Effect of different positions of lumbar traction on pain, function, and range of motion in adults with non-specific low back pain: a randomized clinical trial Authors: Waleed S Mahmoud, Ahmed Alhowimel, Abdalmajed Almotery, Kamel Mustafa Hafiz, Sahar Mahmoud Hassan, Mohamed Mohamed Mazen Published: 2026 ABSTRACT AbstractNon-specific low back pain (NSLBP) represents about 80–95% of all LBP cases, yet the optimal lumbar traction position for effective treatment remains debated. To investigate the effect of different positions during lumbar traction on pain, function, and ROM in adults with NSLBP. Thirty participants were randomly allocated to one of three lumbar traction position groups: supine, side-lying, or prone. Outcome measures included the Oswestry Disability Index (ODI), International Physical Activity Questionnaire short form (IPAQ-SF), Finger-to-Floor Test (FFT), and Numerical Pain Rating Scale (NPRS). Measurements were obtained at baseline, after the third session, and after completion of the 3-week intervention. Within-group analysis showed significant improvements in ODI and NPRS scores over time in all three groups (p < 0.001). The supine group demonstrated earlier within-group improvement after the third session (p = 0.012). IPAQ improved significantly only in the side-lying group (p = 0.01), whereas FFT did not show significant changes in any group (p > 0.05). However, between-group analysis revealed no statistically significant differences among the three traction positions at any measurement point for any outcome (p > 0.05). Lumbar traction was associated with improvements in pain and disability regardless of position. However, no traction position demonstrated superiority over the others. The earlier improvement observed in the supine group should be interpreted cautiously because it was based on within-group change rather than between-group superiority. Future studies with larger samples are needed to clarify whether positional differences are clinically meaningful. Also, these findings are limited to male participants and should not be generalized to female populations.Trial registration: The trial was registered on 02/02/2025, clinicaltrials.gov, Identifier: (NCT06812338).Keywords: Non-specific low back pain, Lumbar traction, Oswestry Disability Index, International Physical Activity QuestionnaireSubject terms: Health care, Medical research, Signs and symptoms Introduction Around 80–95% of low back pain cases are classified as non-specific low back pain (NSLBP), while only a small proportion are linked to clearly identifiable pathologies or nerve-related causes1,2. It is estimated that about half of the population will experience low back pain at some point in their lives3. NSLBP is a complex and multifactorial condition, shaped by a combination of physical, psychological, social, lifestyle, and individual factors4,5. The high prevalence of low back pain gave rise to the development of a wide range of treatment approaches, including exercise therapy, manual therapy, dry needling, and soft tissue mobilization6. Mechanical traction has also remained one of the most commonly used treatment techniques for low back pain, with its application tracing back to the 4th–fifth century BC7,8. Although mechanical traction is widely used in clinical practice, used by approximately 41% of clinicians in the United Kingdom and 77% in the United States9, its therapeutic effectiveness remains uncertain. This controversy is largely attributed to the limited availability of high-quality randomized clinical trials, many of which lack adequate blinding, sufficient sample sizes, clearly defined inclusion criteria, and long-term follow-up6,8. Another explanation for the conflicting evidence is the variation in traction protocols. Previous systematic reviews have identified 37 RCTs that differed markedly in dosage, frequency, traction position, and patient characteristics. These inconsistencies have contributed to the uncertainty regarding the true effectiveness of traction and the optimal method for its application9. Some studies have supported the benefits of traction, which has a positive impact in reducing short-term pain in Patients with LBP10,11, while others have concluded that there is no significant difference in outcome between actual and sham traction12,13. With regard to the optimal method of traction application, previous studies comparing intermittent and continuous traction have generally reported no significant difference when similar traction forces are used13, Other studies comparing different traction forces have reported pain improvement at higher force levels, although no significant differences were observed in straight leg raise range of motion14. Researchers have also explored how body position during traction may affect outcomes. One study compared traction performed in a lying (supine) position with inversion traction at varying angles, finding that greater inversion was associated with more improvement in pain, lumbar flexibility, and trunk muscle strength15. Two other studies have compared the supine with the prone position; both studies found a significant impact on pain and disability in the prone position, suggesting supine and prone positions are better choices when applying traction8,16. Another study compared side-lying with supine-lying; the authors found that the side-lying group induced a significant improvement in back and leg pain as well as in the Oswestry Disability Index (ODI) and score of L5/S1 herniated disc index17. So far, there is limited evidence directly comparing different lumbar traction positions within a single randomized study. Most previous research has focused on comparing only two positions and has often involved populations other than those with NSLBP. Therefore, this study aimed to examine and compare the effects of three lumbar traction positions—supine, side-lying, and prone—on pain, function, physical activity, and range of motion in adults with NSLBP. Understanding whether body position affects treatment outcomes could help improve and refine traction protocols in clinical practice. Methods Study design This three-arm randomized clinical trial was conducted at the outpatient clinic of the College of Applied Medical Sciences at Prince Sattam bin Abdulaziz University between March and September 2025. Only male participants were selected because the clinical setting where the study was conducted has logistical considerations related to patient recruitment and treatment delivery that limit mixed-sex participation. The study protocol was registered at ClinicalTrials.gov under the identifier NCT06812338. Ethical approval was obtained from the Research Ethics Committee of the Department of Health and Rehabilitation Sciences at Prince Sattam bin Abdulaziz University (Approval No. RHPT/024/020). This study was conducted in accordance with the Declaration of Helsinki. It was also designed and reported in accordance with the CONSORT guidelines for randomized clinical trials (Fig. 1)18. All participants provided written informed consent before data collection began. They also agreed to the publication of their data in anonymized form in an online open-access journal. No personally identifiable information is included in this manuscript. Participant’s recruitment flow chart according to the consolidated standards of reporting trials (CONSORT). Sample size calculation An effect size of 0.7 (Cohen’s d) was derived from Bilgilisoy Filiz et al.8, and a significance level of 0.05 was used to calculate the sample size using G*power 3.0.10 software (University of Dusseldorf, Dusseldorf, Germany). The analysis indicated that a minimum of 24 participants was required to fulfill 80% power to reject the null hypothesis. To account for potential dropouts and withdrawals, the sample size was increased by 20%, resulting in a total of 30 participants. Participants All potential candidates underwent the diagnostic triage for low back pain to exclude causes other than NSLBP19. Those who passed the screening process were then assessed according to the inclusion and exclusion criteria. Inclusion criteria: male participants aged 18–65 years experiencing non-specific low back pain lasting over 12 weeks were deemed eligible for inclusion. A cutoff point of 4/10 on NPRS was used to indicate moderate pain that is clinically meaningful and would likely benefit from intervention20. Exclusion criteria: participants were excluded if they had vertebral fractures, axial spondyloarthritis, the presence of osteophytes, cauda equina syndrome, radicular pain, radiculopathy, spinal stenosis, spinal surgeries, malignancy, or inflammatory or infectious cases of LBP21. Of the 30 participants initially enrolled, 15 were using medications. Ten participants were taking antihypertensive and/or antidiabetic medications and continued them throughout the study. Five participants were taking muscle relaxants, which were discontinued before the intervention. The remaining 15 participants were not taking NSAIDs or other analgesic medications. Randomization and blinding To minimize bias, several blinding procedures were implemented. Participants randomly selected one sealed envelope corresponding to Group A, Group B, or Group C under the supervision of an independent researcher who was not involved in assessment or treatment. Outcome assessments were performed by an independent evaluator, and interventions were delivered by three trained therapists, each assigned to one group. Treatment sessions for the three groups were scheduled on different days to reduce contact between participants from different groups. Complete participant blinding to body position was not feasible because of the nature of the intervention. Statistical analyses were performed by an independent researcher who was not involved in treatment administration or outcome assessment. Procedures Assessment procedure Participants were first weighed using a digital scale (Beurer GS12, Germany), and their body weight in kilograms was recorded to determine the traction force22. Participants were first weighed using a digital scale (Beurer GS12, Germany), and their body weight in kilograms was recorded to determine the traction force22. They then completed the Arabic version of the Oswestry Disability Index (ODI), which measures the impact of low back pain on daily activities across 10 sections. Each section is scored from 0 to 5, with higher scores indicating greater disability23. They then completed the Arabic version of the Oswestry Disability Index (ODI), which measures the impact of low back pain on daily activities across 10 sections. Each section is scored from 0 to 5, with higher scores indicating greater disability23. Pain intensity was assessed using the Numerical Pain Rating Scale (NPRS)24, an 11-point scale ranging from 0, indicating no pain, to 10, indicating the worst imaginable pain Pain intensity was assessed using the Numerical Pain Rating Scale (NPRS)24, an 11-point scale ranging from 0, indicating no pain, to 10, indicating the worst imaginable pain In addition, participants completed the Arabic version of the short-form International Physical Activity Questionnaire (IPAQ-SF), which evaluates the type and amount of physical activity performed during the previous 7 days25,26. Although the IPAQ-SF does not directly measure the effects of traction therapy, it was included as a standardized and validated measure of self-reported physical activity and its potential relationship to functional change following treatment27–29. IPAQ scores were calculated using the automated IPAQ-SFS scoring spreadsheet in Excel30. In addition, participants completed the Arabic version of the short-form International Physical Activity Questionnaire (IPAQ-SF), which evaluates the type and amount of physical activity performed during the previous 7 days25,26. Although the IPAQ-SF does not directly measure the effects of traction therapy, it was included as a standardized and validated measure of self-reported physical activity and its potential relationship to functional change following treatment27–29. IPAQ scores were calculated using the automated IPAQ-SFS scoring spreadsheet in Excel30. Finally, to perform the finger-to-floor test (FFT), subjects were asked to stand on a 20 cm tall platform (KUYING adjustable aerobic stepper, 3-level height workout platform, China). The examiner instructed the patient to bend forward as far as possible; the distance between the tip of the middle finger and the platform was measured in cm using the MABIS DMI Healthcare retractable tape measure (60″, model 35-780-010, USA). Choosing FFT was due to its excellent validity (rs = − 0.96) and reliability (ICC = 0.99), in addition to the responsiveness to change and clinical applicability, making it a reliable tool reflecting the participant’s improvements31. Finally, to perform the finger-to-floor test (FFT), subjects were asked to stand on a 20 cm tall platform (KUYING adjustable aerobic stepper, 3-level height workout platform, China). The examiner instructed the patient to bend forward as far as possible; the distance between the tip of the middle finger and the platform was measured in cm using the MABIS DMI Healthcare retractable tape measure (60″, model 35-780-010, USA). Choosing FFT was due to its excellent validity (rs = − 0.96) and reliability (ICC = 0.99), in addition to the responsiveness to change and clinical applicability, making it a reliable tool reflecting the participant’s improvements31. Treatment procedure A Chattanooga 3D ActiveTrac traction device (USA) was used to apply a controlled and measurable traction force. The traction force was initially set at 25% of the participant’s body weight and was gradually increased based on individual tolerance, without exceeding 50% of body weight. Each treatment session lasted 15 min, consisting of a 1-min ramp-up phase, 13 min of traction, and a 1-min ramp-down phase. Traction was delivered in an intermittent mode, with cycles of 30 s on and 10 s off8. The intervention was administered over a 3-week period32. Throughout the intervention, adverse events and any potential harms—such as increased pain, muscle soreness, or worsening of symptoms—were carefully monitored and recorded. Each group received traction in a different body position, with hip and knee angles measured using an Elite Medical Instruments goniometer, as described below: Group A Participants in Group A were positioned supine, with the arms resting beside the body and the palms facing upward. The hips and knees were flexed to 90° and supported on an adjustable stool. A harness was attached around the chest and pelvis33, as shown in Fig. 2. Supine traction position with the hips and knees flexed to 90°. Chest straps provide stabilization, and a pelvic harness secures the participant during lumbar traction. Group B Participants in Group B were positioned side-lying on the side contralateral to the pain. The head rested comfortably on one hand, while the other arm rested on the trunk. A pillow was placed between the knees. The hip was positioned at 80° of flexion and the knee at 90° of flexion, with the harness attached around the chest and pelvis17, as shown in Fig. 3. The patient is side-lying position with the hips flexed at 80° and knees flexed at 90°. Chest straps provide stabilization, while a pelvic attachment secures the patient for lumbar traction. Group C Participants in Group C were positioned prone, facing downward, with the arms resting comfortably beside the body. The neck and shoulders were maintained in a neutral position, and a pillow was placed under the abdomen. A harness was attached around the chest and pelvis34, as shown in Fig. 4. The patient is lying in a prone position with a pillow under the abdomen. The chest and pelvis are stabilized with a harness. Outcome measures Outcome measures were collected at three time points: baseline, after the third session, and following the completion of the intervention. The primary outcomes were the Arabic version of the Oswestry Disability Index (ODI), which evaluates the effect of back pain on functional disability. In addition, the Arabic version of the short-form International Physical Activity Questionnaire (IPAQ) was to assess physical activity levels. The finger-to-floor test (FFT) was to measure lumbar spine mobility and range of motion. Finally, the Numerical Pain Rating Scale (NPRS) was used to determine pain intensity. Statistical analysis Statistical analyses were performed using SPSS version 26.0 (IBM Corp., Chicago, IL, USA). Continuous data are presented as mean ± standard deviation (SD), whereas categorical variables are presented as frequencies and percentages. The Shapiro–Wilk test was used to assess normality, and Levene’s test was used to examine homogeneity of variances. Although an intention-to-treat approach was initially planned, only participants with complete outcome data (n = 25) were included in the final analysis. Missing data were not imputed. Differences among the three groups at each time point were assessed using one-way ANOVA, while changes over time within each group were assessed using repeated measures ANOVA. Post-hoc pairwise comparisons were conducted using Bonferroni adjustment. A p-value < 0.05 was considered statistically significant. Results A total of 30 participants were randomized, and 25 participants were included in the final analysis. Baseline demographic and clinical characteristics were generally comparable across the three groups. The mean age was 42.44 ± 13.33 years in Group A, 44.13 ± 3.68 years in Group B, and 44.63 ± 10.90 years in Group C. Mean body weight was 86.78 ± 24.25 kg, 85.38 ± 9.58 kg, and 86.88 ± 9.69 kg in Groups A, B, and C, respectively. Mean height was 167.78 ± 4.63 cm in Group A, 169.25 ± 6.18 cm in Group B, and 171.80 ± 4.13 cm in Group C. Mean BMI values were 30.52 ± 7.75, 29.81 ± 3.00, and 29.47 ± 3.35 in Groups A, B, and C, respectively. Mean pain duration was 3.72 ± 4.41 years in Group A, 3.20 ± 3.10 years in Group B, and 2.38 ± 1.51 years in Group C. Statistical analysis indicated no significant differences across the groups at baseline, confirming the initial homogeneity in both demographic and clinical parameters, as summarized in Table 1. Table 1Demographic and baseline characteristics for each group.Group A (N = 9)Group B (N = 8)Group C (N = 8)p-valueMean ± SD (CI 95%)Age42.44 ± 13.33 (32.2–52.7)44.13 ± 3.68 (41.05–47.2)44.63 ± 10.9 (35.51–53.73)0.901FWeight (kg)86.78 ± 24.25 (68.13–105.42)85.38 ± 9.58 (77.37–93.38)86.88 ± 9.69 (78.78–94.97)0.979FHeight (cm)167.78 ± 4.63 (164.22–171.34)169.25 ± 6.18 (164.08–174.42)171.8 ± 4.13 (168.3–175.2)0.284FBMI (kg/m2)30.52 ± 7.75 (24.56–36.48)29.81 ± 3 (27.3–32.31)29.47 ± 3.35 (26.67–32.27)0.916FPain duration (years)3.72 ± 4.41 (0.33–7.11)3.20 ± 3.1 (0.61–5.8)2.38 ± 1.51 (1.12–3.63)0.703FODI-baseline14.89 ± 8.54 (8.33–21.45)15.88 ± 4.85 (11.82–19.93)13.88 ± 6.17 (8.71–19.04)0.842FIPAQ-baseline574.22 ± 529.86 (166.93–981.51)653.13 ± 508.76 (227.79–1078.46)903.19 ± 799.81 (234.53–1571.84)0.541FFFT-baseline10.22 ± 12.55 (0.58–19.87)10.88 ± 7.7 (4.44–17.31)8.75 ± 7.78 (2.25–15.25)0.905FNPRS-baseline4.89 ± 0.78 (4.29–5.49)5 ± 0.76 (4.37–5.63)4.63 ± 0.52 (4.19–5.06)0.553FNo. (%)Marital status Single1 (11.1%)8 (100%)8 (100%)0.396x Married8 (88.9%)Medications Non6 (66.7%)3 (37.5%)4 (50%)0.118x HTN and/or DM3 (33.3%)1 (12.5%)3 (37.5%) Muscle relaxant4 (50%)1 (12.5%)Occupation Field1 (11.1%)2 (25%)2 (25%)0.548x Office2 (22.2%)1 (12.5%)2 (25%) Mixed3 (33.3%)5 (62.5%)2 (25%) Retired3 (33.3%)2 (25%)BMI category Normal2 (22.2%)4 (50%)1 (12.5%)0.690x Overweight3 (33.3%)4 (50%)4 (50%) Obese4 (44.4%)3 (37.5%)Pain site Left2 (22.2%)1 (12.5%)3 (37.5%)0.412x Center6 (66.7%)5 (62.5%)2 (25%) Right1 (11.1%)2 (25%)3 (37.5%)Group A, Supine; Group B, Side-lying; Group C, Prone; BMI, Body mass index; ODI, Oswestry Disability Index; IPAQ, International Physical Activity Questionnaire; FFT, Finger-to-floor test; NPRS, Numerical pain rating scale; F, One-way ANOVA; X2, Pearson Chi-square. Demographic and baseline characteristics for each group. Group A, Supine; Group B, Side-lying; Group C, Prone; BMI, Body mass index; ODI, Oswestry Disability Index; IPAQ, International Physical Activity Questionnaire; FFT, Finger-to-floor test; NPRS, Numerical pain rating scale; F, One-way ANOVA; X2, Pearson Chi-square. Between-group comparisons were performed using one-way ANOVA (Table 2). No statistically significant differences were found among the three groups at baseline, after the third session, or after completion of the intervention for ODI, IPAQ, FFT, or NPRS (all p > 0.05). Within-group changes over time were examined using repeated measures ANOVA. Table 2Influence of lumbar traction in different positions on ODI, IPAQ, FFT, and NPRS between and within groups.ODIIPAQFFTNPRSBaselinePost 3rd sessionPost interventionp† (ηp2 )BaselinePost 3rd sessionPost interventionp† (ηp2 )BaselinePost 3rd sessionPost interventionp† (ηp2 )BaselinePost 3rd sessionPost interventionp† (ηp2 )A (N = 9)Mean ± SD (CI 95%)14.89 ± 8.54 (8.33 –21.45)13.11 ± 8.19 (6.81–19.41)8.56 ± 6.91 (3.24–13.87) < .001** (0.876)574.22 ± 529.86 (166.93–981.51)1271.39 ± 1270.64 (294.69–2248.09)1498.61 ± 1661.57 (221.41–2775.81)0.07 (0.346)10.22 ± 12.55 (0.58–19.87)9.67 ± 12.08 (0.38–18.95)8 ± 10.16 (0.19–15.81)0.10 (0.291)4.89 ± 0.78 (4.29–5.49)4.22 ± 0.833 (3.58–4.86)2.78 ± 1.39 (1.71–3.85) < .001** (0.827)B (N = 8)15.88 ± 4.85 (11.82 –19.93)14.5 ± 5.61 (9.81–19.19)10.25 ± 5.09 (5.99–14.51) < .001** (0.886)653.13 ± 508.76 (227.79–1078.46)1626.94 ± 1051.656 (747.73–2506.14)1778.88 ± 1141.42 (824.63–2733.12)0.01** (0.628)10.88 ± 7.7 (4.44–17.31)10.13 ± 7.42 (3.93–16.32)9.63 ± 7.01 (3.77–15.48)0.198 (0.207)5 ± 0.76 (4.37–5.63)4.63 ± 0.744 (4–5.25)3.38 ± 0.74 (2.75–4) < .001** (0.896)C (N = 8)13.88 ± 6.18 (8.71 –19.04)12 ± 7.52 (5.71 –18.29)8.13 ± 6.92 (2.34–13.91) < .001** (0.762)903.19 ± 799.81 (234.53–1571.84)1183.94 ± 905.33 (427.06–1940.81)1132.38 ± 912.13 (369.81–1894.94)0.335 (0.145)8.75 ± 7.78 (2.25–15.25)7.75 ± 7.13 (1.79–13.71)7.25 ± 6.76 (1.6–12.9)0.08 (0.375)4.63 ± 0.52 (4.19–5.06)4 ± 1.2 (3 –5)2.88 ± 1.25 (1.83–3.92).002** (0.726)p††0.8420.7890.7820.5410.6950.6130.9050.8640.8410.5530.4170.55Group A, Supine; Group B, Side-lying; Group C, Prone; BMI, Body mass index; ODI, Oswestry Disability Index; IPAQ, International Physical Activity Questionnaire; FFT, Finger-to-floor test; NPRS, Numerical pain rating scale.†Repeated measures ANOVA (within group comparison).††One-way ANOVA (Between groups comparison).**Significant p < 0.05. Influence of lumbar traction in different positions on ODI, IPAQ, FFT, and NPRS between and within groups. Group A, Supine; Group B, Side-lying; Group C, Prone; BMI, Body mass index; ODI, Oswestry Disability Index; IPAQ, International Physical Activity Questionnaire; FFT, Finger-to-floor test; NPRS, Numerical pain rating scale. †Repeated measures ANOVA (within group comparison). ††One-way ANOVA (Between groups comparison). **Significant p < 0.05. Repeated measures ANOVA revealed significant improvements across the three time points in all groups for both ODI (Group A: F(2,16) = 56.37, p < 0.001, ηp2 = 0.876; Group B: F(2,14) = 54.25, p < 0.001, ηp2 = 0.886; Group C: F(2,14) = 22.35, p < 0.001, ηp2 = 0.762) and NPRS scores (Group A: F(2,16) = 38.37, p < 0.001, ηp2 = 0.827; Group B: F(2,14) = 46.33, p < 0.001, ηp2 = 0.869; Group C: F (1.17,8.21) = 18.544, p = 0.002, ηp2 = 0.726). Post-hoc analysis showed that groups A and B had the largest improvements from baseline to post-intervention (p < 0.001) and from post-3rd session to post-intervention (Group A: p < 0.001; Group B: p ≤ 0.001) for both measures. The baseline to post-3rd session change was significant only in group A (ODI: p = 0.021; NPRS: p = 0.012) but not in group B (ODI: p = 0.084; NPRS: p = 0.239). In group C, significant improvements were seen from baseline to post-intervention (ODI: p = 0.003; NPRS: p = 0.006) and from post-3rd session to post-intervention (ODI and NPRS, p < 0.001), whereas baseline to post-3rd session changes were not significant (ODI: p = 0.3; NPRS: p = 0.285). Repeated measures ANOVA showed no significant changes in IPAQ scores across the three time points in groups A and C groups (Group A: F(1.7, 8.5) = 4.23, p = 0.70, ηp2 = 0.346; Group C: F (2, 14) = 1.185, p = 0.335, ηp2 = 0.145). In contrast, group B showed a significant improvement, F (1.03, 7.22) = 11.805, p = 0.010, ηp2 = 0.628. Post-hoc analysis revealed significant differences between baseline and post-3rd session (p = 0.034) and between baseline and post-intervention (p = 0.030), while the change from post-3rd session to post-intervention was not significant (p = 0.099). No statistically significant within-group changes were observed in FFT scores in any of the three groups (Table 2). In Group A, no statistically significant change over time was observed, F(1.09, 8.74) = 3.29, p = 0.102, ηp2 = 0.291. Similarly, Group B did not demonstrate a significant time effect, F(2, 16) = 3.28, p = 0.064, ηp2 = 0.291. Group C also showed no significant change across time points, F(2, 16) = 3.29, p = 0.064, ηp2 = 0.291. Discussion This study explored how three different lumbar traction positions—supine, side-lying, and prone—affect pain, disability, physical activity, and range of motion in adults with non-specific low back pain (NSLBP). The authors explicitly state that the participants were only males, which necessitates interpreting the results with caution and understanding them as referring only to men. Overall, participants in all three groups experienced improvements in pain and disability over time. However, there were no meaningful differences between the groups, suggesting that none of the traction positions was superior to the others. Authors didn’t find any between-group differences, so we need to be cautious about that conclusion. Because the sample size was relatively small, small-to-moderate differences might have been missed. We can consider these findings an interesting starting point for preliminary or hypothesis-generating work, not definitive proof that the traction positions work in the same way. In contrast, previous literature mostly compared two positions, such as supine versus prone as in Khan et al.16 and Bilgilisoy Filiz et al.8 or side-lying versus supine as in Naguib et al.17. Regardless of the position and type of traction, traction effects have been attributed to biomechanical decompression of spinal structures; the pulling force increases the intervertebral space, reducing pressure on the spinal discs and nerve roots, lowering intradiscal pressure, and encouraging reorganization of disc material and stimulating non-nociceptive mechanoreceptors that modulate pain perception35–37. In addition, the traction eases movement and alleviates pain that originates largely from pain-induced movement inhibition and paraspinal muscular stiffness38–40. Patients typically experience reduced guarding and demonstrate a greater willingness to move as traction decreases discomfort and alleviates muscle tension41. In non-specific low back pain, the improvement in range of motion following lumbar traction appears to reflect secondary benefits in pain relief and increased functional capacity rather than structural spinal correction10. Traction stretches paraspinal muscles and ligaments, relieving muscle spasms and enhancing flexibility, which leads to symptom relief. These changes increase soft tissue extensibility and segmental mobility, thereby contributing to functional improvements in range of motion (ROM), as documented in previous studies42–44. While our findings align to some extent with previous studies, they also highlight notable differences in outcomes across traction positions. For instance, Beattie et al.34 demonstrated significant improvements in pain and function following an 8-week prone traction program in patients with activity-limiting LBP. Similarly, Bilgilisoy Filiz et al.8 reported favorable effects of prone traction, although their study involved patients with radiculopathy and showed more immediate improvements with the prone position, in contrast to our findings where the supine position yielded more immediate effects. Furthermore, Naguib et al.17 found that side-lying traction was superior to supine traction in patients with lumbar disc herniation. Taken together, these findings suggest that although different traction positions may produce beneficial outcomes, their relative effectiveness may vary depending on patient characteristics and study conditions8,17. After analyzing each group, the findings indicated that lumbar traction in the supine position produced significantly better improvements in both ODI and NPRS scores as early as the third session. In comparison, lumbar traction performed in side-lying or prone positions did not show significant benefits until the full treatment course was completed. The rapid progress observed in the supine traction group may be attributed to the biomechanical and neuromuscular advantages offered by this position. Supine traction encourages earlier relaxation of the paraspinal muscles by providing enhanced comfort and stability, which helps to minimize muscle guarding and pain more quickly. Furthermore, it supports a neutral spinal alignment and ensures uniform force distribution, promoting more effective vertebral decompression and modulation of nociceptive signals16,17. Collectively, these advantages may expedite therapeutic outcomes, enabling patients to achieve clinically significant symptom relief sooner43,45. However, although the supine group showed earlier within-group improvement, this finding should be interpreted with caution because no statistically significant between-group superiority was detected. These rapid changes are supported by previous literature indicating that some individuals demonstrate early responsiveness, particularly when the minimal clinically important difference (MCID) is achieved or exceeded—such as a ≥ 11% reduction in ODI scores or a ≥ 2-point reduction in NPRS46,47. In contrast, prone and side-lying positions may require longer exposure to produce comparable levels of muscle relaxation and spinal decompression, which may explain the delayed improvements observed at the end of the treatment period48. Although this does not establish superiority, it may have practical relevance in clinical settings where early symptom relief is valued. Future studies should examine whether early response to traction can help identify patients who are more likely to benefit from treatment. Second, as previously indicated in Table 2, none of the groups demonstrated statistically significant or clinically meaningful improvements in lumbar range of motion (as measured by FFT scores). Previous investigations, such as those by Murugan et al.49, Wegner et al.7, and Wang et al.11 have found that there is frequently little improvement in range of motion after lumbar traction. These results suggest that, although some quantifiable improvements may have occurred, they were insufficient to be considered statistically or clinically significant. This is in line with previous research showing that changes in the finger-to-floor test of less than 4.5 cm do not approach the traditional level of statistical significance (p < 0.05) or the minimal clinically significant difference (MCID) criteria50. Because range-of-motion limitations in NSLBP are frequently influenced by pain-related protective mechanisms, fear of movement, muscle guarding, and altered motor control, which may not change sufficiently over a short intervention period, the lack of significant improvement in lumbar ROM in this context may reflect the distinction between structural and functional effects of traction. The absence of FFT improvement does not necessarily contradict pain reductions, but rather suggests that symptom relief and mobility restoration may not occur simultaneously. As a result, our results underscore the necessity for cautious interpretation of functional gains. Third, only group B (side-lying) demonstrated improvement in physical activity (IPAQ) scores compared to baseline, a finding that is also reported by Wang et al.11. However, this should not be interpreted as evidence of superiority over the other groups, as groups A and C did not show significant changes in this outcome. Previous research suggests that clinical improvements, such as reductions in pain and enhanced physical function, do not consistently translate into increased self-reported physical activity in individuals with chronic pain51. This reflects a recognized discrepancy between clinical outcomes and patient-reported activity levels52,53. In this context, the observed improvement in the side-lying group should be interpreted cautiously, as IPAQ is a self-reported measure that may capture perceived rather than actual changes in activity. Therefore, differences across groups may be influenced by the limitations of the measurement tool and individual variability rather than true positional effects54–57. Future studies are needed to confirm these findings using more objective measures of physical activity, such as accelerometry. By pinpointing the traction positions that help patients reduce pain and functional disability, our findings encourage therapists to align lumbar traction practices with the broader goals of sustainability in health and care within the physical therapy field. Limitations There are a number of limitations to this study. First, because only male participants were included, the findings cannot be generalized to female patients with NSLBP, particularly given the known sex-related differences in pain perception and treatment response. Second, the final analysis included 25 of the 30 randomized participants, which may have decreased statistical power and raised the possibility of type II error, especially for between-group comparisons. Third, the absence of a sham traction or control group limits the ability to determine whether the observed changes were due to specific effects of lumbar traction, placebo effects, therapist–patient interaction, or a natural recovery course. Further studies with appropriate control conditions are warranted to more accurately differentiate and quantify specific therapeutic effects of lumbar traction. Forth, future study would benefit from the use of objective tools like accelerometers, as IPAQ is a subjective self-reported measure that might not correctly reflect real changes in physical activity. Fifth, the relatively short intervention time and lack of long-term follow-up limit our ability to know if the observed improvements were maintained beyond the treatment phase. Future studies should include longer intervention periods and longer follow-up assessments to investigate the durability of treatment effects and establish the long-term effectiveness of lumbar traction applied in different positions. Conclusion Lumbar traction resulted in improvements in pain and disability across all three positions in adults with nonspecific low back pain (NSLBP). However, no statistically significant differences were observed between the supine, side-lying, and prone positions. Consequently, these findings do not indicate the superiority of any particular traction position. Although the supine group appeared to experience earlier improvements in pain and disability, and the side-lying group showed greater enhancement in physical activity, these findings should be interpreted with caution, as no significant differences were found between the two groups. Further research with larger sample sizes and more rigorous methodologies is required to determine whether traction position yields clinically meaningful effects. Acknowledgements The authors are grateful to the Deanship of scientific research at Prince Sattam bin Abdulaziz University, Saudi Arabia, for accomplishing this work. Author contributions WSM designed the study, collected the data, analyzed and interpreted the data, and drafted the manuscript. AAH contributed to study design, data analysis, and critical revision of the manuscript. AAM revised the manuscript for intellectual content. KMH and MMM conducted the statistical analysis, interpreted the data, and revised the manuscript. SH revised the manuscript for intellectual content. All authors reviewed and approved the final version of the manuscript. Funding The authors extend their appreciation to Prince Sattam bin Abdulaziz University for funding this research work through the Project Number (PSAU/2024/03/31744). Data availability The data generated and reported in the present study are available upon reasonable request. Declarations Competing interests The authors declare no competing interests. Clinical trial registration The study protocol was registered at ClinicalTrials.gov under the identifier NCT06812338. Ethics approval statement Ethical approval was obtained from the Research Ethics Committee of the Department of Health and Rehabilitation Sciences at Prince Sattam bin Abdulaziz University (Approval No. RHPT/024/020). This research was conducted in accordance with the ethical standards established by the Declaration of Helsinki. Patient consent statement All participants provided written informed consent before data collection began. They also agreed to the publication of their data in anonymized form in an online open-access journal. No personally identifiable information is included in this manuscript. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The data generated and reported in the present study are available upon reasonable request.