Abstract 4349201: Differential Associations by Sex Between Sleep-Wake Regularity and All-Cause and Cardiovascular Mortality

Abstract

Introduction: Highly irregular sleep-wake timing is prospectively associated with higher all-cause and cardiovascular (CVD) mortality rates. These relationships are hypothesized to reflect harmful effects of inconsistent patterns of exposures (eg, light-dark), behaviors (eg, sleep-wake), and circadian rhythms (ie, intrinsic time-varying physiology). Sex differences span sleep and circadian physiology, yet sex-specific relationships between sleep regularity and mortality have not been assessed. Question: Using the Sleep Regularity Index (SRI), a metric that captures circadian disruption through the average probability that participants are in the same sleep-wake state 24 hours apart, this study examined whether associations between the day-to-day regularity of sleep-wake timing and mortality differ by sex. Methods: The study included UK Biobank participants with week-long accelerometer-derived sleep-wake data. To reduce reverse causality bias, deaths within one year of recordings were excluded. SRI was calculated by sleepreg . All-cause and CVD mortality were ascertained from inpatient hospital records and death registrars. Age-adjusted Cox proportional hazards models estimated adjusted hazards ratios (aHRs) between SRI categories and time to all-cause or CVD mortality without and with SRI × sex interaction term. SRI–sex combinations were compared using estimated marginal means. Results: Among 73,647 adults (41,489 [56.3%] female, mean age 62.5±7.8 y) followed for a mean of 7.5±0.9 y, there were 2631 deaths; 20.5% were from primarily CVD causes. Higher age-adjusted mortality rates were observed with the lowest SRI (<60) compared to the highest (≥90) for both all-cause (aHR=2.4; 95% CI=2.0–2.9) and CVD mortality (3.2; 2.1–4.6) (Fig 1). Significant interaction terms were identified for SRI <60 × male (all-cause) and SRI 60–69 × male (CVD). Higher all-cause and CVD mortality rates were associated among males with SRI <60 versus ≥90 (all-cause 2.9; 2.1–4.0; CVD 3.8; 1.9–7.5) and compared to females with SRI <60 (all-cause 2.1; 1.5–3.0; CVD 2.5; 1.2–5.0) (Fig 2). In contrast, neither all-cause nor CVD mortality differed by sex among individuals with SRI ≥90. Conclusion: Higher mortality rates associated with highly irregular sleep-wake timing are particularly pronounced among males, especially for CVD mortality. The absence of sex differences among those with high sleep regularity suggests that consistent sleep-wake patterns may attenuate heightened male mortality rates.