Adolescent Development of Multiple Learning Systems

Adolescence is a time filled with opportunities for making choices that have not been encountered before. How do adolescents learn to make these decisions? Maturation of learning processes coupled with dynamic changes in brain systems for learning must be studied in order to determine the mechanisms that underlie adolescent decision making. Research in adults has found contributions from multiple learning systems for decision making. One such system learns incrementally from feedback and reinforcement, and depends in part on the striatum. Another system, in the hippocampus, encodes episodes and allows for flexible use of learned information when required by novel contexts. Recent research in adults explores how these systems can cooperate and compete to facilitate decision making. Ongoing research into learning and decision making processes over the course of adolescence has also implicated the striatum in learning and decision making, but how the hippocampus and striatum interact for decision making remains unknown. In this dissertation I investigate contributions of multiple learning systems for learning and decision making in adolescence. I leverage what is known about underlying brain systems for learning and decision making in adults, and consider how changes in these same systems over adolescence might contribute to behavioral shifts in adolescence. Specifically, in the studies included here, I show how developmental trajectories for learning can enhance performance in adolescents for some types of learning and not others. In the first study I ask how do the striatal and hippocampal systems contribute to feedback based learning in adolescence? I show that in adolescents, both the hippocampus and the striatum contribute to probabilistic feedback learning, and that this type of learning is better in adolescents than in adults. This response to feedback in the hippocampus was found to relate to memory accuracy for features of feedback events only in adolescents. Pushing the finding of hippocampal activation in adolescents, in the second study I ask how does learned value influence flexible decision making in adolescence? Adolescents did not show reliable transfer of value, but there were individual differences in this tendency. Thus, in the third study, I ask which brain regions account for individual differences in learning and value transfer? I show that variability in connectivity at rest between the hippocampus and the vmPFC related to the tendency to transfer value in adults. Taken together, these results contribute to a growing body of research in adolescent decision making, and extend upon our understanding of the mechanisms for learning and decision making systems, and how they change over development.