Differentiating autistic children with quantitative encephalography: a 3-month longitudinal study.

The present study used a single-channel quantitative electroencephalographic (EEG) assessment to differentiate autistic children from normal control subjects. One hundred five normal and 17 autistic children participated in the study. In addition to amplitude measures of the frequency bands of delta, theta, alpha, sensorimotor rhythm, and beta and the theta to beta ratio, intra- (6 minutes) and intersessional (3 months) consistencies were also examined. The results indicated that autistic children showed significantly higher quantitative EEG amplitudes in many of the frequency bands than normal children; furthermore, their quantitative EEG activities were found to be relatively unstable within a 6-minute session compared with normal children. Discriminant function analyses revealed that absolute sensorimotor rhythm and beta amplitudes were the best predictors that correctly differentiated autistic children from normal children in the present sample, with a high accuracy rate of 95.2%. In addition, quantitative EEG measurements of normal and autistic children were found to be generally consistent across the 3-month period.