Reproducibility of the Rotor 2INpower Crankset for Monitoring Cycling Power Output: A Comprehensive Analysis in Different Real-Context Situations.

Abstract

To examine the reproducibility (intradevice and interdevice agreement) of the Rotor 2INpower device under a wide range of cycling conditions. Twelve highly trained male cyclists and triathletes completed 5 cycling tests, including graded exercise tests at different cadences (70-100rpm), workloads (100-650W), pedaling positions (seated and standing), and vibration conditions (20-40Hz) and an 8-second maximal sprint (>1000W). An intradevice analysis included a comparison between the power output registered by 3 units of Rotor 2INpower, whereas the power output provided by each one of these units and the gold-standard SRM crankset were compared for the interdevice analysis. Among others, statistical calculations included the standard error of measurement, expressed in absolute (in watts) and relative terms as the coefficient of variation (CV). Except for the graded exercise test seated at 100 rpm/100W (CV = 10.2%), the intradevice analysis showed an acceptable magnitude of error (CV ≤ 6.9%, standard error of measurement ≤ 12.3W) between the 3 Rotor 2INpower. Similarly, these 3 units showed an acceptable agreement with the gold standard in all graded exercise test situations (CV ≤ 4.0%, standard error of measurement ≤ 13.1W). On the other hand, both the intradevice and interdevice agreements proved to be slightly reduced under high cadences (intradevice: CV ≤ 10.2%; interdevice: CV ≤ 4.0%) and vibration (intradevice: CV ≤ 4.0%; interdevice: CV ≤ 3.6%), as well as during standing pedaling (intradevice: CV ≤ 4.1%; interdevice: CV ≤ 2.5%). Although within the limits of an acceptable agreement, measurement errors increased during the sprint tests (CV ≤ 7.4%). Based on these results, the Rotor 2INpower could be considered a reproducible tool to monitor power output in most cycling situations.