Introduction

Anterior cruciate ligament (ACL) injury is one of the common knee injuries in clinical practice, which seriously affects the athletic ability and knee stability of patients. According to statistics, ACL rupture occurs in about 250,000 people per year in the United States1, the incidence rate in Sweden is about 78–81 cases/100,000 people/year2, and the rate of ACL injury in Chinese national-level athletes is about 0.47%3. The incidence of ACL injuries and reconstruction is on the rise as the percentage of youth participating in competitive sports increases, and as middle-aged and older adults remain physically active for longer periods of time. ACL reconstruction has become the primary treatment for this injury. Annual expenditures in the United States for ACL reconstruction and rehabilitation reach $7.6 billion and $17.7 billion, respectively, highlighting the significant burden of disease associated with this injury4. Furthermore, the risk of re-injury following ACL reconstruction remains significant. Studies indicate that after the initial reconstruction, the risk of rupture in the unaffected knee is even higher than the risk of re-rupture in the injured knee5,6. Surgical treatment can restore the physiological structure and function of the patient’s knee joint, but systematic rehabilitation training is still required to regain daily living and athletic abilities. In terms of injury mechanisms, non-contact ACL ruptures predominantly occur during high-dynamic movement scenarios such as landing from a jump or sudden stops and changes of direction7, accounting for approximately 72%–95% of all ACL ruptures8.

Current surface electromyography studies on neuromuscular activity following ACL injury predominantly focus on major motor muscle groups such as the quadriceps femoris, hamstrings, and triceps surae9,10,11. Existing research indicates that insufficient gluteal muscle control can lead to dynamic knee varus, thereby increasing the risk of non-contact ACL injuries12. Therefore, investigating the role of the Gmax in dynamic knee control is crucial. Furthermore, most studies in this field employ relatively low-impact functional tasks such as walking13, running14, and squatting15 for EMG analysis. Given that the most common ACL injury movements are jumping and landing, emergency deceleration, and lateral changes of direction, in order to reproduce muscle EMG data closer to the actual moment of injury, the test movement needs to be as consistent as possible with the injury movement. In addition, most of the research subjects in this field are individuals who have undergone ACL reconstruction surgery, and there is a lack of research on the muscle activation mechanism of patients without reconstruction during high-dynamic movements.

Since ACL injuries are typically not caused by direct external force but are closely related to an individual’s inherent movement patterns, they are considered a type of injury that can be prevented through intervention. This underscores the critical importance of clearly identifying the differences in electromyographic timing and intensity characteristics of lower limb muscle groups between ACL rupture patients and healthy individuals. Simultaneously, to prevent recurrent injuries to the unaffected and affected side in ACL-injured patients, targeted rehabilitation interventions based on the characteristics of electromyographic timing and intensity changes in the lower limb muscle groups of both the affected and unaffected sides following ACL rupture are crucial. Based on the aforementioned research gaps, this study targets patients with chronic non-contact ACL complete tears who have not undergone reconstruction. It utilizes a single-leg jump landing task to analyze the activation timing, activation duration, and activation intensity characteristics of key muscle groups, including the vastus lateralis (VL), vastus medialis (VM), biceps femoris (BF), semitendinosus (ST), and gluteus maximus (Gmax), in both the injured and uninjured limbs, and compares them with corresponding limbs in the control group. Aims to provide a theoretical foundation for guiding targeted preoperative rehabilitation interventions for ACL rupture patients and early ACL injury prevention for sports enthusiasts.

Materials and methods

Participant recruitment

The study used G*Power (version 3.1; https://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-und-arbeitspsychologie/gpower) software for sample size estimation. The level of significance was set at α = 0.05, and the target statistical power was 80%. Based on Cohen’s recommendation16, the effect size (ES) was set at 0.46, and the minimum required sample size was calculated to be 40 individuals (20 in each group) using a two-way ANOVA model. Considering a dropout rate of approximately 10%, the final plan was to include 22 subjects in each group, for a total of 44, to ensure adequate statistical efficacy. All patients with ACL injuries completed testing before undergoing ACL reconstruction. Healthy individuals of the same sex, dominant leg side, and similar level of exercise (Tegner score not differing by more than ± 1 point) were screened by questionnaire to serve as the control group. All enrolled patients received no systematic physical therapy or rehabilitation intervention post-injury. They underwent standardized rehabilitation training at the Knee Sports Injury Department of Sichuan Orthopedic Hospital for three days following admission and prior to surgery. Detailed exercise prescriptions are provided in Supplementary Table 1. The control group received the same volume of exercise training. Considering that adaptive changes occur in the neuromuscular system of patients more than six months after ACL injury, indicating that the neuromuscular reorganization process following ACL injury has reached a certain stage, this allows us to observe the significant impact of ACL injury on electromyographic activity in lower limb muscle groups17. Therefore, we selected a disease duration criterion of more than six months for complete ACL injury. All participants signed informed consent forms prior to the experiment. All methods were carried out in accordance with relevant guidelines and regulations, including the Good Clinical Practice (GCP) guidelines and the Declaration of Helsinki. The study protocol was approved by the Ethics Committee of Sichuan Orthopedic Hospital (Ethics No. KY2020-017-01). The inclusion criteria were: injury group: (1) age 18–35 years old; (2) mode of injury was non-contact injury; (3) the patient’s injured leg was his/her dominant leg. We have determined that the patient’s dominant leg did not change before or after the injury (the dominant leg is defined as the leg preferred for kicking, standing, jumping, or stepping on a box); (4) MRI showed a complete rupture of the ACL alone; (5) the duration of the disease was more than six months; (6) there was no injury to the adjacent joints; and (7) there was no underlying disease such as gout or rheumatoid arthritis. Control group: (1) Age: 18–35 years old, regardless of gender; (2) No motor system diseases in the last six months; (3) No history of lower limb surgery.

Data collection

In this study, a QUALISYS 3D motion capture system (Model Oqus 300, Sweden) was synchronized with a Noraxon wireless surface EMG system (Model Ultium EMG, sampling frequency 2000 Hz, USA) and a Bertec force table (Model FP4060-08, sampling frequency 200 Hz, USA) for motion data acquisition. The force table was used to detect the initial contact (IC) time, defined as the instant when the vertical ground reaction force first exceeded 10N18. EMG acquisition sites were referenced to the SENIAM standardized localization guidelines19 as detailed in Supplementary Table 2. Subjects underwent a 10-minute warm-up prior to testing, consisting of 5 min of jogging with 5 min of dynamic stretching. After warming up, the area to which the electrodes were attached was prepared by cleaning with 75% alcohol, shaving, exfoliating with fine sandpaper, and again cleaning with an alcohol cotton ball and air-drying to minimize skin surface impedance. Subsequently, silver/silver chloride electrodes (model CH3236, CATHAY, China) were affixed with an electrode spacing of 20 mm20, the reference electrode was kept equidistant from the acquisition electrode and secured against dislodgement by means of a muscle patch with medical adhesive tape (see Fig. 1)21. In my test procedure, the one-legged jump landing task was given first priority, followed by the maximal voluntary contraction (MVC) test, with a 5-minute rest period between the two. Subjects were asked to stand with their feet shoulder-width apart on a 30 cm raised platform, with the distance between the platform and the center of the force plate set at 80% of the distance from the subject’s anterior superior iliac spine to the medial ankle22. The EMG signals were recorded before the start of the jump, and the subjects maintained a one-legged standing position on command, stabilized and jumped forward without initial velocity, and remained stationary for 3 s after landing on one foot. Each subject completed 5 tests at 1-minute intervals. Visual Analogue Scale (VAS) was used during testing to assess subjects’ pain levels while performing the single-leg landing task, with no significant discomfort observed throughout the process (see Fig. 2). Maximum voluntary contraction (MVC) testing was performed using the David isometric device, which measures the maximum flexion and extension forces of the anterior and posterior thigh muscle groups, respectively. To minimize interference with EMG results caused by variations in the distance between the instrument and the muscles, we adjusted the instrument’s rotational axis according to the subject’s seated position during measurement, aligning it with the greater trochanter of the femur on the outside of the knee joint being tested. The upper fixture held the thigh as close to the chair surface as possible. Furthermore, the slider was fixed 3 cm above the ankle joint, depending on the patient’s leg length, to ensure maximum strength for each subject. For the Gmax MVC test, the subject was lying prone on a yoga mat with the legs abducted about 30°, the knees flexed 90°, and the hip joint in 0° extension. The experimenter applied resistance to the lower and middle thighs with his bare hands, and the subject tried his best to complete the hip extension and abduction movements under the oral command, with the knee joint slightly off the ground (about 2–3 cm) and holding it for 5 seconds23. All MVC tests were repeated 3 times at 30-second intervals with strong verbal encouragement to elicit maximal force output24. The MVC tests for the BF and ST muscles are shown in Fig. 3.

Fig. 1
figure 1

VL, VM surface EMG acquisition site.

Full size image
Fig. 2
figure 2

One-legged jump landing test.

Full size image
Fig. 3
figure 3

MVC test diagram for BF and ST.

Full size image

Data processing

VL, VM, BF, ST, Gmax muscle activation timing metrics

The raw surface electromyography (SEMG) signals were first visually inspected to ensure that the acquired data were complete and valid. Subsequently, the DC offset was removed using Noraxon EMG (Ultium version; https://www.noraxon.com) software and a fourth-order zero-delay Butterworth high-pass filter (cutoff frequency fc = 15 Hz) was applied to remove motion artifacts. To extract temporal features of muscle activation, the filtered signal was full-wave rectified and a low-pass filter (fc = 20 Hz) was applied to generate a linear envelope signal. By experimenting with cutoff frequencies in the range of 10–25 Hz (in steps of 1 Hz), 20 Hz was finally selected as the low-pass filtering threshold, a parameter that maximizes the retention of key features consistent with changes in muscle tone25. The identification of muscle activation timing was based on a threshold of 15% of the maximum amplitude of the linear envelope signal: when 28 consecutive sampling points (sampling frequency 2000 Hz) exceeded or fell below this threshold, respectively, they were considered as the beginning and the end of muscle activation26. To determine this threshold, we conducted multiple rounds of comparisons and manual calibration within the range of 3%–25%. Specifically, we manually labeled the start and end times of each activation event while testing different threshold settings, comparing the automatically extracted activation sequences with the manually calibrated ones. Through repeated threshold adjustments, we evaluated the accuracy of muscle activation timing at each threshold. Ultimately, 15% was selected as the optimal threshold, as it accurately distinguished genuine muscle activation from background noise and aligned with prior research findings27. During manual calibration, we paid particular attention to how threshold selection affected signal clarity. Too low a threshold caused excessive background noise to be mistaken for muscle activation, while too high a threshold risked missing weaker muscle activation signals. The final determination of 15% as the optimal threshold accurately captured the onset and offset of muscle activation while minimizing noise interference. The final extracted temporal metrics included: the duration of muscle activation (ms); the time difference of activation onset time relative to initial contact (IC) (onset-IC, ms); and the time difference of peak EMG appearance time relative to IC (peak-IC, ms).

VL, VM, BF, ST, Gmax muscle activation strength indicators

Raw SEMG signals were full-wave rectified, smoothed using a moving-window (50 ms) root mean square (RMS) algorithm, and normalized by the peak value recorded during MVC. Muscle activation intensity indices were divided into the following two categories: preparatory muscle activity: peak and mean RMS values within a 100 ms window before IC; reactive muscle activity: peak and mean RMS values within a 100 ms window after IC.

Statistical analysis

Raw data were entered and organized using Microsoft Excel, and statistical analysis was done using GraphPad Prism (version 9.5; https://www.graphpad.com) software. All continuous variables were first assessed for normality by the Shapiro-Wilk test, and Levene’s test for chi-square was used. For variables that satisfied both normal distribution and chi-square, a two-way ANOVA was used, with “group” (ACL injury group vs. control group) as a between-group factor and “limb side” (dominant vs. nondominant leg) as a within-group factor. If main effects or interactions were significant, post hoc two-by-two comparisons were further performed using Fisher’s LSD method. Given the small sample size of this study, employing Fisher’s LSD method as the post-hoc comparison technique is a reasonable choice. This method is well-suited for multiple comparisons in small samples and effectively identifies intergroup differences. Although Fisher’s LSD does not perform p-value correction, we minimized the risk of Type I errors by rigorously verifying all assumptions, including ensuring data normality. We did not employ Bonferroni correction because this method significantly reduces statistical power, particularly in multiple comparisons. It may lead to overly stringent significance levels, thereby increasing the risk of Type II errors (false negatives). Therefore, given the exploratory nature of this study and the limitations of its sample size, we consider a certain degree of Type I error acceptable. This approach helps maintain the sensitivity and statistical power of the analysis. All data are presented as mean ± standard deviation and the significance level was set at p < 0.05.

Results

Baseline characteristics of participants

During the testing process, one subject in the injury group withdrew from the trial because she gave up ACL reconstruction surgery and chose conservative treatment; one subject in the control group was excluded because she was unable to complete all the tests due to scheduling conflicts. In the end, a total of 42 subjects completed all the tests, including 21 in the injury group and 21 in the control group. Among them, 22 were male and 20 were female. The basic information of the subjects is shown in Table 1.

Table 1 Basic information of subjects.
Full size table

Muscle activation duration

Since no significant interaction effect was presented between subject group and test limb, the main effects of the group factor and limb factor were analyzed separately, and the results are shown in Table 2. In terms of muscle activation duration, subject group showed a significant main effect on BF (F(1,40) = 4.732, p = 0.0326), with the duration of activation of the dominant leg in the injury group being significantly longer than that of the dominant leg in the control group (p = 0.0409). A significant main effect was also shown for Gmax (F(1,40) = 5.961, p = 0.0168), which likewise showed a significantly longer duration of sustained activation of the dominant leg in the injury group (p = 0.0469).

Table 2 Duration of sustained activation of each muscle comparison (ms).
Full size table

Characteristics of EMG activation timing by muscle

Since there was no significant interaction effect between subject group and test limb, the main effects of the two factors were analyzed separately in this section, and the results are shown in Table 3; Fig. 4. In the comparison of onset-IC, the subject group showed a significant main effect on BF (F(1,40) = 5.127, p = 0.0351), with onset-IC for the dominant leg in the injury group significantly earlier than that for the control group’s dominant leg (p = 0.0457).ST also showed a significant main effect (F(1,40) = 6.314, p = 0.0239), with the onset-IC of the dominant leg in the injury group being significantly earlier than that of the dominant leg in the control group (p = 0.0277) and that of its own non-dominant leg (p = 0.0405). In addition, Gmax also showed a significant main effect (F(1,40) = 4.053, p = 0.0474), as evidenced by a significantly earlier onset-IC in the dominant leg of the injury group than in the dominant leg of the control group (p = 0.0192). In the comparison of peak-IC, BF showed a significant main effect (F(1,40) = 4.040, p = 0.0415), in which the peak-IC of the dominant leg in the injury group was significantly later than that of the dominant leg in the control group (p = 0.0457).ST also showed a significant main effect (F(1,40) = 7.271, p = 0.0190), with the dominant leg in the injury group having the peak-IC significantly later than the control dominant leg (p = 0.0280).

Table 3 Comparison of onset-IC and peak-IC by muscle (ms)
Full size table
Fig. 4
figure 4

Pre-activation sequence of each muscle before touchdown.

Full size image

Characterization of EMG activation intensity by muscle

For the RMS peak versus mean within 100 ms before and after IC, there was no significant interaction effect between subject group and test limb, so the main effect of subject group versus test limb was analyzed, as shown in Figs. 5 and 6. Specific p values for the multiple comparisons results are detailed in Supplementary Table 3. There was a significant main effect of subject group on the peak RMS of the VL within 100 ms after IC (F(1,40) = 4.742, p = 0.0324), and the dominant leg in the injury group was significantly lower than the dominant leg in the control group (p = 0.0171). There was a significant main effect of subject group on the peak RMS of VM within 100 ms after IC (F(1,40) = 6.007, p = 0.0164), with the dominant leg in the injury group being significantly lower than the dominant leg in the control group (p = 0.0054) and the non-dominant leg in the control group (p = 0.0253). There was a significant main effect of subject group on the mean RMS of VM within 100 ms after IC (F(1,40) = 6.925, p = 0.0102), which was significantly lower for the dominant leg in the injury group (p = 0.0178) and the nondominant leg in the injury group (p = 0.0357) than for the dominant leg in the control group. There was a significant main effect of subject group on the peak RMS of ST within the first 100 ms of IC (F(1,40) = 6.429, p = 0.0132), which was significantly higher in the dominant leg of the injury group than in the dominant leg of the control group (p = 0.0439) and the non-dominant leg of the control group (p = 0.0229). There was a significant main effect of subject group on the mean RMS of ST within the first 100 ms of IC (F(1,40) = 5.563, p = 0.0208), and the dominant leg was significantly higher in the injury group than the dominant leg in the control group (p = 0.0483). There was a significant main effect of subject group on the peak RMS of Gmax within 100 ms after IC (F(1,40) = 13.94, p = 0.0004), with the dominant leg in the injury group being significantly lower than the dominant leg in the control group (p = 0.0003) and the nondominant leg in the control group (p = 0.0122), and the nondominant leg in the injury group being significantly lower than the dominant leg in the control group (p = 0.0081).

Fig. 5
figure 5

Comparison of RMS in each muscle during the first 100 ms of IC. note: DL indicates dominant leg; NDL, nondominant leg; *, indicates p < 0.05 for comparison between the dominant leg in the injury group and the other groups; RMS, root-mean-square amplitude.

Full size image

Discussion

ACL disruption is thought to disrupt pre-existing neuromuscular control strategies in the lower extremity. It has been shown that ACL injury can cause significant alterations in the kinematic, kinetic, and electromyographic characteristics of the lower extremity during high-impact tasks such as jump landings28,29,30, and that these changes reflect a phenomenon of neuromuscular adaptation or reprogramming in response to joint instability aimed at enhancing the dynamic stability of the knee and reducing the risk of re-injury by modulating the order of activation, duration, and peak response. In the present study, we found that muscle activation durations were generally prolonged in the affected leg of the ACL-injured group during a single-leg jump landing task, with BF and Gmax being the most significant. This result is consistent with previous findings regarding prolonged EMG durations in walking, such that the VM, VL, ST, and lateral popliteus showed similar trends31,32. The prolonged activation duration may represent a compensatory strategy for knee instability, increasing joint compressive forces by maintaining longer muscle contractions, thereby enhancing mechanical stability and compensating for limitations caused by ACL deficits, resulting in a kinematic profile similar to that of the healthy side.

In terms of muscle activation timing, the present study observed that the onset of activation (onset-IC) of several key muscle groups in the ACL-injured group was significantly earlier than that of the control group, especially the posterior masseter muscles (BF, ST) and Gmax. This early pre-activation pattern was considered as part of the “pretension conditioning” during landing, which aims to By preactivating the muscles a sufficiently stable base prior to contact with the ground to control knee trajectory and absorb ground reaction forces. In the case of the quadriceps muscle, for example, it has been shown that healthy individuals preactivate the VM and VL 103–114 ms before ground contact, and the injury group in this study showed a similar or even earlier activation sequence. In addition, there is an electromechanical delay (EMD) between the EMG signal and actual tension generation, which ranges from 20 to 100 ms for knee extensors33 and may reach 55–92 ms for hamstrings34, making early activation particularly critical. Analysis of the time of peak EMG appearance (peak-IC) revealed that both BF and ST peaks were significantly delayed in the ACL injury group, which is highly consistent with the hamstring function of action. The hamstrings are the main co-stabilizing muscles of the ACL by generating tibial “posterior drawer force” to reduce anterior displacement. During landing maneuvers, peak tibiofemoral shear occurs approximately 28–30 ms after touchdown25,26. Therefore, a delay in maximal hamstring output to this stage coincides with the most vulnerable window of the ACL, thus forming a dynamic protective barrier. Blackburn et al.34 state that hamstring EMD is 72 ms, which is highly consistent with the result that hamstrings reach peak activity approximately 30–40 ms before touchdown. This is highly consistent with the result that the hamstrings reach peak activity around 30–40 ms before touchdown.

Gmax showed a similar pattern of regulation. Not only was Gmax activated earlier before touchdown in the ACL injury group, its peak activation also occurred around 40 ms after landing, and the abduction moment generated by Gmax was expected to peak 92 ms after landing. This coincides with the electromechanical delay (52 ms on average) in hip abduction proposed by Kim et al.35, suggesting that its abduction output is functioning right at the peak tibiofemoral shear force phase. It has been reported that the peak knee valgus angle mostly occurs 85 ms after landing (range 70–100 ms), and more than 60% of athletes reach their maximal knee valgus angle at this stage36,37. Therefore, early activation of Gmax helps to counteract the tendency of hip adduction and knee valgus and prevents pelvic instability from adversely affecting the knee joint. Overall, patients with ACL injuries show a typical “pre-activation-post-peak” bi-directional strategy in landing tasks: key muscle groups are activated early to pre-establish muscle tone in response to ground impact, and the peak is moderately delayed to match the timing of the peak of the ground reaction force and the emergence of shear forces, thus maximizing their ability to respond to ground impact. This reprogramming of neuromuscular timing is clinically associated with reduced reinjury risk, improved motor control, and enhanced patient functional scores. Therefore, targeted strengthening of the preactivation and timing coordination of the hamstrings and Gmax during rehabilitation training is expected to improve dynamic knee stability and support patients’ safe return to sports.

In the ACL-injured group, the peak EMG values of the quadriceps muscles (VL, VM) were significantly lower than those of the healthy control group within 100 ms after landing, suggesting that the activation of the knee extensor muscles during the touchdown phase was reduced. This feature is referred to as the “quadriceps avoidance strategy”, reflecting the fact that patients may actively reduce extensor output to avoid knee instability, which affects joint braking and postural control38. In contrast, the hamstrings showed a tendency to increase activation, especially ST, which peaked significantly within 100 ms before touchdown, and BF, which tended to increase both before and after touchdown. As a co-stabilizing muscle of the ACL, popliteus mainly limits anterior translation and internal rotation by generating tibial posterior moment, and its enhanced recruitment contributes to knee stability and compensates for the lack of ACL function. This feature fits with the electromechanical delay time of the hamstrings, which ensures the provision of an effective response during the peak phase of shear32. Gmax, although showing a reduced peak after landing in the injury group, has an earlier onset and longer duration, suggesting that it compensates through temporal regulation to maintain hip and pelvic stability. Hamstrings take on more loads when the function of Gmax is reduced, and their enhanced activation may also partly stem from the indirect compensation for gluteal insufficiency39. In addition, ACL injury may interfere with central drive, resulting in decreased efficiency of motor unit recruitment, further exacerbating Gmax underactivation40. Overall, the ACL-injured group showed an activation pattern of “extensor inhibition - flexor strengthening - gluteal sequential compensation”, which reflects the self-regulation mechanism of the neuromuscular system under the condition of impaired joint stability. Clinically, quadriceps avoidance strategies are often associated with subjective instability and decreased functional jumping performance. However, enhanced activation of the hamstrings and Gmax may assume a new dynamic stabilizing function after ACL loss, helping patients maintain joint stability. Therefore, during rehabilitation, long-term quadriceps inhibition should be avoided, and strength and control training of the hamstrings and Gmax should be strengthened to compensate for the mechanical deficiencies caused by ACL loss and improve clinical functional outcomes. Although the increased activation of the hamstrings and Gmax in this study was primarily interpreted as a compensatory response to ACL injury, whether this phenomenon also indicates that these muscle groups have assumed a new stabilizing role after ACL loss warrants further investigation. Numerous studies have shown that the hamstrings and Gmax play an important role in knee stability12,41. Following ACL injury, they may not simply be a compensatory response but rather assume a new dynamic stabilizing function. The hamstrings can unload the ACL by generating posterior tibial shear forces, while the gluteal muscles unload the ACL by resisting knee valgus moments42, a function that is particularly important after ACL injury. To further validate this, future studies should incorporate biomechanical data to investigate the role of the hamstrings and Gmax in knee stability after ACL loss.

This study had a relatively limited sample size (21 patients with ACL injuries and 21 healthy controls). Statistically, this small sample size may have reduced power, leading to some true effects not reaching significance. It may also have magnified the influence of individual differences on the mean, increasing the instability of the results. Physiologically, the small sample size may have masked differences in myoelectric patterns within specific populations. The subjects were primarily younger. Neuromuscular adaptations in younger individuals may manifest as faster preactivation and greater force output, which may underestimate the degree of quadriceps avoidance or hamstring compensation in landing strategies of older individuals. Therefore, the generalizability of the present results to other age groups is limited. Furthermore, all ACL-injured patients in this study had their dominant leg injured. The dominant leg typically possesses greater strength and more mature motor control, potentially demonstrating earlier preactivation and longer sustained activation to maintain task stability. This characteristic may have magnified the observed compensatory trends in the hamstrings and Gmax; conversely, the magnitude and strategy of compensation may differ in the non-dominant leg. Therefore, the omission of the non-dominant leg limits our inferences about the relationship between limb dominance and EMG timing and intensity. Additionally, because this study was primarily exploratory in nature, p-values were not adjusted for multiple comparisons. The absence of p-value correction should be considered when interpreting the results. Future studies with larger and more diverse samples are warranted to validate these findings and, where appropriate, to adopt false discovery rate (FDR)-based approaches such as the Benjamini–Hochberg procedure to enhance statistical rigor, particularly in confirmatory research.

Conclusion

In a single-leg jump landing task, patients with ACL injuries of an average disease duration of 18 months generally had a prolonged muscle activation duration, reflecting compensatory mechanisms to maintain kinematic stability, despite a preactivation sequence consistent with that of healthy individuals. Preactivation of the biceps femoris, semitendinosus, and gluteus maximus muscles in the affected limb occurred with delayed hamstring peaks, potentially contributing to knee stability by modulating tibial motion and reducing excessive anterior translation and internal rotation.

Fig. 6
figure 6

Comparison of RMS of each muscle within 100ms after IC. note: DL indicates dominant leg; NDL indicates nondominant leg; *, indicates p < 0.05 for comparison between the dominant leg and other groups in the injury group; **, indicates p < 0.01 for comparison between the dominant leg and other groups in the injury group; ***, indicates p < 0.001 for comparison between the dominant leg and other groups in the injury group; #, indicates p < 0.05 for comparison between the nondominant leg and other groups in the injury group; ##, indicates p < 0.01 for comparison between the nondominant leg and other groups in the injury group; RMS, indicates root mean square amplitude.

Full size image