Mortality From "Mild" Eating Disorders Greater Than Rates for Anorexia and Bulimia

Pauline Anderson

October 23, 2009

October 23, 2009 — Although "eating disorders not otherwise specified" (EDNOS) are often considered relatively mild, the associated mortality rate for these disorders exceeds that for anorexia nervosa and bulimia, new research shows.

This new finding may change the thinking surrounding EDNOS, which have traditionally been a catch-all label for eating disorders that do not fit into the categories of anorexia or bulimia nervosa, said the study's lead author, Scott J. Crow, MD, professor of psychiatry, University of Minnesota, Minneapolis.

"Some facilities view [EDNOS] as less severe and less in need of treatment than bulimia or anorexia, and that just doesn't make much sense according to our data," he told Medscape Psychiatry.

The researchers also found that suicide was an important cause of death for bulimia and other eating disorders.

The study was published online October 15 in the American Journal of Psychiatry.

Subjects Predominantly Women

The study included 1885 patients who were evaluated at the Outpatient Eating Disorders Clinic at the University of Minnesota between 1979 and 1997. Each patient completed an Eating Disorders Questionnaire to assess demographic variables, eating-disordered behaviors, and treatment experiences, among other things.

Researchers calculated body mass index (BMI) from self-reported weight and height. They defined anorexia as a BMI of 17.5 kg/m2 or less and a fear of weight gain, and bulimia as a BMI of more than 17.5 kg/m2 with recent regular binge-eating episodes plus use of diuretics, laxatives, or enemas; vomiting; fasting; or exercise. All other patients were considered to have an EDNOS.

To obtain information on mortality, researchers accessed the National Death Index database. They classified causes of death into 4 categories: suicide, substance-use related, traumatic, and medical.

Not surprisingly, the majority of participants in the study (95.1%) were women. The most common diagnosis was bulimia (48.1%), followed by EDNOS (42.5%) and anorexia (9.4%).

Mortality Rates Surprising

The EDNOS mortality rate was 5.2%. In comparison, the mortality rate for bulimia was 3.9% — higher than rates reported in other studies. This could be because other studies used a weaker method for establishing mortality and followed-up patients for a shorter period of time, said Dr. Crow. As well, the study may have captured patients who actually had anorexia but who had met criteria for bulimia at the time of their death.

The mortality rate for anorexia (4.0%) was lower than that reported in some other studies, said Dr. Crow.

The elevated mortality rate among patients with an EDNOS was largely contained in certain subpopulations, and the researchers are currently analyzing these data.

Some of these high mortality rates were surprising, said Dr. Crow. "Looking at it from the historical perspective, people have tended to say that you just don't see mortality in bulimia or in EDNOS."

This new information may be useful when considering changes to diagnostic categories in the revised Diagnostic and Statistical Manual of Mental Disorders, 5th edition, said Dr. Crow.

Suicide in Bulimia

Another significant finding was that suicide was an important cause of death in bulimia as well as the other 2 groups (0.9% of deaths in bulimia compared with 0.5% in EDNOS and 0.6% in anorexia). The prevailing opinion has been that the suicide rate among patients with bulimia was lower than for other eating disorders, said Dr. Crow.

This study did not include extensive detail on causes of death among patients with eating disorders, but Dr. Crow said such patients may be at risk from complications such as blood electrolyte problems, heart attack, and starvation in the case of anorexia, and vomiting and excessive laxative use in the case of bulimia.

The diagnoses used in this study were based on patient self-report data from the Eating Disorders Questionnaire that do not provide a specific illness-onset date. Other limitations of the study were that it was conducted at only one site, did not include patients with binge-eating disorder, and as with all attempts to determine cause of death, suicide rates may have been underreported.

New Findings on Bulimia

Asked by Medscape Psychiatry to comment on the findings, Graham Redgrave, MD, assistant director, Eating Disorder Program, John Hopkins University School of Medicine, Baltimore, Maryland, said he found the study "fascinating" and significant because of its size, the inclusion of a large number of patients with bulimia, and its rigorous ascertainment of cause of death.

"In some ways I was most surprised about the bulimia statistics because they have never been shown before; as far as I'm aware, both the increased mortality and increased risk of suicide in bulimia are new findings," he said.

Dr. Redgrave pointed out that because the study included only outpatients, the mortality rates for eating disorders that it uncovered might not reflect the true picture. But they do represent typical patients in the community. "Bulimia is much more likely to be treated in an outpatient center, so it's particularly helpful for us to be able to tell our patients with bulimia that there is in fact an increased [mortality] risk."

Dr. Redgrave agreed that some cases of EDNOS can be serious. He used the example of a woman who quickly drops from a BMI of 35 kg/m2 to a BMI of 20 kg/m2. "That's a loss of 15 BMI points, which may be incredibly stressful and increase risk for death, even though that woman may be at normal weight."

Dr. Crow reports he has received research support from GlaxoSmithKline, Ortho-McNeil, and Pfizer and speakers honoraria from Eli Lilly. Study investigator Carol B. Peterson, PhD, receives royalties from Guilford Press. All other authors have disclosed no relevant financial relationships.

Am J Psychiatry. Published online October 15, 2009.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.