Construct validity and reliability of a French-Canadian translation of the eating in the absence of hunger questionnaire for children and adolescents
Construct validity and reliability of a French-Canadian translation of the eating in the absence of hunger questionnaire for children and adolescents

Construct validity and reliability of a French-Canadian translation of the eating in the absence of hunger questionnaire for children and adolescents

Appl Physiol Nutr Metab. 2024 May 31. doi: 10.1139/apnm-2023-0609. Online ahead of print.

ABSTRACT

Eating in the absence of hunger (EAH) has been associated with overweight and obesity during childhood. The gold standard to assess this behavior is a laboratory-based protocol, but a questionnaire to assess EAH more efficiently in children and adolescents has been developed and validated in English. We assessed construct validity (structural and convergent validity) and reliability (internal consistency and temporal stability) of a French translation of the EAH Questionnaire for Children and Adolescents among French-Canadian youths. We recruited participants in Montreal (Canada) aged 7-15 years old, who completed the questionnaire and provided anthropometric data. We asked participants to complete the questionnaire a second time ~4 weeks later. The questionnaire consists of 14 questions and 3 subscales that assess EAH due to negative affect, fatigue/boredom and external cues. We performed an exploratory factor analysis to test the factor structure and we calculated Cronbach alpha coefficients and intra-class correlations to assess internal consistency and temporal stability, respectively. We assessed associations between EAH and BMI z-score using Pearson correlations. We included 196 participants (50% girls; mean (SD) 11.9 (2.3) years old) for the first completion and 153 for the second completion. The exploratory factor analysis generated the same 3 subscales as the original questionnaire: negative affect (α=0.86; ICC=0.78), fatigue/boredom (α=0.75; ICC=0.70), and external cues (α=0.68; ICC=0.68). Participant’s BMI z-scores were positively associated with the average scores from the negative affect subscale (r=0.19; ρ=0.009). Our results suggest that this questionnaire had an adequate construct validity, internal consistency, and temporal stability.

PMID:38820601 | DOI:10.1139/apnm-2023-0609