Iron indexes and total antioxidant status in response to soy protein intake in perimenopausal women

¹ From the Departments of Food Science and Human Nutrition (JHS, DLA, SBD, and MBR) and Statistics (CTP), Iowa State University, Ames.

² Paper no. J-19145 of the Iowa Agriculture and Home Economics Experiment Station, Ames, IA, and project no. 3602.

³ Supported by the Hatch Act, the state of Iowa, and a special grant from the US Department of Agriculture. The soy-protein isolates (Supro 675 HG and 675 IF) were donated by Protein Technologies International, a DuPont Business, St Louis; whey protein (ProMod) was donated by Ross Laboratories, Columbus, OH; cranberries were donated by Ocean Spray Cranberries, Inc, Lakeville-Middleboro, MA; and the flavorings and extracts were donated by Tone Brothers, Inc, Ankeny, IA.

⁴ Reprints not available. Address correspondence to MB Reddy, 1127 HNSB, Department of Food Science and Human Nutrition, Iowa State University, Ames, IA 50011. E-mail: mbreddy{at}iastate.edu.

ABSTRACT  
Background: Elevated iron stores, oxidative stress, and estrogen deficiency may place postmenopausal women at greater risk of heart disease and cancer than premenopausal women.

Objective: The objective was to determine the effect of soy-protein isolate (SPI) intake and iron indexes on plasma total antioxidant status (TAS) in perimenopausal women after control for other contributing factors.

Design: Perimenopausal women (n = 69) were randomly assigned (double blind) to treatment: isoflavone-rich SPI (SPI+; n = 24), isoflavone-poor SPI (SPI-; n = 24), or whey protein (control; n = 21). Each subject consumed 40 g soy or whey protein daily for 24 wk. Plasma TAS, serum ferritin, serum iron, transferrin saturation, and hemoglobin were measured at baseline, week 12, and week 24.

Results: No significant time-by-treatment interactions on iron indexes or TAS were observed, whereas time had an effect on serum ferritin (P 0.0001) and hemoglobin (P = 0.004) but not on TAS. Multiple regression analysis showed that at week 12, 48% (P 0.0001) of the variability in TAS was accounted for by baseline TAS, alcohol intake, soy intake (soy compared with control; P = 0.016), plasma lipoprotein(a), and dietary iron. At week 24, 47% of the variability in TAS was accounted for by baseline TAS, serum ferritin, serum estrone, dietary zinc, and dietary meat, fish, and poultry.

Conclusions: SPI intake had no significant effect on iron status, but our results suggest that dietary soy protein and low iron stores may protect perimenopausal women from oxidative stress.

Key Words: Antioxidant status • ferritin • iron indexes • lipoprotein(a) • menopause • soy-protein isolate • oxidation • perimenopausal women

INTRODUCTION  
Ample information is available on the iron status of pre- and postmenopausal women, whereas far less is known about the change in iron status or its effect on plasma antioxidant status during the menopausal transition. Oxidative status typically decreases with age (1), whereas body iron stores generally increase, especially in women after menopause (2). High iron stores have been hypothesized to induce oxidative stress because of the ability of iron to catalyze the Haber-Weiss reaction, which produces the hydroxyl radical. This radical is highly reactive with all biomolecules and is considered to be very toxic, causing structural damage to macromolecules (eg, proteins and lipids) and breakage of DNA strands (3). Generation of this radical in excess may deplete antioxidants (4). This disturbance in the equilibrium between prooxidants and antioxidants may contribute to diseases such as heart disease and cancer. Because epidemiologic evidence suggests that elevated iron stores place one at risk for developing heart disease (5, 6) and cancer (7), postmenopausal women may have the same risk as men (8). Antioxidant defense systems, which include enzymes, plasma proteins, and dietary factors (9), react with radicals and minimize their damage (10). These defense systems may be compromised during and after menopause and the concomitant increase in iron stores may increase such oxidative damage (11, 12).

Soy-protein isolate (SPI) is used extensively by the food industry, where it is incorporated into a variety of processed foods (13). SPI is known to markedly inhibit nonheme-iron absorption (14, 15), primarily because of its high phytic acid content (16) and protein moiety (17). Thus, diets rich in soy protein may be inadequate for maintaining optimal iron status in populations at risk of developing iron deficiency, yet be beneficial to groups at risk of developing excess iron stores. In addition, naturally occurring isoflavones in soy may have antioxidant properties because of their ability to donate hydrogen atoms or electrons from their hydroxyl groups to free radicals, making them less reactive (18). However, the ability of these isoflavones in vivo to affect total antioxidant status (TAS) during menopause has not been documented. We hypothesized that consumption of SPI during the menopausal transition may reduce the rate of menopause-associated increases in iron stores, thereby improving the antioxidant status in perimenopausal women.

SUBJECTS AND METHODS  
Research design and treatment
Perimenopausal women were recruited throughout Iowa and bordering states. Subjects were included in the study if they met the following criteria: had 10 hot flushes/wk, had irregular menses or menstrual period cessation, had an elevated follicle-stimulating hormone concentration (30 IU/L), had a body mass index (BMI; in kg/m²) between 20 and 31, were free from chronic disease, were not excessively exercising [<10.46 MJ (<2500 kcal) expenditure/wk], had one or both ovaries remaining, were able to participate for 24 wk, and were willing to be randomly assigned to treatment. Exclusion criteria included chronic disease, routine use of medications, smoking, alcohol abuse, estrogen or hormone replacement therapy during the prior 12 mo, or a history of eating disorders. Women started the study in 1 of 4 waves or cohorts beginning in January 1997, May 1997, September 1997, and March 1998. The study protocol and consent forms were approved by the Iowa State University Human Subjects Review Committee.

In this double-blind, parallel-arm, 24-wk study, 69 women were randomly assigned to 1 of 3 treatment groups: isoflavone-rich SPI (SPI+; n = 24; Protein Technologies International, St Louis), isoflavone-poor SPI (SPI-; n = 24; Protein Technologies International), or whey-protein control (n = 21; Ross Laboratories, Columbus, OH). The women were free-living and were supplied with a total of 40 g protein/d, with 50% of this protein incorporated into a muffin and the other 50% as a powder that subjects incorporated into their food or beverages. Because the muffin and the powder provided 2.09 MJ (500 kcal)/d, subjects were instructed to consume these as a meal replacement and not as a supplement. The muffins were prepared in the Human Metabolic Unit of the Center for Designing Foods to Improve Nutrition at Iowa State University. To control for variability in supplement intake among subjects, each participant was provided with a single daily over-the-counter vitamin and mineral supplement (Spring Valley Sentury-Vite; Leiner Health Products Inc, Carson, CA) and was instructed to not take any other vitamin or supplement and to avoid additional food items containing isoflavones throughout the study.

Data collection
As described by Alekel et al (19), anthropometric measures, fasting blood samples, and 24-h urine samples were collected from each subject at baseline, week 12, and week 24. Anthropometric data included the measurement of height with a stadiometer and weight with a balance-beam scale (Health-o-Meter, Inc, Bridgeview, IL). Dual-energy X-ray absorptiometry (QDR-2000+; Hologic, Inc, Waltham, MA) was used by 2 trained researchers to assess total body composition (lean and fat mass), and the data were analyzed by using software provided by the manufacturer (version 7.10, 1992).

Dietary intake was assessed at baseline, week 12, and week 24 from 5-d food records. To assist the subjects in quantifying portion sizes, 2-dimensional visual aids (1981; Nutritional Consulting Enterprises, Morgan/Posner, Framingham, MA) of food portions were provided. The assessment of dietary intake included macro- and micronutrients. The food records were analyzed by trained nutrition students using the NUTRITIONIST IV computerized nutrient database program (version 4.1, 1995; First Data Bank, San Bruno, CA).

Urine samples were portioned and frozen at -80°C for subsequent analysis of isoflavones (Fujicco, Inc, Kobe, Japan) to monitor compliance. The serum and plasma from each blood sample were stored at -80°C until serum ferritin, serum iron, transferrin saturation, lipids, serum reproductive hormones (follicle-stimulating hormone, estrone, and 17ß-estradiol), and plasma TAS were determined. The hemoglobin concentration was determined in whole blood with the HemoCue system (HemoCue, Inc, Mission Viejo, CA) immediately after the blood samples were drawn. The serum ferritin concentration was determined with the use of an enzyme-linked immunoassay kit (RAMCO Laboratories, Houston). Plasma TAS was determined with the use of an assay kit according to the manufacturer's (Calbiochem-Novabiochem Corp, La Jolla, CA) guidelines. Total antioxidant concentrations were measured with the use of a calorimetric method that relies on the ability of antioxidants in the sample to inhibit the oxidation of ABTS [2,2'-azino-di-(3-ethylbenzthiazoline sulfonate)] to ABTS⁺ by metmyoglobin (a peroxidase). Quest Diagnostics (St Louis), a certified clinical laboratory, measured serum iron, transferrin saturation, 17ß-estradiol, estrone, total cholesterol, LDL cholesterol, HDL cholesterol, and triacylglycerol at each time point. LDL-cholesterol concentrations were calculated by using the Friedewald equation (1972):

RESULTS  
Compliance
Compliance was based on the women's reported intakes and was corroborated by the urinary excretion of isoflavones (19). Eighty-seven percent of the women (60 of 69 women) reported consuming 100% of the muffins and 84% (58 of 69) reported consuming 100% of the powder during the study. The data for one woman from the control group were removed from all analyses because her urinary isoflavone excretion during the treatment period was similar to that of subjects in the SPI+ group. Thus, we concluded that she was consuming isoflavone-containing foods.

Subject characteristics, dietary intake, and serum measures
Characteristics (age and anthropometric data) of the women at baseline are presented in Table 1. Because these characteristics were not significantly different between the 3 treatment groups, the values are the means for the 3 treatment groups combined. The dietary intake of selected nutrients and serum analytes at baseline, week 12, and week 24 are shown in Table 2. Because no significant time-by-treatment interactions on nutrient intakes (except dietary iron, P 0.001) or serum measures were found, values reported at each time point are for women in the 3 treatment groups combined. At baseline, dietary iron intakes were similar between the 3 treatment groups, but at weeks 12 and 24 the control group consumed 31% and 42% less iron, respectively, than did the SPI+ and SPI- groups combined. These differences may have resulted because of the higher iron content of SPI than of whey protein. Some of the subjects consumed no alcohol during the study; therefore, the values are reported as the 75th percentile because the 50th percentile was equal to zero.

View this table:
TABLE 1 . Baseline characteristics of perimenopausal women from all 3 treatment groups combined¹  

View this table:
TABLE 2 . Dietary nutrient intakes and serum analytes at baseline, week 12, and week 24 in perimenopausal women from the 3 treatment groups combined¹  
Iron indexes and antioxidant status
Plasma total antioxidant and hemoglobin concentrations are presented as means ± SEMs, whereas serum ferritin values were not normally distributed and thus are presented as medians with the 25th and 75th percentiles indicating variability (Figure 1). Treatment did not affect serum ferritin or total antioxidants, but had a marginal effect on hemoglobin (F = 3.17, P = 0.052). Time had a significant effect on hemoglobin (P = 0.004) and serum ferritin (P 0.0001) but not on plasma total antioxidants. However, there was no time-by-treatment interaction for serum ferritin (P = 0.41), hemoglobin (P = 0.09), or total antioxidants (P = 0.35). The control group (126 g/L) had a lower (P = 0.052) hemoglobin concentration than did the SPI+ (134 g/L) and SPI- (132 g/L) groups at baseline, but the difference between the 3 groups was no longer significant at week 24. Overall, the mean serum ferritin concentration for the 3 groups combined generally increased from baseline (27.5 µg/L) to week 24 (33.9 µg/L), but there were no significant differences between the groups during the study. Similarly, time and treatment had no significant effect on serum iron or transferrin saturation (Table 3). The mean serum iron concentration varied for the 3 groups from 14.7 to 16.9 µmol/L at baseline and from 14.6 to 15.0 µmol/L at the end of the study, with no significant differences at either time point.

View larger version (12K):
FIGURE 1. . Concentrations of hemoglobin (  

View this table:
TABLE 3 . Transport iron indexes at baseline, week 12, and week 24 in perimenopausal women from the 3 treatment groups¹  
We further explored the relation between plasma TAS and selected factors in Tables 1 and 2 in conjunction with iron indexes by using multiple regression analysis. We used orthogonal contrasts (SPI+ and SPI- groups compared with the control group; SPI+ group compared with SPI- group) to determine the effect of treatment. Factors contributing to plasma TAS at weeks 12 and 24 are presented in Table 4. In the regression model for week 12, the only significant treatment contrast (SPI+ and SPI- groups compared with the control group; P = 0.016) indicated that plasma total antioxidant concentrations were affected positively by soy protein rather than by whey protein. Moreover, the other contrast (ie, the SPI+ group compared with the SPI- group) was not significant, indicating that the effect was due to the soy protein and not to the isoflavones. At week 12, baseline plasma TAS, alcohol intake, soy intake, plasma lipoprotein(a), and dietary iron accounted for 48% of the variability (P 0.0001) in TAS. However, at week 24, baseline plasma total antioxidant concentrations, dietary MFP, serum ferritin, serum estrone, and dietary zinc accounted for 47% of the variability in plasma TAS. To account for differences in iron intake between the treatment groups, dietary iron (P = 0.56) was forced into the model at week 24. The contrast between the SPI+ and SPI- groups and the control group (P = 0.22) was not significant at week 24. With the exception of baseline TAS, the factors influencing TAS were different at the 2 time points.

View this table:
TABLE 4 . Regression analyses: contributors to plasma total antioxidant status (TAS) at weeks 12 and 24¹  

DISCUSSION  
At baseline, 13% of all women were anemic (hemoglobin < 12 g/L); at week 24, 7% of all women were anemic, suggesting that the incidence of anemia declined during the menopausal transition. The increase in hemoglobin was more dramatic in the control group, possibly because of the marginally lower (albeit nonsignificant) hemoglobin values in these subjects at baseline. Regardless of treatment, serum ferritin showed age-associated increases, menopause-associated increases, or both during the study (P 0.0001), which is consistent with the findings of other studies (2, 21). The increase in serum ferritin may have been due to decreased iron losses via menstruation secondary to hormonal changes (22). Alternatively, the increase in serum ferritin may have been influenced by the overall lower mean serum ferritin concentrations observed at baseline in our study (27.5 µg/L) than in premenopausal (37 µg/L) and postmenopausal (71 µg/L) women in the study by Milman et al (22). The overall percentage of women with iron deficiency in our study (ie, serum ferritin < 15 µg/L) was 44.4% at baseline, which decreased to 25.9% by week 24. This percentage of iron deficiency is higher than that found in women of childbearing age in the third National Health and Nutrition Examination Survey (23). Although soy consumption, regardless of its isoflavone content, reduced serum iron (by 11%) and transferrin saturation (by 3–7%), the effect was too small to be significant, perhaps because of the relatively small sample size or high variability.

Oxidative status depends on the balance between prooxidants (eg, iron) and antioxidants (eg, enzymes, plasma proteins, and dietary factors) (24). The plasma total antioxidant concentration is an overall indicator of oxidative status. The women in our study had baseline plasma total antioxidant concentrations ranging from 0.53 to 0.68 mmol/L, apparently lower than values previously reported (1.3–1.8 mmol/L) in humans (25). The use of multiple regression analysis to determine factors that may have influenced plasma TAS showed that baseline plasma total antioxidant concentrations, soy consumption, and alcohol intake favorably contributed to plasma TAS at week 12. The positive association between TAS at baseline and at 12 wk illustrates that women who began the study with a better TAS were better able to maintain this status throughout the study. Soy intake, irrespective of isoflavone content, had a beneficial effect on TAS, perhaps because of the antioxidant property of phytic acid (26, 27), the reduction of nonheme-iron absorption due to both phytic acid (28) and the protein moiety (17) or to both factors. Our previous data also clearly indicated that dietary phytic acid significantly reduces iron stores and decreases lipid peroxidation in iron-overloaded mice (29). The positive effect of alcohol on plasma TAS may have been due to the polyphenol content of the alcoholic beverages. Phenolic compounds have been shown to have an antioxidant effect, as evidenced by their ability to inhibit the oxidation of human LDL in vitro (30). The positive relation of dietary MFP with TAS at week 24 may have been related to the antioxidant properties of conjugated linoleic acid, the only known antioxidant associated primarily with animal foods (31). Conjugated linoleic acid has been shown to inhibit free radical generation (32), thereby providing a plausible explanation for the favorable effect of MFP on plasma TAS that we observed.

Plasma lipoprotein(a) and dietary iron were negatively associated with TAS at week 12, but not at week 24. Lipoprotein(a) is of increasing interest because it is considered both a marker of thrombotic and atherosclerotic risk (33). Hormonal changes that occur during menopause may elevate lipoprotein(a) concentrations (34). Elevated lipoprotein(a) concentrations are associated with an increased risk of atherosclerotic cardiovascular disease, a disease known to be associated with lipid peroxidation. This association may exist because greater lipoprotein(a) concentrations may provide more abundant substrate for reaction with radicals, which in turn may burden the antioxidant system, thereby decreasing plasma TAS. The inverse relation between dietary iron and TAS suggests a prooxidant role of iron in promoting lipid peroxidation in vitro (4).

Serum ferritin, serum estrone, and dietary zinc were negatively related to plasma TAS only at week 24. The negative relation between serum ferritin and TAS supports the findings of epidemiologic studies that indicate that lower iron stores may protect against oxidative stress–related diseases (2, 6). The negative relation between serum ferritin and plasma TAS may have reflected the increase in iron stores by week 24 in most of the perimenopausal women, which may have been related to age, menopause, or both. We could not distinguish between an age- or menopause-related effect on iron stores, nor did we find a relation between serum estrogens and serum ferritin. The lack of a relation between serum estrogen and serum ferritin in our study is consistent with previous findings (22). The inverse relation between serum estrone and TAS is difficult to explain, but may be related to the high fat mass (37%) in these women. Adipose tissue is a primary site for peripheral aromatization of adrostenedione to estrone (35). Thus, we speculated that the relatively high serum estrone concentration in these women compared with postmenopausal women (36) may have reflected adipose tissue conversion from adrostenedione. The inverse relation between serum estrone and TAS may have simply mirrored the estrone concentrations typically reported during menopause. Although physiologic doses of 17ß-estradiol have been shown to inhibit lipoprotein oxidation (37), serum estrone has not been shown to have antioxidant properties. Thus, this inverse relation is probably not one of cause and effect.

In contrast, it was surprising to observe an inverse relation between dietary zinc and plasma TAS because zinc is an essential cofactor for many antioxidant enzymes, such as superoxide dismutase (EC 1.15.1.1). Because zinc also binds to phytic acid (38), dietary zinc competes intraluminally with iron (39) for binding sites on phytic acid, which may have resulted in a greater proportion of iron available for absorption. Thus, as dietary zinc increases it may allow for greater nonheme-iron absorption, thereby increasing serum ferritin and decreasing total antioxidant concentrations. This hypothesis is consistent with the inverse relation we noted between dietary iron and TAS at week 12 and serum ferritin and TAS at week 24.

In conclusion, soy intake had no significant influence on any iron indexes during the 24-wk study period, but had a beneficial effect on TAS at week 12 in these perimenopausal women. In the short term (ie, after 12 wk), SPI (regardless of its isoflavone content) may enhance TAS, but in the long term (ie, 24 wk), low iron stores may protect against oxidative stress by improving TAS. However, in addition to iron status, other factors that we did not assess may play a role in determining TAS in perimenopausal women. Our data also indicate that some perimenopausal women with low serum ferritin may not yet be at risk of compromised oxidative status associated with higher iron stores typically found in postmenopausal women. In addition, the women in the present study who had a better TAS at the beginning of menopause were more likely to maintain this status throughout the study. Future larger scale studies are needed to explore preventive measures aimed at controlling menopause- and age-related increases in body iron and thereby maintaining optimum TAS.

REFERENCES  

1. Jayachandran M, Jayanthi B, Sundaravadivel B, Panneerselvan C. Status of lipids, lipid peroxidation, and antioxidant systems with vitamin C supplementation during aging in rats. J Nutr Biochem 1996; 7:270–5.
2. Berge LN, Bonaa KH, Norday A. Serum ferritin, sex hormones, and cardiovascular risk factors in healthy women. Arterioscler Thromb 1994;14:857–61.
3. Santanam N, Ramachandran S, Parthasarathy S. Oxygen radicals, antioxidants, and lipid peroxidation. Semin Reprod Endocrinol 1998;16:275–80.
4. McCord JM. Human disease, free radicals, and the oxidant/antioxidant balance. Clin Biochem 1993;26:351–7.
5. Salonen JT. The role of iron as a cardiovascular risk factor. Curr Opin Lipidol 1993;4:277–82.
6. Kiechl S, Aichner F, Gerstenbrand F, et al. Body iron stores and presence of carotid atherosclerosis. Arterioscler Thromb 1994;14: 1625–30.
7. Nelson RL, Davis FG, Sutter E, Sobin LH, Kikendall JW, Bowen P. Body iron stores and risk of colonic neoplasia. J Natl Cancer Inst 1994;86:455–60.
8. Sullivan JL. Iron and the sex difference in heart disease risk. Lancet 1981;1:1293–4.
9. Thomas JA. Oxidative stress and oxidant defense. In: Shils ME, Olson JA, Shike M, Ross AC, eds. Modern nutrition in health and disease. 9th ed. Baltimore: Williams & Wilkins, 1999:751–60.
10. Halliwell B. The role of oxygen radicals in human disease, with particular reference to the vascular system. Haemostasis 1993;23(suppl): 118–26.
11. Berliner JA, Heinecke JW. The role of oxidized lipoproteins in atherogenesis. Free Radic Biol Med 1996;20:707–27.
12. Regnström J, Walldius G, Hådell K, et al. Influence of lipoprotein lipids, dietary fat and smoking on macrophage degradation of native and oxidized low density lipoprotein. Atherosclerosis 1994;110: 13–23.
13. Erdman JW, Fordyce EJ. Soy products and the human diet. Am J Clin Nutr 1989;49:725–37.
14. Cook JD, Morck TA, Lynch SR. The inhibitory effect of soy products on nonheme iron absorption in man. Am J Clin Nutr 1981;34: 2622–9.
15. Gillooly M, Bothwell T, Torrance J, et al. The effect of organic acids, phytates and polyphenols on the absorption of iron from vegetables. Br J Nutr 1983;49:331–6.
16. Hurrell RF, Juillerat M-A, Reddy MB, Lynch SR, Dasenko SA, Cook JD. Soy protein, phytate, and iron absorption in humans. Am J Clin Nutr 1992;56:573–8.
17. Lynch SR, Dassenko SA, Cook JD, Juillerat MA, Hurell R. Inhibitory effect of a soybean-protein-related moiety on iron absorption in humans. Am J Clin Nutr 1994;60:567–72.
18. Mitchell JH, Gardner PT, McPhail DB, Morrice PC, Collins AR, Duthie GG. Antioxidant efficacy of phytoestrogens in chemical and biological model systems. Arch Biochem Biophys 1998;360:142–8.
19. Alekel DL, St Germain A, Peterson CT, Hanson KB, Stewart JW, Toda T. Isoflavone-rich soy protein isolate attenuates bone loss in the lumbar spine of postmenopausal women. Am J Clin Nutr 2000;72:844–52.
20. SAS Institute Inc. SAS/STAT users guide, version 8.0. Cary, NC: SAS Institute Inc, 1999.
21. Penckofer S, Schwertz D. Improved iron status parameters may be a benefit of hormone replacement therapy. J Womens Health Gend Based Med 2000;9:141–51.
22. Milman N, Kirchoff M, Jorgensen T. Iron status markers, serum ferritin and hemoglobin in 1359 Danish women in relation to menstruation, hormonal contraception, parity, and postmenopausal hormone treatment. Ann Hematol 1992;65:96–102.
23. Looker A, Dallman P, Carroll M, Gunter E, Johnson C. Prevalence of iron deficiency in the United States. JAMA 1997;227:973–6.
24. Rice-Evans CA, Burdon R. Free radical-lipid interactions and their pathological consequences. Prog Lipid Res 1993;32:71–110.
25. Miller NJ, Rice-Evans C, Davies MJ, Gopinathan V, Milner A. A novel method for measuring antioxidant capacity and its application to monitoring the antioxidant status in premature neonates. Clin Sci (Colch) 1993;84:407–12.
26. Graf E, Eaton JW. Antioxidant functions of phytic acid. Free Radic Biol Med 1990;8:61–9.
27. Ko KM, Godin DV. Ferric ion-induced lipid peroxidation in erythrocyte membrane: effects of phytic acid and butylated hydroxytoluene. Mol Cell Biochem 1990;95:125–31.
28. Reddy MB, Hurrell RF, Juillerat MA, Cook JD. The influence of different protein sources on phytate inhibition of nonheme-iron absorption in humans. Am J Clin Nutr 1996;63:203–7.
29. Guo L. The protective effect of phytic acid on reducing oxidative stress in ß₂-microglobulin deficient mice. Master's thesis. Iowa State University, Ames, 1999.
30. Frankel E, Waterhouse A, Teissedre P. Principal phenolic phytochemicals in selected California wines and their antioxidant activity in inhibiting oxidation of human low-density lipoproteins. J Agric Food Chem 1995;43:890–4.
31. Pariza MW, Park Y, Cook ME. Mechanisms of action of conjugated linoleic acid: evidence and speculation. Proc Soc Exp Biol Med 2000;223:8–13.
32. Steinhart C. Conjugated linoleic acid—the good news about animal fat. J Chem Ed 1996;73:A302–3.
33. Dobroski DR, Loscalzo J. Thrombotic risk factors for atherosclerosis. Coron Artery Dis 1996;7:919–32.
34. Mijatovic V, Kenemans P, Netelenbos JC, et al. Oral 17 ß-estradiol continuously combined with dydrogesterone lowers serum lipoprotein(a) concentrations in healthy postmenopausal women. J Clin Endocrinol Metab 1997;82:3543–7.
35. Nordin BE, Crilly RG, Marshall DH, Barkworth SA. Oestrogens, the menopause and the andrenopause. J Endocrinol 1981;89(suppl): 131P–43P.
36. Nachtigall LE, Raju U, Banerjee S, Wan L, Levitz M. Serum estradiol-binding profiles in postmenopausal women undergoing three common estrogen replacement therapies: associations with sex hormone-binding globulin, estradiol, and estrone levels. Menopause 2000;7:243–50.
37. Sack MN, Rader DJ, Cannon RO III. Oestrogen and inhibition of oxidation of low-density lipoproteins in postmenopausal women. Lancet 1994;343:269–70.
38. Zhou JR, Erdman JW Jr. Phytic acid in health and disease. Crit Rev Food Sci Nutr 1995;35:495–508.
39. Solomons NW, Pineda O, Viteri F, Sandstead HH. Studies on the bioavailability of zinc in humans: mechanism of the intestinal interaction of nonheme iron and zinc. J Nutr 1983;113:337–49.