Biomechanical characteristics are important swimming performance determinants, fundamental in understanding propulsion mechanics in the highly specific hydrodynamic environment. Analyzing swimmers’ force production should be a priority in their training control and research, as an effective propulsion is fundamental for competitive success [1
]. For this purpose, tethered swimming has been one of the most frequently used methods, yielding substantial associations between tethered forces and swimming performance in sprint events [4
]. These studies underpin the notion that movements used in training and testing should be mechanically similar to those used in competition [7
Complementarily, it is known that similar force application from both right and left body sides can positively affect swimming performance, particularly by reducing drag, promoting better body alignment and lowering intracycle velocity variations [8
]. In fact, ~96% of the human population present a perceptible asymmetry level [9
], which may negatively influence exercise performance, mainly in cyclic and continuous sports [10
]. Since elite swimmers spend a huge number of hours in-water and land workouts, they are more vulnerable to overuse injuries, particularly if they present muscular and technical imbalances [12
]. Furthermore, if hydrodynamic and inertial limitations are also considered, swimmers need to be as symmetrical as possible when propelling through the water.
The isokinetic dynamometer is the laboratorial gold standard to evaluate asymmetries and imbalances in force production, permitting analysis of specific muscle groups and movements [12
]. However, even if its use is common in muscular symmetry analysis, it does not allow evaluating specific swimming movements. For this reason, tethered swimming has been conducted to quantify and analyze both force production and asymmetries during swimming [11
]. Surprisingly, it is unclear if measurements obtained from tethered swimming and isokinetic dynamometer are related, and whether swimmers generate a similar level of force production symmetry in-water and dryland conditions.
Using an innovative approach, with closer approximation to real-world swimming characteristics, of assessing shoulder and knee extension force values (hand placement, angular velocities, isolated and combined analysis), the aim of the current study was to determine the degree of association between tethered swimming peak propulsive forces and dry land isokinetic torque and power. We also sought to determine whether there is a similar contralateral symmetry regarding force production in both the above referred conditions and the nature of their symmetry index (SI) relationships. Finally, we determined whether swimming symmetry can be identified by using only the first 10 upper limb cycles during tethered swimming.
Tethered swimming and force production isokinetic variables were moderately to very largely directly related, with correlation values ranging from r = 0.62 to 0.96. The peak torque, mean torque, mean power and mean peak forces values of the first 10 upper limb cycles and 30 s of front crawl and backstroke tethered swimming, for both preferred and nonpreferred body sides, are presented in Table 1
. When analyzed as combined isokinetic forces, the relationships ranged from r = 0.65 to 0.94. In most of the evaluated variables, the non-preferred side presented higher absolute correlations than the preferred side. Values obtained during the first 10 s of tethered swimming presented similar correlation to the isokinetic forces when compared to the 30 s data.
No substantial differences were evident between SI of isokinetic and tethered swimming force production or between first 10 upper limb cycles and 30 s of tethered swimming (Figure 2
). Only in two swimmers, an asymmetrical force production in tethered swimming (SI > 10%) was identified. However, in the isokinetic evaluations, eight swimmers presented asymmetrical force production (at least in one of the studied variables), reduced to three swimmers when the shoulder and knee extension combined variables were evaluated. Absolute values of SI obtained from tethered swimming and isokinetic force variables showed no substantial differences, but a high variability was evident between subjects (Figure 2
). The correlation analyses showed that only the SI of 10 upper limb cycles had a moderate inverse relationship with SI of power of knee extension at 300°/s (r = −0.66) and combined average power SI (r = −0.60). No substantial relationships in the other SI (r = −0.06 to −0.41 and 0.04 to 0.44 with p
= 0.18 to 0.88).
Swimmers showed a similar force production between preferred and nonpreferred body sides (both in tethered front crawl and backstroke swimming) and isokinetic tests (Table 2
= 0.18 to 0.97). In the tethered swimming and isokinetic tests, intraclass correlation between repetitions for both preferred and nonpreferred side were excellent (ICC: 0.989 to 0.997; 95% CI: 0.963 to 0.999 with p
< 0.001). Agreement analyses between SI’s of isokinetic evaluations and 30 s tethered swimming presented a bias standard deviation >9% and the 95% limits of agreement higher than 19%, demonstrating that the SI were independent from each other (see Table 2
). The SI of the first 10 upper limb cycles and 30 s in tethered swimming test showed a very strong relationship (r = 0.99). In addition, comparison between the SI calculated with the first 10 upper limb cycles and 30 s of tethered swimming showed a good agreement with a bias of −0.7 ± 1.2% and 95% limits of agreement of −3.0 and 1.6%, resulting in a consistent behavior of SI during the 30 s effort.
Our results indicate moderate to very strong relationships between in-water and dryland conditions force production, which supports the first hypothesis that forces produced in land condition are closely related with the capability to apply force in-water. We also identified similar force production between preferred and nonpreferred body sides, showing that elite swimmers, as a consequence of their training, exhibit upper and lower body symmetry. Our swimmers presented similar but independent contralateral symmetry during swimming and dryland testing. The very strong relationship between isokinetic and tethered forces supports the assertion that isokinetic ones are related to swimming performance and reinforced that the 30 s tethered swimming test is a good predictor of swimming performance [6
The above outcomes are in accordance with the study that reported an inverse relationship between isokinetic knee extension force and mono-fin swimming 100 m time [25
]. However, some studies indicated that isokinetic force variables were not associated with swimming performance [26
]. These last studies used dryland tests that employed different movements types than those used in swimming (different muscular groups and patterns). Regarding not only isokinetic variables, other studies showed strong relationships (e.g., r = 0.78–0.94) between force variables in dryland exercises (bench press and squad jumps) and force produced during tethered swimming [29
]. Taken together, these data reinforce the notion that forces produced in dryland testing can be transferred to water when patterns of movement angular velocities and amplitude are similar.
The biomechanical asymmetries are useful both for clinical and research purposes, permitting the characterization of the functional imbalances between contralateral limbs [31
]. Theoretically, the front crawl and backstroke intracycle velocity variation can be enhanced by this functional imbalance, increasing the energy cost of swimming [32
] and deteriorating body postures with increasing hydrodynamic drag [10
] at a given swimming speed. The current results demonstrate that elite swimmers are more symmetrical than lower level swimmers (data from [16
]), since we did not observe considerable differences between the force production of contralateral body sides.
However, the mean values of the SI need to be considered with caution, as asymmetry measurements are direction-dependent and a tendency for both preferred and non-preferred predominance will make the mean values tend towards zero. Therefore, comparing absolute values between conditions (in-water and dryland) is preferred. The similar SI in-water and in-dry land conditions observed in this study likely reflects the technical training to build a balanced and efficient swimming technique. This process typically involves a combination of pool-based dryland training enhance bilateral symmetry and muscular proficiency.
Despite similar tethered and isokinetic force production SI, the correlation analyses show the few substantial relationships between almost all tethered and isokinetic SI. The agreement analyses presented a bias of ~10 and 95% limits of agreement ~20% indicating why SI values were not underpinning the change of SI direction in almost all swimmers evaluated. These data indicate that during swimming, one upper limb is typically used for propulsion and the other for control and support. In addition, it appears that the preferred upper limb is the limb of control and support during swimming, while the non-preferred limb is used mainly for propulsion. This pattern of limb control and propulsion explains the higher correlation values on the non-preferred body side.
This outcome highlights that the explanations of asymmetry observed in-water are more related to swimming technique adaptations consequent of the aquatic environment and to a lesser extent to muscular imbalances [34
]. We expected changes on SI across 30 s test in response to the change of energetic system and muscular recruitment [18
]. However, when examined the force (a)symmetry development along the maximal test, very strong direct relationships (r = 0.98) and similar values between the SI from the first 10 upper limb cycles and 30 s where found. The behavior of SI during the first 10 cycles showed high agreement with the 30 s data, enhancing the capability of 10 cycles to indicate likely patters during the overall effort. These results can be justified by the adaptations of swimming technique during the test being more important than the capacity of force production for SI behavior. However, this outcome is contrary to that shown in a previous study [16
] using only front crawl tethered swimming in low-level swimmers. These outcomes indicate that high-level swimmers can maintain symmetry during the effort even when the absolute force produced is diminishing.