Objective: To characterise the association between risk of poor glycaemic control and self-reported and area-level food insecurity among adult patients with type 2 diabetes.
Design: We performed a retrospective, observational analysis of cross-sectional data routinely collected within a health system. Logistic regressions estimated the association between glycaemic control and the dual effect of self-reported and area-level measures of food insecurity.
Setting: The health system included a network of ambulatory primary and speciality care sites and hospitals in Bronx County, NY.
Participants: Patients diagnosed with type 2 diabetes who completed a health-related social need (HRSN) assessment between April 2018 and December 2019.
Results: 5500 patients with type 2 diabetes were assessed for HRSN with 7·1 % reporting an unmet food need. Patients with self-reported food needs demonstrated higher odds of having poor glycaemic control compared with those without food needs (adjusted OR (aOR): 1·59, 95 % CI: 1·26, 2·00). However, there was no conclusive evidence that area-level food insecurity alone was a significant predictor of glycaemic control (aOR: 1·15, 95 % CI: 0·96, 1·39). Patients with self-reported food needs residing in food-secure (aOR: 1·83, 95 % CI: 1·22, 2·74) and food-insecure (aOR: 1·72, 95 % CI: 1·25, 2·37) areas showed higher odds of poor glycaemic control than those without self-reported food needs residing in food-secure areas.
Conclusions: These findings highlight the importance of utilising patient- and area-level social needs data to identify individuals for targeted interventions with increased risk of adverse health outcomes.
Keywords: Community health; Diabetes mellitus; Disease management; Food insecurity; Primary health care; Social needs.