Produce Prescription Subsidy in Diabetes: Pragmatic Trial Results

Key Takeaways

• At 12 months, the produce prescription subsidy was not associated with better glycemic outcomes than usual care.
• No between-group differences were reported in emergency department visits, blood pressure, body mass index, or inpatient visits.
• Results were similar in participants with HbA1c levels of 8% or higher, and monthly subsidy use was moderate.

In a pragmatic randomized trial, adults with diabetes at risk for food insecurity had no better 12-month outcomes with the produce prescription subsidy than with usual care. The adjusted between-arm difference in HbA1c was 0.20 percentage points, with a 95% CI of 0.05 to 0.35, favoring usual care. The subsidy also was not associated with better cardiometabolic measures or health care utilization outcomes at 12 months in this health system program. Both groups received diabetes self-management educational materials.

This 2-arm pragmatic randomized clinical trial was conducted in an integrated academic health system in the southeastern US, with recruitment from June to August 2023 and 12-month follow-up after randomization. Among 9,608 patients assessed for eligibility, 2,155 consented, were randomized, and were studied after meeting criteria of diabetes plus risk for food insecurity. Participants were assigned to PRx or usual care, and both groups received diabetes self-management educational materials. The PRx group also received an $80 monthly debit card for up to 12 months for eligible fresh, frozen, or canned fruits, vegetables, and legumes at grocery retailers. Participants had a mean age of 56 years, 71% were female, and the prespecified primary outcomes were HbA1c and emergency department visits at 12 months.

At 12 months, the adjusted HbA1c difference favored usual care rather than the subsidy group in the between-arm analysis. There were no significant between-group differences in emergency department visits at one year, the other prespecified primary outcome. Researchers also found no significant differences in blood pressure, body mass index, or inpatient visits across the randomized groups. Overall, outcomes did not shift toward improved cardiometabolic status or lower health care use with PRx during follow-up. The investigators concluded that the PRx program did not improve cardiometabolic health or health care utilization over 12 months.

Findings were similar in the subgroup of 651 participants with elevated HbA1c levels (≥8%), with no clear subgroup effect. The cohort was 61% non-Hispanic Black, 32% non-Hispanic White, and 5.3% Hispanic, with a mean baseline HbA1c of 7.48%. Subsidy uptake was moderate, with 433 participants, or 30%, using 80% or more of the monthly benefit. Randomization was stratified by prior mean HbA1c of 8% or higher versus lower than 8%, and data were analyzed from October 2024 to April 2025.