Discriminant validity of the Eating Attitudes Test according to American Psychiatric Association and World Health Organization criteria of eating disorders.

The purpose of this study was to assess the sensitivity, the specificity, and the positive predictive value of the Eating Attitudes Test in a sample of Spanish nonclinical 18-yr.-olds. 304 subjects answered the Eating Attitudes Test-40, 290 of whom were interviewed individually with the Spanish version of the Schedules for Clinical Assessment in Neuropsychiatry. Eating disorders were diagnosed using ICD-10 and DSM-III-R criteria. The prevalence of eating disorders was higher for ICD-10 (5.2%) than for DSM-III-R (2.6%) and only affected the rate of diagnosis in women. According to ICD-10 criteria, the cut-off of 25 was more sensitive (87.5%) than the cut-off of 30 (75%) and varied little in specificity (93.9% vs 97.1%). The positive predictive value of the Eating Attitudes Test cut-off of 30 for eating disorders (ICD-10) was 36%. Our results support the test as useful for identifying eating disturbances in 18-yr.-olds and suggest assessment of a cut-off lower than 30 may be appropriate in the general population if confirmed in further research with a representative sample of adults.