The siesta and mortality in the elderly: effect of rest without sleep and daytime sleep duration

To examine the effects of daytime rest without sleep and duration of the siesta on mortality.

Longitudinal observation.

Population sample of 442 community-dwelling 70-year-old subjects examined both at home and in a geriatric hospital.

N/A.

Overall mortality for those who neither rested nor slept was 9.3%, for those who rested without sleep 10.9%, and for those who slept 19%, p<0.02. Males had higher mortality (19%) than females (10%), p<0.006. Rest without sleep had no effect on mortality: in males, 13%, 21%, 18% for no rest, rest of less than one hour, or more than one hour with respective 8%, 6% and 4% in females. However, daytime sleep in males of more than one hour had 28.2% mortality, whereas sleep of one hour or less had 13.6% mortality, p=0.02. In females mortality rate was not different by sleep duration: 16% and 13.6% for those who slept for one hour or less, or more than one hour, respectively. In a multivariate analysis in females, a siesta of one hour or less was associated with risk odds ratio (ROR) of 4.67 and 95% confidence interval (CI) 1.22-17.80 and of one to two hours, ROR was 5.57 and 95% CI 1.05-29.49. For males, siesta of less than one hour was not associated with excess risk, ROR 0.9 and 95% CI 0.39-2.38; a siesta of one to two hours with ROR of 2.61 and 95% CI 1.01-6.80; and two hours or more ROR 13.6 and 95% CI 0.98-2.10.

Rest without sleep is not associated with excess risk of mortality. However, siesta of one to two hours is associated with increased mortality in males whereas, in women, a siesta of less than one hours confers the excess risk.