Book Description

OBJECTIVES: To summarize the effectiveness of community, workplace, and health care system-based programs and policies aimed at supporting and promoting breastfeeding and determine the association between breastfeeding and maternal health. DATA SOURCES: We searched PubMed(r)/MEDLINE(r), the Cochrane Library, and CINAHL(r) from January 1, 1980, to October 12, 2017, for studies relevant to the effectiveness of health care system-based, workplace, and community breastfeeding programs and policies. For evidence on breastfeeding and maternal health, we updated the 2007 Agency for Healthcare Research and Quality report on this topic and searched the same databases from November 1, 2005, to October 12, 2017. For studies of breastfeeding programs and policies, trials, systematic reviews, and observational studies with a control group were eligible; we excluded primary care-based programs delivered as part of routine care. For studies related to breastfeeding and maternal health, we included systematic reviews, case-control studies, and cohort studies. REVIEW METHODS: Pairs of reviewers independently selected, extracted data from, and rated the risk of bias of relevant studies; they graded the strength of evidence (SOE) using established criteria. We synthesized all evidence qualitatively. RESULTS: We included 128 studies (137 publications) and 10 systematic reviews. Of these, 40 individual studies were relevant to the effectiveness of breastfeeding programs or policies, and the remainder were relevant to one or more maternal health outcomes. Based on evidence from one large randomized controlled trial (RCT) (Promotion of Breastfeeding Intervention Trial [PROBIT], N=17,046) enrolling mothers who intended to breastfeed and nine cohort studies (1,227,182 women), we graded the SOE for the Baby-Friendly Hospital Initiative (BFHI) as moderate for improving rates of breastfeeding duration. Evidence from eight cohort studies of BFHI (135,983 women) also demonstrates improved rates of breastfeeding initiation (low SOE). Low SOE (k=4 studies; 1,532 women) supports the conclusion that health care education or training of staff alone (without additional breastfeeding support services) does not improve breastfeeding initiation rates. Women, Infants and Children (WIC, a Federal supplemental nutrition program) interventions that focus on peer support are effective in improving rates of breastfeeding initiation and duration (low SOE). We found limited evidence for other (community-based) interventions and no comparative studies on workplace or school-based interventions or harms associated with interventions. For maternal health outcomes, low SOE supports the conclusion that ever breastfeeding or breastfeeding for longer durations may be associated with lower rates of breast cancer, epithelial ovarian cancer, hypertension, and type 2 diabetes, but not fractures. Because of heterogeneity and inconsistent results, we found insufficient evidence on whether breastfeeding is associated with postpartum depression, cardiovascular disease, or postpartum weight change. CONCLUSIONS: The body of evidence for breastfeeding programs and policies was diverse in terms of interventions and settings. Current evidence supports the benefit of BFHI for improving rates of breastfeeding initiation and duration; however, evidence from one large RCT (PROBIT) has limited applicability, and observational studies do not clearly establish the magnitude of benefit. For women enrolled in WIC, low SOE supports peer-support interventions for improving breastfeeding outcomes. The identified associations between breastfeeding and improved maternal health outcomes are supported by evidence from observational studies, which cannot determine cause-and-effect relationships.