ROBERTSON, ROBERT J.; MOYNA, NIALL M.; SWARD, KATHY L.; MILLICH, NANCY B.; GOSS, FREDRIC L.; THOMPSON, and PAUL D.
Mixed gender experimental paradigms have been used extensively to study exertional perceptions during both weight-bearing and nonweight-bearing dynamic exercise (2–17, 19–21,24,26–28). A mixed gender paradigm has also been employed to examine the abatement of exertional perceptions after maximal exercise testing (16). These paradigms raise the possibility of systematic gender differences in ratings of perceived exertion (RPE). Differences in RPE between female and male subjects when not accounted for in a mixed gender design can make interpretation and generalization of research findings difficult. However, experimental evidence regarding the possible effect of gender on RPE is inconsistent for both adult and pediatric samples.
Previous investigations have compared RPE between female and male subjects using experimental paradigms that expressed the physiological reference criterion in absolute (i.e., oxygen uptake, (V̇O2), L·min-1, heart rate (HR), beats·min-1) (4,10,13,17,28), and/or relative (i.e., %V̇O2max/peak; %HRmax/peak) (4,8,13,17,21,26) units. These paradigms involved dynamic exercise modalities requiring relative metabolic rates ranging between 25% and 90% of mode specific maximal/peak values for V̇O2 and HR. Evidence both refuting and supporting a systematic gender effect on exertional perceptions is evident in these previous investigations. As an example, differences in RPE for the overall body have been observed between female and male subjects when comparisons were made at a given submaximal V̇O2 during weight-bearing [i.e., treadmill (8,9,13)], partial weight-bearing [i.e., ski machine (13)], and nonweight-bearing [i.e., cycle (9,13,21) and aerobic rider (13)] exercise modes. Conversely, gender differences in RPE have not been observed when the physiological reference criterion involved V̇O2(17,27), HR (4,21,24,26), blood lactic acid concentration (26), and the lactate (8,26) and ventilatory (21) thresholds. However, it is important to note that many of these investigations did not compare RPE between gender at both absolute and relative physiological reference criteria within the same experimental paradigm (12,19,28). Such methodological inconsistencies make it difficult to systematically examine the influence of gender on RPE responsiveness during dynamic exercise. In contrast, the present investigation compared RPE between genders at absolute and relative reference criteria within the same paradigm for a given exercise mode.
Gender differences in RPE have (13) and have not (9) been observed for weight-bearing exercise such as treadmill walking and running. Similar inconsistency in gender specific perceptual responsiveness is apparent when the exercise mode is either partial weight bearing or nonweight bearing. Men and women have been shown to be both similar and dissimilar in their perception of exertion during stationary cycling (6,7,9,14,28), arm ergometry (19), swimming (3,12,26), simulated skiing (13), and semirecumbent riding (13), all of which are modes that either partially or completely support body weight.
Evidence is both limited and inconsistent regarding the effect of gender on differentiated RPE arising from the active limbs and/or respiratory structures. RPE-arms was higher in female than male subjects when comparisons were made at a fixed submaximal power output (PO) (19) and lower in female than male subjects at a self-selected swimming velocity (12). In contrast, gender differences were not observed for differentiated RPE reflecting respiratory function and localized muscular effort (14).
The experimental paradigms employed in the investigations cited above generally compared RPE between female and male subjects at a predetermined physiological and/or physical reference criteria. A consistent effect of gender on both the undifferentiated and differentiated ratings of perceived exertion was not found for either adult or pediatric samples owing principally to three methodological factors: a) whether the reference criterion upon which gender specific RPE comparisons were made was expressed as either absolute [i.e., V̇O2, (L·min-1); HR, (beats·min-1); PO, (W)] or relative [i.e., %V̇O2max; %HRmax; %POpeak ] units; b) the type of physiological (i.e., V̇O2, HR, lactic acid) and/or physical (i.e., PO, velocity) reference criterion employed; and c) whether the ergometric device was either weight bearing (i.e., treadmill), partial weight bearing (i.e., swimming, simulated skiing), or nonweight bearing (i.e., cycle and arm ergometer). The foregoing inconsistencies not withstanding, a trend in experimental evidence suggests that female and male subjects do not differ in their perception of exertion when comparisons are made at relativized physiological measures of aerobic metabolic and/or cardiovascular function (2,3,5,17,26,28). This experimental trend is especially apparent when RPE is compared between female and male subjects at both relative and absolute physiological reference criteria within the same exercise paradigm. Using such a paradigm, Noble et al. observed that gender differences in RPE that were apparent at an absolute V̇O2 were absent when comparisons were made at a given percent of maximal V̇O2 (i.e. %V̇O2max) (17).
It was the expectation of this investigation that undifferentiated (overall body) and differentiated (limbs and chest) RPE would not differ between female and male subjects when comparisons were made at relative aerobic and cardiovascular physiological reference criteria during weight-bearing (treadmill), partial weight-bearing (simulated skiing), and nonweight-bearing (cycle ergometer) dynamic exercise. Oxygen uptake and heart rate were selected as the aerobic and cardiovascular reference criteria because these variables: a) provide a stable indication of metabolic and circulatory function over a wide range of intensities involving multi-limb dynamic exercise (22, p. 238–239) and b) are commonly used in experimental paradigms where physiological measures are employed as reference criteria upon which to compare perceived exertion responses.
Nine, physically active men and 10 physically active women participated as subjects (Table 1). Using a power of 0.80, an α of 0.05, and an effect size of 1.2, it was estimated that nine subjects were required for each experimental cohort in a two-factor design with repeated measures on intensity but not gender (25). It is recognized that this total sample size is comparatively small. However, the actual effect size in the present investigation was ≥ 1.2 for all comparisons using the absolute V̇O2 as the physiological reference criterion. As such, for a power of 0.80, the number of subjects in the experimental cohorts was sufficient to test the null hypothesis at an α of 0.05. Each subject provided written consent to participate in accordance with the University of Pittsburgh’s Institution Review Board policy. All subjects were healthy nonsmokers used no medications and were light to moderate recreational exercisers.
The data reported in this investigation were determined during a series of perceptual estimation trials administered as part of a larger estimation–production experimental paradigm. In the larger paradigm, subjects underwent one RPE scale anchoring and one estimation trial on each of six randomly presented exercise machines, i.e., treadmill, simulated ski machine, cycle ergometer, rowing ergometer, aerobic rider, and stair stepper. Tests on different exercise machines were separated by 1 wk. The overall experimental paradigm presented the six modes in randomized order minimizing the possibility of a test sequence effect on the perceptual and physiological responses. The present investigation reports data obtained during the estimation trial for three of the machines, i.e., treadmill, simulated ski machine, and cycle ergometer. These modes were selected because: a) they represent the three basic types of architecture used in exercise machines, i.e., weight bearing (treadmill), partial weight bearing (ski simulator), and nonweight bearing (cycle); b) have been used extensively in exercise research; and (c) are among the most commonly sold exercise machines for home and commercial use. Subjects did not exercise for 24 h before testing and performed tests in a 4-h postabsorptive state.
Ratings of Perceived Exertion
Ratings of perceived exertion were obtained using the 15-category Borg RPE scale (1). Before each test, subjects read a standard set of perceptual scaling instructions. These instructions followed an established format used in previous investigations (18). Three separate ratings of perceived exertion were obtained in counterbalanced sequence during the final 30 s of each exercise stage. An undifferentiated rating was assessed for the overall body (RPE-O). A differentiated rating was assessed for the perception of respiratory exertion in the chest (RPE-C) and for the perception of peripheral exertion in the legs (RPE-L). In addition, a differentiated RPE was determined for the arms during simulated ski exercise. This response was designated RPE-A (ski) and was only used in data analysis involving simulated ski exercise. The low and high rating standards (i.e., perceptual scale anchors) were established separately for each exercise machine using a continuous incremental maximal/peak exercise test as reported previously (16). Each exercise stage of the anchoring test was 2 min in duration. A rating of 7 (low anchor) was assigned to the lowest exercise intensity and 19 (high anchor) was assigned to the highest exercise intensity. Subjects were instructed to make their subjective assessments of perceived exertion relative to the designated low and high perceptual anchors.
The estimation trials consisted of a continuous, incremental maximal/peak exercise test identical to that used in the anchoring trials. Testing involved three exercise machines: treadmill (Trackmaster, TM500 E/R Pensacola, FL); cross-country ski simulator (Nordic Track Pro, Nordic Track, Chaska, MN); and cycle ergometer (Monark 818E, Monark AB, Varberg, Sweden). During the initial 6 min of the treadmill test, elevation was set at 0% and belt speed was 120.8 m·min-1 for female and 147.6 m·min-1 for male subjects. Every 2 min thereafter, belt speed and elevation were increased by 13.4 m·min-1 and 1.5%, respectively. Gender specific treadmill speeds were used for the warm-up stage to account for absolute differences in maximal aerobic power between the female and male cohorts. Exercise on the ski machine consisted of contralateral arm and leg movements that simulated cross-country skiing. Arm resistance was applied by a disk brake system and leg resistance by a direct-drive flywheel system. Using the manufacturer’s settings on the Pro Model Nordic Track, exercise intensity was incremented in 2-min stages according to the following gender specific protocols. For the female cohort, the initial arms resistance was set at level 2 and incremented by 0.5 level per stage. The initial legs resistance was set at 1.5 units and incremented by 0.5 unit per stage. The initial leg-slide/arm-pull cadence was 104 cycles·min-1 and was incremented by 6 cycles·min-1. For the male cohort, the initial arms resistance was set at level 2.0 and incremented by 1.0 level per stage. The initial legs resistance was set at 1.5 units and was incremented by 1.0 unit per stage. The initial slide/pull cadence was 104 cycles·min-1 and was incremented by 6 cycles·min-1. The cycle test protocol consisted of pedaling without external brake resistance for 2 min after which PO was increased by 25 W at the beginning of each succeeding 2 min exercise stage. Pedal rate was 50 rev·min-1, with the cadence signaled by a metronome. The exercise test for each machine was voluntarily terminated by the subject when he/she could no longer continue owing to fatigue. Subjects were verbally encouraged to exercise until exhaustion.
Maximal/peak oxygen uptake (V̇O2max/peak) was determined by averaging the two highest consecutive 30-s values for a given exercise mode. Criteria for V̇O2max/peak were attainment of at least two of the following: a) a plateau of V̇O2, as indicated by a difference of either < 2.1 mL·kg-1·min-1 or 150 mL·min-1 between the last two stages of the test; b) respiratory exchange ratio (RER) > 1.10; and c) heart rate within ± 5 beats·min-1 of each subject’s age predicted maximal value. Using these criteria, V̇O2max/peak was attained by each subject during the three exercise modes.
Cardiovascular and Respiratory Metabolic Measures
Respiratory-metabolic responses were determined during each exercise test using standard open circuit spirometry. Expired volume was determined with a ventilatory measurement module (VMM) (Interface Associates, Laguna Niguel, CA). Oxygen and carbon dioxide concentrations of expired air were sampled from a 5 L RMC-1 Rayfield mixing chamber (Rayfield Equipment, Waitsfield, VT) and analyzed using a mass spectrometer (MGA 100, Marquette Electronics, Milwaukee, WI). Raw data were reduced by a custom designed data acquisition program and an 8-bit A/D converter. Before testing, the mass spectrometer was calibrated with standard gases of known concentration. Heart rate was measured using a Polar monitor (Polar Sport Tester PE 3000, Polar, Port Washington, NY). Both respiratory-metabolic and heart rate measures were obtained during each minute of the exercise tests.
Data obtained during the final minute of each stage of the three estimation trials (i.e. treadmill, ski machine, and cycle ergometer) were submitted to statistical analysis. Physiological reference criteria (i.e. V̇O2 and HR) were derived by first determining the lowest and highest absolute values that were common to both the female and male data sets. Separate determinations were made for each exercise mode. For both V̇O2 and HR the lowest absolute value held in common by the gender cohorts was equivalent to 65% of maximal/peak responses. Using the overlapping submaximal physiological response ranges for the female and male data sets, each absolute and its respective relative reference criteria were selected. The actual (i.e., absolute and relative) values for each physiological reference criterion are listed by mode in Tables 2–7.
Linear regression analysis was used to determine the relation between RPE and the absolute values for V̇O2 and HR. Analyses were computed separately for RPE-O, RPE-L, RPE-A (Ski), and RPE-C. Regression functions were computed individually for each female and male subject. Each individual regression equation was then used to compute RPE (-O, -L, -A-ski, -C) at the three predetermined absolute reference criteria for V̇O2 and HR.
RPE at the preselected relative reference criteria were also determined individually for each subject using linear regression analysis. For both V̇O2 and HR, the absolute values equivalent to 70, 80, and 90% of the mode specific maximal/peak responses during the estimation trials were submitted to regression analysis using the same equations as above. In all cases, the values used in the analyses fell within the physiological response range that was common to both the female and male data sets. Individual (subject × criterion × mode) regression equations were then generated for RPE-O, RPE-L, RPE-A (ski), and RPE-C.
Each RPE derived from the individual regression equations for both the absolute and relative reference criteria were taken as separate data points and subsequently compiled into data sets for the female and male cohorts. These data sets were then submitted to a two (gender) × three (intensity level) analysis of variance with repeated measures on the second factor. Significant main and interaction effects were analyzed with a Scheffe post hoc procedure.
Maximal/peak responses for V̇O2, HR, and RPE are listed by gender and exercise mode in Table 1. V̇O2max/peak was higher (P < 0.05) for male than female subjects for each exercise mode. Maximal/peak responses for HR and RPE did not differ between gender at any mode.
V̇O2 Reference Criterion
Presented in Tables 2–4 are the comparisons of RPE responses between female and male subjects at absolute and relative V̇O2 criteria. RPE-O, RPE-L, RPE-A (ski), and RPE-C were significantly higher in the female than male cohort when compared at the three absolute V̇O2 (mL·kg-1·min-1 or L·min-1) criteria. When compared at the three mode specific relative (i.e., %V̇O2max/peak) criteria, RPE-O, RPE-L, RPE-A (ski), and RPE-C did not differ between female and male cohorts. These gender comparisons were consistent for the treadmill, simulated ski, and cycle exercise modes.
HR Reference Criterion
Presented in Tables 5–7 are the RPE comparisons between female and male subjects at absolute and relative HR criteria. Differences in RPE-O, RPE-L, RPE-A (ski), and RPE-C were not found between female and male cohorts when comparisons were made at both the preselected absolute (i.e., b·min-1) and relative (i.e. %HR max/peak) reference criteria. These findings were consistent for treadmill, simulated ski, and cycle exercise.
This investigation examined the effect of gender on undifferentiated and differentiated RPE using aerobic metabolic and cardiovascular physiological reference criteria. Comparisons were made at absolute and relative units for V̇O2 and HR during weight-bearing, partial weight-bearing, and nonweight-bearing dynamic exercise. The present findings generally support the premise that female and male subjects do not differ in their perception of physical exertion during dynamic exercise when comparisons are made at a given level of relative physiological strain using aerobic metabolic and/or cardiovascular reference criteria. These findings hold for dynamic exercise intensities that fall between 70% and 90% of mode specific maximal/peak values.
When V̇O2 expressed in absolute units was used as the reference criterion, RPE-O, RPE-L, RPE-A (ski), and RPE-C were higher for the female than male cohorts. However, RPE did not differ between gender when the V̇O2 reference criterion was relativized to mode specific maximal/peak values. These responses may in part be explained by gender differences in maximal aerobic power that were observed during all three dynamic exercise modes. V̇O2max/peak was lower for the female than male cohort during treadmill, ski, and cycle exercise. Therefore, each absolute submaximal V̇O2 reference criterion represented a higher relative aerobic metabolic rate for female than male subjects. In general, the relative aerobic metabolic rate is positively related to the intensity of exertional perceptions during dynamic exercise (23). It follows that when gender comparisons were made at absolute V̇O2 criteria the relative aerobic metabolic strain was greater for the female than male cohort as was RPE. However, the relative aerobic strain was likely similar between genders when comparisons were made at a given submaximal %V̇O2 max/peak. The intensity of exertional perceptions would also be expected to be similar between gender at a given relativized V̇O2 criterion.
Neither the undifferentiated nor differentiated RPE differed between the female and male cohorts when comparisons were made at both absolute and relative HR criteria. These responses were consistent across all three exercise modes. It is likely that both gender cohorts experienced the same relative physiological strain when comparisons were made at a specified submaximal HR. The HR max/peak did not differ between gender cohorts for any exercise mode. Therefore, the three absolute submaximal HRs that were used as the physiological reference for inter-cohort comparison represented the same %HR max/peak for both the female and male subjects. In the present investigation, HR was taken as a cardiovascular index of aerobic metabolic response to dynamic exercise. As the relativized HR was the same between gender cohorts, it was expected that the relative aerobic metabolic strain was also similar between female and male subjects, a response consistent with the V̇O2 data. Recognizing that RPE is linked to the relative aerobic metabolic rate, it would be expected that perceptual responsiveness would not differ between female and male subjects when comparisons were made using the same submaximal HR as a reference criterion for both genders. This conclusion held for the three dynamic exercise modes studied presently.
A number of previous investigations have measured the effect of gender on RPE where submaximal HR was the comparative reference criterion (9,12,13,28). The findings of Eynde and Ostyn (9) are partially consistent with those reported presently. When compared at a given %HR max/peak, RPE did not differ between female and male subjects during both treadmill and cycle ergometer exercise. However, the data presented did not allow comparison of RPE between gender using an absolute submaximal HR criterion.
Investigations by Koltyn et al. (12) and Winborn et al. (28) indicated gender differences in RPE when comparisons were undertaken using absolute submaximal HR as the reference criterion. Koltyn et al. (12) reported that RPE for the overall body and the arms was lower in female than male subjects during a 200-yd swimming trial at 90% of competitive speed. Winborn et al. (28) found no statistically significant difference in RPE between groups of female and male subjects of low and high athletic experience during submaximal exercise intensities equivalent to 30, 50, and 70% of estimated cycle ergometer V̇O2peak. In contrast, HR responses were higher for female than male subjects at each of the submaximal cycle intensities. By extrapolation, these data suggest that RPE was lower for female than male subjects at a given absolute HR reference criterion. However, neither Koltyn et al. (12) nor Winborn et al. (28) reported peak HR responses for the exercise modes that were studied. As such, it is not possible to determine whether the observed gender specific RPE responses persisted when comparisons were made at a relativized HR reference criterion. It follows that a balanced comparison between the findings of these two investigations and those reported presently is not possible.
Kravitz et al. (13) studied gender-specific RPE responsiveness using a multi-modal exercise paradigm. Gender differences in RPE were not found for submaximal exercise on a treadmill, cycle ergometer, and ski machine. Mean submaximal HR was higher in female than male subjects for each of the three modes studied. However, the experimental design compared RPE and HR between gender at self-selected exercise intensities specific to each mode. As such, it was not possible to examine a gender effect on RPE at either an absolute or relative HR criterion, making comparisons with the present investigation difficult.
A methodological issue involving the HR max/peak responses observed presently can also be addressed at this juncture. It is generally assumed that HR max/peak during dynamic exercise is slightly higher in female than male subjects. However, this gender specific response is not invariant (9,13,27). In a recent report involving a large number of subjects (i.e., 2010), differences in HRmax were not found between adult female and male subjects ranging in age from 14 to 77 yr (27). The similarity of HRmax/peak between the female and male cohorts observed presently is consistent with this comparatively large cross-sectional investigation. It is likely that gender specific maximal HR responses during dynamic exercise occur when experimental cohorts differ in body weight, smoking status, and/or chronological age (27). In this regard, the mean difference in age of 4.7 yr between the female and male cohorts used presently likely accounted for a difference in HRmax/peak of approximately 2.8 b·min-1, (22, p. 589), a value well within the variability of measurement reported by Whaley et al. (27).
Three previous investigations have examined a possible gender effect on RPE where perceptual responses were differentiated to the active limbs and chest (12,14,19). Using relative reference criteria, Mihevic and Morgan (14) reported that RPE differentiated according to respiratory exertion and to local muscle exertion did not differ between women and men. In contrast, O’Connor et al. (19) found that RPE for the arms during arm ergometry was greater for female than male subjects when compared at the same absolute power output. Using a swimming paradigm, Koltyn et al. (12) reported lower RPE-Arms in female than male subjects when gender cohorts performed at an intensity equivalent to 90% of their competitive speed for a 200-yd event. In the present investigation, RPE that was differentiated to the active limbs and chest did not differ between gender when comparisons were made at relative reference criteria for V̇O2 and HR. These responses were consistent with experimental expectations and likely reflected the influence of relative aerobic metabolic strain on anatomically regionalized exertional perceptions (23).
The weight-bearing characteristics of an exercise machine are dependent on the extent to which the body’s center of gravity is vertically displaced during locomotion and the associated support provided by the machines structural architecture. In the present investigation, comparisons of RPE between gender were made for responses obtained during treadmill (weight bearing), simulated ski (partial-weight bearing), and cycle (nonweight bearing) exercise modes. It was expected that RPE would be independent of gender when comparisons were made at V̇O2 and HR criteria relativized to maximal/peak values determined separately for each exercise mode. The present findings indicate that when the female and male cohorts were matched for the level of relative aerobic metabolic strain they did not differ in their perception of exertion regardless of the weight-bearing architecture of the exercise machine employed. This is consistent with previous investigations that have employed weight-bearing (8,16), partial weight-bearing (3,26) and nonweight-bearing (2,4,13,24) exercise modes.
Physiological versus Physical Reference Criteria
The present findings indicate that undifferentiated and differentiated RPE do not differ between adult female and male cohorts when comparisons are made using relative aerobic metabolic and cardiovascular reference criteria that are between 70 and 90% of maximal/peak values. In contrast, a number of previous experiments observed gender differences in perceived exertion when comparisons were made using absolute and relative physical units of work (i.e., W; velocity) as the reference criterion (6,12,19). One such experiment recently reported by Cook et al. (7) found that RPE was lower in female than male subjects when gender comparisons were made at cycle POs relativized to peak levels. Comparison of RPE between genders was not undertaken at either an absolute or relative PO in the present investigation. As such, the mechanism accounting for the difference in gender specific perceptual responsiveness between the present investigation and that of Cook et al. (7) cannot be discerned from either experimental paradigm.
The foregoing inconsistencies not withstanding, experimental paradigms can either: a) compare RPE between gender at a given physiological or physical response or b) compare a physiological or physical response between gender at a given RPE. In either paradigm, it seems important to recognize that the effect of gender on RPE may depend on whether either a relativized physiological or a relativized physical reference criterion is employed.
The present findings indicated that adult female and male subjects did not differ in the intensity of their exertional perceptions when comparisons were made at relative levels of aerobic metabolic and cardiovascular strain that were between 70 and 90% of mode specific maximal/peak values. These conclusions held for both undifferentiated and differentiated RPE and were independent of the weight-bearing architecture of the dynamic exercise machines employed. Experimental designs should recognize that gender differences in RPE might depend on whether or not physiological reference criteria are expressed as either relative or absolute measures of aerobic metabolic and cardiovascular strain, especially at comparatively higher dynamic exercise intensities.
Although the data reported presently consistently support all research hypotheses, several methodological issues place limitations on the interpretation and subsequent extrapolation of the findings. First, regression analysis was used to identify the RPE that was equivalent to the predetermined absolute and relative physiological reference criteria. This procedure was in part necessitated by the use of only one RPE response for each 2 min exercise stage in a given modality. The interpolated data derived from the regression model likely attenuated the sensitivity of RPE comparisons between gender. Second, the investigation employed a comparatively small sample within the two gender cohorts. Although sufficient statistical power was available to test the null hypothesis, it is nevertheless possible that selected gender differences in RPE could have been detected for the various modes and intensities if a larger total sample had been employed. Third, it was not possible to match the female and male cohorts with respect to mode specific physical activity history. Differential exposure to the exercise modes employed in the present design could have influenced metabolic efficiency and in turn the physiological assessments.
It is recognized that the generalizability of the present findings regarding differences and similarities in RPE responses between genders is restricted to intensities commonly used in exercise training, i.e., 70–90% V̇O2max/peak. The effect of gender on the exertional experience during multi-modal dynamic exercise undertaken at low to moderate intensities cannot be discerned from the experimental paradigm employed presently. A substantial amount of daily physical activity involving home, occupational, and recreational pursuits is performed at comparatively low intensities. As such, it is recommended that follow-on investigations regarding the effect of gender on RPE employ experimental paradigms that encompass the full aerobic metabolic range.
This research was supported in part by a grant from Nordic Track Inc. Appreciation is extended to Mrs. Donna Farrell for her technical assistance in manuscript preparation.
Address for correspondence: Robert J. Robertson, 107 Trees Hall, University of Pittsburgh, Pittsburgh, PA 15261; E-mail: firstname.lastname@example.org.
1. Borg, G. Psychological basis of perceived exertion. Med. Sci. Sports Exerc. 14: 377–384, 1982.
2. Borg, G., and H. Linderholm. Exercise performance and perceived exertion in patients with coronary insufficiency, arterial hypertension and vasoregulatory asthenia. Acta Med. Scand. 644 (Suppl.187): 17–26, 1970.
3. Burke, E. J., and T. Meade. Perceived exertion during work on a tethered swimming apparatus in age-group swimmers. The Perception of Exertion in Physical Work
, G. Borg and D. Ottoson (Eds.). London: Macmillan Press, 1986, pp. 149–159.
4. Butts, N. K., and D. Crowell. Effect of caffeine ingestion on cardiorespiratory endurance in men and women. Res. Q. Exerc. Sport 56: 301–305, 1985.
5. Butts, N. K., K. M. Knox, and T. S. Foley. Energy costs of walking on a dual-action treadmill in men and women. Med. Sci. Sports Exerc. 27: 121–125, 1995.
6. Cook, D. B., P. J. O’Connor, S. A. Eubanks, J. C. Smith, and M. Lee. Naturally occurring muscle pain during exercise: assessment and experimental evidence. Med. Sci. Sports Exerc. 29: 999–1012, 1997.
7. Cook, D. B., P. J. O’Connor, S. E. Oliver, and Y. Lee. Sex differences in naturally occurring leg muscle pain and exertion during maximal cycle ergometry. Intern. J. Neurosci. 95: 183–202, 1998.
8. DeMello, J. J., K. J. Cureton, R. E. Boineau, and M. M. Singh. Ratings of perceived exertion at the lactate threshold in trained and untrained men and women. Med. Sci. Sports Exerc. 19: 354–362, 1987.
9. Eynde B. V., and M. Ostyn. Rate of perceived exertion and its relationships with cardiorespiratory response to submaximal and maximal muscular exercise. The Perception of Exertion in Physical Work
G. Borg and D. Ottoson (Eds.). London: Macmillan Press, 1986, pp. 327–335.
10. Henriksson, J., H. G. Knuttgen, and F. Bonde-Peterson. Perceived exertion during exercise with concentric and eccentric muscle contractions. Ergonomics 15: 537–544, 1972.
11. Kolkhorst, F. W., S. W. Mittelstadt, and F. A. Dolgener. Perceived exertion and blood lactate concentration during graded treadmill running. Eur. J. Appl. Physiol. 72: 272–277, 1996.
12. Koltyn, K. F., P. J. O’Connor,and W. P. Morgan. Perception of effort in female and male competitive swimmers. Int. J. Sports Med. 12: 427–429, 1991.
13. Kravitz, L., R. A. Robergs, V. H. Hayward, D. R. Wagner, and K. Powers. Exercise mode and gender comparisons of energy expenditure at self-selected intensities. Med. Sci. Sports Exerc. 29: 1028–1035, 1997.
14. Mihevic, P. M., and W. P. Morgan. Perceptual and heart rate sensitivity to changes In exercise intensity. Med. Sci. Sports Exerc. 12: 112, 1980.
15. Meyer, F., O. Bar-Or, and B. Wilk. Children’s perceptual responses to ingesting drinks of different compositions during and following exercise in the heat. Int. J. Sport Nutr. 5: 13–24, 1995.
16. Noble, B. J. Validity of perceptions during recovery from maximal exercise in men and women. Percept. Mot. Skills 49: 891–897, 1979.
17. Noble, B. J., C. M. Maresh, and M. Ritchey. Comparison of exercise sensations between females and males. Med. Sport 14: 175–179, 1981.
18. Noble, B. J., and R. J. Robertson. Perceived Exertion. Champaign, IL: Human Kinetics, 1996, pp. 77–82.
19. O’Connor, P. J., J. S. Raglin, and W. P. Morgan. Psychometric correlates of perception during arm ergometry in males and females. Int. J. Sports Med. 17: 462–466, 1996.
20. Parfitt, G., and R. Eston. Changes in ratings of perceived exertion and psychological affect in the early stages of exercise. Percept. Mot. Skills 80: 259–266, 1995.
21. Purvis, J. W., and K. J. Cureton. Ratings of perceived exertion at the anaerobic threshold. Ergonomics 24: 295–300, 1981.
22. Robergs, R. A., and S. O. Roberts. Exercise Physiology: Exercise, Performance and Clinical Applications.
Boston: Mosby-Yearbook, Inc., 1997, pp. 238–239, 589.
23. Robertson, R. J. Central signals of perceived exertion during dynamic exercise. Med. Sci. Sports Exerc. 14: 390–396, 1982.
24. Sidney, K. H., and R. J. Shephard. Perception of exertion in the elderly, effects of aging, mode of exercise and physical training. Percept. Mot. Skills 44: 999–1010, 1977.
25. Tran, Z. V. Estimating sample size in repeated measures analysis of variance. In:Measurement in Physical Education and Exercise Science
(Special Issue), T. M. Wood (Ed.). Hillsdale, NJ: Lawrence Erlbaum Inc., 1: 89–102, 1997.
26. Ueda, T., and J. Kurokawa. Relationships between perceived exertion and physiological variables during swimming. Int. J. Sports Med. 16: 385–389, 1995.
27. Whaley, M. H., L. A. Kaminsky, G. B. Dwyer, L. H. Getchell, and J. A. Norton. Predictors of over-and underachievement of age-predicted maximal heart rate. Med. Sci. Sports Exerc. 24, 10: 1173–1179, 1992.
28. Winborn, M. D., A. W. Meyers, and C. Mulling. The effects of gender and experience on perceived exertion. J. Sport. Exerc. Psychol. 10: 22–31, 1988.
© 2000 Lippincott Williams & Wilkins, Inc.