A range of modifiable factors, including clinician support, may encourage mothers to continue breast-feeding their infants. A study in a Northern California health maintenance organization found that women whose health care providers had encouraged breast-feeding had reduced odds of discontinuing by the 12th week postpartum (odds ratio, 0.6).1 In contrast, women with depressive symptoms, and those encountering difficulties breast-feeding at school or on the job, had elevated odds of discontinuing (1.2-3.2).

The study involved 1,007 low-risk, breast-feeding mothers, along with their infants, enrolled during their postpartum hospitalization at a managed care hospital in Sacramento in 1996-1997. Enrollment criteria included a birth weight of 2,500-4,600 g and anticipated hospital discharge within 48 hours. Data were collected at three study interviews and through patient chart review. At an in-person interview during hospitalization, nurses asked the women about their confidence in their ability to breast-feed, whether they had taken a breast-feeding class and whether they thought breast-feeding is important. In a telephone interview at two weeks postpartum, a researcher administered a validated, 20-item depression instrument and asked about the father's support of breast-feeding and how much difficulty the woman had had breast-feeding during days 2-3. In another telephone interview, at 12 weeks, women reported whether they had returned to work or school, and whether they had problems breast-feeding in these environments; they also reported on whether a clinician had encouraged their breast-feeding. At all interviews, women reported whether they had stopped exclusive breast-feeding (i.e., fed the infant more than 12 oz. of formula daily) and, if so, when and why. Variables showing significance in preliminary analyses were included in multivariate logistic regression analysis to assess which ones were independently associated with breast-feeding discontinuation.

The study cohort comprised mostly white (62%), married (89%) women; the average age was 28. All but 6% of women had completed high school, and three in 10 had an annual household income exceeding $55,000. Almost all the women had received first-trimester prenatal care. Forty-one percent of participants had not given birth before; for 35%, this was their second child. The majority of participants perceived breast-feeding as very important (94%), were very confident in their ability to breast-feed (68%) and had had no serious problems with breast-feeding the infant at days 2-3 (79%). Slightly more than one-third had taken prenatal breast-feeding courses (36%). Most reported receiving "a great deal" of support for breast-feeding from the baby's father (83%), but recalled receiving no encouragement from clinicians (77%). By week 12, nearly half the mothers had returned to school or work (47%); exactly half of those who had returned reported having problems breast-feeding in these environments.

In the first week, 105 women discontinued breast-feeding. For half these women, the reason cited was lack of milk production (the infant was still hungry after being breast-fed) or infant problems with latching on or sucking. Forty-six percent of the 74 women who discontinued in weeks 2-3 cited insufficient milk production, breast pain or soreness, or return to work or school. Of the 112 women who discontinued in weeks 4-6, and the 53 in weeks 7-9, 62-67% cited insufficient milk production or return to work or school. In the three final study weeks, 58% of 19 women who discontinued cited return to work or school.

In multivariate analysis, women had significantly increased odds of stopping breast-feeding at or before the second week if they had lacked confidence in their ability to breast-feed (odds ratio, 2.8) or had had problems breast-feeding early on (1.5). The lower a woman's educational level, the higher her odds of discontinuing (1.5). Asian women were more likely than white women to discontinue (2.6). Odds of stopping within two weeks were decreased if the woman had considered breast-feeding very important instead of not important (0.3).

The younger a woman was, the less education she had and the lower her household income, the greater her odds of stopping between the two-week and 12-week interviews (odds ratios, 1.2 each). The odds also were elevated for women who were Asian (2.3), had returned to work or school (2.4), had problems breast-feeding at work or school (3.2), or had scored in the lowest quartile on the depression screening instrument (1.2). Participants who had perceived breast-feeding as very important or received encouragement from a provider were less likely than other women to stop between weeks two and 12 (0.2 and 0.6, respectively).

The authors point out that their findings may not be generalizable to all U.S. women, because their cohort had higher levels of education, income and insurance coverage than the general population. In addition, because data on clinician encouragement of breast-feeding relied on retrospective self-reports, recall bias may explain the observed association between encouragement and discontinuation.

Nonetheless, the authors believe that their study "adds to the growing body of literature suggesting that support and encouragement to breastfeed from health care providers is associated with a higher likelihood of breastfeeding continuation." Finally, they note that "attention should also be paid to maternal mental health status...as a potentially modifiable factor in promoting breastfeeding continuation."

—C. Coren

1. Taveras EM et al., Clinician support and psychosocial risk factors associated with breastfeeding discontinuation, Pediatrics, 2003, 112(1):108-115.