点击显示 收起
1 From the Cardiology Division, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Hong Kong, China (Y-HC, K-KL, K-HY, H-TC, C-PL, and H-FT); the Department of Medicine, Tung Wah Hospital, Hong Kong, China (S-WL); the Department of Clinical Biochemistry Unit, Queen Mary Hospital, Hong Kong (ST); and the Department of Medicine, Vanderbilt Epidemiology Center and Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, TN (X-OS)
2 Supported by the Committee of Research and Conference Grants Small Project Funding of University of Hong Kong (project no. 200507176137). 3 Reprints not available. Address correspondence to HF Tse, Department of Medicine, The University of Hong Kong, Room 1928, 19/F, Block K, Queen Mary Hospital, 102 Pokfulam Road, Hong Kong, China. E-mail: hftse{at}hkucc.hku.hk.
ABSTRACT
Background: Epidemiologic studies have suggested that a high phytoestrogen intake is inversely associated with cardiovascular disease risk factors and the incidence of cardiovascular events. However, the relation between the intake of isoflavone, a major component of phytoestrogen, and vascular endothelial function and the atherosclerotic burden remains unclear.
Objective: We aimed to investigate the effects of various dietary soy isoflavone intakes on brachial artery flow-mediated dilation and mean maximum carotid intima–media thickness.
Design: We studied 126 consecutive patients (
Conclusion: In persons at high risk of cardiovascular events, a greater isoflavone intake is associated with better vascular endothelial function and lower carotid atherosclerotic burden.
Key Words: Isoflavone • flow-mediated dilation • carotid intima–media thickness
INTRODUCTION
In the past few years, there has been increasing interest in the potential role of phytoestrogens in the prevention of cardiovascular disease (CVD) (1–4). In epidemiologic studies, populations with higher intakes of soy proteins have a lower incidence of cardiovascular events (5). The mechanisms of benefit of phytoestrogen and its major component—isoflavone—may be multiple (1, 2). Although isoflavone may exert no major direct influence on the lipid profile (6–8), it has potential antioxidant effects (9, 10), and it improves vascular endothelial function (11, 12). However, data on the intake of isoflavone in persons at high risk of cardiovascular events are limited. Furthermore, the relations between the dietary intake of isoflavone and vascular endothelial function and the severity of atherosclerosis have not been investigated. Therefore, the objectives of this study were to determine the dietary isoflavone intake in persons at high risk of cardiovascular events and to investigate the relations between the dietary intake of isoflavone and brachial flow-mediated dilation (FMD) and carotid intima–media thickness (IMT).
SUBJECTS AND METHODS
Subjects
Consecutive patients with documented coronary artery disease, ischemic stroke, or diabetes mellitus who are at high risk of future cardiovascular events were recruited from the Queen Mary Hospital medical outpatient clinic. Patients with dilated cardiomyopathy, significant valvular heart disease, chronic atrial fibrillation, New York Heart Association class III or IV heart failure, significant renal impairment (creatinine > 220 mmol/L), recent myocardial infarction, unstable angina, coronary revascularization, and stroke or acute heart failure within the past 6 mo and patients who declined to participate were excluded. As a result, a total of 126 persons were eligible for this study. All participants had a stable diet pattern and cardiovascular medications for 3 mo before the day of recruitment.
Written informed consent was obtained from each subject. The institutional review board of the University of Hong Kong approved the research protocol.
Demographic, dietary, and laboratory evaluations
Baseline demographic data, CVD risk factors, and cardiovascular medications were documented. CVD risk factors, including tobacco smoking, diabetes, hypercholesterolemia, hypertension, and body mass index (BMI; in kg/m2) were assessed. Fasting blood samples were obtained to measure serum concentrations of creatinine, glucose, total cholesterol, triacylglycerol, LDL, and HDL. Hypertension was defined as either resting systolic or diastolic blood pressure 140 or 90 mm Hg, respectively, at 2 different clinical visits or the use of antihypertensive medication. Diabetes mellitus was defined as serum fasting glucose 7.0 mmol/L or the use of medication. Hypercholesterolemia was defined as fasting total plasma cholesterol of 4.9 mmol/L or the use of medications. BMI was calculated. Smoking status was recorded as smoker (past or current) or nonsmoker. Education level was divided into 5 categories—ie, uneducated, primary, secondary, postsecondary, and university graduate—and was graded on a scale of 0 to 4. A family history of coronary artery disease in first-degree relatives whose disease was diagnosed before age 55 y was considered positive.
The dietary intakes of study subjects were estimated by using a food-frequency questionnaire (FFQ) designed for Chinese populations (13–16). This questionnaire was validated in the Shanghai Women's Health Study (14) and was specially structured to capture isoflavone contents in the diet; food models were used to help subjects in the quantification of their diet. The FFQ was conducted as described previously (13–16). In brief, all subjects were asked how often they ate a specific type or group of soy food, and how many liangs (1 liang = 37.8 g) they ate of the particular food(s) per unit of time (day, week, month, or year) in the 5 y before the reference date. Nutritional conversion into estimated intakes of isoflavone and soy proteins was done by using a custom-made computer software from the Shanghai Women's Health Study (14).
Vascular ultrasound examination
Vascular ultrasound was performed with a high-resolution ultrasound system (Agilent Sonos 5500; Philips, Andover, MA) with the use of a 7.5-MHz linear array transducer by 2 experienced operators without knowledge of the subjects. All of the scanned images were stored digitally and analyzed offline by the same operators, who were blinded to the identity of studied subjects.
Brachial flow-mediated dilatation
Participants were studied in the fasting state, and all vasoactive medications were withheld for 12 h before the study. As previously described (17, 18), longitudinal scans of the brachial artery were obtained at rest, and then FMD was induced by inflation of a pneumatic tourniquet placed on the forearm to a pressure of 250 mm Hg for 5 min. The cuff was then released, and serial imaging of the brachial artery was recorded for 5 min. The brachial artery was allowed to return to baseline. Finally, the brachial artery was measured again 5 min after the administrtion of 400 µg sublingual nitroglycerin spray. FMD was defined as the percentage change in brachial artery diameter from the baseline scan to 1 min after cuff deflation. Interobserver variability testing for FMD measurement found a correlation coefficient of 0.89 (P = 0.012).
Carotid intima–media thickness
All subjects were examined in a supine position, and the IMTs of the carotid arteries were measured as described previously (19, 20). Briefly, carotid IMT was determined by measuring the distance between the lumen-intima and media-adventitia border of the vascular wall with the use of electronic calipers. Each ultrasonic scan was performed in the anterior, lateral, and posterior projections of the right and left carotid arteries. Three IMT measurements were made on the near and far walls of the left and right common carotid arteries, carotid bifurcation, and internal carotid arteries. The mean maximum IMT (mmIMT) was calculated by averaging the values of maximum IMT measured from 12 previously selected segments in the carotid arteries. Plaque thickness was incorporated in IMT measurement. The intraobserver variability study for IMT measurement showed a correlation coefficient of 0.97 (P < 0.001).
Statistical analysis
Continuous variables were expressed as means ± SDs, and those variables with skewed distribution were presented as medians and 25%–75% ranges. Isoflavone and soy protein intakes were categorized by the quartile distribution of the study population in analysis. Statistical comparisons were performed by using Student's t test or Fisher's exact test, as appropriate. Comparisons of variables between different quartiles of isoflavone and soy protein intakes were performed with one-way analysis of variance (ANOVA) with post hoc Bonferroni's correction for multiple comparisons and test for linear trend. Absolute changes and 95% CIs of FMD and IMT were calculated by using univariate and multivariate linear regression analysis. Multivariate analyses were performed with a backward stepwise regression model in which each variable with P 0.05 (according to the univariate analysis) was entered into the model. The crude model included only quartiles of isoflavone and soy protein intakes. All statistical analyses were performed with SPSS software (version 13.0; SPSS Inc, Chicago, IL). P < 0.05 was considered statistically significant.
RESULTS
Baseline characteristics
The baseline characteristics of the study population are shown in Table 1. Their mean age was 66.5 ± 11.1 y; 69% of the subjects were men. All participants were at high risk of cardiovascular events: 94% had documented coronary artery diseases or ischemic stroke (or both), and 44% had diabetes. Significant proportions of subjects had impaired FMD [ie, < 3.5% (53%)] and increased IMT [ie, >1 mm (48%)].
View this table:
TABLE 1. Clinical characteristics of study population1
The daily dietary profiles of the study population as measured by FFQ were presented in Table 2. The mean and median daily isoflavone intakes were 13.7 ± 21.2 mg and 5.5 (range: 2.2–13.3) mg, respectively. The mean and median daily soy protein intakes were 2.6 ± 4.1 g and 1.2 (range: 0.4–2.8) g, respectively.
View this table:
TABLE 2. Dietary profile of the study population1
Effects of dietary intakes of isoflavone and soy protein
After adjustment for age and sex, there were no significant linear trends in systolic and diastolic blood pressure or in serum lipid concentrations with increasing quartiles of isoflavone or soy protein intake (P > 0.05 for all; Table 3). As shown in Figure 1, participants in the 4th quartile of daily isoflavone intake had a significantly higher FMD than did those in the 1st, 2nd, and 3rd quartiles (P < 0.05 for all). Furthermore, participants in the 4th quartile of daily soy protein intake had a significantly higher FMD than did those had the 1st quartile (P < 0.05; Figure 2). There was a significant linear trend of FMD in participants in different quartiles of isoflavone intake (P = 0.002) but not in those in different quartiles of soy protein intake (P = 0.095) (Table 4).
View this table:
TABLE 3. Effects of isoflavone and soy protein intake on blood pressure and serum lipid profile1
FIGURE 1.. Mean (±SEM) values of brachial flow-mediated dilation (FMD) in participants in the 1st (2.57 ± 0.43%; n = 31), 2nd (3.06 ± 0.49%; n = 32), 3rd (2.97 ± 0.35%; n = 32), and 4th (5.06 ± 0.70%; n = 31) quartiles of daily intake of isoflavone. *Significantly different from the 4th quartile, P < 0.05 (one-way ANOVA with Bonferroni's post hoc correction for multiple comparisons).
FIGURE 2.. Mean (±SEM) values of brachial flow-mediated dilation (FMD) in participants in the 1st (2.64 ± 0.43%; n = 31), 2nd (3.07 ± 0.48%; n = 32), 3rd (3.32 ± 0.50%; n = 32), and 4th (4.70 ± 0.62%; n = 31) quartiles of daily intake of soy protein. *Significantly different from the 4th quartile, P < 0.05 (one-way ANOVA with Bonferroni's post hoc correction for multiple comparisons).
View this table:
TABLE 4. Univariate and multivariate predictors of brachial flow-mediated dilation1
Similarly, there was a significant linear trend of carotid IMT in participants in different quartiles of isoflavone intake (P = 0.010) and of soy protein intake (P = 0.05) (Table 5). Nevertheless, there were no significant differences in carotid IMT among participants in different quartiles of daily isoflavone intake (Figure 3) and of daily soy protein intake (Figure 4).
View this table:
TABLE 5. Univariate and multivariate predictors for carotid intima–media thickness1
FIGURE 3.. Mean (±SEM) maximum carotid intima–media thickness (mmIMT) in participants in the 1st (1.17 ± 0.08 mm; n = 31), 2nd (1.04 ± 0.06 mm; n = 32), 3rd (1.11 ± 0.06 mm; n = 32), and 4th (0.98 ± 0.04 mm; n = 31) quartiles of daily intake of isoflavone. No significant differences were observed in mmIMT between quartiles, P > 0.05 for all (one-way ANOVA with Bonferroni's post hoc correction for multiple comparisons).
FIGURE 4.. Mean (±SEM) maximum carotid intima–media thickness (mmIMT) in participants in the 1st (1.14 ± 0.08 mm; n = 31), 2nd (1.12 ± 0.07 mm; n = 32), 3rd (1.05 ± 0.05 mm; n = 32), and 4th (0.99 ± 0.04 mm; n = 31) quartiles of daily intake of soy protein. No significant differences were observed in mmIMT between quartiles, P > 0.05 for all (one-way ANOVA with Bonferroni's post hoc correction for multiple comparisons).
Predictors for flow-mediated dilation and intima-media thickness
Univariate regression analysis found that the use of aspirin, the 4th quartile of isoflavone and soy protein intake, and dietary intakes of calorie, carbohydrate, protein, dietary fiber, and vitamins A, C, and E predicted changes in FMD (P < 0.05; Table 4). However, multivariate regression analysis showed that only the 4th quartile of isoflavone intake was an independent predictor for changes in FMD (Table 4). As compared with the 1st quartile of isoflavone intake, the 4th quartile of intake was associated with an absolute 2.71% increase in FMD (95% CI: 0.44, 4.99; P = 0.020), which accounted for a relative increase of 103%.
As shown in Table 5, univariate regression analysis found that age and a history of hypertension and smoking predicted changes in mmIMT (P < 0.05). The effect of isoflavone intake in the 4th quartile on mmIMT was marginally significant (P = 0.076). Multivariate regression analysis showed that age, history of hypertension and smoking, and the 4th quartile of isoflavone intake were independent predictors of changes in mmIMT (Table 5). As compared with the 1st quartile of isoflavone intake, the 4th quartile was associated with an absolute 0.17-mm decrease in mmIMT (95% CI: –0.34, –0.01; P = 0.043), which accounted for a relative decrease of 14.5%. Similarly, multivariate analysis did not find any significant effects of the different soy protein intakes on FMD and mmIMT (P > 0.05; Tables 4 and 5).
DISCUSSION
Endothelial dysfunction, as reflected by the impairment of endothelium-dependent FMD, has been shown to be associated with adverse clinical outcomes in persons with CVD (21–24). Furthermore, previous studies have shown that an increased carotid IMT was also an independent predictor for cardiovascular events after adjustment for traditional risk factors (23, 25). The results of this study showed that a higher daily intake of isoflavone was associated with a larger FMD and a lower mmIMT in a population with documented CVD or diabetes or both. These findings suggest that isoflavone may exert a vascular-protective effect in those high-risk populations.
Prior experimental studies showed that isoflavone had a stimulatory effect on endothelial function both in vitro (12) and in vivo (11, 26) and that it retarded the progression of atherosclerosis in nonhuman primates (27, 28). However, clinical data on the effects of isoflavone intake on vascular endothelial function and progression of atherosclerosis are quite limited. In subjects with hypercholesterolemia, an increase in soy protein intake improved endothelial function, independent of the change in total and LDL cholesterol (29, 30). In contrast, increased dietary intake of soy protein in normocholesterolemic men and postmenopausal women failed to improve endothelial function (31, 32). In the present study, a high intake of isoflavones in persons at high risk of cardiovascular events was associated with better endothelial function and less carotid atherosclerosis. Furthermore, the intake of isoflavones rather than of soy protein was predictive of brachial FMD and carotid IMT. It is possible that the beneficial effect of soy food on vascular function may be mediated by isoflavone. In fact, the administration of isoflavones has been shown to induce nitric oxide–dependent relaxation in human forearm microcirculation (11) and to improve brachial FMD in healthy postmenopausal women (33, 34).
The exact mechanisms for the vascular-protective effects of isoflavone are unknown. Soy protein with isoflavone has been shown to be associated with significant decreases in serum total and LDL cholesterol and significant increases in serum HDL cholesterol. The therapeutic effect on serum lipids appears to be directly related to the initial serum cholesterol concentration (6, 7, 35, 36). However, whether such a hypocholesterolemic effect could be attributed to isoflavone (37, 38) or to soy protein (7, 35) remains controversial (8). Studies have also suggested that isoflavone may reduce blood pressure (26, 31), but results have not been consistent (32, 39). In the present study, we did not observe any relations between dietary isoflavone intake and the serum lipid profile or systolic or diastolic blood pressure. Although the reasons for these discrepancies remain unclear, they may have to do with the fact that all subjects with hypertension or hyperlipidemia (or both) in the present study had already received antihypertensive agents, statin, or both. As a result, soy protein or isoflavone may have less apparent effects on blood pressure and serum lipids. On the other hand, these findings suggested that the effects of isoflavone on endothelial function and progression of atherosclerosis could be independent of its lipid- or blood pressure–lowering effects.
Several potential mechanisms may account for the cardiovascular benefits of isoflavone. First, isoflavone has an affinity for the estrogen receptor–ß in blood vessels similar to that of estrogen, and isoflavone has been shown to achieve vasodilatation to an extent similar to that of ß-estradiol in human subjects (11). Second, studies have shown that isoflavone prevents LDL oxidation and vascular damage in vitro (26, 40, 41) and reduces lipid peroxidation and increases LDL resistance to oxidation in vivo (9, 10). Third, besides its effects on oxidative stress and insulin resistance, isoflavone has been shown to reduce serum-soluble vascular cell adhesion molecule-1 in animals (42) and to be associated with lower circulating concentrations of tumor necrosis factor- (43) and C-reactive protein (44) in humans. Thus, it may play a favorable role in modulating immune function (45). Finally, isoflavone may affect platelet function by decreasing thromboxane A2 receptor density, thus potentially reducing the risk of platelet aggregation and vascular thrombosis (46). All of these beneficial effects in concert may contribute to the cardiovascular protectiveness of isoflavone.
It is worth noting that the isoflavone intakes in our study population were much higher than those reported from healthy Western populations (47, 48) but lower than those in a healthy Chinese population (13, 16). In the present study, the beneficial effects of isoflavone on FMD and mmIMT trended similarly and were observed only in those participants in the highest quartile of intake, which suggests a possible threshold of isoflavone intake for its vascular-protective effects. Because endothelial dysfunction and the atherosclerosis burden are independent predictors of future cardiovascular events, a higher intake of isoflavone in these high-risk populations may have a beneficial effect in reducing cardiovascular morbidity and mortality. However, in this cross-sectional study, a direct causality of the daily isoflavone intake and vascular assessments cannot be established. Therefore, future randomized controlled studies will be needed to confirm these potential beneficial effects of isoflavone intake on endothelial function and progression of atherosclerosis in persons at high risk of cardiovascular events.
ACKNOWLEDGMENTS
We thank Qiong Li for contributing data analysis to this study.
The authors' responsibilities were as follows—YHC, KKL, KHY, SWL, CPL, and HFT: designed the study, recruited the participants, performed the investigations, collected and performed the statistical processing of the data, and wrote the manuscript; ST: processed the blood samples and performed the laboratory investigations; HTC: performed the assessment of the food-frequency questionnaire; XOS: performed the statistical processing of the data, and wrote the manuscript; and all authors: reviewed and revised the manuscript. None of the authors had a personal or financial conflict of interest.
REFERENCES