Abstract
BACKGROUND: Emerging literature notes the importance of neighborhood-level factors for cancer control behaviors beyond that of individual factors. Markers of neighborhood-level disadvantage have been linked to greater likelihood of nonsalutary cancer control behaviors. There has been less examination of many neighborhood factors simultaneously, which more accurately reflects individuals' daily experiences. We estimated associations of neighborhood deprivation indices with cancer control behaviors, identifying the relative importance of neighborhood-level deprivation index components for these outcomes. METHODS: We used data from the Religion and Health in African Americans study, a national probability sample of African American adults. We separately considered 4 screening and 4 prevention behaviors as outcomes. We constructed neighborhood deprivation indices using census tract-level data and estimated their associations with outcomes using bayesian index models, adjusting for individual-level covariates. We reported odds ratios (ORs), credible intervals, and exceedance probabilities. RESULTS: Participants in our sample engaged in relatively high levels of screening behaviors and lower levels of prevention behaviors. Neighborhood deprivation indices were statistically significantly associated with a greater likelihood of binge drinking (OR = 1.13, exceedance probability = 98.5%), smoking (OR = 1.07, exceedance probability = 99.4%), and insufficient colonoscopy (exceedance probability = 99.9%), Papanicolaou (exceedance probability = 99.7%), and prostate-specific antigen (exceedance probability = 99.1%) screening. Within neighborhood deprivation indices, median household income, percentage of individuals without some college education, and percentage of individuals unemployed received large estimated importance weights. CONCLUSION: We identified statistically significant associations between neighborhood disadvantage and nonsalutary cancer control behaviors as well as important neighborhood-level deprivation index components for each outcome. These and similar findings from future studies should be used to target specific neighborhood factors for specific cancer control behaviors rather than using a one-size-fits-all approach.