Journal Logo


Critical Velocity during Intermittent Running with Changes of Direction


Author Information
Medicine & Science in Sports & Exercise: February 2019 - Volume 51 - Issue 2 - p 308-314
doi: 10.1249/MSS.0000000000001774
  • Free


Endurance capacity is a central determinant of endurance sport performance regardless of whether the activity is intermittent, steady state, an individual activity, or team-based sport. The importance of endurance capacity is especially relevant when the general pace of play or intensity of exercise must be sustained during a game or competitive event. If pace cannot be maintained, the athlete may be beaten by an opponent, fall off race pace, or be removed from the game and replaced by another player (1–4). Therefore, assessments of endurance capacity (i.e., sustained work rate) that are specific to the demands of the athlete’s sport, are of interest to both athletes and coaches.

Perhaps the most common method of determining endurance capacity is a continuous exercise, graded-intensity, treadmill, or cycling ergometer test. Graded exercise testing is well recognized to describe cardiorespiratory fitness and maximal aerobic capacity (V˙O2max) (5,6), with further analysis describing variables, such as gas exchange threshold or “percent delta” calculations (between gas exchange threshold and V˙O2max) (7,8). Despite the obvious value of graded exercise treadmill testing, the protocol is usually constrained to a laboratory setting, typically requires the use of expensive and bulky equipment, is limited to a single person assessment, and often measures only steady state unidirectional movement. Combined, this poses a challenge for practical field testing. Recognizing this limitation, Bangsbo and colleagues (6,9) developed a simple field test (i.e., Yo-Yo Intermittent Recovery Test) to provide an indicator of maximal fitness performance. The intermittent nature of the Yo-Yo test offers direct relevance to sports involving regular intervals of intense intermittent (Int) and change of direction running (CoD). However, both the traditional graded exercise test and the current field tests typically target quantification of maximal exercise capacity and the tolerance to increasing workload (i.e., V˙O2max value, total work performed, or distance covered). As a result, it is difficult to identify the critical work rate that can be sustained over a long period.

For a given individual, the highest sustainable rate of oxidative phosphorylation (critical V˙O2) is closely associated with the highest sustainable work rate (critical power [CP] or critical velocity [CV] as a surrogate in running). Identifying the highest sustainable work rate provides a “threshold” around which athletes and coaches can generate a pacing strategy to fit the event, game, or race needs (10–12). Critical power has been accurately identified when exercise performed just above CP leads to progressive perturbation of metabolism (i.e., elevated blood lactate and/or V˙O2 slow component) and inevitable fatigue, yet exercise performed just below CP results in a stable metabolic profile (13). With regard to Int/CoD activity, shuttle running research suggests Int/CoD tasks of the same average velocity as linear running require greater oxygen demand, primarily attributed to the increase in positive work required to accelerate the center of mass (14,15). Along these lines, an exhaustive “all-out” test protocol recently applied to shuttle running in an effort to describe CV, did not confirm the measured CV met the requisite physiological profile (16).

With this information as background, the goal of this study was to develop a simple field test of short duration that accurately and reliably identifies CV specific to sporting activities requiring Int/CoD. We tested the hypothesis that a custom field test can reliably and accurately identify CV for Int/CoD running. We further tested the hypothesis that CV is specific to the task or type of activity. That is, CV for linear running (17) is not equivalent to the CV associated with a Int/CoD running task.



With institutional review board (St. Charles Medical Center, Bend OR) approval and after written informed consent, a total of 12 young healthy recreationally active adults (10 male, two females, age: 27.5 ± 2.4 yr old; weight: 78.9 ± 10.5 kg; height: 180.9 ± 6.9 cm; body mass index: 24.5 ± 2.4 kg·m−2; means ± SD) volunteered to participate in the present study. Some subjects participated in more than one protocol (number of participants in each protocol are defined below). All subjects were nonsmokers, nonobese, normotensive (resting blood pressure <140/90 mm Hg), and not taking any medications. All exercise sessions were performed with a minimum of 24 h after any previous exercise and subjects were instructed to maintain consistent activity levels, as well as, diet and caffeine habits between and before all data collection sessions. Subjects were instructed to abstain from caffeine and food for at least 2 h before exercise.

CV shuttle test

Figure 1 is a graphical representation depicting the CV shuttle test (CVST). In general, the CVST consisted of exercise intervals with maximal effort out-and-back sprinting for 18.28 m (i.e., 20 yards per direction equating to a total of 36.57 m each interval) for a total of 7 min. The 7 min were segmented into three rounds, each of which contained different rest intervals. Round one was 3 min in duration and permitted 15 s of rest between each down and back interval accomplished (i.e., 15 s per split). Round two was 2 min in duration with rest periods of 10 s per split. Round three was 2 min in duration with no rest permitted between splits (i.e., continuous shuttle effort). Each round started immediately after the conclusion of the previous round.

CVST. Subjects were instructed to run at maximal effort for 18.3 m (~60 ft), change direction and return to the starting line. The test was 7 min in total duration, with three rounds of work, and an imposed rest interval immediately after each return to starting line. Round 1 was 3 min in duration with 15-s rest intervals, round 2 was 2 min in duration with 10-s rest intervals, and round 3 was 2 min in duration with no prescribed rest interval.

Before each CVST, the subject was allowed a duration of 5 to 10 min to complete a moderate warm-up routine of their choice, followed by 5 min of rest. Before the start of each CVST, subjects were instructed to give maximum effort for each sprint and were regularly provided verbal encouragement to deliver maximal effort throughout the CVST. Split times were monitored using a stopwatch with lap function, and cones were used to define the location of test distances. The test was stopped when the subject returned to the starting line at 7 min, or after the subject completed the split in which 7 min had occurred. End velocity (EV) was defined as the average velocity for the final four splits completed in the last round of the CVST. Total distance covered was quantified after test completion by calculating the total number of splits completed multiplied by the fixed course distance.

Protocol 1: critical velocity shuttle test—physiological profile and reliability

To examine the basic physiological response from the CVST and to determine reliability of the EV obtained from the CVST, six subjects performed two CVST within a 2-wk time frame, but no sooner than 48 h between visits. To determine the physiological profile and ensure maximal effort during the CVST; heart rate was recorded for the entirety of each CVST using a Wahoo Tickr X heart rate strap (Dynastream Innovations, Inc. Alberta Canada) and capillary blood lactate was measured by finger-prick (Lactate Scout +; EKF Diagnostics, Cardiff, UK) within 1 min before the test started and immediately after the test was terminated. Testing location and course remained the same during all visits.

Protocol 2: CVST validation

To determine whether EV obtained from the CVST is a valid determinant of CV for Int/CoD running, eight subjects completed three visits. Visit 1 required completion of the CVST (see CVST methodology detailed above and Figure 1) where the test EV was established. During visits 2 and 3 (performed in counterbalanced order), a work rate (i.e., running velocity) corresponding to either 5% above or 5% below CVST EV was computed and prescribed to each subject. The validation course remained the same as that which was used to conduct the CVST. Specifically, subjects were instructed to perform the prescribed work rate to task failure (defined as voluntary cessation or inability to maintain prescribed speed), or for a total of 20 min at which time the validation session ceased. Subjects were paced using a custom LabVIEW software application (LabVIEW, National Instruments, Austin Texas) producing an audible “beep” alerting the subject to make a change of direction at either the start or end cone set, and/or to resume running after a period of rest. The frequency of “beeps” corresponded with the prescribed velocity subjects were required to run between cones. Subjects followed a standardized 3-min warm-up and familiarization period before each validation trial. The warm-up pace was kept at 5% below EV obtained from the CVST to ensure subject were not warming up in the severe intensity domain.

To confirm that EV from the CVST is indeed an accurate determination of the work rate boundary that discriminates an unsustainable metabolic profile from a sustainable metabolic profile, capillary blood lactate was measured every 4 min during the validation trials performed 5% above and below EV (visits 2 and 3). Capillary sample measurement time was recorded and averaged 24.9 ± 5.7 s. Heart rate was monitored continuously throughout each validation visit.

Protocol 3: critical velocity—intermittent/change of direction versus continuous linear running comparison

To determine the relationship between CV obtained from the CVST and CV from traditional linear running testing protocols (17), ten subjects completed the CVST or a running time trial of either 1.6-, 2.4-, or 5-km distances on a flat course with zero abrupt changes in direction on four different lab visits in randomized order. Time trials were spaced at least 3 d apart and completed within a month time span during which subjects were instructed to maintain consistent activity levels. End velocity from the CVST was used to define CV for Int/CoD running as described above and a speed duration curve was established from the time and distance data of the three linear running time trials. Critical velocity was identified as the slope of the three data points (17,18).


All values are reported as means ± S.D. unless otherwise noted and analyses were performed using SigmaPlot (Systat Software, San Jose, CA). Specific hypothesis testing between condition over time was performed using two way repeated-measures ANOVA (time × condition). ANOVA analysis was assessed for homoscedasticity with both Breusch–Pagan and White tests. Data comparison across conditions was made with paired t tests. To demonstrate the magnitude of difference between groups in standard deviation units, Cohen’s d was calculated. The percent coefficient of variation was calculated as standard deviation divided by the mean × 100. Linear regression analysis was performed to determine regression equation, coefficient of determination (R2) and standard error of the estimate (SEE). Intraclass correlation coefficients were calculated to assess reproducibility of measurements. Statistical significance was set a priori at P < 0.05.


Protocol 1: physiology and reliability of the CVST

Figure 2 depicts representative data from one subject demonstrating heart rate and average velocity for each interval within each round of the CVST. On average, blood lactate increased from pre-CVST to post-CVST during both visit 1 (2.4 ± 0.4 mM vs 15.7 ± 1.8 mM, P < 0.05) and visit 2 (2.0 ± 0.3 mM vs 15.0 ± 2.2 mM; P < 0.05), and values were similar for a given time point across visits. Maximum heart rate was similar between visits (177 ± 5 vs 179 ± 7) and reached 93% ± 4% of age-predicted maximal heart rate calculated for each subject. Figure 3 depicts the relationship in CVST EV (separated for each round) relationship between visit 1 and visit 2 (protocol 1: reliability trials). Neither the average total distance covered nor the CVST EV were significantly different between visits (visit 1 = 864 ± 21 m and 3.23 ± 0.13 m·s−1 vs visit 2 = 900 ± 30 m and 3.21 ± 0.15 m·s−1, respectively). Importantly, the SEE for the CVST EV was only 0.046 m·s−1, showing high resolution and very small error (Fig. 3). The percent (%) coefficient of variation in CVST EV was 4.16% for visit 1 and 4.71% for visit 2, with a between visit coefficient of variation of less than 1%.

CVST: Representative subjects velocity and heart rate profile. The running velocity and heart rate response during the CVST in a representative subject. Note the general decline in velocity over the test, yet an observed stable EV profile during the final series of repetitions. Heart rate rose over the test period to near maximal values.
Relation in Running Velocity between CVST visit 1 and CVST visit 2. The ICC and the SEE for running velocities were calculated between CVST visits 1 and 2. ICC, intraclass coefficient.

Protocol 2: validation of critical velocity from the CVST

All subjects successfully completed 20 min during the paced trial at 5% below CV (as determined by CVST) and capillary blood lactate reached stable, steady state values over the paced trial duration. In contrast, all subjects failed to complete 20 min during the paced trial at 5% above CV (average time to exhaustion = 10.6 ± 3.9 min), and capillary blood lactate increased continuously (never stabilizing) before task failure occurred (Fig. 4). The Cohen’s d for a given percent of trial completion was 0.36, 1.23, 1.37, 2.07, and 2.66 for 20%, 40%, 60%, 80%, and 100% of trial completion, respectively; demonstrating a clear difference between the two paced trials.

Capillary blood lactate for critical velocity validation trials. To examine whether CVST EV identifies critical velocity, blood lactate was measured during intermittent Int/CoD running at 5% above or 5% below CVST EV. *P < 0.05.

Protocol 3: comparison of CV from CVST to continuous linear running

Critical velocity was lower (P < 0.05) for the Int/CoD running task (3.40 ± 0.34 m·s−1) compared to linear running (4.13 ± 0.56 m·s−1), with a Cohen’s d of 1.58 (Fig. 5). Despite this difference in CV magnitude, a statistically significant correlation (R2 = 0.72, P < 0.05) and a small SEE (0.19 m·s−1) demonstrates a clear relationship between the CV obtained from CVST and the linear running CV (Fig. 5B).

Comparison of linear running CV to Int/CoD running CV. A, Critical velocity during Int/CoD running is significantly lower than critical velocity of linear running. B, Linear running CV is significantly related to Int/CoD running with an R 2 of 0.72.


The results of the present study describe a novel field test that reliably and accurately determines an individual’s CV during Int/CoD. These findings also demonstrate the importance of sport-specific or task-specific CV identification. Collectively, these findings provide a tool for an athlete, coach, or trainer who intends to monitor sustainable versus unsustainable exercise intensities and the corresponding physiological impact of exercise load.

Continuous short-duration “all-out” test protocols reliably and successfully identify the work rate associated with the boundary separating the sustainable heavy intensity domain from the unsustainable severe intensity domain (19–21). However, common field tests used to quantify endurance capacity for intermittent activities (6,9) do not easily identify sustainable work rates for athletes performing intermittent activity with change of direction movements (10,19,22–24). Therefore, we found it valuable to create and validate a novel field test that would identify this important physiological boundary that applies to tasks routinely executed in specific sports. Pilot experiments in our laboratory suggested that CV is task-specific, and that CV obtained from a traditional linear running test does not appropriately discriminate CV for Int/CoD running. This observation is consistent with prior evidence indicating that change of direction activity is metabolically expensive (14,15). Taken together, the study described here aimed to develop a field test specific for Int/CoD running capable of identifying CV.

Consistent with other maximal effort test profiles (19,21,25), the CVST elicited peak work rates within the first few seconds which was followed by a general decline in velocity across the test duration reaching a plateau in running velocity for the final portions of the test. Capillary blood lactate was significantly greater immediately after test completion and heart rate approached near age-predicted maximums (Fig. 2). Because CVST end test velocity is the sole predictor of CV, coupled with the fact that even small errors in CV can lead to a large influence on the outcome, we aimed to confirm that the CVST was repeatable. The average test–retest within-subject coefficient of variation for EV was less than 1%, the intraclass coefficient was 0.96, and the SEE was only 0.046 m·s−1. Together, this indicates end-test velocity from the CVST is highly reproducible. The similar total distance covered and similar end-test velocity between test 1 and test 2 suggests that subjects were, in fact, performing maximal effort trials over both testing days.

Critical velocity represents a work rate associated with the highest attainable steady state for aerobic energy production and is the demarcation between the sustainable heavy intensity domain and the unsustainable severe intensity domain. Any work rate below CV (heavy domain) is expected to result in a stable metabolic profile. Whereas any work rate above CV (severe domain) will lead to an inexorable rise in blood lactate and V˙O2 as a function of time until V˙O2max is reached, thereby resulting in either task failure or a required reduction in work rate to below CV (19,20,22,24). Observations from protocol 2 demonstrate that CVST end test velocity represents the task-specific CV. Specifically, all participants running at an average velocity above CVST EV displayed an unstable rise in capillary blood lactate and were not able to complete the trial. In contrast, all participants running at an average velocity below CVST EV completed the trial and exhibited a stable blood lactate profile (Fig. 3). Collectively, these results are consistent with the current definition and understanding of CV (13,19,24) and demonstrate a clear ability of the CVST to distinguish between important exercise intensity domains with a single, simple, field test.

The critical threshold phenomenon applies to many different whole body exercise modalities, (i.e., running, swimming, rowing and cycling) as well as isolated single muscle groups, but has predominately been studied performing constant work rate exercise (17,19,24,26–29). Although linear running is a main contributor for physiological demand during competition in sports like soccer (30), further demands are complicated by quick sprint bursts, bouts of rest, and changes of direction (1,14,15). Therefore, a need exists to determine any task-specific differences in CV due to demands incurred by Int/CoD. Our data demonstrate an important difference for CV between the two running tasks. Critical velocity for Int/CoD (from the CVST) was significantly lower than the CV observed for linear running (△ −17% ± 2%). This is not simply because the Int/CoD running task had a lower average velocity but is likely due to the increased physiological demand and energy cost for deceleration and acceleration of the center of mass during rapid change in direction (14,15). This highlights two key findings: 1) critical thresholds are task-specific and sensitive to changes of the modality used to elicit threshold and 2) critical thresholds can apply to a wide array of exercise modalities and future work should needs to identify thresholds that are specific to a given sport.


Despite effective identification of CV for Int/CoD running by the CVST, a few considerations should be acknowledged. Although the coefficient of variation calculations were low, protocol 1 (reliability protocol) included only six subjects. Presumably, increased subject number would increase confidence in test–retest reliability. The three running time trials to determine CV for the linear running task were performed over a month timeframe, which may lead to changes in fitness between trials. Subjects were instructed to have no substantial change in activity level during that period and, although we do not have data confirming there was no change in fitness, the high correlation coefficient for the three trials suggest no significant change in subject fitness.

The key feature of the CVST is simply that the effort must be maximal and be coupled to intermittent running that incorporates changes of direction. This means the exact durations used in the present study for total work (7 min) and rest intervals (15 and 10 s) need not be considered the only option. Our experience from pilot studies suggests subjects found very short, all-out efforts (~3 min) with no rest much less tolerable than efforts slightly longer in duration, inclusive of short rest. To this end, although the test must elicit maximum effort, the test itself must encourage athlete participation and not be so unbearable as to limit athlete motivation and therefore accuracy. In general, the CVST test conditions (work/rest ratio and distance) are modifiable to easily fit within a coaches personal program or preference. However, protocols must be kept consistent to be comparable across subjects and groups. Test administrators should recognize that shortened distances would increase the proportion of change in direction over a given period. The critical power model commonly provides a second variable in addition to the critical power, referred to as the W′ or the finite work capacity that can be performed above the critical power (19,22,24,31). The focus of the present study was to identify the critical work rate and was not aimed at quantifying the W′ during the CVST. The reasoning, in part, was due to the non-linear recharge of W′ when exercise is performed below the critical power (e.g., during the recovery interval), the athlete-specific nature of such reconstitution rates, and the reported variability in W′ measurement accuracy (2,10,31–34). Future studies targeting W′ balance profiles during intermittent sport would presumably strengthen specific time to fatigue predictions for severe exercise encroaching on the limit of tolerance. Nevertheless, the current study provides a simple, reproducible, and valid protocol for critical threshold boundary identification.


The field test described here provides coaches and athletes a simple method to obtain work rate thresholds specific to sport modality, thereby improving training intensity decisions, game strategy decisions, and resultant understanding of the impact of work accomplished during practice and competition. Because work rates performed in the severe exercise intensity domain lead to predictable fatigue, with work rates executed in the heavy domain considered sustainable for a substantial period, identification of critical work rate provides a valuable measure to assist athletes during practice and competition with regards to monitoring and regulating time to fatigue. Taken together, we present a novel and repeatable field test that accurately determines critical work rate during dynamic Int/CoD running activity. The CVST could be used by coaches and athletes to better understand the physiological impact and rate of fatigue during practice and competition.

The authors thank all subjects who participated in the study. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation, and the results of the present study do not constitute endorsement by ACSM.

No conflicts of interest are declared. This project was supported by internal Nike, Inc funding mechanisms.

B. S. K. and E. M. B. contributed to the experimental conception and design, collection, analysis and interpretation of the data, and writing of this article. B. W. W. contributed to the experimental conception and design, data interpretation, critical review and editing of this article. All authors have given final approval of this article. All experiments were performed at Nike World Headquarters, Beaverton, Oregon.


1. Gray AJ, Jenkins DG. Match analysis and the physiological demands of Australian football. Sports Med. 2010;40(4):347–60.
2. Skiba PF, Clarke D, Vanhatalo A, Jones AM. Validation of a novel intermittent W’ model for cycling using field data. Int J Sports Physiol Perform. 2014;9(6):900–4.
3. Abe D, Tokumaru H, Niihata S, et al. Assessment of short-distance breaststroke swimming performance with critical velocity. J Sports Sci Med. 2006;5(2):340–8.
4. Denadai BS, Gomide EB, Greco CC. The relationship between onset of blood lactate accumulation, critical velocity, and maximal lactate steady state in soccer players. J Strength Cond Res. 2005;19(2):364–8.
5. Gupta S, Rohatgi A, Ayers CR, et al. Cardiorespiratory fitness and classification of risk of cardiovascular disease mortality. Circulation. 2011;123(13):1377–83.
6. Krustrup P, Mohr M, Nybo L, Jensen JM, Nielsen JJ, Bangsbo J. The Yo-Yo IR2 test: physiological response, reliability, and application to elite soccer. Med Sci Sports Exerc. 2006;38(9):1666–73.
7. Wasserman K, Whipp BJ, Koyl SN, Beaver WL. Anaerobic threshold and respiratory gas exchange during exercise. J Appl Physiol. 1973;35(2):236–43.
8. Lansley KE, Dimenna FJ, Bailey SJ, Jones AM. A ‘new’ method to normalise exercise intensity. Int J Sports Med. 2011;32(7):535–41.
9. Krustrup P, Mohr M, Amstrup T, et al. The yo-yo intermittent recovery test: physiological response, reliability, and validity. Med Sci Sports Exerc. 2003;35(4):697–705.
10. Jones AM, Vanhatalo A. The ‘Critical Power’ concept: applications to sports performance with a focus on intermittent high-intensity exercise. Sports Med. 2017;47(1 Suppl):65–78.
11. Morton RH, Billat LV. The critical power model for intermittent exercise. Eur J Appl Physiol. 2004;91(2–3):303–7.
12. Poole DC, Burnley M, Vanhatalo A, Rossiter HB, Jones AM. Critical power: an important fatigue threshold in exercise physiology. Med Sci Sports Exerc. 2016;48(11):2320–34.
13. Jones AM, Wilkerson DP, DiMenna F, Fulford J, Poole DC. Muscle metabolic responses to exercise above and below the “critical power” assessed using 31P-MRS. Am J Physiol Regul Integr Comp Physiol. 2008;294(2):R585–93.
14. Zamparo P, Bolomini F, Nardello F, Beato M. Energetics (and kinematics) of short shuttle runs. Eur J Appl Physiol. 2015;115(9):1985–94.
15. Zamparo P, Zadro I, Lazzer S, Beato M, Sepulcri L. Energetics of shuttle runs: the effects of distance and change of direction. Int J Sports Physiol Perform. 2014;9(6):1033–9.
16. Saari A, Dicks ND, Hartman ME, Pettitt RW. Validation of the 3-min all-out exercise test for shuttle running prescription. J Strength Cond Res. 2017. [Epub ahead of print]. doi: 10.1519/JSC.0000000000002120.
17. Hughson RL, Orok CJ, Staudt LE. A high velocity treadmill running test to assess endurance running potential. Int J Sports Med. 1984;5(1):23–5.
18. 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 xFitness. 1994;34(1):23–37.
19. Broxterman RM, Ade CJ, Poole DC, Harms CA, Barstow TJ. A single test for the determination of parameters of the speed-time relationship for running. Respir Physiol Neurobiol. 2013;185(2):380–5.
20. Burnley M, Doust JH, Vanhatalo A. A 3-min all-out test to determine peak oxygen uptake and the maximal steady state. Med Sci Sports Exerc. 2006;38(11):1995–2003.
21. Vanhatalo A, Doust JH, Burnley M. Determination of critical power using a 3-min all-out cycling test. Med Sci Sports Exerc. 2007;39(3):548–55.
22. Jones AM, Vanhatalo A, Burnley M, Morton RH, Poole DC. Critical power: implications for determination of VO2max and exercise tolerance. Med Sci Sports Exerc. 2010;42(10):1876–90.
23. Ozyener F, Rossiter HB, Ward SA, Whipp BJ. Influence of exercise intensity on the on- and off-transient kinetics of pulmonary oxygen uptake in humans. J Physiol. 2001;533(Pt 3):891–902.
24. Poole DC, Ward SA, Gardner GW, Whipp BJ. Metabolic and respiratory profile of the upper limit for prolonged exercise in man. Ergonomics. 1988;31(9):1265–79.
25. Burnley M. Estimation of critical torque using intermittent isometric maximal voluntary contractions of the quadriceps in humans. J Appl Physiol (1985). 2009;106(3):975–83.
26. Cheng CF, Yang YS, Lin HM, Lee CL, Wang CY. Determination of critical power in trained rowers using a three-minute all-out rowing test. Eur J Appl Physiol. 2012;112(4):1251–60.
27. Lauderdale MA, Hinchcliff KW. Hyperbolic relationship between time-to-fatigue and workload. Equine Vet J Suppl. 1999;30:586–90.
28. Tsai MC, Thomas SG. Three-minute all-out test in swimming. Int J Sports Physiol Perform. 2017;12(1):27–35.
29. Kellawan JM, Tschakovsky ME. The single-bout forearm critical force test: a new method to establish forearm aerobic metabolic exercise intensity and capacity. PLoS One. 2014;9(4):e93481.
30. Bloomfield J, Polman R, O’Donoghue P. Physical demands of different positions in FA premier league soccer. J Sports Sci Med. 2007;6(1):63–70.
31. Skiba PF, Chidnok W, Vanhatalo A, Jones AM. Modeling the expenditure and reconstitution of work capacity above critical power. Med Sci Sports Exerc. 2012;44(8):1526–32.
32. Bergstrom HC, Housh TJ, Zuniga JM, et al. Differences among estimates of critical power and anaerobic work capacity derived from five mathematical models and the three-minute all-out test. J Strength Cond Res. 2014;28(3):592–600.
33. Gama MCT, Dos Reis IGM, Sousa FAB, Gobatto CA. The 3-min all-out test is valid for determining critical power but not anaerobic work capacity in tethered running. PLoS One. 2018;13(2):e0192552.
34. Karsten B, Hopker J, Jobson SA, et al. Comparison of inter-trial recovery times for the determination of critical power and W′ in cycling. J Sports Sci. 2017;35(14):1420–5.


Copyright © 2018 by the American College of Sports Medicine