EEG Correlates of the Attentional Blink: Relationship to Autism Symptoms

Background:

In light of findings of superior temporal visual search abilities in individuals with autism spectrum disorders (ASD), we sought to understand neurophysiological mechanisms underlying single-target search and its relationship to the attentional blink (AB). The AB is a phenomenon whereby the second of two targets is frequently missed when it occurs within 200-500 milliseconds (ms) of the first.

Objectives:

There were three main objectives to this study. First, to examine characteristics of the AB in children with ASD and IQ- and age-matched comparison children. Second, to assess the relationship between electrophysiological responses to single color-bound targets and a color-binding AB task in both groups. Third, to assess the relationship between electrophysiological responses and diagnostic criteria for ASD.

Methods:

15 children with ASD (mean age = 11.8 years; mean IQ = 102.7) and 15 IQ- and age-matched comparison children completed an AB color-binding task in which they were asked to identify two purple targets amongst black distractors. Participants also completed a single-target version of this task while electrophysiological responses were recorded. Variables of interest included the accuracy of identifying the first target (T1), the accuracy of identifying the second target given that the first target was identified (T2|T1), the AB depth, and properties of the N2pc and P3 ERP components.

Results:

Analysis of the AB tasks revealed no group differences in target detection accuracy, or in the depth of the AB. ERP results from the single-target detection task revealed no significant differences between groups for the mean amplitudes of the N2pc and P3 components, or peak latency of the P3.

ASD and comparison groups differed in their relationship between the behavioral results on the AB color-binding task and ERP measures on the single-target detection task. For the ASD group, but not the comparison group, the AB depth was positively correlated with P3 peak latency in both the left (r = .732; p < .01) and right (r = .637; p < .05) hemispheres.

Exploratory analyses examining the relationship between ASD diagnostic measures and ERP results revealed a negative correlation between ADOS Social Affect scores and the amplitudes of the N2pc (r = -.728, p < .01) and left-lateralized P3 components (r = -.638, p < .05). Both verbal and nonverbal scores on the ADI were negatively correlated with the mean amplitudes of the N2pc and left-lateralized P3 components.

Conclusions:

Though no differences were found between ASD and comparison groups in any measures of the AB or the ERP components, the relationship between the AB and ERP components differed between the groups. For individuals with ASD, but not comparison individuals, AB depth was positively correlated with the P3 latency, indicating that individuals who are slower to process a single target also show a larger AB to dual-targets. Finally, exploratory analyses suggest that the size of the N2pc and left-lateralized P3 components might be related to social cognition impairments in ASD.