This study examined the relationships of CS performance in soccer with LS, COD and CMJ from the U-13 to Senior category soccer players in both the good and weak sides, as well as the difference in performance across age categories. Moderate to large correlations were found in most relationships between the CS test and LS (5 and 20 meters), COD and CMJ performance, irrespective of age category. Moreover the findings showed very large differences in the CS performance (>2.0 Cohen’s d), on both good and weak sides, between all pairs of age categories, with exception of U-15 and U-17 categories, that presented a moderated difference.
There were considerable differences in CS performance between the age categories, both on the good and the weak side. From a developmental perspective, it is plausible to assume that advancing age, accompanied by the improvement of neuromuscular function and mechanical properties as well as the accumulation of experience, might explain the observed improvements in CS performance [22]. We cannot clearly justify the smaller magnitude of differences between the U-15 and U-17 categories, but it is well known that physical performance in these categories can be influenced by maturational changes. It would be expected large differences in the timing and tempo of maturation between players in these categories. However, the selection of early maturing players particularly in the U-15 category might have caused them to differ little in their physical and physiological characteristics from the U-17 categories, which are no longer strongly influenced by maturation, as most players in this category have already passed peak height velocity. There is evidence that the magnitude of the effect of relative age among young soccer players in the U-15 category may be greater than in the U-17 [23]. Nonetheless, U-17 were still moderately better than U-15 players in CS. Future research should investigate how maturational aspects can influence CS performance, especially in players surrounding the peak height velocity age.
As found in a study with the U-20 category, the correlation between CS and LS performances tended to be higher when distance travelled was longer (i.e. from 5- to 20-m) [14]. For the CS and 5-m LS relationship, it could be seen that the correlation from the U-13 to Senior category for the dominant leg remained stable and non-significant (r from 0.30 to 0.34), with the exception of U-15, where a significant relationship of 0.53 was observed. For the non-dominant leg, the relationship was significant in the U-15 and Senior categories, but both with moderate magnitude (r from 0.46 to 0.56). This suggests that CS performance in these categories depends to a small extent on the player's ability to linearly accelerate (<30% of shared variance).
When we examined the relationship between CS and 20-m LS, a consistent correlation of moderate to high magnitude was observed in CSWS, from U-13 to senior, with a slightly lower magnitude for the U-17 category. Furthermore, the relationship between CSGS and 20-m LS was greater in Seniors than in the other age categories. As our participants in the Senior category are of a higher competitive level, unlike other samples involving semi-professional athletes, it is apparent that more qualified players are more efficient at stabilizing the joints in the frontal plane, allowing them to generate the adequate centripetal force to achieve high speeds in both linear and curvilinear sprinting [14,24,25]. On the other hand, previous studies with U-20, U-17, and U-15 soccer players have shown that these relationships did show a consistent trend in any of the categories [13,14,16]. For example, Filter-Ruger et al. (2022) showed that the CS-LS relationship decreased with age from the U-15 group (r = 0.75 and 0.76) to the U-20 group (r = 0.27 and 0.41) for the weak and good sides, respectively. This is consistent with the idea that as sport specialization increases, abilities that share similar determining factors become more independent and tend to become unrelated. Future studies are necessary to determine whether older and more highly trained players have LS and CS performances more or less related to each other than younger and less specialized players.
The observed correlations between CS and COD performance suggest an increase in the strength of the relationships from U-13 to U-17 (r from 0.20 to 0.65), but they cease to be significant in Senior. However, these correlations are generally lower than those observed between CS and LS [26], with shared variance of ~24% and ~42% for the good and weak sides, respectively. These results are similar to those observed in other studies showing that while LS is more strongly related to CS [13,26], linearly faster athletes are not necessarily faster in curves or in COD trajectories due to differences in kinematic and neuromuscular demands [15]. Our results add to the literature showing different patterns in the magnitude of associations between CS and COD, depending on age category. Future research should seek to understand the mechanisms underlying CS performance in soccer players from a biomechanical and developmental perspective.
The relationship between CS and CMJ has received little attention in the literature. Loturco et al. (2020a) observed relationships of 0.57 (CSGS) and 0.61 (CSWS) in U-20 athletes, while Kobal et al. (2021) found correlations of 0.56 and 0.53 for CSGS and CSWS, respectively, in an elite women's soccer group. Despite the similar correlations found in our study compared to the others (with the exception of U-17, in which the correlation was not significant), it is possible to observe a reduction in the magnitude of the associations from U-13 to U-17, although the relationship between CS and CMJ is still moderate in seniors. The relationship between jumping and sprinting performance has been justified from a mechanical perspective [27]. To achieve high performance in both capacities, high demands on vertical force production are required (i.e., from the beginning of the jumping movement to the take-off and during the acceleration phase of the sprint) to project the body in space and achieve high sprint speed [27]. Interestingly, improvements in jumping ability might be at least partially “transferred” to CS performance at the youngest age, but this hypothesis needs to be tested in future intervention studies.
Overall, the strength of the correlations tended to be higher in the CSWS with the physical attributes assessed (i.e., LS, COD, and CMJ). Running in curvature brings some biomechanical differences compared to straight running, for example, the shortening of the stride length as a biomechanical strategy to maintain the body in a more upright posture [10]. The curvilinear motion prolongs the neuromuscular activity in the outside leg [10], which is probably explained by the necessity to generate ground reaction forces to accelerate the body towards the curvature [28]. By scrutinizing the assessments of our players, it was possible to notice that the fastest CS performance was for most of the athletes (65 out of 121 athletes, 54%) in the non-dominant side, which was confirmed for almost all categories (U-13: 64%, U-17: 61%, and Senior: 53%), except for the U-15 (40%). This implies that the dominant side was the outside limb and therefore the major contributor to the curvilinear performance, as it has been previously demonstrated [10,28].
Despite the important results of our study, some limitations can be raised. This study was conducted in only one club, which limits the generalization of the results for athletes that compete in higher or lower levels. Although it is important to emphasize that the same training “philosophies” are followed (i.e., from U-13 to senior level) in athletes from the same club. Nonetheless, this is the first study to examine the relationship between CS performance and other physical abilities from the younger ages to the Senior category, which can serve as a basis for monitoring the related variables at different age groups. Future studies are desirable to investigate the determinants of CS performance in each category, involving longitudinal design and other relevant measurements (e.g., maturation and biomechanical analysis), as well as to test different training strategies to develop CS performance.