Medically Tailored Meals Work. Medicaid Work Requirements May Undercut Them.

When Massachusetts started delivering home-cooked, dietitian-tailored meals to its sickest Medicaid patients in 2020, the goal was simple: keep people out of the hospital by feeding them properly. Five years of data later, the evidence suggests it worked — and by a margin that is rare in public health. People who received medically tailored meals were hospitalized 31 percent less often [Nature Medicine] and landed in the emergency room 20 percent less than similar patients who didn't [Nature Medicine]. Their per-person healthcare costs dropped by $3,433 during the enrollment period [Nature Medicine]. The cost of the meal program itself was almost entirely offset — 98 percent — by those savings [Nature Medicine]. For patients with heart disease, kidney disease, depression, or diabetes, the meals didn't just improve health. They saved the state money.

That is the finding published Monday in Nature Medicine by researchers at UMass Chan Medical School, the Tufts Friedman School, and Community Servings, based on the largest statewide analysis of medically tailored meals ever conducted inside Medicaid. The study tracked 1,866 meal recipients against 1,372 comparable patients across 11 Massachusetts health systems from 2020 through 2023 [Nature Medicine]. It is the kind of result that makes health economists take notice, because cost-saving interventions inside Medicaid are not common. "It's rare to find anything in medicine that both improves health and saves money," said Dariush Mozaffarian, director of the Food is Medicine Institute at Tufts and a senior author on the study [Nature Medicine]. "It should be a no-brainer to extend similar programs to patients in other states."

The same morning that paper appeared in a journal, STAT News Morning Rounds also carried news of the Trump administration's finalized Medicaid work requirements. The rules, which Congress mandated in last summer's tax bill and which CMS guidance confirmed in December, require adults enrolled through the ACA Medicaid expansion and certain waiver programs to work or volunteer 80 hours per month starting January 1, 2027 [CMS]. The Congressional Budget Office projects those requirements will push 5.3 million people off Medicaid over the next decade [KFF].

The collision is not subtle. The Massachusetts data suggests that nutrition programs inside Medicaid — which serve people with diet-sensitive chronic conditions and food insecurity — keep patients healthier and reduce the expensive acute care that drives Medicaid spending. The work requirements, as framed by HHS Secretary Robert F. Kennedy Jr. and CMS Administrator Dr. Mehmet Oz, are about restoring the dignity of work and lifting people out of poverty [CMS]. Neither CMS nor HHS responded to questions about whether MTM participants would qualify for any exemption, or whether completing meal delivery rounds could count toward the 80-hour requirement. The law exempts people who are medically frail, pregnant, or caring for children under 14. It does not, based on current guidance, explicitly exempt patients enrolled in nutrition programs.

The practical concern is enrollment continuity. The Massachusetts study found that savings materialized primarily after 90 days of enrollment — adults with more than three months in the program had $2,502 less in healthcare expenditures compared to non-participants [Mass.gov]. Shorter enrollments, or churn caused by coverage gaps triggered by work documentation failures, could eliminate the savings window. The authors note that Food is Medicine programs can stabilize patients during acutely challenging periods, but stabilization requires sustained enrollment, and 80 hours of monthly documentation requirements create exactly the kind of administrative friction that causes churn among low-income patients.

Massachusetts, under a Section 1115 Medicaid waiver, built its nutrition program inside accountable care organizations — groups of providers who accept fixed budgets for patient populations and have financial incentive to invest in preventive care. That structural alignment is part of why the program worked, and part of why replicating it elsewhere is not automatic. Other states operating under the new work requirements will be running nutrition pilots inside a more turbulent enrollment environment, with less certainty about whether patients will stay covered long enough for the investment to pay off.

The study has real limitations. Participants were not randomly assigned; people who enrolled in meal programs may have been more motivated to engage with their health than those who didn't. The researchers used propensity overlap weighting and sensitivity analyses to address this, and the results held across multiple statistical approaches. But an independent audit of the underlying data has not been published. The authors acknowledge the limitation openly. For policymakers deciding whether to embed nutrition programs into Medicaid contracts, it is the best evidence yet — not the last word.

What the collision between these two policy directions makes clear is that the argument for medically tailored meals is no longer primarily about compassion or food security. It is about hospital beds and emergency rooms and the per-person cost curves that state Medicaid directors have to manage. Whether that argument survives contact with a work-requirement regime designed around labor-force participation rather than health outcomes is a different question — and one that the data, so far, has not answered.