Psychometric Properties of the Reflective Function Questionnaire for Youth Five-Item Version in Adolescents with Restrictive Eating Disorders

Abstract

Mentalizing difficulties are an established correlate of eating disorders, but there is a need to establish the validity and reliability of easy-to-administer instruments to measure the construct in adolescents. This study evaluated the psychometric properties of a five-item version of the Reflective Function Questionnaire for Youth (RFQY-5) in a sample of adolescents with restrictive eating disorders recruited for a prospective observational study. Adolescents completed the 46-item version of the RFQ-Y at the start of treatment (T1) (n = 171) and nine months later (n = 51). The factor structure, internal reliability and convergent validity of the RFQY-5 were assessed at T1. Sensitivity to change was assessed in a subsample of n = 51 who completed the RFQ-Y nine months after the start of treatment (T2). At T1, the RFQY-5 demonstrated inadequate internal reliability (McDonald’s Omega = 0.657) and poor model fit for the unidimensional structure (Χ2(5) = 29.98; p < 0.001; CFI = 0.801; TLI = 0.602; RMSEA = 0.171; SRMR = 0.070). Convergent validity was partially supported. The RFQY-5 was not sensitive to change. The RFQY-5 was found to have inadequate psychometric properties in adolescents with restrictive eating disorders. Psychometric studies are needed to further develop valid and reliable measures of mentalizing for this population.