Abstract

In 1973 Rimland reported that some autistic children responded favorably to high doses of vitamin B6. Since this finding, different studies were performed to identify apparently B6 responsive subjects and to critically evaluate clinical and biological B6 responsiveness. Magnesium was included because large doses of B6 might increase irritability. 44 patients (mean age 9.3 years) were examined. All selected children had marked autistic symptoms. The children received a complete diagnostic work-up, including psychiatric, psychological, neurological and medical evaluation. Clinical data were scored using an estimate of global clinical state and numerical rating on a 18 item scale (Behavior Summarized Evaluation). In a first open trial 15 out of 44 children exhibited moderate clinical improvement with worsening on termination of the trial. Thirteen responders and 8 non responders were re-tested in a 2-week crossover, double-blind trial and the responses to the open trial were confirmed. Biochemical data analysis revealed that a significant decrease in urinary homovanillic acid (HVA) levels was observed during B6-Mg administration. During B6-Mg treatment, middle latency evoked potentials exhibited a significant increase of amplitude.