Soy Product and Isoflavone Consumption in Relation to Prostate Cancer in Japanese Men

The incidence of prostate cancer is much lower in Asian than in Western populations (1). However, Japanese migrants to the United States and Brazil have an increased incidence (2, 3), and the incidence of latent or clinically insignificant prostate cancer in autopsy studies among men from Asian countries and the United States is similar (4, 5). It has therefore been suggested that environmental factors may play an important role in the progression of prostate cancer. Asian populations consume large quantities of soy food that contained isoflavones such as genistein and daidzein (6). Mean serum or plasma concentrations of isoflavones in Japanese men are 10 to 100 times higher than those in men from the United Kingdom (7) and Finland (8). Moreover, Morton et al. (9) reported a higher concentration of daidzein in the prostatic fluid of Asian men than in Western men. Genistein and daidzein exhibit anticarcinogenic properties and estrogenic activity in vitro and have shown a protective effect against prostate cancer development in some animal studies (6). On these bases, isoflavones have been recognized as key substances that may decrease the incidence of prostate cancer in Asia. However, previous findings from epidemiologic studies regarding isoflavone or soy food intake and prostate cancer are equivocal (10-17).

This inconsistency may be due to errors in exposure measurement and limited variation in soy intake. Some of the previous epidemiologic studies investigated association between prostate cancer and a single soy food only, such as tofu or soy milk, and most were conducted in Western countries, in which physiologically meaningful amounts of soy are not consumed (10-12). Here, we investigated the association between isoflavone intake and risk of prostate cancer in a prospective study in Japanese who consume large amounts of soy.

The Japan Public Health Center–Based Prospective Study was initiated in 1990 for cohort I and in 1993 for cohort II. The study design has been described in detail previously (18). Cohort I included those residents ages 40 to 59 years who had registered their addresses in five public health center areas (Iwate, Akita, Nagano, Okinawa, and Tokyo). Cohort II included those residents ages 40 to 69 years who had registered in six public health center areas (Ibaraki, Niigata, Kochi, Nagasaki, Okinawa, and Osaka). The Tokyo subjects were not included in the data analysis because incidence data were not available. This study was approved by the institutional review board of the National Cancer Center, Tokyo, Japan. The initial cohort consisted of 68,557 men.

At baseline, participants completed a self-administered questionnaire that assessed information on lifestyle factors, medical, and smoking histories. The food frequency questionnaire (FFQ) in the baseline survey had 44 food items for cohort I and 52 food items for cohort II with four (cohort I) or five (cohort II) frequency categories but without standard portions/units. In contrast, the 5-year follow-up survey included a self-administered FFQ, which included lifestyle factors, medical history, and 147 food and beverage items with standard portions/units and nine frequency categories. Owing to this greater detail, the present study therefore used the 5-year follow-up survey as baseline and followed the subjects from 1995 for cohort I and from 1998 for cohort II until 2004. After the 5-year follow-up survey, 128 subjects were found to be ineligible and were excluded because of non-Japanese nationality (n = 28), late report of emigration occurring before the start of the follow-up period (n = 97), incorrect birth data (n = 3), and subjects with self-reported prostate cancer (n = 21), leaving 58,427 men eligible for participation. Among eligible subjects, 46,001 men (79%) returned valid responses to the 5-year follow-up FFQ.

We dealt with two item groups: consumption of miso soup and soy food. Soy food referred to the consumption of “Tofu, Yushidofu (pre-drained tofu), Koyadofu (freeze-dried tofu), Aburaage (deep-fried tofu), Natto (fermented soybean), and soymilk,” for which the major ingredient is soybeans. The questionnaire asked about the usual consumption of 147 foods and beverages during the previous year. The frequency of miso soup consumption was divided into six categories (almost never, 1-3 days/mo, 1-2 days/wk, 3-4 days/wk, 5-6 days/wk, and daily). Portion sizes were specified, and the amounts provided in three categories (less than half, same, and more than 1.5 times). One bowl of miso soup was calculated as 150 mL. Nine frequency categories were used for soy foods (almost never, 1-3 times per month, 1-2 times per week, 3-4 times per week, 5-6 times per week, once a day, 2-3 times per day, 4-6 times per day, and ≥7 times per day). Portion sizes were specified, and the amounts were determined in three categories (less than half, same, and more than 1.5 times). Ten frequency categories were used for soy milk (almost never, 1-3 times per month, 1-2 times per week, 3-4 times per week, 5-6 times per week, 1 glass per day, 2-3 glasses per day, 4-6 glasses per day, 7-9 glasses per day, and >9 glasses per day). The total consumption of miso soup (mL/d) and soy food (g/d) was calculated from these responses, whereas that of isoflavones (daidzein and genistein) was calculated using values in a specially developed food composition table for isoflavones in Japanese foods (19, 20).

Validity was assessed among subsamples using 14- or 28-day dietary records. Spearman's correlation coefficients between the energy-adjusted intake of miso soup and soy food consumption from the questionnaire and from dietary records were 0.54 and 0.53 for cohort I and 0.48 and 0.52 for cohort II, respectively, whereas those for energy-adjusted intake of daidzein and genistein were 0.65 and 0.65 for cohort I and 0.49 and 0.48 for cohort II, respectively. Moreover, Spearman's correlation coefficients for daidzein and genistein between energy-adjusted intakes from FFQ and those from serum concentration were 0.26 and 0.40, respectively, and with those from creatinine-adjusted urinary excretion were 0.22 and 0.33, respectively (21). These correlation coefficients are considered acceptable (22). With regard to the reproducibility of estimations between two questionnaires administered 1 year apart, respective correlation coefficients for the energy-adjusted intake of miso soup, soy food, daidzein, and genistein were 0.80, 0.64, 0.75, and 0.75 for cohort I and 0.75, 0.57, 0.53, and 0.51 for cohort II (23-25).

Among the 46,001 men who responded to the questionnaire, 2,492 who reported extreme total energy intake (<800 or >4,000 kcal) were excluded, leaving 43,509 men for analysis.

Subjects were followed from the 5-year follow-up survey until December 31, 2004. Changes in residence status, including survival, were identified annually through the residential registry in each area or, for those who had moved out of the study area, through the municipal office of the area to which they had moved. Generally, mortality data for residents included in the residential registry are forwarded to the Ministry of Health, Labour, and Welfare and coded for inclusion in the national Vital Statistics. Residency and death registration are required by the Basic Residential Register Law and Family Registry Law, respectively, and the registries are believed to be complete. Here, information on the cause of death was based on death certificates from the respective public health center for those who had not moved out of the original area. Among questionnaire respondents to the 5-year follow-up FFQ, 3,855 men (8.4%) died, 1,492 men (3.2%) moved out of the study area, and 66 men (0.1%) were lost to follow-up during the study period.

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 the local governments responsible for the cancer registries. Cases were coded using the International Classification of Diseases for Oncology, Third Edition (26). Death certificate information was used as a supplementary information source. The proportion of cases of prostate cancer first notified by death certificate was 0.9%. The ratio of incidence to mortality was 7.7. The registration rate as introduced by Parkin et al. (27) was 94.3%. The proportion of case patients with prostate cancer ascertained by death certificate only was 0.6%. These ratios were considered satisfactory for the present study. A total of 307 newly diagnosed prostate cancer cases were identified by December 31, 2004.

Finally, a population-based cohort of 43,509 men (18,105 in cohort I and 25,404 in cohort II) was established for analysis. During the 325,371 person-years of follow-up (167,611 in cohort I and 157,760 in cohort II), 307 cases of prostate cancer were newly diagnosed (156 in cohort I and 151 in cohort II).

Person-years of follow-up were calculated for each man from the date of completion of the 5-year follow-up FFQ to the date of prostate cancer diagnosis, the date of emigration from the study area, or the date of death, whichever came first; or if none of these occurred, follow-up was through to the end of the study period (December 31, 2004). Men who were lost to follow-up were censored at the last confirmed date of presence in the study area. The crude incidence rate for prostate cancer was calculated by dividing the number of prostate cancer cases by the number of person-years. The relative risks (RR) of prostate cancer were calculated in quartile for the categories of miso soup consumption, soy food consumption, and isoflavone intake, with the lowest consumption category as the reference. RRs and 95% confidence intervals (95% CI) were calculated by the Cox proportional hazards model, adjusting for age at 5-year follow-up survey and study area (10 public health center areas) according to the SAS PHREG procedure (version 9.1; SAS Institute, Inc., Cary, NC). For further adjustment, additional possible confounders were incorporated into the model: smoking status (never, former, and current); alcohol intake (almost never, <3-4 days/wk, and >5 days/wk); marital status (yes/no); body mass index; and consumption of dairy foods, vegetables, fruit, and total fatty acids.

We conducted additional analyses according to the stage of prostate cancer. Advanced cases were defined by a diagnosis of extraprostatic or metastatic cancer involving lymph nodes or other organs. If this information was not available, advanced cases were defined as those with a high Gleason score (8-10) or poor differentiation. These criteria were selected to allow the identification of advanced cases with a high likelihood of poor prognosis. The remaining cases were organ localized. In this study, there were 74 advanced cases, 220 localized cases, and 13 (4% of total) cases of undetermined stage.

The trend was assessed by assigning ordinal values for categorical variables. All Ps were two sided, and statistical significance was determined at the P < 0.05 level.

Distribution of subject characteristics at the 5-year follow-up survey according to quartile of energy-adjusted isoflavone consumption is shown in Table 1, in which the results for genistein were used as a surrogate for isoflavones owing to the high correlation among results for genistein, daidzein, miso soup, and soy food. Men with high genistein consumption were slightly older. Body mass index was higher in the highest category than in the other categories. The proportion of current smokers increased as genistein intake increased, whereas alcohol intake decreased. The proportion of men who live with their wife was lower in the lowest category than in the other categories. Screening-detected prostate cancer accounted for >40% of cases in all categories. Dairy food, fruit, vegetable, and total fatty acid consumption were positively correlated with genistein intake. As expected, soy food and miso soup increased as genistein intake increased and were highly correlated with it (P < 0.0001).

Table 2 shows age- and area-adjusted and multivariable RRs and 95% CIs for total prostate cancer by each quartile of genistein, daidzein, miso soup, and soy food consumption. Genistein, daidzein, and soy food consumption slightly decreased the risk of total prostate cancer. Multivariable RRs for the highest versus lowest quartile of genistein, daidzein, and soy food consumption were 0.71 (95% CI, 0.48-1.03), 0.77 (95% CI, 0.52-1.13), and 0.82 (95% CI, 0.57-1.19), respectively. Tests for linear trends were not statistically significant. No statistically significant association was seen between miso soup consumption and total prostate cancer risk (highest versus lowest: RR, 1.04; 95% CI, 0.72-1.50).

We next classified the data according to prostate cancer stage (Table 3). Consumption of genistein decreased the risk of localized prostate cancer in a statistically significant manner (RR, 0.59; 95% CI, 0.38-0.93), although P_trend was not statistically significant. Men in the highest quartile of daidzein also had a decreased the risk of localized prostate cancer, although with only marginal significance (RR, 0.66; 95% CI, 0.42-1.04). Miso soup and soy foods tended to decrease the risk of localized prostate cancer, with respective multivariable RRs for the highest versus lowest quartile of 0.78 (95% CI, 0.51-1.20) and 0.77 (95% CI, 0.50-1.19). In contrast, genistein and daidzein increased the risk of advanced prostate cancer, with respective RRs for men with the highest versus lowest consumption of genistein and daidzein of 1.26 (95% CI, 0.56-2.83) and 1.43 (95% CI, 0.63-3.28). Miso soup was dose-dependently increased the risk of advanced prostate cancer, with multivariable RR for the highest versus lowest quartile of 2.79 (95% CI, 1.19-6.55; P_trend = 0.02). Consumption of soy food was not associated with advanced cancer.

Table 4 shows multivariable RRs and 95% CIs for prostate cancer by stage categorized according to age. We found that the negative association with genistein, daidzein, and soy food and localized prostate cancer became clear when we analysis was restricted to men ages >60 years. Genistein, daidzein, and soy food were dose-dependently associated with a decreased risk of localized prostate cancer, with multivariable RR for the highest versus lowest quartile of 0.52 for genistein (95% CI, 0.30-0.90; P_trend = 0.03), 0.50 for daidzein (95% CI, 0.28-0.88; P_trend = 0.04), and 0.52 for soy food (95% CI, 0.29-0.90; P_trend = 0.01). Miso soup also decreased the risk of localized prostate cancer, although without statistical significance. In contrast, RRs of daidzein and miso soup showed an increased risk of advanced prostate cancer in men ages >60 years old. In particular, miso soup was positively associated with advanced cancer, with a multivariable RR for the highest versus lowest quartile of 2.86 (95% CI, 1.01-8.11). Daidzein tended to increase the risk of advanced cancer (highest versus lowest: RR, 1.49; 95% CI, 0.55-4.03). Genistein was not associated with advanced cancer. Soy food tended to be negatively associated with advanced cancer in men ages >60 years. In men ages ≤60 years, genistein, daidzein, and soy food increased the risk of both localized and advanced prostate cancer. Multivariable RR for the highest versus lowest quartile was 1.18 for genistein, 1.38 for daidzein, and 1.38 for soy food in localized prostate cancer and 2.00 for genistein, 2.46 for daidzein, and 1.48 for soy food in advanced prostate cancer. Miso soup was not associated with localized prostate cancer (RR, 0.82) but was associated with an increased risk of advanced prostate cancer (RR, 2.65). However, none of these values was statistically significant.

To weaken the influence of localized prostate cancer detected by prostate-specific antigen screening, we also analyzed the association between prostate cancer and the four items after excluding screening-detected tumors by stage in men ages >60 years, notwithstanding that screening information was available for only 70% of subjects (Table 5). Results in both localized and advanced prostate cancer were similar to those in Table 4 when screening-detected prostate cancer was included, although the statistical significance in these results was lost. Genistein, daidzein, and miso soup tended to decrease the risk of localized prostate cancer (highest versus lowest: RRs of genistein, daidzein, miso soup, and soy food of 0.52, 0.49, 0.73, and 0.51, respectively) but without statistical significance. In advanced prostate cancer, multivariable RR for the highest versus lowest quartile was 0.85 for genistein, 1.10 for daidzein, 1.97 for miso soup, and 0.73 for soy food; however, these results were not statistically significant. Results in subjects ages ≤60 years were similar to those which included screening-detected cancers. Multivariable RR for the highest versus lowest quartile was 1.28 (95% CI, 0.33-4.97) for genistein, 1.55 (95% CI, 0.42-5.78) for daidzein, and 1.94 (95% CI, 0.58-6.52) for soy food in localized prostate cancer and 2.22 (95% CI, 0.50-9.91) for genistein, 2.93 (95% CI, 0.53-16.29) for daidzein, and 1.82 (95% CI, 0.33-9.91) for soy food in advanced prostate cancer. However, these values were not statistically significant (data not shown).

In the present study, we observed a dose-dependent decrease in the risk of localized prostate cancer with isoflavone consumption. Men with the highest intake of isoflavones (as genistein, ≧32.8 mg/d) had a decreased risk of prostate cancer compared with those with the lowest intake of isoflavones (as genistein, <13.2 mg/d). To our knowledge, this is the first prospective study to report an inverse association between isoflavone and localized prostate cancer in Japanese, whose intake of soy food is high.

Our results support previous studies, which reported that soy food is protective for prostate cancer. Among case-control studies, Sonoda et al. (17) reported that natto (fermented soy) consumption showed a significantly decreasing linear trend for risk of prostate cancer in Japanese; Lee et al. (13) found that the highest intake of tofu and genistein had a statistically significant association with a decreased risk of prostate cancer in Chinese compared with the lowest intake; and Strom et al. (11) reported an inverse association between daidzein intake and prostate cancer risk in American men. Soy foods were also inversely related to prostate cancer in a large multicenter case-control study (12). In prospective studies, Jacobson et al. (10) reported that frequent consumption of soy milk was associated with a decreased risk of prostate cancer in Californian Adventist men. However, no association was seen between tofu consumption and a decreased risk of prostate cancer in Japanese men living in Hawaii (16), nor was tofu or miso soup significantly associated with prostate cancer risk in native Japanese (14). The reason these studies did not show a protective effect of soy foods on prostate cancer may have been due to their evaluation of a single soy food only or their failure to assess specific nutrients such as genistein or daidzein.

Studies in vivo and in vitro experiments have also shown a protective effect of isoflavones against prostate cancer development. Isoflavones possess weak estrogen activity, inhibit tyrosine protein kinases and angiogenesis, and reduce serum testosterone levels (6, 28, 29). Isoflavones also inhibit 5α-reductase, an enzyme that metabolizes testosterone to dihydrotestosterone (30). Any or all of these mechanisms may explain the inverse association between isoflavones and localized prostate cancer seen here. Moreover, our results are plausible because the incidence of prostate cancer in Japanese is much lower than in Western men (1).

However, when the data were analyzed by stage, we found that the results differed between advanced and localized cancer. These results suggest that the effects of isoflavone may differ according to stage. One mechanism by which isoflavones reduce the risk of prostate cancer seems to involve estrogen receptor β in prostate tissue (31), but cancer with higher metastatic potential is associated with the complete or partial loss of estrogen receptor β expression (32-34). Moreover, animal studies in rats showed that the beneficial effects of a soy diet play a role in the early stages of tumor development but have no effect in invasive prostate cancer (35, 36). On this basis, isoflavones may prevent the early stages of prostate cancer development only. Clinically significant localized prostate cancer likely arises from latent cancer and then develop to advanced cancer with high mortality (4, 37). Given that the incidence of latent prostate cancer in Japanese men is the same as in Western men despite a lower incidence of prostate cancer (1, 4, 5), isoflavone may delay the progression of latent prostate cancer.

When we limited analysis to men ages >60 years, the association between isoflavone and localized prostate cancer was strengthened. Hoffman et al. (38) reported that men with cancers detected by prostate-specific antigen screening were more often younger than those men in whom cancer was clinically diagnosed. Our study also showed that the proportion of screening-detected cancers was higher (54.6%) in those men ages ≤60 years than in those ages >60 years (28.1%), although prostate-specific antigen screening information was available for only 70% of subjects. However, although we analyzed the association between localized prostate cancer and isoflavones after excluding screening-detected tumors, results did not change. Isoflavone may be protective for localized prostate cancer only in men ages >60 years and may not have a protective effect in the early stage of prostate cancer in younger men.

Our study has several methodologic strengths. First, it was a prospective design, which diminishes the probability of recall bias that is inherent to case-control studies. Second, we evaluated isoflavone intake using a validated questionnaire, and participants had a large variation in isoflavone consumption. One reason for the inconsistent findings for the association between soy food and prostate cancer in previous studies may be errors in exposure measurements and the small exposure variation in Western subjects. Third, we adjusted possible confounding factors to remove associations with other substances. It is also possible that a lifestyle associated with a high intake of soy food may have contributed to the risk of prostate cancer. In this study, the associations between isoflavones and prostate cancer were strengthened after adjustment for several confounding factors. Fourth, response rate was high (∼80%), and the proportion of subjects lost to follow-up was relatively low (0.1%).

On the other hand, the present study had several limitations. One was our inability to distinguish screening-detected cancer from total prostate cancer. It is possible that men who have health check-ups are more health conscious and may consume more soy food. However, such misclassification, if present, would lead to increase the risk of localized prostate cancer. Therefore, this inability to distinguish would not account for the decreased risk of localized prostate cancer. Another limitation was that the number of advanced prostate cancer cases was small. A larger sample size may have detected the positive effects of isoflavones on advanced prostate cancer with greater precision. Moreover, misclassification of exposure due to changes in isoflavone consumption during the study period might have occurred because we used information on consumption obtained at one point only. If present, however, such misclassification would underestimate the true relative risk.

In summary, we found that isoflavone intake was associated with a decreased risk of localized prostate cancer but tended to be associated with an increased risk of advanced prostate cancer. Recent interest has focused on whether isoflavones have chemopreventive effects. Given that Japanese consume isoflavones regularly throughout life, we do not yet know the period during which the effects of isoflavones on prostate cancer are preventive. Further research is required, including well-designed clinical trials in humans.

Members of the Japan Public Health Center–Based Prospective Study Group (Principal investigator: S. Tsugane): 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, I. Hashimoto, and T. Ikuta (Iwate Prefectural Ninohe Public Health Center, Ninohe); Y. Miyajima, N. Suzuki, S. Nagasawa, Y. Furusugi and N. Nagai (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, M. Suzuki, and K. Matsui (Niigata Prefectural Kashiwazaki, Kashiwazaki and Nagaoka Public Health Center, Nagaoka); M. Doi, A. Terao, Y. Ishikawa, and T Tanoue (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, H. Takaesu, and M. Uehara (Okinawa Prefectural Miyako Public Health Center, Hirara); F. Horii, I. Asano, H. Yamaguchi, K. Aoki, S. Maruyama, M. Ichii, and M. Takano (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, Toon); 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); the late M. Yamaguchi, Y. Matsumura, S. Sasaki, and S. Watanabe (NIH and Nutrition, Tokyo); T. Kadowaki (Tokyo University, Tokyo); Y. Kawaguchi (Tokyo Medical and Dental University, Tokyo); H. Shimizu (Sakihae Institute, Gifu).