Age-related changes in memory alter decision-making

项目总结。 虽然有几项研究发现老年人的决策能力存在缺陷，但其中许多研究 特别关注强化学习及其相关的奖励网络。强化学习理论 声称我们的选择是由基于我们经验的连续平均值的预期决定的，但这一点 可能提供了一种过于简单化的观点。最近的研究表明，在选择的时间做出决策取决于 不仅基于过去经历的平均值，而且还基于对特定个人经历及其 关联的上下文(例如，体验它们的时间和地点)。这种上下文引导的记忆 由赞助商开发的决策抽样(CGMS)模型断言，选择行为是 除传统强化外，还受决策时检索到的记忆内容的影响 学习(例如，最近奖励的影响)。因此，我们认为与年龄相关的记忆缺陷可能 在影响老年人的选择行为方面发挥着重要作用。考虑到与年龄相关的特征良好的 记忆研究方面的不足，特别是在联想记忆和情景记忆领域，这一项目 解决记忆过程在多大程度上影响决策，以及年龄相关的缺陷是否 决策是由于在记忆中观察到的与年龄相关的损伤得到了充分的记录。横跨3个 实验中，我们利用一种神经计算方法来检查记忆过程和 强化学习有助于整个生命周期的决策制定。在这样做的过程中，我们的目标正是 确定与老化过程中的决策失败相关的机制。在目标1中，我们操作内存需求 以及评估个体记忆能力差异如何影响决策的学习内容 战略。在目标2中，我们将使用计算模型来确定发作期与年龄相关的差异 记忆和强化学习影响老年人与年轻人随后的选择行为 人口。在目标3中，我们将使用高分辨率功能磁共振成像来确定 特定的神经计算和网络在试验中支持和解释老年人的选择行为- 以试行为基础。这些拟议研究的结果有助于开发培训干预措施，以促进 健康老龄化和有针对性的学习干预，以支持早期发现和治疗与年龄相关的疾病 老年人群中的认知功能障碍。这项研究与老年人口有关，网址为 大，但也适用于阿尔茨海默病和相关痴呆症患者，他们在 学习、记忆和决策。

PROJECT SUMMARY. While several studies have identified deficits in decision-making abilities in old age, many of these studies focus specifically on reinforcement learning and its associated reward networks. Reinforcement learning theory states that our choices are shaped by expectations based on a running average of our experiences, but this may provide an overly simplistic view. Recent work has shown that decision-making at time of choice depends not only on the average of past experiences, but also on memories of specific individual experiences and their associated contexts (e.g., the time and place in which they were experienced). This context-guided memory sampling (CGMS) model of decision-making, developed by the sponsor, asserts that choice behavior is influenced by memory content that is retrieved when making a decision, in addition to traditional reinforcement learning (e.g., the influence of recent rewards). We propose that age-related deficits in memory, therefore, may play a significant role in influencing choice behavior in older adults. Given the well-characterized age-related deficits in memory research, particularly in the areas of associative and episodic memory, this project addresses to what extent memory processes influence decision-making, and whether age-related deficits in decision-making are due to the well-documented age-related impairments observed in memory. Across 3 experiments, we utilize a neuro-computational approach to examine how memory processes and reinforcement learning contribute to decision-making across the lifespan. In doing so, we aim to precisely identify the mechanisms associated with decision failures in aging. In Aim 1, we manipulate memory demands and learning content to assess how individual differences in memory ability influence decision-making strategies. In Aim 2, we will use computational modeling to identify how age-related differences in episodic memory and reinforcement learning influence subsequent choice behaviors in older vs. younger adult populations. In Aim 3, we will use high-resolution functional magnetic resonance imaging to determine the specific neural computations and networks that support and explain choice behavior in older adults on a trial- by-trial basis. Findings from these proposed studies can help develop training interventions that promote healthy aging and target learning interventions to support early detection and treatment of age-related cognitive dysfunctions in older adult populations. This research is relevant to the older adult population at large, but also to patients with Alzheimer’s disease and related dementias who experience profound deficits in learning, memory, and decision-making.