Fourfold Faster Knee Valgus Development During Single-Leg Landing Compared with Countermovement Jump: A Novel Approach to ACL Injury Risk Assessment

1. Introduction

Rapid decelerations, changes of direction, and landings are associated with ~70% of non-contact anterior cruciate ligament (ACL) injuries [1,2,3,4,5,6,7,8,9,10,11]. Such movements are typically characterized by increased knee valgus [12]. A critical element of ACL injury prevention is risk assessment [13,14,15]. This process involves the administration of tests executed on one or two legs such as various forms of squats, countermovement jumps (CMJ), and landings in an effort to mimic sport-specific situations where many ACL injuries occur. ACL injury risk is then inferred from these tests based on the magnitude of dynamic knee valgus (DKV). Its simplicity has made the squat test popular [16,17,18,19,20,21], however numerous researchers have pointed out the limitation that the squat test does not resemble the velocity profile of real-life sport movements, potentially reducing its sensitivity to predict ACL injuries. Moreover, the bilateral execution of the squat test further limits its utility, as knee injury risk is known to increase predominantly during unilateral weight-bearing tasks, such as single-leg jumps, cutting maneuvers, and changes of direction[1,9].

The single-leg counter-movement jump (SL-CMJ) test is emerging as a simple and clinically perhaps more valid test than the squat test to screen for ACL risk by measuring DKV. Its potential advantage arises from the high loads on the knee and its resemblance to knee joint kinematics observed during on-field movements. DKV was also higher in tasks performed with one compared with two legs (landing, drop jump, squat) [22,23]. Still, both the SL-CMJ and the squat tests can be inaccurate and often fail to identify those with high risks for a future ACL injury [24,25]. DKV is a reasonably accurate predictor of future ACL injuries [13,26]. The biomechanical basis of this prediction is that ACL loading increases with DKV [27,28,29,30]. Biomechanical studies reported that DKV tends to be higher in women than men [23]. Because DKV is known to increase ACL strain, it may partly explain the 2–9-fold higher incidence of non-contact ACL injuries in females compared with males [1,31,32].

In addition to knee valgus, knee flexion magnitude also affects ACL strain, which is highest at low knee flexion angles and progressively decreases as knee flexion increases [30,33,34,35]. Recent in vivo imaging evidence indicates that ACL strain during single-leg jumping tasks is more strongly associated with sagittal-plane knee mechanics, with higher strain observed when the knee remains relatively extended during task execution [36]. Modeling studies suggest that sagittal-plane kinematics and lower-limb muscle actions jointly influence ACL strain during single-leg landing (SLL), consistent with a task-dependent mechanism of ACL loading [37] . Consistent with these findings, video analyses of confirmed non-contact ACL injuries also reveal a tendency toward smaller knee flexion angles during ground contact in injury cases, although not always reaching statistical significance[4].

Taken together, these findings suggest that the combined assessment of DKV magnitude and knee flexion magnitude is warranted, as both reduced knee flexion and increased DKV have been shown to cumulatively increase ACL strain. Specifically, the same peak valgus may be less concerning if it occurs at a greater knee flexion, potentially indicating a lower-risk for an ACL injury. Accordingly, incorporating knee flexion magnitude into the interpretation of valgus-based metrics — such as by normalizing peak knee valgus to peak knee flexion during functional tasks — may provide a more context-sensitive estimation of ACL injury risk. Therefore, the present study examined the dynamic interplay between knee valgus and knee flexion during SLL and SL-CMJ.

SLL and SL-CMJ are candidate tests for ACL risk prediction based on knee kinematics. In particular, SLL might be preferred not only to the squat but even to the SL-CMJ because knee kinematics more closely resembles those observed during on-field tasks [38]. Critically, peak knee flexion and peak knee valgus and the corresponding angular velocities may be closer to on-field conditions during SLL compared with SL-CMJ. This may be because, although peak knee valgus is comparable between SL-CMJ and SLL, peak knee flexion is smaller and knee valgus angular velocity could be greater during the SLL task.

The purpose of the present study was to compare knee kinematics and external loading characteristics between SLL and SL-CMJ tasks, with a specific focus on peak knee valgus, knee flexion, knee valgus angular velocity, and ground reaction forces (GRF). We hypothesized that: 1. peak knee valgus would not differ between SLL and SL-CMJ; 2. during SLL compared with SL-CMJ, knee flexion at peak valgus would be smaller, and 3. knee valgus angular velocity would be higher during SLL than during SL-CMJ. We also performed exploratory correlation analyses among anthropometric measures and knee flexion and knee valgus magnitude and velocity.

2. Materials and Methods

2.1. Participants

We recruited asymptomatic, physically active female college students who were free from injury but demonstrated high DKV (≥ 10°) during the screening procedures (see below). The inclusion criteria were female sex, age between 18 and 30 years and DKV ≥ 10°. Exclusion criteria were current injuries and past surgeries in the spine, hip, knee, and ankle joints, or pain of orthopedic origin. Based on these criteria, 30 participants were ultimately enrolled in the study (age: 21.5 ± 2.21 years; height: 168.8 ± 6.97 cm; mass: 62.6 ± 8.89 kg; exercise training history: 15.1 ± 3.36 years). Participants pursued ground-contact sports 6.2 ± 2.93 hours per week at the club level, but none of them competed at or above national levels. Participants received verbal and written explanations of the experimental procedures and were informed about the potential risks. The study was approved by the University Ethics Committee (approval number: 7961-PTE2019) and was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to participation. Written informed consent was obtained from the study participant depicted in the figures for publication of the images in an online open-access scientific journal.

2.2. Sample Size Calculation

We performed an a priori sample size calculation using G*Power 3.1.9.7, targeting a medium effect size (Cohen’s d ≈ 0.5), α = 0.05, and statistical power (1–β) = 0.80. This analysis indicated that a minimum of 27 participants would be required. To increase statistical power and account for potential dropouts, the final sample included 30 participants.

2.3. Experimental Procedure

One week prior to the experimental session, participants visited the laboratory for a familiarization session to ensure that they were comfortable with and able to correctly execute the experimental tasks, including SLL and SL-CMJ. During this session, they also performed a single-leg squat for screening purposes to identify the leg exhibiting DKV ≥ 10°.

The main experimental session lasted approximately 30 minutes. Participants completed a 5-minute self-paced warm-up on a cycle ergometer, followed by general and lower-extremity specific stretching. Prior to the test trials, participants stood quietly on the force plate while body weight was measured and recorded for normalization of GRF data. They then executed the experimental tests in a fixed order: SL-CMJ followed by SLL, with 3 minutes of inter-trial rest to minimize fatigue and maintain consistent jumping and landing performance.

2.4. Single-Leg CMJ

Participants executed the SL-CMJs on a force platform (Tenzi Ltd., Pilisvörösvár, Hungary) using the leg previously identified as exhibiting greater DKV (Figure 1). Participants stood on one leg with arms akimbo and the non-supporting knee slightly flexed. They were instructed to jump as high as possible while keeping their hands on the hips; no additional instructions regarding jump strategy were provided. Each participant performed three trials, with a 3-minute inter-trial rest period to minimize fatigue and ensure consistent performance. Vertical GRF was recorded at a sampling rate of 420 Hz and normalized to the participant’s body weight measured on the force plate prior to testing, resulting in dimensionless values expressed as multiples of body weight (×BW).

2.5. Single-Leg Landing

Participants executed the SLLs from a 0.46-m-high plyometric box (Capital Sports Plyo Crew Box) using the lower limb previously identified as exhibiting greater DKV (Figure 2). They were instructed to step off the box without jumping, land on the designated leg, and maintain the landing position for 5 s. Each participant completed three trials, with a 3-minute inter-trial rest period to minimize fatigue and ensure consistent performance. Vertical GRF was recorded, and data were processed similarly to that described during SL-SMJ).

2.6. Knee JOINT KINEMATICS

A 3D motion tracking system (Noraxon, Scottsdale, USA) with a sampling frequency of 100 Hz was used to analyze the kinematics of the SL-CMJ and SLL. Inertial sensors comprising accelerometer, gyroscope, and magnetometer units were affixed with Velcro straps on the shank and thigh according to the manufacturer’s recommendations for sensor placement and calibration. Using the sensor’s segmental orientation angle-time data, peak knee valgus angle was considered as the frontal-plane thigh orientation angle. Knee flexion angle at the time point of peak knee valgus was determined during both SL-CMJ and SLL tasks, using the sagittal-plane shank and thigh orientation angle data. To account for the combined contribution of knee valgus and knee flexion to ACL loading, peak knee valgus was normalized to the corresponding knee flexion angle, representing a novel normalization approach in ACL risk assessment. During this procedure, peak knee valgus was divided by the knee flexion angle measured at the time of peak knee valgus, yielding a knee valgus value normalized to knee flexion.

2.7. Data Processing and Statistical Analyses

Data were exported from the measurement systems (force plate and the Noraxon MyoMotion system) and analyzed in MATLAB [39], where the datasets were screened and corrected for technical or transfer errors prior to statistical analysis. The cleaned numerical data were then transferred to Origin for statistical analyses. We present the data as mean ± standard deviation (SD) for normally distributed variables and as median [interquartile range, IQR] for variables deviating from normality. Normality was assessed using the Shapiro–Wilk test. For normally distributed outcomes, differences between the two test conditions (SL-CMJ vs SLL) were evaluated using paired-samples t-tests. When the assumption of normality was violated, paired comparisons were performed using the Wilcoxon signed-rank test. Effect sizes were calculated as Cohen’s d for paired samples (for t-tests) and as r = |Z|/√N for Wilcoxon tests. Statistical significance was set at p < 0.05. All analyses were performed in Origin (OriginLab, Northampton, MA, USA).

3. Results

All variables were normally distributed except knee valgus angular velocity during the single-leg landing task and vertical GRF during the SL-CMJ.

3.1. Comparisons Between SLL and SL-CMJ

Figure 3 shows a typical example of knee joint kinematics in the two tasks, including knee flexion and knee valgus angles and their corresponding angular velocities.

Compared with SL-CMJ, SLL was characterized by lower knee flexion at peak DKV but higher vertical GRF and faster movement dynamics (Table 1). Knee flexion angle at peak DKV was 14.8° (-24%) lower during SLL than during SL-CMJ (p < 0.001, d = 1.69). Peak DKV angle did not differ statistically between tasks (Δ = 1.2°, 6.6%; p > 0.05, d = 0.21). Despite similar peak DKV, SLL showed higher DKV normalized to knee flexion (+0.045; +17%) (p < 0.05, d = 0.42), higher knee flexion angular velocity (300.5°/s; +153%) (p < 0.001, d = 1.95), and higher DKV angular velocity (137.3 °/s; +254%) (p < 0.001, r = 0.87) than SL-CMJ. Vertical GRF was 2.05xBW higher during SLL (+124%) (p < 0.001, r = 0.87).

3.2. Correlation Analyses

Figure 4 shows the Pearson correlation between body weight and the magnitude of DKV during SLL (r = 0.37, p = 0.047); however, no significant correlation was observed between body weight and DKV SL-CMJ (p > 0.05).

Of the four knee flexion and valgus outcomes, the correlation (Figure 5) between body height and knee valgus during SLL approached significance (r = 0.33, p = 0.075). No other correlations between height and knee kinematics were significant (all p >0.05).

4. Discussion

The primary findings of the present study were fourfold. First, DKV magnitude did not differ between the SLL and SL-CMJ tasks. Second, knee flexion at peak knee valgus was significantly smaller during SLL compared with SL-CMJ. Third, when DKV was interpreted in relation to knee flexion—by normalizing DKV magnitude to knee flexion—significantly greater values were observed during SLL. Fourth, the temporal characteristics of DKV development differed markedly between tasks, with DKV evolving at a substantially higher angular velocity during SLL vs. SL-CMJ. Collectively, these findings suggest that although peak DKV magnitude alone appears similar between tasks, SLL is characterized by a biomechanical context involving reduced knee flexion and more rapid DKV development, which may influence how valgus-based screening metrics are interpreted [24,25].

In line with our first hypothesis, DKV magnitude did not differ significantly between SLL and SL-CMJ (Figure 3). This finding is consistent with previous reports showing that peak DKV alone is often insufficient to discriminate between tasks or individuals with differing ACL injury risk [24,25]. Although DKV has been identified as a reasonably accurate predictor of future ACL injury [13,26], several studies have highlighted that reliance on peak valgus magnitude alone may be insufficient for assessment of risk for an ACL injury and mask important biomechanical differences in task execution [25].

Consistent with our second hypothesis, knee flexion angle at peak valgus was significantly smaller during SLL compared with SL-CMJ (Figure 3). This observation is relevant given the well-established relationship between knee flexion magnitude and ACL loading, as ACL strain is highest at relatively small knee flexion angles and progressively decreases as knee flexion increases [30,33,34,35,36]. Accordingly, the smaller knee flexion angles observed during SLL suggest a biomechanically less favorable joint configuration with respect to ACL loading compared with SL-CMJ and enhance the sensitivity of SLL to identify those with an increased risk for an ACL injury.

A key novel finding of the present study was that knee valgus normalized to knee flexion was significantly greater during SLL compared with SL-CMJ. This result directly supports the conceptual framework outlined in the Introduction, which proposed that the ACL strain associated with a given valgus magnitude may be attenuated when valgus occurs at greater knee flexion angles [30,33,34,36]. By integrating valgus and flexion into a single metric, the present findings provide a more context-sensitive interpretation of frontal-plane knee mechanics with respect to screening for ACL injury risks. This approach aligns with emerging evidence suggesting that isolated peak metrics may underestimate ACL injury risk if not interpreted within the broader kinematic context of task execution [36,37].

In agreement with our third hypothesis, knee valgus angular velocity was markedly higher during SLL than during SL-CMJ (Figure 3). This finding highlights the importance of considering not only the magnitude but also the rate at which knee valgus develops, consistent with recent evidence emphasizing the role of neuromuscular timing in female soccer players [40]. Rapid development of valgus has been proposed as a critical factor in ACL injury mechanisms, as high loading rates may exceed the capacity of neuromuscular control to stabilize the knee joint [5].

The substantially higher valgus angular velocity observed during SLL supports previous assertions that landing tasks more closely resemble the high-speed, high-load conditions encountered during on-field injury scenarios than jump-based screening tasks [9,38]. This finding is particularly relevant given prior criticisms that commonly used screening tasks, such as squats and CMJ-based tests, do not adequately reflect the velocity profiles of real-world sporting movements [15,16,17]. Previous evidence suggests that eccentric-only training elicits significantly greater increases in eccentric maximal strength than concentric-only training (≈27% vs. ≈10%) [41]. Consistent with this observation, our previous work demonstrated that high-intensity eccentric strengthening of the hip abductor musculature, and the resulting increases in eccentric hip abduction torque, led to greater reductions in knee valgus during functional tasks (jumping and landing) following a four-week intervention compared with concentric training [42]. The literature consistently indicates that training involving high-velocity contractions is more effective at enhancing rapid force production than low-velocity or isometric strength training [43,44,45,46]. In the present study, the high knee valgus angular velocities observed during the SLL task (~200°/s) suggest that, during landing, the hip abductors must control valgus development through rapid eccentric contractions, underscoring the importance of velocity-specific training considerations in ACL injury prevention.

This finding is biomechanically intuitive, as the landing phase of the SLL occurs at a greater angular velocity than the propulsion phase of the SL-CMJ. Consequently, ground reaction forces were more than twofold greater during SLL than during SL-CMJ propulsion. These results suggest that, during landing, the hip abductor muscles must generate faster eccentric contractions while simultaneously resisting greater external forces to control excessive valgus positioning and potentially reduce ACL injury risk.

Taken together, the present findings highlight two key implications for ACL injury risk assessment and prevention. First, peak knee valgus magnitude should not be interpreted in isolation when screening for ACL injury risk. Rather, valgus magnitude should be considered in conjunction with the knee flexion angle at peak valgus. Normalizing knee valgus to knee flexion may provide a more biomechanically meaningful representation of ACL loading than valgus position alone, as valgus and knee flexion jointly determine the strain experienced by the ligament. Second, the substantially higher knee valgus angular velocity observed during the more mechanically demanding SLL task suggests that rapid frontal-plane femoral motion is an important characteristic of potentially injurious movement patterns. Consequently, conditioning of the hip abductor musculature—which plays a key role in resisting medial femoral displacement—may benefit from incorporating higher-velocity contractions. In accordance with the principle of mode specificity, training at contraction velocities that more closely resemble those observed during high-risk landing tasks may be more effective in improving rapid movement control, although this requires confirmation by randomized trials.

5. Conclusions

ACL injury risk assessment should extend beyond the measurement of peak valgus alone and incorporate valgus normalization to knee flexion. The high valgus angular velocities (~200°/s) observed during SLL call for velocity-specific hip abductor training to reduce risks for an ACL injury. Consequently, conditioning of the hip abductor musculature—which plays a key role in resisting medial femoral displacement—may benefit from incorporating higher-velocity contractions. In accordance with the principle of mode specificity, training at contraction velocities that more closely resemble those observed during high-risk landing tasks may be more effective in improving rapid movement control, although this requires confirmation by randomized trials. These findings may complement existing neuromuscular warm-up and injury-prevention programmes, such as FIFA 11+, PEP, and HarmoKnee, which have been shown to influence phase-specific CMJ performance and neuromuscular activation patterns in youth soccer players [40,47,48,49].