School-based gardening, cooking and nutrition intervention increased vegetable intake but did not reduce BMI: Texas sprouts - a cluster randomized controlled trial

16 elementary schools in Texas randomized to Texas Sprouts (n=8) or control (n=8). Participants: 3rd–5th grade students; baseline N=3,135 (1,412 in intervention; 1,723 in control). Mean age ~9.2 years; 64% Hispanic; 47% male; 69% eligible for free/reduced lunch. Outcomes included anthropometrics (BMI, waist circumference, body fat by bioelectrical impedance), blood pressure, and dietary intake (vegetables, fruits, and sugar-sweetened beverages). Analyses used generalized weighted linear mixed models to account for clustering by school; complete-case and multiple-imputation approaches.

Texas Sprouts program delivered during the school year (9 months) with a 0.25-acre outdoor garden, Garden Leadership Committees, 18 student gardening, nutrition, and cooking lessons (60 minutes each), and nine monthly parent lessons, taught by trained educators.

Vegetable intake rose in the intervention relative to control by about 0.33 to 0.48 servings per day (p-values ~0.002–0.02, depending on analysis). No significant differences for fruit intake, sugar-sweetened beverages, obesity measures (BMI, BMI z-score, BMI percentile, waist circumference, body fat) or blood pressure. Conclusion: The program increased vegetable intake but did not reduce obesity markers or blood pressure within one school year; longer or more intensive, sustained approaches may be needed to influence obesity-related outcomes. Demonstrates feasibility and potential to improve dietary patterns in a low-income, predominantly Hispanic population.

Limitations include limited generalizability (predominantly low-income, Hispanic sample); very low parental engagement (about 7% attended at least one parent lesson; only 88.9% of parent sessions were delivered); weather disruptions led to indoor sessions for ~34% of child lessons; duration was one school year with no long-term follow-up; dietary data were self-reported with moderate validity; limited integration with school foodservice; missing data and variability across sites, though multiple-imputation results aligned with complete-case analyses.