Prediction of performance reduction in self-paced exercise as modulated by the rating of perceived exertion.

purposeRating of perceived exertion (RPE) is a scale of exercise difficulty and has been hypothesized to be a regulator of work rate during self-pacing. The goal of this work was to develop a dynamic prediction of RPE and to characterize the control strategy employed to reduce work rate during self-paced exercise using RPE as feedback.

methodsTraining and test data were acquired from the literature to develop a linear regression of RPE as a function of four physiological variables: core temperature, mean-weighted skin temperature, metabolic rate, and integral of relative oxygen consumption (R (2) = 0.85). A thermoregulatory model was used to predict core and mean-weighted skin temperature. Utilizing self-paced cycling and running data from the literature, we characterized reductions in work rate with a proportional-derivative control algorithm with RPE as feedback.

resultsBland-Altman analysis revealed the necessity to parameterize RPE equations for untrained and endurance-trained individuals. Afterwards, dynamic predictions of RPE were accurate for a wide range of activity levels and air temperatures for walking, running, and cycling (LoA and bias of 2.3 and -0.03, respectively). For self-paced exercise, the control algorithm characterized the trend and magnitude of work rate reductions for cycling and running, and showed regulated RPE to be less conservative for shorter vs. longer duration exercise.

conclusionsA novel methodology to characterize self-paced work intensity, based upon dynamic physiologic response, is provided. The complete model is a useful tool that estimates performance decrements during self-paced exercise and predicts tolerance time for exhaustive fixed-rate exercise.