Abstract
Many health risk factors are intervention targets within communities, but information regarding high-risk subpopulations is rarely available at a geographic resolution that is relevant for community-scale interventions. Researchers and community partners in New Bedford, Massachusetts (USA) collaboratively identified high-priority behaviors and health outcomes of interest available in the Behavioral Risk Factor Surveillance System (BRFSS). We developed multivariable regression models from the BRFSS explaining variability in exercise, fruit and vegetable consumption, body mass index, and diabetes prevalence as a function of demographic and behavioral characteristics, and linked these models with population microdata developed using spatial microsimulation to characterize high-risk populations and locations. Individuals with lower income and educational attainment had lower rates of multiple health-promoting behaviors (e.g., fruit and vegetable consumption and exercise) and higher rates of self-reported diabetes. Our models in combination with the simulated population microdata identified census tracts with an elevated percentage of high-risk subpopulations, information community partners can use to prioritize funding and intervention programs. Multi-stressor modeling using data from public databases and microsimulation methods for characterizing high-resolution spatial patterns of population attributes, coupled with strong community partner engagement, can provide significant insight for intervention. Our methodology is transferrable to other communities.