Colorectal cancer (CRC) is the third most common cancer worldwide, accounting for 1.36 million new cases and 694 000 deaths worldwide in 2012 (Ferlay et al., 2015). Western lifestyle and dietary factors may be responsible for the high incidence of CRC in industrialized countries and for the rapidly increasing rates in some countries undergoing economic transition (Durko and Malecka-Panas, 2014; Torre et al., 2016). It has been estimated that as high as 70% (range: 50–80%) of CRC deaths could be avoided by dietary change (Willett, 1995). A study from Alberta estimated that about 12% of CRC in 2012 were attributable to red and processed meat intake (Grundy et al., 2016).
The potential role of red meat and processed meat intake in CRC risk has been the subject of scientific debate. A consensus statement issued in 2011 by the World Cancer Research Fund (WCRF) and the American Institute for Cancer Research (AICR) concluded that there was convincing evidence to support a positive association between intakes of red and processed meat and CRC (World Cancer Research Fund, 2011). However, a recently published comprehensive meta-analysis concluded that red meat intake does not appear to be an independent predictor of CRC (Alexander et al., 2015).
Most of the previous studies were carried out in North America, western Europe, and certain Asian countries (Alexander et al., 2015). Data are lacking from countries following a Mediterranean diet and countries with low intake of fresh red meat and processed meat. Israel is a Mediterranean country with an overall low Intake of fresh red meat, especially among the Jewish population, and extremely low pork consumption, fresh or processed, because of religious and cultural prohibitions against pork consumption in Jewish and Muslim traditions. Further, the overall red meat intake and the proportion of the various meat subtypes (beef, lamb, and pork) differ significantly between Jews and Arabs in Israel. This study aimed to assess the association between red meat, processed meat, and CRC risk using data from a large Mediterranean population-based study.
Patients and methods
The Molecular Epidemiology of Colorectal Cancer study is an ongoing population-based case–control study of incident CRC in northern Israel (Poynter et al., 2005; Schmit et al., 2016). Recruitment to the Molecular Epidemiology of Colorectal Cancer Study started on 31 March 1998, and all consecutive patients diagnosed with CRC residing in a geographically defined area of northern Israel at the time of diagnosis were eligible to participate in this study. CRC-free controls were selected randomly from the same source population with the use of the Clalit Health Services register. Clalit Health Services is the largest healthcare provider in Israel and covers more than half of the population in Israel. Healthcare coverage in Israel is mandatory and is provided by four groups akin to not-for-profit health maintenance organizations. Thus, all study participants (patients and controls) had a similar health insurance plan and similar access to health services, including prevention and cancer screening. Controls were matched individually on year of birth, sex, ethnicity (Jews vs. Arabs), and residence (primary care clinic). Participants provided IRB-approved written informed consent at the time of enrollment and were interviewed in-person by trained interviewers to obtain information about their dietary intake and lifestyle using a questionnaire that includes a semiquantitative food-frequency questionnaire (FFQ) modified and validated for the Israeli diet (Yizhaki et al., 2008).
The study questionnaire includes demographic information, personal and family history of cancer and relevant diseases, education and health habits including smoking, alcohol consumption, physical activity, height and weight, reproductive history, hormonal and drug exposures, and a detailed nutritional component. Total energy and fiber intake, dietary folate, calcium and vitamin D, and other dietary components were calculated from the food frequency questionnaire.
Specifically for this study, red meat was defined as unprocessed beef, lamb, or pork meat, and processed meat was defined as meat that has been transformed through salting, curing, smoking, fermentation, or other processes and might contain red meat and/or poultry such as sausages, hotdogs, salami, and pastrami (Bouvard et al., 2015).
For this analysis, we excluded 414 participants with excessive missing items in the FFQ (<150 of the 178 requested items) and participants with implausibly low or high energy intake (<600 or >4000 kcal/day). We also excluded 221 non-Jewish/non-Arab participants.
Student t-tests for continuous variable and χ2-tests for categorical variable were used to compare the baseline characteristics between cases and controls and between Jews and Arabs. Logistic regression was used to assess the association between meat intake and CRC risk. Red meat consumption was studied both as a single entity and separately as beef, lamb, or pork. Intake was tested as a continuous variable as well as quartiles of consumption and was compared with nonconsumers according to the frequency of use of the various meat subtypes. Multivariate logistic regression models were adjusted for age, sex, BMI (<25, 25–30, >30 kg/m2), first-degree family history for CRC, smoking (never, former, current), education (≤12 vs. >12 years), sports activity (yes vs. no), total energy intake, fibers, calcium, vitamin D, and folate intake, alcohol intake (any vs. none), vegetable consumption (≥3 vs. <3 serving/day), fruit consumption (≥ 3 vs. <3 serving/day), and regular aspirin intake (daily low dose for ≥1 year), and intake of each meat subtype was mutually adjusted for the intake of other meats.
For a limited period of time, CRC cases were enrolled in the study without matched controls; as a result, the number of cases is higher than the number of controls. Therefore, an unconditional model was used with matching variables (age, sex) included in the model and with stratification by ethnicity because of the major differences in consumption patterns between Jews and Arabs.
All statistical analyses were carried out using the SPSS statistics package, version 23.0 (IBM, New York, New York, USA). For all analyses, a P value of less than 0.05 for the two-tailed tests was considered to be statistically significant.
A total of 10 026 participants were included in the study. Of these, 8615 were Jews (4615 cases and 4000 controls) and 1411 were Arabs (857 cases and 554 controls). The demographic, clinical, and nutritional characteristics of cases and controls stratified by ethnic group (Jews and Arabs) are presented in Table 1.
Fresh red meat consumption, mostly of beef, was higher in Arab controls, with 3.0±1.98 servings/week, than in Jewish controls, with 1.29±1.45 servings/week (P<0.001) (Table 2). Lamb consumption was also higher in Arab controls (0.90±1.03 servings/week) than in Jews (0.06±0.25) (P<0.001), and fresh pork consumption was extremely low in both groups. Processed meat was consumed significantly more frequently by Jewish controls (1.97±2.97 servings/week) than by Arab controls (0.9±1.56 servings/week) (P<0.001) (Table 2).
Red and processed meat intake and colorectal cancer risk in Jews
Using meat intake as a continuous variable, the adjusted odds ratio (OR) for an increase in one serving per week of all red meats was 1.05 [95% confidence interval (CI): 1.01–1.08]. Although the CRC risk was not elevated for beef consumption, OR=1.01 (0.97–1.05), it was significantly increased for lamb, OR=1.46 (1.20–1.79), and pork, OR=1.21 (1.07–1.35), and was very mild for one additional serving of processed meat (usually pork free), OR=1.02 (1.0–1.03) (Table 3).
Among Jews, 23.2% reported no intake of red meats and 32% reported no intake of processed meats at all, allowing us to compare nonconsumers with consumers. The adjusted CRC risk for any red meat was 1.03 (95% CI: 0.93–1.16), and differed between beef intake OR=0.96 (0.86–1.07), lamb intake OR=1.28 (1.10–1.50), pork intake, OR=1.44 (1.24–1.67), and processed meat, OR=1.22 (1.10–1.35) (Table 4). Assessment of the association by comparing highest to lowest quartiles of meat intake yielded an OR=1.13 (95% CI: 0.98–1.31) for all red meat intake, OR=0.95 (0.82–1.11) for beef intake, and OR=1.21 (1.06–1.38) for processed meat intake (Supplementary Table S1, Supplemental digital content 1, http://links.lww.com/EJCP/A205). The variance of lamb and pork intake was low among Jews; therefore, classification into quartiles was not applicable (Table 2 and Supplementary Table S1, Supplemental digital content 1, http://links.lww.com/EJCP/A205).
Red and processed meat intake and colorectal cancer risk in Arabs
The adjusted risks for CRC of an increase of one serving/week of overall red meat consumption, beef consumption, lamb consumption, or pork consumption were all not associated with a significant increase in the Arab population in our study; OR=0.94 (95% CI: 0.88–1.01), OR=0.88 (0.81–0.96), OR=1.02 (0.91–1.16), OR=1.15 (0.89–1.48), respectively. The risk associated with processed meat consumption was also not increased, OR=0.99 (0.91–1.07) (Table 3).
Compared with no intake, the adjusted OR was 1.34 (95% CI: 0.69–2.62) for overall red meat intake, 0.94 (0.61–1.45) for beef, 1.01 (0.75–1.37) for lamb, 1.07 (0.73–1.56) for pork, and 1.04 (0.82–1.33) for processed meat (Table 4).
Assessment of the association with quartiles of meat intake was applicable only for the total consumption of red meats and for beef intake becuase of the low variance of the other red meat subtype intake among Arabs (Table 2 and Supplementary Table S1, Supplemental digital content 1, http://links.lww.com/EJCP/A205). Compared with the lowest quartile of intake, the odds of CRC for those in the highest quartile were 0.69 (95% CI: 0.47–1.01) for overall red meats intake and 0.65 (0.45–0.93) for beef intake (Supplementary Table S1, Supplemental digital content 1, http://links.lww.com/EJCP/A205).
Anatomical sub-site data of CRC were available for 4392 (95.2%) Jewish cases and 805 (93.9%) CRC cases among Arabs. The risk estimates of red meats and processed meats consumption were similar for right colon, left colon, and rectal tumors among Jews and among Arabs. Lamb and pork consumption seemed to have a stronger association specifically with the right colon, but none of the associations reached statistical significance (Table 5 and Supplementary Table S2, Supplemental digital content 1, http://links.lww.com/EJCP/A205).
Our study of the association between red and processed meat and CRC risk, carried out in a country with low overall meat consumption and minimal consumption of pork products, shows only weak associations with the risk of CRC. Beef, pork, lamb, and processed meat consumption were examined independently, and there was no evidence of a meaningful association between beef consumption and the risk of CRC, whereas a modest risk was associated with consumption of lamb, pork, and processed meats.
Previous studies have shown conflicting results with respect to the association between red meat and processed meat intake and CRC risk ranging from no effect, positive association, and negative association (Chao et al., 2005; Norat et al., 2005; Cross et al., 2010; Williams et al., 2010; Egeberg et al., 2013; Bernstein et al., 2015). Studies that examined red unprocessed and processed meats separately detected stronger and more consistent associations between processed meat and CRC (Yizhaki et al., 2008; Bernstein et al., 2015). A consensus statement issued in 2007 by the WCRF and AICR concluded that red meat and processed meat are ‘convincing’ risk factors for CRC (World Cancer Research Fund, 2011). The WCRF and AICR reconfirmed their recommendations in 2011 on the basis of a meta-analysis that showed a significant association between red meat intake and CRC; relative risk (RR)=1.10 (95% CI: 1.00–1.21) and between processed meat intake and CRC; RR=1.17 (95% CI: 1.09–1.5) for the highest intake compared with the lowest intake (Chan et al., 2011). However, a comprehensive meta-analysis published in 2015 showed that red meat intake had a weak and non-significant association with CRC; RR=1.05 (95% CI: 0.98–1.12) for the highest intake compared with the lowest intake (Alexander et al., 2015). The authors concluded that red meat intake does not appear to be an independent risk factor for CRC (Alexander et al., 2015). More recently, in October 2015, the International Agency for Research on Cancer classified processed meat intake as a group I carcinogen and red meat intake as a group 2A (probable) carcinogen (Yizhaki et al., 2008).
The EPIC study, similar to our finding, identified lamb and pork intake, but not beef intake, to be associated significantly with CRC risk (Norat et al., 2005). A Danish study, while finding no association between overall red meat intake and CRC risk, reported pork intake to be associated with an increased risk of rectal cancer, lamb intake with an increased risk of colon cancer, and beef intake with a decreased risk of rectal cancer (Egeberg et al., 2013).
Cancers arising in the proximal colon, distal colon, and rectum arise from different embryonic tissue, and each has distinct clinical behavior and risk factor profiles, likely related to important differences in molecular cancer pathways (Iacopetta 2002; Williams et al., 2010). Red meat and processed meat may have a different impact on the development of the various CRC sites (Chao et al., 2005; Larsson and Wolk, 2006; Cross et al., 2010; Bernstein et al., 2015). We carefully examined the risks of red meat consumption in our large, population-based study and found no significant differences by tumor location.
There may be subtle differences in risk estimates of red meat consumption in Jews and Arabs that are difficult to discern in our large population-based study. For example, among Arabs, we observed a significant, modestly reduced odds ratio of 0.88 (95% CI: 0.81–0.96) for beef consumption in a fully adjusted model, whereas there was no association in Jews OR=1.01 (95% CI: 0.97–1.05). Whether these differences are meaningful is open to discussion, but what is clear is that there is no evidence of an increased risk of CRC from beef consumption in any population in Israel.
Various compounds with carcinogenic potential could explain the associations between meat and colon cancer, including heme iron, heterocyclic amines, polycyclic aromatic hydrocarbons, and nitrites and nitrates (Cross et al., 2010; Miller et al., 2013).
Our study has the strengths of a large size and detailed information in a population-based setup. This allowed us to analyze the differences by type of meat intake while controlling for many relevant potential confounders. However, because of the observational nature of the study, we cannot exclude the possibility of residual confounding, which, given the weak associations identified, may have a large impact on the effect estimate. Observational studies, by their nature, are susceptible to recall bias as cases may have reported past meat consumption differently than controls if meat was perceived to be a risk factor for CRC. Measurement errors associated with FFQs could also lead to nondifferential misclassification of respondent into dietary exposure categories, thereby attenuating risk estimates.
In summary, we identified only weak associations between overall red meat and processed meat intake and CRC risk in this low-consumption population, with evidence supporting the hypothesis that lamb and pork consumption, even at low levels, may be more relevant as risk factors for CRC than beef intake.
This work was supported in part by NIH R01 CA81488.
Conflicts of interest
There are no conflicts of interest.
Alexander DD, Weed DL, Miller PE, Mohamed MA (2015). Red meat
and colorectal cancer
: a quantitative update on the state of the epidemiologic science. J Am Coll Nutr 34:521–543.
Bernstein AM, Song M, Zhang X, Pan A, Wang M, Fuchs CS, et al (2015). Processed and unprocessed red meat
and risk of colorectal cancer
: analysis by tumor location and modification by time. PLoS One 10:e0135959.
Bouvard V, Loomis D, Guyton KZ, Grosse Y, Ghissassi FE, Benbrahim-Tallaa L, et al, International Agency for Research on Cancer Monograph Working Group (2015). Carcinogenicity of consumption of red and processed meat
. Lancet Oncol 16:1599–1600.
Chan DS, Lau R, Aune D, Vieira R, Greenwood DC, Kampman E, et al (2011). Red and processed meat
and colorectal cancer
incidence: meta-analysis of prospective studies. PLoS One 6:e20456.
Chao A, Thun MJ, Connell CJ, McCullough ML, Jacobs EJ, Flanders WD, et al (2005). Meat consumption and risk of colorectal cancer
. JAMA 293:172–182.
Cross AJ, Ferrucci LM, Risch A, Graubard BI, Ward MH, Park Y, et al (2010). A large prospective study of meat consumption and colorectal cancer
risk: an investigation of potential mechanisms underlying this association. Cancer Res 70:2406–2414.
Durko L, Malecka-Panas E (2014). Lifestyle modifications and colorectal cancer
. Curr Colorectal Cancer
Egeberg R, Olsen A, Christensen J, Halkjær J, Jakobsen MU, Overvad K, et al (2013). Associations between red meat
and risks for colon and rectal cancer depend on the type of red meat
consumed. J Nutr 143:464–472.
Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, et al (2015). Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 136:E359–E386.
Grundy A, Poirier AE, Khandwala F, McFadden A, Friedenreich CM, Brenner DR (2016). Cancer incidence attributable to red and processed meat
consumption in Alberta in 2012. CMAJ Open 4:E768–E775.
Iacopetta B (2002). Are there two sides to colorectal cancer
? Int J Cancer 101:403–408.
Larsson SC, Wolk A (2006). Meat consumption and risk of colorectal cancer
: a meta-analysis of prospective studies. Int J Cancer 119:2657–2664.
Miller PE, Lazarus P, Lesko SM, Cross AJ, Sinha R, Laio J, et al (2013). Meat-related compounds and colorectal cancer
risk by anatomical subsite. Nutr Cancer 65:202–226.
Norat T, Bingham S, Ferrari P, Slimani N, Jenab M, Mazuir M, et al (2005). Meat, fish, and colorectal cancer
risk: the European Prospective Investigation into cancer and nutrition. J Natl Cancer Inst 97:906–916.
Poynter JN, Gruber SB, Higgins PD, Almog R, Bonner JD, Rennert HS, et al (2005). Statins and the risk of colorectal cancer
. N Engl J Med 352:2184–2192.
Schmit SL, Rennert HS, Rennert G, Gruber SB (2016). Coffee consumption and the risk of colorectal cancer
. Cancer Epidemiol Biomarkers Prev 25:634–639.
Torre LA, Siegel RL, Ward EM, Jemal A (2016). Global cancer incidence and mortality rates and trends – an update. Cancer Epidemiol Biomarkers Prev 25:16–27.
Willett WC (1995). Diet, nutrition, and avoidable cancer. Environ Health Perspect 103 (Suppl 8):165–170.
Williams CD, Satia JA, Adair LS, Stevens J, Galanko J, Keku TO, et al (2010). Associations of red meat
, fat, and protein intake with distal colorectal cancer
risk. Nutr Cancer 62:701–709.
World Cancer Research Fund (2011). American Institute for Cancer Research: Continuous Update Project Summary Food, nutrition, physical activity, and the prevention of cancer: a global perspective. Washington, DC: World Cancer Research Fund/American Institute for Cancer Research.
Yizhaki D, Rennert H, Rosen G, Rennert G (2008). Validity and reproducibility of a semi-quantitative food frequency questionnaire adapted to an Israeli population. The Open Nutrition Journal 2:9–14.