We also performed an analysis of our data to confirm that the striatal deactivation was not a physiological artifact (Figure S1C). Strikingly, the only brain region commonly active between the time of incentive presentation (Table S1) and the execution of the motor task (Table S2) was bilaterally encompassing ventral striatum (Table S3). Furthermore, additional whole brain analyses did not reveal any brain regions that were directly correlated with Selleck Vorinostat participants’ parabolic behavioral performance or interactions between incentive level and task difficulty (see Supplemental Information for details,

Figure S1E). These analyses provided us with further evidence of the ventral striatum’s integral role in mediating participants’ http://www.selleckchem.com/products/s-gsk1349572.html responses during performance for incentives. The idiosyncratic pattern of striatal activity we observed (i.e., activation at the time of incentive presentation and deactivation at the time of action) resembles that reported for participants experiencing potential monetary gains and losses (Tom et al., 2007 and Yacubian et al., 2006). Tom et al. (2007) found that ventral striatum was activated by the prospect of gains, and deactivated by the prospect of losses, and that such deactivation was strongly correlated with a behavioral measure of loss aversion.

The findings of Tom et al. (2007), in conjunction with our results, led us to develop a new hypothesis regarding the role of ventral striatum in mediating performance decrements for large incentives: deactivation of ventral striatum during motor action reflects evaluation

of the potential loss (of a presumed gain) that would arise from failure to successfully achieve the task. Essentially, larger incentives are framed as larger potential losses, and as these perceived potential losses increase (in the highest incentive conditions) they are manifested as performance decrements. Because this hypothesis is generated in part from a “reverse-inference” (Poldrack, 2006), we needed to obtain additional evidence in order to provide direct empirical support. Our hypothesis led to the following nearly predictions: (1) striatal deactivation at the time of motor action would predict the extent of individuals’ decrements in behavioral performance; (2) activity in ventral striatum during motor action would relate to an individual’s behavioral loss aversion (i.e., the more loss averse a participant, the greater her ventral striatal deactivation during motor action); and (3) a participant’s degree of behavioral loss aversion would be predictive of her propensity to exhibit performance decrements for large incentives, as well as the level of incentive that resulted in peak performance. To test the first prediction, we examined the extent to which a participant’s decrease in performance at the highest incentive level was related to her neural sensitivity to incentive.