Current recommendations for type 2 diabetes treatment are based on targeting HbA1c depending on different factors. HbA1c reflects recent 8-12 weeks glucose control, but amount of data show HbA1c insufficiencies. Continuous glucose monitoring (CGM) is used for insulin dose adjustment but also for control assessment and treatment determination. In our study we compare seven-day non-real time CGM data and HbA1c to find out which is more appropriate for determining drug prescription in patients with type 2 diabetes. Our study includes 85 patients with type 2 diabetes (35 women, 50 men; age 43.93±10.87 years, disease duration 21.91±6.years) , on different treatment regimens (31 on oral therapy, 33 treated with premixed insulin, 21 on multiple insulin injections) . Patients performed Non-Real Time Continuous Glucose Monitoring by using iProTM for seven days and HbA1c was measured at the end of this period. No difference in HbA1c level was found in three observed groups. HbA1c showed moderate negative correlation to percentage of time within the limits (r1=-0.48; r2= -0.654; r3= -0.674) . Based on HbA1c assessment 34.84% of type 2 patients on oral medications, 36.36% of patients on premixed insulin and 19.05% of type 2 patients on multiple daily injection are with good control. CGM defined 54.84% of type 2 patients on oral medication, 18.27% of patients on premixed insulin and 17.80% of type 2 patients on intensified regimen as well controlled. CGM was more precise than HbA1c in defining patients with good control (p<0.05) . HbA1c shows bad correlation with overall glucose excursion (r1=-0.096, r2=0.200, r3=0.377) . No difference was found between studied groups in regard to CGM percentage of time spent above and below the limits and to AUC above upper limit. We conclude that CGM data give more precise information than HbA1c about glucose control and are more useful for defining treatment adjustment. Even performing CGM for short period helps physician to choose the most appropriate drug.


T.Totomirova: n/a. M.Arnaudova: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at