Validation of the youth‐nine item avoidant/restrictive food intake disorder screen

• Published in: European eating disorders review 2024-01, Vol.32 (1), p.20-31
• Main Authors: Billman Miller, Marley G., Zickgraf, Hana F., Murray, Helen Burton, Essayli, Jamal H., Lane‐Loney, Susan E.
• Format: Article
• Language: English
• Subjects: Adolescent; Avoidant Restrictive Food Intake Disorder; avoidant/restrictive food intake disorder (ARFID); Binge-Eating Disorder; Child; children and adolescents; Eating; Eating disorders; Feeding and Eating Disorders - diagnosis; Female; Humans; Male; Retrospective Studies; Validity
• ISSN: 1072-4133; 1099-0968
• DOI: 10.1002/erv.3017

Objective This study assessed the factorial, divergent, and criterion‐related validity of the Youth‐Nine Item Avoidant/Restrictive Food Intake Disorder (ARFID) Screen (Y‐NIAS) in a paediatric clinical sample at initial evaluation for an eating disorder (ED). Method Participants included 310 patients (82.9% female, 77.4% White, Age M = 14.65) from a tertiary ED clinic. Confirmatory factor analysis (CFA) evaluated the three‐factor of the Y‐NIAS. One‐way analysis of variance compared Y‐NIAS scores across diagnoses. A receiver operating curve analysis assessed the ability of each subscale to identify ARFID presentations from the full sample. Two logistic regressions assessed the criterion‐related validity of the obtained Y‐NIAS cut‐scores. Results CFA supported the original three‐factor structure of the Y‐NIAS. Clinically‐elevated scores were observed in all diagnostic groups except for binge‐eating disorder. Subscales were unable to discriminate ARFID cases from other ED diagnoses. Cut scores were identified for picky eating subscale (10) and Fear subscale (9), but not for Appetite subscale. In combination with the ED Examination Questionnaire (EDE‐Q), classification accuracy was moderate for ARFID (62.7%) and other EDs (89.4%). Discussion The Y‐NIAS demonstrated excellent factorial validity and internal consistency. Findings were mixed regarding the utility of the Y‐NIAS for identifying clinically‐significant ARFID presentations from other ED diagnoses. Key points The Y‐NIAS demonstrated acceptable factorial validity, internal consistency, and discriminant validity with a measure of weight/shape‐driven disordered eating. Mean scores on the Y‐NIAS subscales did not significantly differ between those diagnosed with Avoidant/Restrictive Food Intake Disorder (ARFID) and those diagnosed with other eating disorders. Cut scores for the picky eating (PE) and Fear subscales with high sensitivity and acceptable specificity were identified, but positive predictive value (PPV) was low. Despite limited ability to discriminate ARFID from other eating disorders, the Y‐NIAS was able to identify clinically‐elevated scores in all diagnostic groups with restrictive eating disorders, indicating that it may have the potential to serve as a transdiagnostic ED screener.