0827 ASSOCIATIONS BETWEEN EVERYDAY DISCRIMINATION AND SLEEP AMONG AFRICAN AMERICANS IN THE JACKSON HEART STUDY

Abstract

African Americans are more likely to experience poor sleep than non-Hispanic whites. Psychosocial stressors such as discrimination may contribute to the higher prevalence of poor sleep among African Americans; however, this association is underexplored. We investigated longitudinal associations between discrimination and sleep and explored interactions with sex in the Jackson Heart Study (JHS). At two time points (2000–2004 and 2008–2012), participants (N=3749) completed the Williams Everyday Discrimination Scale, self-reported hours of sleep (sleep duration) and rated their sleep quality (1=poor -5=excellent). We fit a series of mixed-effects logistic and linear regression models to examine the association of discrimination (in quartiles) with categorical sleep duration (≤6 hours, or “short” vs. 7 or 8 hours), continuous sleep duration and sleep quality accounting for survey years, demographics, socioeconomic status, and other potential confounders (body mass index, hypertension, diabetes, physical activity, stress, alcohol, social support, smoking, depressive symptoms). At baseline, participants’ mean age was 55.4 ± 12.8 years and 63.5% were female. The prevalence of short sleep was 54.7% and 55.8% at each time point, respectively. Men reported a lower mean sleep duration and reported more experiences with discrimination. The association between discrimination and sleep did not change overtime. In pooled analyses, participants who reported higher levels of discrimination vs. lower levels, had a 43% higher odds (odds ratio (OR)=1.43 confidence interval (CI)=1.26, 1.63) of short sleep, slept 15.3 minutes (-20.63, -9.89) less on average, and had a lower sleep quality score (-0.25 (-0.31, -0.18)). Perceived stress partially explained the associations. In fully-adjusted models, associations were attenuated but persisted. Sex modified the association between everyday discrimination and sleep duration, P<0.10. Women who reported higher levels of discrimination slept 22.3 minutes (-29.93, -14.70) less on average compared to those who reported lower levels of discrimination; there was no association observed among men. Everyday discrimination was associated with shorter sleep duration and poorer sleep quality across two time points, and women were more vulnerable to the association with sleep duration. Future research should identify interventions that may buffer the effects of discrimination on sleep among African Americans. 3R01HL110068-03 and 5T32HL007901-18