Introduction
Regular exercise by persons with diabetes mellitus (DM) contributes to physiological and health benefits on the short and long term.1–3 Maintaining acceptable glucose concentrations during sports activities is challenging, however. The more intense the sport activity and the more extreme the conditions, the larger the challenge of balancing energy intake and insulin use. Therefore, particularly during exercise, timely and accurate blood glucose readings are of importance to enhance chances on good metabolic control, allowing maximal performance.
During strenuous circumstances, an alternative to capillary self-measurement of blood glucose (SMBG) levels performed is interstitial glucose measurements. At present, real-time continuous glucose monitoring (rt-CGM) and flash glucose monitoring (FGM) are available as methods for interstitial glucose measurements. If used appropriately, with rt-CGM the user often only needs to calibrate the sensor at some time points daily and being able to act and react on alarms preset at certain cut-off points. In contrast, FGM users use a factory-calibrated sensor, which with the presently available version in The Netherlands needs scanning for obtaining the required information.4
Some authors report qualitatively comparable glucose results of rt-CGM and FGM,5 although others observed differences in accuracy.6 However, these comparative studies are virtually always done under routine or standardized clinical circumstances. With such standardized testing, it can be assumed that factors such as interstitial fluid dynamics and interstitial glucose concentrations will behave more or less in a stable pattern during the testing periods. In contrast, real-life (intensive) exercise most likely will lead to more frequent and profound changes in interstitial fluid dynamics and thus in interstitial glucose concentration.7 Therefore, performance of interstitial glucose measurements might be quite different during intensive exercise conditions as compared with a performance during normal daily activities (NDA). Furthermore, depending on the technique used to measure interstitial glucose concentrations, the change in interstitial fluid dynamics might also affect reading outcomes differently.
In an earlier study, we found that during NDA and standardized tests FGM reading showed lower glucose concentrations than actually present with capillary measurements in the lower ranges, and higher than found in the higher ranges.8 In contrast, FGM performance proved to be different during intensive exercise in a real-life setting, showing comparatively higher glucose readings than actually present, but also less accuracy throughout the whole measurement spectrum.9 The rt-CGM performance during exercise in that study also was less accurate than hoped for. Both the FGM and rt-CGM devices used in that study used a glucose oxidase based method for measuring glucose.10 11
Recently, an implantable rt-CGM device has been introduced (Eversense), using fluorescence-based glucose measurement technique,12 with an accuracy comparable to other rt-CGM devices.13 The Eversense is an implantable rt-CGM device, needs calibrating twice daily, and ideally functions for 6 months. The Free Style Libre (FSL) FGM system has a transcutaneous probe, is factory calibrated and functions for 14 days.
Given the importance of reliable glucose measurements during intensive exercise, we aimed to compare in a real-life setting the accuracy of two interstitial devices that use both a different method (glucose oxidase based vs fluorescence based) to measure glucose in persons with type 1 diabetes mellitus (T1DM). Furthermore, since differences in accuracy may differ depending on the circumstances and activity level,14 measurements were also performed during a period of NDA with no sports activities in the week after the mountain biking.