Differential Causal Associations of Gut Microbiota, Blood Metabolites, and Immune Cell Phenotypes With Early- and Late-Onset Alzheimer's Disease: A Bidirectional Mendelian Randomization Analysis.

内容摘要

Alzheimer's disease (AD) is a heterogeneous syndrome with distinct genetic and clinical profiles between early-onset AD (EOAD) and late-onset AD (LOAD) subtypes. However, specific causal etiologies linking the gut microbiota-immune-metabolic axis to these subtypes remain poorly understood. We employed a comprehensive bidirectional two-sample Mendelian randomization (MR) framework to systematically investigate the causal associations of gut microbiota, immune cell phenotypes, and blood metabolites with EOAD and LOAD. Large-scale genome-wide association study (GWAS) summary statistics were utilized from the MiBioGen consortium, Sardinian cohort, and Canadian Longitudinal Study on Aging, alongside AD outcome data from the FinnGen consortium. Causal estimates were generated using the inverse variance-weighted method, with rigorous sensitivity analyses including false discovery rate (FDR) correction and Steiger directionality tests to ensure robustness. Our analysis revealed divergent multi-omics signatures for AD subtypes. While the genus Veillonella and myeloid dendritic cells emerged as shared protective factors, the risk profiles were distinct. EOAD susceptibility was primarily driven by adaptive immune dysregulation and lipid metabolism disturbances. In contrast, LOAD risk was predominantly associated with innate immune dysfunction and perturbations in amino acid and gut-derived metabolite turnover, such as hippurate. This study provides genetic evidence that EOAD and LOAD are driven by fundamentally different peripheral mechanisms across the gut-immune-metabolic axis. These findings challenge the monolithic view of AD pathogenesis and underscore the critical necessity of stratifying patients by onset age to develop precision therapeutic interventions.