目的：通过孟德尔随机化（Mendelian randomization，MR）探讨线粒体与癫痫之间的因果关系。方法： 与66种线粒体丰富度显著相关的SNP被用作工具变量。采用逆方差加权平均法（IVW）、MR Egger回归、加 权中位数（Weighted median）确定暴露因子与癫痫发生之间的因果关系。以及使用Cochran’s Q检验、留一 法分析和MR-PRESSO检测结果的稳健性。结果：3个数据集和全身性癫痫有显著因果关系，IVW的结果显 示[丙酮酸脱氢酶（乙酰转移）]激酶同工酶 1 与全身性癫痫的风险升高有关（OR=1.275，95%CI=1.056～ 1.540，PIVW=0.011，PFDR=0.017），rRNA 甲基转移酶 3 与全身性癫痫的风险升高有关（OR=1.158，95%CI= 1.002～1.338，PIVW=0.047，PFDR=0.047），而基本的 MCU 调节剂与全身性癫痫的风险降低有关（OR=0.727， 95%CI=0.578～0.915，PIVW=0.007，PFDR=0.020），5个数据集和局灶性癫痫具有显著因果关系。IVW结果表明 核糖体循环因子与局灶性癫痫的风险升高有关（OR=1.085，95%CI=1.000～1.178，PIVW=0.049，PFDR=0.061）， 蛋白质赖氨酸脱酰基酶 Sirtuin 5 与局灶性癫痫的风险升高有关（OR=1.080，95%CI=1.001～1.167，PIVW= 0.048，PFDR=0.079），补体成分 1 Q 亚组分结合蛋白与局灶性癫痫的风险降低有关（OR=0.917，95%CI= 0.861～0.977，PIVW=0.007，PFDR=0.018），丝氨酸- tRNA连接酶与局灶性癫痫的风险降低有关（OR=0.852，95% CI=0.779～0.934，PIVW=0.001，PFDR=0.003），NFU1 铁硫簇支架同源物与局灶性癫痫的风险升高有关（OR= 1.098，95%CI=1.000～1.204，PIVW=0.049，PFDR=0.049）。敏感性检验确定了其结构的稳健性（均P＞0.05）。结 论：建立了线粒体与癫痫之间的因果关系。

To explore the causal relationship between mitochondria and epilepsy using Mendelian randomization (MR). Methods: Single nucleotide polymorphisms (SNPs) significantly associated with mitochondrial abundance across 66 traits were utilized as instrumental variables. The inverse variance weighted (IVW) method, MR Egger regression, and weighted median approach were employed to determine the causal relationship between exposure factors and the onset of epilepsy. Additionally, Cochran’s Q test, leave-one-out analysis, and MR-PRESSO were used to assess the robustness of the results. Results: Significant causal relationships were identified between three datasets and generalized epilepsy. The IVW results indicated that [pyruvate dehydrogenase (lipoamide) kinase isozyme 1] was associated with an increased risk of generalized epilepsy (OR=1.275, 95%CI=1.056～1.540, PIVW=0.011, PFDR=0.017), rRNA methyltransferase 3 was also linked to a heightened risk of generalized epilepsy (OR=1.158, 95% CI=1.002～1.338, PIVW=0.047, PFDR=0.047), while the essential MCU regulator was associated with a reduced risk of generalized epilepsy (OR=0.727, 95% CI= 0.578～0.915, PIVW=0.007, PFDR=0.020). Five datasets exhibited significant causal relationships with focal epilepsy. The IVW results revealed that ribosome recycling factor was associated with an increased risk of focal epilepsy (OR=1.085, 95% CI=1.000～1.178, PIVW=0.049, PFDR=0.061), protein lysine deacylase Sirtuin 5 was also linked to a heightened risk of focal epilepsy (OR=1.080, 95% CI=1.001～1.167, PIVW=0.048, PFDR=0.079), complement component 1 Q subcomponent-binding protein was associated with a reduced risk of focal epilepsy (OR=0.917, 95% CI=0.861～0.977, PIVW=0.007, PFDR=0.018), serine--tRNA ligase was also linked to a decreased risk of focal epilepsy (OR=0.852, 95% CI=0.779～0.934, PIVW=0.001, PFDR=0.003), and NFU1 iron-sulfur cluster scaffold homolog was associated with an increased risk of focal epilepsy (OR=1.098, 95% CI=1.000～1.204, PIVW=0.049, PFDR=0.049). Sensitivity tests confirmed the robustness of the findings (all P>0.05). Conclusion: A causal relationship between mitochondria and epilepsy has been established.