Prediction of the uncontrolled course of autoimmune type 1 diabetes in children

Abstract

Background. Diabetes mellitus as a disease of civilization is one of the main causes of death in most developed countries. Purpose of the study: to determine the risk factors for the uncontrolled course of autoimmune type 1 diabetes in children. Materials and methods. To determine the type of a newly diagnosed diabetes, 100 children aged 1 to 18 years were examined by evaluating the levels of autoantibodies to glutamic acid decarboxylase, tyrosine phosphatase, zinc transporters using the enzyme immunoassay method (Euroimmun) and detecting genetic defects in b-cell function by polymerase chain reaction and Illumina sequencing technology. To predict the course of autoimmune type 1 diabetes with the help of Wald sequential analysis, 100 clinical, immunological, molecular genetic, instrumental diagnostic parameters were analyzed with relative risk and diagnostic coefficient determination after follow-up observation of 53 children with uncontrolled (main group, n = 22) and controlled course (control group, n = 31). Results. Autoimmune type 1 diabetes was diagnosed in 98 %, genetic defects in b-cell function (maturity-onset diabetes of the young type 5 and Wolfram syndrome type 1) — in 2 % of the examined children aged 1 to 18 years. Seventeen factors with sufficient informative prognostic significance (I ≥ 0.5), which are convenient for use in clinical practice, were identified as predictors for the development of an uncontrolled course of autoimmune type 1 diabetes in children: history of influenza one year before the onset of type 1 diabetes (I = 1.73), increased alanine aminotransferase level (I = 1.27), presence of antibodies to glutamic acid decarboxylase (I = 1.14), manifestation of type 1 diabetes at the age of 1–4 years or more than 6 years (I = 0.98), increased blood urea level (I = 0.86), macrosomia at birth (I = 0.71), an increase in glycated hemoglobin content of more than 8 % at the onset of the disease (I = 0.7), a decrease in alkaline phosphatase level (I = 0.66), heredity for autoimmune illnesses (I = 0.62), concomitant diseases of the cardiovascular system (I = 0.61), hyperglycemia of more than 23 mmol/l (I = 0.56), delayed hospitalization of a child for more than 1 month from the manifestation of type 1 diabetes (I = 0.54), lack of feeding regimen during the 1st year of life (I = 0.52), glucosuria (I = 0.52), history of recurrent acute respiratory infections (I = 0.52), hypoproteinemia (I = 0.51), the presence of chronic foci of infection (I = 0.5). Conclusions. To improve the quality of diagnosis, prediction of the course and personalized treatment of diabetes, especially in patients with a newly diagnosed disease, it is necessary to determine specific autoantibodies against the islet apparatus of the pancreas. In cases with a negative result, a further referral for molecular genetic testing is recommended in order to exclude non-autoimmune types of diabetes. With laboratory verified autoimmune type 1 diabetes, an easy-to-use mathematical model with high valid features is proposed to predict the further course of the disease in children.

Keywords

• autoimmune diabetes
• children
• prediction
• predictors
• wald test