Objective: The purpose of this review of reviews is to summarize the evidence of the effectiveness of school-based interventions in reducing obesity and promoting physical activity and nutrition in children and adolescents.
Methods: The search strategy followed the guidelines set out by the Cochrane Collaboration for conducting systematic reviews. The databases MEDLINE, EMBASE, Cochrane, CINAHL, ERIC, PsycINFO, and Social Science Abstracts were searched from 1990 to January 2002 inclusive. The reference lists of all retrieved articles were reviewed for potentially relevant articles. Seven relevant peer-reviewed journals were hand-searched from 1996 to January 2002 inclusive. In addition to published literature, a search for unpublished material consisted of input from the Ontario Public Heath Association Primary Prevention of Diabetes Committee, key informants, and web site searches of Canadian, American, British, and Australian government agency sites. Two independent reviewers used standardized tools to rate each study for relevance andmethodological quality and to extract data.
Results: Twenty-three articles met the relevance criteria for the prevention of obesity and promotion of physical activity and nutrition. Quality ratings resulted in seven strong reviews, five moderate reviews, and 11 weak reviews. Although findings were inconsistent, certain trends contributed to the effectiveness of school-based programs. Programs that included an environmental change (cafeteria, physical education classes, curricula, lunch, or recess), were multi-faceted, were longer in duration, had frequent booster sessions, and were based on behavioural theory were more effective than programs that did not.
Conclusions: The results of synthesizing the outcomes of 12 reviews are inconclusive in determining which aspects of the design and components of a school-based intervention are most effective for reducing obesity and promoting physical activity and/or nutrition in children and adolescents. Some interventions were more effective than others at modifying one criterion, but not another, and effects were not necessarily seen in both sexes. Interventions were more effective at modifying knowledge than behaviour.