Sleep Temporal Entropy as a Novel Digital Biomarker of Sleep Fragmentation for Cardiometabolic and Mortality Risk

This study introduces Sleep Temporal Entropy (STE), a novel entropy-based digital sleep biomarker derived from Shannon entropy theory, to quantify sleep fragmentation and explore its associations with cardiometabolic disorders and mortality. Unlike traditional metrics (e.g., Sleep Efficiency, Wake After Sleep Onset), STE captures the complexity of transitions across both sleep-wake and sleep-stage boundaries, providing a multidimensional analysis of sleep architecture. Based on two distinct cohorts--a clinic-based cross-sectional population and a community-based longitudinal population--the biomarker demonstrated consistent performance across multiple outcomes. In the Shanghai Sleep Health Study Cohort (SSHSC, n=3,219), STE consistently outperformed established metrics in predicting hypertension, diabetes, and hyperlipidemia. In the Sleep Heart Health Study (SHHS, n=4,862), STE demonstrated robust U-shaped associations with both all-cause and cardiovascular mortality, with REM STE emerging as the most significant and independent predictor. Key innovations of this study include the application of STE in cross-cohort and multi-outcome validations, direct performance comparisons with existing metrics, and the identification of nonlinear health impacts. These findings contribute to advancements in sleep biomarker research, suggesting that STE could provide valuable insights for clinical and research applications in sleep health and cardiometabolic risk assessment.