Disentangling racial/ethnic and socioeconomic differences in self-reported sleep measures: the Boston Area Community Health Survey

Abstract Background Racial/ethnic differences in sleep are widely reported. However, the extent to which socioeconomic factors account for crude variation in sleep parameters by racial/ethnic groups is not clearly understood. Methods We studied 4144 individuals in the Boston Area Community Health Survey (2006-2010), a racially/ethnically balanced population-based cohort of black, Hispanic, and white men and women. Self-reported sleep measures were sleep duration, sleep latency, restless sleep, risk for sleep apnea, and sleep medication use. We assessed changes in the age- and sex-adjusted association between race/ethnicity and sleep measures after additional adjustment for individual socioeconomic factors (income, education, and employment) and lifestyle and behavioral factors. Results Self-identified non-Hispanic black race/ethnicity was significantly associated with higher odds of short sleep duration (fully adjusted odds ratio [OR], 1.93; 95% confidence interval [CI], 1.41-2.64), longer sleep latency (fully adjusted, 20.9% longer; 95% CI, 4.1-41.9), and lower odds of pharmaceutical sleep aid use (fully adjusted OR: 0.47, 95% CI: 0.34, 0.65) than white race/ethnicity. Education level accounted for the most substantial reduction in estimates of the age- and sex-adjusted association between black race/ethnicity and short sleep duration and sleep latency. Having less than a 4-year college education was associated with approximately 55% lower sleep latency than having postgraduate education. No significant associations were observed comparing Hispanic to white participants. Conclusions Significant variation was observed between black and white race/ethnicity in short sleep duration, sleep latency, and sleep aid use. Although considerable variation in sleep by race/ethnicity was explained by education level, additional variation in self-reported sleep by race/ethnicity may be due to other unmeasured socioeconomic, lifestyle, or behavioral factors.