Neural sensitivity to incentive was defined as the slope of the relationship between BOLD percent signal change

and incentive level; a positive neural sensitivity corresponded to neural activation, whereas a negative activity was indicative of deactivation. In keeping with the first prediction, we found significant correlations between levels of striatal deactivation at the time of the motor task and performance decrements at the $100 incentive level (Figure 4B; r = 0.70; p = 0.001). Critically, no significant relationship between neural sensitivity and performance was found at the time of incentive presentation (r = 0.22; p = 0.38). Using LGK-974 cost a cross-product term in a multiple regression model, we also found a significant interaction between neural sensitivity during incentive presentation and the motor task and performance (statistics for interaction term: t(14) = 4.18; p = 0.001). To test the second prediction we recalled a subset of participants

(n = 12) who originally participated in these experiments and tested them on a behavioral loss aversion task. This task was the same as that used by Tom et al. (2007), and allowed us to determine a measure λ, indicating how heavily participants weighed losses compared to gains. This subset of participants was found to have a median λ estimate of 2.09 (interquartile range [IQR] 1.09). These values of λ are similar to those ERK inhibitor screening library reported in previous studies (Bateman et al., 2005, Gachter et al., 2007, Tom et al., 2007 and Tverskey and Kahneman,

1992). We found significant correlations between increasing behavioral loss aversion and striatal deactivation during motor action (Figure 5A; r = 0.60; p = 0.04; Figure S3). Importantly, we did not find a significant correlation between neural sensitivity during incentive until presentation and participants’ behavioral loss aversion (r = 0.30; p = 0.34). We also found a significant interaction between neural sensitivity during incentive presentation and the motor task and loss aversion (statistics for interaction term: t(8) = 2.40 p = 0.05). These results illustrate that differences in behavioral loss aversion were indicative of neural responses during motor action. To test the third prediction, and to reach an adequate sample size to test behavioral correlations, we included an additional 20 participants who performed the motor task, the behavioral loss aversion task, and a risk aversion task outside the fMRI scanner. A group comprised of both the subset of imaging participants (n = 12), and the additional participants (n = 20) had a median λ estimate of 2.10 (IQR 0.85). We found a highly significant (r = 0.53; p = 0.002) relationship between increasing behavioral loss aversion and the proclivity to show performance decrements in the hard difficulty level ( Figure 5B), but not in the easy difficulty level (r = 0.22; p = 0.23). We also found a significant relationship (r = 0.