Neighbourhood characteristics and mortality in the Atherosclerosis Risk in Communities Study.

BACKGROUND This study investigates the relationship between neighbourhood characteristics and mortality (all-cause, cardiovascular disease [CVD], and cancer) in the Atherosclerosis Risk in Communities Study (ARIC). METHODS Analysis was limited to African-American and white participants 45-64 years of age at baseline whose records were linked to census data. Deaths ascertained through 31 December 1999 were included in the analysis. Individual-level characteristics were obtained from the baseline interview. A composite index was used to characterize the neighbourhood socioeconomic environment. Proportional hazards regression was used to estimate the effect of neighbourhood socioeconomic status (SES) index and family income on the survival time. RESULTS The rate of mortality adjusted for age and gender was highest among those who lived in disadvantaged neighbourhoods and were of lower SES. In general, all-cause and CVD mortality rates decreased with increasing neighbourhood SES advantage and family income in all race-gender groups. Although this pattern generally persisted after adjustment for individual socioeconomic factors, statistically significant associations persisted for CVD mortality in whites only (hazard ratio = 1.4, 95% CI: 1.0, 2.0) for most disadvantaged versus most advantaged tertile). When compared with the most affluent participants living in the most advantaged neighbourhoods, the increased risk of all-cause and CVD mortality associated with being poor and living in the most disadvantaged neighbourhoods was equivalent to being 11 and 13 years older at baseline for whites and African Americans, respectively. CONCLUSION Our findings indicate that neighbourhood socioeconomic characteristics are associated with modest increases in CVD mortality in white adults. The lack of neighbourhood effects in African Americans needs to be interpreted with caution due to the limited range in the characteristics of the neighbourhood from which these participants were drawn.