Camara, E., Rodriguez-Fornells, A., & Münte, T. F. (2009). Functional connectivity of reward processing in the brain. Frontiers in human neuroscience, 2, 19.
Goal: The authors sought to address discrepancies in networks of brain regions recruited during the processing of gains and losses.
Methods: They administered a monetary reward gambling fMRI task with added special trials: unexpectedly large gains/losses and same magnitude unexpected stimuli change to adults in order to capture response to uncertainty as well as to the gains and losses. They conducted a univariate analysis to look at the overall pattern of activation for gains v losses. Importantly, they also utilized beta series correlations to examine whole brain functional connectivity with the ventral striatum with regard to specific task-based events.
Results:
- Univariate Results: Similar fronto-parietal-subcortical networks were activated during both the boosted monetary gain and loss trials. This network included: cingulate cortex, superior frontal cortex, inferior parietal lobe, insula, parahippocampal regions, thalamus, caudate, cerebellum, cuneus, and ventral striatum. Importantly, monetary gains showed greater activation than losses.
- Functional Connectivity Results: Similar patterns of functional connectivity with the ventral striatum were observed between the standard gain/loss trials and the boosted gain/loss trials. This network included: hippocampus, insula, and orbitofrontal cortex (OFC). Comparing functional connectivity with the ventral striatum during loss > gain trials, significant differences were found in the medial orbitofrontal cortex with increases in BOLD signal during gains and decreases during losses.
Discussion:
- The authors emphasize the importance of leveraging functional connectivity analyses in addition to standard univariate approaches as different results were found in this study between the two methodologies.
- They mention proposed functional loops between the frontal cortex and basal ganglia as well as between the hippocampus and midbrain related to reward processing.
- They emphasize the role of the insula as connecting many regions commonly implicated in reward and decision making: OFC, ACC, NAcc, amygdala.
- They emphasize the role of the medial OFC in differentially processing gains (increased activation) and losses (decreased activation).
Thoughts:
- I think the authors are overselling a non-significant finding that aligns with some previous literature (the amygdala is more functionally connected to the ventral striatum after losses). That is interesting to note, but should not be one of their main findings.
- The authors caution that there is evidence for both excitatory and inhibitory projections within the ventral striatum and between the ventral striatum and other structures, so that BOLD activity must be interpreted carefully. Importantly, this idea translates beyond the ventral striatum and should be kept in mind for all results. This also underscores the importance of primate work in determining directionality of relationships and underlying neurotransmitter structure. However, GIMME can achieve this as well.
- The regions found here are consistent throughout the reward literature in adults. I am curious how reward networks differ in developmental populations.