The Effects of Relaxation on Cognitive Bias Modification Training in Social Anxiety Disorder

Cognitive models of social anxiety disorder posit that these individuals often interpret social situations in maladaptive ways. Cognitive bias modification of interpretations (CBM-I; Mathews & Mackintosh, 2000) is a promising avenue of treatment that targets these maladaptive interpretations, and aims to train individuals to interpret ambiguous situations in more adaptive ways. However, the effect sizes associated with these paradigms are typically small (Hallion & Ruscio, 2011). Relaxation is a treatment component that has been utilized in previous studies to augment both exposure-based (e.g., Borkovec & Sides, 1979) and cognitive (e.g., Borkovec & Costello, 1993) treatments for anxiety. In the current study, we investigated whether relaxation can similarly enhance the efficacy of CBM-I for social anxiety disorder in terms of both cognitive and behavioral outcomes. Participants were randomly assigned to receive one of three inductions (relaxation, socially anxious thinking, or neutral thinking) prior to completing a CBM-I training that trained them to interpret ambiguous social situations in a benign manner. After assessing participants’ interpretation bias (i.e., cognitive outcome), participants were asked to give an impromptu speech, during which we assessed self-reported anxiety and behavioral avoidance (i.e., willingness to give a speech, length of speech). We additionally assessed participants’ perception of their speech performance in terms of overall speech elements and specific speech behaviors. Contrary to hypotheses, rather than participants in the relaxation condition, participants in the neutral thinking condition evidenced the most adaptive patterns of interpretation bias. There were no between-groups differences on behavioral outcomes, despite differential patterns of interpretation bias. However, relative to participants in the other two groups, participants in the socially anxious thinking condition rated their performance on global speech elements relatively higher than their performance on specific elements of the speech. Combined with evidence from another study in our lab (Stevens et al., in preparation), these findings first suggest that relaxation may not enhance cognitive flexibility. Second, combined evidence from this study and several prior investigations of CBM-I for anxiety-related concerns suggest that these paradigms may not exert effects at the behavioral level, which may pose a serious limitation of these paradigms in clinical practice.