Investigating change in worry and anxiety during cognitive-behavioural therapy and applied relaxation for generalized anxiety disorder

The present study examined data from a controlled clinical trial comparing two types of psychotherapy for Generalized Anxiety Disorder (GAD): Cognitive-Behavioural Therapy (CBT), and Applied Relaxation (AR). The goal of this study was to examine precedence of change in worry and anxiety in CBT and AR. According to the interacting subsystems model of anxiety, change in the cognitive aspect of anxiety (worry) may lead to change in the somatic aspect (physiological anxiety), and vice-versa. It was hypothesized that in CBT, change in worry would precede change in anxiety, and conversely, that in AR change in anxiety would precede change in worry. Twenty participants (CBT n = 10; AR n = 10) completed daily ratings of worry and anxiety during therapy. Using Time-Series Analysis (Tiao & Box, 1981), the causal impact of each variable on the other was assessed. Results showed no differences between the treatment groups; however, the majority of participants (70%) had a bi-directional relationship between worry and anxiety. A contingency analysis showed a near-significant trend suggesting that treatment responders were more likely to have a bi-directional causal relationship between worry and anxiety. These results appear to be consistent with the interactive subsystems model of anxiety.