Neural Correlates of Affective Processing in Adults with Autism Spectrum Disorder

Background: The ability to perceive and interpret emotional faces, integral to successful social functioning, has been found to be impaired in ASD. Magnetoencephalography (MEG) is a non-invasive neuroimaging modality that provides high spatio-temporal resolution of neural activity. Previous MEG studies have shown atypical neural activity during affective face processing in youth with ASD but no studies to date have examined the neural substrates of this ability in adults with ASD.

Objectives: The present study explored the temporal and spatial properties of MEG activation during implicit angry and happy face processing in young adults with and without ASD. We hypothesized that adults with ASD would show atypical activity, particularly in social brain areas.

Methods: We recruited 52 young adults (ASD group: N=26, 8 females, 26.3+4.2 years; Control group: N=26, 8 females, 26.3+4.1 years) to complete an emotional face processing task in the MEG. Participants were presented rapidly (80ms) with an emotional face (happy or angry) concurrently with a scrambled pattern (target) on either side of a central fixation cross. Using a response button box, participants indicated the location (right or left) of the target by pressing the appropriate button. A 3T structural MRI was obtained for each participant for co-registration with MEG data for accurate source localisation. SPM12 was used for MEG data pre-processing and analysis. Empirical Bayes Beamformer estimated activation sources related to happy and angry face processing.

Results: Happy and angry faces elicited greater activation in adults with ASD, relative to controls, in a number of areas including bilateral inferior temporal gyri, bilateral fusiform gyri, and left parahippocampal gyrus. Adults with ASD also showed greater activity in key social processing areas such as the right anterior insula and anterior cingulate cortex. Of particular interest was our finding that, to angry faces specifically, adults with ASD showed greater right amygdala and left superior orbital activity, relative to controls. In contrast, in controls, angry and happy faces elicited greater, relative to adults with ASD, activity in areas including sustained activity in the left cuneus/precuneus and the right post-central gyrus. Furthermore, within group analyses showed more scattered activation patterns to both happy and angry faces in adults with ASD, compared to controls.

Conclusions: This study is the first to provide evidence of distinct patterns of atypical activation to happy and angry faces in adults with ASD using MEG. Differences in neural activity between young adults with and without ASD indicate atypical processing in neural areas involved in face perception, affect processing, threat processing, and emotion processing and social reward may account for deficits in atypical emotional processing in adults with ASD. Within group results suggest adults with ASD show more disorganised processing of emotions. Thus, atypical recruitment of neural regions may contribute to deficits in affective processing early on in young adults with ASD.