Event-Related Potentials Index Atypical Processing of Auditory Tones in Young Children with Autism Spectrum Disorder

Background: Magnetoencephalography (MEG) studies have identified atypical processing of simple tones as a putative biomarker associated with Autism Spectrum Disorder (ASD). Specifically, participants with ASD exhibit temporal delays and hemispheric differences emanating from superior temporal gyrus between 50 and 100ms post-stimulus, and reduced activity at approximately 200ms post-stimulus, relative to controls. Furthermore, these differences are associated with language abilities in ASD. However, these findings have yet to be replicated using other measures, such as electroencephalography (EEG) based event-related potentials (ERPs). EEG and MEG are highly similar in their millisecond-level temporal resolution, yet different in that MEG has generally greater sensitivity to brain activity source locations with the limitation of being relatively insensitive to radially oriented brain activity.

Objectives: To investigate the neural correlates of simple auditory tone processing in children with ASD using EEG-based ERPs.

Methods: Participants are children with ASD aged 3- to 5-years (n=14) and chronological age and nonverbal mental age matched typically developing (TD) children (n=14). Auditory stimuli consist of 400ms long 200 Hz, 600 Hz, and 1000 Hz sine tones, presented at 70dB. The current analyses focus on the N200 and P200 components recorded over temporal and frontal cortex, respectively. ANOVAs with Frequency (200 Hz, 600 Hz, 1000 Hz) and Hemisphere (N200: Left, Right; P200: Left, Medial, Right) as within-subjects factors and Group (ASD, TD) as a between-subjects factor were conducted on the amplitudes and latencies of these components. Correlations between component latency and an index of language delay were also computed.

Results: P200 latencies exhibited a Frequency by Group interaction over frontal cortex (F(2, 25) = 4.70, p < .05, η_p² = .273), whereby latencies in the Medial region were shorter in ASD participants for 200 Hz (ASD: 174ms; TD: 187ms) and 600 Hz (ASD: 156ms; TD: 167ms) tones, but longer for ASD participants for the 1000 Hz tone (ASD: 163ms; TD: 153ms). No other significant Group effects were observed. P200 latencies were also negatively correlated with degree of language delay for the 200 Hz (r = -0.575, p < .05) and 600 Hz (r = -0.601, p< .05) tones in the ASD group.

Conclusions: The preliminary ERP findings of this on-going study replicate the general finding of atypical auditory perceptual processing for simple tones in ASD, albeit with a different pattern than that observed in previous MEG studies. Differences in the measurement properties and capabilities for MEG and EEG may explain the different patterns in the data. Alternatively or additionally, the age of the participants (younger in the current study) may help explain the differences. Regardless, the replication of an ASD versus TD group level difference in simple tone processing as well as correlation of more atypical tone processing with language delays in the ASD group, strongly suggest that the current EEG study findings reflect sensitivity to the biomarker previously identified using MEG. Together, EEG and MEG studies have potential to provide more detailed characteristics and explanations for atypical tone processing in infants, children, and adults with ASD.