Behavioral and Neural Heterogeneity in OCD and Depression

Project Summary This project investigates behavioral and neural heterogeneity in obsessive-compulsive disorder (OCD) and major depressive disorder (MDD) using hypothesis- and data-driven approaches. OCD and MDD are major sources of disability worldwide, with fewer than half of patients responding adequately to first-line treatments. Treatment may be impeded by the fact that OCD and MDD are both highly heterogeneous, with clusters of symptoms likely derived from differing psychological and neurobiological etiologies. The Research Domain Criteria (RDoC) framework seeks to address this problem in part by investigating dimensional components of behavior that span across disorders and more closely align with brain circuitry than does the multifaceted diagnostic phenotype. In the present project, we follow an RDoC approach to the study of heterogeneity by investigating relationships between behavior, clinical symptoms, and brain function using task-based and resting-state fMRI in a sample of 75 OCD patients and 75 MDD patients. Our preliminary data in OCD patients and individuals with depressive symptoms point to interoceptive sensitivity (IS) and perseverative negative thinking (PT) as dimensional components of behavior that are not only transdiagnostic but also explain important within-disorder heterogeneity; the first aim of this proposal builds on these findings by investigating associations of IS and PT with clinical symptoms, behavior, and brain function in OCD and MDD. Complementing this hypothesis-driven approach, the second aim of the project is a data-driven investigation using unsupervised machine learning (ML) to uncover neurobiological profiles that characterize variability in brain function across OCD and MDD. Building on our preliminary analysis identifying three subgroups within OCD distinguished by differential patterns of resting-state functional connectivity, we will investigate neurobiological heterogeneity across OCD and MDD and link clinical and behavioral data to neurobiological subgroups using a comprehensive ‘deep phenotyping’ approach to data acquisition and analysis. The overarching goals of both aims are to identify factors contributing to heterogeneity that can be developed as novel targets for personalized treatments based on an individual patient’s neurobehavioral profile.