Warning: fopen(/home/virtual/epih/journal/upload/ip_log/ip_log_2024-11.txt): failed to open stream: Permission denied in /home/virtual/lib/view_data.php on line 95 Warning: fwrite() expects parameter 1 to be resource, boolean given in /home/virtual/lib/view_data.php on line 96 Dietary mercury intake, the IL23R rs10889677 polymorphism, and the risk of gastric cancer in a Korean population: a hospital-based case-control study
Skip Navigation
Skip to contents

Epidemiol Health : Epidemiology and Health

OPEN ACCESS
SEARCH
Search

Articles

Page Path
HOME > Epidemiol Health > Volume 46; 2024 > Article
Original Article
Dietary mercury intake, the IL23R rs10889677 polymorphism, and the risk of gastric cancer in a Korean population: a hospital-based case-control study
Ji Hyun Kim1orcid, Madhawa Gunathilake1orcid, Jeonghee Lee1orcid, Il Ju Choi2orcid, Young-Il Kim2orcid, Jeongseon Kim1orcid

DOI: https://doi.org/10.4178/epih.e2024051
Published online: May 21, 2024

1National Cancer Center Graduate School of Cancer Science and Policy, National Cancer Center, Goyang, Korea

2Center for Gastric Cancer, National Cancer Center Hospital, National Cancer Center, Goyang, Korea

Correspondence: Jeongseon Kim National Cancer Center Graduate School of Cancer Science and Policy, National Cancer Center, 323 Ilsan-ro, Ilsandong-gu, Goyang 10408, Korea E-mail: jskim@ncc.re.kr
• Received: January 30, 2024   • Accepted: April 10, 2024

© 2024, Korean Society of Epidemiology

This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

prev next
  • 1,574 Views
  • 50 Download
  • OBJECTIVES
    Mercury can stimulate immune responses through T helper 17 (Th17). The gene IL23R is a key factor in Th17 function, which may also contribute to digestive tract diseases. The aim of this study was to identify the associations between dietary mercury and gastric cancer (GC) and to investigate whether the IL23R rs10889677 polymorphism modifies those associations.
  • METHODS
    This case-control study included 377 patients with GC and 756 healthy controls. Dietary mercury intake (total mercury and methylmercury) was assessed using a dietary heavy metal database incorporated into the food frequency questionnaire. IL23R genetic polymorphism rs10889677 (A>C) was genotyped. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using unconditional logistic regression models with adjustments for potential confounders.
  • RESULTS
    A higher dietary methylmercury intake was associated with an elevated risk of GC (OR for the highest vs. lowest tertile [T3 vs. T1], 2.02; 95% CI, 1.41 to 2.91; p for trend <0.001). The IL23R rs10889677 reduced the risk of GC in individuals who carried at least 1 minor allele (OR, 0.62; 95% CI, 0.46 to 0.83; p=0.001; AC/CC vs. AA). Individuals with a C allele exhibited a lower susceptibility to GC through methylmercury intake than those with the AA genotype (OR for the T3 of methylmercury and AA carriers, 2.93; 95% CI, 1.77 to 4.87; and OR for the T3 of methylmercury and AC/CC genotype, 1.30; 95% CI, 0.76 to 2.21; p-interaction=0.013).
  • CONCLUSIONS
    Our findings suggest that a genetic polymorphism, rs10889677 in IL23R, plays a role in modifying the association between dietary methylmercury intake and the risk of GC.
This study aimed to investigate the associations between dietary mercury and gastric cancer (GC) and to determine whether the IL23R rs10889677 polymorphism, located within a predicted binding site for microRNA-lethal-7, may modify these associations. A higher dietary methylmercury intake was associated with an increased risk of GC, while the IL23R rs10889677 polymorphism may modify the detrimental effect of dietary methylmercury on gastric carcinogenesis.
Gastric cancer (GC) was one of the most diagnosed cancers and a major cause of cancer-related deaths worldwide in 2020, ranking fourth in incidence and fifth in mortality [1]. Despite a continuous declining trend in GC incidence and improved survival rates, the burden remains high in East Asian countries, particularly in Korea [1,2]. Therefore, it is crucial to conduct further epidemiological studies to better understand the factors underlying gastric carcinogenesis, including environmental and genetic factors.
As an environmental contaminant, mercury is a persistent and toxic heavy metal harmful to health. It can originate from various emission sources, transfer through different environmental mediums, undergo complex biogeochemical cycling, and accumulate in animals and plants [3]. Mercury-induced toxicity is not confined to a single cellular target; it can cause extensive damage throughout the body, including neurotoxicity, nephrotoxicity, and gastrointestinal toxicity, resulting in ulceration and hemorrhage [4-6]. Diet is the predominant route of mercury exposure in humans, with fish being the main source of mercury intake [7-9]. Scientific literature suggests that regular exposure to mercury or its most toxic organic form, methylmercury (derived from methylation of inorganic mercury by microorganisms), may adversely contribute to digestive disorders including chronic gastritis and colorectal cancer [5,10-12].
Oxidative stress is a common pathway through which genotoxicity is induced [13]. The reactive oxygen species triggered by mercury can cause lesions of the gastric mucosa, damage DNA, and subsequently disrupt gene regulation, cell signal transduction, and cell growth, which can ultimately lead to gastric carcinogenesis and metastasis [14]. The induced formation of free radicals has been implicated in promoting inflammation and damage in the context of autoimmunity [15]. In addition, xenobiotics can stimulate immune responses by binding to immune cell receptors, which are expressed by many immune cells (e.g., T helper 17 [Th17] subsets) [16-18]. While the exact etiology remains elusive, heavy metals (a prominent category of xenobiotics) have been implicated in the alteration of immune cell responses in exposed individuals, thereby contributing to disease susceptibility (e.g., autoimmune diseases and cancers), particularly when accompanied by inflammation-induced sensitization [18,19].
Interleukin 23 receptor (IL-23R) is a key player in the pro-inflammatory signal transduction pathway of the IL-23/IL-17 axis (IL-23→IL-23R→STAT3→Th17→IL-17/IL-17F) and is recognized for its crucial role in inflammatory diseases and cancers [20,21]. It has been hypothesized that genetic variants within this pathway modify cancer risk. Several functional variants in the IL23R gene have been reported to impact cancer susceptibility [22]. Notably, rs10889677 is situated in the 3´-untranslated region (3´-UTR) of IL23R and is located within a predicted binding site for microRNA lethal-7 (miR-let-7) [23]. It has been reported that nucleobase substitution (A> C) may increase the binding affinity of miR-let-7 and can further affect posttranscriptional regulation of IL23R [23]. Several types of cancers (e.g., stomach, colorectum, ovary, breast, lung, nasopharynx, and bladder) have been investigated in relation to this genetic polymorphism, and pooled estimates from meta-analyses noted a reduced risk of cancer among AC/CC genotypes when compared to AA genotypes [22-24]. However, the findings regarding this variant in digestive tract cancers remain unclear [20,23,25-29], and previous studies did not include the potential effects of dietary factors when analyzing this variant [27].
To our knowledge, there is a paucity of epidemiologic evidence examining the impact of dietary mercury exposure or IL23R polymorphism on cancer risk. Moreover, the combined effect of dietary mercury intake and the IL23 variant on gastric carcinogenesis has not been investigated thus far. Therefore, this study investigated the association between dietary mercury intake or IL23R rs10889677 polymorphism and GC risk, and explored whether there was an interaction between dietary mercury intake and the IL23R variant in relation to GC.
Study participants
To conduct a case-control study, we recruited participants from the National Cancer Center Hospital (NCC) in Korea between March 2011 and December 2014. Case participants included those diagnosed with GC within 3 months prior to enrollment and confirmed as invasive carcinoma in the NCC’s Center for Gastric Cancer. Patients with severe chronic diseases and pregnant or breastfeeding women were excluded. Participants who underwent health-screening examinations at the NCC’s Cancer Prevention and Detection Center, which confirmed that they did not have a history of cancer or other severe comorbidities, were recruited as controls. We used age and gender to match controls and cases at a ratio of 2:1. A total of 756 controls and 377 cases with available genotype data were included for analysis (Figure 1). Detailed descriptions of participant recruitment have been outlined elsewhere [30].
Data collection
Information on socio-demographic characteristics and lifestyle factors was collected from each participant using a self-administered questionnaire. The status of Helicobacter pylori infection was determined histologically or serologically with at least a positive result on a rapid urease test (Pronto Dry, Medical Instruments Corp., Solothurn, Switzerland). The participants’ dietary intake data over the 12 months prior to their interview was obtained by a well-trained dietary interviewer using a validated 106-item semiquantitative food frequency questionnaire [31]. The determination of daily total energy and detailed food item intake were based on a combination of intake frequency (never or rarely, 1 times/mo, 2-3 times/mo, 1-2 times/wk, 3-4 times/wk, 5-6 times/wk, 1 times/day, 2 times/day, and 3 times/day) and portion size (small, medium, and large), using a Computer Aided Nutritional analysis program (CAN-PRO 4.0, Korean Nutrition Society, Seoul, Korea). The daily food consumption information was subsequently linked to a database of dietary heavy metals, encompassing total mercury in all forms, including methylmercury. The dietary heavy metal database utilized in this study was developed by analyzing heavy metal content in samples of foods predominantly consumed by the Korean population. The samples were collected from markets in Korea and analyzed using methods specified in the Korean Food Standard Codex. The gold amalgamation method was used to detect total mercury, and gas chromatography with an electron capture detector was used to detect methylmercury. The participants’ food consumption data (g/day) were linked to the mercury levels (μg/g) of listed food items to estimate daily mercury consumption (μg/day). A detailed explanation of the database has been provided elsewhere [32,33].
Genotyping and single nucleotide polymorphism selection
Genomic DNA samples were extracted from the peripheral blood leukocytes of all participants. The genotyping was performed using an Affymetrix Axiom® Exome 319 Array containing 318,983 variants (Affymetrix Inc., Santa Clara, CA, USA). Genotype imputation was performed using the Asian population (n=504) in the 1000 Genome haplotypes phase III integrated variant set release GRch37/hg19 (https://www.1000genomes.org/) as a reference panel. We used SHAPEIT (v2.r837) for phasing and IMPUTE2 (2.3.2) for single nucleotide polymorphism (SNP) imputation. After filtering for an imputation quality score (or INFO score) over 0.6, quality control criteria were applied. Detailed information on the genotyping and quality control steps is mentioned elsewhere [34,35].
The following quality control criteria were used for further exclusions: (1) SNPs with a low call rate (<97%), (2) individuals with a low genotype call rate (<97%), (3) SNPs with a minor allele frequency (MAF) <5%, (4) SNPs showing deviation from Hardy-Weinberg equilibrium (p-values <1 × 10-6), and (5) individuals closely related based on a pairwise identity-by-descent proportion (pi-hat >0.25). The current study focused on rs10889677 polymorphism, which is located in the 3´-UTR of IL23R and within a binding site for miR-let-7 [23].
Statistical analysis
We assessed the differences between cases and controls in sociodemographic, anthropometric, and lifestyle factors and in total energy intake using the chi-square test for categorical variables and the Student t-test for continuous variables. Mercury intake was adjusted for energy using the residual method [36]. We classified dietary mercury intake into tertiles based on the distribution among controls. For the genetic association, we employed the codominant, dominant, and allelic inheritance models. To analyze associations among dietary mercury consumption, IL23R rs10889677, and the risk of GC, we constructed unconditional logistic models to estimate the odds ratios (ORs) and 95% confidence intervals (CIs), considering potential confounding factors. The covariates included continuous age and the categorical gender, body mass index (< 23, 23-< 25, or ≥ 25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education attainment (less than college or college and higher), income (< 200, 200-< 400, or ≥ 400 [×10,000 Korean won/mo]), and first-degree family history of GC (yes or no). Because the expression of IL-23 is closely associated with H. pylori infection [37,38], we included it in the final statistical model. Moreover, we investigated the dose-response effects of dietary mercury in relation to GC risk by using the median value within each tertile of dietary mercury intake to identify a test for trend. To test for the effect of the interaction between dietary mercury and the IL23R variant on GC risk, we employed a likelihood ratio test by comparing models with and without the interaction term (tertile categories mercury *SNP). All statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA), and a 2-sided p-value < 0.05 indicated statistical significance.
Ethics statement
This study was approved by the Institutional Review Board of the National Cancer Center Korea (No. NCC 2021-0181), and written informed consent was obtained from all participants prior to the examination.
The general characteristics of the study participants and their dietary mercury intake are presented in Table 1. Compared to healthy participants, patients with GC exhibited a greater prevalence of positive H. pylori infection (92.6 vs. 61.4%), a first-degree family history of GC (20.4 vs. 12.6%), current smoking status (30.8 vs. 20.4%), and education less than college (75.8 vs. 44.2%). When compared to the controls, GC cases showed less adherence to regular physical activity (36.1 vs. 56.1%) and fewer had a monthly income ≥ 4,000,000 Korean won (23.3 vs. 32.7%). The case participants exhibited a higher daily caloric intake (1,925.2±611.9 vs. 1,717.3±546.9 kcal/day) and a greater mean dietary mercury intake (energy-adjusted total mercury: 14.0±2.3 vs. 13.7±2.4 µg/day; and methylmercury: 13.5±3.5 vs. 12.4±3.3 µg/day) than the controls.
The association between dietary mercury intake (total mercury and methylmercury) and GC risk is reported in Table 2. When compared to the lowest tertile (T1) of dietary total mercury, men participants in the highest intake tertile group (T3) were associated with an elevated risk of GC (model III: OR, 1.73; 95% CI, 1.11 to 2.68; p for trend=0.015). However, in participants overall, the increased OR due to dietary total mercury exposure observed in the crude model disappeared after adjusting for potential covariates. Among women, no clear concentration-response effects were observed with increasing tertiles of dietary total mercury. The results indicated an elevated risk of GC in the second tertile (T2) in both crude and fully adjusted models, but not in the T3. In terms of dietary methylmercury intake, an increased risk of GC was observed in participants overall (OR, 2.02; 95% CI, 1.41 to 2.91; p for trend < 0.001). In the gender-stratified analysis, the proportional association trends of methylmercury intake with GC risk remained consistent in both men (OR, 1.77; 95% CI, 1.14 to 2.74; p for trend=0.010) and women (OR, 2.80; 95% CI, 1.40 to 5.62; p for trend=0.001).
The association between IL23R polymorphism and the risk of GC is shown in Table 3. In the dominant model, we observed a reduced GC risk in individuals carrying at least 1 minor allele C of IL23R rs10889677 polymorphism (OR, 0.62; 95% CI, 0.46 to 0.83; AC/CC vs. AA). When stratifying the participants by gender, the inverse associations remained statistically significant in both men (OR, 0.61; 95% CI, 0.42 to 0.88; AC/CC vs. AA) and women (OR, 0.60; 95% CI, 0.36 to 0.99). In the allelic model, a reduced risk of GC was also observed among participants overall and men specifically, in those carrying the C allele (OR, 0.74; 95% CI, 0.59 to 0.93 for overall; and OR, 0.71; 95% CI, 0.53 to 0.95 for men). In the codominant model, a decreased risk of GC was found in the AC genotype, but not in the CC genotype, when compared to the AA genotype. This trend persisted among overall participants as well as in both men and women (Supplementary Material 1).
We also investigated the associations between dietary mercury intake and the risk of GC using the dominant model of the IL23R rs10889677 variant, with the T1 of mercury intake and the homozygous AA genotype set as the reference. In the analysis of total mercury intake among overall participants, non-significant associations were identified for all groups based on tertiles of total mercury intake and the dominant model, as well as for the interaction between total mercury and the IL23R rs10889677 variant in relation to GC. This trend persisted in the gender-stratified analysis, except for a borderline significant reduction in the risk of GC noted among men with the lowest total mercury and AC/CC genotypes versus the AA genotype (Table 4). In the analysis of dietary methylmercury intake among total participants, individuals carrying the C allele showed a comparatively lower susceptibility to GC from methylmercury than those with the AA genotype (OR for the T3 of methylmercury and AA genotype, 2.93; 95% CI, 1.77 to 4.87; and OR for the T3 of methylmercury and AC/CC genotype, 1.30; 95% CI, 0.76 to 2.21; p-interaction=0.013). When stratified by gender, a higher intake of methylmercury and the presence of a C allele exhibited an attenuated risk of GC in both men and women. However, no significant interaction was found between dietary methylmercury and the IL23R rs10889677 variant for GC risk (Table 5).
The present study investigated the association between dietary mercury intake and GC risk, considering the IL23R rs10889677 genetic variant among the Korean population. The key findings were (1) a higher dietary intake of methylmercury was significantly associated with an elevated risk of GC; (2) participants carrying the C allele for IL23R rs10889677 showed a protective effect on GC; and (3) the adverse impact of dietary methylmercury on GC tends to be less prominent among participants with the C allele for IL23R rs10887677.
We observed that a higher intake of mercury was associated with an elevated risk of GC. The effect of mercury on digestive tract diseases has been reported in several previous studies. A case-control study in Korea using the same dietary heavy metal database also demonstrated an increased risk of colorectal cancer among those who ingested greater amounts of dietary mercury [11]. In a study of 152 patients with chronic gastritis and 149 healthy controls in China, mercury levels in hair showed positive correlations with the severity of chronic gastritis among patients and with seafood intake among controls [10]. Another community-based cross-sectional study of 80 participants was conducted in the Canadian arctic region, where freshwater fish or seafood served as key food sources of methylmercury [5]. In this study, a model of gastric carcinogenesis incorporating intermediate endpoints (e.g., intestinal metaplasia, atrophy, and severe chronic gastritis) was utilized, and the ORs for severe chronic gastritis and atrophy reached their highest levels when hair-methylmercury levels exceeded 1 μg/g, particularly in conjunction with the lowest selenium intake [5]. Furthermore, mercury has been identified as a toxic compound that may promote cancer by inhibiting gap junction intercellular communications and producing inflammatory cytokines [39]. That evidence supports the hypothesis that mercury may contribute to inflammatory responses in the digestive tract, which can progress to carcinogenesis [5,10]. This also aligns with the findings of our study, which emphasized that a higher intake of dietary methylmercury was a significant risk factor for GC, likely due to the rapid absorption of highly lipophilic organic mercury in the gastrointestinal tract where a significant portion of the mercury accumulates in the human body [12,13]. A comprehensive review of epidemiological and experimental toxicology studies on cancer indicated that, although a plausible relationship between mercury-induced toxicity and cancer exists, there were inconsistencies across epidemiological studies (e.g., exposure assessment methods and cancer type). This highlights the need for further investigation [13].
Furthermore, the present study revealed a protective effect of the genetic polymorphism of the IL23R rs10889677 AC/CC or AC genotype on GC risk, but not for CC. According to the SNP database provided by the National Center for Biotechnology Information, a relationship between rs10889677 and GC risk has been documented in several studies, though the results were controversial. According to a case-control study comprising 1,010 cases of GC and 800 healthy controls, individuals carrying the AC or CC genotype were protected against GC when compared with those with the AA genotype (CC: OR, 0.47; 95% CI, 0.31 to 0.71 and AC: OR, 0.81; 95% CI, 0.66 to 0.99) [20]. In another study involving 500 cases and 500 controls, individuals with the CC genotype showed an elevated risk of GC (OR, 2.22; 95% CI, 1.27 to 3.87) compared to AA, whereas those with the AC or AC/CC genotype did not [25]. Non-significant associations with GC were observed in a study of 898 patients with GC and 992 controls (AC or CC) [26] and in another study with 479 cases of GC and 483 controls (AC/CC) [27]. Among case-control studies of other digestive tract cancers, the Chinese population with the AC or CC genotype exhibited a reduced risk of esophageal squamous cell carcinoma [40]. The AA genotype was identified as a risk factor for colorectal cancer in Iranian adults [23], while no correlation was observed in the Tunisian population [28]. Comprehensive findings from meta-analyses incorporating several types of solid cancers have demonstrated that the rs10889677 A> C may play a crucial role in the malignant transformation process; individuals with AC, CC, or AC/CC genotypes showed a lower risk of tumors than those with AA [22-24]. However, it should also be noted that the relationship of this genetic variant to prognosis in patients with GC may differ from its relationship to cancer incidence. One out of 2 studies following the mortality of patients with GC suggested an increased risk with each increment of the C allele (hazard ratio, 1.25; 95% CI, 1.05 to 1.49) [26], while another study reported that survival was not affected by their genotype [27]. A case-only study on breast cancer also showed that the C allele affected the age of cancer onset [41]. Therefore, prospective studies on both cancer incidence and prognosis are warranted to investigate how the IL23R genetic variant affects outcomes differently.
The possible biological mechanisms underlying the modification of cancer incidence by IL23R rs10889677 need to be explored. The IL23R gene encodes IL-23R, which plays a pivotal role in initiating, maintaining, and accelerating the IL-23/IL-17 inflammatory signal transduction pathway and is crucial for tumorigenesis [42-44]. Importantly, rs10889677 is a functional SNP located within the 3´-UTR of the IL23R and miRNA-mRNA hybridization site. It may alter the binding affinity of miR-let-7 and further downregulate IL23R gene expression, consequently inhibiting the translation of the IL-23R protein [23,24,45]. It has been shown that cancer-free adults with the rs10889677 AA genotype exhibited a higher expression of IL-23R in peripheral blood mononuclear cells relative to those with the AC/CC genotype. Indeed, the A allele carriers exhibited higher levels of regulatory T (Treg) cells in vivo and a lower T-cell proliferation rate in vitro when compared to C allele carriers [45]. It is also noteworthy that the imbalance between Th17 and Treg cells, 2 distinct CD4+ T cell subsets, can promote tissue inflammation by secreting the pro-inflammatory cytokine IL-17 and may ultimately contribute to carcinogenesis [46,47].
These findings provided a notable insight into the role of IL23R rs10889677 A>C in altering susceptibility to GC. This can likely be attributed to the reduction in IL-23R expression levels and the modulation of the IL-23/IL-17 inflammatory axis, both known to be implicated in the development of GC [20]. In addition, exposure to toxicants such as methylmercury may trigger immune dysfunction (e.g., cellular signals linked to the development of autoimmunity) and DNA damage, potentially increasing susceptibility to GC [14,18,19]. Collectively, we noted a significant trend wherein individuals carrying the C allele exhibited a comparatively lower susceptibility to GC from methylmercury intake than those with the AA genotype. Although the exact biological role remains unclear, our findings suggest that the IL23R rs10889677 polymorphism may interact with dietary methylmercury to stimulate pro-inflammatory cytokine responses, potentially contributing to the development of GC. Therefore, it is noteworthy that both low methylmercury intake and a specific polymorphism may play a protective role against GC.
The current study is one of the few studies investigating the effect of dietary mercury or a genetic polymorphism related to the IL17/IL23 inflammation pathway axis on the risk of GC. It also represents a novel attempt to explore the impact of an interactive effect between dietary mercury intake and the IL23R genetic variant on GC. Nevertheless, several limitations should be addressed. First, bias may be present due to the case-control study design. In terms of selection bias, the control participants were recruited from individuals who visited the clinic for a health check-up. They may have been more health-conscious and may have adopted healthier behaviors compared to patients with GC or those who were unwilling to receive health screening. Furthermore, our findings can be influenced by recall bias. The case participants may have recalled their past behaviors with greater accuracy since those factors were believed to be related to their disease. Accordingly, differences in the observed rates of exposure to risk factors between the cases and controls could be overstated. Nevertheless, the current study utilized a validated dietary questionnaire that was not influenced by prior knowledge of the research hypotheses [30]. Second, although the dietary mercury intake from each participant was meticulously linked with our heavy metal database, we could not validate the dietary mercury intake with mercury levels from biospecimens [48]. In future studies concerning GC risk, it is essential to validate the correlation between biospecimen mercury levels and dietary mercury intake. Lastly, while we suggested the genetic involvement of the IL17/IL23 pathway in the link between dietary mercury and GC, exploring a complex combination of genes in future studies is required for a thorough understanding of other genetic aspects in these biological processes [49,50].
In conclusion, we observed that a higher dietary intake of methylmercury was associated with an increased risk of GC, while IL23R rs10889677 polymorphism at the predicted miR-let-7 binding site exhibited a protective effect against GC. This genetic variant may modify the detrimental effect of dietary methylmercury on GC. Future large-scale prospective studies, incorporating biospecimen mercury levels and a wide array of genes related to miR-let-7 among diverse ethnicities, are warranted to unravel their roles in GC risk.
Supplementary materials are available at https://doi.org/10.4178/epih.e2024051.

Supplementary Material 1.

Association between IL23R rs10889677 genetic polymorphism and the risk of gastric cancer in the codominant model
epih-46-e2024051-Supplementary-1.docx

Data availability

The datasets used and analyzed in the current study are available from the corresponding author upon reasonable request.

Conflict of interest

The authors have no conflicts of interest to declare for this study.

Funding

This research was supported by a National Research Foundation grant funded by the Ministry of Science and Information Technology (2021R1A2C2008439).

Author contributions

Conceptualization: Kim JH, Gunathilake M, Kim J. Data curation: Choi IJ, Kim YI, Lee J, Kim J. Formal analysis: Kim JH. Funding acquisition: Kim J. Methodology: Kim JH, Gunathilake M, Kim J. Project administration: Lee J, Kim J. Writing – original draft: Kim JH. Writing – review & editing: Kim JH, Gunathilake M, Lee J, Choi IJ, Kim YI, Kim J.

We appreciate Hae Dong Woo and Dong Woo Kim for developing the database of dietary heavy metal concentrations.
Figure 1.
Flowchart of the participant selection process. SQFFQs, semi-quantitative food frequency questionnaires.
epih-46-e2024051f1.jpg
epih-46-e2024051f2.jpg
Table 1.
General characteristics of the study population and their dietary mercury intake, comparing patients with gastric cancer and controls
Characteristics All (n=1,133)
Men (n=743)
Women (n=390)
Controls (n=756) Cases (n=377) p-value1 Controls (n=497) Cases (n=246) p-value1 Controls (n=259) Cases (n=131) p-value1
Age (yr) 53.8±9.0 53.9±9.3 0.947 54.8±8.4 55.0±8.6 0.758 51.9±9.7 51.6±10.1 0.826
Gender
 Men 497 (65.7) 246 (65.3) 0.870 - - - -
 Women 259 (34.3) 131 (34.7) - - - -
Body mass index (kg/m2) 24.0±2.9 23.9±3.1 0.389 24.5±2.7 24.2±3.0 0.288 23.1±3.1 23.1±3.0 0.984
 <23 276 (36.5) 147 (39.0) 0.673 140 (28.2) 84 (34.2) 0.207 136 (52.5) 63 (48.1) 0.727
 23-<25 230 (30.4) 107 (28.4) 160 (32.2) 68 (27.6) 70 (27.0) 39 (29.8)
 ≥25 249 (32.9) 122 (32.4) 197 (39.6) 94 (38.2) 52 (20.1) 28 (21.4)
 Missing 1 (0.2) 1 (0.2) 0 (0.0) 0 (0.0) 1 (0.4) 1 (0.7)
Helicobacter pylori infection
 Negative 292 (38.6) 28 (7.4) <0.001 175 (35.2) 16 (6.5) <0.001 117 (45.2) 12 (9.2) <0.001
 Positive 464 (61.4) 349 (92.6) 322 (64.8) 230 (93.5) 142 (54.8) 119 (90.8)
Family history of gastric cancer in first-degree relatives
 No 659 (87.2) 299 (79.3) <0.001 424 (85.3) 190 (77.2) 0.006 235 (90.7) 109 (83.2) 0.030
 Yes 95 (12.6) 77 (20.4) 71 (14.3) 55 (22.4) 24 (9.3) 22 (16.8)
 Missing 2 (0.2) 1 (0.3) 2 (0.4) 1 (0.2) 0 (0.0) 0 (0.0)
Regular exercise
 Yes 424 (56.1) 136 (36.1) <0.001 279 (56.1) 100 (40.6) <0.001 145 (56.0) 36 (27.5) <0.001
 No 329 (43.5) 241 (63.9) 215 (43.3) 146 (59.4) 114 (44.0) 95 (72.5)
 Missing 3 (0.4) 0 (0.0) 3 (0.6) 0 (0.0) 0 (0.0) 0 (0.0)
Smoking status
 Never smokers 344 (45.5) 151 (40.0) <0.001 96 (19.3) 34 (13.8) <0.001 248 (95.8) 117 (89.4) 0.038
 Ex-smokers 258 (34.1) 110 (29.2) 251 (50.5) 103 (41.9) 7 (2.7) 7 (5.3)
 Current smokers 154 (20.4) 116 (30.8) 150 (30.2) 109 (44.3) 4 (1.5) 7 (5.3)
Alcohol drinking status
 Never drinkers 212 (28.0) 112 (39.7) 0.333 81 (16.3) 42 (17.1) 0.330 131 (50.6) 70 (53.4) 0.863
 Ex-drinkers 58 (7.7) 37 (9.8) 46 (9.3) 31 (12.6) 12 (4.6) 6 (4.6)
 Current drinkers 486 (64.3) 228 (60.5) 370 (74.4) 173 (70.3) 116 (44.8) 55 (42.0)
Education
 Less than college 334 (44.2) 289 (75.8) <0.001 188 (37.8) 187 (76.0) <0.001 146 (56.4) 102 (77.9) <0.001
 College and higher 392 (51.8) 87 (23.1) 281 (56.5) 58 (23.6) 111 (42.9) 29 (22.1)
 Missing 30 (4.0) 1 (0.1) 28 (5.7) 1 (0.4) 2 (0.7) 0 (0.0)
Monthly income (×10,000 Korean won/mo)
 <200 132 (17.5) 120 (31.8) <0.001 74 (14.9) 78 (31.7) <0.001 58 (22.4) 42 (32.1) 0.050
 200-<400 313 (41.4) 132 (35.0) 217 (43.7) 94 (38.2) 96 (37.1) 38 (29.0)
 ≥400 247 (32.7) 88 (23.3) 153 (30.8) 50 (20.3) 94 (36.3) 38 (29.0)
 Missing 64 (8.4) 37 (9.9) 53 (10.6) 24 (9.8) 11 (4.2) 13 (9.9)
Total energy intake (kcal/day) 1,717.3±546.9 1,925.2±611.9 <0.001 1,765.5±542.3 2,033.9±635.8 <0.001 1,624.9±544.9 1,721.1±506.4 0.093
Dietary total mercury, crude (µg/day) 13.2±4.2 14.8±4.5 <0.001 13.4±4.1 15.6±4.7 <0.001 12.6±4.4 13.3±3.7 0.110
Dietary total mercury, energy-adjusted (µg/day)2 13.7±2.4 14.0±2.3 0.033 13.6±2.0 14.1±2.5 0.011 13.8±3.1 13.9±2.1 0.816
Dietary methylmercury, crude (µg/day) 11.8±4.2 14.1±4.5 <0.001 12.5±4.1 15.0±4.6 <0.001 10.5±4.0 12.4±3.8 <0.001
Dietary methylmercury, energy-adjusted (µg/day)2 12.4±3.3 13.5±3.5 <0.001 12.8±3.0 13.7±3.6 <0.001 11.5±3.8 13.1±3.2 <0.001
IL23R rs10889677 genotype
 A/A 361 (47.7) 218 (57.8) 0.004 241 (48.5) 139 (56.5) 0.103 120 (46.3) 79 (60.3) 0.019
 A/C 331 (43.8) 127 (33.7) 219 (44.1) 89 (36.2) 112 (43.2) 38 (29.0)
 C/C 64 (8.5) 32 (8.5) 37 (7.4) 18 (7.3) 27 (10.4) 14 (10.7)

Values are presented as number (%) or mean±standard deviation.

1 The comparisons were made between gastric cancer cases and controls using the chi-square test for categorical variables and the Student t-test for continuous variables.

2 Dietary total mercury and methylmercury were adjusted for total energy intake using the residual method.

Table 2.
Association between dietary mercury intake and the risk of gastric cancer1
Dietary mercury (µg/day) Controls Cases Model I Model II Model III
Total mercury2
 Total (n=1,133)
  T1 (<12.78) 252 (33.3) 105 (27.9) 1.00 (reference) 1.00 (reference) 1.00 (reference)
  T2 (12.78-14.46) 252 (33.3) 126 (33.4) 1.20 (0.88, 1.64) 1.29 (0.91, 1.82) 1.42 (0.99, 2.04)
  T3 (≥14.46) 252 (33.3) 146 (38.7) 1.39 (1.02, 1.89) 1.38 (0.98, 1.93) 1.38 (0.97, 1.97)
  p for trend3 0.035 0.068 0.088
 Men (n=743)
  T1 (<12.90) 166 (33.4) 71 (28.9) 1.00 (reference) 1.00 (reference) 1.00 (reference)
  T2 (12.90-14.39) 166 (33.4) 74 (30.1) 1.04 (0.71, 1.54) 1.19 (0.77, 1.85) 1.41 (0.89, 2.24)
  T3 (≥14.39) 165 (33.2) 101 (41.1) 1.43 (0.99, 2.08) 1.55 (1.02, 2.35) 1.73 (1.11, 2.68)
  p for trend3 0.054 0.040 0.015
 Women (n=390)
  T1 (<12.51) 87 (33.6) 31 (23.7) 1.00 (reference) 1.00 (reference) 1.00 (reference)
  T2 (12.51-14.55) 86 (33.2) 54 (41.2) 1.76 (1.03, 3.00) 1.70 (0.94, 3.05) 2.25 (1.04, 4.89)
  T3 (≥14.55) 86 (33.2) 46 (35.1) 1.50 (0.87, 2.59) 1.39 (0.75, 2.57) 1.96 (0.63, 6.14)
  p for trend3 0.184 0.345 0.282
Methylmercury2
 Total (n=1,133)
  T1 (<11.00) 252 (33.3) 81 (21.5) 1.00 (reference) 1.00 (reference) 1.00 (reference)
  T2 (11.00-13.63) 252 (33.3) 112 (29.7) 1.38 (0.99, 1.93) 1.29 (0.89, 1.86) 1.43 (0.97, 2.11)
  T3 (≥13.63) 252 (33.3) 184 (48.8) 2.27 (1.66, 3.11) 1.98 (1.40, 2.80) 2.02 (1.41, 2.91)
  p for trend3 <0.001 <0.001 <0.001
 Men (n=743)
  T1 (<11.56) 166 (33.4) 62 (25.2) 1.00 (reference) 1.00 (reference) 1.00 (reference)
  T2 (11.56-13.98) 166 (33.4) 67 (27.2) 1.08 (0.72, 1.62) 1.09 (0.69, 1.70) 1.29 (0.81, 2.08)
  T3 (≥13.98) 165 (33.2) 117 (47.6) 1.90 (1.30, 2.76) 1.68 (1.11, 2.56) 1.77 (1.14, 2.74)
  p for trend3 <0.001 0.011 0.010
 Women (n=390)
  T1 (<9.89) 87 (33.6) 19 (14.5) 1.00 (reference) 1.00 (reference) 1.00 (reference)
  T2 (9.89-12.97) 86 (33.2) 48 (36.6) 2.56 (1.39, 4.70) 2.26 (1.17, 4.37) 2.58 (1.28, 5.23)
  T3 (≥12.97) 86 (33.2) 64 (48.9) 3.41 (1.88, 6.16) 2.75 (1.43, 5.30) 2.80 (1.40, 5.62)
  p for trend3 <0.001 0.004 0.001

Values are presented as number (%) or odds ratio (95% confidence interval).

1 Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).

2 The values are presented as tertiles of dietary total mercury or methylmercury intake, which were adjusted for total energy intake using a linear residual regression method.

3 Test for trend was calculated with the median intake for each category of dietary mercury as a continuous variable.

Table 3.
Association between IL23R rs10889677 genetic polymorphism and the risk of gastric cancer1
IL23R rs10889677 Inheritance model Controls Cases Model I Model II Model III
Total (n=1,133) Dominant model
 AA 361 (47.8) 218 (57.8) 1.00 (reference) 1.00 (reference) 1.00 (reference)
 AC/CC 395 (52.2) 159 (42.2) 0.67 (0.52, 0.86) 0.68 (0.52, 0.89) 0.62 (0.46, 0.83)
Allelic model
 A 1,053 (69.6) 563 (74.7) 1.00 (reference) 1.00 (reference) 1.00 (reference)
 C 459 (30.4) 191 (25.3) 0.78 (0.64, 0.95) 0.80 (0.64, 0.99) 0.74 (0.59, 0.93)
Men (n=743) Dominant model
 AA 241 (48.5) 139 (56.5) 1.00 (reference) 1.00 (reference) 1.00 (reference)
 AC/CC 256 (51.5) 107 (43.5) 0.73 (0.53, 0.99) 0.70 (0.50, 0.99) 0.61 (0.42, 0.88)
Allelic model
 A 701 (70.5) 367 (74.6) 1.00 (reference) 1.00 (reference) 1.00 (reference)
 C 293 (29.5) 125 (25.4) 0.82 (0.64, 1.04) 0.79 (0.60, 1.04) 0.71 (0.53, 0.95)
Women (n=390) Dominant model
 AA 120 (46.3) 79 (60.3) 1.00 (reference) 1.00 (reference) 1.00 (reference)
 AC/CC 139 (53.7) 52 (39.7) 0.57 (0.37, 0.87) 0.60 (0.38, 0.96) 0.60 (0.36, 0.99)
Allelic model
 A 352 (68.0) 196 (74.8) 1.00 (reference) 1.00 (reference) 1.00 (reference)
 C 166 (32.0) 66 (25.2) 0.71 (0.51, 1.00) 0.77 (0.54, 1.12) 0.77 (0.52, 1.15)

Values are presented as number (%) or odds ratio (95% confidence interval).

1 Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).

Table 4.
Association between the interaction of IL23R rs10889677 polymorphism with dietary total mercury and the risk of gastric cancer1
Dominant model Dietary mercury (µg/day) Controls Cases Model I Model II Model III
IL23R (rs10889677)
 Total (n=1,133)
  AA T1 (<12.78) 112 (14.8) 55 (14.6) 1.00 (reference) 1.00 (reference) 1.00 (reference)
T2 (12.78-14.46) 127 (16.8) 77 (20.4) 1.24 (0.80, 1.90) 1.47 (0.91, 2.36) 1.60 (0.96, 2.65)
T3 (≥14.46) 122 (16.1) 86 (22.8) 1.44 (0.94, 2.20) 1.40 (0.87, 2.24) 1.36 (0.82, 2.24)
  AC/CC T1 (<12.78) 140 (18.5) 50 (13.3) 0.73 (0.46, 1.15) 0.79 (0.48, 1.30) 0.71 (0.42, 1.21)
T2 (12.78-14.46) 125 (16.5) 49 (13.0) 0.80 (0.50, 1.27) 0.83 (0.50, 1.37) 0.81 (0.47, 1.39)
T3 (≥14.46) 130 (17.2) 60 (15.9) 0.94 (0.60, 1.47) 1.03 (0.63, 1.68) 0.94 (0.56, 1.59)
  p-interaction 0.748 0.896 0.964
 Men (n=743)
  AA T1 (<12.90) 79 (15.9) 40 (16.3) 1.00 (reference) 1.00 (reference) 1.00 (reference)
T2 (12.90-14.39) 81 (16.3) 43 (17.5) 1.05 (0.62, 1.78) 1.25 (0.69, 2.28) 1.44 (0.77, 2.71)
T3 (≥14.39) 81 (16.3) 56 (22.8) 1.37 (0.82, 2.28) 1.31 (0.74, 2.32) 1.32 (0.72, 2.43)
  AC/CC T1 (<12.90) 87 (17.5) 31 (12.6) 0.70 (0.40, 1.23) 0.64 (0.35, 1.20) 0.51 (0.27, 0.99)
T2 (12.90-14.39) 85 (17.1) 31 (12.6) 0.72 (0.41, 1.26) 0.72 (0.38, 1.35) 0.68 (0.35, 1.33)
T3 (≥14.39) 84 (16.9) 45 (18.3) 1.06 (0.63, 1.79) 1.18 (0.65, 2.13) 1.17 (0.63, 2.18)
  p-interaction 0.790 0.398 0.150
 Women (n=390)
  AA T1 (<12.51) 35 (13.5) 17 (13.0) 1.00 (reference) 1.00 (reference) 1.00 (reference)
T2 (12.51-14.55) 43 (16.6) 31 (23.7) 1.48 (0.71, 3.11) 1.61 (0.70, 3.68) 1.84 (0.73, 4.59)
T3 (≥14.55) 42 (16.2) 31 (23.7) 1.52 (0.72, 3.19) 1.59 (0.69, 3.66) 1.58 (0.64, 3.89)
  AC/CC T1 (<12.51) 52 (20.1) 14 (10.7) 0.55 (0.24, 1.27) 0.68 (0.27, 1.69) 0.80 (0.30, 2.14)
T2 (12.51-14.55) 43 (16.6) 23 (17.6) 1.10 (0.51, 2.38) 1.19 (0.51, 2.77) 1.35 (0.53, 3.42)
T3 (≥14.55) 44 (17.0) 15 (11.5) 0.70 (0.31, 1.60) 0.71 (0.28, 1.78) 0.61 (0.23, 1.62)
  p-interaction 0.780 0.538 0.348

Values are presented as number (%) or odds ratio (95% confidence interval).

1 Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).

Table 5.
Association between the interaction of IL23R rs10889677 polymorphism with dietary methylmercury and the risk of gastric cancer1
Dominant model Dietary methylmercury (µg/day) Controls Cases Model I Model II Model III
IL23R (rs10889677)
 Total (n=1,133)
  AA T1 (<11.00) 126 (16.7) 39 (10.3) 1.00 (reference) 1.00 (reference) 1.00 (reference)
T2 (11.00-13.63) 113 (14.9) 59 (15.7) 1.69 (1.05, 2.72) 1.58 (0.94, 2.67) 1.78 (1.03, 3.09)
T3 (≥13.63) 122 (16.1) 120 (31.8) 3.18 (2.05, 4.93) 2.62 (1.62, 4.23) 2.93 (1.77, 4.87)
  AC/CC T1 (<11.00) 126 (16.7) 42 (11.1) 1.08 (0.65, 1.78) 1.05 (0.61, 1.80) 1.04 (0.59, 1.83)
T2 (11.00-13.63) 139 (18.4) 53 (14.1) 1.23 (0.76, 1.99) 1.12 (0.66, 1.88) 1.22 (0.71, 2.11)
T3 (≥13.63) 130 (17.2) 64 (17.0) 1.59 (1.00, 2.54) 1.43 (0.86, 2.39) 1.30 (0.76, 2.21)
  p-interaction 0.016 0.047 0.013
 Men (n=743)
  AA T1 (<11.56) 85 (17.1) 33 (13.4) 1.00 (reference) 1.00 (reference) 1.00 (reference)
T2 (11.56-13.98) 79 (15.9) 33 (13.4) 1.08 (0.61, 1.91) 1.18 (0.63, 2.21) 1.57 (0.80, 3.07)
T3 (≥13.98) 77 (15.5) 73 (29.7) 2.44 (1.46, 4.08) 2.05 (1.16, 3.64) 2.28 (1.25, 4.17)
  AC/CC T1 (<11.56) 81 (16.3) 29 (11.8) 0.92 (0.51, 1.65) 0.90 (0.47, 1.72) 0.86 (0.44, 1.69)
T2 (11.56-13.98) 87 (17.5) 34 (13.8) 1.01 (0.57, 1.77) 0.92 (0.49, 1.72) 0.97 (0.50, 1.87)
T3 (≥13.98) 88 (17.7) 44 (17.9) 1.29 (0.75, 2.21) 1.17 (0.64, 2.15) 1.10 (0.59, 2.07)
  p-interaction 0.121 0.337 0.252
 Women (n=390)
  AA T1 (<9.89) 40 (15.4) 9 (6.9) 1.00 (reference) 1.00 (reference) 1.00 (reference)
T2 (9.89-12.97) 39 (15.1) 26 (19.8) 2.96 (1.23, 7.12) 2.91 (1.12, 7.58) 2.67 (0.95, 7.50)
T3 (≥12.97) 41 (15.8) 44 (33.6) 4.77 (2.06, 11.04) 3.91 (1.55, 9.91) 4.64 (1.68, 12.81)
  AC/CC T1 (<9.89) 47 (18.2) 10 (7.6) 0.95 (0.35, 2.56) 1.04 (0.36, 3.06) 0.98 (0.31, 3.08)
T2 (9.89-12.97) 47 (18.2) 22 (16.8) 2.08 (0.86, 5.03) 1.88 (0.72, 4.91) 2.60 (0.92, 7.36)
T3 (≥12.97) 45 (17.4) 20 (15.3) 1.98 (0.81, 4.83) 1.81 (0.68, 4.81) 1.50 (0.53, 4.26)
  p-interaction 0.168 0.228 0.080

Values are presented as number (%) or odds ratio (95% confidence interval).

1 Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).

Figure & Data

References

    Citations

    Citations to this article as recorded by  

      Figure
      • 0
      • 1
      Related articles
      Dietary mercury intake, the IL23R rs10889677 polymorphism, and the risk of gastric cancer in a Korean population: a hospital-based case-control study
      Image Image
      Figure 1. Flowchart of the participant selection process. SQFFQs, semi-quantitative food frequency questionnaires.
      Graphical abstract
      Dietary mercury intake, the IL23R rs10889677 polymorphism, and the risk of gastric cancer in a Korean population: a hospital-based case-control study
      Characteristics All (n=1,133)
      Men (n=743)
      Women (n=390)
      Controls (n=756) Cases (n=377) p-value1 Controls (n=497) Cases (n=246) p-value1 Controls (n=259) Cases (n=131) p-value1
      Age (yr) 53.8±9.0 53.9±9.3 0.947 54.8±8.4 55.0±8.6 0.758 51.9±9.7 51.6±10.1 0.826
      Gender
       Men 497 (65.7) 246 (65.3) 0.870 - - - -
       Women 259 (34.3) 131 (34.7) - - - -
      Body mass index (kg/m2) 24.0±2.9 23.9±3.1 0.389 24.5±2.7 24.2±3.0 0.288 23.1±3.1 23.1±3.0 0.984
       <23 276 (36.5) 147 (39.0) 0.673 140 (28.2) 84 (34.2) 0.207 136 (52.5) 63 (48.1) 0.727
       23-<25 230 (30.4) 107 (28.4) 160 (32.2) 68 (27.6) 70 (27.0) 39 (29.8)
       ≥25 249 (32.9) 122 (32.4) 197 (39.6) 94 (38.2) 52 (20.1) 28 (21.4)
       Missing 1 (0.2) 1 (0.2) 0 (0.0) 0 (0.0) 1 (0.4) 1 (0.7)
      Helicobacter pylori infection
       Negative 292 (38.6) 28 (7.4) <0.001 175 (35.2) 16 (6.5) <0.001 117 (45.2) 12 (9.2) <0.001
       Positive 464 (61.4) 349 (92.6) 322 (64.8) 230 (93.5) 142 (54.8) 119 (90.8)
      Family history of gastric cancer in first-degree relatives
       No 659 (87.2) 299 (79.3) <0.001 424 (85.3) 190 (77.2) 0.006 235 (90.7) 109 (83.2) 0.030
       Yes 95 (12.6) 77 (20.4) 71 (14.3) 55 (22.4) 24 (9.3) 22 (16.8)
       Missing 2 (0.2) 1 (0.3) 2 (0.4) 1 (0.2) 0 (0.0) 0 (0.0)
      Regular exercise
       Yes 424 (56.1) 136 (36.1) <0.001 279 (56.1) 100 (40.6) <0.001 145 (56.0) 36 (27.5) <0.001
       No 329 (43.5) 241 (63.9) 215 (43.3) 146 (59.4) 114 (44.0) 95 (72.5)
       Missing 3 (0.4) 0 (0.0) 3 (0.6) 0 (0.0) 0 (0.0) 0 (0.0)
      Smoking status
       Never smokers 344 (45.5) 151 (40.0) <0.001 96 (19.3) 34 (13.8) <0.001 248 (95.8) 117 (89.4) 0.038
       Ex-smokers 258 (34.1) 110 (29.2) 251 (50.5) 103 (41.9) 7 (2.7) 7 (5.3)
       Current smokers 154 (20.4) 116 (30.8) 150 (30.2) 109 (44.3) 4 (1.5) 7 (5.3)
      Alcohol drinking status
       Never drinkers 212 (28.0) 112 (39.7) 0.333 81 (16.3) 42 (17.1) 0.330 131 (50.6) 70 (53.4) 0.863
       Ex-drinkers 58 (7.7) 37 (9.8) 46 (9.3) 31 (12.6) 12 (4.6) 6 (4.6)
       Current drinkers 486 (64.3) 228 (60.5) 370 (74.4) 173 (70.3) 116 (44.8) 55 (42.0)
      Education
       Less than college 334 (44.2) 289 (75.8) <0.001 188 (37.8) 187 (76.0) <0.001 146 (56.4) 102 (77.9) <0.001
       College and higher 392 (51.8) 87 (23.1) 281 (56.5) 58 (23.6) 111 (42.9) 29 (22.1)
       Missing 30 (4.0) 1 (0.1) 28 (5.7) 1 (0.4) 2 (0.7) 0 (0.0)
      Monthly income (×10,000 Korean won/mo)
       <200 132 (17.5) 120 (31.8) <0.001 74 (14.9) 78 (31.7) <0.001 58 (22.4) 42 (32.1) 0.050
       200-<400 313 (41.4) 132 (35.0) 217 (43.7) 94 (38.2) 96 (37.1) 38 (29.0)
       ≥400 247 (32.7) 88 (23.3) 153 (30.8) 50 (20.3) 94 (36.3) 38 (29.0)
       Missing 64 (8.4) 37 (9.9) 53 (10.6) 24 (9.8) 11 (4.2) 13 (9.9)
      Total energy intake (kcal/day) 1,717.3±546.9 1,925.2±611.9 <0.001 1,765.5±542.3 2,033.9±635.8 <0.001 1,624.9±544.9 1,721.1±506.4 0.093
      Dietary total mercury, crude (µg/day) 13.2±4.2 14.8±4.5 <0.001 13.4±4.1 15.6±4.7 <0.001 12.6±4.4 13.3±3.7 0.110
      Dietary total mercury, energy-adjusted (µg/day)2 13.7±2.4 14.0±2.3 0.033 13.6±2.0 14.1±2.5 0.011 13.8±3.1 13.9±2.1 0.816
      Dietary methylmercury, crude (µg/day) 11.8±4.2 14.1±4.5 <0.001 12.5±4.1 15.0±4.6 <0.001 10.5±4.0 12.4±3.8 <0.001
      Dietary methylmercury, energy-adjusted (µg/day)2 12.4±3.3 13.5±3.5 <0.001 12.8±3.0 13.7±3.6 <0.001 11.5±3.8 13.1±3.2 <0.001
      IL23R rs10889677 genotype
       A/A 361 (47.7) 218 (57.8) 0.004 241 (48.5) 139 (56.5) 0.103 120 (46.3) 79 (60.3) 0.019
       A/C 331 (43.8) 127 (33.7) 219 (44.1) 89 (36.2) 112 (43.2) 38 (29.0)
       C/C 64 (8.5) 32 (8.5) 37 (7.4) 18 (7.3) 27 (10.4) 14 (10.7)
      Dietary mercury (µg/day) Controls Cases Model I Model II Model III
      Total mercury2
       Total (n=1,133)
        T1 (<12.78) 252 (33.3) 105 (27.9) 1.00 (reference) 1.00 (reference) 1.00 (reference)
        T2 (12.78-14.46) 252 (33.3) 126 (33.4) 1.20 (0.88, 1.64) 1.29 (0.91, 1.82) 1.42 (0.99, 2.04)
        T3 (≥14.46) 252 (33.3) 146 (38.7) 1.39 (1.02, 1.89) 1.38 (0.98, 1.93) 1.38 (0.97, 1.97)
        p for trend3 0.035 0.068 0.088
       Men (n=743)
        T1 (<12.90) 166 (33.4) 71 (28.9) 1.00 (reference) 1.00 (reference) 1.00 (reference)
        T2 (12.90-14.39) 166 (33.4) 74 (30.1) 1.04 (0.71, 1.54) 1.19 (0.77, 1.85) 1.41 (0.89, 2.24)
        T3 (≥14.39) 165 (33.2) 101 (41.1) 1.43 (0.99, 2.08) 1.55 (1.02, 2.35) 1.73 (1.11, 2.68)
        p for trend3 0.054 0.040 0.015
       Women (n=390)
        T1 (<12.51) 87 (33.6) 31 (23.7) 1.00 (reference) 1.00 (reference) 1.00 (reference)
        T2 (12.51-14.55) 86 (33.2) 54 (41.2) 1.76 (1.03, 3.00) 1.70 (0.94, 3.05) 2.25 (1.04, 4.89)
        T3 (≥14.55) 86 (33.2) 46 (35.1) 1.50 (0.87, 2.59) 1.39 (0.75, 2.57) 1.96 (0.63, 6.14)
        p for trend3 0.184 0.345 0.282
      Methylmercury2
       Total (n=1,133)
        T1 (<11.00) 252 (33.3) 81 (21.5) 1.00 (reference) 1.00 (reference) 1.00 (reference)
        T2 (11.00-13.63) 252 (33.3) 112 (29.7) 1.38 (0.99, 1.93) 1.29 (0.89, 1.86) 1.43 (0.97, 2.11)
        T3 (≥13.63) 252 (33.3) 184 (48.8) 2.27 (1.66, 3.11) 1.98 (1.40, 2.80) 2.02 (1.41, 2.91)
        p for trend3 <0.001 <0.001 <0.001
       Men (n=743)
        T1 (<11.56) 166 (33.4) 62 (25.2) 1.00 (reference) 1.00 (reference) 1.00 (reference)
        T2 (11.56-13.98) 166 (33.4) 67 (27.2) 1.08 (0.72, 1.62) 1.09 (0.69, 1.70) 1.29 (0.81, 2.08)
        T3 (≥13.98) 165 (33.2) 117 (47.6) 1.90 (1.30, 2.76) 1.68 (1.11, 2.56) 1.77 (1.14, 2.74)
        p for trend3 <0.001 0.011 0.010
       Women (n=390)
        T1 (<9.89) 87 (33.6) 19 (14.5) 1.00 (reference) 1.00 (reference) 1.00 (reference)
        T2 (9.89-12.97) 86 (33.2) 48 (36.6) 2.56 (1.39, 4.70) 2.26 (1.17, 4.37) 2.58 (1.28, 5.23)
        T3 (≥12.97) 86 (33.2) 64 (48.9) 3.41 (1.88, 6.16) 2.75 (1.43, 5.30) 2.80 (1.40, 5.62)
        p for trend3 <0.001 0.004 0.001
      IL23R rs10889677 Inheritance model Controls Cases Model I Model II Model III
      Total (n=1,133) Dominant model
       AA 361 (47.8) 218 (57.8) 1.00 (reference) 1.00 (reference) 1.00 (reference)
       AC/CC 395 (52.2) 159 (42.2) 0.67 (0.52, 0.86) 0.68 (0.52, 0.89) 0.62 (0.46, 0.83)
      Allelic model
       A 1,053 (69.6) 563 (74.7) 1.00 (reference) 1.00 (reference) 1.00 (reference)
       C 459 (30.4) 191 (25.3) 0.78 (0.64, 0.95) 0.80 (0.64, 0.99) 0.74 (0.59, 0.93)
      Men (n=743) Dominant model
       AA 241 (48.5) 139 (56.5) 1.00 (reference) 1.00 (reference) 1.00 (reference)
       AC/CC 256 (51.5) 107 (43.5) 0.73 (0.53, 0.99) 0.70 (0.50, 0.99) 0.61 (0.42, 0.88)
      Allelic model
       A 701 (70.5) 367 (74.6) 1.00 (reference) 1.00 (reference) 1.00 (reference)
       C 293 (29.5) 125 (25.4) 0.82 (0.64, 1.04) 0.79 (0.60, 1.04) 0.71 (0.53, 0.95)
      Women (n=390) Dominant model
       AA 120 (46.3) 79 (60.3) 1.00 (reference) 1.00 (reference) 1.00 (reference)
       AC/CC 139 (53.7) 52 (39.7) 0.57 (0.37, 0.87) 0.60 (0.38, 0.96) 0.60 (0.36, 0.99)
      Allelic model
       A 352 (68.0) 196 (74.8) 1.00 (reference) 1.00 (reference) 1.00 (reference)
       C 166 (32.0) 66 (25.2) 0.71 (0.51, 1.00) 0.77 (0.54, 1.12) 0.77 (0.52, 1.15)
      Dominant model Dietary mercury (µg/day) Controls Cases Model I Model II Model III
      IL23R (rs10889677)
       Total (n=1,133)
        AA T1 (<12.78) 112 (14.8) 55 (14.6) 1.00 (reference) 1.00 (reference) 1.00 (reference)
      T2 (12.78-14.46) 127 (16.8) 77 (20.4) 1.24 (0.80, 1.90) 1.47 (0.91, 2.36) 1.60 (0.96, 2.65)
      T3 (≥14.46) 122 (16.1) 86 (22.8) 1.44 (0.94, 2.20) 1.40 (0.87, 2.24) 1.36 (0.82, 2.24)
        AC/CC T1 (<12.78) 140 (18.5) 50 (13.3) 0.73 (0.46, 1.15) 0.79 (0.48, 1.30) 0.71 (0.42, 1.21)
      T2 (12.78-14.46) 125 (16.5) 49 (13.0) 0.80 (0.50, 1.27) 0.83 (0.50, 1.37) 0.81 (0.47, 1.39)
      T3 (≥14.46) 130 (17.2) 60 (15.9) 0.94 (0.60, 1.47) 1.03 (0.63, 1.68) 0.94 (0.56, 1.59)
        p-interaction 0.748 0.896 0.964
       Men (n=743)
        AA T1 (<12.90) 79 (15.9) 40 (16.3) 1.00 (reference) 1.00 (reference) 1.00 (reference)
      T2 (12.90-14.39) 81 (16.3) 43 (17.5) 1.05 (0.62, 1.78) 1.25 (0.69, 2.28) 1.44 (0.77, 2.71)
      T3 (≥14.39) 81 (16.3) 56 (22.8) 1.37 (0.82, 2.28) 1.31 (0.74, 2.32) 1.32 (0.72, 2.43)
        AC/CC T1 (<12.90) 87 (17.5) 31 (12.6) 0.70 (0.40, 1.23) 0.64 (0.35, 1.20) 0.51 (0.27, 0.99)
      T2 (12.90-14.39) 85 (17.1) 31 (12.6) 0.72 (0.41, 1.26) 0.72 (0.38, 1.35) 0.68 (0.35, 1.33)
      T3 (≥14.39) 84 (16.9) 45 (18.3) 1.06 (0.63, 1.79) 1.18 (0.65, 2.13) 1.17 (0.63, 2.18)
        p-interaction 0.790 0.398 0.150
       Women (n=390)
        AA T1 (<12.51) 35 (13.5) 17 (13.0) 1.00 (reference) 1.00 (reference) 1.00 (reference)
      T2 (12.51-14.55) 43 (16.6) 31 (23.7) 1.48 (0.71, 3.11) 1.61 (0.70, 3.68) 1.84 (0.73, 4.59)
      T3 (≥14.55) 42 (16.2) 31 (23.7) 1.52 (0.72, 3.19) 1.59 (0.69, 3.66) 1.58 (0.64, 3.89)
        AC/CC T1 (<12.51) 52 (20.1) 14 (10.7) 0.55 (0.24, 1.27) 0.68 (0.27, 1.69) 0.80 (0.30, 2.14)
      T2 (12.51-14.55) 43 (16.6) 23 (17.6) 1.10 (0.51, 2.38) 1.19 (0.51, 2.77) 1.35 (0.53, 3.42)
      T3 (≥14.55) 44 (17.0) 15 (11.5) 0.70 (0.31, 1.60) 0.71 (0.28, 1.78) 0.61 (0.23, 1.62)
        p-interaction 0.780 0.538 0.348
      Dominant model Dietary methylmercury (µg/day) Controls Cases Model I Model II Model III
      IL23R (rs10889677)
       Total (n=1,133)
        AA T1 (<11.00) 126 (16.7) 39 (10.3) 1.00 (reference) 1.00 (reference) 1.00 (reference)
      T2 (11.00-13.63) 113 (14.9) 59 (15.7) 1.69 (1.05, 2.72) 1.58 (0.94, 2.67) 1.78 (1.03, 3.09)
      T3 (≥13.63) 122 (16.1) 120 (31.8) 3.18 (2.05, 4.93) 2.62 (1.62, 4.23) 2.93 (1.77, 4.87)
        AC/CC T1 (<11.00) 126 (16.7) 42 (11.1) 1.08 (0.65, 1.78) 1.05 (0.61, 1.80) 1.04 (0.59, 1.83)
      T2 (11.00-13.63) 139 (18.4) 53 (14.1) 1.23 (0.76, 1.99) 1.12 (0.66, 1.88) 1.22 (0.71, 2.11)
      T3 (≥13.63) 130 (17.2) 64 (17.0) 1.59 (1.00, 2.54) 1.43 (0.86, 2.39) 1.30 (0.76, 2.21)
        p-interaction 0.016 0.047 0.013
       Men (n=743)
        AA T1 (<11.56) 85 (17.1) 33 (13.4) 1.00 (reference) 1.00 (reference) 1.00 (reference)
      T2 (11.56-13.98) 79 (15.9) 33 (13.4) 1.08 (0.61, 1.91) 1.18 (0.63, 2.21) 1.57 (0.80, 3.07)
      T3 (≥13.98) 77 (15.5) 73 (29.7) 2.44 (1.46, 4.08) 2.05 (1.16, 3.64) 2.28 (1.25, 4.17)
        AC/CC T1 (<11.56) 81 (16.3) 29 (11.8) 0.92 (0.51, 1.65) 0.90 (0.47, 1.72) 0.86 (0.44, 1.69)
      T2 (11.56-13.98) 87 (17.5) 34 (13.8) 1.01 (0.57, 1.77) 0.92 (0.49, 1.72) 0.97 (0.50, 1.87)
      T3 (≥13.98) 88 (17.7) 44 (17.9) 1.29 (0.75, 2.21) 1.17 (0.64, 2.15) 1.10 (0.59, 2.07)
        p-interaction 0.121 0.337 0.252
       Women (n=390)
        AA T1 (<9.89) 40 (15.4) 9 (6.9) 1.00 (reference) 1.00 (reference) 1.00 (reference)
      T2 (9.89-12.97) 39 (15.1) 26 (19.8) 2.96 (1.23, 7.12) 2.91 (1.12, 7.58) 2.67 (0.95, 7.50)
      T3 (≥12.97) 41 (15.8) 44 (33.6) 4.77 (2.06, 11.04) 3.91 (1.55, 9.91) 4.64 (1.68, 12.81)
        AC/CC T1 (<9.89) 47 (18.2) 10 (7.6) 0.95 (0.35, 2.56) 1.04 (0.36, 3.06) 0.98 (0.31, 3.08)
      T2 (9.89-12.97) 47 (18.2) 22 (16.8) 2.08 (0.86, 5.03) 1.88 (0.72, 4.91) 2.60 (0.92, 7.36)
      T3 (≥12.97) 45 (17.4) 20 (15.3) 1.98 (0.81, 4.83) 1.81 (0.68, 4.81) 1.50 (0.53, 4.26)
        p-interaction 0.168 0.228 0.080
      Table 1. General characteristics of the study population and their dietary mercury intake, comparing patients with gastric cancer and controls

      Values are presented as number (%) or mean±standard deviation.

      The comparisons were made between gastric cancer cases and controls using the chi-square test for categorical variables and the Student t-test for continuous variables.

      Dietary total mercury and methylmercury were adjusted for total energy intake using the residual method.

      Table 2. Association between dietary mercury intake and the risk of gastric cancer1

      Values are presented as number (%) or odds ratio (95% confidence interval).

      Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).

      The values are presented as tertiles of dietary total mercury or methylmercury intake, which were adjusted for total energy intake using a linear residual regression method.

      Test for trend was calculated with the median intake for each category of dietary mercury as a continuous variable.

      Table 3. Association between IL23R rs10889677 genetic polymorphism and the risk of gastric cancer1

      Values are presented as number (%) or odds ratio (95% confidence interval).

      Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).

      Table 4. Association between the interaction of IL23R rs10889677 polymorphism with dietary total mercury and the risk of gastric cancer1

      Values are presented as number (%) or odds ratio (95% confidence interval).

      Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).

      Table 5. Association between the interaction of IL23R rs10889677 polymorphism with dietary methylmercury and the risk of gastric cancer1

      Values are presented as number (%) or odds ratio (95% confidence interval).

      Model I: crude model; Model II: adjusted for age (continuous), gender (unadjusted in the gender-stratified analysis), body mass index (<23, 23-<25, or ≥25 kg/m2), smoking status (current-, ex-, or non-smoker), drinking status (current-, ex-, or non-drinker), physical activity (yes or no), education level (less than college or college and higher), income (<200, 200-<400, or ≥400 [×10,000 Korean won/mo]), and first-degree family history of gastric cancer (yes or no); Model III: additionally adjusted for Helicobacter pylori infection (positive or negative).


      Epidemiol Health : Epidemiology and Health
      TOP