Alcohol is a probable risk factor with regard to colorectal neoplasm and is metabolized to the carcinogen acetaldehyde by the genetically polymorphic alcohol dehydrogenase 3 (ADH3) enzyme. We evaluated whether the association between alcohol and colorectal adenomas is modified by ADH3 polymorphism. We recruited 433 cases with adenomatous polyps and 436 polyp-free controls among Caucasians undergoing endoscopy between 1995 and 2000. Frequency and amount of habitual alcohol consumption were assessed by beverage type, using a validated self-administered food frequency questionnaire. All participants provided blood for genotyping of ADH3. Multivariate analyses adjusting for gender, age, and indication for endoscopy showed that alcohol increased the risk of colorectal adenomas among women [odds ratio (OR), 1.8; 95% confidence interval (CI), 1.0–3.2, ≥10 versus <1 drink/week]. Among men, the risk of adenomas was increased only for those consuming > 21 drinks/week (OR, 1.8; 95% CI, 0.9–3.8, compared with men drinking < 1 drink/week). Among subjects in the highest tertile of alcohol consumption, those with the ADH3*1/*1 genotype were at higher risk (OR, 1.8; 95% CI, 1.0–3.1) than those with other ADH3 genotypes (OR, 1.2; 95% CI, 0.7–1.9) when compared with those in the lowest tertile with ADH3*1/*2 or ADH3*2/*2 genotypes. In conclusion, our findings are consistent with results of other studies, suggesting that alcohol consumption elevates the risk of adenomatous colorectal polyps. ADH3 polymorphism may modify the association between alcohol consumption and colorectal adenomas.

The majority of colorectal cancers is thought to arise from adenomas (1, 2). Alcohol consumption is a probable risk factor with respect to colorectal adenomas (2) and colorectal cancer (3).

This increased risk might be caused by ethanol via cocarcinogenesis, induction of DNA hypomethylation, or tumor promotion (4) but might also result from its metabolite acetaldehyde, which was qualified as a probable carcinogen to humans (5).

Acetaldehyde is mainly produced from ethanol via oxidation by the ADH3 enzyme (6) and is subsequently detoxified into acetate by ALDH. Ethanol metabolism predominates in the liver, but ADH is also expressed in other tissues, including the colon (7, 8). ADH isoenzymes primarily involved in ethanol oxidation consist of subunits encoded by ADH2 and ADH3 genes (6). In contrast to ADH2(9, 10, 11), the ADH3 gene is highly polymorphic in Caucasians. Of the two allelic variants, ADH3*1 is associated with higher enzyme activity than ADH3*2 and occurs in Caucasians at frequencies of 55–63%. Polymorphism in the ALDH gene is extremely rare in Caucasians (12, 13).

To our knowledge, thus far, no studies with regard to the role of ADH3 genotype in alcohol-associated colorectal neoplasm have been published. Risk of oral cancer was five times higher among heavy drinkers (>57 drinks/week) with the ADH3*1/*1 genotype than in heavy drinkers with other ADH3 genotypes (14). Among alcoholics, the risk of laryngeal cancer, but not of oropharyngeal cancer, was increased only in combination with the ADH3*1/*1 genotype (15). Similarly, the risk of premenopausal breast cancer was increased only in women consuming > 6.5 alcoholic drinks/month with the ADH3*1/*1 genotype (16). However, other studies investigating oral (17, 18), breast (19), and head and neck (20) cancers did not indicate that ADH3 polymorphism plays a major role in the association between alcohol and cancer.

ADH3 polymorphism might be relevant to alcohol-associated colorectal carcinogenesis. Acetaldehyde might accumulate in the colon epithelium (6) where ADH activity is present (8), whereas ALDH activity is relatively low (21). If the rate of acetaldehyde accumulation and the duration of exposure to acetaldehyde depend on ADH3 genotype, the association between alcohol consumption and adenomas in colon and probably also rectum would probably be strongest in subjects with the ADH3*1/*1 genotype. Because acetaldehyde accumulation increases with alcohol consumption, the role of ADH3 polymorphism might be most relevant in heavy drinkers.

In this study, among Dutch Caucasians, we evaluated whether alcohol and ADH3 genotype interplay in the etiology of adenomatous colorectal polyps. We also investigated the effect of ADH3 genotype among those with highest alcohol consumption.

Population.

Cases and controls were recruited among those undergoing endoscopy in the outpatient clinics of eight hospitals in the Netherlands between June 1997 and June 2000. Medical ethical committees of all participating hospitals and of Wageningen University approved the study protocol.

Potential participants were recruited at the time of endoscopy by the endoscopy staff (47%) or were selected at regular intervals using endoscopy reports of all patients who had undergone endoscopy in the previous 3 months, and these were invited by mail (53%). Eligible subjects were Dutch speaking, Caucasian, aged 18–75 years at the time of endoscopy, had no hereditary forms of colorectal cancer (i.e., hereditary nonpoly-posis colorectal cancer, familial adenomatous polyposis, and other hereditary syndromes), chronic inflammatory bowel disease, history of colorectal cancer or (partial) bowel resection, or serious disabling morbidity. After obtaining informed consent, blood samples were drawn for DNA analysis, and dietary and lifestyle questionnaires were administered. Overall response was 54%.

We defined cases as those having had at least one histologically confirmed colorectal adenomatous polyp in their life; controls had no history of any type of polyps. Complete colonoscopy was conducted on 82% of the cases and 60% of the controls; in 7% of cases and 17% of controls, complete visualization of the colon was achieved by sigmoidoscopy followed by colon X-ray.

Among cases, main indications for endoscopy were routine check-up for adenoma recurrence (37%), anal bleeding (27%), and large bowel complaints (15%), whereas controls mainly underwent endoscopy because of large bowel complaints (38%), defecation problems (21%), or anal bleeding (17%).

Medical files were checked for additional information with regard to medical history and information on polyp recurrence, size, localization, histology, and the number of excised polyps.

In retrospect, based on information from questionnaires and medical files, we excluded 170 participants who did not meet the eligibility criteria, mainly because of nonadenomatous or unknown types of polyps (59%). In addition, we used complete information of 64 subjects (40 cases and 24 controls, all having undergone complete visualization of their colon) meeting our criteria, recruited between December 1995 and June 1997, from a preceding study on somatic mutations in colorectal adenomas conducted in one of eight hospitals (22). The recruitment procedures as well as the questionnaires used in this preceding study were essentially the same as those described here for the main study population. This increased the study population from 861 to 925 subjects.

Questionnaires.

All invited subjects received a short questionnaire inquiring about important characteristics such as age, gender, alcohol consumption, education level, and smoking habits. Approximately one-third of the subjects who did not want to participate in the study completed this short questionnaire. Although they were older, they did not differ from the participants with respect to gender, education level, smoking habits, and alcohol consumption.

Participants additionally received dietary and lifestyle questionnaires and were requested to complete these with regard to habits in the year previous to their last endoscopy or complaints.

To assess dietary habits, we used a standardized and validated semiquantitative food frequency questionnaire, described in detail by Ocké et al.(23). By means of this questionnaire, consumption of alcoholic beverages was assessed with respect to beer, white wine, red wine, ports, and liquors separately. Subjects could choose to report average consumption in glasses/day, week, month, or year. Reproducibility of alcohol consumption as assessed by this questionnaire was high for both males and females (r = 0.91) as was its relative validity (r = 0.74 for males, r = 0.87 for females compared with the means of 12 24-h recalls; Ref. 23). Intakes of total energy and of various nutrients and ethanol were calculated by using a computerized version of the Dutch food composition table. A Dutch alcoholic consumption contains ∼10 g of ethanol. Nutrients, except ethanol, were adjusted for total energy intake using the residual regression method (24).

Laboratory Analyses.

Blood samples were stored at −20°C. DNA was isolated from 200 μl of frozen whole blood, using the QIAamp blood kit (Qiagen, Inc.), diluted to a concentration of ∼20 ng/μl, and stored at 4°C until analyzed.

We used a PCR-RFLP method for determination of ADH3 genotype. A 145-bp fragment of exon 8 of the ADH3 gene was amplified using primers described by Groppi et al. (Ref. 25; 5′-GCTTTAAGAGTAAATATTCTGTCCCC-3′ and 5′-AATCTACCTCTTTCCAGAGC-3′). To check for DNA cross-contamination, one in eight samples contained no DNA but contained water instead.

For RFLP analysis, SspI digested the ADH3*1 allele into fragments of 67, 63, and 15 bp and the ADH3*2 allele into fragments of 130 and 15 bp. DNA fragments were separated on an ethidium bromide stained agarose gel (4%) and visualized under UV light.

To control the specificity of ADH3 genotyping, a random sample of primary PCR products was digested by NlaIII, cleaving the ADH1 and ADH2 genes that are closely related to ADH3 but leaving ADH3 intact (25). Laboratory personnel was blinded to case-control status. DNA of 10 participants was not available, and it was not possible to genotype another 8 samples (<1%) for ADH3.

Data Analysis.

Subjects with incomplete dietary data (n = 38) were excluded, as were the 18 subjects of whom ADH3 genotype was not assessed. The analyses thus included 869 subjects: 433 cases and 436 controls. We studied total alcohol consumption in glasses/week, calculated by summing the separately reported intakes of beer, wines, ports, and spirits, as well as total ethanol intake in grams/day from all dietary sources, including small amounts from sauces, puddings, chocolates, and low-alcohol beer. Alcohol consumption was divided in tertiles based on the distribution in the total study population. To additionally evaluate the effect of a combination of ADH3 polymorphism with high alcohol consumption, we defined high alcohol consumption as the consumption of >3 drinks/day because this amount exceeds the recommended daily maximum for both men and women. ADH3*1/*1 was considered as the high-risk genotype and compared with the combination of ADH3*1/*2 and ADH3*2/*2 genotypes.

The analyses were conducted using Statistical Analysis Software (SAS version 6.12, SAS Institute, Cary, NC). All tests of statistical significance were two-sided. To test for linear trend, we modeled the tertile of alcohol intake as a continuous variable in the logistic regression model in which each tertile was assigned its median value. Logistic regression models were used to calculate ORs and 95% CIs. Factors selected as possible confounders were age, gender, body mass index, indication for endoscopy, center, cigarette smoking, physical activity, family history of colorectal cancer, education level, use of nonsteroidal anti-inflammatory drugs, total energy intake, consumption of vegetables, fruit, total meat and red meat, and nutrients related to these food groups. Variables related to colorectal adenomatous polyps as well as to exposure at P < 0.5 (26) were separately entered as covariates in the regression models. None of these changed the OR for alcohol consumption by >10%. However, we included age and indication for endoscopy (complaints-related, screening, and other/unknown) in the multivariate models to control for potential confounding. Analyses of the total population were additionally adjusted for gender. Gender-stratified analyses were conducted because of (a) different male-female ratios between cases and controls, (b) gender-specific patterns of alcohol consumption, and (c) gender-specific differences in alcohol vulnerability (27).

To evaluate the possible interplay between ADH3 genotype and alcohol consumption, the group with a combination ADH3*1/*2 or ADH3*2/*2 genotypes and low-alcohol consumption (lowest tertile) served as reference category.

As different risk factors may operate in different stages of carcinogenesis, we conducted case-case analyses for adenoma recurrence (primary versus recurrent), size (<1 cm versus ≥1 cm), localization (proximal versus distal) and number (multiple versus single) of polyps.

To check whether former adenomas among cases or undetected right-sided polyps in controls could have biased our results, we repeated all analyses after restriction of our study population to cases with first diagnosis of adenomas not longer than 1 year ago (n = 299) and controls with complete visualization of the colon (n = 334).

The case group contained more men than the control group (55 versus 37%). Table 1 shows characteristics of the study population for cases and controls stratified by gender. Among women and men, cases were significantly older than controls, and less often underwent endoscopy because of bowel complaints. Among women, cases had a higher intake of alcohol (especially of spirits and fortified wines), vegetables, fruits, and folic acid and had changed their diet less frequently because of bowel complaints compared with controls. Among men, cases more frequently (had) smoked and consumed less energy and folic acid than controls. There were no differences with respect to ADH3 genotype. The distribution of ADH3 was in Hardy-Weinberg equilibrium.

Alcohol intake from all sources ranged from 0 to 10 glasses/day. In the control group, the median alcohol consumption among men was 9 consumptions/week, whereas the median consumption of alcohol among women was only 1 glass/week. Among cases, median alcohol intake was 10 drinks/week among men and 2.5 drinks/week among women.

Alcohol intake correlated positively with energy intake and education level. Also, alcohol consumption was related to (history of) cigarette smoking and family history of colorectal cancer (data not shown).

In Table 2, risk estimates for the association between the number of drinks/week (in tertiles) and colorectal adenomas are shown. Overall, alcohol intake was weakly but not statistically significantly related to colorectal adenomas. Among women, alcohol consumption significantly increased the risk of colorectal adenomas. When consumers of >21 alcoholic consumptions/week were compared with consumers of <1 glass/week, alcohol appeared to also be a risk factor for men. We did not find increased risks for women consuming > 21 glasses/week, probably because this category only contained 10 cases and 12 controls (data not shown). When using sex-specific tertiles (<4.5, 4.5–15.9, and ≥16 drinks/week for men and <0.2, 0.2–3.9, and ≥4.0 drinks/week for women), results were comparable with those presented in Table 2 (i.e., OR for the third compared with the first tertile, 1.1; 95% CI, 0.6–1.9 and OR, 1.5; 95% CI, 0.9–2.6 for men and women, respectively).

The analysis of alcohol intake from all sources in grams/day, which was slightly different from the daily alcohol intake in grams acquired from alcoholic beverages only, yielded ORs comparable with those presented in Table 2. No specific type of beverage was responsible for the increased risk of colorectal adenomas (data not shown). Case-case analyses did not indicate that alcohol consumption was related to specific adenoma characteristics such as location, size, type, and number of adenomas (data not shown).

After exclusion of cases who had previously been diagnosed with adenomas and of controls whose proximal colon was not examined, results remained similar (OR and 95% CI for women in the highest tertile of alcohol consumption, 2.2 and 1.1–4.2 and for men, 1.0 and 0.5–2.0). After exclusion of those who underwent endoscopy because of large bowel complaints, the gender and age adjusted OR inflated moderately among women (OR, 1.9; 95% CI, 1.0–3.4) but not among men (OR, 1.1; 95%, CI 0.5–2.1).

In Table 3, we show risk estimates for the combined associations of alcohol consumption and ADH3 genotype with colorectal adenomas. The association between alcohol and adenomas was not markedly influenced by ADH3 polymorphism, although the risk of adenomas was highest among subjects who had the ADH3*1/*1 genotype and were in the upper tertile of alcohol consumption (Table 3). When comparing consumers of >21 alcoholic drinks/week to those consuming <1 drink/week, the risk increased most markedly for men with the ADH3*1/*1 genotype (OR, 2.8; 95% CI, 1.0–8.3) and less so for men with different genotypes (OR, 1.6; 95% CI, 0.7–4.1). However, the interaction term was not statistically significant (P = 0.4). All ORs (sex-specific, age adjusted, and multivariate) for interaction varied between 0.95 and 1.05 and were nonsignificant.

In this first study on alcohol consumption and ADH3 genotype in the epidemiology of colorectal adenomas, we observed that alcohol consumption increased the risk of colorectal polyps most markedly among women and that this association may be influenced by ADH3 genotype.

We recruited both cases and controls among those undergoing endoscopy. In the Netherlands, endoscopies are not routinely conducted for screening purposes like in the United States. Consequently, in our study, endoscopies were mainly conducted for bowel pain, anal bleeding, or defecation problems (64%). These complaints may influence dietary patterns. Our study population might also be more health conscious than the general population. This implies that our findings cannot easily be extrapolated to the general population. However, alcohol consumption in our control group was similar to the habitual alcohol consumption we assessed using the same questionnaire in a random sample (n = 1935) from the general population inhabiting the same regions as the controls.4

Of those invited, ∼54% were willing to participate. Selection bias might have occurred if habitual alcohol consumption influenced the probabilities of being invited or of participating. It is not likely that alcohol consumption influenced the chance of being invited because habitual alcohol consumption was unknown at the time of the selection for almost all subjects (>95%). Moreover, participants did not differ in alcohol consumption from those who refused to participate but who completed the short questionnaire.

The control group consisted of significantly more women, possibly because women are more likely than men to undergo endoscopy for major bowel complaints such as irritable bowel syndrome, which was found to be more prevalent in Dutch women than in men (28). Also, cases were older than controls. Gender and age differences between cases and controls were also observed in other case-control studies on colorectal adenomas (29, 30, 31). Cases with history of adenomas might be overrepresented in our study population because these had a higher probability of being invited, and this might have introduced bias. We therefore included indication for endoscopy in our multivariate model. Exclusion of those with a history of adenomas yielded essentially the same results.

Inclusion of controls with bowel complaints probably did not lead to important overestimation of the true associations between alcohol and adenomas, although bowel complaints occurred more often in the control group and were associated to lower alcohol consumption in women. ORs only marginally inflated after exclusion of those undergoing endoscopy because of bowel complaints. We expect no misclassification by inclusion of controls with incomplete visualization of the colon (22%). In theory, these could have proximal adenomas, leading to bias toward the null. However, exclusion of controls with incomplete colonoscopy did not change our results.

Recall bias might have occurred because most cases and controls were aware of their status at the time of completion of the questionnaires. If alcohol would have been known as a risk factor for polyps, cases might have reported lower or higher intake than their true intake of alcohol. However, alcohol consumption is generally believed to increase the risk of several cancers but probably not of colorectal adenomas. Indeed, none of those previously being diagnosed with adenomas indicated to have changed alcohol consumption because of colorectal adenomas.

We assessed alcohol consumption by a food frequency questionnaire. Although a validation study of our questionnaire showed that habitual alcohol intake might be systematically underestimated, especially by men, subjects were appropriately ranked with regard to alcohol consumption (23). Ideally, per beverage type, both the frequency and the number of drinks/occasion should be noted (32). We had no information with respect to drinking patterns or on drinking habits over the years. Recent drinking habits might well reflect those in the past, as in a Dutch cohort, alcohol consumption patterns were found to be relatively stable, especially among men (33). The same was concluded from a follow-up study among British male doctors (34).

Allele frequencies of ADH3*1 and ADH3*2 among controls were 59 and 41%, respectively, which is similar to frequencies reported from other Caucasian populations (10, 11, 16, 19). We correctly amplified ADH3 and not ADH1 or ADH2 in all samples (as checked by digestion of a random sample of PCR products with the restriction enzyme NlaIII). Use of an internal control in RFLP analysis proved that all digestions were successful.

Our finding that alcohol consumption increases the risk of adenomatous colorectal polyps corresponds with the results of most previously conducted studies (29, 30, 31, 35, 36, 37, 38, 39, 40). Among women, the risk of adenomas increased with consumption of ≥10 alcoholic beverages/week, whereas among men, the risk was increased only with consumption of >21 alcoholic beverages/week. It is difficult to compare these results with those obtained in other studies because different cutoff points are used and gender-specific results are not always presented. In general, like in our study population, men consume more alcohol than women do, and the range of alcohol consumption is wider in men than in women (33, 41). A possible explanation of our results is that the threshold for an effect of alcohol on adenomas could be higher in men than in women. Women are more vulnerable to alcohol than men, mainly because of a lower rate of first-pass ethanol metabolism in the stomach (27, 42). Because blood ethanol levels are higher in women than in men at equal consumption and ethanol reaches colonic epithelium via the blood circulation, this might imply that at equal intake of alcohol, the colonic epithelium of women is exposed to higher levels of ethanol and acetaldehyde than that of men. However, the differences we found between men and women might also reflect sex-specific differences in alcohol consumption. Among women, 23% were not exposed to alcohol and 19% consumed ≥10 drinks/week, whereas among men, only 11% were nondrinkers and >50% consumed ≥10 drinks/week. Possibly, the difference in adenoma risk is much greater between no and low exposure to alcohol than the difference between low and medium exposure to alcohol.

Alcohol did not specifically increase adenoma recurrence and other adenoma characteristics. To our knowledge, only Boutron et al.(43) found that alcohol consumption specifically increased risk of large adenomas.

We did not find specific types of beverages to be responsible for the observed increase in risk, which is in line with the conclusion of the World Cancer Research Fund expert committee reading that “the effect generally seems to be related to total ethanol intake, irrespective of the type of drink” (44).

The effect of ethanol is probably cocarcinogenic rather than carcinogenic (4, 45). In contrast, its major metabolite acetaldehyde is a probable carcinogen and was found to form adducts, induce DNA cross-links, chromosomal aberrations, and sister chromatid exchanges in vitro and inhibit DNA repair enzymes (5). Therefore, we hypothesized that polymorphism of the ADH3 gene, encoding the principal enzyme oxidizing ethanol to acetaldehyde, would play a role in the association between alcohol and adenomas. We found stronger associations between alcohol consumption and colorectal adenomas in carriers of the ADH3*1/*1 genotype than in those with other ADH3 genotypes. However, interaction terms for ADH3 genotype and alcohol consumption were not statistically significant, possibly because modest gene-environment interactions can only be studied in populations with several thousands of subjects (46). An alternative explanation is that the role of ADH3 genetic polymorphism might be obscured by ADH production (47) or acetaldehyde production by intestinal microflora (21). Because of these effects, ADH3 genotype might be especially important among heavy drinkers. In contradiction to our expectations, the potential role of ADH3 polymorphism did not become more pronounced with high alcohol consumption, i.e., >21 drinks/week. However, our study population only included very few heavy drinkers or alcoholics, and this may have influenced our results. Alternatively, ADH-catalyzed oxidation of ethanol might be less important because other enzymatic systems are upgraded in heavy drinkers (48).

Thus far, ADH3 polymorphism has not been considered in studies on colorectal adenomas. Other studies with regard to the role of ADH3 polymorphism in the association between alcohol and neoplasm concerned oropharyngeal, laryngeal, head and neck, and breast cancer (14, 15, 16, 17, 18, 19, 20). Three of these seven studies indicated that drinkers with the ADH3*1/*1 genotype are at higher risk of neoplasm than those carrying ADH3*1/*2 and ADH3*2/*2 genotypes (14, 15, 16).

We conclude that alcohol consumption elevates the risk of adenomatous colorectal polyps. Although we found no indications for a statistically significant effect modification by ADH3, we found that high alcohol consumption most markedly increased the risk of colorectal adenomas in subjects with the ADH3*1/*1 genotype. These findings need confirmation in larger studies. Our hypothesis that the influence of ADH3 genotype becomes relevant at high ethanol concentrations should preferably be tested in a large population with higher alcohol consumption. Moreover, exposure of the human colon to ethanol and acetaldehyde and effects of this, as well as the role and impact of ADH synthesis by gastrointestinal bacteria, need to be studied in more detail.

The costs of publication of this article were defrayed in part by the payment of 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

This work was supported by Dutch Cancer Society (KWF/NKB) Grant LUW 96-1374.

3

The abbreviations used are: ADH, alcohol dehydrogenase; ALDH, aldehyde dehydrogenase; OR, odds ratio; CI, confidence interval.

4

E. W. Tiemersma, D. W. Voskuil, M. Kriege, E. Kampman, unpublished data.

Table 1

Characteristics of the study population by gender

Women (N = 471)Men (N = 398)
Cases (n = 196)Controls (n = 275)Cases (n = 237)Controls (n = 161)
Age (yr), mean ± SD 58.6 ± 10.7 49.5 ± 15.0a 59.3 ± 10.6 51.9 ± 12.2a 
Genotype, %     
ADH3*1/*1 35.2 36.7 36.3 34.2 
ADH3*1/*2 48.0 47.6 44.7 46.6 
ADH3*2/*2 16.8 15.6 19.0 19.3 
Alcohol intake, median (25th, 75th percentile)     
 Ethanol (g/day) 3.9 (0.3; 14.6) 1.5 (0.1; 8.7)a 14.2 (4.3; 30.4) 14.5 (3.5; 26.5) 
 Alcoholic consumptions/wk 2.5 (0.2; 10.0) 1.0 (0; 6.0)a 10.0 (2.3; 21.1) 9.3 (2.5; 18.0) 
  Beer 0 (0; 0) 0 (0; 0.1) 0.5 (0; 7.0) 2.0 (0; 7.0) 
  Wine (red, white, rosé) 0.5 (0; 4.0) 0.2 (0; 2.7) 0.5 (0; 5.2) 0.7 (0; 4.0) 
  Spirits (incl port, sherry etc) 0.1 (0; 1.1) 0 (0; 0.5)a 0.9 (0; 6.5) 0.5 (0; 4.0) 
Dietary habits, median (25th, 75th percentile)     
 Total energy (KJ/day) 7,684 (6,539; 8,979) 7,440 (6,419; 8,978) 9,293 (7,942; 10,951) 10,239 (8,435; 11,623)a 
 Vegetables (g/day) 122 (98, 154) 114 (90, 144)a 110 (86, 140) 112 (83, 137) 
 Fruit (pieces/day) 1.5 (1.0, 2.3) 1.0 (0.6, 2.0)a 1.0 (0.4, 2.0) 1.0 (0.4, 2.0) 
 Red meat (g/day) 57 (28, 77) 48 (26, 73) 74 (48, 91) 73 (42, 90) 
 Coffee (cups/day) 4.0 (2.5; 5) 4.0 (2; 6) 4.0 (3; 6) 5.0 (3; 6) 
 Folic acid (μg/day) 187 (162; 221) 175 (153; 204)a 204 (171; 245) 216 (187; 252)a 
 Retinol (μg/day) 554 (392; 785) 499 (346; 744) 736 (524; 1223) 838 (517; 1189) 
Medical history, %     
 Dietary changes because of bowel complaints 18.9 33.8a 17.3 19.3 
 Family history of colorectal cancer 26.5 19.3 20.3 19.3 
 Previous diagnosis of adenomas 45.9 n.a.b 40.1 n.a. 
 Complaints-related endoscopy indication 46.9 78.2a 55.7 74.5a 
Other characteristics     
 Body mass index (kg/m2), mean ± SD 25.8 ± 4.5 25.1 ± 4.5 26.3 ± 3.3 25.9 ± 3.1 
 Cigarette smoking, % never smokers 45.9 52.7 32.1 46.6a 
 High physical activity, % 35.7 34.6 20.3 26.7 
 Low education level, % 44.4 33.1 30.0 26.1 
Women (N = 471)Men (N = 398)
Cases (n = 196)Controls (n = 275)Cases (n = 237)Controls (n = 161)
Age (yr), mean ± SD 58.6 ± 10.7 49.5 ± 15.0a 59.3 ± 10.6 51.9 ± 12.2a 
Genotype, %     
ADH3*1/*1 35.2 36.7 36.3 34.2 
ADH3*1/*2 48.0 47.6 44.7 46.6 
ADH3*2/*2 16.8 15.6 19.0 19.3 
Alcohol intake, median (25th, 75th percentile)     
 Ethanol (g/day) 3.9 (0.3; 14.6) 1.5 (0.1; 8.7)a 14.2 (4.3; 30.4) 14.5 (3.5; 26.5) 
 Alcoholic consumptions/wk 2.5 (0.2; 10.0) 1.0 (0; 6.0)a 10.0 (2.3; 21.1) 9.3 (2.5; 18.0) 
  Beer 0 (0; 0) 0 (0; 0.1) 0.5 (0; 7.0) 2.0 (0; 7.0) 
  Wine (red, white, rosé) 0.5 (0; 4.0) 0.2 (0; 2.7) 0.5 (0; 5.2) 0.7 (0; 4.0) 
  Spirits (incl port, sherry etc) 0.1 (0; 1.1) 0 (0; 0.5)a 0.9 (0; 6.5) 0.5 (0; 4.0) 
Dietary habits, median (25th, 75th percentile)     
 Total energy (KJ/day) 7,684 (6,539; 8,979) 7,440 (6,419; 8,978) 9,293 (7,942; 10,951) 10,239 (8,435; 11,623)a 
 Vegetables (g/day) 122 (98, 154) 114 (90, 144)a 110 (86, 140) 112 (83, 137) 
 Fruit (pieces/day) 1.5 (1.0, 2.3) 1.0 (0.6, 2.0)a 1.0 (0.4, 2.0) 1.0 (0.4, 2.0) 
 Red meat (g/day) 57 (28, 77) 48 (26, 73) 74 (48, 91) 73 (42, 90) 
 Coffee (cups/day) 4.0 (2.5; 5) 4.0 (2; 6) 4.0 (3; 6) 5.0 (3; 6) 
 Folic acid (μg/day) 187 (162; 221) 175 (153; 204)a 204 (171; 245) 216 (187; 252)a 
 Retinol (μg/day) 554 (392; 785) 499 (346; 744) 736 (524; 1223) 838 (517; 1189) 
Medical history, %     
 Dietary changes because of bowel complaints 18.9 33.8a 17.3 19.3 
 Family history of colorectal cancer 26.5 19.3 20.3 19.3 
 Previous diagnosis of adenomas 45.9 n.a.b 40.1 n.a. 
 Complaints-related endoscopy indication 46.9 78.2a 55.7 74.5a 
Other characteristics     
 Body mass index (kg/m2), mean ± SD 25.8 ± 4.5 25.1 ± 4.5 26.3 ± 3.3 25.9 ± 3.1 
 Cigarette smoking, % never smokers 45.9 52.7 32.1 46.6a 
 High physical activity, % 35.7 34.6 20.3 26.7 
 Low education level, % 44.4 33.1 30.0 26.1 
a

Significantly different from cases: P < 0.05 (Fisher’s exact test for categorical variables, t test for age, and body mass index, and Wilcoxon’s rank test for dietary variables).

b

n.a., not applicable.

Table 2

Association between alcohol consumption and risk of adenomatous colorectal polyps

Alcohol consumption (drinks/week)aP for trend
<11–10≥10
All N (cases/controls) 122/163 139/153 172/120  
 Gender and age adjusted OR (95% CI) 1 (REF)b 1.11 (0.78–1.58) 1.54 (1.06–2.24) 0.02 
 Multivariate OR (95% CI)c 1 (REF) 1.00 (0.68–1.46) 1.38 (0.93–2.07) 0.07 
      
Women N (cases/controls) 76/135 69/99 51/41  
 Age adjusted OR (95% CI) 1 (REF) 1.17 (0.75–1.82) 2.19 (1.30–3.68) 0.003 
 Multivariate OR (95% CI)c 1 (REF) 0.99 (0.61–1.60) 1.81 (1.02–3.21) 0.04 
      
Men N (cases/controls) 46/28 70/54 121/79  
 Age adjusted OR (95% CI) 1 (REF) 0.93 (0.50–1.73) 1.07 (0.60–1.91) 0.63 
 Multivariate OR (95% CI)c 1 (REF) 0.96 (0.50–1.85) 1.12 (0.61–2.05) 0.57 
Alcohol consumption (drinks/week)aP for trend
<11–10≥10
All N (cases/controls) 122/163 139/153 172/120  
 Gender and age adjusted OR (95% CI) 1 (REF)b 1.11 (0.78–1.58) 1.54 (1.06–2.24) 0.02 
 Multivariate OR (95% CI)c 1 (REF) 1.00 (0.68–1.46) 1.38 (0.93–2.07) 0.07 
      
Women N (cases/controls) 76/135 69/99 51/41  
 Age adjusted OR (95% CI) 1 (REF) 1.17 (0.75–1.82) 2.19 (1.30–3.68) 0.003 
 Multivariate OR (95% CI)c 1 (REF) 0.99 (0.61–1.60) 1.81 (1.02–3.21) 0.04 
      
Men N (cases/controls) 46/28 70/54 121/79  
 Age adjusted OR (95% CI) 1 (REF) 0.93 (0.50–1.73) 1.07 (0.60–1.91) 0.63 
 Multivariate OR (95% CI)c 1 (REF) 0.96 (0.50–1.85) 1.12 (0.61–2.05) 0.57 
Alcohol consumption (drinks/week)a
<11–20≥21
Men N 47/29 130/104 60/28  
 Age adjusted OR 1 (REF) 0.86 (0.49–1.51) 1.67 (0.85–3.30) 0.04 
 Multivariate ORc 1 (REF) 0.90 (0.50–1.63) 1.77 (0.86–3.64) 0.04 
Alcohol consumption (drinks/week)a
<11–20≥21
Men N 47/29 130/104 60/28  
 Age adjusted OR 1 (REF) 0.86 (0.49–1.51) 1.67 (0.85–3.30) 0.04 
 Multivariate ORc 1 (REF) 0.90 (0.50–1.63) 1.77 (0.86–3.64) 0.04 
a

One alcoholic consumption contains ∼10 g of ethanol.

b

REF, reference group.

c

Adjusted for sex, age, and indication for endoscopy (complaints related, screening, other/unknown).

Table 3

Risk of adenomatous colorectal polyps as a result of alcohol consumption and ADH3 genotypea

Alcohol consumption (drinks/week)b
<11–10≥10
All ADH3*1/*2, *2/*2 N (cases/controls) 72/96 99/105 107/79 
  OR (95% CI) 1 (REF)c 0.97 (0.60–1.56) 1.15 (0.70–1.90) 
 ADH3*1/*1 N (cases/controls) 50/67 40/48 65/41 
  OR (95% CI) 0.94 (0.53–1.64) 0.99 (0.54–1.79) 1.76 (1.00–3.11) 
Women ADH3*1/*2, *2/*2 N (cases/controls) 41/77 51/68 35/29 
  OR (95% CI) 1 (REF) 0.94 (0.51–1.74) 1.38 (0.66–2.87) 
 ADH3*1/*1 N (cases/controls) 35/58 18/31 16/12 
  OR (95% CI) 0.90 (0.46–1.73) 0.92 (0.42–2.05) 2.61 (1.05–6.50) 
Men ADH3*1/*2, *2/*2 N (cases/controls) 31/19 48/37 72/50 
  OR (95% CI) 1 (REF) 1.03 (0.47–2.27) 1.05 (0.50–2.20) 
 ADH3*1/*1 N (cases/controls) 15/9 22/17 49/29 
  OR (95% CI) 1.31 (0.41–4.16) 1.06 (0.42–2.69) 1.50 (0.67–3.34) 
Alcohol consumption (drinks/week)b
<11–10≥10
All ADH3*1/*2, *2/*2 N (cases/controls) 72/96 99/105 107/79 
  OR (95% CI) 1 (REF)c 0.97 (0.60–1.56) 1.15 (0.70–1.90) 
 ADH3*1/*1 N (cases/controls) 50/67 40/48 65/41 
  OR (95% CI) 0.94 (0.53–1.64) 0.99 (0.54–1.79) 1.76 (1.00–3.11) 
Women ADH3*1/*2, *2/*2 N (cases/controls) 41/77 51/68 35/29 
  OR (95% CI) 1 (REF) 0.94 (0.51–1.74) 1.38 (0.66–2.87) 
 ADH3*1/*1 N (cases/controls) 35/58 18/31 16/12 
  OR (95% CI) 0.90 (0.46–1.73) 0.92 (0.42–2.05) 2.61 (1.05–6.50) 
Men ADH3*1/*2, *2/*2 N (cases/controls) 31/19 48/37 72/50 
  OR (95% CI) 1 (REF) 1.03 (0.47–2.27) 1.05 (0.50–2.20) 
 ADH3*1/*1 N (cases/controls) 15/9 22/17 49/29 
  OR (95% CI) 1.31 (0.41–4.16) 1.06 (0.42–2.69) 1.50 (0.67–3.34) 
a

Adjusted for sex, age, and indication for endoscopy (complaints related, screening, other/unknown).

b

One alcoholic consumption contains ∼10 g of ethanol.

c

REF, reference group.

We thank all the people who kindly participated in this study. We thank Elly Monster and Maria van Vugt of the Division of Human Nutrition and Epidemiology, Wageningen University, the Netherlands, for collecting blood samples and assistance with the conduct of this study. The participants were recruited with help from the endoscopy staff of the following Dutch hospitals: Ziekenhuis Gelderse Vallei (Ede/Wageningen), Rivierenland Ziekenhuis (Tiel), University Medical Centre Sint Radboud (Nijmegen), Slingeland Ziekenhuis (Doetinchem), Sint Antonius Ziekenhuis (Nieuwegein), Eemland Ziekenhuis (Amersfoort), Bosch MediCentrum (Den Bosch), and Slotervaart Ziekenhuis (Amsterdam).

1
Fearon E. R., Vogelstein B. A genetic model for colorectal tumorigenesis.
Cell
,
61
:
759
-767,  
1990
.
2
Peipins L. A., Sandler R. S. Epidemiology of colorectal adenomas.
Epidemiol. Rev.
,
16
:
273
-297,  
1994
.
3
Potter J. D. Colorectal cancer, molecules and populations.
J. Natl. Cancer Inst. (Bethesda)
,
91
:
916
-932,  
1999
.
4
Seitz H. K., Pöschl G., Simanowski U. A. Alcohol and cancer.
Recent Dev. Alcohol.
,
14
:
67
-95,  
1998
.
5
International programme on chemical safety. .
Acetaldehyde. Environmental Health Criteria, Vol. 167
, World Health Organization Geneva, Switzerland  
1995
.
6
Seitz H. K., Oneta C. M. Gastrointestinal alcohol dehydrogenase.
Nutr. Rev.
,
56
:
52
-60,  
1998
.
7
Yin S. J., Liao C. S., Lee Y. C., Wu C. W., Jao S. W. Genetic polymorphism and activities of human colon alcohol and aldehyde dehydrogenases: no gender and age differences.
Alcohol. Clin. Exp. Res.
,
18
:
1256
-1260,  
1994
.
8
Estonius M., Svensson S., Höög J. O. Alcohol dehydrogenase in human tissues: localisation of transcripts coding for five classes of the enzyme.
FEBS Lett.
,
397
:
338
-342,  
1996
.
9
Whitfield J. B., Nightingale B. N., Bucholz K. K., Madden P. A. F., Heath A. C., Martin N. G. ADH genotypes and alcohol use and dependence in Europeans.
Alcohol. Clin. Exp. Res.
,
22
:
1463
-1469,  
1998
.
10
Coutelle C., Ward P. J., Quattrocchi P., Fleury B., the French Group for Research on Alcohol and Liver. Population distribution of alcohol dehydrogenase class I in France: comparison with other populations, and distribution with respect to gender and age.
Alcohol Alcohol.
,
33
:
173
-183,  
1998
.
11
Borràs E., Coutelle C., Rosell A., Fernández-Muixi F., Broch M., Crosas B., Hjelmqvist L., Lorenzo A., Gutiérrez C., Santos M., Szczepanek M., Heilig M., Quattrocchi P., Farrés J., Vidal F., Richart C., Mach T., Bogdal J., Jörnvall H., Seitz H. K., Couzigou P., Parés X. Genetic polymorphism of alcohol dehydrogenase in Europeans: the ADH2*2 allele decreases the risk for alcoholism and is associated with ADH3*1.
Hepatology
,
31
:
984
-989,  
2000
.
12
Goedde H. W., Agarwal D. P., Fritze G., Meier-Tackmann D., Singh S., Beckmann G., Bhatia K., Chen L. Z., Fang B., Lisker R., Paik Y. K., Rothhammer F., Saha N., Segal B., Srivastava L. M., Czeizel A. Distribution of ADH2 and ALDH2 genotypes in different populations.
Hum. Genet.
,
88
:
344
-346,  
1992
.
13
Suzuki K., Uchida A., Mizoi Y., Fukunaga T. A study on ADH2 and ALDH2 genotyping by PCR-RFLP and SSCP analyses with description of allele and genotype frequencies in Japanese, Finn and Lapp populations.
Alcohol Alcohol.
,
29 (Suppl. 1)
:
21
-27,  
1994
.
14
Harty L. C., Caporaso N. E., Hayes R. B., Winn D. M., Bravo-Otero E., Blot W. J., Kleinman D. V., Brown L. M., Armenian H. K., Fraumeni J. F., Shields P. G. Alcohol dehydrogenase 3 genotype and risk of oral cavity and pharyngeal cancers.
J. Natl. Cancer Inst. (Bethesda)
,
89
:
1698
-1705,  
1997
.
15
Coutelle C., Ward P. J., Fleury B., Quattrocchi P., Chambrin H., Iron A., Couzigou P., Cassaigne A. Laryngeal and oropharyngeal cancer, and alcohol dehydrogenase 3 and glutathione S-transferase M1 polymorphisms.
Hum. Genet.
,
99
:
319
-325,  
1997
.
16
Freudenheim J. L., Ambrosone C. B., Moysich K. B., Vena J. E., Graham S., Marshall J. R., Muti P., Laughlin R., Nemoto T., Harty L. C., Crits G. A., Chan A. W. K., Shields P. G. Alcohol dehydrogenase 3 genotype modification of the association of alcohol consumption with breast cancer risk.
Cancer Causes Control
,
10
:
369
-377,  
1999
.
17
Bouchardy C., Coutelle C., Ward P. J., Dayer P., Benhamou S. Re: Alcohol dehydrogenase 3 genotype and risk of oral cavity and pharyngeal cancers.
J. Natl. Cancer Inst. (Bethesda)
,
90
:
937
-939,  
1998
.
18
Sturgis E. M., Dahlstrom K. R., Guan Y., Eicher S. A., Strom S. S., Spitz M. R., Wei Q. Alcohol dehydrogenase 3 genotype is not associated with risk of squamous cell carcinoma of the oral cavity and pharynx.
Cancer Epidemiol. Biomark. Prev.
,
10
:
273
-275,  
2001
.
19
Hines L. M., Hankinson S. E., Smith-Warner S. A., Spiegelman D., Kelsey K. T., Colditz G. A., Willett W. C., Hunter D. J. A prospective study of the effect of alcohol consumption and ADH3 genotype on plasma steroid hormone levels and breast cancer risk.
Cancer Epidemiol. Biomark. Prev.
,
9
:
1099
-1105,  
2000
.
20
Olshan A. F., Weissler M. C., Watson M. A., Bell D. A. Risk of head and neck cancer and the alcohol dehydrogenase 3 genotype.
Carcinogenesis (Lond.)
,
22
:
57
-61,  
2001
.
21
Salaspuro M. Microbial metabolism of ethanol and acetaldehyde and clinical consequences.
Addict. Biol.
,
2
:
35
-46,  
1997
.
22
Voskuil D. W., Kampman E., Van Geloof W., Grubben M., Kok F., Van Muijen G., Nagengast F., Vasen H., Van’t Veer P. No major difference in K-ras and p53 abnormalities in sporadic and hereditary nonpolyposis colorectal adenomas.
Dig. Dis. Sci.
,
45
:
2187
-2194,  
2000
.
23
Ocké M. C., Bueno de Mesquita H. B., Goddijn H. E., Jansen A., Pols M. A., Van Staveren W. A., Kromhout D. The Dutch EPIC food frequency questionnaire. I. Description of the questionnaire, and relative validity and reproducibility for food groups.
Int. J. Epidemiol.
,
26 (Suppl. 1)
:
S37
-S48,  
1997
.
24
Willett W. C., Howe G. R., Kushi L. H. Adjustment for total energy intake in epidemiologic studies.
Am. J. Clin. Nutr.
,
65 (Suppl.)
:
1220S
-1228S,  
1997
.
25
Groppi A., Begueret J., Iron A. Improved methods for genotype determination of human alcohol dehydrogenase (ADH) at ADH 2 and ADH 3 loci by using polymerase chain reaction-directed mutagenesis.
Clin. Chem.
,
36
:
1765
-1768,  
1990
.
26
Dales L. G., Ury H. K. An improper use of statistical significance testing in studying covariables.
Int. J. Epidemiol.
,
7
:
373
-375,  
1978
.
27
Baraona E., Abittan C. S., Dohmen K., Moretti M., Pozzato G., Chayes Z. W., Schaefer C., Lieber C. S. Gender differences in pharmacokinetics of alcohol.
Alcohol. Clin. Exp. Res.
,
25
:
502
-507,  
2001
.
28
Boekema P. J., Van Dam van Isselt E. F., Bots M. L., Smout A. J. P. M. Functional bowel symptoms in a general Dutch population and associations with common stimulants.
Neth. J. Med.
,
59
:
23
-30,  
2001
.
29
Kikendall J. W., Bowen P. E., Burgess M. B., Magnetti C., Woodward J., Langenberg P. Cigarettes and alcohol as independent risk factors for colonic adenomas.
Gastroenterology
,
97
:
660
-664,  
1989
.
30
Cope G. F., Wyatt J. I., Pinder I. F., Lee P. N., Heatley R. V., Kelleher J. Alcohol consumption in patients with colorectal adenomatous polyps.
Gut
,
32
:
70
-72,  
1991
.
31
Martínez M. E., McPherson R. S., Annegers J. F., Levin B. Cigarette smoking and alcohol consumption as risk factors for colorectal adenomatous polyps.
J. Natl. Cancer Inst. (Bethesda)
,
87
:
274
-279,  
1995
.
32
Feunekes G. I. J., Van’t Veer P., Van Staveren W. A., Kok F. J. Alcohol intake assessment: the sober facts.
Am. J. Epidemiol.
,
150
:
105
-112,  
1999
.
33
Hajema K. J., Knibbe R. A., Drop M. J. Changes in alcohol consumption in a general population in the Netherlands: a 9-year follow-up study.
Addiction
,
92
:
49
-60,  
1997
.
34
Doll R., Peto R., Hall E., Wheatley K., Gray R. Mortality in relation to consumption of alcohol: 13 years’ observations on male British doctors.
BMJ
,
309
:
911
-918,  
1994
.
35
Kono S., Ikeda N., Yanai F., Shinchi K., Imanishi K. Alcoholic beverages and adenomatous polyps of the sigmoid colon: a study of male self-defence officials in Japan.
Int. J. Epidemiol.
,
19
:
848
-852,  
1990
.
36
Kune G. A., Kune S., Read A., MacGowan K., Penfold C., Watson L. F. Colorectal polyps, diet, alcohol, and family history of colorectal cancer: a case-control study.
Nutr. Cancer
,
16
:
25
-30,  
1991
.
37
Giovannucci E., Stampfer M. J., Colditz G. A., Rimm E. B., Trichopoulos D., Rosner B. A., Speizer F. E., Willett W. C. Folate, methionine, and alcohol intake and risk of colorectal adenoma.
J. Natl. Cancer Inst. (Bethesda)
,
85
:
875
-884,  
1993
.
38
Sandler R. S., Lyles C. M., McAuliffe C., Woosley J. T., Kupper L. L. Cigarette smoking, alcohol, and the risk of colorectal adenomas.
Gastroenterology
,
104
:
1445
-1451,  
1993
.
39
Boutron-Ruault M. C., Senesse P., Faivre J., Couillault C., Belghiti C. Folate and alcohol intakes: related or independent roles in the adenoma-carcinoma sequence?.
Nutr. Cancer
,
26
:
337
-346,  
1996
.
40
Baron J. A., Sandler R. S., Haile R. W., Mandel J. S., Mott L. A., Greenberg E. R. Folate intake, alcohol consumption, cigarette smoking, and risk of colorectal adenomas.
J. Natl. Cancer Inst. (Bethesda)
,
90
:
57
-62,  
1998
.
41
Hupkens C. L., Knibbe R. A., Drop M. J. Alcohol consumption in the European community: uniformity and diversity in drinking patterns.
Addiction
,
88
:
1391
-1404,  
1993
.
42
Frezza M., Di Padova C., Pozzato G., Terpin M., Baraona E., Lieber C. S. High blood alcohol levels in women. The role of decreased gastric alcohol dehydrogenase activity and first-pass metabolism.
N. Engl. J. Med.
,
322
:
95
-99,  
1990
.
43
Boutron M. C., Faivre J., Dop M. C., Quipourt V., Senesse P. Tobacco, alcohol, and colorectal tumors: a multistep process.
Am. J. Epidemiol.
,
141
:
1038
-1046,  
1995
.
44
World Cancer Res. Fund/American Institute for Cancer Res. .
Food, Nutrition and the Prevention of Cancer: a global perspective
,
234
American Institute for Cancer Res. Washington, DC  
1997
.
45
Seitz H. K., Simanowski U. A., Osswald B. R. Epidemiology and pathophysiology of ethanol-associated gastrointestinal cancer.
Pharmacogenetics
,
2
:
278
-287,  
1992
.
46
Foppa I., Spiegelman D. Power and sample size calculations for case-control studies of gene-environment interactions with a polytomous exposure variable.
Am. J. Epidemiol.
,
146
:
596
-604,  
1997
.
47
Nosova T., Jousimies-Somer H., Kaihovaara P., Jokelainen K., Heine R., Salaspuro M. Characteristics of aldehyde dehydrogenases of certain aerobic bacteria representing human colonic flora.
Alcohol. Clin. Exp. Res.
,
21
:
489
-494,  
1997
.
48
Lieber C. S. Ethanol metabolism, cirrhosis and alcoholism.
Clin. Chim. Acta
,
257
:
59
-84,  
1997
.