Neural correlates of face processing in social anxiety disorder and their modulation by therapeutic interventions
Final Report Abstract
The overall aim of the project was to contribute to a better understanding of the neural basis of 1) social anxiety disorder (SAD), 2) cognitive behavioral therapy (CBT) as the primary nonpharmacological intervention for treating this condition and 3) its modulation by a cognitive enhancer (D-cycloserlne: DCS) thought to increase the effect of CBT In SAD. The primary measure obtained during the project was neuroimaging data from a comparatively large number of patients diagnosed with and treated for SAD and a group of matched control participants. Functional scans were collected while patients and controls were engaged in several face processing tasks that were tapping into different neural mechanisms. Most importantly, brain activation was examined In a "social task" involving angry versus neutral faces compared to emotional versus neutral and, secondly, in "novelty task" involving previously familiarized faces or objects versus novel faces or objects. All participants were scanned twice. Patients with SAD additionally received treatment between the scanning sessions with CBT that was either combined with DCS or with placebo. Comparing brain responses as measured by functional magnetic resonance Imaging (fMRI) during the novelty task, between-group differences were found in a number of cortical and subcortical areas, in particular the amygdala. Importantly and to some degree surprisingly, we found that SAD is associated with a reduced capacity to adapt or habituate to familiar faces: Patients with SAD had reduced brain responses to novel versus familiar faces when compared to the control group. Further analysis demonstrated that this effect was due to a higher response to familiar faces in patients. Importantly, these effects were specific to face stimuli since no corresponding findings could be revealed for object stimuli. With respect to the therapy of SAD using CBT, the first research question that we addressed with our data set was if the brain responses collected for the social task prior to treatment would allow predicting the outcome of the treatment. Our reasoning was that such a brain-based prognosis of treatment outcome might in future permit personalized treatment plans for this disorder. We indeed found that particularly brain responses to angry versus neutral faces in two clusters of activation located in the higher visual cortex were significantly correlated with the outcome of treatment with SAD. These brain measures outperformed conventional measures for treatment outcome prediction such as degree of severity, and point toward a possible nearterm clinical application of neuroimaging. Ongoing analyses address the questions of 1) actual changes in the brain after treatment of SAD using CBT, 2) possible differential effects for CBT either in combination with or without DCS and 3) the investigation of additional brain measures collected during the study, for instance those pertaining to brain structure and functional connectivity between brain regions.
