Respiratory Medicine (2012) 106, 243e248

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/rmed

Reference values for the incremental shuttle

walking test
Vanessa S. Probst a,b,*,d, Nidia A. Hernandes a,b,d, Denilson C. Teixeira a,
Josiane M. Felcar a, Rafael B. Mesquita a, Cristiane G. Gonçalves a,
Daniela Hayashi a, Sally Singh c, Fabio Pitta b

a
Centro de Pesquisa em Ciências da Saúde (CPCS), Centro de Ciências Biológicas e da Saúde (CCBS),
Universidade Norte do Paraná (UNOPAR), Av Paris 675, Jd Piza, CEP: 86041-120, Londrina, PR, Brazil
b
Laboratório de Pesquisa em Fisioterapia Pulmonar (LFIP), Departamento de Fisioterapia,
Universidade Estadual de Londrina (UEL), Av Robert Koch 60, Vila Operaria, CEP: 86038-350, Londrina, PR, Brazil
c
University Hospitals of Leicester NHS Trust, Groby Road, Leicester LE3 9QP, United Kingdom

Received 27 May 2011; accepted 30 July 2011

Available online 23 August 2011

KEYWORDS Summary
Incremental shuttle Background: Reference values for the incremental shuttle walking test (ISWT) which are appli-
walking test; cable to the whole population need to be solidly established. This study aimed to determine
Reference value; which anthropometric and demographic variables influence the walking distance achieved in
Statistical regression the ISWT in healthy subjects with a broad age range and to establish a reference equation
for predicting ISWT for that population.
Methods: In a cross-sectional study, 242 healthy subjects (102 male) performed two ISWT and
had their weight, height and body mass index (BMI) measured.
Results: In general, healthy subjects walked 810 [IQR 25e75%: 572e1030] m in the ISWT,
presenting large variability (range 210e1820 m). The walked distance correlated with age (r Z

0.76), height (r Z 0.49) and BMI (r Z
0.23) (p < 0.001 for all), but not with weight
(r Z 0.06, p Z 0.315). A model of stepwise multiple regression showed that gender, age and
BMI were independent contributors to the ISWT in healthy subjects, explaining 71% (p < 0.0001)
of the variability. The derived reference equation was: ISWTpred Z 1449.701
(11.735 
age) þ (241.897  gender)
(5.686  BMI), where male gender Z 1 and female gender Z 0.
Conclusion: In conclusion, the variability of the ISWT is explained largely by gender, age and BMI.
The reference values for the ISWT can be adequately predicted using the equation proposed in this
study.
ª 2011 Elsevier Ltd. All rights reserved.

* Corresponding author. Centro de Pesquisa em Ciências da Saúde (CPCS), Centro de Ciências Biológicas e da Saúde (CCBS), Universidade
Norte do Paraná (UNOPAR), Av Paris 675, Jd Piza, CEP: 86041-120, Londrina, PR, Brazil.
E-mail address: vanessaprobst@uol.com.br (V.S. Probst).
d
These authors contributed equally to this article.

0954-6111/$ - see front matter ª 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.rmed.2011.07.023
244 V.S. Probst et al.

Introduction Committee of the Universidade Norte do Paraná (UNOPAR),

Brazil, and all participants gave written informed consent.
Exercise tests are commonly used both in clinical practice and Inclusion criteria were: subjects of both genders aged
in scientific investigations since exercise intolerance is 18e83 years and absence of any severe and/or unstable
a typical feature in a number of chronic diseases. The disease which could limit the exercise tolerance. Partici-
objective assessment of exercise capacity through a cardio- pants were excluded if they were unable to understand or
pulmonary exercise test provides information about the perform any procedure during the protocol or if they would
causes of reduced exercise tolerance, which can be eventu- like to leave the study for any reason.
ally improved by specific interventions.1 Moreover, outcome After obtaining the informed consent, a questionnaire
parameters of exercise tests are used as guide for workload was applied in order to investigate health status, medica-
prescription during exercise interventions.2 As laboratory tion, smoking habits and whether subjects were engaged in
assessment is not widely available and may be expensive, any regular physical activity. Height (cm) and body weight
field tests such as the incremental shuttle walking test (ISWT) (Kg) were measured and the body mass index (BMI) was
have become increasingly popular. calculated. Spirometry (Pony Cosmed, Italy) was performed
The ISWT is a simple and inexpensive test which evalu- to ensure normal lung function. The test was conducted
ates maximal exercise capacity based on the distance according to international standardization13 and the lung
walked around a 10 m course according to different speeds function parameters obtained were forced vital capacity
dictated by an audio signal.3 This exercise test has been (FVC), forced expiratory volume in the first second (FEV1)
used as a predictor of mortality4 and morbidity5,6 in and the direct measurement of maximal voluntary venti-
patients with chronic respiratory and other diseases, as lation (MVV). Reference values adopted were those repor-
a predictor of exacerbation in patients with pulmonary ted by Pereira and colleagues specifically for the Brazilian
disease,7 as well for assessing benefits of interventions8e10 population.14
and detecting oxygen desaturation during exertion.11
Although the ISWT has become very useful in clinical and Incremental shuttle walking test
research settings, there were no reference equations to
predict the distance walked in the test by healthy subjects. Two incremental shuttle walking tests were performed
Recently, an equation was developed12; however, there are with, at least, 30 min of resting in between them. The tests
some limitations which can compromise the external val- were performed in a 10 m course identified by two cones
idity of that study. Firstly, the sample was composed of placed 0.5 m from each end point.3 The best test, that is,
subjects aged 40e84 years; therefore the resulting equa- the longest walked distance was considered for analysis.
tion is not applicable for young people. Moreover, the Participants should walk (or run) around the course
coefficient of determination of its regression analysis was according to the speed dictated by an audio signal. The
relatively modest, explaining only 50% of the total variance initial walking speed was 0.5 m/s and it increased by
in the ISWT. Considering the importance of the ISWT in 0.17 m/s each minute; the speed increment was always
assessing adequately patients of all ages with cardiopul- indicated by a triple bleep. An adaptation of the modified
monary disease and other health conditions, an equation protocol was used, that is, the audio signals continued until
with higher coefficient of determination and applicable for subjects reach their maximal effort, exceeding 12 levels of
the whole adult population needs to be solidly established. speed proposed by the modified protocol and even running,
Hence, the aims of this study were to determine which if necessary. The authors opted for adapting the protocol in
anthropometric and demographic variables (weight, height, order to avoid a ceiling effect since participants were
age and gender) influence the walking distance in the ISWT healthy subjects and could possibly exceed the 12th level in
of healthy subjects with a broad age range and to establish order to ensure their maximal effort. The tests were
an equation for predicting reference values of the ISWT for executed by a physiotherapist or physiotherapy student, all
this population. familiarized with the ISWT, and the two tests were con-
ducted by the same evaluator. The initial explanation was
standardized and no encouraging phrases were given to the
Methods and materials participants during the test. The ISWT was interrupted in
case of participants presented one of the following condi-
Subjects tions: if the subject could not maintain the required speed
due to dyspnea or fatigue; or if the subject failed to
complete a shuttle in the time allowed for the second time.
In a cross-sectional study, 246 healthy subjects were included
Heart rate (HR), arterial blood pressure, perceived dyspnea
in a convenience sample. They were recruited among
and leg fatigue (modified Borg scale) were assessed before
students and employees of two universities in Londrina
and after the tests. Maximal predicted HR was calculated as
(Brazil), as well as their relatives. Part of the sample of
220 minus age (in years).15
subjects older than 60 years old was also composed by indi-
viduals participating in a project which investigates the
health conditions of the elderly in Londrina, Brazil (EELO Statistical analysis
Project). All subjects had their exercise capacity, pulmonary
function, anthropometric and demographic data evaluated. The statistical analysis was performed using the statistical
Data were collected from November 2008 to December 2010. packages SPSS 17.0 (SPSS Inc., USA) and GraphPad Prism
The study was approved (PP000709) by the Research Ethics 5 (GraphPd Software Inc., USA). The normality of data
Reference values for the ISWT 245

distribution was checked by the ShapiroeWilk test. Data health condition presented by the individual. Among the
were described as median [interquartile range 25%e75%]. medications in use, 45% were drugs known to affect blood
Wilcoxon test was used to compare outcome parameters of pressure and/or heart rate. Approximately 30% of the
the ISWT. ManneWhitney test was used to compare subjects’ subjects were current smokers. Regarding the physical
characteristics between male and female genders. Spearman activity level, 61% of participants were not engaged in
correlation coefficients were calculated and a model of regular physical activity.
multiple linear regression was applied (ISWT distance as
dependent variable; demographic and anthropometric data Variability of the ISWT
as independent variables) since its standardized residual
presented normal distribution (p Z 0.785). The Bland & Considering the best of the two ISWT, subjects walked 810
Altman plot was used to evaluate agreement between the [572e1030] m, presenting large variability (range of
actual ISWT distance and the predicted value. In order to 210e1820 m). Twenty-eight percent of subjects exceeded
verify the reliability of the regression equation, it was the 12th level of the protocol, presenting a walked distance
applied a posteriori in a different group of healthy subjects, higher than 1020 m. Subjects walked a median of 770
composed by 23 individuals selected according to the same [530e1010] m in the first ISWT and 790 [550e1030] m in the
inclusion criteria of the present sample. The level of statis- second ISWT (p < 0.0001) (Table 2). The distance walked
tical significance was considered as p < 0.05. during the second ISWT increased by 20 [
10e70] m
(increment of 3%); 63% of subjects had their best ISWT in
Results the second test and 28% in the first test, whereas 9% of the
individuals had the exact same distance in the two tests.
When comparing genders, male walked farther than female
The study sample included 103 male and 143 female
(1010 [755e1200] m vs. 720 [480e910] m, p < 0.0001,
apparently healthy subjects. Four subjects (1 man and 3
respectively). In the best test, subjects reached 99
women) were excluded because they interrupted the test
[89e105]% of their maximal predicted HR.
abruptly due to severe joint pain. Therefore, the results of
242 subjects are presented. The subjects’ age ranged from
18 to 83 years. They presented normal lung function and, in ISWT determinants and reference equation
general, a normal body composition. As expected, men
were taller and weighed more than women, as well as There were significant correlations between the distance
presented higher distance walked in the ISWT in absolute walked in the ISWT with age (r Z
0.76), BMI (r Z
0.23)
values (Table 1). and height (r Z 0.49) (p < 0.001 for all), but not with
weight (r Z 0.06, p Z 0.315). A model of stepwise multiple
regression showed that gender, age and BMI explained 71%
Health status, medication, smoking habits and
(p < 0.0001) of the variability in the ISWT. Unstandardized
physical activity level coefficients, part correlation and significance are shown in
Table 3. The reference equation for the distance walked in
The comorbidities more prevalent among the study the ISWT was:
subjects were systemic hypertension (34%), followed by
peripheral vascular disease (24%) and arthritis (22%). Other ISWTpred Z1449:701
ð11:735  ageÞ þ ð241:897  genderÞ
reported comorbidities were: stable cardiac disease (13%),

ð5:686  BMIÞ
diabetes mellitus (13%), thyroid disorders (8%), osteoporosis
(6%) and asthma (6%). One hundred and forty three (where male gender Z 1 and female gender Z 0).
participants (41% among those were elderly) needed to take The Bland & Altman plot shows the agreement between
some medication continuously. The medications were the actual ISWT and the predicted value obtained from the
prescribed by physicians and were related to the specific reference equation (Fig. 1). A weak but significant

Table 1 Characteristics of the study subjects.

Total group (n Z 242) Male (n Z 102) Female (n Z 140) p-value
Age (yrs) 50 [31e66] 49 [30e69] 51 [33e66] 0.989
Height (m) 1.64 [1.58e1.71] 1.71 [1.66e1.75] 1.59 [1.55e1.64] <0.0001
Weight (Kg) 69 [60e79] 75 [68e85] 64 [57e72] <0.0001
BMI (kg m
2) 26 [23e29] 27 [23e29] 25 [21e29] 0.146
FEV1/FVC 89 [83e95] 87 [82e93] 89 [85e99] 0.044
FEV1 %pred 97 [89e106] 96 [89e106] 97 [89e106] 0.793
FVC %pred 94 [85e104] 92 [86e103] 94 [85e104] 0.772
MVV %pred 98 [81e110] 103 [89e113] 93 [77e105] <0.0001
ISWT (m) 810 [572e1030] 1010 [755e1200] 720 [480e910] <0.0001
Data are expressed as median [interquartile range 25e75%]. P-value concerns the difference between male and female subjects. BMI:
body mass index; FEV1: forced expiratory volume in the first second; FVC: forced vital capacity; MVV: maximal voluntary ventilation;
ISWT: incremental shuttle walking test.
246 V.S. Probst et al.

Table 2 Outcome parameters of the two incremental shuttle walking tests.

ISWT 1 ISWT 2 p-value
ISWT (m) 770 [530e1010] 790 [550e1030] <0.0001
% HR max (%) 96 [84e102] 99 [90e105] <0.0001
SBP before (mmHg) 120 [110e130] 120 [110e130] 0.483
SBP end (mmHg) 160 [150e180] 165 [150e190] 0.959
DSBP (mmHg) 40 [20e60] 40 [30e60] 0.613
DBP before (mmHg) 80 [70e80] 80 [70e80] 0.908
DBP end (mmHg) 80 [80e90] 80 [80e90] 0.862
DDBP (mmHg) 10 [0e10] 0 [0e10] 0.863
Borg D before (pts) 0 [0e0] 0 [0e0] 0.128
Borg D end (pts) 4 [2e6] 4 [3e6] 0.026
DBorg D (pts) 4 [2e5] 4 [2e6] 0.190
Borg F before (pts) 0 [0e0.5] 0 [0e0.5] 0.541
Borg F end (pts) 2 [0e4] 3 [0e4] 0.199
DBorg F (pts) 1 [0e4] 2 [0e4] 0.400
Data are expressed as median [interquartile range 25e75%]. ISWT: incremental shuttle walking test; HR: heart rate; SBP: systolic blood
pressure; DBP: diastolic blood pressure; Borg D: Borg dyspnea scale; Borg F: Borg fatigue scale.

correlation was observed between the difference of the demographic variables which can be easily applied in clin-
ISWT (actual e predicted) and the mean of the ISWT (actual ical and research settings.
and predicted) (r Z 0.28, p < 0.0001). Recently, a Brazilian research group has published
Since the mean  standard deviation distance walked in a reference equation for the ISWT.12 However, there were
the ISWT was 833  273 m, the lower limit of normal some limitations in their study which could limit the use of
calculated according to the confidence interval was 798 m. the equation. Firstly, the analysis did not include subjects
younger than 40 years old, which is a limitation since some
Reliability of the reference equation cardiopulmonary diseases affect subjects in this age range.
Secondly, the regression coefficient was modest (50%),
The characteristics of the other group composed by 23 increasing the chance of bias in predicting the walking
healthy subjects (10 male) included in the a posteriori distance in the ISWT. In fact, this was confirmed when that
analysis were: age 49  20 years, BMI 24  3.8 kg m
2, FEV1 equation12 was applied in our sample studied a posteriori.
93  15%pred and FVC 91  17%pred. When the reference The predicted value was greatly underestimated when
equation was applied in this group, there was no difference compared with the actual ISWT (579  173 m vs. 839  269 m,
between the actual and predicted ISWT (839  269 m vs. p < 0.0001, respectively), which corresponds to a 45%
838  271 m, respectively; p Z 0.970). Furthermore, there difference. This discrepancy did not happen when the
was a strong correlation between the actual and predicted equation from the present study was used. Finally, those
ISWT (r Z 0.85, p < 0.0001). authors considered for the statistical analysis only the
second test instead of the best test. This approach could
definitely have influenced the results since it was observed in
Discussion our study that almost one in each three subjects was able to
walk farther during the first ISWT.
The present study demonstrated that healthy subjects aged The present results demonstrated that gender, age and
from 18 to 83 years present large variability in the ISWT BMI influence the ISWT in apparently healthy subjects.
(210e1820 m), and that 71% of this variance is explained by These findings are in agreement with previously published
age, gender and body mass index. Moreover, a reference data about walking speed. In a study with 42 male subjects,
equation was established using anthropometric and Pearce et al.16 affirmed that age alone was associated with

Table 3 Multiple linear regression analysis with ISWT as the dependent variable.
Unstandardized 95% Confidence p-value Part correlation
coefficients (B) interval for B
Constant 1449.701 1316.884e1582.519
Age (yrs)
11.735
12.867 to
10.603 <0.0001
0.70
Gendera 241.897 198.491e285.302 <0.0001 0.38
BMI (kg m
2)
5.686
10.832 to
0.539 0.031
0.07
Residual standard deviation Z 167.8 m.
The derived equation from the regression analysis was:
ISWTpred Z 1449.701
(11.735  age) þ (241.897  gender)
(5.686  BMI).
a
Male gender Z 1 and female gender Z 0.
Reference values for the ISWT 247

ISWT correcting values for gender, age and BMI allows

interpreting the results adequately taking into account
each subject’s characteristics, enabling more reliable
intergroup comparisons.
Although a reference equation for the ISWT has been
recently published, some points in the present study update
those findings. Firstly, our sample included a higher number of
subjects (242 vs. 131 subjects) and ages from 18 to 83 years,
being more representative of the vast majority of cardiopul-
monary patient’s population. Secondly, in our analysis, the
anthropometric and demographic variables could explain
a larger part of the variance in the ISWT (71%). Thirdly, our
sample reached a higher HR in terms of % predicted of the
Figure 1 Bland & Ajtman plot of the difference between the
maximal HR, suggesting more effort during the test.
actual and predicted value of the ISWT plotted against the mean
Despite all efforts, some limitations occurred. A few
of the actual and the predicted value of the ISWT for the entire
variables which could also explain the variance in the ISWT
subject group. The central dotted line corresponds to the
were not assessed, such as peripheral muscle force and
average difference between the actual and predicted value of
balance. On the other hand, this study aimed at estab-
the ISWT, whereas the lower and upper dotted lines corresponds
lishing a reference equation using independent variables
to lower and upper limits of agreement, respectively.
which are easily available in clinical settings. Another
limitation was the use of a convenience sample, although
walking speed. Himann et al.17 also found that age was the caution was taken concerning the number of participants in
only significant independent predictor for walking speed each age range, as well as the proportion of male and
when 438 subjects aged 19e102 years walked over an 80-m female subjects. Finally, it can be speculated that the
indoor course. Bohannon18 studied 1923 subjects who had subjects from the present study could have performed even
their gait speed evaluated when walking in a corridor, and better if a third ISWT was performed instead of two.
showed that younger, male and taller individuals walk However, it has been shown that a third ISWT had no
faster. Recently,19 the same author observed that body benefit in terms of walking distance in patients following
composition also influences the walking ability in elderly artery bypass.24 Therefore, we hypothesized that this could
people, demonstrating that higher BMI was significantly also be the case for people without cardiac disease.
associated with limited walking (r Z
0.42). A negative Nevertheless, the necessity of performing a third ISWT still
association between adiposity and walking speed in older remains to be completely elucidated in healthy individuals.
women was also observed by Ortega-Alonso et al.20
Previous studies have shown that patients reach higher Conclusions
peak HR during the ISWT than during the 6-min walking
test, characterizing the ISWT as a maximal exercise In summary, it can be concluded that variability in the
test.3,21 Furthermore, in our study, the peak HR was even performance of the incremental shuttle walking test in
higher than those reported previously,3,21 since partici- healthy people can be explained (71%) by using age, gender
pants achieved 99% of their maximal predicted HR. It can and body mass index. A reference equation could be
be attributed to the fact that our sample was composed of established based on these anthropometric and demo-
healthy subjects, who do not present any ventilatory or graphic variables. Standardization of the ISWT is necessary
cardiovascular limitation, differently from the subjects to ensure maximal effort during the test in order to avoid
included in those studies. In contrast, Jürgensen et al.12 incorrect interpretation of results when considering refer-
reported that healthy subjects reached only 78% of their ence values.
maximal predicted HR when performing the ISWT, which
represents 21% less in comparison to the healthy subjects
from the present study. This possibly represents a meth- Acknowledgments
odological issue in the study of Jürgensen et al., since it
might be that those subjects did not perform their maximal The authors are grateful to physiotherapy students from
effort during the ISWT. the research laboratory for contributing to collection data.
The majority of subjects increased their walking
distance during the second test with an average improve-
ment of 20 m. This is in accordance with results previously
Conflict of interest
reported, in which the magnitude of the improvement in
the second test ranges from 20 to 40 m.3,22,23
Authors have no conflict of interest to disclose.

Strengths and limitations References

The clinical importance of our findings is evident since the 1. Palange P, Ward SA, Carlsen KH, Casaburi R, Gallagher CG,
use of absolute values in assessing patients usually induces Gosselink R, et al. Recommendations on the use of exercise
bias in the results. The use of a reference equation for the testing in clinical practice. Eur Respir J 2007;29:185e209.
248 V.S. Probst et al.

2. Nici L, Donner C, Wouters E, ZuWallack R, Ambrosino N, 13. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R,
Bourbeau J, et al. American Thoracic Society/European Coates A, et al. Standardisation of spirometry. Eur Respir J
respiratory Society statement on pulmonary rehabilitation. Am 2005;26:319e38.
J Respir Crit Care Med 2006;173:1390e413. 14. Pereira CAC, Neder JA. Diretrizes para testes de função pul-
3. Singh SJ, Morgan MD, Scott S, Walters D, Hardman AE. Devel- monar. J Bras Pneumol 2002;28:S2e237.
opment of a shuttle walking test of disability in patients with 15. Karvonen MJ, Kentala E, Mustala O. The effects of training on
chronic airways obstruction. Thorax 1992;47:1019e24. heart rate; a longitudinal study. Ann Med Exp Biol Fenn 1957;
4. Ringbaek T, Martinez G, Brondum E, Thogersen J, Morgan M, 35:307e15.
Lange P. Shuttle walking test as predictor of survival in chronic 16. Pearce ME, Cunningham DA, Donner AP, Rechnitzer PA,
obstructive pulmonary disease patients enrolled in a rehabili- Fullerton GM, Howard JH. Energy cost of treadmill and floor
tation program. J Cardiopulm Rehabil Prev 2010;30:409e14. walking at self-selected paces. Eur J Appl Physiol Occup
5. Geddes D, Davies M, Koyama H, Hansell D, Pastorino U, Pepper J, Physiol 1983;52:115e9.
et al. Effect of lung-volume-reduction surgery in patients with 17. Himann JE, Cunningham DA, Rechnitzer PA, Paterson DH. Age-
severe emphysema. N Engl J Med 2000;343:239e45. related changes in speed of walking. Med Sci Sports Exerc
6. Murray P, Whiting P, Hutchinson SP, Ackroyd R, Stoddard CJ, 1988;20:161e6.
Billings C. Preoperative shuttle walking testing and outcome 18. Bohannon RW. Population representative gait speed and its
after oesophagogastrectomy. Br J Anaesth 2007;99:809e11. determinants. J Geriatr Phys Ther 2008;31:49e52.
7. Emtner MI, Arnardottir HR, Hallin R, Lindberg E, Janson C. 19. Bohannon RW. Body mass index and mobility of older home
Walking distance is a predictor of exacerbations in patients care patients. Physiother Theory Pract; 1 November 2010. doi:
with chronic obstructive pulmonary disease. Respir Med 2007; 10.3109/09593985.2010.523058. Published online first.
101:1037e40. 20. Ortega-Alonso A, Sipila S, Kujala UM, Kaprio J, Rantanen T.
8. Evans RA, Singh SJ, Collier R, Williams JE, Morgan MD. Pulmo- Body fat and mobility are explained by common genetic and
nary rehabilitation is successful for COPD irrespective of MRC environmental influences in older women. Obesity 2008;16:
dyspnoea grade. Respir Med 2009;103:1070e5. 1616e21.
9. Murray MP, Turnbull K, Macquarrie S, Hill AT. Assessing response 21. Green DJ, Watts K, Rankin S, Wong P, O’Driscoll JG. A
to treatment of exacerbations of bronchiectasis in adults. Eur comparison of the shuttle and 6 minute walking tests with
Respir J 2009;33:312e8. measured peak oxygen consumption in patients with heart
10. Hillerdal G, Lofdahl CG, Strom K, Skoogh BE, Jorfeldt L, failure. J Sci Med Sport 2001;4:292e300.
Nilsson F, et al. Comparison of lung volume reduction surgery 22. McKeough ZJ, Leung RW, Alison JA. Shuttle walk tests as
and physical training on health status and physiologic outcome measures: are two incremental shuttle walk tests and
outcomes: a randomized controlled clinical trial. Chest 2005; two endurance shuttle walk tests necessary? Am J Phys Med
128:3489e99. Rehabil 2011;90:35e9.
11. Sandland CJ, Morgan MD, Singh SJ. Detecting oxygen desatu- 23. Jolly K, Taylor RS, Lip GY, Singh S. Reproducibility and safety of
ration in patients with COPD: incremental versus endurance the incremental shuttle walking test for cardiac rehabilitation.
shuttle walking. Respir Med 2008;102:1148e52. Int J Cardiol 2008;125:144e5.
12. Jurgensen SP, Antunes LC, Tanni SE, Banov MC, Lucheta PA, 24. Fowler SJ, Singh SL, Revill S. Reproducibility and validity of the
Bucceroni AF, et al. The incremental shuttle walk test in older incremental shuttle walking test in patients following coronary
Brazilian adults. Respiration 2011;81:223e8. artery bypass. Physiotherapy 2005;91:22e7.