Coffee consumption and risk of colorectal cancer in a population-based prospective cohort of Japanese men and women

Authors

  • Kyung-Jae Lee,

    1. Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan
    2. Department of Preventive Medicine, Soonchunhyang University College of Medicine, Seoul, Korea
    Search for more papers by this author
  • Manami Inoue,

    Corresponding author
    1. Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan
    • Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan
    Search for more papers by this author
    • Fax: +81-3-3547-8578.

  • Tetsuya Otani,

    1. Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan
    Search for more papers by this author
  • Motoki Iwasaki,

    1. Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan
    Search for more papers by this author
  • Shizuka Sasazuki,

    1. Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan
    Search for more papers by this author
  • Shoichiro Tsugane,

    1. Epidemiology and Prevention Division, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo, Japan
    Search for more papers by this author
  • for the JPHC Study Group

    1. Department of Preventive Medicine, Soonchunhyang University College of Medicine, Seoul, Korea
    Search for more papers by this author

Abstract

We prospectively examined the association between coffee consumption and the risk of developing colorectal cancer in a large population-based cohort study (the JPHC Study) of Japanese men and women. Data were analyzed from a population-based cohort of 96,162 subjects (46,023 men and 50,139 women). A total of 1,163 incident colorectal cancers were identified during the follow-up period, including 763 cases of colon cancer and 400 of rectal cancer. We observed a significant inverse association between coffee consumption and the risk of developing invasive colon cancer among women. Compared with those who almost never consumed coffee, women who regularly consumed 3 or more cups of coffee per day had a RR of 0.44 (95% CI = 0.19–1.04; p for trend = 0.04) after adjustment for potential confounding factors. However, no significant association was found for rectal cancer in women. In men, no significant decrease was observed in any colorectal cancer site. Further, additional analyses on the association of green tea consumption with colorectal cancer risk found no significant association in men or women. These findings suggest that coffee consumption may lower the risk of colon cancer among Japanese women. © 2007 Wiley-Liss, Inc.

Colorectal cancer is one of the most common cancers in more developed countries, with nearly 945,000 new cases diagnosed worldwide each year and 492,000 deaths.1 In Japan, the incidence and mortality of colorectal cancer has recently increased, and it is now a leading cause of morbidity and mortality, with about 36,000 deaths in 2000.2 This recent increase in Japan may be associated with changes in environmental factors such as dietary habits or lifestyle factors.

Coffee drinking is popular in many parts of the world. In Japan, coffee has been popularly consumed since its initial introduction, although coffee consumption is relatively new in comparison to green tea consumption. Given the widespread consumption of coffee and high incidence of colorectal cancer in developed countries, the possible role of coffee consumption in the etiology of colorectal carcinogenesis has drawn interest,3 and the relation between coffee consumption and colorectal cancer risk has been investigated over the last 4 decades. However, results from epidemiological studies have been inconsistent.3, 4 A meta-analysis showed a lower risk of colorectal cancer associated with substantial consumption of coffee,3 but the results were inconclusive owing to inconsistencies between case–control and prospective studies. Case–control studies have tended to support an inverse association between coffee consumption and risk of colorectal cancer,5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 whereas the albeit limited number of prospective cohort studies have not supported any significant reduction in risk.16, 17, 18, 19, 20, 21, 22, 23, 24, 25 Moreover, most previous studies have been conducted in Western countries.16, 17, 18, 19, 20, 21, 22, 23, 24

Here, we prospectively examined the association between coffee consumption and risk of developing colorectal cancer in a large population-based prospective study of Japanese men and women.

Material and methods

Study population

The basis for this investigation lies with the Japan Public Health Center-based Prospective Study (JPHC Study), which was established in 1990–1994. This study comprised 11 prefectural public health center (PHC) areas and included questions on physical activity. Details of the study design have been provided elsewhere.26 The study protocol was approved by the institutional review board of the National Cancer Center, Japan. For the present analysis, the Katsushika PHC area (metropolitan Tokyo) and a part of the Suita PHC area were excluded since their study population was defined differently to the others, and no data for cancer incidence were available for the Katsushika PHC area.

In the baseline survey, 99,808 subjects were identified as the study population. Of these, 29 were found to be ineligible during the follow-up period and excluded owing to non-Japanese nationality (n = 18) and late report of emigration before the start of the follow-up period (n = 11). As a result, a population-based cohort of 99,779 subjects was established for this study.

The self-administered baseline questionnaire included items on sociodemographics, personal medical history, smoking, and alcohol drinking habits and diet (response rate: 79%). Excluded from the study were those who reported a history of cancer (e.g., colorectal, gastric, lung, liver, breast, uterine, and other cancer) (n = 2,135), were diagnosed with cancers (n = 240) or were found by late report to have emigrated before the start of the follow-up survey (n = 165), or reported missing data for coffee consumption (n = 1,077). Finally, 96,162 eligible subjects (46,023 men and 50,139 women) were included in the analysis.

Follow-up

Subjects were followed from the baseline survey until December 31, 2002. Residence status was confirmed annually through the residential registry maintained by the respective study area municipality; for those who moved out of the area, residence status was confirmed through the municipal office of the area to which they had moved. The occurrence of cancer was identified by active patient notification from major local hospitals in the study area and data linkage with population-based cancer registries, with permission from each of the local governments responsible for the cancer registries. Death certificate information was used as a supplementary information source. The site and histology of each case were coded using the International Classification of Diseases for Oncology, Third Edition (ICD-O-3).27

As of December 2002, a total of 1,163 cases of newly diagnosed colorectal cancer were identified during the follow-up period based on ICD-O-3 (code: C18-C20). Subsite-specific analyses were conducted as follows: C180–C189 for colon cancer, and C199 and C209 for rectal cancer. Finally, 763 cases of colon cancer and 400 of rectal cancer were included in the analysis.

Assessment of exposure

The self-administered baseline questionnaire included a section on nonalcoholic beverages such as coffee and green tea consumption. Information on coffee drinking was obtained in terms of the frequency and amount of coffee consumption and divided into the following categories: almost never, 1–2 days/week, 3–4 days/week, 1–2 cups/day, 3–4 cups/day, and 5 or more cups/day. The validity of the food frequency questionnaire for both men and women has been reported previously.28 The validity of coffee consumption was assessed by volunteer samples from the cohort using a dietary record for 28 days (7-day dietary records in 4 seasons) or 14 days. The rank correlation coefficient between questionnaire data and data from the dietary records was 0.42 for men and 0.38 for women, while that for green tea consumption was 0.57 for men and 0.63 for women.26 The type of coffee consumed (decaffeinated or caffeinated) was not included in the questionnaire.

Statistical Analysis

Person-years were counted from the date of response to the baseline survey until one of the following endpoints: the date of diagnosis of colorectal cancer, the date of emigration from the study area, the date of death, or the end of the study period, which ever came first. Persons who were lost to follow-up were censored on the last confirmed date of their presence in the study area.

Incident cases of colorectal cancer were the main outcome measure. In addition, to estimate the incidence rate of invasive colorectal cancer, incident cases were classified into 2 groups according to the depth of tumor invasion, i.e., invasive cases extending through the mucosal layer corresponding to code 3 (malignant, primary site) and noninvasive cases within the mucosal layer corresponding to carcinoma in situ in ICD-O-3.29

Relative risk (RR) and 95% confidence intervals (95% CI) were used to describe the RR of colorectal cancer incidence associated with coffee consumption. Subgroup analyses were done to investigate the relation of coffee consumption in men and women. Additional subsite-stratified analyses were done to examine associations with coffee consumption, and their respective RRs were estimated.

Cox proportional hazards modeling was used to examine the association between coffee consumption and the risk of developing colorectal cancer with person-years as the time scale. The assumptions for the Cox proportional hazards model were checked and found to hold. We adjusted for age at baseline (continuous) and study area (10 PHC areas), and then additionally adjusted for a family history of colorectal cancer (yes or no), smoking status (never, former, or current), alcohol intake (none, occasional, or regular), and BMI (in quartiles). Further adjustment was made for intake of green vegetables (at least 3 times per week, or less), beef (at least 3 times per week, or less), pork (at least 3 times per week, or less), green tea (none, less than everyday, everyday), black tea (none, less than everyday, everyday), and Chinese tea (none, less than everyday, everyday). These variables are either known or suspected risk or preventive factors for colorectal cancer.26, 30, 31, 32, 33, 34 The significance of interactions between coffee consumption and these risk factors for colorectal cancer was determined by comparing the model of coffee consumption × the respective variable with the model that assessed the main effect only. Statistical analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC). We used ordinal exposure values as a continuous variable to calculate the tests for trend. All P values were based on two-sided tests and were considered to be statistically significant at less than 0.05.

Results

During 10 years of follow-up (average follow-up period: 10 years) for 96,162 subjects (46,023 men and 50,139 women), a total of 1,163 cases of colorectal cancer developed in 962,985 person-years. Mean age at baseline was 51.9 years among men and 52.3 years among women. Baseline characteristics of study subjects according to coffee consumption in men and women are shown in Table I. Men who drank 3 or more cups of coffee/day were younger and more likely to report being more physically active and to have lower regular alcohol drinking than men who almost never drank coffee, but to have higher current smoking, lower consumption of green vegetables and green tea, and greater consumption of beef. Similar trends were found for women when the same analyses were performed. However, unlike in men, alcohol drinking in women increased as coffee consumption increased, although the proportion of regular drinkers was relatively low (Table I).

Table I. Baseline Characteristics of Study Subjects by Sex in the JPHC Study
 Coffee consumption (cups/day)
Almost never<11–23+p1
  • 1

    Based on Cochran-mantel-Haenszel statistics.

Men (n = 46,023)
 No. of subjects1435313956116616053 
 Person-years14138013915411342658061 
 Incidence rate (per 100,000)193.80155.94139.30129.17 
 Mean age (years)53.7852.2250.9848.84 
 Mean BMI (kg/m2)23.5323.6323.5123.40 
 Family history of colorectal cancer (%)1.171.271.461.340.09
 Current smokers (%)43.2449.8155.9771.08<0.0001
 Regular drinkers (%)68.7368.3267.4060.15<0.0001
 Leisure-time physical activity (%, more than once a week)20.5123.5327.5227.76<0.0001
 Green vegetable intake (%, more than 3 times/week)60.2858.5856.4151.76<0.0001
 Beef intake (%, more than 3 times/week)7.117.668.3110.54<0.0001
 Pork intake (%, more than 3 times/week)21.7623.1721.4023.230.35
 Green tea intake (%, daily)75.5274.9574.7969.17<0.0001
 Black tea intake (%, daily)2.011.803.623.24<0.0001
 Chinese tea intake (%, daily)10.638.9812.1710.370.02
Women (n = 50,139)
 No. of subjects1659314896140754575 
 Person-years16930015284614310845710 
 Incidence rate (per 100,000)109.2784.4072.6741.57 
 Mean age (years)55.1852.5450.2347.73 
 Mean BMI (kg/m2)23.6923.5823.3923.18 
 Family history of colorectal cancer (%)1.131.171.420.980.34
 Current smokers (%)4.354.266.8816.26<0.0001
 Regular drinkers (%)8.209.7613.4116.50<0.0001
 Leisure-time physical activity (%, more than once a week)16.6519.5420.7119.16<0.0001
 Green vegetable intake (%, more than 3 times/week)68.8268.6867.8363.98<0.0001
 Beef intake (%, more than 3 times/week)5.566.798.1211.07<0.0001
 Pork intake (%, more than 3 times/week)21.3926.9525.1827.20<0.0001
 Green tea intake (%, daily)76.8477.6274.8966.29<0.0001
 Black tea intake (%, daily)2.222.254.964.15<0.0001
 Chinese tea intake (%, daily)12.3311.4814.8416.07<0.0001

Associations between coffee consumption and the risk of incident colorectal cancer in men are presented in Table II. No significant association was observed, nor was any significant trend seen by site of colorectal cancer.

Table II. Relative Risk for Colorectal Cancer in Relation to Coffee Consumption in Men in the JPHC Study
 Coffee consumption (cups per day)
Almost never<11–23+p for trend
  • Adjusted for age and study area.

  • 1

    Additionally adjusted for BMI, smoking status, alcohol drinking, family history of colorectal cancer, physical activity, and intake of green vegetables, beef, pork, green tea, Chinese tea and black tea.

Total colorectal cancer
 No of cases (n = 726)27621715875 
 RR (95% CI)1.00 (reference)0.89 (0.74–1.06)0.94 (0.77–1.15)1.02 (0.79–1.33)0.83
 RR (95% CI)11.00 (reference)0.96 (0.78–1.17)0.94 (0.75–1.18)1.10 (0.82–1.47)0.91
Invasive colorectal cancer
 No of cases (n = 444)16213910142 
 RR (95% CI)1.00 (reference)0.97 (0.77–1.22)1.02 (0.79–1.32)0.98 (0.69–1.38)0.99
 RR (95% CI)11.00 (reference)1.02 (0.79–1.32)1.04 (0.78–1.38)1.08 (0.74–1.56)0.68
Colon cancer
 No of cases (n = 477)17415610047 
 RR (95% CI)1.00 (reference)1.01 (0.81–1.26)0.95 (0.74–1.22)1.03 (0.74–1.44)0.93
 RR (95% CI)11.00 (reference)1.12 (0.87–1.42)0.94 (0.70–1.25)1.15 (0.80–1.66)0.83
Invasive colon cancer
 No of cases (n = 271)90956026 
 RR (95% CI)1.00 (reference)1.20 (0.90–1.61)1.12 (0.81–1.57)1.13 (0.72–1.77)0.47
 RR (95% CI)11.00 (reference)1.26 (0.91–1.75)1.07 (0.73–1.57)1.23 (0.76–2.00)0.50
Rectal cancer
 No of cases (n = 249)102615828 
 RR (95% CI)1.00 (reference)0.68 (0.49–0.93)0.91 (0.66–1.27)1.01 (0.65–1.55)0.82
 RR (95% CI)11.00 (reference)0.69 (0.48–0.99)0.94 (0.64–1.36)1.01 (0.61–1.66)0.92
Invasive rectal cancer
 No of cases (n = 173)72444116 
 RR (95% CI)1.00 (reference)0.68 (0.47–1.00)0.90 (0.60–1.33)0.79 (0.45–1.37)0.39
 RR (95% CI)11.00 (reference)0.72 (0.47–1.10)0.99 (0.64–1.54)0.91 (0.49–1.68)0.86

Table III shows the RR of colorectal cancer by coffee consumption in women. Women who regularly consumed 3 or more cups of coffee per day had a 32% lower risk of colorectal cancer than those almost never consumed coffee, although the difference was not statistically significant (RR = 0.68; 95% CI = 0.40–1.15). Subsite-specific analyses showed a nonsignificant inverse association between coffee consumption and the risk of developing colon cancer. As compared with those in the lowest level, women in the highest level (3 or more cups/day) had a RR of 0.51 (95% CI = 0.26–0.99) with adjustment for age and study area and of 0.60 (95% CI = 0.31–1.19) after further adjustment for potential confounding factors. However, no significant trend was found for rectal cancer.

Table III. Relative Risk for Colorectal Cancer in Relation to Coffee Consumption in Women in the JPHC Study
 Coffee consumption (cups/day)
Almost never<11–23+p for trend
  • Adjusted for age and study area.

  • 1

    Additionally adjusted for BMI, smoking status, alcohol drinking, family history of colorectal cancer, physical activity, and intake of green vegetables, beef, pork, green tea, Chinese tea, and black tea.

Total colorectal cancer
 No. of cases (n = 437)18512910419 
 RR (95% CI)1.00 (reference)0.90 (0.72–1.13)0.97 (0.75–1.24)0.64 (0.40–1.05)0.21
 RR (95% CI)11.00 (reference)0.92 (0.71–1.19)1.01 (0.76–1.33)0.68 (0.40–1.15)0.42
Invasive colorectal cancer
 No. of cases (n = 299)132827213 
 RR (95% CI)1.00 (reference)0.81 (0.61–1.06)0.93 (0.69–1.26)0.60 (0.33–1.08)0.17
 RR (95% CI)11.00 (reference)0.82 (0.60–1.12)0.95 (0.68–1.32)0.58 (0.30–1.10)0.23
Colon cancer
 No. of cases (n = 286)129836410 
 RR (95% CI)1.00 (reference)0.84 (0.64–1.11)0.87 (0.64–1.20)0.51(0.26–0.99)0.07
 RR (95% CI)11.00 (reference)0.87 (0.64–1.19)0.94 (0.67–1.33)0.60(0.31–1.19)0.27
Invasive colon cancer
 No. of cases (n = 187)8853406 
 RR (95% CI)1.00 (reference)0.77(0.55–1.09)0.75 (0.50–1.11)0.40 (0.17–0.93)0.02
 RR (95% CI)11.00 (reference)0.79 (0.54–1.16)0.74 (0.48–1.14)0.44 (0.19–1.04)0.04
Rectal cancer
 No. of cases (n = 151)5646409 
 RR (95% CI)1.00 (reference)1.04 (0.70–1.54)1.17 (0.76–1.80)0.92 (0.44–1.90)0.74
 RR (95% CI)11.00 (reference)1.04 (0.67–1.62)1.16 (0.72–1.88)0.84 (0.36–1.94)0.87
Invasive rectal cancer
 No. of cases (n = 112)4429327 
 RR (95% CI)1.00 (reference)0.87(0.54–1.40)1.32(0.81–2.14)1.05(0.46–2.41)0.43
 RR (95% CI)11.00 (reference)0.88(0.52–1.50)1.42(0.83–2.42)0.90 (0.34–2.41)0.46

We conducted additional analyses restricted to 943 cases of invasive colorectal cancer. Among women, the significant inverse association was essentially limited to colon cancer and was strengthened: those who regularly consumed 3 or more cups of coffee per day had a nearly 60% lower risk of colon cancer than those who almost never consumed coffee (RR = 0.44; 95% CI = 0.19–1.04). A statistically significant decrease in the risk of invasive colon cancer was associated with increasing levels of coffee consumption among women (p for trend = 0.04). However, no significant trend was found for invasive rectal cancer (Table III).

Additionally, we also investigated the significance of interactions between coffee consumption and potential confounding factors, including age, smoking, alcohol drinking, body mass index, physical activity, green vegetable intake and green tea intake in relation to the risk of colorectal cancer. None of the tests for statistical interaction between coffee consumption and these variables produced significant results in this study. Further, no significant association was observed on stratified analysis confined to nonsmoking and nondrinking men, who albeit accounted for only 7.3% of all men (data not shown).

To avoid potential bias because of preclinical conditions, we performed further analyses after excluding colorectal cancer cases diagnosed during the first 2 year of follow-up. However, the tendencies were essentially unchanged after this exclusion (data not shown).

Further, additional analyses to investigate the association of green tea consumption with colorectal cancer risk found no significant association in men or women (Table IV).

Table IV. Relative Risk for Colorectal Cancer in Relation to Green Tea Consumption in Men and Women in the JPHC Study
 Green tea consumption (cups per day)
Almost never<11–23–45+p for trend
  • Adjusted for age and study area.

  • 1

    Additionally adjusted for BMI, smoking status, alcohol drinking, family history of colorectal cancer, physical activity, and intake of green vegetables, beef, pork, coffee, Chinese tea and black tea.

Men
 Total colorectal cancer
  No. of cases (n =724)8673151212202 
  RR (95% CI)1.00 (reference)0.78 (0.57–1.06)0.91 (0.70–1.20)0.98(0.76–1.28)0.86(0.66–1.12)0.75
  RR (95% CI)11.00 (reference)0.80 (0.56–1.14)1.00 (0.73–1.35)1.04(0.78–1.40)0.96(0.71–1.29)0.56
 Colon cancer
  No. of cases (n = 476)5542109143127 
  RR (95% CI)1.00 (reference)0.70 (0.47–1.04)1.05 (0.75–1.47)1.06 (0.76–1.46)0.85 (0.61–1.17)0.80
  RR (95% CI)11.00 (reference)0.73 (0.47–1.14)1.09 (0.75–1.59)1.06 (0.74–1.53)0.92 (0.63–1.33)0.75
 Rectal cancer
  No. of cases (n = 248)3131426975 
 RR (95% CI)1.00 (reference)0.92 (0.56–1.52)0.68 (0.42–1.09)0.86 (0.55–1.34)0.89 (0.57–1.37)0.84
 RR (95% CI)11.00 (reference)0.93 (0.52–1.67)0.81 (0.47–1.39)1.01 (0.60–1.67)1.04 (0.63–1.72)0.57
Women
 Total colorectal cancer
  No. of cases (n = 434)535470118139 
  RR (95% CI)1.00 (reference)0.98 (0.67–1.44)0.80 (0.56–1.15)0.91 (0.65–1.27)0.96 (0.69–1.34)0.96
  RR (95% CI)11.00 (reference)1.01 (0.67–1.52)0.81 (0.55–1.22)0.96 (0.66–1.40)1.02 (0.70–1.47)0.79
 Colon cancer
  No. of cases (n = 284)3633427697 
  RR (95% CI)1.00 (reference)0.85 (0.53–1.37)0.69 (0.44–1.08)0.82 (0.54–1.24)0.92 (0.62–1.37)0.95
  RR (95% CI)11.00 (reference)0.97 (0.57–1.63)0.81 (0.49–1.35)0.99 (0.62–1.57)1.10 (0.70–1.73)0.45
 Rectal cancer
  No. of cases (n = 150)1721284242 
  RR (95% CI)1.00 (reference)1.26 (0.66–2.39)1.04 (0.56–1.93)1.09 (0.60–1.98)1.03 (0.57–1.86)0.84
  RR (95% CI)11.00 (reference)1.09 (0.56–2.13)0.81 (0.42–1.57)0.92 (0.49–1.71)0.85 (0.45–1.61)0.55

Discussion

In this large population-based prospective study of Japanese men and women, we found a significant inverse association between coffee consumption and the risk of developing invasive colon cancer among women. A nonsignificant inverse association was observed between coffee consumption and overall colorectal cancer. The study also found a significant dose-response relationship between increasing levels of coffee consumption and a lower risk of invasive colon cancer among women. In contrast to these results; however, no significant associations were found among men.

Meta-analyses of coffee consumption and colorectal cancer risk revealed conflicting results in case–control and cohort studies.3, 4 The combined results of 12 case–control studies suggested that coffee consumption is associated with a reduced risk of colorectal cancer (OR for high versus low consumption = 0.72; 95% CI = 0.61–0.84),5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 whereas the combined results of 5 prospective studies revealed no association,16, 17, 18, 19, 20 albeit that the number of cases in each study was small (RR = 0.97; 95% CI = 0.73–1.29). On this basis, the overall results from 17 studies found a RR of 0.76 (95% CI = 0.66–0.89, p for trend < 0.001) for high versus low level of coffee consumption. Although these results remain open to discussion owing to these inconsistencies between cohort and case–control studies, it seems unlikely that a constant methodological artifactin the case–control studies would account for the relative consistency of the results in diverse countries and settings. Moreover, consistent with the inverse association between coffee consumption and colon cancer among women in the present study, recent results from another prospective cohort study in Japan also indicated that daily coffee consumption significantly decreased the risk of colon cancer in women in comparison with those who almost never consumed coffee (RR = 0.43; 95% CI = 0.22–0.85; p for trend < 0.01), but not in men.25 However, this previous study used fewer coffee consumption categories and a smaller number of cancer cases than the present study.

Further, although previous prospective studies showed no association between coffee consumption and colorectal cancer,16, 17, 18, 19, 20, 21, 22, 23 these were conducted mainly in Western countries. Owing to insufficient numbers of coffee nondrinkers, investigators had to include both never drinkers and those who drank less than 1 cup/day as the reference group. Compared to these prospective studies in Western countries, the recent results in Japan were obtained using relatively low level as the reference group.25 The reference group in this study was those who almost never consumed coffee, which may partially explain why they did not show the protective effect of coffee consumption on colorectal cancer seen in previous studies.

The present study showed a significant inverse association between coffee consumption and colon cancer among women. Although previous studies have investigated the association with colorectal cancer by anatomical subsite,6, 9, 11, 13, 15, 35 results have been inconsistent. The inverse association with colon cancer observed here is consistent with some of these.11, 13, 35 A recent report using data from a prospective study in Japan showed a significant inverse association between coffee consumption and colon cancer in women,25 although with a smaller number of incident cases and fewer coffee consumption categories than the present study. Although we accounted for lifestyle factors and dietary habits that can be associated with colorectal cancer as potential confounders, adjustment for these variables did not alter our results. Moreover, a previous report using data from the JPHC Study revealed no statistically significant association between dietary fiber intake and colorectal cancer.36 To avoid potential bias because of preclinical conditions, we performed additional analyses after excluding cases diagnosed within the first 2 years of follow-up, but the tendencies were essentially unchanged. Notwithstanding the known limitations of our study, we consider that our results are meaningful and not the result of chance. One possible hypothesis might be that coffee reduces the concentration of 2 potential intestinal carcinogens in the bowel, namely bile acids and neutral sterol,37 which would appear consistent with the restriction of the association here to the colon.4, 35

Potential biological mechanisms may underlie the inverse association between coffee consumption and colon cancer observed here. These include a reduction in cholesterol, bile acid and neutral sterol secretion in the colon by substances in coffee, on the basis that bile acids are promoters of colon carcinogenesis.37 Coffee consumption may increase colonic motility, which may be inversely related to risk via the reduced epithelial exposure to colonic carcinogens,38 although this interpretation remains open to discussion.39 Coffee consumption might also lower the risk of hyperglycemia, which would in turn decrease colon cancer risk by preventing the risk of diabetes mellitus.40, 41 Other plausible mechanisms may relate to the antioxidant properties of coffee beverages,42, 43 given that coffee beans contain several phenolic compounds with antioxidant properties such as caffeinic acid and chlorogenic acid, or to the potential antimutagenic properties of some coffee components against heterocyclic amines and other mutagenic agents.44 Finally, the coffee diterpenes cafestol and kahweol have been shown to produce a broad range of biochemical effects which reduce the genotoxicity of several carcinogens.45

In contrast, we observed no association in the risk of colorectal cancer among men. The reason for this null finding is unclear, but might be at least in part attributable to residual confounding effects in men, particularly smoking or alcohol drinking. Indeed, previous results from the JPHC study showed a statistically significant association of colorectal cancer risk with both smoking and alcohol drinking in men but not women.30 Therefore, the role of smoking and alcohol drinking may be strongly associated with the risk among men in this study. One way to reduce potential confounding by these factors is to restrict the analysis to never smokers and nondrinkers. We examined the association between coffee consumption and colorectal cancer risk in nonsmoking and nondrinking men, but no significant association was observed on stratified analysis. Moreover, only 7.3% of men were nonsmokers and nondrinkers. Although the potential for true sex differences in response to coffee consumption cannot be ruled out, results from a meta-analysis which showed no strong sex difference, apart from small differences reported in some studies, are likely more significant.3

Diabetes may be a potential confounder of the coffee-colon cancer relationship. However, a previous report using data from the JPHC Study showed no clear association between colon cancer and diabetes in women (RR = 0.93; 95% CI = 0.49–1.75) and a slightly increased risk of colon cancer in men (RR = 1.36; 95% CI = 1.00–1.85).46 Further, additional adjustment for diabetes in the multivariate analyses did not alter the rate ratios, as shown in the following: for colorectal cancer among men, RR (95% CI) = 1.00, reference; 0.96 (0.77–1.19), <1 cup per day; 0.89 (0.69–1.13), 1–2 cups per day; 0.98 (0.70–1.36), 3+ cups per day; p for trend = 0.805. For colorectal cancer among women, RR (95% CI) = 1.00, reference; 0.81 (0.61–1.09), <1 cup per day; 0.88 (0.65–1.19), 1–2 cups per day; 0.47 (0.24–0.90), 3+ cups per day; p for trend = 0.251.

The main limitation of this study was its use of self-reports. Because the data were collected before diagnosis; however, any imprecision was likely to have caused an underestimation rather than overestimation of associations. Further, since the data were based on the baseline survey only, changes in coffee intake that arose from symptoms related to a subsequent diagnosis of colorectal cancer after the start of the study may have resulted in misclassification. Such misclassification, if any, is probably nondifferential and would likely lead to an underestimation of the results. With regard to other limitations, although unmeasured confounders may not have been controlled in this study, lifestyle factors and dietary habits previously identified as potential confounders for colorectal cancer were accounted for,26, 30, 31, 32, 33, 34 with the expectation that statistical adjustment might be to some degree possible if information on associated variables were available. On the other hand, whereas we did not have data on the type of coffee consumed (e.g. filtered or boiled, decaffeinated or not), we note that decaffeinated coffee is very uncommon in Japan. Moreover, it remains unclear whether caffeine itself or some other substances in coffee beans are responsible for the protective effect on colorectal cancer,4 although a significant reduction in the risk of rectal cancer with decaffeinated coffee has been reported in a previous study.24 Further, because green tea is a major source of caffeine intake in the Japanese population, we additionally analyzed the association between green tea consumption and colorectal cancer risk, but found no association (data not shown).

Notwithstanding the known limitations of observational studies, the present study has the following strengths: it was a prospective population-based study with a 10-year follow-up period; information on coffee consumption was collected before the subsequent onset of cancer, eliminating the exposure recall bias inherent to case–control studies; study subjects were selected from the general population, and response rate to the questionnaire was high while the proportion of losses to follow-up was negligible; and finally, adjustment was made for lifestyle factors and dietary habits previously identified as possibly associated with colorectal cancer.26, 30, 31, 32, 33, 34

In conclusion, we observed a significant inverse association between coffee consumption and the risk of developing invasive colon cancer among women. Although we cannot deny the possibility of a chance finding because of the aforementioned limitations and unmeasured confounding factors which may not have been controlled, these findings suggest that coffee consumption may lower the risk of colon cancer among Japanese women.

Acknowledgements

Mr. K.-J. Lee was awarded a Visiting Scientist Fellowship from the Foundation for Promotion of Cancer Research in Japan. The authors thank all staff members in each study area for their unfailing efforts in conduction the baseline and follow-up surveys. We also express our gratitude to the Iwate, Aomori and Okinawa Cancer Registries for providing the incidence data.

Appendix

Members of the Japan Public Health Center-based Prospective Study (JPHC Study, principal investigator: S. Tsugane) Group are: S. Tsugane, M. Inoue, T. Sobue, and T. Hanaoka, Research Center for Cancer Prevention and Screening, National Cancer Center, Tokyo; J. Ogata, S. Baba, T. Mannami, and A. Okayama, National Cardiovascular Center, Suita; K. Miyakawa, F. Saito, A. Koizumi, Y. Sano, and I. Hashimoto, Iwate Prefectural Ninohe Public Health Center, Ninohe; Y. Miyajima, N. Suzuki, S. Nagasawa, and Y. Furusugi, Akita Prefectural Yokote Public Health Center, Yokote; H. Sanada, Y. Hatayama, F. Kobayashi, H. Uchino, Y. Shirai, T. Kondo, R. Sasaki, Y. Watanabe, and Y. Miyagawa, Nagano Prefectural Saku Public Health Center, Saku; Y. Kishimoto, E. Takara, T. Fukuyama, M. Kinjo, M. Irei, and H. Sakiyama, Okinawa Prefectural Chubu Public Health Center, Okinawa; K. Imoto, H. Yazawa, T. Seo, A. Seiko, F. Ito, and F. Shoji, Katsushika Public Health Center, Tokyo; A. Murata, K. Minato, K. Motegi, and T. Fujieda, Ibaraki Prefectural Mito Public Health Center, Mito; K. Matsui, T. Abe, M. Katagiri, and M. Suzuki, Niigata Prefectural Kashiwazaki and Nagaoka Public Health Center, Kashiwazaki and Nagaoka; M. Doi, A. Terao, and Y. Ishikawa, Kochi Prefectural Chuo-higashi Public Health Center, Tosayamada; H. Sueta, H. Doi, M. Urata, N. Okamoto, and F. Ide, Nagasaki Prefectural Kamigoto Public Health Center, Arikawa; H. Sakiyama, N. Onga, and H. Takaesu, Okinawa Prefectural Miyako Public Health Center, Hirara; F. Horii, I. Asano, H. Yamaguchi, K. Aoki, S. Maruyama, and M. Ichii, Osaka Prefectural Suita Public Health Center, Suita; S. Matsushima and S. Natsukawa, Saku General Hospital, Usuda; M. Akabane, Tokyo University of Agriculture, Tokyo; M. Konishi, and K. Okada, Ehime University, Ehime; H. Iso, Osaka University, Suita; Y. Honda and K. Yamagishi, Tsukuba University, Tsukuba; H. Sugimura, Hamamatsu University, Hamamatsu; Y. Tsubono, Tohoku University, Sendai; M. Kabuto, National Institute for Environmental Studies, Tsukuba; S. Tominaga, Aichi Cancer Center Research Institute, Nagoya; M. Iida and W. Ajiki, Osaka Medical Center for Cancer and Cardiovascular Disease, Osaka; S. Sato, Osaka Medical Center for Health Science and Promotion, Osaka; N. Yasuda, Kochi University, Nankoku; S. Kono, Kyushu University, Fukuoka; K. Suzuki, Research Institute for Brain and Blood Vessels Akita, Akita; Y. Takashima, Kyorin University, Mitaka; E. Maruyama, Kobe University, Kobe; M. Yamaguchi, Y. Matsumura, S. Sasaki, and S. Watanabe, National Institute of Health and Nutrition, Tokyo; T. Kadowaki, Tokyo University, Tokyo; Y. Kawaguchi, Tokyo Medical and Dental University, Tokyo; and H. Shimizu, Sakihae Institute, Gifu.

Ancillary