Understanding the Error in Psychopathology: Notable Intraindividual Differences in Neural Variability of Performance Monitoring

Abstract

BackgroundAbnormal performance monitoring is a possible transdiagnostic marker of psychopathology. Research on neural indices of performance monitoring, including the error-related negativity (ERN), typically examines group and interindividual (between-person) differences in mean/average scores. Intraindividual (within-person) variability in activity captures the capacity to dynamically adjust from moment to moment, and excessive variability appears maladaptive. Intraindividual variability in ERN represents a unique and largely unexamined dimension that might impact functioning. We tested whether psychopathology group differences (major depressive disorder, generalized anxiety disorder, obsessive-compulsive disorder) or corresponding psychiatric symptoms account for intraindividual variability in single-trial ERN scores. MethodsHigh-density electroencephalogram was recorded during a semantic flanker task in 51 participants with major depressive disorder, 44 participants with generalized anxiety disorder, 31 participants with obsessive-compulsive disorder, and 56 psychiatrically healthy participants. Time-window mean ERN amplitude was scored 0–125 ms following participant response across four frontocentral sites. Multilevel location-scale models were used to simultaneously examine interindividual and intraindividual differences in ERN. ResultsAnalyses indicated considerable intraindividual variability in ERN that was common across groups. However, we did not find strong evidence to support relationships between ERN and psychopathology groups or transdiagnostic symptoms. ConclusionsThese findings point to important methodological implications for studies of performance monitoring in healthy and clinical populations—the common assumption of fixed intraindividual variability (i.e., residual variance) may be inappropriate for ERN studies. Implementation of multilevel location-scale models in future research can leverage between-person differences in intraindividual variability in performance monitoring to gain a rich understanding of trial-to-trial performance monitoring dynamics.