Skip to main content
NCBI home page
Springer logoLink to Springer
. 2026 Jun 29;126(10):5225–5234. doi: 10.1007/s00421-026-06290-x

Rates of maximal blood lactate accumulation in the extreme intensity exercise domain: implications for training prescription

Ozgur Ozkaya 1,2,✉, Hakan As 1,2, Arda Peker 1,2, Mark Burnley 3, Andrew M Jones 4
PMCID: PMC13633953  PMID: 42371065

Abstract

Purpose

This study investigated the rate of blood lactate accumulation (vLa) across six maximal sprint efforts and identified a breakpoint (vLaBP) in the vLa–power relationship.

Methods

Twelve physically active males completed 5, 10, 15, 20, 25, and 30-s all-out sprints in randomised order on separate days, with resistances set at 10 to 12.5% of body mass. The vLa was calculated as the change in blood lactate between the resting and peak value (∆La) divided by the test duration. A piecewise linear regression was used to identify the vLaBP and its corresponding power output (P-vLaBP) in the vLa–power relationship. Pulmonary V̇O2 responses were modelled using V̇O2 data obtained from exercise and the subsequent 15-min recovery period.

Results

A distinct P-vLaBP was identified at 795 ± 135 W and 11.2 ± 1.5 W·kg−1, corresponding to a vLa of 0.50 ± 0.10 mM·s–1, ∆La of 6.50 ± 1.27 mM, and sprint duration of 13.3 ± 1.9 s. The vLa did not increase significantly during the 30, 25, 20, and 15-s all-out sprint tests (0.41 ± 0.04 to 0.48 ± 0.10 mM·s⁻1; p > 0.05) but increased markedly in the 10 and 5-s tests (0.58 ± 0.14 mM·s⁻1, p < 0.05; 0.7 ± 0.20 mM·s⁻1, p < 0.01). V̇O2 on-kinetics below the P-vLaBP (30 and 15 s) was best described by linear function (R2 ≥ 0.96; SEE ≤ 272 mL·min–1; p < 0.001), while the V̇O2 data could not be modelled for sprints performed above the P-vLaBP (10 and 5 s).

Conclusion

The present study demonstrated that the P-vLaBP separates ranges of power outputs with discrete V̇O2 kinetics which are characteristic of different subzones within the extreme intensity exercise domain.

Keywords: All-out sprint, Breakpoint in vLa, Cycle ergometry, Maximal lactate formation rate, V̇O2 kinetics, vLamax

Introduction

Within the exercise sciences, exercise intensity domains are typically classified as moderate, heavy, severe, and extreme (Hill et al. 2002; Poole et al. 2005; Burnley and Jones 2007; Poole and Jones 2012; Bishop et al. 2025). The moderate, heavy and severe intensity exercise domains are characterised by distinct dynamic behaviours of V̇O2 kinetics and blood lactate concentration (La), and these behaviours are consistent within each domain even at different absolute power outputs (Whipp and Mahler 1980; Gaesser and Poole 1996; Poole et al. 2005; Burnley and Jones 2007). However, V̇O2 and blood La kinetics have been less well characterised for the extreme intensity exercise domain. This is surprising given that, for a healthy and physically active individual, the extreme intensity exercise domain incorporates a potential range of power outputs almost twice as large as the sum of the moderate, heavy, and severe intensity domains (Ozkaya et al. 2023).

Notably, a recent study reported that V̇O2 kinetics change significantly as the power output increases within the extreme intensity exercise domain, and recognised three subzones, termed extreme-1, extreme-2 and extreme-3 (Ozkaya et al. 2023). Based on this framework, the V̇O2 kinetics in the extreme-1 intensity exercise zone is exponential, but it is not possible to attain maximal O2 uptake (V̇O2max) at task failure. In the extreme-2 zone, V̇O2 kinetics are best described by a linear function, while V̇O2 cannot be modelled due to the very brief exercise duration in the extreme-3 zone. Given the breadth of the extreme intensity domain, it has been suggested that these sub-categories may help in understanding physiological responses to exercise and in guiding training. Nevertheless, although a subzone approach within the extreme intensity exercise domain has been proposed, supported by discrete V̇O2 on- and off-response kinetics, methods for determining the transition intensities (i.e., as the power output or running speed) that separate the subzones within the domain have not yet been established.

Recently, there has been increasing interest in the rate of blood lactate accumulation (vLa) during short-term maximal exercise and especially its potential role in the evaluation of athletes’ performance (Quittmann et al. 2020, 2021; Yang et al. 2023; Dunst et al. 2023; Meixner et al. 2024a, b; Langley et al. 2024; Thron et al. 2024; Quittmann 2026). It has been reported that end-exercise peak blood La obtained from 5–10-s maximal efforts is significantly lower than that measured following 30–60-s sprint durations (Heck et al. 2003; Dunst et al. 2023; Langley et al. 2024). However, those 5–10-s short sprint bursts lead to a higher maximal lactate accumulation rate (vLamax) (Mavroudi et al. 2023; Langley et al. 2024; Porter and Langley 2025; Quittmann 2026). If there is an inverse relationship between longer sprint durations that elicit a higher end-exercise peak blood La response and shorter sprint durations that elicit a higher vLa response, this might be detectable as a breakpoint in the vLa–power relationship (vLaBP). This breakpoint would occur at a specific power output (P-vLaBP) and have a corresponding specific sprint duration (t-vLaBP). It is possible that this P-vLaBP may represent a transition between exercise in which there are adequate V̇O2 data that may be described by a linear function (i.e., extreme-2) and exercise in which V̇O2 cannot be modelled due to the very brief exercise duration dominated by breath-holding consistent with a Valsalva manoeuvre (i.e., extreme-3).

The aim of this study was, therefore, to investigate blood La dynamics during recovery from maximal sprint exercise of different durations and to calculate the corresponding vLa. We hypothesised that plotting the vLa values against the corresponding mean power outputs might reveal a P-vLaBP, which separates the extreme-3 from extreme-2 subzone based on both blood La dynamics and V̇O2 kinetics.

Materials and methods

Participants

This study received ethical approval from the Ege University Ethics Committee (Protocol no: 22-2T/16, 2 October 2022) and conformed to the principles outlined in the Declaration of Helsinki (1983). Prior to participation, all procedures, along with potential risks and benefits, were clearly explained to the volunteers, and written informed consent was obtained in accordance with the approved documentation. The study recruitment was open to both sexes, but only male participants volunteered to take part. A total of twelve healthy, physically active males took part in the study (mean ± SD: age: 26.6 ± 6.6 years; height: 1.76 ± 0.07 m; body mass: 71.3 ± 12.9 kg). None of the participants reported any cardiovascular, pulmonary, metabolic or neuromuscular disorders. Participants were instructed to maintain consistent dietary habits throughout the study period, refrain from strenuous exercise in the 24 h preceding each testing session and avoid caffeine or alcohol intake during this same time frame.

Experimental overview

This study utilised a counter-balanced, randomised, within-subject crossover (repeated measures) design to examine different sprint durations, integrating blood lactate measurements, vLa–power relationships, and V̇O2 on- and off-kinetics. The study protocol comprised eight laboratory visits per participant, each separated by 24–48 h, and was completed over a period of 2–3 weeks. A familiarisation session was undertaken before the sprint trials to minimise single practice effects. Following a ramp incremental exercise test, each participant performed six all-out sprint tests (i.e., 5, 10, 15, 20, 25, and 30 s) on separate days. The calculated vLa values were plotted against the corresponding power outputs to define a breakpoint (i.e., vLaBP) in the relationship. The vLamax was also determined for each participant. Time-dependent V̇O2 responses were modelled to describe V̇O2 kinetics. All exercise tests were conducted in a well-ventilated laboratory, under standardised conditions (20–22 °C temperature, ~20.8% O2, ~0.05% CO2, and 50–60% relative humidity). The time of the day allocated for testing was also standardised (1.5 ± 0.4 h) to control for diurnal variance.

Procedures

Ramp incremental tests

Ramp incremental tests were performed on an electromagnetically braked cycle ergometer (Monark LC7, Vansbro, Sweden) with a sampling interval of 1 Hz. The manufacturer certifies that measurement error is below 2% when the ergometer is regularly calibrated. Based on previous studies indicating that an optimal knee angle at the pedal’s bottom dead centre position is approximately 150° (Bohlmann 1981; de Vey Mestdagh 1998), the seat height was adjusted to provide a knee angle of 150° for each participant. Ramp incremental tests commenced with 4 min of cycling without resistance before the power output was increased by 30 W·min−1. Participants were asked to maintain their preferred cadence for as long as possible (i.e., approximately 70–90 rpm). The tests were terminated when participants’ cadence fell more than 10 rpm below their preferred cadence for longer than 10 s, despite verbal encouragement. Pulmonary gas exchange and ventilation data were measured breath-by-breath throughout the ramp tests using a metabolic cart (Cosmed Quark CPET, Rome, Italy). Data were initially filtered to remove occasional errant breaths resulting from e.g., swallows, coughs, or sighs. V̇O2 data points lying more than four standard deviations from the local mean (five breaths) were deleted and were smoothed using a 5-breath moving average. Data were then exported in 5-s binned averages. The highest 30-s moving average of V̇O2 (i.e., V̇O2max) and the corresponding power output (i.e., P-V̇O2max) were determined for each participant.

Familiarisation sessions

Familiarisation sessions for the all-out sprint tests were performed on a mechanically braked cycle ergometer (Monark Peak Bike 894E, Varberg, Sweden). During these sessions, all participants accelerated the flywheel to its maximum speed, and 7–8 s sprints were performed without additional frictional resistance. The performance criterion in this first stage of familiarisation was reaching peak flywheel speed within 5–6 s depending on flywheel inertia (Ozkaya 2013). Participants who met the criterion in two consecutive trials were given 5 min of rest before proceeding. In the second stage of familiarisation, following the pre-acceleration phase, a load corresponding to 10% of body mass was applied to the system, once the cadence reached 120 rpm. Subsequent all-out trials were considered successful when the time to peak power ≤ 2 s was achieved in two consecutive trials (Ozkaya 2013). Participants fulfilling these criteria were deemed to be familiarised and were admitted to the all-out sprint sessions.

All-out sprint sessions

Since we know that a traditional Wingate-type 30-s sprint is characterised by a linear behaviour in V̇O2 kinetics (Beneke et al. 2002; Ozkaya et al. 2014) and is typically performed within the extreme-2 intensity exercise zone (Ozkaya et al. 2023), the longest all-out sprint test duration was set to 30-s Wingate-type task. In this phase of the study, each participant completed six all-out sprint tests (5, 10, 15, 20, 25, and 30 s) on separate days in a randomised cross-over design. To obtain valid vLa estimations, resistance was systematically increased by 0.5% over the typical Wingate load (i.e., 10% of body mass) for each test, and all tests were performed all-out without pacing. Body mass was measured before each testing session using a calibrated digital scale (Seca 217, Hamburg, Germany) with participants wearing minimal clothing and no shoes. Accordingly, the resistance applied for 30, 25, 20, 15, 10, and 5-s sprints corresponded to 10%, 10.5%, 11%, 11.5%, 12%, and 12.5% of body mass, respectively. Participants completed a 5-min warm-up at ~100 W at their preferred cadence (70–90 rpm), followed by a 5-min rest. Each test then began with 4 min of no-resistance cycling at ~50 rpm. During the final 30 s of the baseline period, participants stopped pedalling and awaited the start command. All tests were performed in a seated position, starting from a static position with the dominant leg at the top of the first pedal stroke. A verbal countdown (“3–2–1–go”) signalled the start of each sprint, at which point participants were encouraged to produce maximal effort from the first pedal stroke. Once cadence exceeded 120 rpm, the ergometer automatically applied the pre-set resistance (i.e., 10 to 12.5% of body mass), initiating data acquisition. Participants were verbally encouraged to maintain maximal effort throughout the sprint. Power output (i.e., W and W·kg−1) and cadence (i.e., rpm) were recorded continuously. The highest instantaneous power production during the sprint was considered the Pmax, while the mean power output measured during each sprint was recorded as the Pavg. The highest 30-s moving average of V̇O2 measured during the sprints and the following 15-min recovery periods were also analysed from 5-s binned averages.

Estimation of the vLa and vLamax

Blood La values were analysed, using 20-μl blood samples, which were collected from the fingertip before, at the end, and at the 1st, 3rd, 5th, 7th and 10th min of the recovery periods, by the enzymatic amperometric method (Biosen C-line, EKF Diagnostics, GmbH, Barleben, Germany). A resting blood La sample was obtained upon arrival at the laboratory, before the warm-up, after ≥ 5 min of seated rest. The change in blood La (i.e., ∆La in mM) between the resting (Larest) and post-exercise peak La values (Lapeak) obtained from all six sprint tests were calculated. The vLa (in mM·s–1) was considered the ∆La divided by the sprint duration (ttest) (Eq. 1). vLamax was defined as the highest vLa value obtained across the six sprint tests.

graphic file with name d33e887.gif 1

where vLa is the lactate accumulation rate; Lapeak is the peak blood La response to each sprint obtained from post-exercise measurements; Larest is the resting blood La measurement; and ttest is the sprint duration. Note that the vLamax was considered the highest vLa value obtained across the six sprints.

Estimation of breakpoints in the relationship between vLa and power output

Calculated vLa values (on the y-axis) were plotted against the Pavg during the sprints (on the x-axis) to derive a vLaBP for each participant. A two-segment linear regression model (SigmaPlot; Systat Software Inc., San Jose, CA) was used to mathematically determine the vLaBP. Regression analysis essentially fits two separate linear regression lines to different domains of the data defining the gradient of the first and second segment, the change in gradient from the first to second segments, and the breakpoint position (Muggeo 2003). The power output at the breakpoint position (i.e., P-vLaBP) and the exercise duration corresponding to the P-vLaBP (i.e., t-vLaBP) were then recorded for each participant individually. Breakpoint positions which were determined statistically were then visually verified by five sports physiologists, who were asked to independently determine breakpoint positions in the vLa–power output relationships (Ozkaya et al. 2025).

Analysis of pulmonary V̇O2 kinetics

The breath-by-breath V̇O2 data were initially modelled using a mono-exponential function. However, this approach failed to provide an adequate fit, likely due to the rapid termination of the exercise. In contrast, the V̇O2 kinetics was more appropriately described by a linear model, which yielded a superior goodness of fit (Ozkaya et al. 2023) (Eq. 2):

graphic file with name d33e977.gif 2

where V̇O2(t) is the V̇O2 responses achievable at any time during exercise; V̇O2(base) is the baseline V̇O2 during the last 1 min of no-resistance cycling; and a is the slope of the line.

Furthermore, post-exercise V̇O2 kinetics was described by exponential decay models based on data obtained from 15-min recovery period. Thus, the behaviour of the V̇O2 kinetics over time was fitted by single and double exponential decay models (Eqs. 3 and 4):

graphic file with name d33e1003.gif 3

where V̇O2(t) is the V̇O2 response achieved at any time during post-exercise; V̇O2(base) is the baseline V̇O2 during the last 1 min of no-resistance cycling; and Afc is the amplitude of the fast component, τfc is the time constant of the fast component, and TDfc is the time delay of the post-exercise V̇O2 response.

graphic file with name d33e1026.gif 4

where V̇O2(t) is the V̇O2 response achieved at any time during post-exercise; V̇O2(base) is the baseline V̇O2 during the last 1 min of no-resistance cycling; Afc is the amplitude of the fast component, τfc is the time constant of the fast component, and TDfc is the time delay of the fast component; and Asc is the amplitude of the slow component, τsc is the time constant of the slow component, and TDsc is the time delay of the slow component.

Linear, single and double exponential decay models were applied to describe the V̇O2 kinetics using Sigma-Plot 14.0 software (Systat Software Inc., San Jose, CA, USA). Regression analyses were used to test the fitting behaviour of the V̇O2 data based on the standard error of the estimate (SEE) and the coefficient of determination (R2). On- and off- V̇O2 kinetics, which could not be modelled due to very short exercise durations in response to very high-power outputs, were assessed visually by five sports physiologists, who were asked to independently determine the V̇O2 shape over time (Ozkaya et al. 2023).

Statistical analyses

All statistical analyses were performed using SPSS Statistics 26 (IBM Corp, Armonk, NY, USA). Normality was tested with the Shapiro-Wilk test, and homogeneity of variance was assessed by Levene’s test. Sphericity was evaluated using Mauchly’s test: when assumptions were violated, degrees of freedom were corrected using Greenhouse-Geisser (ε < 0.75) or Huynh-Feldt (ε > 0.75) as appropriate. Differences in power output, V̇O2, La, and time related variables across sprint tests were then analysed using a repeated-measures ANOVA. Pairwise comparisons were conducted with Bonferroni adjustment for multiple comparisons. Effect size (ES) values for repeated measures ANOVA were computed using partial eta squared (Inline graphic) and categorised as follows: (i) small ES when 0.04 < Inline graphic ≤ 0.25; (ii) moderate ES when 0.25 < Inline graphic ≤ 0.64; and (iii) large ES when Inline graphic > 0.64 (Ferguson 2009). A Pearson product-moment correlation was computed to evaluate the relationship between ∆La and vLa derived from the sprint tests. Correlation magnitudes were interpreted using the qualitative thresholds proposed by Hopkins et al. (2009): trivial: (r < 0.10), small (0.10 ≤ r < 0.30), moderate (0.30 ≤ r < 0.50), large (0.50 ≤ r < 0.70), very large (0.70 ≤ r < 0.90), and nearly perfect (r ≥ 0.90). Significance levels were set at p ≤ 0.05. Data are presented as mean ± SD.

Results

The mean time to limit of tolerance at the end of the ramp incremental exercise test was 11.6 ± 1.7 min. The mean V̇O2max of the group was 50.2 ± 5.1 mL·min–1·kg–1 and the P-V̇O2max corresponded to 334 ± 56 W and 4.7 ± 0.4 W·kg−1. The highest Pmax (1027 ± 212 W; 14.5 ± 2.6 W·kg−1) and vLamax (0.70 ± 0.20 mM·s–1) were achieved during the 5-s all-out sprint test for each participant. Peak cadence values obtained from the six all-out efforts performed from 30 to 5 s were 134 ± 8 rpm, 138 ± 8 rpm, 136 ± 7 rpm, 132 ± 9 rpm, 137 ± 8 rpm and 137 ± 8 rpm, respectively. The two-segment linear regression model identified a distinct P-vLaBP at 795 ± 135 W and 11.2 ± 1.5 W·kg−1 (R2 = 0.98 ± 0.03; SEE% = 2.90 ± 1.78; p ≤ 0.01). This corresponded to a ∆La of 6.50 ± 1.27 mM, a vLa of 0.50 ± 0.10 mM·s–1, and a t-vLaBP of 13.3 ± 1.9 s (Table 1). Figure 1 shows an individual vLaBP, identified as the intersection point derived from plotting the vLa values against the Pavg from six all-out sprint efforts, obtained from a representative participant.

Table 1.

Parameters associated with the 5, 10, 15, 20, 25, and 30-s all-out sprints

Variables Exercise duration vLaBP p-value Inline graphic
5-s sprint 10-s sprint 15-s sprint 20-s sprint 25-s sprint 30-s sprint
Pmax (W·kg−1) 14.37 ± 2.61 13.78 ± 2.50 13.40 ± 1.99 13.03 ± 1.92a 13.03 ± 1.94a 12.60 ± 1.47ab 13.36 ± 1.92 p < 0.001 0.36
Pavg (W·kg−1) 13.07 ± 2.38 11.93 ± 1.75a 10.88 ± 1.44ab 10.04 ± 1.37abc 9.52 ± 1.15abcd 8.64 ± 0.76abcde 11.25 ± 1.47a p < 0.001 0.84
∆La (mM) 3.49 ± 1.02 5.76 ± 1.44a 7.18 ± 1.47ab 9.30 ± 1.72abcd 11.16 ± 1.50abcde 12.22 ± 1.21abcdef 6.50 ± 1.27a p < 0.001 0.95
vLa (mM·s–1) 0.70 ± 0.20 0.58 ± 0.14a 0.48 ± 0.10ab 0.47 ± 0.09ab 0.45 ± 0.06ab 0.41 ± 0.04ab 0.50 ± 0.10a p < 0.001 0.66
Open in a new tab

Pmax: the highest instantaneous power production during the sprints; Pavg: the mean power output measured during the sprints; ∆La: delta blood lactate response; vLa: the rate of blood lactate accumulation; Inline graphic: partial eta squared. Data are presented as mean ± SD.

aSignificantly different from 5-s sprint, p < 0.05

bSignificantly different from10-s sprint, p < 0.05

cSignificantly different from vLaBP, p < 0.05

dSignificantly different from 15-s sprint, p < 0.05

eSignificantly different from 20-s sprint, p < 0.05

fSignificantly different from 25-s sprint, p < 0.05

Fig. 1.

Fig. 1

Open in a new tab

Assessment of the breakpoint in the relationship between the rate of blood lactate accumulation (vLa) and power output derived from six all-out sprints (from 30 to 5 s) in a representative participant. The vLaBP was determined as the intersection of the two regression lines (black arrow). Data points are shown with a fill gradient from light to dark colour in proportion to the decreasing sprint durations

There was a significant main effect of sprint duration on ∆La and vLa (∆La: p < 0.001, Inline graphic = 0.95; and vLa: p < 0.001; Inline graphic = 0.66). During the 30 to 15-s all-out sprints, vLa responses did not increase significantly (0.41 ± 0.04 to 0.48 ± 0.10 mM∙s−1; p > 0.05); however, a marked increase in vLa was achieved during the 10-s (0.58 ± 0.14 mM∙s−1; p < 0.05) and especially the 5-s sprints (0.70 ± 0.20 mM∙s−1; p < 0.01). Additionally, there was a significant difference in vLa between the 10 s and 5 s sprint tests (p < 0.05). Figure 2 illustrates the negative correlation between vLa and ∆La responses across the sprint efforts.

Fig. 2.

Fig. 2

Open in a new tab

Scatterplot of the relationship between the rate of blood lactate accumulation (vLa) and change in blood lactate concentration (∆La) based on group-averaged values across six all-out sprints (from 5 to 30 s). The figure shows the very large negative correlation between vLa and ∆La

V̇O2 on-kinetics of the all-out sprints performed below the P-vLaBP (i.e., for 15-s, 20-s, 25-s and 30-s all-out sprints) was best described by a linear function (R2 ≥ 0.96; SEE ≤ 272 mL·min–1; p < 0.001), while the V̇O2 data could not be modelled for all-out sprints performed above the P-vLaBP (i.e., for 5-s and 10-s all-out sprints). As expected, for the 5-s and 10-s shorter bursts performed above the P-vLaBP, all participants preferred to hold their breath (Fig. 3). Moreover, V̇O2 responses continued to increase during the first 5–20 s of recovery periods for the 30, 25, 20, and 15-s sprints. Following the achievement of the peak response, V̇O2 decreased exponentially during recovery and were described by double exponential model (R2 ≥ 0.96; SEE ≤ 148 mL·min–1; p < 0.001). Similarly, because the first breath was seen within the first moments of recovery, the highest V̇O2 values were always attained within the recovery period following the 5 and 10-s sprints. However, there was an M-shape in V̇O2 during the recovery period and those response kinetics could not be modelled by traditional functions following the 5 and 10-s sprints.

Fig. 3.

Fig. 3

Open in a new tab

Pulmonary V̇O2 responses to six all-out sprint tests performed at durations ranging from 30 to 5 s in a representative participant. Breath-by-breath V̇O2 responses were interpolated to 1-s intervals. V̇O2 kinetics was characterised by an apparently linear profile during those sprints performed at 15–30 s; however, during 5 and 10-s sprints, participant frequently exhibited breath-holding behaviour, with peak V̇O2 typically reached during recovery and displaying an M-shaped profile. B: baseline; E: exercise; R: recovery

Discussion

The purpose of the present study was to investigate whether individual P-vLaBP derived from the vLa–power relationship from six all-out sprint tests corresponds to the boundary which separates the extreme-3 from the extreme-2 subzone. Consistent with our hypothesis, the P-vLaBP was found to separate a range of power outputs wherein V̇O2 kinetics during exercise could be best described with a linear function from a range of power outputs wherein V̇O2 kinetics could not be modelled at all. The upper boundary of the extreme-2 intensity exercise zone may therefore be considered to represent the highest all-out effort beyond which the vLa shifts from limited to rapidly accelerating behaviour. The vLaBP corresponded to a mean power output of 795 W and 11.2 W·kg−1 (i.e., approximately 238% of P-V̇O2max and 78% of Pmax). Accordingly, the extreme-3 exercise zone includes exercise durations up to approximately 13 s.

The extreme-3 intensity exercise zone of our participants had a range of approximately 232 W and 3.3 W·kg−1 between Pmax and P-vLaBP (i.e., 1027 – 795 W; 14.5 – 11.2 W·kg−1). Within this zone, which comprises very short-term maximal efforts, participants typically prefer to hold their breath, and the peak V̇O2 is reached during the recovery period, with the recovery V̇O2 response resembling an M-shape (Lamarra 1990; Ozkaya et al. 2023). It is known that breath holding during such short-term, extreme intensity exercise reflects the Valsalva manoeuvre, which significantly increases systolic blood pressure and supports the expression of strength and power (Narloch and Brandstater 1995). If the total exercise test duration incorporates a significant period of breath holding, it is not possible to measure or model V̇O2 data during exercise (Ozkaya et al. 2023). In contrast, exercise performed at power outputs below the P-vLaBP transition (i.e., those in the 15–30 s duration range), performed within the extreme-2 exercise zone, resulted in an apparently linear V̇O2 response during exercise along with greater metabolic acidosis. It has been previously suggested that the linearity in V̇O2 kinetics for extreme-2 exercise may be due to a combination of V̇O2 reflecting a very high demand but being constrained by the V̇O2max and the relatively short exercise duration (Ozkaya et al. 2023). Clearly, however, maximal exercise in this time range enables a high rate of anaerobic glycolysis, as reflected by a greater end-exercise blood La concentration.

On the other hand, a recent systematic review fitted four separate bi-exponential models to cycling data, excluding the 3- and 60-s data points, and predicted that the shortest test duration for eliciting vLamax is approximately 4.5–5 s (Langley et al. 2025). At shorter exercise durations, the time-equivalent accounting for predominantly phosphocreatine (PCr) driven ATP resynthesis with minimal glycolytic activation and negligible blood La accumulation (tPCr) is likely to have a disproportionate impact, potentially inflating variability in vLamax estimates. Indeed, unlike V̇O2max, which is typically operationalised as the highest V̇O2 value attained within a standardised time interval (i.e., 1-min), vLamax is calculated by the ∆La divided by sprint duration. As exercise intensity increases and sprint duration shortens, this ratio-based computation may produce a vLamax value that is disproportionately high and which may not accurately reflect the underlying glycolytic rate. This issue could be considered one of the most important limitations to the ability to obtain a vLamax response and to use this value as an indicator of maximal glycolytic power. Leaving these debates aside, our findings suggest that, rather than attempting to determine a single vLamax value, it may be more appropriate to frame the phenomenon as a vLa breakpoint (i.e., vLaBP), with the corresponding power output representing a transition that separates the extreme-2 and extreme-3 subzones.

A further methodological issue in the vLamax literature concerns the tPCr and the variability in how it is defined and calculated. It has been claimed that tPCr refers to the beginning of the sprint performance that is dominated by PCr breakdown without a significant activation of glycolysis or accumulation of blood La (Mader 1984, 1994; Heck et al. 2003; Yang et al. 2023). There are different assumptions when calculating the tPCr. For example, for the 100 m, 200 m, and 400 m sprints (i.e., approximately 10–12-s, 20–25-s and 45–60-s duration), it is suggested that tPCr be fixed at 3-s, 4-s, and 8-s time periods, respectively (Mader 1994; Heck et al. 2003). Meixner et al. (2024a) suggest the use of a standard tPCr of 3.5 s. In another method, the time to reach Pmax is considered to be the tPCr duration (Heck et al. 2003; Manunzio et al. 2016). On the other hand, it has also been suggested that the tPCr should be defined as the time window from the onset of exercise to the point at which power output has decreased by 3.5% after the Pmax is reached (Weber 2003; Heck et al. 2003; Langley et al. 2024). As another example, Yang et al. (2023) have recently suggested that the contribution of oxidative energy metabolism should be taken into account to reveal a more genuine vLamax. However, it is important to note that this time interval approach in which glycolytic activity does not occur is merely hypothetical assumption (Gaitanos et al. 1993; Brooks 2018), and none of these methods has yet gained widespread acceptance or evolved into a broadly endorsed consensus. In a recent review, Wackerhage et al. (2025) highlighted several conceptual and methodological limitations associated with the estimation of vLamax, suggesting that calculating a peak vLa value (i.e., vLapeak) without subtracting a tPCr may offer a more valid alternative. Indeed, in empirical settings, it has been suggested to assume “tPCr = 0” and to calculate vLamax by dividing ∆La by test duration (Quittmann 2026), given that phosphagen contribution can only be estimated with limited precision (Hirvonen et al. 1987; Gastin 2001; Quittmann 2026) and that accelerated glycolysis and lactate accumulation begin immediately at exercise onset (Gastin 2001; Brooks et al. 2022; Quittmann 2026). Accordingly, we took this approach in the present study.

From a practical perspective, the P-vLaBP could be useful in discriminating and prescribing sprint interval training (SIT) and repeated sprint training (RST). Over the past two decades, the physiological and performance benefits of SIT (i.e., 20–30-s all-out sprints, performed at approximately 150–200% of P-V̇O2max) have been consistently demonstrated (Burgomaster et al. 2005; Gibala et al. 2006). Similar improvements have been described even following very short-duration, high-intensity intermittent exercise of 8–12 s, i.e., RST (Buchheit and Laursen 2013). In the current study, the t-vLaBP, corresponding to an effort time of approximately 13 s, may be considered a boundary separating SIT and RST prescriptions. Indeed, exercise involving comparatively longer sprint bouts and yielding high end-exercise blood La appears more closely aligned with SIT, while much shorter bouts, often accompanied by breath-holding and characterised by low end-exercise blood La responses despite a high lactate accumulation rate, are consistent with RST.

In terms of limitations, it should be noted that post-exercise La was sampled at the 1st, 3rd, 5th, 7th and 10th min. Future work should sample blood more frequently (e.g., every minute) to refine breakpoint estimation. Additionally, further investigations should consider a broader range of populations, including well-trained or elite athletes, females, and older adults, and should also evaluate the validity of the approach for other exercise modalities. The reliability of vLaBP determination could also be investigated in future studies.

Conclusions

The present study identified a P-vLaBP that separated the power outputs that provide the greatest vLa from those that generate higher levels of La. This P-vLaBP boundary was found to occur at approximately 13-s of all-out effort for the physically active participants in this study. The all-out sprints performed below the P-vLaBP resulted in apparently linear V̇O2 response kinetics, while it was not possible to measure or model V̇O2 data for the all-out sprints performed above the P-vLaBP. The results of this study may be useful in differentiating extreme-2 from extreme-3 exercise and potentially in the prescription of SIT and RST.

Acknowledgements

The authors would like to express their gratitude to the participants for their valuable contribution to this study.

Abbreviations

Afc

Amplitude of the fast component

Asc

Amplitude of the slow component

ES

Effect size

La

Blood lactate concentration

Lapeak

Peak blood La value measured in recovery period

Larest

Resting blood La value

Pavg

Mean power output

Pmax

Highest instantaneous power production

P-vLaBP

Power output corresponding to the vLaBP

P-V̇O2max

Power output corresponding to the V̇O2max

PCr

Phosphocreatine

RST

Repeated sprint training

R2

Coefficient of determination

SD

Standard deviation

SEE

Standard error of estimation

SIT

Sprint interval training

TDfc

Time delay of the fast component

TDsc

Time delay of the slow component

tPCr

Time equivalent to account for energy resynthesis via PCr

ttest

Sprint duration

t-vLaBP

Exercise duration corresponding to the vLaBP

vLa

Rate of blood La

vLaBP

Breakpoint in the relationship between the vLa and power output

vLamax

Maximal vLa value

vLapeak

Peak vLa value

V̇O2

Pulmonary O2 uptake

V̇O2(base)

V̇O2 obtained from the last 1 min of no-resistance cycling

V̇O2(t)

V̇O2 responses achievable at any time during exercise

V̇O2max

Maximal O2 uptake

Inline graphic

Partial eta squared

τfc

Time constant of the fast component

τsc

Time constant of the slow component

∆La

Change in blood La between the resting and peak

Author contributions

The authors confirm contribution to the paper as follows: OO, AMJ, and MB contributed to the study conception and design. Material preparation, data collection and analysis were performed by HA and AP. The first draft of the manuscript was written by OO, AMJ, and MB and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

Open access funding provided by the Scientific and Technological Research Council of Türkiye (TÜBİTAK). The funding was provided by AixTech Technology Ltd.

Data Availability

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interest

The authors declare they have no conflicts of interests.

Footnotes

Publisher’s Note

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

References

  1. Beneke R, Pollmann C, Bleif I et al (2002) How anaerobic is the wingate anaerobic test for humans? Eur J Appl Physiol 87:388–392. 10.1007/s00421-002-0622-4 [DOI] [PubMed] [Google Scholar]
  2. Bishop DJ, Beck B, Biddle SJH et al (2025) Physical activity and exercise intensity terminology: a joint American College of Sports Medicine (ACSM) expert statement and Exercise and Sport Science Australia (ESSA) consensus statement. J Sci Med Sport 28:980–991. 10.1016/j.jsams.2024.11.004 [DOI] [PubMed] [Google Scholar]
  3. Bohlmann JT (1981) Injuries in competitive cycling. Phys Sportsmed 9:117–124. 10.1080/00913847.1981.11711083 [DOI] [PubMed] [Google Scholar]
  4. Brooks GA (2018) The science and translation of lactate shuttle theory. Cell Metab 27:757–785. 10.1016/j.cmet.2018.03.008 [DOI] [PubMed] [Google Scholar]
  5. Brooks GA, Arevalo JA, Osmond AD et al (2022) Lactate in contemporary biology: a phoenix risen. J Physiol 600:1229–1251. 10.1113/JP280955 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Buchheit M, Laursen PB (2013) High-intensity interval training, solutions to the programming puzzle: Part I: cardiopulmonary emphasis. Sports Med 43:313–338. 10.1007/s40279-013-0029-x [DOI] [PubMed] [Google Scholar]
  7. Burgomaster KA, Hughes SC, Heigenhauser GJF et al (2005) Six sessions of sprint interval training increases muscle oxidative potential and cycle endurance capacity in humans. J Appl Physiol 98:1985–1990. 10.1152/japplphysiol.01095.2004 [DOI] [PubMed] [Google Scholar]
  8. Burnley M, Jones AM (2007) Oxygen uptake kinetics as a determinant of sports performance. Eur J Sport Sci 7:63–79. 10.1080/17461390701456148 [DOI] [PubMed] [Google Scholar]
  9. de Vey Mestdagh K (1998) Personal perspective: in search of an optimum cycling posture. Appl Ergon 29:325–334. 10.1016/S0003-6870(97)00080-X [DOI] [PubMed] [Google Scholar]
  10. Dunst AK, Hesse C, Feldmann A, Holmberg HC (2023) A novel approach to determining the alactic time span in connection with assessment of the maximal rate of lactate accumulation in elite track cyclists. Int J Sports Physiol Perform 18:157–163. 10.1123/ijspp.2021-0464 [DOI] [PubMed] [Google Scholar]
  11. Ferguson CJ (2009) An effect size primer: a guide for clinicians and researchers. Prof Psychol Res Pract 40:532–538. 10.1037/a0015808 [DOI] [Google Scholar]
  12. Gaesser GA, Poole DC (1996) The slow component of oxygen uptake kinetics in humans. Exerc Sport Sci Rev 24:35–71. 10.1249/00003677-199600240-00004 [DOI] [PubMed] [Google Scholar]
  13. Gaitanos GC, Williams C, Boobis LH, Brooks S (1993) Human muscle metabolism during intermittent maximal exercise. J Appl Physiol 75:712–719. 10.1152/jappl.1993.75.2.712 [DOI] [PubMed] [Google Scholar]
  14. Gastin PB (2001) Energy system interaction and relative contribution during maximal exercise. Sports Med 31:725–741. 10.2165/00007256-200131100-00003 [DOI] [PubMed] [Google Scholar]
  15. Gibala MJ, Little JP, van Essen M et al (2006) Short-term sprint interval versus traditional endurance training: similar initial adaptations in human skeletal muscle and exercise performance. J Physiol 575:901–911. 10.1113/jphysiol.2006.112094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Heck H, Schulz H, Bartmus U (2003) Diagnostics of anaerobic power and capacity. Eur J Sport Sci 3:1–23. 10.1080/17461390300073302 [DOI] [Google Scholar]
  17. Hill DW, Poole DC, Smith JC (2002) The relationship between power and the time to achieve VO2max. Med Sci Sports Exerc 34:709–714. 10.1097/00005768-200204000-00023 [DOI] [PubMed] [Google Scholar]
  18. Hirvonen J, Rehunen S, Rusko H, Härkönen M (1987) Breakdown of high-energy phosphate compounds and lactate accumulation during short supramaximal exercise. Eur J Appl Physiol Occup Physiol 56:253–259. 10.1007/BF00690889 [DOI] [PubMed] [Google Scholar]
  19. Hopkins WG, Marshall SW, Batterham AM, Hanin J (2009) Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 41:3–13. 10.1249/MSS.0b013e31818cb278 [DOI] [PubMed] [Google Scholar]
  20. Lamarra N (1990) Variables, constants, and parameters: clarifying the system structure. Med Sci Sports Exerc 22:88–95 [PubMed] [Google Scholar]
  21. Langley J, Haase R, Nitzsche N, Porter M (2025) Methodological approaches in testing maximal lactate accumulation rate - νLamax: a systematic review. J Sci Cycl 14:9–9. 10.28985/1425.jsc.09 [DOI] [Google Scholar]
  22. Langley JO, Ng SC, Todd EE, Porter MS (2024) VLamax: determining the optimal test duration for maximal lactate formation rate during all-out sprint cycle ergometry. Eur J Appl Physiol 124:2461–2472. 10.1007/s00421-024-05456-9 [DOI] [PubMed] [Google Scholar]
  23. Mader A (1984) Eine Theorie zur Berechnung der Dynamik und des steady state von Phosphorylierungsszzustand und Stoffwech selaktivität der Muskelzelle als Folge des Energiebedarfs
  24. Mader A (1994) Aussagekraft der Laktatleistungskurve in Kombination mit anaeroben Tests zur Bestimmung der Stoffwechselkapazität. In: Stellenwert der Laktat bestimmung in der Leistungsdiagnostik. pp 133–52
  25. Manunzio C, Mester J, Kaiser W, Wahl P (2016) Training intensity distribution and changes in performance and physiology of a 2nd place finisher team of the Race across America over a 6 month preparation period. Front Physiol 7:642. 10.3389/fphys.2016.00642 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mavroudi M, Kabasakalis A, Petridou A, Mougios V (2023) Blood lactate and maximal lactate accumulation rate at three sprint swimming distances in highly trained and elite swimmers. Sports (Basel) 11:87. 10.3390/sports11040087 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Meixner B, Nusser V, Koehler K et al (2024) Reliability of power output, maximal rate of capillary blood lactate accumulation, and phosphagen contribution time following 15-s sprint cycling in amateur cyclists. Physiol Rep 12:e16086. 10.14814/phy2.16086 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Meixner BJ, Nusser V, Koehler K et al (2024) Relationship of peak capillary blood lactate accumulation and body composition in determining the mechanical energy equivalent of lactate during sprint cycling. Eur J Appl Physiol 124:3399–3407. 10.1007/s00421-024-05529-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Muggeo VMR (2003) Estimating regression models with unknown break-points. Stat Med 22:3055–3071. 10.1002/sim.1545 [DOI] [PubMed] [Google Scholar]
  30. Narloch JA, Brandstater ME (1995) Influence of breathing technique on arterial blood pressure during heavy weight lifting. Arch Phys Med Rehabil 76:457–462. 10.1016/S0003-9993(95)80578-8 [DOI] [PubMed] [Google Scholar]
  31. Ozkaya O (2013) Familiarization effects of an elliptical all-out test and the Wingate test based on mechanical power indices. J Sports Sci Med 12:521–525 [PMC free article] [PubMed] [Google Scholar]
  32. Ozkaya O, As H, Peker A et al (2025) Resolving differences between MLSS and CP by considering rates of change of blood lactate during endurance exercise. Med Sci Sports Exerc 57:217–226. 10.1249/MSS.0000000000003548 [DOI] [PubMed] [Google Scholar]
  33. Ozkaya O, Colakoglu M, Kuzucu EO, Delextrat A (2014) An elliptical trainer may render the Wingate all-out test more anaerobic. J Strength Cond Res 28:643–650. 10.1519/JSC.0b013e3182a20f77 [DOI] [PubMed] [Google Scholar]
  34. Ozkaya O, Jones AM, Burnley M et al (2023) Different categories of VO2 kinetics in the “extreme” exercise intensity domain. J Sports Sci 41:2144–2152. 10.1080/02640414.2024.2316504 [DOI] [PubMed] [Google Scholar]
  35. Poole DC, Jones AM (2012) Oxygen uptake kinetics. Compr Physiol 2:933–996. 10.1002/cphy.c100072 [DOI] [PubMed] [Google Scholar]
  36. Poole DC, Kindig CA, Behnke BJ, Jones AM (2005) Oxygen uptake (VO2) kinetics in different species: a brief review. Comp Exerc Physiol 2:1–15. 10.1079/ECP200445 [DOI] [Google Scholar]
  37. Porter M, Langley J (2025) The relationship between muscle oxygen saturation kinetics and maximal blood lactate accumulation rate across varying sprint cycle durations. Eur J Sport Sci 25:e12242. 10.1002/ejsc.12242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Quittmann OJ (2026) Maximal lactate accumulation rate (c˙Lamax): current evidence and future directions for exercise testing and training. Eur J Appl Physiol 126:1–36. 10.1007/s00421-025-06022-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Quittmann OJ, Appelhans D, Abel T, Strüder HK (2020) Evaluation of a sport-specific field test to determine maximal lactate accumulation rate and sprint performance parameters in running. J Sci Med Sport 23:27–34. 10.1016/j.jsams.2019.08.013 [DOI] [PubMed] [Google Scholar]
  40. Quittmann OJ, Schwarz YM, Mester J et al (2021) Maximal lactate accumulation rate in all-out exercise differs between cycling and running. Int J Sports Med 42:314–322. 10.1055/a-1273-7589 [DOI] [PubMed] [Google Scholar]
  41. Thron M, Woll A, Doller L et al (2024) Physiological and locomotor profiling enables to differentiate between sprinters, 400-m runners, and middle-distance runners. J Strength Cond Res 38:1419–1427. 10.1519/JSC.0000000000004801 [DOI] [PubMed] [Google Scholar]
  42. Wackerhage H, Kabasakalis A, Seiler S, Heck H (2025) Is the vLamax for glycolysis what the VO2max is for oxidative phosphorylation? Sports Med. 10.1007/s40279-025-02259-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Weber S (2003) Berechnung leistungsbestimmender parameter der metabolischen Aktivität auf zellulärer Ebene mittels fahrrader gometrischer Untersuchungen
  44. Whipp BJ, Mahler M (1980) Dynamics of gas exchange during exercise. In: West JB (ed) Pulmonary gas exchange. Academic Press, New York, pp 33–96 [Google Scholar]
  45. Yang W-H, Park S-Y, Kim T et al (2023) A modified formula using energy system contributions to calculate pure maximal rate of lactate accumulation during a maximal sprint cycling test. Front Physiol 14:1147321. 10.3389/fphys.2023.1147321 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.

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

RESOURCES