Individual Differences in Dopamine Are Associated with Reward Discounting in Clinical Groups But Not in Healthy Adults

Abstract

Some people are more willing to make immediate, risky, or costly reward-focused choices than others, which has been hypothesized to be associated with individual differences in dopamine (DA) function. In two studies using PET imaging, one empirical (Study 1: N = 144 males and females across 3 samples) and one meta-analytic (Study 2: N = 307 across 12 samples), we sought to characterize associations between individual differences in DA and time, probability, and physical effort discounting in human adults. Study 1 demonstrated that individual differences in DA D2-like receptors were not associated with time or probability discounting of monetary rewards in healthy humans, and associations with physical effort discounting were inconsistent across adults of different ages. Meta-analytic results for temporal discounting corroborated our empirical finding for minimal effect of DA measures on discounting in healthy individuals but suggested that associations between individual differences in DA and reward discounting depend on clinical features. Addictions were characterized by negative correlations between DA and discounting, but other clinical conditions, such as Parkinson's disease, obesity, and attention-deficit/hyperactivity disorder, were characterized by positive correlations between DA and discounting. Together, the results suggest that trait differences in discounting in healthy adults do not appear to be strongly associated with individual differences in D2-like receptors. The difference in meta-analytic correlation effects between healthy controls and individuals with psychopathology suggests that individual difference findings related to DA and reward discounting in clinical samples may not be reliably generalized to healthy controls, and vice versa.SIGNIFICANCE STATEMENT Decisions to forgo large rewards for smaller ones due to increasing time delays, uncertainty, or physical effort have been linked to differences in dopamine (DA) function, which is disrupted in some forms of psychopathology. It remains unclear whether alterations in DA function associated with psychopathology also extend to explaining associations between DA function and decision making in healthy individuals. We show that individual differences in DA D2 receptor availability are not consistently related to monetary discounting of time, probability, or physical effort in healthy individuals across a broad age range. By contrast, we suggest that psychopathology accounts for observed inconsistencies in the relationship between measures of DA function and reward discounting behavior.

Keywords: PET; decision making; delay discounting; dopamine; effort; probability.

Figure 1.

Average DA D2R availability. Average voxelwise whole-brain binding potential for (A) Sample 1 collected using [18F]fallypride in young adults, (B) Sample 2 collected using [18F]fallypride across the adult life span, and (C) Sample 3 collected using [11C]FLB 457 across the adult life span. Sagittal images use the cortical BP_ND color scale. Axial images use the striatal BP_ND color scale. The differences in binding potential between cortical and striatal regions depend on the radiotracer and mean age of the sample.

Figure 2.

Correlations between reward discounting and D2R availability. Correlation plots represent associations between D2R availability (PVC) and proportion of smaller sooner/higher probability/less effortful choices. Individual subject data points are depicted for time (turquoise), probability (pink), and effort (green). Solid lines indicate regression slopes for [18F]fallypride. Dotted lines indicate regression slopes for [11C]FLB 457.

Figure 3.

Meta-analytic comparison of associations between individual differences in DA and reward discounting. Left, Forest plot represents variation in effect sizes according to clinical status (healthy, addiction, and other psychopathology). Values represent correlation coefficients, r, for display purposes; positive values indicate a positive correlation between DA function and greater discounting (e.g., more immediate choices). Black diamonds represent individual study effects (diamond size depicts the weight in the meta-analysis). Horizontal lines indicate 95% CIs of the individual effects, noted on the right). Gray diamonds represent 95% CIs of the factor coefficients from the clinical status term. Bottom right, Funnel plot. Plotted points indicate individual effects. Points indicate the residuals of the psychopathology groups and their associated study precision (SE). When the effect residuals lie within the unshaded area, it implies that heterogeneity in the main effect is successfully accounted for by the interaction model. Points within the unshaded region correspond to p values > 0.10. p values in the light gray and dark gray regions correspond to p values between 0.10 and 0.05 and between 0.05 and 0.01, respectively.