A large-scale causal analysis of gut microbiota and delirium: A Mendelian randomization study.

BACKGROUND: Several studies have linked gut microbiota to human brain activity. This study used Mendelian randomization (MR) to investigate the causal relationship between gut microbes and delirium.

METHODS: MR was used to select SNPs from large-scale GWAS summary data on 211 gut microbiota taxa and delirium. Inverse variance weighting method (IVW), weighted median, and MR-Egger tests were used for statistical analyses. Outliers were assessed using the leave-one-out method. To avoid horizontal pleiotropy, we performed the MR-PRESSO and MR-Egger intercept tests. Cochran's Q and I2 values for IVW and MR-Egger were used to assess heterogeneity.

RESULTS: IVW suggested that genetic prediction of the family Desulfovibrionaceae (1.784 (1.267-2.512), P = 0.001), order Desulfovibrionales (1.501 (1.058-2.128), P = 0.023), and genus Candidatus Soleaferrea (1.322 (1.052-1.659), P = 0.016) increased the risk of delirium, but the family Oxalobacteraceae (0.841 (0.722-0.981), P = 0.027), and genera Holdemania (0.766 (0.620-0.946), P = 0.013), Ruminococcus gnavus (0.806 (0.661-0.982), P = 0.033), and Eggerthella (0.815 (0.667-0.997), P = 0.047) reduced the risk of delirium.

LIMITATIONS: (1) Limited sample size, (2) inability to assess gut microbiota interactions, and (3) limited to European populations.

CONCLUSION: Our results suggest that presence of the microbial family Desulfovibrionaceae, order Desulfovibrionales, and genus Candidatus Soleaferrea increased the risk of delirium, whereas the Oxalobacteraceae family, and the genera Holdemania, Ruminococcus gnavus, and Eggerthella decreased the risk of delirium. However, the potential of gut probiotic interventions in the prevention of perioperative delirium should be emphasized.