Genetically predicted optimal mediator models for oxidative stress, immunity and intracerebral hemorrhage

The influence of oxidative stress (OS) and cellular traits on ICH onset remains uncertain. We aimed to investigate the mediating role of cellular traits between OS biomarkers and ICH using multiple Mendelian randomization (MR) methods.

We included summarized data from seven sources to select 19 OS biomarkers, including trace elements, metabolites, proteins, functional composite structures, and 24-hour dietary quantification. All biomarkers were detected in blood and recorded in Online follow-up using standardization methods. Firstly, we estimated the causal relationships between 19 OS biomarkers and multiple sources of ICH using univariable MR and meta-analyses. Concurrently, two-step MR was employed to determine the mediating role of cellular traits between OS and ICH. Subsequently, multivariable MR and MR Bayesian model averaging were applied to identify the optimal mediation models. Additionally, gene enrichment analyses were performed to explore pathways linking OS to ICH.

Significant causal relationships exist between OS biomarkers and ICH, including lobar and deep ICH. In optimal mediation models, CD11b on CD66b++ myeloid cell potentially mediated the causal effect of plasma albumin on ICH, monocyte percentage of white cells mediated the causal effect of catalase on ICH, and CD19 on unswitched memory B cell potentially mediated the causal effect of kynurenine on lobar ICH. Pathway analyses suggested that OS contributes to the ICH onset by disrupting vascular endothelial function.

The immunity may mediate part of the causality of OS on ICH. A better understanding of the OS and immunity may help improve the prevention and monitoring of ICH.