pedi13299-sup-0001-Supinfo.docxWord 2007 document , 248.5 KB

Supplementary Table S1 Characteristics of the study population according to gender, age and pubertal status.

Supplementary Table S2. Laboratory measured HbA1c, mean glucose concentration calculated from CGM data collected from the 4 weeks preceding the HbA1c measurement and the corresponding GMI value and HbA1c-GMI discordance of the entire study population and of subgroups of patients according to the type of CGM device used and insulin therapy administration.

Supplementary Table S3. Laboratory measured HbA1c, mean glucose concentration calculated from CGM data collected from the 4 weeks preceding the HbA1c measurement and the corresponding GMI value and HbA1c-GMI discordance of the entire study population and of subgroups of patients according to gender, age and pubertal status.

Supplementary Table S4 . Binary logistic regression analysis. Dependent variables: presence of HbA1c-GMI discordance ≥0.1, ≥0.5 and ≥ 1.0. Independent variables: age, gender, BMI (continuous), pubertal status (prepubertal, pubertal, postpubertal), type of CGM device (isCGM vs. rtCGM), type of insulin therapy administration (MDI vs. CSII), hemoglobin (continuous), anemia (present vs. absent), autoimmune diseases (present vs. absent).

Supplementary Table S5. Multiple regression analysis. Dependent variables: presence of HbA1c-GMI discordance with GMI calculated from the mean glucose value of CGM data collected from the 4 weeks preceding the HbA1c measurement. Independent variables: age, gender, BMI (continuous), pubertal status (prepubertal, pubertal, postpubertal), type of CGM device (isCGM vs. rtCGM), type of insulin therapy administration (MDI vs. CSII), hemoglobin (continuous), anemia (present vs. absent), autoimmune diseases (present vs. absent). R2 = 0.011;p = 0.549.

Supplementary Table S6. Binary logistic regression analyses. Dependent variables: presence of HbA1c-GMI discordance ≥0.1, ≥0.5 and ≥ 1.0 with GMI calculated from the mean glucose value of CGM data collected from the 4 weeks preceding the HbA1c measurement. Independent variables: age, gender, BMI (continuous), pubertal status, type of CGM device (isCGM vs rtCGM), type of insulin therapy administration (MDI vs CSII), hemoglobin (continuous), anemia (present vs absent), autoimmune diseases (present vs absent).

Supplementary Table S7. The absolute value of HbA1c-GMI discordance and the proportion of patients who were discordant at the second visit calculated in a subsample composed by 271 patients compared to the HbA1c-GMI discordance with GMI calculated from the mean glucose value of CGM data collected for the overall study population from the 12 weeks preceding the HbA1c measurement of the enrolment visit.

Supplementary Table S8. GMI values (% and mmol/mol) calculated for various CGM-derived mean glucose concentrations (mg/dL and mmol/L) according to the regression equation derived from data of the entire study population.

Supplementary Figure S1. Bland–Altman Plot with HbA1c-GMI discordance plotted on the y-axis and the average of HbA1c and GMI value plotted on the x-axis. Regression equation: HbA1c-GMI discordance (%) = 1.340 (95%CI 0.963–1.718) - 0.0178 (95%CI -0.228 – −0.129) x (average of HbA1c and GMI).