Predicting the Development of Type 1 Diabetes Using Blood Proteins
A set of altered proteins are identified that can predict a condition known as islet autoimmunity, a precursor for developing Type 1 diabetes. This discovery is done by researchers at the Department of Energy's Pacific Northwest National Laboratory and their colleagues.
This work marks the beginning, not the end, of a search for a way to predict who will develop the disease. More work needs to be done to verify the results and test whether the findings apply to everyone, not just the children who were in the study and were genetically predisposed to developing the disease.
‘A machine learning analysis predicts if the individual would remain in autoimmunity or develop Type 1 diabetes 6 months before the appearance of autoantibody.#type1 diabetes #autoimmunity #artifical intelligence’
A biomarker that detects pending autoimmunity would help physicians monitor a patient's condition, perhaps to detect worsening health and to prompt faster medical care even before symptoms appear.
The results are the culmination of a nine-year study led by PNNL scientist Thomas Metz that looked at hundreds of proteins in more than 8,000 blood samples from nearly 1,000 children at six study sites in North America and Europe (). The findings are published in the journal Cell Reports Medicine.
The children have been participants in a larger umbrella study known as TEDDY-The Environmental Determinants of Diabetes in the Young. TEDDY includes children who, because of their genetic makeup, are more likely than others to develop Type 1 diabetes. The study, nearing its 20th year, seeks to understand why some children develop the disease while others don't.
New Way to Know Who Will Develop Type 1 Diabetes Before Occurrence
In the latest study, they analyzed blood plasma samples of nearly 1,000 children from their birth up to the age of 6. Researchers identified a set of 83 proteins whose combination of changes predicted which children went on to develop either islet autoimmunity or Type 1 diabetes ().Right now, there is no way to know if or when either islet autoimmunity or diabetes will occur in people who are genetically predisposed. Doctors do know that when a patient develops at least two islet autoantibodies, they have islet autoimmunity and will develop diabetes-but they do not know what triggers autoimmunity or when diabetes will manifest.
This allows us to learn more about what causes the immune system to turn on the body. This could help us tease out and understand the mechanisms at play in the development of diabetes better than we do currently and provide potential targets for intervention.
Researchers created a machine learning algorithm that analyzed the vast amount of data-information on hundreds of proteins from nearly 1,000 children, with multiple blood samples from each taken from their birth until they were 6 years old.
The scientists did the study in two phases. In the discovery phase, the team studied 2,252 blood samples from 184 children. There the team identified 376 proteins that were altered in patients who later developed islet autoimmunity or Type 1 diabetes. Then the team performed a more in-depth validation study, looking at 6,426 blood samples from 990 children.
The 83 blood proteins the team pinpointed in the validation study are key to several important processes in the body, including antigen presentation, complement and blood clotting, inflammatory signaling, and metabolism ().
The list of proteins matches up well with the proteins that the team knows are active in the pancreas of patients who have diabetes. The team identified those proteins previously in work conducted through the Human Islet Research Network.
Researchers hope to continue the study by analyzing additional blood samples that have already been collected from the same children when they were older. Those blood samples were collected through TEDDY until the children turned 15.
In other studies related to diabetes and inflammation, PNNL scientists are studying exactly how our body's insulin-producing cells become vulnerable to an autoimmune attack-how insulin-producing cells in the pancreas are targeted and killed by the body. Researchers are also looking at how omega-3 fatty acids can protect those cells.
References:
- Plasma protein biomarkers predict the development of persistent autoantibodies and type 1 diabetes 6 months prior to the onset of autoimmunity - (https:www.cell.com/cell-reports-medicine/fulltext/S2666-3791(23)00212-4?)
- Biomarker discovery study design for type 1 diabetes in The Environmental Determinants of Diabetes in the Young (TEDDY) study - (https:onlinelibrary.wiley.com/doi/abs/10.1002/dmrr.2510)
- Predictive Modeling of Type 1 Diabetes Stages Using Disparate Data Sources - (https:diabetesjournals.org/diabetes/article/69/2/238/16226)
Source: Eurekalert