How does the Kids SIP smart ER program impact the sugar‐sweetened beverage intake of students: An investigation beyond total treatment effect in randomized controlled trial
Naveen Abedin et al.
Abstract
This study develops and empirically estimates a structural framework to decompose the causal pathways of multilevel behavioral interventions targeting adolescent health behaviors. We apply this framework to the Kids SIPsmartER (KSS) program, a 6-month, school-based intervention evaluated through a clustered randomized control trial in rural Appalachia to reduce sugar-sweetened beverage (SSB) consumption among 7th-grade students. KSS includes a classroom-based student component (KSS-S) grounded in the Theory of Planned Behavior and health literacy, and a caregiver component (KSS-C) delivered through two-way text messaging to modify household beverage practices. Using structural equation modeling grounded in a Stackelberg household production framework, we estimate how behavioral strategies, health literacy, and caregiver-child decision-making jointly shape students' SSB intake. Stronger behavioral intentions are associated with a 5.30 oz/day reduction (p<0.01), and the caregiver decision index (capturing household rules, availability, and role modeling) corresponds to a 16.41 oz/day reduction. KSS-S reduces intake by 6.42 oz/day (p<0.05), largely through a direct pathway (6.23 oz/day, p<0.05). KSS-C reduces intake by 4.22 oz/day, primarily through caregiver-mediated effects (3.36 oz/day). Combined, both components reduced students' SSB consumption by 10.64 oz/day (36% from baseline) (p<0.01). We use the estimated structural parameters to simulate hypothetical SSB tax scenarios, finding that KSS achieves reductions comparable to those generated by modest tax rates. This suggests that school-based behavioral interventions like KSS can function as complementary or alternative policy tools. By identifying causal mechanisms and enabling counterfactual policy simulations within a unified framework, the structural approach provides richer policy insights than average treatment effects alone.
Evidence weight
Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40
| F · citation impact | 0.50 × 0.4 = 0.20 |
| M · momentum | 0.50 × 0.15 = 0.07 |
| V · venue signal | 0.50 × 0.05 = 0.03 |
| R · text relevance † | 0.50 × 0.4 = 0.20 |
† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.