Abstract
Purpose
We evaluated the association of coffee and tea drinking with risk of the urinary tract cancer in Finnish men, with high coffee consumption, using data from the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study.
Methods
The ATBC trial conducted from 1985 to 1993, enrolled 29,133 male smokers. We used Cox proportional hazards regression models to estimate hazard ratios (HRs) and confidence intervals (CIs), using men who drank >0 but <1 cup coffee/day and tea non-drinkers as our referent group for coffee and tea analyses, respectively.
Results
During 472,402 person-years of follow-up, 835 incident cases of bladder cancer and 366 cases of RCC were ascertained. For bladder cancer, we observed no association for coffee consumption (HR ≥4 vs >0 to <1 cups/day = 1.16, 95% CI=0.86–1.56) and a borderline statistically significant inverse association for tea consumption (HR ≥1 vs 0 cup/day=0.77, 95% CI=0.58–1.00). For RCC, we observed no association for coffee (HR ≥4 vs >0 to <1 cuPs/day = 0.85, 95% CI=0.55–1.32) or tea consumption (HR ≥1 vs 0 cup/day=1.00, 95% CI=0.68– 1.46). We found no impact of coffee preparation on coffee-cancer associations.
Conclusion
Coffee drinking was not associated with urinary tract cancers risk. Further research on tea and bladder cancer are warranted.
Keywords: coffee, tea, bladder cancer, Renal Cell Carcinoma, caffeine, catechin
Introduction
After water, coffee and tea are the two most commonly consumed drinks worldwide. Coffee drinking has been consistently and inversely associated with liver 1 and endometrial cancer 2. A pooled analysis including 13 prospective cohort studies found suggestive inverse associations between coffee drinking and renal cell carcinoma (RCC) 3, but adverse associations between coffee drinking and risk of bladder cancer in case-control studies 4 have raised concerns that coffee consumption could be harmful. A meta-analysis including 34 case-control studies and six prospective cohort studies observed a positive association between coffee drinking and bladder cancer, particularly among men 4, but another meta-analysis found no evidence of a coffee-bladder cancer association 5. Results have also been inconsistent for smokers versus nonsmokers with some studies only reporting a positive association between coffee and bladder cancer among nonsmokers 5 and others only reporting a positive association among smokers 6. Smoking is a known risk factor for bladder cancer, and incidence rates in men are highest in Europe and North America, with global trends reflecting the prevalence of tobacco smoking 7.
Coffee contains many phenolic compounds and diterpenes, such as cafestol and kahweol, which may reduce the genotoxicity of some carcinogens 8. Unfiltered coffees, including boiled coffee, contain relatively high levels of cafestol and kahweol but low or negligible levels are found in instant or filtered coffees 9. Few studies collect data on coffee preparation method, and no prospective studies have evaluated the impact of preparation method on the association of coffee drinking with cancer of the urinary tract. A meta-analysis highlighted the need for prospective studies investigating the impact of coffee preparation method 5.
Tea drinking has been inversely associated with certain types of cancer including oral cancer 10, and tea contains polyphenols such as catechins with antioxidant properties 11that have been shown to prevent tumor progression by interfering in cell signaling in experimental studies 11. However, epidemiologic studies of tea drinking and cancer of the urinary tract have yielded inconsistent results 3, 12, and few studies have considered the association among smokers. A pooled analysis including 13 prospective cohort studies found suggestive inverse associations between tea drinking and renal cell carcinoma (RCC) 3, and a report by the World Cancer Research Fund and the American Institute for Cancer Research concluded that there is limited suggestive evidence of an inverse association between tea and bladder cancer 13.
Per capita consumption of coffee in Finland is among the highest in the world 14, and both tobacco smoking and male sex are major risk factors for bladder cancer 7. Thus, the Alpha-Tocopherol, Beta-Carotene Cancer Prevention study (ATBC), a cohort of Finnish male smokers with a high prevalence of heavy coffee drinking who also provided information on coffee preparation method and tea drinking, provided a unique opportunity to study the association of coffee and tea drinking with bladder cancer and RCC.
Materials and Methods
Study design
The ATBC Trial, which investigated whether vitamin supplementation could prevent lung cancer in smokers, was conducted in Finland from 1985 to 1993 and enrolled 29,133 male smokers. The design of the ATBC Trial and subsequent cohort study have been described elsewhere 15. In brief, male smokers were eligible to participate if they smoked at least five cigarettes/day, had no history of cancer, or a serious disease limiting long term follow-up, and were not using vitamin supplements at date of randomization. Eligible men who gave informed consent were randomized to one of four groups using a 2*2 factorial design: Beta Carotene (20 mg/day) and/or Alpha Tocopherol (50 mg/day). The trial ended in April 1993; however, follow-up continued for cancer incidence and death using the Finnish Cancer Registry and the Registry of Cause of Death, respectively, providing complete ascertainment of cases 16. The study protocol was approved by the institutional review board of the National Cancer Institute in the US and the National Institute for Health and Welfare in Finland 15.
For our analysis, we excluded men with a self-reported history of renal failure (n=57), as this condition may alter fluid consumption, those with extreme energy intake, defined as more than two interquartile ranges above the 75th percentile or below the 25th percentile of energy intake (n=213), and those missing data on coffee or tea consumption (n=2022). Our final analytic sample included 26,841 men.
Outcome Ascertainment
Incident cancer cases were ascertained via the Finnish Cancer Registry and medical records 17, 18. Bladder cancers were defined as ICD-9 code 188 or 233.7, and RCC as ICD-9 code 189.0. Renal pelvis/ureter cancers (ICD-9 code 189.1 and 189.2) were not included in our case definition due to small sample size (N=62). Follow-up time (i.e., person years) was calculated from date of randomization (i.e. baseline assessment) to date of death, date of bladder cancer or RCC diagnosis, or end of follow-up (i.e., December 31, 2015), whichever came first. Because bladder cancer and RCC have different risk factors, most notably tobacco smoking, we analyzed these outcomes separately.
Exposure Assessment
A validated 276-item food frequency questionnaire (FFQ), which included foods commonly consumed in Finland, and picture book illustrating portion sizes were used to collect information about usual dietary intake during the previous year 19. Coffee and tea consumptions were assessed separately as cups per day or per week. Cup size was ascertained using pictures depicting common portion sizes. Data on frequency and portion size were used to estimate the number of grams of coffee and tea typically consumed per day with typical cup sizes ranging from 70 to 170 grams for coffee and 110 to 220 grams for tea. We converted grams to cups using USDA conversion tables, estimating that 237 grams of coffee or tea equals one 8 fluid ounce cup 20, to enable comparison with prior research on coffee and cancer 6, 20. A study conducted to evaluate the reproducibility and validity of the ATBC dietary assessment very high intraclass correlation coefficients for coffee consumption (ranging from 0.72 to 0.79), comparing data from food records and FFQs 19. In 1990, approximately two to five years after baseline, participants reported how they usually prepared their coffee using four pre-defined categories: filtered (i.e., pouring water over ground, roasted coffee within a filter), boiled (i.e., pouring water on ground coffee, and drinking when the grounds settle), instant, or a combination of the methods, and no coffee drinking. In total, 20,603 men provided data on coffee preparation. Approximately 71% and 21% of the participants reported usually drinking filtered and boiled coffee, respectively. Type of tea was not specified in the questionnaire as black tea was the predominant type of tea consumed in Finland at the time of the study.
Demographic data, past medical history, and smoking status were obtained at the time of randomization. Body mass index (BMI) was computed from measured height and weight (kg/m2). Hypertension was defined as blood pressure readings at baseline of systolic blood pressure >140 mmHg and/or diastolic blood pressure >90 mmHg, or self-reported history of hypertension. The baseline questionnaire queried occupational exposure to lead refining, nickel refining, copper smelting, steel production, oil refining, gas manufacture, manufacture of chromium pigments, arsenic production or foundry work. We adjusted for these potential bladder cancer-related occupational exposures 21, using dummy variables in our multivariable-adjusted models.
Statistical Analysis
Coffee drinkers were categorized into one of six groups (non-drinkers, >0 but <1, ≥1, >2, ≥3, or ≥4 cups/day), and tea drinkers were categorized into one of three groups (non-drinkers, <1, or ≥1 cup/day). Coffee drinkers who reported consuming more than zero but less than one cup of coffee per day were used as the referent group for coffee analyses since few men in this cohort were coffee nondrinkers. However, there were a considerable number of tea non-drinkers; thus, analyses of tea consumption use tea non-drinkers as the referent group. For continuous analyses, we estimated the hazard ratios (HRs) associated with a one cup per day increase in either coffee or tea consumption.
Multivariable Cox proportional hazards regression models were used to estimate HRs and 95% confidence intervals (CIs) for categories of coffee or tea consumption and cancers of the urinary tract. Linear trends across categories were evaluated using the median consumption of each category. We used person-years as the underlying time metric. Using age as the underlying metric yielded similar results (data not shown).
We adjusted multivariable models for potential confounders including: age (years), BMI (kg/m2), level of education/training (less than or equal to a high school education or more than a high school education), smoking duration (years), smoking intensity (cigarettes per day), alcohol (grams/day), history of diabetes (yes or no), frequency of physical activity in leisure time (<1/week, 1–2/week, or ≥3/week), energy intake (kcal/day), nutrient density adjusted fruit consumption (grams/day), and nutrient density adjusted vegetable consumption (grams/day). For RCC analyses, we additionally adjusted for the presence of hypertension. We also mutually adjusted for coffee and tea consumption in tea and coffee models, respectively.
In secondary analyses of coffee drinking and cancers, we collapsed the two highest coffee consumption categories to preserve sample size. We tested for potential effect modification by smoking status, and method of preparation by comparing models with and without the interaction term between coffee or tea consumption (continuous) and continuous pack-years or each level of coffee preparation method using the likelihood ratio test. To evaluate the impact of smoking on risk estimates, we stratified by median pack-years. Next, to explore the impact of coffee preparation method, we adjusted for consumption of filtered, boiled, instant, or a combination of coffee types using categorical variables in a subset of participants (n=20,603) with information on method of preparation. We compared the HRs from the multivariable adjusted model to the model additionally adjusted for preparation method, using likelihood ratio test. We also stratified by method of coffee preparation defined as either filtered or boiled (the most common methods). Finally, to evaluate potential reverse causality, we excluded participants with less than two years of follow-up. We observed no deviations from the proportional hazards assumption using the Schoenfeld residuals test (P=0.24 for coffee; P=0.16 for tea consumption).
P values of less than 0.05 were considered statistically significant, and all tests were two-sided. All statistical analyses were carried out using STATA software (version 14, STATA Corp, College Station, TX, USA).
Results
During 472,402 person-years of follow-up, 835 incident cases of bladder cancer and 366 incident cases of RCC were ascertained. Median follow-up time was 17.6 years, and mean age at baseline was 57.2 years. Approximately 98% of participants reported drinking coffee and 36% reported drinking tea at baseline. Mean consumption among drinkers was 2.3 cups/day for coffee and 0.7 cups/day for tea. The mean (±SD) number of cigarettes smoked per day was 20.4 (±8.8), and the mean number of years smoked was 35.9 (±8.4). Participants who reported heavier coffee drinking tended to be younger, smoke more cigarettes per day, drink less alcohol and less tea, be less educated, have a lower prevalence of hypertension, and consume more calories and vegetables but less fruit per day (Table 1). Participants who reported heavier tea drinking tended to smoke for fewer years, be more educated, drink less coffee, have a higher prevalence of hypertension, and consume more calories, fruits, and vegetables per day (Table 1).
Table 1.
Baseline characteristics of the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study by coffee and tea consumption (N=26,841)
| Coffee consumption (cups/day)1 | Tea consumption (cups/day)1 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Non-Drinkers | <1 | >1 | >2 | >3 | >4 | P value | Non-drinker | <1 | >1 | P value | |
| No. of participants | 670 | 3086 | 7175 | 8042 | 4472 | 3396 | 17204 | 7266 | 2371 | ||
| Randomization group,% | 0.26 | 0.33 | |||||||||
| Placebo | 29.0 | 24.8 | 25.1 | 24.9 | 24.8 | 25.6 | 24.8 | 25.7 | 25.5 | ||
| AT2 | 21.3 | 25.2 | 25.6 | 24.1 | 25.3 | 25.5 | 25.2 | 24.0 | 25.7 | ||
| BC3 | 23.7 | 25.3 | 24.6 | 25.2 | 25.2 | 25.1 | 25.0 | 25.4 | 24.2 | ||
| AT+BC | 26.0 | 24.7 | 24.7 | 25.8 | 24.8 | 23.8 | 25.0 | 24.9 | 24.6 | ||
| Age, years4 | 57.2±5.0 | 57.6±5.0 | 57.8±5.2 | 57.3±5.1 | 56.5± 4.8 | 56.2± 4.7 | <0.0001 | 57.2±5.0 | 57.2±5.1 | 57.1±5.1 | 0.73 |
| BMI4, kg/m2 | 25.9±4.2 | 26.4±3.8 | 26.3±3.8 | 26.2±3.7 | 26.3± 3.7 | 26.4± 3.8 | 0.01 | 26.3±3.8 | 26.2±3.7 | 26.2±3.7 | 0.52 |
| Years of smoking4 | 36.1±8.2 | 35.7±9.0 | 36.2±8.6 | 35.8±8.5 | 35.7±7.9 | 36.1±7.9 | 0.001 | 36.2±8.3 | 35.4±8.6 | 35.4±8.6 | <0.0001 |
| No. of cigarette/day4 | 20.7±9.8 | 19.2±9.0 | 19.0±8.2 | 20.1±8.5 | 21.5±8.7 | 23.5±9.4 | <0.0001 | 20.8±8.3 | 19.7±8.6 | 19.4±8.9 | 0.06 |
| Alcohol4, g/day | 23.6±26.6 | 26.7±28.1 | 19.7±21.8 | 15.9±18.8 | 15.1±18.4 | 13.5±18.4 | <0.0001 | 17.4±21.1 | 18.9±21.7 | 18.7±22.5 | <0.0001 |
| Education, greater than high school, % | 64.5 | 72.1 | 70.6 | 65.6 | 64.4 | 59.0 | <0.0001 | 62.5 | 74.6 | 71.9 | <0.0001 |
| History of diabetes, % | 5.1 | 5.1 | 4.2 | 3.8 | 4.3 | 4.3 | 0.05 | 4.4 | 4.0 | 3.7 | 0.12 |
| Physical activity in leisure time, < 1/week (the lowest category) | 53.3 | 51.1 | 47.2 | 51.0 | 55.3 | 57.7 | <0.0001 | 53.4 | 48.1 | 49.1 | <0.0001 |
| Family history of bladder cancer, % | 1.1 | 1.2 | 1.1 | 1.2 | 1.3 | 1.4 | 0.91 | 1.2 | 1.1 | 1.6 | 0.38 |
| Tea drinker | 61.2 | 60.4 | 42.9 | 31.5 | 25.4 | 18.4 | <0.0001 | ||||
| Coffee drinker | 98.5 | 98.1 | 88.7 | <0.0001 | |||||||
| Energy4, kcal/day | 2526±724 | 2475±688 | 2534±671 | 2669±672 | 2815±704 | 2942±776 | <0.0001 | 2643±711 | 2701±704 | 2724±717 | <0.0001 |
| Fruit consumption4, grams/day | 88.0±84.0 | 97.2±82.9 | 97.6±81.0 | 93.2±80.7 | 89.1±83.4 | 85.3±88.1 | <0.0001 | 88.5±82.6 | 101.8±81.5 | 99.0±83.9 | <0.0001 |
| Vegetable consumption, grams/day4 | 274.9±274.9 | 282.8±101.6 | 296.1±100.8 | 294.3±101.9 | 293.8±105.8 | 287.4±110.7 | <0.0001 | 287.6±102.9 | 302.6±104.1 | 291.6±105.0 | <0.0001 |
| Presence of hypertension, % | 23.6 | 25.4 | 21.4 | 17.7 | 16.4 | 13.7 | <0.0001 | 18.8 | 19.0 | 20.8 | <0.0001 |
One 8 oz cup=237 grams of coffee or tea
Alpha-Tocopherol
Beta-Carotene
Mean±SD
We observed no statistically significant associations for coffee drinking (HR ≥4 vs >0 to <1 cups/day = 1.16, 95% 0=0.86–1.56, P trend=0.29, Table 2) and a non-significant inverse association for tea drinking (HR >≥1 vs 0 cup/day=0.77, 95% CI=0.58–1.00, P trend=0.05, Table 3) with bladder cancer in multivariable adjusted models. In analyses of continuous coffee and tea consumption, we observed no statistically significant associations between either coffee (HR1 cup/day increase=1.03; 95% CI 0.98–1.08) or tea (HR1 cup/day increase=0.92; 95% CI 0.82–1.03) consumption and bladder cancer.
Table 2.
Hazard ratios (95% Confidence Intervals) for cancer of urinary system according to coffee consumption in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study
| Model adjustment | Coffee consumption (cups/day)1 | |||||||
|---|---|---|---|---|---|---|---|---|
| Non-drinker | <1 | >1 | >2 | >3 | >4 | P trend | 1 cup/day increase | |
| Bladder cancer | ||||||||
| No of cases/person-years | 17/10,452 | 80/51,228 | 214/124,128 | 256/143,952 | 149/81,838 | 119/61,468 | ||
| Unadjusted model | 1.06 (0.631.79) | 1 | 1.09 (0.841.41) | 1.12 (0.641.75) | 1.13 (0.861.48) | 1.20 (0.911.60) | 0.21 | 1.03 (0.991.08) |
| Multivariable adjusted2 | 1.03 (0.611.74) | 1 | 1.04 (0.811.35) | 1.06 (0.821.37) | 1.10 (0.831.46) | 1.16 (0.861.56) | 0.29 | 1.03 (0.981.08) |
| Multivariable adjusted3 | 1.03 (0.611.74) | 1 | 1.02 (0.791.32) | 1.02 (0.791.32) | 1.05 (0.791.40) | 1.10 (0.811.49) | 0.52 | 1.02 (0.971.07) |
| Renal cell carcinoma | ||||||||
| No of cases/person-years | 9/10,720 | 42/52,462 | 92/121,975 | 113/144,756 | 65/80,496 | 45/61,128 | ||
| Unadjusted model | 1.05 (0.512.16) | 1 | 0.90 (0.621.29) | 0.95 (0.671.35) | 0.95 (0.641.40) | 0.88 (0.581.34) | 0.69 | 0.99 (0.921.06) |
| Multivariable adjusted2,4 | 1.03 (0.502.12) | 1 | 0.86 (0.601.25) | 0.92 (0.641.32) | 0.94 (0.631.40) | 0.85 (0.551.32) | 0.68 | 0.99 (0.921.07) |
| Multivariable adjusted3,4 | 1.03 (0.502.12) | 1 | 0.87 (0.601.27) | 0.94 (0.651.36) | 0.96 (0.641.45) | 0.88 (0.561.38) | 0.82 | 0.99 (0.921.08) |
One 8 oz cup=237 grams of coffee
Adjusted for age (continuous), BMI (kg/m2, continuous), level of education (high school or less, greater than high school), years of smoking (continuous), cigarettes smoked per day (continuous), alcohol (grams/day, continuous), history of diabetes (yes or no), frequency of physical activity in leisure time (<1/week, 1–2/week, ≥3/week), Energy (kcal/day, continuous), nutrient density adjusted fruit consumption (grams/day, continuous), nutrient density adjusted vegetable consumption (grams/day, continuous)
Additionally adjusted for tea consumption (drinker, non-drinker)
Additionally adjusted for presence of hypertension
Table 3.
Hazard ratios (95% Confidence Intervals) for cancer of urinary system according to tea consumption in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study
| Model adjustment | Tea consumption (cups/day)1 | ||||
|---|---|---|---|---|---|
| Non-drinker | <1 | >1 | P trend | 1 cup/day increase | |
| Bladder cancer | |||||
| No. of cases/ person-years | 553/297,629 | 223/132,968 | 59/42,441 | ||
| Unadjusted model | 1 | 0.88 (0.76–1.03) | 0.73 (0.56–0.96) | 0.02 | 0.90 (0.80–1.01) |
| Multivariable adjusted | 1 | 0.91 (0.78–1.07) | 0.77 (0.58–1.00) | 0.046 | 0.92 (0.82–1.03) |
| Multivariable adjusted | 1 | 0.92 (0.78–1.08) | 0.77 (0.58–1.02) | 0.07 | 0.92 (0.82–1.04) |
| Renal cell carcinoma | |||||
| No. of cases/ person-years | 226/292,468 | 109/130,788 | 31/42,678 | ||
| Unadjusted model | 1 | 1.07 (0.85–1.35) | 0.96 (0.66–1.39) | 0.86 | 0.99 (0.85–1.15) |
| Multivariable adjusted2 | 1 | 1.11 (0.88–1.40) | 1.00 (0.68–1.46) | 0.95 | 1.01 (0.86–1.17) |
| Multivariable adjusted3 | 1 | 1.10 (0.87–1.40) | 0.97 (0.65–1.44) | 0.92 | 0.99 (0.84–1.16) |
One 8 oz cup=237 grams of tea
Adjusted for age (continuous), BMI (kg/m2, continuous), level of education (high school or less, greater than high school), years of smoking (continuous), cigarettes smoked per day (continuous), alcohol (grams/day, continuous), tea consumption (drinker, non-drinker), history of diabetes (yes or no), frequency of physical activity in leisure time (<1/week, 1–2/week, ≥3/week), Energy (kcal/day, continuous), nutrient density adjusted fruit consumption (grams/day, continuous), nutrient density adjusted vegetable consumption (grams/day, continuous)
Additionally adjusted for coffee consumption (cups/day)
In multivariable adjusted models, we observed no statistically significant associations for either coffee drinking (HR ≥4 vs >0 to <1 cups/day = 0.85, 95% CI=0.55–1.32, P trend=0.68, Table 2), or tea drinking (HR ≥1 vs 0 cup/day=1.00, 95% CI=0.68–1.46, P trend=0.95, Table 3) with RCC in multivariable adjusted models. In analyses of continuous coffee and tea consumption, we observed no statistically significant associations between either coffee (HR1 cup/day increase =0.99; 95% CI 0.92–1.07) or tea (HR1 cup/day increase=1.01; 95% CI 86–1.17) consumption and RCC risk. Additional adjustment for occupation or randomization group did not substantially alter risk estimates in analyses of coffee or tea (data not shown).
Smoking status, defined using pack-years, did not modify coffee- or tea-bladder cancer associations (P interaction=0.66 and 0.73, respectively, Table 4).
Table 4.
Hazard ratios (95% Confidence Intervals) for bladder cancer according to coffee consumption in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study stratified by smoking status (N=26,841)
| Model adjustment | Coffee (cups/day)1 | Tea (cups/day)1 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Non drinker | <1 | >1 | >2 | >3 | P trend | P interaction | Non drinker | <1 | >1 | P trend | P interaction | |
| <35 pack-years | 0.66 | 0.73 | ||||||||||
| No of cases | 9 | 34 | 115 | 125 | 107 | 249 | 109 | 32 | ||||
| Multivariable adjusted1 | 1.45 (0.69 3.02) | 1 | 1.33 (0.91 1.96) | 1.36 (0.92 2.00) | 1.44 (0.97 2.15) | 0.20 | 1 | 0.85 (0.68 1.07) | 0.79(0.55 1.15) | 0.19 | ||
| >35 pack-years | ||||||||||||
| No of cases | 8 | 46 | 99 | 131 | 161 | 304 | 114 | 27 | ||||
| Multivariable adjusted2 | 0.75 (0.35 1.59) | 1 | 0.82 (0.28 1.17) | 0.83 (0.29 1.17) | 0.90 (0.53 1.26) | 0.91 | 1 | 0.96 (0.77 1.20) | 0.73 (0.49 1.09) | 0.13 | ||
One 8 oz cup =237 grams of coffee or tea
Adjusted for age (continuous), BMI (kg/m2, continuous), level of education (high school or less, greater than high school), years of smoking (continuous), cigarettes smoked per day (continuous), alcohol (grams/day, continuous), tea consumption (drinker, non-drinker), history of diabetes (yes or no), frequency of physical activity in leisure time (<1/week, 1–2/week, ≥3/week), Energy (kcal/day, continuous), nutrient density adjusted fruit consumption (grams/day, continuous), nutrient density adjusted vegetable consumption (grams/day, continuous)
Next, we considered the impact of coffee preparation method on the association between coffee and cancers of the urinary tract (Table 5). Among 20,603 men with information on preparation method, 71% drank filtered coffee, 21% drank boiled coffee, and 8% drank instant coffee or were not coffee drinkers. Baseline characteristics of the subset of men with data on preparation method were similar to the overall cohort 20. Overall, adjustment for coffee preparation method did not alter the risk estimates for bladder cancer significantly (P for likelihood ratio test =0.71). We found no statistical evidence of effect modification by preparation method (P interaction= 0.15) (Table 5). For the RCC analysis, adjustment for coffee preparation method substantially altered risk estimates (i.e., >10% change in HR estimates). However, we found no statistical evidence of effect modification by preparation method for RCC analysis (P interaction= 0.66; Table 5). Finally, restricting the analysis to those who were followed for at least two years did not meaningfully alter HR estimates (data not shown).
Table 5.
Hazard ratios (95% Confidence Intervals) for cancer of urinary system according to coffee consumption in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study among those with information on coffee preparation method (N=20,603)
| Coffee (cups/day)1 | ||||||
|---|---|---|---|---|---|---|
| <1 | >1 | >2 | >3 | P trend | P interaction | |
| Bladder cancer | ||||||
| All methods of preparing coffee | 0.15 | |||||
| No of cases | 67 | 174 | 215 | 229 | ||
| Multivariable adjusted2 | 1 | 0.99 (0.74–1.31) | 1.02 (0.78–1.36) | 1.11 (0.84–1.48) | 0.18 | |
| Multivariable adjusted3 | 1 | 0.98 (0.73–1.30) | 1.01 (0.77–1.34) | 1.10 (0.83–1.47) | 0.31 | |
| Filtered method of preparing coffee | ||||||
| No of cases | 51 | 129 | 160 | 171 | ||
| Multivariable adjusted2 | 1 | 0.83 (0.60–1.16) | 0.89 (0.64–1.22) | 0.99 (0.72–1.38) | 0.62 | |
| Boiled method of preparing coffee | ||||||
| No of cases | 9 | 34 | 41 | 49 | 0.65 | |
| Multivariable adjusted2 | 1 | 1.45 (0.69–3.04) | 1.28 (0.61–2.66) | 1.36 (0.65–2.84) | ||
| Renal cell carcinoma | ||||||
| All methods of preparing coffee | 0.66 | |||||
| No of cases | 32 | 72 | 94 | 80 | ||
| Multivariable adjusted2 | 1 | 0.86 (0.56–1.30) | 0.95 (0.63–1.43) | 0.81 (0.53–1.24) | 0.46 | |
| Multivariable adjusted3 | 1 | 0.98 (0.63–1.54) | 1.10 (0.71–1.70) | 0.95 (0.60–1.49) | 0.84 | |
| Filtered method of preparing coffee | ||||||
| No of cases | 20 | 57 | 79 | 62 | ||
| Multivariable adjusted2 | 1 | 0.95 (0.57–1.59) | 1.14 (0.69–1.88) | 0.90 (0.53–1.51) | 0.97 | |
| Boiled method of preparing coffee | ||||||
| No of cases | 4 | 9 | 10 | 13 | ||
| Multivariable adjusted2 | 1 | 0.90 (0.27–2.95) | 0.76 (0.23–2.49) | 0.93 (0.29–3.03) | 0.88 | |
One 8 oz cup=237 grams of coffee
Adjusted for age (continuous), BMI (kg/m2, continuous), level of education (high school or less, greater than high school), years of smoking (continuous), cigarettes smoked per day (continuous), alcohol (grams/day, continuous), history of diabetes (yes or no), frequency of physical activity in leisure time (<1/week, 1–2/week, ≥3/week), Energy (kcal/day, continuous), nutrient density adjusted fruit consumption (grams/day, continuous), nutrient density adjusted vegetable consumption (grams/day, continuous)
Additionally adjusted for coffee preparation method
Discussion
In this prospective cohort study of male Finnish smokers, we found no evidence of an association between coffee consumption and bladder cancer risk, and method of preparation did not appear to modify estimates. We observed an inverse association between tea and bladder cancer risk that approached statistical significance but found no evidence of an association for coffee or tea drinking with RCC risk.
For bladder cancer, our results are consistent with other prospective studies such as the Prostate, Lung, Colorectal, and Ovarian cancer screening trial (PLCO) Study and the NIH-AARP Study, which found no association between coffee drinking and risk of bladder cancer following careful adjustment for potential confounders, particularly cigarette smoking 6, 22. However, our results are not consistent with the results of two Japanese cohorts, which observed an inverse association between coffee drinking and bladder cancer 23. Coffee drinkers in the Japanese cohorts were more highly educated, were more likely to be employed, and had lower rates of chronic disease compared to the coffee-nondrinkers. Thus, the inverse association with bladder cancer may reflect residual confounding by socioeconomic class or healthy lifestyle. In contrast, ATBC participants were heavy smokers; consequently, a weak inverse association between coffee and bladder cancer may have been masked by residual confounding by smoking. Previously observed adverse associations between coffee drinking and bladder cancer are largely restricted to case-control studies 5, which are more vulnerable to recall and selection bias than cohort studies.
We observed limited evidence for an inverse association between tea drinking and bladder cancer risk. A meta-analysis of eight studies, seven of which were cohort studies, observed a lower risk of bladder cancer among tea drinkers in western countries only 10. However, another meta-analysis of 25 case-control studies and seven cohort studies found no association for tea drinking overall, or by type (i.e., green or black), with bladder cancer 12. Similarly, investigators found no association between tea drinking and risk of bladder cancer in a large US cohort study 22.
For RCC, a pooled analysis of 13 prospective studies found suggestive inverse associations for both coffee and tea drinking with RCC 3. However, a meta-analysis of 16 case-control and 6 cohort studies reported no association between coffee consumption and RCC 24, which is consistent with our findings.
Both coffee and tea contain phenolic compounds with antioxidant potential 25, 26 and caffeine, which may alter cell cycle function or induce apoptosis 27. The inverse, albeit not statistically significant, association between tea and bladder cancer may be due to antioxidants in tea, such as catechins, that are protective against cancer and have been shown to concentrate in the bladder 11. Catechins have been shown to induce apoptosis, inhibit metastases and angiogenesis, and enhance antioxidant effects 11. On the other hand, despite detailed adjustment for smoking and other potential confounders, we cannot rule out residual confounding by unknown or poorly measured risk factors such as arsenic in drinking water 13.
Although the data in this study were collected prospectively, mitigating important sources of bias, this study is observational, and relied on self-reported consumption; therefore, results should be interpreted with caution. Furthermore, our study assessed diet and lifestyle factors at baseline; thus, we were unable to account for changes in these exposures over the course of follow-up. Coffee and tea consumption were assessed by self-report using an FFQ, which although susceptible to measurement-error, has been shown to capture coffee consumption reliably 19, 20. Finally, our study included only male smokers, which is a population at high risk of bladder cancer; however, our results may have limited generalizability. The strengths of this study include the large sample size, the prospective design with extended follow-up and rigorous case ascertainment, the high prevalence of coffee drinking and data on method of coffee preparation, and the collection of detailed data on smoking and other potential confounders.
In summary, we found no association between coffee and tea consumption with RCC. Coffee consumption was also not associated with bladder cancer, although a borderline association between tea drinking and bladder cancer warrants further investigation in diverse populations with higher tea consumption.
Acknowledgment
The authors express their gratitude to the participants in the ATBC study for their corporation.
Funding: This work was supported by the Intramural Research Program of the National Institutes of Health (NIH), Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, Department of Health and Human Services.
Abbreviations used
- (FFQ)
food-frequency-questionnaire
- (ATBC)
Alpha-Tocopherol Beta-Carotene Cancer Prevention
- (RCC)
Renal Cell Carcinoma
Footnotes
Conflict of interest: none declared
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Sang LX, Chang B, Li XH, Jiang M. Consumption of coffee associated with reduced risk of liver cancer: a meta-analysis. BMC Gastroenterol 2013;13: 34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Je Y, Giovannucci E. Coffee consumption and risk of endometrial cancer: findings from a large up-to-date meta-analysis. Int J Cancer 2012;131: 1700–10. [DOI] [PubMed] [Google Scholar]
- 3.Lee JE, Hunter DJ, Spiegelman D, Adami HO, Bernstein L, van den Brandt PA, Buring JE, Cho E, English D, Folsom AR, Freudenheim JL, Gile GG, et al. Intakes of coffee, tea, milk, soda and juice and renal cell cancer in a pooled analysis of 13 prospective studies. Int J Cancer 2007;121: 2246–53. [DOI] [PubMed] [Google Scholar]
- 4.Wu W, Tong Y, Zhao Q, Yu G, Wei X, Lu Q. Coffee consumption and bladder cancer: a meta-analysis of observational studies. Sci Rep 2015;5: 9051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zhou Y, Tian C, Jia C. A dose-response meta-analysis of coffee consumption and bladder cancer. Prev Med 2012;55: 14–22. [DOI] [PubMed] [Google Scholar]
- 6.Loftfield E, Freedman ND, Inoue-Choi M, Graubard BI, Sinha R. A Prospective Investigation of Coffee Drinking and Bladder Cancer Incidence in the United States. Epidemiology 2017;28: 685–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Antoni S, Ferlay J, Soerjomataram I, Znaor A, Jemal A, Bray F. Bladder Cancer Incidence and Mortality: A Global Overview and Recent Trends. Eur Urol 2017;71: 96–108. [DOI] [PubMed] [Google Scholar]
- 8.Cavin C, Holzhaeuser D, Scharf G, Constable A, Huber WW, Schilter B. Cafestol and kahweol, two coffee specific diterpenes with anticarcinogenic activity. Food Chem Toxicol 2002;40: 1155–63. [DOI] [PubMed] [Google Scholar]
- 9.Urgert R, Vanderweg G, Kosmeijerschuil TG, Vandebovenkamp P, Hovenier R, Katan MB. Levels of the Cholesterol-Elevating Diterpenes Cafestol and Kahweol in Various Coffee Brews. J Agr Food Chem 1995;43: 2167–72. [Google Scholar]
- 10.Zhang YF, Xu Q, Lu J, Wang P, Zhang HW, Zhou L, Ma XQ, Zhou YH. Tea consumption and the incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Eur J Cancer Prev 2015;24: 353–62. [DOI] [PubMed] [Google Scholar]
- 11.Conde VR, Alves MG, Oliveira PF, Silva BM. Tea (Camellia sinensis (L.)): a putative anticancer agent in bladder carcinoma? Anticancer Agents Med Chem 2015;15: 26–36. [DOI] [PubMed] [Google Scholar]
- 12.Weng H, Zeng XT, Li S, Kwong JS, Liu TZ, Wang XH. Tea Consumption and Risk of Bladder Cancer: A Dose-Response Meta-Analysis. Front Physiol 2016;7: 693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.World Cancer Research Fund/American Institute for Cancer Research. Diet N, Physical Activity and Cancer: a Global Perspective. Continuous Update Project Report 2018. Available at dietaandcancerreport.org.
- 14.Ministry of Foreign Affaires; CBI Trade statistics: coffee. Available online: https://www.cbi.eu/sites/default/files/market.../trade-statistics-coffee-2016.pdf (accessed on 7 May 2018); 2016.
- 15.The alpha-tocopherol, beta-carotene lung cancer prevention study: design, methods, participant characteristics, and compliance. The ATBC Cancer Prevention Study Group. Ann Epidemiol 1994;4: 1–10. [DOI] [PubMed] [Google Scholar]
- 16.Kyllonen LE, Teppo L, Lehtonen M. Completeness and accuracy of registration of colorectal cancer in Finland. Ann Chir Gynaecol 1987;76: 185–90. [PubMed] [Google Scholar]
- 17.Leinonen MK, Miettinen J, Heikkinen S, Pitkaniemi J, Malila N. Quality measures of the population-based Finnish Cancer Registry indicate sound data quality for solid malignant tumours. Eur J Cancer 2017;77: 31–9. [DOI] [PubMed] [Google Scholar]
- 18.Korhonen P, Malila N, Pukkala E, Teppo L, Albanes D, Virtamo J. The Finnish Cancer Registry as follow-up source of a large trial cohort-accuracy and delay. Acta Oncol 2002;41: 381–8. [DOI] [PubMed] [Google Scholar]
- 19.Pietinen P, Hartman AM, Haapa E, Rasanen L, Haapakoski J, Palmgren J, Albanes D, Virtamo J, Huttunen JK. Reproducibility and validity of dietary assessment instruments. I. A self-administered food use questionnaire with a portion size picture booklet. Am J Epidemiol 1988;128: 655–66. [DOI] [PubMed] [Google Scholar]
- 20.Lai GY, Weinstein SJ, Albanes D, Taylor PR, McGlynn KA, Virtamo J, Sinha R, Freedman ND. The association of coffee intake with liver cancer incidence and chronic liver disease mortality in male smokers. Br J Cancer 2013;109: 1344–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cumberbatch MG, Cox A, Teare D, Catto JW. Contemporary Occupational Carcinogen Exposure and Bladder Cancer: A Systematic Review and Meta-analysis. JAMA Oncol 2015;1: 1282–90. [DOI] [PubMed] [Google Scholar]
- 22.Hashibe M, Galeone C, Buys SS, Gren L, Boffetta P, Zhang ZF, La Vecchia C. Coffee, tea, caffeine intake, and the risk of cancer in the PLCO cohort. Br J Cancer 2015;113: 809–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sugiyama K, Sugawara Y, Tomata Y, Nishino Y, Fukao A, Tsuji I. The association between coffee consumption and bladder cancer incidence in a pooled analysis of the Miyagi Cohort Study and Ohsaki Cohort Study. Eur J Cancer Prev 2017;26: 125–30. [DOI] [PubMed] [Google Scholar]
- 24.Wijarnpreecha K, Thongprayoon C, Thamcharoen N, Panjawatanan P, Cheungpasitporn W. Association between coffee consumption and risk of renal cell carcinoma: a meta-analysis. Intern Med J 2017;47: 1422–32. [DOI] [PubMed] [Google Scholar]
- 25.Conde VR, Alves MG, Oliveira PF, Silva BM. Tea (Camellia Sinensis (L.)): a Putative Anticancer Agent in Bladder Carcinoma? Anticancer Agents Med Chem 2014. [DOI] [PubMed] [Google Scholar]
- 26.Correa TA, Monteiro MP, Mendes TM, Oliveira DM, Rogero MM, Benites CI, Vinagre CG, Mioto BM, Tarasoutchi D, Tuda VL, Cesar LA, Torres EA. Medium light and medium roast paper-filtered coffee increased antioxidant capacity in healthy volunteers: results of a randomized trial. Plant Foods Hum Nutr 2012;67: 277–82. [DOI] [PubMed] [Google Scholar]
- 27.Bode AM, Dong Z. The enigmatic effects of caffeine in cell cycle and cancer. Cancer Lett 2007;247: 26–39. [DOI] [PMC free article] [PubMed] [Google Scholar]