Sleep duration predicts subsequent long-term mortality in patients with type 2 diabetes: a large single-center cohort study

Background: Sleep duration is associated with mortality. However, prior studies exploring whether sleep duration predicts subsequent long-term mortality in patients with diabetes are limited. This study aims to examine whether metabolic factors affect the associations between baseline sleep duration and subsequent risks of all-cause, expanded, and non-expanded cardiovascular disease (CVD) mortalities among patients with type 2 diabetes (T2D).

Methods: A total of 12,526 T2D patients aged 30 years and older, with a follow-up period ≥ 3 years, were identified from the Diabetes Case Management Program of a medical center in Taiwan. Sleep duration was measured using computerized questionnaires by case managers, and the time frame for this question was 1 month prior to the interview date. Sleep duration in relation to subsequent mortality from all causes, expanded CVD, and non-expanded CVD was examined using Cox proportional hazard models.

Results: Within 10 years of follow-up, 2918 deaths (1328 CVD deaths and 1590 non-CVD deaths) were recorded. A J-shaped association was observed for all-cause, expanded CVD, and non-expanded CVD mortalities, and the lowest risks were observed for patients with 5-7 h of sleep. The significant joint effects included sleep duration of more or less than 7 h with age ≥ 65 years [adjusted HRs: 4.00 (3.49-4.60)], diabetes duration ≥ 5 years [1.60 (1.40-1.84)], age at diabetes diagnosis ≤ 45 years [1.69 (1.38-2.07)], insulin use [1.76 (1.54-2.03)], systolic blood pressure/diastolic blood pressure > 130/85 mmHg [1.24 (1.07-1.43)], triglyceride ≥ 150 mg/dL [1.38 (1.22-1.56)], HbA1c ≥ 7% [1.31 (1.13-1.52)], and body mass index < 27 kg/m² [1.31 (1.17-1.45)] for all-cause mortality.

Conclusion: A J-shaped association was observed between sleep duration and all-cause and expanded CVD mortality, and a sleep duration of 5-7 h had the lowest mortality risk. Sleep duration also showed significant synergistic interactions with diabetes duration but shared an antagonistic interaction with age and obesity.