Longitudinal Analysis of Diffusion Tensor Imaging to Discover Adolescent Alcohol Use Effect PROJECT ABSTRACT Alcohol abuse is the third leading preventable cause of death in the United States. A signature injury of Alcohol Use Disorder (AUD) is in the white-matter (WM) microstructure and its constituents, which enable connectivity of proximal and distal brain structures and functional integration. Despite the progress in understanding alcohol’s effects in adults, still unclear is the causal direction between abnormal WM maturation and initiation of heavy yet non-dependent drinking during adolescence. To gain insight into regional development of connectivity, the National Consortium on Alcohol and NeuroDevelopment in Adolescence (NCANDA), a prospective longitudinal study of 831 youth to track normal and alcohol-related deviant neurodevelopmental trajectories, acquired diffusion tensor imaging (DTI) data annually. Unraveling alcohol effects from the complex dynamic course of adolescent neurodevelopment requires highly sensitive analysis approaches. In this project, I propose a new type of longitudinal DTI analysis, in which a unified trajectory model is used both for explicitly capturing biologically-plausible variation in DTI measurements across visits within each subject and for estimating a group-level developmental trajectory. Such an approach will result in longitudinally consistent DTI measurements enabling accurate characterization of microstructural development in normal adolescents. Based on this approach, I will test the hypothesis that precursors of drinking onset and the disruption effects induced by initiation of alcohol use are related to different brain regions. The analysis will identify the precursors by comparing developmental trajectories prior to drinking onset between the no-to-low and heavy drinking cohorts and will identify the disruption by comparing trajectories after drinking onset. The alcohol- induced disruption to WM maturation will be further stratified with respect to age, highlighting the heightened vulnerability in younger adolescents. Lastly, the proposed longitudinal analysis also facilitates further tracking of microstructural remodeling following abstinence to identify reversible or persistent alcohol-related injury. Revealed imaging phenotypes linked to adolescent heavy drinking could point to potential causes and precursors of AUD, which could in turn improve diagnosis and prevention in clinical settings. Moreover, researchers will be able to use the proposed longitudinal approach as a general tool to answer their neuroscientific questions in the context of seeking developmental change.