A hallmark symptom of Substance Use Disorders (SUDs) is persistent choice for drugs at the expense of other, more adaptive rewards, yet little is known about the neurobehavioral factors guiding this key behavioral pathology. Building on our preliminary studies, this application seeks to bridge the gap between decision science and addiction neuroscience with the overall aim of identifying how choices for drugs and other rewards in the `addicted' brain differ from those made by drug users who are more readily able to control their use. Given the ever increasing prevalence of cannabis use, we will focus on cannabis smokers (?4x/month), recruiting large samples of individuals with (n=60) and without (n=60) Cannabis Use Disorder (CUD). Participants will undergo a 3-day inpatient protocol comprising fMRI and behavioral choice tasks and laboratory cannabis self-administration. We will investigate two key processes of decision-making during choices for cannabis and an alternative reinforcer (opportunities to play a game of chance): 1) neural encoding of subjective value (SV) signals during choices; and 2) reinforcement learning processes guiding choices. In the healthy brain, valuation circuitry (e.g. ventromedial Prefrontal Cortex) has reliably been shown to quantitatively represent the reward values of available options, calculated in a `common currency' during choice. No research has disambiguated SV encoding during decisions about drugs and other rewards in human drug users, despite the overvaluation of drugs and undervaluation of alternatives thought to contribute critically to the development and manifestation of SUDs. Based on preliminary data, we anticipate that individuals with CUD will show enhanced SV signaling for cannabis and blunted SV signals for the alternative reward compared to non-CUD participants. Our investigation of reinforcement learning will assess the relative contribution of model-based and model-free learning systems, thought to underlie more deliberative and more habitual choice behavior respectively, to decisions about cannabis and the alternative reward. Model-based learning entails use of a cognitive model of associations between choices and outcomes, and facilitates deliberative choice. Model-free learning uses only the reward history to guide choices, and is a feature of habitual decisions. We expect that decisions made for drug reinforcement will recruit more model-free learning in CUD participants, contributing to persistent maladaptive choices for drugs over other rewards. To assess the clinical relevance of these decision processes, we will also investigate associations between SV signals, reinforcement learning, and functional impairment measured by self- and peer-report. Thus, by delineating the decision processes differentiating more controlled drug use from more entrenched, maladaptive patterns of use, this project will address a central challenge for addiction neuroscience: understanding the neurobehavioral mechanisms guiding the expression of a cardinal symptom of SUDs.
Despite substantial recent gains in scientific understanding about the neurobiology of choice in the healthy brain, little is known about the neural processes underlying decisions to use drugs in people with problematic drug use. This project, focusing on cannabis smokers, will investigate the brain and behavioral mechanisms of choices for cannabis and an alternative reward in cannabis smokers with and without Cannabis Use Disorder. Results of this research will contribute to understanding about the brain processes underlying addictions, information that may lead to the development of more targeted treatments.