REVIEW OF LITERATURE

BACK BELT USE FOR PRELOW BACK PAIN: A SYST

Carlo Ammendolia, DC,a Michael S. Kerr, PhD,b a

ABSTRACT

Background: Back pain continues to be the lead

Recently, there has been a renewed interest in the us

Objectives: To examine the literature and evalu

occupational LBP.

Methods: MEDLINE, CINAHL, EMBASE, and HEALTH

2003. Studies were included if participants were material han

ause

lace.

the

e of

is a

ntify

and critically appraise published epidemiologic studies

evaluating back belt use. Although there have been

previous reviews on this topic, this review differs in that

it evaluates both clinical trials and observational studies and

were not included in other recent reviews on the topic.

a ReseCanada; Research

of Medical Sciences, Faculty of Medicine,

for Work and Health, 481 University Ave, Suite 800, Toronto,Ontario, Canada M5G 2E9 (e-mail: cammendolia@iwh.on.ca). In Canada, like other industrialized countries, back

injuries account for 20% to 30% of all lost time claims,

the largest single claims category in most workers6,7

Paper submitted September 25, 2003; in revised form December12, 2003.0161-4754/$30.00includes the largest and most recent randomized controlled

trial (RCT)4 evaluating back belts. The inclusion of

observational studies was deemed important because of

the limited number of published high-quality trials, the

difficulty in conducting such trials in the workplace, and

the existence of influential cohort studies on the topic that3,5

Health, Toronto, Canada.c Professor of Medicine and Rheumatology Division Director,

University of Toronto, Toronto, Canada; Senior Scientist, Institutefor Work and Health, Toronto, Canada; Director, Clinical Decision-Making and Health Care Division, University Health Network,Toronto, Canada; Rheumatologist, Mount Sinai Hospital, Toronto,Canada; Reasearch Chair.Submit requests for reprints to: Carlo Ammendolia, DC, InstituteUniversity of Toronto, Canada.b Assistant Professor, School of Nursing, Unversity of Western

Ontario, London, Ontario, Canada; Scientist, Institute for Work andCanadian MemorialCandidate, Institute128

Copyright D 20doi:10.1016/j.jmsystematic review of the literature, conducted to ideAssociate, Graduate Education and Research,Chiropractic College, Toronto, Canada; PhDlost time of reported LBP among workers who wore back belts compared with those who did not. The quality of the

evidence was scored independently by 2 reviewers using a double rating method, first according to research design

followed by an internal validity rating. Final synthesis of the evidence was performed in which the evidence was classified

as good, fair, conflicting, or insufficient.

Results: Ten epidemiologic studies meeting inclusion criteria were identified. Of 5 randomized controlled trials,

3 showed no positive results with back belt use; 2 cohort studies had conflicting results; and 2 nonrandomized controlled

studies and 1 survey showed positive results.

Conclusions: Currently, because of conflicting evidence and the absence of high-quality trials, there is no conclusive

evidence to support back belt use to prevent or reduce lost time from occupational LBP. (J Manipulative Physiol Ther

2005;28:128-134)

Key Indexing Terms: Low Back Pain; Prevention; Effectiveness; Back Belts; Review Literature

Back pain continues to be the leading overall c

of morbidity and lost productivity in the workp

Recently, there has been a renewed interest in

use of back belts by industry to reduce the incidenc

occupational low back pain (LBP).1-3 This paperarch Associate, Institute for Work and Health, Toronto,05 by National University of Health Sciences.pt.2005.01.009fectiveness of back belt use for the primary prevention of

STAR were searched for relevant articles published up to July

dlers, and outcomes included the incidence and/or duration ofVENTION OF OCCUPATIONALEMATIC REVIEW

nd Claire Bombardier, MDc

ing overall cause of morbidity and lost productivity in the workplace.

e of back belts by industry to reduce occupational low back pain (LBP).

ate the efcompensation jurisdictions. Disability resulting from

LBP is the most common chronic health problem in adults

Ammendolia et alJournal of Manipulative and Physiological Therapeutics

Back Belts for Low Back Pain PreventionVolume 28, Number 2129younger than 45 years and second only to arthritis in those

aged 45 to 65 years.8 In terms of costs, LBP is estimated to

be the most costly ailment of working age adults.9,10

Table 1. Levels of evidence and quality ratings of individualstudiesT

Levels of evidence

A. Research design rating

I Evidence from at least 1 RCT

II-1 Evidence from controlled trial(s) without randomization

II-2 Evidence from cohort or case-control analytic studies,

preferably from more than one center or research group

II-3 Evidence from comparisons between times or places with or

without the intervention; dramatic results from uncontrolled

studies could be included here

III Opinions of respected authorities, based on clinical

experience; descriptive studies or reports of expert committees

B. Quality (internal validity) ratingy

Good A study that meets all design-specific criteria well

Fair A study that does not meet (or it is not clear that it meets) at

least one design-specific criterion but has no known bfatal flawQPoor A study that has at least one design-specific fatal flaw, or an

accumulation of lesser flaws to the extent that the results

of the study are not deemed able to inform conclusions

T The Canadian Task Force methodology is described in Woolf et al29 oravailable from the Canadian Task Force Web site (http:/www.ctfphc.org),

under History and Methods.

y General design-specific criteria by study type are outlined in Harriset al.30There are 3 main categories of potential risk factors for

occupational LBP: individual, biomechanical, and psycho-

social. For a given individual, the strongest risk factor is a

previous history of LBP, along with the severity of the

previous episode.9

The most consistent associations among biomechanical

risk factors have been exposure to lifting or carrying heavy

loads; whole body vibration; and frequent bending and

twisting.1,9,11-14 Although the literature on psychosocial risk

factors for LBP is less consistent,1,14-17 there is growing

empirical evidence linking psychosocial stressors, such as

perceived high workload, time pressure, lack of intellectual

discretion, and job dissatisfaction, with an increased risk of

occupational LBP.9,11,15,18-20

There are 2 primary prevention strategies for occupa-

tional LBP: those directed at the individual worker and

those directed at the workplace. Workplace strategies

involve modification of the work site to suit the worker,

such as ergonomic job redesign.

Programs directed at the worker are by far the most

common preventive strategy in industry.21 Such programs

include education (eg, improving knowledge regard-

ing lifting techniques and injury awareness), exercise

(eg, improving strength and overall fitness), preemployment

screening efforts to detect risk factors such as smoking,

obesity, previous LBP, and psychosocial factors, and the use

of mechanical back supports (eg, belts or corsets).13,17,21,22There are no standards for back belts with respect to

materials or design. For example, many are modeled on

weight-lifting belts and can be made of plastic or elastic. All

are designed to support the lumbar spine and abdomen.

There are a number of mechanisms of action postulated; the

belts may prevent back pain by providing:

1. An increase in abdominal pressure, which provides an

extensor moment, thus decreasing the force required

by the back muscles and relieving the compressive

forces on lumbar intervertebral disks.23,24

2. An increase in spinal rigidity by limiting end range

movement, which protects the spine from extreme

ranges of movement, thereby minimizing shearing

forces.25,26

3. An increased proprioceptive awareness (such as a

reminder for using proper safety measures) and an

increased sense of security.25,27,28

Despite the recent increase in their popularity as a

preventive measure, the effectiveness of mechanical back

supports such as back belts remains uncertain. The main

objectives of this review are to evaluate the evidence on the

effectiveness of back belt use for primary prevention or

reduction of lost time from occupational LBP and to develop

clinical practice recommendations for primary health care

providers for their use.

METHODS

The published English literature up to July 2003 was

identified with a computerized search of MEDLINE,

Table 2. Main methodologic quality criteria

Internal validity criteria

1. Was there appropriate assembly and maintenance of comparable

groups?

2. Was there adequate follow-up?

3. Were interventions clearly defined?

4. Were equal, reliable and valid outcome measures used?

5. Were the analyses/sample size appropriate and was intention-to-

treat analysis used?CINAHL, EMBASE, and HEALTHSTAR databases using

the following keywords: back, lumbar, spine, belts, supports,

braces, orthotic devices, prevention, and occupational.

Pertinent references from articles obtained from the above

search were reviewed. Studies were included if the study

participants were material handlers (eg, exposed to lifting),

and outcome measures included the incidence and/or

duration of lost time of reported LBP among workers who

wore back belts compared with those who did not. Although

the primary population of interest is asymptomatic workers,

studies that included workers with a previous history of LBP

were not excluded. No restriction was made on the style of

back belt used.

The evidence was reviewed systematically using the

US Preventive Services Task Force.30 The internal validity

c

ri

130 Journal of Manipulative and Physiological TherapeuticsAmmendolia et alFebruary 2005Back Belts for Low Back Pain Preventionof each study was assessed independently by 2 authors

(CA, MK). The main internal validity items are summarized

in Table 2. Consensus was used to resolve disagree-

ments in scoring.

RESULTS

Ten epidemiologic studies meeting inclusion criteria

were identified. These include 5 RCTs,4,31-34 2 non-

RCTs,35,36 2 cohort studies,37,38 and 1 survey.39 A summary

of the study quality evaluations is shown in Table 3.methodology of the Canadian Task Force on Preventive

Health Care (Table 1).29 The quality of the evidence was

scored using a double rating method first according to

research design, RCTs were given the highest rating (level 1

evidence), followed by an internal validity rating (good, fair,

or poor) using design-specific criteria developed with theTable 3. Methodologic quality of studies on the effectiveness of ba

Internal validity crite

Study 1 2 3

Walsh and Schwartz33 + ?Reddell et al31 + + +

Alexander et al34 + +Van Poppel et al32 ? + +

Kraus et al4 ? + ?

Anderson, 199335 + +Thompson et al36 +Wassel et al38 ? + +

Kraus et al37 + +Mitchell et al39 + ++, criterion met; , criterion not met; ?, not reported.Controlled TrialsWalsh and Schwartz33 randomly assigned 90 warehouse

workers into 3 equal groups to receive either, no inter-

vention, a 1-hour training session in lifting techniques and

back pain prevention, or 1 hour of training and a back belt

for use during working hours. No group was assigned the

back belt only.

The results revealed a significant decrease in lost time

(2.5 days) in the group receiving training plus back belts

(P = .03). In the subgroup analysis, the authors suggest the

reduction in lost time seen in the group receiving training

plus back belts was limited to workers with previous LBP

(P = .02). The main weaknesses of this study include the

lack of adjustment for the apparent baseline differences in

days lost among the groups and the failure to assess group

similarity with respect to history of LBP at baseline.

Reddell et al31 studied 642 out of an initial 896 selected

airline baggage handlers who were randomly assigned to4 groups: back belt only; 1-hour training class (on proper

lifting) only; back belt and 1-hour training class; and no

intervention. There were no significant differences in injury

rates, lost workdays, or Workers Compensation Board costs

among the groups. However, the results are inconclusive

because patients were analyzed according to use of back

belts, and not on group assignment; there was high

noncompliance (58% discontinued back belt use) and high

nonparticipation rate (28%).

Alexander et al34 randomly assigned 60 health care

workers to wear a belt or to be in a control group and then

followed them for 3 months. There were only 3 self-reported

low back injuries at follow-up, 1 in the belted group and 2 in

the control. The difference was not statistically significant

(P = .53). The small number of injuries in this study

suggests the sample size may have been too small or the

follow-up period was too short to detect any significant

difference in injury rates between the 2 groups.

In another RCT, van Poppel et al32 randomly assigned a

total of 315 airline cargo workers (within preexisting work

groups) to 1 of 4 groups: education and lumbar support;

k belts

on

4 5 Study design Overall rating

+ ? I Fair

I Poor? ? I Fair

? + I Fair

? + I Fair

+ ? I-2 Poor

? ? I-2 Poor

? + II-2 Good

? II-3 Poor ? III Pooreducation only; lumbar support only; or no intervention.

After 6 months there was no significant difference in self-

report LBP incidence (risk difference 1%; 95% CI, 10 to13) or sick leave for LBP (risk difference 4%; 95% CI,

3 to 11) among workers in the groups assigned to wearlumbar supports compared with those who were not. In a

small subgroup of workers with LBP at baseline (15%), the

group with lumbar supports had fewer days with LBP per

month than did the group without (mean of 3.1 vs 8.4 days,

P = .03). A limitation of the study was the randomization

process, where assignment was made by workgroups

instead of the individual worker, which may have poten-

tially introduced confounding if there were undetected

systematic differences between the work groups. In addi-

tion, there was potential recall bias and selection bias due to

high noncompliance.

In the most recent RCT, Kraus et al4 randomized

9 agencies, employing 12,772 home attendants, into

In a retrospective survey among workers at an air force39

Ammendolia et alJournal of Manipulative and Physiological Therapeutics

Back Belts for Low Back Pain PreventionVolume 28, Number 2131base, Mitchell et al reported a marginally significant

protective effect with back belt use (OR 0.60; 95% CI,

0.36-1.00) for the first low back injury. The limitations in

this study include the lack of randomization, back belt users

and nonusers were not matched in time, and the use of self-

report data on back belt use and injuries.

In a historical cohort study, Kraus et al compared the low

back injury rates of 36,000 retail workers before and after

mandatory back support use policy periods implemented

over a 6-year period.37 The results indicated an incidence

density ratio (or ratio of the number of low back injury

claims per million working hours before and after back sup-

port policy implementation) of 1.52 (95% CI, 1.36-1.69).

The authors reported that 34% of injuries could have been

prevented if all subjects had worn a back support.

The conclusions of the Kraus study must be interpreted

with caution. Because the comparison groups were not

matched in time, it was not possible to control for unknown

confounders such as a change in hiring practices, job duties,

claims handling, safety regulations, and workers compen-

sation claims policies. In addition, the gradual increase in

the use of forklifts and pallets during the 6-year study could

have been an important co-intervention contributing to the

significant decrease in the incidence density ratio.

The largest prospective study on back belt use was3 groups: back belts only, lifting advice only, and control.

Over a 28-month period the authors reported a marginally

significant lower back injury rate among the back belt group

when compared with the control group (rate ratio 1.36; 95%

CI, 1.02-1.82). There was no significant difference when the

back belt group was compared with the advice-only group

(rate ratio 1.22; 95% CI, 0.70-2.11). The limitation to this

study was the inability to adjust for cluster randomization

and known risk factors when comparing the back belt and

control groups because the control group data were based on

crude (unadjusted) injury data only because of a lack of

baseline data. Therefore, it is not certain if the difference

observed was due to confounding. Other weaknesses

include no comparison of preintervention injury rates and

no data on previous LBP.

In 2 nonrandomized trials, Anderson et al35 found a

reduction in back injury rates and Thompson et al36 found a

decrease in back injuries and back pain, respectively, among

workers who wore a back support compared with those who

did not. The positive results found in these 2 nonrandomized

trials raise questions because the groups being compared

were not assessed for factors associated with increased risk

of LBP. Known and unknown confounding factors within

work sites may have influenced the results.

Observational Studiesconducted by Wassel et al38 who compared the incidence

of low back injuries and self-report LBP among a sampleof 13,873 material handlers employed at either 1 of

89 retail stores who had a mandatory back belt policy or

1 of 71 stores that had a voluntary policy. When

controlling for potential confounding factors the authors

found no difference in back injury claims (OR 1.22; 95%

CI, 0.87-1.70) or self-report episodes of LBP (OR 0.97;

95% CI, 0.83-1.13) among the stores with mandatory or

voluntary back belt use policy.

The limitations of this study included the high non-

participation rate (32%) and the high number of workers

who did not complete follow-up interviews (33%). Lack of

randomization with the inability to control for unknown

confounders and recall bias from self-report interviews may

have been other potential sources of bias.

Potential HarmsPotential negative effects of back belt use were discussed

in most studies that were reviewed. In 1 trial, 20% of

workers felt that the belt rubbed, pinched, or bruised their

ribs; 15% stated the belt caused problems during sitting or

driving; and 20% said that it was too hot or caused excessive

sweating.31 Reduced movement, uncomfortable sitting, and

excessive heat were complaints expressed by workers in

another study.32 The negative comments may have con-

tributed to the high noncompliance rate for back belt use in

these 2 studies.31,32 Post hoc analysis of noncompliant back

belt users in 1 study31 indicated a significantly higher

number of lost workdays (P b .0181) when compared withcompliant back belt or control groups. However, a similar

finding was not observed in the compliance subgroup

analysis from another study.32 Two studies evaluating

abdominal strength change found no significant loss of

abdominal strength among back belt users.32,33

Synthesis of Key EvidenceOf 5 RCTs reviewed, 3 failed to show positive results

with back belt use.31,32,34 One RCT showed decreased time

loss in workers who received both training and used a back

belt.33 A review of their subgroup analysis suggests this

effect is seen only among workers with a previous history of

LBP. Another RCT reported a marginally significant

decrease in low back injury rates among employees

receiving a back belt compared with the control group.4

This analysis however was based on unadjusted data not

controlling potential confounders. All RCTs reviewed had

methodological flaws, some of which, such as lack of

blinding, are inherent to workplace studies and the type of

intervention used, whereas others, such as inappropriate

randomization, lack of intention-to-treat analysis, and

inadequate follow-up times, are related more to the study

designs. A priori sample size calculation for lost time was

conducted in only one RCT.32 Based on this calculation, only2 RCTs appeared to have had a sufficient sample size.31,32

Only one RCT performed a sample size calculation for the

When considering only good- and fair-quality studies the

The publication of the Canadian Centre for Occupational

Health and Safety regarding back belt use44 refers to the

132 Journal of Manipulative and Physiological TherapeuticsAmmendolia et alFebruary 2005Back Belts for Low Back Pain Preventionevidence remains conflicting. There were 2 level I studies

evaluating lost time from LBP with back belt use, both rated

fair quality. One showed a negative result,32 the other

positive.33 However, for the prevention of LBP the evidence

appears weighted against the use of back belts with

4 negative studies (3 level I studies of fair quality32-34 and

1 level II-2 study of good quality)38 and only 1 positive

study (a level I study of fair quality).

In the subgroup analysis of 2 fair-quality trials, work-

ers with a previous history of LBP33 or LBP at baseline32

had significantly reduced time off if randomized to wear a

back belt.

DISCUSSION

Based on this review, the evidence for the effectiveness

of back belt use in preventing the incidence or reducing lost

time for occupational LBP among material handlers is

conflicting and limited in both quantity and quality.

In contrast, in 2 recent literature reviews of back belt

clinical trials, 1 concluded there was moderate evidence3

and the other strong and consistent evidence5 that lumbar

supports are not effective for primary prevention. Another

recent review of back belts, which included both clinical

trials and cohort studies published since 1995, found the

evidence for primary prevention to be inconclusive.40 None

of these reviews included the most recent positive RCT.4

Moreover, in one of these reviews,3 2 of the 7 trials

included32,33 were considered to be of high quality, which

was not concordant with our quality designations. Both

these trials were downgraded to fair quality, 132 because of

inappropriate randomization (ie, work-group level only) andincidence of low back injuries.4 This RCT had an 80% power

to detect a 30% or more decline in the low back injury rate.

No sample size calculations were performed for assessing

change in the incidence of LBP in any of the reviewed RCTs.

Sample size estimates for this outcome, based on mean

incidence values found in 1 trial,32 suggests only 2 trials31,32

had sufficient sample size to detect a reduction in the

incidence rate of less than 70% (using a power of 80%).

None of the RCTs reviewed were considered to be of

bgoodQ quality. Four were considered of bfairQ quality4,32-34

and 131 was considered bpoor.QThe only study with a good-quality rating was the recent

large cohort.38 This study failed to show any benefit to back

belt use for the main outcomes or in the subgroup analysis.

Lost time in this study was not assessed.

The remaining studies found positive results with back

belt use.35-37,39 These studies, however, were found to have

significant weaknesses in both methods and analysis, as

reflected in their poor-quality scores.high noncompliance and the other33 because of the lack of

comparability of the groups at baseline.NIOSH review43 and recommends that back belts should

not be used as a primary workplace prevention approach.

In contrast, the recent ergonomics regulation of the

Occupational Safety and Health Administration classified

lumbar supports as personal protective equipment and

suggests they may prevent back injuries in certain indus-

trial settings.45

Controversy over back belt use also extends into the area

of treatment, where results from RCTs are also conflict-

ing.3,46 However, the positive results demonstrated in

2 RCTs47,48 and in the subgroup analysis in 2 trials

reviewed here32,33 suggest that patients with a history of

LBP may benefit from back belt use. The lack of consistent

conclusions from the studies included for this review is not

surprising given the conflicting laboratory evidence on how

back belts are thought to prevent LBP.22-25,27,43,49

Conclusions from studies reporting adverse effects were

limited in both quantity and by the methods used. NIOSH

has suggested that a false sense of security may accompany

back belt use, which may lead to increased risk taking

behaviors (such as excessive lifting).50 Other potential risks

include cardiovascular strain,51,52 back muscle weakening,53

and abdominal hernia.23 However, these possible risks have

been extrapolated from studies evaluating various physio-

logical parameters such as intra-abdominal pressure, electro-

myogram, heart rate, and blood pressure, and we were

unable to find any epidemiologic evidence to support these

possible adverse effects of back belt use.

CONCLUSION

Implications for Clinical PracticeIn general, the majority of the evidence presented in this

review, and the evidence presented in earlier reviews of the

topic, indicates that individual workers presenting with no

prior history of LBP are unlikely to benefit from the use of a

back belt. Those with a previous history of LBP may

experience some potential benefit from back belt use.

However, before back belt prescription, individuals should

be screened for cardiovascular risk and receive training onEarlier reviews on the topic concluded there was

insufficient evidence to make recommendation for or

against the use of back belts for the prevention of occupa-

tional low back injuries.22,27,41,42

National Institute for Occupational Safety and Health

(NIOSH), following the results of their most recent back

belt study, concluded there was no evidence to support the

use of back belts as a preventive measure.38 This further

supports their previous review and recommendations on

back belt use.43lifting mechanics.28,52 Although there is some laboratory

evidence suggesting possible concern for the adverse effects

Carolina at Chapel Hill; Dr Harry Shannon, Program in

1997;23:243-56.2. National Institute for Occupational Safety and Health. No

2001;26:778-87.

Ammendolia et alJournal of Manipulative and Physiological Therapeutics

Back Belts for Low Back Pain PreventionVolume 28, Number 21336. Spengler DM, Bigos SJ, Martin NA, Zeh J, Fisher L,Nachemson AL. Back injuries in industry: A retro-spective study, I: Overview and cost analysis. Spine 1986;11:241-51.

7. National Work Injuries Statistics Program, Association ofevidence that back belts reduce injury seen in landmark studyof retail users. [Press Release]. Washington (DC)7 Centers forDisease Control and Prevention; 2000 [Cited 2000 Dec 7].Available from: http://www.cdc.gov/niosh/beltinj.html.

3. Jellema P, van Tulder MW, van Poppel MNM, Nachenson AL,Bouter LM. Lumbar supports for prevention and treatment oflow back pain. A systematic review within the framework ofthe Cochrane Back Review Group. Spine 2001;26:377-86.

4. Kraus JF, Schaffer KB, Rice R, Maroosis J, Harper J. A fieldtrial of back belts to reduce the incidence of acute low backinjuries in New York City home attendants. Int J OccupEnviron Health 2002;8:97-104.

5. Linton SJ, van Tulder MW. Preventive interventions for backand neck pain problems. What is the evidence? SpineOccupational Health and Environmental Medicine, McMas-

ter University, Hamilton, Ontario; and Dr Howard Vernon

from the Canadian Memorial Chiropractic College, Toronto,

Ontario, for reviewing a draft form of this report and

Dr Elaine Wang and Nadine Wathen for their guidance

and encouragement.

REFERENCES1. Burdorf A, Sorock G. Positive and negative evidence of risk

factors for of back disorders. Scand J Work Environ Healthof long-term use,28 these possible risks have not been

proven. However, given the combination of questionable

benefits and potential for negative effects, if back belts are

to be prescribed, it should only be for short-term use.28

Research AgendaThe overall level and quality of this evidence on the topic

remains limited and conflicting. Well-conducted RCTs into

the efficacy of back belts are still needed. Although it is

recognized that rigorous workplace effectiveness trials are

logistically very difficult to conduct, it is recommended that

future studies should include a large bat-riskQ population;individual worker randomization; appropriate control

groups; long follow-up; high compliance rate; and the use

of validated outcome measures, including a special focus on

those with a prior LBP history.

ACKNOWLEDGMENTS

We acknowledge Dr Timothy Carey, Cecil G. Sheps

Center for Health Services Research, University of NorthWorkers Compensation Boards of Canada Mississauga (Can-ada); 1999.8. Bigos SJ, Bowyer OR, Braen GR, Brown K, Deyo R. Acutelow back problems in adults. Clinical Practice Guidelines No14. Rockville (Md)7 Agency for Health Care Policy andResearch, Public Health Service, US Dept Of Health andHealth Services; 1994 [Publication no 95-0642].

9. Frank JW, Brooker A, DeMaio S, Kerr MS, Maetzel A,Shannon HS, et al. Disability resulting from occupational lowback pain part II: What do we know about secondaryprevention? A review of the scientific evidence on preventionafter disability begins. Spine 1996;21:2918-29.

10. Salkever DS. Morbidity cost: National estimates and economicdeterminants. NCHSR Research Summary Series; 1985 DHHS:Publ No (PHS) (86-3393):13.

11. Bernard BP, editor. Musculoskeletal Disorders and WorkplaceFactors. A Critical Review of Epidemiological Evidence forWork-related Musculoskeletal Disorders of the Neck, UpperExtremity, and Low Back. Cincinnati (Ohio)7 US Departmentof Health and Human Service, NIOSH; 1997. p. 6-34.

12. Hoogendoorn WE, van Poppel MNM, Bongers PM, Koes BW,Bouter LM. Physical load during work and leisure time as riskfactors for back pain. Scand J Work Environ Health1999;25:387-403.

13. Garg A, Moore JS. Prevention strategies and the low back inindustry. Occup Med 1992;7:629-40.

14. Kerr MS, Frank JW, Shannon HS, Norman RWK, Wells RP,Neumann WP, et al. Biomechanical and psychological riskfactors for low back pain at work. Am J Public Health 2001;91:1069-75.

15. Bigos SJ, Battie MC, Spengler DM, Fisher LD, Fordyce WE,Hansson TH, et al. A prospective study of work perceptionsand psychosocial factors affecting the report of back injury.Spine 1991;16:1-6.

16. Bongers PM, de Winter CR, Kompier MAJ, Hildebrandt VH.Psychosocial factors at work and musculoskeletal disease.Scand J Work Environ Health 1993;19:297-312.

17. Burdorf A. Exposure assessment of risk factors for disorders ofthe back in occupational epidemiology. Scand J Work EnvironHealth 1992;18:1-9.

18. Nachemson A, Vingard E. Influences of individual factors andsmoking on neck and low back pain. In: Nachemson A,Jonsson E, editors. Neck and Back Pain: The ScientificEvidence of Causes, Diagnosis and Treatment. Philadelphia(Pa)7 Lippincott Williams & Wilkins; 2000. p. 79-95.

19. Hoogendoorn WE, van Poppel MNM, Bongers PM, Koes BW,Bouter LM. Systematic review of psychosocial factors at workand private life as risk factors for back pain. Spine2000;25:2114-25.

20. Institute of Medicine. Musculoskeletal disorders and theworkplace: Low back and upper extremities. Washington(DC)7 National Academy Press; 2001.

21. Garg A, Moore JS. Epidemiology of low-back pain in industry.Occup Med: State of the Art Rev 1992;7:593-608.

22. Lahad A, Malter AD, Berg AO, Deyo RA. The effectiveness offour interventions for the prevention of low back pain. JAMA1994;272:1286-91.

23. Harman EA, Rosenstein RM, Frykman PN, Nigro GA. Effectsof a belt on intra-abdominal pressure during weight lifting.Med Sci Sports Exerc 1989;21:186-90.

24. Woodhouse M.L, McCoy RW, Redondo DR, Shall LM.Effects of back support on intra-abdominal pressure andlumbar kinetics during heavy lifting. Hum Factors 1995;37:582-90.

25. Van Poppel MNM, de Looze MP, Koes BW, Smid T, Bouter

LM. Mechanisms of action of lumbar supports. A systematicreview. Spine 2000;25:2103-13.

26. McGill S, Seguin J, Bennett G. Passive stiffness of the uppertorso in flexion, extension, lateral bending, and axial rotation:Effect of belt wearing and breath holding. Spine 1994;19:696-704.

27. Minor SD. Use of back belts in occupational settings. PhysTher 1996;76:403-8.

28. McGill S. Update on the use of back belts in industry: Moredata- same conclusions. In: Karwowski W, Marras W, editors.Occupational Ergonomics Handbook. Boca Raton (Fla)7 CRCPress; 1999. p. 1353-8.

29. Woolf SH, Battista RN, Anderson GM, Logan AG, Wang E.Members of the Canadian Task Force on the Periodic HealthExamination. Assessing the clinical effectiveness of preventivemanoeuvres: Analytic principals and systematic methods inreviewing evidence and developing clinical practice recom-

issued back belts in areas of high risk for back injury. J OccupMed 1994;36:90-4.

40. Gatty C, Turner M, Buitendorp D, Batman H. The effective-ness of back pain and injury prevention programs in theworkplace. Work 2002;20:257-66.

41. Van Poppel MN, Koes BW, Smid T, Bouter LM. A systematicreview of controlled clinical trials on the prevention of backpain in industry. Occup Environ Med 1997;54:841-7.

42. Karras BE, Conrad K. Back injury prevention interventionsin the workplace. Am Assoc Occup Health Nurs 1996;44:189-96.

43. National Institute for Occupational Safety and Health. Work-place use of back belts. Cinninati (OH): Centers for DiseaseControl and Prevention, US Dept of Health and HumanServices, 1994 [Pub No 94-122].

44. Canadian Centre for Occupational Health and Safety. Back

134 Journal of Manipulative and Physiological TherapeuticsAmmendolia et alFebruary 2005Back Belts for Low Back Pain Prevention30. Harris RP, Helfand M, Woolf SH, Lohr KN, Mulrow CD,Teutsch SM, et al. Current methods of the U.S PreventiveServices Task Force A review of the process. Am J Prev Med2001;20:21-35.

31. Reddell CR, Congleton JJ, Huchingson RD, Montgomery JF.An evaluation of a weightlifting belt and back injuryprevention training class for airline baggage handlers. ApplErgon 1992;23:319-29.

32. Van Poppel MNM, Koes BW, van der Ploeg T, Smid T, BoutarLM. Lumbar supports and education for the prevention of lowback pain in industry: A randomized controlled trial. JAMA1998;279:1789-94.

33. Walsh NE, Schwartz RK. The influence of prophylacticorthoses on abdominal strength and low back injury in theworkplace. Am J Phys Med Rehabil 1990;69:247-50.

34. Alexander A, Woolley SM, Bisesi M, Schaub E. Theeffectiveness of back belts on occupational back injuries andworker perception. Prof Saf 1995;40:22-6.

35. Anderson CK, Morris TL, Vechin DC. The effectiveness ofusing lumbar support belt. Dallas (Tex)7 Advanced Ergonom-ics; 1993.

36. Thompson L, Pati AB, Davidson H, Hirsh D. Attitudes andback belts in the workplace. Work 1994;4:22-7.

37. Kraus JF, McArthur DL, Samaniego L. Reduction of acute lowback injuries by use of back supports. Int J Occup EnvironHealth 1996;2:264-73.

38. Wassell JT, Gardner LI, Landsittel DP, Johnston JJ, JohnstonJM. A prospective study of back belts for prevention of backpain and injury. JAMA 2000;284:2727-32.

39. Mitchell LV, Lawler FH, Bowen D, Mote W, Asundi P,Purswell J. Effectiveness and cost-effectiveness of employer-belts. Document 0365I.wpf 250; Aug 11, 1995.45. Occupational Safety and Health Administration. Ergono-

mic program: Final rule. Fed Regist 2000;(November 14):68262-870.

46. Koes BW, van den Hoogen HMM. Efficacy of bed rest andorthosis on low back pain. A review of randomised clinicaltrials. Eur J Phys Med Rehabil 1994;4:86-93.

47. Million R, Nilsen KH, Jayson MIV, Baker RD. Evaluation oflow back pain and assessment of lumbar corset with andwithout back supports. Ann Rheum Dis 1981;40:449-54.

48. Valle-Jones JC, Walsh H, OHara H, Davey NB, Hopkin-Richards H. Controlled trail of a back support (lumbo-train) inpatients with non-specific back pain. Curr Med Res Opin1992;12:604-12.

49. Genaidy AM, Simmons RJ. Can back supports relieve the loadon the lumbar spine for employees engaged in industrialoperations? Ergonomics 1995;38:996-1010.

50. National Institute for Occupational Safety and Health (NIOSH).Back belts: Do they prevent injury? Centers for DiseaseControl and Prevention, US Dept Of Health and HumanServices, 1994 [Pub No 94-127].

51. Hunter GR, McGuirk J, Mitrano N, Pearman P, Thomas B,Arrington R. The effects of a weight training belt on bloodpressure during exercise. J Appl Sport Sci Res 1989;3:13-8.

52. Perkins MS, Bloswick DS. The use of back belts to increaseintra abdominal pressure as a means of preventing low backinjuries: A survey of the literature. Int J Occup Environ Health1995;1:326-35.

53. Holmstrom E, Moritz U. Effect of lumbar belts on trunkmuscle strength and endurance: A follow-up study of con-struction workers. J Spinal Disord 1992;5:260-6.mendations. J Clin Epidemiol 1990;43:891-905.

Back belt use for prevention of occupational low back pain: A systematic reviewMethodsResultsControlled trialsObservational studiesPotential harmsSynthesis of key evidence

DiscussionConclusionImplications for clinical practiceResearch agenda

AcknowledgmentsReferences