School-based obesity interventions in the metropolitan area of Rio De Janeiro, Brazil: pooled analysis from five randomised studies.

School-based studies, despite the large number of studies conducted, have reported inconclusive results on obesity prevention. The sample size is a major constraint in such studies by requiring large samples. This pooled analysis overcomes this problem by analysing 5926 students (mean age 11·5 years) from five randomised school-based interventions. These studies focused on encouraging students to change their drinking and eating habits, and physical activities over the one school year, with monthly 1-h sessions in the classroom; culinary class aimed at developing cooking skills to increase healthy eating and attempts to family engagement. Pooled intention-to-treat analysis using linear mixed models accounted for school clusters. Control and intervention groups were balanced at baseline. The overall result was a non-significant change in BMI after one school year of positive changes in behaviours associated with obesity. Estimated mean BMI changed from 19·02 to 19·22 kg/m2 in the control group and from 19·08 to 19·32 kg/m2 in the intervention group (P value of change over time = 0·09). Subgroup analyses among those overweight or with obesity at baseline also did not show differences between intervention and control groups. The percentage of fat measured by bioimpedance indicated a small reduction in the control compared with intervention (P = 0·05). This large pooled analysis showed no effect on obesity measures, although promising results were observed about modifying behaviours associated with obesity.