Background Adolescence is a period marked by critical changes in behaviours and body composition that place adolescents at an increased risk of becoming overweight and obese. Health education in schools has the potential to improve health behaviours by encouraging critical thinking about these issues. To develop sustainable interventions to prevent obesity, it is necessary to understand whether educational interventions during adolescence are effective, and which intervention elements are associated with improvements in BMI.

We addressed the question: do school-based educational interventions improve BMI in adolescence, and what intervention features are associated with effectiveness?

Methods We carried out a systematic review and meta-analysis of published intervention studies. We searched MEDLINE, PsycINFO, CINAHL, and ERIC from 2006 to 2017. Two independent researchers assessed titles and abstracts, extracted data and assessed risk of bias in the included studies, following the PRISMA guidance. Review inclusion criteria were: a) health education intervention, studies conducted in schools in high-income countries that included a control group b) participants aged 10–19 years, c) studies reporting BMI/BMI Z-score at baseline and follow-up. A meta-analysis was conducted combining findings of studies presenting data on BMI z-score (n=13). Heterogeneity was assessed using Cochran’s Q and the percentage of variability due to heterogeneity was quantified using I².

Results Searches identified 29,174 publications, of which 312 studies papers were selected as potentially meeting inclusion criteria. Twenty-five studies met the inclusion criteria. Most (n=18) were delivered by teachers in classroom settings, followed by researchers, schools nurses and students. Additional methods used in the interventions included out of class components, physical activity sessions, digital interventions and parent involvement. Eleven of the included studies were effective in improving BMI or BMI Z-score at follow-up periods from two months to two years post-intervention. Intervention features associated with effectiveness were the provision of training for teachers prior to intervention, and involvement of parents or families. Data from 13 of the 25 studies were included in a meta-analysis, using a random effects model given the high level of heterogeneity (I²=65.1%). The overall pooled estimate of change in BMI Z-score in the intervention group, compared with the control group, demonstrated that BMI improved in adolescents exposed to the intervention [β=-0.10, 95% CI (-0.14, -0.05); p<0.001].

Conclusion This systematic review demonstrates that school-based educational interventions have the potential to prevent obesity in adolescence. Mediating factors associated with effectiveness included providing training for teachers prior to the intervention and involving parents or families.