Shared genetic architecture between Alzheimer's disease and psychiatric disorders revealed by multi-trait genome-wide analyses

OBJECTIVE: Clarify the shared genetic architecture between Alzheimer’s disease (AD) and major psychiatric disorders. METHODS: We integrated large GWAS summary datasets derived predominantly from individuals of European ancestry for AD and eight psychiatric disorders; estimated genome-wide genetic correlations with LDSC/HDL; mapped local genetic correlations with SUPERGNOVA; ran pairwise Multi-trait analysis of GWAS (MTAG (AD with ADHD, BIP, MDD, PTSD, SCZ)) followed by FUMA; performed cross-trait colocalization with HyPrColoc and trait–eQTL colocalization using GTEx v8 (49 tissues). RESULTS: HDL identified significant genome-wide correlations for five AD–psychiatric pairs, led by AD–MDD; LDSC was significant for AD–MDD. We mapped 18 locally correlated regions. MTAG yielded 33 AD-associated loci (118 SNPs), including 12 novel (e.g., RAB27B/rs12968702, PTCH1/rs3824488, EP300/rs12157997), and 336 psychiatric-trait signals across 265 loci. HyPrColoc detected 74 AD–psychiatric colocalized regions (40 with PP ≥ 0.8), with 13 driven by a single candidate causal variant. Trait–eQTL colocalization prioritized 25 genes in 122 associations; brain-tissue signals implicated P4HTM, GPX1, CCDC71, and—at an AD–SCZ locus—ADAM10. CONCLUSION: AD shares substantial pleiotropic architecture with psychiatric disorders, particularly MDD. Integrative multi-trait association, colocalization, and tissue-specific eQTL evidence highlights convergent mechanisms—exosome biology (RAB27B), astrocytic Ca²⁺ signaling (P4HTM), and non-amyloidogenic APP processing (ADAM10)—and nominates testable therapeutic targets.