This study marks the pioneering exploration of the potential causal links between sepsis and cortical volume across diverse brain regions and also endeavors to assess the reverse causal effects of cortical volume on sepsis and its subtypes by using the bidirectional MR method. Supported by a range of statistical approaches, the results revealed that Streptococcal sepsis risk exhibited negative correlations with the cortical volume of the Left Precentral Gyrus and Right Cingulate Gyrus, while a positive correlation emerged with the Right Supramarginal Gyrus. Similarly, Pneumococcal septicemia demonstrated negative associations with cortical volume in the Left Hippocampus and Right Lingual Gyrus. Furthermore, Other sepsis risk displayed a negative association with the Left I-IV Cerebellum. The inverse MR analysis highlighted that genetic predisposition towards the Right Cuneal Cortex and Right Inferior Frontal Gyrus displayed negative correlations with Streptococcal sepsis, whereas genetic predisposition towards the Right Thalamus was negatively associated with Pneumococcal septicemia.
Previous investigations into the impact of sepsis on cortical volume primarily relied on observational neuroimaging studies, leaving questions about the causal relationships between sepsis and specific cortical regions largely unanswered. Prospective studies have observed reduced cortical volume in sepsis-induced brain dysfunction, particularly in regions like the cerebral cortex, hippocampus, and amygdala (12). Another MRI examination has additionally revealed neuronal loss in areas such as the insula, cingulate cortex, frontal lobe, precuneus, and thalamus (13). Notably, as far as the hippocampus is concerned, several studies have documented hippocampal atrophy in sepsis patients compared to healthy controls (19, 20). Another animal experiment also observed that bacteremia with meningitis plays a vital role in the development of hippocampal damage in pneumococcal meningitis (21), which is consistent with our observations. The hippocampus is renowned for its key role in memory, spatial navigation, and cognition (22). Sepsis is widely acknowledged to induce sepsis-associated encephalopathy and cognitive impairments (23), often leaving survivors with long-term memory issues, reduced attention spans, and deficits in verbal fluency and executive function upon hospital discharge (24). The potential relationship between sepsis and hippocampal volume offers a promising avenue for elucidating this connection. The results of our bidirectional MR analysis further support a causal relationship between sepsis and its subtypes and the volumes of specific cortical regions. However, there remains a scarcity of research investigating the impact of sepsis on cortical volumes in other specific brain regions, temporarily leaving our pioneering work somewhat unsupported by prior studies. Therefore, further investigations are imperative to determine the connection between sepsis and particular cortical volumes.
The underlying mechanism by which multiple sepsis subtypes are negatively correlated with specific regional cortical volumes may be attributable to sepsis-induced neuroinflammation, as shown in Fig. 4. Sepsis is characterized by a dysregulated systemic inflammatory response (25), closely linked to long-term cognitive deficits (26). During sepsis, systemic endotoxemia and pro-inflammatory cytokines disrupt the blood-brain barrier by triggering microglia overactivation and endothelial barrier dysfunction (27). Increased blood-brain barrier permeability renders the central nervous system vulnerable to neurotoxic factors (28), including free radicals and inflammatory mediators, leading to oxidative stress and brain cell dysfunction. Pro-inflammatory cytokines infiltrate the brain parenchyma, inducing shifts in oxidative stress levels and impairing brain cell function (29). These cytokines bind to receptors on the surface of brain cells, intensifying the inflammatory response (30). Sepsis affects different brain regions unequally, with heightened susceptibility in the cortex and hippocampus (30), where affected neurons can undergo apoptosis (31). Mitochondrial dysfunction also plays a significant role in sepsis-induced brain injury, with damaged mitochondria generating ROS and RNS and releasing DAMPs (32, 33). These effects can cause structural damage to cell membranes and induce inflammation, leading to apoptosis of neurons and changes in cortical volume (34). This MR study reveals a positive causal relationship between streptococcal sepsis and cortical volume in the right marginal superior gyrus, contradicting previous research (35), and warranting further study for conclusive accuracy.
Inverse MR analysis revealed negative correlations between genetic susceptibility to streptococcal sepsis and the volumes of the right cuneate cortex and right inferior frontal gyrus. Similarly, genetic susceptibility to pneumococcal sepsis exhibited negative correlations with right thalamus volumes. These associations may be linked to the neurovascular coupling hypothesis, where blood perfusion is finely regulated to meet the increased metabolic demands of a larger cerebral cortex (36). Given that severe sepsis and septic shock are characterized by significantly reduced overall perfusion, resulting in hypotension, unequal distribution of regional blood flow, and inadequate tissue perfusion (37), an increased cerebral perfusion may contribute to greater sepsis tolerance. In fact, MR results in our study may not precisely mirror clinical associations, as they significantly rely on the strength of the original GWAS data (38).
This MR has several notable strengths. Foremost, it stands as the pioneering endeavor to scrutinize the causal impact of sepsis and its subtypes on cortical volume across diverse regions through MR analysis. The credibility of MR results often surpasses those from traditional observational research because it mitigates biases stemming from confounders and reverse causality. Additionally, our summary data were sourced from the most comprehensive GWAS meta-analysis encompassing populations of European ancestry to date which not only enhanced our statistical robustness but also yielded compelling conclusions. Furthermore, multiple sensitivity analyses demonstrated that our findings remained unaffected by pleiotropy and heterotropy, confirming the robustness of our conclusions.
However, some limitations of the study can't be ignored. It's important to acknowledge several limitations in our study. Firstly, our findings are derived from GWAS data conducted primarily on individuals of European ancestry, which raises questions about the generalizability of these results to other ethnic groups. Nevertheless, the homogeneity of our participant pool helped minimize the risk of population-mixing confounding. Secondly, it's crucial to note that the OR value we present may not necessarily reflect the magnitude of the effect but rather the presence of a causal relationship. Additionally, while our partial MR analysis didn't yield significant results, it could provide valuable insights. Lastly, our study lacks direct observational and mechanistic investigations to substantiate our findings. To comprehensively understand the relationship between sepsis and cortical volume, further research is warranted to explore the impact of sepsis on the immune system, blood-brain barrier, and the structure and function of the nervous system.