基于孟德尔随机化的精神分裂症与药物滥用和酒精使用障碍的因果效应研究  

Causal effects of schizophrenia on substance abuse and alcohol use disorder by Mendelian randomization

作　　者：霍康明 崔明湖[2] 张中文[1] Huo Kangming;Cui Minghu;Zhang Zhongwen(School of Public Health,Binzhou Medical University,Yantai 264003,China;Binzhou Medical University Hospital,Binzhou 256603,China)

机构地区：[1]滨州医学院公共卫生学院,山东烟台264003 [2]滨州医学院附属医院,山东滨州256603

出　　处：《中国医院统计》2024年第6期401-407,共7页Chinese Journal of Hospital Statistics

基　　金：山东省统计科学应用研究课题重点项目(2023TJZD005);大学生创新创业训练计划国家级项目(202210440051,202310440115);滨州医学院2021年教学改革与研究项目(JYKTMS2021059);2022年滨州医学院青年骨干教师培养计划项目。

摘　　要：目的评估精神分裂症与药物滥用和酒精使用障碍的因果效应,明确疾病与行为之间的因果关联,为改进临床治疗、改善预后提供新的思路与方案。方法使用精神分裂症、药物滥用和酒精使用障碍的全基因组关联研究的汇总统计数据,在满足孟德尔随机化工具变量三大假设的条件下,剔除具有混淆因素的工具变量,并对数据进行方向协同处理。以逆方差加权法为主要分析方法,加权中值法、MR Egger法、简单模式和加权模式为补充分析方法,对数据进行双样本孟德尔随机化分析。运用Cochran Q检验、MR-Egger intercept分析和leave-one-out法进行敏感性分析,检验其异质性和多效性,评估结果的稳定性与可靠性。结果为排除弱工具变量影响,使工具变量与暴露强相关,采用比以往研究的P值阈值(5×10^(-8))更严格的P值阈值(5×10^(-10)),共纳入64个单核苷酸多态性位点作为工具变量。逆方差加权法估计精神分裂症患者患药物滥用的风险(OR)是一般人群的1.19倍(95%CI:1.126~1.261,P=1.499×10^(-9)),其他4种分析方法的结果也表明精神分裂症会增加药物滥用的风险。对于精神分裂症与2组酒精使用障碍的数据,逆方差加权法评估精神分裂症患者患酒精使用障碍的风险(OR)分别是一般人群的1.17倍(95%CI:1.105-1.239,P=7.398×10^(-8))和1.13倍(95%CI:1.072~1.187,P=3.124×10^(-6))。加权中值法、简单模式和加权模式的评估结果同样为精神分裂症会增加酒精使用障碍的风险提供了证据。结论精神分裂症与药物滥用和酒精使用障碍存在正向因果效应,为后续的临床研究提供了重要思路和依据。Objective To assess the causal effect of schizophrenia on substance abuse and alcohol use disorder,to clarify the causal associations between diseases and behaviors,and to provide new ideas and protocols for improving clinical treatment and prognosis.Methods Summary statistics from genome-wide association studies of schizophrenia,substance abuse,and alcohol use disorder were used to exclude instrumental variables with confounding factors and directionally harmonize the data,provided that the three major assumptions of Mendelian randomization of instrumental variables were met.Two-sample Mendelian randomization of the data was performed using the inverse variance weighted method as the main analytical method,with the weighted median method,MR Egger method,simple mode and weighted mode as supplementary analytical methods.Sensitivity analyses were performed using Cochran Q test,MR-Egger intercept analysis and leave-one-out method to test for heterogeneity and pleiotropy and to assess the stability and reliability of the results.Results In order to exclude the effect of weak instrumental variables and to make instrumental variables strongly correlated with exposure,a more stringent threshold than that of previous studies was used,and a total of 64 single nucleotide polymorphisms loci were included as instrumental variables with the P-value threshold set at 5 × 10^(-10).The inverse variance weighted methoi estimated that the risk(OR) of developing substance abuse in schizophrenia was 1.19 times higher than that of the general population(95% CI:1.126-1.261,P=1.499 × 10^(-9)),and the results of the other four analyses also indicated that schizophrenia increased the risk of substance abuse.For the data on schizophrenia and two groups of alcohol use disorder,the inverse variance weighted method estimated that the risk(OR) of alcohol use disorder in patients with schizophrenia was 1.17 times(95% CI:1.105-1.239,P=7.398 × 10^(-8)) and 1.13 times(95% CI:1.072-1.187,P=3.124 ×10^(-6)) that of the general population,respecti

关 键 词：精神分裂症 药物滥用 酒精使用障碍 因果效应 孟德尔随机化 

分 类 号：R74[医药卫生—神经病学与精神病学]