Alternative methods for estimating maximum lactate steady state velocity in physically active young adults

The aim of this study was to compare the velocities found in the protocols used to measure the indirect individual anaerobic threshold (IATind), glucose threshold (GT) and critical velocity (CV) with the gold standard, the maximum lactate steady state (MLSS) protocol. Fourteen physically active young adults (23±3.1 years; 72±10.97 kg; 176±7 cm; 21±5.36% body fat) performed a 3000-m track running test to determine IATind using the prediction equation and an incremental test on a treadmill to determine GT. The CV was identified by linear regression of the distance-time relationship based on 3000-m and 500-m running performance. The MLSS was identified using two to five tests on different days to identify the intensity at which there was no increase in blood lactate concentration greater than 1 mmol/L between the 10th and 30th minute. A significant difference was observed between mean CV and MLSS (P≤0.05) and there was a high correlation between MLSS and IATind (R 2=0.82; P≤0.01) and between MLSS and GT (R2=0.72; P≤0.01). The Bland-Altman method showed agreement between MLSS and IATind [mean difference -0.24 (confidence interval -1.72 to 1.24) km/h] and between MLSS and GT [0.21 (-1.26 to 1.29) km/h]. We conclude that the IATind and GT can predict MLSS velocity with good accuracy, thus making the identification of MLSS practical and efficient to prescribe adequate intensities of aerobic exercise.


INTRODUCTION
The maximum lactate steady state (MLSS) is defined as the exercise intensity at which a dynamic metabolic balance exists between maximum lactate production and maximum lactate removal.The corresponding workload can be maintained for a prolonged period of time in the absence of continuous accumulation of lactate in the bloodstream.The velocity corresponding to the MLSS is a key factor for the evaluation of endurance training and provides positive results when its intensity is used for training prescription in physically active individuals 1,2 .
One of the methods used to determine MLSS consists of performing 30-minute exercise tests on different days at intensities ranging from 50-90% VO 2 max 3 .However, other authors have used less invasive methods for the determination of MLSS that do not require blood collection or only a small number of blood samples 4 .Protocols that are less time consuming and that optimize the prediction of MLSS use alternatives such as the glucose threshold (GT) 5 , critical velocity (CV) 6 , and indirect individual anaerobic threshold (IAT ind ) 7 .In addition to their relevant cost-benefit relationship, these protocols are practical because they can be applied on ergometers, require little space, and can be used in different populations [8][9][10] .Several studies have proposed alternative methods for estimating MLSS; however, the results are conflicting and few studies involve non-athletes.
The aim of the present study was to compare the velocities obtained with different protocols (GL, CV and IAT ind ) with the MLSS velocity in non-athletes.

METHODOLOGICAL PROCEDURES
This was a cross-sectional, prospective and quantitative study.After anthropometric assessment, all participants performed three tests: 1) two track tests to determine CV; 2) one test on a treadmill to determine GT velocity, and 3) one test on a treadmill to determine MLSS velocity.All tests were performed at an interval of 48 to 72 hours.

Participants
Fourteen healthy men without any physical or clinical exercise restriction volunteered to participate in the study.The participants were non-athlete, university students recruited on the campus of a university in Bauru, São Paulo, Brazil (23 ± 3.1 years; 72 ± 10.97 kg; 1.76 ± 0.07 m; 21 ± 5.36% body fat).The procedures, risks and benefits of the study were explained to the participants before they signed the free informed consent form.The experimental procedures were approved by the local Ethics Committee (Permit No. 32/2008).

Study design
On the first visit to the laboratory, the height, body weight and skinfolds of the participants were measured.Body density was estimated using the 3-site skinfold protocol proposed by Pollock 11 and the equation of Siri 12 for assessing body composition.All subjects were asked not to perform any strenuous physical activity 48 hours prior to the tests.
The sample size was estimated using the table proposed by Hulley 13 , considering an alpha value (two-sided) of 0.05, beta error of 0.05, and expected correlation coefficient of r=0.80.
The experiment consisted of two initial tests on a running track to determine CV, one test on a treadmill to determine GT, and another test to determine MLSS.The IAT ind was calculated using the following formula: where V3000 is the mean velocity (m/min) achieved in the initial 3000-m test.
The tests were performed on a treadmill (Movement LX-150) and on a running track with charcoal flooring and 100-m marks.The subjects were asked to have their meals 3 hours before the tests.

Test 1
A linear model of the distance-time relationship obtained based on the 3000-m and 500-m running performance on the track was used to determine CV.The slope of the linear regression line defined the value of CV 14 .

Test 2
For the determination of GT, the subject was submitted to an incremental test on a treadmill (Movement LX-150) at an inclination of 1% 15 and initial velocity of 65% of V3000, followed by increments of 0.5 km/h at intervals of 3 minutes in each stage.There was an interval of 30 seconds in each stage for blood collection from the ear lobe.The GT was defined as the velocity when glucose levels were minimal during the test 8 .Blood glucose was measured with the OneTouch Ultra-2 ® glucose meter (Johnson & Johnson ® ) and is expressed as mg/dL.

Test 3
Two to five 30-minute tests were performed for the determination of MLSS.In the first test, a velocity that was 5% below the GT obtained in the previous test was used.Subsequently, the velocity was increased by 5% in each test until the increase in lactate was higher than 1 mmol/l between the 10 th and 30 th minute, considering the previous velocity as the corresponding MLSS 16 .The tests were separated by resting intervals of 48-72 hours.

Blood collection and analysis
Blood samples were collected with a disposable lancet by puncture of the ear lobe, previously disinfected with alcohol, using disposable gloves.The blood was collected into heparinized capillary tubes calibrated to contain 25 µL arterial blood and the tubes were transferred to Eppendorf tubes containing 50 µL 1% sodium fluoride.Before the collection of each sample, the site was cleaned to prevent contamination with sweat or other materials that would make the blood samples unusable.All samples were stored in a freezer for subsequent analysis.Blood lactate concentrations were measured with a lactate analyzer based on an electroenzymatic method (YSI 1500 Sports, Yellow Springs Instruments, OH, USA) and are expressed as mmol/L.

Statistical analysis
First, the data were submitted to descriptive analysis (mean ± standard deviation).Analysis of variance for repeated measures was used to determine possible differences in IAT ind , GT, CV and MLSS velocity.Correlations between IAT ind , GT, CV and MLSS were estimated using Pearson's correlation test.A level of significance of P≤0.05 was adopted.Bland-Altman analysis was used to evaluate the agreement between methods.Statistical analysis was performed using the Statistical Package for the Social Sciences 17.0 (SPSS) and Bland-Altman plots were constructed using the Medcalc 12.3.0.0 program (both for Windows).

RESULTS
The overall results of the parameters studied are shown in Table 1.A significant difference was observed between mean CV and MLSS (P≤0.05).
Despite the high correlation between all methods used to estimate MLSS, Bland-Altman analysis revealed agreement only between MLSS and IAT ind [mean difference -0.24 (confidence interval -1.72 to 1.24) km/h] and MLSS and GT [0.21 (-1.26 to 1.29) km/h] (Figure 1).The mean (± standard deviation) lactate concentration in the MLSS test was 5 ± 1.8 mmol/L.

DISCUSSION
Conventional statistical approaches have shown the absence of significant differences between the MLSS, IAT ind and GT methods.This suggests that the IAT ind and GT methods can be used to estimate MLSS.In contrast, CV was not a good predictor of MLSS, as indicated by the significant difference between MLSS and CV, overestimating MLSS despite the observation of a good correlation.
During high-intensity activities, lactate accumulation is due to production being higher than its removal 17 .The MLSS is a measure of the exercise intensity at which a dynamic balance exists between maximum lactate production and maximum lactate removal 3,17 .It can be defined as the highest lactate concentration in blood and its respective workload can be maintained over time without continuous blood lactate accumulation 1,3,18 .The MLSS is an individual indicator of exercise intensity 19 and is therefore important for the prescription, monitoring and evolution of aerobic exercise intensity 1 .However, the protocol used to identify the MLSS requires that the participants are available for 2 to 6 constant workload tests at different intensities on different days, with each test lasting 30 minutes.This procedure is not feasible for participants enrolled in a training program and researchers often do not have the time or material needed for the tests.Therefore, investigators have tried to develop methods to optimize the estimation of MLSS [20][21][22] .
Simões et al. 23 proposed an indirect test to estimate MLSS in endurance runners based on the linear relationship between mean 3-km running velocity and velocity at the individual anaerobic threshold.The data obtained in the present study suggest that the method proposed by Simões et al. 23 is suitable to estimate the intensity associated with MLSS in physically ac- tive subjects.No significant difference was observed between the velocity at IAT ind and MLSS (Figure 1) and the methods showed a high correlation (r=0.91;R 2 =0.82;P≤0.01) and agreement [-0.24 (-1.72 to 1.24) km/h].These results suggest that MLSS can be estimated safely based on a single 3000-m run on a running track, thus providing a faster and noninvasive procedure that does not requires expenses with materials.
There was also no significant differences between the velocities at GT and MLSS, with the observation of a high correlation (r=0.85;R 2 =0.72;P≤0.01) and agreement [0.21 (-1.26 to 1.69) km/h].In view of the easy execution and low cost of the identification of GT, this test also has great practical applicability.Sotero et al. 8 found a high correlation (r=0.96;P≤0.01) and agreement [1.7 (8.5) m/min] between GT and MLSS, indicating that blood glucose concentration is a good predictor of MLSS in physically active subjects.
Many of the studies estimating MLSS based on the CV have obtained values that overestimated MLSS, a fact indicating that this method is not adequate for the estimation of MLSS 4,24,25 .Other factors that support the low accuracy of CV in estimating MLSS are the fact that the intensity determined for CV may vary depending on the distance covered 26 and that the mathematical model used for the determination of CV shows a variation of 18% 27 .Taken together, the results suggest that the model of CV used in this study is not adequate to estimate MLSS in physically active young male subjects.
One important limitation of this study was the application of the test to estimate CV on a track, considering the technical differences between running on a treadmill and on a track.The alternative protocols used for the determination of MLSS have advantages and disadvantages.According to Azevedo et al. 28 , the protocol applied should taken into consideration the methodology used, experience of the examiner, and population studied.However, the use of methods such as IAT ind and GT has a high practical value, considering the importance of identifying the second threshold for training prescription.

CONCLUSION
The present results permit us to conclude that MLSS can be estimated in young and physically active subjects using IAT ind and GT.Both methods were found to be a good option for the determination and prescription of exercise intensities, particularly aerobic exercise.However, the present results showed that CV is not a valid method to estimate MLSS.

Figure 1 .
Figure 1.Bland-Altman plots of agreement between MLSS and mean IATind, GT and CV velocities.Continuous lines indicate the mean differences and dashed lines indicate the 95% limits of agreement between measures.