OBJECTIVE: This study investigated the relation between clients' attachment patterns and the therapeutic alliance in two psychotherapies for bulimia nervosa. METHOD: Data derive from a randomized clinical trial comparing cognitive-behavioral therapy (CBT) and psychoanalytic psychotherapy (PPT) for bulimia nervosa. Client attachment patterns were assessed with the Adult Attachment Interview. Independent raters scored audiotapes of early, middle, and late therapy sessions for 68 clients (175 sessions) using the Vanderbilt Therapeutic Alliance Scale. RESULTS:Client attachment security was found to be a significant (p = .007) predictor of alliance levels at the three measured time points, with clients higher on attachment security developing stronger alliances with their therapists in both treatments as compared to clients higher on attachment insecurity. No evidence was found to support a hypothesized interaction whereby dismissing clients would develop weaker alliances in PPT and preoccupied clients would develop weaker alliances in CBT. CONCLUSIONS: As the first study to examine client attachment and therapeutic alliance using observer-based instruments, this study supports the theoretical assumption that clients with secure attachment patterns are likely to develop stronger alliances with their therapist across different treatment settings.
RCT Entities:
OBJECTIVE: This study investigated the relation between clients' attachment patterns and the therapeutic alliance in two psychotherapies for bulimia nervosa. METHOD: Data derive from a randomized clinical trial comparing cognitive-behavioral therapy (CBT) and psychoanalytic psychotherapy (PPT) for bulimia nervosa. Client attachment patterns were assessed with the Adult Attachment Interview. Independent raters scored audiotapes of early, middle, and late therapy sessions for 68 clients (175 sessions) using the Vanderbilt Therapeutic Alliance Scale. RESULTS: Client attachment security was found to be a significant (p = .007) predictor of alliance levels at the three measured time points, with clients higher on attachment security developing stronger alliances with their therapists in both treatments as compared to clients higher on attachment insecurity. No evidence was found to support a hypothesized interaction whereby dismissing clients would develop weaker alliances in PPT and preoccupied clients would develop weaker alliances in CBT. CONCLUSIONS: As the first study to examine client attachment and therapeutic alliance using observer-based instruments, this study supports the theoretical assumption that clients with secure attachment patterns are likely to develop stronger alliances with their therapist across different treatment settings.