Computational Modeling of Ambiguity Aversion in Internalizing Disorders: Neural and Behavioral Markers, and a Test of Target Engagement - PROJECT SUMMARY Uncertainty is a part of daily life, but those with difficulty tolerating uncertainty are at risk for anxiety-related and other internalizing disorders. A key contributor to difficulty with uncertainty is an elevated aversion to choices that have ambiguous outcomes (aka ambiguity aversion [AA]). Studies, however, have not tested (a) AA as a transdiagnostic dimension of internalizing psychopathology, and (b) the time course of the brain’s responses to ambiguous choices. This study thus uses computational modeling, electrophysiology, and a novel experimental manipulation to test whether AA is a behavioral and neural mechanism of internalizing psychopathology in a transdiagnostic sample of 140 adult participants with moderate to high intolerance of uncertainty. We will test whether computational parameters of behavioral AA (Aim 1) and event-related potentials [ERPs, neurophysiological markers with millisecond resolution) to ambiguity (Aim 2) exhibit convergent validity (associations with anxiety and depression dimensions, intolerance of uncertainty) and discriminant validity (lack of associations with sensation seeking). Aim 3 will further test AA as a mechanism of internalizing disorders by testing whether an experimental manipulation (i.e., a brief behavioral intervention) “moves” AA. This experimental approach not only provides a robust test of AA as a mechanism within internalizing psychopathology, but also informs future clinical trials that could, for example, use an uncertainty-focused intervention as an adjunct treatment to existing interventions. These aims are novel in two additional ways. First, collecting ERPs related to AA is novel and cost-effective (relative to prior neuroscience studies which have used fMRI), and will provide insight about the time course about the brain’s responses to ambiguity (e.g., whether AA is driven by early [i.e.,~100ms], or later [after 300ms] reactions to ambiguity). Second, while past studies have examined AA in only single disorders, this project will examine associations with transdiagnostic dimensions that theoretically relate to AA (convergent validity) and dimensions that should theoretically not relate to AA (discriminant validity), a rigorous approach that will ultimately have larger clinical impacts. In sum, this project has the potential to identify a transdiagnostic intervention target using an easily administered behavioral and brain assessment of AA, a goal that is directly in line with NIMH’s Strategic Objectives. This K23 will also provide the applicant with the training and mentored research experience necessary for his long-term goal of studying disrupted brain and computationally-derived behavioral mechanisms of internalizing disorders. Training will focus on 1) computational modeling of decision-making behavior, 2) experimental target engagement approaches to interventions, 3) advanced electrophysiological methods and dimensional, transdiagnostic “RDoC” models of psychopathology, and 4) development of grantsmanship, lab management and mentorship skills. This training will be conducted at Northwestern University and facilitated by a team of experts in computational, psychophysiological, and therapeutic approaches, all studying mechanisms of change in internalizing disorders.
