Objective
This study presents a psychometric evaluation of the avoidant/restrictive food intake disorder (ARFID) module 2.0 for the Eating Disorder Examination (EDE), its child (ChEDE), and parent version. Within a pediatric sample seeking treatment for restrictive feeding or eating disorders and non‐treatment‐seeking controls, the module's interrater reliability, parent–child agreement, and its convergent, divergent, and discriminant validity were examined.
Method
The child, adult, and/or parent version of the German ARFID module was administered to N = 176 children and adolescents (0–17 years) and their parents, as were the (Ch)EDE, well‐established measures on food‐avoidance behaviors, food variety, and body esteem, and objective anthropometric measures.
Results
Across all versions of the ARFID module, substantial to almost perfect interrater reliability was shown. Parent–child agreement for ARFID diagnosis was substantial. Based on medium‐to‐large associations between interview‐assessed avoidant/restrictive food intake and questionnaire‐assessed food‐avoidance behaviors, food variety, and objective weight status, the module showed high convergent validity, especially for the child and parent version. Low associations of avoidant/restrictive food intake with weight and shape concern demonstrated divergent validity. Individuals with ARFID differed significantly from those with anorexia nervosa and individually matched controls in a range of clinical characteristics, indicating discriminant validity.
Discussion
This comprehensive validation supports the EDE ARFID module to be a valuable measure for the assessment and diagnosis of ARFID in 0–17‐year‐olds based on self‐ and parent‐report. Validation of the ARFID module against other interview‐based measures on ARFID and its evaluation in an adult sample are pending.
Public Significance
Based on good reliability and validity of the avoidant/restrictive food intake disorder (ARFID) module for the Eating Disorder Examination (EDE) in its child, adult, and parent version, the present study paves the way for the clinical and research use of the interview‐based EDE ARFID module for assessing ARFID across childhood and adolescence.