Genetically Proxied Associations Among 473 Gut Microbial Taxa, 1,400 Circulating Metabolites, and Mental and Behavioral Disorders: A Mendelian Randomization Study.

Mental and behavioural disorders substantially impair everyday functioning, yet the biological mechanisms through which the gut microbiome may contribute to psychiatric vulnerability remain insufficiently characterized. We jointly analysed genome-wide association study data for 473 gut microbial taxa, approximately 1,400 circulating metabolites, and seven mental and behavioural disorders, including schizophrenia, obsessive-compulsive disorder, post-traumatic stress disorder, attention-deficit/hyperactivity disorder, chronic depression, bulimia nervosa, and hypersomnia. Mendelian randomization was used to estimate genetically proxied associations among these traits, and two-step MR was used to evaluate candidate statistical mediation by circulating metabolites. We identified FDR-supported associations for eight microbial taxa across four disorders: UBA8904, Dorea, and Ruminococcus A sp000432335 with ADHD; Pseudomonas aeruginosa, Bifidobacterium kashiwanohense, and CAG-273 sp003534295 with chronic depression; Syntrophorhabdia with bulimia nervosa; and Aneurinibacillales with hypersomnia. Twelve circulating metabolites were statistically consistent with partial mediation of candidate microbiota-disorder associations. Glycocholate glucuronide (1) accounted for an estimated 5.95% of the Syntrophorhabdia-bulimia nervosa association, whereas 5-dodecenoate (12:1n7) was consistent with mediation of 8.28% of the inverse association between CAG-273 sp003534295 and chronic depression. Target enrichment prioritized lipid metabolism and neuroactive ligand-receptor interaction pathways for further study. The inverse association between genetically predicted 5-dodecenoate and depression was reproduced in an independent depression GWAS. These findings nominate candidate microbiota-metabolite pathways for mechanistic investigation; however, residual horizontal pleiotropy and linkage-disequilibrium-related confounding cannot be excluded, and the results should not be interpreted as definitive causal evidence.