Abstract
STUDY OBJECTIVES: To examine associations between perceived neighborhood social environment (PNSE) and sleep, mediated by physical activity (PA) and psychosocial stressors. METHODS: A sample (n = 4705) of African Americans (mean age 55.0 years; 63.4% female) enrolled in the 2000-2004 Jackson Heart Study (JHS). Four self-reported sleep outcomes were analyzed: sleep duration (minutes/night), sleep quality (high/low), short sleep (short ≤ 6 h vs recommended = 7-8 h), and long sleep (long ≥ 9 h vs recommended). PNSE factors included violence (e.g. robbery), problems (trash/litter), and social cohesion (trusting neighbors). PA and psychosocial stressors (lifetime and everyday discrimination, perceived stress, and depressive symptoms) were tested as mediators. With bootstrap-generated 95% bias-corrected confidence intervals (BC CIs), linear regression was used to test for mediation adjusting for covariates. RESULTS: Neighborhood violence and problems were associated with sleep duration, mediated through PA (B = -1.97, 95%BC CI = -3.76, -0.60; B = -1.23, 95%BC CI = -2.55, -0.27, respectively), lifetime discrimination (B = 2.61, 95%BC CI = 0.93, 4.80; B = 2.25, 95%BC CI = 0.93, 3.94), perceived stress (B = -3.08, 95%BC CI = -6.20, -0.41; B = -2.17, 95%BC CI = -4.33, -0.28), and depressive symptoms (B = -2.22, 95%BC CI = -5.09, -0.25; B = -1.94, 95%BC CI = -4.10, -0.35). Social cohesion was positively associated with sleep duration, mediated through PA, lifetime discrimination, and perceived stress. Similar patterns were demonstrated for binary outcomes. Yet, effect sizes were relatively small. PNSE was neither directly nor indirectly associated with sleep outcomes by everyday discrimination. CONCLUSIONS: Each PNSE factor was associated with sleep outcomes, mediated by PA and psychosocial stressors. Further research should emphasize effective community efforts to decrease adverse neighborhood conditions and psychosocial factors and increase PA; thereby reducing CVD events for African Americans.