The amount of time patients were in a safe blood-sugar range of between 70 and 180 mg/DL over a 14-day period was as good a predictor of neuropathy, retinopathy and nephropathy as the standard approach using hemoglobin A1c levels, the researchers found.

“The landmark 10-year, 1,440-person Diabetes Control and Complications Trial (DCCT), published in 1993, established hemoglobin A1c as the gold standard for evaluating the risk for complications from type 1 diabetes. However, the use of continuous glucose monitoring is on the rise and there is no study of the magnitude of the DCCT to affirm CGM-based metrics as standard for evaluating diabetes control,” said Boris Kovatchev, PhD, director of the UVA Center for Diabetes Technology. “The lack of long-term large-scale CGM data has a number of clinical and regulatory implications; for example, CGM is still not accepted as a primary outcome from diabetes drug studies.”

Using Landmark Diabetes Data

The DCCT obtained hemoglobin A1c readings from participants either monthly or every three months, plus a blood-sugar profile every three months. The data is available, upon request, in the archives of the National Institute of Diabetes and Digestive and Kidney Diseases.

Using advanced machine learning techniques to process the DCCT data sets, the researchers were able to create virtual continuous glucose monitor traces for all participants and for the duration of their participation in the trial.

The researchers found that 14 days of data from the virtual continuous glucose monitors had a similar ability to predict diabetes complications as hemoglobin A1c readings. In addition to the time spent in a safe blood-sugar range of 70 to 180 mg/DL, the researchers also found that other continuous glucose monitor readings also accurately predicted diabetes complications. These readings included the time spent in “tight range” (between 70 and 140 mg/DL) as well as the time spent above 140 mg/DL, above 180 mg/DL and above 250 mg/DL.

With continuous glucose monitors now in frequent use by patients with diabetes, these new findings could aid patients in managing their diabetes as well as help researchers continue to advance diabetes care.

“A study of the magnitude of the DCCT done with continuous glucose monitoring in addition to hemoglobin A1c would be prohibitively time-consuming and expensive,” Kovatchev said. “Virtualizing a clinical trial to fill in the gaps in old, sparse data using advanced data science methods is the next best thing we can do today.”