Exercise programmes failed to get people with chronic pain moving more

Physical activity is a cornerstone of managing chronic musculoskeletal pain, from osteoarthritis to low back pain, and countless programmes exist to encourage it. But there is a gap between joining a programme and actually moving more through the day, and most evaluations rely on people reporting their own activity, which is notoriously generous. Wearable devices such as accelerometers and pedometers remove that flattery. This review asked a bracingly simple question: when movement is measured objectively, do physical activity interventions actually increase how much people with chronic pain move, compared with little or no intervention?

The authors systematically searched five databases plus the main clinical trial registries for randomised or quasirandomised controlled trials in people with chronic musculoskeletal pain. Eligible trials had to measure physical activity objectively, with devices such as accelerometers or pedometers, and compare an activity intervention against no or minimal intervention. Analyses were run separately for short-term follow-up of up to 3 months, intermediate follow-up between 3 and 12 months, and long-term follow-up of 12 months or more. Eight published trials were included, and the team also logged 6 registered trials whose results were not yet available.

Pooling 6 trials at short-term follow-up showed no significant effect of the interventions on objectively measured activity: the average difference was 0.34 standardised units, with a range of plausible values that crossed zero. The intermediate and long-term analyses came out similarly non-significant. Grading the whole body of evidence, the authors rated it as low quality. Their conclusion was correspondingly careful: physical activity interventions may lead to little or no difference in how much people with chronic musculoskeletal pain actually move, and the pooled estimate is likely to change once the 6 registered trials report.

This finding complicates a comfortable assumption in pain management: that prescribing or delivering an activity programme translates into more movement. Trials showing improvements in pain or self-reported function do not necessarily mean people moved more, and this review suggests the behavioural change itself may be the hard part. Because the evidence base was small and rated low quality, the result is provisional rather than final, and the authors explicitly flag that pending trials could shift the pooled effect in either direction. It is a prompt for intervention designers as much as a verdict on exercise.

If you live with chronic joint or back pain and have been through an activity programme, this is a reason to judge it by what your feet and wrists record rather than by enrolment or good intentions. It is emphatically not evidence that movement fails to help pain; the question here was narrower, whether programmes change the amount people move, and the answer so far is barely. Worth knowing, too, that this is a live area: more trials were registered than published when this review was done, so the picture may sharpen.