Transdiagnostic connectome-based prediction of craving

Abstract Craving is a central construct in the study of motivation and human behavior and is also a clinical symptom of substance and non-substance-related addictive disorders. Thus, craving represents a target for transdiagnostic modeling. We applied connectome-based predictive modeling (CPM) to functional connectivity data in a large (N=274) transdiagnostic sample of individuals with and without substance-use-related conditions, to predict self-reported craving. CPM is a machine-learning approach used to identify neural ‘signatures’ in functional connectivity data related to a specific phenotype. Functional connectomes were derived from three guided imagery conditions of personalized appetitive, stress, and neutral-relaxing experiences. Craving was rated before and after each imagery condition. CPM successfully predicted craving, thereby identifying a transdiagnostic ‘craving network’ comprised primarily of the posterior cingulate cortex, hippocampus, visual cortex, and primary sensory areas. Findings suggest that craving may be associated with difficulties directing attention away from internal self-related processing, represented in the default mode network.