Iron supplementation of breastfed infants


EE Ziegler, Fomon Infant Nutrition Unit, Department of Pediatrics, A-136 MTF, Oakdale Research Park, 2501 Crosspark Road, Coralville, IA 52241-8802, USA. E-mail:, Phone: +1-319-335-4570, Fax: +1-319-335-4856.


Reported here are three studies performed with the objective of finding ways to improve the iron status of breastfed infants and to prevent iron deficiency (ID). Participating infants were exclusively breastfed until 4 months of age; thereafter, they could receive complementary foods and, in some studies, supplemental formula. In the first study, infants were given medicinal iron between the ages of 1 and 5.5 months. During this period, iron status improved and ID was prevented; however, these benefits did not continue after the intervention ceased. In the second study, infants received medicinal iron or an equivalent amount of iron from an iron-fortified cereal between the ages of 4 and 9 months. Again, iron supplementation largely prevented ID from occurring, while non-anemic ID and ID anemia occurred in the control group as well as in the intervention groups before the intervention began. In the third study, infants received dry cereals fortified with electrolytic iron or with ferrous fumarate between the ages of 4 and 9 months. The cereals were equally effective in providing relative protection from ID. The results of these three studies indicate it is possible to protect breastfed infants from ID and IDA.


When high need for a nutrient meets with a diet that is low in the nutrient in question, there is a risk of deficiency. The breastfed infant exemplifies this constellation with regard to iron. The need for iron is high due to rapid growth, and the diet (breast milk and typical complementary foods) is low in iron. During the first few months of life, the breastfed infant is protected from iron deficiency (ID) by the usually generous amount of iron the infant receives from the mother. This birth iron endowment (BIE) provides the infant for some time after birth with the iron needed for growth. This independence from dietary iron ends when the iron endowment becomes exhausted, which usually happens around 4–6 months of age. The infant then needs iron from exogenous sources to meet its needs for growth and the replacement of inevitable losses. Breast milk contains only small amounts of iron (0.2–0.4 mg/L), which, in spite of being highly bioavailable,1,2 meet only a fraction of the infant's needs, estimated at 6.9 mg per day between 7 and 12 months of age.3 Given these facts, it is not too surprising that ID and even iron deficiency anemia (IDA) can occur in breastfed infants. For the present discussion, ID is defined as a state with exhausted iron stores, indicated by a plasma ferritin concentration of <10 µg/L. When ID is accompanied by evidence of limited tissue iron availability, such as anemia, ID is considered severe.


Iron deficiency in the first six months of life

It is well documented that some breastfed infants develop ID before 6 months of age, a condition known as early ID. Makrides et al.4 reported that, in Australia, 15% of 6-month-old breastfed infants had ID and 1% had IDA. In Sweden, Domellöf et al.5 found the prevalence of IDA among 5-month-old infants to be about 1% and Lind et al.6 found IDA in 2% of 6-month-old infants. ForNorway, Hay et al.7 reported the prevalence of “low iron status” to be 4% at 6 months. For Chile, Lozoff et al.8 reported that 3.6% of predominantly breastfed infants had IDA at 5–6 months of age, while in Turkey, Arvas et al.9 found IDA in 9.5% of breastfed infants at 4 months of age. Denmark seems to be an exception in that no ID was found at 6 months of age.10 In our own studies of breastfed infants in the United States, early ID and IDA have been observed with regularity (see below).11–13 It is thus evident that not all breastfed infants are protected from ID by their BIE during the first 6 months of life.

A measure of the size of an infant's BIE may be obtained from the concentration of ferritin in cord blood or blood obtained within 2 months of birth. Based on ferritin concentrations, which show variation as large as 20-fold,11–17 the size of the BIE is known to vary greatly. Infants born with a low BIE would be expected to be at high risk of ID before the age of 6 months. Tamura et al.15 showed that infants born with cord ferritin concentrations in the lowest quartile (<76 µg/L) were at risk of impaired mental and psychomotor development at 5 years of age. Georgieff et al.18 reported that infants born with low iron stores have poorer iron nutritional status at 9 months of age than infants born with normal stores. Our own observations11–13 seem to be the first to show that a low BIE can lead to ID in the first 6 months of life. Without exception, infants who developed IDA in the first 6 months of life were born with a low BIE, whereas those who developed ID only were not uniformly born with a very low BIE. Although the cause of variation in BIE is not known, it must be assumed that low maternal iron status plays a role, albeit a role of unknown magnitude.

Iron deficiency late in the first year of life

During the second 6 months of life, most infants have used up the iron provided to them by their mothers. Only the infants born with the largest BIEs continue to derive protection against ID.16 After the BIE has been used up, iron status reflects the balance between iron intake and iron use for growth. Given the low iron content of breast milk, it is not too surprising that an association between the duration of breastfeeding and the occurrence of ID has been documented in several localities, including Canada,19 Korea,20 Turkey,9 Mexico,21,22 and the United States.23 In our own studies of breastfed infants about 6% of infants developed ID between 6 and 12 months of age.11,12 In contrast with infants who developed ID before 6 months of age, these infants were not born with particularly low BIE and ID was mostly mild, i.e., without anemia. Typical complementary foods are low in iron content and many infants do not regularly consume meats or iron-fortified foods that are the most reliable sources of iron for infants.24

Iron deficiency during the second year of life

During the second year of life the majority of infants have been weaned off the breast and those who continue to be breastfed consume a variety of other foods. Iron needs continue to be relatively large and are not always met by the typical toddler diet. That ID is relatively common in the second year of life is thus not too surprising and is well documented in the literature7,20,25–30; its prevalence has also not changed much in recent decades.29 In one of our own studies11 the prevalence of ID at 15 and 18 months of age was 12%, and in another study12 37% of toddlers between the ages of 12 and 24 months had ID. In neither study did we observe any IDA during the second year of life.

Iron supplementation and growth

In some studies a negative effect of iron supplementation on the growth of infants has been observed.31,32 The effect on growth was seen only in iron-replete infants and not in those with low or depleted iron stores. We similarly observed small but significant effects on growth.11,12 Although the effects on growth are small, they suggest that iron supplementation should be undertaken only with good reason and probably not in infants with replete iron stores.


Given the potential of severe ID to impair the neurocognitive development of infants,33–35 prevention of ID in breastfed infants is of considerable importance. To be useful, preventive measures must not only be effective, they should also be free of adverse effects and should be simple to administer. Our group has undertaken several studies that examined different means of iron supplementation of breastfed infants. In each of the studies, infants were breastfed exclusively until 4 months of age. After 4 months infants could receive complementary foods and, with the exception of the DryCereal study described below, could receive some formula. Breastfeeding continued up to 5.5 months in all infants; in many infants it continued much longer, even into the second year of life in some. In each study, infants visited the study center every 4 weeks and had capillary blood drawn at predetermined ages. In the first study (EarlyFe), medicinal iron was given between the ages of 1 and 5.5 months. In the second study (FeCereal) medicinal iron drops or a wet-pack cereal were given between 4 and 9 months of age. The third study (DryCereal) compared two sources of iron (electrolytic iron and ferrous fumarate) as fortificants of dry infant cereals, which were given between the ages of 4 and 9 months.

The EarlyFe study11 was a prospective double-blind study that was designed to answer the question of whether iron supplementation from an early age can augment iron stores, which at that age tend to be large, and whether such augmentation, if it occurred, was carried forward beyond the period of actual supplementation. In other words, could early iron supplementation also protect against iron deficiency later in the first year of life? Breastfed infants were enrolled at 1 month of age and randomly assigned to a daily supplement of 7 mg of iron in the form of ferrous sulfate, or to a placebo of similar appearance and taste. The intervention, from 1 to 5.5 months of age, was completed by 31 infants in the iron group and 32 infants in the placebo group. Infants were followed up to 18 months.

Figure 1 shows that plasma ferritin levels were significantly increased during the supplementation period, indicative of augmentation of iron stores. However, soon after the iron supplementation ceased, the difference in ferritin levels disappeared, indicating that augmentation of iron stores did not extend much beyond the period of actual supplementation. Plasma concentration of sTfR was significantly lower at 5.5 months in the iron group than in the placebo group, but there was no difference in hemoglobin concentration at any time. Small differences in relative distribution width and mean corpuscular volume were observed during, as well as after, the intervention period. In the placebo group, one infant developed IDA and one developed ID by 5.5 months of age, whereas in the Fe group no infant developed ID. After the intervention, one infant in the Fe group developed IDA at 9 months. Six subjects (among 52 followed to 18 months) developed ID at 15 and/or 18 months (prevalence 12%), but none developed IDA. Although weight gain among females was significantly slower in the Fe group, overall, there was no significant effect of iron supplementation on weight gain or length gain. Of note is that plasma ferritin concentrations tracked very strongly from 1 month all the way to 18 months.

Figure 1.

Plasma ferritin concentrations of breastfed infants receiving medicinal iron (Fe) or placebo from 1 to 5.5 months of age. Reproduced with permission from Ziegler et al.11

The FeCereal study12 was a randomized open-label study designed to examine, in comparative fashion, two regimens of providing iron to infants from 4 to 9 months of age. The primary aim of this study was to examine iron nutrition during the period when infants begin to exhaust their BIE and become dependent on exogenous sources of iron. Around the age of 4 months infants begin to receive complementary foods, but many of these foods are poor sources of iron.24 We hypothesized that the regular consumption of iron either in the form of medicinal iron or in the form of iron-fortified cereal would protect infants from ID and maintain iron status at a higher level than that of control infants lacking regular consumption of iron. Exclusively breastfed infants were randomized at 4 months to either medicinal iron (FeMed) in the amount of 7 mg per day, or to a wet-pack cereal (FeCer) that provided 7 mg of iron each day, or to a control group in which infants received complementary foods entirely at the parents’ discretion. Medicinal iron and the wet-pack cereal both provided iron from ferrous sulfate.

Iron status was improved by medicinal iron and wet-pack cereal to about the same degree. Plasma ferritin concentrations in both intervention groups were significantly higher than in the control group (Figure 2). As in the EarlyFe study, the effect on iron status was limited to the intervention period, with no effect on iron status soon after termination of the intervention. One infant developed IDA at the age of 2 months and was not randomized. In the control group, two infants developed IDA (by 5.5 months) and six infants developed ID. The 14.3% prevalence was significantly (P = 0.016) greater than in the combined FeMed and FeCer groups, for which it was 2.5%. At 12 months of age, 12 infants (9.6%) had ID. During the second year of life, 37% of subjects had ID on one, two, or three occasions, but none had IDA. The prevalence of ID was significantly (P < 0.001) higher in the control group than in the combined intervention groups, suggesting there was still some effect from the intervention. As noted in the earlier study, plasma ferritin tracked strongly from 1 month to 24 months. Growth (weight and length gain) was significantly lower (P = 0.027 for weight, P = 0.011 for length) in the FeMed group compared to the control group.

Figure 2.

Plasma ferritin concentrations of breastfed infants receiving medicinal iron (FeMed) or iron-fortified cereal (FeCer) from 4 to 9 months of age. Reproduced with permission from Ziegler et al.12

Our third study, the DryCereal study,13 was a randomized double-blind study designed to compare ferrous fumarate as fortification iron of a dry infant cereal with electrolytic iron, the widely used source of fortification iron. Because of its organoleptic properties, electrolytic iron is widely used to fortify infant cereals, but its bioavailability is considered to be poor.36,37 Ferrous fumarate, on the other hand, is more available and is suitable as a fortificant of cereals in that it does not cause discoloration or changes of taste.38 We therefore hypothesized that cereal fortified with ferrous fumarate would be more efficacious in maintaining iron status and preventing ID than cereal fortified with electrolytic iron. Exclusively breastfed infants were enrolled at age 2 months and were assigned at random to one of the two cereals, which were fed daily between the ages of 4 and 9 months. The amount of cereal fed was greater than the amount that free-living infants typically consume. Three infants developed IDA by 4 months of age; they were not randomized and were treated. Three additional infants developed ID at 4 months and these infants were randomized.

As Figure 3 shows, ferritin concentrations were quite similar during the intervention. The small persistent differences in favor of ferrous fumarate were not statistically significant. During the intervention, none of the study infants developed IDA, but four infants developed ID (two each with fumarate and electrolytic iron), including the two infants who already had ID at the time of randomization. The prevalence of ID during the intervention (4.2%) was lower than the prevalence observed in the control group of the FeCereal study (14.3%), suggesting that iron-fortified dry cereal is effective in providing relative protection of infants from ID. It was concluded that electrolytic iron and ferrous fumarate were equally effective as fortification sources of iron in infant cereals.

Figure 3.

Plasma ferritin concentrations of breastfed infants receiving dry cereal fortified with electrolytic iron or ferrous fumarate. None of the differences were statistically significant. Reproduced with permission from Ziegler et al.13

Iron deficiency during the first year of life

ID and IDA were observed during the first 6 months of life in each of our supplementation studies of breastfed infants. ID and IDA occurred mostly, though not exclusively, among infants who did not receive supplemental iron, i.e., infants in the placebo group of the EarlyFe study, infants in the control group of the FeCereal study, and infants in most studies before starting interventions at 4 months of age (N = 291). Data for infants who developed IDA (N = 8) are summarized in Table 1. The prevalence of IDA was 2.75%. Since the 10th centile for plasma ferritin (PF) was 124 µg/L at age 1 month and 64 µg/L at age 2 months, all but one infant with IDA had early PF below the 10th centile and, thus, were born with a low BIE. The sole exception was infant 9,905 who at 1 month had a PF of 222 µg/L. However, since this infant had an elevated C-reactive protein level of 3.2 mg/dL, an acute reaction may have raised the PF, thereby masking the presence of low iron stores. Thus, a low BIE seems to be a prerequisite for IDA in the first 6 months of life. All infants who developed IDA received treatment with iron. Non-anemic ID also occurred in the first 6 months of life (N = 8) and data are summarized in Table 1. The prevalence of ID without anemia was 2.75%. Contrary to infants who developed IDA, these infants did not all have low PF at 1 or 2 months of age. None of these infants received iron treatment and none went on to develop IDA later. During the second 6 months of life, we did not observe IDA, but non-anemic ID occurred in about 5.6% of infants. At 12 months, ID was present in 6.8% of infants.

Table 1. Occurrence of iron-deficiency anemia (IDA) and iron deficiency (ID) in breastfed infants during the first 6 months of life.
Group and subject no.PF at month 1 or 2 (µg/L)Age (months)PF (µg/L)Hb (g/L)
  • Siblings.

  • † 

    Elevated C-reactive protein.

  • ‡ 

    PF at 2 months of age.

  • Abbreviations: Hb, hemoglobin; PF, plasma ferritin.



Breastfed infants can develop ID, including IDA, which represents severe ID. In the three studies reviewed here, IDA occurred in 2.75% of breastfed infants in the first 6 months of life. All of these infants were born with a low birth iron endowment. An additional 2.75% of infants developed ID in the first 6 months of life. In the second 6 months of life, 5.6% of infants developed ID but no infant developed IDA. Medicinal iron prevented ID and IDA but had a small yet significant negative effect on growth. The feeding of iron-fortified cereals, both dry and wet-pack, almost completely prevented ID. Our data suggest it is possible to prevent IDA and almost all ID in breastfed infants.


Funding.  The studies reviewed here were supported by the National Institutes of Health (grant HD40315).

Declaration of interest.  The authors have no relevant interests to declare.