0568 Multidimensional Sleep Health and Mortality: The Multi-Ethnic Study of Atherosclerosis

Abstract

Abstract Introduction Prior research on the association of sleep with mortality has tended to focus on single sleep dimensions such as insufficiency. However, sleep health is conceptualized to consist of multiple inter-related dimensions: prior work suggests that simultaneous consideration of multiple sleep dimensions may better predict health outcomes than individual sleep measures considered in isolation. We tested the hypothesis that better multidimensional sleep health is associated with lower mortality in the Multi-Ethnic Study of Atherosclerosis, an ethnically-diverse prospective cohort study. Methods At baseline (2010-2013), sleep was measured objectively by at-home polysomnography and 7-day wrist actigraphy, additionally by validated questionnaires. A composite sleep score was constructed in a prior project by summary index whose interpretation is the sum/count of 13 sleep attributes, selected a priori, which met clinical criteria for favorability. We used proportional hazards models to estimate the association between sleep health scores and all-cause mortality, adjusting for socio-demographics, and lifestyle/behavioral factors such as smoking status, alcohol consumption, and Body Mass Index. Results 1,757 participants were followed for a median of 5.9 years (Q1-Q3, 5.4–6.4 years) until death or last contact (128 deaths). After adjustment, each unit increase in composite sleep health scores (indicating healthier sleep) were associated with a 12% lower mortality risk (HRsummary_index: 0.88; 95% CI: 0.80, 0.96). This result was further supported by a similar association with mortality of a sleep score constructed by Principal Components Analysis. In post-hoc analyses, achievement of favorable actigraphic-estimated sleep regularity (midpoint sd <30 minutes; duration sd <90 minutes), sleep duration (6-8 hours), and Apnea-Hypopnea Index (≤15 events/hour) were significantly associated with lower mortality. Associations with mortality were in the expected direction for % N3, % R, and sleep fragmentation. Conclusion More favorable multidimensional sleep is associated with reduced mortality risk. Of specific sleep dimensions, sleep regularity, duration, and obstructive sleep apnea are suggested to be public health targets to potentially reduce risk of premature mortality. Support (If Any) This work was supported by the following National Institute of Health (NIH) grants: 5T32HL007901-19, 5K01HL138211-05, K01HL143034, R01HL098433. SR was partly funded by the NIH National Heart, Lung, and Blood Institute (NHLBI) R35HL135818.