There is a growing interest in the analysis of different methods for monitor fatigue during resistance training sessions. This study aimed to (1) analyse the relationships between the percentage of performed repetitions with respect to the maximum possible number (%REP), RPE and magnitude of velocity loss (VL), and (2) examine whether a multiple regression analysis with the RPE and VL as predictor variables could improve the goodness of fit to predict %REP in the bench press exercise performed in a Smith machine. Seven men performed a repetition maximum test, on 3 separate testing sessions, against 3 different absolute loads based on a target mean velocity (MV) according to an individual load-velocity profile (≈1.00, ≈0.70, and ≈0.50 m/s). MV, VL, %REP and RPE were collected and used for analysis. Based upon quadratic polynomial regression analysis strong relationships were reported between the RPE and %REP (r2 = 0.89 and SEE = 9.85%) and between the VL and %REP (r2 = 0.91 and SEE = 9.85%). Multiple regression analysis with the RPE and VL as predictor variables improved the goodness of fit (r2 = 0.94 and SEE = 7.18%) of the model to predict %REP. These results suggest that both RPE and VL are useful variables to accurately estimate %REP in the bench press exercise.
Keywords: Velocity-based training; exercise prescription; level of effort; subjective scale.