BackgroundRecent national surveys in Brazil have demonstrated a decrease in the consumption of traditional food and a parallel increase in the consumption of ultra-processed food, which has contributed to a rise in obesity prevalence in all age groups. Environmental factors, especially familial factors, have a strong influence on the food intake of preschool children, and this has led to the development of psychometric scales to measure parents’ feeding practices. The aim of this study was to test the validity of a translated and adapted Comprehensive Feeding Practices Questionnaire in a sample of Brazilian preschool-aged children enrolled in private schools.MethodsA transcultural adaptation process was performed in order to develop a modified questionnaire (43 items). After piloting, the questionnaire was sent to parents, along with additional questions about family characteristics. Test-retest reliability was assessed in one of the schools. Factor analysis with oblique rotation was performed. Internal reliability was tested using Cronbach’s alpha and correlations between factors, discriminant validity using marker variables of child’s food intake, and convergent validity via correlations with parental perceptions of perceived responsibility for feeding and concern about the child’s weight were also performed.ResultsThe final sample consisted of 402 preschool children. Factor analysis resulted in a final questionnaire of 43 items distributed over 6 factors. Cronbach alpha values were adequate (0.74 to 0.88), between-factor correlations were low, and discriminant validity and convergent validity were acceptable.ConclusionsThe modified CFPQ demonstrated significant internal reliability in this urban Brazilian sample. Scale validation within different cultures is essential for a more comprehensive understanding of parental feeding practices for preschoolers.

We perform importance sampling for a randomized matrix multiplication algorithm by Drineas, Kannan, and Mahoney and derive probabilities that minimize the expected value (with regard to the distributions of the matrix elements) of the variance. We compare these optimized probabilities with uniform probabilities and derive conditions under which the actual variance of the optimized probabilities is lower. Numerical experiments with query matching in information retrieval applications illustrate that the optimized probabilities produce more accurate matchings than the uniform probabilities and that they can also be computed efficiently.

No abstract

IntroductionChildren’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.MethodsTranscultural 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.ResultsConfirmatory 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 (ρ = −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 (ρ = 0.09–0.40). Test–retest reliability was acceptable (intraclass correlation coefficient = 0.45–0.77).ConclusionSince 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 six-factor model of the CFPQ is valid to measure parental feeding behaviors of school-aged children in urban Brazilian settings.

In this study, the risk variables identified for the use of the maternal restrictive feeding practices to control the child's weight can be used to provide education and guidance interventions in health and education networks directed to groups with similar characteristics to those of the studied population.