Provision of adequate iron in fortified foods is an essential aspect of the nutritional approach to preventing iron deficiency in children throughout the world (1,2). Although ferrous sulfate is generally considered the optimal source of iron in fortification strategies, it is not practical to add this form of iron to all foods. Elemental iron is also widely used, but depending on its physical characteristics, it may have poor bioavailability.
Increasingly, both in the United States and in the developing countries, ferrous fumarate is being selected as an iron fortificant. Ferrous fumarate appears to provide a compromise between other forms of iron in terms of cost, bioavailability, and sensory/shelf-life characteristics. However, absorption of iron from ferrous fumarate may not be as good as from ferrous sulfate (3,4). Furthermore, an important aspect of the use of any iron source as a fortificant is the ability to have its absorption enhanced by other food ingredients, especially ascorbic acid. Whereas ascorbic acid is readily demonstrated in all age groups to enhance ferrous sulfate absorption, available data for ferrous fumarate are limited and conflicting (5–8).
Therefore, our goal was to evaluate the effects of juice containing ascorbic acid on ferrous fumarate absorption using a population of well-nourished children, ages 4 to 7 years. We chose an age group in which a school or preschool feeding program may use a fortified food product and a population at low risk for health problems such as parasite infestation that may affect the results. Our hypothesis was that iron absorption from ferrous fumarate would be significantly enhanced in these children by orange juice containing ascorbic acid versus apple juice.
SUBJECTS AND METHODS
Healthy children, 4.0 to 7.9 years of age, in the greater Houston area were recruited through public advertising and via a database of families who have participated in previous studies. Subjects were selected to reflect the approximate racial and ethnic distribution of the greater Houston metropolitan population. Subjects were eligible for enrollment if they had no underlying serious health conditions, had a body mass index between the 3rd and 97th percentile for age and sex, were not born prematurely (<37 weeks gestation), had a birth weight >2500 g, and were not taking regular medications. Any vitamin and mineral supplements were discontinued at least 2 weeks before enrollment and throughout the study period. The institutional review board of Baylor College of Medicine and affiliated hospitals approved the protocol and informed written consent was obtained from 1 parent of each child.
The 57Fe fumarate was prepared in collaboration with one of the major commercial suppliers of iron fortification compounds, Dr Paul Lohmann GmbH KG (Emmerthal, Germany). The procedure used was similar to the industrial production of the equivalent fortification compounds but was done as a small-scale laboratory procedure. Elemental iron isotopically enriched with 57Fe (96%) was purchased from Nippon Sanso Europe (Plaisir, France). For the preparation of 57Fe fumarate, 57Fe was dissolved in sulfuric acid (20%) and combined with an aqueous solution of sodium fumarate. The iron fumarate precipitate was washed with ethanol until no residual sulfate was detected, dried, and grounded in an agate mortar to a fine powder. The 58Fe (93% enrichment) was purchased from Trace Sciences International (Toronto, Ontario, Canada) and converted to the sulfate as previously described (3).
Muffins were prepared with 4 mg of elemental iron added to each muffin (12.5 mg ferrous fumarate). The muffins were individually prepared in the metabolic research kitchen at the US Department of Agriculture/Agricultural Research Service Children's Nutrition Research Center in Houston, TX. A well was made in the center of the batter using a toothpick, and the premeasured iron was dropped from a plastic bullet into the well. Approximately 0.5 mL of deionized water was used to rinse the bullet of any residual isotope. The toothpick was used to stir the isotope into the batter. The muffins were then baked, cooled, weighed, and stored in vacuum-sealed bags (Food Saver, Neosho MO) in a freezer.
Mineral Absorption Study
For all of the study visits, subjects arrived after a 2- to 3-hour fast and were admitted to the Metabolic Research Unit at the US Department of Agriculture/Agricultural Research Service Children's Nutrition Research Center. Anthropometric data and vital signs were obtained at each visit and the relevant medical history reviewed.
At the first visit, subjects were given 100 mL of orange juice containing 50 mg ascorbic acid and 1 mg 58Fe (as ferrous sulfate) as a reference dose. Subjects fasted 2 hours after consuming the juice and reference dose. At the second visit, which occurred 1 to 3 days after the first visit, the subjects were given a meal consisting of 1 or 2 muffins (1 labeled with 4 mg 57Fe as ferrous fumarate) and were randomized to 1 of the 2 juices: either 50 mL orange juice containing approximately 25 mg ascorbic acid or an equal volume of apple juice that did not contain any ascorbic acid. All of the subjects were offered a second muffin identical to the first except without additional iron. Subjects were observed fasting for 30 minutes after the meal and then were discharged home to continue fasting for a total of 2 hours.
Fourteen days later subjects were readmitted. Blood was drawn for measurement of a complete blood count, serum ferritin, serum hepcidin, and iron isotope ratios. The subjects then consumed a meal identical to the one consumed in the previous visit, except it included the other juice (if apple juice without ascorbic acid was given in the first meal, then orange juice with ascorbic acid was given in the second meal).
Fourteen days later a final blood sample was taken for iron isotope ratio measurement to assess the incorporation of 57Fe from the second muffin meal. The amount of iron incorporated was corrected for the incorporation from the first meal.
Two nonconsecutive 24-hour dietary recalls were obtained by telephone by the study dietitian from each parent to determine their child's intake during the study. These were not part of the original study design and were performed shortly after the third visit. Dietary intake data were collected and analyzed using Nutrition Data System for Research software version 2007, developed by the Nutrition Coordinating Center, University of Minnesota, Minneapolis.
Iron Status Measurements
Serum ferritin, hemoglobin, and hematocrit were measured once, at the third visit. Serum ferritin was determined in our laboratory using the Fer-Iron II procedure, a 2-site immunoradiometric assay from Ramco Laboratories, Stafford, TX. Antibody to ferritin is coated on the surface of plastic beads. Ferritin present in the serum binds to the antibody-coated beads. Radiolabeled antibody in turn binds to the ferritin on the solid-phase antibody. The solid phase is washed and counted in a gamma counter. The amount of radiolabel bound to the solid phase is directly proportional to the concentration of ferritin present in the serum. Hemoglobin and hematocrit were sent to a commercial laboratory (Quest Diagnostics, Houston, TX) and determined by cytometry.
Iron Absorption Measurements
Iron isotope ratios were measured using inductively coupled plasma mass spectrometry as previously described (9) and used to measure incorporation of isotope into red blood cells (3). Iron isotope ratios from the first blood draw were used to assess the reference dose absorption (from the 58Fe incorporation) and absorption of 57Fe fumarate from the first muffin (from 57Fe incorporation). Data from the second blood draw was used to measure 57Fe incorporation from the second muffin meal, correcting for the residual isotope enrichment in the blood from the first meal with 57Fe.
Red blood cell iron incorporation was calculated using the hemoglobin, estimated blood volume (70 mL/kg), and isotope enrichment (10,11). This value for estimated blood volume was selected as a value between the one usually used for infants (80 mL/kg) and that of adults (65 mL/kg). It is recognized to be an approximation. Values were converted to estimated iron absorption by dividing by 0.9 based on the assumption that 90% of absorbed iron was incorporated into red blood cells. Because each subject served as his or her own control both for the 57Fe absorption and for comparison with the reference dosing, small variations in calculated blood volume or the accuracy of the 90% value would have minimal effects on the findings from the study, although they need to be considered in comparisons of absorption values between studies and population groups.
Hepcidin was measured by Intrinsic LifeSciences LLC of La Jolla, CA via competitive enzyme-linked immunosorbent assay as previously described (12). Samples were obtained from subjects 14 days after consumption of a meal consisting of 1 muffin labeled with 4 mg 57Fe as ferrous fumarate. Antibody to human hepcidin was purified on staphylococcal protein A columns (Thermo Fisher Scientific, Rockford, IL); 96-well plates were coated with the antibody and incubated with 100 μL of 1:20 dilution of serum in Tris-buffered saline containing 0.05% Tween-20, with 10 ng/mL of biotinylated hepcidin-25 added as the tracer. Duplicate 10-μL aliquots of plasma were diluted 1:20 in phosphate-buffered saline, pH 7.4, and bioactive serum hepcidin measured using previously described rabbit polyclonal antibodies (Intrinsic LifeSciences LLC). High-performance liquid chromatography–purified, synthetic, bioactive hepcidin (Bachem Bioscience Inc, King of Prussia, PA) was used as reference material for construction of duplicate 12-point standard curves. Curve fitting was performed using GraphPad Software (San Diego, CA).
Statistical Analysis and Sample Size Determination
Differences in iron absorption with orange juice versus apple juice were assessed by analysis of variance with meal (with orange juice or with apple juice) and order (orange juice first or apple juice first) and number of muffins consumed as independent variables. Data were also analyzed by analysis of covariance with measures of iron status (serum ferritin and/or reference dose iron absorption, serum hepcidin), nutrient intake, body size (height or weight), and age as covariates.
The relation between iron absorption and body size (height and weight), age, and iron status (serum ferritin and/or reference dose iron absorption, serum hepcidin) were assessed by linear and multiple regression analyses.
Serum ferritin and serum hepcidin were log transformed (to base 10) before analysis to normalize the distribution. Statistical significance was assumed at P < 0.05. Data are reported as mean ± standard error of the mean unless otherwise stated.
We hypothesized that fractional ferrous fumarate iron absorption would be enhanced by orange juice by a difference equal to two thirds of the standard deviation of the expected iron absorption with apple juice (3). Evaluation of 20 subjects had a power of 80% to detect this difference; therefore, 22 subjects were recruited to allow for patient attrition.
A total of 22 subjects were enrolled, with 21 subjects (11 girls and 10 boys) completing the study. One male subject was dropped because of failure to collect the blood samples. There were no significant differences between the sexes for any baseline or outcome variables, and sex was not a significant factor in any of the analysis and was removed from the analysis of covariance analyses.
The mean weight at the time of the first visit was 20.7 ± 0.7 kg and height was 115.4 ± 1.6 cm. Age was 6.1 ± 0.3 years, with a minimum of 4.6 years and a maximum of 7.9 years. Although not recruited by age groups, we have shown the baseline characteristics by age in Table 1 to demonstrate the similarities between age groups except for the trend toward higher ferritin in the older age group.
The nutrient composition of the muffin is shown in Table 2. The average weight of the muffins with the isotope was 27 ± 2 g. Most children consumed 2 muffins at each visit (76% did so at the first visit and 81% at the second visit). The apple juice did not contain any polyphenols (specifically, daidzein, genistein, glycitein, coumestrol, biochanin A, or formononetin) that may have reduced iron absorption. The orange juice contained 1.69 mg citric acid per 100 g (13).
The supplemental meals provided 4 mg of added 57Fe ferrous fumarate and either 0 or 25 mg of supplemental ascorbic acid. Each muffin (before adding ferrous fumarate) provided only 0.15 mg of iron and 0.28 mg of ascorbic acid and, thus, consumption of the second muffin increased iron intake by <5%. Therefore, the overall ascorbic acid:iron molar ratio was almost exactly 2:1 in this study varying by <0.1 based on the number of unlabeled muffins ingested. There was no statistical difference in absorption of reference dose, ferrous fumarate with orange juice, or with apple juice related to number of muffins consumed.
Effects of Orange Juice on Ferrous Fumarate Absorption
Orange juice increased iron absorption from 5.5% ± 0.7% to 8.2% ± 1.2%, P < 0.001 (Table 3). We evaluated various covariates to determine their relation with both the actual amount of iron absorbed from each meal and the individual differences in iron absorption between the 2 doses. Covariates with no significant effect (P > 0.1) included sex, study order (either study orange juice or apple juice first), and nutrient intakes, including iron. However, methodological limitations should be noted as described in the Subjects and Methods section related to iron intake assessment.
Unexpectedly, we identified a significant relation with measures of body size, age, iron absorption, and the differential absorption of iron with orange juice versus apple juice. There was a highly significant relation, P < 0.01, such that older and larger children had a beneficial effect of orange juice on ferrous fumarate absorption and younger children in the study did not. This effect was not significantly related to body mass index. Figure 1 shows the relation between age and the difference between absorption with and without ascorbic acid. Although the relation appears linear, the data have been separated into children 4.5 to 5.9 years old and children 6.0 to 7.9 years old. A significant beneficial effect of the orange juice was seen only in the older group (Fig. 2).
To further explore these relations, we evaluated age, height, and weight and their relations to the absorption of each muffin, the reference dose absorption, and the difference in absorption between studies. These results are shown in Table 3. We found that body size and age were significantly related to both the absorption of iron from ferrous fumarate with orange juice and the difference between the 2 doses. Iron status (as assessed by log-transformed serum ferritin) was significantly inversely related to iron absorption from the meal containing apple juice (r = −0.67, P = 0.001) but not as closely related to the absorption of iron from the meal containing orange juice (r = −0.38, P = 0.09).
Serum hepcidin averaged 40 ± 22 ng/mL (standard deviation) and was significantly lower in boys than girls, averaging 22 ± 13 ng/mL in 10 boys and 59 ± 12 ng/mL (standard deviations) in 11 girls, P = 0.048. Serum hepcidin was correlated to serum ferritin, y = 0.46x + 16.6, r = 0.78, P < 0.001, where y = serum ferritin and x = serum hepcidin. Log-transformed serum ferritin was related to log-transformed serum hepcidin by the relation y = 0.32 x + 1.02, r = 0.65, P = 0.002. Five of the 21 subjects had values for hepcidin below the limit of detection of 5 ng/mL and values of 2.5 ng/mL were used for these subjects as the mid-range of 0 and the lower limit of detection. Of the 5 subjects with a serum ferritin <12 ng/mL, 4 also had hepcidin values below the range of detection. One subject with a serum ferritin concentration of 10 ng/mL had a hepcidin of 19 ng/mL.
The log-transformed serum hepcidin was not significantly inversely related to iron absorption from the meal containing apple juice (r = −0.32, P = 0.15), from the meal containing orange juice (r = −0.11, P = 0.62), or from the reference dose absorption (r = 0.345, P = 0.13).
We found that giving orange juice provides a modest dose of ascorbic acid leading to an ascorbic acid:iron molar ratio of 2:1, significantly increasing the absorption of ferrous fumarate in a group of young children with good iron status. We cannot exclude a possible contribution to these effects from other components of orange juice, such as citric acid (13); however, it is likely that ascorbic acid was primarily responsible for this effect. Our results add to the limited and conflicting present literature on ascorbic acid and ferrous fumarate absorption (5–8). In young infants, adding 50 mg of ascorbic acid to a large amount of ferrous fumarate (30 mg of iron, ascorbic acid:iron molar ratio 0.5:1) had no effect on iron absorption (5), nor did increasing the ascorbic acid:iron molar ratio from 3.2:1 to 6.3:1 (6). In adults, 1 study showed no effect of ascorbic acid on ferrous fumarate absorption from a chocolate-flavored drink (7). The only previous study to show an effect of ascorbic acid was Fidler et al (8), in which iron absorption was significantly increased from a ferrous fumarate–fortified infant cereal by the addition of ascorbic acid. Of note, the test cereal was similar to that previously studied by Davidsson et al (6), in which no effect of ascorbic acid on iron absorption was seen in infants.
In the present study, iron status (as assessed by log-transformed serum ferritin) was significantly inversely related to iron absorption from ferrous fumarate given without ascorbic acid, but not to the absorption of iron from a small meal containing ascorbic acid. Although serum hepcidin was correlated with serum ferritin in the study subjects, the use of serum hepcidin as a marker of iron status did not lead to a similar relation in this small study or improve the sensitivity of serum ferritin as a marker of iron absorption.
Hepcidin and ferritin both show a similar response to changes in iron stores and inflammatory processes; however, changes in hepcidin take place in a matter of hours, whereas ferritin changes are much slower. In adults, there is a sex difference in serum hepcidin, with lower levels in adult females, most likely because of lower body iron stores in women (12). However, we found the opposite sex difference in the children in our study, with average serum hepcidin greater in girls than boys. Indeed, the mean hepcidin values observed in this study closely agree with the median hepcidin values observed in adult women (65 ng/mL), whereas the mean values for boys were 5-fold lower than the median value in adult men (112 ng/mL) (12). The reason for this difference is uncertain because all of our subjects were healthy and prepubertal, but it may relate to an increased requirement for iron in young boys as compared with young girls. The normal ranges of serum hepcidin have not yet been definitely established in this age group for either sex and will require further studies.
The relation between ferritin and iron absorption is in keeping with a similar study we carried out in 3- to 6-year-old children, in which the absorption of ferrous sulfate was significantly inversely correlated with serum ferritin from an apple juice–containing meal, but not from an orange juice–containing meal (14). This suggests that in well-nourished populations, ascorbic acid or other components of orange juice are less critical determinants of iron absorption. It is important to note that our subjects were neither anemic nor iron deficient, so we cannot say whether this effect would be maintained across a wider range of iron status.
We chose to compare orange juice with apple juice because these are common types of products that are given with iron-fortified foods. The use of pure supplements of ascorbic acid would not be in keeping with practical intervention programs in this age group. Nonetheless, we cannot strictly exclude that other components of orange juice, such as citric acid, contributed to this effect along with the ascorbic acid. We also cannot exclude the possibility that polyphenols in apple juice inhibited iron absorption, but these were negligible in the processed apple juice we provided (15).
An unexpected finding was the significant relation between age or body size and iron absorption. These results should be interpreted with caution because evaluation of results based on age or body size was not a part of the initial study design and the study was not specifically powered to identify differences in effects based on age or body size. Nonetheless, this secondary analysis showed that the age did not significantly affect ferrous fumarate absorption without ascorbic acid (apple juice), but iron absorption with ascorbic acid (orange juice) increased as age increased. A similar effect was seen for body size (height or weight). Because age, height, and weight are so closely correlated, it is not possible to determine which factors were responsible for the relation. We could not identify any dietary factors or a confounding effect of iron status that may explain this relation, but this study was not designed to evaluate these factors. Further studies would be needed to consider the effects of age and body size on these relations.
It seems unlikely that specific dietary components changed markedly during the 4- to 8-year age range. Rather, it is more likely that intestinal maturation gradually changed over time leading to a greater benefit of ascorbic acid in older children. The factors related to intestinal maturation could include gastric acidity or hormonal factors, including maturation of metal ion transporters (16). Although it is generally believed that ascorbic acid enhances iron absorption by increasing solubility or forming easily absorbed iron chelates, it is also possible that other mechanisms, poorly elucidated, are involved (17).
Although it is difficult to compare iron absorption values between studies because of differences in iron status, subject characteristics, test meals, and iron loads, the iron absorption from ferrous fumarate in this study was relatively good and similar to absorption of ferrous sulfate from a 5-mg iron dose in iron-replete children of a similar age (14). This contrasts with a previous study we carried out in younger, less iron-replete children, in which iron absorption was significantly higher from ferrous sulfate than from ferrous fumarate (3). In that study absorption of ferrous fumarate was much lower than the present study. Other data also suggest that iron absorption in children may be lower from ferrous fumarate than from ferrous sulfate (18).
In a previous study (18), the average absorption of 4 mg ferrous fumarate in a fortified food given with apple juice (without ascorbic acid) was 0.22 mg iron (5.5% of 4 mg). This is approximately 30% of the median total requirement for absorbed iron for children 4 to 8 years (0.74 mg/day) (19). The average absorption of ferrous fumarate with orange juice containing ascorbic acid in the present study was 0.33 mg iron (8.2% of 4 mg). This is approximately 44% of the median total requirement for absorbed iron for children 4 to 8 years old (0.74 mg/day) (19).
In a similar study involving ferrous sulfate with and without ascorbic acid in 3- to 6-year-old children, the absorption of 5 mg ferrous sulfate in a fortified food with and without ascorbic acid was 0.33 mg and 0.38 mg, respectively (14). These results are relatively similar between iron sources, although a small relative benefit for ferrous sulfate appears to persist.
From a global perspective, the children in this study were iron replete and at low risk of parasite infection. Therefore, our data should not be used to evaluate the use of ferrous fumarate in those populations without caution. As such, we chose not to attempt to compare against populations using normalization to a fixed ferritin (eg, 40 ng/mL), although we have provided the raw data so that this comparison can be made if desired.
In summary, we have demonstrated a significant enhancing effect of orange juice on absorption of ferrous fumarate in young children. However, the effect of ascorbic acid may depend, in part, on the subjects' age and serum ferritin level. Policy makers will need to take these factors into account when considering the appropriate iron fortificant for a population and whether ascorbic acid should be included. Ideally, iron absorption should be assessed directly in the target population and should be reevaluated as the iron status of the target population changes.
We gratefully acknowledge the contributions of Sopar Seributra, RN; the staff of the Metabolic Research Unit; Gloria Orozco and Shannon Mitchell, MD, for their help with study visits; Maria Hamzo for sample analysis; and Richard Hurrell, PhD, for providing the isotopically labeled ferrous fumarate.
1. Singh M, Sanderson P, Hurrell RF, et al
. Iron bioavailability: UK food standards agency workshop report. Br J Nutr 2006; 96:985–990.
2. Abrams SA. New approaches to iron fortification
: role of bioavailability studies. Am J Clin Nutr 2004; 80:1104–1105.
3. Perez-Exposito AB, Villalpando S, Rivera JA, et al
. Ferrous sulfate is more bioavailable among preschoolers than other forms of iron in a milk-based weaning food distributed by PROGRESA, a national program in Mexico. J Nutr 2005; 135:64–69.
4. Tondeur MC, Schauer CS, Christofides AL, et al
. Determination of iron absorption
from intrinsically labeled microencapsulated ferrous fumarate
(sprinkles) in infants with different iron and hematologic status by using a dual-stable isotope method. Am J Clin Nutr 2004; 80:1436–1444.
5. Zlotkin SH, Schauer CS, Agyei SO, et al
. Demonstrating zinc and iron bioavailability from intrinsically labeled microencapsulated ferrous fumarate
and zinc gluconate sprinkles in young children. J Nutr 2006; 136:920–925.
6. Davidsson L, Kastenmayer P, Szajewska H, et al
. Iron bioavailability in infants from an infant cereal fortified with ferric pyrophosphate or ferrous fumarate
. Am J Clin Nutr 2000; 71:1597–1602.
7. Hurrell RF, Reddy MB, Dassenko SA, et al
. Ferrous fumarate
fortification of a chocolate drink powder. Br J Nutr 1991; 65:271–283.
8. Fidler MC, Davidsson L, Zeder C, et al
. Iron absorption
from ferrous fumarate
in adult women is influenced by ascorbic acid but not by Na2EDTA. Br J Nutr 2003; 90:1081–1085.
9. Kastenmayer P, Davidsson L, Galan P, et al
. A double stable isotope technique for measuring iron absorption
in infants. Br J Nutr 1994; 71:411–424.
10. Chen Z, Griffin IJ, Plumlee LM, et al
. High resolution inductively coupled plasma mass spectrometry allows rapid assessment of iron absorption
in infants and children. J Nutr 2005; 135:1790–1795.
11. Abrams SA, Wen J, O'Brien KO, et al
. Determination of iron absorption
from intrinsically labeled microencapsulated ferrous fumarate
(sprinkles) in infants with different iron and hematologic status by using a dual-stable-isotope method. Biol Mass Spectrom 1994; 23:771–775.
12. Ganz T, Olbina G, Girelli D, et al
. Immunoassay for human hepcidin
. Blood 2008; 112:4292–4297.
13. Penniston KL, Nakada SY, Holmes RP, et al
. Quantitative assessment of citric acid in lemon juice, lime juice, and commercially-available fruit juice products. J Endourol 2008; 22:567–570.
14. Shah M, Griffin IJ, Lifschitz CH, et al
. Effect of orange and apple juices on iron absorption
in children. Arch Pediatr Adolesc Med 2003; 157:1232–1236.
15. Seeram NP, Aviram N, Zhang Y, et al
. Comparison of antioxidant potency of commonly consumed polyphenol-rich beverages in the United States. J Agric Food Chem 2008; 56:1415–1422.
16. Lonnerdal B. Trace element nutrition of infants–molecular approaches. J Trace Elem Med Biol 2005; 19:3–6.
17. Wienk KJ, Marx JJ, Santos M, et al
. Dietary ascorbic acid raises iron absorption
in anaemic rats through enhancing mucosal iron uptake independent of iron solubility in the digesta. Br J Nutr 1997; 77:123–131.
18. Sarker SA, Davidsson L, Mahmud H, et al
. Helicobacter pylori infection, iron absorption
, and gastric acid secretion in Bangladeshi children. Am J Clin Nutr 2004; 80:149–153.
19. Institute of Medicine, Food and Nutrition Board. Dietary Reference Intakes for Vitamin A, Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel, Silicon, Vanadium, and Zinc
. Washington, DC: National Academy Press; 2000.