Computational Phenotyping of Effort-Based Decision Making in Unmedicated Adults With Remitted Depression

Reduced motivation is a core feature of major depressive disorder (MDD). Yet, the extent to which this deficit persists in remitted MDD (rMDD) remains unclear. Here, we examined effort-based decision making as one aspect of amotivation in rMDD using computational phenotyping to characterize decision-making processes and strategies.

Unmedicated adults with rMDD (n = 40) and healthy control (HC) participants (n = 68) completed the Effort Expenditure for Rewards Task. Repeated-measures analysis of variance and computational modeling-including hierarchical drift diffusion modeling and subjective value modeling-were applied to quantify decision-making dynamics in effort allocation across different reward magnitudes and probabilities.

Relative to HC participants, participants with rMDD made overall fewer hard task choices, with an attenuated effect when accounting for anhedonia. However, specific to high reward, high probability conditions, participants with rMDD chose to expend effort more often than HC participants. This was supported by the drift diffusion model results revealing that participants with rMDD showed a drift rate biased toward selecting the easy task, counteracted by heightened influence of reward probability and magnitude. Probed with the subjective value model, this was not driven by group differences in decision strategies with respect to magnitude and probability information use.

Collectively, these findings suggest that while individuals with rMDD exhibit persistent motivational deficits, they retain a heightened sensitivity to high-value rewards, requiring more substantial or certain rewards to engage in effortful tasks. This pattern may reflect impairments in reward processing and effort-cost computations, contributing to motivational dysfunction. Targeting reward sensitivity and effort allocation could be valuable for interventions aimed at preventing MDD relapse.