Project Summary. Much of human interaction is based on trust. Aging has been associated with deficits in trust- related decision making, likely further exacerbated in age-associated neurodegenerative disease (Alzheimer's disease/AD), possibly underlying the dramatically growing public health problem of elder fraud. Optimal trust- related decision making and avoiding exploitation require the ability to learn about the trustworthiness of social partners across multiple interactions, but the role that learning plays in determining age deficits in trust decisions is currently unknown. To address this gap, this project will (i) characterize basic cognitive processes and neural mechanisms in learning to trust and distrust in healthy aging and in older individuals with subjective cognitive decline (SCD) and a family history of AD, representing an `early' preclinical AD group; and (ii) probe the malleability of these processes with training to form the foundation for future clinical intervention toward reducing exploitation vulnerability in aging. The proposed work is conceptually embedded in the Changes in Integration for Social Decisions in Aging (CISDA) framework. This framework describes how the integration of decision- relevant information is impacted by trajectories of change in theory of mind, memory systems, and social- emotional processing with age. Two innovative trust-learning paradigms – the Social Iowa Gambling Task (sIGT) and the FLorida-Arizona Gambling Task (FLAG) – will be leveraged to test CISDA predictions across three experiments and complimented by an ecologically valid transfer task assessing elder fraud susceptibility. The proposed research addresses three goals. Aim 1/Study 1: Confirm age deficits in learning to trust in an adult lifespan sample that also includes older individuals with SCD and determine the extent to which social cues of trustworthiness bias trust-related decisions and learning in older age and individuals with SCD. Further, this study will use computational modeling to isolate specific learning biases (social cue, loss aversion, and recency) within the CISDA framework. Aim 2/Study 2: Use fMRI versions of the two new learning paradigms to confirm altered anterior cingulate cortex, insula, and amygdala activity, and their interplay, as neural mechanisms of age- associated learning deficits. Aim 3/Study 3: Probe the malleability of the underlying neurocircuitry of trust- learning deficits in aging. This study will utilize real-time fMRI neurofeedback to train healthy older adults in anterior cingulate cortex up-regulation toward enhanced trust-related learning in aging and confirm critical mechanisms of experience-dependent social decisions in aging. This project's interdisciplinary approach encompasses experimental and affective aging, neuroeconomics, and computational neuroscience. Collectively, this research will advance the basic science of social decision making in aging and determine the malleability of under...