PROJECT SUMMARY
Anhedonia—the diminished capacity to experience pleasure from normally pleasurable stimuli—is a
transdiagnostic symptom of multiple psychiatric disorders, including depression, schizophrenia and post-
traumatic stress disorder. The recent and rapid growth of anhedonia research is fueled by its alarming clinical
presentation: anhedonia is independently associated with increased risk of suicide, treatment resistance, and
reduced quality of life. Major depressive disorder and schizophrenia—disorders of which anhedonia is a cardinal
symptom—are linked with increasingly high levels of economic burden related to substantial health care costs
and unemployment. There is currently no clear biological definition of anhedonia. As a result, clinicians rely on
self-report measures with no clear established connection to its underlying neurobiology. While behavioral
reward processing deficits and dysfunctional reward circuitry have been observed in anhedonia, in-the-moment
reward responding is frequently preserved, suggesting that memory for the value of the experience may be
compromised. This prompts the central question of this proposal: to what extent is anhedonia a memory
problem? We propose to test a memory-based account for anhedonia as part of our goal to biologically define
the construct. We note that while it is unlikely that memory is the only basis for anhedonia (certainly there are
clear experiential aspects), it may be an understudied and underappreciated player. Critically, a well-defined
memory contribution can help identify novel treatment targets and pave the path to improving clinical practice
with biologically informed decision-making. We use a computational psychiatry approach, combining
mathematical modeling of behavior in novel paradigms with advanced neuroimaging and AI tools to identify and
validate biomarkers relevant to the prevention and treatment of anhedonia. We unite computational models of
value, reinforcement learning, and episodic memory, bridging across the RDoC domains of Positive Valence and
Cognitive Systems. Our aims are to: (1) Test the impact of anhedonia on value-modulated episodic memory and
its neural mechanisms using high-resolution whole brain fMRI; (2) Test the impact of anhedonia on memory-
guided decisions for reward and the associated neural mechanisms using high-resolution whole brain fMRI; and
(3) Test the impact of anhedonia on structural and functional connectivity measures as well as autonomic
regulation. We will also use AI/ML tools to create a multimodal library of predictive biomarkers for anhedonia.
Our ultimate goal is to develop a comprehensive, mechanistic, and actionable memory-based account for
anhedonia using new paradigms, computational models, high-resolution neuroimaging, as well as artificial
intelligence approaches to develop novel interventions and improve clinical practice.