Objective: To cross-sectionally and prospectively investigate the association between irregular sleep patterns, a potential marker for circadian disruption, and metabolic abnormalities.

Research design and methods: In the Multi-Ethnic Study of Atherosclerosis, participants completed 7-day actigraphy at exam 5 (2010-2013) and were prospectively followed throughout exam 6 (2016 to 2017). Sleep regularity was quantified by the 7-day SD of actigraphy-assessed sleep duration and sleep onset timing. Metabolic abnormalities were defined by 1) the National Cholesterol Education Program Adult Treatment Panel III criteria and 2) a data-driven clustering of metabolic factors.

Results: In the exam 5 cross-sectional analysis adjusted for sociodemographic and lifestyle factors (n = 2,003), every 1-h increase in the sleep duration SD was associated with 27% (95% CI 1.10, 1.47) higher odds of metabolic syndrome, and every 1-h increase in the sleep timing SD was associated with 23% (95% CI 1.06, 1.42) higher odds. The associations remained significant with additional adjustment for sleep-related factors including sleep duration. In the prospective analysis (n = 970), the corresponding fully adjusted odds ratio (OR) was 1.27 (95% CI 0.97, 1.65) for sleep duration and 1.36 (1.03, 1.80) for sleep timing. Compared with the cluster of few metabolic changes, every 1-h increase in sleep variability was associated with almost doubled odds for the cluster characterized by incidence of multiple metabolic abnormalities (OR 1.97 [95% CI 1.18, 3.30] for sleep duration and OR 2.10 [95% CI 1.25, 3.53] for sleep timing).

Conclusions: Increased variability in sleep duration and timing was associated with higher prevalence and incidence of metabolic abnormalities even after consideration of sleep duration and other lifestyle factors.