Analysis of impact of type 1 diabetes on colorectal cancer by using two-sample Mendelian randomization
10.3969/j.issn.1674-8115.2024.06.011
- VernacularTitle:两样本孟德尔随机化法分析1型糖尿病对结直肠癌的影响
- Author:
Yang YU
1
;
Dan MENG
;
Yiwen QIU
;
Jian YUAN
;
Yingjie ZHU
Author Information
1. 上海中医药大学附属龙华医院肿瘤一科,上海 200032
- Keywords:
colorectal cancer;
type 1 diabetes;
Mendelian randomization
- From:
Journal of Shanghai Jiaotong University(Medical Science)
2024;44(6):755-761
- CountryChina
- Language:Chinese
-
Abstract:
Objective·To investigate the potential causal relationship between type l diabetes and colorectal cancer by using Mendelian randomization(MR).Methods·Two-sample bidirectional MR was used to investigate the causal relationship between type 1 diabetes and colorectal cancer.All research data were collected from the IEU Open GWAS Project database.The dataset of type l diabetes included 9 266 cases and 15 574 controls,with correlation analysis in 12 783 129 single nucleotide polymorphisms(SNPs);the dataset of colorectal cancer included 5 657 cases and 372 016 controls,with correlation analysis in 29 999 696 SNPs.The instrumental variables SNPs were screened.The results derived from the inverse-variance weighted(IVW)method were used as the main indicator of effect.The results derived from other four methods,namely MR-Egger regression,weighted median,simple mode,and weighted mode,were used as reference.Sensitivity was analyzed with the leave-one-out method.Heterogeneity was analyzed with Cochran's Q test by using both IVW and MR-Egger methods.Pleiotropy was analyzed with MR-pleiotropy function,and Steiger test was used for directional research.The colocation analysis was used to find out whether the causal relationship between type 1 diabetes and colorectal cancer was caused by the same SNP.The genetic correlation between 2 diseases was analyzed by using the linkage disequilibrium score regression(LDSC).All tests were analyzed by using R language software(version 4.3.1).Results·After being screened,a total of 33 instrumental variables(SNPs)were used.The heterogeneity test results showed that there was heterogeneity among the SNPs(IVW and MR-Egger:P＜0.05),so the effect evaluation was based on the results of the random effect model.MR analysis showed that type 1 diabetes had a significant causal effect on colorectal cancer(P＜0.05)by using IVW,MR-Egger,weighted median and weighted mode.Sensitivity analysis showed that the results were stable.Pleiotropy was not detected in pleiotropy test(P＞0.05).Steiger test showed that the effect of type l diabetes on colorectal cancer was not interfered with by the reverse effect.Reverse MR analysis showed no causal effect of colorectal cancer on type l diabetes(P＞0.05).The results of colocalization analysis showed that the probability of H4 hypothesis was 45.7％,and the causal relationship between the 2 diseases was not caused by the same SNP in the gene sequences.LDSC analysis demonstrated that there was no genetic correlation between the two diseases.Conclusion·Type 1 diabetes may promote colorectal cancer,but colorectal cancer has no effect on type 1 diabetes.
