School-based diet and physical activity interventions show modest effects on child weight in Africa

The study designs were predominantly quasi-experimental (70%) or cluster randomized controlled trials (18.2%), with modest to low risk of bias. When nine intervention comparisons were pooled in a meta-analysis, the effect on BMI-for-age was not statistically significant (MD = 0.03, 95% CI: −0.24 to 0.29; p = 0.813), despite moderate heterogeneity (I² = 69.0%). Individual studies did report modest improvements in adiposity measures and maintenance of normal weight status in intervention groups. The review found that multi-component programs combining nutrition and physical activity showed reductions in a few comparisons, suggesting that integrated approaches may hold promise. A major limitation was the geographic concentration in South Africa, which reduces generalizability across the African continent.

The heterogeneity in results (I² = 69.0%) points to substantial variation between studies—intervention intensity, duration, fidelity, and local context all appear to matter. Several individual trials did show adiposity improvements, which suggests the pooled null finding may reflect implementation gaps or study quality variation rather than true ineffectiveness. The multi-component nature of most interventions makes it difficult to isolate whether physical activity, nutrition education, or behavioral support drove any benefits. Critically, only 63.6% of studies were published in the last decade, and 72.7% were from South Africa; evidence from West Africa, East Africa, and other regions remains sparse. The review explicitly highlights the need for more methodologically robust designs and geographic diversity before drawing strong conclusions about what works across the continent.