B Thurfjell1, B Edlund, H Arinell, B Hägglöf, I Engström. 1. Department of Neuroscience, Child and Adolescent Psychiatry, Uppsala University, University Hospital, Sweden. barbro.thurfjell@bupinst.uu.se
Abstract
OBJECTIVE: To evaluate the Swedish version of the Eating Disorder Inventory (EDI-C) for the internal consistency, factor structure and ability to discriminate between adolescent girls with and without a known eating disorder (ED). METHODS: Completely answered EDI-C questionnaires were available from 201 girls diagnosed with an ED and from 2073 schoolgirls. The girls were adolescents attending grade 7-12. RESULTS: Internal consistency was 0.70-0.91 for the ED group. Eleven factors with eigenvalue > 2.2 explained 56% of the variance and showed satisfying correspondence to the proposed constructs of EDI-C. A discriminant analysis classified 85.6% of the cases correctly. The three most important discriminating subscales were Drive for Thinness, Asceticism and Body Dissatisfaction. All subscales except Maturity Fears showed significantly higher mean values for the ED group. CONCLUSIONS: The results indicate that the psychometric properties of EDI-C are comparable to the properties of EDI-2. EDI-C should be preferred for adolescents since its statements suit this age group.
OBJECTIVE: To evaluate the Swedish version of the Eating Disorder Inventory (EDI-C) for the internal consistency, factor structure and ability to discriminate between adolescent girls with and without a known eating disorder (ED). METHODS: Completely answered EDI-C questionnaires were available from 201 girls diagnosed with an ED and from 2073 schoolgirls. The girls were adolescents attending grade 7-12. RESULTS: Internal consistency was 0.70-0.91 for the ED group. Eleven factors with eigenvalue > 2.2 explained 56% of the variance and showed satisfying correspondence to the proposed constructs of EDI-C. A discriminant analysis classified 85.6% of the cases correctly. The three most important discriminating subscales were Drive for Thinness, Asceticism and Body Dissatisfaction. All subscales except Maturity Fears showed significantly higher mean values for the ED group. CONCLUSIONS: The results indicate that the psychometric properties of EDI-C are comparable to the properties of EDI-2. EDI-C should be preferred for adolescents since its statements suit this age group.