Autistic Traits and Social Anxiety As Unique Predictors of Neural Attentional Responses during Facial Emotion Identification

Background:

Individuals with autism spectrum disorder (ASD) exhibit impairments in the ability to perceive and respond to social cues, including facial emotion identification (e.g., Gross, 2004; Keehn et al., 2013). One reason for this deficit may be impairments in neural processing that occurs during emotion identification. Recent research using event-related potential (ERP) components of EEG has demonstrated both that autistic individuals show different neural processing when attending to emotional faces than that of non-ASD individuals (e.g., Sokhadze et al., 2015) and that neural processing is associated with differences in social skills (Hileman et al., 2011). Additionally, social anxiety is often co-morbid with ASD and, in individuals with ASD, social anxiety contributes to impaired facial emotion recognition (Corden et al., 2008). Research is needed to examine how traits related to autism and to social anxiety uniquely contribute to facial emotion identification.

Objectives:

The current study tested whether neural attention to faces depicting different emotions would be differentially affected by traits related to autism and social anxiety.

Methods:

Participants were 41 non-ASD undergraduate students (48.8% men; M_age = 19.7 years). Participants completed two self-report measures, the Autism Quotient (AQ, Baron-Cohen et al., 2001), and the Social Phobia and Anxiety Inventory (SPAI-23 Roberson-Nay et al., 2007), which assess autistic behaviors and social anxiety, respectively. While EEG was recorded from 65 electrodes, participants completed an oddball task in which deviant faces representing a change in emotion (i.e., happiness, anger, surprise, and fear) were presented within a series of faces displaying neutral emotions.

Results:

As is typical with an oddball task, visual inspection revealed a P3 ERP component which was quantified at electrode Pz between 300ms and 650ms. Separate analyses were conducted for each of the four oddball emotions: happiness, anger, surprise, and fear. Regression analyses were chosen to examine the unique contribution of autistic traits and social anxiety. These analyses revealed that, for happy stimuli, the AQ was a significant predictor of P3 amplitude, B = -.95, t = -2.35, p = .043, such that participants with more autistic traits showed lower amplitude P3s to happy faces. In addition, for fear faces, both AQ, B = -.94, t = -3.04, p = .019, and social anxiety, B = 1.07, t = 2.48, p = .042, significantly predicted P3 amplitudes. Interestingly, whereas higher autistic traits corresponded with smaller P3 amplitudes to the fear faces (see Figure 1), higher social anxiety was related to larger P3 amplitudes to the fear faces (see Figure 2).

Conclusions:

This study revealed that neural attention to faces depicting different emotions was uniquely predicted by traits related to autism and social anxiety. In fact, although measures of autistic traits and social anxiety are highly related, they were associated with different patterns of brain responses to fearful faces. This work sheds some light on differences in neural activation that impact the ability to perceive and respond to social cues. Future research should continue to examine how traits related to autism and to social anxiety uniquely contribute to facial emotion identification.