In a cross-sectional analysis of 936 U.S. adults paired with single-cell RNA sequencing in exercised mice, higher physical activity levels associated with significantly lower DNAm-predicted PhenoAge (β = −0.014) and reduced circulating β2-microglobulin (β2M) levels (β = −0.006). Mediation modeling revealed β2M accounts for 37.67% of the total association between physical activity and epigenetic aging deceleration. Mouse scRNA-seq localized these effects primarily to B cells and myeloid cells, implicating immune, inflammatory, mitochondrial, and circadian pathway remodeling as downstream mechanisms.
This finding meaningfully advances the mechanistic picture of exercise-mediated longevity. β2M has previously been identified as a pro-aging blood factor — elevated in aged plasma and causally linked to cognitive decline in parabiosis experiments — making it a plausible effector rather than mere bystander. That exercise suppresses it enough to explain over a third of its epigenetic age benefit is a quantitatively striking result. The subgroup pattern — stronger effects in males and higher-BMI individuals — suggests β2M reduction may be especially relevant where baseline inflammaging is elevated, potentially pointing toward a therapeutic window. Limitations are notable: the human arm is observational with modest n=936, cross-sectional design precludes causality, and self-reported activity carries measurement error. Mouse-to-human scRNA-seq translation remains inferential. Still, the multi-modal design combining population epidemiology with cellular transcriptomics elevates this above typical associative work — an incrementally important, mechanistically grounded contribution to exercise-longevity biology.
