uptake, cardiac output, blood gas tensions, localization of glucose transporter 4 and Wijnberg, I.D., Franssen, H., Van der Kolk,
pH and lactate concentrations during and fatty acid translocase/CD36 in equine J.H., Breukink, H.J., 2002. Quantitative
after exercise in the horse. In: Snow D.H., skeletal muscle. Am. J. Vet. Res. 65, motor unit action potential analysis of
Persson, S.G.B., Rose, R.J. (Eds.), Equine 951–956. skeletal muscles in the Warmblood horse.
exercise physiology. Granta Editions, Van der Kolk, J.H., Wensing, T., Kalsbeek, Equine Vet. J. 34, 556–561.
Cambridge, UK, pp. 470–486. H.C., Breukink, H.J., 1995. Lipid Wijnberg, I.D., Van Dam, K.G., Graaf-
Thornton, J., Pagan, J., Persson, S.G.B., 1987. metabolism in horses with Roelfsema, E., de, Keizer, H.A., Van
The oxygen cost of weight loading and hyperadrenocorticism. J. Am. Vet. Med. Ginneken, M.M., Barneveld, A., et al.,
inclined treadmill exercise in the horse. In: Assoc. 206, 1010–1012. 2008. (Over)training effects on quantitative
Gillespie, J.R., Robinson, N.E. (Eds.), Veldhuis, J.D., Keenan, D.M., Pincus, S.M., electromyography and muscle enzyme
Equine exercise physiology 2. ICEEP 2008. Motivations and methods for activities in Standardbred horses. J. Appl.
Publications, Davis, California, analyzing pulsatile hormone secretion. Physiol. 105, 1746–1753.
pp. 206–215. Endocr. Rev. 29, 823–864. Wijnberg, I.D., Van der Kolk, J.H., Franssen,
Tielens, A.G.M., van.Hellemond, J.J., 2007. Waller, A.P., Lindinger, M.I., 2010. Nutritional H., Breukink, H.J., 2003. Needle
Anaerobic mitochondria: properties and aspects of post exercise skeletal muscle electromyography in the horse compared
origins. In: Martin, W., Müller, M. (Eds.), glycogen synthesis in horses: a comparative with its principles in man: a review. Equine
Origin of mitochondria and review. Equine Vet. J. 42, 274–281. Vet. J. 35, 9–17.

441
This page intentionally left blank
 CHAPTER 19 

Mechanical analysis and scaling

Antonie J. Van den Bogert, Liduin S. Meershoek, At L. Hof

rigid bodies, whereas Joseph Louis Lagrange (France/Italy, 1736–

Mechanical analysis 1813) formulated equations for linked multibody systems which
are the basis of modern robotics. The laws of dynamics have proved
useful to predict motion resulting from known forces (forward
Introduction dynamics), or to estimate the forces that were the cause of an
Visual examination of equine gait has been the main clinical tool observed motion (inverse dynamics). Since Elftman introduced the
for diagnosis of lameness in horses during the past two or three inverse dynamics analysis in biomechanics (1939) it has been
millennia. Similarly, trainers and breeders routinely examine gait applied successfully in human movement research. Examples can
visually in order to predict sport performance. Recent developments be found in orthopedics (Andriacchi, 1993), in motor control
in gait measurement techniques (Clayton, 1996) have made it pos- (Winter & Eng, 1995) and in sport (Yeadon & Challis, 1994). It
sible to perform gait analysis in a quantitative manner, which results provides an opportunity for interpretation of gait that is not pos-
in better reproducibility, improved temporal and spatial resolution, sible with kinematic or kinetic measurements alone. This opportu-
and is less dependent on the experience and judgment of the clini- nity has also been explored for basic questions on equine gait
cian or trainer. The separate analysis of either movements or forces, (Bartel et al., 1978; Schryver et al., 1978; Clayton et al., 1998).
or kinetics respectively, can enhance our understanding of equine Inverse dynamic analysis treats the locomotor system as a system
gait. However, an even more powerful method is to analyze move- of linked rigid segments, with hypothetical torque (or moment)
ments and forces simultaneously using Newton’s laws of dynamics. motors at each joint representing the action of muscles. The analysis
Kinematic variables, such as limb trajectories, joint angles and combines kinematic and force platform data to derive these joint
angular velocities, have been used frequently in the past two decades moments throughout the movement cycle (Fig. 19.1). The methods
to quantify equine gait (e.g. van Weeren et al., 1993). These vari- have been well established in the literature, both in two dimensions
ables contain essentially the same information that is presented to (Winter, 1990) and in three dimensions (Vaughan, 1984; van den
the eye of a clinician, the only difference being the quantitative Bogert, 1994). Software to perform inverse dynamic analysis on any
nature of the information and the increased spatio-temporal resolu- user-specified locomotor system, allowing for horses as well as
tion. This resolution is important, because equine gait is very repro- humans, is available from the major manufacturers of gait analysis
ducible and small or fast changes in movement, invisible to the equipment.
human eye, are relevant. Until now, however, automated kinematic The moment at the center of rotation of a joint is directly related
gait analysis has not resulted in a better ability to quantify and to the combination of muscle forces acting across the joint. These
characterize lameness, as compared to visual examination by an joint moments by themselves provide useful information about
experienced clinician (Back et al., 1993). However, some success movement. Even more so than the human, the horse, with its four
has been reported in prediction of performance using gait analysis multi-jointed limbs frequently acting as closed-chain mechanisms,
(Deuel & Park, 1993; Back et al., 1995). may redistribute its joint moments without visual gait changes. For
A different type of gait information is obtained by using kinetic instance, a visually identical hind limb extension in late stance may
variables. Kinetic variables are variables related to forces. Internal be accomplished by only hip extensor muscles, only knee extensor
kinetic variables such as tendon forces (Riemersma et al., 1996) and muscles, or any combination of these. Inverse dynamic analysis
bone strain (Hartman et al., 1984; Biewener et al., 1988) have been allows us to ‘see’ these differences in muscle coordination. One
measured in horses. These measurements are, however, invasive and further useful step in inverse dynamic analysis is the quantification
limited to research applications. For routine gait evaluation, the of joint power profiles. This allows investigation of the sources of
only non-invasive kinetic measuring device is the force platform, mechanical power for movement. We will show in this chapter how
which measures forces between hoof and ground. After initial work joint moment and joint power can be used to estimate force and
by Pratt and O’Connor (1976), clinical applications were explored power output in musculotendinous structures.
(Merkens & Schamhardt, 1988; Dow et al., 1991). These studies It must be stressed that results of inverse dynamic analysis are
have shown that, although force plate signals do not always provide sensitive to the protocol for data collection and the methods for
enough information for full diagnostics, they are useful to quantify data analysis. Since the horse is not a set of perfect rigid body seg-
lameness and to indicate the affected limb. ments, and kinematic data are contaminated by measurement errors
Isaac Newton (England, 1642–1727) formulated the basic laws and soft tissue deformation, results can only be an approximation.
of dynamics, establishing a relationship between force, mass and This is acceptable, as long as the methodology is well understood
translational movement: F = m · a. Leonhard Euler (Germany, and used consistently so that results can be compared between
1707–1783) further developed these laws to describe rotation of studies.

443
 19 Mechanical analysis and scaling

Fig 19.1  Inverse dynamic analysis uses

a linked segment model, kinematics
m,l and force platform data to calculate
joint moments, forces and powers. (See
text for details.)

X,Y
M
a, α
P

GRF

An indication of the potential applications of inverse dynamic properties, mass (m), location of the center of mass and moment
analysis can be obtained from examples in the human literature. A of inertia (I), of the segments are constant (Winter, 1990). In
well-known finding is that knee extensor moments during gait are mechanical terms the moment of inertia is the rotational equivalent
significantly reduced after anterior cruciate ligament injury. Inverse to the mass. While the mass determines the resistance to linear
dynamic analysis provides a tool to monitor the patient’s progress accelerations, the moment of inertia determines the resistance to
during rehabilitation and decide when surgical treatment is required angular accelerations. The force needed for a certain (linear) accel-
(Andriacchi & Birac, 1993). Using out-of-sagittal plane joint eration equals the mass times the acceleration (F = m · a). In a
moments, it has been shown that a certain bracing method reduces similar way the moment needed for a certain angular acceleration
the medial compartment loads in the knee joint (Lindenfeld et al., equals the moment of inertia times the angular acceleration (M =
1997). Especially noteworthy is the success of gait analysis, includ- I · α). This moment of inertia is determined by the spatial distribu-
ing inverse dynamic and joint power analysis, in surgical decision tion of the mass within the segment. The mass of the segment
making for children with cerebral palsy (Rose et al., 1993). In equals the total mass between the joints; it represents not only
research, useful information for understanding of injury mecha- the bone but also the soft tissue surrounding it. The mass of
nisms and design of joint replacements is obtained from inverse all segments together is therefore equal to the body mass of the
dynamic analysis of muscle and joint forces during various activities horse. Figure 19.2 gives a graphical representation of a linked-
(Paul, 1971; van den Bogert et al., 1999). Similar clinical applica- segment model.
tions can be found for horses when using inverse dynamics to The action of the muscles is represented by the moment they
analyze equine gait. It has, for example, been used successfully generate around the joint. This moment generating function is
to analyze the influence of heel wedges in horses with tendinitis explained in Figure 19.3. If a muscle is activated it generates a tensile
(Clayton et al., 2000). force at the bones. Due to this force the bones are compressed at
Before such applications are possible, it is imperative that valid the joint. This causes a bone-to-bone force from both bones onto
test protocols are established and normative data are collected each other. So, because of the muscular activity there are two forces
(Ounpuu et al., 1991). Movements of horses are remarkably planar acting on each bone: from the muscle at the attachment site and
and the function of most muscles, especially those in the distal one from the other bone at the joint. Since these two forces are
limbs, is limited to flexion and extension in the sagittal plane. equal in magnitude and opposite in direction they form a couple.
Therefore, two-dimensional analysis is sufficient for most purposes. The moment associated with this couple equals the force times the
This chapter presents in detail the procedure for inverse dynamic distance between the two forces. Since the two forces are equal in
analysis for sagittal plane movement, with specific reference to magnitude and opposite in direction they cancel each other out and
development of protocols for equine applications. Some applica- the muscle action can be represented by the moment (Elftman,
tions require a full three-dimensional analysis. A short introduction 1939). In a similar way the forces exerted by ligaments can also be
of this more complex analysis therefore concludes this chapter. represented by moments. It should be noted that the bone-to-bone
force of Figure 19.3 does not represent the complete joint contact
force; it is only the part of this force that is caused by the activity
Inverse dynamic analysis of the muscle.
Linked segment model
For the inverse dynamic calculations a simplified model of the
Inverse dynamic calculations
horse is used. This model is called a linked-segment model since it In inverse dynamic calculations the movement of the linked
consists of rigid segments, which are linked to each other. The seg- segment model is used to calculate the underlying forces (Elftman,
ments can rotate in the joints that link the segments. In a two- 1939; Winter, 1990). The calculations are based on the principles
dimensional model, these joints are assumed to be ideal hinge of Newton’s laws of motion. In order to apply these principles the
joints: there is no friction, there is no translation possible and the linked segment model is split into separate segments. The interac-
only movement is a pure rotation around a fixed point, the center tion between the segments is summarized in a net joint force and
of rotation. Since the segment is assumed to be rigid, its length, or a net joint moment. The net joint force is the resultant of all forces
the distance between the joints, is constant. Furthermore the inertial between the two segments. The net joint moment is the sum of the

444
 Mechanical analysis

Fig 19.2  Linked-segment model of a horse (moment

of inertia).
m,l
m,l

m,l

m,l
m,l

m,l m,l
m,l
m,l

m,l m,l
m,l
m,l

m,l m,l m,l

m,l

m,l m,l
m,l m,l
m,l m,l m,l
m,l

Mfet
COR
FXfet

-FYcoffin -Fcoffin
COM
FYfet
Ffet
G

-FXcoffin
COR

-Mcoffin
GRF GRFY
Fig 19.3  Muscle activity causes a bone-to-bone force at the joint. The
muscle action can thus be represented by a moment.

FXcoffin Mcoffin
moments of all muscles and ligaments crossing the joint. All
moments and forces acting on each segment are depicted in a free
body diagram. The moments are the net joint moments at both COM
joints. The forces are the gravitational force, the net joint forces at
both joints and, sometimes, an external force. The most important
external force is the ground reaction force (GRF) acting on the hoof G
segment during stance. Other external forces can be the weight of a
rider or the force needed to pull a load. All those external forces are
measured. Furthermore, the linear and angular accelerations are Fcoffin FYcoffin
calculated from the (measured) movements of the segments. The GRFX POA
inverse dynamic calculations can then start by analyzing either the
most proximal or the most distal segment. Fig 19.4  Free body diagrams of the hoof and pastern segments. POA,
The free body diagram of the most distal segment, the hoof point of application of the COM (segmental center of mass); COR, joint
segment, is drawn in Figure 19.4. The amplitude and direction of center of rotation.

445
 19 Mechanical analysis and scaling

8000 N from the hoof on the pastern is the opposite of the coffin joint
moment. The net joint force at the fetlock joint can now be calcu-
Mcoffin lated in a similar way as was previously done for the coffin joint.
The horizontal acceleration is used to calculate the horizontal com-
COR ponent of the net joint force and the vertical acceleration is used to
FXcoffin calculate the vertical component (see Fig 19.6):
COM 3.0cm
ΣFX = mpast ⋅ aXpast
Y ΣFY = mpast ⋅ aYpast
4.5cm Finally, the angular acceleration is used to calculate the net joint
FYcoffin moment:
1750 N POA ΣM = Ipast ⋅ α past
After the pastern segment, the metacarpal segment can be analyzed
0.2cm in the same manner, again using the principle of action and reac-
0.8cm tion. This can be continued, segment after segment, until the most
proximal segment, most often the head segment. The only forces
on this segment are the gravitational force and the net joint force
X
at the distal side. The only moment is the distal net joint moment.
Fig 19.5  Free body diagram of the hoof segment. In theory these forces and this moment should balance the linear
and angular accelerations of the final segment. However, most often
a residual force and moment are found. This is caused by measure-
ment errors and non-rigidity of the segments. Although errors are
present in all calculated joint forces and moments, they are largest
the coffin joint moment and force are not yet known. They are for the final joint since they accumulate during the calculations. In
therefore drawn in an arbitrary direction. After the calculations the order to prevent this accumulation of errors, and to remove the
real direction and magnitude will be known. According to Newton’s residual moment and force at the final segment, an alternative
second law the sum of all forces on the segment must equal its mass method has been developed (Kuo, 1998). This method solves
times acceleration: the equations for all segments simultaneously. Because there are
ΣFX = mhoof ⋅ aXhoof more equations than there are unknown forces and moments,
where the subscript x denotes the horizontal components of the they cannot be solved exactly. However an approximate solution
force (see Fig. 19.4), mhoof is the mass of the hoof segment and aXhoof can be calculated using a least squares method. In this way the
is the horizontal acceleration of the center of mass of the hoof. errors are distributed evenly over all joints. The errors of the final
Using this equation the horizontal components of the net joint joint are therefore smaller than in the segment-after-segment
force can be calculated from measurements, as is illustrated in approach, however, the errors of the distal joints will be larger.
Figure 19.5. Complete formulas can be found below. A similar Furthermore, the simultaneous least squares method is only useful
equation can be written for the vertical direction to calculate the if the whole horse is analyzed. If only part of the horse is analyzed,
vertical component of the net joint force at the coffin joint: e.g. only one limb, both methods will give the same result.
ΣFY = mhoof ⋅ aYhoof
where the subscript Y denotes the vertical components of the force Measurement of input variables
and acceleration. Of course, the gravitational force (G) should now
In order to perform inverse dynamic calculations several input vari-
be taken into account.
ables are needed. These variables can be sub-divided into three
All forces, except the gravitational force, generate a moment rela-
categories: inertial properties, movement data and external force
tive to the center of mass. The magnitude of the moment is deter-
data. The inertial properties are mass, moment of inertia and loca-
mined by the amplitude of the force and its perpendicular distance
tion of the center of mass for all segments. These properties have
from the center of mass. Moments that tend to rotate the segment
been measured in cadaver segments and are represented as regres-
counterclockwise are defined positive whereas moments that tend
sion equations (van den Bogert et al., 1989; Buchner et al., 1997).
to rotate clockwise are defined negative. For the hoof of Figure 19.4
This allows scaling to horses of different size by using individual
this means that positive moments tend to flex the coffin joint. Note
body mass and/or segment lengths. The position of the center of
that this association between flexion or extension and the sign of
mass is represented in a segment-based coordinate system. The
the moment is dependent on the way the free body diagram is
origin of this coordinate system is located at the proximal joint
drawn since it is reversed for mirrored images. The sum of the net
center of rotation, the x-axis runs through the distal joint center of
coffin joint moment and the moments of all forces must equal the
rotation and the y-axis is perpendicular to the x-axis and points
moment of inertia times the angular acceleration of the hoof
cranially or dorsally (Fig. 19.7).
segment (αhoof):
The movement data consist of the position of the joints and the
ΣM = Ihoof ⋅ α hoof angular and linear acceleration of all segments. The measurement
From this equation the net joint moment can be calculated. Now of position data is discussed in Chapter 2. However, for inverse
all forces and moments acting on the hoof segment are known and dynamic calculations not only the joint angles but also the position
the analysis of this segment is complete. The calculations can be of the joints should be known. The easiest way to measure these
continued by analyzing the next segment, the pastern segment. The positions is by putting the markers on the joint centers of rotation
movement of the pastern joint is often neglected and the remaining (Leach & Dyson, 1988). Joint moments are sensitive to joint center
part of the digit is represented by one segment. location, so it is important to know the inaccuracies introduced by
The free body diagram of the pastern segment (Fig. 19.4, upper this procedure. Alternatively, two markers are applied to each
part) contains the moment and force acting from the hoof on the segment and the position of the joint center of rotation relative to
pastern. According to Newton’s third law (action equals minus reac- these markers is determined in a separate measurement with addi-
tion) the force from the hoof on the pastern is the exact opposite tional markers on the joints. For some joints this procedure can
of the force from the pastern on the hoof. Similarly, the moment limit the skin movement artifacts because the markers can be

446
 Mechanical analysis

Mfet Fig 19.6  Free body diagram of the hoof and pastern
COR segment.

FXfet
5.0cm
7989 N
COM

FYfet
G 6.0cm

1750 N
COR
211 Nm
4.0cm
7.0cm
8000 N

211 Nm

COR
Y 1750 N
COM

POA 7989 N
1750 N

applied on places with minimal skin movement (van Weeren et al., system. This can be performed in a separate session to prevent
1992a). Further discussion of the skin movement problem in interference with the normal data collection.
equine movement analysis may be found in Chapter 2 of this
volume. After the movements are measured they must be differenti-
ated twice to obtain the accelerations. This can be done with a Net joint moment and muscle force
spline function, as described in Chapter 2, or with a finite difference
The net joint moment is generated by the muscles. The polarity of
method (see below). When using the finite difference method exces-
the moment indicates whether flexors or extensors are active,
sive noise must first be removed from the data (see Fig. 2.5). This
whereas the amplitude is a measure of the amount of activity. Based
can be done by filtering or by averaging several trials.
on the net joint moments the activity of muscle groups (flexors and
The external force data consist of the amplitude, direction and
extensors of the different joints) can be analyzed. However, some-
point of application of the external forces, most often the GRF. The
times it is desirable to estimate forces of individual muscles.
GRF can be measured with a force plate, force shoe or instrumented
Depending on the number of muscles crossing the joint, the muscle
treadmill as described in Chapter 2. For inverse dynamic calcula-
forces can either be calculated or estimated from the net joint
tions not only the magnitude and direction of the GRF must be
moment (Fig. 19.8).
measured but also the point of application. Force plates, although
In some joints there is only one muscle that can flex the joint.
very accurate with respect to the force amplitude, often have system-
The deep digital flexor muscle (DDF), for instance, is the only
atic errors in the point of application (Bobbert & Schamhardt,
muscle that can flex the coffin joint. (The navicular ligaments can
1990). These errors can be corrected with a calibration procedure
generate a flexor moment but only if the joint is (hyper) extended.)
in which static loads are applied at known positions. Instrumented
In a flexed joint the moment generated by DDF will equal the net
treadmills or force shoes give the opportunity to measure consecu-
coffin joint moment:
tive strides. However, the accuracy of these systems should be evalu-
ated carefully. Mcoffin = FDDF ⋅ dDDF
The movement and GRF data must be aligned and synchronized. where dDDF is the moment arm of the deep digital flexor – the per-
Some motion analysis systems can capture the force data directly pendicular distance between the joint center of rotation and the line
and do not need additional synchronization. For other systems of action of the muscle (Fig. 19.8, Fig. 19.9A). This moment arm
synchronization pulses from the video camera can be sampled can be measured from radiographs (Jansen et al., 1993) or from
simultaneously with the force data to synchronize the force data. longitudinal sections of an in vitro limb. For some muscles the
Alternatively, a counter operated by the force sampling equipment moment arm depends on the joint angle. A model containing the
and visible in the camera image can be used to synchronize the origin, insertion and possible curvatures of the muscle can then be
motion data. The coordinate systems of the force plate and motion used to calculate the moment arm from the joint angles (van den
analysis systems can be aligned by putting markers on the edges of Bogert & Sauren, 1989). When calculating FDDF from Mcoffin using
the force plate and measuring them with the motion analysis the equation above, it is assumed that there is no other active

447
 19 Mechanical analysis and scaling

1
4
3 6

13
7

5
8
14

z
C
y

A
x 15
9

10
B 16
11 17
12 18

Fig 19.7  Definition of the segments and their coordinate systems. Red lines represent the segment boundaries. The schematic enlargement shows the
coordinate system. The z axis is used only in three-dimensional analyses.
Reprinted from Buchner, H.H.F., Savelberg, H.H.C.M., Schamhardt, H.C., Barneveld, A., 1997. Inertial properties of Dutch Warmblood horses. J. Biomech. 653–658, with permission from
Elsevier.

FDDF force. This type of analysis has shown that desmotomy of the distal
accessory ligament causes a decrease in the DDF force that does not
211 Nm disappear within 6 months (Buchner et al., 1996; Becker et al.,
1998).
COR
For joints with only one muscle at each side, the calculation of
3.3 cm muscle force is simple. However, most joints, e.g. the hip joint of
Figure 19.9B, are crossed by more than one muscle. In order to
calculate muscle forces from the net moments at these joints,
assumptions must be made about the distribution of the moment
among the available muscles. To estimate muscle forces the
moments are most often distributed in such a way that a certain
criterion or cost function is minimized. Computer algorithms are
available to solve the distribution problem by minimizing a given
cost function. This procedure is called optimization. Some com-
Fig 19.8  Deep digital flexor force. monly used cost functions are the sum of all muscle forces, the sum
of the squared or cubed muscle forces and the highest muscle force.
In order to correct for size differences between the muscles the force
is sometimes replaced by stress: muscle force divided by physiologi-
muscle or force-generating ligament at the coffin joint. The antago- cal cross-sectional area. Other cost functions might include mea-
nistic muscles (the digital extensors) are assumed to be inactive. If sures for muscle fatigue, energy costs or joint contact forces
these muscles, contrary to the assumption, are active they will gen- (Crowninshield & Brand, 1981; Dul et al., 1984; Glitsch & Baumann,
erate an extensor moment, which should be compensated by the 1997).
DDF. The actual FDDF will then be higher than the calculated force. A major problem in optimization is that it is difficult to validate
The calculated force is therefore the lower limit of the actual muscle the cost function. The real muscle forces can only be found when

448
 Mechanical analysis

FDDF

FTR1
FRO1 FRO2
Mcoffin
FOT
COR FSV3
dDDF FTR2

FSS

FSV1
FSV2
Mcoffin = FDDF • dDDF

∑FX,TRi + ∑FX,ROi + ∑FX,SVi + FX,OT + FX,SS = FXscapula

∑FY,TRi + ∑FY,ROi + ∑FY,SVi + FY,OT + FY,SS = FYscapula
and
FGS
FGM ∑(FTRi * dTRi) + ∑(FROi * dROi) + ∑(FSVi * dSVi) + FOT * dOT + FSS * dSS = Mscapula
while e.g.
dGM
∑(FTRi2) + ∑(FROi2) + ∑(FSVi2) + FOT2 + FSS2 is minimized
dGS
C

dST
dSM FBI
dBI

FSM

FST

Fig 19.9  Different ways to estimate muscle forces (F) from net joint moments
(M) and muscular moment arms (d). (A) With only one muscle the force can be
calculated from the moment. (B) With several muscles, optimizations are needed
to distribute the moment. (C) The muscles of the shoulder synsarcosis generate
FGM * dGM + FGS * dGS + FBI * dBI + FST * dST + FSM * dSM = Mhip both the net moment and the net forces. Muscles with large origins are
represented by several (numbered) force vectors; in the formulas the influence of
while e.g. these vectors are summed. BI, biceps femoris; COR, coffin joint center of rotation;
FGM2 + FGS2 + FBI2 + FST2 + FSM2 is minimized DDF, deep digital flexor; GM, gluteus medius; GS, gluteus superficialis; OT,
omotransversarius; RO, rhomboideus; SM, semimembranosis; SS, subscapularis;
B ST, semitendinosus; SV, serratus ventralis; TR, trapezius.

the cost function represents the principles by which the nervous These more sophisticated dynamic optimization models, as opposed
system solves the distribution problem. These principles, however, to static optimizations that do not consider muscle properties (both
are not known. Another problem is that realistic boundary condi- based on inverse dynamic analysis of movement!), have not been
tions must be included in the optimizations. For each muscle, the used other than in research, presumably because more complicated
force it can generate is limited to a certain maximum. This maximum calculations are required. However, they will result in more realistic
is influenced by the momentary length and the contraction velocity estimates of muscle force. In rare cases, results from optimizations
of the muscle fibers. Predicted forces should of course not exceed have been compared to direct measurements using invasive
this maximum force. Furthermore, muscle force can only increase methods. In cats, the load sharing among synergistic muscle was
and decrease gradually, muscles cannot be either ‘on’ or ‘off’ not correctly predicted by static optimization (Herzog & Leonard,
momentarily. Dynamic models of muscle in which these properties 1991) when compared to tendon force transducer data.
are incorporated can be used in optimization based on inverse Most muscles can be represented by moment generators. This is
dynamics (Thunnissen, 1993). possible because the bones are connected by joints that can be

449
 19 Mechanical analysis and scaling

modeled as hinge or ball joints and can generate the bone-to-bone estimated by static optimization of muscle forces and compared to
force at a known joint center. However in the shoulder synsarcosis measurements from an instrumented prosthesis (Brand et al.,
there is no joint (Dyce et al., 1987). Therefore the muscles of the 1994). It was found that the model calculations overestimated
shoulder synsarcosis are not moment generators. Furthermore, the the joint contact force, mainly due to incorrect moment arms in
movement of the scapula and relative to the trunk is not limited to the model.
rotation around a fixed point. Due to the absence of a joint, sub-
stantial translations are also possible. The linked-segment model
will become more realistic if the hinge joint between scapula and Continued calculations: power
trunk is replaced by a free movement and the muscles are repre- The muscles are the motors of the body. The moments and forces
sented by force generators. The net moment and net force can still generated by the muscles have been analyzed in the previous para-
be calculated relative to a fixed point on the scapula – this can be graphs. However, a motor does not only generate moments or
any point, for instance the center of mass. However, the point of forces; it also produces or converts mechanical energy. The muscles
application of the net force on the trunk is not fixed any more. The obtain their energy from the oxidation of foods and convert it to
forces of the muscles can still be calculated exactly if there are three mechanical energy. The energy is distributed over the segments and
active muscles in the model (since there are three equations of is used internally, to accelerate or elevate the segments, or externally,
motion for a body segment), or using optimization methods if e.g. to overcome air resistance or to pull a load. Energy production,
there are more than three. The only difference in the equations is distribution and use can be analyzed with calculations based on
that the muscles should not only generate the net moment between inverse dynamic results. If a force acts on a moving object, energy
scapula and trunk but also the net force. Therefore the position of is produced or absorbed. The amount of energy equals the force
the muscle should not be represented by the moment arm but by times displacement of the point of application in the direction of
its actual line of action between trunk and scapula (Fig. 19.9C). the force. Power is the time derivative; it equals energy production
divided by time, which is equal to force times velocity of the point
of application:
Net joint force and joint contact force
P = FX ⋅ v X + FY ⋅ v Y
The net joint force, calculated from the equations of motion, rep-
For rotational movements the power equals moment times angular
resents the sum of all forces between the two segments. It should
velocity (P = M · ω). Note that power will be calculated in Watts if
not be confused with the joint contact force, which can be measured
the moment is expressed in Newton-meters and the angular velocity
between the bones in the joint. The joint contact force is only one
in radians per second. In power analyses of linked-segment models,
of the forces between the segments; other forces are the muscle and
the powers associated with all forces and moments on the segments
ligament forces. The real contact force therefore equals the differ-
are calculated. The sum of these powers equals the rate of change
ence between net joint force and the summed muscle (and liga-
of the kinetic and potential energy of the entire system. This power
ment) forces (see Fig. 19.10):
balance follows from Newton’s laws of motion and may be used to
Fcontact = Fnet − ΣFmuscle determine the contributions of each joint to the movement of the
When calculating the joint contact force, the forces of all active entire system (van Ingen Schenau & Cavanagh, 1990).
muscles should be taken into account. In calculating muscle forces For most forces and moments the power calculation is straight-
from the net joint moment it is often assumed that there are no forward. However, the GRF needs special attention. The point of
antagonistic co-contractions. This is implied by most of the static application of the GRF shifts forward during stance. It is tempting
optimization methods. These co-contractions will cancel each other to multiply the velocity of this shift with the GRF to obtain the
out with respect to the joint moment and are therefore ‘invisible’ power. However this is not the powerflow associated with the GRF
to a mechanical analysis. However both muscles add a bone-to- (van Ingen Schenau & Cavanagh, 1990). On a hard, non-deformable
bone force and increase the joint contact force. Ignoring antagonis- surface the GRF does not perform work. The GRF is a distributed
tic co-contractions therefore results in an underestimation of the force, which is summarized to one resultant force with a certain
joint contact force. Similarly, the joint contact force will be influ- point of application. In Figure 19.11 the GRF is represented as five
enced by the way the net joint moment is distributed over muscles smaller forces acting on different positions of the hoof surface.
with different moment arms. Muscles with a larger moment arm During stance the amplitude of these five partial forces changes,
need a smaller force to generate a similar moment than muscles
with a smaller moment arm. The joint contact force will therefore
be smaller if the contribution of muscles with a larger moment arm
is higher. In humans, the contact force in the hip joint was

6394 N
FXcontact
45°

FYcontact

Fig 19.11  Shifts and movements of the point of application of the GRF on
a hard surface (above) and on a deformable surface (below). (See text for
Fig 19.10  Joint contact force. details.)

450
 Mechanical analysis

whereas their position stays constant. This causes the point of appli- two adjacent segments. When adding the powers of the different
cation of the resultant GRF to shift forward. However, none of the segments the powers associated with the net joint forces cancel each
partial forces generates power since their velocity is zero. The resul- other out. The only power, therefore, is the power associated with
tant force, which is simply the summation of all partial forces, does the net joint moment. At each joint the powers for the two segments
not generate power. In reality there are, of course, more than five are summed to one joint power. This power equals the net joint
partial forces, but the same reasoning applies. On a deformable moment multiplied by the difference in angular velocity between
surface, like sand, the situation is different. When the hoof pene- the segments:
trates the surface the point of application not only shifts due to
Pjoint = Mjoint ⋅ (ω prox − ω dist )
amplitude changes but also moves due to hoof movements. The
position of the five partial forces changes and they generate power This power is generated by the muscles crossing the joint and is
(see Fig. 19.11). Therefore, on deformable surfaces, power can be therefore a measure of the activity of the muscles. The joint powers
associated with the GRF. This power is related to the movement of of all joints can be summed to obtain the total power production.
the surface; it equals the velocity of the hoof surface times the This power is used for external power production (e.g. pulling a
amplitude of the GRF. This power is used to deform the surface. load, air resistance) and kinetic and potential energy changes of the
This example illustrates that valid power results are only obtained segments. Power production at the joints can therefore be compared
if a force is multiplied by the velocity of a point on a physical body with external power losses due to air resistance, friction, etc. In
to which the force is applied. The velocity of a point that is not human research this approach has been applied successfully in
rigidly attached to a mass is not suitable. movements where most of the power is used externally and where
When analyzing the powers of the linked-segment model, two internal loss of power is minimal. Examples can be found for bicy-
different approaches are used: a segment oriented approach cling, swimming and speed skating (van Ingen Schenau & Cava-
(Elftman, 1939; Winter & Robertson, 1978; Robertson & Winter, nagh, 1990). In walking and running the external power production
1980) and a joint-oriented approach (van Ingen Schenau & Cava- is very small; most power is used to accelerate and decelerate the
nagh, 1990). The segment-oriented approach is used to analyze limbs and to provide a shock-absorbing support against gravity.
internal use and the distribution of energy, or power-flow, whereas Apart from the high-speed gallop, where air resistance might
the joint-oriented approach is used to analyze power production and become substantial, the same holds for the gaits of a horse. Although
external use. In the segment-oriented approach two power-flows are it is not useful to apply this approach to the equine walk and trot,
distinguished at both the proximal and distal end of the segment: it can be used when the external power output is substantial, e.g.
the power-flow associated with the net joint force and the power- in pulling a load, during accelerating and in jumping.
flow associated with the net joint moment. The power-flow associ- Apart from the external power production the joint-oriented
ated with the net joint force is a passive power-flow from the analyses might give insight into muscle function during movement
adjacent segment, or, if negative, a passive power-flow into the (de Koning et al., 1991; Devita & Skelly, 1992). Positive joint power
adjacent segment. At the joint, the passive power-flow into indicates power production that might originate from concentric
the proximal segment equals the passive power-flow out of the activity of the muscles crossing the joint. Large positive power pro-
distal segment (or reversed). This is caused by the fact that the net duction peaks indicate important propulsive muscles whereas
joint forces on the distal and proximal segments are each other’s timing of the different peaks indicates the coordination of the
opposites while the velocities at the point of application, the center movement. Furthermore, negative joint power indicates power
of rotation, are equal. The power-flow associated with the net joint absorption, which might be related to eccentric muscle contraction.
moment is an active power-flow generated by the muscles. At the These contractions are assumed to be a major cause of muscle sore-
joint, the angular velocities of the two segments are normally dif- ness and a risk for muscle injuries (Jones & Round, 1990). Power
ferent and therefore the power-flow associated with the net joint analyses can therefore be used to find muscles that are at high risk
moment is different for the two segments. This difference is gener- for injuries. This type of analysis has been used for the equine
ated (or absorbed) by the muscles. forelimb during walk (Colborne et al., 1998). Moment peaks of the
All four power-flows, passive and active proximal and distal flow, carpal and fetlock joints were of similar magnitude whereas the
can either be positive or negative. Most often some of them are coffin joint moment was much smaller. The carpal joint power was
positive whereas others are negative and power-flows from one almost zero due to the small movements of this joint. The fetlock
segment to the other. However the sum of the four flows is not joint had alternating periods of power absorption and production
always zero. If the sum is positive, the energy of the segment whereas the coffin joint predominantly absorbed power.
increases. This energy is converted into potential or kinetic energy The two approaches, segment-oriented and joint-oriented power
as the height or the velocity of the segment increases. Similarly, a analyses, are closely related. They emphasize different aspects of the
negative power-flow is converted from potential and/or kinetic power balance. The joint-oriented approach emphasizes the exter-
energy when the height and/or velocity decreases. nal power production and the places where this power is produced
Using this kind of analysis the power-flow within the limb can whereas the segment-oriented approach emphasizes the power-flow
be analyzed. This has been done for the distal front limb during within the body and the internal use (see Fig. 19.12). In both
walk (Colborne et al., 1997). During the major part of the stance approaches power production and absorption have simply been
phase the power-flow of the metacarpal and pastern segments was related to concentric and eccentric contraction of muscles at the
distal to proximal. This probably reflects the propulsion of the joint. However the biological reality is a bit more complex for two
proximal part of the body. During final stance the power-flowed in reasons: elastic structures can temporarily store energy and most
at both ends of the segments and was probably used to accelerate muscles cross more than one joint (van Ingen Schenau & Cavanagh,
and elevate the segments for the swing phase. Care must be taken 1990).
with this kind of analysis because results are dependent on the Elastic structures, like tendons and ligaments, store energy when
velocity of the coordinate system in which the movements are elongated. This energy is released when they return to their normal
represented. This will influence the results when measuring on length. The long and stiff tendons in the distal limbs of horses are
a treadmill. We recommend that the analysis be performed only able to store substantial amounts of elastic energy (Alexander &
in a coordinate system attached to the ground, which implies a Bennet-Clark, 1977). Furthermore, the deep palmar ligament of the
moving coordinate system for treadmill analysis. carpus might also absorb a substantial amount of energy during
In the joint-oriented approach of power analyses the power-flows hyperextension of the carpus. Power absorption is therefore not
of all segments are summed (van Ingen Schenau & Cavanagh, always associated with eccentric muscle contraction but can also be
1990). The power associated with the net joint force is opposite for caused by elastic energy storage in tendons and ligaments. The

451
 19 Mechanical analysis and scaling

0% – 20% 20% – 45% 45% – 60% 60% – 90% 90% – 100%

Fig 19.12  Five parts of stance phase.

subsequent power production can originate from the release of as rigid struts over which the body vaults – the kinetic energy and
elastic energy instead of concentric muscle contraction. If elastic potential energy changes are each other’s opposites. In the bounc-
properties of tendons are known, and tendon force is estimated ing gaits like trot, the limbs act as springs and the body center of
from inverse dynamic analysis, the amount of energy storage can mass moves like a bouncing ball – the kinetic and potential energy
be estimated separately. increase and decrease simultaneously. Equine gallop might have
Power production and absorption are usually analyzed at a joint both bouncing and pendulum aspects (Minetti et al., 1999). Apart
level but many muscles span more than one joint. These muscles from this classification of gaits, attempts have been made to relate
can therefore absorb power at one joint and simultaneously produce the changes in mechanical energy to metabolic energy consump-
power at another joint. This has, for instance, been described for tion (Cavagna & Kaneko, 1977; Minetti et al., 1999). The major
the gastrocnemius muscle in human vertical jumping: at the end of problem is that when moving at a constant speed the mechanical
push-off, the muscle generates positive power at the ankle and nega- energy at the end of the stride equals the mechanical energy at the
tive power at the knee. This function of biarticular muscles is beginning of the stride and no net work is performed. Neverthe-
referred to as power transport between joints (Gregoire et al., 1984). less, to enable analysis, negative and positive changes of energy of
It is assumed to be a major function of these muscles. Power trans- the segments are treated differently. Either all positive changes are
port can take place during isometric contraction without power added, ignoring negative powers, or different efficiencies are used
production by the transporting muscle. Therefore, no muscle fibers to add positive and negative power production. However the effi-
are needed to transport power and tendons can also transport ciencies of positive and negative muscular work are speculative.
power. In the lower limbs of horses there are several more or less Furthermore, energy can be stored temporarily in elastic structures
tendinous bi- or polyarticular muscles. It has been hypothesized or it can be transferred between segments and probably also
that the proximal muscles generate the power and that this power between limbs. The relative importance of these processes is not
is transported to the distal joints by the tendinous biarticular known. Results of internal power analyses, therefore, are largely
muscles (van Ingen Schenau, 1994). The major advantage of this dependent on the assumptions of internal power transport, energy
system is that the lower limbs are very lightweight while the joints storage and efficiency. This necessitates careful validation of the
can still be actively moved and can contribute to the propulsion. mechanical energy analysis, and estimation of error bounds due to
The lower limbs undergo large accelerations during gait; if assumptions regarding power transfer, efficiency and elastic storage
they are lightweight the accelerations will cost less energy and (Thys et al., 1996).
can be performed more rapidly, enabling a faster and more eco-
nomical gait.
To prevent the problems with biarticular power transport, power
Accuracy of inverse dynamic calculations
production can also be analyzed at the level of the muscles, using All direct and indirect measurements of biological variables are
the muscle force estimation methods described previously. The esti- influenced by certain sources of errors and, therefore, have a
mated muscle force is multiplied by the contraction velocity of the limited accuracy. Inverse dynamic calculations are no exception to
muscle. The contraction velocity can be obtained from kinematic this rule. The errors influencing the accuracy of inverse dynamic
data (Riemersma et al., 1988; van den Bogert et al., 1988) or by calculations can be subdivided in two categories: errors in the mea-
multiplying the joint angular velocity by the moment arm of the surement of input variables and errors originating from model
muscle. For human cycling, energy expenditure predictions obtained assumptions. The most important measurement errors are caused
from power analysis at the muscle level differed greatly from predic- by estimation of the inertial properties from in vitro data, misalign-
tions using power analysis at the joint level (Neptune & van den ment of markers with the joint rotation center, skin displacement
Bogert, 1998). This is due to the fact that joint power does not relative to the underlying bones, the use of noisy data to calculate
consider power transport by biarticular muscles or antagonistic accelerations, inaccuracies in the measurement of the point of
co-contraction. Estimates of energy expenditure should therefore application of the GRF and misalignment of force and movement
not be obtained from analyses at the joint level but from analyses data. Further details on some of these inaccuracies may be found
at the muscle level. Power analysis at the joint level can still be used in Chapter 2. The major assumptions influencing the accuracy of
to describe and analyze the control of movement. the calculations are the rigidity of the segments and the fixed rota-
Apart from the segment-oriented and joint-oriented approaches tion point. If muscle forces are calculated, two major assumptions
there is a third power analysis approach, which is used in gaits are often added: the absence of antagonistic co-contractions and
without substantial external power production. This approach does the cost function used to distribute the moment among the avail-
not use the joint moments and forces but focuses on mechanical able muscles. The influence of all these errors and assumptions on
energy changes. During a stride cycle both the potential and the the final accuracy depends, of course, on the magnitude of the
kinetic energy of the segments change constantly. Based on com- errors. This magnitude is determined by the equipment used and
parison of patterns of these energy changes gaits have been classi- the conditions of the actual measurements. However, some general
fied as either inverted pendulum or bouncing gaits (Cavagna & considerations can be made on the order of the errors and their
Kaneko, 1977). In inverted pendulum gaits like walk, the limbs act influence on the final results.

452
 Mechanical analysis

The assumptions of a fixed rotation center can be questioned for distribution between the medial and lateral compartments of the
the stifle and carpal joints, whereas the alignment of the markers knee (Lindenfeld et al., 1997).
with the rotation center is difficult in the coffin joint and the proxi- Extension of inverse dynamic analysis to three dimensions is
mal joints. Even relatively small errors in the coffin joint marker mathematically straightforward (Vaughan et al., 1992; van den
placement can seriously influence the results for this joint because Bogert, 1994). The equation of motion for force is the same, with
of the small moment arms in the hoof segment. Radiographs the only difference that force is now a vector with three, rather than
should therefore be used to validate and, if necessary, correct this two components. The equation of motion for moments has an extra
marker position. Due to similar small moment arms, measurement term in three dimensions, and uses a matrix I for the moment of
errors in the point of application of the GRF relative to the hoof inertia rather than a single scalar value:
will also seriously influence the coffin joint results (Bartel et al.,
∑ M = I ⋅ α + ω × (I ⋅ ω)
1978). These errors can arise from systematic errors in the point of
application or from misalignment of the coordinate systems of the Joint moments M are vectors in three dimensions, as are angular
force and movement data. The systematic errors in the point of velocities w and angular accelerations α. In the force equation,
application should be corrected (Bobbert & Schamhardt, 1990) vectors are usually expressed in a global, ground-based coordinate
and the two coordinate systems should be aligned carefully. system. In the moment equation, however, all vectors and matrices
Although the errors in the point of application, in the alignment must be expressed relative to a segment-fixed coordinate system.
and in the marker placement will influence the accuracy for other The final term is proportional to the square of the magnitude of the
joints, the relative influence is much smaller due to the larger angular velocity and accounts for centrifugal effects when the axis
moment arms. of rotation is not aligned with an axis of symmetry (principal axis)
The accuracy of the inertial properties differs for the proximal of the body segment. Note the cross product symbol ‘×’, which
and distal segments (Buchner et al., 1997). In general the proxi- makes this moment perpendicular to the axis of rotation. The
mal segments have a large mass, are not quite rigid and it is diffi- inertia matrix I is a 3 × 3 matrix which is diagonal (only three ele-
cult to separate them from each other. The distal limb segments, ments are non-zero) when the axes of the segment coordinate
in contrast, have a small mass, are quite rigid and the boundary system are aligned with the principal axes; otherwise I has 9 non-
between these segments is clear. Therefore, the inertial data for zero elements that reflect the non-symmetrical distribution of the
the distal segments is quite accurate and does not seriously influ- mass. Three-dimensional inertial properties for equine segments
ence the accuracy of the results. Furthermore, the accelerations of have been measured with respect to anatomically defined coordi-
the distal segments are very small, at least during stance, and nate systems (Buchner et al. 1997). When expressed in a suitable
errors in the inertial data or the accelerations hardly influence the segment-fixed coordinate system, the three components of the
results for these segments. Similarly, the total ignorance of the moment vector M can be interpreted as flexion–extension,
accelerations, as is done in a quasi-static analysis, seems to be abduction–adduction and internal–external rotation moment,
acceptable for these segments. During the swing phase, however, respectively. The angular velocity vector w must be decomposed
the results are determined completely by the accelerations. Errors along the same three axes in order to obtain the joint power by
in these accelerations or the inertial data seriously influence the taking the dot product of moment and angular velocity.
accuracy. Skin displacements for the distal joints are small and Instrumentation for collecting three-dimensional movement and
will not influence the accuracy. However, for the proximal joints GRF data is now standard equipment in biomechanical laborato-
they are very large, influence the accuracy and should be cor- ries. However, problems arise from modeling assumptions and the
rected, using the available correction algorithms (van Weeren standardization of the experimental protocol. These problems may
et al., 1992b). lead to inaccuracies that are, in some cases, larger than the magni-
The effect of random errors on the final results can be determined tude of the results of the calculations. In those cases, a three-
with Monte Carlo simulation (van den Bogert et al., 1994). In this dimensional analysis does not provide useful additional information
simulation the analysis is performed multiple times, with multiple compared to a sagittal plane analysis.
copies of the input data that are modified by adding random errors In the experimental protocol, three markers must be attached to
according to known probability distributions. The standard devia- each segment in order to obtain all six degrees of freedom (three
tion of these multiple results provides an estimate of the error in translations and three rotations) for each segment. It is no longer
the final result. Systematic errors, such as skin movement, are not possible to have markers coincide with the joint center in three
random and should be modeled as a hypothetical function of time dimensions, since the joint center is inside the body. Markers must
or joint angle in such an error analysis. therefore be attached to standardized anatomical landmarks, rela-
tive to which the location of the joint center is known. Suitable
protocols have been developed for human analysis (Bell et al.,
Three-dimensional analysis 1990; Vaughan et al., 1992), and first attempts have been made for
The methodology presented above applies to a two-dimensional equine analysis (Nicodemus et al., 1999).
analysis, usually in the sagittal plane. Movement of horses seems Soft tissue movement may be more of a problem in three dimen-
remarkably planar, but for some applications the small medio­ sions because of the indirect method for locating joint centers. Joint
lateral movements may also be important. In trotters, these out of moments are known to be sensitive to errors in the joint center loca-
plane movements were reproducible, highly individual, and pre- tion (Burdett, 1982). Also, three markers may not be representative
sumably related to performance (van Weeren et al., 1993). It is of the motion of the entire body segment when the segment under-
therefore conceivable that analysis of the sources of those move- goes complex soft tissue movements during impacts. This may lead
ments is useful. Furthermore, the magnitude of out-of-plane to errors in the inertial terms of the equations of motion, the terms,
moments is presently unknown. In human walking, it has been which are proportional to masses and moments of inertia. On seg-
shown that a two-dimensional analysis underestimates the average ments such as the thigh, using more than three markers should be
joint power at the hip by 23% because the hip abductor moment considered. Least-squares techniques can then provide the best esti-
is ignored (Eng & Winter, 1995). Similar underestimations might mate of three-dimensional rigid segment movement that is consis-
be obtained in horses and a full three-dimensional analysis should tent with data from all markers (Soderkvist & Wedin, 1993). Note
be undertaken to determine for which movements and which appli- that there are two conflicting requirements for a marker placement
cations the out-of-plane moments can be neglected. Finally, out-of- protocol: for location of joint centers, markers should closely follow
plane moments might provide useful information about joint the bone movement; for accurate estimation of inertial forces,
loading. In humans, adductor–abductor moments reflect load markers should follow the ‘average’ of the mass distribution, of

453
 19 Mechanical analysis and scaling

Mfet
FXfet COR

ypast -yfet
-FYcoffin -Fcoffin
COM
FYfet
Ffet
G ypast -ycoff

COR -FXcoffin

-Mcoffin
GRF GRFY Xfet -Xpast
Xcoff -Xpast

Mcoffin
FXcoffin COR COR
yhoof -ycoff
COM

yhoof -yGRF

Fcoffin FYcoffin
GRFx POA
Xcoff -yhoof
A B XGRF -Xhoof

Fig 19.13  Free body diagrams of the hoof and pastern segments. (A) Free body diagrams; (B) x and y coordinates of the points of application of the forces.

which the muscles can be an important component. Different Formulas for inverse dynamic calculations
marker sets for these two calculations should be considered.
Accelerations for certain out-of-plane movements in the horse Hoof segment
may be small relative to the errors associated with differentiation
The sum of all forces on the segment must equal its mass times
of noisy movement data (see Chapter 2). It is therefore important
acceleration. In the horizontal direction this gives the following
to have good estimates of these errors, so that the accelerations that
equation:
are smaller than the noise can simply be neglected, rather than
introduce extra noise into the results. ∑ FX = mhoof ⋅ aXhoof ⇔ GRFx + FXcoffin = mhoof ⋅ aXhoof
In modeling, the body is typically represented as rigid segments
or:
linked by ball joints. For estimation of muscle forces, three
moment arms must be known for each muscle, one for rotation FXcoffin = mhoof ⋅ aXhoof − GRFX
about each axis of the segment coordinate system. Three- For symbols see Figure 19.13 and the symbol list below.
dimensional models of the muscle path from origin to insertion For the vertical direction the gravitational force (G) should be
(e.g. Brand et al., 1982) are typically used to determine the instan- taken into account:
taneous moment arms, as well as the orientation of muscle force
vectors during movement. Interactive software tools are now avail- ∑ FY = mhoof ⋅ aYhoof ⇔ GRFY + FYcoff = Ghoof = mhoof ⋅ aYhoof
able to develop such models of the three-dimensional musculosk- or:
eletal anatomy (Delp & Loan, 1995).
FYcoffin = mhoof ⋅ aYhoof − GRFY − mhoof ⋅ g
Many joints in horses and humans are hinge joints where rota-
tions other than flexion and extension are prevented by the articu- The sum of the net coffin joint moment and the moments of all
lar surfaces and ligaments. Such joints have an axis of rotation forces must equal the moment of inertia times the angular accelera-
rather than a center of rotation. The ‘joint center’ used for the tion of the hoof segment:
inverse dynamic analysis lies on this axis; typically, the midpoint of ∑ M = Ihoof ⋅ ahoof
the part of the axis within the bone is used. The moment compo-
nents associated with internal–external rotation and abduction– or:
adduction components do not represent muscle function. Mcoffin + GRFX ⋅ (yhoof − yGRF ) + GRFY ⋅ ( xGRF − xhoof )
Interpreting these moments as muscle moments may lead to severe + FXcoffin ⋅ (yhoof − ycoff ) + FYcoffin ⋅ ( xcoff − xhoof ) = Ihoof ⋅ α hoof
overestimation of muscle forces and joint loading (Burdett, 1982;
or:
Glitsch & Baumann, 1997). Instead, these moments can indicate
load distributions in articular contact and ligaments (Lindenfeld Mcoffin = Ihoof ⋅ α hoof − GRFX ⋅ (yhoof − yGRF ) − GRFY ⋅ ( xGRF − xhoof )
et al., 1997). − FXcoffin ⋅ (yhoof − ycoff ) − FYcoffin ⋅ ( xcoff − xhoof )

454
 Scaled energetics of locomotion

Note the different subtraction of the x and y coordinates: the x Symbols

coordinate of the joint center of rotation is subtracted from the
x coordinate of the point of application of the force whereas α angular acceleration of the segment
the y coordinate of the point of application is subtracted from the a linear acceleration of the segment
y coordinate of the center of rotation. In this way moments F force
produced by the forces will automatically get a correct sign: positive Fcoffin net coffin (coff) joint force
for counterclockwise and negative for clockwise moments. Ffetlock net fetlock (fet) joint force
GRF ground reaction force
g gravitational acceleration: 9.81 m/s2
Pastern segment I moment of inertia
Horizontal and vertical forces: m segmental mass
∑ FX = mpast ⋅ α Xpast ⇔ FXfet − FXcoffin = mpast ⋅ α Xpast M moment
Mcoffin net coffin (coff) joint moment
∑ FY = mpast ⋅ α Ypast ⇔ FYfet − FYcoffin + Gpast = mpast ⋅ α Ypast Mfetlock net fetlock (fet) joint moment
Note that the net coffin joint force (the force from the pastern of t time
the hoof) is inverted to obtain the force from the hoof on the v velocity
pastern. x horizontal coordinate of point of application or center of
Moments: mass
∑ M = Ipast ⋅ α past X horizontal component of force or acceleration (subscript)
y vertical coordinate of point of application or center of
or: mass
Mfet − Mcoffin + FXfet ⋅ (ypast − yfet ) + FYfet ⋅ ( xfet − xpast ) Y vertical component of force or acceleration (subscript).
− FXcoffin ⋅ (ypast − ycoff ) − FYcoffin ⋅ ( xcoff − xpast ) = Ipast ⋅ α past

Velocity and accelerations Scaled energetics of locomotion

Velocity and acceleration can be calculated using a finite difference
method. The difference in position between frame number i and Introduction
i + 1 is divided by the time difference to obtain an approximation
Horses come in quite different sizes (Fig. 19.15), raising the ques-
of the velocity at frame i + 1/2 (see Fig. 19.14):
tion as to how the performance of a small and large horse should
∆x xi + l − xi be compared. In Figure 19.15, the two breeds of horses shown are
Vi +1/ 2 = = about the biggest (British Shire) and smallest (Argentinean Falla-
∆t t i + l − t i
bella) that exist. Imagine a situation in which the Shire (height
Subsequently the difference in velocity between i + 1/2 and i − 1/2 1.91 m, weight 1000 kg) moves alongside the Fallabella (height
is used to calculate the acceleration: 0.80 m, weight 60 kg). The Fallabella will be trotting, maybe even
galloping, while the Shire is still walking. Obviously, the speeds of
∆v v i + l / 2 − v i −1 / 2 x i + l − x i − ( x i − x i −1 ) x i + l − 2 x i + x i −1
ai = = = both horses should be scaled to their size, but how should this be
∆t ∆t (∆t )2 achieved? Is it proportional to weight? This is not likely, since in
For power calculations the velocity at i instead of i + 1/2 must be that case the Shire would move 1000/60 = 17 times as fast as the
known; this velocity can be calculated from the difference in posi- Fallabella. Is it, then, proportional to height? This seems more
tion at frame i − 1 and i + 1: plausible, but it will be shown that it is not correct either. Questions
of this type will be examined in this chapter. It will be seen that the
∆x x i + l − x i −1 x i + l − x i −1 outcome depends not only on the question asked – a common situ-
vi = = =
∆t t i + l − t i −1 2∆t ation – but also on the assumptions made. Recognizing assump-
tions is a trickier problem.
There is a vast amount of literature in the field of scaling. It has
been produced by generations of biologists and a few engineers,
who have always been fascinated by questions like why an ant can
carry 20 times its own weight, and why a horse evidently cannot.
There are several books on this subject (Alexander & Goldspink,
1977; McMahon & Bonner, 1983; McMahon, 1984; Schmidt-
Nielsen, 1984; Pennycuick, 1992), all of which make fascinating
Xi+1 reading, as their authors are talented writers.

Geometric scaling

PROBLEM
Xi
The first problem to be handled is the most elementary. Assume that
animal B is in all aspects twice as large as animal A. A consequence is
that B can take steps that are twice as a long as A. Can it also run
Xi-1 ∆t ∆t twice as fast? And what stride frequency will it use?

ti-1 ti ti+1
In a first approximation it will be assumed that the bigger animal
Fig 19.14  Finite differences can be used to calculate velocity and is just an enlarged version of the smaller one, that all linear mea-
acceleration. sures (leg length, leg diameter, trunk circumference, etc.) are

455
 19 Mechanical analysis and scaling

Fig 19.15  The smallest and biggest breeds of

horse: the Fallabella, height 0.80 m, mass 60 kg,
with foal (weight at birth 8 kg); and the Shire,
height 1.91 m, mass 1000 kg.
Courtesy of Dr. Agaath Kooi, Noorder Zoo at Emmen, The
Netherlands.

Table 19.1  Dimensionless numbers for length, time, speed/

velocity and acceleration

Length Time Velocity Acceleration

l t v a
lˆ = t̂ = v̂ = â =
l0 l0 /g l0 g g

as our horses don’t walk on the moon, where g = 2 m/s2, it may be

assumed to be constant. Acceleration is calculated as distance
divided by time squared (unit: m/s2). When distance increases, the
time scale should thus be made longer as well. When one animal
has height l1 and the other height l2, all temporal variables, stance
time, stride time, etc., should be scaled in the proportion t1/l2. The
argument goes as follows:
ma1 ma2 a1 a2 l1 l2
= ⇒ = ⇒ 1 = 2
mg mg g g gt 2 gt 2

Fig 19.16  Geometric scaling. Two horses of different size, but with the This leads to:
same proportions.
t1 l1
=
t2 l2

proportional to a single length measure. An example is given in Time and temporal variables should thus be scaled as the square
Figure 19.16. The big horse has been obtained by enlarging the root of height. A consequence is that speed and velocity should be
small one on a photocopier. For the reference length, the height at scaled inversely with the square root of height:
the withers will be taken and denoted as l. A second assumption to v1 l1 / t1 l1 l2 l
be made relates to the consideration of which two categories of = = = 1
v2 l2 / t1 l2 l1 l2
forces are the most important to our problem? For overground
locomotion, the answer is acceleration forces (= ma) and gravita- A method to account for this scaling with size is to use so-
tional forces (= mg) (Hof, 1996). called dimensionless numbers, the definition of which is given in
Having made this assumption, our scaling should now be made Table 19.1.
such that for animals of different sizes ma and mg should scale in It can easily be verified that these numbers are indeed dimension-
proportion. One reason is that forces are vector quantities, they have less by noting that the factors in the denominator have the same
both a magnitude and a direction, and the direction of the forces units as the quantity in the numerator. The dimensionless number
should be the same irrespective of the size. The constant of gravity for speed (or sometimes the square of it, which can be confusing)
g is practically the same (9.81 m/s2) anywhere on earth, and as long is known as the ‘Froude number.’ Examples of the possibilities of

456
 Scaled energetics of locomotion

these dimensionless numbers are given in Figures 19.17 and 19.18, A practical application was given by Back et al. (1995) in a study
both from Alexander (1977). Figure 19.18 gives dimensionless in which kinematic variables were employed in 4-month-old foals
stride length (the distance traveled in a complete walking stride) as in order to predict the same variables in adult (26-month-old)
a function of dimensionless speed for a number of very diverse horses.
animals. It can be seen that by using dimensionless numbers, i.e.
by accounting for geometric scaling, the differences due to size are
eliminated. The data for all animals could be expressed in a single ANSWER
formula:
If animal B is twice as big as animal A, it will make strides twice as
λˆ = 2.3vˆ 0.6 long. Animal A, on the other hand, can make faster steps. Unjust as it
Figure 19.18 shows a diagram with the gaits adopted as a function is, it turns out that A does not make the steps with twice the
of speed. It can be seen that all animals change from walking to frequency of B, but at a rate proportional to the square root of 2, i.e.
trotting or running at a dimensionless speed of around 0.8. They about 1.4 times as quickly. As a consequence, a corresponding speed
change from trot to gallop somewhere between 1.3 and 2.0. of B, e.g. the speed at which the horse changes from walk to trot or
An example might be helpful. Imagine the Shire going along with from trot to gallop, is 1.4 times as fast as in A. If horses of unequal
the Fallabella at a speed of 2.5 m/s (9 km/h). Dimensionless speed size are to be compared, speeds and temporal variables should thus
for Shire is 2.5/ 1.91 × 9.81 = 0.58, a walking speed. Dimensionless be corrected by a factor equal to the square root of the ratio of the
speed for Fallabella, on the other hand, is 2.5/ 0.80 × 9.81 = 0.89, respective heights at the withers.
and it will be trotting. A small dog, with a height of 30 cm, may gallop
at this speed, as its dimensionless speed 2.5/ 0.3 × 9.81 = 1.46.
The mouse (height 5 cm), which is chased by the dog, has
vˆ = 2.2/ 0.05 × 9.81 = 3.1, and is galloping at top speed. Stride
Differences in proportion
length λ of the big horse can be estimated from the above equation
as λ = 1.91 × λˆ = 1.91 × 2.3 × (0.58)0.6 = 1.91 × 1.65 = 3.16 m. Stride PROBLEM
frequency is thus 2.5 / 3.16 = 0.79 strides/s. The mouse, at the other
extreme, will have a stride length of only 23 cm, but a stride fre- Horses of different breeds are not just bigger or smaller copies of
quency of 11 strides/s. each other, there are substantial differences in morphology. This
section will explore what difference it makes whether a horse is
compact or slender. Which type can carry heavier loads?

5
In the preceding section the scaling was purely geometric: every
measure was multiplied by the same scaling factor. When looking
at a picture of various horse breeds, it is obvious that this is only
^ an approximate picture. Even a layman can see differences in the
λ 2
proportions of Shetland ponies and Thoroughbreds, and the expert
can distinguish many dozens of breeds. In a purely geometric
scaling, as used in the previous section, it will hold that mass (m)
1
is specific mass times volume, while volume is proportional to l0 to
0 0.2 0.5 1 2 5 the third power. The overall relation can be written as m ∝ l03.
In the subsequent parts of this chapter some mathematics cannot
u^ be avoided. Many of the scaling relations that follow will be
Fig 19.17  Relationship between stride length and speed. Dimensionless expressed in the so-called allometric form, as functions of the form
stride length plotted against dimensionless speed in logarithmic y = axb. As a reminder, some properties of functions with exponents
coordinates. Walk, orange circles; trot, red circles; canter, blue squares. The have been given in Table 19.2.
line gives the regression. Data are from Muybridge (1887) on horses, mostly One factor that distinguishes between breeds is whether they are
with riders. of slender or compact build. This might be expressed in a ratio
Alexander, R. McNeil; Principles of Animal Locomotion. ©2003 by Princeton University Press. d/l0, in which d is a characteristic width measure, e.g. the thickness
Reprinted by permission of Princeton University Press. of the midshaft of a forelimb or the circumference of the trunk, and

Fig 19.18  Diagram showing the ranges of speed, expressed as

Man W R the dimensionless number, at which mammals commonly use
their various gaits. W, walk; R, run; T, trot; G, gallop or canter;
Horse W G H, hop; P, pentapedal gait (slow gait, typical for kangaroos).
T
Alexander, R. McNeil; Principles of Animal Locomotion. ©2003 by Princeton
Gnu W G University Press. Reprinted by permission of Princeton University Press.

Gazelle W T G

Cat W T G

Mouse T G

Kangaroo P H

0 1 2 3 4
u^

457
 19 Mechanical analysis and scaling

and the general trend suggests that elastic similarity scaling is fol-
Table 19.2  Some properties of power functions, as used
lowed. Within the horse family, however, the differences in d/l0 ratio
in this chapter
between breeds are more interesting than a general trend over the
widest possible range of animal sizes.
ax b ⋅ cx d = acx b+d
A second application for the above equation is to assess maximum
xa acceleration or sprinting power. According to Newton’s second law
= x a −b
xb a = F/m, and thus maximum acceleration should scale according to
the above equation.
n
xb = xb/n
While data on the height l0 and the mass are commonly available,
( x a )b = x ab a problem is to find data on a measure related to thickness d. A way
to circumvent this problem is to use the body mass equation and
x0 =1
replace d in the subsequent equations by m1/ 2l0−1/ 2. This results in the
alternative forms:
Supported force:
l0 is a characteristic length measure, such as the height at the withers, Fs ∝ m3 / 2l0−5 / 2 force factor
as used in the first section. If the form of our horses may be approxi-
and:
mated with a stack of cylinders, this leads to the following propor-
tionalities for volume or body mass and surface area, respectively: Fs
∝ m1 / 2l0−5 / 2 force/weight
mg
Body mass: m ∝ d l0 2

It can be seen that an ordering according to the force factor is in

Surface area: S ∝ dl0 agreement with common sense. It increases with body mass, and
An important item to be scaled is muscle force. It is generally known thus with muscle mass, and is highest for heavy horses with rela-
from physiology that the force of a muscle is proportional to tively short limbs. The Belgian horse, with Fs = 9400, is the cham-
its physiological cross-section, around 20 N/cm2. Determining the pion in force. The Shire, which has about the same weight, is
physiological cross-sectional area of a muscle is not an easy task, as second in force, due to its longer legs, Fs = 6300. The force/weight
most muscles have quite a complex architecture. It is plausible, factor is highest in Shetland ponies, Fs/m = 13, with their small
however, to assume that it is proportional to the cross-section of mass and short legs. The Quarter Horse also stands out for its
the limb, thus with d2: high acceleration, Fs/m = 8.6. Thoroughbreds, on the other hand,
Muscle force: Fm ∝ d 2 have force/weight factors around 6; they are selected for speed at
longer distances. Standardbreds have factors in between these
The moment that is exerted by the muscle is the product of force extremes.
and moment arm. The moment arm can be assumed proportional The biggest land animal, the African elephant, is the strongest,
to d, so that: but not per kg of body weight, Fs/m = 4.1. In this respect it is far
Moment: M ∝ d 3 surpassed by the ant (mass 2 mg, leg length 3 mm), which sports
When a bent limb has to support a force Fs, the muscle has to supply an Fs/m of almost 3000, and indeed is able to carry loads many
a moment equal to Fsx, in which x is the perpendicular distance times its body weight.
between the joint that is spanned by the muscle and the Table 19.3 gives Fs and Fs /m for a growing Warmblood foal. As
line between the joints proximal and distal of it. When the angle it seems, foals are born with excessively long limbs, compared to
of the limb is not related to size, it can be assumed that x is pro-
portional to l, and thus:
Supported force: Fs ∝ d 3 /l0
Here we are confronted with a problem. In standing, walking or Table 19.3  Size and calculated force factor for a growing Dutch
running the supported force will be equal to 1–2 times body weight, Warmblood foal*
respectively. According to the equation above, body mass, and thus
body weight, is proportional to d2l0. Thus, supported weight and Age Mass Height Fs/1000 Fs/m
body weight depend in different ways on d and l0: (months) (kg) (m)
Fs d 3 /l0 d
∝ 2 = 2 0 57 1.013 0.42 7.31
mg d l0 l0
This means that bigger animals need higher stresses in their muscles 1 101 1.109 0.78 7.76
to support their weight. At some point it becomes too much, and 3 178 1.239 1.39 7.81
that may explain why whales are too big to walk. Horses are far
from the largest species, fortunately, and even elephants are able to 6 262 1.352 2.00 7.62
walk. There are some effects, nevertheless. Small animals walk and 9 331 1.428 2.47 7.47
run with strongly angulated limbs and stand in a rather crouched
position, while large animals keep their legs much straighter. A lying 12 395 1.492 2.89 7.31
dog or cat can quickly jump into a standing position, but a horse
15 452 1.532 3.31 7.32
(or an elephant for that matter) needs considerably more time and
effort to assume a standing posture. 18 494 1.565 3.58 7.25
A possible ‘solution’ to this problem would be to make width d
21 510 1.584 3.65 7.15
proportional to l02, ‘constant stress scaling’ as it is called, in contrast
to true geometric scaling in which all sizes scale in the same propor- 24 519 1.606 3.62 6.97
tion: d ∝ l0. A third alternative is ‘elastic similarity scaling’ (McMahon
& Bonner, 1983; McMahon, 1984) which leads to a proportionality 27 534 1.624 3.67 6.88
d 2 ∝ l03. It can indeed be observed that large animals (elephants)
Data on height and weight from Smolders (1989).
have in general proportionally thicker legs than small ones (mice),

458
 Scaled energetics of locomotion

their weight. When the force/weight ratio is calculated, however, it Kangaroo rat Spring hare Pony
appears to remain quite constant during the growth process, from Ground squirrel Dog
57 kg at birth to an almost full-grown 534 kg.
5
The proportionalities also indicate that, although large horses can
draw or carry heavier loads than small ones (high force factor),
small ones are stronger per kg of body weight. This may explain 4
why donkeys and mules are favored as pack animals over larger
horses. On the other hand, for drawing a heavy cart or plowing, 3
large horses with thick legs are best. Even without the help of bio-

WN–1
mechanics, equine experts have known this for ages. The Shire horse
in Figure 19.15, for example, is said to descend from the horses that 2
carried the knights in their heavy armor in the Middle Ages. It stands
out more as a riding than a pulling horse, in contrast to the Belgian. 1

0
0 1 2 3 4 5 6 7 8
ANSWER Speed (ms–1)

In this section the consequences of the fact that animals have Fig 19.19  Energy consumption per newton of body weight, as a function
different proportions of thickness and length were considered. of speed for (in order of size) a kangaroo rat, a ground squirrel, a spring
Muscle force is proportional to cross-sectional area of the muscle. hare, a dog and a pony.
When this effect is accounted for, it turns out that: Kram, R., Taylor, C.R., 1990. Energetics of running: a new perspective. Reprinted by permission
1. large animals are stronger than small ones, but that this increase from Macmillan Publishers Ltd, Nature Publishing Group, ©1990.
is less than proportional to body weight, therefore;
2. large animals can carry less load per kg of body weight.
These two effects are expressed in, respectively, a force factor
Fs and the force divided by body mass Fs/m. The latter factor When Kram and Taylor (1990) first published this figure, the phe-
explains why an ant can carry many times its body weight, while nomenon had already been known for a considerable time.
a horse cannot. Table 19.3 gives both factors for a growing (Kram and Taylor used body weight mg instead of body mass m,
Warmblood foal. which is more correct, but makes their formulae somewhat more
difficult to compare with those of others.) It is to their merit that
they provided an interpretation, by observing that while E*trans /mg
decreases, step length lc increases by about the same factor (Fig.
19.20B). The latter will not come as a great surprise to the atten-
Energy cost of locomotion tive reader, as step length lc will be proportional to leg length or
A running horse consumes energy; this is obvious to the observer height l0, and in geometric scaling it would hold that l0 ∝ m1/3.
and to the horse. The laws of mechanics do not provide an easy In fact the exponent is slightly less, 0.30. This may be because of
answer, however, to the question how much energy is needed per the fact that small animals run with more angulated legs, while
meter of progression. According to Newton’s first law, in an ideal big animals keep their legs straighter. The overall consequence is
frictionless world the cost of locomotion on the level should be nil. that the product (E*trans lc)/mg is constant (Fig. 19.20C), and
Only uphill locomotion would require energy equal to mgΔh, body is equal to 0.183± 0.045  J/N or m. This means that for all
weight (mg) times rise (Δh). Going upward with a speed of rise vrise animals the energy cost per newton of body weight is the same
would thus require a power: per step: for each step an energy is needed equal to what is
required to make an upward movement of about 20  cm (0.183±
Prise = mgvrise
0.045  m):
For level walking, theory is of no help, and we should rely on mea-
surement. A convenient way to measure energy consumption is to *
Ptrans mg
measure oxygen consumption and recalculate this to the power of = 0.183 ⋅
* v lc
transport Ptrans in watts. In the following an asterisk (*) will be used
to denote work and energy values derived from oxygen consump-
This is a comprehensive summary of a great amount of data. It
tion. Figure 19.19 gives data on running for animals of different
nicely sums up why large animals need less energy to cover a certain
sizes, in which power has already been divided by body weight (in
* distance; it is simply because they can make bigger steps. It further
newtons, N). Two things are apparent from this figure: first, Ptrans
suggests that animals with long legs have an advantage, a fact that
increases linearly with speed in running, without intercept of the
* breeders of whippets and racing horses will readily admit. To give
vertical axis. This means that Ptrans divided by speed v, i.e. the energy
an explanation for these simple facts, on the other hand, is not so
per meter traversed, E*trans, is independent of speed. The second
easy. In the following section some considerations will be given.
aspect is that E*trans depends on size; after dividing by body mass m
One reservation should be made at this point. Kram and Taylor
it is smaller for big animals (see Fig. 19.20A). With m in kg and
(1990) found their experimental relationship from experiments on
E*trans in J, the relation can be expressed as: cost of transport per unit
widely different animals (see Fig. 19.19). The decrease of energy
of body weight, in running; step length, defined as the distance the
consumption per unit mass and distance and the increase of step
body moves forward in a single foot contact; the product of both,
length with body size (Fig. 19.20) are undoubtedly valid for the
E*trans1c /mg, is constant and equal to 0.183/pm 0.045 J/N. Same
range of animals studied. More recent work (Griffin et al., 2004)
animals as shown in Figure 19.19:
has shown that such a relation does not apply within a species. They
E*trans /m = 8.8m−025 determined cost of transport for horses of three different sizes
Another way to express the same relationship is: (Table 19.4). It turns out that among horses cost of transport is
practically independent of body mass over the ranges encompassing
*
Ptrans all horse breeds. The cost for trotting is least in the Arabian
= E*trans = 8.8m0.75
v racing horse.

459
 19 Mechanical analysis and scaling

10
There is not a simple mechanical reasoning to predict how much
energy is needed for horizontal overground locomotion. It can be
measured experimentally in a reasonably simple way, however, by
Slope = –0.25 assessment of oxygen consumption. It turns out that big animals
N–1 m–1

consume less energy (or oxygen) per kilogram of body mass and per
1.0
meter traveled. This can be expressed in a remarkably simple way: the
(gross) energy needed to make a simple step is equal to body weight
times 20 cm, irrespective of size.

0.1
0.1 1 10 100 1000 10,000
A Muscle work and power

1.0
This section attempts to provide some arguments for the decrease in
energy consumption per meter and per kg, as presented in the
previous one. This will require quite a complicated reasoning that
may be omitted on first reading.
m

0.1
Slope = 0.30
Muscles can exert force, and when they are allowed to shorten,
they can do work: force times shortening. The unit of work is the
joule (J). The rate at which the work is done is called power, mea-
0.01 sured in watts (W) or joule/s. The relations between force, work
0.1 1 10 100 1000 10,000 and velocity in muscle are depicted in Figure 19.21. When a
muscle shortens at a certain speed, the force it develops decreases
B
according to the so-called Hill relation (Hill, 1938), (Fig. 19.21A).
As long as there is no lengthening or shortening, the muscle devel-
1.0 ops its ‘isometric’ force F0. With increasing shortening speed
muscle force becomes less, until at a speed v0 force has decreased
0.8 to zero. In lengthening (negative speed in the figure), force can
increase somewhat above isometric, maybe 20–40%. This relation
0.6 has often been observed in all kinds of muscle. It explains the
N–1

common observation that a light load can be moved faster than a

0.4 heavy one.
Slope = 0.04 Mechanical power is force times shortening speed: P = Fv. It can
0.2 be calculated for muscle from the Hill relation (Fig. 19.21B). Power
is zero at two points, when speed is zero, i.e. when the muscle is
0 isometric, and at (and above) the maximum speed, when force is
0.1 1 10 100 1000 10,000 zero. The maximum power is developed at an intermediate speed,
C Body weight (N) around 0.3v0. At that point force has declined to 0.3F0. Power is
thus (with some rounding-off) equal to about 0.1F0v0, with a fairly
Fig 19.20  As a function of body weight mg. (A) Cost of transport per unit flat maximum. An active muscle not only can do mechanical work
of body weight, Etrans *
/mg, in running; (B) step length lc, defined as the but it can generate heat as well, a kind of unavoidable waste, not
distance the body moves forward in a single foot contact; (C) the product of different from any other motor. Two components of heat develop-
*
both, Etrans lc /mg, is constant and equal to 0.183 ± 0.045 J/N. Same animals as ment can be discerned: a constant part and one that increases with
shown in Figure 19.19. shortening speed (Fig. 19.21C). In Figure 19.21C the mechanical
Kram, R., Taylor, C.R., 1990. Energetics of running: a new perspective. Reprinted by permission power of Figure 19.21B has also been drawn, reminding the reader
from Macmillan Publishers Ltd, Nature Publishing Group, ©1990. that both mechanical work and heat are forms of energy. From this
figure can be seen that the ratio (mechanical power)/(total energy
delivered) is not constant for all speeds. It is, of course, zero for
Table 19.4  Data on cost of transport (CoT)*
v = 0 and v = v0 and it has again a maximum in between, but now
at a speed around 0.2v0. At this optimum speed, mechanical work
is generated at an efficiency of about 25%. In this respect a horse
Leg length Mass CoT walk CoT trot does better than a motorcar, which has an efficiency no greater than
(m) (kg) (J.kg−1m−1) (J.kg−1m−1) 15%. Note that energy and power here are net values, to be differ-
entiated from the gross values with an asterisk in the preceding
Miniature 0.73 112 2.23 ± 0.47 2.75 ± 0.48
paragraph.
Arabian 1.24 448 1.98 ± 0.06 2.08 ± 0.09 In the muscle force equation it has already been noted that
maximum isometric muscle force F0 is proportional to muscle
Draught 1.41 715 1.91 ± 0.44 2.39 ± 0.30 physiological cross section, about 20 N/cm2. In a single contrac-
tion, muscle can shorten to about 50% of its initial fiber length,
*Horses of three different sizes (3 each) were trained to walk and trot on a
treadmill. For walking CoT seems to decrease slightly with mass (α m−0.085), but this
but if this is to be done at a reasonably efficient speed, force is no
trend is not significant. For trotting the CoT is lowest for the Arabian racing horse. higher than 0.3F0. As energy is force times shortening, 1 cm3 of
muscle (length 0.01 m) can thus produce 20 N × 0.3 × 0.005 m =
From Griffin et al. (2004), mean ± SD.
0.03 J. With a specific mass of muscle fibers around 1 g/cm3, this

460
 Scaled energetics of locomotion

1.0 0.10

0.8 0.08

0.6 0.06
F/F0

F/F0
0.4 0.04

0.2 0.02

0 0
0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0
A v/v0 B v/v0

0.4 0.8

0.3 A+av+Fv 0.6

Es/F0v0

0.2 0.4
Q

A+av

0.1 0.2
A

0 0
0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0
C v/v0 D v/v0

Fig 19.21  Force, velocity and power in muscle. (A) Muscle force as a function of shortening speed is shown relative to, the force in an isometric contraction.
Shortening speed expressed relative to the muscle’s maximum shortening speed. (B) Mechanical power (work per unit time). (C) Energy production in the
form of activation heat, shortening heat and mechanical power Fv. (D) Efficiency of work production. Only the efficiency of muscle force production is
shown, from ATP to, the processes leading to ATP production reduce the maximal efficiency to about 25%.

gives an energy output of 30 J/kg of muscle for a single contrac- 102

tion. In a real muscle, part of the mass may be due to connective
and fat tissue, which may make an estimate of 20 J/kg more realis-
tic. Maximum energy output of a muscle is thus directly propor-
tional to its mass.
Remembering that the work needed to raise the center of gravity
by an amount Δh in a jump is mgΔh, this explains why animals of
any size, from fleas to horses, can jump about the same height, from
V0/I0 (/s)

0.5–1.0 m. For an 80-kg human, the legs have mass of about 30 kg. 101
Assuming that muscle mass is 20 kg, one would predict an energy
output of 20 × 20 = 400 J and a jumping height of 400/(80.10) =
0.50 m. This is exactly the height well-trained volleyball players can
jump in a standing jump. Horses seem to perform better; they are
reported to jump up to 2.00 m. Taking into account the height of
the trunk in standing, ca. 1 m, a rise of 1 m is still performed. Such
jumps will be made only after a run, however, and with a run and
a fiberglass jumping pole humans can reach almost 6 m. (With a 100
final running speed of 10 m/s, the human athlete has gained a 10-2 10-1 100 101 102 103
kinetic energy of 1 2mv2 = 4000 J. This corresponds to a rise in center Body mass (kg)
of gravity of 5.0 m.) Fig 19.22  Maximal shortening speed divided by muscle fiber length, of
Maximum shortening speed v0 is proportional to muscle length: slow (soleus, blue circles) and fast (extensor digitorum longus, red circles)
a longer muscle contains more sarcomeres in series. In physiology muscles of mouse (data points on left), rat (data points in center) and cat
it is therefore usual to give relative muscle speed in muscle lengths (data points on right). Bigger animals have slower muscles per unit muscle
per second: v/l0. In Figure 19.22 a number of data on the maximum Fv length F0v0. The lines are drawn according to the equation below. No
muscle shortening speed v0/l0 have been given for animals of differ- liability is accepted for extrapolations of these data to the size of horses.
ent sizes, and for a fast muscle (extensor digitorum longus) and a Data from Close (1972).

461
 19 Mechanical analysis and scaling

slow muscle (soleus). It can be seen that maximum shortening Speed records
speed consistently decreases with size of the animal: 20
v0 /l0 = c v m−0.16 18
with cv = six muscle lengths/s for slow muscles like soleus and 16
A
cv = 14 for fast muscles and body mass in kg. For horses between
200 and 1000 kg one might thus expect v0/l0 to be 2–2.6/s for slow, 14 B
and 4.6–6.0/s for fast muscles. 12

Speed (m/s)
We will turn our attention at first to the exponent of m. The data
of Figure 19.22 reflect activities at the cellular level: myosin ATP-ase 10
activity. On the other hand, it is very practical. Remember, from
8
Table 19.1 that in geometric scaling all speeds should be scaled as C
l10/ 2, and next that m ∝ l3. Taken together, this means that v/l0 should 6
scale as m−1/6 = m−0.167, which is in very good agreement with the
measured data of Figure 19.21. It seems that muscle intrinsic speeds, 4
related to enzyme activities, are closely adapted to the speeds that
2
can be expected in an animal of the given size. These size-related
differences are probably genetically determined, as they are already 0
present in newborn animals and are rather insensitive to training. 0 50 100 150 200 250 300 350
It is not known whether this size dependency occurs not only Time (s)
between species of widely different size but also within species, e.g.
Fig 19.23  Speed records for horse and man. (A) World records of horse
horses.
racing, 0.25 mile up to 3 miles; (B) course records of the trotting course
In a similar reasoning as for the work, the maximum mechanical Duindigt, The Netherlands, distances from 900–3600 m; (C) world records
power that can be generated by 1 cm3 tissue, area 1 cm2 and length for men running in 1983, distances 100–2000 m and 30 km (shown at far
1 cm can now be calculated: right).
P(1 cm3 ) = 0.1F0 /l0 = 0.1 ⋅ 20 ⋅ 0.01 ⋅ c v m−0.16 = 0.02 ⋅ c v m−0.16 Data from McWhirter (1984).

Calculated for a muscle mass of m kg, this leads to:

Pmuscle ,peak = 20c v m0.84
Maximum aerobic capacity
The power that can be sustained for more than a few minutes is
For horses between 200 and 1000 kg, this results in Pmuscle,peak =
determined by the aerobic capacity, the maximum oxygen con-
40–50 W/kg for slow and 90–120 W/kg for fast muscle. These are
sumption. Taylor et al. (1981) have determined a general equation
peak power values. In a series of cyclic contractions, averages of
for aerobic capacity as a function of body mass. Recalculated in W,
one-third of the above will be the maximum.
and with body mass m in kg, they found:
Even when it just generates force, without shortening, muscle
generates heat and consumes energy. Data on this heat production/
*
Pmax = 39m0.79 (all animals)
resting energy consumption are very scarce, but there are good Remarkably, horses (and dogs) are in an exceptional position, as
reasons to assume that it is proportional to the maximal work pro- their maximum aerobic capacity is substantially above the average
duction. Thus: line.
For a 500-kg racing horse one would expect, on the basis of the
*
Piso ∝ F0c v m−0.16
equation above, a maximal capacity of 11 W/kg. In fact Thorough-
In Table 19.1 it was seen that time should be scaled by l0 , and as bred horses reach 50–60 W/kg, five times as high (150–170 mLO2/
in geometric scaling l0 ∝ m1/ 3, by temporal factors, like step time kg/min). The exceptional capacities of horses can also be seen from
tc should be proportional to m1/6. Kram and Taylor’s (1990) argu- Figure 19.23: they can run about twice as fast as humans.
ment was that:
F0c v
*
Piso ∝ The maximal aerobic power production, the power that can be
tc generated for longer than a few minutes, can be measured as the
To arrive at running cost of locomotion, they assumed that (a) the maximal oxygen consumption. It increases with body weight, not in
average weight to be supported by muscles is equal to body proportion but as mass to the power 0.79. This means that an animal
weight, and (b) muscle contraction in running is nearly isometric. twice as heavy as a smaller one has not twice the aerobic power, but
The latter seems rather improbable, but there are good arguments only 1.73 times as much.
for it, related to the role of muscle elasticity in locomotion. Even
if it is not completely true, if there is indeed muscle work done, it
To arrive at the maximum aerobic speed one might theoretically
should be considered that this work would still be proportional to
divide maximum aerobic capacity by cost of transport:
F0 and to intrinsic muscle speed v0. The simple expressions thus
require a complex explanation, related to intramuscular enzyme *
Pmax
activities. vmax = *
Etrans
*
For the given values for Pmax = 50 W/kg and E*trans = 2.09 J/kg/m from
The simple experimental fact that energy consumption in locomotion Table 19.4 for Thoroughbreds, one would then arrive at a vmax of no
per meter and per kg body weight decreases with increasing body less than 24 m/s. Unfortunately, this value is way too high (see Fig.
size, has probably to do with the fact that the muscles of large 19.23); maximum racing velocities are around 15 m/s. Here our
animals are slower than those of small ones. This is an innate fine theoretical arguments seem to get stuck. Probably E*trans at
property and related to specific enzyme activities. It is functional, as maximum speed is considerably higher than the optimum values
the large limbs of big animals move more slowly. of Table 19.4, which were obtained on a treadmill that did not allow
higher speeds than 5 m/s.

462
 References

completely futile. Scaling theories give a frame of reference for the

Expressions are given for the maximum speed of horses, unloaded problems discussed: the order of magnitude of the effect is taken
and loaded with a rider, respectively. The outcome is that maximum into account, based on very elementary considerations, and often
speed decreases with mass and increases with leg length, in such a rough approximations. For details one should go further into the
way that those animals of similar shape but different size will run problem. An example is the theory of geometric scaling discussed
equally fast, but that animals with proportionally longer legs have an earlier in this chapter. According to this theory stride time and step
advantage. time in animals of different sizes should be scaled to the square
root of a characteristic length. Sound physical arguments are given
for this. This theory explains some relevant findings, e.g. why your
Dachshund is galloping while you are still walking. Other findings
are not explained, e.g. why Standardbreds trot and Thoroughbreds
Concluding remarks gallop. Scaling theory should certainly be used when one tries to
relate findings on locomotion patterns in rats to the situation in
After this brief journey around the outposts of biology and engi- horses. Without doubt, there will remain major differences between
neering, it is hoped that the reader has been as much entertained rats and horses, but the agreements will emerge more clearly. Any
by the reading as the author has with the writing. Hesitatingly, the predictions from it will be inferior to the insights from experts, but
question of practical applicability of the above might be brought the charming thing is that even simple models can produce realistic
forward. In our opinion such applicability is small, but not predictions.

References
Alexander, R.McN., Bennet-Clark, H.C., Bobbert, M.F., Schamhardt, H.C., 1990. Close, R.I., 1972. Dynamic properties of
1977. Storage of elastic strain energy in Accuracy of determining the point of force mammalian skeletal muscles. Physiol. Rev.
muscle and other tissues. Nature 265, application with piezoelectric force plates. 52, 129–197.
114–117. J. Biomech. 23, 705–710. Colborne, G.R., Lanovaz, J.L., Sprigings, E.J.,
Alexander, R.McN., Goldspink, G., 1977. Brand, R.A., Crowninshield, R.D., Wittstock, Schamhardt, H.C., Clayton, H.M., 1997.
Mechanics and energetics of animal C.E., Pedersen, D.R., Clark, C.R., van Power flow in the equine forelimb. Equine
locomotion. Chapman and Hall, Krieken, F.M., 1982. A model of lower Vet. J. (Suppl 23): 37–40.
London. extremity muscular anatomy. J. Biomech. Colborne, G.R., Lanovaz, J.L., Sprigings, E.J.,
Andriacchi, T.P., 1993. Functional analysis of Eng. 104, 304–310. Schamhardt, H.C., Clayton, H.M., 1998.
pre and post-knee surgery: total knee Brand, R.A., Pedersen, D.R., Davy, D.T., Kotzar, Forelimb joint moments and power in the
arthroplasty and ACL reconstruction. J. G.M., Heiple, K.G., Goldberg, V.M., 1994. walking stance phase of horses. Am. J. Vet.
Biomech. Eng. 115, 575–581. Comparison of hip force calculations and Res. 59, 609–614.
Andriacchi, T.P., Birac, D., 1993. Functional measurements in the same patient. Crowninshield, R.D., Brand, R.A., 1981. A
testing in the anterior cruciate ligament- J. Arthroplasty 9, 45–51. physiologically based criterion of muscle
deficient knee. Clin. Orthop. 40–47. Buchner, H.H.F., Savelberg, H.H.C.M., Becker, force prediction in locomotion. J. Biomech.
Back, W., Barneveld, A., van Weeren, P.R., van C.K., 1996. Load redistribution after 14, 793–801.
den Bogert, A.J., 1993. Kinematic gait desmotomy of the accessory ligament of de Koning, J.J., de Groot, G., van Ingen
analysis in equine carpal lameness. Acta the deep digital flexor tendon in adult Schenau, G.J., 1991. Coordination of leg
Anat. (Basel) 146, 86–89. horses. Vet. Quart. 18, S70–74. muscles during speed skating. J. Biomech.
Back, W., Schamhardt, H.C., Hartman, W., Buchner, H.H.F., Savelberg, H.H.C.M., 24, 137–146.
Bruin, G., Barneveld, A., 1995. Predictive Schamhardt, H.C., Barneveld, A., 1997. Delp, S.L., Loan, J.P., 1995. A graphics-
value of foal kinematics for the locomotor Inertial properties of Dutch Warmblood based software system to develop and
performance of adult horses. Res. Vet. Sci. horses. J. Biomech. 30, 653–658. analyze models of musculoskeletal
59, 64–69. Burdett, R.G., 1982. Forces predicted at the structures. Comput. Biol. Med. 25,
Bartel, D.L., Schryver, H.F., Lowe, J.E., Parker, ankle during running. Med. Sci. Sports 21–34.
R.A., 1978. Locomotion in the horse: a Exerc. 14, 308–316. Deuel, N.R., Park, J., 1993. Gallop kinematics
procedure for computing the internal Cavagna, G.A., Kaneko, M., 1977. Mechanical of Olympic three-day event horses. Acta
forces in the digit. Am. J. Vet. Res. 39, work and efficiency in level walking and Anat. (Basel) 146, 168–174.
1721–1727. running. J. Physiol. (Lond.) 268, 647–681. Devita, P., Skelly, W.A., 1992. Effect of landing
Becker, C.K., Savelberg, H.H., Buchner, H.H., Clayton, H.M., 1996. Instrumentation and stiffness on joint kinetics and energetics in
Barneveld, A., 1998. Long-term techniques in locomotion and lameness. the lower extremity. Med. Sci. Sports Exerc.
consequences of experimental desmotomy Vet. Clin. North Am.: Equine Pract. 12, 24, 108–115.
of the accessory ligament of the deep 337–350. Dow, S.M., Leendertz, J.A., Silver, I.A.,
digital flexor tendon in adult horses. Am. J. Clayton, H.M., Lanovaz, J.L., Schamhardt, Goodship, A.E., 1991. Identification of
Vet. Res. 59, 347–351. H.C., Willemen, M.A., Colborne, G.R., subclinical tendon injury from ground
Bell, A.L., Pedersen, D.R., Brand, R.A., 1990. A 1998. Net joint moments and powers in reaction force analysis. Equine Vet. J. 23,
comparison of the accuracy of several hip the equine forelimb during the stance 266–272.
center location prediction methods. phase of the trot. Equine Vet. J. 30, Dul, J., Townsend, M.A., Shiavi, R., Johnson,
J. Biomech. 23, 617–621. 384–389. G.E., 1984. Muscular synergism – I. On
Biewener, A.A., Thomason, J.J., Lanyon, L.E., Clayton, H.M., Willemen, M.A., Lanovaz, J.L., criteria for load sharing between synergistic
1988. Mechanics of locomotion and Schamhardt, H.C., 2000. Effects of a heel muscles. J. Biomech. 17, 663–673.
jumping in the horse (Equus): in vivo wedge in horses with superficial digital Dyce, K.M., Sack, W.O., Wensing, C.J.G., 1987.
stress in the tibia and metatarsus. J. Zool. flexor tendinitis. Vet. Comp. Orthop. Textbook of veterinary anatomy. Saunders
214, 547–565. Traum. 13, 1–8. company, Philadelphia.

463
 19 Mechanical analysis and scaling

Elftman, H., 1939. Forces and energy changes Merkens, H.W., Schamhardt, H.C., 1988. Smolders, G., 1989. Groei en ontwikkeling
in the leg during walking. Am. J. Physiol. Evaluation of equine locomotion during van rijpaarden. (GrowthÆand development
125, 339–356. different degrees of experimentally induced of riding horses.) PR Praktijkonderzoek 2,
Eng, J.J., Winter, D.A., 1995. Kinetic analysis lameness. II: Distribution of ground 26–28.
of the lower limbs during walking: what reaction force patterns of the concurrently Soderkvist, I., Wedin, P.A., 1993. Determining
information can be gained from a loaded limbs. Equine Vet. J. (Suppl. 6): the movements of the skeleton using
three-dimensional model? J. Biomech. 28, 107–112. well-configured markers. J. Biomech. 26,
753–758. Minetti, A.E., Ardig, O.L., Reinach, E., Saibene, 1473–1477.
Glitsch, U., Baumann, W., 1997. The F., 1999. The relationship between Taylor, C.R., Maloiy, G.M.O., Weibel, E.R.,
three-dimensional determination of mechanical work and energy expenditure et al., 1981. Design of the mammalian
internal loads in the lower extremity. of locomotion in horses. J. Exp. Biol. 202, respiratory system. III Scaling maximum
J. Biomech. 30, 1123–1131. 2329–2338. aerobic capacity to body mass: wild and
Gregoire, L., Veeger, H.E., Huijing, P.A., van Muybridge, E., 1887. Animals in Motion;, domestic mammals. Resp. Physiol. 44;:
Ingen Schenau, G.J., 1984. Role of modern reprint 1979. Dover, New York. 25–37.
mono- and biarticular muscles in Neptune, R.R., van den Bogert, A.J., 1998. Thunnissen, J., 1993. Muscle force prediction
explosive movements. Int. J. Sports Standard mechanical energy analyses do during human gait. PhD thesis, University
Med. 5, 301–305. not correlate with muscle work in cycling. of Twente.
Griffin, T. M., Kram, R., Wickler, S. J., Hoyt, D. J. Biomech. 31, 239–245. Thys, H., Willems, P.A., Saels, P., 1996. Energy
F., 2004. Biomechanical and energetic Nicodemus, M.C., Lanovaz, J.L., Corn, C., cost, mechanical work and muscular
determinants of the walk-trot transition in Clayton, H.M., 1999. The application of efficiency in swing-through gait with elbow
horses. J. Exp. Biol. 207, 4215-4223. virtual markers to a joint coordinate system crutches. J. Biomech. 29, 1473–1482.
Hartman, W., Schamhardt, H.C., Lammertink, for equine three-dimensional motions. Van den Bogert, A.J., 1994. Analysis and
J.L., Badoux, D.M., 1984. Bone strain in the Proc. 16th Equine Nutr. Physiol. Symp. pp. simulation of mechanical loads on the
equine tibia: an in vivo strain gauge 24–25. human musculoskeletal system: a
analysis. Am. J. Vet. Res. 45, 880–884. Ounpuu, S., Gage, J.R., Davis, R.B., 1991. methodological overview. Exerc. Sport Sci.
Herzog, W., Leonard, T.R., 1991. Validation of Three-dimensional lower extremity joint Rev. 22, 23–51.
optimization models that estimate the kinetics in normal pediatric gait. J. Pediatr. Van den Bogert, A.J., Hartman, W., Sauren,
forces exerted by synergistic muscles. Orthop. 11, 341–349. A.A.H.J., 1989. Measurement and
J. Biomech. 24, 31–39. Paul, J.P., 1971. Load actions on the human computation of inertial parameters in the
Hill, A.V., 1938. The heat of shortening and femur in walking and some resultant horse. In: van den Bogert, A.J. (Eds.),
the dynamic constants of muscle. Proc. stresses. Exp. Mech. 11, 121–125. Computer simulation of locomotion in the
Royal Soc. Series B. 126, 136–195. Pennycuick, C.J., 1992. Newton rules horse. PhD thesis, University of Utrecht,
Hof, A.L., 1996. Scaling gait data to body size. biology. A physical approach to pp. 67–89.
Gait Posture 4, 222–223. biological problems. Oxford University Van den Bogert, A.J., Read, L., Nigg, B.M.,
Jansen, M.O., van den Bogert, A.J., Riemersma, Press, Oxford. 1999. An analysis of hip joint loading
D.J., Schamhardt, H.C., 1993. In vivo Pratt Jr., G.W., O’Connor Jr., J.T., 1976. Force during walking, running, and skiing. Med.
tendon forces in the forelimb of ponies at plate studies of equine biomechanics. Am. Sci. Sports Exerc. 31, 131–142.
the walk, validated by ground reaction J. Vet. Res. 37, 1251–1255. Van den Bogert, A.J., Sauren, A.A.H.J., 1989.
force measurements. Acta Anat. (Basel) Riemersma, D.J., van den Bogert, A.J., Jansen, Implementation of curves
146, 162–167. M.O., Schamhardt, H.C., 1996. Tendon musculotendinous structures in rigid body
Jones, D.A., Round, J.M., 1990. Skeletal strain in the forelimbs as a function of gait dynamics. In: van den Bogert, A.J. (Eds.),
muscle in health and disease. Manchester and ground characteristics and in vitro Computer simulation of locomotion in the
University Press, Manchester. limb loading in ponies. Equine Vet. J. 28, horse. PhD thesis, University of Utrecht,
Kram, R., Taylor, C.R., 1990. Energetics of 133–138. pp. 37–55.
running: a new perspective. Nature 346, Riemersma, D.J., van den Bogert, A.J., Van den Bogert, A.J., Schamhardt, H.C.,
265–267. Schamhardt, H.C., Hartman, W., 1988. Hartman, W., 1988. Relationship between
Kuo, A.D., 1998. A least-squares estimation Kinetics and kinematics of the equine simultaneously measured muscle
approach to improving the precision of hind limb: in vivo tendon strain and activation, length, force and power in the
inverse dynamics computations. joint kinematics. Am. J. Vet. Res. 49, horse. In: Proc. 6th Congr. Europ. Soc.
J. Biomech. Eng. 120, 148–159. 1353–1359. Biomech. p C 32
Leach, D.H., Dyson, S., 1988. Instant centres Robertson, D.G., Winter, D.A., 1980. Van den Bogert, A.J., Smith, G.D., Nigg, B.M.,
of rotation of equine limb joints and their Mechanical energy generation, absorption 1994. In vivo determination of the
relationship to standard skin marker and transfer amongst segments during anatomical axes of the ankle joint complex:
locations. Equine Vet. J. (Suppl 6): walking. J. Biomech. 13, 845–854. an optimization approach. J. Biomech. 27,
113–119. Rose, S.A., DeLuca, P.A., Davis, R.B.D., 1477–1488.
Lindenfeld, T.N., Hewett, T.E., Andriacchi, T.P., Ounpuu, S., Gage, J.R., 1993. Kinematic Van Ingen Schenau, G.J., 1994. Proposed
1997. Joint loading with valgus bracing in and kinetic evaluation of the ankle after actions of bi-articular muscles and the
patients with varus gonarthrosis. Clin. lengthening of the gastrocnemius fascia in design of hind limbs of bi- and
Orthop. 344, 290–297. children with cerebral palsy. J. Pediatr. quadrupeds. Hum. Movem. Sci. 13,
McMahon, T.A., 1984. Muscles, reflexes and Orthop. 13, 727–732. 665–681.
locomotion. Princeton University Press, Schmidt-Nielsen, K., 1984. Scaling: Why is Van Ingen Schenau, G.J., Cavanagh, P.R., 1990.
Princeton. Animal Size so Important? Cambridge Power equations in endurance sports.
McMahon, T.A., Bonner, J.T., 1983. University Press, Cambridge. J. Biomech. 23, 865–881.
On size and life. WH Freeman & Co., Schryver, H.F., Bartel, D.L., Langrana, N., Van Weeren, P.R., Jansen, M.O., van den
New York. Lowe, J.E., 1978. Locomotion in the Bogert, A.J., Barneveld, A., 1992a. A
McWhirter, N., 1984. The guiness book of horse: kinematics and external and internal kinematic and strain gauge study
records, Dutch ed. Luitingh, Utrecht. [In forces in the normal equine digit in the of the reciprocal apparatus in the
more recent editions, data on horse racing walk and trot. Am. J. Vet. Res. 39, equine hind limb. J. Biomech. 25,
are absent.] 1728–1733. 1291–1301.

464
 References

Van Weeren, P.R., van den Bogert, A.J., Back, Vaughan, C., Davis, B., O’Connor, J., Winter, D.A., Eng, P., 1995. Kinetics: our
W., Bruin, G., Barneveld, A., 1992b. 1992. Dynamics of Human gait, first ed. window into the goals and strategies of the
Correction models for skin displacement in Human Kinetics Publishers, Champaign, central nervous system. Behav. Brain Res.
equine kinematic gait analysis. J. Equine Illinois. 67, 111–120.
Vet. Sci. 12, 178–192. Vaughan, C.L., 1984. Computer Winter, D.A., Robertson, D.G., 1978.
Van Weeren, P.R., van den Bogert, A.J., Back, simulation of human motion in Joint torque and energy patterns
W., Bruin, G., Barneveld, A., 1993. sports biomechanics. Exerc. Sport in normal gait. Biol. Cybern. 29,
Kinematics of the Standardbred trotter Sci. Rev. 12, 373–416. 137–142.
measured at 6, 7, 8 and 9 m/s on a Winter, D., 1990. Biomechanics and motor Yeadon, M.R., Challis, J.H., 1994. The future
treadmill, before and after 5 months of control of human movement, 2nd edn. of performance-related sports biomechanics
prerace training. Acta Anat. (Basel) 146, Wiley, New York. research. J. Sports Sci. 12, 3–32.
154–161.

465
This page intentionally left blank
 C H A P T ER 20 

Modeling, simulation and animation

Adam Arabian, Jonathan Merritt, Franca Jonquiere, Antonie J. Van den Bogert,
Hilary M. Clayton

Development of a computer model supports, or even entire complex mechanisms such as automotive
engines could be examined for fit and accuracy before any real
monetary investment was made by the company.
Introduction
Computer modeling is generally defined as the process of creating Computer modeling and biomechanics
a representation of an object or system within the computer. The
origins lie in tools developed by engineers to assist them in the Even before solid modeling became an industry standard, ana­
industrial design process. These models allowed the engineer to tomists and biomechanists had recognized the opportunities it
create each component of a product, assemble them ‘virtually’ in represented. One of the earliest uses of computers to model a mus-
the computer, and ensure a good fit before manufacturing. Using culoskeletal system represented an approximation of the human
computer mathematical analysis tools, it is also possible to predict body as a series of rigid links with simple hinge or pin joints
the forces that the component will see and make certain that it will (Chaffin, 1969). Although this was useful in visualization of human
not fail under that loading. motion, it was not a tool for examining specific bones or joints. It
For biological systems, computer models are useful for many of was not until medical imaging devices such as computed tomogra-
the same reasons. Instead of having to perform in vivo studies, a phy (CT) and magnetic resonance imaging (MRI) scanners were
computer representation of an anatomical structure can be created used as the basis for the geometry that truly representative anatomy
and examined in ways that would be prohibitively difficult in a live could be created in a computerized model.
animal. The advantages are obvious; if the model is reliable, it can After the techniques of modeling had been proven in human
be used to examine many biomechanical systems including the studies, the veterinary community began to use the same methods
motion of a complex joint, such as the tarsus, the loading and to explore the complexities of motion in animals. An early equine
motion in regions that are not visible to the human eye, such as study using the fundamentals of joint motion developed for human
within the hoof capsule, or the loading that causes a bone to frac- protocols modeled the entire forelimb distal to the metacarpopha-
ture. These can be studied by modeling, which does not require the langeal joint (Cheung & Thompson, 1993), but subsequent research
use of live subjects. has typically focused on a single joint, such as the carpus (Les et al.,
1997) or fetlock (Thompson & Cheung, 1993).

History of computer modeling

The origins of computer modeling can be traced back to a program
Applications of computer models
developed in 1963 at the Massachusetts Institute of Technology There are four typical applications of computer models in biome-
(Encarnacao et al., 1990). Its capabilities were limited to what chanics: visualization, kinematics, dynamics and finite element
would be called a simple drawing program in modern terms, but analysis.
many of the fundamental drawing subroutines and standards were
established by this work. Although developments were made in
subsequent years, it wasn’t until the late 1970s that computer-based
Visualization
design moved from the realm of academic theory into mainstream The potential of computer models in visualization represents one
science. As the increasing power of computers allowed for more of the great emerging applications of this technology. A well-known
advanced applications, a new technique was developed in the 1980s example is the use of CT or MRI data by a veterinarian or physician
called ‘solid modeling.’ In this process a three-dimensional repre- to produce a three-dimensional recreation of a bone or joint, rather
sentation of a structure is created within a computer, rather than than looking at it as a series of two-dimensional projections. This
the older method of simply having a number of two-dimensional functionality is included in a number of commercial medical
projections. imaging products. The operator uses specific areas as ‘edges’ of
Using this new technology, designers could view parts of a system structures and the computer rebuilds a visualization of the bone,
and their interactions without the cost and time involved in creating soft tissue or any other object of interest. This type of modeling is
a prototype. Assemblies of independent components, structural useful for viewing different structures and the relationships between

467
 20 Modeling, simulation and animation

them in live subjects, but some of its most promising practical

applications arise from the ability to produce physical objects by
the processes of rapid prototyping and machining.
Rapid prototyping is a method by which a sample part is
made from (typically) a soft plastic or other suitable material that
can be held and examined, not just viewed on a computer screen.
The process consists of creating a solid model in the computer, then
sending the information to a rapid prototyping machine, which
recreates it as a series of layers of the material. In industry, entire
prototype engines have been constructed this way to demonstrate
new technology and concepts that would be difficult to otherwise
describe. Similarly, anatomical structures can be recreated and
examined in the same way as a dissected specimen. This has been
explored in humans as a method of preparing surgeons to perform
operations, including neurological and orthopedic procedures
(McGurk et al., 1997).
Machining is a process by which a product is manufactured from
a permanent material. For orthopedic procedures, this allows pre-
liminary evaluation of surgeries and the manufacturing of custom
implants. Using CT data obtained preoperatively, the joint can be
recreated precisely in the computer and a surgeon can have an exact
fit for the orthopedic implant prepared prior to surgery or even
constructed during the procedure (Mulier et al., 1989).
One of the most exciting areas into which visualization tech-
niques are expanding is the field of virtual reality. The concept of
virtual reality has become something of a buzzword of the media
in recent years due to proliferation of home computers, increased
distance interaction on the internet, and incredible advances in
computer graphics. The true concept of virtual reality is that, by
recreating a situation in the processor of a computer and by using
highly realistic graphics, the human brain can be fooled into believ-
ing that the computer-generated images are real. Although techno-
logical limitations still prevent complete immersion in a virtual
reality environment, some early demonstrations of its medical
applications have been performed, such as the National Library of Fig 20.1  Persival, the Ecole Nationale d’Equitation’s mechanical horse
Medicine’s Visible Human Project. This is an ambitious program to simulator.
develop a complete database of human anatomy based on cryosec-
tions and CT scans of male and female cadaver specimens. The
results of this research have already been used to develop educa-
tional and research tools for audiences ranging from school-age Kinematics (movement)
children to medical professionals. Kinematics is defined as the observation of motion in a mechanical
Another focus for the development for visualization tools in system. It encompasses a spectrum that ranges from the rotation of
medicine is in preoperative planning. An entire surgery can be a horse’s limbs as it gallops across a polo field, to the movements
simulated and performed any number of times to rehearse the of a player’s arm and mallet as a shot is made, to the way the polo
mechanics of the procedure prior to a single incision being made ball arcs through the air after it has been hit. Kinematic studies have
on a live patient. However, the use of a purely anatomical model described the horse’s movement at various gaits and during jumping.
cannot completely prepare the student for all situations and com- This information can be used to create an accurate simulation of
plications that may arise during a procedure and that necessitate a the motion of a particular joint that shows how the various com-
rapid and accurate response by the surgeon. However, in the same ponents interact and allows the user to visualize the effects of
way that pilots can practice emergency maneuvers and landings in disease, lameness or surgeries. Other types of computer simulations
jet simulators, critical conditions could be simulated to prepare a might show deformation of the structures within the hoof during
student for surgical emergencies. weight bearing, or the actions of the rider in giving a horse the aids
The application of virtual reality in the field of equine biome- for different movements.
chanics is quite advanced. An early example is ‘Persival,’ a mechani- It is important to recognize that kinematics is limited to the
cal horse simulator (Fig. 20.1) developed by Patrick Galloux of the process of observation; it does not have the ability to make infer-
Ecole Nationale d’Equitation in Saumur, France (Galloux et al., ences. For example, kinematics can be used to study how a bone,
1994). His group has constructed a mechanical simulator that recre- joint, limb or animal moves based on a set of data describing that
ates the movements of a horse at a walk, trot and canter, as well as movement, but it cannot predict the results of a perturbation of the
during jumping. Virtual reality is used to allow the rider to follow system, such as an injury or lameness. Using computer models as
a show jumping course by watching a screen in front of the kinematic simulations is, to a great extent, a specific application
simulator or by wearing virtual reality glasses. The rider sees the of the previously discussed visualization tools. However, it goes
obstacles, and sensors are used to detect and react to the rider’s aids. beyond simple visualization because the kinematic information
It has been found that by using this simulator, the time required to represents a bridge between visualization, a purely external observa-
train a rider was shorter than when using a real horse, and the tion, and dynamics, an understanding of the forces that are respon-
novice rider was able to learn safely and without the need for sible for creating the movement and the forces that arise as a
‘lesson’ horses. A horse simulator is also available that allows the consequence of the movement. Halvorsen et al. (2008) used filter-
ride to experience and replay a virtual dressage test (http:// ing for optimally tracking kinematics. Bobbert & Santamaría (2005)
www.racewood.com/dressage.php). and Bobbert et al. (2007) used body kinematics to calculate limb

468
 Development of a computer model

dynamics (Bobbert et al., 2007) and energetics (Bobbert & Sant- both clinical and research settings to study the function and dys-
amaria, 2005). function of the limb at the different joints (Buchner et al., 1996;
Clayton et al., 1998). Brown et al. (2003) developed a detailed
musculoskeletal model of the distal equine forelimb to study the
Kinetics (dynamics) influence of musculoskeletal geometry (i.e. muscle paths) and
As stated above, kinematics is a passive observation of how a system muscle physiology (i.e. force–length properties) on the force-
moves. Dynamics uses kinematic data, combined with additional and moment-generating capacities of muscles crossing the carpal
information, to evaluate the forces within the system. Having cal- and metacarpophalangeal joints. They predicted that the suspen-
culated the dynamics, much more complex simulations can be sory and check ligaments contributed more than half of the total
performed. Two methods are primarily used for the development support moment developed about the MCP joint in the model.
of dynamic results: inverse dynamics, and forward or direct dynam- Additional information can be added to the model to increase the
ics. In general, inverse dynamics is a technique by which internal level of sophistication. For example, knowledge of the orientation
forces are estimated using data from external observations, while of anatomic structures such as the line of action of tendons can
forward dynamics predicts external forces and motions by prescrib- allow for a more detailed estimation of internal muscle and tendon
ing internal forces. An application of computer modeling using forces (Riemersma et al., 1988; Jansen et al., 1993). Merritt et al.
inverse dynamics involves the calculation of net joint moments and (2008) developed a two-dimensional model of the musculoskeletal
joint powers from knowledge of the kinematics, ground reaction system of the distal limb region and applied it to kinematic and
forces (GRFs) and morphological data using a link-segment model. kinetic data from walking and trotting horses. The forces in major
In this type of model, the complex structure of the limbs is simpli- tendons and joint reaction forces were calculated. The components
fied by representing the limb segments as rigid bodies linked by of the joint reaction forces caused by wrapping of tendons around
hinge joints (Schryver et al., 1978; Hjertén & Drevemo, 1987; sesamoid bones were found to be of similar magnitude to the reac-
Clayton et al., 1998; Colborne et al., 1998; Lanovaz et al., 1999) tion forces between the long bones at each joint. This finding high-
(Fig. 20.2). The musculature that drives the limb has a complex lighted the importance of taking into account muscle-tendon
geometry and many mechanical and material properties that are wrapping when evaluating joint loading in the equine distal fore-
difficult to quantify. In order to study the effect of muscle action limb (Figs 20.3–20.6).
without defining all of the specific parameters of the system, the net Swanstrom et al. (2005) developed a musculoskeletal model of
muscular action within the limb is represented in the model as net the Thoroughbred forelimb and a dynamic simulation of the
torques or net joint moments acting at each of the joints. Zarucco motion of the distal segments during the stance phase of high-speed
et al. (2003) developed and tested an experimental model for (18 m/s) gallop. The musculoskeletal model was comprised of
in-vivo short-term recording of peak isometric forces of the digital segment, joint, muscle-tendon, and ligament information. The
flexor muscles in the forelimb of adult horses and thus provided dynamic simulation incorporated a proximal forward-driving force,
further insight to functional implications of the complex architec- a distal ground reaction force model, muscle activations, and initial
ture of these muscles. Realistic inertial parameters such as mass and positions and velocities. A simulation of the gallop after transection
moment of inertia are assigned to each segment of the computer of an accessory ligament demonstrated increased soft tissue strains
model. Kinematic data and GRFs are measured and used as inputs
to the model. An inverse dynamics solution is used to calculate the
torques required to cause the observed motion and forces. This ‘net
effect’ of muscle action derived from the model can be applied in

Metacarpus
Interosseous
ligament (lL)

Fetlock joint
Superficial
digital flexor Pastern
tendon (SDFT)
Coffin joint

Deep digital
flexor tendon
Hoof
(DDFT)

Fig 20.3  Two-dimensional, sagittal plane model of the equine forelimb.

The model contained three rigid body segments and three major tendinous
structures.
Reprinted from Merritt J.S., Davies H.M.S., Burvill C., Pandy M.G., 2008. Influence of
muscle-tendon wrapping on calculations of joint reaction forces in the equine distal
forelimb, Journal of Biomedicine and Biotechnology, Hindawai Publishing Corporation, with
Fig 20.2  Link-segment model of the equine forelimb. kind permission from Jonathan Merritt.

469
 20 Modeling, simulation and animation

simulation was found to accurately recreate the ground reaction

force patterns measured in research studies.
Musculoskeletal models have been used to investigate underlying
concepts of equine locomotion that are often broadly accepted, but
not understood in detail. Merritt et al. (2010) investigated the func-
tion of the fetlock joint and found that the metacarpus was loaded
Metacarpus in compression by the combined actions of the proximal sesamoid
bones and the first phalanx. Although this kind of compressive
loading was already understood to exist in the metacarpus, the study
Fsdf Fddf illustrated the extent to which this loading depended upon the con-
Fil
certed function of the entire forelimb, rather than being a feature of
the fetlock alone. Harrison et al. (2010) utilized an extensive forelimb
Fpses model to investigate elastic energy storage and release in the distal
Fmcpj forelimb. Elastic energy storage had long been considered an impor-
Fddf tant feature of locomotion in large animals, but these authors were
Fil Fsdf Fil able to quantify its role in detail for the first time. Even in the
Proximal First phalanx muscular structures of the distal forelimb, such as the superficial and
sesamoid Fsdf deep digital flexor muscles, these authors reported that between 69–
bones Second phalanx 90% of the total work was done by elastic energy release, rather than
by active muscle contraction, with the amount of elastic energy
Third phalanx storage increasing at faster speeds. Studies of this kind help to close
Fnav Fdipj critically important gaps in overall understanding, by providing the
Fddf
detailed mechanisms of action for features of locomotion, which
Fddf were previously understood in only the broadest terms.
The availability of a complete and accurate model of even a single
Navicular bone joint opens up incredible opportunities. Should a specific treatment
be recommended, such as a tenectomy, the outcome can be pre-
dicted by incorporating into the model the loss of the forces trans-
Ground reaction
force mitted by that tendon. The model could then be used to predict the
effectiveness of the treatment, allowing for a more educated and
Fig 20.4  Forces considered when calculating the joint reaction forces at confident recommendation of the likely benefits. This application
the fetlock and coffin joints. The forces were the ground reaction force, the was validated by a study in which the human lower leg was recre-
force exerted by the distal phalanx on the middle phalanx (Fdipj), the force ated in a computer to evaluate orthopedic surgical procedures. It
exerted by the navicular bone on the middle phalanx (Fnav), the DDFT force was found that the model accurately predicted the effects of tendon
(Fddf), the SDFT force (Fsdf), the IL force (Fil), the force exerted by the transfers and lengthening procedures (Delp et al., 1990). It is within
proximal phalanx on the distal metacarpus (Fmcpj), and the force exerted the realm of possibility that a sufficiently complex model of the
by the proximal sesamoid bones on the distal metacarpus (Fpses). equine limb could be created that it would be capable of assessing
Reprinted from Merritt J.S., Davies H.M.S., Burvill C., Pandy M.G., 2008. Influence of the effects of orthopedic shoeing, changes in the physical properties
muscle-tendon wrapping on calculations of joint reaction forces in the equine distal of the footing, muscular atrophy, or surgical procedures such as
forelimb, Journal of Biomedicine and Biotechnology, Hindawai Publishing Corporation, with tenectomy or arthrodesis.
kind permission from Jonathan Merritt.

in the remaining support structures of the distal forelimb, consis- Finite element analysis
tent with those previously reported from in vitro studies. Forward The stress on a structure can be measured through a dynamic analy-
or direct dynamics utilizes data describing muscle activation pat- sis as described above or it can be computed through application
terns and the force of muscular contraction to determine the posi- of mechanical measurement techniques such as strain gauges.
tion of each segment of the limb. Although this method is Having determined the stress on a structure, it is of great interest to
theoretically ideal, as it is the most accurate representation of what know what effect this might have on the structure in terms of
is actually occurring in the system, it is practically difficult to causing deformation (strain) or stimulating biological adaptation.
measure these inputs with the accuracy needed to produce a model For example, hoof deformation at impact could be used to evaluate
that represents the internal forces and interactions. A variation on how energy is dampened or dissipated. Observation of the habitual
this type of dynamic analysis was performed for an entire horse in patterns of stress in a bone can be related to the way in which that
1987 using a mechanical mechanism analysis package (Van den bone adapts to its environment and reconstructs itself. The com-
Bogert, 1989). The horse was represented by a series of rigid bodies plexity of these types of calculations makes it almost impossible to
connected by springs, dampers and torque producers representing accurately determine or calculate by hand the way in which forces
tendons and ligaments. The innovative aspect of this simulation was will be distributed across a particular structure.
that, instead of simply prescribing the motion of a segment by In the field of engineering, this problem has been solved by the
kinematic data, an intelligent model was created that prescribed the development of the process called finite element analysis (FEA) or
limitations of joints, the reaction forces created by the tension in finite element modeling (FEM). To understand the process of FEA,
ligaments and muscle activations. The model was analyzed and consider a very simple piece of geometry, a cube. It is relatively easy
validated through both kinematic and ground reaction force data. to predict the way a cube deforms based on the forces applied and
However, in order to simulate the entire horse, a number of assump- the properties of the material from which the cube is made. A steel
tions had to be made due to the complexity of the project and the cube deforms in a specific way, which is different from an alumi-
limitations in computer power at the time. The entire system was num or wood cube.
assumed to be purely two dimensional with the segments of the Most objects are more complicated than simple cubes. They are
horse moving in the sagittal plane. Complex joints, such as the made up of very complex surfaces and curves, which make it diffi-
carpus and tarsus, were approximated as simple hinges, and it was cult, if not impossible, to predict how the structure will deform
necessary to completely ignore certain joints, such as the coffin joint under stress. The process of FEA consists of breaking up a piece of
and the articulation of the spine. In spite of these assumptions, the complex geometry into a series of simple shapes, such as cubes or

470
 Development of a computer model

Walk Trot
12 12
10 10

Force (N/kg bwt)

Force (N/kg bwt)

8 8
6 6
4 4
2 2
Distal phalanx 0 0
0 20 40 60 80 100 0 20 40 60 80 100
Stance (%) Stance (%)

10 10

8 8
Force (N/kg bwt)

Force (N/kg bwt)

6 6

4 4

2 2
Navicular bone
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Stance (%) Stance (%)

80 80
70 70
60 60
Force (N/kg bwt)

Force (N/kg bwt)

50 50
40 40
30 30
20 20
10 10
Proximal phalanx 0 0
0 20 40 60 80 100 0 20 40 60 80 100
Stance (%) Stance (%)

80 80
70 70
60 60
Force (N/kg bwt)

Force (N/kg bwt)

50 50
40 40
30 30
20 20
Proximal 10 10
sesamoid bones 0 0
0 20 40 60 80 100 0 20 40 60 80 100
Stance (%) Stance (%)

Fig 20.5  Magnitudes of calculated joint reaction force components from long bones and sesamoid bones in the coffin and fetlock joints for walking and
trotting during the stance phase of the stride. The magnitudes are shown as solid lines with dashed lines to represent ±1SD. The stance phase of the stride
was the time during which the vertical component of the ground reaction force was greater than 50N. 0% of stride was the time of first hoof–ground
contact, 100% of stride was the time of last hoof–ground contact.
Reprinted from Merritt J.S., Davies H.M.S., Burvill C., Pandy M.G., 2008. Influence of muscle-tendon wrapping on calculations of joint reaction forces in the equine distal forelimb, Journal of
Biomedicine and Biotechnology, Hindawai Publishing Corporation, with kind permission from Jonathan Merritt.
pyramids. These subdivisions, called elements, react to stress in a in three directions (up/down, left/right and forward/backward).
mathematically predictable way. The entire structure is represented When we take into account the fact that a finite element model can
by numerous elements, which are each described in terms of the consist of thousands of these elements, it becomes evident that a
motion of their corners, or in more complex models, the motions powerful computer is an essential tool in performing this type of
of the midpoints between the corners as well. When the model is analysis.
analyzed, the deformation of one element defines the force on the Biological structures are even more complex than most products
next element, and subsequently its force on the next, and so on. manufactured in industry, making finite element analysis invalu-
The minimal requirement is to analyze each corner of each element able for predicting their behavior under various loading conditions.

471
 20 Modeling, simulation and animation

35

30

25
Moment arm (mm)

20

15

10

0
190 195 200 205 210 215 220
Fetlock angle (degrees)

Model SDFT Brown SDFT

Model DDFT Brown DDFT
Fig 20.6  Moment arms from the forelimb model compared with those
published by Brown et al. 2003. Moment arms for the forelimb model were
calculated using a simulated tendon excursion method.
Reprinted from Merritt J.S., Davies H.M.S., Burvill C., Pandy M.G., 2008. Influence of
muscle-tendon wrapping on calculations of joint reaction forces in the equine distal
forelimb, Journal of Biomedicine and Biotechnology, Hindawai Publishing Corporation, with
kind permission from Jonathan Merritt.

For example, by creating a computer model of a particular bone, Fig 20.7  FE models, presentation of results. The 5° raised heels (upper
and applying forces to the bone from data known a priori, the type diagram), and 5° lowered heels (lower diagram) shown in lateral view:
and location of fractures in that bone may be predicted. displacement is shown as deformed geometry, colours represent von Mises
A number of equine studies have been performed using finite stresses with red representing maximum stress and blue representing
element analysis. One of the earliest was an analysis of the structure minimum stress. Red arrows indicate the maximum stress zone in the
of the hoof. The intent of the project was to examine the role of proximal dorsal wall; blue arrows indicate the maximum displacement of
mechanical factors in lameness and degenerative diseases within the the dorsal wall. Comparing the 5° raised heels model with the 5° lowered
hoof capsule (Hogan et al., 1991). A two-dimensional finite element heels model, less stress and smaller deformations are found in the 5° raised
analysis was performed, and the deformation results were examined heels model.
to determine motions that were not visible through direct observa- Reprinted with permission from Hinterhofer, C.H., Stanek, C.H., Haider, H. The effect of flat
tion. The results indicated that during weight-bearing the coffin bone horseshoes, raised heels and lowered heels on the biomechanics of the equine hoof assessed
moved downward toward the sole, and rotated away from the dorsal by finite element analysis (FEA) in Transboundary and Emerging Diseases 2001 (2), 73–82.
hoof wall (Fig. 20.7). Additionally, approximations of the stresses Copyright © 2001, John Wiley and Sons.
induced in the laminar tissue during loading were determined. The
stress values within the hoof derived from this and similar studies (Hinterhofer et al., 2000) (Fig. 20.7). Maximal displacement was
have given a much more comprehensive understanding of the behav- calculated in the hoof capsule shod with a regular horseshoe
ior of the hoof. The results were of particular interest because, at that without a clip. Minimal displacement was found in the capsule with
time, there was little direct information describing the motion of the a toe clip and 2 side clips placed behind the 3rd nail. All models
coffin bone with respect to the hoof wall during loading. showed higher displacements when calculated with a loose nail
Due to improvements in computing capabilities and finite fixation (Hinterhofer et al., 2001).
element programs, Les et al. (1997) were able to develop a more A more complex, thirty-two component finite element model of
comprehensive three-dimensional model (Fig. 20.8) than that of horse and donkey digits with over 106 elements was used to compare
Hogan et al. (1991). Using CT data, an accurate model of the meta- von Mises stress levels in the deep and superficial digital flexor
carpus was created in the computer, and it was then subdivided into tendons in the horse (respectively, 1.34 MPa and 0.56 MPa) and
a series of hexahedral elements. This model was compared to ex vivo the donkeys (respectively, 0.78 MPa and 0.27 MPa). The same
loading results, and the accuracy of the FEA was validated. This model was used to evaluate the effects of weight-bearing on capsular
study was an excellent example of the opportunities opened up by deformation patterns (Collins et al., 2009).
FEM. Once a particular model has been validated, many different Pollock et al. (2008a,b) defined a mathematical model to predict
loading conditions can be simulated and analyzed within the com- strains in the humerus at stance within ± 2 standard deviations of
puter and an accurate representation of the stress response can experimental strains at four of these locations and predicted negli-
be determined without the need to use additional live subjects. gible strains at the remaining two locations, which is consistent
Hinterhofer et al. (2000, 2001) studied the use of finite element with experimental findings.
analysis in evaluating the effect of trimming and shoeing on hoof
mechanics in the horse. Stresses are high in the material surround- Development of a computer model:
ing the quarter nails, in the heels and in the proximal dorsal wall.
Raising the heels by 5° resulted in significantly (p < 0.05) lower
a horse history
stress and displacement values. The model with heels lowered by Although the end result of computer models can vary greatly
5° yielded the highest stress and displacement values, and the FE between applications, the steps involved in the process are quite
model with the regular horseshoe lay between the two extremes similar. Visualization requires the greatest accuracy in the overall

472
 Development of a computer model

Proximal

Medial

Palmar
A

Fig 20.8  Finite element model of the equine metacarpus.

Reprinted from Les, C.M., Keyak, J.H., Stover, S.M., Taylor, K.T., 1997. Development and
validation of a series of three-dimensional finite element models of the equine metacarpus.
Journal of Biomechanics 30 (7), 737–742, with permission from Elsevier.

form and appearance of structures. When attempting to simulate a

surgical procedure, it is critical that the anatomy appear as realistic
as possible. The model must accurately predict what the surgeon
B
sees during the procedure if it is to have any practical value. In order
to reproduce anatomical structures accurately, it is necessary to use
very advanced medical imaging techniques. When studying skeletal Fig 20.9  A three-dimensional reconstruction of the first lumbar vertebra
structures, CT scans can be used to build a representation of struc- from CT images. (A) Spline curves of successive CT slices were lined up to
tures with high X-ray absorbance (Fig. 20.9). Soft tissue structures generate the surface contour of the bone. (B) The bone surface was then
are typically visualized more clearly using MRI images. Regardless generated from the spline curves and filled with bone.
of the source of data, the process of recreating the region in question Reprinted from Groesel, M., Gfoehler, M., Peham, C., 2009. Alternative solution of virtual
is the same. The resulting images are stored as a series of pictures biomodeling based on CT-scans. Journal of Biomechanics 12 (4), 2006–2009, with
that are then analyzed to locate the edge of the structure of interest. permission from Elsevier.
These interface locations are stored in the computer and are used
to define the edges. Graphics tools are subsequently used to rebuild
the edges into a solid structure. By this method, almost any complex maintaining an isometric virtual ligament and maintaining contact
anatomical structure can be recreated for a visual examination. between the appropriate articular surfaces, creating a variable
Zarucco et al. (2006) developed a three-dimensional generic mus- moment arm for the tendons. The simulation of the proximal sesa-
culoskeletal model of the equine forelimb comprised of bony moid bones was compared to movement recorded in vitro. The
segment, muscle-tendon, and ligament information, and based on paths and origins used for the deep digital flexor tendon (DDFT),
high-resolution computed tomographic (CT) and T1-weighted superficial digital flexor tendon (SDFT) and suspensory ligament
magnetic resonance (MR) images from an isolated forelimb of a (SL) were altered and the effects on their calculated strains during
Thoroughbred racehorse. Image fusion was achieved through coreg- trot stance were examined. It was shown that within the anatomi-
istration of CT and MR images with an image analysis program cally realistic spectrum, changes to tendon paths can have an appre-
(analyze) by adjustment of the relative position and orientation of ciable effect on calculated strains; however the origin sites chosen
fiducial markers visible in both modalities until the mutual infor- are not as influential as changes to paths at the metacarpo-phalangeal
mation between the images was maximized. Three-dimensional joint.
surfaces of the bones and origin/insertion sites, centroid paths and These techniques are also useful in kinematic and dynamic mod-
volumes of the muscle-tendon and ligamentous structures were eling. Regardless of how it may appear, joint motion is never as
obtained from the multimodal (CT/MR) images using semiauto- simple as a two-dimensional rotation. Using the techniques speci-
mated and manual segmentation combined with sagittal and trans- fied above, a more comprehensive representation can be constructed
verse color-cryosection anatomic images obtained from three other which will allow for a much more accurate study of the dynamics
cadaveric equine forelimbs. Once bony and soft-tissue structures of joint motion, and hence a better understanding of pathologies.
were reconstructed in the same coordinate system, data were Finite element models perhaps require the most comprehensive
imported to a software package for interactive musculoskeletal and accurate descriptions of skeletal and accessory structures and
modeling (SIMM). their properties. The accuracy of the analysis is directly dependent
Lawson et al. (2007) created a subject-specific distal forelimb upon how closely the model approximates the original geometry.
model using bones extracted from CT scans to examine movement Although this analysis technique is inherently an approximation
from in-vivo invasive-marker motion capture. The movements of due to its very nature of subdividing curved surfaces into linear
the sesamoid bones were simulated using the constraints of segments, without an accurate framework on which to build the

473
 20 Modeling, simulation and animation

structure of the finite element model, the analysis will be inherently

flawed. The model itself is only as good as the information from Q
which it was created, and without a great deal of precision in the
original model, the results will be useless.
If a kinematic or dynamic model is being developed, another
major step is necessary to the completion of the simulation. Exter-
nal data, such as kinematic information describing the motion of φ
the horse and GRFs, are necessary for a comprehensive model. There
are, however, two ways in which these data can be used. The first is
to apply the information as the perturbation input to the model.
By specifying the kinematic and force input, data such as joint LS
torques, moments and temporal responses may be analyzed. The
second method is to use the external data to verify a solution arrived
at by other means. In the full body equine simulation of Van den
Bogert (1989), an intelligent system was used to describe the
motion of the segments of the horse. A standard muscular model
controlled some joints with activation timing based on the percent
of stride, while others were controlled by constraints based on Fig 20.10  Typical example of a human knee joint flexion simulation. ϕ, the
joint angle; LS, the leg segment length; Q, the quadriceps muscle force.
approximate joint angles. Since the force data were independent of
the solution, they could be used to confirm the veracity of the solu-
tion. With a model of the equine hind limb, the function of the
‘passive stay apparatus’ could be explained. A model of the forelimb allow full experimental control and access to all mechanical vari-
was used to determine force distributions in the digital flexor ables for measurement. In the case of human or equine movement,
tendons, and to experiment with various methods to change the one possible option currently available is to build a model of the
force distribution. Simulations of this type are valuable tools in system under investigation, and then perform experiments on the
basic research on functional anatomy and etiology of injuries. model. Although physical and mechanical models have been used,
An application with much potential is the optimization of sports computer models (also numerical models) are almost exclusively
performances using simulation methods (Van den Bogert & used for this purpose.
Schamhardt, 1993). Van den Bogert & Nigg (2007) defined simulation as follows:
simulation is the process of performing experiments on a numerical
model. The term experiment implies that it is crucially important
Development of a computer simulation: that the model reflects the causality of the system that is investi-
a human perspective gated. The input of the model should precede the output in a
causative sense. Once such a numerical model is developed, varia-
tions on the original experiments can be performed quickly and at
Introduction low cost. It should, however, be kept in mind that the experiment
Research on equine movement does not always use horses and the is always performed on the model only, and conclusions may not
same accounts for humans. In most areas of biomedical research, apply to the human system when the model is not valid, and the
animal models are used when rigorous control of experimental model is only valid within the scope of its original design and vali-
conditions is required, measurements require invasive techniques, dation. This is a central problem in modeling and simulation. It is
or when the use of human or equine subjects is not ethically accept- important that a model replicates those features of the system that
able (Van den Bogert et al., 1989; Van den Bogert & Schamhardt, are essential for the question that is studied. This requires careful
1993). design of the model and appropriate validation tests once the
In the context of biomechanics of human movement, animal model is finished. In biomechanics, simulation is applied in four
models are typically used to study basic biological mechanisms. A different areas: dynamics of movement, tissue mechanics, fluid flow,
good example is research on muscle mechanics, where full experi- and measuring techniques. In dynamics of movement, simulations
mental control over muscle length and activation is required, and allow the experimenter to apply controlled perturbations to the
direct measurements of variables such as muscle force, muscle fiber anatomy, muscle coordination, and control system, and observe the
length, and even heat production can be done (Fig. 20.10). This resulting changes in movement and forces.
approach would be possible in animal models and not in human Tissue mechanics can be studied using finite element models
subjects. There is, however, a strong pressure on researchers and (FEM), which are the numerical equivalent of partial differential
their institutes to make use of simulation techniques instead from equations (PDE), which describe the local relationships between
an animal welfare perspective and thus reduce the use and number stress and strain in tissue. These methods have been applied to
of testing animals. bone, cartilage, tendon, ligament, muscle, and even (animal) brain
For more applied questions, related to specific human injuries or tissue during impact.
specific human movement tasks, animal models are not appropriate Biomechanical measuring techniques often suffer from errors,
because the dynamics and anatomy are not comparable. For and simulation may be used to determine how errors in the raw
instance, the study of knee injuries in distance runners has a basic data are propagated to the final result of the calculation.
biological component and a mechanical component. Basic biology,
or in this case the reaction of cartilage to mechanical loads, can be Designing musculoskeletal models
studied in animal models, e.g. in the horse. It is, however, obviously
impossible to use animal models to learn how to reduce forces in Movement simulations can be driven using a variety of methods,
the knee joint with appropriate orthotics or other mechanical inter- summarized below from Van den Bogert & Nigg (2007). In simula-
ventions. Furthermore, in human subjects, the force in the joint tions of human movement during free fall or space flight, move-
cannot be measured; other factors contributing to injury cannot be ment of the body can be produced by well-coordinated limb
controlled, and ethical problems arise in the study of more acute, movements (Passerello & Huston, 1971). Since the required muscle
serious injuries. Analogous to the use of animal models, it is desir- forces are extremely small, it is appropriate to consider joint rota-
able to study such movement-related questions with methods that tions as the cause of the movement (Yeadon, 1993). The required

474
 Development of a computer simulation: a human perspective

as 100% higher than direct measurements using an instrumented

CE SEE prosthesis (Brand et al., 1994).
The contribution of a muscle force to a generalized force can be
determined using the principle of virtual work. When the system
undergoes a hypothetical small change in the generalized coordi-
PEE nates, the muscle will undergo a virtual length change. The principle
of virtual work states that the work done by the muscle should be
LCE equal to the work done by the generalized forces, i.e. the dot product
of the generalized force vector and the generalized displacement.
LM If a generalized coordinate is a joint angle, the corresponding
generalized force will be the moment of the muscles at that joint.
Fig 20.11  The three-component Hill muscle model. CE, contractile If the generalized coordinates are segment orientations, the muscle
element; PEE, parallel elastic element; SEE, series elastic element. moment will contribute to the generalized forces for both segments.
The partial derivatives, which are the moment arms, may be derived
from an anatomical model, for instance, using straight lines, or
muscle forces may be computed, but are not relevant to the problem measured directly in cadaver specimens or in vivo using three-
as long as they do not exceed the force-generating capacity of the dimensional imaging techniques. Moment arms can be simply mea-
muscles. sured as the distance from the muscle’s line of action to the joint
When muscle forces are included in a simulation of movement, center, requiring assuming a fixed joint center. A more elegant
it is usually desirable to include several well-known mechanical method is to collect data of muscle length at various joint angles
properties of muscle (Hill, 1938): the force–length relationship, the (or other generalized coordinates), and use regression analysis to
force–velocity relationship, and the activation dynamics. A muscle find a mathematical relationship describing it as a function of the
model, which includes these properties, will allow the muscle forces generalized coordinates. Partial derivatives may then be computed
to respond realistically to changes in length and neural stimulation. analytically. This method was first used by Grieve et al. (1978) to
Muscle models exist in various levels of complexity, but for move- determine moment arms of the ankle plantarflexors, and applied
ment simulation, the three-component Hill model (Fig. 20.11) has later to determine moment arms for hip and ankle muscles (Spoor
been used almost exclusively (Winters, 1990; Zajac, 1989). The et al., 1990; Visser et al., 1990).
model consists of a contractile element (CE) and two non-linear The coupling between skeleton and muscles requires computa-
elastic elements, the parallel elastic element (PEE) and the series tion of muscle lengths from the bone positions in the skeleton,
elastic element (SEE). The CE represents the muscle fibers, the SEE computation of muscle force from the state equations of the muscles
represents tendon and other elastic tissue in series with the fibers, and the application of the muscle forces to the skeleton. For a
and the PEE represents the passive properties of the fibers and complete musculo-skeletal model, passive forces should also be
elastic tissue surrounding the muscle fibers. The elastic elements are considered. Ligaments, when their function is to guide joint kine-
described by a mathematical force–length relationship, which has matics, need not be included, since their function has been incor-
the property that no force is generated below a certain length (the porated in the joint model. For instance, the cruciate ligaments in
slack length), and stiffness increases as force increases. The stiffness the knee can be assumed to be inextensible during normal move-
parameter can be scaled such that the tendon strain at maximal ments, resulting in a moving center of rotation at the intersection
muscle force is about 3–5% or about half the level where damage of the cruciates. These guiding ligaments do not perform mechani-
occurs (Winters, 1990). It should be kept in mind that the three- cal work, and can be omitted from the equations of motion. They
component Hill model is based on a limited set of experimental will, however, have an effect on the articular contact force. To model
observations, such as isometric, isotonic, and isokinetic contrac- this, joints can no longer be considered to be kinematic mecha-
tions, usually performed at maximal activation. Interactions nisms and deformation needs to be modeled (Blankevoort &
between the effects of length change and activation may, therefore, Huiskes, 1996). Internal passive force also occurs at the end of the
be poorly predicted. range of motion of a joint. This may result from stretching liga-
The Hill model, in spite of its limitations, is presently considered ments, compressing muscles, or a combination of both.
the most practical muscle model for simulation of human move- The net effect on the movement can be modeled as an extra joint
ment. When muscles are connected to a skeleton with many degrees moment, which is a function of one or more joint angles (Riener
of freedom, the muscle model needs to be coupled to the equations & Edrich, 1999). Passive tension in ligaments spanning several
of motion, resulting in a combined system of equations represent- joints, which occurs in some animals (Van Ingen Schenau &
ing the musculo-skeletal system (Van den Bogert et al., 1998). Bobbert, 1993), can also be modeled as a simple passive muscle
The output of a muscle model is the magnitude of muscle force, model with only a SEE and coupled to the skeleton in the same way
a scalar. The equations of motion require a force vector (magnitude as muscles. Interactions between skeletal dynamics and muscle
and direction) and also a point of application, so that the moment dynamics can only be observed in a model in which both are rep-
can be computed. Also, the muscle model requires the instanta- resented and coupled.
neous length of the muscle as input. All of these require a model External forces are forces acting between the system of interest
of the musculoskeletal anatomy. One method for doing this is to and the environment. Common external forces are: gravity, contact
model the muscle as a straight line of action between origin and forces, and aerodynamic forces. Gravity is simply modeled as a
insertion. If the kinematic state variables are known, the position constant force vector of magnitude (mg) and downward direction,
vectors of origin, and insertion, can be calculated from the position applied to the center of mass of each rigid body segment in the
and orientation of the two body segments. The same vector, but in model, where m is the mass of the body segment and g is the accel-
opposite direction, is applied to the origin point. Straight line eration of gravity (9.81 m/s2). Modeling the contact as a kinematic
muscle models have limitations, and usually require additional connection like a joint can often include contact forces. This was,
wrapping points to model muscles and tendons that follow a curved for example, appropriate for the foot-ground contact forces in a
path over underlying tissue (Delp & Loan, 1995). Failure to prop- simulation of vertical jumping (Pandy & Zajac, 1991), provided that
erly model the path of the muscle may result in underestimation of the simulation is terminated when tensile forces in the connection
moment arms and hence, overestimation of muscle forces required begin to occur.
for movement. This was a problem in early estimates of articular When simulation is used as a scientific tool, experiments are
contact forces in the human hip. Model calculations were as much performed on a numerical model. Direct validation of a numerical

475
 20 Modeling, simulation and animation

model is usually difficult because the same experiments cannot be Statistical analysis showed no significant effect for the population
done on human subjects. This could be because the experiments as a whole, confirming earlier results on human subjects. Thus,
are related to severe injury, because human subjects are not statistical analysis prevented incorrect generalizations, which could
sufficiently reproducible, because humans are too fatigable or have been made when just one model had been used. In this case,
because the outcome is a variable, which cannot be measured. the model was sensitive to movement style. Sensitivities to other
Most invasive research in horses is also precluded by modern variations within the human population, such as anatomical varia-
ethical standards. tions, may be detected or eliminated similarly, by creating a popula-
How then do we ensure the validity of scientific studies using tion of models with the appropriate range of parameter values.
computer models? Only indirect methods are available. For this Note, however, that modeling and simulation are sometimes used
reason, the term validation may be too strong and evaluation specifically to determine the influence of inter-subject variations,
should be used instead. First of all, a model should be consistent and in such case the statistical approach is not appropriate. In
with observations that can be made on humans. When optimiza- summary, simulation experiments only tell the truth about the
tion of performance is carried out, and a realistic movement is model that was used. Generalization to the human population is
obtained, the model is generally considered to be valid because no always hazardous and requires extensive validation and careful
movement data were used to develop the simulation. When solving examination of the results.
the tracking problem, it is expected that, after optimization, all
variables are within two standard deviations of the mean. If this is Practical applications
not the case, the model is unable to perform the movement in
a sufficiently realistic manner. This should be reason to closely The field of biomechanical movement simulation has matured suf-
examine the model and make improvements where necessary. ficiently to allow its use in answering certain basic and applied
Passing this test, however, does not guarantee a valid model. Due questions on human movement, but for the equine species this
to the redundancy of the locomotor system, the model could have development is still in its infancy, despite the work of Van den
found a different solution than the human to achieve the same Bogert et al. (1989) and Van den Bogert & Schamhardt (1993).
external movement and force variables. In that case, additional The best-developed areas of application are gait and sports injuries.
predictions must be elicited from the model, which can then be In gait, the functional role of muscles has been identified by solving
compared to measurements that were not used for development of the tracking problem in a forward dynamic model, followed by an
the model. This is especially important when the tracking problem induced acceleration analysis (Neptune et al., 2001). Minimum
is solved. We recommend testing the response of the model to energy optimizations have produced realistic movements, suggest-
controlled interventions and compare that response to results of the ing that minimal energy is the governing principle of human gait
same experiments on humans. Even when the final application is a (Anderson & Pandy, 2001). In sports medicine, simulation has
study on severe injuries, it is often still possible to evaluate the shown how the effect of foot orthoses on knee joint mechanics can
model dynamics using non-destructive perturbation tests. Care differ between subjects (Neptune et al., 2000). Simulation has
must be taken that these experiments test those aspects of the model perhaps its greatest impact in the area of acute injuries, where no
that are important for the final application. For development of a human experimentation is possible, although this would open per-
model for knee ligament injuries, perturbations of initial conditions spectives for the use of the equine species being a model for these
were used to evaluate the validity of the model (McLean et al., human injuries. Nonetheless, there is a long history of increasingly
2003). Finally, a model should not be overly sensitive to errors in realistic passive human movement simulations in vehicle collisions.
model parameters. Critical model parameters can be identified by With active muscle models and optimization in realistic musculo-
sensitivity analysis: each parameter is adjusted by a small amount skeletal models, these techniques are now becoming feasible for
and the change in the results of the simulation is examined. In some studies on knee and ankle ligament injuries during sports (Wright
cases, this will show that certain model parameters are too critical et al., 2000; McLean et al., 2004). Although the basic methodolo-
and the results of a simulation study would depend entirely on a gies are quite straightforward, the complexity of modeling required
random error in such a parameter. In certain cases, optimization for these applications is still beyond the capabilities of most labo-
methods are helpful. In a quasi-static model of the knee joint, it ratories. Van den Bogert & Nigg (2007) expect, however, that com-
was found that the behavior was sensitive to the lengths of the liga- mercial and user-friendly software will become available to make
ments (Blankevoort & Huiskes, 1996). Solving the tracking problem the technology more accessible in the near future for human appli-
then eliminated these unknown parameters. Ligament lengths, cations and eventually on the long run for use in the horse; in
which minimized the difference between simulated and measured the equine species animation seems a more likely application in the
movements, were found and these values were used for subsequent short term.
applications of the model.
Another powerful safeguard against overly sensitive models is
statistical analysis. Experiments on humans or animals are never Development of computer animation:
performed on a single individual because one individual may not
be representative of the population. Statistical analysis is performed an equine perspective
to ensure valid generalizations. When using complex musculo-
skeletal models, the same principle should apply. These models
have many degrees of freedom, many natural frequencies, and are
Introduction
often unstable and chaotic. Results from a single model could well The movement of horses has always been an object of intense inter-
be completely irrelevant. est, not only in modeling and simulation (Van den Bogert et al.,
In a simulation study on the effect of shoe hardness on impact 1989; Van den Bogert & Schamhardt, 1993), but also in animation.
forces in running, both positive and negative responses were found In the Scientific American of 1861 was a report on a horse race on
in a group of models (Wright et al., 1998). By examining the model, stage, in which horses were galloping on a treadmill and a synchro-
this could be attributed to two mechanisms, which worked in nously rolling curtain behind the horses was depicting the scenery
opposite directions. Impact forces tend to increase initially with from a local racetrack (Fig. 20.12). The early Muybridge experi-
harder material. This then increases the rate of knee flexion, result- ments in trying to capture the movement of horses (and other
ing in a better shock absorption by the body. In subjects with animals) on film, started over a debate about the question whether
a certain movement style, the latter mechanism resulted in trotting horses have an airborne moment where all four hooves are
overcompensation. off the ground or not (Muybridge, 1899). To prove this point

476
 Development of computer animation: an equine perspective

Fig 20.12  Treadmill for animation: a horse race on stage showing (A) the foreground display and (B) the mechanical background. Scientific American (1861).

Muybridge began his experiments with line triggered cameras. disciplines interact and should work together as there are many
His photographs were the first that definitely showed all phases in cross links between animation for film and animation in the inter-
the horse’s gaits. But the interest in understanding the movement est of equine medicine.
of quadrupeds has not become any less intense over the later
years (Van de Panne, 2000). Many films and other forms of media
entertainment include the movement of all kinds of (virtual) crea-
tures, either in live-motion or in other forms of special effects. These
History
include animated animals, either drawn by hand or by computer, As stated above, there has been a large background in the study of
and other CGI (computer generated imagery) effects (Hahn, 2008). locomotion among scientists. It has only been in the last 120 years
Many older and more modern films have featured animated and or so that certain advanced film tools became available to interested
life-like movements of horses and horse-like creatures, such as scientists to capture the motion of live animals. As mentioned
Spirit: Stallion of the Cimarron (Dreamworks Animation SKG, Glen- before, one of the earliest scientists was Eadweard Muybridge
dale CA, USA) and Avatar (20th Century Fox, Los Angeles, USA). (Kingston upon Thames, UK: 1830–1904). He is considered one of
On the other hand, veterinary medicine is also becoming more and the pioneers in the study of animal movement. Muybridge wanted
more interested in understanding, capturing and simulating equine to analyze the specific footfall sequences of horses in different gaits.
movements. Many different devices have been developed to capture For this endeavor he used a series of twelve mounted photographic
equine motion and visualize it in a way that makes in-depth analy- cameras. These cameras were then triggered one at a time as the
sis of equine gaits possible. Thus, subtle lameness and asymmetrical horse passed them at a walk, trot or canter, thus taking 24 photo-
use of limbs can be detected and treated earlier or more severe graphs within the short period of time it took for the horse to pass.
injury may even be prevented. This is an area where many Muybridge gained fame with his recording of equine movement on

477
 20 Modeling, simulation and animation

photographic film, thus proving horses do have moments in the and features such as Betty Boop and Snow White (Walt E. Disney:
trot and canter strides when all 4 feet are off the ground. He is often 1901 (Chicago)–1966 (Los Angeles)). More recently a computer
credited as being a pioneer in the science of biomechanics. Although replaced the rotoscope, but the principle remained unchanged and
he started his experiments with an extensive focus on equine was a great help for animators in being able to capture difficult and
motion, his studies have included almost all animals and any move- life-like movements, such as dancing. In a way, rotoscoping can be
ment imaginable. Muybridge was also among the early inventors of seen as a component of a larger technique: ‘motion capture’. Already
precursors of film movement, with his zoöpraxiscope. His photo- in classical animation there was a demand for perfecting the move-
graphs of animal movement and analysis of animal gaits are used ment of both animals and humans in drawings, and also a great
even today as a reference for animators who try to draw correct and need for a proper understanding of the placement of limbs, the
life-like animal motions. correct footfall and angles of the joints. Muybridge’s photo series
Animation is also based on the photographic principle of film. were helpful, but not sufficient reference material. Animators need
To the human eye, projected images that follow each other very fast a lot of data for reference of locomotion and other actions. This is
(24 per second), seem to become connected and are perceived as a where the motion capture technique appeared on the animation
single, moving image. Almost everyone will have had some experi- scene. It is basically a matter of recording movement in live-action
ence with this in their youth through flipping books where images on film, then translating these movements into a model. This is
drawn on successive pages of a book are flipped through at a certain often done by attaching certain markers (such as reflecting dots or
speed at which the images seem to come alive and ‘move’. The same strips) to certain moving points on the subject’s body and then
principle applies to an animated film: different images are drawn translating the motions of these markers into a model that can be
and captured on a film frame, with slightly different poses on each used for quantitative gait analysis or used to create an animated
drawing. This means that for every second of animated film, 24 version of the subject. These recorded movements can be animated
drawings will have to be made. This traditional way of animation characters (Gleicher, 1999).
is called ‘classical or key frame animation’ (Thornton, 2004). For
every frame a pencil or ink drawing is made and then inked and Modern animation
captured on one frame of film. When all these separate frames/
drawings are projected at a speed of 24 frames per second, they Besides ‘key frame animation’ there is also the possibility of ‘com-
become a moving animated film. In these early animation films, puter animation’. This form is based upon the same principle as
there was already much interest in trying to capture the movement classical animation, but instead of hand-drawing the image on cells,
of animals as realistically as possible and different techniques were the images are now created in the computer. The advantage of this
developed for recreating life-like animal movements. In the old days is that the computer, being programmed with certain algorithms,
this implied first filming a desired motion, such as walking or trot- can ‘fill in’ missing frames that are between certain key-frames. Key
ting, and then projecting it and capturing this motion in drawings frames are also used in classical animation. They show images of
by slowing down the film. One of the first developed techniques certain key positions of the animated object but movement based
for helping this process along was called ‘rotoscoping’. This process only on key frames is jerky so the moments between keyframes need
consists of first filming the action played out by real life actors or to be filled with images showing intermediate positions between
animals, projecting the filmed images onto a frosted glass pane and the key frame moments. These images are aptly called ‘in-betweens’.
tracing over them onto paper, frame by frame. The projector used One of the advantages of computer animation is that it cuts down
for the projection of the still images was a rotoscope (Gleicher, the time needed to draw the in-betweens. Computers can also be
1999). used for motion capture techniques.
The rotoscope was invented in 1915 by animator Max Fleischer
(Fig. 20.13), who developed it for his animation series Out of the
Inkwell. It has been used extensively and in many animation films Motion capture
Nowadays the basis for motion capture is still the same as in the
early days of classical animation, but a variety of technologies are
available for tracking and recording the real-time motion (Menasche,
1999) using multiple markers placed onto animal models or using
markerless techniques. The resulting data can be used to generate
the movement of a computer-animated model. As Gleicher (1999)
puts it aptly ‘Motion capture creates a representation that distills
the motion from the appearance. It encodes the motion in a form
that is suitable for the kinds of processing or analysis that we need
to perform’. The technique has developed further and further, to a
point where it is now even possible to capture and recreate very
subtle and life-like facial expressions with motion capture tech-
niques. This is what was used in films like Avatar, King Kong and
Lord of the Rings.
There are many advantages to motion capture as compared to
classical animation: the whole process of capturing movement is
more rapid, sometimes even using real-time results. This greatly
reduces the cost of animating complex and long movements and
allows a large amount of data to be collected in a single session. It
facilitates understanding and analyzing complex actions and realis-
tic physical interactions such as secondary motions, weight and
exchange of forces.
Many systems are available for motion capture. There is a distinc-
tion between real-time, on-line systems where the animation is
produced instantly, sometimes also referred to as performance ani-
Fig 20.13  The rotoscope was invented in 1915 by animator Max Fleischer. mation, and systems that are not real time (Gleicher, 1999). Real-
From Cabarga L (1988). The Fleischer Story. Revised Edition. DaCapo Press: New York. time motion capture may have applications in veterinary diagnostics,

478
 Development of computer animation: an equine perspective

but it needs to be further developed to become more user-friendly

and less expensive. For animation purposes, the non-real time
motion capture method is more often used as it allows more pos-
sibilities for animating production purposes. Many different types
of markers (acoustic, inertial, LED, magnetic or reflective) or com-
binations of any of these can be used to track movements of the
points of interest. The same techniques are used in equine motion
analysis.
There is also a dichotomy between optical and non-optical
systems. Optical systems use data captured from image sensors to
triangulate the three-dimensional position of a marker between one
or more cameras with over-lapping fields of view. Non-optical
systems do not use external cameras or markers that are ‘seen’ and
then translated into data, instead the data are captured directly
without an optical data receiver in between. They are based upon
several different techniques (Menasche, 1999) and have recently
become more popular. Optical and non-optical systems can be
further divided into passive and active systems (Mundermann
et al., 2006).
Fig 20.14  A typical example of a so-called ‘exo skeleton’ motion capture
technique.
Optical systems From www.solomonia.com.

Passive markers systems: passive optical systems use highly reflective

markers that reflect light back toward the camera’s lens. The camera
can be adjusted so only the bright reflective markers will be recorded, restrictions to this technique when applied to animal movement
ignoring other materials such as fabric. Passive systems do not (Fig. 20.14).
require the subject to wear wires or electronic equipment. The Magnetic systems use transmitters that establish magnetic fields
markers are usually attached directly to the skin. within a space, and then use sensors that can determine their
Active optical systems do not use reflective markers but instead emit position and orientation within the space based on these fields
light information in the form of laser light, light patterns or light (Gleicher, 1999).
pulses, often using small LED illuminators. The system can illumi- For entertainment purposes there are some downsides to the use
nate one LED at a time in quick succession or multiple LEDs. These of motion capture. There is the problem of attaching markers to a
illuminated LEDs can then be used to triangulate movements. hairy, moving and sometimes even wild animal. These markers have
Markerless systems. When using markerless systems, the subject to be small enough to not influence the movement of the joints
does not need to wear markers or a special suit or skeletal system. that they are attached to, but sturdy and noticeable enough to be
Special computer algorithms are created to analyze multiple ‘seen’ by cameras and other capturing systems. As with the use of
streams of optical input and identify specific forms, making it pos- motion capture and analysis for other purposes, artifacts due to skin
sible to track them. The data can then be processed using different displacement (Van Weeren et al., 1990) or neurosensory stimula-
techniques, for example the motion of an animal can be extracted tion (Mundermann et al., 2006; Clayton et al., 2008) can yield
from a video and then be controlled by using a system called active motion patterns that do not reflect natural patterns of movement.
contouring (Skrba et al., 2008). The model is scaled to match the Another challenge is to persuade the animal to perform the desired
form of the animal in the video and it is aligned with the video. movement in a restricted space that the cameras or other capturing
The contour of this video model is then anchored to the horse systems are able to cover. Dogs and horses can be trained to perform
model. This means that the model changes the position of its specific required movements and can also be taught to walk and
limbs when the horse in the video does this too. Another method run on treadmills. Another limitation is that real animals can only
that uses markerless, optical techniques is statistical analysis of perform movements that follow the laws of physics, which may not
video images (Skrba et al., 2008). In this method, live video be sufficient for entertainment purposes. However, motion capture
footage of a moving animal is segmented into binary images in techniques can be used as a basis and supplemented by computer
order to isolate the foreground subject from the background. It can animation and CGI to achieve what is physically impossible for a
be applied to the recognition of animal movement within wildlife live animal. There is also the disadvantage of cost: although the
(Skrba et al., 2008). This technique and other similar ones still actual time needed to animate a movement can be greatly reduced,
have their limitations in creating a life-like motion capture model it is necessary to buy or develop specific hardware and software that
since the data are not always suitable for use in the simulation of are required to obtain and process the data. This obviously requires
three-dimensional animal characters. Nonetheless Wilhelms & van financial investment. Finally, traditional animation techniques,
Gelder (2003) already indicated that automatic tracking with active such as secondary motion or manipulating the shape of the char-
contours could be used successfully for moderately slow move- acter, as with squash and stretch animation techniques, must be
ments and where the background is relatively simple. This is rela- added later.
tively easily arranged in the controlled circumstances of a clinical New techniques are constantly being developed that can be
setting. applied to film, animation, and games. It is anticipated that these
advances will facilitate the application of motion capture tech-
niques for veterinary uses by eliminating the need for a treadmill,
Non-optical systems the attachment of multiple markers and very controlled locomotion
Mechanical motion capture systems are often referred to as exo-skeleton circumstances for useful analysis of gaits (Fig. 20.15).
motion capture systems, due to the way the sensors are attached to There are many great ways in which the animation and motion
the body. Instead of small markers, a skeletal-like structure is capture techniques that are being developed for the entertainment
attached to the subject’s body. When the subject moves their body industry can be applied for veterinary diagnostic purposes and vice
and joints, the exo-skeleton moves simultaneously, mirroring the versa. A great example of this interdisciplinary approach to equine
placements of limbs and angles of joints. There are obviously some movement is the use of a live-action horse for the creation of a mix

479
 20 Modeling, simulation and animation

of CGI adapted live-action movement and a digitally animated learn and benefit from each other. Seeing how an animated horse
horse model in a feature film. The way the entertainment industry is created from the movements of a real one, and how this animated
and the veterinary world can work together in a mutually beneficial model can be used to perform actions that the original horse never
way is shown in the example of the film SINT, a Dutch horror film did, may help us to see new approaches to clinical gait analysis and
made in 2010 (http://www.sintdefilm.com) that uses footage of a lameness examination.
live-action galloping horse to create a spectacular chase sequence The scenes were storyboarded well ahead of shooting and were
over the roofs of Amsterdam. It is interesting to briefly discuss how used as a basis to guide the horse selection and training. A suitable
this process was created as it vividly demonstrates the intersection horse was selected on the basis of being easy going and willing to
of science and entertainment and the way these two disciplines can keep up a steady rhythm in each gait while being ridden on the
treadmill. For the recording session, the horse was positioned on
the treadmill and the rider was attached to a safety harness. After
some trial runs, the horse and rider were filmed at different gaits
(walk, trot and canter) and from many different angles (Fig. 20.16).
The treadmill area was shielded in green covers to provide an even
green background to facilitate replacing the background with the
rooftops of Amsterdam later on. Additional horses were filmed in
a large film studio, running past a 50 × 10 m green screen, or on
set pieces made to look like rooftops.
The horse and rider and their movements were then cut out of
the (green) environment around them. This was done frame by
frame and created a cut-out figure of horse and rider without any
background. The next step involved matching the camera angle of
the horse footage to footage of the rooftops of Amsterdam. This
involved re-calibrating all the frames using three-dimensional
camera tracking software, resulting in a new virtual camera. The next
stage involved creating a whole new, CGI background involving
rooftops and snow. This CGI background was then filmed with the
virtual camera created earlier. The moving, cut-out horse and rider
were inserted into this newly created background, then moving ele-
ments were added like hoof prints, snow reacting to the horse and
falling snow. Finally, color correction was applied to the whole
sequence. The final result shows a spectacular ride of the horse and
rider galloping over snowy rooftops in Amsterdam, while being
chased by police cars on the streets below (Fig. 20.17).
In addition to footage of the live-action horse and rider in this
chase sequence, a small part of the scene consists of a computer
animated horse model, instead of the live action horse. This is the
Fig 20.15  Dr. Hilary Clayton rides her horse MSU Magic J, who is wearing a part where the horse loses its footing and slides down a diagonal
blanket covered with predefined, anatomical markers, in front of an infrared rooftop, hitting a chimney, regains its footing and continues on at
motion capture system to develop an animation of the gaits of dressage full speed. This was an action that could not safely be performed by
horses. a real horse, so the sequence had to be animated completely. This

Fig 20.16  Recording on treadmill in small green room: (A) camera stationary setup, (B) treadmill, (C) camera swing setup, (D) light setup and (E) video
result.

480
 Development of computer animation: an equine perspective

Fig 20.16  Continued

481
 20 Modeling, simulation and animation

Fig 20.16  Continued

process is a lot more time-consuming to create and thus more expen- Relations and hierarchies between the joints are programmed, often
sive than the live-action movements. A completely animated horse based on real anatomical ‘laws’. Examples are the directions and
model can be used to create almost life-like movements, including ranges of motion and whether the movement of one joint, influ-
ones that have not been seen before. The creativity and correct ana- ences movement of other joints. For example, the effect of the
tomical knowledge of the animators is the essential ingredient in reciprocal apparatus in the equine hind limb that synchronizes
making these animated movements, as there often is no example flexion of the stifle, hock and fetlock joints. After the IK skeleton is
that makes the desired movement exactly the way it has to be made. finished, the model can be viewed from any direction as it moves.
The way to create an animated horse model starts with the basics, Just having the IK skeleton of a moving horse is not sufficient, so
often going back to anatomical books. Skrba et al. (2008) provides the next step involves adding ‘body’ to the model and adding ‘skin’
an overview of animation methods that are used for quadruped to the body in a process called modeling, or creating a mesh model.
animation. The first step involves creating an animation skeleton, Horse meshes can be created by an artist or bought online. The IK
often called IK (inverse kinematics) skeletons (Fig. 20.18). These IK skeleton is bound to a mesh and rules are created determining how
skeletons are not as precise and detailed as a real skeletal structure, much of the mesh is influenced by each IK bone. Finally, controls
but they are a way to control the animated model and making it are created in a process called ‘rigging’ that enables the animator to
move according to physical laws. They are a simplified version of control the horse’s ‘acting’ intuitively.
the anatomical skeleton; not every joint has to be included, only After these rules or laws have been programmed, the creative work
the primary moving joints. Skrba et al. (2008) defines it as a hier- begins. For each frame it has to be decided which parts of the model
archy of local reference frames, each frame corresponding to a joint. need to be moved, in which direction and how much. This is a

482
 Development of computer animation: an equine perspective

Fig 20.17  Story board (A) of the rear view and screen shot (B) of the front view of a spectacular scene in which horse and rider gallop over snowy rooftops
in Amsterdam while being chased by police cars on the streets below.
From www.sintdefilm.com.

Fig 20.18  The ‘Rhett’ model by Tim Mayo and Chris Carson. (A) Defining the IK skeleton, (B) adding a mesh body to the skeleton, (C) setting bone influence
rules and (D) tuning rigging controls.
From www.noxlabs.com. 483
 20 Modeling, simulation and animation

Fig 20.18  Continued

484
 Development of computer animation: an equine perspective

Fig 20.18  Continued

combination of intense anatomical knowledge, lots of data and If we consider the potentials for veterinary use, then the possibilities
reference materials (often Muybridge’s photographs are consulted are exciting. Imagine shooting video of a patient with a suspected
for specific gaits and body posturing) and talent of the animator. lameness, loading it into a computer program and seeing a model
After the main movements have been programmed, the computer of the movement of that horse and being able to view it from every
can fill in the in-betweens. After this the animation process still angle or monitoring progression of lameness by comparing or over-
involves a lot of cleaning-up and addition of small details, such as lapping images taken at different points in time or quantify the
rippling of the skin, light fall, movement of hair, mane and tail, that reduction of lameness after treatment. As long as we are able to see
give the animation a feeling of reality. the potential in the developments of the entertainments industry
All this work combined results in a 2-min chase scene with (Moeslund et al., 2006) and have the ability and willingness to look
footage that was partly shot of a real horse galloping on the tread- over the boundaries of veterinary medicine, then the possibilities
mill, in the studio and partly made up of an animated horse model. for future collaborations are endless.

References
Anderson, F.C., Pandy, M.G., 2001. Dynamic changes in jumping. J. Exp. Biol. 208, the deep digital flexor tendon in adult
optimization of human walking. J. 249–260. horses. Vet. Quart 18 (Suppl. 2), 570–
Biomech. Eng. 123, 381–390. Brand, R.A., Pedersen, D.R., Davy, D.T., Kotzar, 574.
Blankevoort, L., Huiskes, R., 1996. Validation G.M., Heiple, K.G., Goldberg, V.M., 1994. Chaffin, D.B., 1969. A computerized
of a three-dimensional model of the knee. Comparison of hip force calculations and biomechanical model development and
J. Biomech. 29 (7), 955–961. measurements in the same patient. J. used in studying gross body actions. J.
Bobbert, M.F., Gómez Álvarez, C.B., Van Arthroplasty. 9, 45–51. Biomech. 2, 429–441.
Weeren, P.R., Roepstorff, L., Weishaupt, Brown, N.A., Pandy, M.G., Kawcak, C.E., Cheung, T.K., Thompson, K.N., 1993.
M.A., 2007. Validation of vertical ground McIlwraith, C.W., 2003. Force- and Development of a three dimensional
reaction forces on individual limbs moment-generating capacities of muscles electronic solids model of the lower
calculated from kinematics of horse in the distal forelimb of the horse. J. Anat. forelimb of the horse. Vet. Radiol.
locomotion. J. Exp. Biol. 210, 1885–1896. 203, 101–113. Ultrasound 34, 331–333.
Bobbert, M.F., Santamaría, S., 2005. Buchner, H.H.F., Savelberg, H.H.C.M., Becker, Clayton, H.M., Lanovaz, J.L., Schamhardt,
Contribution of the forelimbs and hind C.K., 1996. Load redistribution after H.C., Willemen, M.A., Colborne, G.L.,
limbs of the horse to mechanical energy desmotomy of the accessory ligament of 1998. Net joint moments and powers in

485
 20 Modeling, simulation and animation

the equine forelimb in the stance phase of and lowered heels on the biomechanics of during walking and trotting. Am. J. Vet.
the trot. Equine Vet. J. 30, 384–389. the equine hoof assessed by finite element Res. 71, 508–514.
Clayton, H.M., White, A.D., Kaiser, L.J., analysis (FEA). J. Vet. Med. A Physiol. Moeslund, T.B., Hilton, A., Kruger, V., 2006. A
Nauwelaerts, S., Lavagnino, M., Stubbs, Pathol. Clin. Med. 47, 73–82. Survey of advances in vision-based human
N.C., 2008. Short-term habituation of Hinterhofer, C., Stanek, C., Haider, H., 2001. motion capture and analysis. Computer
equine limb kinematics to tactile Finite element analysis (FEA) as a model to Vision and Image Understanding 104,
stimulation of the coronet. Veterinary predict effects of farriery on the equine 90–126.
Comparative Orthopaedics and hoof. Equine Vet. J. 33 (Suppl.), 58–62. Mulier, J.C., Mulier, M., Brady, L.P., et al. 1989.
Traumatology 21, 211–214. Hjertén, G., Drevemo, S., 1987. A method to A new system to produce intraoperatively
Colborne, G.R., Lanovaz, J.L., Sprigings, E.J., analyse forces and moments in the custom femoral prosthesis from
Schamhardt, H.C., Clayton, H.M., 1998. extremities of the horse during the stance measurements taken during the surgical
Forelimb joint moments and power during phase at the trot. In: Robinson, N.E., procedure. Clin. Orthop. Related Res. 249,
the walking stance phase of horses. Am. J. Gillespie, J.R. (Eds.), Equine exercise 97–112.
Vet. Res. 59, 609–614. physiology 2. Davis, ICEEP Publications, Mundermann, L., Corazza, S., Andriacchi, P.,
Collins, S.N., Murray, R., Kneissl, S., Stanek, California, pp. 587–598. 2006. The evolution of methods for the
C., Hinterhofer, C., 2009. 32-Component Hogan, H.A., Wichtmann, A.D., Hood, D.M., capture of human movement leading to
finite element models of a horse and 1991. Finite element analysis of the markerless motion capture for
donkey digit. Equine Vet. J. 41 (3), internal deformation and stresses in the biomechanical applications. Journal of
219–224. equine distal digit. Proc. 10th Ann. Mtg. NeuroEngineering and Rehabilitation 3, 6.
Delp, S.L., Loan, J.P., 1995. A Graphics-based Assoc. Equine Sports Med. 43. Muybridge, E., 1899. Animals in motion.
software system to develop and analyze Jansen, M.O., Bogert, A.J., van den, Republished (1957), L.S. Brown Editor,
models of musculo-skeletal structures. Riemersma, D.J., Schamhardt, H.C., 1993. Dover Publications, New York, USA.
Comput. Biol. Med. 25, 21–34. In vivo tendon forces in the forelimb of Neptune, R.R., Kautz, S.A., Zajac, F.E., 2001.
Delp, S.L., Loan, J.P., Hoy, M.G., Zajac, F.E., ponies at the walk, validated by ground Contributions of the individual ankle
Topp, E.L., Rosen, J.M., 1990. An interactive reaction force measurements. Acta. Anat. plantar flexors to support, forward
graphics-based model of the lower 146, 162–167. progression and swing initiation during
extremity to study orthopaedic surgical Lanovaz, J.L., Clayton, H.M., Colborne, G.R., walking. J. Biomech. 34, 1387–1398.
procedures. IEEE Trans. Biomed. Engin. 37, Schamhardt, H.C., 1999. Forelimb Neptune, R.R., Wright, I.C., van den Bogert,
757–767. kinematics and net joint moments during A.J., 2000. The influence of orthotic devices
the swing phase of the trot. Equine Vet. J. and vastus medialis strength and timing on
Encarnacao, J.L., Linder, R., Schlechtendahl,
30 (Suppl.), 235–239. patellofemoral loads during running. Clin.
E.G., 1990. Computer-Aided Design.
Fundamentals and System Architectures, Lawson, S.E., Chateau, H., Pourcelot, P., Biomech. 15, 611–618.
second ed. Springer-Verlag, Berlin. Denoix, J.M., Crevier-Denoix, N., 2007. Pandy, M.G., Zajac, F.E., 1991. Optimal
Sensitivity of an equine distal limb model muscular coordination strategies for
Galloux, P., Richard, N., Dronka, T., et al.
to perturbations in tendon paths, origins jumping. J. Biomech. 24 (1), 1–10.
1994. Analysis of equine gait using three
and insertions. J. Biomech. 40, 2510–2516. Passerello, C.E., Huston, R.L., 1971. Human
dimensional accelerometers fixed on the
saddle. Equine Vet. J. 17 (Suppl.), 44–47. Les, C.M., Keyak, J.H., Stover, S.M., Taylor, attitude control. J. Biomech. 4 (2),
K.T., 1997. Development and validation of 95–102.
Gleicher, M., 1999. Animation from
three dimensional finite element models of Pollock, S., Hull, M.L., Stover, S.M., Galuppo,
observation: motion capture and motion
the equine metacarpus. J. Biomech. 30, L.D., 2008a. A musculoskeletal model of
editing. Computer Graphics 33, 51–54.
737–742. the equine forelimb for determining
Grieve, D.W., Pheasant, S., Cavanagh, P.R., McGurk, M., Amis, A.A., Potamianos, P., surface stresses and strains in the humerus-
1978. Prediction of gastrocnemius length Goodger, N.M., 1997. Rapid prototyping part I. Mathematical modeling. J. Biomech.
from knee and ankle joint posture. techniques for anatomical modeling in Eng. 130, 041006.
Biomechanics VI-A (eds. Asmussen, E. and medicine. Annals Royal Coll. Surg. England
Jorgensen, K.). University Park Press, Pollock, S., Stover, S.M., Hull, M.L., Galuppo,
79, 169–174. L.D., 2008b. A musculoskeletal model of
Baltimore, pp. 405–412.
McLean, S.G., Huang, X., Su, A., Van Den the equine forelimb for determining
Groesel, M., Gfoehler, M., Peham, C., 2009. Bogert, A.J., 2004. Sagittal plane surface stresses and strains in the
Alternative solution of virtual biomodeling biomechanics cannot injure the ACL during humerus-part II. Experimental testing and
based on CT-scans. Journal of sidestep cutting. Clin. Biomech. 19, model validation. J. Biomech. Eng. 130,
Biomechanics 42, 2006–2009. 828–838. 041007.
Hahn, D., 2008. The alchemy of animation. McLean, S.G., Su, A., van den Bogert, A.J., Riemersma, D.J., Schamhardt, H.C., Hartman,
Disney Enterprises Inc., New York. 2003. Development and validation of a W., Lammertink, J.L.M.A., 1988. Kinetics
Halvorsen, K., Johnston, C., Back, W., Stokes, 3-D model to predict knee joint loading and kinematics of the equine hind limb: In
V., Lanshammar, H., 2008. Tracking the during dynamic movement. J. Biomech. vivo tendon loads and force plate
motion of hidden segments using Eng. 125, 864–874. measurements in ponies. Am. J. Vet. Res.
kinematic constraints and kalman filtering. Menasche, A., 1999. Understanding motion 49, 1344–1352.
J. Biomech. Eng. 130, 011012. capture for computer animation and video Riener, R., Edrich, T., 1999. Identification of
Harrison, S.M., Whitton, R.C., Kawcak, C.E., games. Academic Press, San Diego, CA. passive elastic joint moments in the lower
Stover, S.M., Pandy, M.G., 2010. Merritt, J.S., Davies, H.M., Burvill, C., Pandy, extremities. J. Biomech. 32, 539–544.
Relationship between muscle forces, joint M.G., 2008. Influence of muscle- Schryver, H.F., Bartel, D.L., Langrana, N., Lowe,
loading and utilization of elastic strain tendonwrapping on calculations of joint J.E., 1978. Locomotion in the horse:
energy in equine locomotion. J. Exp. Biol. reaction forces in the equine distal kinematics and external and internal forces
213, 3998–4009. forelimb. J. Biomed. Biotech. in the normal equine digit in the walk and
Hill, A.V., 1938. The heat of shortening and doi:10.1155/2008/165730. trot. Am. J. Vet. Res. 39, 1728–1733.
the dynamic constants of muscle. Proc. Merritt, J.S., Pandy, M.G., Brown, N.A., Burvill, Skrba, L., Reveret, L., Hetroy, F., Cani, M.P.,
Royal Soc. Series B. 126, 136–195. C.R., Kawcak, C.E., McIlwraith, C.W., et al., O’Sullivan, C., 2008. Quadruped
Hinterhofer, C., Stanek, C., Haider, H., 2000. 2010. Mechanical loading of the distal end Animation. The Eurographics Association
The effect of flat horseshoes, raised heels of the third metacarpal bone in horses 2008.

486
 References

Spoor, C.W., van Leeuwen, J.L., Meskers, musculo-skeletal system, third ed. J.Wiley & and Woo, S.L.-Y.). Springer Verlag, New
C.G.M., Titulaer, A.F., Huson, A., 1990. Sons, Toronto, Canada, pp. 624–664. York.
Estimation of instantaneous moment arms Van den Bogert, A.J., Schamhardt, H.C., 1993. Wright, I.C., Neptune, R.R., van den Bogert,
of lower-leg muscles. J. Biomech. 23 (12), Multi-body modelling and simulation of A.J., Nigg, B.M., 1998. Passive regulation of
1247–1259. animal locomotion. Acta Anat. 146, impact forces in heel-toe running. Clin.
Swanstrom, M.D., Zarucco, L., Hubbard, M., 95–102. Biomech. 13, 521–531.
Stover, S.M., Hawkins, D.A., 2005. Van den Bogert, A.J., Schamhardt, H.C., Wright, I.C., Neptune, R.R., van den Bogert,
Musculoskeletal modeling and dynamic Crowe, A., 1989. Simulation of A.J., Nigg, B.M., 2000. The influence of foot
simulation of the thoroughbred equine quadrupedal locomotion using a positioning on ankle sprains. J. Biomech.
forelimb during stance phase of the gallop. dynamic rigid body model. J. Biomech. 22, 33, 513–519.
J. Biomech. Eng. 127, 318–328. 33–41. Yeadon, M.R., 1993. The biomechanics of
Thompson, K.N., Cheung, T.K., 1993. A finite Van Ingen Schenau, G.J., Bobbert, M.F., 1993. twisting somersaults, part 3: aerial twist.
element model of the proximal sesamoid The global design of the hind limb in J. Sports Sci. 11, 209–218.
bones of the horse under different loading quadrupeds. Acta. Anat. 146, 103–108. Zajac, F.E., 1989. Muscle and tendon:
conditions. Vet. Comp. Orthop. Traumatol. Van Weeren, P.R., Van den Bogert, A.J., properties, models, scaling, and application
7, 35–39. Barneveld, A., 1990. A quantitative analysis to biomechanics and motor control. Crit.
Thornton, T.L., 2004. Computer Animation of of skin displacement in the trotting horse. Rev. Biomed. Eng. 17, 359–411.
Quadrupedal Locomotion. Thesis Texas Equine Veterinary Journal 9 (Suppl.), Zarucco, L., Swanstrom, M.D., Driessen, B.,
A&M University. 101–109. Hawkins, D., Hubbard, M., Steffey, E.P., et
Van de Panne, M., 2000. Control for simulated Visser, J.J., Hoogkamer, J.E., Bobbert, M.F., al., 2003. An in vivo equine forelimb
human and animal motion. Annual Huijing, P.A., 1990. Length and moment model for short-term recording of peak
Reviews in Control, 24, 189–199. arm of human leg muscles as a function of isometric force in the superficial and deep
Van den Bogert, A.J., 1989. Computer knee and hip-joint angles. Eur. J. Appl. digital flexor muscles. Vet. Surg. 32,
simulation of locomotion in the horse. Physiol. 61 (5–6), 453–460. 439–450.
PhD thesis, University of Utrecht. Wilhelms, J., Van Gelder, A., 2003. Combining Zarucco, L., Wisner, E.R., Swanstrom, M.D.,
Van den Bogert, A.J., Gerritsen, K.G.M., Cole, vision and computer graphics for video Stover, S.M., 2006. Image fusion of
G.K., 1998. Human muscle modelling from motion capture. The Visual Computer 19, computed tomographic and magnetic
a user’s perspective. J. Electromyogr. 360–376. resonance images for the development
Kinesiol. 8, 119–124. Winters, J.M. 1990. Hill-based muscle models: of a three-dimensional musculoskeletal
Van den Bogert, A.J., Nigg, B.M., 2007. Chapter a systems engineering perspective. Multiple model of the equine forelimb.
4.8. Simulation. In: Nigg, B.M., Herzog, W. muscle systems: biomechanics and Vet. Radiol. Ultrasound. 47,
(Eds.), Biomechanics of the movement organization (eds. Winters, J.M 553–562.

487
This page intentionally left blank
 Index

Page numbers followed by ‘f’ indicate figures Alias effect, 61, 63f Asia Minor, 3
and ‘t’ indicate tables. Aliasing, 35, 63 Aspartate, 77–78
Allometric form, 457 Assyrian culture, 3
Amino acids, 77–78, 425–426, 429 Asymmetry, 359–360, 360f
A AMP deaminase, 285 Ataxia, 53–54, 188, 188f
Abduction, 37, 38f Anaerobic capacity, 283, 290, 427–428, 428f Athenian Academy, 3
Absolute elevation, 363 Anaerobic fatigue, 430 Atlanto-occipital joint, 201
Absolute relative weight, 365–366 Anaerobic glycolysis, 422 Atlas, 199
Abu Bakr ibn el-bedr al Baytar, 5 Anaerobic pathways, 281–283 ATP:ADP ratio, 422
Acceleration regulation, 422–423 ATPase staining technique, 74–75
calculations, 455, 455f Anaerobic phosphorylation, 422 ATP synthetase, 420
jumping, 329–330 Anaerobic power, 427–431, 428f Aubin, 91
in thoroughbred racehorses, 308 Anaerobic threshold, 283, 429 Auer, Jörg, 23
Accelerometers, 50–52, 51f, 190–191 Analgesia, 193 Australia, studies in, 24
Acetylcholine (ACh), 74, 288 Analog to digital converter (ADC), 61 Auto-correlation, 68
Acetyl coenzyme A (Acetyl CoA), 420 Ancient cultures, 3 Axis, 199
Actin, 74–75 Anconeus muscle, 103 Axons, 76
Action, 166 Andalusian horses, 89–91 Aysan, Ihsan, 16
Action potentials, 74 André, Marcel, 17
Active neurophysiological response time, 155 Anesthesia, diagnostic, 192–193
Active optical systems, 479 Angular limb deformities, 168, 169f
B
Acupuncture, 407–408 Angular variables, 37 Back, 18–19
Adaptation Angular velocities, 453 anatomy, 199–205, 204f
functional, 267 Animal physiotherapy: assessment, treatment and biomechanics, 205–206
of muscles, 278–279, 284–287, 285f rehabilitation of animals, 382 bones, 199–205, 200f
Adduction, 37, 38f Animation, 476–485 coordination and locomotion, 206
Adductor magnus/adductor brevis, 128t, 130 history, 477–478, 477f–478f effect of the saddle on the motion of,
Adenosine diphosphate (ADP), 74, 419–420, modern, 478–485 342–343
420f motion capture, 478–479 function, 199–211
Adenosine triphosphate (ATP), 74, 281–283, non-optical systems, 479–485, 479f–485f length, 217–218
282f, 419–420, 420f, 423, 424t optical systems, 479 movement
Adipose tissue, 424–425 Anterior cruciate ligament injury, 444 in canter, 209–211, 210f
Advanced non-enzymatic glycation end Anti-aliasing filter, 63 in trot, 207–209, 207f–209f, 211
products (AGEs), 268 Anti-gravity muscles, 82 in walk, 206–207, 206f, 211
Advanced placements Antique world, 3–4 pain/dysfunction, 216–220
gait classification, 85 Apsyrtos, 4 causes, 384
racing sports, 308 Arab conquest, 5 clinical signs of, 387
riding methodology, 357–358 Arabian horses, 127–129 differences between horses with and
trot, 88 Ariel Performance Analysis System, 39 without back pain, 218–219
Aepli, A., 14 Aristotle, 3–4 effect of head and neck position,
Aerial locomotion, 11 Arthritis, 167–168, 175 219–220, 220f
Aerobic capacity, maximal, 419, 462–463 Arthrogryposis, 253 effect of local analgesia, 220
Aerobic glycolysis, 420, 421f Arthrosis, 167–168 electromyography, 388–389
Aerobic pathways, 281–283, 423 Articular cartilage functional assessment, 387–388
regulation, 422–423 biology, 268–270, 268f–270f kinematic examination of horses with
Aerobic phosphorylation, 420 cross-links, 268 back pain, 218
Aerobic power, 426–427, 426f developmental aspects, 270–271 kinematic examination of normal horses,
maximum, 427 effect of exercise on 216–218, 217f
Afanasieff, S., 15 during growth and development, kinematics, 388
Aggrecan, 268 271–273, 271f–272f objective measurements, 388
Agriculture, advent of, 2 in young adult animals, 273–274 patho-anatomical diagnosis, 386–397
Air mountain range, 3 loading, 268, 269f pressure algometry, 390–394, 390f,
Air resistance, 451 remodeling, 269–270 392t–393t
Akhi Hizam al-Furusiyah wa al-Khayl, 5 response to exercise, 268–274 prevalence, 384–386
Alexander, R. McNeill, 21–22 role of, 268 ultrasonography, 394–397, 395f–396f
Alexander the Great, 3–4, 4f Artificial neural networks (ANNs), 69, research on, 20
Alfort veterinary school, 6, 20–21, 21f 191–192 Back, Willem, 19

489
 Index

Back at the knee conformation, 233 Borelli, Giovanni Alphonso, 5–6, 5f lameness, 182–183, 187–188
Backin® program, 20 Borissov, V. M., 15 net moment, 107, 110f
Back lifts, 345, 345f Bouley shoe, 162, 163f net power, 110f
Badoux, Dick, 16, 19 Bourgelat, Claude, 6 three-dimensional forelimb analysis,
Bagley’s bundle, 73, 79 Box-walking, 374 119–120, 120f
Balance, 147 Brachiocephalicus, 101t, 102, 102f, 201 Cartilage see Articular cartilage
Balance of Borelli, 14f neck elevation, 208 Cartilage oligomeric matrix protein (COMP),
Bandages, 193 at walk, 206–207 273–274, 294
Bands, 369 Breakover, 122 Cathepsin B, 273
Banking, 165 Breeding, locomotory responses, welfare Cathepsin D, 273
Bantoiu, C., 12 implications, 376–377 Cave art, 1–3, 2f
Barneveld, Ab, 20 Bucephalus, 3–4 Cavendysh, William, 6
Barrey, Eric, 20 Buchmann, J., 12–13 Center of gravity, 5–6, 14, 14f
Barrier, G., 12, 13f Buchner, Florian, 19 Central brain mechanisms, 79
Bar shoes, 162f Bucked shins, 309 Central pattern generators (CPGs), 73, 77–78,
Baum, Hermann, 12 Buck knee, 233 82, 82f
Bayer, J., 7, 9f Buckle transducer, 49 environmental constraints, 78
Bearing phase, 120–122 Bulk epiphyseal trabecular bone (BETB), 274 extremity movements in tetrapods, 77–78
Becker, C. K., 19 Bursae, 201 hind limb, 77
Behavior Bushmen culture, 2 neural principles, 77, 78f
in neck dysfunction Byzantine Empire, 4 proprioceptive afferents, 78
normal grazing, 370 Cerebellar ataxia, 253
problem, 375 Cerebellar nuclei
response to pain, 391
C Cerebellum, 79, 82
stereotypical, 374 Calcineurin, 288 Cervical spine
Belgian studies, 22 Calf knee conformation, 233, 251–253 facet joints, 386
Bell boots, 163 ‘C’ allele, 248 intervertebral disc disease, 386
Bench kneed conformation, 233 Camcorders, 39 Cervical vertebral stenosis, 188
Bending force, 47 Cameras, 61 CGI motion capture, 479–482, 480f–483f
Beta-oxidation, 422 for optoelectronic systems, 40 Chambon, 345, 345f
Bethcke, 12 for photographic systems, 39 Chateau, Henry, 21
B4GALT7, 258–259 Canadian research, 22 Chestnut, 147
Biceps brachii, 103–104, 106 Cancellous bone, 275 Chieffi, Armando, 14
Biceps femoris, 128t, 129–130, 129f Canter, 91–92 China, 3
Biceps tendon, 103 classification, 86t Chondrocytes, 268–270, 272
Bilateral lameness, 183 collected, 325–326, 325t, 327f Chondrodysplasia
Biomechanics, computer modeling and, dressage horses, 325–326, 325t, 326f–327f complex, 254–257, 256t–257t
467–472 extended, 325, 325t simple, 257–259, 258f
Bio signals, 61, 62f footfall sequence, 140 Christian church, 5
Bits, 343–344, 344f forelimb, 113, 115f Chubb, 15
Biventer cervicis, 203 ground reaction forces, 326, 326f Cine film, 10–12
Björck, G., 16 hind limb, 127, 140–142, 141f Cinematography, 39
Blank joint, 270–271 kinematic analysis, 113 Circadian rhythms, 372
Blood gas analysis, 433 kinematic variables, 114t Circulation, 419
Blood oxygen-carrying capacity, 419 lead change see Lead changes Citrate, 422
Blood volume measurement, 433 medium, 325, 325t Clayton, Hilary, 22–23
B-mode ultrasonography, 396 neck and back movement in, 209–211, 210f Cleidocephalicus, 201
Bobbert, Maarten, 20 pirouette, 326–327, 327f Clicking phenomenon, 107
Body, conformation, 234–236 rider movements at, 353 ClinSeat, 53
Bone(s), 274–278 transitions Clips, 164
diaphysis, 274 to halt, 92–93 Codamotion, 40
epiphysis, 274 to trot, 92–93, 327 CODA-3® system, 19–20
extracellular matrix, 274–275 from walk, 93f Coffin bone, 147
growth plate closure times, 167t to walk, 92–93 Coffin joint, 150f
markers, 33 working, 325, 325t angulation, 156–157
measuring strain in, 47–48, 48f Capillaries, increased density, 284–286 in lameness, 178–179, 179f, 179t
metaphysis, 274 Capriole, 327 COL23A1, 258–259
mineral densities, 275–277 Carbohydrates, 419–420 Colborne, Bob, 22–23
morphology and physiology of, 274–275 storage, 424 Collagen, 269–270, 270f
neck and back, 200f, 201 utilization, 425–426 bone strength, 274–275
need for exercise, 275–276 Cardiac output, 419 cross-links, 268
response to exercise Carnus, Captain, 14–15 fibrils, 268, 268f, 272–273
general, 275 Carpal flexion test, 192 in tendons, 291, 294, 294f
in the horse, 276–278 Carpal joint Collateromotion, 116
strain gauges, 47–48, 48f in anesthetized horse, 107 Collection, 94, 360–366
Bone spavin, 168, 251–254 arthritis, 175 at canter, 325–326, 325t, 327f
Bony landmarks in kinematic analysis, 33 forelimb analysis, 106–107 forward stepping of hind limbs relative to
Borcherdt, Werner, 12 joint angle, 110f hind quarters, 364

490
 Index

head and neck positions, 361, 361f Cross correlation (CC), 68 Distal interphalangeal (DIP) joint
horse-rider interaction in, 360–366 Cross-links, 268 forelimb, 109–110, 112f, 116–119, 117f–118f,
at trot, 88, 323, 323t–324t, 325f, 361–364, Croup 121–122
363t–364t conformation, 238 hind limb, 137, 139f
at walk, 88, 322, 361, 362f, 362t in lameness, 181 Distal phalanx, 147
Common digital extensor, 104t, 105 Cryotherapy, 405, 406f Domestication
Common digital extensor tendon, 149f Cutting horses, 335, 335t of the horse, 1–2
Compact cortical bone, 275 Cytochrome c oxidase, subunit 4, isoform 2 of other animals, 2–3
Compensatory lameness, 181–182 gene (COX4I2), 247 Dopamine, 77–78
Complementary medicine, 381–382 Dörrer, H., 12
Complexus, 203 Dorsal longitudinal ligament, 201
Computed tomography (CT) analysis,
D Draw reins, 220, 222, 345
467–468, 472–473 Dark Ages, 5 Dressage horses, 320–327
Computer animation, 478 Dartfish, 39, 40f back movement in, 217–218
Computer models/modeling, 467 Davies, Helen, 24 breed selection, 89–91
applications, 467 Dawes, H. W., 17 canter in, 325–326, 325t, 326f–327f
and biomechanics, 467–472 Death rates, 382–383 lead changes, 326, 326t
finite element analysis, 470–472, Decision making, 69, 69f pirouette, 326–327, 327f
472f–473f Decline of horses, 15–16 capriole, 327
history of, 467 Deep cerebellar nuclei, 79 conformation, 232, 327
horse history, 472–474, 473f Deep digital flexor muscle, 104–105, 104t effect of training on, 94
kinematics, 468–469 force, 447–448, 448f gaits, 322–327, 322f
kinetics, 469–470, 469f–472f hind limb, 129f genetics, 320–321
visualization, 467–468 Deep digital flexor tendon, 147–148, 149f half pirouette at walk, 323
see also Computer simulation and coffin joint angle, 156–157 passage, 324, 324f, 324t
Computer simulation, 474–476, 474f compressive force on navicular bone, 187, piaffe, 324–325, 324t
applications, 476 187f rider effects, 321–322
designing musculoskeletal models, contraction, 168 transitions, 327
474–476, 475f wedges and, 164, 165f trot in, 88, 323, 323t–324t, 325f
Conformation, 229–244 Degenerative joint disease (DJD), 254 walk in, 88, 322, 322f, 322t
ancient work on, 4 Degenerative lameness-causing disorders, Drevemo, Stig, 18
conformational lameness-causing disorders, 254 Dušek, J., 17
251–253 Dehydration, 435 Dutch Warmbloods, ground reaction force,
deviations of limb and toe axis, 232–233, De la Croix, Magne, 14, 15f 42
232f, 233t Dendrite bundles, 76–77 Du Teil, Lenoble, 7
dressage horses, 327 Denoix, Jean-Marie, 20 Dwarfism, 257–259, 258f
forelimbs, 236–238, 236f–237f, 237t Dens, 199 Dynamic control, 399
genetics of, 245 Dermis, 147 Dynamic optimization models, 449
and growth, 230–232 Descending projections, 78–79 Dynamics, 443
head, neck and body, 234–236, 234f, Desmotomy, 193–194, 194f
235t De Solleysel, Jacques, 6
hind limbs, 238–241, 238f–240f Detraining, 286–287
E
history and tradition, 229 Deuel, Nancy, 23 Economy of locomotion, 429–430, 430f
jumping horses, 332 Development, 73–84 Edison, Thomas, 10–11
of pacers, 311 Developmental lameness-causing disorders, EGA system, 191
and performance, 246 254–257 Egg bar shoes, 161–163, 162f, 167
performance and soundness, 232 Diagnostic anesthesia, 192–193 Egyptian culture, 3
predicting performance and soundness by Diagonal dissociation, 310, 365 18th Century, 5–6
analysis of, 241–242, 241f Dietary alterations, effects on energy substrate Elastic materials deformation, 47
quantitative analysis, 230, 231f utilization, 425–426 Elastic recoil, 108, 133, 136
sport horses, 312 Differentiation, 63–64, 64f Elastic similarity scaling, 458
subjective evaluation, 229 definition of, 63 Elbow joint
of trotters, 311 increasing, decreasing and finding local conformation, 237–238
Constantinople, 4 maximum or minimum, 64 extensor moment, 107
Contact, 358–359 phase-plane analysis, 64 forelimb analysis, 106, 109f
Convolution, 68–69 physical concept of Newton, 63–64 joint angle, 109f
Coordinate transformation, 65 in time and frequency domain, 67 net moment, 106, 109f
Coordination in sport horses, 315–316 Digestible energy (DE), 423–424 net power, 109f
Coronet, 147 Digit, third, 147 Electrocardiography, telemetered, 432–433
Corticospinal tract (CST), 73–74, 79 Digital cushion, 156 Electrogoniometry, 38–39, 39f
Cost of locomotion, 459–460, 459f–460f, Digital extensors, 129f Electromagnetic kinematic analysis systems,
460t Digital flexion test, 192 40–41
Cow hocked conformation, 232, 254 Digital joints, 119t, 120, 121t Electromyography (EMG), 53–55, 73
Craniocaudal hoof balance, 148–150 Digital signal processing (DSP), 61 amplitude of signal, 55
Creatine kinase, 422 Diocletianus, Emperor, 4 back pain diagnosis, 388–389
Creatine phosphate, 424t Direct linear transformation (DLT), 36 electrodes, 54
Cretans, 3 Distal accessory ligament (DAL), 156–157 the electromyogram, 54–55, 54f
Cro-Magnon race, 1–2 Distal check ligament, 193–194, 194f equipment, 54

491
 Index

during locomotion in adult animals, 80–81, Exercise Flexion tests, 192, 192f
80f–81f articular cartilage response to, 268–274 Flex-o-meter, 192, 192f
muscle fatigue, 75 bone response to, 274–278 Flexor carpi radialis, 104, 104t
muscle potentials, 54 endogenous growth hormone responses to, Flexor carpi ulnaris, 104, 104t, 107
myoelectric signals, 54 435 Flexor digitorum lateralis, 128t, 130
neck and trunk muscles at trot, 207, energy pathways in, 423, 423t Flexor digitorum medialis, 128t, 130
208f–209f exercise-based treatment techniques, Flexor digitorum profundus, 128t, 130
neck and trunk muscles at walk, 399–401, 401f Flexor digitorum superficialis, 128t, 130
206–207 facilitation-based exercise therapy, 401–404, Flexural limb deformities, 168, 168f
percutaneous techniques, 54 402f–403f, 404t, 405f Flight arc, 157–158, 158f
rider skill assessment, 353 head and neck position Flock of Birds, 40–41
sampling frequency, 61 impact on welfare, 374 Flying trot, 91, 91f
suppleness, 358 muscles response to, 278–290, 279t Foals
surface, 54 oxygen uptake, 426–427 grazing, 370
Electron transport, 420 tendons and ligaments response to, kinetic activity, 369–370
Electro-physical intervention, 405–408 290–296 play, 371
Electrotherapy, 406–407 thermoregulatory consequences of, 435–436 stretching, 370
Ellenberger, Wilhelm, 12 Exercise bands, 404, 405f Foot
Emotional motor system (EMS), 79 Exertional rhabdomyolysis, 253–254 conformation, 251–253
Endurance racing, 311–312 Exo-skeleton motion capture systems, 479, functional anatomy, 147–148, 149f–150f
Energetic efficiency, 161–162, 306 479f nerve supply, 148, 152f
Energetics of locomotion, scaled see Scaled Expiration, 95–96 proprioception, 148, 153f
energetics of locomotion Extension, 37, 38f uneven, 242, 242f, 251–253
Energy Extension phase, 80 vascular supply, 148, 151f
cost of locomotion, 459–460, 459f–460f, Extensor carpi radialis, 103–104, 104t, 107 Footfall sequences, 87f–88f, 90f, 92, 92f, 140
460t Extensor digitorum lateralis, 128t Footing, 164–165, 166f
expenditure, 426–431 Extensor digitorum longus, 128t, 130 Force plates, 43–45, 44f
measuring, 431–435 Extracellular matrix (ECM), 291 bad trials, 44
forelimb, 105 Extracorpral shock wave therapy (ESWT), 193 calibration, 44
hind limb, 127, 140 Extremity muscles, 82 installation, 44
pathway in exercising horse, 423, 423t lameness, 191
production, 419–423 selection of, 43–44
stores, 424, 424t, 451–452
F surface materials, 44–45
use on treadmill, 32 Facet joints, 201 in treadmills, 44–45, 45f
Energy substrates, 423–426 Facilitation-based exercise therapy, 401–404, variables, 44
effects of dietary alterations on, 425–426 402f–403f, 404t, 405f Force shoes, 45–46, 46f, 191
Enlightenment, 6 FADH, 420, 421f, 422 Force transducers, 49
Eohippus, 267, 369, 370f Fat, 424–425, 424t Forelimbs, 99–125
Epidermal lamellae, 147, 149f Fatigue, 430–431 canter, 113, 115f
Equestrian sports limb coordination changes, 114–115 computer modeling, 469–470, 469f
history of, 17 in thoroughbred racehorses, 309 conformation, 236–238, 236f–237f, 237t
performance in, 305–340 Fatty acids, 419–420, 424t energy storage, 105
popularity of, 1 utilization, 422 gallop, 114–115
see also specific sports Feeding, 372–373 ground reaction force, 99
Equinalysis, 39 Femur, 239, 239f joint angles, 99, 100f, 113f
Equine Gait Trax Digital Motion Analysis Fetlock joint, 149f kinematics, 99, 114t, 115–122
System, 39 angulation, 156–157 in sport horses, 312, 313f–314f
Equine motor neuron disease (EMND), extension, 185 lameness, 177, 177f, 177t, 180–182
53–54 hyperextension, 183 musculotendinous architecture, 99–105
Equine myotonic dystrophy, 53–54 in forelimb lameness, 178–179, extrinsic muscles, 101–103, 101t, 102f
Equine protozoal myelitis, 188 179f–180f, 179t, 185 intrinsic muscles, 103–105, 104t
Equitation science, 377 reduction in lameness, 185 net joint moment, 110–111
Equusense, 51 FGFR1/FGFR2, 259 protraction, 157
Ergot, 147 Fiberoptic transducers, 49 sagittal plane analysis, 105–115
Ethology, 369 Fibroblast-like cells, 291 segments, 100f, 100t
Euler, Leonhard, 443 Fibrocartilage, 147–148 stance phase, 110–111
Euler angles, 37 Fibronectin, 273 stride, 99
Europe, 17–22 Fibrotic myopathy, 189, 189f swing phase, 110
Eventing, 333–334 Fibularis tertius, 127, 130 terminology, 99
dressage performance, 333 Filtering three-dimensional kinematics, 115–122
effect of added weight on jumping, 334, signal processing, 66–69, 66f trot, 106–111, 107f, 116–120, 117f–118f
334t special filters, 68–69 walk, 111–113, 116–120, 117f–118f
endurance performance, 333–334 Finite difference method, 446–447, 455, 455f Forging, 235
speed performance, 333–334 Finite element analysis (FEA), 470–472, Forssell, G., 14–15
Evolution, 369, 370f 472f–473f Forward dynamics, 443
Excess postexercise oxygen Flavin adenine dinucleotide (FAD), 420 Founder, chronic, 167
consumption(EPOC), 429 Fleischer, Max, 478, 478f Fourier analysis, 65–66, 66f
Excitatory amino acids, 77–78 Flexion, 37, 38f Four point trimming, 152

492
 Index

Foxtrot, 91 Gamgee, Joseph, 7 Greek civilization, 3–4

Franke, Horst, 12–13 Gap junctions, 294 Gregory, W. K., 15
Franks, 5 Gastrocnemius, 128t, 129f, 130 Grogan, J. W., 17
Frederick II, 5 Gene expression, 246–248, 249f, 249t–250t, Ground properties, 164–165, 166f
Fredricson, Ingvar, 17–18 287–288, 288f Ground reaction force (GFR), 41–43, 43f,
Free body diagram, 445–446, 445f–447f, 454f Genetics, 245–265 444–445
Freedom of the shoulders, 237–238 complex chondrodysplasia: osteochondrosis, in canter, 326, 326f
Free fatty acids, 285–286 254–257, 256t–257t computer modeling, 469
French supremacy, 12 conformational lameness-causing disorders, data, 447
Frequency domain see Time domain-frequency 251–253 force components, 42, 42f
domain dressage horses, 320–321 standard patterns, 42
Fresians, dwarfism in, 257–259, 258f historical background, 245 force plate see Force plates
Frog, 147 neuromuscular lameness-causing disorders, force shoes, 45–46, 46f
proprioception, 148 253–254 forelimb, 99
Froude number, 315, 456–457 new genomic technologies, 246–248, 249f, galloping, 45–46
Fructose-1,6-bisphosphate (F-1,6-BP), 420 249t–250t hind limb, 139
Fructose-6-phosphate (F-6-P), 420 selection for locomotor performance, hoof balance, 150–152
FSA, 53 246–248 indirect measurement methods, 46–47
Functional adaptation, 267 selection for locomotor soundness, jumping, 329, 329f, 331, 331f
Functional proprioceptive taping technique, 250t–252t, 251–259, 253f in lameness, 43, 178, 184, 184f
404, 406f simple chondrodysplasia: dwarfism, navicular disease, 187
Fuzzy clustering, 69 257–259, 258f normalization, 43
Fuzzy logic, 69 Genieser, D., 16 in passage, 324
Fuzzy sets, 69 Genome mapping, 246–248 in piaffe, 324–325
Genome-wide association studies (GWAS), postural sway analysis, 43, 44f
247 power calculations, 450–451, 450f
G Geometric scaling, 455–457, 456f–457f, 456t standardization, 36–37
GABA, 77–78 Germany, horse population in, 16 in thoroughbred racehorses, 308
Gait analysis, 444 Germany supremacy, 12–15 trot, 42–43
biological variability, 31–32 Gingerich, D. A., 23 trotters, 310
measurement techniques, 31–60 (see also Girths, 343 walking, 42–43, 112
specific gaits; specific techniques) Girtler, D., 19 Group housing, 373
revolution in, 7–12 Glanders, 6 Growth and conformation, 230–232
Gait diagram Global Coordinate System (GCS), 33–34, 35f Growth hormone, 434–435, 435f
Goiffon and Vincent, 7, 7f–8f Glucose, 74, 420 Growth hormone-releasing hormone (GHRH),
Lecoq, 7, 9f Glucose-6-phosphate, 420 434
Gaits, 85–97 Glucose-1-phosphate (G-1-P), 420 Growth plate closure times, 167t
asymmetry, 85 Glucose transporter 4 (GLUT-4), 285–286, 420 Grzimek, Bernhard, 15–16
classification, 85–92, 86t, 89f L-Glutamate, 77–78 Günther, Karl, 7
description of, 85–92 Gluteal muscles, 129f Gyroscopes, 50–52
development, 94 Gluteus medius, 127, 128t, 129
footfall sequences, 87f–88f Gluteus profundus, 128t, 129
genetics of, 245 Gluteus superficialis, 128t, 129
H
ontogeny, in sport horses, 314–315, 314f Glycerol-3-phosphate dehydrogenase, 285 HAAS (head acceleration asymmetry),
running, 85, 86t Glycogen, 285–286, 419–420, 423t–424t 190
slow, 88 Glycogen branching enzyme deficiency Half pirouette at walk, 323
stepping, 85 (GBED), 253 Hall effect transducers, 49
symmetry, 85 Glycogen synthase (GS), 424 Halt transitions
training effect, 94 Glycolysis from canter, 92–93
transitions, 92–93, 327 aerobic, 420, 421f from walk, 92–93
treadmill influence on, 94–95, 95f anaerobic, 422 Hamstrings, 127
variables, 90t Glycolytic pathway, 423 Hands of riders, 355, 356f–357f
walking, 85, 86t Glycosaminoglycans (GAGs), 270f, 274, 291 Han dynasty, 3
western pleasure, 335, 335t Goiffon, 6–7, 7f–8f Hard tissues
see also specific gaits Goniograms, 38–39 measuring strains in, 47–48, 48f
Galenus, 4 Goodman, Neville, 7 see also Bone(s)
Gallop, 91–92 Goodship, Allen, 21–22 Harness, 3
classification, 86t Goubaux, A., 12, 13f Hartman, W., 19
footfall sequence, 87f–88f, 92f Gracilis, 128t, 129f Haunches, lowering, 365
forelimb, 114–115 Grassfoot phenomenon, 241f–242f, 242 Haussler, Kevin, 23
ground reaction force, 45–46 Gravitational force, 444–446, 454–455 Head
heritabilities, 90t Gray, J., 15 conformation, 234–236
hind limb, 127 Grazing motion in thoroughbred racehorses,
kinematic analysis, 114–115 normal behavior, 370 308
lead change, 91 restricted, impact on locomotion, 373 movement
rotary, 91–92, 92f Great Britain in lameness, 181–182, 181f–182f,
stride parameters, 94f horse population in, 16 185–186, 206
transverse, 91–92, 92f research activities in, 21 in trot, 207

493
 Index

position, 211–212, 212f–214f, 220–222, center of pressure, 156–157 heart bar, 162f, 163
220f, 221t central reference point, 150–152 historical aspects, 147, 148f
in collection, 361, 361f cracks, 166 manipulations, 161–164, 162t
impact on locomotion, 375–376 flight arc, 157–158, 158f clips, 164
Heart bar shoes, 162f, 163 general anatomy, 147–148, 149f–150f heel caulks, 164
Heart rate measurement, 432–433 ground interaction, 164–165, 166f hoof pads, 163
Heat therapy, 405–406 horn, 147 length of shoes, 162–163, 162f–163f
Heel caulks, 164 impact, 152–156, 155f nails, 164
Heels, 147 initial ground contact, 152 rims, 164
underrun, 149–150 interaction with the ground, 137–138 toe grabs, 164
Hematocrit, 433 lameness, 175–176 toe of the shoe, 163–164, 163f
Hemispherical markers, 33 landing pattern in lameness, 178 wedges, 164, 165f
Hereditary equine regional dermal asthenia length, 161 weight of shoes, 161–162, 161f
(HERDA), 253 manipulations, 158–161, 159f–160f width of shoes, 163
Hereditary multiple exostoses, 253 mechanics rolled toe/rocker toe, 167
Heritabilities during locomotion, 152–158 Seattle, 161–162
gallop, 90t standing horse, 148–152 Housing, impact on locomotion, 373
trot/trotting, 88–91, 90t mechanism, 156, 156f–157f Howell, A. Brazier, 15
walk/walking, 88, 90t, 322 movement in lameness, 177–178 Humerus, 236–238
Hethiter, 3 nerve supply, 148, 152f Hunagdi, Shih, 3
Hexokinase (HK), 420 practical applications to performance, Huns, 2–3
High-speed video cameras, 39 166 Hunting response, 405
High-velocity, low amplitude thrust (HVLA) pressure mats, 52, 52f Hyaline cartilage, 147–148
techniques, 398–399 proprioception, 148, 153f Hydrotherapy, 404
Hildebrand, Milton, 14, 17 segments, free body diagram, 445–446, Hydroxyapatite, 274–275
Hill muscle model, 475, 475f 445f–447f, 454–455, 454f Hydroxylysylpyridinoline, 270f
Hind limbs, 127–145 shape, 148–149 5-Hydroxytryptamine (5-HT), 77–78
canter, 127, 140–142, 141f sliding of, 154–155 Hypercapnea, 433
conformation, 238–241, 238f–240f strain, 156 Hyperkalemic periodic paralysis (HYPP),
energy generation, 140 three-dimensional analysis, 121t 53–54, 253–254
gallop, 127 trajectory in lameness, 178 Hyperthermia, 253
joint angle-time, 141f trimming Hypothalamic-pituitary-adrenocortical (HPA)
kinematics, 305 four point, 152 axis, 434
in sport horses, 312–313 manipulations, 158–161, 159f Hypoxemia, 433
lameness, 177, 181–182, 182f vascular supply, 148, 151f Hypoxia inducible factor 1, alpha subunit
musculotendinous architecture, 127–131, wall, 147 (basic helix-loop-helix transcription
128t, 129f accelerometers, 51 factor) gene (HIF1A), 248
net joint moments, 140 cracks, 166 Hyracotherium, 267, 369, 370f
protraction/retraction, 139–140, 364 growth, 147, 159–160
rotation, 232–233 strain gauges, 48, 48f
sagittal plane analysis, 131–142, 133t–134t strains, 156
I
segments, 133f, 133t see also Horseshoes Icelandic horses, 88, 90f
stay apparatus, 130–131 Hoof pads, 163 I (inverse) K (kinematics) skeletons, 482,
trot, 127, 133–138, 141f Hoof-pastern axis, 148, 154f 483f–485f
walk, 134t, 138–140, 141f HoofTM System, 52, 52f Iliacus, 128t, 205
Hind quarters lowering, 365, 365f Hormone profile, 434–435 Iliocostal system, 203
Hip joint, 131, 133–135 Horn, 147 Immunodeficiency, combined, 253
extension, 139 tubules, 147 Impar ligament, 147–148, 149f
extensor moment, 134–135 Horse bands, 369 Impulsion, 359
flexion, 139 H(orse) INDEX, 43, 191 Indirect mobilization techniques, 401–402,
joint angle, 135f Horse racing see Racing/racehorses 402f
net moment, 134–135, 135f Horse-rider interaction, 341–368 Inertial sensors, 50–52
net power, 135f basic rider movements, 347–353, 347f Inferior olivary nucleus, 79
Hittites, 2–3 bits, 343–344 Infraspinatus, 103
Hock angles, 239–240, 240f collection, 360–366 Injury
Hock flexion test, 192 hind quarters lowering/increased hind joint prevalence of, 383
Hoggar mountain range, 3 angulation, 365, 365f thoroughbred racehorses, 309
Hoof reins, 345–347 Inspiration, 95–96
angle, 158–160, 159f–160f rider skill, 353–357 Insulin, 420
balance, 147–152, 155f rider’s weight influence, 357 Integration, 64, 64f, 67–68, 68f
craniocaudal, 148–150 riding methodology, 357–360 Intensive management, impact on
four point trimming, 152 saddle, 341–343 locomotion, 372
mediolateral, 150–152 Horseshoes Interference, 166
bearing surface, 149–150 bar, 162f, 163, 167 Interlimb coordination, 85–97
bony skeleton, 147 Bouley, 162, 163f see also Gaits
breakover, 157, 157f, 159, 163–164 effect on breakover, 157 International Conference on Equine Exercise
capsule, 147, 149f effect on hoof mechanism, 156, 157f Physiology (ICEEP), 18
cartilages, 147–148 egg bar, 161–163, 162f, 167 Interspinous ligaments, 201

494
 Index

Intertransverse ligaments, 201 suspension, 328f, 330, 330f forelimb, 105–115

Intervertebral discs, 201 terminology, 328, 328f ground reaction force see Ground reaction
Intralimb coordination, 105, 131, 138–139, training welfare issues, 376–377 force (GFR)
314 water jumps, 332 hind limb, 131–142
Intramuscular electromyography, 53–54 Junctional epidermolysis bullosa (JEB), 253 inverse dynamic analysis, 47
Intra-osseous LEDs, 19 lameness, 183–186
Intrathoracic pressure in neck dysfunction, 215 limb, in thoroughbred racehorses,
Inverse dynamic analysis (IDA), 47, 443–454
K 308–309
accuracy, 452–453 Kadletz, Max, 13 normalization, 43
applications, 444 Kaegi system, 191 shod versus unshod horses, 162t
calculations, 444–446, 445f–447f Kaemmerer, K., 16 shoe manipulations, 162t
computer modeling, 469 Kalman filters, 69 variables, 443
formulas, 454–455 Keegan, Kevin, 23 KinTrak/Orthotrak, 39
input variables measurement, 446–447, Keller, Professor, 12 Knee
448f Keratan sulphate, 274 bench, 233
joint contact force, 450, 450f Kersjes, A. W., 19 buck, 233
linked segment model, 444, 444f–445f Khan, Genghis, 2–3 calf, 233, 251–253
muscle force, 447–450, 448f–449f Kinematic crosstalk, 38 extensor moments, 444
navicular disease, 187 Kinematic research, 19–20, 20f Knezevic, Peter, 16, 19
net joint force, 450 Kinematics/kinematic analysis, 31, 33–38 Kobluk, Calvin, 23
net joint moment, 447–450 back function, 216–218 Kolda, J., 14–15
power calculations, 450–452, 450f back pain diagnosis, 388 Kronacher, C., 12–13
three-dimensional, 453–454 calibration, 34–35 Kroon, H. M., 14–15, 15f
Italy, 5 canter, 113 Krüger, L., 16
Ivanov, S., 15 computer modeling, 468–469 Krüger, Wilhelm, 13–14, 14f, 16
digitization, 35
following back pain treatment, 399
J forelimb, 99, 105–122, 114t
L
Jansen, 19 in sport horses, 312, 313f–314f Lacertus fibrosus, 103, 107
Japan, studies in, 24 gallop, 114–115 Lactate analysis, 432
Jeffcott, Leo, 18, 24 hind limb, 131–142 Lactate dehydrogenase, 285
Jog, 335, 335t in sport horses, 312–313 Lagrange, Joseph Louis, 443
Joint angle patterns kinematic data, 37 Lamellar bodies, 148
in lameness, 178–179 lameness, 176–183 Lameness, 166–168
mixed lameness, 183 lameness indicators, 190–191 accelerometers, 190–191
Joint angles, 36 markers, 33–34, 34f–35f assessment, 176
changes, 38 methods, 38–41 (see also specific asymmetry, 176–177
conformation analysis, 241 methods) and back movement, 222
forelimb, 99, 100f movement in relation to rider skill, 355 and back pain, 384
hind limb, 131, 131f multi-planar analysis, 37 bilateral, 183
increased hind joint, 365 neck and back movement, 207, 209 carpal joint, 182–183
see also specific joints normalization, 36–37 classification of, 175–176, 176f
Joint contact force, 450, 450f sagittal plane analysis, 37 compensatory, 181–182
Joint coordinate system (JCS) forelimb, 105–115 conformational lameness-causing disorders,
forelimb analysis, 115–116, 116f hind limb, 131–142, 133t–134t 251–253
kinematic crosstalk, 38 sampling frequency, 35 definitions, 175
three-dimensional, 37–38, 38f shoeing for the first time, 161 degenerative lameness-causing disorders,
Joint(s) smoothing, 35, 36f 254
disorders, 268 stick figures, 33, 33f developmental lameness-causing disorders,
manipulation/mobilization, 398–399, 398f three-dimensional analysis, 34, 37–38, 254–257
moments, 443, 453 38f diagnostic aids, 192–193
Jump/jumping, 92, 327–333 digital joints, 119t, 120 EGA system, 191
acceleration, 329–330 forelimb, 115–122 force plates/shoes, 191, 191f
approach, 328–330, 328f–330f tarsal joint at trot, 142, 142f–143f forelimb, 177, 177f, 177t, 180–182
conformation for, 332 trot, 116–120, 119t, 120f gait adaptation in, 175–197
departure, 328f, 331–332, 331f walk, 116–122, 119t ground reaction force, 43
effect of added weight, 334, 334t transformation, 36 head and trunk movement, 181–182,
effect of a rider, 333 treadmill, 32 181f–182f, 185–186
effect of early training, 332–333, 333f trot, 106–111 hind limb, 177, 181–182, 182f
ground reaction forces, 329, 329f, 331, variables, 443 hoof-landing pattern, 178
331f walk, 111–113 hoof movement, 177–178
injuries, 383 Kinetic research, 19 hoof pressure mats, 52
landing, 328f, 331–332, 331f Kinetics/kinetic analysis, 31, 41–47 hoof trajectory, 178
lift-off, 328–330, 328f–330f computer modeling, 469–470, 469f incidence, 383
mechanics, 328–332, 328t, 329f–331f early activity, 369–370 joint angle patterns, 178–179
predictability of performance, 332–333 foals, 369–370 kinematics indicators, 190–191
puissance technique, 332 force plate see Force plates kinematics of, 176–183
stride characteristics, 328, 328t force shoes, 45–46, 46f kinetics of, 183–186

495
 Index

limb movement, 178–179 Load Martin, Paul, 12

load redistribution, 185–186, 186f redistribution in lameness, 185–186, 186f Mary Anne McPhail Equine Performance
load reduction reduction Center, 22–23
in lame limb, 184–185, 184f in lame limb, 184–185, 184f Maximum accumulated oxygen deficit
mechanisms of, 185 mechanisms of, 185 (MAOD), 428, 428f
locomotion analysis, clinical use of, Loading Mechanical analysis, 443–455
189–192 articular cartilage, 268, 269f Mechanical horse simulator, 468, 468f
mixed, 182–183 limbs, 156 Mechanical intervention, 405–408
models, 175 tendons, 292–293 Mechanical motion capture systems, 479, 479f
neuromuscular lameness-causing disorders, Local analgesia, effect on back movement, 220 Mechanical nociceptive threshold (MNT), 390,
253–254 Locomotion 393–394, 393t
peculiar, 188–189 development of, 369–373 Medean culture, 3
prevalence of, 383 economy of, 429–430, 430f Mediolateral hoof balance, 150–152
specific, 186–189 effect on management regimes on, 373–374 Membrane fatty acid translocase (FAT/CD36),
stride variables, 176, 177t energy cost of, 459–460, 459f–460f, 460t 422
supporting limb, 176–182 impact of different training methods, Merkens, H. W., 19
swinging limb, 178, 182–183 375–376 Mesencephalic locomotor region, 79
temporal stride pattern, 176–178 impact of housing/restricted grazing on, Mesopotamia, 3
therapeutic aids, 193–194 373 Metabolic energetics, 419–441
timing changes, 185 impact of intensive management on, 372 energy expenditure, 426–431
vertical displacement of body, 185 problem behaviors, 375 energy production, 420–423
see also Soundness Locomotor system development, 267 energy substrates, 423–426
Laminitis, 166–167, 187, 187f Locomotory responses, 376–377 thermoregulatory consequences of exercise,
Landing, 120 Longevity, 382–384 435–436
Lanovaz, Joel, 22–23 Longissimus muscle, 81, 209 training programs, 431–435
Lanyon, Lance, 21–22 in canter, 211 Metacarpal bones, 149f
Laterality see Sidedness in stance phase, 209, 211 segments, free body diagram, 446
Lateral rectus capitis, 201–202 in walk, 206 third, three-dimensional analysis, 121, 121t
Latissimus dorsi, 101t, 102–103, 102f Longissimus system, 203 Metacarpophalangeal (MCP) joint, 111–112,
Laughlin, Harry H., 15 Longus capitis, 202–203 111f, 118f, 119–120
Leach, Doug, 22 Longus cervicis, 202–203 effect of exercise on, 273
Lead changes, 91 Loose-boxes, 373 forelimb analysis, 108–109
in dressage horses, 326, 326t Loose-housing, 373 Metacarpus analysis, 472
before jumping, 331 Lope, 335, 335t Metatarsophalangeal (MTP) joint, 136–137,
in thoroughbred racehorses, 308 Low palmar nerve block, 148 138f, 140f
Leclerc, George, Comte de Buffon, 6 Lumbosacral spine, 400 Michigan, studies in, 22–23, 23f
Lecoq, F., 7, 9f Middle Ages, 1, 5
Le Hello, P., 11–12 Middle phalanx, 147
Leonardo da Vinci, 5
M Missouri, studies in, 23
Leucine, 425 Machining, 468 Mitochondria, 419
Life spans, 382 Maennicke, E. M., 16 Mixed lameness, 182–183
Ligamenta flava, 201 Magdalenian period, 1–2 M-mode ultrasonography, 396
Ligaments Magne de la Croix, P., 14, 15f Moment, 443
injury, 383 Magnetic motion capture systems, 479 generation in muscle, 444, 445f
measuring strains in, 48–49 Magnetic resonance imaging (MRI), 467–468, generators, 449–450
of the neck and back, 201, 202f 472–473 of inertia, 444
response to exercise, 290–296 Magnetic resonance imaging (MRI)-diffusion joint rotation, 446
Limbic motor system (LMS), 79 tractography, 73 relative to centre of mass, 446
Limb(s) Magnetometers, 50–52 see also Net joint moment
angular deformities, 168, 169f Maillard reaction, 268 Mononeural innervation, 76
deformities Maitland mobilization technique, 398–399, Monte Carlo simulation, 453
angular, 168, 169f 398f Moskovits, 13
flexural, 168, 168f Malignant hyperthermia, 253 Motion analysis see Kinematics/kinematic
deviations, 232–233, 232f, 233t Mamillary processes, 199–200 analysis
flexural deformities, 168, 168f Management regimes, effect on locomotion Motion Analysis System, 39–40, 40f
kinetics, in thoroughbred racehorses, and welfare, 373–374 Motion capture, 478–479
308–309 Maneuverability in western sports, 334 Motoneurones, 76–78, 77f
loading, 156 Manipulation/mobilization, 398–399, 398f, Motor control, 399–401
movement in lameness, 178–179 401f Motor unit, 76–78, 212, 280–281
numerical modelling of the, 49 Manter, John T., 15 Motor unit action potentials (MUAPs), 54–55
power outputs, 419 Manual therapy, 397–408 Mounted soldiers, 2–3
protraction, 157, 178 Marey, Etienne Jules, 7–12, 11f Mounting, 342
retraction, 178 Markerless systems, 479 Movement
Linear scoring system, 229, 230f, 246 Markers data, 447
Linear variables, 37 back function assessment, 217f efficiency in racehorses, 306, 306f
Linked segment models, 444, 444f–445f kinematic analysis, 33–34, 40 muscles and, 74–75
Lipolysis, 424–425 skin see Skin, markers in relation to rider skill, 355–357
Liquid metal strain gauges, 49 three-dimensional kinetic analysis, 116 Moving average, 61–62, 63f

496
 Index

Mulligan mobilization-with-movement Myosin, 74–75 Neural transmitters, 77–78

technique, 398–399 Myositis, 53–54 Neurobiology, 73–84
Multifidus muscle, 209, 399–400 Myostatin gene (MSTN), 247–248 Neuromotor control, 399–401
anatomy, 204–205 Myotonia, 53–54 Neuromuscular electrical stimulation (NMES),
in canter, 211 Myotonia congenita, 53–54 407
EMG, 81 Myotubes, 75 Neuromuscular functionality, neck dysfunction
at walk, 206–207 Neuromuscular lameness-causing disorders,
Multi-planar analysis, 37 253–254
Muscle cells, 75
N Newton, Isaac, 443
Muscle enzymes, 212, 284–285 NADH, 420, 421f, 422 Newton concept, 63–64
Muscle fibers, 74, 280–281, 283f Nails, 164 Newton’s laws of motion, 444–445
in different breeds, 75 National Library of Medicine’s Visible Human Nicolescu, 12
distribution, 75 Project, 468 Nicotinamide adenine dinucleotide (NAD),
hind limb, 127 Navicular bone, 147–148, 149f, 167, 187, 420
histology, 74–75 187f Niki, Y., 24
hypertrophy, 75, 287–288 Navicular bursa, 149f 19th century, 6–15
mechanical force, 74 Navicular disease, 186–187, 187f Nomura, 17
resting, 54 genetics, 254 Non-esterified fatty acids (NEFAs), 422, 424t
size, 284 lameness patterns, 176 Non-steroidal anti-inflammatory drugs
type I, 74–75 shoe wedges, 164 (NSAIDs), 193
type II, 74–75 Navicular ligament, 147–148, 149f North America see United States
types, 74–75 Navicular syndrome, 167 Nuchal ligament, 201, 202f
type transitions, 284 Neck Numerical modelling of the limbs, 49
Muscle potentials, 54 anatomy, 199–205, 204f Nutrition and muscle development, 75
Muscle(s), 278–290, 280f biomechanics, 205–206 Nyquist-Shannon sample theorem, 61–63,
adaptation to training, 278–279, 284–290, bones, 199–205, 200f 62f
285f, 289t–290t conformation, 234–236, 234f
biopsy, 279, 280f–281f, 433–434 coordination and locomotion, 206
body mass, 127 dysfunction, 211–216
O
buffering capacity, 286 head and neck positions, 212f–214f, 221t Obliquus capitis caudalis, 201–202
contraction, 74, 279–283, 419–420 markers Obliquus capitis cranialis, 201–202
electromyography, 53–54 measurements from neck Obliquus externus abdominis, 206–207, 209
force of, 101 discussion, 216 Obliquus internus abdominis, 211
detraining, 286–287 interpretations Ogrizek, A., 12–13
development, 75 neuromuscular functionality assessment Omotransversarius, 101t, 102, 102f, 201
exercise physiology, 279–284 radiographical analysis Onset of blood lactate accumulation (OBLA),
fatigue, 75 respiratory functionality assessment 429
force, 279–280, 447–450, 448f–449f, short-term stress parameters evaluation Ontrack, 39
460–462, 461f video analysis Optical motion capture, 33
forelimb, 99–105 function, 199–211 Optimization models, 449
extrinsic, 101–103, 101t, 102f hyperflexion, 21, 211, 375 Optoelectronic kinematic analysis systems,
intrinsic, 103–105, 104t motion in thoroughbred racehorses, 308 39–40, 40f
function studies, 49 movement Optotrak Certus, 40
hind limb, 127–131, 128t, 129f in canter, 209–211, 210f Orthopedics, 192–194
increased capillary density, 284–286 in trot, 207–209, 207f–209f, 211 Osteoarthritis, 397
increased mass in response to training, in walk, 206–207, 206f, 211 Osteochondrosis, 254–257, 256t–257t
288–290 muscle activity and neck orientation, Osteochondrosis dissecans (OCD), 254–255,
innervation, 76 208–209, 208f–209f 256t–257t
metabolic changes, 284–286 osseous pathologies, 386 Osteons, 275
moment generation, 444, 445f pain, prevalence, 384–386 Over at the knee, 233
movements, 74–75 position, 211–212, 212f–214f, 220–222, Overo-lethal-white syndrome, 253
neck and back, 201–205, 204f 220f, 221t Overreach distance, 364
overtraining, 286 in collection, 361, 361f Over-reaching, 235
physiological adaptations, 286 impact on locomotion, 375–376 Overtraining, 286, 374–377, 375f
power, scaling, 460–462, 461f Neolithic period, 2 behavior problems in, 375
recruitment, 55 Nerve blocks impact of different training methods,
response to exercise, 278–290, 279t diagnostic, 192–193 375–376
suppleness, 358 low palmar, 148 welfare implications, 376–377
work, scaling, 460–462, 461f palmar digital, 148 Overtraining syndrome, 430
Muscle-twitch, 74 Nerve conduction velocities, 334 Oxidation, 420
Musculoskeletal injury/disorders, 267–268, Nerves of the neck and back, 201–205 Oxygen
382–384 Net joint force, 450 availability, 420
Musculoskeletal models, designing, 474–476, Net joint moment, 47, 447–450 debt, 429
475f see also specific joints deficit, 428
Muybridge, Eadweard, 7–12, 10f, 477–478 Net joint power, 47 postexercise consumption, 429
Myocytes, 75 Net total oxygen cost (NTOC), 429 supply, 422
Myofibrils, 74 Neural network see Artificial neural networks transport, 419
Myokinase reaction, 422 (ANNs) transport chain, 426, 426f

497
 Index

Oxygen uptake Phosphorylation Pyruvate dehydrogenase kinase isozyme 4,

aerobic power, 426–427, 426f aerobic, 420 mitochondrial gene (PDK4),
measuring, 432 anaerobic, 422 247
postexercise, 429 Photographic kinematic analysis systems, 39, Pyruvate dehydrogenase (PDH), 422–423
at rest, 426–427 40f Pyruvate kinase, 422
during submaximal exercise, 426–427 Physical Therapy and Massage for the Horse,
training programs, 432 382
Physiological cross sectional area (PCSA), 101
Q
Physiotherapy, 381–382 Quadratus lumborum, 205
P Physiotherapy: the science behind the profession, Quadriceps femoris, 128t, 130
P3 fracture, 166 382 Qualisys AB, 18–19
Pace, 91 Piaffe, 88, 324–325, 324t Quarter horses
Pacers, 311 Pie gait diagram, 85, 87f–88f effect of urging, 307
conformation, 311 Piezoelectric force transducers ground reaction force, 42
sidedness, 311 force plates, 43 hind limb musculature, 127–129
stride variables, 310t, 311 force shoes, 45–46 racing, 306–307
Pain scales, 391 Pirouette, canter, 326–327, 327f sidedness, 307
Palestinians, 3 Plaiting, 163 stride variables, 306–307, 307t
Palmar digital arteries, 148 Plantar digital arteries, 148
Palmar digital nerve, 148 Plantar digital nerve, 148
Palmar digital nerve blocks, 148 Plasma volume (PV), 433
R
Palmgren, A., 14–15 Plato, 3 Raabe, Captain, 7
Palo Alto, 9–10, 10f Play, 371, 371f Racing/racehorses, 305–312
Palpation scoring system, 391 Pliance system, 53, 53f efficiency of movement, 306, 306f
Parapatellar fibrocartilage, 131 Pliohippus, 369, 370f endurance racing, 311–312
Paso, 88 Pneumatic gate recorder, 90f history of, 17
Paso Finos, 88 Polyneural innervation, 76 hoof angle, 158–159
Passage, 88, 324, 324f, 324t Polysaccharide storage myopathy (PSSM), influence of training, 94
Passive markers systems, 479 253–254 injuries in, 383
Passive stretching techniques, 397–398 Popliteus, 128t Quarter Horses, 306–307, 307t
Pastern joint, 149f Positional release, 398 selection for performance, 246
conformation, 238, 240 Postural adjustments, 78–79, 81–82 speed, 305–306
segments, free body diagram, 445–446, Postural control development, 81 Standardbreds, 309–311
445f, 447f, 454f, 455 Postural sway analysis, 43, 44f pacers, 311
Pastoralism, advent of, 2 Pourcelot, Philippe, 21 trotters, 309–311
Patella Power stride frequency, 305–306
congenital luxation, 254 absorption, 451–452 stride length, 305, 306f
fixation, 168 calculations, 450–452, 450f Thoroughbreds, 307–309, 307f, 307t
locking mechanism, 131 external production, 451 Rack, 88
Pectineus, 128t flow, 451 Radescu, 12
Pectoral muscles, 101t, 102, 102f negative joint, 451 Radiographical analysis of head and neck
Peculiar lameness, 188–189 positive joint, 451 position
Pelagonius, 4 production, 451–452 Rapid prototyping, 468
Pelvic rotation, 134, 238 Power vs time-to-fatigue (P:TTF), 428–429, Ras-mitogen-activated protein kinase,
Pelvis, angle of, 238–239, 238f 429f 288
Pennation angle, 100–101 Pratt Jr, George, 23 Rathor, 16
Pentosidine, 268 Preconditioning, 292–293 Ratzlaff, Marc, 23
Performance Prehistoric times, 1–3 Reaction time in western sports, 334
effect of conformation, 232 Preoperative planning, 468 Rectus abdominis, 205–206, 209
effect of hoof on, 166 Pressure algometry, back pain diagnosis, at canter, 211
in equestrian sports, 305–340 390–394, 390f, 392t–393t at walk, 206
predicting Pressure mats, 52–53 Rectus capitis dorsalis major, 201–202
by conformation analysis, 241–242, 241f hoof, 52, 52f Rectus capitis dorsalis minor, 201–202
jumping horses, 332–333 saddle, 53, 53f Rectus femoris, 128t
selection for, 246–248 Pritchard, 17 Red cell volume (RCV), 433
Peroxisome proliferator-activated receptor PROP1, 258–259 Red nucleus, 78–79
gamma, coactivator 1 alpha gene Proportion differences, 457–459, 458t Reduction, 420
(PPARGC1A), 248 Proprioception, hoof/foot, 148, 153f Rehabilitation, 381–417
Persian culture, 3 ProReflex® system, 18, 20, 20f, 39–40 definition, 381–382
Persival mechanical horse simulator, 468, Protein, 425–426 exercise and facilitation techniques,
468f Proteoglycans (PGs), 268–269, 272 399–404
Peters, F., 7 Proximal interphalangeal (PIP) joint, 118f, historical background, 381–382
Pettigrew, J. B., 7 119, 121–122 longevity and musculoskeletal disorders,
Phalanges, 149f Przewalski-type horse, 1–2, 2f 382–384
Phase-plane analysis, 64 Psoas major, 128t, 205 manual therapy, 397–408
Phosphagen pool depletion, 430–431 Psoas minor, 128t, 205 mechanical and electro-physical
Phosphocreatine reaction, 422 Puissance jumping technique, 332 intervention, 405–408
Phosphofructokinase (PFK), 285, 420 Purkinje cells, 79 patho-anatomical diagnosis, 386–397

498
 Index

Reins, 345–347 joint angle, 108f

auxiliary, 345, 345f
S net moment, 106, 108f
contact, 358–359 SAAS (sacral acceleration asymmetry), 190 net power, 108f
tension, 345–347, 346f Sacroiliac lesions, 385–386, 401 Shoulders, freedom of the, 237–238
measurement of, 50, 50f–51f Sacrum, 200 Shung dynasty, 3
Relative weight bearing, 365–366 Saddle, 341–343 Sickle hocks, 239–240, 254
Relaxation, 358 effect on motion of the back, 342–343 Sidedness
Renaissance, 5–6 fit, 343 development of, 371–372, 372f
Research girths, 343 pacers, 311
centers and activities, 17–24 pads, 343 quarter horses, 307
revival of, 17 panels, 343 trotters, 310–311
Respiratory chain, 420 poorly fitted, 386 Side reins, 220
Respiratory coupling, 95–96, 95f pressure in standing horse, 341–342, 342f Signals, 61–71
Respiratory functionality in neck dysfunction pressure mats, 53, 53f definition of, 61, 62f
Retroreflective material, 33 saddle mat, 341 differentiation, 63–64, 64f
Rhabdomyolysis syndrome (RHA), 253 types, 343 definition of, 63
Rhett model, 482, 483f–485f Saddlery, Chinese design, 3 increasing, decreasing and finding local
Rhomboideus, 101t, 102, 102f, 201 Sagittal plane analysis, 37 maximum or minimum, 64
Rhythm(s) forelimb, 105–115 phase-plane analysis, 64
of activity, 372 hind limb, 131–142, 133t–134t physical concept of Newton, 63–64
gait, 357–358 Sahara Desert, 3 in time and frequency domain, 67,
Ribs, 199–200 Sampling frequency 67f–68f
Richter, O., 16 choosing the, 61–63, 62f processing, 63–69
Rider motion of a wheel, 61, 63f artificial neural network, 69
and back health, 222–223 resampling and normalization, 61–62 decision making, 69, 69f
basic movements, 347–353, 347f Sarcomeres, 74, 280 differentiation, 63–64
effect on dressage horses, 321–322 Sartorius, 128t differentiation in time and frequency
effect on jumping horses, 333 Saskatoon, studies in, 22 domain, 67, 68f
effect on sport horses, 320, 321f Satellite cells, 74 filtering, 66–69, 66f–67f
horse-rider interaction see Horse-rider Scaled energetics of locomotion, 455–463 Fourier analysis, 65–66, 65f–66f
interaction energy cost of locomotion, 459–460, fuzzy logic, 69
skill, 353–357 459f–460f, 460t integration, 64, 64f
in general, 353 geometric scaling, 455–457, 456f–457f, integration in time and frequency
the hand, 355, 356f–357f 456t domain, 67–68, 68f
horse movement in relation to, maximum aerobic capacity, 462–463 system analysis in time and frequency
355–357 muscle work and power, 460–462, 461f domain, 68–69
at trot, 353–354 proportion differences, 457–459, 458t time curve analysis, 63–65, 64f
at walk, 354–355, 354f Scalenus, 202–203 transformation of the coordinate system,
urging by, 307 Scalping, 235 65, 65f
weight of, 357 Scapula sampling frequency choice, 61–63
Riding methodology, 357–360 conformation, 236, 236f Silver, I. A., 21–22
asymmetry basics, 359–360 forelimb analysis, 106 SIMI Motion, 39
contact, 358–359 Schamhardt, Henk, 19–20 Simulation see Computer simulation
impulsion, 359 Schauder, Wilhelm, 12, 16 Single gene diseases, 253
rhythm, 357–358 Schmaltz, Reinhold, 12 Single nucleotide polymorphism (SNP)
straightness, 359 Schmidt, H., 12–13 database, 247
suppleness/relaxation, 358 Schtserbakow, N. M., 15 Skin
Riemersma, D.J., 19 Scythians, 2–3 displacement, 105, 131
Rims, 164 Seattle shoe, 161–162 markers, 33 (see also Markers)
Rinderpest, 6 Seeherman, Howard, 23 forelimb, 105, 106f
Rock art, 1–3, 2f Semimembranosus, 127, 128t, 129f, 130 hind limb, 131, 132f–133f
Rocker toe shoes, 167 Semispinalis capitis, 203 Slijper, E.J., 16
Rolled toe shoes, 167 Semitendinosus, 127, 128t, 129f, 130 Slow gait, 88
Rollkur, 21, 211, 375 Sensory integration, 402–404, 403f, 404t, Socrates, 3
Roman Empire, 4 405f Software packages, kinematic analysis,
Rooney, Jim, 23 Serratus ventralis cervicis, 101t, 102, 102f, 201 39–40
Rosette gauges, 47–48 Serratus ventralis thoracis, 101–102, 101t, 102f, Sole of the foot, 147
Rösiö, Birger, 12 201 Soleus, 128t
Rotary gallop, 91–92, 92f Sesamoid bone, 149f Somatostatin (SRIF), 434
Rotation, kinematic analysis, 37, 38f 17th Century, 6 Sonomicrometry, 49
Rotoscope, 478, 478f Shins, bucked, 309 Soundness
Royal Veterinary College, 6 Shivers, 53–54 effect of conformation, 232
RSFootScan, 52 Shoes see Horseshoes predicting, by conformation analysis,
Rubber bands, 345, 345f Short pastern bone, 147 241–242, 241f
Rubrospinal tract (RST), 78–79 Shoulder joint selection for, 250t–252t, 251–259,
Ruffus, Jordanus, 5 extensor moment, 106 253f
Ruini, Carlo, 5 forelimb analysis, 106, 108f Spavin see Bone spavin
Running walk, 88 in forelimb lameness, 179 Spavin flexion test, 192

499
 Index

Speed Stockholm, 17–19 Swing duration

event horses, 333–334 Stoss, A. O., 12 hind limb at walk, 139
racehorses, 305–306 Straightness, 359 shortening, in lameness, 176
scaling, 462f Straight sesamoidean ligament, 149f in sport horses, 312
Spherical markers, 33 Strain see Tissue strain standardization, 36–37
Spinalis cervicis, 203 Strain gauges, 47 Swinging limb lameness, 178, 182–183
Spinalis thoracis, 203 bone, 47–48, 48f Swing phase, 80
Spinal manipulation, 399 force plates, 43 hind limb at walk, 140
Spinal stability, 399 force shoes, 45 trot, 99, 110
Spinal system, 203 hoof mechanism, 156 walk, 112–113
Spinous processes, 199–200 hoof wall, 48, 48f Symbols, 455
Splenius, 102f, 202–203, 206–207 rein tension, 50 Symmetry indices, 190, 190t, 192
Spline, 67 tendons and ligaments, 48–49 Sympathetic-adrenal medullary (SAM) axis,
Spondylosis, 217–218 Stratul, J., 12 434
Sport horses, 312–320 Stretching Synsarcosis, 201
breed differences, 313–314 foals, 370 Synthesis, 82
conformation, 312 passive, 397–398
coordination, 315–316 Stride
dynamic similarity, 315 forelimb, 99
T
effect of a rider, 320, 321f variables in lameness, 176, 177t ‘T’ allele, 248
effect of training, 316–320, 319f, velocity-related changes in variables in, Tang dynasty, 3
320t 93–94, 93f Tarsal joint, 135–136, 137f
forelimb kinematics, 312, 313f–314f Stride duration degenerative joint disease, 254
hind limb kinematics, 312–313 hind limb at walk, 139 in forelimb lameness, 179, 180f
linear similarity, 315 influence of rider, 357 lameness, 187–188
normalization of temporal variables, 315, in piaffe, 324 three-dimensional kinematics at trot, 142,
315f, 316t–318t in sport horses, 312 142f–143f
ontogeny of gait, 314–315, 314f Standardbreds, 311 Tarsocrural joint, 142
quality of movement, 312 standardization, 36–37 TCA cycle, 422–423
stride and swing duration, 312 Stride frequency, 93–94, 93f Temperature, effect on oxygen uptake, 427
Stability in western sports, 334 changes in gallop, 94f Temporal data, 37
Stabilogram, 43, 44f racehorses, 305–306 Temporal stride pattern, 176–178
Stables, 373 Stride length, 93, 93f Tendinitis, 188
Stallion testing, 317 changes in gallop, 94f Tendonitis, 175
Stalls, 373 endurance horses, 311–312 Tendons
Stance duration racehorses, 305, 306f biomechanics, 292–293, 293f
ground force reaction, 43 Standardbreds, 311 blood vessels, 291–292
hind limb at walk, 139 in trot, 323 collagen fibrils, 268, 268f, 272–273
lengthening, in lameness, 176, in walk, 322 collagen in, 291, 294, 294f
177t Stringhalt, 53–54, 189, 189f composition and structure, 291–292, 292f
standardization, 36–37 Strubelt, 12 conditioning, 291
trot, 323 Subchondral bone (SCB), 274 crimp, 291, 294
Stance phase, 80 Subclavius, 101t, 102 cyclical loading, 292–293
sport horses, 313 Sukhanov, V. B., 17 effects of ageing on, 294
thoroughbred racehorses, 308 Superficial digital flexor muscle effects of exercise on, 293–295
trot, 99, 110–111 forelimb, 104–105, 104t forelimb, 99–105
trotters, 310 hind limb, 127, 129f injury, 291, 383
walk, 112–113 tendinitis, 188 load-elongation curve, 48
Standardbreds Superficial digital flexor tendon, 149f measuring strains in, 48–49, 50f
conformation, 232 contraction, 168 response to exercise, 290–296
influence of training, 94 effect of exercise on, 295 strains, 292, 308–309
racing, 309–311 injuries, 291 stress, 292
see also Pacers; Trotters wedges and, 164 see also specific tendons
Stanford, Leland, 9–10 Suppleness, 358 Tenoblasts, 291
Static optimization models, 449 Supporting limb lameness, 176–182 Tenocytes, 291
Statistical analysis/tests, 31–32 Supraspinatus, 103 Tensor fascia lata, 128t, 129f, 130
choosing, 69, 69f Supraspinous ligament, 201 Terrestrial locomotion, 11
computer simulation, 476 Surface electromyography (EMG), Tetanic contraction, 74
Stay apparatus, hind limb, 130–131 54 Thalamic nuclei, 79
Steinmann pins, 20, 33 Surface type, 165 Thermoregulation, 435–436
Step cycle, 78, 80–81 Suspension phase, 176–177, 177t Thoracic sling, 101–102
Stepping pace, 88 Suspensory ligament, 149f Thoracolumbar spine, 399–400
Stereotypical behaviors, 374 forelimb, 105 lesions, 384–385
Sternal lift, 401–402, 402f hind limb, 127 movements, 208f–209f, 209
Sternocephalic muscle, 202–203 strain, 193 Thoroughbreds
neck elevation, 208 Sweating, 435 acceleration, 308
at walk, 206–207 Swedish Equine Biomechanics Research conformation, 232
Stifle joint, 135, 136f, 239 Group, 18, 18f fatigue, 309

500
 Index

forelimb motion, 115 Transport three-dimensional kinematic analysis of

genetics, 247–248, 249f role of horse, 1 tarsal joint, 142, 142f–143f
ground reaction forces, 308 welfare implications, 376 transitions
head and neck motion, 308 Transverse gallop, 91–92, 92f from canter, 92–93, 327
hind limb musculature, 127–129 Transverse processes, 199–200 from walk, 327
influence of training, 94 Transversospinalis system, 203 to walk, 92–93, 327
injury, 309 Transversus abdominis, 399–400 variables changes with training stages,
lead changes, 308 Trapezius, 101t, 102, 102f 90t
limb kinetics, 308–309 Traquenard, 85–87, 91 working, 88, 323, 323t
racing, 307–309 Treadmills Trunk
stride parameters, 94 energy expenditure measurement, 431–432 acceleration, 51
stride variables, 307–308, 307f, 307t evaluation, 32 motion, markers for, 33–34
track surface, 309 habituation, 32, 32f movement
training, 309 force plates in, 44–45, 45f in canter, 209–211, 210f
Three-dimensional analysis, 34, 37–38, 38f influence on gait, 94–95, 95f in lameness, 180–181, 182f, 185–186
digital joints, 119t, 120 mounted studies, 32 in trot, 207–209, 207f–209f, 211
forelimb, 115–122 Tricarboxylic acid (TCA) cycle, 422–423 in walk, 206–207, 206f, 211
inverse dynamic analysis, 453–454 Triceps brachii, 103–104, 106 muscle activity, 208f–209f, 209
tarsal joint at trot, 142, 142f–143f Trigger points, 407 muscles, 75, 82
trot, 116–120, 119t, 120f Triglycerides, 424–425 Twitch, 407
walk, 116–122, 119t Triose phosphate isomerase (TPI), 420
Tibialis caudalis, 128t, 130 Tropocollagen, 291
Tibialis cranialis, 128t, 130 Troponin, 74
U
Time curve analysis, 63–65 Trotters, 309–311 UK see Great Britain
Time domain-frequency domain, 65–66, 66f conformation, 311 Ulnaris lateralis, 104, 104t
differentiation in, 67, 67f–68f flying trot, 91, 91f Ultrasonography
integration in, 67–68, 68f ground reaction forces, 310 back pain diagnosis, 394–397, 395f–396f
system analysis in, 68–69 reproducibility of gait, 310 B-mode, 396
Tissue strain, 47–52 rhythm irregularities, 91 kinematic analysis systems, 41, 41f
accelerometers, 50–52, 51f sidedness, 310–311 M-mode, 396
calculation of, 49–50 stride variables, 309–310, 310t therapeutic, 406
gyroscopes, 50–52 track design, 311 Ultrasound probes, tendon strain, 49, 50f
inertial sensors, 50–52 trot in, 91 Underrun heels, 149–150
magnetometers, 50–52 Trot/trotting, 88–91 Uneven feet, 242, 242f, 251–253, 253f
measurement of rein tension, 50, 50f–51f asymmetry, 359–360 Ungual cartilages, 147–148
measuring classification, 86t United Kingdom see Great Britain
in hard tissues, 47–48, 48f collected, 88, 323, 323t–324t, 325f, United States, 16, 22–23
in tendons and ligaments, 48–49, 50f 361–364, 363t–364t Uppsala, 17–19
sonomicrometry, 49 dressage horses, 323, 323t–324t, 325f Urging by rider, 307
strain transducers, 47 extended, 88, 323, 323t Utrecht, 19–20
Toe axes deviations, 232–233, 232f, 233t flying, 91, 91f
Toe clips, 164 footfall sequence, 87f–88f, 90f
Toe grabs, 164 forelimb, 106–111, 107f, 116–120, 117f–118f
V
Toe-in conformation, 233 foxtrot, 91 Valgus deviation, 168, 169f
Toelt, 88, 90f ground reaction force, 42–43 Van den Bogert, A. J., 19
Toe-out conformation, 232–233 in harness trotters, 91 Van der Plank, 14–15
Tokuriki, M., 24 heritabilities, 88–91, 90t Van Weeren, 19
Total blood volume (TBV), 433 hind limb, 127, 133–138, 141f Variability, 31–32
Townsend, Hugh, 22 kinematic analysis, 106–111 Vastus intermedius, 128t
Track design/surface, 309, 311 three-dimensional, 116–120, 119t, 120f Vastus lateralis, 128t
effect on oxygen uptake, 427 kinematic variables, 114t Vastus medialis, 128t, 131
TrackEye® system, 18, 39 load carrying at, 357 Vectordynamograms, 191, 191f
Training medium, 88, 323, 323t Velocity, 36–37
effect on gait, 94 neck and back movement in, 207–209, calculations, 455, 455f
effect on jumping horses, 332–333, 333f 207f–209f, 211, 217–218 changes in gallop, 94f
effect on muscles, 286–287, 287f pneumatic gate recorder, 90f forelimb, 112
impact of different methods on locomotion, rhythm irregularities, 91 hind limb, 140
375–376 rider movements at, 347–348, 348f in lameness, 176
locomotory responses, welfare implications, rider skill at, 353–354 related changes in stride variables, 93–94,
376–377, 377f rising 93f
metabolic energetics, 431–435 asymmetry in, 359–360, 360f walk, 112, 140
overtraining see Overtraining rider movements at, 348–349, 354 Ventral rectus capitis, 201–202
sport horses, 316–320, 319f, 320t sitting, rider movements at, 347–348, 348f, Vertebrae, 199, 200f
thoroughbred racehorses, 309 354 cervical, 199
Trait mapping, 246–248 in sport horses, 312, 313f coccygeal, 200
Transcutaneous electrical nerve stimulation stance duration, 323 lumbar, 200
(TENS), 406–407 in a straight line, 116–120, 119t sacral, 200
Transitions, 92–93, 327 stride length, 323 thoracic, 199–200

501
 Index

Vertebral joints, 201 half pirouette at, 323 Wehner, R., 12–13, 17
Vestibular disease, 188 heritabilities, 88, 90t, 322 Weight bearing, relative, 365–366
Vestibular nuclei, 79 hind limb, 134t, 138–140, 141f Welfare
Vestibulum development, 81 kinematic analysis, 111–113 breeding/training locomotory responses,
Veterinarians, Roman Empire, 4 three-dimensional, 116–122, 119t 376–377
Veterinary College of Philadelphia, 6 kinematic variables, 114t effect on management regimes on,
Veterinary education load carrying at, 357 373–374
19th century, 6–15 medium, 322 impact of exercise on, 374
start of, 6 neck and back movement in, 206–207, Wentink, G. H., 19
Veterinary medicine, antique cultures, 1, 3 206f, 211, 217–218 Western sports, 334–335
Vicon MOTUS, 39 rider movements at, 349–353 cutting, 335, 335t
Vicon MX, 39–40 lateral and rotational, 350–353, gaits, 335, 335t
Video analysis, head and neck position, 351f–352f maneuverability, 334
215 sagittal plane, 349–350, 349f, 350t, 351f reaction time, 334
Video cameras, 39 rider skill in, 354–355, 354f stability, 334
Videography, 39 running, 88 Wheel motion, 61, 63f
Vienna, 19 in a straight line, 116–120, 119t White line, 147
Vincent, 6–7, 7f–8f stride length, 322 Wiechert, F., 12–13
Virtual reality, 468, 468f transitions Wilson, Alan, 21–22
Visible Human Project, 468 from canter, 92–93 Winging, 163
Visualization, computer models, 467–468 to canter, 93f Withers
Von Hochstätter, Conrad, 7 to halt, 92–93 height at, 234–235, 235t
Von Lengerken, 17 from trot, 92–93, 327 in lameness, 180–181
Von Lettow-Vorbeck, First Lieutenant, 16 to trot, 327 Wobbler disease, 188, 254
turning sharply at, 120–122 Wu, emperor, 3
velocity, 112
W Walter, K., 12, 16
WAAS (Withers Acceleration ASymmetry), 190 Warfare, 15–17
X
Wagener, H., 12–13 Warmbloods Xenophon, 4, 245
Walk/walking, 88 conformation, 246 Xsensor, 53
in adult animals, 80 selection for dressage, 89–91
classification, 86t as sport horses, 313–314
collected, 88, 322, 361, 362f, 362t Water immersion, 405
Z
dressage horses, 322, 322f, 322t Water jumps, 332 Zebris system, 41, 41f
extended, 88, 322 Weaving, 374 ZNF346, 258–259
footfall sequence Wedges, 164 Zoöpraxiscope, 10–11
forelimb, 111–113, 116–120, 117f–118f heel/toe, 164, 165f Zurich, 21
free, 322 in navicular syndrome, 167 Zurich, studies in, 22f
ground reaction force, 42–43, 112 side, 164 Zwaenepoel, M., 14–15

502