"Sleep Duration and Metabolic Syndrome in U.S. Adults: Evidence of a U-Shaped Association From NHANES 2005-2020".

Abstract

PurposeTo examine the association between sleep duration and metabolic syndrome (MetS) prevalence in U.S. adults.DesignCross-sectional analysis of National Health and Nutrition Examination Survey (NHANES) data.SettingUnited States.Sample14 982 adults (≥18 years) from 7 NHANES cycles (2005-2020) with complete data on sleep, MetS components, and covariates.MeasuresSelf-reported sleep duration (short: <7 h, normal: 7-9 h, long: >9 h). MetS was defined by NCEP ATP III criteria. Covariates included age, sex, race/ethnicity, sedentary time, and smoking.AnalysisSurvey-weighted logistic regression modeled sleep as continuous linear and quadratic terms. Sensitivity analyses included restricted cubic splines and component-wise assessment.ResultsA U-shaped association was observed between sleep duration and MetS. Compared with normal sleep, both short (<7 h) and long (>9 h) sleep were associated with higher odds of MetS in fully adjusted models. Linear (OR = 0.68, 95% CI: 0.56-0.82) and quadratic (OR = 1.03, 95% CI: 1.01-1.04) terms confirmed nonlinearity (P < 0.001). The model-derived sleep duration associated with the lowest predicted odds of MetS was 7.52 h/night (95% CI: 6.81-8.24 h). Age, sedentary time, and smoking independently increased MetS risk. The U-shaped pattern persisted across most individual MetS components and was supported by spline analyses.ConclusionBoth short and long sleep durations were significantly associated with higher odds of metabolic syndrome, highlighting sleep duration as a potential indicator for cardiometabolic risk screening.