Abstract
The authors examined sociodemographic, lifestyle, and comorbidity factors that could confound or mediate U-shaped associations between sleep duration and health in 6,472 United Kingdom adults from the Whitehall II Study (1997-1999) and 3,027 US adults from the Western New York Health Study (1996-2001). Cross-sectional associations between short (<6 hours) and long (>8 hours) durations of sleep across several correlates were calculated as multivariable odds ratios. For short sleep duration, there were significant, consistent associations in both samples for unmarried status (United Kingdom: adjusted odds ratio (AOR) = 1.49, 95% confidence interval (CI): 1.15, 1.94; United States: AOR = 1.49, 95% CI: 1.10, 2.02), body mass index (AORs were 1.04 (95% CI: 1.01, 1.07) and 1.02 (95% CI: 1.00, 1.05)), and Short Form-36 physical (AORs were 0.96 (95% CI: 0.95, 0.98) and 0.97 (95% CI: 0.96, 0.98)) and mental (AORs were 0.95 (95% CI: 0.94, 0.96) and 0.98 (95% CI: 0.96, 0.99)) scores. For long sleep duration, there were fewer significant associations: age among men (AORs were 1.08 (95% CI: 1.01, 1.14) and 1.05 (95% CI: 1.02, 1.08)), low physical activity (AORs were 1.75 (95% CI: 0.97, 3.14) and 1.60 (95% CI: 1.09, 2.34)), and Short Form-36 physical score (AORs were 0.96 (95% CI: 0.93, 0.99) and 0.97 (95% CI: 0.95, 0.99)). Being unmarried, being overweight, and having poor general health are associated with short sleep and may contribute to observed disease associations. Long sleep may represent an epiphenomenon of comorbidity.