Interaction between vitamin E intake and a COMT gene variant on colorectal cancer risk among Korean adults: a case-control study

OBJECTIVES Previous human trials have not supported the anticarcinogenic effect of vitamin E despite biological plausibility and considerable epidemiological evidence. A possible explanation for this inconsistency is the interactive effect of the catechol-O-methyltransferase (COMT) gene and supplemental vitamin E on cancer. We examined whether a COMT gene variant modulates the effect of dietary vitamin E intake on colorectal cancer (CRC) risk. METHODS In this case-control study of Korean adults (975 cases and 975 age- and sex-matched controls), dietary vitamin E density (mg/1,000 kcal) was measured using a semiquantitative food frequency questionnaire, COMT single nucleotide polymorphism (SNP) rs740603 (A>G) was genotyped, and CRC was verified histologically. We estimated odds ratios (ORs) and 95% confidence intervals (CIs) using unconditional logistic regression models with adjustments for potential confounders. RESULTS Higher vitamin E density was associated with a lower risk of CRC (highest vs. lowest quartiles: OR, 0.72; 95% CI, 0.55 to 0.96; p-for-trend=0.002). When stratified by COMT SNP rs740603 genotype, the inverse association between vitamin E density and CRC risk was confined to those with at least 1 A allele (≥median vs. <median: OR, 0.63; 95% CI, 0.51 to 0.78). The interaction between rs740603 and vitamin E density was significant (p-for-interaction=0.020). No direct association was observed between COMT SNP rs740603 and CRC risk (OR, 1.08; 95% CI, 0.83 to 1.41). CONCLUSIONS Our findings support a role for a genetic polymorphism in COMT in modifying the association between dietary vitamin E intake and CRC.

Interaction between vitamin E intake and a COMT gene variant on colorectal cancer risk among Korean adults: a case-control study zyme activity may have confounded the association between vitamin E and cancer risk [9].In 2 large randomized controlled trials of vitamin E (alpha-tocopherol) supplementation in populations of European ancestry, polymorphisms in the catechol-O-methyltransferase (COMT) gene significantly modified the effect of vitamin E on total cancer incidence [9].The interactive effects did not reach statistical significance when focusing on specific cancer types, but the pattern of vitamin E being beneficial only among rs4680 Met-allele homozygotes was generally consistent for breast cancer and colorectal cancer (CRC) [9].Questions for further research included whether the pharmacogenetic effect of vitamin E and COMT on cancer prevention extends to dietary vitamin E intake and to specific cancer types.
Therefore, we examined whether a polymorphism in COMT modulates the effect of dietary vitamin E intake on CRC risk in a case-control study of Korean adults.CRC is the third most common cancer both in Korea and worldwide, and effective and feasible prevention strategies are necessary [10].We focused on the single-nucleotide polymorphism (SNP) rs740603 (A > G), which was shown to modify the effect of dietary vitamin E intake on waist circumference in a Spanish case-control study [11].

Study population
CRC case-control data were collected at the National Cancer Center Korea.Case participants were recruited among patients who were newly diagnosed with adenocarcinoma of the colon or rectum based on endoscopic biopsies.There were 2 rounds of data collection: first, from August 2010 to August 2013; and second, from January 2018 to September 2020.Control participants were recruited from cancer-free individuals who came for a health checkup at the National Cancer Center Korea from October 2007 to December 2021 [12].
The study participant selection process is described in Figure 1.From 1,492 cases and 10,707 controls who completed the general questionnaire and food frequency questionnaire (FFQ), we were able to successfully obtain rs740603 genotype information in 1,119 cases and 1,443 controls.We then excluded those with implausible energy intake ( < 500 or > 4,000 kcal/day regardless of sex [13,14]; 8 cases and 1 control), missing body mass index (BMI) information (1 case and 84 controls), or missing education information (1 case and 54 controls).Among the remaining 1,109 cases and 1,304 controls, controls were matched to cases by sex and 5-year age group at a 1:1 ratio using a published macro [15].As a result, 975 cases and 975 controls were matched successfully and were included in our analysis.

Data collection
Trained interviewers administered the general questionnaire on socio-demographic, lifestyle, and health characteristics and the 106-item semiquantitative FFQ.To minimize recall bias, we contacted cases within a few days after the first hospital admission and asked them to report their lifestyles and habitual diets before CRC diagnosis.The FFQ was developed to assess Korean adults' diets and has been validated against 12-day dietary records across  seasons [16].Daily energy intake and vitamin E intake (i.e., sum of α-, β-, γ-, δ-tocopherols and α-, β-, γ-, δ-tocotrienols) were calculated using the Computer Aided Nutrition Analysis Program version 4 (Can-Pro 4.0; Korean Nutrition Society, Seoul, Korea).
To adjust for energy and measurement errors associated with the FFQ [17], vitamin E density was calculated as the ratio of dietary vitamin E intake to total dietary energy intake (mg/1,000 kcal).Vitamin C density was calculated using the same method.We also calculated the Korea Healthy Eating Index, which represents overall dietary quality by measuring adherence to the Dietary Guidelines for Koreans [18].Anthropometric measurements and a blood draw were conducted by trained health technicians on the same day or within a few days of the administration of the questionnaire.BMI was calculated as weight in kilograms divided by the square of height in meters.Obesity was defined as BMI> 25 kg/m 2 based on the Asianspecific cut-off [19].If height and body weight measurements were missing (567 cases and 376 controls), self-reported height and weight were used to calculate the BMI.Among 599 participants who had both measured and self-reported anthropometry data, the correlation coefficients between self-reported and measured height and weight were very high ( ≥ 0.98).

Single-nucleotide polymorphism genotyping
From blood samples, DNA was extracted using the MagAttract DNA Blood M48 Kit (Qiagen, Hilden, Germany) and the BioRobot M48 automatic extraction equipment (Qiagen).SNP genotyping was performed using the Illumina MEGA-Expanded Array (Illumina Inc., Hayward, CA, USA) consisting of 123K SNPs.To increase genome coverage, we imputed untyped SNPs, including rs740603, using the Michigan imputation server with the 1000 Genome Project phase 3 East Asian ancestry as a reference panel.

Statistical analysis
Descriptive statistics for cases and controls are presented as numbers (%) for categorical variables or means ± standard deviations for continuous variables.The significance of differences between cases and controls was determined based on the chi-square test for categorical variables and the t-test for continuous variables.
We used unconditional logistic regression models with adjustment for covariates selected based on the previous literature, biological plausibility, and collinearity to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association between vitamin E density or the SNP rs740603 and CRC risk.Vitamin E density was divided into quartiles based on the distribution of controls, and the rs740603 genotype was combined based on a dominant model, with the A allele regarded as a dominant allele because the presence of the A allele was associated with a lower waist circumference and interacted with vitamin E intake [11].We then examined diet-gene interactions by conducting logistic regression analyses stratified by the rs740603 genotype and assessing the statistical significance of the interaction by using the Wald test for cross-product terms of vitamin E density and the rs740603 genotype.Vitamin E density was divided into 2 groups ( < and ≥ median based on the distribution in controls) to secure sufficient statistical power.Model 1 included matching factors (age and sex), total energy intake, and first-degree family history of CRC.Model 2 additionally included socioeconomic and lifestyle factors, such as smoking, drinking, education, and obesity.
In sensitivity analyses, the rs740603 genotype was also tested in both recessive and codominant models (Supplementary Materials 1 and 2).Given the possible synergy between vitamins E and C [20], we also examined joint associations of vitamin E density ( < and ≥ median based on the distribution in controls) and vitamin C density ( < and ≥ median based on the distribution of controls) with CRC risk stratified by rs740603 genotype (Supplementary Material 3).We also tested the inclusion of the Korean Healthy Eating Index score in addition to all covariates in model 2, but did not present the results due to concerns about overadjustment bias.
All analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).All statistical tests were 2-sided and used a significance level of 0.05.

Ethics statement
This study was approved by the Institutional Review Board of the National Cancer Center Korea (No. NCC 2022-0118), and written informed consent was obtained from all study participants.

RESULTS
The general characteristics of the case participants and control participants are presented in Table 1.Compared with controls, cases were slightly younger, were less likely to have college degrees and be never-drinkers, and were more likely to have a first-degree family history of CRC.Cases had higher intakes of total energy and vitamin E but had a lower vitamin E density than controls.No significant differences in smoking, obesity, Korean Healthy Eating Index score, or rs740603 genotype between cases and controls were noted.
We first examined the association between vitamin E density and the risk of CRC (Table 2).Statistically significant inverse associations between vitamin E density and CRC risk were observed when adjusted for age, sex, total energy intake, and first-degree family history of CRC without socioeconomic and lifestyle factors (model 1; Q4 vs. Q1: OR, 0.60; 95% CI, 0.46 to 0.78; p-for-trend < 0.001) or with socioeconomic and lifestyle factors (model 2; Q4 vs. Q1: OR, 0.72; 95% CI, 0.55 to 0.96; p-for-trend = 0.002).
We also examined the relationship between COMT SNP rs740603 (G/G vs. A/A or A/G; dominant model) and the risk of CRC (Table 3).No significant association was observed (model 2; OR, 1.08; 95% CI, 0.83 to 1.41).No significant association was noted in either the recessive or codominant model (Supplementary Material 1).
We then examined whether the association between vitamin E density and CRC risk varied according to the SNP rs740603 genotype (Table 4).The inverse association was confined to those with at least one A allele (model 2; p-for-interaction= 0.02).Compared with lower vitamin E density ( < median), higher vitamin E density ( ≥ median) was associated with an OR of 0.63 (model 2; 95% CI, 0.51 to 0.78) among those with at least 1 A allele (A/A or A/G).Vitamin E density was not associated with CRC risk among those with the G/G genotype (model 2; OR, 1.13; 95% CI, 0.66 to 1.91).The results were consistent in the codominant model, with significant associations between vitamin E density and CRC risk only noted in those with A/A and A/G genotypes (Supplementary Material 2).In the recessive model, significant associations be-tween vitamin E density and CRC risk were observed not only in those with the A/A genotype but also in those with the G/G or A/     3  1.00 (reference) 0.88 (0.54, 1.43) 0.611 Model 2 41.00 (reference) 1.13 (0.66, 1.91) 0.662 A/A+A/G No. of cases/controls 498/398 321/420 Model 1 3  1.00 (reference) 0.56 (0.46, 0.69) <0.001 Model 2 4  1.00 (reference) 0.63 (0.51, 0.78) <0.001 Values are presented as odds ratio (95% confidence interval).CRC, colorectal cancer; COMT, catechol-O-methyltransferase; SNP, single-nucleotide polymorphism.
1 Dominant model.Lower and higher vitamin E groups had vitamin E density at or below and above the median (5.03 mg/1,000 kcal), respectively.
3 Adjusted for age, sex, total energy intake and first-degree family history of CRC; p-for-interaction=0.060.
G genotype, which may have been driven by a significant association among those with the A/G genotype.
In the joint analysis of vitamin C and vitamin E density, vitamin E largely drove the associations.Briefly, significant associations were observed only between the higher dietary vitamin E density group, with either lower or higher vitamin C density, and CRC risk among participants with at least 1 A allele (Supplementary Material 3).

DISCUSSION
Despite biological plausibility [1] and considerable epidemiological evidence [6,[21][22][23] suggesting that vitamin E plays a protective role in CRC development [8], human trials have not supported the anticarcinogenic effect of supplemental vitamin E. A possible explanation for this inconsistency is the interactive effect of COMT and supplemental vitamin E on cancer [9].Our results suggest that the interaction observed among participants of European ancestry [9] may be applicable to dietary vitamin E and CRC among participants of Asian ethnicity.In our study of Korean adults, dietary vitamin E intake was protective against CRC development only in carriers of the COMT rs740603 A allele, but not in those who were homozygotes for the G allele.
The enzyme encoded by COMT catalyzes the O-methylation of various compounds, including catecholamines, catechol estrogens, and dietary polyphenols.Therefore, the COMT enzyme has been implicated in multiple neuropsychiatric disorders [24,25] and metabolic disorders [11,[26][27][28].However, insignificant associations have been reported between COMT and CRC incidence and survival [29,30], suggesting that COMT might interact with environmental factors in modulating carcinogenesis.Indeed, in our study, COMT rs740603 was not directly associated with CRC, but rather modified the association between dietary vitamin E and CRC risk.The SNP rs740603 is located within intron 1 and is approximately 6.1 kb 5' to rs4680, the most studied functional variant.This polymorphism was identified as a potentially functional variant associated with a psychiatric phenotype [31] and had interactive effects with dietary vitamin E intake on waist circumference [11].We were unable to examine whether rs4680 has similar effects to those of rs740603 due to a lack of genotype data, but a weak correlation has been found between rs4680 and rs740603 in East Asian populations (1000 Genomes; R 2 = 0.11) [32].Future studies could investigate whether rs4680 or other variants correlated with rs740603 may be involved in the interaction between rs740603 and dietary vitamin E on CRC.
The mechanisms underlying this interaction between vitamin E and COMT need further investigation, but the interaction may be related to oxidative stress management based on the following data.The knockdown of COMT expression was shown to increase estradiol-induced or catechol estrogen-induced reactive oxygen species, microsatellite instability, and neoplastic transformation of endometrial glandular cells [33].In platelets, the COMT enzyme was shown to methylate quercetin, which is an antioxidant [34].
A recent report on the interactive effect of polymorphisms of IL-10 and COMT may have implications regarding the potential interplay between the dopaminergic system and the inflammatory system [35].However, the possible role of vitamin C in this interaction proposed by a previous study [9] based on its association with the COMT gene [36], as well as its synergistic interaction with vitamin E [20], was not supported in our study, suggesting that vitamin E may be involved in a unique mechanism.
In the absence of stratification by rs740603, we observed that dietary vitamin E intake was inversely associated with CRC risk, which is consistent with many previous epidemiological studies [6,[21][22][23].Because the G allele is a minor allele in the Asian population (MAF of 0.41 in this study sample and 0.40 in the East Asian population from 1000 Genomes), null associations in G allele homozygotes seem diluted when all rs740603 genotypes are combined.Regarding the null findings from previous human trials [8] and some epidemiological studies [37,38], we speculate that the composition (only alpha-tocopherol) and high dose of supplemental vitamin E provided in the trial [1] or the large variation in linkage disequilibrium patterns across populations (especially between European and East Asian populations) [39] may have affected the results.
Our data should be understood within the context of the study's strengths and limitations.The strengths of our study include histological verification of CRC patients and the use of a validated FFQ.In addition, we adjusted for a range of plausible confounders measured by a comprehensive questionnaire and clinical assessments, although there is always a possibility of residual confounding.
One of the major limitations of this study was the absence of other populations to test the replicability of our results.Moreover, the case-control study design employed in our study is susceptible to recall bias and selection bias.To minimize recall bias, we interviewed CRC patients immediately after hospital admission for surgery and asked them to report their habitual diet and lifestyle prior to diagnosis.To minimize selection bias, we selected control participants from those who visited the same hospital for health check-ups supported by the Korea National Screening Program; however, the control participants who voluntarily came for checkups may have been more health-conscious than the case participants.Another major limitation is the lack of supplemental vitamin E intake information.As dietary supplements are used by approximately 25% of Korean adults [40], we suggest that future investigations should be conducted with complete information on dietary and supplemental vitamin E information to confirm our findings.Of note, our vitamin E estimate was a simple sum of four tocopherols (α, β, γ, δ) and four tocotrienols (α, β, γ, δ) in milligrams per day.Future studies may also explore the different chemical forms of vitamin E given the growing attention on the anticarcinogenic effects of γand δ-tocopherols [41].
In conclusion, the COMT genetic variant modified the association between dietary vitamin E intake and CRC; specifically, vitamin E was inversely associated with CRC risk only in carriers of the COMT rs740603 A allele.These findings add key evidence to the literature on nutrient-gene interactions affecting CRC risk and may inform future precision nutrition research and practice in cancer prevention.

Table 1 .
Characteristics of colorectal cancer cases and controls 1Differences between cases and controls were tested using the chi-square test for categorical variables and the t-test for continuous variables.

Table 2 .
Association between vitamin E density and colorectal cancer risk

Table 3 .
Association between the rs740603 genotype and CRC risk Values are presented as odds ratio (95% confidence interval).CRC, colorectal cancer; COMT, catechol-O-methyltransferase; SNP, single-nucleotide polymorphism.1 Dominant model.2Adjustedforage, sex, total energy intake and first-degree family history of CRC.3 Adjusted for age, sex, total energy intake, first-degree family history of CRC, smoking, drinking, education, and obesity.

Table 4 .
Association between vitamin E density and CRC risk, stratified by COMT rs740603 genotype