|Home | About | Journals | Submit | Contact Us | Français|
High intake of meat, particularly red and processed meat, has been associated with an increased risk of a number of common cancers, such as breast, colorectum, and prostate in many epidemiological studies. Heterocyclic amines (HCAs) are a group of mutagenic compounds found in cooked meats, particularly well-done meats. HCAs are some of most potent mutagens detected using the Ames/salmonella tests and have been clearly shown to induce tumors in experimental animal models. Over the past 10 years, an increasing number of epidemiological studies have evaluated the association of well-done meat intake and meat carcinogen exposure with cancer risk. The results from these epidemiologic studies were evaluated and summarized in this review. The majority of these studies have shown that high intake of well-done meat and high exposure to meat carcinogens, particularly HCAs, may increase the risk of human cancer.
High intake of meat, particularly red and processed meat, has been associated with an increased risk of a number of common cancers, such as cancers of the breast, colorectum, and prostate in many epidemiological studies (Key et al, 2002; Larsson et al, 2006), including several recent large prospective cohort studies (Norat et al, 2005; Chao et al, 2007; Cho et al, 2007). Dietary fat, particularly saturated fat, has long been suspected to be responsible for the meat-cancer association. This hypothesis, however, has not been supported by many prospective cohort studies (Hunter et al, 1996; Willett, 2005). Cumulative evidence from recent animal studies and some human studies has implicated certain meat carcinogens, such as heterocyclic amines (HCAs), in the pathogenesis of human cancer (Cross et al, 2004; Knize et al, 2005; Felton et al, 2007).
Heterocyclic amines (HCAs) are a group of mutagenic compounds found in cooked meats, particularly in well-done meats (Knize et al, 2005; Felton et al, 2007). These compounds are formed during high-temperature cooking of meat from the reaction of creatine or creatinine, amino acids, and sugar (Cross et al, 2004; Knize et al, 2005; Felton et al, 2007). More HCAs are formed when meats are cooked at higher temperatures and for longer periods of time. HCA formation also depends on the type of meats and cooking method used (Cross et al, 2004; Knize et al, 2005). In most cases, 2-amino-1-methyl-6-phenylimidazo (4,5-b) pyridine (PhIP) and 2-amino-3,8-dimethylimidazo (4,5-f) quinoxaline (MeIQx) are the most mass-abundant HCAs detected in cooked meat (Cross et al, 2004; Knize et al, 2005). Other carcinogens also detectable in certain cooked meats include 2-amino-3,4,8-trimethylimidazo (4,5-f) quinoxaline (DiMeIQx) and polycyclic aromatic amines (PAHs), such as benzo(a)pyrene (BaP). HCAs have been documented as some of the most potent mutagens detected using the Ames/salmonella test (Felton et al, 2007). In animal studies, many HCAs have clearly been shown to increase the occurrence of tumors in multiple sites, including mammary glands, lung, colon, forestomach, and prostate (Adamson et al, 1995; Ito et al, 1991; Shirai et al, 1995). Recently, HCA-DNA adducts or putative HCA-DNA adducts have been detected in a variety of human tissues and organs including the breast, colorectum, and prostate, suggesting that human tissues are also vulnerable to the attack from these carcinogens (Turesky RJ et al, 2004; Malfatti MA et al, 2006; Zhu J et al, 2006; Tang et al, 2007a; Tang et al, 2007b), and thus exposure to these carcinogens may increase the risk of cancer in humans.
Over the past two decades, an increasing number of epidemiological studies have evaluated the possible role of HCA exposure in the pathogenesis of human cancers. Because HCAs are formed during high temperature cooking of meats, many epidemiological studies have used the intake level of meats prepared with high temperature cooking methods (such as pan-frying, baking, grilling/barbequing) as surrogate measures of HCA exposure. Many earlier studies included only a few items of high-temperature cooked meats in the food frequency questionnaire, which was inadequate for a comprehensive assessment of dietary intake levels of HCAs. Furthermore, since the level of HCAs in cooked meats also depends, to a large extent, on the temperature and length of time that meats are cooked, studies that failed to assess meat doneness levels may have considerable errors in their assessments of HCA exposure.
Having recognized the limitation of earlier studies, more recent epidemiologic studies have attempted to take into consideration of cooking methods and preference for doneness level in estimating HCA exposure. Some studies used color photographs for cooked meats that reflect a range of doneness levels from rare to very well-done in order to standardize the assessment of preferred meat doneness level for each meat included in dietary surveys (Sinha, 2002). The validity and reliability of these newly-developed tools to assess usual HCA exposures, however, remain to be determined. In a recent study, Kobayashi et al (2007) found that the level of HCA exposure estimated using food frequency questionnaire data was moderately correlated with that measured in hair samples. This study suggests that usual intake levels of HCAs can be measured using food frequency questionnaires in epidemiologic studies.
To evaluate the evidence for a potential role of well-done meat intake and dietary exposure to HCAs in the etiology of human cancer, we systematically searched for and reviewed epidemiological studies that have been published and listed in MEDLINE databases since 1996 when the meat-specific questionnaire and carcinogen database became available. The electronic search of MEDLINE databases was conducted using the following terms “heterocyclic amines”, “HCA”, “well-done meat”, “meat”, “cooked meat”, “cancer”, and “risk factors”. The references included in the recovered papers were also reviewed in order to identify papers that were missed in the initial MEDLINE search. We included in the review only original studies that assessed both meat intake by high temperature cooking methods and meat doneness level. Provided in Table 1 is a summary of 21 such studies, which met our criteria. Each study is listed only once in the table, along with relevant results, if any, reported in separate publications from the same study. Inevitably, a few studies may not have been identified and thus, may not be included in this review. However, these few missed studies, if any, should not affect the conclusion of this review, given the large number of studies evaluated.
Colorectal cancer has been the focus of seven studies (Augustsson et al, 1999; Kampman et al, 1999; Le Marchand et al, 2001; Nowells et al, 2002; Butler et al, 2003; Murtaugh et al, 2004; Navarro et al, 2004). With the exception of an early study conducted in Sweden (Augustsson et al, 1999), all other studies provided some evidence for a positive association of well-done meat intake and HCA exposure with the risk of colorectal cancer. In a large case-control study conducted in the U.S., Kampman (1999) reported a 30% elevated risk in association with mutagen index, a summary measure of HCA exposure. Furthermore, the association between mutagen index and colon cancer risk was found to be modified by NAT2 genotypes. The gene-diet interaction was replicated in a subsequent study, in which a 3-fold elevated risk was found among those who preferred well-done meat and had rapid CYP1A2 and NAT2 phenotypes (Le Marchand et al, 2001). Several recent studies have provided additional support for a positive association of colorectal cancer with HCA exposure (Nowell et al, 2002; Butler et al, 2003; Murtaugh et al, 2004; Navarro et al, 2004;). The meat-specific questionnaire was used in two of these studies to assess well-done meat intake and HCA exposure (Nowell et al, 2002; Butler et al, 2003). Both studies reported a moderate to strong association with well-done meat intake and certain HCAs, including MeIQx, DiMeIQx and PhIP. A dose-response relation was statistically significant for MeIQx in one study (Nowell et al, 2002), and for DiMeIQx in another study among African Americans (Butler et al, 2003). The level of DiMeIQx is typically lower than MeIQx in well-done meat, and reasons for the inconsistent findings of these two studies are unclear.
In addition to research on colorectal cancer, some epidemiological studies have evaluated well-done meat intake and HCA exposure in relation to adenomatous polyps, precursors of colorectal cancer (Probst-Hensch et al, 1997; Sinha et al, 1999; Sinha et al, 2001; Ishibe et al, 2002; Tiemersma et al, 2004; Gunter et al, 2005; Sinha et al, 2005; Wu et al, 2006; Shin et al, 2007a; Shin et al, 2007b). With the exception of two studies (Tiemersma et al, 2004; Gunter et al, 2005), all other studies reported positive associations of colorectal polyps with intake of well-done meat and/or exposure to HCAs. The potential for recall bias to have affected the results of these studies should be smaller than case-control studies of cancers, because most colorectal polyps are asymptomatic, and the dietary assessments for many of these studies were conducted before polyp diagnosis. Overall, the results from polyp studies are consistent with those of colorectal cancer studies, providing additional evidence for a significant role of HCA exposure in the etiology of colorectal tumors.
Three studies have published results on the associations between intake of high temperature cooked meat and/or HCA exposure and breast cancer risk (De Stefani et al, 1997; Zheng et al, 1998; Delfino et al, 2000; Sinha et al, 2000b). With the exception of a small hospital-based case-control study, in which patients with a benign breast disease were included as controls, two other studies reported positive associations with well-done meat intake and HCA exposure (De Stefani et al, 1997, Zheng et al, 1998; Sinha et al, 2000b). The meat-specific questionnaire was used in the case-control study conducted among postmenopausal Iowa women, in which a highly significant dose-response relationship was observed between meat doneness score and breast cancer risk (p for trend, 0.001) (Zheng et al, 1998). Women who consistently ate well-done meat had a 4.6-fold (95% CI 1.4–15.7) elevated risk of breast cancer. Using data from this study, dietary exposure levels of major HCAs were estimated (Sinha et al, 2000b). A clear dose-response relationship was observed for the risk of breast cancer with exposure to PhIP but not other HCAs evaluated in the study. Furthermore, the study also suggested that the association between well-done meat intake and breast cancer risk may be modified by genetic polymorphisms in the NAT1, NAT2, GSTM1, GSTT1, and SULT1A1 genes that encode metabolizing enzymes for HCA activation or detoxification (Zheng et al, 1999, 2001, 2002; Deitz et al, 2000).
Four published studies, including two cohort studies, have evaluated the association of intake of meat cooked at high temperature with prostate cancer risk (Cross et al, 2005; Norrish et al, 1999; Rovito et al, 2005; Koutros et al, 2008). In a large prospective cohort study conducted at 10 U.S. centers, usual meat intake by cooking methods and doneness level assessed for approximately 29,000 adult men who participated in an NCI-sponsored cancer screening trial (Cross et al, 2005). During the follow-up period, 868 incident prostate cancer cases were identified. Although high intake of red or white meat was not associated with an elevated risk of prostate cancer, a clear dose-response relation was found for intake of very well-done meat and exposure to PhIP. In another cohort study conducted in the United States, a positive association was also found to be associated with intake of well or very well-done meat (Koutros et al, 2008). However, no apparent association was found for HCA exposure derived using the dietary intake data collected in the study. The results were similar to those reported from an earlier case-control study conducted in New Zealand that reported a positive association with well-done beefsteak but not HCA exposure. The only report showing no association between HCA exposure or well-done meat intake and prostate cancer risk was from a small hospital-based, case-control study, which included some prevalent cases. As a result, potential biases in subject selection and exposure assessment may be a major concern for that study.
Pancreatic cancer is one of the deadliest malignancies in humans. To date, three epidemiological studies, including a large cohort study, have evaluated the potential role of HCA exposure in the etiology of this cancer. All three studies carried out a detailed assessment of well-done meat intake and HCA exposure using the meat-specific questionnaire. The first study was conducted in Minnesota, included 193 cases and 674 controls, and reported a positive association of pancreatic cancer risk with intake of grilled or barbequed red meat (Anderson et al, 2002). In a subsequent analysis of the data from this study, clear dose-response relationships were found between cancer risk and dietary exposure to several meat carcinogens, including PhIP, DiMeIQx, and BaP (Anderson et al, 2002). These findings were replicated in a larger study published recently, in which preference for well-done red meat and chicken and high exposure to HCAs were found to be associated with an increased risk of pancreatic cancer (Li et al, 2007). Using data from a large cohort study involving approximately 33,000 subjects, Stolzenberg-Solomon et al (2007) evaluated the association of pancreatic cancer risk with meat intake and meat carcinogen exposure. High intakes of red meat, high-temperature cooked meat, and well or very well-done meat were found to be associated with an elevated risk of pancreatic cancer among men but not among women. A 2-fold increased risk of pancreatic cancer was observed among men whose usual diet was characterized to have overall high mutagenic activity. High exposure to a number of HCAs, including PhIP was associated with an elevated risk of pancreatic cancer.
To date, four studies have evaluated the association of well-done meat intake and HCA exposure with cancers of the stomach and esophagus (Ward et al, 1997; De Stefani et al, 2001; Terry et al, 2003). No association was found in a small hospital-based, case-control study conducted in Uruguay for stomach cancer (De Stefani et al, 2001). In a large population-based case-control study conducted in Sweden, Terry et al. reported that high HCA exposure was associated with an elevated risk of squamous cell carcinoma of the esophagus but not adenocarcinoma of the esophagus or gastric cardia. On the other hand, a study in the U.S. reported that the risk of adenocarcinoma at these sites was elevated among those with high intake of gravy or a preference for well-done meat (Ward et al, 1997). Furthermore, a positive association was found in a study conducted in Uruguay for cancers of the upper aerodigestive tract, including esophageal cancer. These studies provide some evidence for a possible role of HCA exposure in the etiology of esophageal and gastric cancers, the two most common malignancies of the upper digestive tract.
Only one study has evaluated the association of well-done meat intake with lung cancer risk (Sinha et al, 1998), in which high intake of well-done red meat or fried red meat was associated with an increased risk of lung cancer. Additional analyses of data from this study showed a positive dose-response association of MeIQx exposure and lung cancer risk. These results are consistent with the carcinogenic effect of MeIQx in animal models and suggest that HCA exposure may contribute to the development of lung cancer in humans.
Of the many HCAs identified to date, PhIP is the most abundant HCA detected in human diets, followed by MeIQx and DiMeIQx. Of the 13 studies included in this review that evaluated PhIP exposure, eight have shown a positive association of PhIP exposure with cancer risk (Table 2), with a clear dose-response relationship reported in the majority of these studies. The five studies showing no association with PhIP are those conducted for cancer of the colorectum (Augustsson et al, 1999; Butler et al, 2003), prostate (Norrish et al, 1999; Koutros et al, 2008), and lung (Sinha et al, 2000a). It is of note that the HCA database used in the prostate cancer study conducted by Norrish (1999) was very limited, compared to the database that is commonly used in recently epidemiological studies (Sinha, 2002). None of the HCAs assessed in that study, including total HCAs, PhIP, MeIQx, and DiMeIQx, were found to be associated with the risk of prostate cancer. The case-control study conducted in Sweden reported that high dietary exposure to PhIP was associated with a significantly reduced risk of colon cancer (Augustsson et al 1999). This finding was unexpected and contrary to all other studies of colorectal cancer. Compared to PhIP, results for association analyses of dietary exposure to other HCAs and BaP with cancer risk have been less consistent (Table 2). The level of MeIQx and DiMeIQx in cooked meat is much lower than PhIP, and the effect under such a low exposure level may be difficult to detect in epidemiological studies. Human exposure to BaP may come from multiple sources. Among smokers and those with heavy occupational and residential BaP exposures, dietary sources of BaP exposure could be relatively minor, which could affect the evaluation of the association of dietary BaP exposure with cancer risk in epidemiological studies. It is also possible that these carcinogens may be less well quantified by dietary surveys than PhIP, which would affect the evaluation of their associations with cancer risk in epidemiological studies. Additional research is needed to improve exposure assessment and quantify the association of meat carcinogen exposure and cancer risk.
Numerous in vitro and in vivo experiments have clearly demonstrated that HCAs are potent mutagens and can induce tumors of multiple sites in animal models. Over the past two decades, particularly over the past 10 years, an increasing number of epidemiological studies have evaluated the association of well-done meat intake and HCA exposure with cancer risk in humans. Several studies have also evaluated the interaction between HCA exposure or well-done meat intake and genetic polymorphisms in carcinogen-metabolizing enzymes in the risk of cancer. However, the sample size for those studies was, in general, small and statistical power was limited. Nevertheless, most previous studies, particularly those studies with detailed assessments of meat intake by cooking methods and meat doneness levels, have reported that high intake of well-done meat and/or high exposure to certain HCAs may be associated with the risk of cancers, including cancers of the colorectum, breast, prostate, pancreas, lung, stomach, and esophagus. Results from these studies are consistent with data from in vitro and in vivo experiments and suggest that exposure to HCAs through consumption of well-done meat may increase the risk of certain cancer in humans.
This work is supported in part by NIH grant R01CA100374 and the Physicians Committee for Responsible Medicine. The authors would like to thank Brandy Sue Venuti for technical assistance in the preparation of this review.