A critique of RPE as a basis of exercise prescription

Abstract

I read with interest the recent article of Scherr et al. (2013) examining the relationships between Borg’s 6–20 scale rating of perceived exertion (RPE), ear-lobe lactate concentrations and heart rates in a large sample of sedentary and athletic adults with an average age of 28 years. The authors suggested that ‘‘moderate exercise is performed at an RPE of 11–12’’ and they concluded that ‘‘Borg’s RPE can be seen as a valid and inexpensive tool for monitoring exercise intensity for primary and secondary disease prevention.’’ Some previous investigators, also, have suggested that an RPE of 11–14 provides a useful basis for the prescribing of therapeutic exercise. Looking at the graph relating heart rate to RPE (Fig. 2) for the subjects studied by Scherr et al. (2013), their conclusion seems reasonable enough in terms of the average response. At an RPE of 11, the heart rate averages about 137 beats min, and at an RPE of 12 the average is about 146 beats min. Using the heart rate reserve method (Karvonen et al. 1957), such figures would correspond to about 55 and 62 % of the individual’s maximal oxygen intake. However, as with most of the prediction procedures that are used in clinical physiology, it is unwise to assess validity based upon the average response; advice must be given to the individual patient. Let us suppose that a person is undergoing cardiac rehabilitation, and the clinician has stressed that it would be unwise for the intensity of therapeutic exercise to exceed 65 % of the patient’s maximal oxygen consumption. If the prescription is based upon advice to exercise at an RPE of 11–12, then on average, the oxygen consumption will fall nicely within the desired range, at 55–62 % of maximal. However, a closer examination of Fig. 2 shows that individual heart rates cover a broad range. If the graph shows standard deviations, as appears to be the case from a statement on p. 154, then 5 % of individuals who are at an RPE of 12 would have been exercising either at a homeopathic heart rate of less than 100 beats min or at a dangerously high rate of 192 beats min, close to maximal aerobic effort. Moreover, RPE-based prescriptions would become even more tenuous if an attempt was made to apply information collected during standardized laboratory exercise on a cycle ergometer or treadmill to free activity in the community (Ceci and Hassmen 1991). It might be argued that fallibility of the RPE-based predictions is somewhat exaggerated by the present analysis, since the sample of Scherr et al. (2013) included both athletic and sedentary individuals. But when their sample was sub-divided on the basis of athletic involvement, as in their Fig. 3, the 95 % range of heart rates predicted for an RPE of 12 remained almost equally broad (92–180 beats/ min). The range of ages of their subjects was substantial (17–44 years), and some narrowing of the range of predictions might have been obtained if the reduction of maximal heart rate with age had been taken into consideration; unfortunately, this was not done. One might also wonder if greater validity would have been observed if the RPE prescription had been related to direct measures of oxygen consumption, rather than to heart rates. However, Shephard et al. (1996) reported that when exercising at an RPE of 13, directly measured oxygen consumption values (expressed as a percentage of maximal aerobic effort) continued to show a standard deviation of 12 %, still much too broad a range to be of help in prescribing exercise for Communicated by Klaas Westerterp/Hakan Westerblad.