Exploring Modifiable Risk Factors: Insights from Mendelian Randomization Analyses of Gastric Cancer in East Asian Populations

doi:10.21203/rs.3.rs-4355148/v1

Download PDF

Research Article

Exploring Modifiable Risk Factors: Insights from Mendelian Randomization Analyses of Gastric Cancer in East Asian Populations

https://doi.org/10.21203/rs.3.rs-4355148/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: The incidence of gastric cancer (GC) shows strong geographic variation, with the highest incidence occurring in East Asia. Epidemiological studies have linked lifestyle, diet, and inflammatory factors to the risk of GC. However, their causal relationship is subject to debate due to the potential presence of bias. Addressing these uncertainties is vital for guiding effective preventive strategies.

Methods: We used genetic variants as instruments via two-sample univariate and multivariate Mendelian randomization (MR) analyses to examine the relationships between 40 potentially modifiable risk factors and gastric cancer in 6563 patients with gastric cancer and 195745 controls. These population data came from a genome-wide association study of people of Asian ancestry and were obtained from BioBank Japan(BBJ).

Results: Our multivariable MR analyses provided suggestive evidence of a potential association between genetically predicted concentrations of serum hemoglobin (OR_SD 0.62 [95% CI 0.41 ~ 0.93]; p=0.02), lactate dehydrogenase (OR_SD 0.62 [95% CI 0.41 ~ 0.93]; p<0.001) and alkaline phosphatase (OR _SD0.80 [95% CI 0.73 ~ 0.88]; p <0.001) and a decreased risk of GC. Furthermore, our study revealed a causal link between type 2 diabetes mellitus (OR_SD 0.83, 95% CI=0.73~0.93, P value=0.002) and GC incidence.

Conclusions: This analysis identified several potential modifiable factors for gastric cancer, including hemoglobin, lactate dehydrogenase, alkaline phosphatase and T2DM. These findings should be considered when formulating strategies for the primary prevention of GC, thereby informing evidence-based public health policies.

Gastric cancer

Mendelian randomization

Modifiable factors Hemoglobin

Type 2 diabetes mellitus

Gastric cancer (GC) is an important cancer worldwide and nearly one in every 13 deaths globally, with more than one million new cases in 2020[1]. The incidence of GC shows strong geographic variation, with the highest incidence occurring in East Asia, especially in Japan and Korea[2, 3]. The formation of GCs is usually a multifactorial process, and the common risk factors for GC include lifestyle factors, genetic factors, and other dietary factors[4]. Most available evidence for causal relationships between potentially modifiable factors and the risk of GC comes from observational studies, which are susceptible to confounding bias and reverse causation[5]. Thus, interest in developing public health programs to reduce the incidence of GC by targeting modifiable risk factors is increasing.

Modifiable factors usually refer to personal behaviors, lifestyles, and environmental factors associated with disease or health and play important roles in disease onset, development, and prevention[6]. For example, regarding Helicobacter pylori infection, smoking status, alcohol consumption, etc. a recent study provided evidence that people with poor lifestyles (smoking, drinking) have a greater risk of GC[7]. With national H. pylori screening and eradication, trends in GC incidence rates may be affected[8]. For most other risk factors, however, the evidence is too inconclusive to reliably establish causal associations.

Mendelian randomization (MR) is a natural experiment in which germline genetic variants are used as proxies for putative risk factors[9]. In MR, researchers use genetic variation as a tool for natural random assignment to infer a causal relationship between the effects of a modifiable factor of interest (e.g., drug use, lifestyle) and a specific outcome (e.g., disease occurrence, biomarker changes). MR has high scientific value because it involves the use of single nucleotide polymorphisms (SNPs) as an instrumental variable for exposure, which can effectively prevent confounding factors from relevant trait variations[9]. Therefore, we used a two-sample MR approach to examine whether several modifiable factors affect the risk of GC from a genetic perspective.

Most existing Mendelian randomization studies have used European or North American populations, and data from genome-wide analyses based on GC populations are insufficient. There are only a few MR studies of GC in Asia, a region with a high incidence of GC. Therefore, this study aimed to investigate the potential causal relationships between modifiable factors and GC incidence using data from Biobank Japan (BBJ).

Study design

We obtained ethical approval from the ethics committees of the RIKEN Yokohama Research Institute and the University of Tokyo Institute of Medical Sciences and obtained written informed consent from all participants. Our study utilized publicly available summary statistics from the International Working Unit (IEU) open GWAS (https://gwas.mrcieu.ac.uk/). Additionally, no institutional review board approval was required due to the use of nonpersonal data. A schematic summary of the study design is given in Fig. 1. Our MR methods were predicted on the basis of the assumptions that a single-nucleotide polymorphism (SNP) is valid for an MR survey, three hypotheses should be met: 1) the correlation hypothesis, indicating SNP correlation with risk factors; 2) the independence hypothesis, signifying SNP independence from potential confounders; and 3) the exclusion-limitation hypothesis, suggesting that the SNP solely impacts outcomes via the risk factor for interest [10](Fig. 2).

Date resources

In this two-sample MR analysis, genetic variants were employed as instrumental variables to investigate the potential associations between 40 modifiable risk factors and GC incidence. A comprehensive search of PubMed was conducted prior to this study to identify 10 subgroups, including those related to serum lipids, hepatic function, renal function, coagulation function, blood pressure, serum ion levels, cytometry, myocardial, menstrual enzymes, and blood glucose potentially impacting GC risk, as assessed in GC. The GC cohort consisted of 6563 patients and 195745 controls. All exposure and outcome data were sourced from Biobank Japan (BBJ), the largest East Asian biospecimen repository, encompassing 200,000[11] Japanese individuals aged > 20 to 89 years; these individuals were recruited and tracked at 12 medical institutions from 2003 to 2018.

Selection of instrumental variables

For our instrumental variables, the following criteria were applied: 1) significant association with GC (P < 5 × 10^− 8), 2) independence (linkage disequilibrium, r² < 0.001; within 10,000 kb), 3) suitable effect allele frequency (≥ 5%), and 4) nonpalindromic SNPs (A/T or C/G) with effect allele frequencies between 40% and 70%. We calculated R² to explain the variance of the instrumental variables and the F-statistic to test the first hypothesis, defining a strong instrumental variable with an F-statistic > 10[12, 13].

Mendelian randomization analysis

Given the SNPs associated with GC, corresponding modifiable factor GWAS data were extracted from BBJ and reconciled to ensure that effect alleles were linked with elevated GC risk. All GWAS data were harmonized to ensure that the effect allele (versus the noneffect allele) was associated with a higher GC risk (Supplementary Table S3). The inverse variance weighting (IVW) method served as the primary approach to compute odds ratios (ORs) and 95% confidence intervals (95% CIs) for each increase in the log odds of genetically forecasted GC. We considered this a causal association (P > 0.05) only when the IVW estimate aligned in direction and was statistically significant according to at least one sensitivity analysis and when no evidence of multiple effects was demonstrated. Additionally, the robustness of the main results was verified using the MR‒Egger method, weighted median, and weighted mode.

Sensitivity analyses

For the second and third assumptions, we performed a sensitivity analysis to test the assumptions. Sensitivity analyses were also conducted to ensure MR result robustness and mitigate IV pleiotropy bias. These analyses included heterogeneity tests, genetic horizontal pleiotropy assessments, and leave-one-out analysis. Heterogeneity in causal effects between IVs was assessed using IVW and MR‒Egger estimates, with Cochran's Q statistic gauging heterogeneity, and Q statistics significant at a p value < 0.05 can imply the presence of heterogeneity. Random-effects IVW was employed if SNP heterogeneity was observed (P < 0.05 by Cochran's Q test); if the p value was > 0.05, the MR-PRESSO[14] method was applied to remove outliers before conducting the Q test again. Scatter plots were generated to visualize outcomes. Genetic pleiotropy was estimated via the MR‒Egger regression intercept and MR-PRESSO tests. SNP exposure and effect estimate results were plotted to identify abnormal genetic variation. The leave-one-out method evaluates SNP influence by iteratively removing each SNP and examining the remaining SNPs. We constructed leave-one-out plots to evaluate the effect of each SNP on GC and to identify SNPs that may significantly affect the associations. Funnel plots, commonly used in meta-analyses, were applied for visual asymmetry inspection, suggesting violation of the MR hypothesis via horizontal pleiotropy[15]. We considered causal conclusions to be relatively reliable when MR analyses showed consistent directions of causal estimation among methods and no horizontal pleiotropy was found by the MR-PLEIOTROPY test (p > 0.05).

Multivariate MR analyses

Because of the genetic similarity of the modifiable factors between each subgroup and the SNPs potentially sharing coregulatory pathways, multivariate Mendelian randomization (MVMR)[16] was employed, a recently developed approach facilitating simultaneous assessment of correlated but distinct exposures by incorporating genetic variants from each risk factor into the same model. To adjust for different traits in each subgroup, we performed an MVMR analysis. For each exposure, the instruments are selected, and all exposures for those SNPs are regressed against the outcome together, weighting for the inverse variance of the outcome. After harmonization was performed, we clumped the data based on a threshold of r² < 0.001. The main estimation method performed was also the IVW method.

We performed all the MR analyses, and MVMR was performed using the Two Sample MR package (version 0.5.6) and the MR-PRESSO package in R (version 1.0). A P value less than 0.05 for the 2-sided test was considered to indicate statistical significance.

The 40 potentially modifiable risk factors for GC that we studied included 3 traits related to serum lipids, 9 to hepatic function, 3 to renal function, 2 to coagulation function, 2 to blood pressure, 5 to serum ion levels, 2 cytometry factors, 2 myocardial factors ,2 menstrual enzymes and 2 related to glucose and metabolism. The detailed data sources are shown in Supplementary Table S1. All the exposure data were harmonized to ensure that the effect allele (versus the noneffect allele) was associated with a higher GC risk (Supplementary Table S2). The number of SNPs used for these potentially modifiable risk factors ranged from 1 to 101 for genetic tools. The F-statistic ranged from 30 to 5768, suggesting the presence of a relatively strong instrumental variable (Supplementary Table S2).

In the primary SVMR analysis, with the use of the random effects IVW estimate, a total of 9 causative relationship traits were identified at p < 0.05. The MR‒Egger method, weighted median, and weighted mode were used as complements (Supplementary Table S3). To visualize the effect sizes and 95% confidence intervals, a forest plot was generated. (Fig. 3) .We noted a suggestive association between genetically predicted low serum alanine aminotransferase (OR_SD 0.72 [ 95% CI 0.53 ~ 0.97 ] ; p = 0.0307)、low gamma glutamyl transferase (OR_SD 0.86 [ 95% CI 0.76 ~ 0.98 ] ; p = 0.0191)and low alkaline phosphatase(ALP) (OR_SD0.80 [ 95% CI 0.73 ~ 0.88 ] ; p<0.001) concentrations and increased risk of GC. Systolic blood pressure (OR_SD 0.65 [95% CI 0.45 ~ 0.95]; p = 0.0253) and diastolic blood pressure (OR_SD 0.62 [ CI 0.43 ~ 0.89]; p = 0.0094) were associated with GC risk. A suggestive association based on 15 SNPs was noted between serum calcium concentrations and decreased risk of GC (OR_SD 1.36 [95% CI 1.06 ~ 1.73]; p = 0.0135). When we included information about all genetic variants associated with various cytometry factors, basophil count was suggested to be associated with GC (OR_SD 0.79 [95% CI 0.68 ~ 0.91]; p = 0.0018). According to our restricted analysis of glucose and metabolism, there was a suggestive association between genetically predicted increased serum hemoglobin A1c concentrations and a decreased risk of GC (OR_SD 0.79 [95% CI 0.65 ~ 0.97]; p = 0.0221), and the IVW hazard ratio per doubling of the probability of T2DM for GC was 0.86 [95% CI 0.82 ~ 0.91]. In the sensitivity analyses, heterogeneity was detected for some modifiable factors (p < 0.05) by the Cochran Q test (Supplementary Table S4). Furthermore, MR-PRESSO detected some outliers, and we removed the corresponding SNPs manually and repeated MR-PRESSO analysis. The specific outliers were as follows: one SNP (rs62525059) for systolic blood pressure, one SNP (rs4951254) for basophil count, and four SNPs (rs174594, rs2748427, rs147538848, and rs2237896) for hemoglobin A1c. (Only one SNP was suitable for use as an instrumental variable for, which meant that sensitivity analysis could not be performed.) Funnel and scatter plots were drawn to visualize the results (Supplementary Table S1 and S2). In addition, a leave-1-out plot indicated that no SNP had a significant effect on these results (Supplementary Table S1 and S2). However, causal estimates were consistent among the MR models, and MR pleiotropy analysis indicated that no horizontal pleiotropy was detected except for gamma glutamyl transferase (p = 0.0141) (Supplementary Table S5).

In MVMR analyses, the associations for each genetic instrument of each exposure were conditioned on one another. Therefore, we performed MVMR on each subgroup (ten subgroups in total) to reduce the risk of violating MR assumptions. According to the IVW method, 4 causative factors were associated with the risk of GC, namely, genetically predicted T2DM (OR_SD 0.83, 95% CI = 0.73 ~ 0.93, p = 0.002) and alkaline phosphatase (ALP) (OR _SD0.80 [95% CI = 0.73 ~ 0.88]; p < 0.001), which retained its association with GC incidence. Notably, in subgroup 8 (cytometry), after accounting for various cells, hemoglobin was suggestively associated with GC (OR_SD 0.62 [95% CI 0.41 ~ 0.93]; p = 0.02), but we noted no association with UVMR (OR _SD0.68 [95% CI 0.43 ~ 1.06]; p <0.09). Similarly, when creatine kinase and lactate dehydrogenase (LDH) were examined together in the MVMR cohort, suggestive associations were noted between genetically predicted LDH (OR_SD 0.62 [95% CI 0.41 ~ 0.93]; p < 0.001) and a lower risk of GC (Supplementary Table S6). In addition, the positive associations for genetically predicted levels of serum calcium (OR_SD 1.20 [95% CI 0.85 ~ 1.70]; p = 0.3056), alanine aminotransferase (OR_SD 0.88 [95% CI 0.56 ~ 1.37]; p = 0.5657), and diastolic blood pressure (OR_SD 0.46 [95% CI 0.06 ~ 3.50]; p = 0.4541) became inconclusive in MVMR analyses adjusted for genetically predicted levels of other related ion traits. We are cautious about assuming that the MVMR results are more independent because they are more effective at preventing confounders in the exposure–outcome association. A comparison of the four traits with univariate results is detailed in Supplementary Table S7. To visually represent effect sizes and 95% confidence intervals, we also created a forest plot. (Fig. 4).

This Mendelian randomization study, utilizing genetic variants as proxies for potential risk factors, presents suggestive evidence of associations between increased hemoglobin, alkaline phosphatase, and lactate dehydrogenase levels and a reduced risk of GC. Additionally, indications of potential links between T2DM and GC risk were also observed.

Our application of the MR method allowed us to examine the relationship between hemoglobin and GC risk. In the context of MVMR, we found an association between genetically predicted elevated serum hemoglobin concentrations and a decreased risk of GC. However, our reverse MR analysis did not yield the same association. In patients with cancer, anemia is a common complication and a significant indicator of cancer risk[17]. There are various classifications of anemia, of which the common ones are iron deficiency anemia and pernicious anemia. Hemoglobin levels serve as a critical marker. Extensive studies have demonstrated that anemia is associated with an increased risk of GC. Numerous retrospective, prospective, case‒control studies have demonstrated that iron deficiency and pernicious anemia are associated with an increased risk of GC and a poor prognosis[18–20]. These conclusions are consistent with our findings. We found an association between genetically predicted increased serum hemoglobin concentrations and a reduced risk of GC (OR_SD 0.62 [95% CI 0.41 ~ 0.93]; p = 0.02). Over the past few years, many aspects of the pathophysiology of anemia in cancer have improved. Notably, the pathophysiology of anemia in cancer has been increasingly understood. Investigations into the role of nitrite concentration and nitrosylation reactions in GC have contributed to the understanding of this relationship[21, 22]. Recent views suggest that dysregulation of iron metabolism increases cancer risk and promotes tumor growth, and perturbations in the mechanisms controlling transcripts of iron-regulated proteins, including differences in miRNA, methylation, and acetylation, have been observed in cancer cells[23]. In addition, a prospective study showed that low vitamin B12 increases the risk of GC[24]. Although there is biological plausibility for the relationship between GC and iron intake, our study did not categorize the type of anemia, and due to the lack of GWAS data on serum iron and vitamin B12 in the BBJ database, additional data may be needed for further studies in the future. Our study suggested that anemia, as a modifiable factor, is expected to improve the body’s immunity and anemia status from a nutritional point of view by maintaining adequate nutrient intake, especially by increasing the number of foods rich in iron, vitamin B12, and folic acid, which may reduce the risk of gastric cancer.

Several prospective studies have highlighted alkaline phosphatase activity as an independent risk factor for cancer development. While bone metastasis due to GC is rare, ALP is often used as a marker for such metastasis[25]. In our study, genotyping and imputed genetic data were employed to elucidate the potential causal association between ALP and GC risk. Surprisingly, we observed an association between increased ALP levels and a reduced risk of GC. However, further investigations are essential for elucidating the intricate biological mechanisms underlying these associations.

Our findings of an association between genetically predicted LDH concentrations and GC risk (OR_SD 0.62 [95% CI 0.41 ~ 0.93]; p < 0.001) are concordant with the findings of several observational and cell-related experimental studies. Recent studies suggest that high LDH expression may be an independent risk factor for the prognosis of gastric cancer and that LDH-A might be a potential therapeutic target in gastric cancer[26–28]. LDH-A is a kind of isoform of LDH that is responsible for transforming pyruvate into lactate, an important energy-producing step in cancer cells.[29] However, the molecular mechanism of LDH-A expression and function in GC remains unknown. Opinions have suggested that the HER2-HIF-1α-LDHA axis may promote the progression and metastasis of GC.[30]

T2DM is a chronic progressive disease characterized by years of insulin resistance and hyperinsulinemia preceding the development of hyperglycemia and has been linked to various outcomes, including increased mortality in cancer patients[31]. Most observational studies have suggested an increased risk of GC in patients with T2DM. An umbrella review of meta-analyses of observational studies revealed that the risk for GC was greater in people with diabetes than in those without diabetes control; however, the results were not strongly significant[32]. In addition, a study suggested that T2DM is associated with an increased risk of GC among patients in whom Helicobacter pylori was eradicated, in particular, for gastric cardia cancer and in those with suboptimal diabetes mellitus [33]. The cause of bias in these studies remains unclear due to possible confounding factors and reverse causality in observational studies. Evidence from a number of current studies has supported the role of hyperglycemia and insulin resistance in promoting GC development[34, 35]. However, the causal relationship between diabetes mellitus and the risk of GC is currently inconclusive. Previously, MR analysis also did not reveal strong evidence supporting associations between diabetes and the risk of various cancers[36]. Interestingly, our study revealed the reverse association: genetically predicted groups with T2DM exhibited a reduced risk of GC. The discrepancy between our genetic predictions and these findings could be attributed to the use of these medications. The incidence and prevalence of diabetes are increasing in Japan, driven by an increase in T2DM in recent years. Oral antidiabetic agents were the most common treatment approach (51.4%), and the commonly used drugs were biguanides[37]. The use of several drugs, including metformin, statins, and angiotensin converting enzyme inhibitors/angiotensin receptor blockers (ACEIs/ARBs), has been reported to reduce the incidence of GC[38–40]. A propensity score analysis suggested that metformin improves the survival and recurrence rate of patients with diabetes and gastric cancer[41]. One of the common hypoglycemic drugs, metformin, may have multiple antitumor effects, including regulating miRNA levels in diabetic patients and decreasing the levels of cell cycle proteins in several cancer cell lines; additionally, metformin inhibits cancer cell proliferation by decreasing the expression of EGFR and blocking IGF-1R signaling[42]. More compelling evidence may need to be explored from a drug-targeting perspective.

This study has several limitations. First, because the GC GWAS and espouses GWAS were performed only from BBJ form the IEU open GWAS, we did not have access to another sample population to estimate the effects of SNPs discovered in our GWAS. As our SNPs were discovered and effects estimated in the same population, the effects could have been overestimated due to “winner's curse” phenomena[43]. We were not able to correct for sample overlap. Second, the study was limited by a lack of access to individual-level data, as we had access only to GWAS summary statistics, and the lack of stratification on T2DM status and typing of LDH and ALP may dilute any causal effect. Interpreting the magnitude of estimates for the effect of T2DM on GC risk requires caution. It is possible that only exposure to T2DM within a specific window of time (e.g. During drug treatment) affects GC risk. Third, the sample size and the number of variants included in the analysis were relatively small. For robust insights, combining MR findings with other epidemiologic methodologies is recommended.

In summary, our MR study showed that genetically predicted GC was positively associated with hemoglobin and T2DM in Japanese populations. The findings remained consistent and robust when we used MVMR approaches and when we performed a series of sensitivity analyses. The analysis serves as a valuable reference point for primary prevention strategies targeting GC in the population. Strategies involving the diabetic population and early medication use to control blood sugar levels may offer benefits to GC patients. Additionally, prioritizing the nutritional status of the body and addressing anemia-related causes are crucial for preventing GC. However, further MR studies are needed to explore modifiable risk factors for GC in other East Asian populations.

Acknowledgments

We express our gratitude to the BioBank Japan for providing publicly available summarized data from genome-wide association studies on gastric cancer and a tatol 40 modifiable factors, including activated partial thromboplastin time, alanine aminotransferase, albumin, etc. We also thank the MRC Integrative Epidemiology Unit (IEU) for the publicly available summarized genome-wide association studies data on this study.

Author contributions

Study conception and design: Hao Liu, Wenjun He, Xin Tan, Weihao Yang.

Data acquisition and analysis: Hao Liu, Wenjun He, Xin Tan, Weihao Yang.

Drafting the article and figures: Hao Liu, Wenjun He, Xin Tan, Weihao Yang.Donghua Huang, Hengyi, Zhang

Reviewing the article: Hao Liu, Wenjun He, Xin Tan, Weihao Yang, Donghua Huang, Hengyi, Zhang

The work reported in the paper has been performed by the authors, unless clearly specified in the text.

Availability of data and materials

We thank the participants and researchers for providing the publicly available summary data used in this study. Our study utilized publicly available summary statistics from the International Working Unit (IEU) open GWAS (https://gwas.mrcieu.ac.uk/).

Funding

None

The authors declare no potential conflict of interests.

Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71:209–49.
Smyth EC, Nilsson M, Grabsch HI, van Grieken NC, Lordick F. Gastric cancer. Lancet. 2020;396:635–48.
Suzuki A, Katoh H, Komura D, Kakiuchi M, Tagashira A, Yamamoto S, et al. Defined lifestyle and germline factors predispose Asian populations to gastric cancer. Sci Adv. 2020;6:eaav9778.
Wang F-H, Zhang X-T, Li Y-F, Tang L, Qu X-J, Ying J-E, et al. The Chinese Society of Clinical Oncology (CSCO): Clinical guidelines for the diagnosis and treatment of gastric cancer, 2021. Cancer Commun (Lond). 2021;41:747–95.
Verduijn M, Siegerink B, Jager KJ, Zoccali C, Dekker FW. Mendelian randomization: use of genetics to enable causal inference in observational studies. Nephrol Dial Transplant. 2010;25:1394–8.
Kulhánová I, Znaor A, Shield KD, Arnold M, Vignat J, Charafeddine M, et al. Proportion of cancers attributable to major lifestyle and environmental risk factors in the Eastern Mediterranean region. Int J Cancer. 2020;146:646–56.
Rawla P, Barsouk A. Epidemiology of gastric cancer: global trends, risk factors and prevention. Gastroenterology Rev. 2018;14:26–38.
Lin Y, Zheng Y, Wang H-L, Wu J. Global Patterns and Trends in Gastric Cancer Incidence Rates (1988-2012) and Predictions to 2030. Gastroenterology. 2021;161:116-127.e8.
Sekula P, Del Greco M F, Pattaro C, Köttgen A. Mendelian Randomization as an Approach to Assess Causality Using Observational Data. J Am Soc Nephrol. 2016;27:3253–65.
Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018;362:k601.
Hirata M, Kamatani Y, Nagai A, Kiyohara Y, Ninomiya T, Tamakoshi A, et al. Cross-sectional analysis of BioBank Japan clinical data: A large cohort of 200,000 patients with 47 common diseases. J Epidemiol. 2017;27:S9–21.
Levin MG, Judy R, Gill D, Vujkovic M, Verma SS, Bradford Y, et al. Genetics of height and risk of atrial fibrillation: A Mendelian randomization study. PLoS Med. 2020;17:e1003288.
Davies NM, Holmes MV, Davey Smith G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018;362:k601.
Verbanck M, Chen C-Y, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50:693–8.
Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44:512–25.
Zheng J, Baird D, Borges M-C, Bowden J, Hemani G, Haycock P, et al. Recent Developments in Mendelian Randomization Studies. Curr Epidemiol Rep. 2017;4:330–45.
Knight K, Wade S, Balducci L. Prevalence and outcomes of anemia in cancer: a systematic review of the literature. Am J Med. 2004;116 Suppl 7A:11S-26S.
Hung N, Shen C-C, Hu Y-W, Hu L-Y, Yeh C-M, Teng C-J, et al. Risk of Cancer in Patients with Iron Deficiency Anemia: A Nationwide Population-Based Study. PLOS ONE. 2015;10:e0119647.
Vannella L, Lahner E, Osborn J, Annibale B. Systematic review: gastric cancer incidence in pernicious anaemia. Alimentary Pharmacology & Therapeutics. 2013;37:375–82.
A. KOKKOLA S-MS R HAAPIAINEN, P SIPPONEN, P PUOLAKKAINEN, H JÄRVINEN. The Risk of Gastric Carcinoma and Carcinoid Tumours in Patients with Pernicious Anaemia: A Prospective Follow-Up Study. Scandinavian Journal of Gastroenterology. 1998;33:88–92.
Ruddell WSJ, Bone ES, Hill MJ, Walters CL. PATHOGENESIS OF GASTRIC CANCER IN PERNICIOUS ANÆMIA. The Lancet. 1978;311:521–3.
Dicato M, Plawny L, Diederich M. Anemia in cancer. Annals of Oncology. 2010;21:vii167–72.
Manz DH, Blanchette NL, Paul BT, Torti FM, Torti SV. Iron and cancer: recent insights. Annals of the New York Academy of Sciences. 2016;1368:149–61.
Miranti EH, Stolzenberg-Solomon R, Weinstein SJ, Selhub J, Männistö S, Taylor PR, et al. Low vitamin B12 increases risk of gastric cancer: A prospective study of one-carbon metabolism nutrients and risk of upper gastrointestinal tract cancer. International Journal of Cancer. 2017;141:1120–9.
Lim SM, Kim YN, Park KH, Kang B, Chon HJ, Kim C, et al. Bone alkaline phosphatase as a surrogate marker of bone metastasis in gastric cancer patients. BMC Cancer. 2016;16:385.
Namikawa T, Ishida N, Tsuda S, Fujisawa K, Munekage E, Iwabu J, et al. Prognostic significance of serum alkaline phosphatase and lactate dehydrogenase levels in patients with unresectable advanced gastric cancer. Gastric Cancer. 2019;22:684–91.
Sun X, Sun Z, Zhu Z, Guan H, Zhang J, Zhang Y, et al. Clinicopathological significance and prognostic value of lactate dehydrogenase A expression in gastric cancer patients. PLoS One. 2014;9:e91068.
Liu X, Yang Z, Chen Z, Chen R, Zhao D, Zhou Y, et al. Effects of the suppression of lactate dehydrogenase A on the growth and invasion of human gastric cancer cells. Oncology Reports. 2015;33:157–62.
Kolev Y, Uetake H, Takagi Y, Sugihara K. Lactate dehydrogenase-5 (LDH-5) expression in human gastric cancer: association with hypoxia-inducible factor (HIF-1alpha) pathway, angiogenic factors production and poor prognosis. Ann Surg Oncol. 2008;15:2336–44.
Zhu W, Ma L, Qian J, Xu J, Xu T, Pang L, et al. The molecular mechanism and clinical significance of LDHA in HER2-mediated progression of gastric cancer. Am J Transl Res. 2018;10:2055–67.
Giovannucci E, Harlan DM, Archer MC, Bergenstal RM, Gapstur SM, Habel LA, et al. Diabetes and Cancer: A Consensus Report. CA: A Cancer Journal for Clinicians. 2010;60:207–21.
Tsilidis KK, Kasimis JC, Lopez DS, Ntzani EE, Ioannidis JPA. Type 2 diabetes and cancer: umbrella review of meta-analyses of observational studies. BMJ. 2015;350:g7607.
Cheung KS, Chan EW, Chen L, Seto WK, Wong ICK, Leung WK. Diabetes Increases Risk of Gastric Cancer After Helicobacter pylori Eradication: A Territory-Wide Study With Propensity Score Analysis. Diabetes Care. 2019;42:1769–75.
Tsilidis KK, Kasimis JC, Lopez DS, Ntzani EE, Ioannidis JPA. Type 2 diabetes and cancer: umbrella review of meta-analyses of observational studies. BMJ. 2015;350:g7607.
Tseng C-H, Tseng F-H. Diabetes and gastric cancer: The potential links. World J Gastroenterol. 2014;20:1701–11.
Goto A, Yamaji T, Sawada N, Momozawa Y, Kamatani Y, Kubo M, et al. Diabetes and cancer risk: A Mendelian randomization study. Int J Cancer. 2020;146:712–9.
Neville SE, Boye KS, Montgomery WS, Iwamoto K, Okamura M, Hayes RP. Diabetes in Japan: a review of disease burden and approaches to treatment. Diabetes Metab Res Rev. 2009;25:705–16.
Tseng C-H. Metformin reduces gastric cancer risk in patients with type 2 diabetes mellitus. Aging (Albany NY). 2016;8:1636–49.
Ma Z, Wang W, Jin G, Chu P, Li H. Effect of statins on gastric cancer incidence: a meta-analysis of case control studies. J Cancer Res Ther. 2014;10:859–65.
Sugimoto M, Yamaoka Y, Shirai N, Furuta T. Role of renin-angiotensin system in gastric oncogenesis. J Gastroenterol Hepatol. 2012;27:442–51.
Greenhill C. Gastric cancer. Metformin improves survival and recurrence rate in patients with diabetes and gastric cancer. Nat Rev Gastroenterol Hepatol. 2015;12:124.
Lan W-H, Lin T-Y, Yeh J-A, Feng C-L, Hsu J-T, Lin H-J, et al. Mechanism Underlying Metformin Action and Its Potential to Reduce Gastric Cancer Risk. International Journal of Molecular Sciences. 2022;23:14163.
Kraft P. Curses--winner’s and otherwise--in genetic epidemiology. Epidemiology. 2008;19:649–51; discussion 657-658.

No competing interests reported.

SupplementaryFiles.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Exploring Modifiable Risk Factors: Insights from Mendelian Randomization Analyses of Gastric Cancer in East Asian Populations

Status:

Version 1

Abstract

Figures

Background

Methods

Study design

Date resources

Selection of instrumental variables

Mendelian randomization analysis

Sensitivity analyses

Multivariate MR analyses

Results

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1