- βAverage glucose from CGM predicted HbA1c with 83% accuracy in children with type 1 diabetes
- βThe most recent 4 weeks of CGM data were stronger predictors than the full 12-week period
- βCombining average glucose with time-above-range created the most reliable HbA1c prediction model
| Journal | Annals of pediatric endocrinology & metabolism |
| Year | 2026 |
| Authors | Lee, Yang, Kim |
| PMID | 41260209 |
What Was Studied β and Why It Matters
For children and teens living with type 1 diabetes, the HbA1c blood test has long been the gold standard for measuring average blood sugar over roughly three months. But HbA1c has real limitations β it's a single snapshot taken at a clinic visit, it can miss dangerous highs and lows, and it doesn't reflect day-to-day patterns. Continuous glucose monitors (CGMs) capture all of that detail in real time.
Researchers in Korea wanted to know: can the data already streaming from a child's CGM device actually predict what their HbA1c result will be β before they even walk into the clinic? Their findings, published in Annals of Pediatric Endocrinology & Metabolism, suggest the answer is a strong yes.
What the Researchers Did
The team analyzed CGM records from 85 children and teenagers (ages 2β18) with type 1 diabetes, all wearing Dexcom G6 or G7 sensors. They divided the 12 weeks of CGM data before each HbA1c test into five time windows and looked at several key CGM metrics, including:
- Time-in-Range (TIR): How often blood sugar stayed in the healthy target zone
- Time-Above-Range (TAR): How often blood sugar ran too high
- Average Glucose: The overall mean blood sugar reading
- Coefficient of Variation: A measure of how much blood sugar bounced around
What They Found
The most important discovery was about timing: CGM data from just the 4 weeks before an HbA1c test was a better predictor of the result than data from the full 12 weeks. Recent glucose trends matter more than older ones.
Among all the metrics tested, average glucose was the single strongest predictor of HbA1c, explaining 83% of the variation in results. When average glucose was combined with time-above-range, the prediction model became even more accurate and reliable.
Time-in-range also consistently showed that higher TIR was linked to lower (better) HbA1c β reinforcing what many diabetes care teams already encourage families to focus on.
What This Means for Your Child's Care
This research has practical implications for families managing pediatric type 1 diabetes every day:
- Your CGM data is powerful. The numbers on your child's CGM app aren't just for daily decisions β they're a running preview of the HbA1c result ahead.
- The last month matters most. If your child has had a rough stretch, there's still time before the next clinic visit to bring average glucose down and improve outcomes.
- Fewer surprises at the clinic. Families and care teams can use recent CGM trends to anticipate HbA1c results and adjust insulin plans proactively β rather than reacting after the fact.
Keeping your child's CGM supplies consistent and reliable is essential to capturing this data. MDS Diabetes carries CGM sensors, transmitters, and accessories to help ensure there are no gaps in monitoring.
Bottom Line
In children with type 1 diabetes, the CGM data from the 4 weeks before an HbA1c test β particularly average glucose and time spent running high β can predict that HbA1c result with remarkable accuracy. This means families and doctors don't have to wait for lab results to know how management is going. Real-time CGM data, reviewed consistently, is one of the most actionable tools available for keeping kids healthy long-term.
