Back

Static and dynamic noise abolish altruistic biases in Parkinson's disease independent of deep brain stimulation

Hubschmid, F.; Sellitto, M.; Bahners, B. H.; Pirenne, T.; Kalenscher, T.; schnitzler, A.; van Wingerden, M.; Florin, E.

2026-10-11 neurology
10.64898/2026.10.08.26365079 medRxiv
Show abstract

There are known alterations in decision-making attributed to Parkinson's disease. Yet little is known about how those alterations affect decisions in social contexts. We aimed to assess social decision-making in healthy human volunteers as well as people with Parkinson's disease treated with deep brain stimulation of the subthalamic nucleus. Twenty healthy controls and twenty-two people with Parkinson's in both the stimulation ON and OFF conditions took part in a social discounting task that measured the degree of generosity toward others as a function of social-emotional distance. Using a computational approach, we modelled mechanisms of social discounting across three different models that differed in how they mapped choice valuation onto observed decisions. People with Parkinson's less often made generous choices than controls. Across all analyses, we found converging evidence that this effect was attributable to increased static and dynamic decision noise, i.e. inhibitory and attentional processes, but not changes in prosocial valuation. While stimulation sped up decisions, there was no effect of deep brain stimulation on social decision-making. Present results indicate that people with Parkinson's are more prone to random processes such as increased impulsivity and attentional lapses during choices in social situations. Such changes attributed to the disease appear to evade clinical care with dopaminergic compensation and deep brain stimulation.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.