© 2009 Canadian Medical Association or its licensors

Interpret results with caution The conclusions drawn by the authors of the March 31 meta-analysis of the effect of school-based physical activity interventions on body mass index (BMI) in childrenare troubling. Their analysis found that school-based physical activity interventions im-plemented in the general school popula-tion do not improve BMI. This says nothing about effects of physical activity on the subgroup of children who are overweight or obese; it says nothing about the role of physical activity in pre-venting weight gain in children of healthy weight; and it cannot account for changes in body composition among children (a limitation they note). Physical activity should be an inte-gral part of population-based health promotion efforts. We should be work-ing to enhance these efforts in schools. We should avoid drawing conclusions (with policy implications) about the impact of physical activity on child-hood obesity using data and method-ological approaches that have inherent limitations