Abstract
Background:
Equol, an isoflavone (ISF)-derived metabolite by the gut microbiome in certain individuals termed as equol-producers, might be the key anti-atherogenic component of ISFs. Our objective was to determine the association between equol-producing status and aortic atherosclerosis assessed as aortic calcification (AC).
Methods:
This population-based study of 302 Japanese men aged 40–49, free of cardiovascular disease, examined serum levels of equol and ISFs, AC in the entire aorta by electron-beam computed tomography with Agatston method, and cardiovascular risk factors. We defined equol-producers as individuals with serum levels of equol ≥20 nM and prevalent AC as an AC score ≥10. We analyzed the association between equol-producing status and AC using Tobit and logistic regressions. We performed age-stratified analyses since age was a significant effect-modifier.
Results:
The 60th to 90th percentile AC scores were 4 to 243 in equol-producers and 15 to 444 in non-producers, respectively. Overall, equol-producers (41% of the sample) had lower AC scores (−209, [95% confidence interval (CI): −455, 36]) and odds of AC (odds ratio (OR): 0.7 [95% CI: 0.4, 1.3]), although not statistically significant, compared to non-producers after controlling for cardiovascular risk factors. Among men aged 46–49, equol-producers had significantly lower AC scores (−428 [95% CI: −827, −29]). Furthermore, there were null associations between serum levels of ISFs and both AC score and the odds of AC.
Conclusion:
In middle-aged Japanese men, equol-producers had a non-significantly lower burden of aortic atherosclerosis than non-producers whereas ISFs had a null association. Studies with larger sample sizes in both sexes are warranted.
Keywords: Equol, soy isoflavones, atherosclerosis, aortic calcification, Japanese, epidemiology
1. Introduction
Soy isoflavones (ISFs) are non-steroidal phytoestrogens regularly consumed in East Asian countries whereas their intake in the US is very limited (an average of 25–50 mg/day in East Asia vs. <2 mg/day in the US [1–3]). As compared to estrogen which preferably binds to estrogen receptor α that are abundant in reproductive tissues, ISFs show a greater affinity toward estrogen receptors β (ER β) which are expressed in many systems including the vascular system [4, 5]. Several studies in East Asian countries reported that dietary intake of ISFs is significantly and inversely associated with incident coronary heart disease (CHD) [6–8]. However, a randomized clinical trial in the US among 350 postmenopausal women showed that ISFs intervention for 2.7 years had a null treatment effect on atherosclerosis, overall [9]. This discrepancy may be due to the higher capacity of producing equol after consuming ISFs among people living in the East Asian countries than in the US; such individuals are referred to as “equol-producers”. Equol is a metabolite of an ISF daidzein by the gut bacteria [10, 11]. Compared to ISFs, equol is more biologically active, a more potent antioxidant, and has a similar or greater affinity to ERβ and thus may have a higher anti-atherogenic effect [10, 12, 13]. In fact, a nested case-control study within a prospective cohort study in China demonstrated that urinary equol but not ISFs or their other metabolites is significantly associated with incident CHD [7]. Interestingly, 50 to 70% of the population residing in East Asian countries are equol-producers in contrast to 20 to 30% in Western countries. This is due to the differences in microbiome but not genetics [10, 11] and suggests that equol may be the key anti-atherogenic component of ISFs. High microbial diversity caused by a healthy eating pattern may enhance the quality of equol-producing bacteria and thus boost the production of equol [14].
Aortic calcification (AC) and coronary artery calcification (CAC) are biomarkers of atherosclerosis. While CAC is a powerful predictor of future cardiovascular events and significantly improves the classification of cardiovascular risk status [15], several studies have suggested that both presence and progression of AC significantly predict future cardiovascular events independent of CAC [16–18]. To date, no study has examined the association between equol producing status and AC.
This study aimed to assess the association between equol producing status and AC in a sample of middle-aged Japanese men from the ERA JUMP (Electron-Beam Tomography, Risk Factor Assessment Among Japanese and U.S. Men in the Post-World War II Birth Cohort) study. The ERA JUMP Study reported that Japanese men in Japan had a significantly lower odds of AC presence compared to White and Japanese Americans (32%, 49%, and 43%, respectively) despite similar or greater exposure to traditional cardiovascular risk factors in Japanese in Japan, including hypertension, diabetes, hypercholesterolemia and especially smoking [19, 20]. The Pathobiological Determinants of Atherosclerosis in Youth study has reported that smoking is a stronger determinant of atherosclerosis in the aorta than in the coronary artery [21]. Therefore, significantly lower AC in Japanese in Japan may suggest some protective factors in this population. In this study, we hypothesize that equol-producers had a significantly lower degree of AC than non-producers in middle-aged men in Japan.
2. Material and methods
2.1. Study population
ERA JUMP was a population-based study of 926 men aged 40 to 49 that was established to compare the levels of subclinical atherosclerosis and explore their risk factors in Japanese in Japan, White and Japanese Americans. This age group was selected because exposure to traditional risk factors was similar or worse in Japanese men in Japan than American men [19]. From 2002 to 2007, 313 Japanese men from Kusatsu, Shiga, Japan; 310 white men from Allegheny County, Pennsylvania; and 303 Japanese American men from a representative sample of offspring of fathers who participated in the Honolulu Heart Program, Honolulu, Hawaii were randomly selected. All participants were free of cardiovascular disease (CVD) or other severe diseases [22]. We originally intended to study the association between equol-producing status and AC in the three populations. However, sub-sample of White (n=57) and Japanese Americans (n=60) showed extremely low levels of ISFs and almost zero equol-producers which prohibited us to analyze the data in these two populations. Therefore, this present study is limited to the sample of Japanese in Japan.
The current study is a cross-sectional study among Japanese men from Japan. Of 313 men, we excluded men with missing AC scores (n=3) or blood levels of equol (n=8), resulting in a total sample size of 302. All participants provided informed consent. The study was approved by the Institutional Review Boards of the Shiga University of Medical Science, Otsu, Japan, and the University of Pittsburgh, Pittsburgh, USA.
2.2. Measurement of covariates
All covariates were measured at the baseline of the study. Physical examination, a questionnaire for lifestyle habits, and laboratory assessment were conducted on all participants [22]. Bodyweight and height were measured while the participant was wearing light clothing barefoot. Body mass index (BMI) was calculated as weight (kg) divided by the square of the height (m2). Blood pressure (BP) was measured after the participant emptied his bladder and sat quietly for 5 minutes. The BP was the average of the two measurements on the right arm with an automated sphygmomanometer (BP-8800, Colin Medical Technology, Komaki, Japan) using an appropriately sized cuff. Hypertension was defined as systolic blood pressure (SBP) greater than or equal to 140 mmHg, diastolic blood pressure greater than or equal to 90 mmHg, or use of anti-hypertensive medications [23]. The frequency of soy product intake in the past year was assessed through the food frequency questionnaire (FFQ). Alcohol intake was measured as grams of daily ethanol intake [24]. Pack-years of smoking were calculated as smoking years multiplied by the number of cigarettes per day divided by 20. The use of medication (antihypertensive, anti-diabetic, and lipid-lowering) was reported as “yes” or “no”. Blood samples were collected in the morning after a 12-hour fast. All the blood samples were stored at −80 degrees Celsius and shipped on dry ice to the Heinz Nutrition Laboratory, University of Pittsburgh (glucose, lipid panel, daidzein, genistein, and equol), and University of Vermont (C-reactive protein (CRP)). Serum glucose was determined using a hexokinase glucose-6- phosphate-dehydrogenase enzymatic assay. Type 2 diabetes was defined as a fasting serum glucose level greater than or equal to 126 mg/dL or use of anti-diabetic medications [25]. Serum low-density lipoprotein cholesterol (LDL-C) was determined using standardized protocols from the Center for Disease Control and Prevention [22] and estimated by the Friedewald equation [26]. When triglyceride level exceeded 400mg/dL, LDL-C was measured directly using an automated spectrophotometric assay [LDL Direct Liquid Select (Equal Diagnostics, Exton US)]. Hypercholesterolemia was defined as a fasting serum LDL-C level greater than or equal to 140mg/dL [27] or the use of lipid-lowering medication.
2.3. Measurement of serum ISFs and equol
Serum daidzein and genistein samples were collected and measured at baseline; equol sample was collected and stored at baseline but was measured in 2013. The modified Pumford method was used in the measurements [28]. Daidzein-d4, genistein-d4 (Cambridge Isotope Laboratories), and equol-d4 (Medical Isotopes Inc.) was added and the samples were incubated with beta-glucuronidase. The samples were extracted with diethyl ether, dried under N2, and silylated. The samples were analyzed by gas chromatography-mass spectrometry in the selected ion monitoring mode. Ions monitored (m/z) were: 425/482, 234/470, and 555 for daidzein, equol, and genistein, respectively; 428/485, 236/474, and 559 for daidzeind4, equol-d4, and genistein-d4, respectively. The coefficient of variation was 10, 15, and 5 percent for daidzein, genistein, and equol, respectively.
2.4. Measurement of AC
For AC measurements, an electron-beam computed tomographic (EBCT) scanner (GE Medical Systems, South San Francisco, California) was used. All images were saved to optical discs and shipped to the Cardiovascular Institute, University of Pittsburgh Medical Center, and were read by one trained reader blinded to the participants’ characteristics. For AC evaluation, 6-mm-thick transverse contiguous images were obtained from the level of the aortic arch to the iliac bifurcation using 300-ms exposure time. Calcium was considered to present in the aorta when at least three contiguous pixels of 130 Hounsfield Units were present on a 30-cm matrix. AC was quantified by the Agatston method [29]. The within-reader intra-class correlation was 0.98 [20].
2.5. Statistical analysis
Characteristics of participants were expressed as medians (interquartile range) for continuous variables and as percentages for categorical variables by equol producing status. No continuous variables except for BMI conformed to the normal distributions. The difference in characteristics between equol-producers and non-producers was tested by the Kruskal-Wallis test when the variable was continuous and by the Chi-square test when the variable was categorical. Equol-producers were defined as the blood level of equol of ≥20 nM. We defined the AC presence as AC scores of ≥10 units because scores <10 was considered artifact [30]. A cut-off point of 200 was used to distinguish “moderate” from “more advanced” AC.
To examine the association of equol producing status with the AC score, Tobit conditional regression was used because it is suited to the right-skewed AC score distribution with many zeros [31, 32]. The Tobit conditional regression thus provides an estimate with the combination of the following two regressions: i) logistic regression of the odds of AC ≥10, and ii) linear regression of log-transformed AC when AC ≥10 [33]. We also performed logistic regression to examine the association of equol producing status with prevalent AC (AC ≥10 vs. AC<10). Models were performed in the following orders: i) crude, ii) adjusted for age and the traditional CVD risk factors (diabetes, pack-year of smoking, hypercholesterolemia, and hypertension [34]), iii) further adjusted for other confounders (BMI, CRP, and alcohol consumption), iv) further adjusted for daidzein or genistein. A p-value <0.05 was considered statistically significant.
We assessed the modification impact of all risk factors (age, diabetes, BMI, pack-year of smoking, LDL-C, SBP, and CRP) by including an interaction term between each of them and the equol producing status in the fully adjusted model. The p-value for the modification effect of age was 0.09. Thus, we performed age-stratified analyses by using the median value of age (i.e., 45) as the cut-off point (40–45 vs. 46–49 years).
The correlations between levels of equol, daidzein, and genistein were assessed by Spearman correlation coefficients due to their skewed distributions. The association of the log-transformed ISFs (the sum of daidzein and genistein) with the AC score and AC presence were also assessed with linear regressions and logistic regression, respectively.
To illustrate the predicted probabilities of having more advanced AC (≥200) or moderate AC (≥10 and <200) in equol-producers and non-producers, a figure was depicted using proportional odds models with all risk factors controlled. All analyses were performed using R Statistical Software Version 4.0.2 (Foundation for Statistical Computing, Vienna, Austria). The plot was drawn by using the “effects v4.2–0” package in R.
3. Results
Among the 302 participants, 41% (n=125) were equol-producers. The characteristics of equol-producers and non-producers (n=177) are described in Table 1. There was no significant difference between equol-producers and non-producers in the frequency of pack-year of smoking, consumption of alcohol or soy products, hypertension, or hypercholesterolemia. Equol-producers compared to non-producers had significantly lower diabetes prevalence and AC scores at 70th percentile and above (Table 1).
Table 1.
Characteristics of Japanese men aged 40–49 years by equol producing status in the ERA JUMP study from 2002 to 2004, Kusatsu, Shiga, Japan (n=302)
| Equol-producers | Non-producers | ||||
|---|---|---|---|---|---|
| n=125 | n=177 | ||||
| Age, years | 46 | 43–48 | 45 | 43–47 | |
| Body mass index, kg/m2 | 24 | 22–25 | 23 | 21–25 | |
| Diabetes, % | 2 * | 9 * | |||
| Hypertension, % | 22 | 30 | |||
| Hypercholesterolemia, % | 45 | 42 | |||
| C-reactive protein, mg/dL | 0.3 | 0.2–0.7 | 0.3 | 0.2–0.7 | |
| Pack-year of smoking | 18 | 3–29 | 19 | 4–29 | |
| Alcohol consumption, g/day | 18 | 2–42 | 14 | 3–43 | |
| Soy product consumption frequency, % | |||||
| Daily | 9 | 4 | |||
| 2–5 times/week | 39 | 42 | |||
| 0–4 times/month | 52 | 54 | |||
| Equol, nM | 55* | 25–138 | 0.1 * | 0.1–5 | |
| Isoflavones, nM | 484 | 200–1227 | 472 | 182–1028 | |
| Daidzein, nM | 92 | 34–231 | 82 | 36–199 | |
| Genistein, nM | 394 | 171–923 | 357 | 156–784 | |
| AC score ≥10, % | 29 | 34 | |||
| AC score | |||||
| 50th percentile | 0 | 0 | |||
| 60th percentile | 0 | 0 | |||
| 70th percentile | 4 * | 15 * | |||
| 80th percentile | 66 * | 95 * | |||
| 90th percentile | 243 * | 444 * | |||
| 100th percentile | 1224 * | 4139 * | |||
AC: aortic calcification, IQR: interquartile ranges.
Median values and IQR are shown for continuous variable. Median values are show for AC score across percentiles.
indicates the difference by equol producer status is significant at 0.05 level. The differences were tested by Kruskal–Wallis test (continuous variables) or chi-square test (categorical variables)
Definition: equol-producers: blood level of equol ≥20 nM; diabetes: fasting serum glucose ≥126 mg/dL or taking anti-diabetes medication; hypertension: systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg or taking anti-hypertension medication; hypercholesterolemia: low-density lipoprotein-cholesterol ≥140 mg/dL or taking lipid medication
When stratified by age, participants aged 46–49 years (n=163) compared to 40–45 years (n=139) had significantly higher prevalent hypertension, pack-year of smoking, genistein levels, AC scores, and the odds of prevalent AC (Table 2). Only among participants aged 40–45 years, equol-producers had significantly lower prevalent diabetes and levels of daidzein and genistein than non-producers. The frequency of consuming soy products was not significantly different by equol-producing status in either age group.
Table 2.
Characteristics of Japanese men by age groups and equol producing status from 2002 to 2007 (n=302)
| Age 40–45 years (n=139) | Age 46–49 years (n=163) | |||||
|---|---|---|---|---|---|---|
| Overall | Equol-producer | Non-producer | Overall | Equol-producer | Non-producer | |
| Age, years | 42 | 42 | 42 | 47 | 47 | 47 |
| Body mass index, kg/m2 | 24 | 24 | 24 | 23 | 24 | 23 |
| Diabetes, % | 6 | 0 § | 9 § | 7 | 4 | 9 |
| Hypertension, % | 18 * | 13 | 20 | 33 * | 28 | 37 |
| Hypercholesterole mia, % | 39 | 42 | 38 | 46 | 46 | 46 |
| C-reactive protein, mg/dL | 0.3 | 0.3 | 0.7 | 0.3 | 0.3 | 0.3 |
| Pack-year of smoking | 14 * | 11 | 15 | 22 * | 21 | 23 |
| Alcohol consumption, g/day | 14 | 14 | 26 | 20 | 20 | 16 |
| Soy product consumption frequency, % | ||||||
| Daily | 6 | 7 | 3 | 6 | 10 | 3 |
| 2–5 times/week | 37 | 33 | 44 | 43 | 43 | 44 |
| 0–4 times/month | 57 | 60 | 53 | 51 | 48 | 53 |
| Equol, nM | 7.0 | 39 § | 0.1 § | 12 | 70 § | 0.1 § |
| Equol-producers, % | 38 | / | / | 44 | / | / |
| Isoflavones, nM | ||||||
| Daidzein, nM | 75 | 71 § | 154 § | 95 | 98 | 90 |
| Genistein, nM | 294 * | 252 § | 525 § | 405 * | 484 | 384 |
| AC score | ||||||
| 50th percentile | 0 | 0 | 0 | 0 | 0 | 0 |
| 60th percentile | 0 * | 0 | 0 | 5 * | 1.2 § | 14 § |
| 70th percentile | 0 * | 0 | 0 | 49 * | 30 | 63 |
| 80th percentile | 11 * | 5 § | 12 § | 126 * | 99 | 199 |
| 90th percentile | 113 * | 122 | 90 | 497 * | 358 | 645 |
| 100th percentile | 947 * | 629 | 957 | 4139 * | 1224 § | 4139 § |
AC: aortic calcification.
Median values are shown for continuous variable
Definition: equol-producers: blood level of equol ≥20 nM; diabetes: fasting serum glucose ≥126 mg/dL or taking anti-diabetes medication; hypertension: systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg or taking anti-hypertension medication; hypercholesterolemia: low-density lipoprotein-cholesterol ≥190 mg/dL or taking lipid medication
indicates the difference by age categories was significant at 0.05 level. The difference was tested by Chi-square test (categorical variable) or Kruskal–Wallis test (continuous variable).
indicates the difference by equol-producing status within each age category was significant at 0.05 level.
Serum levels of daidzein and genistein were significantly correlated (Spearmen correlation coefficient =0.94, p<0.001). Serum level of equol was significantly correlated with daidzein and genistein in equol-producers (Spearmen correlation coefficients =0.28, 0.21, respectively, p<0.001). Daidzein was detected in all participants.
Table 3 describes the association of equol producing status with AC. Equol-producers compared to non-producers had lower AC scores (−209, 95% CI: −455, 36, Model 2) and lower odds of AC presence (0.7, 95% CI: 0.4, 1.3, Model 2) after controlling for all risk factors, although not statistically significant. The results are robust even if we choose different cut-off points of equol-producers (Supplementary Table 1). Further adjustment for serum levels of daidzein or genistein did not materially change the results (data not shown). When the above association was examined by age sub-groups, equol-producers had significantly lower AC scores in participants aged 46–49 years (−428, 95% CI: −827, −29, p-value = 0.03, Model 2) but not in 40–45 years (8, 95% CI: −194, 210, Model 2).
Table 3.
Mean AC score, odds of prevalent AC (≥10) in equol producers as compared to non-producers (n=302)
| Overall Sample | Age 40–45 years | Age 46–49 years | |
|---|---|---|---|
| n=302 | n=163 | n=139 | |
| Coefficients (95% CI) | |||
| Crude | −174 (−441,93) | 22 (−195, 239) | −440 (−852, −27) * |
| Model 1 | −193 (−451, 51) | 5 (−195, 205) | −370 (−763, 23) |
| Model 2 | −209 (−455, 36) | 8 (−194, 210) | −428 (−827, −29) * |
| ORs (95% CI) | |||
| Crude | 0.8 (0.5, 1.3) | 1.0 (0.4, 2.0) | 0.6 (0.3, 1.1) |
| Model 1 | 0.8 (0.4, 1.3) | 0.9 (0.4, 2.0) | 0.6 (0.3, 1.4) |
| Model 2 | 0.7 (0.4, 1.3) | 0.5 (0.2, 1.2) | 0.5 (0.2, 1.2) |
CI: confidence interval, ORs: odds ratios, AC: aortic calcification
For overall sample: Model 1, adjusted for age, pack-year of smoking, hypercholesterolemia, diabetes, and hypertension; Model 2, Model 1 + body mass index, alcohol intake, and C-reactive protein
indicates that we reject the null hypothesis at a 0.05 significance level
Supplementary Figure 1 shows that the probability of having a more advanced AC (AC≥200) increased by age and the probability of no AC presence decreased by age. However, equol-producers had a slightly slower increase in the probability of having a more advanced AC accompanying aging than non-producers.
The serum level of ISFs was not associated with AC score or prevalent AC (Table 4). In the overall sample, 1 unit increase in log-transformed ISFs had 39 (95% CI: −187, 264, Model 2) higher AC scores. The OR for prevalent AC was 1.0 (95%CI: 0.6, 1.8, Model 2). The associations were also non-significant in participants aged 40–45 and 46–69. When the associations for daidzein or genistein were analyzed separately, similar null associations were observed (data not shown).
Table 4.
Mean AC score, odds of prevalent AC with 1 unit increase in log-transformed ISFs (n=302)
| Overall Sample | Age 40–45 years | Age 46–49 years | |
|---|---|---|---|
| n=302 | n=163 | n=139 | |
| Coefficients (95% CI) | |||
| Crude | 69 (−182, 320) | −65 (−261, 131) | 141 (−271, 554) |
| Model 1 | 30 (−198, 257) | −71 (−248, 106) | 243 (−153, 641) |
| Model 2 | 39 (−187, 264) | −75 (−254, 104) | 284 (−119, 686) |
| ORs (95% CI) | |||
| Crude | 1.1 (0.7, 1.8) | 0.7 (0.4, 1.5) | 1.4 (0.7, 2.8) |
| Model 1 | 1.0 (0.6, 1.8) | 0.7 (0.3, 1.4) | 1.8 (0.8, 4.3) |
| Model 2 | 1.0 (0.6, 1.8) | 0.7 (0.3, 1.4) | 1.9 (0.8, 4.6) |
CI: confidence interval, AC: aortic calcification, ISFs: soy isoflavones (daidzein + genistein), ORs: odds ratios
For overall sample: Model 1, adjusted for age, pack-year of smoking, hypercholesterolemia, diabetes, and hypertension; Model 2, Model 1 + body mass index, alcohol intake, and C-reactive protein.
For age-specific analyses, age was dropped from models
4. Discussion
This population-based cross-sectional study among Japanese men aged 40–49 years showed that equol-producers tended to have lower AC scores, lower odds of prevalent AC, and lower odds of being in a higher AC category than equol-non-producers. The difference in AC scores between equol-producers and non-producers reached statistical significance in participants aged 46–49 years. Meanwhile, serum levels of ISFs in overall or age-stratified analyses showed null associations with AC suggesting that equol may be the key anti-atherogenic factor of ISFs.
Several previous observational studies in humans and RCTs in monkeys support the equol hypothesis. A nested case-control study of CHD in 1,130 Chinese women found a significant inverse association of incident CHD with equol (OR: 0.46, 95% CI: 0.24, 0.89, comparing the lowest to highest quartile) but not with ISFs and their other metabolites [7]. We reported that that equol-producing status, but not serum levels of ISFs, was significantly and inversely associated with prevalent CAC (OR: 0.1, 95%CI: 0.01, 0.9) in Japanese men in ERA JUMP [30]. RCTs in monkeys showed that supplementation of ISFs significantly reduced the progression of atherosclerosis than placebo [35, 36]. Given that all monkeys are equol-producers, equol may contribute to the anti-atherogenic property of ISFs in these RCTs. Although secondary analyses of three RCTs of ISFs in Western countries reported no significant association of equol producing status with the progression of atherosclerosis [9, 37, 38], the negative results are very likely to be a lack of statistical power due to low rates of equol-producers in Westerners [9, 37], as well as short duration of intervention [37, 38].
We observed that equol-producers had significantly lower AC scores only among participants aged 45–49 years but not in younger age group. Although the reasons for this are unknown, this may be partly due to higher AC scores with a larger variation in older than younger age groups. On the other hand, we observed null associations between ISFs and AC in overall and age-stratified analyses. Previous observational studies of the association of dietary intake of ISFs and CHD in Asian populations showed inconsistent results. Dietary intake of ISFs was significantly inversely associated with CHD in 40,462 Japanese in Japan [6] whereas it was not in 63,257 Chinese in Singapore [39]. Much higher dietary intake of ISFs in Japanese than in Singapore Chinese may in part account for this different result. Unfortunately, equol or equol producing status was not reported in these studies.
The serum levels of daidzein and genistein were highly correlated. This observation is consistent with the results of a previous study in Japan [40]. This is expected because soy foods in the Japanese diet contain both daidzein and genistein to some extent [41]. We observed equol producers had a significantly lower rate of diabetes than non-producers. A recent meta-analysis of 15 prospective cohort studies showed that dietary intake of ISFs is inversely associated with the risk of diabetes [42]. This may raise a possibility that the observed inverse association of equol producing status with AC is mediated through diabetes. However, neither adjusting for diabetic status nor excluding diabetic participants materially changed the results.
The significant difference in AC scores by 428 units between equol-producers and non-producers in the higher age group may suggest a substantial improvement in CVD risks. The Multi-Ethnic Study of Atherosclerosis (MESA) reported that the hazard ratios (HRs) for non-fatal and fatal CVD events for abdominal AC at 242 to 1,437 units compared to 0 to 241 units were 1.87 (p=0.07) and 3.77 (p=0.002), respectively [17]. MESA also reported that the presence of thoracic AC at baseline was significantly associated with incident CHD during a mean follow-up of 4.5 years. The association remained significant even after adjusting for traditional CVD risk factors and CAC in women but not in men [18]. The Framingham Heart Study (Offspring and Third Generation cohorts) showed that the HRs (95% CI) for major CHD and major CVD incidence for each standard deviation increase in the log-transformed abdominal AC was 1.95 (1.27, 3.00) and 1.50 (1.11, 2.05), respectively [43].
Age-adjusted CHD mortality in Japan was less than a third of that in the US despite a lifetime exposure to traditional risk factors similar or worse in Japanese [44]. This low CHD mortality in Japan was unlikely to be due to genetic susceptibility because migrant studies of Japanese to the US documented a dramatic rise in CHD rates [44]. The finding from ERA JUMP are consistent with these observations: the level of subclinical atherosclerosis was significantly lower in Japanese in Japan than White and Japanese Americans despite similar levels of CVD risk factors [22]. This finding led us to investigate the difference in environmental factors, especially diet. The Japanese diet is characterized by a very high dietary intake of long-chain n-3 fatty acids and ISFs. We reported that the difference in blood levels of long-chain n-3 fatty acids partly accounted for the difference in levels of subclinical atherosclerosis between Japanese and Americans [22, 45]. The current study suggests that high dietary intake of ISFs may not be sufficient but proper processing of ISF to equol may be crucial.
Although this study examined the association of equol with atherosclerosis, other vascular effects of equol have been reported. A cross-over RCT of equol for 12 weeks among 54 overweight or obese subjects in Japan reported a significant improved cardio-ankle vascular index, a biomarker of arterial stiffness [46]. Three small RCTs of equol in the US and Japan reported significant reduction in vasomotor symptoms and hot flashes in peri- and postmenopausal women [47–49]. We recently reported that among 91 cognitively normal older adults in Japan, equol-producers had significantly lower white matter lesion volume (WML) in the brain, a biomarker of cerebral small vessel disease examined by structural brain magnetic resonance, compared to non-producers. Equol-producing status was determined 6–9 years before the imaging study. As with the current study, however, serum levels of ISFs did not relate to WML [40].
The association between equol producing status and AC remained similar after adjusting for traditional risk factors, suggesting that equol may exert its effect through pathways other than the traditional risk factors. These pathways include antioxidation, anti-inflammation, vasorelaxation and anti-calcification properties of equol [4, 5, 13]. Preclinical studies showed that equol exerts many of these effects through ERβ [50].
Our study has several limitations. First, we cannot establish causality due to the cross-sectional study design. However, equol producing status is reported to be stable over years [51] and it is unlikely that the presence of AC affected equol producing status. Second, because no optimal cut-off points for equol-producers and prevalent AC have been established, we chose cut-off points that are consistent with previous studies [52]. Nevertheless, sensitivity analyses using different cut-off points for equol producing status and prevalent AC did not materially change the results (data not shown). Third, all the participants were men. While men and women may have different association between equol and atherosclerosis, further studies in both sexes are needed. However, as previous epidemiological studies examined urinary or serum equol and their association with health outcomes have been performed in women, our study is among the first to show an association in men. Fourth, our findings may not be attributed purely to equol but phenotypes related to equol-producing status. Finally, our sample size was relatively small. The OR of AC presence was 0.7 comparing equol-producers to non-producers in the present study. To detect a significant difference of 0.7 with 80% power, we need 1000 more participants in each group.
Our study has several strengths. First, this was a community-based study that utilized a random sampling method. Second, daidzein was detected in all the participants and thus misclassification of equol-producing status was unlikely. Third, we collected blood equol as well as daidzein and genistein to delineate the association between not only equol but also ISFs with AC.
In conclusion, we found that equol-producers had non-significantly lower AC compared to non-producers whereas ISFs had a null association with AC in middle-aged Japanese men. Research on the association of equol producing status and atherosclerosis is almost impossible in Western countries because of the very limited dietary intake of ISFs and the low rate of equol-producers in Western countries. Therefore, additional studies in East Asian countries with larger sample sizes and a wider age range in both sexes are warranted.
Supplementary Material
Highlights.
Equol-producers had a lower burden of aortic atherosclerosis than non-producers
There was no association between serum soy isoflavones and aortic atherosclerosis
Equol may be the key anti-atherogenic factor of soy isoflavones
Funding
The research was funded by the National Institutes of Health (R01 HL68200 and RF1 AG051615) in the US, the Ministry of Education, Culture, Sports, Science, and Technology ((A) 25253046, (A) 23249036, (B) 24790616, (B) 21790579, (C) 21590688, (C) 23590790, and (C) 23590791) in Japan and small grant from the Department of Epidemiology, University of Pittsburgh, USA.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of interest
None declared.
References
- [1].Klein MA, Nahin RL, Messina MJ, Rader JI, Thompson LU, Badger TM, et al. Guidance from an NIH workshop on designing, implementing, and reporting clinical studies of soy interventions. The Journal of nutrition. 2010;140:1192S–204S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Valentin-Blasini L, Sadowski MA, Walden D, Caltabiano L, Needham LL, Barr DB. Urinary phytoestrogen concentrations in the U.S. population (1999–2000). J Expo Anal Environ Epidemiol. 2005;15:509–23. [DOI] [PubMed] [Google Scholar]
- [3].Liu W, Tanabe M, Harada KH, Koizumi A. Levels of urinary isoflavones and lignan polyphenols in Japanese women. Environ Health Prev Med. 2013;18:394–400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Yuan JP, Wang JH, Liu X. Metabolism of dietary soy isoflavones to equol by human intestinal microflora–implications for health. Molecular nutrition & food research. 2007;51:765–81. [DOI] [PubMed] [Google Scholar]
- [5].Mayo B, Vázquez L, Flórez AB. Equol: a bacterial metabolite from the daidzein isoflavone and its presumed beneficial health effects. Nutrients. 2019;11:2231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Kokubo Y, Iso H, Ishihara J, Okada K, Inoue M, Tsugane S. Association of Dietary Intake of Soy, Beans, and Isoflavones With Risk of Cerebral and Myocardial Infarctions in Japanese Populations. Circulation. 2007;116:553–2562. [DOI] [PubMed] [Google Scholar]
- [7].Zhang X, Gao Y-T, Yang G, Li H, Cai Q, Xiang Y-B, et al. Urinary isoflavonoids and risk of coronary heart disease. International journal of epidemiology. 2012;41:1367–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Zhang X, Shu XO, Gao Y-T, Yang G, Li Q, Li H, et al. Soy food consumption is associated with lower risk of coronary heart disease in Chinese women. The Journal of nutrition. 2003;133:2874–8. [DOI] [PubMed] [Google Scholar]
- [9].Hodis HN, Mack WJ, Kono N, Azen SP, Shoupe D, Hwang-Levine J, et al. Isoflavone soy protein supplementation and atherosclerosis progression in healthy postmenopausal women: a randomized controlled trial. Stroke. 2011;42:3168–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Setchell KD, Brown NM, Lydeking-Olsen E. The clinical importance of the metabolite equol—a clue to the effectiveness of soy and its isoflavones. The Journal of nutrition. 2002;132:3577–84. [DOI] [PubMed] [Google Scholar]
- [11].Setchell KD, Clerici C. Equol: history, chemistry, and formation. The Journal of nutrition. 2010;140:1355S–62S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Jackman KA, Woodman OL, Sobey CG. Isoflavones, equol and cardiovascular disease: pharmacological and therapeutic insights. Current medicinal chemistry. 2007;14:2824–30. [DOI] [PubMed] [Google Scholar]
- [13].Zhang X, Veliky CV, Birru RL, Barinas-Mitchell E, Magnani JW, Sekikawa A. Potential Protective Effects of Equol (Soy Isoflavone Metabolite) on Coronary Heart Diseases—From Molecular Mechanisms to Studies in Humans. Nutrients. 2021;13:3739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Yoshikata R, Myint KZ, Ohta H, Ishigaki Y. Inter-relationship between diet, lifestyle habits, gut microflora, and the equol-producer phenotype: baseline findings from a placebo-controlled intervention trial. Menopause. 2019;26:273–85. [DOI] [PubMed] [Google Scholar]
- [15].McClelland RL, Jorgensen NW, Budoff M, Blaha MJ, Post WS, Kronmal RA, et al. 10-year coronary heart disease risk prediction using coronary artery calcium and traditional risk factors: derivation in the MESA (Multi-Ethnic Study of Atherosclerosis) with validation in the HNR (Heinz Nixdorf Recall) study and the DHS (Dallas Heart Study). Journal of the American College of Cardiology. 2015;66:1643–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Kälsch H, Mahabadi AA, Moebus S, Reinsch N, Budde T, Hoffmann B, et al. Association of progressive thoracic aortic calcification with future cardiovascular events and all-cause mortality: ability to improve risk prediction? Results of the Heinz Nixdorf Recall (HNR) study. European Heart Journal-Cardiovascular Imaging. 2019;20:709–17. [DOI] [PubMed] [Google Scholar]
- [17].Criqui MH, Denenberg JO, McClelland RL, Allison MA, Ix JH, Guerci A, et al. Abdominal aortic calcium, coronary artery calcium, and cardiovascular morbidity and mortality in the Multi-Ethnic Study of Atherosclerosis. Arteriosclerosis, thrombosis, and vascular biology. 2014;34:1574–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Budoff MJ, Nasir K, Katz R, Takasu J, Carr JJ, Wong ND, et al. Thoracic aortic calcification and coronary heart disease events: the multi-ethnic study of atherosclerosis (MESA). Atherosclerosis. 2011;215:196–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Sekikawa A, Horiuchi B, Edmundowicz D, Ueshima H, Curb J, Sutton-Tyrrell K, et al. A “natural experiment” in cardiovascular epidemiology in the early 21st century. Heart. 2003;89:255–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].El-Saed A, Curb JD, Kadowaki T, Okamura T, Sutton-Tyrrell K, Masaki K, et al. The prevalence of aortic calcification in Japanese compared to white and Japanese-American middle-aged men is confounded by the amount of cigarette smoking. International journal of cardiology. 2013;167:134–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].McGill HC Jr, McMahan CA, Malcom GT, Oalmann MC, Strong JP. Effects of serum lipoproteins and smoking on atherosclerosis in young men and women. Arteriosclerosis, thrombosis, and vascular biology. 1997;17:95–106. [DOI] [PubMed] [Google Scholar]
- [22].Sekikawa A, Curb JD, Ueshima H, El-Saed A, Kadowaki T, Abbott RD, et al. Marine-derived n-3 fatty acids and atherosclerosis in Japanese, Japanese-American, and white men: a cross-sectional study. Journal of the American College of Cardiology. 2008;52:417–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Weber MA, Schiffrin EL, White WB, Mann S, Lindholm LH, Kenerson JG, et al. Clinical practice guidelines for the management of hypertension in the community: a statement by the American Society of Hypertension and the International Society of Hypertension. Journal of hypertension. 2014;32:3–15. [DOI] [PubMed] [Google Scholar]
- [24].Mahajan H, Choo J, Masaki K, Fujiyoshi A, Guo J, Hisamatsu T, et al. Association of alcohol consumption and aortic calcification in healthy men aged 40–49 years for the ERA JUMP Study. Atherosclerosis. 2018;268:84–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Mellitus D. Diagnosis and classification of diabetes mellitus. Diabetes care. 2005;28:S5–S10. [DOI] [PubMed] [Google Scholar]
- [26].Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clinical chemistry. 1972;18:499–502. [PubMed] [Google Scholar]
- [27].Inadera H, Hamazaki T. Cholesterol controversy: cutoff point of low-density lipoprotein cholesterol level in Guidelines by Japan Atherosclerosis Society. Nihon eiseigaku zasshi Japanese journal of hygiene. 2010;65:506. [DOI] [PubMed] [Google Scholar]
- [28].Pumford SL, Morton MM, Turkes A, Griffiths K. Determination of the isoflavonoids genistein and daidzein in biological samples by gas chromatography-mass spectrometry. Annals of clinical biochemistry. 2002;39:281–92. [DOI] [PubMed] [Google Scholar]
- [29].Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. Journal of the American College of Cardiology. 1990;15:827–32. [DOI] [PubMed] [Google Scholar]
- [30].Ahuja V, Miura K, Vishnu A, Fujiyoshi A, Evans R, Zaid M, et al. Significant inverse association of equol-producer status with coronary artery calcification but not dietary isoflavones in healthy Japanese men. British Journal of Nutrition. 2017;117:260–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [31].Reilly MP, Wolfe ML, Localio AR, Rader DJ. Coronary artery calcification and cardiovascular risk factors: impact of the analytic approach. Atherosclerosis. 2004;173:69–78. [DOI] [PubMed] [Google Scholar]
- [32].Tobin J Estimation of relationships for limited dependent variables. Econometrica: journal of the Econometric Society. 1958:24–36. [Google Scholar]
- [33].Martin SS, Qasim AN, Mehta NN, Wolfe M, Terembula K, Schwartz S, et al. Apolipoprotein B but not LDL cholesterol is associated with coronary artery calcification in type 2 diabetic whites. Diabetes. 2009;58:1887–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Greenland P, Knoll MD, Stamler J, Neaton JD, Dyer AR, Garside DB, et al. Major risk factors as antecedents of fatal and nonfatal coronary heart disease events. Jama. 2003;290:891–7. [DOI] [PubMed] [Google Scholar]
- [35].Clarkson TB, Anthony MS, Morgan TM. Inhibition of postmenopausal atherosclerosis progression: a comparison of the effects of conjugated equine estrogens and soy phytoestrogens. The Journal of Clinical Endocrinology & Metabolism. 2001;86:41–7. [DOI] [PubMed] [Google Scholar]
- [36].Anthony MS, Clarkson TB, Bullock BC, Wagner JD. Soy protein versus soy phytoestrogens in the prevention of diet-induced coronary artery atherosclerosis of male cynomolgus monkeys. Arteriosclerosis, thrombosis, and vascular biology. 1997;17:2524–31. [DOI] [PubMed] [Google Scholar]
- [37].Curtis PJ, Potter J, Kroon PA, Wilson P, Dhatariya K, Sampson M, et al. Vascular function and atherosclerosis progression after 1 y of flavonoid intake in statin-treated postmenopausal women with type 2 diabetes: a double-blind randomized controlled trial. The American of Clinical Nutrition. 2013;97:936–42. [DOI] [PubMed] [Google Scholar]
- [38].Liu Z-m, Ho SC, Chen Y-m, Ho S, To K, Tomlinson B, et al. Whole soy, but not purified daidzein, had a favorable effect on improvement of cardiovascular risks: a 6-month randomized, double-blind, and placebo-controlled trial in equol-producing postmenopausal women. Molecular nutrition & food research. 2014;58:709–17. [DOI] [PubMed] [Google Scholar]
- [39].Talaei M, Koh W-P, van Dam RM, Yuan J-M, Pan A. Dietary soy intake is not associated with risk of cardiovascular disease mortality in Singapore Chinese adults. The Journal of nutrition. 2014;144:921–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [40].Sekikawa A, Higashiyama A, Lopresti BJ, Ihara M, Aizenstein H, Watanabe M, et al. Associations of equol-producing status with white matter lesion and amyloid-β deposition in cognitively normal elderly Japanese. Alzheimer’s & Dementia: Translational Research & Clinical Interventions. 2020;6:e12089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [41].Yamamoto S, Sobue T, Sasaki S, Kobayashi M, Arai Y, Uehara M, et al. Validity and reproducibility of a self-administered food-frequency questionnaire to assess isoflavone intake in a Japanese population in comparison with dietary records and blood and urine isoflavones. The Journal of nutrition. 2001;131:2741–7. [DOI] [PubMed] [Google Scholar]
- [42].Tang J, Wan Y, Zhao M, Zhong H, Zheng J-S, Feng F. Legume and soy intake and risk of type 2 diabetes: A systematic review and meta-analysis of prospective cohort studies. The American journal of clinical nutrition. 2020;111:677–88. [DOI] [PubMed] [Google Scholar]
- [43].Hoffmann U, Massaro JM, D’Agostino RB Sr, Kathiresan S, Fox CS, O’Donnell CJ. Cardiovascular event prediction and risk reclassification by coronary, aortic, and valvular calcification in the Framingham Heart Study. Journal of the American Heart Association. 2016;5:e003144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Sekikawa A, Miyamoto Y, Miura K, Nishimura K, Willcox BJ, Masaki KH, et al. Continuous decline in mortality from coronary heart disease in Japan despite a continuous and marked rise in total cholesterol: Japanese experience after the Seven Countries Study. International journal of epidemiology. 2015;44:1614–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [45].Sekikawa A, Miura K, Lee S, Fujiyoshi A, Edmundowicz D, Kadowaki T, et al. Long chain n-3 polyunsaturated fatty acids and incidence rate of coronary artery calcification in Japanese men in Japan and white men in the USA: population based prospective cohort study. Heart. 2014;100:569–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [46].Usui T, Tochiya M, Sasaki Y, Muranaka K, Yamakage H, Himeno A, et al. Effects of natural S-equol supplements on overweight or obesity and metabolic syndrome in the Japanese, based on sex and equol status. Clinical endocrinology. 2013;78:365–72. [DOI] [PubMed] [Google Scholar]
- [47].Aso T Equol improves menopausal symptoms in Japanese women. J Nutr. 2010;140:1386S–9S. [DOI] [PubMed] [Google Scholar]
- [48].Ishiwata N, Melby MK, Mizuno S, Watanabe S. New equol supplement for relieving menopausal symptoms: randomized, placebo-controlled trial of Japanese women. Menopause. 2009;16:141–8. [DOI] [PubMed] [Google Scholar]
- [49].Jenks BH, Iwashita S, Nakagawa Y, Ragland K, Lee J, Carson WH, et al. A pilot study on the effects of S-equol compared to soy isoflavones on menopausal hot flash frequency. J Womens Health (Larchmt). 2012;21:674–82. [DOI] [PubMed] [Google Scholar]
- [50].Muthyala RS, Ju YH, Sheng S, Williams LD, Doerge DR, Katzenellenbogen BS, et al. Equol, a natural estrogenic metabolite from soy isoflavones: convenient preparation and resolution of R-and S-equols and their differing binding and biological activity through estrogen receptors alpha and beta. Bioorganic & medicinal chemistry. 2004;12:1559–67. [DOI] [PubMed] [Google Scholar]
- [51].Setchell KD, Clerici C. Equol: pharmacokinetics and biological actions. The Journal of nutrition. 2010;140:1363S–8S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [52].Morton MS, Arisaka O, Miyake N, Morgan LD, Evans BA. Phytoestrogen concentrations in serum from Japanese men and women over forty years of age. The Journal of nutrition. 2002;132:3168–71. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.