AUTHOR=Mais Laís Amaral , Warkentin Sarah , Latorre Maria do Rosário Dias de Oliveira , Carnell Susan , Taddei José Augusto de Aguiar Carrazedo 

TITLE=Validation of the Comprehensive Feeding Practices Questionnaire among Brazilian Families of School-Aged Children

JOURNAL=Frontiers in Nutrition

VOLUME=Volume 2 - 2015

YEAR=2015

URL=https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2015.00035

DOI=10.3389/fnut.2015.00035

ISSN=2296-861X

ABSTRACT=Introduction Children’s eating behaviors are influenced by parents, who are the first nutritional educators. The Comprehensive Feeding Practices Questionnaire (CFPQ) was developed to measure feeding practices among parents, but has not yet been validated in Brazil, where child obesity rates are steeply increasing. The aim of the study was to test the validity of the CFPQ among Brazilian parents of school-aged children and propose a new version of the instrument. Methods Transcultural adaptation included translation into Portuguese, back-translation, content validity, testing for semantic equivalence, and piloting. Questionnaire data were obtained for 659 parents of 5-to-9-year-olds. Confirmatory and exploratory factor analyses and psychometric analyses (tests for internal consistency, factor correlations, item-discriminant and convergent validity, and test-retest reliability) were conducted. Results Confirmatory factor analysis demonstrated a poor fit of the data to the original 12-factor model. Exploratory factor analysis generated a 6-factor model composed of 42 items: Healthy Eating Guidance, Monitoring, Restriction for Weight Control, Restriction for Health, Emotion Regulation/Food as Reward, and Pressure. This factor solution was supported by internal consistency tests (α=0.71-0.91) and factor correlations (rho=-0.16 to 0.32). Item-discriminant and convergent validity tests showed that parents who used coercive practices had more overweight children and were more concerned about their child’s weight (rho=0.09 to 0.40). Test-retest reliability was acceptable (ICC=0.45 to 0.77). Conclusions Since parental practices are highly culturally- and age group-sensitive, it is essential to conduct careful evaluations of questionnaires when introduced into specific age groups within new cultural settings. This modified 6-factor model of the CFPQ is valid to measure parental feeding behaviors of school-aged children in urban Brazilian settings.