In the Multi-Ethnic Study of Atherosclerosis (n=1,620), we found prospective evidence that multidimensional sleep health is associated with cardiovascular events and mortality. In Cox Hazard models, we estimated that a 1-standard deviation increase in in a composite Sleep Health Score (obtained by PCA) was associated with an average of i) 19% and ii) 27% lower risk of experiencing an i) incident cardiovascular event (HR: 0.81 [0.66-1.00]; n events = 89) or ii) mortality (HR: 0.73 [0.61-0.87]; n events = 112) at median follow-up times of i) 5.85 and ii) 5.92 years.

Post-hoc analyses revealed that participants with higher composite Sleep Health Scores evinced more favorable sleep across all metrics: greater objective sleep duration, regularity in timing and duration, and sleep continuity (less fragmentation, fewer minutes of wake after sleep onset and thus greater sleep maintenance efficiency), a greater proportion of the night spent in N3 and Rapid Eye Movement sleep, less severe sleep disordered breathing (AHI), sleep midpoints closer to midnight (timing), and more favorable after-sleep evaluations in sleep quality and alertness.

We also estimated the component contributions of each sleep dimension to outcomes. Our secondary analyses found evidence that, in our sample, the sleep health–cardiovascular event relationship appears to be driven by i) sleep disordered breathing, ii) % Rapid Eye Movement, and plausibly iii) duration variability, but less so by a) total sleep time, b) timing regularity, c) continuity, and d) subjective evaluations of sleep (quality, alertness). The sleep health–mortality relationship in our sample appears to be driven by i) sleep disordered breathing, ii) duration variability, iii) sleep duration, and plausibly iv) placement in the 24-hour day (timing), but less so by a) sleep latency, b) % REM, c) % N3, d) continuity, and e) subjective evaluations of sleep (quality and alertness).