On the utility of a rotating swim bench as a freestyle swimming emulator

Abstract

The swim bench is an isokinetic ergometer intended to aid competitive swimming training by replicating the underwater pull of the freestyle stroke. Yet limited literature addresses the biomechanical fidelity of the swim bench relative to in-water swimming (Olbrecht & Clarys, 1983; Piovezan et al., 2019). Further, the lack of body roll on the swim bench may limit simulation fidelity. Accordingly, the Kayak Pro SwimFast swim bench includes a rotating bench setting. Specific changes to a competitive swimmer’s kinematics and muscle activation patterns with a rotating setting is unknown. The purposes of this study were assessment of the influences of swim bench settings on kinematics and muscle recruitment, and exploration of the similarity of kinematic data between swim bench and in-water data sets. Fifteen collegiate and/or national level, male, competitive swimmers completed 8 sets (4 rotating and 4 fixed) of 30 second continuous freestyle stroke pulling on a Kayak Pro SwimFast swim bench. Surface electromyography of 12 right upper limb muscles and bilateral upper limb and torso kinematics were collected. Time-series swim bench kinematic and electromyographic data were compared using statistical parametric mapping, enabling holistic evaluation. The swim bench kinematics were compared to existing in-water data from McCabe (2008). Few kinematic and electromyographic differences existed between the rotating and fixed swim bench settings. Briefly, left shoulder elevation was higher on the rotating swim bench setting nearing the end of the push leading into the recovery phases of the freestyle swimming stroke (p = 0.08). The left shoulder axial rotation approached significance at during the push phase, with a higher internal rotation angle on the rotating setting. Right wrist radial deviation was greater on the fixed setting during the recovery phase (p = 0.024). Infraspinatus achieved greater activations on the fixed bench than the rotating during the late pull to early push phases (p = 0.012). Despite the device’s roll design, no differences existed in shoulder roll between the rotating or fixed setting and, regardless of bench setting, participants laterally flexed the torso, potentially as compensation for the overall lack of roll allowance. The similarities between settings indicated that the rotating setting may not substantially augment the realism of stroke mechanics on the swim bench. Compared to in-water swimming, the swim bench produced similar elbow flexion angles and maximum vertical depth at the third distal phalanx. However, entry phase duration phase decreased on the swim bench, while the pull, push, and recovery phases increased (p < 0.0001). Additionally, the stroke length, mediolateral stroke width range, and vertical stroke depth range, and total shoulder roll decreased (p = 0.0002, p < 0.0001, p < 0.0001, p < 0.0001). The differences between the stroke mechanics, lack of entry phase, and addition of the lateral torso flexion on the swim bench are notable considerations for swim bench use in training and research. Swimmers could develop associated habits that reduce swimming economy, and the results suggest that using the swim bench in training may not extrapolate to in-water swimming.