Asymmetrical lower-limb injury risk during the volleyball spike jump of male elite volleyball players

2.1 Participants

A total of 11 elite male collegiate volleyball athletes were recruited between 10 January and 12 February 2025 for this study. Participants had a mean age of 20.4 ± 1.7 years, body mass 79.4 ± 7.1 kg, height 1.89 ± 0.10 m, and training experience 9.0 ± 1.5 years. Inclusion criteria were: (i) currently competing at the collegiate level, (ii) a minimum of 5 years of systematic volleyball training and participation in official competitions, and (iii) the ability to perform spike jumps with full approach and take-off technique. Exclusion criteria were: (i) any musculoskeletal injury or pain in the lower limbs or trunk within the previous 6 months, (ii) a history of lower-limb surgery, and (iii) inability to complete the jump tasks according to experimental requirements. All participants were right-side dominant and reported no musculoskeletal injuries in the 6 months preceding data collection. Prior to participation, all athletes provided written informed consent. The study protocol was approved by the Jeonbuk National University Institutional Review Board (JBNU2024-09-015-002). All participants competed at the elite collegiate level, which supports the ecological validity of the findings for high-performance volleyball settings.

2.2 Preparation for testing

Before the experiment, participants performed familiarization trials to become accustomed to the experimental procedures. An adjustable-height device [7, 10] was used to suspend a ball in front of the net, and the ball was released on contact to standardize the SPJ task and encourage maximal jump height. Each foot was positioned on a separate force platform to allow bilateral ground-reaction-force recording, and successful spikes were required to land in a designated target area (Fig. 1).

Fig. 1.Experimental site setting. Schematic setup of the SPJ test on a standard volleyball court. Athletes performed an approach and spike over the net, stepping on a force plate placed on the attack path close to the 3-m attack line to record the ground-reaction forces during the plant and take-off. Successful trials were required to land in the predefined target area.

After a standardized warm-up, athletes performed familiarization trials under investigator supervision until they were comfortable with the task. They were permitted to choose their preferred approach distance and angle and to adopt a self-selected starting position.

Each participant was required to perform five valid SPJ. Their goal was to maximize jump height and to strike the suspended ball into a designated target area on the opposite court. A trial was considered valid if both feet contacted the two separate force plates during take-off and the ball was hit into the target area. Trials were deemed invalid if the net was touched, as this could affect jump performance or was considered a rule violation.

A 1-min rest interval was provided between trials to minimize fatigue, and participants were instructed to perform approach jumps that replicated typical match conditions.

2.3 Data collection

A 13-camera infrared motion capture system (Prime17w, OptiTrack, NaturalPoint Inc., Corvallis, OR, USA) recorded the trajectories of 57 retroreflective markers (14 mm diameter) at 240 Hz. Marker placement followed previously described protocols [19]. Two AMTI force platforms (OR6-6-2000, AMTI Inc., Watertown, MA, USA) simultaneously recorded ground reaction forces at 1200 Hz. Kinematic and kinetic data were synchronized in Motive 2.2.0 (OptiTrack, NaturalPoint Inc., Corvallis, OR, USA).

2.4 Data processing

Kinematic and kinetic data were processed in Visual3D x64 v2024.10.4 (C-Motion, Inc., Rockville, MD, USA) using a 15-segment skeletal model (feet, shanks, thighs, pelvis, trunk, head, upper arms, forearms, hands). Joint centers were defined using bilateral reflective markers. Hip joint centers were estimated in Visual3D using its built-in Cartesian Optoelectronic Dynamic Anthropometer (CODA) pelvis model (Charnwood Dynamics Ltd., Leicestershire, UK), which uses regression-based equations to compute hip joint center locations, i.e., hip joint locations. The phases of the SPJs were defined using combined kinematic and kinetic criteria, consistent with previous work, and are illustrated in (Fig. 2, Ref. [9]) [10].

Fig. 2.Phases of the SPJs. Approach phase: from the start of the approach to the first ground contact of the dominant foot. Planting phase: from the end of the approach phase to the instant of the lowest CoM. Push-off phase: from the end of the planting phase to the last ground contact at take-off; this interval includes the ground contact of the non-dominant foot. Take-off: the final contact with either force plate (FP1 or FP2), identified using a 20 N [9] ground reaction force threshold. FP: force plate; CoM: center of mass.

Net joint torques and reaction forces for the lower extremities were computed via inverse dynamics. Marker trajectories and ground reaction forces were filtered using a fourth-order zero-lag Butterworth filter (6 Hz for kinematics [21], 20 Hz for kinetics [22]). Energy flow was analyzed following Robertson and Winter [23] using two variables: joint force power (JFP) and joint torque power (JTP). JFP represents energy transfer between adjacent segments, where a positive value indicates energy inflow and a negative value indicates energy outflow (Eqn. 1):

JTP represents energy generation, absorption, or transfer via joint torques: where .

When the proximal and distal STPs have the same sign (or one is zero), JTP is interpreted as net energy generation (positive) or absorption (negative) at that joint. In the present study, JTP was analyzed in this sense as joint-level energy generation rather than decomposed into a separate “torque-mediated transfer” term. The transfer component of JTP has been reported to contribute minimally during explosive jump tasks and was not examined independently here [18].

For each leg, JFP and JTP were integrated over the take-off phase, defined from the instant when whole-body CoM vertical velocity reached zero to the instant when ground reaction force (GRF) on the force plate dropped below 20 N [9]. The resulting integrals were interpreted as total transferred energy (from JFP) and total generated energy (from positive JTP) for that phase.

To compare inter-limb strategy, the percentage contribution of each joint (hip, knee, ankle) to total transferred energy and to total generated energy was calculated. All instantaneous power terms (JFP, JTP) and their time integrals were normalized to body mass (W/kg for power, J/kg for work-like integrals).

2.5 Statistical analysis

Descriptive statistics (mean ± standard deviation (SD)) were calculated for all variables. Normality was assessed with the Shapiro-Wilk test, and all variables met the normality assumption (p > 0.05). Accordingly, two-tailed paired-samples t-tests were used for all limb comparisons with α = 0.05.

Effect sizes for paired comparisons were reported as Cohen’s d = t/√n (based on the SD of the paired differences). The sign convention was DL-N-DL; therefore, negative Cohen’s d values indicate higher means in the non-dominant leg. Ninety-five percent confidence intervals (95% CI) for Cohen’s d were obtained from the noncentral t distribution.

A sensitivity power analysis for the paired-samples t-test (α = 0.05, two-tailed; power = 0.80; N = 11) was performed in G*Power (version 3.1.9.7; Universität Kiel, Kiel, SH, Germany), indicating that the design could detect paired differences of approximately d ≥ 0.94 [24]. Most between-limb effects met or exceeded this magnitude. The knee-joint work difference during energy generation showed a slightly smaller effect (d = 0.92) but remained statistically significant (p = 0.013) and large in practical terms, indicating that this asymmetry is likely meaningful despite marginally lower a priori power.

All statistical analyses were performed in IBM SPSS Statistics, version 27 (IBM Corp., Armonk, NY, USA). p-values are reported to three decimals (p < 0.001 where appropriate); units are given in the corresponding tables.

3.1 Approach velocity and jump height

The mean jump height based on CoM displacement was 0.65 ± 0.06 m, and the approach velocity was 3.66 ± 0.25 m/s, which is comparable to previously reported values in elite male volleyball players performing SPJs (0.67 ± 0.07 m; 3.71 ± 0.33 m/s) [7].

3.2 Energy generation differences between legs

The DL produced significantly greater total mechanical work than the N-DL (p = 0.002; d = 1.27, very large). Knee (p = 0.013; d = 0.92, large) and ankle joint work (p = 0.003; d = 1.18, large) were significantly higher in the DL, whereas hip work showed no significant difference (p = 0.335; d = 0.31, small).

Hip peak power was significantly higher in the N-DL (p < 0.001; d = −2.22, extremely large), while knee (p = 0.903; d = 0.04, trivial) and ankle (p = 0.250; d = 0.37, small) peak power showed no significant differences.

Joint contribution percentages did not differ significantly across the legs for hip (p = 0.568; d = −0.18, small), knee (p = 0.839; d = 0.06, trivial), or ankle (p = 0.556; d = 0.18, small) (Table 1).

Interpretation: The DL greater total mechanical work, concentrated at the knee and ankle, indicates higher cumulative extensor demand during take-off. Repeated exposure to such propulsive demands may increase mechanical loading associated with patellar tendon stress and Achilles tendon overload.

3.3 Energy transfer differences between legs

The N-DL transferred significantly more total energy than the DL (p = 0.005; d = −1.08, large). Joint-specific analysis showed greater energy transfer in the N-DL at the hip (p = 0.006; d = −1.04, large), knee (p = 0.009; d = −0.97, large), and ankle (p = 0.006; d = −1.04, large).

Peak transfer power was also significantly higher in the N-DL at the hip (p < 0.001; d = −1.71, very large), knee (p = 0.008; d = −1.00, large), and ankle (p = 0.008; d = −0.99, large).

Joint contribution percentages (%) did not differ significantly between legs for the hip (p = 0.423; d = −0.25, small), knee (p = 0.465; d = 0.23, small), or ankle (p = 0.900; d = 0.04, trivial) (Table 2).

Interpretation: The N-DL’s role as the primary conduit for horizontal-to-vertical energy transfer places it under greater deceleration and frontal-plane control demands. This loading pattern is consistent with mechanisms of lateral ankle sprain and ACL strain, particularly when rapid approach speed or unanticipated opponent contact challenges neuromuscular control. The high peak transfer power at the hip further highlights the need for strong gluteal and trunk stabilizer function to limit knee valgus and protect the ACL during explosive take-offs.

4.1 Energy generation

The DL produced significantly more total energy during take-off (p = 0.002, d = 1.27), driven by an earlier onset and longer duration of power generation (Fig. 3). Because the CoM is positioned closer to the DL at plant, this leg must initiate and sustain vertical acceleration for longer [7, 25]. This concentrates high-magnitude loading on the knee extensors and ankle plantarflexors of the DL, increasing mechanical demand on the quadriceps patellar tendon complex and the triceps surae Achilles tendon complex [9, 17]. Repetition of this high extensor output across many spikes may increase exposure to mechanical conditions associated with patellar tendon stress and Achilles overuse [26, 27].

Interestingly, although the DL hip began producing energy before the CoM reached its lowest point, its peak power was lower than that of the N-DL (p < 0.001, d = −2.22). This likely reflects coordination rather than weakness. First, arm swing and trunk inclination can prolong hip extension while simultaneously increasing anterior trunk momentum [28, 29]; if poorly controlled, this may elevate lumbar and hip joint stress even as it improves jump height [30]. Second, when the DL knee extensors are producing large concentric torque, the rectus femoris must generate high force. Because the rectus femoris spans both the hip and knee, its high activation contributes a flexor moment at the hip, which effectively opposes the net hip extensor moment generated by the gluteal and hamstring muscles [31]. At the same time, large quadriceps forces acting in a flexed-knee posture increase patellofemoral joint reaction force due to a high compressive load on the patellofemoral joint [32]. In addition, when the knee is flexed, the hamstrings are mechanically disadvantaged: they are shortened over the knee and thus less able to generate hip extension torque or to resist anterior tibial translation, which places greater anterior shear demand on the ACL [33, 34, 35]. Taken together, these interactions suggest that the DL’s role in propulsion is achieved primarily through knee- and ankle-dominant output, rather than through maximal instantaneous hip power.

4.2 Energy transfer

The N-DL exhibited significantly greater total energy transfer than the DL (p = 0.005; d = −1.08), indicating that it functions primarily as the braking and redirection limb during the initial take-off phase. The N-DL manages the conversion of approach-run horizontal momentum into vertical motion, generating a large horizontal impulse and decelerating the body, while controlling frontal-plane motion [10, 25]. This exposes the N-DL to high multiplanar loads, including valgus and rotational moments at the knee and inversion stresses at the ankle, all of which are consistent with established mechanisms for non-contact ACL injury, patellofemoral overload, and lateral ankle sprain—especially under fatigue or perturbation [36, 37, 38, 39].

In double-leg SPJs, this redistribution of braking demand to the N-DL appears to permit the DL to focus on vertical propulsion. In contrast, during single-leg SPJs, the plant leg must both brake and propel, which is associated with higher horizontal ground reaction forces and less efficient conversion of horizontal to vertical impulse [9, 40]. The current findings, therefore, suggest that in the bilateral SPJs, the N-DL serves as the primary braking limb that stabilizes and reorients the system, while the DL capitalizes on that stability to generate lift. Functionally, this means the N-DL accumulates repeated high deceleration loads similar to single-leg tasks, which may increase exposure to conditions linked to ankle instability and ACL strain over time.

4.3 Task-specific specialization

Although the DL and N-DL differed in the total energy generated and transferred, they showed almost identical hip, knee, ankle contribution ratios and followed the same proximal-to-distal coordination sequence [41, 42]. This indicates that the two limbs do not use fundamentally different motor patterns; instead, the same coordination template is expressed for different functional roles; the N-DL brakes and redirects horizontal momentum, and the DL drives vertical propulsion [7, 10, 25].

In this pattern, the DL is the limb that produces more total mechanical energy and sustains this output for longer, which is consistent with a predominantly propulsive task. By contrast, the N-DL shows greater inter-joint energy transfer and more evident multiplanar involvement, indicating that it is the limb that handles the redirection of the approach momentum and the control of body alignment during take-off, rather than simply “pushing” vertically.

Because these roles expose the two limbs to different dominant loading conditions, training should mirror the limb-specific demands rather than force artificial symmetry. For the DL, repeated high knee- and ankle-extensor output warrants (i) eccentric quadriceps and triceps-surae strengthening and (ii) exercises that maintain hip contribution to propulsion, to avoid a purely knee-dominant pattern that may increase patellofemoral or patellar tendon loading. For the N-DL, drills that emphasize single-leg balance in the frontal plane, hip-abductor/external-rotator strength, and controlled redirection of approach velocity are appropriate, because they directly target the multiplanar energy-transfer role identified in this limb.

These training directions should be interpreted as load-management suggestions derived from the observed energy-generation and energy-transfer patterns, not as injury prediction. Prospective and fatigue-modeling studies are still required to establish dose-response relationships between limb-specific loading and actual injury occurrence.

4.4 Limitations

This study inferred limb-specific mechanical loading exposure from biomechanical variables obtained during the SPJ, rather than from prospectively observed injury incidence. We did not collect longitudinal or prospective injury data, so we cannot claim that the identified loading profiles cause or predict injury; rather, they should be interpreted as indicating exposure to mechanical conditions that have been associated with patellar tendon stress, patellofemoral overload, ACL-related valgus and rotational loading, and lateral ankle sprain/instability in previous reports.

The sample consisted of 11 elite male collegiate volleyball athletes, which improves ecological validity for competitive play but limits generalizability and raises the possibility that some very large effect sizes were inflated by individual variability.

Future studies should incorporate subject-specific musculoskeletal modeling and data-driven ligament-fatigue analyses to evaluate how repeated exposure to these loading conditions may contribute to progressive tissue-level vulnerability over time [43].