Introduction
Physical training load can be described as the dose of training completed by an athlete during an exercise bout (18). There are a variety of methods available to quantify the physical training load undertaken by athletes. Traditionally, training programs have been described based on a measure of external training load, which is a measure of training load independent of individual internal characteristics. For example, in an endurance athlete, a coach may prescribe a physical training session with respect to a desired training distance (m) or training time (minutes) (e.g., 10-km run in 40 minutes). However, it is the relative physiological stress imposed on the athlete (internal training load) and not the external training load that determines the stimulus for training adaptation (31). Indeed, if 2 athletes with different fitness characteristics and performance abilities completed the same external training load, 1 athlete would inevitably find the training session more difficult than the other. Therefore, it is important for coaches to monitor internal training load so that training programs can be tailored to the needs of individual athletes.
There have been many attempts by researchers to develop a suitable method for quantifying internal training load that incorporates both training duration and individual training intensity. At present, the most commonly used methods for quantifying internal training load use heart rate (HR) as a measure of exercise intensity. For example, Banister et al. (5) proposed the TRIMP method as a means to quantify the internal training load undertaken by athletes. This method is based on a product of training duration and a weighting factor determined by an individual's HRmean for each exercise bout (4,5). Other similar methods have calculated a measure of internal training load using the knowledge of HR values around common inflection points such as ventilatory (20) or lactate thresholds (17). However, the application of HR as a measure of training load has several limitations. For example, HR response may be a relatively poor method for evaluating intensity during very high-intensity exercise such as plyometrics, resistance training, and interval training. Furthermore, HR methods can require the use of expensive equipment and operators with technical expertise and knowledge for interpretation.
To overcome the limitations associated with the use of HR information, Foster et al. (12) proposed a simple method (session-RPE [sRPE]) for quantifying internal training load in athletes. This method requires subjects to subjectively rate the intensity of the entire training session using a rating of perceived exertion (RPE) according to the category ratio scale (CR-10 scale) of Borg et al. (7). This intensity value is then multiplied by the training duration (minutes) to create a single measure of internal training load in arbitrary units. It has been suggested that the sRPE method may better reflect the internal training load placed on athletes than either HR or blood lactate measures during non–steady-state exercise (8) and has been shown to compare favorably with more complicated methods of quantifying internal training load in endurance (11,25), team sport (1,17), and resistance-trained athletes (9).
The previous methods have been shown to be useful for quantifying the internal training load in athletes across a range of exercise intensities and activities. However, at present, only 2 studies reported on the reproducibility of physiological data and sRPE during standardized training sessions (14,28) and only 1 has compared these methods using measures of oxygen consumption (V[Combining Dot Above]O2) (14). This study demonstrated strong correlations between V[Combining Dot Above]O2 (r = 0.98), HR (r = 0.96), and sRPE (modified 10-point scale) (r = 0.88) measures of exercise intensity in repeated trials of 30-minute steady-state exercise at 3 different intensity levels (easy effort = ∼40–50% V[Combining Dot Above]O2peak, moderate effort = ∼60–70% V[Combining Dot Above]O2peak, and hard effort = ∼80–90% V[Combining Dot Above]O2peak). Moreover, the correlation between the modified 10-point RPE scale and V[Combining Dot Above]O2 was suggesting good construct validity of this measure. More recently, Scott et al. (28) reported that both CR-10 and CR-100-derived sRPE methods have good construct validity for assessing training load in AF during 38 Australian Football training sessions. Moreover, they also reported poor levels of reliability for both the CR-10 and CR-100 sRPE during short bouts of intermittent running over a narrow range of exercise intensities in the field. To date, no study has compared the validity and reliability of the common methods for quantifying training load using a well-controlled laboratory study. Therefore, the purpose of this investigation was to compare the criterion validity and test-retest reliability of common methods for quantifying internal training load. It was hypothesized that the sRPE would be a valid and reliable measure of training load for endurance exercise.
Methods
Experimental Approach to the Problem
Each subject completed 18 randomly assigned physical training sessions during a 6-week period. All physical training sessions and testing procedures were undertaken in the environmentally controlled Human Performance Research Laboratory. In the month preceding the investigation, all subjects were familiarized with the equipment and all testing methods used in the study. The physical training sessions consisted of 3 steady state (SS) and 3 interval (INT) sessions that were each completed 3 times during a 6-week period. All physical training sessions were supervised by the principal researcher and were completed on an electronically braked cycle ergometer (ERG 601; Bosch, Berlin, Germany). Steady-state sessions were 18 minutes in duration and were performed at 35, 50, and 65% of maximum work rate (Wmax). These exercise intensities were designed to provide training sessions that would be performed at easy (RPE, <3), moderate (RPE, 3–5), and difficult (RPE, >5) training intensities. Interval sessions were performed at 50, 60, and 70% of Wmax with a work to rest ratio of 1:0.5 and matched for total work with the 50% SS session. These physical training sessions were designed to measure the physiological and psychological effects of interval training compared with steady-state training. At least 24 hours of recovery was required before each training session. A summary of the exercise protocol can be seen in Table 1.
;)
Table 1: Summary of exercise protocols used in this study.
Oxygen consumption (V[Combining Dot Above]O2) and HR were measured continuously throughout all physical training sessions. The mean V[Combining Dot Above]O2 (V[Combining Dot Above]O2mean) was used as the criterion measure of internal training load. Blood lactate concentration and RPE measures were taken after 6, 12, and 18 minutes for the SS sessions and after each interval for the INT sessions. After each exercise bout, subjects were required to rate the perceived difficulty of the entire training session (global RPE) according to the category ratio (CR-10) scale of Borg et al. (7). No feedback regarding workload or physiological response was provided to the subject during any exercise session.
Subjects
Ten (5 men and 5 women) recreational athletes (mean ± SD, V[Combining Dot Above]O2max: 37.0 ± 4.3 ml·kg−1·min−1; age: 23.8 ± 8.4 years; age range: 18-43 years) volunteered to participate in this study. The subjects trained at least 3 times a week in addition to competing in their chosen sport at least once per week. Before the commencement of the study, all subjects were fully informed of the potential risks and benefits associated with participation. Written informed consent was obtained by each subject, and ethical approval was granted by the University Human Research Ethics Committee for all experimental procedures.
Procedures
Each subject performed an incremental cycle test to exhaustion to establish maximal oxygen consumption (V[Combining Dot Above]O2max), Wmax, maximum heart rate (HRmax), and ventilatory thresholds (VTs). An incremental test was also performed after the investigation period to adjust for changes in physiological characteristics. The incremental cycle test was performed on an electronically braked cycle ergometer (ERG 601; Bosch). Subjects were instructed to standardize food and fluid intake before each testing session.
The incremental cycle test began with a power output of 20 W and the workload increased by 25 W·min−1 until volitional fatigue (21). The subjects were required to keep the pedal cadence of 70–80 revolutions per minute (rpm) constant throughout the duration of the test. The test was terminated when a pedal cadence could not be maintained at ≥70 rpm. Oxygen uptake was measured continuously throughout the incremental test using a Physio-Dyne Gas Analysis System (Physio-Dyne; Fitness Instrument Technologies, Quogue, NY, USA). The gas analysis system was calibrated before and after each test with reference and calibration gases of known concentrations. The pneumotach was calibrated with ambient air using a 3 L syringe (Hans Rudolph, Inc., Kansas City, MO, USA). Maximum oxygen uptake was considered the highest 30-second average of oxygen volume recorded during the last minute of the exercise. Heart rate was recorded throughout the test via Polar Team HR monitors (Polar Electro Oy, Kempele, Finland) and downloaded to a computer using Polar Advantage Software (Polar). Maximum HR was taken as the highest HR recorded during the incremental test. Wmax was determined using the equation:
where Wfinal (W) is the power output during the final stage completed, t(s) is the amount of time reached in the final uncompleted stage, T(s) is the duration of each stage, and Winc is the workload increment (13).
The VT was determined using the criteria of an increase in both V[Combining Dot Above]E × V[Combining Dot Above]O2−1 with no concomitant increase in V[Combining Dot Above]E × V[Combining Dot Above]CO2−1. The respiratory compensation point (RCP) was determined using the criteria of an increase in both the V[Combining Dot Above]E × V[Combining Dot Above]CO2−1 (2). These thresholds were visually determined by 3 experienced investigators, and a consensus-derived value was used. Capillarized blood samples (30 μl) were taken from the fingertip to assess [BLa−] and analyzed using a Lactate Scout Portable Lactate Analyser (SensLab GmbH, Leipzip, Germany).
To monitor the psychophysiological recovery process between each physical training session, each subject was required to complete the Total Quality of Recovery (TQR) Questionnaire and a Well-Being Questionnaire before each exercise session. The TQR Questionnaire was applied according to the recommendations of Kenttä and Hassmén (19). The Well-Being Questionnaire required subjects to record subjective ratings of quality of sleep, fatigue, stress, and muscle soreness on a Likert scale of 1-7 from very, very low, or good (1 point) to very, very high, or bad (7 point) (15). These questionnaires were implemented to ensure that each subject had recovered from the previous training session. If a questionnaire revealed that a subject was in a negatively recovered state, a further 24-hour recovery was prescribed before undertaking the training session. Data were not reported.
A variety of methods were used to quantify the internal training load of the subjects during each exercise bout. For example, the HR-based TRIMP method proposed by Banister (4) (Banister's TRIMP) was used to quantify internal training load. This method was calculated using the following equation:
where D = duration of training session, b = 1.67 for women and 1.92 for men and (ΔHR ratio) is determined using the following equation:
where HRrest = the average HR during rest and HRex = the average HR during exercise (4,23).
The HR-based method proposed by Lucia (20) (Lucia's TRIMP) was also used to quantify internal training load in this study. This method involves multiplying the time spent in 3 HR zones (zone 1: below VT, zone 2: between VT and RCP; and zone 3: above RCP) by a coefficient relative to each intensity zone (k = 1 for zone 1, k = 2 for zone 2, and k = 3 for zone 3) and then summating the results. This method is summarized by the following equation:
The sRPE method proposed by Foster et al. (12) was also used as a method for quantifying the internal training load. This method requires athletes to subjectively rate the intensity of the entire training session using a RPE according to the category ratio scale (CR-10) of Borg et al. (7).
The RPE value is then multiplied by the total duration (minutes) of the training session. To ensure the athletes report a RPE for the entire training session (global RPE), the measure of RPE is taken 30 minutes after the completion of the session. Standard instructions and anchoring procedures were explained during the familiarization process (24). This method can be calculated using the following equation:
where D is the duration of the entire training session and RPE is the global RPE (Borg CR-10) (12).
Similarly, session-RPE(worktime) (sRPEwt) was also used as a method to quantify internal training load. This method differs from sRPE in that the global RPE is only multiplied by the sum of duration actually spent performing physical training and not the duration of the entire training session. This method has previously been used to quantify training load in athletes undertaking resistance training (9,30). The total oxygen consumption (total V[Combining Dot Above]O2) was taken as the criterion measure of internal training load. This method was calculated as the mean relative V[Combining Dot Above]O2 for each exercise bout multiplied by the exercise duration. In addition to these methods for quantifying internal training load, the total work (kilojoules) performed during each training session was calculated and used as a measure of external training load.
Statistical Analyses
All data are expressed as mean ± SD unless otherwise stated. A 1-way analysis of variance (ANOVA) was used on each dependent variable to determine the session-to-session variability in HR, [BLa−], V[Combining Dot Above]O2, and RPE. When a significant F value was found, the Scheffe post hoc test was applied. As recommended by Hopkins (16), the typical error as a coefficient of variation (CV) and intraclass correlation coefficients (ICCs) were calculated to establish the reliability of HR, [BLa−], V[Combining Dot Above]O2, and RPE responses. The CV (typical error) is defined as the ratio of the SD to the mean (16). The ICC (3,1) was derived from a mixed model according to the methods of Shrout and Fleiss (29). Pearson's correlations were used to examine the relationships of exercise intensity (expressed as a percentage of HRmax) and V[Combining Dot Above]O2max with the corresponding SD. The following criteria were adopted to interpret the magnitude of the correlation (r) between measures: <0.1 trivial, 0.1–0.3 small, 0.3–0.5 moderate, 0.5–0.7 large, 0.7–0.9 very large, and 0.9–1.0 almost perfect. The level of statistical significance was set at 0.05.
Results
Maximal oxygen uptake (V[Combining Dot Above]O2max) for this group of subjects was significantly different after the 6-week study (pre: 37.0 ± 4.3 ml·kg−1·min−1 vs. post: 40.3 ± 3.2 ml·kg−1·min−1; p ≤ 0.05). Physiological and psychological data were recorded from a total of 180 training sessions. Individual correlations between the total V[Combining Dot Above]O2 and commonly used methods for quantifying physical training load were determined from 18 individual training sessions and are shown in Table 2.
Table 2: Individual correlations coefficients between total V[Combining Dot Above]O2 and common methods for quantifying physical training load.
One-way ANOVA revealed significant main effects between methods for quantifying physical training load. Scheffe post hoc analysis demonstrated significant differences between sRPE and external work (p < 0.001) and sRPEwt and external work (p ≤ 0.05).
Individual correlations between %V[Combining Dot Above]O2max and each of the methods for quantifying training intensity were also determined from 18 individual training sessions. The individual correlation coefficients for methods to quantify training intensity can be seen in Table 3.
Table 3: Individual correlations coefficients between %V[Combining Dot Above]O2max and common methods for quantifying training intensity.
The intrarater reliability for each of the methods for quantifying training load and training intensity between trials 1–2 and 2–3 are shown in Tables 4 and 5, respectively. Using the %CV and %ICC from trials 2–3, each of the HR methods for quantifying training load showed a poor level of reliability (Banister's TRIMP [15.6% CV, 0.818 ICC] and Lucia's TRIMP [10.7% CV, 0.733 ICC]). A poor level of reliability was also shown for sRPE (28.1% CV, 0.763 ICC) and sRPEwt (28.1% CV, 0.735 ICC). A good level of reliability was shown for HR as a measure of exercise intensity (3.9% CV, 0.862% ICC). However, only a moderate level of reliability was shown for %V[Combining Dot Above]O2max (6.1% CV, 0.835 ICC) and RPE 6-20 (8.5% CV, 0.765 ICC). A poor level of reliability was shown for the CR-10 scale (28.1% CV, 0.766 ICC).
Table 4: Measures of reliability (with 95% confidence intervals) of methods for evaluating training load and intensity from trial 1 to trial 2 on a cycle ergometer (n = 10).*
Table 5: Measures of reliability (with 95% confidence intervals) of methods for evaluating training load and intensity from trial 2 to trial 3 on a cycle ergometer (n = 10).*
Discussion
The purpose of this study was to evaluate the criterion validity and reliability of common methods for quantifying training load during endurance-based exercise. A number of previous studies have compared subjective (RPE-based) and objective (HR-based) methods for quantifying internal training load (10,11,17,26). However, this is the first study to compare internal vs. external methods for quantifying training load against measures of the total oxygen consumption (V[Combining Dot Above]O2).
The present results reveal large to almost perfect correlations between each method for quantifying training load (internal and external) and total V[Combining Dot Above]O2. The strongest correlation occurred between the external work and total V[Combining Dot Above]O2. This result was expected considering the well-established strong positive relationship between these variables. The HR methods of both Banister and Lucia also produced strong positive correlations with the total V[Combining Dot Above]O2. These results suggest that the HR methods may provide good alternative methods for quantifying training load where measures of external work are not easily defined (e.g., swimming and running). Interestingly, the Banister and Lucia TRIMP produced significantly lower correlations with the total V[Combining Dot Above]O2 than did measures of HR alone when compared with %V[Combining Dot Above]O2max. This finding suggests that the calculation of a TRIMP score may harbor an increased potential for error associated with the strength of the weighting factors and/or the calculation of VT and RCP. Further refinement of these methods may improve the use of HR as a method for quantifying training load.
In agreement with the recent research (14), large-to-very large correlations were also observed between sRPE and total V[Combining Dot Above]O2. Furthermore, a significant difference was also observed between sRPE and external work when compared with the total V[Combining Dot Above]O2. The reduced strength in the correlation between sRPE and V[Combining Dot Above]O2 compared with external work and the HR-based methods may be attributed to psychobiological nature of the CR-10 scale for monitoring exercise intensity (6). For example, the CR-10 scale was developed to reflect physiological (oxygen uptake, ventilation, HR, circulating glucose concentration, and glycogen depletion) and psychological responses to exercise. Therefore, the reduced strength in the correlation between sRPE and total V[Combining Dot Above]O2 may suggest that factors other than V[Combining Dot Above]O2 affect global training load. For example, the muscle damage caused from a previous training bout may influence perception of effort (22). This investigation also showed the correlation between sRPE and total V[Combining Dot Above]O2 (r = 0.75) to improve with the calculation of sRPEwt (r = 0.79). Previous investigations have shown sRPEwt to be a valid method for quantifying internal training load in resistance training (9,30). The present findings further support the use of sRPEwt as a valid method for quantifying internal training load in endurance-based exercise.
Previous studies have suggested that RPE may provide a more valid method for quantifying training intensity when both aerobic and anaerobic systems are activated (e.g., intermittent exercise) (3,17). This investigation revealed a significantly greater correlation between HR and %V[Combining Dot Above]O2max compared with RPE and %V[Combining Dot Above]O2max (r = 0.92 and r = 0.80, respectively). These results suggest that HR provides the most accurate field-based method for assessing exercise intensity when exercise is performed at an intensity less than V[Combining Dot Above]O2max. However, because the obtainment of HR information requires expensive equipment and can be tedious to interpret, the present authors support the use of RPE as a valid and practical alternative for quantifying exercise intensity. Furthermore, previous studies have suggested RPE to be a valid method for quantifying exercise intensity during ultra-high intensity exercise (e.g., resistance training, plyometric training, and intervals training at an intensity greater than V[Combining Dot Above]O2max) where HR measures may be inappropriate (9,11,30). Collectively, these findings suggest that RPE to be a versatile method for quantifying training intensity. Further studies are required investigating the validity of methods for quantifying training intensity in athletes performing supramaximal exercise before this hypothesis can be confirmed.
The second purpose of this study was to determine the reliability of common methods for quantifying training load. The results show poor levels of reliability for each of the HR-based methods (Banister's TRIMP [15.6% CV, 0.818% ICC] and Lucia's TRIMP [10.7% CV, 0.733% ICC]) for quantifying internal training load. Interestingly, the use of HR as a measure of exercise intensity showed good levels of reliability (3.9% CV, 0.862% ICC). These results indicate that the weighting factors used to determine the TRIMP scores may also reduce the reproducibility of these HR-based methods. Therefore, a refinement in these weighting factors may increase the reliability of these methods.
The present results also show poor levels of reliability for sRPE as a measure of internal training load (28.1% CV, 0.763% ICC). These results may be attributed to the poor reliability of the CR-10 scale for quantifying training intensity (28.1% CV, 0.766% ICC). The present findings are in accordance with several previous investigations showing sRPE to have poor reliability (14.9% CV) when used to quantify internal training load in athletes undertaking resistance training (9), (31.9% CV) in small sided soccer games (27), and standardized intermittent running drills in Australian football players (28). There are several possible explanations for the poor level of reliability in this study, including the possibility that more trials were required to familiarize the subjects with the scale. This is supported by the small improvement in %CV and ICCs from trials 2–3 (Table 5) compared with trials 1–2 (Table 4). It is also possible that fitness levels may affect the reliability of sRPE measures, but this contention is yet to be elucidated.
Interestingly, the RPE 6-20 scale showed moderate levels of reliability (8.5% CV, 0.765% ICC). The improvement in reliability in the RPE 6-20 scale may be attributed to the fact that the 6-20 scale is a ratio scale and/or is more sensitive than the CR-10 scale (e.g., 15-point scale compared with a 10-points scale). These results indicate that the reliability of sRPE may be improved if the measure of intensity was based on the RPE 6-20 scale. However, because the CR-10 scale has been suggested to be particularly useful for monitoring intensity during high-intensity, intermittent-based activities (17), substituting the CR-10 scale for the RPE 6-20 scale may only benefit athletes undertaking endurance-based exercise at an intensity less than V[Combining Dot Above]O2max.
In summary, this study showed that measures of external work, HR-based methods, and RPE-based methods each provided a valid method for quantifying training load in endurance-based exercise. However, of these methods, external work correlate best with the total V[Combining Dot Above]O2. These results suggest that a measure of external work has the greatest aptitude for measuring training load in recreational endurance athletes. A comparison between our internal training load methods showed the HR-based methods to correlate better with the total V[Combining Dot Above]O2 when compared with the RPE-based methods. However, when interpreting these findings, it is important to remember that factors other than V[Combining Dot Above]O2 affect global training load. External work, HR, and RPE were all valid methods for quantifying training intensity during endurance-based exercise at an intensity less than V[Combining Dot Above]O2max. However, previous studies have suggested that RPE may be more appropriate for quantifying training intensity during supramaximal exercise. Future work should assess the validity and reliability of sRPE for assessing training loads when work is greater than V[Combining Dot Above]O2max.
Poor levels of reliability were reported for each of the HR-based TRIMP methods for quantifying internal training load. Because HR alone was shown to have good reliability, the poor level of reliability in the TRIMP methods was attributed to inappropriate weighting factors or errors in determining VT and RCP. The sRPE was also shown to have poor reliability, most likely because of the poor reliability of the CR-10 scale for quantifying training intensity. Substituting the CR-10 scale with the 6-20 scale improves the reliability of the sRPE method. Regardless of the scale used however, and in accordance with our previous work (28), it may be that the relatively poor reliability of both sRPE scales used in this study that sRPE may not be sensible to detecting small changes in exercise intensity.
Practical Applications
There are a variety of methods available to coaches for determining the individual stress placed on an athlete from an exercise bout. At present, measures of V[Combining Dot Above]O2 are thought to provide the most valid method for quantifying the internal training load in athletes undertaking endurance-based exercise. Unfortunately, obtaining direct V[Combining Dot Above]O2 information in the field is mostly impractical. Of the alternative methods for quantifying training load, measures of external work still remain the most related to V[Combining Dot Above]O2 and therefore offers the best method for quantifying training load in athletes undertaking endurance-based exercise (e.g., road cycling). External work also provides athletes with instantaneous feedback regarding training intensity without the lag time associated with HR or the familiarization required with RPE. However, if an athlete is undertaking endurance-based exercise where external work cannot be easily calculated (e.g., running, rowing, and cross-country skiing), then HR-based methods provide the most valid alternative.
The RPE-based methods are the simplest and cheapest of all methods and may be the most valid methods for quantifying training load in athletes undertaking supramaximal exercise. However, the RPE-based methods correlate least with V[Combining Dot Above]O2 during endurance-based exercise below V[Combining Dot Above]O2max, are subjective, and require greater familiarization. In conclusion, each of the methods for quantifying training load presented in this paper has advantages and limitations in quantifying the training load from an exercise bout. Weighing up the cost of the equipment, the level/goals of the athlete, and the intensity performed by the athlete will determine which method is most appropriate for each individual.
References
1. Alexiou H, Coutts AJ. A comparison of methods used for quantifying internal training load in women soccer players. Int J Sports Physiol Perform 3: 320–330, 2008.
2. Amann M, Subudhi AW, Foster C. Predictive validity of ventilatory and lactate thresholds for cycling time trial performance. Scand J Med Sci Sports 16: 27–34, 2006.
3. Bangsbo J. The physiology of soccer with special reference to intense intermittent exercise. Acta Physiol Scand Suppl 151: 1–156, 1994.
4. Banister EW. Modeling elite athletic performance. In: Physiological Testing of Elite Athletes. Green H.J., McDougal J.D., Wenger H.A., eds. Champaign, IL: Human Kinetics, 1991. pp. 403–424.
5. Banister EW, Calvert TW, Savage MV, Bach T. A systems model of training for athletic performance. Aust J Sports Med 7: 57–61, 1975.
6. Borg G. Psychophysical bases of perceived exertion. Med Sci Sports Exerc 14: 377–381, 1982.
7. Borg GAV, Hassmen P, Langerstrom M. Perceived exertion in relation to heart rate and blood lactate during arm and leg exercise. Eur J Appl Physiol 65: 679–685, 1985.
8. Coutts AJ, Rampinini E, Marcora S, Castagna C, Impellizzeri FM. Heart rate and blood lactate correlates of perceived exertion during small-sided soccer games. J Sci Med Sport 12: 79–84, 2009.
9. Day M, McGuigan MR, Brice G, Foster C. Monitoring exercise intensity during resistance training using the session-RPE scale. J Strength Cond Res 18: 353–358, 2004.
10. Foster C. Monitoring training in athletes with reference to overtraining syndrome. Med Sci Sports Exerc 30: 1164–1168, 1998.
11. Foster C, Florhaug JA, Franklin J, Gottschall L, Hrovatin LA, Parker S, Doleshal P, Dodge C. A new approach to monitoring exercise training. J Strength Cond Res 15: 109–115, 2001.
12. Foster C, Hector LL, Welsh R, Schrager M, Green MA, Snyder AC. Effects of specific versus cross-training on running performance. Eur J Appl Physiol Occup Physiol 70: 367–372, 1995.
13. Halson SL, Bridge MW, Meeusen R, Busschaert B, Gleeson M, Jones DA, Jeukendrup AE. Time course of performance changes and fatigue markers during intensified training in cyclists. J Appl Physiol (1985) 93: 947–956, 2002.
14. Herman L, Foster C, Maher MA, Mikat RP, Porcari JP. Validity and reliability of the session RPE method for monitoring exercise training intensity. S Afr J Sports Med 18: 14–17, 2006.
15. Hooper SL, Mackinnon LT, Hanrahan S. Mood states as an indication of staleness and recovery. Int J Sports Psychol 28: 1–12, 1997.
16. Hopkins WG. Measures of reliability in sports medicine and science. Sports Med 30: 1–15, 2000.
17. Impellizzeri FM, Rampinini E, Coutts AJ, Sassi A, Marcora SM. Use of RPE-based training load in soccer. Med Sci Sports Exerc 36: 1042–1047, 2004.
18. Impellizzeri FM, Rampinini E, Marcora SM. Physiological assessment of aerobic training in soccer J Sports Sci 23: 583–592, 2005.
19. Kenttä G, Hassmén P. Overtraining and recovery: A conceptual model. Sports Med 26: 1–16, 1998.
20. Lucía A, Hoyos J, Santalla A, Earnest C, Chicharro JL. Tour de France versus Vuelta a Espana: Which is harder? Med Sci Sports Exerc 35: 872–878, 2003.
21. Lucia AL, Parado A, Durantez A, Hoyos J, Chicharro JL. Physiological differences between professional and elite road cyclists. Int J Sports Med 19: 342–348, 1998.
22. Marcora SM, Bosio A. Effect of exercise-induced muscle damage on endurance running performance in humans. Scand J Med Sci Sports 17: 662–671, 2007.
23. Morton RH, Fitz-Clarke JR, Banister EW. Modeling human performance in running. J Appl Physiol (1985) 69: 1171–1177, 1990.
24. Noble BJ, Robertson RJ. Percieved Exertion. Champaign, IL: Human Kinetics, 1996.
25. Oliveira Borges T, Bullock N, Duff C, Coutts AJ. Methods for quantifying training in sprint kayak. J Strength Cond Res 28: 474–482, 2013.
26. Ozkan A, Kin-Isler A. The reliability and validity of regulating exercise intensity by rating of perceived exertion in step dance lessions. J Strength Cond Res 21: 296–300, 2007.
27. Rampinini E, Impellizzeri FM, Castagna C, Abt G, Chamari K, Sassi A, Marcora SM. Factors influencing physiological responses to small-sided soccer games. J Sports Sci 25: 659–666, 2007.
28. Scott TJ, Black CR, Quinn J, Coutts AJ. Validity and reliability of the session-RPE method for quantifying training in Australian football: A comparison of the CR10 and CR100 scales. J Strength Cond Res 27: 270–276, 2013.
29. Shrout PE, Fleiss JL. Intraclass correlations: Uses in assessing rater reliability. Psychol Bull 86: 420–428, 1979.
30. Sweet TW, Foster C, McGuigan MR, Brice G. Quantitation of resistance training using the session rating of perceived exertion method. J Strength Cond Res 18: 796–802, 2004.
31. Viru A, Viru M. Nature of training effects. In: Exercise and Sport Science. Garret W.E., Kirkendall D.T., eds. Philadelphia, PA: Lippincott Williams and Wilkins, 2000. pp. 67–95.
