Vol. 10, 439–446, May 2001 Cancer Epidemiology, Biomarkers & Prevention 439 Systematic Review of the Prospective Cohort Studies on Meat Consumption and Colorectal Cancer Risk: A Meta-Analytical Approach1 Manjinder S. Sandhu,2 Ian R. White, and confounded or modified by other factors cannot be Klim McPherson excluded. Department of Public Health and Primary Care, Institute of Public Health, University of Cambridge, Strangeways Research Laboratory, Cambridge, CB1 8RN [M. S. S.]; Medical Research Council Biostatistics Unit, Institute of Introduction Public Health, University of Cambridge, Cambridge, CB2 2SR [I. R. W.]; and The relation between meat consumption and colorectal cancer Cancer and Public Health Unit, London School of Hygiene and Tropical Medicine, London, WC1E 7HT [K. M.], United Kingdom risk remains controversial (1–10). Subsequent to the report of the National Academy of Sciences, “Diet and Health” (11), which implicated red meat as a causative factor in the etiology Abstract of colorectal cancer, two subsequent reports have reviewed the epidemiological evidence on meat and colorectal cancer risk (4, The relation between meat consumption and colorectal 3 cancer risk remains controversial. In this report, we 7). The report of the WCRF concluded: “The evidence shows quantitatively reviewed the prospective observational that red meat probably increases risk and processed meat pos- studies that have analyzed the relation between meat sibly increases risk of colorectal cancer” (7). The report from consumption and colorectal cancer. We conducted COMA judged that “there is moderately consistent evidence electronic searches of MEDLINE, EMBASE, and from cohort studies of a positive association between the con- CANCERLIT databases through to the end of June 1999 sumption of red or processed meat and risk of colorectal can- and manual searches of references from retrieved cer” (4). A WHO consensus statement reached a similar con- articles. We used both fixed and random-effects meta- clusion, stating that “consumption of red meat is probably analytical techniques to estimate the overall association associated with increased colorectal cancer risk,” but also stated and to investigate possible sources of heterogeneity that epidemiological studies on meat and colorectal cancer risk among studies. Thirteen studies were eligible for inclusion are not consistent (2). in the meta-analysis. Pooled results indicate that a daily Both the COMA and WCRF reports made dietary recom- increase of 100 g of all meat or red meat is associated mendations based on their qualitative assessments of the epi- with a significant 12–17% increased risk of colorectal demiological literature. The WCRF report recommended lim- cancer. The marginally significant between-study iting “intake of red meat to less than 80 g daily.” The COMA heterogeneity for all meat and red meat was explained by report, targeted at the population of the United Kingdom, ad- vised that consumption of red and processed meat for those a number of study-level covariates. A significant 49% ϳ increased risk was found for a daily increase of 25 g of consuming population average levels ( 90 g/day for the United Kingdom population) should not rise. It also recommended that processed meat. The individual study estimates for Ͼ processed meat showed no detectable heterogeneity. On people who are consuming high levels ( 140 g cooked weight/ the basis of this quantitative review of prospective day) should consider a reduction. studies, the overall association between meat consumption In contrast, other researchers have noted that “it remains and risk of colorectal cancer appears to be positive, with uncertain whether meat is a risk factor for cancer” (12) and that marginal heterogeneity between studies. The finding for the current prospective evidence on meat and colorectal cancer processed meat and data from experimental studies risk “is now clearly negative for this association” (3). It has also suggests that it may also be an important predictor of been suggested that any association between high meat con- colorectal cancer risk. However, because only a few of the sumption and colorectal cancer may be as a result of deficien- studies reviewed here attempted to examine the cies in other protective dietary factors, such as vegetables and independent effect of meat intake on colorectal cancer fruits (1, 13). References to studies in many of these reports, risk, the possibility that the overall association may be reviews, and discussion articles are incomplete and limited; therefore, we aimed to review prospective cohort studies that have investigated the relation between meat and colorectal cancer risk. In addition, given that quantitative recommenda- Received 9/1/00; revised 2/7/01; accepted 2/22/01. tions—based on qualitative assessments of the literature—have The costs of publication of this article were defrayed in part by the payment of been made, we sought to use quantitative review methods to page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 M. S. S. is supported by the United Kingdom Medical Research Council. 2 To whom requests for reprints should be addressed, at Department of Public Health and Primary Care, Institute of Public Health, University of Cambridge, Strangeways Research Laboratory, Wort’s Causeway, Cambridge, CB1 8RN, 3 The abbreviations used are: WCRF, World Cancer Research Fund; COMA, United Kingdom. Phone: 01223-740168; Fax: 01223-740177; E-mail: Chief Medical Officer’s Committee on Medical Aspects of Food; OR, odds ratio; [email protected]. HCA, heterocyclic amine. Downloaded from cebp.aacrjournals.org on September 28, 2021. © 2001 American Association for Cancer Research. 440 Meat Consumption and Colorectal Cancer Risk Table 1 List of studies used in the review and meta-analysis and their selected characteristics Author, year Years of follow-up; % Start of No. of cases; Dietary assessment;a Country Age at entry; sex Adjustments published completed follow-up follow-up No. in cohort Quantilesb Bjelke et al., 1980 Norway 45–74; M 5; not stated 1968 41;c 12,166 50-item FFQ; Q3 Age (29) Gaard et al., 1996 Norway 20–54;M&F Mean 11.4; 83 1977 143;c 55,535 80-item FFQ;d Q4 Age; attained age (21) Giovannucci et al., USA 40–75; M 6; 95 1986 205;c 47,949 131-item FFQ;d Q5 Age; energy intake 1994 (22) Goldbohm et al., Netherlands 55–69;M&F 3.3; 95 1986 293;c 3,123 150-item FFQ;d Q4 Age; total calories 1994 (23) &Q5 Hirayama et al., Japan Ͼ40;M&F 17;80 1965 725; 265,113 9-item FFQ (7 food Age 1990 (28) groups & 2 beverages); Q4 Hsing et al., 1998 USA Ն35; M 20; 77 1966 145; 13,606 35-item FFQ; Q5 Total calories; age; (27) smoking; alcohol Kato et al., 1997 USA 34–65; F Mean 7.1; (97 in New 1985– 100;c 14,727 70-item FFQ; Q4 Total calories; age; place (20) York State recruits) 1991 of enrollment; education Knekt et al., 1999 Finland 15–99;M&F 24;notstated 1967 73;c 9,990 1 year dietary Age; sex; municipality; (17) history interview; smoking; energy Q4 intake Phillips et al., USA Ͼ30;M&F 21;notstated 1960 172; 25,439 21-item FFQ; Q3 Age; sex 1985 (26) Sellers et al., 1998 USA 55–69; F 9; 76 1986 241;c 26,937 127-item FFQ;d Q3 Age; energy intake (18) Singh et al., 1998 USA 25–104;M&F 6;97 1977 179;c 32,051 55-item FFQ;d Q3 Age; body mass index; (19) sex; smoking; physical activity; family history; alcohol; aspirin use Thun et al., 1992 USA 30–110;M&F 6;98 1982 1,150; 5,746 42-item FFQ (32 None (24)e food items & 10 beverages); Q5 Willett et al., 1990 USA 34–59; F 6; 96 1980 150;c 88,751 61-item FFQ;d Q5 Age; energy intake (25) a Food Frequency Questionnaire. b Quantiles. Q3 ϭ tertiles; Q4 ϭ quartiles; Q5 ϭ quintiles. c Incident cases. d Undertook validation of dietary assessment. e Nested case control. both confirm the current recommendations and to investigate of June 1999. A search strategy that included both truncated possible sources of heterogeneity among studies. free-text and exploded MeSH terms was used. MeSH headings included “colorectal,” “colon,” “bowel,” “rectum,” “diet,” Materials and Methods “meat,” “cancer,” “neoplasm,” “prospective,” “follow-up,” or “cohort” and their variants. All references that matched the Inclusion Criteria. We sought to include both published and unpublished prospective cohort studies that contained risk es- inclusion criteria were retrieved and the references of those timates of colorectal cancer associated with meat consumption. articles were checked for other relevant publications. Refer- A broad definition of “meat” was used, which was taken to ences contained in recent reviews of the literature were also include red meat, lamb, beef, pork, and processed meats, such consulted (4, 7, 8). Finally, principal investigators responsible as sausages, meat burgers, ham, bacon and other meat products, for the collated studies and authors of recent reviews were but which, where possible, excluded white meat, such as poul- contacted for any unpublished or missed research. try. Eligible outcomes were colon or colorectal cancer inci- Data Extraction and Classification. Rate ratios, 95% confi- dence or mortality. dence intervals, and various study characteristics were ex- Exclusion Criteria. We excluded case-control and ecological tracted from the original reports and included in the meta- studies. Case-control studies, where diet is assessed after the analysis (Table 1). The extracted data for each study were then onset of disease, may be subject to information (recall) and sent to the original investigator for review and to request any selection bias, and inaccurate or biased measurements of die- additional data that were required for the meta-analysis.