Construtos dimensionais, categóricos e híbridos dos transtornos alimentares: uma avaliação empírica da estrutura latente do Eating Disorder Examination – Questionnaire (EDE-Q)

• Ano de defesa: 2016
• Autor(a) principal: Paraventi, Felipe [UNIFESP]
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Paulo (UNIFESP)
• Palavras-chave em Português: Eating disorder; Psychometry; Hybrid model; EDE-q; Psicometria; Model híbrido; Transtornos alimentares; Estrutura
• Link de acesso: https://sucupira.capes.gov.br/sucupira/public/consultas/coleta/trabalhoConclusao/viewTrabalhoConclusao.jsf?popup=true&id_trabalho=3638997; http://repositorio.unifesp.br/handle/11600/46680

Background – The EDE-Q is a widely used instrument to assess eating disorders symptoms and behaviors, but evidence regarding its latent structure is not consistent. The present study aims to explore EDE-Q’s inner configuration through a mixture categorical-dimensional approach. Method – Data were taken from the second cohort of the The Women’s Eating, Health and Lifestyle longitudinal study (n=706). A dimensional approach included traditional Confirmatory Factor Analysis (CFA) and the Bayesian Confirmatory Factor Analysis (BCFA) on eight different models as previously reported in literature. Latent Class Analysis was performed to explore categorical classes underlying the EDE-Q’s items. Investigators also assessed the discriminant validity of the best-fitting solution for six distal outcomes. Results – None of the dimensional models were confirmed after CFA or even the more flexible procedure (i.e. BCFA). LCA derived three distinct classes: healthy people, subclinical group and ill population. All classes returned good discriminant validity regarding general psychopathological distress, body-mass index and regular binge eating. Among the EDE-Q’s twenty-two items, “Weight Dissatisfaction”, “Shape Dissatisfaction” and “Feeling of fatness” had the highest endorsement probabilities. Conclusions – Our results support a categorical phenomenon underlying EDE-Q’s structure; consequently, use of EDE-Q’s global score or even its subscales may not be valid. This work suggests that weight/shape dissatisfaction may have greater influence than eating behaviors to the categorization of eating disorders.