Association of Sleep Duration With All- and Major-Cause Mortality Among Adults in Japan, China, Singapore, and Korea

Key Points Question What is the association between sleep duration and mortality outcomes in individuals from East Asia? Findings In this cohort study of 322 721 adult participants, when compared with 7 hours, all other sleep durations were associated with a significantly increased mortality risk. Age and sex were significant modifiers of the associations between sleep duration and death from all causes (age), cardiovascular disease (sex), cancer (age and sex), and other causes (age and sex), and age was a modifier of the association only among men. Meaning These findings suggest that sleep duration recommendations for East Asian populations may need to be considered in the context of sex and age.

smoker), alcohol consumption (non-drinker/past drinker, current drinker <150g ethanol/week, and current drinker ≥150g ethanol/week), physical activity (low: once per week or < 1 hour per week; intermediate: 1-4 days per week or 1-4 hours per week; high: almost daily or ≥ 5 hours per week) according to each cohort's questionnaire, history of diabetes, history of hypertension (prevalent hypertension, systolic blood pressure ≥ 140, or diastolic blood pressure ≥90), and menopausal status (for women only) were all based on self-reported information. For the SWHS, information on menopausal status and prevalent hypertension was derived from both the baseline questionnaire and the third follow-up survey.

All-cause mortality
Age was a statistically significant effect modifier for all-cause mortality among men (p<0.001) but not women (p=0.62). When stratifying analyses by age, younger men were at an increased risk of all-cause mortality with all sleep duration categories compared to the referent 7 hours (eTable 4). Among older men, there was an increased risk only with sleep durations of 9 hours, and ≥10 hours in both the minimallyand multivariable adjusted model. BMI was not an effect modifier for all-cause mortality in men (p=0.33) or women (p=0.67). Analyses stratified by BMI are shown in eTables 6 and 7.

Cancer mortality
Among men, age (p<0.001) but not BMI (p=0.88) was a statistically significant effect modifier of the association between sleep duration and cancer mortality. In age-stratified analyses, younger men had an increased risk of cancer mortality with sleep durations of ≤5 hours, 8 hours, 9 hours, and ≥10 hours in both the minimally-and multivariable adjusted models. Sleep duration was unrelated to risk of cancer mortality among older men. In women, neither age (p=0.38) nor BMI (p=0.64) were effect modifiers of the association between sleep duration and cancer mortality.

Other-cause mortality
In men, age (p=0.002) but not BMI (p=0.27) was a statistically significant effect modifier of the association between sleep duration and mortality from other causes. Age-stratified analyses revealed increased risks for younger men with sleep durations of ≤5 hours, 8 hours, 9 hours, and ≥10 hours in both models. Older men were at an increased risk of other-cause mortality only with sleep durations of 9 hours, and ≥10 hours in both models.
Neither age (p=0.17) nor BMI (p=0.29) were effect modifiers for other-cause mortality in women. eTable 1. Participating cohort characteristics * * Cohort characteristics are shown according to the final study population (i.e., individuals who are complete cases without any prevalent cardiovascular disease (CVD) or cancer at baseline, and who did not die within the first five years of follow-up Model 1 was adjusted for age, marital status, and study area (JPHC Study only) Model 2 was further adjusted for body mass index, smoking, alcohol intake, physical activity, history of diabetes, hypertension, and menopausal status *p<0.05, **p<0.01, ***p<0.001 vs reference eTable 4 HRs and CIs for mortality according to sleep duration in men stratified by age (younger/older than 65 years of age) Model 1 was adjusted for marital status, and study area (JPHC Study only) Model 2 was further adjusted for body mass index, smoking, alcohol intake, physical activity, history of diabetes, and hypertension *p<0.05, **p<0.01, ***p<0.001 vs reference eTable 5 HRs and CIs for mortality according to sleep duration in women stratified by age (younger/older than 65 years of age)