Key Points
Question
Does adding a nonrigid lumbar belt to usual care improve disability and pain in adults with nonspecific low back pain?
Findings
In this randomized clinical trial of 168 adults with nonspecific low back pain lasting 1 to 6 months, the belt group had greater improvement in Oswestry Disability Index at week 12 than usual care. Mean pain at rest and during activity also decreased more in the belt group, and fewer belt-group patients used back pain medication.
Meaning
These findings suggest nonrigid lumbar belts may be a nonpharmacological option to help manage pain and functional disability in non-specific low back pain.
This randomized clinical trial tests whether a soft and extensible lumbar belt reduces functional disability and pain in patients with nonspecific low back pain.
Abstract
Importance
Nonspecific low back pain is highly prevalent and disabling. Most current guidelines state that evidence is insufficient to recommend lumbar belts for this condition.
Objective
To determine whether a soft and extensible lumbar belt reduces functional disability and pain in patients with nonspecific low back pain.
Design, Setting, and Participants
This multicenter, open-label, randomized clinical trial was conducted at 17 medical centers across France from February 26, 2021, to March 28, 2024. Participants were adults with nonspecific low back pain lasting 1 to 6 months. Data were analyzed from August 7, 2024, to June 18, 2025.
Interventions
Patients were randomized using an interactive web response system to wear a nonrigid lumbar belt for 12 weeks in addition to usual care or to receive usual care without a device.
Main Outcomes and Measures
The primary outcome was the between-group difference in mean change in the Oswestry Disability Index (ODI) score from baseline to week 12. Outcomes were assessed at baseline, week 4, and week 12; pain was also reported weekly using online questionnaires.
Results
At total of 168 patients were randomized (mean [SD] age, 49.0 [13.7] years; 101 women [60.1%]), with 86 participants in the belt group and 82 participants in the usual care group. The primary analysis included 155 patients with at least 1 postrandomization assessment (81 patients in the belt group; 74 patients in the control group). ODI improvement from baseline to week 12 was greater in the belt group (change, −10.0 [95% CI, −13.0 to −7.0] points) than in the control group (change, −5.3 [95% CI, −8.6 to −2.1] points), with a between-group difference of −4.7 (95% CI, −8.4 to −0.9) points (P = .01). Reductions in pain at rest and during activity were also greater in the belt group, with between-group differences of −8.7 (95% CI, −16.1 to −1.3) points (P = .02) and −10.0 (95% CI, −18.3 to −1.6) points (P = .02), respectively. Over 12 weeks, medication use was lower in the belt group than in the control group (41 patients [50.6%] vs 50 patients [67.6%]; P = .03). No serious device-related adverse events occurred.
Conclusions and Relevance
In this randomized clinical trial, adding a nonrigid lumbar belt to usual care resulted in improved pain and functional disability in patients with nonspecific low back pain lasting 1 to 6 months. These findings suggest that lumbar belts may be considered as a nonpharmacological option for symptom management.
Trial Registration
ClinicalTrials.gov Identifier: NCT04701073
Introduction
Low back pain (LBP) is a leading cause of disability worldwide, and up to 80% of individuals experience LBP during their lifetime.1 Although acute LBP often improves within 6 weeks, recurrence can occur in up to one-third of individuals by 12 months,2 affecting quality of life.3 LBP also carries a substantial economic burden due to health care costs and productivity loss.2,4
Approximately 90% of patients with LBP are diagnosed with nonspecific LBP, defined as pain not attributable to a recognizable, specific pathology.2,5 Current evidence supports a biopsychosocial approach, as pain and disability are influenced by biological, psychological, social, and lifestyle factors, including physical inactivity, smoking, and obesity.2,4,6 Management generally emphasizes education, advice to remain active, exercise therapy, physical therapy, and, when appropriate, psychological or multidisciplinary interventions. Pharmacological options, such as nonsteroidal anti-inflammatory drugs (NSAIDs), may provide short-term pain relief but may be limited by adverse effects.7
Lumbar supports, which may be rigid, semirigid, or nonrigid or soft and vary in extensibility or elasticity,8 are commonly used as adjunctive nonpharmacological options. Proposed mechanisms include improved postural control, proprioceptive feedback, and reduced mechanical load,9,10,11 although the clinical relevance of these effects remains uncertain.
Evidence supporting lumbar supports remains inconsistent. Studies have varied in device type, population, duration of use, and outcome measures,12 limiting interpretation and meta-analysis. European guidelines either do not recommend lumbar supports or do not make a recommendation,13,14 and recent World Health Organization guidelines conditionally recommend against routine use of braces, belts, or supports for chronic LBP because of very low-certainty evidence and potential harms with long-term use, including dependence, fear-avoidance, and deconditioning.15 However, concerns that lumbar supports reduce muscle strength or delay recovery have not been consistently supported.16 Some randomized trials and reviews suggest that soft and extensible belts may improve disability and pain in selected patients.17,18,19,20,21 Therefore, further randomized evidence is needed to clarify whether short-term use of a nonrigid lumbar belt as an adjunct to usual care provides clinically relevant benefits without safety concerns.
This randomized clinical trial assessed the effectiveness and safety of an adjustable, nonrigid lumbar belt for the treatment of nonspecific LBP. We hypothesized that wearing the lumbar support, in addition to usual care, would improve functional disability at 12 weeks.
Methods
Study Design
This prospective, multicenter, open-label, parallel group, randomized clinical superiority trial was conducted from February 26, 2021, to March 28, 2024, in 17 medical centers across France. The study was approved by the Committee for the Protection of Persons Ile de France IV, and registered on ClinicalTrials.gov on January 7, 2021. All participants provided written informed consent. The trial protocol and statistical analysis plan are provided in Supplement 1. The trial is reported according to the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.22
Study Participants
Eligible patients were aged 18 to 75 years, had nonspecific LBP lasting 1 to 6 months, were covered by French social security, and reported mean pain at rest or during activity over the previous 72 hours of at least 40 out of 100 on a visual analogue scale (VAS). Exclusion criteria included wearing a lumbar belt on the inclusion day, spinal or lower limb surgery in the previous 2 years or planned during the study, epidural corticosteroid injection within 1 month or planned during the study, lumbar radicular syndrome, pregnancy or planned pregnancy, cognitive or psychiatric disorder compromising participation or assessment, size outside the belt range, allergy to belt materials, contraindication to belt use, or participation in another clinical trial affecting outcomes.
Before inclusion, physicians recorded clinical and demographic characteristics, and classified nonspecific LBP as subacute, recently chronic, or recurrent. Recurrent LBP was defined pragmatically as recurrent painful episodes or re-exacerbation in patients with a known history of LBP.
Intervention and Randomization
The investigated device was an adjustable, nonrigid lumbar belt (LombaStab; Thuasne). It has 2 back panels connected to front panels by elasticated fabric and handles to adjust tightness. Thermal sensors were attached to objectively assess daily belt wear.
Eligible patients were randomized 1:1 at baseline to the belt or usual care group via a centralized interactive web response system using site-stratified, computer-generated lists with permuted blocks of 4. The allocation sequence was generated and implemented independently of enrolling physicians, who had no access to the sequence. The trial was not blinded.
Patients allocated to the belt group were instructed to wear the belt during daytime activities for 4 to 8 hours per day throughout the 12-week study period, adjust it for comfortable support without excessive tightness, and not wear it while sleeping. Both groups were advised to remain physically active, use usual pain medication, and to avoid oral or epidural corticosteroids during the study.
Follow-up visits occurred at 4 weeks (T1) and 12 weeks (T2). Patients also completed weekly online questionnaires.
Outcomes
Functional disability was assessed using the Oswestry Disability Index (ODI; French version 2.1) at baseline, T1, and T2.23 The primary outcome was the between-group difference in absolute change in ODI score from baseline to T2. Additional categorical ODI outcomes included improvement of at least 30%, improvement of less than 30%, and no improvement or worsening.
Minimum, maximum, and mean pain intensity over the previous 72 hours, at rest and during activity, were assessed using a 0 to 100 VAS at baseline, T1, and T2. Mean pain at rest and during activity over the previous 72 hours was also assessed weekly. Physical activity was assessed using the short International Physical Activity Questionnaire (IPAQ-SF).24 Physicians assessed lumbopelvic-femoral mobility with the finger-to-floor test and lumbosacral mobility with the Schober-Macrae test.25,26
Global change was assessed from clinician and patient perspectives. Physicians assessed overall condition using the Clinical Global Impression-Severity scale (CGI-S) and improvement using the Clinical Global Impression-Improvement scale (CGI-I).27 Patient-perceived change was assessed using the Patient Global Impression of Change (PGI-C).28 Quality of life was assessed using EuroQol (EQ-5D).29 Medication use, consultations, sick leave, and adverse events were recorded. At T2, the belt group completed an online questionnaire rating belt satisfaction and acceptability from 0 to 10, with higher score indicating greater satisfaction and acceptability.
Sample Size
The sample size calculation used a target between-group ODI difference of 10 points, assumed SD of 20, 90% power, and 2-sided α = .05, requiring 86 patients per group. Assuming 15% dropout, 203 patients were planned. The 10-point value was a design assumption and not an established minimally important between-group difference.
Statistical Analyses
Analyses were performed in the intention-to-treat population (ITT), defined as all randomized patients, the full analysis set (FAS; defined as randomized patients with ≥1 postrandomization efficacy assessment), the completers set (defined as FAS patients completing T2 with ODI available), and the per-protocol population (defined as FAS patients without major protocol deviations). The primary analysis used the FAS. Missing T2 ODI values were imputed using last observation carried forward. The absolute ODI change from baseline to T1 and T2 was compared using a mixed model for repeated measures, implemented in SAS software version 9.4 (SAS Institute) using the PROC MIXED process, including treatment group, visit, and treatment × visit interaction as fixed effects, with baseline ODI and center as covariates. A participant-specific random intercept was included to account for between-participant variability, and the within-participant covariance across visits was modeled using an unstructured covariance matrix. Sensitivity analyses used the same model in completers and per-protocol populations and multiple imputation in the ITT population.
Pain outcomes were analyzed separately using mixed models for repeated measures and EQ-5D using analysis of covariance adjusted for baseline score. Categorical variables were analyzed using Fisher exact or χ2 tests, and continuous variables using t tests or Wilcoxon tests depending on distribution. No multiplicity adjustment was applied to secondary outcomes; these analyses are exploratory. Data were analyzed from August 7, 2024, to June 18, 2025.
Results
Study Population and Adherence
A total of 168 patients (mean [SD] age, 49.0 [13.7] years; 101 women [60.1%]) were randomized across 15 active centers (mean [range], 11.2 [1-27] participants per center) among 17 participating centers, including 86 participants in the belt group and 82 participants in the control group. The FAS included 81 belt-group patients and 74 controls (Figure 1). Availability of ODI data at baseline, T1, and T2 is reported in eTable 1 in Supplement 2. Among randomized patients, baseline characteristics were similar between groups (Table 1). Most patients had recurrent nonspecific LBP, and the current episode had lasted approximately 3 months.
Figure 1. Flow Diagram of the Study Participants.
Table 1. Demographic and Clinical Characteristics of the Study Population in the Intention-to-Treat Group at Baseline.
| Characteristic | Patients, No. (%) | ||
|---|---|---|---|
| Belt group (n = 86) | Control group (n = 82) | Total (N = 168) | |
| Age, mean (SD), y | 48.9 (14.5) | 49.4 (13.1) | 49.0 (13.7) |
| Sex | |||
| Male | 38 (44.2) | 29 (35.4) | 67 (39.9) |
| Female | 48 (55.8) | 53 (64.6) | 101 (60.1) |
| BMI | |||
| <18.5 | 1 (1.2) | 2 (2.4) | 3 (1.8) |
| 18.5 to <25 | 43 (50.0) | 41 (50.0) | 84 (50.0) |
| 25 to <30 | 28 (32.6) | 23 (28.0) | 51 (30.4) |
| ≥30 | 14 (16.3) | 16 (19.5) | 30 (17.9) |
| Occupation | |||
| Working | |||
| Overall | 69 (80.2) | 63 (76.8) | 132 (78.6) |
| Nonsedentary profession | 23 (33.3) | 20 (31.7) | 43 (32.6) |
| Mixed profession | 17 (24.6) | 16 (25.4) | 33 (25.0) |
| Sedentary profession | 29 (42.0) | 27 (42.9) | 56 (42.4) |
| Retired | 15 (17.4) | 13 (15.9) | 28 (16.7) |
| Unemployed | 2 (2.3) | 4 (4.9) | 6 (3.6) |
| On benefitsa | 0 | 2 (2.4) | 2 (1.2) |
| Active in sports | |||
| Yes | 45 (52.3) | 45 (54.9) | 90 (53.6) |
| No | 41 (47.7) | 37 (45.1) | 78 (46.4) |
| Medical history | |||
| Cardiac disorder | 1 (1.2) | 2 (2.4) | 3 (1.8) |
| Endocrine disorder | 3 (3.5) | 8 (9.8) | 11 (6.5) |
| Gastrointestinal disorder | 6 (7.0) | 6 (7.3) | 12 (7.1) |
| Neoplasm | 5 (5.8) | 3 (3.7) | 8 (4.8) |
| Neurological disorder | 5 (5.8) | 8 (9.8) | 13 (7.7) |
| Psychiatric disorder | 6 (7.0) | 7 (8.5) | 13 (7.7) |
| Respiratory disorder | 4 (4.7) | 3 (3.7) | 7 (4.2) |
| Hypertension | 11 (12.8) | 9 (11.0) | 20 (11.9) |
| Type 2 diabetes | 1 (1.2) | 1 (1.2) | 2 (1.2) |
| Low back pain type | |||
| Subacute | 16 (18.6) | 15 (18.3) | 31 (18.5) |
| Recently chronic | 19 (22.1) | 19 (23.2) | 38 (22.6) |
| Recurrent | 51 (59.3) | 48 (58.5) | 99 (58.9) |
| Low back pain locationb | |||
| Highly localized | 65 (75.6) | 63 (76.8) | 128 (76.2) |
| Widespread | 21 (24.4) | 20 (24.4) | 41 (24.4) |
| Duration of current episode of back pain, mean (SD), mo | 3.1 (1.5) | 3.0 (1.6) | 3.1 (1.6) |
Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).
This category refers to participants engaged in unpaid activities, such as volunteering.
Pain location categories were not mutually exclusive. One participant in the control group selected both highly localized and widespread pain; therefore, percentages may sum to more than 100%.
Thermal sensor data were available for 52 of 86 belt-group patients (60.5%). Between baseline and T2, the belt was worn for a mean (SD) of 5.5 (1.8) days per week by diaries and 4.5 (2.0) days per week by sensors. On wear days, mean (SD) daily wear time was 4.9 (2.2) hours by diaries and 3.8 (1.4) hours by sensors. Diary and sensor data were significantly correlated for wear frequency, expressed as percentage of days with belt wear (Spearman r = 0.698 [95% CI, 0.520-0.813]; P < .001), and for total hours of belt wear (Spearman r = 0.545 [95% CI, 0.315-0.709]; P < .001) over baseline to T2
Primary Outcome
From baseline to T2, adjusted mean ODI change by −10.0 (95% CI, −13.0 to −7.0) points in the belt group and −5.3 (95% CI, −8.6 to −2.1) points in controls, with a between-group difference of −4.7 (95% CI, −8.4 to −0.9) points (P = .01) (Table 2). At T2, 49 of 81 belt-group patients (60.5%) achieved at least 30% ODI improvement compared with 30 of 74 (40.5%) controls, corresponding to an absolute difference of 20.0 percentage points. Improvement of less than 30% occurred in 17 belt-group patients (21.0%) and 20 controls (27.0%), and no improvement or worsening in 15 belt-group patients (18.5%) and 24 controls (32.4%) (P = .04).
Table 2. Adjusted ODI and Pain Scores at T0 and Adjusted Changes From Baseline to T1 and T2.
| Time point | Mean (SD) | Between-group difference | ||
|---|---|---|---|---|
| Belt group (n = 81) | Control group (n = 74) | Between-group Mean (95% CI), points | P value | |
| ODI | ||||
| T0 score | 25.9 (1.6) | 27.8 (1.7) | −1.9 (−5.7 to 1.9) | .33 |
| Difference: T1 − T0 | −9.1 (1.4) | −5.9 (1.5) | −3.2 (−6.5 to 0.01) | .05 |
| Difference: T2 − T0 | −10.0 (1.5) | −5.3 (1.6) | −4.7 (−8.4 to −0.9) | .01 |
| Pain at rest | ||||
| Mean pain | ||||
| T0 score | 44.5 (2.3) | 44.5 (2.5) | −0.07 (−5.6 to 5.4) | .98 |
| Difference: T1 − T0 | −19.8 (3.0) | −7.0 (3.2) | −12.8 (−20.1 to −5.6) | <.001 |
| Difference: T2 − T0 | −19.2 (3.1) | −10.5 (3.3) | −8.7 (−16.1 to −1.3) | .02 |
| Minimum pain | ||||
| T0 score | 22.1 (2.5) | 25.5 (2.6) | −3.4 (−9.4 to 2.7) | .27 |
| Difference: T1 − T0 | −6.8 (3.1) | −1.5 (3.3) | −5.3 (−12.5 to 1.9) | .15 |
| Difference: T2 − T0 | −6.1 (3.3) | −4.9 (3.4) | −1.2 (−9.0 to 6.5) | .76 |
| Maximum pain | ||||
| T0 score | 63.4 (2.6) | 63.8 (2.8) | −0.4 (−6.6 to 5.8) | .90 |
| Difference: T1 − T0 | −24.5 (3.8) | −13.2 (4.0) | −11.3 (−20.3 to −2.3) | .01 |
| Difference: T2 − T0 | −22.8 (4.0) | −17.1 (4.2) | −5.7 (−15.5 to 4.0) | .25 |
| Pain during activity | ||||
| Mean pain | ||||
| T0 score | 56.7 (2.2) | 57.9 (2.3) | −1.2 (−6.5 to 4.0) | .65 |
| Difference: T1 − T0 | −23.8 (3.2) | −11.4 (3.4) | −12.5 (−20.1 to −4.8) | .002 |
| Difference: T2 − T0 | −23.8 (3.4) | −13.8 (3.6) | −10.0 (−18.3 to −1.6) | .02 |
| Minimum pain | ||||
| T0 score | 35.3 (2.9) | 40.1 (3.1) | −4.9 (−12.0 to 2.3) | .18 |
| Difference: T1 − T0 | −16.7 (3.5) | −13.8 (3.7) | −3.0 (−11.2 to 5.3) | .48 |
| Difference: T2 − T0 | −13.9 (3.9) | −11.8 (4.1) | −2.1 (−11.7 to 7.5) | .67 |
| Maximum pain | ||||
| T0 score | 75.7 (2.0) | 76.2 (2.1) | −0.6 (−5.3 to 4.2) | .81 |
| Difference: T1 − T0 | −29.1 (3.7) | −16.1 (3.9) | −13.0 (−22.0 to −4.1) | .005 |
| Difference: T2 − T0 | −29.4 (3.9) | −18.8 (4.1) | −10.6 (−20.0 to −1.3) | .03 |
Abbreviations: ODI, Oswestry Disability Index; T0, baseline; T1, 4 weeks; T2, 12 weeks.
At T1, the adjusted between-group difference in ODI change was −3.2 (95% CI, −6.5 to 0.0) points (P = .05). Sensitivity analyses in the completers and per-protocol populations, as well as the ITT multiple-imputation analysis, were consistent with the primary FAS analysis (eTable 2 in Supplement 2).
Secondary Outcomes
Mean pain scores generally decreased more in the belt group than in controls, with differences observed at T1 and persisting at T2 (Table 2 and Figure 2). At T2, adjusted between-group differences in change from baseline were −8.7 (95% CI, −16.1 to −1.3) points for mean pain at rest and −10.0 (95% CI, −18.3 to −1.6) points for mean pain during activity. Maximum pain during activity also favored the belt group at both T1 and T2 (Table 2). In weekly analyses, model-based mean activity-related pain fell below 40 points in the belt group only from week 3.
Figure 2. Line Graphs of Weekly Pain Outcomes From Baseline to Week 12.
P values were derived from mixed models for repeated measures. For change-from-baseline analyses, P values compare the absolute change from baseline between groups at each weekly time point.
Key secondary outcomes are reported in Table 3. No statistically significant between-group difference was detected for change in physical activity. Finger-to-floor distance improved more in the belt group, with a mean between-group difference of −4.2 (95% CI, −0.30 to −8.28) cm, whereas Schober-Macrae did not differ.
Table 3. Secondary Outcomes at T2 or Change From Baseline.
| Secondary outcome | Belt group (n = 81) | Control group (n = 74) | P value |
|---|---|---|---|
| EQ-5D-5L index at T2, | |||
| Adjusted mean (95% CI) | 0.691 (0.638 to 0.743) | 0.630 (0.578 to 0.681) | .11a |
| Missing data, No. | 17 | 9 | NA |
| VAS general health at T2, | |||
| Mean (SD) | 70.1 (21.0) | 67.0 (18.1) | .37b |
| Missing data, No. | 17 | 9 | NA |
| CGI-S evolution from T0 to T2, No. (%) | |||
| Improvement | 39 (54.9) | 23 (33.3) | .03c |
| No change | 23 (32.4) | 35 (50.7) | |
| Deterioration | 9 (12.7) | 11 (15.9) | |
| Missing data, No. | 10 | 5 | NA |
| CGI-I at T2, No. (%) | |||
| Improvement | 51 (71.8) | 31 (44.9) | <.001c |
| No change/stabilization | 16 (22.5) | 21 (30.4) | |
| Deterioration | 4 (5.6) | 17 (24.6) | |
| Missing data, No. | 10 | 5 | |
| PGI-C at T2, No. (%) | |||
| Strong improvement | 17 (26.6) | 12 (18.5) | .02c |
| Moderate improvement | 19 (29.7) | 9 (13.8) | |
| Slight improvement | 15 (23.4) | 30 (46.2) | |
| No change/deterioration | 13 (20.3) | 14 (21.5) | |
| Missing data, No. | 17 | 9 | NA |
| Finger-to-floor distance, change from T0 to T2 | |||
| Mean (SD) | −6.9 (12.6) | −2.7 (10.9) | .04b |
| Median (IQR) | −4.0 (−12.0 to 0.0) | 0.0 (−8.0 to 2.3) | |
| Missing data, No. | 12 | 6 | NA |
| Schober-McRae index, change from T0 to T2 | .29b | ||
| Mean (SD) | 0.4 (2.1) | 0.0 (2.5) | |
| Median (IQR) | 0.5 (0.0 to 1.0) | 0.0 (−1.0 to 1.0) | |
| Missing data, No. | 14 | 8 | NA |
| Physical activity, change from T0 to T2, MET-min/wk | |||
| Mean (SD) | 1063.7 (7180.1) | −1098.2 (7520.1) | .10b |
| Median (IQR) | 939.0 (−1722.0 to 3840.0) | 80.0 (−3542.3 to 2043.0) | |
| Missing data, No. | 19 | 10 | NA |
| Pain medication use, No. (%) | |||
| Any | 41 (50.6) | 50 (67.6) | .03d |
| Analgesics | 24 (29.6) | 25 (33.8) | .58d |
| NSAIDs | 15 (18.5) | 27 (36.5) | .01d |
| Opioids, level 2 | 9 (11.1) | 16 (21.6) | .08d |
| Opioids, level 3 | 1 (1.2) | 1 (1.4) | >.99d |
| Missing data, No. | 0 | 0 | NA |
| Sick leave between T0 and T2 | |||
| No. (%) | 8 (9.9) | 7 (9.5) | >.99 |
| Missing data, No. | 0 | 0 | |
| Sick leave duration | .33e | ||
| Mean (SD) | 48.0 (33.3) | 60.4 (36.0) | |
| Median (IQR) | 52.5 (18.0 to 76.5) | 85.0 (23.0 to 90.0) | |
| NA or no sick leave, No. | 73 | 67 | NA |
| Medical consultations related to current low back pain between T0 and T2 | .27d | ||
| No. (%) | 27 (33.3) | 31 (41.9) | |
| Missing data, No. | 0 | 0 | NA |
Abbreviations: CGI-I, Clinical Global Impression–Improvement; CGI-S, Clinical Global Impression–Severity; EQ-5D-5L, EuroQol 5-Dimension 5-Level questionnaire; MET, metabolic equivalent of task; NA, not applicable; NSAID, nonsteroidal anti-inflammatory drug; PGI-C, Patient Global Impression of Change; T0, baseline; T2, week 12; VAS, visual analog scale.
Analysis of covariance adjusted for treatment group, center, and T0 EQ-5D-5L index.
t test.
Fisher exact test.
χ2 test.
Wilcoxon test.
Physician-rated CGI-S indicated improvement from baseline to T2 in 39 of 71 belt-group patients (54.9%) and 23 of 69 controls (33.3%) (P = .03). Physician-rated CGI-I classified 51 of 71 belt-group patients (71.8%) and 31 of 69 controls (44.9%) as improved at T2 (P < .001). Patient-reported PGI-C showed moderate to strong improvement in 36 of 64 belt-group patients (56.3%) and 21 of 65 controls (32.3%) (P = .02). EQ-5D index and VAS current health scores did not differ between groups, although mobility, pain or discomfort, and usual activities favored the belt group (Table 3).
Fewer belt-group patients than controls used LBP medication during follow-up (41 of 81 patients [50.6%] vs 50 of 74 patients [67.6%]; P = .03), particularly NSAIDs (15 of 81 patients [18.5%] vs 27 of 74 patients [36.5%]; P = .01). No statistically significant between-group difference was detected for level 2 opioid use, sick leave, or consultations.
A total of 63 belt-group patients (73.3%) completed the satisfaction survey. Mean (SD) ratings were 6.8 (2.8) for overall benefit, 6.7 (3.0) for pain relief, and 6.3 (3.0) for comfort.
Two serious adverse events occurred: 1 inguinal hernia in the belt group and 1 positional vertigo in controls. Neither was considered related to the study by the physician. Six belt-group patients reported mild device-related adverse events: discomfort, musculoskeletal, or skin irritation.
Discussion
In this open-label randomized clinical trial, adding a nonrigid lumbar belt to usual care resulted in greater improvement in ODI after 12 weeks than usual care alone. The adjusted mean between-group difference of −4.7 ODI points was smaller than the 10-point target difference used for sample size calculation. In addition, 60.5% of belt-group patients achieved at least 30% ODI improvement, a threshold considered clinically meaningful at the individual level,30 compared with 40.5% of controls. Because the ODI assesses disability related to daily activities, these findings suggest a greater improvement in patient-perceived functional limitations in the belt group.
Secondary outcomes supported these findings. Mean pain at rest and during activity generally decreased more in the belt group over 12 weeks, with adjusted between-group differences at T2 of −8.7 and −10.0 points, respectively. The difference appeared particularly relevant for movement-related pain, as mean activity-related pain fell below 40 points in the belt group only from week 3 onward. These findings may suggest an early improvement in movement-related pain, which could help patients remain active and reduce fear of movement. However, weekly pain analyses were secondary and should be interpreted cautiously. The greater reduction in maximum pain during activity also suggests that the belt may reduce pain peaks during effort. Potential mechanisms underlying pain relief with lumbar belts have been described previously.8,9,10,11 Although no between-group difference was observed for the overall EQ-5D index or current health VAS, several EQ-5D dimensions, including mobility, pain or discomfort, and usual activities, favored the belt group, consistent with the observed improvements in ODI and pain.
The lower proportion of patients using LBP-related medication in the belt group, particularly NSAIDs, further supports the potential analgesic benefit of lumbar belts as a nonpharmacological option. This is consistent with previous evidence suggesting that soft lumbar belts may provide immediate pain relief at rest.31 Given the known adverse effects associated with NSAIDs7 and the absence of serious device-related adverse events in this study, lumbar belts may represent a useful complementary option for pain relief in nonspecific LBP.
The study objectively assessed adherence using thermal sensors. According to sensor data, patients wore the belt for a mean of 4.5 days per week and approximately 4 hours per wear-day, although diaries slightly overestimated use. A prespecified subgroup analysis comparing good vs poor adherence showed numerically greater ODI improvement with good adherence but was not statistically significant; further dedicated exposure-response analyses using sensor data are planned.
Our results are consistent with previous studies on nonrigid lumbar belts.17,18,32 However, other trials, including a large study by Oleske et al33 among workers with recent work-related LBP, did not show benefit.33 Differences may relate to design features. Oleske et al33 included predominantly acute work-related LBP, with most participants reporting onset within 2 weeks, used a 12-month follow-up (during which spontaneous recovery is likely), and relied on self-reported adherence, which declined over time. Future trials should carefully define symptom duration, baseline severity, device type, comparator, wearing instructions, adherence monitoring, and follow-up duration. Whether lumbar belts contribute to earlier return to work or reduced economic burden remains to be tested in dedicated studies. Methodological strengths of this trial include its multicenter randomized clinical design, standardized eligibility criteria and intervention instructions, use of validated patient-reported and clinical outcome measures, prospective weekly online pain assessments limiting recall bias, objective adherence monitoring with thermal sensors to support assessment of intervention fidelity, and sensitivity analyses conducted to assess the robustness of the primary findings.
Limitations
The study has limitations. Recruitment was lower than planned because of the COVID-19 pandemic, which may have reduced power. The open-label design and absence of a sham-device control are important limitations, particularly for patient-reported outcomes. Placebo, expectation, or nocebo effects cannot be excluded, and the trial cannot distinguish device-specific from contextual effects. Thus, findings should be interpreted as the overall effect of prescribing a nonrigid lumbar belt in clinical care. Assessors were not formally blinded, which may have affected clinician-rated global impression scales. Patients were not asked to rate pain or disability without wearing the belt, so outcomes reflect real-time use. Prior lumbar belt use was not systematically collected and may have influenced expectations or adherence. Multiple secondary outcomes were analyzed without multiplicity adjustment and are exploratory. Furthermore, baseline ODI indicated moderate disability, limiting generalizability to more severe LBP.
Conclusions
In this randomized clinical trial of patients with nonspecific LBP lasting 1 to 6 months, adding a nonrigid lumbar belt to usual care resulted in improved pain and functional disability over 12 weeks. These findings apply to belt use for 4 to 8 hours per day throughout follow-up, particularly during physical activity. Lumbar belts may be considered as a nonpharmacological option to help manage pain and functional disability in nonspecific LBP.
Trial Protocol and Statistical Analysis Plan
eTable 1. Availability of ODI data at each time point
eTable 2. Sensitivity analysis of absolute change from baseline in ODI score using MMRM with multiple imputation in the randomized population
Data Sharing Statement
References
- 1.Rubin DI. Epidemiology and risk factors for spine pain. Neurol Clin. 2007;25(2):353-371. doi: 10.1016/j.ncl.2007.01.004 [DOI] [PubMed] [Google Scholar]
- 2.Hartvigsen J, Hancock MJ, Kongsted A, et al. ; Lancet Low Back Pain Series Working Group . What low back pain is and why we need to pay attention. Lancet. 2018;391(10137):2356-2367. doi: 10.1016/S0140-6736(18)30480-X [DOI] [PubMed] [Google Scholar]
- 3.Stubbs B, Koyanagi A, Thompson T, et al. The epidemiology of back pain and its relationship with depression, psychosis, anxiety, sleep disturbances, and stress sensitivity: data from 43 low- and middle-income countries. Gen Hosp Psychiatry. 2016;43:63-70. doi: 10.1016/j.genhosppsych.2016.09.008 [DOI] [PubMed] [Google Scholar]
- 4.Foster NE, Anema JR, Cherkin D, et al. ; Lancet Low Back Pain Series Working Group . Prevention and treatment of low back pain: evidence, challenges, and promising directions. Lancet. 2018;391(10137):2368-2383. doi: 10.1016/S0140-6736(18)30489-6 [DOI] [PubMed] [Google Scholar]
- 5.Malik KM, Nelson AM, Chiang TH, Imani F, Khademi SH. The specifics of non-specific low back pain: re-evaluating the current paradigm to improve patient outcomes. Anesth Pain Med. 2022;12(4):e131499. doi: 10.5812/aapm-131499 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Maksimova MY, Kotlyar YA. Chronic nonspecific back pain. Hum Physiol. 2022;48:968-974. doi: 10.1134/S0362119722080096 [DOI] [Google Scholar]
- 7.George SZ, Lentz TA, Goertz CM. Back and neck pain: in support of routine delivery of non-pharmacologic treatments as a way to improve individual and population health. Transl Res. 2021;234:129-140. doi: 10.1016/j.trsl.2021.04.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Im SC, Seo SW, Kang NY, Jo H, Kim K. The effect of lumbar belts with different extensibilities on kinematic, kinetic, and muscle activity of sit-to-stand motions in patients with nonspecific low back pain. J Pers Med. 2022;12(10):1678. doi: 10.3390/jpm12101678 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shahvarpour A, Preuss R, Larivière C. The effect of extensible and non-extensible lumbar belts on trunk postural balance in subjects with low back pain and healthy controls. Gait Posture. 2019;72:211-216. doi: 10.1016/j.gaitpost.2019.06.013 [DOI] [PubMed] [Google Scholar]
- 10.Bernier E, Driscoll M. Numerical investigation of intra-abdominal pressure and spinal load-sharing upon the application of an abdominal belt. J Biomech. 2023;161:111863. doi: 10.1016/j.jbiomech.2023.111863 [DOI] [PubMed] [Google Scholar]
- 11.Soisson O, Lube J, Germano A, et al. Pelvic belt effects on pelvic morphometry, muscle activity and body balance in patients with sacroiliac joint dysfunction. PLoS One. 2015;10(3):e0116739. doi: 10.1371/journal.pone.0116739 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Landauer F, Trieb K. An indication-based concept for stepwise spinal orthosis in low back pain according to the current literature. J Clin Med. 2022;11(3):510. doi: 10.3390/jcm11030510 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Airaksinen O, Brox JI, Cedraschi C, et al. ; COST B13 Working Group on Guidelines for Chronic Low Back Pain . Chapter 4. European guidelines for the management of chronic nonspecific low back pain. Eur Spine J. 2006;15(Suppl 2)(suppl 2):S192-S300. doi: 10.1007/s00586-006-1072-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.National Institute for Health and Care Excellence . Recommendations—low back pain and sciatica in over 16s: assessment and management. Updated November 30, 2016. Accessed November 5, 2024. https://www.nice.org.uk/guidance/NG59/chapter/Recommendations#non-invasive-treatments-for-low-back-pain-and-sciatica
- 15.World Health Organization . WHO Guideline for Non-Surgical Management of Chronic Primary Low Back Pain in Adults in Primary and Community Care Settings. World Health Organization; 2023. [PubMed] [Google Scholar]
- 16.Azadinia F, Ebrahimi E, Kamyab M, Parnianpour M, Cholewicki J, Maroufi N. Can lumbosacral orthoses cause trunk muscle weakness: a systematic review of literature. Spine J. 2017;17(4):589-602. doi: 10.1016/j.spinee.2016.12.005 [DOI] [PubMed] [Google Scholar]
- 17.Gignoux P, Lanhers C, Dutheil F, Boutevillain L, Pereira B, Coudeyre E. Non-rigid lumbar supports for the management of non-specific low back pain: a literature review and meta-analysis. Ann Phys Rehabil Med. 2022;65(1):101406. doi: 10.1016/j.rehab.2020.05.010 [DOI] [PubMed] [Google Scholar]
- 18.Calmels P, Queneau P, Hamonet C, et al. Effectiveness of a lumbar belt in subacute low back pain: an open, multicentric, and randomized clinical study. Spine (Phila Pa 1976). 2009;34(3):215-220. doi: 10.1097/BRS.0b013e31819577dc [DOI] [PubMed] [Google Scholar]
- 19.Schott C, Zirke S, Schmelzle JM, Kaiser C, Fernández LAI. Effectiveness of lumbar orthoses in low back pain: review of the literature and our results. Orthop Rev (Pavia). 2018;10(4):7791. doi: 10.4081/or.2018.7791 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Oleiwi MA, Shah SZA, Bilal H, et al. Efficacy of orthotic support in mitigating low back pain and disability in low back pain sufferers. J Back Musculoskelet Rehabil. 2023;36(5):1111-1125. doi: 10.3233/BMR-220200 [DOI] [PubMed] [Google Scholar]
- 21.Akbar MS, Syafar M, Thamrin Y, Aras D, Maidin A, Arsyad M. Impact of lumbar support on pain reduction in low back pain patients: a systematic review and meta-analysis of randomized control trials. Narra J. 2025;5(1):e2165. doi: 10.52225/narra.v5i1.2165 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hopewell S, Chan AW, Collins GS, et al. CONSORT 2025 explanation and elaboration: updated guideline for reporting randomised trials. BMJ. 2025;389:e081124. doi: 10.1136/bmj-2024-081124 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Vogler D, Paillex R, Norberg M, de Goumoëns P, Cabri J. Validation transculturelle de l’Oswestry disability index en français. Ann Readapt Med Phys. 2008;51(5):379-385. doi: 10.1016/j.annrmp.2008.03.006 [DOI] [PubMed] [Google Scholar]
- 24.Lee PH, Macfarlane DJ, Lam TH, Stewart SM. Validity of the International Physical Activity Questionnaire Short Form (IPAQ-SF): a systematic review. Int J Behav Nutr Phys Act. 2011;8:115. doi: 10.1186/1479-5868-8-115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ekedahl H, Jönsson B, Frobell RB. Fingertip-to-floor test and straight leg raising test: validity, responsiveness, and predictive value in patients with acute/subacute low back pain. Arch Phys Med Rehabil. 2012;93(12):2210-2215. doi: 10.1016/j.apmr.2012.04.020 [DOI] [PubMed] [Google Scholar]
- 26.Macrae IF, Wright V. Measurement of back movement. Ann Rheum Dis. 1969;28(6):584-589. doi: 10.1136/ard.28.6.584 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Busner J, Targum SD. The clinical global impressions scale: applying a research tool in clinical practice. Psychiatry (Edgmont). 2007;4(7):28-37. [PMC free article] [PubMed] [Google Scholar]
- 28.Scott W, McCracken LM. Patients’ impression of change following treatment for chronic pain: global, specific, a single dimension, or many? J Pain. 2015;16(6):518-526. doi: 10.1016/j.jpain.2015.02.007 [DOI] [PubMed] [Google Scholar]
- 29.Szende A, Janssen B, Cabases J, eds. Self-Reported Population Health: An International Perspective Based on EQ-5D. Springer; 2014. doi: 10.1007/978-94-007-7596-1 [DOI] [PubMed] [Google Scholar]
- 30.Ostelo RW, Deyo RA, Stratford P, et al. Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change. Spine (Phila Pa 1976). 2008;33(1):90-94. doi: 10.1097/BRS.0b013e31815e3a10 [DOI] [PubMed] [Google Scholar]
- 31.Shahvarpour A, Preuss R, Sullivan MJL, Negrini A, Larivière C. The effect of wearing a lumbar belt on biomechanical and psychological outcomes related to maximal flexion-extension motion and manual material handling. Appl Ergon. 2018;69:17-24. doi: 10.1016/j.apergo.2018.01.001 [DOI] [PubMed] [Google Scholar]
- 32.Valle-Jones JC, Walsh H, O’Hara J, O’Hara H, Davey NB, Hopkin-Richards H. Controlled trial of a back support (‘Lumbotrain’) in patients with non-specific low back pain. Curr Med Res Opin. 1992;12(9):604-613. doi: 10.1185/03007999209111527 [DOI] [PubMed] [Google Scholar]
- 33.Oleske DM, Lavender SA, Andersson GB, Kwasny MM. Are back supports plus education more effective than education alone in promoting recovery from low back pain?: results from a randomized clinical trial. Spine (Phila Pa 1976). 2007;32(19):2050-2057. doi: 10.1097/BRS.0b013e3181453fcc [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial Protocol and Statistical Analysis Plan
eTable 1. Availability of ODI data at each time point
eTable 2. Sensitivity analysis of absolute change from baseline in ODI score using MMRM with multiple imputation in the randomized population
Data Sharing Statement
