TORONTO – Fluoxetine failed to prevent relapse of anorexia nervosa in the largest controlled medication trial to date exploring the issue, Dr. B. Timothy Walsh reported at the annual meeting of the American Psychiatric Association.
“Antidepressants don't offer patients with anorexia nervosa very much,” said Dr. Walsh, director of the eating disorders research unit of the New York State Psychiatric Institute, New York City. “We can't say for sure that an antidepressant other than fluoxetine wouldn't have an effect, but it would surprise me.”
The study represents yet another failure to find an efficacious medical treatment for anorexia. These failures imply that anorexia is unlike any other psychiatric condition, Dr. Walsh said.
“Interventions that are useful for other sorts of possibly related disorders don't work terribly well for anorexia nervosa,” he said. The only evidence-based treatment for anorexia is cognitive-behavioral therapy (CBT), which helps, but “not a lot,” he added.
The investigators, whose results were later published, compared fluoxetine with placebo in a clinical trial involving 93 patients who had completed intensive treatment and maintained a body mass index (BMI) of at least 19 kg/m
Subjects were randomly assigned to receive 20 mg daily of fluoxetine (49 patients), with a goal of increasing to 60 mg, or placebo (44 patients). They were monitored by a psychiatrist for dosing, adverse effects, and general medical status. They also received CBT.
The subjects underwent assessments of depression, anxiety, self-esteem, and quality of life every month. They were followed for 50 weeks, or until they either relapsed, defined as falling back to a BMI of 16.5 for 2 weeks, or dropped out of the study.
At 25 weeks, 48% of the 49 fluoxetine-treated subjects had either dropped out or relapsed, versus 39% of the placebo controls. At 50 weeks, 58% of the fluoxetine-treated subjects and 55% of the controls had either dropped out or relapsed.
To pick up on any possible benefit of the drug, Dr. Walsh and his colleagues also considered just those who relapsed, and those who relapsed plus those who clinically appeared to be in trouble when they dropped out–either because their BMI had dropped to 17 or because they had begun binging and purging again, at least twice a week.
At 50 weeks, 27% of the fluoxetine-treated patients and 29% of the placebo controls were documented relapsers.
Considering as failures both those with documented relapse and those who clinically appeared to be not doing well when they dropped out, the percentages were 49% fluoxetine and 51% placebo.
Dr. Walsh said he also looked only at those with depression, and at those who purged or restricted food only, and treatment still made no difference in those particular patients. “However we cut it, we can find no evidence that fluoxetine prolongs time to relapse following successful initial treatment, and we've looked pretty hard,” Dr. Walsh said.
In an editorial comment accompanying the published report, Dr. Scott J. Crow of the University of Minnesota, Minneapolis, said, “While the results of previous relapse prevention trials have been mixed, the report by Walsh and colleagues has many strengths and appears convincingly negative” (JAMA 2006;295:2659–60).
The study amply demonstrates that antidepressant therapy, “a fairly common treatment practice for this illness,” provides no benefit, Dr. Crow noted.
Additional reporting was done by contributing writer Mary Ann Moon.
'We can find no evidence that fluoxetine prolongs time to relapse … and we've looked pretty hard.' DR. WALSH