ERP Signatures of Rule Violation and Association with Repetitive Behavior in ASD

Background: Although “insistence on sameness” was described as a fundamental feature of autism in Leo Kanner’s original description, research into the underlying mechanisms of rigid behavior has been relatively scant. Understanding how this rigidity develops may inform targets for etiological research and for treatment. In the current study, we used event related potentials (ERP) to explore the neural time course of rule violation in ASD.

Objectives:  To investigate ERP signatures in response to rule violation in ASD children and controls. Based on previous research in healthy adults we hypothesized differences in the P300 wave (associated with decision making) and a frontal slow wave from around 500-700ms. 

Methods:  We used the CyberShape rule violation task developed by Michael Crowley as an analog to the CyberBall social exclusion task. The participant and two virtual partners are supposed to throw to another player indicated by a target shape that matches the shape in the player’s glove. But in alternating blocks of trials, one of the virtual players consistently throws to the “wrong” player.

Thirty four children 11-17 years of age, including 17 diagnosed with ASD and 17 age- and IQ-matched healthy controls, participated in the CyberShape task while being monitored using EGI’s Geodesic128 channel EEG nets and Netstation software. A parent of each participant completed questionnaires regarding dimensional measures of autism symptoms (the Social Responsiveness Scale) and anxiety (the Spence Children’s Anxiety Scale). 

Results:  No significant difference was found between groups for reaction times. ERPs were analyzed across three conditions: trials in a congruent block where there was no rule violation; trials in an incongruent block where one virtual partner threw against the rules to the other virtual player; and an incongruent condition where one virtual partner threw incorrectly to the participant. A repeated measures ANOVA using a 2 (diagnostic group) x 3 (condition) design wave found a significant main effect (highest activity for incorrect throws to the participant and lowest activity for the congruent trials) but the interaction with group was not significant. Nonetheless, for the ASD group only there was a significant negative correlation (r = -.61) between mean P300 activity and the Mannerisms subscale of the SRS, suggesting that greater repetitive and restricted behavior is associated with decreased P300 response.

A frontal negativity between 300-400ms (faster than we had hypothesized) did not show an overall significant main effect but the group x condition interaction was significant. Visual inspection of the data shows that a clear main effect in the control group was obscured by a lack of differential response to condition. 

Conclusions:  Our results indicate that the CyberShapes task can reliably elicit neural signatures related to decision making and monitoring, and there is evidence these are related to everyday symptoms of rigid repetitive behavior in ASD. Limited ability to differentiate among different conditions of rule violation or congruence may indicate arise due to inability to extinguish response from the incongruent trials and suggest confusion induced by rule changes in the environment.