Three-dimensional kinematics in backstroke swimming: reliability, isokinetic strength and asymmetries at sub-maximal and maximal paces
Item statusRestricted Access
Embargo end date23/06/2023
Dias, Margarida Ferreira
Research in front crawl (FC) swimming has revealed that: a) a change of pace affects stroke kinematics, b) asymmetries in some variables may affect performance and, c) the arms are estimated to contribute up to 85% of the propulsion, making the isokinetic strength measures and their relationships with performance important. Backstroke (BK) swimming is expected to present different kinematic characteristics due to the constraint of the shoulder movement but has been much less investigated. Very few researchers have investigated BK with 3D motion analysis and no studies have explored any complex kinematics in the three areas mentioned above. Generally, in swimming research it has been often assumed that performance is highly reproducible, and that single trials are sufficient to provide an accurate insight on swimming technique. However, it would be of interest to assess the reliability of swimming kinematics for repeated trials in maximal (M) and sub-maximal (SM) BK paces, as such data could provide confidence in the experimental results of a study. Therefore, the current thesis was designed to address the four aims: assess reliability of kinematic variables during M and SM paces in BK; investigate key kinematic variables within and between paces in BK and explore their relationship with performance; quantify key strength variables and identify any kinematic or strength asymmetries, and assess their relationship with performance. Following a pilot study, the aim of experimental Study One was to determine the reliability and repeatability of key kinematic variables at SM pace (replicating a 200-m pace) and M pace (maximal effort). Seven participants (mixed-sex) of national level, who specialised in BK completed four trials in each pace. A two-way repeated measures ANOVA was conducted to determine systematic bias; typical error and coefficient of variation (CV) to assess within subject variation; and test-retest reliability assessed using an intraclass correlation (ICC 3,1), so the number of trials to achieve a minimum of ICC of 0.80 could be established. Results showed that two trials were sufficient to provide acceptable reliability for the majority of the variables at both paces. The results of experimental Study One determined the variables to be investigated and the number of trials within the experimental study. Experimental Study Two focused on kinematic (at both paces) and strength measures. Differences between paces, asymmetries and relationships with performance (defined as mean velocity in the direction of motion [v]) were explored. Twelve swimmers specialising in BK participated in this study. For the kinematic assessment, they completed two BK trials in both a SM and a M pace. A paired t-test and Pearson’s product-moment correlations were used. The duration of the pull and push phases, swimming v and stroke frequency (SF), and trunk twist (TT) on the non-preferred (NP) side, and elbow angle on the preferred side (P) were significantly greater at M pace. The duration of the entry and release phases, normalised stroke length, and hip roll (HR) were greater at SM pace, with no changes for shoulder roll (SR) or other elbow joint angle variables. Faster swimmers had higher SR and total SR at M pace whereas slower swimmers had greater maximum elbow angle at SM pace. Swimmers with greater velocity of the hand on the NP side showed greater SR. To achieve higher swimming v, swimmers should prolong the push phase, and reduce the elbow extension at the end of the push phase. Finally, an increase in TT within the range observed in the study, did not impair performance. For the isokinetic strength assessment of the shoulder, the relationship between performance and peak torque and work to body mass was explored, during abduction (Abd), adduction (Add), external (ER) and internal (IR) rotation. The participants completed multiple maximal concentric repetitions at 60°/sec and at 180°/sec. Results showed that most movements were positively correlated with performance, particularly on the P side. This may suggest that implementing regular strength training in such movements could increase the velocity of the hand and ultimately improve performance. For the asymmetry analysis the magnitude of differences between P and NP sides, as well as dominant (D) and non-dominant (ND) was assessed, and the symmetry index (SI) calculated for kinematic and isokinetic strength variables. The relationship between the magnitude of SI and performance was also investigated. No significant differences were found between NP and P sides, but all variables were significantly higher for the D side. Nevertheless, even though participants presented various levels of asymmetry, this did not impact overall performance. The results suggest that for competitive swimmers of national level, the magnitude of asymmetries observed did not negatively affect performance.