Perseveration and shifting in obsessive–compulsive disorder as a function of uncertainty, punishment, and serotonergic medication

Abstract

Background The nature of cognitive flexibility deficits in obsessive–compulsive disorder (OCD), historically tested with probabilistic reversal learning tasks, remains elusive. Here, a novel deterministic reversal task and inclusion of an unmedicated patients illuminated the role of fixed versus uncertain rules/contingencies and of serotonergic medication. Additionally, our understanding of probabilistic reversal was enhanced through theoretical computational modeling of cognitive flexibility in OCD. Methods We recruited 49 patients with OCD, of whom 21 were unmedicated, and 43 healthy controls, matched for age, IQ, and gender. Participants were tested on two tasks: a novel visuo-motor deterministic reversal learning task with three reversals (feedback rewarding/punishing/neutral) measuring accuracy/perseveration and a two-choice visual probabilistic reversal learning task with uncertain feedback and a single reversal measuring win stay and lose shift. Bayesian computational modeling provided measures of learning rate, reinforcement sensitivity, and stimulus stickiness. Results Unmedicated OCD patients were impaired at the deterministic reversal task under punishment only at the first and third reversals, compared with both controls and medicated OCD patients, who had no deficit. Perseverative errors were correlated with OCD severity. On the probabilistic reversal task, unmedicated patients were only impaired at reversal, whereas medicated patients were impaired at both the learning and reversal stages. Computational modeling showed that the overall change was reduced feedback sensitivity in both OCD groups. Conclusions Both perseveration and increased shifting can be observed in OCD, depending on test conditions including the predictability of reinforcement. Perseveration was related to clinical severity and remediated by serotonergic medication.