The onset of type 1 diabetes could become easier to detect early, thanks to a recent study. An international team led by William Hagopian, from the Pacific Northwest Research Institute in Seattle, have developed a cost-effective model to screen and predict type 1 diabetes onset in high-risk children. 

Type 1 diabetes is a serious and long-term disease in which the body’s immune system destroys the pancreatic beta cells that produce insulin, disrupting sugar uptake. The resulting insulin deficiency can initiate diabetic ketoacidosis, a life-threatening illness whose symptoms include nausea, dehydration, excessive urine production, and severe abdominal pain. 

Current screening approaches rely on monitoring a set of immune response markers known as beta-cell specific autoantibodies. These approaches are effective but require frequent measurement, which is costly. The onset of type 1 diabetes, which can occur in infants, is also linked to metabolic status, genetic risk, family history, and environmental factors, but these have been ignored in screening and prediction approaches. Therefore, the elements that trigger this onset are unclear, making who will develop T1D and at what age difficult to predict.

Using data from The Environmental Determinants of Diabetes in the Young (TEDDY) cohort, which screened newborns at six centers in the US and Europe, the researchers devised a simple model that can assess the susceptibility of high-risk children to diabetes onset during their first ten years of life. They combined autoantibody monitoring with other aspects known to increase disease risk.

The team produced a three-variable model incorporating autoantibody testing, genetic risk, and family history using TEDDY data from 7,798 high-risk children who were closely followed from birth to 9.3 years of age. They found that combining genetic risk factors with early and frequent autoantibody monitoring enhanced the efficiency of newborn screening to prevent ketoacidosis. 

The model best predicted diabetes in children older than two years over an eight year period, outperforming conventional autoantibody-based screening. The researchers make the case that their model could reduce testing frequency and costs by facilitating adaptive strategies, allowing children to opt out of close follow-up, despite having a high genetic risk, if they maintain low onset probabilities throughout the screening period. 

The team plan to expand their assessment to other populations with distinct genetic backgrounds and environments to further validate their model.

References