Shared host-genetic architecture between gut microbiota and internalizing psychopathology

Whether gut microbial composition is causally linked to mental illness, or merely correlated with it, remains unresolved. Using genetic variants as natural instruments (Mendelian randomization, MR), we tested the genetically predicted effects of 211 gut microbial taxa on nine psychiatric and psychopathology-related phenotypes, using the largest available genome-wide association studies. Across 1 898 valid tests, seven taxon-outcome associations passed false-discovery-rate correction (FDR < 0.05), and all fell on the internalizing spectrum (depression, neuroticism and insomnia) rather than on bipolar disorder, schizophrenia or ADHD; they included a protective association of the Mollicutes/Tenericutes clade with depression (β = -0.073, p = 1.5 ×10 ^-6 ) and of Butyrivibrio with neuroticism, and a deleterious association of Betaproteobacteria with neuroticism. Conservative tests tempered any per-locus causal reading: Bayesian colocalization gave PP.H4 < 0.05 at every locus and CAUSE found 0 of 45 pairs genuinely causal; yet the direction of effect matched the protective-versus-deleterious hypothesis in 33 of 45 pairs (binomial p = 1.2 ×10 ^-3 ). Modelling the shared genetics of the nine phenotypes placed these taxa specifically on a latent internalizing factor (correlation 0.48 with a separate psychotic factor) and, in a bifactor model, on internalizing-specific genetic variance beyond a general psychopathology factor. Selected gut microbial taxa and internalizing psychopathology therefore appear to share host genetics rather than a direct microbe-to-disorder causal chain. We release the full analysis as an open resource for larger microbiome studies, brain-tissue follow-up and experimental tests of candidate mechanisms.