Soy isoflavones are phytoestrogens found in soybeans. Phytoestrogens are plant-derived nonsteroidal compounds that possess estrogen-like biologic activities. These compounds reportedly display weak estrogenic and antiestrogenic properties.1–3 The 2 primary isoflavones found in soy are genistein and daidzein. Structural similarities allow isoflavones to bind to estrogen receptors.4
It has been hypothesized that soy isoflavones may play a role in the etiology of estrogen-related diseases and several epidemiologic studies have been conducted; however, findings have been complicated and inconsistent.5–7 A prospective study in Japan, where isoflavone intake is known to be relatively high, showed a protective effect on postmenopausal breast cancer.5 On the other hand, a nested case–control study in the United Kingdom, where intake is relatively low, showed that serum and urinary isoflavone levels were associated with increased breast cancer risk.6 A recent meta-analysis found a small reduction in breast cancer risk associated with soy intake.7 However, the authors suggested that the results should be interpreted cautiously due to potential exposure misclassification, confounding, lack of a dose-response pattern and the possibility of adverse effects of soy constituents.
Endometriosis is a benign, proliferative disease in which tissue similar to endometrial tissue is found outside the uterus—usually in the pelvic cavity, but sometimes in distant organs. Endometriosis is commonly accompanied by pelvic pain and infertility. Both genetic and environmental factors may contribute.8 The reported prevalence of largely asymptomatic endometriosis found in women undergoing tubal ligation is about 4%, ranging from 1% to 7%.9 Progression of endometriosis is considered estrogen-dependent.10 Soy isoflavones might thus be expected to affect the risk and severity of endometriosis. However, few studies have investigated the effects of soy isoflavones on endometriosis.
Several studies have recently described associations between estrogen receptor (ESR) gene polymorphisms and endometriosis.11–13 Genistein and daidzein reportedly display much greater affinity for ESR2 than for ESR1,14 suggesting that the estrogenic or antiestrogenic properties of soy isoflavones may occur preferentially through ESR2. Although functional variability of ESR2 gene polymorphisms could feasibly be associated with response to soy isoflavones, whether ESR2 gene polymorphisms exert altered phenotypic effects on endometriosis through interactions with soy isoflavones is not known.
The present study investigated whether urinary genistein and daidzein are associated with risk and severity of endometriosis, and whether polymorphisms in the ESR2 gene are associated with response to soy isoflavones.
Study Protocol and Ethics
This study was part of a case–control study conducted on a Japanese population to investigate associations between genetic and environmental factors in endometriosis.15 We recruited consecutive female patients age 20 to 45-year-old who attended the Department of Obstetrics and Gynecology at Jikei University School of Medicine Hospital for infertility in 1999 or 2000. Since pregnancy commonly results in complete resolution of minimal or mild endometriosis, women who had given birth or lactated were ineligible, leaving a total of 159 women who met the criteria. After excluding 15 women who did not give consent, 5 who did not undergo blood screening or laparoscopic examination, and 1 whose DNA sample was not available, a total of 138 women were available for the study (participation rate = 87%). No participants had undergone therapy before laparoscopic examination.
All study protocols were approved by the Institutional Review Boards of Jikei University, National Cancer Center and National Institute for Environmental Studies. All participants provided written informed consent before laparoscopic examination.
Before the laparoscopic examination, participants were interviewed by a single trained interviewer using a structured questionnaire to collect information on demographic factors, age, height, weight, medical history for themselves and their families, reproductive and menstrual history, oral contraceptive use, food- and alcohol-consumption frequency, and smoking history.
Participants collected first morning urine sample using a paper cup and plastic tube, and gave a fasting blood sample before the laparoscopic examination. Blood samples were divided into plasma and buffy layers. All biologic samples were stored at −80°C until analysis.
Diagnosis of Endometriosis
Laparoscopy is necessary for definitive diagnosis of endometriosis. In the present study, all participants underwent diagnostic laparoscopy, and stage of endometriosis was determined by trained gynecologists in accordance with the revised classifications of the American Fertility Society.16 Endometriosis was absent in 59 women (43%), Stage I in 21 women (15%), Stage II in 10 women (7%), Stage III in 23 women (17%) and Stage IV in 25 women (18%). Current theories of endometriosis suggest that what is defined as minimal/mild endometriosis may actually represent a normal physiologic process. Furthermore, a lack of consistency between laparoscopic and histologic diagnosis has been reported, particularly for minimal/mild endometriosis.17 Considering the more severe stages as a separate category thus appears logical.18 Based on surgically or pathologically confirmed disease status, we classified cases into 2 subgroups: early (Stage I–II) or advanced endometriosis (stage III–IV). Women without endometriosis were defined as controls.
Determination of Soy Isoflavone Levels
Urinary levels of soy isoflavones offer a useful biomarker for dietary intake and plasma concentration of isoflavones.19–21 The present study measured urinary levels of genistein and daidzein as markers for dietary intake of soy isoflavones. A total of 30 mL of first-morning urine was collected before laparoscopic examination. Genistein and daidzein levels were analyzed using high-performance liquid chromatography with a coulometric array detector in accordance with the modified methods of Gamache and Acworth.22
Concentrations of genistein and daidzein were determined by linear regression of peak height for each standard, and were adjusted according to recovery rate of the internal standard. The regression coefficient of peak height and concentration calculated for soy isoflavones revealed a linearity range of 0–8.0 μg/mL, with correlation coefficient values >0.995. Voltametric response for the standard solution displayed coefficients of variation of 2.7%–8.4% for intraday variation and 11.1%–12.2% for interday variation. Recovery rates of soy isoflavones in urine samples ranged between approximately 85% and 100%. Detection limits were 3.22 ng/mL for genistein and 4.14 ng/mL for daidzein.
Concentrations of urinary genistein and daidzein were adjusted by urinary creatinine concentration to correct for variability in urine dilution (μmol/g Cre). All measurements were performed by investigators blinded to case–control status.
Genotyping of ESR2 Gene Polymorphism
The ESR2RsaI polymorphism, comprising a G-to-A change at nucleotide 1082 in exon 5, was genotyped using polymerase chain reaction (PCR) restriction fragment length polymorphism methods.23 Blood samples were obtained before laparoscopic examination. Genomic DNA samples were extracted from peripheral white blood cells using a standard protease K method. PCR products were digested using 5 U of RsaI restriction enzyme at 37°C for 8 hours, then electrophoresed on a 3% agarose gel containing ethidium bromide.
In this study, ESR2RsaI polymorphism is represented by the r and R alleles, with R indicating the presence of corresponding restriction sites, and r indicating the absence of restriction sites. For quality control, blinded control samples were inserted to validate genotyping identification procedures. Concordance for blinded samples was 100%. Genotyping was conducted by investigators blinded to case–control status.
To assess differences between cases and controls, basic characteristics and possible risk factors for endometriosis were compared using Student t test and the χ2 test. Spearman correlation coefficients between urinary level of genistein and daidzein were calculated. We calculated odds ratios (ORs) and 95% confidence intervals (CIs) for urinary levels by an unconditional logistic regression model following classification into medians or quartiles based on control distribution. Linear trends for ORs were tested in the unconditional logistic regression model by treating the categories as ordinal variables. We evaluated trends for median values according to disease stage to assess associations between urinary levels of genistein and daidzein and disease stage.
To compare observed and expected genotype frequencies, we tested for Hardy-Weinberg equilibrium by using an exact test. ESR2RsaI polymorphism was classified into 2 subgroups according to the presence of corresponding restriction sites: r/r genotype; and R/r + R/R genotype. ORs and 95% CIs were calculated for associations between ESR2RsaI polymorphism and endometriosis using the unconditional logistic regression model.
To investigate whether the ESR2RsaI genotype modified the effect of urinary levels of genistein or daidzein, we calculated ORs and 95% CIs of endometriosis according to a combination of subgroups for the ESR2RsaI genotype and urinary isoflavones, using the unconditional logistic regression model. A low level of urinary isoflavones in combination with R/r + R/R genotype was considered as the reference group. Interactions between ESR2RsaI polymorphism and urinary isoflavones in the risk of endometriosis were tested with the Wald test using product terms between urinary genistein or daidzein and genotypes.
The present study was designed to have 80% power to detect a decrease in risk of two-thirds at the 5% level of significance. All statistical tests were based on 2-tailed probabilities. We adjusted ORs and 95% CIs for possible confounding factors of endometriosis, namely age (continuous), menstrual cycle (continuous), and duration of menstrual bleeding (less than 7 days or 7 days or more).9,10 We used SPSS for Windows software version 11.0 (SPSS JAPAN, Tokyo, Japan) for statistical analyses.
Baseline Characteristics and Possible Risk Factors for Endometriosis
Table 1 shows baseline characteristics and possible risk factors for endometriosis in controls and cases. No important differences in mean age or body mass index were identified between groups. Distribution of menstrual bleeding, hypermenorrhea, and smoking also did not differ substantially. The advanced endometriosis group had a shorter mean menstrual cycle length than controls (controls, 30.7 ± 6.1 days; advanced endometriosis, 28.3 ± 3.0 days) and was more likely to have menstrual cramps and dyspareunia.
Effect of Urinary Isoflavones on Endometriosis
Table 2 shows risk of endometriosis according to median or quartile levels of urinary isoflavones. In controls, median isoflavone level was 3.24 μmol/g Cre for genistein and 4.01 μmol/g Cre for daidzein. The Spearman correlation coefficient between genistein and daidzein was 0.84. Urinary genistein and daidzein levels were inversely associated with advanced endometriosis (P for trend = 0.01 and 0.06, respectively) but not with early endometriosis. For advanced endometriosis, the adjusted odds ratio for the highest quartile group was 0.21 (95% CI = 0.06–0.76) for genistein and 0.29 (0.08–1.03) for daidzein when compared with the lowest group.
Table 3 shows the trends of median values for urinary isoflavones according to disease stage. An inverse relationship with stage of endometriosis was observed for both genistein levels (P for trend = 0.01) and daidzein levels (P for trend = 0.07).
Associations Between ESR2RsaI Polymorphism and Endometriosis
Table 4 shows the genotypic distribution of ESR2RsaI polymorphism and associations with risk of endometriosis. The ESR2RsaI r/r genotype was predominant. Allele frequencies of ESR2RsaI polymorphism were 0.77 for the r allele and 0.23 for the R allele. In addition, the distribution of ESR2RsaI polymorphism was in Hardy–Weinberg equilibrium (P = 0.26). The ESR2RsaI r/r genotype was associated with reduced risk of early endometriosis compared with the R/r + R/R genotype (OR = 0.30; CI = 0.11–0.85). The association was weaker for advanced endometriosis (0.67; 0.29–1.55).
Interactions Between ESR2RsaI Polymorphism and Urinary Isoflavones in the Risk of Endometriosis
Table 5 shows ORs and 95% CIs of endometriosis for combinations of ESR2RsaI genotype and urinary isoflavone levels. Compared with subjects with the ESR2RsaI R/r + R/R genotype and a low genistein level, ORs of advanced endometriosis were lower among the 3 other groups. The adjusted OR was 0.10 (95% CI = 0.02–0.48) for subjects with ESR2RsaI R/r + R/R genotype with high genistein level; 0.32 (0.10–1.04) for subjects with ESR2RsaI r/r genotype with low genistein level; 0.27 (0.08–0.92) for subjects with ESR2RsaI r/r genotype with high genistein level. A significant interaction was noted between ESR2RsaI polymorphism and genistein levels in risk of advanced endometriosis (P for interaction = 0.03). Interactions between ESR2RsaI polymorphism and genistein level were not observed in early endometriosis. Although a similar pattern was observed for ORs of both early and advanced endometriosis for the combinations of ESR2RsaI genotype and urinary daidzein level, these may have been due to chance.
The present study showed an inverse association between urinary isoflavones and the risk of advanced endometriosis. This association was stronger for genistein than daidzein. In addition, there was statistical evidence for interaction between urinary genistein and ESR2 gene polymorphisms.
The reduced risk of endometriosis following ingestion of soy isoflavones may be attributable to antiestrogenic properties of these compounds. A previous study showed that prolonged exposure to genistein results in decreased levels of estrogen receptor mRNA in addition to decreased response to estradiol stimulation.24 Plasma levels of isoflavones can be 10,000- to 100,000-fold higher than those of estradiol.25 When the relative binding affinity of 17β-estradiol was set at 100 in solid-phase competition experiments, relative binding affinity for ESR2 was 87 for genistein and 0.5 for daidzein.14 Although the elimination half-life from blood and urine is reportedly 7–8 hours for both genistein and daidzein,26 long-term soy diets may modify the physiologic effects of estrogens. Given these facts, a lower prevalence of endometriosis might be expected in Japanese populations compared with Western countries, as with breast cancer. Nevertheless, the prevalence of endometriosis in the Japanese general population remains unclear due to the need for surgical diagnosis.
Our finding showed that the strength of association was stronger for genistein than for daidzein. One possible explanation is the difference in their binding affinities to ESR2. A second possibility is based on the difference in metabolism between genistein and daidzein. Daidzein can be metabolized to equol and O-desmethylangolites by intestinal bacteria, and these metabolites are absorbed, enter the circulation, and are excreted in urine. Although equol has been suggested to possess stronger estrogenic properties than genistein, some individuals are capable of equol production whereas others are not, probably because of differences in gut microflora. This difference might play a role in the weaker associations for daidzein than genistein.27
ESR2 plays important roles in endometrial function, in addition to the well-known role of ESR1 in endometrial proliferation and differentiation.28 The ESR2RsaI polymorphism does not cause amino acid changes, but may well be associated with altered ligand-binding affinity or transcriptional activity. Genes containing single nucleotide polymorphisms (SNPs) can cause different structural folds in mRNA,29 and these mRNA variants may possess different biologic functions during interactions with other cellular components. Altered estrogen or soy isoflavone signal transduction thanks to ESR2 gene polymorphisms may be directly responsible for interindividual susceptibility to and severity of endometriosis.
The present study found evidence of an interaction between urinary genistein and ESR2 gene polymorphisms. Isoflavones may play a more effective role among the ESR2RsaI R/r + R/R genotype than the r/r genotype, although the latter itself is likely to be protective for endometriosis. This result should be interpreted cautiously, however, because of the relatively small number of subjects—a major limitation of this study. When the number of subjects studied is not large and the expected difference is small, actual differences are quite likely to pass undetected. Inconsistent results between early and advanced endometriosis might be attributable to the lack of sufficient numbers and possible misclassification in the early endometriosis group. Alternatively, the observed interactions may have occurred merely by chance.
A second issue is our definition of cases and controls. In accordance with the revised classifications of the American Fertility Society, we defined women without endometriosis as controls and women with early (Stage I–II) and advanced endometriosis (Stage III–IV) as cases,16 although there is no clear criterion for dichotomizing cases. The present study did not show a persuasive inverse association between urinary isoflavones and the risk of early endometriosis, although a strong protective effect was found for advanced endometriosis. Further analysis, however, did show an inverse association between urinary isoflavones and the severity of endometriosis. This finding may be reasonable given that endometriosis occurs in a continuum of severity.
A third issue is measurement of urinary levels of isoflavones. The present study measured urinary excretion of genistein and daidzein as markers of soy isoflavone consumption. Urinary excretion of soy isoflavones is reportedly related to annual dietary intake of soy isoflavones.19 Since we collected spot urine samples, intraindividual variation in urinary isoflavones cannot be ignored. Such misclassification, however, is probably nondifferential and would lead to a null result.
Participants in the present study were infertile. They might therefore have changed their diet due to their symptoms or in attempt to become pregnant. If a change in diet was more likely among patients with advanced endometriosis than the controls, our findings might have been the result of the change in diet. In addition, given reports that factors associated with endometriosis differ between parous women (who experienced neither primary nor secondary infertility) and nulliparous infertile women,30,31 the influence of urinary isoflavone levels on endometriosis risk between the 2 groups may have differed. Therefore, our present findings may be limited to infertile women.
In conclusion, in a case–control study in infertile Japanese women, we found that higher urinary level of isoflavones was associated with a reduced risk of advanced endometriosis. Although the interaction between urinary genistein and ESR2 gene polymorphisms supported the mechanism for a role of isoflavones in the etiology of endometriosis, further studies with a large number of subjects are needed to confirm these findings.
The authors are grateful for the collaboration of Amanda Sue Niskar (Israel Center for Disease Control, Gertner Institute) in designing the study protocol. In addition, the authors wish to thank Hiroaki Itoh (Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan) for his helpful comments.
1. Tham DM, Gardner CD, Haskell WL. Clinical review 97: potential health benefits of dietary phytoestrogens: a review of the clinical, epidemiological, and mechanistic evidence. J Clin Endocrinol Metab
2. Wagner JD, Anthony MS, Cline JM. Soy phytoestrogens: research on benefits and risks. Clin Obstet Gynecol
3. Gruber CJ, Tschugguel W, Schneeberger C, et al. Production and actions of estrogens. N Engl J Med
4. Limer JL, Speirs V. Phyto-oestrogens and breast cancer chemoprevention. Breast Cancer Res
5. Yamamoto S, Sobue T, Kobayashi M, et al. Soy, isoflavones, and breast cancer risk in Japan. J Natl Cancer Inst
6. Grace PB, Taylor JI, Low YL, et al. Phytoestrogen concentrations in serum and spot urine as biomarkers for dietary phytoestrogen intake and their relation to breast cancer risk in European prospective investigation of cancer and nutrition-norfolk. Cancer Epidemiol Biomarkers Prev
7. Trock BJ, Hilakivi-Clarke L, Clarke R. Meta-analysis of soy intake and breast cancer risk. J Natl Cancer Inst
8. Kennedy S. The genetics of endometriosis. J Reprod Med
9. Cramer DW, Missmer SA. The epidemiology of endometriosis. Ann NY Acad Sci
10. Vigano P, Parazzini F, Somigliana E, et al. Endometriosis: epidemiology and aetiological factors. Best Pract Res Clin Obstet Gynaecol
11. Georgiou I, Syrrou M, Bouba I, et al. Association of estrogen receptor gene polymorphisms with endometriosis. Fertil Steril
12. Kitawaki J, Obayashi H, Ishihara H, et al. Oestrogen receptor-alpha gene polymorphism is associated with endometriosis, adenomyosis and leiomyomata. Hum Reprod
13. Wang Z, Yoshida S, Negoro K, et al. Polymorphisms in the estrogen receptor beta gene but not estrogen receptor alpha gene affect the risk of developing endometriosis in a Japanese population. Fertil Steril
14. Kuiper GG, Lemmen JG, Carlsson B, et al. Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology
15. Tsukino H, Hanaoka T, Sasaki H, et al. Associations between serum levels of selected organochlorine compounds and endometriosis in infertile Japanese women. Environ Res
16. American Fertility Society. Revised American Fertility Society classification of endometriosis. Fertil Steril
17. Marchino GL, Gennarelli G, Enria R, et al. Diagnosis of pelvic endometriosis with use of macroscopic versus histologic findings. Fertil Steril
18. Zondervan KT, Cardon LR, Kennedy SH. What makes a good case-control study? Design issues for complex traits such as endometriosis. Hum Reprod
19. Maskarinec G, Singh S, Meng L, et al. Dietary soy intake and urinary isoflavone excretion among women from a multiethnic population. Cancer Epidemiol Biomarkers Prev
20. Arai Y, Uehara M, Sato Y, et al. Comparison of isoflavones among dietary intake, plasma concentration and urinary excretion for accurate estimation of phytoestrogen intake. J Epidemiol
21. Yamamoto S, Sobue T, Sasaki S, 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. J Nutr
22. Gamache PH, Acworth IN. Analysis of phytoestrogens and polyphenols in plasma, tissue, and urine using HPLC with coulometric array detection. Proc Soc Exp Biol Med
23. Sundarrajan C, Liao WX, Roy AC, et al. Association between estrogen receptor-beta gene polymorphisms and ovulatory dysfunctions in patients with menstrual disorders. J Clin Endocrinol Metab
24. Wang TT, Sathyamoorthy N, Phang JM. Molecular effects of genistein on estrogen receptor mediated pathways. Carcinogenesis
25. Manach C, Williamson G, Morand C, et al. Bioavailability and bioefficacy of polyphenols in humans. I. Review of 97 bioavailability studies. Am J Clin Nutr
. 2005;81(1 Suppl):230S–242S.
26. King RA, Bursill DB. Plasma and urinary kinetics of the isoflavones daidzein and genistein after a single soy meal in humans. Am J Clin Nutr
27. Atkinson C, Frankenfeld CL, Lampe JW. Gut bacterial metabolism of the soy isoflavone daidzein: exploring the relevance to human health. Exp Biol Med (Maywood)
28. Lecce G, Meduri G, Ancelin M, et al. Presence of estrogen receptor beta in the human endometrium through the cycle: expression in glandular, stromal, and vascular cells. J Clin Endocrinol Metab
29. Shen LX, Basilion JP, Stanton VP Jr. Single-nucleotide polymorphisms can cause different structural folds of mRNA. Proc Natl Acad Sci USA
30. Missmer SA, Hankinson SE, Spiegelman D, et al. Reproductive history and endometriosis among premenopausal women. Obstet Gynecol
31. Missmer SA, Hankinson SE, Spiegelman D, et al. Incidence of laparoscopically confirmed endometriosis by demographic, anthropometric, and lifestyle factors. Am J Epidemiol