M6. Relationship Between Peripheral Amino Acid Concentration and Diagnosis and Severity of Schizophrenia Using Structural Equation Modeling Analysis

Background: Altered peripheral levels of some amino acids (eg, glycine and D-serine) have previously been reported to play a role in schizophrenia. However, the elucidation of the role of amino acids in the pathophysiology of schizophrenia requires a powerful statistical approach by analyzing a set of amino acids to provide a good fit, taking into account their interactive behaviors. Thus, we utilized structural equation modeling (SEM) for the analyses of peripheral amino acid levels. Methods: Fifty-one patients with schizophrenia (males, 31; age, 52.6 ± 10.0) and 37 normal controls (males, 17; age, 50.2 ± 10.2) were recruited between 2010 and 2014. All patients were treated at Dokkyo Medical University Hospital, Mori Hospital, or Shimotsuga General Hospital. The patients were assessed by 2 independent psychiatrists using the DSM-IV-TR criteria; control subjects were confirmed to be free of mental illness. The severity of illness was evaluated using the Positive and Negative Syndrome Scale. The peripheral plasma levels of glutamate, glutamine, 2 enantiomers of serine (L-serine, D-serine), glycine, histidine, aspartate, asparagine, threonine, alanine, proline, valine, leucine, and isoleucine were measured using high-performance liquid chromatography. To adjust the interactive effects of a set of amino acids, age, sex, and smoking status, SEM was conducted for all participants to investigate whether individual amino acid phases account for the diagnosis of schizophrenia. In addition, SEM was conducted to determine the relationship between the severity of schizophrenia and concentration of each amino acid, in which the same factors were adjusted. The study was approved by the ethics committee of the Dokkyo Medical University School of Medicine. Results: SEM revealed that the concentrations of glycine, L-serine, asparagine, leucine, isoleucine, and threonine significantly accounted for the diagnosis of schizophrenia. Furthermore, the use of SEM to test the relationship of the severity of schizophrenia and concentration of amino acids demonstrates that glycine is related to positive and negative PANSS scores. Conclusion: SEM proved to be a suitable and powerful method to evaluate the interactive effects of a set of amino acids and showed that some key amino acids are involved in the pathophysiology of schizophrenia. However, some factors that could affect the system were not involved in this analysis. To obtain more accurate results, detailed examination is warranted.