Research

Original Investigation

Prevention of Low Back Pain

A Systematic Review and Meta-analysis
Daniel Steffens, PhD; Chris G. Maher, PhD; Leani S. M. Pereira, PhD; Matthew L Stevens, MScMed (Clin Epi);
Vinicius C. Oliveira, PhD; Meredith Chapple, BPhty; Luci F. Teixeira-Salmela, PhD; Mark J. Hancock, PhD

Invited Commentary
IMPORTANCE Existing guidelines and systematic reviews lack clear recommendations for Supplemental content at
prevention of low back pain (LBP). jamainternalmedicine.com

OBJECTIVE To investigate the effectiveness of interventions for prevention of LBP.

DATA SOURCES MEDLINE, EMBASE, Physiotherapy Evidence Database Scale, and Cochrane
Central Register of Controlled Trials from inception to November 22, 2014.

STUDY SELECTION Randomized clinical trials of prevention strategies for nonspecific LBP.

DATA EXTRACTION AND SYNTHESIS Two independent reviewers extracted data and assessed
the risk of bias. The Physiotherapy Evidence Database Scale was used to evaluate the
risk-of-bias. The Grading of Recommendations Assessment, Development, and Evaluation
system was used to describe the quality of evidence.

MAIN OUTCOMES AND MEASURES The primary outcome measure was an episode of LBP, and
the secondary outcome measure was an episode of sick leave associated with LBP. We
calculated relative risks (RRs) and 95% CIs using random-effects models.

RESULTS The literature search identified 6133 potentially eligible studies; of these, 23
published reports (on 21 different randomized clinical trials including 30 850 unique
participants) met the inclusion criteria. With results presented as RRs (95% CIs), there was
moderate-quality evidence that exercise combined with education reduces the risk of an
episode of LBP (0.55 [0.41-0.74]) and low-quality evidence of no effect on sick leave (0.74
[0.44-1.26]). Low- to very lowquality evidence suggested that exercise alone may reduce the
risk of both an LBP episode (0.65 [0.50-0.86]) and use of sick leave (0.22 [0.06-0.76]). For
education alone, there was moderate- to very lowquality evidence of no effect on LBP (1.03
[0.83-1.27]) or sick leave (0.87 [0.47-1.60]). There was low- to very lowquality evidence that
back belts do not reduce the risk of LBP episodes (1.01 [0.71-1.44]) or sick leave (0.87
[0.47-1.60]). There was low-quality evidence of no protective effect of shoe insoles on LBP
(1.01 [0.74-1.40]).
Author Affiliations: Musculoskeletal
CONCLUSION AND RELEVANCE The current evidence suggests that exercise alone or in Division, The George Institute for
combination with education is effective for preventing LBP. Other interventions, including Global Health, Sydney Medical
School, The University of Sydney,
education alone, back belts, and shoe insoles, do not appear to prevent LBP. Whether
Sydney, Australia (Steffens, Maher,
education, training, or ergonomic adjustments prevent sick leave is uncertain because the Stevens); Department of
quality of evidence is low. Physiotherapy, Federal University of
Minas Gerais, Belo Horizonte, Brazil
(Steffens, Pereira, Oliveira,
Teixeira-Salmela); Discipline of
Physiotherapy, Medicine and Health
Sciences, Macquarie University,
Sydney, Australia (Chapple, Hancock).
Corresponding Author: Daniel
Steffens, PhD, Musculoskeletal
Division, The George Institute for
Global Health, Sydney Medical
School, The University of Sydney,
GPO Box 5389, Sydney,
JAMA Intern Med. doi:10.1001/jamainternmed.2015.7431 New South Wales, Australia 2000
Published online January 11, 2016. (dsteffens@georgeinstitute.org.au).

(Reprinted) E1

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Research Original Investigation Prevention of Low Back Pain

L
ow back pain (LBP) is one of the most burdensome health prevent future episodes of LBP; (3) compared intervention group
problems worldwide,1 generating enormous costs in treat- with groups that received no intervention, placebo, or minimal
ments and time lost from work.2 The global point preva- intervention; and (4) reported a measure of a new episode of LBP
lence of LBP is 12%; with the aging population, the number of (eg, episode of LBP or episode of sick leave due to LBP). Studies
people affected is likely to increase over the coming years.3 A that used a quasi-randomized design or reported the comparison
key contributor to the burden is the high recurrence rate: ap- of 2 prevention strategies (eg, exercise vs lumbar support) were
proximately one-half of patients experience a recurrence of LBP excluded. No restrictions were placed on the setting or context
within 1 year after recovering from a previous episode.4-6 It is of the included studies, languages, or date of the RCT report.
therefore important to know whether it is possible to prevent
LBP and, if so, which interventions are most effective. Data Extraction and Synthesis
Although there have been several systematic reviews of We assessed the quality of the trials methods using the PEDro
strategies to prevent LBP, most have major limitations. Many scale15,16 by either downloading the available scores from the
of the existing reviews are out-of-date,7,8 report data from ran- PEDro database (http://www.pedro.org.au) or rating the trial our-
domized clinical trials (RCTs) of symptomatic participants,9 do selves. Scores on the PEDro scale range from 0 (very low meth-
not consider the strength of evidence (eg, using the Grading odologic quality) to 10 (high methodological quality); meth-
of Recommendations Assessment, Development, and Evalu- odologic quality was not an inclusion criterion of this review.
ation [GRADE] system),8,10 are restricted to a particular type Two independent reviewers (D.S. or M.J.H. with V.C.O. or
of intervention11 or setting, or do not follow a prespecified, pub- M.C.) extracted the characteristics and intervention out-
licly accessible protocol.7,8 comes of each trial using a standardized data extraction form.
Therefore, a comprehensive, high-quality review that in- When possible, we extracted the raw outcomes (number of per-
cludes the most recent publications is needed to provide a cur- sons having an episode of LBP) for each group (intervention
rent overview of the effectiveness of prevention strategies. The and control) and calculated the estimates of treatment effect
aim of this systematic review was to evaluate the evidence on using methods recommended in the Cochrane Handbook for
the effectiveness of interventions for prevention of episodes Systematic Reviews of Interventions, Version 5.1.0.17
of LBP and use of sick leave due to LBP. To evaluate the overall quality of the evidence, we used the
GRADE system.18 The GRADE classification was downgraded
from high quality by 1 level for each factor that we encoun-
tered: (1) design limitation (>25% of participants from studies
Methods with low methodologic quality: PEDro score <7), (2) inconsis-
Literature Search tency of results (wide variance of point estimates across stud-
The PRISMA Statement was used to guide the conduct and re- ies or large heterogeneity between trials: I2 >50%), and (3) im-
porting of the study.12 This study searched the following elec- precision (<400 participants for each outcome). We did not
tronic databases from the earliest record to November 22, 2014: consider the indirectness criterion in this review because we in-
MEDLINE, EMBASE, Physiotherapy Evidence Database cluded a specific population with relevant outcomes and di-
(PEDro), and the Cochrane Central Register of Controlled Trials. rect comparisons. A GRADE profile was completed for each
A sensitive search strategy was used based on the recommen- pooled estimate and for single trials comparing an LBP preven-
dations of the Cochrane Back Review Group13 for randomized tion strategy with controls. When only single RCTs were avail-
controlled trials and back pain as well as search terms for able, evidence from studies with fewer than 400 participants
prevention.14 The full search strategy is outlined in eTable 1 in was downgraded for inconsistency and imprecision (ie, sparse
the Supplement. The reference lists of relevant reviews and data) and rated as low-quality evidence. These trials could be
trials were screened for additional studies, and we also used further downgraded to very lowquality evidence if limita-
citation tracking of all included trials. tions of study design were found (PEDro score <7). Two review-
During the first screening, 2 reviewers (D.S. or M.J.H. with ers (D.S. or M.J.H. with V.C.O. or M.C.) judged whether these fac-
V.C.O. or M.C.) evaluated the titles and abstracts of each cita- tors were present for each outcome. The quality of evidence was
tion and excluded clearly irrelevant studies. For each poten- defined as (1) high (further research is unlikely to change our
tially eligible study, 2 reviewers (D.S. or M.J.H. with V.C.O. or confidence in the estimate of effect and there are no known or
M.C.) examined the full-text article and assessed whether the suspected reporting biases: all domains are fulfilled); (2) mod-
study fulfilled the inclusion criteria. In cases of disagree- erate (further research is likely to have an important effect on
ment, a decision was made by consensus or, if necessary, a third our confidence in the estimate of effect and might change the
reviewer (C.G.M.) was consulted. estimate: 1 of the domains is not fulfilled); (3) low (further re-
search is likely to have an important effect on our confidence
Study Selection in the estimate of effect and is likely to change the estimate: 2
We included RCTs assessing the effectiveness of prevention strat- of the domains are not fulfilled); and (4) very low (we are un-
egies for nonspecific LBP. To be eligible, trials needed to meet the certain about the estimate: 3 of the domains are not fulfilled).19
following criteria: (1) included participants without LBP at study
entry or at least 1 outcome was not present at baseline (eg, some Statistical Analysis
participants had mild LBP, but all were working and the study out- Outcome data were extracted for short-term (follow-up evalu-
come was an episode of work absence due to LBP); (2) aimed to ations 12 months) and long-term (follow-up evaluations >12

E2 JAMA Internal Medicine Published online January 11, 2016 (Reprinted) jamainternalmedicine.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Prevention of Low Back Pain Original Investigation Research

of the trials focused largely or completely on working-age popu-

Figure 1. Flow Diagram of Review Process
lations. The sample size of the trials ranged from 30 to 4325
participants. A comprehensive description of each trial is pro-
7153 Records identified through 35 Records identified through
database searching other sources vided in Table 1.
1508 MEDLINE
Methodologic quality assessment was conducted using the
4894 EMBASE
711 Cochrane Central PEDro scale. The mean (SD) score was 5.1 (1.5), with the key
40 PEDro problem items being blinding, concealed allocation, and loss
to follow-up (eTable 2 in the Supplement).
Estimates of the effects of LBP prevention strategies on
6133 Records after duplicates LBP episode or sick leave due to LBP were calculated for 21
removed
trials. The number of events, sample size, and RRs (95% CIs)
for the trials are presented in eTable 3 in the Supplement.
5974 Records excluded based
on title and abstract Trials were grouped according to the prevention strategy,
outcome (episode of LBP or sick leave), and follow-up time
point (short- or long-term). Table 2 provides a summary of
159 Full-text articles assessed
for eligibility the findings and GRADE quality ratings.

136 Full-text articles excluded Exercise vs Control, Minimal Intervention, or Supplement

61 Not an RCT Four trials reporting data on 898 participants were included
16 No LBP outcome
55 LBP at baseline in the meta-analysis to estimate the short-term (ie, 12
1 No control group months) efficacy of exercise on incident cases of LBP (pre-
1 Retraction
2 Pregnancy
sented as RR [95% CI]).21,25,26,42 The pooled results provide
low-quality evidence of a protective effect of exercise (0.65
[0.50-0.86]). In the long-term (ie, >12 months), the pooled
23 RCTs included in review
(31 112 unique participants) results of 2 trials (334 participants) provide very lowquality
ev idence of no effec t of exerc ise (1.04 [0.73-1.49])
LBP indicates lower back pain; PEDro, Physiotherapy Evidence Database; and (Figure 2).21,33 Two trials presented data from 128 partici-
RCT, randomized clinical trial. pants and provide very lowquality evidence that exercise
reduces the risk of sick leave due to LBP in the long-term
months) follow-up. When multiple time points fell into the (0.22 [0.06-0.76]) (Figure 3).30,42
same category, we used the longest follow-up period.
Trials considered homogeneous were grouped accord- Exercise and Education vs Control, Minimal Intervention,
ing to the prevention strategy, comparison group, outcome or Supplement
(LBP episode and sick leave), and outcome assessment time The effect of exercise and education was investigated in 4 trials
points (short-term and long-term). We calculated relative (442 participants) at short-term follow-up,22,35,39,42 and in 2
risks (RRs) and 95% CIs and used the random-effects model trials (138 participants) at long-term follow-up (LBP
to pool estimates for each analysis obtained with Compre- episode).30,40 The pooled results (presented as RR [95% CI])
hensive Meta-analysis, version 2.2.064 (Biostat). For trials of 4 trials provide moderate-quality evidence that exercise and
that did not report the sample size at the end of the education reduce the risk of an episode of LBP at short-term
follow-up period, we calculated the RR using the baseline follow-up (0.55 [0.41-0.74]). The long-term results are based
sample size. on 2 trials30,40 and provide low-quality evidence of a protec-
tive effect (0.73 [0.55-0.96]) (Figure 2).
For prevention of sick leave due to LBP, 3 trials (228
participants)22,39,42 presented short-term data and 2 trials
Results (138 participants)30,40 presented long-term data. The pooled
The initial electronic database search identified 6133 poten- results (presented as RR [95% CI]) provide low-quality evi-
tially eligible studies. After screening citations by title and dence of no protective effect at short-term follow-up (0.74
abstract, we considered 159 potentially eligible studies for [0.44-1.26]) or long-term follow-up (0.72 [0.48-1.08])
inclusion and retrieved full-text articles. Twenty-three (Figure 3).
published reports (21 different RCTs including 30 850 unique
participants) met the inclusion criteria and were included Education vs Control, Minimal Intervention, or Supplement
i n t h i s r e v i e w. 2 0 - 4 2 T w o RC Ts w e r e r e p o r t e d i n 4 The efficacy of education compared with control was inves-
articles 22,30,39,40 (2 with 12-month data 22,39 and 2 with tigated in 3 trials (2343 participants) at short-term follow-up
36-month data 30,40 ). Figure 1 outlines the flow of RCTs and in 2 trials (13242 participants) at long-term follow-up (LBP
through the review. episode). The pooled results (presented as RR [95% CI]) pro-
The included trials investigated 6 different LBP preven- vide moderate-quality evidence of no protective effect of edu-
tion strategies: exercise, education, exercise and education, cation at either short-term follow-up (1.03 [0.83-1.27])37,41,42
back belts, shoe insoles, and other prevention strategies. Most or long-term follow-up (0.86 [0.72-1.04])20,34 (Figure 2). In

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online January 11, 2016 E3

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Research Original Investigation Prevention of Low Back Pain

Table 1. Characteristics of the Randomized Clinical Trials Included in Review of Low Back Pain Prevention Strategies
Time of Duration of
Source Participants Outcome Intervention and Control Sessions Intervention
George 4325 Army soldiers; LBP episode that resulted in Traditional exercise: traditional lumbar exercises for the 5 Times/wk 12 wk
et al,20 mean (SD) age, the patient seeking of rectus abdominus and oblique abdominal muscles for 5 min
2011 22.0 (4.2) y; male health care Education: evidence-based information on LBP and 1 Time/wk Single
(71%) educational book for 45 min session
Core exercise: core stabilization exercises for transverse 5 Times/wk 12 wk
abdominus, multifidus, and the erector spinae for 5 min
Helewa 402 University LBP episode: continuous or Exercise: abdominal muscle strength exercises 7 Times/wk 24 mo
et al,21 employees and students, intermittent pain resulting for 5 min
1999 hospital staff, and in moderate to severe Education: classes on spinal anatomy, pathophysiology, 3 Times/wk 3 Sessions
London residents; mean limitation of function posture, lifting techniques, and general fitness for 90 min (baseline,
(SD) age, 38.4 (9.2) y; lasting >2 d 1- and 2-y
male (47%) follow-up)
Lnn 81 Participants recruited LBP episode: recurrences Exercise and education: active back school-didactic 2 Sessions/wk 20 Sessions
et al,22 through local media Sick leave: due to episodes session included anatomy, biomechanics, pathology, and for 7 wk and (13 wk)
1999 advertisement and of LBP basic ergonomic principles related to the spinal column 1/wk for 6 wk;
referral from other and pelvis; practical session included bending the knee each session
health professionals; and hip joints, while keeping the lumbar segments near 60 min
mean (range) age, midposition and using short-lever arms during functional
39.4 (19.2-49.8) y; 46% exercises and obstacle course simulations; strength
males training of leg muscles and muscles between the upper
body and pelvis; stretching exercises for the calf muscles,
hamstrings, rectus femoris, and hip flexors
Control group: no intervention)
Mattila 220 Finnish defense LBP episode: requiring a Shoe insoles: customized insoles made from firm-density Daily service 6 mo
et al,23 forces; mean age 19.0 y; visit to the physician and polyethylene, and the hard plastic shell was time
2011 100% male suspension from duty for at three-quarters the length of the foot
least 1 d Control: no intervention
Milgrom 404 New recruits LBP episode: presence of Semirigid shoe insoles: semirigid biomechanical orthoses Unclear 14 wk
et al,24 beginning elite infantry LBP Soft shoe insoles: soft biomechanical orthoses
2005 training; mean (SD), Control: simple shoe inserts, without supportive or
18.8 (0.7) y; 100% male shock-absorbing qualities
Moore 30 Outpatients of the LBP episode: incidence of Exercise: 6 calisthenic exercises to strengthen and stretch 15 min/d 12 mo
et al,25 Brown Cancer Center, self-reported LBP the pelvis-spineattached muscles that move lumbar and
2012 University of Louisville; lumbosacral joints and control upright, 2-legged balance
mean (range) age, Control: no intervention
49.0 (43-63) y; 23%
male
Sihawong 563 Office workers; LBP episode: LBP lasting Exercise: muscle stretching and endurance training Twice daily 12 mo
et al,26 mean (SD) age, >24 h during the past (repeatedly contracted each muscle [ie, erector spinae, (5 d/wk for
2014 37.1 (10.4) y; 31% male month multifidus, quadratus lumborum, and transversus 30 s each time)
abdominis] 10 times and rested for 60 s between muscle
contractions)
Control: no intervention
Allen and 47 Employees of the LBP episode: back injury Education: training in biomechanics and proper lifting Volunteers 6 mo
Wilder,27 Veterans Administration techniques were asked
1996 Hospital; age and sex not Back belts: training on proper use of back belts to wear the
specified back support
belts while on
duty whenever
they were
lifting
patients
Daltroy 3597 US postal workers; LBP episode: occurrence of Education: safe lifting and handling; posture while sitting, 90 min 2 Sessions
et al,28 mean (SD) age, LBP injury standing, and lying down; pain management; stretching
1997 42.5 (12.3) y; 66% male and strengthening exercises; group discussion of barriers
to implementation; on-site work-station ergonomic
analysis
Control: no intervention
Driessen 3047 Employees of 4 LBP episode: DMQ asked Ergonomic program: implementation of ergonomic 6h 1 Session
et al,29 Dutch companies; mean about the presence of LBP measurers aimed to prevent LBP
2011 (SD) age, 42.0 (21.8) y; in the previous 3 mo (l, no, Control: no intervention
59% male never; 2, yes, sometimes; 3,
yes, regularly; 4, yes,
always); prevalence was
determined by combining
the categories 1 and 2 as
no LBP and categories 3
and 4 as LBP

(continued)
41,42
addition, a single trial (3597 participants) not included in the Two trials (366 participants) presented short-term
meta-analysis because it did not report raw data provides mod- data on sick leave prevention. The pooled results provide
erate-quality evidence of no protective effect of education at very lowquality evidence of no protective effect of educa-
long-term follow-up (rate ratio, 1.11 [95% CI, 0.90-1.37]) (eTable tion on sick leave due to LBP at short-term follow-up (RR,
3 in the Supplement).28 0.87 [95% CI, 0.47-1.60]) (Figure 3).

E4 JAMA Internal Medicine Published online January 11, 2016 (Reprinted) jamainternalmedicine.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Prevention of Low Back Pain Original Investigation Research

Table 1. Characteristics of the Randomized Clinical Trials Included in Review of Low Back Pain Prevention Strategies (continued)
Time of Duration of
Source Participants Outcome Intervention and Control Sessions Intervention
Glomsrd 81 Participants recruited LBP episode: recurrence of Exercise and education: active back school-didactic 2 Sessions/wk (20 Sessions)
et al,30 from referrals and episodes session included anatomy, biomechanics, pathology, and for 7 wk; 13 wk
2001 advertisement; mean Sick leave: due to episodes basic ergonomic principles related to the spinal column 1 session/wk
(SD) age, 39.8 (6.4) y; of LBP and pelvis; practical session included bending the knee for 6 wk;
46% male and hip joints, while keeping the lumbar segments near each session
midposition and using short lever arms during functional 60 min
exercises and obstacle course simulations; strength
training of leg muscles and muscles between the upper
body and pelvis; stretching exercises for the calf muscles,
hamstrings, rectus femoris, and hip flexors
Control: no intervention
Gundewall 69 Nurses and nurses Sick leave: work absence Exercise: back muscle exercises to increase endurance, 6 Times/mo 13 mo
et al,31 aides; mean (SD) age, due to LBP isometric strength and functional coordination for 20 min
1993 37.5 (10.5) y; 1% male Control: no intervention
IJzelenberg 489 Workers from Sick leave: absent from Education, training, and ergonomic adjustments: Unclear Unclear
et al,32 physically demanding work during the past 6 mo individually tailored education and training, immediate
2007 jobs; mean (SD) age, and 12 mo due to back pain treatment of acute LBP, and advice on ergonomic
41.3 (9.7) y; 97% male adjustment of the workplace
Usual care: Dutch guidelines for the health care of
patients with LBP
Kellett 111 Employees of Sick leave: attributable to Exercise and education: warm-up, stretching, 2 Times/wk 18 mo
et al,33 kitchen unit production; LBP strengthening, cardiovascular, coordination exercises and for 20-35 min
1991 mean (SD) age, cool down; one-third of the classes started with 10-min
41.7 (10.1) y; 70% male lecture on theories of back pain prevention, eg, reducing
bed rest and increasing activities, eg, swimming
Control: no intervention
Kraus 12772 Home care LBP episode: acute-onset, Back belt: stretch nylon back belts Unclear 28 mo
et al,34 attendants; mean physician-diagnosed injury Education: information on LBP health
2002 (range) age, NS (18-65 to the lower back that Control: no intervention
y); 5% male occurred during a
work-related activity
Larsen 314 Military conscripts; LBP episode: No. of persons Education: back school lesson consisted of the theory Single 40-min Single
et al,35 mean (SD) age, who reported having based on a booklet43 session session
2002 21.0 (1.5) y; 100% male consulted the military Exercise: 15 passive prone extensions of the back Twice daily 10 mo
medical physician with back Control: no intervention
problems
Larsen 146 Military conscripts; LBP episode: self-reported Shoe insoles: custom-made biomechanical shoe orthoses Whenever 3 mo
et al,36 mean (range) age, NS back problems Control: no intervention wearing
2002 (18-24 y); 99% male their military
boots
Lavender 2144 Workers from LBP episode: self-reported Education: lifting training; participants were 5 Sessions 10 mo
et al,37 distribution centers that back injury instrumented with motion-capture sensors to quantify the for 30 min
2007 require lifting; mean dynamic moments (torque) vector acting on lumbar spine Unclear Single
(range) age, (L5/S1) session
33.5 (18-65) y; 96% Video training: demonstrating various lifting techniques
male
Schwellnus 1388 New military LBP episode: overuse back Shoe insoles: neoprene-impregnated with nitrogen Daily 9 wk
et al,38 recruits; mean (SD) age, injury bubbles covered with stretch nylon
1990 18.5 (1.2) y; sex NS Control: standard military footwear
Soukup 77 Outpatients from LBP episode: resulting in Exercise and education: Mensendieck exercises and 20 Sessions 13 wk
et al,39 medical and professional management biomechanical/ ergonomic, back anatomy, pain for 60 min
1999 physiotherapist Sick leave: LBP resulting in mechanisms, and working posture education
practices; mean (SD) use of sick leave Control: no intervention
age, 37.7 (8.0) y; 47%
male
Soukup 77 Outpatients from LBP episode: resulting in Exercise and education: Mensendieck exercises and 20 Sessions 13 wk
et al,40 medical and professional management biomechanical/ ergonomic, back anatomy, pain for 60 min
2001 physiotherapist Sick leave: LBP resulting in mechanisms and working posture education
practices; mean (SD) use of sick leave Control: no intervention
age, 37.7 (8.0) y; 47%
male
van Poppel 624 Airline employees; LBP episode: in the past Back belts: lumbar support with adjustable elastic side Wear at all 6 mo
et al,41 mean (SD) age, month pulls with Velcro fasteners and flexible stays times during
1998 35.1 (7.8) y; sex NS Sick leave: time lost from Education: lifting instructions work hours
work in the past month 1 Session for 2 h 3 Sessions
and 2 sessions for 12 wk
for 1.5 h
Warming 181 Hospital nurses LBP episode: perceived LBP Education: patient transfer technique based on the law of Working hours Two 6-wk
et al,42 Copenhagen; mean (SD) Sick leave: due to LBP physics and the natural movement pattern of moving 1 sessions
2008 age, 34.8 (9.3) y; sex NS body part at a time 2 Times/wk 16 Sessions
Exercise: physical fitness training: aerobic fitness and for 60 min for 8 wk
strength training
Control: no intervention

Abbreviations: DMQ, Dutch musculoskeletal questionnaire; LBP, low back pain; NS, not specified.

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online January 11, 2016 E5

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Research Original Investigation Prevention of Low Back Pain

Table 2. Summary of Findings and Quality of Evidence Assessment

Summary of Findings Quality of Evidence Assessment (GRADE)

Outcome Timea No. of Participants RR (95% CI) Limitation Inconsistency Imprecision Quality
Exercise vs Control/Minimal Intervention/or as Supplement
LBP episode Short-term 89821,25,26,42 0.65 (0.50-0.86) Limitation Inconsistency No imprecision Low
LBP episode Long-term 33421,33 1.04 (0.73-1.49) Limitation Inconsistency Imprecision Very low
Sick leave Long-term 12831,42 0.22 (0.06-0.76) Limitation Inconsistency Imprecision Very low
Exercise and Education vs No/Minimal Intervention/or as Supplement
LBP episode Short-term 44222,35,39,42 0.55 (0.41-0.74) Limitation No inconsistency No imprecision Moderate
LBP episode Long-term 13830,40 0.73 (0.55-0.96) Limitation No inconsistency Imprecision Low
Sick leave Short-term 22822,39,42 0.74 (0.44-1.26) Limitation No inconsistency Imprecision Low
Sick leave Long-term 13830,40 0.72 (0.48-1.08) Limitation No inconsistency Imprecision Low
Education vs Nothing/Minimal Intervention/or as Supplement
LBP episode Short-term 234337,41,42 1.03 (0.83-1.27) Limitation No inconsistency No imprecision Moderate
LBP episode Long-term 13 24220,34 0.86 (0.72-1.04) Limitation No inconsistency No imprecision Moderate
LBP episode Long-term 359728 Rate ratio (95% CI), Limitation No inconsistency No imprecision Moderateb
1.11 (0.90-1.37)
Sick leave Short-term 36641,42 0.87 (0.47-1.60) Limitation Inconsistency Imprecision Very low
Back Belt vs Nothing/Minimal Intervention/or as Supplement
LBP episode Short-term 32927,41 1.01 (0.71-1.44) Limitation Inconsistency Imprecision Very low
LBP episode Long-term 847234 0.85 (0.64-1.14) Limitation Inconsistency Imprecision Very lowa
Sick leave Short-term 28241 1.44 (0.73-2.86) No limitation Inconsistency Imprecision Lowa
Shoe Insole vs Nothing/Minimal Intervention/or as Supplement
LBP episode Short-term 183323,24,36,38 1.01 (0.74-1.40) Limitation Inconsistency No imprecision Low
Others Prevention Strategies vs Control
LBP episode Short-term 304729 OR (95% CI), Limitation No inconsistency No imprecision Moderatea
1.23 (0.97-1.57)
Sick leave Short-term 36032 0.95 (0.51-1.76) Limitation Inconsistency Imprecision Very lowa
Abbreviations: GRADE, Grading of Recommendations Assessment, evaluations longer than 12 months.
Development and Evaluation; LBP, low back pain; OR, odds ratio; b
Quality of evidence was evaluated using a single trial.
RR, relative risk.
a
Short-term, follow-up evaluations of 12 months or less; long-term, follow-up

Back Belts vs Control, Minimal Intervention, or Supplement Other LBP Prevention Strategies
The efficacy of back belts over control to prevent LBP epi- Two trials reported the short-term effect of other prevention
sodes (short- and long-term) or sick leave owing to LBP (short- strategies vs control for LBP episode (3047 participants),29 and
term) was reported in 3 trials.27,34,41 For episodes of LBP, pool- sick leave due to LBP (360 participants).32 An ergonomic pro-
ing of 2 trials (329 participants) (presented as RR [95% CI]) gram (moderate-quality evidence) was not more effective than
provides very lowquality evidence of no short-term effect of control in reducing episodes of LBP at short-term follow-up
back belts over controls (1.01 [0.71-1.44]) (Figure 2).27,41 At long- (odds ratio, 1.23 [95% CI, 0.97-1.57]) (Table 2). It is unclear
term follow-up, a single trial (8472 participants) provides mod- whether sick leave due to LBP can be prevented by educa-
erate-quality evidence that back belts do not reduce the risk tion, training, and ergonomic adjustments since there was very
of LBP episodes when compared with controls (0.85 [0.64- lowquality evidence (RR, 0.95 [95% CI, 0.51-1.76]) (Figure 3).
1.14]) (Figure 2).34 For sick leave owing to LBP, a single trial (282
participants) provides low-quality evidence of no effect of back
belts compared with controls at short-term follow-up (RR, 1.44
[95% CI, 0.73-2.86]) (Figure 3).41
Discussion
Statement of Principal Findings
Shoe Insole vs Control, Minimal Intervention, or Supplement The results of this systematic review and meta-analysis indi-
Four trials reported data from 1833 participants on the short- cate that exercise in combination with education is likely to
term efficacy of shoe insoles compared with controls.23,24,36,38 reduce the risk of LBP. Exercise alone may reduce the risk of
For prevention of episodes of LBP, there is low-quality evi- an episode of LBP and sick leave; however, it is uncertain
dence that shoe insoles are not superior to control at short- whether the effects persist beyond 1 year. Education alone, back
term follow-up (RR, 1.01 [95% CI, 0.74-1.40]) (Figure 2). One belts, shoe insoles, and ergonomic adjustments probably do
trial reported the efficacy of semirigid shoe insole vs control not prevent an episode of LBP or sick leave due to LBP. It is un-
and soft shoe insole vs control.24 Only the group from the semi- certain whether education, training, or ergonomic adjust-
rigid shoe insole was included in the meta-analysis. ments prevent LBP owing to the very low quality of evidence.

E6 JAMA Internal Medicine Published online January 11, 2016 (Reprinted) jamainternalmedicine.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Prevention of Low Back Pain Original Investigation Research

Figure 2. Relative Risk for Low Back Pain Episode in Controlled Trials on Efficacy of Low Back Pain Prevention Strategies

Intervention Control

No. of No. of No. of No. of Favors Favors

Source Events Patients Events Patients RR (95% CI) Intervention Control Weight, %
Exercise vs control (short-term)
Helewa et al,21 1999 28 113 37 157 1.05 (0.69-1.61) 36
Warming et al,42 2008 14 35 22 33 0.60 (0.37-0.96) 31
Moore et al,25 2012 0 13 10 17 0.06 (0.00-0.96) 1
Sihawong et al,26 2014 23 261 53 269 0.45 (0.28-0.71) 32
Pooled effect: I2 = 67.9% 0.65 (0.50-0.86)
Exercise vs control (long-term)
Kellett et al,33 1991 8 37 14 48 0.74 (0.35-1.58) 21
Helewa et al,21 1999 35 101 45 148 1.14 (0.79-1.64) 79
Pooled effect: I2 = 0% 1.04 (0.73-1.49)
Exercise and education vs control (short-term)
Soukup et al,39 1999 11 34 20 35 0.57 (0.32-1.00) 25
Lnn et al,22 1999 11 38 20 35 0.51 (0.29-0.90) 24
Larsen et al,35 2002 9 101 28 113 0.36 (0.18-0.73) 17
Warming et al,42 2008 14 35 29 51 0.70 (0.44-1.13) 34
Pooled effect: I2 = 0% 0.55 (0.41-0.74)
Exercise and education vs control (long-term)
Soukup et al,40 2001 18 31 27 35 0.75 (0.53-1.07) 50
Glomsrd et al,30 2001 20 37 27 35 0.70 (0.50-0.99) 50
Pooled effect: I2 = 0% 0.73 (0.55-0.96)
Education vs control (short-term)
van Poppel et al,41 1998 50 142 49 140 1.01 (0.73-1.38) 35
Lavender et al,37 2007 66 957a 76 1020a 0.93 (0.67-1.27) 34
Warming et al,42 2008 22 33 29 51 1.17 (0.83-1.65) 31
Pooled effect: I2 = 0% 1.03 (0.83-1.27)
Education vs control (long-term)
Kraus et al,34 2002 89 4300a 109 4635a 0.88 (0.67-1.16) 31
George et al,20 2011 300 1995 406 2312 0.86 (0.75-0.98) 69
Pooled effect: I2 = 0% 0.86 (0.72-1.04)
Back belt vs control (short-term)
Allen and Wilder,27 1996 0 23 3 24 0.15 (0.01-2.73) 2
van Poppel et al,41 1998 48 134 51 148 1.04 (0.76-1.43) 98
Pooled effect: I2 = 40.6% 1.01 (0.71-1.44)
Back belt vs control (long-term)
Kraus et al,34 2002 77 3837a 109 4635a 0.85 (0.64-1.14) 100
Shoe insoles vs control (short-term)
Schwellnus et al,38 1990 2 237 24 1151 0.40 (0.10-1.70) 4
Larsen et al,36 2002 9 58 9 63 1.09 (0.46-2.55) 10
Milgrom et al,24 2005 16 51 18 53 0.92 (0.53-1.61) 23
Mattila et al,23 2011 24 73 42 147 1.15 (0.76-1.74) 40
Pooled effect: I2 = 0% 1.01 (0.74-1.40)

0.1 1.0 10
RR (95% CI)

a
Studies are ordered chronologically within prevention strategies. Short-term Only the baseline sample size was available.
indicates follow-up of 12 months or less; long-term, follow-up evaluation of
more than 12 months.

Strengths and Weaknesses of the Study This review was designed to be comprehensive with a robust
The strengths of this review include the use of a prespecified search strategy; however, it is possible that not all studies were
protocol registered on PROSPERO, inclusion of all prevention identified. Some identified trials did not have the term prevention
strategies from any setting, the use of the GRADE system to in either the title or the abstract.27,33,34 For several prevention
evaluate the overall quality of the evidence, and the use of a strategies, we could identify only a small number of trials; this
highly sensitive search strategy to identify LBP prevention combined with the quality of the trials means the level of
trials. We assessed trials methodologic quality with the PEDro evidence for several prevention strategies is very low or low.
scale, which has been shown to have acceptable reliability and
validity.15,16,44 All scores were available online at the PEDro Comparison With Other Studies
website. These scores were rated by experienced PEDro To our knowledge, this review is the first to have included a va-
researchers, which provided less chance of errors. riety of LBP prevention strategies and conducted a meta-analysis

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online January 11, 2016 E7

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Research Original Investigation Prevention of Low Back Pain

Figure 3. Relative Risk for Sick Leave in Controlled Trials on Efficacy of Low Back Pain Prevention Strategies

Intervention Control

No. of No. of No. of No. of Favors Favors

Source Events Patients Events Patients RR (95% CI) Intervention Control Weight, %
Exercise vs control (long-term)
Gundewall et al,31 1993 1 28 12 32 0.10 (0.01-0.69) 39
Warming et al,42 2008 2 35 5 33 0.38 (0.08-1.81) 61
Pooled effect: I2 = 11.7% 0.22 (0.06-0.76)
Exercise and education vs control (short-term)
Lnn et al,22 1999 7 38 11 35 0.59 (0.26-1.34) 39
Soukup et al,39 1999 10 34 11 35 0.94 (0.46-1.91) 51
Warming et al,42 2008 2 35 5 51 0.58 (0.12-2.84) 11
Pooled effect: I2 = 0% 0.74 (0.44-1.26)
Exercise and education vs control (long-term)
Glomsrd et al,30 2001 12 37 18 35 0.63 (0.36-1.11) 47
Soukup et al,39 1999 13 31 18 35 0.82 (0.48-1.38) 53
Pooled effect: I2 = 0% 0.72 (0.48-1.08)
Education vs control (short-term)
van Poppel et al,41 1998 12 142 17 140 0.70 (0.35-1.40) 73
Warming et al,42 2008 5 33 5 51 1.55 (0.48-4.93) 27
Pooled effect: I2 = 22.5% 0.87 (0.47-1.60)
Back belts vs control (short-term)
van Poppel et al,41 1998 17 134 13 148 1.44 (0.73-2.86) 100
Education, training, and ergonomics vs control (short-term)
IJzelenberg,32 2007 18 185 18 175 0.95 (0.51-1.76) 100

0.1 1.0 10
RR (95% CI)

Studies are ordered chronologically within prevention strategies. Short-term indicates follow-up of 12 months or less; long-term, follow-up evaluation of more than
12 months.

of RCTs. Several reviews have investigated the effectiveness of study: shoe insoles are not effective for the prevention of back
an exercise and/or education program on LBP prevention. All are pain. The most recent review by Chuter et al10 included 6 trials;
out-of-date, included at least 1 RCT with symptomatic partici- our review included 4. We excluded 2 trials because the partici-
pants at baseline (ie, the trial evaluated treatment, not preven- pants were symptomatic at the time of study entry.53,54
tion), and presented data descriptively.7-9,45-49 The most recent
review we know of investigating the effectiveness of exercise for Meaning of the Study
preventing a LBP episode,11 presented data from 3 trials. One was Although our review found evidence for both exercise alone
included in the meta-analysis of the current review (ie, exercise (35% risk reduction for an LBP episode and 78% risk reduction
vs control),33 one was excluded because the trial included symp- for sick leave) and for exercise and education (45% risk reduc-
tomatic participants at baseline,50 and one was included in a dif- tion for an LBP episode) for the prevention of LBP up to 1 year,
ferent LBP prevention strategy (ie, exercise and education vs we also found the effect size reduced (exercise and education)
control).39 That review by Choi et al11 reported a 50% (2 RCTs with or disappeared (exercise alone) in the longer term (>1 year). This
130 patients) reduction in future LBP episodes when compared finding raises the important issue that, for exercise to remain
with no intervention, which is a larger effect than our estimate protective against future LBP, it is likely that ongoing exercise
of a 35% reduction (4 RCTs with 898 patients). is required. Prevention programs focusing on long-term behav-
Previous reviews investigating the efficacy of exercise on ior change in exercise habits seem to be important.
the prevention of LBP episodes have not distinguished be-
tween studies that included education with the exercise from
those just including exercise.11,45,46 In our review, the combi-
nation of exercise and education was effective at long-term fol-
Conclusions
low-up (RR, 0.73 [95%CI, 0.55 to 0.96]), while exercise alone The results of this systematic review and meta-analysis of RCTs
was not (RR, 1.04 [95% CI, 0.73 to 1.49]), suggesting that the indicate that exercise in combination with education is likely
distinction between exercise alone and exercise combined with to reduce the risk of LBP and that exercise alone may reduce the
education may be important. risk of an episode of LBP and sick leave due to LBP, at least for
The present reviews finding that back belts do not prevent the short-term. The available evidence suggests that educa-
LBP is consistent with results of a previous systematic review.51 tion alone, back belts, shoe insoles, and ergonomics do not pre-
There are a few previous systematic reviews10,45,52 investigat- vent LBP. It is uncertain whether education, training, or ergo-
ing the use of shoe insoles in the prevention of an LBP episode. nomic adjustments prevent sick leave due to LBP because the
Findings from these reviews are in line with the results of our quality of evidence is very low.

E8 JAMA Internal Medicine Published online January 11, 2016 (Reprinted) jamainternalmedicine.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Prevention of Low Back Pain Original Investigation Research

ARTICLE INFORMATION 11. Choi BK, Verbeek JH, Tam WW, Jiang JY. 26. Sihawong R, Janwantanakul P, Jiamjarasrangsi
Accepted for Publication: November 9, 2015. Exercises for prevention of recurrences of low-back W. A prospective, cluster-randomized controlled
pain. Cochrane Database Syst Rev. 2010;(1): trial of exercise program to prevent low back pain in
Published Online: January 11, 2016. CD006555. office workers. Eur Spine J. 2014;23(4):786-793.
doi:10.1001/jamainternmed.2015.7431.
12. Moher D, Liberati A, Tetzlaff J, Altman DG; 27. Allen SK, Wilder K. Back belts pay off for nurses.
Author Contributions: Drs Steffens and Maher had PRISMA Group. Preferred reporting items for Occup Health Saf. 1996;65(1):59-62.
full access to all the data in the study and take systematic reviews and meta-analyses: the PRISMA
responsibility for the integrity of the data and the 28. Daltroy LH, Iversen MD, Larson MG, et al.
statement. BMJ. 2009;339:b2535. A controlled trial of an educational program to
accuracy of the data analysis.
Study concept and design: Steffens, Maher, Stevens, 13. Furlan AD, Pennick V, Bombardier C, van Tulder prevent low back injuries. N Engl J Med. 1997;337
Teixeira-Salmela, Hancock. M; Editorial Board, Cochrane Back Review Group. (5):322-328.
Acquisition, analysis, or interpretation of data: 2009 Updated method guidelines for systematic 29. Driessen MT, Proper KI, Anema JR, Knol DL,
Steffens, Maher, Pereira, Oliveira, Chapple, reviews in the Cochrane Back Review Group. Spine Bongers PM, van der Beek AJ. The effectiveness of
Teixeira-Salmela, Hancock. (Phila Pa 1976). 2009;34(18):1929-1941. participatory ergonomics to prevent low-back and
Drafting of the manuscript: Steffens, Maher, 14. Burton AK, Balagu F, Cardon G, et al; COST B13 neck painresults of a cluster randomized
Teixeira-Salmela, Hancock. Working Group on Guidelines for Prevention in Low controlled trial. Scand J Work Environ Health. 2011;
Critical revision of the manuscript for important Back Pain. Chapter 2. European guidelines for 37(5):383-393.
intellectual content: All authors. prevention in low back pain : November 2004. Eur 30. Glomsrd B, Lnn JH, Soukup MG, B K,
Statistical analysis: Steffens, Maher, Chapple. Spine J. 2006;15(suppl 2):S136-S168. Larsen S. Active back school, prophylactic
Administrative, technical, or material support: 15. Maher CG, Sherrington C, Herbert RD, Moseley management for low back pain: three-year
Steffens, Oliveira, Teixeira-Salmela, Hancock. AM, Elkins M. Reliability of the PEDro scale for follow-up of a randomized, controlled trial.
Study supervision: Steffens, Maher, Pereira, rating quality of randomized controlled trials. Phys J Rehabil Med. 2001;33(1):26-30.
Teixeira-Salmela, Hancock. Ther. 2003;83(8):713-721. 31. Gundewall B, Liljeqvist M, Hansson T. Primary
Conflict of Interest Disclosures: None reported. 16. de Morton NA. The PEDro scale is a valid prevention of back symptoms and absence from
measure of the methodological quality of clinical work: a prospective randomized study among
REFERENCES trials: a demographic study. Aust J Physiother. hospital employees. Spine (Phila Pa 1976). 1993;18
1. GBD 2013 Risk Factors Collaborators. Global, 2009;55(2):129-133. (5):587-594.
regional, and national comparative risk assessment 17. Higgins JPT, Green S, eds. Cochrane Handbook 32. IJzelenberg H, Meerding WJ, Burdorf A.
of 79 behavioural, environmental and occupational, for Systematic Reviews of Interventions, Version Effectiveness of a back pain prevention program:
and metabolic risks or clusters of risks in 188 5.1.0. Cochrane Collaboration. http://www a cluster randomized controlled trial in an
countries, 1990-2013: a systematic analysis for the .cochrane-handbook.org. Updated March 2011. occupational setting. Spine (Phila Pa 1976). 2007;32
Global Burden of Disease Study 2013 [published Accessed February 12, 2015. (7):711-719.
online September 10, 2015]. Lancet.
18. Atkins D, Best D, Briss PA, et al; GRADE Working 33. Kellett KM, Kellett DA, Nordholm LA. Effects of
2. Hoy D, March L, Brooks P, et al. Measuring the Group. Grading quality of evidence and strength of an exercise program on sick leave due to back pain.
global burden of low back pain. Best Pract Res Clin recommendations. BMJ. 2004;328(7454):1490. Phys Ther. 1991;71(4):283-291.
Rheumatol. 2010;24(2):155-165.
19. Guyatt GH, Oxman AD, Vist GE, et al; GRADE 34. Kraus JF, Schaffer KB, Rice T, Maroosis J,
3. Hoy D, Bain C, Williams G, et al. A systematic Working Group. GRADE: an emerging consensus on Harper J. A field trial of back belts to reduce the
review of the global prevalence of low back pain. rating quality of evidence and strength of incidence of acute low back injuries in New York
Arthritis Rheum. 2012;64(6):2028-2037. recommendations. BMJ. 2008;336(7650):924-926. City home attendants. Int J Occup Environ Health.
4. Stanton TR, Henschke N, Maher CG, Refshauge 20. George SZ, Childs JD, Teyhen DS, et al. Brief 2002;8(2):97-104.
KM, Latimer J, McAuley JH. After an episode of psychosocial education, not core stabilization, 35. Larsen K, Weidick F, Leboeuf-Yde C. Can passive
acute low back pain, recurrence is unpredictable reduced incidence of low back pain: results from prone extensions of the back prevent back
and not as common as previously thought. Spine the Prevention of Low Back Pain in the Military problems? a randomized, controlled intervention
(Phila Pa 1976). 2008;33(26):2923-2928. (POLM) cluster randomized trial. BMC Med. 2011;9: trial of 314 military conscripts. Spine (Phila Pa 1976).
5. Carey TS, Garrett JM, Jackman A, Hadler N. 128. 2002;27(24):2747-2752.
Recurrence and care seeking after acute back pain: 21. Helewa A, Goldsmith CH, Lee P, Smythe HA, 36. Larsen K, Weidich F, Leboeuf-Yde C. Can
results of a long-term follow-up study: North Forwell L. Does strengthening the abdominal custom-made biomechanic shoe orthoses prevent
Carolina Back Pain Project. Med Care. 1999;37(2): muscles prevent low back paina randomized problems in the back and lower extremities?
157-164. controlled trial. J Rheumatol. 1999;26(8):1808-1815. a randomized, controlled intervention trial of 146
6. Pengel LH, Herbert RD, Maher CG, Refshauge 22. Lnn JH, Glomsrd B, Soukup MG, B K, Larsen military conscripts. J Manipulative Physiol Ther.
KM. Acute low back pain: systematic review of its S. Active back school: prophylactic management for 2002;25(5):326-331.
prognosis. BMJ. 2003;327(7410):323. low back pain: a randomized, controlled, 1-year 37. Lavender SA, Lorenz EP, Andersson GB. Can a
7. van Poppel MN, Hooftman WE, Koes BW. An follow-up study. Spine (Phila Pa 1976). 1999;24(9): new behaviorally oriented training process to
update of a systematic review of controlled clinical 865-871. improve lifting technique prevent occupationally
trials on the primary prevention of back pain at the 23. Mattila VM, Sillanp P, Salo T, Laine HJ, related back injuries due to lifting? Spine (Phila Pa
workplace. Occup Med (Lond). 2004;54(5):345-352. Menp H, Pihlajamki H. Orthotic insoles do not 1976). 2007;32(4):487-494.
8. Maher CG. A systematic review of workplace prevent physical stress-induced low back pain. Eur 38. Schwellnus MP, Jordaan G, Noakes TD.
interventions to prevent low back pain. Aust J Spine J. 2011;20(1):100-104. Prevention of common overuse injuries by the use
Physiother. 2000;46(4):259-269. 24. Milgrom C, Finestone A, Lubovsky O, Zin D, of shock absorbing insoles: a prospective study. Am
9. Linton SJ, van Tulder MW. Preventive Lahad A. A controlled randomized study of the J Sports Med. 1990;18(6):636-641.
interventions for back and neck pain problems: effect of training with orthoses on the incidence of 39. Soukup MG, Glomsrd B, Lnn JH, B K, Larsen
what is the evidence? Spine (Phila Pa 1976). 2001; weight bearing induced back pain among infantry S. The effect of a Mensendieck exercise program as
26(7):778-787. recruits. Spine (Phila Pa 1976). 2005;30(3):272-275. secondary prophylaxis for recurrent low back pain:
10. Chuter V, Spink M, Searle A, Ho A. The 25. Moore C, Holland J, Shaib F, Ceridan E, a randomized, controlled trial with 12-month
effectiveness of shoe insoles for the prevention and Schonard C, Marasa M. Prevention of low back pain follow-up. Spine (Phila Pa 1976). 1999;24(15):1585-
treatment of low back pain: a systematic review in sedentary healthy workers: a pilot study. Am J 1591.
and meta-analysis of randomised controlled trials. Med Sci. 2012;344(2):90-95. 40. Soukup MG, Lnn J, Glomsrd B, B K, Larsen
BMC Musculoskelet Disord. 2014;15:140. S. Exercises and education as secondary prevention

jamainternalmedicine.com (Reprinted) JAMA Internal Medicine Published online January 11, 2016 E9

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016

 Research Original Investigation Prevention of Low Back Pain

for recurrent low back pain. Physiother Res Int. 45. Bigos SJ, Holland J, Holland C, Webster JS, randomized controlled clinical trials. Eur Spine J.
2001;6(1):27-39. Battie M, Malmgren JA. High-quality controlled 2012;21(12):2520-2530.
41. van Poppel MN, Koes BW, van der Ploeg T, Smid trials on preventing episodes of back problems: 50. Donchin M, Woolf O, Kaplan L, Floman Y.
T, Bouter LM. Lumbar supports and education for systematic literature review in working-age adults. Secondary prevention of low-back pain: a clinical
the prevention of low back pain in industry: Spine J. 2009;9(2):147-168. trial. Spine (Phila Pa 1976). 1990;15(12):1317-1320.
a randomized controlled trial. JAMA. 1998;279(22): 46. Bell JA, Burnett A. Exercise for the primary, 51. van Duijvenbode IC, Jellema P, van Poppel MN,
1789-1794. secondary and tertiary prevention of low back pain van Tulder MW. Lumbar supports for prevention
42. Warming S, Ebbehj NE, Wiese N, Larsen LH, in the workplace: a systematic review. J Occup and treatment of low back pain. Cochrane Database
Duckert J, Tnnesen H. Little effect of transfer Rehabil. 2009;19(1):8-24. Syst Rev. 2008;(2):CD001823.
technique instruction and physical fitness training 47. Dawson AP, McLennan SN, Schiller SD, Jull GA, 52. Sahar T, Cohen MJ, Neeman V, et al. Insoles for
in reducing low back pain among nurses: a cluster Hodges PW, Stewart S. Interventions to prevent prevention and treatment of back pain. Cochrane
randomised intervention study. Ergonomics. 2008; back pain and back injury in nurses: a systematic Database Syst Rev. 2007;(4):CD005275.
51(10):1530-1548. review. Occup Environ Med. 2007;64(10):642-650.
53. Tooms RE, Griffin JW, Green S, Cagle K. Effect
43. McKenzie R. Treat Your Own Back. Waikanae, 48. Lahad A, Malter AD, Berg AO, Deyo RA. The of viscoelastic insoles on pain. Orthopedics. 1987;10
New Zealand: Spinal Publications; 1987. effectiveness of four interventions for the (8):1143-1147.
44. Macedo LG, Elkins MR, Maher CG, Moseley AM, prevention of low back pain. JAMA. 1994;272(16):
1286-1291. 54. Faun P, Klund S, Andreasen I, Jrgensen U.
Herbert RD, Sherrington C. There was evidence of Soreness in lower extremities and back is reduced
convergent and construct validity of Physiotherapy 49. Demoulin C, Marty M, Genevay S, by use of shock absorbing heel inserts. Int J Sports
Evidence Database quality scale for physiotherapy Vanderthommen M, Mahieu G, Henrotin Y. Med. 1993;14(5):288-290.
trials. J Clin Epidemiol. 2010;63(8):920-925. Effectiveness of preventive back educational
interventions for low back pain: a critical review of

E10 JAMA Internal Medicine Published online January 11, 2016 (Reprinted) jamainternalmedicine.com

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a University Library of Southern Denmark User on 01/12/2016