A Multi-Reward Framework to Improve Translation in Alcohol Use Disorder

Alcohol use disorder (AUD) is a major contributor to global disease burden and a leading cause of preventable death, yet available treatments remain modestly effective. Although rodent models are widely used to study alcohol-related behaviors and guide pharmacotherapy development, many treatments with strong preclinical efficacy fail clinically. Here, we propose a reward-component framework that separates ethanol intake into four components: pre-oral, oral, post-oral peripheral, and post-oral central drug rewards. Synthesizing behavioral, pharmacological, and neurobiological evidence, we find that the component primarily sustaining high ethanol intake in rodents remains unclear, with intake reflecting mixed oral and post-oral rewards. In contrast, human AUD is predominantly sustained by post-oral central drug reward, creating a translational misalignment in pharmacotherapy testing. Pharmacotherapies act across multiple components, and translational failure may arise from misalignment between the components sustaining drinking in rodent models and human AUD. This framework provides a basis for developing more predictive preclinical models and more effective pharmacological interventions.