P194 Withdrawal symptoms of electrical brain stimulation in a probabilistic decision making task

Abstract

Question It has been shown that transcranial electrical brain stimulation (TES) can improve many aspects of cognition, including decision making and learning. However, it has not been studied whether brain is capable of adapting itself to perform at least equally well without TES, after initially learning the task under influence of TES. We used a probabilistic learning task to investigate this question. Methods Participants ( n = 10) took part in two groups of active ( n = 5) and sham ( n = 5) transcranial direct current stimulation (tDCS). Each participant attended two experimental sessions. In both sessions participants were asked to perform a probabilistic decision making task. In this task participants adapted to changes in reward contingencies. Participants were presented with two options of which one of them was designated as the better one, leading to higher possibility of rewarding than punishing feedback. Participants were asked to maximise their gain by choosing the better option. The contingencies changed over the course of the trials. Consequently participants had to adjust to the changes in the environment. In the first session, participants in the active and sham tDCS groups, received 15 min and 16 s of anodal tDCS over their left dorsolateral prefrontal cortex, respectively. For all participants sham stimulation was administered in the second session. Number of rewards in both sessions was recorded and their difference was considered for analysis. Results Comparison of difference of acquired rewards between the two sessions showed that participants in the active group tend to perform worse than those in the sham group. Conclusions This result shows that learning under the influence of TES leads to adaptation which induces changes that might not be efficient without TES in a later phase. In more general terms, this result indicates that learning a task under the influence of TES leads to creation of a model which might not be valid anymore without TES.