Reward Learning in Prolonged Grief: Evidence from the Probabilistic Reward Task

Background Deficits in reward processing have been proposed as a mechanism contributing to the persistence of prolonged grief disorder (PGD). The probabilistic reward task (PRT) is a well-validated behavioural paradigm for examining reward learning. In this study, we used the PRT to investigate whether reward learning is affected in bereaved adults with probable PGD compared to bereaved controls. Method Bereaved adults with probable PGD (n = 55) and bereaved controls (n = 56) completed the PRT. Reward learning was assessed using signal detection indices, reaction time (RT) distribution analyses, and hierarchical drift diffusion modelling (HDDM) to quantify latent decision parameters. Associations with prolonged grief severity and anhedonia were examined. Results Participants with probable PGD responded more slowly but did not differ reliably from bereaved controls in response bias, discriminability, accuracy, or HDDM parameters. Grief severity and hedonic capacity were not associated with task performance. Conclusion The findings did not indicate a broad deficit in learning from probabilistic monetary reinforcement among participants with probable PGD. Future research should examine whether reward-related difficulties in prolonged grief are more evident in valuation or motivation, particularly for social and attachment-related rewards.