A computational mechanism linking momentary craving and decision-making in alcohol drinkers and cannabis users

Craving and maladaptive choices are intertwined across addictive disorders, yet the specific computational mechanisms mediating their interactions remain elusive. Here we tested a hypothesis that momentary craving and reinforcement learning influence each other during substance-related decision-making. Two substance-using groups with moderate to high addiction risk levels (alcohol drinkers and cannabis users; total n = 132) performed a decision-making task in which they received a group-specific addictive cue or monetary outcomes and reported moment-to-moment subjective craving. Computational modeling revealed that momentary craving biased substance-specific learning rate in both groups, but in opposite directions. In addition, expected values and outcomes jointly influenced elicited craving across groups and decision contexts. Finally, regressions incorporating model-derived parameters best predicted alcohol, but not cannabis, addiction risk scores, supporting the selective utility of using these model-based parameters in making clinical predictions. Together, these findings provide a computational framework that accounts for the interaction between craving and maladaptive choices across addictive domains. This study investigated the computational mechanisms linking momentary craving and decision-making in people with moderate to high addiction risk levels for alcohol or cannabis use, uncovering different mechanisms of alcohol and cannabis addiction risk and implying the need for substance-specific interventions.