Electrophysiological Correlates of Audio-Visual Integration of Spoken Words in Typical Development and Autism Spectrum Disorder

Background: In a previous electrophysiological (ERP) study examining audio-visual (AV) integration of speech in typically developing adults we found a speeding up and attenuation of the auditory N1 component specific to an AV condition with informative or predictive lip movements.  This N1 attenuation correlated with an earlier increased fronto-polar negativity (FPN) raising the possibility of a top-down modulation effect.  The present study examines ERP correlates of AV integration of spoken words in typically developing (TD) adolescents and adolescents with autism spectrum disorder (ASD).  There are a number of reasons why we might expect to see differences in an autistic population, including atypical unimodal auditory processing (e.g. Bomba & Pang, 2004), atypical unimodal visual processing, particularly with regards to face processing (e.g. McPartland et al, 2004), and multi-sensory processing differences (e.g. Bebko et al, 2006; Magnee et al, 2008).

Objectives: The present study will examine whether the adult pattern of AV integration effects are replicated in a group of TD adolescents and whether the ERP correlates of multisensory integration are different in adolescents with ASD compared to TD adolescents.

Methods: ERPs were recorded from 19 TD adolescents and 14 adolescents with ASD while they were presented with monosyllabic words in one of five conditions:  audio-only (A), visual-only (V), audio-visual with face (AVF), audio-visual with scrambled face (AVS), and visual-only scrambled face (VS).  Significant multisensory interactions [AVF - (A+V)] > 0 were examined.

Results: Preliminary results suggest that in both autism and typical development, only approximately half of participants show the adult pattern of AV integration.  Amongst those showing the FPN and N1 attenuation effects there are differences between the autistic and typically developing adolescents.  Namely, TD adolescents show AV integration effects in the P50 but adolescents with ASD do not, the amplitude of the FPN effect was greater in the TD group than the ASD group, and the N1 attenuation and FPN effects were correlated in the TD adolescents but not the ASD adolescents.  Spectral analysis of the EEG is currently being conducted to attempt to elucidate differences between subgroups (both TD and ASD) and differences between TD and ASD adolescents.

Conclusions: The electrophysiological correlates of AV integration appear to undergo a developmental maturation well into young adulthood, in both TD and ASD individuals.  ERP correlates also imply that the AV integration process is different in ASD.