For decades, it was believed that Type 1 Diabetes (T1D) was a sudden autoimmune reaction occurring in childhood or adolescence. However, modern medicine is finding increasing evidence that the pathogenesis of the disease begins much earlier and may originate as early as the prenatal period. The primary challenge for scientists has been identifying the pathological process at a stage when the pancreatic cells responsible for insulin synthesis are still viable.
Swedish researchers studied the umbilical cord blood proteins of approximately 17,000 newborns and established that predicting the disease is possible from the very first minutes of birth.
Fundamentals of Type 1 Diabetes Pathogenesis
Type 1 Diabetes causes the destruction of pancreatic -cells, which in turn leads to an insulin deficiency in the body. Genetic factors, specifically HLA DR-DQ genes, determine 40–50% of the risk of developing the disease. Among environmental factors, enteroviral infections experienced during childhood, as well as respiratory diseases and stress experienced by the mother during pregnancy, are significant.
During the inflammatory process, cytokines activate the STAT1, IRF1, and NF-B signaling pathways. -cells attempt to adapt, but this process often ends in failure. Since autoantibody testing does not reveal the early stages of the disease, the goal of this study was precisely to eliminate this diagnostic gap.
Methodology and the ABIS Cohort
Researchers began observing 16,683 newborns from the ABIS (All Babies in Southeast Sweden) cohort between 1997 and 1999. The uniqueness of this cohort is determined by the possibility of studying the dynamics of children’s health status over almost three decades. Through national registries, the diagnosis of Type 1 Diabetes was monitored until 2023, revealing 167 cases of the disease. Using machine learning algorithms and proteomic analysis, scientists identified key proteins in umbilical cord blood. These proteins represent early prognostic markers for the development of the pathology.
Prenatal Factors and Prognostic Accuracy
Within the scope of the study, 44 prenatal and perinatal factors were evaluated. In addition to family history, cesarean section, stays in the neonatal intensive care unit (NICU), and gastroenteritis experienced by the mother during pregnancy were found to correlate with the risk of developing the disease. Furthermore, the use of antibiotics and psychotropic medications were recorded as significant factors.
Nevertheless, an analysis based solely on clinical data showed low prognostic accuracy. This result confirms that clinical indicators, when used in isolation, cannot provide a comprehensive assessment of disease risk.
Key Markers in Umbilical Cord Blood
As part of the study, 386 immune proteins were measured in umbilical cord blood. Analysis showed that in children who were later diagnosed with Type 1 Diabetes, the concentrations of HLA-DRA, IDS, SCGB3A2, and CTSC proteins were significantly elevated. Conversely, high levels of TIMP3, CD40LG, and other protective proteins were recorded in the control group (healthy individuals).
Genetic analysis confirmed that proteins such as HLA-DRA, CD40LG, and TIMP3 participate in the regulation of cytokine signaling pathways. This process directly determines the immune response of leukocytes, which plays a decisive role in the early stage of disease development.
Age of Diagnosis and Specific Protein Signals
The most distinct changes were revealed in children who were diagnosed with Type 1 Diabetes before the age of 5. In cases identified in adults (18–24 years), a completely different picture was recorded: in this group, levels of IL-20 and CCL4 proteins were elevated.
In early-age cases, the activation of T-cells, NF-B signaling pathways, and the enhancement of immune mechanisms in response to certain infections (toxoplasmosis, yersiniosis) emerged. These data indicate that the pathogenesis of the disease and its early markers are significantly dependent on the age at which the diagnosis is manifested.
Advantages of Protein Markers
The study showed that even in the absence of HLA risk alleles, the concentrations of HLA-DRA and IDS proteins were still high in children who later developed Type 1 Diabetes. This indicates that protein markers provide more extensive information about the risk of the disease than genetic data alone.
Analysis conducted using machine learning (XGBoost model) confirmed that the prognostic accuracy based solely on proteomic data reached 0.89 (AUC). This result highlights the fact that studying biomarkers is a much more effective means for early identification of the disease than genetic screening alone.
Interconnection of Toxins, Metabolites, and Proteins
The research revealed a close link between immune proteins and environmental pollutants such as PFOS (perfluorooctane sulfonate). It was established that stearic acid triggers an increase in the levels of IDS and HLA-DRA proteins. Biologically, each of these components plays a specific role in the development of the disease.
Specifically, HLA-DRA enhances antigen presentation, while IDS facilitates insulin secretion from pancreatic islet cells. In turn, the TIMP3 protein protects pancreatic islets and limits tumor necrosis factor (TNF), representing a kind of defense mechanism of the organism.
New Opportunities for Prevention
Proteomic analysis of umbilical cord blood provides the opportunity to determine the risk of the disease immediately upon birth (especially in the case of patients under 5 years of age). This process implies the reduction of prenatal inflammation and the prevention or significant “postponement” of the formation of insulin dependence. Accordingly, preventive measures for Type 1 Diabetes begin as early as the prenatal period.
Source: Nature Communications
