Evidence Check: What 14 Randomized Trials Really Say About “Food as Medicine”

Quick Take for Physicians: Across 14 randomized trials of Food-as-Medicine interventions, diet quality and food security reliably improved; clinical endpoints (e.g., HbA1c, BP) were mixed—likely due to short duration and small samples.

The statement maps a path to larger, longer, standardized trials—and describes how clinicians can act now.

‍

What the AHA Statement Covers

The scientific statement catalogues U.S. RCTs of produce prescriptions, medically tailored groceries, and medically tailored meals (MTMs) for noncommunicable diseases. Outcomes spanned diet quality, food security, clinical markers, and utilization.

🔗 Circulation PDF
🔗 AHA Summary

‍

Key Findings at a Glance

• Diet quality: Most trials improved fruit/vegetable intake and HEI components in intervention arms.
  Source
• Food security: Consistent reductions vs. controls, especially with produce Rx and grocery benefits.
  Source
• Clinical markers: Mixed results for HbA1c/BP/weight during short follow-up; several diabetes trials show limited short-term A1c change.
  Comparator RCT
• Heterogeneity: Variation in intensity (meals/week), duration (8–24 weeks), and wraparound services (RD coaching) limits pooling and power.
  Source

How to Act in Clinic While Evidence Scales
‍

1) Identify High-Yield Patients

• Diabetes, hypertension, ASCVD risk, or HF plus food insecurity or low diet quality.
• Functional barriers to shopping/cooking → consider MTMs over groceries/produce alone.

2) Prescribe the Right Intervention

• Produce Rx with clear weekly targets and simple education.
• Medically tailored groceries for cardiometabolic profiles.
• MTMs (7–14/week for 8–12 weeks) for patients with IADL limitations or post-discharge windows.

3) Add Wraparound Supports

• Dietitian touchpoints every 2–4 weeks (telehealth fine).
• Reminders, delivery reliability, and simple resistance activity to preserve lean mass.

4) Measure What Matters

• Diet quality (HEI), food security status, weight/BP; HbA1c or time-in-range for diabetes cohorts.
• Utilization: ED visits, unplanned admissions; program persistence.

‍

Fast onboarding • Evidence-aligned programs • Outcome tracking

‍

FAQs

Why are clinical endpoints mixed if diet quality improves?

Many trials are short and underpowered. Behavior change appears quickly; downstream biomarker and utilization changes often need longer duration and larger samples.
‍

Which intervention should I pick first?

Match to barriers: produce Rx or groceries for mild–moderate needs; MTMs for patients with food insecurity plus functional limitations or post-discharge risk.
‍

What should we track to show value?

Diet quality, food security, weight/BP, HbA1c (where relevant), and ED/admission rates; also track program persistence and patient-reported function.