Effects of exercise and physical activity on circulating inflammaging- and immunosenescence-related biomarkers in older adults and adults with chronic diseases: an umbrella review

Background Chronic low-grade inflammation and immunosenescence accompany ageing and chronic disease. Exercise and physical activity are proposed to modify circulating inflammatory biomarkers, but review-level evidence is dispersed across populations and conditions. Methods We searched PubMed, Web of Science, Embase, Cochrane Library, Scopus, and SPORTDiscus through 1 March 2026 for systematic reviews, meta-analyses, and overviews on exercise or physical activity and circulating inflammaging- or immunosenescence-related biomarkers in older adults and adults with chronic diseases. We reported the review in accordance with PRISMA 2020 and PRIOR. Methodological quality was appraised with A MeaSurement Tool to Assess systematic Reviews, version 2 (AMSTAR-2), primary-study overlap with the corrected covered area, and certainty with GRADE. We performed no de novo meta-analysis or meta-regression; this umbrella review is a structured narrative synthesis of quantitative review-level estimates rather than a quantitative reanalysis. Results Twenty-one reviews were included: 14 core-eligible (full Population, Intervention, Comparator, Outcomes, and Study design [PICOS] match, contributing to the primary synthesis and conclusions) and seven evidence-map-only (partial PICOS match, used only for mapping and background). Among core-eligible reviews reporting standardised mean differences, exercise was associated with small reductions in tumour necrosis factor-alpha (TNF-alpha; point estimates − 1.42 to − 0.05, two of five confidence intervals excluding the null) and C-reactive protein (CRP; − 0.65 to − 0.47, two of three excluding the null), whereas interleukin-6 (IL-6) estimates were inconsistent (− 0.81 to + 0.09, two of five excluding the null). Heterogeneity was frequently high ( I 2 up to 94%). Overall primary-study overlap was slight (corrected covered area [CCA] = 0.0091), but very high within the multiple sclerosis cluster (CCA = 0.4048). AMSTAR-2 confidence was critically low in 11 reviews, low in nine, and moderate in one; GRADE certainty was very low for every biomarker construct. A post hoc restricted descriptive sensitivity analysis excluding critically low reviews attenuated the apparent signal. Conclusion Exercise and physical activity may produce small reductions in some circulating inflammatory biomarkers, most consistently TNF-alpha and CRP, but effects on IL-6 are inconsistent, and certainty is very low. Effects on immunosenescence and cellular-senescence markers remain an evidence gap. This biomarker evidence alone is insufficient to justify prescribing exercise specifically to lower inflammatory biomarkers and does not alter broader physical-activity recommendations based on clinical and functional outcomes. PROSPERO : CRD420261410061