Validation of a Child Version of the Three-Factor Eating Questionnaire – A Psychometric Tool for the Evaluation of Eating Behaviour

• Main Author: Yabsley, Jaime-Lee
• Other Authors: Chaput, Jean-Philippe
• Format: Others
• Language: en
• Published: Université d'Ottawa / University of Ottawa 2018
• Subjects: children; overweight; obesity; eating behaviours; psychometric properties; food and taste preferences
• Online Access: http://hdl.handle.net/10393/37977; http://dx.doi.org/10.20381/ruor-22235

Introduction: Currently, 1 in 7 children are classified as obese, which represents an obesity rate two times higher than that of the last 25 years. Part of the solution to address the positive energy balance underlying weight gain is to target the specific eating behaviours and factors that lead to food intake. One widely used tool to measure eating behaviour is the Three-Factor Eating Questionnaire (TFEQ). Objective: The primary objective of this study was to validate scores of the 21-item Child version of the Three-Factor Eating Questionnaire (TFEQ-R21 C), by examining validity evidence and reliability of TFEQ-R21 C responses in a sample of Canadian children and adolescents. The secondary aim was to examine the associations between the TFEQ-R21 C factors and body mass index (BMI) z-scores and food/taste preferences. Methods: The participants consisted of a sample of 158 children, 63 boys (mean age: 11.5 ± 1.6 years) and 95 girls (mean age: 11.9 ± 1.9 years), recruited from English schools in the Ottawa area. To assess eating behaviour, participants filled out the TFEQ-R21 C, the Power of Food Scale, and the Leeds Food Preference Questionnaire. Height and weight measurements were taken using a stadiometer and a digital scale. An exploratory factor analysis with oblique rotation and an item analysis were conducted to determine the factor structure and validity of the questionnaire. A median split on Cognitive Restraint (CR), Internal Uncontrolled Eating (UE 1), External Uncontrolled Eating (UE 2), and Emotional Eating (EE) was used to dichotomize factor-based scores into high and low categories for each factor, to allow for group comparisons. Bivariate correlations explored relationships between weight, BMI and BMI z-score, and food and taste preference, by sex and age group. To determine if BMI, BMI z-scores, and food/taste preferences were associated with factor scores of the TFEQ-R21 C, two-way ANOVAs were conducted. Results: The exploratory factor analysis replicated the Emotional Eating (EE) and Cognitive Restraint (CR) scales of the original TFEQ-R21, whereas the global factor of Uncontrolled Eating (UE) produced two subscales: Internal Uncontrolled Eating (UE 1) and External Uncontrolled Eating (UE 2). Item 17 did not load onto any of the factors and was subsequently removed. The four-factor model, with item 17 removed (FFEQ-R21 C: 20-item Child version Four-Factor Eating Questionnaire), accounted for 41.2% of the common variance in the data and showed good internal consistency (α= 0.81). The factors of UE 1 (r= 0.27, p<0.001), UE 2 (r= 0.36, p<0.0001), and CR (r= 0.20, p= 0.04) correlated significantly with EE. Younger children reported higher UE 1 scores [F(1,143)= 3.99, p=0.048, f2= 0.028] and CR scores [F(1,143)= 3.99, p= 0.001, f2= 0.089]. Boys who reported a high UE 1 scores had a significantly higher weight [F(1,58)= 6.44, p=0.014, f2= 0.117 ] and BMI z-scores [F(1,58)= 4.45, p=0.039, f2=0.083], compared to those who reported low UE 1 scores. Children with overweight or obesity [F(1,143)= 2.75, p<0.001. f2= 0.035] reported higher EE scores, compared to children of normal weight. Children with high UE 1 scores reported greater preference for high protein and fat foods, and high fat savoury (HFSA) and high fat sweet (HFSW) foods, compared to those with low UE 1 scores. Higher preference for high protein, fat, and carbohydrate foods, and HFSA, HFSW, and low fat savoury foods (LFSA) foods was found in children with high UE 2 scores, compared to those with low UR 2 scores. Children and adolescents with low CR scores reported greater preference for high protein, carbohydrate, and fat foods, compared to those who reported high CR scores. Discussion: This study showed adequate reliability and validity evidence of the TFEQ-R21 C scores, and that the questionnaire is best represented by a 20-item four-factor model in our sample. The FFEQ-R21 C was able to identify relevant eating behaviour traits associated with higher BMI z-scores and food preferences in both sexes and age groups, which were mainly in accordance with previous findings in children and adolescents. These results support the utility of the questionnaire for the assessment and identification of problematic eating behaviour and food preferences in the Canadian pediatric population. Younger children reported higher influence of the psychological constructs of eating behaviour (CR, UE 1, UE 2, and EE), compared to older children. This study provides preliminary evidence that FFEQ-R20 is a reliable and valid self-report tool to measure eating behaviour in children and adolescents to characterize those at higher risk for excess weight. However, further research is needed to examine the validity of the questionnaire in larger samples and in other geographical locations across Canada, as well as the inclusion of extraneous variables such as parental eating behaviour, socioeconomic status, and physical activity levels.