Age‐Related Differences in Jumping and Sprinting Performance and Force Production Capacities in Young Soccer Players

ABSTRACT

The purpose of this study was to evaluate the evolution of jump and sprint force‐production capacities with maturation in young soccer players. One hundred sixteen young elite male soccer players aged 11–17 years were assigned to six different groups according to their maturity status. The force–velocity (F–V) profiles in jumping and sprinting performances were compared among groups. The results show a significant, progressive, and linear improvement in the majority of F–V profile determinants of jumping and sprinting performances in parallel with maturation. When comparisons were carried out between consecutive groups, significant differences between PHV−1 and PHV groups in H_max (p < 0.05), H₆₀ (p < 0.05), VT‐F0 (p < 0.05), and VT‐P_max (p < 0.05) were observed but disappeared when these variables were expressed relative to body mass. Regarding the F–V profile determinants of sprinting performance, we observed significantly shorter sprint times in PHV compared to PHV−1 and in PHV+2 in comparison with the PHV+1 group. These between‐group differences were associated with significant greater HZT‐F0 kg⁻¹ (p < 0.05) and HZT‐P_max kg⁻¹ (p < 0.001). However, significantly more negative value of HZT‐Sfv kg⁻¹ was observed in PHV−1 compared to PHV−2 (p < 0.05), in PHV+1 in comparison with PHV (p < 0.05) and in PHV+3 when compared to the PHV+2 group (p < 0.05). Thus, these findings revealed an alternation of accelerated periods of development in force and velocity qualities, expressing sprinting performance, with maturation during the physical development process of young soccer players. Trainers and coaches should be aware of the distinct mechanical characteristics related to each maturity stage and design specific programs accordingly.

Summary.

• The biological age‐related evolution of mechanical components associated with jumping and sprinting performance is different during the maturation process.

• The changes in jumping performance and its vertical mechanical characteristics are mainly due to the growth and maturation‐related body mass gain.

• The mechanical components associated with sprinting performance reveal an alternation of accelerated periods of development in force and velocity qualities with maturation during the physical development process of young soccer players.

• A better understanding of the mechanical characteristics' evolution specific to each maturation status would be useful to optimize the talent selection process. This should also be considered by practitioners when designing appropriate and individualized programs to optimize training‐related adaptations in young soccer players.

1. Introduction

Despite their short duration compared to total match time, jumping and sprinting performances are amongst the most important and decisive athletic motor skills in youth soccer (Castagna et al. 2003). Furthermore, they constitute useful indicators of athletic performance and talent‐identification markers discriminating elite from nonelite players (Stølen et al. 2005).

In recent years, there has been an increasing interest to assess athletes' force production capacities in sprinting and jumping based on the force–velocity (F–V) and power–velocity relationships (Jiménez‐Reyes et al. 2017; Samozino et al. 2008, 2016) due to their strong link to physical performance in sports, including soccer (Jiménez‐Reyes et al. 2018, 2019). These relationships are usually established to assess the athlete's mechanical capabilities profile (Samozino et al. 2014, 2016). Several devices have been commonly used to evaluate them such as force platform, cycle ergometer, and running treadmill (Armstrong et al. 2019; Cormie et al. 2007). Recently, simple methods for evaluating force, velocity, and power output during the vertical jump test against different loads (Morin and Samozino 2016) and/or maximal horizontal sprint test have been validated by Samozino et al. (2008), (2016) shown their practicality in field conditions and ability to overcome the multiple constraints related to the abovementioned devices (i.e., high cost, complexity of use, and time consuming related to data analysis). The principal mechanical determinants of the F–V profile typically studied comprised the maximal theoretical force (F0), maximal theoretical velocity (V0), slope of the F–V relationship (Sfv), and maximal power output (P_max) (Morin et al. 2011; Samozino et al. 2016). Assessing these mechanical determinants of jump and sprint performance has recently been shown to be reliable in youth (Runacres et al. 2019) and has already been carried out in young soccer players (Baumgart et al. 2018; Fernández‐Galván et al. 2021, 2022). They can be expressed as absolute or relative to body mass in order to control for body mass gain influence with age and maturation (Malina, Bouchard, et al. 2004).

From childhood to early adulthood, the locomotor system of young soccer players undergoes maturity‐related morphological and neural changes that influence its motor function and, consequently, mechanical output and physical performance (Malina, Bouchard, et al. 2004). For instance, it has been shown that both sprinting and jumping performance change through the maturation process in young soccer players (Lloyd et al. 2015; Mendez‐Villanueva et al. 2011). Despite their importance in understanding the evolution of jumping and sprinting performance in relation to physical and physiological development, there is little research regarding biological age‐related variations in horizontal force production capacities (i.e., sprinting F–V profile outputs) underlying these performance changes throughout the maturation process in this cohort (Fernández‐Galván et al. 2021, 2022). The first study of Fernández‐Galván et al. 2021 aimed to examine and compare the relationship between sprinting and jumping F–V parameters with either maturity offset or chronological age. The results showed similar correlations between the mechanical components of the F–V profile of sprint and jump performance and each of the maturity offset and chronological age. The second one (Fernández‐Galván et al. 2022) focused on analyzing the influence of maturation on the sprint F–V profile in young soccer players by comparing three groups before, around, and after the peak height velocity (PHV). The authors pointed out that variables related to strength quality develop more from the pre‐PHV to mid‐PHV stage, whereas those related to speed qualities improve to a greater extent from the mid‐PHV to post‐PHV. Although the interesting results of these recent studies, none of them has focused on a macroscopic monitoring of the evolution of F–V profile determinants in jumping and sprinting, stage‐by‐stage, throughout the maturation process. Given the growing and substantial interest in the most accurate and effective training methods for each stage of football player maturity, further research is needed to accurately identify the specific maturation‐related adaptations of F–V mechanical determinants leading to improved jumping and sprinting performance across different maturation stages. Such information could provide more comprehensive and meaningful data and reference values to be used for better talent‐identification and training monitoring.

Previous studies have demonstrated strong relationships between sprinting performance with both force and velocity capacities (Buchheit et al. 2014; Fernández‐Galván et al. 2021, 2022), whereas jumping performance seems to be specifically reliant on force production capacity (Fernández‐Galván et al. 2021). Based on the literature, the evolution of force and velocity qualities with age and maturation is different. Force production capacity increases progressively and linearly during the maturational process, with maximal peak gains around and shortly after peak height velocity (PHV) (G. Beunen and Malina 1988; Malina, Bouchard, et al. 2004). However, velocity qualities seem to develop substantially at post‐PHV (Moran et al. 2017), coinciding with important changes in musculotendinous tissue (Lloyd et al. 2013) along with an increased efficiency of the stretch‐shortening cycle (Lloyd et al. 2020).

Assessing each mechanical determinant of the vertical and horizontal F–V relationships during the maturational process could provide more information on the biological age‐related evolution of these determinants associated with jumping (Fernández‐Galván et al. 2021) and sprinting (Armstrong et al. 2019; Fernández‐Galván et al. 2022). This could allow a better understanding of how each of the two qualities of force and velocity evolve independently but also and predominantly jointly (F–V relationship or F–V profile) with maturation. Indeed, previous studies have shown that optimal balance between the force and velocity output (optimal F–V profile) would allow a good expression of power, which would determine jump and sprint performances (Jiménez‐Reyes et al. 2017; Morin and Samozino 2016). However, the different evolution of force and velocity qualities during growth could result in different profiles, sometimes out of balance with the optimal force–velocity profile, which would partly explain the difference in power output capability according to age and maturation status. Thus, a better understanding of the evolution of this profile could be very useful to identify more precisely strengths and weaknesses of each stage across the growth, and subsequently, the most accurate training methods to implement in order to optimally improve the F–V profile determinants of jumping and sprinting performances for each maturity status.

Therefore, the current study aimed to investigate the biological age‐related evolutions of the mechanical determinants of jump and sprint F–V profiles in elite male young soccer players. We hypothesized a progressive linear increase in jumping performance and its underlying mechanical determinants with maturation, mainly due to biological age‐related improvements in body mass and force production capacity. However, improved sprinting performance with maturation was assumed to rely on different accelerated periods of development with greater spurts of gains in force around PHV and in velocity after PHV.

2. Methods

2.1. Study Design and Participants

A comparative experimental design comprising several categories of young soccer players was set up to assess differences related to biological development in sprinting and jumping performance along with the associated mechanical determinants. One hundred thirty‐two elite male young soccer players belonging to the same first‐category soccer academy from 6 different age categories were enrolled in this study. They usually followed training sessions, which included a technical and tactical skills development program, as well as exercises aimed at general physical improvement, without relying on any one particular physical quality. Just before the end of the previous season, all the assessed athletes were instructed to refrain from any type of physical activity during the off‐season, which lasted for 1 month, regardless of their age category. The measurements were then taken at the beginning of the annual training season (during the month of July, after only 2 weeks of training) to limit differences in the training status between players. Testing sessions were completed at the same time of the day (from 9 to 11 a.m.), inside the gym for jumping and outside on artificial turf for sprinting performance. After screening, 16 players were excluded because they were not injury free during the 6 months before the study (88% retention). As a result, 116 players (aged 10.7–17.8 years) participated in this study and were assigned, according to their maturity status, to one of the following six biological age groups: PHV−2: (−3 years < predicted age from/to peak height velocity [PAPHV] ≤ −2 years), PHV−1: (−2 years < PAPHV ≤ −0.5 years), PHV: (−0.5 years < PAPHV ≤ +0.5 years), PHV+1: (+0.5 years < PAPHV < +2 years), PHV+2: (+2 years ≤ PAPHV < +3 years), and PHV+3 group: (+3 years ≤ PAPHV < +4 years). Athletes and their parents were informed of the benefits and possible risks of the investigation and agreed to participate in the current research and parental consents were signed. The experimental protocol was fully approved by the local Ethics Committee (CPP SUD N° 0438/2022).

2.2. Testing Procedures

Before the study and prior to tests, all players completed a 2‐week orientation period (two sessions/week) to become familiar with the form and technique of each performance test and to minimize the learning effect. During this period, participants were allowed to practice as many trials of the test protocols as needed in order to demonstrate consistent technical execution. Thereafter, athletes were asked to avoid strenuous activity during the 48 h preceding the test session.

2.2.1. Anthropometrics

On the first testing day, maturity estimation and anthropometric measurements were conducted at the same time for all the players. Standing and seated height (centimeters) were measured using a stadiometer (SC126; Holtain, CrymychWales, UK), whereas body mass (kilograms) was measured using a balance beam scale (Seca 700; Seca, Hamburg, Germany). These data, in addition to date of birth, were incorporated into a sex‐specific regression equation to predict years from/to PHV as a measure of maturity offset. The equation has previously been validated for boys with standard error of estimates reported as 0.57 years (Equation 1) (Mirwald et al. 2002). This assessment is the most commonly used indicator of somatic maturation in the sports field, although potential limitations associated with the offset equations have been discussed (Moore et al. 2015; Malina et al. 2006).

$$\text{Maturity}\text{offset}=-9.236+\left(0.0002708\times \text{leg}\text{length}\times \text{sitting}\text{height}\right)+\left(-0.001663\times \text{age}\times \text{leg}\text{length}\right)+\left(0.007216\times \text{age}\times \text{sitting}\text{height}\right)+(0.2292\times \text{body}\text{mass}\text{by}\text{height}\text{ratio})$$ (1)

Other measures including lower limb length (from iliac crest to tip of the toes) and initial height (distance between the ground and the right lower limb iliac crest at an approximately 90° knee angle squat position) were also taken to determine the push‐off distance (Samozino et al. 2008). The same experimenter conducted all measurements.

2.2.2. Jumping Assessment

After several warm‐ups, including a 5 min jog at a self‐selected comfortable pace, a 5min series of dynamic stretching (hip flexion/extension, hip abduction/adduction, and butt kicks), and some unloaded jumps and accelerations, participants performed maximal squat jumps (SJs) under different loading conditions (without load [H_max] and against two extra loads corresponding to 30% [H₃₀] and 60% [H₆₀] of body mass in a randomized order). Before each jump, participants were instructed to stand up straight with their hands positioned on their hips for unloaded conditions and on the free‐weight bar (7 kg) placed on their shoulders during loaded jumps. Hand positions were to be maintained throughout the movements. Participants were asked to maintain a crouch position (flexed knee at approximately 90°) for 2 s before jumping as fast as possible to reach their maximum height. Countermovement was prohibited and carefully checked for each trial. Participants were asked to land with their lower limbs extended and maximal foot plantar flexion. If any instructions were not respected, the trial was repeated. Two valid trials were performed with each load, with a 3 min recovery between attempts and 5 min between load conditions to avoid effects of fatigue. All trials were recorded via an iPad 11 Pro (Apple, USA) with a high‐speed camera at a quality of 1080p positioned at frontal plane at approximately 2.5 m from the participant. Recorded measurements were analyzed using My Jump 2 (v.5.0.5), designed for high‐speed video analysis (240 fps). Best trial for each load condition was retained for the F–V relationship assessment through a previously validated linear computation model (Samozino et al. 2008). The intraclass correlation coefficients (ICC) with 95% confidence interval [lower 95% CI–upper 95% CI] and typical error of measurement (TEM) were 0.898 [0.846–0.933] and TEM = 0.0038 for unloaded SJ, 0.880 [0.819–0.921] and TEM = 0.0044 for 30% of body mass SJ, and 0.860 [0.790–0.908] and TEM = 0.0037 for 60% of body mass SJ. Mechanical determinants extracted from the F–V relationship included maximal theoretical vertical force at null velocity (VT‐F0, representing force production capacities at low velocities in N or N kg⁻¹), maximal theoretical lower‐limb extension velocity until which force can be produced (VT‐V0, representing force production capacity at high velocities in m s⁻¹), maximal vertical power output (VT‐P_max in W or W kg⁻¹), and the slope of the linear F–V relationship (VT‐Sfv, in N s m⁻¹, or N s m⁻¹ kg⁻¹) (Samozino et al. 2008).

2.2.3. Sprinting Assessment

On the second day after completing the same warm‐up exercises as the first day, athletes started from a three‐point crouching position (staggered‐stance) with their right hand on the track and performed two maximal effort 30 m linear sprints, with 5 min rest in between. No starting signal was given, and athletes were allowed to begin their sprints in their own time. Participants were instructed to refrain from backward movement prior to the start, progressive acceleration (maximal effort was to be maintained throughout the sprint) and deceleration before crossing the last 30 m marker.

The same iPad 11 Pro used for SJs was mounted to a tripod and positioned at a distance of 10 m perpendicularly to the sprint direction directly in front of the 15 m marker to record each sprint test. Videos were analyzed using MySprint to provide split times in milliseconds at 5, 10, 15, 20, 25, and 30 m. Intraclass correlation coefficients (ICC) with 95% confidence interval [lower 95% CI–upper 95% CI] and typical error of measurement (TEM) were the following for 5 m (ICC = 0.696 [0.578–0.786] and TEM = 0.021), 10 m (ICC = 0.870 [0.812–0.911] and TEM = 0.014), 15 m (ICC = 0.926 [0.891–0.950] and TEM = 0.011), 20 m (ICC = 0.871 [0.812–0.912] and TEM = 0.019), 25 m (ICC = 0.951 [0.928–0.967] and TEM = 0.011), and 30 m (ICC = 0.955 [0.933–0.970] and TEM = 0.013). According to the method proposed by Samozino and colleagues (Samozino et al. 2016), split times from the best attempt along with anthropometric characteristics and weather conditions (barometric pressure [mmHg], air temperature [°C], and wind velocity [mph]) can be used to assess the F–V relationship and associated mechanical determinants (maximal theoretical horizontal force [HZT‐F_0, representing horizontal force production capacities at low velocities, in N or N kg⁻¹], theoretical maximal running velocity until which horizontal force can be produced [HZT‐V_0, representing horizontal force production capacities at high velocities, in m s⁻¹], maximal propulsive power output [HZT‐P_max, in W or W kg⁻¹], and Sfv [in N s m⁻¹ or N s m⁻¹ kg⁻¹]). These measurements were provided using an Excel computation model validated in a previous study (Morin and Samozino 2016) based on the reduction of velocity through a linear load–velocity relationship.

2.3. Statistics

Data are presented as means and standard deviations (SDs) in the text, tables, and figure. The reliability of jump and sprint measures was assessed using the ICC. ICC values lower than 0.5, between 0.5 and 0.75, between 0.75 and 0.90, and > 0.90 were considered to have poor, moderate, good, and excellent reliability, respectively (Koo and Li 2016). The typical error of measurement (TEM) was computed as an absolute reliability index (Weir 2005). TEM values smaller than 10% of the average test–retest value were considered to indicate excellent absolute reliability. After normal distribution of data examination using the Shapiro–Wilk test, a one‐way between‐groups analysis of variance (ANOVA) was used to compare jump and sprint performances and their different mechanical determinants between groups. If a statistically significant group effect was detected, Bonferroni post hoc test was performed to locate the pairwise differences between mean values. For each test, the significance level was set to p < 0.05. Effect sizes between consecutive groups were calculated using Cohen's d. Given the small size and the unequal groups in the current study, we applied a correction by Hedge's g to avoid a biased estimation of the population ES (Hedges and Olkin 1985). According to Cohen, ES can be classified as small (0.00 ≤ d ≤ 0.49), medium (0.50 ≤ d ≤ 0.79), and large (d ≥ 0.80). Data analysis was performed using Statistica (version 8.0, StatSoft Inc., Tulsa, OK).

3. Results

3.1. Anthropometrics

All anthropometric characteristics increased significantly in parallel with age and maturation.

Significant greater height, sitting height, and body mass were observed in consecutive groups, namely in PHV−1, PHV, and PHV+1 compared to PHV−2 (p < 0.01), PHV−1 (p < 0.001), and PHV (p < 0.01) groups, respectively (Table 1). Leg length and training experience significantly increased in PHV−1 and PHV in comparison with PHV−2 (p < 0.05) and PHV−1 (p < 0.05) groups, respectively. However, no significant differences in anthropometric characteristics were found between PHV+1–PHV+2 and PHV+2–PHV+3 groups (Table 1).

TABLE 1.

Mean and SD (±) for anthropometric measurements by the biological age group in elite youth soccer players.

	PHV−2 (n = 19)	PHV−1 (n = 20)	PHV (n = 18)	PHV+1 (n = 20)	PHV+2 (n = 19)	PHV+3 (n = 20)
Chronological age (year)	11.2 ± 0.3	13.2 ± 0.3***	14.2 ± 0.2***	15.4 ± 0.2***	16.3 ± 0.3***	17.4 ± 0.4***
Predicted from/to APHV (year)	−2.4 ± 0.4	−1.0 ± 0.4***	0.1 ± 0.3***	1.5 ± 0.6***	2.3 ± 0.5***	3.3 ± 0.6***
Height (cm)	149.0 ± 6.2	156.1 ± 4.4**	165.9 ± 3.3***	174.3 ± 5.3***	177.4 ± 4.8	179.8 ± 5.7
Leg length (cm)	95 0.5 ± 4.6	99 0.7 ± 4.0*	105.6 ± 3.6***	107.4 ± 3.8	108.8 ± 3.5	111.1 ± 3.6
Sitting height (cm)	75.3 ± 3.3	79.1 ± 2.4**	84.6 ± 2.1***	90.2 ± 3.7***	92.1 ± 2.6	93.1 ± 3.5
Leg length at 90° (cm)	63.2 ± 4.4	65.9 ± 4.2	71.4 ± 7.4	70.3 ± 5.3	70.5 ± 5.0	74.9 ± 4.8
Body mass (kg)	39.6 ± 4.9	46.2 ± 4.9*	53.5 ± 4.4*	69.7 ± 4.7**	68.4 ± 7.7	74.8 ± 9.1
BMI (kg m−2)	17.8 ± 1.5	19.0 ± 1.3	19.4 ± 1.2	20.9 ± 2.4	21.7 ± 1.8	23.1 ± 2.3
Training experience (year)	4.5 ± 1.3	6.4 ± 1.5**	8.2 ± 1.0*	8.4 ± 1.3	8.9 ± 2.0	10.4 ± 1.7

Note: Significant difference with respect to the previous biological age group (*p < 0.05, **p < 0.01, and ***p < 0.001).

Abbreviation: APHV, age from/to peak height velocity.

3.2. Jump Performance

It is worth noting that the PHV−2 group was excluded from the between‐groups comparison of jump performance F–V profiles because of the difficulty of most players to perform the loaded jump correctly (probably due to their young age and low training experience), preventing us to properly establish their F–V relationship. This was done to avoid drawing inconsistent conclusions. For other age groups, results showed significantly greater jump performance at different load conditions (p < 0.05), VT‐P_max (p < 0.05) and significantly more negative values of VT‐Sfv (p < 0.05) in all groups from PHV to PHV+3 in comparison with PHV−1. With the exception of VT‐Sfv, these variables were also significantly greater in PHV+3 when compared to PHV (p < 0.01) and PHV+1 (p < 0.05) groups. However, there were no significant differences in VT‐F0 kg⁻¹, VT‐P_max kg⁻¹, and VT‐V0 between the different groups (p > 0.05).

When comparisons were carried out between consecutive groups, significant differences were found between PHV−1 and PHV groups in H_max (p < 0.05), H₆₀ (p < 0.05), VT‐F0 (p < 0.05), and VT‐P_max (p < 0.05). However, only VT‐F0 was significantly different between PHV and PHV+1 (p < 0.05) groups, and only VT‐P_max was significantly different between PHV+2 and PHV+3 groups (p < 0.05) (Table 2).

TABLE 2.

Comparison of mechanical components of the jump force–velocity profile by pairs of consecutive biological age groups.

	PHV−1	PHV	PHV+1	PHV+2	PHV+3
Hmax (0%)	0.23	0.27*	0.28	0.30	0.32
(m)	(0.03)	(0.04)	(0.04)	(0.04)	(0.03)
ES		0.36	0.07	0.13	0.13
Hmax (30%)	0.14	0.17	0.18	0.20	0.22
(m)	(0.03)	(0.04)	(0.03)	(0.03)	(0.02)
ES		0.24	0.04	0.15	0.26
Hmax (60%)	0.09	0.12*	0.13	0.14	0.15
(m)	(0.03)	(0.03)	(0.03)	(0.02)	(0.02)
ES		0.24	0.06	0.08	0.10
VT‐F0	1215.1	1520.1*	1842.1*	1997.6	2146.5
(N)	(225.5)	(273.7)	(454.9)	(350.4)	(190.3)
ES		0.35	0.30	0.08	0.11
VT‐F0	26.4	28.3	28.8	29.1	29.0
(N kg−1)	(4.7)	(4.0)	(4.7)	(3.3)	(2.9)
ES		0.11	0.03	0.02	0.02
VT‐V0	3.1	3.4	3.1	3.2	3.7
(m s−1)	(0.7)	(1.0)	(0.9)	(0.8)	(0.8)
ES		0.10	0.05	0.03	0.14
VT‐Pmax	912.8	1250.8*	1398.9	1586.0	1973.5*
(W)	(162.1)	(363.8)	(360.8)	(397.0)	(435.8)
ES		0.51	0.11	0.12	0.26
VT‐Pmax	19.7	23.2	21.9	23.1	26.3
(W kg−1)	(2.6)	(5.7)	(4.3)	(4.6)	(3.3)
ES		0.33	0.06	0.07	0.18
VT‐Sfv	−424.7	−501.3	−650.6	−665.7	−604.2
(N s m−1)	(160.1)	(201.0)	(294.1)	(235.6)	(131.4)
ES		0.13	0.19	0.01	0.07
VT‐Sfv	−9.3	−9.4	−10.1	−9.7	−8.2
(N s m−1 kg−1)	(3.6)	(3.6)	(4.0)	(3.1)	(2.0)
ES		0.01	0.05	0.03	0.13

Note: Significant difference with respect to the previous biological age group (*p < 0.05).

Abbreviation: ES, effect size with the consecutive previous group.

3.3. Sprint Performance

A significant, progressive, and linear improvement (from the PHV group) in sprint performances (p < 0.05), HZT‐F0 (p < 0.001), HZT‐V0 (p < 0.01), HZT‐P_max (p < 0.001), and HZT‐Sfv (p < 0.05) was observed in parallel with maturation.

When comparisons were carried out between consecutive groups, we recorded significant shorter 5 m sprint time (p < 0.05), significant more negative value of HZT‐Sfv kg⁻¹ (p < 0.05), and significant greater HZT‐F0 kg⁻¹ (p < 0.05) in PHV−2 in comparison with the PHV−1 group. No other significant differences in the remaining variables were found between these two groups.

When comparing PHV with the PHV−1 group, we observed significantly shorter sprint times over all distances (p < 0.05) and significantly greater HZT‐F0 (p < 0.001), HZT‐F0 kg⁻¹ (p < 0.05), HZT‐P_max (p < 0.001), and HZT‐P_max kg⁻¹ (p < 0.001) in the PHV group. In contrast, HZT‐Sfv was found at lower value in PHV in comparison with the PHV−1 group (p < 0.05). HZT‐V0 and HZT‐Sfv kg⁻¹ did not change significantly between these two groups (p > 0.05).

No significant differences of any variables were found between PHV and PHV+1 groups, except for HZT‐Sfv kg⁻¹, which was at a less negative value in the latter group (p < 0.05).

Significant sprint performance differences were revealed mainly at shorter distances (5–20 m) between PHV+1 and PHV+2 groups (p < 0.05), which were accompanied by significant greater values of HZT‐F0 (p < 0.001), HZT‐F0 kg⁻¹ (p < 0.01), HZT‐P_max (p < 0.001), and HZT‐P_max kg⁻¹ (p < 0.01) and significant more negative value of HZT‐Sfv (p < 0.01) in the PHV+2 group.

When PHV+3 was compared to PHV+2, significantly greater HZT‐V0 (p < 0.001) and HZT‐Sfv kg⁻¹ (p < 0.05) were observed in the oldest group. No significant differences of the remaining variables were found between these two groups (Table 3).

TABLE 3.

Comparison of mechanical components of the sprint force–velocity profile by pairs of consecutive biological age groups.

	PHV−2	PHV−1	PHV	PHV+1	PHV+2	PHV+3
Sprint‐5 m	1.48	1.55*	1.45**	1.48	1.39**	1.41
(S)	(0.07)	(0.07)	(0.08)	(0.07)	(0.07)	(0.06)
ES		0.25	0.33	0.11	0.31	0.07
Sprint‐10 m	2.38	2.42	2.27***	2.29	2.16**	2.16
(S)	(0.08)	(0.10)	(0.12)	(0.09)	(0.08)	(0.08)
ES		0.14	0.35	0.06	0.34	0.01
Sprint‐15 m	3.19	3.23	3.03***	3.01	2.87**	2.84
(S)	(0.09)	(0.14)	(0.15)	(0.11)	(0.09)	(0.06)
ES		0.08	0.31	0.03	0.33	0.09
Sprint‐20 m	3.97	4.00	3.77***	3.71	3.53*	3.47
(S)	(0.12)	(0.20)	(0.18)	(0.14)	(0.11)	(0.08)
ES		0.05	0.27	0.08	0.32	0.17
Sprint‐25 m	4.73	4.75	4.51*	4.39	4.17	4.08
(S)	(0.16)	(0.25)	(0.34)	(0.17)	(0.11)	(0.10)
ES		0.03	0.22	0.09	0.30	0.21
Sprint‐30 m	5.51	5.50	5.15***	5.05	4.83	4.67
(S)	(0.20)	(0.32)	(0.24)	(0.22)	(0.14)	(0.12)
ES		0.00	0.26	0.11	0.25	0.31
HZT‐F0	281.9	301.4	393.0***	425.6	529.3***	531.6
(N)	(39.9)	(42.3)	(61.0)	(58.9)	(78.6)	(86.6)
ES		0.12	0.50	0.14	0.42	0.00
HZT‐F0	7.15	6.44*	7.33*	6.72	7.74**	7.11
(N kg−1)	(0.73)	(0.56)	(0.99)	(0.63)	(0.80)	(0.63)
ES		0.23	0.36	0.17	0.38	0 0.21
HZT‐V0	6.79	7.04	7.47	8.02	8.18	9.06***
(m s−1)	(0.35)	(0.59)	(0.44)	(0.60)	(0.30)	(0.61)
ES		0.17	0.17	0.32	0.07	0.71
HZT‐Pmax	479.5	534.1	734.6***	857.2	1083.1***	1198.3
(W)	(78.6)	(90.7)	(129.1)	(160.8)	(167.6)	(163.6)
ES		0.16	0.51	0.25	0.34	0.18
HZT‐Pmax	12.12	11.34	13.69***	13.46	15.81**	16.04
(W kg−1)	(1.20)	(1.35)	(2.02)	(1.53)	(1.57)	(1.09)
ES		0.15	0.40	0.03	0.37	0.04
HZT‐Sfv	−41.5	−42.5	−52.7*	−53.1	−64.7**	−59.2
(N s m−1)	(5.6)	(6.8)	(8.4)	(6.3)	(9.7)	(12.5)
ES		0.04	0.35	0.01	0.44	0.15
HZT‐Sfv	−1.06	−0.92*	−0.99	−0.84*	−0.95	−0.79*
(N s m−1 kg−1)	(0.14)	(0.12)	(0.15)	(0.11)	(0.11)	(0.11)
ES		0.23	0.13	0.26	0.24	0.38

Note: Significant difference with respect to the previous biological age group (*p < 0.05, **p < 0.01, and ***p < 0.001).

Abbreviation: ES, effect size with the consecutive previous group.

4. Discussion

The aim of the current study was to examine the evolution of jumping and sprinting performance and their mechanical determinants across maturation in young elite male soccer players. In accordance with our hypothesis, the results regarding the maturation‐related differences in jump performance and associated mechanical components are mainly due to body mass gain throughout growth and maturation, which is at the origin of F0 and P_max increases. Regarding the mechanical determinants of sprint performance, they demonstrated different accelerated periods of development in force and velocity qualities with maturation during the physical development process of young soccer players.

4.1. Jumping Performance

It is worth noting that the maximal jump heights measured here is directly dictated by the mechanical impulse generated in the direction of the movement. Impulse is equal to the product of mean net vertical external force and force application time. This highlights the importance of analyzing vertical force production capacities (e.g., the jumping force–velocity profile) to identify the key factors influencing vertical jump performance. Our resulting jump heights at different loads were significantly higher in all groups from PHV (circa‐PHV) to PHV+3 (post‐PHV) in comparison with the PHV−1 group (pre‐PHV). PHV+3 also performed significantly higher jumps than PHV and PHV+1 groups. In line with these results, previous studies have reported higher jumping performance in young soccer players at post‐ > circa‐ > pre‐PHV, suggesting an increase of jumping performance with growth and maturation (Lloyd et al. 2015; Malina, Eisenmann, et al. 2004b). However, the current study is the first to examine force production capacities underlying jumping performance across maturation process. We originally found significantly greater values of VT‐F0 and VT‐P_max and significant lower values of VT‐Sfv in all groups from PHV to PHV+3, when compared to the PHV−1 group. In addition, PHV+3 displayed significantly higher values of VT‐F0 than PHV and PHV+1 and of VT‐P_max when compared to PHV, PHV+1, and PHV+2 groups. This underlies that the F–V profile progressively tends to be more oriented toward force production capacities at low velocity with growth and maturation. These findings are explained by structural adaptations that occur due to a gradual increase of circulating androgenic hormones from puberty onward (PHV) (Malina 1969; Viru et al. 1999). These hormones have a determinant role in muscle glycogen synthesis and hypertrophy, leading to an improvement in muscle mass, force production capacity, and its different expressions (F0 and P_max) (Meylan et al. 2014; Reyes et al. 2018). However, when these F–V profile components were expressed relative to body mass (VT‐F0 kg⁻¹, VT‐P_max kg⁻¹, and VT‐Sfv kg⁻¹), no significant between‐group differences were revealed. This result indicates that significant increases of muscle mass (and so body mass) with age and maturation (Malina 1969) plays a relevant role in increasing VT‐F0, VT‐P_max, and decreasing VT‐Sfv, in accordance with the linear relationship previously reported between body mass and strength (Jaric et al. 2005). Moreover, V0 did not show any significant difference between groups, indicating that the gradual increase of jump performance with age and maturation is mainly dependent on VT‐F0 development rather than VT‐V0. This result is in line with a recent study that established a strong relationship between F0, VT‐Sfv, and both chronological age and maturity offset (Fernández‐Galván et al. 2021), confirming that F0 is the major mechanical determinant of jumping performance evolution across age and maturation.

When two consecutive biological age groups were compared, only the PHV (circa‐PHV) group revealed significantly greater performances of both unloaded (H_max) and loaded jumps (H₆₀) with respect to the previous age group (PHV−1; pre‐PHV) (Table 2). Significantly, higher VT‐F0 and VT‐P_max were also observed in PHV compared to the PHV−1 group. As explained above, these results are due to the rapid gains in muscle mass, which coincide with the onset of pubertal growth spurts in athletes in the PHV group (Malina 1969; Viru et al. 1999). Although PHV+1 (post‐PHV) also showed significantly greater VT‐F0 in comparison with PHV, no significant differences in jumping performance were observed. In accordance with these results, previous studies suggest that maximal gains in muscular strength occur after PHV, which is probably related to these adolescent spurts in muscle mass that occur shortly after PHV (G. Beunen and Malina 1988; Malina, Bouchard, et al. 2004). In our results, body size that increases gradually up to the PHV+1 group point in this direction, suggesting that growth mainly ends at this biological age group, although some increase still carries on. Although we also recorded a significantly higher VT‐P_max in PHV+3 compared to PHV+2, no significant change in jump performance between the two groups were recorded.

4.2. Sprinting Performance

In accordance with previous studies that determined sprint performance in young athletes is influenced by age and maturation (Malina, Eisenmann, et al. 2004; Mendez‐Villanueva et al. 2011), we also demonstrated a significant change in sprint performance across groups PHV−2 to PHV+2. Contrary to results observed in jumping performance, all mechanical components associated with sprint performance improved with the process of growth and maturation. Some of them, especially those related to force quality (HZT‐F0, HZT‐F0 kg⁻¹, HZT‐P_max, HZT‐P_max kg⁻¹, and HZT‐Sfv), plateaued around the maturity stage corresponding to PHV+2 (Table 3).

Regarding the changes between consecutive age groups, we observed a significantly shorter sprint time, mainly at short distance (5 m), in PHV−2 when compared to the PHV−1 group. Previous studies also show a decline of sprint performance in the year before PHV (G. P. Beunen et al. 1988; G. Beunen and Malina 1988; Philippaerts et al. 2006). This phenomenon of a temporary decline in performance just before the approximate age of PHV is called “adolescent awkwardness” resulting from a disruption of motor coordination due to differential timings of growth spurts in both leg and trunk length (Malina, Bouchard, et al. 2004). We also observed significant less negative value of HZT‐Sfv kg⁻¹ and a significant lower HZT‐F0 kg⁻¹ in the PHV−1 group. This decline in the ratio of force/body mass in the PHV−1 group was expected given the strong relationship previously reported between acceleration and HZT‐F0 kg⁻¹ during short‐distance sprinting (Buchheit et al. 2014). Moreover, the lack of significant difference in HZT‐V0 and HZT‐P_max was logically associated with similar running performances over distances > 10 m, as perfect correlations had already been reported between these horizontal mechanical components and maximal sprinting speed (Buchheit et al. 2014). These results indicated that force relative to body mass became temporarily less efficient in soccer players who were just before the onset of puberty (PHV−1). This deficiency could be explained by a predominant growth spurt in trunk length, which would increase body mass without substantial force development of propulsive muscles. Another possible explanation is that the likely increase in force production capacity of muscles occurs in those that are not particularly involved in the motor pattern of the acceleration phase of sprinting.

In addition, the PHV−1 group also showed significant longer sprint time over all distances, compared to PHV group, as well as significant lower values of HZT‐F0, HZT‐F0 kg⁻¹, HZT‐P_max, and HZT‐P_max kg⁻¹ and significant more negative value of HZT‐Sfv. Given the improvement of F0 is usually associated with an increase in strength, both were expected to develop significantly from the PHV−1 (pre‐) to PHV (circa‐PHV) group as maximal gains in muscle strength and power have been shown to coincide with PHV in young soccer players (Malina, Bouchard, et al. 2004; Philippaerts et al. 2006). The greater ESs in HZT‐F0, HZT‐P_max, and HZT‐P_max kg⁻¹ observed between PHV−1 and PHV groups, when compared to the other between‐groups' ESs, confirm the results of those former studies (Table 3). However, no significant changes in horizontal mechanical components expressing maximal running speed (HZT‐V0) were observed between these different groups. These findings showed that PHV−1 year is a critical age and maturity stage during which a high deficit of absolute and relative force to body mass is apparent when compared to both inferior and superior biological age groups. A specific training program aimed at improving coordination and horizontal force (e.g., horizontally directed resistance exercises) should be implemented in the exercise regimen of this biological age group. Not only would it enhance their body awareness (coordination) during maturation due to growth spurts in leg and trunk length, it would improve their horizontal application of force, thereby reducing this decline in acceleration performance associated with this maturity stage (Arcos et al. 2014; Randell et al. 2010).

Although no significant difference was found for any variable between PHV and PHV+1 groups (except for HZT‐Sfv kg⁻¹, which did not impact sprint performance), significantly differences in sprint performance, mainly at shorter distances (5–20 m), were revealed between PHV+1 and PHV+2 groups. Significantly greater values of HZT‐F0, HZT‐F0 kg⁻¹, HZT‐P_max, and HZT‐P_max kg⁻¹ and significantly more negative value of HZT‐Sfv in PHV+2 were also recorded. These results indicate that players in PHV+1 also need to particularly improve their acceleration phase, force, and power output. Practitioners should reemphasize the development of maximal horizontal force production capacity at this biological age by implementing horizontally directed conditioning training. Several exercises that have already been verified to enhance propulsive force, power, and therefore, sprinting performance mainly at short distances include: resistance exercises performed in the horizontal axis (Arcos et al. 2014), resisted‐sled training (Morin et al. 2017), horizontal jump (Ramírez‐Campillo et al. 2015), and combined low‐load‐high‐velocity resistance and plyometric exercises (Maio Alves et al. 2010; Zghal et al. 2019).

When compared to its younger consecutive group, PHV+2 did not demonstrate biological age‐related improvements in sprint performance over long distances (above 20 m) nor mechanical components expressing the maximum running speed (HZT‐V0). It also displayed significantly lower values of these mechanical components in comparison to the PHV+3 group. The implementation of training protocols to develop maximum running speed, such as unresisted specific sprint (Zafeiridis et al. 2005), assisted (elastic cord assistance, downhill, etc.) (Bartolini et al. 2011; Ebben et al. 2008), or sprint‐specific plyometric training (Kotzamanidis 2006), could be beneficial for this maturity stage (young soccer players which are at ∼2‐year post‐PHV) to improve overall sprint performance.

The gradual and significant increase of HZT‐V0 and HZT‐Sfv kg⁻¹ after the onset of pubertal development (PHV) confirm the assumption of previous studies highlighting that adolescents at post‐PHV are characterized by changes in musculotendinous tissue and limb growth (Lloyd et al. 2013) along with an increased efficiency of the stretch‐shortening cycle (Lloyd et al. 2020), potentially resulting in an improvement in V0 (Moran et al. 2017). When pairs of consecutive biological age groups were compared, only the PHV+3 group revealed significantly greater value of HZT‐V0 and significant less negative value of HZT‐Sfv kg⁻¹ when compared to the PHV+2 group. Larger ESs (ranging between 0.38 and 0.71) were also observed between PHV+3 and PHV+2 compared to the other between‐group ESs for these two mechanical components. This suggests that PHV+3 (∼17 years) could represent the optimal age and maturation‐related gain in both maximum running speed quality and its mechanical components, at least for the biological age range tested (PHV−2–PHV+3). Despite these significant differences of mechanical components expressing maximal sprinting speed between PHV+2 and PHV+3 groups, we did not find any significant difference in sprint running performance over the full distance from 5 to 30 m. It is possible that a significant difference could have been observed over a greater distance, as maximal velocity for team sport athletes is typically achieved during a longer sprint of 30–40 m (Young et al. 2008). The increasingly larger ESs with increasing sprint distance observed between PHV+2 and PHV+3 groups might support this hypothesis.

Contrary to HZT‐V0, which increases linearly with maturation, and substantially in the PHV+3 group, HZT‐F0 kg⁻¹ exhibits a sawtooth‐shaped biological age‐related evolution with significant marked peaks of increase in PHV and PHV+2. In contrast, HZT‐Sfv kg⁻¹ displayed significant less negatives values in PHV−1, PHV+1 and PHV+3 groups, indicating different spurts of gains in velocity qualities at biological age corresponding to these groups (Figure 1). These results suggest that accelerated periods of development in force and velocity qualities with age and maturation alternately occur during the physical development of young soccer players, with greater gains in force in PHV and PHV+2 and in velocity qualities in PHV−1, PHV+1, and PHV+3. The relationships between these specific mechanical characteristics and maturation status should be considered by practitioners when designing appropriate and individualized programs to optimize training‐related adaptations in young soccer players. For instance, a focus on exercises that improve qualities, which are less naturally developed, would allow young male soccer players to benefit from adaptations related to both natural development and deliberate training. Based on our findings, that require further longitudinal research to be definitely supported, we recommend emphasizing the development of maximal horizontal force production capacity (by implementing resistance exercises performed in the horizontal axis, resisted‐sled training, horizontal jump, or combined resistance/plyometric training) at biological age stages corresponding to PHV−1, PHV+1, and PHV+3 categories. On the contrary, favoring training methods that improve velocity qualities, such as unresisted specific sprint, assisted (elastic cord assistance, downhill, etc.), or sprint‐specific plyometric training would be very beneficial for maturational stages corresponding to the PHV−2, PHV, and PHV+2 categories. However, it should be noted that, despite our large sample size compared to many other studies, it remains very small compared to the relative number of youth academy soccer players worldwide. Therefore, more data are necessary, especially from longitudinal studies, since a difference in the training history between the different biological age categories might have an effect, rather than the effect of maturation per se, on the evolution of jumping and sprinting performances and their underlying mechanical determinants. Indeed, although all our tested players belong to the same academy with almost similar technical, tactical and physical training strategies applied by the same coaches for at least 5 years, they belong to six different age categories involving different levels of competition in terms of time, pace, combativeness, etc., which also require different training volume and intensity. Furthermore, since force–velocity variables are integrative indexes of external force production capacities, they are affected by neural underlying mechanisms, as activation dynamics or intermuscular coordination, which are known to change with the maturation status (Armstrong et al. 2019; Driss and Vandewalle 2013; Van Praagh and Doré 2002; Lazaridis et al. 2010). This may partly explain some specifics evolution of force–velocity variables.

FIGURE 1.

Representation of the different mechanical components evolution of the sprint force–velocity profile during the maturation process. Top graph: HZT‐V0 (m s⁻¹; open triangle) and relative HZT‐F0 (N kg⁻¹; open circle). Bottom graph: relative HZT‐Sfv (N s kg⁻¹ m⁻¹; open rhombus). All values are expressed as means ± SD. Significant difference with respect to the previous biological age group (+: p < 0.05, ++: p < 0.01, and +++: p < 0.001).

5. Limitations

Despite the relevance of our findings, our study presents some limitations. First, our decision to exclude the PHV−2 group from the between‐groups comparison of jump performance F–V profiles has narrowed our assessed age range, thereby limiting our insight of how mechanical determinants of this performance evolve with maturation. Second, the linear model used for the F–V relationship assessment was based on jumps at only three different points corresponding to three different loads (at 0%, 30%, and 60% of body mass). This could have affected the accuracy of the established F–V profile, and therefore, the extrapolated mechanical variables. However, for our analysis, we were careful to only include the participants showing a linear F–V relationship with a coefficient of determination > 0.95. Third, the Mirwald's equation method used to predict somatic maturity has been shown to present some limitations, making it less accurate, particularly in individuals who are already far beyond their peak height velocity (PHV) (Moore et al. 2015; Malina et al. 2006). Therefore, caution should be exercised when interpreting the study results based on this equation.

6. Conclusion and Practical Applications

Current findings in young elite male soccer players suggest the existence of specific mechanical characteristics related to each maturation status, which explains the differences in jump and sprint performances between groups. Our observations revealed an alternation of accelerated periods of development in force and velocity qualities with growth and maturation during the physical development process of young soccer players. It seems that certain stages of maturity would be more favorable to greater gains in force, whereas others could show broader developmental spurts in terms of velocity qualities. A greater understanding of these differences occurring at various stages of growth and maturation would be useful to optimize the talent selection process for young soccer players. More importantly, it could provide interesting data to practitioners in order to implement an appropriate and individualized training stimulus based on the mechanical characteristics relative to each maturity status to optimize training‐related adaptations in young soccer players.

Conflicts of Interest

The authors declare no conflicts of interest.