Long-term glycemic control and glucose variability assessed with continuous glucose monitoring in a pediatric population with type 1 diabetes: Determination of optimal sampling duration

Background: No studies have assessed if 2-week of continuous glucose monitoring (CGM) data provide good estimation of long-term glycemic control and glucose variability (GV) in pediatric patients with type 1 diabetes (T1D) as in adults. Methods: Six hundred fifty-four T1D pediatric patients were enrolled and 12-weeks of CGM data, before HbA1c measurement, were collected. Metrics of glycemic control and GV in incremental sampling periods were calculated. The agreement between metrics calculated in the sampling periods and the full 12-week period was assessed with correlation analysis (R2), median relative absolute difference (RAD) or absolute difference in the entire study populations and subjects stratified by age, pubertal status, insulin therapy (MDI,CSII), type of CGM (intermittently scanned [isCGM], real-time [rtCGM]), and HbA1c level. Results: Correlations with metrics of the full 12-week period improved by extending the sampling periods. R2 values close to 0.90 using 4-week period were significantly higher than 2-week period, particularly for coefficient of variation, mean glucose SD, percentage of time below the range <70 mg/dL. A significant difference was found comparing the median RAD of 2- and 4-week, especially for mean glucose and coefficient of variation. Similar results were obtained analyzing subjects according to age and pubertal status, whereas in patients with HbA1c ≤7%, using rtCGM and CSII significant correlations were found for 2-week period. Conclusions: In T1D pediatric subjects, 4-week CGM data better reflects long-term glycemic control and GV in MDI and isCGM users. The 2-week period may be acceptably accurate in CSII and rtCGM users, especially in those with good glycometabolic control.