Recent Treatment Research in Bulimia Nervosa

Treatment and research in bulimia nervosa have recently been reviewed by the Cochrane Database System.^1-5 After reviewing 19 trials that compared antidepressants with placebo, Bacaltchuk et al. concluded that the use of a single antidepressant was clinically effective for the treatment of bulimia nervosa when compared to placebo. Single antidepressant therapy produced a greater remission rate but also a higher dropout rate. In terms of dropout, fluoxetine was judged to be a more acceptable treatment than the tricyclic antidepressants.

Psychotherapy vs. Antidepressants

When Bacaltchuk and colleagues evaluated the efficacy of antidepressants compared with psychological treatments and their combination, the authors concluded that psychotherapy was superior. Using a conservative statistical approach, there was a clinically relevant difference. However, the difference was not significant. They concluded that the number of trials might be insufficient to show a significance of 20% absolute risk reduction in efficacy, and that psychotherapy was better accepted by patients. The authors also concluded that the efficacy of combined treatments was superior to single-treatment approaches. However, when antidepressants were combined with psychotherapy, the acceptability of the psychological approach was reduced. These reviews also found a small body of evidence for the efficacy of cognitive behavioral therapy (CBT) in the treatment of bulimia nervosa and “similar syndromes.” The quality of the trials was highly variable, and the sample sizes were often small. The researchers concluded that more trials were needed, particularly for evaluating patients with binge eating disorder (BED) and eating disorders not otherwise specified (EDNOS). They also suggested more trials to evaluate other psychotherapies and the effectiveness of less- intensive psychotherapies.

Relative to recently published drug trials, researchers reported that a placebo-controlled comparison of meclobemide, a MAO type A selective and reversible inhibitor, and pill placebo failed to show that the active drug was effective at the usual dosage.⁶ In contrast, ondansetron, a peripherally active antagonist of the serotonin receptor 5-HT3, which is commonly prescribed for cancer chemotherapy nausea prophylaxis, was superior to placebo.⁷ Overall, however, it appears that if drug therapy is to be used with bulimia nervosa patients, selective serotonin reuptake inhibitors (SSRIs) remain the treatment of choice.

Fluoxetine, which is the only Food and Drug Administration-approved drug for bulimia nervosa, was evaluated in two parallel, multicenter, double-blind, randomized, placebo-controlled trials. Further analysis of those data indicate that fluoxetine, 60 mg/day, was effective in treating bulimia nervosa, regardless of the presence or absence of comorbid depression.⁸ Open-label studies suggested that fluoxetine may be useful in adolescents with bulimia nervosa.⁹ A placebo-controlled maintenance trial among persons who responded to fluoxetine found evidence that continued treatment with the drug in patients with bulimia nervosa improved their outcome and decreased the likelihood of relapse, compared to placebo. However, the dropout rate in this trial was quite high.¹⁰

In a separate study, Walsh and colleagues treated 22 patients with bulimia nervosa who had not responded or had relapsed following a course of CBT or interpersonal therapy (IPT) with placebo or fluoxetine.¹¹ Significant improvement was demonstrated in those treated with the active drug, despite the lack of response to psychotherapy.

A double-blind, placebo-controlled trial of topiramate in bulimia nervosa has been published.¹² Sixty-four outpatients were included in the intent-to-treat analysis. Those in the active drug group had significant improvements in both binge and purge symptoms compared to the placebo group. Self-esteem, eating attitudes, anxiety, and body image were all improved in the active drug group.¹³

CBT and Exposure Therapy

There have also been a number of informative psychotherapy articles in the last few years. Bulik and her co-workers examined the utility of exposure to pre-binge cues, exposure to pre-purge cues, or a relaxation-training condition added to eight sessions of CBT.¹⁴ Results indicated that CBT was a highly effective treatment for bulimia nervosa. The authors concluded that exposure and response prevention for bulimia nervosa was an expensive and logistically difficult treatment that did not appear to add any significant benefits in proportion to the amount of effort required to implement it. The authors also found that the frequency of binge eating and the character trait of self-directedness may be useful predictors of those individuals likely to respond to a brief course of CBT for BN.¹⁵ Subsequently, they reported that post-treatment binge eating, food restriction, and urge to binge predicted poor outcome at one year, whereas the character trait of self-directedness predicted good outcome for treatment with CBT.¹⁶ At the three-year follow-up, 85% of the patients no longer met criteria for bulimia nervosa, and 69% did not meet criteria for any other eating disorder. No differential effects between relaxation and exposure therapy were found at three-year follow-up.

CBT Non-responders

Secondary treatments for CBT non-responders were examined in a multi-site trial.¹⁸ Initially, 194 women meeting DSM-III-R criteria for bulimia nervosa were treated with CBT at one of three sites. Analysis revealed that poor outcome was predicted by a reduction in purging of less than 70% by the sixth treatment session, allowing for identification of a substantial portion of prospective failures. The results of the secondary treatment for CBT non-responders were reported subsequently.¹⁹ Thirty-seven subjects completed such treatments, and 25 dropped out or were withdrawn from the study. The abstinence rate for subjects assigned to secondary treatment with IPT was 16%, and 10% for those assigned to medication management. The authors concluded that the dropout rates were excessively high and the response rates low, and that offering lengthy sequential treatments appears to have little utility.

Another study examined the relationship between the relative efficacy of fluoxetine and self-help therapy in a randomized placebo-controlled trial.²⁰ The singular and combined effects of fluoxetine and manual-based self-help were examined in 91 adult outpatient women with bulimia nervosa. Fluoxetine and the self-help manual were found to be effective in reducing the frequency of vomiting episodes and improving the response rates for vomiting and binge-eating episodes. Furthermore, both factors were shown to work additively. Keel and colleagues examined the 10-year outcome in a cohort of women who had been treated with CBT and/or antidepressant therapy at baseline.²¹ At long-term follow-up, the data demonstrated that both treatments resulted in improved psychosocial functioning. A further analysis of a study in which outpatient bulimia nervosa patients were assigned to CBT, supportive therapy, antidepressants, placebo, or medication alone indicated that high baseline frequencies of binge eating and vomiting as well as a history of substance abuse or dependence were negative prognostic indicators. CBT was significantly more effective than supportive therapy in reducing binge eating and vomiting frequency.²²

CBT, Interpersonal Therapy, and Guided Self-Help

A multicenter trial comparing cognitive behavioral therapy and interpersonal therapy found that CBT produced significantly more rapid improvement in patients with bulimia nervosa, although there were no significant differences at the conclusion of treatment.²³ The authors concluded that CBT should be considered the preferred psychotherapeutic treatment.

In a study comparing group and individual cognitive behavioral therapies, Chen and colleagues assigned 22 patients to group CBT and 22 other patients to individual CBT.²⁴ The effects were equal on most measures, and a greater proportion of individuals were abstinent from bulimic behavior following individual CBT. However, there were no significant differences between the two groups at follow-up.

In another study, Davis et al. investigated the efficacy of the step-care approach using group psychoeducation, followed by CBT.²⁵ The outcome suggested limited support for offering CBT to subjects who had completed the initial trial of group psychoeducation. In a dismantling study, Hsu et al. compared CBT to nutritional therapy and the combination against a support control group in outpatients with bulimia nervosa.²⁶ The results indicated that CBT therapy, either alone or in combination with nutritional therapy, is the preferred treatment.

Dialectical Behavioral Therapy and Self-Help

Safer, Telch, and Agras compared 20 weeks of dialectic behavioral therapy (DBT) to 20 weeks of a waiting-list control group of outpatients with bulimia nervosa, and found a highly significant decrease in target symptoms in the DBT group.²⁷ In a series of studies, self-help has been examined by investigators at the Maudsley Hospital in London.²⁸ Sixty-two patients with DSM-IIR-defined bulimia nervosa were randomly assigned either to use of a self-care manual plus eight sessions of CBT therapy (guided self-help) or to 16 weekly sessions of CBT. Both treatments appeared effective, and the authors concluded that guided self-help, incorporated with the use of a self-care manual, could be as effective as CBT. In a subsequent report, Treasure and colleagues described a study involving 125 patients with bulimia nervosa.²⁹

The first phase of treatment involved four sessions of either CBT or motivational enhancement. There were no differences between the two treatments, in terms of improvement in bulimic symptoms. A third report detailed a four-year follow-up of patients who had originally been assigned to the self-help manual plus eight CBT sessions (guided self-help) or CBT, finding equivalency between the two treatment cells.³⁰

Another group led by Walsh studied guided self-help and fluoxetine in a primary care setting, finding a high dropout rate for self-help and no evidence for efficacy, while fluoxetine was associated with better retention and symptomatic improvement compared to patients given a placebo.³¹

Carter and Olmsted and their colleagues compared the ability of a self-help manual addressing symptoms of bulimia nervosa with a self-help manual focusing on self-assertion skills.³² A subgroup of patients did indeed benefit, and the two self-help manuals had similar effects.

Durand and King compared the effectiveness of the general-practice-based self-help approach to a specialist outpatient treatment, and again found relative equivalency in outcomes for patients in both groups.³³ Another related trial showed that guided imagery was effective for treating patients with bulimia nervosa.³⁴

Still A Need for Novel Strategies

In summary, the treatment literature for bulimia nervosa has continued to develop and to show efficacy for both drug and psychotherapy approaches. However, a significant subgroup of patients—in many studies the majority of patients—remain symptomatic. One hopes that the next generation of studies will introduce novel strategies that are either more effective overall, or that are effective for targeted subgroups to which patients can be matched.