In this population-based case-control study, we found that dietary consumption of soy foods and tea was inversely associated with endometrial cancer risk. Three polymorphisms (rs1065779, rs752760, and rs1870050), in both single marker and haplotype analyses, were found to be related to the risk of endometrial cancer and their effects were modified by tea consumption.
Our findings of an inverse association of endometrial cancer with intake of soy food and tea are consistent with previous studies (12
), including our earlier reports from subsets of the SECS (12
). Soy foods are rich in isoflavones, which have estrogenic or anti-estrogenic properties due to the similarity of their structure to estradiol (7
). It has been suggested that these phytoestrogens, as well as tea polyphenols, decrease estrogen biosynthesis and produce anti-estrogenic effects through lowering the activity of aromatase (16
). In addition to this anti-estrogenic effect, polyphenols also have other anti-carcinogenic properties, such as an antioxidant effect (31
Aromatase is a crucial enzyme catalyzing three consecutive hydroxylation reactions converting C19 androgens to aromatic C18 estrogenic steroids, namely, androstenedione and testosterone to estrone and estradiol, respectively. The CYP19A1 gene codes aromatase and its transcription is regulated by tissue-specific promoters (32
). The SNPs rs1870050 and rs752760 are located in the first exon, close to promoter I.1, the major promoter for the placenta. The SNP rs28566535 is located in the first exon close to promoter I.4, which is the promoter for adipose tissue, bone, and skin. The SNP rs1065779 is located in intron 9, 53 bp upstream of exon 10, which may affect transcription or expression of aromatase. The SNP rs700519 (Arg264Cys) is located in exon 7. By causing an amino acid change from Arg to Cys, this SNP may result in a change of enzyme activity. Previous studies have found that variants in the CYP19A1 gene lead to altered aromatase activity and estrogen levels (18
), which are related to risk of hormone-related cancers (34
), including endometrial cancer (21
). We also found that several polymorphisms, SNPs rs1065779, rs752760 and rs1870050, were associated with risk of endometrial cancer (23
), suggesting that these polymorphisms may play a critical role in the development of endometrial cancer.
In this study, we observed significant interactions between three CYP19A1
genetic polymorphisms and tea consumption. Although these findings are new, they are biologically possible. It has been shown that green tea catechins and black tea polyphenol theaflavins decrease aromatase enzyme activity (17
). Our finding that the associations of rs1065779, rs752760 and rs1870050 with endometrial cancer are more pronounced in tea drinkers is in agreement with these results. It is possible that SNP rs1065779 and rs752760, two polymorphisms close to the promoter for the placenta, may play an important role in the pathogenesis of endometrial cancer. SNP rs1065779, through its effect on transcription or the expression of aromatase, may also be involved in the carcinogenesis of endometrial cancer. The tea and single SNP interactions were also confirmed in the haplotype analyses. Of note, these interactions were most evident under the recessive model. Given that the functionality of SNPs and associated haplotypes are not well understood and that the genetic regulation of aromatase has not been extensively studied in endometrial carcinoma, further studies with a more comprehensive SNP coverage of the CYP19A1
gene are needed to evaluate our findings.
We did not observe a significant interaction between intake of soy foods and any of the five polymorphisms in the CYP19A1
gene. These results are not surprising because previous studies have relatively consistently found that the inhibitory activity of isoflavones on aromatase, if any (30
), was much weaker than that of other polyphenols (30
). On the other hand, in vitro
studies have found that soy isoflavones can inhibit the activity of 17β-hydroxysteroid dehydrogenase type I (17
), a key enzyme in catalyzing estrone (E1) to the biologically more active estradiol (E2) (2
). We have previously reported from the same study that soy consumption may interact with polymorphisms in the 17
gene in relation to endometrial cancer risk (43
). Taken together, these findings suggest that isoflavones and tea polyphenols may modify the associations of genetic polymorphisms in different estrogen-related genes with endometrial cancer risk.
As with all case-control studies, the possibility for recall bias could not be completely eliminated. We tried to minimize recall bias by shortening the interval between diagnosis and interview for cases and by asking participants to ignore any dietary change over the preceding year. Furthermore, it is unlikely that recall bias would depend on an individual’s genotype. In this study, we collected information on tea consumption according to the amount of tea leaves consumed because Chinese people drink tea by putting loose tea leaves in a cup and repeatedly adding water using the same tea leaves. Variability in the amount of water added to the tea may have introduced measurement errors with regard to tea consumption. Likewise, soy intake was also subject to measurement errors. Misclassification of tea and soy consumption, which is most likely to be non-differential, may bias the results towards the null. Although we adjusted for many potential confounding factors, we still cannot exclude the possibility that residual confounding or related dietary patterns may partially explain our results. Finally, we only included five polymorphisms in this study. These SNPs were chosen based on literature review of published data and the potential functionality of the SNPs. Thus, our study is not as comprehensive as other studies which have applied the haplotype tagging SNP approach (44
). Therefore, we cannot exclude the possibility that there may be other polymorphisms that interact with dietary polyphenols and that these unstudied polymorphisms may be responsible for the gene-tea interaction observed in this study.
Strengths of this study included the population-based study design, high response rate, high DNA sample donation rate, and the low frequency of hysterectomy in the population (3.6 percent), which minimized the selection bias. The homogeneous ethnic background (>98 percent Han Chinese) of the study participants avoided potential confounding from ethnicity. The large sample size enabled us to explore potential gene-environment interactions.
In summary, this population-based study suggests that consumption of soy food and tea was related to a reduced risk of endometrial cancer. Tea consumption may interact with genotypes of the CYP19A1 gene in the etiology of endometrial cancer.