Human

Objectives. To review the evidence on the association between breastfeeding and infant and child health outcomes, including the extent to which these associations vary by the intensity, duration, mode, and source of breastmilk consumption. In this review, breastfeeding refers to feeding breastmilk whether directly from the breast or other means and includes breastmilk from pasteurized donor milk.

Data sources. Systematic literature searches in MEDLINE, Embase and CINAHL for English-language articles published from 2006 to August 14, 2024. We identified additional studies from reference lists and technical experts.

Review methods. We worked with our sponsor and a panel of technical experts to identify the outcomes of interest for this review. The evidence for more than 20 outcomes was synthesized, including outcomes related to infectious diseases, asthma and allergic conditions, oral health, autoimmune gastrointestinal conditions, endocrine conditions, cardiovascular disease (CVD), childhood cancer, cognitive development, and infant mortality. We relied on existing systematic reviews (ESRs) for all outcomes and conducted bridge searches for newer primary studies since the search date of the most recent and relevant ESR. Studies were evaluated for eligibility and quality, and data were abstracted on study design, demographics, breastfeeding exposures and referents, and outcomes. We synthesized the evidence by outcome, summarizing the results of ESRs alongside those of newer primary studies. No meta-analyses were conducted given the combination of ESR and primary study evidence and heterogeneity in exposures and outcomes; but figures were created to visually display point estimates across studies.

Results. A total of 29 ESRs and 145 primary studies were included. The cumulative number of studies included for each outcome varied from only 4 studies examining the relationship between breastfeeding and type 2 diabetes to more than 180 studies reporting on the relationship between breastfeeding and obesity-related outcomes. We rated the strength of evidence as “Low” or “Moderate” for most outcomes, given limitations of the underlying evidence base, along with concerns related to heterogeneity of the study designs, and the consistency and precision of results. An association indicating a reduced risk from “more” versus “less” breastfeeding was most apparent for otitis media, asthma, obesity in childhood, and childhood leukemia. A protective association of breastfeeding was also found for severe respiratory and gastrointestinal infections in younger children, allergic rhinitis, malocclusion, inflammatory bowel disease, type 1 diabetes, rapid weight gain and growth, systolic blood pressure, and infant mortality, including sudden unexpected infant death, although our confidence in these findings was lower. There was no apparent association for the outcomes of atopic dermatitis, celiac disease, and cognitive ability. An association indicating an increased risk of dental caries was noted for breastfeeding 12 months or longer. There was insufficient evidence to draw conclusions about the relationship with food allergies and type 2 diabetes and no data on coronavirus disease 2019 (COVID-19) or CVD endpoint outcomes (i.e., events or mortality). While nearly all outcomes had evidence on ever (versus never) breastfeeding, exclusive (versus nonexclusive or no) breastfeeding, and longer durations (versus shorter or no) of any or exclusive breastfeeding, the exposure comparisons and categorizations reported in the ESRs and primary evidence made it extremely difficult to examine the nuances of these relationships. There was no clear “threshold” of breastfeeding that appeared to be most beneficial for any outcome. Furthermore, there were little data on how the relationships varied by mode of breastfeeding or source of breastmilk.

Conclusions. Breastfeeding is associated with beneficial effects for several infant and child outcomes, although there are limitations to the data that preclude high certainty in the findings. Further research that addresses the limitations of existing studies is needed to continue to inform national guidelines and initiatives.