A new umbrella review gathered 21 earlier reviews to ask whether exercise reliably lowers inflammatory biomarkers in older adults and people living with chronic disease. The answer was a small, very uncertain yes for two markers and a firm no to the idea that inflammatory biomarkers alone justify an exercise prescription.
Inflammatory biomarkers have become the most popular explanation for why exercise appears to protect aging bodies, and an umbrella review published in Frontiers in Immunology on September 29, 2026 set out to test how well that explanation actually holds together.
The review was written by Guangwei Chai and Jinhong Xue of Hebei Sport University in Shijiazhuang, China, with Bin Xie of Hebei Chemical and Pharmaceutical College. Instead of pooling individual trials, the team collected the systematic reviews and meta-analyses that had already done that pooling, then graded how trustworthy each of those reviews was.
Searches ran across PubMed, Web of Science, Embase, the Cochrane Library, Scopus and SPORTDiscus through March 1, 2026. Twenty-one reviews met the criteria. Fourteen matched the full population, intervention, comparator, outcomes and study-design framework and fed the primary synthesis. Seven were used only to map the surrounding evidence.
What came back was modest. Exercise was associated with small reductions in tumor necrosis factor alpha and C-reactive protein, while interleukin 6 moved inconsistently. Certainty, graded using GRADE, was very low for every biomarker the authors examined.
At a Glance
- Published September 29, 2026 in Frontiers in Immunology, volume 17, as an umbrella review.
- Twenty-one reviews were included: 14 core-eligible and seven used only for evidence mapping.
- Tumor necrosis factor alpha point estimates ranged from minus 1.42 to minus 0.05, with two of five confidence intervals excluding the null.
- C-reactive protein estimates ran from minus 0.65 to minus 0.47, with two of three confidence intervals excluding the null.
- Interleukin 6 estimates spanned minus 0.81 to plus 0.09, with two of five excluding the null.
- Heterogeneity between studies reached an I2 value of up to 94 percent.
- GRADE certainty was very low for every one of the inflammatory biomarkers assessed.
What an Umbrella Review of Inflammatory Biomarkers Can Show
An umbrella review sits one level above a meta-analysis. Rather than gathering trials, it gathers reviews, then asks how consistent those reviews are with one another and how carefully each was conducted. The design earns its place when a question has been answered many times and the answers keep disagreeing.
That description fits the exercise and inflammation literature closely. Dozens of reviews have reported that training lowers inflammatory biomarkers across different populations, different exercise doses and different comparison groups, and the pattern has been treated in fitness coverage as more or less settled.
The authors deliberately avoided running a fresh pooled analysis. They describe the work as a structured narrative synthesis of quantitative review-level estimates rather than a quantitative reanalysis, and they reported it in line with the PRISMA 2020 and PRIOR guidance for overviews.
Where the Inflammatory Biomarkers Actually Moved
Among the core-eligible reviews reporting standardized mean differences, tumor necrosis factor alpha showed point estimates running from minus 1.42 to minus 0.05. Two of five confidence intervals excluded the null, which also means three of them did not.
C-reactive protein looked steadier. Point estimates ran from minus 0.65 to minus 0.47, and two of three confidence intervals excluded the null. Both inflammatory biomarkers drifted in the direction exercise researchers expect, and both drifted by small amounts.
Interleukin 6 was the outlier. Estimates spanned minus 0.81 to plus 0.09, with two of five confidence intervals excluding the null. In plain language, some reviews found a drop, others found nothing at all, and no single direction of travel emerged from the set.
Why Certainty Stayed Very Low
Methodological confidence was appraised using AMSTAR-2, a standard checklist for judging how well a systematic review was built. Eleven of the included reviews were rated critically low, nine were low, and exactly one reached moderate. None was rated high.
GRADE certainty came out very low for every one of the inflammatory biomarkers examined. That rating is not a formality. It means the true effect could plausibly sit some distance from the reported estimate, in either direction, once better studies arrive.
A post hoc restricted descriptive sensitivity analysis that excluded the critically low reviews attenuated the apparent signal. When the weakest evidence was set aside, the picture became fainter rather than clearer, which is the opposite of what a robust finding usually does.
The highest I2 heterogeneity value recorded across the pooled estimates, meaning the included studies often disagreed with each other far more than chance alone would explain.
Overlap, and Why It Matters Here
Umbrella reviews of inflammatory biomarkers carry a specific hazard. If several of the included reviews analyze the same underlying trials, the same data gets counted repeatedly and a weak result can look like independent replication.
The authors measured this with the corrected covered area. Overall overlap was slight, at 0.0091. Inside the multiple sclerosis cluster, however, it reached 0.4048, which is very high. Readers of that subgroup are largely reading the same primary studies described several times over.
Publication bias was another gap. It was often unassessed in the original reviews and could not be reevaluated at review level, so the possibility that null results never reached print remains open.
What Inflammatory Biomarkers Do Not Settle
The conclusion the authors reached is unusually direct for a review of this kind. They write that this biomarker evidence alone is insufficient to justify prescribing exercise specifically to lower inflammatory biomarkers, and that it does not alter broader physical-activity recommendations based on clinical and functional outcomes.
That distinction is worth holding onto. Exercise still has a large and well-replicated evidence base built on outcomes people can feel and clinicians can count, including mortality, disease incidence, strength, mobility and independence. None of that rests on a blood panel.
The review also flags a genuine gap. Effects on immunosenescence and cellular-senescence markers, the aging of the immune system itself rather than its background noise, remain largely unstudied. The inflammatory biomarkers that dominate this literature are the easy ones to measure, not necessarily the important ones.
How This Fits With Other Exercise Research
The pattern here rhymes with other biomarker work covered on this site. A 2025 pooled analysis of physical activity and oxidative stress markers found similarly modest and variable movement across a large pooled sample, with the same caution about study quality.
Where exercise research has been strongest is on outcomes rather than intermediates. Work on structured exercise and survival in people with cancer and on cardiorespiratory fitness and long-term disease risk both rest on hard endpoints, and both produced far larger and more consistent effects than anything in this biomarker set.
For older adults specifically, the practical literature has moved toward very short, repeatable bouts. Reviews of exercise snacking for strength and mobility focus on function rather than inflammation, which may be the more useful frame for anyone deciding how to spend a training week.
What This Means for Your Own Training
Nothing in this review argues for doing less. It argues against a specific claim: that you can judge whether your training is working by watching inflammatory biomarkers on a blood test.
A few things follow for general readers, none of them individualized advice:
- Treat a single CRP or interleukin 6 reading as weak feedback on a training program, not a verdict on it.
- Judge progress by outcomes you can observe, such as walking speed, stair tolerance, strength, sleep and how you feel on the third day of a busy week.
- Keep meeting general activity guidance, which is built on clinical endpoints rather than on inflammatory biomarkers.
- If a clinician has ordered inflammatory markers for a diagnosed condition, interpret those results with that clinician rather than against fitness literature.
Anyone with a chronic condition should also remember that most of the reviews here studied supervised programs in clinical populations, which is a different setting from a commercial gym and a different starting point from most readers.
What Would Settle It
The authors are explicit about the shape of the missing study. It would be a large, well-powered trial with a prespecified primary biomarker, a consistent comparator, standardized assay methods and enough follow-up to see whether early changes in inflammatory biomarkers persist.
Until something like that exists, the honest summary is the one the review offers. Exercise and physical activity may produce small reductions in some circulating inflammatory biomarkers, most consistently tumor necrosis factor alpha and C-reactive protein, but the effects on interleukin 6 are inconsistent and the certainty is very low.
The Fitness Living Takeaway
An umbrella review of 21 reviews found exercise linked to only small, very-low-certainty reductions in inflammatory biomarkers, with tumor necrosis factor alpha and C-reactive protein the most consistent and interleukin 6 unresolved.
The finding does not weaken the case for exercise, which rests on mortality, disease and function data rather than on blood markers. It weakens the case for using inflammatory biomarkers as the reason to exercise, or as the scoreboard for whether training is working. The authors say so themselves, and they call immunosenescence markers an outright evidence gap.
Research & Sources
- Frontiers in Immunology: Effects of exercise and physical activity on circulating inflammaging- and immunosenescence-related biomarkers
- Study DOI
- Images: George Pagan III / Unsplash; john livingstone / Unsplash
This article summarizes peer-reviewed research for general information and is not individualized medical or exercise advice.
