CHOOSE-OCD

BACKGROUND

Patients with obsessive-compulsive disorder (OCD) face intrusive thoughts (obsessions) and repetitive behaviors (compulsions) that provide temporary relief but lead to significant long-term negative consequences. These patients exhibit decision-making difficulties, notably marked indecision and excessive caution, even when advantageous choices are obvious. Reinforcement learning (RL), a relevant theoretical framework for modeling decision-making, suggests that OCD may result from an altered learning policy, leading to an overestimation of positive reinforcement or an underestimation of negative reinforcement. Dysfunctions in the lateral orbitofrontal cortex, involved in credit assignment for actions, confirm impairments in action-outcome contingency learning. Furthermore, OCD is increasingly viewed as a disorder of doubt and uncertainty, with difficulties in updating action values in volatile environments. This project aims to investigate whether OCD patients have difficulty adapting their choices in contexts where action-outcome relationships change.

---

GOING FURTHER WITH THE SUPPORT OF LabEx CerCoG

To this end, 25 OCD patients and 25 healthy controls, matched for age, sex, and education level, will participate in a reversal reinforcement learning task. This task requires participants to learn and revise action-outcome associations in the presence of reversals, and their choices will be analyzed using reinforcement learning (RL) and counterfactual learning (CL) models. In parallel, pupillary dilation will be measured to study arousal related to exploratory versus exploitative choice strategies. The protocol, which has already received ethical approval, is in its pilot phase, with recruitment scheduled to begin in September 2025. The project is coordinated by GIN (Grenoble) with LJK for modeling and the Paris Neuromodulation Institute for patient access.