6 Rev Bras Med Esporte _ Vol. 11, Nº 1 – Jan/Fev, 2005 1. Study Group and Research in Metabolism, Nutrition and Exercise – Phys- ical Education and Sports Center – Londrina State University – Londri- na, PR. 2. Biosciences Institute – Department of Physical Education – São Paulo State University – Rio Claro, SP. 3. Laboratory of Motor Evaluation – Juiz de Fora Federal University, MG. Received in 3/9/04. 2nd version received in 28/11/04. Approved in 30/11/04. Correspondence to: Fábio Yuzo Nakamura, Grupo de Estudo e Pesquisa em Metabolismo, Nutrição e Exercício, Centro de Educação Física e Despor- tos, Universidade Estadual de Londrina, Rod. Celso Garcia Cid, km 380, Campus Universitário – 86051-990 – Londrina, PR – Brazil. E-mail: fabioy_ nakamura@yahoo.com.br Use of perceived exertion in determining critical velocity in deep water running Fábio Yuzo Nakamura1,2, Miriely Regina Gancedo1, Leandro Albuquerque da Silva1, Jorge Roberto Perrout de Lima3 and Eduardo Kokubun2 ORIGINAL ARTICLE Key words: Perceived exertion. Critical velocity. Deep water running. ENGLISH VERSION ABSTRACT The linear relation between exercise intensity and the increase rate of the neuromuscular activity assessed by electromyography allows the estimation of the fatigue threshold, which would be the intensity that could be maintained indefinitely without presenting increase on the electromyography signals along time. It has been hypothesized that the subjetive perception of exertion would be- have similarly to the neuromuscular activity and that a perceived exertion threshold (PET) identified similarly to the electromyographic fatigue threshold could coincide with the critical velocity (CritV). Thirteen individuals from both genders (23.0 ± 2.5 years), in a 15 m long x 2.5 m deep swimming pool performed three deep water running exhaustive tests for the determination of the parameters of the critical velocity model, reporting the perceived exertion (6- 20 points in Borg scale) each 15 m. For the PET identification, the straight lines inclination coefficients of the increase on the per- ceived exertion in time (ordinate) and the velocities used (abscis- sa) were adjusted into a linear function that provided a point in the velocity axis where, theoretically, the perceived exertion would be indefinitely stable. The CritV was estimated through the equations used in the critical velocity model. For comparison purposes of the CritV and PET estimations and their associations, the repeated measures analysis of variance ANOVA was used (p < 0.05) and the Pearson correlation was calculated. The data obtained for the CritV determination fulfilled the criteria adopted for the model’s validity and CritV and PET presented no statistical difference (0.23 ± 0.02 m/s x 0.24 ± 0.03 m/s), being significantly correlated (r = 0.85). These results suggest that the PET seems to represent the maxi- mum exercise intensity in which the physiological and psychophys- ical variables would attain stability and that this index may be used in the CritV determination. INTRODUCTION Perceived exertion scales were created with the objective of establishing relations between the subjetive perception of exer- tion and the external load or physiological stress objective data. According to Borg(1), the exertion perception is a result of the inte- gration of afferent signals originated from both the skeletal mus- cles (peripheral) and from the cardiorespiratory system (central). During the performance of high-intensity exercises(2), in other words, exercises performed at intensities above the critical power (CritP), the metabolic acidosis seems to be an etiological agent common to both types of sensorial activities(3,4). The decrease on the tissue pH causes muscular fatigue and the decrease on the blood pH is associated with a ventilation increase. Both responses require higher afferent neuromotor activity for both skeletal mus- cles and ventilatory muscles(3,5), which would be subjectively per- ceived by the individual as a progressive effort to maintain the same muscular work rate. Thus, it seems that the exertion perception involves both feedback and feed forward(5). Studies show that the electromyographic activity of the knee extensor muscles undergoes progressive increase along time in high-intensity exercises in cycle ergometer(6,7). The electromyo- graphic activity increase rate is proportional to the exercise inten- sity. The linear relation between exercise intensity and electromyo- graphic activity increase rate allows the estimation of the fatigue threshold, which is the intercept of the linear regression in the intensity axis. The fatigue threshold would be the exercise intensi- ty that could be maintained indefinitely without alterations on the neuromuscular activity, in other words, without increases on the electromyographic signals along time. In the present study, it has been hypothesized that the subjetive perception of exertion would behave similarly to the neuromuscular activity objective data. Thus, a perceived exertion threshold could be estimated through proce- dure similar to the electromyographic fatigue threshold. Le Chevalier et al.(7) suggest that the fatigue threshold coincides with the CritP determined through the inclination of the work-time linear relation in ergometer adapted for unilateral knee extension exercise. This intensity was originally described as that in which the exertion of isolated muscles is maintained without exhaustion(8). Moritani et al.(9) extended this concept to the cycle ergometer. In modalities such as running(10) and swimming(11,12), the work is re- placed by distance and the power by velocity. One adopts, through this conversion, a linear relation between the respective pairs of variables. Thus, the critical velocity may be estimated (CritV). The CritV corresponds to a limit exertion intensity that could be maintained with VO2 (13) and lactate(14) steady state. Above this in- tensity, these variables reach peak values, thus predicting the ex- haustion occurrence. The occurrence of exhaustion in running at supra-CritV coincides with the full exhaustion of the anaerobic run- ning capacity (AnaerRC). The AnaerRC is described as the maxi- mum distance to be covered at the expense of the anaerobic me- tabolism. As the anaerobic metabolism mobilization is the main source of H+ ions accumulation in tissues and hence the decrease on the pH in the organism, one may infer that the AnaerRC exhaustion rate would be the responsible for the increase on the perceived exer- tion in high-intensity exercises. Thus, the maximum intensity able to be maintained without increases on the perceived exertion along time should coincide with the CritV. Therefore, the objective of the Rev Bras Med Esporte _ Vol. 11, Nº 1 – Jan/Fev, 2005 7 present study is to compare CritV obtained in the deep water run- ning with the perceived exertion threshold (PET) that we have pro- posed as alternative way to determine the maximal exercise inten- sity that allows stabilization of the physiological and psychophysical variables. METHODS Subject Thirteen young individuals from both genders with 23.0 ± 2.5 years of age, 170 ± 12 cm of height and 64.3 ± 15.3 kg of body mass participated in this study. All individuals signed the free con- sent form in order to participate in the study. None of them pre- sented previous experience with deep water running. Familiarization with deep water running and Borg scale The swimming pool used in this study was 15 m long and 2.5 m deep, with water temperature kept near to 28oC. It was consid- ered as favorable for the deep water running practice because the individuals could not touch the ground while maintaining their heads out of water. The volunteers underwent two to four sessions of deep water running familiarization. These sessions were aimed at instructing them the correct deep water running technique with the objective of presenting reliable results during tests for the determination of the parameters of the critical velocity model and PET. The deep water running technical instructions were given by two appraisers. One of them remained out of the swimming pool while the other monitored movements in the underwater environment. Once par- ticipants wore floating vests, their relaxed bodies trended to be positioned vertically in relation to the bottom of the swimming pool, with heads above the surface. During running, the trunk should simulate traditional running in order for the body to start dislocat- ing. The incursions amplitude of hip, knees and ankles joints should be large and fingers should be united in the sagittal plane in order to simulate swimming strokes. Their elbows should lie along their trunks. Each familiarization session lasted about 20 minutes. The exercise intensity was selected by the participant himself. They were advised to select a comfortable intensity. At the end of each session, shortly after a rest period, the participants were asked to cover 60 m at maximum water displacement velocity. When the deep water running technique seemed to be effective and the per- formance in both consecutive 60-m tests was not over than 5% different, the participants were able to start the tests to determine CritV, AnaerRC and PET. Still during familiarization, instructions about the use of the Borg 15-points scale(1) were given to be used in further stage. The scale was presented to participants during the deep water running prac- tice, and they attributed a numerical value in the corresponding scale to their general exertion perception in that moment. The scale presents verbal attributes next to the numbers to facilitate the se- lection so that participants would familiarize with the use of the scale and memorize the relation between the verbal attributes and the numerical values they should report (example: 7-very, very light; 17-very intense). Tests to determine CritV and AnaerRC All tests used to determine the parameters of the critical veloc- ity model were performed in different days. Before each test, the participant performed quick warm-up exercise in rhythm freely selected. The tests intensities (0.25 to 0.35 m/s) were established individually, so that exhaustion would only occur within range of one to ten minutes(15). In order to control velocity, cones were placed at the border of the swimming pool each 5 m. Sound signals, per- formed with the aid of a whistle, indicated the moment in which the individual should be aligned with the cone. When reaching the end of the swimming pool, the participant turned around to per- form the same exercise all the way back with no interruption. The voluntary abdication or the incapacity of maintaining the rhythm indicated for more than two cones were considered as exhaus- tion. In case two consecutive errors of rhythm were made, the participant was encouraged to correct them in the next cone. If the errors were successfully corrected, the test was carried out until one of the exhaustion criteria occurred. If a participant could not correct the rhythm requested, the distance covered by the last two cones (10 m) was deduced from the total distance covered. This variable, the test total time and the mean velocity were ad- justed to the equations predicted through the critical velocity mod- el in order to estimate CritV and AnaerRC. These equations are described below: Time = AnaerRC/(velocity – CritV) (equation 1) Distance = AnaerRC + (CritV. time) (equation 2) Velocity = CritV + [AnaerRC. (1/time)] (equation 3) PET estimation During the three exhaustive tests, the individuals were instruct- ed to report the perceived exertion according to the Borg 15-points scale(1), each 15 m covered. The angular coefficient of the linear regression between time as independent variable and individual perceived exertion values attributed during each test was deter- mined through linear regression. The angular coefficients of all straight lines (perceived exertion increase rate) obtained with this procedure were used to estimate the parameters of the linear lin- ear regression in function of the mean velocity of the three tests. PET was identified through the projection of this straight line on the velocity axis (x), in other words, it corresponded to the velocity in which the perceived exertion increase rate was “0”. Figures 1 and 2 exemplify these procedures in a representative individual. Statistical treatment CritV and AnaerRC were estimated through equations 1, 2 and 3 and through the linear and non-linear regression procedures. The respective standard error of estimate (SEE) and the determination errors (r2) associated to each equation were also estimated. The comparison between the CritV estimations provided by the three equations and the PET value was performed through repeated measures analysis of variance (ANOVA). The same statistical tests were adopted to compare the AnaerRC estimations. The Scheffé post hoc test for multiple comparisons was used to identify differ- ences between means. The Pearson correlation coefficient was used to verify associations between CritV and PET estimations. The significance level adopted in all analyses was of p < 0.05. RESULTS Table 1 shows the velocity mean values, distance and time of the three exhaustive tests used to determine CritV and AnaerRC. TABLE 1 Mean values (± SD) of velocity in exhaustive rectangular tests, distance and duration. Results are listed in increasing order of velocity used in each test Velocity (m.s-1) Distance (m) Time (s) 0.26 ± 0.02 184.6 ± 68.7 699.1 ± 274.5 0.31 ± 0.02 085.4 ± 29.5 271.2 ± 083.4 0.35 ± 0.03 049.3 ± 19.5 142.7 ± 058.3 Figure 1 shows that the increase on the perceived exertion along each test occurred linearly. In some cases, visual inspection showed that the first perceived exertion value reported did not fit to the linear tendency of the next points used in the linear regression. 8 Rev Bras Med Esporte _ Vol. 11, Nº 1 – Jan/Fev, 2005 When this occurred, the first point was excluded from the analy- sis. There were other occasions in which the last points repeated on 20, which is the maximum value. This generally occurred in the longest test. These points also weakened the linearity tendency of the rest of the curve. Thus, these points were also rejected so that the linear portion of the curve was described with better fitting of the regression equation. The PET obtained, according to example of a case shown by figure 2, was of 0.24 ± 0.03 m/s. Table 3 contains the correlation matrix between CritV and PET estimations. All correlations were high, but the correlations be- tween PET and the different CritV estimations, which ranged be- tween 0.85 and 0.88, are emphasized. -0,01 0 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08 0,15 0,2 0,25 0,3 0,35 0,4 LEP Fig. 2 – Perceived exertion threshold (PET) determination through linear relation between perceived exertion increase rate and deep water running velocity from a representative subject P er ce iv ed e xe rt io n in cr ea se r at e Velocity (m/s) Table 2 shows the mean values obtained for CritV, AnaerRC, the respective SEE and r2. Significant differences were observed be- tween the AnaerRC estimations (p < 0.05). The estimation provid- ed by equation 1 was different from that provided by equation 3 (p < 0.05). The CritV estimations were compared between each oth- er, along with the PET value (figure 3). No significant differences between each other were detected, although a tendency (p = 0.054) for this difference was observed, according to the preestablished significance level. TABLE 2 Critical velocity (CritV), anaerobic running capacity (AnaerRC), standard error of estimate (SEE) and determination coefficient (r2) provided by equations (1, 2 and 3) predicted through the critical power model (mean ± SD) Equation CritV AnaerRC r2 m/s SEE m SEE 1 0.23 ± 0.02 0.005 ± 0.000 19.1 ± 7.8* 3.3 ± 2.6 0.983 ± 0.026 2 0.24 ± 0.03 0.009 ± 0.006 17.0 ± 7.5* 3.9 ± 2.7 0.999 ± 0.001 3 0.25 ± 0.03 0.025 ± 0.041 15.0 ± 7.6* 2.7 ± 1.4 0.965 ± 0.042 * * * * * Significant difference in relation to equation 3. TABLE 3 Correlation matrix of the estimations of CritV resulting from the application of the three equations of the critical velocity model and PET Equation 1 Equation 2 Equation 3 PET Equation 1 1.00 Equation 2 0.98 1.00 Equation 3 0.92 0.98 1.00 PET 0.85 0.88 0.88 1.00 Note – All correlations are significant (p < 0.05). DISCUSSION There are no published studies that estimated CritV and Anaer- RC in deep water running in literature. Most works on the modali- ty, unlike the present study, use the static deep water running model, in which the individual remains attached to the border through the floating vest, and the exertion intensity is controlled by the steps frequency or through physiological markers, such as the heart rate. To review these studies, Wilder and Brennan(16), and Reilly et al.(17) are recommended. The parameters of the critical velocity model were determined in modalities related to the deep water running and their physio- logical meaning was established. According to Wakayoshi et al.(11,18) and Kokubun(12), the CritV obtained in swimming is not different from the anaerobic threshold, calculated through fixed concentra- tion of 4 mM of blood lactate in test with progressive velocities. The correlation between both variables in the studies mentioned was high (r = 0.85 to 0.89). Besides, the CritV seems to be coinci- dent with the maximal lactate steady-state directly measured(12,18). In running, Sid-Ali et al.(14) and Hill and Ferguson(13) found similar Fig. 1 – Increase on the perceived exertion along time in three exhaustive rectangular tests from a representative subject B or g sc al e (6 – 2 0) Time (s) 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 0 50 100 150 200 250 300 0,25 m/s 0,33 m/s 0,30 m/s Linear (0,25 m/s) Linear (0,30 m/s) Fig. 3 – Mean values of CritV estimated through the three equations of the critical velocity model and PET LEPEQ.3EQ.2EQ.1 ,27 ,26 ,25 ,24 ,23 ,22 ,21 IC95% média Ve lo ci ty (m /s ) mean Rev Bras Med Esporte _ Vol. 11, Nº 1 – Jan/Fev, 2005 9 results with regard to the meaning of CritV. In the second study, it was additionally demonstrated that the CritV is the maximal inten- sity possible to be maintained without the elevation of the VO2 up to its maximum value. The meaning of AnaerRC is not fully clear, once the expected correlation between this index and the peak plasma lactate after exhaustive test was not confirmed by Housh et al.(19). Still, the AnaerRC would theoretically be an indicative of anaerobic capacity(20). Albuquerque da Silva (data not published) showed that, in the deep water running, the CritV is not significantly different from the anaerobic threshold, determined through the fixed concentration of 3.5 mM in incremental test. The correlation between these vari- ables ranged from 0.79 to 0.90, depending on the equation used in the CritV estimation. Therefore, the physiological meaning of CritV in the deep water running seems to agree with that established in other modalities. The CritV estimations are not different from each other in the present study. Besides, the mean SEE ranged from 2 to 10% of the CritV. According to Hill(21), these are important conditions in which the parameter’s validity and accuracy are assured. Howev- er, a significant difference was observed (p = 0.05) between the AnaerRC estimations. The SEE associated to this variable ranged from 14 to 23%. Both factors made the AnaerRC validity and accu- racy infeasible as indicative of anaerobic capacity, according to cri- teria elaborated by Hill and Smith(20). Still, the adjustment of the experimental data to all equations was satisfactory, considering the high values of r2, above 0.96 on mean. Much has been discussed on the neurophysiological mechanisms responsible for the exertion perception. Cafarelli(5) summarizes the discussion into three thinking lines with their respective explana- tory hypotheses. Hypothesis 1 supports the feed forward primacy and the theory that a “copy” of the motor impulses is transmitted to the sensorial cortex, carrying the information about the degree of muscular activation. This information would be the main constit- uent of the exertion perception. Hypothesis 2 supports that the feedback coming from chemical and mechanical receptors in the joints, tendons, parts of the cardiorespiratory system and the skel- etal muscles would produce the peripheral sensation, translated into exertion perception in high levels of the central nervous sys- tem. Hypothesis 3 proposes that the afferent muscles would be compared to the efferent motors and that this comparison of infor- mation would cause immediate compensations by counterbalanc- ing the fatigue. This hypothesis supports that the sensorial cortex monitors both the efferent and the afferent signals in order to gen- erate the exertion perception. Several works have attempted to establish perceptive respons- es to exertion performed at metabolic thresholds (ventilatory, lac- tate)(22-24), to regulate lactacidemia in tests with variable loads through perceived exertion(25), and also to control other variables such as the heart rate and exercise intensity with the use of per- ceived exertion scales(26). Particularly, about the determination of the metabolic thresholds that has been objective of this study, the use of the exertion perception has presented some problems. The results from Mahon et al.(24) indicate that the perceived exertion rated by the Borg(1) 15-points scale at the ventilatory threshold in- tensity presents significant inter-individual variability, besides be- ing significantly different when children and adults were compared. However, a study by Hill et al.(22) demonstrated that the training affects the ventilatory threshold in adults, but does not modify the associated exertion perception. Weltman(27) gathered findings that emphasize that the 14 and 16-17 exertion perception is associated to the blood lactate fixed concentrations of 2.0 and 4.0 mM, fre- quently indicated as indicative of lactate and anaerobic thresholds, respectively. The approach presented in this study is different from the works mentioned above with regard to the detection of one of the thresholds – CritV – once we did not intend to establish a fixed value for the perceived exertion at this intensity. The proposal was to search for an original application for the Borg(1) scale, which takes into consideration the perceived exertion variation rate as alterna- tive to the CritV individualized determination. Indeed, as verified in our results, the perceived exertion value may change along a rect- angular exercise, what invalidates the attempt of establishing a fixed value associated to the exertion intensity in high-intensity exercise. For the same step frequency, the perceived exertion is signifi- cantly higher in the deep water running than in the field running(28). This probably occurs because in this situation, the VO2 and the heart rate are quite higher in the deep water running. Both vari- ables seem to be relevant in the prediction of the perceived exer- tion in deep water running through the multiple regression analy- sis(29). Other important variables are the minute-ventilation and ventilatory frequency (central) and the legs movement velocity (pe- ripheral). This information, integrated in the sensorial cortex and in the respiratory centers, probably determines the evolution of the perceived exertion in exhaustive rectangular tests and, thus, would influence the PET estimation. In the present study, PET provided good CritV indirect estima- tion, once both presented values close to each other, and the cor- relation between each other ranged from 0.85 to 0.88. It was pre- supposed that the perceived exertion would be increased in exhaustive tests (figure 1) along with the use of the AnaerRC. There- fore, the maximum intensity in which, theoretically, no increase on the perceived exertion would occur, would coincide with the inten- sity in which AnaerRC would not be used. In this context, the PET was determined based on the intersection of the linear regression of the relation between the perceived exertion increase rate and the velocity used in exhaustive tests (figure 2). One supposes that the regions from the central nervous system responsible for or- ganism’s perceptive representation is sensitive to the anaerobic reserve depletion, represented by the AnaerRC. In fact, this hy- pothesis is in agreement with Caffareli(5), Kostka and Cafarelli(3), and Robertson et al.(4), who established through their findings that H+ accumulation, both in the exercised muscles and blood, would be the main responsible for the increase on the perceived exer- tion. Most H+ accumulation in exercise is direct result of the anaer- obic metabolism mobilization. The role the feed forward plays in the perceived exertion mani- festation cannot be ignored, once the electromyographic activity that indicates the motor drive for the muscular activation increas- es along high-intensity rectangular exertions(6). This increase may be attributed to the incapacity of the motor stimuli at the begin- ning of the exercise to remain effective until the end of the exer- cise due to alterations on the peripheral cellular environment (ADP and H+ accumulation). The increase on the electromyographic ac- tivity would be a result of the communication between efferent and afferent information, whose comparison and integration would modulate the perceived exertion. The coincidence between the fatigue threshold and CritP(7) reinforces this hypothesis 3. From the practical point of view of the use of metabolic thresh- olds for the prescription of aerobic exercises, the PET identifica- tion seems to be promising, once the results of this study suggest that the increase on the exertion perception would somehow be determined through the use of the anaerobic capacity. A limitation of this study was the lack of an objective measure of the use and depletion of the anaerobic energetic supplies concomitantly to the perceived exertion data collecting. Although the PET identifies the same metabolic threshold already described through other meth- ods based on the heart rate, ventilation and on the blood lactate concentration, it showed to be a useful tool for the prescription and control of the aerobic training. The use of any equipment or the employment of invasive procedures would be useless. How- ever, its physiological meaning and applicability on the aerobic train- ing deserve further investigations. 10 Rev Bras Med Esporte _ Vol. 11, Nº 1 – Jan/Fev, 2005 CONCLUSIONS The results of the present study suggest that the perceived ex- ertion may be used in the CritV determination. The PET was not significantly different in relation to the CritV (p > 0.05). Besides, the correlations between both variables were quite high (r = 0.85 to 0.88). Thus, PET seems to represent the maximum exercise intensity where physiological and psychophysical variables would find stability. New studies must be conducted in order to assess the PET validity in other experimental conditions. All the authors declared there is not any potential conflict of inter- ests regarding this article. REFERENCES 1. Borg GAV. Psychophysical bases of perceived exertion. Med Sci Sports Exerc 1982;14:377-81. 2. Gaesser GA, Poole D. The slow component of oxygen uptake kinetics in hu- mans. Exerc Sport Sci Rev 1996;24:35-70. 3. Kostka CE, Cafarelli E. Effect of pH on sensation and vastus lateralis electromyo- gram during cycling exercise. J Appl Physiol 1982;52:1181-5. 4. Robertson RJ, Falkel JE, Drash AM, Swank AM, Metz KF, Spungen SA, et al. Effects of blood pH on peripheral and central signals of perceived exertion. Med Sci Sports Exerc 1986;18:114-22. 5. Cafarelli E. Peripheral contributions to the perception of effort. Med Sci Sports Exerc 1982;14:382-9. 6. DeVries HA, Moritani T, Nagata A, Magnussen K. The relation between critical power and neuromuscular fatigue as estimated from electromyographic data. Ergonomics 1982;25:738-91. 7. Le Chevalier JM, Vandewalle H, Thépaut-Mathieu C, Stein JF, Caplan L. Local critical power is an index of local endurance. Eur J Appl Physiol 2000;81:120-7. 8. Monod H, Scherrer J. The work capacity of a synergic muscular group. Ergo- nomics 1965;8:329-38. 9. Moritani T, Nagata A, DeVries HA, Muro M. Critical power as a measure of phys- ical work capacity and anaerobic threshold. Ergonomics 1981;24:339-50. 10. Hughson RL, Orok CJ, Staudt LE. A high velocity treadmill running test to asses endurance running potential. Int J Sports Med 1984;5:23-5. 11. Wakayoshi K, Yoshida T, Udo M, Kasai T, Moritani T, Mutoh Y, et al. A simple method for determining critical speed as swimming fatigue threshold in com- petitive swimming. Int J Sports Med 1992;13:367-71. 12. Kokubun E. Velocidade crítica com estimador do limiar anaeróbio na natação. Rev Paul Educ Fís 1996;10:5-20. 13. Hill DW, Ferguson CS. A physiological description of critical velocity. Eur J Appl Physiol 1999;79:290-3. 14. Sid-Ali B, Vandewalle H, Chaïr K, Moreaux A, Monod H. Lactate steady state and distance-exhaustion time relationship in running. Arch Int Physiol Biochim 1991; 99:297-301. 15. Poole DC. Letter to the editor-in-chief. Med Sci Sports Exerc 1986;26:1335-40. 16. Wilder RP, Brennan DK. Physiological responses to deep water running in ath- letes. Sports Med 1993;16:374-80. 17. Reilly T, Dowzer CN, Cable NT. The physiology of deep-water running. J Sports Sci 2003;21:959-72. 18. Wakayoshi K, Yoshida T, Udo M, Harada T, Moritani T, Mutoh Y, et al. Does critical swimming velocity represent exercise intensity at maximal lactate steady- state? Eur J Appl Physiol 1993;66:90-5. 19. Housh TJ, Johnson GO, McDowell SL, Housh DJ, Pepper ML. The relationship between anaerobic running capacity and peak plasma lactate. J Sports Med Phys Fitness 1992;32:117-22. 20. Hill DW, Smith JC. A method to ensure the accuracy of estimates of anaerobic capacity derived using the critical power concept. J Sports Med Phys Fitness 1994;34:23-37. 21. Hill DW. The critical power concept: a review. Sports Med 1993;16:237-54. 22. Hill DW, Cureton KJ, Grisham SC, Collins MA. Effect of training on the rating of perceived exertion at the ventilatory threshold. Eur J Appl Physiol 1987;56:206- 11. 23. Kolkhorst FW, Mittelstadt SW, Dolgener FA. Perceived exertion and blood lac- tate concentration during graded treadmill running. Eur J Appl Physiol 1996;72: 272-7. 24. Mahon AD, Gay JA, Stolen KQ. Differentiated ratings of perceived exertion at ventilatory threshold in children and adults. Eur J Appl Physiol 1998;18:115-20. 25. Stoudmire NM, Wideman L, Pass KA, McGinnes CL, Gaesser GA, Weltman A. The validity of regulating blood lactate concentration during running by ratings of perceived exertion. Med Sci Sports Exerc 1996;28:490-5. 26. Ceci R, Hassmén P. Self-monitored exercise at different RPE intensities in tread- mill vs field running. Med Sci Sports Exerc 1991;23:732-8. 27. Weltman A. The blood lactate response to exercise. 1st rev. ed. Champaign: Human Kinetics, 1995. 28. Brown SP, Chitwood LF, Beason KR, McLemore DR. Perceptual responses to deep water running and treadmill exercise. Percept Mot Skills 1996;83:131-9. 29. Brown SP, Chitwood LF, Beason KR, McLemore DR. Physiological correlates with perceived exertion during deep water running. Percept Mot Skills 1996;83: 155-62.