Evaluating Three Psychotherapeutic Approaches for Anorexia Nervosa

Virginia W. McIntosh, PhD and co-workers in New Zealand and at the University of North Carolina recently evaluated three different psychotherapeutic approaches for patients with anorexia nervosa (AN).While many randomized controlled studies have examined therapies for adults with bulimia nervosa, few controlled randomized studies have examined treatment for AN (Am J Psychiatry 2005;162:741).

Dr. McIntosh and colleagues studied 56 women with AN who were randomly assigned to one of three treatments: cognitive behavior therapy (CBT), interpersonal therapy, or a control treatment approach that combined clinical management and supportive psychotherapy (nonspecific supportive clinical management). All patients were treated with 20 sessions over at least 20 weeks, and were assessed after the 10th week of therapy and after her final session.

Some surprising results

Thirty-five women completed therapy. Among the 21 others who did not complete the study, dropouts occurred over the course of therapy. Reasons for leaving the study included dislike of the therapy, moving from the area, and for unexplained reasons. Four were hospitalized for weight loss or complications of AN (3 in the interpersonal psychotherapy group, and 1 assigned to nonspecific supportive clinical management). One hospitalized patient died later, probably from medical complications of AN.

At the end of the study, 9% of the women had a very good outcome and 21% more had improved considerably; 70% made only small gains or no gains. The surprise for the researchers was that contrary to their hypothesis, patients who received nonspecific supportive clinical management had an outcome as good or better than the outcomes of those who received specialized psychotherapy. Interpersonal psychotherapy was the least effective of the three approaches. Possible explanations for this include the relative lack of focus on symptoms, the relatively long time taken to decide on the problem area, which reduced the middle phase of interpersonal psychotherapy, and the lack of reactivity of the symptoms of AN. Possible reasons that CBT wasn’t more effective included the large amount of psychoeducational material and extensive skills that had to be acquired, inability to generate alternatives to fixed cognitions stemming from the ego-syntronic nature of AN, which made it difficult to work toward change.

According to the authors, nonspecific supportive clinical management may be a particularly effective treatment for acute AN. Its effectiveness may have been due to the fact that clinicians experienced with treating eating disorders followed a detailed treatment manual, and included a detailed discussion of ways to increase food choices in quantities in order to gain weight. The remaining session content was based on issues the patient chose to present, which may have given patients a sense of autonomy and control, improving the therapeutic alliance, empathy, positive regard, and support for this patient group.