Examining the validity and consistency of the Adult Eating Behaviour Questionnaire-Español (AEBQ-Esp) and its relationship to BMI in a Mexican population

• Published in: Eating and weight disorders Vol. 27; no. 2; pp. 651 - 663
• Main Authors: Hunot-Alexander, Claudia, Arellano-Gómez, Laura Patricia, Smith, Andrea D., Kaufer-Horwitz, Martha, Vásquez-Garibay, Edgar M., Romero-Velarde, Enrique, Fildes, Alison, Croker, Helen, Llewellyn, Clare H., Beeken, Rebecca J.
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01-03-2022; Springer Nature B.V
• Subjects: Adult; Adults; Body Mass Index; Eating behavior; Feeding Behavior - psychology; Humans; Medicine; Medicine & Public Health; Mexico; Middle Aged; Original; Original Article; Psychiatry; Questionnaires; Reproducibility of Results; Surveys and Questionnaires; Young Adult; Mexico; Appetite; Obesity; Appetitive traits; Eating behaviour; Behavioural susceptibility theory; Adults; Weight
• ISSN: 1590-1262; 1124-4909; 1590-1262
• DOI: 10.1007/s40519-021-01201-9

Purpose
Appetitive traits in adults and their associations with weight can be measured using the Adult Eating Behaviour Questionnaire (AEBQ). The aim of this study was to confirm the factor structure of the Spanish AEBQ (AEBQ-Esp) in a Mexican sample and explore associations between the eight traits with body mass index (BMI).
Method
A sample of 1023 adults, mean age of 36.8 ± 12.8 years, was recruited from Guadalajara, Mexico. Researchers weighed and measured participants, and they completed the AEBQ-Esp either online or in paper format and reported sociodemographic data. To test two alternative factor structures (eight factors including Hunger; seven factors excluding Hunger), confirmatory factor analysis (CFA) was used. Internal reliability was assessed using Cronbach’s alpha; test–retest reliability was assessed using intra-class correlation coefficients. Multivariate linear regressions were used to test for associations between the AEBQ subscales and BMI, adjusted for age, sex, format of AEBQ responses, education, marital and employment status.
Results
A seven-factor structure was the best model fit using CFA, excluding the Hunger subscale but similar to the original AEBQ. Internal reliability was good for all subscales (Cronbach’s
α
 = 0.70–0.86), and the intra-class correlation coefficient (0.70–0.91) reflected good test–retest reliability. In the fully adjusted models, Satiety Responsiveness [
β
 = − 0.61; (− 1.01, − 0.21)] and Slowness in Eating [
β
 = − 0.70; (− 1.01, − 0.39)] were negatively associated with BMI, and Emotional Over-Eating [
β
 = 0.94; (0.62, 1.27)] was positively associated with BMI.
Conclusions
The AEBQ-Esp (excluding Hunger) appears to be a valid and reliable psychometric questionnaire for measuring appetitive traits in a Mexican Spanish-speaking population. Some traits appear to be associated with BMI in adulthood and warrant further exploration.
Level of evidence
Level III evidence obtained from well-designed cohort or case–control analytic studies. Although this was just an observational study, it was well designed and provided new evidence.