King's College London

Research portal

Accuracy of Continuous Glucose Monitoring During Exercise in Type 1 Diabetes Pregnancy

Research output: Contribution to journalArticlepeer-review

Kavita Kumareswaran, Daniela Elleri, Janet M. Allen, Karen Caldwell, Marianna Nodale, Malgorzata E. Wilinska, Stephanie A. Amiel, Roman Hovorka, Helen R. Murphy

Original languageEnglish
Article numberN/A
Pages (from-to)223-229
Number of pages7
JournalDiabetes Technology and Therapeutics
Issue number3
PublishedMar 2013

King's Authors


Background: Performance of continuous glucose monitors (CGMs) may be lower when glucose levels are changing rapidly, such as occurs during physical activity. Our aim was to evaluate accuracy of a current-generation CGM during moderate-intensity exercise in type 1 diabetes (T1D) pregnancy.

Subjects and Methods: As part of a study of 24-h closed-loop insulin delivery in 12 women with T1D (disease duration, 17.6 years; glycosylated hemoglobin, 6.4%) during pregnancy (gestation, 21 weeks), we evaluated the Freestyle Navigator (R) sensor (Abbott Diabetes Care, Alameda, CA) during afternoon (15:00-18:00 h) and morning (09:30-12:30 h) exercise (55 min of brisk walking on a treadmill followed by a 2-h recovery), compared with sedentary conditions (18:00-09:00 h). Plasma (reference) glucose, measured at regular 15-30-min intervals with the YSI Ltd. (Fleet, United Kingdom) model YSI 2300 analyzer, was used to assess CGM performance.

Results: Sensor accuracy, as indicated by the larger relative absolute difference (RAD) between paired sensor and reference glucose values, was lower during exercise compared with rest (median RAD, 11.8% vs. 18.4%; P <0.001). These differences remained significant when correcting for plasma glucose relative rate of change (P <0.001). Analysis by glucose range showed lower accuracy during hypoglycemia for both sedentary (median RAD, 24.4%) and exercise (median RAD, 32.1%) conditions. Using Clarke error grid analysis, 96% of CGM values were clinically safe under resting conditions compared with only 87% during exercise.

Conclusions: Compared with sedentary conditions, accuracy of the Freestyle Navigator CGM was lower during moderate-intensity exercise in pregnant women with T1D. This difference was particularly marked in hypoglycemia and could not be solely explained by the glucose rate of change associated with physical activity.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454