Biomarkers Related to the Hypotheses of the Pathophysiology of Schizophrenia Spectrum Disorders

Schizophrenia spectrum disorders (SSD) represent a multidimensional and severe mental illness, characterized by abnormalities spanning molecular processes and dysregulated pathways. The pursuit of effective molecular biomarkers for aiding in the diagnosis of this complex disorder has unfolded over the past century. In this chapter, the author conducts a thorough review aimed at scrutinizing the metabolic biomarkers selected over the last 12 years (2011–2023) for the development of diagnostic models for SSD, assessed through receiver operating characteristic (ROC) curve analysis. The study encompasses 32 articles, comprising data from over 2800 SSD patients. The assessment of models using the selected biomarkers yielded area under the curve (AUC) values ranging from 71% to 100%. It is noteworthy that 34% of the articles omitted the presentation of goodness-of-fit and model predictability values. In instances where these values were provided, some did not consistently align with the excellent performance suggested by the AUC; instead, they often indicated overfitting of the model. Nevertheless, out of 84 metabolites that originated from the studies and were subjected to enrichment analysis, 19 significantly matched with the disease signature for SSD (FDR 1.51E-30). These findings underscore the complexity of the landscape of metabolic biomarkers for SSD and the importance of a comprehensive evaluation of model performance.