Background: The A1c-Derived Average Glucose (ADAG) study was a multicenter observational study that used continuous glucose monitoring and seven-point daily self-monitored blood glucose (SMBG) to determine the relationship between glycated hemoglobin (HbA1c) and mean average glucose. The estimated average glucose in mg/dL was calculated by the formula 28.7xHbA1c(%)-46.7. Using the SMBG results from 378 ADAG study participants, it was determined the average fasting blood glucose (BG) for predefined target HbA1c groups between 5.5-8.5%.
Aims: To compare fasting plasma glucose (FPG) and HbA1c measurements (venous sample) from an outpatient population followed in our Hospital with SMBG results (capillary sample) from ADAG study.
Method: We analyzed data from nonpregnant adults between 18 and 70 years old and HbA1c values between 5.5-8.5% (sample characteristics similar to ADAG study). FPG was performed by an enzymatic-ultraviolet assay and HbA1c by ionic exchange high-performance liquid chromatography, according to international standards.
Results: We included 10520 patients, 53.6% female, mean age 55.97±10.93years, Hb 13.70±1.70g/dL, HbA1c 6.33±0.80% and FPG 117.21±39.49mg/dL. In our sample, the mean FPG in patients with HbA1c levels of 5.5-6.49%, 6.5-6.99%, 7.0-7.49%, 7.5-7.99% and 8.0-8.5% was 101.77mg/dL (95%CI 101.28-102.27), 130.33mg/dL (95%CI 128.40-132.26), 141.65mg/dL (95%CI 139.01-144.30), 158.20mg/dL (95%CI 154.38-162.01) and 177.18mg/dL (95%CI 172.39-181.97), respectively. FPG were significantly lower in female when HbA1c levels were between 5.5-7.49% (p<0.05), but sex distribution in our sample was similar to ADAG study (54% female). In fasting state, mean venous plasma glucose from our population was always lower comparing with mean capillary BG from ADAG study. The difference between our results and ADAG study was 20.23mg/dL, 11.67mg/dL, 10.35mg/dL, 8.8mg/dL and 0.82mg/dL when HbA1c levels range from 5.5-6.49%, 6.5-6.99%, 7.0-7.49%, 7.5-7.99% and 8.0-8.5%, respectively. The increased in the HbA1c group diminished the discrepancy between our results and the comparative study. The reason for the discrepant results is not completely understood, but may due to different measurement methods and eventual differences between populations.
Discussion: Several factors may interfere with glucose and HbA1c measurements. Our results highlight the difficulty in comparing glucose levels from different measuring procedures. Empirical data may help patients and physicians to optimize glucose monitoring targets.

Fasting plasma glucose and HbA1c measurement: a comparative analysis with A1c-derived average glucose study