Multifactorial, biomarker-based model for assessing the state of patients with schizophrenia

Abstract

Relevance. Objective comparison of biological markers and real clinical presentation is especially difficult in mental disorders, which are classified according to a large number of diagnostic criteria and a wide variety of symptoms. Therefore, the development of an effective system of biochemical markers and assessment of their relationship to optimize the diagnosis and treatment of schizophrenia are relevant.The aim of the study was to develop a statistical model that combines known and tested biochemical markers for mental illnesses in patients with schizophrenia.Materials and methods. The study included 47 women aged 18–50 years (median age – 22 years) with the diagnosis of schizophrenia (ICD-10, F20) and 25 healthy women of the same age. The model was based on the functional activity of complement, thrombodynamics parameters, markers of inflammation, glutamate and energy metabolism, and antioxidant defense, which were shown to be associated with the severity of schizophrenia. The listed markers were evaluated in plasma, platelets, and erythrocytes of sick and healthy individuals.Results. Statistical software found pair correlations and features of the distribution of all markers as random variables in the examined groups and evaluated correlations between pairs of markers. Ten biomarkers were identified and united into a system that was adequately described by the logistic regression model. The model was evaluated using the Pearson’s test (χ2(11) = 57.6, p = 0.001) and calculation of correct predictions (91 and 80%) for samples of patients and healthy people, respectively.Conclusion. Calculating the logistic equation resulted in the probability that the patient has schizophrenia involving the immune system, hemostasis, and oxidative stress. This model can be considered as a new formalized approach to the preclinical diagnosis of mental illnesses.