the infant. I shall look carefully at these areas to help pave the way to a more solid basis for modern breastfeeding medicine. Firstly, I shall look at evolutionary theory for human breastfeeding and consider in general terms the implications for optimal nutritional care of breastfed infants. Secondly, I shall show how HM composition has been incorrectly translated into dietary intake in a large body of past flawed work that resulted in misleading data. Implementing such data as a model for infant formula appears to have increased the risk of obesity and cardiovascular disease (CVD) in formula-fed infants. Finally, most studies that examine the benefits of HM are observational and potentially confounded. So, this body of data needs to be backed by experimental evidence. Here, I shall use preterm infants as a model, since numerous RCTs and physiological studies over 40 years have compared exclusive HM feeding versus cow’s milk exposure. Unexpectedly diverse immediate beneficial effects span the field of neonatology, and long-term programmed effects have been shown for cognition, brain structure, risk factors for CVD, structural development of the heart and lungs, bone health, and atopy. These data add much weight to the evidence, obtained in full-term infants using observational study designs, that HM feeding in early life may fundamentally and permanently change the biology of the organism. Breastfeeding is emerging as a major evidence-based field of medical and public health practice.
© 2019 Nestlé Nutrition Institute, Switzerland/S. Karger AG, Basel
Introduction
Given 7 billion people on the planet and 134 million births/year, the recommendation that all babies should be breastfed constitutes a colossal public health intervention. All public health interventions should be rooted in sound scientific evidence and I shall consider some modern advances in the science and understanding of this important field.
I shall focus on 3 important pillars in breastfeeding medicine: (1) that human milk (HM) is the product of 200 million years of mammalian evolution; (2) that HM composition is the gold standard for infant nutritional requirements; and (3) that HM has numerous clinical benefits for the infant.
Finally, I shall emphasize the great importance of breastfeeding as an evidence-based clinical and public health intervention.
Breastfeeding and Mammalian Evolution
With 200 million years of mammalian evolution, breast milk has evolved major diversity – for instance a 2% concentration of fat in mare’s milk contrasts with over 40% fat in the milk of the harp seal, where the offspring must survive extreme cold. Nevertheless, the application of evolutionary biology to human breastfeeding requires some special considerations with potential implications for practice [1].
Until relatively recently, humans lived in hunter-gatherer societies, but, in a short period, as intelligent primates, humans changed their environment dramatically, whereas our genes are still ancient. The consequent mismatch between our genes and environment is known as “evolutionary discordance.” As Cordain et al. [2] noted for adult humans, the principal phenotypic manifestation of evolutionary discordance is disease. Thus, it is proposed that the high incidence of obesity and cardiovascular disease (CVD) in modern humans is due to the mismatch between genetic adaptation and our modern diet – an example of evolutionary discordance.
The question of relevance here is whether human breastfed infants are affected by evolutionary discordance and how this should be managed to complement the considerable value of breastfeeding identified later in this article. Thus, modern mothers eat less green leafy plants than our ancestors and presumably have less vitamin K in their breast milk [3]. This may explain the past occurrence of late vitamin K deficiency bleeding in modern breastfed infants – a condition that had a high incidence of intracranial bleeding. Thus, all babies now receive prophylactic vitamin K after birth. A further example is that a consequence of recent migration of human populations into less light-exposed areas of the globe is increased propensity to vitamin D deficiency, which may require vitamin D prophylaxis. An intriguing hypothesis to explain the occurrence of early iron deficiency anemia comes from the observation that piglets put in a concrete pen develop iron deficiency since pig’s milk is relatively low in iron and a concrete pen prevents iron intake from soil [4]. Hallberg [5] speculated that early human infants might have eaten soil to supplement the iron received from human breast milk, but with environmental change and modern public health, modern infants no longer consume iron from soil.
One consequence of the major recent change in the diet of humans is that the n-6/n-3 fatty acid ratio in the diet of hunters-gatherers is believed to be around 1: 1 whereas with a modern Western diet this ratio is around 15: 1, reflecting a relatively low n-3 fatty acid status in modern mothers [6]. The impact of supplementing the diet of a lactating mother with n-3 fatty acids is not established but does at least raise the hypothesis for future testing that nutritional status of the offspring might be further optimized by dietary care of breastfeeding mothers.
In summary, current evidence (see later) shows that breastfeeding is superior to its substitutes on numerous health grounds. Nevertheless, given the evolutionary aspects considered, it is in the interests of population health to identify areas in which nutritional care of breastfeeding mothers or their babies could further improve outcome – a principle already in practice in relation to the use of prophylactic vitamin K and vitamin D in infancy.
Breast Milk Composition as the Gold Standard for Infant Nutritional Needs
HM composition has generally been regarded as a gold standard for deriving infant nutritional requirements – for instance in situations where artificial feeding is required. This has certainly been a most helpful concept.
However, for breast milk to be a valid gold standard, it is critical that accurate data are obtained using appropriate methodology. This latter aspect is the one that is discussed in this section since it will be argued that despite intensive work on the composition of breast milk, misleading data have been derived in the past that have misdirected nutrition practice in ways that have had adverse impact on babies and their long-term health.
In 1953, Hoobler et al. [7] were able to summarize no less than 1,500 scientific publications on the composition of HM. In 1977, the UK Department of Health added further to this list: an official publication on the nutrient content of breast milk obtained by complete expression of one breast in mothers from 4 UK cities [8]. These data were proposed to provide a basis for infant nutritional needs and a model for the design of infant formulas. It was at this stage that this and past studies on breast milk composition were challenged as methodologically flawed [9].
One major difficulty in the study of breast milk content is obtaining representative samples of breast milk for analysis. Breast milk fat, and hence energy content, varies greatly during a feed, between breasts, and throughout lactation. Our own data show that during the course of a breastfeed, breast milk fat content doubles, and milk flow, statistically at least, decreases in a curvilinear manner [10]. We suggested that obtaining representative values for milk fat might be best derived by studying milk composition and milk flow during the feeding process, yet this had never been done. In the absence of such knowledge, we hypothesized that milk obtained for analysis in the traditional manner by unphysiological manual or mechanical expression of the breast – so-called expressed breast milk (EBM) – might differ greatly in its fat and energy content compared to the milk obtained by the baby during physiological breastfeeding – termed by Lucas “suckled