Functional and structural connectivity underlying reinforcement learning in young and older adults

We base many of our choices on previous experiences. Learning from the results of our actions to improve our behavior and increase the resulting reward is called reinforcement learning. In this thesis, I investigated how brain areas communicate the necessity and implementation of such behavioral adjustments, and how these connectivity patterns change with age. The results indicated that in young adults, oscillatory brain activity over the frontal cortex, especially in the theta band (4-8 Hz), signals the need for behavioral adjustment. Inter-site synchronization of oscillatory dynamics (a measure of functional connectivity) seemed to underlie the communication between the frontal learning network and other task-relevant brain areas, to implement this adjustment and optimize future behavior. Brain connectivity patterns that were related to learning in young adults were less predictive of learning success in older adults (65 years and older). Instead, both functional connectivity in especially anterior prefrontal cortex, and structural connectivity between the striatum and anterior and lateral prefrontal cortex, became more important for learning in this age group. Together, the results in this thesis highlight the importance of integration of information in the brain to achieve successful reinforcement learning. Although processing in and connectivity between central parts of the learning network decrease with aging, older adults seem to be able to compensate by utilizing connections with other task-appropriate brain areas. Depending on mental and physical health, intelligence and working memory, and task demands, this can allow older adults to still reach the same performance level as their younger counterparts.