A medically tailored meal programme has tested a gradual reduction in deliveries while retaining support for healthy eating at home. The American Heart Association released the preliminary findings on 2 November, ahead of its Scientific Sessions meeting on 6-9 November.
The comparison involved 110 adults with type 2 diabetes and high blood pressure. One group received ten meals a week for eight months. The other received ten for three months, followed by a monthly reduction, reaching five meals a week in the final month. Both groups received dietitian counselling and a heart-health curriculum, including cooking demonstrations and ingredients for home preparation. Read the AHA announcement.
FoodEdge view: the service is part of the product
For teams developing food-as-medicine services, our reading is that the delivery schedule belongs in the product brief alongside the food specification. A planned handover to home cooking could change what a provider needs to supply: meals, skills, ingredients and continued contact may each have a role. This study cannot separate their contributions.
The practical test is whether a service can reduce deliveries while preserving outcomes and access to suitable food, once the cost of the accompanying support is included. A smaller meal bill alone would leave that question unanswered.
The next evidence to examine is the full between-group analysis, retention and total programme costs. Five meals a week in the final month does not mean deliveries were halved throughout the programme. Any commercial model built on a 50% saving would outrun these findings.
Sources and reporting
Source context: US participants and programme. This is a service-design question for other markets, with local effectiveness and payment arrangements still to be tested. AHA funded the project through its Health Care by Food initiative. A public AHA-distributed copy preserves the release. These copies share one origin.
Desk report on the AHA release and project information. The conference findings remain preliminary; no participant-level data were analysed by FoodEdge.
AI tools supported source research and drafting during the editorial process.
