Common and distinct neural correlates of facial emotion processing in social anxiety disorder and Williams syndrome: A systematic review and voxel-based meta-analysis of functional resonance imaging studies

Background: Social Anxiety Disorder (SAD) and Williams-Beuren Syndrome (WS) are two conditions which seem to be at opposite ends in the continuum of social fear but show compromised abilities in some overlapping areas, including some social interactions, gaze contact and processing of facial emotional cues. The increase in the number of neuroimaging studies has greatly expanded our knowledge of the neural bases of facial emotion processing in both conditions. However, to date, SAD and WS have not been compared. Methods: We conducted a systematic review of functional magnetic resonance imaging (fMRI) studies comparing SAD and WS cases to healthy control participants (HC) using facial emotion processing paradigms. Two researchers conducted comprehensive PubMed/Medline searches to identify all fMRI studies of facial emotion processing in SAD and WS. The following search key-words were used: emotion processing; facial emotion; social anxiety; social phobia; Williams syndrome; neuroimaging; functional magnetic resonance; fMR1 and their combinations, as well as terms specifying individual facial emotions. We extracted spatial coordinates from each study and conducted two separate voxel-wise activation likelihood estimation meta-analyses, one for SAD and one for WS. Results: Twenty-two studies met the inclusion criteria: 17 studies of SAD and five of WS. We found evidence for both common and distinct patterns of neural activation. Limbic engagement was common to SAD and WS during facial emotion processing, although we observed opposite patterns of activation for each disorder. Compared to HC, SAD cases showed hyperactivation of the amygdala, the parahippocampal gyrus and the globus pallidus. Compared to controls, participants with WS showed hypoactivation of these regions. Differential activation in a number of regions specific to either condition was also identified: SAD cases exhibited greater activation of the insula, putamen, the superior temporal gyrus, medial frontal regions and the cuneus, while WS subjects showed decreased activation in the inferior region of the parietal lobule. Conclusions: The identification of limbic structures as a shared correlate and the patterns of activation observed for each condition may reflect the aberrant patterns of facial emotion processing that the two conditions share, and may contribute to explaining part of the underlying neural substrate of exaggerated/diminished fear responses to social cues that characterize SAD and WS respectively. We believe that insights from WS and the inclusion of this syndrome as a control group in future experimental studies may improve our understanding of the neural correlates of social fear in general, and of SAD in particular. (C) 2014 Elsevier Ltd. All rights reserved.