Investigating Resting State Functional Connectivity in Cannabis Use Disorder Individuals Using Human Connectome Project Data

Abstract

PURPOSE: The triple-network model of psychopathology theorizes psychological disorders manifest from aberrant functional connectivity in three major brain networks, the central executive network, default mode network, and salience network. To date, no research has used this framework to investigate resting-state functional connectivity in cannabis users using fMRI technology. Since attentional deficits have been associated with cannabis use, the dorsal attention network was an additional network of interest. METHODS: Using Human Connectome Project Young Adults (HCP-YA) data, 34 CUD individuals were matched to 34 controls using propensity score matching, resulting in 68 participants with brain data (Mage: 27.2, 17.6% female). Functional connectivity was assessed using CONN, a MATLAB-based extension package for Statistical Parameter Mapping 12 (SPM12) dedicated to assessing and displaying functional connectivity using fMRI technology. A seed-to-voxel technique was used, with a priori regions of interest (ROI) derived from CONN’s library of ROI. Significant clusters exceeded a voxel threshold of p < .0001 FDR- correction, and p < .001 cluster threshold. RESULTS: CUD individuals had aberrant functional connectivity across the default mode network, salience network, and dorsal attention network. A consistent finding across networks was weaker anti-correlation with the occipital cortex. Both the default mode network and dorsal attention network exhibited weaker positive connectivity with surrounding areas and pre-/post- central gyrus. The salience network in CUD individuals uniquely exhibited greater connectivity, with greater positive connectivity between the right supramarginal gyrus and left inferior frontal gyrus / precentral gyrus / central opercular cortex, and greater anti-correlation between the left insula and right postcentral gyrus. CONCLUSION: The triple-network approach to CUD revealed systemic differences across networks; but may not be the best model for understanding CUD biomarkers. The results highlight functional brain connectivity with problematic cannabis usage.