A return to the environment in epidemiology: multilevel analysis of diabetes mellitus in a large urban agglomeration of Argentina.

OBJECTIVE: The burden of disease due to diabetes mellitus has shown an increasing trend worldwide. Given the potential of physical structures to create opportunities for the adoption of individual behaviors, this study aimed to analyze the relationship between neighborhood food environments and green spaces and the prevalence of diabetes mellitus in the Mar del Plata-Batán urban agglomeration in Argentina, while also considering individual characteristics, from 2013 to 2018. METHODS: A cross-sectional design with a multilevel approach was used. Individual data from the National Risk Factor Survey were linked to data on food environments and green spaces aggregated at the census tract level. Multilevel logistic regression models were used, with individuals nested within census tracts; odds ratios were estimated for independent variables with 95% confidence intervals. RESULTS: A total of 946 observations distributed across 81 census tracts were analyzed. After adjustment for individual variables, living in census tracts with a high density of healthy food outlets was associated with a 43% reduction in the odds of having diabetes [OR 0.57 (0.35-0.93)]. Public green space area per inhabitant in the neighborhood was not significantly associated with the odds of diabetes. The proportion of variability attributable to the contextual level ranged from 2.35 to 1.20%. CONCLUSIONS: The neighborhood food environment was associated with diabetes mellitus in the studied agglomeration. These results may encourage the development of longitudinal studies with greater power to demonstrate the causal role of the food environment in the occurrence of diabetes.