Introduction: The association of cardiovascular (CV) disease with inadequate sleep is well-documented, with evidence pointing toward a J-shaped curve, where the worst CV risk is found at both extremes (high and low) of sleep duration. Few studies have examined this association in a diverse population, or with subclinical measures of CV risk or atherosclerosis.
Methods: We examined data from the Bogalusa Heart Study, a long-term community-based observational study of a biracial cohort, with first measurements in childhood in 1973. Men and women (n=1245) who underwent a measure of carotid intima-media thickness (IMT), calculated as the composite IMT (cIMT), mean of six segments (common and internal carotid, and carotid bifurcation, each left and right), and completed a questionnaire on sleep duration, were included. Analyses of variance (ANOVA) and multivariable analyses of covariance (ANCOVA) were employed to model the association between cIMT and sleep duration, adjusted for age, sex, race, smoking status (at time of IMT measurement), shift work, physical activity, BMI, systolic blood pressure, HDL/LDL cholesterol ratio, diabetes, hypertension, and previous heart attack. Sleep duration was the weighted average of weekday and weekend sleep hours reported, and were categorized as: less than 6 hours, 6-7, 7-8, 8-9, 9-10, and more than 10 hours per night.
Results: One-way ANOVA found a significant difference in mean cIMT between the categories of sleep duration (p<0.01). Multivariable ANCOVA found those who slept less than 6 hours had mean cIMT approximately 0.06 mm greater than those who slept 8 hours (p=0.05), and those who slept 9 hours had 0.1 mm smaller mean cIMT than those who slept 8 hours (p=0.05). Overall, the model explained 16.1% of the variation in cIMT.
ConclusionS: Sleep duration is significantly associated with cIMT, with the lowest mean cIMT found in those who slept 9 hours per night.