弓形虫感染与精神分裂症的两样本孟德尔随机化研究

目的 采用两样本孟德尔随机化方法探究弓形虫感染和精神分裂症之间的因果关系。 方法 利用汇总的大样本GWAS数据提取与弓形虫血清抗体密切相关的遗传位点作为工具变量，分别运用MR-Egger回归、加权中位数和逆方差加权法进行孟德尔随机化分析，以OR值及95%CI评价弓形虫感染与精神分裂症之间是否存在关联。采用Egger-intercept法检验水平多效性，“leave-one-out”法进行敏感性分析。 结果 从GWAS汇总数据共提取12个SNPs作为工具变量，MR-Egger回归、加权中位数和逆方差加权法计算得OR值和95%CI分别为0.964（95%CI0.937~0.993）、0.982（95%CI0.954~1.011）、0.981（95% CI 0.957~1.005），所得结果基本一致，且统计效能Power为100%。逆方差加权法结果显示两者之间的关联无统计学意义，尚不能认为弓形虫感染与精神分裂症之间存在因果关系。MR-Egger法显示截距为0.012，P=0.097，截距接近于0，即所选的工具变量不存在水平多效性。“leave-one-out”敏感性分析显示结果稳定，不存在对结果有强影响的工具变量。 结论 孟德尔随机化分析结果显示，弓形虫感染与精神分裂症之间不存在因果关联。

Analysis of the association between toxoplasma gondii infection and schizophrenia with two-sample Mendelian randomization method

To explore the causal relationship between toxoplasma gondii and schizophrenia using Mendelian randomization (MR) analysis. Methods SNP was extracted from GWAS data, and was employed as instrument variable (IV) closely related to Toxoplasma gondii infection. Odds ratio and 95% confidence interval were used as outcome indicator. The causal relationship between Toxoplasma gondii infection and schizophrenia was analyzed by two-sample MR method including MR-Egger, weighted median estimator (WME), and inverse variance weighted (IVW). MR-egger regression method was carried out to test genetic pleiotropy. The leave-one-out method was performed for sensitivity analysis. Results 12 SNPs were picked up for IV, odds ratio with 95% confidence interval calculated by MR-egger regression. WME and IVW were 0.964 (0.937-0.993), 0.982(0.954-1.011) and 0.981(0.957-1.005) respectively, and statistical Power was 100%. IVM outcome indicated that Toxoplasma gondii was not associated with the risk of schizophrenia. The MR-egger regression results suggested that genetic pleiotropy was unlikely to bias our outcome (the intercept=-0.036, P=0.097). There was no specific single SNP examined by leave-one-out sensitivity analysis. Conclusion The result of two-sample Mendelian randomization analysis reveals no casual effects between toxoplasma gondii infection and schizophrenia.