Anxiety in Autism Spectrum Disorder Is Associated with Abnormal Prefrontal Cortex Activity

Background: Autism spectrum disorder (ASD) is often characterized by intense emotional responsiveness and poor emotion regulation.  Emotion dysregulation often takes the form of increased anxiety in ASD. While very little is known about the neurobiology of anxiety in ASD, the literature on pediatric anxiety disorders has converged on a model where the emotion functions of subcortical structures (namely amygdala) are poorly regulated by prefrontal cortex (PFC).  This model has yet to be robustly tested in samples of individuals with ASD who have co-occurring anxiety disorders.  

Objectives: The present study tests the hypothesis that co-occurring anxiety disorders in ASD are associated with abnormal PFC activity, amygdala activity, or both.

Methods: Data presented here are part of an ongoing study examining the psychophysiology of anxiety and emotion regulation deficits in ASD.  To date, 21 participants with ASD have been scanned using a selective attention facial identity task based on that of Vuilleumier and colleagues.  The task involves the simultaneous presentation of two pairs of faces and houses.  Participants were asked to selectively attend to either the face or house pairs and make a subordinate-level identity judgment.  Faces were either neutral or angry.  Analyses to date have examined the main effects of selective attention (face versus house judgment) and emotional valence (angry versus neutral faces).  All participants met criteria for ASD on a combination of the ADOS, SCQ, a developmental interview (typically the ADI-R), and clinical judgment.  Our sample also received in-depth phenotyping for anxiety following the Anxiety Disorders Interview Schedule (ADIS).  A total of 11 participants met our study criteria for anxiety (the ASD+Anxiety group, as opposed to ASD Alone). There were no significant differences in age or gender between ASD+Anxiety and ASD Alone groups.

Results: Selective attention to faces was associated with decreased dorsal and ventromedial PFC activity in the ASD+Anxiety group (versus ASD Alone).  Conversely, perception of angry versus neutral faces was associated with increased ventromedial PFC (anterior cingulate cortex) activity in the ASD+Anxiety group.  At the present sample size, no significant difference was identified in amygdala activation between the ASD+Anxiety and ASD Alone groups.

 Conclusions: The data from this ongoing study indicate that anxiety in ASD is associated with abnormal PFC activity.  Furthermore, the data suggests that this population may be relatively emotionally disengaged to faces, but hypervigilant when presented with emotionally charged faces.  This has real-world implications for social and communication development in individuals with both ASD and anxiety.