Amplitude normalisation for intramuscular fine-wire electromyography of the leg during cycling

Electromyographic (EMG) signal amplitude is typically normalised to reference values obtained from maximal (MVC) or submaximal voluntary isometric contractions (sMVC), or the maximum measured EMG amplitude (MAX). Guidelines for amplitude normalisation of EMG recordings from the leg during dynamic tasks are incomplete. The purpose of this study was to determine the most suitable method of amplitude normalisation for intramuscular fine-wire electromyographic (fEMG) data of the leg during cycling. To do this we compared the statistical sensitivity of fEMG data of cycling normalised to MVC, sMVC and MAX using an index of data variance, root mean square error (RMSE), and the least significant difference (LSD). Recordings normalised to MVC and sMVC reference values were highly variable between cyclists (RMSE MVC 91.2–271.8%, sMVC 46.7–176.5%) and LSD was high (MVC 127.1–282.2%, sMVC 79.7–231.6%). Normalisation to MAX was least variable between cyclists (RMSE MAX 12.8–27.9%, p < 0.01) and LSD was less for MAX than for MVC and sMVC (17.01–32.22%, p < 0.01). Together with evidence of greater repeatability of data normalised to MAX from our previous studies, our findings of less population variability and LSD, and the resulting greater sensitivity of statistical comparisons, indicate that normalisation to the maximum measured EMG amplitude is most suitable for fEMG data of the leg during cycling