All Eating Disorders Can Be Deadly, Anorexia the Worst

Deborah Brauser

July 27, 2011

July 27, 2011 — Patients with an eating disorder of any type have a significantly increased risk for death, but anorexia nervosa appears to be particularly deadly and linked to the highest mortality and suicide rates, new research shows.

In a new meta-analysis, similarly elevated mortality rates were found for those with bulimia nervosa and eating disorder not otherwise specified (EDNOS). However, the rate was even higher for those with anorexia nervosa, with a weighted annual rate of 5 deaths per 1000 person-years. Of those who died, 1 in 5 did so by committing suicide.

In addition, an older age at first presentation for those with anorexia, especially between the ages of 20 and 29 years, was found to be a significant predictor of mortality.

"It was not surprising to find out that mortality in eating disorders, particularly [anorexia], was high. It was, however, surprising to find out the high levels of deaths by suicides among this population," lead author Jon Arcelus, PhD, from the Eating Disorders Service in Leicester and Loughborough University, United Kingdom, told Medscape Medical News.

Dr. Jon Arcelus

"This study gives clinicians some information about predictive factors they can use in their day-to-day work. It should also give the primary care doctors and the general population a message that [eating disorders] are serious and the quicker they are treated the better," added Dr. Arcelus.

The study was published in the July issue of the Archives of General Psychiatry.

Previous Research 'Confusing'

Most previous research looking at mortality and eating disorders have focused on those with anorexia, with the standardized mortality ratio (SMR) varying widely, according to the investigators.

This study gives clinicians some information about predictive factors they can use in their day-to-day work. It should also give the primary care doctors and the general population a message that [eating disorders] are serious and the quicker they are treated the better.

They note that other studies have suggested a low mortality risk for bulimia, which is "surprising" because of the medical complications associated with purging behaviors.

"We were aware that eating disorders, particularly [anorexia] had high mortality rates. However, the message from the literature was very confusing. Our aim was to clarify this, to come with the best available figure of mortality, and to investigate whether we could say something about predictive factors," said Dr. Arcelus.

His team evaluated data from 36 peer-reviewed articles that included mortality rates for patients with eating disorders and were published between January 1966 and September 2010.

The studies reported outcomes of specific disorders during person-years, including 166,642 total for anorexia, 32,798 for bulimia, and 22,644 for EDNOS.

The researchers examined both weighted mortality (deaths per 1000 person-years) and SMRs (ratio of observed to expected deaths).

Highest Mortality for Anorexia

Results showed that the highest mortality rates were found for those with anorexia (weighted mortality, 5.1; SMR, 5.86).

Of the 12,808 total patients with anorexia, 639 died (mean follow-up period, 12.82 years). Among these, 1.3 deaths per 1000 person-years were from suicide.

The weighted mortality rates and SMRs were 1.74 and 1.93, respectively, for bulimia and 3.31 and 1.92 for EDNOS. There were 57 deaths among 2585 total patients with bulimia and 59 deaths among 1879 patients with EDNOS. However, only 6 of the studies chosen reported EDNOS mortality data.

"Despite the relatively small number of studies, the examination of this group is important given that these patients represent such a large proportion of patients observed in practice," explain the investigators.

Although age at first assessment was found to be a significant mortality risk factor for those with anorexia (P = .01), prognostic factors were not assessed for bulimia or EDNOS because of the small number of studies looking at these specific disorders.

Finally, there was no significant differences in observed mortality between the bulimia and EDNOS group, but a 2.7-fold higher rate was found for the anorexia group compared with the bulimia group.

"Future robust studies should inform physicians of the predictive factors associated with mortality rate in patients with EDNOS and [bulimia]; so far, late presentation of [anorexia] appears to be the only clear predictor of death among these disorders," write the investigators.

They also note that the SMR they found for patients with anorexia was "much higher" than for other psychiatric disorders, reporting that past studies have found SMRs in male and female patients of 2.8 and 2.5 for those with schizophrenia, 1.9 and 2.1 for those with bipolar disorder, and 1.5 and 1.6 in those with unipolar disorder, respectively.


"This study adds to the burgeoning literature on the extent of mortality rates with all eating disorders," Kathryn Zerbe, MD, director of the Oregon Psychoanalytic Institute in Portland, told Medscape Medical News.

"As the study concludes, and we have really suspected this for some time, eating disorders have the highest mortality rate of really any psychiatric illness," said Dr. Zerbe.

She noted that this meta-analysis looked only at studies where the specific eating disorder was known.

"One of our concerns in mental health, especially for women, is the number of EDNOS patients who are never diagnosed; and their mortality rate is charted as something else. So how many of these get missed in general medicine? If they're never diagnosed, they never get into a program and are never written about."

Dr. Zerbe, who was not involved with this study, has twice been on the American Psychiatric Association's Practice Guidelines Task Force for eating disorders, including the most recent edition. She explained that the diagnostic criteria for EDNOS "has shifted" over the years.

This paper raises awareness that all eating disorders across the board must be treated as life-threatening illnesses.

"It used to be a catch-all term. As research has gone on, we actually expanded it to include what I call 'subclinical anorexia.' And I see a lot of these people in my practice, those who don't meet all the criteria for [anorexia or bulimia], including those who binge but don't purge," she said.

"What is nice about this paper is that it raises awareness that all [eating disorders] across the board must be treated as life-threatening illnesses."

The study authors and Dr. Zerbe have disclosed no relevant financial relationships.

Arch Gen Psychiatry. 2011;68:724-731. Abstract


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.