Abstract
BACKGROUND: Prediction of type 1 diabetes mellitus (T1DM) at the preclinical stage allows for timely initiation of preventive therapeutic interventions and may prevent disease progression. AIM: This work aimed to evaluate the potential of predicting T1DM based on autoantibody concentrations and their changes. METHODS: A prospective longitudinal cohort study was conducted in three regional children’s hospitals: in Nizhny Novgorod, the Chuvash Republic, and the Republic of Mari El. The study included children aged 0–18 years hospitalized with newly diagnosed T1DM between 2017 and 2020, as well as their healthy siblings (enrolled concurrently). Data from 517 participants were analyzed: 314 children with newly diagnosed T1DM and 203 healthy siblings. Regression modeling was applied for the analysis of repeated measurements. Antibodies to glutamate decarboxylase, tyrosine phosphatase, and zinc transporter 8 were determined. RESULTS: Among healthy siblings, a high risk of developing T1DM was associated with: elevated baseline concentrations of all three antibodies (57.5–92 times higher than reference values on average); a significant and rapid decrease in glutamate decarboxylase and tyrosine phosphatase concentrations −23.29 and −43.3 IU/mL per month, respectively; and a slight and very slow decrease in zinc transporter 8 concentration −5.3 U/mL per month. CONCLUSION: Modeling the longitudinal profiles of glutamate decarboxylase, tyrosine phosphatase, and zinc transporter 8 may serve as the basis for the development of more advanced and precise diagnostic systems. This approach appears promising but requires further investigation.

This publication has 12 references indexed in Scilit: