Computational Neuroscience to Decode Higher-Order Fear in PTSD

University Of Texas At Austin, Austin TX

Investigators

Abstract

PROJECT SUMMARY/ABSTRACT The objective of this K23 is to support new mentored training in the computational neuroscience of higher-order threat learning and memory in posttraumatic stress disorder (PTSD) as the candidate transitions to an independent career as research faculty specializing in integrating lab work with real-world experience sampling in larger, heterogenous research samples to further translational efforts in conceptualizing and treating trauma and anxiety psychopathology. PTSD affects ~6% of the US population annually and is associated with substantial health and economic burden. RDoC-aligned threat conditioning paradigms have substantially contributed to etiological models of PTSD and exposure therapies, yet some patients fail to respond to treatment or experience relapse. One potential explanation is that exposure primarily addresses directly associated stimuli and experiences. Consequently, reductions in fear tend not to spread to more nebulous and indirect higher order networks of generalized threat associations commonly seen in pathological anxiety. Prolific nonhuman animal work has produced a neurobiological model of the mechanisms of higher-order threat learning, yet systematic translation to humans is almost nonexistent. Thus, we lack a compelling human neural model needed to anchor translational efforts that integrate higher-order learning principles into the theoretical framework of exposure. During this award, the candidate will build on initial functional MRI training and incorporate new training in advanced computational decoding applications of multivariate pattern analysis (MVPA) to test the complex interplay of aversive learning and memory mechanisms that result in dense and difficult-to-treat higher-order threat associations. Aim 1 will a) apply MVPA to test how memory integration mechanisms facilitate threat generalization across higher-order pathways in implicated medial temporal structures (basolateral amygdala, hippocampus, perirhinal cortex) and medial prefrontal cortex; b) determine the long-term durability of higher- order threat generalization biases memory retrieval at 24-hour or 1-month after initial learning; and c) test the hypothesis that higher-order threat generalization is heightened in PTSD compared with participants without psychopathology. Aim 2 will extend this work and address heterogeneity inherent to the PTSD diagnosis via subgroup and dimensional anxiety-related trait modeling. Aim 3 uses these lab-based neural metrics of threat generalization to predict real-world daily experience of PTSD and anxiety-related symptoms and physiology, which is needed to bring neural models in better alignment with the clinical reality of pathological anxiety. To facilitate this work, the candidate will receive extensive supervised training in longitudinal ambulatory assessment designs employing state-of-the-art wearable technology and experience sampling methods. The proposed work can advance neurobiological models of anxiety-related psychopathology and set the stage for addressing prominent treatment limitations. This award lays the groundwork for the candidate to achieve these research and training goals and to contribute to efforts optimizing treatments using lab and real-world data.