Impaired outcome-based decision-making in human cocaine addiction

Despite rising global rates for cocaine addiction, treatments options are limited. Psychological treatments remain the standard option but have only modest effectiveness. One major barrier to treatment success is that patients often make disadvantageous decisions that go against recovery, highlighting the need for strategies to support decision-making. Converging evidence indicates that decision-making deficits in cocaine addiction are driven by impaired processing of outcomes and risks. We investigated how patients with cocaine addiction use explicit outcomes and risk information to guide decision-making. One-hundred-and-five patients with moderate-to-severe cocaine use disorder (CUD) and 80 healthy control volunteers completed a decision-making task in which reward, loss, and probability information were provided explicitly. A subset of participants underwent diffusion-weighted magnetic resonance imaging to investigate the orbitofronto-striatal white matter tracts that underpin decision-making. We used generalized linear mixed-effects models to determine how rewards, losses, and probability influence decisions, and combined probabilistic tractography with principal component analysis to characterize the white matter tracts. We found that patients incorporated explicit probability information into their decisions to the same degree as control participants but showed reduced integration of outcomes. Notably, patients were less likely to integrate losses into their decisions than rewards. Although orbitofronto-striatal white matter tracts did not differ between groups, individual variability predicted the extent to which reward and loss information guided decisions, irrespective of diagnosis. Overall, these findings reveal fundamental challenges in using outcomes, particularly negative ones, to guide decision-making in patients with CUD, emphasizing the need for additional support to promote adaptive choices during recovery.