LAUSR

OBJECTIVE Our aim was to examine sex- and age-specific relationships of sleep behaviors with all-cause mortality rates. METHODS A retrospective cohort study was conducted among 5288 adults (≥50 years) from the 2005-2008 National Health and Nutrition Examination Surveys who were followed-up for 54.9 ± 1.2 months. Sleep duration was categorized as < 7 h, 7-8 h and >8 h. Two sleep quality indices were generated through factor analyses. 'Help-seeking behavior for sleep problems' and 'diagnosis with sleep disorders' were defined as yes/no questions. Sociodemographic covariates-adjusted Cox regression models were applied to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). RESULTS A positive relationship was observed between long sleep and all-cause mortality rate in the overall sample (HR = 1.90, 95% CI: 1.38, 2.60), among males (HR = 1.48, 95% CI: 1.05, 2.09), females (HR = 2.32, 95% CI: 1.48, 3.61) and elderly (≥65 years) people (HR = 1.80, 95% CI: 1.30, 2.50). 'Sleepiness/sleep disturbance' (Factor I) and all-cause mortality rate were positively associated among males (HR = 1.22, 95% CI: 1.03, 1.45), whereas 'poor sleep-related daytime dysfunction' (Factor II) and all-cause mortality (HR = 0.75, 95% CI: 0.62, 0.91) were negatively associated among elderly people. CONCLUSIONS Sex- and age-specific relationships were observed between all-cause mortality rate and specific sleep behaviors among older adults.