The impact of a primary health care intervention on infant feeding practices: a cluster randomised controlled trial in Brazil.

BACKGROUND: Proper feeding practices in early life can enhance the full human potential development of children. We aimed to evaluate the impact of a primary health care intervention on infant feeding practices among children from low-income families. METHODS: A cluster randomised controlled trial was conducted in Porto Alegre, Brazil. Healthcare centres were randomised into intervention (n = 9) and control (n = 11) groups. In intervention sites, health workers were trained in accordance with the national guidelines. Infant feeding practices were assessed in children at 6 months (n = 617) and 12 months (n = 516) of age. Feeding practice quality was assessed using the Infant and Child Feeding Index (ICFI). Additionally, we evaluated the introduction of nonrecommended foods. RESULTS: At 6 months, the mean ICFI score was higher in the intervention group [MD = 0.22; 95% confidence interval (CI) = 0.24-1.11]. The prevalence of infants who met the recommendation for meat into the food-frequency score was higher in the intervention than the control group [relative risk (RR) = 1.63; 95% CI = 1.26-2.11]. At 12 months of age, the ICFI mean (MD = 0.23; 95% CI = 0.35-0.56) and the prevalence of children who met the recommendation for dietary diversity (RR = 1.11; 95% CI = 1.01-1.22) and meal frequency (RR = 4.68; 95% CI = 1.34-16.36) were higher in the intervention group, although only among children who had more than seven follow-up appointments during the first year of life. The children from intervention group had a significant delay for added sugar (MD = 0.51; 95% CI = 0.13-0.89), tea (mean = 0.47; 95% CI = 0.13-0.82), jelly (MD = 0.35; 95% CI = 0.11-0.58) and filled cookies (MD = 0.29; 95% CI = 0.06-0.52) compared to the control group. CONCLUSIONS: The health workers' training was effective with respect to improving infant feeding practices.