强直性脊柱炎可能增加精神障碍的风险：基于两样本孟德尔随机化方法

Ankylosing spondylitis may increase the risk of mental disorders: a two-sample Mendelian randomization study

摘要:

目的 采用两样本孟德尔随机化（two-sample Mendelian randomization，TSMR）分析方法，探讨强直性脊柱炎（ankylosing spondylitis，AS）与精神障碍之间潜在的因果关联，从遗传学角度为AS患者早期预防精神障碍的发生发展提供参考。方法 通过对非重叠全基因组关联研究公开数据进行筛选，以AS作为暴露数据，精神障碍（双相情感障碍和重度抑郁症、神经性厌食症、阿尔茨海默病、焦虑病）作为结局数据，采用逆方差加权法（inverse-variance weighted，IVW）作为该研究评估因果效应的主要分析方法，MR-Egger法、加权中位数法（weighted median）、加权模型（weighted mode）、简单模型（simple mode）等作为补充方法，并通过Cochran’s Q、MR-Egger-intercept进行敏感性分析。结果 ①IVW结果表明遗传预测的AS与双相情感障碍和重度抑郁症（OR=1.055，95%CI=1.019~1.093，P=0.003）、神经性厌食症（OR=1.370，95%CI=1.068~1.759，P=0.013）呈正向因果关系；与阿尔茨海默病（OR=0.976，95%CI=0.955~0.997，P=0.029）呈负向因果关系；与焦虑症（OR=1.034，95%CI=0.906~1.179，P=0.620）之间未发现因果关联。②敏感性分析未发现存在异质性以及水平多效性（P>0.05），表明研究结果没有产生偏倚，留一法也提示结果稳健。结论 本研究提供了AS与精神障碍之间因果关系的新证据，AS可能与双相情感障碍和重度抑郁症及神经性厌食症风险增加有关，为精神障碍的遗传研究提供了新的见解，但未来仍需更大样本量的全基因组关联研究数据通过进一步的孟德尔随机化研究来验证这种关系。

Abstract:

Objective To investigate the possible causal link between ankylosing spondylitis（AS） and mental disorders through a two-sample Mendelian randomization（MR） analysis，and to offer genetic insights into the early prevention of the development and progression of mental disorders in patients with AS.Methods Using the public data from non-overlapping genome-wide association studies，AS was considered as the exposure variable，and mental disorders （bipolar affective disorder and major depression，anorexia nervosa，Alzheimer’s disease，and anxiety disorder） as the outcome variables. We used inverse variance weighting（IVW） as the main analysis to assess the causal effect，with the MR-Egger method，weighted median method，weighted mode，and simple mode for supplementary causal analyses. The Cochran’s Q test and MR-Egger intercept test were used for sensitivity analysis.Results The IVW results revealed that genetically predicted AS had positive causal associations with bipolar affective disorder and major depression（odds ratio ［OR］=1.055，95%CI=1.019-1.093，P=0.003） and anorexia nervosa（OR=1.370，95%CI=1.068-1.759，P=0.013） and a negative causal association with Alzheimer’s disease（OR=0.976，95%CI=0.955-0.997，P=0.029），but with no causal association with anxiety disorder（OR=1.034，95%CI=0.906-1.179，P=0.620）. The sensitivity analysis detected no heterogeneity or horizontal pleiotropy（P>0.05），which indicated no bias in the results. The leave-one-out analysis suggested that the results were robust.Conclusion This study provides new evidence for causal relationships between AS and mental disorders. AS may increase the risk of bipolar affective disorder and major depression and anorexia nervosa. These findings offer new insights into the genetic research of mental disorders. However，the conclusions need further MR-based validation with larger-sample genome-wide association data.