Department of Psychology, University at Albany, State University of New York, United States of America.
Department of Child and Adolescent Psychiatry and Behavioral Sciences, Children's Hospital of Philadelphia, Perelman School of Medicine, University of Pennsylvania, United States of America.
Eat Behav. 2020 Dec;39:101447. doi: 10.1016/j.eatbeh.2020.101447. Epub 2020 Oct 22.
The factor structure of the Eating Disorder Examination-Questionnaire (EDE-Q) has proven difficult to replicate, including in vegans, whose eating behaviors differ from omnivores in important ways. We sought to assess fit of data from vegans and omnivores with the most recently proposed brief three-factor model of the EDE-Q, which retains only seven of the original 28 EDE-Q items. We examined fit indices of the EDE-Q brief three-factor model in vegans (i.e., individuals refraining from all animal products, n = 318) and omnivores (i.e., individuals not restricting intake of animal products, n = 200) in single-group confirmatory factor analyses (CFA). Configural and metric invariance across the two groups was examined in multi-group CFA. Data from omnivores exhibited good model fit. Fit in vegans was slightly worse, but still adequate and superior to alternative models. Findings from multi-group CFA supported configural, but not metric invariance across the two groups. We document satisfactory fit of data from vegans and omnivores with the EDE-Q brief three-factor model, suggesting that it is better suited for quantifying disordered eating than the original four-factor, full three-factor, or alternative two-, full one-, and brief one-factor versions, including in individuals who abstain from animal products.
饮食障碍问卷（EDE-Q）的因子结构难以复制，包括素食者，他们的饮食行为与杂食者有很大的不同。我们试图用 EDE-Q 最近提出的简短三因素模型来评估素食者和杂食者的数据拟合情况，该模型仅保留了原始 28 个 EDE-Q 项目中的 7 个。我们在单组验证性因素分析（CFA）中检查了素食者（即完全避免所有动物产品的个体，n=318）和杂食者（即不限制动物产品摄入的个体，n=200）的 EDE-Q 简短三因素模型的拟合指数。在多组 CFA 中检查了两组之间的组态和度量不变性。杂食者的数据表现出良好的模型拟合。素食者的拟合稍差，但仍然足够，并且优于替代模型。多组 CFA 的结果支持两组之间的组态不变性，但不支持度量不变性。我们记录了素食者和杂食者的数据与 EDE-Q 简短三因素模型的良好拟合，表明它比原始的四因素、全三因素或替代的两因素、全一因素和简短一因素版本更适合量化饮食障碍，包括那些不食用动物产品的个体。
