Exercise Duration Affects Session Ratings of Perceived Exertion as a Function of Exercise Intensity

Abstract

Session ratings of perceived exertion (sRPE) are considered a practical marker of whole session exercise intensity, but its relationship to exercise volume has remained unclear. We analyzed the effects of exercise duration at different intensities on overall and differentiated sRPE. Sixteen males (M_age = 22.6, SD = 2.2 years; M_height = 176.4, SD = 5.8 cm; M_weight = 74.0, SD = 5.9 kg; and M_{body fat} = 9.4, SD = 2.2%) performed 15 and 30 minute runs at speeds associated with RPE levels of two (weak), three (moderate) and five (strong) on Borg’s CR-10 scale during a previous graded exercise test. We used Foster’s scale to access sRPE 30 minutes after each trial. Significant increases in sRPE were found with increases in running speed (p < 0.01, η _G ² = 0.48) and duration (p < 0.01, η _G ² = 0.16), with a significant speed X duration interaction (p < 0.01, η _G ² = 0.10). In addition, there was a significant effect for sRPE type (p = 0.01, η _G ² = 0.05) in that overall sRPE was slightly lower than sRPE differentiated to legs and higher than sRPE differentiated to breathing through the trials. Changes in sRPE from 15 to 30-minute trials were minimal for the slow speed and weak sRPE (Cohen´s dz = 0.04 – 0.25) but got higher at the moderate (Cohen´s dz = 0.88 – 1.06) and strong (Cohen´s dz = 1.94 – 2.50) speeds and sRPEs. Thus, exercise duration affects sRPE in an intensity dependent manner. This finding has practical relevance for prescribing exercise, suggesting a need to target specific training loads or aims to optimize trainees’ retrospective perceptions of the exercise experience.

Keywords

session-RPE training load volume running exercise prescription

Introduction

Internal training load (ITL) can be defined as the degree of psychophysiological stress induced by a given dose of exercise, with its magnitude dependent on the interaction between the individual exerciser’s characteristics, environmental conditions and exercise features. Quantifying ITL is an essential step towards modeling dose-response relationships to exercise in a fashion that prevents injury/overtraining and improves both performance (Bourdon et al., 2017; Haddad et al., 2017) and health (Iellamo et al., 2014; Volterrani & Iellamo, 2016).

As a practical alternative to objective physiological measurements such as heart rate (HR), blood lactate and oxygen uptake, it has been suggested that ITL scores could be computed from exercisers’ retrospective perceptions of the completed exercise bouts, using the exercisers’ session ratings of perceived exertion (sRPE) multiplied by their total exercise duration (in minutes) (Foster, 1998; Foster et al., 1995, 2001). In this perspective, sRPE can be considered an extension of Borg’s on-task ratings of perceived exertion (RPE) (G. Borg, 2000; G. A. Borg, 1982). sRPEs provide a single intensity measure for the whole exercise bout, while exercise duration reflects the level of exercise volume.

Given its simple and inexpensive nature, sRPE data collection has become very popular in the last two decades, and it has been widely adopted by scientists, coaches and exercise practitioners for ITL quantification (Eston, 2012; Foster et al., 2017; Haddad et al., 2017). In addition, increasing interest has been devoted to sRPE as a marker of relative whole session intensity in non-athletic populations (Haile et al., 2013; Iellamo et al., 2014; Kilpatrick et al., 2009), with some authors advocating its potential role in public health settings and in making large scale decisions regarding future exercise behavior (Haile et al., 2013; Kilpatrick et al., 2009).

Based on the idea that differentiated central and peripheral on-task RPE might be associated with specific physiological changes during exercise (Pandolf, 1982; Robertson & Noble, 1997), studies differentiating sRPE into breathing and muscular components of an exercise bout have been proliferating in the scientific literature (Green et al., 2009; Green et al., 2011; Mann et al., 2019; McLaren et al., 2016; Ribeiro et al., 2013). The rationale for this approach is to go beyond procuring only an overall sRPE (sRPE-O) and gain more detailed information on perceptual signal dominance for a whole exercise session (Ribeiro et al., 2013). This method provides a more sensitive evaluation of training load (McLaren et al., 2016). Regardless of sRPE type, a logical assumptive premise of the sRPE method has been that retrospective perceptions are not inherently associated with exercise volume (Foster et al., 2001; Herman et al., 2006), otherwise sRPE per se could provide a satisfactory measure of training load without having to factor in the additional effects of prolonged duration or volume (Barroso et al., 2015; Green et al., 2009).

Data presented by Foster et al. (2001) on ITL scores for 12 volunteers following 30, 60 and 90 minutes of cycloergometry at 90% of individual anaerobic threshold, revealed that increases in exercise duration led to only slight increases in sRPE-O taken 30 minutes after exercise. Similarly, Green et al. (2009) subjected 10 volunteers to 20, 30 and 40 minutes of treadmill exercise at ∼ 70% VO₂max and found no significant differences between trials regarding sRPE-O and differentiated sRPE, despite a trend toward higher values of sRPE-O in longer exercise bouts. Although these results corroborated the idea that volume has little influence on sRPE, they led to suspicions that more demanding exercise paradigms (i.e., prolonged and/or more intense exercise) might yield new results (Green et al., 2009) and help explain the significant upward directions of sRPE-O that had been found with increased exercise volume during swimming (Barroso et al., 2015; Fusco et al., 2020). Nevertheless, given the single intensity research paradigms and different exercise modes adopted in the above studies, the question of whether there is an intensity-dependent effect of exercise duration on sRPE remains unanswered. Thus, our purpose in the present study was to analyze the effects of treadmill exercise duration at different intensities on both overall and differentiated sRPE among recreationally active participants. Given the evidence that physiological parameters and RPE increases during constant speed submaximal running in an intensity-dependent manner (Kilpatrick et al., 2009; Monteiro et al., 2019; Steed et al., 1994), and that volume can affect sRPE in some circumstances (Barroso et al., 2015; Fusco et al., 2020), we hypothesized that a duration effect would emerge on sRPE, but only after more, rather than less, intense running bouts.

Method

Experimental Approach

We used a crossover methodology, and our participants were healthy individuals who were randomly assigned to different exercise intervention groups. After measuring the clinical outcome, we changed the exercise sequence. In a within-subject repeated measures design, our independent variables were: (a) three treadmill speeds (weak, moderate and strong); (b) two trial durations (15 and 30 minutes); and (c) three different types of sRPE (overall, breathing and legs). Our dependent variables were on-task RPE, sRPE and HR values.

Participants

After the research protocol was approved by the university institutional ethics committee (Protocol # 29393814.0.0000.5526), 16 recreationally active healthy male students (M_age = 22.6, SD = 2.2 years; M_height = 176.4, SD = 5.8 cm; M_weight = 74.0, SD = 5.9 kg and M_{body fat} = 9.4, SD = 2.2%) volunteered and signed informed consent forms to participate in this study, conducted according to the national norms for research involving human participants (Resolution 466/12; December 12, 2012). Prior to participation, volunteers completed the Physical Activity Readiness Questionnaire (PAR-Q) and a medical history questionnaire. Study inclusion criteria were: (a) age 18–30 years, (b) negative responses to all questions on the PAR-Q, (c) absence of any known medical restriction to exercise, (d) no reported use of ergogenic and/or psychoactive drugs, and (e) self-reported engagement in at least 150 minutes of aerobic physical activity per week during the prior six months.

Procedures

Participants reported to the laboratory on nine occasions separated by 48 – 72 hours, with the whole protocol concluded within three – four weeks for a single volunteer. The first visit included anthropometric measurements by means of a calibrated weight scale with a stadiometer and a skinfold caliper, and a 10-minute supine rest period during which we used a HR monitor (Polar S610, Polar Electro, Kempele, Finland) to take a resting HR assessment. At the end of the first and again at the second visit, volunteers were carefully instructed regarding the proper use of perceived exertion scales and shown the instrument’s combined memory and exercise anchoring procedures (Haile et al., 2014). The following sessions included a maximal graded exercise test and six constant speed trials on a treadmill (Suprema Fit - Rotaxx, Brazil) with inclination fixed at 1%.

Experimental sessions were carried out at approximately the same time of the day (±1 hour) for each participant with ambient temperature kept at 20 – 22 °C. Volunteers were asked to use the same lightweight running kit, maintain hydration habits and eat a light meal at least three hours before each test. They also were instructed to avoid intense physical activity and alcohol for 24 hours, and caffeine for three hours prior to experimental sessions. On all occasions, there were warm-up and cool-down periods consisting of three minutes of jogging and walking, respectively.

Measures

Perceived Exertion and Heart Rate Measures

In each of the different experimental sessions, the Borg’s CR10 scale (G. Borg, 2000) and Foster’s scale (Foster et al., 2001) were employed for on-task RPE and sRPE evaluations, respectively. Borg’s CR10 is a category-ratio, single-item self-report scale for perceived exertion, with verbal anchors ranging from zero (Nothing at all) to 10 (Extremely strong), also enabling the respondent to indicate higher values by denoting “Absolute maximum” if onés perceived exertion is higher than previously experienced. Foster’s scale is an 11-point single-item scale with verbal anchors ranging from zero (Rest) to 10 (Maximal). Satisfactory validity and reliability of perceived exertion measures using such scales in different exercise paradigms have been previously attested (Haddad et al., 2017; Herman et al., 2006; Soriano-Maldonado et al., 2015). Both RPE and sRPE measures included overall perceptions (RPE-O and sRPE-O) and perceptions differentiated to legs (RPE-L and sRPE-L) and breathing (RPE-B and sRPE-B). For overall ratings, volunteers were instructed to focus on whole-body exertion instead of relying on any particular peripheral or central cues. On the other hand, for differentiated ratings participants were asked to separately focus on legs and chest/breathing perceptions. In addition to perceptual measures, five second average HR values were continuously recorded during all exercise sessions through the use of a previously cited heart rate monitor; these data were later transferred and analyzed in a dedicated software program (Polar Pro Trainer 5 for Windows, Polar Inc., Kempele, Finland).

Graded Exercise Test

The graded exercise test started at seven km/h, and speed was increased by 1 km/h every five minutes until the attainment or slightly surpassing an on-task RPE-O of five. From this point, speed was increased by 1 km/h every minute until the participant reported exhaustion. Participants verbally rated their RPE-O within the last 15 seconds of each stage and at the moment of test termination. Speeds corresponding to two (2), three (3) and five (5) points on the Borg’s CR-10 scale during this test were taken as weak, moderate and strong speeds, respectively. We set speeds by the Borg’s own CR-10 scale, considering that the numbers related “weak”, “moderate” and “strong” were suggested as perceptual references for light, moderate and vigorous exercise categories (Norton et al., 2010). In cases where such RPE values were not exactly reported, target speeds were computed by means of linear speed-RPE interpolations, considering the immediate inferior and superior values. The highest five second average HR and RPE-O observed on this test were considered maximal HR and maximal RPE-O.

Constant Speed Trials

Running bouts of 15 and 30 minutes were performed at weak, moderate and strong speeds, as determined from the graded exercise test. On each occasion, treadmill speed was individually set to a predetermined value following the warm-up, and speed was kept constant throughout the defined trial duration until decreased for the cool down period. Although participants were previously told about speed and duration they would perform, the display on the treadmill was covered during these efforts, so that participants received no information about elapsed times. Overall and differentiated RPE were accessed in the last 15 seconds of each 3-minute period during the trials; mean and peak values were retained for analysis. Mean and peak HR, expressed as percentages of HR reserve (%HRR), were taken as objective intensity measures. Overall and differentiated sRPE were taken with participants seated quietly in the laboratory after a 30-minute recovery period, during which they had ad libitum access to water and bathroom. We adopted counterbalancing for the order of constant speed trials and type of RPE and sRPE assessment.

Statistical Analysis

Data analysis was carried out with a statistical software package (Statistica 7.0, Stat Soft, Tulsa, OK, USA), and the results were presented as means (Ms) and standard deviations (SDs). We used a two-way ANOVA for repeated measures (three speeds x two durations) for HR analysis, whereas we used a three-way ANOVA for repeated measures (three speeds x two durations x three RPE types) for both RPE and sRPE analyses. When sphericity violation was found by the Mauchlýs test, we used Huynh-Feldt (H-F) corrections for degrees of freedom, with post hoc comparisons performed by Scheffé´s test. We set statistical significance at p < 0.05 for all inferential analysis. As effect size estimates, we computed generalized eta squared (η _G ²) from each ANOVA table (Bakeman, 2005), and we calculated raw and standardized mean differences (Cohen´s dz) as well as their 95% confidence intervals (95%CI) between sRPE taken after 15 and 30 minute trials (Dankel & Loenneke, 2018). While raw differences enabled inferences on the magnitude of changes in Foster’s scale sRPE units, Cohen´s dz values were presented as complimentary statistics in order to enable future comparisons between studies and meta-analyses, given the variety of scales currently in use for sRPE evaluation.

Results

Peak speed on the graded test was 14.5 (SD = 1.5) km/h, while maximal HR and RPE-O were 195 (SD = 7) bpm and 10.5 (SD = 1.8), respectively. Weak, moderate and strong speeds for the 15 and 30-minute trials were 8.2 (SD =1.0), 9.2 (SD = 1.0) and 10.4 (SD = 0.9) km/h.

Significant effects of speed and duration were observed for mean HR [F_(1.33,19,99) = 52.55, p < 0.01, η _G ² = 0.48 and F_(1,15) = 24.15, p < 0.01, η _G ² = 0.06] and peak HR [F_(1.35,20.28) = 81.83, p < 0.01, η _G ² = 0.52 and F_(1,15) = 48.01, p < 0.01, η _G ² = 0.09], with a significant speed x duration interaction found for the later [F_(2,30) = 5.04, p = 0.01, η _G ² = 0.01]. Post-hoc comparisons showed that mean and peak HR increased significantly (p < 0.01) from weak to moderate and strong speeds in both durations, and except for the weak speed, HR values were significantly higher in the longer trials. These results are shown in Table 1.

Table 1.

Means (and SDs) for Mean and Peak Heart Rate (HR) in the Constant Speed Trials.

	15 min			30 min
	Weak	Moderate	Strong	Weak	Moderate	Strong
Mean HR (%HRR)	64 (9)	71 (7)*	80 (6)*^#	67 (8)	75 (6)*^§	84 (6)*^#§
Peak HR (%HRR)	69 (7)	78 (7)*	83 (8)*^#	69 (11)	80 (8)*^§	87 (9)*^#§

Note. %HRR = percentage of heart rate reserve. Scheffé honestly significant difference post-hoc test: *p < 0.05 from weak speed within duration; ^#p < 0.05 from moderate speed within duration; ^§p < 0.05 from 15 min within speed.

Concerning mean and peak RPE, significant main effects were found for speed [F_{(1.4, 20.9)} = 49.56, p < 0.01, η _G ² = 0.45 and F_{(2, 30)} = 68.46, p < 0.01, η _G ² = 0.46], duration [F_{(1, 15)} = 21.98, p < 0.01, η _G ² = 0.18 and F_(1,15) = 24.04, p < 0.01, η _G ² = 0.25] and RPE type [F_(1.2,18.1) = 7.96, p < 0.01, η _G ² = 0.02 and F_(1.2,18.5) = 5.47, p = 0.02, η _G ² = 0.01]. In addition, a statistically significant speed x duration interaction was found for peak RPE [F_(2,30) = 4.79, p = 0.01, η _G ² = 0.06]. According to post-hoc comparisons, RPE measures were significantly greater (p < 0.05) in the strong compared to weak and moderate speeds in both durations. Except for mean RPE-O, significant differences (p < 0.05) were also found between weak and moderate speed trials of longer duration. Moreover, RPE measures in moderate and strong speed efforts lasting 30 minutes were significantly greater (p < 0.05) than those in 15-minute efforts, with no significant differences observed between RPE types within any trial (p > 0.05). These results are shown in Table 2.

Table 2.

Means (and SDs) for Mean and Peak RPE in the Constant Speed Trials.

	15 min			30 min
	Weak	Moderate	Strong	Weak	Moderate	Strong
Mean RPE
RPE-O	2.4 (0.8)	2.5 (0.6)	3.8 (1.0)*^#	2.8 (0.7)	3.4 (0.8)^§	5.2 (1.5)*^#§
RPE-B	2.2 (0.8)	2.4 (0.6)	3.8 (1.1)*^#	2.6 (0.8)	3.4 (1.0)*^§	5.1 (1.5)*^#§
RPE-L	2.5 (0.8)	2.8 (0.9)	4.2 (1.2)*^#	2.9 (0.8)	3.7 (1.2)*^§	5.5 (1.4)*^#§
Peak RPE
RPE-O	3.0 (0.9)	3.3 (0.9)	5.4 (1.4)*^#	3.8 (1.7)	5.2 (1.6)*^§	8.7 (2.3)*^#§
RPE-B	2.8 (1.0)	3.1 (0.9)	5.5 (1.5)*^#	3.6 (1.8)	5.3 (2.2)*^§	8.6 (2.3)*^#§
RPE-L	3.1 (1.0)	3.8 (1.7)	6.1 (2.2)*^#	4.1 (2.1)	5.7 (2.6)*^§	8.8 (2.5)*^#§

Note. RPE-O = overall RPE; RPE-B = breathing RPE; RPE-L = legs RPE. Scheffé honestly significant difference post hoc test: *p < 0.05 from weak speed within duration and RPE type; ^#p < 0.05 from moderate speed within duration and RPE type; ^§p < 0.05 from 15 min trial within speed and RPE type.

Figure 1 shows overall and differentiated sRPE after the constant speed trials. Significant main effects were found for speed [F_(2,30) = 61.54, p < 0.01, η _G ² = 0.48], duration [F_(1,15) = 45.43, p < 0.01, η _G ² = 0.16] and sRPE type [F_(1.17,17.60) = 7.25, p = 0.01, η _G ² = 0.05]. In addition, a significant speed X duration interaction was observed [F_(2,30) = 15.45, p < 0.01, η _G ² = 0.10]. Post-hoc analysis revealed that within duration and type, sRPE after strong speed trials were significantly greater (p < 0.05) than those following weak and moderate trials, whereas significant differences between the later speeds were restricted to 30-minute bouts. Comparisons between 15 and 30-minute trials within speeds and sRPE types showed extended duration to significantly increase sRPE-O, sRPE-B and sRPE-L after moderate and strong runs (p < 0.05), but not weak runs. Except for the higher sRPE-L vs sRPE-B (p < 0.01) after the 30-minute strong speed bout, no significant differences were found between sRPE types in any other trial. Raw and standardized mean changes in sRPE from 15 to 30-minute trials at different speeds are shown in Table 3.

Figure 1.

Session Ratings of Perceived Exertion (sRPE) Taken After 15 and 30-Minute Trials at Weak, Moderate and Strong Speeds. Note: sRPE-O = overall sRPE; sRPE-B = breathing sRPE; sRPE-L = legs sRPE. Scheffé honestly significant difference post hoc test: *p < 0.05 from weak and moderate speeds within duration and sRPE type; ^#p < 0.05 from weak speed within duration and sRPE type; ^§p < 0.05 from 15 minute trial within speed and sRPE type; ^¥p < 0.05 from sRPE-B within speed and duration.

Table 3.

Raw and Standardized (Cohen’s dz) Mean Changes in Session Ratings of Perceived Exertion (sRPE) From 15 to 30 min Trials at Different Speeds.

	Raw (95%CI)	Cohen’s dz (95%CI)
Weak
sRPE-O	0.03 (−0.40 to 0.46)	0.04 (−0.45 to 0.53)
sRPE-B	0.06 (−0.34 to 0.46)	0.08 (−0.41 to 0.57)
sRPE-L	0.19 (−0.21 to 0.59)	0.25 (−0.25 to 0.74)
Moderate
sRPE-O	1.06 (0.48–1.64)	0.98 (0.37–1.57)
sRPE-B	0.88 (0.37–1.39)	0.92 (0.32–1.49)
sRPE-L	1.06 (0.42–1.70)	0.88 (0.29–1.45)
Strong
sRPE-O	2.13 (1.28–2.98)	1.33 (0.64–2.00)
sRPE-B	1.94 (1.15–2.73)	1.31 (0.62–1.97)
sRPE-L	2.50 (1.54–3.46)	1.39 (0.68–2.07)

Note. 95% CI = 95% confidence interval.

Discussion

To our knowledge this is the first study to address the effects of exercise duration at different intensities on sRPE. Our main results were that extending treadmill running from 15 to 30 minutes increased both overall and differentiated sRPE after moderate and strong runs, but not after weak speed trials among recreationally active participants. These findings corroborated previous speculations that duration can play an important role in onés retrospective perceptual evaluation of a more demanding exercise bout.

With small effect sizes and mean increases in sRPE of less than 0.2 points on the Foster’s scale, effects of duration were minimal and non-significant at weak speed, with both 15 and 30-minute trials rated between weak to moderate. Effect sizes grew in parallel to exercise intensity, with significant increases of ∼ 1 and 2.2 points in sRPE (data pooled by sRPE type) at moderate and strong speeds, respectively. Such increases changed retrospective perceptions that were midway between weak and moderate, to a point between moderate and somewhat strong after moderate speed, and from midway between somewhat strong and strong, to between strong and very strong after strong speed exercise. These results clearly show an intensity-dependent effect of duration on sRPE, extending an understanding of sRPE from that based on previous studies that used only single intensity research designs (Barroso et al., 2015; Foster et al., 2001; Fusco et al., 2020; Green et al., 2009).

The sRPE dependence on exercise duration may present a fundamental problem for ITL assessment with sRPE (Herman et al., 2006), since our data imply that it is possible to overvalue exercise volume when multiplying duration by sRPE (seemingly already sensitive to volume). As far as we know, two other comparative studies (both involving swimmers subjected to interval training) showed significant effects of exercise volume on overall sRPE. Barroso et al. (2015) found that sRPE taken 30 minutes after exercise was higher after 20×100 meter swims compared to 10×100 meter swim repetitions at critical speed, and Fusco et al. (2020) recently reported increases in sRPE accessed 10 minutes after each four sets of 10×100 (91.4 meter) swims, thought to represent a high-volume training session. Barroso et al. (2015) presented no data for on-task markers of exercise intensity, but critical swimming speed is known to lie above the moderate intensity exercise domain (Dekerle et al., 2010; Ribeiro et al., 2010). Fusco et al. (2020) found no significant changes in speed, HR and blood lactate throughout the sets, speculating that the sRPE drift could be related to progressive muscle glycogen depletion.

Although comparisons between studies are not straightforward, given their different methodologies, past studies together with ours suggest that a volume effect on sRPE may occur in different exercise paradigms. Nevertheless, the idea that sRPE per se could provide a measure of ITL may be too simplistic and warrants caution. Even though it is affected by exercise volume, sRPE is still mainly dependent on intensity, as denoted by the three times greater effect size observed for running speed compared to that for exercise duration in our study (η _G ² = 0.48 vs 0.16). Secondly, given the parallel rises in retrospective perceptions and on-task psychophysiological strain due to exercise prolongation, it is still plausible that the effects of duration on sRPE were mediated, at least in part, by increased exercise intensities.

Milanez et al. (2011) stated that sRPE may be influenced by many interacting factors in a way at least as complex as for on-task RPE, for which the role of interactions between afferent (i.e., cardiopulmonary and muscular interoceptors) and efferent (i.e., corollary discharges from central motor command) mechanisms have been under debate (Abbiss et al., 2015; Pageaux, 2016). Irrespective of the underlying mechanisms, however, on-task RPE rises in association with physiological indices (i.e., HR, VO₂ and blood lactate) in both graded (Dantas et al., 2015; Monteiro et al., 2019; Soriano-Maldonado et al., 2015) and constant speed running (Kilpatrick et al., 2009; Monteiro et al., 2019; Steed et al., 1994), since the magnitude of RPE increases are greater in bouts of higher intensity. Although sRPE seems to track average physiological perturbations brought about in an exercise session (Green et al., 2009; Haddad et al., 2017; Herman et al., 2006), its underlying neurophysiological mechanisms remain to be determined.

In the only other investigation conducted on a treadmill, Green et al. (2009) found no significant effects of duration (despite a trend toward higher values with increased duration) on overall or differentiated sRPE taken 20 minutes after 20, 30 and 40-minute runs at ∼ 70% VO₂max. They also showed that separate trials did not differ significantly on on-task physiological measures (i.e, HR and blood lactate) or perceptual responses. A close look at their results, however, revealed that, in all trials, both mean RPE (3.2 to 4.0) and sRPE (4.1 to 4.7) values were within the “somewhat easy” range on the OMNI pictorial scale by Utter et al. (2004).

In the present study, instead of an objective physiological marker we set speeds by the Borg’s CR-10 scale, considering that the numbers related “weak”, “moderate” and “strong” were suggested as perceptual references for light, moderate and vigorous exercise categories (Norton et al., 2010). By using the above verbal anchors we also sought to provide our volunteers readily understandable terms and cover an intensity range that is usually experienced by exercise practitioners. This procedure enabled volunteers to complete all trials and resulted in distinct sRPE responses. The ∼ 1 km/h increases from weak to moderate and strong speeds were followed by increases in mean and peak HR in both exercise durations. Though to a small extent, exercise prolongation also led to elevations in cardiovascular demand that reached statistical significance at higher speeds. According to different positions (Garber et al., 2011; Norton et al., 2010) from mean HR data, the trials ranged from the lower to the upper limits of vigorous exercise. Of note, however, a more diverse picture emerged in terms of perceptual categories of intensity.

Consistent with previous investigations (Green et al., 2011; Kilpatrick et al., 2009; Monteiro et al., 2019), we found that RPE increased over time in all trials. Thus, overall and differentiated peak ratings were at least one intensity category above their mean ratings in all sessions. The fact that RPE in constant speed trials did not precisely reproduce those observed in the graded exercise is in line with Monteiro et al. (2019) who found that RPE obtained during a graded treadmill test typically overestimated RPE at the beginning of constant speed submaximal exercise, while treadmill test RPE underestimated initial constant speed submaximal exercise RPE after approximately 20 minutes.

As in previous studies (Milanez et al., 2011; Ribeiro et al., 2013) we accessed RPE and sRPE with different scales in order to avoid a possible influence of repeated on-task measures on post exercise ratings. But even though Borg’s CR10 (G. Borg, 2000; G. A. Borg, 1982) and Foster’s (Foster et al., 2001) scales have separate numerical and psychometric properties (Eston, 2012) that preclude direct statistical comparisons, mean RPE and sRPE provided very similar categorical responses that, in turn, were systematically below those related to peak on-task perceptions.

The similarity between mean RPE and sRPE for exercise intensity categories supports the idea that sRPE provides a quantitative evaluation of the intensity component for an entire exercise session. Nevertheless, comparisons between on-task and session perceptions in aerobic exercise have led to contradictory conclusions in past research. For instance, it has been argued that sRPE may best reflect the final minutes (Kilpatrick et al., 2009) as well as the most intense segment (Kilpatrick et al., 2012) of the exercise session, while Green et al. (2009) found sRPE to lie between mean and peak RPE. In addition, Haile et al. (2013) showed sRPE to be greater than mean on-task RPE on self-selected but not on imposed intensity cycloergometer exercise even when these two conditions had the same physiological workload. Although the extent to which sRPE is influenced by any momentary or averaged on-task perceptions in different paradigms remains in question, it is worth noting that in the above studies mean RPE and sRPE values fell within the range of same verbal descriptors in almost all experimental conditions.

Overall and differentiated sRPE responses to protocol manipulations were virtually the same, albeit we observed a significant type effect with a small effect size, such as for on-task perceptions. Overall sRPE was slightly but consistently lower than sRPE-L and higher than sRPE-B through the trials. Yet, the only significant post-hoc difference was for the 30-minute bout at the highest speed, after which sRPE-L was 1.2 points higher than sRPE-B (p < 0.05) and 0.9 point higher than sRPE-O (p < 0.01). These findings are at least partially supported by those from Green et al. (2009) in that sRPE-B was comparatively blunted versus sRPE-O and sRPE-L after submaximal treadmill bouts.

It should be noted that knowledge of differentiated sRPE after running exercise is still in its infancy, with studies scarce and available results equivocal (Green et al., 2011; Mann et al., 2019; McLaren et al., 2016). In this context, the modest dominance of sRPE-L over sRPE-B, as well as its amplification in the more demanding exercise bout, add interesting data to this relatively unexplored topic. Contrary to suggestions that differentiating sRPE might provide minimal additional information over sRPE-O (Green et al., 2009; Green et al., 2011), in our opinion the participants’ ability to discern between peripheral and central retrospective perceptions offers a refined look at how a given session is internalized, potentially enabling a more accurate examination of the dose-response nature of acute and chronic exercise from both practical and theoretical perspectives.

Limitations and Directions for Future Research

Despite the novelty of our findings, some weaker characteristics of our experimental design must be highlighted. First, our results may not be generalizable to participant samples other than recreationally active young adults or exercise modes beyond treadmill running. Second, the running speeds and durations represented in this study were but a small sample of what exercisers can experience in real training/conditioning settings. Although not detected by our experimental design it remains possible that a volume effect would occur even at low intensity exercise over a longer duration. Finally, our lack of physiological measures other than HR precluded any exploration of other potential physiological mediators of sRPE (i.e., EMG, blood lactate, VO₂ and heat gain). All of these limitations warrant further investigation in future studies.

Conclusion

In conclusion, this study showed that exercise duration can affect sRPE in an intensity dependent manner. This finding has important implications for prescribing specific training loads that optimize the exercise experience. Additionally, those interested in using overall and/or differentiated sRPE should be aware that these measures can be affected by exercise volume, especially for exercise at higher intensities. In practice, changes in sRPE due to exercise duration can lead to disproportional increases in internal training load scores as indicated by sRPE. This effect should also be considered when prescribing specific training loads to optimize exercisers’ retrospective perceptions of the exercise experience.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Eduardo da Silva Alves

Author Biographies

Raille Silva de Jesus, is a professor at Universidade Paulista, and Department of Health and Wellness Sciences at Universidade Potiguar, in Natal, Rio Grande do Norte, Brazil. She received her master degree in Physical Education from the Universidade Federal do Rio Grande do Norte, and her current scientific interests include physical exercise and executive functions.

Rebecca Évelyn Santos Batista, is a master degree candidate at Escola de Educação Física e Esporte of the Universidade de São Paulo, in São Paulo, Brazil. She received her degree in Physical Education from the Universidade Estadual de Santa Cruz, and her current scientific interests include epistemology and history of Physical Education and sociology of sports.

Vinícius Moura Eça Santos, is a professor at União Metropolitana de Educação e Cultura, in Itabuna, Bahia, Brazil. He received his master degree in Health Sciences from the Universidade Estadual de Santa Cruz, and his current scientific interests include affective and perceptual responses to exercise.

David Ohara, is a professor at Department of Health Sciences of the Universidade Estadual de Santa Cruz, in Ilhéus, Bahia, Brazil. He received his PhD in Physical Education from the Universidade Estadual de Londrina, and his current scientific interests include sports science, physical activity and sedentary behaviour.

Eduardo da Silva Alves, is a professor at Department of Health Sciences of the Universidade Estadual de Santa Cruz, in Ilhéus, Bahia, Brazil. He received his PhD in Sciences from the Universidade Federal de São Paulo, and his research interests include sports, exercise science and exercise physiology.

Luiz Fernando Paulino Ribeiro, is a professor at Department of Health Sciences of the Universidade Estadual de Santa Cruz, in Ilhéus, Bahia, Brazil. He received his PhD in Motricity Sciences from the Universidade Estadual Paulista, and his current scientific interests include affective and perceptual responses to exercise.

References

Abbiss

C. R.

Peiffer

J. J.

Meeusen

Skorski

(2015). Role of ratings of perceived exertion during self-paced exercise: What are we actually measuring? Sports Medicine (Auckland, N.Z.), 45(9), 1235–1243. https://doi.org/10.1007/s40279-015-0344-5

Bakeman

(2005). Recommended effect size statistics for repeated measures designs. Behavior Research Methods, 37(3), 379–384. https://doi.org/10.3758/BF03192707

Barroso

Salgueiro

D. F.

do Carmo

E. C.

Nakamura

F. Y.

(2015). The effects of training volume and repetition distance on session rating of perceived exertion and internal load in swimmers. International Journal of Sports Physiology and Performance, 10(7), 848–852. https://doi.org/10.1123/ijspp.2014-0410

Borg

(2000). Escalas de Borg para a dor e o esforço percebido [Borg scales for pain and perceived exertion] (1^a Edição). Editora Manole.

Borg

G. A.

(1982). Psychophysical bases of perceived exertion. Medicine and Science in Sports and Exercise, 14(5), 377–381.

Bourdon

P. C.

Cardinale

Murray

Gastin

Kellmann

Varley

M. C.

Gabbett

T. J.

Coutts

A. J.

Burgess

D. J.

Gregson

Cable

N. T.

(2017). Monitoring athlete training loads: Consensus statement. International Journal of Sports Physiology and Performance, 12(Suppl 2), S2161–S2170. https://doi.org/10.1123/IJSPP.2017-0208

Dankel

S. J.

Loenneke

J. P.

(2018). Effect sizes for paired data should use the change score variability rather than the pre-test variability. Journal of Strength and Conditioning Research. Advance online publication. https://doi.org/10.1519/JSC.0000000000002946

Dantas, J. L., Doria, C., Rossi, H., Rosa, G., Pietrangelo, T., Fanò-Illic, G., & Nakamura, F. Y. (2015). Determination of blood lactate training zone boundaries with rating of perceived exertion in runners. Journal of Strength and Conditioning Research, 29(2), 315–320. https://doi.org/10.1519/JSC.0000000000000639

Dekerle

Brickley

Alberty

Pelayo

(2010). Characterising the slope of the distance-time relationship in swimming. Journal of Science and Medicine in Sport, 13(3), 365–370. https://doi.org/10.1016/j.jsams.2009.05.007

10.

Eston

(2012). Use of ratings of perceived exertion in sports. International Journal of Sports Physiology and Performance, 7(2), 175–182. https://doi.org/10.1123/ijspp.7.2.175

11.

Foster

(1998). Monitoring training in athletes with reference to overtraining syndrome. Medicine and Science in Sports and Exercise, 30(7), 1164–1168. https://doi.org/10.1097/00005768-199807000-00023

12.

Foster

Florhaug

J. A.

Franklin

Gottschall

Hrovatin

L. A.

Parker

Doleshal

Dodge

(2001). A new approach to monitoring exercise training. Journal of Strength and Conditioning Research, 15(1), 109–115.

13.

Foster

Hector

L. L.

Welsh

Schrager

Green

M. A.

Snyder

A. C.

(1995). Effects of specific versus cross-training on running performance. European Journal of Applied Physiology and Occupational Physiology, 70(4), 367–372. https://doi.org/10.1007/BF00865035

14.

Foster

Rodriguez-Marroyo

J. A.

de Koning

J. J.

(2017). Monitoring training loads: The past, the present, and the future. International Journal of Sports Physiology and Performance, 12(Suppl 2), S22–S28. https://doi.org/10.1123/ijspp.2016-0388

15.

Fusco

Knutson

King

Mikat

R. P.

Porcari

J. P.

Cortis

Foster

(2020). Session RPE during prolonged exercise training. International Journal of Sports Physiology and Performance, 15(2), 292–294. https://doi.org/10.1123/ijspp.2019-0137

16.

Garber

C. E.

Blissmer

Deschenes

M. R.

Franklin

B. A.

Lamonte

M. J.

Lee

I.-M.

Nieman

D. C.

Swain

D. P.

, & American College of Sports Medicine. (2011). American College of Sports Medicine position stand. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: Guidance for prescribing exercise. Medicine and Science in Sports and Exercise, 43(7), 1334–1359. https://doi.org/10.1249/MSS.0b013e318213fefb

17.

Green

J. M.

Laurent

C. M.

Bacon

N. T.

Oneal

E. K.

Davis

J. K.

Bishop

P. A.

(2011). Crossmodal session rating of perceived exertion response at low and moderate intensities. Journal of Strength and Conditioning Research, 25(6), 1598–1604. https://doi.org/10.1519/JSC.0b013e3181ddf6a2

18.

Green

Matthew, McIntosh

J. R.

Hornsby

Timme

Gover

Mayes

J. L.

(2009). Effect of exercise duration on session RPE at an individualized constant workload. European Journal of Applied Physiology, 107(5), 501–507. https://doi.org/10.1007/s00421-009-1153-z

19.

Haddad

Stylianides

Djaoui

Dellal

Chamari

(2017). Session-RPE method for training load monitoring: Validity, ecological usefulness, and influencing factors. Frontiers in Neuroscience, 11, 612. https://doi.org/10.3389/fnins.2017.00612

20.

Haile

Gallagher

J. M.

Robertson

R. J.

(2014). Perceived exertion laboratory manual: From standard practice to contemporary application. Springer.

21.

Haile

Goss

F. L.

Robertson

R. J.

Andreacci

J. L.

Gallagher

Nagle

E. F.

(2013). Session perceived exertion and affective responses to self-selected and imposed cycle exercise of the same intensity in young men. European Journal of Applied Physiology, 113(7), 1755–1765. https://doi.org/10.1007/s00421-013-2604-0

22.

Herman

Foster

Maher

M. A.

Mikat

R. P.

Porcari

J. P.

(2006). Validity and reliability of the session RPE method for monitoring exercise training intensity. South African Journal of Sports Medicine, 18(1), 14–17.

23.

Iellamo

Manzi

Caminiti

Vitale

Massaro

Cerrito

Rosano

Volterrani

(2014). Validation of rate of perceived exertion-based exercise training in patients with heart failure: Insights from autonomic nervous system adaptations. International Journal of Cardiology, 176(2), 394–398. https://doi.org/10.1016/j.ijcard.2014.07.076

24.

Kilpatrick

M. W.

Bortzfield

A. L.

Giblin

L. M.

(2012). Impact of aerobic exercise trials with varied intensity patterns on perceptions of effort: An evaluation of predicted, in-task, and session exertion. Journal of Sports Sciences, 30(8), 825–832. https://doi.org/10.1080/02640414.2012.671954

25.

Kilpatrick

M. W.

Robertson

R. J.

Powers

J. M.

Mears

J. L.

Ferrer

N. F.

(2009). Comparisons of RPE before, during, and after self-regulated aerobic exercise. Medicine & Science in Sports & Exercise, 41(3), 682–687. https://doi.org/10.1249/MSS.0b013e31818a0f09

26.

Mann

R. H.

Williams

C. A.

Clift

B. C.

Barker

A. R.

(2019). The validation of session rating of perceived exertion for quantifying internal training load in adolescent distance runners. International Journal of Sports Physiology and Performance, 14(3), 354–359. https://doi.org/10.1123/ijspp.2018-0120

27.

McLaren

S. J.

Graham

Spears

I. R.

Weston

(2016). The sensitivity of differential ratings of perceived exertion as measures of internal load. International Journal of Sports Physiology and Performance, 11(3), 404–406. https://doi.org/10.1123/ijspp.2015-0223

28.

Milanez

V. F.

Spiguel Lima

M. C.

Gobatto

C. A.

Perandini

L. A.

Nakamura

F. Y.

Ribeiro

L. F. P.

(2011). Correlates of session-rate of perceived exertion (RPE) in a karate training session. Science & Sports, 26(1), 38–43. https://doi.org/10.1016/j.scispo.2010.03.009

29.

Monteiro

Cunha

Brasil

Joi

Farinatti

(2019). Rates of perceived exertion obtained from cardiopulmonary exercise testing are not reproduced during prolonged aerobic bouts. Journal of Exercise Physiology Online, 22(4), 29–39.

30.

Norton

Sadgrove

(2010). Position statement on physical activity and exercise intensity terminology. Journal of Science and Medicine in Sport, 13(5), 496–502. https://doi.org/10.1016/j.jsams.2009.09.008

31.

Pageaux

(2016). Perception of effort in exercise science: Definition, measurement and perspectives. European Journal of Sport Science, 16(8), 885–894. https://doi.org/10.1080/17461391.2016.1188992

32.

Pandolf

K. B.

(1982). Differentiated ratings of perceived exertion during physical exercise. Medicine and Science in Sports and Exercise, 14(5), 397–405.

33.

Ribeiro

L. F. P.

Alves

V. V.

da Silva

L. H.

Fontes

E. B.

(2013). Overall and differentiated session ratings of perceived exertion at different time points following a circuit weight training workout. Journal of Exercise Science & Fitness, 11(1), 19–24. https://doi.org/10.1016/j.jesf.2013.03.001

34.

Ribeiro

L. F. P.

Lima

M. C. S.

Gobatto

C. A.

(2010). Changes in physiological and stroking parameters during interval swims at the slope of the d-t relationship. Journal of Science and Medicine in Sport, 13(1), 141–145. https://doi.org/10.1016/j.jsams.2008.10.001

35.

Robertson

R. J.

Noble

B. J.

(1997). Perception of physical exertion: Methods, mediators, and applications. Exercise and Sport Sciences Reviews, 25, 407–452.

36.

Soriano-Maldonado

Ruiz

J. R.

Álvarez-Gallardo

I. C.

Segura-Jiménez

Santalla

Munguía-Izquierdo

(2015). Validity and reliability of rating perceived exertion in women with fibromyalgia: Exertion-pain discrimination. Journal of Sports Sciences, 33(14), 1515–1522. https://doi.org/10.1080/02640414.2014.994661

37.

Steed

Gaesser

G. A.

Weltman

(1994). Rating of perceived exertion and blood lactate concentration during submaximal running. Medicine and Science in Sports and Exercise, 26(6), 797–803. https://doi.org/10.1249/00005768-199406000-00021

38.

Utter

A. C.

Robertson

R. J.

Green

J. M.

Suminski

R. R.

McAnulty

S. R.

Nieman

D. C.

(2004). Validation of the adult OMNI scale of perceived exertion for walking/running exercise. Medicine and Science in Sports and Exercise, 36(10), 1776–1780. https://doi.org/10.1249/01.mss.0000142310.97274.94

39.

Volterrani

Iellamo

(2016). Cardiac rehabilitation in patients with heart failure: New perspectives in exercise training. Cardiac Failure Review, 2(1), 63–68. https://doi.org/10.15420/cfr.2015:26:1