You are here

Article Text

Article menu

Download PDFPDF

Original articles
A novel submaximal cycle test to monitor fatigue and predict cycling performance
  1. R P Lamberts,
  2. J Swart,
  3. T D Noakes,
  4. M I Lambert
  1. UCT/MRC Research Unit for Exercise Science and Sports Medicine, Department of Human Biology, University of Cape Town, Sport Science Institute of South Africa, Newlands, South Africa
  1. Correspondence to Robert Lamberts, UCT/MRC Research Unit for Exercise Science and Sports Medicine, Sport Science Institute of South Africa, PO Box 115, Newlands 7725, South Africa; robert.lamberts{at}uct.ac.za

Abstract

Objective The purpose of this study was to determine the reliability and predictive value of performance parameters, measured by a new novel submaximal cycle protocol, on peak power and endurance cycling performance in well-trained cyclists.

Methods Seventeen well-trained competitive male road racing cyclists completed four peak power output (PPO) tests and four 40-km time trials (40-km TT). Before each test, all cyclists performed a novel submaximal cycle test (Lamberts and Lambert Submaximal Cycle Test (LSCT)). Parameters associated with performance such as power, speed, cadence and rating of perceived exertion (RPE) were measured during the three stages of the test when cyclists rode at workloads coinciding with fixed predetermined heart rates. Heart rate recovery (HRR) was measured after the last stage of the test.

Results Parameters measured during the second and third stages of the LSCT were highly reliable (intraclass correlation range: R=0.85−1.00) with low typical error of measurements (range: 1.3−4.4%). Good relationships were found between the LSCT and cycling performance measured by the PPO and 40-km TT tests. Mean power had stronger relationships with measures of cycling performance during the second (r=0.80−0.89) and third stages (r=0.91−0.94) of the LSCT than HRR (r=0.55−0.68).

Conclusions The LSCT is a reliable novel test which is able to predict peak and endurance cycling performance from submaximal power, RPE and HRR in well-trained cyclists. As these parameters are able to detect meaningful changes more accurately than VO2max, the LSCT has the potential to monitor cycling performance with more precision than other current existing submaximal cycle protocols.

https://doi.org/10.1136/bjsm.2009.061325

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Monitoring changes in the performance of well-trained cyclists is necessary to determine the optimal balance between the training load and recovery. Two tests widely used to monitor and predict cycling performance are the peak power output (PPO) test and the 40-km time trial (40-km TT) test.1,–,3 Both these tests are reliable and able to detect small meaningful differences (<1%) in the day-to-day performance of well-trained cyclists, when the tests are conducted under well-controlled circumstances.4 However, the maximal and high-intensity effort associated with both tests may interfere with the prescribed training program and preparation for races of well-trained and elite cyclists. Therefore, it is impractical to conduct these tests on a weekly or even monthly basis as would be required for ongoing monitoring. As a result, these tests are often only conducted two or three times each year during the rest period, the precompetition period and sometimes once during the competition period.

    In an attempt to monitor athletes and predict performance more regularly, submaximal tests have been used since the 1950s. The Cooper test is an example of one of the first submaximal tests to be widely used to predict performance of runners. In this test, subjects are asked to cover as much distance as possible in 12 min. The total distance is used to predict maximal oxygen uptake capacity (VO2max) and endurance exercise capacity.5 A variety of protocols (walking and running) have evolved from this test. Examples of these adapted submaximal tests are the 6-min walk test,6 7 the shuttle walk test8 and the Rockport Fitness Walk test.9 More recently, a submaximal running test (Heart rate Interval Monitoring System) has been developed to monitor changes in training status through the measurement of heart rate recovery (HRR).10 11

    The development of submaximal cycle protocols also began in the 1950s with the Åstrand test12 and later the Physical Work Capacity 170 test.13 Although sex- and age-specific equations were developed to improve the accuracy of the VO2max predicted in the general population from these tests,14 15 the predicted VO2max in well-trained cyclists still seems to be underestimated by approximately 8.1 ml/kg/min.16 However, this may not be relevant for monitoring changes in performance associated with training, as VO2max has a limited accuracy to detect meaningful changes in well-trained cyclists.4 17

    The development of devices such as cycle ergometers and heart rate monitors has provided the opportunity to measure a wider range of variables which may be associated with performance including—for example, power, cadence, cycling efficiency, HRR and heart rate variability.4 18 This offers the potential of developing a new type of submaximal cycle test which uses variables other than VO2max as a predictor of cycling performance. If these variables which have been shown to have lower typical errors of measurements (TEMs)4 are indeed able to predict cycling performance, such a test has the potential to monitor changes in cycling performance more accurately.

    Therefore, the aim of this study was to determine the reliability and predictive value of parameters measured during a novel submaximal cycle protocol,4 on peak power and endurance cycling performance in well-trained cyclists. Based on the linear relationship between heart rate, exercise intensity and oxygen consumption,19 we hypothesised that other variables such as power are able to predict cycling performance accurately.

    Methods

    Recruitment

    Seventeen well-trained competitive male road racing cyclists, who were all engaged in preseason base training and between the ages of 18 and 40 years, were recruited for the study. All subjects completed a Physical Activity Readiness Questionnaire15 and were interviewed about their training history, after which an informed consent was signed. Inclusion criteria for the study were a minimum of six training hours per week over the 6-week period before the trial and a minimum competitive cycling background of at least 3 years. The study was approved by the Ethics and Research Committee of the Faculty of Health Sciences of the University of Cape Town. The data presented in this paper were part of a larger study of which a part has already been published.4 However, whereas the previous publication focused on measurement error around 40-km TTs and PPO tests, this study focuses on the reliability and predictive power of the Lamberts and Lambert Submaximal Cycle Test (LSCT).

    Study design

    As previously described,4 each subject completed four PPO tests (including respiratory gas analysis for oxygen consumption) and four 40-km TT tests over a period of five consecutive weeks (fig 1). Before each test, subjects were questioned about “sport injuries”, “sleeping patterns”, “use of medication”, “caffeine use”, “general fatigue” and “muscle soreness”.4 In addition, the LSCT (see fig 2) was performed each test. The initial 40km TT test, PPO test and 2 LSCTs were used to familiarise the subjects and to establish the correct intensity of the warm-up protocol (see LSCT) and were not included in the data analysis. Subjects were allowed to postpone both tests once during the 5-week period if they were able to provide a valid reason. Valid reasons for postponement included not feeling physically well (eg, sore throat, coughing), not being able to start the testing at the required time due to unforeseen events (eg, traffic holdups, power cuts) or inadequate sleep prior to the test, as described in Lamberts et al.4 Before each test, all subjects were questioned about possible confounding performance factors such as, for example, the usage of caffeine. During the testing period, the cyclists were asked to maintain their training load constant, so that their training status was constant throughout the study.

    Figure 1
    Figure 1

    A schematic presentation of a testing week including the recovery times between the peak power output (PPO) and 40-km time trial (TT) tests.

    Figure 2
    Figure 2

    Representation of the data of an arbitrary cyclists' heart rate response and power profile during the submaximal cycle test (LSCT). *5-min period over which performance parameters were analysed, #2-min period over which performance parameters were analysed.

    Preliminary testing

    Anthropometric data including height, weight and seven skinfolds (triceps, biceps, supra-iliac, subscapular, calf, thigh and abdomen)20 were obtained at the start of the study. Body fat was determined as the sum of seven skinfolds and also as a percentage of body mass.21 Before each testing and training session, cyclists were weighed to check for any significant changes in body mass which might reflect changes in their hydration status.10 22

    Calibration protocol

    The rear wheel tire of the subject's own bicycle was inflated to 800 kPa, and the bicycle was mounted, by a rear axle quick-release mechanism, to a cycle ergometer (Computrainer Pro 3D, RacerMate, Seattle, Washington, USA). Before the start of a standardised warm-up protocol, the contact pressure of the load generator against the rear wheel was calibrated to 0.88–0.93 kg. Six minutes into the warm-up protocol, by which time the tire had warmed-up, the load generator was re-calibrated to 0.88–0.93 kg as recommended by Davison et al.23

    LSCT

    The LSCT, which has been developed to monitor changes in training status4 and detect the symptoms of overreaching, was used as the warm-up protocol before all performance tests. The total duration of the LSCT is 17 min, during which time subjects are asked to cycle at intensities which elicit target heart rates of 60% (stage 1), 80% (stage 2) and 90% (stage 3) of their maximum heart rate (HRmax) (fig 2). Target heart rates for each of the different stages of the LSCT were calculated from the first PPO test in which HRmax was determined.4 Following calibration, a flat-course profile mimicking normal road circumstances was loaded into the Computrainer system. Thereafter, subjects were asked to change their front derailleur to the “small” chain ring, after which the LSCT was started. During the LSCT, subjects were allowed to alter their cadence and/or change gears on their rear derailleur within each stage in an attempt to elicit the correct target heart rate (within 1 beat/min).In stage 1 of the LSCT (0:00–6:00 min:second), subjects were asked to elicit a target heart rate equal to 60% of their personal HRmax. Following this stage, subjects were asked to change their front derailleur to the “large” chain ring, while the Computrainer system was recalibrated for accuracy purposes (6:00–6:30 min:second). During stage 2 of the LSCT (6:30–12:30 min:second), subjects were asked to elicit a target heart rate equal to 80% of their HRmax. Directly after this stage (stage 3; 12:30–15:30 min:second), subjects increased the workload to elicit a heart rate of 90% of HRmax. Immediately thereafter, subjects were asked to stop cycling and sit straight up, so that HRR data could be captured over the final 90 seconds of the LSCT (15:30–17:00 min:second). Although HRR analysis was performed over a 60-second period, as described before by Lamberts et al,10 the heart rate data were measured over 90 seconds to ensure that there were no missing data. During the LSCT, power, speed and cadence were measured continuously, while a rating of perceived exertion (RPE)23 for each stage was recorded 30 seconds before the end of each stage. Target heart rates for each of the different stages of the LSCT were calculated from the first PPO test in which HRmax was determined.4

    PPO test

    The PPO tests, which included respiratory gas analysis (VO2max), were performed as previously described by Lamberts et al4 and started exactly 8 min after the LSCT. Maximal PPO was determined as the mean power output during the final minute of the PPO test, whereas VO2max (ml/kg/min) was determined as the highest recorded reading for 30 seconds during this test.4

    Forty-km time trial

    The 40-km TT tests were started 3 min after completing the LSCT on a simulated 40-km flat TT course during which any type of feedback, except for completed distance, was withheld from the subject. No verbal encouragement was given during the time trial, except for during the last kilometre when the distance was counted down in 100-m sections and during the last hundred metres in 10-m sections.4 Thirty minutes after finishing the 40-km TT, subjects gave an overall RPE for the 40-km TT (6–20 Borg-scale).23

    Data collection and analysis

    Power output, speed, cadence and elapsed time were measured during all trials and stored by the Computrainer software at a rate of 34 Hz. Heart rate during these tests were measured with Suunto T6 heart rate monitors (Suunto Oy, Vantaa, Finland), and data were stored every 2 seconds. Oxygen uptake (VO2) and CO2 production (VCO2) were measured with an on-line breath-by-breath gas analyser and pneumotach (Oxycon, Viasys, Hoechberg, Germany) and stored as average values over eight breaths.

    Due to the slow half-life of changes in heart rate24 and subjects having to make fine adjustments to their workload to reach a specific heart rate during the warm-up, the heart rate and performance data in the first minute of each stage of the LSCT were excluded from analysis. Therefore, average performance and heart rate values were calculated over a 5-min period (from 1.00 to 6.00 and from 7:30 to 12:30 (min:second)) for stages 1 and 2, respectively, and for a 2-min period (13:30–15.30 min:second) for stage 3. Performance and heart rate data from the PPO test and the 40-km TT test were analysed for the full period of the data capture.

    Analysis of performance data was performed using CyclingPeaks analysis software (WKO+ edition, Version 2.1, 2006; Lafayette, Colorado, USA) and the Computrainer coaching software (Version 1.5.308; RacerMate, Seattle, Washington, USA). Heart rate data were analysed with Suunto Training Manager (Version 2.1.0.3; Suunto Oy, Vantaa, Finland).

    Statistical analysis

    The data were analysed with STATISTICA version 8.0 (Stat-soft Inc, Tulsa, Oklahoma, USA) for any statistical significance (p<0.05). All data are expressed as mean (SD).The mean physiological responses of all six submaximal cycle tests, conducted before each PPO and 40-km TTs, were compared by using a repeated measures analysis of variance. Reliability for each variable was also assessed by calculating interclass correlation coefficient (ICC) and the 95% CI.25 TEM and TEM as a percentage (%), expressed as a % of the mean score, were calculated with 90% CI, using a spreadsheet downloaded from http://www.newstats.org.

    Results

    The data set of two cyclists were excluded from analysis because they either did not complete the trial or unintentionally began using a medically prescribed drug which potentially enhances performance.26 The subject characteristics of the remaining 15 cyclists are shown in table 1, while subjects trained 10 (3) h/week with a competitive cycling history of 7 (3) years.4

    View this table:
    Table 1

    General characteristics of 15 well-trained cyclists

    Nine cyclists completed the testing within 4 weeks, while six riders completed the testing within 5 weeks. A questionnaire which was completed before each test revealed that none had consumed caffeine within the last 3 h before the test, and all subjects had maintained a similar dietary pattern. Furthermore, none reported any symptoms of general fatigue or muscle soreness, slept well the night before each test, were wearing the same cycling outfits during all tests and made no changes to their bicycle setups.

    All testing was performed on the same day of the week and at about the same time (within 1 h) with stable climatic conditions (21.7 (0.7)°C, 52 (4)% relative humidity, 101.8 (0.7) kPa). The body mass of the cyclists remained stable within 1.0 (0.4)% (0.8 (0.3) kg) throughout the study.

    Lamberts and Lambert Submaximal Cycle Test (LSCT)

    All subjects were able to closely regulate their heart rate by adjusting their exercise intensity during the three different stages of the LSCT. Mean fluctuations in heart rates were 0 (1) beats (range, −3 to 2 beats) at 60% of HRmax, 0 (1) beats (range, −1 to 2 beats) at 80% of HRmax and 0 (1) beats (range, −1 to 1 beat) while cycling at 90% of HRmax. Workload intensities corresponded to 31 (5)%, 60 (4)% and 80 (3)% of PPO, respectively, while RPEs increased progressively with each workload (8 (2), 11 (1) and 16 (1), respectively) as expected.

    Mean physiological responses of the cyclists during the LSCT are shown in table 2. In addition, intraclass correlation coefficients (ICC), TEM and TEM as coefficient of variation (%) for the physiological responses are shown in table 2. The highest ICC and lowest TEM were found during stage 3 (90% of HRmax) of the LSCT. The highest overall ICC (R=1.00) occurred for mean power measured during the last stage at 90% of HRmax with a TEM of 4.6 W (1.5%). The TEM of speed and cadence during this stage was 0.5 km/h (1.5%) and 3.5 rpm (3.7%), respectively. Absolute and relative HRR after the LSCT and their respective ICC and TEMs are also shown in table 2.

    View this table:
    Table 2

    Mean physiological changes within the LSCTs (n=6), which were performed before each PPO (n=3) and 40-km TT tests (n=3)

    PPO test and the 40-km TT

    The mean physiological responses of the cyclists during the three PPO tests and 40-km TTs are presented in table 3. High ICCs and low TEMs were found during the PPO test for PPO (R=1.00, TEM=0.9%), relative PPO (R=1.00, TEM=1.1%) and maximum oxygen uptake (VO2max) (R=0.99; TEM=2.2%). During the 40-km TT, there were high ICC and low TEMs for 40-km TT time (R=0.99; TEM=0.7%), average power (R=1.00; TEM=1.7%) and speed (R=1.00, TEM=0.6%). The lowest ICCs were found for cadence (R=0.93, TEM=3.5%) and RPE (R=0.90, TEM=2.8%). The average heart rate during the 40-km TT was 92% of HRmax.

    View this table:
    Table 3

    Mean time and physiological changes during the three peak power output tests and 40-km TTs

    Relationships between LSCT and performance parameters

    Relationships between mean power during the second (80% of HRmax) and third stage (90% of HRmax) of the LSCT and the performance parameters during the PPO test (VO2max and PPO) and the 40-km TT (40-km TT time and average 40-km TT power) are shown in fig 3. Although good relationships occurred between mean PPO during the second stage (80% of HRmax) and the performance parameters, stronger relationships were found between these performance parameters and mean power output during the third stage (90% of HRmax).

    Figure 3
    Figure 3

    The predictive value of the mean power output when cycling at 80% and 90% maximum heart rate (HRmax) during the LSCT on performance parameters maximal oxygen uptake capacity (VO2max), peak power output (PPO), average 40-km time trial power and 40-km TT time.

    The relationships between HRR and the performance parameters are shown in fig 4. Weak relationships occurred between HRR and peak power and endurance performance. Although the relationship between performance parameters and HRR was significant, relationships between the performance parameters and the mean power during the second (80% of HRmax) and third stages (90% of HRmax) of the LSCT were substantially better than the HRR relationships.

    Figure 4
    Figure 4

    Correlation between heart rate recovery (HRR) measured as part of the LSCT and absolute and relative performance parameters maximal oxygen uptake capacity (VO2max), peak power output (PPO), average 40-km time trial power and 40-km TT time.

    Discussion

    The first relevant finding of this study was that the LSCT test was truly submaximal, as confirmed by the workloads (31%, 60% and 80% of PPO, respectively) and RPEs (8, 12 and 16, respectively). In addition, the low TEM in the RPEs during the first stage of the LSCT, low TEM in 40-km TT and PPO results4 and no reports of fatigue and/or muscle soreness before the LSCT (which was performed twice per week) indicate that the LSCT does not interfere with “normal” training habits. This was also confirmed by the anecdotal comments we received from the subjects. These factors suggest that the test has positive practical attributes for monitoring cyclists on a regular basis. In contrast, a maximal or near maximal endurance test, which by definition has an exhausting character, would interfere with the cyclists' training and racing programme and, therefore, cannot be administered as frequently as a submaximal test. All cyclists were able to adhere to the protocol for the LSCT and maintained their heart rates within a few beats of their target heart rate for each workload. As the exercise intensity increased, the variation in heart rate decreased, with the variation being reduced to within 1 beat at the workload corresponding to 90% of HRmax. This is in accordance with earlier published research work in our laboratory on physically active subjects undergoing an incremental submaximal running test.10 27

    The second finding of this study was that all performance parameters measured during the second and third stages of the LSCT were highly repeatable (ICC range: 0.85–1.00) with relatively low TEM (TEM range: 1.3–4.4%).18 The overall highest repeatability with the lowest TEM occurred during the third stage of the LSCT (90% of HRmax), where mean power output and mean speed had the lowest TEMs (1.5% and 1.3%, respectively). In addition, the measurement of HRR was also reliable. HRR, which is a marker of autonomic function,28,–,30 has recently been associated with a change in prescribed training load11 and training status.31 HRR, with an unchanged training status over the duration of the testing period, varied by only 5 beats/min (see also table 3). This is similar to what has been reported before10 27 32 and smaller than what was measured as a result of an improved training status,18 33 34 indicating that there is sufficient precision in the measurement of HRR to detect small meaningful changes.

    The third finding of this study describes the relationship between mean power and HRR measured during the LSCT, and peak power (PPO) and endurance (40-km TT) cycle performance. Although good relationships were found between the parameters measured during the second stage of the LSCT (80% HRmax) and PPO and 40-km TT performance (r=0.80–0.89), an even better relationship was found with mean power during the third stage of the LSCT (90% of HRmax) (r=0.91–0.94).

    Based on the understanding that the heart rate decreases during exercise at a similar workload as training status improves,35 it is a given that during exercise at the same heart rate, workload will increase. This is supported by Lucía et al, who showed that submaximal power, at ventilatory thresholds 1 and 2 and lactate threshold increased as training status improved, while the heart rates at these stages remained unchanged.1 While no single parameter has been identified as a reliable marker of training-induced fatigue or symptoms of overreaching, it is generally accepted that non-functional overreaching is associated with a decrement in performance.36 In addition, most overtraining studies report a decrease in submaximal and maximal heart rates with the manifestation of fatigue,37,–,40 which would possibly lead to an increase in power during stages 2 and 3 of the LSCT because the predetermined submaximal heart rates during each of these workloads would be harder to elicit. Based on the good correlations between average power during stages 2 and 3 of the LSCT and performance parameters, a change in mean power could possibly reflect changes in training status in a practical, meaningful way.

    In contrast to power, HRR showed a weaker relationship to PPO and 40-km TT performance. This can possibly be explained by the homogeneity of our cyclists who were all well trained and in which genetic polymorphisms in the acetylcholine receptor M2 gene (CHRM2) can explain interindividual differences in HRR.41 Although there was a relatively weak relationship between HRR and cycling performance, changes in HRR have recently been shown to track well with changes in performance parameters4 and seem to be associated with a blunted improvement in endurance performance parameters.31 Therefore, a combination of a change in HRR with a change in power and RPEs during the LSCT might provide more useful information for monitoring fatigue and predicting performance than other current submaximal cycle protocols which rely on changes in predicted VO2max.

    Summary and conclusion

    In summary, the goal of this study was to identify markers arising from submaximal exercise which would accurately reflect cycling performance. Such markers would provide a practical method for coaches and scientists to accurately monitor cycling performance. The LSCT, in which heart rate is fixed at a predetermined submaximal level, has the potential to detect subtle changes in performance as a result of training-induced fatigue. In addition, monitoring the cumulative fatigue could contribute to detecting the status of functional or nonfunctional overreaching. Collectively, these factors will be useful in maintaining the balance between training load and recovery with the purpose of achieving the most optimal training status. In particular, the LSCT could indicate the development of fatigue by a combination of three parameters: (1) increased RPE levels, as subjects will have to work harder to elicit their heart rates to 90% of HRmax; (2) a sudden strong increase in mean power, as a higher workload is needed to elicit the target heart rates; and (3) a change in HRR based on deregulation of the autonomic nervous system.11 39 The precision of these markers in detecting and tracking fatigue will have to be determined in future research.

    What is already known on this topic

    • ▶ Submaximal cycle tests such as the Åstrand test and Physical Work Capacity 170 are reliable tests which are able to predict of VO2max in a general population.

    • ▶ However, changes in VO2max do not track with changes in cycling performance in well-trained cyclists. Therefore, these tests have a limited application and are not applicable for measuring small changes in cycling performance.

    What this study adds

    • ▶ This study presents a novel submaximal cycle protocol (LSCT) which is reliable, has a low typical error of measurement, is able to predict cycling performance with reasonable accuracy and possibly can monitor the accumulation of fatigue.

    • ▶ Based on these findings, the LSCT has the potential to monitor cycling performance and training-induced fatigue with sufficient precision to be able to detect small but meaningful changes.

    Acknowledgments

    The authors would like to thank all cyclists who participated this study. This study was funded by the Medical Research Council of South Africa, Discovery Health and the University of Cape Town/MRC Research Unit for Exercise Science and Sports Medicine.

    References

    1. ↵
      1. Lucía A,
      2. Hoyos J,
      3. Pérez M,
      4. et al
      . Heart rate and performance parameters in elite cyclists: a longitudinal study. Med Sci Sports Exerc 2000;32:1777–82.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Jeukendrup A
      . High-performance cycling. Champaign, Illinois, USA: Human Kinetics Publishers, Inc, 2002.
    3. ↵
      1. Mujika I,
      2. Padilla S
      . Physiological and performance characteristics of male professional road cyclists. Sports Med 2001;31:479–87.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Lamberts RP,
      2. Swart J,
      3. Woolrich RW,
      4. et al
      . Measurement error associated with performance testing in well-trained cyclists; application to the precision of monitoring changes in training status. Int SportMed J 2009;10:33–44.
      OpenUrlWeb of Science
    5. ↵
      1. Cooper KH
      . A means of assessing maximal oxygen intake. Correlation between field and treadmill testing. JAMA 1968;203:201–4.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Butland RJ,
      2. Pang J,
      3. Gross ER,
      4. et al
      . Two-, six-, and 12-minute walking tests in respiratory disease. Br Med J (Clin Res Ed) 1982;284:1607–8.
      OpenUrlFREE Full Text
    7. ↵
      ATS statement: guidelines for the six-minute walk test. Am J Respir Crit Care Med 2002;166:111–17.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Singh SJ,
      2. Morgan MD,
      3. Scott S,
      4. et al
      . Development of a shuttle walking test of disability in patients with chronic airways obstruction. Thorax 1992;47:1019–24.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Kline GM,
      2. Porcari JP,
      3. Hintermeister R,
      4. et al
      . Estimation of VO2max from a one-mile track walk, gender, age, and body weight. Med Sci Sports Exerc 1987;19:253–9.
      OpenUrlPubMedWeb of Science
    10. ↵
      1. Lamberts RP,
      2. Lemmink KA,
      3. Durandt JJ,
      4. et al
      . Variation in heart rate during submaximal exercise: implications for monitoring training. J Strength Cond Res 2004;18:641–5.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Borresen J,
      2. Lambert MI
      . Changes in heart rate recovery in response to acute changes in training load. Eur J Appl Physiol 2007;101:503–11.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Åstrand PO,
      2. Ryhming I
      . A nomogram for calculation of aerobic capacity (physical fitness) from pulse rate during sub-maximal work. J Appl Physiol 1954;7:218–21.
      OpenUrlFREE Full Text
    13. ↵
      1. Haber P,
      2. Schlick W,
      3. Schmid P,
      4. et al
      . [Estimation of the performance spectrum of healthy adolescents by using the PWC 170 (Physical Work Capacity)]. Acta Med Austriaca 1976;3:164–6.
      OpenUrlPubMed
    14. ↵
      1. Hartung GH,
      2. Blancq RJ,
      3. Lally DA,
      4. et al
      . Estimation of aerobic capacity from submaximal cycle ergometry in women. Med Sci Sports Exerc 1995;27:452–7.
      OpenUrlPubMedWeb of Science
    15. ↵
      1. Whaley MH,
      2. Brubaker PH,
      3. Otto RM
      American College of Sports Medicine. Submaximal exercise tests. In: Whaley MH, Brubaker PH, Otto RM, eds. ACSM's guidelines for exercise testing and prescription. Baltimore, Maryland, USA: Lippincott Williams & Wilkins, 2007:70–4.
    16. ↵
      1. Hartung GH,
      2. Krock LP,
      3. Crandall CG,
      4. et al
      . Prediction of maximal oxygen uptake from submaximal exercise testing in aerobically fit and nonfit men. Aviat Space Environ Med 1993;64:735–40.
      OpenUrlPubMed
    17. ↵
      1. Noakes TD
      . Testing for maximum oxygen consumption has produced a brainless model of human exercise performance. Br J Sports Med 2008;42:551–5.
      OpenUrlFREE Full Text
    18. ↵
      1. Lamberts RP,
      2. Swart J,
      3. Noakes TD,
      4. et al
      . Changes in heart rate recovery after high-intensity training in well-trained cyclists. Eur J Appl Physiol 2009;105:705–13.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Arts FJ,
      2. Kuipers H
      . The relation between power output, oxygen uptake and heart rate in male athletes. Int J Sports Med 1994;15:228–31.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. MacDougall JD,
      2. Wenger HA,
      3. Green HS
      1. Ross WD,
      2. Marfell-Jones MJ
      . Kinanthropometry. In: MacDougall JD, Wenger HA, Green HS, eds. Physiological testing of the high performance athlete. Champaign, Illinois, USA: Human Kinetics, 1991:223–308.
    21. ↵
      1. Durnin JV,
      2. Womersley J
      . Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br J Nutr 1974;32:77–97.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Montain SJ,
      2. Coyle EF
      . Influence of graded dehydration on hyperthermia and cardiovascular drift during exercise. J Appl Physiol 1992;73:1340–50.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Davison RC,
      2. Corbett J,
      3. Ansley L
      . Influence of temperature and protocol on the calibration of the computrainer electromagnetically braked cycling ergometer. J Sports Sci 2007;25:257–8.
      OpenUrl
    24. ↵
      1. Achten J,
      2. Jeukendrup AE
      . Heart rate monitoring: applications and limitations. Sports Med 2003;33:517–38.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Vincent WJ
      . Statistics in kinesiology. Champaign, Illinois, USA: Human Kinetics Publishers, Inc, 1995.
    26. ↵
      1. Swart J,
      2. Lamberts RP,
      3. Lambert MI,
      4. et al
      . Exercising with reserve: evidence that the CNS regulates prolonged exercise performance. Br J Sports Med 2009;43:782–88.
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Lamberts RP,
      2. Lambert MI
      . Day-to-day variation in heart rate at different levels of submaximal exertion: implications for monitoring training. J Strength Cond Res 2009;23:1005–10.
      OpenUrlPubMedWeb of Science
    28. ↵
      1. Buchheit M,
      2. Papelier Y,
      3. Laursen PB,
      4. et al
      . Noninvasive assessment of cardiac parasympathetic function: postexercise heart rate recovery or heart rate variability? Am J Physiol Heart Circ Physiol 2007;293:H8–10.
      OpenUrlFREE Full Text
    29. ↵
      1. Buchheit M,
      2. Gindre C
      . Cardiac parasympathetic regulation: respective associations with cardiorespiratory fitness and training load. Am J Physiol Heart Circ Physiol 2006;291:H451–8.
      OpenUrlAbstract/FREE Full Text
    30. ↵
      1. Borresen J,
      2. Lambert MI
      . Autonomic control of heart rate during and after exercise: measurements and implications for monitoring training status. Sports Med 2008;38:633–46.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Lamberts RP,
      2. Swart J,
      3. Capostagno B,
      4. et al
      . Heart rate recovery as a guide to monitor fatigue and predict changes in performance parameters. Scand J Med Sci Sports 2010;20:449–57.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Brisswalter J,
      2. Legros P
      . Daily stability in energy cost of running, respiratory parameters and stride rate among well-trained middle distance runners. Int J Sports Med 1994;15:238–41.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Buchheit M,
      2. Millet GP,
      3. Parisy A,
      4. et al
      . Supramaximal training and postexercise parasympathetic reactivation in adolescents. Med Sci Sports Exerc 2008;40:362–71.
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Sugawara J,
      2. Murakami H,
      3. Maeda S,
      4. et al
      . Change in post-exercise vagal reactivation with exercise training and detraining in young men. Eur J Appl Physiol 2001;85:259–63.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Wilmore JH,
      2. Stanforth PR,
      3. Gagnon J,
      4. et al
      . Endurance exercise training has a minimal effect on resting heart rate: the HERITAGE Study. Med Sci Sports Exerc 1996;28:829–35.
      OpenUrlPubMedWeb of Science
    36. ↵
      1. Meeusen R,
      2. Duclos M,
      3. Gleeson M,
      4. et al
      . Prevention, diagnosis and treatment of the overtraining syndrome. Eur J of Sport Sci 2006;6:1–14.
      OpenUrlCrossRef
    37. ↵
      1. Halson SL,
      2. Bridge MW,
      3. Meeusen R,
      4. et al
      . Time course of performance changes and fatigue markers during intensified training in trained cyclists. J Appl Physiol 2002;93:947–56.
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. Jeukendrup AE,
      2. Hesselink MK,
      3. Snyder AC,
      4. et al
      . Physiological changes in male competitive cyclists after two weeks of intensified training. Int J Sports Med 1992;13:534–41.
      OpenUrlPubMedWeb of Science
    39. ↵
      1. Lehmann M,
      2. Dickhuth HH,
      3. Gendrisch G,
      4. et al
      . Training-overtraining. A prospective, experimental study with experienced middle- and long-distance runners. Int J Sports Med 1991;12:444–52.
      OpenUrlPubMedWeb of Science
    40. ↵
      1. Urhausen A,
      2. Gabriel HH,
      3. Weiler B,
      4. et al
      . Ergometric and psychological findings during overtraining: a long-term follow-up study in endurance athletes. Int J Sports Med 1998;19:114–20.
      OpenUrlPubMedWeb of Science
    41. ↵
      1. Hautala AJ,
      2. Rankinen T,
      3. Kiviniemi AM,
      4. et al
      . Heart rate recovery after maximal exercise is associated with acetylcholine receptor M2 (CHRM2) gene polymorphism. Am J Physiol Heart Circ Physiol 2006;291:H459–66.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Competing interests None.

    • Ethics approval This study was conducted with the approval of the Ethics committee, University of Cape Town, ref no: 265/2006.

    • Provenance and peer review Not commissioned; externally peer reviewed.