Exploring mediators of accelerometer assessed physical activity in young adolescents in the HEalth In Adolescents study – a group randomized controlled trial

BACKGROUND There is a shortage of information about the factors that mediate physical activity intervention effects which involve youth. The purpose of this study was to examine whether personal, social and physical-environmental factors mediated the intervention effect on physical activity and whether gender and weight status moderated mediated effects in the Health In Adolescents Study - a school-based intervention to promote healthy weight development among young adolescents. METHODS Participating schools were randomized to Control (n = 25) and Intervention (n = 12). The intervention components to enhance physical activity targeted change through theoretically informed mediators embedded in a social-ecological framework. Accelerometer assessed physical activity (mean count per minute) and self-efficacy, enjoyment, perceived social support from parents, teachers and friends and perceived environmental opportunities were measured by questionnaires at baseline and post-intervention after 20 months among 700 11-13 year-old adolescents (Intervention = 485; Control = 215). The product-of-coefficient test was used to examine mediation. RESULTS No mediating effect of any of the hypothesized mediators was identified and gender and weight status did not moderate any mediated effects with the exception of weight status that moderated the mediated effect of enjoyment. Few intervention effects were seen on the mediators, except for a positive change in social support from teachers among girls and the normal weight, and a negative effect on enjoyment and self-efficacy among the overweight. However, change in enjoyment, self-efficacy, perceived social support from friends and environmental opportunities were associated with change in mean count per minute with some variation across the investigated subgroups, and thus show evidence of being potential mediators of physical activity change in adolescents. CONCLUSIONS While no mediation effects were observed, change in both personal and social-environmental factors predicted change in physical activity behavior. Hence, a social- ecological approach targeting a wide range of determinants to promote change in physical activity holds promise. Overweight and normal weight adolescents may not respond in the same way to school-based physical activity interventions. Therefore, strategies to better reach the overweight seem needed. Future studies should continue to identify mediating and moderation mechanisms in physical activity change in adolescents.