Dietary Fiber and Risk of Colorectal Cancer in the Japan Collaborative Cohort Study

In the early 1970s, Burkitt hypothesized that dietary fiber may protect against colorectal cancer, based on the observation that colorectal cancer was rare in rural Africans, and they ate a diet rich in fiber from unrefined grains and/or leafy vegetables (1). Since then, many case-control studies reported an inverse association between dietary fiber intake and the risk of colorectal cancer (2-6). Furthermore, many animal models show different inhibitory effects of various types of fiber on colon tumor development (7-9). On the contrary, many prospective studies showed no protective effects of fiber (10-20), although a recent international cohort study in Europe has reported an inverse association between fiber intake and colorectal cancer risk (21, 22). In addition, increasing fiber intake did not reduce adenoma recurrence in large intervention trials (23-25). Thus, Burkitt's hypothesis has not fully been supported.

The short-term study of recurrent adenomas in the clinical trials, however, may have little relevance to the progression of adenomas to colorectal cancer (26). As for cohort studies on this issue, all have been conducted in Western populations (10-20) except for one in Japan (27). Because food sources of dietary fiber greatly vary (21) and genetic susceptibility to colorectal cancer may differ between ethnic groups, prospective studies are warranted also in non-Western populations. Le Marchand et al. (4) suggested that the risk reduction associated with fiber from vegetables was larger in Japanese than in Caucasians.

We have been conducting a large cohort study named as the Japan Collaborative Cohort (JACC) Study for the evaluation of cancer risk sponsored by the Monbusho, the Ministry of Education, Culture, Sports, Science and Technology of Japan (JACC Study; refs. 28, 29). The study involves subjects throughout Japan with a relatively low level of dietary fiber intake; median, 13.3 g/day in 2003 (30). The potential protective effects of dietary fiber may have a threshold intake level, so it may be informative to conduct investigations in populations with a low intake.

Japan is one of the countries with the highest incidence rate of colorectal cancer (31); the rate per 100,000 population (standardized to the World Population) was 49.9 in men and 27.2 in women in 1999 (32). Furthermore, immigrant studies suggest the higher genetic susceptibility to colorectal cancer risk in Japanese; Japanese Americans showed a higher incidence rate than U.S. Caucasians (31). Our cohort, therefore, may be appropriate to address the issue. Thus, we examined the association of dietary fiber intake with colorectal cancer risk by analyzing the data from the JACC Study.

The JACC Study for evaluation of cancer risk sponsored by the Monbusho started in 1988 to 1990, during which period, 110,792 male and female inhabitants aged 40 to 79 years completed a baseline questionnaire. The details of this study are described elsewhere (28, 29). In brief, participants were enrolled from 45 study areas throughout Japan, mostly from general populations or examinees of municipal health check-ups. Informed consent for participation was obtained individually from each participant, except in a few study areas where it was provided at the group level after the aim of the study and confidentiality of the data had been explained to community leaders. The Ethical Board of the Nagoya University School of Medicine approved the protocol of this investigation, including the procedures used to obtain informed consent.

Potential participants for the present analysis were restricted to 60,569 men and women who lived in 22 study areas where information on cancer incidence is available, and a food frequency questionnaire (FFQ) to estimate food and nutrient intake was included in the baseline questionnaire.

The baseline questionnaire covered lifestyle factors, including dietary habits, smoking and drinking, and physical activity, as well as medical history, education, family history of cancer, and height and weight.

The dietary component of the questionnaire included 40 food items (33). For 33 foods or dishes, we asked about the average intake frequency without specifying portion size information. We used five response choices: almost never, 1 to 2 times a month, 1 to 2 times a week, 3 to 4 times a week, and almost everyday. For rice, miso soup (soup of fermented soybean paste with soybean curd, vegetables, and/or seaweeds, etc.), and four nonalcoholic beverages, the number of bowls or cups consumed per day was inquired. The frequency of alcohol consumption was asked with the usual amount on one occasion. Nutrient intakes were computed using the Japanese food composition table (34) assuming the standard portion sizes. Values for total, water-soluble, and insoluble fiber, obtained by enzymatic-gravimetric methods by Prosky et al. (35), were derived from the food composition table.

Energy-adjusted intakes of nutrients, including total, soluble, insoluble, fruit, vegetable, and bean fiber, were calculated by the residual method (36). Natural logarithms of energy and nutrient intakes were used to improve the normality of their distribution.

The FFQ was validated by referring to four 3-day weighed dietary records over a 1-year period as a standard (33). We reanalyzed data from the validation study to consider skewed distributions of nutrient intakes and within-person variation in intakes (37, 38). The de-attenuated correlation coefficients for energy-adjusted intakes between the FFQ and dietary records were 0.46 for total dietary fiber, 0.42 for soluble fiber, 0.47 for insoluble fiber, 0.33 for fruit fiber, 0.41 for vegetable fiber, and 0.37 for bean fiber. The ratios of mean intakes estimated by the FFQ to those calculated from the dietary records were 0.60, 0.51, 0.58, 0.84, 0.53, and 0.50 for total, soluble, insoluble, fruit, vegetable, and bean fiber, respectively (part of total dietary fiber was not classified as soluble or insoluble fiber in the food composition table). We did not estimate intake of cereal fiber by the FFQ because the correlation coefficient was <0.3.

Of the 60,569 potential participants, we excluded 159 men and women with a history of colorectal cancer, 17,221 without sufficient responses to the FFQ to estimate nutrient intake (judged by predefined criteria), and 74 with an implausibly high or low intake of total energy (<500 or >3,500 kcal/day), leaving 16,636 men and 26,479 women (71.2% of potential participants) eligible for the analysis. Participants included in the analysis were younger (mean age ± SD, 56.4 ± 10.1 years for men and 56.7 ± 9.9 years for women) and likely to have a family history of colorectal cancer in parents and/or siblings (2.4% in men and 2.8% in women) than those excluded (61.0 ± 10.2 years for men and 62.7 ± 9.4 years for women, 1.9% in men and 2.1% in women, respectively). In women, the analytic cohort tended to be more highly educated (beyond high school, 11.4%) and included fewer ever smokers (6.6%) compared with those left out (9.1% and 9.0%, correspondingly). Other baseline characteristics, including drinking habits, body mass index (BMI), walking time, time of exercise, and engagement in sedentary work were comparable between the two groups.

We used population registries in the municipalities to determine the vital and residential status of the participants. Registration of death is required by the Family Registration Law in Japan and is adhered to nationwide. For logistical reasons, we discontinued the follow-up of those who had moved out of their given study areas.

We ascertained the incidence of cancer by means of a linkage with the records of population-based cancer registries, supplemented by a systematic review of death certificates. In some study areas, medical records in local major hospitals were also reviewed. In 3 out of 22 areas, population-based cancer registries were not available. Therefore, hospital-based cancer registries or inpatient records of hospitals treating cancer patients were used to collect information on cancer incidence in such areas. The mortality and incidence data were coded following the rules of the 10th Revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10). We defined colorectal cancer as C18, C19, and C20 of the ICD-10 code, colon cancer as C18, and rectal cancer as C20.

The follow-up was conducted from the time of the baseline survey through the end of 1997 except for one area (to the end of 1994). During the study period, only 2.8% (n = 1,190) of the subjects were lost to follow-up due to moving away. In the analytic cohort, the proportion of death-certificate-only registrations was 3.8% (17 of 443 cases) for colorectal cancer. The mortality-to-incidence ratio was 0.24, which is comparable with those available from representative population-based cancer registries in Japan (0.20 to 0.53; ref. 31).

BMI at baseline was calculated from reported height and weight: BMI = (weight in kilograms)/(height in meters)². The age- and sex-adjusted mean intake of dietary fiber by bowel movement frequency was computed using the analysis of covariance. We counted person-time of follow-up for each participant from the date of filling out the baseline questionnaire to the date of diagnosis of colorectal cancer, the date of death from any cause, the date of emigration outside the study area, or the end of the follow-up period, whichever came first. For cases identified only with a death certificate, the date of death was assumed to be that of diagnosis. Those who died from causes other than colorectal cancer or who moved out of their study areas were treated as censored cases.

The rate ratios (RR) with 95% confidence intervals (95% CI) for colorectal cancer over sex-specific quartiles of energy-adjusted intakes of total, soluble, insoluble, fruit, vegetable, or bean dietary fiber (the RRs for the second, third, and highest quartiles versus the lowest) were estimated using proportional hazards models (39) adjusted for age, sex, and other potential confounders selected based on prior knowledge (40-42). The RRs were adjusted for age (as a continuous variable), sex, area (Hokkaido and Tohoku, Kanto, Chubu, Kinki, Chugoku, or Kyushu), educational level (attended school until the age of ≤15, 16-18, or ≥19 years), family history of colorectal cancer in parents and/or siblings (yes or no), alcohol consumption [never drink, ex-drinkers, or current drinkers who consume <2 or ≥2 Japanese drinks (<46 g or ≥46 g of ethanol) per day for men, and never drink, ex-drinkers, or current drinkers for women), smoking (never smoke, ex-smokers, or current smokers), BMI (<20.0, 20.0-24.9, or ≥25.0 kg/m²), daily walking habits (≤30 or >30 min/day), exercise (seldom or never, or 1-2, 3-4, or ≥5 h a week), sedentary work (yes or no), consumption of beef (almost never, 1-2 times a month, 1-2 times a week, ≥3 times a week) and pork (almost never, 1-2 times a month, 1-2 times a week, ≥3 times a week), energy intake (as a continuous variable), and energy-adjusted intakes of folate, calcium, and vitamin D (sex-specific quartile for each). We considered walking time because that was the major physical activity in the study population (43). The RRs and 95% CIs were also computed for colon, rectal, and colorectal cancer among men, women, and both sexes combined. Missing values for each covariate were treated as an additional category in the variable and were included in the proportional hazards model. As a basis for the trend tests, the dietary fiber intake was scored from 0 to 3 according to the sex-specific quartile of intake, and the score was included in the model (21).

We repeated all the analyses after excluding the first 2 years of follow-up in which 79 cases of colorectal cancer were diagnosed. All P values were two-sided, and all the analyses were done using the Statistical Analysis System (44).

Mean intake of total dietary fiber was 100% and 80% higher in the highest quartile than in the lowest for men and women, respectively (Table 1). Men and women with higher dietary fiber intake were older, more likely to be highly educated, to walk more than 30 min/day, to exercise more than 1 h/week, and to engage in sedentary work, whereas the proportions of current drinkers and smokers decreased with an increasing total fiber intake. For dietary variables, individuals with higher intake of dietary fiber more frequently consumed pork but less frequently ate beef. We further found increasing trends in intakes of folate, calcium, and vitamin D with an increasing level of total fiber intake. The major sources of dietary fiber were miso soup (18.2%), rice (14.1%), fruits other than oranges (8.7%), and green leafy vegetables (7.2%).

During the 327,273 person-years of follow-up (mean ± SD, 7.6 ± 1.9 years), 443 cases of incident colorectal cancer were documented. In all participants, we found a decreasing trend in risk of colorectal cancer with increasing intake of total dietary fiber after adjustment for potential confounding factors (Table 2). The multivariate RRs across quartiles were 1.00, 0.96 (95% CI, 0.72-1.27), 0.72 (0.53-0.99), and 0.73 (0.51-1.03; P_trend = 0.028). This trend was exclusively detected for colon cancer: the corresponding RRs were 1.00, 0.90 (95% CI, 0.64-1.26), 0.56 (0.38-0.83), and 0.58 (0.38-0.88; P_trend = 0.002). The inverse associations were similarly found between intake of total dietary fiber and the risk of colorectal or colon cancer among men. In women, those with high intakes of total dietary fiber were at a decreased risk of colorectal or colon cancer, although the trend in RRs with increasing intakes did not reach statistical significance. The multivariate RRs for female rectal cancer were greater than unity for higher levels of fiber intakes. These figures, however, were unstable because of the small number of cases (n = 39). The major confounders that changed the age- and sex-adjusted RRs for the highest quartile of total dietary fiber for colorectal cancer by more than 5% were area and energy-adjusted intake of calcium for men and women combined, area, alcohol consumption, and energy-adjusted intakes of calcium and vitamin D for men, and area for women.

When dietary fiber was classified into water-soluble and insoluble fiber, the reduction in risk of colorectal or colon cancer associated with a higher intake was observed in the multivariate RRs for both kinds of fiber (Table 3). The risk of rectal cancer was not correlated with soluble or insoluble dietary fiber. It was difficult to estimate the RRs for soluble and insoluble fiber controlled for each other because the intakes of these two kinds of fiber were very highly correlated (Spearman's correlation coefficient = 0.95), and this adjustment resulted in inflated CIs; the multivariate RRs for colorectal cancer across quartiles of intakes were 1.00, 0.76 (95% CI, 0.51-1.13), 0.78 (0.46-1.32), and 0.69 (0.36-1.32; P_trend = 0.32) for soluble dietary fiber and 1.00, 1.27 (95% CI, 0.85-1.89), 0.94 (0.56-1.59), and 1.02 (0.54-1.94; P_trend = 0.85) for insoluble dietary fiber among men and women combined.

For food sources of fiber (Table 4), bean fiber intake was somewhat inversely correlated with the risk of colorectal cancer (P_trend for multivariate RRs = 0.055), particularly that of colon cancer (P_trend for multivariate RRs = 0.037). On the other hand, the intake of vegetable fiber tended to be associated with rectal cancer risk (P_trend for multivariate RRs = 0.060). The positive association of fruit fiber with age- and sex-adjusted RRs for colorectal cancer disappeared after adjustment for potential confounding factors.

All the findings in Tables 2–4 remained essentially the same when we excluded the first 2 years of follow-up from the analyses (data not shown). The multivariate RRs (95% CI) in men and women combined over quartiles of intake of total dietary fiber were 1.00, 0.95 (0.69-1.30), 0.74 (0.52-1.04), and 0.70 (0.48-1.02; P_trend = 0.029) for colorectal cancer; 1.00, 0.92 (0.64-1.34), 0.62 (0.40-0.94), and 0.56 (0.35-0.90; P_trend = 0.005) for colon cancer; and 1.00, 0.92 (0.49-1.71), 1.13 (0.61-2.12), and 1.08 (0.54-2.16; P_trend = 0.69) for rectal cancer.

In this cohort of Japanese population, dietary fiber intake was inversely associated with colorectal cancer risk. The association was stronger for the risk of colon cancer, and no clear relationship was observed for rectal cancer risk. In addition, no material differences appeared in the strength of associations with the risk between soluble and insoluble dietary fiber.

Many possible mechanisms have been proposed to explain putative risk-reducing effect of increased dietary fiber intake (7-9). When entering the large bowel, fiber increases stool bulk and dilutes fecal carcinogens. It also shortens fecal transit time and, thus, reduces the contact of the colon epithelium to carcinogens in stool. Indeed, severe constipation (bowel movement every 6 days or less) was linked with the risk of female colorectal cancer in the JACC Study (45). In the present population, the intake of total dietary fiber decreased with decreasing frequency of bowel movement; the age- and sex-adjusted mean intakes ± SE were 10.49 ± 0.02, 9.84 ± 0.04, 9.39 ± 0.10, and 9.24 ± 0.23 g/day for bowel movement frequency of once or more than once a day, once every 2 to 3 days, once every 4 to 5 days, and once every 6 days or more, respectively (P_trend < 0.0001). The bowel movement frequency, therefore, may be considered as an intermediate factor in the causal pathway. Furthermore, dietary fiber can bind to bile acids that produce carcinogenic metabolites.

Moreover, fiber acts as a substrate for bacterial fermentation, and short-chain fatty acids are produced. These fatty acids prevent the conversion from primary to the more toxic secondary bile acids by lowering colonic pH. The butyrate, one of the major short-chain fatty acids, has specific anticarcinogenic effects; it reduces cell proliferation and induces apoptosis. Fermentable fiber has the potential to selectively enhance the growth of the flora such as those of certain lactic acid bacteria and bifidobacteria. Recently, hyperinsulinemia has been related to colorectal carcinogenesis (46, 47). Dietary fiber can delay the absorption of starch, reducing the consequent postprandial hyperinsulinemia (41).

The majority of case-control studies reported inverse associations between dietary fiber intake and colorectal cancer risk. A combined analysis of 13 case-control studies yielded summary odds ratios for colorectal cancer of 1.00, 0.79, 0.69, 0.63, and 0.53 across the quintiles of intake compared with the lowest one (2). Large-scale case-control studies after this pooled analysis also supported the protective effects of fiber (3-6).

The findings of prospective studies, however, are rather inconsistent. Park et al. (48) pooled 13 cohort studies and concluded that high dietary fiber intake was not related to a reduced risk of colorectal cancer; the summary relative risk was 0.94 (95% CI, 0.86-1.03) for the highest versus lowest quintiles of intake. On the contrary, the European Prospective Investigation into Cancer and Nutrition (EPIC) reported a lower risk of colorectal cancer with higher intake of total fiber (21, 22).

The discrepancy between the studies may, in part, be attributable to the difference in the extent of diversity of diet among the study population. In the EPIC, the analytic cohort included participants from eight countries and showed a very large variety of diet (21).

In our cohort, the low-level intake of dietary fiber may be related to the emergence of risk-reducing effect of dietary fiber. In fact, the above-mentioned pooled analysis of prospective studies (48) suggested an increase in the risk associated with very low fiber intake. The pooled RR comparing <10 g/day versus ≥10 g/day of dietary fiber intake was 2.16 (95% CI, 1.12-4.16) after correction for measurement error. Although the 40-item FFQ in the JACC Study may underestimate the intake, the data from the National Nutrition Survey (30) shows the low median intake of dietary fiber in Japan (13.3 g/day in 2003). Because the 25-percentile value was lower than 10 g/day in the survey (9.7 g/day), one fourth of Japanese might benefit from increasing intake of dietary fiber to prevent colorectal cancer if the results of the pooled analysis of cohorts (48) are true. The Japan Public Health Center–Based Prospective Study, another nationwide cohort study in Japan, did not find a clear dose-response relationship between dietary fiber and the risk of colorectal cancer, but reported an elevated risk in the subgroup of women with a very low intake of fiber [the lowest subtertile of the lowest quintile (mean, 8.3 g/day for the lowest quintile); ref. 27]. The low dietary fiber content in refined rice, the staple among Japanese, may result in the low intake of fiber in this population (7).

Moreover, the major food contributors to fiber will vary between countries, and the combination of miso soup and rice may be specific to Japan (27). These foods might be rich in some kinds of dietary fiber that reduce the risk of colorectal cancer. Indeed, bean fiber intake was somewhat inversely correlated with the risk in the present study as reported by Lin et al. (19). We cannot, however, rule out the possible confounding by some biofactors concentrated in the major sources of dietary fiber.

In this study, the inverse association of dietary fiber with the risk was predominantly found for colon cancer rather than rectal cancer. This may be biologically plausible because the rectum is empty most of the time (49), reducing the above-mentioned potential risk-reducing effects of fiber. The colon predominance in risk reduction in the present study is in line with some case-control (50) and cohort studies (21, 22).

Our analyses did not find a material difference in the effects of colorectal cancer risk for soluble and insoluble fiber. Of the five case-control studies that addressed such a difference, four are consistent with our results (3-5, 51), whereas Slattery et al. (6) reported a stronger inverse association of insoluble fiber with rectal cancer risk than that of soluble fiber. Wakai et al. (50) also found a decrease in the risk of colon cancer specifically associated with insoluble fiber in a case-control study. Soluble fiber is most effective for delaying starch absorption, thus reducing the glycemic load (41). On the other hand, insoluble fiber is most efficient for reducing the transit time of stool, provides bulk, dilutes fecal contents, and is converted to short-chain fatty acids, including butyrate as well as soluble fiber (7). The relative importance of soluble and insoluble fiber may depend on the underlying mechanism of colorectal carcinogenesis in the populations studied. In addition, it would not be easy for observational studies to separate the effects of soluble and insoluble fiber because intakes of these two kinds of fiber may be highly correlated as in the present study.

The percentage of individuals with a positive family history of colorectal cancer in parents and/or siblings was 2.4% in men and 2.8% in women in the present study. Considering the higher incidence rates in Japan than in the United States (31), these figures seem to be very low compared with those in the United States, where the percentage often exceeds 10% (17, 19, 20). In Japan, however, the low proportion is not specific to the JACC Study. The percentage was 1.3% for both men and women in the Japan Public Health Center-Based Prospective (JPHC) Study I (52). The proportions of participants who reported a family history of colorectal cancer in parents were 1.3% in men and 1.2% in women in the JPHC Study II (52), against 1.7% and 1.9% in our population, respectively. The corresponding figures for history in siblings were 0.5% and 0.5% in the JPHC Study II (52) and 0.8% and 0.9% in the present study. One possible reason for the low percentages in Japan is the rapid increase in incidence rates in the country (32). Another one may be the underreporting of family history. The sensitivity of self-reporting of past cancer history was previously reported to be quite low in Japan (36%; ref. 53), so that underreporting of family history may have occurred.

The strengths of the present study include its prospective design and large sample size. We assessed dietary intake before the diagnosis of colorectal cancer; thus, any errors of recall should have been nondifferential between cases and noncases.

Some methodologic limitations of our study, however, need consideration. First, because dietary habits could have changed during the follow-up, one-time administration of the FFQ at baseline might not sufficiently reflect the long-term exposure. However, the Nurses' Health Study has found that findings from baseline diet questionnaire were consistent with those from data updated with follow-up FFQs (20).

Second, possible residual confounding cannot be ruled out. For example, we could not consider the use of nonsteroidal anti-inflammatory drugs, which may confound the association of diet with colorectal cancer risk (41). In this study, physical activity was rather roughly measured, although it seemed not to be a major confounder. Third, the intakes of dietary fiber are likely to be underestimated due to the limited number of food items in our FFQ, as shown by the relatively low ratios of mean intakes estimated by the FFQ to those derived from the dietary records in the validation study. Finally, the dietary data from the 40-item FFQ did not allow us to estimate intakes of fiber from some food sources such as cereals. Further studies with comprehensive dietary instruments are required to address these issues.

In conclusion, this prospective study supported potential protective effects of dietary fiber intake against colorectal cancer and mainly against colon cancer. The highest intake may be related to a 25% decrease in the risk of colorectal cancer, compared with the lowest intake. The role of dietary fiber in the prevention of colorectal cancer seems to remain inconsistent. Further investigations in various populations are warranted.

The present investigators involved, with the co-authorship of this paper, in the JACC Study and their affiliations are as follows: Dr. Akiko Tamakoshi (present chairman of the study group), Nagoya University Graduate School of Medicine, Nagoya, Japan; Dr. Mitsuru Mori, Sapporo Medical University School of Medicine, Sapporo, Japan; Dr. Yutaka Motohashi, Akita University School of Medicine, Akita, Japan; Dr. Ichiro Tsuji, Tohoku University Graduate School of Medicine, Sendai, Japan; Dr. Yosikazu Nakamura, Jichi Medical School, Shimotsuke, Japan; Dr. Hiroyasu Iso, Graduate School of Medicine, Osaka University, Suita, Japan; Dr. Haruo Mikami, Chiba Cancer Center, Chiba, Japan; Dr. Yutaka Inaba, Juntendo University School of Medicine, Tokyo; Dr. Yoshiharu Hoshiyama, University of Human Arts and Sciences, Iwatsuki, Japan; Dr. Hiroshi Suzuki, Niigata University School of Medicine, Niigata, Japan; Dr. Hiroyuki Shimizu, Gifu University School of Medicine, Gifu, Japan; Dr. Hideaki Toyoshima, Nagoya University Graduate School of Medicine, Nagoya, Japan; Dr. Kenji Wakai, Aichi Cancer Center Research Institute, Nagoya, Japan; Dr. Shinkan Tokudome, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan; Dr. Yoshinori Ito, Fujita Health University School of Health Sciences, Toyoake, Japan; Dr. Shuji Hashimoto, Fujita Health University School of Medicine, Toyoake, Japan; Dr. Shogo Kikuchi, Aichi Medical University School of Medicine, Nagakuto, Japan; Dr. Akio Koizumi, Graduate School of Medicine and Faculty of Medicine, Kyoto University, Kyoto, Japan; Dr. Takashi Kawamura, Kyoto University Center for Student Health, Kyoto, Japan; Dr. Yoshiyuki Watanabe, Kyoto Prefectural University of Medicine Graduate School of Medical Science, Kyoto, Japan; Dr. Tsuneharu Miki, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan; Dr. Chigusa Date, Faculty of Human Life and Environment, Nara Women's University, Nara, Japan; Dr. Kiyomi Sakata, Wakayama Medical University, Wakayama, Japan; Dr. Takayuki Nose, Tottori University Faculty of Medicine, Yonago, Japan; Dr. Norihiko Hayakawa, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan; Dr. Takesumi Yoshimura, Fukuoka Institute of Health and Environmental Sciences, Dazaifu, Japan; Dr. Akira Shibata, Kurume University School of Medicine, Kurume, Japan; Dr. Naoyuki Okamoto, Kanagawa Cancer Center, Yokohama, Japan; Dr. Hideo Shio, Moriyama Municipal Hospital, Moriyama, Japan; Dr. Yoshiyuki Ohno, Asahi Rosai Hospital, Owari-Asahi, Japan; Dr. Tomoyuki Kitagawa, Cancer Institute of the Japanese Foundation for Cancer Research, Tokyo; Dr. Toshio Kuroki, Gifu University, Gifu, Japan; and Dr. Kazuo Tajima, Aichi Cancer Center Research Institute, Nagoya, Japan.