Neural Signatures of Bipolar Disorder across Depression and Remission

University Of California Los Angeles, Los Angeles CA

Investigators

Linked publications & trials

Abstract

Project Summary Bipolar disorder is characterized by variability in mood symptoms and cognitive impairment which leads to functional impairment both during depression and even during periods of remission. A major clinical challenge is that patients exhibit substantial variability in treatment response despite having the same clinical diagnosis, which suggests heterogeneity in underlying neurobiological dysfunction. In this R03, our aims are to de?ne speci?c frontal- cingulate cognitive control and amygdala-frontal emotion regulatory neural clusters in brain activation and connectivity using dimensional tasks of cognitive control and acute threat. We will also develop neurocognitive and emotional pro?les of the neural clusters we identify. To accomplish our aims, we will analyze an existing dataset of functional magnetic resonance imaging (fMRI) scans, obtained in previous NIMH studies at the UCLA Mood Disorders Research Program, from bipolar disorder subjects (65 remitted and 21 depressed) and 65 healthy controls. We will use a novel approach, called Dimensional Neural Clustering, to classify fMRI data. Our approach is based on techniques used in other serious illnesses with multiply- determined outcomes, such as breast cancer, that have revolutionized treatment by identifying subgroups and made invaluable life-saving contributions. After computing brain activation and connectivity associated with the neural circuits underlying cognitive control (activated by response inhibition) and acute threat (activated by emotion regulation), we will use Dimensional Neural Clustering to identify subgroups of patients. Dimensional Neural Clustering differs from previous machine-learning approaches because we will use neurobiology to de?ne subgroups of patients, rather than to classify patients based on a clinical diagnosis. We hypothesize that patients can be classi?ed based on inhibitory and regulatory neural clusters. Moreover, the neural clusters will be associated with differential phenotypic expression characterized by neurocognitive performance, mood/anxiety symptoms, and the ability to reappraise/suppress emotion. Our demonstration of neural subtypes of bipolar disorder will provide a basis for a neurobiologically-driven approach to understanding bipolar disorder, on which we can elaborate in future dimensional work. Our ?ndings will provide a foundation for future work to personalize treatment.