Using fMRI to Identify Neural Mechanisms of Response to Cognitive Behavioral Therapy for Anxiety in ASD

Background: Cognitive Behavioral Therapy (CBT) can be helpful for anxiety in ASD but, little is known about the neural mechanisms of this potentially effective intervention.

Objectives: This presentation discusses a randomized-controlled trial to identify neural mechanisms of CBT efficacy for anxiety treatment in children with ASD, and to present the results of an open study of fMRI correlates of emotion regulation before and after CBT for anxiety in ASD.

Methods: Our ongoing, randomized-controlled study has been designed to test the effects of CBT on the fMRI signatures of fear processing and emotion regulation in children with ASD. Children are randomized to CBT conducted using the CBT manual or to a supportive therapy control condition and anxiety outcomes are assessed by the Pediatric Anxiety Rating Scale administered by an independent evaluator. Children also complete neurocognitive tasks of emotion regulation and socio-emotional processing before and after treatment. Prior to initiating the randomized trial, we also conducted an open study with 10 children with ASD (age rage 10 to 12; 3 girls, 7 boys) who completed the face perception and emotion regulation tasks during fMRI before and after CBT for anxiety. During the fMRI tasks children were required to view faces and shapes and to either passively look at the emotional or neutral images or to decrease their emotional response to aversive emotional images.

Results: A whole-brain fMRI analysis of the emotion regulation task revealed increased activation in vmPFC, vlPFC, and dlPFC after CBT. We also observed reduction of left amygdala and bilateral insula activation in the look-negative vs. look-neutral contrast (emotional reactivity) after treatment. The magnitude of the effects of CBT on the levels of BOLD activation in the contrasts of interest calculated as the Cohen’s d effect size for the difference in post- to pre-treatment activation divided by the pooled standard deviation ranged from 0.68 to 1.23, indicating moderate to large effect sizes. Mean differences between faces vs shapes condition of the face perception task were computed from beta values extracted from the structurally defined left and right amygdala and vmPFC for the pre- and post-treatment scans. Paired-samples t-test indicated that there was a decrease in activity in the bilateral amygdala (Cohen’s d=0.66 and 0.51 for left and right amygdala, respectively) and an increase in activity in the vmPFC (Cohen’s d=0.51) relative to baseline.

Conclusions: CBT can engage the neural circuitry of emotion regulation in children with ASD and co-occurring anxiety. Clinical implications of understanding neural mechanisms to enhance the effectiveness of CBT for anxiety symptoms in ASD are discussed.