Building a multi-factor etiological model of the emergence of general psychopathology (the "P factor") in adolescence with multi-modal neuroimaging in ABCD

Abstract There is substantial evidence that psychopathology is structured hierarchically. In addition to two major specific factors, internalizing and externalizing, there is a single overarching general factor, the “P factor”, that explains a sizable share of variance in psychiatric symptoms. The P factor model represents a major recent advance in our understanding of the architecture of psychopathology. However, there is a critical gap in our current knowledge: We have little understanding of the neurodevelopmental etiological factors that produce the P factor during youth. We have an ideal opportunity to address this gap with the Adolescent Brain Cognitive Development (ABCD) longitudinal study (n=11,875; 4 biennial waves of data over the course of this five-year grant). We have formulated a Dual Dysmaturation Model in which the P factor arises from altered maturation during adolescence in two systems: executive control systems (leading to globally reduced higher-order cognition and inhibitory control) and impulse generation systems (leading to globally elevated impulse generation). We have also undertaken a comprehensive analysis of psychopathology data in ABCD to derive and validate a superordinate general psychopathology factor (“P factor”). Our overarching aim in this project is to build on these results and delineate the multi-factor etiology of the P factor in youth in ABCD, integrating knowledge across socio-environmental, psychological, neural, and genetic levels of analysis. More specifically we seek to achieve four aims. Aim 1 uses a latent growth modeling approach to quantify co- development of psychological variables (including executive functions, negative emotions, and aggressive impulses) with the emergence of the P factor over adolescence. In Aim 2, we use advanced methods to fractionate brain imaging maps into a small number of cohesive components. We then use latent growth modeling to identify brain components that co-develop with the P factor. For Aim 3, we delineate genetic factors that contribute to the P factor. For this aim, we identify brain components that mediate the relationship between polygenic risk for the P factor and the emergence of the P factor in late adolescence. For the Aim 4, we integrate the preceding factors (psychological, neural, and genetic) with additional socio-environmental variables to build an overall nomological network for the emergence of the P factor, and we distinguish this network from analogous networks for the emergence of internalizing and externalizing specific factors. By bringing together the seminal ABCD dataset and advanced multivariate multi-modal neuroimaging methods, this project will give us important new mechanistic insights into the multi-factor neurodevelopmental etiology of the P factor. This knowledge is a key input to downstream research programs, such as programs that seek to identify high-risk youth or to develop interventions that mitigate or block the emergence of psychopathology.