Multivariate patterns between brain network properties, polygenic scores, phenotypes, and environment in preadolescents

Abstract

AbstractThe brain network is an infrastructure for cognitive and behavioral processes. Genetic and environmental factors influence the development of the brain network. However, little is known about how specific genetic traits and children’s brain network properties are related. Furthermore, insight into the holistic relationship of brain network properties with genes, environment, and phenotypic outcomes in children is still limited. To fill these knowledge gaps, we investigated the multivariate associations between the brain network properties and three domains using a large youth sample (the ABCD study, N=9,393, 9-10 years old): (i) genetic predisposition of various traits, (ii) phenotypic outcomes, and (iii) environmental factors. We constructed structural brain networks using probabilistic tractography and estimated nodal and global network measures such as degree and network efficiency. We then conducted sparse canonical correlation analysis with brain network measures and polygenic scores of 30 complex traits (e.g., IQ), phenotypic traits (e.g., cognitive ability), and environmental variables. We found multivariate associations of brain network properties with (i) genetic risk for psychiatric disorders, (ii) genetic influence on cognitive ability, and (iii) the phenotype of cognitive ability-psychopathology in preadolescents. Our subsequent mediation analysis using the latent variables from the canonical correlation analysis showed that the influence of genetic factors for cognitive ability on the cognitive outcomes was partially mediated by the brain network properties. Taken together, this study shows the key role of the development of the brain structural network in children in cognitive development with its tight, likely causal, relationship with genetic factors. These findings may shed light on future studies of the longitudinal deviations of those gene-environment-brain network relationships in normal and disease conditions.