• Continuous glucose monitor-measured mean glucose and time above range may increase 10-year predicted cardiovascular disease risk.
• CVD risk was highest for adults with high glycemic burden and high variability.

Continuous glucose monitoring may be used to predict cardiovascular disease risk among adults without prevalent type 2 diabetes or CVD, according to findings published in Diabetes Care.

In an analysis of CGM data from 1,356 people without diabetes who participated in the Framingham Heart Study, researchers examined associations between both glycemic measures and CGM profiles and 10-year CVD risk. Bahar Bakhshi, MS, a PhD candidate in molecular and translational medicine at Boston University School of Medicine, said the study showed how glycemic patterns found in CGM data could provide additional information into a person’s cardiometabolic risk.

“CGM phenotyping may improve diabetes prevention and further enhance cardiometabolic risk stratification in individuals without diabetes,” Bakhshi told Healio. “However, additional large-scale longitudinal investigations are needed to confirm these findings.”

Adults without type 2 diabetes or CVD underwent an oral mixed-meal tolerance test and wore a Dexcom G6 CGM for up to 10 days (55.8% women; mean age, 59.3 years). Ten-year predicted CVD risk was calculated using the American Heart Association’s PREVENT equations. Adults with a 10-year risk of less than 5% were considered low risk for CVD; 5% to less than 7.5% was defined as medium risk, and 7.5% or higher was categorized as high risk.

Increased 2-hour postprandial plasma glucose, mean glucose and CGM-measured time above range with glucose of more than 140 mg/dL were associated with higher 10-year predicted CVD risk. Two CGM composite scores, J-index and GRADE, were also tied to increased 10-year predicted CVD risk.

Higher 2-hour postprandial plasma glucose, mean glucose, time above range with glucose of more than 140 mg/dL and three CGM composite scores were associated with higher odds of having a 7.5% or higher predicted CVD risk compared with having a less than 5% CVD risk. There were no glycemic measures linked to higher odds for medium CVD risk of 5% to less than 7.5% compared with low risk of less than 5%.

In a secondary analysis, higher 2-hour postprandial plasma glucose, CGM measures of glucose and three CGM composite scores were associated with increased levels of fasting triglycerides, HDL cholesterol and triglyceride-glucose index, as well as greater odds for hypertension and dyslipidemia.

Researchers developed four phenotypes based on glycemic excursions and time above range with glucose of more than 140 mg/dL: low glycemic burden and variability (n = 491), low glycemic burden and high variability (n = 188), high glycemic burden and low variability (n = 187) and high glycemic burden and high variability (n = 490). The high glycemic burden/high variability group had a worse cardiometabolic risk profile than the other three groups. Additionally, 10-year predicted CVD risk was higher for those with worse glycemic profiles (P < .001). Adults with high glycemic burden/high variability and those with high glycemic burden/low variability had 6% to 7% higher 10-year predicted CVD risk than adults with low glycemic burden and low variability.

Bakhshi said the researchers could not assess long-term cardiometabolic outcomes because the CGM data were collected from 2022 to 2025. Additionally, Bakhshi said, the glycemic profiles and data-driven approaches should be externally validated in future research.

“There are various different directions for the field to move forward, including but not limited to large-scale longitudinal analyses investigating CGM summary measures and glycemic risk profiles with type 2 diabetes and CVD development, in addition to identifying complementary multimodal phenotyping that may be required to fully leverage the potential clinical utility of CGM for characterizing glycemic heterogeneity and informing cardiometabolic risk,” Bakhshi said.

Bahar Bakhshi, MS, can be reached at baharb@bu.edu or LinkedIn @BaharBakhshi.