Improving the implementation of an adolescent obesity prevention intervention in under-resourced schools

Guided by the Consolidated Framework for Implementation Research ( CFIR 2.0 ), this study examined the implementation and impact of an adapted, school-based, multi-component adolescent obesity prevention intervention called the Pr eventing O besity U sing D igital-Assisted M ovement and E ating (ProudMe) across two phases, with refinements made in between. Twelve middle schools from two districts ( N = 266 students) participated in the study. Phase 1 included six schools (ProudMe: n = 3 schools, 33 students; control: n = 3 schools, 46 students), and phase 2 included six schools (ProudMe: n = 5 schools, 165 students; control: n = 1 school, 22 students). ProudMe comprised four components: ProudMe Cafeteria (i.e., simplified Smarter Lunchrooms strategies), ProudMe PE (i.e., a purposeful physical education [PE] curriculum), ProudMe Tech (i.e., AI-assisted behavioral counseling), and ProudMe PD (i.e., a half-day staff professional development [PD] with continual support). Mixed methods data were used to assess three implementation outcomes (i.e., penetration, fidelity, sustainability). Pre-to-post changes in cafeteria scores, health behaviors (physical activity, screen time, fruit/vegetable intake), and BMI z-scores were compared across groups. Phase 2 demonstrated stronger implementation outcomes than phase 1 across ProudMe components. Penetration index was high in phase 2 and/or showed marked improvement. Fidelity index was high for ProudMe Cafeteria and ProudMe PD in both phases, while a moderate to high level of modification was observed with ProudMe PE implementation. Phase 2 demonstrated positive sustainability (8.4 ± 1.83 on a 10-point Likert scale). For impact, the cafeteria scores gained more in phase 2 ( M diff = + 8.3, p = 0.10; d = 1.53) than those in phase 1 ( M diff = + 4.67, p = 0.06; d = 1.14) versus a decline in control schools ( M diff = − 2.36, p = 0.61; d = − 0.36). No significant group differences were observed for student health behaviors or BMI z-scores. Qualitative results further suggested that phase 2 reflected improved perceived feasibility, contextual fit, and implementer buy-in across components. CFIR analysis revealed more facilitators and fewer barriers in phase 2 than phase 1, although time constraints, testing schedules, technology-related disruptions, and low parental consent remained persistent barriers. The refined ProudMe trial (phase 2) yielded improved implementation outcomes and measurable impact on school cafeteria policy and environmental changes. Student-level behavioral and weight status change may require longer duration to emerge.