Validity and reliability of the G-Cog device for kinematic measurements.

The aim of this study was to test the validity and the reliability of the G-Cog which is a new BMX powermeter allowing for the measurements of the acceleration on X-Y-Z axis (250 Hz) at the BMX rear wheel. These measurements allow computing lateral, angular, linear acceleration, angular, linear velocity and the distance. Mechanical measurements at submaximal intensities in standardized laboratory conditions and during maximal exercises in the field conditions were performed to analyse the reliability of the G-Cog accelerometers. The performances were evaluated in comparison with an industrial accelerometer and with 2 powermeters, the SRM and PowerTap. Our results in laboratory conditions show that the G-Cog measurements have low value of variation coefficient (CV=2.35%). These results suggest that the G-cog accelerometers measurements are reproducible. The ratio limits of agreement of the rear hub angular velocity differences between the SRM and the G-Cog were 1.010 × ÷ 1.024 (95%CI=0.986-1.034) and between PowerTap and G-Cog were 0.993 × ÷ 1.019 (95%CI=0.974-1.012). In conclusion, our results suggest that the G-Cog angular velocity measurements are valid and reliable compared with SRM and PowerTap and could be used to analyse the kinematics during BMX actual conditions.