Data from continuous glucose monitors can predict nerve, eye, and kidney damage caused by type 1 diabetes. New research From the University of Virginia Diabetes Technology Center. The researchers said this could help patients with type 1 diabetes from experiencing blindness, diabetic neuropathy and other complications.
Researchers wanted to determine whether distance (TIR) is a viable marker of glycemic control. Time and researchers noted that the FDA does not accept continuous glucose monitoring metrics as endpoints of efficacy in clinical trials of new drugs or devices.
In 10 years published in 1993, a 1,440 diabetes control controlled trial (DCCT) established hemoglobin A1C as the gold standard for assessing the risk of complications of type 1 diabetes.
A study at the University of Virginia: It's published online in Diabetes and treatment In January 2025, they evaluated metrics from diabetic control and continuous glucose monitoring data from complications in type 1 diabetes patients.
“Studies of DCCT size performed with continuous glucose monitoring in addition to hemoglobin A1C will be time consuming and expensive,” says Boris Kovatchev, director of UVA Diabetes Technology Center, PhD, in A. I mentioned it. News release. “Virtualizing clinical trials to bridge the gaps in old sparse data using advanced data science methods is the next best thing you can do today.”
Previous studies have demonstrated an association between continuous glucose monitoring and time within the range measured in diabetic complications.
To determine whether the range time may be a marker of glycemic control and diabetes complications, Kovatchev and his colleagues advanced to process datasets from previous diabetes control and complication trials. We used machine learning technology. They were able to create virtual continuous glucose monitor traces for all participants.
They found that patients with a safe glucose range of 70-180 mg/dL over 14 days was a predictor of neurological disorders, retinopathy and nephropathy, as well as hemoglobin A1C levels.
“Looking at the foreseeable future, new technologies such as machine learning, generative artificial intelligence, and Silico modeling are being accepted in the diabetes research community. Classical clinical trials proposed here,” the researchers wrote in the paper. I wrote it. “All of this can reduce the current burden of classic clinical trials for participants and clinical teams.”
