Promoting healthy sleep patterns for diabetes prevention: joint effects of sleep regularity and sleep duration in a prospective Chinese cohort.

BACKGROUND: Sleep regularity, defined as the consistency of sleep-wake timing across days, is increasingly recognized as an important dimension of sleep health. While inadequate sleep duration is a known risk factor for type 2 diabetes (T2D), the independent contribution of irregular sleep remains poorly understood. Evidence is particularly limited in Asian populations using objective sleep measures. OBJECTIVES: To evaluate the association between accelerometer-derived sleep regularity and incident T2D, and to examine whether meeting recommended sleep duration mitigates the risk associated with irregular sleep patterns. METHODS: We analyzed data from 8,974 insured Chinese adults aged 40-79 years using wearable-linked medical insurance records between August 2020 and June 2023. Raw accelerometry data were collected over 7 consecutive days and processed to compute the Sleep Regularity Index (SRI). Individuals were classified into regular (SRI > 87.3), moderately irregular (71.6-87.3), and irregular (< 71.6) categories. Sleep duration was categorized as adequate or inadequate based on age-specific recommendations. Incident T2D was identified from hospital electronic health records and prescription registries during a median 3.0-year follow-up. Cox proportional hazards models were used to estimate adjusted hazard ratios (HRs) for T2D across SRI categories and in joint exposure models. RESULTS: Compared to individuals with irregular sleep, those with moderately irregular (HR 1.22; 95% CI: 1.08-1.39) and regular sleep (HR 1.22; 95% CI: 1.06-1.40) patterns had significantly higher risk of developing T2D. A non-linear association was observed, with the highest risk at mid-range SRI levels. Meeting sleep duration recommendations did not mitigate the elevated risk associated with irregular sleep. Notably, individuals with regular sleep but insufficient duration had a lower, though non-significant, risk of T2D (HR 0.77; 95% CI: 0.56-1.07). CONCLUSIONS: In this database-based Chinese cohort, sleep regularity showed a non-linear association with incident T2D risk, independent of sleep duration. Public health efforts to promote consistent sleep timing may offer additional benefits for diabetes prevention beyond sleep duration alone.