|
|
||||||||
Original Research |
School of Biological & Chemical Sciences, Deakin University, Burwood (E.L.A., M.J.B.) AUSTRALIA
International Health & Development Unit, Alfred Hospital, Monash University, Prahran (F.D.) Victoria, AUSTRALIA
Address reprint requests to: Madeleine J. Ball, M.D., Biomedical Sciences, School of Biological & Chemical Sciences, Deakin University, 221 Burwood Hwy Burwood, Victoria, 3125, AUSTRALIA. E-mail: mjbikr{at}deakin.edu.au
| ABSTRACT |
|---|
|
|
|---|
Design: A randomized cross over dietary intervention study.
Setting: Free-living individuals studied at Deakin University.
Subjects: Forty-five free-living healthy males aged 35 to 62 years completed the dietary intervention. Three subjects were non-compliant and excluded prior to analysis.
Interventions: A diet containing 150 grams of lean meat per day was compared to a diet containing 290 grams of tofu per day in an isocaloric and isoprotein substitution. Each dietary period was one month duration.
Results: Analysis of the seven-day diet record showed that diets were similar in energy, protein, carbohydrate, total fat, saturated and unsaturated fat, polyunsaturated to saturated fat ratio, alcohol and fiber. Total cholesterol and triglycerides were significantly lower, and in vitro LDL oxidation lag phase was significantly longer on the tofu diet compared to the meat diet. The hemostatic factors, factor VII and fibrinogen, and lipoprotein(a) were not significantly affected by the tofu diet.
Conclusions: The increase in LDL oxidation lag phase would be expected to be associated with a decrease in coronary heart disease risk.
Key words: tofu, LDL, oxidation, hemostatic factors, CHD
| INTRODUCTION |
|---|
|
|
|---|
Thrombosis is also an important contributor to atherosclerosis and may be the event which precipitates a myocardial infarction [9]. The hemostatic factors, factor VII and fibrinogen, are important in this process. Fibrinogen concentration is also a major determinant of plasma viscosity [10]. Cross-sectional studies [11] have shown a relationship between CHD and plasma levels of factor VII and fibrinogen. Hemostatic factors can be affected by diet, particularly the fat content [9], and it is important to determine whether other dietary factors, such as replacement of animal protein by vegetable proteins, have an effect.
Lipoprotein (a) (Lp(a)) has also been identified as an independent risk factor for CHD [12]. The physiological role of Lp(a) is not known, but it is thought to deliver cholesterol to cells at wound sites, a function that may become pathological when plasma levels of Lp(a) are high. Lp(a) is very similar to LDL, except it contains apolipoprotein(a) (apo(a)) attached to apolipoprotein B-100 by a disulfide bond. The apo(a) is homologous to plasminogen, which binds to the endothelial cell plasminogen receptors and interferes with clot lysis and inhibits the rate of plasmin generation [12]. Lp(a) is highly inheritable but has been shown to be altered by nicotinic acid and estrogen [13, 14]. Most of the dietary studies have shown little effect of changes in diet, including changes in dietary fat, on Lp(a) [14], although trans fatty acid intake appears to increase Lp(a) levels [15, 16]. In a cross sectional study conducted in our laboratory, female vegetarians tended to have lower Lp(a) levels than age-gender matched omnivore controls [17]; thus, it is relevant to assess whether soy proteins affect this lipoprotein.
A dietary intervention study was designed to investigate the effect of tofu, a soy product, versus meat on serum lipids, Lp(a), factor VII and fibrinogen, and the in vitro susceptibility of LDL to copper sulfate induced oxidation.
| METHODS AND MATERIALS |
|---|
|
|
|---|
Deakin University Ethics Committee approved the study protocol and informed written consent was obtained from each participant.
Study and Design
The study was designed to compare the effects of two diets using a randomised crossover design. Prior to commencement of the dietary intervention, all subjects were given detailed written and verbal instructions on how to complete a weighed food record accurately and were instructed to record their habitual diets for seven days, including two weekend days. This gave subjects practice on completing a food record and provided details of the subjects normal energy intakes and patterns of food consumption to aid in manipulating their diet. The subjects were then randomly assigned to a study diet for one month, after which they returned to their habitual diet for a two week washout period before undertaking the second diet for one month.
The diets were designed to be similar in energy, protein, fat, carbohydrate, alcohol and dietary fiber, with only the source of protein changing from an animal source to a plant source. Subjects consumed similar vegetarian breakfasts, lunches and snacks on both diets. During the meat diet, subjects consumed 150 grams (raw weight) of cooked lean red meat, with all visible fat removed, each day. Fifteen grams of polyunsaturated margarine was prescribed. Subjects avoided all soy products during this diet. The tofu diet was designed to replace 90 to 100% of the animal protein with 290 grams of tofu. Subjects were advised to add tofu to meals and snacks throughout the day to reduce the amount required to eat in one meal. To minimise the differences in monounsaturated (MUFA), polyunsaturated (PUFA) and saturated fatty acids (SFA) between the two diets, five grams of butter, five grams of lard and eight mL of olive oil were also prescribed daily on the tofu diet. The nutrients that are supplied from the lean meat and PUFA margarine and the tofu with butter, lard and olive oil are given in Table 1. All tofu was supplied from Blue Lotus Foods P/L (Kilsyth, Australia) and analysed for its genistein and daidzein content. The tofu and fats were supplied free, and recipes and specially prepared tofu biscuits were also provided to help subjects consume the tofu. Dietary counselling was provided weekly to help with dietary manipulation and to assess and improve compliance. All subjects were instructed to maintain their usual exercise patterns for the duration of the study. Subjects were instructed to advise the investigators if they felt hungry or satisfied and if they experienced any weight change. During the last week of each diet the subjects completed a seven-day diet record using accurate scales, or household measures when weighing was not possible [18]. All diets were coded and analysed using System for Online Dietary Analysis (SODA) Version 5 (Cottesloe, Australia) using NUTTAB 91 database with updated nutrient composition of tofu and margarine.
|
Laboratory Analysis
The concentrations of serum TC, TG, and high density lipoprotein (HDL) cholesterol were measured by an enzymatic colorimetric test using commercial kits from Boehringer Mannheim (Mannheim, Germany) on a Hitachi 704 autoanalyser (Tokyo, Japan). LDL cholesterol was calculated using the Friedewald equation [19]. Serum cholesteryl ester fatty acids were measured on samples from 11 subjects by gas chromatography using the method described by Sinclair [20].
The isoflavone analysis of the tofu was based on the method of Coward et al. [21]. Freeze dried tofu was frozen under liquid nitrogen and crushed to a powder in a pestle and mortar. One hundred mg of the powdered tofu was weighed out in duplicate in 20 mL vials. Ten mL of 80% aqueous methanol (Crown Scientific, Melbourne, VIC, Australia) were added to the sample and heated at 65°C for two hours. The vials were then centrifuged at 3000 rpm for 10 minutes. A one mL sample was aliquoted into the HPLC vial for analysis. Phytoestrogens (daidzein and genistein) were isolated by reverse-phase high performance liquid chromatography (Shimadzu system LC 10A) as previously described [22].
Urinary concentrations of genistein and daidzein were measured after isoflavonoids were deconjugated using glucuronidase (Sigma Chemical Company, St Louis, USA) and extracted with diethyl ether (BDH Chemicals, Melbourne, Australia), by reverse-phase HPLC (Shimadzu system LC10A) according to the method of Eldridge [23]. Creatinine was measured colorimetrically using reagents from Boehringer Mannheim (Mannheim, Germany) and urine controls from Bio-Rad (Anaheim, California) on a Hitachi 704 autoanalyser (Tokyo, Japan).
LDL was isolated by ultracentrifugation at 15°C at 100,00rpm for two hours in a Beckman Optima TLX-100 ultracentrifuge with a TLA-100.4 rotor (California, USA). The LDL band was extracted through the side of the tube by a needle. EDTA was leached from the LDL by dialysis in deoxygenated PBS buffer pH 7.4. The Sigma Bicinchoninic Acid protein assay kit, using the Lowry method [24], was used to determine the protein concentration of the LDL sample. Fifty µg of LDL, oxygenated PBS and 32µL of 1mM CuCl were added to a 2mL quartz cuvette [25], and the oxidation reaction products, conjugated dienes, were monitored continuously at 37°C on a Shimadzu UV-1601 spectrophotometer (Tokyo, Japan) at 234nm according to the method of Esterbauer [26]. Fig. 1 shows the oxidation process. The lag phase is determined as the number of minutes from time zero to the time the intercept of the tangent of the slope of the absorbance curve in the propagation phase intercepts the horizontal axis [27].
|
Lp(a) was calculated from the measurement of apo(a) that is bound to Lp(a), as there is no direct method to measure Lp(a). Apolipoprotein was measured using a Mercodia Apo(a) RIA kit (Uppsala, Sweden). According to the kit, one Unit of apo(a) is equal to 0.7 mg Lp(a).
Statistical Analyses
Statistical analysis was performed using SPSS version 8.0 (SPSS Corp., Chicago, Ill). The mean value of the samples collected three to four days apart was used. General linear model was used to investigate the overall effect of the two diets, any carry over effects or order effects [29]. The relationship between the change in urinary excretion of genistein and CHD risk factors was assessed by Spearmans correlation. The relationship between LDL oxidation lag phase and serum linoleic acid was also analysed using Spearmans correlation.
| RESULTS |
|---|
|
|
|---|
|
Mean (SD) urine excretion of genistein and daidzein per µmol creatinine was significantly higher on the tofu diet (p < 0.001): being 202 (196) ng/µmol and 401 (254) ng/µmol on the tofu diet and 12 (11) ng/µmol and 20 (24) ng/µmol on the meat diet, respectively. There was no significant correlation between the difference in urinary genistein excretion and change in any of the CHD risk factors measured.
Results of the CHD risk factors measured on the two diets are given in Table 3. TC and TG were significantly lower on the tofu diet than on the meat diet, but LDL cholesterol and LDL:HDL ratio were not significantly different. HDL was significantly lower on the tofu diet as reported and has been discussed elsewhere [30]. The blood lipid values at the commencement of each diet were not significantly different. Lp(a), factor VII and fibrinogen were not significantly different between diets. The LDL oxidation lag phase was significantly longer on the tofu diet.
|
| DISCUSSION |
|---|
|
|
|---|
Assessment of subjects food records and the high urinary isoflavone concentrations on the tofu diet indicated that subjects complied well with the prescribed diets. The macronutrient intake was not significantly different except for the protein source and dietary cholesterol. The later was significantly lower on the tofu diet, but it is doubtful that a mean difference of 90mg/day in the context of a low fat diet (total fat < 33% of energy) and a low dietary cholesterol diet (<300mg/day) had an impact on our results, except lipoprotein levels. Factor VII was not affected by dietary cholesterol in studies by Miller and Bladbjerg [31, 32], although an association was found in the Rotterdam cross-sectional study [9]. Fibrinogen levels do not appear to be altered by dietary cholesterol intake [10, 33]. In normocholesterolemic men, dietary cholesterol had no effect on Lp(a) levels [34]. According to Howells equation [35], this difference in dietary cholesterol could only cause a 0.05mmol/L difference in TC.
Serum cholesteryl ester fatty acids analysed in a subgroup to provide an independent assessment of the difference in fat between the two diets, revealed linoleic acid to be 8.5% higher on the tofu diet. This may indicate a small difference in intake in the few days prior to blood sampling, although the dietary records indicated similar intake. This small difference may have contributed to decreasing cholesterol on the tofu diet, if it replaced SFA, but would not explain the total change. The small increase in serum linoleic acid could suggest a greater concentration within the LDL particle, which would be expected to increase the susceptibility of the LDL to oxidation [36]. However the present results show a decrease in susceptibility to in vitro oxidation, with an increased lag phase on the tofu diet. The susceptibility of LDL to undergo oxidative modification is also related to antioxidant concentrations within the LDL particle, but the dietary difference between tofu and meat is unlikely to cause major changes in this parameter.
Phytoestrogens have been identified as having antioxidant activity [37], which could alter the susceptibility of LDL to oxidative modification. A small number of studies have investigated the effect of isoflavones on LDL oxidation [3843]. In the study by Hodgson et al. [38], different concentrations of isoflavonoids (genistein, daidzein and a metabolic products of daidzein, equol) added to serum all inhibited oxidative modification at 234nm, with addition of one µM of genistein, increasing the lag time by 57%. De Whalley found that added flavonoids enhanced the resistance of LDL to oxidation [40]. Kapiotis found that the addition of genistein and, to a lesser extent, daidzein inhibited oxidative modification of LDL in a cell mediated and a copper induced cell free system and that genistein could also protect isolated endothelial cells from damage by oxidised LDL [42]. Pretreatment with genistein did not have this effect indicating that genistein had to be present to exert an effect. Results from our current study of an increase in LDL lag phase indicate that the isoflavones may have entered the LDL particle to inhibit oxidation. In a study where a tablet containing 80mg of genistein and daidzein was given to subjects daily there was no effect on LDL oxidation or on lipoprotein levels [39]. However, subjects fed soycreme showed a greater amount of lipoprotein peroxidation products after dialysis of their LDL into copper dichloride [41]. Tikkanen et al. [43] investigated the effect of 57mg of genistein and daidzein daily provided in the form of a soy bar on in vitro copper sulfate initiated LDL oxidation in six healthy subjects. No change was seen in TC or TG concentrations after the two weeks of soy supplement, but there was a 20-minute increase in the lag phase after the soy period, although the LDL particle contained only small quantities of the phytoestrogens. The lag phase in the current study was 6.77 minutes (16%) longer on the soy diet [than on] the lean meat diet; and the 290g of tofu consumed daily supplied 119.8mg of genistein and daidzein. The larger increase in lag phase in the Tikkanen study may be explained by the small numbers, as there is a large variability in response to isoflavonoids between subjects [44]. It may also be that the slightly higher linoleic acid levels on our tofu diet reduced the effect.
The effect of soy phytoestrogens on LDL oxidation may be through their ability to act as free radical scavengers [38] although Tikkanen et al. [43] suggest the molar ratio of phytoestrogens to LDL was too small to explain the effects as free radical scavengers. They suggested several other possibilities including phytoestrogens ability to reduce hydroperoxide formation or that soy protein binds to apolipoprotein B of the LDL and thus prevents copper sulfate from binding and initiating the oxidation. Alternatively, the effect on oxidation may be through an undetected metabolite or due to soy altering the LDL particle.
Isoflavonoids have also been implicated as the component of soy that alters lipoprotein levels, but the effects of isolated phytoestrogens are inconsistent [39, 45]. Studies have fairly consistently shown that soy protein compared to casein decreases TC, LDL cholesterol and TG concentrations without significantly affecting HDL cholesterol concentration [4]. Our results showed that tofu decreased TC and TG concentrations, and there was a trend for LDL cholesterol to be lower on the tofu diet, but LDL:HDL ratio was not significantly different. Our study may not have achieved the significant reduction in LDL cholesterol found in many studies in the meta-analysis [4], because of the smaller amount of soy protein consumed, 35.7 grams of soy protein compared to 47 grams, or the fact that the comparison was of soy with meat proteins, rather than with casein as in most studies.
The effect of soy protein on hemostatic factors has not been investigated previously, and in this study it did not appear to influence factor VII and fibrinogen concentrations. A dietary intervention study by Hostmark et al. [46] found that a three-week vegetarian diet decreased fibrinogen concentration, but in this study there were many dietary changes including those involving fat which could have had an effect. A study investigating the effect of dietary fat found that stearic acid, but not myristic acid, increased factor VII [47].
Several authors initially suggested that Lp(a) was not affected by dietary change [13, 14], but further studies have identified a positive relationship with dietary intake of trans fatty acids [15]. Estrogens, as hormone replacement therapy can also lower Lp(a) levels [14, 48]. Hodgson et al. [45] found no effect on Lp(a) when postmenopausal women were given a tablet with 55mg of isoflavones rather than a placebo. Results from the current study suggest that the isoflavones genistein and daidzein do not influence Lp(a) levels.
| CONCLUSION |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
Received May 26, 1999. Accepted July 14, 2000.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. Zhan and S. C Ho Meta-analysis of the effects of soy protein containing isoflavones on the lipid profile Am. J. Clinical Nutrition, February 1, 2005; 81(2): 397 - 408. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Vega-Lopez, K.-J. Yeum, J. L Lecker, L. M Ausman, E. J Johnson, S. Devaraj, I. Jialal, and A. H Lichtenstein Plasma antioxidant capacity in response to diets high in soy or animal protein with or without isoflavones Am. J. Clinical Nutrition, January 1, 2005; 81(1): 43 - 49. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Atkinson, W. Oosthuizen, S. Scollen, A. Loktionov, N. E. Day, and S. A. Bingham Modest Protective Effects of Isoflavones from a Red Clover-Derived Dietary Supplement on Cardiovascular Disease Risk Factors in Perimenopausal Women, and Evidence of an Interaction with ApoE Genotype in 49-65 Year-Old Women J. Nutr., July 1, 2004; 134(7): 1759 - 1764. [Abstract] [Full Text] |
||||
![]() |
M. Sagara, T. Kanda, M. NJelekera, T. Teramoto, L. Armitage, N. Birt, C. Birt, and Y. Yamori Effects of Dietary Intake of Soy Protein and Isoflavones on Cardiovascular Disease Risk Factors in High Risk, Middle-Aged Men in Scotland J. Am. Coll. Nutr., February 1, 2004; 23(1): 85 - 91. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. L. Ashton, J. D. Best, and M. J. Ball Effects of Monounsaturated Enriched Sunflower Oil on CHD Risk Factors Including LDL Size and Copper-Induced LDL Oxidation J. Am. Coll. Nutr., August 1, 2001; 20(4): 320 - 326. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |