Lower extremity kinematic analysis in male athletes with unilateral anterior cruciate reconstruction in a jump-landing task and its association with return to sport criteria

Twenty-seven athletes with ACLR (23.8 ± 3.3 years) and 15 healthy athletes with no history of any lower extremity injury (24.6 ± 3.1 years) participated in this cross-sectional study. Out of 27 participants with unilateral ACLR, 14 passed the RTS criteria (passed-RTS group), while 13 did not (failed-RTS group). Table 1 presents the characteristics of the participants of the three groups. All Participants were male football players, recruited from the local sport physiotherapy clinics and football clubs over a period of 6 months. Inclusion criteria were: age 18 to 40 years, unilateral ACL reconstruction at least 6 months before the study beginning [5, 17], ability to perform regular jumping and pivoting activities (≥50 h per year) before the injury [18]. The exclusion criteria were a meniscal tear of greater than grade I [19], osteochondral defect, a complete tear of other knee ligaments, a history of other lower extremity joint surgery and neuromuscular disorders in the hip, knee, ankle and lower back that limits the ability of athletes to perform sports activities. Participants in the control group had no history of lower limb and back injury or surgeries. The ACLR participants completed a progressive rehabilitation treatment supervised by licensed sports physiotherapists, and had full knee ROM, knee joint effusion of grade trace or zero based on the stroke test [20] and ability to perform the single-hop test without pain. A sport physiotherapist screened participant for the inclusion/exclusion criteria. All participants read and signed an informed consent approved by the university ethics committee (ethics approval number: IR.AJUMS.REC.1396.579).

Participants underwent RTS and kinematic assessments in two separate testing sessions. In the first session, their demographic data including age, weight, height, and Tegner score (before the injury and 6 months after ACLR) [21] were recorded. They were also asked to fill the Knee Outcome Survey–Activities of Daily Living Scale (KOS-ADL) and the Global Knee Rating Scale questionnaires for each limb separately and without any specific cutoff [22]. An experienced sports physiotherapist performed the RTS testing. Participants were allocated to the failed and passed RTS group based on the outcomes of four commonly accepted RTS criteria including maximal isometric quadriceps strength, four functional single-legged hop tests (single, triple, crossed and 6-m-time-to-hop), the KOS-ADL and the Global Knee Rating Scale questionnaires. These particular tests are considered important to estimate the functional/neuromuscular status of athlete who want return to their pre-injury level of activities [3]. Quadriceps strength test was performed using a hand held dynamometer according to the standard belt-stabilized method [23]. For the hop tests for distance, athletes stood on one leg with their toes behind the starting line, hopped forward as far as possible and landed on the same leg. For the single and triple hop tests, athletes performed a single hop, and three consecutive hops, respectively. For the crossed hop test, they performed three consecutive hops while crossing the midline with each hop. The distance of hop was recorded for each test in centimeters (cm). For the 6-m hop test, they hopped on one leg as quickly as possible over a distance of 6 m. The tests were performed in random order and for both legs to calculate the limb symmetry index (LSI) (\( \frac{\mathrm{involved}}{\mathrm{uninvolved}} \)) [24]. To pass the RTS criteria, each participant had to gain a minimum LSI of 0.90 on all RTS criteria. All RTS tests were performed at least 6 months after ACL reconstruction [5].

A 7-camera (240 Hz) motion analysis system (Qualysis, Gothenburg, Sweden) was used to capture the 3D position of 36 reflective markers attached to the participants’ lower extremity while they performed a bilateral jump-landing task. The landing task adopted in this study (see below) is a common test to investigate biomechanical landing safety in ACLR athletes [25]. The markers placed bilaterally on the specific anatomic landmarks of the feet, ankles, shanks, knees, thighs, and pelvis to track segments’ motion during each trial. A three-second upright standing static trial was recorded before the landing trials, to align the participant with the laboratory coordinate system and to build subjects’ static reference model for kinematic analyses.

All participants performed a light intensity pedaling on a stationary bicycle for 5-min to warm-up before participating in the jump-landing test. For the jump-landing task, the participants jumped from a 30 cm box and landed at a distance of 50% of body height away from the box (as the landing area) followed by a maximal vertical jump [9]. Each participant performed 2 practice trials, followed by 10 actual test trials. A trial was considered successful if the individual jumped bilaterally from the box, landed in the landing area. Participants could rest at least 1 min between the consecutive trials. Participants wore a pair of standard athletic shoes that has previously been adjusted to place the reflective markers.

Ten successful jump-landing tests for each limb were used for further analyses using Visual3D software (C-motion Inc., Kingston, Canada). Data was low pass filtered with a cutoff frequency of 6 Hz followed by identifying the initial contact using the lower leg segment’s vertical velocity. We concentrated on the initial contact of the landing task because previous studies demonstrated that ACL injury/re-injury mostly occurs at the initial contact of landing [8, 26, 27]. That is, the minimum vertical velocity of the lateral malleolus marker [16] was used as the initial contact moment. Moreover, manual inspection was used to verify correct identification and adjustment when required. Cardan angles were used to define the joint angles. The sequence of rotation was x-y-z, based on the recommendations of the International Society of Biomechanics [28]. Hip, knee and ankle kinematics in the three anatomical planes of motion were examined at initial contact, and the average of the 10 landing trials was used for statistical analysis. The positive sign assigned to the hip and knee flexion, adduction, and internal rotation, ankle dorsiflexion, and ankle eversion.

Statistical analysis was completed using IBM SPSS Statistics Version 25.0 (SPSS, Inc., Chicago, IL), and statistical significance was set at p < 0.05. Normality of distribution of kinematic data was checked using Shapiro–Wilk normality tests. The limb symmetry indices for the functional, quadriceps strength and questionnaires tests (Table 2), and demographic characteristics (Table 1) were compared between the three groups using one-way analysis of variances (ANOVAs). The kinematic variables were compared between the limbs (involved/uninvolved) and the groups (passed-RTS, failed-RTS and healthy) using nine separate mixed ANOVAs. Post-hoc tests with Bonferroni adjustment were performed to locate the source of significance.

No significant limb-by-group interactions were found for the hip, knee, and ankle 3D kinematics (p > 0.05). The main effect of group was significant for hip abduction/adduction (F = 10.541, p < 0.001), hip rotation (F = 5.081, p = 0.011) and ankle eversion/inversion (F = 11.384, p < 0.001). The post-hoc comparisons revealed lower hip abduction angle in both failed (involved limb 10.6 ° ± 4.2, p < 0.001) and passed RTS (involved limb 11.6° ± 3.1, p = 0.006) groups compared to the healthy group (non-dominant limb 16.6° ± 3.7). However, hip abduction was not different between the failed- and passed-RTS groups (p > 0.05). With regard to hip transverse plane kinematics, post-hoc test found lower hip external rotation in the failed-RTS group (involved limb 12.7° ± 6.1) compared to the healthy group (non-dominant limb 15.7° ± 8.3, p = 0.01), but no significant differences between the failed- and passed-RTS groups (p = 0.152), and also between the passed-RTS and healthy groups (p = 0.923).

There were no significant differences between the groups in knee kinematics (p > 0.05). Ankle inversion in the failed-RTS (involved limb 10.2° ± 5.1) was significantly lower than both the passed-RTS (involved limb 12.1° ± 4.8, p = 0.05) and the healthy (non-dominant limb 22.1° ± 8.2, p < 0.001) groups. Finally, there was no significant difference between the passed-RTS and the healthy groups for ankle inversion/eversion (p = 0.117).

The main effect of limb was only significant for hip abduction/adduction (F = 9.383, p = 0.006). The involved limb of failed-RTS had lower hip abduction compared to the non-involved limb (mean difference = 5.45° ± 6). The passed-RTS group and the healthy athletes had relatively similar hip abduction/adduction (Table 4 and Fig. 1).

In the passed-RTS group, despite appropriate improvement in the muscle strength and functional performance (LSI ≥ 90%), lower hip abduction angle was found compared to the healthy group. This result is consistent with previous studies that reported a tendency toward hip adduction during walking and landing in individuals with ACLR who returned to play [25, 32]. Research suggest an association between ACL injury and excessive hip adduction and internal rotation during multiplanar movement, e.g., landing [8, 33]. Thus, reduced hip abduction and external rotation seen in ACLR athletes who passed RTS criteria may indicate that current strength and function-based RTS criteria is not sufficient to identify athletes at risk of ACL re-injury and highlights the need for lower extremity kinematic analysis together with current RTS criteria. Altered hip kinematics in athletes who passed RTS may also suggest the need for more attention to the strengthening of the hip and core stabilizers, and neuromuscular training following ACLR as positive effects of hip strengthening and neuromuscular training on changing frontal plane hip motion reported in previous studies [34‐36].

The failed- and passed-RTS groups were only statistically different in the ankle inversion, with no significant differences in the hip and knee kinematics. Consistent with our finding, Chang et al. (2019) reported no significant difference in knee kinematics between female athletes who pass and fail the RTS criteria. The similar hip and knee kinematics in the two groups, even with different RTS scores, could indicate that functional or strength asymmetry is not associated with the kinematic characteristics of lower extremity in the ACLR athletes. This is in agreement with the results of a recent study by Xergia et al. (2015) who found no significant correlation between functional asymmetry assessed by LSI for the hop test and lower extremity kinematics during landing [37]. However, similar hip and knee kinematics of failed- and passed-RTS groups in our study is contrary to the finding of Di Stasi et al. (2013) who found different hip and knee kinematics in the passed- and failed-RTS groups during treadmill walking [5]. This controversy may be due to the differences in the task evaluated, as it has been shown that the kinematics of gait and landing tasks are not associated [38].

Similar lower extremity kinematics in failed- and passed-RTS groups indicates that passed RTS athletes still have biomechanical risk factors for ACL re-injury similar to the failed group despite passing the return to sport criteria. Therefore, kinematic analysis together with current RTS may provide more detailed information for increased safety to return to sport decision making.

In our study, participant in the failed-RTS group landed with less ankle inversion when compared with the passed-RTS group. Because of linked motion between lower extremity joints during close kinetic chain tasks such as landing, changes in the ankle frontal plane motion may impact hip and knee joint kinematics [39, 40] and may increase the risk of ACLR injury/re-injury [41]. Previous studies reported that Change in ankle motion during landing may contribute to Transverse plane hip and frontal plane knee positioning when contacting the ground, which are known to increase the risk of lower extremity injury [42, 43].

Hip abduction (mean diff = 4.5°), hip external rotation (mean diff = 5.6°), and ankle inversion (mean diff = 7.8°) were significantly lower in the failed-RTS group compared to the healthy group. Previous studies also reported significant differences between the ACLR and healthy groups in peak hip flexion [17, 24], hip abduction/adduction [25] and frontal plane knee motion [44]. Decreased hip abduction, external rotation and ankle inversion may increase tendency toward dynamic valgus phenomena in the initial phase of landing which is an important risk factor to ACL re-injury [8]. Paterno el al. (2010) in a prospective study also reported that excessive inward motion of lower extremity in frontal plane during landing results in three times increased risk of ACL re-injury following ACL reconstruction [8]. These kinematic changes in the lower extremity joints in the failed RTS group, even several months after the reconstruction surgery and rehabilitation, may further highlight the need for more attention to movement education/training and neuro-biomechanics of motion during ACLR rehabilitation.

We found lower hip abduction angle in the ACL reconstructed limb compared to the intact limb in the failed-RTS athletes. This may explain the higher rate of re-injury in the ACL reconstructed limb in individuals with a history of unilateral ACLR. That is, increased hip adduction, which can lead to a valgus knee position, is known as an important risk factor that can predict the risk of ACL re-injury [8]. Adam et al. (2018), in a meta-analysis, also showed that individuals with ACLR land with increased hip adduction in the involved limb compared to the intact limb [25].

One limitation of our study is that we did not evaluate hip muscles strength. This was because the focus of this study was to investigate the association between RTS criteria and lower extremity kinematics. Several studies, however, reported a relationship between hip strength and lower extremity kinematics during challenging activities such as landing [45, 46], step decent [47] and hopping [48]. Future studies should consider hip muscle strength while comparing lower extremity kinematics in the ACLR knee. Moreover, this study included only male participants which limit generalizability of our outcomes to females with ACLR. This was mainly due to the low rate of female ACLR athletes in our available population. Also considering kinetic data in conjunct with kinematics in the future studies can improve our knowledge about landing biomechanics as a suggested task to evaluation of athlete readiness to return to sport after ACLR and its relationship with current RTS criteria. Finally, Current RTS testing is performed under pre-planned/anticipated circumstances. There is a plethora of studies investigating the effects of unplanned/unanticipated jump landings/cuttings on lower limb biomechanics [49‐51]. As team sport athletes interacting in a variable and unforeseen environment, such tests may better mimic the realistic affordances (increase ecological validity).