A year-long trial of produce subsidies shows that access alone may not be enough to shift cardiometabolic outcomes, raising new questions about how Food is Medicine programs should be designed and targeted.
Study: Produce Prescription Subsidy for Patients With Diabetes. Image credit: Pressmaster/Shutterstock.com
A recent paper published in JAMA Internal Medicine examined health outcomes following the implementation of a produce prescription subsidy program in patients with T2D at risk of food insecurity.
Food access shapes diabetes outcomes nationwide
T2D affects over 30 million people in the United States, and its clinical outcomes are strongly influenced by the availability and accessibility of nutritious food, often described as food security. Food insecurity occurs when nutritious food is not readily accessible due to socioeconomic factors, leading individuals to rely more heavily on inexpensive, energy-dense foods that are often less nutritious.
In response, public health policymakers have promoted the concept of “Food is Medicine” (FIM), which includes the use of produce prescription (PRx) programs to improve access to medically appropriate foods. These programs have gained traction among healthcare systems and policymakers in part because they offer participants flexibility in food choice while posing fewer logistical challenges than prepared-meal delivery programs.
Despite growing enthusiasm, strong evidence supporting the effectiveness of PRx programs remains limited. At least one prior randomized trial found no significant benefit, and methodological constraints in earlier studies have made it difficult to draw firm conclusions about the impact of Food is Medicine interventions.
The current study sought to evaluate whether a PRx subsidy could improve cardiometabolic health outcomes and healthcare use among patients with T2D identified as at risk for food insecurity.
Monthly produce subsidy tested over 12 months
This pragmatic randomized clinical trial enrolled 2,177 participants, of whom 2,155 were included in the final analysis. Eligible participants had T2D and were considered at risk for food insecurity based on prior food insecurity screening responses, Medicaid insurance status, or residence in census block groups with high area deprivation index scores. The average age of participants was 56 years, 71 % were female, 61 % were non-Hispanic Black, 32 % were non-Hispanic White, and 5.3 % were Hispanic.
Participants were stratified according to baseline hemoglobin HbA1c levels, using a threshold of less than 8 % versus 8 % or greater. Within each stratum, participants were randomized in a 2:1 ratio to receive either the intervention or usual care.
Those in the intervention group received a restricted-use debit card with $80 per month in spending for 12 months to purchase eligible fruits, vegetables, and legumes. Eligible canned and frozen items were required to contain no added sugars, salt, or fats, and unused funds did not carry over from one month to the next. The control group received usual care. Both groups were provided with diabetes self-management educational materials.
The primary outcomes were changes in HbA1c levels and the likelihood of any emergency department visits over the 12-month follow-up period. Secondary outcomes included body mass index, blood pressure, and inpatient hospital visits.
Produce subsidy fails to improve HbA1c
At baseline, the mean HbA1c was 7.49 %. Over the course of the study, adherence to the subsidy varied. Approximately 30 % of participants in the intervention group used 80 % or more of their subsidy for at least ten months, 14 % used between 60 % and 79 %, and 36 % used less than 60 %. Notably, 21 % of those assigned to the intervention never used the subsidy, largely because they did not complete enrollment procedures.
After 12 months, the adjusted between-group difference in mean HbA1c was 0.20 percentage points in favor of the control group. Although this difference reached statistical significance, it was not considered clinically meaningful. There were no significant differences between the intervention and control groups in the likelihood of any emergency department visits, inpatient admissions, blood pressure measurements, or body mass index. Among participants with baseline HbA1c levels of 8 % or higher, results were similar, with an adjusted between-group difference of 0.39 percentage points again favoring usual care.
Overall, the findings indicate that a produce prescription subsidy alone did not result in clinically meaningful improvements in cardiometabolic outcomes or healthcare utilization among patients with T2D at risk of food insecurity. The authors emphasized that these results do not necessarily demonstrate that PRx programs are ineffective. Instead, they suggested that the absence of benefit may reflect limited intervention intensity, imperfect targeting of individuals with confirmed food insecurity or cost-related barriers, and suboptimal enrollment and adherence.
The study’s findings are consistent with other research showing the limited impact of standalone grocery subsidy programs, while more comprehensive Food is Medicine interventions that incorporate social needs screening and care management, such as North Carolina’s Healthy Opportunities Pilots program, have been associated with reduced emergency department utilization.
Strengths and Limitations
This trial was notable for its large sample size, pragmatic design, and use of clinically observable outcomes within routine care settings. However, measuring outcomes through routine clinical encounters may have introduced variability in data collection. Hispanic participants were underrepresented, potentially limiting generalizability. In addition, the study did not directly measure household food insecurity status during follow-up, diet quality, or patient-centered outcomes such as quality of life.
The authors also noted that identifying participants based on risk proxies rather than confirmed food insecurity may have diluted potential effects. Furthermore, the fixed subsidy amount was not scaled to household size, and the intervention did not include additional behavioral or care-management support, as present in some more successful programs.
Food Is Medicine may require broader support
In this large pragmatic trial, an $80-per-month produce prescription subsidy did not yield clinically meaningful improvements in glycemic control or healthcare use among patients with type 2 diabetes who were at risk for food insecurity.
The findings suggest that standalone produce subsidy programs may be insufficient to meaningfully alter cardiometabolic outcomes, and that future research should explore how Food is Medicine interventions can be better targeted and integrated into broader strategies that address the structural drivers of food insecurity and health inequities.
Download your PDF copy clicking here.
Blood metabolites reveal who will develop type 2 diabetes years before diagnosis