Factors underlying probabilistic and deterministic stimulus-response learning performance in medicated and unmedicated patients with Parkinson’s disease.

Abstract

Prior studies have not tested individual differences or effects of medication dosage on stimulus-response learning in patients with Parkinson's disease (PD). In the current study, we investigated the effects of motor variables (including tremor, akinesia, and disease duration) as well as dopaminergic medication dosage on learning in unmedicated PD patients, medicated PD patients, and healthy controls. We tested all subjects on probabilistic and deterministic learning tasks, and also collected awareness measures data using postexperimental questionnaires. Importantly, we tested learning performance in tremor-dominant and akinesia-dominant PD patients, and further correlated learning performance with disease duration and medication dosage. Results show that akinesia-dominant patients were more impaired than tremor-dominant patients at probabilistic reward- but not punishment-based learning, which is in agreement with prior studies of the relationship between akinesia and basal ganglia dysfunction. We also found no difference between medicated and unmedicated PD patients in reward- or punishment-based deterministic learning, but medicated patients were better than unmedicated patients at reward-based probabilistic learning. Our results show that awareness measures explain differences among probabilistic and deterministic learning performance. Moreover, we found that disease duration and motor severity are inversely correlated, and medication dosage is positively correlated, with reward-based probabilistic learning. Our results suggest that stimulus-response learning performance in patients with PD depends on the type of learning (probabilistic vs. deterministic), medication status (on vs. off medication, dopaminergic medication dosage), disease duration as well as motor severity and subtype in PD patients (tremor- vs. akinesia-dominant).