Emily Lippincott and Dr. Renata Forste, Sociology

Breast-feeding is the healthiest way for a mother to ensure that her newborn child gets the best nutrition possible; in addition, breast milk provides maternal antibodies to the child that are highly effective in preventing disease (2). Hence the slogan “Breast is Best” was created to encourage mothers to breastfeed their children; unfortunately, there are still a larger number of women who have resisted this message and continue to choose other methods of feeding. For instance, white women are two times more likely to breast-feed than their black counterparts, and college educated women are three times more likely to breast-feed than women who have not graduated from high school (1). Given the health benefits, why did the “Breast is Best” slogan seem to reach a certain part of the population and miss others? Because breast-milk has been documented as the best thing for a mother to give her child, it is essential to understand the factors underlying the decision to breast-feed.

Building upon past research, I model the effects of maternal characteristics, including attitudinal measures, education level, and the effects of birth characteristics on the likelihood of breast-feeding. A unique contribution of this paper is the inclusion of not only maternal characteristics, but also birth characteristics in a single model. The primary aim of this paper, therefore, is to identify the effects of various factors on a woman’s choice to breast-feed. Factors such as race, education, religious affiliation, maternal attitude, geographic location, birth order of infant, type of delivery, and health of infant at birth are included in the model to see if there are any associations among these variables and a woman’s choice to breast-feed.

The analyses are based on data from the 1995 National Survey of Family Growth (NSFG). The National Survey of Family Growth collected data from women of child-bearing age in the United States.

The dependent variable is a dichotomous measure coded 1 if the respondent breast-fed and coded 0 if the respondent did not breast-feed. The independent variables include maternal characteristics such as education, religious affiliation, race, marital status, poverty level, and geographical location.

Because the dependent variable is dichotomous, the model is estimated using logistic regression. The results of this technique express the log odds of breast-feeding versus not breast-feeding as a linear function of a set of explanatory variables.

Of the respondents in the data set, 57 % breast-fed their children. Several factors significantly determine whether or not a woman will choose to breast-feed. Factors such as race, education level, and region of the United States are significant. The results show that whites are more likely to breastfeed than Blacks. Hispanics, however, when compared with Blacks, are just as likely to breast-feed.

Education level is a highly significant factor in a woman’s choice to breast-feed. The findings suggest that someone with a high school education or less is less likely to breast-feed than someone with a college education.

Finally, geographic region is highly significant. Women in the West were the most likely to breast-feed compared to other regions of the country.

Future attempts to increase breast-feeding in the United States should focus on race, region, and early education to target low breast-feeding groups and develop effective programs. Increasing the knowledge and resources of targeted women may help to increase the breast-feeding population and subsequently help more children receive the health benefits of a mother’s breast milk.

References

• Smith, David P. 1985. “Breastfeeding in the United States.” Social Biology 32:53-60.
• Winikoff, Beverly, Edward C. Baer. 1980. “The obstetrician’s opportunity: Translating “breast is best” from theory into practice.” American Journal of Obstetrics and Gynecology 138:105-117. 226