Genetic Polymorphism of Enzymes Involved in Xenobiotic Metabolism and the Risk of Colorectal Cancer

Environmental factors such as smoking cigarette, diets and alcohol may interact with genetic factors, which put one individual at a greater or lesser risk of a particular cancer than another. Advances in molecular biology have allowed many allelic variants of several drug metabolizing enzymes so that individuals with the susceptible genotypes can be determined easily. Many pieces of research have focused on the relationship between the distribution of polymorphic variants of different forms of the metabolic enzymes and colorectal cancer susceptibility because of importance roles of the metabolic enzymes in the activation of many procarcinogens or chemicals. In this respect five groups of the metabolic enzymes, cytochrome P450 (CYP) 1 Al /CYPIA2, glutathione S-transferases (GSTs), N-acetyltransferases (NATs), aldehyde dehydrogenase 2 (ALDH2) and methylenetetrahydrofolate reductase (MTHFR), have been discussed here. A positive association between development of colorectal cancer and the mutant homozygous genotype in Mspl polymorphism of CYPIA1 gene has been reported in Japanese in Hawaii. The relation between genetic polymorphisms in GSTs and cancer risk has also taken an interest. At least nine studies have demonstrated the relation between the GST polymorphisms and colorectal cancer. Two of these studies suggested an increased risk of approximately 2-fold among those with the GSTM1 null genotype, while others found no risk increase. None of these studies examined the combined effect of CYPIA1 and GST polymorphisms. Either NAT2 or CYPIA2 alone have been slightly associated with colorectal cancer. When CYPIA2 and NAT2 phenotype were combined, a significant increased risk (odds ratio of 2.8) was seen among well done meat consumers with the rapid-rapid phenotype. Two published studies have found that the risk of colorectal cancer can be enhanced (2-3 fold) in alcohol drinkers with heterozygous genotype of ALDH2 in two Japanese populations recently. Findings from three published studies suggested that the mutant genotype of MTHFR inversely slightly associated with colorectal cancer. Although some of genetic polymorphisms discussed here have not shown statistically significant increase/decrease in risk, individuals with differing genotypes may have different susceptibilities to colorectal cancer, based on environmental factors. Further studies are needed to identify risk groups more specific and to determine factors of importance in colorectal cancer development. J Epidemiol, 2000 ; 10 : 349360


Genetic
Polymorphism of Enzymes Involved in Xenobiotic Metabolism and the Risk of Colorectal Cancer

Chikako Kiyohara
Environmental factors such as smoking cigarette, diets and alcohol may interact with genetic factors, which put one individual at a greater or lesser risk of a particular cancer than another.Advances in molecular biology have allowed many allelic variants of several drug metabolizing enzymes so that individuals with the susceptible genotypes can be determined easily.Many pieces of research have focused on the relationship between the distribution of polymorphic variants of different forms of the metabolic enzymes and colorectal cancer susceptibility because of importance roles of the metabolic enzymes in the activation of many procarcinogens or chemicals.
A positive association between development of colorectal cancer and the mutant homozygous genotype in Mspl polymorphism of CYPIA1 gene has been reported in Japanese in Hawaii.The relation between genetic polymorphisms in GSTs and cancer risk has also taken an interest.At least nine studies have demonstrated the relation between the GST polymorphisms and colorectal cancer.Two of these studies suggested an increased risk of approximately 2-fold among those with the GSTM1 null genotype, while others found no risk increase.
None of these studies examined the combined effect of CYPIA1 and GST polymorphisms.
Either NAT2 or CYPIA2 alone have been slightly associated with colorectal cancer.When CYPIA2 and NAT2 phenotype were combined, a significant increased risk (odds ratio of 2.8) was seen among well done meat consumers with the rapid-rapid phenotype.
Two published studies have found that the risk of colorectal cancer can be enhanced (2-3 fold) in alcohol drinkers with heterozygous genotype of ALDH2 in two Japanese populations recently.Findings from three published studies suggested that the mutant genotype of MTHFR inversely slightly associated with colorectal cancer.
Although some of genetic polymorphisms discussed here have not shown statistically significant increase/decrease in risk, individuals with differing genotypes may have different susceptibilities to colorectal cancer, based on environmental factors.Further studies are needed to identify risk groups more specific and to determine factors of importance in colorectal cancer development.

INTRODUCTION
Colorectal cancer is the fourth and the second leading cause of cancer death in male and female Japanese, respectively 1).Genes that interact with environmental or dietary factors, including cigarette smoking, may play a key role in colorectal carcinogenesis.The attributable risk for metabolic polymorphisms for carcinogens or substances derived from diet with a modest risk increase (e.g., the odds ratio (OR) for the null genotype of glutathione S-transferase (GST) M1 is 2.0 2) may be higher than that of rare mutations in DNA mismatch repair genes such as hMSHI and hMSH2 with much higher risk increase (mutation in those gene are frequently found in the germline cells of hereditary nonpolyposis colorectal cancer familes a 3.4) .In addition studying gene-environment interactions in relation to risk of cancer may be valuable because positive findings would clearly implicate the substrates with which the gene interacts as disease-causing exposures, clarify cancer etiology and point to preventive dietary and other environmental modifications.
Cigarette smoking, consumption of diets high in red meat and alcohol use probably increase the risk of colorectal cancer 5).Also cooking meat at high temperatures possibly increase the cancer risk 5).Compounds in cigarette smoke or diets are modified by drug metabolizing enzymes and some of the metabolites may be the cause of colorectal cancer.
Polycyclic aromatic hydrocarbons (PAHs) and other tobacco-related carcinogens are activated by phase I enzyme cytochrome P450 (CYP) 1A1 and detoxified by phase II enzyme GSTs.The metabolic balance between phase I and phase H enzymes may be of importance to determine genetic susceptibility to colorectal carcinogenesis as well as lung carcinogenesis.
IARC judged some kinds of the heterocyclic amines (HCAs) to be possible or probable carcinogens 6).Humans are exposed to HCAs when they consume fish or meat cooked at very high temperature.HCAs are metabolically activated by CYP1A2 7) and further activated by N-acetyltransferase (NAT2) 7).
Mitochondrial aldehyde dehydrogenase (ALDH2) eliminates most of the acetaldehyde produced during alcohol metabolism, with two alleles ALDH2 *1 and ALDH2 *2 8).A mutated ALDH2 homozygotes (ALDH2 *2/ALDH2 *2) contributes to decreased activity of the enzyme while the wild-type ALDH2*1/ALDH2*1 genotype produced tolerance to alcohol use 9).It is likely that individuals with the wild-type genotype of ALDH2 gene have decreased risk of alcohol-induced colorectal cancer.
In epidemiological studies, high levels of dietary folate have been inversely associated with colorectal cancer 10,11) MTHFR is a key regulatory enzyme in the metabolism of folate.The homozygous mutant genotype of MTHFR gene, which shows decreased enzyme activity, was recently shown to be related to decreased risk of colorectal cancer in three studies [12][13][14].
In this paper, we discuss the relationship between genetic polymorphism of enzymes involved in xenobiotic metabolism and colorectal cancer, with special emphasis on the most investigated genes/enzymes of CYPJAJ, CYP1A2, GSTs, NATs ALDH2 and MTHFR.

CYPIAI/GST POLYMORPHISMS AND CIGARETTE SMOKING
Although cigarette smoking has been clearly implicated as a cause of a number of cancers, evidence for an association between cigarette smoking and colorectal cancer appears inconsistent.More recent studies have tended to find cigarette smokers to be at higher risk for colorectal cancer, however 15-1n.In addition, cigarette smoking has been consistently associated with colorectal adenomas, the precursor of colorectal cancer 18.19)The influence of smoking on risk for cancers would most likely be caused by PAHs.These carcinogens can readily reach the mucosa of the colon through the circulatory system 21).There is also evidence of constituents of cigarette smoke reaching other organs through the circulation.Aromatic-DNA adducts have been identified in human endometorium 21) and breast 22. 23).If PAHs in cigarette smoke are involved in colorectal carcinogenesis, genetic polymorphisms in the carcinogen metabolizing enzymes a re likely to influence the risk of colorectal cancer.
Two phases exist in the metabolism of most chemical carcinogens.In the phase I reaction, the carcinogens exert their effect only after being metabolically activated to intermediate metabolites, which are capable of binding to DNA and causing mutations.The most ubiquitous phase I catalysts are the CYPs.CYPIAI, one form of CYPs, is active toward PAHs such as benzo(a)pyrene (BP).PAH metabolism has been described to take place in the colon 24).CYPIAI gene is predominantly expressed in many extrahepatic tissues.To date, four polymorphic sites in CYPIAI gene have been reported and among those Mspl polymorphism may contribute individual cancer susceptibility as genetic modifiers of cancer risk.MspI polymorphism (T to C transition) in the 3'-flanking region of CYPIAI gene 25), which can be classified into 3 genotypes, predominant homozygous alleles (genotype A), heterozygote (genotype B) and homozygous rare alleles (genotype C).Increased inducibility of CYPIAI enzymatic activity was recently observed among subjects with genotype C 26).The frequency of genotype C allele is 0.33 in Japanese population 26.27), and less than 0.10 in Caucasian populations [28][29][30][31][32][33].Thus, the frequencies of genotype C in Caucasians and Japanese population are less than 1 % and 10 %, respectively.A small case-control study suggested that genotype C was positively associated with in situ colorectal cancer in Japanese in Hawaii (P = 0.008), while no such association was found in Caucasians probably due to the lack of power 28).The OR for genotype C compared with genotypes A and B combined was 7.9 (95% confidence interval (CI), 1.4-44.4) in Japanese in Hawaii.
Following phase I reaction, phase II enzymes such as GSTs are responsible for detoxification of activated forms PAH epoxides.GSTs are constitutively found in a wide variety of tissues, with different characteristic patterns of GST isozymes.
GSTs also form a superfamily of genes consisting of four distinct families, named alpha, mu, pi and theta.Certain genes within the GSTM, GSTT and GSTP subfamilies (GSTMI, GS7T1 and GSTPI genes, respectively) are polymorphic in humans.The phenotypic absence of GSTMI and GS7T1 activity is due to homozygosity for deletion of these genes, termed the null genotype 34.35).Two genetic polymorphisms at the GSTP1 locus results from a single base pair substitution in exon 5 (1lelO5Val) and exon 6 (Alal14Val) 36).In vitro cDNA expression study suggests that substitution of these amino acid reduces enzyme activity 37).
The homozygous deletion of GSTMI gene has been shown to occur in approximately 50% of the populations of various ethnic origins 2.38-44), while the homozygous deletion of GSTT1 gene has distributed between 10 and 64 % of various ethnic groups 36.[38][39][40]42).The frequency of the GS7T1 null genotype in Caucasian populations is 30% or less but that in Oriental populations may be similar to the frequency of the GSTMI null genotype.The GSTP1 polymorphism in exon 6 is less common than that in exon 5. Individuals with homozygous for the 105 valine allele (the mutant allele) are most common among African-Americans (18%) and least common among Japanese (2%) with European-Americans (11%) intermediate between these groups 36).
Most interest in the possible consequences of GST polymorphisms have focused on the polymorphism at the GSTMI gene loci.In an earlier study, the GSTMI null genotype was found to yield a 1.8-fold increased risk for colon cancer 2).In the second study, an excess of the nulled individuals were seen in colorectal cancer but this was not a significant.When the patients were divided into cancers occurring in the proximal or distal colon, the null genotype became a significant risk factor among the subgroup with distal colorectal tumors (OR, 2.03; 95% CI, 1.06-3.90)38).Therefore, two 2,38) of eight 2. [38][39][40][41][42][43][44] showed approximately 2-fold increased risk for colorectal cancer (Table 1).Five published studies 38-40,42.44)have examined the relationship between GSTTI null genotype and colorectal cancer risk.One 39) of these studies showed the GSTTI null genotype was related to significantly increased risk colorectal cancer with an OR of 1.9 (95% CI, 1.3-2.8).Two studies 36,41) have examined the frequency of GSTPI polymorphism in exon 5; neither study reported any association between GSTP1 polymorphism and susceptibility to colorectal carcinogenesis.Genetically determined susceptibility to smoking-related cancers may depend on the metabolic balance between phase I and phase II enzymes.BP, one of the most typical PAHs, is metabolized to ultimate carcinogen BP 7,8-diol-9,10-oxide by phase I enzyme CYPIAI.Subsequently, the ultimate carcinogen can be metabolized further to innocuous water soluble metabolites through conjugation with glutathione by phase II enzymes GSTs.It is likely that individuals with more reactive phase I enzymes and less efficient phase II enzymes might be at higher risk of cancer than individuals with the opposite combination.Smoking-related risk of colorectal cancer may be more accurately estimated when genetic susceptibility is allowed for as regards both CYPIAI and GST genotypes.To date, none of these studies examined the combined effect of CYPIAI and GST polymorphisms.

2.87) 57
).An important role of the CYPIA2 phenotype in colon carcinogenesis is recently suggested by one study 57) while a number of studies have verified whether individuals with the NAT2 rapid acetylator phenotype or genotype are predisposed to the development of colorectal cancer.So far nineteen stud-ies have investigated the relation between rapid NAT2 genotype/phenotype and colorectal cancer risk (Tables 2-1 and 2-2).Four 74-77 of seven 57.74-79), phenotyping studies showed apparently 1.8-to 2.5-fold increased risk while one 68) of twelve 41,43.44,61-69)genotyping studies found a significant increased risk of 1.6fold (Tables 2-1 and 2-2).Moreover inheritance of rapid NAT] genotype has conferred a 1.9-fold (95% CI, 1.2-3.2) increased risk for colorectal cancer in an English population 61).None of those phenotyping studies have measured NATI enzymatic activity.Although rapid NAT2 genotype has not been a significant risk factor in this population, among those with rapid NAT2 genotype, rapid NAT] genotype has shown a 2.8-fold (95%CI, 1.4-5.7)increased risk relative to slow NAT] genotype 61).NAT genes may confer differential susceptibility to the effect of red meat intake on colorectal cancer risk.Among men aged 60 or older who were rapid genotypes for both NATI and NAT2, consumption of more than one serving of red meat per day was associated with a relative risk of 5.82 (95% CI, 1.11-30.6)compared with the referent consumption 69).However, they demonstrated that NAT] or NAT2 or combined genotypes itself failed to show any significant risk increase.
There is also evidence that rapid metabolizers for both CYPIA2 and NAT2 have increased susceptibility to colon cancers (OR, 2.79; 95% CI, 1.69-4.47)57).Among well done red  meat consumers, the more pronounced OR of 2.79 (95% CI, 1.69-4.47)for the combined rapid-rapid phenotype strongly indicated that the proposed metabolic pathway and the level of exposure to HCAs could play a significant role in the risk for colorectal cancer.

ALDH2 POLYMORPHISM AND ALCOHOL USE
The evidence that alcohol increases the risk of upper aerodigestive tract cancers such as esophageal cancer is convincing 5).High alcohol use probably increases the risk of colorectal cancer and the effect generally seems to be related to total ethanol intake, irrespective of the type of alcohol-bevearge 5).
Taken orally, ethanol is readily absorbed from the gastrointestinal tract and diffuses rapidly and uniformly throughout the body water.The direct carcinogenic effects of alcohol on several sites of cancer and alcohol-related physical alternations have been attributed to acetaldehyde rather than to alcohol itself.The major pathway for the disposition of ethanol is its oxidation in the liver to acetaldehyde and hydrogen.Other tissues such as kidneys, stomach and intestine oxidize ethanol to a small extent 80, 81).Although ethanol is also oxidized in the large intestine to acetaldehyde via colonic bacteria 82), the bacterial capacity of oxidation is low 83).Two principal enzymes, the cytosolic alcohol dehydrogenase (ADH) and the major non-ADH pathway CYP2EJ enzyme, are responsible for the oxidation of ethanol.The further oxidation of acetaldehyde to acetate,which is converted to carbon dioxide via the citric acid cycle, is catalyzed by the ALDH2.Most of the acetaldehyde generated during alcohol metabolism is eliminated by ALDH2 9).
Considerable interindividual variations as well as ethnic differences in alcohol metabolism rate have been reported 84).Protein structural studies have revealed that a point mutation in exon 12 is responsible for the inactivation of ALDH2 isozyme 85).The presence of a single ALDH2 *2 allele results in the defi-cient ALDH2 enzymatic activity 86).Those with the wild-type homozygote (ALDH2 *1/ALDH2 *1) have great tolerance to alcohol and can drink alcoholic beverages.Those with the mutant homozygote (ALDH2*2/ALDH2*2) are highly intolerant to alcohol and consequently do not drink because the alcohol-flush reaction in the result of excessive acetaldehyde accumulation and the unpleasant symptoms tend to reduce alcohol consumption.The heterozygotes have blood acetaldehyde concentrations -6 times higher than the wild-type homozygotes 87).Takeshita et al. reported that the mean amount of alcohol consumption in the heterozygotes is about one half that in the wild-type homozygotes 88).As the genotype of ALDH2 serves as an indicator of acetaldehyde exposure after alcohol consumption 87), alcohol drinking may be more harmful for the heterozygotes than the wild-type homozygotes.
Oriental populations of Mongoloid origin show up to 50% isozyme deficiency, whereas none of the Caucasoid and the Negroid populations screened have this isozyme abnormality 89).In a Japanese population, frequencies of wild-type homozygotes (ALDH2*1/ALDH2*1) and carriers with the mutant allele ALDH2 *2 are 56 and 44 %, respectively 90).
A positive association between upper aerodigestive tract cancers and ALDH2 polymorphism has been described 91.92).The levels of blood acetaldehyde related to the ALHD2 genotypes may be associated with the risk of colorectal cancer as well as upper aerodigestive tract cancers 9l, 92).Two case-control studies 93.94) in Japan reported an increased risk of colorectal cancer with the ALDH2*1/ALDH2*2 as compared with subjects with the ALDH2*1/ALDH2*1 genotype (Table 3).The first study 93) conducted among alcoholics showed a significantly increased OR of 3.35 in the presence of the ALDH2 *2 allele.All subjects with ALDH2 *2 allele were ALDH2 *1/ALDH2 *2 heterozygotes and the daily alcohol consumption did not differ between colon cancer and cancer-free alcoholics.An increased risk of colon cancer was also observed for the carriers with ALDH2 *2 allele after adjustment for alcohol consumption in the second study 94).They suggested that both drinkers with the heterozygote and non-drinkers with the mutant homozygote for ALDH2 gene may be at high risk for colon cancer.Although ALDH2 *2 allele actually affects drinking behavior, there exists considerable variation in the amounts consumed by the wild-type homozygotes and the heterozygotes 88).An interaction between the ALDH2 genotypes and alcohol consumption should be assessed in both studies.If the effect of alcohol consumption is limited to a genetically predisposed subgroup, evaluation of alcohol without genetic information may lead to it being dismissed as a cause of disease because the overall effect may be diluted when susceptible and nonsusceptible individuals are combined.
The relationship between alcohol use and colorectal cancer is less clear in Caucasians than in Orientals of Mongoloid origin.This disagreement may be partly due to the ethnic difference in ALDH2 *2 allele frequency.Japanese may be more sensitive to alcohol-induced colorectal carcinogenesis than Caucasians.
Much work remains in elucidation of the underlying mechanisms that link the ALDH2 *2 allele and colorectal cancer.

MTHFR POLYMORPHISM AND FOLATE / ALCOHOL
Various naturally occurring constituents in vegetables have been studied for their chemopreventive potential against colorectal cancer.Evidence that diets rich in vegetables protects against colorectal cancer is convincing 5).
Folate, which is rich in fresh leafy or cruciferous vegetables such as spinach, lettuce, cabbage, broccoli, brussels sprouts and cauliflower, has recently been the subject of much research interest.Animal and in vitro cell studies suggested a key role for folate in colon carcinogenesis 10.11)In epidemiological studies, high levels of dietary folate 95.96) or high blood folate levels 97)have been inversely associated with colorectal cancer.Furthermore, diets high in folate particularly in combination with low alcohol use, are associated with decreased risk of colorectal cancer 95,96).
Folate is essential for regenerating methionine, the methyl donor for DNA methylation, and for producing the purines and pyrimidines required for DNA synthesis.Decreased availability of folate may contribute to aberrations in DNA methylation.This imbalance in methylation of DNA is thought to result in abnormal expression of oncogenes and tumor suppressor genes 98).Evidence has been accumulated as regards the role for disturbances in DNA methylation in colon carcinogenesis ".On the other hand, methyl (folate) deficiency may also cause interference with the thymidylate biosynthesis and result in development of deoxynucleotide pool imbalances 100).This leads to increase of the uracil (U) content in DNA 100) and removal of this abnormal base might labilize DNA to strand breaks 101) Chromosome breaks appear to be important in nearly all human cancers and are especially common in colorectal cancer 102).
MTHFR is a critical enzyme regulating the metabolism of folate.MTHFR catalyzes the biologically irreversible reduction of 5,10-methyleneTHF to 5-methyl THF, the major circulatory form of folate and carbon donor for the remethylation of homocysteine to methionine.A variant of the human MTHFR gene that results in alanine to valine substitution has been described at by 677.This is an autosomal recessive mutation.This mutation codes for thermolabile enzyme with reduced MTHFR activity, resulting in elevated plasma homocysteine levels 103).Individuals with homozygous for this variant have been reported as having 30% of normal enzyme activity and heterozygotes have been reported as having 65% of normal enzyme activity 103).
The allele frequency of the mutant genotype of MTHFR in Japanese populations is estimated to be 0.33 105), which is comparable to that in European and American Caucasian populations [12][13][14].In contrast, the mutation has a very low prevalence in African-Americans, for whom the mutation was absent in homozygosity105).
Epidemiologic evidence, including that from prospective studies done in males 12.13), suggested that subjects with the homozygous mutation were consistently shown to have an approximately two-fold decrease in the risk of colorectal cancer compared with the wild-type homozygous and heterozygous genotype combined (Table 4).As shown in Table 4, the decreased risk associated with the mutant genotype was seen only in those consuming little or no alcohol 12.13).Alcohol has been hypothesized as having several possible mechanisms that may influence carcinogenesis, including one that may indirectly alter DNA methylation patterns by affecting the intestinal absorption, hepatobiliary metabolism and renal excretion of folate 106).The third case-control study 14) estimated that ORs for the mutant genotype was 0.8 (95% CI, 0.6-1.1)for males and 0.9 (95% CI, 0.6-1.2) for females as compared with that for the wild-type homozygote and showed a weak protective effect on colon cancer risk among individuals with the mutant homozygous genotype.The magnitude of the association was less than that reported previously.
Individuals with the mutant homozygote have lower levels of plasma 5-methyl THE because the conversion of 5,10methylene THE to 5-methyl THF, which is essentially irreversible under physiological conditions, is less efficient.Thus they may be less prone to get into imbalances of nucleotide pools during DNA synthesis.Low levels of 5-methyl THE probably result in lower levels of cellular methionine and Sadenosylmethionine, which is required for DNA methylation, and then lead to aberrant DNA methylation.When dietary methyl supply is enough, individuals with the mutant homozygous genotype are at reduced risk of colorectal cancer probably because higher levels of 5,10-methylene THE may prevent alternation of DNA synthesis.
In contrast, when the methyl content in dietary intake is low or depleted by alcohol consumption, the mutant homozygotes may be less able to compensate, leading to potentially oncogenic alternations in DNA methylation; the protective association of the MTHFR polymorphism is thus eliminated.Since alcohol may deplete 5-methyl THF, individuals with the mutant homozygote would be predicted to be particularly disadvantageous.
Thus far, all of three studies, to greater or lesser degrees, showed a protective effect of the mutant genotype on colorectal carcinogenesis.Dietary supply may be particular critical among the mutant homozygotes of MTHFR gene.

CONCLUSIONS
The application of molecular epidemiology to the study of genetic polymorphism of enzymes involved in xenobiotic metabolism has met with some difficulty.As discuss here, there are numerous conflicting reports on the association between different genotypes such as NAT genotypes and colorectal cancer risk.It is evident that there are large interethnic differences in those genes, as exemplified here by comparing allele frequencies in some populations.In the CYP1A1 locus, the detrimental mutations frequently found in the CYP1A1 gene among Japanese populations cannot be appreciably identified among Caucasian populations.This is also the case for ALDH2 gene.Thus connection of between cancer and a particular polymorphic site in one ethnic group might be of limited value of as a genetic marker for cancer in another ethnic group and extrapolations should be avoided.It is concluded that, at the present stage, a few of the polymorphic sites stated here can yet be used as biomarkers for increased/decreased colorectal cancer risk.To obtain a better understanding of genetic susceptibility to environmental factors, studies are needed to consider genetic characteristics of the population that may alter genetic susceptibility in combination with high and low risk environmental factors.
NS indicates that ORs were not significantly different from a value 1.-: not done a The combination of rapid genotype and frequent fired meat consumption versus the remainder combinations.b The combination of rapid NAT2/rapid NAT] genotypes and more frequent red meat intake versus the combination of rapid NAT2/rapid NAT] genotypes and less frequent red meat intake, among men aged 60 years old or older .

Table 1 .
Relation of GST polymorphisms to colorectal cancer .

Table 2 -
1. Relation of NAT2 phenotype to colorectal cancer.NS indicates that ORs were not significantly different from a value 1.

Table 2 -
2. Relation of NAT polymorphisms to colorectal cancer.

Table 3 .
Relation of ALDH2 polymorphism to colorectal cancer.a The heterozygotes versus the wild-type homozygotes.b Alcoholic men aged 40 years or older.c Adjusted for age at admission, daily alcohol consumption and amount of cigarette smoking.d Adjusted for age, daily alcohol consumption, amount of cigarette smoking and dietary factors.and Colorectal Cancer

Table 4 .
Relation of MTHFR polymorphism to colorectal cancer.OR 1: Multivariate-adjusted risk.OR 2: Risk for the combination of genotype and alcohol use.NS: P value-was.not less than 0.05.a The mutant and heterozygous genotypes versus wild-type genotype.b The mutant homozygotes versus the wild-type homozygotes and heterozygotes combined.