Absorption of zinc from wheat products fortified with iron and either zinc sulfate or zinc oxide 1 3

Transcription

1 Absorption of zinc from wheat products fortified with iron and either zinc sulfate or zinc oxide 1 3 Daniel López de Romaña, Bo Lönnerdal, and Kenneth H Brown ABSTRACT Background: Several chemical forms of zinc have been proposed for food fortification, but information is needed on their absorption from common cereals having varied phytate content. Objective: The goal was to measure zinc absorption from wheat products fortified with iron sulfate and either zinc sulfate or zinc oxide. Design: Adult volunteers received either low-phytate bread (n = 11) or higher-phytate porridge (n = 11) once weekly on 2 or 3 occasions. The foods were fortified with 1 of the 2 zinc salts (60 mg elemental Zn/kg wheat flour) during week 1 and with the other during week 2, in random order. 65 Zn in the same chemical form as the fortificant was incorporated in each food to assess zinc absorption with the use of whole-body counting. The porridge group received an additional test meal fortified with zinc oxide during week 3, but the 65 Zn tracer was given as an oral solution of 65 ZnCl 2. Results: Zinc absorption from bread (13.8%; 95% CI: 11.8%, 16.2%) was significantly (P < 0.001) greater than from porridge (6.4%; 5.5%, 7.6%), presumably because of the greater phytate content of the porridge. With control for food type, there were no significant differences in zinc absorption from meals fortified with zinc sulfate or zinc oxide (P =0.24). When the porridge was fortified with zinc oxide and labeled with 65 ZnCl 2, absorption of the tracer (8.9%; 7.1%, 11.0%) was significantly (P = 0.007) greater than when 65 ZnO was incorporated in the porridge (5.6%; 4.5%, 6.9%). Conclusions: Either zinc oxide or zinc sulfate can be used to fortify wheat products consumed by presumably healthy persons. Isotopic tracers used to assess the absorption of mineral fortificants should have the same chemical form as the fortificant. Am J Clin Nutr 2003;78: KEY WORDS radioisotopes Iron, zinc, wheat, fortification, zinc absorption, INTRODUCTION Recent analyses of national food supplies suggest that the diets of 20% of the global population contain inadequate amounts of zinc in relation to theoretical requirements (1). Zinc deficiency has been associated with poor growth (2), depressed immune function (3), increased susceptibility to and severity of infection (4, 5), adverse outcomes of pregnancy (6), and neurobehavioral abnormalities (7). In many developing countries, zinc deficiency is due to the low consumption of animal source foods, which are rich in zinc, and a high intake of cereals and legumes, which contain substantial amounts of phytate (myo-inositol hexaphosphate), a compound known to inhibit zinc absorption (8). One strategy for controlling zinc deficiency is fortification of an appropriate food vehicle with an absorbable zinc salt. Five zinc compounds are currently listed as generally recognized as safe, or GRAS, by the US Food and Drug Administration: zinc sulfate, zinc chloride, zinc gluconate, zinc oxide, and zinc stearate. At present, little information is available on the bioavailability of these zinc compounds in fortified foods, and there is no consensus regarding the most appropriate form to use in fortification programs. Absorbability of the different zinc salts presumably depends on their solubility in aqueous solution, with zinc sulfate and zinc chloride being very soluble and zinc oxide being almost insoluble at neutral ph. This difference is particularly important when gastric acid output is low (9), which may occur more frequently in malnourished children in developing countries because of the effects of malnutrition (10) and Helicobacter pylori infection on gastric acid production (11). Of the GRAS zinc salts, zinc sulfate and zinc oxide are the best prospects for fortification programs because of their relatively low cost. Of the 2, zinc oxide is considerably cheaper and more stable than is zinc sulfate, but zinc sulfate may be better absorbed because of its greater solubility at a neutral ph (12). Several studies have shown that whole-body counting of the zinc radioisotope 65 Zn added to food as an extrinsic label can be used to measure zinc absorption from food products provided to adults (13 16). On theoretical grounds, it would seem appropriate to provide the zinc tracer in the same chemical and physical form as the fortificant if the purpose of the study is to quantify the absorption of the fortificant. This might be particularly important for a fortificant such as zinc oxide because of its poor solubility in water. The present studies were completed to compare the absorption of zinc from relatively low- or high-phytate wheat products fortified with zinc sulfate or zinc oxide. We also compared zinc 1 From the Program in International Nutrition, Department of Nutrition, University of California, Davis. 2 Supported by The Bill and Melinda Gates Foundation. 3 Reprints not available. Address correspondence to D López de Romaña, Instituto de Investigación Nutricional, Avenida La Molina 685, La Molina, Lima, Perú. dromana@iin.sld.pe. Received October 15, Accepted for publication February 28, Am J Clin Nutr 2003;78: Printed in USA American Society for Clinical Nutrition 279

2 280 LÓPEZ DE ROMAÑA ET AL absorption from porridge that was fortified with zinc oxide and labeled with either zinc oxide or zinc chloride to determine whether the results were affected by the chemical form of the tracer. SUBJECTS AND METHODS Subjects Twenty-two adult male volunteers were recruited from the University of California, Davis, student population by placing advertisements on campus. Subjects were admitted into the study if they were reportedly healthy and not habitually taking any mineral supplements. The sample size estimate was based on withinsubject comparisons of the fractional absorption of zinc by type of fortificant. Using results of a previous study by Sandström et al (16), which found a 4% SD in the fractional absorption of zinc, we estimated that a sample size of 11 subjects (for each meal type) would be adequate to detect a 5% absolute difference in within-subject fractional absorption of zinc by type of fortificant (two-sided, paired t test; level of significance = 0.05, power = 0.80), assuming a worst-case scenario of no intraindividual correlation. The protocol was approved by the Human Subjects Committee and the Radioactive Drug Research Committee at the University of California, Davis. Preparation of isotopes and fortificants Isotopes and fortificants were prepared by the Brookhaven National Laboratory. The 65 ZnSO 4 solution and 65 ZnO in powder form were both synthesized from 65 ZnCl 2. Specifically, 65 ZnSO 4 was prepared by dissolving 403 mg unlabeled zinc metal in 4 ml concentrated HCl, reducing the volume on a hotplate, and adding ml 65 ZnCl 2 and 2 ml concentrated H 2 SO 4. This solution was again reduced in volume, and concentrated H 2 SO 4 (2 ml) was added 2 additional times to remove chloride, and the mixture was taken to dryness. The resulting 65 ZnSO 4 crystals were reconstituted in 10 ml water and buffered with 164 mg sodium acetate before the ph was adjusted to 2 3 with sodium hydroxide and the solution diluted to a final volume of 20 ml. 65 ZnO was prepared from the labeled 65 ZnCl 2 solution, prepared as above. After the solution was dried, the resulting white crystals were dissolved in 10 ml water, and concentrated ammonium hydroxide was added until the resulting solution reached a ph of 8.4 to precipitate zinc hydroxide. The supernatant fluid was removed after centrifugation (2000 g, 5 min, room temperature), and the precipitate was rinsed twice more with water. The resulting zinc hydroxide crystals were then collected on number 2 filter paper and converted to zinc oxide by heating in a porcelain boat in a 400 C furnace. Unlabeled zinc sulfate and zinc oxide were synthesized from cold zinc chloride by following the same procedures as for the radioisotope preparations. Finally, individual 65 ZnCl 2 doses were prepared by dilution from a 65 ZnCl 2 stock solution. Preparation and labeling of the diets Loaves of bread were prepared in a bread machine from unenriched, white-wheat (70% extraction) flour (500 g), water (300 ml), salt (10 g), sugar (10 g), margarine (10 g), and yeast (15 g). Subjects were served 80-g portions of bread, which contained 50 g wheat flour. Porridge was prepared by cooking wheat farina (85% extraction) flour (500 g), water (3 L), and salt (15 g). The subjects were served 412-g portions, which also contained 50 g wheat flour, to which they could add 8.4 g (2 tsp) sugar. The bread and porridge meals were fortified per kilogram of wheat flour with 30 mg Fe as iron sulfate and 60 mg Zn as either zinc sulfate or zinc oxide. The amount of iron added to the flour is that which is currently used in many iron fortification programs in developing countries. For both the breads and the porridges, the ferrous sulfate, zinc sulfate, and 0.6 mg 65 ZnSO 4 per dose (0.5 Ci per dose) were added directly to the water during the preparation of the dough or porridge, and the zinc oxide powder and 0.6 mg 65 ZnO per dose (0.5 Ci per dose) were added directly to the flours before mixing. Study design At baseline, each subject s height and weight were measured and a blood sample was collected for analysis of hemoglobin, serum ferritin, and plasma zinc concentrations. The volunteers were then randomly divided into 2 groups, and those in each group received either the bread meals (n = 11) or the porridge meals (n = 11) once a week on 2 occasions. The respective meals were fortified with 1 of the 2 zinc salts during week 1 and with the other during week 2, in random order. The subjects who received porridge continued in the study for 1 more week so that their zinc absorption could be assessed when the porridge was fortified with zinc oxide and the zinc tracer was provided as a separate oral solution of ZnCl 2 rather than added directly to the food in the same chemical form as the fortificant. Zinc absorption was estimated by using the method of Arvidsson et al (13). Each week, the empty chamber of the whole-body counter (Radiobiology Laboratory, University of California, Davis) was assayed to determine the background level of radioactivity. The whole-body counter was equipped with two cm sodium iodide crystals and a multichannel analyzer (ND-66; Nuclear Data, Schaumburg, IL). Whole-body radioactivity was then counted for 15 min before and immediately after each test meal was consumed, and the difference between the pre- and postprandial counts was used as the total dose of 65 Zn consumed. The counting was then repeated 7 d later (on day 8) to assess retention of the previously consumed dose. These latter measurements also served as baseline values for the second diet period for each subject. On day 8 the procedure was repeated, with the subjects consuming the corresponding meal. On day 15 of the study, the whole procedure was repeated again for the subjects who received the porridge diet to assess the effect of the chemical form of the tracer. Absorption of 65 Zn on days 8 and 15 for both groups and on day 22 for the porridge group was estimated by correcting for endogenous excretion from days 0 8, 9 15, and 15 22, by using a previously published mean retention function (R = 0.15e 0.08t e t, where R = retention and t=the time since the oral dose of the isotope), which was developed from repeated measurements of whole-body retention in a group of healthy subjects who received an intravenous injection of 65 Zn (13). Given that the whole-body counter technique measures the percentage of zinc that is retained in the body on the day that the counting is done, this correction for excretion of endogenous zinc provides an estimate of the percentage of zinc that was originally absorbed. Statistical analysis Data analysis was performed with SAS software (SAS for WINDOWS, release 8.1; SAS Institute Inc, Cary, NC). The major outcome variable, zinc absorption (as a percentage of intake), was log transformed to conform more closely to a normal distribution.

3 ABSORPTION OF ZINC FORTIFICANTS 281 TABLE 1 Baseline characteristics of the study subjects 1 Value Age (y) 26.2 ± 9.4 ( ) Weight (kg) 80.5 ± 15.2 ( ) Height (cm) ± 6.4 ( ) Hemoglobin (g/dl) 15.8 ± 0.6 ( ) Serum ferritin ( g/l) (76, 153) Plasma zinc ( g/dl) ± 29.3 ( ) 1 x ± SD; range in parentheses. n = Geometric x ; 95% CI in parentheses. The 2 diets and 2 types of fortificants were compared with each other by using two-factor, repeated-measures analysis of variance, with diet as a between-subject factor, the form of zinc as a withinsubject factor, and an interaction term. In the porridge group, the absorption of zinc from the porridge fortified with zinc oxide and labeled with zinc oxide was compared with absorption from the same porridge labeled with zinc chloride by using a paired t test. The relation between mean log zinc absorption and log serum ferritin was examined by using Pearson s correlation. RESULTS Subjects The baseline characteristics of the study subjects are shown in Table 1. Mean values for all indicators of iron and zinc status were in the normal range. All subjects were reportedly healthy at the beginning of the study and remained so throughout the study. Diets The composition of the study diets, as estimated from the US Department of Agriculture food-composition table (17) and published data on the phytate content of foods (18), is presented in Table 2. All meals were fortified with 7.4 mg ferrous sulfate, which provided 1.5 mg elemental iron. The meals fortified with zinc sulfate contained 13.2 mg of the fortificant, and those fortified with zinc oxide contained 4.5 mg of the fortificant. Samples of all final diets were analyzed for their iron and zinc contents by atomic absorption spectrometry. The breads fortified with zinc sulfate TABLE 2 Composition of the test meals 1 Food components Amount Energy Protein Phytate g kcal g mg Bread Wheat flour Margarine Sugar Yeast Water Total Porridge Farina flour Sugar Water Total Estimated from the US Department of Agriculture food-composition tables (Internet: FIGURE 1. Geometric mean (± 1 SD) absorption of zinc by type of meal and zinc fortificant. Zinc absorption from bread was significantly greater than from porridge (P < 0.001, repeated-measures ANOVA). There was no significant difference by fortificant (P = 0.24), and there was no significant interaction (P = 0.28). contained 1.9 mg elemental iron and 3.7 mg elemental zinc per serving, and the breads fortified with zinc oxide contained 1.5 mg elemental iron and 3.1 mg elemental zinc per serving. Their molar ratios of phytate to zinc were 0.5. The porridges fortified with zinc sulfate had 2.7 mg elemental iron and 3.1 mg elemental zinc per serving, and those fortified with zinc oxide had 2.9 mg elemental iron and 3.3 mg elemental zinc per serving. The molar ratios of phytate to zinc of the porridges were 12. Zinc absorption All calculations were done by using geometric means for zinc absorption to account for skewed data. The mean absorption of zinc from the breads was significantly greater than from the porridges (Figure 1). When the type of diet was controlled for, there were no significant differences in mean zinc absorption between meals fortified with zinc sulfate and those fortified with zinc oxide (P = 0.24), and there was no significant interaction between diet and type of fortificant (P = 0.28). Finally, the mean absorption of zinc from the porridge fortified with zinc oxide and labeled with a simultaneously administered oral solution of zinc chloride was significantly greater than absorption from the porridge that was fortified with zinc oxide and labeled with the same zinc compound (Figure 2). When diet type was controlled for, there was no significant correlation between an individual s serum ferritin concentration and mean zinc absorption (P = 0.26). DISCUSSION The present studies were carried out to compare zinc absorption from test meals fortified with 2 different zinc salts: zinc sulfate and zinc oxide. The test meals were prepared from iron- and zinc-fortified wheat products (bread or porridge) that contained relatively low or high amounts of phytate. Although the fractional absorption of zinc from the bread meals was nearly two-fold that from the porridges, there were no significant differences in zinc absorption by type of fortificant, regardless of the estimated phytate-to-zinc ratio of the study diet. The porridges contained substantially more phytate (phytate:zinc = 12) than did the breads

4 282 LÓPEZ DE ROMAÑA ET AL FIGURE 2. Geometric mean (± 1 SD) and individual paired values for the absorption of zinc from porridge fortified with zinc oxide, by type of zinc tracer (P = 0.007, paired t test). (phytate:zinc = 0.5), and this likely explains the differences in zinc absorption from the 2 types of foods. Because other components of the diets also varied, however, other factors may have contributed to these differences. Two types of methods have been used previously to compare zinc absorption from different zinc compounds. One method, the oral-zinc-tolerance test, can be used to measure the relative absorption of zinc from aqueous solutions of different zinc compounds. Using this technique, Prasad et al (12) found that zinc was better absorbed from zinc sulfate than from zinc oxide, but English-Westcott et al (19) reported that there were no differences in zinc absorption from the 2 compounds. Interestingly, a third group of investigators completed oral-zinc-tolerance tests of zinc absorption by using zinc acetate, another water-soluble zinc salt, and zinc oxide, both before and after treating subjects to inhibit their gastric acid secretion (9). Whereas there were no differences in zinc absorption in the untreated subjects, there was greater absorption from zinc acetate than from zinc oxide after induced hypochlorhydria. Thus, differences in the zinc absorption of watersoluble and water-insoluble zinc salts may become apparent only in persons with impaired gastric acid secretion. No information on gastric function was reported in the first 2 aforementioned studies, so it is possible that this factor explains their contradictory results. Using an alternative study design based on stable-isotope tracers, other investigators also compared zinc absorption from test meals fortified with iron and either zinc sulfate or zinc oxide. In one study in Indonesia, no differences were found in fractional zinc absorption from wheat dumplings that were fortified with 1 of the 2 zinc salts (20). Likewise, in another study in Mexico, no differences in zinc absorption were reported from maize diets fortified with either zinc sulfate or zinc oxide (JL Rosado, unpublished observations, 2000). Thus, despite the theoretical disadvantage of zinc oxide because of its poor solubility in water, it appears that this zinc salt is absorbed as well as zinc sulfate when provided in a food matrix. Nevertheless, because gastric function was not assessed in these study subjects, it is conceivable that differences might occur in persons with hypochlorhydria. This may be of particular concern in populations with high rates of malnutrition or H. pylori infection, both of which can produce secondary impairment of gastric acid secretion (10, 11). Therefore, caution is warranted when choosing zinc fortificants in these settings, and further studies are needed to assess zinc absorption from meals fortified with different zinc salts in persons with hypochlorhydria. Unlike the previous tracer studies described, the present study used radioisotopic 65 Zn tracers and the whole-body counter technique. Advantages of this technique are the ability to confirm quantitatively the consumption of the tracer and the relatively low cost of radioisotope studies compared with stable-isotope studies. Although the whole-body counting technique requires application of an assumption regarding the fecal excretion of absorbed zinc, data are available from a previous study to permit this correction (13). In this former study, an equation was developed to correct for the excretion of absorbed zinc on the basis of the results of repeated whole-body counting after intravenous administration of 65 Zn. In the current study, we applied this same, previously developed equation to correct for losses of absorbed zinc. The present study also examined whether the chemical form of the isotopic tracer affected the estimation of zinc absorption from fortified foods. Notably, the estimated absorption of zinc from wheat porridge fortified with zinc oxide was slightly, but significantly, greater when zinc chloride was used as the tracer than when zinc oxide was the tracer. It is possible that the water-soluble zinc chloride tracer was better absorbed than the insoluble zinc oxide tracer, so it is preferable to use either the same chemical form of zinc as the fortificant that is being evaluated or a tracer with similar solubility. It is also conceivable that the different method of delivery of the tracer was responsible for the observed differences, because the zinc oxide tracer was added directly to the porridge whereas the zinc chloride solution was provided as a separate beverage along with the meal. Nevertheless, a recent study found that there were no differences in zinc absorption when tracers were added directly to the meal either 16 h before consumption or shortly before serving (21). Therefore, in the present study, it seems more likely that zinc absorption was affected by the chemical form of the tracer rather than the manner of delivery. It is also possible that the present comparison of zinc oxide and zinc chloride tracers was influenced by the nonrandom order in which these were given. However, examination of a possible sequence effect during the first 2 study periods, when the order of treatments was randomly assigned, did not support this conclusion. We found no correlation between iron status, as assessed by serum ferritin, and zinc absorption. Iron homeostasis is regulated primarily by the expression of a divalent metal transporter (DMT1), which is located in the apical membrane of the small intestine. The iron status of the enterocyte strongly affects DMT1 expression and regulates the transport of iron across the mucosa (22). Although DMT1 is known to be an iron transporter, it was originally found to transport other divalent cations, including zinc (23). Because ferritin expression and DMT1 expression are inversely correlated, we decided to examine whether a relation between serum ferritin and zinc absorption might occur. No such correlation was found, suggesting that iron status and DMT1 expression may not affect zinc absorption. In summary, the results of the present study suggest that either zinc sulfate or zinc oxide can be used to fortify wheat products. Because of the lower cost of zinc oxide, this salt may be considered preferable in most situations. However, because of theoretical concerns regarding the absorption of zinc oxide by persons with impaired gastric acid secretion, additional information would

5 ABSORPTION OF ZINC FORTIFICANTS 283 be desirable for this subgroup. In future tracer studies of the absorption of zinc fortificants added to different food products, the same zinc salt should be used for both the tracer and the fortificant to ensure that these are metabolized similarly. We appreciate the valuable assistance of Janet M Peerson (Program in International Nutrition, UC Davis), who guided the statistical analyses, and Leonard F Mausner (Brookhaven National Laboratory), who was responsible for the synthesis of the isotopic tracers. DLR, BL, and KHB participated in the study design, data analysis, and writing of the manuscript. In addition, DLR was responsible for data collection. None of the authors had a personal financial interest in the results, and none had any involvement in a company whose policies might be affected by the findings of the study. REFERENCES 1. Brown KH, Wuehler SE, Peerson JM. The importance of zinc in human nutrition and estimation of the global prevalence of zinc deficiency. Food Nutr Bull 2001;22: Brown KH, Peerson JM, Rivera J, Allen LH. Effect of supplemental zinc on the growth and serum zinc concentrations of prepubertal children: a meta-analysis of randomized controlled trials. Am J Clin Nutr 2002;75: Fraker PJ, Gershwin ME, Good RA, Prasad AS. Interrelationships between zinc and immune function. Fed Proc 1986;45: Bhutta ZA, Black RE, Brown KH, et al (Zinc Investigator s Collaborative Group). Prevention of diarrhea and pneumonia by zinc supplementation in children in developing countries: a pooled analysis of randomized controlled trials. J Pediatr 1999;135: Bhutta ZA, Black RE, Brown KH, et al (Zinc Investigator s Collaborative Group). Therapeutic effects of oral zinc in acute and persistent diarrhea in children in developing countries: pooled analysis of randomized controlled trials. Am J Clin Nutr 2000;72: Caulfield LE, Zavaleta N, Shankar AH, Merialdi M. Potential contribution of maternal zinc supplementation during pregnancy to maternal and child survival. Am J Clin Nutr 1998;68: Black MM. Zinc deficiency and child development. Am J Clin Nutr 1998;68(suppl):464S 9S. 8. Gibson R. Zinc nutrition in developing countries. Nutr Res Rev 1994; 7: Henderson LM, Brewer GJ, Dressman JB, et al. Effect of intragastric ph on the absorption of oral zinc acetate and zinc oxide in young healthy volunteers. JPEN J Parenter Enteral Nutr 1995;19: Gilman RH, Partanen R, Brown KH, et al. Decreased gastric acid secretion and bacterial colonization of the stomach in severely malnourished Bangladeshi children. Gastroenterology 1988;94: Dale A, Thomas JE, Darboe MK, Coward WA, Harding M, Weaver LT. Helicobacter pylori infection, gastric acid secretion, and infant growth. J Pediatr Gastroenterol Nutr 1998;26: Prasad AS, Beck FWJ, Nowak J. Comparison of absorption of five zinc preparations in humans using oral zinc tolerance test. J Trace Elem Exp Med 1993;6: Arvidsson B, Cederblad A, Björn-Rasmussen E, Sandstrom B. A radionuclide technique for studies of zinc absorption in man. Int J Nucl Med Biol 1978;5: Dawson-Hughes B, Seligson FH, Hughes VA. Effects of calcium carbonate and hydroxyapatite on zinc and iron retention in postmenopausal women. Am J Clin Nutr 1986;44: Sandström B, Kivistö B, Cederblad A. Absorption of zinc from soy protein meals in humans. J Nutr 1987;117: Sandström B, Arvidsson B, Cederblad A, Bjorn-Rasmussen E. Zinc absorption from composite meals. I. The significance of wheat extraction rate, zinc, calcium, and protein content in meals based on bread. Am J Clin Nutr 1980;33: US Department of Agriculture. Nutrient database for standard reference, release 15. Internet: nut_search.pl (accessed 10 August 2001). 18. Harland BF, Oberleas D. Phytate in foods. World Rev Nutr Diet 1987; 52: English-Westcott JL, Hambidge KM, Ellenbogen L. A comparison of zinc sulfate and oxide absorption in humans using an oral zinc tolerance test. FASEB J 1991;5:A933 (abstr). 20. Herman S, Griffin IJ, Suwarti S, et al. Co-fortification of iron-fortified flour with zinc sulfate, but not zinc oxide, decreases iron absorption in Indonesian children. Am J Clin Nutr 2002;76: Fredlund K, Rossander-Hulthen L, Isaksson M, Almgren A, Sandberg AS. Extrinsic labeling of zinc and calcium in bread. Appl Radiat Isot 2002;57: Tallkvist J, Bowlus CL, Lönnerdal B. Functional and molecular responses of human intestinal Caco-2 cells to iron treatment. Am J Clin Nutr 2000;72: Gunshin H, Mackenzie B, Berger UV, et al. Cloning and characterization of a mammalian proton-coupled metal-ion transporter. Nature 1997;388:482 8.