European Journal of Clinical Nutrition (2002) 56, Suppl 1, S19–S24 ß 2002 Nature Publishing Group All rights reserved 0954–3007/02 $25.00 www.nature.com/ejcn ORIGINAL COMMUNICATION consumption and of the large bowel{

AS Truswell1*

1Human Nutrition, The University of Sydney, Sydney, Australia

Since the major reviews on by the World Cancer Research Fund (WCRF) and by the British Department of Health’s Committee on Medical Aspects of Food Policy (COMA) in 1997 and 1998, additional epidemiological studies relating (red) meat consumption and colorectal cancer have been published or found by search. These are collected here. Thirty adequate case – control studies have been published up to 1999 (from 16 different countries). Twenty of them found no significant association of (red) meat with colorectal cancer. Of the remaining 10 studies reporting an association, some obtained statistical significance only in rectal or colon , another only in men, not women, or found a stronger association with pasta and rice, or used an inadequate food list in the food frequency questionnaire. Fifteen cohort studies have now been published.{ Only in three were significant associations of (red) meat found with colorectal cancer. Two of these positive studies were from the same group in the USA (relative risk 1.7). The results of the third positive study appear to conflict with data from part of the vegetarians follow up mortality study. Here, five groups of vegetarians (in three different countries) with socially matched controls were followed up (total 76 000 people). Mortality from colorectal cancer was not distinguishable between vegetarians and controls. While it is still possible that certain processed or (with a variety of added ingredients) or meats cooked at very high temperature carry some risk, the relationship between meats in general and colorectal cancer now looks weaker than the ‘probable’ status it was judged to have by the WCRF in 1997. European Journal of Clinical Nutrition (2002) 56, Suppl 1, S19 – S24

Keywords: colorectal cancer; meat; case – control studies; cohort studies; vegetarians

In 1978 four of us wrote the Conclusions from the Marabou The question of meat consumption and colorectal cancer symposium on Food and Cancer. It included this: ‘There is became highly controversial in 1997 when the Chairman of some epidemiological evidence in international comparisons the World Cancer Research Fund (WCRF) singled out meat as suggesting that consumption is correlated with cancer a dietary risk factor when the WCRF’s thorough and impor- of the large intestine; frying or broiling could produce a tant review (World Cancer Research Fund, 1997) was carcinogenic product but this evidence is not firm enough to launched. It is a difficult judgement whether the evidence justify recommendations to the general public. Cancer of the in that review amounts to a ‘probable’ or ‘possible’ association rectum should be distinguished from the more common site of with colorectal cancer. The judgement of WCRF in the colon. Evidence associating cancer of the rectum with was ‘probable’, although I would say only ‘possible’ from beer consumption is inconsistent’ (Truswell et al, 1978). The thorough scrutiny of what is in the report. Then the final consensus remained much the same as this until 1997, ie recommendations from the 670 page review highlighted the possibly a relationship but more research needed and not risk of meat as if it were a new discovery: ‘If eaten at all, limit enough evidence to make negative recommendations to the intake of red meat to less than 80 grams (3 ounces) daily. It is general public. preferable to choose fish, poultry or meat from non-domes- ticated animals in place of red meat.’ In the summary matrix in the WCRF report (yellow or green blocks on grey squares) the association with colorectal *Correspondence: AS Truswell, Human Nutrition Unit, Biochemistry cancer was the only one considered ‘probable’ for meat (four Building, G08, The University of Sydney, Sydney, NSW 2006, Australia. E-mail: [email protected] other types of cancer were judged to have only ‘possible’ {This paper was presented at Erna¨hrungsforum 2000, Fleischverzehr: associations). Evolution und Fortschritt, Hamburg, 17 – 18 October 2000. A British expert COMA committee reported a few months { A 15th cohort study by Pietinen et al (1999) should be added as a later (Department of Health, 1998). They concluded that postscript to Table 2. With 27 111 subjects in Finland and 185 colorectal cancer cases they found no association with red meat, fried meat or all there is moderate evidence that lower red meat and pro- meats in the ATBC trial (Cancer Causes and Control 10, 387 – 396). cessed meat consumption would reduce the risk of colorectal Meat consumption and cancer AS Truswell S20 cancer. However, they thought the evidence that lower set out more clearly in this review than in the others. It is a consumption of meat would reduce the risk of breast useful resource. Baghurst et al, conclude ‘To the extent that cancer, lung cancer, prostate cancer and pancreatic cancer the consumption of large amounts of meat may simply was weak. They recommended that adult individual’s con- reduce the opportunities for an individual to eat a wide sumption of red and processed meat should not rise, and variety of foods from plant sources’ (associated with reduced higher consumers should consider a reduction; as a conse- cancer risk) ‘a high meat diet may be associated with a higher quence of this the population average would fall. The com- risk of cancer. However, the evidence, both epidemiologic mittee were concerned that any general recommendations and mechanistic, that red meat is intrinsically carcinogenic regarding red or processed meat should not compromise and deserves more attention than any other food as a public those for whom an intake of red meat, in moderation, is health concern, is very weak’. making an important contribution to micronutrient status. All the published epidemiological studies of meat and In a figure in the COMA report’s section on colorectal colorectal cancer that I could find are reviewed below. cancer 10 of the 17 case – control studies shown had no significant association with meat and in another two studies risk of rectal cancer was higher and risk of colorectal cancer The case – control studies lower than 1.0. Although not clearly set out, the text states Neither the WCRF nor the COMA reviews included all the that five cohort studies reported no association and two published epidemiological studies and several have appeared reported a significant association of colorectal cancer with since they were written, the largest (in number of subjects) in red meat. 1999. Table 1 contains 30 case – control studies that had at At about the same time as these two well-publicised least 99 cases and reported on meat intake (Benito et al, 1990; general reviews on diet and cancer were published, P Bidoli et al, 1992; Centonze et al, 1994; Dales et al, 1979; Baghurst, an Australian epidemiologist, and colleagues com- Gerhardsson et al, 1991; Graham et al, 1978; Haenszel et al, pleted a thorough review devoted to red meat and cancer (at 1973, 1980; Hu et al, 1991; Iscovich et al, 1992; Kampman any site, Baghurst et al, 1997). The data are summarised and et al, 1995, 1999; Kato et al, 1990; Kune et al, 1987; La Vecchia

Table 1 The main case – control studies reporting meat and colo-rectal cancer

Author Number of cases Country WCRF COMA Result

Benito et al (1990) 286 Majorca Yes — C and R meat Bidoli et al (1992) 248 Italy Yes I R þ beef & ; C-ve Centonze et al (1994) 118 Italy Yes H NS Dales et al (1979) 99 US Blacks — INS Gerhardsson et al (1991) 559 Sweden Yes H C and R þ meats Graham et al (1978)) 586 USA (NY) — INS Haenszel et al (1973) 179 Japanese in Hawaii — L CRC þ meat Heanszel et al (1980) 588 Japan ——NS Hu et al (1991) 336 China Yes — NS Iscovich et al (1992) 110 Argentina Yes H NS Kampman et al (1995) 312 Netherlands Yes H NS Kampman et al (1999) 1542 USA Multicentre ——NS Kato et al (1990) 221 Japan — INS Kune et al (1987) 715 Australia Yes — C þ beef men, not women R þ lamb women, not men La Vecchia et al (1988) 575 N. Italy Yes — RR 2.0 CRC for meat but 3.0 for pasta þ rice La Vecchia et al (1996) 1326 N. Itlay — H CRC þ (1.6) for red meat Lee et al (1989) 203 Singapore Yes H NS Lyon et al (1988) 246 Utah, USA Yes — NS Macquart-Moulin et al (1986) 399 France Yes I NS Manousos et al (1983) 100 Greece Yes — CRC þ beef and lamb Miller et al (1983) 542 Canada Yes — R þ beef (?) Muscat et al (1994) 511 USA ——NS Peters et al (1989) 146 Los Angeles, USA Yes — NS Peters et al (1992) 746 Los Angeles, USA Yes H NS Steinmetz et al (1993) 200 Australia Yes H NS Tajima et al (1985) 100 Japan ——NS Tuyns et al (1988) 818 Belgium Yes I C þ beef; R not Van den Brandt et al (1990) 656 Netherlands ——NS Young & Wolf (1988) 554 Wisconsin, USA Yes — NS Zaridze et al (1993) 217 Russia Yes I NS

30 12 958 16 20 16

C colon cancer; R rectal cancer; NS not statistically significant; L low; I intermediate; H high quality score; CRC colorectal cancer.

European Journal of Clinical Nutrition Meat consumption and cancer AS Truswell S21 et al, 1988, 1996; Lee et al, 1989; Lyon & Mahony, 1988; The cohort studies Macquart-Moulin et al, 1986; Manousas et al, 1983; Miller et al, Cohort studies are likely to be more reliable than case – 1983; Muscat & Wynder, 1994; Peters et al, 1989, 1992; control studies. The same dietary assessment method is Steinmetz & Potter, 1993; Tajima & Tominaga, 1985; Tuyns used for all — those that turn out to be cases and the much et al, 1988; van der Brandt et al, 1990; Young & Wolf, 1988; larger number who are controls. Recording of the antecedent Zaridze et al, 1993). There are other case – control studies that diet is far less dependent on memory and cannot be distorted did not report meat intake or had fewer than 99 cases. in response to the diagnosis. There are 14 cohort studies Numbers of cases are shown in the table. The studies usually (Hirayama, 1990; Phillips & Snowdon, 1985; Heilbrun et al, had more controls than cases. Most report red meat or beef; 1989; Willett et al, 1990; Thun et al, 1992; Bostick et al, 1994; some report meat (total). All the studies estimated food Giovannucci et al, 1994; Goldbohm et al, 1994; Knekt et al, intakes with (different) food frequency questionnaires. 1994; Gaard et al, 1996; Kato et al, 1997; Hsing et al, 1998; Of these 30 good sized studies, 10 are not in the WCRF Singh & Fraser, 1998; Jensen et al, 1999) in Table 2 based on a review but its table of case – control studies contains three total of 1 492 308 subjects. This is a larger set than seven studies with 50 or fewer cases and four studies whose out- cohort studies for meat and colorectal cancer in the WCRF come was adenomas or polyps, not carcinoma. Fourteen of review and nine cohort studies reporting fresh meat or the studies in Table 1 are not in the British COMA meta- processed meat in the British COMA review. analysis figure of case – control studies (Figure 5.21). I have In Table 2 there were significant associations of red meat, omitted one study in that table (Martinez) which scored only not white meat, with colon cancer in US nurses (Willett et al, 2=10 for dietary assessment (UK Nutritional Epidemiology 1999) and in US health professional men (Giovannucci et al, Group, 1993). 1994). Both these studies were made by the same team, As can be seen in the table, 20 of the 30 studies found no headed by Professor Willett at Harvard. The third cohort significant association of red meat or meat with colorectal showing a significant association was the Adventist Health cancer. In at least seven of them the relative risks were under Study (Singh & Fraser, 1998) and association was with red 1.00. Of the 10 studies reporting an association, three and white meats. The top quintile in the Nurses cohort (all obtained statistical significance only if colon cancer was females) was 134 g or more of red meat per day; in the US separated from rectal cancer, another only if men were Health Professionals (men) it was averaged 129 g of red meat separated from women. La Vecchia et al, found a stronger per day but in the Adventists (Singh and Fraser) the top association of pasta and rice with colorectal cancer than beef tertile was (any) meat one or more times per week. The and veal! In the remaining five studies with positive associa- Adventist Health Study participated in the follow-up study tions, two used hospital controls, which are probably not as of five vegetarian groups discussed below (Key et al, 1998). In reliable as matched community controls (used by the major- that report the number of subjects is given as 28 952. GE ity of studies) and one used a food frequency questionnaire Fraser is the second author. It is surprising that the mortality with only 29 items. To conclude on the case – control studies, rate from colorectal cancer was 1.01 against omnivores, with the majority found no significant association of red meat or 104 cases. This alternative look at the data raises doubts meat and colorectal cancer, and the minority that reported about the meat results in Singh and Fraser’s paper. some significant association do not seem to suggest that the In the other 11 studies in Table 1 cancer rates were lower studies with null findings were flawed. in meat eaters in four studies (Hirayama, 1990; Phillips &

Table 2 Cohort studies reporting meat intake and colon or colorectal cancer

Study Results Subjects Cancer cases

Hirayama (1990) Japan Meat significantly down 265 000 726a Phillips & Snowdon (1985) USA, California SDAs Meat NS, slightly down 25 493 215a Heilbrun et al (1989) Hawaii Red meat NS, even 8 006 162b Willett et al (1990) US nurses Red meat RR 1.77, significant 88 751 150b Thun et al (1992) American Cancer Society Red meat NS, beef slightly down 764 343 1150b Bostick et al (1994) IOWA women, USA Red meat NS trend even or down 32 215 212b Giovannucci et al (1994) US health proefssionals Red meat RR 1.71, significant 49 949 205b Goldbohm et al (1994) Netherlands Meats NS fresh meats trend down 120 852 215b Knekt et al (1994) Finland men Fried meat NS trend even 9 990 73a Gaard et al, (1996) Norway Meat NS (excluding ), trend down in men, up in women 50 535 143b Kato et al (1997) USA New York women Red meat NS (RR 1.23) 14 727 100a Hsing et al (1998) USA Lutheran Brotherhood Red meat NS trend up 17 633 145a Singh & Fraser (1998) USA California SDAs Total, red and white meat all significantly up 32 051 157b Jensen et al (1999) Seven Countries Study Meat NS (Rr 1.03) 12 763 165b

aColorectal cancer; bcolon cancer. RR, relative risk.

European Journal of Clinical Nutrition Meat consumption and cancer AS Truswell S22 Snowdon 1985; Thun et al, 1992; Goldbohm et al, 1994) and friend or relative who ate meat. The fifth group of vegetar- significantly so in the Hirayama cohort (Hirayama, 1990). In ians was a relatively small one in Heidelberg, Germany. In the remaining seven cohorts the cancer rates were even or the combined series of 76 000 people there were significantly non-significantly up in subjects with the higher quantiles of fewer deaths from ischaemic heart disease in the vegetarians red meat or meat intake. (although they smoked less). However their average death The British COMA report groups processed meats with red rate ratio for colorectal cancer was 0.99, indistinguishable meat. Willett et al (1990) and Giovannucci et al (1994) both from 1.00 in the vegetarians. This combined prospective reported an association of ‘processed meats’ with colorectal study is different from the cohort studies in Table 2. In cancer but the relative risks were lower than for red meat and many of those studies degree of meat eating was based on in the men the probability was just over 0.05. In the Nether- a single food frequency questionnaire. Avowed vegetarians lands cohort (Goldbohm et al, 1994) a significant association are more committed long term. Here the statistics are not with processed meat was also seen (not seen with other about quantities of estimated meat intake but simply yes=no meat). This was thought to be mainly sausages. As the about meat eating. authors say ‘Processed meat differs from fresh meat in that To sum up, most case – controls have not found signifi- it has been cured with the addition of (salt, cantly more colorectal cancer in higher meat consumers. nitrate and smoke) and other additives ... it may be that, in Eleven of 14 cohort studies have not found significantly this population, the circumstances in which processed meats more colorectal cancer in high meat eaters, and a large are eaten, as a sandwich without vegetables and often with- combined follow-up of five groups of vegetarians found the out fruits in the same meal are important determinants for overall death rate from colorectal cancer no different from the risk’ (Goldbohm et al, 1994). Processed meats were not that in socially matched controls. reported as significantly associated with colorectal cancer in After a workshop in Adelaide, Australia in December 1998 any of the other cohort studies. (Report of an Expert Workshop, 1999; which I chaired) with One must conclude that most cohort studies have not epidemiologists, anti-cancer association officials, colorectal found more colorectal cancer in people who were originally researchers and public health nutritionists, the 16 partici- recorded as eating more red meat or meat. The reason why pants all signed this consensus statement: ‘The workshop the two Harvard cohorts (Willett et al, 1990; Giovannucci et confirms the nutritional benefits of eating a balanced and al, 1994) stand out as different is not clear but the red meat varied diet. After reviewing the evidence the consensus view intake of the US Nurses top quintile is very large, at 134 g per was that the consumption of a moderate amount of meat, as day or more, and only this top quintile was significantly part of a diet containing adequate cereals and grain foods, associated with colorectal cancer. The relative risks in vegetables and fruit, is not associated with an increased risk ascending quintiles of red meat intake were: 1.0, 1.16, of bowel cancer. Meat is a good source of iron, vitamin B-12, 1.25, 1.13 and 1.77. protein, zinc and other nutrients. Avoid charring and heavy browning when frying or grilling meat, poultry or fish.’ This last sentence is a sort of insurance policy. We know that very small amounts of carcinogenic and mutagenic Five combined prospective studies of vegetarians heterocyclic amines and polycyclic aromatic hydrocarbons and socially matched controls are produced on the surface of overcooked meat — not only In general, randomised controlled trials are more reliable red meat (Barrington et al, 1990; Thomson, 1999), also pork than cohort studies. Yet it is never going to be possible to (Sinha et al, 1998), but also chicken (Sinha et al, 1995) and have a randomised controlled trial in which a large group of fish (Yamaizumi et al, 1980). However, their calculated daily people are asked to eat (red) meat, who they are being intakes in humans are of the order of 1000 – 5000 times decided by random numbers, and the same number of smaller than those found to induce cancer in animals other people are asked not to eat meat — and this is done (Baghurst, 1999; Thomson, 1999). in a community at risk of developing colorectal cancer, and In the most thorough epidemiological study of estimated continued for five or more years in thousands of people. The temperatures of cooking protein-rich foods and its relation nearest we can ever come to a randomised controlled trial of to cancer, Augustsson et al, (1999) in Sweden found that meat eating is a controlled follow-up of vegetarians, but the people with colorectal cancer or rectal cancer did not appear control subjects should be matched for social, educational to have ingested more heterocyclic amines than controls. and economic status. So a collaborative analysis of five The Australian National Health and Medical Research prospective studies of vegetarians in USA, UK and Germany Council published Guidelines for the Prevention, Early (Key et al, 1998) should provide valuable insight. Two of the Detection and Management of Colorectal Cancer in 1999 groups were Seventh Day Adventists in California followed (National Health and Medical Research Council, 1999), the over two different time periods: vegetarian and non-vegetar- work of several panels of experts. After considering the report ian. Thirdly, health food shoppers in Britain were surveyed of our workshop and the result of the Australian Polyp to find who ate meat and who did not. Fourthly, members of Prevention Trial (MacLennan et al, 1995), they removed the UK Vegetarian Society were asked to recruit as a control a meat from their list of dietary risk factors. The inconsistent

European Journal of Clinical Nutrition Meat consumption and cancer AS Truswell S23 evidence about meat intake and colorectal cancer is dis- Centonze S, Boeing H, Leoci C, Guerra V & Misciagna G (1994): cussed in the text. Dietary habits and colorectal cancer in a low-risk area. Results from a population-based case – control study in southern Italy. In the Australian Polyp Prevention Project (MacLennan Nutr. Cancer 21: 233 – 246. et al, 1995), 400 patients who had had one or more colorectal Dales LG, Friedman GD, Ury HK, Grossman S & Williams SR (1979): A polyps removed were followed for 4 y on combinations of six case – control study of relationships of diet and other traits to color- different diets (low fat or not; and b-carotene or not; and ‘All ectal cancer in American blacks. Am. J. Epidemiol. 109:132– 144. Department of Health (1998): Report of the Working Group on Diet Bran’ or not). Patients on extra wheat bran plus low fat (but and Cancer of the Committee on Medical Aspects of Food and no restriction of lean red meat) did not have recurrence at Nutrition Policy. Nutritional Aspects of the Development of Cancer colonoscopy of large adenomas, which happened on the London: The Stationery Office. other diets in some patients. In this intervention trial the Gaard M, Tretli S & Løken EB (1996): Dietary factors and risk of colon cancer: a prospective study of 50,535 young Norwegian men and effect of fat on large intestine neoplasms was thus separated women. Eur. J. Cancer Prev. 5: 445 – 454. from that of meat. Gerhardsson DVM, Hagman U, Peters RK, Steineck G & Overvik E An expert workshop in New Zealand chaired by Professor (1991): Meat, cooking methods and colorectal cancer: a case- C Tasman Jones (Report of Workshop, 2000; with 24 partici- referent study in Stockholm. Int. J. Cancer 49: 520 – 525. Giovannucci E, Rimm EB, Stampfer MJ, Colditz GA, Ascherio A & pants) agreed to a similar statement to the Australian one: Willett WC (1994): Intake of fat, meat and fiber in relation to risk ‘There is no convincing evidence from published epidemio- of colon cancer in men. Cancer Res. 54, 2390 – 2397. logical studies that moderate intakes of lean red meat Goldbohm RA, van den Brandt PA, van’t Veer P, Brants , Dorant increase the risk of colorectal cancer when eaten as part of E, Sturmans F & Hermus RJJ (1994): A prospective cohort study on the relation between meat consumption and the risk of colon a mixed diet including carbohydrates (breads and cereals) cancer. Cancer Res. 54, 718 – 723. vegetables and fruit, and dairy products. Red meat is a good Graham S, Dayal H, Swanson M, Mittelman A & Wilkinson G (1978): source of protein and micronutrients, especially iron, zinc Diet in the epidemiology of cancer of the colon and rectum. J. Natl. Cancer Inst. and vitamin B-12. Products of metabolism and some meth- 61, 709 – 714. Haenszel W, Berg JW, Segi M, Kurihara M & Locke FB (1973): Large ods of cooking meat (eg, charring, grilling) produce chemi- bowel cancer in Hawaii Japanese. J. Natl. Cancer Inst. 51, 1765 – cals with the potential to damage DNA. The relevance of 1779. these chemicals to colorectal cancer needs clarification.’ Haenszel W, Locke FB & Segi M (1980) A case – control study of large bowel cancer in Japan. J. Natl. Cancer Inst. 64,17– 22. The Australian Cancer Council published a popular cook- Heilbrun LK, Normura A, Hankin JH & Stemmerman GN (1989): Diet book in 1999 (Leng et al, 1999). On its front cover it has three and colorectal cancer with special reference to fiber intake. Int. J. pieces of lightly cooked, low-fat red meat surrounded by Cancer 44,1– 6. several different lightly cooked vegetables on a tortilla base. Hirayama T (1990): Lifestyle and Mortality: a Large-scale Census-based Study in Japan. Basel: Karger. According to the recipe, it provides 112 – 150 g red meat per Hsing AW, McLaughlin JK, Chow WH, Schuman LM, Chien HTC, person. The introductory press release opens. ‘The Australian Gridley G, Bjelke E, Wacholder S & Blot WJ (1998): Risk factors for Cancer Society today launched the Lifetrack Cookbook that colorectal cancer in a prospective study among US white men. Int. blows away a few myths, including that eating red meat J. Cancer 77,549– 553. Hu JF, Liu YY, Yu YK, Zhao TZ, Liu SD & Wang QQ (1991): Diet and causes cancer.’ cancer of the colon and rectum: a case-control study in China. Int. J. Epidemiol 20, 362 – 367. Iscovich JM, L’Abbe KA, Castelleto R, Calzona A, Bernedo A & Chopita NA (1992): Colon cancer in Argentina. I. Risk from References intake of dietary items. Int. J. Cancer 51,851– 857. Augustsson K, Skog K, Ja¨gerstad M, Dickman PW & Steineck G. Jensen MCJF, Bueno-de-Mesquita HB, Buzina R, Fidanza F, Menotti A, (1999): Dietary heterocyclic amines and cancer of the colon, Blackburn H, Nissinen AM, Kok FJ & Kromhut D (1999): Dietary rectum, bladder, and kidney: a population-based study. Lancet fiber and plant foods in relation to colorectal cancer mortality: the 353, 703 – 707. Seven Countries Study. Int. J. Cancer 81, 174 – 179. Baghurst PA (1999): Polycyclic aromatic hydrocarbons and hetero- Kampman E, Verhoeven D, Sloots L & van’t Veer P (1995): Vegetable cyclic amines in the diet: the role of red meat. Eur. J. Cancer Prev. 8: and animal products as determinants of colon cancer risk in Dutch 193 – 199. men and women. Cancer Causes Control 6, 225 – 234. Baghurst P, Record S & Syrette J (1997): Does red meat cause cancer? Kampman E, Slattery ML, Bigler J, Leppert M, Samowitz W, Caan BJ & Aust. J. Nutr. Diet 54(Suppl 4), S1 – S44. Potter JD (1999): Meat consumption, genetic susceptibility, and Barrington PJ, Baker RSU, Truswell AS, Bonin AM, Ryan AJ & Paulin colon cancer risk: a United States multicenter case – control study. AP (1990): Mutagenicity of basic fractions derived from lamb and Cancer Epidemiol. Biomarkers & Prev. 8,15– 24. beef cooked by common household methods. F. Chem. Toxiol. 28: Kato I, Tominaga S, Matsuura A, Yoshii Y, Shirai M & Kobayashi SA 141 – 146. (1990): A comparative case-control study of colorectal cancer and Benito E, Obrador A, Stiggelbout A, Bosch FX, Mulet M & Munoz NA. adenoma. Jpn J. Cancer Res. 81, 1101 – 1108. (1990): A population-based case – control study of colorectal Kato I, Akhmedkhanov A, Koeing K, Toniolo PG, Shore RE & Riboli E. cancer in Majorca. I. Dietary factors. Int. J. Cancer 45:69– 76. (1997): Prospective study of diet and female colorectal cancer: the Bidoli E, Franceschi S, Talawini R, Barra S & La Vecchia C. (1992): New York University Women’s Health Study. Nutr. Cancer 28, Food consumption and cancer of the colon and rectum in North- 276 – 281. Eastern Italy. Int. J. Cancer 50: 223 – 229. Key TJ, Fraser GE, Thorogood M, Appleby PN, Beral V, Reeves G, Burr Bostick RM, Potter JD, Kushi LH, Sellers TA, Steinmetz KA, McKenzie ML, Chang-Claude J, Frentzel-Beyme R, Kuzma JW, Mann JI & DR, Gapstur SM & Folsom AR (1994): Sugar, meat, and fat intake McPherson K (1998): Mortality in vegetarians and non-vegetarians: and non-dietary risk factors for colon cancer incidence in Iowa a collaborative analysis of 8300 deaths among 76,000 men and women (United States). Cancer Cause Control 5:38– 52. women in five prospective studies. Public Health Nutr. 1,33– 41.

European Journal of Clinical Nutrition Meat consumption and cancer AS Truswell S24 Knekt P, Steineck G, Ja¨rvinen R, Hakulinen T & Aromaa A (1994): Singh PN & Fraser GE (1998): Dietary risk factors for colon cancer in a Intake of fried meat and risk of cancer: a follow-up study in low-risk population. Am. Epidemiol. 148, 761 – 774. Finland. Int. J. Cancer. 59, 756 – 760. Sinha R, Knize MG, Salmon CP, Brown ED, Rhodes D, Felton JS, Kune S, Kune GA & Watson LF (1987): Case – control study of dietary Levander OA & Rothman N (1998): Heterocyclic amine content of etiological factors: the Melbourne colorectal cancer study. Nutr. pork products cooked by different methods and to varying degrees Cancer. 9,21– 42. of doneness. Food Chem. Toxicol. 36, 289 – 297. La Vecchia C, Negri E, Decarli A, D’Avanzo B & Galotti L (1988): A Sinha R, Rothman N, Brown ED, Salmon CP, Knize MG, Swanson CA, case – control study of diet and colo-rectal cancer in northern Italy. Rossi SC, Mark SD, Levander OA & Felton JS (1995): High con- Int. J. Cancer 41, 492 – 498. centrations of the carcinogen 2-amino-1 methyl-6-phenylimi- La Vecchia C, Ferranoni M, Mezzetti M, Enard L, Negri E & Franceschi dazo-(4,5) pyridine (PhIP) occur in chicken but are dependent S (1996): Attributable risks for colorectal cancer in northern Italy. on the cooking method. Cancer Res. 55, 4516 – 4519. Int. J. Cancer. 66,60– 64. Steinmetz KA & Potter JD (1993): Food group consumption and Lee HP, Gourley L, Duffy SW, Esteve J, Lee J & Day NE (1989): colon cancer in the Adelaide case-control study. II. Meat, poultry, Colorectal cancer and diet in an Asian population — case — con- seafood, dairy foods and eggs. Int. J. Cancer. 53, 720 – 727. trol study among Singapore Chinese. Int. J. Cancer 43, 1007 – 1016. Tajima K & Tominaga S (1985): Dietary habits and gastro-intestinal Leng V et al (1999): Eat! Enjoy! The Lifetrack Cookbook. Melbourne: cancers: a comparative case – control study of stomach and large Australian Cancer Society and Anti-Cancer Council of Victoria. intestinal cancers in Nagoya, Japan. Jpn. J. Cancer Res. 76, 705 – Lyon JL & Mahony AW (1988): Fried foods and risk of colon cancer. 716. Am. J. Epidemiol. 128, 1000 – 1006. Thomson BM (1999): Heterocyclic amine levels in cooked meat and MacLennan R, Macrae F, Bain C, Battistutta D, Chapuis P, Gratten H, the implication for New Zealanders. Eur. J. Cancer Prev. 8, 201 – Lambert J, Newland RC, Ngu M, Rusell A, Ward M & Wahlqvist ML 206. (1995): Randomized trial of intake of fat, fiber, and beta carotene Thun MJ, Calle EE, Namboodiri MM et al (1992): Risk factors for fatal to prevent colorectal adenomas. J. Natl. Cancer Inst. 87, 1760 – colon cancer in a large prospective study. J. Natl. Cancer Inst. 84, 1766. 1491 – 1499. Macquart-Moulin G, Riboli E, Cornee J, Charnay B, Berthezene P & Truswell AS, Asp N-G, James WPT & MacMahon B (1978): Conclu- Day N (1986): Case – control study on colorectal cancer and diet in sions of Marabou Symposium ‘Food and Cancer’. Nutr. Rev. 36, Marseilles. Int. J. Cancer. 38, 183 – 191. 313 – 314. Manousas O, Day NE, Trichopoulos D, Gerovassilis F & Tzonou A Tuyns AJ, Kaaks R & Haelterman M (1988): Colorectal cancer and the (1983): Diet and colorectal cancer: a case-control study in Greece. consumption of foods: a case-control study in Belgium. Nutr. Int. J. Cancer. 32,1– 5. Cancer. 11, 189 – 204. Miller AB, Howe GR, Jain M, Craib KJ, Harrison L (1983): Food items UK Nutritional Epidemiology Group (1993): Diet and Cancer. A and food groups as risk factors in a case-control study of diet and Review of the Epidemiological Literature. London: The Nutrition colo-rectal cancer. Int. J. Cancer. 32, 155 – 161. Society. Muscat JE & Wynder EL (1994): The consumption of well-done red van den Brandt PA, Goldbohm RA, Van Loon AJ & Kok FJ (1990): meat and the risk of colorectal cancer. Am. J. Public Health 84, Cross-sectional versus longitudinal investigations of the diet- 856 – 858. cancer relation. Epidemiology 1, 402 – 404. National Health & Medical Research Council (1999): The Prevention, Willett WC, Stampfer MJ, Colditz GA, Rosner BA & Speizer FE (1990): Early Detection and Management of Colorectal Cancer. Clinical practice Relation of meat, fat and fiber intake to the risk of colon cancer in guidelines. Canberra: NH&MRC. a prospective study among women. New Engl. J. Med. 232, 1664 – Peters RK, Garabrant D, Yu M & Mack T (1989): A case – control study 1672. of occupational and dietary factors in colorectal cancer in young World Cancer Research Fund (1997): Food, Nutrition and the Prevention men by subsite. Cancer Res. 49, 5459 – 5468. of Cancer: a Global Perspective. Washington, DC: American Institute Peters RK, Pike MC, Garabrant D & Mack TM (1992): Diet and colon for Cancer Research. cancer in Los Angeles County, California. Cancer Causes Control 3, Yamaizumi Z, Shiomi T Kasai H, Nishimura S, Takahashi Y, Nagao M 457 – 473. & Sugimura T (1980): Detection of potent mutagens, Trp-P-1 and Phillips RL & Snowdon DA (1985): Dietary relationships with fatal Trp-P-2 in broiled fish. Cancer Lett. 9,75– 83. colorectal cancer among Seventh Day Adventists. J. Natl. Cancer Young TB & Wolf DA (1988): Case – control study of proximal and Inst. 74, 307 – 317. distal colon cancer and diet in Wisconsin. Int. J. Cancer 42, 167 – Report of an Expert Workshop (1999): Report of an Expert Workshop 175. Meat Intake and Colorectal Cancer Risk, convened in December Zaridze D, Filipchenko V, Kustov V, Serdyuk V & Duffy S (1883): Diet 1998 in Adelaide, South Australia. Eur. J. Cancer Prev. 8, 175 – 211. and colorectal cancer: results of two case – control studies in Report of Workshop (2000): Red meat and colorectal cancer. N.Z. Russia. Eur. J. Cancer 29A, 112 – 115. Med. J. 113, 195 – 196.

European Journal of Clinical Nutrition