Methods: We obtained data from the Health & Retirement Study. We included respondents aged 50 years and older who had no recorded history of CVD as of 2006 (N=8,826). The respondents were followed for 10 years (2006-2016). Multilevel Cox proportional hazards models were estimated with CVD incidence as the outcome variable and time-variant social environment factors (i.e., perceived neighborhood social cohesion, perceived neighborhood physical disorder, and county-level crime rates) as exposures, after controlling for sociodemographic factors (i.e., age, sex, education, and family income), CVD-related risk/protective factors (i.e., alcohol consumption, smoking status, depression symptoms, moderate-to-vigorous physical activity level, hypertension, and diabetes), and neighborhood socioeconomic status. To account for differential neighborhood effects by race/ethnicity shown in previous literature, we stratified the data by race/ethnicity in all analyses.
Results: Perception of neighborhood social cohesion had a significant inverse association with CVD incidence for non-Hispanic Blacks (HR=0.80, 95% CI=0.68-0.95) and Hispanics (HR=0.80, 95% CI=0.64-0.99) when adjustment for age, sex, education, family income, and neighborhood socioeconomic status. Additional adjustment for CVD-related risk/protective factors attenuated the magnitude of the association, but the association remained statistically significant (Blacks: HR=0.83, 95% CI=0.71-0.98; Hispanics: HR=0.81, 95% CI=0.65-1.00). Perception of neighborhood physical disorder was positively associated with CVD incidence for non-Hispanic Blacks (HR=1.19, 95% CI=1.02-1.39) and Hispanics (HR=1.29, 95% CI=1.04-1.60) when adjustment for age, sex, education, income, and neighborhood socioeconomic status. The association remained significant among both non-Hispanic Blacks (HR=1.18, 95% CI=1.01-1.38) and Hispanics (HR=1.28, 95% CI=1.03-1.58) after additional adjustment for behavioral and biomedical factors. Local crime was not significantly associated with CVD incidence for a any of the three racial/ethnic groups.
Conclusions/Implications: The results showed the importance of perceived neighborhood social cohesion and physical disorder in CVD incidence among non-Hispanic Blacks and Hispanics, but not non-Hispanic Whites. The findings suggest that efforts to intervene on perceived social environments might benefit Black and Hispanic adults’ cardiovascular health and reduce health disparities. Further research is needed to identify pathways through which Blacks and Hispanics benefit from living in favorable social environments and the reason(s) why the benefits do not accrue among Whites.