In recent years, interest has increased in studying the repeated sprint ability (RSA) of footballers as a fundamental variable in team sports. RSA are characterized by high intensity actions interspersed with short recovery times.1 Considering the actual football competitive demands, players cover more distance at high speed, therefore need better performance to withstand high intensity.2 Although, RSAs don't usually appear frequently during full-time play due to the high energy demand this implies. RSA can be considered as an independent variable in the soccer training process, as it works as a means to develop acceleration, speed, explosive power of the legs, aerobic power and high running performance.3 In addition, it directly intervenes in muscle metabolism by improving maximum oxygen uptake, phosphocreatine (PCr) restoration and the glycolytic pathway. This is decisive for the physical performance as it improvement the recovery during sprints.3 In football, RSA testing is typically characterized by short-distance not exceeding 20 meters, with short recovery periods (<30 seconds). This is important since the combination of very high intensity actions together with insufficient recovery periods generates stresses on muscle metabolic processes1 and results in the development of neuromuscular fatigue.

When evaluating performance using an RSA, external load indicators such as best time, total time, and fatigue index are often measured. However, to understand the response to muscle energy demands, some internal load parameters such as heart rate (HR) and maximum oxygen consumption (VO2max) should be also monitored. These internal load variables together with the external load variables would give us a better overview of the decrease in RSA due to fatigue accumulation.4 HR is a physiological variable widely used in football, but HR cannot determine muscle energy capacity during intermittent high-intensity activities such as sprinting and passive recovery.5 For this reason, the development of theoretical models and software allow knowing other variables as energy cost (EC) and metabolic power (MP), which indirectly assess energy metabolism through the distance covered, speed and accelerometry during training and matches.6 It has been shown that MP analysis can be useful to adequately assess professional player´s performance limitations, especially during high-intensity actions.6,7 Furthermore, MP represents the required ATP hydrolysis rate to develop the muscular work needed during running.8

Non-invasive technology such as near infrared spectroscopy (NIRS) is used to evaluate muscle metabolism, which allows the measurement of muscle oxygen saturation (SmO2).9 The interest in evaluating SmO2 is due to its ability to discriminate metabolic changes within the muscle related to oxygen transport during high-intensity actions. In repeated-sprint activities, it has been proposed to measure the mean rates of desaturation (during the sprint) and re-saturation (during recovery between sprints) as indicators of metabolic performance.10 Furthermore, slower reoxygenation indicates that there is a slower recovery of intramuscular phosphates that are required for high-intensity exercise at levels prior to sprinting exercise lower performance.10 Besides, some factors such as high values ​​of body fat % and less muscle mass (body composition) affect performance of aerobic and anaerobic capacity, intermittent endurance capacity, RSA, small-side games and football specific activities.11,12 Likewise, the muscle oxygenation capacity is influenced by adipose tissue,13 because greater adipose tissue affects muscle oxygen trapping during exercise.

Therefore, we hypothesize that SmO2 is related to energy cost, metabolic power, heart rate, and performance during RSA. We also consider the body fat % as a variable that intervenes in performance due to its relationship with a better muscle oxygenation capacity and force production. This study aimed to 1) explore the potential relationship between energy cost and metabolic power with SmO2 dynamics during a RSA; 2) determine the degree of influence of fat percentage on muscle oxygen transport during RSA in footballers.

This study followed a cross-sectional observational design aiming to relationships between RSA and SmO2 studied, using physiological response of HR, %HR and mechanical responses based on time and individual speed during RSA. All tests were carried out on a heated sports court with an ambient temperature of 16–19 °C and relative humidity of 40%–50%, respectively. Data were obtained during the preseason. RSA stimuli were performed with the following criteria to avoid possible biases: a) a minimum of 48 hours of rest, after the last training was a recovery load [i.e., evaluations were carried out on Tuesday or Wednesday, depending on the day a game was played (Saturday or Sunday)], and the test was carried out before training to try to guarantee maximum recovery and (b) participants were instructed not to consume alcohol or caffeine 24 h before each test and maintain habitual sleep habits to avoid a decrease in performance. The participants were divided into six workgroups to guarantee considerable measurement time for each player.

A descriptive analysis of the variables was performed during the test expressed as mean ± SD. Analysis of variance, followed by Tukey post hoc tests, was used to compare the eight sprints of the RSA test. This approach allowed us to make comparisons between the baseline (Second Sprint) and the other sprints. This is due to the fact that the first sprint is always different from the other sprints, because an intramuscular pressure occurs that abruptly decreases the SmO2. Measurement error was expressed in “typical percentage error” (TE) and “minimum detectable change” (MDC). The typical error was calculated by dividing the SD of the difference score by √2. This specific percentage error is a coefficient of variation and is considered highly reliable if less than 5%. MDC values [also referred to as the “smallest detectable difference (SMD)], which reflects the magnitude of change necessary to provide confidence that the change was not resultant of random variation or measurement error, were calculated as 1.96*√2*TE.24 Then, a comparative analysis of the mechanical variables, HR% and ▽% SmO2 is applied for the three groups of fat percentage (level 1 = <9%; level 2 = 9.1 -11.5, level 3 => 11.6). Furthermore, to analyze the influence of the Muscle Oxygenation variables with variables of EC, MP, and performance, Multiple regression (Stepwise) analyses were performed. This was determined by the researcher, based on the factor within-subjects (behavior of the variable), Pearson's correlation coefficient >0.50, and a significance p-value of <0.05 (Supplementary Table 1), along with this the percentage of prediction between variables with R2. Data were analyzed using IBM SPSS Statistics V.22.0.

Table 1 shows the descriptive values of the sample, represented by the mean and standard deviation. Table 2 represents the mean ± standard deviation for each of the sprints. First, differences are observed between the first and second sprints in all the external load variables (Time, Speed, Energy Cost, and Metabolic power). Likewise, the HR obtained changes in each of the sprints. And the desaturation rate there were only changes between 1st sprint and 3rd sprint (Δ= 2.17 p= 0.035) and 8 (Δ= 2.40 p= 0.010), the re-saturation rate between 1st sprint and 3rd, 4th, and 8th sprints and the ▽%SmO2 with between 1st sprint and all other sprints.

Changes from 2ndsprint were also observed to identify the important points, where we found a decrease in performance (time and speed) from sprint 4 (Δ= -2.53 p= 0.018) and (Δ= 0.12 p= 0.021) respectively. Likewise, the decrease in EC (Δ= 0.88 p= 0.025) and MP (Δ= 7.23 p= 0.027) occurs from the 5th sprint. Regarding the SmO2 variables, they were identified in the re-saturation rate at the 5th sprint (0.166-0.024) and at the 8th sprint (Δ= 0.174 p=0.030) and the ▽%SmO2 at the 8th sprint (Δ= -12.92 p= 0.028), all in compared to the second sprint.

Also, considering that all subjects responded to the MDC, we found that performance decreases from the 5th sprint to the last sprint, and its associated speed begins to drop from the 6th sprint. Regarding SmO2 variables, the MDC of the reoxygenation rate was only at the 8th sprint and the oxygen extraction (▽%SmO2) at the 4th and 8th sprint. The response was highly variable for the SmO2, and there is no clear answer about the most significant changes. Likewise, in terms of heart rate, all the subjects responded in the first three sprints, then there are no significant changes due to entering high-intensity zones.

Fig. 2 shows the evolution of oxygen extraction, heart rate, energy cost, and metabolic power during RSA. The results were divided into a group of body fat %. First, in relation to oxygen extraction (▽%SmO2), from the 4th sprint, greater oxygen extraction is observed with lower levels of body fat: Tukey's multiple comparisons between Group level 1 vs. Group level 3 (Δ= -13; 95% CI= -20 to -6, p= 0.014) and Group level 1 vs. Group level 2 (Δ= 9.6; 95% CI= 2.0 to 17; p= 0.018).

Fig. 2.

Influence of body fat % in the response of muscle oxygen extraction, heart rate, energy cost and metabolic power during the sprint-repetitions.

Note: 1) RSA test graphs: a) muscle oxygen extraction, b) heart rate, c) metabolic power and d) energy cost. 2) 95% confidence interval: a) muscle oxygen extraction, b) heart rate, c) metabolic power and d) energy cost.

Difference between low body fat % (level 1) vs higher body fat % (level 2 and level 3), (*) p value <0.05 statistically significant. Difference between the body fat % (level 2) vs the body fat % (level 3), (+) p value <0.05 statistically significant.

(0.23MB).

Heart rate shows a difference in the first sprints: Tukey's multiple comparisons between Group level 1 vs. Group level 3 (Δ= -3.9; 95%CI= -6.2 to -1.7; p= 0.003), Group level 2 vs. Group level 3 (Δ = 2.1; 95%CI= 0.29 to 3.9; p= 0.026) and Group level 2 vs. Group level 1 (Δ= 6.0; 95% CI= 3.5 to 8.5; p= 0.000) from the 5th sprint there are no differences between groups where the high intensity zones relative to HR reserve % begins.

Regarding EC and MP, the difference is observed in the 6th, 7 and 8 sprint: Tukey's multiple comparisons between Group level 3 vs. Group level 1 (Δ = -4.6 95% CI = -6.2 to -2.9 p = 0.000) and Group level 3 vs. Group level 2 (Δ = -3.6 95% CI = -5.4 to -1.8 p = 0.001).

Fig. 3 presents the box-and-whisker plot of the equivalent index's mean values and anaerobic index through the body fat % groups. In graph a) a difference of the equivalent index is observed between-group level 1 vs group level 3 (Δ = 1.01 p = 0.034) and in graph b) that represents the anaerobic index differences were found between group level 1 vs group level 3 (Δ = -0.06 p = 0.043).

Fig. 3.

Differences of equivalent Index and Anaerobic Index by groups of body fat % levels.

(0.08MB).

Table 3 represents the multiple linear regression explanation models for external load and performance based on SmO2 and heart rate variables. It is shown that SmO2 better predicted energy cost, metabolic power, speed, and total time than heart rate during the test, with a prediction percentage of 57% (EC), 61% (MP), 59% (Speed) and 59% (TT sec).

This is the first study that related SmO2 dynamics with variables of energy cost and metabolic power associated with speed. %Body fat stands out as a contaminating performance variable that can influence the ability of muscle oxygen extraction to maintain high speed during RSA.

Regarding Table 2, it indicates that the performance decrease in EC, MP, total time, and muscle oxygenation begins from the 4th and 5th sprint, results are similar to previous studies,10 this is due to the time factor (> 1 m) of the test as a causative agent of fatigue and decrease in PCr and glycolytic pathways25,26 necessary to satisfy the energy demand of the sprint. Also, during the first sprints there is a decrease of Intramuscular oxygen partial pressure (IPO2), which causes immediate desaturation. Then the oxidative phosphorylation pathway begins and an increase in blood flow due to the hyperemic response; this phenomenon is associated with the ability to remove and buffer H+ ions and the ability to replace intramuscular PCr.27 Interestingly, these HR stages no longer detect the MDC and have begun a slow increase in VO2 and HR kinetics due to the high-intensity. Likewise, in ▽%SmO2 a decrease in the values ​​is observed as the test progresses compared to the 1st sprint. Other studies have explored the communication between the decline in SmO2 and the increase in HR as a function of athletes' speed.28,29 However, they do not interpret muscle oxygenation with changes in metabolic pathways because HR discriminates the training zones, but SmO2 decreases and increases with movement, regardless of slow HR kinetics during high-intensity exercise.10,28 This is one of the problems of interpretation of data variability, and this does not allow an adequate analysis of performance in scientific studies; authors have discarded it as a performance factor.32 while other authors promote it as an observable variable product of the impact of speed.30–32 In this regard, our research found that SmO2 variables better predicted performance: EC, PM, and total time. However, our study used the MDC, with which we can know athletes respond to the true change (%) in each of the sprints. The ▽% SmO2 was observed as the most sensitive parameter, where we found differences in each sprint and not in the desaturation rate and re-saturation rate as previously described.10

Consequently, we were able to observe in Fig. 2 a greater ▽% SmO2 with a lower body fat, (<9% fat) from the 4th sprint compared to group levels 2 and 3 (>9% fat). This can be explained for two reasons: 1) it has been observed higher SmO2 values ​​in subjects with greater adipose tissue,13 however this is not very likely, because the sample obtained a mean calf skinfold 9.3 mm (see Table 1), which is considered low and with good reproducibility and sensitivity for measurements with the MOXY instrument33 and 2) the activation of type II fibers is better in high trained subjects with less body fat %34 in addition, have a greater capacity to extract oxygen through the glycolytic pathway, and not the type I fibers that depend more of the HHB or increase SmO2 that indicates the “oxidative phosporylation”,35–37 this is more likely to have occurred in our study, because similarly, together with the extraction of muscle oxygen, a better performance was obtained in the equivalent index and anaerobic index (See Fig. 3). In general, anaerobic systems can produce high metabolic power but have limited energy capacity. In contrast, the aerobic system has a large capacity, but is characterized by considerably less peak power and inertia.38

In this same context, lower HR values ​​were observed in group level 1 (body fat %), this means that a greater ▽%SmO2 by glycolytic route compensates the cardiac output,39 studies such as Casazza et al.,40 and Vasquez-Bonilla et al.,41 support this foundation since subjects with a lower body fat can metabolize ATP energy faster through the anaerobic metabolism (glucose) in a short time; this also delays the activation of type I fibers that are dependent on oxygen and increase the blood flow42 at the end of the RSA tests.10 However, without measurements of the contributions of the energy pathways,43 it is difficult to know the performance of one metabolic substrate or another.

Finally, for the performance prediction employing SmO2, it depends on the body fat % as a variable that contaminates performance; heart rate may not have much relevance in interpreting the decrease in EC, and MP is currently used. Recent studies such as that of Paquette et al., (2018) highlighted the importance of using peripheral adaptations during short and long events since they were able to predict performance through muscle oxygenation by multiple regression analysis. The ▽% SmO2 with NIRS is a predictor of metabolic performance that approximates VO2max and energy expenditure. However, we suggest that with portable NIRS we will obtain better results in the information when interpreting the physiological adaptations of peripheral form and fatigue resistance.12 However, we need more studies as in real game situations.

Considering the results of this study, it's proposed to measure SmO2 dynamics and ▽% SmO2 as a physiological variable for to find potential differences in sport performance, because when the energy cost and metabolic power begins to decrease, so does the increase in muscle oxygen extraction during repeated sprints in footballers.

Therefore, SmO2 seems more promising than %RH as an indicator of internal load during high-intensity running. In addition, we address the influence of having a low body fat % during the season, since it greatly influences the players speed, together with the ability to oxygenate the muscles.

The study design was approved by the Bioethical and Biosecurity Commission of the University of Extremadura (Reg. Code 131/2018).

This study has been supported by the Government of Extremadura with a grant from the European Regional Development Fund (Ref: GR21189).

The data that support the findings of this study are available on request from the corresponding author, [initials]. The data are not publicly available due to [restrictions e.g. their containing information that could compromise the privacy of research participants].