CGM Average Glucose to HbA1c Conversion
An average daily glucose of 105 mg/dL on CGM corresponds to an estimated HbA1c of approximately 5.2-5.3%. 1
Direct Calculation Using Standard Conversion
Based on the American Diabetes Association's established correlation table from the ADAG study, the relationship between mean glucose and HbA1c shows that:
- An HbA1c of 5% corresponds to a mean glucose of 97 mg/dL (95% CI: 76-120 mg/dL) 1
- An HbA1c of 6% corresponds to a mean glucose of 126 mg/dL (95% CI: 100-152 mg/dL) 1
Your CGM average of 105 mg/dL falls between these two reference points, placing the estimated HbA1c at approximately 5.2-5.3%. 1
Alternative Calculation Using GMI Formula
The Glucose Management Indicator (GMI) provides a more modern calculation method using the validated formula: GMI (%) = 3.31 + 0.02392 × [mean glucose in mg/dL]. 1, 2
However, this GMI formula was developed and validated in diabetic populations with higher glucose ranges, which explains why it yields a slightly higher estimate than the ADAG table for glucose values in the normal range. 2
Important Clinical Caveats
The ADAG table is more appropriate for your glucose level of 105 mg/dL because:
- The ADAG study included non-diabetic individuals and captured the full physiologic range of glucose values 1
- The GMI formula was specifically developed for diabetic populations using modern CGM technology and may overestimate HbA1c at near-normal glucose levels 2
- The correlation coefficient (r = 0.92) in the ADAG study was strong but still allows for individual variation 1
Key limitations to consider:
- The 95% confidence interval for any given mean glucose is wide—for a mean glucose around 100 mg/dL, the corresponding HbA1c could range from approximately 4.2% to 6.7% 1
- Individual factors affecting red blood cell turnover (hemoglobinopathies, anemia, recent blood loss) can cause discordance between CGM-derived estimates and laboratory HbA1c 1, 3
- The ADAG data were based on 2,700 glucose measurements over 3 months using early CGM technology combined with capillary glucose, so generalizability to modern CGM systems has some uncertainty 1
- Mismatches between CGM-derived estimates and laboratory HbA1c >40 mg/dL occur more than 5% of the time due to stable non-glycemic factors 3
Clinical Recommendation
For a CGM average of 105 mg/dL, report an estimated HbA1c of approximately 5.2%, with the understanding that the actual laboratory value could reasonably range from 4.8% to 5.6% based on individual biological variation. 1 If precise HbA1c determination is clinically important, obtain a laboratory measurement rather than relying solely on CGM-derived estimates. 1, 3