Siesta and mortality in a Mediterranean population: a community study in Jerusalem.

Abstract

In light of contradictory studies relating to the health effects of siesta, we investigated the practice of siesta in relation to all-cause and cardiovascular mortality, controlling for recognized predictors of mortality. Prospective study with 9 to 11 years of mortality follow-up. West Jerusalem neighborhood of Kiryat Yovel. 1859 residents (55% women), aged at least 50 years in 1985-87 (85% response). Data on daytime napping, its average duration, and duration of night sleep were collected in addition to a rich array of covariates. We identified 405 deaths (206 women) during an average 10-year follow-up undertaken by linkage with the National Population Registry. Cox regression was used to control for potential confounders. In multivariable models among men, there was a weak association with all-cause mortality (hazard ratio [HR], 1.36; 95%CI, 0.93-1.97), which was restricted to ages 65 to 74 years (HR, 2.21; 95%CI, 1.28-3.80; P for age interaction, 0.008). Exclusion of patients with chronic conditions attenuated the association. In women, siesta was not associated with mortality. Long siesta (>2h), a marker of daytime somnolence, was strongly associated with excess risk of all-cause and cardiovascular mortality in men. The weak overall relation of siesta with mortality in men that was attenuated upon exclusion of patients with chronic conditions, the absence of internal consistency in the association across age and sex groups, and the association with long daytime naps tend to argue against a causal role for siesta. Nevertheless, the data are also compatible with long siestas conferring excess risk. However, without persuasive data, recommendations related to this traditional practice should not be made.