Determinants of the introduction of early complementary feeding before and after the third month of life: a multinomial analysis.

The introduction of early complementary feeding (ECF) is determined by different factors depending on when it occurs. The objective of this study was to analyze factors associated with the introduction of ECF in two different moments of the infant's life: from zero to three and from four to five months of age. A cohort with 3,306 dyads studied in the BRISA survey in São Luis/MA in 2010 was used. Questionnaires were applied at birth and at follow-up when the infants were 15 to 36 months of age of women with more than 20 weeks of gestational age, residing in this municipality. A multivariate model of multinomial logistic regression was used to verify associations between independent variables and ECF at 0 to 3 months and at 4 to 5 months of age. A hierarchical analysis model was used to select variables for confounding adjustment. Variables with a P-value <0.05 were considered significant. For ECF introduced between 0-3 months, the variables "use of pacifier", "maternal paid activity", "smoking", and "postpartum pregnancy" were identified as risk factors. The variables "use of pacifier" and "maternal paid activity" remained associated as a risk for ECF introduced from 4-5 months. The variable 'mother without partner' (RR=1.26 and P=0.04) represented a risk factor for ECF only for the 4-5 months period. Although each period presented specific risk factors, the use of pacifier and maternal professional activity were associated in the two periods studied, indicating their importance for the introduction of ECF.