Action Schools! BC (AS! BC) was scaled-up from an efficacy trial to province-wide delivery across 11 years (2004-2015). In this study we: (1) describe strategies that supported implementation and scale-up; (2) evaluate implementation (teachers' physical activity (PA) delivery) and student's PA and cardiorespiratory fitness (CRF) within a cluster randomized controlled trial during years 2 and 3 of scale-up; and (3) assess relationships between teacher-level implementation and student-level outcomes. We classified implementation strategies as process, capacity-building or scale-up strategies. Elementary schools (n = 30) were randomized to intervention (INT; 16 schools; 747 students) or usual practice (UP; 14 schools; 782 students). We measured teachers' PA delivery (n = 179) using weekly logs; students' PA by questionnaire (n = 30 schools) and accelerometry (n = 9 schools); and students' CRF by 20-m shuttle run (n = 25 schools). INT teachers delivered more PA than UP teachers in year 1 (+33.8 min/week, 95% CI 12.7, 54.9) but not year 2 (+18.8 min/week, 95% CI -0.8, 38.3). Unadjusted change in CRF was 36% and 27% higher in INT girls and boys, respectively, compared with their UP peers (year 1; effect size 0.28-0.48). Total PA delivered was associated with change in children's self-reported MVPA (year 1; r = 0.17, p = 0.02). Despite the 'voltage drop', scaling-up school-based PA models is feasible and may enhance children's health. Stakeholders must conceive of new ways to effectively sustain scaled-up health promoting interventions if we are to improve the health of students at a population level. Clinical Trials registration: NCT01412203.
Keywords: children; effectiveness; fitness; implementation strategies; intervention; physical activity; scale-up; school.