Data from continuous glucose monitors can estimate potential damage to nerves, eyes, and kidneys associated with type 1 diabetes, according to research conducted by the University of Virginia Center for Diabetes Technology. This indicates that healthcare professionals might utilize information from these devices to prevent conditions such as blindness, diabetic neuropathy, and other significant diabetes-related issues.
Data from continuous glucose monitors can estimate potential nerve, eye, and kidney damage linked with type 1 diabetes, researchers at the University of Virginia Center for Diabetes Technology have discovered. This implies that healthcare providers could leverage data from these devices to help protect patients from serious complications, including blindness and diabetic neuropathy.
The researchers found that tracking the duration patients maintained a healthy blood sugar level between 70 and 180 mg/DL over a two-week period was just as effective in predicting neuropathy, retinopathy, and nephropathy as the conventional method that relies on measuring hemoglobin A1c levels.
“The groundbreaking 10-year, 1,440-participant Diabetes Control and Complications Trial (DCCT), published in 1993, established hemoglobin A1c as the benchmark for assessing complication risks from type 1 diabetes. However, the increased usage of continuous glucose monitors has not yet been studied on a scale comparable to the DCCT, leading to uncertainties about CGM-based metrics becoming standard for evaluating diabetes management,” explained Boris Kovatchev, PhD, the director of the UVA Center for Diabetes Technology. “The absence of extensive, long-term CGM data presents various clinical and regulatory challenges; for instance, CGM results are not yet recognized as primary outcomes in diabetes medication studies.”
Utilizing Notable Diabetes Research Data
The DCCT collected hemoglobin A1c measurements from participants either monthly or quarterly, along with a blood sugar profile every three months. This data can be accessed from the archives of the National Institute of Diabetes and Digestive and Kidney Diseases upon request.
By applying advanced machine learning techniques to analyze the data from the DCCT, researchers were able to simulate virtual continuous glucose monitor readings for all trial participants throughout the study.
They discovered that 14 days’ worth of information from these virtual monitors could predict diabetes complications nearly as well as hemoglobin A1c levels. Besides the time spent in a healthy blood sugar range of 70 to 180 mg/DL, additional continuous glucose monitor metrics, such as the duration spent in a “tight range” (from 70 to 140 mg/DL), as well as the time above thresholds of 140 mg/DL, 180 mg/DL, and 250 mg/DL, also accurately forecasted potential diabetes complications.
With continuous glucose monitors now widely used by individuals with diabetes, these findings may assist patients in managing their condition while also aiding researchers in the ongoing improvement of diabetes treatment.
“Conducting a study equivalent in scale to the DCCT using continuous glucose monitoring alongside hemoglobin A1c would be prohibitively costly and time-consuming,” Kovatchev noted. “Using advanced data science techniques to create virtual simulations of a clinical trial allows us to bridge gaps in historical sparse data, representing the best alternative available to us at this time.”
