The anterior cruciate ligament (ACL) plays an integral role in the mechanical stability of the knee, functioning as the primary restraint to sagittal plane tibial displacement and providing restraint to frontal and transverse plane movements.1 The ACL also contributes to knee joint sensorimotor control and proprioception.2 Previous research showed that ACL injury negatively impacts knee joint sensorimotor control and proprioception.3
The overall incidence of ACL injury is reported to be higher in female athletes, with most injuries reported as a non-contact mechanism. Agel et al.4 reported on ∼4 times higher risk of ACL injury in female athletes, while a recent report by Waldén et al.5 reported an ∼3 times higher risk in female soccer athletes.
ACL reconstruction surgery is typically recommended following complete rupture.6, 7 The aim is to restore mechanical stability to the joint so that the athlete can return to sports. However, controversy exists as to whether return to full sporting participation is the correct option for every athlete.7 A study by Waldén et al.8 reported that professional soccer players, upon return to sport after surgery had a significantly higher risk of subsequent knee joint injury when compared to players without a previous history of ACL injury. Furthermore, the development of post-traumatic osteoarthritis9 (OA) and ACL re-injury10, 11 are of great concern following return to sporting participation after surgery.
The development of post-traumatic OA could be influenced by altered lower limb kinematic profiles.9 Previous research showed that aberrant lower limb kinematic profiles during jump landings puts female athletes at risk for initial ACL injury.12 Recent publications showed that kinematic and sensorimotor deficits are present during walking and single leg landings in ACL reconstructed (ACL-R) athletes.13–17 However, the extent to which these aberrant lower limb kinematic profiles are present during a higher velocity sport specific task, such as a drop vertical jump (DVJ) following ACL reconstruction surgery in female athletes who have returned to full sporting participation is still largely unknown.
Thus, our aim was to investigate hip and knee joint kinematic profiles in a group of ACL reconstructed (ACL-R) female athletes during the performance of a DVJ. We hypothesized that when compared to a non-injured sex, age, and activity matched control group, ACL-R female athletes would display altered lower limb kinematic profiles.
This investigation employed a laboratory based case–control design. The independent variable was group (Controls vs. ACL-R). The dependent variables were hip joint adduction–abduction angle, hip joint flexion–extension angle, hip joint internal–external rotation angle, knee joint adduction–abduction angle, knee joint flexion-extension angle, and knee joint internal–external rotation angle. Subjects reported to the laboratory on one single occasion for testing (Figs. 1–4).
Fourteen recreational female athletes (age, 20.8 ± 1.1 years; height, 1.65 ± 0.06 m; body mass, 65.4 ± 7.4 kg; BMI, 23.8 ± 2.4 kg/m2) without previous history of knee joint injury volunteered to participate.
Fourteen recreational female athletes (age, 23.0 ± 3.4 years; height, 1.64 ± 0.05 m; body mass, 64.85 ± 8.67 kg; BMI, 24.0 ± 2.4 kg/m2) who had previously undergone ACL reconstruction for a non-contact ACL injury volunteered to participate. Eight had surgical stabilization using a bone-patellar tendon-bone auto-graft; the remainder had a hamstring auto-graft procedure. The mean time from surgical stabilization to the study was 4.4 years.
International Knee Documentation Committee Subjective Knee Form (IKDC) and Knee Injury and Osteoarthritis Outcome Score (KOOS)
All subjects completed the IKDC questionnaire and the KOOS subscales (KOOSpain, KOOSsymptoms, KOOSactivities of daily living (ADL), KOOSsport, KOOSknee-related quality of life (KQoL)).
Each subject was instrumented with the CODA bilateral lower limb gait set-up. Prior to the attachment of lower limb markers and wands, each subject had specific anthropometric measures recorded for the calculation of internal joint centers at the hip, knee, and ankle joints as described in Monaghan et al.18, 19 The markers and marker wands were then applied by the same investigator for all subjects and were positioned as follows: the lateral aspect of the knee joint line in the median frontal plane, the anterior aspect of the lateral malleolus, the posterior inferior lateral aspect of the heel, and the lateral aspect of the fifth metatarsal head. Wands with anterior and posterior markers attached were positioned on the pelvis and sacrum, the thigh, and the shank. The thigh wand was aligned perpendicular to the knee joint line using a T-shaped frame (Charnwood Dynamics Ltd, Leicestershire, UK). The tibial wand was aligned perpendicular to the ankle joint line using a goniometer after medial malleolus and lateral malleolus points are projected onto the floor. The markers were fixed to the skin and the wands with double sided adhesive tape. This set-up was previously utilized in our laboratory.18–22 A neutral stance trial was used to align the subject with the laboratory coordinate system and to function as a reference position for subsequent kinematic analysis. Kinematic data acquisition was made at 200 Hz using 3 CODA mpx1 units (Charnwood Dynamics Ltd, Leicestershire, UK), which were fully integrated with 4 AMTI (Watertown, MA) walkway embedded force-plates. The CODA mpx1 units were time synchronized with the force-plates.
Drop Vertical Jump
Each subject performed three DVJ trials consisting of starting on top of a 35-cm box with the feet positioned shoulder-width apart and hands placed on the hips. They were instructed to place one foot forward and drop off the 35-cm box, landing simultaneously on both feet, one each on two adjacent force plates. Upon landing, subjects immediately performed a maximum vertical jump, while keeping their hands on their hips. The technique is similar to that previously described by Ford et al.23 Each subject was allowed a maximum of five practice trials for familiarization with the technique.
Kinematic data were calculated by comparing the angular orientations of the coordinate systems of adjacent limb segments using the angular coupling set “Euler Angles” to represent clinical rotations in 3D. Marker positions within a Cartesian frame were processed into rotation angles using vector algebra and trigonometry (CODA mpx30 User Guide, Charnwood Dynamics Ltd, Leicestershire, UK). Joint angular displacements were calculated for the hip and knee in the sagittal, frontal, and transverse planes. Kinematic data were analyzed using the CODA software. The point of initial contact (IC) with the force-plate to 200 ms post-IC was extracted from the initial landing phase of the DVJ. Kinematic data for each DVJ trial were extracted and converted to a Microsoft Excel file. The average of three trials for each subject was utilized for further analysis with group profiles then being calculated (i.e. Controls vs. ACL-R). This technique was previously used in our laboratory.18–22 Time-averaged profiles were subsequently used for analysis, as well as the magnitude of the peak hip and knee joint angular displacements.
Between-subjects ANOVAs were used to test for between group differences in the IKDC and KOOS subscale scores. Associated effect sizes were calculated using the formula: sum of squares between-groups/total sum of squares,24 and quantified according to Cohen25 as 0.01 = small effect size, 0.06 = medium effect size, and 0.14 = large effect. Between-subjects ANOVAs were used to test for between group differences in magnitude of the peak hip and knee joint angular displacements. Time-averaged profiles for hip and knee joint kinematics were calculated for each participant, with group mean profiles then subsequently calculated. Differences in ACL-R and control group time-averaged profiles were tested for significance using independent two-sided t-tests for each datum point. All data were analyzed using Predictive Analytics SoftWare (Version 18, SPSS, Inc., Chicago, IL) with significance set at p < 0.05.
IKDC and KOOS
ACL-R subjects differed significantly from the control group subjects on the IKDC (p < 0.01) and the KOOS pain (p < 0.05), symptoms (p < 0.01), sport (p < 0.01), and KOOS KQoL (p < 0.01) (Table 1). The associated effects sizes were large. No significant difference was found between the ACL-R and control subjects on the KOOSADL (p > 0.05).
Table 1. IKDC and KOOS Subscale Results
Values are presented as mean ± SD.
Significantly different from control participants.
The ACL-R group exhibited a significantly increased peak hip adduction (p = 0.01) and internal rotation (p = 0.03). No between group difference (p = 0.35) was noted for peak hip flexion (Table 2). The ACL-R group also exhibited a significantly decreased peak knee adduction (p = 0.03) and flexion (p < 0.01). No between group difference (p = 0.33) was noted for peak knee joint internal rotation.
Table 2. Peak Hip and Knee Joint Angular Displacements
Values are presented as mean ± SD.
Hip adduction–abduction; adduction is positive, abduction is negative.
Hip flexion–extension; flexion is positive, extension is negative.
Hip internal–external rotation; internal rotation is positive, external rotation is negative.
Knee adduction–abduction; adduction is positive, abduction is negative.
Knee flexion–extension; flexion is positive, extension is negative.
Knee internal–external rotation; internal rotation is positive, external rotation is negative.
Significantly different from control participants.
The hip joint adduction-abduction and internal–external rotation kinematic profiles differed significantly between the ACL-R and control groups; the ACL-R group exhibited a more adducted (p < 0.05) and internally rotated (p < 0.05) position of the hip during the entire 200 ms time frame following IC (Figs. 1 and 2).
The knee joint adduction-abduction and flexion-extension kinematic profiles differed significantly as well. At IC, the ACL-R group displayed an abducted (valgus) knee position in contrast to the adducted position of the knee at IC in the control group. From 78 to 104 ms and 160 to 200 ms post-IC, the ACL-R group displayed a significantly less (p < 0.05) adducted position of the knee (Fig. 3). They also displayed significantly less (p < 0.05) flexion of the knee from 106 to 182 ms post-IC (Fig. 4).
Our primary hypothesis was supported by the results, which indicate that even upon return to sporting participation, female ACL-R athletes exhibit altered lower limb kinematic profiles when compared to an age, sex, and activity matched control group.
IKDC and KOOS
The IKDC is a reliable and valid knee specific measurement of symptoms, function, and sports activity appropriate for a wide variety of knee problems. It was designed as an evaluative instrument to measure improvements or deterioration,26 and was validated for use in an ACL reconstruction population.27 In our study, a significant difference (p < 0.01) of >9 points was observed for the IKDC scores of the ACL-R and control groups. This difference is the threshold for identifying a clinically meaningful effect,26 while the associated effect size was large.
The KOOS questionnaire was designed to evaluate short and long term symptoms and functions in patients with a variety of knee injuries that could possibly result in OA,28 and has been validated for use with an ACL reconstruction population.28 Unlike the IKDC, the KOOS has separate scores for different aspects of the questionnaire: pain, symptoms, activities of daily living, sport and recreation function, and knee related quality of life. Separating the KOOS into 5 scores makes it easier to follow the phases of rehabilitation and helps clarify the major concerns of the patient.29 In our study, a significant difference (p < 0.01) between the ACL-R and control groups was observed on four of the subscales: (KOOSpain; KOOSsymptoms; KOOSsport; and KOOSKQoL). The associated effect sizes were all large, indicating clinical meaningfulness of the differences. No significant between-group difference (p > 0.05) was observed on the KOOS ADL subscale.
We decided to use both the IKDC and 5 KOOS subscales. A recent study by Hambly and Griva30 supports our approach. The observation of no significant between-group difference for the KOOSADL is not unexpected; Hambly and Griva30 reported that the items for the KOOSADL were neither viewed as being particularly important by participants, nor were they frequently experienced. Furthermore the KOOSADL questions involve low stress tasks such as climbing stairs or taking off socks that would pose little difficulty to our female athletes. Our results show that although ACL reconstruction allows for a complete return to activities of daily living, the injury continues to affect patients in several aspects of life, especially during more demanding tasks, for a long time post-reconstruction.
Hip Joint Kinematics
During the DVJ, the ACL-R athletes were characterized by a more adducted and less externally rotated position of the hip for the entire time frame from IC to 200 ms post-IC. Furthermore, the ACL-R group displayed a significantly greater peak hip adduction (p = 0.01) and internal rotation (p = 0.03) when compared to the control group.
At IC, the ACL-R group displayed a relatively neutral hip position in the frontal plane. However, following IC the ACL-R group progressively displayed an increase in hip joint adduction that peaked at ∼136 ms post-IC. The control group remained in an abducted position throughout the entire 200 ms post-IC.
At IC, the ACL-R group displayed a relatively neutral hip position in the transverse plane. Approximately 32 ms post-IC, the ACL-R group began to move towards a position of hip joint internal rotation, which continued until 130 ms post-IC. The control group remained in a position of hip joint external rotation throughout the entire 200 ms post-IC.
Recently the idea of a multi-planar ACL injury mechanism has been put forth,31 with the authors suggesting that future work should examine the potential contribution of proximal structures on knee joint biomechanics. Our study suggests that even after ACL reconstruction and return to sport, female athletes display altered hip joint frontal and transverse plane kinematics during DVJs. Our results are supported by Paterno et al.,32 who indicated that hip joint transverse plane motion was an important predictor of second ACL injury following primary ACL reconstruction. Furthermore, Ireland33 illustrated the potential role of aberrant hip joint frontal plane motion in ACL injury, whereby the hip joint assumes an excessively adducted position. Further support is put forth by Myer et al.34 who emphasize the importance of lateral and posterior hip joint neuromuscular control in preventing knee joint valgus. We feel that the kinematic deficits observed at the hip in our study could be contributing to the increased abducted (valgus) position immediately following IC observed at the knee joint.
Knee Joint Kinematics
During the DVJ, the ACL-R athletes were characterized by a less adducted and less flexed knee position during specific time points following IC. At IC, the ACL-R group demonstrated a position of knee abduction that continued until ∼24 ms post-IC. Then throughout the remaining time they were characterized by a less adducted knee position when compared to the control group (significant from 78 to 104 ms post-IC and 146 to 200 ms post-IC). The ACL-R group demonstrated a position of less knee joint flexion during the time period from 106 to 182 ms post-IC.
Our findings in relation to the kinematic profile at the knee are consistent with previous studies that have examined ACL-injury mechanisms.12, 32, 35 A recent model-based image matching study identified that knee joint abduction during the first 40 ms post-IC is likely a key factor in the injury mechanism.35 Furthermore, the ACL-R group displayed a less adducted knee position throughout the landing, up to a difference of 8° from controls. This is consistent with findings of Hewett et al.,12 who in a prospective study identified that athletes who subsequently developed an ACL injury had on average 8 more degrees of abduction.
Furthermore, the ACL-R group were characterized by a position of less knee joint flexion during the time period from 106 to 182 ms post-IC. A position of less flexion has been implicated as a contributing factor to ACL injury32 based on the premise that a decrease in flexion during loading increases the anterior shear force at the proximal end of the tibia.36–38 Thus, our results support the contribution of a sagittal plane knee joint kinematic deficit as being a risk factor for future knee injury.
Our study shows that lower limb kinematic deficits still exist in female athletes following ACL reconstruction. The sensorimotor system comprises all of the sensory, motor, and central integration and processing components that govern the maintenance of joint homeostasis during dynamic movement, and thus is responsible for overall functional joint stability.39 The ACL contributes to sensorimotor control and proprioception,2 thus ACL injury influences lower limb sensorimotor control and joint stability. Previous research showed that aberrant lower limb kinematic profiles and sensorimotor control increase the likelihood of initial ACL injury.12 We believe that the altered lower limb kinematic profiles observed in our study may increase this risk following ACL reconstruction. This hypothesis can only be supported by a prospective study, however a recent study32 found that altered hip and knee joint sensorimotor control are predictors of second ACL injury, which supports our hypothesis.
Our results indicate that the kinematic deficits present following ACL reconstruction are multi-planar in nature, effecting frontal, sagittal and transverse plane movements. Thus, concerning rehabilitation, clinicians should complete a thorough evaluation of athlete sensorimotor control that addresses those specific multi-planar deficits. Evidence in the literature indicates that specific neuromuscular training can decrease the incidence of first time ACL injury in female athletes, with the proposed mechanism being attributed to an improvement in lower limb biomechanics.40 Future research is necessary to determine whether similar neuromuscular training can enhance sensorimotor control and reduce the risk of recurrent injury. Based on our findings, we would advocate the use of dynamic stabilization training, with emphasis placed on the correct execution of sports applicable techniques incorporating tri-planar hip and knee joint control.
A limitation of our study was that we did not report on trunk and ankle joint kinematics. In accordance with the recommendations of Quatman et al.,31 we recommend that future studies concentrate on investigating how proximal and distal neuromuscular control deficits can influence knee joint biomechanics and their role in ACL injury mechanisms. Furthermore, we did not perform sub-group analysis of reconstruction method or graft type. We would suggest that future studies examine the influence of reconstruction methods and graft type on lower limb kinematics and future injury incidence.
In conclusion, our results indicate that female ACL-R athletes still exhibit altered lower limb kinematic profiles during jump landing even after return to sport. These kinematic deficits involve frontal, sagittal and transverse plane movement, and could increase the risk of future injury. Future research should establish whether these altered hip and knee joint kinematic profiles can be positively influenced through a neuromuscular training protocol.