Estrogen exposures play a critical role in the development of endometrial cancer. Genetic variation in the estrogen metabolism UGT1A1 gene may modify the effect of estrogenic exposures on endometrial cancer risk. We tested this hypothesis in a population-based case-control study of 1,047 endometrial cancer cases and 1,035 controls who completed an in-person interview and were genotyped for the UGT1A1 polymorphisms rs2070959 (A/G), rs887829 (G/A), and rs8175347 (6/7 TA repeats). Estrogen exposure-related factors evaluated include menstrual characteristics, oral contraceptive use, body mass index, waist-hip ratio, and soy food intake. Conditional logistic regression was used to calculate odds ratios and 95% confidence intervals. The homozygote variant genotype (G/G) of the rs2070959 polymorphism was significantly associated with a reduced risk of endometrial cancer (odds ratio, 0.5; 95% confidence interval, 0.3-0.8). No significant associations between endometrial cancer risk and genotype were seen for the rs887829 and rs8175347 polymorphisms. Analysis of the joint effects of genotype and markers of estrogen exposure found the lowest risk of endometrial cancer among those with the homozygous variant genotype of the rs2070959 polymorphism and who were postmenopausal, had low body mass index, and had low soy food intake, although a test for multiplicative interaction was not significant. Taken together, these data suggest that the G/G genotype (rs2070959) in the UGT1A1 gene may decrease the risk of endometrial cancer and that this effect is most evident among women with low levels of endogenous estrogen exposure or with low soy food intake. (Cancer Epidemiol Biomarkers Prev 2008;17(3):563–70)

Estrogen exposures play a crucial role in endometrial carcinogenesis as women with higher levels of estrogen are at increased risk for development of the disease (1-7). Specifically, endogenous estrogens, such as unopposed estrogen therapy, as well as markers of endogenous estrogen exposure, such as nulliparity, early menarche, late menopause, and obesity, have all been shown to increase endometrial cancer risk (1-9). Contrarily, phytoestrogens (estrogens derived from plant foods) have been shown to be inversely related to risk presumably through their ability to competitively bind estrogen receptors, to stimulate production of sex hormone-binding globulin, and to inhibit certain estrogen synthesis enzymes (10-13).

Estradiol and estrone, the primary estrogens in women, are converted into a myriad of catechol estrogens and eliminated via various processes, including hydroxylation, oxidation, detoxification, sulfination, and glucuronidation (14). Catechol estrogens, depending on the compound, can have either tumorigenic or antitumorigenic activity. For example, 4-hydroxyestradiol possesses potent hormonal activity via its activation of the estrogen receptor (15). In CD-1 mice, 4-hydroxyestradiol has also been shown to more strongly induce endometrial adenocarcinoma compared with 2-hydroxyestradiol (16). Conversely, 2-methyoxyestradiol has antitumorigenic activity as it is an inhibitor of tumor cell proliferation and angiogenesis (17, 18).

Given the evidence for the role of estrogen metabolites in carcinogenesis, research has focused on the various proteins and pathways that modulate estrogen metabolism and elimination, including catechol-O-methyltransferase, cytochrome P450 enzymes, glutathione S-transferases, and N-acetyltransferases (19-25). Other important but perhaps less well-studied members of the estrogen metabolism pathway in terms of genetic variation and endometrial cancer risk are UDP-glucuronosyltransferases. UDP-glucuronosyltransferases are a family of membrane-bound enzymes involved in the inactivation and elimination of catechol estrogens through glucuronidation. One family member, UGT1A1, encodes a UDP-glucuronosyltransferase, an enzyme of the glucuronidation pathway that transforms small lipophilic molecules, such as steroids, bilirubin, hormones, and drugs, into water-soluble secreted molecules. UGT1A1 has been identified as a major contributor of the glucuronidation activity associated with estrogens and catechol estrogens (Fig. 1). Thus, functional genetic variants in the UGT1A1 gene may be associated with altered endometrial cancer risk (19, 21, 22, 26-28). Several studies have looked at polymorphisms in the UGT1A1 gene in relation to breast cancer risk, and a promoter variant has been linked with increased risk of breast cancer (29-32). However, to date, only two studies have evaluated the role of polymorphisms in the UGT1A1 gene and endometrial cancer risk with inconsistent results (33, 34). We tested the hypothesis that variation in the UGT1A1 gene may be associated with risk of endometrial cancer in a large population-based case-control study, the Shanghai Endometrial Cancer Study (SECS), and evaluated the effect of the gene variants in concert with various estrogenic exposures.

Figure 1.

Role of UGT1A1 in the estrone and estradiol metabolism pathway.

Figure 1.

Role of UGT1A1 in the estrone and estradiol metabolism pathway.

Close modal

Study Population

The SECS includes a total of 1,204 population-based incident endometrial cancer cases and 1,212 community controls recruited from 1997 to 2003 and has been described previously (35). Institutional review board approval for this study was garnered from the relevant review boards in both China and the United States, and informed consent was obtained from each participant before inclusion into the study. Of the 1,458 eligible endometrial cancer cases and 1,629 controls, 1,204 (82.6%) cases and 1,212 (74.4%) controls completed in-person interviews. Among the 254 nonparticipant cases, 135 (9.3%) refused to participate, 66 (4.5%) died before study enrollment, 39 (2.7%) could not be located, and the remaining 14 (1.0%) could not be interviewed due to health or communication issues. Among nonparticipant eligible controls, 340 (20.9%) declined to be interviewed, 61 (3.7%) did not live in Shanghai during the study period, and 16 (1.1%) did not participate in the study due to health or communication problems. Interviews were conducted by retired health professionals using a structured questionnaire, which collected detailed information on demographic factors, menstrual and reproductive history, hormone use, prior disease history, family history of cancer, and lifestyle factors, including alcohol and tobacco use and physical activity history. Additionally, information regarding the participants' usual dietary habits over the preceding 5 years was collected during the interview using a validated food frequency questionnaire. The food frequency questionnaire listed 71 food items and covered more than 85% of the foods commonly consumed in Shanghai (36).

Current weight and height of each participant were measured, and a blood sample or buccal cell sample was taken during the in-person interview. Among the study participants, 860 cases and 861 controls donated a blood samples, whereas 285 cases and 285 controls provided a buccal cell sample (obtained through either a mouthwash or buccal swab method). There was a very low DNA yield from the 93 cases and 88 controls who provided a buccal swab, and their samples were not included in the genotyping. As such, DNA samples for 87.3% (n = 1,052) of cases and 87.3% (n = 1,058) were included in the current study.

Single Nucleotide Polymorphism Selection and Laboratory Methods

Polymorphisms evaluated in this study were selected in 2005 based on their known or potential functionality. The rs8175347 polymorphism represents TA repeats in the promoter. Six repeats (also known by the designation *1) have been shown to be associated with high transcriptional activity, whereas seven repeats (*28) have been associated with lower activity. This polymorphism was also selected given its previous associations with altered cancer risk in the literature (29, 30, 37). The rs2070959 polymorphism was selected as the variant allele and can lead to a threonine-to-alanine amino acid change dependent on splicing. The rs887829 was selected due to its location in the promoter. Genomic DNA was extracted either from buffy coat fractions or from buccal cells using a QIAamp DNA mini kit (Qiagen) per the manufacturer's protocol. Allelic discrimination of the rs2070959 and rs887829 polymorphisms in the UGT1A1 gene was determined using TaqMan genotyping assays on the ABI PRISM 7900 Sequence Detections Systems (Applied Biosystems). The assay ID for the rs887829 polymorphism was C_2669357_10. For the rs2070959 polymorphism, primers were CCTACCATCTGTGTACCTCTTCAG (forward) and TGTAGCACCTGGGAATGTAGGA (reverse) and the probes were VIC-CTGGAGCATACATTC and FAM-TGGAGCATGCATTC. The final volume for each reaction was 5 μL consisting of 2.5 μL TaqMan Universal PCR Master Mix (Applied Biosystems), 0.25 μL primers/probes mix, and 5.0 ng genomic DNA. The PCR cycling profile consisted of an initial denaturation step at 95°C for 10 min and 50 cycles at 95°C for 15 s and 60°C for 1 min. The fluorescence level was measured with an ABI PRISM 7900HT sequence detector (Applied Biosystems). Genotypes were determined by ABI SDS software.

UGT1A1 TA tandem repeat (rs8175347) genotyping was conducted using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (Sequenom). The method includes an initial step of PCR amplification of the UGT1A1 gene region containing the polymorphism followed by allele-specific primer extension using extension primers. The assay for the UGT1A1 tandem repeat (rs8175347) was done using homogeneous MassEXTEND primer extension reaction. This MassEXTEND reaction generates allele-specific primer extension products that are typically several bases longer than the original MassEXTEND primer. A specifically designed MassEXTEND primer generated extension products of varying lengths corresponding to the number of (TA)n repeats; 8 to 14 bases were added to the MassEXTEND primer. Genotypes of n = 5, 5/6, 5/7, 5/8, 6, 6/7, 6/8, 7, 7/8, and 8 were distinguishable in this assay; however, only the 6/6, 6/7, and 7/7 genotypes were present in our study population. Primer sequences were as follows: forward primer CCTTTGCTCCTGCCAGAGG, reverse primer AAACATTAACTTGGTGTATCGATT, and MassEXTEND primer CGCCCTCTCCTACTTATATA.

The laboratory staff was blind to the identity of the subjects. Quality-control samples were included in each of the genotyping assays. Each 384-well plate contained four water, eight CEPH 1347-02 DNA, eight blinded quality-control DNA samples (where sample identity is unknown to technician), and eight unblinded quality-control DNA samples (duplicates of samples on the plate). The concordance rates for the quality-control samples were 98.9% for rs2070959, 100.0% for rs887829, and 100.0% for rs8175347. In addition, we genotyped rs887829 in 45 DNA samples of the Chinese participants used in the International HapMap Project and 24 DNA samples used in the Perlegen genotype database and rs2070959 in 24 DNA samples used by Perlegen as an additional quality control. The genotypes of the samples generated from our study were compared with data downloaded from HapMap (http://www.hapmap.org) and/or Perlegen (http://genome.perlegen.com). The concordance rates between the data generated in our laboratory and the data from the Perlegen database were 100%.

Among those who provided a DNA sample, genotyping data were obtained from 1,003 (95.3%) cases and 1,004 (94.9%) controls for the rs2070959 polymorphism, 1,039 (98.8%) cases and 1,027 (97.1%) controls for the rs887829 polymorphisms, and 1,030 (97.9%) cases and 998 (94.3%) controls for the rs8175347 polymorphism. The primary reasons for incomplete genotyping were insufficient DNA used for the assay and unsuccessful PCR amplification.

Soy Food Intake

Total soy food intake was measured by soy protein and soy isoflavone intake using information collected with the food frequency questionnaire, which specifically asked about consumption of soymilk, tofu, soybean sprouts, fresh soy beans, and other soy-containing products. Study participants were asked how often they consumed these foods followed by a question on how many liang (1 liang = 50 g) they ate of the particular food(s) per unit of time (day, week, month, or year) in the preceding 5 years. Soy protein intake was estimated by multiplying the amount of soy food consumed with the amount of protein in that food according to the Chinese Food Composition Tables (38). Isoflavone intake was computed using published data on the isoflavone content of each soy food (39, 40).

Statistical Analysis

χ2 statistics were used to compare differences in the distribution of demographic and risk factors between those with and without genotyping information as well as compare case-control differences among those genotyped for UGT1A1 polymorphisms. Multivariable conditional logistic regression was used to estimate odds ratios (OR) and 95% confidence intervals (95% CI) to measure the strength of association while adjusting for confounders. Factors considered as potential confounders included age, education, first-degree family history of cancer, years of menstruation, age at menarche, age at menopause, number of pregnancies, menopausal status, oral contraceptive use, hormone replacement therapy, physical activity, body mass index (BMI), total caloric intake, vegetable intake, soy intake, marital status, and income.

The estimates from fully adjusted models did not differ substantially from those in minimally adjusted models. Final models presented were deemed most parsimonious and were adjusted for age (continuous variable), education (no formal education, elementary education, middle school, high school, college), first-degree family history of cancer (yes/no), years of menstruation (quartile), parity (continuous), menopausal status (pre/post), oral contraceptive use (ever/never), regular physical activity (yes/no), BMI (quartiles), total caloric intake (quartiles), vegetable intake (quartiles), and soy intake (quartiles).

Single nucleotide polymorphisms in the UGT1A1 gene were evaluated using a three-level categorization (homozygote-referent, heterozygote, homozygote-minor allele variant). Multiplicative interactions between the genotype and estrogen exposure variables were evaluated by comparing the difference of the log likelihoods between a model with the main effects and a model with both the main effects and the interaction term. Haplotypes for three polymorphic sites (in the order: rs2070959, rs8878929, and rs8175347) were generated using the Haploview program, which employs an expectation-maximization algorithm to estimate haplotypes (41). ORs and 95% CIs for the association between haplotypes and endometrial cancer risk were generated using the Hapstat program (42). Associations between genotypes, haplotypes, and endometrial cancer risk were evaluated under additive, dominant, and recessive genetic models. Statistical analyses were conducted using SAS software (version 9.1). Linkage disequilibrium was calculated using the GENETICS package in the free software R (version 1.9.1; ref. 43).

Selected demographic and risk factor characteristics of the subjects genotyped for UGT1A1 polymorphisms in this study are presented in Table 1. Cases and controls were similar with respect to age and educational status. Compared with controls, cases were more likely to have a family history of cancer, have a lower age at menarche, have a later age at menopause, use hormone replacement therapy, have fewer live births, have a higher BMI and waist-hip ratio, and have higher daily caloric intake while they were less likely to have ever used oral contraceptives or engage in regular physical activity. There were no appreciable differences in the distribution of demographic or risk factors between the entire study population and those with genotyping data (data not shown).

Table 1.

Comparison of cases and controls with genotyping data by selected descriptive characteristics, the SECS, 1997-2003

Subject characteristicsCases (n = 1,047)Controls (n = 1,035)P*
Age (y), x ± SD 54.7 ± 8.5 54.9 ± 8.5 0.70 
Middle school education or more (%) 78.5 77.6 0.38 
Regular smoker (%) 3.2 3.5 0.69 
Regular alcohol consumption (%) 3.1 5.5 <0.01 
First-degree relative with cancer (%) 35.3 29.1 <0.01 
Age at menarche, x ± SD 14.5 ± 1.7 14.8 ± 1.8 <0.01 
No. live births, x ± SD 1.7 ± 1.2 1.8 ± 1.7 0.03 
Postmenopausal (%) 56.7 61.9 0.02 
Age at menopause,x ± SD 50.3 ± 3.6 49.0 ± 3.6 <0.01 
Years of menstruation, x ± SD 32.8 ± 4.9 30.7 ± 5.3 <0.01 
Ever used oral contraceptives (%) 18.2 25.1 <0.01 
Ever used hormone replacement therapy (%) 4.7 4.3 0.62 
BMI, median (25th, 75th percentile) 25.2 (22.9, 28.1) 23.5 (21.4, 26.0) <0.01 
Waist-to-hip ratio (25th, 75th percentile) 0.84 (0.80, 0.87) 0.81 (0.78, 0.85) <0.01 
Engaged in regular physical activity (%) 28.1 33.8 <0.01 
Daily energy intake (kcal), x ± SD 1,792.75 ± 458.9 1,749.5 ± 461.9 0.02 
Daily intake of soy (g), x ± SD 153.1 ± 126.4 159.1 ± 136.5 0.59 
Daily intake of all vegetables (g), x ± SD 331.9 ± 204.3 325.5 ± 197.0 0.60 
Daily intake of legumes§ (g), x ± SD 43.7 ± 38.5 45.2 ± 38.3 0.20 
Daily intake of Allium foods (g), x ± SD 7.5 ± 16.3 8.2 ± 18.4 0.14 
Total lifetime tea consumption (liang), x ± SD 897.6 ± 1,022.6 859.7 ± 1,041.8 0.92 
Daily intake of soy (g/day), median (25th, 75th percentile) 110.9 (55.0, 224.7) 113.6 (53.3, 238.8) 0.61 
Daily intake of legumes,§ median (25th, 75th percentile) 42.9 (18.2, 55.6) 35.6 (19.3, 58.9) 0.22 
Daily intake of isoflavones, median (25th, 75th percentile) 31.5 (17.3, 46.0) 31.2 (16.8, 47.1) 0.79 
Subject characteristicsCases (n = 1,047)Controls (n = 1,035)P*
Age (y), x ± SD 54.7 ± 8.5 54.9 ± 8.5 0.70 
Middle school education or more (%) 78.5 77.6 0.38 
Regular smoker (%) 3.2 3.5 0.69 
Regular alcohol consumption (%) 3.1 5.5 <0.01 
First-degree relative with cancer (%) 35.3 29.1 <0.01 
Age at menarche, x ± SD 14.5 ± 1.7 14.8 ± 1.8 <0.01 
No. live births, x ± SD 1.7 ± 1.2 1.8 ± 1.7 0.03 
Postmenopausal (%) 56.7 61.9 0.02 
Age at menopause,x ± SD 50.3 ± 3.6 49.0 ± 3.6 <0.01 
Years of menstruation, x ± SD 32.8 ± 4.9 30.7 ± 5.3 <0.01 
Ever used oral contraceptives (%) 18.2 25.1 <0.01 
Ever used hormone replacement therapy (%) 4.7 4.3 0.62 
BMI, median (25th, 75th percentile) 25.2 (22.9, 28.1) 23.5 (21.4, 26.0) <0.01 
Waist-to-hip ratio (25th, 75th percentile) 0.84 (0.80, 0.87) 0.81 (0.78, 0.85) <0.01 
Engaged in regular physical activity (%) 28.1 33.8 <0.01 
Daily energy intake (kcal), x ± SD 1,792.75 ± 458.9 1,749.5 ± 461.9 0.02 
Daily intake of soy (g), x ± SD 153.1 ± 126.4 159.1 ± 136.5 0.59 
Daily intake of all vegetables (g), x ± SD 331.9 ± 204.3 325.5 ± 197.0 0.60 
Daily intake of legumes§ (g), x ± SD 43.7 ± 38.5 45.2 ± 38.3 0.20 
Daily intake of Allium foods (g), x ± SD 7.5 ± 16.3 8.2 ± 18.4 0.14 
Total lifetime tea consumption (liang), x ± SD 897.6 ± 1,022.6 859.7 ± 1,041.8 0.92 
Daily intake of soy (g/day), median (25th, 75th percentile) 110.9 (55.0, 224.7) 113.6 (53.3, 238.8) 0.61 
Daily intake of legumes,§ median (25th, 75th percentile) 42.9 (18.2, 55.6) 35.6 (19.3, 58.9) 0.22 
Daily intake of isoflavones, median (25th, 75th percentile) 31.5 (17.3, 46.0) 31.2 (16.8, 47.1) 0.79 
*

From χ2 test (categorical variables) or t test (continuous variables).

Among postmenopausal women.

Soy: soy protein, tofu, soy bean sprouts, fresh soy beans, and other soy products.

§

Legumes: fresh soybeans, fresh broad beans, yard long beans, green beans, hyacinth beans/snow peas, peanuts dry soy beans, soy bean sprouts, mung beans, and mung bean sprouts.

Allium foods: onions, garlic and garlic shoots, heads of garlic, Chinese chives, and green onions.

The distribution of UGT1A1 genotypes for rs2070959, rs887829, and rs8175347 are presented in Table 2. Among controls, rs8175347 frequencies were not consistent with Hardy-Weinberg equilibrium (HWE), whereas, among cases, rs2070959 and rs8175347 frequencies were not consistent with HWE. Genotyping of control samples found a 98.9% concordance rate for rs2070959, and there was no evidence of genotyping errors such as preferential amplification of one allele over the other. Of note, the distribution of rs2070959 genotypes differed substantially from those presented by HapMap, but not from those reported by Perlegen, for the Chinese Han population. For example, HapMap genotype frequencies for the rs2070959 polymorphism were 8.9% for AA, 44.4% for AG, and 46.7% for GG, whereas the frequencies were 31.0% for AA, 33.3% for AG, and 6.0% for GG among controls in our study population of Han Chinese. To clarify these inconsistencies, we genotyped (44) Chinese samples used in HapMap and found the consistency rate was only 11%. We additionally genotyped (24) Chinese samples provided by Perlegen and the consistency rate of our data with Perlegen was 100%. Perlegen and HapMap polymorphism rates are consistent for Caucasian and African populations but not for the Asian population. We contacted Cold Spring Harbor Laboratories, which informed us that there are some genotype frequencies published in HapMap that may not be correct, and this may be one such example.

Table 2.

Genotype frequencies for polymorphisms in the UGT1A1 gene, the SECS, 1997-2003

GenotypesLocationPositionCases (n = 1,047)Controls (n = 1,035)Age-adjusted OR (95% CI)Fully Adjusted OR (95% CI)*
rs2070959 Varies; dependent on splicing Chr. 2 234383941     
    A/A   62.2 61.0 1.0 1.0 
    A/G   34.9 33.0 1.0 (0.8-1.2) 1.0 (0.9-1.3) 
    G/G   2.9 6.0 0.5 (0.3-0.8) 0.5 (0.3-0.8) 
      HWE P 0.02 0.09 Ptrend 0.10 0.27 
rs887829 Promoter Chr. 2 234450570     
    G/G   78.2 81.1 1.0 1.0 
    G/A   20.7 17.2 1.2 (1.0-1.6) 1.3 (1.0-1.6) 
    A/A   1.1 1.7 0.8 (0.4-1.6) 0.6 (0.3-1.4) 
      HWE P 0.42 0.09 Ptrend 0.22 0.30 
rs8175347 Promoter Chr. 2 234450870     
    6/6   78.2 81.0 1.0 1.0 
    6/7   19.5 16.6 1.2 (1.0-1.6) 1.3 (1.0-1.6) 
    7/7   2.3 2.4 1.1 (0.6-1.9) 1.0 (0.6-1.9) 
      HWE P 0.01 <0.0001 Ptrend 0.18 0.19 
Haplotypesb   Frequency among controls    
    AG6    0.692 1.0 1.0 
    AA7    0.083 1.0 (0.7-1.4) 1.0 (0.7-1.3) 
    GG6    0.195 0.9 (0.7-1.2) 0.9 (0.7-1.2) 
GenotypesLocationPositionCases (n = 1,047)Controls (n = 1,035)Age-adjusted OR (95% CI)Fully Adjusted OR (95% CI)*
rs2070959 Varies; dependent on splicing Chr. 2 234383941     
    A/A   62.2 61.0 1.0 1.0 
    A/G   34.9 33.0 1.0 (0.8-1.2) 1.0 (0.9-1.3) 
    G/G   2.9 6.0 0.5 (0.3-0.8) 0.5 (0.3-0.8) 
      HWE P 0.02 0.09 Ptrend 0.10 0.27 
rs887829 Promoter Chr. 2 234450570     
    G/G   78.2 81.1 1.0 1.0 
    G/A   20.7 17.2 1.2 (1.0-1.6) 1.3 (1.0-1.6) 
    A/A   1.1 1.7 0.8 (0.4-1.6) 0.6 (0.3-1.4) 
      HWE P 0.42 0.09 Ptrend 0.22 0.30 
rs8175347 Promoter Chr. 2 234450870     
    6/6   78.2 81.0 1.0 1.0 
    6/7   19.5 16.6 1.2 (1.0-1.6) 1.3 (1.0-1.6) 
    7/7   2.3 2.4 1.1 (0.6-1.9) 1.0 (0.6-1.9) 
      HWE P 0.01 <0.0001 Ptrend 0.18 0.19 
Haplotypesb   Frequency among controls    
    AG6    0.692 1.0 1.0 
    AA7    0.083 1.0 (0.7-1.4) 1.0 (0.7-1.3) 
    GG6    0.195 0.9 (0.7-1.2) 0.9 (0.7-1.2) 
*

Conditional logistic regression adjusted for age, education, first-degree family history of cancer, years of menstruation, parity, menopausal status, oral contraceptive use, physical activity, BMI, total caloric intake, vegetable intake, and soy intake.

An apparent association was found for the rs2070959 genotype with endometrial cancer (OR, 0.5; 95% CI, 0.3-0.8) for minor allele homozygous carriers. This association persisted after multivariable adjustment. There was no evidence of gene effects for the rs887829 and rs8175347 genotypes among all women in the study population (Table 2). Also presented are the results for UGT1A1 haplotype analysis. Only three haplotypes occurred at a frequency of greater than 5% among controls. No significant associations were observed for any of the major haplotypes and endometrial cancer risk under dominant, recessive, or additive models (only the results from the additive model are presented in Table 2).

Because the UGT1A1 gene is an important enzyme in the estrogen metabolism pathway, we theorized that the effect of the genetic variants on endometrial cancer risk may vary with levels of estrogenic exposures. Table 3 presents the joint effect of UGT1A1 genotypes and indicators of endogenous estrogen exposure on endometrial cancer risk. We found that the effect of the rs2070959 homozygous variant genotype was most evident among subgroups defined by lower levels of endogenous estrogen, including postmenopausal women, those with low BMI, and those who menstruated for fewer than 32 years, although a test for multiplicative interaction was not significant. There was no evidence of a significant multiplicative interaction between the other two genotypes and any of the endogenous estrogenic exposures, with the exception of the 7/7 genotype in the rs8175347 polymorphism and oral contraceptive use, although the limited sample size of the homozygous variant genotype may be responsible for this result. We also evaluated age at diagnosis, family history of cancer, smoking status, and parity on their interaction with genotype on endometrial cancer risk. However, there was no evidence of modification of the genotype effect on endometrial cancer risk by any of these factors.

Table 3.

Joint effects of endogenous markers of estrogen exposure and UGT1A1 polymorphisms on endometrial cancer risk, the SECS, 1997-2003

All women
VariablesCases/controlsOR (95% CI)Cases/controlsOR (95% CI)Pinteraction
  Postmenopausal*  Premenopausal*   
rs2070959 AA 356/374 1.0 268/239 1.1 (0.8-1.5) 0.19 
 AG 190/204 0.9 (0.7-1.2) 160/127 1.3 (0.9-1.8)  
 GG 16/41 0.4 (0.2-0.8) 13/19 0.8 (0.3-1.7)  
rs887829 GG 455/514 1.0 357/319 1.2 (0.9-1.7) 0.49 
 GA 125/108 1.3 (1.0-1.8) 90/69 1.4 (0.9-2.1)  
 AA 9/12 0.8 (0.3-1.9) 3/5 0.5 (0.1-2.5)  
rs8175347 6/6 450/505 1.0 355/303 1.4 (1.0-1.9) 0.17 
 6/7 115/102 1.3 (1.0-1.8) 86/64 1.6 (1.0-2.4)  
 7/7 17/13 1.4 (0.6-3.0) 7/11 0.7 (0.3-2.0)  
       
  Low BMI (less than median)  High BMI (greater than or equal to median)   
rs2070959 AA 298/403 1.0 326/210 2.2 (1.7-2.7) 0.61 
 AG 172/222 1.1 (0.9-1.4) 178/109 3.0 (2.1-4.3)  
 GG 13/42 0.6 (0.2-1.7) 16/18 1.6 (0.5-5.0)  
rs887829 GG 402/556 1.0 410/277 2.1 (1.7-2.6) 0.57 
 GA 89/119 1.1 (0.8-1.5) 126/58 3.0 (2.1-4.3)  
 AA 5/11 0.6 (0.2.-1.7) 7/6 1.6 (0.5-5.0)  
rs8175347 6/6 403/544 1.0 402/264 2.2 (1.8-2.2) 0.48 
 6/7 85/114 1.1 (0.8-1.5) 116/52 3.1 (2.1-4.4)  
 7/7 10/14 1.0 (0.4-2.2) 14/10 2.0 (0.8-4.5)  
       
  Low waist-hip ratio (less than median)b  High waist-hip ratio (greater than or equal to median)b   
rs2070959 AA 287/402 1.0 337/211 1.6 (1.3-2.1) 0.90 
 AG 175/232 1.0 (0.8-1.3) 175/99 1.8 (1.4-2.5)  
 GG 14/38 0.5 (0.3-0.9) 15/22 0.8 (0.4-1.7)  
rs887829 GG 392/553 1.0 420/280 1.7 (1.4-2.2) 0.36 
 GA 97/120 1.2 (0.9-1.7) 118/57 2.2 (1.5-3.2)  
 AA 3/12 0.3 (0.1-1.3) 9/5 1.9 (0.6-5.9)  
rs8175347 6/6 389/539 1.0 416/274 1.8 (1.4-2.2) 0.78 
 6/7 89/113 1.2 (0.9-1.6) 112/53 2.3 (1.6-3.3)  
 7/7 13/16 1.2 (0.5-2.6) 11/8 1.4 (0.5-3.6)  
       
  Ever oral contraceptive use  Never oral contraceptive use   
rs2070959 AA 121/164 1.0 503/449 1.1 (0.9-1.4) 0.71 
 AG 58/73 0.9 (0.6-1.4) 292/258 1.2 (0.9-1.5)  
 GG 4/16 0.4 (0.1-1.1) 25/44 0.6 (0.4-1.1)  
rs887829 GG 137/209 1.0 675/624 1.4 (1.1-1.8) 0.08 
 GA 50/48 1.6 (1.0-2.6) 165/129 1.6 (1.1-2.2)  
 AA 3/2 1.7 (0.2-11.6) 9/15 0.7 (0.3-1.8)  
rs8175347 6/6 135/202 1.0 670/606 1.5 (1.2-1.9) 0.04 
 6/7 47/47 1.6 (1.0-2.6) 154/119 1.7 (1.2-2.4)  
 7/7 6/2 4.5 (0.8-24.3) 18/22 1.1 (0.6-2.2)  
       
  <32 y of menstruation§  ≥32 y of menstruation§   
rs2070959 AA 250/338 1.0 374/275 1.6 (1.3-2.1) 0.65 
 AG 141/181 1.0 (0.7-1.3) 209/150 1.7 (1.3-2.2)  
 GG 11/36 0.4 (0.2-0.8) 18/24 0.9 (0.5-1.7)  
rs887829 GG 327/459 1.0 485/374 1.7 (1.4-2.1) 0.69 
 GA 80/102 1.1 (0.8-1.6) 135/75 2.5 (1.8-3.5)  
 AA 6/7 1.2 (0.4-3.7) 6/10 0.8 (0.3-2.2)  
rs8175347 6/6 322/440 1.0 483/368 1.8 (1.5-2.2) 0.94 
 6/7 74/99 1.1 (0.8-1.5) 127/67 2.7 (1.9-3.8)  
 7/7 12/12 1.4 (0.6-3.3) 12/12 1.3 (0.6-2.9)  
All women
VariablesCases/controlsOR (95% CI)Cases/controlsOR (95% CI)Pinteraction
  Postmenopausal*  Premenopausal*   
rs2070959 AA 356/374 1.0 268/239 1.1 (0.8-1.5) 0.19 
 AG 190/204 0.9 (0.7-1.2) 160/127 1.3 (0.9-1.8)  
 GG 16/41 0.4 (0.2-0.8) 13/19 0.8 (0.3-1.7)  
rs887829 GG 455/514 1.0 357/319 1.2 (0.9-1.7) 0.49 
 GA 125/108 1.3 (1.0-1.8) 90/69 1.4 (0.9-2.1)  
 AA 9/12 0.8 (0.3-1.9) 3/5 0.5 (0.1-2.5)  
rs8175347 6/6 450/505 1.0 355/303 1.4 (1.0-1.9) 0.17 
 6/7 115/102 1.3 (1.0-1.8) 86/64 1.6 (1.0-2.4)  
 7/7 17/13 1.4 (0.6-3.0) 7/11 0.7 (0.3-2.0)  
       
  Low BMI (less than median)  High BMI (greater than or equal to median)   
rs2070959 AA 298/403 1.0 326/210 2.2 (1.7-2.7) 0.61 
 AG 172/222 1.1 (0.9-1.4) 178/109 3.0 (2.1-4.3)  
 GG 13/42 0.6 (0.2-1.7) 16/18 1.6 (0.5-5.0)  
rs887829 GG 402/556 1.0 410/277 2.1 (1.7-2.6) 0.57 
 GA 89/119 1.1 (0.8-1.5) 126/58 3.0 (2.1-4.3)  
 AA 5/11 0.6 (0.2.-1.7) 7/6 1.6 (0.5-5.0)  
rs8175347 6/6 403/544 1.0 402/264 2.2 (1.8-2.2) 0.48 
 6/7 85/114 1.1 (0.8-1.5) 116/52 3.1 (2.1-4.4)  
 7/7 10/14 1.0 (0.4-2.2) 14/10 2.0 (0.8-4.5)  
       
  Low waist-hip ratio (less than median)b  High waist-hip ratio (greater than or equal to median)b   
rs2070959 AA 287/402 1.0 337/211 1.6 (1.3-2.1) 0.90 
 AG 175/232 1.0 (0.8-1.3) 175/99 1.8 (1.4-2.5)  
 GG 14/38 0.5 (0.3-0.9) 15/22 0.8 (0.4-1.7)  
rs887829 GG 392/553 1.0 420/280 1.7 (1.4-2.2) 0.36 
 GA 97/120 1.2 (0.9-1.7) 118/57 2.2 (1.5-3.2)  
 AA 3/12 0.3 (0.1-1.3) 9/5 1.9 (0.6-5.9)  
rs8175347 6/6 389/539 1.0 416/274 1.8 (1.4-2.2) 0.78 
 6/7 89/113 1.2 (0.9-1.6) 112/53 2.3 (1.6-3.3)  
 7/7 13/16 1.2 (0.5-2.6) 11/8 1.4 (0.5-3.6)  
       
  Ever oral contraceptive use  Never oral contraceptive use   
rs2070959 AA 121/164 1.0 503/449 1.1 (0.9-1.4) 0.71 
 AG 58/73 0.9 (0.6-1.4) 292/258 1.2 (0.9-1.5)  
 GG 4/16 0.4 (0.1-1.1) 25/44 0.6 (0.4-1.1)  
rs887829 GG 137/209 1.0 675/624 1.4 (1.1-1.8) 0.08 
 GA 50/48 1.6 (1.0-2.6) 165/129 1.6 (1.1-2.2)  
 AA 3/2 1.7 (0.2-11.6) 9/15 0.7 (0.3-1.8)  
rs8175347 6/6 135/202 1.0 670/606 1.5 (1.2-1.9) 0.04 
 6/7 47/47 1.6 (1.0-2.6) 154/119 1.7 (1.2-2.4)  
 7/7 6/2 4.5 (0.8-24.3) 18/22 1.1 (0.6-2.2)  
       
  <32 y of menstruation§  ≥32 y of menstruation§   
rs2070959 AA 250/338 1.0 374/275 1.6 (1.3-2.1) 0.65 
 AG 141/181 1.0 (0.7-1.3) 209/150 1.7 (1.3-2.2)  
 GG 11/36 0.4 (0.2-0.8) 18/24 0.9 (0.5-1.7)  
rs887829 GG 327/459 1.0 485/374 1.7 (1.4-2.1) 0.69 
 GA 80/102 1.1 (0.8-1.6) 135/75 2.5 (1.8-3.5)  
 AA 6/7 1.2 (0.4-3.7) 6/10 0.8 (0.3-2.2)  
rs8175347 6/6 322/440 1.0 483/368 1.8 (1.5-2.2) 0.94 
 6/7 74/99 1.1 (0.8-1.5) 127/67 2.7 (1.9-3.8)  
 7/7 12/12 1.4 (0.6-3.3) 12/12 1.3 (0.6-2.9)  
*

ORs adjusted for age, alcohol consumption, education, first-degree relative with cancer, oral contraceptive use, years of menstruation, parity, physical activity, total caloric intake, soy intake, and total vegetable intake.

ORs adjusted for age, alcohol consumption, education, first-degree relative with cancer, oral contraceptive use, years of menstruation, parity, menopausal status, physical activity, total caloric intake, soy intake, and total vegetable intake.

ORs adjusted for age, alcohol consumption, education, first-degree relative with cancer, years of menstruation, parity, menopausal status, physical activity, total caloric intake, soy intake, and total vegetable intake.

§

ORs adjusted for age, alcohol consumption, education, first-degree relative with cancer, oral contraceptive use, parity, menopausal status, physical activity, total caloric intake, soy intake, and total vegetable intake.

The joint associations of UGT1A1 polymorphisms and soy food intake with endometrial cancer are presented in Table 4. Similar to the results observed with endogenous estrogen exposures, significant reductions in endometrial cancer risk were associated with carriers of the rs2070959 homozygous variant genotype. This association was more evident among subjects with the lowest soy protein and isoflavone intake. Similarly, rs887829 and rs8175347 polymorphisms were not related to the risk of endometrial cancer among any subgroup defined by soy protein or isoflavone intake. The gene-disease association did not differ by tea or Allium food intake (data not shown).

Table 4.

Joint effects of dietary exposures and UGT1A1 polymorphisms on endometrial cancer risk, the SECS, 1997-2003

All women
VariablesCases/controlsOR (95% CI)Cases/controlsOR (95% CI)Pinteraction
  Low soy protein intake (less than median)  High soy protein intake (greater than or equal to median)   
rs2070959 AA 317/297 1.0 307/316 0.8 (0.6-1.0) 0.07 
 AG 191/168 1.1 (0.8-1.4) 159/163 0.8 (0.6-1.1)  
 GG 9/35 0.3 (0.1-0.6) 20/25 0.7 (0.4-1.4)  
rs887829 GG 415/416 1.0 397/417 0.9 (0.7-1.1) 0.84 
 GA 115/90 1.2 (0.9-1.7) 100/87 1.1 (0.8-1.5)  
 AA 5/8 0.5 (0.2-1.7) —/— —  
rs8175347 6/6 415/403 1.0 390/405 0.9 (0.7-1.1) 0.86 
 6/7 105/84 1.2 (0.9-1.7) 96/82 1.1 (0.8-1.6)  
 7/7 10/10 1.0 (0.4-2.6) 14/14 0.9 (0.4-1.9)  
       
  Low legume intake (less than median)  High legume intake (greater than or equal to median)   
rs2070959 AA 298/403 1.0 326/210 1.1 (0.8-1.5) 0.42 
 AG 172/222 0.9 (0.7-1.2) 178/109 1.3 (0.9-1.8)  
 GG 13/42 0.4 (0.2-0.8) 16/18 0.8 (0.3-1.7)  
rs887829 GG 402/556 1.0 410/277 1.2 (0.9-1.7) 0.10 
 GA 89/119 1.3 (1.0-1.8) 126/58 1.4 (0.9-2.1)  
 AA 5/11 0.8 (0.3-1.9) 7/7 0.5 (0.1-2.5)  
rs8175347 6/6 403/544 1.0 402/264 1.4 (1.0-1.9) 0.76 
 6/7 85/114 1.3 (1.0-1.8) 116/52 1.6 (1.0-2.4)  
 7/7 10/14 1.4 (0.6-3.0) 14/10 0.7 (0.3-2.0)  
       
  Low isoflavone intake (less than median)  High isoflavone intake (greater than or equal to median)   
rs2070959 AA 298/403 1.0 326/210 1.1 (0.8-1.5) 0.42 
 AG 172/222 0.9 (0.7-1.2) 178/109 1.3 (0.9-1.8)  
 GG 13/42 0.4 (0.2-0.8) 16/18 0.8 (0.3-1.7)  
rs887829 GG 402/556 1.0 410/277 1.2 (0.9-1.7) 0.10 
 GA 89/119 1.3 (1.0-1.8) 126/58 1.4 (0.9-2.1)  
 AA 5/11 0.8 (0.3-1.9) 7/7 0.5 (0.1-2.5)  
rs8175347 6/6 403/544 1.0 402/264 1.4 (1.0-1.9) 0.76 
 6/7 85/114 1.3 (1.0-1.8) 116/52 1.6 (1.0-2.4)  
 7/7 10/14 1.4 (0.6-3.0) 14/10 0.7 (0.3-2.0)  
All women
VariablesCases/controlsOR (95% CI)Cases/controlsOR (95% CI)Pinteraction
  Low soy protein intake (less than median)  High soy protein intake (greater than or equal to median)   
rs2070959 AA 317/297 1.0 307/316 0.8 (0.6-1.0) 0.07 
 AG 191/168 1.1 (0.8-1.4) 159/163 0.8 (0.6-1.1)  
 GG 9/35 0.3 (0.1-0.6) 20/25 0.7 (0.4-1.4)  
rs887829 GG 415/416 1.0 397/417 0.9 (0.7-1.1) 0.84 
 GA 115/90 1.2 (0.9-1.7) 100/87 1.1 (0.8-1.5)  
 AA 5/8 0.5 (0.2-1.7) —/— —  
rs8175347 6/6 415/403 1.0 390/405 0.9 (0.7-1.1) 0.86 
 6/7 105/84 1.2 (0.9-1.7) 96/82 1.1 (0.8-1.6)  
 7/7 10/10 1.0 (0.4-2.6) 14/14 0.9 (0.4-1.9)  
       
  Low legume intake (less than median)  High legume intake (greater than or equal to median)   
rs2070959 AA 298/403 1.0 326/210 1.1 (0.8-1.5) 0.42 
 AG 172/222 0.9 (0.7-1.2) 178/109 1.3 (0.9-1.8)  
 GG 13/42 0.4 (0.2-0.8) 16/18 0.8 (0.3-1.7)  
rs887829 GG 402/556 1.0 410/277 1.2 (0.9-1.7) 0.10 
 GA 89/119 1.3 (1.0-1.8) 126/58 1.4 (0.9-2.1)  
 AA 5/11 0.8 (0.3-1.9) 7/7 0.5 (0.1-2.5)  
rs8175347 6/6 403/544 1.0 402/264 1.4 (1.0-1.9) 0.76 
 6/7 85/114 1.3 (1.0-1.8) 116/52 1.6 (1.0-2.4)  
 7/7 10/14 1.4 (0.6-3.0) 14/10 0.7 (0.3-2.0)  
       
  Low isoflavone intake (less than median)  High isoflavone intake (greater than or equal to median)   
rs2070959 AA 298/403 1.0 326/210 1.1 (0.8-1.5) 0.42 
 AG 172/222 0.9 (0.7-1.2) 178/109 1.3 (0.9-1.8)  
 GG 13/42 0.4 (0.2-0.8) 16/18 0.8 (0.3-1.7)  
rs887829 GG 402/556 1.0 410/277 1.2 (0.9-1.7) 0.10 
 GA 89/119 1.3 (1.0-1.8) 126/58 1.4 (0.9-2.1)  
 AA 5/11 0.8 (0.3-1.9) 7/7 0.5 (0.1-2.5)  
rs8175347 6/6 403/544 1.0 402/264 1.4 (1.0-1.9) 0.76 
 6/7 85/114 1.3 (1.0-1.8) 116/52 1.6 (1.0-2.4)  
 7/7 10/14 1.4 (0.6-3.0) 14/10 0.7 (0.3-2.0)  

NOTE: ORs adjusted for age, alcohol consumption, education, first-degree relative with cancer, oral contraceptive use, years of menstruation, parity, menopausal status, physical activity, total caloric intake, and total vegetable intake.

In this analysis, we sought to elucidate the role of polymorphisms in the UGT1A1 gene in endometrial cancer risk as well as clarify their potential for interaction with exogenous phytoestrogen and endogenous estrogen exposure. Our results from this population-based case-control study of Chinese women indicate that the variant allele (G) in the rs2070959 polymorphism was significantly associated with reduced risk of endometrial cancer. Furthermore, we found that this effect appeared to be more evident among women who were postmenopausal, had low BMI and waist-hip ratio, and had low soy protein, legume, and isoflavone intake.

There are two previous studies of genetic variation in UGT1A1 and endometrial cancer risk. The Nurses' Health Study found a significant decrease in risk among carriers of the homozygous variant genotype for the rs8175347 polymorphism, and the association was more pronounced among premenopausal women (34). It was also shown in this study that UGT1A1 had a preference for the glucuronidation of 2-hydroxyestradiol, suggesting that decreased expression of UGT1A1 could lead to a decreased risk of endometrial cancer by altering the conjugation of 2-hydroxy estrogen metabolites. In our study, we did not find that the rs8175347 polymorphism was related to the risk of endometrial cancer.

Our results were consistent with that of the Women's Insights and Shared Experiences Study, which also showed a null association with this polymorphism (33). The rs8175347 promoter polymorphism has been associated previously with altered levels of E1 and E2 (31, 45). Because Chinese women typically have a lower level of estrogen compared with their Western counterparts (44, 46), the estrogenic effects of this promoter polymorphism may obviate any reduction in the risk of endometrial cancer associated with alterations in the metabolism of 2-hydroxyestradiol (34). Future research should consider incorporating direct measures of endogenous hormones in assessing the gene-disease association.

To our knowledge, this is the first study to investigate the role of the rs2070959 polymorphism in endometrial cancer risk. The frequency of the minor allele (G) was 6.0% among our population-based controls (mainly composed of Han Chinese). Although the frequency of the homozygous variant genotype (G/G) is low among the Han Chinese population and therefore less likely to have an large effect at the population level among this group, this polymorphism may make a more relevant contribution to endometrial cancer risk in the African and Caucasian/European populations where the G allele occurs at frequency of 15% and 46.7%, respectively. Further laboratory follow-up is needed to ascertain the functional implications of the amino acid change. Furthermore, the effect of the variant rs2070959 polymorphism (G) was most evident among women with low estrogen levels and low consumption of soy food. Given existing evidence that isoflavones may act through multiple mechanisms to reduce carcinogenesis and cancer progression, it is possible that the effect of this polymorphism is obscured by the competing pathways under conditions of high estrogen or polyphenol exposure (47-49).

There are a few limitations that should be noted in this study. First, we had modest power to detect associations in some subgroups, particularly among ever users of oral contraceptives with homozygote variant genotype. Second, the rs8175347 polymorphism was not in HWE among controls. We did, however, exclude the possibility of genotyping errors for this polymorphism. Furthermore, because this polymorphism is in strong linkage disequilibrium (correlation, 96%) with the rs887829 polymorphism, we are able to use this polymorphism as a surrogate for conclusions about the association of this area of the gene with endometrial cancer risk. There was a departure from HWE among cases for both rs8175347 and rs2070959 polymorphisms; however, given that genotyping errors were ruled out, it is possible that this departure is due to a real association with endometrial cancer. Third, we included three polymorphisms in this study, which may not be as comprehensive as studies now employing the haplotype-tagging single nucleotide polymorphism approach. These polymorphisms were selected on a literature review of published data and for their potential for functionality in 2005. As such, we cannot exclude the possibility that there may be other polymorphisms that interact with endogenous estrogen exposures or soy foods and that these unstudied polymorphisms may be responsible for the gene-environment interactions observed in this study. Finally, limited sample size (of class II endometrial cancers) precluded us from evaluating the association between the genotypes and cancer pathology.

This study possesses several strengths, including its population-based design, high response rate, high DNA collection rate, and extensive quantification of dietary intake. In addition, this study population was composed almost entirely of Han Chinese (>98%), which reduces the likelihood of confounding by ethnicity. Furthermore, this population has higher consumption of soy food compared with Western populations, which allowed for an in-depth evaluation of these exposures and gene-diet interaction. In addition, a very small proportion of our study population received hormone replacement therapy, which allowed for the investigation of dietary and endogenous markers of estrogen without the influence exogenous hormones. Similarly in this population, the prevalence of smoking, an exposure associated with lower estrogen levels, is <3%.

In summary, the results from this population-based study of Chinese women suggest that the rs2070959 polymorphic variant (G) was associated with reduced risk of endometrial cancer, and this reduction in risk appears to be more evident among women with low levels of endogenous estrogen exposure or with low consumption of soy food.

Grant support: National Cancer Institute/USPHS grant R01 CA92585.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. Fan Jin for contributions in implementing the study in Shanghai, Drs. Ji-Rong Long and Nobuhiko Kataoka and Regina Courtney for contributions in the genotyping, and Bethanie Hull for assistance in the preparation of this article. The study would not have been possible without the support of the study participants and research staff of the SECS.

1
Beral V, Bull D, Reeves G. Endometrial cancer and hormone-replacement therapy in the Million Women Study.
Lancet
2005
;
365
:
1543
–51.
2
Herrinton LJ, Weiss NS. Postmenopausal unopposed estrogens. Characteristics of use in relation to the risk of endometrial carcinoma.
Ann Epidemiol
1993
;
3
:
308
–18.
3
Cushing KL, Weiss NS, Voigt LF, et al. Risk of endometrial cancer in relation to use of low-dose, unopposed estrogens.
Obstet Gynecol
1998
;
91
:
35
–9.
4
Weiderpass E, Adami HO, Baron JA, et al. Risk of endometrial cancer following estrogen replacement with and without progestins.
J Natl Cancer Inst
1999
;
91
:
1131
–7.
5
Parazzini F, La VC, Bocciolone L, et al. The epidemiology of endometrial cancer.
Gynecol Oncol
1991
;
41
:
1
–16.
6
Grady D, Gebretsadik T, Kerlikowske K, et al. Hormone replacement therapy and endometrial cancer risk: a meta-analysis.
Obstet Gynecol
1995
;
85
:
304
–13.
7
Persson I, Adami H. Endometrial cancer. In: Adami I, Hunter D, Tricholpolos D, editors. Textbook of cancer epidemiology. New York: Oxford University Press; 2002.
8
Kaaks R, Lukanova A, Kurzer MS. Obesity, endogenous hormones, and endometrial cancer risk: a synthetic review.
Cancer Epidemiol Biomarkers Prev
2002
;
11
:
1531
–43.
9
Wagner JD, Anthony MS, Cline JM. Soy phytoestrogens: research on benefits and risks.
Clin Obstet Gynecol
2001
;
44
:
843
–52.
10
Messina MJ, Persky V, Setchell KD, et al. Soy intake and cancer risk: a review of the in vitro and in vivo data.
Nutr Cancer
1994
;
21
:
113
–31.
11
Adlercreutz CH, Goldin BR, Gorbach SL, et al. Soybean phytoestrogen intake and cancer risk.
J Nutr
1995
;
125
:
757
–70S.
12
Baker ME. Endocrine activity of plant-derived compounds: an evolutionary perspective.
Proc Soc Exp Biol Med
1995
;
208
:
131
–8.
13
Molteni A, Brizio-Molteni L, Persky V. In vitro hormonal effects of soybean isoflavones.
J Nutr
1995
;
125
:
751
–6S.
14
Zhu BT, Conney AH. Functional role of estrogen metabolism in target cells: review and perspectives.
Carcinogenesis
1998
;
19
:
1
–27.
15
Zhu BT, Fu JH, Xu S, et al. Different biochemical properties of nuclear and microsomal estrone-3-sulfatases: evidence for the presence of a nuclear isozyme.
Biochem Biophys Res Commun
1998
;
246
:
45
–9.
16
Newbold RR, Liehr JG. Induction of uterine adenocarcinoma in CD-1 mice by catechol estrogens.
Cancer Res
2000
;
60
:
235
–7.
17
Zhu BT, Conney AH. Is 2-methoxyestradiol an endogenous estrogen metabolite that inhibits mammary carcinogenesis?
Cancer Res
1998
;
58
:
2269
–77.
18
Mooberry SL. Mechanism of action of 2-methoxyestradiol: new developments.
Drug Resist Updat
2003
;
6
:
355
–61.
19
Bailey LR, Roodi N, Verrier CS, et al. Breast cancer and CYPIA1, GSTM1, and GSTT1 polymorphisms: evidence of a lack of association in Caucasians and African Americans.
Cancer Res
1998
;
58
:
65
–70.
20
Haiman CA, Hankinson SE, De VI, et al. Polymorphisms in steroid hormone pathway genes and mammographic density.
Breast Cancer Res Treat
2003
;
77
:
27
–36.
21
Siegelmann-Danieli N, Buetow KH. Constitutional genetic variation at the human aromatase gene (Cyp19) and breast cancer risk.
Br J Cancer
1999
;
79
:
456
–63.
22
Taioli E, Trachman J, Chen X, et al. A CYP1A1 restriction fragment length polymorphism is associated with breast cancer in African-American women.
Cancer Res
1995
;
55
:
3757
–8.
23
Tao MH, Cai Q, Xu WH, et al. Cytochrome P450 1B1 and catechol-O-methyltransferase genetic polymorphisms and endometrial cancer risk in Chinese women.
Cancer Epidemiol Biomarkers Prev
2006
;
15
:
2570
–3.
24
Doherty JA, Weiss NS, Freeman RJ, et al. Genetic factors in catechol estrogen metabolism in relation to the risk of endometrial cancer.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
357
–66.
25
Olson SH, Bandera EV, Orlow I. Variants in estrogen biosynthesis genes, sex steroid hormone levels, and endometrial cancer: a HuGE review.
Am J Epidemiol
2007
;
165
:
235
–45.
26
Duell EJ, Millikan RC, Pittman GS, et al. Polymorphisms in the DNA repair gene XRCC1 and breast cancer.
Cancer Epidemiol Biomarkers Prev
2001
;
10
:
217
–22.
27
Millikan RC, Pittman GS, Tse CK, et al. Catechol-O-methyltransferase and breast cancer risk.
Carcinogenesis
1998
;
19
:
1943
–7.
28
Millikan RC, Pittman GS, Newman B, et al. Cigarette smoking, N-acetyltransferases 1 and 2, and breast cancer risk.
Cancer Epidemiol Biomarkers Prev
1998
;
7
:
371
–8.
29
Adegoke OJ, Shu XO, Gao YT, et al. Genetic polymorphisms in uridine diphospho-glucuronosyltransferase 1A1 (UGT1A1) and risk of breast cancer.
Breast Cancer Res Treat
2004
;
85
:
239
–45.
30
Guillemette C, Millikan RC, Newman B, et al. Genetic polymorphisms in uridine diphospho-glucuronosyltransferase 1A1 and association with breast cancer among African Americans.
Cancer Res
2000
;
60
:
950
–6.
31
Guillemette C, De VI, Hankinson SE, et al. Association of genetic polymorphisms in UGT1A1 with breast cancer and plasma hormone levels.
Cancer Epidemiol Biomarkers Prev
2001
;
10
:
711
–4.
32
Shatalova EG, Walther SE, Favorova OO, et al. Genetic polymorphisms in human SULT1A1 and UGT1A1 genes associate with breast tumor characteristics: a case-series study.
Breast Cancer Res
2005
;
7
:
R909
–21.
33
Rebbeck TR, Troxel AB, Wang Y, et al. Estrogen sulfation genes, hormone replacement therapy, and endometrial cancer risk.
J Natl Cancer Inst
2006
;
98
:
1311
–20.
34
Duguay Y, McGrath M, Lepine J, et al. The functional UGT1A1 promoter polymorphism decreases endometrial cancer risk.
Cancer Res
2004
;
64
:
1202
–7.
35
Xu WH, Dai Q, Xiang YB, et al. Interaction of soy food and tea consumption with CYP19A1 genetic polymorphisms in the development of endometrial cancer.
Am J Epidemiol
2007
;
166
:
1420
–30.
36
Shu XO, Yang G, Jin F, et al. Validity and reproducibility of the food frequency questionnaire used in the Shanghai Women's Health Study.
Eur J Clin Nutr
2004
;
58
:
17
–23.
37
Shatalova EG, Loginov VI, Braga EA, et al. Association of polymorphisms in SULT1A1 and UGT1A1 genes with breast cancer risk and phenotypes in Russian women.
Mol Biol Mosk
2006
;
40
:
263
–70.
38
Chinese Academy of Medical Sciences. Food composition tables. Beijing: People's Health Publishing House; 1991.
39
Franke AA, Hankin JH, Yu MC, et al. Isoflavone levels in soy foods consumed by multiethnic populations in Singapore and Hawaii.
J Agric Food Chem
1999
;
47
:
977
–86.
40
Chen Z, Zheng W, Custer LJ, et al. Usual dietary consumption of soy foods and its correlation with the excretion rate of isoflavonoids in overnight urine samples among Chinese women in Shanghai.
Nutr Cancer
1999
;
33
:
82
–7.
41
Barrett JC, Fry B, Maller J, et al. Haploview: analysis and visualization of LD and haplotype maps.
Bioinformatics
2005
;
21
:
263
–5.
42
Lin DY, Zeng D, Millikan R. Maximum likelihood estimation of haplotype effects and haplotype-environment interactions in association studies.
Genet Epidemiol
2005
;
29
:
299
–312.
43
The comprehensive R archive network. Available from: http://cran.r-project.org/; 2006.
44
MacMahon B, Cole P, Brown JB, et al. Urine oestrogen profiles of Asian and North American women.
Int J Cancer
1974
;
14
:
161
–7.
45
Sparks R, Ulrich CM, Bigler J, et al. UDP-glucuronosyltransferase and sulfotransferase polymorphisms, sex hormone concentrations, and tumor receptor status in breast cancer patients.
Breast Cancer Res
2004
;
6
:
R488
–98.
46
Shimizu H, Ross RK, Bernstein L, et al. Serum oestrogen levels in postmenopausal women: comparison of American whites and Japanese in Japan.
Br J Cancer
1990
;
62
:
451
–3.
47
Magee PJ, Rowland IR. Phyto-oestrogens, their mechanism of action: current evidence for a role in breast and prostate cancer.
Br J Nutr
2004
;
91
:
513
–31.
48
Cline JM, Hughes CL, Jr. Phytochemicals for the prevention of breast and endometrial cancer.
Cancer Treat Res
1998
;
94
:
107
–34.
49
Burton JL, Wells M. The effect of phytoestrogens on the female genital tract.
J Clin Pathol
2002
;
55
:
401
–7.