Lou Jiaqi, Xiang Ziyi, Zhu Xiaoyu, Fan Youfen, Li Jiliang, Jin Guoying, Cui Shengyong, Huang Neng
Burn Department, Ningbo No. 2 Hospital, Ningbo, China.
Institute of Pathology, Faculty of Medicine, University of Bonn, Bonn, Germany.
Medicine (Baltimore). 2025 Apr 25;104(17):e42199. doi: 10.1097/MD.0000000000042199.
Whether there is a causal relationship between circulating levels of systemic inflammatory regulators and sepsis remains unclear. To determine whether genetically predicted circulating levels of cytokines are associated with risk of sepsis, a bidirectional two-sample Mendelian randomization (MR) analysis based on the a STROBE-compliant cross-sectional observational study was conducted utilizing gene-wide association study (GWAS) data. Selected with rigor, single-nucleotide polymorphisms served as instrumental variables for subsequent MR analysis. The preferred method for the MR analysis was the inverse-variance weighted approach. However, for comprehensive sensitivity analyses, 6 additional MR methods were employed. Cochrane's Q test was performed to examine heterogeneity. A leave-one-out method ensured the stability of MR results. Our findings suggest an inverse association between the levels of beta-nerve growth factor (BNGF) and the risk of sepsis development (OR = 0.769, 95% CI = 0.599-0.987, P = .039). In contrast, higher levels of TNF-related apoptosis-inducing ligand and vascular endothelial growth factor A (VEGF-A) are positively correlated with sepsis risk (OR = 1.094, 95% CI = 1.012-1.183, P = .025; OR = 1.182, 95% CI = 1.016-1.375, P = .031, respectively). Reverse MR Analysis indicated that sepsis risk is linked with lower circulating levels of adenosine deaminase and Interleukin-17A (β = -0.043, 95% CI = -0.085 to -0.002, P = .042; β = -0.061, 95% CI = -0.108 to -0.013, P = .012, respectively), and also with higher circulating levels of BNGF, delta/notchlike epidermal growth factor-related receptor, fibroblast growth factor 23, leukemia inhibitory factor, monocyte chemoattractant protein-1, and osteoprotegerin (β = 0.056, 95% CI = 0.015-0.096, P = .007; β = 0.137, 95% CI = 0.035-0.240, P = .009; β = 0.118, 95% CI = 0.020-0.216, P = .018; β = 0.136, 95% CI = 0.020-0.252, P = .022; β = 0.143, 95% CI = 0.043-0.242, P = .005; β = 0.116, 95% CI = 0.010-0.222, P = .031, respectively). Sum up, our study provides evidence supporting a bidirectional causal relationship between sepsis and genetically predicted circulating levels of systemic inflammatory regulators.
全身炎症调节因子的循环水平与败血症之间是否存在因果关系仍不清楚。为了确定基因预测的细胞因子循环水平是否与败血症风险相关，我们基于一项符合STROBE标准的横断面观察性研究，利用全基因组关联研究（GWAS）数据进行了双向两样本孟德尔随机化（MR）分析。经过严格筛选，单核苷酸多态性被用作后续MR分析的工具变量。MR分析的首选方法是逆方差加权法。然而，为了进行全面的敏感性分析，我们还采用了另外6种MR方法。进行Cochrane's Q检验以检查异质性。留一法确保了MR结果的稳定性。我们的研究结果表明，β-神经生长因子（BNGF）水平与败血症发生风险呈负相关（OR = 0.769，95% CI = 0.599 - 0.987，P = 0.039）。相反，较高水平的肿瘤坏死因子相关凋亡诱导配体和血管内皮生长因子A（VEGF - A）与败血症风险呈正相关（OR = 1.094，95% CI = 1.012 - 1.183，P = 0.025；OR = 1.182，95% CI = 1.016 - 1.375，P = 0.031，分别）。反向MR分析表明，败血症风险与较低的腺苷脱氨酶和白细胞介素 - 17A循环水平相关（β = -0.043，95% CI = -0.085至 -0.002，P = 0.042；β = -0.061，95% CI = -0.108至 -0.013，P = 0.012，分别），并且还与较高的BNGF、δ/Notch样表皮生长因子相关受体、成纤维细胞生长因子23、白血病抑制因子、单核细胞趋化蛋白 - 1和骨保护素循环水平相关（β = 0.056，95% CI = 0.015 - 0.096，P = 0.007；β = 0.137，95% CI = 0.035 - 0.240，P = 0.009；β = 0.118，95% CI = 0.020 - 0.216，P = 0.018；β = 0.136，95% CI = 0.020 - 0.252，P = 0.022；β = 0.143，95% CI = 0.043 - 0.242，P = 0.005；β = 0.116，95% CI = 0.010 - 0.222，P = 0.031，分别）。总之，我们的研究提供了证据支持败血症与基因预测的全身炎症调节因子循环水平之间的双向因果关系。
