Two Community-Based Initiatives Fail to Increase Breast-Feeding Levels Among Disadvantaged Women

Susan London

First published online:

| DOI: https://doi.org/10.1363/4112709

Offering community-based support for breast-feeding to relatively socioeconomically disadvantaged women during and after pregnancy does not improve rates of breast–feeding, according to a pair of cluster–randomized controlled trials. In a trial conducted in Scotland, the proportion of women who were breast-feeding their infants 6–8 weeks after birth was about the same (roughly three in 10) in areas that expanded their breast-feeding support group programs and those that did not.1 And in a trial conducted in the United Kingdom, the proportion of women who started breast-feeding their newborns before leaving the hospital was the same (seven in 10) whether women had attended antenatal clinics that offered peer support or clinics that did not.2

Breast-Feeding Group Support

In a trial conducted during 2002–2007 in 14 Scottish areas covered by a national surveillance program, half of the areas undertook an intervention to at least double their number of breast-feeding support groups and invite all pregnant and breast-feeding women to attend as part of routine primary care. The groups met weekly, were facilitated by female health professionals and focused on women's issues. The other half of the areas did not alter their number of breast-feeding groups. Health visitors and midwives collected data on breast-feeding, defined as an infant's receipt of any breast milk, and the women completed questionnaires about their satisfaction with breast-feeding and perceived social support.

Analyses were based on birth records of 9,635 eligible women in the intervention areas and 8,968 eligible women in the control areas. Women's median age was about 29. More than a quarter of the women overall were in the most deprived of five categories of socioeconomic status.

Intervention areas increased their total number of breast-feeding groups from 10 to 27, whereas control areas maintained their original 10 breast-feeding groups. Only about one in 10 of all women in intervention areas attended the groups, and they started attending a median of 36 days after giving birth. Women attending groups in intervention areas were older than women in those areas who began breast-feeding but did not attend groups; they had higher incomes than women attending general postnatal groups in control areas.

The proportion of women who were breast-feeding their infants at age 6–8 weeks, the trial's main outcome, did not differ significantly between the intervention areas and the control areas (26% vs. 30%); moreover, within the intervention areas, the proportion did not change after the intervention. The intervention and control areas were also statistically indistinguishable in terms of the proportions of women breast-feeding their infants immediately after birth (51% vs. 53%), at 5–7 days (42% vs. 45%) and at 8–9 months (21% vs. 20%).

Women in the intervention areas and their counterparts in the control areas gave similarly high ratings to their satisfaction with breast-feeding (median scores of 118 and 119, respectively, on a 150-point scale) and their perceived level of social support (median score of 4.25 in each group on a five-point scale). The intervention had an average annual cost of about $20,144 per area. The average cost per woman of attending one support group meeting was nearly the same as the average cost of one home visit by a health visitor ($54 vs. $47).

The investigators speculate that the lack of improvement in breast-feeding rates with expansion of group support may have been due to suboptimal group attendance, especially during pregnancy; the older age and higher incomes of attendees in the intervention areas (because the likelihood of breast-feeding increases both with age and with income); or the fairly high level of social support in the population generally. Given the ineffectiveness of this intervention, "resources may be better directed to the first two weeks after birth, when the highest proportion of women stop breast feeding," they conclude.

Breast-Feeding Peer Support

A 2007 U.K. trial conducted in 66 antenatal clinics in a socioeconomically deprived urban area of Birmingham assessed the impact of peer support for breast-feeding. Half of the clinics added a community-based peer support breast-feeding intervention to usual care, while the other half continued with usual care alone. In the intervention clinics, pregnant women were matched with peer support workers of similar race, ethnicity, and social and demographic background, who offered two sessions (at 24–28 weeks' and roughly 36 weeks' gestation), during which they gave advice about, information on and support for breast-feeding; the workers also provided postnatal follow-up to women who did start breast-feeding. At least one session was intended to take place in the woman's home. Initiation of breast-feeding, defined as a newborn infant's receipt of any breast milk before hospital discharge, was ascertained from hospital maternity records.

Analyses were based on 1,083 women attending the clinics offering peer support and 1,315 women attending the clinics providing only usual care. The large majority of women (80%) were 21–35 years old; 91% belonged to racial or ethnic minority groups. More than two-thirds of the women were in the most deprived of 10 socioeconomic categories.

In the peer support group, 80% of the women were in fact offered this support, and 74% accepted the offer, but only 42% of this group received the intended two sessions. First sessions lasted an average of 13 minutes, and most sessions overall occurred in the clinic.

The majority of women initiated breast-feeding, and the proportion did not differ between the peer support and usual care groups (69% vs. 68%). In a multivariate analysis, women had elevated odds of initiating breast-feeding if they were of a minority race or ethnicity (odds ratios, 1.6–6.5). On the other hand, women had reduced odds of initiating breast-feeding if they were multiparous (0.6) and if they had a cesarean delivery (0.7).

Discussing the lack of effectiveness of the peer support intervention, the investigators note that the sessions may not have been numerous or long enough and did not achieve the in-home contact intended. They add that improving on the level of support provided by usual antenatal care in the United Kingdom might require a more intensive home-based intervention, which would have considerable cost if provided to all women. "Peer support might be more effective if targeted at specific groups, such as those women not planning to breast feed … or those for whom routine advice on breast feeding is less accessible because of linguistic difficulties," they comment.

A Call for Integration

Taken together, the trials' results challenge recommendations for general community-based support strategies as a means to promote breast-feeding, according to the author of an accompanying editorial.3 Instead, women need detailed information about how to breast-feed and how to overcome early problems with lactation. Additionally, single interventions are unlikely to work in isolation because multiple factors influence breast-feeding practices, especially among disadvantaged populations, he contends. "Effective interventions will have a higher chance of producing results if embedded in a national or local plan that is tailored to specific needs," the editorialist concludes.

—S. London


1. Hoddinott P et al., Effectiveness of policy to provide breastfeeding groups (BIG) for pregnant and breastfeeding mothers in primary care: cluster randomised controlled trial, BMJ, 2009, 338:a3026, <http://www.bmj.com/cgi/reprint/338/jan30_1/a3026>, accessed Mar. 10, 2009.

2. MacArthur C et al., Antenatal peer support workers and initiation of breast feeding: cluster randomised controlled trial, BMJ, 2009, 338:b131, <http://www.bmj.com/cgi/reprint/338/jan30_1/b131>, accessed Mar. 10, 2009.

3. Cattaneo A, Promoting breast feeding in the community: breastfeeding groups and peer counselling must be integrated into wider programmes, BMJ, 2009, 338:a2657, <http://www.bmj.com/cgi/content/full/338/jan30_1/a2657>, accessed Mar. 10, 2009.