Effects of unilateral and bilateral eccentric training on lower limb maximal strength, jump, and sprint performance in trained adolescent male basketball players: a randomized controlled trial

Abstract

Background

Basketball requires a solid foundation of lower limb strength and explosive power. Unilateral and bilateral resistance training modes may differ in improving sport-specific performance, but their relative effects and potential advantages in trained adolescent athletes remain unclear. This pilot study aims to explore the application effects of these two methods on specific performance indicators.

Objective

This exploratory study aimed to compare the effects of a seven-week unilateral eccentric training (UET) and bilateral eccentric training (BET) on lower limb maximal strength, jump, and sprint performance in trained adolescent male basketball players.

Methods

Twelve participants (n=12) were randomly assigned to either a UET or BET group. In addition to completing six hours of regular basketball training per week, both groups performed unilateral or bilateral eccentric resistance training three times a week for seven weeks. Three categories of performance indicators were comprehensively assessed pre- and post-intervention: 1) lower limb maximal strength (estimated values: Advantage Unilateral Bulgarian Squat [AUBS], Advantage Unilateral Deadlift [AUD], Bilateral Squat [BS], and Bilateral Deadlift [BD]); 2) jump performance (Standing Triple Jump [STJ], Standing Advantage Unilateral Reach Height [SAURH], and Standing Bilateral Reach Height [SBRH]); and 3) sprint performance (30-meter sprint). A Bayesian mixed-effects model was used for an exploratory analysis of the posterior probabilities of performance changes.

Results

One. Between-group analyses revealed that neither the UET nor the BET group demonstrated a statistically significant between-group advantage across all tested indicators (posterior probabilities < 97.5%). 2. Regarding lower limb maximal strength, within-group exploratory analyses indicated that only the UET group showed a high probability of improvement in the estimated AUD (98.25% probability, g = 12.01，95% CrI [1.04 – 22.87]). The other strength indicators (AUBS, BS, BD) did not exhibit clear high-probability improvement trends in either group. 3. Regarding jump performance, the UET group demonstrated extremely high improvement potential in SAURH (99.74% probability, g = 9.12, 95% CrI [3.41 – 14.81]) and SBRH (98.43% probability, g = 7.60, 95% CrI [0.96 – 14.29]). Concurrently, the BET group also showed a high probability of improvement in SBRH (97.56% probability, g = 6.76, 95% CrI [0.04 – 13.79]). However, neither group displayed a high-probability adaptive enhancement in the STJ metric. 4. Regarding sprint performance, only the UET group exhibited a high probability of improvement in the 30-meter sprint time (99.03% probability, g = 0.27, 95% CrI [0.05 – 0.48]), whereas the BET group did not present a similar high-probability trend.

Conclusion

Although both UET and BET interventions demonstrated within-group potential to improve specific sports performance metrics (e.g., partial maximal strength, jump, and sprint capabilities) in trained adolescent basketball players, neither modality established a statistically significant absolute between-group advantage. Future large-scale studies, strictly controlling for training volume and concurrent sport-specific loads, are required to further validate these exploratory findings.

Trial registration

(Retrospectively registered): Chinese Clinical Trial Registry, Registration number: ChiCTR2500114472, Reg Date: 2025-12-12.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13102-026-01691-z.

Introduction

Basketball involves frequent repetitive actions such as jumping, sprinting, turning, accelerating, decelerating, and change of direction (COD) [1], all of which impose substantial eccentric loads on athletes. Consequently, eccentric strength is essential for basketball players. Recent studies have consistently demonstrated that eccentric training can enhance lower-limb explosive power and overall performance in basketball players [2–5]. The favorable adaptations observed during eccentric exercise may be attributed to increased cross-bridge formation, greater tendon stiffness, reduced active motor unit recruitment, and selective activation of fast-twitch muscle fibers [6]. Moreover, eccentric training has been shown to reduce the risk of tendon injuries [7]. Therefore, it is a critical component in improving performance and preventing injuries in basketball players.

Basketball exhibits a high degree of both unilateral and bilateral movement patterns. Skills such as layups, running, blocking, and changing direction are predominantly executed using unilateral force. Research has indicated that unilateral training is well aligned with the specific demands of the sport [8]. Unilateral training involves exercises that utilize only a single-limb movement pattern [9]. Compared with bilateral training, unilateral training can more effectively improve strength, jumping ability, sprint performance, and balance [10–13]. In particular, it has a greater impact on enhancing unilateral explosive power [9, 14, 15]. This may be because unilateral training provides a smaller base of support for the lower limbs, requiring greater multi-joint neuromuscular coordination and stability. Furthermore, unilateral exercises engage stabilizing and synergistic muscles to a greater extent, thereby promoting a more stable state during unilateral movements [16–19]. The widespread application of unilateral training may also be attributed to the cross-transfer theory, which posits that resistance training of one limb can enhance the strength and skills of the contralateral, untrained limb [20]. In practical settings, this allows coaches to design more targeted and diverse lower-limb performance programs for athletes.

Additionally, the bilateral force deficit theory underscores the distinctive advantages of unilateral training. The bilateral force deficit refers to the phenomenon in which the maximal force generated during bilateral contraction is less than the combined maximal force of the left and right limbs contracting individually [21, 22]. Consequently, unilateral training can play an important role in preventing and correcting bilateral force imbalances [23, 24]. Despite its unique advantages, the effectiveness and underlying mechanisms of unilateral versus bilateral training remain subjects of debate within the academic community. Moreover, research findings on unilateral and bilateral eccentric training vary considerably [13], and few studies have specifically examined their effects on basketball players. Therefore, the present study aims to provide meaningful insights and practical recommendations.

This study aimed to compare the effects of a 7-week unilateral versus bilateral eccentric resistance training on lower limb maximal strength, jump, and sprint performance in trained adolescent male basketball players. We hypothesized that: (1) both UET and BET would elicit a high probability of within-group improvements in maximal strength, jump, and sprint performance; and (2) UET would induce a greater magnitude of improvement compared to BET in between-group comparisons.

Methods

Participants

Based on sample sizes from previous related studies [25], 16 school team members volunteered to participate in this experiment. The inclusion criteria were strictly defined as follows: (1) full-time enrolled students; (2) at least 2 years of structured resistance training experience; and (3) no history of severe illness or musculoskeletal injury within the past 6 months. Exclusion criteria included: (1) an intervention attendance rate of less than 80%; and (2) any injury sustained during the study period that could affect athletic performance.

During the baseline screening, 4 players were excluded due to potential injury risks or failure to meet the criteria. Ultimately, 12 trained adolescent male basketball players were randomly assigned to either the unilateral eccentric training (UET) group or the bilateral eccentric training (BET) group. All participants had over 4 years of basketball-specific training experience and more than 2.5 years of resistance training history, having secured top-three finishes in provincial tournaments that same year. According to the participant classification framework proposed by McKay et al., the subjects in this study were explicitly classified as “Tier 2: Trained/Developmental” [26].

During the 7-week intervention period, all participants also completed routine basketball-specific training and matches three times per week, lasting 2 h per session (i.e., concurrent training). Aside from the prescribed experimental intervention and scheduled basketball training, participants were instructed to refrain from any additional strenuous physical conditioning. Prior to the commencement of the study, all participants and their legal guardians were fully informed of the experimental procedures, potential risks, and expected benefits, and provided written informed consent. This study strictly adhered to the ethical guidelines of the Declaration of Helsinki and was formally approved by the Ethics Committee of Beijing Sport University (approval number: 2024433 H) (Table 1).

Table 1.

Participants baseline characteristics

Parameter	UET(n = 6)	BET(n = 6)
Age (years)	16.16 ± 0.75	15.30 ± 0.51
Height (cm)	182.16 ± 5.15	180.16 ± 6.82
Body mass (kg)	78.66 ± 11.39	79.33 ± 17.95

Abbreviations: UET Unilateral Eccentric Training group, BET Bilateral Eccentric Training group

An a priori power analysis was conducted using G*Power software (Version 3.1, Kiel University, Germany) to determine the appropriate sample size(Two-tailed). Based on previous literature investigating similar interventions [25], an expected effect size of d ≥ 0.90 for the primary outcome was utilized. With an alpha level of 0.05 and a desired statistical power of 0.80, the analysis indicated that a minimum sample size of 11 participants was required. To ensure sufficient statistical power while accounting for potential dropouts, a total of 12 participants (n = 12) were recruited for this study.

Experimental design

The entire experimental period spanned 10 weeks. Prior to the formal intervention, all 12 participants completed a 1-week familiarization period to fully adapt to the training protocols and standardized testing procedures. Subsequently, in strict accordance with the CONSORT (Consolidated Standards of Reporting Trials) guidelines, the participants were randomly assigned to either the unilateral eccentric training (UET) group or the bilateral eccentric training (BET) group, utilizing a random sequence generation and allocation concealment mechanism via Excel software. Given the inherent nature of exercise interventions, participants were fully aware of their assigned training content, rendering a double-blind setup unattainable. Consequently, this study employed a single-blind design, where blinding was strictly implemented exclusively for the researcher (K.X) responsible for test supervision and data analysis. Both groups completed the 7-week intervention (three sessions per week) under the direct supervision of a strength and conditioning coach at the same facility using identical equipment. During each session, the UET group specifically performed unilateral eccentric resistance training, whereas the BET group executed bilateral eccentric resistance training. Comprehensive assessments of lower limb maximal strength, jump, and sprint performance were conducted for both groups immediately before and after the intervention period.

Training protocol

Given that the participants were in the developmental stage of adolescence and required to complete 6 h of concurrent basketball-specific training weekly, the load design of this study specifically aimed to mitigate the risk of overtraining. All training sessions commenced with a standardized 10-minute warm-up, comprising core activation (planks, bodyweight lunges, single-leg balance with eyes closed) and dynamic stretching (knee-to-chest calf raises, single-leg Superman stretches, the “World’s greatest stretch,” and walking stretches).To isolate the eccentric stimulus and minimize neuromuscular interference from concentric contractions, both groups actively performed only the eccentric phase (requiring a controlled, uniform descent over 5 s) for all exercises. The subsequent concentric phase was entirely assisted and lifted by training partners. Regarding training volume and set configuration, this study referenced the experimental design logic of Zhao et al. [27]: The UET group performed 5 sets per leg for each unilateral exercise (i.e., 5RM × 5 reps × 5 sets per leg, totaling 10 sets per exercise); the BET group performed 5 sets of bilateral exercises (5RM × 5 reps × 5 sets). Rest periods were 2 min between sets and 3 min between exercises, followed by standardized post-training cool-down stretches.

It must be candidly acknowledged that this design—intended to ensure that the unilateral limb in the UET group received an identical localized set volume as the BET group—inevitably resulted in a mismatch in the total absolute training volume between the two groups (the total number of sets in the UET group was doubled). This inequality in training volume constitutes a significant methodological limitation in the experimental control of this study. (Table 2).

Table 2.

Intervention training protocols

Group	UET	BET
Content	Left-SEBS X5, Right-SEBS X5, Left-SED X5, Right-SEDX5	DES X5, DED X5
Load and sets	5RM 5SET	5RM 5SET
Rest intervals	Rest for 2 min between sets and 3 min between exercises
Requirements	The eccentric contraction part is required to be completed in 5 s. The concentric part is lifted with the help of a partner.

Abbreviations: UET Unilateral Eccentric Training group, BET Bilateral Eccentric Training group, SEBS Single-leg eccentric bulgaria squat, SED Single-leg eccentric deadlift, DES Double-leg eccentric squat, DED Double-leg eccentric deadlift

Testing methods and procedures

Before and after the 7-week intervention, participants underwent comprehensive assessments encompassing lower limb maximal strength, jump, and sprint performance. Prior to testing, participants completed a standardized warm-up of a 5-minute jog and 5 min of dynamic stretching. They were also required to abstain from alcohol for 48 hours and avoid caffeine or any supplements for 8 hours before testing.

Lower limb maximal strength

It is crucial to clarify that, to align mechanically with the “eccentric-only” intervention modality, all maximal strength assessments (both unilateral and bilateral) in this study evaluated solely the eccentric (lowering) phase of the movements. Participants did not perform the concentric force-production phase to stand up. The traditional Mayhew formula:1RM = 52.2 + 41.9e^(-0.055 × reps) [28]. was used to estimate maximal eccentric strength. Given that this formula was originally established based on complete concentric-eccentric bilateral compound movements, applying it to estimate pure eccentric and unilateral asymmetrical movements (e.g., Bulgarian split squats) presents inherent limitations in construct validity. Therefore, the 1RM data obtained in this section can only be regarded as a “submaximal eccentric resistance load index.”

Advantage Unilateral Bulgarian Squat (AUBS)

Participants warmed up with moderate weights for 2–3 sets. Standing with feet shoulder-width apart, the dominant leg was positioned forward while the non-dominant leg rested on a 30-cm box, with the barbell across the middle trapezius. Participants executed only a controlled eccentric squat until the thigh of the dominant leg was parallel to the floor (the rear thigh also had to lower in tandem). If significant instability occurred, three spotters immediately lifted the barbell to terminate the test. Participants had 3 attempts; the optimal weight and repetitions from the eccentric control were recorded for estimation.

Advantage Unilateral Deadlift (AUD)

Participants stood on their dominant leg with the barbell held in front of the hips. Maintaining a braced core, they performed only the eccentric descent by slightly flexing the hip and knee, extending the non-dominant leg backward until the torso was parallel to the floor. The attempt was void if the non-dominant leg touched the ground. Participants had 3 attempts to record their best pure eccentric performance.

Bilateral Squat (BS)

Participants stood with feet shoulder-width apart, barbell on the rear trapezius. With a braced core, they performed a controlled eccentric squat until the thighs were parallel to the floor; upon reaching the lowest point, spotters took control of the barbell. Participants had 3 attempts to record their best pure eccentric performance.

Bilateral Deadlift (BD)

Participants stood with fixed feet, barbell in front of the hips. Keeping the bar close to the body, they slightly flexed the hips and knees to execute only the eccentric descent until the torso was parallel to the floor. Participants had 3 attempts to record their best pure eccentric performance.

Jump performance

Standing Triple Jump (STJ)

Participants stood with feet together behind the starting line. They initiated the first jump with both feet, split their legs in the air, landed on one leg, and immediately transitioned into the second jump. They then landed on the opposite leg, transitioning into the third jump, landing squarely on both feet and holding a stationary position. Participants had 3 attempts with 2 min of rest between attempts; the best score was recorded [29].

Standing Advantage Unilateral Reach Height (SAURH)

Using the identified dominant leg, participants stood on one leg with the other suspended. They performed a rapid downward countermovement and exploded vertically to touch the highest point possible on the testing vane. Participants had 3 attempts; the best score was recorded.

Standing Bilateral Reach Height (SBRH)

Participants stood with both feet flat. They performed a rapid countermovement via a squat and immediately executed a maximal vertical jump with an arm swing to reach the highest point. Participants had 3 attempts; the best score was recorded.

Sprint performance

30-meter sprint

The sprint time was measured using the SmartSpeed timing system [30]. Participants completed a 30-meter linear sprint; false starts or deviating from the lane resulted in a voided attempt and a retest. Participants had 3 attempts, and the fastest time was recorded.

Statistical analysis

Statistical analysis was performed by the author K.X., who was not involved in experimental procedures i.e., blinded to group allocation. Data are presented as mean ± standard deviation (SD). Within the week prior to the intervention, two baseline tests separated by 48 h were conducted on a subset of participants, utilizing the coefficient of variation (CV) and intraclass correlation coefficient (ICC) to evaluate the reliability and consistency of each metric [31], calculated as ICC (3,1) (two-way mixed-effects model, consistency type). To reduce reliance on the binary interpretation of P-values, Bayesian mixed-effects models were employed to estimate the probability of performance improvements from baseline within each group. Posterior distributions of performance changes, along with their 95% credible intervals (CrI), were derived using the brms [32] package, assuming Gaussian likelihoods and default weakly informative priors. This Bayesian framework offers advantages in small-sample settings by providing richer probabilistic inferences beyond traditional null hypothesis significance testing. Additionally, we calculated the posterior probability that performance improvements exceeded zero (> 0), offering a direct quantification of the likelihood of a positive effect.

Results

The results of the Bayesian mixed-effects model analysis indicated that after the 7-week intervention, neither the UET group nor the BET group demonstrated a statistically significant between-group advantage across all eight tested indicators, including lower limb maximal strength, jump, and sprint performance (the high-probability threshold was set at > 97.5%, and the posterior probabilities for all between-group comparisons were below this critical value). However, exploratory within-group analyses revealed potential trends in improving specific neuromuscular performances for both groups:

Lower limb maximal strength

Among the estimated indicators of lower limb maximal strength, only the UET group exhibited an extremely high probability of improvement in AUD (98.25%), with an average improvement effect of 12.01 kg (g = 12.01, 95% CrI [1.04–22.87]). The other maximal strength indicators (e.g., AUBS, BS, BD) did not show clear high-probability improvement trends in either group.

Jump performance

Regarding jump performance, the UET group showed the highest probability of improvement in SAURH, reaching 99.74%, with an average increase of 9.12 cm (g = 9.12, 95% CrI [3.41–14.81]). Concurrently, the UET group also had a high probability of improvement (98.43%) in SBRH, with an average increase of 7.60 cm (g = 7.60, 95% CrI [0.96–14.29]). Notably, the BET group also presented a high probability of within-group improvement in the SBRH indicator (97.56%), with an average increase of 6.76 cm (g = 6.76, 95% CrI [0.04–13.79]). Neither group showed a high probability of improvement in STJ.

Sprint performance

In terms of 30-meter linear sprint performance, only the UET group demonstrated an extremely high probability of improvement (99.03%), with the absolute sprint time shortened by an average of 0.27 s (g = 0.27, 95% CrI [0.05–0.48]). In contrast, the BET group did not present a high-probability improvement trend for this indicator. (Table 3, 4).

Table 3.

Outcomes measure’s reliability

Category	Variable	Intraclass correlation coefficient (95% CI)	Coefficient of variation (%)
Maximum strength	AUBS	0.99 [0.97–0.99]	1.60%
	AUD	0.99 [0.98–0.99]	1.64%
	BS	0.98 [0.95–0.99]	1.85%
	BD	0.97 [0.94–0.99]	1.50%
Jump	STJ	0.95 [0.90–0.97]	1.67%
	SAURH	0.82 [0.68–0.91]	1.37%
	SBRH	0.80 [0.65–0.90]	1.45%
Sprint	30 M-Sprint	0.90 [0.82–0.95]	1.51%

Abbreviations: CI confidence interval, STJ standing triple jump, SAURH standing advantage unilateral reach height, AUBS advantage unilateral bulgaria squat, AUD advantage unilateral deadlift, SBRH standing bilateral reach height, BS bilateral squat, BD bilateral deadlift

Table 4.

Probability of test indicators between groups and between groups under Bayesian mixed model

	Maximum strength				Jump			Sprint
	AUBS (kg)	AUD (kg)	BS (kg)	BD (kg)	STJ (cm)	SAURH (cm)	SBRH (cm)	30 M-Sprint (s)
BET	84.01%	80.25%	92.86%	96.86%	60.26%	96.33%	97.56%	89.26%
UET	95.73%	98.25%	93.09%	96.95%	69.88%	99.74%	98.43%	99.03%
UETvsBET	71.71%	84.43%	48.29%	52.24%	57.74%	85.78%	57.95%	82.01%

Abbreviations: UET Unilateral Eccentric Training group, BET Bilateral Eccentric Training group, STJ standing triple jump, SAURH standing advantage unilateral reach height, AUBS advantage unilateral bulgaria squat, AUD advantage unilateral deadlift, SBRH standing bilateral reach height, BS bilateral squat, BD bilateral deadlift

Discussion

The results of the Bayesian mixed-effects model indicated that after a 7-week intervention, there was no statistically significant between-group advantage (posterior probabilities < 97.5%) between the unilateral eccentric training (UET) and bilateral eccentric training (BET) groups in improving eight performance indices, including lower limb maximal strength, jumping, and sprinting performance, in trained adolescent basketball players. This finding suggests that UET cannot be asserted as superior to BET in terms of overall intervention efficacy. However, as a pilot exploratory study with a very limited sample size [33], within-group analyses revealed potential trends for both groups in improving specific performance metrics. Such exploratory trends based on posterior probabilities are not only limited by sample size but may also be profoundly influenced by specific variables in the experimental design, warranting cautious interpretation. The following discussion delves into these findings, categorized by different neuromuscular performance traits. (Table 5).

Table 5.

Means and 95% credible intervals calculated using the Bayesian mixture model

Category	Variable	BET	UET	UET vs. BET
Maximum strength	AUBS	7.36 [-8.10–22.33]	13.25 [-1.87–28.88]	5.90 [-16.14–27.55]
	AUD	4.47 [-6.66–15.47]	12.01 [1.04–22.87]	7.54 [-7.71–23.18]
	BS	8.56 [-3.21–20.16]	8.35 [-3.22–20.40]	-0.21 [-16.48–16.51]
	BD	8.79 [-0.48–18.33]	9.06 [-0.42–18.38]	0.27 [-13.24–13.44]
Jump	STJ	7.36 [-52.92–66.14]	15.08 [-44.50–72.59]	7.71 [-77.43–90.94]
	SAURH	4.96 [-0.48–10.58]	9.12 [3.41–14.81]	4.16 [-3.71–12.03]
	SBRH	6.76 [0.04–13.79]	7.60 [0.96–14.29]	0.84 [-9.17–10.17]
Sprint	30 M-Sprint	0.13 [-0.08–0.35]	0.27 [0.05–0.48]	0.13 [-0.17–0.43]

Abbreviations: UET Unilateral Eccentric Training group, BET Bilateral Eccentric Training group, STJ standing triple jump, SAURH standing advantage unilateral reach height, AUBS advantage unilateral bulgaria squat, AUD advantage unilateral deadlift, SBRH standing bilateral reach height, BS bilateral squat, BD bilateral deadlift

Lower limb maximal strength

This study defined squat and deadlift tests as maximal strength indicators because these movements occur at very low velocities and primarily reflect the muscle’s maximal force production capacity, which is fundamentally different from explosive power that emphasizes the rate of force development. The test results showed no significant differences between UET and BET in most between-group comparisons. Regarding within-group trends, only the UET group exhibited an extremely high probability of improvement in the unilateral deadlift (AUD) test (98.25%, g = 12.01, 95% CrI [1.04–22.87]), whereas bilateral squat (BS), bilateral deadlift (BD), and unilateral Bulgarian split squat (AUBS) did not show clear high-probability enhancements in either group.

Previous studies have shown that short-term unilateral and bilateral strength training produce similar effects on promoting muscle hypertrophy [34, 35], but unilateral multi-joint training often leads to preliminary adaptations in neuromuscular coordination due to the instability of the movement pattern [9, 35]. Some literature indicates that unilateral patterns can stimulate higher neural excitability and greater potential for fast-twitch muscle fiber recruitment [36]. For instance, Belegišanin et al. [37] found a larger effect size in the single-leg knee extension test following unilateral and bilateral eccentric flywheel training. However, the potential advantage of the UET group in AUD in this study must be examined in light of the limitation of volume mismatch. In the study, each leg in the UET group received 5 sets of high-intensity stimulus (10 sets of total load), whereas the BET group only performed 5 sets. Therefore, the higher probability of AUD performance improvement within the UET group is highly likely driven partially by the double limb-specific stimulus and longer time under tension, rather than an absolute advantage of the unilateral mode itself.

Furthermore, this study utilized the Mayhew formula to estimate unilateral and bilateral eccentric maximal strength (1RM). It must be acknowledged that such formulas were originally validated based on traditional bilateral, compound movements involving complete concentric and eccentric cycles. The force-velocity relationship of pure eccentric contractions and unilateral asymmetrical movements (e.g., Bulgarian split squat) differs fundamentally from bilateral movements. Therefore, the data in this study can only be considered as “predicted submaximal resistance load indices,” which further limits confirmatory inferences regarding absolute maximal strength values Fig. 1.

Fig. 1.

Effects of UET and BET on Maximum strength

Jump performance

Regarding jump performance, the results indicated that neither group demonstrated a high probability of improvement in the standing vertical jump (STJ). However, in the standing arm-up reach height (SAURH), the UET group showed an extremely high probability of improvement (99.74%, g = 9.12, 95% CrI [3.41–14.81]); in the standing bilateral reach height (SBRH), both the UET group (98.43%, g = 7.60, 95% CrI [0.96–14.29]) and the BET group (97.56%, g = 6.76, 95% CrI [0.04–13.79]) exhibited a high probability of within-group improvement potential.

The improvement in unilateral jumping for SAURH aligns with the principle of training specificity [38, 39]. That is, training methods that closely replicate the biomechanical and neuromuscular demands of a target movement may lead to more effective performance transfer [40]. The smaller base of support in the UET group may force stabilizing muscle groups into adaptive contractions, thereby providing better dynamic stability during the take-off and landing phases [13]. For the SBRH, a bilateral explosive power indicator, both groups showed improvement potential. Some scholars have proposed that unilateral training might alleviate bilateral strength deficits through the “cross-transfer” effect [41, 42], thereby enhancing bilateral jumping performance [43]. However, given that this study did not collect any surface electromyography (EMG) or muscle biopsy data, attributing the enhancement of jumping performance directly to deep neural mechanisms (such as fast-twitch fiber recruitment or cross-transfer) is an overextension. We can only speculate that this change might be related to preliminary improvements in neural coordination.

It is worth noting that the participants were adolescents with an average age of 15–16 years, a physiological window near peak height velocity (PHV). During this stage, endogenous hormone secretion leads to natural leaps in neural recruitment capacity and muscle strength. Because this study lacked a pure non-intervention control group and did not control for maturity offset, the observed improvements in jumping ability not only include the intervention effect but are inevitably confounded by the natural growth and development effects of adolescence Fig. 2.

Fig. 2.

Effects of UET and BET on jump and sprint performance

Sprint performance

In the 30-meter linear sprint test, only the UET group demonstrated an extremely high probability of improvement within the group (99.03%, g = 0.27, 95% CrI [0.05–0.48]), with a trend towards a shorter absolute time, indicating a potential enhancement in sprint capacity; the BET group did not show a similar high-probability trend.

This exploratory finding is consistent with the views of some previous reviews [44]. From a biomechanical perspective, the alternating single-leg force production pattern of horizontal sprinting has a higher mechanical correlation with unilateral training [45]. Scholars such as Zhang et al. pointed out that the positive effect of unilateral training on short sprints may primarily stem from the optimization of coordination between the nervous system and muscles [13]. Because unilateral training involves relatively lighter loads and more forgiving joint angles, it can safely reshape motor unit recruitment strategies during early training [13, 44, 46].

However, when interpreting this phenomenon, the “concurrent training interference effect” must be considered. During the 7-week intervention, participants also had to complete up to 6 h of high-intensity basketball-specific training per week. The extensive running, jumping, and aerobic expenditure might inhibit the neuromuscular adaptations that resistance training would otherwise elicit. This study lacked quantitative monitoring of external basketball loads through inertial measurement units (IMUs) or internal loads via session rating of perceived exertion (sRPE). Consequently, we cannot rule out the possibility that the UET and BET groups experienced heterogeneous loads during regular basketball training, introducing significant uncertainty to the fluctuation of sprint results. Future studies should further verify the true effect of unilateral eccentric training on sprint performance while controlling for sport-specific training loads.

Limitations

As an exploratory pilot experiment, this study has several limitations that must be emphasized, which lead to a high degree of uncertainty in the results.

Sample inference and participant classification limitations

First, the extremely small sample size (n = 12) and the lack of an a priori minimal detectable effect size (MDES) derivation severely limit the precision of statistical inferences and increase the risk of false positives or exaggerated effect sizes. Second, according to current international classification frameworks, the participants in this study should be strictly defined as “Tier 2: Trained/Developmental” rather than high-level athletes, which limits the extrapolation of the findings to populations with higher competitive levels [26].

Volume mismatch

This is a major confounding factor in the experimental design. The unilateral limbs of the UET group bore double the load volume of the BET group (10 sets per leg vs. 5 sets for both legs). Thus, the higher probability of improvement observed within the UET group is highly likely due, in part, to the greater absolute training volume imposed on the limbs, rather than an absolute advantage of the unilateral modality itself.

Confounding effects of maturity and concurrent training

The participants were near peak height velocity (PHV). Lacking a non-intervention control group and without controlling for biological maturity offset, this study cannot fully isolate the confounding impact of adolescent natural growth on performance enhancements. Additionally, the participants completed 6 h of high-intensity basketball training weekly. Without objective load monitoring (e.g., IMU or sRPE), the interference effect caused by this concurrent training may have had an uncontrollable and heterogeneous impact on performance adaptations.

Construct validity and measurement limitations

This study used theMayhew formula to estimate the maximal strength (1RM) of pure eccentric and unilateral asymmetrical movements. However, because the load-velocity curves of unilateral movements differ fundamentally from traditional bilateral compound movements, the validity of this formula is unverified, meaning the resulting data can only serve as submaximal resistance load indices [47]. Moreover, the lack of baseline lower limb symmetry assessments and the relatively short intervention period (7 weeks) might have been insufficient to induce deep, structural adaptations in tendons and muscles.

Future directions

Future studies should expand the sample size and extend the intervention period based on strictly matched training volumes, the inclusion of a non-intervention control group, and the use of objective load monitoring tools. Simultaneously, it is recommended to incorporate neurophysiological measurement tools, such as surface electromyography (EMG), to investigate the exact mechanisms of neuromuscular adaptations to different modes of eccentric training through a more rigorous design.

Practical application

Basketball is a team sport that heavily relies on complex unilateral and bilateral force production patterns. The preliminary exploration of this study suggests that both unilateral and bilateral eccentric training have the potential to improve specific performances in trained adolescent basketball players at the within-group level. Considering that the two training modalities did not exhibit significant between-group differences statistically, and each possesses unique biomechanical value, we do not recommend strength and conditioning coaches view them as mutually exclusive or opposing training methods. Instead, coaches should formulate mixed training strategies based on sport-specific demands: BET can be utilized to build foundational bilateral synergistic force capacity and overall absolute strength; concurrently, UET can be interspersed, leveraging its relatively lighter loads and flexible joint angles to provide specific stimuli to individual limbs or to correct bilateral strength asymmetries. This complementary training configuration will help safely and comprehensively diversify training variations and promote the long-term physical development of adolescent athletes.

Conclusions

Based on a 7-week pilot intervention study on trained adolescent basketball players, Bayesian mixed-effects model analysis indicated that unilateral eccentric training (UET) and bilateral eccentric training (BET) did not exhibit any statistically significant between-group advantages in improving lower limb maximal strength, jump, and sprint. Although exploratory within-group trend analyses showed potential improvement trends for UET and BET in certain specific indicators (e.g., UET for unilateral reach, both groups for bilateral reach), constrained by the extremely small sample size, volume mismatch, and concurrent training interference, this study cannot draw the confirmatory conclusion that UET is superior to BET in overall intervention efficacy. Future practical and academic inquiries should consider UET and BET as two parallel and equally valuable, effective complementary training modalities.

Authors' contributions

Bitai Wu conceived the study and led the writing and revision of the entire manuscript, including addressing all related issues. Jianzhong Hu conducted a comprehensive review of the entire text and provided accurate revision directions. Shaojuan Hu, Zhifeng Liu, and Anqi Chen restructured and revised the entire abstract, results, limitations, and discussion sections for the second revision. Mingyue Yin, Kai Xu, and Meiling Tao contributed to data analysis, image processing, and manuscript review. Leijiao Yue participated in the experimental operations and reviewing of the manuscript. Zhiquan Song and Hengxian Liu participated in reviewing the manuscript.

Funding

This study received no funding.

Data availability

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of [Beijing Sport University] (No. 2024433H). Written informed consent was obtained. The study complied with the Declaration of Helsinki. Our research involved minors, individuals under the age of 16. Before the experiment began, all the parents or guardians of the underage participants signed the informed consent form regarding the experiment.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.