Introduction

In its 2009 review of human carcinogens, a working group of the International Agency for Research on Cancer (IARC) concluded that there is sufficient evidence in humans to affirm the carcinogenicity of alcohol consumption. Sufficient evidence is considered available to associate alcohol consumption with various types of cancer, including cancers of the oral cavity, pharynx and larynx, esophagus, colorectum, and female breast, and hepatocellular carcinoma [1•]. In that evaluation, the authors suggested that acetaldehyde associated with alcohol consumption also causes both esophageal and upper aerodigestive tract cancers [2].

The association between alcohol consumption and colorectal cancer risk has been studied closely for a number of decades in various populations. Recent systematic reviews and meta-analyses have provided clues to quantitative evaluation of the association. In addition, increased interest in genetic studies has attracted further attention to possible gene-environment interactions on the effect of alcohol exposure on colorectal cancer.

We review epidemiological evidence obtained to date for the association between alcohol consumption and colorectal cancer risk and summarize possible mechanisms of alcohol-related carcinogenesis.

Epidemiological Evidence for Alcohol Consumption and Colorectal Cancer Risk

A substantial number of epidemiological studies have observed an increased risk of colorectal cancer with higher consumption of alcohol. The evaluation of the World Cancer Research Fund and American Institute for Cancer Research (WCRF/AICR) report in 2007 [3] reported ample, generally consistent evidence on this association, with apparent dose-response and plausible mechanisms. This report also mentioned that consumption of more than 30 g of ethanol per day from alcohol drinks is a convincing cause of colorectal cancer in men, and probably in women.

A recently detailed systematic review and meta-analysis of this association targeting 61 epidemiological studies (27 cohort studies and 34 case-control studies) in various populations provided strong evidence for the association between moderate and heavy alcohol drinking and colorectal cancer risk, in which the summary relative risk (RR) was 1.21 (95 % confidence interval (95 % CI) 1.13–1.28) for moderate drinking and 1.52 (95 % CI 1.27–1.81) for heavy drinking [4•]. In that review, epidemiological results were found generally to show consistently similar associations between anatomic subsite, namely colon and rectum, and between sexes. This review also supported the positive dose-response association by meta-analyses, with RRs of 1.07 (95 % CI 1.03–1.4), 1.38 (95 % CI 1.28–1.5), and 1.82 (95 % CI 1.41–2.35) for daily alcohol consumption of 10, 50, and 100 g, and pointed to a stronger association in heavy drinkers in Asian (RR 1.81, 95 % CI 1.33–2.46) than in western populations [4•]. Several, recent, large-scale pooled analyses have evaluated the association with greater sensitivity in settings of combined population studies. The EPIC study [5] showed that both lifelong and baseline alcohol intake elevated colon and rectal cancer risk, with risk apparently increasing with alcohol drinking more than 30 g per day and a tendency for higher risk at the distal site. The study also reported a tendency to a somewhat higher colorectal cancer risk for beer than for wine. Another pooled analysis by the Pooling Project of Prospective Studies of Diet and Cancer showed somewhat different result for type of alcohol [6], in which risk was higher for wine than beer or liquor, albeit that differences among specific beverages were not statistically meaningful.

These pooled analyses from western populations [5, 6] and another pooled analysis from Japanese populations [7] all showed a positive association, although the magnitude of relative risk was much higher in the Japanese population (pooled RR 1.42 (95 % CI 1.21–1.66), 1.95 (95 % CI 1.53–2.49), 2.15 (95 % CI 1.74–2.64), and 2.96 (95 % CI 2.27–3.86), for 23 to <46 g, 46 to <69 g, 69 to <92 g, and ≥92 g respectively, compared with nondrinkers) [7]. These higher RRs in Japanese populations might be due to their higher prevalence of the slow-metabolizing aldehyde dehydrogenase variant [8], which causes elevated blood levels of acetaldehyde.

Because the prevalence of drinking differs among countries depending on culture, the proportion of colorectal cancer cases attributable to alcohol drinking, called the population-attributable fraction (PAF), is a helpful means of estimating the burden in each country. At a country level, the recent Global PAF of alcohol consumption for colorectal cancer was estimated at 4.6 % in men and 1.7 % in women [9], and by country at 11.2 % in men and 2.7 % in women in France [10], 15.5 % and 6.9 % in the United Kingdom [11], 2.3 % in men and 0.3 % in China [12], and 32.9 % in men and 2.1 % in women in Japan [13], with these cases regarded as having been preventable by the avoidance of alcohol drinking.

The association between alcohol consumption and colorectal cancer varies by dietary factors, including folate status. Low folate consumption is suggested to increase the risk of colorectal cancer [14]. The EPIC study showed that the association between alcohol intake at baseline and colorectal cancer risk was stronger in subjects with low folate intake, with a borderline significant interaction [5]. A similar result was observed in another pooled analysis [6].

Possible Mechanism

Alcohol drinking results in acetaldehyde exposure, and the carcinogenicity of acetaldehyde associated with alcoholic beverage to humans has now been confirmed as Group 1 by the IARC Working Group [1•]. However, the specific mechanisms of the effect of alcohol intake on colorectal carcinogenesis in humans have yet to be fully defined. One notable mechanism identified to date is considered to relate to the folate pathway. Dietary folate influences DNA methylation, synthesis, and repair [14]. Folate has the potential to decrease carcinogenesis by various mechanisms in relation to DNA synthesis and methylation [14]. Alcohol may induce DNA hypomethylation [15] through its antifolate effect, and subsequently induce folate deficiency [16], under which the condition of high alcohol combined with low folate may markedly increase the risk of colorectal cancer. Acetaldehyde itself is responsible for colorectal carcinogenesis via its inhibition of the intestinal absorption of potentially anticarcinogenic nutrients, such as folate [17] and calcium [18].

Genetic Polymorphisms Relating to Alcohol Metabolism

Ethanol is generally oxidized in the liver by alcohol dehydrogenases (ADH) to acetaldehyde, which is regarded as carcinogenic in humans (Group 1), and is then further detoxified to acetate by aldehyde dehydrogenases (ALDH). Genetic variants that cause functional differences in these enzyme activities result in differences in acetaldehyde exposure [19]. East Asians, such as Japanese, have a relatively high prevalence of low or no ALDH2 activity, which is considered to affect cancer etiology among drinkers [19]. For ALDH2 polymorphism, ALDH2*1 and ALDH2*2 are the active and inactive subunits of the alleles. ALDH2-deficient phenotypes are ALDH2*2 homozygous (null ALDH2 activity) and ALDH2 heterozygous (6 % residual activity) [20]. Nevertheless, several published papers on colorectal cancer in Asian populations did not observe a consistent effect of ALDH2 polymorphism and alcohol consumption on colorectal cancer risk [21, 22].

The enzyme 5,10-methylenetetrahydrofolate reductase (MTHFR) plays a crucial role in folate metabolism. MTHFR acts as catalyzer for the conversion of 5,10-methylenetetrahydrofolate, which is required for purine and thymidine syntheses, to 5-methyltetrahydrofolate, which is the primary circulatory form of folate. Folate is crucial for methionine synthesis. A common mutation (677C→T) in MTHFR reduces its enzyme activity and induces lower levels of 5-methyltetrahydrofolate [23]. Compared with the CC genotype, the TT or CT genotype results in a substantial reduction in MTHFR activity [24]. A review on this variant and colorectal cancer risk reported that the TT genotype was associated with a moderately reduced risk in most studies [25].

In general, although the role of genetic polymorphisms associated with colorectal cancer remains inconclusive due to a limited number of studies and inconsistent results, several clues to understanding the mechanism of colorectal carcinogenesis in relation to alcohol are available, namely with regard to genetic polymorphisms such as ALDH2 and MTHFR.

Conclusions

The positive association between alcohol consumption and colorectal cancer risk is convincing. Although the site-specific mechanism of the influence of alcohol intake on colorectal carcinogenesis and the effect of gene-environment interaction requires clarification, avoidance of excess alcohol drinking will substantially contribute to reducing the risk and burden of colorectal cancer attributable to alcohol drinking.