Motivation and reward learning : neural mechanisms, changes in depression and following pharmacological treatment

• Main Author: Scholl, Jacqueline
• Other Authors: Harmer, Catherine ; Rushworth, Matthew
• Published: University of Oxford 2015
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.730593

The primary goal of this thesis was the understanding of the cognitive and computational mechanisms of reward and effort guided processing and their change in depression and commonly used/putative treatments. In chapter 2, I developed a new paradigm that allowed me to examine the neural basis of concurrent learning about rewards and effort in a complex, naturalistic environment. Measuring brain activity using functional magnetic resonance imaging, I proposed a prefrontal cortical mechanism for overcoming behavioural biases generated by irrelevant reward features (Scholl et al., 2015). In chapter 3, I used this paradigm to measure the effects of prolonged administration of a commonly used antidepressant, a serotonin re-uptake inhibitor, which has previously been hypothesized to enhance learning and neural plasticity. I found that it increased neural learning signals for both reward and effort in a widespread network of brain areas, suggesting a general learning enhancing effect (Scholl et al., under review). In chapter 4, I measured whether a putative novel treatment for depression, the partial NMDA agonist d-cycloserine, could also influence learning. While I found this not to be the case, it led to a marked improvement in how participants integrated the available information to make decisions (Scholl et al., 2014). In chapter 5, I tested how these different aspects of reward-guided behaviour, some of which were changed in my pharmacological manipulation, are related to changes in depression. We found that dysphoric participants did not differ from healthy controls in how quickly they learnt appetitive or aversive/effortful information. However, when making decisions they used this learnt information less than healthy controls to determine their decisions. Together, these studies revealed that complex decisions require additional cognitive and neural mechanisms that go beyond those seen in the simplest learning tasks. These diverse mechanisms can be computationally teased apart, rely on different neural systems and can thus be influenced separately. Lastly, we found that common treatments of depression did not necessarily target the same mechanisms affected by depression, suggesting potential new avenues for new treatments.