Alcohol intake, ADH1B and ADH1C genotypes, and the risk of colorectal cancer by sex and subsite in the Netherlands Cohort Study

Alcohol intake, ADH1B and ADH1C genotypes, and the risk of colorectal cancer by sex and subsite... Abstract The alcohol–colorectal cancer (CRC) association may differ by sex and ADH1B and ADH1C genotypes. ADH enzymes oxidize ethanol to acetaldehyde, both of which are human carcinogens. The Netherlands Cohort Study includes 120 852 participants, aged 55–69 years at baseline (1986), and has 20.3 years follow-up (case-cohort: nsubcohort = 4774; ncases = 4597). The baseline questionnaire included questions on alcohol intake at baseline and 5 years before. Using toenail DNA, available for ~75% of the cohort, we successfully genotyped six ADH1B and six ADH1C SNPs (nsubcohort = 3897; ncases = 3558). Sex- and subsite-specific Cox hazard ratios and 95% confidence intervals for CRC were estimated comparing alcohol categories, genotypes within drinkers and alcohol categories within genotype strata. We used a dominant genetic model and adjusted for multiple testing. Alcohol intake increased CRC risk in both sexes, though in women only in the (proximal) colon when in excess of 30 g/day. In male drinkers, ADH1B rs4147536 increased (distal) colon cancer risk. In female drinkers, ADH1C rs283415 increased proximal colon cancer risk. ADH1B rs3811802 and ADH1C rs4147542 decreased CRC risk in heavy (>30 g/day) and stable drinkers (compared to 5 years before baseline), respectively. Rs3811802 and rs4147542 significantly modified the alcohol-colon cancer association in women (Pfor interaction = 0.004 and 0.02, respectively). A difference in associations between genotype strata was generally clearer in men than women. In conclusion, men showed increased CRC risks across subsites and alcohol intake levels, while only colon cancer risk was increased in women at heavy intake levels. ADH1B rs3811802 and ADH1C rs4147542 significantly modified the alcohol–colon cancer association in women. Introduction Alcohol intake is a known risk factor for colorectal cancer (CRC), both the colon and rectum, and various other cancers including cancers of the oral cavity, pharynx, larynx, esophagus, liver and female breast (1,2). Cancer risk has been shown to increase as the volume of alcohol consumed increases (3). Sex differences may exist, with CRC risk being more strongly affected by alcohol intake in men than women (4) and a dose–response relation being less apparent in women than men (5). While this may be due to a restriction of range effect because women consume less alcohol than men, part of these differences may also simply be due to the scarcity of studies on the alcohol–CRC association in women (4). The alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) gene families encode for enzymes responsible for the breakdown of alcohol in the body. Ethanol is oxidized by ADH to acetaldehyde, which is in turn oxidized by ALDH to acetate (6,7). The formation of acetaldehyde starts in the mouth and continues along the digestive tract, with the main production of acetaldehyde occurring in the liver and the gut (1). Both ethanol and acetaldehyde have been classified as human carcinogens by the International Agency for Research on Cancer (IARC) (6). Genetic variants [i.e. single nucleotide polymorphisms (SNPs)] in ADH and ALDH genes have been shown to affect enzyme activities resulting in a slower or faster conversion of ethanol to acetaldehyde and acetaldehyde to acetate (7,8). The ALDH2 rs671 gene variant strongly determines the slow rate at which most of the ethanol-derived acetaldehyde is oxidized, resulting in acetaldehyde accumulation when carrying the minor allele, especially when also carrying ADH genotypes associated with fast ethanol to acetaldehyde conversion, e.g. for ADH1B rs1229984. This results in the facial flushing syndrome known in Asian populations (7). In Caucasians, however, the rs671 allele causing slow oxidation of acetaldehyde is largely absent (9). ADH1B and ADH1C, on the other hand, have functional polymorphisms in Caucasian populations (7), but most previous epidemiological studies on SNPs in these genes and the risk of CRC had low sample sizes (10,11). In the absence of ALDH alleles causing slow oxidation of acetaldehyde, increased cancer risks may be explained by fast conversion of ethanol to acetaldehyde in the liver, resulting in higher systemic levels of acetaldehyde, potentially reaching the colorectum. Alternatively, slow conversion of ethanol to acetaldehyde in the liver results in ethanol circulating the blood for longer periods of time. Circulating ethanol may expose the colorectum to locally formed levels of acetaldehyde, depending on ADH activity in the colonic mucosa and conversion by intestinal bacteria (12–15). Therefore, in the absence of data on how much ethanol or acetaldehyde accumulates in the colorectum, an association between alcohol intake and CRC risk in both slow and fast ethanol metabolizers is conceivable. The objective of this study was twofold in that we first aimed to investigate the alcohol–CRC association by sex and subsite, specifically addressing the limited evidence in women, and secondly we aimed to investigate whether this association was modified by genetic variants in ADH1B and ADH1C. ADH1B and ADH1C variants were also studied in relation to CRC risk by sex and subsite in drinkers, where an effect may be expected. The study setting was a prospective cohort including 120 852 participants, who were followed up for 20.3 years (16). A former study in this cohort after 13.3 years follow-up found no evidence of an influence of alcohol on CRC risk, overall and by subsite, though the number of cases among the heavy drinkers was rather limited (i.e. ≥30 g of alcohol per day), hampering sex-specific analyses (17). Another study in this cohort, investigating the alcohol–CRC association by ADH1C rs698 genotype strata after 7.3 years follow-up found no apparent evidence of effect modification, but the power was limited there as well (18). Therefore, we reinvestigated the alcohol–CRC association by sex and subsite and possible effect modification by rs698 and other ADH1B and ADH1C tagSNPs after 20.3 years of follow-up. Recently, Dutch recommendations with respect to alcohol intake for cancer prevention have been modified: from a maximum of two drinks daily for men and one drink daily for women to no alcohol intake for both sexes (2,19), and if consuming alcohol, no more than one drink daily (19). This revised recommendation may be particularly important for specific subgroups with an unfavorable genetic background. Materials and methods Design and study population The Netherlands Cohort Study (NLCS) is a prospective cohort study which was initiated in 1986 and consists of 120 852 men and women who were aged 55–69 years old at baseline (16). Study participants completed a baseline self-administered questionnaire on dietary habits, lifestyle factors and other risk factors for cancer. For efficiency reasons, the NLCS uses a case-cohort approach in which cases are enumerated from the entire cohort and the person-time at risk is estimated from a subcohort (20). This subcohort, consisting of 5000 participants, was randomly selected immediately after baseline, independent of any exposure. After exclusion of participants who reported a history of cancer (other than skin cancer) at baseline, 4774 subcohort members were left. Follow-up of vital status and migration for these participants was done through the Central Bureau of Genealogy and the municipal population registries (>99.9% completeness). Follow-up for incident cancer cases was performed by record linkage to the Dutch Cancer Registry and PALGA (the Netherlands pathology database) (21,22) (>96% completeness) (23). After 20.3 years follow-up, 4597 CRC cases (ICD-O-3 codes C18-C20) were identified from the original cohort. The NLCS was approved by the institutional review boards of the Netherlands Organisation for Applied Scientific Research TNO (Zeist) and Maastricht University (Maastricht). Exposure assessment The baseline questionnaire included a 150-item semi-quantitative food frequency questionnaire (FFQ) containing questions on diet and alcohol intake. In addition to the questionnaire, participants in the NLCS were asked to return toenail clippings. Roughly 90 000 (~75%) of NLCS participants provided toenail clippings which were used as a source of DNA for genotyping (24,25). Using 20.3 years follow-up in the NLCS, DNA samples were available for 3558 CRC cases and 3897 subcohort members. Alcohol intake and covariates Alcohol intake during the year preceding the start of the study was measured by questions on six different types of alcohol: beer; red wine; white wine; sherry and other fortified wines; liquor types containing on average 16% alcohol; (Dutch) gin, brandy and whiskey. Participants were asked about the number of glasses they consumed during each drinking session and their usual frequency of alcohol drinking. Additionally, for the categories ‘beer’ and ‘other alcoholic beverages’, participants were asked to recall if they drank more, less, or the same amount of alcohol 5 years before baseline. The total amount of daily alcohol intake (g/day) was calculated using the information about how often the participants drank alcohol, the number of glasses they consumed during each drinking session and the types of alcohol they drank (i.e. their alcoholic content). We defined two alcohol categories: light-moderate alcohol intake as drinking >0 to <30 g of alcohol per day (>0 to <3 glasses of alcohol per day) and heavy alcohol intake as drinking 30 or more grams of alcohol per day (≥3 glasses of alcohol per day). Information on other covariates that were considered potential confounders on the basis of previous research was also available from the baseline questionnaire. The FFQ was validated against a 9-day diet record (26) and was tested for reproducibility (27). The adjusted Spearman correlation coefficient between mean daily alcohol intake assessed by the FFQ and that estimated from the 9-day diet record was 0.89 for all participants and 0.85 for users of alcoholic beverages. The absolute amount of alcohol reported in the FFQ by users of alcoholic beverages was, on average, 86% of that reported in the 9-day diet record (26). Selection and genotyping of tagging SNPs Tagging SNPs (tagSNPs) within ADH1B and ADH1C (including 5 kb up- and downstream) were selected as to potentially cover all of the genetic variation in these genes with a minor allele frequency of 5% or higher. In total, 13 tagSNPs were identified using the HapMap CEU (Utah Residents with Northern and Western European Ancestry) population, an r2 threshold of 0.8 and aggressive tagging. Seven of these (i.e. rs1159918, rs2075633, rs1693439, rs9307239, rs4147536, rs3811802, rs17033) represented 84% of the genetic variation in ADH1B and six (i.e. rs698, rs1662033, rs3114046, rs4147542, rs283415, rs4699741) represented 96% of the genetic variation in ADH1C (28). SNPs were genotyped using 50 ng of toenail DNA per participant, which was carried out using the iPLEXTM assay for the MassARRAY® system (Agena Bioscience GmbH, Hamburg, Germany). Samples had a mean call rate of 97.1% (as based on the 13 SNPs in ADH1B and ADH1C studied here and 10 other SNPs that were included in the assay). SNP call rates were 94% or higher, except for rs4147542, which had a SNP call rate of 87%. Two SNPs, ADH1C rs4699741 and ADH1C rs9307239, violated Hardy–Weinberg equilibrium. When using a significance threshold of 0.05, one in twenty SNPs may be expected to show a violation on the basis of chance alone. Although two SNPs exceed this expectation by chance and we cannot check conditions needed for Hardy–Weinberg, e.g. random mating, all SNPs were genotyped using a single assay, which makes it unlikely that these violations represent genotyping errors. Therefore, we conservatively refrained from excluding these SNPs from the analysis. Genotyping for ADH1B rs17033 was unsuccessful (i.e. only the T allele was amplified) and therefore not included in our analyses as originally intended. Consequently, we used six ADH1B tagSNPs covering 76% of the genetic variation in ADH1B and six ADH1C tagSNPs covering 96% of the genetic variation in ADH1C. Statistical analysis Statistical analyses were carried out using Cox regression to calculate hazard ratios (HR) and corresponding 95% confidence intervals (95% CI) for CRC by sex and subsite. Standard errors were estimated using the Huber–White sandwich estimator to account for additional variance introduced by sampling the subcohort from the entire cohort. ADH1B and ADH1C genotypes were defined according to a dominant model for reasons of power. Categories of total alcohol intake (0.1–29 and ≥30 g/day) were compared relative to abstaining (0 g/day). Trends were evaluated by including categorical variables as continuous variables in the Cox regression model. The proportional hazards assumption was tested using the scaled Schoenfeld residuals and by visually inspecting the −log–log transformed hazard curves. Multiplicative interactions were tested using the Wald test. All tests (two-tailed) were performed using Stata (version 14) and differences were regarded as statistically significant at P < 0.05. Marginal effects of alcohol intake on CRC Multivariable-adjusted models were used to study the alcohol–CRC associations. The covariates included were either a priori-selected risk factors based on the literature or variables that changed the HRs by at least 10% (using a backwards stepwise procedure). This resulted in the following confounder set: age (years), BMI (kg/m2), smoking (never/ex/current), family history of CRC (yes/no), meat intake (g/day), processed meat intake (g/day), folate intake (µg/day) and physical activity based on baseline non-occupational physical activity (min/day) (29). Participants with incomplete or inconsistent questionnaires and missing information on alcohol intake and/or the predefined confounding factors were excluded from the analysis, leaving 4125 subcohort members and 3996 CRC cases (see Figure 1). Figure 1. View largeDownload slide Flow diagram of available subcohort members and colorectal cancer cases, Netherlands Cohort Study, 1986–2006. FU, follow-up. AAnalysis 1 is on the marginal effects of alcohol intake on CRC. BAnalysis 2 is on the associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers. CAnalysis 3 is on effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs. Figure 1. View largeDownload slide Flow diagram of available subcohort members and colorectal cancer cases, Netherlands Cohort Study, 1986–2006. FU, follow-up. AAnalysis 1 is on the marginal effects of alcohol intake on CRC. BAnalysis 2 is on the associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers. CAnalysis 3 is on effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs. Associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers We studied individual SNPs in relation to CRC risk in drinkers. Although small amounts of ethanol are produced endogenously, especially in the gastrointestinal tract (1), an effect of SNPs in alcohol-metabolizing genes may be expected in drinkers foremost as this is the group where the substrate (alcohol) is available. We conservatively refrained from adjusting for factors other than age because it is unlikely that ADH1B and ADH1C genotypes are influenced by CRC risk factors in lifestyle and diet. Participants were excluded from the analysis if no toenail DNA sample was available, the sample call rate was less than 90%, the baseline questionnaire was incomplete or inconsistent, or information on alcohol intake was missing (see Figure 1).This resulted in 2526 subcohort members and 2491 CRC cases. Effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs Multivariable-adjusted models were used to study effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs, using the confounder set as described for the alcohol–CRC analyses. After excluding participants without available DNA samples, with a sample call rate of less than 90%, with incomplete or inconsistent questionnaires and without complete information on alcohol intake and/or the predefined confounding factors, 3150 subcohort members and 2985 CRC cases were left for analysis (see Figure 1). Multiple testing Because multiple tests were conducted within each gene, we applied the false discovery rate (FDR) control method of Benjamini and Hochberg (30,31) to address the issue of multiple testing. To accomplish this, P-values calculated from our analyses were ranked in ascending order. Gene- and endpoint-specific Benjamini adjusted P-values were calculated by dividing the P-value rank order by the total number of P-values and then multiplying this number by the FDR [i.e. the recommended 20% (32)]. If the original P-value was less than 0.05 and fell below the adjusted P-value, it was considered significant. Sensitivity analyses Drinking patterns over a longer duration of time may influence CRC risk differently or more profoundly than when evaluated on a single time point. For instance, stronger alcohol–CRC associations may be expected in those with a relatively constant long-term exposure to alcohol. As the NLCS has data available on alcohol intake 5 years before baseline, we conducted sensitivity analyses using these data. This included restricting the analyses on alcohol–CRC associations and effect modification to participants who reported to have had the same alcohol intake 5 years before baseline, which included abstainers on both occasions (i.e. the stable subgroup). For the SNP-CRC analyses in drinkers, this included restriction to those drinking equal amounts of alcohol 5 years before baseline (i.e. the stable drinkers). We also evaluated whether there may be a threshold level of alcohol intake at which individual SNPs start to influence CRC risk by stratifying SNP–CRC associations on alcohol intake level (light-moderate and heavy). Furthermore, because changes in reported alcohol intake may indicate underlying reasons such as health issues or exposure misclassification, possibly due to underreporting, we repeated the alcohol–CRC analyses restricting once to baseline drinkers who reported drinking less alcohol 5 years before baseline and once to baseline drinkers who reported drinking more alcohol 5 years before baseline. Finally, we checked for the risk of protopathic bias by excluding the first 2 years of follow-up (with no essential changes in results), as this may be especially likely when investigating alcohol intake in relation to cancer risk. Results Baseline characteristics Table 1 shows the baseline characteristics of subcohort members and CRC cases. As regards alcohol intake, men were less often abstainers as compared to women and men were more likely to consume higher levels of alcohol than women, especially male CRC cases. Table 1. Distribution of potential confounders and alcohol intake among subcohort members and CRC cases in the NLCS, 1986–2006   Male subcohort  Male CRC cases  Female subcohort  Female CRC cases    N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  Age (years)    61.4 (4.2)    61.8 (4.2)    61.5 (4.3)    62.1 (4.1)  Alcohol intake (g/day)   0  329 (14.5)    303 (12.1)    731 (32.7)    599 (31.7)     0.1–4  479 (21.1)    549 (22.0)    806 (36.0)    692 (36.6)     5–14  621 (27.3)    630 (25.2)    417 (18.6)    348 (18.4)     15–29  505 (22.2)    593 (23.7)    207 (9.3)    168 (8.9)     ≥30  339 (14.9)    426 (17.0)    77 (3.4)    84 (4.4)    BMI (kg/m2)    25.0 (2.6)    25.3 (2.7)    25.1 (3.6)    25.1 (3.5)  Smoking   Never  300 (12.9)    313 (12.3)    1431 (58.9)    1190 (58.7)     Ex  1175 (50.4)    1463 (57.3)    491 (20.2)    438 (21.6)     Current  856 (36.7)    777 (30.4)    509 (20.9)    398 (19.6)    Family history of CRC   Yes  118 (5.1)    219 (8.6)    134 (5.5)    189 (9.3)     No  2213 (94.9)    2336 (91.4)    2298 (94.5)    1840 (90.7)    Meat intake (g/day)    104.9 (44.1)    105.3 (43.2)    92.2 (41.1)    90.6 (40.9)  Processed meat intake (g/day)    15.6 (16.9)    16.3 (16.8)    10.3 (11.9)    10.2 (11.3)  Folate intake (µg/day)    222 (77)    219 (72)    195 (71)    194 (71)  Non-occupational physical activity (min/day)   ≤30  447 (19.5)    431 (17.1)    639 (26.8)    596 (29.9)     >30–60  710 (30.9)    776 (30.8)    737 (30.9)    598 (30.0)     >60–90  419 (18.3)    497 (19.7)    518 (21.7)    451 (22.6)     >90  719 (31.3)    815 (32.4)    490 (20.6)    351 (17.6)      Male subcohort  Male CRC cases  Female subcohort  Female CRC cases    N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  Age (years)    61.4 (4.2)    61.8 (4.2)    61.5 (4.3)    62.1 (4.1)  Alcohol intake (g/day)   0  329 (14.5)    303 (12.1)    731 (32.7)    599 (31.7)     0.1–4  479 (21.1)    549 (22.0)    806 (36.0)    692 (36.6)     5–14  621 (27.3)    630 (25.2)    417 (18.6)    348 (18.4)     15–29  505 (22.2)    593 (23.7)    207 (9.3)    168 (8.9)     ≥30  339 (14.9)    426 (17.0)    77 (3.4)    84 (4.4)    BMI (kg/m2)    25.0 (2.6)    25.3 (2.7)    25.1 (3.6)    25.1 (3.5)  Smoking   Never  300 (12.9)    313 (12.3)    1431 (58.9)    1190 (58.7)     Ex  1175 (50.4)    1463 (57.3)    491 (20.2)    438 (21.6)     Current  856 (36.7)    777 (30.4)    509 (20.9)    398 (19.6)    Family history of CRC   Yes  118 (5.1)    219 (8.6)    134 (5.5)    189 (9.3)     No  2213 (94.9)    2336 (91.4)    2298 (94.5)    1840 (90.7)    Meat intake (g/day)    104.9 (44.1)    105.3 (43.2)    92.2 (41.1)    90.6 (40.9)  Processed meat intake (g/day)    15.6 (16.9)    16.3 (16.8)    10.3 (11.9)    10.2 (11.3)  Folate intake (µg/day)    222 (77)    219 (72)    195 (71)    194 (71)  Non-occupational physical activity (min/day)   ≤30  447 (19.5)    431 (17.1)    639 (26.8)    596 (29.9)     >30–60  710 (30.9)    776 (30.8)    737 (30.9)    598 (30.0)     >60–90  419 (18.3)    497 (19.7)    518 (21.7)    451 (22.6)     >90  719 (31.3)    815 (32.4)    490 (20.6)    351 (17.6)    View Large Marginal effects of alcohol intake on CRC Alcohol intake was positively associated with CRC risk in both men and women (Table 2). In women, however, only colon cancer risks, in particular proximal colon cancer risk, were increased, but not until alcohol intake exceeded 30 g/day [HRheavy drinkers versus abstainers (95% CI) = 1.52 (1.03–2.24), 1.70 (1.09–2.66) and 1.25 (0.71–2.20) for colon, proximal colon and distal colon cancer, respectively]. In men, colon and rectal cancer risks were both increased and there were also non-significantly increased risks in men who consumed a light-moderate amount of alcohol versus abstainers [HRlight-moderate drinkers versus abstainers (95% CI) = 1.21 (0.97–1.50), 1.23 (0.93–1.63), 1.23 (0.93–1.63) and 1.10 (0.82–1.48) for colon, proximal colon, distal colon and rectal cancer, respectively]. In addition, there was a statistically significant positive linear trend across alcohol intake categories in men, except for proximal colon cancer. When restricting to the stable subgroup, more pronounced associations were observed between alcohol intake and CRC risk in men, while statistically significant associations were no longer observed in women, but this may be explained by limited power. Remarkably, male heavy drinkers as compared to light-moderate drinkers reporting more alcohol intake 5 years before baseline had decreased CRC risks across subsites. Possibly, men who still reported to be heavy drinkers at baseline endure alcohol better than light-moderate drinkers who reported more alcohol intake 5 years before baseline. In the subanalysis in male drinkers reporting less alcohol intake 5 years before baseline, HRs were around the null for heavy drinkers as compared to light-moderate drinkers. In female drinkers who reported more and those who reported less alcohol intake 5 years before baseline, (non-significantly) increased risks of CRC were observed for heavy drinkers as compared to light-moderate drinkers, though the power was limited. Table 2. Hazard ratios and 95% confidence intervals for the association between alcohol intake and colorectal cancer risk in men and women stratified by subsite in the NLCS, 1986–2006   PT at risk  Colorectum  Colon  Proximal Colon  Distal Colon  Rectum      N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  Men  Alcohol intake   Abstainers (0 g/day)  4548  267  1  (ref)  176  1  (ref)  81  1  (ref)  88  1  (ref)  69  1  (ref)   Light-moderate (0.1–29 g/day)  23 074  1625  1.22  (1.00, 1.48)  1050  1.21  (0.97, 1.50)  485  1.23  (0.93, 1.63)  538  1.23  (0.93, 1.63)  388  1.10  (0.82, 1.48)   Heavy (≥30 g/day)  4852  394  1.40  (1.10, 1.78)  246  1.35  (1.03, 1.77)  109  1.30  (0.92, 1.83)  126  1.42  (1.01, 1.98)  110  1.45  (1.02, 2.07)   P for trend      0.007      0.03      0.14      0.04      0.03    Stable subgroupb   Abstainers (0 g/day)  3514  211  1  (ref)  139  1  (ref)  65  1  (ref)  68  1  (ref)  52  1  (ref)   Light-moderate (0.1–29 g/day)  14 413  991  1.17  (0.93, 1.46)  628  1.13  (0.88, 1.46)  291  1.11  (0.81, 1.54)  324  1.21  (0.88, 1.68)  242  1.14  (0.80, 1.62)   Heavy (≥30 g/day)  2615  241  1.58  (1.17, 2.14)  155  1.58  (1.13, 2.21)  69  1.47  (0.97, 2.23)  78  1.70  (1.11, 2.59)  66  1.73  (1.11, 2.70)   P for trend      0.003      0.008      0.08      0.02      0.02    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  3754  317  1  (ref)  207  1  (ref)  92  1  (ref)  111  1  (ref)  80  1  (ref)   Heavy (≥30 g/day)  778  42  0.56  (0.34, 0.92)  23  0.45  (0.25, 0.82)  12  0.51  (0.25,1.08)  10  0.37  (0.17,0.84)  11  0.57  (0.26,1.23)  Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2588  151  1  (ref)  104  1  (ref)  46  1  (ref)  51  1  (ref)  28  1  (ref)   Heavy (≥30 g/day)  989  67  1.12  (0.71, 1.77)  47  1.08  (0.66, 1.78)  21  1.04  (0.52,2.08)  25  1.22  (0.67,2.23)  14  1.45  (0.65,3.23)  Women  Alcohol intake   Abstainers (0 g/day)  11 373  531  1  (ref)  400  1  (ref)  232  1  (ref)  154  1  (ref)  93  1  (ref)   Light-moderate (0.1–29 g/day)  23 546  1101  1.01  (0.87, 1.17)  825  1.02  (0.87, 1.20)  481  1.02  (0.84,1.23)  324  1.06  (0.85,1.33)  199  0.99  (0.75,1.31)   Heavy (≥30 g/day)  1253  78  1.46  (1.02, 2.10)  58  1.52  (1.03, 2.24)  38  1.70  (1.09,2.66)  18  1.25  (0.71,2.20)  11  1.03  (0.51,2.07)   P for trend      0.29      0.23      0.21      0.46      1.00    Stable subgroupb   Abstainers (0 g/day)  9296  441  1  (ref)  336  1  (ref)  193  1  (ref)  132  1  (ref)  72  1  (ref)   Light-moderate (0.1–29 g/day)  13 059  632  1.01  (0.84, 1.20)  475  1.02  (0.84, 1.24)  286  1.06  (0.85,1.33)  180  1.00  (0.77,1.31)  112  1.06  (0.75,1.49)   Heavy (≥30 g/day)  701  46  1.52  (0.95, 2.42)  34  1.52  (0.91, 2.54)  21  1.59  (0.88,2.88)  11  1.33  (0.64,2.76)  6  1.12  (0.44,2.85)   P for trend      0.38      0.37      0.27      0.73      0.71    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2818  113  1  (ref)  83  1  (ref)  47  1  (ref)  30  1  (ref)  24  1  (ref)   Heavy (≥30 g/day)  127  7  1.51  (0.48, 4.79)  5  1.53  (0.43, 5.48)  4      1      1      Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2427  101  1  (ref)  77  1  (ref)  44  1  (ref)  32  1  (ref)  19  1  (ref)   Heavy (≥30 g/day)  255  16  1.72  (0.75, 3.95)  11  1.57  (0.63, 3.93)  6  2.03  (0.63, 6.55)  5  1.31  (0.41, 4.15)  3        PT at risk  Colorectum  Colon  Proximal Colon  Distal Colon  Rectum      N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  Men  Alcohol intake   Abstainers (0 g/day)  4548  267  1  (ref)  176  1  (ref)  81  1  (ref)  88  1  (ref)  69  1  (ref)   Light-moderate (0.1–29 g/day)  23 074  1625  1.22  (1.00, 1.48)  1050  1.21  (0.97, 1.50)  485  1.23  (0.93, 1.63)  538  1.23  (0.93, 1.63)  388  1.10  (0.82, 1.48)   Heavy (≥30 g/day)  4852  394  1.40  (1.10, 1.78)  246  1.35  (1.03, 1.77)  109  1.30  (0.92, 1.83)  126  1.42  (1.01, 1.98)  110  1.45  (1.02, 2.07)   P for trend      0.007      0.03      0.14      0.04      0.03    Stable subgroupb   Abstainers (0 g/day)  3514  211  1  (ref)  139  1  (ref)  65  1  (ref)  68  1  (ref)  52  1  (ref)   Light-moderate (0.1–29 g/day)  14 413  991  1.17  (0.93, 1.46)  628  1.13  (0.88, 1.46)  291  1.11  (0.81, 1.54)  324  1.21  (0.88, 1.68)  242  1.14  (0.80, 1.62)   Heavy (≥30 g/day)  2615  241  1.58  (1.17, 2.14)  155  1.58  (1.13, 2.21)  69  1.47  (0.97, 2.23)  78  1.70  (1.11, 2.59)  66  1.73  (1.11, 2.70)   P for trend      0.003      0.008      0.08      0.02      0.02    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  3754  317  1  (ref)  207  1  (ref)  92  1  (ref)  111  1  (ref)  80  1  (ref)   Heavy (≥30 g/day)  778  42  0.56  (0.34, 0.92)  23  0.45  (0.25, 0.82)  12  0.51  (0.25,1.08)  10  0.37  (0.17,0.84)  11  0.57  (0.26,1.23)  Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2588  151  1  (ref)  104  1  (ref)  46  1  (ref)  51  1  (ref)  28  1  (ref)   Heavy (≥30 g/day)  989  67  1.12  (0.71, 1.77)  47  1.08  (0.66, 1.78)  21  1.04  (0.52,2.08)  25  1.22  (0.67,2.23)  14  1.45  (0.65,3.23)  Women  Alcohol intake   Abstainers (0 g/day)  11 373  531  1  (ref)  400  1  (ref)  232  1  (ref)  154  1  (ref)  93  1  (ref)   Light-moderate (0.1–29 g/day)  23 546  1101  1.01  (0.87, 1.17)  825  1.02  (0.87, 1.20)  481  1.02  (0.84,1.23)  324  1.06  (0.85,1.33)  199  0.99  (0.75,1.31)   Heavy (≥30 g/day)  1253  78  1.46  (1.02, 2.10)  58  1.52  (1.03, 2.24)  38  1.70  (1.09,2.66)  18  1.25  (0.71,2.20)  11  1.03  (0.51,2.07)   P for trend      0.29      0.23      0.21      0.46      1.00    Stable subgroupb   Abstainers (0 g/day)  9296  441  1  (ref)  336  1  (ref)  193  1  (ref)  132  1  (ref)  72  1  (ref)   Light-moderate (0.1–29 g/day)  13 059  632  1.01  (0.84, 1.20)  475  1.02  (0.84, 1.24)  286  1.06  (0.85,1.33)  180  1.00  (0.77,1.31)  112  1.06  (0.75,1.49)   Heavy (≥30 g/day)  701  46  1.52  (0.95, 2.42)  34  1.52  (0.91, 2.54)  21  1.59  (0.88,2.88)  11  1.33  (0.64,2.76)  6  1.12  (0.44,2.85)   P for trend      0.38      0.37      0.27      0.73      0.71    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2818  113  1  (ref)  83  1  (ref)  47  1  (ref)  30  1  (ref)  24  1  (ref)   Heavy (≥30 g/day)  127  7  1.51  (0.48, 4.79)  5  1.53  (0.43, 5.48)  4      1      1      Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2427  101  1  (ref)  77  1  (ref)  44  1  (ref)  32  1  (ref)  19  1  (ref)   Heavy (≥30 g/day)  255  16  1.72  (0.75, 3.95)  11  1.57  (0.63, 3.93)  6  2.03  (0.63, 6.55)  5  1.31  (0.41, 4.15)  3      Results were not shown when less than five cases were available. CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAdjusted for age, BMI, smoking, family history of CRC, meat intake, processed meat intake, folate intake and non-occupational physical activity. bParticipants who reported drinking the same amount of alcohol intake at baseline and 5 years before baseline, including those who reported to be abstainers on both occasions. View Large Associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers Tables 3 and 4 show associations between ADH1B and ADH1C tagSNPs and CRC risk overall and by subsite in male and female drinkers, respectively, as analyzed according to a dominant model. Only FDR significant results will be mentioned below. ADH1B rs4147536 was associated with the risk of colon cancer and distal colon cancer in male drinkers [HRCA/AA versus CC (95% CI) = 1.25 (1.05–1.48) and 1.32 (1.07–1.62), respectively]. ADH1C rs283415 was associated with the risk of proximal colon cancer in female drinkers [HRTC/CC versus TT (95% CI) = 1.39 (1.08–1.80)]. Restricting these analyses to the stable drinkers revealed a statistically significant association between ADH1C rs4147542 and CRC risk in women [HRTC/CC versus TT (95% CI) = 0.73 (0.57–0.93)]. Stratifying these analyses by alcohol intake amount (i.e. light-moderate and heavy) to evaluate a potential threshold level of alcohol intake at which individual SNPs start to influence CRC risk revealed a statistically significant association between ADH1B rs3811802 and CRC risk in women who were heavy drinkers at baseline (>30 g/day) [HRAG/GG versus AA (95% CI) = 0.19 (0.07–0.50)], while no significant associations were observed in light-moderate drinkers. The results of both sensitivity analyses are presented in Supplementary Tables 1 and 2, available at Carcinogenesis Online. Table 3. Hazard ratios and 95% confidence intervals for the association between single nucleotide polymorphisms in ADH1B and ADH1C and risk of overall colorectal cancer and subtypes in male drinkers in the NLCS, 1986–2006   Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8918  638  1 (ref)  407  1 (ref)  185  1 (ref)  209  1 (ref)  155  1 (ref)    CA + AA  13 547  960  1.01 (0.87, 1.18)  632  1.05 (0.88, 1.24)  301  1.11 (0.89, 1.38)  318  1.02 (0.83, 1.26)  244  1.05 (0.83, 1.32)    P-value (adjusted P-value)      0.89 (0.17)    0.59 (0.15)    0.36 (0.08)    0.86 (0.18)    0.69 (0.15)   rs2075633  TT  11 627  799  1 (ref)  519  1 (ref)  254  1 (ref)  251  1 (ref)  200  1 (ref)    TC + CC  10 879  806  1.08 (0.93, 1.25)  524  1.08 (0.91, 1.27)  234  0.98 (0.79, 1.21)  278  1.19 (0.97, 1.46)  202  1.08 (0.86, 1.35)    P-value (adjusted P-value)      0.32 (0.07)    0.38 (0.07)    0.84 (0.18)    0.10 (0.03)    0.49 (0.10)   rs1693439  GG  19 243  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    GA + AA  3247  237  1.03 (0.84, 1.28)  159  1.07 (0.85, 1.36)  84  1.25 (0.94, 1.67)  73  0.95 (0.71, 1.28)  59  1.02 (0.75, 1.40)    P-value (adjusted P-value)      0.76 (0.13)      0.55 (0.12)    0.13 (0.05)    0.75 (0.15)    0.88 (0.20)   rs9307239  CC  8625  557  1 (ref)  362  1 (ref)  174  1 (ref)  182  1 (ref)  145  1 (ref)    CT + TT  13 841  1046  1.17 (1.00, 1.37)  680  1.17 (0.99, 1.40)  313  1.12 (0.90, 1.40)  347  1.19 (0.96, 1.48)  256  1.10 (0.87, 1.39)    P-value (adjusted P-value)      0.04 (0.02)    0.07 (0.03)    0.30 (0.07)    0.11 (0.05)    0.41 (0.05)   rs4147536  CC  14 259  975  1 (ref)  617  1 (ref)  295  1 (ref)  305  1 (ref)  251  1 (ref)    CA + AA  8221  630  1.16 (0.99, 1.35)  426  1.25 (1.05, 1.48)  193  1.20 (0.96, 1.49)  224  1.32 (1.07, 1.62)  151  1.06 (0.84, 1.34)    P-value (adjusted P-value)      0.06 (0.03)    0.01b (0.02)    0.11 (0.03)    0.01b (0.02)    0.60 (0.13)   rs3811802  AA  6245  454  1 (ref)  295  1 (ref)  139  1 (ref)  150  1 (ref)  119  1 (ref)    AG + GG  16 261  1150  0.96 (0.81, 1.14)  747  0.96 (0.80, 1.16)  349  0.94 (0.75, 1.20)  378  0.96 (0.76, 1.20)  283  0.91 (0.71, 1.16)    P-value (adjusted P-value)      0.65 (0.12)    0.66 (0.17)    0.63 (0.13)    0.72 (0.13)    0.44 (0.07)  ADH1C   rs698  TT  7777  549  1 (ref)  368  1 (ref)  180  1 (ref)  182  1 (ref)  134  1 (ref)    TC + CC  14 729  1053  1.01 (0.86, 1.18)  674  0.96 (0.81, 1.14)  308  0.89 (0.71, 1.11)  346  1.00 (0.81, 1.24)  267  1.05 (0.83, 1.33)    P-value (adjusted P-value)      0.93 (0.20)    0.65 (0.20)    0.31 (0.15)    1.00 (0.20)    0.69 (0.08)   rs1662033  TT  10 409  748  1 (ref)  496  1 (ref)  238  1 (ref)  248  1 (ref)  188  1 (ref)    TG + GG  12 079  857  0.97 (0.83, 1.13)  547  0.93 (0.78, 1.10)  250  0.88 (0.71, 1.08)  281  0.96 (0.78, 1.18)  214  0.97 (0.77, 1.22)    P-value (adjusted P-value)      0.68 (0.13)    0.38 (0.10)    0.22 (0.13)    0.69 (0.08)    0.79 (0.13)   rs3114046  CC  19 235  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    CT + TT  3271  238  1.03 (0.84, 1.27)  159  1.07 (0.85, 1.35)  84  1.24 (0.93, 1.65)  73  0.95 (0.71, 1.27)  60  1.04 (0.76, 1.41)    P-value (adjusted P-value)      0.77 (0.15)    0.58 (0.17)    0.14 (0.10)    0.72 (0.10)    0.83 (0.15)   rs4147542  TT  10 907  797  1 (ref)  514  1 (ref)  245  1 (ref)  258  1 (ref)  202  1 (ref)    TC + CC  9628  754  1.07 (0.92, 1.25)  488  1.08 (0.91, 1.28)  219  1.01 (0.81, 1.26)  254  1.12 (0.91, 1.38)  188  1.06 (0.84, 1.33)    P-value (adjusted P-value)      0.37 (0.12)    0.40 (0.12)    0.94 (0.20)    0.28 (0.03)    0.63 (0.05)   rs283415  TT  6975  500  1 (ref)  334  1 (ref)  168  1 (ref)  160  1 (ref)  123  1 (ref)    TC + CC  15 531  1104  0.99 (0.84, 1.16)  709  0.95 (0.79, 1.13)  320  0.85 (0.68, 1.06)  369  1.03 (0.83, 1.29)  279  1.02 (0.80, 1.30)    P-value (adjusted P-value)      0.88 (0.17)    0.56 (0.15)    0.15 (0.12)    0.77 (0.13)    0.90 (0.18)   rs4699741  TT  19 601  1432  1 (ref)  931  1 (ref)  437  1 (ref)  470  1 (ref)  353  1 (ref)    TC + CC  2905  173  0.80 (0.63, 1.01)  112  0.79 (0.61, 1.03)  51  0.77 (0.55, 1.08)  59  0.83 (0.60, 1.14)  49  0.92 (0.66, 1.30)    P-value (adjusted P-value)      0.06 (0.03)    0.08 (0.03)    0.13 (0.08)    0.25 (0.02)    0.66 (0.07)    Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8918  638  1 (ref)  407  1 (ref)  185  1 (ref)  209  1 (ref)  155  1 (ref)    CA + AA  13 547  960  1.01 (0.87, 1.18)  632  1.05 (0.88, 1.24)  301  1.11 (0.89, 1.38)  318  1.02 (0.83, 1.26)  244  1.05 (0.83, 1.32)    P-value (adjusted P-value)      0.89 (0.17)    0.59 (0.15)    0.36 (0.08)    0.86 (0.18)    0.69 (0.15)   rs2075633  TT  11 627  799  1 (ref)  519  1 (ref)  254  1 (ref)  251  1 (ref)  200  1 (ref)    TC + CC  10 879  806  1.08 (0.93, 1.25)  524  1.08 (0.91, 1.27)  234  0.98 (0.79, 1.21)  278  1.19 (0.97, 1.46)  202  1.08 (0.86, 1.35)    P-value (adjusted P-value)      0.32 (0.07)    0.38 (0.07)    0.84 (0.18)    0.10 (0.03)    0.49 (0.10)   rs1693439  GG  19 243  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    GA + AA  3247  237  1.03 (0.84, 1.28)  159  1.07 (0.85, 1.36)  84  1.25 (0.94, 1.67)  73  0.95 (0.71, 1.28)  59  1.02 (0.75, 1.40)    P-value (adjusted P-value)      0.76 (0.13)      0.55 (0.12)    0.13 (0.05)    0.75 (0.15)    0.88 (0.20)   rs9307239  CC  8625  557  1 (ref)  362  1 (ref)  174  1 (ref)  182  1 (ref)  145  1 (ref)    CT + TT  13 841  1046  1.17 (1.00, 1.37)  680  1.17 (0.99, 1.40)  313  1.12 (0.90, 1.40)  347  1.19 (0.96, 1.48)  256  1.10 (0.87, 1.39)    P-value (adjusted P-value)      0.04 (0.02)    0.07 (0.03)    0.30 (0.07)    0.11 (0.05)    0.41 (0.05)   rs4147536  CC  14 259  975  1 (ref)  617  1 (ref)  295  1 (ref)  305  1 (ref)  251  1 (ref)    CA + AA  8221  630  1.16 (0.99, 1.35)  426  1.25 (1.05, 1.48)  193  1.20 (0.96, 1.49)  224  1.32 (1.07, 1.62)  151  1.06 (0.84, 1.34)    P-value (adjusted P-value)      0.06 (0.03)    0.01b (0.02)    0.11 (0.03)    0.01b (0.02)    0.60 (0.13)   rs3811802  AA  6245  454  1 (ref)  295  1 (ref)  139  1 (ref)  150  1 (ref)  119  1 (ref)    AG + GG  16 261  1150  0.96 (0.81, 1.14)  747  0.96 (0.80, 1.16)  349  0.94 (0.75, 1.20)  378  0.96 (0.76, 1.20)  283  0.91 (0.71, 1.16)    P-value (adjusted P-value)      0.65 (0.12)    0.66 (0.17)    0.63 (0.13)    0.72 (0.13)    0.44 (0.07)  ADH1C   rs698  TT  7777  549  1 (ref)  368  1 (ref)  180  1 (ref)  182  1 (ref)  134  1 (ref)    TC + CC  14 729  1053  1.01 (0.86, 1.18)  674  0.96 (0.81, 1.14)  308  0.89 (0.71, 1.11)  346  1.00 (0.81, 1.24)  267  1.05 (0.83, 1.33)    P-value (adjusted P-value)      0.93 (0.20)    0.65 (0.20)    0.31 (0.15)    1.00 (0.20)    0.69 (0.08)   rs1662033  TT  10 409  748  1 (ref)  496  1 (ref)  238  1 (ref)  248  1 (ref)  188  1 (ref)    TG + GG  12 079  857  0.97 (0.83, 1.13)  547  0.93 (0.78, 1.10)  250  0.88 (0.71, 1.08)  281  0.96 (0.78, 1.18)  214  0.97 (0.77, 1.22)    P-value (adjusted P-value)      0.68 (0.13)    0.38 (0.10)    0.22 (0.13)    0.69 (0.08)    0.79 (0.13)   rs3114046  CC  19 235  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    CT + TT  3271  238  1.03 (0.84, 1.27)  159  1.07 (0.85, 1.35)  84  1.24 (0.93, 1.65)  73  0.95 (0.71, 1.27)  60  1.04 (0.76, 1.41)    P-value (adjusted P-value)      0.77 (0.15)    0.58 (0.17)    0.14 (0.10)    0.72 (0.10)    0.83 (0.15)   rs4147542  TT  10 907  797  1 (ref)  514  1 (ref)  245  1 (ref)  258  1 (ref)  202  1 (ref)    TC + CC  9628  754  1.07 (0.92, 1.25)  488  1.08 (0.91, 1.28)  219  1.01 (0.81, 1.26)  254  1.12 (0.91, 1.38)  188  1.06 (0.84, 1.33)    P-value (adjusted P-value)      0.37 (0.12)    0.40 (0.12)    0.94 (0.20)    0.28 (0.03)    0.63 (0.05)   rs283415  TT  6975  500  1 (ref)  334  1 (ref)  168  1 (ref)  160  1 (ref)  123  1 (ref)    TC + CC  15 531  1104  0.99 (0.84, 1.16)  709  0.95 (0.79, 1.13)  320  0.85 (0.68, 1.06)  369  1.03 (0.83, 1.29)  279  1.02 (0.80, 1.30)    P-value (adjusted P-value)      0.88 (0.17)    0.56 (0.15)    0.15 (0.12)    0.77 (0.13)    0.90 (0.18)   rs4699741  TT  19 601  1432  1 (ref)  931  1 (ref)  437  1 (ref)  470  1 (ref)  353  1 (ref)    TC + CC  2905  173  0.80 (0.63, 1.01)  112  0.79 (0.61, 1.03)  51  0.77 (0.55, 1.08)  59  0.83 (0.60, 1.14)  49  0.92 (0.66, 1.30)    P-value (adjusted P-value)      0.06 (0.03)    0.08 (0.03)    0.13 (0.08)    0.25 (0.02)    0.66 (0.07)  CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAge-adjusted. bSignificant after adjusting for multiple testing. View Large Table 4. Hazard ratios and 95% confidence intervals for the association between single nucleotide polymorphisms in ADH1B and ADH1C and risk of overall colorectal cancer and subtypes in female drinkers in the NLCS, 1986–2006   Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8159  366  1 (ref)  270  1 (ref)  157  1 (ref)  109  1 (ref)  66  1 (ref)    CA + AA  11 718  517  0.98 (0.82, 1.18)  384  0.99 (0.81, 1.20)  234  1.03 (0.81, 1.31)  141  0.90 (0.68, 1.19)  99  1.04 (0.75, 1.45)    P-value (adjusted P-value)      0.82 (0.15)    0.88 (0.20)    0.80 (0.17)    0.45 (0.10)    0.82 (0.18)   rs2075633  TT  9702  457  1 (ref)  338  1 (ref)  216  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  10 233  429  0.89 (0.75, 1.07)  318  0.89 (0.74, 1.09)  177  0.78 (0.62, 0.99)  136  1.13 (0.86, 1.49)  76  0.81 (0.59, 1.13)    P-value (adjusted P-value)      0.22 (0.05)    0.27 (0.05)    0.04 (0.02)    0.38 (0.08)    0.22 (0.02)   rs1693439  GG  16 820  746  1 (ref)  545  1 (ref)  328  1 (ref)  208  1 (ref)  141  1 (ref)    GA + AA  3135  140  1.01 (0.79, 1.29)  111  1.09 (0.84, 1.42)  65  1.06 (0.77, 1.46)  42  1.08 (0.75, 1.57)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.95 (0.20)    0.51 (0.10)    0.70 (0.15)    0.67 (0.12)    0.71 (0.17)   rs9307239  CC  7622  348  1 (ref)  258  1 (ref)  158  1 (ref)  94  1 (ref)  67  1 (ref)    CT + TT  12 313  536  0.95 (0.79,1.14)  396  0.95 (0.77, 1.16)  234  0.91 (0.72, 1.16)  155  1.02 (0.77, 1.35)  98  0.90 (0.65, 1.26)    P-value (adjusted P-value)      0.58 (0.10)    0.59 (0.13)    0.45 (0.10)    0.91 (0.20)    0.55 (0.12)   rs4147536  CC  12 688  563  1 (ref)  426  1 (ref)  249  1 (ref)  171  1 (ref)  98  1 (ref)    CA + AA  7247  321  0.99 (0.82, 1.19)  228  0.93 (0.75, 1.14)  144  1.00 (0.78, 1.27)  77  0.78 (0.58, 1.06)  67  1.19 (0.85, 1.66)    P-value (adjusted P-value)      0.90 (0.18)    0.47 (0.08)    0.99 (0.20)    0.11 (0.07)    0.32 (0.03)   rs3811802  AA  6101  260  1 (ref)  195  1 (ref)  115  1 (ref)  78  1 (ref)  46  1 (ref)    AG + GG  13 854  625  1.06 (0.87, 1.29)  460  1.05 (0.84, 1.29)  278  1.07 (0.83, 1.39)  171  0.97 (0.72, 1.31)  119  1.14 (0.79, 1.64)    P-value (adjusted P-value)      0.53 (0.08)    0.68 (0.18)    0.59 (0.12)    0.84 (0.17)    0.47 (0.08)  ADH1C   rs698  TT  7385  304  1 (ref)  225  1 (ref)  126  1 (ref)  95  1 (ref)  62  1 (ref)    TC + CC  12 553  580  1.13 (0.94, 1.36)  430  1.13 (0.92, 1.39)  266  1.25 (0.98, 1.61)  155  0.97 (0.73, 1.28)  102  0.97 (0.69, 1.37)    P-value (adjusted P-value)      0.20 (0.07)    0.23 (0.07)    0.08 (0.07)    0.81 (0.15)    0.88 (0.17)   rs1662033  TT  9720  410  1 (ref)  300  1 (ref)  169  1 (ref)  124  1 (ref)  80  1 (ref)    TG + GG  10 215  476  1.11 (0.93, 1.33)  356  1.13 (0.93, 1.38)  224  1.27 (1.00, 1.60)  126  0.97 (0.74, 1.28)  85  1.01 (0.73, 1.41)    P-value (adjusted P-value)      0.26 (0.10)    0.21 (0.05)    0.05 (0.03)    0.83 (0.17)    0.93 (0.20)   rs3114046  CC  16 820  745  1 (ref)  544  1 (ref)  328  1 (ref)  207  1 (ref)  141  1 (ref)    CT + TT  3135  141  1.02 (0.79, 1.30)  112  1.10 (0.85, 1.44)  65  1.06 (0.77, 1.46)  43  1.11 (0.77, 1.61)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.90 (0.18)    0.46 (0.13)    0.70 (0.17)    0.56 (0.07)    0.71 (0.10)   rs4147542  TT  9297  443  1 (ref)  324  1 (ref)  205  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  9566  402  0.87 (0.73, 1.05)  300  0.89 (0.73, 1.09)  169  0.79 (0.62, 1.01)  123  1.04 (0.79, 1.38)  71  0.77 (0.55, 1.08)    P-value (adjusted P-value)      0.15 (0.05)    0.26 (0.08)    0.06 (0.05)    0.77 (0.12)    0.13 (0.03)   rs283415  TT  6906  272  1 (ref)  200  1 (ref)  109  1 (ref)  87  1 (ref)  55  1 (ref)    TC + CC  13 049  614  1.20 (0.99, 1.46)  456  1.22 (0.99, 1.50)  284  1.39 (1.08, 1.80)  163  1.00 (0.75, 1.33)  110  1.06 (0.75, 1.51)    P-value (adjusted P-value)      0.06 (0.02)    0.07 (0.02)    0.01b (0.02)    0.99 (0.18)    0.72 (0.12)   rs4699741  TT  17 642  769  1 (ref)  577  1 (ref)  347  1 (ref)  217  1 (ref)  136  1 (ref)    TC + CC  2314  117  1.19 (0.91, 1.57)  79  1.08 (0.80, 1.46)  46  1.06 (0.73, 1.52)  33  1.18 (0.78, 1.78)  29  1.66 (1.06, 2.58)    P-value (adjusted P-value)      0.20 (0.08)    0.62 (0.18)    0.77 (0.18)    0.43 (0.05)    0.03 (0.02)    Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8159  366  1 (ref)  270  1 (ref)  157  1 (ref)  109  1 (ref)  66  1 (ref)    CA + AA  11 718  517  0.98 (0.82, 1.18)  384  0.99 (0.81, 1.20)  234  1.03 (0.81, 1.31)  141  0.90 (0.68, 1.19)  99  1.04 (0.75, 1.45)    P-value (adjusted P-value)      0.82 (0.15)    0.88 (0.20)    0.80 (0.17)    0.45 (0.10)    0.82 (0.18)   rs2075633  TT  9702  457  1 (ref)  338  1 (ref)  216  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  10 233  429  0.89 (0.75, 1.07)  318  0.89 (0.74, 1.09)  177  0.78 (0.62, 0.99)  136  1.13 (0.86, 1.49)  76  0.81 (0.59, 1.13)    P-value (adjusted P-value)      0.22 (0.05)    0.27 (0.05)    0.04 (0.02)    0.38 (0.08)    0.22 (0.02)   rs1693439  GG  16 820  746  1 (ref)  545  1 (ref)  328  1 (ref)  208  1 (ref)  141  1 (ref)    GA + AA  3135  140  1.01 (0.79, 1.29)  111  1.09 (0.84, 1.42)  65  1.06 (0.77, 1.46)  42  1.08 (0.75, 1.57)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.95 (0.20)    0.51 (0.10)    0.70 (0.15)    0.67 (0.12)    0.71 (0.17)   rs9307239  CC  7622  348  1 (ref)  258  1 (ref)  158  1 (ref)  94  1 (ref)  67  1 (ref)    CT + TT  12 313  536  0.95 (0.79,1.14)  396  0.95 (0.77, 1.16)  234  0.91 (0.72, 1.16)  155  1.02 (0.77, 1.35)  98  0.90 (0.65, 1.26)    P-value (adjusted P-value)      0.58 (0.10)    0.59 (0.13)    0.45 (0.10)    0.91 (0.20)    0.55 (0.12)   rs4147536  CC  12 688  563  1 (ref)  426  1 (ref)  249  1 (ref)  171  1 (ref)  98  1 (ref)    CA + AA  7247  321  0.99 (0.82, 1.19)  228  0.93 (0.75, 1.14)  144  1.00 (0.78, 1.27)  77  0.78 (0.58, 1.06)  67  1.19 (0.85, 1.66)    P-value (adjusted P-value)      0.90 (0.18)    0.47 (0.08)    0.99 (0.20)    0.11 (0.07)    0.32 (0.03)   rs3811802  AA  6101  260  1 (ref)  195  1 (ref)  115  1 (ref)  78  1 (ref)  46  1 (ref)    AG + GG  13 854  625  1.06 (0.87, 1.29)  460  1.05 (0.84, 1.29)  278  1.07 (0.83, 1.39)  171  0.97 (0.72, 1.31)  119  1.14 (0.79, 1.64)    P-value (adjusted P-value)      0.53 (0.08)    0.68 (0.18)    0.59 (0.12)    0.84 (0.17)    0.47 (0.08)  ADH1C   rs698  TT  7385  304  1 (ref)  225  1 (ref)  126  1 (ref)  95  1 (ref)  62  1 (ref)    TC + CC  12 553  580  1.13 (0.94, 1.36)  430  1.13 (0.92, 1.39)  266  1.25 (0.98, 1.61)  155  0.97 (0.73, 1.28)  102  0.97 (0.69, 1.37)    P-value (adjusted P-value)      0.20 (0.07)    0.23 (0.07)    0.08 (0.07)    0.81 (0.15)    0.88 (0.17)   rs1662033  TT  9720  410  1 (ref)  300  1 (ref)  169  1 (ref)  124  1 (ref)  80  1 (ref)    TG + GG  10 215  476  1.11 (0.93, 1.33)  356  1.13 (0.93, 1.38)  224  1.27 (1.00, 1.60)  126  0.97 (0.74, 1.28)  85  1.01 (0.73, 1.41)    P-value (adjusted P-value)      0.26 (0.10)    0.21 (0.05)    0.05 (0.03)    0.83 (0.17)    0.93 (0.20)   rs3114046  CC  16 820  745  1 (ref)  544  1 (ref)  328  1 (ref)  207  1 (ref)  141  1 (ref)    CT + TT  3135  141  1.02 (0.79, 1.30)  112  1.10 (0.85, 1.44)  65  1.06 (0.77, 1.46)  43  1.11 (0.77, 1.61)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.90 (0.18)    0.46 (0.13)    0.70 (0.17)    0.56 (0.07)    0.71 (0.10)   rs4147542  TT  9297  443  1 (ref)  324  1 (ref)  205  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  9566  402  0.87 (0.73, 1.05)  300  0.89 (0.73, 1.09)  169  0.79 (0.62, 1.01)  123  1.04 (0.79, 1.38)  71  0.77 (0.55, 1.08)    P-value (adjusted P-value)      0.15 (0.05)    0.26 (0.08)    0.06 (0.05)    0.77 (0.12)    0.13 (0.03)   rs283415  TT  6906  272  1 (ref)  200  1 (ref)  109  1 (ref)  87  1 (ref)  55  1 (ref)    TC + CC  13 049  614  1.20 (0.99, 1.46)  456  1.22 (0.99, 1.50)  284  1.39 (1.08, 1.80)  163  1.00 (0.75, 1.33)  110  1.06 (0.75, 1.51)    P-value (adjusted P-value)      0.06 (0.02)    0.07 (0.02)    0.01b (0.02)    0.99 (0.18)    0.72 (0.12)   rs4699741  TT  17 642  769  1 (ref)  577  1 (ref)  347  1 (ref)  217  1 (ref)  136  1 (ref)    TC + CC  2314  117  1.19 (0.91, 1.57)  79  1.08 (0.80, 1.46)  46  1.06 (0.73, 1.52)  33  1.18 (0.78, 1.78)  29  1.66 (1.06, 2.58)    P-value (adjusted P-value)      0.20 (0.08)    0.62 (0.18)    0.77 (0.18)    0.43 (0.05)    0.03 (0.02)  CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAge-adjusted. bSignificant after adjusting for multiple testing. View Large Effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs Table 5 shows the associations between alcohol intake and CRC risk in men and women, stratified by genotype, as analyzed according to a dominant model. The alcohol–CRC associations observed in genotype strata generally aligned with overall alcohol–CRC associations. HRs around the null were observed in one stratum and a pattern of increasing risks across alcohol categories was observed in the other stratum in men, while a difference in associations between genotype strata was less clear in women. Statistically significantly increased HRs for CRC were only present when comparing heavy drinkers with abstainers. Most interactions were not significant after FDR correction except for the interactions between alcohol intake and ADH1B rs3811802 and ADH1C rs4147542 in women. Risk was strongly increased in heavy (but not light-moderate) drinkers versus abstainers in female rs3811802 AA carriers, although the CI around this estimate is large [HR (95% CI) 5.72 (2.24–14.63)], while the risk across alcohol categories in female rs3811802 AG/GG carriers remained almost unchanged. In female rs4147542 TT carriers, there was a significant positive trend in CRC risk with increasing alcohol intake [HRlight-moderate drinkers versus abstainers (95% CI) 1.25 (0.98–1.59) and HRheavy drinkers versus abstainers (95% CI) 1.78 (1.01–3.14)], while a decreased risk was observed for light-moderate drinkers and an increased risk for heavy drinkers relative to abstainers in female rs4147542 TC/CC carriers [HR (95% CI) 0.78 (0.60–1.02) and HR (95% CI) 1.74 (0.92–3.31), respectively]. After restricting the analysis to the stable subgroup (Supplementary Table 3, available at Carcinogenesis Online), only the interaction between alcohol intake and ADH1C rs4147542 in relation to CRC risk in women remained significant after FDR correction. In addition, in men, results were more pronounced, showing stronger increased HRs for CRC in heavy drinkers as compared with abstainers and more significant positive trends. Table 5. Hazard ratios and 95% confidence intervals for the association between alcohol intake and colorectal cancer risk by ADH1B and ADH1C genotypes in men and women in the NLCS, 1986–2006 Gene  SNP  Allele  Alcohol intake            Abstainers  Light-moderate (0.1–29 g/day)  Heavy (≥30 g/day)            N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  P for trend  P for interaction (adjusted P-value)  Men   ADH1B  rs1159918  CC  88/1232  1  (ref)  499/7191  0.92  (0.65, 1.31)  116/1445  1.05  (0.68, 1.64)  0.75        CA/AA  117/2071  1  (ref)  743/10 709  1.22  (0.91, 1.64)  182/2269  1.41  (0.98, 2.02)  0.07  0.41 (0.10)   ADH1B  rs2075633  TT  95/1596  1  (ref)  625/9152  1.14  (0.81, 1.59)  154/1960  1.34  (0.89, 2.01)  0.14        TC/CC  110/1707  1  (ref)  624/8769  1.09  (0.80, 1.49)  144/1774  1.20  (0.80, 1.79)  0.38  0.99 (0.20)   ADH1B  rs1693439  GG  185/2880  1  (ref)  1073/15305  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        GA/AA  20/424  1  (ref)  175/2599  1.12  (0.53, 2.35)  52/515  1.98  (0.82, 4.81)  0.08  0.28 (0.05)   ADH1B  rs9307239  CC  73/1083  1  (ref)  434/6872  0.88  (0.61, 1.28)  108/1504  1.01  (0.64, 1.58)  0.86        CT/TT  132/2208  1  (ref)  814/11 009  1.22  (0.91, 1.62)  189/2230  1.37  (0.96, 1.96)  0.09  0.40 (0.08)   ADH1B  rs4147536  CC  127/1985  1  (ref)  760/11 432  0.99  (0.75, 1.31)  177/2395  1.09  (0.76, 1.55)  0.59        CA/AA  78/1319  1  (ref)  489/6483  1.29  (0.89, 1.88)  121/1338  1.53  (0.96, 2.46)  0.08  0.42 (0.12)   ADH1B  rs3811802  AA  59/1007  1  (ref)  350/5062  1.23  (0.80, 1.88)  82/1013  1.29  (0.75, 2.22)  0.38        AG/GG  146/2297  1  (ref)  898/12 859  1.04  (0.79, 1.37)  216/2721  1.21  (0.87, 1.69)  0.22  0.80 (0.17)   ADH1C  rs698  TT  69/1270  1  (ref)  419/6312  1.33  (0.90, 1.97)  105/1197  1.75  (1.07, 2.86)  0.03        TC/CC  136/2034  1  (ref)  827/11 609  1.02  (0.77, 1.36)  193/2537  1.11  (0.78, 1.57)  0.53  0.35 (0.10)   ADH1C  rs1662033  TT  97/1773  1  (ref)  576/8380  1.33  (0.96, 1.86)  144/1678  1.69  (1.11, 2.57)  0.02        TG/GG  108/1530  1  (ref)  673/9541  0.95  (0.68, 1.31)  154/2056  1.02  (0.68, 1.51)  0.86  0.17 (0.05)   ADH1C  rs3114046  CC  185/2880  1  (ref)  1073/15297  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        CT/TT  20/424  1  (ref)  176/2623  1.13  (0.54, 2.38)  52/515  1.88  (0.78, 4.51)  0.10  0.27 (0.08   ADH1C  rs4147542  TT  107/1260  1  (ref)  631/8763  0.85  (0.61, 1.18)  144/1723  0.98  (0.65, 1.48)  0.95        TC/CC  94/1706  1  (ref)  570/7672  1.28  (0.91, 1.79)  149/1575  1.61  (1.06, 2.46)  0.03  0.12 (0.03)   ADH1C  rs283415  TT  65/1243  1  (ref)  385/5654  1.40  (0.94, 2.08)  93/1104  1.66  (1.00, 2.75)  0.05        TC/CC  140/2060  1  (ref)  863/12 266  1.00  (0.75, 1.32)  205/2630  1.12  (0.79, 1.58)  0.46  0.37 (0.12)   ADH1C  rs4699741  TT  179/2738  1  (ref)  1121/15601  1.05  (0.83, 1.35)  258/3230  1.17  (0.86, 1.58)  0.31        TC/CC  26/565  1  (ref)  128/2319  1.19  (0.65, 2.19)  40/503  1.61  (0.75, 3.44)  0.22  0.50 (0.15)  Women   ADH1B  rs1159918  CC  147/3395  1  (ref)  335/7422  1.08  (0.82, 1.43)  20/392  1.43  (0.73, 2.81)  0.38        CA/AA  234/5460  1  (ref)  456/10 666  0.99  (0.79, 1.24)  38/572  1.75  (1.03, 2.97)  0.33  0.68 (0.15)   ADH1B  rs2075633  TT  199/4735  1  (ref)  412/8725  1.13  (0.89, 1.43)  28/570  1.27  (0.72, 2.23)  0.26        TC/CC  181/4120  1  (ref)  382/9420  0.94  (0.73, 1.21)  30/394  2.10  (1.13, 3.91)  0.46  0.16 (0.03)   ADH1B  rs1693439  GG  324/7497  1  (ref)  668/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        GA/AA  57/1358  1  (ref)  126/2793  1.11  (0.69, 1.81)  11/157  1.57  (0.54, 4.60)  0.46  0.96 (0.18)   ADH1B  rs9307239  CC  148/3546  1  (ref)  309/6954  1.08  (0.81, 1.44)  28/377  1.80  (0.95, 3.38)  0.18        CT/TT  233/5292  1  (ref)  483/11 191  0.98  (0.78, 1.22)  30/587  1.39  (0.79, 2.43)  0.71  0.62 (0.13)   ADH1B  rs4147536  CC  236/5557  1  (ref)  513/11 574  1.08  (0.86, 1.34)  32/637  1.37  (0.80, 2.33)  0.30        CA/AA  145/3298  1  (ref)  279/6572  0.94  (0.70, 1.25)  26/328  2.03  (1.03, 3.99)  0.45  0.32 (0.07)   ADH1B  rs3811802  AA  107/2509  1  (ref)  224/5709  0.96  (0.68, 1.37)  22/137  5.72  (2.24, 14.63)  0.13        AG/GG  274/6313  1  (ref)  569/12 457  1.04  (0.85, 1.28)  36/827  1.09  (0.67, 1.77)  0.63  0.004b (0.02)   ADH1C  rs698  TT  139/3189  1  (ref)  273/6696  0.94  (0.70, 1.27)  17/349  1.50  (0.73, 3.06)  0.86        TC/CC  240/5666  1  (ref)  520/11 453  1.09  (0.88, 1.36)  40/615  1.70  (1.02, 2.84)  0.11  0.60 (0.17)   ADH1C  rs1662033  TT  193/4346  1  (ref)  367/8777  0.92  (0.72, 1.19)  25/440  1.59  (0.87, 2.92)  0.84        TG/GG  188/4508  1  (ref)  427/9369  1.12  (0.87, 1.44)  33/525  1.63  (0.92, 2.87)  0.13  0.44 (0.13)   ADH1C  rs3114046  CC  324/7497  1  (ref)  667/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        CT/TT  57/1358  1  (ref)  127/2793  1.12  (0.69, 1.82)  11/157  1.55  (0.53, 4.54)  0.45  0.95 (0.20)   ADH1C  rs4147542  TT  182/4890  1  (ref)  399/8448  1.25  (0.98, 1.59)  29/458  1.78  (1.01, 3.14)  0.02        TC/CC  180/3474  1  (ref)  356/8747  0.78  (0.60, 1.02)  28/405  1.74  (0.92, 3.31)  0.58  0.01b (0.02)   ADH1C  rs283415  TT  134/2923  1  (ref)  245/6263  0.87  (0.64, 1.19)  14/324  1.27  (0.59, 2.72)  0.67        TC/CC  247/5932  1  (ref)  549/11 903  1.12  (0.90, 1.39)  44/640  1.83  (1.11, 3.02)  0.05  0.19 (0.07)   ADH1C  rs4699741  TT  326/7548  1  (ref)  692/16 037  1.00  (0.83, 1.20)  50/850  1.52  (0.97, 2.37)  0.39        TC/CC  55/1307  1  (ref)  102/2129  1.43  (0.83, 2.46)  8/114  2.08  (0.61, 7.12)  0.12  0.82 (0.18)  Gene  SNP  Allele  Alcohol intake            Abstainers  Light-moderate (0.1–29 g/day)  Heavy (≥30 g/day)            N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  P for trend  P for interaction (adjusted P-value)  Men   ADH1B  rs1159918  CC  88/1232  1  (ref)  499/7191  0.92  (0.65, 1.31)  116/1445  1.05  (0.68, 1.64)  0.75        CA/AA  117/2071  1  (ref)  743/10 709  1.22  (0.91, 1.64)  182/2269  1.41  (0.98, 2.02)  0.07  0.41 (0.10)   ADH1B  rs2075633  TT  95/1596  1  (ref)  625/9152  1.14  (0.81, 1.59)  154/1960  1.34  (0.89, 2.01)  0.14        TC/CC  110/1707  1  (ref)  624/8769  1.09  (0.80, 1.49)  144/1774  1.20  (0.80, 1.79)  0.38  0.99 (0.20)   ADH1B  rs1693439  GG  185/2880  1  (ref)  1073/15305  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        GA/AA  20/424  1  (ref)  175/2599  1.12  (0.53, 2.35)  52/515  1.98  (0.82, 4.81)  0.08  0.28 (0.05)   ADH1B  rs9307239  CC  73/1083  1  (ref)  434/6872  0.88  (0.61, 1.28)  108/1504  1.01  (0.64, 1.58)  0.86        CT/TT  132/2208  1  (ref)  814/11 009  1.22  (0.91, 1.62)  189/2230  1.37  (0.96, 1.96)  0.09  0.40 (0.08)   ADH1B  rs4147536  CC  127/1985  1  (ref)  760/11 432  0.99  (0.75, 1.31)  177/2395  1.09  (0.76, 1.55)  0.59        CA/AA  78/1319  1  (ref)  489/6483  1.29  (0.89, 1.88)  121/1338  1.53  (0.96, 2.46)  0.08  0.42 (0.12)   ADH1B  rs3811802  AA  59/1007  1  (ref)  350/5062  1.23  (0.80, 1.88)  82/1013  1.29  (0.75, 2.22)  0.38        AG/GG  146/2297  1  (ref)  898/12 859  1.04  (0.79, 1.37)  216/2721  1.21  (0.87, 1.69)  0.22  0.80 (0.17)   ADH1C  rs698  TT  69/1270  1  (ref)  419/6312  1.33  (0.90, 1.97)  105/1197  1.75  (1.07, 2.86)  0.03        TC/CC  136/2034  1  (ref)  827/11 609  1.02  (0.77, 1.36)  193/2537  1.11  (0.78, 1.57)  0.53  0.35 (0.10)   ADH1C  rs1662033  TT  97/1773  1  (ref)  576/8380  1.33  (0.96, 1.86)  144/1678  1.69  (1.11, 2.57)  0.02        TG/GG  108/1530  1  (ref)  673/9541  0.95  (0.68, 1.31)  154/2056  1.02  (0.68, 1.51)  0.86  0.17 (0.05)   ADH1C  rs3114046  CC  185/2880  1  (ref)  1073/15297  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        CT/TT  20/424  1  (ref)  176/2623  1.13  (0.54, 2.38)  52/515  1.88  (0.78, 4.51)  0.10  0.27 (0.08   ADH1C  rs4147542  TT  107/1260  1  (ref)  631/8763  0.85  (0.61, 1.18)  144/1723  0.98  (0.65, 1.48)  0.95        TC/CC  94/1706  1  (ref)  570/7672  1.28  (0.91, 1.79)  149/1575  1.61  (1.06, 2.46)  0.03  0.12 (0.03)   ADH1C  rs283415  TT  65/1243  1  (ref)  385/5654  1.40  (0.94, 2.08)  93/1104  1.66  (1.00, 2.75)  0.05        TC/CC  140/2060  1  (ref)  863/12 266  1.00  (0.75, 1.32)  205/2630  1.12  (0.79, 1.58)  0.46  0.37 (0.12)   ADH1C  rs4699741  TT  179/2738  1  (ref)  1121/15601  1.05  (0.83, 1.35)  258/3230  1.17  (0.86, 1.58)  0.31        TC/CC  26/565  1  (ref)  128/2319  1.19  (0.65, 2.19)  40/503  1.61  (0.75, 3.44)  0.22  0.50 (0.15)  Women   ADH1B  rs1159918  CC  147/3395  1  (ref)  335/7422  1.08  (0.82, 1.43)  20/392  1.43  (0.73, 2.81)  0.38        CA/AA  234/5460  1  (ref)  456/10 666  0.99  (0.79, 1.24)  38/572  1.75  (1.03, 2.97)  0.33  0.68 (0.15)   ADH1B  rs2075633  TT  199/4735  1  (ref)  412/8725  1.13  (0.89, 1.43)  28/570  1.27  (0.72, 2.23)  0.26        TC/CC  181/4120  1  (ref)  382/9420  0.94  (0.73, 1.21)  30/394  2.10  (1.13, 3.91)  0.46  0.16 (0.03)   ADH1B  rs1693439  GG  324/7497  1  (ref)  668/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        GA/AA  57/1358  1  (ref)  126/2793  1.11  (0.69, 1.81)  11/157  1.57  (0.54, 4.60)  0.46  0.96 (0.18)   ADH1B  rs9307239  CC  148/3546  1  (ref)  309/6954  1.08  (0.81, 1.44)  28/377  1.80  (0.95, 3.38)  0.18        CT/TT  233/5292  1  (ref)  483/11 191  0.98  (0.78, 1.22)  30/587  1.39  (0.79, 2.43)  0.71  0.62 (0.13)   ADH1B  rs4147536  CC  236/5557  1  (ref)  513/11 574  1.08  (0.86, 1.34)  32/637  1.37  (0.80, 2.33)  0.30        CA/AA  145/3298  1  (ref)  279/6572  0.94  (0.70, 1.25)  26/328  2.03  (1.03, 3.99)  0.45  0.32 (0.07)   ADH1B  rs3811802  AA  107/2509  1  (ref)  224/5709  0.96  (0.68, 1.37)  22/137  5.72  (2.24, 14.63)  0.13        AG/GG  274/6313  1  (ref)  569/12 457  1.04  (0.85, 1.28)  36/827  1.09  (0.67, 1.77)  0.63  0.004b (0.02)   ADH1C  rs698  TT  139/3189  1  (ref)  273/6696  0.94  (0.70, 1.27)  17/349  1.50  (0.73, 3.06)  0.86        TC/CC  240/5666  1  (ref)  520/11 453  1.09  (0.88, 1.36)  40/615  1.70  (1.02, 2.84)  0.11  0.60 (0.17)   ADH1C  rs1662033  TT  193/4346  1  (ref)  367/8777  0.92  (0.72, 1.19)  25/440  1.59  (0.87, 2.92)  0.84        TG/GG  188/4508  1  (ref)  427/9369  1.12  (0.87, 1.44)  33/525  1.63  (0.92, 2.87)  0.13  0.44 (0.13)   ADH1C  rs3114046  CC  324/7497  1  (ref)  667/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        CT/TT  57/1358  1  (ref)  127/2793  1.12  (0.69, 1.82)  11/157  1.55  (0.53, 4.54)  0.45  0.95 (0.20)   ADH1C  rs4147542  TT  182/4890  1  (ref)  399/8448  1.25  (0.98, 1.59)  29/458  1.78  (1.01, 3.14)  0.02        TC/CC  180/3474  1  (ref)  356/8747  0.78  (0.60, 1.02)  28/405  1.74  (0.92, 3.31)  0.58  0.01b (0.02)   ADH1C  rs283415  TT  134/2923  1  (ref)  245/6263  0.87  (0.64, 1.19)  14/324  1.27  (0.59, 2.72)  0.67        TC/CC  247/5932  1  (ref)  549/11 903  1.12  (0.90, 1.39)  44/640  1.83  (1.11, 3.02)  0.05  0.19 (0.07)   ADH1C  rs4699741  TT  326/7548  1  (ref)  692/16 037  1.00  (0.83, 1.20)  50/850  1.52  (0.97, 2.37)  0.39        TC/CC  55/1307  1  (ref)  102/2129  1.43  (0.83, 2.46)  8/114  2.08  (0.61, 7.12)  0.12  0.82 (0.18)  CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAdjusted for age, BMI, smoking, family history of CRC, meat intake, processed meat intake, folate intake and non-occupational physical activity. bSignificant after adjusting for multiple testing. View Large Supplementary Tables 4–7, available at Carcinogenesis Online, show the associations between alcohol intake and the risk of CRC by subsite, i.e. the colon, proximal colon, distal colon and rectum, in men and women, stratified by genotype. After FDR correction, the only statistically significant interactions observed were those between alcohol intake and ADH1B rs3811802 and ADH1C rs4147542 in relation to (proximal) colon cancer in women, consistent with the interactions observed for CRC in women. In men, a difference in associations was observed between genotype strata, when considering alcohol intake in relation to the risk of colon and proximal colon cancer, and, in particular, rectal cancer, but less so or not in relation to distal colon cancer. In women, the power was limited in analyses for distal colon and rectal cancer, hampering a proper comparison. Discussion This study addressed the lack of evidence regarding alcohol intake and CRC risk in women and found alcohol to be a CRC risk factor in men and women. Associations with CRC differed by sex. Alcohol intake increased CRC risk (non-)significantly at light-moderate and heavy intake levels across subsites in men. Only when alcohol intake exceeded 30 g/day, we observed increased colon cancer risks, particularly for the proximal colon, in women. We studied associations in ADH1B and ADH1C genetic subgroups, because these may be particularly susceptible to the deleterious effects of alcohol on CRC risk. ADH1B rs3811802 and ADH1C rs4147542 modified the association between alcohol intake and the risk of colorectal, colon and proximal colon cancer in women after FDR correction. The alcohol–CRC associations observed in genotype strata generally aligned with overall alcohol–CRC associations. A difference in associations between genotype strata was generally clearer in men than women but significant effect modification was only present in women. Restricting to participants with equal alcohol intake amounts 5 years before baseline resulted in (more) significant positive linear trends across alcohol intake categories within genotype strata in men but not women. Furthermore, ADH1B rs4147536 and ADH1C rs283415—which was in strong linkage disequilibrium (LD) with the commonly investigated ADH1C rs698 (r2 = 0.9) in our data—were associated with an increased cancer risk at colon subsites in male and female drinkers, respectively, after FDR correction. ADH1B rs3811802 and ADH1C rs4147542 were associated with a decreased CRC risk in female heavy and stable drinkers, respectively, after FDR correction. These results substantiate the interplay between alcohol intake, ADH1B and ADH1C in relation to CRC risk. A potential sex difference in intake levels at which alcohol intake increases CRC risk may in part be explained by differences in drinking pattern. Men perhaps are more likely to be more regular consumers than women. Regular alcohol exposure may be especially deleterious and may also increase CRC risk at light-moderate levels (33). In addition, sex differences in first-pass metabolism of alcohol (i.e. presystemic elimination of ethanol in, predominantly, the stomach and liver (34)) and ADH activity could lead to sex differences in the CRC risk associated with alcohol intake (35). Women have prolonged, higher blood ethanol concentrations than men upon similar intake levels due to differences in elimination of alcohol (i.e. the volume distribution is higher in men than women) (35). However, based on this, one would expect women to be affected by alcohol at lower intake levels than men, whereas we found a non-significant association with CRC risk at light-moderate alcohol intake levels in men but not women. A more plausible explanation, therefore, may be that there are interactions between alcohol intake and sex-specific factors. For example, women might be protected from the adverse effects of alcohol at light-moderate intake levels through a positive relationship between alcohol intake and estradiol levels (35). Increased estradiol levels were found to be protective against CRC in women (36,37). Alternatively, as suggested by Klatsky et al. (38), an increased risk of cancer among light-moderate drinkers may be due in part to the underreporting of heavy alcohol intake. This could explain the associations observed with light-moderate alcohol intake in men as the percentage of heavy drinkers is higher in men than women (even though underreporting may be expected more in women than men due to social desirability standards). Finally, the fact that a sex difference was very pronounced in relation to rectal cancer could indicate that there may have been some residual confounding by smoking, even after adjusting for smoking. Smoking has been more strongly associated with rectal cancer than colon cancer risk (39), and smoking correlated more strongly with alcohol intake in men than women [88% of male drinkers as compared to 48% of female drinkers in the subcohort were ever smokers]. We found that the association between alcohol intake and (proximal) colon cancer risk in women was significantly modified by ADH1B rs3811802 and ADH1C rs4147542 after FDR correction. The SNPs selected in this study were not selected on the basis of that these were strong causal variants per se, but on the basis of that these common SNPs (minor allele frequency >5%) tag the genetic variation in ADH1B and ADH1C. Considering that tagSNPs generally confer only minor risks, which may be explained by imperfect correlations with true causal variants and gene–gene interactions, it is difficult to show significance in gene-environment interaction studies, even with large sample sizes (40). Therefore, it may be considered remarkable that two SNPs modified the association between alcohol intake and (proximal) colon cancer risk in women after FDR correction. Especially ADH1C rs4147542 is noteworthy in this regard since it can be linked to functional evidence: it is an expression quantitative trait locus (eQTL) for ADH1C in several tissues including the transverse colon (41), and has been reported to be a methylation quantitative trait locus (mQTL) (42). DNA methylation might, in part, underlie our finding of a gene–environment interaction between rs4147542 and alcohol intake which was specific to proximal colon cancer risk and to women, in which the CpG island methylator phenotype (CIMP) is present more often (43,44). Curtin et al. (45) showed that ADH1C rs698 was associated with CRCs positive for CIMP in those with low folate intake. Alcohol may influence DNA methylation levels via influencing one-carbon metabolism (46), and folate is an important methyl donor. On the other hand, other studies have not specifically linked alcohol intake to CIMP in CRC (47–49). Of the four other tagSNPs (ADH1B rs4147536 and rs3811802 and ADH1C rs283415 and rs4147542) that were associated with CRC risk in drinkers—also suggesting interplay between alcohol intake, ADH1B and ADH1C, and CRC risk—rs283415 is in strong LD with the commonly investigated rs698, for which functional evidence is available. ADH1C rs698 C-allele carriers who also carry the rs1693482 A-allele, encoding Ile350Val and Arg272Gln substitutions, respectively (50), have a ~2.5 times slower alcohol metabolizing rate (51,52) and were found to be at an increased risk of alcohol dependence in Asian populations (7). However, the evidence linking rs698 to cancer was judged inconclusive by IARC due to the small number of studies (1). A meta-analysis of 35 case–control studies comparing rs698 slow with faster alcohol metabolizers found an association with the risk of cancer overall in African and Asian but not European populations (53). This suggests ethnicity is an important factor to take into account. For example, in Caucasian populations there is uncertainty around whether slow or fast alcohol metabolizers are at an increased CRC risk. Although Caucasians carry ADH alleles affecting ethanol oxidation, they lack ADH alleles causing very fast oxidation of ethanol and also lack ALDH alleles causing slow oxidation of acetaldehyde. As such, conflicting results may have emerged from Caucasian studies on rs698 (54,55) in the absence of data on how much ethanol or acetaldehyde accumulates in the colorectum, as explained in the introduction. Strengths of the present study include the population-based prospective design and long follow-up, yielding large case numbers and making selection and information bias unlikely. In addition, the NLCS contains information on alcohol intake at baseline as well as 5 years before baseline, allowing us to investigate whether drinking patterns or fluctuations in alcohol intake affected the studied associations. Information on potential confounders was based on a single baseline measurement, and although changes over time cannot be excluded, the NLCS population has been found stable in its dietary habits (16). Importantly, the elaborate available baseline information enabled us to adjust for a large set of relevant confounders. This is essential considering that individuals who consume higher levels of alcohol may in general have an unhealthier lifestyle than those who have lower intake levels. Furthermore, the high genotyping quality may also be considered as a strength. In conclusion, as opposed to men who showed increased CRC risks across subsites and alcohol intake levels, alcohol intake only increased colon cancer risk in women and only at heavy intake levels. ADH1B rs3811802 and ADH1C rs4147542 modified the alcohol–CRC association in women. These data indicate that alcohol may be a particularly important CRC risk factor in specific genetic subgroups. Previous literature indicates a functional role of rs4147542, supporting our finding of an effect of this variant on alcohol-associated colorectal carcinogenesis and strengthening our confidence in covering relevant genetic variation in ADH1B and ADH1C. Supplementary material Supplementary data are available at Carcinogenesis online. Funding This work was supported by the European Foundation for Alcohol Research (EA 14 39 to C.C.J.M.S., P.v.d.B. and M.P.W.), the Biobanking and Biomolecular Research Infrastructure Netherlands (to M.P.W.) and the Health Foundation Limburg (to M.P.W.). Abbreviations ADH alcohol dehydrogenase ALDH aldehyde dehydrogenase CI confidence interval CRC colorectal cancer FFQ food frequency questionnaire FDR false discovery rate FFQ food frequency questionnaire NLCS Netherlands Cohort Study SNP single nucleotide polymorphism Acknowledgements We are indebted to the participants of this study and wish to thank the Netherlands Cancer Registry and the Netherlands nationwide registry of pathology (PALGA). We also thank Drs. A. Volovics, and A. Kester for statistical advice; S. van de Crommert, H. Brants, J. Nelissen, C. de Zwart, M. Moll, W. van Dijk and A. Pisters for data management; H. Hoofs, H. van Montfort, T. van Moergastel, L. van den Bosch, R. Schmeitz and J. Berben for programming assistance; L. Jonkers, J. Goessens, K. Lemmens and S. Lumeij for the laboratory work involved; and the Biobank Maastricht UMC+ for sample storage. Conflict of Interest Statement: None declared. References 1. International Agency for Research on Cancer. ( 2012) IARC Monographs on the Evaluation of Carcinogenic Risks to Humans: Personal Habits and Indoor Combustions: Consumption of Alcoholic Beverages . International Agency for Research on Cancer, Lyon, France. PubMed PubMed  2. WCRF|Continuous Update Project. Alcohol and Cancer. http://www.wcrf.org/int/cancer-facts-figures/link-between-lifestyle-cancer-risk/alcohol-cancer ( 18 December 2017, date last accessed). 3. Rehm, J.et al.  ( 2003) The global distribution of average volume of alcohol consumption and patterns of drinking. Eur. Addict. Res ., 9, 147– 156. Google Scholar CrossRef Search ADS PubMed  4. WCRF|Continuous Update Project. Colorectal cancer . http://www.wcrf.org/sites/default/files/Colorectal-Cancer-2011-Report.pdf ( 18 December 2017, date last accessed). 5. Fedirko, V.et al.  ( 2011) Alcohol drinking and colorectal cancer risk: an overall and dose-response meta-analysis of published studies. Ann. Oncol ., 22, 1958– 1972. Google Scholar CrossRef Search ADS PubMed  6. Baan, R.et al.  ; WHO International Agency for Research on Cancer Monograph Working Group. ( 2007) Carcinogenicity of alcoholic beverages. Lancet. Oncol ., 8, 292– 293. Google Scholar CrossRef Search ADS PubMed  7. Hurley, T.D.et al.  ( 2012) Genes encoding enzymes involved in ethanol metabolism. Alcohol Res ., 34, 339– 344. Google Scholar PubMed  8. Seitz, H.K.et al.  ( 2007) Molecular mechanisms of alcohol-mediated carcinogenesis. Nat. Rev. Cancer , 7, 599– 612. Google Scholar CrossRef Search ADS PubMed  9. Zerbino, D.Ret al.  (2018) Ensembl 2018. Nucleic Acids Res., 46, D754–D761. doi:10.1093/nar/gkx1098. 10. Ferrari, P.et al.  ( 2012) Alcohol dehydrogenase and aldehyde dehydrogenase gene polymorphisms, alcohol intake and the risk of colorectal cancer in the European Prospective Investigation into Cancer and Nutrition study. Eur. J. Clin. Nutr ., 66, 1303– 1308. Google Scholar CrossRef Search ADS PubMed  11. Martínez, C.et al.  ( 2010) Variability in ethanol biodisposition in whites is modulated by polymorphisms in the ADH1B and ADH1C genes. Hepatology , 51, 491– 500. Google Scholar CrossRef Search ADS PubMed  12. Jokelainen, K.et al.  ( 1994) In vitro acetaldehyde formation by human colonic bacteria. Gut , 35, 1271– 1274. Google Scholar CrossRef Search ADS PubMed  13. Jokelainen, K.et al.  ( 1996) In vitro alcohol dehydrogenase-mediated acetaldehyde production by aerobic bacteria representing the normal colonic flora in man. Alcohol. Clin. Exp. Res ., 20, 967– 972. Google Scholar CrossRef Search ADS PubMed  14. Pestalozzi, D.M.et al.  ( 1983) Immunohistochemical localization of alcohol dehydrogenase in the human gastrointestinal tract. Gastroenterology , 85, 1011– 1016. Google Scholar PubMed  15. Seitz, H.K.et al.  ( 1996) Alcohol dehydrogenase in the human colon and rectum. Digestion , 57, 105– 108. Google Scholar CrossRef Search ADS PubMed  16. van den Brandt, P.A.et al.  ( 1990) A large-scale prospective cohort study on diet and cancer in The Netherlands. J. Clin. Epidemiol ., 43, 285– 295. Google Scholar CrossRef Search ADS PubMed  17. Bongaerts, B.W.et al.  ( 2008) Alcohol consumption, type of alcoholic beverage and risk of colorectal cancer at specific subsites. Int. J. Cancer , 123, 2411– 2417. Google Scholar CrossRef Search ADS PubMed  18. Bongaerts, B.W.et al.  ( 2011) Alcohol consumption, alcohol dehydrogenase 1C (ADH1C) genotype, and risk of colorectal cancer in the Netherlands Cohort Study on diet and cancer. Alcohol , 45, 217– 225. Google Scholar CrossRef Search ADS PubMed  19. Gezondheidsraad( 2015) Richtlijnen goede voeding 2015 . Den Haag: Gezondheidsraad: publicatienr. 2015/24. https://www.gezondheidsraad.nl/sites/default/files/201524_richtlijnen_goede_voeding_2015.pdf ( 18 December 2017, date last accessed). 20. Prentice, R.L. ( 1986) A case-cohort design for epidemiologic cohort studies and disease prevention trials. Biometrika , 73, 1– 11. Google Scholar CrossRef Search ADS   21. Casparie, M.et al.  ( 2007) Pathology databanking and biobanking in The Netherlands, a central role for PALGA, the nationwide histopathology and cytopathology data network and archive. Cell. Oncol ., 29, 19– 24. Google Scholar PubMed  22. Van den Brandt, P.A.et al.  ( 1990) Development of a record linkage protocol for use in the Dutch Cancer Registry for Epidemiological Research. Int. J. Epidemiol ., 19, 553– 558. Google Scholar CrossRef Search ADS PubMed  23. Goldbohm, R.A.et al.  ( 1994) Estimation of the coverage of Dutch municipalities by cancer registries and PALGA based on hospital discharge data. Tijdschr Soc Gezondheidsz , 72, 80– 4. 24. van Breda, S.G.et al.  ( 2007) Toenails: an easily accessible and long-term stable source of DNA for genetic analyses in large-scale epidemiological studies. Clin. Chem ., 53, 1168– 1170. Google Scholar CrossRef Search ADS PubMed  25. Hogervorst, J.G.et al.  ( 2014) DNA from nails for genetic analyses in large-scale epidemiologic studies. Cancer Epidemiol. Biomarkers Prev ., 23, 2703– 2712. Google Scholar CrossRef Search ADS PubMed  26. Goldbohm, R.A.et al.  ( 1994) Validation of a dietary questionnaire used in a large-scale prospective cohort study on diet and cancer. Eur. J. Clin. Nutr ., 48, 253– 265. Google Scholar PubMed  27. Goldbohm, R.A.et al.  ( 1995) Reproducibility of a food frequency questionnaire and stability of dietary habits determined from five annually repeated measurements. Eur. J. Clin. Nutr ., 49, 420– 429. Google Scholar PubMed  28. de Bakker, P.I.et al.  ( 2005) Efficiency and power in genetic association studies. Nat. Genet ., 37, 1217– 1223. Google Scholar CrossRef Search ADS PubMed  29. Simons, C.C.et al.  ( 2013) Physical activity, occupational sitting time, and colorectal cancer risk in the Netherlands cohort study. Am. J. Epidemiol ., 177, 514– 530. Google Scholar CrossRef Search ADS PubMed  30. Benjamini, Y.et al.  ( 1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B Stat. Methodol ., 57, 289– 300. 31. Benjamini, Y.et al.  ( 2001) Controlling the false discovery rate in behavior genetics research. Behav. Brain Res ., 125, 279– 284. Google Scholar CrossRef Search ADS PubMed  32. Smith, N.L.et al.  ( 2007) Association of genetic variations with nonfatal venous thrombosis in postmenopausal women. JAMA , 297, 489– 498. Google Scholar CrossRef Search ADS PubMed  33. Cao, Y.et al.  ( 2015) Light to moderate intake of alcohol, drinking patterns, and risk of cancer: results from two prospective US cohort studies. BMJ , 351, h4238. Google Scholar CrossRef Search ADS PubMed  34. Oneta, C.M.et al.  ( 1998) First pass metabolism of ethanol is strikingly influenced by the speed of gastric emptying. Gut , 43, 612– 619. Google Scholar CrossRef Search ADS PubMed  35. Erol, A.et al.  ( 2015) Sex and gender-related differences in alcohol use and its consequences: contemporary knowledge and future research considerations. Drug Alcohol Depend ., 156, 1– 13. Google Scholar CrossRef Search ADS PubMed  36. Lin, J.H.et al.  ( 2013) Association between sex hormones and colorectal cancer risk in men and women. Clin. Gastroenterol. Hepatol ., 11, 419– 424.e1. Google Scholar CrossRef Search ADS PubMed  37. Roshan, M.H.et al.  ( 2016) The role of testosterone in colorectal carcinoma: pathomechanisms and open questions. EPMA J ., 7, 22. Google Scholar CrossRef Search ADS PubMed  38. Klatsky, A.L.et al.  ( 2014) Moderate alcohol intake and cancer: the role of underreporting. Cancer Causes Control , 25, 693– 699. Google Scholar CrossRef Search ADS PubMed  39. Liang, P.S.et al.  ( 2009) Cigarette smoking and colorectal cancer incidence and mortality: systematic review and meta-analysis. Int. J. Cancer , 124, 2406– 2415. Google Scholar CrossRef Search ADS PubMed  40. Simonds, N.I.et al.  ( 2016) Review of the gene-environment interaction literature in cancer: what do we know? Genet. Epidemiol ., 40, 356– 365. Google Scholar CrossRef Search ADS PubMed  41. GTEx Portal. https://gtexportal.org/home/eqtls/bySnp?snpId=rs4147542&tissueName=All ( 18 December 2017, date last accessed). 42. Zhang, H.et al.  ( 2014) Identification of methylation quantitative trait loci (mQTLs) influencing promoter DNA methylation of alcohol dependence risk genes. Hum. Genet ., 133, 1093– 1104. Google Scholar CrossRef Search ADS PubMed  43. Hawkins, N.et al.  ( 2002) CpG island methylation in sporadic colorectal cancers and its relationship to microsatellite instability. Gastroenterology , 122, 1376– 1387. Google Scholar CrossRef Search ADS PubMed  44. Weisenberger, D.J.et al.  ( 2006) CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer. Nat. Genet ., 38, 787– 793. Google Scholar CrossRef Search ADS PubMed  45. Curtin, K.et al.  ( 2007) Genetic polymorphisms in one-carbon metabolism: associations with CpG island methylator phenotype (CIMP) in colon cancer and the modifying effects of diet. Carcinogenesis , 28, 1672– 1679. Google Scholar CrossRef Search ADS PubMed  46. Varela-Rey, M.et al.  ( 2013) Alcohol, DNA methylation, and cancer. Alcohol Res ., 35, 25– 35. Google Scholar PubMed  47. Razzak, A.A.et al.  ( 2011) Alcohol intake and colorectal cancer risk by molecularly defined subtypes in a prospective study of older women. Cancer Prev. Res. (Phila) ., 4, 2035– 2043. Google Scholar CrossRef Search ADS PubMed  48. Schernhammer, E.S.et al.  ( 2011) B vitamins, methionine and alcohol intake and risk of colon cancer in relation to BRAF mutation and CpG island methylator phenotype (CIMP). PLoS One , 6, e21102. Google Scholar CrossRef Search ADS PubMed  49. Slattery, M.L.et al.  ( 2007) Diet and lifestyle factor associations with CpG island methylator phenotype and BRAF mutations in colon cancer. Int. J. Cancer , 120, 656– 663. Google Scholar CrossRef Search ADS PubMed  50. Osier, M.V.et al.  ( 2002) A proline-threonine substitution in codon 351 of ADH1C is common in Native Americans. Alcohol. Clin. Exp. Res ., 26, 1759– 1763. Google Scholar CrossRef Search ADS PubMed  51. Bosron, W.F.et al.  ( 1986) Genetic polymorphism of human liver alcohol and aldehyde dehydrogenases, and their relationship to alcohol metabolism and alcoholism. Hepatology , 6, 502– 510. Google Scholar CrossRef Search ADS PubMed  52. Höög, J.O.et al.  ( 1986) The gamma 1 and gamma 2 subunits of human liver alcohol dehydrogenase. cDNA structures, two amino acid replacements, and compatibility with changes in the enzymatic properties. Eur. J. Biochem ., 159, 215– 218. Google Scholar CrossRef Search ADS PubMed  53. Xue, Y.et al.  ( 2012) ADH1C Ile350Val polymorphism and cancer risk: evidence from 35 case-control studies. PLoS One , 7, e37227. Google Scholar CrossRef Search ADS PubMed  54. Giovannucci, E.et al.  ( 2003) Methylenetetrahydrofolate reductase, alcohol dehydrogenase, diet, and risk of colorectal adenomas. Cancer Epidemiol. Biomarkers Prev ., 12, 970– 979. Google Scholar PubMed  55. Homann, N.et al.  ( 2009) Alcohol and colorectal cancer: the role of alcohol dehydrogenase 1C polymorphism. Alcohol. Clin. Exp. Res ., 33, 551– 556. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Carcinogenesis Oxford University Press

Alcohol intake, ADH1B and ADH1C genotypes, and the risk of colorectal cancer by sex and subsite in the Netherlands Cohort Study

Loading next page...
 
/lp/ou_press/alcohol-intake-adh1b-and-adh1c-genotypes-and-the-risk-of-colorectal-s77lqaxmGi
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com.
ISSN
0143-3334
eISSN
1460-2180
D.O.I.
10.1093/carcin/bgy011
Publisher site
See Article on Publisher Site

Abstract

Abstract The alcohol–colorectal cancer (CRC) association may differ by sex and ADH1B and ADH1C genotypes. ADH enzymes oxidize ethanol to acetaldehyde, both of which are human carcinogens. The Netherlands Cohort Study includes 120 852 participants, aged 55–69 years at baseline (1986), and has 20.3 years follow-up (case-cohort: nsubcohort = 4774; ncases = 4597). The baseline questionnaire included questions on alcohol intake at baseline and 5 years before. Using toenail DNA, available for ~75% of the cohort, we successfully genotyped six ADH1B and six ADH1C SNPs (nsubcohort = 3897; ncases = 3558). Sex- and subsite-specific Cox hazard ratios and 95% confidence intervals for CRC were estimated comparing alcohol categories, genotypes within drinkers and alcohol categories within genotype strata. We used a dominant genetic model and adjusted for multiple testing. Alcohol intake increased CRC risk in both sexes, though in women only in the (proximal) colon when in excess of 30 g/day. In male drinkers, ADH1B rs4147536 increased (distal) colon cancer risk. In female drinkers, ADH1C rs283415 increased proximal colon cancer risk. ADH1B rs3811802 and ADH1C rs4147542 decreased CRC risk in heavy (>30 g/day) and stable drinkers (compared to 5 years before baseline), respectively. Rs3811802 and rs4147542 significantly modified the alcohol-colon cancer association in women (Pfor interaction = 0.004 and 0.02, respectively). A difference in associations between genotype strata was generally clearer in men than women. In conclusion, men showed increased CRC risks across subsites and alcohol intake levels, while only colon cancer risk was increased in women at heavy intake levels. ADH1B rs3811802 and ADH1C rs4147542 significantly modified the alcohol–colon cancer association in women. Introduction Alcohol intake is a known risk factor for colorectal cancer (CRC), both the colon and rectum, and various other cancers including cancers of the oral cavity, pharynx, larynx, esophagus, liver and female breast (1,2). Cancer risk has been shown to increase as the volume of alcohol consumed increases (3). Sex differences may exist, with CRC risk being more strongly affected by alcohol intake in men than women (4) and a dose–response relation being less apparent in women than men (5). While this may be due to a restriction of range effect because women consume less alcohol than men, part of these differences may also simply be due to the scarcity of studies on the alcohol–CRC association in women (4). The alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) gene families encode for enzymes responsible for the breakdown of alcohol in the body. Ethanol is oxidized by ADH to acetaldehyde, which is in turn oxidized by ALDH to acetate (6,7). The formation of acetaldehyde starts in the mouth and continues along the digestive tract, with the main production of acetaldehyde occurring in the liver and the gut (1). Both ethanol and acetaldehyde have been classified as human carcinogens by the International Agency for Research on Cancer (IARC) (6). Genetic variants [i.e. single nucleotide polymorphisms (SNPs)] in ADH and ALDH genes have been shown to affect enzyme activities resulting in a slower or faster conversion of ethanol to acetaldehyde and acetaldehyde to acetate (7,8). The ALDH2 rs671 gene variant strongly determines the slow rate at which most of the ethanol-derived acetaldehyde is oxidized, resulting in acetaldehyde accumulation when carrying the minor allele, especially when also carrying ADH genotypes associated with fast ethanol to acetaldehyde conversion, e.g. for ADH1B rs1229984. This results in the facial flushing syndrome known in Asian populations (7). In Caucasians, however, the rs671 allele causing slow oxidation of acetaldehyde is largely absent (9). ADH1B and ADH1C, on the other hand, have functional polymorphisms in Caucasian populations (7), but most previous epidemiological studies on SNPs in these genes and the risk of CRC had low sample sizes (10,11). In the absence of ALDH alleles causing slow oxidation of acetaldehyde, increased cancer risks may be explained by fast conversion of ethanol to acetaldehyde in the liver, resulting in higher systemic levels of acetaldehyde, potentially reaching the colorectum. Alternatively, slow conversion of ethanol to acetaldehyde in the liver results in ethanol circulating the blood for longer periods of time. Circulating ethanol may expose the colorectum to locally formed levels of acetaldehyde, depending on ADH activity in the colonic mucosa and conversion by intestinal bacteria (12–15). Therefore, in the absence of data on how much ethanol or acetaldehyde accumulates in the colorectum, an association between alcohol intake and CRC risk in both slow and fast ethanol metabolizers is conceivable. The objective of this study was twofold in that we first aimed to investigate the alcohol–CRC association by sex and subsite, specifically addressing the limited evidence in women, and secondly we aimed to investigate whether this association was modified by genetic variants in ADH1B and ADH1C. ADH1B and ADH1C variants were also studied in relation to CRC risk by sex and subsite in drinkers, where an effect may be expected. The study setting was a prospective cohort including 120 852 participants, who were followed up for 20.3 years (16). A former study in this cohort after 13.3 years follow-up found no evidence of an influence of alcohol on CRC risk, overall and by subsite, though the number of cases among the heavy drinkers was rather limited (i.e. ≥30 g of alcohol per day), hampering sex-specific analyses (17). Another study in this cohort, investigating the alcohol–CRC association by ADH1C rs698 genotype strata after 7.3 years follow-up found no apparent evidence of effect modification, but the power was limited there as well (18). Therefore, we reinvestigated the alcohol–CRC association by sex and subsite and possible effect modification by rs698 and other ADH1B and ADH1C tagSNPs after 20.3 years of follow-up. Recently, Dutch recommendations with respect to alcohol intake for cancer prevention have been modified: from a maximum of two drinks daily for men and one drink daily for women to no alcohol intake for both sexes (2,19), and if consuming alcohol, no more than one drink daily (19). This revised recommendation may be particularly important for specific subgroups with an unfavorable genetic background. Materials and methods Design and study population The Netherlands Cohort Study (NLCS) is a prospective cohort study which was initiated in 1986 and consists of 120 852 men and women who were aged 55–69 years old at baseline (16). Study participants completed a baseline self-administered questionnaire on dietary habits, lifestyle factors and other risk factors for cancer. For efficiency reasons, the NLCS uses a case-cohort approach in which cases are enumerated from the entire cohort and the person-time at risk is estimated from a subcohort (20). This subcohort, consisting of 5000 participants, was randomly selected immediately after baseline, independent of any exposure. After exclusion of participants who reported a history of cancer (other than skin cancer) at baseline, 4774 subcohort members were left. Follow-up of vital status and migration for these participants was done through the Central Bureau of Genealogy and the municipal population registries (>99.9% completeness). Follow-up for incident cancer cases was performed by record linkage to the Dutch Cancer Registry and PALGA (the Netherlands pathology database) (21,22) (>96% completeness) (23). After 20.3 years follow-up, 4597 CRC cases (ICD-O-3 codes C18-C20) were identified from the original cohort. The NLCS was approved by the institutional review boards of the Netherlands Organisation for Applied Scientific Research TNO (Zeist) and Maastricht University (Maastricht). Exposure assessment The baseline questionnaire included a 150-item semi-quantitative food frequency questionnaire (FFQ) containing questions on diet and alcohol intake. In addition to the questionnaire, participants in the NLCS were asked to return toenail clippings. Roughly 90 000 (~75%) of NLCS participants provided toenail clippings which were used as a source of DNA for genotyping (24,25). Using 20.3 years follow-up in the NLCS, DNA samples were available for 3558 CRC cases and 3897 subcohort members. Alcohol intake and covariates Alcohol intake during the year preceding the start of the study was measured by questions on six different types of alcohol: beer; red wine; white wine; sherry and other fortified wines; liquor types containing on average 16% alcohol; (Dutch) gin, brandy and whiskey. Participants were asked about the number of glasses they consumed during each drinking session and their usual frequency of alcohol drinking. Additionally, for the categories ‘beer’ and ‘other alcoholic beverages’, participants were asked to recall if they drank more, less, or the same amount of alcohol 5 years before baseline. The total amount of daily alcohol intake (g/day) was calculated using the information about how often the participants drank alcohol, the number of glasses they consumed during each drinking session and the types of alcohol they drank (i.e. their alcoholic content). We defined two alcohol categories: light-moderate alcohol intake as drinking >0 to <30 g of alcohol per day (>0 to <3 glasses of alcohol per day) and heavy alcohol intake as drinking 30 or more grams of alcohol per day (≥3 glasses of alcohol per day). Information on other covariates that were considered potential confounders on the basis of previous research was also available from the baseline questionnaire. The FFQ was validated against a 9-day diet record (26) and was tested for reproducibility (27). The adjusted Spearman correlation coefficient between mean daily alcohol intake assessed by the FFQ and that estimated from the 9-day diet record was 0.89 for all participants and 0.85 for users of alcoholic beverages. The absolute amount of alcohol reported in the FFQ by users of alcoholic beverages was, on average, 86% of that reported in the 9-day diet record (26). Selection and genotyping of tagging SNPs Tagging SNPs (tagSNPs) within ADH1B and ADH1C (including 5 kb up- and downstream) were selected as to potentially cover all of the genetic variation in these genes with a minor allele frequency of 5% or higher. In total, 13 tagSNPs were identified using the HapMap CEU (Utah Residents with Northern and Western European Ancestry) population, an r2 threshold of 0.8 and aggressive tagging. Seven of these (i.e. rs1159918, rs2075633, rs1693439, rs9307239, rs4147536, rs3811802, rs17033) represented 84% of the genetic variation in ADH1B and six (i.e. rs698, rs1662033, rs3114046, rs4147542, rs283415, rs4699741) represented 96% of the genetic variation in ADH1C (28). SNPs were genotyped using 50 ng of toenail DNA per participant, which was carried out using the iPLEXTM assay for the MassARRAY® system (Agena Bioscience GmbH, Hamburg, Germany). Samples had a mean call rate of 97.1% (as based on the 13 SNPs in ADH1B and ADH1C studied here and 10 other SNPs that were included in the assay). SNP call rates were 94% or higher, except for rs4147542, which had a SNP call rate of 87%. Two SNPs, ADH1C rs4699741 and ADH1C rs9307239, violated Hardy–Weinberg equilibrium. When using a significance threshold of 0.05, one in twenty SNPs may be expected to show a violation on the basis of chance alone. Although two SNPs exceed this expectation by chance and we cannot check conditions needed for Hardy–Weinberg, e.g. random mating, all SNPs were genotyped using a single assay, which makes it unlikely that these violations represent genotyping errors. Therefore, we conservatively refrained from excluding these SNPs from the analysis. Genotyping for ADH1B rs17033 was unsuccessful (i.e. only the T allele was amplified) and therefore not included in our analyses as originally intended. Consequently, we used six ADH1B tagSNPs covering 76% of the genetic variation in ADH1B and six ADH1C tagSNPs covering 96% of the genetic variation in ADH1C. Statistical analysis Statistical analyses were carried out using Cox regression to calculate hazard ratios (HR) and corresponding 95% confidence intervals (95% CI) for CRC by sex and subsite. Standard errors were estimated using the Huber–White sandwich estimator to account for additional variance introduced by sampling the subcohort from the entire cohort. ADH1B and ADH1C genotypes were defined according to a dominant model for reasons of power. Categories of total alcohol intake (0.1–29 and ≥30 g/day) were compared relative to abstaining (0 g/day). Trends were evaluated by including categorical variables as continuous variables in the Cox regression model. The proportional hazards assumption was tested using the scaled Schoenfeld residuals and by visually inspecting the −log–log transformed hazard curves. Multiplicative interactions were tested using the Wald test. All tests (two-tailed) were performed using Stata (version 14) and differences were regarded as statistically significant at P < 0.05. Marginal effects of alcohol intake on CRC Multivariable-adjusted models were used to study the alcohol–CRC associations. The covariates included were either a priori-selected risk factors based on the literature or variables that changed the HRs by at least 10% (using a backwards stepwise procedure). This resulted in the following confounder set: age (years), BMI (kg/m2), smoking (never/ex/current), family history of CRC (yes/no), meat intake (g/day), processed meat intake (g/day), folate intake (µg/day) and physical activity based on baseline non-occupational physical activity (min/day) (29). Participants with incomplete or inconsistent questionnaires and missing information on alcohol intake and/or the predefined confounding factors were excluded from the analysis, leaving 4125 subcohort members and 3996 CRC cases (see Figure 1). Figure 1. View largeDownload slide Flow diagram of available subcohort members and colorectal cancer cases, Netherlands Cohort Study, 1986–2006. FU, follow-up. AAnalysis 1 is on the marginal effects of alcohol intake on CRC. BAnalysis 2 is on the associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers. CAnalysis 3 is on effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs. Figure 1. View largeDownload slide Flow diagram of available subcohort members and colorectal cancer cases, Netherlands Cohort Study, 1986–2006. FU, follow-up. AAnalysis 1 is on the marginal effects of alcohol intake on CRC. BAnalysis 2 is on the associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers. CAnalysis 3 is on effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs. Associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers We studied individual SNPs in relation to CRC risk in drinkers. Although small amounts of ethanol are produced endogenously, especially in the gastrointestinal tract (1), an effect of SNPs in alcohol-metabolizing genes may be expected in drinkers foremost as this is the group where the substrate (alcohol) is available. We conservatively refrained from adjusting for factors other than age because it is unlikely that ADH1B and ADH1C genotypes are influenced by CRC risk factors in lifestyle and diet. Participants were excluded from the analysis if no toenail DNA sample was available, the sample call rate was less than 90%, the baseline questionnaire was incomplete or inconsistent, or information on alcohol intake was missing (see Figure 1).This resulted in 2526 subcohort members and 2491 CRC cases. Effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs Multivariable-adjusted models were used to study effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs, using the confounder set as described for the alcohol–CRC analyses. After excluding participants without available DNA samples, with a sample call rate of less than 90%, with incomplete or inconsistent questionnaires and without complete information on alcohol intake and/or the predefined confounding factors, 3150 subcohort members and 2985 CRC cases were left for analysis (see Figure 1). Multiple testing Because multiple tests were conducted within each gene, we applied the false discovery rate (FDR) control method of Benjamini and Hochberg (30,31) to address the issue of multiple testing. To accomplish this, P-values calculated from our analyses were ranked in ascending order. Gene- and endpoint-specific Benjamini adjusted P-values were calculated by dividing the P-value rank order by the total number of P-values and then multiplying this number by the FDR [i.e. the recommended 20% (32)]. If the original P-value was less than 0.05 and fell below the adjusted P-value, it was considered significant. Sensitivity analyses Drinking patterns over a longer duration of time may influence CRC risk differently or more profoundly than when evaluated on a single time point. For instance, stronger alcohol–CRC associations may be expected in those with a relatively constant long-term exposure to alcohol. As the NLCS has data available on alcohol intake 5 years before baseline, we conducted sensitivity analyses using these data. This included restricting the analyses on alcohol–CRC associations and effect modification to participants who reported to have had the same alcohol intake 5 years before baseline, which included abstainers on both occasions (i.e. the stable subgroup). For the SNP-CRC analyses in drinkers, this included restriction to those drinking equal amounts of alcohol 5 years before baseline (i.e. the stable drinkers). We also evaluated whether there may be a threshold level of alcohol intake at which individual SNPs start to influence CRC risk by stratifying SNP–CRC associations on alcohol intake level (light-moderate and heavy). Furthermore, because changes in reported alcohol intake may indicate underlying reasons such as health issues or exposure misclassification, possibly due to underreporting, we repeated the alcohol–CRC analyses restricting once to baseline drinkers who reported drinking less alcohol 5 years before baseline and once to baseline drinkers who reported drinking more alcohol 5 years before baseline. Finally, we checked for the risk of protopathic bias by excluding the first 2 years of follow-up (with no essential changes in results), as this may be especially likely when investigating alcohol intake in relation to cancer risk. Results Baseline characteristics Table 1 shows the baseline characteristics of subcohort members and CRC cases. As regards alcohol intake, men were less often abstainers as compared to women and men were more likely to consume higher levels of alcohol than women, especially male CRC cases. Table 1. Distribution of potential confounders and alcohol intake among subcohort members and CRC cases in the NLCS, 1986–2006   Male subcohort  Male CRC cases  Female subcohort  Female CRC cases    N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  Age (years)    61.4 (4.2)    61.8 (4.2)    61.5 (4.3)    62.1 (4.1)  Alcohol intake (g/day)   0  329 (14.5)    303 (12.1)    731 (32.7)    599 (31.7)     0.1–4  479 (21.1)    549 (22.0)    806 (36.0)    692 (36.6)     5–14  621 (27.3)    630 (25.2)    417 (18.6)    348 (18.4)     15–29  505 (22.2)    593 (23.7)    207 (9.3)    168 (8.9)     ≥30  339 (14.9)    426 (17.0)    77 (3.4)    84 (4.4)    BMI (kg/m2)    25.0 (2.6)    25.3 (2.7)    25.1 (3.6)    25.1 (3.5)  Smoking   Never  300 (12.9)    313 (12.3)    1431 (58.9)    1190 (58.7)     Ex  1175 (50.4)    1463 (57.3)    491 (20.2)    438 (21.6)     Current  856 (36.7)    777 (30.4)    509 (20.9)    398 (19.6)    Family history of CRC   Yes  118 (5.1)    219 (8.6)    134 (5.5)    189 (9.3)     No  2213 (94.9)    2336 (91.4)    2298 (94.5)    1840 (90.7)    Meat intake (g/day)    104.9 (44.1)    105.3 (43.2)    92.2 (41.1)    90.6 (40.9)  Processed meat intake (g/day)    15.6 (16.9)    16.3 (16.8)    10.3 (11.9)    10.2 (11.3)  Folate intake (µg/day)    222 (77)    219 (72)    195 (71)    194 (71)  Non-occupational physical activity (min/day)   ≤30  447 (19.5)    431 (17.1)    639 (26.8)    596 (29.9)     >30–60  710 (30.9)    776 (30.8)    737 (30.9)    598 (30.0)     >60–90  419 (18.3)    497 (19.7)    518 (21.7)    451 (22.6)     >90  719 (31.3)    815 (32.4)    490 (20.6)    351 (17.6)      Male subcohort  Male CRC cases  Female subcohort  Female CRC cases    N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  N (%)  Mean (SD)  Age (years)    61.4 (4.2)    61.8 (4.2)    61.5 (4.3)    62.1 (4.1)  Alcohol intake (g/day)   0  329 (14.5)    303 (12.1)    731 (32.7)    599 (31.7)     0.1–4  479 (21.1)    549 (22.0)    806 (36.0)    692 (36.6)     5–14  621 (27.3)    630 (25.2)    417 (18.6)    348 (18.4)     15–29  505 (22.2)    593 (23.7)    207 (9.3)    168 (8.9)     ≥30  339 (14.9)    426 (17.0)    77 (3.4)    84 (4.4)    BMI (kg/m2)    25.0 (2.6)    25.3 (2.7)    25.1 (3.6)    25.1 (3.5)  Smoking   Never  300 (12.9)    313 (12.3)    1431 (58.9)    1190 (58.7)     Ex  1175 (50.4)    1463 (57.3)    491 (20.2)    438 (21.6)     Current  856 (36.7)    777 (30.4)    509 (20.9)    398 (19.6)    Family history of CRC   Yes  118 (5.1)    219 (8.6)    134 (5.5)    189 (9.3)     No  2213 (94.9)    2336 (91.4)    2298 (94.5)    1840 (90.7)    Meat intake (g/day)    104.9 (44.1)    105.3 (43.2)    92.2 (41.1)    90.6 (40.9)  Processed meat intake (g/day)    15.6 (16.9)    16.3 (16.8)    10.3 (11.9)    10.2 (11.3)  Folate intake (µg/day)    222 (77)    219 (72)    195 (71)    194 (71)  Non-occupational physical activity (min/day)   ≤30  447 (19.5)    431 (17.1)    639 (26.8)    596 (29.9)     >30–60  710 (30.9)    776 (30.8)    737 (30.9)    598 (30.0)     >60–90  419 (18.3)    497 (19.7)    518 (21.7)    451 (22.6)     >90  719 (31.3)    815 (32.4)    490 (20.6)    351 (17.6)    View Large Marginal effects of alcohol intake on CRC Alcohol intake was positively associated with CRC risk in both men and women (Table 2). In women, however, only colon cancer risks, in particular proximal colon cancer risk, were increased, but not until alcohol intake exceeded 30 g/day [HRheavy drinkers versus abstainers (95% CI) = 1.52 (1.03–2.24), 1.70 (1.09–2.66) and 1.25 (0.71–2.20) for colon, proximal colon and distal colon cancer, respectively]. In men, colon and rectal cancer risks were both increased and there were also non-significantly increased risks in men who consumed a light-moderate amount of alcohol versus abstainers [HRlight-moderate drinkers versus abstainers (95% CI) = 1.21 (0.97–1.50), 1.23 (0.93–1.63), 1.23 (0.93–1.63) and 1.10 (0.82–1.48) for colon, proximal colon, distal colon and rectal cancer, respectively]. In addition, there was a statistically significant positive linear trend across alcohol intake categories in men, except for proximal colon cancer. When restricting to the stable subgroup, more pronounced associations were observed between alcohol intake and CRC risk in men, while statistically significant associations were no longer observed in women, but this may be explained by limited power. Remarkably, male heavy drinkers as compared to light-moderate drinkers reporting more alcohol intake 5 years before baseline had decreased CRC risks across subsites. Possibly, men who still reported to be heavy drinkers at baseline endure alcohol better than light-moderate drinkers who reported more alcohol intake 5 years before baseline. In the subanalysis in male drinkers reporting less alcohol intake 5 years before baseline, HRs were around the null for heavy drinkers as compared to light-moderate drinkers. In female drinkers who reported more and those who reported less alcohol intake 5 years before baseline, (non-significantly) increased risks of CRC were observed for heavy drinkers as compared to light-moderate drinkers, though the power was limited. Table 2. Hazard ratios and 95% confidence intervals for the association between alcohol intake and colorectal cancer risk in men and women stratified by subsite in the NLCS, 1986–2006   PT at risk  Colorectum  Colon  Proximal Colon  Distal Colon  Rectum      N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  Men  Alcohol intake   Abstainers (0 g/day)  4548  267  1  (ref)  176  1  (ref)  81  1  (ref)  88  1  (ref)  69  1  (ref)   Light-moderate (0.1–29 g/day)  23 074  1625  1.22  (1.00, 1.48)  1050  1.21  (0.97, 1.50)  485  1.23  (0.93, 1.63)  538  1.23  (0.93, 1.63)  388  1.10  (0.82, 1.48)   Heavy (≥30 g/day)  4852  394  1.40  (1.10, 1.78)  246  1.35  (1.03, 1.77)  109  1.30  (0.92, 1.83)  126  1.42  (1.01, 1.98)  110  1.45  (1.02, 2.07)   P for trend      0.007      0.03      0.14      0.04      0.03    Stable subgroupb   Abstainers (0 g/day)  3514  211  1  (ref)  139  1  (ref)  65  1  (ref)  68  1  (ref)  52  1  (ref)   Light-moderate (0.1–29 g/day)  14 413  991  1.17  (0.93, 1.46)  628  1.13  (0.88, 1.46)  291  1.11  (0.81, 1.54)  324  1.21  (0.88, 1.68)  242  1.14  (0.80, 1.62)   Heavy (≥30 g/day)  2615  241  1.58  (1.17, 2.14)  155  1.58  (1.13, 2.21)  69  1.47  (0.97, 2.23)  78  1.70  (1.11, 2.59)  66  1.73  (1.11, 2.70)   P for trend      0.003      0.008      0.08      0.02      0.02    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  3754  317  1  (ref)  207  1  (ref)  92  1  (ref)  111  1  (ref)  80  1  (ref)   Heavy (≥30 g/day)  778  42  0.56  (0.34, 0.92)  23  0.45  (0.25, 0.82)  12  0.51  (0.25,1.08)  10  0.37  (0.17,0.84)  11  0.57  (0.26,1.23)  Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2588  151  1  (ref)  104  1  (ref)  46  1  (ref)  51  1  (ref)  28  1  (ref)   Heavy (≥30 g/day)  989  67  1.12  (0.71, 1.77)  47  1.08  (0.66, 1.78)  21  1.04  (0.52,2.08)  25  1.22  (0.67,2.23)  14  1.45  (0.65,3.23)  Women  Alcohol intake   Abstainers (0 g/day)  11 373  531  1  (ref)  400  1  (ref)  232  1  (ref)  154  1  (ref)  93  1  (ref)   Light-moderate (0.1–29 g/day)  23 546  1101  1.01  (0.87, 1.17)  825  1.02  (0.87, 1.20)  481  1.02  (0.84,1.23)  324  1.06  (0.85,1.33)  199  0.99  (0.75,1.31)   Heavy (≥30 g/day)  1253  78  1.46  (1.02, 2.10)  58  1.52  (1.03, 2.24)  38  1.70  (1.09,2.66)  18  1.25  (0.71,2.20)  11  1.03  (0.51,2.07)   P for trend      0.29      0.23      0.21      0.46      1.00    Stable subgroupb   Abstainers (0 g/day)  9296  441  1  (ref)  336  1  (ref)  193  1  (ref)  132  1  (ref)  72  1  (ref)   Light-moderate (0.1–29 g/day)  13 059  632  1.01  (0.84, 1.20)  475  1.02  (0.84, 1.24)  286  1.06  (0.85,1.33)  180  1.00  (0.77,1.31)  112  1.06  (0.75,1.49)   Heavy (≥30 g/day)  701  46  1.52  (0.95, 2.42)  34  1.52  (0.91, 2.54)  21  1.59  (0.88,2.88)  11  1.33  (0.64,2.76)  6  1.12  (0.44,2.85)   P for trend      0.38      0.37      0.27      0.73      0.71    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2818  113  1  (ref)  83  1  (ref)  47  1  (ref)  30  1  (ref)  24  1  (ref)   Heavy (≥30 g/day)  127  7  1.51  (0.48, 4.79)  5  1.53  (0.43, 5.48)  4      1      1      Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2427  101  1  (ref)  77  1  (ref)  44  1  (ref)  32  1  (ref)  19  1  (ref)   Heavy (≥30 g/day)  255  16  1.72  (0.75, 3.95)  11  1.57  (0.63, 3.93)  6  2.03  (0.63, 6.55)  5  1.31  (0.41, 4.15)  3        PT at risk  Colorectum  Colon  Proximal Colon  Distal Colon  Rectum      N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  N cases  HRa  95% CI  Men  Alcohol intake   Abstainers (0 g/day)  4548  267  1  (ref)  176  1  (ref)  81  1  (ref)  88  1  (ref)  69  1  (ref)   Light-moderate (0.1–29 g/day)  23 074  1625  1.22  (1.00, 1.48)  1050  1.21  (0.97, 1.50)  485  1.23  (0.93, 1.63)  538  1.23  (0.93, 1.63)  388  1.10  (0.82, 1.48)   Heavy (≥30 g/day)  4852  394  1.40  (1.10, 1.78)  246  1.35  (1.03, 1.77)  109  1.30  (0.92, 1.83)  126  1.42  (1.01, 1.98)  110  1.45  (1.02, 2.07)   P for trend      0.007      0.03      0.14      0.04      0.03    Stable subgroupb   Abstainers (0 g/day)  3514  211  1  (ref)  139  1  (ref)  65  1  (ref)  68  1  (ref)  52  1  (ref)   Light-moderate (0.1–29 g/day)  14 413  991  1.17  (0.93, 1.46)  628  1.13  (0.88, 1.46)  291  1.11  (0.81, 1.54)  324  1.21  (0.88, 1.68)  242  1.14  (0.80, 1.62)   Heavy (≥30 g/day)  2615  241  1.58  (1.17, 2.14)  155  1.58  (1.13, 2.21)  69  1.47  (0.97, 2.23)  78  1.70  (1.11, 2.59)  66  1.73  (1.11, 2.70)   P for trend      0.003      0.008      0.08      0.02      0.02    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  3754  317  1  (ref)  207  1  (ref)  92  1  (ref)  111  1  (ref)  80  1  (ref)   Heavy (≥30 g/day)  778  42  0.56  (0.34, 0.92)  23  0.45  (0.25, 0.82)  12  0.51  (0.25,1.08)  10  0.37  (0.17,0.84)  11  0.57  (0.26,1.23)  Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2588  151  1  (ref)  104  1  (ref)  46  1  (ref)  51  1  (ref)  28  1  (ref)   Heavy (≥30 g/day)  989  67  1.12  (0.71, 1.77)  47  1.08  (0.66, 1.78)  21  1.04  (0.52,2.08)  25  1.22  (0.67,2.23)  14  1.45  (0.65,3.23)  Women  Alcohol intake   Abstainers (0 g/day)  11 373  531  1  (ref)  400  1  (ref)  232  1  (ref)  154  1  (ref)  93  1  (ref)   Light-moderate (0.1–29 g/day)  23 546  1101  1.01  (0.87, 1.17)  825  1.02  (0.87, 1.20)  481  1.02  (0.84,1.23)  324  1.06  (0.85,1.33)  199  0.99  (0.75,1.31)   Heavy (≥30 g/day)  1253  78  1.46  (1.02, 2.10)  58  1.52  (1.03, 2.24)  38  1.70  (1.09,2.66)  18  1.25  (0.71,2.20)  11  1.03  (0.51,2.07)   P for trend      0.29      0.23      0.21      0.46      1.00    Stable subgroupb   Abstainers (0 g/day)  9296  441  1  (ref)  336  1  (ref)  193  1  (ref)  132  1  (ref)  72  1  (ref)   Light-moderate (0.1–29 g/day)  13 059  632  1.01  (0.84, 1.20)  475  1.02  (0.84, 1.24)  286  1.06  (0.85,1.33)  180  1.00  (0.77,1.31)  112  1.06  (0.75,1.49)   Heavy (≥30 g/day)  701  46  1.52  (0.95, 2.42)  34  1.52  (0.91, 2.54)  21  1.59  (0.88,2.88)  11  1.33  (0.64,2.76)  6  1.12  (0.44,2.85)   P for trend      0.38      0.37      0.27      0.73      0.71    Reported drinking more alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2818  113  1  (ref)  83  1  (ref)  47  1  (ref)  30  1  (ref)  24  1  (ref)   Heavy (≥30 g/day)  127  7  1.51  (0.48, 4.79)  5  1.53  (0.43, 5.48)  4      1      1      Reported drinking less alcohol 5 years before baseline   Light-moderate (0.1–29 g/day)  2427  101  1  (ref)  77  1  (ref)  44  1  (ref)  32  1  (ref)  19  1  (ref)   Heavy (≥30 g/day)  255  16  1.72  (0.75, 3.95)  11  1.57  (0.63, 3.93)  6  2.03  (0.63, 6.55)  5  1.31  (0.41, 4.15)  3      Results were not shown when less than five cases were available. CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAdjusted for age, BMI, smoking, family history of CRC, meat intake, processed meat intake, folate intake and non-occupational physical activity. bParticipants who reported drinking the same amount of alcohol intake at baseline and 5 years before baseline, including those who reported to be abstainers on both occasions. View Large Associations between ADH1B and ADH1C tagSNPs and CRC risk in drinkers Tables 3 and 4 show associations between ADH1B and ADH1C tagSNPs and CRC risk overall and by subsite in male and female drinkers, respectively, as analyzed according to a dominant model. Only FDR significant results will be mentioned below. ADH1B rs4147536 was associated with the risk of colon cancer and distal colon cancer in male drinkers [HRCA/AA versus CC (95% CI) = 1.25 (1.05–1.48) and 1.32 (1.07–1.62), respectively]. ADH1C rs283415 was associated with the risk of proximal colon cancer in female drinkers [HRTC/CC versus TT (95% CI) = 1.39 (1.08–1.80)]. Restricting these analyses to the stable drinkers revealed a statistically significant association between ADH1C rs4147542 and CRC risk in women [HRTC/CC versus TT (95% CI) = 0.73 (0.57–0.93)]. Stratifying these analyses by alcohol intake amount (i.e. light-moderate and heavy) to evaluate a potential threshold level of alcohol intake at which individual SNPs start to influence CRC risk revealed a statistically significant association between ADH1B rs3811802 and CRC risk in women who were heavy drinkers at baseline (>30 g/day) [HRAG/GG versus AA (95% CI) = 0.19 (0.07–0.50)], while no significant associations were observed in light-moderate drinkers. The results of both sensitivity analyses are presented in Supplementary Tables 1 and 2, available at Carcinogenesis Online. Table 3. Hazard ratios and 95% confidence intervals for the association between single nucleotide polymorphisms in ADH1B and ADH1C and risk of overall colorectal cancer and subtypes in male drinkers in the NLCS, 1986–2006   Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8918  638  1 (ref)  407  1 (ref)  185  1 (ref)  209  1 (ref)  155  1 (ref)    CA + AA  13 547  960  1.01 (0.87, 1.18)  632  1.05 (0.88, 1.24)  301  1.11 (0.89, 1.38)  318  1.02 (0.83, 1.26)  244  1.05 (0.83, 1.32)    P-value (adjusted P-value)      0.89 (0.17)    0.59 (0.15)    0.36 (0.08)    0.86 (0.18)    0.69 (0.15)   rs2075633  TT  11 627  799  1 (ref)  519  1 (ref)  254  1 (ref)  251  1 (ref)  200  1 (ref)    TC + CC  10 879  806  1.08 (0.93, 1.25)  524  1.08 (0.91, 1.27)  234  0.98 (0.79, 1.21)  278  1.19 (0.97, 1.46)  202  1.08 (0.86, 1.35)    P-value (adjusted P-value)      0.32 (0.07)    0.38 (0.07)    0.84 (0.18)    0.10 (0.03)    0.49 (0.10)   rs1693439  GG  19 243  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    GA + AA  3247  237  1.03 (0.84, 1.28)  159  1.07 (0.85, 1.36)  84  1.25 (0.94, 1.67)  73  0.95 (0.71, 1.28)  59  1.02 (0.75, 1.40)    P-value (adjusted P-value)      0.76 (0.13)      0.55 (0.12)    0.13 (0.05)    0.75 (0.15)    0.88 (0.20)   rs9307239  CC  8625  557  1 (ref)  362  1 (ref)  174  1 (ref)  182  1 (ref)  145  1 (ref)    CT + TT  13 841  1046  1.17 (1.00, 1.37)  680  1.17 (0.99, 1.40)  313  1.12 (0.90, 1.40)  347  1.19 (0.96, 1.48)  256  1.10 (0.87, 1.39)    P-value (adjusted P-value)      0.04 (0.02)    0.07 (0.03)    0.30 (0.07)    0.11 (0.05)    0.41 (0.05)   rs4147536  CC  14 259  975  1 (ref)  617  1 (ref)  295  1 (ref)  305  1 (ref)  251  1 (ref)    CA + AA  8221  630  1.16 (0.99, 1.35)  426  1.25 (1.05, 1.48)  193  1.20 (0.96, 1.49)  224  1.32 (1.07, 1.62)  151  1.06 (0.84, 1.34)    P-value (adjusted P-value)      0.06 (0.03)    0.01b (0.02)    0.11 (0.03)    0.01b (0.02)    0.60 (0.13)   rs3811802  AA  6245  454  1 (ref)  295  1 (ref)  139  1 (ref)  150  1 (ref)  119  1 (ref)    AG + GG  16 261  1150  0.96 (0.81, 1.14)  747  0.96 (0.80, 1.16)  349  0.94 (0.75, 1.20)  378  0.96 (0.76, 1.20)  283  0.91 (0.71, 1.16)    P-value (adjusted P-value)      0.65 (0.12)    0.66 (0.17)    0.63 (0.13)    0.72 (0.13)    0.44 (0.07)  ADH1C   rs698  TT  7777  549  1 (ref)  368  1 (ref)  180  1 (ref)  182  1 (ref)  134  1 (ref)    TC + CC  14 729  1053  1.01 (0.86, 1.18)  674  0.96 (0.81, 1.14)  308  0.89 (0.71, 1.11)  346  1.00 (0.81, 1.24)  267  1.05 (0.83, 1.33)    P-value (adjusted P-value)      0.93 (0.20)    0.65 (0.20)    0.31 (0.15)    1.00 (0.20)    0.69 (0.08)   rs1662033  TT  10 409  748  1 (ref)  496  1 (ref)  238  1 (ref)  248  1 (ref)  188  1 (ref)    TG + GG  12 079  857  0.97 (0.83, 1.13)  547  0.93 (0.78, 1.10)  250  0.88 (0.71, 1.08)  281  0.96 (0.78, 1.18)  214  0.97 (0.77, 1.22)    P-value (adjusted P-value)      0.68 (0.13)    0.38 (0.10)    0.22 (0.13)    0.69 (0.08)    0.79 (0.13)   rs3114046  CC  19 235  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    CT + TT  3271  238  1.03 (0.84, 1.27)  159  1.07 (0.85, 1.35)  84  1.24 (0.93, 1.65)  73  0.95 (0.71, 1.27)  60  1.04 (0.76, 1.41)    P-value (adjusted P-value)      0.77 (0.15)    0.58 (0.17)    0.14 (0.10)    0.72 (0.10)    0.83 (0.15)   rs4147542  TT  10 907  797  1 (ref)  514  1 (ref)  245  1 (ref)  258  1 (ref)  202  1 (ref)    TC + CC  9628  754  1.07 (0.92, 1.25)  488  1.08 (0.91, 1.28)  219  1.01 (0.81, 1.26)  254  1.12 (0.91, 1.38)  188  1.06 (0.84, 1.33)    P-value (adjusted P-value)      0.37 (0.12)    0.40 (0.12)    0.94 (0.20)    0.28 (0.03)    0.63 (0.05)   rs283415  TT  6975  500  1 (ref)  334  1 (ref)  168  1 (ref)  160  1 (ref)  123  1 (ref)    TC + CC  15 531  1104  0.99 (0.84, 1.16)  709  0.95 (0.79, 1.13)  320  0.85 (0.68, 1.06)  369  1.03 (0.83, 1.29)  279  1.02 (0.80, 1.30)    P-value (adjusted P-value)      0.88 (0.17)    0.56 (0.15)    0.15 (0.12)    0.77 (0.13)    0.90 (0.18)   rs4699741  TT  19 601  1432  1 (ref)  931  1 (ref)  437  1 (ref)  470  1 (ref)  353  1 (ref)    TC + CC  2905  173  0.80 (0.63, 1.01)  112  0.79 (0.61, 1.03)  51  0.77 (0.55, 1.08)  59  0.83 (0.60, 1.14)  49  0.92 (0.66, 1.30)    P-value (adjusted P-value)      0.06 (0.03)    0.08 (0.03)    0.13 (0.08)    0.25 (0.02)    0.66 (0.07)    Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8918  638  1 (ref)  407  1 (ref)  185  1 (ref)  209  1 (ref)  155  1 (ref)    CA + AA  13 547  960  1.01 (0.87, 1.18)  632  1.05 (0.88, 1.24)  301  1.11 (0.89, 1.38)  318  1.02 (0.83, 1.26)  244  1.05 (0.83, 1.32)    P-value (adjusted P-value)      0.89 (0.17)    0.59 (0.15)    0.36 (0.08)    0.86 (0.18)    0.69 (0.15)   rs2075633  TT  11 627  799  1 (ref)  519  1 (ref)  254  1 (ref)  251  1 (ref)  200  1 (ref)    TC + CC  10 879  806  1.08 (0.93, 1.25)  524  1.08 (0.91, 1.27)  234  0.98 (0.79, 1.21)  278  1.19 (0.97, 1.46)  202  1.08 (0.86, 1.35)    P-value (adjusted P-value)      0.32 (0.07)    0.38 (0.07)    0.84 (0.18)    0.10 (0.03)    0.49 (0.10)   rs1693439  GG  19 243  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    GA + AA  3247  237  1.03 (0.84, 1.28)  159  1.07 (0.85, 1.36)  84  1.25 (0.94, 1.67)  73  0.95 (0.71, 1.28)  59  1.02 (0.75, 1.40)    P-value (adjusted P-value)      0.76 (0.13)      0.55 (0.12)    0.13 (0.05)    0.75 (0.15)    0.88 (0.20)   rs9307239  CC  8625  557  1 (ref)  362  1 (ref)  174  1 (ref)  182  1 (ref)  145  1 (ref)    CT + TT  13 841  1046  1.17 (1.00, 1.37)  680  1.17 (0.99, 1.40)  313  1.12 (0.90, 1.40)  347  1.19 (0.96, 1.48)  256  1.10 (0.87, 1.39)    P-value (adjusted P-value)      0.04 (0.02)    0.07 (0.03)    0.30 (0.07)    0.11 (0.05)    0.41 (0.05)   rs4147536  CC  14 259  975  1 (ref)  617  1 (ref)  295  1 (ref)  305  1 (ref)  251  1 (ref)    CA + AA  8221  630  1.16 (0.99, 1.35)  426  1.25 (1.05, 1.48)  193  1.20 (0.96, 1.49)  224  1.32 (1.07, 1.62)  151  1.06 (0.84, 1.34)    P-value (adjusted P-value)      0.06 (0.03)    0.01b (0.02)    0.11 (0.03)    0.01b (0.02)    0.60 (0.13)   rs3811802  AA  6245  454  1 (ref)  295  1 (ref)  139  1 (ref)  150  1 (ref)  119  1 (ref)    AG + GG  16 261  1150  0.96 (0.81, 1.14)  747  0.96 (0.80, 1.16)  349  0.94 (0.75, 1.20)  378  0.96 (0.76, 1.20)  283  0.91 (0.71, 1.16)    P-value (adjusted P-value)      0.65 (0.12)    0.66 (0.17)    0.63 (0.13)    0.72 (0.13)    0.44 (0.07)  ADH1C   rs698  TT  7777  549  1 (ref)  368  1 (ref)  180  1 (ref)  182  1 (ref)  134  1 (ref)    TC + CC  14 729  1053  1.01 (0.86, 1.18)  674  0.96 (0.81, 1.14)  308  0.89 (0.71, 1.11)  346  1.00 (0.81, 1.24)  267  1.05 (0.83, 1.33)    P-value (adjusted P-value)      0.93 (0.20)    0.65 (0.20)    0.31 (0.15)    1.00 (0.20)    0.69 (0.08)   rs1662033  TT  10 409  748  1 (ref)  496  1 (ref)  238  1 (ref)  248  1 (ref)  188  1 (ref)    TG + GG  12 079  857  0.97 (0.83, 1.13)  547  0.93 (0.78, 1.10)  250  0.88 (0.71, 1.08)  281  0.96 (0.78, 1.18)  214  0.97 (0.77, 1.22)    P-value (adjusted P-value)      0.68 (0.13)    0.38 (0.10)    0.22 (0.13)    0.69 (0.08)    0.79 (0.13)   rs3114046  CC  19 235  1367  1 (ref)  884  1 (ref)  404  1 (ref)  456  1 (ref)  342  1 (ref)    CT + TT  3271  238  1.03 (0.84, 1.27)  159  1.07 (0.85, 1.35)  84  1.24 (0.93, 1.65)  73  0.95 (0.71, 1.27)  60  1.04 (0.76, 1.41)    P-value (adjusted P-value)      0.77 (0.15)    0.58 (0.17)    0.14 (0.10)    0.72 (0.10)    0.83 (0.15)   rs4147542  TT  10 907  797  1 (ref)  514  1 (ref)  245  1 (ref)  258  1 (ref)  202  1 (ref)    TC + CC  9628  754  1.07 (0.92, 1.25)  488  1.08 (0.91, 1.28)  219  1.01 (0.81, 1.26)  254  1.12 (0.91, 1.38)  188  1.06 (0.84, 1.33)    P-value (adjusted P-value)      0.37 (0.12)    0.40 (0.12)    0.94 (0.20)    0.28 (0.03)    0.63 (0.05)   rs283415  TT  6975  500  1 (ref)  334  1 (ref)  168  1 (ref)  160  1 (ref)  123  1 (ref)    TC + CC  15 531  1104  0.99 (0.84, 1.16)  709  0.95 (0.79, 1.13)  320  0.85 (0.68, 1.06)  369  1.03 (0.83, 1.29)  279  1.02 (0.80, 1.30)    P-value (adjusted P-value)      0.88 (0.17)    0.56 (0.15)    0.15 (0.12)    0.77 (0.13)    0.90 (0.18)   rs4699741  TT  19 601  1432  1 (ref)  931  1 (ref)  437  1 (ref)  470  1 (ref)  353  1 (ref)    TC + CC  2905  173  0.80 (0.63, 1.01)  112  0.79 (0.61, 1.03)  51  0.77 (0.55, 1.08)  59  0.83 (0.60, 1.14)  49  0.92 (0.66, 1.30)    P-value (adjusted P-value)      0.06 (0.03)    0.08 (0.03)    0.13 (0.08)    0.25 (0.02)    0.66 (0.07)  CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAge-adjusted. bSignificant after adjusting for multiple testing. View Large Table 4. Hazard ratios and 95% confidence intervals for the association between single nucleotide polymorphisms in ADH1B and ADH1C and risk of overall colorectal cancer and subtypes in female drinkers in the NLCS, 1986–2006   Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8159  366  1 (ref)  270  1 (ref)  157  1 (ref)  109  1 (ref)  66  1 (ref)    CA + AA  11 718  517  0.98 (0.82, 1.18)  384  0.99 (0.81, 1.20)  234  1.03 (0.81, 1.31)  141  0.90 (0.68, 1.19)  99  1.04 (0.75, 1.45)    P-value (adjusted P-value)      0.82 (0.15)    0.88 (0.20)    0.80 (0.17)    0.45 (0.10)    0.82 (0.18)   rs2075633  TT  9702  457  1 (ref)  338  1 (ref)  216  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  10 233  429  0.89 (0.75, 1.07)  318  0.89 (0.74, 1.09)  177  0.78 (0.62, 0.99)  136  1.13 (0.86, 1.49)  76  0.81 (0.59, 1.13)    P-value (adjusted P-value)      0.22 (0.05)    0.27 (0.05)    0.04 (0.02)    0.38 (0.08)    0.22 (0.02)   rs1693439  GG  16 820  746  1 (ref)  545  1 (ref)  328  1 (ref)  208  1 (ref)  141  1 (ref)    GA + AA  3135  140  1.01 (0.79, 1.29)  111  1.09 (0.84, 1.42)  65  1.06 (0.77, 1.46)  42  1.08 (0.75, 1.57)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.95 (0.20)    0.51 (0.10)    0.70 (0.15)    0.67 (0.12)    0.71 (0.17)   rs9307239  CC  7622  348  1 (ref)  258  1 (ref)  158  1 (ref)  94  1 (ref)  67  1 (ref)    CT + TT  12 313  536  0.95 (0.79,1.14)  396  0.95 (0.77, 1.16)  234  0.91 (0.72, 1.16)  155  1.02 (0.77, 1.35)  98  0.90 (0.65, 1.26)    P-value (adjusted P-value)      0.58 (0.10)    0.59 (0.13)    0.45 (0.10)    0.91 (0.20)    0.55 (0.12)   rs4147536  CC  12 688  563  1 (ref)  426  1 (ref)  249  1 (ref)  171  1 (ref)  98  1 (ref)    CA + AA  7247  321  0.99 (0.82, 1.19)  228  0.93 (0.75, 1.14)  144  1.00 (0.78, 1.27)  77  0.78 (0.58, 1.06)  67  1.19 (0.85, 1.66)    P-value (adjusted P-value)      0.90 (0.18)    0.47 (0.08)    0.99 (0.20)    0.11 (0.07)    0.32 (0.03)   rs3811802  AA  6101  260  1 (ref)  195  1 (ref)  115  1 (ref)  78  1 (ref)  46  1 (ref)    AG + GG  13 854  625  1.06 (0.87, 1.29)  460  1.05 (0.84, 1.29)  278  1.07 (0.83, 1.39)  171  0.97 (0.72, 1.31)  119  1.14 (0.79, 1.64)    P-value (adjusted P-value)      0.53 (0.08)    0.68 (0.18)    0.59 (0.12)    0.84 (0.17)    0.47 (0.08)  ADH1C   rs698  TT  7385  304  1 (ref)  225  1 (ref)  126  1 (ref)  95  1 (ref)  62  1 (ref)    TC + CC  12 553  580  1.13 (0.94, 1.36)  430  1.13 (0.92, 1.39)  266  1.25 (0.98, 1.61)  155  0.97 (0.73, 1.28)  102  0.97 (0.69, 1.37)    P-value (adjusted P-value)      0.20 (0.07)    0.23 (0.07)    0.08 (0.07)    0.81 (0.15)    0.88 (0.17)   rs1662033  TT  9720  410  1 (ref)  300  1 (ref)  169  1 (ref)  124  1 (ref)  80  1 (ref)    TG + GG  10 215  476  1.11 (0.93, 1.33)  356  1.13 (0.93, 1.38)  224  1.27 (1.00, 1.60)  126  0.97 (0.74, 1.28)  85  1.01 (0.73, 1.41)    P-value (adjusted P-value)      0.26 (0.10)    0.21 (0.05)    0.05 (0.03)    0.83 (0.17)    0.93 (0.20)   rs3114046  CC  16 820  745  1 (ref)  544  1 (ref)  328  1 (ref)  207  1 (ref)  141  1 (ref)    CT + TT  3135  141  1.02 (0.79, 1.30)  112  1.10 (0.85, 1.44)  65  1.06 (0.77, 1.46)  43  1.11 (0.77, 1.61)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.90 (0.18)    0.46 (0.13)    0.70 (0.17)    0.56 (0.07)    0.71 (0.10)   rs4147542  TT  9297  443  1 (ref)  324  1 (ref)  205  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  9566  402  0.87 (0.73, 1.05)  300  0.89 (0.73, 1.09)  169  0.79 (0.62, 1.01)  123  1.04 (0.79, 1.38)  71  0.77 (0.55, 1.08)    P-value (adjusted P-value)      0.15 (0.05)    0.26 (0.08)    0.06 (0.05)    0.77 (0.12)    0.13 (0.03)   rs283415  TT  6906  272  1 (ref)  200  1 (ref)  109  1 (ref)  87  1 (ref)  55  1 (ref)    TC + CC  13 049  614  1.20 (0.99, 1.46)  456  1.22 (0.99, 1.50)  284  1.39 (1.08, 1.80)  163  1.00 (0.75, 1.33)  110  1.06 (0.75, 1.51)    P-value (adjusted P-value)      0.06 (0.02)    0.07 (0.02)    0.01b (0.02)    0.99 (0.18)    0.72 (0.12)   rs4699741  TT  17 642  769  1 (ref)  577  1 (ref)  347  1 (ref)  217  1 (ref)  136  1 (ref)    TC + CC  2314  117  1.19 (0.91, 1.57)  79  1.08 (0.80, 1.46)  46  1.06 (0.73, 1.52)  33  1.18 (0.78, 1.78)  29  1.66 (1.06, 2.58)    P-value (adjusted P-value)      0.20 (0.08)    0.62 (0.18)    0.77 (0.18)    0.43 (0.05)    0.03 (0.02)    Allele  PT at risk  Colorectum  Colon  Proximal colon  Distal colon  Rectum        N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  N cases  HRa 95% CI  ADH1B   rs1159918  CC  8159  366  1 (ref)  270  1 (ref)  157  1 (ref)  109  1 (ref)  66  1 (ref)    CA + AA  11 718  517  0.98 (0.82, 1.18)  384  0.99 (0.81, 1.20)  234  1.03 (0.81, 1.31)  141  0.90 (0.68, 1.19)  99  1.04 (0.75, 1.45)    P-value (adjusted P-value)      0.82 (0.15)    0.88 (0.20)    0.80 (0.17)    0.45 (0.10)    0.82 (0.18)   rs2075633  TT  9702  457  1 (ref)  338  1 (ref)  216  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  10 233  429  0.89 (0.75, 1.07)  318  0.89 (0.74, 1.09)  177  0.78 (0.62, 0.99)  136  1.13 (0.86, 1.49)  76  0.81 (0.59, 1.13)    P-value (adjusted P-value)      0.22 (0.05)    0.27 (0.05)    0.04 (0.02)    0.38 (0.08)    0.22 (0.02)   rs1693439  GG  16 820  746  1 (ref)  545  1 (ref)  328  1 (ref)  208  1 (ref)  141  1 (ref)    GA + AA  3135  140  1.01 (0.79, 1.29)  111  1.09 (0.84, 1.42)  65  1.06 (0.77, 1.46)  42  1.08 (0.75, 1.57)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.95 (0.20)    0.51 (0.10)    0.70 (0.15)    0.67 (0.12)    0.71 (0.17)   rs9307239  CC  7622  348  1 (ref)  258  1 (ref)  158  1 (ref)  94  1 (ref)  67  1 (ref)    CT + TT  12 313  536  0.95 (0.79,1.14)  396  0.95 (0.77, 1.16)  234  0.91 (0.72, 1.16)  155  1.02 (0.77, 1.35)  98  0.90 (0.65, 1.26)    P-value (adjusted P-value)      0.58 (0.10)    0.59 (0.13)    0.45 (0.10)    0.91 (0.20)    0.55 (0.12)   rs4147536  CC  12 688  563  1 (ref)  426  1 (ref)  249  1 (ref)  171  1 (ref)  98  1 (ref)    CA + AA  7247  321  0.99 (0.82, 1.19)  228  0.93 (0.75, 1.14)  144  1.00 (0.78, 1.27)  77  0.78 (0.58, 1.06)  67  1.19 (0.85, 1.66)    P-value (adjusted P-value)      0.90 (0.18)    0.47 (0.08)    0.99 (0.20)    0.11 (0.07)    0.32 (0.03)   rs3811802  AA  6101  260  1 (ref)  195  1 (ref)  115  1 (ref)  78  1 (ref)  46  1 (ref)    AG + GG  13 854  625  1.06 (0.87, 1.29)  460  1.05 (0.84, 1.29)  278  1.07 (0.83, 1.39)  171  0.97 (0.72, 1.31)  119  1.14 (0.79, 1.64)    P-value (adjusted P-value)      0.53 (0.08)    0.68 (0.18)    0.59 (0.12)    0.84 (0.17)    0.47 (0.08)  ADH1C   rs698  TT  7385  304  1 (ref)  225  1 (ref)  126  1 (ref)  95  1 (ref)  62  1 (ref)    TC + CC  12 553  580  1.13 (0.94, 1.36)  430  1.13 (0.92, 1.39)  266  1.25 (0.98, 1.61)  155  0.97 (0.73, 1.28)  102  0.97 (0.69, 1.37)    P-value (adjusted P-value)      0.20 (0.07)    0.23 (0.07)    0.08 (0.07)    0.81 (0.15)    0.88 (0.17)   rs1662033  TT  9720  410  1 (ref)  300  1 (ref)  169  1 (ref)  124  1 (ref)  80  1 (ref)    TG + GG  10 215  476  1.11 (0.93, 1.33)  356  1.13 (0.93, 1.38)  224  1.27 (1.00, 1.60)  126  0.97 (0.74, 1.28)  85  1.01 (0.73, 1.41)    P-value (adjusted P-value)      0.26 (0.10)    0.21 (0.05)    0.05 (0.03)    0.83 (0.17)    0.93 (0.20)   rs3114046  CC  16 820  745  1 (ref)  544  1 (ref)  328  1 (ref)  207  1 (ref)  141  1 (ref)    CT + TT  3135  141  1.02 (0.79, 1.30)  112  1.10 (0.85, 1.44)  65  1.06 (0.77, 1.46)  43  1.11 (0.77, 1.61)  24  0.91 (0.58, 1.45)    P-value (adjusted P-value)      0.90 (0.18)    0.46 (0.13)    0.70 (0.17)    0.56 (0.07)    0.71 (0.10)   rs4147542  TT  9297  443  1 (ref)  324  1 (ref)  205  1 (ref)  114  1 (ref)  89  1 (ref)    TC + CC  9566  402  0.87 (0.73, 1.05)  300  0.89 (0.73, 1.09)  169  0.79 (0.62, 1.01)  123  1.04 (0.79, 1.38)  71  0.77 (0.55, 1.08)    P-value (adjusted P-value)      0.15 (0.05)    0.26 (0.08)    0.06 (0.05)    0.77 (0.12)    0.13 (0.03)   rs283415  TT  6906  272  1 (ref)  200  1 (ref)  109  1 (ref)  87  1 (ref)  55  1 (ref)    TC + CC  13 049  614  1.20 (0.99, 1.46)  456  1.22 (0.99, 1.50)  284  1.39 (1.08, 1.80)  163  1.00 (0.75, 1.33)  110  1.06 (0.75, 1.51)    P-value (adjusted P-value)      0.06 (0.02)    0.07 (0.02)    0.01b (0.02)    0.99 (0.18)    0.72 (0.12)   rs4699741  TT  17 642  769  1 (ref)  577  1 (ref)  347  1 (ref)  217  1 (ref)  136  1 (ref)    TC + CC  2314  117  1.19 (0.91, 1.57)  79  1.08 (0.80, 1.46)  46  1.06 (0.73, 1.52)  33  1.18 (0.78, 1.78)  29  1.66 (1.06, 2.58)    P-value (adjusted P-value)      0.20 (0.08)    0.62 (0.18)    0.77 (0.18)    0.43 (0.05)    0.03 (0.02)  CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAge-adjusted. bSignificant after adjusting for multiple testing. View Large Effect modification of the alcohol–CRC association by ADH1B and ADH1C tagSNPs Table 5 shows the associations between alcohol intake and CRC risk in men and women, stratified by genotype, as analyzed according to a dominant model. The alcohol–CRC associations observed in genotype strata generally aligned with overall alcohol–CRC associations. HRs around the null were observed in one stratum and a pattern of increasing risks across alcohol categories was observed in the other stratum in men, while a difference in associations between genotype strata was less clear in women. Statistically significantly increased HRs for CRC were only present when comparing heavy drinkers with abstainers. Most interactions were not significant after FDR correction except for the interactions between alcohol intake and ADH1B rs3811802 and ADH1C rs4147542 in women. Risk was strongly increased in heavy (but not light-moderate) drinkers versus abstainers in female rs3811802 AA carriers, although the CI around this estimate is large [HR (95% CI) 5.72 (2.24–14.63)], while the risk across alcohol categories in female rs3811802 AG/GG carriers remained almost unchanged. In female rs4147542 TT carriers, there was a significant positive trend in CRC risk with increasing alcohol intake [HRlight-moderate drinkers versus abstainers (95% CI) 1.25 (0.98–1.59) and HRheavy drinkers versus abstainers (95% CI) 1.78 (1.01–3.14)], while a decreased risk was observed for light-moderate drinkers and an increased risk for heavy drinkers relative to abstainers in female rs4147542 TC/CC carriers [HR (95% CI) 0.78 (0.60–1.02) and HR (95% CI) 1.74 (0.92–3.31), respectively]. After restricting the analysis to the stable subgroup (Supplementary Table 3, available at Carcinogenesis Online), only the interaction between alcohol intake and ADH1C rs4147542 in relation to CRC risk in women remained significant after FDR correction. In addition, in men, results were more pronounced, showing stronger increased HRs for CRC in heavy drinkers as compared with abstainers and more significant positive trends. Table 5. Hazard ratios and 95% confidence intervals for the association between alcohol intake and colorectal cancer risk by ADH1B and ADH1C genotypes in men and women in the NLCS, 1986–2006 Gene  SNP  Allele  Alcohol intake            Abstainers  Light-moderate (0.1–29 g/day)  Heavy (≥30 g/day)            N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  P for trend  P for interaction (adjusted P-value)  Men   ADH1B  rs1159918  CC  88/1232  1  (ref)  499/7191  0.92  (0.65, 1.31)  116/1445  1.05  (0.68, 1.64)  0.75        CA/AA  117/2071  1  (ref)  743/10 709  1.22  (0.91, 1.64)  182/2269  1.41  (0.98, 2.02)  0.07  0.41 (0.10)   ADH1B  rs2075633  TT  95/1596  1  (ref)  625/9152  1.14  (0.81, 1.59)  154/1960  1.34  (0.89, 2.01)  0.14        TC/CC  110/1707  1  (ref)  624/8769  1.09  (0.80, 1.49)  144/1774  1.20  (0.80, 1.79)  0.38  0.99 (0.20)   ADH1B  rs1693439  GG  185/2880  1  (ref)  1073/15305  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        GA/AA  20/424  1  (ref)  175/2599  1.12  (0.53, 2.35)  52/515  1.98  (0.82, 4.81)  0.08  0.28 (0.05)   ADH1B  rs9307239  CC  73/1083  1  (ref)  434/6872  0.88  (0.61, 1.28)  108/1504  1.01  (0.64, 1.58)  0.86        CT/TT  132/2208  1  (ref)  814/11 009  1.22  (0.91, 1.62)  189/2230  1.37  (0.96, 1.96)  0.09  0.40 (0.08)   ADH1B  rs4147536  CC  127/1985  1  (ref)  760/11 432  0.99  (0.75, 1.31)  177/2395  1.09  (0.76, 1.55)  0.59        CA/AA  78/1319  1  (ref)  489/6483  1.29  (0.89, 1.88)  121/1338  1.53  (0.96, 2.46)  0.08  0.42 (0.12)   ADH1B  rs3811802  AA  59/1007  1  (ref)  350/5062  1.23  (0.80, 1.88)  82/1013  1.29  (0.75, 2.22)  0.38        AG/GG  146/2297  1  (ref)  898/12 859  1.04  (0.79, 1.37)  216/2721  1.21  (0.87, 1.69)  0.22  0.80 (0.17)   ADH1C  rs698  TT  69/1270  1  (ref)  419/6312  1.33  (0.90, 1.97)  105/1197  1.75  (1.07, 2.86)  0.03        TC/CC  136/2034  1  (ref)  827/11 609  1.02  (0.77, 1.36)  193/2537  1.11  (0.78, 1.57)  0.53  0.35 (0.10)   ADH1C  rs1662033  TT  97/1773  1  (ref)  576/8380  1.33  (0.96, 1.86)  144/1678  1.69  (1.11, 2.57)  0.02        TG/GG  108/1530  1  (ref)  673/9541  0.95  (0.68, 1.31)  154/2056  1.02  (0.68, 1.51)  0.86  0.17 (0.05)   ADH1C  rs3114046  CC  185/2880  1  (ref)  1073/15297  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        CT/TT  20/424  1  (ref)  176/2623  1.13  (0.54, 2.38)  52/515  1.88  (0.78, 4.51)  0.10  0.27 (0.08   ADH1C  rs4147542  TT  107/1260  1  (ref)  631/8763  0.85  (0.61, 1.18)  144/1723  0.98  (0.65, 1.48)  0.95        TC/CC  94/1706  1  (ref)  570/7672  1.28  (0.91, 1.79)  149/1575  1.61  (1.06, 2.46)  0.03  0.12 (0.03)   ADH1C  rs283415  TT  65/1243  1  (ref)  385/5654  1.40  (0.94, 2.08)  93/1104  1.66  (1.00, 2.75)  0.05        TC/CC  140/2060  1  (ref)  863/12 266  1.00  (0.75, 1.32)  205/2630  1.12  (0.79, 1.58)  0.46  0.37 (0.12)   ADH1C  rs4699741  TT  179/2738  1  (ref)  1121/15601  1.05  (0.83, 1.35)  258/3230  1.17  (0.86, 1.58)  0.31        TC/CC  26/565  1  (ref)  128/2319  1.19  (0.65, 2.19)  40/503  1.61  (0.75, 3.44)  0.22  0.50 (0.15)  Women   ADH1B  rs1159918  CC  147/3395  1  (ref)  335/7422  1.08  (0.82, 1.43)  20/392  1.43  (0.73, 2.81)  0.38        CA/AA  234/5460  1  (ref)  456/10 666  0.99  (0.79, 1.24)  38/572  1.75  (1.03, 2.97)  0.33  0.68 (0.15)   ADH1B  rs2075633  TT  199/4735  1  (ref)  412/8725  1.13  (0.89, 1.43)  28/570  1.27  (0.72, 2.23)  0.26        TC/CC  181/4120  1  (ref)  382/9420  0.94  (0.73, 1.21)  30/394  2.10  (1.13, 3.91)  0.46  0.16 (0.03)   ADH1B  rs1693439  GG  324/7497  1  (ref)  668/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        GA/AA  57/1358  1  (ref)  126/2793  1.11  (0.69, 1.81)  11/157  1.57  (0.54, 4.60)  0.46  0.96 (0.18)   ADH1B  rs9307239  CC  148/3546  1  (ref)  309/6954  1.08  (0.81, 1.44)  28/377  1.80  (0.95, 3.38)  0.18        CT/TT  233/5292  1  (ref)  483/11 191  0.98  (0.78, 1.22)  30/587  1.39  (0.79, 2.43)  0.71  0.62 (0.13)   ADH1B  rs4147536  CC  236/5557  1  (ref)  513/11 574  1.08  (0.86, 1.34)  32/637  1.37  (0.80, 2.33)  0.30        CA/AA  145/3298  1  (ref)  279/6572  0.94  (0.70, 1.25)  26/328  2.03  (1.03, 3.99)  0.45  0.32 (0.07)   ADH1B  rs3811802  AA  107/2509  1  (ref)  224/5709  0.96  (0.68, 1.37)  22/137  5.72  (2.24, 14.63)  0.13        AG/GG  274/6313  1  (ref)  569/12 457  1.04  (0.85, 1.28)  36/827  1.09  (0.67, 1.77)  0.63  0.004b (0.02)   ADH1C  rs698  TT  139/3189  1  (ref)  273/6696  0.94  (0.70, 1.27)  17/349  1.50  (0.73, 3.06)  0.86        TC/CC  240/5666  1  (ref)  520/11 453  1.09  (0.88, 1.36)  40/615  1.70  (1.02, 2.84)  0.11  0.60 (0.17)   ADH1C  rs1662033  TT  193/4346  1  (ref)  367/8777  0.92  (0.72, 1.19)  25/440  1.59  (0.87, 2.92)  0.84        TG/GG  188/4508  1  (ref)  427/9369  1.12  (0.87, 1.44)  33/525  1.63  (0.92, 2.87)  0.13  0.44 (0.13)   ADH1C  rs3114046  CC  324/7497  1  (ref)  667/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        CT/TT  57/1358  1  (ref)  127/2793  1.12  (0.69, 1.82)  11/157  1.55  (0.53, 4.54)  0.45  0.95 (0.20)   ADH1C  rs4147542  TT  182/4890  1  (ref)  399/8448  1.25  (0.98, 1.59)  29/458  1.78  (1.01, 3.14)  0.02        TC/CC  180/3474  1  (ref)  356/8747  0.78  (0.60, 1.02)  28/405  1.74  (0.92, 3.31)  0.58  0.01b (0.02)   ADH1C  rs283415  TT  134/2923  1  (ref)  245/6263  0.87  (0.64, 1.19)  14/324  1.27  (0.59, 2.72)  0.67        TC/CC  247/5932  1  (ref)  549/11 903  1.12  (0.90, 1.39)  44/640  1.83  (1.11, 3.02)  0.05  0.19 (0.07)   ADH1C  rs4699741  TT  326/7548  1  (ref)  692/16 037  1.00  (0.83, 1.20)  50/850  1.52  (0.97, 2.37)  0.39        TC/CC  55/1307  1  (ref)  102/2129  1.43  (0.83, 2.46)  8/114  2.08  (0.61, 7.12)  0.12  0.82 (0.18)  Gene  SNP  Allele  Alcohol intake            Abstainers  Light-moderate (0.1–29 g/day)  Heavy (≥30 g/day)            N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  N cases/PT at risk  HRa  95% CI  P for trend  P for interaction (adjusted P-value)  Men   ADH1B  rs1159918  CC  88/1232  1  (ref)  499/7191  0.92  (0.65, 1.31)  116/1445  1.05  (0.68, 1.64)  0.75        CA/AA  117/2071  1  (ref)  743/10 709  1.22  (0.91, 1.64)  182/2269  1.41  (0.98, 2.02)  0.07  0.41 (0.10)   ADH1B  rs2075633  TT  95/1596  1  (ref)  625/9152  1.14  (0.81, 1.59)  154/1960  1.34  (0.89, 2.01)  0.14        TC/CC  110/1707  1  (ref)  624/8769  1.09  (0.80, 1.49)  144/1774  1.20  (0.80, 1.79)  0.38  0.99 (0.20)   ADH1B  rs1693439  GG  185/2880  1  (ref)  1073/15305  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        GA/AA  20/424  1  (ref)  175/2599  1.12  (0.53, 2.35)  52/515  1.98  (0.82, 4.81)  0.08  0.28 (0.05)   ADH1B  rs9307239  CC  73/1083  1  (ref)  434/6872  0.88  (0.61, 1.28)  108/1504  1.01  (0.64, 1.58)  0.86        CT/TT  132/2208  1  (ref)  814/11 009  1.22  (0.91, 1.62)  189/2230  1.37  (0.96, 1.96)  0.09  0.40 (0.08)   ADH1B  rs4147536  CC  127/1985  1  (ref)  760/11 432  0.99  (0.75, 1.31)  177/2395  1.09  (0.76, 1.55)  0.59        CA/AA  78/1319  1  (ref)  489/6483  1.29  (0.89, 1.88)  121/1338  1.53  (0.96, 2.46)  0.08  0.42 (0.12)   ADH1B  rs3811802  AA  59/1007  1  (ref)  350/5062  1.23  (0.80, 1.88)  82/1013  1.29  (0.75, 2.22)  0.38        AG/GG  146/2297  1  (ref)  898/12 859  1.04  (0.79, 1.37)  216/2721  1.21  (0.87, 1.69)  0.22  0.80 (0.17)   ADH1C  rs698  TT  69/1270  1  (ref)  419/6312  1.33  (0.90, 1.97)  105/1197  1.75  (1.07, 2.86)  0.03        TC/CC  136/2034  1  (ref)  827/11 609  1.02  (0.77, 1.36)  193/2537  1.11  (0.78, 1.57)  0.53  0.35 (0.10)   ADH1C  rs1662033  TT  97/1773  1  (ref)  576/8380  1.33  (0.96, 1.86)  144/1678  1.69  (1.11, 2.57)  0.02        TG/GG  108/1530  1  (ref)  673/9541  0.95  (0.68, 1.31)  154/2056  1.02  (0.68, 1.51)  0.86  0.17 (0.05)   ADH1C  rs3114046  CC  185/2880  1  (ref)  1073/15297  1.07  (0.84, 1.36)  246/3219  1.16  (0.86, 1.56)  0.33        CT/TT  20/424  1  (ref)  176/2623  1.13  (0.54, 2.38)  52/515  1.88  (0.78, 4.51)  0.10  0.27 (0.08   ADH1C  rs4147542  TT  107/1260  1  (ref)  631/8763  0.85  (0.61, 1.18)  144/1723  0.98  (0.65, 1.48)  0.95        TC/CC  94/1706  1  (ref)  570/7672  1.28  (0.91, 1.79)  149/1575  1.61  (1.06, 2.46)  0.03  0.12 (0.03)   ADH1C  rs283415  TT  65/1243  1  (ref)  385/5654  1.40  (0.94, 2.08)  93/1104  1.66  (1.00, 2.75)  0.05        TC/CC  140/2060  1  (ref)  863/12 266  1.00  (0.75, 1.32)  205/2630  1.12  (0.79, 1.58)  0.46  0.37 (0.12)   ADH1C  rs4699741  TT  179/2738  1  (ref)  1121/15601  1.05  (0.83, 1.35)  258/3230  1.17  (0.86, 1.58)  0.31        TC/CC  26/565  1  (ref)  128/2319  1.19  (0.65, 2.19)  40/503  1.61  (0.75, 3.44)  0.22  0.50 (0.15)  Women   ADH1B  rs1159918  CC  147/3395  1  (ref)  335/7422  1.08  (0.82, 1.43)  20/392  1.43  (0.73, 2.81)  0.38        CA/AA  234/5460  1  (ref)  456/10 666  0.99  (0.79, 1.24)  38/572  1.75  (1.03, 2.97)  0.33  0.68 (0.15)   ADH1B  rs2075633  TT  199/4735  1  (ref)  412/8725  1.13  (0.89, 1.43)  28/570  1.27  (0.72, 2.23)  0.26        TC/CC  181/4120  1  (ref)  382/9420  0.94  (0.73, 1.21)  30/394  2.10  (1.13, 3.91)  0.46  0.16 (0.03)   ADH1B  rs1693439  GG  324/7497  1  (ref)  668/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        GA/AA  57/1358  1  (ref)  126/2793  1.11  (0.69, 1.81)  11/157  1.57  (0.54, 4.60)  0.46  0.96 (0.18)   ADH1B  rs9307239  CC  148/3546  1  (ref)  309/6954  1.08  (0.81, 1.44)  28/377  1.80  (0.95, 3.38)  0.18        CT/TT  233/5292  1  (ref)  483/11 191  0.98  (0.78, 1.22)  30/587  1.39  (0.79, 2.43)  0.71  0.62 (0.13)   ADH1B  rs4147536  CC  236/5557  1  (ref)  513/11 574  1.08  (0.86, 1.34)  32/637  1.37  (0.80, 2.33)  0.30        CA/AA  145/3298  1  (ref)  279/6572  0.94  (0.70, 1.25)  26/328  2.03  (1.03, 3.99)  0.45  0.32 (0.07)   ADH1B  rs3811802  AA  107/2509  1  (ref)  224/5709  0.96  (0.68, 1.37)  22/137  5.72  (2.24, 14.63)  0.13        AG/GG  274/6313  1  (ref)  569/12 457  1.04  (0.85, 1.28)  36/827  1.09  (0.67, 1.77)  0.63  0.004b (0.02)   ADH1C  rs698  TT  139/3189  1  (ref)  273/6696  0.94  (0.70, 1.27)  17/349  1.50  (0.73, 3.06)  0.86        TC/CC  240/5666  1  (ref)  520/11 453  1.09  (0.88, 1.36)  40/615  1.70  (1.02, 2.84)  0.11  0.60 (0.17)   ADH1C  rs1662033  TT  193/4346  1  (ref)  367/8777  0.92  (0.72, 1.19)  25/440  1.59  (0.87, 2.92)  0.84        TG/GG  188/4508  1  (ref)  427/9369  1.12  (0.87, 1.44)  33/525  1.63  (0.92, 2.87)  0.13  0.44 (0.13)   ADH1C  rs3114046  CC  324/7497  1  (ref)  667/15 372  1.04  (0.86, 1.25)  47/808  1.61  (1.02, 2.54)  0.21        CT/TT  57/1358  1  (ref)  127/2793  1.12  (0.69, 1.82)  11/157  1.55  (0.53, 4.54)  0.45  0.95 (0.20)   ADH1C  rs4147542  TT  182/4890  1  (ref)  399/8448  1.25  (0.98, 1.59)  29/458  1.78  (1.01, 3.14)  0.02        TC/CC  180/3474  1  (ref)  356/8747  0.78  (0.60, 1.02)  28/405  1.74  (0.92, 3.31)  0.58  0.01b (0.02)   ADH1C  rs283415  TT  134/2923  1  (ref)  245/6263  0.87  (0.64, 1.19)  14/324  1.27  (0.59, 2.72)  0.67        TC/CC  247/5932  1  (ref)  549/11 903  1.12  (0.90, 1.39)  44/640  1.83  (1.11, 3.02)  0.05  0.19 (0.07)   ADH1C  rs4699741  TT  326/7548  1  (ref)  692/16 037  1.00  (0.83, 1.20)  50/850  1.52  (0.97, 2.37)  0.39        TC/CC  55/1307  1  (ref)  102/2129  1.43  (0.83, 2.46)  8/114  2.08  (0.61, 7.12)  0.12  0.82 (0.18)  CI, confidence interval; HR, hazard ratio; NLCS, Netherlands Cohort Study; PT, person-time; ref, reference. aAdjusted for age, BMI, smoking, family history of CRC, meat intake, processed meat intake, folate intake and non-occupational physical activity. bSignificant after adjusting for multiple testing. View Large Supplementary Tables 4–7, available at Carcinogenesis Online, show the associations between alcohol intake and the risk of CRC by subsite, i.e. the colon, proximal colon, distal colon and rectum, in men and women, stratified by genotype. After FDR correction, the only statistically significant interactions observed were those between alcohol intake and ADH1B rs3811802 and ADH1C rs4147542 in relation to (proximal) colon cancer in women, consistent with the interactions observed for CRC in women. In men, a difference in associations was observed between genotype strata, when considering alcohol intake in relation to the risk of colon and proximal colon cancer, and, in particular, rectal cancer, but less so or not in relation to distal colon cancer. In women, the power was limited in analyses for distal colon and rectal cancer, hampering a proper comparison. Discussion This study addressed the lack of evidence regarding alcohol intake and CRC risk in women and found alcohol to be a CRC risk factor in men and women. Associations with CRC differed by sex. Alcohol intake increased CRC risk (non-)significantly at light-moderate and heavy intake levels across subsites in men. Only when alcohol intake exceeded 30 g/day, we observed increased colon cancer risks, particularly for the proximal colon, in women. We studied associations in ADH1B and ADH1C genetic subgroups, because these may be particularly susceptible to the deleterious effects of alcohol on CRC risk. ADH1B rs3811802 and ADH1C rs4147542 modified the association between alcohol intake and the risk of colorectal, colon and proximal colon cancer in women after FDR correction. The alcohol–CRC associations observed in genotype strata generally aligned with overall alcohol–CRC associations. A difference in associations between genotype strata was generally clearer in men than women but significant effect modification was only present in women. Restricting to participants with equal alcohol intake amounts 5 years before baseline resulted in (more) significant positive linear trends across alcohol intake categories within genotype strata in men but not women. Furthermore, ADH1B rs4147536 and ADH1C rs283415—which was in strong linkage disequilibrium (LD) with the commonly investigated ADH1C rs698 (r2 = 0.9) in our data—were associated with an increased cancer risk at colon subsites in male and female drinkers, respectively, after FDR correction. ADH1B rs3811802 and ADH1C rs4147542 were associated with a decreased CRC risk in female heavy and stable drinkers, respectively, after FDR correction. These results substantiate the interplay between alcohol intake, ADH1B and ADH1C in relation to CRC risk. A potential sex difference in intake levels at which alcohol intake increases CRC risk may in part be explained by differences in drinking pattern. Men perhaps are more likely to be more regular consumers than women. Regular alcohol exposure may be especially deleterious and may also increase CRC risk at light-moderate levels (33). In addition, sex differences in first-pass metabolism of alcohol (i.e. presystemic elimination of ethanol in, predominantly, the stomach and liver (34)) and ADH activity could lead to sex differences in the CRC risk associated with alcohol intake (35). Women have prolonged, higher blood ethanol concentrations than men upon similar intake levels due to differences in elimination of alcohol (i.e. the volume distribution is higher in men than women) (35). However, based on this, one would expect women to be affected by alcohol at lower intake levels than men, whereas we found a non-significant association with CRC risk at light-moderate alcohol intake levels in men but not women. A more plausible explanation, therefore, may be that there are interactions between alcohol intake and sex-specific factors. For example, women might be protected from the adverse effects of alcohol at light-moderate intake levels through a positive relationship between alcohol intake and estradiol levels (35). Increased estradiol levels were found to be protective against CRC in women (36,37). Alternatively, as suggested by Klatsky et al. (38), an increased risk of cancer among light-moderate drinkers may be due in part to the underreporting of heavy alcohol intake. This could explain the associations observed with light-moderate alcohol intake in men as the percentage of heavy drinkers is higher in men than women (even though underreporting may be expected more in women than men due to social desirability standards). Finally, the fact that a sex difference was very pronounced in relation to rectal cancer could indicate that there may have been some residual confounding by smoking, even after adjusting for smoking. Smoking has been more strongly associated with rectal cancer than colon cancer risk (39), and smoking correlated more strongly with alcohol intake in men than women [88% of male drinkers as compared to 48% of female drinkers in the subcohort were ever smokers]. We found that the association between alcohol intake and (proximal) colon cancer risk in women was significantly modified by ADH1B rs3811802 and ADH1C rs4147542 after FDR correction. The SNPs selected in this study were not selected on the basis of that these were strong causal variants per se, but on the basis of that these common SNPs (minor allele frequency >5%) tag the genetic variation in ADH1B and ADH1C. Considering that tagSNPs generally confer only minor risks, which may be explained by imperfect correlations with true causal variants and gene–gene interactions, it is difficult to show significance in gene-environment interaction studies, even with large sample sizes (40). Therefore, it may be considered remarkable that two SNPs modified the association between alcohol intake and (proximal) colon cancer risk in women after FDR correction. Especially ADH1C rs4147542 is noteworthy in this regard since it can be linked to functional evidence: it is an expression quantitative trait locus (eQTL) for ADH1C in several tissues including the transverse colon (41), and has been reported to be a methylation quantitative trait locus (mQTL) (42). DNA methylation might, in part, underlie our finding of a gene–environment interaction between rs4147542 and alcohol intake which was specific to proximal colon cancer risk and to women, in which the CpG island methylator phenotype (CIMP) is present more often (43,44). Curtin et al. (45) showed that ADH1C rs698 was associated with CRCs positive for CIMP in those with low folate intake. Alcohol may influence DNA methylation levels via influencing one-carbon metabolism (46), and folate is an important methyl donor. On the other hand, other studies have not specifically linked alcohol intake to CIMP in CRC (47–49). Of the four other tagSNPs (ADH1B rs4147536 and rs3811802 and ADH1C rs283415 and rs4147542) that were associated with CRC risk in drinkers—also suggesting interplay between alcohol intake, ADH1B and ADH1C, and CRC risk—rs283415 is in strong LD with the commonly investigated rs698, for which functional evidence is available. ADH1C rs698 C-allele carriers who also carry the rs1693482 A-allele, encoding Ile350Val and Arg272Gln substitutions, respectively (50), have a ~2.5 times slower alcohol metabolizing rate (51,52) and were found to be at an increased risk of alcohol dependence in Asian populations (7). However, the evidence linking rs698 to cancer was judged inconclusive by IARC due to the small number of studies (1). A meta-analysis of 35 case–control studies comparing rs698 slow with faster alcohol metabolizers found an association with the risk of cancer overall in African and Asian but not European populations (53). This suggests ethnicity is an important factor to take into account. For example, in Caucasian populations there is uncertainty around whether slow or fast alcohol metabolizers are at an increased CRC risk. Although Caucasians carry ADH alleles affecting ethanol oxidation, they lack ADH alleles causing very fast oxidation of ethanol and also lack ALDH alleles causing slow oxidation of acetaldehyde. As such, conflicting results may have emerged from Caucasian studies on rs698 (54,55) in the absence of data on how much ethanol or acetaldehyde accumulates in the colorectum, as explained in the introduction. Strengths of the present study include the population-based prospective design and long follow-up, yielding large case numbers and making selection and information bias unlikely. In addition, the NLCS contains information on alcohol intake at baseline as well as 5 years before baseline, allowing us to investigate whether drinking patterns or fluctuations in alcohol intake affected the studied associations. Information on potential confounders was based on a single baseline measurement, and although changes over time cannot be excluded, the NLCS population has been found stable in its dietary habits (16). Importantly, the elaborate available baseline information enabled us to adjust for a large set of relevant confounders. This is essential considering that individuals who consume higher levels of alcohol may in general have an unhealthier lifestyle than those who have lower intake levels. Furthermore, the high genotyping quality may also be considered as a strength. In conclusion, as opposed to men who showed increased CRC risks across subsites and alcohol intake levels, alcohol intake only increased colon cancer risk in women and only at heavy intake levels. ADH1B rs3811802 and ADH1C rs4147542 modified the alcohol–CRC association in women. These data indicate that alcohol may be a particularly important CRC risk factor in specific genetic subgroups. Previous literature indicates a functional role of rs4147542, supporting our finding of an effect of this variant on alcohol-associated colorectal carcinogenesis and strengthening our confidence in covering relevant genetic variation in ADH1B and ADH1C. Supplementary material Supplementary data are available at Carcinogenesis online. Funding This work was supported by the European Foundation for Alcohol Research (EA 14 39 to C.C.J.M.S., P.v.d.B. and M.P.W.), the Biobanking and Biomolecular Research Infrastructure Netherlands (to M.P.W.) and the Health Foundation Limburg (to M.P.W.). Abbreviations ADH alcohol dehydrogenase ALDH aldehyde dehydrogenase CI confidence interval CRC colorectal cancer FFQ food frequency questionnaire FDR false discovery rate FFQ food frequency questionnaire NLCS Netherlands Cohort Study SNP single nucleotide polymorphism Acknowledgements We are indebted to the participants of this study and wish to thank the Netherlands Cancer Registry and the Netherlands nationwide registry of pathology (PALGA). We also thank Drs. A. Volovics, and A. Kester for statistical advice; S. van de Crommert, H. Brants, J. Nelissen, C. de Zwart, M. Moll, W. van Dijk and A. Pisters for data management; H. Hoofs, H. van Montfort, T. van Moergastel, L. van den Bosch, R. Schmeitz and J. Berben for programming assistance; L. Jonkers, J. Goessens, K. Lemmens and S. Lumeij for the laboratory work involved; and the Biobank Maastricht UMC+ for sample storage. Conflict of Interest Statement: None declared. References 1. International Agency for Research on Cancer. ( 2012) IARC Monographs on the Evaluation of Carcinogenic Risks to Humans: Personal Habits and Indoor Combustions: Consumption of Alcoholic Beverages . International Agency for Research on Cancer, Lyon, France. PubMed PubMed  2. WCRF|Continuous Update Project. Alcohol and Cancer. http://www.wcrf.org/int/cancer-facts-figures/link-between-lifestyle-cancer-risk/alcohol-cancer ( 18 December 2017, date last accessed). 3. Rehm, J.et al.  ( 2003) The global distribution of average volume of alcohol consumption and patterns of drinking. Eur. Addict. Res ., 9, 147– 156. Google Scholar CrossRef Search ADS PubMed  4. WCRF|Continuous Update Project. Colorectal cancer . http://www.wcrf.org/sites/default/files/Colorectal-Cancer-2011-Report.pdf ( 18 December 2017, date last accessed). 5. Fedirko, V.et al.  ( 2011) Alcohol drinking and colorectal cancer risk: an overall and dose-response meta-analysis of published studies. Ann. Oncol ., 22, 1958– 1972. Google Scholar CrossRef Search ADS PubMed  6. Baan, R.et al.  ; WHO International Agency for Research on Cancer Monograph Working Group. ( 2007) Carcinogenicity of alcoholic beverages. Lancet. Oncol ., 8, 292– 293. Google Scholar CrossRef Search ADS PubMed  7. Hurley, T.D.et al.  ( 2012) Genes encoding enzymes involved in ethanol metabolism. Alcohol Res ., 34, 339– 344. Google Scholar PubMed  8. Seitz, H.K.et al.  ( 2007) Molecular mechanisms of alcohol-mediated carcinogenesis. Nat. Rev. Cancer , 7, 599– 612. Google Scholar CrossRef Search ADS PubMed  9. Zerbino, D.Ret al.  (2018) Ensembl 2018. Nucleic Acids Res., 46, D754–D761. doi:10.1093/nar/gkx1098. 10. Ferrari, P.et al.  ( 2012) Alcohol dehydrogenase and aldehyde dehydrogenase gene polymorphisms, alcohol intake and the risk of colorectal cancer in the European Prospective Investigation into Cancer and Nutrition study. Eur. J. Clin. Nutr ., 66, 1303– 1308. Google Scholar CrossRef Search ADS PubMed  11. Martínez, C.et al.  ( 2010) Variability in ethanol biodisposition in whites is modulated by polymorphisms in the ADH1B and ADH1C genes. Hepatology , 51, 491– 500. Google Scholar CrossRef Search ADS PubMed  12. Jokelainen, K.et al.  ( 1994) In vitro acetaldehyde formation by human colonic bacteria. Gut , 35, 1271– 1274. Google Scholar CrossRef Search ADS PubMed  13. Jokelainen, K.et al.  ( 1996) In vitro alcohol dehydrogenase-mediated acetaldehyde production by aerobic bacteria representing the normal colonic flora in man. Alcohol. Clin. Exp. Res ., 20, 967– 972. Google Scholar CrossRef Search ADS PubMed  14. Pestalozzi, D.M.et al.  ( 1983) Immunohistochemical localization of alcohol dehydrogenase in the human gastrointestinal tract. Gastroenterology , 85, 1011– 1016. Google Scholar PubMed  15. Seitz, H.K.et al.  ( 1996) Alcohol dehydrogenase in the human colon and rectum. Digestion , 57, 105– 108. Google Scholar CrossRef Search ADS PubMed  16. van den Brandt, P.A.et al.  ( 1990) A large-scale prospective cohort study on diet and cancer in The Netherlands. J. Clin. Epidemiol ., 43, 285– 295. Google Scholar CrossRef Search ADS PubMed  17. Bongaerts, B.W.et al.  ( 2008) Alcohol consumption, type of alcoholic beverage and risk of colorectal cancer at specific subsites. Int. J. Cancer , 123, 2411– 2417. Google Scholar CrossRef Search ADS PubMed  18. Bongaerts, B.W.et al.  ( 2011) Alcohol consumption, alcohol dehydrogenase 1C (ADH1C) genotype, and risk of colorectal cancer in the Netherlands Cohort Study on diet and cancer. Alcohol , 45, 217– 225. Google Scholar CrossRef Search ADS PubMed  19. Gezondheidsraad( 2015) Richtlijnen goede voeding 2015 . Den Haag: Gezondheidsraad: publicatienr. 2015/24. https://www.gezondheidsraad.nl/sites/default/files/201524_richtlijnen_goede_voeding_2015.pdf ( 18 December 2017, date last accessed). 20. Prentice, R.L. ( 1986) A case-cohort design for epidemiologic cohort studies and disease prevention trials. Biometrika , 73, 1– 11. Google Scholar CrossRef Search ADS   21. Casparie, M.et al.  ( 2007) Pathology databanking and biobanking in The Netherlands, a central role for PALGA, the nationwide histopathology and cytopathology data network and archive. Cell. Oncol ., 29, 19– 24. Google Scholar PubMed  22. Van den Brandt, P.A.et al.  ( 1990) Development of a record linkage protocol for use in the Dutch Cancer Registry for Epidemiological Research. Int. J. Epidemiol ., 19, 553– 558. Google Scholar CrossRef Search ADS PubMed  23. Goldbohm, R.A.et al.  ( 1994) Estimation of the coverage of Dutch municipalities by cancer registries and PALGA based on hospital discharge data. Tijdschr Soc Gezondheidsz , 72, 80– 4. 24. van Breda, S.G.et al.  ( 2007) Toenails: an easily accessible and long-term stable source of DNA for genetic analyses in large-scale epidemiological studies. Clin. Chem ., 53, 1168– 1170. Google Scholar CrossRef Search ADS PubMed  25. Hogervorst, J.G.et al.  ( 2014) DNA from nails for genetic analyses in large-scale epidemiologic studies. Cancer Epidemiol. Biomarkers Prev ., 23, 2703– 2712. Google Scholar CrossRef Search ADS PubMed  26. Goldbohm, R.A.et al.  ( 1994) Validation of a dietary questionnaire used in a large-scale prospective cohort study on diet and cancer. Eur. J. Clin. Nutr ., 48, 253– 265. Google Scholar PubMed  27. Goldbohm, R.A.et al.  ( 1995) Reproducibility of a food frequency questionnaire and stability of dietary habits determined from five annually repeated measurements. Eur. J. Clin. Nutr ., 49, 420– 429. Google Scholar PubMed  28. de Bakker, P.I.et al.  ( 2005) Efficiency and power in genetic association studies. Nat. Genet ., 37, 1217– 1223. Google Scholar CrossRef Search ADS PubMed  29. Simons, C.C.et al.  ( 2013) Physical activity, occupational sitting time, and colorectal cancer risk in the Netherlands cohort study. Am. J. Epidemiol ., 177, 514– 530. Google Scholar CrossRef Search ADS PubMed  30. Benjamini, Y.et al.  ( 1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B Stat. Methodol ., 57, 289– 300. 31. Benjamini, Y.et al.  ( 2001) Controlling the false discovery rate in behavior genetics research. Behav. Brain Res ., 125, 279– 284. Google Scholar CrossRef Search ADS PubMed  32. Smith, N.L.et al.  ( 2007) Association of genetic variations with nonfatal venous thrombosis in postmenopausal women. JAMA , 297, 489– 498. Google Scholar CrossRef Search ADS PubMed  33. Cao, Y.et al.  ( 2015) Light to moderate intake of alcohol, drinking patterns, and risk of cancer: results from two prospective US cohort studies. BMJ , 351, h4238. Google Scholar CrossRef Search ADS PubMed  34. Oneta, C.M.et al.  ( 1998) First pass metabolism of ethanol is strikingly influenced by the speed of gastric emptying. Gut , 43, 612– 619. Google Scholar CrossRef Search ADS PubMed  35. Erol, A.et al.  ( 2015) Sex and gender-related differences in alcohol use and its consequences: contemporary knowledge and future research considerations. Drug Alcohol Depend ., 156, 1– 13. Google Scholar CrossRef Search ADS PubMed  36. Lin, J.H.et al.  ( 2013) Association between sex hormones and colorectal cancer risk in men and women. Clin. Gastroenterol. Hepatol ., 11, 419– 424.e1. Google Scholar CrossRef Search ADS PubMed  37. Roshan, M.H.et al.  ( 2016) The role of testosterone in colorectal carcinoma: pathomechanisms and open questions. EPMA J ., 7, 22. Google Scholar CrossRef Search ADS PubMed  38. Klatsky, A.L.et al.  ( 2014) Moderate alcohol intake and cancer: the role of underreporting. Cancer Causes Control , 25, 693– 699. Google Scholar CrossRef Search ADS PubMed  39. Liang, P.S.et al.  ( 2009) Cigarette smoking and colorectal cancer incidence and mortality: systematic review and meta-analysis. Int. J. Cancer , 124, 2406– 2415. Google Scholar CrossRef Search ADS PubMed  40. Simonds, N.I.et al.  ( 2016) Review of the gene-environment interaction literature in cancer: what do we know? Genet. Epidemiol ., 40, 356– 365. Google Scholar CrossRef Search ADS PubMed  41. GTEx Portal. https://gtexportal.org/home/eqtls/bySnp?snpId=rs4147542&tissueName=All ( 18 December 2017, date last accessed). 42. Zhang, H.et al.  ( 2014) Identification of methylation quantitative trait loci (mQTLs) influencing promoter DNA methylation of alcohol dependence risk genes. Hum. Genet ., 133, 1093– 1104. Google Scholar CrossRef Search ADS PubMed  43. Hawkins, N.et al.  ( 2002) CpG island methylation in sporadic colorectal cancers and its relationship to microsatellite instability. Gastroenterology , 122, 1376– 1387. Google Scholar CrossRef Search ADS PubMed  44. Weisenberger, D.J.et al.  ( 2006) CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer. Nat. Genet ., 38, 787– 793. Google Scholar CrossRef Search ADS PubMed  45. Curtin, K.et al.  ( 2007) Genetic polymorphisms in one-carbon metabolism: associations with CpG island methylator phenotype (CIMP) in colon cancer and the modifying effects of diet. Carcinogenesis , 28, 1672– 1679. Google Scholar CrossRef Search ADS PubMed  46. Varela-Rey, M.et al.  ( 2013) Alcohol, DNA methylation, and cancer. Alcohol Res ., 35, 25– 35. Google Scholar PubMed  47. Razzak, A.A.et al.  ( 2011) Alcohol intake and colorectal cancer risk by molecularly defined subtypes in a prospective study of older women. Cancer Prev. Res. (Phila) ., 4, 2035– 2043. Google Scholar CrossRef Search ADS PubMed  48. Schernhammer, E.S.et al.  ( 2011) B vitamins, methionine and alcohol intake and risk of colon cancer in relation to BRAF mutation and CpG island methylator phenotype (CIMP). PLoS One , 6, e21102. Google Scholar CrossRef Search ADS PubMed  49. Slattery, M.L.et al.  ( 2007) Diet and lifestyle factor associations with CpG island methylator phenotype and BRAF mutations in colon cancer. Int. J. Cancer , 120, 656– 663. Google Scholar CrossRef Search ADS PubMed  50. Osier, M.V.et al.  ( 2002) A proline-threonine substitution in codon 351 of ADH1C is common in Native Americans. Alcohol. Clin. Exp. Res ., 26, 1759– 1763. Google Scholar CrossRef Search ADS PubMed  51. Bosron, W.F.et al.  ( 1986) Genetic polymorphism of human liver alcohol and aldehyde dehydrogenases, and their relationship to alcohol metabolism and alcoholism. Hepatology , 6, 502– 510. Google Scholar CrossRef Search ADS PubMed  52. Höög, J.O.et al.  ( 1986) The gamma 1 and gamma 2 subunits of human liver alcohol dehydrogenase. cDNA structures, two amino acid replacements, and compatibility with changes in the enzymatic properties. Eur. J. Biochem ., 159, 215– 218. Google Scholar CrossRef Search ADS PubMed  53. Xue, Y.et al.  ( 2012) ADH1C Ile350Val polymorphism and cancer risk: evidence from 35 case-control studies. PLoS One , 7, e37227. Google Scholar CrossRef Search ADS PubMed  54. Giovannucci, E.et al.  ( 2003) Methylenetetrahydrofolate reductase, alcohol dehydrogenase, diet, and risk of colorectal adenomas. Cancer Epidemiol. Biomarkers Prev ., 12, 970– 979. Google Scholar PubMed  55. Homann, N.et al.  ( 2009) Alcohol and colorectal cancer: the role of alcohol dehydrogenase 1C polymorphism. Alcohol. Clin. Exp. Res ., 33, 551– 556. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

Journal

CarcinogenesisOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off