Association Between Diabetes Management Behaviors and State-Level Social Vulnerability: A Cross-sectional Study.

Introduction: Type 2 diabetes mellitus remains a major public health concern in the United States. Previous research demonstrates that social determinants of health, including inequities in living conditions, directly affect the biological and behavioral outcomes associated with diabetes control and prevention. New tools such as the social vulnerability index (SVI) can be used to assess these inequities. This paper aims to assess the impact of state-level social vulnerability on diabetes management behaviors.

Methods: We used data from the Centers for Disease Control/Agency for Toxic Substances and Disease Registry's Social Vulnerability Index (SVI) and the Behavioral Risk Factor Surveillance System (BRFFS) Survey. The final sample included data from 2016 to 2021 and 89,643 people diagnosed with type 2 diabetes. We averaged the SVI and its component indices for each state and categorized them into terciles. The outcome measures included whether the respondents received diabetes education, conducted self-monitoring of blood glucose levels, and received medical care for diabetes. We ran logistic regressions to test the impact of the SVI on the outcomes, adjusting for state and fixed time-fixed effects and covariates. We weighted according to BRFFS guidelines.

Results: Individuals in states with moderate social vulnerability were 1.87 times more likely to receive diabetes education than those in states with the lowest vulnerability (p < 0.01). Conversely, those in states with the highest vulnerability were 1.48 times more likely to see a doctor for diabetes care than those in the least vulnerable states (p =  < 0.01). However, individuals in states with moderate vulnerability are less likely to self-monitor their blood sugar levels (OR = 0.74, p = 0.03).

Conclusions: People in states with different levels of social vulnerability exhibit differences in behaviors affecting diabetes management, particularly in activities not captured by standard diabetes quality measures. Differences are more apparent in activities that are not part of diabetes quality measures.