Eating disorder diagnoses may be too narrow

  • Share
  • Read Later

The majority of patients diagnosed with eating disorders do not have bulimia or anorexia, according to criteria from the current Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), but instead suffer from what are known as “eating disorders not otherwise specified” (EDNOS)—illnesses effectively defined by what they aren’t. Yet, insufficient study of these disorders, together with a lack of understanding of how their severity compares with that of anorexia and bulimia may mean that patients aren’t getting the care they need, according to new research published in the journal Pediatrics. In a study of 1,310 female patients between the ages of 8 and 19 who had been diagnosed with either anorexia, bulimia or EDNOS, researchers from Stanford University School of Medicine and Duke University Medical Center analyzed how specific health outcomes varied in patients with different diagnoses. They found that, though nearly two thirds of patients studied were correctly diagnosed as suffering from EDNOS, of those patients, more than 61% met criteria for hospitalization, and most had more significantly deteriorated health than patients diagnosed with bulimia.

In order to narrow the findings to more specific subcategories of EDNOS, researchers divided patients into groups whose symptoms were consistent with partial anorexia or partial bulimia. Among those with partial bulimia, there were three additional subdivisions: partial bulimia binge/purge, characterized by binging and purging, but not at the frequency of a bulimia diagnosis; partial bulimia binge only, similar to binge eating disorder, but not meeting the frequency of that diagnosis; and partial bulimia purge only, patients who don’t alter eating habits, but purge regularly. For those with partial anorexia, there were also three subcategories: partial anorexia with low weight but no menstrual symptoms; partial anorexia with low weight, lack of menstruation but no acknowledgment of psychiatric criteria; and partial anorexia in patients whose body weight was higher than the 85th percentile for their peers, but lower than the 90th. Lastly, researchers also categorized EDNOS patients who didn’t fit into either of the other categories but had lost 25% of their pre-illness body weight, but did not weigh less than 85% of their peers. (The previous edition of the DSM had included patients who met that criteria as anorexic, but the current issue of the manual does not.)

Researchers found that patients in this last category had more significant adverse health outcomes than those in other EDNOS categories, and even more severe than some patients diagnosed as anorexic. In general, they also found that patients with partial bulimia subcategories of EDNOS tended to be younger, suffer the disease for a shorter period of time, weigh less and lose weight more quickly than those who met the diagnostic criteria for bulimia. Yet, in terms of health outcomes, there was little significant difference between those diagnosed with bulimia and those the researchers identified as partially bulimic. Researchers found that, in many instances, though patients with widely varying symptoms were lumped together as having an “eating disorder not otherwise specified” they often tended to have fewer symptoms and health outcomes in common with one another than they did with patients suffering from bulimia or anorexia.

The findings support “the idea that EDNOS is too heterogeneous a category,” the authors argue, adding that “EDNOS patients who narrowly miss criteria for [anorexia] or [bulimia] are often medically compromised and in need of treatment.” Future research is necessary to better determine how patients who do not meet diagnostic criteria for anorexia or bulimia should be diagnosed and treated, the authors say, but in the meantime, the study results “imply the need to better delineate predictors of complications and medical protocols separately,” the conclude.