CAUSAL RELATIONSHIPS BETWEEN STRUCTURAL BRAIN IMAGING MARKERS IN ADOLESCENCE AND PSYCHIATRIC DISORDERS IN ADULTHOOD

Abstract Background Adolescent brain imaging markers have been identified as potential predictors of adult psychiatric disorders, but it remains unclear whether brain development during childhood and adolescence causally results in psychiatric disorders. Aims & Objectives The goal of our research is to better understand the causal relationship between youth’ s cerebral cortex and three primary psychiatric disorders: bipolar disorder (BIP), major depressive disorder (MDD), and schizophrenia (SCZ). Method We implemented a two-sample Mendelian randomization (MR) approach to estimate the causal effects of adolescent cortical thickness (CT) and cortical surface area (SA) of 34 regions of interest (ROIs) on the three mental disorders. We integrated imaging and genotypic data from the Adolescent Brain Cognitive Development (ABCD) Study and the summary results of genome-wide association studies (GWAS) from the ENIGMA consortium (N=33,992) and the Psychiatric Genomics Consortium (BIP: 41,917 cases and 371,549 controls; MDD: 170,756 cases and 329,443 controls; SCZ:74,776 cases and 101,023 controls) to conduct the causal analysis. We computed the genetic correlations (rg) between the cortical measures and mental disorders. We employed MR to investigate the causal relationship between ROIs that are genetically associated with the mental disorders. Results The global measure of SA was positively correlated with BIP (rg: 0.068 and P-value: 0.037) and inversely correlated with MDD (rg: -0.080 and P-value: 0.009). We identified significant genetic correlations between SA of seven ROIs and BIP, and SA of twelve ROIs and MDD. We identified that SA of the superior frontal as a potential causal region for BIP (OR: 1.22 and 95% CI: 1.10-1.36) based on MR causal analysis. Other ROIs with nominally significant causal effects (P <0.05) included SA of the caudal middle frontal on BIP, and the SA of the four regions (medial orbitofrontal, lateral orbitofrontal, inferior temporal and lateral occipital) on MDD. Discussion & Conclusion Our findings suggest that larger SA in early adolescence may causally increase the risk of BIP, with superior frontal being the potential causal ROI. Additionally, we identified potential ROIs such as medial and lateral orbitofrontal regions, which are negatively associated with MDD.