Causal relationship between inflammatory factors and cerebral small vessel disease: Univariate, multivariate, and summary‐data‐based mendelian randomization analysis

• Published in: Brain and behavior Vol. 14; no. 2; pp. e3399 - n/a
• Main Authors: Qiao, Tian‐Ci, Tian, Hao‐Yu, Shan, Shi‐Zhe, Shan, Li‐Li, Peng, Zheng‐Yu, Ke, Jia, Li, Meng‐Ting, Wu, Yang, Han, Yan
• Format: Journal Article
• Language: English
• Published: United States John Wiley & Sons, Inc 01.02.2024; Wiley
• Subjects: cerebral small vessel disease; Cerebral Small Vessel Diseases - diagnostic imaging; Cerebral Small Vessel Diseases - genetics; Confounding (Statistics); Datasets; Disease; DNA methylation; E-Selectin; Estimates; Gene expression; Genome-Wide Association Study; Genomes; Humans; inflammatory factors; mendelian randomization; Mendelian Randomization Analysis; Risk Factors; Software; summary‐data‐based mendelian randomization; summary-data-based mendelian randomization; cerebral small vessel disease; inflammatory factors; mendelian randomization
• ISSN: 2162-3279; 2162-3279
• DOI: 10.1002/brb3.3399

Objective To explore the impact of inflammatory factors on the incidence of cerebral small vessel disease (CSVD), we performed a mendelian randomization (MR) study to analyze the causal relationship between multiple inflammatory factors and CSVD imaging markers and utilized summary‐data‐based mendelian randomization (SMR) analysis to infer whether the impact of instrumental variables (IVs) on disease is mediated by gene expression or DNA methylation. Methods Using public databases such as UKB and IEU, and original genome‐wide association studies, we obtained IVs related to exposure (inflammatory factors) and outcome (CSVD imaging markers). We performed the inverse variance weighted, weighted median, and MR‐Egger methods to assess causal effects between exposure and outcome in univariate MR analysis. To evaluate their heterogeneity, a series of sensitivity analyses were conducted, including the Cochrane Q test, MR‐Egger intercept test, MR‐Presso, and leave‐one‐out analysis. We also applied mediation and multivariate MR analysis to explore the interactions between positive exposures on the same outcome. Additionally, we conducted the SMR, which utilizes instruments within or near relevant genes in blood or brain tissues, to elucidate the causal associations with CSVD markers. Results ABO Univariate MR of multiple cohorts revealed that the risk of small vessel stroke (SVS) increases with elevated levels of TNF‐related apoptosis‐inducing ligand (TRAIL, OR, 1.23, 95% CI, 1.08–1.39) and interleukin‐1 receptor‐like 2, (IL‐1RL2, OR, 1.29, 95% CI, 1.04–1.61). IL‐18 was a potential risk factor for extensive basal ganglia perivascular space burden (BGPVS, OR, 1.02, 95% CI, 1.00–1.05). Moreover, the risk of extensive white matter perivascular space burden (WMPVS) decreased with rising levels of E‐selectin (OR, .98, 95% CI, .97–1.00), IL‐1RL2 (OR, .97, 95% CI, .95–1.00), IL‐3 receptor subunit alpha (IL‐3Ra, OR, .98, 95% CI, .97–1.00), and IL‐5 receptor subunit alpha (IL‐5Ra, OR, .98, 95% CI, .97–1.00). Mediation and multivariate MR analysis indicated that E‐selectin and IL‐3Ra might interact during the pathogenesis of WMPVS. SMR estimates showed that TRAIL‐related IVs rs5030044 and rs2304456 increased the risk of SVS by increasing the expression of gene Kininogen‐1 (KNG1) in the cerebral cortex, particularly in the frontal cortex (βsmr = .10, Psmr = .003, FDR = .04). Instruments (rs507666 and rs2519093) related to E‐selectin and IL‐3Ra could increase the risk of WMPVS by enhancing DNA methylation of the gene ABO in blood tissue (βsmr = .01–.02, Psmr = .001, FDR = .01–.03). Conclusion According to MR and SMR analysis, higher levels of TRAIL increased the risk of SVS by upregulating gene expression of KNG1 in brain cortex tissues. In addition, protective effects of E‐selectin and IL‐3a levels on WMPVS were regulated by increased DNA methylation of gene ABO in blood tissue.