Race, place, and health: how does neighborhood racial/ethnic context affect health behavior and risk factors

This dissertation builds upon current debates on the detrimental versus protective effects of racial/ethnic residential isolation and immigrant concentration on health in the United States (US), and empirically examines: (1) whether neighborhood-level ethnic density, racial diversity, and immigrant concentration are negatively or positively associated with health risks; (2) for whom and under what conditions are these residential patterns health-detrimental or health-protective; (3) what are the structural and psychosocial pathways underlying the detrimental or protective effects of these residential patterns; (4) whether neighborhood influences are susceptible to sample selection bias. Individual-level data from the 2003-2008 National Health and Nutrition Examination Survey and the 2006 and 2008 Southeastern Pennsylvania Household Health Survey were merged with census-tract profiles obtained from the 2005-2009 American Community Survey estimates, the 2000 Decennial Census and Geographic Information System-based built-environment data. Multilevel analysis, mediation analysis, and Propensity Score Matching method were performed to answer these research questions. Results largely confirmed the salient impact of neighborhood racial/ethnic context on individual health risks, while selection bias was also evident. Neighborhood racial isolation and ethnic concentration showed detrimental health effects, whereas racial diversity showed positive effects. The observed effect of immigrant concentration was likely due to neighborhood selection bias. Effect modification and underlying pathways were complex and were dependent on the specific neighborhood contextual predictors.