Relationship between mental disorders and non-traumatic cerebral hemorrhage: cross-sectional analysis and mendelian randomization

Background The causality between mental disorders and spontaneous non-traumatic cerebral hemorrhage (NCH) has not been sufficiently explored. This study aims to investigate the associations of a broad spectrum of 16 mental disorders with NCH through Mendelian randomization (MR) analysis and to assess the association in a real-world study. Methods Univariate analysis and mediational MR analysis were employed to investigate the potential association between mental disorders and NCH. Genetic instruments associated with mental disorders were identified through a genome-wide association study (GWAS) using the FinnGen database of European populations. Furthermore, the MR results were corroborated based on the real population (MIMIC-IV database). The correlation between mental disorders and the risk of NCH was examined using logistic regression models, which progressively added confounding factors for adjustments. Results Univariate MR analysis showed that genetically predicted dementia was significantly associated with NCH (OR = 1.090, 95% CI [1.049–1.133], p = 1.16 × 10 −5 ); However, no statistically significant associations were observed between NCH and other mental disorders, including attention-deficit/hyperactivity disorder, Lewy body dementia, depression, eating disorders, gender identity disorder, habit and impulse disorders, obsessive-compulsive disorder, persistent mood disorders, phobic anxiety disorders, post-traumatic stress disorder, sleep disorders, somatoform disorders, anxiety disorders, autism spectrum disorders, and bipolar disorder. Reverse MR analysis further indicated that there is no reverse causal relationship between dementia and NCH. When several confounding factors were taken into account, observational studies revealed that dementia was linked to an increased risk of NCH (OR = 1.31, 95% CI [1.07–1.59], p = 0.007). Additionally, we did not observe any intermediate factors that can mediate dementia and NCH. Conclusions Results of both MR analyses and observational studies showed that dementia was associated with the risk of NCH. Furthermore, reverse MR analysis revealed no causal effect of NCH on dementia.