Electrophysiological Indices of Performance Monitoring Across the Anxiety- and Obsessive-Compulsive Spectrum

Abstract

Errors are a ubiquitous part of human experiences, associated with potential risks but also serving as resources for improving performance. The neural monitoring of performance can be measured through event-related brain potentials (ERPs), such as the error-related negativity (ERN) and the correct-response negativity (CRN). Meta-analyses have identified an altered ERN as a trait-like risk marker for both internalizing and externalizing disorders, with increased amplitudes observed in anxiety and obsessive-compulsive disorders. However, this association is not consistent across all related diagnostic categories, often attributed to differing underlying transdiagnostic factors (e.g., trait worry). Challenging the assumed trait-like nature of the ERN, symptom fluctuations (e.g., state worry) may also influence the ERN. Furthermore, there is limited understanding of how an increased ERN contributes to psychopathology, along with only few studies evaluating multiple disorder categories and testing for effect specificity. This dissertation aims to refine the understanding of electrophysiological indices of performance monitoring across the anxiety and obsessive-compulsive spectrum by employing a transdiagnostic and dimensional approach that includes various diagnostic groups and ERPs. Disentangling the associations of the ERN/CRN with trait and state worry, the preregistered randomized-controlled longitudinal study 1 (n = 90) did not indicate intraindividual variability of the ERN/CRN due to the experimental manipulation of state worry. Instead, higher levels of trait worry were linked to an enhanced ERN/CRN, particularly in women. Study 2 investigated the role of the ERN/CRN as risk markers using a preregistered cross-sectional design (n = 156). This study found no significant ERN/CRN differences between participants with obsessive-compulsive disorder, social anxiety disorder, or specific phobia, as well as control participants. However, after reclassifying the sample, study 2 found trend-level significant larger ERN/CRN amplitudes in clinical compared to nonclinical participants, as well as significant larger amplitudes for those with a family risk for internalizing symptoms. Additionally, joint analyses of study 1 and 2 (n = 246) confirmed the ERN/CRN link with trait worry for women. Finally, the prospective study 3 addressed the pathways from an enhanced ERN/CRN to the development of internalizing psychopathology by employing mediation analyses during the COVID-19 pandemic (n = 113). Results indicated that an increased pre-pandemic ERN/CRN were indirectly linked to anxiety, obsessive-compulsive, and depressive symptoms during the pandemic, mediated by a heightened COVID-19 risk perception and increased stress. As a synthesis of the available literature and the current results, this dissertation proposes an integrative model of error-related brain activity and psychopathology. This model identifies an altered ERN as a neural vulnerability marker linked to both the internalizing and externalizing spectrum through latent dimensions underlying multiple clinical phenotypes, such as trait worry for anxiety- and obsessive-compulsive symptoms or impulsivity for substance use and hyperactivity. Moreover, it considers potential moderators and mediators of these associations, such as biological sex, temperament, adverse life events, and interpersonal stress. However, given the heterogeneous CRN literature, it seems premature to incorporate an altered CRN as an additional model component. Nevertheless, the model offers a neuroscientifically informed framework with the potential to not only guide future research but also to facilitate the development of targeted interventions based on knowledge from brain potentials, ultimately aimed at preventing mental health impairments and reducing the risk of psychopathological conditions.