EAT-26: The Eating Attitude Test
I have always had great interest in eating disorders and therefore I thought it would be nice to share some information about the Eating Attitude Test (EAT), or EAT-26.
Before I go in detail about the test, I would like to briefly talk about the eating disorders. According to DSM-5, there are 3 types of eating disorder: Anorexia Nervosa, Bulimia Nervosa, and Binge Eating Disorder. They are serious mental illnesses that are often chronic and difficult to treat. In fact, mortality rates for Anorexia Nervosa are the highest of any psychiatric disorder.
A person with eating disorder may start out with minor difference in their dietary whether less or more than their usual dietary intakes, but at some point, they urge to eat less or more uncontrollably. With serious disturbances in eating behavior, they are also obsessed with weight regulation like inappropriate compensatory behaviors in order to prevent weight gain. (e.g., self-induced vomiting, misuse of laxatives or other medications, fasting, or excessive exercise)
On the McGill Mental Health Service website, I found very striking facts about eating disorders:
- “61% of female American university students have some kind of subclinical eating problem, including chronic dieting, binge/purging and subclinical bulimia” (Mintz & Betz, 1988, as quoted in Yager & O’Dea, 2008).
- “9–21% of male university students suffer from disordered eating and 2% meet the criteria to be diagnosed with clinical Bulimia Nervosa” (O’Dea and Abrahams, 2002, as quoted by Yager & O’Dea, 2008).
Although eating disorders among females are 2 ½ times greater than among males, both genders suffer from eating disorders significantly and chronically. In a study done at Princeton University, scientists found that among patients with lifelong eating disorder problems, 53% say that their disorders first emerged during college.
Enough of information about eating disorder, I want to introduce the EAT (Eating Attitude Test), which was developed by David Garner et al., in response to concerns raised at an National Institute of Mental Health consensus panel after recognizing a need for screening large populations to detect anorexia nervosa or those who are at risk. The EAT-26 is the shorter yet effective version of the original test, EAT-40, which was published in 1982 by Garner et al.. In fact, the EAT-26 is highly correlated (0.98) with the original version.
EAT-26 is probably the most widely used standardized self-report measure of symptoms and characteristics of eating disorders. They have been translated into different languages and used worldwide over 30 years. According to EAT-26, those who score 20 or higher on the test is highly recommended to see a qualified professional to see if they meet the diagnostic criteria for an eating disorder. However, scoring below 20 does not mean that one is completely safe from diagnosis. For example, some people with binge eating disorder may score lower on the EAT-26 but may have a serious eating disorder. Therefore, the score should not be used to conclude whether you have eating disorder or not.
As Garfinkel mentions, “The EAT has excellent psychometric properties, including good sensitivity and specificity. According to Mann et al. (1983), a threshold of 20 (on the EAT-26) yielded a sensitivity of 88%, and a specificity of 96%. The test retest reliability of the EAT is also very good (0.84). Another strength is that this test has been widely used in variety of cultures and age groups.”(2). The EAT-26 is used in both clinical and non-clinical setting in order to figure out whether an individual needs to be referred to a specialist. The test is also available online and can be easily accessed. I have included a link to EAT-26 at the bottom of my post so give it a try if you are interested and it’s free! But I want to emphasize again that this is not for diagnosing an eating disorder. This should be considered as the first step in a two-stage screening process.
Although the EAT-26 has good psychometric properties of reliability and validity, and reasonable sensitivity and specificity for the eating disorders, there are some limitations to the instrument. The EAT-26 is good at identifying a core psychopathology but it may fail to recognize some significant symptoms in attitude and behavior because they fluctuate considerably. A self-report instrument, such as EAT-26, is often influenced by a person’s attitude. For example, if a patient is asked to fill out the form in a clinical setting where other people are present, the social expectations may evoke a different response compared to filling out the survey at home. Not surprisingly, about 15% of people with anorexia may score below the cutoff point due to their denial of illness.
The prevalence of eating disorders is relatively low in the population, which results in inefficient screening. EAT high scorers include many people without full-syndrome anorexia or bulimia nervosa. Consequently, the positive predictive value is relative low. In order for a test to be efficient in detection or high in its sensitivity and specificity, a disorder must have a prevalence of approximately 20%. Even though the specificity are good for the EAT, they are not efficient in detecting eating disorders with prevalence of 2–4% of populations of young women. Comorbidity may also influence the result and interfere the accuracy of screening. For example, people with type 1 DM (Diabetes mellitus) scored higher than control group in EAT.
Although EAT may not be effective in detecting specific eating disorders such as anorexia nervosa or bulimia nervosa, it is still valuable in screening for eating disturbance in general, in the population at risk overall, or in specific sub-groups [e.g., athlete team, college students with type 1 DM (Diabetes mellitus)]. In order to use the test more efficiently, one must understand the limitations of the assessment.
Eating Attitudes Test (EAT-26):
