The main objective of this systematic review was to synthesize and evaluate the available literature regarding the associations between lower inter-limb asymmetry and endurance running performance in healthy populations. According to the risk of bias assessment, all included studies were of moderate to strong quality. To compare and evaluate the association of inter-limb asymmetry with running performance (and/or its determinants), it was necessary to differentiate between dimensions to quantify asymmetry. Therefore, this review addressed the link between functional, morphological, kinematic and kinetic inter-limb asymmetry with running performance, separately. It is important to note that the limited available literature on the topic alongside the high heterogeneity in terms of asymmetry assessments and running metrics, as well as the different mathematical equations for calculating asymmetry magnitude, made it difficult to compare studies and even impossible to conduct a meta-analysis. This discrepancy across test protocols and outcome measures resulted in inconsistent findings highlighting the task, metric, test occasion and individual specific nature of inter-limb asymmetry and its magnitude [7–12].
3.7 Functional asymmetry and running performance
Functional asymmetry was most commonly assessed using strength measures (e.g., isometric strength) [35, 46]. This is unsurprising since strength training-induced neuromuscular adaptations have been demonstrated to enhance running economy (i.e., 2–8%) as well as time trial performance and maximal sprint velocity in middle and long-distance runners [53–55]. Moreover, larger magnitudes of inter-limb strength asymmetry have also been associated with increased gait asymmetry (r = 0.44), indicating a transfer from functional assessments to sport-specific measures [56].
Blagrove et al. [35] examined the relationship between the magnitude of isometric muscle strength asymmetry (i.e., quarter-squat, hip extension and hip abduction) and running performance as well as running economy in male and female competitive middle- and long-distance runners. In general, this study observed group mean asymmetry values ranging between 4.6–8.4%, resulting in negligible associations between inter-limb strength asymmetry and running economy (r = − 0.02 to 0.13) and running performance (r = − 0.26 to 0.13). However, a larger magnitude of inter-limb asymmetry of 10% was found in hip abduction torque asymmetry for the female endurance runners. This inter-limb asymmetry in abduction strength was significantly positively correlated (r = 0.85) with running economy (i.e., energy cost in kJ.kg− 75.km− 1), indicating the potential negative impact of larger inter-limb asymmetry magnitudes on running performance as higher energy costs are detrimental to endurance performances [57]. Similarly, Tabor et al. [46] reported that reductions in the magnitude of the sum of muscle torque in knee and hip flexors and extensors were negatively correlated with running velocity in female middle-distance runners (β = -6.64). However, given the task-dependent nature of functional asymmetry [7], it may not be advisable to add together different strength measures to determine strength asymmetry. Therefore, this latter result should be interpreted with caution.
3.8 Morphological asymmetry and running performance
Previous research documented a negative relationship between asymmetry in various traits (e.g., nostrils and ears) and running performance in middle-distance runners [58]. However, the existing literature on the association between the magnitude of morphological asymmetry and (determinants of) running performance (e.g., metabolic cost) seems to be limited to only one study in untrained, occasional and skilled runners. A first important finding of this study conducted by Semanti at al. [47] was the moderate and positive correlation (r = 0.606) between anatomical asymmetry (i.e., side-to-side differences in volume of the lower limbs measured by magnetic resonance imaging) and dynamical asymmetry (i.e., body centre of mass displacements), indicating that runners with greater magnitudes of morphological asymmetry tend to exhibit more pronounced asymmetrical running patterns. Moreover, this latter study showed that training status moderated this relationship, as more experienced runners showed smaller magnitudes of dynamic asymmetry at higher running velocities compared to their untrained peers. However, this study did not report a significant correlation between anatomical asymmetry and metabolic cost. The authors speculated that certain physiological adaptations may compensate for the relatively small anatomical asymmetry magnitudes observed (i.e., 0.77 to 0.83%), regardless of training status [47].
Given the scarcity of literature on the link between morphological inter-limb asymmetry and running performance, it is difficult to draw clear conclusions. However, it is important to note that none of the studies included in this review addressed leg length discrepancies. This is probably because leg length differences are typically reported as an absolute difference between the right and left lower limb, rather than as a relative asymmetry score (i.e., expressed as a percentage). As such, larger absolute leg length differences (> 2 cm) have been reported to increase energy expenditure during walking, to increase oxygen consumption during submaximal running and to impair running economy [59, 60]. Moreover, and although this seems to be individual specific, absolute leg length differences have been positively associated with a more pronounced gait asymmetry (r = 0.29 to 0.51) [61, 62]. In contrast, leg length differences smaller than 1 cm do not appear to be significantly associated with running economy [48, 63]. These results support the notion that the magnitude of morphological asymmetry between the lower limbs could affect running performance [61].
3.9 Kinematic asymmetry and running performance
Despite the wide range of kinematic asymmetry magnitudes observed (i.e., 3–54%), the narrative review by Carpes et al. [15] concluded in 2010 that the available studies failed to establish significant relationships between kinematic asymmetry and running performance. Due to the recently growing interest on the topic, several studies attempted to investigate the association between kinematic inter-limb asymmetry and determinants of running performance as well as personal records. For instance, two studies included in the current systematic review indicated that inter-limb asymmetry in ground contact times (i.e., the average time each foot spends in contact with the ground while running) was correlated with impaired running economy (r = 0.808) and metabolic power (i.e., energy cost, based on O2 consumption and CO2 production [52]) (β = 0.78) [47, 48]. As discussed in a recent review by Moore et al. [64], there is still debate on whether short or long contact times are favourable in view of running performance. Whereas short ground contact times are suggested to impose a higher metabolic cost due to the need for faster force production [65, 66], longer ground contact times are suggested to increase the metabolic cost during the increased deceleration, resulting in a lengthened braking phase [67]. However, the findings in our review indicate that inter-limb asymmetry in ground contact times has a negative impact on energy depletion and running economy, potentially impairing running performance. Similarly, step time asymmetry (i.e., including both the ground contact time and the subsequent aerial time) was significantly positively correlated (β = 0.35) with metabolic power [47]. This result is consistent with previous research in which larger asymmetric step times were associated with increased metabolic power in walking [68]. Beck et al. [47] attributed these findings to reduced mechanical energy conservation in asymmetric step times, resulting in increased muscle mechanical work per step and an increased metabolic rate.
Studies investigating the association of asymmetry in trajectories of body centre of mass and (determinants of) running performance revealed equivocal results. Whilst asymmetry in body centre of mass displacements was moderately negatively related to mechanical efficiency (r = -0.66), no significant association was found with metabolic cost [45, 50]. Differences in duration of the running protocol have been postulated as a possible explanation for these discrepancies. Melo et al. [50] argued that longer distance protocols (e.g., 10 km) are more suitable for detecting kinematic asymmetry, which may not be evident in shorter running bouts. Moreover, variations in running experience [69], running intensity [70, 71] and muscle fatigue [72] could also explain these differences in findings.
Previous research showed that peak ankle dorsiflexion (i.e., maximal ankle dorsiflexion angle during stance phase) later in stance was positively related to running economy and thus possibly affecting running performance [73]. In the study by Stiffler-Joachim et al. [49], inter-limb asymmetry in peak ankle dorsiflexion was the only kinematic variable that was significantly and negatively correlated with within-season personal records (β = -6.1, CI: [-12.9, 0.7]. Every 1° increase in peak ankle dorsiflexion asymmetry was related to a 7.6 s decrease in the best running time on 8km for male and 6 km for female distance runners. Whilst the underlying mechanism for this finding is unclear, it should be noted that the magnitude for peak dorsiflexion was quantified as the absolute value of the inter-limb differences and not as a percentage. Lastly, Tabor et al. [46] demonstrated that swing phase asymmetry can impair running velocity in intermediate and advanced middle-distance runners. Given that a shorter support phase has been related to increased running velocity [67, 74], it seems plausible that asymmetry in the extension of the swing phase could impair running velocity [46].
3.10 Kinetic asymmetry and running performance
Stiffler-Joachim et al. [49] documented varying kinetic asymmetry percentages in National Collegiate Athletic Association (NCAA) Division I runners, ranging from 3% for peak vertical ground reaction force up to 20% for average vertical loading rate. Propulsive impulse asymmetry has been reported to be related to impaired race performance in distance running [49]. Given that metabolic cost during running is to a large extent determined by propulsive impulse [64], practitioners should not only consider to improve runners’ overall propulsive impulse but also aim to minimize side-to-side differences in this respect. In contrast, average vertical loading rate, braking impulse and peak vertical loading rate were not found to be significant predictors for race performance [64]. This could potentially be attributed to the fact that these associations were investigated in elite runners, who exhibited low overall asymmetry scores for these particular metrics.
Regarding the relationship between kinetic inter-limb asymmetry and determinants of running performance, results indicated that stance average vertical ground reaction force, peak propulsive ground reaction force and leg stiffness asymmetry were positively associated with metabolic power in recreational runners [47]. This suggests that more pronounced kinetic inter-limb asymmetry could eventually increase energy expenditure while running and thus potentially have an adverse effect on running performance. [47]. In contrast, peak ground reaction force asymmetry was not found to be significantly associated with metabolic power, demonstrating the variable nature of asymmetry [47]. This variability in asymmetry metrics and their associations with running performance was further emphasized in the study conducted by Mo et al. [51]. The latter study indicated that the association between inter-limb kinematic asymmetry and running performance highly depends on the velocity of the running test, the running experience of the participants and the parameter of interest assessed. Consistent with previous research, the magnitude of asymmetry not only varied considerably within kinetic variables, but also appeared to be more pronounced compared to kinematic variables [51, 75].
3.11 Limitations and strengths
Although this is the first systematic review to provide a holistic view on the available evidence concerning lower inter-limb asymmetry and the association with (determinants of) running performance in endurance runners, this research effort is not without limitations.
First, it is difficult to draw a definitive conclusion due to the high heterogeneity among the included studies. More specifically, the heterogeneity was evident in a variety of dimensions, (e.g., functional, morphological, kinematic or kinetic), assessments, equations and metrics being used to assess and express inter-limb asymmetry, as well as in a diversity of population characteristics (e.g., sex, age and/or training status of participants. Furthermore, caution is warranted when interpreting these results because of the scarcity of eligible studies, making comparisons between study results difficult and less robust. Moreover, only a quarter of the studies documenting Pearson’s or Spearman’s rank order correlations also reported an assessment of normality on their raw data. Since falsely (i.e., with non-normal data) using a Pearson’s correlation coefficient highly increases type I error rates, justification for the use of parametric statistics by means of normality tests (e.g., Shapiro-Wilk test) is essential [76]. Lastly, only one study reported the reliability of the asymmetry metric of interest. Given the inherently high variably nature of inter-limb asymmetry [13], a good test-rest as well as inter-rater reliability is necessary to ensure the quality of the data.
3.12 Directions for future research
By analogy with the work of Afonso et al. [13], the synopsis of literature on the topic presented in the current systematic review is important to identify a research agenda highlighting some key areas for future research on inter-limb asymmetry in endurance runners (see Fig. 2 for an illustrative overview):
-
In literature, inter-limb asymmetry is often not reported using sport-specific and field-based assessments in endurance runners. Whilst functional asymmetry is generally measured using maximal (isometric) strength, (repeated) hop tests are presumed to have a greater ecological validity for assessing inter-limb asymmetry in runners due to their ability to measure various facets related to the stretch and shortening cycle [77]. Notably, storing and returning mechanical energy in the process of elastic energy utilization plays a key role in the metabolic energy-saving mechanism, and consequently running economy [78]. In this regard, leg stiffness (i.e., resistance to deformation of the limb) and reactive strength (i.e., the ability to effectively use the stretch and shortening cycle as well as the energy produced by the muscle-tendon complex) have been proposed to be important neuromuscular factors contributing to the elastic energy utilization [64, 79, 80]. Given that these factors can be measured using (repeated) hop tests and rebound jump protocols, moderate to large correlations between a countermovement jump and running economy have been previously reported [60]. Hence, future research should consider investigating the effect of functional inter-limb asymmetry in leg stiffness and reactive strength using unilateral (repeated) hop tests. Moreover, the use of sport-specific, valid and reliable field-based assessments of functional asymmetry is needed to enhance the ecological validity and applicability among practitioners.
-
Disparities in asymmetry outcomes and magnitudes between different types of runners underscore the necessity for practitioners to account for inter-individual differences. Variables such as type of running (e.g., track versus road running), training status of runners (e.g., trained versus untrained) and injury history of runners (i.e., injured versus non-injured) should be considered when assessing lower inter-limb asymmetries [51, 81–83]. Although larger magnitudes of inter-limb asymmetry are expected in novice endurance runners than in elite endurance runners [51], the results of the present review indicate that – based on the included studies – 68% of research has been conducted in competitive runners and only 22% in recreational runners and 10% in novice runners. Therefore, addressing a more diverse range of running populations in terms of training status, age and sex, while acknowledging the high inter- and intra-variability, is warranted.
-
The direction of asymmetry is highly variable between tasks and between test occasions [84, 85]. For instance, a distance runner may favour his right limb on a first test occasion whilst favouring his left limb on a second test occasion. Given that asymmetry is a ratio metric, reporting kappa values is highly recommended to assess differences in the direction of asymmetry between different tasks and/or test occasions.
-
Several factors relating to test protocols, such as running velocity, test intensity or fatigue, will likely induce intra-individual differences in asymmetry [51, 70, 86]. In addition, also environmental factors such as the running underground, air humidity and ambient temperature may possibly lead to different asymmetry magnitudes and/or running performances. This accentuates the need for a standardized approach under stable conditions when evaluating inter-limb asymmetry.
-
Recognizing the highly variable nature of inter-limb asymmetry, researchers are urged to report the reliability of their tests and related outcome measures (e.g., test-retest or inter-rater reliability) to mitigate the impact of fluctuations on asymmetry due to test errors. A standardized approach for expressing asymmetry magnitude across studies is also needed, preferably using “stronger” and “weaker” limb instead of “right” and left “limb” [87].
** PLEASE INSERT FIGURE 2 ABOUT HERE **