Little is known about the association between equol and bioavailable testosterone (BT) in adults. In this study, we examined the associations of urinary equol concentrations with serum concentrations of total, bioavailable and free testosterone (FT), dehydroepiandrosterone sulfide (DHEAS), free androgen index (FAI) and sex hormone-binding globulin (SHBG). This cross-sectional study included 1,904 women with a mean age of 59.7 years. Urinary equol concentrations were measured using high-performance liquid chromatography–tandem mass spectrometry (HPLC–MS/MS). The serum androgenic indices and SHBG were also determined. Overall, urinary equol tended to be inversely associated with bioactive forms of androgenic indices (BT, FT or FAI) but not with total testosterone (TT) or DHEAS. Urinary equol was also positively associated with SHBG. In multi-covariate-adjusted analyses stratified by menopausal status, graded and inverse associations between urinary equol and bioactive forms of androgenic indices (BT, FT and FAI) were observed in postmenopausal women (all p-trends < 0.05), but not in premenopausal women. A significant positive association between urinary equol and SHBG was observed only in postmenopausal women. No significant associations were observed between urinary equol and TT or DHEAS in either group. A path analysis indicated that these associations of equol with androgens in postmenopausal women might be mediated by SHBG. In conclusion, urinary equol exhibited graded and inverse associations with BT or FT, but not TT in women. However, further longitudinal studies of human patients are needed to confirm these results and overcome the limitations of cross-sectional studies.

ANDROGEN DEFICIENCY has been favorably associated with the decreased risk of some chronic diseases, including cardiovascular diseases [1], type 2 diabetes mellitus [2] and cancer [3] in women. Slightly less than 50% of the total testosterone (TT) in circulation is bound tightly to sex hormone-binding globulin (SHBG) [4], while approximately 50% and 2% of TT exists in a loose albumin-bound or free state [5]. These latter states retain biological activity [6] and are classified as bioavailable testosterone (BT) [4]. Notably, BT, but not TT, plays predominant roles in the development of chronic diseases [2, 7, 8]. Therefore, it is important to explore BT as a modifiable factor in humans. Although dietary factors have been identified as important modifiable factors in the risks of chronic diseases [9], few human studies have associated dietary factors with BT levels.

Phytoestrogens, which share chemical structural similarities with estrogens [10], exert estrogenic or anti-estrogenic effects by binding to estrogen receptors (ER-α and -β) [11]. Soy isoflavones (e.g., genistein, daidzein and glycetein) are major sources of phytoestrogens in standard diets. Previous studies of the effects of isoflavones in soy protein or isolated extracts on the circulating levels of TT and FT in women have reported inverse [12] and null [13] associations. Equol, a major gut flora-dependent metabolite of daidzein [14], has a longer half-life [10] and stronger estrogenic/anti-estrogenic properties, compared with daidzein and genistein [11, 15]. A few in vitro and animal studies revealed that equol exerted modulating hormone-like effects (e.g., testosterone [16-20], and SHBG [21, 22]) and could alter the blood levels of androgens [23]. Accordingly, equol might affect blood androgens in humans. In light of the estrogenic effects of equol, higher circulating levels of equol might prevent against bone loss [24] and some chronic diseases (e.g., type 2 diabetes mellitus [25], cardiometabolic risk markers [26] and coronary heart disease [27]) in postmenopausal women due to estrogen deficiency. On the other hand, equol might decrease the estrogen-related cancer (e.g., breast cancer [28]) in women via its anti-estrogenic effects. Therefore, elevated circulating/urinary equol concentrations might be beneficial to women’s health. To date, however, few studies have verified the associations between equol and androgens in women. Although in vitro and animal studies indicated the modulating hormone-like effects of equol, observational [29, 30] but not interventional [31, 32] studies showed inverse associations between equol and androgens in women. It remains uncertain whether null associations between equol and androgens, which were mainly reported in studies on Western populations, could be attributable to a limited effect size, shorter interventional period (3 days [31, 32]) or lower proportion of equol excretors (20–35% [33]). No previous study has directly examined the relationship between circulating equol and BT in women, and therefore the associations between equol and androgens remain largely unclear.

In this community-based cross-sectional study, we examined the potential associations of urinary equol with BT and other androgenic indices in a sample of middle-aged and older Chinese women.

This community-based cross-sectional study was based on the Guangzhou Nutrition and Health Study (GNHS). The GNHS was a prospective study designed to explore the determinants (e.g., lifestyle, dietary and biochemical indices) of common chronic diseases in Guangzhou, Guangdong province, China. The study design was described in detail previously [34]. As shown in Fig. 1, 4,048 community-dwelling residents who had lived in Guangzhou for at least 5 years were enrolled between 2008 and 2013. The participants had a mean age of 60.7 years. They were followed up approximately every 3 years until May 31, 2017. Of the participants, 3,169 were recruited between 2008 and 2010 and 2,465 completed the first follow-up visit between 2011 and 2013. A further 879 participants were recruited and completed the baseline survey between 2012 and 2013. A total of 2,144 subjects were excluded for the following reasons: (1) self-reported chronic renal dysfunction (n = 4) or cancer (n = 19); (2) refusal (n = 710); (3) emigration or lost to follow-up (n = 423); (4) death or serious disease (n = 25); (5) incomplete data (n = 81); and (6) male sex (n = 882). The present study included 1,904 women, of whom 1,666 were postmenopausal.

Fig. 1

Flow chart of the study subjects

The GNHS was registered at ClinicalTrials.gov (ID: NCT03179657) and approved by the Ethics Committee of the School of Public Health of Sun Yat-sen University. All procedures were performed in accordance with the relevant laws and guidelines and the Code of Ethics of the World Medical Association (Declaration of Helsinki). Consent has been obtained from each subject after full explanation of the purpose and nature of all procedures used. Written informed consent was obtained from all participants.

Face-to-face interviews based on a structured questionnaire were conducted by trained interviewers. We collected detailed information from all participants, including socio-demographic characteristics, lifestyle factors, physical activity and menopausal status (women only). Marital status was classified as unmarried (single, divorced or widowed) or being married. Daily physical activity was estimated using a physical activity questionnaire containing 19 items, and the metabolic equivalent (MET) intensities were calculated after excluding sleeping and sitting [35]. Each participant’s body hipline and waistline were measured and used to calculate the waist to hip ratio (WHR). Participants who had drunk wine ≥1 times daily for at least 6 consecutive months were defined as current drinkers. Participants who had a smoked >5 cumulative packs of cigarettes during the past year were defined as smokers.

At the examination site, fasting venous blood samples were collected from subjects during recruitment or follow-up visits and subjected to an analysis of sex hormones. Within 2–4 h of collection, blood samples were centrifuged at 1,500 × g and 4°C for 15 minutes to separate the serum. All samples were stored at –80°C. The serum concentrations of TT, dehydroepiandrosterone sulfide (DHEAS) and SHBG were detected using chemiluminescent enzyme immunometric assay kits from Abbott Corporation (Chicago, IL, USA) and an ARCHITECT I2000 automated analyzer (Abbott Corporation). The intra-assay coefficients of variation (CV) of TT, DHEAS and SHBG were 5.90%, 4.79% and 4.64%, respectively. The FT and BT levels and free androgen index (FAI) were calculated as described by Vermeulen et al. [36].

Urinary equol levels were measured using high-performance liquid chromatography–tandem mass spectrometry (HPLC–MS/MS) methods reported by previous studies [37, 38], with some modifications. Briefly, a 200 μL aliquot of urine was mixed with 3 μL of β-glucuronidase (Helix pomatia; Sigma, St. Louis, MO, USA) in 300 μL of sodium acetate buffer (0.2 mol/L, pH = 5.5). The mixture was vortexed (10 minutes) and then incubated for 30 min at 37°C. After a 10-min cool-down, the mixture was combined with 1 mL of ethyl acetate via shaking for 30 min and centrifuged at 4,000 × g and 4°C for 15 min, after which the supernatant was collected. After repeated extractions, the supernatant was dried under nitrogen and the residue was then reconstituted with 50 μL of ethyl acetate. Next, the urinary concentration of equol was analyzed using triple quad-liquid chromatography/mass spectrometry (Agilent Technologies, Santa Clara, CA, USA) with an Eclipse Plus C18 column (3.5 μm, 2.1 mm × 150 mm; Agilent Technologies). The mobile phase comprised acetonitrile (A) and a 0.1% aqueous solution of formic acid (B). The elution gradient was as follows: 45% A at 0–1 min, 45–80% A at 1–2.5 min, 80% A at 2.5–5 min and 80–45% A at 5–8 min, followed by a 6-min re-equilibration of the gradient. The flow rate was 1μL/min, and the sample injection volume was 5 μL.

The urinary creatinine level was measured using a Roche-Cobas Mira Plus clinical chemistry auto analyzer (Roche Diagnostics, Basel, Switzerland). For normalization, the equol concentrations were divided by the creatinine concentrations and were expressed as mg/mol of creatinine. The CVs of the equol and creatinine measurement methods were 8.67% and 1.20%, respectively.

The baseline characteristics of the participants were presented as means (standard deviations, SDs) or medians (interquartile ranges) for continuous variables and as frequencies (percentages) for categorical variables after stratification by menopausal status. The androgenic indices and SHBG concentrations were subjected to logarithmic transformation to achieve approximately normal distributions, and the resulting Z-scores were used in path analyses.

The subjects were categorized into tertiles according to urinary equol concentrations. Multivariate analyses of covariance (ANCOVAs) were performed to compare the mean differences in androgenic indices (TT, FT, BT, DHEAS and FAI) and SHBG levels and to test for trends according to the urinary equol tertiles. The Bonferroni test was used for pair-wise comparisons between tertiles. The percentage differences in the levels of androgenic indices (TT, BT, FT, DHEAS and FAI) and SHBG between the top and bottom urinary equol tertiles were calculated.

Stratified analyses were conducted according to the menopausal status (postmenopausal vs. premenopausal). Model 1 was adjusted for age. For the overall population and premenopausal women, Model 2 was adjusted for age, WHR, alcohol consumption, smoking status, marriage, physical activity, energy intake and oral estrogen use. For postmenopausal women, Model 2 was further adjusted for the years since menopause.

Path analysis was used to evaluate the mediating effects of SHBG on the associations of equol with androgenic indices (BT, FT and FAI) and was adjusted for the multi-covariates listed in Model 2. In the data analysis, bootstrapping was performed using the SPSS add-on PROCESS (Version 2.16.3) for mediation analysis, according to the method reported by Hayes [39]. Equol was entered as variable “X” (i.e., predictor), while the androgenic indices (BT, FT and FAI) were entered as ”Y” variables (i.e., outcomes). SHBG was entered as variable ”M” (i.e. the mediator). A two-tailed p value < 0.05 was considered statistically significant. All analyses were performed using IBM SPSS Statistics 21.0 for Windows (IBM, Inc., Armonk, NY, USA).

The baseline characteristics of the total study population and of postmenopausal and premenopausal women are listed in Table 1. The total population included 1,904 women, including 1,666 postmenopausal and 238 premenopausal women. The mean ages of the total, postmenopausal and premenopausal groups were 59.7, 60.0 and 58.3 years, respectively. The majority of women in this study were married, did not smoke or drink and did not use oral estrogen supplements. Higher median urinary equol concentrations were observed in postmenopausal women than in total and premenopausal women (0.42 vs. 0.33 and 0.02 mg/mol creatinine, respectively).

Overall, urinary equol tended to be inversely associated with the bioactive forms of androgenic indices (BT, FT or FAI), but not with TT and DHEAS, across the two models. Urinary equol also exhibited a positive association with SHBG. Among total women (Table 2), the age-adjusted mean differences between urinary equol tertiles 3 and 1 were –2.66% for BT, –5.61% for FT, –16.00% for FAI and 1.73% for SHBG. After further adjusting for the other covariates in Model 2, significant differences of –4.09% for FT and –11.82% for FAI were observed between tertiles 3 and 1 (p-trends = 0.046–0.049).

In analyses stratified by menopausal status, graded and inverse associations between urinary equol and the bioactive forms of androgenic indices (BT, FT and FAI) were observed in postmenopausal women according to both Model 1 and Model 2 (all p-trends < 0.05) (Table 3), but were not observed in premenopausal women (Table 4). Similarly, a significant positive association between urinary equol and SHBG was observed only in postmenopausal women. No significant associations were observed with TT and DHEAS in either group.

A path analysis revealed that the above-mentioned associations of equol with androgenic indices in postmenopausal women might be mediated by SHBG. An analysis with adjustments for multiple covariates demonstrated that urinary equol might directly influence SHBG (Fig. 2). The SHBG concentration correlated positively with a higher urinary equol concentration (all r = 0.053, p < 0.05) but inversely with the androgenic indices (r_BT = –0.615, r_FT = –0.624 and r_FAI = –0.758, all p < 0.001).

Fig. 2

Path analyses of the associations of equol and the potential mediator (SHBG) with androgenic indices (BT, FT and FAI) in postmenopausal women.

FT, free testosterone; BT, bioavailable testosterone; FAI, free androgen index; SHBG, sex hormone-binding globulin. All variables are presented as the Z-scores of log-transformed values. Age, WHR, alcohol consumption, smoking status, marriage, physical activity, energy intake, oral estrogen use and the years since menopause were adjusted in the analysis.

*p < 0.05, *** p < 0.001.

In this community-based cross-sectional study of women, we observed graded and inverse associations of urinary equol with concentrations of circulating BT or FT, but not TT. The inverse associations appeared to be mediated (at least in part) by SHBG. To the best of the authors’ knowledge, this study was the first to examine the associations between urinary equol and bioactive androgens in a sample of women before and after stratification by menopausal status. Our results suggested that equol might influence the blood levels of androgens, and particularly the bioactive forms.

One meta-analysis of 36 interventional studies reported the null effects of soy protein and soy isoflavones on the concentrations of TT, SHBG or BT in men [40]. Although another study reported similar null effects of soy protein and/or isoflavone interventions on estrogen and SHBG levels in both pre- and postmenopausal women, isoflavones reduced the levels of follicle-stimulating hormone and luteinizing hormone (secondary outcomes) in premenopausal women [41]. However, the effects of isoflavones on the circulating androgen levels in women remained largely unknown. It also remains uncertain whether these null effects of soy protein and/or isoflavones on androgens and estrogens could be attributable to a small effect size, short intervention period (generally 4–26 weeks [40, 41]) or lower proportion of equol excretors (20–35% [33]) in Western populations.

Equol has a longer half-life and higher apparent bioavailability than its precursor daidzein [42, 43] and exhibits superior estrogenic properties relative to other isoflavone metabolites [15]. Accordingly, interest in the health benefits of equol is increasing. Higher equol concentrations might prevent against bone loss [24] and some chronic diseases (e.g., type 2 diabetes mellitus [25], cardiometabolic risk markers [26] and coronary heart disease [27]) in postmenopausal women due to estrogen deficiency by its estrogenic effects. Besides, equol also might inhibit the occurrence and development of the estrogen-related cancer (e.g., breast cancer [28]) in women via its anti-estrogenic effects. However, few studies have examined the associations of circulating equol with androgens or SHBG. In women, three previous studies examined differences in the blood TT or FT levels between equol excretors and non-excretors [30-32]. In one cross-over trial of 14 premenopausal women who received 0.15, 1.0 and 2.0 mg of isoflavones/kg/day for 9 days, separated by approximate wash out intervals of 3 weeks, Duncan et al. observed lower concentrations of TT in equol excretors than in non-excretors (mean values: 0.70–0.73 vs. 1.04–1.05 nmol/L) [30]. However, no significant differences in serum TT and FT levels were observed between equol excretors and non-excretors in a study of 199 premenopausal women challenged with a single soy bar per day for 3 days [31] or a study of 152 women challenged with 93 mg of daidzein per day [32]. A cross-sectional study of 194 pregnant women reported a negative correlation between maternal urinary equol and TT levels (r = –0.16, p < 0.05) [29]. However, previous studies demonstrated that bioactive T, rather than TT, played a predominant role in the development of chronic diseases [3, 7, 8].

This study was the first to examine the associations of urinary equol with bioactive androgens in total, postmenopausal and premenopausal women. Notably, we observed inverse associations of equol with BT or FT, but not TT in postmenopausal women. Our results suggest that this highly bioactive metabolite of daidzein might be a determinant of circulating BT and FT concentrations in postmenopausal women. The inconsistent results of our and previous studies may be attributable to several factors. Interestingly, the inverse associations were more likely to be observed in Asian populations ([29] and this study than in Western populations [30-32]), in studies with a longer intervention period [30] (vs. short periods [31, 32]) or with BT (vs. TT in this study). In addition, some reasons might explain the nonsignificant associations of urinary equol with androgenic indices and SHBG in premenopausal women. Firstly, the small sample size (n = 238) in premenopausal women. Next, estrogen levels were higher in premenopausal women than that in postmenopausal women. Equol tended to exert anti-estrogenic effects in premenopausal women with high estrogen levels, while had estrogenic effects in postmenopausal women with low estrogen levels [43, 44]. A previous study observed that the production of SHBG was mediated by estrogen actions, and demonstrated the hypothesis of estrogen dependent stimulation of SHBG production [45]. Therefore, significant associations of equol with androgenic indices or SHBG would be more likely to be found in postmenopausal women than that in premenopausal women. Given the limited available evidence, further longitudinal studies, and particularly randomized controlled trials (RCTs), are needed to confirm the effects of equol.

Several biological mechanisms might explain the observed association between equol and BT in postmenopausal women. One in vitro study showed that equol decreased the mRNA expression of CYP11A1, a key enzyme in the conversion of cholesterol to pregnenolone. As pregnenolone is the steroid precursor of testosterone, the downregulation of CYP11A1 reduced the production of testosterone in an in vitro ovarian follicle culture system [16]. Additionally, equol directly diminished the ability of antral follicles to produce testosterone by increasing the ratio of Bax to Bcl2 mRNA in these follicles and thus dysregulating the apoptotic pathway [16]. Equol also inhibited the secretion of testosterone from antral follicles by reducing the levels of estrogens, thus inhibiting the production of cyclic adenosine monophosphate [46, 47].

This study of the testosterone status in women expanded beyond the traditional biomarkers to include SHBG. Notably, we observed a positive association between the levels of urinary equol and SHBG. Both in vitro and in vivo evidence from previous studies demonstrated that equol could increase SHBG production in HepG2 liver carcinoma cells [21, 22, 30]. This increased production was attributed to upregulated SHBG mRNA expression in response to elevated triiodothyronine levels [48]. Ultimately, the levels of bound-testosterone and BT may increase and decrease, respectively, in response to elevated SHBG levels. Our path analysis also identified mediating role of SHBG on the associations of equol with various androgenic indices (BT, FT and FAI) in postmenopausal women.

This study had several strengths. To the best of our knowledge, this was the first study to evaluate the associations of urinary equol with serum BT levels in women. Previous studies indicated that BT, but not TT, was positively associated with the development of chronic diseases in women, such as non-alcoholic fatty liver disease, type 2 diabetes mellitus and cancer [3, 7, 8]. Our results might elucidate the health effects of equol on chronic diseases via BT and may support interventions intended to modulate the levels of BT and SHBG. Next, the relatively large study size permitted us to detect potential associations. Third, the urinary equol concentrations of participants consuming normal diets would more closely represent habitual equol exposure than the concentrations in subjects challenged with isoflavones. Fourth, the observation of similar results for different bioavailable androgenic indices reduced the probability of falsely significant results.

However, some limitations of our study should be considered. First, we did not examine causality because of the cross-sectional study design. However, equol reaches a steady-state in healthy adults (terminal elimination half-life: 7–8 h), and this state is mainly influenced by dietary isoflavones rather than sex hormones. Moreover, populations with equol synthesis ability were relatively steady [33]. Therefore, cause-and-effect correlations were less likely to be inverted. Second, we collected only first-morning urine samples, which could not accurately reflect equol concentrations in subjects compared with 24 hr collected urinary samples. However, a high correlation was found between overnight spot urine and 24 hr collected urinary samples for equol levels (r = 0.93) [49]. Third, the androgenic indices and SHBG levels were measured only once in our study. Day-to-day fluctuations in these levels might attenuate the underlying associations. Finally, FT and BT were calculated in this study, rather than directly detected. However, previous studies confirmed the existence of strong correlations between the calculated and measured values of FT (r = 0.97) [36] and BT (r = 0.88) [50].

Generally, our study of community-dwelling women identified graded and inverse associations of the urinary equol concentrations with the concentrations of BT or FT, but not TT. Further longitudinal studies, and particularly RCTs, are warranted to validate these results and overcome the limitations of cross-sectional studies.

ANCOVAs, analyses of covariance; BT, bioavailable testosterone; CV, coefficients of variation; DHEAS, dehydroepiandrosterone sulfide; ER, estrogen receptors; FAI, free androgen index; FT, free testosterone; GNHS, Guangzhou Nutrition and Health Study; HPLC–MS/MS, high-performance liquid chromatography-tandem mass spectrometry; MET, metabolic equivalent; RCTs, randomized controlled trials; SHBG, sex hormone-binding globulin; TT, total testosterone; SDs, standard deviations; WHR, waist to hip ratio.

We thank all staff and participants in the GNHS cohort study gratefully. This work was supported by the National Science Foundation of China (grant numbers 81773416, 81730090 and 81472965); and the 5010 Program for Clinical researches of the Sun Yat-sen University (grant number 2007032). The funders had no role in the design, analysis or writing of this article.

None of the authors have any potential conflicts of interest associated with this research.

Yuming Chen conceived and designed the research; Hongli Dong, Feng Xiong, Qingwei Zhong, Yihong Li and Meng Liu collected the data; Hongli Dong performed the data analysis and drafted this paper, Yuming Chen, Xinyi Tang and Wenhua Ling critically revised the manuscript, and Yuming Chen had primary responsibility for final content. All authors read and approved the final manuscript.

• 1   Zhao  D,  Guallar  E,  Ouyang  P,  Subramanya  V,  Vaidya  D, et al. (2018) Endogenous sex hormones and incident cardiovascular disease in post-menopausal women. J Am Coll Cardiol 71: 2555–2566.
• 2   Kalish  GM,  Barrett-Connor  E,  Laughlin  GA,  Gulanski  BI (2003) Association of endogenous sex hormones and insulin resistance among postmenopausal women: results from the postmenopausal estrogen/progestin intervention trial. J Clin Endocrinol Metab 88: 1646–1652.
• 3   Michels  KA,  Brinton  LA,  Wentzensen  N,  Pan  K,  Chen  C, et al. (2019) Postmenopausal androgen metabolism and endometrial cancer risk in the women’s health initiative observational study. JNCI Cancer Spectr 3: pkz029.
• 4   Dunn  JF,  Nisula  BC,  Rodbard  D (1981) Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma. J Clin Endocrinol Metab 53: 58–68.
• 5  Pardridge WM (1986) Serum bioavailability of sex steroid hormones. Clin Endocrinol Metab 15: 259–278.
• 6   Vermeulen  A,  Verdonck  L,  Kaufman  JM (1999) A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 84: 3666–3672.
• 7   Goodman-Gruen  D,  Barrett-Connor  E (2000) Sex differences in the association of endogenous sex hormone levels and glucose tolerance status in older men and women. Diabetes Care 23: 912–918.
• 8   Klisic  A,  Kavaric  N,  Jovanovic  M,  Soldatovic  I,  Gligorovic-Barhanovic  N, et al. (2017) Bioavailable testosterone is independently associated with fatty liver index in postmenopausal women. Arch Med Sci 13: 1188–1196.
• 9   Yakoob  MY,  Micha  R,  Khatibzadeh  S,  Singh  GM,  Shi  P, et al. (2016) Impact of dietary and metabolic risk factors on cardiovascular and diabetes mortality in South Asia: analysis from the 2010 global burden of disease study. Am J Public Health 106: 2113–2125.
• 10  Setchell KD (1998) Phytoestrogens: the biochemistry, physiology, and implications for human health of soy isoflavones. Am J Clin Nutr 68: 1333s–1346s.
• 11   Bhathena  SJ,  Velasquez  MT (2002) Beneficial role of dietary phytoestrogens in obesity and diabetes. Am J Clin Nutr 76: 1191–1201.
• 12   Nagata  C,  Inaba  S,  Kawakami  N,  Kakizoe  T,  Shimizu  H (2000) Inverse association of soy product intake with serum androgen and estrogen concentrations in Japanese men. Nutr Cancer 36: 14–18.
• 13   Tsuji  M,  Tamai  Y,  Wada  K,  Nakamura  K,  Hayashi  M, et al. (2012) Associations of intakes of fat, dietary fiber, soy isoflavones, and alcohol with levels of sex hormones and prolactin in premenopausal Japanese women. Cancer Causes Control 23: 683–689.
• 14   Xu  X,  Harris  KS,  Wang  HJ,  Murphy  PA,  Hendrich  S (1995) Bioavailability of soybean isoflavones depends upon gut microflora in women. J Nutr 125: 2307–2315.
• 15   Hwang  CS,  Kwak  HS,  Lim  HJ,  Lee  SH,  Kang  YS, et al. (2006) Isoflavone metabolites and their in vitro dual functions: they can act as an estrogenic agonist or antagonist depending on the estrogen concentration. J Steroid Biochem Mol Biol 101: 246–253.
• 16   Mahalingam  S,  Gao  L,  Gonnering  M,  Helferich  W,  Flaws  JA (2016) Equol inhibits growth, induces atresia, and inhibits steroidogenesis of mouse antral follicles in vitro. Toxicol Appl Pharmacol 295: 47–55.
• 17   Bae  M,  Woo  M,  Kusuma  IW,  Arung  ET,  Yang  CH, et al. (2012) Inhibitory effects of isoflavonoids on rat prostate testosterone 5alpha-reductase. J Acupunct Meridian Stud 5: 319–322.
• 18   Hu  GX,  Zhao  BH,  Chu  YH,  Zhou  HY,  Akingbemi  BT, et al. (2010) Effects of genistein and equol on human and rat testicular 3 beta-hydroxysteroid dehydrogenase and 17 beta-hydroxysteroid dehydrogenase 3 activities. Asian J Androl 12: 519–526.
• 19   Opalka  DM,  Kaminska  B,  Piskula  MK,  Puchajda-Skowronska  H,  Dusza  L (2006) Effects of phytoestrogens on testosterone secretion by Leydig cells from Bilgoraj ganders (Anser anser). Br Poult Sci 47: 237–245.
• 20   Lund  TD,  Munson  DJ,  Haldy  ME,  Setchell  KD,  Lephart  ED, et al. (2004) Equol is a novel anti-androgen that inhibits prostate growth and hormone feedback. Biol Reprod 70: 1188–1195.
• 21   Loukovaara  M,  Carson  M,  Palotie  A,  Adlercreutz  H (1995) Regulation of sex hormone-binding globulin production by isoflavonoids and patterns of isoflavonoid conjugation in HepG2 cell cultures. Steroids 60: 656–661.
• 22   Martin  ME,  Haourigui  M,  Pelissero  C,  Benassayag  C,  Nunez  EA (1996) Interactions between phytoestrogens and human sex steroid binding protein. Life Sci 58: 429–436.
• 23   Loutchanwoot  P,  Srivilai  P,  Jarry  H (2014) Lack of anti-androgenic effects of equol on reproductive neuroendocrine function in the adult male rat. Horm Behav 65: 22–31.
• 24   Tousen  Y,  Ezaki  J,  Fujii  Y,  Ueno  T,  Nishimuta  M, et al. (2011) Natural S-equol decreases bone resorption in postmenopausal, non-equol-producing Japanese women: a pilot randomized, placebo-controlled trial. Menopause 18: 563–574.
• 25  Dong HL, Tang XY, Deng YY, Zhong QW, Wang C, et al. (2019) Urinary equol, but not daidzein and genistein, was inversely associated with the risk of type 2 diabetes in Chinese adults. Eur J Nutr Apr 5. [Epub ahead of print]
• 26   Shi  L,  Ryan  HH,  Jones  E,  Simas  TA,  Lichtenstein  AH, et al. (2014) Urinary isoflavone concentrations are inversely associated with cardiometabolic risk markers in pregnant U.S. women. J Nutr 144: 344–351.
• 27   Zhang  X,  Gao  YT,  Yang  G,  Li  H,  Cai  Q, et al. (2012) Urinary isoflavonoids and risk of coronary heart disease. Int J Epidemiol 41: 1367–1375.
• 28   Minatoya  M,  Kutomi  G,  Asakura  S,  Otokozawa  S,  Sugiyama  Y, et al. (2013) Equol, adiponectin, insulin levels and risk of breast cancer. Asian Pac J Cancer Prev 14: 2191–2199.
• 29   Nagata  C,  Iwasa  S,  Shiraki  M,  Ueno  T,  Uchiyama  S, et al. (2006) Associations among maternal soy intake, isoflavone levels in urine and blood samples, and maternal and umbilical hormone concentrations (Japan). Cancer Causes Control 17: 1107–1113.
• 30   Duncan  AM,  Merz-Demlow  BE,  Xu  X,  Phipps  WR,  Kurzer  MS (2000) Premenopausal equol excretors show plasma hormone profiles associated with lowered risk of breast cancer. Cancer Epidemiol Biomarkers Prev 9: 581–586.
• 31   Atkinson  C,  Newton  KM,  Stanczyk  FZ,  Westerlind  KC,  Li  L, et al. (2008) Daidzein-metabolizing phenotypes in relation to serum hormones and sex hormone binding globulin, and urinary estrogen metabolites in premenopausal women in the United States. Cancer Causes Control 19: 1085–1093.
• 32   Frankenfeld  CL,  McTiernan  A,  Tworoger  SS,  Atkinson  C,  Thomas  WK, et al. (2004) Serum steroid hormones, sex hormone-binding globulin concentrations, and urinary hydroxylated estrogen metabolites in post-menopausal women in relation to daidzein-metabolizing phenotypes. J Steroid Biochem Mol Biol 88: 399–408.
• 33   Setchell  KD,  Cole  SJ (2006) Method of defining equol-producer status and its frequency among vegetarians. J Nutr 136: 2188–2193.
• 34   Liu  J,  Shi  WQ,  Cao  Y,  He  LP,  Guan  K, et al. (2014) Higher serum carotenoid concentrations associated with a lower prevalence of the metabolic syndrome in middle-aged and elderly Chinese adults. Br J Nutr 112: 2041–2048.
• 35   Ainsworth  BE,  Haskell  WL,  Herrmann  SD,  Meckes  N,  Bassett Jr  DR, et al. (2011) 2011 Compendium of physical activities: a second update of codes and MET values. Med Sci Sports Exerc 43: 1575–1581.
• 36   Vermeulen  A,  Verdonck  L,  Kaufman  JM (1999) A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 84: 3666–3672.
• 37   Wang  L,  Lu  W,  Chang  C,  Xie  F,  Wei  H, et al. (2009) The simultaneous analysis of six isoflavone metabolites in human urine by HPLC. Acta Nutrimenta Sinica 31: 181–184 (In Chinese).
• 38   He  J,  Feng  Y,  Ouyang  HZ,  Yu  B,  Chang  YX, et al. (2013) A sensitive LC–MS/MS method for simultaneous determination of six flavonoids in rat plasma: application to a pharmacokinetic study of total flavonoids from mulberry leaves. J Pharm Biomed Anal 84: 189–195.
• 39  Hayes AF (2013) Introduction to mediation, moderation, and conditional process analysis: a regression-based approach (1^st). New York: 85–122.
• 40   Hamilton-Reeves  JM,  Vazquez  G,  Duval  SJ,  Phipps  WR,  Kurzer  MS, et al. (2010) Clinical studies show no effects of soy protein or isoflavones on reproductive hormones in men: results of a meta-analysis. Fertil Steril 94: 997–1007.
• 41   Huber  J,  Imhof  M,  Schmidt  M (2010) Effects of soy protein and isoflavones on circulating hormone concentrations in pre- and post-menopausal women: a systematic review and meta-analysis. Hum Reprod Update 16: 110–111.
• 42   Setchell  KD,  Clerici  C (2010) Equol: history, chemistry, and formation. J Nutr 140: 1355S–1362S.
• 43   Setchell  KD,  Clerici  C (2010) Equol: pharmacokinetics and biological actions. J Nutr 140: 1363S–1368S.
• 44   Kurzer  MS,  Xu  X (1997) Dietary phytoestrogens. Annu Rev Nutr 17: 353–381.
• 45   Gebhart  VM,  Jirikowski  GF (2014) Estrogen dependent expression of sex hormone binding globulin in PC 12 cells. Steroids 81: 26–30.
• 46   Tang  X,  Zhu  X,  Liu  S,  Wang  S,  Ni  X (2011) Isoflavones suppress cyclic adenosine 3',5'-monophosphate regulatory element-mediated transcription in osteoblastic cell line. J Nutr Biochem 22: 865–873.
• 47   Nakamura  I,  Kusakabe  M,  Swanson  P,  Young  G (2016) Regulation of sex steroid production and mRNAs encoding gonadotropin receptors and steroidogenic proteins by gonadotropins, cyclic AMP and insulin-like growth factor-I in ovarian follicles of rainbow trout (Oncorhynchus mykiss) at two stages of vitellogenesis. Comp Biochem Physiol A Mol Integr Physiol 201: 132–140.
• 48   Mercier-Bodard  C,  Baville  F,  Bideux  G,  Binart  N,  Chambraud  B, et al. (1989) Regulation of SBP synthesis in human cancer cell lines by steroid and thyroid hormones. J Steroid Biochem 34: 199–204.
• 49   Franke  AA,  Hebshi  SM,  Pagano  I,  Kono  N,  Mack  WJ, et al. (2010) Urine accurately reflects circulating isoflavonoids and ascertains compliance during soy intervention. Cancer Epidemiol Biomarkers Prev 19: 1775–1783.
• 50   Fabbri  E,  An  Y,  Gonzalez-Freire  M,  Zoli  M,  Maggio  M, et al. (2016) Bioavailable testosterone linearly declines over a wide age spectrum in men and women from the baltimore longitudinal study of aging. J Gerontol A Biol Sci Med Sci 71: 1202–1209.