Biomarkers for Predicting Clinical Deterioration in Schizophrenia-Spectrum Disorders: A Systematic Review

Background/Objectives: Psychotic relapse affects over 80% of individuals with schizophrenia-spectrum disorders, driving long-term disability and hospitalization. Clinical relapse management relies on symptomatic monitoring without objective neurobiological tools to guide individualized antipsychotic decisions. Methods: This systematic review synthesizes evidence on neurophysiological, blood-based, molecular, neuroimaging, and digital biomarkers for relapse prediction in schizophrenia-spectrum disorders. Results: Following the PRISMA 2020 guidelines, five databases were searched through March 2026 for longitudinal biomarker studies. Quality was assessed using the Newcastle-Ottawa Scale and PROBAST; findings were synthesized narratively due to substantial heterogeneity. From the 6812 citations screened, 21 studies were included across clinical high-risk, first-episode, and established illness populations. Conclusions: Mismatch negativity and P300 event-related potential (P300) showed the most consistent associations with relapse vulnerability, with mismatch negativity demonstrating relative independence from antipsychotic effects. Inflammatory and neuroendocrine markers-interleukin-6, C-reactive protein, and cortisol awakening response-predicted poor treatment response in multiple longitudinal investigations. Peripheral blood gene expression profiling identified TCF4 network dysregulation as a candidate molecular marker of impending relapse. Neuroimaging models did not outperform standard clinical variables. Digital phenotyping showed ecological promise but remains methodologically nascent. No single biomarker achieves sufficient accuracy for clinical implementation. Neurophysiological and inflammatory markers are the most tractable candidates for monitoring protocols. Future research should prioritize multimodal longitudinal designs, external validation, and systematic antipsychotic confounding control. Among the biomarkers reviewed, mismatch negativity and the interleukin-6/cortisol awakening response combination represent the most tractable candidates for pilot clinical implementation, particularly in specialized early psychosis services and antipsychotic dose-reduction research contexts; no biomarker currently achieves sufficient accuracy for routine use in maintenance treatment decisions.