Differential encoding of losses and gains in the human striatum

Abstract

Studies on human monetary prediction and decision making emphasize the role of the striatum in encoding prediction errors for financial reward. However, less is known about how the brain encodes financial loss. Using Pavlovian conditioning of visual cues to outcomes that simultaneously incorporate the chance of financial reward and loss, we show that striatal activation reflects positively signed prediction errors for both. Furthermore, we show functional segregation within the striatum, with more anterior regions showing relative selectivity for rewards and more posterior regions for losses. These findings mirror the anteroposterior valence-specific gradient reported in rodents and endorse the role of the striatum in aversive motivational learning about financial losses, illustrating functional and anatomical consistencies with primary aversive outcomes such as pain.

Figure 1.

Experimental design. Visual cues were presented for 3.5 s and followed immediately with the outcome depicting the outcome amount, which was displayed for 1.5 s. For the analysis, events were marked at the time of the outcome, and linear contrasts were performed between the different outcome types. p, Pence.

Figure 2.

Behavioral results. a, Preference scores. One-way repeated-measures ANOVA; F _(4,92) = 5.572; p = 0.0005; post hoc two-tailed t test yielded significant differences between univalent reward and neutral and univalent loss and neutral (p < 0.05). b, Mean pupillometry, average across all trials across learning, in a trial-specific manner. We looked for a basic effect of conditioning between the rewarding, aversive, and neutral cue, which is a standard measure of conditioning. Repeated-measures ANOVA revealed a significant effect of trial type. F _(2,19) = 3.342; p < 0.05. The post hoc t tests showed a significant effect (increased amplitude of light reflex) for both rewarding and aversive cues when compared with the neutral cue (p < 0.05). Error bars indicate SEM. p, Pence.

Figure 3.

fMRI simple bivalent, univalent contrasts. a, Aversive prediction error, right ventral striatum [Montreal Neurological Institute (MNI) coordinates (x, y, z): −16, 0, −10; z = 3.74; 46 voxels at p < 0.005]. This contrast also revealed a peak in the right anterior insula [data not shown; MNI coordinates (x, y, z): 30, 18, −12; z = 3.60]. Yellow corresponds to p < 0.005; magenta corresponds to p < 0.001. b, Reward prediction error, right ventral striatum [MNI coordinates (x, y, z): −16, 6, −6; z = 3.38; 28 voxels at p < 0.005]. Yellow corresponds to p < 0.005; magenta corresponds to p < 0.001. c, Sagittal view showing the two peaks, reward (green) and aversive (red).

Figure 4.

fMRI temporal difference (TD) model. a, Left, Aversive TD error, right mid striatum [Montreal Neurological Institute (MNI) coordinates (x, y, z): −20, −4, 6; z = 3.89; p < 0.005; 21 voxels]. Yellow corresponds to p < 0.005; magenta corresponds to p < 0.001. Right, The image is also shown in sagittal section (in red). b, Left, Appetitive TD error, right ventral striatum (nucleus accumbens) [MNI coordinates (x, y, z): 10, 6, −1; z = 3.13; shown at p < 0.005; 15 voxels] and left ventral striatum (nucleus accumbens) [MNI coordinates (x, y, z): −12, 6, −18; z = 3.62; 14 voxels]. Yellow corresponds to p < 0.005; magenta corresponds to p < 0.001. Right, The image is also shown in sagittal section (in green).