Low iron content of cow's milk
The iron content of CM is about 0.5 mg/L and is thus comparable to that of human milk. Like human milk, CM alone cannot meet the iron needs of infants or toddlers regardless of the amount of CM consumed. The Institute of Medicine18 has estimated that the growing infant needs to have a net iron gain of 0.7 mg per day, and therefore needs to take in 7 mg of iron every day; the toddler needs to take in 11 mg/day. If the availability of iron from CM were assumed to be 10%, the infant would have to consume 14 L of CM every day to meet iron requirements. Clearly, neither infants nor toddlers can depend on CM to meet their iron needs. Rather, infants and toddlers depend on other foods or iron supplements to meet the body's demands for iron. In contrast, formulas that have iron concentrations between 6 and 12 mg/L easily meet infants' iron needs. Most complementary foods are low in iron unless they are fortified with iron. Fruits and vegetables are also low in iron content. Only meats and meat-containing infant foods contain appreciable amounts of iron, but they are only sparingly consumed by infants. Infant cereals and, to some extent, cereals for older children are, at least in the United States, fortified with iron and are the primary source of iron for infants and toddlers. Given this dependency on certain complementary foods to ensure the necessary iron intake, it is not too surprising that infants and toddlers fed CM are at risk of ID.
Occult intestinal blood loss
Healthy infants lose small quantities of blood in their feces at all times. Using the radiochromium (51Cr) technique, in which erythrocytes are withdrawn from the subject, labeled in vitro with 51Cr, and then reinjected into the subject before feces are collected and examined, quantitative determination can be made of the amount of blood lost in feces. Using this technique, Elian et al.19 found that blood loss averaged 0.59 mL/day in infants hospitalized for various non-intestinal infectious illnesses. Each infant had measurable blood loss. Wilson et al.20 were the first to recognize that the feeding of CM to infants and toddlers can lead to an increase in intestinal blood loss. In a subgroup of a cohort of infants and toddlers with ID, they found a clear and reproducible increase in fecal blood loss associated with the feeding of CM. Infants with consistent CM-induced blood loss lost, on average, 1.7 mL of blood each day. When the same subjects were fed formula (milk-based or soy-based), blood loss decreased to 0.3 mL/day. Blood loss of 1.7 mL/day is equivalent to iron loss of 0.53 mg/day, which, if it is not causing outright negative iron balance, makes it, at the least, very difficult for the infant to achieve the 0.7 mg/day net iron gain that is necessary to prevent iron depletion.
To answer the question of whether iron-sufficient infants can also have occult intestinal blood loss when fed CM, Fomon et al.21 randomly assigned healthy 4-month-old infants to pasteurized CM, heat-treated CM (heat-treated to the same extent as ready-to-feed infant formula), or iron-fortified infant formula. Between 112 and 196 days of age, stool specimens were collected weekly and tested for the presence of blood using the guaiac test. During the first 4 weeks of the trial, 39% of infants fed pasteurized CM had one or more guaiac-positive stools and 17% of all stools were guaiac positive. In contrast, among infants fed formula or heat-treated CM, only 9% of infants had one or more guaiac-positive stools (P < 0.01) and only 2.4% of all stools were guaiac positive (P < 0.001). There were no differences in iron nutritional status, possibly because iron intake was generous in all groups.
The study by Fomon et al.21 provided no information about the quantity of blood lost during feeding with CM, and it involved relatively young infants. A subsequent study of similar design employed a quantitative method for the determination of fecal hemoglobin.22 Also, it initiated the feeding of CM at a later age (5 ½ months). One group of healthy infants was fed CM and a second group was fed iron-fortified milk-based formula. Infants had initially been breastfed or had been fed milk-based or soy-based formulas, but for at least 1 month before entering the trial all infants were fed a milk-based formula. As in the earlier study, the percentage of guaiac-positive stools rose from a baseline of 3% to 30% during the first month of feeding with CM (P < 0.01), whereas among infants fed with formula the proportion of guaiac-positive stools remained low (5%). The mean stool concentration of hemoglobin increased from 622 µg/g dry stool at baseline to a mean of 3,598 (standard deviation 10,479) µg/g dry stool during the first month of feeding with CM, as illustrated in Figure 1. In infants fed with formula, stool hemoglobin concentration did not change from baseline and remained significantly (P < 0.01) lower than in infants fed with CM. Again, a large minority (38%) of infants fed with CM reacted to the milk (responders) and accounted for all of the observed increase in mean stool hemoglobin. In the majority of infants (non-responders) stool hemoglobin concentrations were indistinguishable from those of formula-fed control infants.
Figure 1. Fecal hemoglobin concentration in infants fed formula and infants fed CM beginning at 168 days of age. Drawn from data of Ziegler et al.22
Download figure to PowerPoint
One responder infant in the study of Ziegler et al.22 had massive blood loss (average iron loss of 2.04 mg/day) and developed iron deficiency anemia after just 1 month on CM. The remaining nine responders lost, on average, 0.24 mg of iron per day, an amount that is not trivial in infants whose ability to gain 0.7 mg of iron per day is already compromised by low iron intake due to being fed CM. It should be noted that intestinal blood loss was clinically silent and that feeding-related behaviors did not differ between responders and non-responders. Also of note is that infants who were fed milk-based formula from birth were less likely (P = 0.059) to respond to CM with intestinal blood loss than infants who were breastfed from birth.
Subsequent studies23,24 answered the question of whether CM-induced intestinal blood loss also occurs in older infants. Infants were fed CM for 2 months starting at 7 ½ months, 9 months, or 12 months of age. About half of the study participants were initially breastfed for various lengths of time, while the other half of infants were fed formula only (including soy-based formula). While being fed with CM beginning at 7 ½ months (224 days) of age, fecal hemoglobin concentrations rose significantly (P < 0.05), but the increase was, on average, less pronounced than at 5 ½ months of age. Among the 26% of infants who were classified as responders, average fecal hemoglobin was 3,010 µg/g dry stool (Figure 2). Among the responders were two infants with persistent high fecal hemoglobin concentrations that averaged 7,430 and 8,156 µg/g dry stool, respectively.
Figure 2. Summary of results from three studies in which CM was introduced at different ages.22–24 Open bars represent infants fed with CM who did not respond to CM. Hatched bars represent those responding to CM with increased fecal hemoglobin. Numbers on top of the bars indicate the percentage of responders.
Download figure to PowerPoint
When CM was introduced at 9 months (280 days) of age, 29% of infants showed a response and had stool hemoglobin concentrations that were significantly (P < 0.01) higher than at baseline. But when CM was introduced at 12 months of age, it produced no increase in fecal hemoglobin above baseline (Figure 2). Two infants (7%) showed a response, but it was quite mild and short-lived. Of note is that baseline fecal hemoglobin concentrations showed a marked increase with age. They averaged 657 µg/g dry stool at 5 ½ months of age and rose steadily until they reached 1,395 µg/g dry stool at 9 months of age and 1,194 µg/g dry stool at 12 months of age (Figure 2). The increase probably reflects the increasing presence of unabsorbed food heme (for example, from meat) rather than elevated baseline blood loss, but the latter possibility cannot be excluded.
Inhibition of iron absorption by components of cow's milk
In contradistinction to heme iron, the absorption of non-heme iron, which constitutes the bulk of dietary iron, is subject to inhibition by substances that are common in the diet, including casein and calcium.25 Casein and calcium are present in CM in much higher concentrations than in human milk. In adult subjects, iron absorption from human milk was found to be 37.3% while it was 15.5% from CM.26 When the calcium concentration of human milk was raised to the same level as in CM, iron absorption from human milk was only 19.2%. Thus, calcium explained about 70% of the difference in iron bioavailability between human milk and CM.26 No similar studies have been reported for infants. Hurrell et al.27 showed that casein and other milk proteins strongly inhibit the absorption of iron. Thus, it is evident that CM, with its high calcium and casein content, is likely to have an adverse impact on the availability of the already modest amounts of iron that infants and toddlers ingest with complementary foods.
Iron fortification of cow's milk
Being a rich source of protein and minerals, CM plays an important role in the overall nutrition of infants and young children. In some countries the government provides CM to low-income populations free of charge or at subsidized prices. The adverse effect of CM on iron status detracts from the value of CM as a nutritional staple. Fortification of CM with iron offsets the negative effect of CM on iron metabolism. In addition, fortification takes advantage of the widespread use of CM, using it as a vehicle for general iron supplementation as part of efforts to combat ID. Iron fortification of CM is technically feasible and has no effect on the organoleptic properties of the milk.
Several studies have documented the effectiveness of iron fortification of CM. In Mexico, a program of distributing iron-fortified CM has been shown to be effective in reducing the prevalence of anemia and ID.28,29 Chile has had a milk fortification program for several decades and its effectiveness has been documented.30,31 Similarly, in Indonesia the distribution of iron-fortified milk has led to a reduction in the prevalence of anemia among infants and young children.32 Although these reports demonstrate that iron fortification of CM is associated with better iron nutritional status than unfortified CM, it remains unclear whether fortification completely offsets the adverse effects of CM. One variable that seems crucial in this regard is the amount of iron added to the milk. Nevertheless, it is established that iron fortification of CM is feasible and has proven efficacy for improving the iron status of infants and children.