In Massachusetts, nearly 1,900 Medicaid members with diabetes, heart disease, and depression received a radical form of health care: 10 dietitian-designed meals delivered to their homes each week. Now, the first large statewide analysis reveals what researchers have long suspected: food really is medicine.
The study, published in Nature Medicine, found that Medicaid participants who received these medically tailored meals—customized home-delivered meals prepared for people with diet-sensitive conditions—experienced 31% fewer hospitalizations and 20% fewer emergency department visits compared to similar eligible members who did not receive the meals. The program not only improved health outcomes; it also produced net cost savings for Medicaid for people with conditions including heart disease, chronic kidney disease, diabetes, and depression.
The research team, led by cardiologist Dariush Mozaffarian at Tufts University's Food is Medicine Institute, analyzed Medicaid claims data from 2020 to 2023 across 11 health care systems in Massachusetts. They carefully compared 1,866 people who received meals with similar eligible members who didn't, accounting for differences in demographics, health conditions, and prior health care use. All meals were prepared and delivered by Community Servings, a Boston-based nonprofit. Participants received meals for an average of about six months, though the program lasted anywhere from three to 33 months depending on individual need.
Each participant started with an initial consultation with a registered dietitian nutritionist who tailored the meal plan to their specific medical needs and dietary preferences. Participants then received a weekly mix of breakfasts, lunches, dinners, and snacks—a concrete, practical intervention for people who often face both food insecurity and activity limitations that make meal preparation difficult.
What makes these findings particularly striking is the speed of improvement. Reductions in hospitalizations and emergency visits occurred within months while participants were actively receiving meals, suggesting the program's effects are not incremental but immediate. Importantly, the program did not reduce necessary care, such as primary care visits—it eliminated the costly emergencies without cutting people off from appropriate medical supervision.
The researchers ran multiple statistical checks to confirm their findings, including analyzing data from before the meal program began to ensure that differences between groups were not already present. The results held across all these approaches. Medicaid patients who received meals for longer periods experienced the largest improvements in health care costs, indicating a dose-response relationship: more months of medically tailored meals correlated with greater savings.
As the first state to broadly offer medically tailored meals in Medicaid, Massachusetts provided what researchers call "an important opportunity to evaluate the real-world impact of such a program." Now, at least a dozen other U.S. states are rolling out similar pilot projects, suggesting the model could reshape how we think about health care for the 71 million Americans covered by Medicaid. For people living with diet-sensitive conditions and the systems trying to serve them, this finding arrives as both validation and invitation: the most powerful medicine might be sitting right on the dinner table.