An objective assessment of the totality and quality of the evidence based on a comprehensive systematic review is required to support an update of a Dietary Reference Intake (DRI) nutrient review. The scanning process described herein is not intended to replace the need for a comprehensive systematic review. Rather, an evidence scan, as operationalized in this report, enables a cost-effective and objective means of assessing whether there is sufficient new evidence to merit a formal reexamination of a nutrient or nutrients to support a DRI review.
This evidence scan was carried out to elicit new information on a range of micro- and macronutrients in human milk and the volume of milk consumed by breastfeeding infants. The scanning process required continuous evaluation and adjustment to obtain the most relevant evidence to inform development of future DRI values. Although the committee was not charged to make recommendations, it offers comments on the approach to the evidence-scan process, and discusses its findings and interpretation of the evidence in this chapter.
Design of the Search Process
The committee made a deliberate choice to be inclusive in the design of its search. An example of this approach was the committee’s choice to go beyond the United States and Canada to accept data from countries that were ranked as high-income or high middle-income by the World Bank. The committee used its own judgment to decide whether to exclude reports from high middle-income countries that would have not been in this classification when the investigation was done, often decades ago. The committee chose to use these two groups of countries as they helped it to distinguish environments that might produce results for milk volume and composition that would be similar to those obtained among contemporary American women.
The committee further acknowledges that contemporary American women have likely received prenatal supplements during gestation and may still have been taking them at the time of milk sampling. Very few reports included information on women’s prior or current use of prenatal or other supplements. As a result, it is possible that the milk composition values reported here reflect this exposure and, thus, these results must be interpreted in light of this possibility.
Selection of Inclusion and Exclusion Criteria
The committee selected its inclusion and exclusion criteria on the basis of identifying studies (or a subset of data within them) that would provide the most reliable estimate of milk volume (measured as infant milk intake, not milk production) and nutrient composition. The selected criteria reflect the expertise of the committee as well as that of subject-matter experts
who provided information about approaches for milk sampling and analysis to the committee in a public session. This approach was useful because it eliminated studies that, while well established in the literature, were not relevant to the committee’s task because of small sample sizes (often fewer than 10 women per group) or methods now considered less rigorous than currently accepted methodologies. The criteria for milk sampling further allowed the committee to discriminate among studies that included data for some nutrients that could be used and data for other nutrients that could not. This process reinforced the importance the committee placed on the proper sampling of milk for the nutrient(s) to be analyzed. (See Chapter 2 and Appendix D for more information about milk collection and analytical methods.) It should be noted, however, that there are currently insufficient data to determine optimal sampling protocols for all nutrients.
The Screening Process
The number of records retrieved in the committee’s search, even after the criteria were refined and revised iteratively, was much larger than anticipated. However, the abstract-screening process, as expected, revealed a large number of records that did not meet the committee’s inclusion criteria. For example, many studies focused on oligosaccharides in human milk. These were not included because there are, at present, no DRI values for these milk components. Among studies on trace minerals, some were eliminated when the research objective was analysis of toxic levels of minerals or contaminants (e.g., lead) while others were eliminated because care was not exercised in using trace mineral-free collection containers.
Notably, a high proportion of studies were conducted outside of the United States and Canada. At the stage of full-text screening many of these studies were excluded when the committee determined that the populations studied across low- and middle-income countries were not comparable to those in the United States and Canada. Among the remaining studies reviewed, a report was deemed relevant only when it described participants (e.g., mothers of term infants who served as the comparison group for mothers of preterm infants were included in the study population) who met the criteria. In numerous other cases, the population studied, methods used, or the data obtained were reported with insufficient rigor for the findings to be useful to the committee. Altogether, the majority of studies were excluded owing to inadequate methods at both the full-text screen (26 percent) and the final full-text rescreen (86 percent). Thus, at the final, full-text rescreening, only 126 of the 398 fully evaluated articles were determined eligible for final assessment.
A striking feature of Tables 3-1, 3-2, and 3-3 is the disproportionate number among the types of relevant studies reported. The committee found only 32 eligible reports that included data on milk volume since 1980, compared to 102 eligible reports on milk composition in the same time period. Given the public health messages about the importance of human milk and the increasing proportion of women who breastfeed or feed pumped milk, the committee considered this result disproportionately low. Additionally, given the recommendation to introduce complementary foods to infants’ diets at about 6 months of age, and the fact that only a minority of U.S. infants are exclusively breastfed at and after this age, the committee expected to find few reports of the measurement of milk volume in infants older than 6 months. Indeed, only three
such reports met all of the committee’s inclusion criteria, except that the mothers should be exclusively breastfeeding their infants (Neville et al., 1988, and Stuff and Nichols, 1989, in Table 3-2; and Krebs et al., 1994, in Table 3-3). Krebs et al. (1994) found that 93 percent of infants were not exclusively breast fed by 7 months of age. Neville et al. (1988), the only report that included longitudinal data, documented the expected decline in milk volume from 6–12 months of age. While it has long been known that milk composition also changes as milk volume declines after the introduction of complementary foods, the committee notes that additional research will be needed to define the volume and composition of milk produced by women for infants 6 months of age and older.
In another observation, the committee found that despite increased interest in essential fatty acids within the research community, only three of the identified studies met the screening criteria. Exclusion of other studies on this subject was based primarily on the fact that relative amounts (rather than absolute concentrations) of the fatty acids were reported without also reporting total lipids. As described above, in some studies, collection methods were not optimized (e.g., not a full breast expression) and/or the method (e.g., creamatocrit) used for measuring total lipid was not deemed adequate by the committee. This finding highlights that some data uses, such as DRIs, require calculation of the total amount of fatty acids that a nursing infant consumes. This means that it is important to determine both the total amount of lipid in the milk sample as well as the absolute amounts of each fatty acid per unit volume of milk.
Assessment of Findings Relative to Methodologies Used
Among reports published since 1980, only a small number met the committee’s inclusion criteria. Given the many new methods that have been developed for measuring these components (see Appendix D), researchers are now poised for a reexamination of these essential nutrients. The committee found a similar number of studies on trace minerals (81),1 macrominerals (80),2 and electrolytes (80),3 in human milk. There were 24 studies on macronutrients. Two aspects of the scanning process merit further comment related to the methodologies reported and limitations in the evidence. First, one of the committee’s criteria was that the lactating women studied should be “healthy.” Some study authors asserted that participants were healthy but did not provide further details about what this meant and how it was determined. Other authors were clear about the selection process for their study population(s), and they provided data about their characteristics (e.g., age, body mass index, absence of specific conditions and/or behaviors, such as smoking). The committee decided to accept the authors’ description of the subjects as being “healthy.” These considerations meant that some of the reviewed reports could contain data on participants who may not meet the criteria as defined by a DRI committee (such as those who smoke or have obesity).
The problem of defining what is meant by “healthy” also applied to infants. The committee also accepted the authors’ description of the infant as being healthy when this was
1 Trace minerals are classified as minerals required in the diet in daily amounts of 100 milligrams or less. These include copper, zinc, selenium, iodine, chromium, fluoride, manganese, molybdenum, and others.
2 Macrominerals are classified as minerals required in the diet in daily amounts greater than 100 milligrams. These include calcium, phosphorus, magnesium, sodium, potassium, chloride, and sulfur.
3 Electrolytes are charged minerals whose functions include maintenance of intra- and extracellular fluid balance, maintenance of metabolic pH, and nerve transduction. Electrolytes include sodium, calcium, potassium, chloride, phosphate, and magnesium.
provided. However, the criteria for adequate growth in infants changed during the 40-year scope of this evidence scan and were poorly described or not at all by the study investigators, thus it was difficult for the committee to discern if some of the infants studied would be considered to be growing adequately according to current growth standards.
Second, there was substantial variation in the quality of the methods used for analysis of milk composition. This variation meant that some of the articles, such as those with more limited descriptions of how the work was done, may be adequate while others may not. Thus, articles with limited details should be used with caution. The committee’s criteria related to how milk was sampled and the methods that were used to analyze it. These criteria did not require reporting of the quality-control procedures used in the analyses, for example. This factor will therefore require attention in future applications of the evidence presented in this report.
The committee adhered to the DRI framework when it reported milk volume and composition values separately from birth to 5.9 and from 6 to 12 months. Given that the composition of colostrum differs greatly from that of more mature milk, the committee eliminated reports or parts of reports that included data on colostrum. Consequently, as can be seen from the values in Table 3-1, the concentrations of some nutrients vary widely within the period from birth to 5.9 months and the period from 6 to 12 months, respectively. For other non-DRI applications, users of these data may find it advantageous to subdivide the age ranges further.
In this evidence scan, which covered over a 40-year period, a total of 126 studies were identified as relevant to milk composition and volume based on the committee’s criteria. From among these, the committee identified only 8 final articles that included data on both the volume of milk consumed by the infant and its nutrient composition combined. As is clear from this evidence scan, both milk volume and milk composition are highly variable among individuals. Information on both, preferably from the same mother–infant dyad, are required to estimate an infant’s nutrient requirement accurately.
To inform development of future DRI values for infants from birth through 12 months, for at least some nutrients, it is possible that sufficient numbers of participants have been studied so that new DRIs s could be developed from the data available. For other nutrients additional research will be needed before sufficient data are available to consider a revision of the current DRIs (see Figures 3-2 and 3-3).
In developing its criteria for methodologic rigor, it became clear to the committee that more research would be needed on the question of what constitutes an appropriate milk sampling strategy, as this is poorly understood for many nutrients. In addition, many intriguing components of human milk have been identified in recent years, and the research community is gaining a better understanding of the implications of some milk components (e.g., oligosaccharides) for infant health. However, for others (e.g., pluripotent stem cells), their importance is not yet understood. Thus, attention to these several kinds of components of human milk will be important in the future.