On April 10, 2026, The New York Times published a feature asking a question that would have seemed fringe a decade ago: can food actually be medicine? Reporter Kim Severson's answer, drawn from researchers, clinicians, and patients across the country, is that the evidence is no longer ambiguous. Medical schools are adding culinary training to their curricula. Thirteen states have started using Medicaid dollars to pay for medically tailored meals. And patients who receive food-based interventions are showing up in the data with meaningfully better outcomes. If you're a Project FoodBox member, you're already part of what that reporting describes.
The Times piece profiles Chuck Self, a retired Boston police officer managing Type 2 diabetes and heart disease who was facing amputation when his doctor prescribed medically tailored meals and groceries. The food, combined with other treatments, helped him lose weight and reduce the number of medications he takes. His story is not an outlier — it reflects what controlled research is finding at scale. Severson notes that one study cited in the article showed a 16 percent reduction in monthly healthcare costs and a nearly 50 percent reduction in hospital admissions among people who consistently ate medically tailored meals. These are the kinds of outcomes that move health systems and insurers, and they are increasingly the basis on which programs like Project FoodBox are funded and expanded.
The peer-reviewed findings from Project FoodBox's own partnership with UC Irvine Health follow the same pattern. Participants managing diabetes saw average A1C levels fall from 8.5% to 7.5% — a one-point drop that corresponds to roughly a 40% reduction in the risk of diabetes complications, including kidney damage, nerve damage, and vision loss. That is a clinically significant result by any standard measure. Separately, a 2025 program survey of more than 3,000 members found that participants reported eating 1.17 fewer fast-food meals per week (p<.001) and a measurable reduction in weekly symptom frequency (p<.01) — gains that held through the end of the program. The UCI Health findings are peer-reviewed; the survey results are program-reported. Both point in the same direction.
The Times article draws a clear distinction between general nutrition advice and food that is prescribed with the same clinical specificity as medication. As Corby Kummer of the Aspen Institute told Severson, the goal is not to help doctors throw better dinner parties — it's to understand how specific diseases like Type 2 diabetes, renal failure, and cardiac conditions can be managed through targeted food interventions. That specificity is what Project FoodBox is built around. Boxes are designed by registered dietitians and matched to a member's condition: the Renal box is built around kidney-safe produce, the Cardiac and Low Sodium boxes are calibrated for heart and blood pressure management, and the Diabetes box emphasizes low-glycemic vegetables. Each delivery contains 15 to 18 pounds of fresh produce at no cost to the member. The tailoring is not incidental — it is the mechanism.
The Times also traces the roots of this work to the AIDS epidemic of the 1980s, when volunteers began delivering nutrient-dense meals to help combat H.I.V. wasting syndrome. Community organizations expanded that model over the decades, eventually calibrating meals to Type 2 diabetes, cardiac disease, and kidney conditions. Project FoodBox operates in that same lineage — now at a scale of more than 97 million pounds of produce delivered since 2020, reaching more than 5 million people across Medi-Cal in California and Medicaid partnerships in New York. What began as crisis response has matured into evidence-based clinical practice.
The national conversation the Times is documenting is not abstract from where you sit as a member. The program you're enrolled in exists because enough evidence has accumulated to convince health plans and Medicaid administrators that medically tailored food belongs inside a treatment plan — not alongside it as an afterthought. Every A1C reading, every reported change in how often you reach for fast food, every week of improved symptoms contributes to that evidence base. You are not just receiving a benefit. You are part of an ongoing demonstration that this approach works.
If you found this useful, share it with a patient, caregiver, or clinician who manages chronic conditions — especially anyone on Medi-Cal in California or Medicaid in New York who may not know this program exists. You can read the full New York Times feature (paywall).