Neighborhood Environments and Cardiovascular Disease

BACKGROUND:

There is abundant evidence of persistent differences in cardiovascular disease morbidity and mortality by socioeconomic status (SES). The determinants of SES-related differences in CVD outcomes and risk factors have not been fully established. Previous work in this area has focused predominantly on individual-level SES indicators, but recently attention has shifted to the role of neighborhood or community-level variables in shaping health outcomes, independently of individual-level SES. Several epidemiologic studies have suggested that neighborhood characteristics may influence the distribution of disease risk, but the role of both neighborhood-level and individual level SES variables in shaping individual-level outcomes and risk factors has been rarely addressed in epidemiologic studies of CVD.

DESIGN NARRATIVE:

Associations of neighborhood socioenvironmental characteristics with CVD prevalence and incidence in middle-aged and elderly populations were investigated using data from the ARIC Study and CHS. Associations of neighborhood socioenvironmental characteristics with CVD risk factors and risk factor trends in young and middle-aged adults were investigated using data from the CARDIA and ARIC studies. CARDIA and ARIC data were also used to explore the contributions of neighborhood characteristics to racial differences in CVD risk factors. Census defined areas were used as proxies for neighborhoods. Participants were linked to their census-tract and block-group of residence using their home address, and neighborhood characteristics were obtained from the 1990 U.S. Census. The three data sets were analyzed separately. After exploratory and descriptive analyses, regression models were used to investigate associations of neighborhood characteristics with the outcomes before and after controlling for individual-level SES and other relevant covariates. Appropriate statistical methods (mixed effects models) were used to account for the multilevel structure of the data (individuals nested within neighborhoods and repeated measures nested within individuals), and the potential violations of the assumption of independence of observations that might arise from it.