TY - JOUR T1 - When Less is More: Enhanced Statistical Learning After Disruption of Bilateral DLPFC JF - bioRxiv DO - 10.1101/198515 SP - 198515 AU - Géza Gergely Ambrus AU - Karolina Janacsek AU - Anna B. C. Trimborn AU - Gyula Kovács AU - Dezso Nemeth Y1 - 2017/01/01 UR - http://biorxiv.org/content/early/2017/10/04/198515.abstract N2 - Human learning depends on multiple cognitive systems related to dissociable brain structures. These systems interact not only in cooperative but also in competitive ways in optimizing performance. Previous studies have shown that manipulations reducing the engagement of frontal lobe-mediated control and attentional processes can lead to improved performance in striatum-related learning processes. These studies, however, used correlational methods only. The aim of the present study was to directly test the causal role of the prefrontal cortex (PFC) in striatum-mediated implicit statistical learning and its consolidation using brain stimulation. Healthy young adults were trained on a probabilistic sequence learning task. 1 Hz transcranial magnetic stimulation (TMS) or sham stimulation of both the left and right dorsolateral PFCs (DLPFC) was applied intermittently during the learning session to disrupt frontal lobe functions. To assess the lasting effects of TMS on learning and consolidation, statistical learning performance (i.e., increased reaction times for sequences with high as compared to low probabilities) was tested ten minutes, two hours, and 24 hours later. In line with the predictions of a competitive relationship between DLPFC functions and statistical learning, the DLPFC stimulation group showed better performance compared to the sham group after the 24-hour consolidation period. This finding suggests that the disruption of DLPFC induced qualitative changes in picking up statistical regularities during learning that became salient in behavior after a stabilization period. Our results support an antagonistic relationship between the brain networks of automatic and controlled processes.Significance statement Here we demonstrate the role of the dorsolateral prefrontal cortex (DLPFC) in statistical learning, which is a fundamental learning mechanism of the brain, using non-invasive brain stimulation. Participants were trained on a probabilistic sequence learning task while applying transcranial magnetic stimulation on both the left and right DLPFC in order to disrupt frontal lobe functions. This is the first time when bilateral TMS was applied in a cognitive neuroscience study. According to the results, the DLPFC disruption led to long-term enhanced performance providing causal evidence for the models of competition between DLPFC functions and procedural learning. Thus, our work supports the view that an antagonistic relationship exists between the brain networks of automatic and controlled processes. ER -