Objective To explore the correlation between body mass index(BMI)and gynecological benign and malignant diseases(endometrial cancer,ovarian cancer,cervical cancer,uterine leiomyoma,and endometriosis)using a two-sample Mendelian randomization(MR)method.Methods BMI data encoded as ieu-b-40 from the IEU OpenGWAS public database were used in this study,and there were 681 275 samples from European and East Asian populations.A total of 506 single nucleotide polymorphism(SNP)related to BMI were screened using R language software.The two-sample analysis was applied with BMI as the exposure factor.Significant SNPs in GWAS data were selected as instrumental variables which were ensured independent of each other.The causal relationship between exposure factor of gene prediction(BMI)and outcome indicators were explored by separate analysis using the onset risk of 5 gynecological benign and malignant diseases as outcome factors.Statistical analysis was performed by the TwoSampleMR package in R language(4.3.1)software.MR analysis was performed by inverse variance weighted(IVW),MR Egger regression,weighted median(WM),simple mode(SM),and weighted mode(WM).The results were expressed by as odds ratio(OR)and 95％Cl.Sensitivity analysis was conducted by heterogeneity test,horizontal pleiotropy analysis,and leave-one-out method to evaluate the potential bias of MR analysis and significant impact of an instrumental variable on outcome variables.Results The 506 SNPs were strongly correlated with BMI,and their F values were all greater than 10.Reverse MR analysis of endometrial cancer identified 16 SNPs closely related to endometrial cancer,and they were strong instrumental variables.The results of IVW,MR Egger,SM,WM,and WM suggested that genetically predicted BMI may be involved in the occurrence and development of endometrial cancer,and there was a positive causal relationship between BMI and the occurrence and development of endometrial cancer(OR＞1).However,no genetic predictive causal association was found between BMI and other gynecological benign and malignant diseases(ovarian cancer,cervical cancer,uterine leiomyoma,and endometriosis).Furthermore,the IVW results of reverse MR analysis showed that the occurrence and development of endometrial cancer may not affect changes in BMI values(OR=1.020,95％CI:0.969-1.073,P=0.45).The MR-Egger-intercept analysis did not detect potential horizontal pleiotropy(P=0.19 for endometrial cancer,P=0.99 for ovarian cancer,P=0.73 for cervical cancer,P=0.76 for uterine leiomyoma,and P=0.08 for endometriosis),indicating that the instrumental variable did not significantly affect the outcomes of the five gynecological diseases through pathways other than BMI.Meanwhile,no horizontal pleiotropy was detected in the reverse MR analysis of endometrial cancer and BMI(P=0.09).The results of Cochran's Q heterogeneity test showed potential heterogeneity in both forward and reverse MR analyses of BMI and endometrial cancer(forward MR Egger:P=6.48×10-4,IVW:P=5.88×10-4;reverse MR Egger:P=0.05,IVW:P=1.78×10-7),but BMI was not significantly heterogeneous with other gynecological diseases.The sensitivity test results of the leave-one-out method showed that the MR analysis results generated by sequentially removing individual SNPs were basically consistent with the MR analysis results of including all SNPs,with an estimated error line of about 0,indicating that there was no impact of a single SNP on the overall causal estimation.The funnel plot showed that the distribution of included SNPs was basically balanced and not affected by potential factors.Conclusion There is a positive causal relationship between gene prediction of BMI and the occurrence and development of endometrial cancer.BMI may have a certain effect on controlling the occurrence and development of endometrial cancer.
关键词
体质指数/妇科良恶性疾病/孟德尔随机化/子宫内膜癌
Key words
Body mass index/Gynecological benign and malignant diseases/Mendelian randomization/Endometrial cancer
维普
万方数据