Background: The aim of this study is to analyze the relationship
between the accumulated training load parameters (i.e., acute (AWL), chronic
(CWL), acute: chronic workload ratio (ACWR), training monotony (TM), and training
strain (TS)) and sprint performance variations in elite adolescent soccer
players, taking into account the maturation status of the players. Besides, we
aimed to use regression models with mentioned parameters, sprint level, and peak
height velocity (PHV) as predictors to explain variations in sprint performance
during the in-season. Methods: Twenty-seven U16 soccer players (age:
15.5
Soccer is an intermittent sport characterized by interspersed multiple high-intensity short activities (e.g., running and sprinting) with predominantly low-intensity activity (e.g., standing and walking) demands [1, 2]. Even from a young age, modern soccer requires high levels of physical fitness development [3, 4]. According to time–motion analysis, elite professional adult soccer players cover a total distance of approximately 10–12 km at an average intensity close to the anaerobic threshold (80–90% of maximum heart rate) [5, 6], and they perform 1350 activities every 4–6 seconds during the game. Approximately 150 to 250 of these activities are short, intense, and explosive activities associated with maximal sprint, acceleration, and change of direction [7, 8]. On the other hand, the activity profiles of young soccer players (distance covered, high-intensity activity and sprinting) during the match are low. It was shown in a study that elite young soccer players between the ages of 13–18 covered a distance of approximately 6.5–9.0 km during the match, and high-intensity activity was carried out with 670–970 m of this distance, and 190–670 m was the sprint distance [9]. Considering the above values, although energy is supplied by the aerobic system for most of the soccer game, during the performance of continuous explosive activities the anaerobic system works actively, such us keeping control of the ball against defensive pressure, jumping, tackling, kicking, turning, sprinting, changing of direction during the game [6, 10, 11]. Therefore, soccer players need to have well-developed aerobic and anaerobic metabolisms in order to meet and sustain the necessary physical and physiological demands, in turn providing the best performance during the match [7, 12, 13].
Considering that the most decisive movements in soccer take place in areas
smaller than 10 m
During the season, it is recommended that the applied workloads should be sufficient to improve the physical performance quality of the players [18]. Monitoring the training load is seen as an important factor to determine whether the athletes are adapting to the training program, to optimize the training process, and to minimize the risk of non-functional overreaching, disease and/or injury [32, 33]. Sports and exercise scientists recognize that “training load” includes of both “external” and “internal” domain [32, 34]. External training load is defined as the activity profiles of players or physical work during the training sessions (for example, total distance covered, acceleration, deceleration or metabolic power), while internal training load includes all psychophysiological responses that occur during execution of the exercise predicted in response to external training load (for example, degree of perceived exertion (RPE), heart rate (HR)) [33, 35]. session-RPE (s-RPE) is an easy-to-use [36, 37], and the most common valid/reliable method for measuring internal training load and accumulation between sessions in team sports [38]. It was previously demonstrated that sRPE was associated with the HR-derived measures of training intensity in professional soccer players [39]. Besides s-RPE, recent studies have shown that parameters derived from the internal and external training load of soccer players are also frequently used in the monitoring training load throughout the season. These parameters are the acute (AWL), chronic (CWL), acute: chronic workload ratio (ACWR), training monotony (TM), and training strain (TS). Haddad et al. [38] stated that these parameters mentioned above can be calculated from the session-RPE method data of a training microcycle. Higher TM scores indicate lower standard deviations of the mean, i.e., small variations within a week, while higher training strain points out larger acute loads applied with small variations during the week [40]. These high scores may be associated with disease incidence, poor performance, and the onset of overtraining [36, 38]. With this, the use of the ACWR to understand changes in the load and how these changes relate to risk of injur, has received increasing scientific attention [41, 42]. ACWR is calculated by dividing the AWL (the workload of the week preceding the injury, fatigue component) by the CWL (the average workload of the four weeks preceding the injury, fitness component) [41, 42]. Considering the training intensity parameters mentioned above, coaches can determine the physical and physiological effects of training sessions on players.
Furthermore, Nobari et al. [43] emphasized that accumulated training load and maturation status play a critical role in the physical capacity changes observed across the season, especially sprinting which was demonstrated to improve naturally with age, reported that improvements in performance result from changes in neuromuscular mechanisms related to growth and maturity [44]. Therefore, coaches need to take these two factors into consideration in order to carefully interpret the fitness variations in their players and to adjust the types of training they will perform according to the maturation level of the players. As far as we know, there is no study examining the relationship between the accumulated training load (AWL, CWL, TM, TS, and ACWR) and the changes in sprint performance, that also takes into account the maturation factor in elite young soccer players. Considering the advantage of having a good sprint performance in the soccer game, it is extremely important to optimize the training load-rest relationship throughout the season, and to improve the parameters related to speed. The aim of this study is to analyze the relationship between the accumulated training load parameters and sprint performance variations in elite adolescent soccer players, taking into account the maturation status of the players. According to the relevant literature examining the relationship between training load variables and different physical fitness characteristics (except sprinting) [4, 13, 43]. As a result, based on the literature presented [4, 45, 46, 47, 48], we hypothesized that the accumulated training load and maturation maybe partially explain variation of sprint performance during the competitaion season in elite youth soccer players.
Twenty-seven U16 soccer players (age: 15.5

The dominant training microcycle during the competition season.
In this study there was a cohort with monitoring the daily workload for 15 weeks in the competitive season: early-season (EaS) weeks (w) W1 to W5; mid-season (MiS) W6 to W10; and end-season (EnS) W11 to W15 (Fig. 2). Participants were assessed on anthropometric measurements, maturity and sprint performance by the same group of researchers during the complete study, at the same time of the day (8–11 Am) [49]. The first evaluations were performed at 16 °C and 27% humidity and the second stage evaluations were performed at 12 °C and 35% humidity. All tests and exercises were performed on natural grass.

Research outline of the weekly monitoring on training and match load and assessed sessions during the competition season. EaS, early-season; Mid, mid-season; EnS, end-season; wCL, weekly accumulated chronic workload; TS, training sessions; A.U., arbitrary unit.
Height was measured with a portable stadiometer (Seca model 213, Hamburg,
Germany). Body mass was performed using portable weighing scales (Seca model 813,
United Kingdom). This data was used to distinguish the maturity offset and age at
PHV of the subjects, the down formula was used [50], as follows: Maturity offset
= –9.236 + 0.0002708 (leg length
The intensity of training sessions was estimated using the Borg CR-10 rate of perceived exertion (RPE) scale [50]. Thirty minutes after the end of the training session each player reported his RPE for each session confidentially without knowledge of other players’ ratings. As a measure of internal load, the session-RPE was derived by multiplying RPE and session duration (min) [36]. Players were previously familiarized with the scale during two years at the club.
Additional, workload (WL) parameters were calculated. A total load of daily
training during the week was considered as weekly AWL; the uncoupled formula [51]
was used to obtain the weekly chronic (CWL) and acute-chronic workload ratio
(ACWR); weekly training monotory (TM) (weekly AWL
Each participant performed two maximal 30-m sprints, measured with one pair of the electronic timing system sensors (Newtest Oy, Finland) mounted on tripods that were set at hip height and was positioned 3 m apart facing each other on either side of the starting line. The participants commenced the sprint from a standing start, 0.5 cm behind the first timing gate. Between two trials recovery was 3 minutes. The best time was recorded for analysis. Tests were performed outdoor and on natural grass.
Data were analyzed in SPSS Version 25 (IBM SPSS Inc., Chicago, IL, USA) except
for multiple linear regression and Akaike information criterion (AIC), which were
calculated using Graph-Pad Prism 9 (GraphPad Software Ind, San Diego, California,
CA, USA). Results are expressed as mean
In Table 1 significant positive correlations were shown between Sprint EaS with
Sprint EnS (r = 0.965; p
Variable | ||||||||||||||||||
PHV ( |
1 | |||||||||||||||||
SPRINT1 ( |
–0.206 | 1 | ||||||||||||||||
SPRINT2 ( |
–0.169 | 0.965** | 1 | |||||||||||||||
AWL1 ( |
–0.070 | 0.413 | 0.327 | 1 | ||||||||||||||
AWL2 ( |
0.014 | 0.548* | –0.387 | –0.429** | 1 | |||||||||||||
AWL3 ( |
–0.110 | –0.236 | –0.226 | 0.001 | 0.061 | 1 | ||||||||||||
CWL1 ( |
–0.071 | –0.107 | –0.108 | 0.285* | –0.299* | –0.168 | 1 | |||||||||||
CWL2 ( |
0.070 | –0.584** | –0.543* | 0.242* | –0.134 | 0.093 | 0.091 | 1 | ||||||||||
CWL3 ( |
0.224 | 0.093 | 0.031 | –0.201 | 0.176 | 0.188 | –0.263* | –0.002 | 1 | |||||||||
ACWR1 ( |
0.138 | 0.254 | 0.216 | 0.124 | 0.095 | –0.018 | 0.182 | 0.198 | 0.187 | 1 | ||||||||
ACWR2 ( |
0.268 | 0.151 | 0.130 | –0.023 | 0.465** | –0.279** | –0.086 | 0.099 | 0.045 | 0.718** | 1 | |||||||
ACWR3 ( |
0.405 | 0.458* | 0.355 | 0.001 | –0.244* | 0.011 | 0.046 | 0.043 | 0.138 | 0.006 | –0.194 | 1 | ||||||
TM1 ( |
0.156 | 0.579* | 0.463* | 0.601** | –0.374* | 0.108 | 0.197 | 0.070 | –0.071 | 0.015 | –0.247 | –0.044 | 1 | |||||
TM2 ( |
0.351 | –0.294 | –0.218 | –0.500** | 0.447** | –0.015 | –0.285* | –0.151 | 0.056 | 0.094 | 0.153 | –0.103 | –0.438** | 1 | ||||
TM3 ( |
0.053 | 0.216 | 0.190 | 0.208 | –0.365** | 0.058 | 0.239 | –0.111 | –0.108 | 0.015 | –0.088 | –0.009 | –0.063 | 0.014 | 1 | |||
TS1 ( |
–0.184 | 0.513* | 0.419 | 0.685** | –0.419** | –0.034 | 0.294 | 0.083 | –0.217 | 0.025 | –0.233 | –0.081 | 0.943** | –0.453** | –0.018 | 1 | ||
TS2 ( |
0.438 | –0.350 | –0.283 | –0.518** | 0.472** | 0.025 | –0.351 | –0.118 | 0.126 | 0.102 | 0.187 | –0.107 | –0.451** | 0.966** | 0.023 | –0.476** | 1 | |
TS3 ( |
0.095 | –0.092 | 0.074 | 0.052 | –0.231* | 0.064 | 0.191 | –0.337 | –0.023 | 0.024 | –0.186 | 0.039 | –0.087 | –0.075 | 0.516 | –0.086 | –0.081 | 1 |
AWL = the accumulated acute workload in the season; CWL = the accumulated chronic
workload in the season; ACWR = the accumulated acute: chronic workload ration in
the season; TM = the accumulated training monotony in the season; TS = the
accumulated training strain in the season; PHV, Peak height velocity 1:
early-season; 2: mid-season; 3: end-season; * Represent demonstrated significance
in correlation between two parameters at p |
Descriptive workload and sprint results and comparison between EaS, MiD and EnS
are presented in Table 2. Regarding data, there was no difference between EaS vs.
MiD (p
Variables | EaS (Mean |
MiD | EnS (Mean |
EaS vs. MiD | MiD vs. EnS | Eas vs. Ens | ||||||
(Mean |
p | CI (95%) | Effect size | p | CI (95%) | Effect size | p | CI (95%) | Effect size | |||
Sprint (s) | 4.22 |
— | 4.14 |
— | — | — | — | — | — | 0.04, 0.10 | –0.28 (–0.39; –0.18) | |
AWL (A.U.) | 1615.5 |
1606.3 |
1407.6 |
0.99 | –209.4, 197.6 | 0.06 (–0.27; 0.40) | 0.124 | –26.1, 317.4 | –0.34 (–0.63; –0.05) | 0.113 | –22, 301.4 | –0.36 (–0.61; –0.10) |
CWL (A.U.) | 1660.4 |
1591.1 |
1398.3 |
0.377 | –31.2, 138.2 | –0.34 (–1.91; –1.11) | 100.7, 288.9 | –1.51 (–1.91; –1.11) | 154.5, 342.2 | –0.80 (–1.05; –0.56) | ||
ACWR (A.U.) | 1.05 |
0.94 |
0.94 |
0.202, 0.413 | –3.02 (–3.63; –2.40) | 0.022 |
–0.357, –0.022 | –2.29 (–3.03; –1.56) | 0.373 | –0.307, 0.070 | –0.18 (–0.46; 0.11) | |
TM (A.U.) | 1.25 |
1.22 |
1.57 |
0.99 | –0.127, 0.207 | –0.01 (–0.47; 0.27) | 0.99 | –0.294, 0.190 | –0.30 (–0.70; 0.10) | 0.99 | –0.230, 0.205 | 0.13 (–0.23; 0.49) |
TS (A.U) | 2074.4 |
2134.9 |
1912.1 |
0.99 | –519.4, 480.5 | 0.01 (–0.35; 0.36) | 0.255 | –123.6, 737.7 | –0.42 (–0.71; ––0.13) | 0.195 | –88.3, 663.5 | –0.31 (–0.59; –0.03) |
AWL = the accumulated acute workload in the season; CWL = the accumulated chronic
workload in the season; ACWR = the accumulated acute: chronic workload ration in
the season; TM = the accumulated training monotony in the season; TS = the
accumulated training strain in the season; CI, Confidence interval; EaS,
early-season; Mid, mid-season; EnS, end-season; |
Multiple linear regression analyses were performed to predict the percentage of
change in sprint performance based on workload and maturity (Table 3 and Fig. 3).
The analysis of sprint showed that there were significant (F (4, 14) = 6.70,
p = 0.01), with a R
Variables | Beta | Estimate | |t| | p value | 95% CI for estimated | Total predict |
Sprint (%) | –13.37 | 5.41 | –18.6, –8.07 | R | ||
ACWR (A.U.) | 0.9621 | 4.09 | 0.001** | 0.45, 1.46 | Estimated R | |
TM (A.U.) | 0.5423 | 2.22 | 0.04* | 0.01, 1.06 | p: 0.003 | |
TS (A.U.) | –0.0001 | 2.24 | 0.04* | –0.001, 0.0004 | AIC value: 15.9 | |
PHV (years) | 0.2818 | 0.72 | 0.48 | –1.11, 0.55 | ||
AWL = the accumulated acute workload in the season; CWL = the accumulated chronic
workload in the season; ACWR = the accumulated acute: chronic workload ration in
the season; TM = the accumulated training monotony in the season; TS = the
accumulated training strain in the season; PHV, peak height velocity; COD,change
of direction; % = the percentage of change in between assessments from
early-season to after-season; AIC, Akaike information criterion, and CI,
confidence interval; * Represent demonstrated significance at p |

Prediction of the percentage of change in (a) sprint and residual plots in (b) sprint of multiple linear regression analysis. Note: PHV, Peak height velocity.
The aim of this study was to analyze the relationships between training WL parameters with variations in sprint performance in under-16 soccer players. The present study revealed that sprint performance improved in EnD compared to EaS independent of maturation, agreeing with our original hypothesis. Furthermore, there were significant variations in workload parameters (CWL and AWCR) over a soccer season. Additionally, significant correlations were found between the sprint performance, and the accumulated workload parameters, which is also in line with our hypothesis. Lastly, sprint performance can be estimated by ACWR, TM and TS values during the 15-week competitive season in young soccer players.
Analyzing the probability of associations between accumulated training load and changes in sprint performance helps determine whether training load is a determinant of these changes or if there are other factors that coaches should be aware of [53]. Having good physical capacity during the season also increases tolerance to training load. In one study, Malone et al. [18] expressed that well-developed lower body strength, repeated sprint ability, and speed performance provide better tolerance to higher workloads in team athletes, and are associated with a lower risk of injury. Moreover, previous study indicated that athletes who were slower at 5-m, 10-m and 20-m running distances were at higher risk of injury compared to faster athletes [54]. The present study revealed that the 30 m sprint performance improved during the competitive soccer season (EaS–EnS period). Suporting our results, recent studies demonstrated that sprint performance gradually improved over the course of the season in elite youth soccer players [4, 55]. On the contrary, previous studies found that sprint performance (10 m, 30 m) did not change significantly during a season in elite female soccer players, which is not compatible with the results of our study [53, 56]. Furthermore, multiple linear regression analysis revealed that maturity had no significant effect on the change in velocity performance during the season in the current study. Our hypothesis that maturation has a significant effect on the improvement in sprint performance was rejected (estimate = 0.28, t = 0.72, p = 0.48). Consistent with our results, recent studies showed that maturation did not significantly affect sprint performance [57, 58]. In contrast, some studies reported that maturation was effective in improving sprint performance in young soccer players [30, 59]. Similarly, Nobari et al. [4] found a strong correlation between the development of speed variables and PHV during the season in young soccer players and as a result, they empahized that maturation had a significant effect on the improvement in sprint performance. The reason why the improvement in sprint performance is independent of maturation can be explained as follows; the development of certain speed and power traits during growth and maturation may depend on the stage of development of physiological determinants or mechanisms that support these particular traits [21], such as myelination of motor nerves and neural maturation [43]. Moreover, Myers et al. [59] pointed out that measures of relative stifness and relative maximal strength had significant influence on the development of maximum sprint speed in males, independent of maturity in youths.
In literature, there are some studies that test the relationship between WL and
variations in physical and physiological variables during the competition season
in young soccer players. For instance, it was previously noted that sRPE during
the pre-season period were positively and largely associated with (r =
0.70–0.75) variations on 30–15 intermittent fitness test performance in
professional soccer players [60]. Another study conducted by Nobari et
al. [13] stated that a large and moderate relationship was found between
accumulated daily loads during one week and peak power and change of direction at
different periods of the season. Moreover, the same authors proposed that the CWL
and accumulated TM values could be utilised to better clarify the physical
capacities of young soccer players. Additionally, another study showed that there
were large correlations between cardiorespiratory performance (maximal aerobic
speed) and accumulated RPE, and accumulated sRPE [61]. As far as we know, there
is no study to examine the relationships between changes the accumulated workload
parameters (AWL, CWL, ACWR, TM and TS) and sprint performance over a soccer
season in youth soccer players. Therefore, the present data showed that the
percentage of change in sprint performance can be predicted by accumulated
workload parameters such as the ACWR, TM and TS. With the exception of PHV, these
three variables were observed to be significant predictors of the percentage
change in sprint performance during the 15-week competitive season. Contrary to
our findings, a previous study reported that there was no significant
relationship between the ACWR value and the improvement in sprint performance
[61], and another study noted that there was no significant relationship between
sRPE and fitness status (including 10 m and 30 m sprint performance) in elite
female soccer players [20]. Also in these studies, ACWR value is widely used to
predict injury risk [19, 33], and a recent study suggested that it can be used as
a performance monitoring tool for team sports athletes as well as injury
prediction [62]. In our study, it was observed that ACWR value was significantly
higher in EaS compared to MiD, and significantly higher in MiD compared to EnS
(0.94–1.05 A.U). In other words, we can say that the ACWR value is high in EaS
and MiD, and gradually decreases towards EnS. There are some findings in these
studies that support our results. For instance, Clemente et al. [63] is
in support of our results, stating that elite volleyball players had a high
training load during the early season period. In another study Nobari et
al. [6] demonstrated that ACWR of elite youth soccer players ranged from
0.90–1.14 A.U. throughout the competitive season. Additionally, Hulin et
al. [62] stated that high WL ratios (
Furthermore, TM is a measure of daily training variability [39], and variations
in training play a critical role in the prevention of monotony formation and the
realization of supercompensation. TS, like TM, is also related to level of
training compliance, and can increase the incidence of infectious diseases and
injuries during periods of high load associated with high monotony [55]. The
present study revealed that the significant improvement in sprint performance
throughout the season was predicted by the TM and TS variables. In favour of our
study, Stochi de Oliveira and Borin [55] reported low monotony values (1.4–1.7
A.U.) during the 20-week season in futsal players leading to an increase in the
height of the CMJ and thus a lower extremity strength preformance. The same
researchers suggested that distribution ratios of neuromuscular training and
tactical technical training throughout the season, as well as TM, provide
positive adaptations in lower extremity power performance. Furthermore, another
study stated that proper WL distribution or variations prevented maladjustment
from sports training and optimized athletic performance (maintaining positive
adaptations throughout the training cycle) [65]. Additionally, the present study
observed that there were no significant variations in TM and TS values during the
15-week competitive season. Our results were supported in the previous study on
professional soccer players conducted by Lu et al. [66] stated no
significant changes in sRPE-based TM or TS over four weeks. In our study, it is
seen that the TM values in the EaS, MiD and EnD periods are around 1.25, 1.22 and
1.57, respectively. Nobari et al. [67] reported that TM values in young
soccer players varied between 1.19–1.06 A.U. during 20 weeks, whereas TS values
varied between 1196.36 and 1735.53 A.U. According to Nobari et al. [67],
we can say that TM and TS values are lower than our study throughout the season.
In an another study, Nobari et al. [8] indicated that TM values average
1.2 A.U. during the season in under-16 soccer players. Another study conducted by
Miloski et al. [68] found the highest TM and TS values during the season
to be 1.61
Although we tried to have the same number of training sessions and compensation for all players during the season, this can be one of the limitations of the present study, since different number of games could affect the training workload. Another limitation of the study may be the lack of evaluation of external load monitoring with GPS [13, 69]. It has been suggested that external load monitoring should be done in future studies.
The present study revealed that sprint performance improved throughout the season in young soccer players, with significant intra-season variations, especially in CWL and ACWR load variables (Eas and Mid). In addition, it was observed that maturation did not have a significant effect on the change in sprint performance. This study clearly showed that there could be a relationship between sprint performance and accumulated wokload variables, and that the significant change in sprint performance can be explained by load variables such as AWCR, TM, and TS. With the repetition of such studies, increasing the sample size in different ages and sports branches, along with taking into account different genders. As in this study, it was observed that in-season load optimization and adjustment of variability promoted sprinting performance increase, especially in young soccer players. This information can assist coaches in talent selection and optimal design and development of training programs for different workload variables throughout the competitive season period.
Study Design—HN, HİC, EM-P. Data Collection—HN, HİC, SK, MEÖ. Data Analysis—HN, EM-P. Writing Original Draft—HİC, SK, MEÖ. Manuscript Review and Editing—HN, HİC, SK, MEÖ, EM-P. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.
Players and their parents received a clear explanation of the study. Experimental procedures were approved by the Ethics Committee of the University of Mohaghegh Ardabili (09.03.2020) and the recommendations of Human Ethics in Research were followed according to the Helsinki Declaration. Written informed consent was obtained from both the players and their parents before beginning the investigation.
Not applicable.
This research received no external funding.
The authors declare no conflict of interest. HN and EM-P are serving as the Guest editors of this journal. We declare that HN and EM-P had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to AT.
Publisher’s Note: IMR Press stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.