This paper pooled the results of many earlier studies to see whether sugar affects children. It combined 16 reports that together held 23 studies, all of which compared the same people with and without sugar. Most of the studies gave children a drink or food with sugar (mostly sucrose) on one occasion and a sugar-free placebo on another. One study instead gave children full diets for three periods of three weeks each.
Across 14 different types of measurement, the average effects were very small and could point either way. They ranged from a tiny drop on direct observations of behavior (-0.14) to a small rise on academic tests (+0.30). For all 14 measures, the range the study was confident about still included zero, meaning the results fit with sugar having no real effect.
The people studied were varied. They included children with typical development, children with ADHD, children whose parents said they reacted to sugar, delinquent teenagers, psychiatric inpatients, and people with Prader-Willi syndrome, with ages from 2 to 30. In the trials, sugar doses were 1.25 to 5.6 grams per kilogram of body weight, or fixed amounts of 35 to 78 grams.
The authors note the small number of studies gave low power, so a small effect, or an effect in some children, cannot be ruled out. Anyone with questions about their own situation should talk to a clinician.
Source: Wolraich M. The Effect of Sugar on Behavior or Cognition in Children. JAMA (1995). doi:10.1001/jama.1995.03530200053037