Dopaminergic modulation of neural mechanisms underlying dynamical cost-benefit valuations and (dysfunctional) motivational processes
Final Report Abstract
In project 1, we used a placebo-controlled, within-subjects pharmacological study design and behavioural testing to assess the role of the neurotransmitter dopamine in dynamic effort-reward integration during action. We show that boosting dopamine function with 150mg levodopa increases an implicit dissociation between high and low rewards, reflected in enhanced motor vigour for high rewards. Conversely, blocking dopamine with 1.5mg haloperidol attenuates a dissociation between reward levels, reflected in reduced strategic effort-reward integration. The latter also affected subjective but not objective fatigue, an effect most pronounced in subjects with low trait motivation. Our findings show that modulating a homeostatic dopamine balance distinctly alters implicit and explicit effort allocation to maximize reward during action, identifying dopamine system integrity as potential neurobiological mediation of a vulnerability to disorders of motivation. In project 2, we used functional neuroimaging, computational modelling of behaviour and eye-tracking to assess how the brain processes uncertainty and predictiveness, with a particular focus on anxiety. In the first study, we used functional MRI and computational modelling to assess how humans make inference under certainty. We demonstrate that activity in hippocampus increases when subjects rely more on previously acquired internal representations compared to upcoming novel evidence. Moreover, response in superior parietal cortex reflected a precision-weighted prediction error signal (i.e., the distance between observations and expectations) that was modulated by hippocampal activity. Together with evidence from domains such as memory, planning, and self-projection, our results support a view that a critical role of hippocampus is to reflect the relevance of acquired internal representations compared to upcoming novel evidence. In the second stud, we used functional MRI to show that separate subregions of the human amygdala distinctly contribute to aversive experience. We demonstrate that activity within basolateral amygdala (BLA) scales with increasing levels of threat, which translates into heightened subjective anxiety. Trait-anxious individuals are characterised by elevated and undifferentiated threat signals in BLA, identifying this subregion as potential neurobiological mediator of a vulnerability to anxiety disorders. In the third study, we used computational modelling and eye-tracking to assess how humans allocate attention to learn efficiently in aversive and uncertain environments. Our results show that attention is influenced by aversive value (the likelihood of an unpleasant event) but not by uncertainty (akin to confidence in its probability). Moreover, this relationship is bidirectional such that attention biases value updates for attended stimuli, leading to a subjective overestimation of threat. Our findings shed light on the computational mechanism underlying biased attention in psychopathology and support a role for learning in the expression of threat-related attentional biases in anxiety disorders.
Publications
- (2019). The role of the hippocampus in weighting expectations during inference under uncertainty. Cortex. 115:1-14
Rigoli F, Michely J, Friston KJ, Dolan RJ
(See online at https://doi.org/10.1016/j.cortex.2019.01.005)
