Nutrient Secretion in Milk

Nutrient needs during lactation depend primarily on the volume and composition of milk produced and on the mother's initial nutrient needs and nutritional status. Among women exclusively breastfeeding their infants, the energy demands of lactation exceed prepregnancy demands by approximately 640 kcal/day during the first 6 months post partum compared with 300 kcal/day during the last two trimesters of pregnancy (NRC, 1989). In contrast, the demand for some nutrients, such as iron, is considerably less during lactation than during pregnancy.

Table 9-1 provides the estimated daily output of various nutrients in human milk compared with the increments in nutrient intakes for lactating women as specified in the Recommended Dietary Allowances (RDAs) (NRC, 1989). Each nutrient output is given as a range, reflecting milk volumes from 600 to 1,000 ml/day and the average concentration of each nutrient in human milk, as described in Chapter 6 (Table 6-1). Because there is variability in the milk concentrations of some nutrients, both between women and in the same woman at different stages of lactation, the estimates in Table 9-1 are provided primarily as an illustration of the relative levels of secretion in milk. They reflect the nutrient output by women exclusively breastfeeding a single infant; therefore, they are underestimates for those women breastfeeding twins or triplets and overestimates for those whose infants are given substantial amounts of other milks or solid foods.

TABLE 9-1

Estimated Secretion of Nutrients in Mature Human Milk Compared with Increments in Recommended Dietary Allowances (RDAs) for Lactating Women.

Previous editions of Recommended Dietary Allowances (e.g., NRC, 1980) presented RDAs for lactating women as increments to be added to the RDAs for nonpregnant, nonlactating women. In contrast, the 1989 RDAs for lactating women are presented as absolute amounts. This provides convenient numbers that apply to all lactating women regardless of age and reflects the limited precision of the published data. The absolute values in the 1989 RDAs represent the RDAs for nonpregnant, nonlactating women aged 25 to 50 plus increments, which are shown in Table 9-1. For many nutrients, it is evident that the RDA increments were designed to exceed the estimated daily output during lactation. The RDAs for lactating women are the same for mothers of all ages, but some of the RDAs for nonpregnant, nonlactating women differ for women of different age groups. The difference is most notable for vitamin D and calcium. For those two nutrients, no increment for lactation is included for women younger than 25. Their prepregnancy RDAs are considerably higher than those of the older group (10 µg compared with 5 µg for vitamin D, and 1,200 mg with 800 mg for calcium), and they are identical to the older women's RDAs for these nutrients during lactation.

Daily outputs of energy, vitamin A, vitamin B₁₂, iron, and iodine in milk tend to exceed the recommended increments. The allowance for energy is based on the assumption that lactating women can draw on fat stores deposited during pregnancy to help support milk production. Thus, gradual weight loss is expected during lactation. For women who are underweight or whose weight gain during pregnancy was low, a 650-kcal/day increase in energy intake during the first 6 months of lactation is recommended. Total energy needs during lactation depend greatly on the level of physical activity, as described in Chapter 5.

In estimating output for several nutrients, the subcommittee used values for milk concentrations that were higher than from those used in deriving the RDAs. These are identified in Table 9-1, along with other comments regarding the rationale for the RDA increments for lactation.

Long-Term Nutrient Output During Lactation

In addition to examining the daily output of nutrients in milk, it is useful to estimate the overall nutrient outputs resulting from various durations of lactation. Obviously, these estimates depend on the timing and degree to which human milk is replaced with other foods over the period of weaning. If the mother ingests lower amounts of a nutrient than she requires to meet her own needs and to cover nutrients secreted in the milk, she can draw upon body stores. To illustrate the possible impact of inadequate intake on the mother, Table 9-2 lists the estimated total body content of protein, calcium, and folate; the average amounts of iron and vitamin A stored by women; and the estimated outputs during partial or exclusive breastfeeding for various durations. Energy output in milk is not listed: changes in maternal body stores cannot be easily computed since energy-supplying nutrients could originate in either fat or lean tissue.

TABLE 9-2

Estimated Long-Term Demand of Lactation for Selected Nutrients.

As shown in Table 9-2, protein output in milk during 6 months of exclusive breastfeeding totals approximately 1,500 g (1.5 kg). For a 60-kg woman with 25% body fat, lean body mass is 45 kg, or about 11 kg of protein. Assuming that the efficiency of conversion of body protein to milk protein approximately equals the conversion of dietary protein to milk protein (˜70% [NRC, 1989]), a woman consuming only the RDA for protein for a nonpregnant, nonlactating woman would need to mobilize about 19% of her lean tissue to support 6 months of milk production. In the absence of adequate calcium intake, calcium output in milk represents 2 to 8% of total body calcium, as described in Chapter 8. Although iron losses in milk during 6 months of exclusive breastfeeding are equivalent to approximately 14% of the average woman's iron stores, this is only about half of what is ordinarily lost through menstruation (NRC, 1989). Thus, unless there was excessive blood loss at delivery, the total demand for iron during lactation is reduced while the woman is still amenorrheic compared with the demand when the woman is nonpregnant and nonlactating. When menstruation resumes, the combined demands of milk production and menstruation could draw heavily on iron reserves if dietary iron intake is low.

Vitamin A reserves vary greatly among women and may be precariously low in women whose habitual intake of the vitamin is marginal. In the U.S. population, on average, women are expected to have vitamin A stores of approximately 200 mg (NRC, 1989)—more than the total output of many women over usual periods of lactation (see Table 9-2). However, for a woman with much smaller stores whose diet barely meets her own vitamin A requirement, the loss of the vitamin in milk could theoretically deplete vitamin A stores within a few months.

Since folate tends to be excreted in the urine if intake exceeds demand, body stores of folate are not large. Thus, with limited dietary sources of folate, reserves would be depleted after only a few months of lactation.

The estimates discussed above are meaningful only in the context of dietary information. For example, total protein output in milk during 6 months of exclusive breastfeeding is high compared with total body protein, but this is not of great concern in the United States, where protein intake is usually generous. Calcium output during lactation may be of greater concern for reasons discussed in Chapter 8.

Predicted Average Intakes by U.S. Women During Lactation

One approach recommended for estimating the likelihood of nutrient deficiencies is to compare the distribution of intakes of each nutrient with the hypothetical distribution of requirements in a given population (NRC, 1986). If this method could be used, it would avoid the use of arbitrary cutoff values to define deficient intakes. As noted in Chapter 4, however, data on the usual dietary intakes of lactating women are not sufficient for this method to be applied, and the data that are available were generally obtained from relatively affluent groups, who are unlikely to be representative of the population as a whole.

For these reasons, and because energy needs are highly variable and energy intake is difficult to quantify, the subcommittee opted to evaluate the likelihood of nutrient shortfalls during lactation based on the nutrient density (nutrient intake per 1,000 kcal) of the average woman's diet in the United States. Total energy intakes by women are generally underestimated surveys; the nutrient density approach avoids this potential bias. However, the calculated nutrient densities could also be biased if underreporting is greater for certain foods than for others. No data are available to evaluate this possibility.

Table 9-3 provides nutrient densities for protein, minerals, and vitamins as determined from nationally representative samples of U.S. women aged 19 to 50 and the total nutrient intakes that would be predicted from those densities at three different levels of energy intake (nutrient density × kcal of energy = total intake). The three levels of energy intake chosen for the calculations are 2,700 kcal (the estimated energy needs for lactation listed in Recommended Dietary Allowances [NRC, 1989]), 2,200 kcal (a value closer to the average reported intakes of lactating women, as described in Chapter 4), and 1,800 kcal (a level that might occur if a lactating woman were actively restricting food intake in order to lose weight). Intakes of specific nutrients at these three levels of energy intake can be compared with the RDAs for lactation, also listed in Table 9-3.

TABLE 9-3

Estimated Mean Nutrient Intakes by U.S. Women at Three Energy Levels Compared with the Recommended Dietary Allowances (RDAs) for Lactation.

Comparisons of Estimated and Recommended Intakes

Several patterns become apparent when comparing estimated intakes with the RDAs shown in Table 9-3. At an intake of 2,700 kcal, average predicted intakes are below the RDAs for only two nutrients: calcium and zinc. At an intake of 2,200 kcal, however, predicted intakes of calcium, magnesium, zinc, thiamin, vitamin B₆, and vitamin E fall below the RDA. At 1,800 kcal, predicted intakes fall below the RDAs for all the nutrients listed above plus riboflavin, folate, phosphorus, and iron.

Several cautions are necessary in interpreting these patterns:

• The RDAs for most nutrients include a wide margin of safety, so intakes below the RDAs do not necessarily indicate inadequacy. Of the nutrients for which predicted average intake is below the RDA, most intake estimates range from 73 to 97% of the RDA, except intakes for zinc, which are 67 and 55% of the RDA (at 2,200 and 1,800 kcal, respectively) and for calcium and vitamin B₆, which are 60 and 67%, respectively (each at 1,800 kcal). There is also a wide range in the nutrient densities used to generate the averages shown in Table 9-3. Even if the average predicted intake of a given nutrient is at or above the RDA, the intakes of a substantial proportion of the population would fall below that level.
• The margin of safety in the RDA varies by nutrient. As shown in Table 9-1, for example, the increment in the RDA for vitamin B₆ for lactation is more than five times the estimated secretion of this vitamin in milk. In contrast, the increment for folate is less than two times the estimated secretion.
• Overt signs of deficiency are extremely rare in the United States, even for nutrients with small safety margins.
• Nutrient densities of diets consumed by lactating women are likely to differ from those for U.S. women as a whole. Inspection of data for lactating women shown in Chapter 4 indicates that densities of calcium and vitamin A, for example, are 40 to 50% higher in the diets of the lactating women surveyed than the values shown in Table 9-3, possibly because greater quantities of milk products are consumed. It could be argued either than women who choose to breastfeed are women who ordinarily have high nutrient intakes, or that women consume more nutrient-rich foods when breastfeeding. Whatever the case, higher than average densities of some nutrients in diets of the few lactating women studied to date are not grounds for complacency, because the samples in these studies were generally not randomly selected. There is no evidence that the same nutrient intakes would pertain to breastfeeding women in less affluent or less educated groups.

Keeping these cautions in mind, this subcommittee attempted to determine which nutrients are most likely to be in short supply in the diets of lactating women, relative to the RDAs, and the consequences of any such shortfalls.

Nutrient Concerns for Selected Groups of Lactating Women

The discussion above focuses on nutrients likely to be consumed by lactating women in the United States in amounts lower than the RDAs. Within certain age, income, and ethnic groups, however, the diet may be characterized by nutrient densities higher or lower than the averages shown in Table 9-3. Intake data from national surveys are available by age, race (white, black, and other) (e.g., USDA, 1987), and income (above and below the poverty level) (e.g., NCHS, 1983; USDA, 1988). Such data may be useful in identifying potential nutrient inadequacies in selected groups.

Following are four examples of groups whose diets, on average, have nutrient densities lower than the values shown in Table 9-3. An energy intake of 2,200 kcal was assumed when calculating recommended nutrient densities.

• Diets of adolescents (aged 15 to 17) typically contain less iron (an average of 5.7 mg/1,000 kcal) (NCHS, 1983) than recommended during lactation (6.8 mg/1,000 kcal).
• Diets of adolescents with family incomes below the poverty level have a low vitamin A content (1,500 IU/1,000 kcal) (NCHS, 1983) compared with a desired density of 1,950 IU/1,000 kcal.
• On average, the diets of black women contain about 30% less calcium, 20% less magnesium, and 20% less vitamin A than average diets consumed by white women (USDA, 1987).
• Diets of low-income adult women are characterized by lower densities of calcium and vitamin A than are typical of diets of women above the poverty level (NCHS, 1983).

Thus, special care should be taken to ensure that breastfeeding women in such groups have access to a nutrient-dense diet.

Women with restricted eating patterns will have undesirably low intakes of certain nutrients. This applies to those whose total food and energy intake is low (unless nutrient density is unusually high) and to those who avoid foods that are major sources of nutrients, such as calcium-rich dairy products, vitamin D-fortified milk, animal foods (for vitamin B₁₂), or fruits and vegetables (for folate and vitamin C).

In some cultural groups, beliefs regarding foods that should not be consumed by lactating women may affect dietary patterns (Baumslag, 1986), but the influence of restrictive food beliefs on nutrient intake is not well documented.