Abstract
Numerous relationship process variables are positively associated with psychotherapy outcomes, but less is known about the therapeutic relationship in mind-body interventions (e.g., yoga). This study examined components of the therapeutic relationship (alliance, credibility/expectations) for patients with generalized anxiety disorder randomized to 12-weeks of group Kundalini Yoga (yoga), cognitive behavior therapy (CBT), or stress education (SE) in groups of 3-6. 147 participants (71% women, M age=33.2±13.3) completed the Working Alliance Inventory (WAI) and Credibility and Expectancy Questionnaire (CEQ) (yoga: n=56, CBT: n=65, SE: n=26) and outcome measures of worry, anxiety, and depression. WAI included subscales of goal, task, and bond. ANOVA assessed differences in WAI and CEQ scores by treatment at midpoint (Week 6) and endpoint (Week 12). Longitudinal linear mixed effects models assessed effects of midpoint WAI by treatment on self-report outcomes. Mean WAI and CEQ scores were significantly higher for CBT than SE by midpoint. The largest differences in WAI subscale scores compared to SE were in task for both CBT and yoga. No significant differences emerged between CBT and yoga. WAI at midpoint was not a significant predictor of midpoint to posttreatment change in worry, anxiety, or depression, nor was the effect of WAI on change in these outcomes different between treatment groups. Results suggest patient ratings of relationship process variables were similar between yoga and CBT, but different from SE in group treatment for anxiety, and that components of the relationship may be optimized within yoga by augmenting goal and task, perhaps by borrowing elements from CBT-based treatments.