Exercise only slightly lowers inflammation markers, and the evidence is shaky

Ageing bodies tend to run a constant, low-level inflammatory hum, and the immune system itself grows less responsive over time. Both processes are linked to chronic disease, so the idea that exercise could dial them down is attractive and widely repeated. Individual reviews have examined the question in different groups, from older adults to people with long-term conditions, but their findings were scattered and sometimes contradictory. This umbrella review set out to gather all of that review-level evidence in one place and judge how strong it really is.

The authors searched six major databases up to 1 March 2026 for systematic reviews and meta-analyses covering exercise or physical activity and circulating markers of inflammation and immune ageing in older adults and adults with chronic diseases. Twenty-one reviews made the cut: 14 matched the question fully and drove the conclusions, while seven were used only for background mapping. Each review's methodological quality was formally appraised and the certainty of the evidence was graded. Importantly, the team ran no new pooled calculations of their own; this was a structured narrative reading of what the existing reviews reported.

Exercise was associated with small reductions in TNF-alpha, with point estimates ranging from minus 1.42 to minus 0.05, though only two of five estimates clearly excluded no effect at all. C-reactive protein showed a similar pattern, with estimates from minus 0.65 to minus 0.47 and two of three excluding no effect. Interleukin-6 was inconsistent, spanning minus 0.81 to plus 0.09. The underlying studies disagreed with each other substantially, 11 of the reviews were rated critically low quality and nine low, and certainty was judged very low for every marker. When the weakest reviews were set aside, the apparent signal weakened further.

Exercise is routinely described as anti-inflammatory, and this review neither demolishes nor confirms that framing: the direction of effect generally favours exercise, but the evidence quality is too poor to lean on. It also exposes a genuine gap, since markers of immune cell ageing and cellular senescence have barely been studied at review level. The authors are explicit that this biomarker evidence alone cannot justify prescribing exercise specifically to lower inflammatory markers. Crucially, they also note it does not change physical activity recommendations built on clinical and functional outcomes, which rest on far firmer ground.

If you exercise partly because you have read it fights inflammation, the honest position is that the blood-marker evidence is thin and uncertain, even if it leans the right way. That is not a reason to move less: the established benefits of activity for strength, fitness, mood and long-term health do not depend on these markers. It is, however, a reason to be sceptical of programmes or products sold on the promise of taming your inflammatory markers through specific exercise protocols. Better designed reviews may yet firm the picture up.