The Effects of Reverse Nordic Exercise Training on Measures of Physical Fitness in Youth Male Soccer Players

Abstract

This study aimed to evaluate the impact of an 8-week reverse Nordic exercise training (RNET) program on physical fitness in male youth soccer players. A total of 35 players participated in the study and were divided into two groups: the RNET group (n = 19, age 16.39 ± 0.46 years) and the active control group (CG: n = 16, age 16.53 ± 0.48 years). To assess fitness changes, participants were tested on linear sprint speed (5, 10, and 20 m sprints), change-of-direction (CiD) speed (505-CiD), vertical jump (countermovement jump [CMJ]), horizontal jump (standing long jump [SLJ]), drop jump (20 cm drop jump [DJ-20]), and repeated sprint ability (RSA). Significant group-by-time interactions were observed (effect size, [ES] = 0.70 to 1.37), with substantial improvements in the RNET group across linear sprint, CiD, and jumping performances (ES = 0.61 to 1.47), while no significant changes were noted in the CG. However, no significant group-by-time interactions were observed for RSA parameters. Individual response analysis revealed that 63–89% of RNET group exhibited improvements exceeding the smallest worthwhile change (SWC_0.2) threshold. These results suggest that the RNET program is both effective and safe for enhancing physical fitness in male youth soccer players.

1. Introduction

Eccentric training, characterized by muscle lengthening under tension, is a well-established approach for enhancing muscle strength and overall physical fitness (Douglas et al., 2017; Hessel et al., 2017; Hody et al., 2019; Roig et al., 2009). Eccentric training is widely recognized for eliciting greater increases in muscle mass and strength compared to other types of muscle actions, such as concentric, isometric, and concentric–eccentric actions (Baroni et al., 2015; Coratella et al., 2022; Hessel et al., 2017) while also being more energy-efficient (Hessel et al., 2017). Therefore, the combination of high force production and low energy cost makes eccentric training particularly well-suited for athletic development (Hessel et al., 2017). Moreover, eccentric training imposes substantial mechanical stress on muscles, stimulating adaptations that enhance both muscular and tendinous properties, ultimately contributing to improvement in overall physical fitness (Douglas et al., 2017).

In this regard, eccentric training plays a crucial role in practically all sports that incorporate jumping, running, and change in direction (CiD) speed, as it is a fundamental component of the stretch-shortening cycle (Chaabene et al., 2018; Vogt & Hoppeler, 2014). During training sessions and competitive matches, soccer players perform a wide range of repeated high-intensity actions, including jumping, kicking, accelerating, decelerating, sprinting, and changing directions at various angles (Stolen et al., 2005). Although growth and maturation naturally enhance key components of physical fitness, targeted training interventions can further optimize their development. Indeed, systematic participation in physical fitness programs during early childhood and adolescence has been shown to support long-term physical development (Lloyd et al., 2016), thereby enhancing the likelihood of athletic success in later stages of an athlete’s career. Eccentric training appears to be an effective tool toward this goal. However, its application with youth soccer players remains limited (Fiorilli et al., 2020; Moran et al., 2022), underscoring the need for further research and the development of practical guidelines to ensure its safety and effective integration into youth training programs.

Most earlier studies in soccer were conducted with young players (Fiorilli et al., 2020), often utilizing training tools such as the flywheel inertial device (Krommes et al., 2017). Furthermore, despite the well-established importance of Eccentric training, the variety of exercises available for integration into training programs remains relatively limited. This is particularly relevant in soccer, where eccentric movements are essential for meeting the sport’s physical demands, such as rapid decelerations, CiD speed, and high-intensity actions (Chaabene et al., 2018). Among the limited available studies, Moran et al. (2022) investigated the effects of Nordic hamstring exercise (NHE) training on various measures of physical fitness, revealing beneficial effects (effect size [ES] = 1.84 to 3.02) in male youth soccer players. Meanwhile, Negra et al. (2024) explored an innovative eccentric training method based on horizontal speed deceleration. The same authors revealed a large (ES = 1.01 to 1.70) effect of an 8-week horizontal speed deceleration training program on various measures of physical fitness including linear sprint, jumping ability, CiD speed, and repeated sprint ability among male youth handball players. Moreover, Bouguezzi et al. (2024) examined the effects of an 8-week eccentric training program centered on the Reverse Nordic Exercise (RNE) on physical fitness in youth karate athletes. In this context, RNE is a bodyweight movement that focuses on the eccentric contraction of the quadriceps and hip flexors, without engaging hip hinge mechanics. The authors reported positive effects, with small-to-large improvements in measures of linear sprint, jumping ability, and CiD speed.

To date, the potential of the RNE in youth soccer remains underexplored, and this study represents an important step toward understanding its role in improving physical performance in this population. To the best of our knowledge, aside from the work of Bouguezzi et al. (2024) with youth karate athletes, no prior study has previously investigated the effects of the RNE training (RNET) on measures of physical fitness in youth male soccer players. Therefore, this study aimed to assess the impact of RNET on various physical fitness parameters in male youth soccer players. We hypothesized that participants who completed the RNET program would display greater improvements in physical fitness compared to those who followed their routine soccer training (Bouguezzi et al., 2024).

2. Methods

2.1. Experimental Approach to the Problem

To assess the effects of an 8-week bi-weekly in-season RNET program on various measures of physical fitness, a team of postpubertal male soccer players was randomly assigned to either a RNET group or an active control group (CG). The RNET group replaced 10–20 min of low-intensity technical drills with the RNE protocol on two training days per week (Tuesdays and Thursdays), while the CG continued their regular training routine. The RNET was implemented right after the warm-up. Following the RNET portion of each session, players in the experimental group completed the rest of their regular soccer training. The study was conducted during the second half of the in-season period (March–May 2024) of the year, which consisted of five training sessions per week, each lasting approximately 90 to 120 min, along with one official match on the weekend. Training sessions typically included a combination of technical drills (e.g., passing, dribbling, and shooting) and tactical exercises (e.g., small- and large-sided games focused on positional play and team strategy). Two weeks before the initial assessments, participants attended a familiarization session to become acquainted with all testing procedures. Physical fitness assessments were conducted before and after the intervention to monitor changes. The test battery included linear sprint tests over distances of 5, 10, and 20 m; change in direction (CiD) speed measured using the 505 test; vertical jump evaluations via the countermovement jump (CMJ) and a 20 cm drop jump (DJ-20); horizontal jump performance assessed with the standing long jump (SLJ); and repeated sprint ability (RSA). Testing was conducted over three separate days: on the first day, linear sprint speed and CiD speed tests were performed, jump testing took place on the second day, and RSA testing was conducted on the third day. Each testing session lasted approximately 60 to 90 min. All evaluations were consistently conducted by a co-author and expert in the field of strength and conditioning (YN), ensuring a high level of proficiency and consistency All assessments were performed under standardized conditions between 7:00 and 8:30 a.m., at least 48 h after the last training session or match, and under stable environmental conditions (temperature 16–18 °C, no wind).

Participants

Figure 1 displays a Consolidated Standards of Reporting Trials (CONSORT) diagram of the levels of reporting and participant flow for the study (Schulz et al., 2010). We performed an a priori sample size calculation for the SLJ test, contrasting intervention and control groups. We set the type I error rate at 0.05 and the statistical power at 80%. The estimated effect size of Cohen’s d = 0.97 is based on a recent study by (Abdelkader et al., 2022). The analysis indicated that a total of 17 participants would represent a sufficient sample. To account for potential participant attrition, thirty-five postpubertal male soccer players were randomly assigned by blinded evaluators to either the experimental training (ET) group (n = 19; age = 16.39 ± 0.46 years; maturity offset = 2.43 ± 0.47 years) or the active control group (CG) (n = 16; age = 16.53 ± 0.48 years; maturity offset = 2.51 ± 0.40 years). All participants were considered experienced players with an average of 6.0 ± 1.8 years of systematic soccer training, which included four to five training sessions per week. Additionally, all participants were in good health and had no musculotendinous injuries during the season leading up to the start of the study. All participants were engaged in their regular in-season soccer training and competition schedules throughout the study period. No additional organized physical activities were permitted outside of the intervention and regular soccer training. All anthropometric measurements (

Table 1. Anthropometric characteristics of the included participants.

	RNET Group (n = 19)	CG (n = 16)
Age (years)	16.39 ± 0.46	16.53 ± 0.48
Height (cm)	176.21 ± 6.46	175.94 ± 5.16
Body mass (kg)	63.42 ± 6.10	65.70 ± 7.55
Maturity offset (years) *	2.43 ± 0.47	2.51 ± 0.40
Playing experience (years)	6.1 ± 1.7	5.9 ± 1.9

Notes: Data are presented as means and standard deviations; RNET = Reverse Nordic exercise training; CG = control group; *: as years from peak height velocity.

) were conducted by the same researcher to ensure consistency. Body height was measured from the floor to the vertex of the head using a stadiometer (accuracy: 0.1 cm; Hotain, London, UK), and body mass was recorded using a digital scale (accuracy: 0.1 kg; Tanita BF683W, Munich, Germany). During all measurements, participants were barefoot and wore shorts only.

The maturity offset method was used to assess the biological maturity of participants (Moore et al., 2015). The following prediction equation was applied:

Maturity offset = −7.999994 + (0.0036124 × age × height).

Prior to the start of the study, all procedures and any potential risks were thoroughly explained. Written informed consent was obtained from parents or legal guardians, along with assent from the participants. The study protocol received approval from the Institutional Review Committee of the Higher Institute of Sport and Physical Education of Ksar Saïd (approval number: 201D0045, 19 January 2024) and was carried out in accordance with the most recent revision of the Declaration of Helsinki.

2.2. Linear Sprint Speed Time

Twenty-meter linear sprint performance was assessed at 5, 10, and 20 m intervals using a single beam electronic timing system (Wittygate, Microgate, SRL, Bolzano, Italy). Participants started in a standing split stance position with their lead foot 0.3 m behind the first infrared photoelectric gate, which was placed 0.75 m above the ground to ensure that it captured trunk movement and avoided false signals through limb motion (Nikolaidis et al., 2016). In total, four single-beam photoelectric gates were used. No rocking or false steps were permitted before starting. The between-trial recovery time was three minutes. The best performance out of two trials was used for further analysis. The between-trial intraclass correlation coefficient (ICC) was 0.88, 0.95, and 0.91 for the 5 m, 10 m, and 20 m sprint, respectively.

2.3. The 505 Change in Direction Test

The 505 CiD speed test was administered using the protocol previously outlined by (Negra et al., 2024). using an electronic timing system (Microgate, Bolzano, Italy). Players assumed a standing position 10 m from the start line, ran as quickly as possible through the start/finish line, pivoted 180° at the 15 m line indicated by a cone marker, and returned as fast as possible through the start/finish line. To ensure proper execution of the test, a researcher was positioned at the turning line and if the participant changed direction before reaching the turning point, or turned off the incorrect foot, the trial was disregarded and reattempted after the recovery period. A between-trial rest period of three minutes was provided. The best performance out of two trials was used for further analysis. The between-trial ICC was 0.98.

2.4. Countermovement Jump

During the CMJ, participants started from a standing position and performed a fast downward movement by flexing the knees and hips before rapidly extending the legs and performing a maximal vertical jump (Hulse et al., 2013). During the test, participants were instructed to maintain their arms akimbo. Jump height was recorded using an optoelectric system (Optojump, Microgate, SRL, Bolzano, Italy). A rest period of 1 min was allowed between trials. This was consistent across all trials for all jump-hop tests. The best out of two trials was retained for further analysis. The between-trial ICC was 0.91.

2.5. Standing Long Jump

The participants positioned behind the starting line in a standing position with feet shoulder-width apart line and arms loosely hanging down. On the command ready, set, go, participants executed performed a fast downward movement before jumping at maximal effort in the horizontal direction. Participants had to land with both feet at the same time and were not allowed to fall forward, sideward, or backward. The horizontal distance between the starting line and the heel of the rear foot was recorded via a tape measure to the nearest 1 cm (Ramírez-Vélez et al., 2017). Three trials were performed, and the best time was taken for further analysis with a rest period of 90s was allowed between trials. The between-trial ICC was 0.87.

2.6. 20 cm Drop Jump

From a 20 cm platform above the ground, participants positioned standing with arms akimbo and the leading leg straight to avoid any initial upward propulsion. To ensure the validity of the test, the participants were instructed to leave the platform with knees and ankles fully extended and to land in a similarly extended position with short ground contact followed by a maximal vertical jump thrust (Prieske et al., 2019). Three trials were performed, and the best time was taken for further analysis with a rest period of 90s was allowed between trials. The between-trial ICC was 0.86.

2.7. Repeated Sprint Ability Test

The RSA test was assessed via the same photocell system used for the linear speed and 505 CiD tests (Microgate, Bolzano, Italy). Immediately after a standardized warm-up, participants completed a preliminary single shuttle-sprint test (20 + 20 m with 180° CiD). The first trial provided the criterion score for the actual shuttle-sprint test (Padulo et al., 2015). Participants then rested for five minutes before starting the RSA test. During the first sprint, participants had to achieve at least 97.5% of their criterion score, otherwise, they rested for five minutes and then restarted the test (Padulo et al., 2015). We used such an approach to determine if participants adopted a coping strategy for performance. Of note, all participants attained their criterion score during the first sprint. All performed six 20 m shuttle sprints with 180° turns, separated by 25 s of passive recovery (Padulo et al., 2015). Three seconds prior to the commencement of each sprint, players were asked to adopt the ready position using a split stance, with their front foot 0.3 m behind the starting line, until the next start signal. From the starting line, they sprinted for 20 m and touched the second line with one foot before performing a 180° CiD and returning to the starting line as quickly as possible. Participants were instructed to complete all sprints as fast as possible. The RSAbest time, RSA average time, and RSAtotal time were determined. Due to the fatigue induced by the test, only one maximal attempt was made, i.e., no ICC was calculated.

2.8. The Eccentric Training Program

The RNET program was designed and implemented by two co-authors (YN and HC), both experienced practitioners and researchers in the fields of sports science. The first author supervised all training sessions, ensured proper exercise execution, and provided standardized instructions throughout the 8-week intervention. A work-to-rest ratio of 1:10 was implemented, as outlined in

Table 2. Reverse Nordic exercise training program.

Week	Session per Week	Sets	Reps	Eccentric Phase Duration (s)	Work-to-Rest Ratio	Recommendation
1	2	2	6	3–5	~1–10	Ensure smooth, controlled movement with emphasis on knee and hip alignment.
2	2	4	6	3–5	~1–10
3	2	4	6	3–5	~1–10	Maintain control and form while progressively increasing the range of motion
4	2	4	8	3–5	~1–10
5	2	4	10	3–5	~1–10	Increase the range of motion while maintaining proper knee tracking
6	2	4	10	3–5	~1–10
7	2	4	10	3–5	~1–10	Full range of motion with good technique
8	2	4	10	3–5	~1–10

. Players were instructed to perform the eccentric phase of each movement over a duration of 3 to 5 s. While this tempo was not rigorously enforced, consistent verbal cues emphasized a controlled descent to the point of maximal inclination, ensuring correct hip and knee alignment. The intervention did not involve any form of external assistance or additional load. Progression was achieved by systematically increasing the number of sets and repetitions throughout the 8-week period. The structure of the ET protocol is presented in Table 2 and aligns with the methodologies applied in two recent studies by Bouguezzi et al. (2024, 2025), which utilized comparable training designs in similar athlete populations. The RNET was performed at the beginning of the soccer training session. Before every eccentric training session, a standardized 8–12 min warm-up was completed including low-intensity running, coordination exercises, dynamic movements (i.e., lunges, skips), sprints, and dynamic stretching for the lower-limb muscles. At the beginning of each training week, the first eccentric training session was performed at least 48 h after the last soccer match that was scheduled on the weekend. The second ET session was completed 48 h after the first session (i.e., Tuesday and Thursday).

2.9. Statistical Analyses

Data are presented as means and standard deviations (SD). Normality assumption was tested using the Shapiro–Wilk test. To establish the effect of the interventions on the dependent variables, a 2 (group: ET and CG) × 2 (time: pre, post) ANOVA with repeated measures was determined for each parameter. When group × time interactions reached the level of significance (i.e., significant F value), group-specific post hoc tests (i.e., paired t-tests) were used. To determine the magnitude of the training effect, ES were determined by converting partial eta-squared to Cohen’s d. According to Hopkins et al. (2009), ES values are classified as trivial (<0.2), small (0.2–0.6), moderate (0.6–1.2), large (1.2–2.0), very large (2.0–4.0), and extremely large (>4.0). Between-trial reliability was assessed using the ICC. The smallest worthwhile change (SWC) was calculated as 0.2 * SD pooled where SD represents the pooled standard deviation of pre-training scores. The alpha level of significance was set at p < 0.05. All data analyses were performed using SPSS 25.0 (SPSS, Inc., Chicago, IL, USA).

3. Results

The findings of this study are included in

Table 3. Measures of physical fitness before and after training in both groups.

	RNET (n = 19)					CG (n = 16)					ANOVA
	Pretest		Posttest			Pretest		Posttest			p-Value (ES)
	M	SD	M	SD	Δ	M	SD	M	SD	Δ	Time	Group × Time
Linear sprint speed
5 m sprint (s)	1.17	0.06	1.06	0.08	9.10	1.18	0.09	1.15	0.10	2.69	<0.001 (1.35)	<0.05 (0.73)
10 m sprint (s)	1.94	0.07	1.82	0.10	6.40	1.90	0.09	1.86	0.08	2.04	<0.001(1.36)	<0.05 (0.71)
20 m sprint (s)	3.29	0.15	3.16	0.12	4.10	3.18	0.12	3.16	0.22	0.42	<0.01(0.85)	<0.05 (0.70)
Change in direction speed
505-CiD speed test (s)	2.44	0.11	2.35	0.09	3.62	2.46	0.09	2.45	0.10	0.32	<0.01 (1.20)	<0.01 (1.01)
Muscle Power
CMJ (cm)	28.94	4.31	31.67	3.94	9.44	33.82	4.18	34.81	4.63	2.92	<0.001 (1.73)	<0.05 (0.81)
SLJ (cm)	1.85	0.19	2.04	0.12	10.21	1.91	0.20	1.92	0.17	0.62	<0.001 (1.55)	<0.001 (1.37)
DJ-20 (cm)	27.92	5.58	30.27	3.76	8.45	32.99	4.83	32.84	4.77	0.45	0.08(0.63)	<0.05 (0.71)
Repeated sprint ability
RSAbest (s)	7.22	0.47	7.22	0.27	0.01	7.12	0.17	7.19	0.17	0.78	>0.05 (0.22)	>0.05 (0.22)
RSAmean (s)	7.52	0.29	7.47	0.29	0.68	7.40	0.23	7.37	0.35	0.37	>0.05 (0.36)	>0.05 (0.11)
RSAtot (s)	45.10	1.73	44.79	1.73	0.68	44.37	1.38	44.21	2.12	0.37	>0.05 (0.36)	>0.05 (0.11)

RNET: reverse Nordic exercise training, CG: control group, CiD: change in direction, CMJ: countermovement jump, SLJ: standing long jump, DJ-20; Drop jump 20 cm, RSA: repeated sprint ability; RSA_best: repeated sprint ability best time; RSA_mean: repeated sprint ability mean time; RSA_tot: repeated sprint ability total time.

. The adherence rate to training was 96% for both groups. There were no statistically significant baseline differences between groups (all p > 0.05) with regard to chronological age, height, body mass, maturity offset, APHV, soccer experience (i.e., training background) (Table 1). No baseline between-group differences were detected for all physical fitness variables. No injuries were documented during or following the intervention, indicating the safety of this training approach.

3.1. Linear Sprint-Time

Findings indicated significant group × time interaction for the 5, 10, and 20 m sprint test (F_(1,33) = 4.41, 4.17, and 3.98, ES = 0.73, 0.71, and 0.70, respectively, all p < 0.05) (Table 3). For the 5 m sprint performance, the post hoc analyses showed a large pre–post-performance improvement in the ET group (∆9.10%; ES = 1.24; p < 0.001). Similarly, post hoc analysis revealed moderate improvements in 10 and 20 m sprint test (∆6.40, 4.10%; and ES = 0.92 and 0.74, respectively, all p < 0.01). The CG revealed non-significant changes from pre-to-post-test in all sprint intervals (ES = 0 to 0.25; all p > 0.05)

In terms of the individual responses, results indicated that 89% of the ET group (n = 17) improved 5 and 10 m sprint performance to a level that was greater than the SWC_0.2. Regarding 20 m sprint, 68% (n = 13) of the ET improved their performance to a level that exceeds the SWC_0.2. Regarding the CG, 56% (n = 9), 43% (n = 7), and 43% (n = 7) improved 5-, 10-, and 20 m sprint performance to a level that exceeds the SWC_0.2 thresholds.

3.2. Jumping Performance

For CMJ-height, a significant group × time interaction (F_(1,33) = 5.46, ES = 0.8, p < 0.05) was observed (Table 3). Post hoc analyses demonstrated a moderate CMJ-height improvement for the ET group (∆9.44%; ES = 1.18; p < 0.001). The CG revealed no significant pre-to-post change (∆2.92%; ES = 0.46; p > 0.05).

In terms of individual responses, our statistical calculations showed that 78% of the ET group (n = 15) improved the CMJ-height to a level that exceeded the SWC_0.2 with only 31% (n = 5) of the CG.

For the SLJ test, a significant group × time interaction (F_(1,33) = 15.38, ES = 1.37, p < 0.01) was noted. Post hoc analyses indicated a moderate pre-to-post improvement for the ET group (∆10.21%; ES = 1.47; p < 0.01). However, the CG did not reveal any significant change (∆0.62%; ES = 0.12 p > 0.05). Additionally, results revealed that 78% of the ET group (n = 15) improved the SLJ performance to a level that was greater than the SWC_0.2 with only 25% (n = 4) of the CG.

Regarding the DJ-20, a significant group × time was observed (F_(1,33) = 4.18, ES = 0.71; p < 0.05). Post hoc analyses pointed to a moderate DJ-20 performance improvement for the ET group (∆7.15%, ES = 0.61; p < 0.05) while the CG failed to reach a significant pre-to-post change (∆0.76%; ES = 0.10; p > 0.05). Moreover, results revealed that 63% of the ET group (n = 12) improved the DJ-20 test performance to a level that was greater than the SWC_0.2 sprint test with only 12.5% (n = 2) of the CG.

3.3. Change in Direction Test

Results indicated a significant group × time interaction (F_(1,33) = 8.43, ES = 1.20, p < 0.01) (Table 3). Post hoc analyses demonstrated a moderate 505 CiD performance improvement from pre-to-post for the ET group (∆−3.62%; ES = 0.98; p < 0.01). No significant pre-to-post changes were found for the CG (∆0.32%; ES = 0.11; p > 0.05). Our statistical calculation showed that 78% of the ET group (n = 15) improved 505 CiD performance to a level that was greater than the SWC_0.2. However, only 50% (n = 8) of the CG have shown improvement in the 505 CiD performance that exceeds the SWC_0.2.

3.4. Repeated Sprint Ability

Results indicated no significant group × time interaction for all the RSA outcomes (F_(1,33) = 0.39, 0.98, and 0.98, ES = 0.22, 0.11, and 0.11 for the RSA_{best time}, RSA_{mean time}, and RSA_{total time}, respectively; all p > 0.05). In addition, 47% of the ET group (n = 8) improved the RSA_best performance to a level that was greater than the SWC_0.2. For the RSA_{mean time} and RSA_{total time}, 57% of the ET group (n = 11) showed a level of improvement that exceeds the SWC_0.2. Regarding the CG, our statistical analysis revealed that 6.25% (n = 1), 18.75% (n = 3), and 18.75% (n = 3) improved RSA_best, RSA_{mean time}, and RSA _total, respectively, to a level that exceeds the SWC_0.2

4. Discussion

This study examined the effects of an 8-week RNET program on various components of physical fitness in postpubertal male soccer players. The main finding indicated that an in-season RNET executed alongside regular soccer-specific training generated beneficial effects on measures of linear sprint, CiD speed, and jumping performance in postpubertal male soccer players. Conversely, soccer-specific training alone did not yield any significant changes in the same fitness measures.

Linear sprint performance is critical and plays a crucial role in soccer games, as quick bursts of speed are often necessary during decisive moments, such as scoring a goal (Faude et al., 2012). The findings of this study revealed that 8 weeks of RNET resulted in moderate-to-large within-group improvements in short sprint performance over distances up to 20 m (5 m: ES = −1.24 [−9.10%]; 10 m: ES = −0.92 [−6.40%]; 20 m: ES = 0.74 [−4.10%]) with no significant changes in the CG. Additionally, 68 to 89% of the RNET group improved linear sprint performance across all distances to a level that exceeded the SWC_0.2 compared to only 43 to 56% in the CG. These results are in agreement with those of Bouguezzi et al. (2024) who showed a large (∆10.09%; ES = 1.63) pre–post-performance improvement in the 10 m sprint performance following an 8-week of RNET in youth karate athletes. Further, Krommes et al. (2017), demonstrated that 10 weeks of Nordic hamstring exercise (NHE)-based training results in significant linear sprint performance improvement in elite male soccer players (i.e., 5 m [Δ7.5%] and 10 m [Δ4.6%]) with no changes in CG. Likewise, Moran et al. (2022) reported moderate improvements in 10 (ES = 0.84) and 40 m (ES = 0.89) sprint performance following 8 weeks of NHE training in male youth soccer players aged 16 years. Although not directly assessed in this study, it is plausible to assume that the observed improvements could be attributed to enhanced strength and power of the quadriceps and hip flexor muscles, which are essential for short-distance running (Sugiura et al., 2008). Furthermore, RNET likely boosts the nervous system’s efficiency in activating muscles, leading to improved motor unit recruitment and firing rates (Alonso Fernández et al., 2019). This enhancement can result in increased force production during sprinting.

Regarding CMJ-height performance, our results suggest that RNET induced a large improvement (∆9.44%; ES = 1.18). However, no change was noted in the CG. Individual performance analysis indicates that 78% of the RNET group improved their CMJ performance to a level that exceeded the SWC_0.2 compared to only 31% in the CG. Likewise, the RNET program induced large (10.21%; ES = 1.47) and small (∆7.15%, ES = 0.35) improvements in SLJ and DJ-20, respectively. The pre–post individual changes indicate that 78% of the RNET group (n = 15) improved the SLJ test performance to a level that was greater than the SWC_0.2 with only 25% (n = 4) of the CG. In terms of the DJ-20, 63% of the RNET group (n = 12) improved their performance to a level that exceeded SWC_0.2 with only 12.5% (n = 2) of the CG. These results are in line with those recently reported by Bouguezzi et al. (2025), who reported a small CMJ-height improvement (d = 0.35) following an 8-week eccentric training program comprising NHE and RNE. Additionally, Maroto-Izquierdo et al. (2017) conducted a systematic review and meta-analysis, revealing significant differences in training-induced adaptations favoring eccentric overload flywheel resistance training compared to a control condition, with an effect size of 0.46 in vertical jumping performance. There is evidence that the RNE is an effective eccentric exercise that targets quadriceps strength (Weldon et al., 2025). Although not directly assessed in this study, it is plausible to argue that the observed improvements in jump performance are associated with enhanced strength and power of the quadriceps. This could partly be due to improved motor unit recruitment and firing rates, thereby enhancing force and power capabilities.

The ability to quickly change direction while sprinting is a key performance characteristic in youth male soccer players (Mirkov et al., 2010). Our results showed a moderate improvement in CiD performance (∆−3.62%; ES = 0.98) following RNET. In comparison, there were no changes in the CG. Interestingly, 78% of participants in the RNET group showed improvement in their CiD speed performance exceeding the SWC_0.2, while only 50% of those in the CG demonstrated similar progress. Supporting these findings, Moran et al. (2022) reported a large effect (ES = 1.04) in 505 CiD performance after an 8-week NHE-based training integrated into regular soccer training, compared to regular training with no NHE. More recently, Chaabene et al. (2022) conducted a systematic review and meta-analysis examining the effects of flywheel resistance training versus traditional resistance training on CiD speed performance in male athletes. Their findings indicate that both training methods are effective; however, flywheel resistance training demonstrated larger effects (ES = 0.64) compared to traditional resistance training. Mounting evidence highlights the importance of eccentric strength of knee extensors for optimal CiD performance, particularly during the deceleration phase (Chaabene et al., 2018). The observed improvements in CiD speed in the current study likely stem from the well-documented benefits of eccentric training on knee extensor strength (Alonso Fernández et al., 2019; Brughelli et al., 2010). Greater knee extensor strength facilitates more effective braking, which is crucial for quick horizontal deceleration during CiD movements. In fact, rapid deceleration is a key factor for executing successful CiD maneuvers, allowing a quicker re-acceleration in a new direction, resulting in better overall CiD speed performance (Dos’Santos et al., 2018). Moreover, research indicates that increased eccentric strength leads to improved joint stability and more efficient force transfer through the joints, further optimizing CiD performance (Ramirez-Campillo et al., 2021).

This study is the first to report the effect of RNET on RSA parameters in male youth soccer players. The findings of this study revealed no improvements in all RSA parameters (i.e., RSA_best, RSA_average, and RSA_total) after an 8-week RNET. While RNE primarily targets the strength and power of knee extensors and hip flexors, these adaptations may not directly transfer to specific speed and endurance improvements (Taylor et al., 2015). RNE focuses on eccentric strength and muscle control, which are important but might not translate effectively to the repeated rapid accelerations and decelerations involved in RSA. Other training modalities that emphasize sprinting mechanics and aerobic conditioning might be more effective for improving RSA.

This study has a number of limitations that readers should consider. Firstly, although the overall training exposure was comparable between the RNET group and the CG, it would have been beneficial to monitor training load throughout the 8-week period using external measures (e.g., total distance covered) and/or internal measures (e.g., rate of perceived exertion). Second, the eccentric strength of the knee extensors was not directly assessed, for example, using an isokinetic device. This limitation should be addressed in future studies. Third, our research did not incorporate any direct physiological measures (e.g., electromyography) or biomechanical assessments (e.g., vertical ground reaction force). These limitations render the statements about the mechanisms of adaptations rather speculative. Fourth, players’ positions were not considered in the randomization or analysis. Given the potential impact of positional demands on performance and training responses, this factor may have introduced variability in the outcomes and should be controlled in future studies. A further limitation is the lack of monitoring of perceptual and motivational factors, which are highly relevant for understanding participants’ engagement and for the practical implementation of the program in real-world settings. Future research should consider including these dimensions to better understand the feasibility and long-term sustainability of the intervention. Furthermore, the duration of the RNET sessions was relatively short. As such, future studies should explore the effects of varying session durations on performance outcomes. It is also recommended that future investigations examine the medium- and long-term effects of RNET interventions, as well as their applicability across different age groups and in female athletes, to enhance the generalizability of the findings.

5. Conclusions

The implementation of a short-duration (i.e., 8 weeks) RNET alongside regular soccer-specific training resulted in notable improvements in various measures of physical fitness, including sprinting, jumping, and CiD speed, exceeding the gains achieved with soccer-specific training alone. Importantly, no injuries were reported during or after the intervention, highlighting the safety of this training approach. These findings underscore the benefits and safety of incorporating RNET into the training regimen of youth soccer players during the competitive season.