ABSTRACT
Objective
Our aim was to evaluate the diagnostic accuracy of the Eating Disorder Examination Questionnaire (EDE-Q) scores and establish general and context-specific cut-off points for anorexia nervosa (AN) in Australian women.
Methods
The sample included 67 women with AN, 41 women with a history of AN and restored weight, 100 women with self-reported AN, 27 sisters of individuals with AN, 73 healthy controls, and 597 community controls. ROC analysis assessed the EDE-Q’s accuracy and determined cut-off scores in different contexts.
Results
The EDE-Q showed varying discriminative ability, with cut-off scores differing by context. In the most realistic scenario–diagnosed AN versus community controls drawn from the general population, reflecting typical screening contexts–, the Area Under the Curve (AUC) was 0.88, and a cut-off score of 2.775 yielded sensitivity of 0.91 and specificity of 0.71, recommended as the general cut-off for AN in Australian women.
Conclusion
These findings indicate the EDE-Q is useful for identifying individuals at risk of AN but has limited diagnostic capacity due to moderate specificity. While the EDE-Q can guide initial screening using a 2.775 cut-off on the Global score, it is primarily a screening tool and should be complemented by further assessment to confirm diagnosis. Overall, our study provides evidence-based guidance for establishing a general EDE-Q cut-off for AN and informs the definition of context-specific thresholds.