Pacifiers are widely used in many developing countries. Although their use is not recommended for breastfeeding infants, there have been no published reports on the association between pacifier use and risk of early weaning. In a study of 354 infants in Brazil, mothers were asked about pacifier use at age 1 month, duration of breastfeeding, and introduction of other foods. Among 249 children still breastfed at 1 month, the risk that a child would be weaned at any age between 1 and 24 months was higher in pacifier users than in non-users (hazard ratio 3.0, 95% CI 2.0-4.6). The association remained even after adjustment for the child's age, sex, birthweight, socioeconomic status, and age at introduction of bottle-feeding.
PIP: Epidemiologists analyze data on 354 children less than 2 years old living in slum areas of Pelotas, Brazil, to determine the relationship between pacifier use and breast feeding duration. By the age of 1 month, 67% of the infants use a pacifier, and at 3 months and older, 80% do. Breast feeding lasts for a median of 3.5 months. By age 1 month, 94% of the children have been given water and/or herbal teas and 27% are fed formula. The 166 pacifier users at 1 month have a greater likelihood of being weaned between 1 and 24 months than are the 83 nonusers (hazard ratio = 3). Further, among the 186 children who are at least 6 months old, nonusers of pacifiers are significantly less likely to be weaned at 6 months than are users of pacifiers (24% vs. 65%; p .001). Among children who use pacifiers, part-time users of pacifiers are significantly less likely to be weaned at 6 months than are full-time users (59% vs. 72%; p .001), suggesting a dose-response relationship. The association between pacifier use and weaning continues even when the researchers adjust for bottle feeding. The only other factor linked with weaning is age of introduction to bottle feeding (hazard ratio for bottle feeding 1 month = 1.6). Infants who use pacifiers at 1 month are 1.3 times more likely to have been treated for diarrhea than nonusers. The mean weight for length z score of pacifier users is much more negative than nonusers (-0.45 vs. -.21; p = .055). Possible explanations for these results are the pacifiers are causally associated with weaning, used as a means to begin weaning, or mark the beginning of breast feeding, pacifier use may decrease nipple stimulation and thereby reduce milk production. Further research is needed to understand the role of pacifiers in weaning, and breast feeding promotion efforts must incorporate the information obtained from such studies.