Accuracy of continuous glucose monitoring during differing exercise conditions

Depending on intensity, exercise may induce a strong hormonal and metabolic response, including acid-base imbalances and changes in microcirculation, potentially interfering with the accuracy of continuous glucose monitoring (CGM). The present study aimed at comparing the accuracy of the Dexcom G4 Platinum (DG4P) CGM during continuous moderate and intermittent high-intensity exercise (IHE) in adults with type 1 diabetes (T1DM).

Ten male individuals with well-controlled T1DM (HbA1c 7.0 ± 0.6% [54 ± 6 mmol/mol]) inserted the DG4P sensor 2 days prior to a 90 min cycling session (50% VO2peak) either with (IHE) or without (CONT) a 10s all-out sprint every 10 min. Venous blood samples for reference glucose measurement were drawn every 10 min and euglycemia (target 7 mmol/l) was maintained using an oral glucose solution. Additionally, lactate and venous blood gas variables were determined.

Mean reference blood glucose was 7.6 ± 0.2 mmol/l during IHE and 6.7 ± 0.2 mmol/l during CONT (p<0.001). IHE resulted in significantly higher levels of lactate (7.3 ± 0.5 mmol/l vs. 2.6 ± 0.3 mmol/l, p<0.001), while pH values were significantly lower in the IHE group (7.27 vs. 7.38, p=0.001). Mean absolute relative difference (MARD) was 13.3 ± 2.2% for IHE and 13.6 ± 2.8% for CONT suggesting comparable accuracy (p=0.90). Using Clarke Error Grid Analysis, 100% of CGM values during both IHE and CONT were in zones A and B (IHE: 77% and 23%; CONT: 78% and 22%).

The present study revealed good and comparable accuracy of the DG4P CGM system during intermittent high intensity and continuous moderate intensity exercise, despite marked differences in metabolic conditions. This corroborates the clinical robustness of CGM under differing exercise conditions.

ClinicalTrials.gov NCT02068638.