The Journal of Nutrition Vol. 129 No. 1 January 1999, pp. 181-187

⁵⁹Fe Is Retained from an Elemental ⁵⁹Fe Powder Supplement without Effects on ⁶⁵Zinc, ⁴⁷Calcium and ⁶⁷Copper in Young Pigs^1,2,3,4

Kurt R. Zinn⁵, Tandra R. Chaudhuri, James M. Mountz, Gerrit J. van den Berg, Dennis T. Gordon^**, and Gary L. Johanning^*

Department of Radiology and ^* Department of Nutrition Science, University of Alabama at Birmingham Medical Center, Birmingham, AL 35294;  Department of Radiochemistry, Interfaculty Reactor Institute, Delft University of Technology, 2629 JB Delft, The Netherlands; and ^** Department of Cereal Sciences, North Dakota State University, Fargo, ND 58105-5728

	ABSTRACT

Abstract Introduction Methods Results Discussion References

In vivo counting with the use of a germanium detector evaluated the retention of an elemental ⁵⁹Fe powder supplement while measuring potential interactions with zinc, calcium and copper. Effects of dietary iron and zinc on in vivo retentions of ⁵⁹Fe, ⁶⁵Zn, ⁶⁷Cu and ⁴⁷Ca were studied in young pigs. In Experiment 1, 4-d-old piglets fed a cereal-based diet were randomly assigned to one of four treatment groups (2 × 2 factorial arrangement, n = 5 per group). Variables were dietary iron source (either elemental iron or FeSO₄, each at 100 mg iron/kg diet) and the dosage form of radioactive iron (either elemental ⁵⁹Fe powder or ⁵⁹FeSO₄). Experiment 2 (2 × 3 factorial arrangement) was performed using two levels of iron (100 and 200 mg/kg, as elemental iron) and three levels of zinc (25, 50 and 100 mg/kg). Piglets were also dosed with ⁴⁷Ca, ⁶⁵Zn and ⁶⁷Cu; all radioisotopes were measured for 8 d. Apparent absorption of elemental ⁵⁹Fe powder was 13 ± 1%, whereas ⁵⁹Fe sulfate was significantly (P < 0.05) higher at 26 ± 1%. The FeSO₄ diet decreased ⁶⁵Zn retention in Experiment 1, in contrast to the elemental iron diet, which did not have this effect in either experiment. Apparent ⁶⁵Zn absorption averaged 44 ± 2, 35 ± 1 and 27 ± 2% for the three levels of zinc (25, 50 and 100 mg/kg), respectively. Retention of ⁴⁷Ca was not affected by dietary iron or zinc; retention of ⁶⁷Cu was not affected by dietary iron. The data demonstrate good bioavailability of elemental iron without effects on zinc, copper and calcium.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Iron deficiency is a common nutritional problem in infants (Ziegler and Fomon 1996), and weaning cereals are routinely fortified with iron to address this problem. Elemental iron is the most common chemical form of iron added for this purpose in ready-to-eat breakfast cereals (Anderson 1985). Elemental iron fortification of cereals has been shown to prevent symptoms associated with iron deficiency, including anemia (Walter et al. 1993) and therefore remains an important and reliable way of supplementing infants with iron. However, the reported bioavailability of elemental iron has been variable, ranging from 10 to 90% of the bioavailability of iron sulfate (Patrick 1985). In addition, the iron present in iron-fortified cereals remains controversial because there may be a competitive reduction in absorption of other essential trace elements such as zinc and copper. The possible reduction in absorption depends not only on body stores of iron and the chemical form of this nutrient, but also on the molar ratio of iron with respect to the other elements. These interactions may be most important for infants because zinc is required for their proper growth (Walravens et al. 1989), and a negative effect of iron on zinc absorption might limit the supply of zinc to the body during this critical period of development.

One experiment showed a competitive effect of iron on zinc absorption when 25:1 molar ratios of iron to zinc were ingested as inorganic aqueous solutions (Solomons and Jacob 1981). However, when iron was present at normal fortification levels and zinc was present as a natural matrix component of foods, the inhibitory effect of iron on zinc absorption was not observed (Sandström et al. 1985, Solomons and Jacob 1981, Valberg et al. 1984). A recent review of studies involving iron and zinc absorption interactions discussed these conflicting results (Fairweather-Tait 1995).

A good model for the human infant is the newborn pig (Bustad and McClellan 1966, Newport and Henschel 1984, Pond and Houpt 1978). Because piglets are born without iron stores, this newborn more closely approximates the iron status of an infant that is 6-9 mo old (Blood et al. 1979). Use of the newborn piglet model allows the application of radioisotope techniques to study trace-element interactions. Advances in -ray detectors and analysis have made it possible to measure multiple, -emitting radioisotopes simultaneously. This technology is possible using -ray spectroscopy with high energy resolution germanium detectors that are capable of resolving -ray peaks that differ in energy by a few kiloelectronvolts. Application of this technique for measurements of in vivo retentions of -emitting, mineral radiotracers has been previously reported (Gordon et al. 1993, Zinn and Morris 1988, Zinn et al. 1995, 1997). Another application of the technique was to study the bioavailability of zinc, iron, manganese and selenium in liquid formulas ingested by 4-d-old pigs using simultaneously administered ⁶⁵Zn, ⁵⁹Fe, ⁵⁴Mn and ⁷⁵Se, with fecal measurements made by a germanium detector (Atkinson et al. 1993).

The purpose of this study was to apply the technique of high resolution -ray spectroscopy to compare iron supplements and their potential interactions on copper, zinc and calcium in newborn pigs. This was accomplished by simultaneously and repeatedly measuring the in vivo retentions of ⁵⁹Fe-labeled iron supplements and ⁶⁷Cu, ⁶⁵Zn and ⁴⁷Ca with which the piglets were dosed. Unique in this study was the intrinsic labeling of elemental iron powder, which is the most common method to iron-fortify infant cereals. By administering multiple -emitting radioisotopes in the same animals, it was possible to test whether dietary levels of iron and zinc influenced the retentions of ⁵⁹Fe, ⁶⁷Cu, ⁶⁵Zn and ⁴⁷Ca. Rice meal cereal was the major component of the test diets in these experiments. The levels of iron in the test diets were above the recommended requirement for pigs (NRC 1988) and included a range of iron:zinc molar ratios between 1.1 and 9.8. For comparison, the amount of diets fed to the pigs was equivalent to an infant consuming 3-6 daily servings (45-60 g) of completely iron-fortified (elemental electrolytic form) rice meal cereal (Gerber Products, Fremont, MI).

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Experimental design and diets.  Experiment 1 was a 2 × 2 factorial arrangement of treatments (n = 5 per group) that tested two variables as follows: 1) the dietary iron source, either elemental iron powder or FeSO₄, and 2) the radioactive dosage form of iron, either ⁵⁹Fe in the form of elemental iron powder or iron sulfate (FeSO₄). For the latter, each dose contained 0.30 MBq ⁵⁹Fe mixed together with ⁴⁷Ca (0.21 MBq), ⁶⁵Zn (0.24 MBq) and ⁶⁷Cu (0.33 MBq). The piglets were dosed twice, 8 d apart. After dosing, -ray emissions from the radiotracers were measured simultaneously at 6 and 19.5 h, and 2.8, 4.8, 6.8 and 7.8 d. Only the retention data for the first dosing are included in this report.

Diets for the first experiment are summarized in Table 1 and differed only in the chemical form of iron (at 100 mg/kg of diet). The major ingredient of the diet was rice meal cereal, provided by Gerber Products. Two types of rice meal cereal were used. One was the commercial, iron-fortified product as packaged for sale; the other was the same cereal produced without the iron fortification. Both rice meals were analyzed extensively before use to establish that there were no differences in elemental or nutritional composition other than iron concentration.

The second experiment was a 2 × 3 factorial arrangement of treatment groups, using two levels of iron (100 and 200 mg/kg) and three levels of zinc (25, 50, and 100 mg/kg). The diet compositions are summarized in Table 1. Each diet had the same chemical form of iron (elemental iron) and zinc (zinc carbonate). All of the iron in the second experiment was provided by the iron-fortified rice meal cereal. In the second experiment, all piglets were dosed one time with the same ⁵⁹Fe-powder diet slurry (0.25 MBq/dose), containing ⁶⁵Zn (0.27 MBq) and ⁴⁷Ca (0.31 MBq). In vivo retentions of radioisotopes were measured at 4.1, 4.8, 5.9 and 6.7 d after dosing.

Diets containing 100 and 200 mg of elemental iron/kg diet will be designated Fe 100 Diet and Fe 200 Diet, respectively. Similarly, diets containing 25, 50 and 100 mg zinc/kg diet will be designated Zn 25 Diet, Zn 50 Diet and Zn 100 Diet, respectively. The diet containing iron sulfate as the main source of iron (diet B in Table 1) will be designated FeSO₄ Diet.

The elemental iron powder (A-131, OMG Americas, Research Triangle Park, NC; formerly SCM Metal Products, SCM Chemicals, Cleveland, OH) used in these experiments is also commonly referred to as electrolytic iron because it is produced by electrolytic deposition, followed by mechanical grinding. It met all requirements for electrolytic iron by the U.S. Food and Drug Administration, as specified by reference to the NRC (1981). The powder is irregular in shape and the fernlike structure results in high surface area. Subsieve analysis showed the following range of particle sizes: 30-44 µm (7.8%), 20-30 µm (15.7%), 10-20 µm (34.8%) and <10 µm (39.5%).

Statistical comparisons of treatment groups were made using two-way ANOVA (General Linear Model, Statistical Analysis System, SAS, Release 6.11, SAS Institute, Cary, NC). When the F-test was significant (P < 0.05) the means were compared using least-square estimates of marginal means (LSMEANS statement in SAS). Linear regression analyses of retention data (percentage retained vs. time) were also performed using SAS. Values are means ± sem.

Animals.  Newborn pigs were used for both experiments. Male piglets, a Dekalb hybrid breed (specific pathogen free), for Experiment 1 were obtained from W. B. Smith Feedmill (Columbia, MO). For Experiment 2, male Dekalb hybrid piglets were obtained from Jim Bagwell Farms (Bagley, AL). In the first experiment, the piglets were 1 d old; in the second experiment, the piglets were 4 d old. The piglets did not receive iron injections, were not castrated, and did not have their tails docked or eye teeth removed. All of these normal management practices were withheld specifically as requested by the investigators. Pigs were housed in wire-bottomed cages and received care in accordance with NRC guidelines (1996). The temperature of the room was maintained at 32.2°C. Initially, 4-6 piglets were kept in each cage to maintain normal body temperature. Pigs were fed the same diet during an adjustment phase before the test diets were fed. In the case of the first experiment, Diet B (Table 1) was fed for 3 d, whereas for the second experiment, Diet E was fed during the 4-d adjustment phase. During the adjustment phase, the piglets were taught to eat the slurried diets from pans placed in the cages. For the first 3 d, the slurry was 20% solids, and thereafter increased to 30% solids. The slurried diets were prepared fresh immediately before feeding. A slurried diet that was prepared fresh every 6 h was fed ad libitum to the piglets. Additional fresh water was provided after each feeding and each animal was carefully monitored. Eighteen hours before the radioactive dosing, the piglets were weighed, randomized according to treatment, placed in individual cages and switched to their respective test diets.

Radioisotopes and dosing.  All radioisotopes were produced by neutron irradiation of targets at the University of Missouri Research Reactor (Columbia, MO). ⁵⁹Fe (0.21 MBq/mg Fe) was produced by a 120-h irradiation of the elemental iron powder in a flooded can in a reflector irradiation position (position H1). The powder was weighed and vacuum sealed in a high purity quartz vial before irradiation. The flooded can ensured that the temperature of the vial during irradiation did not rise above the reactor pool temperature (~37°C). After irradiation, the activated iron was subjected to -ray spectroscopic analysis. The major -emitting radioimpurities included ⁵¹Cr, ⁵⁴Mn and ⁶⁰Co; the sum of these three radioimpurities totaled 3.5% of the ⁵⁹Fe activity. The radioimpurities did not cause any interference with required -ray measurements in the study. The major radioisotope produced from irradiation of natural elemental iron was ⁵⁵Fe (half-life = 2.7 y), which does not emit any -rays during decay. At the end of neutron irradiation, the amount of ⁵⁵Fe produced was 60% greater than the amount of ⁵⁹Fe activity. It is mentioned here for informational purposes only because the ⁵⁵Fe must be considered for disposal of radioactive waste resulting from animal experiments. The ⁵⁹Fe sulfate dosage formulation was prepared by dissolving the neutron-irradiated iron powder in concentrated sulfuric acid and evaporating completely. The ⁵⁹Fe sulfate was reconstituted with deionized water; sufficient acid remained to maintain an acidic pH of 1.88. The solution was sterile filtered.

⁴⁷Calcium (0.56 MBq/µg Ca) was produced by irradiation of 38% enriched ⁴⁶Ca (Ca nitrate) for 1 wk in the flux trap of the reactor. Following irradiation, the calcium nitrate (containing ⁴⁷Ca) was dissolved in water. ⁶⁵Zinc (40.7 MBq/mg Zn) was produced by irradiation of high purity zinc metal (99.9999%) in the flux trap of the reactor for 306 h. The zinc metal was dissolved in concentrated HCl and evaporated to dryness, converting the chemical form to ZnCl₂. ⁶⁷Cu (37 MBq/µg Cu) was produced by neutron irradiation of an enriched ZnO target (93% ⁶⁷Zn) with chemical separation as described (Zinn et al. 1994). After purification, the ⁶⁷Cu was evaporated to dryness (CuCl₂) and reconstituted in deionized water (pH 5.2). The stock solutions were sterile filtered using a 0.2-µm syringe filter.

The radioisotopes were mixed together as individual doses in a slurry of 10% unfortified rice meal (3 mL) 4 h before the pigs were dosed. Each dose was inside a polyethylene scintillation vial, and the radioisotopes were precisely measured in a calibrated geometry using a germanium detector. This detector was calibrated with a mixed -ray standard traceable to The National Institute of Standards and Technology (Gaithersburg, MD). The pigs were provided with and allowed to consume 50% of a meal before dosing; the remaining 50% of the meal was provided and consumed after dosing. The pigs were dosed by gastric gavage. First, the disposable gavage tube (4 mm, 41 cm long, 8890-701215, Sherwood Medical, St. Louis, MO), equipped with a 3-way stopcock valve, was inserted into the stomach and stomach contents were aspirated. Next, the dose was given, followed by a 5-mL rinse with deionized water. After dosing, the gastric gavage tube was folded and placed inside the scintillation vial. The dosing syringe was also rinsed with deionized water, which was placed in the scintillation vial. The syringe was checked to ensure that it was free of radioactivity. Finally, the scintillation vials and washings were measured. By subtracting the measurements, before and after dosing, the precise activity administered to each piglet was calculated.

In vivo measurement of radioisotopes.  The amount of radioactivity retained in the piglets was measured by placing each piglet individually in front of a germanium detector and acquiring a 8000-channel -ray spectrum (energy range 60-2000 keV). The energy resolution for germanium detectors (expressed as full-width half-maximum for a given -ray) is typically <2 keV for the 1332 keV ⁶⁰Co peak (Ehman and Vance 1991). The detectors in this study were capable of resolving all -rays emitted from the radiotracers (Zinn 1997). Initial measurements required 2 min of counting time; 10 min per measurement was needed at the end of the study. These in vivo measurements were done without anesthesia. The piglets were placed in a tubular, elastic orthopedic stockinet (10.16 cm diameter, HRI 8137-009074, Johnson and Johnson, New Brunswick, NJ) that was closed at both ends. The piglets could breathe through the sock; the darkness had a calming effect and the elastic pressure on their feet seemed to minimize movement. The material was disposable and also helped to contain the radioactivity. The piglet, inside the stockinet, was placed on his back inside a plastic box (Rubbermaid, Wooster, OH) secured on a table. One end of the plastic box was cut out and covered with padding; the piglet's neck rested on the edge. The sides of the box prevented the animal from twisting and the legs were extended upward. A picture of the counting position is given in Figure 1 and shows the location of the germanium detector, which was centered below the animal. The distance from the top of the detector to the bottom of the pig holder was 5 cm.

View larger version (116K): [in this window] [in a new window]   Fig 1. Picture showing the position of the pig with respect to the -ray detector during measurements of the whole-body radioisotope retention. Piglets were Dekalb hybrids fed a cereal-based diet with differing iron sources and dosage forms of radioactive iron. During actual measurements, the stockinet was closed at both ends.

In the first experiment, an intrinsic germanium detector (30% efficiency, Canberra, Meriden, CT) was used, with data acquisition through a ND9900 -ray spectroscopy system. The detector in the second experiment was a lithium drifted germanium detector (GeLi, 11.6% efficiency, Ortec, Oak Ridge, TN); data were acquired with a genie-PC -ray spectroscopy system (Canberra).

A phantom was counted separately and yielded ~2% error over the course of the studies. The reproducibility of these measurements served as a quality control index for the experiments each time the pigs were measured. The phantom consisted of a 1-L plastic bottle filled with water acidified with HCl and containing the same mixture of radioisotopes with which the pigs had been dosed. The ⁵⁹Fe in the phantom was the soluble iron sulfate form. The absolute counting efficiency of the phantom was measured in the pig counting geometry. Because the phantom closely resembled the body geometry of the piglets, the phantom efficiency data were used to calculate the retentions of the radioisotopes in the piglets.

Tissue and blood analyses.  Tissues from the first experiment were analyzed after the pigs were killed. The piglets were 20 d old and had received two doses of radioisotopes. The liver was analyzed for radioactivity and for elemental concentrations of iron, zinc and copper by flame atomic absorption. In both experiments, blood hematocrit and hemoglobin levels were determined. Total blood hemoglobin levels were quantitated using a commercial kit, Catalog # 525-A, from Sigma Diagnostics (St. Louis, MO). Blood plasma was also analyzed for ceruloplasmin activity after the method of Schosinsky et al. (1974) as modified by Lee et al. (1988). Ceruloplasmin activity was expressed in katals, which are defined as the moles of o-dianisidine oxidized per second.

	RESULTS

Abstract Introduction Methods Results Discussion References

Animal performance.  All piglets in the experiments gained weight and were healthy. In both experiments, there was no effect of diet on weight gain. In the first experiment, the mean weight of the 5-d-old pigs (day of first radioisotope dose) was 1900 (±108) g. At the end of the experiment, the mean weight of the 20-d-old pigs was 4600 (±200) g. In the second experiment, the mean weight of the 8-d-old pigs was 2100(±54) g, and increased to a mean weight of 2800 (±97) g at 15 d of age. At 15 d, the mean weight of the piglets in the second experiment was ~10% less than that of corresponding piglets in the first experiment. The specific pathogen-free status of piglets in Experiment 1 may have contributed to a greater weight gain.

Retention of ⁵⁹Fe.  The nonradioactive, dietary iron (both chemical form and level) had no significant effect on the retention of the ⁵⁹Fe in either experiment. Only the chemical form of the radioactive ⁵⁹Fe dosage significantly (P < 0.05) affected the ⁵⁹Fe retention. The ⁵⁹Fe retention data were determined using both the 192- and 1099-keV -rays and found to be equivalent. This fact demonstrated the accuracy and reproducibility of the -ray spectroscopy technique. The 192-keV peak, because of its lower energy -ray, was subject to more attenuation and had to be measured in the presence of a large -ray emission of ⁶⁷Cu at 185 keV. The 185-keV peak did not interfere with the measurement of the 192 keV -ray of ⁵⁹Fe.

The retention data for elemental ⁵⁹Fe powder were extrapolated by regression analysis back to the time of dosing to determine the apparent absorption (Heth and Hoekstra 1965). By regression analysis, the calculated value was 13.7 ± 1.1% (r2 = 0.97) for Experiment 1, not significantly different from 11.8 ± 1.0% (r2 = 0.94) for Experiment 2. The retention data for ⁵⁹Fe-sulfate were also extrapolated by regression analysis back to the time of dosing to determine the apparent absorption (Heth and Hoekstra 1965). By regression analysis, the calculated value was 26.5 ± 1.3% (r2 = 0.99) for Experiment 1.

Retention of ⁶⁵Zn.  In Experiment 1, as expected, there was no effect of the dosage form of ⁵⁹Fe on the retention of ⁶⁵Zn, a consideration because the two radiotracers were mixed together for dosing. The level of elemental iron in the diet (100 vs. 200 mg/kg diet) also did not influence the retention of ⁶⁵Zn. Data for ⁶⁵Zn retention, as well as extrapolations to determine apparent absorption, are presented in Figure 2.

View larger version (26K): [in this window] [in a new window]   Fig 2. In vivo retentions of 65Zn in Dekalb hybrid piglets fed a cereal-based diet as affected by (A) dietary iron source (Experiment 1) and (B) level of dietary zinc (Experiment 2). Data are means ± SEM, n = 10 per group (A), and n = 6 per group (B). The dotted lines are the regression lines for the last four data points, using all piglets in each treatment group. Fe 100 Diet refers to the diet supplemented with electrolytic Fe powder (100 mg/kg diet), whereas Fe SO4 Diet refers to the diet supplemented with Fe sulfate (100 mg/kg diet). Zn 25 Diet, Zn 50 Diet and Zn 100 Diet refer to diets supplemented with zinc carbonate (25, 50 and 100 mg/kg diet, respectively). 65Zn retention (A) was significantly (P < 0.05) reduced for piglets fed the FeSO4 Diet compared with those fed the Fe 100 Diet. Piglets showed significantly (P < 0.05) less 65Zn retention as dietary Zn increased (B).

The FeSO₄ Diet in Experiment 1 had a significant effect on the ⁶⁵Zn retention, compared with the Fe 100 Diet (elemental iron), although there was no difference in the apparent absorption (Fig. 2A). Upon extrapolation of the retention data by regression analysis, the value for piglets consuming this diet was 27.3 ±1.6% (r² = 0.94), not different than the 28.5% (r2 = 0.99) value determined for the Fe 100 Diet. However, the rate of ⁶⁵Zn elimination as determined by the slope of the retention curve was significantly greater for piglets consuming the FeSO₄ Diet. The slope was 1.6 ± 0.3 for this group, compared with a slope of 1.10 ± 0.03 for the Fe 100 group. The greater rate of elimination in the FeSO₄ group led to significantly less retention of the ⁶⁵Zn dose at 7.8 d after dosing. In absolute terms, the retention of ⁶⁵Zn in piglets consuming the Fe 100 Diet averaged 19.8 ± 1.2% at 7.8 d after dosing. The ⁶⁵Zn retention of the piglets fed the FeSO₄ Diet was significantly (P < 0.05) lower, and averaged 15.5 ± 1.04% at the same time.

The level of zinc in the diet significantly influenced the retention of the ⁶⁵Zn dose. Data are presented in Figure 2B for the three levels of dietary zinc in Experiment 2. As the level of dietary zinc decreased, the percentage of absorption of the ⁶⁵Zn dose increased. By regression analysis, the apparent ⁶⁵Zn absorptions for piglets consuming the Zinc 100 , Zinc 50 and Zinc 25 Diets were 26.0 ± 2.2% (r² = 0.86), 35.0 ± 1.1% (r² = 0.96) and 44.5 ± 2.2% (r2 = 0.89), respectively.

Retention of ⁴⁷Ca and ⁶⁷Cu.  The retention of ⁴⁷Ca was not significantly affected by dietary iron (form or level) or dietary zinc (Fig. 3). By regression analysis, the apparent ⁴⁷Ca absorption was 53 ± 1.6% (r2 = 0.94) in Experiment 1 and 60.8 ± 1% (r² = 0.97) in Experiment 2. Retention of ⁶⁷Cu was not affected by the dietary iron form in Experiment 1 (Fig. 4). The apparent absorption yielded by regression analysis was 42 ± 1.5% (r² = 0.99).

View larger version (24K): [in this window] [in a new window]   Fig 3. In vivo retentions of 47Ca in Dekalb hybrid piglets fed a cereal-based diet as affected by dietary iron source. Data are means ± SEM, n = 10 per group. The dotted lines are the regression lines for the last four data points, using all piglets in each experiment. Fe 100 Diet refers to the diet supplemented with electrolytic Fe powder (100 mg/kg diet), whereas Fe SO4 Diet refers to the diet supplemented with Fe sulfate (100 mg/kg diet). Dietary iron source did not have a significant effect on 47Ca retention in the piglets.

View larger version (24K): [in this window] [in a new window]   Fig 4. In vivo retention of 67Cu in Dekalb hybrid piglets fed a cereal-based diet as affected by dietary iron source. Data are means ± SEM, n = 10 per group. The dotted lines are the regression lines for the last four data points, using all piglets in the experiment. Fe 100 Diet refers to the diet supplemented with electrolytic Fe powder (100 mg/kg diet), whereas Fe SO4 Diet refers to the diet supplemented with Fe sulfate (100 mg/kg diet). Dietary iron source did not have a significant effect on 67Cu retention in the piglets.

Tissue analyses in Experiment 1.  A summary of the results for the iron, zinc and copper analyses of liver samples from Experiment 1 are reported in Table 2. These values agree closely with values per gram liver of 100 µg (1.79 µmol), 204 µg (3.12 µmol) and 197 µg (3.10 µmol) reported for iron, zinc and copper, respectively, in 3-wk-old pigs fed a high Fe diet (300 mg/kg diet) (Gipp et al. 1974). Dietary iron, regardless of form or level, had no effect on the copper and zinc concentrations in the liver. However, piglets consuming the iron sulfate diet had significantly higher (P < 0.05) levels of iron in the liver, compared with pigs fed elemental iron.

Because liver samples were analyzed for both radioactivity and elemental concentrations, the tissue specific activity for each radioisotope could be determined. The radioactivity measurement was calculated as the percentage of total dose per gram of liver, whereas the elemental concentration was given in micrograms of element per gram of liver. Therefore, division of the former by the latter (for the same pig) yielded %dose per microgram of element in the liver. This unit was subsequently converted to %dose per gram, or milligram, depending on the radioisotope. Specific activity results are presented in Table 2. The zinc and copper specific activities were not affected by dietary iron. However, the dietary iron and ⁵⁹Fe-dosage form both significantly affected the ⁵⁹Fe specific activity. ⁵⁹Fe specific activity was lowest for pigs dosed with ⁵⁹Fe-powder and was not affected by dietary iron source. Significantly higher ⁵⁹Fe specific activity was found for pigs dosed with the ⁵⁹Fe-sulfate compared with pigs fed the iron powder. Moreover, for pigs dosed with ⁵⁹Fe-sulfate, the ⁵⁹Fe specific activity was highest (83.7 ± 15.2) for pigs consuming the Fe 100 Diet; this contrasts with the lower value (55.7 ± 8.4) for pigs consuming the iron sulfate diet.

Blood analyses.  Blood hematocrit and hemoglobin levels were not affected by the experimental test diets in either of the two experiments. In Experiment 1, the hematocrit averaged 0.34 when the pigs were 4 d old, 0.37 at 12 d of age and 0.41 at termination at 20 d of age. The hemoglobin averaged 93, 102 and 118 g/L, respectively, for these same ages. In the second experiment, the hematocrit and hemoglobin averaged 0.29 and 111 g/L, respectively. Plasma ceruloplasmin levels in Experiment 1 were not affected by dietary iron. The mean for pigs fed the elemental iron diet was 163 ± 20 kkat/L, not different than 170 ± 20 kkat/L for pigs fed the iron sulfate diet. These values were measured at the termination of the study.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Our experiments showed that elemental ⁵⁹Fe powder was bioavailable; its apparent absorption was 50% that of ⁵⁹Fe sulfate. Dietary iron source did not affect retention of ⁴⁷Ca or ⁶⁷Cu (Figs. 3 and 4). There are few reports of effects of iron on absorption of calcium, but an increased Fe:Cu ratio was reported to reduce the hepatic storage of copper in pigs and guinea pigs (O'Dell 1989). We did not observe an effect of iron on copper concentrations in the liver.

Pigs have a gastrointestinal system similar to that of humans and should serve as a good model for trace mineral absorption by human infants. In addition, the young pig is a good model for the older human infant because hemoglobin levels in piglets at birth are 90-110 g/L, but decline to 40-50 g/L at 10 d of age (Blood et al. 1979). Human infants at birth have substantial iron reserves which, as in young pigs, become depleted with age. Our study demonstrated that elemental iron was readily available, with an apparent absorption of 12-14%. This compares with apparent and true fractional absorption values for iron of 12.6 and 38%, respectively, which have been previously reported for young pigs consuming supplemental iron in the form of ferrous sulfate (Atkinson et al. 1993, Morais et al. 1996). In a previous study, human infants receiving an iron-fortified rice meal cereal had a significantly higher hemoglobin concentration than infants receiving unfortified cereal (Walter et al. 1993). Thus, in both human infants and in a young pig model, elemental iron incorporated into rice meal cereal appears to be a readily available source of iron.

The results of Experiment 1 suggested that the iron sulfate diet, but not the elemental iron diet, decreased ⁶⁵Zn retention, likely due to the greater bioavailability of the iron sulfate diet. This prompted us in Experiment 2 to feed diets containing levels of iron powder that could potentially have an adverse effect on retention of zinc. The results of Experiment 2 indicate that iron powder did not inhibit zinc retention, even when iron was fed at levels considerably higher than the NRC iron requirement of 100 mg/kg diet for 5- to 10-kg pigs (NRC 1988). The level of zinc required to maintain total body tissue levels similar to those in nursing pigs has been estimated to be between 14 and 20 mg/kg diet, when baby pigs were fed a casein diet (Shanklin et al. 1968). The lowest level of zinc in the diets fed in this study (25 mg/kg diet) was thus near the lowest level required to maintain body zinc stores in the pigs, and was well below the NRC requirement of 100 mg/kg diet for 5- to 10-kg pigs (NRC 1988). Even at this low dietary zinc level, high dietary iron as iron powder did not adversely affect ⁶⁵Zn retention or trace element accumulation in the liver. Because the iron sulfate-supplemented diet inhibited ⁶⁵Zn retention to a greater extent than did the iron powder-fortified diet in Experiment 1, it appears that the iron powder diet may be more beneficial than the iron sulfate diet in preventing a negative Fe-Zn interaction. We hypothesize that, relative to zinc salts present in the pig diets, elemental iron powder may be less readily solubilized in the stomach and upper gastrointestinal tract, and therefore compete to a lesser extent for zinc retention than would a more soluble iron salt such as iron sulfate. The greater liver iron values of piglets consuming the iron sulfate diet, relative to those consuming the elemental iron diet, may be a partial reflection of the greater solubility of iron in the iron sulfate diet.

There was good agreement between the two experiments for the apparent zinc absorption for animals consuming the same zinc diet (Zn 100). Animals fed this diet had an apparent zinc absorption of 28% in Experiment 1 and 26% in Experiment 2, values closely approximating the zinc absorption of 28-31% reported in infants (Fairweather-Tait et al. 1995). The apparent zinc absorptions increased for animals consuming the Zn 50 and Zn 25 diets in Experiment 2, to 35 and 45%, respectively. However, in terms of total mass of zinc absorption, animals consuming the Zn 100 diet had the greatest zinc absorption. This finding is of particular significance given the importance of zinc to growth and the fact that 20-40% of infants (6-12 mo of age) consume <67% of the RDA for zinc; the same is true for 80% of toddlers (Lönnerdal 1994). Our data in the newborn pig model confirm that supplemental zinc is absorbed without any negative effect on iron or calcium metabolism.

An important finding of the studies reported here is that high levels of dietary iron supplied from elemental iron-fortified rice meal cereal did not adversely affect zinc retention or tissue zinc accumulation in a young pig model. The highest dietary iron level fed to the piglets would approximate an infant consuming six servings per day of completely iron-fortified rice meal cereal (15 g per serving, 500 mg iron/kg cereal) without negative effects on zinc retention. This level of cereal consumption is considerably greater than the recommended two servings per day, suggesting that consumption even in excess of the recommendation would likely still not have an adverse effect on zinc retention. Furthermore, iron fortification at these levels did not appear to affect the metabolism of calcium or copper.

	FOOTNOTES

Manuscript received 4 February 1998. Initial reviews completed 20 March 1998. Revision accepted 5 October 1998.

	ACKNOWLEDGMENTS

We thank Sandra Bartholmey for her expert assistance with the design of the experiments and her thoughtful review of the manuscript. We also thank Qi Wu and Cheryl A. Smyth for assistance with animal care and manuscript formatting.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Anderson, R. H. (1985) Breakfast cereals and dry milled corn products. In: Iron Fortification of Foods (Clydesdale, F. M. & Wiemer, F. L., eds.), p. 116. Academic Press, Orlando, FL.
• Atkinson S. A., Shah J. K., Webber C. E., Gibson I. L., Gibson R. S. A multi-element isotopic tracer assessment of true fractional absorption of minerals from formula with additives of calcium, phosphorus, zinc, and iron in young piglets. J. Nutr. 1993; 123:1586-1593
• Blood, D. C., Henderson, J. A. & Radostits, O. M. (1979) Iron deficiency. In: Veterinary Medicine: A Textbook of Diseases of Cattle, Sheep, Pigs, and Horses, 5th ed., pp. 883-887. Lea & Febiger, Philadelphia, PA.
• Bustad L. K. & McClellan R. O. (1966) Swine in Biomedical Research. Frayn Printing Company, Seattle, WA.
• Ehman, W. D. & Vance, D. E. (1991) Solid-state semiconductor detectors. In: Radiochemistry and Nuclear Methods of Analysis, pp. 234-235. John Wiley and Sons, New York, NY.
• Fairweather-Tait S. J. Iron-zinc and calcium-iron interactions in relation to zinc and iron absorption. Proc. Nutr. Soc. 1995; 54:465-473[Medline]
• Fairweather-Tait S. J., Wharf S. G., Fox T. E. Zinc absorption in infants fed iron-fortified weaning food. Am. J. Clin. Nutr. 1995; 62:785-789[Abstract/Free Full Text]
• Gipp W. F., Pond W. G., Kallfelz F. A., Tasker J. B., van Campen D. R., Krook L., Visek W. J. Effect of dietary copper, iron and ascorbic acid levels on hematology, blood and tissue copper, iron and zinc concentrations and ⁶⁴Cu and ⁵⁹Fe metabolism in young pigs. J. Nutr. 1974; 104:532-541
• Gordon, D. T., Zinn, K., Stoops, D., Trokey, D., Gusy, R., Peluso, M. & Ratliff, V. (1993) The effect of increasing dietary iron on the retention of ⁵⁹Fe, ⁶⁵Zn, and ⁶⁷Cu in the growing rat. In: Trace Elements in Man and Animals-TEMA 8 (Anke, M., Meissner, D. & Mills, C. F., eds.), pp. 625-629. Verlag Media Touristik, Gersdorf, Germany.
• Heth D. A., Hoekstra W. G. Zn-65 absorption and turnover in rats. J. Nutr. 1965; 85:367-374[Medline]
• Lee D., Schroeder J., Gordon D. T. Enhancement of Cu bioavailability in the rat by phytic acid. J. Nutr. 1988; 118:712-717
• Lönnerdal, B. (1994) Choose foods with iron, zinc and calcium. In: Dietary Guidelines for Infants. Pediatric Basics 69: 25-29.
• Morais M. B., Feste A., Miller R. G., Lifschitz C. H. Effect of resistant and digestible starch on intestinal absorption of calcium, iron and zinc in infant pigs. Pediatr. Res. 1996; 39:872-876[Medline]
• National Research Council (1981) The Food Chemicals Codex, 3rd ed., pp. 151-152. National Academy Press, Washington, DC.
• National Research Council (1988) Nutrient Requirements of Swine, 9th ed. National Academy Press, Washington, DC.
• National Research Council (1996) Guide for the Care and Use of Laboratory Animals. National Academy Press, Washington, DC.
• Newport M. J., Henschel M. J. Evaluation of the neonatal pig as a model for infant nutrition: effects of different proportions of casein and whey protein in milk on nitrogen metabolism and composition of digesta in the stomach. Pediatr. Res. 1984; 18:658-662[Medline]
• O'Dell B. Mineral interactions relevant to nutrient requirements. J. Nutr. 1989; 119:1832-1838
• Patrick, J. (1985) Elemental sources. In: Iron Fortification of Foods (Clydesdale, F.M. & Wiemer F. L., eds.), p. 116. Academic Press, Orlando, FL.
• Pond, W. G. & Houpt, K. A. (1978) Nutrition. In: The Biology of the Pig, pp. 276-335. Cornell University Press, Ithaca, NY.
• Sandström B., Davidsson L., Cederblad A., Lönnerdal B. Oral iron, dietary ligands and zinc absorption. J. Nutr. 1985; 115:411-414
• Schosinsky K. H., Lehman H. P., Beeler M. F. Measurement of ceruloplasmin from its oxidase activity in serum by use of o-dianisidine dihydrochloride. Clin. Chem. 1974; 20:1564-1567[Abstract]
• Shanklin S. H., Miller E. R., Ullrey D. E., Hoeffer J. A., Luecke R. W. Zinc requirements of baby pigs on casein diets. J. Nutr. 1968; 96:101-108
• Solomons N. W., Jacob R. A. Studies on the bioavailability of zinc in humans: effects of heme and nonheme iron on the absorption of zinc. Am. J. Clin. Nutr. 1981; 34:475-482[Abstract/Free Full Text]
• Valberg L. S., Flanagan P. R., Chamberlain M. J. Effects of iron, tin, and copper on zinc absorption in humans. Am. J. Clin. Nutr. 1984; 40:536-541[Abstract/Free Full Text]
• Walravens P. A., Hambidge K. M., Koepfer D. M. Zinc supplementation in infants with a nutritional pattern of failure to thrive: a double-blind, controlled study. Pediatrics 1989; 83:532-538[Abstract/Free Full Text]
• Walter T., Dallman P. R., Pizarro F., Velozo L., Pena G., Bartholmey S. J., Hertrampf E., Olivares M., Letelier A., Arredondo M. Effectiveness of iron-fortified infant cereal in prevention of iron deficiency anemia. Pediatrics 1993; 91:976-982[Abstract/Free Full Text]
• Ziegler E. E., Fomon S. J. Strategies for the prevention of iron deficiency: iron in infant formulas and baby foods. Nutr. Rev. 1996; 54:348-354[Medline]
• Zinn K. R. (1997) Radiotracers for nutritionally essential mineral elements, In: Handbook of Nutritionally Essential Mineral Elements (O'Dell, B. & Sunde, R.A., eds.), pp. 661-679. Marcel Dekker, New York, NY.
• Zinn K. R., Chaudhuri T. R., Cheng T. P., Morris J. S., Meyer W. A. Production of no-carrier-added ⁶⁴Cu from zinc metal irradiated under boron shielding. Cancer 1994; 73:774-778[Medline]
• Zinn, K. R., Gordon, D., Stoops, D., Fernando, L. & van den Berg, G. (1995) Retention and interactions of iron powder or iron sulfate in a pig model. FASEB J. 9: A976 (abs.).
• Zinn K. R., Johanning G. L., Smyth C. A., Wu, Q. & Mountz, J. M. (1997) Nutritional interaction between iron and zinc determined in young pigs by in vivo gamma-ray spectroscopy. FASEB J. 11: A444 (abs.).
• Zinn K. R. & Morris, J. S. (1988) High-resolution gamma-ray spectroscopy as an in vivo tool for following the zinc-selenium interaction, In: Trace Elements in Man and Animals 6 (Hurley, L. S., Keen, C. L., Lönnerdal B. & Rucker, R. B., eds.), pp. 591-592, Plenum Press, New York, NY.

0022-3166/99 $3.00 ©1999 American Society for Nutritional Sciences