Researchers at the University of Virginia Center for Diabetes Technology have discovered that data from continuous glucose monitors can be utilized to anticipate nerve, eye, and kidney damage instigated by type 1 diabetes. This discovery implies that medical professionals might use the data from these devices to prevent patients’ sight loss, diabetic neuropathy, and other severe diabetes complications.
The study revealed that the length of time patients’ blood sugar levels remained within the safe range of 70 to 180 mg/DL over a fortnight was as effective in predicting neuropathy, retinopathy, and nephropathy as the traditional method using hemoglobin A1c levels.
Historically, the Diabetes Control and Complications Trial (DCCT), a significant ten-year-long study involving 1,440 individuals, published in 1993, endorsed hemoglobin A1c as the benchmark for determining the risk of complications from type 1 diabetes. However, Boris Kovatchev, PhD, director of the UVA Center for Diabetes Technology, notes that the use of continuous glucose monitoring is increasingly prevalent, but there are no studies as comprehensive as the DCCT to confirm CGM-based metrics as the standard for assessing diabetes control. There are clinical and regulatory implications due to the absence of extensive, long-term CGM data; for instance, CGM is not yet accepted as a primary outcome from diabetes drug studies.
The DCCT collected hemoglobin A1c readings from participants either monthly or quarterly, along with a blood sugar profile every three months. This data is accessible in the National Institute of Diabetes and Digestive and Kidney Diseases’ archives. The researchers managed to create virtual continuous glucose monitor traces for all participants and for their entire participation duration in the trial by applying advanced machine learning techniques to the DCCT data sets.
The researchers discovered that data from the virtual continuous glucose monitors for 14 days predicted diabetes complications as effectively as hemoglobin A1c readings. The findings also revealed that other readings from the continuous glucose monitor, such as the time spent in “tight range” (70-140 mg/DL), and the time spent above 140 mg/DL, 180 mg/DL, and 250 mg/DL, accurately predicted diabetes complications.
These findings could be instrumental in helping patients manage their diabetes and aid researchers in furthering diabetes care, considering that continuous glucose monitors are now frequently used by patients with diabetes. Kovatchev points out that conducting a study as extensive as the DCCT with continuous glucose monitoring in addition to hemoglobin A1c would be both time-consuming and costly. He suggests that utilizing advanced data science methods to virtualize a clinical trial and fill in the gaps in old, sparse data is the most feasible approach currently.
Comments are closed for this post.