Eating Disorders Common Across All Ages, Sexes, Ethnicities

Batya Swift Yasgur, MA, LSW

April 25, 2018

Eating disorders (EDs) affect people of all ages, sexes, and ethnic/racial groups and are associated with significant psychosocial impairment, new research shows.

Investigators analyzed data from more than 35,000 US adults and found that women had significantly higher odds of having EDs than men and that whites had greater odds of haivng anorexia nervosa (AN) than their non-Hispanic black and Hispanic counterparts.

On the other hand, the odds of having bulimia nervosa (BN) did not significantly differ by race/ethnicity, and fewer non-Hispanic black than Hispanic and white respondents had lifetime binge eating disorder (BED).

All three disorders were associated with substantial psychosocial impairment.

"Although eating disorders may not be as prevalent as some other psychiatric disorders, such as depression, anxiety, or alcohol and drug use disorders, they are common and found in men and women across ethnic/racial groups and occur throughout the lifespan," lead author Tomoko Udo, PhD, assistant professor, Department of Health Policy, Management, and Behavior, University at Albany School of Public Health, New York, told Medscape Medical News.

"Binge eating disorder, a new 'formal' diagnosis in the DSM-5, is important to screen for and identify, as it is associated with substantially increased risk of obesity, and all eating disorders are associated with impairments in psychosocial functioning and thus represent an important public health problem," said Udo, speaking not only for herself but also on behalf of her coauthor, Carlos M. Grilo, PhD.

The study which was published online April 17 in Biological Psychiatry.

New Prevalence Estimates

"There exist few nationally representative population-based data on the prevalence of EDs" in the United States, the authors write.

Previous studies used DSM-IV criteria to assess the prevalence of EDs, the authors observe. Moreover, ED prevalence rates across ethnic/racial groups were calculated by pooling data from several different samples.

"Data from large-scale nationally-representative samples assessed with diagnostic interview is required to update prevalence estimates of EDs in the US," they add.

This is particularly necessary because of updated diagnostic criteria of EDs in the DSM-5. Changes include BED becoming a formal diagnosis and lowering the frequency of bingeing in diagnosing BED.

"We felt it was important to obtain new prevalence estimates in a larger and representative sample, especially because the DSM-5 included several changes to the criteria for EDs from the earlier DSM-IV," said Udo.

"Many researchers and clinicians expected higher estimates than those found in earlier studies as a result of 'loosening' of diagnostic criteria for EDs," she noted.

The current study used data from the 2012-2013 National Epidemiologic Survey Alcohol and Related Conditions (NESARC-III) (n = 36,309 respondents aged ≥18 years) assessed with lay-administered diagnostic interviews.

Respondents provided sociodemographic information (age, sex, ethnicity, education, and income).

They also reported height and weight, which were used to calculate body mass index (BMI).

The NIAAA Alcohol Use Disorder and Associated Disabilities Interview Schedule–5 (AUDADIS-5) was used to assess AN, BN, and BED. Data included age at onset and age at the most recent episode.

AUDADIS was also used to assess impairment in social function due to EDs. The factors that were assessed included interference with normal daily activities, serious problems getting along with others, and serious problems fulfilling responsibilities.

Most Common EDs

The researchers created specific ED diagnostic groups (AN, BN, BED) on the basis of DSM-5 criteria, using the NESARC-III responses to relevant AUDADIS-5 questions and rescoring the NESARC-III variable data to create the diagnostic categories for the present study.

Prevalence estimates of lifetime ED showed AN and BED to be the most common EDs, with prevalence rates were 0.80% (SE, 0.07%) and 0.85% (SE, 0.05%). The prevalence rate of BN was 0.28% (SE, 0.03%).

Prevalence estimates of 12-month AN, BN, and BED were 0.05% (SE, 0.02%), 0.14% (SE, 0.02%), and 0.44% (SE, 0.04%), respectively.

Of the 0.22% of respondents who reported comorbid EDs (ie, having a lifetime diagnoses of two or more specific EDs), 0.01% reported comorbid AN plus BN, 0.02% reported AN plus BED, 0.13% reported BN plus BED, and 0.05% reported all three EDs.

For all three EDs, the prevalences of lifetime and 12-month diagnoses were significantly greater among women than men, with unadjusted estimates of 1.42%, 0.46%, and 1.25% in women vs 0.12%, 0.08%, and 0.42% in men for AN, BN, and BED, respectively.

There were also significant differences in prevalence between ethnicities/races in lifetime AN; adjusted odds ratios (AORs) were significantly lower for non-Hispanic black and Hispanic respondents in comparison with non-Hispanic white respondents.

However, AORs of 12-month AN were significantly lower for Hispanic than for non-Hispanic white respondents. There were no cases of 12-month AN among non-Hispanic black respondents.

Although the lifetime prevalence and the 12-month prevalence of BN did not differ significantly by race/ethnicity, AORs for lifetime BED were significantly lower for non-Hispanic black than for non-Hispanic white respondents.

There were no racial differences in 12-month BED, nor was educational level significantly associated with differences in ED prevalence.

By contrast, higher income categories were associated with significantly increased odds of lifetime AN.

After adjusting for age, sex, ethnicity/race, and educational level, AN and BED were more prevalent among younger groups than among persons aged 60 years or older.

No "Overpathologizing"

All eating disorders were associated with high rates of impaired psychosocial function.

For lifetime diagnoses, rates of any impairment in social function were significantly greater for persons with lifetime BN (61.4%) and BED (53.7%) in comparison with persons with AN (30.7%).

On the other hand, the only significant psychosocial difference in 12-month diagnosis was for BN; respondents with BN reported greater difficulties than those with BED in getting along with others.

For lifetime as well as 12-month diagnoses, people with AN had significantly lower current BMI than did those with BN and BED. For lifetime and 12-month diagnoses, those with BN had a significantly lower current BMI than did those with BED.

Moreover, persons with lifetime AN had significantly greater odds of being underweight or of being of normal weight compared with persons with lifetime BED. In addition, their odds of being overweight or obese were significantly reduced compared with persons with lifetime BED, which was associated with significantly increased odds of obesity and extreme obesity.

The authors note that their prevalence estimates "are at odds with critics' view of DSM-5, who used BED as an illustration of over-pathologizing."

Udo acknowledged being "surprised" by the findings.

"We thought we would see increased prevalence in all EDs because of changes in the diagnostic requirements, particularly greater increases for BN and BED, which we did not observe, and in fact, our estimates for both BN and BED were lower than previous studies and what we expected."

Better Screening Needed

Commenting on the study for Medscape Medical News, Kendrin Sonneville, ScD, RD, assistant professor of nutritional sciences and research assistant professor, Center for Human Growth and Development, University of Michigan School of Public Health, Ann Arbor, who was not involved with the study, said that "studies like this one, conducted in a national sample, help us understand who is suffering with eating disorders and who may not be getting help for their eating disorder."

A "key message" is that "people of all ages, gender, and racial/ethnic groups are affected by EDs" and "experience substantial impairment, with many suffering for an extended period of time," she said.

Sonneville urged practicing clinicians and health systems to "identify strategies to detect eating disorders early." She noted that early detection and referral to treatment are "necessary for improving health and reducing suffering of people with EDs."

Udo added, "We believe our study highlights the complexity of assessing and recognizing eating disorders, while documenting their clinical and public health impact."

Because specialty treatment for EDs is "not as easily accessible in many communities as it is for other psychiatric and medical conditions, improving access to care is an important priority for healthcare systems," she said.

Dr Grilo's work was supported in part by the National Institutes of Health. Dr Grilo has disclosed no relevant financial relationships. His coauthors' disclosures are listed in the original article. Dr Udo and Dr Sonneville have disclosed no relevant financial relationships.

Biol Psychiatry. Published online April 17, 2018. 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.