Combining serum protein concentrations to diagnose schizophrenia: a preliminary exploration.

Objective: It is difficult for clinicians to diagnose schizophrenia solely based on interviews. We explored the diagnostic efficiency and predictive capability of serum biomarkers for schizophrenia.

Method: Levels of β nerve growth factor (β-NGF), brain-derived neurotrophic factor (BDNF), interleukin 6 (IL-6), tumor necrosis factor α (TNF-α), interferon γ (IFN-γ), calcium binding protein S100β, myelin basic protein (MBP), and glial fibrillary acidic protein (GFAP) were measured in the sera of 278 schizophrenia patients, 240 depression and bipolar disorder patients, and 260 healthy controls. DSM-IV-TR criteria were used as the diagnostic criteria for schizophrenia and depressive and bipolar disorders. The diagnostic efficiency was high in patients with schizophrenia compared with the healthy controls. Receiver operating characteristic (ROC) curve analysis was used to ascertain the diagnostic efficiency of the 8 proteins. Data were collected between July 2010 and December 2012.

Results: One-way analysis of variance significantly demonstrated lower serum BDNF, MBP, and GFAP levels (F = 16.504, P < .001; F = 207.209, P < .001; F = 33.668, P < .001, respectively) but higher serum IL-6 and S100β concentrations (F = 15.250, P < .001; F = 12.751, P < .001, respectively) among patients with schizophrenia. ROC analysis of the discriminant scores of the serum β-NGF, BDNF, IL-6, S100β, MBP, and GFAP levels resulted in significant discrimination between the schizophrenia and control groups (AUC = 0.922) and the depressive/bipolar disorder and control groups (AUC = 0.762).

Conclusions: Serum levels of 6 proteins (but not TNF-α and IFN-γ) contribute most to the diagnosis of schizophrenia. These proteins may prove to be useful adjuncts for the clinical assessment of this disease.