To the Editor.
—Dr Wohlraich and colleagues1 concluded on the basis of a meta-analysis of 23 studies that "sugar does not affect the behavior or cognitive performance of children." However, several factors argue for a more qualified interpretation of the data.The authors based their conclusion on the mean effect observed in each of the studies examined. That approach averaged children in whom sugar had no effect, a stimulatory effect, or a quieting effect. If only a minority of children in the studies were truly sensitive to sugar, it could easily have been obscured by this averaging. In one study2 (not included in the meta-analysis) of 21 boys, the majority became less active when exposed to sugar, but three became more active. Limiting the meta-analysis to studies in which means, SDs, or other statistical measures were provided and using those measures as the sole indication of whether sugar