Eating disorder examination-questionnaire – 7: Construct validity in a sample of portuguese overweight women

Abstract

Introduction

Although the Eating Disorder Examination Questionnaire (EDEQ; Fairburn et al. 2008) is the most used instrument worldwide for the assessment of eating disorders symptoms, its factorial structure considerably varies, which limits its construct validity. Using exploratory factor analisys in data from a sample of overweight women, our group found a three-factors structure of the EDEQ Portuguese version (Peixoto et al. 2013), Although it was in accordance with other psychometric studies (eg. Peterson et al 2007), it was different from the original matrix. Further investigation regarding its factor structure has been conducted, with studies supporting a modified seven-item-three-factors structure (dietary restraint, shape/weight overvaluation, body dissatisfaction) with improved psychometric properties (Grilo et al. 2013, 2015), including with Portuguese samples (Machado et al. 2018; Santos et al. 2019).

Objectives

To analyze if the EDEQ version composed of seven items and three factors is replicated in a Portuguese sample of overweight women.

Methods

The EDEQ was administered to an outpatient sample of 276 women (Mean age= 43.85±11.89 years; Mean BMI=32.82±5.43 Kg/height²) attending a weight loss treatment consultation in a public hospital.

Results

Confirmatory factor analysis (CFA) revealed an adequate fit of the EDEQ-7 second order model with three dimensions (χ²/df=1.5497; RMSEA=.0452, CFI=.9955, TLI=.9914, GFI=.xxx; p<.001). The EDEQ7 Cronbach’s alphas for the total and its dimensions were α<.70.

Conclusions

Given its good psychometric properties, the overlap of the measurement model with those found with different samples and the reduced number of items, the EDEQ7 will be very useful both in research and clinical settings with/for overweight women.