Skip to main content
NCBI home page
Springer logoLink to Springer
. 2026 Jul 30;126(10):5101–5106. doi: 10.1007/s00421-026-06331-5

Maximal aerobic speed: methodological implications and unintended consequences

Fernando Gabe Beltrami 1,✉, Øyvind Støren 2, Jan Helgerud 3
PMCID: PMC13633949  PMID: 42530639

Abstract

Maximal aerobic speed (MAS) represents the minimal speed where a further increase in speed will not result in a further increase in oxygen consumption (V̇O2). MAS is determined from the relationship between maximal V̇O2 (V̇O2max) and running economy, the latter also expressed as the O2 cost of running. There is no standardized method for determination of MAS in the literature, but three approaches are more dominant: the first approach consists of determining a loss of linearity between V̇O2 and speed. The second approach performs a linear extrapolation from the relationship between submaximal V̇O2 and speed towards V̇O2max. The third approach divides V̇O2max by the O2 cost of running measured over one or more speeds. We explore the assumptions and flaws of the different approaches. We show that the first approach lacks sufficient objectivity. The second approach, based on linear extrapolation, is prone to significant error, particularly when changes in O2 cost happen at the lower or higher speeds tested, to the extent that higher sub-maximal V̇O2 for a given speed results in improved MAS and vice-versa. We suggest that MAS should be assessed using a simple division of V̇O2max by the O2 cost of running determined from a race-relevant speed, which can be easily integrated into a single testing session.

Introduction

Maximal aerobic speed (MAS) represents the minimal speed where a further increase in speed will not result in an increase in oxygen consumption (V̇O2), or, in other words, the minimal speed required to achieve maximal V̇O2 (V̇O2max) (Lacour et al. 1991). Although in this article MAS is discussed within the context of running, the concept can be readily used in other situations, such as skiing (Støren et al. 2023) or cycling (Støren et al. 2014), using power instead of speed when appropriate. MAS is determined from the relationship between V̇O2max and running economy, the latter calculated either as V̇O2 at sub-maximal running speeds or converted into the O2 cost of running (Cr, in ml kg− 1km− 1) by dividing V̇O2 by speed, paying attention to the appropriate conversion of the units of distance and time. MAS, therefore, has no regards to the actual capacity to maintain that speed or power for prolonged periods or the capacity to sustain steady-state, as is the case for variables such as critical speed or maximal lactate steady state (Nixon et al. 2021).

The concept of MAS nonetheless has relevance for endurance events (Morgan et al. 1989; Jones 1998), as it integrates the two main determinants of performance, the third one being the fraction of V̇O2max sustained during an event, although the utility of the latter for differentiating athletes of different levels has been questioned recently, because athletes of different levels show a similar percentage of V̇O2max at the lactate threshold (Johansen et al. 2025). It has also been argued that the fractional utilization of V̇O2max is a result of the event time (and therefore performance), and not a determinant of it (Støa et al. 2010). Theoretically, if two athletes have the same MAS the one who runs an event at a higher percentage of their MAS will have a higher overall speed, but this can only be assessed retrospectively, and the percentage of V̇O2max at the lactate threshold does not seem suited for discriminating the faster between these two athletes (Johansen et al. 2025).

Another benefit of the quantification of MAS is that it bypasses the issue of how to best express V̇O2max or running economy, whether using absolute values, relative to body mass or the usage of different exponents in allometric scaling (Berg 2003; Helgerud et al. 2010). As long as the same denotation is used for both V̇O2max and the submaximal VO2, the resulting MAS will be the same, providing a metric which is expressed in the same units as performance itself, i.e., a speed. Besides using similar units, it is important that both V̇O2max and Cr are calculated and expressed in a way that minimizes variability and improves the practical value of MAS, including that changes in the underlying components of MAS affect it in a predictable manner. MAS has been calculated in several ways (Thron et al. 2024), as also reviewed by Billat and Koralsztein (1996), and no standard mathematical definition of MAS exists in the literature. Here, we consider some methodological implications of the different approaches and show how the choice of calculation could affect the results obtained on a similar dataset. We finish with a recommendation for standardizing the assessment of MAS.

First approach – vV̇O2max

Billat and colleagues (1995) defined vV̇O2max as “the lowest speed which elicited a V̇O2 value equal to VO2max. Subsequently, Billat and colleagues (1996) and Hanon and colleagues (2007) defined vV̇O2max as “the first stage during an incremental test where V̇O2max was attained (mean of the 3 consecutive and highest values) and where the subsequent increase in V̇O2 was less than 2.1 ml kg− 1 min− 1 for an increase in speed of 1 km h− 1 (the original protocol employed 3-min stages and 1 km h− 1 increase in speed between stages when approaching V̇O2max) (Billat et al. 1995). Moreover, V̇O2max must be associated with a blood lactate, a respiratory exchange ratio and a heart rate equals respectively, to 8 to 12 mmol.l− 1, 1.1 and minimum 90% of the theoretical maximal HR (220-age)”. While both these definitions are particularly prone to small fluctuations in data within an exercise test (Myers and Ashley 1997; Beltrami et al. 2014), the latter definition is particularly problematic. If it is accepted that the O2 cost of running is independent of speed for a given runner, as has been repeatedly demonstrated (Brueckner et al. 1991; Di Prampero et al. 1993; Fletcher et al. 2009), then for an athlete with an O2 cost of running of 220 ml kg−1km− 1 the expected increase in V̇O2 for a 1 km h− 1 change in speed is 3.7 ml kg−1 min− 1. However, if the O2 cost of running is 170 ml kg−1 km− 1, within what is reported in high-level athletes (Jones et al. 2021), the expected increase in V̇O2 for a 1.0 km h− 1 change in speed will be only 2.6 ml kg−1 min− 1. Clearly, using an absolute 2.1 ml kg−1 min− 1 threshold for defining loss of linearity in V̇O2 increases does not work universally.

Although the work of Taylor and colleagues (1955) is used by Billat et al. (1995) to justify the fixed threshold of 2.1 ml kg− 1 min− 1 (Billat and Koralsztein 1996) to define the attainment of V̇O2max, Taylor et al. based this number on the fact that the average V̇O2 change between stages for their protocol (which was an increase in treadmill incline of 2.5% at each stage, not 1.0 km h−1 change in speed) was 4.2 ml kg−1 min−1. The authors then decided that differences in V̇O2 of less than half the expected values could be reasonably expected to represent a physiological plateau. Thus, a correct interpretation of that work would be to define V̇O2max as the minimal speed where changes in V̇O2 are less than half of what would be expected (Beltrami et al. 2014). This approach, however, is also problematic during continuous tests, as multiple stages can have successive differences which would satisfy this criterion (Beltrami et al. 2014). The utilization of secondary criteria based on respiratory exchange ratio, heart rate or blood lactate levels does little to ensure the attainment of V̇O2max, as these can both be attained long before V̇O2max (Poole et al. 2008), or not achieved at all despite maximal efforts. Given the limitations of the method, we consider this approach to determining MAS less objective and more error prone than the other two.

Lastly, for avoidance of confusion within the literature, we note that both Daniels and Danies (1992) and Morgan and colleagues (1989) used the term speed at V̇O2max and the abbreviation vV̇O2max in their work. These authors, however, defined and calculated vV̇O2max as described in our second approach (see section below), which is now termed MAS. Therefore, the vV̇O2max of Daniels and Danies (1992) and Morgan et al. (1989) is not equivalent and should not confused with the more recent definition of vV̇O2max used by Billat and colleagues (1996).

Second approach

The second approach to determining MAS is based on the extrapolation of the relationship between sub-maximal, steady-state V̇O2 and running speed to the value of the measured V̇O2max (blue line in Fig. 1). Here, whether V̇O2 is expressed absolute, relative to body mass or using allometric scaling (Di Prampero et al. 1993; Støren et al. 2012; Rønnestad et al. 2020), is of no consequence to the calculated MAS. While this method is popular in the literature, it is particularly prone to large oscillations in MAS in response to small changes in sub-maximal V̇O2, whether these are biological or errors in measurement. Consider the three theoretical incremental tests with steady-state V̇O2 measurements during sub-maximal states from the same athlete who achieves a V̇O2max, of 65 ml kg− 1 min− 1 in all three attempts of the test, shown in Fig. 1; Table 1. Using this method, an increase in V̇O2 during the first stage of the test (2.0 ml kg−1 min− 1 higher, implying a worse running economy at this stage) will improve MAS by 3.5% (red line in Fig. 1). This is because the effect of an increase in V̇O2 at the early parts of the extrapolation will be to flatten the relationship between V̇O2 and speed. The opposite will happen if V̇O2 decreases during the first stage (green line in Fig. 1): MAS will decrease. Although a lower V̇O2 would imply better running economy for that particular speed, it increases the steepness of the relationship between speed and V̇O2. Given the nature of linear interpolations, both the magnitude and direction of a given change in V̇O2 depends on where in the line the change occurs, as shown in Table 2. It can be seen that a similar change (in the example of Table 2 an increase in submaximal V̇O2 of 2.0 ml kg− 1 min− 1) will have an effect ranging from + 3.5 to −3.8% on the estimated MAS. The magnitude of the effect will be more pronounced the further the data is from the middle in either direction, as the linear regression is more sensitive to these data points. While Cr typically does not change between speeds at group level (Di Prampero et al. 1993; Fletcher et al. 2009; Jones et al. 2021), substantial within individual variability can be present between stages of a test. This is well illustrated by Fig. 2 in Jones et al. (2021), where Cr oscillated at times by more than 10.0 ml kg− 1 km− 1 in world-class runners between the different speeds of a single test. Such oscillations, whether biological or technical, can cause substantial variability in tests responses over time and mask or enhance true changes in MAS.

Fig. 1.

Fig. 1

Open in a new tab

Three (theoretical) tests performed on the same athlete, with equal V̇O2max, of 65.0 ml kg−1 min−1 in all three occasions, demarked by the dotted black line. Note how a change in V̇O2 at sub-maximal speeds has a large effect on the resulting MAS. By the nature of linear regressions, changes in sub-maximal V̇O2 at either extreme of the line will have larger effects on MAS. Changes in V̇O2 in the slower speeds, as seen in the example, will have contradictory effects on MAS: a higher submaximal V̇O2 will result in higher MAS, and vice-versa

Table 1.

Physiological data from Fig. 1

Speed (km h− 1) Test 1 Test 2 Test 3
V̇O2 (ml kg−1 min−1) C (ml kg− 1 km− 1) V̇O2 (ml kg−1 min−1) C (ml kg− 1 km− 1) V̇O2 (ml kg−1 min−1) C (ml kg− 1 km− 1)
12 39.0 195 41.0 205 37.0 185
13 42.0 194 42.0 194 42.0 194
14 45.7 196 45.7 196 45.7 196
15 48.8 195 48.8 195 48.8 195
16 52.3 196 52.3 196 52.3 196
V̇O2max (ml kg−1 min−1) 65.0 65.0 65.0
MAS (Method 2, in km h− 1) 19.8 20.5 19.3
Cr (average all speeds) (ml kg− 1 km− 1) 195.2 197.2 193.2
MAS (Method 3, in km h− 1) 20.0 19.8 20.2
Open in a new tab

Values for V̇O2max in ml･kg-1･min-1 need to be multiplied by 60 before division by Cr in ml･kg-1･km-1 to obtain a speed in km/h

MAS maximal aerobic speed, V̇O2 oxygen uptake, V̇O2max maximal oxygen uptake, Cr O2 cost. LInes highlighted in bold show different values between the different tests.

Table 2.

Effects of changes in V̇O2 and VO2max on MAS

Speed (km h− 1) V̇O2 (ml·kg−1·min−1)
Baseline case Change stage 1 Change stage 2 Change stage 3 Change stage 4 Change stage 5
12.0 39.0 41.0 39.0 39.0 39.0 39.0
13.0 42.0 42.0 44.0 42.0 42.0 42.0
14.0 45.7 45.7 45.7 47.7 45.7 45.7
15.0 48.8 48.8 48.8 48.8 50.8 48.8
16.0 52.3 52.3 52.3 52.3 52.3 54.3
Delta MAS for a 5% increase in V̇O2max (from 65.0 to 68.2 ml · kg−1· min−1)
MAS (method 2, %) + 4.9% + 5.4% + 5.2% + 5.0% + 4.7% + 4.6%
MAS (method 3, %) + 5.0% + 5.0% + 5.0% + 5.0% + 5.0% + 5.0%
V̇O2max (ml kg−1 min−1) 65 65 65 65 65 65
MAS (method 2, km h− 1) 19.8 20.5 (+ 3.3%) 20.1 (+ 1.2%) 19.7 (−0.6%) 19.4 (−2.3%) 19.1 (−3.8%)
MAS (method 3, km h− 1) 20.0 19.8 (−1.0%) 19.8 (−1.0%) 19.8 (−1.0%) 19.8 (−1.0%) 19.8 (−1.0%)
Open in a new tab

The different cases at the top present a baseline case and the effect of a change in sub-maximal V̇O2 of + 2.0 ml･kg-1･min-1 at each of the different sub-maximal stages according to method 2 and 3 for determination of MAS. Values in bold are different from baseline case. The bottom of the table shows the relative delta (%) in MAS resulting from a 5% increase in V̇O2max in each case. As the different cases have different slopes, the changes in V̇O2max will vary the magnitude of change in MAS when method 2 is used. This is not the case with method 3, where the changes in V̇O2max are equal in direction and magnitude to the changes in V̇O2max, as no linear interpolation is used

MAS maximal aerobic speed, V̇O2 oxygen uptake, V̇O2max maximal oxygen uptake

A similar issue is present with regards to variations in V̇O2max. Because the slope of the V̇O2 x speed relationship is not necessarily 1.0, a given change in VO2max will have a smaller (if the slope is higher than 1.0) or larger (if the slope is lower than 1.0) effect on the calculated MAS. An example of this situation is shown in bottom of Table 2, where it can be seen that a 5% increase in V̇O2max will increase MAS by between 4.6 and 5.4% depending on the slope of the V̇O2 x speed relationship.

Third approach

Instead of relying on a linear extrapolation of submaximal V̇O2 to VO2max, the third approach divides V̇O2max by Cr (Di Prampero et al. 1993; Rønnestad et al. 2020) to obtain a speed. Care must be taken to use the appropriate units: V̇O2max in ml kg−1 min−1 needs to be multiplied by 60 to yield V̇O2max in ml kg− 1 h− 1 before a division by Cr results in a speed in km h− 1. Here, both Cr and V̇O2max can also be calculated gross (i.e. including) or net (i.e. deducting) of the resting V̇O2 (Brueckner et al. 1991). Using the same data as in Fig. 1, the average Cr across the tested sub-maximal speeds and the resulting MAS calculated using method 3 is shown in Table 1 (bottom two lines). Here, the 2.0 ml kg− 1 min− 1 change in O2 cost during the first stage of the test only results in a 1% increase in Cr, since the submaximal V̇O2 was constant for the other four speeds. Accordingly, the calculated MAS only varies by 1%, in the same direction and magnitude as the changes in Cr. This approach also makes the resulting MAS independent of where a change in CR occurs within the range of speeds selected, whether of biological or technical nature. As shown in Table 2, an increase in submaximal V̇O2 of 2.0 ml kg−1 min− 1 in any of the speeds used to determine Cr would have a similar effect on MAS (a decrease of 1.0%), whereas the magnitude of the effect ranged between + 3.5 – −3.8% with the linear interpolation method. Likewise, any changes in V̇O2max (Table 2) will have an effect which is similar in direction and magnitude in the calculated MAS.

One source of confusion with this method might be which stages are to be used when determining Cr. Although Cr is stable over a range of speeds (Helgerud et al. 2010), this needs to be standardized so that researchers and practitioners do not fall into the trap of using different numbers at every new measurement for a given athlete, or go outside the range where the relationship between V̇O2 and speed is linear. This issue is addressed in the final section of the manuscript (Practical recommendations).

Discussion

We examined the implications and limitations of different methodologies currently available to determine MAS during running in athletes. The first approach, based on the minimal speed where V̇O2max is reached, lacks sufficient accuracy in the quantification and should not be recommended. The second approach, based on linear extrapolation of the V̇O2 x speed relationship, is prone to unintended results, including that higher V̇O2 values in the early stages of a test can lead to higher MAS and vice-versa, a finding which is hard to justify from a physiological standpoint. Calculating MAS according to the third approach, by dividing V̇O2max by Cr (determined from a single relevant speed or averaged from a range of relevant speeds) retains the desirable properties of the calculation, integrating running economy and V̇O2max into a single value. Values then only change in the expected direction, i.e., MAS will decrease if either V̇O2max decreases or Cr increases, or vice-versa. Moreover, the magnitude of the changes will be directly proportional to the changes in both V̇O2max and Cr, with no speed exerting greater impact on the results, in lines with the notion that Cr is constant across speeds. Central to both methods 2 and 3, evidently, is the need to collect appropriate values of sub-maximal V̇O2 or Cr, paying attention that the data is collected at speeds where true steady-state can be achieved in both V̇O2 and CO2 output (Gaesser and Poole 1996). While there are also arguments in the literature about whether running economy or Cr should be determined using V̇O2 or metabolic power (in Joules) (Fletcher et al. 2009; Shaw et al. 2014), this discussion is not relevant to the current work, as the difference between the calculation methods remains the same.

Rejection of the slope-based estimation of MAS does not mean that different athletes cannot have different V̇O2 vs. speed slopes. Different slopes are in fact expected for athletes with different Cr. Daniels and Daniels (1992) had already noted that athletes with different specializations showed different slopes in their V̇O2 x speed relationship, with middle distance athletes showing flatter slopes compared with long distance athletes. They put forward the idea that running economy adapts better for the speed that is most relevant to the athlete. We note, however, that no subsequent study attempted to validate these findings, which come from a small set of well-trained individuals. Regardless of whether training can or cannot change the slope of V̇O2 x speed in runners, determination of MAS via the extrapolation method will be exposed to undesirable variation. When using low, non-relevant speeds in the extrapolation, these might have a disproportional influence in the calculation of MAS. An example of this can found in the work of Jones (1998), following an Olympic runner over 5 years: the MAS of the athlete (calculated using linear interpolation of V̇O2 from speeds between 14.0 and 19.0 km h− 1) increased by 2.6% between 1992 and 1993 (from 19.5 to 20.0 km h− 1). This happened even though V̇O2max decreased by 3.6% (from 68.5 to 66.0 ml kg− 1 min− 1). Running economy, reported only at 16.0 km h− 1, improved by 3.8% (from 53.0 to 51.0 ml kg− 1 min− 1), thus barely enough to offset the drop in V̇O2max. Therefore, the only explanation for the increase in MAS is that the slope of the V̇O2 x speed relationship must have flattened significantly, which could only have happened if either V̇O2 increased at the lower sub-maximal speeds used (which would be interpreted as a worse running economy at those speeds) or if it decreased at the higher speeds used, or a combination of both. Deriving MAS using method 3, from the race-relevant speeds, could have provided a clearer picture of the mechanisms behind the changes in MAS, or indeed have given an altogether different MAS value, as in the example from Table 1.

Practical recommendation

One possible solution for robust determination of MAS is to use only a range of speeds which are relevant to the population at hand. If an athlete competes at speeds ranging between 18.0 and 20.0 km h− 1, what is the value of allowing the variations in Cr at much lower speeds, for example 14.0 km h− 1, influence the resulting MAS? It could be argued that since endurance athletes have a speed range of 2.0–3.0 km h− 1 from 5-km races to the marathon, depending on their level, this speed range should be used for determination of MAS, thus decoupling it from lower, less relevant speeds from an incremental test. Cr has been shown to be stable over a range of speeds between 70 and 90% of V̇O2max (Helgerud et al. 2010), and this represents speeds typically encountered during endurance events. If this range of speed is within that of endurance performances, the prerequisite of achieving steady-state will also be met, although this needs checking for each athlete at every speed tested.

Given that MAS is commonly measured in athletes or endurance-trained individuals, for whom performance level is known or can be reasonably estimated, we propose that quantification of MAS for research studies can be performed based on a speed expected to yield approximately 75% of V̇O2max for athletes in endurance events lasting 2 or more hours, based on expected marathon speed, and close to 90% of V̇O2max for athletes focusing on shorter events. For the assessment of individual athletes, it makes most sense to simply use the speed which is more race relevant. In either case, this assessment can be easily integrated into a single day testing session where a preliminary warm-up is followed by 5-min bouts at the target speed, which is then followed by a ramp test to determine V̇O2max. To improve the accuracy of measures, the speeds used for determination of Cr should be tested at least in duplicate and results averaged together.

Author contributions

FGB conceived and drafted the manuscript. JH and OS provided critical commentary and contributed new ideas and inputs to the original draft. All authors approved the final version of the manuscript.

Funding

Open access funding provided by Swiss Federal Institute of Technology Zurich. No funding was received for preparation of this manuscript.

Declarations

Conflict of interest

The authors declare that they have no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Beltrami FG, Wong DP, Noakes TD (2014) High prevalence of false-positive plateau phenomena during VO < inf>2max testing in adolescents. J Sci Med Sport 10.1016/j.jsams.2013.07.012 [DOI] [PubMed] [Google Scholar]
  2. Berg K (2003) Endurance training and performance in runners: research limitations and unanswered questions. Sport Med 33:59–73 [DOI] [PubMed] [Google Scholar]
  3. Billat LV, Koralsztein JP (1996) Significance of the velocity at VO2max and time to exhaustion at this velocity. Sports Med 22:90–108 [DOI] [PubMed] [Google Scholar]
  4. Billat V, Renoux JC, Pinoteau J et al (1995) Times to exhaustion at 90, 100 and 105% of velocity at VO2 max (maximal aerobic speed) and critical speed in elite long-distance runners. Arch Physiol Biochem 103:129–135. 10.3109/13813459508996126 [DOI] [PubMed] [Google Scholar]
  5. Brueckner JC, Atchou G, Capelli C et al (1991) The energy cost of running increases with the distance covered. Eur J Appl Physiol Occup Physiol 62:385–389. 10.1007/BF00626607 [DOI] [PubMed] [Google Scholar]
  6. Daniels J, Daniels N (1992) Running economy of elite male and elite female runners. Med Sci Sports Exerc 24:483–489 [PubMed] [Google Scholar]
  7. Di Prampero PE, Capelli C, Pagliaro P et al (1993) Energetics of best performances in middle-distance running. J Appl Physiol 74:2318–2324. 10.1152/jappl.1993.74.5.2318 [DOI] [PubMed] [Google Scholar]
  8. Fletcher JR, Esau SP, MacIntosh BR (2009) Economy of running: beyond the measurement of oxygen uptake. J Appl Physiol 107:1918–1922. 10.1152/japplphysiol.00307.2009 [DOI] [PubMed] [Google Scholar]
  9. Gaesser GA, Poole DC (1996) The slow component of oxygen uptake kinetics in humans. Exerc Sport Sci Rev 24:35–71 [PubMed] [Google Scholar]
  10. Hanon C, Vivier L, Thomas C (2007) Time course of velocity and oxygen uptake during 1500 meters realised with a strategy of best chronometric performance. New Stud Athl 22:15–21 [Google Scholar]
  11. Helgerud J, Støren O, Hoff J (2010) Are there differences in running economy at different velocities for well-trained distance runners? Eur J Appl Physiol 108:1099–1105. 10.1007/s00421-009-1218-z [DOI] [PubMed] [Google Scholar]
  12. Johansen JM, Støa EM, Sunde A et al (2025) Elite Aerobic Endurance Performance: Is It Really Related to Lactate Threshold Expressed Relative to Peak Oxygen Uptake? Int J Sports Physiol Perform 20:875–879. 10.1123/ijspp.2024-0528 [DOI] [PubMed] [Google Scholar]
  13. Jones AM (1998) A five year physiological case study of an Olympic runner. Br J Sports Med 32:39–43. 10.1136/bjsm.32.1.39 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Jones AM, Kirby BS, Clark IE et al (2021) Physiological demands of running at 2-hour marathon race pace. J Appl Physiol 130:369–379. 10.1152/japplphysiol.00647.2020 [DOI] [PubMed] [Google Scholar]
  15. Lacour JR, Padilla-Magunacelaya S, Chatard JC et al (1991) Assessment of running velocity at maximal oxygen uptake. Eur J Appl Physiol Occup Physiol 62:77–82. 10.1007/BF00626760 [DOI] [PubMed] [Google Scholar]
  16. Morgan DW, Baldini FD, Martin PE, Kohrt WM (1989) Ten kilometer performance and predicted velocity at VO2max among well-trained male runners. Med Sci Sports Exerc 21:78–83. 10.1249/00005768-198902000-00014 [DOI] [PubMed] [Google Scholar]
  17. Myers J, Ashley E (1997) Dangerous curves: A perspective on exercise, lactate, and the anaerobic threshold. Chest 111:787–795 [DOI] [PubMed] [Google Scholar]
  18. Nixon RJ, Kranen SH, Vanhatalo A, Jones AM (2021) Steady-state V˙ O 2 above MLSS: evidence that critical speed better represents maximal metabolic steady state in well-trained runners. Eur J Appl Physiol 121:3133–3144. 10.1007/s00421-021-04780-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Poole DC, Wilkerson DP, Jones AM (2008) Validity of criteria for establishing maximal O2 uptake during ramp exercise tests. Eur J Appl Physiol 102:403–410. 10.1007/s00421-007-0596-3 [DOI] [PubMed] [Google Scholar]
  20. Rønnestad BR, Rømer T, Hansen J (2020) Increasing oxygen uptake in well-trained cross-country skiers during work intervals with a fast start. Int J Sports Physiol Perform 15:383–389. 10.1123/ijspp.2018-0360 [DOI] [PubMed] [Google Scholar]
  21. Shaw AJ, Ingham S, Folland JP (2014) The valid measurement of running economy in runners. Med Sci Sports Exerc 46:1968–73. 10.1249/MSS.0000000000000311 [DOI] [PubMed] [Google Scholar]
  22. Støa EM, Støren Ø, Enoksen E, Ingjer F (2010) Percent utilization of VO2 max at 5-km competition velocity does not determine time performance at 5 km among elite distance runners. J strength Cond Res 24:1340–1345. 10.1519/JSC.0b013e3181cc5f7b [DOI] [PubMed] [Google Scholar]
  23. Støren Ø, Bratland-Sanda S, Haave M, Helgerud J (2012) Improved VO2max and time trial performance with more high aerobic intensity interval training and reduced training volume: a case study on an elite national cyclist. J Strength Cond Res 26:2705–2711. 10.1519/JSC.0b013e318241deec [DOI] [PubMed] [Google Scholar]
  24. Støren Ø, Rønnestad BR, Sunde A et al (2014) A time-saving method to assess power output at lactate threshold in well-trained and elite cyclists. J Strength Cond Res 28:622–9. 10.1519/JSC.0b013e3182a73e70 [DOI] [PubMed] [Google Scholar]
  25. Støren Ø, Sunde A, Helgerud J et al (2023) Maximal aerobic and anaerobic power and time performance in 800 m double poling ergometer. Eur J Appl Physiol 123:1199–1208. 10.1007/s00421-023-05149-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Taylor HL, Buskirk E, Henschel A (1955) Maximal oxygen intake as an objective measure of cardio-respiratory performance. J Appl Physiol 8:73–80 [DOI] [PubMed] [Google Scholar]
  27. Thron M, Düking P, Ruf L et al (2024) Assessing anaerobic speed reserve: A systematic review on the validity and reliability of methods to determine maximal aerobic speed and maximal sprinting speed in running-based sports [DOI] [PMC free article] [PubMed]

Articles from European Journal of Applied Physiology are provided here courtesy of Springer

RESOURCES