Characterizing Decision-Making in Anorexia Nervosa Under Conditions of Risk and Ambiguity using Computational Neuroimaging

PROJECT SUMMARY/ABSTRACT
Anorexia nervosa (AN) is the deadliest of all mental illnesses and is associated with significant medical
consequences and functional impairment. Despite significant research efforts, the etiology of the disorder
remains unknown, and outcomes from existing psychological treatments remain suboptimal. Some prior work in
AN has suggested alterations in decision-making, sensitivity to risk, and altered activity in frontostriatal circuits
may be involved in the maintenance of the disorder. However, this work has been mixed in its findings, likely
secondary to limited consideration of subcomponents of decision-making and/or relevant contextual influences
on this process. Drawing from work in behavioral economics and computational neuroscience, alongside clinical
evidence that many outcomes related to food and weight are uncertain/ambiguous, we propose that the decision-
making profile of individuals with AN may be better characterized by ambiguity aversion, such that a sensitivity
toward ambiguous outcomes interrupts successful value-based decision-making and results in maintenance of
dietary restraint despite significant functional and personal cost. To test this hypothesis, we will gather behavioral
choice and neural data from females with AN, restricting type (n=65, ages 16-22) and healthy control (HC)
subjects (n=65) while completing a well-validated decision-making task that allows separation of risky and
ambiguous decisions. Additionally, the AN group will provide measures of symptoms at baseline and 6-month
follow-up to explore how performance on the task relates to the persistence of symptoms. We hypothesize that
the AN group will demonstrate increased ambiguity aversion compared to HC and choice behavior inconsistent
with prevailing models of value-based decision-making (Aim 1). On the neural level, we expect that, compared
to HC, individuals with AN will demonstrate differential activation in frontostriatal regions during ambiguous trials,
rather than risk (Aim 2). Lastly, we expect that neural and behavioral parameters gauging ambiguity aversion
will relate to BMI, dietary restriction, and cognitive symptoms of AN both cross-sectionally and over time, such
that individuals demonstrating greater aversion to ambiguous decisions will have lower body weights and greater
engagement in symptoms, supporting our hypothesized maintenance mechanism (Aim 3). Overall, the current
project will provide an important next step toward a more precise characterization of the cognitive profile of
individuals with AN. In pursuing this knowledge, consistent with the goal of the R15 AREA award, the current
project will simultaneously provide graduate and undergraduate students at Hofstra University with invaluable
hands-on experiences in neuroimaging, computational methods, and clinical research.

Grant Number: 1R15MH128896-01A1
NIH Institute/Center: NIH
Principal Investigator: Emily Barkley-Levenson