Inter-limb asymmetry: longitudinal monitoring and associations with speed and change of direction speed in elite academy soccer players

Inter-limb asymmetry has been a popular topic of investigation in recent years, with the majority of studies reporting the prevalence of asymmetry during different test protocols in athlete and non-athlete populations, and between genders. However, such information does little to inform practitioners as to whether asymmetry should be of any concern. To more fully examine the current body of evidence pertaining to asymmetry, a systematic review was completed as part of this thesis, in an attempt to determine the association between asymmetry and measures of athletic performance. Results showed that asymmetry was often associated with reduced athletic performance, especially when measured during the sport-specific task.
The findings of the systematic review also highlighted some important considerations for future research. Specifically, it was observed that the association between asymmetry and measures of athletic performance has been reported at single time points only, with a distinct lack of data to examine how asymmetry varies over time and if a change in asymmetry corresponds to changes in athletic performance. In addition, limited studies had investigated the link between asymmetry and fatigue. This information would help practitioners by determining if measurement of asymmetry is useful as part of the ongoing monitoring process.
The aim of study 1 was to use the unilateral isometric squat, unilateral countermovement (CMJ) and unilateral drop jumps (DJ), in a test-retest design, to determine test reliability, the magnitude of asymmetry for both the mean and best scores, and the consistency of asymmetry direction. Within and between-session reliability showed good to excellent relative reliability for all tests and metrics (intraclass correlation coefficient [ICC] = 0.81-0.98) and for the most part, acceptable absolute reliability (coefficient of variation [CV] = 2.3-13.7%). When calculated from the best trial, significant differences in asymmetry were present for impulse at 0.3 s during the isometric squat (p = 0.04; effect size [ES] = -0.60) and for ground contact time during the DJ (p = 0.04; ES = 0.54). No significant differences in asymmetry were evident when calculating from mean scores. The level of agreement indicating how consistently asymmetry favoured the same limb between test sessions was fair to substantial for the isometric squat, moderate to substantial for the CMJ, and fair to moderate for the DJ. Given the test-retest design of this study, it was concluded that most metrics in each test are reliable for detecting asymmetry, although the isometric squat did show higher within-session CV values than the jump tests. In addition, given no training intervention was conducted, it is suggested that the average of all trials was a more appropriate method to calculate asymmetry.
Study 2 carried forward the unilateral CMJ and DJ tests, in addition to 5, 10, 30 m and 505 change of direction speed (CODS) tests to track seasonal variation in performance and asymmetry during pre, mid and end of season time points in a group of 18 under-23 elite academy soccer players. The unilateral isometric squat was not carried forward for the remainder of testing procedures due to time constraints in a professional soccer club setting. Associations between jumping asymmetry and speed/CODS tests were examined at each time point, and changes in asymmetry and changes in speed/CODS performance were also examined through associative analysis. When assessing the relationship between asymmetry and performance tests, no meaningful correlations were evident at pre or mid-season (ρ = -0.32 to 0.37). However, at the end of season, significant relationships were found between DJ height asymmetry and 5 m (ρ = 0.63; p < 0.008), 10 m (ρ = 0.62; p < 0.008) and 505 on the right limb (ρ = 0.65; p < 0.008). When assessing relationships between changes in asymmetry and changes in performance tasks, no significant relationships were found between changes in asymmetry and changes in speed or CODS performance (ρ = -0.44 to 0.56). A median split technique was also used to create high and low asymmetry groups for pre, mid, and end of season, respectively. At all time points, significant differences in asymmetry were found between groups for all jump metrics (p < 0.01). For speed/CODS tests, significant differences were reported at the end-season time point between groups when using DJ height asymmetry for 10 m (high asymmetry = 1.84 ± 0.13; low asymmetry = 1.72 ± 0.07; p < 0.05; ES = -1.15), 505 left (high asymmetry = 2.26 ± 0.05; low asymmetry = 2.19 ± 0.09; p < 0.05; ES = -0.96) and 505 right (high asymmetry = 2.30 ± 0.11; low asymmetry = 2.18 ± 0.05; p < 0.01; ES = -1.40). No other significant differences in speed or CODS were present between groups. Despite these findings at the end of the season suggesting significant relationships, the cumulative results of this study and specifically the inconsistencies shown, indicate that asymmetry is largely independent of speed and CODS, both at single time points and when monitored over time.
Seasonal variation of jump scores and asymmetry were also reported. Unilateral jump data showed significant reductions in CMJ height and concentric impulse at mid-season on both limbs and for peak force on the left limb only. DJ height showed no meaningful changes between time points; however, ground contact time and reactive strength index (RSI) showed significant improvements at the end of the season, compared to both previous time points. When monitoring asymmetry, the group mean value showed no significant differences throughout the season with corresponding trivial to small ES (range = -0.60 to 0.55). However, poor to substantial levels of agreement were reported across the season for the direction of asymmetry, in both jump tests. These data indicate that when monitoring the magnitude of asymmetry alone, group mean values do not reflect the potential inconsistencies in limb dominance over time. By also monitoring the direction of asymmetry, this more accurately highlights its task and variable nature, and allows practitioners to account for inherent changes in limb dominance throughout the season.
The final experimental chapter (study 3), examined the effects of acute fatigue on asymmetry in elite under-18 male soccer players. A repeated measures design was used, where unilateral CMJ and unilateral DJ tests were performed before and immediately after five soccer matches. Global positioning system (GPS) data were also collected for each match to assess relationships and interactions between asymmetry and in-game soccer actions. Unilateral CMJ height and concentric impulse showed significant reductions post-matches (p < 0.01; ES: -0.67 to -0.69), but peak force did not (ES: -0.05 to -0.13). DJ height and reactive strength also showed significant reductions post-matches (p < 0.01; ES: -0.39 to -0.58). No significant reductions in asymmetry were present at the group level, but individual responses were highly variable. Match related variables were almost always not associated with asymmetry. However, significant correlations were evident between post-match reactive strength asymmetry and relative high speed running only (ρ = 0.44; p < 0.008). These findings indicate that data derived from unilateral jump tests are more sensitive than asymmetry scores in their ability to detect a real change immediately post soccer competition. Thus, practitioners should be cautious about using asymmetry as a marker to determine acute fatigue following soccer match-play.
In conclusion, the findings from this thesis suggest that: i) it may be more favourable to calculate asymmetry scores from an average of all trials, rather than from the best trial; ii) monitoring the group mean value (magnitude) disguises the inherent variability associated with asymmetry; iii) monitoring the direction of asymmetry allows practitioners to account for individual variation; iv) although relationships between asymmetry and speed/CODS/in-game soccer actions do exist, they are not consistent over time and in-response to acute fatigue from soccer match-play. Cumulatively, and given the highly varied response of asymmetry, individual monitoring is recommended but further research is required to more fully understand the usefulness of this approach. Specifically, relationships with injury and a more mechanistic approach to understanding why asymmetry is present, is suggested.