The study focused on five aging indicators that gerontologists consider especially informative. Telomere length and telomerase activity capture the integrity of the protective caps on chromosomes, which erode as cells divide and are widely studied as hallmarks of cellular aging. Global DNA methylation reflects the epigenetic chemical tags that accumulate and shift across the genome over a lifetime. Appendicular lean mass, the muscle mass of the arms and legs, tracks the physical frailty that accompanies aging, while the frailty index aggregates deficits in health and function into a single vulnerability score. Because these markers span the spectrum from chromosomes to whole-body function, any consistent effect of exercise across them would carry considerable weight.

To quantify that effect, the researchers conducted a meta-analysis of 32 studies encompassing 49 randomized controlled trials and 5,189 participants. Randomized trials are the gold standard for causal inference because participants are assigned by chance to exercise or control conditions, minimizing the confounding that plagues observational studies. The team harmonized the volume of physical activity across trials using metabolic equivalent tasks, or METs, a standard unit that expresses the energy cost of an activity relative to resting metabolism. This harmonization allowed them to place trials with wildly different exercise prescriptions, from walking programs to structured resistance training, onto a common dose scale.

The pooled results were consistent in direction. Physical activity significantly increased telomere length, with a standardized mean difference of 0.18, and produced a larger boost to telomerase activity, the enzyme that rebuilds telomeres, with a standardized mean difference of 0.46. Exercise also increased appendicular lean mass with a standardized mean difference of 0.29 and reduced the frailty index with a standardized mean difference of minus 0.13. In practical terms, the effect sizes are modest but meaningful: they suggest that regular activity nudges several independent aging systems in the direction of youthfulness rather than decline, an effect visible across molecular, cellular, and whole-body measures simultaneously.

Perhaps the most striking finding came from the dose-response analysis. Using restricted cubic spline regression, a flexible statistical technique that can model curved rather than straight-line relationships, the researchers mapped how aging indicators respond as physical activity volume increases. The curves for different markers followed a similar shape, rising steeply at low activity levels before flattening out. The estimated peak benefit occurred at approximately 11.5 marginal MET-hours per week, with confidence intervals of roughly 10.9 to 12.7 for telomere length and 12.1 to 19.7 for appendicular lean mass. In everyday terms, that dose corresponds to a moderate but sustainable routine, well within reach of most adults, and the flattening of the curves suggests that piling on far more exercise yields diminishing returns for these aging markers.

Meta-analysis of randomized trials is powerful, but the researchers went further by asking whether genetics supports a causal link. They turned to Mendelian randomization, a technique that exploits the random allocation of gene variants at conception as a natural experiment. Because genetic variants that predispose people to higher levels of moderate-to-vigorous physical activity are inherited independently of lifestyle, socioeconomic factors, and disease, they can serve as proxies for a lifetime exposure to exercise. If people genetically predisposed to be more active also show younger biological aging profiles, that pattern is difficult to explain by reverse causation or confounding, the two great weaknesses of observational research.

The Mendelian randomization analysis drew on genome-wide association studies and found genetic support for causal effects of moderate-to-vigorous physical activity on leukocyte telomere length, appendicular lean mass, the frailty index, and four epigenetic clocks, including the Hannum clock, PhenoAge, GrimAge, and intrinsic epigenetic age acceleration. Epigenetic clocks are mathematical models that estimate biological age from patterns of DNA methylation, and they have become among the most closely watched tools in geroscience because they can, in principle, measure the pace of aging in a single blood sample. The convergence between the trial evidence and the genetic evidence is what gives the new study its force: two independent lines of investigation point in the same direction.

To probe the mechanisms behind these effects, the team performed a mediation analysis integrating multi-omics data from proteomics, metabolomics, and immunomics. This approach searches for molecules whose genetically predicted levels lie on the causal pathway between physical activity and aging outcomes. The analysis identified genetically inferred mediators including ELANE, a neutrophil elastase enzyme, and NEU1, neuraminidase 1, pointing toward inflammatory and glycoprotein-processing pathways as potential conduits through which exercise influences aging biology. Identifying such molecular intermediaries matters because they represent candidate targets for future interventions that could mimic the anti-aging benefits of exercise in people unable to be physically active.

The researchers also examined whether genetically predicted physical activity influences the biological age gap, the difference between biological and chronological age, across nine major human organs. The analysis linked predicted moderate-to-vigorous activity to mitigation of the pulmonary biological age gap, suggesting that the lungs may be among the organs most responsive to exercise at the molecular level. Organ-specific aging is an emerging frontier in the field, as different tissues appear to age at different rates within the same person, and pinpointing which organs benefit most from a given intervention could eventually allow more personalized prescriptions.

The authors are careful to frame their findings as evidence-based guidance rather than a mandate for extreme training. The estimated peak intensity of roughly 11.5 marginal MET-hours per week offers a concrete benchmark for clinicians and public health authorities, and the flattening dose-response curves reinforce a message that has emerged repeatedly in exercise science: consistency at moderate doses appears to capture most of the available benefit. By triangulating randomized trial data, genetic causal inference, and multi-omics mediation analysis, the study provides an unusually robust case that physical activity does not merely correlate with slower aging but plausibly drives it, while opening a window onto the molecular machinery, from telomeres to inflammatory enzymes, through which a daily walk or a session of resistance training writes itself into our biology.

Subject of Research: The effects of physical activity volume on biological aging indicators including telomere length, epigenetic clocks, lean mass and frailty, assessed by meta-analysis and Mendelian randomization.

Article Title: Physical activity volume and aging indicators: a meta-analysis and genetic causality assessment

Article References: Liu, J., Wen, J., Ji, N., Zhang, J., Liu, H., Liu, Z., Liu, Z., Xia, Z., Luo, P., & Cheng, Q. (2026). Physical activity volume and aging indicators: a meta-analysis and genetic causality assessment. BMC Medicine. https://doi.org/10.1186/s12916-026-05153-8

Image Credits: AI Generated

DOI: 10.1186/s12916-026-05153-8

Keywords: physical activity, aging, telomere length, telomerase activity, epigenetic clocks, Mendelian randomization, meta-analysis, frailty index, appendicular lean mass, dose-response, multi-omics, biological age gap