Continuous glucose monitor data predicts type 1 diabetes complications

BORIS KOVATCHEV and PHD are the directors of the UVA Diabetes Technology Center.

Virginia University Diabetes Researcher was discovered that data from continuous glucose monitors could predict nerves, eyes, and kidney damage caused by type 1 diabetes. This suggests that doctors can use the device data to save patients from blindness, diabetic neuropathy, and diabetes complications that change other life.

Patients have a safe blood sugar range of 70 to 180 mg/dl over 14 days, as well as standard approaches using hemoglobin A1c levels, forecasting neuropathy, retinopathy, and nephropathy. I found it as good as it was. 。

“The 1,440 diabetes -controlling test (DCCT), which was released in 1993, has established Hemoglobin A1C as a gold standard to evaluate the risk of type 1 diabetes. The use of glucose monitoring is increasing, and there is no study of DCCT size that checks CGM -based metric as a standard for evaluating diabetes control. ” “Lack of long -term large -scale large -scale CGM data has many clinical and coordinating meanings. For example, CGM has not been accepted as a major result from diabetes research.”

Use of landmark diabetes data

DCCT acquired hemoglobin A1C measurements from participants every month or every three months, and gained a blood sugar profile every three months. Data is an archive of national diabetes, digestive and kidney disease, and can be used in response to requests.

Using advanced machine learning methods for processing DCCT datasets, researchers were able to create virtual continuous glucose monitor traces for all participants and create them over the exam. 。

Researchers have found that 14 -day data from virtual continuous glucose monitor has the ability to predict the complications of diabetes as a measured hemoglobin A1c. In addition to the time spent within the safe blood sugar range of 70-180 mg/dl, researchers have found that the measured values ​​of other continuous glucose monitors also accurately predict the complications of diabetes. These measured values ​​include the time spent in “tight range” (70-140 mg/DL) and time exceeding 180 mg/dl and exceeding 140 mg/dl, exceeding 250 mg/dl. It was.

The continuous glucose monitor used by diabetic patients can help patients manage diabetes and to continue to care for diabetes.

“In addition to Hemoglobin A1C, research on DCCT -monitoring performed on continuous glucose monitoring will take time and expensive,” says Kovatchev. “Virtualization of clinical trials to fill the gaps of old -fashioned data using advanced data science methods is the best thing we can do today.”

Announced survey results

The results of the research are so Disclosure Journal diabetes and treatment. The author of this article is Benjamin Robo, Kiara Fabris, Mohamadreza Ganju, Anas El Fati, Mark D. Breton, Lauren Kanapka, Craig Colman, Tedge Colman, Roy W. Beck, Kobachev is. Disclosure from researchers can be found in a dissertation.