EEG Abnormalities in Adolescents with ASD: A Power and Coherence Analysis

Background: Social and cognitive difficulties observed in autism could be partly explained by an atypical integration and perceptual processing associated with abnormalities in cerebral activity. Indeed, Resting-state electroencephalography (EEG) studies conducted in adults with Autism Spectrum Disorder (ASD) show a U-shaped model of power alteration with excessive power in low-frequency (delta and theta) and high-frequency bands (beta and gamma) contrasting with reduced alpha power. In addition, other EEG studies report altered brain connectivity with both hyper- and hypo-connectivity. In children with ASD, some results recently obtained do not fit with these models with for example and contrary to adults, reduced theta observed before 13 years old. Hence, modifications may occur after 13 years or during adolescence leading to adult models but there are too few studies conducted during this period to conclude.

Objectives: To specify cerebral particularities during resting state with high-density EEG, among adolescents with high functioning ASD aged from 10 to 18 years old compared to controls.

Methods: EEG was recorded in a sample of 17 adolescents with ASD compared to 27 age and IQ-matched typically developing adolescents. We examined eyes-closed and eyes-open EEG relative power and coherence in resting state for beta, theta, alpha and beta bands. Both short and long range functional connectivity was assessed with electroencephalographic coherence between pairs of electrodes in a high-density electrode array.

Results: We observed an important variability between participants for both relative power and coherence. We reported a U-shaped profile of relative power alterations in the ASD group compared to controls: delta, theta and beta relative power were increased and alpha relative power was reduced in adolescents with ASD. We also noticed long range hypo-connectivity from frontal to posterior regions for all band frequencies that contrasted with hyper-connectivity in posterior regions only for alpha band.

Conclusions: Our findings validate the U-shape profile in adolescents with ASD with increased power at low (delta, theta) and high (beta) frequencies and reduced alpha power. This profile of activity is associated with abnormal long range connectivity in all frequency bands that is consistent with difficulties in cognitive integration and regulation from anterior brain regions on posterior ones. In addition, we reported a posterior and short-range hyper-connectivity in alpha band. This excessive connectivity may participate to the atypical perceptual processing that characterizes ASD people. Further analysis combining EEG and behavioral data collected in this study will be crucial to identify EEG correlates of both perceptual and cognitive processing in adolescents with autism.