Evaluating a medicaid policy to provide medically tailored groceries to patients with type 2 diabetes

The National Institute of Diabetes and Digestive and Kidney Diseases awarded Harvard Pilgrim Health Care, Inc. $827,060 on August 21, 2026, under the Diabetes, Digestive, and Kidney Diseases Extramural Research program (CFDA 93.847) to evaluate California's Medicaid medically tailored grocery program for adults with type 2 diabetes.

The project will conduct a rigorous randomized encouragement design study enrolling 600 adults with type 2 diabetes who are Medicaid participants and regular customers at Albertsons-owned stores across three California counties. Participants will be randomly assigned to receive encouragement to apply for the medically tailored grocery program, which provides $35–$50 in groceries (7–10 meals) weekly, or to a control group receiving nutrition education. At baseline and six-month follow-up, all participants will complete three 24-hour dietary recalls to measure Healthy Eating Index scores and assess changes in dietary quality and metabolic health. After the six-month follow-up, the control group will also be invited to participate in the program. The research will use instrumental variable analyses to determine whether the medically tailored grocery program improves dietary quality and household grocery purchasing patterns.

The award period runs through June 30, 2031, with place of performance in Boston, Massachusetts. Albertsons Companies provides program implementation and food delivery as a partner organization.

Abstract

Since 2021, california's medicaid program-the largest in the country-has offered food is medicine programs to members with chronic diseases. These programs could improve the diet quality and metabolic health of adults with type 2 diabetes (T2D), especially those with lower income who face barriers to accessing and affording healthy food. Medicaid managed care organizations (mcos) in several california counties participate in this initiative and offer a medically tailored grocery (MTG) program that provides patients with $35-$50 worth of groceries (7-10 meals) each week. Program implementation and food delivery is provided by albertsons companies, the second-largest supermarket company in the US. We will partner with albertsons to rigorously evaluate the effects of this real-world MTG program on diet quality, metabolic health, and healthfulness of household grocery purchases among adults with T2D using a randomized encouragement design. We will recruit 600 adults with T2D who self-report medicaid participation and are regular customers at albertsons-owned stores in 3 california counties with an MCO that offers the MTG program. We will randomly select half of them to receive encouragement to apply to participate in the MTG program and the other half to a control group (nutrition education). After 6 months, when follow-up diet and metabolic health measures have been completed in both groups, we will also encourage the control group to apply to participate in the program. In aim 1, we will collect three 24-hour dietary recalls from all 600 participants before the MTG program begins (baseline) and another three 24-hour recalls at 6-month follow-up. We will calculate healthy eating index (HEI) scores, a measure of overall dietary quality, at each time point. We will use instrumental variable analyses to determine if the MTG program improved HEI scores (primary outcome) as well as food security status, nutrition knowledge, and self- efficacy to eat healthfully (secondary outcomes) versus control. In aim 2, we will leverage albertsons' extensive network of pharmacies to collect measures of glucose control (HBA1C) and weight (CO-primary outcomes) and blood pressure (secondary outcome) at baseline and 6-month follow-up from all 600 participants. We will use instrumental variable analyses to determine if the MTG program led to improvements in these health measures. In aim 3, albertsons will provide US with longitudinal grocery purchase data from all trial participants who enrolled in the MTG program from 12 months before program initiation through 18-month follow-up. Households will be identified by unique customer loyalty ID. We will use interrupted time series analysis, a quasi-experimental method, to estimate the long-term effect of MTG program participation on ofcom nutrient profile model scores, a measure of the healthfulness of grocery purchases. This research will provide one of the first rigorous evaluations of a real-world MTG program for adults with T2D, offered through the nation's largest medicaid program. This evidence is critical to inform active discussions about food is medicine programs.