Scuba diving is today considered a possible activity in T1D following an appropriate protocol, while the performance of most recently available rtCGM systems has not yet been tested in this setting. Aim of this study is to verify if prolonged and consecutives immersions in water and exposure to high and variable pressures can alter the subsequent performance of the sensor, in terms of accuracy.
11 T1DM subjects participating at the camp “Diabete Sommerso 2019” were enrolled (mean age 45yr, mean diabetes duration 22yr, mean A1c 7.5%). Subjects wore a rtCGM 24 hours before the first dive and used their own glucometer to check glycemia and take all therapeutic decisions.
We collected data of 88 dives (64 at depths <18mt, 24 at 18-30mt). For the assessment of rtCGM accuracy we compared 328 capillary glycaemic values to the corresponding CGM data using MARD and Pearson’s correlation and analysing first overall data, then different glucose ranges (TIR, TAR, TBR) and dive depth (<18 or 18-30mt): despite results shown an excellent Pearson correlation in overall data (R=0.85, p<0.05) and still a good one in subgroup analysis, the MARD values revealed a worsening of accuracy (overall MARD 13.4) compared with routine use.
In our experience rtCGM system used during multiple diving at depth <30mt has shown a good univariate correlation, despite a not optimal accuracy. It still could represent a valid aid to capillary controls needed to safely deal with diving in people with T1D, but it requires caution in interpreting results.