Motivational Interviewing and Cognitive Behavioural Therapy for Generalized Anxiety Disorder

The proposed study builds on the findings of the clinical trial by Westra, Arkowitz and Dozois (2009), and aims to: (1) provide a powerful test of the value of adding MI to CBT for high severity GAD (as pilot data has indicated that individuals with severe GAD may preferentially benefit from the addition of MI to CBT), (2) employ treatment procedures generalizable to clinical practice, and (3) examine the mechanisms underlying any additive treatment benefit. 106 individuals with a principal diagnosis of GAD and of high worry severity will be randomly assigned to receive either 4 sessions of MI followed by 11 sessions of CBT (MI-CBT arm) or 15 sessions of CBT (CBT alone arm). Both groups will also receive 2 follow-up CBT treatment sessions at 1 and 3 months post-treatment. Each therapist will deliver all treatment components, treatments will appear seamless to clients, and therapists will be nested within treatment group to control for allegiance effects and avoid cross-contamination of the therapies. In order to increase generalizability to clinical practice, in the CBT phase for the MI-CBT arm, therapists will shift to MI in the presence of empirically derived markers of resistance and shift back to CBT when resistance has diminished. In the CBT alone arm therapists will respond to resistance using manualized recommendations derived from leading CBT theorists for the management of resistance. Anxiety and related symptoms, motivation, and treatment engagement will be assessed at various points during treatment, and at 6 and 12 months post-treatment. A multi-method approach to assessment will be used, including self-report, clinician-report, clinician-administered, observer-rated, and interview-based measures. Importantly, we will include not only traditional outcome measures, but also process measures (e.g., observer-coded interpersonal behavior, interview-based derivation of client experiences) to investigate possible mechanisms underlying any additive treatment benefit. In addition, all therapy sessions will be videotaped and assessed by independent raters for protocol adherence.

The specific hypotheses are as followed:

Hypothesis 1: Worry, Anxiety, \& Related Symptoms. The MI-CBT arm will show greater reductions in worry, anxiety, and related symptoms (depression, disability, maladaptive beliefs about worry, life satisfaction) and a greater percentage of treatment responders (using clinical significance criteria) across the treatment period and at follow-up (FU), compared to the CBT alone arm.

Hypothesis 2: Motivation and Resistance. Over the course of early treatment (i.e., sessions 1 to 4), the MI-CBT arm will demonstrate greater increases in motivation compared to the CBT alone arm. In addition, the MI-CBT arm will exhibit lower in-session resistance during early (sessions 5 \& 7), middle (sessions 10 \& 12), and late (sessions 15 \& 17) phases of treatment.

Hypothesis 3: Homework Compliance. The MI-CBT arm will show greater client- and therapist-rated homework compliance throughout CBT treatment compared to the CBT alone arm.

Hypothesis 4a: Interpersonal Processes between Clients and Therapists. Compared to CBT alone clients, MI-CBT clients and therapists will report higher working alliance quality during both the initial MI sessions and subsequent phases of CBT. Furthermore, moment-to-moment interpersonal processes between client and therapist during early, middle, and late treatment are expected to be characterized by more adaptive processes (e.g., higher affiliation and less hostility) in the MI-CBT vs. CBT alone arms.

Hypothesis 4b. Interpersonal Processes During Resistance. During episodes of client resistance (identified with resistance coding), therapist and client exchanges in MI-CBT are expected to be characterized by more adaptive processes (i.e., higher affiliation, less hostility, and greater client autonomy-preservation) compared to CBT alone. Further, when a small subset of clients (5 per treatment group) are interviewed about their experiences of resistance episodes, we expect client accounts in the MI-CBT arm to be characterized more positively (e.g., more supportive of client autonomy, more conducive to increased engagement vs. disengagement with treatment) compared to CBT alone.

Hypothesis 5: Mediation of Therapeutic Outcomes. We expect that increased motivation and reduced resistance in MI-CBT, compared to CBT alone, will lead to higher levels of engagement in the CBT (better homework compliance and therapeutic alliance), which will result in greater worry reduction.