Distinct neural activations correlate with maximization of reward magnitude versus frequency

Abstract

Choice selection strategies and decision-making are typically investigated using multiple-choice gambling paradigms that require participants to maximize expected value of rewards. However, research shows that performance in such paradigms suffers from individual biases towards the frequency of gains such that users often choose smaller frequent gains over larger rarely occurring gains, also referred to as melioration. To understand the basis of this subjective tradeoff, we used a simple 2-choice reward task paradigm in 186 healthy human adult subjects sampled across the adult lifespan. Cortical source reconstruction of simultaneously recorded electroencephalography suggested distinct neural correlates for maximizing reward magnitude versus frequency. We found that activations in the parahippocampal and entorhinal areas, which are typically linked to memory function, specifically correlated with maximization of reward magnitude. In contrast, maximization of reward frequency was correlated with activations in the lateral orbitofrontal cortices and operculum, typical areas involved in reward processing. These findings reveal distinct neural processes serving reward frequency versus magnitude maximization that can have clinical translational utility to optimize decision-making.

Keywords: entorhinal cortex; melioration; memory; orbitofrontal cortex; reward expected value.

Fig. 1

Reward task and associated behavior. A) Task schematic showing central fixation for 0.5 s followed by 2 choice doors. Post-response, fixation reappeared for 0.5 s, followed by presentation of the chosen door for 0.5 s, then gain or loss trial reward feedback provided for 0.5 s, and finally, cumulative feedback of all gains/losses up to the present trial shown for 0.5 s. Reward distributions for the door choices are presented in Table 1. B) Win-Stay and Lose-Shift difference as seen as a function of the 2 block types and RareL/RareG choices with ^*:P < 0.05, ^**:P < 0.01, ^***:P < 0.005. RareG choices only in the ΔEV block showed significantly greater Win-Stay versus Lose-Shift behavior. C) Reward EV performance (Perf) and reward frequency bias (Bias) measures are significantly correlated. D) Perf but not Bias differ by age categories, specifically middle-aged adults (26–60 years old) had significantly greater EV maximization (Perf) than older adults (>60 years old). The bar graphs present median and MAD error bars of the Perf and Bias variables for the 3 age groups.

Fig. 2

Neural correlates of reward payoff magnitude maximization. A) shows the ERSPs for RareG trials on the ΔEV and Δ₀EV blocks at frontal (F3, F4) and parietal (P3, P4) electrodes, P < 0.05 corrected. ERSPs are baseline corrected and time-locked to choice door presentation at 0–500 ms; immediate reward presented at 500–1,000 ms, and cumulative reward presented at 1,000–1,500 ms. Event-related synchronizations (ERS) is observed in the lower theta/alpha frequency bands and event-related de-synchronizations (ERD) is observed in the beta band. B) shows the ΔEV minus Δ₀EV ERSP difference topography maps, P < 0.05 corrected. C) shows the ΔEV minus Δ₀EV source difference maps, P < 0.05 corrected. The significant predictors of Perf performance: alpha activity in the left parahippocampal cortex during the choice presentation period and beta activity in the right entorhinal cortices during the cumulative reward period are encircled. D) Significant neurobehavioral interactions in the Perf model are shown for beta activity in the right entorhinal cortex during the cumulative reward period; activity in this ROI interacted with age, depression, and inattention.

Fig. 3

Neural correlates of reward gain frequency maximization. A) shows the ERSPs for RareL versus RareG trials on the Δ₀EV block corresponding to the Bias measure at frontal (F3, F4) and parietal (P3, P4) electrodes, P < 0.05 corrected. ERSPs are baseline corrected and time-locked to choice door presentation at 0–500 ms; immediate reward presented at 500–1,000 ms, and cumulative reward presented at 1,000–1,500 ms. B) shows the RareL minus RareG ERSP difference topography maps on the Δ₀EV block, P < 0.05 corrected. C) shows the RareL minus RareG source maps, P < 0.05 corrected. The significant predictors of reward gain frequency Bias: beta activations in the left lateral orbitofrontal cortex and right pars opercularis during the choice presentation period are encircled.