Search for plasma biomarkers in drug-free patients with bipolar disorder and schizophrenia using metabolome analysis

Aim: There is an urgent need for diagnostic biomarkers of bipolar disorder (BD) and schizophrenia (SZ); however, confounding effects of medication hamper biomarker discovery. In this study, we conducted metabolome analyses to identify novel plasma biomarkers in drug-free patients with BD and SZ. Methods: We comprehensively analyzed plasma metabolites using capillary electrophoresis time-of-flight mass spectrometry in patients with SZ (n = 17), BD (n = 6), and major depressive disorder (n = 9) who had not received psychotropics for at least 2 weeks, and in matched healthy controls (n = 19). The results were compared with previous reports, or verified in an independent sample set using an alternative analytical approach. Results: Lower creatine level and higher 2-hydroxybutyric acid level were observed in SZ than in controls (uncorrected P = 0.016 and 0.043, respectively), whereas they were unaltered in a previously reported dataset. Citrulline was nominally significantly decreased in BD compared to controls (uncorrected P = 0.043); however, this finding was not replicated in an independent sample set of medicated patients with BD. N-methyl-norsalsolinol, a metabolite of dopamine, was suggested as a candidate biomarker of BD; however, it was not detected by the other analytical method. Levels of betaine, a previously reported candidate biomarker of schizophrenia, were unchanged in the current dataset. Conclusion: Our preliminary findings suggest that the effect of confounding factors, such as duration of illness and medication, should be carefully controlled when searching for plasma biomarkers. Further studies are required to establish robust biomarkers for these disorders.