Implementation and Effectiveness Outcomes of a Pilot Comparative Effectiveness Randomized Controlled Trial Evaluating a Food Is Medicine Program Among At-Risk Pediatric Populations

<b><i>Introduction:</i></b> Food Is Medicine (FIM) programs have demonstrated effectiveness in improving diet quality and food insecurity. There remains a lack of evidence of their impacts in pediatric populations. This pilot comparative effectiveness randomized control trial assessed two FIM strategies on implementation and health outcomes in Medicaid-eligible children aged 5-12 years with a BMI ≥85th percentile. <b><i>Methods:</i></b> Participants (<i>n</i> = 150) were enrolled for a 32-week intervention through recruitment at two urban pediatric primary care clinics in Houston, TX. Participants were randomized 1:1:1 into three arms: (1) biweekly $25 produce vouchers + nutrition education, (2) biweekly produce home delivery + nutrition education, or (3) wait-listed control (<i>n</i> = 50 per arm). Implementation outcomes included retention, redemption, dosage, reach, fidelity, and acceptability. Child outcome measures included diet, food security, BMI z-scores, hemoglobin A1c, liver panels, and lipid panels. Multilevel mixed-effects regression models were used to assess the effectiveness of the intervention on outcomes. <b><i>Results:</i></b> On average, voucher participants redeemed $353 out of $400 (88%) of their funds, and 100% of the home delivery group received ∼18 lb (52 servings) of produce per week. Parents found the program helpful in reducing grocery costs (voucher: 95%, delivery: 76%). Compared to the control group, voucher group participants had a significant decrease in aspartate aminotransferase (-5.50, 95% confidence interval: -9.43, -1.57, <i>p</i> = 0.006) from baseline to post-intervention. <b><i>Conclusions:</i></b> This pilot study found FIM programs are both feasible and well accepted among at-risk populations, with slightly higher acceptability of the voucher model. More adequately powered studies with a stringent design are needed to test their effectiveness.