The association between gut microbiota and delirium: A systematic review.

BACKGROUND: Postoperative delirium (POD) is an acute neurocognitive complication associated with prolonged hospitalization, mortality, and long-term functional decline. Gut microbiota may contribute to POD through gut-brain axis pathways involving inflammation, barrier dysfunction, neurotransmission, and metabolism, but the available evidence remains heterogeneous. OBJECTIVE: To synthesize human and animal evidence on the association between gut microbiota alterations and delirium, particularly POD, and to evaluate the supporting mechanistic evidence. METHODS: Following PRISMA guidance, PubMed, Embase, CINAHL, MEDLINE, CNKI, Wanfang, and SinoMed were searched from inception to 2 August 2024. Reference lists were manually screened. Ten studies, including five human studies and five animal studies, were included. Because of substantial heterogeneity in populations, delirium definitions, surgical or clinical contexts, and microbiota profiling methods, findings were synthesized narratively and stratified by study type, microbial diversity, differentially abundant taxa, and reported functional pathways. RESULTS: Human studies reported recurrent enrichment of potentially pro-inflammatory taxa, particularly Proteobacteria and Enterobacteriaceae, and lower abundance of selected potentially protective or SCFA-associated genera, including Lactobacillus and Olsenella. However, alpha-diversity findings were inconsistent across studies. Animal studies showed surgery- or anesthesia-related microbiota alterations involving taxa such as Akkermansia, together with metabolite-related pathways involving indole-3-propionic acid and palmitic amide. Mechanistic evidence for inflammation, gut-barrier disruption, and metabolite-neuroimmune pathways was stronger in animal models than in human observational studies. CONCLUSIONS: Gut microbiota alterations are associated with delirium and POD-related phenotypes, but current evidence may not establish causality. Interpretation is limited by the small number of human studies, heterogeneous delirium ascertainment and microbiota profiling, and incomplete control of perioperative confounders. Standardized longitudinal clinical studies and intervention trials are needed before microbiota- or metabolite-based strategies can be recommended.