Computational and neural correlates of Gain-loss differences in intertemporal choice

Intertemporal choice exhibits systematic differences between gain and loss contexts; however, the computational mechanisms underlying these differences and their neural correlates remain incompletely understood. In the present study, fifty-three healthy adults completed a delay discounting task involving both gain and loss contexts while undergoing functional magnetic resonance imaging (fMRI). At the behavioral level, the proportion of choosing smaller-sooner options in the gain context was significantly higher than the proportion of choosing larger-later options in the loss context. Drift diffusion model (DDM) analyses revealed that gain-loss context was significantly associated with multiple components of decision dynamics, including higher drift rates, stronger starting-point biases, and longer non-decision times in the gain context, whereas no significant contextual differences were observed in decision boundary. At the neural level, intertemporal choices in the gain context were accompanied by increased activation in the bilateral hippocampus and right insula. Together, these findings provide converging computational and neural evidence for gain-loss asymmetries in intertemporal decision making.