OBJECTIVETo explore the association between sleep duration and stroke in adults.

METHODSBaseline data of 57 704 subjects who were aged 30-79 years and enrolled into China Kadoorie Biobank (CKB) study from Tongxiang county, Zhejiang province were analyzed. Multiple logistic regression analysis was conducted to investigate the putative association between sleep duration and stroke after adjusting for potential confounders.

RESULTSThe mean age of the subjects was (53.15 ± 10.20) years in males and (51.72 ± .69) years in females, respectively. There were 14.43% of males and 15.30% of females reporting sleep durations ≤ 6 hours per day and 5.39% of males and 5.95% of females reporting long duration of sleep (≥ 10 hours per day). The prevalence of stroke was 0.92% in males compared with 0.44% in females. The prevalence of stroke showed a U-shaped distribution with sleep duration. Compared with 7 hours sleep duration per day, long sleep duration (≥ 10 hours per day) was associated with stroke. The odds ratios (OR) were 2.11 (95%CI: 1.32-3.37) for males and 2.13 (95%CI: 1.24-3.65) for females after adjusting for age, socioeconomic status, health behaviors and health status. No statistical significant association was found between short sleep duration and stroke. Meanwhile, frequent sleep snoring was found to be associated with stroke in females (OR=1.63, 95% CI: 1.11-2.40).

CONCLUSIONLonger sleep duration was found to be associated with higher risk of stroke in both males and females. Frequent sleep snoring would increase the risk of stroke in females.