The primary assessment of anaerobic performance on land is based on the Wingate Anaerobic Test (WAnT) (Bar-Or, 1987), in which the obtained power of exercise is presented as a function of the trial time or covered distance. The maximal power level and the time to reach and maintain the velocity peak may indicate phosphagen capacity and the average power determines glycolytic capacity (Papoti et al., 2013). A previous study reported a strong correlation (r = 0.83) between maximal power measured on the cyclo-ergometer during WAnT and freestyle swimming time for 50 m (Hawley and Williams, 1991). The relationship between the time needed to swim 50 m of freestyle and mean power in WAnT performed with upper limbs (r = 0.63) and lower limbs (r = 0.76) were established by Hawley et al. (1992). Duche et al. (1993) found a strong correlation between the time necessary to swim 50 m of freestyle and the average power obtained in the WAnT test (r = 0.68). However, it is known that the specificity of the water environment, i.e., water density, buoyant force, or hydrodynamic drag forces (Truijens and Toussaint, 2005) largely influences aquatic exercise (Kimura et al., 1990). Thus, the diagnostic value of WAnT during land-based conditions is questionable for assessing the exercise capacity of swimmers because it does not reflect the actual free-swimming conditions (Smolka and Ochmann, 2013; Soares et al., 2014).

The evaluation of anaerobic performance in swimmers was also conducted using in-water tests. Shionoya et al. (1999), measured the net mechanical power calculated from the product of tethered force and swimming velocity in real-time (during swimming) with an ergometer located on land. Smolka and Ochmann (2013), used nonchronometric approaches during in-water tests similar to the WAnT. They analyzed a series of instantaneous swimming velocities over 100 m based on data recorded with a video camera. Bátorová et al. (2021) and Demarie et al. (2019), analyzed anaerobic performance using Inertial Measurement Units (IMU). Fernandes et al. (2008) and Neiva et al. (2011), introduced anaerobic critical velocity (AnCV), as a parameter to evaluate and monitor anaerobic training among swimmers.

The fully-tethered swimming method was also used to assess the anaerobic performance of swimmers in the water through a 30 s maximal front crawl swimming test (Nagle Zera et al., 2021). The swimmer does not move forward in fully-tethered swimming, and the propulsive movements generate zero horizontal velocity (Samson et al., 2019). Kjendlie and Thorsvald (2006) showed that a fully-tethered swimming power test was highly reliable (Cronbach’s α = 0.99). However, Samson et al. (2019) reported that the hand’s orientation is perpendicular to the surface of the water earlier than in free swimming, and velocity and acceleration are different in fully-tethered vs free swimming. Furthermore, fully-tethered swimming sustains the strength potential of a swimmer rather than the ability to apply force effectively (Ruiz- Navarro et al., 2020), leading to the overestimation of force (Santos et al., 2023). This gives rise to the question of underestimating the measurement of the kinematic of swimming stroke (e.g., acceleration and maximal velocity) which are significant predictors of an adolescent swimmer’s performance (Sokołowski et al., 2022) and consequently may result in an unrealistic view of the bioenergy of exercise (Thompson et al., 2004).

Another method helpful in evaluation of swimming performance is semi-tethered swimming. According to this approach, a swimmer moves forward in the water subjected to an external load, while the forward motion induces relative streamwise water flows around the body. Moreover, the swimmer maintains more natural swimming mechanics. It makes the semi-tethered swimming test more specific to the free-swimming condition compared to fully-tethered swimming (Hancock et al., 2015) depending on the load (Cuenca-Fernández et al., 2020). The semi-tethered method has been used to evaluate the net power output which was calculated by multiplying the speed and force data produced against an external load (Dominguez-Castells et al., 2013; Kimura et al., 2013). However, the athletes often swim with almost a constant velocity that makes the sum of the mean propulsive and resistive forces during one stroke cycle nearly zero. Hence, the power calculated in semi-tethered swimming considers the power against the external load and not the propulsive power produced by the swimmer (Gonjo et al., 2021). The different use of semi-tethered swimming might be useful to measure the velocity with different external loads to examine the interaction between load and velocity (Gonjo et al., 2020; Olstad et al., 2020; Gonjo et al., 2021). The maximum velocity at zero load corresponds to maximal velocity during free-swimming. Moreover, the force exerted by the swimmer at fully tethered swimming should match the magnitude of the tethered force produced by the maximum load at zero velocity. The semi-tethered swimming is a reliable methodology and can potentially be used to assess strength and velocity capabilities during swimming (Olstad et al., 2020).

Short time and maximum intensity of exercise are the main criteria for the rate of anaerobic pathway contributions, such as split into phosphocreatine and anaerobic glycolysis (Driss and Vandewalle, 2013). The activation of phosphagen energy sources contribute mainly to the velocity increase for the approximately first 3 seconds of exercise (Hirvonen et al., 1987). The sum of the time to reach v_max (tv_{to max} [s]) and maintain (tv_{at max} [s]) maximal velocity is included in the time of maximal phosphagen transformations (Hawley, 1995). During longer swimming sprints, the maximal velocity (v_max) and minimum velocity (v_min) were strong determinants of the time obtained in 100 m front crawl swimming (r = −0.90, r = −0.92, respectively) (Smolka and Ochmann, 2013). These variables related to anaerobic metabolism (phosphagen and glycolytic) show a significant relationship with swimming performance at distances of 50 m and 100 m (Vitor Fde and Böhme, 2010; Neiva et al., 2011). Furthermore, after only a few seconds of exhaustive work, the power produced by the ATP-PCr system decays, so that after 6 seconds it provides only half of the total energy requirements (Hawley, 1995). After 6–10 s of maximal exercise, the contribution of anaerobic energy from PCr and anaerobic glycolysis is essentially equivalent (Hargreaves and Spriet, 2020). Hence, the observation of free-swimming velocity patterns and fatigue over time during a maximal effort could provide information about a swimmer’s anaerobic potential and the dynamics of the ATP-PCr cycle to lactic acid transition (Smolka and Ochmann, 2013; Soares et al., 2014). The presented arguments allow for the assumption that a semi-tethered swimming test can be used as an objective tool for indirectly measuring a swimmer’s potential to perform at maximal anaerobic effort. The potential to perform maximal effort is defined as the work capacity (anaerobic) lasting approximately 15 s (Driss and Vandewalle, 2013).

Adolescents have unique physiological features that greatly differ from those of adults, including a lower ventilatory efficiency during progressive exercise (Cooper et al., 1987) and a lower anaerobic peak power (Van Praagh and Doré, 2002). Moreover, anaerobic capacity is known to increase during puberty so the relative anaerobic contribution to short-distance swims would likely be lower in prepubertal versus adult swimmers (Mezzaroba and Machado, 2014). Hence, adolescents have a lower work capacity and therefore are less efficient than adults (Ratel and Blazevich, 2017). In this cohort, anaerobic metabolism is in its developmental phase (Olbrecht, 2000), which increases the susceptibility to fatigue (Zając et al., 2023). Additionally, in adolescents, inexperienced swimmers, fatigue reactions are also a function of deficits in the quality of swimming technique (Bassan et al., 2016). Different swimming velocity profiles were also linked to the age and competitive level (Smolka and Ochmann, 2013; Soares et al., 2014; Nagle Zera et al., 2021). Consequently, a different swimming velocity profile is to be expected in adolescent swimmers than in adult swimmers.

Most investigations in the area of an indirect assessment of anaerobic performance in swimming concern adult swimmers (Soares et al., 2014; Nagle Zera et al., 2021; Ruiz-Navarro et al., 2022). Few studies exist which include adolescent subjects (Zamparo et al., 2000; Strzala and Tyka, 2009; Mezzaroba and Machado, 2014; Dekerle, 2021). There is a lack of research that has used field-based swim tests to examine the swimming velocity profiles in order to objectively monitor the current state of performing maximal anaerobic effort among adolescent swimmers, which is a unique feature of this study. Furthermore, based on a literature review, the current state of knowledge regarding the interpretation of the diagnostic value of various methods for assessing the anaerobic potential of swimmers show discrepancies. Moreover, there is a small number of scientific reports that determine the reliability of using a motorized resistance device with adolescent swimmers.

Therefore, the current study aimed to evaluate the reliability of a sprint performance test determining an individual profile of instantaneous swimming velocity from semi-tethered front crawl swimming to indirectly assess an adolescent swimmer’s current potential to perform the maximal anaerobic effort. It was hypothesized that the sprint performance test with semi-tethered front crawl swimming is a reliable method to indirectly indicate a potential to perform the maximal anaerobic effort among adolescent swimmers.

Figure 3 shows an individual velocity profile based on a polynomial curve from which the variables describe the potential to perform the maximal anaerobic effort of adolescent swimmers can be read: maximal velocity (v_max), minimum velocity (v_min), the distance (D_to v_max) and time (tv_{to max}) a swimmer need to reach maximal velocity and the distance (D_at v_max) and time (tv_{at max}) at which a swimmer can maintain maximal swimming velocity according to the criterion of a 5% decrease of v_max, and time to reach 15 m (t_trial15).

Figure 4 illustrates an example (one representative of a girl and a boy) of normalized data (range from v_min·v_max ⁻¹ to 1) with an individual velocity profile based on a third-degree polynomial curve.

Table 1 shows the relative and absolute reliability indicators (variability estimates (CV), relative reliability (ICC), and absolute reliability (SEM, MDC) for parameters plotting the velocity profile (v_max, v_min and t_trial15) and the variables describing this profile (tv_{to max}, tv_{at max}, D_to v_max, D_at v_max and FI).

For the parameters plotting the individual velocity profile, CV showed low variance for v_max, v_min and t_trial15 (CV<5%). ICC_2,1 for v_max, v_min and trial t_trial15, demonstrated respectively good (ICC_2,1: 0.88) very good (ICC_2,1: 0.95), and excellent (ICC_2,1: 0.98) rate of relative reliability. The absolute reliability indicators, i.e., SEM and MDC were higher for v_max (SEM: 0.07 m·s^-1; MDC: 0.20 m·s^-1) than v_min (SEM: 0.05 m·s^-1; MDC: 0.14 m·s^-1). However, in the t_trial15 SEM was 0.35 s, while MDC 0.97 s.

The variables describing the velocity profile (Table 1) showed: high variance of CV for tv_{to max}, tv_{at max}, D_to v_max and D_at v_max. ICC_2,1 was moderate for D_to v_max (ICC_2,1: 0.50) and weak for tv_{to max}, tv_{at max}, and D_at v_max. Results concerning the absolute reliability indicators showed that tv_{at max} (SEM: 1.72 s; MDC: 4.76 s) was higher than tv_{to max} (SEM: 0.73 s; MDC: 2.01 s). D_at v_max (SEM: 1.76 m; MDC: 4.87 m) was higher than D_to v_max (SEM: 0.76 m; MDC: 2.12 m). The relative and absolute reliability indicators for fatigue index (FI) showed high variance (CV>5%), low rate of relative reliability (ICC: 0.48), and SEM and MDC indicators for FI were 2.73% and 7.56%, respectively.

Bland-Altman index revealed an acceptable agreement (LoA ≤5%) with no significant differences between the two measurements (trial #1 and #2) in the semi-tethered swimming test for vmax (p = 0.77), vmin (p = 0.81), and t_trial15 (p = 0.22) (Figure 5).

In this study, the reliability of a front crawl sprint performance test based on the estimation of the individual instantaneous swimming velocity profiles in semi-tethered swimming was assessed. Among the analyzed variables, v_max, v_min, and t_trial15 had reliable outcomes indicating low variability (CV<5%) and excellent relative reliability (ICC_2,1 from 0.88 to 0.98) and absolute reliability (SEM from 0.07 to 0.35 and MDC from 0.14 to 0.97). Also, the Bland-Altman analysis showed an acceptable agreement (LoA ≤5%) with no significant differences between trial #1 and #2 in the test with semi-tethered swimming for v_max (p = 0.77), v_min (p = 0.81), and t_trial15 (p = 0.22) (Figure 5). It can be assumed that v_max, v_min, and t_trial15 of the velocity profiles qualify them as objective indirect measures of an adolescent swimmer’s current potential to perform maximal anaerobic effort. This is partly consistent with Soares et al. (2014) where maximal swimming velocity (v_max) is considered an indirect anaerobic performance metrics over short distances (50 m). Furthermore, Smolka and Ochmann (2013) reported that v_max and v_min were determinants of the sprint time in 100 m swimming (r = −0.90, r = −0.92, respectively).

When interpreting these results, attention should be made on how to obtain the v_max and improve the v_min. A reflection of these adaptations is the maintenance of the fatigue index (FI) level or its reduction, which may increase swimming performance at sprint distances. In addition, the variables describing the velocity profile, tv_{to max}, tv_{at max}, D_to v_max, D_at v_max, and also FI were characterized by high variability (CV>5%) and slightly worse reliability in the relative dimensions (ICC from 0.01 to 0.50) and absolute SEM (0.73–2.73), and MDC (2.01–7.56). The high variability and low reliability in the following variables (tv_{to max}, tv_{at max}, D_to v_max, D_at v_max, FI) could be caused by starting the velocity measurement at a point located 3 m from the pool wall, where the study participants obtained different swimming velocities.

In the results from Demarie et al. (2019), strong correlations were found between the average power obtained in WAnT on a cycle ergometer and the average velocity obtained in the swimming test with IMU sensors for 75 m freestyle in a short (r = 0.809) and long course pool (r = 0.700). Additionally, the best times achieved by swimmers in the 50 m and 100 m freestyle races strongly correlated with the average swimming velocity obtained in the 75 m test (r = 0.659–0.952) and the average power in WAnT (r = 0.736–0.855) (Demarie et al., 2019). However, Dotan (2019) referred to the results of McKie et al. (2018) and suggested that the confrontation of the results of anaerobic tests carried out in a specific environment and under laboratory conditions (WAnT) is inappropriate and methodologically incorrect from the point of view of movement modality. Therefore, there are premises to believe that the semi-tethered swimming test enables keeping the natural water conditions, a similar muscular activity (Bollens et al., 1988), swimming stroke, and physiological responses to free swimming (Morouço et al., 2014; Ruiz-Navarro et al., 2020).

The adolescent swimmers in the current study performed their tests within the time characteristic of anaerobic tests (12.67 ± 2.27 s for swimming 15 m) (Table 1). Similarly, short tests lasting 10–15 s were used in the studies by Yeater et al. (1981) and Costill et al. (1986), with fully tethered swimming to assess the individual maximal power values representative of the ATP-PCr catabolism rate. Thus, a shorter test distance to assess the anaerobic potential opens the possibility of obtaining a higher maximal velocity (Inbar et al., 1976; Zajac et al., 1999). In this context, considering the psychological factors of effort maximization (Hatfiel and Landers, 1987; Zajac et al., 1999), the current argument seems to confirm the alternative application of the semi-tethered swimming test in water instead of using the Wingate test.

It seems that the basis for the lower t_trial15 and v_max values (Table 1) is the shorter test duration time presented in the individual velocity profiles (polynomial curve) (Figures 2, 3). The discussed difference may also be due to too short of a rest interval between the first and second trial. The duration was adopted based on research conducted on adult swimmers (Hancock et al., 2015), which draws attention to the potentially greater susceptibility of adolescent swimmers to fatigue reactions (Bailey et al., 1995). However, it is presumed that 70% of the ATP and PCr are restored within 30 s, 90% of the rest value reaches within 2 min, and 100% within 3–5 min after an “all-out” effort (Hultman et al., 1967). Based on this suggestion, the low values of t_trial15 and v_max (Table 1) obtained by the tested swimmers may result from the fact that due to their age and inexperience, they build their exercise capacity along with the elimination of deficits in the quality of their swimming technique (Lätt et al., 2009). In general, it should be assumed that the explanation of the problem discussed here requires further research, considering to extend both the water test duration and recovery time between performance trials.

Energetic contribution to a sprint-swim performance over 22 s in adults has been estimated, i.e.,: 38% from ATP-PCr system, 48% from glycolytic system, and 4% from the aerobic system (Rodríguez and Mader, 2021), with a greater contribution of the first two systems. As performance times decrease to 25 m, there is a larger aerobic contribution for longer performance times (slower swimmers, i.e., adolescents). Moreover, during the prepubertal stage, the anaerobic contribution to short-distance swims is likely to be lower compared to pubertal individuals (Mezzaroba and Machado, 2014). Hence, field-based anaerobic tests should be complemented with additional laboratory-based measurements, e.g., maximal accumulated oxygen deficit (Dekerle, 2021). In the past, anaerobic potential to produce high-power output was measured during maximal 30 s cycling and arm cranking tests among 13-year-old boys and girls (Hawley et al., 1992) and adolescent swimmers aged 16 (Strzala and Tyka, 2009). A correlation of around r = 0.650 between high-power output and swimming performance <100 m distances suggests sprint performance test (to 50 m) may be appropriate to evaluate the anaerobic potential of adolescent swimmers. The field-based semi-tethered test used in this study enables assessment of the current state of performing maximal anaerobic effort in sprint swimming through individual velocity profiles. This approach allows for monitoring the adaptive changes during the training process at any given time and is critical in the context of improving training routines. This leads to the belief that the results obtained from the created velocity profiles during semi-tethered swimming may be the basis for modifying the quantitative and qualitative components of the training load to increase the exercise capacity of adolescent swimmers in sprint races. The practical value of this study indicates that using the field-based semi-tethered test, a coach can compare two swimmers with the same trial time and different trajectories of velocity profiles. Furthermore, the simplicity of such tests encourages their use in swimming profiling.

To conclude, the changes in instantaneous swimming velocity in time/distance during front crawl semi-tethered sprint test with maximal intensity can be analyzed as individual mathematical models of velocity profiles in adolescent swimmers. The findings revealed that the v_max, v_min and t_trial15 estimated from these profiles were considered reliable and can be assigned specificity characteristics. Consequently, the applied procedure allows objective and individualized quantification of the selected parameters to indirectly evaluate the current level of adolescent swimmers’ potential to sustain maximal effort in the anaerobic condition. The observation of the swimming velocity profiles including reliable parameters allows for monitoring the adaptive changes in the current state of performing maximal anaerobic effort matters during the training process. As a result of a field-based semi-tethered test, the coach may identify the capacity for improvement and provide guidelines to athletes for the preparation of specific training sessions for adolescent swimmers. The results obtained from the created velocity profiles may occur point of reference for modifying the quantitative and qualitative components of the training load in order to enhance the exercise capacity of adolescent swimmers in sprint performance. Future studies should further explore these aspects to provide coaches and athletes with more detailed and valuable information to better inform training decisions.