Home delivery of healthy groceries for 6 months significantly improved blood sugar levels among individuals with type 2 diabetes (T2D) independent of any changes in medication, results of a new study showed.

“Medically tailored groceries have generated tremendous interest among health systems, policymakers, and payers as a way to address food insecurity and support chronic disease management,” said lead author Claudia Nau, PhD, a researcher with the Kaiser Permanente Southern California Department of Research & Evaluation.

The idea is intuitive, she added. “If we want patients to manage conditions like diabetes, it’s important not only to prescribe medications but also to ensure they have access to healthy foods,” she said.

The findings of the study were published in Circulation.

Although “Food Is Medicine” initiatives are emerging as ways to address these barriers, data on their effectiveness are limited, the researchers noted.

Nau and colleagues enrolled 460 Medicaid-insured adults in Southern California with T2D between November 2021 and July 2022. Participants had elevated A1c based on at least two measurements ≥ 7.5% in the past year.

The mean age of the participants was 59.2 years, 64.8% were women, and 85.2% identified as Hispanic; 50.2% did not have a high school degree. Mean A1c at baseline was 9.40, and mean BMI was 34. Approximately 90% were prescribed at least one oral diabetes medication, two thirds were on insulin, and 75.8% had high blood pressure.

The intervention consisted of weekly home delivery of healthy foods selected to help improve diabetes control, along with phone-based nutritional counseling and an option for meal planning and dietary assessment. Participants were randomly assigned in a 1:1:1 ratio to usual care (n = 153) or lower (n = 153) or higher volume (n = 154) of grocery delivery. The primary outcome was the change in A1c from baseline to 6 months.

The multivariate analysis included age, sex, race, ethnicity, number of household members, self-reported health, Elixhauser comorbidity index, BMI, hypertension, food insecurity, nutrition insecurity, enrollment month, and change in diabetes medications.

Complete data were available for 438 patients, 293 who received the intervention (low- or high-volume grocery delivery) and 145 who received usual care.

After 6 months, A1c decreased by 0.66 points in the overall intervention group and 0.25 points in the control group compared to baseline, a significant treatment difference of -0.40 points (95% CI, -0.73 to -0.08; P = .016). Similar reductions were seen in both low- and high-volume intervention groups.

In addition, the odds of food security and nutrition security increased significantly from baseline in the intervention group. The treatment difference between the intervention group and the control group for food security was 2.12 (95% CI, 1.13-3.99; P = .020); for nutrition security, the difference was 3.65 (95% CI, 1.84-7.25; P < .001).

A total of 244 (83.3%) participants in the intervention group reported eating most or all of the food provided, and 63 (21.5%) participated in telenutrition counseling. No significant changes were observed in the secondary outcomes of hypertension or BMI, and no adverse events were reported.

The extent of the baseline food and nutrition insecurity and the significant improvement was striking, said Nau. This highlights the need for whole person care beyond what is typically considered medical care.

“Most of what makes us healthy or unhealthy does not happen in the doctor’s office; it happens at home,” she emphasized. “Our findings suggest that providing healthy foods as part of routine diabetes care can help patients improve blood sugar control.”

The results were limited by several factors, including the inability of participants to choose specific food items within curated categories, the inability to account for the possible impact of medication changes, and the reliance on electronic health records for data on secondary outcomes of BMI and hypertension, the researchers noted.

“However, the results support the potential of programs that address nutrition and food access to complement traditional medical treatment for patients with diabetes,” Nau said. “Future research should identify which program designs are most effective, scalable, and sustainable,” she added.

Evidence of Benefits May Promote Wider Adoption

“Food-as-medicine programs are increasingly being piloted, adopted, and evaluated despite limited evidence from randomized trials,” said Charles E. Leonard, PharmD, MSCE, MPH, associate professor of epidemiology at the Perelman School of Medicine at the University of Pennsylvania in Philadelphia, who was not involved in the study.

The notable finding in the current study was the modest but potentially clinically meaningful improvement in A1c, especially in the wake of a recent systematic review that showed inconsistent evidence for the ability of food is medicine interventions to improve glycemic outcomes, he said.

“Improving access to healthy food may complement evidence-based diabetes care, particularly for disadvantaged patients facing barriers to obtaining such foods, but major barriers of reimbursement, sustainability, and scale remain,” Leonard noted.

Broader adoption of an intervention such as home grocery delivery will likely depend on demonstrating meaningful long-term benefits for patients, health systems, and payers, he explained. Looking ahead, more research is needed to show whether short-term improvements in a laboratory-based measure translate into fewer diabetes complications, hospitalizations, and healthcare expenditures over time and which patients are most likely to benefit, he added.

Disclosure information for the authors is available in the original study publication.

Leonard disclosed consulting for Novo Nordisk, Biohaven, Moderna, TriNetX, SciPinion, Cure, and Shine Lawyers and receiving honoraria from the National Institutes of Health, the University of Florida, ”la Caixa” Foundation, and the American College of Clinical Pharmacy Foundation. His spouse reported being employed by Johnson and Johnson. Neither Leonard nor his spouse reported owning stock in Johnson and Johnson.