点击显示 收起
1 From the US Department of Agriculture, Agricultural Research Service, Grand Forks Human Nutrition Research Center, Grand Forks, ND
2 The US Department of Agriculture, Agricultural Research Service, Northern Plains Area, is an equal opportunity/affirmative action employer, and all agency services are available without discrimination. Mention of a trademark or proprietary product does not constitute a guarantee or warranty of the product by the US Department of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable.
3 Supported by the US Department of Agriculture.
4 Reprints not available. Address correspondence to JR Hunt, USDA-ARS GFHNRC, PO Box 9034, Grand Forks, ND 58202-9034. E-mail: jhunt{at}gfhnrc.ars.usda.gov.
See corresponding editorial on page 799.
ABSTRACT
Background: Research conducted before genotyping was possible suggested that subjects heterozygous for the genetic mutation associated with hemochromatosis absorbed nonheme iron more efficiently than did control subjects when tested with a fortified meal. Heme-iron absorption in these subjects has not been reported.
Objective: We compared the absorption of heme and nonheme iron from minimally or highly fortified test meals between HFE C282Y-heterozygous and wild-type control subjects.
Design: After prospective genotyping of 256 healthy volunteers, 11 C282Y-heterozygous and 12 wild-type control subjects were recruited, and their iron absorption was compared by using a hamburger test meal with or without added iron and ascorbic acid. After retrospective genotyping of 103 participants in previous iron-absorption studies, 5 C282Y-heterozygous subjects were compared with 72 wild-type control subjects.
Results: HFE C282Y-heterozygous subjects did not differ significantly from wild-type control subjects in their absorption of either heme or nonheme iron from minimally or highly fortified test meals. No differences were detected in blood indexes of iron status (including serum ferritin, transferrin saturation, and non-transferrin-bound iron) or in blood lipids or transaminases, but heterozygotes had significantly greater, although normal, fasting glucose concentrations than did wild-type control subjects. Compound heterozygotes (those who had both HFE C282Y and H63D mutations) absorbed more nonheme (but not heme) iron from meals with high (but not low) iron bioavailability.
Conclusions: HFE C282Y-heterozygous subjects did not absorb dietary iron more efficiently, even when foods were highly fortified with iron from ferrous sulfate and ascorbic acid, than did control subjects. Iron fortification of foods should not pose an additional health risk to HFE C282Y heterozygotes.
Key Words: Nutrient-gene interactions nonheme-iron absorption heme-iron absorption diet iron bioavailability iron fortification hemochromatosis serum ferritin non-transferrin-bound iron glucose
INTRODUCTION
Hemochromatosis is a disorder of excessive iron accumulation characterized by increased heme- and nonheme-iron absorption (1). In 1996, specific genetic mutations in the HFE gene were identified (2) in hemochromatosis patients of Northern European origin. Although it is not known how the HFE protein influences iron absorption, it may influence the affinity of plasma transferrin for its receptor, as part of an HFE, ß2-microglobulin, and transferrin receptor complex in the basolateral membrane of developing enterocytes (3). More than 80% of hemochromatosis patients are homozygous for a C282Y mutation in this gene, and a smaller proportion are compound heterozygous for both the C282Y mutation and an H63D mutation (2).
Identification of the C282Y mutation raised questions about the possible health effects of this common mutation for heterozygous carriers, who include 9.54% of non-Hispanic whites in the United States (4). Heterozygosity has been associated with shorter life expectancy in a Danish study (5), but not in other studies (6-8), and it has even been associated with longevity in some Italian women (9, 10). Increased risk of cardiovascular disease (CVD) has been reported by some (11-13) but not other (7, 14, 15) researchers. A greater risk of type 2 diabetes in C282Y heterozygotes found in one study (16) was unconfirmed in another (17). Genetic studies in healthy populations have identified many symptom-free C282Y homozygotes, which shows that the clinical phenotypic penetrance of the mutation is quite low (18-20). Heterozygosity has been inconsistently associated with slightly higher transferrin saturation and serum ferritin concentrations (18, 20-26). Non-transferrin-bound iron (NTBI) was reported in one small study to be greater in heterozygous subjects than in wild-type control subjects, but there was no difference in serum ferritin or transferrin saturation between these 2 groups (27).
Before C282Y genotyping was available, iron absorption by heterozygous carriers of hemochromatosis was tested by identifying as heterozygotes the children of patients with hemochromatosis or two siblings who shared a single HLA haplotype (1). No difference was found between heterozygous and control subjects in the absorption of nonheme iron from a hamburger meal (1). However, when the meal was fortified with 20 mg Fe (as ferrous sulfate) and 100 mg ascorbic acid, the heterozygous subjects absorbed 2.5 times as much nonheme iron as the control subjects absorbed. Because of the high frequency of the HFE C282Y mutation, such an increase in absorption may have public health implications for food fortification policies. The objective of the current study was to compare iron absorption in heterozygotes identified by the C282Y mutation with that in wild-type control subjects by feeding test meals similar to those used by Lynch et al (1). Secondary objectives were to examine iron absorption related to other HFE C282Y and H63D polymorphisms and to test for differences in clinical chemistry, including an increase in NTBI in subjects who were heterozygous for the C282Y mutation.
SUBJECTS AND METHODS
General protocol
The absorption of dietary iron in healthy subjects who are heterozygous for the HFE C282Y mutation and in wild-type control subjects was compared by using 2 sets of data. First, 256 healthy volunteers were prospectively genotyped (tested for both the HFE C282Y and H63D mutations), which led to the enrollment of 11 HFE C282Y-heterozygous and 12 wild-type control subjects for tests of heme- and nonheme-iron absorption from test meals prepared with and without extra fortification with iron (as ferrous sulfate) and ascorbic acid. Second, 103 participants in previous iron-absorption studies were retrospectively genotyped, which allowed the comparison of 5 subjects who were heterozygous for the C282Y mutation, 2 compound-heterozygous subjects, and subjects with other HFE genotypes with wild-type control subjects who had neither mutation.
Subjects
Healthy men and nonpregnant women were recruited through public advertising and selected after an interview and blood analysis helped ascertain that they had no apparent underlying disease and had blood hemoglobin concentrations of 120 g/L. Although race was not designated as a selection criterion, all volunteers were white and non-Hispanic. The participants gave written informed consent. The University of North Dakota's Radioactive Drug Research Committee and Institutional Review Board and the US Department of Agriculture's Human Studies Review and Radiological Safety committees reviewed and approved this human study.
Dietary treatments
Prospective genotyping study
Iron absorption was tested from weighed standard or fortified test meals similar to those used by Lynch et al (1). The standard meal consisted of ground beef (113 g), a refined-wheat bun (53 g), French-fried potatoes (68 g), tomato ketchup (20 g), and a vanilla milkshake (150 g). The refined-wheat bun was commercially enriched with iron according to common US practice for refined-wheat products (20 mg iron/460 g flour). The standard meal contained 906 kcal and 5.1 mg Fe, of which 1.2 mg was in the heme form. The fortified meal was similar, but it omitted the French-fried potatoes and milkshake and added 200 g orange juice supplemented with 100 mg ascorbic acid and 20 mg Fe as ferrous sulfate. The orange juice was prepared and mixed with supplements on the same morning that the meals were served. The fortified meal contained 590 kcal and 25.1 mg iron, of which 1.2 mg was in the heme form.
Retrospective genotyping study
Subjects with mutations of interest had previously participated in iron-absorption studies with 3 different sets of dietary treatments. The subjects' iron absorption was tested 1 from a 2-d menu low in dietary iron bioavailability (no meat, small amounts of poultry and fish, and 20 mg ascorbic acid at each meal, with considerable black tea and phytic acid-containing foods) before and after 10 wk of consumption of the same diet (28, 29); 2 from a higher-bioavailability test meal containing beef, refined-wheat roll, French-fried potatoes, and milkshake before and after 12 wk of daily supplementation with a placebo (30); and 3 from 2 test meals of beef, refined-wheat roll, French-fried potatoes, and milkshake, with and without cheese, which were followed by a complete gastrointestinal lavage procedure 8 h after each meal (31). No C282Y mutations were identified in subjects formerly treated with a high-iron-bioavailability diet (28, 29) or with iron supplements (30). Subjects fasted for 8 h before and 4 h after single test meals and ate only the food provided when tested with complete menus for 2 d.
Iron-absorption measurements
Heme- and nonheme-iron absorption was measured by extrinsically labeling the foods (in the final stages of preparation) with 18.5 or 37 kBq 55Fe-labeled hemoglobin and 37 kBq 59FeCl3. Radiolabeled hemoglobin was obtained by intravenously injecting 74 MBq 55Fe into an iron-deficient, pathogen-free rabbit, bleeding the animal 2 wk later, isolating the red cells by centrifugation, and removing the stroma by lysing and further centrifugation (32). For the low-bioavailability diet involving 2 d of meals, the isotopes were added to the diet in proportion to the heme- and nonheme-iron contents of the meals, which yielded constant specific activities (ratios of 55Fe to dietary heme iron and of 59Fe to nonheme iron) for all 6 meals. The tracers were added to the foods that were the best sources of that form of iron in each meal. Meat, poultry, and fish dishes were cooked in advance, cooled, radiolabeled, and then minimally reheated in the microwave just before service. All labeled meals were consumed at the research center.
Nonheme-iron absorption was determined by whole-body scintillation counting that detected only the gamma-emitting 59Fe radioisotope. The custom-made whole-body counter used 32 crystal NaI(Tl) detectors, each 10 cm x 10 cm x 41 cm, arranged in 2 planes above and below the participant, who lay supine. The initial total-body activity was determined before any unabsorbed isotope was excreted but 1 h after the subjects finished the single meals (or 2 meals of the 2-d menu, with calculated adjustments for the fraction of the total activity administered). The percentage of nonheme-iron absorption was measured as the portion of initial whole-body activity that remained after 2 wk. Heme-iron absorption was ascertained by measuring the blood retention of 55Fe (together with 59Fe; 33) after 2 wk and by estimating total blood volume on the bases of body height, sex, and weight (34, 35) and of the assumption that 80% of the absorbed heme iron was incorporated into blood (36). All isotope determinations included corrections for physical decay and background activity measured 12 d before isotope administration.
Genotyping
Genomic DNA was extracted from buccal smears (for retrospective genotyping) or blood (prospective genotyping study) by using a DNA isolation kit (Qiagen, Valencia, CA). HFE C282Y and H63D mutations were ascertained by using a two-round polymerase chain reaction (PCR; semi-nested primers) approach (37, 38) and enzymatic digestion of the products similar to that in previous reports (2, 39). Briefly, genomic DNA was used as a first-round PCR DNA template, and then 0.5 µL of the first-round PCR product was used as the second-round PCR DNA template. As described by Feder et al (2), for the C282Y mutation, the first-round PCR primers were 5'-TGGCAAGGGTAAACAGATCC-3' and 5'-CCAATGAACAAGATGACGACAA-3', and the second-round PCR primers were 5'-TGGCAAGGGTAAACAGATCC-3' and 5'-CTCAGGCACTCCTCTCAACC-3'. For the H63D mutation, the first-round PCR primers were 5'-ACATGGTTAAGGCCTGTTGC-3' and 5'-CCTTGCTGTGGTTGTGATTTTC-3', and the second-round PCR primers were 5'-ACATGGTTAAGGCCTGTTGC-3' and 5'-GCCACATCTGGCTTGAAATT-3' (2). DNA spanning C282Y (560 base pairs) and H63D (295 base pairs) mutations was amplified in the first round of PCR. The C282Y and H63D mutations were detected by amplification of 389 and 208 base pairs, respectively, in the second-round PCR.
Negative controls were included in each PCR. The PCR conditions were as follows: 94 °C for 2 min; then 25 cycles of 94 °C for 45 s, 60 °C for 45 s, and 72 °C for 1 min; and, finally, 72 °C for 2 min. Aliquots of second-round PCR products were digested with Rsa I and Mbo I endonuclease restriction enzymes to identify the C282Y and H63D mutations, respectively, as described previously (39). The enzymatic digestion product was subsequently separated by electrophoresis in 3% agarose 2000 (Gibco-BRL, Carlsbad, CA) gels. Each subject's genotype was determined independently by 2 staff members.
Clinical chemistry
Hemoglobin was measured by using a Celldyne 3500 System (Abbott Laboratories, Abbott Park, IL). Serum iron was determined colorimetrically by using a Cobas Fara Chemistry Analyzer (Hoffman-LaRoche Inc, Nutley, NJ) with a commercial chromogen (Ferene; Raichem Division of Hemagen Diagnostics, San Diego). Iron-binding capacity was similarly measured after the addition of a known amount of ferrous iron to the serum sample under alkaline conditions. The percentage of transferrin saturation was calculated from serum iron and total-iron-binding capacity. Serum ferritin concentrations were measured with the use of an enzyme-linked immunosorbent assay (Abbott Laboratories; or Diagnostic Products Corporation, Los Angeles). Serum NTBI was measured on subsamples of blood obtained at 3 different times from each subject and was detected colorimetrically after the addition of nitriloacetic acid and treatment with cobalt (40).
Colorimetric methods were used to determine glucose, aspartate aminotransferase (AST), -glutamyl transpeptidase (GGT), triacylglycerol (41), total cholesterol (42), and HDL cholesterol (42) from fasting blood serum (all with the use of the Cobas Fara Chemistry Analyzer). VLDL cholesterol was estimated from triacylglycerol (43), and LDL cholesterol was calculated by difference (LDL = Total HDL VLDL).
Statistical analysis
Iron-absorption and serum ferritin data were logarithmically transformed, and geometric means are reported. Power curves were used to plot relations between iron absorption and serum ferritin concentrations (these relations are linear when the data are logarithmically transformed).
For statistical tests of the effect of genotype, the known inverse relation between iron absorption and serum ferritin was used to normalize iron-absorption values to an arbitrary serum ferritin concentration of 40 µg/L, as follows:
RESULTS
Prospective genotyping data set
The genotyping results from this convenience sampling in Grand Forks, ND, did not differ significantly (by chi-square analysis) from a representative sampling of the non-Hispanic white US population (4; Table 1). Of 256 subjects prospectively screened, 20 (8%) were C282Y heterozygous, without the H63D mutation. Of these 20 subjects, 12 agreed to participate in the iron-absorption study. Reasons for nonparticipation included anemia (blood hemoglobin < 120 g/L; n = 2); pregnancy (n = 1); lack of willingness to follow experimental protocol (n = 2); and loss to follow-up contact (n = 3). One subject was subsequently eliminated because of elevated hematocrit and referral for possible polycythemia, which left 11 subjects (8 females and 3 males) who were C282Y heterozygous. Twelve control subjects (wild-type, with neither the HFE C282Y nor H63D mutation; 9 females and 3 males) of roughly similar serum ferritin concentrations, age, sex, and body mass index (BMI; in kg/m2) were selected. These control subjects were not paired with heterozygous subjects for analysis because the initial serum ferritin measurements were later judged to be from an unreliable commercial source and because several months had elapsed between the initial serum ferritin measurements and the iron-absorption experiment. When it was concluded that the initial serum ferritin analyses were unreliable, this measurement was repeated with a different commercial kit, but only in those subjects who had participated in the iron-absorption measurements.
View this table:
TABLE 1. Comparison of genotyping results in the Grand Forks, ND, area with results representative of the non-Hispanic white US population (NHANES III)1
Iron absorption was not significantly affected by HFE C282Y heterozygosity. As expected, absorption of both forms of iron was inversely related to serum ferritin concentrations (Figure 1). After normalization of iron absorption to a serum ferritin concentration of 40 µg/L, there was no significant difference (by analysis of variance) in the absorption of either heme or nonheme iron by subjects with the 2 different genotypes, whether absorption was tested with the standard meal or with the meal fortified with additional iron (as ferrous sulfate) and ascorbic acid (Figure 2). Normalized absorption of nonheme iron in the heterozygous and wild-type subjects was, respectively, 10.9% and 7.5% from the standard meal and 7.7% and 7.5% from the fortified test meal. Normalized absorption of heme iron in the 2 groups was, respectively, 36% and 33% from the standard meal and 29% and 32% from the fortified test meal.
FIGURE 1.. Heme- and nonheme-iron absorption from a hamburger meal fortified with additional iron as ferrous sulfate and ascorbic acid in subjects who were heterozygous for the C282Y mutation (; n = 11) and wild-type control subjects (; n = 12). The 2 groups of subjects absorbed nonheme and heme iron similarly from the hamburger meal. For nonheme iron, regression coefficients were R2 = 0.85 (P < 0.01) in the C282Y-heterozygous subjects and R2 = 0.43 (P < 0.05) in the wild-type control subjects. For heme iron, regression coefficients were R2 = 0.32 (P < 0.05) in the C282Y-heterozygous subjects and R2 = 0.12 (NS) in the wild-type control subjects.
FIGURE 2.. Absorption of heme and nonheme iron as tested with the standard meal and the fortified meal in C282Y-heterozygous (; n = 11) or wild-type control (; n = 12) subjects. When absorption was normalized to a serum ferritin of 40 µg/L (see text), there was no significant difference (P > 0.05) ikn nonheme- and heme-iron absorption between the 2 groups of subjects, whether tested with the standard or the fortified meal.
Several clinical chemistry measurements were compared between the C282Y-heterozygous (n = 11) and wild-type control (n = 12) subjects (Table 2). The control subjects had been selected for serum ferritin measurements, age, sex, and BMI that were similar to those of the C282Y-heterozygous subjects at the time of initial genotyping (several months before). At the time of the iron-absorption measurements, the 2 groups did not differ significantly, according to t tests, in any of the iron status measurements (Table 2). These included NTBI concentration, which was 0.65 ± 0.12 and 0.68 ± 0.08 µmol/L for the heterozygous and control subjects, respectively. No differences were found in serum lipid measurements or transaminases (AST or GGT). Of the clinical chemistry measurements conducted, only fasting serum glucose concentrations differed between the 2 groups (4.6 ± 0.1 and 4.1 ± 0.1 mmol/L for heterozygous and control subjects, respectively; P < 0.02 by t test; Table 2).
View this table:
TABLE 2. Comparison of age, BMI, and clinical chemistry results between prospectively genotyped C282Y-heterozygous and wild-type control subjects1
Retrospective genotyping data set
Only 5 subjects who were heterozygous for the C282Y mutation were identified from the retrospective genotyping (Table 1). These subjects did not absorb either heme or nonheme iron more efficiently than did the wild-type control subjects who had neither mutation. This conclusion was based on t tests (C282Y-heterozygous compared with wild-type control subjects) when the absorption data within a single experiment were normalized to a serum ferritin concentration of 40 µg/L. Iron absorption had been measured twice in each subject, and the data were highly reproducible between measurements (eg, the absorption results at week 10, shown in Figure 3, were similar to those at week 0).
FIGURE 3.. Heme- and nonheme-iron absorption in wild-type control, C282Y-heterozygous, H63D-heterozygous, and compound-heterozygous subjects after 10 wk of consuming a low-bioavailability diet (28). Retrospective genotyping of subjects whose iron absorption had been tested after 10 wk provided data consistent with no difference in either nonheme- or heme-iron absorption between the C282Y-heterozygous and wild-type subjects. The iron absorbed from this low-bioavailability diet did not differ significantly between the single compound-heterozygous subject and the wild-type control subjects. The trend lines fit only the data of the 15 wild-type subjects (nonheme iron: R2 = 0.61, P < 0.01; heme iron: R2 = 0.48, P < 0.01). Similar measurements at week 0 gave similar results.
Two subjects were compound heterozygous for both the C282Y and H63D mutations, and their nonheme-iron absorption depended on the bioavailability of the test diet. The nonheme-iron absorption in one subject was substantially inhibited by the low-iron-bioavailability diet, which is similar to the results in the wild-type subjects (Figure 3). When the absorption data were normalized to a serum ferritin concentration of 40 µg/L (not shown), nonheme-iron absorption from the low-bioavailability diet in this compound-heterozygous subject did not differ significantly from that in the wild-type control subjects. The other compound-heterozygous subject was tested with high-bioavailability meals and appeared to absorb nonheme iron from both a cheeseburger meal (Figure 4) and a similar hamburger meal (not shown) more efficiently than did the wild-type control subjects. When absorption measurements were normalized to a serum ferritin concentration of 40 µg/L, the compound-heterozygous subject absorbed significantly more of the nonheme iron from these high-bioavailability meals than did the 6 wild-type control subjects: 24% compared with 7.5 ± 4.7%, respectively, for the cheeseburger meal (P < 0.01; t test) and 22% compared with 7.9 ± 4.8%, respectively, for the hamburger meal (P < 0.02; t test). Heme-iron absorption was not affected by the compound-heterozygous condition (Figures 3 and 4). There was no indication of any difference in iron absorption between the wild-type control subjects and the single subject who was homozygous for the H63D mutation (Figure 4) or the subjects who were heterozygous for that mutation (Figures 3 and 4).
FIGURE 4.. Nonheme- and heme-iron absorption from a high-bioavailability cheeseburger meal, followed by gastrointestinal lavage (31), in wild-type control, H63D-homozygous, H63D-heterozygous, and compound-heterozygous subjects. Retrospective genotyping of these subjects indicated greater iron absorption from a high-bioavailability meal in the single compound-heterozygous subject than in the wild-type control subjects (see t test results in text). Iron absorption in the subject who was homozygous for the H63D mutation did not differ from that in the wild-type control subjects. The trend lines fit only the data of the wild-type subjects. Although these relations were significant in Figure 3 and in larger studies (1, 30), neither regression was significant in these 6 subjects. Nonheme iron: R2 = 0.33; heme iron: R2 = 0.21.
As indicated in Subjects and Methods, heme-iron absorption in this study was determined by measuring blood 55Fe 2 wk after isotope administration and assuming 80% erythrocyte incorporation of the absorbed isotope. Similar overall results were obtained if heme-iron absorption was determined without an assumption of 80% erythrocyte incorporation but with the assumption that the erythrocyte incorporation of 55Fe from administered heme iron was identical to that of the nonheme 59Fe, which was ascertained from blood and whole-body measurements. However, this latter calculation method resulted in an apparent outlier: one woman with a low serum ferritin concentration (9 µg/L) who was C282Y heterozygous had very low measurements of erythrocyte 59Fe incorporation (56% and 58% at 0 and 10 wk, respectively; calculated by dividing the fractional blood retention by the fractional whole-body retention of 59Fe). The use of this low erythrocyte incorporation of the nonheme radiotracer resulted in unusually high estimations of heme-iron absorption (78% and 58% at 0 and 10 wk, respectively). If erythrocyte incorporation of 80% was used in the absorption calculation, the estimated heme-iron absorption in this female subject was not unusually high (55% and 42% at 0 and 10 wk, respectively; the latter value is shown in Figure 3
DISCUSSION
The results indicate that subjects who are heterozygous for the HFE C282Y mutation do not absorb dietary iron any more efficiently than do wild-type control subjects. These results confirm the finding of Lynch et al (1) that there is no difference between these groups in nonheme-iron absorption from a standard hamburger meal, but they are in contrast to the same group's finding that heterozygous subjects absorb more nonheme iron from a meal fortified with additional iron (from ferrous sulfate) and ascorbic acid than do wild-type control subjects. Prospective power analyses had predicted that 1015 subjects from each genotype would provide 90% power to detect an interaction (genotype x level of fortification) of the magnitude observed by Lynch et al (1)ie, similar nonheme-iron absorption from the minimally fortified meal, but more than twice as much nonheme-iron absorption from the fortified meal. However, the present data gave no indication of this interaction (Figure 2). This difference in results is likely related to the method of genotyping. Because the heterozygous subjects of Lynch et al were identified by their relation to hemochromatosis patients and by HLA assessment, they were more likely 1 to have additional unidentified genetic, ethnic, or lifestyle characteristics that resulted in a phenotypic expression of excessive iron accumulation, or 2 to be mistakenly typed as heterozygous when some may have been compound heterozygous or C282Y homozygous. Moirand et al (48) found that, of subjects whose heterozygous status was inferred by HLA assessment, 3% were C282Y homozygous, 20% were compound heterozygous, and only 70% were C282Y heterozygous. Such misclassification when more sensitive genotyping was not possible may have accounted for the positive results of Lynch et al (1).
This is the first report of heme-iron absorption by heterozygous subjects. None of the tested C282Y and H63D polymorphisms influenced heme-iron absorption in this study, including the 2 measurements in each of 2 subjects who were compound heterozygous.
The consistently elevated nonheme-iron absorption in the compound-heterozygous subject whose iron absorption was tested with a high-bioavailability meal is consistent with the increased risk of clinical hemochromatosis experienced by compound-heterozygous subjects (2). However, the consistent lack of elevation in nonheme-iron absorption in another compound-heterozygous subject tested with a low-iron-bioavailability diet suggests that, even in those genetically prone to absorb iron more efficiently, the iron must be available to the intestinal mucosa in a reduced, soluble form. Although limited conclusions can be drawn from these 2 subjects, our findings are consistent with an interaction between genotype and diet, which suggests that compound-heterozygous subjects absorb nonheme iron more efficiently than do wild-type subjects, but only if the dietary iron is at least moderately bioavailable.
The present negative results for heme- and nonheme-iron absorption by C282Y-heterozygous subjects are consistent with the minimal differences in iron status indexes such as serum ferritin and transferrin saturation observed in this study (Table 2) and other, much larger studies (20-24). The iron status indexes of the C282Y-heterozygous subjects we identified by recruitment from the general population covered a normal range and even included 2 women with iron deficiency anemia, although that condition eliminated them from the absorption testing. The present data do not confirm the finding in a report by de Valk et al (27) of more NTBI in 22 heterozygous subjects (0.51 µmol/L; range: 0.75 to 2.75 µmol/L) than in 17 control subjects (0.30 µmol/L; range: 1.50 to 0.60 µmol/L). Both our study and that of Valk et al included a small number of subjects. The range of measurements was narrower in the present study (Table 2) and, unlike the measurements reported by de Valk et al (who did not report sample variation), did not include negative values. On the basis of the variation observed in the present study, we had 99% statistical power to detect the difference of 0.8 µmol/L in NTBI that was reported by de Valk et al (27), but we did not detect a difference (Table 2). The difference observed by de Valk et al may have been affected by the method of identification of heterozygous subjects. If subjects were identified from among the relatives of patients with hemochromatosis, rather than from the general population, they may have had additional, predisposing genetic factors that influenced the penetrance of the HFE C282Y mutation and the resulting control of iron metabolism.
The finding that the C282Y-heterozygous subjects had significantly higher fasting serum glucose concentrations than did control subjects (Table 2) was unexpected. Although heterozygous subjects have been reported to have a greater risk of type 2 diabetes than do control subjects (16), this was refuted by a recent meta-analysis of HFE mutation frequencies in subjects with type 2 diabetes and in control subjects (17). Beutler et al (18) reported no greater incidence of high fasting blood sugar (>6.7 mmol/L) among either C282Y-homozygous, C282Y-heterozygous, or wild-type control subjects, which is consistent with the results of the present study (Table 2). The fasting glucose concentrations that we found should be interpreted with caution, by recognizing that this one minor difference in blood glucose arose from multiple t test comparisons (Table 2).
In conclusion, these results indicate no significant difference in heme- or nonheme-iron absorption by heterozygous carriers of the HFE C282Y genetic mutation associated with hemochromatosis, whether tested by using common meals or meals highly fortified with iron as ferrous sulfate and ascorbic acid. The results are consistent with greater absorption of nonheme but not heme iron from meals with high iron bioavailability in subjects who are compound heterozygous for the C282Y and H63D mutations. Although a substantial proportion of the US population is heterozygous for the C282Y mutation, the present results do not suggest any greater risk to these subjects due to consumption of foods that are highly fortified with this highly bioavailable form of iron or with ascorbic acid.
ACKNOWLEDGMENTS
We gratefully acknowledge the contributions of members of our human studies research team, particularly Carol Ann Zito, who conducted the blood radioiron analyses and, together with Melissa Phelps and Cheryl Stjern, performed the genotyping assays. In addition, Emily J Nielsen managed the volunteer recruitment and scheduling; Bonita Hoverson supervised the planning and service of the controlled diets; Sandy K Gallagher supervised the clinical laboratory analyses; Glenn I Lykken designed and consulted in the use of the whole body counter; and LuAnn K Johnson performed the statistical analyses. We are especially grateful to the volunteers for their dedication and service.
JRH conceived and designed the study, supervised the iron-absorption measurements, and wrote the manuscript. HZ supervised genotyping measurements and contributed to the manuscript. Both were directly funded by the US Department of Agriculture Agricultural Research Service and had no conflict of interest.
REFERENCES
Related articles in AJCN: