Researchers at the Harvard T.H. Chan School of Public Health linked rest-activity patterns captured by commercial wearables to biological aging trajectories, according to a peer-reviewed study published in Nature Communications. The investigation evaluated multi-year Fitbit tracking data alongside clinical biomarker-derived PhenoAge measurements from 2,222 participants, covering a total of 8,447 person-years drawn from the All of Us Research Program.
Jinjoo Shim and Jukka-Pekka Onnela of the university's Department of Biostatistics applied high-dimensional digital phenotyping to assess circadian rest-activity rhythms across three dimensions: intensity, timing, and stability. The data revealed that individuals with higher rhythm intensity experienced 26% to 46% lower odds of accelerated biological aging.
Biological aging trajectories also reflected sex-specific differences across the tracked activity metrics. Associations involving the timing and regularity of rest-activity rhythms proved stronger in female participants. Among male participants, accelerated aging was characterized by a biphasic instability pattern that featured activity surges in the early morning followed by rebounds in the late evening.
The authors noted that disrupted rest-activity rhythms have previously been associated with aging and chronic disease, but longitudinal evidence from free-living populations had remained lacking. Linking population-scale digital phenotyping to PhenoAge biological aging data demonstrates the potential of wearable-derived digital biomarkers for aging-related risk assessments and future healthy-aging research.
The peer-reviewed paper was submitted on September 30, 2025, accepted on July 21, 2026, and published on August 22, 2026. The authors declared no competing interests, and the research project received no funding.
