Abstract 1642

Background. One of the goals of Healthy People 2000 is to reach a longer duration of exclusive breastfeeding. In the development of initiatives for this goal it is important to assess those factors associated with the length of breastfeeding that could be modified.

Aim. To determine the prevalence of exclusive breastfeeding in a middle-class population and to identify the factors influencing its duration in the first 6 months post partum.

Methods. A longitudinal prospective study of 599 mothers and their healthy full term infants was performed from 1 July 1996 to 31 January 1997. All mothers roomed-in with their infants and 96% of the newborns were put on the breast in the first 6 hours. Twenty six medical, demographic and psychosocial variables were recorded before discharge from the Hospital. Mothers were asked by telephone at 2 weeks, 4 weeks and then monthly for 6 months how they were feeding their infants. ANOVA was used in the comparison between each one of the variables and the duration of exclusive breastfeeding. A regression model was used for variables significantly associated by univariate analysis and the prognostic value of each variable to predict the duration of breastfeeding was tested by failure-time models.

Results. Complete feeding data were available on 539 infants (90%). At birth 97% were exclusively breastfed and 3% were bottled fed. At 4 weeks, 4 months and 6 months post partum the exclusive breastfeeding rates were 83%, 38% and 19% respectively. Median value of the duration of full breastfeeding was 4 months (lower quartile 2 and upper quartile 5). A longer duration of exclusive breastfeeding was found to be significantly associated with higher maternal education (p<0.003), previous breastfeeding (p<0.001), no nipple problems (p<0.001), good mother-infant bonding (p<0.05), maternal attitude toward breastfeeding (p<0.001) and social support (mainly from husbands; p<0.001). These associations persisted after controlling for confounders.

Conclusions. Our findings suggest that certain modifiable maternal factors associated with a longer duration of breastfeeding may be identified in the perinatal period and may contribute to the development of more effective breastfeeding policies.