Glucose-specific signaling effects on delay discounting in intertemporal choice

We propose that decisions related to resource management (e.g., intertemporal choice between a smaller-and-sooner reward and a larger-and-later reward) are sensitive to and regulated by fluctuating blood glucose levels. Circulating glucose affects intertemporal choice by means of signaling body energy condition instead of serving as a replenishing resource for effortful cognitive processing. We intend to dissociate calorie-supplying functions from glucose-unique anticipatory effects on behavioral resource management, measured by delay discounting in making intertemporal choices. Regarding the anticipatory functions of the glucose–insulin system in regulating the degree of delay discounting, we tested three predictions: First, we predict that the signaling effects of circulating glucose on delay discounting do not need to be dose-dependent as long as glucose fluctuation indicates a directional trend in body energy budget. Second, such effects of glucose fluctuation on delay discounting are phagic (appetite related) instead of dipsian (thirst related). Third, this glucose–insulin signaling system requires glucose as the specific input, thus is insensitive to other forms of sugar that are not insulin regulated. In Study 1, fasting participants were randomly assigned to one of five groups: water consumption, zero-consumption, and three glucose consumption (18g, 36g, and 72g cane sugar/250ml water) groups. The participants competed two sets of intertemporal choice questions with varying delay discounting rates before and after a beverage intervention. The results showed that the rate of delay discounting was negatively correlated to blood glucose levels. The effects of circulating glucose on delay discounting closely followed the changes in blood glucose levels showing a plateau on both dose-response curves (i.e., the sugar dose-blood glucose level curve and the sugar does-delay discounting curve). Secondly, the effects of circulating glucose on delay discounting were significant only in the glucose ingestion group, but not in the zero consumption and the water consumption groups, suggesting that the behavioral effects were in fact related to hunger-reduction instead of thirst-reduction. Study 2 revealed that glucose ingestion, but not water or another form of sugar (xylitol matched to glucose either for sweetness or for calories), reduced delay discounting, making future options more attractive. This result suggests that signaling of body energy budget is indeed glucose-unique. Our results suggest a forecasting mechanism of the glucose-insulin system for both metabolic and behavioral regulations of resource acquisition and allocation.