Nineteen studies were included in the review. Ten U.S. cohort studies (although some also used cross-sectional analyses for selected outcomes) were included (six studies were randomised cohorts and 4 were observational/non-randomised cohorts) and study samples ranged from 1041 to 5106 participants. Six international cohort studies were included and sample sizes were not stated. Three additional physical education (PE) and food service cohort studies (although one also used cross-sectional analyses for selected outcomes) were included for comparison purposes.
In the 10 U.S. studies, the weighted ER was 40% (range 9% to 68%), i.e. significant positive effects were observed for 108 of 272 comparisons reported between treatment and comparison groups.
The unweighted ER was 38% (range 15% to 73%).
In the 6 international studies, the weighted ER was 22% (range 9% to 100%), i.e. significant positive effects were observed for 50 of 209 comparisons reported between treatment and comparison groups. The unweighted ER for international studies was 47% (range 10% to 100%).
In the 3 PE and food service studies, the weighted ER was 23% (range 0% to 57%), i.e. significant positive effects were observed for 16 of 71 comparisons reported between treatment and comparison groups. The unweighted ER was 27% (range 0% to 57%).
Across the 16 studies, the overall weighted ER was 31% (range 8% to 83%), i.e. significant positive effects were observed for 158 of the 502 comparisons reported between treatment and comparison groups. The unweighted ER was 43% (range 2% to 98%).
Positive effects were observed more frequently for smoking (80%), cognitive (65%), and fitness (36%) outcomes with lower rates observed for blood pressure (18%) and adiposity measures (16%).