ABSTRACT
Introduction
The Eating Disorder Examination-Questionnaire (EDE-Q) is one of the most widely used measures of eating disorder psychopathology; however, its original four-factor structure is not supported in data-driven investigations. This study examined the latent structure of the EDE-Q in a large sample of inpatients with anorexia nervosa (AN) at the beginning and end of hospitalization.
Method
Inpatients with AN (ages 13–64 years) were included. Exploratory factor analysis (EFA) was conducted using EDE-Q data collected at admission (n = 320). Confirmatory factor analysis (CFA) was conducted using data collected just prior to discharge (n = 301). Reliability was assessed using internal consistency indices and construct validity of the proposed EDE-Q subscales was explored using correlational analyses with the eating disorder inventory (EDI) and the eating pathology symptoms inventory (EPSI).
Results
The original four-factor structure was not supported. A three-factor model (Body Image Dissatisfaction, Discomfort with Body Exposure, and Restrictive Eating Concerns) emerged from the EFA and was supported in CFA with improved model fit (CFI = 0.928, TLI = 0.915, RMSEA = 0.087, SRMR = 0.055). The model showed acceptable internal consistency and convergent validity.
Discussion
The EDE-Q was found to have a three-factor structure, comprising two body image-related domains and one domain assessing restriction and concern about eating. Findings support the emergence of a distinct factor capturing discomfort with body exposure, independent of body dissatisfaction. This highlights the need for empirically derived subscales that better disentangle overlapping eating disorder constructs.