1109-P: The Effect of Neighborhood Food Environment on Diabetes Risk: A Multilevel Analysis of the Veterans Administration Diabetes Risk Cohort

Abstract

Diabetes causes significant morbidity and mortality among adults in the US, yet incidence varies across the country, suggesting that neighborhood factors may contribute to geographic disparities in diabetes. In this longitudinal study, we examined the association between neighborhood food environment and risk of newly diagnosed (“incident”) diabetes in different communities (high density urban, low density urban, suburban, and rural). We created a retrospective dynamic, national cohort of 6,082,018 diabetes free US veterans identified from the Veterans Affairs electronic health records, enrolled between 2008 and 2016, and followed through 2018. Incident diabetes was defined as: 2 clinical encounters with diabetes ICD codes, a documented prescription for diabetes medication other than metformin or acarbose alone, or 1 clinical encounter with diabetes ICD codes and 2 elevated A1C (≥6.5%). Five-year average counts of fast-food restaurants and supermarkets relative to all other food outlets were used to generate two census tract-level neighborhood measures. We examined the association between food environment and incident diabetes using piecewise exponential survival models with county-level random effects. The average age of cohort participants was 59 years; the majority were non-Hispanic white (76.3%) and male (92.2%). Relative density of fast-food restaurants was associated with a modest increased risk of diabetes in all community types; adjusted hazard ratios (aHR) in high density urban, low density urban, and suburban communities were each 1.1, in rural communities was 1.2 (p < 0.05 for all). Relative density of supermarkets was associated with lower diabetes risk in suburban and rural communities only (aHR 0.7, 0.8, respectively, p <0.05). Neighborhood fast food and supermarkets influenced diabetes risk among US veterans in multiple community types, suggesting potential avenues for action to address the diabetes burden. Disclosure R. Kanchi: None. G. Imperatore: None. B. Elbel: None. L. Thorpe: None. P. M. Lopez: None. P. E. Rummo: None. D. C. Lee: None. S. Adhikari: Research Support; Self; Johnson & Johnson. M. D. Schwartz: None. S. Avramovic: None. K. R. Siegel: None. D. B. Rolka: None. Funding Centers for Disease Control and Prevention