Effectiveness, Mediators and Moderators of Remote Live vs. Pre‐Recorded Physical Exercise Training for Chronic Low Back Pain

Background

Remote physical exercise training offers a more accessible alternative for the management of chronic low back pain (CLBP). However, the effectiveness of various remote delivery methods for this population is unclear.

Methods

This study was a randomized controlled trial comparing live interactive training via videoconference to individual pre‐recorded video training sessions in patients with CLBP. Ninety‐four individuals with CLBP were randomly assigned to a waitlist or to a 12‐week general exercise program delivered by a certified kinesiologist, either through live interactive or pre‐recorded video sessions. The training intensity for the 3 weekly one‐hour sessions was adapted to each participant's fitness level.

Results

A total of 57 patients (average age 48.4 ± 12.5 years, 36 women) completed the study, with comparable rates of attrition in the three groups. Those performing exercise, irrespective of modality, experienced significant and clinically worthwhile (> 20%) reductions in pain intensity ratings compared to waitlist. Similarly, exercise training (vs. waitlist) decreased movement‐evoked pain. No differences were observed between both forms of training for pain intensity, perceived difficulty, effort, fatigue, satisfaction or pain evoked during the training sessions. The therapeutic effects of physical exercise training were moderated by sex, with women reporting less pain relief, and mediated by improvements in self‐perceived general health, increased levels of physical activity and greater gait speed (4‐m gait task).

Conclusions

Altogether, these results support that, independently of the mode of delivery (live vs. pre‐recorded), remote exercise training is an accessible alternative to effectively alleviate CLBP.

Significance Statement

Live and pre‐recorded remote physical exercise training were equally effective in reducing pain intensity and movement‐evoked pain. Across modalities, the clinical effects of remote exercise appeared to be mediated by increased engagement in physical activity, faster gait speed, and improvements in self‐rated general health. Together, these findings suggest that exercise training may reduce pain by improving general health–related behaviours and beliefs.

2.1. Study Design

This study was a randomized controlled clinical trial conducted fully online from April 2020 to July 2024. This remote format was chosen to support feasible and flexible long‐term participation across diverse geographical locations in a post‐COVID19 context.

Participants were individuals with CLBP from Quebec (Canada), who were randomly assigned to either a 12‐week multimodal exercise training program via interactive live videoconferences or individual pre‐recorded videos, or else to a waitlist of the same duration. The study was registered at ClinicalTrials.gov (NCT05895630). All procedures were approved by the institutional ethics committee (CER VN20‐21‐07) and exercise interventions were reported following the Consensus on Exercise Reporting Template (Slade et al. 2016), an extension to the TIDieR checklist. Finally, trial results are reported according to the CONSORT guidance (see Data S1).

2.2. Study Participants

Participants were adults between 18 and 80 years of age with CLBP who were able to consent, could communicate, read, and understand French or English, with a stable internet connection and access to a computer or a tablet with a functioning webcam. Patients were recruited from databases of multiple locations across the province of Québec, including participant pools from the Quebec Pain Research Network, the Centre de Recherche de l'Institut Universitaire de Gériatrie de Montréal, the Association Québécoise de la Douleur Chronique, Mathieu Roy's laboratory, as well as from patients attending two partner rehabilitation clinics, and to a lesser extent, through advertisements. Patients and members of the public were involved exclusively at this stage of the trial to support dissemination of the study within their respective networks.

Interested participants were scheduled for a phone call with a member of the research team, who conducted a structured survey to determine the participant's eligibility (Figure 1). Eligible participants were subsequently sent a copy of the informed consent form by email, which was reviewed with the investigator, signed, and returned before participation in any study procedures. Eligible participants were contacted via email to provide instructions and links to the LimeSurvey (LimeSurvey GmbH, Germany) questionnaires that were used as outcome measures (see the Section 2.6). In addition, before randomization, they were scheduled for an online physical assessment via videoconference (see Section 2.6). Participants were instructed to complete the study questionnaires in the days preceding their physical assessment. Participants who successfully met the inclusion criteria after completing the questionnaires and physical assessment were randomized. See Figure 1 for a representation of the experimental protocol.

For inclusion, participants had to meet the criteria for CLBP diagnosis according to the Canadian version of the National Institutes of Health minimum dataset (NIH‐md) for CLBP research (Angarita‐Fonseca et al. 2023): LBP persisting for at least 3 months or pain recurring for at least half of the days in the past 6 months. Participants were excluded if they reported pain intensity < 4/10 (0: no pain, 10: worst pain imaginable) as per the Initiative on Methods, Measurement and Pain Assessment in Clinical Trials (IMMPACT) recommendations (Langford et al. 2023); if they had been diagnosed with any psychiatric or neurologic condition (not including depression or anxiety); if a specific or neuropathic origin for their CLBP was identified (Bardin et al. 2017), including non‐healed fractures, spinal stenosis, osteochondritis and low back sprains; if they received any of the following treatments from 3 months prior to their participation to study completion: facet block, spinal injections, radiotherapy or chemotherapy; and if they had any known contraindication to physical activity (assessed through the use of the physical activity readiness questionnaire for everyone–PAR‐Q+ (Warburton et al. 2011)).

2.3. Group Allocation

Once eligibility was confirmed, participants were randomly allocated to one of three groups in a 1:1:1 ratio. The randomization sequence was generated by the study coordinator prior to participant enrolment using a Google Sheet randomization algorithm (Google Inc., USA), and allocation was implemented according to this pre‐specified sequence. Each of the groups would be exposed to either a live interactive videoconference‐based exercise training group, an individual pre‐recorded video‐based exercise training group, or a waitlist control group (see Section 2.4 and Figure 1). Participants assigned to the latter were offered the possibility to participate in one of the exercise training groups at the end of the waitlist period.

2.4. Exercise Training

Both exercise training programs consisted of three 60‐min weekly sessions, offering a total of approximately 180 min of training per week. Participants were instructed to prepare a space in their home to conduct the exercises, ideally a clear area of about 2 × 2 m, preferably with a mat, while wearing comfortable clothing and running or similar shoes. Each participant was categorized into one of three levels of training intensity, based on their performance according to a mobility decisional tree including a series of balance, mobility and functional tests conducted during the initial online physical assessment (see Section 2.7). This categorization was used to adapt the number of repetitions or the duration of exercises.

The exercise interventions were designed by the Laboratoire du muscle et de sa fonction (Université du Québec à Montréal). Three certified kinesiologists, each with prior experience in adapted physical exercise training for CLBP, were responsible for delivering live training or recording the same training sessions. Both trainings consisted of a combination of exercises (aerobic, resistance and flexibility) of low to moderate intensity, with kinesiologists progressively adjusting the intensity throughout the 12 weeks of intervention, based on each participant's capacity. Each of the three sessions focused more on one of the three components (first session: flexibility; second session: aerobic; third session: resistance). None of the performed exercises required any type of technical gear, beyond elastic bands or small weights, which could be replaced with other objects available at the participant's home. For a detailed account of the exercises used, please refer to Data S2.

The live interactive exercise training (‘live’ from hereon) was conducted in groups of 5 to 10 people via the Zoom videoconference platform (Zoom Video Communications Inc., USA), on a fixed schedule. The link to every session was sent by email the night before, plus a reminder email prior to every week. The training sessions had a duration of 60 min, followed by an optional 15 min of virtual social interaction with the kinesiologist and other group members. This training mode allowed the kinesiologist to correct participants in real time during the sessions while creating a sense of belonging. When participants missed a group session, a video recording of the class was made available for them via a website.

The pre‐recorded video exercise training (‘pre‐recorded’) was conducted individually without interactions (except for the initial contacts with the research team) via a secured web platform provided by the Laboratoire du muscle et de sa fonction. Participants were instructed to complete three sessions every week for 12 weeks, with a day of rest between sessions. When missing a video session, participants were advised to make up for it on the following day or during the weekend. Participants were able to contact a person from the research team by phone or email to ask questions related to the sessions. The training sessions had the same duration (60 min), content and intensity progression as the live session. Compared to live sessions (specific schedule and group session), this training mode allowed for a more flexible pace and timing of each session.

Participants in the waitlist control group were instructed to continue their lifestyle habits throughout the duration of the study, and to notify the research team if they started a new treatment during the waitlist period, which would result in their exclusion. At the end of the waitlist period, they were offered access to the pre‐recorded videos. No remuneration was offered for study participation, but all interventions were provided free of charge.

2.5. Primary Outcome: Pain Intensity

The primary outcome was average LBP intensity for the previous 7 days. It was assessed online as an item of the NIH‐md before and after training or waitlist using an 11‐point numerical rating scale (NRS; 0 = no pain, 10 = worst pain imaginable).

The NIH‐md was promoted as a standardized measure to assess the multidimensionality of CLBP across studies (Deyo et al. 2014). Beyond pain intensity, it includes a total of 40 items and multiple domains to characterize key medical history features (including demographics, involvement in workers' compensation or legal claims, work status, education, comorbidities, and previous treatment history), to stratify CLBP based on its impact, and to assess physical function, psychological function, and sleep disturbance. The majority of items included in the NIH‐md were drawn from the short‐form Patient Reported Outcomes Measurement Information System (PROMIS) (Deyo et al. 2014). This study used the cross‐cultural adaptation to the Canadian setting, which has shown good‐to‐excellent internal consistency across most of its domains (Angarita‐Fonseca et al. 2023).

2.6. Secondary Outcomes: Psychosocial Questionnaires

Secondary outcomes were selected to capture both the multidimensional impact of CLBP and potential biopsychosocial and demographic factors that may influence or explain response to exercise training. These include pain impact, physical function, emotional distress, sleep disturbance, pain‐related beliefs, personality traits, performance on basic functional tasks, and movement‐evoked pain. Secondary outcomes were assessed both through NIH‐md domains and through additional validated questionnaires.

By means of the NIH‐md, multiple CLBP domains were assessed before and after training or waitlist. These included five items related to the CLBP history, 4 items on current and past medical interventions for CLBP, four items on pain interference (which generated a subscale rated from 4 to 20), four items on physical function (generating another subscale rated equally 4–20), four items on emotional distress or depression (another 4–20 subscale), four items on sleep disturbance (a last 4–20 subscale), two items on absenteeism due to CLBP, one item on kinesiophobia and one on pain catastrophizing, two items on substance abuse, one item on smoking and two items to calculate the patient's body‐mass index. The Pain Impact Stratification score, comprising the pain intensity item and subscales on pain interference and physical function, was used to categorize individuals with ‘mild’ (8–27 points), ‘moderate’ (28–34 points), or ‘severe’ (35–50 points) CLBP impact. This score has shown moderate to strong correlations with other common tools used to assess LBP‐related disability, like the Oswestry and Roland‐Morris questionnaires (Deyo et al. 2014). A single‐item self‐assessment of current general health status rated from 0 (worst possible) to 100 (best possible general health status) was derived from the EQ‐5D quality of life questionnaire (Balestroni and Bertolotti 2012; Granet et al. 2023b) and included after the NIH‐md.

In addition to the NIH‐md, a series of validated questionnaires were used to assess multiple dimensions of psychosocial health and personality traits before and after training or waitlist (Figure 1). To comprehensively assess catastrophizing and kinesiophobia as psychological constructs potentially influencing the effects of exercise training, participants completed the Pain Catastrophizing Scale (PCS) and the Tampa Scale for Kinesiophobia (TSK‐11). The Pain Catastrophizing Scale (Sullivan et al. 1995) assesses catastrophic thoughts related to pain. This scale has been validated in multiple chronic pain populations, with acceptable levels of reliability and validity (Sullivan et al. 2001). In addition, fear of movement was measured through the Tampa Scale of Kinesiophobia (Tkachuk and Harris 2012; Vlaeyen et al. 1995), which is a reliable and valid measure of fear of movement and (re)injury for individuals with chronic pain.

Depression and anxiety symptoms are frequently comorbid with CLBP, and as such, each was evaluated through additional questionnaires: the Beck Depression Inventory II (BDI) and the State–Trait Anxiety Inventory (STAI; (Beck et al. 1996; Spielberger et al. 1971)). Both questionnaires have shown good construct validity and internal consistency across chronic pain populations (Harris and D'Eon 2008; Novy et al. 1993). To assess sleep outcomes, we used the Insomnia Severity Index (ISI; (Bastien et al. 2001) scored from 0 to 28), which has demonstrated moderate accuracy for screening insomnia in patients with LBP (Alsaadi et al. 2013).

Finally, the Big Five Inventory (McCrae and John 1992) was used to explore personality traits across five dimensions and their potential role in mediating or moderating the effects of exercise training.

2.7. Secondary Outcomes: Physical Function

Once the questionnaires were completed and before initiating the training program or waitlist period, participants underwent an online physical assessment with one of the study's investigators via videoconference, as in (Peyrusque et al. 2022), which also served to confirm eligibility. The same investigator performed the baseline and endpoint physical assessments for each individual participant. Only participants who successfully met the inclusion criteria after completing the physical assessment were randomized. The same physical assessment was repeated at the end of the training or waiting period (Figure 1).

Physical function was evaluated through the Short Physical Performance Battery (SPPB; (Di Iorio et al. 2007; Guralnik et al. 1994)), including a balance test measured in seconds, a 30‐s chair stand repetition test, and the 4‐m gait speed test at comfortable and maximal paces (Bohannon and Wang 2019). The timed‐up and go (TUG; (Gautschi et al. 2016)) test and the Takai sit‐to‐stand test completed the assessment (Takai et al. 2009). Pain evoked by each functional task was measured by asking participants to rate their pain at rest and immediately after each task, using the same 11‐point NRS. Task‐specific movement‐evoked pain was computed by subtracting the score at rest from that after each task (Butera et al. 2024; Leemans et al. 2022). As recommended, a total composite or aggregate score (ranging from −50 to +50) was obtained with the addition of the scores corresponding to each of the five individual tasks. Finally, the Rapid Assessment of Physical Activity (RAPA, (Azfar et al. 2019) scored from 1 to 7) was completed at baseline and at the end of the study. Combining scores from the functional tasks and the questionnaires, a decision tree was designed to determine the participant's level of physical fitness (based on (Granet et al. 2023b)). According to the score obtained, participants were assigned to one of three levels of exercise training intensity.

2.8. Tertiary Outcomes: Individual Session Evaluations

Before and after each individual training session, participants were provided two links to complete two short questionnaires enquiring about how they felt overall, their level of fatigue, and their LBP intensity while standing, sitting and walking, using the same 11‐point NRS (both before and after training). In addition, after each session, they were asked to rate their level of satisfaction, perceived difficulty, effort and pain experienced with regards to the session. All these dimensions were rated on 0 to 10 scales, except satisfaction, which was rated on a Likert scale ranging from 1 (not satisfied) to 4 (very satisfied). These questionnaires were used to monitor attendance, understand individual participant's response to exercise, and surveil adverse reactions. Adverse reactions were identified based on increases (≥ 3 points or repeated increases) in pain intensity, fatigue or negative perceptions of the session, which were considered indicators of an adverse exercise experience. The 3‐point pain threshold was selected ad hoc, informed by previously established thresholds for clinically important changes in pain intensity (van der Roer et al. 2006).

2.9. Sample Size Calculation

We hypothesized that both methods of exercise training would be effective in reducing pain intensity when compared to the waitlist control. A recently completed study in our laboratory (Bergevin et al. 2026) found that a very similar exercise training intervention (three hourly sessions every week for 14 weeks) in the gym, for the same population (CLBP), was effective at reducing pain intensity compared to a waitlist with very large effect sizes (f = 0.67). Given the uncertainty regarding whether remote training would achieve comparable effects to in‐person, we opted for a more conservative effect size estimate (f = 0.5) for our power calculations. Based on this effect size, a sample of 15 participants per group would be associated with a statistical power of 1‐β = 0.90 (G*Power 3.1.9.4121). Our dropout rate for the previous study was ~20%. We therefore aimed to enrol a minimum of 54 participants in total (18 per group) to reach our desired sample size.

2.10. Statistical Analysis

Analyses were performed using RStudio (v2024.04.2; RStudio Team, USA) and JASP (v0.18.3, JASP Team, The Netherlands). The primary and secondary outcomes were analysed using two distinct linear mixed‐effects models (LMM), accounting for repeated measures nested within participants. The first model examined differences between exercise training and waitlist, while the second assessed differences between live and pre‐recorded (reference category). Models examining the primary outcome included sex and age as covariates. Assumptions of normality and homoscedasticity were checked using residual plots and scatterplots of residuals vs. fitted values, and the level of significance was set at α < 0.05. Bootstrapped confidence intervals were calculated for the estimates using 5000 iterations at the participant level.

All randomized participants' data were included in an initial intention‐to‐treat analysis of the primary outcome of pain intensity ratings. Logistic regression with intervention groups, age, sex and baseline pain intensity as predictors was used to assess the pattern of data missingness (Schafer and Graham 2002). Missing data were handled through multiple imputation procedures using the mice package in R. Five imputations were generated using predictive mean matching (Vink et al. 2015) with 50 iterations. Imputed values were aggregated by calculating their mean across imputations. To confirm the robustness of the intention‐to‐treat findings, a per‐protocol analysis followed, including only participants who adhered to the assigned intervention protocol. Numbers needed to treat for a minimal worthwhile effect on pain intensity (Hansford et al. 2024) were calculated. For exploratory purposes, pain intensity ratings were also compared to those from a similar CLBP population participating in an in‐person exercise intervention delivered by our group; details are provided in Data S2. Finally, demographic and baseline clinical variables were introduced into the LMMs as second‐level (between‐subjects) predictors to assess their potential roles as moderators of the effects of exercise training on pain intensity ratings.

To assess participation in exercise sessions and participants' perceptions, we compared the responses to the questionnaires provided before and after each training session between the live and pre‐recorded groups using non‐parametric non‐paired Mann–Whitney tests. It should be noted that participation could not be systematically tracked in the pre‐recorded group, as completion of the session questionnaires was voluntary. We conducted the same analysis to compare the same variables between participants completing the study and those not, so as to better understand which factors may have contributed to dropout rates.

Finally, we ran multilevel mediation analyses to assess the role of psychological and physical measures in exercise‐induced pain relief. Two separate analyses were conducted for each mediator, one examining the relationship between exercise training vs. waitlist (predictor) and pain intensity ratings (outcome) over time, and a second assessing significant mediators in the relationship between live vs. pre‐recorded (predictor) and pain intensity ratings (outcome), also over time. To this end, we employed linear mixed‐effects models fitted by restricted maximum likelihood (REML), which enabled us to account for participant‐level variations and temporal fluctuations. The nlme package was used following the approach proposed to estimate random effects for the a, b, and c' paths (Bauer et al. 2006). This method allowed us to compute the covariance between the random effects of the a and b paths to estimate the indirect effect. Finally, custom code was used to obtain bootstrapped distributions of the indirect effects (ab path) using 5000 iterations at the participant level (Leeden et al. 2008), and to compute the confidence intervals for the direct, indirect and total effects.

3.1. Recruited Sample

The team had initial contact with 977 persons and 71% of these did not follow‐up or answer to subsequent contacts (see Figure 2 for the CONSORT flowchart). Two‐hundred‐eighty‐seven were screened for eligibility and 42% were excluded, most often due to concurrent health conditions or insufficient pain intensity. One‐hundred‐twenty‐four underwent the online physical assessment, out of which, 24% were excluded, mostly because they reported insufficient pain intensity at that time. Finally, 94 patients were randomized to one of the three groups, although 9 of them dropped out before initiating the protocol for a variety of reasons (e.g., training schedule, group assignment, unwillingness to take part in the waitlist). Therefore, 85 patients participated in the protocol and 67% completed it. The rate of attrition did not differ between groups (𝜒² = 2.37; p = 0.30). About half of the participants not completing the training or waitlist period were lost to follow‐up (n = 15), only 3 for reasons related to the training program itself (schedule or internet connection), and 10 for health reasons unrelated to CLBP, including COVID19 infection and pregnancy (see Figure 2). For a number of participants, the specific reason for not completing the program could not be determined. The number of exercise training sessions attended could only be reliably tracked for those on the live training program (18.94 sessions on average for the group, 25.72 for those completing the training).

Details on sociodemographic variables of all patients completing their participation are available in Table 1. Baseline scores on psychosocial questionnaires are available in Table 2.

3.2. Primary Outcome: Pain Intensity

For the intention‐to‐treat analysis, none of the variables significantly predicted missingness (p's > 0.05), which is consistent with the assumption that the data were missing completely at random. Missing data points for pain intensity ratings post‐intervention were handled using multiple imputation and the imputed values were aggregated. The intention‐to‐treat models showed that pain intensity significantly decreased for those who participated in exercise training when compared to waitlist (B = −1.94, 95% CI [−2.81 to −1.08], t ₈₄ = −4.48, p < 0.001). However, there were no differences between both forms of training (B = 0.25, 95% CI [−0.70 to 1.20], t ₈₁ = 0.52, p = 0.60). Adjusting for sex and age did not improve model fit, nor alter the primary treatment effect estimates, which remained significant (in both models: B = −1.94, t ₈₂ = −4.42, p < 0.001).

These findings were confirmed with the per‐protocol analysis (see Figure 3A–C), which also showed a decrease in pain intensity after exercise vs. waitlist (B = −2.43, 95% CI [−3.44 to −1.40], t ₅₅ = −4.59, p < 0.001), and no differences between live and pre‐recorded training (B = −0.51, 95% CI [−1.61 to 0.59], t ₃₈ = −0.89, p = 0.37). The rest of the analyses were conducted with the per‐protocol dataset. Supporting Information S2 contains additional analyses comparing both forms of remote exercise training to in‐person delivery.

A total of 31 patients engaging in physical exercise training achieved the smallest worthwhile effect for physical exercise (16/21 patients in the pre‐recorded and 15/18 in the live training groups). The numbers needed to treat for exercise training to reach the smallest worthwhile pain reduction were 1.75 participants (1.85 and 1.64 participants in the pre‐recorded and live groups, respectively). An exploratory comparison with in‐person exercise training is reported in Data S2.

3.3. Secondary Outcomes

3.4. Tertiary Outcomes

Self‐reported participation on both forms of exercise training was similar on average (18.94 vs. 14.37, live vs. pre‐recorded sessions; Mann–Whitney U = 634, p = 0.4); however, voluntary reporting of participation in pre‐recorded sessions warrants cautious interpretation. The mean perceived difficulty was similar for both forms of training (live: 6.42 vs. pre‐recorded: 6.52, U = 394, p = 0.57), and so was perceived effort (live: 5.88 vs. pre‐recorded: 6.31, U = 357, p = 0.25), pain in the low back (live: 4.20 vs. pre‐recorded: 4.41, U = 400, p = 0.63), fatigue (live: 4.04 vs. pre‐recorded: 4.34, U = 393, p = 0.56), and satisfaction (live: 3.38 vs. pre‐recorded: 3.21 out of 4, U = 349, p = 0.21) with the training sessions. In an attempt to understand whether the reasons for dropping out were related to dissatisfaction or adverse responses to exercise training, these were compared between those completing training and those dropping out. On average, the latter completed 6.73 sessions in 18.02 days, when compared to 25.72 sessions in 69.13 days for those completing their training. The mean perceived difficulty was similar for those who dropped out vs. those completing training (6.75 vs. 6.28 out of 10, respectively, Mann–Whitney U = 363, p = 0.38), and so was perceived effort (6.23 vs. 6.04 out of 10, respectively, U = 397, p = 0.72), pain (4.42 vs. 4.21 out of 10, respectively, U = 401, p = 0.77), fatigue (4.71 vs. 3.86 out of 10, respectively, U = 300, p = 0.06) and satisfaction (3.25 vs. 3.32 out of 4, respectively, U = 344, p = 0.24). No specific adverse reactions to exercise were identified through this system.

3.5. Moderators of the Clinical Effects

The pre‐intervention sociodemographic and clinical characteristics were investigated as potential moderators of the effect of exercise training on pain intensity scores. The models revealed that females benefited less from physical exercise (B = 2.58, 95% CI [0.44 to 4.67], t ₁₅₃ = 2.39, p = 0.020). In addition, symptom duration shorter than 5 years and no history of work absenteeism (≥ 1 month) due to CLBP were also associated with a reduced effect of exercise on pain ratings (B = 2.28, 95% CI [0.13–4.36], t ₅₁ = 2.10, p = 0.04; and B = –3.03, 95% CI [−5.79 to −0.21], t ₂₉ = −2.16, p = 0.039, respectively). Finally, the personality trait ‘openness to experience’, was also found to moderate the effects of exercise training on pain intensity (B = −0.39, 95% CI [−0.72 to −0.04], t ₄₂ = −2.10, p = 0.041). Higher openness scores were associated with better response to exercise.

We found no moderating effect of age, income, level of education, employment status, smoking or substance use habits, opioid medication use, BMI, pain frequency, presence of leg pain, presence of pain in other body sites, psychosocial symptoms or other personality traits.

3.6. Mediators of Clinical Effects

All questionnaire and physical assessment variables were examined as potential mediators through multilevel mediation analyses to first examine the relationship between both exercises compared to waitlist and pain relief, and to compare both exercise training delivery modes to one another thereafter. Self‐rated general health status, 4‐m gait speed test (both at comfortable and maximum pace), and the RAPA scores were found to significantly mediate the effects of exercise on pain intensity.

4.1. Worthwhile Effects of Remote Exercise Training

Both forms of remote exercise reduced pain intensity by over 20% relative to no intervention (−2.69 points for live and −2.12 points for pre‐recorded training), exceeding the smallest worthwhile effect. This patient‐centered threshold has been proposed as a robust marker of clinical significance reflecting patients' sentiment that the benefits of exercise outweigh its risks, costs, and inconveniences (Hansford et al. 2024; Innocenti et al. 2024). The numbers needed to treat to reach this threshold were relatively low for both forms of exercise (1.85 for pre‐recorded and 1.64 for live). Moreover, exercise training mitigated the impact of CLBP in our sample. Building upon these arguments, our exploratory analysis showed that remote training may yield relief of similar magnitude to in‐person gym sessions (see Data S2). These results align with the emerging literature portraying remote exercise training as a compelling alternative for managing CLBP (Shi et al. 2024).

In contrast with previous reports from remote and in‐person exercise interventions (Bergevin et al. 2026; Ozden et al. 2022), exercise training was not effective in reducing fear of movement or catastrophic pain beliefs in our participants. This could reflect a statistical power issue; however, it is also possible that these benefits of exercise training rely, at least partly, on appropriate supervision and follow‐up by trained clinicians (Ozden et al. 2022), which were not systematically integrated into interventions.

4.2. Nonsuperiority of Either Mode of Remote Exercise Training

Training via pre‐recorded videos or live videoconference sessions led to comparable reductions in most clinical outcomes. A priori, live exercise training may offer several advantages that could enhance its clinical benefits, such as supervision, social facilitation, and a fixed schedule that promotes habit formation. Supervision could facilitate better and safer performance, thus greater pain relief when compared to unsupervised methods (Mataran‐Penarrocha et al. 2020). However, initial contact with a professional and visual support through the provision of videos, as in our pre‐recorded intervention, may suffice to reassure participants. This could explain the noninferiority and similar number of dropouts observed with this format when compared to live training.

Interacting with peers and the instructor in live training was also expected to have a positive impact on psychosocial and functional metrics (Granet et al. 2023a). Yet, we detected no differences for any of these outcomes. Regular clinician contact and follow‐up may represent important components of exercise interventions contributing to improvements in functional performance and psychosocial outcomes. However, this remains speculative.

Despite null effects on task performance, exercise training reduced pain evoked by functional tasks but not pain at rest. For three out of five tasks, reductions in evoked pain exceeded a proposed cutoff for minimal clinically important changes (≤ −1.1; Fleagle et al. 2024), irrespective of training mode. Thus, physical exercise, independent of delivery mode, may specifically target mechanisms contributing to movement‐evoked pain (Leemans et al. 2022).

4.3. Moderators of Response

On average, participants in both training formats reported similar levels of effort, fatigue, pain and satisfaction after sessions. Given that outcomes were also comparable, it remains important to elucidate the factors moderating treatment response. We found that females experienced less relief from exercise, a moderation effect that replicates our previous findings on in‐person training (Bergevin et al. 2026). Females tend to obtain less gain from rehabilitation programs, partially due to a higher burden from domestic, employment and caregiving responsibilities (Ott et al. 2022). Biological mechanisms, including sex differences in the activation of endogenous opioid pathways (Bruehl et al. 2020) and the anti‐hyperalgesic effects of exercise training (Lesnak et al. 2022), could also account for these sex differences.

Exercise was also more effective for those with a history of absenteeism due to CLBP. Workplace interventions were also found to be more effective for those with a history of LBP‐related sick leave (Steenstra et al. 2009). These individuals may be more aware of physical corrections needed to prevent LBP recurrence, which may translate to exercise training. Openness scores also moderated the effects of exercise. Individuals with relatively higher levels of openness may be more prone to engage in physical activity and to experience enhanced affective responses (Wilson and Dishman 2015), which may positively influence exercise expectancy and outcomes (Yue et al., n.d.). Finally, we found greater benefits of exercise in CLBP of longer duration, which may reflect exercise training targeting physical deconditioning associated with chronicity (Roren et al. 2023).

4.4. Mediators of Response

Self‐perceived improvements in general health mediated the effect of exercise training on pain intensity. CLBP and chronic pain more broadly have been linked to poor general health (Hestbaek et al. 2003; Mantyselka et al. 2003). Thus, targeting a key determinant of general health, i.e., physical activity (Bize et al. 2007), could plausibly contribute to CLBP reduction. Consistent with this idea, RAPA scores also mediated the effects of exercise on pain intensity. Given that low physical activity levels are prevalent among individuals with CLBP (Alzahrani et al. 2019), increasing regular physical activity may improve general health, thereby alleviating CLBP (Ambrose and Golightly 2015; Geneen et al. 2017).

In addition, increases in gait speed during a functional test also mediated the effects of exercise training. Lower walking speed and volume are independently associated with CLBP (Haddadj et al. 2025; Smith et al. 2022). While gait speed may serve as a surrogate for general health (Studenski 2019), regular walking alone can reduce back pain recurrence (Pocovi et al. 2024). Therefore, these results may suggest that a general remote exercise program, by increasing physical activity levels, can enhance individuals' perception of their overall health, which in turn may lead to reductions in pain intensity. These effects appear to be independent of the training delivery mode.

Data S1: CONSORT and CERT checklists.

Data S2: Supplementary Methods and Results.

1. Alsaadi, S. M. , McAuley J. H., Hush J. M., et al. 2013. “Detecting Insomnia in Patients With Low Back Pain: Accuracy of Four Self‐Report Sleep Measures.” BMC Musculoskeletal Disorders 14: 196. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Alzahrani, H. , Mackey M., Stamatakis E., Zadro J. R., and Shirley D.. 2019. “The Association Between Physical Activity and Low Back Pain: A Systematic Review and Meta‐Analysis of Observational Studies.” Scientific Reports 9: 8244. [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Ambrose, K. R. , and Golightly Y. M.. 2015. “Physical Exercise as Non‐Pharmacological Treatment of Chronic Pain: Why and When.” Best Practice & Research. Clinical Rheumatology 29: 120–130. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Angarita‐Fonseca, A. , Page M. G., Meloto C. B., et al. 2023. “The Canadian Version of the National Institutes of Health Minimum Dataset for Chronic Low Back Pain Research: Reference Values From the Quebec Low Back Pain Study.” Pain 164: 325–335. [DOI] [PMC free article] [PubMed] [Google Scholar]
5. Azfar, S. M. , Murad M. A., Azim S., and Baig M.. 2019. “Rapid Assessment of Physical Activity and Its Association Among Patients With Low Back Pain.” Cureus 11: e6373. [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Balestroni, G. , and Bertolotti G.. 2012. “EuroQol‐5D (EQ‐5D): An Instrument for Measuring Quality of Life.” Monaldi Archives for Chest Disease 78: 155–159. [DOI] [PubMed] [Google Scholar]
7. Bardin, L. D. , King P., and Maher C. G.. 2017. “Diagnostic Triage for Low Back Pain: A Practical Approach for Primary Care.” Medical Journal of Australia 206: 268–273. [DOI] [PubMed] [Google Scholar]
8. Bastien, C. H. , Vallieres A., and Morin C. M.. 2001. “Validation of the Insomnia Severity Index as an Outcome Measure for Insomnia Research.” Sleep Medicine 2: 297–307. [DOI] [PubMed] [Google Scholar]
9. Bauer, D. J. , Preacher K. J., and Gil K. M.. 2006. “Conceptualizing and Testing Random Indirect Effects and Moderated Mediation in Multilevel Models: New Procedures and Recommendations.” Psychological Methods 11: 142–163. [DOI] [PubMed] [Google Scholar]
10. Beck, A. T. , Steer R. A., and Brown G. K.. 1996. Beck depression inventory.
11. Bergevin, M. , Bendas A., Bobeuf F., et al. 2026. “Unraveling the Pain Trajectory in Chronic Low Back Pain Patients During a Physical Exercise Training Program.” Journal of Science and Medicine in Sport 29, no. 1: 42–50. [DOI] [PubMed] [Google Scholar]
12. Bize, R. , Johnson J. A., and Plotnikoff R. C.. 2007. “Physical Activity Level and Health‐Related Quality of Life in the General Adult Population: A Systematic Review.” Preventive Medicine 45: 401–415. [DOI] [PubMed] [Google Scholar]
13. Bohannon, R. W. , and Wang Y. C.. 2019. “Four‐Meter Gait Speed: Normative Values and Reliability Determined for Adults Participating in the NIH Toolbox Study.” Archives of Physical Medicine and Rehabilitation 100: 509–513. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Bruehl, S. , Burns J. W., Koltyn K., et al. 2020. “Are Endogenous Opioid Mechanisms Involved in the Effects of Aerobic Exercise Training on Chronic Low Back Pain? A Randomized Controlled Trial.” Pain 161: 2887–2897. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Butera, K. A. , Chimenti R. L., Alsouhibani A. M., et al. 2024. “Through the Lens of Movement‐Evoked Pain: A Theoretical Framework of the “Pain‐Movement Interface” to Guide Research and Clinical Care for Musculoskeletal Pain Conditions.” Journal of Pain 25: 104486. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Cecchi, F. , Pasquini G., Paperini A., et al. 2014. “Predictors of Response to Exercise Therapy for Chronic Low Back Pain: Result of a Prospective Study With One Year Follow‐Up.” European Journal of Physical and Rehabilitation Medicine 50: 143–151. [PubMed] [Google Scholar]
17. da Silva, T. , Mills K., Brown B. T., Herbert R. D., Maher C. G., and Hancock M. J.. 2017. “Risk of Recurrence of Low Back Pain: A Systematic Review.” Journal of Orthopaedic and Sports Physical Therapy 47: 305–313. [DOI] [PubMed] [Google Scholar]
18. de Roode, A. , Heymans M. W., van Lankveld W., and Staal J. B.. 2024. “The Impact of Contextual Effects in Exercise Therapy for Low Back Pain: A Systematic Review and Meta‐Analysis.” BMC Medicine 22: 484. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Deyo, R. A. , Dworkin S. F., Amtmann D., et al. 2014. “Report of the NIH Task Force on Research Standards for Chronic Low Back Pain.” Journal of Pain 15: 569–585. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Di Iorio, A. , Abate M., Guralnik J. M., et al. 2007. “From Chronic Low Back Pain to Disability, a Multifactorial Mediated Pathway: The InCHIANTI Study.” Spine 32: E809–E815. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Eysenbach, G. 2005. “The Law of Attrition.” Journal of Medical Internet Research 7: e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Fatoye, F. , Gebrye T., Fatoye C., et al. 2020. “The Clinical and Cost‐Effectiveness of Telerehabilitation for People With Nonspecific Chronic Low Back Pain: Randomized Controlled Trial.” JMIR mHealth and uHealth 8: e15375. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Fernandes, L. G. , Devan H., Fioratti I., Kamper S. J., Williams C. M., and Saragiotto B. T.. 2021. “At My Own Space, Pace and Place: A Systematic Review of Qualitative Studies of Enablers and Barriers to Telehealth Interventions for People With Chronic Pain.” Pain 163, no. 2: e165–e181. [DOI] [PubMed] [Google Scholar]
24. Ferreira, M. L. , de Luca K., Haile L. M., et al. 2023. “Global, Regional, and National Burden of Low Back Pain, 1990–2020, Its Attributable Risk Factors, and Projections to 2050: A Systematic Analysis of the Global Burden of Disease Study 2021.” Lancet Rheumatology 5: e316–e329. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Fleagle, T. R. , Post A. A., Dailey D. L., et al. 2024. “Minimal Clinically Important Change of Movement Pain in Musculoskeletal Pain Conditions.” Journal of Pain 25: 104507. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Fritz, J. M. , Ford I., George S. Z., et al. 2024. “Telehealth Delivery of Physical Therapist‐Led Interventions for Persons With Chronic Low Back Pain in Underserved Communities: Lessons From Pragmatic Clinical Trials.” Frontiers in Pain Research 5: 1324096. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Garofano, M. , Del Sorbo R., Calabrese M., et al. 2025. “Remote Rehabilitation and Virtual Reality Interventions Using Motion Sensors for Chronic Low Back Pain: A Systematic Review of Biomechanical, Pain, Quality of Life, and Adherence Outcomes.” Technologies 13: 186. [Google Scholar]
28. Gautschi, O. P. , Smoll N. R., Corniola M. V., et al. 2016. “Validity and Reliability of a Measurement of Objective Functional Impairment in Lumbar Degenerative Disc Disease: The Timed up and Go (TUG) Test.” Neurosurgery 79: 270–278. [DOI] [PubMed] [Google Scholar]
29. Gell, N. , Hoffman E., and Patel K.. 2021. “Technology Support Challenges and Recommendations for Adapting an Evidence‐Based Exercise Program for Remote Delivery to Older Adults: Exploratory Mixed Methods Study.” JMIR Aging 4: e27645. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Geneen, L. J. , Moore R. A., Clarke C., Martin D., Colvin L. A., and Smith B. H.. 2017. “Physical Activity and Exercise for Chronic Pain in Adults: An Overview of Cochrane Reviews.” Cochrane Database of Systematic Reviews 4: CD011279. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Gilanyi, Y. L. , Shah B., Cashin A. G., et al. 2024. “Barriers and Enablers to Exercise Adherence in People With Nonspecific Chronic Low Back Pain: A Systematic Review of Qualitative Evidence.” Pain 165: 2200–2214. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Grande, G. H. D. , Vidal R. V. C., Salini M. C. R., Christofaro D. G. D., and Oliveira C. B.. 2025. “Barriers and Facilitators to Physical Activity and Exercise Among People With Chronic Low Back Pain: A Qualitative Evidence Synthesis.” Journal of Orthopaedic and Sports Physical Therapy 55: 312–330. [DOI] [PubMed] [Google Scholar]
33. Granet, J. , Peyrusque E., Ruiz F., et al. 2023a. “Online Physical Exercise Intervention in Older Adults During Lockdown: Can We Improve the Recipe?” Aging Clinical and Experimental Research 35: 551–560. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Granet, J. , Peyrusque E., Ruiz F., et al. 2023b. “Web‐Based Physical Activity Interventions Are Feasible and Beneficial Solutions to Prevent Physical and Mental Health Declines in Community‐Dwelling Older Adults During Isolation Periods.” Journals of Gerontology. Series A, Biological Sciences and Medical Sciences 78: 535–544. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Guralnik, J. M. , Simonsick E. M., Ferrucci L., et al. 1994. “A Short Physical Performance Battery Assessing Lower Extremity Function: Association With Self‐Reported Disability and Prediction of Mortality and Nursing Home Admission.” Journal of Gerontology 49: M85–M94. [DOI] [PubMed] [Google Scholar]
36. Haddadj, R. , Nordstoga A. L., Nilsen T. I. L., et al. 2025. “Volume and Intensity of Walking and Risk of Chronic Low Back Pain.” JAMA Network Open 8: e2515592. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Hansford, H. J. , Jones M. D., Cashin A. G., et al. 2024. “The Smallest Worthwhile Effect on Pain Intensity of Exercise Therapy for People With Chronic Low Back Pain: A Discrete Choice Experiment Study.” Journal of Orthopaedic & Sports Physical Therapy 54, no. 7: 477–485. [DOI] [PubMed] [Google Scholar]
38. Harris, C. A. , and D'Eon J. L.. 2008. “Psychometric Properties of the Beck Depression Inventory‐Second Edition (BDI‐II) in Individuals With Chronic Pain.” Pain 137: 609–622. [DOI] [PubMed] [Google Scholar]
39. Hayden, J. A. , Ellis J., Ogilvie R., et al. 2021. “Some Types of Exercise Are More Effective Than Others in People With Chronic Low Back Pain: A Network Meta‐Analysis.” Journal of Physiotherapy 67: 252–262. [DOI] [PubMed] [Google Scholar]
40. Hayden, J. A. , van Tulder M. W., and Tomlinson G.. 2005. “Systematic Review: Strategies for Using Exercise Therapy to Improve Outcomes in Chronic Low Back Pain.” Annals of Internal Medicine 142: 776–785. [DOI] [PubMed] [Google Scholar]
41. Hayden, J. A. , Wilson M. N., Stewart S., et al. 2020. “Exercise Treatment Effect Modifiers in Persistent Low Back Pain: An Individual Participant Data Meta‐Analysis of 3514 Participants From 27 Randomised Controlled Trials.” British Journal of Sports Medicine 54: 1277–1278. [DOI] [PubMed] [Google Scholar]
42. Hestbaek, L. , Leboeuf‐Yde C., and Manniche C.. 2003. “Is Low Back Pain Part of a General Health Pattern or Is It a Separate and Distinctive Entity? A Critical Literature Review of Comorbidity With Low Back Pain.” Journal of Manipulative and Physiological Therapeutics 26: 243–252. [DOI] [PubMed] [Google Scholar]
43. Innocenti, T. , Schleimer T., Salvioli S., Giagio S., Ostelo R., and Chiarotto A.. 2024. “In Trials of Physiotherapy for Chronic Low Back Pain, Clinical Relevance Is Rarely Interpreted, With Great Heterogeneity in the Frameworks and Thresholds Used: A Meta‐Research Study.” Journal of Physiotherapy 70: 51–64. [DOI] [PubMed] [Google Scholar]
44. Jiang, X. , Tang L., Zhang Y., et al. 2024. “Does Sedentary Time and Physical Activity Predict Chronic Back Pain and Morphological Brain Changes? A UK Biobank Cohort Study in 33,402 Participants.” BMC Public Health 24: 2685. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Korownyk, C. S. , Montgomery L., Young J., et al. 2022. “PEER Simplified Chronic Pain Guideline: Management of Chronic Low Back, Osteoarthritic, and Neuropathic Pain in Primary Care.” Canadian Family Physician 68: 179–190. [DOI] [PMC free article] [PubMed] [Google Scholar]
46. Langford, D. J. , Baron R., Edwards R. R., et al. 2023. “What Should Be the Entry Pain Intensity Criteria for Chronic Pain Clinical Trials? An IMMPACT Update.” Pain 164: 1927–1930. [DOI] [PMC free article] [PubMed] [Google Scholar]
47. Leeden, R. , Meijer E., and Busing F. M.. 2008. “Resampling multilevel models.” In Handbook of Multilevel Analysis, 401–433. Springer. [Google Scholar]
48. Leemans, L. , Polli A., Nijs J., Wideman T., den Bandt H., and Beckwee D.. 2022. “It Hurts to Move! Intervention Effects and Assessment Methods for Movement‐Evoked Pain in Patients With Musculoskeletal Pain: A Systematic Review With Meta‐Analysis.” Journal of Orthopaedic and Sports Physical Therapy 52: 345–374. [DOI] [PubMed] [Google Scholar]
49. Lesnak, J. B. , Fahrion A., Helton A., et al. 2022. “Resistance Training Protects Against Muscle Pain Through Activation of Androgen Receptors in Male and Female Mice.” Pain 163: 1879–1891. [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Mantyselka, P. T. , Turunen J. H., Ahonen R. S., and Kumpusalo E. A.. 2003. “Chronic Pain and Poor Self‐Rated Health.” JAMA 290: 2435–2442. [DOI] [PubMed] [Google Scholar]
51. Mataran‐Penarrocha, G. A. , Lara Palomo I. C., Antequera Soler E., et al. 2020. “Comparison of Efficacy of a Supervised Versus Non‐Supervised Physical Therapy Exercise Program on the Pain, Functionality and Quality of Life of Patients With Non‐Specific Chronic Low‐Back Pain: A Randomized Controlled Trial.” Clinical Rehabilitation 34: 948–959. [DOI] [PubMed] [Google Scholar]
52. McCrae, R. R. , and John O. P.. 1992. “An Introduction to the Five‐Factor Model and its Applications.” Journal of Personality 60, no. 2: 175–215. [DOI] [PubMed] [Google Scholar]
53. Meyerowitz‐Katz, G. , Ravi S., Arnolda L., Feng X., Maberly G., and Astell‐Burt T.. 2020. “Rates of Attrition and Dropout in App‐Based Interventions for Chronic Disease: Systematic Review and Meta‐Analysis.” Journal of Medical Internet Research 22: e20283. [DOI] [PMC free article] [PubMed] [Google Scholar]
54. Miller, C. T. , Owen P. J., Than C. A., et al. 2022. “Attempting to Separate Placebo Effects From Exercise in Chronic Pain: A Systematic Review and Meta‐Analysis.” Sports Medicine 52: 789–816. [DOI] [PubMed] [Google Scholar]
55. Novy, D. M. , Nelson D. V., Goodwin J., and Rowzee R. D.. 1993. “Psychometric Comparability of the State‐Trait Anxiety Inventory for Different Ethnic Subpopulations.” Psychological Assessment 5: 343–349. [Google Scholar]
56. Ott, J. , Champagne S. N., Bachani A. M., and Morgan R.. 2022. “Scoping sex and gender in Rehabilitation: (Mis)representations and Effects.” International Journal for Equity in Health 21: 179. [DOI] [PMC free article] [PubMed] [Google Scholar]
57. Ozden, F. , Sari Z., Karaman O. N., and Aydogmus H.. 2022. “The Effect of Video Exercise‐Based Telerehabilitation on Clinical Outcomes, Expectation, Satisfaction, and Motivation in Patients With Chronic Low Back Pain.” Irish Journal of Medical Science 191: 1229–1239. [DOI] [PubMed] [Google Scholar]
58. Peyrusque, E. , Granet J., Pageaux B., Buckinx F., and Aubertin‐Leheudre M.. 2022. “Assessing Physical Performance in Older Adults During Isolation or Lockdown Periods: Web‐Based Video Conferencing as a Solution.” Journal of Nutrition, Health & Aging 26: 52–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
59. Pocovi, N. C. , Lin C. C., French S. D., et al. 2024. “Effectiveness and Cost‐Effectiveness of an Individualised, Progressive Walking and Education Intervention for the Prevention of Low Back Pain Recurrence in Australia (WalkBack): A Randomised Controlled Trial.” Lancet 404: 134–144. [DOI] [PubMed] [Google Scholar]
60. Pratap, A. , Neto E. C., Snyder P., et al. 2020. “Indicators of Retention in Remote Digital Health Studies: A Cross‐Study Evaluation of 100,000 Participants.” npj Digital Medicine 3: 21. [DOI] [PMC free article] [PubMed] [Google Scholar]
61. Qaseem, A. , Wilt T. J., McLean R. M., and Forciea M. A.. 2017. “Noninvasive Treatments for Acute, Subacute, and Chronic Low Back Pain: A Clinical Practice Guideline From the American College of Physicians.” Annals of Internal Medicine 166: 514–530. [DOI] [PubMed] [Google Scholar]
62. Raiszadeh, K. , Tapicer J., Taitano L., Wu J., and Shahidi B.. 2021. “In‐Clinic Versus Web‐Based Multidisciplinary Exercise‐Based Rehabilitation for Treatment of Low Back Pain: Prospective Clinical Trial in an Integrated Practice Unit Model.” Journal of Medical Internet Research 23: e22548. [DOI] [PMC free article] [PubMed] [Google Scholar]
63. Roren, A. , Daste C., Coleman M., et al. 2023. “Physical Activity and Low Back Pain: A Critical Narrative Review.” Annals of Physical and Rehabilitation Medicine 66: 101650. [DOI] [PubMed] [Google Scholar]
64. Schafer, J. L. , and Graham J. W.. 2002. “Missing Data: Our View of the State of the Art.” Psychological Methods 7: 147–177. [PubMed] [Google Scholar]
65. Shi, W. , Zhang Y., Bian Y., et al. 2024. “The Physical and Psychological Effects of Telerehabilitation‐Based Exercise for Patients With Nonspecific Low Back Pain: Prospective Randomized Controlled Trial.” JMIR mHealth and uHealth 12: e56580. [DOI] [PMC free article] [PubMed] [Google Scholar]
66. Shiri, R. , and Falah‐Hassani K.. 2017. “Does Leisure Time Physical Activity Protect Against Low Back Pain? Systematic Review and Meta‐Analysis of 36 Prospective Cohort Studies.” British Journal of Sports Medicine 51: 1410–1418. [DOI] [PubMed] [Google Scholar]
67. Simmich, J. , Ross M. H., and Russell T.. 2024. “Real‐Time Video Telerehabilitation Shows Comparable Satisfaction and Similar or Better Attendance and Adherence Compared With In‐Person Physiotherapy: A Systematic Review.” Journal of Physiotherapy 70: 181–192. [DOI] [PubMed] [Google Scholar]
68. Sivertsson, J. , Sernert N., and Ahlund K.. 2024. “Exercise‐Based Telerehabilitation in Chronic Low Back Pain ‐ a Scoping Review.” BMC Musculoskeletal Disorders 25: 948. [DOI] [PMC free article] [PubMed] [Google Scholar]
69. Slade, S. C. , Dionne C. E., Underwood M., and Buchbinder R.. 2016. “Consensus on Exercise Reporting Template (CERT): Explanation and Elaboration Statement.” British Journal of Sports Medicine 50: 1428–1437. [DOI] [PubMed] [Google Scholar]
70. Smith, J. A. , Stabbert H., Bagwell J. J., Teng H. L., Wade V., and Lee S. P.. 2022. “Do People With Low Back Pain Walk Differently? A Systematic Review and Meta‐Analysis.” Journal of Sport and Health Science 11: 450–465. [DOI] [PMC free article] [PubMed] [Google Scholar]
71. Spielberger, C. D. , Gonzalez‐Reigosa F., Martinez‐Urrutia A., Natalicio L. F., and Natalicio D. S.. 1971. “The State‐Trait Anxiety Inventory.” Revista Interamericana de Psicología/Interamerican Journal of Psychology 5, no. 3–4: 145–158. [Google Scholar]
72. Steenstra, I. A. , Knol D. L., Bongers P. M., Anema J. R., van Mechelen W., and de Vet H. C.. 2009. “What Works Best for Whom? An Exploratory, Subgroup Analysis in a Randomized, Controlled Trial on the Effectiveness of a Workplace Intervention in Low Back Pain Patients on Return to Work.” Spine 34: 1243–1249. [DOI] [PubMed] [Google Scholar]
73. Studenski, S. 2019. “Gait Speed Reveals Clues to Lifelong Health.” JAMA Network Open 2: e1913112. [DOI] [PubMed] [Google Scholar]
74. Sullivan, M. , Bishop S., and Pivik J.. 1995. “The Pain Catastrophizing Scale: Development and Validation.” Psychological Assessment 7: 524–532. [Google Scholar]
75. Sullivan, M. J. , Thorn B., Haythornthwaite J. A., et al. 2001. “Theoretical Perspectives on the Relation Between Catastrophizing and Pain.” Clinical Journal of Pain 17: 52–64. [DOI] [PubMed] [Google Scholar]
76. Tabacof, L. , Baker T. S., Durbin J. R., et al. 2022. “Telehealth Treatment for Nonspecific Low Back Pain: A Review of the Current State in Mobile Health.” PM & R: The Journal of Injury, Function, and Rehabilitation 14: 1086–1098. [DOI] [PubMed] [Google Scholar]
77. Takai, Y. , Ohta M., Akagi R., Kanehisa H., Kawakami Y., and Fukunaga T.. 2009. “Sit‐To‐Stand Test to Evaluate Knee Extensor Muscle Size and Strength in the Elderly: A Novel Approach.” Journal of Physiological Anthropology 28: 123–128. [DOI] [PubMed] [Google Scholar]
78. Tankha, H. , Gaskins D., Shallcross A., et al. 2024. “Effectiveness of Virtual Yoga for Chronic Low Back Pain: A Randomized Clinical Trial.” JAMA Network Open 7: e2442339. [DOI] [PMC free article] [PubMed] [Google Scholar]
79. Tkachuk, G. A. , and Harris C. A.. 2012. “Psychometric Properties of the Tampa Scale for Kinesiophobia‐11 (TSK‐11).” Journal of Pain 13: 970–977. [DOI] [PubMed] [Google Scholar]
80. van der Roer, N. , Ostelo R. W., Bekkering G. E., van Tulder M. W., and de Vet H. C.. 2006. “Minimal Clinically Important Change for Pain Intensity, Functional Status, and General Health Status in Patients With Nonspecific Low Back Pain.” Spine 31: 578–582. [DOI] [PubMed] [Google Scholar]
81. Vink, G. , Lazendic G., and van Buuren S.. 2015. “Partitioned Predictive Mean Matching as a Multilevel Imputation Technique.” Psychological Test and Assessment Modeling 57: 577. [Google Scholar]
82. Vlaeyen, J. W. S. , Kole‐Snijders A. M. J., Boeren R. G. B., and van Eek H.. 1995. “Fear of Movement/(Re)injury in Chronic Low Back Pain and Its Relation to Behavioral Performance.” Pain 62: 363–372. [DOI] [PubMed] [Google Scholar]
83. Wallace, L. M. , Falla D., Rushton A., and Heneghan N. R.. 2022. “Group and Individual Telehealth for Chronic Musculoskeletal Pain: A Scoping Review.” Musculoskeletal Care 20: 245–258. [DOI] [PubMed] [Google Scholar]
84. Warburton, D. E. , Jamnik V. K., Bredin S. S., and Gledhill N.. 2011. “The Physical Activity Readiness Questionnaire for Everyone (PAR‐Q+) and Electronic Physical Activity Readiness Medical Examination (ePARmed‐X+).” Health & Fitness Journal of Canada 4: 3–17. [Google Scholar]
85. Wilson, K. E. , and Dishman R. K.. 2015. “Personality and Physical Activity: A Systematic Review and Meta‐Analysis.” Personality and Individual Differences 72: 230–242. [Google Scholar]
86. Wood, L. , Foster N. E., Dean S. G., Booth V., Hayden J. A., and Booth A.. 2024. “Contexts, Behavioural Mechanisms and Outcomes to Optimise Therapeutic Exercise Prescription for Persistent Low Back Pain: A Realist Review.” British Journal of Sports Medicine 58: 222–230. [DOI] [PubMed] [Google Scholar]
87. Wun, A. , Kollias P., Jeong H., et al. 2021. “Why Is Exercise Prescribed for People With Chronic Low Back Pain? A Review of the Mechanisms of Benefit Proposed by Clinical Trialists.” Musculoskeletal Science & Practice 51: 102307. [DOI] [PubMed] [Google Scholar]
88. Yue, Q. , Jiang H.‐M., Yun D.‐T., and Liu J.. n.d. “Why Do Exercisers With Higher Trait Conscientiousness or Openness Undergo Positive Exercise Experience? The Mediating Effects of Outcome Expectancy.” Social Science Research. 10.2139/ssrn.4791409. [DOI] [Google Scholar]

Supplementary Materials

Data Availability Statement

The data analysis script is available at https://github.com/GeversMontoro/clbpeo2025/. The deidentified data collected in the experiment will be available upon request. Data and code underpinning this secondary analysis are available from the corresponding author upon reasonable request.