OBJECTIVE: based on the theory of social determination of the health-disease process, the objective of this study was to know the duration of overall breast feeding and the factors associated with its interruption, according to the social reproduction profiles of the families (ways of living and working). METHODS: this is a cross-sectional study carried out with a representative sample of 261 children under two years of age, living in Itupeva city, Sao Paulo, Brazil. From a theoretical-methodologicaloperational basis, three social homogeneous groups (GSH) were established, according to similar conditions of working and living. Overall breast feeding duration was calculated from survival tables. Variables with p<0.20 in bivariate Wilcoxon test were then introduced in multiple Cox Regression model in order to find associated aspects to the breast feeding duration. RESULTS: the breast feeding duration medians of the 3 GSH were 6.7 months, 7.1 months and 9.9 months, with no statistically significant difference (p=0.31). The multivariate analysis showed that the sequence of birth (p=0.018), pacifier use (p<0.001) and bottle-feeding (p<0.001) were associated to the interruption of breast feeding. CONCLUSIONS: there was no association between breast feeding duration and the social groups studied. However, the social determinants on breast feeding duration should be considered in actions that promote breast feeding. These actions also need to include information about harmful effects to the duration of breastfeeding when offering pacifiers and bottle-feedings.
Breast feeding; Social group; Social conditions
