Reproducibility of Blood Lactate Concentration Rate under Isokinetic Force Loads

(1) Background: Maximum isokinetic force loads show strongly increased post-load lactate concentrations and an increase in the maximum blood lactate concentration rate (V˙Lamax), depending on load duration. The reproducibility of V˙Lamax must be known to be able to better assess training-related adjustments of anaerobic performance using isokinetic force tests. (2) Methods: 32 subjects were assigned to two groups and completed two unilateral isokinetic force tests (210° s−1, Range of Motion 90°) within seven days. Group 1 (n = 16; age 24.0 ± 2.8 years, BMI 23.5 ± 2.6 kg m−2, training duration: 4.5 ± 2.4 h week−1) completed eight repetitions and group 2 (n = 16; age 23.7 ± 1.9 years, BMI 24.6 ± 2.4 kg m−2, training duration: 5.5 ± 2.1 h week−1) completed 16 repetitions. To determine V˙Lamax, capillary blood (20 µL) was taken before and immediately after loading, and up to the 9th minute post-load. Reproducibility and variability was determined using Pearson and Spearman correlation analyses, and variability were determined using within-subject standard deviation (Sw) and Limits of Agreement (LoA) using Bland Altman plots. (3) Results: The correlation of V˙Lamax in group 1 was r = 0.721, and in group 2 r = 0.677. The Sw of V˙Lamax was 0.04 mmol L−1 s−1 in both groups. In group 1, V˙Lamax showed a systematic bias due to measurement repetition of 0.02 mmol L−1 s−1 in an interval (LoA) of ±0.11 mmol L−1 s−1. In group 2, a systematic bias of −0.008 mmol L−1 s−1 at an interval (LoA) of ±0.11 mmol L−1 s−1 was observed for repeated measurements of V˙Lamax. (4) Conclusions: Based on the existing variability, a reliable calculation of V˙Lamax seems to be possible with both short and longer isokinetic force loads. Changes in V˙Lamax above 0.11 mmol L−1 s−1 due to training can be described as a non-random increase or decrease in V˙Lamax.


Introduction
Maximum anaerobic loads result in increased lactate accumulation and, depending on the load duration (10 s vs. 30 s), in higher concentrations of post-load lactate [1]. Anaerobic performance can be determined using the blood lactate concentration. The anaerobic performance expressed by lactate concentration rate represents the stress of glycolysis [2]. The change in lactate concentration is determined over load duration. To avoid underestimating the rate of lactic acid formation, it is necessary to consider the alactacid time interval of the load duration. Up to now, the efficiency of anaerobic energy supply has been primarily investigated using running and bicycle ergometer loads [1,[3][4][5][6]. This has already shown good to very good reproducibility of maximum power (P max ) [4,5]. Adam et al. [5] found very good reproducibility ofVLa max in anaerobic bicycle sprints over 15 s. When the load duration increases, Heck and Schulz [3] assume a reduction ofVLa max . Empirical evidence of this was provided by Hauser [7].
Isokinetic force tests to determine muscle performance are important in sports practice and research [8]. Performing isokinetic force tests at maximum performance may show the performance of anaerobic metabolism. The strong metabolic stress of the anaerobic energy metabolism at maximum isokinetic loads is characterized by strongly increased post-load lactate concentrations and a resulting maximum lactate concentration rate (VLa max ) [9].VLa max increases with accelerations in movement velocity under isokinetic force loads [10]. Studies on anaerobic performance under isokinetic force loads have shown good to excellent reproducibility of performance and torque [11][12][13][14][15]. Furthermore, Bosquet et al. [16] showed a high correlation of the "total work" between bicycle sprints (Wingate-Test) and isokinetic leg extension movements. Similar results were obtained when comparing a Wingate-Test and a sprint test in running [6,17].
Compared to an isokinetic bicycle sprint, it is possible to make statements about the local performance of muscles using an isokinetic strength test. This can be important for sports with high strength components and dominant anaerobic demands on energy metabolism, such as martial arts, gymnastics, and strength sports, such as CrossFit or weight lifting.
However, due to the lower active muscle content in the isokinetic force test compared to the bicycle sprint and therefore changed lactate invasion and elimination, the absolute glycolysis rates are not directly comparable [9]. A unilateral isokinetic strength test with one leg primarily stresses the leg muscles. A bicycle test also involves parts of the trunk, in addition to both legs. It is therefore questionable whether the results of high reproducibility forVLa max in the bicycle sprint can be transferred to isokinetic force tests. So far, no investigations on the reproducibility ofVLa max under isokinetic force loads have been carried out. Knowledge of the reproducibility and thus the variability ofVLa max in isokinetic force tests is important to explain the changes in local anaerobic performance. This makes it easier to interpret the effects of training intervention (especially local resistance training) on anaerobic performance. SinceVLa max appears to depend on the duration of an anaerobic test, it should be clarified how variability presents itself at different load durations. The aim of the present study was to investigate the reproducibility ofVLa max under isokinetic force loads as a function of load duration (t load ).

Materials and Methods
Thirty two trained male subjects (age 23.9 ± 2.4 years, body height 177.0 ± 8.3 cm, body mass 76.3 ± 11.5 kg, BMI 24.1 ± 0.1 kg m −2 , practiced sports: soccer, endurance, martial arts, strength training) agreed to voluntary participation in the study after receiving oral and written information. The study was approved by the local ethical committee (V-297-17-HS-Lactate-18102018) and conducted in accordance with the Declaration of Helsinki. The subjects were then assigned to two groups and completed two single-legged isokinetic force tests (CON-TREX ® MJ, PHYSIOMED ® , Schnaittach, Germany) with unilateral knee flexion and extension movements (Range of Motion 90 • ) within seven days of each other. The upper body was immobilized with belts and the non-active leg was also immobilized. The subjects were instructed to perform each movement (flexion and extension) at maximum power. The movement velocity was chosen as 210 • s −1 , since this was the highestVLa max was found in pretests. Group  Before each measurement, two capillary blood samples were taken from the earlobe to determine resting lactate (RL). Immediately after loading, capillary blood (20 µL) was drawn every 30 s until the third minute post-loading. Subsequently, capillary blood was taken every minute until the 9th minute post-loading. The calculation (1) ofVLa max was done according to the model by Mader [2] and Heck and Schulz [3]. The maximum blood lactate in the post-loading period (La max ), resting lactate (RL), test duration (t load ), and the alactic time interval (t alac ) were used to determineVLa max [3]. The alactic time interval (Figure 1) is the range from the start of the test to the time at which the maximum power (P max ) has dropped by 3.5% [5,18].
Sports 2018, 6, x FOR PEER REVIEW 3 of 8 Lamax [3]. The alactic time interval (Figure 1) is the range from the start of the test to the time at which the maximum power (Pmax) has dropped by 3.5% [5,18]. The Shapiro-Wilk test was used to check the normal distribution of the variables Lamax, Lamax, tload, RL, Pmax and talac. A paired t-test was performed to test for significant differences in RL and tload between Test 1 and Test 2. For normally distributed data, linear correlations were tested using Pearson's correlation. The Spearman's rho was used for data that showed no normal distribution. Cohen and Manion [19] was used to assess the linear relationship between Test 1 and Test 2. Bland and Altman plots (BA plots) were used to test the dependence of the measurement differences between Test 1 and Test 2 for homoscedasticity [20]. The Limits of Agreement (LoA) were calculated from the mean difference (MWDiff) ± standard deviation * 1.96. Non-normally distributed differences for talac and Pmax were calculated using the non-parametric LoA method (medianDiff) based on the 2.5 (Q0.025) and 97.5 percentiles (Q0.975) [21]. The measurement error as within-subject standard deviation (Sw) was calculated to determine the variability of the variables Lamax, Lamax, Pmax, and talac [22]. The Sw (Sw-norm in %) was normalized by dividing the Sw by the group mean * 100. The significance level was 5%.
In both groups, high correlations between the tests were determined for mean maximum torque, mean maximum power, and maximum power ( Table 1). The physiological quantities Lamax and Lamax of group 1 showed a correlation between both tests of r = 0.688 and r = 0.721. In group 2, there was a correlation of r = 0.821 (Lamax) and r = 0.677 ( Lamax). These can be considered high linear The Shapiro-Wilk test was used to check the normal distribution of the variablesVLa max , La max , t load , RL, P max and t alac . A paired t-test was performed to test for significant differences in RL and t load between Test 1 and Test 2. For normally distributed data, linear correlations were tested using Pearson's correlation. The Spearman's rho was used for data that showed no normal distribution. Cohen and Manion [19] was used to assess the linear relationship between Test 1 and Test 2. Bland and Altman plots (BA plots) were used to test the dependence of the measurement differences between Test 1 and Test 2 for homoscedasticity [20]. The Limits of Agreement (LoA) were calculated from the mean difference (MW Diff ) ± standard deviation * 1.96. Non-normally distributed differences for t alac and P max were calculated using the non-parametric LoA method (median Diff ) based on the 2.5 (Q 0.025 ) and 97.5 percentiles (Q 0.975 ) [21]. The measurement error as within-subject standard deviation (S w ) was calculated to determine the variability of the variables La max ,VLa max , P max , and t alac [22]. The S w (S w-norm in %) was normalized by dividing the S w by the group mean * 100. The significance level was 5%.

T-Tests and Correlations
In group 1, an RL of 0.86 ± 0.22 mmol L −1 was present before Test 1 and 0.91 ± 0.35 mmol L −1 in Test 2 (p > 0.05). Group 2 showed an RL of 0.83 ± 0.41 mmol L −1 before Test 1 and an RL of 0.76 ± 0.18 mmol L −1 (p > 0.05) before Test 2. In both groups, t load did not differ significantly between the two tests (p > 0.05).
In both groups, high correlations between the tests were determined for mean maximum torque, mean maximum power, and maximum power ( Table 1). The physiological quantities La max anḋ VLa max of group 1 showed a correlation between both tests of r = 0.688 and r = 0.721. In group 2, there was a correlation of r = 0.821 (La max ) and r = 0.677 (VLa max ). These can be considered high linear correlations. With regard to t alac , a high correlation of r > 0.665 was observed in group 1 between test 1 and test 2. Group 2 showed a mean linear correlation of r = 0.482 between test 1 and test 2. Table 1. Test-Retest results (mean and standard deviation) of both groups and the correlation coefficient r (p-value).

Groups
Group 1

Variability of Measurement
The S w variability for La max , P max , andVLa max ranged between 5% and 16% ( Table 2). t alac showed S w between 40% and 50%. The variability ofVLa max in both groups was 0.04 mmol L −1 s −1 (S w-norm : 13.47% to 15.78%). Figure 2 shows the systematic biasVLa max quantities using BA plots for both groups. For group 2, a systematic bias on the measurement repetition (MW Diff ) was −0.38 mmol L −1 with an interval (LoA) of ± 1.24 mmol L −1 was found based on the BA plots for La max . The MW Diff of P max was determined as 3.68 watts at an LoA of ± 72.1 watts. The t alac showed an MW Diff of 0.11 s with an LoA of ± 6.70 s. The systematic bias on measurement repetition forVLa max was −0.008 mmol L −1 s −1 at an interval (LoA) of ±0.11 mmol L −1 s −1 .
The Bland and Altman plots did not show a tendency in the measurement differences with respect to the mean value of the variables for either group. Thus homoscedasticity of the measurement differences was assumed.

Discussion
The aim of the study was to investigate the reproducibility of Lamax under isokinetic force loads. For this purpose, single-legged maximum leg flexion and extension movements with different load durations were performed on two groups of test subjects using a test and retest design. With respect to the target variable Lamax, the results of the linear correlation showed a high correlation in the test with eight repetitions and a moderate correlation in the test with 16 repetitions. The variability of the Sw can be considered as low for the investigated physiological values. Adam et al. [5] found similar results for maximum bicycle sprints.

Discussion
The aim of the study was to investigate the reproducibility ofVLa max under isokinetic force loads. For this purpose, single-legged maximum leg flexion and extension movements with different load durations were performed on two groups of test subjects using a test and retest design. With respect to the target variableVLa max , the results of the linear correlation showed a high correlation in the test with eight repetitions and a moderate correlation in the test with 16 repetitions. The variability of the S w can be considered as low for the investigated physiological values. Adam et al. [5] found similar results for maximum bicycle sprints.
VLa max is estimated on the basis of values, which are, in part strongly dependent on coordinative, motivational, and metabolic factors. This physiological size can potentially be affected by numerous determinants. Maximum lactate and duration of the t alac is crucial with comparable load time and preload lactate concentration. In this context, t alac showed the highest variability compared to the other values determiningVLa max . The data suggest that the variability between subjects is similar to that between repeated measurements.
In terms of t alac , it would be difficult to distinguish between intraindividual changes and interindividual characteristics. t alac is the time interval at which the maximum power has been reduced by 3.5%. If the power maximum occurs too late or the power drop can be delayed, t alac increases and consequentlyVLa max increases. For approx. 10 s of running load, Heck and Schulz [3] assume a t alac of approx. four seconds, which increases with increasing running distance, and thus depends on the running strategy. At maximum isokinetic force loads, Nitzsche et al. [9] calculated a t alac of approx. three seconds. From a physiological point of view, t alac represents the time interval which is secured by "high-energy substrates" creatine phosphate (PCr) and adenosine triphosphate (ATP) to generate or maintain a maximum performance. At loads that provide maximum performance over a short time, PCr and ATP (5 to 3.5 mmol L −1 ) drop rapidly, which leads to a drop in performance due to the lower flow rate of phosphates via anaerobic glycolysis [3,23]. Therefore, P max cannot physiologically occur if t alac is delayed. The present study showed lower t alac at eight repetitions and higher t alac at 16 repetitions. Between Test 1 and Test 2, group 1 (8 reps) had a higher linear correlation than group 2 (16 reps). This suggests that the subjects in group 2 (16 reps) may not have achieved maximum performance at the beginning of the test, but may have delayed it. Therefore,VLa max tends to be overestimated.
Since all test subjects were familiar with the test through preliminary examinations, coordinative reasons as well as learning and training effects can be excluded. Wittekind et al. [24] were able to determine "pacing" strategies for "all out" loads. They observed reduced performance at the beginning of longer loading periods (45 s) compared to shorter loading periods (10 s). It is possible that a strategy has been used here, consciously or unconsciously, with the aim of achieving high performance with low performance loss over the entire test period. This influencing factor for determiningVLa max should be investigated in the future using a blinded design. Adam et al. [5] found differences in t alac of up to one second on average during three repetitions of an isokinetic bicycle sprint. In the present study, the differences averaged less than 0.5 s. Introducing a time constant to minimize the variability ofVLa max would be conceivable for t alac . DeterminingVLa max would thus be less strongly influenced by a distorted t alac .
Methodologically, it should be noted that direct comparisons of the variability ofVLa max between the two loading periods are difficult to assess, as they were two independent groups of test persons. The allocation was randomized, but was not yet balanced due to the small sample size. A dependent design would make the comparison and its conclusions more reliable. In particular, the two groups of test persons showed different performances despite comparable training frequencies. With regard to performance, there was a significantly greater variability between the subjects in group 1, possibly having an influence on the higher correlations ofVLa max compared to group 2. The variability of La max may be dependent on carbohydrate intake and the level of the glycogen stores [25,26]. Since no dietary protocol was kept in the present study, there is uncertainty about possible influences of the test persons' diet.

Conclusions
A reliable reproducibility ofVLa max at isokinetic force loads is given both during short (8 reps.) and longer loads (16 reps.). To minimize any distortions of t alac by possible strategies of power output, using shorter load durations is preferable with regard to practical application. Furthermore, a constant for t alac may control the influence of the power output. Changes in theVLa max of over 0.11 mmol L −1 s −1 due to training can be described as a non-random increase or decrease in theVLa max .