Hierarchical MMN subcomponents in schizophrenia: Predictive coding biomarkers and clinical translation

Abstract

Mismatch negativity (MMN), a neurophysiological marker of auditory predictive coding, offers unparalleled insights into schizophrenia's mechanistic heterogeneity. This narrative review synthesizes 25 years of evidence (2000-2025) from PubMed, Web of Science, and Scopus (keywords: mismatch negativity, schizophrenia, predictive coding, EEG biomarkers) to propose that early (∼150 ms) and late (∼250 ms) MMN subcomponents reflect distinct hierarchical disruptions. Early MMN deficits, tied to NMDA receptor hypofunction and theta oscillation abnormalities, correlate with cognitive impairment and functional decline. Late MMN impairments, associated with prefrontal predictive hierarchy failures, align with positive symptoms and dopaminergic dysregulation. By stratifying schizophrenia into neurophysiologically defined subtypes, MMN subcomponents enable personalized interventions: NMDA modulators for early deficits, neurofeedback for late abnormalities. We critique the "one-MMN-fits-all" approach, advocating for paradigms that disentangle adaptation (e.g., roving oddball) from predictive coding (local-global tasks). A dual-pathway model positions early MMN as a neurodevelopmental trait marker and late MMN as a state marker of psychosis progression. This framework bridges basic neuroscience and clinical psychiatry, advancing MMN as a translational tool for biomarker-guided care-a fitting tribute to Professor Näätänen's enduring legacy.

Keywords: Biomarkers; EEG; Mismatch negativity; Predictive coding; Schizophrenia.