Knowledge of educators from day care centers about infant feeding

OBJECTIVE: To assess daycare centers educators' knowledge on feeding during the first year of life. METHODS: Cross-sectional study comprising 137 educators from eight daycare centers of São Paulo city (Brazil) selected from the universe of 36 daycare centers linked to Santo Amaro Coordination. Selection was based on Beghin's methodology. In order to assess educators' knowledge on feeding a structured, previously codified and tested questionnaire was built with true/false and multiple choices questions and applied. Chi-square test was used to test association between selected variables, being significant p<0.05. RESULTS: Educators' ages ranged from 19 to 66, with a median of 38 years; 41.3% had completed university education. Regarding errors for questions on breastfeeding, artificial lactation, complementary feeding and the sum of all questions about infant feeding, the median and standard deviation (sd) were respectively: 3 (sd=1.75); 2 (sd=0.87); 2 (sd= 0.96) and 7 (sd=2.65). CONCLUSIONS: Continuous action programs and improvement of intersectorial actions involving multidisciplinary teams and local daycare centers are required, since solely the knowledge incorporated into practice is not sufficient to promote good feeding and healthy eating habits for the proper growth and development of the assisted children.OBJETIVO: Avaliar o conhecimento das educadoras de creches sobre a alimentação nos primeiros anos de vida. MÉTODOS: Estudo transversal com 137 educadoras de oito creches do município de São Paulo, selecionadas no universo de 36 creches pertencentes à Coordenadoria de Santo Amaro, segundo critérios descritos a partir da metodologia proposta por Beghin. Para avaliar o conhecimento sobre alimentação das educadoras, foi construído e aplicado um questionário estruturado e pré-codificado com questões de múltipla escolha e verdadeiro e falso durante as reuniões mensais realizadas nas creches. No estudo das associações, utilizou-se o teste do qui-quadrado, selecionando variáveis com nível de significância p