Determination of steroid metabolome as a possible tool for laboratory diagnosis of schizophrenia.

Metabolomic studies represent a promising tool for early diagnosis of schizophrenia. The aim of this study was to find differences in the steroid spectrum in patients and controls, and to assess the diagnosis of schizophrenia by building a predictive model based on steroid data. Thirty-nine serum steroids (22 neuroactive steroids and their metabolites and 17 polar conjugates) representing steroid metabolome were measured by gas chromatography-mass spectrometry in 22 drug-naive (first episode) schizophrenia patients (13 men and 9 women) before and after six-month treatment with atypical antipsychotics. The results were compared to the data from healthy subjects (22 males, 25 females). In summary the following significant differences were found: (1) In both sexes higher levels of pregnenolone sulfate and sulfated 5α- as well as 5β-saturated metabolites of C21-steroids in progesterone metabolic pathway were found in patients, pointing to decreased activity of sulfatase. (2) In a few instances decreased levels of the respective 5α-metabolites of C21 steroids were found in patients. (3) As C19 steroids concern, in both sexes there were considerably lowered levels of 5β-reduced metabolites in patients. On the other hand, with only a few exceptions, the treatment did not significantly influence most steroid levels. Further, to assess the relationships between schizophrenia status and steroid levels and to build the predictive model of schizophrenia, multivariate regression with reduction of dimensionality (the method of orthogonal projections to latent structures, OPLS) was applied. Irrespective of the small number of patients, use of this model enabled us to state the diagnosis of schizophrenia with almost 100% sensitivity. Our findings suggest that the assessment of steroid levels may become a valid and accurate laboratory test in psychiatry. A limitation of our study is the absence of subjects with a diagnosis other than schizophrenia, so we cannot conclude whether the results are specific for schizophrenia. On the other hand, steroid metabolome model may be used as a diagnostic tool for further studies.