Shared and Distinct Features of Dynamic Emotional Face Processing in Autism Spectrum Disorder and Schizophrenia As Measured By Neural and Behavioral Responses

Background: Social impairment is a shared feature of autism spectrum disorder (ASD) and schizophrenia (SCZ). Nonverbal communication, including emotional facial expression, is a central component in understanding the social world. Examinations of electroencephalography (EEG) and eye-tracking (ET) have revealed similar brain and behavior atypicalities in individuals with ASD and SCZ in response to emotional faces. Collecting EEG and ET simultaneously during changes in emotional expressions allows for the examination of brain-behavior relationships in more ecologically valid contexts.

Objectives: To contrast event-related potentials (ERPs) to dynamic emotional faces in adults with typical development (TD), ASD, and SCZ and examine their relation to behavior, eye-tracking, and affective expression.

Methods: The TD (n=21), ASD (n=18), and SCZ (n=15) groups were matched on age (range: 18-33 years). EEG was collected using a 128-channel Geodesic Sensor Net. Concurrently, ET was collected with an EyeLink-1000 remote camera system. Participants underwent an interactive paradigm during ET-EEG collection in which they viewed a neutral face, and, contingent upon participant gaze to the eyes, the expression changed to happy or fearful. The N170 ERP component was extracted at emotional expression onset. Behavior, time looking at facial features, and a measure of affective blunting, the Unchanging Facial Expression item from the Scale for the Assessment of Negative Symptoms, were collected.

Results: The TD, ASD, and SCZ groups did not differ on N170 amplitude or latency response to dynamic emotional faces. The ASD and SCZ groups were combined into a clinical group in which an interaction effect between N170 amplitude and group approached significance [F(1, 53)=3.66, p=.06]. A significant difference in time spent looking at the left eye emerged across the TD, ASD, and SCZ groups [F(1, 51)=3.80, p=.03]. The ASD group looked to the left eye significantly less than the SCZ group (p=.01) which survived Bonferroni correction. Time spent looking at eyes in the fearful condition was related to faster right N170 latency in the ASD group (r_s=-.52, p=.03), and a trend was present in the SCZ group (r_s=-.45, p=.09), but not the TD group (r_s=.21, p=.36). This relationship in the happy condition approached significance in the ASD group (r_s=-.45, p=.06) only. A marginally significant relationship between faster right N170 latency in the fear condition and affective expression, as measured by the Unchanging Facial Expression item, emerged in the ASD group (r_s=.557 p=.06) only.

Conclusions: Examination of individual diagnostic groups and a single clinical group compared to normative social function revealed both common and distinct elements of brain-behavior responses to emotional faces between clinical groups. Together, individuals with clinical levels of social dysfunction exhibited distinct neural response from TD individuals suggesting atypical emotional face processing may not be unique to ASD. Yet individuals with ASD and SCZ had distinct viewing patterns of emotional faces and showed distinct brain-behavior relationships from one another and the TD group. Furthermore, the association between observed affect and emotional face processing may serve as a bridge between clinical phenotype and underlying biological response, specifically, in ASD.