Wearable-Derived Diurnal Alignment Between Physical Activity and Device Temperature Predicts Future Disease and...

Circadian rhythms coordinate physiology with the 24 h light-dark cycle, and their disruption contributes to diseases spanning metabolic, cardiovascular, and neuropsychiatric domains. Whether the temporal coherence between wearable-derived activity and temperature rhythms predicts long-term health outcomes in free-living humans remains unknown. Here, analyzing week-long concurrent wrist-worn acceleration and device temperature recordings from approximately 90,000 UK Biobank participants (median age 63 years), we decompose the circular cross-correlation between behavioral and device temperature signals into three alignment features, including 24 h coupling strength (M₂₄), phase deviation from expected antiphase (D₂₄), and 12 h harmonic magnitude (M₁₂). Over 7-11 years of prospective follow-up, higher M₂₄ is associated with lower risk of type 2 diabetes, cardiovascular disease, depression, sleep apnea, and all-cause mortality, whereas higher D₂₄ is associated with increased cardiometabolic risk. Higher M₁₂ was associated with a lower risk of gastrointestinal and psychiatric conditions. Technical replication in the SHARE cohort supported the portability of the feature-extraction framework across device protocols. These findings highlight wearable-derived cross-domain diurnal alignment as a scalable, prospective predictor of disease risk, with potential implications for population health surveillance.