Abstract
This cross-sectional population study examined associations of sleep duration and morning-evening type with sociodemographic and cardiometabolic disease in adults participating in the UK Biobank study (N = 439 933). Multivariable Poisson regression models of sleep duration and morning-evening type with a robust error variance were generated to estimate adjusted prevalence ratios and their 95% confidence intervals. All models were adjusted for sex, race, college attendance, employment status and age. Twenty five percent of the sample reported short sleep; 27% were morning, 64% intermediate and 9% evening type. Black ethnicity emerged as most strongly associated with sleep behavior. Short sleep was twice as prevalent, and morning versus intermediate type was 1.4 times more prevalent in Black than White participants. The greater prevalence of short sleep and morning type among Blacks suggests that sleep-based approaches to improving cardiometabolic outcomes may require a more multidimensional approach that encompasses adequate sleep and circadian alignment in this population.</p>