Green tea and black tea consumption in relation to colorectal cancer risk: the Singapore Chinese Health Study

Transcription

1 Carcinogenesis vol.28 no.10 pp , 2007 doi: /carcin/bgm171 Advance Access publication August 27, 2007 Green tea and black tea consumption in relation to colorectal cancer risk: the Singapore Chinese Health Study Can-Lan Sun, Jian-Min Yuan 1, Woon Puay Koh 2, Hin-Peng Lee 2 and Mimi C.Yu 1 Division of Population Sciences, City of Hope National Medical Center and Beckman Research Institute, 1500 East Duarte Road, Duarte, CA 91010, USA, 1 Cancer Center, University of Minnesota, Minneapolis, MN 55455, USA and 2 Department of Community, Occupational and Family Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore , Singapore To whom correspondence should be addressed. Tel: þ ext ; Fax: þ ; casun@coh.org The relationships between green tea and black tea consumption and colorectal cancer risk were examined within the Singapore Chinese Health Study, a prospective cohort study of diet and cancer involving men and women. Intake of green tea and black tea was assessed through in-person interviews. Incident cancer cases and deaths among cohort members were identified through record linkage of the cohort database with respective databases from the nationwide Singapore Cancer Registry and the Singapore Registry of Births and Deaths. The proportional hazard regression method was used to examine the associations between intake of green and black tea separately and colorectal cancer risk with adjustment for potential confounders. After an average of 8.9 years of follow-up, 845 colorectal cancer cases were identified. Subjects who drank green tea exhibited a statistically non-significant increase in risk [relative risk (RR) , 95% confidence interval (CI) ] relative to non-drinkers of green tea. This risk increase was mainly confined to men (RR , 95% CI ); the comparable RR in women was 0.89 (95% CI ). In men, the green tea colorectal cancer association was noted mainly in those with advanced disease (Duke C or D) (RR , 95% CI ), and the association was dose dependent (P for trend ). This latter association was especially strong within the colon subsite (RR , 95% CI ; P for trend, ). Irrespective of gender, intake of black tea was not associated with risk of colorectal cancer (RR , 95% CI ) in this Asian population. characteristic aroma and red-orange color. An intermediate stage of enzymatic oxidation yields Oolong tea. About 78% of the tea production worldwide is black tea, which is the main tea beverage consumed in the USA and Europe. Green tea, which is the main tea beverage in Japan and parts of China, accounts for 20% of worldwide production. The remaining 2% of tea production is Oolong tea which is consumed mainly in Southern China and Taiwan. From mutagenicity to tumor development, using different carcinogens, different animal species and cell lines, most of the in vitro and non-human in vivo experiments support both green and black tea as chemopreventive agents for the development of colorectal cancer. On the other hand, the overall evidence from epidemiologic studies did not support an association between either green tea or black tea consumption and colorectal cancer risk reduction (2). Epidemiologic data on green tea are sparse, especially data from prospective studies with sufficient sample size and meaningful range of exposure. Although a greater number of epidemiologic studies have examined black tea intake and colorectal cancer, they were mainly conducted in Western populations (2). The current study utilized the Singapore Chinese Health Study, a population-based prospective cohort study of middle-age and older Chinese men and women, to examine green tea and black tea consumption in relation to colorectal cancer risk. The wide range in consumption of both green tea and black tea among Singapore Chinese provides the present study with a unique opportunity to carry out a comprehensive investigation on the independent and interactive effects of black tea and green tea as well as the overall effect of any tea intake in the development of colorectal cancer. Materials and methods Study population The subjects were participants of the Singapore Chinese Health Study, a population-based, prospective investigation of diet and cancer risk (3). The cohort is drawn from permanent residents or citizens of Singapore who reside in government housing estates (86% of Singaporeans resided in such facilities during the enrollment period). We restricted study subjects to the two major dialect groups of Chinese in Singapore: the Hokkien, who originated from southern Fujian province, and the Cantonese, who came from central Guangdong province (these are contiguous provinces in Southeastern China). The eligible age range for cohort enrollment was years. Between April 1993 and December 1998, subjects (85% of eligible subjects) were recruited. The mean age of cohort subjects at enrollment was 56.5 years. Among Introduction cohort participants, 1936 subjects were excluded because they had a baseline history of invasive cancer other than non-melanoma skin cancer. Another subject was excluded because of a history of in situ bladder cancer at baseline. Tea is a popular drink around the world, second only to water. Both green and black teas are derived from the leaves of the plant, Camellia Thus, the present study included subjects. sinensis. The fresh tea leaves contain large amounts of naturally occurring compounds called polyphenols which account for tea s pun- Baseline interview gency and unique flavor. The four primary polyphenols (also called At recruitment, each subject was interviewed in person by a trained interviewer using a structured questionnaire. The questionnaire elicited demographic information, lifetime tobacco use, current physical activity profile, occupational catechins) in fresh tea leaves are epigallocatechin gallate (EGCG), epigallocatechin (EGC), epicatechin gallate (ECG) and epicatechin, exposure, medical history, family history of cancer and menstrual and reproductive history (women only). Dietary information (frequency and portion with EGCG being most abundant. These four catechins account for up to 30% dry weight of the fresh tea leaves (1). Varying methods of size) during the past 12 months also was obtained using a 165-item food processing the tea leaves after harvest lead to three major types of tea. frequency questionnaire (FFQ). The FFQ was subsequently validated against To make green tea, fresh tea leaves are steamed or pan dried at high a series of 24-h dietary recalls (3). Daily intakes of 100 nutritive/nonnutritive dietary components were computed via linkage of the FFQ response temperature right after plucking, resulting in minimal oxidation of the catechins in the tea leaves. Alternatively, fresh tea leaves are rolled or to the Singapore Food Composition Table that we developed in conjunction crushed to encourage oxidation and polymerization of the catechins in with the Singapore Cohort Study (3). For green tea and black tea separately, study subjects were asked to choose a process commonly known as fermentation to make black tea. This their intake frequency during the past 12 months from nine predefined categories: never or hardly ever, one to three times a month, once a week, two to fermentation process results in the generation of other distinct polyphenols such as thearubigins and theaflavins that give black tea the three times a week, four to six times a week, once a day, two to three times a day, four to five times a day and six or more times a day. Non-green tea Abbreviations: CI, confidence interval; ECG, epicatechin gallate; EGCG, drinkers were subjects who reported never or hardly ever in response to the epigallocatechin gallate; FFQ, food frequency questionnaire; RR, relative risk. question regarding green tea intake frequency. The remaining green tea Ó The Author Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org 2143

2 C.-L. Sun et al. Table I. Distribution of selected characteristics among non-tea drinkers, green tea and black tea drinkers in women, the Singapore Chinese Health Study, Non-tea drinkers (%) Green tea drinkers (%) Black tea drinkers (%) Monthly Weekly Daily Monthly Weekly Daily No. of subjects Dialect group Cantonese Hokkien Education No formal education Primary school Secondary or higher Smoking status Never Light a Heavy a Alcohol drinking Non-drinker ,7 drinks/week þ drinks/week Coffee drinking Never/monthly/weekly cup/day þ cups/day Moderate activity No Yes b Family history of colorectal cancer No Yes Diabetes No Yes Body mass index (kg/m 2 ), Mean ± SD 23 ± 3 23 ± 3 23 ± 3 24 ± 3 23 ± 3 23 ± 3 23 ± 3 Red meat (g/day), Mean ± SD 24 ± ± ± ± ± ± ± 22 Fruits (g/day), Mean ± SD 168 ± ± ± ± ± ± ± 198 Vegetables (g/day), Mean ± SD 100 ± ± ± ± ± ± ± 72 a Heavy smokers were those who began smoking before age 15 and smoked 13 cigarettes per day. Light smokers were all other smokers. b Moderate activities included brisk walking, bowling, bicycling on level ground, tai chi or chi kung. drinkers were grouped into three frequency levels (monthly, weekly, daily). Non-drinkers and drinkers of black tea were similarly defined. In terms of total tea intake (i.e. tea intake irrespective of type), subjects who never or hardly ever drank either type of tea were labeled as non-tea drinkers. Subjects who reported drinking either green tea or black tea daily were categorized as daily tea drinkers. For the remaining subjects, those reporting either green tea or black tea intake on a weekly basis were categorized as weekly tea drinkers. All remaining subjects were categorized as monthly tea drinkers. Follow-up Identification of incident colorectal cancer cases and deaths among cohort members were accomplished by record linkage of the cohort database with respective databases from the population-based Singapore Cancer Registry and the Singapore Registry of Births and Deaths. The nationwide cancer registry has been in place since 1968 and has been shown to be comprehensive in its recording of cancer cases (4,5). To date, only 17 subjects were lost to follow-up due to migration out of Singapore. As of 31 December 2004 [an average of 8.9 (median 5 9.6) years of follow-up], 845 cohort participants who were cancer free at baseline had developed invasive colorectal cancer (516 colon cancers and 329 rectal/rectosigmoid cancers). Histological information on 821 (97.2%) colorectal cancer cases was confirmed via manual review of pathology reports. The cancer diagnoses of 14 cases (1.6%) were confirmed via review of medical records by a medically trained research staff. The remaining 10 cases (1.2%) were identified through death certificates. Statistical analyses For each study subject, person-years were counted from the date of baseline interview to the date of cancer diagnosis, date of death, date of last contact (for those 17 subjects who migrated out of Singapore) or 31 December 2004, whichever occurred first. Proportional hazards regression methods (6) were used to examine the tea and colorectal cancer association with adjustment 2144 for potential confounders. The strength of the association was measured by relative risks (RRs) and their corresponding 95% confidence intervals (CIs) and two-sided P values. The linear trend tests were based on the ordinal values of the tea intake frequency categories (0, 1, 2 and 3). A two-sided P value,0.05 would be considered statistically significant. Q statistics (7) were used to test for homogeneity of the RRs between the colon versus the rectum subsite. A two-sided P value,0.10 would be considered statistically significant. In all analyses, we considered the following factors as potential confounders: gender (in analyses involving both genders), age at baseline interview (years), year of interview ( ), dialect group (Cantonese, Hokkien), level of education (no formal education, primary school, secondary school or higher), cigarette smoking (never, light, heavy; see definition below) (8), alcohol consumption (non-drinker, less than seven drinks per week, seven or more drinks per week) (8), coffee drinking (never/occasionally, one cup per day, two or more cups per day), body mass index (,20, 20 to,24, 24 to,28, 28þ kg/m 2 ), familial history of colorectal cancer (no, yes), history of diabetes (no, yes) (9), weekly moderate physical activity (no, yes; see definition below), intakes of total calories (Kcal), total fat (% of total calories), dietary fiber (mg/1000 Kcal), calcium (mg/1000 Kcal) and vitamin C (mg/1000 Kcal). For cigarette smoking, heavy smokers were defined as those who started to smoke before 15 years of age and smoked 13 or more cigarettes per day. The remaining smokers were categorized as light smokers. Moderate physical activities included brisk walking, bowling, bicycling on level ground, tai chi or chi kung. Subjects who engaged in such activities at least 30 min/week were classified into the weekly moderate physical activity category. Rate of replacement hormone use in women in our cohort was low (6%), as were use of vitamin supplements (6%) and use of non-steroidal antiinflammatory drugs (1%) irrespective of gender. Additional inclusion of these covariates did not materially change any of the study results. Similarly, results with or without the inclusion of intakes of fruit, vegetables and red meats as

3 Relationships between green tea and black tea consumption and colorectal cancer risk Table II. Distribution of selected characteristics among non-tea drinkers, green tea and black tea drinkers in men, the Singapore Chinese Health Study, Non-tea drinkers (%) Green tea drinkers (%) Black tea drinkers (%) Monthly Weekly Daily Monthly Weekly Daily No. of subjects Dialect group Cantonese Hokkien Education No formal education Primary school Secondary or higher Smoking status Never Light a Heavy a Alcohol drinking Non-drinker ,7 drinks/week þ drinks/week Coffee drinking Never/monthly/weekly /day þ/day Moderate activity No Yes b Family history of colorectal cancer No Yes Diabetes No Yes Body mass index (kg/m 2 ), Mean ± SD 23 ± 3 23 ± 3 23 ± 3 24 ± 3 23 ± 3 23 ± 3 23 ± 3 Red meat (g/day), Mean ± SD 33 ± ± ± ± ± ± ± 29 Fruits (g/day), Mean ± SD 183 ± ± ± ± ± ± ± 200 Vegetables (g/day), Mean ± SD 101 ± ± ± ± ± ± ± 70 a Heavy smokers were those who began smoking before age 15 and smoked 13 cigarettes per day. Light smokers were all other smokers. b Moderate activities included brisk walking, bowling, bicycling on level ground, tai chi or chi kung. covariates were comparable. The presented results are derived from regression models that did not include these covariates. Statistical computing was conducted using SAS version 9.1 (SAS Institute, Cary, NC) and STATA statistical software (College Station, TX). The study protocol was approved by the Institutional Review Boards of the University of Minnesota and the National University of Singapore. Results The distributions of selected demographic and lifestyle factors by green tea and black tea consumption levels in men and women are presented in Tables I and II. Men were more likely to drink either green tea or black tea compared with women. In both men and women, Cantonese were more likely to be green tea drinkers whereas Hokkien were more likely to be black tea drinkers. Tea drinkers (regardless of type) had higher levels of education, were less likely to smoke but more likely to drink alcoholic beverages, less likely to drink coffee and were more likely to engage in moderate physical activities. Among men, more green tea drinkers than non-tea drinkers reported a history of diabetes. As of 31 December 2004, 845 incident cases of invasive colorectal cancer cases (470 males and 375 females) were identified after an average of 8.9 years of follow-up (SD 5 2.3). There were 516 colon cancer cases and 329 rectal cancer cases. Majority of the cancers were adenocarcinoma (n 5 747, 88%). The mean age at cancer diagnosis was 67 years (SD 5 8.0). The mean interval between entry into the study and cancer diagnosis was 5.5 years (range: 1 month to 11.6 years). Three hundred and twenty-two cases (38.1%) were classified as localized (Duke A or B) and 472 (55.9%) as advanced disease (Duke C or D). Fifty-one cases (6%) could not be staged due to insufficient clinical and histopathological information. Total tea consumption was not associated with colorectal cancer risk (RR , 95% CI ) (Table III). The null association did not differ by subsites; the RR (95% CI) for colon cancer was 1.04 ( ) and for rectal cancer, 1.04 ( ) (P for heterogeneity between two sites ). We then examined tea intake by type (green versus black tea). Neither green tea consumption (RR , 95% CI ) nor black tea consumption (RR , 95% CI ) was associated with colorectal cancer risk, and these null effects were seen in both the colon and rectum subsites (data not shown). However, when tea intake in relation to colorectal cancer risk was examined separately in men and women, men who reported drinking green tea monthly or more frequently showed a statistically significant 30% increase in colorectal cancer risk (RR , 95% CI ). No such association was noted in women (RR , 95% CI ). Black tea consumption was not associated with colorectal cancer risk in either men (RR , 95% CI ) or women (RR , 95% CI ) (Table III). We further examined the tea colorectal cancer association in men according to subsite (colon versus rectum) and stage of disease (localized versus advanced) (Table IV). There was no evidence that risk of localized disease was associated with green tea intake (RR , 95% CI ). On the other hand, there was a highly statistically significant, dose-dependent positive association between green tea intake and risk of advanced colorectal cancer (RR for monthly, weekly and daily drinkers were 1.38, 1.35 and 1.85, respectively, 2145

4 C.-L. Sun et al. Table III. Green tea and black tea consumption in relation to colorectal cancer risk, the Singapore Chinese Health Study, Person-years Total Male Female Cases RR (95% CI) Cases RR (95% CI) Cases RR (95% CI) Any tea Non-drinker Drinker ( ) ( ) ( ) Monthly ( ) ( ) ( ) Weekly ( ) ( ) ( ) Daily ( ) ( ) ( ) P trend 0.51 Green tea Non-drinker Drinker ( ) ( ) ( ) Monthly ( ) ( ) ( ) Weekly ( ) ( ) ( ) Daily ( ) ( ) ( ) P trend Black tea Non-drinker Drinker ( ) ( ) ( ) Monthly ( ) ( ) ( ) Weekly ( ) ( ) ( ) Daily ( ) ( ) ( ) P trend RRs were adjusted for sex (for total), age at baseline interview (years), year of interview ( ), dialect group (Cantonese, Hokkien), education (no formal education, primary school, secondary or higher), family history of colorectal cancer (yes, no), history of diabetes (yes, no), cigarette smoking (never, light, heavy), alcohol drinking (non-drinker, less than seven drinks per week, seven or more drinks per week), coffee drinking (never/occasionally, one cup per day, two or more cups per day), weekly moderate physical activity (no, yes), body mass index (,20, 20 to,24, 24 to,28, 28þ kg/m 2 ), total energy (Kcal), total fat (% Kcal), dietary fiber (mg/1000 Kcal), calcium (mg/1000 Kcal) and vitamin C (mg/1000 Kcal). For green tea, black tea intake (cups per month) was included in the model. Conversely, for black tea, green tea intake (cups per month) was included in the model. Table IV. Green tea and black tea consumption in relation to colorectal cancer risk in men, by subsite and stage of disease, the Singapore Chinese Health Study, Colorectal Colon Rectum Localized Advanced Localized Advanced Localized Advanced N RR (95% CI) N RR (95% CI) N RR (95% CI) N RR (95% CI) N RR (95% CI) N RR (95% CI) Green tea intake Non-drinker Drinker ( ) ( ) ( ) ( ) ( ) ( ) Monthly ( ) ( ) ( ) ( ) ( ) ( ) Weekly ( ) ( ) ( ) ( ) ( ) ( ) Daily ( ) ( ) ( ) ( ) ( ) ( ) P trend , Black tea intake Non-drinker Drinker ( ) ( ) ( ) ( ) ( ) ( ) Monthly ( ) ( ) ( ) ( ) ( ) ( ) Weekly ( ) ( ) ( ) ( ) ( ) ( ) Daily ( ) ( ) ( ) ( ) ( ) ( ) P trend N, number of cancer cases. RRs were adjusted for age at baseline interview (years), year of interview ( ), dialect group (Cantonese, Hokkien), education (no formal education, primary school, secondary or higher), family history of colorectal cancer (yes, no), history of diabetes (yes, no), cigarette smoking (never, light, heavy), alcohol drinking (non-drinker, less than seven drinks per week, seven or more drinks per week), coffee drinking (never/occasionally, one cup per day, two or more cups per day), weekly moderate physical activity (no, yes), body mass index (,20, 20 to,24, 24 to,28, 28þ kg/m 2 ), total energy (Kcal), total fat (% Kcal), dietary fiber (mg/1000 Kcal), calcium (mg/1000 Kcal) and vitamin C (mg/1000 Kcal), as well as black tea intake (cups per month). P trend ). This latter association was primarily present in the colon subsite (RR for monthly, weekly and daily drinkers were 1.11, 1.69 and 2.26, respectively, P trend, ). There were no associations between black tea consumption and colorectal cancer in men by subsite or stage of disease. We also examined the effect of duration of follow-up on the green tea colorectal cancer association in men. Among men with 5 or fewer 2146 years of follow-up, the RR for green tea drinkers compared with nongreen tea drinkers was 1.39 (95% CI ), similar to those with 6 10 years of follow-up (RR , 95% CI ). Among these men with relatively short duration (under 10 years) of follow-up, the associations with green tea intake was primarily confined to those with advanced stage disease (RR , 95% CI for those with 5 years of follow-up and RR ,

5 Relationships between green tea and black tea consumption and colorectal cancer risk Table V. Green tea in relation to colorectal cancer risk in men by length of follow-up, the Singapore Chinese Health Study, to 5 years 6 to 10 years.10 years No. of non-drinkers/no. of drinkers RR (95% CI) No. of non-drinkers/no. drinkers RR (95% CI) No. of non-drinkers/no. of drinkers RR (95% CI) Overall men 92/ ( ) 110/ ( ) 24/ ( ) Men advanced disease only 41/ ( ) 62/ ( ) 15/ ( ) Colon 24/ ( ) 31/ ( ) 7/ ( ) Rectum 17/ ( ) 31/ ( ) 8/ ( ) Men localized disease only 46/ ( ) 40/ ( ) 6/ ( ) RRs were adjusted for age at baseline interview (years), year of interview ( ), dialect group (Cantonese, Hokkien), education (no formal education, primary school, secondary or higher), family history of colorectal cancer (yes, no), history of diabetes (yes, no), cigarette smoking (never, light, heavy), alcohol drinking (non-drinker, less than seven drinks per week, seven or more drinks per week), coffee drinking (never/occasionally, one cup per day, two or more cups per day), weekly moderate physical activity (no, yes), body mass index (,20, 20 to,24, 24 to,28, 28þ kg/m 2 ), total energy (Kcal), total fat (% Kcal), dietary fiber (mg/1000 Kcal), calcium (mg/1000 Kcal) and vitamin C (mg/1000 Kcal), as well as black tea intake (cups per month). 95% CI for those with 6 10 years of follow-up). In contrast, men with.10 years of follow-up exhibited a reduced risk of colorectal cancer (RR , 95% CI ) (Table V). Discussion Using data from the Singapore Chinese Health Study, we demonstrated that green tea but not black tea consumption was associated with an increased risk of colorectal cancer in men. This positive green tea/colorectal cancer risk association was mainly confined to patients with advanced colorectal cancer (Duke C or D) and was more evident for the colon subsite. No association between colorectal cancer and either green or black tea was observed in women. Although experimental studies have supported the notion of a protective effect of green tea or tea catechins on colorectal carcinogenesis (10 17), epidemiologic data on this association in humans are still limited. Only eight studies have examined green tea and colorectal cancer; six were conducted in Japan where few people were non-daily drinkers of green tea (2). All four cohort studies were conducted in Japan. In our recent meta-analysis (2), the summary RR for colorectal cancer in green tea drinkers, based on the four cohort studies, was 0.97 (95% CI ). Our finding of a RR of 1.12 (95% CI ) in total subjects is compatible with the meta-analysis results (P for heterogeneity ). Two of the four previously published cohort studies provided gender-specific estimates (pooled summary RR , 95% CI for men). Again, our finding in men was compatible with the published literature (P for heterogeneity ). Recently, we examined and reported on the relationship between levels of EGC and 4#-O-methyl-EGC in urine samples collected at baseline and subsequent risk of colorectal cancer within the Shanghai Cohort Study, a prospective investigation of men aged years followed since 1986 (18). Due to the repeated thawing of urine specimens in early cases, only colorectal cancer patients with 5 years of follow-up were included in this study. We observed a statistically significant, dose-dependent inverse association between colorectal cancer risk and baseline urinary EGC and 4#-O-methyl-EGC. Consistent with our earlier observation in Shanghai, the present study noted a 50% reduction in risk among men with.10 years of followup. Due to the small number of subjects in this subgroup, the reduced risk estimate was not statistically different from the null hypothesis. Interestingly, the present study suggests that green tea may contain substances that exert a late-stage effect on colorectal cancer in men, primarily impacting on tumor progression and metastasis. We noted a dose-dependent, statistically significant association between green tea intake and advanced colorectal cancer (Duke C or D) that is only evident among subjects with,10 years of follow-up. There is limited experimental support for a promotional effect of green tea on colorectal cancer progression/invasion. In one study in which green tea was administered after carcinogen exposure, green tea-treated rats experienced increased colon carcinoma formation (19). In another study, EGCG increased the intra- and extracellular production of pro-matrix-metalloproteinase (MMP)-7 protein in a dose- and timedependent manner (20). MMP-7 messenger RNA expression has been shown to correlate with colorectal cancer progression (21). Finally, two recent studies show that ECG and EGCG inhibit hypoxia-inducible factor-1a degradation in breast (22) and prostate cancer cell lines (23). It is known that hypoxia develops during tumor progression, including colorectal cancer (24), due to the inability of existing vascular system to supply the growing tumor with adequate amounts of oxygen (25). Over-expression of hypoxia-inducible factor-1a has been observed in primary and metastatic tumors of common human cancers, including colon cancer (26). Recent findings from the Women s Health Initiative Trial (27,28) corroborated reports of observational studies (29,30) in support of progestin, and possibly other female sex hormones, having a direct, beneficial effect on colorectal carcinogenesis. Recently, we noted that among post-menopausal female participants of this Singapore Chinese cohort, green tea drinkers exhibited statistically significant, reduced levels of plasma estrogens relative to non-tea drinkers (31). Thus, a sex hormone-mediated pathway may be involved in the observed positive association between green tea intake and late-stage colorectal cancer, which seems to be restricted to the male gender. This is the first prospective study of black tea and colorectal cancer risk in an Asian population. Consistent with the literature, we failed to note any relationship between black tea intake and colorectal cancer. Although the overall levels of polyphenols are similar between green tea and black tea (32), the contents of catechins are vastly different between the two types of tea. The main polyphenols in black tea are thearubigins and theaflavins; the catechins in black tea are only onethird to one-tenth the comparable levels in green tea (33). Therefore, it is not surprising that black tea fails to exhibit a similar, positive association with colorectal cancer risk as green tea. While we cannot totally rule out the possibility of chance finding, we have, however, adjusted for all possible confounders in our statistical analysis. The positive green tea colorectal cancer association in men persisted after adjustment for cigarette smoking, alcohol consumption and history of diabetes. There is some suggestion that coffee intake might be negatively associated with colorectal cancer risk (34,35). Further adjustment for coffee consumption did not materially change the association. This study has several strengths. To our knowledge, there is no widespread, organized colorectal cancer screening in Singapore, in contrast to several government-subsidized, high-profiled, nationwide mammography and pap smear screening programs among ageeligible citizens in the past 5 and 3 years, respectively. Therefore, the potential of screening-identified cases affecting the validity of the observed associations is quite unlikely. Dietary assessment was 2147

6 C.-L. Sun et al. conducted using a validated, semi-quantitative FFQ. The study population is genetically homogeneous since they are full-blooded descendents of natives from two contiguous prefectures in southern China. Finally, all exposure assessments took place prior to cancer diagnosis, and therefore, could be presumed to be free of recall bias. In conclusion, we observed a novel, positive association between green tea consumption and risk of advanced colorectal cancer in men. On the other hand, for the subgroup of subjects with long duration of follow-up (.10 years), the results are compatible with our earlier findings of a beneficial effect of green tea intake on colorectal cancer risk. Thus, this study suggests a dual (i.e. deleterious versus beneficial) role of green tea on colorectal carcinogenesis that is dependent on the time of exposure (early versus late). Our findings require confirmation in other population-based cohorts and experimental studies are needed to elucidate the mechanisms behind this apparent latestage, progression-/invasion-related effect of green tea on colorectal carcinogenesis. At present, there should be caution in recommending green tea as a health-promoting compound for men. Funding National Cancer Institute, Bethesda, MD (R01 CA55069, R35 CA53890, R01 CA80205 and R01 CA98497). Acknowledgements We thank Ms Siew-Hong Low of the National University of Singapore for supervising the fieldwork of the Singapore Chinese Health Study and Ms Kazuko Arakawa of the University of Southern California for the development and management of the cohort study database. We also thank the Singapore Cancer Registry for assistance with the identification of cancer outcomes. Conflict of Interest Statement: None declared. References 1.Graham,H.N. (1992) Green tea composition, consumption, and polyphenol chemistry. Prev. Med., 21, Sun,C.L. et al. (2006) Green tea, black tea and colorectal cancer risk: a meta-analysis of epidemiologic studies. Carcinogenesis, 27, Hankin,J.H. et al. (2001) Singapore Chinese Health Study: development, validation, and calibration of the quantitative food frequency questionnaire. Nutr. Cancer, 39, Waterhouse, J. et al. (eds) (1976) Cancer Incidence in Five Continents. International Agency for Research on Cancer, Lyon. 5. Parkin, D.M. et al. (eds) (2003) Cancer Incidence in Five Continents. International Agency for Research on Cancer, Lyon. 6.Cox,D. (1972) Regression models and life tables. J. Roy. Stat. Soc. (B), 34, DerSimonian,R. et al. (1986) Meta-analysis in clinical trials. Control. Clin. Trials, 7, Tsong,W.H. et al. (2007) Cigarettes and alcohol in relation to colorectal cancer: the Singapore Chinese Health Study. Br. J. Cancer, 96, Seow,A. et al. (2006) Diabetes mellitus and risk of colorectal cancer in the Singapore Chinese Health Study. J. Natl Cancer Inst., 98, Apostolides,Z. et al. (1996) Inhibition of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) mutagenicity by black and green tea extracts and polyphenols. Mutat. Res., 359, Weisburger,J.H. et al. (1996) Tea polyphenols as inhibitors of mutagenicity of major classes of carcinogens. Mutat. Res., 371, Hernaez,J.F. et al. (1998) Antimutagenic activity of tea towards 2-hydroxyamino-3-methylimidazo[4,5-f]quinoline: effect of tea concentration and brew time on electrophile scavenging. Mutat. Res., 402, Huber,W.W. et al. (1997) Chemoprotection against the formation of colon DNA adducts from the food-borne carcinogen 2-amino-1-methyl-6- phenylimidazo[4,5-b]pyridine (PhIP) in the rat. Mutat. Res., 376, Schut,H.A. et al. (2000) Tea as a potential chemopreventive agent in PhIP carcinogenesis: effects of green tea and black tea on PhIP-DNA adduct formation in female F-344 rats. Nutr. Cancer, 36, Jia,X. et al. (2001) Effects of green tea on colonic aberrant crypt foci and proliferative indexes in rats. Nutr. Cancer, 39, Metz,N. et al. (2000) Suppression of azoxymethane-induced preneoplastic lesions and inhibition of cyclooxygenase-2 activity in the colonic mucosa of rats drinking a crude green tea extract. Nutr. Cancer, 38, Xu,M. et al. (1996) Protection by green tea, black tea, and indole-3- carbinol against 2-amino-3-methylimidazo[4,5-f]quinoline-induced DNA adducts and colonic aberrant crypts in the F344 rat. Carcinogenesis, 17, Yuan,J.M. et al. (2006) Urinary biomarkers of tea polyphenols and risk of colorectal cancer in the Shanghai Cohort Study. Int. J. Cancer, 120, Hirose,M. et al. (2001) Green tea catechins enhance tumor development in the colon without effects in the lung or thyroid after pretreatment with 1,2-dimethylhydrazine or 2,2#-dihydroxy-di-n-propylnitrosamine in male F344 rats. Cancer Lett., 168, Kim,M. et al. (2005) ( )-Epigallocatechin-3-gallate promotes pro-matrix metalloproteinase-7 production via activation of the JNK1/2 pathway in HT-29 human colorectal cancer cells. Carcinogenesis, 26, Ishikawa,T. et al. (1996) Matrilysin is associated with progression of colorectal tumor. Cancer Lett., 107, Zhou,Y.-D. et al. (2004) Hypoxia-inducible factor-1 activation by (-)- epicatechin gallate: potential adverse effects of cancer chemoprevention with high-dose green tea extracts. J. Nat. Prod., 67, Thomas,R. et al. (2005) Epigallocatechin gallate inhibits HIF-1alpha degradation in prostate cancer cells. Biochem. Biophys. Res. Commun., 334, Goethals,L. et al. (2006) Hypoxia in human colorectal adenocarcinoma: comparison between extrinsic and potential intrinsic hypoxia markers. Int. J. Radiat. Oncol. Biol. Phys., 65, Hockel,M. et al. (2001) Tumor hypoxia: definitions and current clinical, biologic, and molecular aspects. J. Natl Cancer Inst., 93, Zhong,H. et al. (1999) Overexpression of hypoxia-inducible factor 1alpha in common human cancers and their metastases. Cancer Res., 59, Anderson,G.L. et al. (2004) Effects of conjugated equine estrogen in postmenopausal women with hysterectomy: the Women s Health Initiative randomized controlled trial. JAMA, 291, Chlebowski,R.T. et al. (2004) Estrogen plus progestin and colorectal cancer in postmenopausal women. N. Engl. J. Med., 350, Grodstein,F. et al. (1999) An incentive to start hormone replacement: the effect of postmenopausal hormone replacement therapy on the risk of colorectal cancer. Am. J. Med., 106, Hebert-Croteau,N. (1998) A meta-analysis of hormone replacement therapy and colon cancer in women. Cancer Epidemiol. Biomarkers Prev., 7, Wu,A.H. et al. (2005) Tea and circulating estrogen levels in postmenopausal Chinese women in Singapore. Carcinogenesis, 26, Wang,Z.Y. et al. (1994) Inhibitory effects of black tea, green tea, decaffeinated black tea, and decaffeinated green tea on ultraviolet B light-induced skin carcinogenesis in 7,12-dimethylbenz[a]anthraceneinitiated SKH-1 mice. Cancer Res., 54, Balentine,D.A. et al. (1997) The chemistry of tea flavonoids. Crit. Rev. Food Sci. Nutr., 37, Michels,K.B. et al. (2005) Coffee, tea, and coffeine consumption and incidence of colon and rectal cancer. J. Natl Cancer Inst., 97, Tavani,A. et al. (2004) Coffee, decaffeinated coffee, tea and cancer of the colon and rectum: a review of epidemiological studies, Cancer Causes Control, 15, Received March 16, 2007; revised July 9, 2007; accepted July 19,