牙周炎与干燥综合征的因果关系：一项孟德尔随机化研究

Bidirectional casual effects between periodontitis and Sjögren's syndrome: a Mendelian randomization study

摘要

目的 利用孟德尔随机化方法（Mendelian randomization，MR）探索牙周炎与干燥综合征（Sjögren's syndrome，SS）的双向因果关系。方法 选取符合同一人种且来自不同地区的牙周炎（N = 45 563）及SS（N = 214 435）的全基因组关联研究（genome-wide association study，GWAS）数据。采用逆方差加权（inverse variance weighted，IVW）、MR-Egger回归、加权中位数（weighted median，WM）法评估因果效应;采用Cochran's Q检验、MR-Egger回归截距项、MR-PRESSO、留一法进行敏感性分析，评估结果的稳定性和可靠性。结果 筛选后的SS和牙周炎GWAS数据分别来自芬兰、英国，且均为欧洲人种数据。采用IVW（OR = 1.017，95%CI = 0.956~1.082）、MR-Egger（OR = 0.985，95%CI = 0.956 ~ 1.082）、WM（OR =1.021， 95%CI = 0.948 ~ 1.099）三种方法均未发现SS对牙周炎存在的效应;反之，三种方法（IVW，OR = 1.024， 95%CI = 0.852~1.230;MR-Egger，OR = 0.978， 95% CI = 0.789~1.212;WM，OR = 1.024，95% CI = 0.846 ~ 1.260）亦未发现牙周炎对SS的因果效应。各项敏感性分析表明结果稳定可靠。Cochran's Q检验、MR-PRESSO表明纳入的工具变量——单核苷酸多态性（single nucleotide polymorphism，SNP）之间无显著异质性;MR-Egger回归截距项显示纳入的SNP不存在多效性;留一法未发现显著影响结果的SNP。结论 MR分析不支持牙周炎与SS之间存在双向因果关联。

Abstract

Objective To evaluate the bidirectional association between periodontitis and Sjögren's syndrome using the Mendelian randomization (MR) method. Methods Genome-wide association study (GWAS) data of periodontitis (N = 45 563) and Sjögren's syndrome (N = 214 435) were selected to meet the requirements of the same ethnicity and different regions. Inverse variance-weighted (IVW), MR-Egger, and weighted median (WM) tests were used to evaluate the causal effect. Cochran's Q statistics, MR-Egger intercept, MR-PRESSO and leave-one-out analysis were used as sensitivity analyses to assess the stability and reliability of the results. Results After screening, the GWAS data of Sjögren's syndrome were based on the Finnish region, and the periodontitis GWAS data were based on the UK region, both of which originated from European ancestry. Using IVW (OR = 1.017, 95% CI = 0.956-1.082), MR-Egger (OR = 0.985, 95% CI= 0.956-1.082), and WM (OR =1.021, 95% CI = 0.948-1.099), no causal effect of Sjögren's syndrome on periodontitis was found using any of the three methods. Conversely, no causal effect of periodontitis on Sjögren's syndrome was found (IVW, OR = 1.024, 95% CI = 0.852-1.230; MR-Egger, OR = 0.978, 95% CI = 0.789-1.212; WM, OR = 1.024, 95% CI = 0.846-1.260). The sensitivity analyses indicated that the results were stable and reliable. Cochran's Q test and MR-PRESSO revealed that there was no significant heterogeneity among the instrumental variables, which included single nucleotide polymorphisms (SNPs). The intercept of MR-Egger regression indicated no pleiotropy in the included SNPs. No individual SNP was found that significantly affected the results using the leave-one-out method. Conclusion This study does not support a bidirectional causal effect between periodontitis and Sjögren's syndrome.

关键词

牙周炎 / 干燥综合征 / 孟德尔随机化 / 全基因组关联分析 / 因果关联 / 双向因果 / 混杂因素 / 逆方差加权

Key words

periodontitis / Sj?gren's syndrome / Mendelian randomization / genome-wide association study / causal association / bidirectional casual association / confounder / inverse variance-weighted

Author summay

谢培莉，本科，Email：[e-mail]