Abstract

Metabolic syndrome (MetS) is a common and clinically important comorbidity in schizophrenia (SCZ), but the biological basis remains unclear. We aimed to characterize the gut microbial and plasma metabolites associated with comorbid MetS in SCZ and to evaluate their potential as integrated biomarkers. A total of 210 participants were enrolled, including 79 patients with SCZ+MetS, 71 patients with SCZ alone, and 60 matched controls. Gut microbiota was analyzed by 16S rRNA gene sequencing, and plasma metabolites were profiled using GC-MS and UPLC-MS. Differential analysis, microbiota-metabolite-clinical phenotype association analysis, and classifier modelling based on gut microbiota and plasma metabolites were mainly performed. We identified thirteen gut microbial genera, including Raoultella , Muribaculum , Pantoea , Paramuribaculum , and Acinetobacter , that showed higher levels in SCZ+MetS patients than in patients with SCZ alone and controls. Amino acids and related derivatives including branched-chain amino acids, L-Glutamic acid, and 3-Methyl-2-Oxovaleric acid showed higher levels in SCZ+MetS patients than in both patients with SCZ alone and control, whereas the levels of phosphatidylcholine were lower relative to the other two groups. Muribaculum , Pantoea , and Raoultella were positively correlated with plasma amino acids and related derivatives including branched-chain amino acids, L-Glutamic acid, and 3-Methyl-2-Oxovaleric acid. A random forest classifier integrating gut microbiota and plasma metabolites achieved good predictive ability for SCZ+MetS, with an AUC of 0.975. In summary, SCZ+MetS is characterized by distinct gut dysbiosis and some disruptions in amino acid and phospholipid metabolism, which may serve as clinically relevant biomarkers of this comorbidity.