Long sleep duration is associated with higher mortality in older people independent of frailty: a 5-year cohort study

Objectives: Although general adult population studies show a U-shaped association between sleep duration and mortality, prolonged rather than short sleep duration was more consistently associated with higher mortality in older populations. Failing health or frailty is a possible mechanism. Thus, we examined the relationship among sleep duration, frailty status, and mortality in an elderly cohort.

Methods: A total of 3427 community-living adults 65 years or older were examined for general health, mood, subjective sleep measures (insomnia, napping, sleep apnea, nighttime sleep duration, sleep medications), frailty, and 5-year mortality.

Results: After 5 years, 12.9% of men and 4.5% of women had died. Mean nighttime sleep duration was 7.3 hours. Proportion of participants who slept 10 or more hours increased with increasing frailty. Age-adjusted hazard ratio (HR) for 5-year mortality of long nighttime sleep (≥ 10 hours) was 2.10 (95% confidence interval [CI] 1.33-3.33) in men, and 2.70 (95% CI 0.98-7.46) in women. The HR in men was attenuated (HR 1.75; 95% CI 1.09-2.81) after adjustment for frailty and other covariates, whereas that of women strengthened (HR 2.88; 95% CI 1.01-8.18). Mortality increased sharply with nighttime sleep of 10 hours or more. Nighttime sleep of 10 or more hours (HR 1.75, men; HR 2.88, women) and frailty (HR 2.43, men; HR 2.08, P = .08 in women) were independently associated with 5-year mortality after full adjustment for covariates.

Conclusion: Frailty and long nighttime sleep duration of 10 or more hours were independently associated with 5-year mortality in older adults.