Causal effects of sepsis on structural changes in cerebral cortex: A Mendelian randomization investigation

Previous research has shown a strong correlation between sepsis and brain structure. However, whether this relationship represents a causality remains elusive. In this study, we employed Mendelian randomization (MR) to probe the associations of genetically predicted sepsis and sepsis-related death with structural changes in specific brain regions. Genome-wide association study (GWAS) data for sepsis phenotypes (sepsis and sepsis-related death) were obtained from the IEU OpenGWAS. Correspondingly, GWAS data for brain structural traits (volume of the subcortical structure, cortical thickness, and surface area) were derived from the ENIGMA consortium. Inverse variance weighted was mainly utilized to assess the causal effects, while weighted median and MR-Egger regression served as complementary methods. Sensitivity analyses were implemented with Cochran Q test, MR-Egger regression, and MR-PRESSO. In addition, a reverse MR analysis was carried out to assess the possibility of reverse causation. We identified that genetic liability to sepsis was normally significantly associated with a reduced surface area of the postcentral gyrus (β = -35.5280, SE = 13.7465, P = .0096). The genetic liability to sepsis-related death showed a suggestive positive correlation with the surface area of fusiform gyrus (β = 11.0920, SE = 3.6412, P = .0023) and posterior cingulate gyrus (β = 3.6530, SE = 1.6684, P = .0286), While it presented a suggestive negative correlation with surface area of the caudal middle frontal gyrus (β = -11.4586, SE = 5.1501, P = .0261) and frontal pole (β = -1.0024, SE = 0.4329, P = .0206). We also indicated a possible bidirectional causal association between genetic liability to sepsis-related death and the thickness of the transverse temporal gyrus. Sensitivity analyses verified the robustness of the above associations. These findings suggested that genetically determined liability to sepsis might influence the specific brain structure in a causal way, offering new perspectives to investigate the mechanism of sepsis-related neuropsychiatric disorders.