肥胖与胰腺癌因果关系的双样本孟德尔随机化研究  

Two-sample Mendelian randomization analysis of the causal relationship between obesity and pancreatic cancer

作　　者：严安 周剑宇[2] 向伟 朱卫东 肖体君[4] 高文哲 朱红伟[1] 伍美容[5] YAN An;ZHOU Jianyu;XIANG Wei;ZHU Weidong;XIAO Tijun;GAO Wenzhe;ZHU Hongwei;WU Meirong(Department of Hepatopancreatobiliary SurgeryⅡ,the Third Xiangya Hospital,Central South University,Changsha 410013,China;Department of Gastrointestinal Surgery I,the Third Xiangya Hospital,Central South University,Changsha 410013,China;Department of General Surgery,the Second Affiliated Hospital of Shaoyang University,Shaoyang,Hunan 422000,China;Department of Nursing,the Third Xiangya Hospital,Central South University,Changsha 410013,China;Second Department of General Surgery,Yueyang Hospital Affiliated to Hunan Normal University,Yueyang,Hunan 414000,China)

机构地区：[1]中南大学湘雅三医院肝胆胰外科Ⅱ,湖南长沙410013 [2]中南大学湘雅三医院胃肠外科Ⅰ,湖南长沙410013 [3]湖南师范大学附属岳阳医院普外二科,湖南岳阳414000 [4]邵阳学院附属第二医院普通外科,湖南邵阳422000 [5]中南大学湘雅三医院护理部,湖南长沙410013

出　　处：《中国普通外科杂志》2024年第8期1287-1299,共13页China Journal of General Surgery

基　　金：国家自然青年科学基金资助项目(82203494);湖南省自然青年基金资助项目(2022JJ40741);湖南省自然科学基金资助项目(2021JJ70051);湖南省卫生健康委科研计划基金资助项目(202204015341);湖南省教育厅科学研究基金资助项目(22A0534);湖南省中医药科研计划基金资助项目(B2024117);中南大学湘雅三医院汇智育才计划基金资助项目(YX202203)

摘　　要：背景与目的:胰腺癌是一种预后极差的恶性肿瘤,而肥胖作为广泛存在的健康问题,被认为与多种癌症相关。本研究利用孟德尔随机化方法(MR)分析肥胖与胰腺癌之间因果关系,为胰腺癌的预防和干预提供科学依据。方法:从IEU数据库下载了8项与人体肥胖相关测量指标的单核苷酸多态性(SNP)数据作为工具变量,并从芬兰R10数据库下载了一项胰腺癌的全基因组关联研究(GWAS)数据。研究采用逆方差加权法(IVW)、孟德尔随机化Egger法(MR Egger)和Cochran Q检验来评估数据的异质性。同时通过IVW、MR Egger、加权中值法、加权模式法、贝叶斯加权孟德尔随机化以及约束最大似然法来评估数据的潜在因果关系。此外,还使用MR PRESSO和留一法来识别并排除异常的SNP,以确保结果的准确性。结果:采用6种MR分析方法对潜在因果联系进行评估,发现胰腺癌与体质量指数(BMI)、基础代谢率、臀围、躯干脂肪量、躯干去脂质量、躯干预测质量和全身去脂体质量存在潜在因果关系(均P<0.05),且均未发现水平多效性和异质性证据。6种不同MR分析方法中β值的一致方向,进一步证实了BMI、臀围和躯干脂肪量的增加与胰腺癌风险增加呈正相关关系,为结果提供稳健性支持。结论:肥胖与胰腺癌存在因果关系,提示通过改善肥胖状态可以降低胰腺癌的风险。Background and Aims:Pancreatic cancer is a malignant tumor with a very poor prognosis,and obesity,as a widespread health issue,is thought to be associated with various cancers.This study uses Mendelian randomization(MR)to analyze the causal relationship between obesity and pancreatic cancer to provide scientific evidence for preventing and intervening in pancreatic cancer.Methods:Single nucleotide polymorphism(SNP)data for eight obesity-related anthropometric measures were downloaded from the IEU database as instrumental variables.Genome-wide association study(GWAS)data for pancreatic cancer were also obtained from the Finnish R10 database.The study employed inverse variance weighting(IVW),Mendelian randomization-Egger(MR-Egger),and Cochran's Q test to assess data heterogeneity.Potential causal relationships were further evaluated using IVW,MR-Egger,weighted median,weighted mode,Bayesian weighted Mendelian randomization,and constrained maximum likelihood Meta-analysis.MR-PRESSO and leave-one-out analyses were also used to identify and exclude outlier SNPs,ensuring result accuracy.Results:Using 6 MR analysis methods to assess potential causal relationships,the study found that pancreatic cancer has a possible causal relationship with body mass index(BMI),basal metabolic rate,hip circumference,trunk fat mass,trunk fat-free mass,trunk predicted mass,and whole body fat-free mass(all P<0.05).No evidence of horizontal pleiotropy or heterogeneity was found.The consistent direction ofβvalues across the 6 different MR analysis methods further confirmed that BMI,hip circumference,and trunk fat mass were positively correlated with an increased risk of pancreatic cancer,providing robustness to the results.Conclusion:There is a causal relationship between obesity and pancreatic cancer,suggesting that improving obesity status may reduce the risk of pancreatic cancer.

关 键 词：胰腺肿瘤 肥胖症 多态性 单核苷酸 孟德尔随机化分析 

分 类 号：R735.9[医药卫生—肿瘤]