Journal Logo

Special Communications: Methods

The validity of the lactate minimum test for determination of the maximal lactate steady state


Author Information
Medicine& Science in Sports & Exercise: August 1998 - Volume 30 - Issue 8 - p 1304-1313
  • Free


Peak aerobic power (V˙O2max) has long been associated with competitive running capability (11,18,46). However, in athletes with similar values for V˙O2max, submaximal physiological factors have been identified that are better able to discriminate differences in endurance running performance. For example, in groups of runners homogenous for V˙O2max, running economy has been shown to be highly correlated with competitive success (10). More recently, physiological measures such as the running velocity corresponding to V˙O2max (V-V˙O2max) (35), and performance measures such as the maximal running velocity attained in an incremental treadmill test (Vmax) (36), which relate to both the V˙O2max and the running economy characteristics of subjects, have been proposed as useful measures in work with elite runners.

The maximal lactate steady state (MLSS) can be defined as the highest blood lactate concentration at which a balance exists between the rate of appearance of lactate in the blood and the rate of removal of lactate from the blood during constant-load exercise of approximately 30 min duration (6,25,49). At exercise intensities below MLSS, a relative homeostasis is maintained, and ATP resynthesis takes place almost solely via aerobic pathways (16). Below MLSS, blood lactate falls over time following an initial rise or reaches an elevated but stable level (1,25,45). In contrast, at exercise intensities exceeding MLSS, blood lactate, minute ventilation, and V˙O2 never attain a steady state but continue to rise until exercise is terminated by fatigue (39,45). Given that the MLSS theoretically demarcates the exercise intensity above which anaerobic metabolism makes an increasingly important contribution to ATP resynthesis (16,43), it is reasonable to consider the running velocity associated with MLSS (V-MLSS) to be of importance in the determination of endurance running success (24,25,31). Submaximal physiological measures made during incremental exercise tests including the lactate threshold (Tlac) and the ventilatory threshold (Tvent) (13,40,42,51), and the "onset of blood lactate accumulation" at 4 mM blood lactate (OBLA) (25,31,48), have been identified as factors that powerfully predict endurance running performance. However, it is not clear how these measures relate to the MLSS of prolonged constant-velocity exercise.

Recently, Tegtbur et al. (52) reported the development of a new test to demarcate V-MLSS. The test, known as the lactate minimum test, involved subjects performing a short period of supramaximal exercise to induce hyperlactemia before commencing a standard incremental treadmill test. The blood lactate minimum ([Lac]BMIN) was defined as the velocity at which a curve fitted to the "U-shaped" blood lactate data derived from the incremental test reached a nadir. This lactate minimum point was supposed to represent a point of equilibrium between lactate production and removal, i.e., net clearance below[Lac]BMIN and net production above [Lac]BMIN. Tegtbur et al. (52) found that when 25 endurance runners ran for 8 km at the velocity at which the [Lac]BMIN occurred, they exhibited an elevated but steady [Lac]B. However, when the subjects ran at a velocity 0.68 km·h−1 above [Lac]BMIN, this resulted in significant accumulation of blood lactate, which prevented completion of the 8-km run in 11 subjects. From their data, Tegtbur et al. (52) proposed that the [Lac]BMIN test provides an accurate estimate of the V-MLSS.

The [Lac]BMIN test would seem to offer significant advantages over existing methods that might be used for estimation of the V-MLSS. Unlike the subjective selection of threshold points (as for Tlac and Tvent), a procedure that has been criticized for poor inter-reviewer reliability (21,60), the [Lac]BMIN may be determined objectively with the application of a zero-gradient tangent to a spline function fitting the blood lactate response to the incremental exercise test. Also, in contrast to methods utilizing reference blood lactate concentrations such as OBLA, the [Lac]BMIN is apparently unaffected by conditions of glycogen depletion (52).

Despite its appeal, the [Lac]BMIN test has yet to be independently validated against V-MLSS or compared with other physiological variables related to the accumulation of blood lactate during submaximal exercise. Further, the relationship between [Lac]BMIN and endurance performance has not been established. Therefore, the objective of this study was to examine the validity of the [Lac]BMIN test for the estimation of V-MLSS and to compare [Lac]BMIN with Tlac, Tvent, and OBLA. A secondary purpose of this study was to identify which of the plethora of maximal and submaximal physiological measures currently in routine use in sports physiology laboratories correlates most strongly with the 8-km running performance.



Thirteen male subjects volunteered to take part in this study in exchange for individual test results and advice relating to enhancing existing training programs. The subjects gave their written informed consent to the experimental procedures after having the possible benefits and risks of participation in the study fully explained to them. This study was approved by the Chelsea School Ethics Committee, University of Brighton. The mean ± SD age, height, body mass, and sum of four skinfolds (17) of the subjects were 24.9 ± 6.3 yr, 1.83 ± 0.7 m, 72.4 ± 5.5 kg, and 28.0 ± 8.0 mm, respectively. The subjects had been training for, and competing in, distance running races regularly over the past 2-10 yr and were well-trained, running an average of 46 ± 19 miles every week. The test battery was completed during the British spring, when the subjects had completed the majority of their off-season aerobic conditioning. Most of the subjects had competed at cross-country through the winter and were shortly to commence a specific preparation program for the track or road racing seasons. All subjects were experienced in laboratory exercise testing procedures and were fully habituated to treadmill running. Subjects were instructed to arrive at the laboratory in a rested and fully hydrated state, at least 2 h postprandial. They were forbidden to train strenuously in the 24 h preceding a test session. Before each test session, subjects performed a full individual warm-up regimen, and subjects wore the same running shoes and lightweight running kit for all tests.

Experimental Methods

The subjects performed 5 tests. These were: 1) the [Lac]BMIN test after Tegtbur et al. (52); 2) a series of five constant-velocity treadmill runs, of 30 min duration, for determination of the MLSS and the V-MLSS; 3) a multistage incremental treadmill protocol for determination of V-Tlac, V-Tvent, V-OBLA, and V˙O2max and for estimation of the running velocity at V˙O2max(V-V˙O2max); 4) an incremental treadmill protocol for determination of maximal treadmill velocity (Vmax); and 5) an 8-km time trial on flat roads. Details of the procedures and protocols are given below.

The tests were presented in random order and were performed on separate days with all tests completed within 2-4 wk. For each subject, laboratory testing took place at the same time of day and within 3 h of the time at which the simulated race was run (09:30 a.m.) to minimize the effect of diurnal biological variability.

Treadmill tests took place on a Woodway ELG2 treadmill (CardioKinetics, Salford, UK). During all treadmill running, except for the procedure used for the determination of V˙O2max when the treadmill was set to higher grades (see Specific Methods), the treadmill was set at a grade of 1%. This procedure has been shown to produce equivalent V˙O2 values for treadmill running and outdoor running over a wide range of velocities(29). The use of an appropriate treadmill grade to compensate for the lack of air resistance during treadmill running is especially important when results from treadmill tests are used to explain or predict outdoor running performance(29).

Respiratory variables were determined by collection of expired air in to Douglas bags over a timed period. Subjects breathed through a Salford respiratory valve of 80 mL deadspace and Falconia tubing (inspiratory and expiratory resistance for valve/tubing arrangement <3.0 cm H2O and <1.0 cm H2O, respectively, at 300 L·min−1). Expired air was analyzed for the concentrations of O2 and CO2 by sampling through a paramagnetic transducer (Servomex Series 1100, Crowborough, UK) and an infrared analyzer(Servomex model 1490), respectively. The O2 and CO2 analyzers were calibrated before and after each test using certified gases (British Oxygen Co., London, UK). Volumes were determined by dry gas meter (Harvard Ltd., Edenbridge, UK) previously calibrated against a wet spirometer (Tissot) and checked regularly against a precision 7-L gas syringe (Hans Rudolph Inc., Kansas). Heart rate (HR) was determined with an ECG-calibrated Polar Sport tester HR monitor (Polar Electro, Kempele, Finland). Fingertip capillary blood samples (20 μL) were made during short (10-20 s) interruptions in exercise, and whole blood lactate concentration([Lac]B) was determined in duplicate using an automated analyzer (Analox GM7, UK).

Specific Methods

Test 1: Blood lactate minimum test. The [Lac]BMIN was determined after the procedures furnished by Tegtbur et al. (52). Briefly, subjects performed two supramaximal treadmill runs of 300 and 200 m (range 18.5-25.5 km·h−1) at 5% treadmill grade, with a 1-min rest interval between runs to induce hyperlactemia. Following 8 min of slow walking to maximize [Lac]B, subjects ran 5-6 exercise stages of 3-min duration, commencing with a running velocity 2 km·h−1 below the subject's predicted current race velocity for 16 km. Fingertip blood samples for determination of [Lac]B were collected at the completion of each stage before subjects resumed running at a 1.0 km·h−1 higher velocity. The V-[Lac]BMIN(the velocity corresponding to the nadir on the [Lac]B vs running velocity plot) was determined from the zero gradient tangent to a spline function fitting the exercise [Lac]B data (Fig. 1).

Figure 1
Figure 1:
Determination of blood lactate minimum velocity in a typical subject.

Test 2: Constant-velocity tests for determination of the V-MLSS. Subjects performed five constant-velocity treadmill runs of 30 min duration for the determination of individual V-MLSS (Fig. 2a). Running velocities were selected according to the ability of the subject but were found to span approximately 65-90% of individual V˙O2max. These runs varied in velocity by 0.5 km·h−1 and were presented in random order on separate days. The [Lac]B was determined at rest and every 5 min throughout the runs until 30 min had elapsed or exhaustion was reached. Expired air was collected in a Douglas bag over the final minute preceding blood sampling (i.e., minutes 4-5, 9-10, 14-15, 19-20, 24-25, and 29-30) for determination of V˙O2. The mean HR over the final 30 s preceding blood sampling was recorded. The V-MLSS was defined as the highest velocity at which [Lac]B did not increase by more than 1.0 mM between minutes 10 and 30 of the constant velocity runs (6,21,49). Two subjects sustained minor injuries, and one subject chose to withdraw from the study before this test, so V-MLSS was only determined in 10 subjects.

Figure 2
Figure 2:
The responses of blood lactate (a) and oxygen uptake (b) to a series of constant-velocity treadmill runs used for determination of the maximal lactate steady state velocity in a typical subject. The solid line in panel b shows the V˙O2max established in the incremental treadmill test.

Test 3: Incremental test for determination of V-Tlac, V-Tvent,V-OBLA, and V˙O2max. V˙O2max and the running velocity at Tlac, Tvent, and OBLA were determined using a multistage incremental treadmill protocol similar to that described and validated by Weltman et al. (55). Subjects completed between 6 and 9 (generally 7) submaximal running stages of 3-min duration. The running velocities used for these submaximal stages were selected according to the ability of the subject, with the velocity of the first stage being equal to 2 km·h−1 below the subject's predicted current race velocity for 16 km. The submaximal stages were found to require approximately 60-95% of individual V˙O2max. Following each stage, a 20-μL sample of fingertip blood was collected before subjects resumed running at a 0.5 km·h−1 higher velocity. Expired air was collected into a Douglas bag over the last 60 s of each stage. When HR exceeded 95% of age-predicted maximum or [Lac]B exceeded 4 mM, the treadmill gradient was increased by 1% every minute until volitional exhaustion for determination of V˙O2max. During this period, expired air was collected over the last 45 s of each minute, and the highest V˙O2 recorded was accepted as V˙O2max. The V˙O2max determined with this protocol is not significantly different from V˙O2max determined with more conventional protocols of approximately 10 min duration (28).

Data were appropriately plotted, coded, and then presented blindly, in duplicate, to two reviewers for determination of V-Tlac (the velocity immediately preceding a sudden and sustained increase in [Lac]B) and V-Tvent (the velocity immediately preceding the first loss of linearity in plots of V˙CO2 against V˙O2(5)). The mean selection of the two reviewers was used in analysis. The reliability of data plot assessment was high both between reviewers (r = 0.95) and within the same reviewer (r = 0.95). The V-OBLA was determined by interpolation.

From this test, individual regression equations of V˙O2 on submaximal running velocity (of the form V = m·V˙O2 + c) were generated and solved for V˙O2max to provide an estimate of the running velocity at V˙O2max(V-V˙O2max). Although Morgan et al. (35) used data collected over a wide range of submaximal exercise in their study, we chose to use only data collected below Tvent in the generation of the regression equations. This approach was taken to avoid possible complications arising from upward curvilinearity in the V˙O2-running velocity relationship at high exercise intensities(23). The regression equations utilized a minimum of 4 sub-tvent data points.

Test 4: Maximal velocity test. For determination of maximal treadmill velocity (Vmax), a fast incremental treadmill protocol was utilized. The test started at a running velocity of 12.0 km·h−1, and running velocity was increased by 0.5 km·h−1 every minute. The Vmax was defined as the running velocity attained during the final completed 1-min stage.

Test 5: Performance trial. The 8-km performance trial took place on a wind-still day using an out-and-back course on a flat seafront promenade. Subjects were informed that the run should be an all-out time trial and were started alphabetically by surname at 30-s intervals so that "drafting" and other tactical considerations were avoided. The performance trial was run, as far as possible, under race conditions. Runners were marshalled over the entire course and were given "split times" at 2, 4, and 6 km. As an incentive, runners were awarded prizes according to their finishing time. No subject reported experiencing a "bad run," and performance times were in keeping with personal best times over similar distances. Two subjects set personal best times for 8 km during the performance trial.

Statistical Analysis

The relationship between the measured physiological variables and 8-km race performance was examined using Pearson Product Moment correlation coefficients. A repeated measures ANOVA with Tukey HSD post hoc tests was used for pairwise comparisons. The 5% level was chosen a priori to represent statistical significance. Data are presented as mean ± SEM unless otherwise indicated.


Mean± SD V˙O2max determined from test 3 was 60.7 ± 4.0 mL·kg−1·min−1. Nine subjects exhibited a plateau in V˙O2 with increasing work intensity, and for all subjects RER exceeded 1.10 at exhaustion. Mean ± SD HRmax was 193 ± 7 beats·min−1.

The relationship of V-[Lac]BMIN to V-MLSS and other measures of blood lactate accumulation during submaximal exercise. Comparisons were made between the V-[Lac]BMIN(test 1) and the V-MLSS (test 2), V-Tlac, V-Tvent, and V-OBLA (test 3).

The V-[Lac]BMIN, V-MLSS, V-Tlac, V-Tvent, and V-OBLA could be determined in 100, 100, 85, 100, and 69% of subjects, respectively (Table 1). The V-[Lac]BMIN(14.9 ± 0.2 km·h−1) was not significantly different from V-Tlac (15.1 ± 0.3 km·h−1) or V-Tvent (14.9± 0.3 km·h−1), but was significantly lower than V-OBLA(16.1 ± 0.2 km·h−1) and V-MLSS (15.7 ± 0.3 km·h−1)(P < 0.05) (Fig. 3). The V-MLSS was not significantly different from V-Tlac(Fig. 4), V-Tvent, or V-OBLA. The V-Tlac was not significantly different from the V-Tvent. The correlations between [Lac]BMIN and other laboratory measures of submaximal and maximal physiological function are shown in Table 2. The correlation between V-MLSS and V-[Lac]BMIN(r = 0.61; Fig. 2) was weaker than the correlations between V-MLSS and V-Tlac (r = 0.94; Fig. 4), V-Tvent(r = 0.77), and V-OBLA (r = 0.93).

Individual values for running velocity during the 8-km performance trial and at a number of physiological measures.
Figure 3
Figure 3:
The relationship between the running velocity at lactate minimum and the maximal lactate steady state.
Figure 4
Figure 4:
The relationship between the running velocity at lactate threshold and the maximal lactate steady state.
Inter-relationships between some laboratory measures of lactate accumulation during submaximal exercise.

Heart rate at [Lac]BMIN(173 ± 3 beats·min−1) was not significantly different from HR at Tlac (170 ± 4 beats·min−1) or HR at MLSS (176 ± 4 beats·min−1) but was significantly lower than HR at OBLA (181 ± 4 beats·min−1) (P < 0.05) and significantly higher than HR at Tvent (168 ± 2 beats·min−1)(P < 0.05). Heart rate at Tlac was not significantly different from HR at Tvent. The [Lac]B at [Lac]BMIN(4.5 ± 0.3 mM) was not significantly different from [Lac]B at MLSS (4.4 ± 0.4 mM), but both were significantly higher than [Lac]B at Tlac (2.5 ± 0.2 mM) (P < 0.05).

Physiological correlates to 8-km running performance. The relationship between 8-km running performance (test 5) and various maximal and submaximal physiological variables was investigated. Submaximal measures examined included V-[Lac]BMIN(test 1), V-MLSS (test 2), and V-Tlac, V-Tvent, and V-OBLA (test 3). Maximal measures included V˙O2max and the estimated velocity at V˙O2max(V-V˙O2max) (test 3) and maximal velocity (Vmax) (test 4).

Table 1 shows individual values for some of the physiological measures made and for 8-km performance. The mean running velocity for the 8-km run (16.0 ± 0.3 km·h−1) was not significantly different from V-OBLA or V-MLSS but was significantly higher than [Lac]BMIN, V-Tlac, and V-Tvent (P < 0.05) and significantly lower than V-V˙O2max(18.1 ± 0.4 km·h−1) and Vmax (19.0 ± 0.3 km·h−1) (P < 0.05). The Vmax was significantly higher than the V-V˙O2max (P < 0.05). The strongest correlations with 8-km performance were found with V-V˙O2max (r = 0.93; N= 13; Fig. 5), V-MLSS (r = 0.92; N = 10; Fig. 6), and V-Tlac (r = 0.93; N = 11). Variables that were less strongly correlated included V˙O2max (r = 0.69; N = 13), Vmax(r = 0.82; N = 13), V-[Lac]BMIN (r = 0.83;N = 13), V-Tvent (r = 0.81; N = 13), and V-OBLA (r= 0.81; N = 9).

Figure 5
Figure 5:
The relationship between 8-km performance velocity and the estimated running velocity at V˙O2max(V-V˙O2max).
Figure 6
Figure 6:
The relationship between 8-km performance velocity and the running velocity at the maximal lactate steady state.


The relationship of V-[Lac]BMIN to V-MLSS and other measures of blood lactate accumulation during submaximal exercise. The V-[Lac]BMIN did not differ significantly from V-Tlac or V-Tvent. It was also of interest that HR and [Lac]B at [Lac]BMIN did not differ significantly from HR and [Lac]B at MLSS. However, V-[Lac]BMIN was significantly lower than V-MLSS (Fig. 3). There are two possible explanations for this discrepancy. First, the lengthy nature of V-MLSS determination and the random order of the tests meant that, in some cases, as much as 2-3 wk elapsed between V-MLSS and V-[Lac]BMIN determination. Given this, it is possible that changes in aerobic fitness occurred over the course of the study. However, we consider this possibility unlikely given the well-trained nature of the subject group and the fact that the test administration interfered minimally with their training program. Second, the criterion used for determination of V-MLSS (no more than a 1.0 mM increase in [Lac]B between minutes 10 and 30 of the constant-velocity run) may have been too lenient and could have led to an over-estimation of V-MLSS. In actuality, the running velocity that could be defined as eliciting MLSS produced an increase in [Lac]B between minutes 10 and 30 of the run, which was much less than 1.0 mM in several cases (Fig. 2a). However, the use of a more stringent criterion, such as an increase in [Lac]B of less than 0.5 mM between minutes 10 and 30 of the constant-velocity runs, might have reduced the V-MLSS slightly and affected the conclusions drawn from the results.

Although the V-MLSS is considered the "gold standard" or criterion measure for the assessment of endurance capacity, the V-MLSS has rarely been assessed in a truly independent fashion. More frequently, the physiological responses to constant-load exercise relative to predetermined "threshold" values have been investigated (1,20,32,33,37,43,47,52,59). Even when MLSS has been measured independently and in randomized order, it has been defined in different ways by different authors and has been measured using different numbers and/or different durations of constant-load exercise bouts. For example, Haverty et al. (24) determined V-MLSS as the highest running velocity at which blood lactate did not rise by more than 0.2 mM between minute 10 and minute 20 of a series of 20-min constant-velocity tests. However, a 20-min test is unlikely to be sufficiently long to ascertain the existence of a steady state in blood lactate (1,34). Heck et al. (25) originally defined V-MLSS as the maximal running velocity that produced an increase in blood lactate of less than 1.0 mM during the last 20 min of a 25-min run preceded by 3 min of warm-up. More recently, Beneke and von Duvillard (6), in rowers, cyclists, and triathletes, and Snyder et al. (49), in runners and cyclists, used the identical definition of MLSS as used in the present study, i.e., the maximal exercise intensity producing <1.0 mM increase in [Lac]B between minutes 10 and 30 of the constant-load tests. This criterion for MLSS determination would seem to offer a reasonable compromise given the small error inherent in capillary blood sampling and assay and the changes in muscle substrate and plasma volume that might be expected with exercise of this duration and intensity. We propose that this definition of MLSS should be adopted in future work in this area.

Practically, the primary advantage to the [Lac]BMIN test is the ease and objectivity with which V-[Lac]BMIN can be determined. The minimum point on a U-shaped curve can be determined with great precision, and identical determinations of V-[Lac]BMIN were made by the reviewers for all subjects. A potential disadvantage to the [Lac]BMIN test is the requirement for bouts of supramaximal exercise before the incremental portion of the test. Risk of accident or injury can be reduced by setting the treadmill gradient to 5-6% and reducing the running velocity. In general, subjects found the sprint exercises to be acceptable following a thorough warm-up, although some did complain of muscle soreness on subsequent days.

The [Lac]BMIN test is purported to identify the exercise intensity at which a balance exists between the rate of entry of lactate into the blood and the rate of lactate removal from the blood (i.e., the V-MLSS). It is considered that the rate of lactate clearance is greater than the rate of entry of lactate into the blood before the nadir on the[Lac]B-running velocity curve and that the rate of lactate entry into the blood exceeds the capacity for its clearance beyond this point(Fig. 1)(52). For this to be valid, it must be assumed that the measured [Lac]B at any exercise intensity reflects the metabolic demand of that exercise intensity. However, if blood lactate clearance from the sprint exercises exhibits an element of time dependency before and during the incremental running phase of the [Lac]BMIN test, then the measured [Lac]B will depend both on the metabolic demand of the previous exercise intensity and the overall blood lactate recovery kinetics. A number of factors may then be envisaged to affect V-[Lac]BMIN. These include the duration and intensity of the initial sprints, the duration and intensity of the recovery period following the sprints, the initial running velocity selected for the incremental test, and the stage duration and the incremental rate utilized in the incremental test. Given these methodological and conceptual concerns, it is possible that the close relationship observed in the present study between V-[Lac]BMIN and other established measures of blood lactate accumulation such as V-Tlac was fortuitous. It should be stressed that the V-[Lac]BMIN produced a weaker correlation with V-MLSS (r = 0.61) and a worse estimate of V-MLSS (SEE = 0.75 km·h−1) than did V-Tlac, V-Tvent, or V-OBLA. It was also noticeable that the running velocity at [Lac]BMIN did not vary greatly(Fig. 3), suggesting that the [Lac]BMIN test has poor discriminatory power between subjects. Clearly, until the [Lac]BMIN test protocol and the assumptions underpinning it are validated, estimation of the V-MLSS by more established methods is to be preferred.

The lengthy nature of V-MLSS determination creates a need for this variable to be reliably estimated from the response of the athlete to a single incremental exercise test. In the present study, V-Tlac and V-Tvent were slightly but not significantly lower than V-MLSS, and both these measures provided close approximations of V-MLSS (SEE for V-Tlac = 0.33 km·h−1 and SEE for V-Tvent = 0.61 km·h−1). Further, prolonged running at the velocities corresponding to Tlac and Tvent always produced a steady state in blood lactate and V˙O2. In nine out of ten subjects, the V-MLSS was the same as, or was just 0.5 km·h−1 higher than, V-Tlac (Table 1, Fig. 4). Yamamoto et al. (59), reported that a 30-min cycle exercise at the power output at Tvent resulted in stable[Lac]B of around 3 mM, whereas a 30-min cycle exercise at 104.9% of the power output at Tvent resulted in a significant increase in the measured [Lac]B between minutes 15 and 30. Aunola and Rusko (1) and Ribeiro et al. (43) both reported that steady state values of blood lactate could be attained during prolonged constant-load work at power outputs up to and including, but not above, the power output at Tlac. The results of the present study extend the findings of previous researchers (1,43,59) and indicate that V-Tvent and V-Tlac provide reasonable working estimates of V-MLSS. In situations in which V-MLSS must be directly determined, our results indicate that if V-Tlac is known, then it should be possible to verify V-MLSS from only two constant-velocity runs, performed at 0.5 km·h−1 and 1.0 km·h−1 above V-Tlac. Our study also reinforces earlier work that demonstrated coincidence between Tlac and Tvent(14,54).

Several groups have criticized the accuracy of subjective assessment of data plots (21,60), whereas others have argued that the blood lactate response to incremental exercise is continuous (15,27), with the implication that no threshold phenomenon exists. In the present study, the reviewers found that V-Tlac was impossible to determine in two subjects because a distinct disproportionate increase in blood lactate was not obvious, and threshold selection remained arbitrary following log-log data transformation (4). However, in the other 11 subjects, very clear points of breakaway in blood lactate at 1.8-3.5 mM were observed. Similarly, the two reviewers of our data were able to identify V-Tvent in all subjects with high reliability coefficients both within(r = 0.95) and between (r = 0.95) reviewers. These results support previous studies that have demonstrated high reliability in subjective data interpretation (8,14). They may also indicate that problems in the subjective identification of "thresholds" in lactate and ventilation are minimized with the use of a multistage treadmill protocol with small increments in running velocity, as in the present study and in the study of Weltman et al. (55).

Greater objectivity in assessing the extent of submaximal blood lactate accumulation during an incremental exercise test is afforded by calculation of the point on the [Lac]B curve that gives the maximal distance to a line joining the two end-points of the curve (Dmax method) (9) or, more commonly, by interpolation of the work intensity corresponding to a certain absolute [Lac]B such as 4 mM (V-OBLA) (25). However, measures that rely upon absolute reference values of blood lactate are known to be sensitive to conditions of glycogen depletion and supercompensation, and they do not reflect inter-individual differences in the kinetics of blood lactate accumulation (22,26). In the present study, the [Lac]B at MLSS was 4.35± 0.4 mM (range = 3.0-6.1 mM). Although V-OBLA determined from the incremental test was highly correlated with V-MLSS (r = 0.93), V-OBLA was higher than V-MLSS for all subjects (mean difference = 0.8 km·h−1; range = 0.3-1.3 km·h−1). Exercise above V-MLSS will lead to markedly different metabolic and perceptual responses compared with exercise below V-MLSS (30,39,47), and time to fatigue is difficult to estimate (1,37). These considerations lead us to reject the use of V-OBLA in the prescription of V-MLSS exercise training. However, the close relationship between V-OBLA and the 8-km performance trial suggests that V-OBLA may have value in the prediction of performance capability in events of this duration and in providing a high intensity training stress.

The V-OBLA could not be determined in four subjects. In three subjects, [Lac]B did not reach 4 mM despite running at velocities requiring approximately 90% V˙O2max, whereas in one subject [Lac]B was never less than 4 mM even at running velocities requiring only 60-70% V˙O2max. Davies and Young (12) have reported that individuals with high proportions of type I muscle fibers may not reach 4 mM blood lactate during incremental arm work. Similar considerations have led Borch et al. (7) to recommend 3 mM as a more appropriate reference level of [Lac]B for use with subjects of greater aerobic fitness. However, this approach does not circumvent other problems associated with the use of absolute [Lac]B in the evaluation of endurance capacity in this subject group. It should be pointed out that our use of whole blood as opposed to plasma (frequently used in the early exercise literature) in the blood lactate assay would have increased the exercise intensity associated with V-OBLA (19). However, most sports physiology laboratories now utilize arterialized (capillary) whole blood lactate assays in the physiological assessment of elite athletes (53) so that assay of this medium is appropriate in the context of the present study. It is well established that the pattern of change of lactate during exercise and recovery is similar in whole blood and plasma so that selection of "threshold" criteria is unlikely to differ significantly between these media (44,57).

In summary, the V-[Lac]BMIN was significantly different from V-MLSS and was less closely related to V-MLSS than the other measures studied. The V-[Lac]BMIN had poor discriminatory power between subjects. Therefore, despite methodologic advantages to the [Lac]BMIN test including the speed and objectivity of assessment, the validity of the test must be questioned both on conceptual (concerns noted above) and practical grounds. Of the measures studied, it is proposed that V-Tlac is the best method for the estimation of V-MLSS. The V-Tlac was not significantly different from V-MLSS, and V-Tlac and V-MLSS were closely correlated (Fig. 4). To support the identification of V-Tlac, it is recommended that V-Tvent be assessed concurrently. The V-Tvent was not significantly different from V-Tlac and could be determined in all subjects tested using the V-slope assessment method (5).

Physiological correlates to 8-km running performance. Despite the relative homogeneity of the subject group (cv. 8-km race velocity = 7.2%), there were a number of physiological variables that were closely related to 8-km running performance. The V-V˙O2max was more closely related to V-8 km than was V˙O2maxper se. However, this should not be considered surprising given that V-V˙O2max represents a composite of both the V˙O2max and the running economy characteristics of subjects(35). These results suggest that 8-km running performance requires a great aerobic energy contribution and that subjects ran at a similar percentage(x = 89 ± 2%) of their V-V˙O2max for the 8-km distance(Fig. 5). The V-8 km was significantly higher than V-MLSS, indicating that 8-km race performance involves some anaerobic contribution to ATP resynthesis and the progressive accumulation of acid products of metabolism.

The strongest correlations with performance were found with V-V˙O2max (r = 0.93) and V-MLSS (r = 0.92), and these physiological variables were closely correlated with one another (r = 0.94). This supports the conclusion of Morgan et al. (35) that V-V˙O2max is a powerful predictor of endurance running performance, which can be a useful noninvasive measure of endurance capacity when blood lactate assay is undesirable or unfeasible. The assessment of [Lac]B over a range of submaximal exercise intensities may be of most value in the monitoring of changes in fitness status brought about by endurance training. In this respect, the relationship of the lactate variable to performance may be of less importance than confidence in the reliability of the measure.

Our results do not concur with the recommendation of Noakes et al. (36) that Vmax achieved in a fast incremental treadmill test be routinely measured in the physiological assessment of runners. The specialist marathon and ultramarathon runners in the study of Noakes et al. (36) were of relatively heterogeneous fitness(V˙O2max range = 54-84 mL·kg−1·min−1), and this may have been a factor in the strong correlations obtained between Vmax and a number of performance measures. In our subjects of lower but more homogeneous fitness, Vmax was less strongly correlated with 8-km performance than were a number of other laboratory measures. Another factor in the discrepancy between the present study and that of Noakes et al. (36) may be the ability of the subjects to run with skill at high treadmill velocities. Clearly, training status (i.e., elite vs nonelite) should influence the choice of tests used in the physiological evaluation of runners.

All physiological measures related to the accumulation of blood lactate during submaximal running, including V-[Lac]BMIN, were closely related to 8-km performance. This is in line with an extensive literature that indicates a strong association between the rate of blood lactate accumulation and endurance exercise performance (40,48,50,61). Interestingly, subjects were found to utilize very similar percentages of V˙O2max at each of the measures relating to blood lactate accumulation during submaximal exercise, e.g., for V-Tlac (80.3 ± 0.9% V˙O2max). The V-MLSS showed a particularly strong association with 8-km performance (Fig. 6), underlining the importance of V-MLSS assessment or estimation in the physiologic profiling of endurance athletes (41).

It was of great interest that for prolonged moderate to heavy intensity treadmill running above V-MLSS, neither V˙O2(Fig. 2b) nor V˙E demonstrated steady-state characteristics in any subject. Although the V˙O2 "slow component" is well established in the exercise literature for cycle ergometry (2,39,45), to our knowledge, this paper is the first to demonstrate the existence of a slow component in V˙O2 for treadmill running. Although the mechanism(s) underpinning the slow component remain incompletely understood (38,56), Barstow et al. (3) have recently demonstrated that the relative magnitude of the slow component during heavy cycle exercise is significantly positively related to the proportion of type II fibers in the vastus lateralis. This suggests that the contraction of the type II fibers, which are operative in this work rate domain and which are less efficient (i.e., lower P:O ratio) than the type I fiber (58), is of some importance in the mediation of the slow component phenomenon.

In conclusion, the results of this study indicate that 8-km running performance is closely related to a number of physiological variables, most notably V-V˙O2max, V-MLSS, and V-Tlac. The best estimate of the V-MLSS was provided by V-Tlac (r = 0.94; SEE = 0.33 km·h−1). The V-[Lac]BMIN of Tegtbur et al. (52) was less closely related to V-MLSS (r = 0.61; SEE= 0.75 km·h−1) and had poor discriminatory power between subjects. Advocacy of the [Lac]BMIN test must await a full experimental critique of the test protocol and the completion of thorough investigations into possible time-dependent components of lactate production and clearance.


1. Aunola, S., and H. Rusko. Does anaerobic threshold correlate with maximal lactate steady-state? J. Sports Sci. 10:309-323, 1992.
2. Barstow, T. J., and P. A. Mole. Linear and nonlinear characteristics of oxygen uptake kinetics during heavy exercise. J. Appl. Physiol. 71(6):2099-2106, 1991.
3. Barstow, T. J., A. M. Jones, P. Nguyen, and R. Casaburi. Influence of muscle fiber type and pedal frequency on oxygen uptake kinetics of heavy exercise. J. Appl. Physiol. 81(4):1642-1650, 1996.
4. Beaver, W. L., K. Wasserman, and B. J. Whipp. Improved detection of lactate threshold during exercise as a log-log transformation. J. Appl. Physiol. 59:1936-1940, 1985.
5. Beaver, W. L., K. Wasserman, and B. J. Whipp. A new method for detecting anaerobic threshold by gas exchange. J. Appl. Physiol. 60:2020-2027, 1986.
6.Beneke, R., and S. P. Von Duvillard. Determination of maximal lactate steady state response in selected sports events. Med. Sci. Sports Exerc. 28(2):241-246, 1996.
7. Borch, K. W., F. Ingjer, S. Larsen, and S. E. Tomten. Rate of accumulation of blood lactate during graded exercise as a predictor of "anaerobic threshold." J. Sports Sci. 11:49-55, 1993.
8. Caiozzo, V. J., J. A. Davis, J. F. Ellis, J. L. Azus, R. Vandagriff, C. A. Prietto, and W. C. Mcmaster. A comparison of gas exchange indices used to detect the anaerobic threshold. J. Appl. Physiol. 53(5):1184-1189, 1982.
9. Cheng, B., H. Kuipers, A. C. Snyder, H. A. Keizer, A. Jeukendrup, and M. Hesselink. A new approach for the determination of ventilatory and lactate thresholds. Int. J. Sports Med. 13(7):518-522, 1992.
10.Conley, D. L., and G. S. Krahenbuhl. Running economy and distance running performance of highly trained athletes. Med. Sci. Sports Exerc. 12(5):357-360, 1980.
11. Costill, D. L., H. Thomason, and E. Roberts. Fractional utilisation of the aerobic capacity during distance running. Med. Sci. Sports 5(4):248-252, 1973.
12. Davies, C. T. M., and K. Young. Effect of temperature on the contractile properties and muscle power of the triceps surae in humans. J. Appl. Physiol. 55:191-195, 1983.
13.Davis, J. A., M. H. Frank, B. J. Whipp, and K. Wasserman. Anaerobic threshold alterations caused by endurance training in middle-aged men. J. Appl. Physiol.46(6):1039-1046, 1979.
14. Davis, J. A., P. Vodak, J. H. Wilmore, J. Vodak, and P. Kurtz. Anaerobic threshold and maximal aerobic power for three modes of exercise. J. Appl. Physiol. 41(4):544-550, 1976.
15.Dennis, S. C., T. D. Noakes, and A. N. Bosch. Ventilation and blood lactate increase exponentially during incremental exercise. J. Sports Sci. 10:437-449, 1992.
16. Di Prampero, P. E. Energetics of muscular exercise. Rev. Physiol. Biochem. Pharmacol. 89:143-222, 1981.
17. Durnin, J. V. G. A., and J. Womersley. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br. J. Nutr. 32:77-97, 1974.
18. Foster, C. VO2max and training indices as determinants of competitive running performance. J. Sports Sci. 1:13-22, 1983.
19. Foxdal, P., A. Sjodin, B. Ostman, and B. Sjodin. The effect of different blood sampling sites and analyses on the relationship between exercise intensity and 4 mmol l−1 blood lactate concentration. Eur. J. Appl. Physiol. 63:52-54, 1991.
20. Foxdal, P., A. Sjodin, and B. Sjodin. Comparison of blood lactate concentrations obtained during incremental and constant intensity exercise. Int. J. Sports Med. 17(5):360-365, 1996.
21. Gladden, L. B., J. W. Yates, R. W. Stremel, and B. A. Stamford. Gas exchange and lactate anaerobic thresholds: inter- and intra-evaluator agreement. J. Appl. Physiol. 58(6):2082-2089, 1985.
22.Gollnick, P. D., W. M. Bayly, and D. R. Hodgson. Exercise intensity, training, diet, and lactate concentration in muscle and blood. Med. Sci. Sports Exerc. 18:334-340, 1986.
23. Hansen, J. E., R. Casaburi, D. M. Cooper, and K. Wasserman. Oxygen uptake as related to work rate increment during cycle ergometer exercise. Eur. J. Appl. Physiol. 57:140-145, 1988.
24. Haverty, M., W. L. Kenney, and J. L. Hodgson. Lactate and gas exchange responses to incremental and steady state running. Br. J. Sports Med. 22(2):51-54, 1988.
25. Heck, H., A. Mader, G. Hess, S. Mucke, S. Muller, and W. Hollmann. Justification of the 4 mmol/l lactate threshold. Int. J. Sports Med. 6:117-130, 1985.
26.Hughson, R. L., and H. J. Green. Blood acid-base and lactate relationships studied by ramp work tests. Med. Sci. Sports Exerc. 14(4):297-302, 1982.
27. Hughson, R. L., K. H. Weisiger, and G. D. Swanson. Blood lactate concentration increases as a continuous function in progressive exercise. J. Appl. Physiol. 62(5):1975-1981, 1987.
28. Jones, A. M., and J. H. Doust. A comparison of three treadmill protocols for the determination of maximal aerobic power in runners. J. Sports Sci. 14:89, 1996.
29. Jones, A. M., and J. H. Doust. A 1% treadmill grade most accurately reflects the energetic cost of outdoor running. J. Sports Sci. 14:321-327, 1996.
30. Katch, V., A. Weltman, S. Sady, and P. Freedson. Validity of the relative percent concept for equating training intensity. Eur. J. Appl. Physiol. 39:219-227, 1978.
31. Kindermann, W., G. Simon, and J. Keul. The significance of the aerobic-anaerobic transition for the determination of work load intensities during endurance training. Eur. J. Appl. Physiol. 42:25-34, 1979.
32. Mclellan, T. M., and G. C. Gass. Metabolic and cardiorespiratory responses relative to the anaerobic threshold. Med. Sci. Sports Exerc. 21(2):191-198, 1989.
33. Mclellan, T. M., and J. S. Skinner. Submaximal endurance performance relative to the ventilation thresholds. Can. J. Appl. Sport Sci. 10(2):81-87, 1985.
34. Mognoni, P., M. D. Sirtori, F. Lorenzelli, and P. Cerretelli. Physiological responses during prolonged exercise at the power output corresponding to the blood lactate threshold. Eur. J. Appl. Physiol. 60:239-243, 1990.
35. Morgan, D. W., F. D. Baldini, P. E. Martin, and W. M. Kohrt. Ten kilometer performance and predicted velocity at VO2max among well-trained male runners Med. Sci. Sports Exerc. 21(1):78-83, 1989.
36. Noakes, T. D., K. H. Myburgh, and R. Schall. Peak treadmill velocity during the VO2max test predicts running performance J. Sports Sci. 8:35-45, 1990.
37.Orok, C. J., R. L. Hughson, H. J. Green, and J. A. Thomson. Blood lactate responses in incremental exercise as predictors of constant load performance. Eur. J. Appl. Physiol. 59:262-267,1989.
38. Poole, D. C. Role of exercising muscle in slow component of VO2. Med. Sci. Sports Exerc. 26(11):1335-1340, 1994.
39. Poole, D. C., S. A. Ward, G. W. Gardner, and B. J. Whipp. Metabolic and respiratory profile of the upper limit for prolonged exercise in man. Ergonomics 31(9):1265-1279, 1988.
40. Powers, S. K., S. Dodd, R. Deason, R. Byrd, and T. Mcknight. Ventilatory threshold, running economy and distance running performance of trained athletes. Res. Q. Sports Exerc. 54(2):179-182, 1983.
41. Priest, J. W., and R. D. Hagan. The effects of maximum steady state pace training on running performance. Br. J. Sports Med. 21(1):18-21, 1987.
42. Rhodes, E. C., and D. C. Mckenzie. Predicting marathon time from anaerobic threshold measurements. Physician Sports Med. 12(1):95-98, 1984.
43. Ribeiro, J. P., V. Hughes, R. A. Fielding, W. Holden, W. Evans, and H. G. Knuttgen. Metabolic and ventilatory responses to steady state exercise relative to lactate thresholds. Eur. J. Appl. Physiol. 55:215-221, 1986.
44. Robergs, R. A., J. Chwalbinska-Moneta, J. B. Mitchell, D. D. Pascoe, J. Houmard, and D. Costill. Blood lactate threshold differences between arterialised and venous blood. Int. J. Sports Med. 11(6):446-451, 1990.
45. Roston, W. L., B. J. Whipp, J. A. Davis, R. M. Effros, and K. Wasserman. Oxygen uptake kinetics and lactate concentration during exercise in humans. Am. Rev. Respir. Dis. 135:1080-1084, 1987.
46. Saltin, B., and P-O. Astrand. Maximal oxygen uptake in athletes.J. Appl. Physiol. 23:353-358, 1967.
47. Simon, J., J. L. Young, B. Gutin, D. K. Blood, and R. B. Case. Lactate accumulation relative to the anaerobic and respiratory compensation thresholds. J. Appl. Physiol. 54(1):13-17, 1983.
48. Sjodin, B., and I. Jacobs. Onset of blood lactate accumulation and marathon running performance. Int. J. Sports Med. 2:23-26, 1981.
49. Snyder, A. C., T. Woulfe, R. Welsh, and C. Foster. A simplified approach to estimating the maximal lactate steady state. Int. J. Sports Med. 15(1):27-31, 1994.
50. Tanaka, K., Y. Matsuura, S. Kumagai, A. Matsuzaka, and K. Hirakoba. Relationship of anaerobic threshold and onset of blood lactate accumulation with endurance performance. Eur. J. Appl. Physiol. 52:51-56, 1983.
51. Tanaka, K., Y. Matsuura, A. Matsuzaka, K. Hirakoba, S. Kumagai, S. O. Sun, and K. Asano. A longitudinal assessment of anaerobic threshold and distance-running performance. Med. Sci. Sports Exerc. 16(3):278-282, 1984.
52. Tegtbur, U., M. W. Busse, and K. M. Braumann. Estimation of an individual equilibrium between lactate production and catabolism during exercise.Med. Sci. Sports Exerc. 25(5):620-627, 1993.
53.Thoden, J. S. Testing aerobic power. In: Physiological Testing of the High-Performance Athlete, 2nd Ed. J. D. MacDougall, H. A. Wenger, and H. J. Green(Eds.). Champaign, IL: Human Kinetics, 1991, pp. 136-139.
54.Wasserman, K., B. J. Whipp, S. N. Koyal, and W. L. Beaver. Anaerobic threshold and respiratory gas exchange during exercise. J. Appl. Physiol. 35:236-243, 1973.
55. Weltman, A., D. Snead, P. Steim, R. Seip, R. Schurrer, R. Rutt, and J. Weltman. Reliability and validity of a continuous incremental treadmill protocol for the determination of lactate threshold, fixed blood lactate concentrations and VO2max. Int. J. Sports Med. 11(1):26-32, 1990.
56. Whipp, B. J. The slow component of O2 uptake kinetics during heavy exercise Med. Sci. Sports Exerc. 26(11):1319-1326, 1994.
57. Williams, J. R., N. Armstrong, and B. J. Kirby. The influence of the site of sampling and assay medium upon the measurement and interpretation of blood lactate responses to exercise. J. Sports Sci. 10:95-107, 1992.
58. Willis, W. T., and M. R. Jackman. Mitochondrial function during heavy exercise. Med. Sci. Sports Exerc. 26(11):1347-1354, 1994.
59. Yamamoto, Y., M. Miyashita, R. L. Hughson, S-I. Tamura, M. Shinohara, and Y. Mutoh. The ventilatory threshold gives maximal lactate steady state. Eur. J. Appl. Physiol. 63:55-59, 1991.
60. Yeh, M. P., R. M. Gardner, T. D. Adams, F. G. Yanowitz, and R. O. Crapo. "Anaerobic threshold": problems of determination and validation. J. Appl. Physiol. 55(4):1178-1186, 1983.
61. Yoshida, T., M. Chida, M. Ichioka, and Y. Suda. Blood lactate parameters related to aerobic capacity and endurance performance.Eur. J. Appl. Physiol. 56:7-11, 1987.


© Williams & Wilkins 1998. All Rights Reserved.