Longitudinal changes in network engagement during cognitive control in cocaine use disorder

Abstract Background Cocaine use disorder (CUD) is characterized by poor cognitive control and has limited empirically supported treatment options. Furthermore, an understanding of brain mechanisms underlying CUD is at a relatively early stage. Thus, this study aimed to investigate longitudinal alterations in functional neural networks associated with cognitive control in cocaine use disorder (CUD). Methods Secondary analysis was performed on data from 44 individuals who participated in three randomized clinical trials for CUD and completed an fMRI Stroop task both at baseline and post-treatment. Independent component analysis (ICA) was performed to assess changes in functional network engagement and investigate associations with cocaine-use behaviors. Mixed linear models were performed to test for longitudinal effects on network engagement and relationships with baseline patterns of cocaine use (i.e., past-month frequency and lifetime years of use) and periods of abstinence/use between scans (i.e., percent negative urine toxicology and maximum days of contiguous abstinence). Results Six functional networks were identified as being related to cognitive control and/or exhibiting changes in engagement following treatment. Results indicated that engagement of amygdala-striatal, middle frontal and right-frontoparietal networks were reduced over time in CUD. Less change in the amygdala-striatal network was associated with greater lifetime years of cocaine use. Additional analyses revealed that negative toxicology results and achievement of continuous abstinence were associated with greater engagement of the right-frontoparietal network. Conclusions Neural systems that underlie cognitive control may change over time in individuals with CUD. A longer history of cocaine-use may hinder changes in network activity, potentially impeding recovery.