Sleep duration among black and white Americans: results of the National Health Interview Survey

INTRODUCTION: Epidemiologic studies have shown the importance of habitual sleep duration as an index of health and mortality risks. However, little has been done to ascertain ethnic differences in sleep duration in a national sample. This study compares sleep duration in a sample of black and white participants in the National Health Interview Survey (NHIS). METHOD: Data were collected from 29,818 Americans (age range 18-85 years) who participated in the 2005 NHIS. The NHIS is a cross-sectional household interview survey that uses a multistage area probability design, thus permitting representative sampling of U.S. households. During face-to-face interviews conducted by trained interviewers from the U.S. Census Bureau, respondents provided demographic data and information about physician-diagnosed chronic conditions, estimated habitual sleep duration and functional capacity, and rated their mood. RESULTS: Fisher's exact test results indicated that blacks were less likely than whites to report sleeping 7 hours (23% vs. 30%; chi2 = 94, p < 0.0001). Blacks were more likely to experience both short sleep (< or = 5 hours) (12% vs. 8%, chi2 = 44, p < 0.0001) and long sleep (> or = 9 hours) (11% vs. 9%, chi2 = 23, p < 0.0001). Logistic regression analysis, adjusting for differences in sociodemographic factors, depression, functional capacity and medical illnesses, demonstrated that black ethnicity was a significant predictor of extreme sleep duration (Wald = 46, p < 0.0001; OR = 1.35, 95% CI: 1.24-1.47). DISCUSSION: Independent of several sociodemographic and medical factors, blacks had more prevalent short and long sleep durations, suggesting greater variation in habitual sleep time. Therefore, blacks might be at increased risks of developing medical conditions associated with short and long sleep.