Inter and Intra Hemispheric EEG Coherence in Adolescents with ASD Related to Social Function

Background: While differences in neural connectivity among those with ASD is well-established, there is a paucity of studies that have examined brain connectivity across an intervention for ASD. Electroencephalogram (EEG) is a flexible tool to assess neural differences related to ASD (Coben, Mohammad-Rezazadeh, & Cannon, 2014). People with ASD have been characterized as having relatively greater right-versus-left-frontal hemisphere activity (Burnette et al., 2011; Moncrieff, 2010). Further, at rest, people with autism show functional under-connectivity in anterior-posterior connections (Cherkassy et al, 2006).

Objectives: The primary objective of the current study was to examine EEG coherence in the alpha band and relations to social behavior among adolescents with ASD before and after a social skills intervention.

Methods: Fifty-nine (59) adolescents between the ages of 11-16 years participated in this study. Autism was confirmed with the ADOS and IQ was assessed with the KBIT-2; an IQ of ≥ 70 was necessary for participation. Participants were part of a RCT of the PEERS^® intervention, which is a manualized, social skills treatment for youth with ASD. Both before and after intervention, adolescents or caregivers completed the Quality of Socialization Questionnaire (QSQ), Social Responsiveness Scale (SRS) and Test of Adolescent Social Skills Knowledge (TASSK).Electroencephalogram (EEG) was performed in a three-minute eyes-open resting state paradigm. The Magnitude Squared Coherence (MSC) was calculated with MatLab using the Minimum Variance Distortionless Response method (MVDR; Capon, 1969) to examine the relation between two signals. Six coherence measures (frontal, temporal, and parietal regions in each hemisphere respectively) were calculated for each participant at pre- and post-intervention. The alpha band (8-12 Hz) was examined because it is associated with neural activity at rest. After distribution analysis, Spearman rank-order (rho) correlations were performed of EEG coherence and social outcome scores before and after intervention.

Results: At pre-, a positive relation was found between the Temporal Right-Temporal Left coherence and the QSQ Total, (r_s (59) = .36, p < .05). At post-, results revealed negative links between the SRS Autistic Mannerisms Subscale and Frontal Left-Frontal Right (r_s (59) = -.39, p < .01), Temporal Right-Temporal Left (r_s (59) = -.31, p < .05), Parietal Right-Parietal Left (r_s (59) = -.36, p < .01), and between QSQ Total and Frontal Right-Parietal Right (r_s (59) = -.32, p < .05).

Conclusions: Findings at pre-intervention provide further support for the under-connectivity theory of ASD. Furthermore, results demonstrate additional associations between coherence and social behavior were found at post test. Across intervention, as autism severity symptoms decrease, the neural connectivity between specified brain regions increases. These findings will be examined further by comparing additional participants as well as comparing data to typically developing adolescents as well as a delayed treatment group. Findings of this study have implications for the structure of interventions for ASD across settings as well as gaining an understanding of neural and behavioral plasticity in ASD.