Effects of a “Food Is Medicine” Intervention on Glucose Control Among Medicaid-Insured Patients With Type 2 Diabetes: A Randomized Controlled Trial

BACKGROUND: Although medically tailored groceries (MTG) hold promise to improve food and nutrition security and health among patients with diet-related illnesses and social needs, few randomized trials have been performed. METHODS: We enrolled 460 Medicaid-insured adults with type 2 diabetes and elevated hemoglobin A1c (HbA1c) between November 2021 and July 2022 in a randomized controlled trial of MTG in southern California. Eligibility was based on at least 2 HbA1c measurements ≥7.5% in the past year. Participants were randomized 1:1:1 to usual-care (N=153), lower-dose MTG (N=153), or higher-dose MTG (N=154) for 6 months, with the prespecified primary assessment comparing the combined MTG groups with control. MTG was provided as weekly home deliveries of healthy produce, scaled to household size ($100 to $170 per month for the lower dose and $135 to $210 per month for the higher dose), along with matched recipes and telenutrition counseling. The primary outcome was a change in HbA1c at 6 months. Secondary outcomes included changes in food security and nutrition security, measured by surveys, and hypertension and body mass index, measured during clinical care encounters; evaluation of dose-response comparing the lower-dose and higher-dose arms; and effects in key population subgroups. RESULTS: At baseline, mean (SD) age was 59.2 (13.2) years, 284 (64.8%) were women, 373 (85.2%) reported Hispanic ethnicity, 254 (58%) reported food insecurity, and mean (SD) HbA1c was 9.40 (1.53). Among intervention participants, 244 (83.3%) reported eating most or all food provided, and 63 (21.5%) engaged in telenutrition counseling. Compared with baseline, HbA1c declined by 0.66 points in the intervention and 0.25 points in the control group, a treatment difference of –0.40 points (95% CI, –0.73, –0.08; P =0.016). Higher-dose and lower-dose MTG produced similar reductions. Odds of food security and nutrition security increased by 2.12 (95% CI, 1.13, 3.99; P =0.020) and 3.65 (95% CI, 1.84, 7.25; P <0.001), respectively. Hypertension and body mass index did not significantly change. Results were consistent in prespecified subgroup analyses by sex, education, baseline food security, baseline nutrition security, and baseline HbA1c level. Findings were similar in analyses adjusting for baseline demographics, food insecurity, nutrition insecurity, self-reported health, comorbidities, and medication changes. CONCLUSIONS: In this randomized trial among Medicaid-insured, racially/ethnically diverse patients with type 2 diabetes, MTG lowered HbA1c and improved food and nutrition security. REGISTRATION: URL: https://clinicaltrials.gov ; Unique identifier: NCT05407376.