Do Physical Therapy and Yoga Improve Pain and Disability through Psychological Mechanisms? A Causal Mediation Analysis of Adults with Chronic Low Back Pain.

The analysis includes only 230 of the 320 people randomised, and those left out because data were missing were more likely to be male, non-White and less well educated, so the sample is not the trial population. Thirty percent of the original sample had dropped out by 52 weeks, with more loss to follow-up in the yoga and physical therapy groups than in the education group, and that difference between groups can bias the comparison. Fewer than half of the participants met the adherence targets set before the study, which the authors say may push the treatment effects toward showing no difference. The psychological measures were taken only at 12 weeks, so the study cannot say whether earlier changes mattered, and the analysis assumes the psychological change came before the pain and disability change. It could be the other way round, or the two could move together. The education control may not have been inactive: the book taught breathing, meditation and flare-up management, which may work through the same psychological routes as physical therapy and yoga, making the comparison a hard one to win. Participants were not selected for having psychological distress, so many started with low fear-avoidance beliefs and only mild anxiety or depression, and the treatments were not designed to change psychological factors. Mediation analysis only holds if nothing unmeasured influences both the mediator and the outcome. The authors checked this for their one significant finding and it held up, but it cannot be proved. The participants were mostly female, 80% non-White, drawn from underserved Boston neighbourhoods with most incomes under $30,000 a year, so the results may not carry to other settings.

The authors state that they have no affiliations with or financial involvement in any organisation or entity with a direct financial interest in the subject matter or materials discussed. No funder is named in the article text.

Do not read this as evidence that treating stress, or adding psychological treatment, is how physical therapy works for back pain. Only one of many mediator pathways reached statistical significance, it was found in a secondary analysis of 230 of the 320 randomised participants, and the interventions were not designed to change psychological factors in the first place.