Summary with Links to the full article (p. 1342) & review p. 1309.:
A large, long-term study extends the finding of high death rates in anorexia nervosa to bulimia nervosa and other eating disorders. Crow et al. (p. 1342) determined diagnoses for 1,885 outpatients with eating disorders evaluated between 1979 and 1997 and searched the National Death Index for matches through 2004. The crude mortality rates for the patients with diagnoses of anorexia nervosa, bulimia nervosa, and "eating disorder not otherwise specified" were 4.0%, 3.9%, and 5.2%, respectively. Compared to national mortality data for demographically similar groups, the rate for eating disorder not otherwise specified was significantly elevated, suggesting that this diagnosis does not indicate a less severe disorder. In addition, 13 of the 84 deaths identified were due to suicide, and eight of these were among the patients with bulimia nervosa. These findings are discussed by Dr. Walter Kaye in an editorial on p. 1309.
OBJECTIVE: Anorexia nervosa has been consistently associated with increased mortality, but whether this is true for other types of eating disorders is unclear. The goal of this study was to determine whether anorexia nervosa, bulimia nervosa, and eating disorder not otherwise specified are associated with increased all-cause mortality or suicide mortality. METHOD: Using computerized record linkage to the National Death Index, the authors conducted a longitudinal assessment of mortality over 8 to 25 years in 1,885 individuals with anorexia nervosa (N=177), bulimia nervosa (N=906), or eating disorder not otherwise specified (N=802) who presented for treatment at a specialized eating disorders clinic in an academic medical center. RESULTS: Crude mortality rates were 4.0% for anorexia nervosa, 3.9% for bulimia nervosa, and 5.2% for eating disorder not otherwise specified. All-cause standardized mortality ratios were significantly elevated for bulimia nervosa and eating disorder not otherwise specified; suicide standardized mortality ratios were elevated for bulimia nervosa and eating disorder not otherwise specified. CONCLUSIONS: Individuals with eating disorder not otherwise specified, which is sometimes viewed as a "less severe" eating disorder, had elevated mortality risks, similar to those found in anorexia nervosa. This study also demonstrated an increased risk of suicide across eating disorder diagnoses.
My Two Cents:
1. My usual complaint: The population from this study only includes people who are from an OP clinic in the University of Minnesota. The racial make-up looks alright - everyone got in there in one way or another. 5% of the subjects were male. Still, it would have been nice to see a pool of different OP & IP clinics.
2. The National Death Index - Good idea. That eliminates the task of having to follow-up with subjects and loosing track of them.
3. Well well well, Suicide is NOT the main cause of death. Only 15% of the AN deaths, 23% of the BN deaths, and 9% of the EDNOS deaths were suicides. Most of the deaths were medical. Wow.
4. The authors do mention that the results should be taken for what they are - results based on an outpatient population. I appreciate that. They go on to say that the OP clinic was not by any means the Mecca of ED treatment - they mention a similar analysis done with AN patients who were seen at a local ER that had a high mortality rate, suggesting that the "less sick" (ie, in need of a lower level of care) patients are seen OP first.
"The common perception of eating disorder not otherwise specified as a "less severe" eating disorder thus seems unwarranted."
"Eating disorders are still not considered a serious form of mental illness in some states and countries, resulting in a health care crisis for those currently suffering from these disorders, as well as their families. This is puzzling, since there is considerable evidence that eating disorders are highly heritable, biologically based, severe mental disorders...
As the authors note, it is well recognized that substantial medical complications are associated with vomiting, laxative abuse, and other purging behaviors. Moreover, the high suicide rate in bulimia nervosa is consistent with this disorder’s association with impulsivity and high comorbidity with mood, anxiety, and substance use disorders. In summary, the elevated mortality risks for bulimia nervosa and eating disorder not otherwise specified were similar to those for anorexia nervosa. These findings underscore the severity and public health significance of all types of eating disorders." - Dr. Kaye's review.
Tell me if the links don't work. I'll fix them. You can also download the full PDF files at the AJP page.
EDIT: For some reason it looks like the full text of the actual study isn't up anymore. That fast...