Sugar rush or sugar crash? Experimental evidence on the impact of sugary drinks in the classroom

462 preschool children, across two field experiments

Preschool children (approximately ages 4-6), Belgium. Well-nourished.

Randomized field experiment assigning children to a sugary drink or an artificially sweetened drink, with behaviour and achievement measured before and after consumption, replicated across two experiments. Randomization and blinding of the drink are strengths. The reported findings are entirely sex-heterogeneous with no reported overall main effect - boys become briefly "relaxed" then restless and score worse, while girls are behaviourally unaffected and score better. Opposite-signed sex subgroups with no main effect is the classic signature of noise-mining, and the outcome measures are acute classroom observations rather than standardized achievement.

One sugary drink produced an initial relaxing effect in boys followed by increased restlessness; girls' behaviour was unaffected. Achievement effects were negative for boys and positive for girls. Results were reported as robust across two field experiments.

None - randomized. The threat is multiplicity and subgroup selection, not heredity.