Hyperkinesis and food additives: testing the Feingold hypothesis.

Teacher ratings, objective classroom and laboratory observational data, attention-concentration, and other psychological measures obtained on 36 school-age, hyperactive boys under experimental and control diet conditions yielded no support for the Feingold hypothesis. Parental ratings revealed positive behavioral changes for the experimental diet; however, they seemed primarily attributable to one diet sequence. Parents' behavioral ratings on ten hyperactive, preschool boys indicated a positive response to the experimental diet; again, laboratory observations showed no diet effect.