Avaliação psicométrica da versão em português brasileiro do Eating Disorder Examination Questionnaire (EDE-Q) em homens adultos cisgêneros gays e bissexuais

Abstract

Researches assert that sexual and gender minority men present high risk of eating disorders compared to their heterosexual counterparts. However, one of the most applied measures of disordered-eating attitudes and behaviors, the Eating Disorder Examination Questionnaire (EDE-Q), hasn‘t had its psychometric properties evaluated in Brazilian cisgender gay and bisexual adult men. Thus, the present study aimed to evaluate the psychometric properties of the Brazilian Portuguese version of the EDE-Q in a sample of Brazilian cisgender gay and bisexual adult men. This is a mixed and sequential methodological research, with a psychometric focus, in which were included cisgender gay and bisexual adult men from Brazil with ages between 18 and 50 years, of any race/ethnicity. After the approval of the Internal Review Board of the Federal University of Juiz de Fora (approval number: 4.690.224), the participants, through Google Forms, answered to a sociodemographic questionnaire as well as measures of disordered-eating attitudes and behaviors, body-ideal internalization, muscle dysmorphia symptoms, drive for muscularity, selfobjectification, and body appreciation. We evaluated the factorial and convergent validities, as well as the internal consistency and the test-retest reliability of the EDEQ. One thousands and four hundred and nine Brazilian gay and bisexual adult men took part, with the average age of 26.96 years (SD = 5.29 years) and an average BMI of 25.99 kg/m² (SD = 5.26). The factor structure of the Brazilian EDE-Q was evaluated through a two-step approach, with samples divided as follows: exploratory factor analysis (EFA; n = 704) and confirmatory factor analysis (CFA; n = 705). The results of an EFA and CFA revealed a unifactorial structure with 22 items, supporting the factorial validity of the EDE-Q. The scores of the EDE-Q showed strong and positive correlations with the self-objectification (rho = .56; p < .001) and the bodyideal internalization (rho = .50; p < .001). Moreover, the scores of the EDE-Q presented moderated and positive correlation with muscle dysmorphia symptoms (rho = .45; p < .001); as well as a weak and positive correlation with the drive for muscularity (rho = .22; p < .001). At last, it was observed a strong and negative correlation with the body appreciation (rho = - .62; p < .001). Such correlations support the convergent validity of the EDE-Q. Regarding reliability, it was observed an adequate internal consistency (McDonald‘ ω = .92, IC 95% = [.91, .93]) and twoweek test-retest stability (intraclass correlation coefficient = .86, IC 95% = [.82, .88], p < .001). It is concluded that the EDE-Q presented good psychometric evidence of validity and reliability for Brazilian cisgender gay and bisexual adult men. The present study contributes to the clinical and epidemiological practice, providing an essential instrument for the evaluation of disordered-eating attitudes and behaviors in a highly stigmatized population.