Abstract
Background: Abundant pre-clinical data suggest that consumption of cruciferous vegetables might protect against bladder cancer. While small-scale clinical evidence supports this hypothesis, population-level data is lacking. We tested the hypothesis that consumption of cruciferous vegetables is associated with a lower risk of bladder cancer in a large population-based study. Methods: We investigated the association between dietary consumption of cruciferous vegetables and the risk of bladder cancer in the NIH-American Association of Retired Persons (AARP) Diet and Health Study. Diet at baseline was collected with self-administered food-frequency questionnaires. Bladder cancer diagnoses were identified through linkage with state cancer registries. Hazard ratio (HR) and 95% confidence intervals (CI) were estimated with Cox proportional hazards models. Results: Our analysis included 515,628 individuals. Higher intake of cruciferous vegetables, both overall and when stratified by variety (broccoli vs. brussels sprouts vs. cauliflower), were not associated with bladder cancer risk for men or women. A history of smoking did not affect the results. Conclusions: Our study shows no association between dietary consumption of cruciferous vegetables and incident bladder cancer.
Keywords: Cruciferous vegetables, broccoli, bladder cancer, NIH-AARP, diet
Introduction
Tremendous interest exists in dietary interventions to reduce the development of cancer. Cruciferous vegetables are a family of vegetables that includes broccoli, cabbage, brussels sprouts, collard greens, kale, cauliflower, and turnips. In the past decade, studies have linked the consumption of cruciferous vegetables to reduced risk of cardiovascular mortality and development of Type II diabetes [1-3]. Based on promising pre-clinical data, cruciferous vegetables have emerged as a dietary substance that might also protect against cancer. Various studies, including: cross-sectional, case control, cohort, and prospective studies, also suggest that consumption of cruciferous vegetables may protect against development of several cancers, including: breast, lung, pancreatic, and prostate cancer [4-8].
Pre-clinical studies suggest that cruciferous vegetables may also protect against bladder cancer. Specifically, cruciferous vegetables are rich in isothiocyanates, compounds that have been shown to have anticarcinogenic effects against bladder cancer in animal models and in vitro studies [9-15]. Pre-clinical studies utilizing isothiocyanate in rat bladder cancer models demonstrated inhibition of bladder cancer development and progression [15]. Similarly, several other in vitro and in vivo studies have supported these results [16-18].
However, epidemiological studies intended to detect an association between cruciferous vegetable intake and bladder cancer in patients has shown conflicting results. Several small cohort and case control studies have found that cruciferous vegetables are associated with a decreased risk for bladder cancer [9,19,20], while others studies have found no association [9,21,22]. However, these studies were limited by relatively small sample sizes, thus lacking statistical power to detect an association between cruciferous vegetable consumption and bladder cancer risk.
We sought to clarify the conflicting evidence in the literature by exploring the relationship between cruciferous vegetable intake and bladder cancer based in a large United States population. We used data from the NIH-AARP Diet and Health Study, which provides dietary data on a prospective cohort including over 500,000 subjects. We tested whether consumption of the cruciferous vegetables broccoli, brussels sprouts, and cauliflower was associated with the risk of incident bladder cancer.
Methods
Study population
The NIH-American Association of Retired Persons (AARP) Diet and Health Study, initiated in 1995-1996, includes subjects (age 50-71) who lived in California, Florida, Pennsylvania, New Jersey, North Carolina, Louisiana, or in the cities of Atlanta or Detroit. After obtaining informed consent, 3.5 million AARP members were mailed baseline dietary questionnaires of which 567,169 (16.2%) were completed. The Special Studies Institutional Review Board of the US National Cancer Institute has reviewed and approved the NIH-AARP Diet and Health study. The data included in this study was obtained after formal review and approval of the research by the NIH-AARP Diet and Health Study (# 201410-0010). All data was de-identified by the NIH-AARP Diet and Health Study prior to provision to our site and was therefore ruled exempt from local IRB review.
Information on cancer stage was obtained from state cancer registries. For bladder cancer, we identified in situ, localized, low grade, high grade, advanced, and fatal bladder cancer. Fatal cases were those who died of bladder cancer during follow up.
From the 567,169 respondents who returned the baseline dietary questionnaire, we excluded those who withdrew from the study, those whose questionnaires were completed by proxy respondents, participants who had submitted duplicate questionnaires and participants with cancer diagnosed prior to study entry. After these exclusions, 515,628 individuals were included in our analysis.
Diet ascertainment
A self-administered food-frequency questionnaire (FFQ) was used to assess baseline diet based on reporting of 124 foods, including 23 vegetables or their derivatives, during the previous year. In regard to broccoli, brussels sprouts and cauliflower, patients were asked the frequency of their consumption using 10 predefined categories of intake, ranging from “never” to “2+ times per day”, as well as their portion sizes, from “less than ¼ cup” to “more than 1 cup”. The FFQ has been previously validated using 2 subsequent, 24-hour recalls in a subset of the cohort [7,23,24].
Cancer ascertainment
Cases of incident bladder cancer were identified through December 31, 2011, through linkage of the NIH-AARP cohort database the 8 state cancer registry databases and the National Death Index Plus. The registries are certified to have >90% case ascertainment within 24 months of cancer diagnosis by the North American Association of Central Cancer Registries. Participants were tracked annually by matching with the National Change of Address database and through processing of undeliverable mail by the U.S. Postal Service, and through direct responses from participants and other address update services. Deaths from bladder cancer was ascertained through December 30, 2011 by linkage with the National Death Index Plus and the Social Security Administration death master file [25-27].
Statistical analysis
We first examined the normality of all variables. Descriptive statistics were presented using Mean ± SD. Anthropometric parameters were compared among the quintiles using Student’s t-test, Wilcoxon Signed Rank, or Chi-Square as appropriate. Within each quintile, univariate and multivariate Cox-proportional hazard models were performed. Hazard ratios were displayed with 95% Confidence Interval for different types of bladder cancers, adjusted by marital status, education levels, race, sex, body mass index (BMI), entry age, history of smoking and any first-degree relatives with cancer, and dietary-related variables specifically food energy, total meat consumption, and daily fiber (CSFII) from vegetables. All statistical tests were two-sided and p-values < 0.05 were considered significant. Statistical analysis was performed using SAS, version 9.4 (SAS Institute, Inc., Cary, NC, USA).
Results
Cohort description
Our cohort included 515,628 individuals (312,461 men and 203,167 women). The cohort reported consuming 0.27 ± 0.34 servings of cruciferous vegetables per day. 8257 reported no consumption. Age, marital status, race, and income were similar across quintiles of cruciferous vegetables. Similarly, BMI, smoking status, and level of physical activity was similar across groups (Table 1).
Table 1.
Baseline characteristics of people with bladder cancer in the NIH-AARP diet and health study
| Total Participants: 515,628 | Quintiles of Total Cruciferous Vegetables Intake | ||||
|---|---|---|---|---|---|
|
| |||||
| 1 | 2 | 3 | 4 | 5 | |
| By quintile n (%) | 112,853 (21.9) | 96,858 (18.8) | 99,356 (19.3) | 103,921 (20.2) | 102,640 (19.9) |
| Gender n (%) | |||||
| Female | 36,209 (17.8) | 35,931 (17.7) | 39,624 (19.6) | 43,773 (21.6) | 47,630 (23.4) |
| Male | 76,644 (24.5) | 60,927 (19.5) | 59,732 (19.1) | 60,148 (19.3) | 55,010 (17.6) |
| Age mean ± SD | 62 ± 5 | 62 ± 6 | 62 ± 5 | 62 ± 5 | 62 ± 5 |
| Age at 1/1/1985 mean ± SD | 50 ± 6 | 50 ± 5 | 50 ± 5 | 50 ± 5 | 50 ± 5 |
| BMI mean ± SD | 27.2 ± 5.0 | 27.2 ± 5.0 | 27.1 ± 5.0 | 27.1 ± 5.1 | 27.1 ± 5.4 |
| BMI n (%) | |||||
| Below Normal | 1,223 (24.5) | 884 (17.7) | 860 (17.2) | 988 (19.8) | 1,046 (20.9) |
| Normal Weight | 35,134 (20.9) | 30,958 (18.4) | 32,675 (19.4) | 34,743 (20.6) | 34,884 (20.7) |
| Overweight | 49,207 (22.5) | 41,946 (19.2) | 42,405 (19.4) | 43,588 (20.0) | 41,184 (18.9) |
| Obese | 22,233 (21.9) | 19,245 (18.9) | 19,460 (19.2) | 20,247 (19.9) | 20,348 (20.0) |
| Severely Obese | 1,935 (20.0) | 1,673 (17.3) | 1,763 (18.2) | 1,967 (20.3) | 2,359 (24.3) |
| Birth Cohort n (%) | |||||
| 1925-1930 | 38,342 (22.5) | 32,387 (18.9) | 32,447 (18.9) | 34,666 (20.2) | 33,481 (19.5) |
| 1931-1935 | 31,301 (21.2) | 27,897 (18.9) | 28,723 (19.5) | 29,868 (20.3) | 29,736 (20.2) |
| 1936-1940 | 25,687 (21.4) | 22,511 (18.8) | 23,291 (19.4) | 24,277 (20.2) | 24,158 (20.1) |
| 1941-1945 | 17,523 (22.8) | 14,063 (18.3) | 14,895 (19.4) | 15,110 (19.7) | 15,265 (19.9) |
| Marital Status n (%) | |||||
| Married | 77,559 (21.8) | 68,488 (19.2) | 70,660 (19.8) | 73,151 (20.5) | 66,754 (18.7) |
| Widowed | 11,663 (21.2) | 9,967 (18.1) | 10,190 (18.5) | 11,030 (20.0) | 12,307 (22.3) |
| Divorced | 15,223 (22.0) | 12,482 (18.0) | 12,576 (18.1) | 13,175 (19.0) | 15,876 (22.9). |
| Separated | 1,426 (23.5) | 1,077 (17.7) | 1,052 (17.3) | 1,127 (18.6) | 1,390 (22.9). |
| Never Married | 5,595 (23.2) | 4,167 (17.3) | 4,230 (17.5) | 4,723 (19.6) | 5,413 (22.4) |
| Education n (%) | |||||
| < 12 years | 9,428 (28.7) | 6,243 (19.0) | 5,588 (17.0) | 5,725 (17.4) | 5,861 (17.8) |
| High school/ Some College | 63,180 (23.3) | 52,358 (19.3) | 51,417 (19.0) | 53,248 (19.6) | 50,869 (18.8) |
| Completed college or more | 36,345 (18.5) | 35,641 (18.2) | 39,717 (20.2) | 42,118 (21.5) | 42,478 (21.6) |
| Race n (%) | |||||
| Non-Hispanic White | 102,786 (21.9) | 89,526 (19.1) | 91,747 (19.5) | 95,453 (20.3) | 90,075 (19.2) |
| Non-Hispanic Black | 3,143 (15.5) | 3,305 (16.2) | 3,443 (16.9) | 4,044 (19.9) | 6,410 (31.5) |
| Hispanic | 3,376 (34.0) | 1,622 (16.3) | 1,565 (15.8) | 1,479 (14.9) | 1,897 (19.1) |
| Asian/Pacific Islander | 995 (15.5) | 913 (14.2) | 1,147 (17.8) | 1,318 (20.5) | 2,069 (32.1) |
| Pacific Islander | 115 (18.1) | 108 (17.0) | 106 (16.7) | 114 (18.0) | 191 (30.1) |
| American Indian/Alaskan Native | 350 (23.3) | 266 (17.7) | 224 (14.9) | 283 (18.9) | 377 (25.1) |
| Unknown | 2,088 (29.1) | 1,118 (15.6) | 1,124 (15.7) | 1,230 (17.1) | 1,621 (22.6) |
| Median Household Income (dollars) | 47,328 | 48,071 | 49,265 | 48,992 | 48,664 |
| Smoking status n (%) | |||||
| Never | 36,925 (20.5) | 32,342 (18.0) | 35,283 (20.0) | 37,723 (21.0) | 37,842 (21.0). |
| Former | 53,826 (21.2) | 48,458 (19.2) | 49,159 (19.4) | 51,325 (20.3) | 50,337 (20.0) |
| Current | 17,075 (27.4) | 12,548 (20.2) | 11,282 (18.1) | 11,138 (17.9) | 10,211 (16.4) |
| Physical activity n (%) | |||||
| < 3 time/month | 45,047 (27.5) | 33,199 (20.3) | 30,750 (18.8) | 29,009 (17.7) | 25,753 (15.7) |
| 1-2 times/week | 22,969 (20.8) | 21,481 (19.5) | 22,318 (20.2) | 23,058 (20.9) | 20,418 (18.5) |
| 3-4 times/week | 25,057 (18.3) | 24,868 (18.2) | 27,172 (19.9) | 29,691 (21.7) | 29,829 (21.8) |
| ≥ 5 times/week | 17,892 (18.1) | 16,353 (16.5) | 18,219 (18.4) | 21,190 (21.4) | 25,397 (25.6) |
| Other Dietary Factors mean ± SD | |||||
| Food energy (kcal) | 1,666 ± 941 | 1,729 ± 816 | 1,836 ± 834 | 1,941 ± 860 | 2,247 ± 1345 |
| Daily fiber from vegetables (g/day) | 3.8 ± 2.7 | 4.7 ± 2.5 | 5.7 ± 2.6 | 7.1 ± 2.9 | 11.5 ± 7.1 |
| Total meat (g/day) | 109 ± 96 | 116 ± 80 | 126 ± 109 | 136 ± 90 | 157 ± 162 |
Abbreviations: n: number; SD: Standard Deviation; kcal: kilocalories; g: grams.
Bladder cancer incidence and cruciferous vegetable consumption
A total of 8,567 cases of incident bladder were identified with 7,313 (85.4%) in men and 1,254 (14.6%) in women) during 15 years of follow up (Table 2). In situ carcinoma accounted for 2,610 (30.5%) of cases, localized carcinoma for 2,259 (26.4%) of cases, low grade carcinoma for 3,907 (45.6%) of cases, high grade carcinoma for 3,264 (38.1%) of cases, advanced carcinoma for 439 (5.12%) and fatal carcinoma for 1,233 (14.4%) of cases. It is important to note that one individual diagnosis may be overlapping in several bladder cancer sub categories (for example, the cancer was both low grade and in situ). We found no significant associations between cruciferous vegetable consumption and bladder cancer. Specifically, there was no trend for bladder cancer incidence across quintiles from very low to high cruciferous vegetable consumption. While confidence intervals for individual quintiles in bladder cancer subsets suggest significant associations, the absence of a dose-response relationship is shown in the non-significant p-values across the quintiles arguing against biological effects. In addition, no associations were seen when examining individual cruciferous vegetables or by bladder cancer type (Table 2).
Table 2.
Hazard ratio for bladder cancer in relation to consumption of cruciferous vegetables stratified by quintile of reported consumption
| Broccoli | Cauliflower/Brussel Sprouts | Composite/Total Cruciferous | |||||
|---|---|---|---|---|---|---|---|
|
|
|
|
|||||
| n | Fully Adjusted HR (CI) | n | Fully Adjusted HR (CI) | n | Fully Adjusted HR (CI) | ||
| All Bladder Cancer | Quintile 1 | 2,081 | ref | 1,920 | ref | 2,016 | ref |
| Quintile 2 | 1,362 | 0.991 (0.920, 1.066) | 1,478 | 1.007 (0.936, 1.084) | 1,714 | 1.022 (0.953, 1.095) | |
| Quintile 3 | 1,933 | 1.007 (0.941, 1.077) | 1,800 | 1.003 (0.936, 1.076) | 1,613 | 0.986 (0.917, 1.060) | |
| Quintile 4 | 2,058 | 1.024 (0.956, 1.097) | 1,756 | 1.019 (0.949, 1.094) | 1,667 | 1.021 (0.949, 1.100) | |
| Quintile 5 | 1,133 | 1.083 (0.989, 1.185) | 1,613 | 1.064 (0.984, 1.149) | 1,557 | 1.082 (0.990, 1.182) | |
| p trend (p value) | 0.214 | 0.479 | 0.262 | ||||
| High Grade Bladder Carcinoma | Quintile 1 | 798 | ref | 770 | ref | 785 | ref |
| Quintile 2 | 543 | 1.026 (0.911, 1.154) | 545 | 0.927 (0.824, 1.042) | 679 | 1.066 (0.953, 1.191) | |
| Quintile 3 | 739 | 1.025 (0.919, 1.144) | 707 | 0.97 (0.868, 1.085) | 589 | 0.955 (0.849, 1.074) | |
| Quintile 4 | 789 | 1.037 (0.927, 1.160) | 620 | 0.911 (0.811, 1.023) | 615 | 1.002 (0.887, 1.131) | |
| Quintile 5 | 395 | 1.068 (0.920, 1.239) | 622 | 1.053 (0.930, 1.193) | 596 | 1.184 (1.025, 1.367) | |
| p trend (p value) | 0.754 | 0.550 | 0.224 | ||||
| Low Grade Bladder Carcinoma | Quintile 1 | 946 | ref | 844 | ref | 898 | ref |
| Quintile 2 | 601 | 0.974 (0.873, 1.088) | 683 | 1.045 (0.937, 1.165) | 854 | 0.99 (0.891, 1.099) | |
| Quintile 3 | 884 | 0.993 (0.898, 1.098) | 811 | 1.03 (0.928, 1.144) | 762 | 1.028 (0.925, 1.143) | |
| Quintile 4 | 929 | 1.01 (0.913, 1.119) | 852 | 1.105 (0.995, 1.226) | 787 | 1.05 (0.943, 1.170) | |
| Quintile 5 | 547 | 1.04 (0.912, 1.187) | 717 | 1.029 (0.916, 1.155) | 706 | 0.995 (0.874, 1.133) | |
| p trend (p value) | 0.729 | 0.762 | 0.754 | ||||
| Fatal Bladder Carcinoma | Quintile 1 | 333 | ref | 279 | ref | 314 | ref |
| Quintile 2 | 195 | 0.877 (0.724, 1.064) | 207 | 1.002 (0.826, 1.216) | 241 | 0.95 (0.791, 1.141) | |
| Quintile 3 | 268 | 0.873 (0.731, 1.043) | 283 | 1.076 (0.897, 1.291) | 220 | 0.901 (0.744, 1.092) | |
| Quintile 4 | 279 | 0.893 (0.745, 1.071) | 226 | 0.911 (0.750, 1.107) | 237 | 0.984 (0.809, 1.197) | |
| Quintile 5 | 158 | 1.023 (0.808, 1.296) | 238 | 1.146 (0.934, 1.406) | 221 | 1.103 (0.873, 1.394) | |
| p trend (p value) | 0.599 | 0.874 | 0.896 | ||||
| Localized Bladder Carcinoma | Quintile 1 | 594 | ref | 553 | ref | 554 | ref |
| Quintile 2 | 372 | 0.992 (0.863, 1.141) | 377 | 0.85 (0.738, 0.979) | 460 | 1.056 (0.924, 1.206) | |
| Quintile 3 | 481 | 0.911 (0.798, 1.040) | 453 | 0.873 (0.763, 0.998) | 398 | 0.913 (0.791, 1.053) | |
| Quintile 4 | 534 | 0.995 (0.872, 1.136) | 442 | 0.888 (0.774, 1.018) | 407 | 0.985 (0.851, 1.140) | |
| Quintile 5 | 278 | 0.998 (0.836, 1.191) | 434 | 1.023 (0.883, 1.184) | 440 | 1.234 (1.041, 1.462) | |
| p trend (p value) | 0.491 | 0.491 | 0.059 | ||||
| Advanced Bladder Carcinoma | Quintile 1 | 105 | ref | 113 | ref | 105 | ref |
| Quintile 2 | 87 | 1.423 (1.044, 1.941) | 63 | 0.743 (0.534, 1.034) | 97 | 1.157 (0.858, 1.560) | |
| Quintile 3 | 92 | 1.1 (0.807, 1.500) | 96 | 0.973 (0.726, 1.303) | 71 | 0.943 (0.681, 1.307) | |
| Quintile 4 | 103 | 1.233 (0.903, 1.682) | 87 | 0.932 (0.688, 1.262) | 97 | 1.251 (0.912, 1.715) | |
| Quintile 5 | 52 | 1.282 (0.854, 1.925) | 80 | 0.93 (0.663, 1.306) | 69 | 1.07 (0.718, 1.594) | |
| p trend (p value) | 0.634 | 0.620 | 0.785 | ||||
| In Situ Bladder Carcinoma | Quintile 1 | 646 | ref | 573 | ref | 593 | ref |
| Quintile 2 | 403 | 0.976 (0.854, 0.115) | 466 | 1.043 (0.915, 1.189) | 540 | 1.067 (0.941, 1.210) | |
| Quintile 3 | 599 | 1.02 (0.903, 1.151) | 542 | 0.983 (0.866, 1.116) | 501 | 1.031 (0.906, 1.174) | |
| Quintile 4 | 603 | 0.968 (0.854, 1.098) | 553 | 1.052 (0.926, 1.195) | 512 | 1.035 (0.906, 1.184) | |
| Quintile 5 | 359 | 1.125 (0.958, 1.322) | 476 | 1.007 (0.874, 1.161) | 464 | 1.04 (0.886, 1.222) | |
| p trend (p value) | 0.271 | 0.722 | 0.764 | ||||
Abbreviations: n: number; HR: Hazard Ratio; CI: Confidence Interval; ref: reference.
Smoking history, bladder cancer risk and cruciferous vegetable consumption
Given the association between smoking and bladder cancer, we performed a stratified analysis based on smoking status to determine if the association between cruciferous vegetable consumption and incident bladder cancer changed. For all categories of smoking status, there was no association between cruciferous vegetable consumption and bladder cancer (Table 3).
Table 3.
Risk of Bladder Cancer by smoking status across quintiles of cruciferous vegetable consumption
| Broccoli | Cauliflower/Brussel Sprouts | Composite/Total Cruciferous | |||||
|---|---|---|---|---|---|---|---|
|
|
|
|
|||||
| n | Fully Adjusted HR (CI) | n | Fully Adjusted HR (CI) | n | Fully Adjusted HR (CI) | ||
| Never Smokers | Quintile 1 | 32,680 | ref | 38,345 | ref | 36,925 | ref |
| Quintile 2 | 27,922 | 0.998 (0.833, 1.197) | 31,542 | 1.010 (0.848, 1.203) | 32,342 | 0.968 (0.817, 1.146) | |
| Quintile 3 | 42,096 | 1.024 (0.868, 1.207) | 38,811 | 1.048 (0.888, 1.238) | 35,283 | 0.943 (0.795, 1.119) | |
| Quintile 4 | 47,274 | 1.010 (0.857, 1.191) | 35,174 | 1.090 (0.920, 1.292) | 37,723 | 0.949 (0.797, 1.132) | |
| Quintile 5 | 30,143 | 1.086 (0.880, 1.341) | 36,243 | 1.164 (0.970, 1.397) | 37,842 | 1.102 (0.898, 1.353) | |
| p trend (p value) | 0.663 | 0.493 | 0.421 | ||||
| Former Smokers | Quintile 1 | 51,745 | ref | 52,507 | ref | 53,826 | ref |
| Quintile 2 | 39,951 | 0.941 (0.856, 1.035) | 43,404 | 1.013 (0.924, 1.111) | 48,458 | 1.014 (0.928, 1.108) | |
| Quintile 3 | 59,398 | 1.013 (0.930, 1.103) | 54,076 | 0.981 (0.898, 1.072) | 49,159 | 0.989 (0.903, 1.084) | |
| Quintile 4 | 62,732 | 1.037 (0.951, 1.131) | 51,935 | 1.013 (0.927, 1.108) | 51,325 | 1.019 (0.928, 1.119) | |
| Quintile 5 | 39,279 | 1.054 (0.941, 1.181) | 51,183 | 1.029 (0.933, 1.135) | 50,337 | 1.069 (0.956, 1.195) | |
| p trend (p value) | 0.444 | 0.670 | 0.599 | ||||
| Current Smokers | Quintile 1 | 16,527 | ref | 15,667 | ref | 17,075 | ref |
| Quintile 2 | 11,332 | 1.134 (0.972, 1.323) | 10,714 | 0.982 (0.834, 1.156) | 12,548 | 1.087 (0.934, 1.266) | |
| Quintile 3 | 14,905 | 0.958 (0.822, 1.117) | 12,828 | 1.037 (0.888, 1.212) | 11,282 | 1.003 (0.851, 1.182) | |
| Quintile 4 | 12,135 | 0.972 (0.824, 1.145) | 11,911 | 0.969 (0.823, 1.140) | 11,138 | 1.095 (0.925, 1.296) | |
| Quintile 5 | 7,355 | 1.203 (0.970, 1.491) | 11,134 | 1.089 (0.909, 1.303) | 10,211 | 1.088 (0.882, 1.343) | |
| p trend (p value) | 0.261 | 0.851 | 0.408 | ||||
Abbreviations: n: number; HR: Hazard Ratio; CI: Confidence Interval; ref: reference.
Discussion
Our report represents the largest analysis of cruciferous vegetable consumption and incident bladder cancer to date. We found no significant relationship between cruciferous vegetable consumption and risk of bladder cancer, regardless of the type of vegetable (broccoli, brussels sprouts and cauliflower, or composite) and across all grades and stages of bladder cancer including: in situ, localized, low grade, high grade, and fatal cancer. While several analyses reached statistical significance, the lack of a consistent dose-dependent association argues against a biological explanation. Findings were similar for both men and women, and after stratifying by smoking status.
Prior to our study, evidence from epidemiological studies had been mixed. A small cohort study examining the intake of cruciferous vegetables in 239 patients with bladder cancer showed an inverse relationship between bladder cancer mortality and increased broccoli intake [15]. Another small case-control study similarly found an inverse relationship between the risk of bladder cancer and consumption of cruciferous vegetables [14]. Meta analyses performed across several studies-including prospective cohort and case-control studies-support that cruciferous vegetables are associated with a decreased risk for bladder cancer [9,19,20]. However, these studies were limited by the relatively small number of cases in each study and the limited number of studies available to be included in the meta-analyses. The most recent meta-analysis across 5 cohort and 5 case control studies, totaling in 5,772 individuals, reported a significantly decreased risk of incident bladder cancer in association with cruciferous vegetable intake [19]. However, the decrease in risk was found to be significant only in the case control studies, but not the cohort studies.
Conversely, several other studies showed no association between risk of bladder cancer and cruciferous vegetable consumption [9]. These include a prospective study examining cruciferous vegetables intake and bladder cancer risk in smokers, a prospective population based cohort study of 82,000 people, and a meta-analysis of fruits and vegetables intake [9,21,22,28].
Our study, the largest to date, also showed no relationship between cruciferous vegetable intake and bladder cancer. The major strength of our study is its large size, including over 500,000 participants from diverse areas across the United States. The size of the study population allows detection of association between specific vegetables and Cruciferae broadly if any relationships truly exist with bladder cancer with a high degree of certainty. The large size also allowed for assessments of associations across bladder cancer subtypes, which no previous study has had the power to detect. Our large population and data collection also allowed us to assess whether the effects of cruciferous vegetables was modified by smoking status. Since sulforaphane, a bioactive micronutrient found at high levels in Cruciferae induces enzymes that reduce and inactivate carcinogens [11-13], stratification for smoking status could unveil potential associations. However, smoking status did not modify the lack of association between dietary cruciferous vegetable consumption and bladder cancer. Finally, our study was also a prospective cohort study, which minimizes the risk of recall bias, a common problem in previous dietary assessment studies.
It is important to note that there was a strong scientific rationale for the hypothesis that cruciferous vegetables may offer protection against bladder cancer. Several in vitro and in vivo experimental studies have looked at the potential anticancer properties of cruciferous vegetables. The leading hypothesis was that isothiocyanates-a uniquely abundant metabolite found in cruciferous vegetables-is the bioactive compound responsible for anticancer activity. In vitro studies have found that isothiocyanates can induce apoptosis and cell cycle arrest in human bladder cancer cells [16,18]. In vivo studies have also supported this hypothesis, with long term rat studies showing decrease in incidence and progression of bladder cancer in rats fed with both broccoli sprout extracts, and other studies finding that broccoli isothiocyanates can inhibit established bladder cancer in xenograft tumor models [16-18]. Isothiocyanates’ proposed mechanism of action include: modulation of carcinogen metabolizing enzymes, cell cycle apoptosis, and epigenetic modulation [9,11].
There are several explanations for the conflicting results of preclinical in vitro and vivo experiments and our large-scale human study. In the in vitro and in vivo studies, isothiocyanates or vegetable extracts are delivered directly to cells or animals in their pure form and in a controlled manner. In contrast, epidemiological studies cannot elucidate to this detail of consumption, and we cannot be certain if the quantities of cruciferous vegetables consumed translates to therapeutic concentrations of isothiocyanates, especially after considering bioavailability and metabolism. In addition, Abbaoui et al also note how cruciferous vegetables are consumed can significantly change the amount of isothiocyonates and other bioactives an individual is exposed to [9]. For example, quantities of isothiocyanates in vegetables are significantly reduced by cooking and storage processes [29-31]. Therefore, the amount of isothiocyanates an individual consumes through cruciferous vegetables may not be enough to ever reach therapeutic levels to protect against bladder cancer.
Our study included several noteworthy limitations. First, dietary information was collected at a single point in time and would not capture changes in diet over time. Therefore, we could not assess how cumulative and longitudinal consumption of cruciferous vegetables affect bladder cancer risk. Second, nearly all dietary assessment studies include the possibility of reporting error, where the reported amount of consumption by the individual is not accurate. Third, considering the latency period of bladder cancer, our follow up time of 15 years may not have been sufficient to fully capture an association as some bladder cancer cases may have not yet had time to manifest or be diagnosed. Still, if a meaningful protective effect exists, one would expect to see some suggestion of an association by 15 years. Fourth, there remains the possibility that some cases of bladder cancer were undiagnosed, impacting their inclusion in our data analyses. However, we do not expect this to affect our data analysis and subsequent results dramatically as the number cases included in our data analysis was quite large.
In summary, contrary to preclinical data and some smaller studies in humans, our analysis showed that increasing consumption of cruciferous vegetables is not associated with reduced risk of bladder cancer. We encourage patients to consume cruciferous vegetables for their other health benefits, especially for cardiovascular health; however, they should not expect this diet to reduce their risk for bladder cancer.
Conclusion
Based on the current analysis, intake of cruciferous vegetables is not associated with bladder cancer risk.
Acknowledgements
This research was supported [in part] by the Intramural Research Program of the NIH, National Cancer Institute. Cancer incidence data from the Atlanta metropolitan area were collected by the Georgia Center for Cancer Statistics, Department of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, Georgia. Cancer incidence data from California were collected by the California Cancer Registry, California Department of Public Health’s Cancer Surveillance and Research Branch, Sacramento, California. Cancer incidence data from the Detroit metropolitan area were collected by the Michigan Cancer Surveillance Program, Community Health Administration, Lansing, Michigan. The Florida cancer incidence data used in this report were collected by the Florida Cancer Data System (Miami, Florida) under contract with the Florida Department of Health, Tallahassee, Florida. The views expressed herein are solely those of the authors and do not necessarily reflect those of the FCDC or FDOH. Cancer incidence data from Louisiana were collected by the Louisiana Tumor Registry, Louisiana State University Health Sciences Center School of Public Health, New Orleans, Louisiana. Cancer incidence data from New Jersey were collected by the New Jersey State Cancer Registry, The Rutgers Cancer Institute of New Jersey, New Brunswick, New Jersey. Cancer incidence data from North Carolina were collected by the North Carolina Central Cancer Registry, Raleigh, North Carolina. Cancer incidence data from Pennsylvania were supplied by the Division of Health Statistics and Research, Pennsylvania Department of Health, Harrisburg, Pennsylvania. The Pennsylvania Department of Health specifically disclaims responsibility for any analyses, interpretations or conclusions. Cancer incidence data from Arizona were collected by the Arizona Cancer Registry, Division of Public Health Services, Arizona Department of Health Services, Phoenix, Arizona. Cancer incidence data from Texas were collected by the Texas Cancer Registry, Cancer Epidemiology and Surveillance Branch, Texas Department of State Health Services, Austin, Texas. Cancer incidence data from Nevada were collected by the Nevada Central Cancer Registry, Division of Public and Behavioral Health, State of Nevada Department of Health and Human Services, Carson City, Nevad. We are indebted to the participants in the NIH-AARP Diet and Health Study for their outstanding cooperation. We also thank Sigurd Hermansen and Kerry Grace Morrissey from Westat for study outcomes ascertainment and management and Leslie Carroll at Information Management Services for data support and analysis.
Disclosure of conflict of interest
None.
References
- 1.Jia X, Zhong L, Song Y, Hu Y, Wang G, Sun S. Consumption of citrus and cruciferous vegetables with incident type 2 diabetes mellitus based on a meta-analysis of prospective study. Prim Care Diabetes. 2016;10:272–280. doi: 10.1016/j.pcd.2015.12.004. [DOI] [PubMed] [Google Scholar]
- 2.Wang PY, Fang JC, Gao ZH, Zhang C, Xie SY. Higher intake of fruits, vegetables or their fiber reduces the risk of type 2 diabetes: a meta-analysis. J Diabetes Investig. 2016;7:56–69. doi: 10.1111/jdi.12376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zhang X, Shu XO, Xiang YB, Yang G, Li H, Gao J, Cai H, Gao YT, Zheng W. Cruciferous vegetable consumption is associated with a reduced risk of total and cardiovascular disease mortality. Am J Clin Nutr. 2011;94:240–246. doi: 10.3945/ajcn.110.009340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Liu B, Mao Q, Cao M, Xie L. Cruciferous vegetables intake and risk of prostate cancer: a meta-analysis. Int J Urol. 2012;19:134–141. doi: 10.1111/j.1442-2042.2011.02906.x. [DOI] [PubMed] [Google Scholar]
- 5.Nomura SJO, Hwang YT, Gomez SL, Fung TT, Yeh SL, Dash C, Allen L, Philips S, Hilakivi-Clarke L, Zheng YL, Wang JH. Dietary intake of soy and cruciferous vegetables and treatment-related symptoms in Chinese-American and non-Hispanic White breast cancer survivors. Breast Cancer Res Treat. 2018;168:467–479. doi: 10.1007/s10549-017-4578-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Petimar J, Wilson KM, Wu K, Wang M, Albanes D, van den Brandt PA, Cook MB, Giles GG, Giovannucci EL, Goodman GE, Goodman PJ, Hakansson N, Helzlsouer K, Key TJ, Kolonel LN, Liao LM, Mannisto S, McCullough ML, Milne RL, Neuhouser ML, Park Y, Platz EA, Riboli E, Sawada N, Schenk JM, Tsugane S, Verhage B, Wang Y, Wilkens LR, Wolk A, Ziegler RG, Smith-Warner SA. A pooled analysis of 15 prospective cohort studies on the association between fruit, vegetable, and mature bean consumption and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev. 2017;26:1276–1287. doi: 10.1158/1055-9965.EPI-16-1006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wright ME, Park Y, Subar AF, Freedman ND, Albanes D, Hollenbeck A, Leitzmann MF, Schatzkin A. Intakes of fruit, vegetables, and specific botanical groups in relation to lung cancer risk in the NIH-AARP Diet and Health Study. Am J Epidemiol. 2008;168:1024–1034. doi: 10.1093/aje/kwn212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wu QJ, Yang G, Zheng W, Li HL, Gao J, Wang J, Gao YT, Shu XO, Xiang YB. Pre-diagnostic cruciferous vegetables intake and lung cancer survival among Chinese women. Sci Rep. 2015;5:10306. doi: 10.1038/srep10306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Abbaoui B, Lucas CR, Riedl KM, Clinton SK, Mortazavi A. Cruciferous vegetables, isothiocyanates, and bladder cancer prevention. Mol Nutr Food Res. 2018;62:e1800079. doi: 10.1002/mnfr.201800079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Abbaoui B, Telu KH, Lucas CR, Thomas-Ahner JM, Schwartz SJ, Clinton SK, Freitas MA, Mortazavi A. The impact of cruciferous vegetable isothiocyanates on histone acetylation and histone phosphorylation in bladder cancer. J Proteomics. 2017;156:94–103. doi: 10.1016/j.jprot.2017.01.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bhamre S, Sahoo D, Tibshirani R, Dill DL, Brooks JD. Temporal changes in gene expression induced by sulforaphane in human prostate cancer cells. Prostate. 2009;69:181–190. doi: 10.1002/pros.20869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Brooks JD, Paton VG, Vidanes G. Potent induction of phase 2 enzymes in human prostate cells by sulforaphane. Cancer Epidemiol Biomarkers Prev. 2001;10:949–954. [PubMed] [Google Scholar]
- 13.Jones SB, Brooks JD. Modest induction of phase 2 enzyme activity in the F-344 rat prostate. BMC Cancer. 2006;6:62. doi: 10.1186/1471-2407-6-62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Tang L, Zirpoli GR, Guru K, Moysich KB, Zhang Y, Ambrosone CB, McCann SE. Consumption of raw cruciferous vegetables is inversely associated with bladder cancer risk. Cancer Epidemiol Biomarkers Prev. 2008;17:938–944. doi: 10.1158/1055-9965.EPI-07-2502. [DOI] [PubMed] [Google Scholar]
- 15.Tang L, Zirpoli GR, Guru K, Moysich KB, Zhang Y, Ambrosone CB, McCann SE. Intake of cruciferous vegetables modifies bladder cancer survival. Cancer Epidemiol Biomarkers Prev. 2010;19:1806–1811. doi: 10.1158/1055-9965.EPI-10-0008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Abbaoui B, Riedl KM, Ralston RA, Thomas-Ahner JM, Schwartz SJ, Clinton SK, Mortazavi A. Inhibition of bladder cancer by broccoli isothiocyanates sulforaphane and erucin: characterization, metabolism, and interconversion. Mol Nutr Food Res. 2012;56:1675–1687. doi: 10.1002/mnfr.201200276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Munday R, Mhawech-Fauceglia P, Munday CM, Paonessa JD, Tang L, Munday JS, Lister C, Wilson P, Fahey JW, Davis W, Zhang Y. Inhibition of urinary bladder carcinogenesis by broccoli sprouts. Cancer Res. 2008;68:1593–1600. doi: 10.1158/0008-5472.CAN-07-5009. [DOI] [PubMed] [Google Scholar]
- 18.Tang L, Zhang Y. Dietary isothiocyanates inhibit the growth of human bladder carcinoma cells. J Nutr. 2004;134:2004–2010. doi: 10.1093/jn/134.8.2004. [DOI] [PubMed] [Google Scholar]
- 19.Liu B, Mao Q, Lin Y, Zhou F, Xie L. The association of cruciferous vegetables intake and risk of bladder cancer: a meta-analysis. World J Urol. 2013;31:127–133. doi: 10.1007/s00345-012-0850-0. [DOI] [PubMed] [Google Scholar]
- 20.Yao B, Yan Y, Ye X, Fang H, Xu H, Liu Y, Li S, Zhao Y. Intake of fruit and vegetables and risk of bladder cancer: a dose-response meta-analysis of observational studies. Cancer Causes Control. 2014;25:1645–1658. doi: 10.1007/s10552-014-0469-0. [DOI] [PubMed] [Google Scholar]
- 21.Vieira AR, Vingeliene S, Chan DS, Aune D, Abar L, Navarro Rosenblatt D, Greenwood DC, Norat T. Fruits, vegetables, and bladder cancer risk: a systematic review and meta-analysis. Cancer Med. 2015;4:136–146. doi: 10.1002/cam4.327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Xu C, Zeng XT, Liu TZ, Zhang C, Yang ZH, Li S, Chen XY. Fruits and vegetables intake and risk of bladder cancer: a PRISMA-compliant systematic review and dose-response meta-analysis of prospective cohort studies. Medicine (Baltimore) 2015;94:e759. doi: 10.1097/MD.0000000000000759. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Pelser C, Mondul AM, Hollenbeck AR, Park Y. Dietary fat, fatty acids, and risk of prostate cancer in the NIH-AARP diet and health study. Cancer Epidemiol Biomarkers Prev. 2013;22:697–707. doi: 10.1158/1055-9965.EPI-12-1196-T. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Thiebaut AC, Jiao L, Silverman DT, Cross AJ, Thompson FE, Subar AF, Hollenbeck AR, Schatzkin A, Stolzenberg-Solomon RZ. Dietary fatty acids and pancreatic cancer in the NIH-AARP diet and health study. J Natl Cancer Inst. 2009;101:1001–1011. doi: 10.1093/jnci/djp168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Davenport MT, Zhang CA, Leppert JT, Brooks JD, Eisenberg ML. Vasectomy and the risk of prostate cancer in a prospective US cohort: data from the NIH-AARP Diet and Health Study. Andrology. 2019;7:178–183. doi: 10.1111/andr.12570. [DOI] [PubMed] [Google Scholar]
- 26.Eisenberg ML, Park Y, Hollenbeck AR, Lipshultz LI, Schatzkin A, Pletcher MJ. Fatherhood and the risk of cardiovascular mortality in the NIH-AARP Diet and Health Study. Hum Reprod. 2011;26:3479–3485. doi: 10.1093/humrep/der305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Lamberts RW, Guo DP, Li S, Eisenberg ML. The relationship between offspring sex ratio and vasectomy utilization. Urology. 2017;103:112–116. doi: 10.1016/j.urology.2016.11.039. [DOI] [PubMed] [Google Scholar]
- 28.Larsson SC, Andersson SO, Johansson JE, Wolk A. Fruit and vegetable consumption and risk of bladder cancer: a prospective cohort study. Cancer Epidemiol Biomarkers Prev. 2008;17:2519–2522. doi: 10.1158/1055-9965.EPI-08-0407. [DOI] [PubMed] [Google Scholar]
- 29.Holst B, Williamson G. A critical review of the bioavailability of glucosinolates and related compounds. Nat Prod Rep. 2004;21:425–447. doi: 10.1039/b204039p. [DOI] [PubMed] [Google Scholar]
- 30.McNaughton SA, Marks GC. Development of a food composition database for the estimation of dietary intakes of glucosinolates, the biologically active constituents of cruciferous vegetables. Br J Nutr. 2003;90:687–697. doi: 10.1079/bjn2003917. [DOI] [PubMed] [Google Scholar]
- 31.Vallejo F, Tomas-Barberan F, Garcia-Viguera C. Health-promoting compounds in broccoli as influenced by refrigerated transport and retail sale period. J Agric Food Chem. 2003;51:3029–3034. doi: 10.1021/jf021065j. [DOI] [PubMed] [Google Scholar]