ASPPH logo


Member Research & Reports

Member Research & Reports

UAlbany Study Takes One of the First Widespread Views of Eating Disorders in the United States Since Criteria Changed in 2013

A new University at Albany School of Public Health study takes the first comprehensive look at the prevalence of eating disorders in the United States since changes were made to diagnostic criteria, and some results are not what researchers and clinicians were expecting.

Noting that national rates of eating disorders hadn’t been studied since 2007, Dr. Tomoko Udo, an assistant professor of health policy, management and behavior, partnered with Dr. Carlos Grilo of Yale University School of Medicine to study the prevalence of three disorders — anorexia nervosa, bulimia nervosa and binge-eating disorder as they are classified in the “Diagnostic and Statistical Manual of Mental Disorders” (DSM-5).

The DSM-5 is used by mental health professionals in the U.S. as the standard classification of mental disorders. It was updated in 2013 to replace an earlier edition, DSM-4, marking the first significant revision since previous diagnostic criteria for eating disorders was established in 1994. According to Dr. Udo, in addition to recognizing binge-eating disorder as a formal diagnosis for the first time, the DSM-5 loosened diagnostic criteria — possibly making it easier for someone to be diagnosed with an eating disorder and leading many professionals to believe that the prevalence of eating disorders would be higher than previous estimates.


Drs. Udo and Grilo used data from the National Epidemiologic Survey Alcohol and Related Conditions (NESARC), a cross-sectional database of a nationally representative sample of the civilian, non-institutionalized population of the United States aged 18 years and older.

The study pool consisted of 36,309 American adults who completed computer-assisted, face-to-face personal interviews between April 2012 and June 2013 as part of NESARC. Study participants had reported their sociodemographic characteristics (age, sex, ethnicity, education and income), and participants body mass index was calculated using self-reported height and weight statistics. The participants also answered questions about body image, eating habits and behavior.

Drs. Udo and Grilo created eating disorder diagnostic groups based on the updated criteria from the DSM-5. For example, to be classified as anorexic, a participant would have a low weight in addition to reporting an obsession with their body shape, while for a diagnosis of bulimia nervosa, a subject would have noted regularly trying to prevent weight-gain by extreme methods such as vomiting or using medication.


Using multiple statistical analysis methods to interpret the collected data, Drs. Udo and Grilo found:

A key takeaway from the research, according to Dr. Udo, is that these numbers are relatively consistent with or even lower than the previous study done in 2007. “Many researchers and clinicians expected higher estimates than earlier studies as a result of ‘loosening’ of diagnostic criteria for eating disorders, but this research found that isn’t the case” she said.

Regardless of the estimates being lower than expected, Dr. Udo stresses that eating disorders are an important public health topic and more studies are needed to explain the reason for these unexpected findings and to better understand the impact of changes in diagnostic criteria.

“Though the estimates aren’t as high as anticipated, it’s clear that eating disorders are still common among both men and women and across racial groups” said Dr. Udo. “They occur throughout the lifespan and are associated with serious impairments in psychosocial functioning; as such, they are an important public health problem and screening should be done across all sociodemographic groups.”

The full research, is published in the peer-reviewed journal Biological Psychiatry.