Up to 3 percent of young women have bulimia nervosa as defined in the Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM–IV). The disorder usually includes binge eating followed by self-induced vomiting and incorporates a sense of loss of control over the behavior and distortions of body image. The condition tends to become chronic, and 20 to 50 percent of patients have symptoms for at least five years despite treatment.

Faris and colleagues postulated that afferent vagal circuits could play a role in bulimia because the latter are essential to normal appreciation of satiety and control of pain thresholds. They studied the effect of treating bulimia nervosa with ondansetron, a selective serotonin receptor antagonist. Ondansetron was developed to control vomiting associated with chemotherapy and radiation therapy. The drug acts to decrease afferent vagal neuro-transmission.

The investigators selected women who met diagnostic criteria for bulimia nervosa and reported more than seven episodes of binge eating per week and vomiting for at least six months plus loss of control over the behavior. The mean age of the patients was 29 years, and the mean duration of illness was 11.8 years. Most of the patients reported multiple unsuccessful previous attempts at treatment.

The extensive initial assessment included physical and psychological examinations, including testing for depression. The patients selected for the study had a normal physical examination, electrocardiogram and blood count. They had a body mass index of 17.5 to 23.5 kg per m² (38 lb, 8 oz to 51 lb, 11 oz per m²), and they were not pregnant and did not have any serious medical conditions. Patients were not psychotic, bipolar or suicidal. They had no known history of drug or alcohol abuse, and they were not taking any psychoactive medications.

During the first week of the study, the 26 women recorded all eating behaviors to establish a baseline. In the second week, all patients received placebo tablets but were told that assignment to placebo or active treatment had taken place and that their assignment could be changed from week to week. Beginning in the third week, the patients were randomly assigned to receive ondansetron (14 patients) or an identical-looking placebo capsule (12 patients) for four weeks.

Throughout the study period, the patients maintained a diary of all eating and purging events, other symptoms and medication use (i.e., the times when the study drug or placebo were taken). Daily contact was maintained with a research assistant, and patients were seen each week by a psychiatrist. Weight, height and pill counts were measured at the weekly visits.

Of the ondansetron group, 13 patients completed the study (one patient dropped out because of accidental injury). All 12 patients using placebo completed the study.

During the placebo week, all patients showed a 20 to 23 percent reduction in episodes of bingeing and vomiting behaviors. After four weeks of treatment, the mean number of bingeing and vomiting episodes in the ondansetron group had dropped from 12.8 (placebo week) to 6.5 per week. In the placebo group, the mean number of episodes was 13.4 during the first week and 13.2 after four weeks. Patients in the ondansetron group also showed a significant improvement in the number of normal meals consumed and a significant decrease in the time spent in bulimic practices. Patients treated with ondansetron reported an average decrease of 7.6 hours per week in time spent in bulimic behaviors, compared with an average decrease of 2.3 hours in the placebo group.

The authors concluded that ondansetron appeared to normalize several aspects of satiation and control of eating behaviors.

editor's note: The toll of bulimia nervosa on patients and their families can be enormous. Because these patients are often of normal weight or only moderately obese, the diagnosis may be missed, or patients may not be believed when they ask for help. Once the problem is recognized, the available treatment strategies are limited and offer only modest success.

This study may open up a new area of therapy, but medication alone is never likely to be successful in patients with bulimia. Patients need a multifaceted approach to develop a healthy concept of food, relearn appropriate eating practices and achieve appropriate self-image and self-esteem. In this process, medication can be a powerful adjunctive treatment.

The results obtained in this study were modest. Even the successful patients continued to invest significant time and energy in abnormal behaviors related to food. Nevertheless, the results are encouraging in that they indicate a possible new line of treatment and verify a physiologic component to this distressing condition. For many patients, knowing that bulimia nervosa is not a completely psychologic condition can be a relief and an incentive to persevere with all aspects of treatment.—a.d.w.