A Systematic Integration of Current Evidence on the Role of Error-Related Negativity as a Biomarker in Psychopathology using Pairwise and Network Meta-Analyses

This meta-analysis integrates data from 139 studies involving 10,084 participants, identified through a systematic literature search, to examine neural error processing across various mental disorders—including obsessive-compulsive disorder (OCD), anxiety disorders, depressive disorders (MDD), attention-deficit/hyperactivity disorder (ADHD), autism spectrum disorder (ASD), substance-related and addictive disorders (SUD), and psychotic disorders. Using pairwise and network meta-analyses, we identified robust alterations in error-related brain responses measured by the error-related negativity (ERN). OCD (g = -0.43) and anxiety disorders (g = -0.23) exhibited heightened error-related neural responses (larger ERN), whereas ADHD (g = 0.26), ASD (g = 0.30), and psychotic disorders (g = 0.67) demonstrated blunted responses (smaller ERN). No significant differences in ERN were detected in MDD (g = 0.00) or SUD (g = 0.00) overall. Patients with psychotic disorders exhibited the greatest blunting of error processing, while those with OCD showed the strongest amplification—both significantly differing from healthy participants and all other clinical groups. These findings support the transdiagnostic potential of the ERN as a reliable neurocognitive marker. ERN alterations in anxiety disorders, OCD, and psychotic disorders may serve as endophenotypes or neural risk markers that precede symptom onset and offer predictive validity, aligning with established endophenotype criteria. Increased ERN signals may reflect amplified prediction errors, leading to over-control and risk-avoidant behaviors, while decreased ERN responses indicate hypoactive error processing, impairing adaptive responses to errors and feedback. Overall, ERN alterations hold promise as biomarkers of cognitive control dysregulation, with important implications for personalized interventions and early risk detection.