Large-N Rat Data Enables Phenotyping of Risky Decision-Making: A Retrospective Analysis of Brain Injury on the Rodent Gambling Task

Abstract

Decision-making is substantially altered after brain injuries. Patients and rats with brain injury are more likely to make suboptimal, and sometimes risky choices. Such changes in decision-making may arise from alterations in how sensitive individuals are to outcomes. To assess this, we compiled and harmonized a large dataset from four studies of TBI, each of which evaluated behavior on the Rodent Gambling Task (RGT). We then determined whether the following were altered: (1) sensitivity to overall contingencies, (2) sensitivity to immediate outcomes, or (3) general choice phenotypes. Overall sensitivity was evaluated using the matching law, immediate sensitivity by looking at the probability of switching choices given a win or loss, and choice phenotypes by k-means clustering. We found significant reductions in sensitivity to the overall outcomes and a bias toward riskier alternatives in TBI rats. However, the substantial individual variability led to poor overall fits in matching analyses. We also found that TBI caused a significant reduction in the tendency to repeatedly choose a given option, but no difference in win- or loss-specific sensitivity. Finally, clustering revealed 5 distinct decision-making phenotypes and TBI reduced membership in the "optimal" type. The current findings support a hypothesis that TBI reduces sensitivity to contingencies. However, in the case of tasks such as the RGT, this is not a simple shift to indiscriminate or less discriminate responding. Rather, TBI rats are more likely to develop suboptimal preferences and frequently switch choices. Treatments will have to consider how this behavior might be corrected.

Keywords: Iowa Gambling Task (IGT); controlled cortical impact (CCI); impulsivity; rat; statistical approaches.

FIGURE 1

Overall choice of the four RGT options shown as mean (solid line) and one standard deviation band (fill). Only sessions after 10 are shown to maximize visualization of stable choice preference, which represents approximately weeks 4–8 post-injury. There were significant Injury effects on each choice option (P1 and P2: p < 0.001; P3: p = 0.048; P4: p = 0.022) and P1 had a slight, but significant decline over time (p = 0.001). Analyses indicated substantial individual variation in choice.

FIGURE 2

Individual subject fits to the matching law show Sham (left, black) and TBI (right, red) subjects. A wide degree of sensitivity was present across subjects. The matching law only described a very small subset of animals.

FIGURE 3

Aggregate matching law parameters and group-level fits. (A) TBI reduced sensitivity (p < 0.001), and (B) shifted bias toward suboptimal options (p < 0.001). (C) The matching law described TBI rats more poorly than Sham (p < 0.001), however the matching law did not describe data well at the individual or (D,E) at the aggregate level.

FIGURE 4

Density plots of distributions of the probability of staying with a choice in TBI vs. Sham animals. (A) Overall tendency to stay on a choice (regardless of outcome) was higher in Sham than TBI rats (p = 0.008). Sham rats displayed a bimodal set of peaks around 40 and 90% likelihood of staying on an outcome. (B) Breakdown of tendency to stay depending on whether the prior outcome was a win (solid) or loss (dashed). The same overall differences were present in TBI (p = 0.008), and losses reduced the probability of staying (leftward shift in curve; p = 0.016) but TBI rats did not show differential sensitivity to wins or losses (p = 0.966). (C) When broken down to each choice option, TBI rats were more likely to stay with P1 (p = 0.001), but less likely to stay with P2 (p < 0.001) regardless of outcome. (D) When choice option data were analyzed depending on whether prior outcome was a win (solid) or loss (dashed), similar overall effects in tendency to stay were observed with TBI rats more likely to stay with P1 (p < 0.001) and P4 (p = 0.021), but less likely for P2 (p < 0.001). There were no differential effects in sensitivity to wins vs. losses (p’s > 0.134).

FIGURE 5

Individual subject’s probability of staying with a choice across total trials, winning trials, and losing trials in Sham (left, black) and TBI (right, red) subjects. There was considerable variability in the tendency to stay with a choice across subjects.

FIGURE 6

Normalized preference (z score) for a given option plotted against each other, with clusters superimposed on top. Points represent individual rats. An Optimal phenotype (green) can be seen as rats that have high P2 values and relatively low values of all others. An Exploratory phenotype (blue) can be seen with values across all options around 0 (average). Two risky phenotypes can be seen, one which highly prefers P3 (light red), and another which prefers P4 (dark red). Finally, a small cluster of indeterminate rats (yellow) can be seen with unique preference for P1.

FIGURE 7

Phenotypes broken down by Injury. (A) TBI rats were significantly less likely to be in the Optimal phenotype and instead were increased across the other phenotypes. Only TBI rats were classified into the Indeterminate phenotype. (B) TBI was significantly different than Sham (p = 0.018) in the Optimal phenotype, with lower P2 and higher P4 choice. (C) TBI was not significantly different than Sham but had high variability in the Exploratory phenotype. (D) TBI was not significantly different than Sham but had high variance in the Risky (P4) phenotype as well. (E) TBI was significantly different than Sham (p = 0.001), with higher P1 and lower P3 preference in the Risky (P3) phenotype. (F) Only TBI animals were present in the Indeterminate phenotype.