[A prospective cohort study on the association of sleep duration and fasting plasma glucose with risk of stroke].

Objective: To investigate the association of sleep duration and fasting plasma glucose with the risk of stroke. Methods: A Prospective cohort study. Two towns in Xin'an County, Henan Province were selected as study sites. Using natural villages as the sampling unit, cluster random sampling was used to recruit 20 194 participants in 2007-2008 for the baseline study, and two follow-up studies were completed in 2013-2014 and 2018-2020, respectively. A total of 13 741 eligible participants were included in the analysis. Multivariate Cox proportional hazards regression model was used to assess the association of sleep duration and fasting plasma glucose with stroke risk and the restricted cubic spline was applied to fit the dose-response relationships. Results: The age [M (Q1, Q3)] of 13 741 participants was 51.0 (41.0, 60.0) years, and 5 308 (38.63%) participants were male. Over a median follow-up of 10.18 years, 1 612 incident stroke events were observed. Compared with participants sleeping 7 to<8 h/d, the risk of stroke increased for those sleeping<6 and≥10 h/d, with HR (95%CI) of 1.44 (1.04-2.00) and 1.37 (1.13-1.66), respectively. A U-shaped relationship between sleep duration and stroke risk was observed (Pnonlinear<0.001). Compared to people with normal fasting glucose, the HR (95%CI) of stroke risk was 1.58 (1.36-1.84) for diabetes patients. A nonlinear association between fasting plasma glucose level and stroke risk was observed (Pnonlinear=0.036). Compared to individuals with sleep duration of 7 to<9 h/d and normal fasting glucose, the HR (95%CI) of stroke risk was 1.60 (1.18-2.16) for those with sleep duration<7 h/d and prediabetes, and 2.07 (1.27-3.39) for those with sleep duration<7 h/d and diabetes; the HR (95%CI) of stroke risk was 1.18 (1.01-1.39) for those with sleep duration≥9 h/d and prediabetes, and 1.84 (1.50-2.26) for those with sleep duration≥9 h/d and diabetes. But no significant interaction between sleep duration and fasting plasma glucose was found on stroke risk (P>0.05). Conclusions: Inadequate sleep duration, prolonged sleep duration and elevated fasting plasma glucose level were associated with increased risk of stroke, and the stroke risk was further increased when sleep duration<7 or≥9 h/d coexisted with abnormal fasting plasma glucose states.