Tel.: 81-3-3542-2511, ext. 3378, Fax: +81-3-3547-8578
Intake of n-3 and n-6 polyunsaturated fatty acids and development of colorectal cancer by subsite: Japan Public Health Center–based prospective study†
Version of Record online: 1 APR 2011
Copyright © 2010 UICC
International Journal of Cancer
Volume 129, Issue 7, pages 1718–1729, 1 October 2011
How to Cite
Sasazuki, S., Inoue, M., Iwasaki, M., Sawada, N., Shimazu, T., Yamaji, T., Takachi, R., Tsugane, S. and for the Japan Public Health Center–Based Prospective Study Group (2011), Intake of n-3 and n-6 polyunsaturated fatty acids and development of colorectal cancer by subsite: Japan Public Health Center–based prospective study. Int. J. Cancer, 129: 1718–1729. doi: 10.1002/ijc.25802
Members of the JPHC 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, A. Okayama and Y. Kokubo, National Cardiovascular Center, Suita; K. Miyakawa, F. Saito, A. Koizumi, Y. Sano, I. Hashimoto, T. Ikuta and Y. Tanaba, 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, Y. Miyagawa, Y. Kobayashi and M. Machida, 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, F. Shoji and R. Saito, Katsushika Public Health Center, Tokyo; A. Murata, K. Minato, K. Motegi and T. Fujieda, Ibaraki Prefectural Mito Public Health Center, Mito; T. Abe, M. Katagiri, M. Suzuki and K. Matsui, Niigata Prefectural Kashiwazaki and Nagaoka Public Health Center, Kashiwazaki and Nagaoka; M. Doi, A. Terao, Y. Ishikawa and T. Tagami, Kochi Prefectural Chuo-higashi Public Health Center, Tosayamada; H. Doi, M. Urata, N. Okamoto, F. Ide and H. Sueta, 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, K. Okada, and I. Saito, Ehime University, Toon; H. Iso, Osaka University, Suita; Y. Honda, K. Yamagishi, S. Sakurai and N. Tsuchiya, 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, W. Ajiki and A. Ioka, 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; K. Nakamura, Niigata University, Niigata; S. Kono, Kyushu University, Fukuoka; K. Suzuki, Research Institute for Brain and Blood Vessels Akita, Akita; Y. Takashima and M. Yoshida, Kyorin University, Mitaka; E. Maruyama, Kobe University, Kobe; M. Yamaguchi, Y. Matsumura, S. Sasaki and S. Watanabe, National Institute of Health and Nutrition, Tokyo; T. Kadowaki, Tokyo University, Tokyo; M. Noda and T. Mizoue, International Medical Center of Japan, Tokyo; Y. Kawaguchi, Tokyo Medical and Dental University, Tokyo; and H. Shimizu, Sakihae Institute, Gifu.
- Issue online: 26 JUL 2011
- Version of Record online: 1 APR 2011
- Accepted manuscript online: 30 NOV 2010 10:47AM EST
- Manuscript Accepted: 3 NOV 2010
- Manuscript Received: 23 JUL 2010
- Grants-in-Aid for Scientific Research for Young Scientists (A) (Ministry of Health, Labor, and Welfare of Japan, Japan Society for the Promotion of Science). Grant Number: 19689014
- Grant-in-Aid for Cancer Research (19 shi-2)
- Third Term Comprehensive 10-Year Strategy for Cancer Control (H21-Sanjigan-Ippan-003) (Ministry of Education, Culture, Sports, Science, and Technology of Japan)
- Management Expenses Grants from the Government to the National Cancer Center
- n-3 polyunsaturated fatty acids;
- n-6 polyunsaturated fatty acids;
- colorectal cancer;
- prospective study;
To date, epidemiologic studies investigating intake of n-3 and n-6 polyunsaturated fatty acids and risk of colorectal cancer are limited, and results remain inconsistent. This is the first prospective study to show the association by subsite (proximal colon, distal colon, rectum). To clarify the role of n-3 and n-6 polyunsaturated fatty acids intake in colon carcinogenesis, we conducted a large, population-based prospective study, characterized by high fish consumption and a wide range of n-3 polyunsaturated fatty acids intakes. Subjects were followed from response to a lifestyle questionnaire in 1995–1999 through 2006. During 827,833 person-years of follow-up (average 9.3 years), we identified 1,268 new colorectal cancer cases (521 colon and 253 rectal for men; 350 colon and 144 rectal for women). Compared to the lowest quintile, the relative risk and 95% confidence interval of developing cancer among the fifth quintile of marine n-3 polyunsaturated fatty acids intake were 0.60 and 0.31–1.14, respectively (p for trend = 0.04) in the colon in women and 0.35 and 0.14–0.88 (p for trend = 0.05) and 1.82 and 0.79–4.20 (p for trend = 0.16) in the proximal and distal colon, respectively, in men. For rectal cancer, the dose response for marine n-3 polyunsaturated fatty acids s was unclear; rather, we observed U-shaped associations in men and women. We found no evidence that n-6 polyunsaturated fatty acids increases or the n-3/n-6 ratio decreases the risk of colorectal cancer. Our results suggest that intake of marine n-3 polyunsaturated fatty acids may be inversely related to the risk of cancer in the proximal site of the large bowel.
Conclusive evidence for the effectiveness of n-3 fatty acids (FAs) against cancer has not been obtained and the mechanisms by which the FAs contribute to the prevention of various cancers have not been fully established.1 The hypothesized mechanisms regarding the possible role of n-3 polyunsaturated fatty acids (PUFAs) in the etiology of colon carcinogenesis includes an anti-inflammatory effect through the suppression of arachidonic acid (AA; n-6 PUFA)–derived proinflammatory eicosanoids and influences on transcription factor activity, gene expression, signal transduction, production of free radicals and reactive oxygen species, insulin sensitivity and so on.2 Owing to the proposed competitive role of n-3 and n-6 PUFAs through inflammation, the composition of these PUFAs was suggested to be a biologically plausible target.
However, an advisory report from the American Heart Association now recognizes that n-6 FAs are a beneficial part of a heart-healthy eating plan.3 According to the report, based on a tracing study, the extent of conversion of linoleic acid (LA; n-6 PUFA) to AA is very low (about 0.2%) and human studies showed that high plasma levels of n-6 PUFAs (mainly AA) were associated with decreased serum proinflammatory markers, particularly interleukin (IL)-6 and IL-1 receptor antagonists, and increased levels of anti-inflammatory markers, particularly transforming growth factor-β. Mechanisms of improving insulin resistance, reducing the risk of diabetes and lowering blood pressure are also presented. Through the preceding mechanisms, especially anti-inflammation or improvement in insulin sensitivity, it is possible that n-6 PUFA has a rather beneficial effect on the risk of developing colorectal cancer (CRC).
To date, epidemiologic studies investigating intake of n-3 and n-6 PUFAs and risk of CRC are limited, and results remain inconsistent.4 This may be due to inaccuracy in dietary assessment, an insufficient amount or variety of intake, possible potentially carcinogenic substances in fish, and so on. One of the studies is our previous report on fish, marine n-3 PUFAs such as eicosapentaenoic acid (EPA) or docosahexaenoic acid (DHA) and CRC risk; in this cohort, we used a concise baseline questionnaire, including 44 or 52 food items, and found no association at all.5 In this report, we used a more detailed questionnaire (138 food items) at the 5-year follow-up to investigate the association between n-3 and n-6 PUFAs as a whole, as well as specific PUFAs, such as DHA, EPA, docosapentaenoic acid (DPA), and α-linolenic acid (ALA). The substantial growth in number of observed cases (705 to 1,268) also enabled us to conduct a more informative analysis. In this large, population-based prospective study using a detailed dietary assessment tool, we attempted to clarify the role of n-3 and n-6 PUFAs in colon carcinogenesis.
Material and Methods
The Japan Public Health Center (JPHC)–Based Prospective Study was started in 1990 for cohort I and in 1993 for cohort II. Subjects were all registered Japanese residents in 11 public health center areas who were aged 40–69 years at the beginning of the baseline survey. Details of the study design have been described previously.6 The institutional review board of the National Cancer Center, Tokyo, Japan, approved the study. The participants in our study were subjects in the JPHC study who responded to the 5-year follow-up questionnaire covering lifestyle factors, including food intake, in 1995–1999 at ages 45–74 years. This follow-up survey was used as the starting point in our study. We did not include one PHC area in the analysis because it had no incidence data and, therefore, we identified 133,323 subjects as the study population. After excluding subjects with non-Japanese nationality (n = 51), a late report of emigration occurring before the starting point (n = 184), ineligibility due to incorrect birth date (n = 7), or duplicate enrollment (n = 4), we established a population-based cohort of 133,077 subjects. After exclusion of 12,056 subjects who had died, moved away from the study area, or been lost to follow-up before the starting point, 121,021 subjects were left as eligible. Among them, 98,466 subjects responded to the questionnaire, yielding a response rate of 81.4%.
We asked subjects to reply to a lifestyle questionnaire that covered sociodemographic characteristics, medical history, smoking and drinking habits, diet, and so on. We designed the food frequency questionnaire (FFQ) to estimate dietary intake from 138 food items and validated it for the estimation of various nutrients and food groups.7 The FFQ asked subjects about their usual intake of 138 food items during the previous year in standard portions/units and nine frequency categories (never, 1–3 times/mo, 1–2 times/wk, 3–4 times/wk, 5–6 times/wk, once/d, 2–3 times/d, 4–6 times/d and ≥7 times/d). Nineteen items regarded fish and shellfish, including salted fish, dried fish, canned tuna, salmon or trout, bonito or tuna, cod or flat fish, seabream, horse mackerel or sardine, mackerel pike or mackerel, dried small fish, salted roe, eel, squid, octopus, prawn, short-necked clam or crab shell, vivipara, chikuwa (fish paste product) and kamaboko (fish paste product). Standard portion sizes for each food item were small (50% smaller), medium (within 50% of the standard) and large (50% larger). We calculated food intake in grams per day by multiplying frequency by standard portion size for each food item. We calculated daily intake of n-3 PUFAs and the specific PUFA, that is, EPA, DHA, DPA, ALA and n-6 PUFA, using a FA composition table of Japanese foods.8 For some food that was not included in the table, the study used a developed (substituted) FA composition table.9 EPA, DPA and DHA were summarized as marine n-3 PUFAs.
We assessed validity among subsamples using the 138-item FFQ with 14-day (for one subtropical area) or 28-day (for other areas) weighted dietary records.10 Based on 102 men and 113 women in cohort I, Spearman rank correlation coefficients between each marine n-3 PUFA intake estimated from the FFQ and intake estimated from dietary records were as follows: EPA, 0.38 and 0.45; DPA, 0.32 and 0.39; DHA, 0.34 and 0.37, for energy-adjusted value in men and women, respectively. For ALA, total n-3, and total n-6, the values were 0.27 and 0.25; 0.21 and 0.34; and 0.30 and 0.21, respectively. For the n-6/n-3 ratio, the value for the crude estimate was 0.40 and 0.37 in men and women, respectively.
We excluded subjects who had been diagnosed or reported as having cancer before the starting point (n = 4,097), who had missing data regarding FA intake (n = 1,182), or who reported extreme total energy intake (upper 2.5% or lower 2.5%). The final analysis included 88,574 subjects (41,382 men and 47,192 women).
Follow-up and identification of CRC cases
We followed subjects from the 5-year follow-up survey (1995–1999) until December 31, 2006; the average follow-up period was 9.3 years. We identified changes in residence status, including survival, annually through the residential registry in each area or, for those who had moved from the area, through the municipal office of the area to which they had moved. Mortality data for persons in the residential registry are forwarded to the Ministry of Health, Labor, and Welfare and are coded for inclusion in the national Vital Statistics database. Residency registration and death registration are required by the Basic Residential Register Law and Family Registry Law, respectively, and the registries are thought to be complete. During the follow-up period in our study, 8,040 (9.1%) subjects died, 4,167 (4.7%) moved away from the study area and 264 (0.3%) were lost to follow-up.
We identified incident data for CRC by active patient notification from major local hospitals in the study area and from data linkage with population-based cancer registries. We coded CRC cases according to the International Classification of Diseases for Oncology, 3rd edition11 (C18–C20). We conducted analyses of site-specific cancers: C18 for colon cancer (C18.0–C18.5 for proximal colon cancer and C18.6–C18.7 for distal colon cancer) and C19 and C20 for rectal cancer. According to the depth of tumor invasion, invasive cases were defined by cancer over a mucosal layer. In our cancer registry system, the proportion of cases for which information was available from death certificates only was 2.5% for CRC.
We calculated person-years of follow-up for each subject from the starting point to the date of cancer diagnosis, date of emigration from the study area, date of death, or end of the follow-up (December 31, 2006), whichever came first. We censored subjects lost to follow-up at the last confirmed date of presence in the study area. The present analysis accrued a total of 380,682 and 447,151 person-years for men and women, respectively.
We calculated relative risks (RRs) and 95% confidence intervals (CIs) of developing CRC for the categories of energy-adjusted FA consumption in quintiles for men and women separately, with the lowest consumption category as the reference. A residual model was used for energy adjustment.12 We used Cox proportional hazards models with adjustment for potential confounding variables such as age in years (<49, 50–54, 55–59, 60–64, 65–69 and >70); PHC area; body mass index (BMI; <24.9, 25–26.9, 27–29.9 and >30); smoking status (never, past and current); alcohol drinking in grams of ethanol per week (none, occasional, 1–149, 150–299, 300–449 and ≥450 g/wk); diabetes mellitus (DM; medication use or history); physical activity in metabolic equivalent task (MET)-hours per day (quartile); screening examinations (fecal occult blood test, barium enema, or colonoscopy) for CRC; and quintiles of total calorie, energy-adjusted intake of calcium, vitamin D, fiber and red meat. The effects of the interaction of each FA and selected environmental factors such as calcium, fiber and vitamin D were also assessed by adding a multiplicative interaction term to the model. We calculated trend p by assigning a median value in each category. All p values are two-sided, and statistical significance was determined at the p < 0.05 level. We performed all statistical analyses with SAS software, version 9.1 (SAS Institute, Cary, NC).
During 827,833 person-years of follow-up, we identified 1,268 new CRC cases (521 colon and 253 rectal for men; 350 colon and 144 rectal for women). Proximal and distal colon cancer developed in 213 and 281 men and 204 and 125 women, respectively.
Table 1 shows the baseline characteristics of the study subjects according to quintile of marine n-3 PUFA, total n-3 PUFA and total n-6 PUFA in men and women (only the first, third and fifth quintiles are listed). Men and women with a high intake of these PUFAs were more likely to be old; to have a history of or current use of medications for DM; to have undergone CRC screening; and to have a higher intake of vitamin D, fiber, fish, vegetables, dressing, cooking oil, fats and other oils. In general, the proportions of overweight subjects and red meat intake were higher among those with high total n-3 PUFA or n-6 PUFA intake or low marine n-3 PUFA intake. Current smoking was less frequent among those with a high intake of these PUFAs, except marine PUFAs in men. Calcium intake was distributed differently between men and women. A U-shaped distribution was seen for total energy intake among those with marine n-3 PUFA and total n-3 PUFA intakes.
Associations of marine n-3 PUFAs, ALA and total n-3 and n-6 PUFAs, and their ratios and risk of colon and rectal cancer in men and women are shown separately (Tables 2 and 3). In men, no association was found between these PUFAs and colon cancer overall (Table 2). When cancers were restricted to invasive tumors, similar results were found (data not shown). However, when cases were subdivided by tumor location, we observed statistically significant risk reduction for EPA, DPA and marine PUFA and proximal colon cancer; p values for trend were 0.008, 0.02 and 0.047, respectively. Although not statistically significant, the RR tended to decrease with ALA intake, and finally, total n-3 PUFA was statistically significantly associated with a reduced risk of proximal colon cancer. Compared to the lowest quintile, the RRs (95% CIs) of the second, third, fourth and fifth quintiles of total n-3 PUFA intake were 0.75 (0.43-1.32), 0.56 (0.29-1.08), 0.46 (0.22-0.97) and 0.42 (0.18-0.98), respectively (p for trend = 0.0496). Also, we observed reduced risk for n-6 PUFA intake (p for trend = 0.04). In contrast, no association was observed between these PUFAs and distal colon cancer. For rectal cancer, we observed the lowest RR for the third quintile across all n-3 PUFAs, which was statistically significant for EPA, 0.51 (0.28–0.92), and marine PUFA, 0.51 (0.27–0.95), respectively. A higher level of intake of these PUFAs did not further reduce the risk of rectal cancer. We observed no apparent association for total n-6 PUFA and the n-3/n-6 ratio.
For women, we observed a statistically significant reduced risk among marine n-3 PUFAs and each specific marine n-3 PUFA and colon cancer (Table 3). Compared to the lowest quintile, the RR was about half for EPA, DHA and DPA. We obtained similar results for invasive colon cancer. Similar to men, the inverse associations for marine n-3 PUFAs and their specific PUFA were more prominent in proximal colon compared to distal colon, whereas some of the trend failed to reach statistical significance; for proximal colon cancer, p values for trend of EPA, DHA, DPA and marine n-3 PUFA were 0.07, 0.13, 0.02 and 0.19, respectively. These trends became statistically significant when all colon cancers were combined. ALA showed no association with colon cancer, and finally, for total n-3 PUFAs, although a reduced risk was suggested by increased intake, it failed to reach statistical significance (p = 0.15). For rectal cancer, we observed the lowest RR for the second quintile across all n-3 marine PUFAs, but this was not statistically significant. We observed neither association for total n-3 PUFA and rectal cancer nor for total n-6 PUFA, the n-3/n-6 ratio, and colon and rectal cancer.
Further analyses to check the effect of stratification by smoking status and interaction with selected factors such as calcium, fiber, and vitamin D intakes were conducted. When subjects were limited to those who had never smoked, the RR was generally attenuated; for example, the RR and 95% CI of the highest quintile of marine n-3 PUFAs for developing rectal cancer among men became 0.79 and 0.16–3.87. However, the results are based on a relatively small sample size (61 rectal cancers), and these subjects were kept in the analysis, as shown in Table 2. Among potential factors, we selected a priori calcium and fiber as luminal modifiers, based on findings from a recent large case–control study including more than 1,000 CRC cases13 and vitamin D based on our previous report on the risk of CRC.14 For men, an interaction was suggested between calcium and n-3 PUFAs, and vitamin D and some types of marine n-3 PUFAs (EPA and DPA) among rectal cancer cases. For women, an interaction was suggested between calcium and ALA among rectal cancer cases and vitamin D and n-3 PUFAs among colon cancer cases. However, adding interaction terms to the model essentially did not alter the results.
In this large, population-based prospective study, which is characterized by high fish consumption and a wide range of n-3 PUFA intakes, marine n-3 PUFAs and specific PUFAs were inversely related to the risk of colon cancer in the proximal site in men. A similar trend was observed in women and it became statistically significant when total colon cancers were combined.
Environmental factors that could favor the development of proximal or distal colon tumors include diet, physical activity, smoking, cholecystectomy and so on.15 Fermentation reactions leading to short-chain FA production are up to eightfold higher in the proximal compared to the distal colon.16 On the other hand, levels of the promutagenic lesion O6-methyldeoxyguanosine, a marker of exposure to N-nitroso compounds, are higher in the normal distal compared to the proximal colonic DNA of patients with CRC.17 Therefore, it is possible that the main effects of PUFAs are more strongly related to the proximal than the distal parts of the large bowel, and a harmful effect, if any, is likely to appear in the distal site. To date, six cohort studies have investigated the relationship between marine n-3 PUFAs and CRC risk, and two studies showed a decreased risk by analyzing marine n-3 PUFA intake18 or serum levels of DPA and DHA,19 but not all.5, 20–22 Among them, only two studies5, 18 demonstrated the results separately for the colon and rectum, which showed similar results by site. No study separated the colon by proximal or distal regions. This is the first prospective study to show the different results by subsite.
In contrast with the linear trend shown in marine n-3 PUFAs and colon cancer, the dose response was not clear; rather, we observed a U-shaped association for rectal cancer in men and women. As mentioned earlier, some harmful effect, if any, may be apt to appear in the distal part of the large intestine. Several causes of such effects include cigarette smoking, chemical contaminants in fish, presence of DM, and heterocyclic aromatic amines. Recently, CRC has been included among tobacco-related cancers by the International Agency for Research on Cancer.23 Interestingly, the association is known to be more pronounced in rectal cancer. As stated earlier, the results were generally attenuated when subjects were limited to those who had never smoked, which suggest that smoking might explain, to some extent, the present findings. Nevertheless, when we calculated the RR with adjustment for cigarette smoking, the effect of residual confounding may still exist. However, considering the number of cigarettes per day (<20 or ≥20) for current smokers and lifetime smoking (pack-years) for ever smokers in the adjustment for smoking status did not alter the results from the original ones. Furthermore, treating age, BMI, physical activity (METs-hour) as well as nutrient data as continuous variables also did not alter the results. According to a recent report based on both separate and combined analyses of three large cohort studies in the United States, the Nurses' Health Study, the Nurses' Health Study 2, and the Health Professionals Follow-Up Study, higher intake of long-chain n-3 PUFAs and fish increases the risk of type 2 DM.24 Long-chain n-3 PUFAs can lower glucose utilization, and the authors suggest that increased circulating concentrations of glucose or interruption of insulin-signaling pathways by toxins, such as dioxins and methyl mercury, may contribute to the association. DM is a known risk factor for CRC, and it is possible that some harmful effects may exist through this mechanism. Furthermore, it is also possible that heterocyclic amines, formed as a byproduct of reactions during the cooking of fish at high temperatures, have posed a potential risk for development of CRC.25–27 A review based on a large body of literature spanning numerous cohorts from many countries and with different demographic characteristics did not provide evidence to suggest a significant association between n-3 FAs and cancer incidence.28 Our significant findings among colon cancers may be largely due to the wide range of fish intake in our population.
Regarding n-6 PUFAs, we generally observed no association but rather an inverse association for the proximal colon in men. This is in line with previous studies,29 but not all. In a meta-analysis of LA, main n-6 PUFAs and CRC risk, based on 11 case–control studies, the pooled odds ratio per 21.3 g LA intake was calculated as 0.92 (0.85–1.08) for all subjects, 1.05 (0.90–1.23) for men and 0.80 (0.66–0.98) for women.29 Similarly, based on four cohort studies, the combined RR with high compared to low intakes of LA was calculated as 0.92 (0.70–1.22). On the other hand, significant or nonsignificant increased risks posed by n-6 PUFAs were observed in studies not only for CRC20, 22 but also for breast cancer in specific situations.30, 31 Several reasons are plausible to explain the discrepancies among studies. Dietary sources of n-3 and n-6 PUFAs include several common foods and, therefore, the effects of n-6 PUFAs may be masked by those of n-3 PUFAs. Actually, on the basis of our previous investigation, the top food in cumulative percentage contribution for n-3 and n-6 PUFAs assessed by the dietary record was “vegetable oils/vegetable oil, mixed,” which contributed 25.2% and 35.2% of n-3 and n-6 PUFAs, respectively.10 Therefore, when further adjusted for ALA or n-3 PUFA intake, the association in the proximal colon among men was attenuated, and the trend was no longer significant (data not shown). In addition to cooking oil, n-6 PUFAs were contributed by many kinds of lean foods traditionally consumed among Japanese such as rice, tofu (soybean curd) and miso. This is in contrast with the contributors of n-6 PUFAs reported in the U.S. cohort: salad dressing, peanut butter and margarine.20 By means of observational study, it is actually difficult to completely eliminate the effects of other nutrients included in these foods. Furthermore, because these foods are consumed on a daily basis and might have small between-person variabilities, the association, if any, might have been attenuated. Because of the marginally decreased risk of CRC for n-3 PUFAs and no association for n-6 PUFAs, the ratio also did not indicate any association.
Our study has several possible limitations. First, we could not adjust for the effect of use of nonsteroidal anti-inflammatory drugs, which may act as anti-inflammatory agents through common pathways to n-3 PUFAs. However, considering the various mechanisms involved in PUFA intake and CRC risk, the extent of this influence might not be so large. Second, although the validity of PUFA intakes was reasonably high for each marine n-3 PUFA, the validities of ALA, total n-3 PUFAs, and total n-6 PUFA intakes were relatively low. This may be due to the difficulty of assessing “vegetable oils,” which are the main common source of these PUFAs. Therefore, caution is needed in interpreting results for these PUFAs. Third, because we conducted multiple comparisons in the analyses, including the differences between colon and rectum and between proximal and distal colon, an interaction effect, some results may be explained, in part, by chance. However, the observed associations were generally consistent for both men and women and also could be reasonably explained by the previously mentioned mechanisms.
The advantage of the study was its prospective design, which enabled us to avoid exposure recall bias. We selected subjects from the general population, we kept the sample size large, the response rate for the questionnaire was acceptable for studies of settings like this, and the number of subjects lost to follow-up was negligible. In addition, the cancer registry was of sufficient quality to reduce the misclassification of the outcome. Japanese people consume much more fish than Western individuals and the Japanese diet has a greater variation of n-3 marine PUFA intake. Based on data from 28 countries with diet recalls, weighing records, or FFQ, in comparison with the United States and most of the European countries, the mean daily intake of combined EPA and DHA among adults was about five times and two to three times higher in Japan, respectively.32 Furthermore, using random sampling methods in six regions in Japan, serum n-3 PUFA levels varied significantly by region, which corresponded to the differences in fish consumption.33 In a previous report from our study based on the baseline questionnaire, we observed no association for fish, EPA, DHA and the n-3/n-6 ratio.5 Owing to the greater detail of the FFQ used in the present analysis compared to that used for the previous report, the variation of marine-based n-3 PUFAs became larger; the median of EPA and DHA intake for the highest group in the present study almost doubled that for the previous one, and the validity of n-6 PUFAs was improved enough to conduct the analysis. The substantial growth in number of observed cases (705–1,268) also allowed us to conduct more informative analysis by subsite of colon (proximal and distal).
In conclusion, our results from a population with high fish consumption and a wide range of n-3 PUFA intakes suggest that PUFAs of marine origin may be inversely related to the risk of cancer in proximal sites of the large bowel.
The authors thank all the staff members in each study area for their painstaking efforts to conduct the survey and follow-up. The authors' responsibilities were as follows: ST was principal investigator; MI conducted the study, managed the data collection; SS helped to conduct the study, analyzed and interpreted the data, and prepared the manuscript. MI, NS, TS, TY, and RT helped to conduct the study. All authors provided critical suggestions for revision of the manuscript.
- 3Omega-6 fatty acids and risk for cardiovascular disease: a science advisory from the American Heart Association Nutrition Subcommittee of the Council on nutrition, physical activity, and metabolism; council on cardiovascular nursing; and council on epidemiology and prevention. Circulation 2009; 119: 902–7., , , , , , , , .
- 4World Cancer Research Fund/American Institute for Cancer Research. Food, nutrition, physical activity, and the prevention of cancer: a global perspective. Washington, DC: American Institute for Cancer Research, 2007. 517 p.
- 8Science and Technology Agency, eds. Fatty acids, cholesterol, vitamin E composition tables of Japanese foods (in Japanese). Tokyo: Ishiyaku Shuppan, 1990. 182 p.
- 10Validity of a self-administered food frequency questionnaire used in the 5-year follow-up survey of the JPHC Study Cohort I to assess fatty acid intake: comparison with dietary records and serum phospholipid level. J Epidemiol 2003; 13( 1 Suppl): S64–81., , , , .
- 11World Health Organization. International classification of diseases for oncology, 3rd edn. Geneva, Switzerland: World Health Organization, 2000. 490 p.
- 12Nutritional epidemiology, 2nd edn. New York: Oxford University Press, 1998. 514 p..
- 31Do both heterocyclic amines and omega-6 polyunsaturated fatty acids contribute to the incidence of breast cancer in postmenopausal women of the Malmö diet and cancer cohort? Int J Cancer 2008; 123: 1637–43., , , , , .