Three-year trajectories of sleep duration and mortality in patients with type 2 diabetes-a hospital-based retrospective cohort study.

OBJECTIVES:Most studies have shown that a single item of self-reported sleep duration is related to mortality risk. However, the long-term effect of sleep duration on mortality remains unclear in patients with diabetes. This study aimed to examine the associations of 3-year trajectory patterns of sleep duration with all-cause and expanded cardiovascular disease mortality in patients with type 2 diabetes. METHODS:Patients with type 2 diabetes and self-reported sleep duration during a 3-year interval were included. Expanded cardiovascular disease was defined as death due to cardiovascular disease, diabetes, and kidney diseases. Cox's proportional hazards models were employed to examine the associations between sleep duration patterns and mortality after controlling for sociodemographic factors, lifestyle behaviors, diabetes-related variables, diabetic complications, and medication use. RESULTS:A total of 7591 patients were included for analysis, and 995 deaths (13.11%) and 424 expanded cardiovascular disease deaths (5.59%) were observed during a mean follow-up of 8.51 years. Five trajectory patterns of sleep duration were identified: cluster 1: "constant 7- to 8-hour group" (50.03%); cluster 2: "constant low group" (19.68%); cluster 3: "high with decreasing trend group" (3.08%); cluster 4: "low with fluctuation group" (1.28%); and cluster 5: "constant high group" (25.93%). Compared with cluster 1, clusters 3 and 4 were associated with increased risks of all-cause mortality (1.41, 1.08-1.84; 1.44, 1.01-2.05), and cluster 5 was associated with high risks of all-cause and expanded cardiovascular disease mortality (1.26, 1.08-1.46; 1.42, 1.12-1.79). CONCLUSIONS:Sleep duration trajectories with constant high or unstable patterns may be associated with higher mortality.