Attention retraining in social anxiety disorder: an fMRI study
Sawyer, Alice Tyer
MetadataShow full item record
Research suggests that patients with social anxiety disorder (SAD) have an attentional bias toward socially threatening stimuli, and recent studies have shown that computerized interventions designed to train attention away from such stimuli decrease attentional bias and SAD symptomatology. The current study sought to replicate findings from previous attention retraining studies and to examine neural mechanisms underlying attentional biases in SAD using functional magnetic resonance imaging (fMRI). Thirty-two SAD patients were randomized to complete either eight 15-minute sessions of a probe detection task designed to train attention away from disgust faces (n=16), or a placebo control task (n=16). Before and after these sessions, patients completed an fMRI probe detection task. Sixteen matched healthy controls also completed this fMRI task on one occasion. Study hypotheses were as follows: (a) post-intervention, SAD patients in the retraining condition would show greater reductions in attentional bias and SAD symptomatology compared to patients in the placebo condition; (b) SAD patients would show greater amygdala activation, and less prefrontal cortex (PFC) activation, when viewing negative faces than healthy controls; and (c) post-intervention, SAD patients in the retraining condition would show less amygdala activation, and greater PFC activation, when viewing negative faces than patients in the placebo condition. Results showed no between-group differences in attentional bias or SAD symptomatology post-intervention, with both groups showing significant symptom reduction. However, attentional bias change was significantly correlated with symptom change across the entire SAD sample (N=32) and was predictive of Liebowitz Social Anxiety Scale scores at post-intervention. Neuroimaging results showed hypo-activation in the orbitofrontal cortex and anterior cingulate cortex at pre-treatment for the SAD group compared to healthy controls. At post-treatment, this difference was no longer significant across the entire SAD group (N=32). Finally, results indicated that activation at pre-treatment in the posterior cingulate cortex/precuneus was significantly correlated with symptom change across the entire SAD sample. These results suggest that SAD patients may not be engaging higher-level cortical regions as readily as healthy controls and add to the recent growing body of research suggesting that attention retraining may not be an effective treatment for patients with SAD.