Cross-cultural adaptation and validation of the Spanish version of the Exercise and Eating Disorders Questionnaire

Abstract

Purpose

This study aimed to adapt and assess the validity and reliability of the Spanish version of the Exercise in Eating Disorders Questionnaire (EED-Q) in order to diversify and offer a more comprehensive, effective, and standardized assessment of maladaptive exercise (ME) in ED.

Methods

The EED-Q is a self-reported questionnaire that assesses eating disorders (ED) patients' attitudes towards exercise. Based on the four-factor model of the original version, the EED-Q was adapted through forward and back-translation and inconsistencies were addressed through a committee of experts. Then, the EED-Q Spanish version (S-EED-Q) was administered to 172 patients with eating disorders (age = 15.28 ± 1.64 years). An exploratory factor analysis was computed to assess the construct validity. Inter-item correlations, item-factor correlations, McDonald's Omega, and Cronbach's Alpha were estimated to test the internal consistency (reliability). In addition, convergent validity was tested by relating EED-Q and the Eating Disorders Inventory 2 (EDI-2) scores, discriminant validity was assessed comparing EED-Q item-factor correlations, and divergent validity was conducted by analyzing EED-Q factor correlations.

Results

The S-EED-Q revealed significant generalized correlations among the scale items and showed good reliability scores (McDonald's Omega and Cronbach's alpha >0.7) except for Factor 2 (McDonald's Omega = 0.63 and Cronbach's alpha = 0.58). After eliminating items 8 and 15 due to their low factor loadings, the EFA revealed a robust empirical factor structure, adequate to the theoretical model, with good levels of total explained variance (65%). Convergent, discriminant and divergent validity showed good performance: results showed expected correlations between EED-Q and EDI-2, all items achieved higher item-factor correlations in their theoretical factor than in the others, and all factor-factor correlations were as expected.

Conclusion

This study is the first to adapt and validate the S-EED-Q. The psychometric properties of the S-EED-Q compared to the original version were supported with some limitations. Although the psychometric properties of the scale are adequate and the construct, convergent, discriminant and divergent validity are endorsed, some of the original items are questionable. Likewise, the items of the positive and healthy exercise factor require an in-depth revision.