Exercise lowered inflammation, but volume wasn't the lever

Low-grade systemic inflammation is a common thread linking excess weight, the postmenopausal transition and cardiometabolic risk, and exercise is one of the few levers known to influence it. What has been much less clear is which part of an exercise programme does the work: the total amount accumulated each week, how many weeks it runs for, or how hard the sessions are. Guidelines have historically leaned on weekly volume targets. Testing volume, duration and intensity against inflammatory markers in one population makes it possible to see which parameter actually tracks the response.

Five databases were searched for randomised controlled trials up to January 2026, with the review restricted exclusively to overweight and obese postmenopausal women and pre-registered on PROSPERO. Thirty trials with a combined 2,124 participants were pooled using random-effects models, with standardised mean differences calculated for C-reactive protein, interleukin-6, TNF-alpha and adiponectin. Meta-regression then tested dose-response relationships for total exercise volume in MET-minutes per week, intervention duration in weeks, and exercise intensity standardised to a percentage of maximum heart rate. Leave-one-out sensitivity analyses checked how much any single trial drove the pooled estimates.

Exercise significantly reduced TNF-alpha, with a standardised mean difference of -0.47, and CRP, with a standardised mean difference of -0.36. Changes in interleukin-6 and adiponectin were not statistically significant. The CRP effect proved highly stable in leave-one-out analysis, ranging only between -0.41 and -0.28 as individual studies were dropped, and the TNF-alpha reduction was similarly robust. Meta-regression found no definitive linear dose-response for total exercise volume or intervention duration on any biomarker. A borderline trend suggested higher intensity was associated with greater TNF-alpha reductions, at p = 0.063.

The two significant reductions add weight to the view that structured exercise measurably lowers inflammatory markers in this group. The absence of any volume or duration gradient complicates the assumption that accumulating more weekly minutes yields proportionally more anti-inflammatory benefit. The authors interpret their intensity trend as pointing to a threshold effect, but a p-value of 0.063 is a hypothesis rather than a finding, and trials designed to compare intensities directly would be needed to test it. Why interleukin-6 and adiponectin did not shift also remains unexplained.

This is a reason to think exercise influences inflammatory markers in overweight and obese postmenopausal women fairly reliably, particularly TNF-alpha and CRP. It is also a reason to be sceptical that simply logging more weekly minutes scales the effect up, since no such gradient appeared here. The hint that harder sessions matter more is genuinely interesting but statistically unresolved, so it should not be read as a prescription for intensity. Marker changes are also a step removed from health outcomes such as disease events.