The proportion of U.S. infants who are ever breast-fed is close to the target set in Healthy People 2010; however, the proportions still being breast-fed at six and 12 months of age, and being exclusively breast-fed when they are six months old, are far below the national goals.1 College-educated mothers come closer than any other subgroup to meeting the Healthy People goals, but even their level of exclusive breast-feeding at six months falls short of the target. Analyses of data from the 2003 and 2004 National Immunization Surveys reveal that the most consistent predictors of whether levels of breast-feeding meet the national goals are participation in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), residence in western states and living in an area with a high prevalence of conditions that are considered unhealthy for newborns.

In all, the two surveys yield a sample of more than 61,000 children aged 19–35 months; with weighting, the data are representative of all U.S. youngsters in this age-group. The analysts used the full sample to calculate descriptive characteristics, identify subgroups that met the Healthy People goals and assess the likelihood that infants had ever been breast-fed. They used the subsample of those who had ever been breast-fed to examine characteristics associated with having been breast-fed at six and 12 months.

Overall, 71% of children were ever breast-fed—close to the 75% targeted in Healthy People. However, only 36% were breast-fed when they were six months old, whereas the national goal is 50%. Some 16% were exclusively breast-fed at six months, and 18% were breast-fed at one year of age; the target for each of these is 25%. Examination of 36 maternal, child and environmental characteristics showed that most subgroups were within 10 percentage points of the goal for any breast-feeding and for breast-feeding at 12 months. Only eight, however, were that close to the goal for breast-feeding at six months; half were within 10 points of the target for exclusive breast-feeding at that age.

Results of logistic regression analysis indicated that the odds that a child had ever been breast-fed were significantly reduced if the mother was not a college graduate (odds ratios, 0.4–0.6), was unmarried (0.6–0.7) or was enrolled in WIC (0.7); they were higher among those living in the West than among Northeasterners (2.1). Child characteristics also were associated with the likelihood of breast-feeding. Compared with white children, blacks had a lower likelihood of having been breast-fed (0.7), and Hispanics had a higher likelihood (2.2). Firstborn children were more likely than others ever to have been breast-fed (1.1). A measure indicating how unhealthy a newborn’s environment is (taking into account state and metropolitan levels of teenage and nonmarital childbearing, low birth weight, and late or no prenatal care, among other factors) was inversely related to a child’s likelihood of having been breast-fed (0.98).

Patterns of associations were largely similar in analyses examining the likelihood that infants who were ever breast-fed were still breast-fed at six months of age. The odds of this outcome were reduced among those whose mothers were not college graduates (odds ratios, 0.6–0.7), were not married (0.7–0.8) or were WIC participants (0.8), and they fell with increasingly unhealthy environment (0.98); the odds were elevated for residents of western states (1.3) and for Hispanics (1.2). Additionally, the likelihood that breast-feeding had continued at least to age six months was reduced if the infant was a firstborn (0.8), if the mother was younger than 30 (0.5–0.8) or if the baby was a boy (0.9).

Exclusive breast-feeding at six months had fewer predictors than any breast-feeding at that age. Infants whose mothers had an incomplete college education, were formerly married, were in their 20s or participated in WIC had reduced odds of receiving only breast milk at six months (odds ratios, 0.8–0.9), as did males and firstborns (0.9 for each). Again, the more unhealthy an environment was for newborns, the lower the likelihood of breast-feeding (0.98). The odds of this outcome were elevated for non-Hispanic children who belonged to a racial group other than white or black, residents of western states and, notably, children whose mothers reported a household income that was below the poverty level (1.1–1.2).

A final model, examining the likelihood that infants who had ever been breast-fed were still being breast-fed when they were a year old yielded results much like those from the model studying any breast-feeding. Infants born to women who lacked a college degree, were not married, were younger than 30, participated in WIC or lived in an area with an unhealthy environment had reduced odds of being breast-fed, as did firstborns, blacks and males. Residence in the West was again associated with elevated odds of breast-feeding.

While acknowledging a number of study limitations, the analysts also note that their findings are consistent with those of other research. They therefore conclude that “many women start [breast-feeding], but too few continue,” and they suggest that deficiencies in maternal and environmental resources are at least partly to blame. “If mothers who lack resources … do not receive support to breastfeed,” they write, “the health and well-being gap between children born to women with resources and those born to disadvantaged mothers will only widen.”


1. Forste R and Hoffmann JP, Are US mothers meeting the Healthy People 2010 breastfeeding targets for initiation, duration, and exclusivity? The 2003 and 2004 National Immunization Surveys, Journal of Human Lactation, 2008, 24(3):278–288.