In search of reward: Are prediction errors metabolically scaled in women and men?

发现新的食物来源是任何觅食动物的关键挑战之一。虽然代谢状态塑造动机是众所周知的，因为饥饿的动物更渴望获得奖励，但饥饿在人类奖励学习中的作用在很大程度上是难以捉摸的。同样，尽管肥胖者奖励学习改变的证据不断出现，但其假定机制在很大程度上仍是未知的，文献中的显著不一致阻碍了统一理论的进展。在这里，我们建议通过对女性和男性的多巴胺依赖性奖励学习作为代谢状态（“深度表型”）的函数进行广泛评估来解决这一差距。为此，我们将使用新开发的智能手机游戏，在一个月内（即女性的一个月经周期）反复评估80名参与者的食物和金钱奖励强化学习。同时，我们将通过连续葡萄糖监测记录葡萄糖水平，以捕获血糖状态的调节。我们将通过两个神经成像会议来补充这些详细的记录，在那里我们操纵代谢状态，并使用fMRI评估奖励预测错误信号的变化。基于初步证据和临床前研究，我们预测饥饿状态与更高的学习率和更快地减少新动作突发事件的预测错误相关。此外，我们预测，这种学习中的“饥饿边缘”在超重/肥胖（代谢状态× BMI相互作用）中会变钝。重要的是，我们将通过检查奖励信号在月经周期中代谢的程度来利用女性荷尔蒙和代谢波动的放大。我们预计，如果能量代谢的变化与奖励学习的变化有关，那么雌二醇和孕酮的更大波动也会在奖励学习中得到回应。总的来说，拟议的项目将揭示一个关键机制，通过多巴胺传输将学习和动机与代谢需求联系起来。进一步深入了解由于能量代谢的变化而导致的奖励学习的重要差异，可能有助于改善患有以奖励功能改变为特征的疾病（如肥胖症）的女性和男性的治疗。通过详细说明潜在的性别和周期对奖励信号代谢比例的影响，拟议的项目也可能为饮食或情绪障碍的病因学提供新的见解。

Discovering novel sources of food is one of the key challenges for any foraging animal. Although it is well-established that metabolic states shape motivation as hungry animals are more eager to attain rewards, the role of hunger in human reward learning is largely elusive. Likewise, despite emerging evidence on altered reward learning in obesity, the putative mechanisms are still largely unknown, and striking inconsistencies in the literature hamper progress towards a unified theory. Here, we propose to address this gap by conducting an extensive assessment of dopamine-dependent reward learning as a function of metabolic states (“deep phenotyping”) in women and men. To this end, using a newly developed smartphone game, we will repeatedly assess reinforcement learning for food and monetary rewards in 80 participants over one month (i.e. one menstrual cycle in women). Concurrently, we will record glucose levels via continuous glucose monitoring to capture modulations by glycemic states. We will complement these detailed recordings by two neuroimaging sessions, where we manipulate metabolic states and assess changes in reward prediction error signals using fMRI. Based on preliminary evidence and preclinical research, we predict that a hungry state is associated with higher learning rates and quicker decreases of prediction errors for new action contingencies. Moreover, we predict that this “hunger edge” in learning is blunted in overweight/obesity (Metabolic State × BMI interaction). Critically, we will capitalize on amplified hormonal and metabolic fluctuations in women by examining to what extent reward signals are metabolically scaled across the menstrual cycle. We expect that if changes in energy metabolism are linked to changes in reward learning, then greater fluctuations in estradiol and progesterone would be echoed in reward learning as well. Collectively, the proposed project would shed new light on a key mechanism linking learning and motivation with metabolic demands via dopamine transmission. Further insights into vital differences in reward learning due to changes in energy metabolism may help improve the treatment of women and men suffering from disorders characterized by altered reward function such as obesity. By detailing potential sex and cycle effects on the metabolic scaling of reward signals, the proposed project may also provide novel insights into the etiology of eating or mood disorders.