Abstract
BACKGROUND: Young people spend half of their days in school, but evidence concerning the influence of school environment on the physical activity (PA) of pupils is still inconsistent. A better understanding of potential correlates of PA on the school-level and their possible interaction with individual aspects is needed to improve the development of more effective interventions. METHODS: We used data from the 2009/10 German Health Behaviour in School-aged Children study (HBSC)-sample (n=5,005 students aged 11-15 years) including self-reported moderate to vigorous intensity PA as well as a variety of biological, demographic and behavioral correlates and matched them with school-level data from the national school principals' HBSC questionnaire. We analyzed the associations of individual- and school-level correlates with MVPA by gender-specific multi-level regression. RESULTS: Only a small share of the overall variation in student's PA was attributable to the school-level. Consequently, the associations of individual-level correlates with PA were stronger than those of the school-level. Our analysis revealed significant associations of individual-level (i.e. age, consumption of softdrinks, overweight) as well as school-level correlates (i.e. the availability of a football ground and a swimming pool) with MVPA. We also observed some gender-specific findings especially for the school level correlates. Cross-level interactions between individual- and school-level were not apparent. CONCLUSIONS: Our findings indicate the usefulness of applying an ecological framework to understand and explain complex health behaviors like PA. As we found gender-specific association it might be important to acknowledge that boys and girls have specific needs to be more physically active. Further research should also take other features/elements of the school environment and neighborhood as well as socio-cognitive correlates into account to advance the field.