QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Online Supporting Material Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fomon, S. J. Articles by Ziegler, E. E. Search for Related Content PubMed PubMed Citation Articles by Fomon, S. J. Articles by Ziegler, E. E. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB IRON

---

Human Nutrition and Metabolism

Absorption and Loss of Iron in Toddlers Are Highly Correlated^1,2

Department of Pediatrics, University of Iowa, Iowa City, IA and ^* Department of Preventive Medicine and Community Health, University of Texas Medical Branch, Galveston, TX

³To whom correspondence should be addressed. E-mail: samfomon{at}aol.com.

	ABSTRACT

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
For estimating the requirements for dietary iron, it is important to know the amount of iron that is lost from the body. Inevitable losses of iron have been determined in adult humans but not in infants or children. We administered ⁵⁸Fe, the least abundant stable isotope of iron, to free-living infants at 168 d of age (5.6 mo) and followed them to age 26 mo. There was no dietary restriction after isotope administration. Blood was obtained at regular intervals for determination of isotopic enrichment and indices of iron status. We estimated the quantity of circulating iron, noncirculating active iron, and storage iron at each age. The administered isotope equilibrated with total body iron by 13 mo of age. From 13 to 26 mo of age, we estimated inevitable loss and absorption of iron from the change in tracer abundance in circulating iron. The rate of decrease of tracer abundance was proportional to addition of tracee, i.e., absorption of iron. Conversely, the rate of decrease in quantity of tracer was proportional to removal of tracee, i.e., loss of iron. From 13 to 26 mo of age, iron absorption was (mean ± SD) 0.49 ± 0.13 mg/d and inevitable iron loss was 0.25 ± 0.12 mg/d. Intersubject variability of iron loss and iron absorption was high, and iron loss and absorption were highly correlated (r = 0.789, P < 0.001). Iron stores were low throughout the study and decreased significantly from 13 to 26 mo of age, suggesting that iron absorption from the diet was inadequate to maintain or increase iron nutritional status. The data suggest that, in this cohort, which may be representative, the intake of bioavailable iron from 13 to 26 mo of age was insufficient to maintain iron nutritional status.

KEY WORDS: • iron absorption • iron loss • toddlers • stable isotope

For an individual to remain in good iron nutritional status, the amount of iron absorbed must at least equal the amount of iron lost. For the growing individual, it is necessary not only to absorb enough iron to replace inevitable iron loss but, in addition, to absorb the amount of iron needed for growth. The iron needed for growth can be estimated with an approach that has been applied to infants (1) and to adolescents after the peak of the adolescent growth spurt (2) and that should be applicable to any age. Loss of iron, on the other hand, has been measured only in adult men (3) and in adolescents of either gender beyond the peak of the growth spurt (2).

Loss of iron occurs mainly from the gastrointestinal tract but also from the skin and in secretions. It is generally thought that endogenous fecal iron (in contradistinction to exogenous, i.e., nonabsorbed dietary iron) represents iron that is lost from the body and that its excretion is not regulated, i.e., is inevitable. The inevitable iron loss by adult men was determined by Green et al. (3), who administered the long-lived radioisotope, ⁵⁵Fe, and then allowed the isotope to reach equilibration with total body iron. Thereafter, the fractional disappearance of the isotope from the circulation (where it is present almost entirely in hemoglobin) is the same as the fractional loss of iron from the body. Green et al. (3) measured the disappearance of circulating isotope, beginning 10 mo after isotope administration, over a period of several years and estimated the loss of iron by adult men to be 1 mg/d. Evidence supporting the assumption that, after sufficient time, an iron isotope becomes completely distributed throughout the body iron pool was provided by the study of an individual who had been given ⁵⁵Fe 36 mo previously and subjected to repeated phlebotomy (4).

In a study of adolescents after the peak of the adolescent growth spurt (2), we labeled the subjects with the least abundant stable isotope of iron, ⁵⁸Fe. After an equilibration period of at least 1.56 y, we measured the abundance of the isotope in circulating erythrocytes during an observation period of >3 y in each subject. From the fractional disappearance of ⁵⁸Fe, we estimated inevitable iron loss to be 0.70 mg/d by men and 0.84 mg/d by women. These losses represented 48% of the estimated requirement for absorbed iron by men and 73% of the estimated requirement for absorbed iron by women.

Little information is available regarding iron loss by younger subjects. The present report concerns a study in which we administered ⁵⁸Fe to infants at age 168 d (5.6 mo). The administered isotope became distributed within total body iron by 13 mo of age, and we were able to determine inevitable iron loss and iron absorption during the interval from 13 to 26 mo of age.

	SUBJECTS AND METHODS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
    Study design. Normal infants were enrolled by 154 d of age and were given ⁵⁸Fe orally within 4 d of 168 d of age. Blood was obtained just before isotope administration and within 4 d of each of the following ages: 182, 196, 224, 252, 280 d and 11, 13, 15, 17, 20, 23, and 26 mo. Intake of iron was restricted from d 154 until isotope administration but was unrestricted thereafter. The study protocol was reviewed and approved by the Institutional Review Board of the University of Iowa and a parent provided written informed consent.

    Subjects. The subjects were term infants of either gender (n = 35) with birth weight >2500 g; they were considered normal by their physicians and the investigators. By chance, all were Caucasian. Many had participated in other studies in our unit. Two infants were breast-fed at the time of enrollment, with 1 receiving small amounts of supplemental formula. Of the infants fed formula at the time of enrollment (n = 33), 5 had been breast-fed initially for periods of 2 wk to 2 mo. At enrollment, infants were given low-iron (1.8 mg/L) formula (Similac®, Ross Products Division). Beginning 2 wk after isotope administration, there were no restrictions regarding feeding or dietary supplements. Formula-fed infants were fed iron-fortified formula (12 mg Fe/L, label claim) for at least several months.

    Tracer (isotope) administration. The tracer, isotopically enriched ⁵⁸Fe, was obtained in elemental form from Cambridge Isotope Laboratory (⁵⁸Fe abundance 91–94 atom %). A precisely weighed quantity of isotopically enriched ⁵⁸Fe was converted to ferrous sulfate and made up to volume as previously described (5). Each infant was admitted to the Lora N. Thomas Pediatric Metabolic Unit within 4 d of age 168 d and given 2 doses of ⁵⁸Fe, each dose 3 h after a feeding and 1 h before the next feeding. Each dose consisted of 0.9 mg of iron and 0.8 mg of ⁵⁸Fe-enriched ferrous sulfate in 5 mL of a 5% glucose solution containing 10 mg of ascorbic acid. Each dose was delivered from a syringe directly into the back of the oral cavity to decrease the likelihood of regurgitation. The syringe was rinsed twice with a 5% glucose solution and the rinses were administered to the infant. Infants were observed in the metabolic unit for 1 h after the 2nd dose to detect regurgitation if it occurred. No regurgitation occurred.

    Blood collection, processing, and analysis. Capillary blood was collected by heel stick, using a disposable spring-loaded device (Tenderfoot, International Technidyne); 1 mL of blood was collected into a heparin-treated tube. Hemoglobin concentration (Hb)⁴ was measured immediately on whole blood using a Coulter AcT diff Hematology Analyzer (Coulter). The blood was then centrifuged at 2000 x g for 15 min to obtain plasma for determination of ferritin concentration (PF; by RIA using the Quantimune kit, Bio-Rad Laboratories). Plasma C-reactive protein (CRP) was quantified by a two-site ELISA developed in our laboratory, and following the general enzyme immunoassay protocol described by Flowers et al. (6). The capture antibody was sheep anti-human CRP (ICN Cappel Research), the primary detection antibody was rabbit anti-human CRP (EMD Biosciences Calbiochem), and the secondary detection antibody was alkaline phosphatase–conjugated goat anti-rabbit IgG (Sigma-Aldrich). Packed erythrocytes were saved for isotope ratio determination.

    Isotope analyses. Isotope ratios (IR_58/57) were determined in erythrocytes as previously described (7). Isotope ratios were determined on a PlasmaQuad 3 inductively coupled plasma MS (ICPMS; VG Elemental). The PlasmaQuad 3 ICPMS instrument was modified for measurement of iron isotope ratios as follows: 1) use of the PlasmaScreen option accompanied by a very low plasma radio frequency power setting; 2) optimization of argon flow rates for minimization of background intensities at m/z = 58 and 57 with simultaneous maximization of analyte intensity for cobalt at m/z = 59; 3) minimization of the nickel contribution to m/z = 58 by using copper rather than the customary nickel sampler and skimmer cones; 4) use of the Cetac U-6000AT ultrasonic nebulizer for sample introduction; 5) replacement of the original equipment discrete dynode electron multiplier detector and its vacuum chamber by a continuous dynode detector and housing that possesses very low electrical impedance, very high linearity of response, and minimal gain fatigue during exposure to high intensities. The ICPMS instrument is sequestered in a clean environment at constant temperature and humidity. In addition, the sample preparation procedures that are used minimize the contributions of Cr, CaOH, and Ni to m/z = 54, 57 and 58, respectively, by selective preconcentration of iron using ultrapure solvents and a chelating reagent highly specific for iron (III). Dilute hydrochloric acid, rather than the customary dilute nitric acid, is used for sample introduction to the ICPMS instrument to minimize formation of ArN⁺ in the plasma and its contribution to background at m/z = 54. Parameters for operation of the PlasmaQuad 3 ICPMS instrument in the laboratory of Dr. Serfass resulted in a worst-case precision of measurement for IR_58/54 of 0.5% relative SD (rsd). The "internal precision of measurement," the mean value achieved by 10 sequential measurements of the same sample, was 0.33rsd.

Data analysis

    Estimation of body iron compartments. Total body iron (Fe_tot) was conceptualized as consisting of 3 compartments, circulating iron (Fe_circ), noncirculating active iron (Fe_nca), and storage iron (Fe_stor). Thus, Fe_tot = Fe_circ + Fe_nca + Fe_stor, where all components are in mg (1 mg Fe = 17.9 µmol), and Fe_circ represents iron in circulating erythrocytes and ignores the small amounts of iron present in circulating iron-containing enzymes and the iron in circulating transport proteins (e.g., transferrin, ferritin). Because circulating non-Hb iron is such a small percentage of circulating iron, we contend that little error is introduced by ignoring it. It was calculated as Fe_circ = BV x Hb x 3.47, where BV is blood volume in L, assumed to be 0.65 L/kg body weight, Hb is hemoglobin concentration in g/L, and 3.47 is the concentration of iron (mg/g) in hemoglobin. Fe_nca consists of iron in myoglobin and in iron-containing enzymes. Fe_nca was assumed by analogy with normal adult men using representative values as suggested by Dallmann et al. (8), which resulted in a value of 455 mg. For a 75 kg adult, Fe_nca is therefore 6 mg/kg.

Storage iron (Fe_stor) consists of iron in ferritin and hemosiderin. In the absence of inflammation, PF is an index of Fe_stor (9–11). Storage iron was thus calculated from PF on the assumption that, after correction for body size, PF bears the same relation to storage iron in infants (excluding the early months of life) and toddlers as in adults. Although this assumption is untested, we showed in most of our studies of infants (12–16) and in a study of preadolescent children (17) that erythrocyte incorporation of administered ⁵⁸Fe is inversely correlated with PF. Thus, we estimated Fe_stor on the basis of analogy with adults using the equation describing the relation between PF and storage iron that Hallberg et al. (18) derived from the data of Skikne et al. (19). This equation predicts the absence of storage iron when PF is 15.9 µg/L.

For purposes of estimating Fe_stor, it was necessary to obtain PF values that were reflective of storage iron. Because ferritin is an acute-phase reactant, PF may not be reflective of iron nutritional status during inflammation or infection. We had planned to use CRP as an indication of the presence of an acute-phase reaction and to ignore values for PF obtained concurrently with elevated CRP. However, in analyzing a large amount of data from another study, it was evident that in some infants, PF was not elevated in the presence of elevated CRP and in others, PF was elevated when CRP was normal. We presume that these circumstances reflect either laboratory errors in one of the measurements or a lack of concurrence between the timing of increase and subsequent decrease of the 2 acute-phase reactants (CRP and PF). We chose to identify spuriously elevated PF without considering CRP. We established the rule that a PF value was classified as spurious if it was 50% higher than the mean of the preceding and the immediately following values. For the purpose of calculating Fe_stor, such spurious values were replaced by "adjusted" values (PF_adj), calculated as the mean of the adjacent values. In the case of a spurious value at age 26 mo, the 2 preceding values were used to obtain a corrected value.

Mean daily gain in total body iron (Fe_gain) during the period 13–26 mo was calculated as the difference in Fe_tot at the beginning and end of the period divided by the number of days in the period.

    Isotope data. The first step was calculation of tracer abundance (⁵⁸A) from measured isotope ratios. Tracer abundance was defined as concentration of tracer isotope in excess of natural abundance isotope. Thus, calculation of tracer abundance included correction for natural abundance of ⁵⁸Fe. Abundance at time t (⁵⁸A^t) was expressed in mass units (mg/g) and was calculated as

where ⁵⁸A^tr is the abundance (molar) of ⁵⁸Fe (tracer) in the enriched ⁵⁸Fe preparation, ⁵⁷A^N is the abundance (molar) of ⁵⁷Fe in natural iron (0.021191), IR^t_58/57 is the measured ⁵⁸Fe /⁵⁷Fe isotope ratio (molar) in the blood sample at time t, IR^N_58/57 is the natural ⁵⁸Fe /⁵⁷Fe isotope ratio (molar) (0.133032), ⁵⁷A^tr is the ⁵⁷Fe abundance (molar) in the enriched ⁵⁸Fe preparation, IR^tr_58/57 is the ⁵⁸Fe /⁵⁷Fe isotope ratio (molar) of the enriched ⁵⁸Fe preparation, 57.933 is the molar mass of ⁵⁸Fe, and 55.845 is the molar mass of natural Fe.

The amount of tracer at time t (⁵⁸Fe^t_tot) was calculated as

where Fe^t_tot is Fe_tot at time t.

Loss and absorption of iron between 13 and 26 mo of age

    Principle. After a tracer has equilibrated with the tracee pool, a decrease (dilution) of tracer concentration (abundance) can occur only through addition of tracee. The decrease in tracer abundance is therefore proportional to the addition of tracee (by iron absorption). Conversely, a decrease in the amount of tracer (= abundance x pool size) can occur only if the tracer is removed, which can occur only in association with tracee. The decrease in the amount of tracer is therefore proportional to the amount of tracee removed (iron loss).

    Calculations. The linear regression of log⁵⁸A^t against time was calculated for each subject for the period 13–26 mo. The slope (k_abs) of the regression reflects the fractional absorption of iron, expressed as fraction of total body iron per unit of time. The mean quantity of iron absorbed each day (mg/d) between 13 and 26 mo (Fe_abs) was calculated as

where _tot is mean total body iron between 13 and 26 mo, calculated as amount at the midpoint (19.5 mo) of the linear regression of Fe_tot against time.

The linear regression of log ⁵⁸Fe^t_tot against time was calculated for each subject for the period 13 to 26 mo. The slope (k_loss) of the regression reflects the fractional loss of iron, expressed as fraction of total body iron per unit of time. The mean amount of iron lost each day (mg/d) between 13 and 26 mo (Fe_loss) was calculated as

    Initial erythrocyte incorporation. Erythrocyte incorporation of tracer (⁵⁸Fe_inc), expressed as a percentage of the dose, was determined from tracer abundance determined 28 d after isotope administration using calculations described previously (7).

    Statistical analysis. Data were analyzed using SAS version 6.12 (SAS Institute). Changes in Fe status were compared by paired t tests, and differences were considered significant with = 0.05. Slopes representing fractional change in log⁵⁸A and log⁵⁸Fe_tot were determined using least-squares regression. Slopes of the regression of isotope abundance on age were assessed for gender effects and homogeneity using mixed-model ANOVA (PROC MIXED). Values presented are means ± SD.

	RESULTS

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
Of the 35 infants administered ⁵⁸Fe at 168 d of age, 5 subjects did not complete the study as planned because the parents found it difficult to keep the scheduled appointments. Three of these subjects were withdrawn before 13 mo of age, but Subjects 7814 and 8106 were not withdrawn until after the 23-mo visit and their data are included. Unless specified otherwise, data for the last-mentioned 2 subjects are included with the 26-mo data for all other subjects. In 1 subject who had the lowest initial erythrocyte incorporation of ⁵⁸Fe (1.18%), tracer abundance was below the detection limit after 17 mo of age, and data for that subject are not included. Thus, results presented here concern 31 subjects, 15 boys and 16 girls, who completed the study to at least 23 mo of age. Between 13 and 26 mo of age, 6 blood samples were available from 17 subjects, 5 samples from 11 subjects, and 4 samples from 3 subjects. Six subjects had PF (unadjusted) <10 µg/L at one or more time points between 13 and 26 mo of age and thus were iron deficient by the conventional classification.

    Body iron compartments. Hemoglobin concentration changed little throughout the period of observation (Table 1). Final Hb values (data not shown) at age 23 or 26 mo were <110 g/L in only 2 of the 31 subjects. As was anticipated, PF_adj decreased from age 168 d (5.6 mo) to age 13 mo (P < 0.001), and there was a further decrease from age 13 to 26 mo (P = 0.027). Iron stores (Fe_stor), whether expressed in milligrams or as a percentage of Fe_tot, also decreased with increasing age (Table 1). Iron stores were depleted (defined as Fe_stor = 0, corresponding to PF 15.9 µg/L) in 9 subjects at 13 mo of age and in 12 subjects at 26 mo. Six subjects were depleted at both ages. The 12 subjects with depleted iron stores at 26 mo included 4 of the subjects who had PF < 10 µg/L at some time between 13 and 26 mo. At age 168 d, Fe_circ accounted for 74.6% of Fe_tot, Fe_nca for 16.0% and Fe_stor for 9.4%. By age 26 mo, Fe_circ had increased to 79.6% and Fe_stor had decreased to 3.1% of Fe_tot.

View larger version (11K): [in this window] [in a new window]   FIGURE 1 Time course of enrichment of circulating iron (log58At) plotted against age in a representative subject (#7818).

View larger version (26K): [in this window] [in a new window]   FIGURE 2 Fitted values for regressions of log58Fettot on age of individual infants.

View larger version (11K): [in this window] [in a new window]   FIGURE 3 Regression of absorbed iron (Feabs, mg/d) on inevitable iron loss (Feloss, mg/d) (r = 0.789, P < 0.001, slope = 0.824).

    Initial erythrocyte incorporation of iron. Erythrocyte incorporation of tracer (⁵⁸Fe_inc) 28 d after administration was 14.6% of the dose with a range from 1.5 to 41.1%. Erythrocyte incorporation of tracer (⁵⁸Fe_inc) was inversely correlated at a borderline level of significance with (unadjusted) PF at 168 d of age (r = –0.511, P = 0.051). Erythrocyte incorporation of tracer was not significantly correlated with Fe_loss (r = 0.354, P = 0.051) nor with Fe_abs (r = 0.160, P = 0.341).

	DISCUSSION

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 
We administered ⁵⁸Fe, the least abundant stable isotope of iron, to infants at 168 d of age and obtained blood samples at regular intervals until the subjects were 26 mo old. The administered isotope had equilibrated with Fe_tot by 13 mo of age. Thus, from 13 to 26 mo of age, ⁵⁸Fe was a valid tracer for Fe_tot; during this age interval, we were able to determine endogenous loss of iron and absorption of iron by following the ⁵⁸Fe abundance in circulating erythrocytes. This is the first determination of iron loss and absorption in toddlers using isotopes.

    Iron loss from 13 to 26 mo. We determined that iron loss as a fraction of Fe_tot was 22.7 ± 10.5%/y. In absolute terms, iron loss was 0.25 ± 0.12 mg/d. Moreover, iron loss had very high intersubject differences, with an almost 8-fold range of values (0.07–0.55 mg/d).² Because the mixed-model analysis excluded the possibility that the values were derived from a single population, the variability could not be attributed to random errors. If iron loss is indeed "inevitable," the observed variability suggests that iron loss is mediated by more than one mechanism. Alternatively, if there is only one mechanism involved, the results indicate that it is operating at different levels of efficiency, possibly the result of genetic variation. It is, of course, also possible that rather than being inevitable, iron loss is regulated or that it is influenced by other factors. One such factor might be the amount of iron absorbed. The high degree of correlation between and Fe_loss and Fe_abs (Fig. 3) is consistent with this possibility.

Endogenous iron loss was determined in adults (3) and in adolescents (2) using methods comparable to that used in the present study, i.e., after equilibration of isotopic tracers with total body iron. With body weight slightly <12 kg for the interval 13 to 26 mo of age, mean iron loss by toddlers in the present study was 0.022 mg/(kg · d). Per unit of body weight, this was somewhat more than the loss by men and adolescents. Assuming that the mean weight of the men studied by Green et al. (3) was 70 kg, the loss of 1 mg/d was 0.014 mg/(kg · d). Adolescent boys weighing 56 kg (2) lost 0.70 mg/d, which is also 0.014 mg/(kg · d). Because loss of iron from the gastrointestinal tract makes the largest contribution to iron loss, the age-related difference in iron loss is probably explained by greater gastrointestinal loss per unit of body weight by toddlers than by older subjects.

There are no previous reports on endogenous iron loss by toddlers and only 1 report on iron loss by infants. Belsten et al. (20) administered ⁵⁸Fe intravenously to infants at 6, 9, and 12 mo of age and determined fecal excretion of the isotope 7–14 d after administration. Endogenous fecal iron was assumed to derive from an "exchangeable" iron pool, whose size was estimated by 2 different methods. At 12 mo of age, endogenous iron loss was 3 µg/(kg · d) by one method and 8 µg/(kg · d) by the other. However, because the isotope could not have equilibrated with total body iron and it is not known whether it had equilibrated with the exchangeable iron pool, the results are difficult to evaluate.

Most fecal iron derives from desquamated enterocytes, which contain iron predominantly in the form of ferritin. Ferritin is indeed present in feces, but it is not known what percentage of endogenous iron is accounted for by ferritin iron. Feces contain small amounts of heme, mostly in the form of hemoglobin, which accounts for some of the fecal iron. We demonstrated (21) that in the majority of infants, especially older infants, fecal iron loss in the form of hemoglobin amounts to 2–4 µg/(kg · d). With cow’s milk feeding, fecal blood loss can increase considerably. The greatest loss of hemoglobin iron induced by cow’s milk that we observed was 250 µg/(kg · d), although in most infants with unequivocal cow’s milk–induced blood loss, fecal loss of hemoglobin iron was much less. Nevertheless, these observations make it appear possible that fecal blood loss could explain some of the variability of fecal iron loss by toddlers. Whether this is indeed the case cannot be verified because there are no reports of fecal blood loss by toddlers.

    Iron absorption from 13 to 26 mo. Iron absorption as a fraction of Fe_tot was 44.0 ± 10.2%/y. In absolute terms, iron absorption (Fe_abs) was 0.49 ± 0.13 mg/d. As with iron loss, iron absorption had high intersubject variability, with a 3-fold range of values (0.24–0.79 mg/d).² The agreement between iron absorption and the sum of loss and gain of iron was good, suggesting that our estimates of loss and absorption were reasonable. Subjects with greater Fe_loss had greater Fe_abs (Fig. 3). One interpretation of this relation is that greater loss of iron somehow stimulates greater iron absorption. This is the more plausible explanation for the relation than the alternative explanation mentioned earlier. It is consistent with the well-established inverse relation between PF and iron absorption, which was also confirmed in the present study. Although this relation is well established, nothing is known about the mechanism(s) involved.

    Prompt erythrocyte incorporation of administered isotope. Incorporation of an administered isotope into erythrocytes has been used extensively as a surrogate of iron absorption even in infants, although we showed (16) that in infants, <80% of newly absorbed iron is promptly (within 4 wk) incorporated into erythrocytes. In the present study, incorporation 28 d after isotope administration was 14.6% of the dose.

Erythrocyte incorporation of administered isotope was borderline correlated with iron loss between 13 and 26 mo of age, an observation suggesting that indices of iron metabolism determined in infancy may be predictive of indices during y 2 of life. If this is confirmed in future studies, it would suggest that there are intersubject differences in iron metabolism among "normal" subjects, i.e., subjects without iron overload. It is, of course, well appreciated that mutations in a number of proteins involved in iron metabolism can lead to iron overload. Our observations suggest that there are identifiable differences in iron metabolism among subjects who remain in iron balance. Whether these differences have clinical consequences is not known; however, it is reasonable to speculate that differences in inevitable iron loss may affect an individual’s risk of developing iron deficiency.

    Adequacy of iron absorption. During y 1 of life, Fe_stor decreases as Fe_circ increases. Because blood volume continues to expand during y 2 of life, Fe_circ continues to increase, but little is known about changes in Fe_stor. In the cohort of toddlers in the present study, there was a slight but significant decrease in Fe_stor from age 13 to 26 mo. At 13 mo of age, 9 of 31 subjects were iron deficient (Fe_stor = 0) and by 23 or 26 mo of age, 12 of 31 subjects were iron deficient. Thus, it is evident that iron absorption during y 2 of life lagged behind the needs for growth and replacement of losses.

What we observed in our study may reflect findings in toddlers generally. Iron deficiency is acknowledged to be prevalent during y 2 of life (22) as observed in the present study. The estimated iron intake by toddlers of 10 mg/d (23,24) appears to be insufficient to maintain iron stores. The quantity of iron consumed is, perhaps, less important than its rather low bioavailability from the toddler diet. The majority of toddlers consume some meat, principally chicken and turkey (25), but the amounts are unknown and it seems unlikely that heme iron makes a major contribution to total iron intake. The sources of the nonheme iron are not well identified, but a considerable portion of the dietary iron may come from iron-fortified "noninfant" cereals, which most toddlers consume at least once daily (25). The iron fortificant of these cereals is "reduced iron" (26). Reduced iron is a member of the family of elemental iron powders which, as used in food fortification, are thought to be of low bioavailability (27–33). Perhaps the most important factor affecting iron absorption by toddlers is their widespread consumption of cow’s milk (24,34). Both calcium (35–37) and cow’s milk proteins (38) are potent inhibitors of iron absorption.

	ACKNOWLEDGMENTS

 
Assistance with the statistical analysis by Hannes Ledolter, Ph.D., is gratefully acknowledged. Janice M. Jeter, RN, and her staff recruited subjects, administered isotope, drew blood, and carried out all other clinical aspects of the study.

	FOOTNOTES

 
¹ Supported by U.S. Public Health Service grants HD 32938 and HD40315.

² Individual values for plasma ferritin concentration at 168d, erythrocyte incorporation of ⁵⁸Fe at 168d, and iron loss, iron absorption and iron gain from 13 to 26 mo are available in Supplemental Table 1 with the online posting of this paper at www.nutrition.org.

⁴ Abbreviations used: ⁵⁸A, mass abundance of ⁵⁸Fe in excess of natural abundance (unless otherwise stated as molar abundance ⁵⁷A^tr, ⁵⁸A^tr, ⁵⁷A^N); CRP, C-reactive protein; Fe_circ, mass Fe in circulation; Fe_gain, mass gain in Fe; ⁵⁸Fe_inc, prompt (within 4 wk) erythrocyte incorporation of ⁵⁸Fe; Fe_loss, mass loss of Fe; Fe_nca, mass noncirculating active Fe; Fe_stor, mass storage Fe; Fe_tot, mass total body iron; ⁵⁸Fe_tot, mass ⁵⁸Fe in excess of natural abundance; Hb, hemoglobin; ICPMS, inductively coupled plasma MS; IR_58/57, molar isotope ratio ⁵⁸Fe/⁵⁷Fe; PF, plasma ferritin; rsd, relative SD.

Manuscript received 23 August 2004. Initial review completed 30 September 2004. Revision accepted 27 December 2004.

	LITERATURE CITED

TOP ABSTRACT SUBJECTS AND METHODS RESULTS DISCUSSION LITERATURE CITED

 

1. Fomon, S. J. (1993) Nutrition of Normal Infants 1993:239-260 Mosby St. Louis, MO.

2. Fomon, S. J., Drulis, J. M., Nelson, S. E., Serfass, R. E., Woodhead, J. C. & Ziegler, E. E. (2003) Inevitable iron loss by human adolescents, with calculations of the requirement for absorbed iron. J. Nutr. 133:167-172.[Abstract/Free Full Text]

3. Green, R., Charlton, R., Seftel, H., Bothwell, T., Mayet, F., Adams, B., Finch, C. & Layrisse, M. (1968) Body iron excretion in man. A collaborative study. Am. J. Med. 45:336-353.[Medline]

4. Charlton, R. W., Fatti, L. P., Lynch, S. R., Torrance, J. D. & Bothwell, T. H. (1980) Equilibration of tracer radioiron with body iron. Clin. Sci. (Lond.) 58:93-100.[Medline]

5. Janghorbani, M., Ting, B.T.G. & Fomon, S. J. (1986) Erythrocyte incorporation of ingested stable isotope of iron (⁵⁸Fe). Am. J. Hematol. 21:277-288.[Medline]

6. Flowers, C. A., Kuizon, M., Beard, J. L., Skikne, B. S., Covell, A. M. & Cook, J. D. (1986) A serum ferritin assay for prevalence studies of iron deficiency. Am. J. Hematol. 23:141-151.[Medline]

7. Fomon, S. J., Ziegler, E. E., Nelson, S. E, Serfass, R. E. & Frantz, J,. A. (1995) Erythrocyte incorporation of iron by 56-day-old infants fed a ⁵⁸Fe-labeled supplement. Pediatr. Res. 38:373-378.[Medline]

8. Dallmann, P. R., Yip, R. & Oski, F. A. (1993) Iron deficiency and related nutritional disorders. Nathan, D. G. Oski, F. A. eds. Hematology of Infancy and Childhood 4th ed. 1993:413-450 W. B. Saunders Philadelphia, PA. .

9. Jacobs, A., Miller, F., Worwood, M., Beamish, M. R. & Wardrop, C. A. (1972) Ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. Br. Med. J. 4:206-208.

10. Walters, G. O., Miller, F. M. & Worwood, M. (1973) Serum ferritin concentration and iron stores in normal subjects. J. Clin. Pathol. 26:770-772.[Abstract/Free Full Text]

11. Cook, J. D., Lipschitz, D. A., Miles, L.E.M. & Finch, C. A. (1974) Serum ferritin as a measure of iron stores in normal subjects. Am. J. Clin. Nutr. 27:681-687.[Abstract]

12. Fomon, S. J., Janghorbani, M., Ting, B.T.G., Ziegler, E. E., Rogers, R. R., Nelson, S. E., Ostedgaard, L. S. & Edwards, B. B. (1988) Erythrocyte incorporation of ingested 58-iron by infants. Pediatr. Res. 24:20-24.[Medline]

13. Fomon, S. J., Ziegler, E. E., Rogers, R. R., Nelson, S. E., Edwards, B. B., Guy, D. G., Erve, J. C. & Janghorbani, M. (1989) Iron absorption from infant foods. Pediatr. Res. 26:250-254.[Medline]

14. Fomon, S. J., Ziegler, E. E. & Nelson, S. E. (1993) Erythrocyte incorporation of ingested ⁵⁸Fe by 56-day-old breast-fed and formula-fed infants. Pediatr. Res. 33:573-576.[Medline]

15. Fomon, S. J., Ziegler, E. E., Serfass, R. E., Nelson, S. E. & Frantz, J. A. (1997) Erythrocyte incorporation of iron is similar in infants fed formulas fortified with 12 mg/L or 8 mg/L of iron. J. Nutr. 127:83-88.[Abstract/Free Full Text]

16. Fomon, S. J., Serfass, R. E., Nelson, S. E., Rogers, R. R. & Frantz, J. A. (2000) Time course of and effect of dietary iron level on iron incorporation into erythrocytes by infants. J. Nutr. 130:541-545.[Abstract/Free Full Text]

17. Woodhead, J. C., Drulis, J. M., Nelson, S. E., Janghorbani, M. & Fomon, S. J. (1991) Gender-related differences in iron absorption by preadolescent children. Pediatr. Res. 29:435-439.[Medline]

18. Hallberg, L., Hulthén, L. & Garby, L. (1998) Iron stores in man in relation to diet and iron requirements. Eur. J. Clin. Nutr. 52:623-631.[Medline]

19. Skikne, B. S., Flowers, C. H. & Cook, J. D. (1990) Serum transferrin receptor: a quantitative measure of tissue iron deficiency. Blood 75:1870-1876.[Abstract/Free Full Text]

20. Belsten, J. L., Phillips, C. E., Fairweather-Tait, S. J., Eagles, J. & Aggett, P. J. (1997) Endogenous iron losses in 6–12 month old infants. Fischer, P.W.F. L’Abbé, M. R Cockell, K. A. Gibson, R. S. eds. Trace Elements in Man and Animals 1997:509-511 NRC Research Press Ottawa, Canada. .

21. Ziegler, E. E., Fomon, S. J., Nelson, S. E., Rebouche, C. J., Edwards, B. B., Rogers, R. R. & Lehman, L. J. (1990) Cow milk feeding in infancy: further observations on blood loss from the gastrointestinal tract. J. Pediatr. 116:11-18.[Medline]

22. Kleinman, R. E. eds. Pediatric Nutrition Handbook 5th ed. 2004:308 American Academy of Pediatrics Elk Grove Village, IL. .

23. Devaney, B., Ziegler, P., Pac, S., Karwe, V. & Barr, S. I. (2004) Nutrient intakes of infants and toddlers. J. Am. Diet Assoc. 104(suppl.):S14-S21.

24. Briefel, R. R., Reidy, K., Karwe, V., Jankowski, L. & Hendricks, K. (2004) Toddlers’ transition to table foods: impact on nutrient intakes and food patterns. J. Am. Diet Assoc. 104(suppl.):S38-S44.

25. Fox, M. K., Pac, S., Devaney, B. & Jankowski, L. (2004) Feeding infants and toddlers study: what foods are infants and toddlers eating?. J. Am. Diet Assoc. 104(suppl.):S22-S30.

26. General Mills, Consumer Service Information (2004) Kellogg label claim 2004 http://www.kelloggs.com. [accessed 2004].

27. Elwood, P. C., Newton, D., Eakins, J. D. & Brown, D. A. (1968) Absorption of iron from bread. Am. J. Clin. Nutr. 21:1162-1169.[Abstract]

28. Cook, J. D., Minnich, V., Moore, C. V., Rasmussen, A., Bradley, W. B. & Finch, C. A. (1973) Absorption of fortification iron in bread. Am. J. Clin. Nutr. 26:861-872.[Abstract]

29. Hurrell, R. F. (1984) Bioavailability of different iron compounds used to fortify formulas and cereals: technological problems. Stekel, A. eds. Iron Nutrition in Infancy and Childhood 1984:147-178 Raven Press New York, NY. .

30. Hurrell, R. F., Furniss, D. E., Burri, J., Whittaker, P., Lynch, S. R. & Cook, J. D. (1989) Iron fortification of infant cereals: a proposal for the use of ferrous fumarate or ferrous succinate. Am. J. Clin. Nutr. 49:1274-1282.[Abstract/Free Full Text]

31. Hurrell, R. F. (1992) Prospects for improving the iron fortification of foods. Fomon, S. J. Zlotkin, S. eds. Nutritional Anemias 1992:193-208 Raven Press New York, NY. .

32. Hallberg, L., Brune, M. & Rossander, L. (1986) Low bioavailability of carbonyl iron in man: studies on iron fortification of wheat flour. Am. J. Clin. Nutr. 43:59-67.[Abstract/Free Full Text]

33. Fomon, S. J. (1987) Bioavailability of supplemental iron in commercially prepared dry infant cereals. J. Pediatr. 110:660-661.[Medline]

34. Skinner, J. D., Ziegler, P., Pac, S. & Devaney, B. (2004) Meal and snack patterns of infants and toddlers. J. Am. Diet. Assoc. 104(suppl.):S65-S70.

35. Dawson-Hughes, B., Seligson, F. H. & Hughes, V. A. (1986) Effects of calcium carbonate and hydroxyapatite on zinc and iron retention in postmenopausal women. Am. J. Clin. Nutr. 44:83-88.[Abstract/Free Full Text]

36. Hallberg, L., Brune, M., Erlandsson, M., Sandberg, A.-S. & Rossander-Hultén, L. (1991) Calcium: effect of different amounts on nonheme- and heme-iron absorption in humans. Am. J. Clin. Nutr. 53:112-119.[Abstract/Free Full Text]

37. Hallberg, L., Rossander-Hultén, L., Brune, M. & Gleerup, A. (1992) Bioavailability in man of iron in human milk and cow’s milk in relation to their calcium contents. Pediatr. Res. 31:524-527.[Medline]

38. Hurrell, R. F., Lynch, S. R., Trinidad, T. P., Dassenko, S. A. & Cook, J. D. (1989) Iron absorption in humans as influenced by bovine milk proteins. Am. J. Clin. Nutr. 49:546-552.[Abstract/Free Full Text]

This article has been cited by other articles:

				M. F. Lynch, I. J. Griffin, K. M. Hawthorne, Z. Chen, M. G. Hamzo, and S. A. Abrams Iron Absorption Is More Closely Related to Iron Status Than to Daily Iron Intake in 12- to 48-Mo-Old Children J. Nutr., January 1, 2007; 137(1): 88 - 92. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Online Supporting Material Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fomon, S. J. Articles by Ziegler, E. E. Search for Related Content PubMed PubMed Citation Articles by Fomon, S. J. Articles by Ziegler, E. E. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB IRON