Abstract
BACKGROUND: Poor sleep is a potential risk factor for metabolic syndrome (MetS), and its relationship with MetS may vary by race/ethnicity and menopausal status among women. METHODS: We used Sister Study enrollment data from 2003 to 2009 to investigate the cross-sectional associations between multiple subjective sleep characteristics and having ≥ 3 prevalent metabolic abnormalities consistent with MetS among white, black, and Hispanic/Latina women. Self-reported sleep characteristics included average sleep duration (short [< 7 h] vs. recommended [7-9 h]), sleep debt (≥ 2-h difference between shortest and longest sleep duration, napping ≥ 3 times/week, and insomnia symptoms (difficulty falling or staying asleep). We used Poisson regression with robust variance to estimate adjusted prevalence ratios (PRs) and 95% confidence intervals (CIs) to compare MetS prevalence between women with poor sleep (e.g., short sleep, sleep debt, frequent napping, or insomnia symptoms [all yes vs. no]) and non-poor sleep within menopausal status categories (premenopausal or postmenopausal). We adjusted for sociodemographic characteristics, mental health, and health behaviors. RESULTS: Among 38,007 eligible women (13,988 premenopausal, 24,019 postmenopausal), mean age was 55 ± 8.8 years, racial/ethnic composition was 86.63% white, 8.53% black, and 4.84% Hispanic/Latina, and 12% had MetS. Associations between certain poor sleep characteristics [i.e., short sleep (PR(premenopausal) = 1.23 [95% CI 1.06-1.42], PR(postmenopausal) = 1.09 [1.02-1.16], p(short sleep*menopause) = 0.0070) and insomnia symptoms (PR(premenopausal) = 1.21 [1.05-1.41], PR(postmenopausal) = 1.11 [1.05-1.18], p(insomnia symptoms*menopause) = 0.035)] and prevalent MetS were stronger among premenopausal compared to postmenopausal women, but did not vary by race/ethnicity. Associations between concurrent short sleep/insomnia symptoms and MetS were stronger among white and Hispanic/Latina postmenopausal women compared to their black counterparts. Menopausal status and race/ethnicity did not modify positive associations for other poor sleep characteristics. CONCLUSIONS: Poor sleep was positively associated with MetS prevalence. Associations between individual poor sleep characteristics (i.e., short sleep, insomnia symptoms) were stronger among premenopausal compared to postmenopausal women but did not vary by race/ethnicity.