Untargeted 1H NMR-based metabolomics unveils distinct circulating biochemical signatures between treatment-resistant and non-treatment-resistant schizophrenia patients: a pilot study

Schizophrenia is a severe psychiatric disorder that affects approximately 1% of the population. Despite the availability of antipsychotic therapies, about 30% of patients develop treatment-resistant schizophrenia (TRS), defined by a lack of response to at least two different antipsychotic trials. Although several genetic and environmental factors have been proposed to explain treatment resistance, metabolomic studies investigating circulating metabolites in TRS remain limited. In this pilot cross-sectional study, we conducted untargeted ¹H NMR-based metabolomics to profile serum metabolites in TRS versus non-TRS patients. Notably, multivariate analysis revealed distinct serum metabolome profiles between the two groups. Additionally, Variable Importance in Projection (VIP) analysis and robust volcano plots showed significant differences between TRS versus non-TRS patients in metabolites involved in lipid and amino acid metabolism. Specifically, serine and glycine emerged as key discriminating molecules, prompting a complementary targeted HPLC analysis in the same serum samples. Although no significant group differences were observed in L-serine, D-serine, the D-serine/total serine ratio, or glycine levels, we found a positive correlation between D-serine levels and cognitive performance, particularly in the area of executive function, across the entire patient cohort. Additionally, a significant correlation between glycine and disorganization symptoms was found selectively in TRS patients. In conclusion, our study offers new insights into potential biomarkers for TRS, highlighting serine-glycine pathway as a possible crossroad between systemic dysmetabolism, NMDA receptor dysfunction, and cognitive impairment in TRS.