Diabetes prevalence has more than doubled in the past two decades, and now nearly 1 in 10 Americans has diabetes. After a similar steep increase, obesity rates among US adults have been stable in the past 10 years, but strong disparities persist and rates are still rising in some groups. Prevention efforts to date have mostly focused on individual-level risk factors. However, a growing body of research has linked diabetes, obesity and other cardiometabolic risk factors to a range of community characteristics, suggesting that environmental factors can have a direct beneficial or deleterious influence on disease risk. Numerous studies have observed significant and plausible associations between diabetes and obesity prevalence and attributes of the social and physical environment. Some of these community factors are amenable to intervention, yet the true magnitudes of their relative contributions, remains unclear, especially after adjusting for population characteristics. In response to FOA-DP-17-001, we aim to collaborate with other funded Centers to examine modifiable community characteristics that may together explain stark disparities in diabetes, obesity and other cardiometabolic conditions in the United States using ecologic and multi-level study designs. Specifically, we aimtousecounty-levelprevalencedatafromtheBehavioralRiskFactorSurveillanceSystem(BRFSS)and individual-levelelectronicmedicalrecorddatafromaverylargenationalVeteran?sAdministration(VA) cohort (~4.6 million patients) to examine the relationship between key community factors pertaining to the food and housing environment and 3 outcomes: diabetes, obesity and discordance between county-level diabetes and obesity prevalence, an approach that allows us to examine unique contextual risk factors for each condition. Expanding on prior work, we propose developing novel community measures that may contribute to diabetes and obesity disparities. Our primary food environment measures to be refined, shared with other Centers, and analyzed pertain to relative ?food swamp? measures (neighborhoods in which cheap, energy- dense and pre-prepared food is a dominant option). Our primary housing environment measures to be refined, shared and analyzed are ?housing cost burden? measures. We will use causal inference techniques to strengthen our multilevel analyses (e.g. instrumental variable analysis and negative control outcomes). We will also perform causal mediation analysis and primary data collection to examine potential mechanistic pathways on our main study datasets. Primary data will be collected through a planned social determinant of health survey in partnership with the New York City Health Department. Capitalizing on (1) the breadth of expertise among New York University researchers in the areas of health surveillance, food and housing policy research, and causal inquiry methods; (2) access to a very large cohort of VA patients; and (3) our demonstrated track record of leadership and collaboration in other CDC-funded networks, this study offers a powerful opportunity to identify community disease determinants for use in policy and systems change to reduce disparities.
Research has linked chronic diseases and their risk factors to community characteristics, suggesting that alteration of social and environmental characteristics may be mutable factors in reducing risk of chronic disease disparities. The proposed application led by NYU School of Medicine aims to collaborate with other funded Centers to examine modifiable community characteristics that may contribute to diabetes and other cardiometabolic disparities in the United States using ecologic, spatial and multi-level study designs. Specifically, we will use county-level data from the Behavioral Risk Factor Surveillance System and individual- level electronic medical record data from a very large Veteran?s Administration cohort (~4.6 million patients) to examine relationships between food and housing environments and 3 outcomes: diabetes, obesity and diabetes-obesity discordance profile.
