Individual Differences in Anxiety Symptoms Predict Amygdala Function in ASD

Background:   Although many of the biological systems supporting social functions also support basic emotion processes, there have been few attempts to date to understand the "social brain" in ASD in terms of these processes.  The absence of data on emotion systems in ASD is highly problematic, as anxiety disorders are common in ASD, but lead to a distinct set of predictions on amygdala and medial prefrontal cortex function that do not fit squarely within social brain models of ASD.  Individual differences in anxiety symptoms may prove critical to our understanding of how deficits in social brain structures contribute to the clinical presentation of ASD.

Objectives:   The present data test for a relationship between anxiety symptoms and social brain function (namely amygdala) among individuals with ASD. 

Methods:   Data from multiple functional MRI studies of ASD with and without co-occurring anxiety symptoms are presented.  Although data collection is ongoing, the present sample includes 42 children with ASD ranging in age from 7 to 17 years (mean age = 12.3 years).  FMRI tasks use faces as stimuli, though data collection using non-face affective stimuli is currently underway.  In addition to fMRI data, extensive neuropsychological and psychodiagnostic data were collected, including parent-report data on anxiety. 

Results:   Data collected to date indicate that increased anxiety is associated with increased amygdala activation during face perception in children in ASD.  The relationship appears strongest when considering anxiety dimensionally  (i.e., questionnaire measures including the parent-report version of the Screen for Child Anxiety Related Emotional Disorders).  The strongest amygdala/anxiety correlations were for panic symptoms and separation anxiety, r=.30 and r=.49, respectively (p-values of .046 and .001).  These correlations remained after removing variance associated with core ASD symptoms (measured via the Social Responsiveness Scale and the Social Communication Questionnaire).

Conclusions:   The present data suggest that increased anxiety is related to amygdala hyperactivity among individuals with ASD.  A model is proposed where amygdala activity reflects a hybrid signal of social approach (diminished in ASD) and emotional arousal (heightened among individuals with ASD and co-occurring anxiety).  Tests of this model will be greatly facilitated by increased attention paid to mood and anxiety symptoms in ASD, including the development of better diagnostic tools for measuring these symptoms.