Re-evaluation and revision of the Eating Habits Questionnaire.

Introduction: The Eating Habits Questionnaire (EHQ) is a key tool in evaluating orthorexia nervosa, an obsession with healthy eating. However, the evaluation process of EHQ has witnessed considerable variation, with one item notably excluded from the last phase of its development. This study undertakes a thorough re-evaluation of the English version of the EHQ, focusing on its original 35 items, within two diverse populations (fashion models and university students) where English serves predominantly as a second language.

Methods: Through an online survey involving 163 female models and 243 non-models (mean age=26.0 years (SD=4.7) and mean age=25.1 years (SD=5.0), respectively) with various ethnic background (proportion from non-English speaking EU countries: 28.5% and 90.4%, respectively). Participants completed the EHQ along with the Eating Disorder Inventory (EDI) and SCOFF questionnaire.

Results: Using confirmatory factor analysis, we analysed the factorial validity of EHQ subscales, eliminating items that did not align with the factor structure. After eliminating 17 items from the original 35-item questionnaire, the fit of the model for EHQ-18 was acceptable. Cronbach's alpha values indicated acceptable reliability. The EHQ-18 problem sub - scale showed significant positive correlations with all EDI subscales, while all EHQ-18 subscales demonstrated significant positive correlations with the EDI Drive for Thinness subscale. Comparison of the groups based on the SCOFF threshold revealed positive and significant differences across all subscales.

Conclusion: We examined the impact of replacing an item in the EHQ-21 previously during its development process, and found that this potentially influenced the resulting factor structure. We recommend a new version of EHQ, the 18 item EHQ-18, supported by the analysis of the factorial and convergent validity, as well as the reliability. Furthermore, the findings suggest a potential discriminant validity of EHQ-18 in a diverse population, mostly speaking English as a second language.