Study objectives: Short and/or long sleep duration are associated with cardiometabolic disease risk and may be differentially experienced among minorities and the socioeconomically disadvantaged. The present study examined nationally representative data along multiple dimensions of race/ ethnicity and socioeconomic status.

Design: Cross-sectional.

Setting: Survey.

Patients or participants: 2007-2008 NHANES (N = 4,850).

Interventions: None.

Measurements and results: Self-reported sleep duration was classified as very short (< 5 h), short (5-6 h), normative (7-8 h) and long (≥ 9 h). Population-weighted multinomial logistic regression analyses examined race/ ethnicity, country of origin, language, income, education, health insurance, and food security, controlling for all others as well as age, sex, marital-status, and overall self-rated health. Outcome was self-reported sleep duration, relative to normative sleep duration. Blacks/African Americans were more likely than whites to report very short (OR = 2.34, P < 0.001) and short (OR = 1.85, P < 0.001) sleep. Mexican Americans reported less long sleep (OR = 0.36, P = 0.032). Other Hispanics/ Latinos reported more very short sleep (OR = 2.69, P = 0.025). Asians/ Others reported more very short (OR = 3.99, P = 0.002) and short (OR = 2.08, P = 0.002) sleep. Mexico-born adults reported less short sleep (OR = 0.63, P = 0.042). Spanish-only speakers reported less very short sleep (OR = 0.32, P = 0.030). Lower income groups reported more very short sleep versus > $75,000. Compared to college graduates, increased very short sleep was seen among all lower education levels. Those with public insurance reported more very short (OR = 1.67, P = 0.31) and long (OR = 1.83, P = 0.011) sleep versus uninsured. Very low food security was associated with very short (OR = 1.86, P = 0.036) and short (OR = 1.44, P = 0.047) sleep.

Conclusions: Minority status and lower socioeconomic position were associated with shorter self-reported sleep durations.