The report in the Journal of the American Medical Assn. sent researchers back to square one in the search for drugs to treat the intractable disorder, which affects mostly women and adolescent girls.
"This is not good news," said Marsha Marcus, an eating disorder specialist at the University of Pittsburgh who was not involved in the study.
Get developments in medicine, nutrition and fitness delivered to your inbox with our The Health Report newsletter. Sign up »
"What this highlights for me is that we really desperately need new treatments."
People with anorexia are obsessed with their body weight and diet to the point where they become dangerously thin. Half the people with the disorder binge on food and then purge by vomiting or using laxatives or diuretics. The other half restricts the food they eat and excessively exercises.
The disorder affects 1% to 2% of females in the U.S. and leads to death in an estimated 10% of cases.
Standard treatment for the disorder is behavioral therapy, and in severe cases patients are hospitalized to restore their weight. The relapse rate in the worst cases is high — up to half of hospitalized patients are readmitted within a year.
The study set out to determine whether Prozac, generically known as fluoxetine, would enhance the benefits of behavioral therapy. Researchers looked at Prozac because it was commonly prescribed for anorexia nervosa, although the drug had not been approved for the condition.
Dr. B. Timothy Walsh of Columbia University and lead author of the report said there were reasons to believe Prozac would be effective.
The drug had been shown to prevent relapses in patients with bulimia, an eating disorder with some similarities to anorexia, he said. People with bulimia binge-eat and purge but maintain a normal weight. In addition, Walsh said, a small previous study in anorexia nervosa patients had hinted that Prozac could postpone relapses.
The latest research, conducted at Columbia University and the University of Toronto, included 93 females ages 16 to 45 with the eating disorder. Patients entered the study after regaining weight to a minimum body-mass index of 19 after intensive inpatient or outpatient treatment.
BMI is a standard measure that takes into account a person's weight and height. A BMI below 18.5 is considered underweight for an adult. A 122-pound, 5-foot-8 adult would have a BMI of 18.5.
All patients in the study received intensive behavioral therapy consisting of 50, 45-minute sessions with experienced psychologist or psychiatrists. In addition, all patients could choose up to five supplemental sessions with family or friends.
About half the patients in the study received Prozac. The others were given placebos. Each patient remained in the study for one year.
There was no significant difference between the groups in the percentages of patients who maintained their body weight and completed the study. Only 26.5% of the Prozac group and 31.5% of the placebo group managed to complete the 52-week study and maintain a BMI of at least 18.5.
Researchers also said there were no statistically significant differences between the groups on standard measures for depression, self-esteem, quality of life and body dissatisfaction, among other things.
"What was remarkable was the absence of even a hint of benefit," Walsh said.
The result was disappointing, he said. "We don't have any evidence that any medication is a useful treatment, and that is not a good position to be in."
Dr. Scott J. Crow of the University of Minnesota, in an editorial accompanying the report, called for more anorexia research.
Most new weight and appetite drugs are to "diminish, not increase food intake," he wrote.