Occupational creativity and liability to psychiatric disorders: a bidirectional Mendelian randomization study

Abstract

The long-standing “mad genius” hypothesis proposes a link between creativity and psychiatric illness, but existing genetic evidence has largely relied on genetic correlations and polygenic associations rather than approaches designed to evaluate directional effects. We conducted bidirectional two-sample Mendelian randomization (MR) using a machine-learning-derived, occupation-level occupational-creativity GWAS from 241,736 European-ancestry UK Biobank participants and GWAS summary statistics for schizophrenia (SCZ), bipolar disorder (BD), major depressive disorder (MDD), attention-deficit/hyperactivity disorder (ADHD), autism spectrum disorder (ASD), and anorexia nervosa (AN). Random-effects inverse-variance weighted MR was the primary estimator, supported by sensitivity analyses for heterogeneity, pleiotropy, directionality, and education- and cognition-related genetic architecture. In forward univariable MR, genetically predicted occupational creativity was inversely associated with ADHD liability (β = −0.518; 95% CI, − 0.821 to − 0.214; FDR q = 0.0031) and MDD liability (β = −0.273; 95% CI, − 0.445 to − 0.102; FDR q = 0.0049). The ASD estimate was positive but sensitivity dependent. No robust forward evidence was observed for SCZ, BD, or AN. Reverse analyses suggested inverse associations of ADHD and MDD liability with genetically predicted occupational creativity, but heterogeneity and pleiotropy diagnostics supported cautious interpretation. Education-adjusted multivariable MR attenuated the ADHD and MDD associations, indicating that the univariable estimates were not clearly independent of education-related genetic architecture. These findings challenge a simple genetic “mad genius” model and suggest that associations between genetically predicted occupational creativity and psychiatric liability are heterogeneous and partly reflect education-related and occupational-selection pathways rather than direct effects of creativity itself.