Metabolomics analysis reveals BMI-related metabolic characteristics in schizophrenia patients

Obesity is highly prevalent in schizophrenia (SCZ) and contributes to unfavorable clinical and cognitive outcomes, yet its underlying metabolic mechanisms remain unclear. We applied non-targeted metabolomics to characterize body mass index (BMI) related metabolic signatures in chronic SCZ.

A total of 347 chronic SCZ patients were stratified into high BMI (n = 187) and low BMI (n = 160) groups. The Positive and Negative Symptom Scale (PANSS) and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) scales were used to evaluate psychopathology and cognition. Fasting serum samples were collected and analyzed using non-targeted metabolomics via ultra-high-performance liquid chromatography-high-resolution mass spectrometry. Multivariate and univariate statistical approaches, pathway enrichment analyses, receiver operating characteristic (ROC) analyses, and linear regression models were applied.

Fifty-eight metabolites differed significantly between high and low BMI groups (primarily involving 12 amino acids and 8 fatty acids). Pathway analyses revealed significant perturbations in Arginine biosynthesis (p < 0.001), Nicotinate and nicotinamide metabolism (p = 0.008) and Phenylalanine metabolism (p = 0.037). Panels combining differential amino acids and fatty acids demonstrated moderate discriminative performance for high BMI status (AUC > 0.7). Moreover, C-Mannosyl Valine and C-Mannosyl Phenylalanine were associated with cognitive impairment in the low BMI group (all p < 0.05), whereas 2-Aminoheptanoic acid were associated with negative symptoms severity in the high BMI group (p < 0.001).

Distinct metabolic profiles characterize BMI heterogeneity in SCZ and are differentially associated with cognitive impairment and negative symptoms.