The multivariate genetic architecture of psychiatric and insulin resistance multimorbidity
Nina Roth Mota,Giuseppe Fanelli,Izel Erdogan,Marieke Klein,Emma Sprooten,Abel Fothi,I. Hyun Ruisch,Geert Poelmans,Kazi Asraful Alam,Jan Haavik,Theresa Wimberley,Martina Arenella,Alessandro Serretti,Chiara Fabbri,Barbara Franke,Janita Bralten
DOI: https://doi.org/10.1101/2024.10.02.24314704
2024-10-02
Abstract:Psychiatric disorders frequently co-occur with insulin resistance (IR)-related conditions, including obesity, type 2 diabetes mellitus (T2DM), and metabolic syndrome (MetS). Although pairwise genetic correlations have been observed, the shared genetics underlying this multimorbidity remains underexplored. Here, we investigate the joint genetic architecture of psychiatric-IR multimorbidity, explore tissue-specific gene expression associations, and identify potential underlying biological mechanisms and repurposable drugs. We applied genomic structural equation modeling (SEM) to genome-wide association study (GWAS) data (N=9,725-933,970) from five psychiatric disorders (attention-deficit/hyperactivity disorder, anorexia nervosa, major depressive disorder, obsessive-compulsive disorder, and schizophrenia) and three IR-related conditions (MetS, obesity, T2DM). Factor analyses revealed a 2-factor solution, where one of the factors was composed by all psychiatric disorders (excluding schizophrenia) and IR-related conditions (the Psych-IR factor), representing the shared genetics of these psychiatric and IR-conditions. This factor showed genetic correlations with the inferior temporal, lateral occipital, and total cortical brain surface areas. A multivariate GWAS of the Psych-IR factor identified 150 risk loci and 366 associated genes (128 novel). The significant gene-set associations included the insulin binding and the Notch signaling pathways, while the gene-property tissue expression implicated the cerebellum, brain cortex, and pituitary gland, particularly involving the brain during prenatal development stages. Transcriptome-wide SEM (T-SEM) assessed tissue-specific gene expression associations and identified 499 genes (191 novel), including MHC-related genes. Drug repurposing analysis using PharmOmics suggested eight potential candidates, including memantine and rosiglitazone. Associated genes derived from the Psych-IR factor multivariate GWAS and T-SEM results were combined for enrichment analyses, which highlighted the involvement of the chr16p11.2 region, BDNF signaling, and lipid metabolism. The identified Psych-IR factor offers novel insights into the shared genetic and biological mechanisms underlying psychiatric-IR multimorbidity, providing a foundation for future research on precision medicine and prevention approaches.