LOS ANGELES – There is very little evidence – and in some cases none at all – to support common pharmacologic treatments for anorexia and bulimia nervosa, according to Michael Strober, Ph.D.
Eating disorders "are the black hole of pharmacotherapy in psychiatry. There’s not much at this point in time we have to offer," Dr. Strober, director of the eating disorders program at the UCLA Neuropsychiatric Institute and Hospital, reported at a psychopharmacology update sponsored by the American Academy of Child and Adolescent Psychiatry.
For anorexia nervosa, evidence is stronger for family-based therapy in which families are taught to help a child eat and maintain weight.
Cognitive-behavioral therapy (CBT) is more effective than drugs for bulimia nervosa and remains its treatment of choice, said Dr. Strober, who estimated he’s treated more than 8,000 eating disorder patients during a career dating back to 1975.
Although selective serotonin reuptake inhibitors (SSRIs) are widely used to treat anorexia, "there is absolutely not a speck of evidence that these drugs, in the acutely malnourished state, are effective," or that they help maintain weight gain, said Dr. Strober, who also serves as director of the adolescent mood disorders program at the institute.
There is slight evidence for the atypical antipsychotic, olanzapine (Zyprexa).
In a trial funded by the drug’s maker, Eli Lilly, 16 anorexia patients were randomized to outpatient day treatment plus olanzapine, and 18 to day treatment plus placebo.
The mean body mass index was low-normal in both groups at the end of 13 weeks (19.66 kg/m2 in the placebo group vs. 20.30 in the olanzapine group). However, placebo patients started with a lower mean BMI of 15.93, compared with 16.39 among olanzapine patients.
Also, at week 13, 55.6% of placebo patients but 87.5% of olanzapine patients achieved the target BMI of 18.5 (Am. J. Psychiatry 2008;165:1281-8).
The differences were "not very dramatic," but were statistically significant, Dr. Strober said.
"We have no idea what the mechanism is. We don’t know if it’s mediated by improved psychological state or weight gain associated with olanzapine," he said.
Given the lack of data, Dr. Strober said that he and his colleagues typically refeed anorexia inpatients for a few weeks, and only then add atypical antipsychotics if they aren’t managing well.
"I generally think [they] have more of an effect than the SSRIs, although we have patients receiving SSRIs who report less anxiety and psychological discomfort with weight gain," Dr. Strober said.
In general, however, "I have never seen a robust clinical effect of these agents," he said.
For bulimia, antidepressants have been shown to help.
"Regardless of the antidepressant used, in the majority of published studies – we really don’t know how many unpublished studies have been buried – active drug separates from placebo," Dr. Strober said.
"What this means is that about 60%-65% of people randomized to active treatment report a response (defined as a 50% or greater reduction in the frequency of binge eating)," compared with placebo responses of about 35%-40%, he said.
"We get remission rates – which means no bingeing or purging the last 2 weeks of the trial – of about 10% on active drug, maybe 0%-3% on placebo," he said.
However, about 50% of responders "have a recurrence of their symptoms" within 5 months, "so these are not very effective drugs for the majority of patients," Dr. Strober said.
CBT does better: "After 12-16 weeks, we get response rates (defined in the same way) that are roughly 65%-70%. You get about 35%-38% that are fully remitted, and that response tends to be more durable," he said.
Evidence also supports family-based therapy (FBT) for anorexia nervosa.
In a recent study, 121 subjects aged 12-18 years with mild to moderate illness – the trial excluded patients below 75% of their ideal body weight – were randomized to either individual therapy or FBT.
At the end of treatment and at 12 months’ follow-up, body weights were normal in about half of FBT patients but only in about a quarter of individual therapy patients (Arch. Gen. Psychiatry 2010;67:1025-32).
"There’s a signal here, I think – a powerful signal," Dr. Strober said. "This is a meaningful difference."
Family-based therapy "is an important treatment. We don’t really know what the ingredient is; we just know that this has been helpful for a number of families, but you really have to be very skilled in working with these people," he said.
Dr. Strober said he has no disclosures.