Continuous Glucose Monitors (CGMs), which monitor blood glucose every 15 minutes and can be used to diagnose diabetes and prediabetes, are rapidly becoming popular as health trackers in wearable technology.
A recent study found that with just 12 hours’ worth of glucose profile data, researchers could categorise Type 2 diabetes, prediabetes, and patients without impaired glucose tolerance. The research results were recently presented in New Orleans and Los Angeles at the 36th Conference on Neural Information Processing Systems (NeurIPS).
CGMs are very user-friendly and benefit diabetic patients who may take them home for routine insulin measurement, according to Dr. Richa Chaturvedi, Senior Consultant, Endocrinologist and Diabetologist at Indraprastha Apollo Hospital in New Delhi.
Dr. Chhavi Agrawal, Associate Consultant, Endocrinologist, Fortis Escorts, Delhi, claims that CGMs have been proven to be a clinically useful tool for minimising bouts of hypo- and hyperglycemia in addition to monitoring blood sugar levels. She claims that using CGM has been proven to be successful in lowering HbA1c levels and minimising glycemic fluctuation. The monitor is currently used to monitor diabetic patients who are pregnant as well as Type 1 and Type 2 diabetics who require numerous daily injections. These subgroups of diabetes patients are “excellent candidates for CGM” because they “need more frequent monitoring of plasma glucose and likely to exhibit greater variations in blood glucose when compared to other diabetics.”
Dr. Agrawal goes on to say that as CGM accuracy has increased, so has the requirement for routine calibration. For patients receiving insulin infusions, the monitors are now employed in hospital management, particularly in the intensive care unit.
The CGM device offers a real-time record of blood sugar levels as well as trends in blood sugar levels in relation to diet and daily activities. As a result, it aids in forecasting how blood glucose would react to different foods or activities, helping both the user and the treating physician gain a better understanding of diabetes, according to the expert.
Data from 436 Indian participants in the study were available. Each participant supplied information on their sex, age, and body mass index while using a CGM device on average for 12 days (BMI). A1C readings of 6.5 percent and above were classified by the researchers as Type 2 diabetes, 5.5 to 6.5 percent as prediabetes, and less than 5.5 percent as healthy. There were 177 participants who were healthy, 87 who had prediabetes, and 172 who had Type 2 diabetes. Doctors confirmed the diagnoses.
Based on various blood glucose level time periods, the researchers developed AI prediction models. They contrasted models based on data windows of 12, 24, 72, 168, and 288 hours.