PROJECT SUMMARY/ABSTRACT The etiology of eating disorders (ED) is complex, yet largely unknown, resulting in a profound lack of effective treatments and a “crisis in care”. Common to EDs are alterations in the motivation to eat, ranging from extreme food restriction and weight loss, to binge eating coupled with compensatory strategies like self-induced vomiting. Despite the traditional emphasis on diagnostic differentiation based on these physical symptoms, they often overlap, and, along with significant diagnostic crossover (e.g., from anorexia nervosa to bulimia nervosa) over time, suggest shared features that are not well captured by current diagnostic criteria. Persistence of restricted eating, binge eating and/or purging despite negative consequences, along with evidence of altered reward and punishment sensitivity in ED, raise the question of whether a failure to appropriately process and/or learn from rewarding and/or punishing experiences might contribute to repeated engagement in maladaptive approach and avoidance behavior and illness maintenance. This is the first study to apply a multi-dimensional framework of reward processing to ED, by examining how the interplay of RDoC-based Positive Valence measures of `liking' (i.e., the hedonic impact of reward consumption), `wanting' or incentive salience (i.e., motivation to pursue a reward), and learning (i.e., the acquisition of reward-outcome contingencies), which are associated with distinct frontostriatal neurocircuitry, differ across ED subtype and correspond to clinical symptoms at baseline and one year later. We will study 150 demographically-matched women with ED (50 AN-restricting type (AN-R), 50 AN- binge eating/purging type (AN-BP), 50 bulimia nervosa (BN)) and 50 healthy controls (HC) aged 18-35. During fMRI, participants will complete 1) a modified monetary incentive delay (MID) task to assess group differences in both neural anticipation (`wanting') and receipt (`liking') of rewarding and aversive disorder-specific (taste) and generalized (money) stimuli (Aim 1), and 2) a probabilistic associative learning task to assess decision-making and instrumental learning from monetary wins and losses (Aim 2). Aim 3 will examine interactions between `liking', `wanting' and learning and associations with symptoms at study entry and 1 year later. An Exploratory Aim will examine associations of dopamine function, as measured by neuromelanin MRI (NM-MRI), with ED diagnosis and brain response to `liking', `wanting', and learning to further inform mechanistic models of reward in ED. This study is innovative and significant in several ways: 1) it adopts a multi-dimensional framework of reward processing to examine independent and interactive contributions of understudied, yet critically important constructs (e.g., `liking', `wanting', learning) in ED, 2) it assesses the role of stimulus modality (taste, money) and valence in `liking' and `wanting', and 3) relates these constructs to actual sympt...