Background:Previous studies revealed inconsistent findings regarding the association between sleep duration and all-cause and disease-specific mortality. This study aimed to clarify the association of sleep duration with mortality using a large population-based prospective cohort study from the USA.

Methods: We used data from the National Health Interview Survey (2004-2014) linked to National Death Index records to 31 December 2015. A total of 284 754 participants aged ≥18 years were included. Self-reported sleep duration (average time slept in a 24-hour period) was categorised into seven groups: ≤4 hours, 5 hours, 6 hours, 7 hours (reference), 8 hours, 9 hours and ≥10 hours. Study outcomes included all-cause, cardiovascular disease-specific and cancer-specific mortality. Cox proportional hazards models were used to examine the association between sleep duration and mortality.

Results: During a median follow-up of 5.25 years, we identified 20 872 deaths, of which 4 129 were cardiovascular disease-related and 5 217 were cancer-related. Compared with 7 hours/day of sleep, both short and long sleep durations were associated with an increased risk of all-cause mortality (≤4 hours: HR=1.46, 95% CI=1.33-1.61; 5 hours: HR=1.22, 95% CI=1.13-1.32; 6 hours: HR=1.10, 95% CI=1.05-1.17; 8 hours: HR=1.22, 95% CI=1.17-1.28; 9 hours: HR=1.41, 95% CI=1.31-1.51; ≥10 hours: HR=2.00, 95% CI=1.88-2.13). Similar results were observed for cardiovascular disease-specific and cancer-specific mortality.

Conclusions: Our study indicates that both short (≤6 hours/day) and long (≥8 hours/day) sleep durations increase the risk of mortality compared with sleep of 7 hours/day. A normal sleep duration (about 7 hours) every day is recommended for health benefits.