Giuliana Da Prato, Italy

University and Hospital of Verona Endocrinology, Diabetes and Metabolism Department of Medicine

Presenter of 1 Presentation

ORAL PRESENTATION SESSION

ACCURACY OF RT-CGM SYSTEMS DURING CONTINUOUS AND INTERVAL EXERCISE IN ADULTS WITH TYPE 1 DIABETES

Abstract

Background and Aims

Real-time glucose monitoring systems (rt-CGMs) play an important role in the treatment decisions of subjects with type 1 diabetes (T1D), but their accuracy may be lower during exercise. We assessed the accuracy of several recent rt-CGMs in T1D subjects during both moderate continuous (CON) and interval training (IT) exercise.

Methods

In 22 patients, already using a rt-CGM, a second different sensor was applied. Participants performed, in a random order, a 30’ CON session, and a 30’ IT session. Data recorded by rt-CGMs were compared with the corresponding blood glucose values, measured by a glucose analyzer. The accuracy of rt-CGMs was assessed by the Sensor Bias (SB), the Mean Absolute Relative Difference (MARD) and the Clarke error grid.

Results

A total of 2355 plasma-sensor glucose paired points were collected. Both average plasma and interstitial glucose did not significantly differ during CON and IT. However, plasma glucose change at the end of exercise was greater during CON than during IT. During CON, the sensors overestimated plasma glucose more than during IT, as shown by both SB and MARD. Classifying the clinical performance according to the Clarke error grid, no differences were found between the exercise sessions and >97% of values were in the A+B zone.

Conclusions

In conclusion, continuous exercise affects rt-CGMs accuracy to a greater extent than interval training,likely due to larger acute glycemic variations. However, this phenomenon does not compromise clinical reliability of recent CGMs. Larger studies, should investigate a broader spectrum of activities in terms of intensity and type of exercise.

Hide