Differences in hip–knee joint coupling during gait after anterior cruciate ligament reconstruction

Clinical Biomechanics 32 (2016) 64–71

Contents lists available at ScienceDirect

Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech

Differences in hip–knee joint coupling during gait after anterior cruciate ligament reconstruction☆ Timothy C. Gribbin a,1, Lindsay V. Slater a,⁎,1, C. Collin Herb a,1, Joseph M. Hart a,1, Ryan M. Chapman a,1, Jay Hertel a,1, Christopher M. Kuenze b,1 a b

Department of Kinesiology, University of Virginia, Charlottesville, VA 22904, USA Department of Kinesiology, Michigan State University, East Lansing, MI 48824, USA

a r t i c l e

i n f o

Article history: Received 5 August 2015 Accepted 19 January 2016 Keywords: Variability Knee injury Biomechanics

a b s t r a c t Background: After anterior cruciate ligament injury, patients have increased risk for developing degenerative osteoarthritis, potentially due to the kinematic changes that persist after surgical reconstruction. Current research only describes single joint kinematic differences rather than the way in which two joints behave concurrently, termed joint coupling. The purpose of this study was to compare knee motion relative to hip motion in anterior cruciate ligament reconstructed and healthy limbs during walking and jogging. Methods: Thirty-seven recreationally active volunteers (22 reconstructed, 15 healthy) walked and jogged at 4.83 km/h and 9.66 km/h respectively. Vector coding methods were used to calculate stride-to-stride variability, magnitude, and vector angle of 6 joint couples during walking and jogging: hip frontal–knee frontal planes, hip frontal–knee sagittal, hip frontal–knee transverse, hip sagittal–knee frontal, hip sagittal–knee transverse, and hip transverse–knee frontal planes. Findings: The hip sagittal–knee frontal and hip sagittal–knee transverse joint couples had decreased variability during mid-stance, and all other couples had increased variability during the stance phase in the reconstructed group. The reconstructed group had decreased magnitude of joint excursion in the hip frontal–knee sagittal couple during all phases of gait during walking. Vector angles of the hip frontal–knee transverse couple increased in the reconstructed group during the loading, middle, and terminal stance phases, and swing phase of gait during walking. Interpretation: The increased variability and decreased magnitude of joint excursion indicate that movement patterns were less consistent during walking gait despite employing a more constrained system during movement in the reconstructed limb compared to healthy controls. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Anterior cruciate ligament (ACL) injuries are common, leading to an estimated 200,000 ACL reconstructions (ACLRs) in the United States each year (Brophy et al., 2009). After injury, ACLR is the most common treatment for active individuals (Gianotti et al., 2009) with the purpose of restoring joint stability to return an individual back to activity. However, only about 65% of individuals who undergo ACLR return to their previous level of activity, and only about half of patients return to any competitive level of sport (Ardern et al., 2014). Two of the most commonly reported reasons for not returning to sport are fear of re-injury and functional deficits in the knee (Ardern et al., 2014). After ACLR, individuals have also reported lower scores on the Knee

☆ There was no funding source for this project. ⁎ Corresponding author at: University of Virginia, 210 South Emmett Street, Memorial Gymnasium 222, Charlottesville, VA 22903, USA. E-mail address: [email protected] (L.V. Slater). 1 This author has no conflicts.

http://dx.doi.org/10.1016/j.clinbiomech.2016.01.006 0268-0033/© 2016 Elsevier Ltd. All rights reserved.

Instability and Osteoarthritis Outcome Score (KOOS) compared to healthy controls, specifically with lower scores in the return to sport and knee-related quality of life sections of the questionnaire (Lohmander et al., 2004). In this same study, 56% of ACLR individuals showed radiographic evidence of osteoarthritis during a 12 year follow-up after injury (Lohmander et al., 2004), which is consistent with previous literature reporting osteoarthritis rates after ACLR to be anywhere from 10%–90% of patients (Gillquist and Messner, 1999; Lohmander and Roos, 1994; Lohmander et al., 2004, 2007; Myklebust and Bahr, 2005; Øiestad et al., 2009). Patients with reconstructed knees also exhibit kinematic changes during gait. Although current research (Georgoulis et al., 2010; Hall et al., 2012; Knoll et al., 2004; Scanlan et al., 2010; Tagesson et al., 2014) suggests that knee sagittal and frontal plane kinematics are not significantly different between ACLR and healthy limbs, patients with ACLR display increased anterior tibial translation by up to 5.5 mm, and an offset of 1.3° to 3.0° of tibial external rotation (Georgoulis et al., 2010; Scanlan et al., 2010; Tagesson et al., 2014). Along with the changes in transverse motion at the knee joint, patients also display

T.C. Gribbin et al. / Clinical Biomechanics 32 (2016) 64–71

approximately 6° decreased hip flexion in the ACLR limb during stance phase of gait (Di Stasi et al., 2013; Noehren et al., 2013). Oftentimes, individuals continue to experience these alterations despite completion of rehabilitation programs and return to activity progressions (Ingersoll et al., 2008a). These changes in the knee transverse plane along with changes in hip kinematics during gait may be related to alterations in knee joint loading patterns. The constraint of the joint injury and reconstruction may lead to variable coordination patterns. Increased variability in movement and coordination patterns may potentially contribute to degenerative conditions of the knee joint such as osteoarthritis (Andriacchi and Dyrby, 2005). The dynamic systems theory states that the body will operate in the most efficient way based on the presenting constraints. Constraints are defined as anything that provides boundaries to the way a system organizes itself to accomplish a task (Davids et al., 2003). A constraint does not determine how the system organizes itself, but dictates the number of options available. These include organismic constraints, including those presented as a result of injury, environmental constraints, such as climate or playing surface, and task constraints, such as walking, jogging, jumping, and cutting (Davids et al., 2003). Additionally, the body has multiple degrees of freedom in which to complete a task. Taking advantage of these degrees of freedom, an individual can alter their movement strategy if one combination of movements either fails, or is no longer available due to newly imposed constraints (Latash and Turvey, 1996; McKeon and Hertel, 2006). Variability in task accomplishment can prevent the stressing of the same structures in the body over time. This may limit the repetitive loading of tissues and prolong the structural integrity. However, it is suggested that there may be an optimal amount of variability for different tasks (Riley and Turvey, 2002; Stergiou et al., 2006; van Emmerik and van Wegen, 2002), indicating that a task may require more consistent motion, and too much variability could be disadvantageous. Consistent loading at the knee joint during gait leads to cartilage growth in response to the chronic loading of the same area of the knee (Andriacchi, 2013). Increased stride-to-stride variability during gait after ACLR may play a role in cartilage degeneration. Female athletes with ACLR exhibit increased lower extremity variability during a side-step cutting task analyzed using vector coding methods (Pollard et al., 2015), implying that athletes with history of ACLR display altered neuromuscular control compared to healthy athletes. Similarly, female athletes also display greater ankle–hip relative phase variability during a unilateral dynamic postural rhythmic coordination task (Kiefer et al., 2013). This analysis was performed using Continuous Relative Phase (CRP) methods. The use of CRP to analyze coupled joint motion has some limitations (Miller et al., 2010), one of which being that it assumes that the data being analyzed is sinusoidal which may not be appropriate for all functional movement tasks (Hamill et al., 2000). Conversely, vector coding methods use the angular position of each segment, allowing direct comparison of joint segmental motion (Hamill et al., 2000; Miller et al., 2010). Vector coding methods may aid in developing a more thorough understanding of variability in hip-to-knee motion during gait in those with a history of ACLR. This is an important step in better appreciating the multiplanar mechanisms contributing to the development of osteoarthritis after ACLR. Therefore, the purpose of this study was to compare knee stride-to-stride variability between healthy and ACLR knees using vector coding methods. We hypothesize that recreationally active individuals with ACLR will display increased lower extremity variability during walking and jogging tasks compared to limb matched healthy controls. 2. Methods

65

Table 1 Mean (SD) for subject demographics for the ACLR and healthy groups.

Age (years) Gender Height (cm) Weight (kg) Body mass index

ACLR

Healthy

22.5 (5) 12 M/10 F 172.9 (7.1) 74.1 (15.5) 24.6 (4.0)

21.3 (3.3) 5 M/10 F 163.8 (1.8) 63.9 (12.5) 23.6 (3.4)

presents the ACLR group's surgical characteristics. Based on previous evidence that indicates a lack of impact of limb dominance on hip and knee kinematics (Sadeghi et al., 2000), ACLR limbs were matched by side of reconstruction to healthy limbs. The subjects were 18–40 years old, recreationally active, with a body mass index less than 35. Subjects were excluded if they reported a history of joint sprain within the past 6 weeks, cardiopulmonary disorder, neurological disorder, or were unable to complete 30 min of aerobic exercise. The subjects in the ACLR group were at least 6 months post unilateral, primary ACLR using either a bone-patellar tendon-bone or hamstring tendon autograft. The ACLR subjects were cleared by a physician or rehabilitation professional for return to full physical activity at the time of participation. Subjects were excluded if they had a multiple ligament reconstruction, failed meniscal repair, significant chondral resurfacing procedure, surgical complication, or history of graft failure. Informed consent was obtained for all participants. All methods were approved by the university's institutional review board (Kuenze et al., 2014). 2.2. Gait analysis We used a 12-camera motion analysis system (Vicon Motion Systems, Inc., Lake Forest, CA, USA) with retroreflective markers attached bilaterally to the ASIS, PSIS, lateral mid-thigh, lateral femoral condyle, and lateral malleolus, consistent with plug-in gait marker setup (Kadaba et al., 1990). Static trials were collected to calibrate the marker setup and provide reference for walking and jogging analysis. All subjects walked on the treadmill at a self-selected pace for 5 min to warm-up and to acclimatize to the treadmill and marker setup. Kinematic and kinetic data were collected while subjects walked and jogged on the treadmill at a speed of 4.83 km/h and 9.66 km/h respectively (Drewes et al., 2009a; McKeon et al., 2009) for three 15-s trials in order to ensure at least 10 full gait cycles were completed (Kuenze et al., 2014). The first 10 strides of the first complete trial were then selected for analysis in both walking and jogging conditions. Kinematic data were sampled at 250 Hz. Ground reaction force data were collected using a multi-axis strain gauge force plate imbedded under a custom-built treadmill (AMTI OR 6-7, Watertown, MA, USA) and sampled at 1000 Hz. A 60 N threshold was used to determine initial contact and toe-off. Kinematic and kinetic data were collected using VICON Workstation software (Version 5.0; VICON Motion Systems, Inc., Lake Forest, CA, USA) and extracted for analysis using a custom LabVIEW program (National Instruments, version 8.2.1; Austin, TX, USA) (Kuenze et al., 2014). Kinematic and kinetic variables were reduced to 101 data points representative of 0–100% of the gait cycle (Kuenze et al., 2014). The gait cycle during walking was divided into four phases: loading response/early stance phase was defined as 0%–20%, mid-stance phase

Table 2 Mean (SD) for ACL reconstruction group injury and surgical characteristics. Hamstring autograft Patella BTB

2.1. Subjects 22 ACL reconstructed (10 females, 12 males) and 15 (10 females, 5 males) healthy volunteers participated in this study (Table 1). Table 2

Gender 7 male/5 female Months from surgery 37.3 (26.3) Meniscectomy 3 med/4 lat

Total

5 male/5 female 12 male/10 female 24.5 (15.6) 31.5 (23.5) 2 med/2 lat 5 med/6 lat

BTB = bone-patellar tendon-bone autograft, Med = medial, Lat = lateral.

66

T.C. Gribbin et al. / Clinical Biomechanics 32 (2016) 64–71

was defined as 21%–40%, late stance phase was defined as 41%–60%, and swing phase was defined as 61%–100% (Kuenze et al., 2014; Novacheck, 1998; Ounpuu, 1994). During jogging, stance phase was defined as 0%–40%, and swing phase as 41%–100% of the gait cycle. 2.3. Joint coupling analysis Vector coding is a method that quantifies coupled motion between two segments (Tepavac and Field-Fote, 2001). Magnitude of excursion and vector angle are calculated for each consecutive data point throughout the gait cycle. Kinematic motion of the hip is plotted on the x-axis and motion of the knee is plotted on the y-axis. Magnitude and vector angle values were calculated using previously established vector coding methods using the following equation (Ferber et al., 2005; Heiderscheit et al., 2002; Herb et al., 2013; Mullineaux and Uhl, 2010; Sparrow et al., 1987; Tepavac and Field-Fote, 2001). The values of x and y refer to hip and knee motion respectively. Magnitude ¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2 ðxiþ1 −xi Þ2 þ yiþ1 −yi

ð1Þ

Vector angle values were adjusted to fall between 0° and 90°, as described by previous researchers using the following equation (Ferber et al., 2005; Herb et al., 2013): Vector Anglei¼abs½tan−1 yiþ1

−yi =xiþ1 −xi 

ð2Þ

In short, magnitude of joint excursion is determined by calculating the distance traveled by each segment between consecutive points, while the vector angle of consecutive points was calculated using a horizontal reference angle. Vector angles less than 45° indicate more hip motion relative to knee motion, angles greater than 45° indicate more knee motion relative to hip motion, and angles equal to 45° indicate equal hip and knee motion in the respective planes of the couple being analyzed. The magnitude and vector angle values are calculated across the gait cycle and used to determine the variability (Herb et al., 2013). Intersegmental variability coefficient, also known as vector coding variability (VCV), is the consistency of magnitude and vector angle quantified on a scale from 0 (no variability) to 1 (maximum variability) (Mullineaux and Uhl, 2010). 2.4. Statistical analysis Six joint couples were selected for analysis: hip frontal and knee frontal planes, hip frontal and knee sagittal planes, hip frontal and knee transverse planes, hip sagittal and knee frontal planes, hip sagittal and knee transverse planes, and hip transverse and knee frontal planes, similar to previous literature (Pollard et al., 2005). VCV, magnitude and vector angle were extracted from kinematic data using custom a MatLab processing code (MATLAB 2013a, The MathWorks, Inc., Natick, MA, USA). Means of VCV, magnitude and vector angle as well as associated 90% CIs were calculated throughout the gait cycle for both ACLR and healthy groups (Hopkins et al., 2009). Time epochs throughout the gait cycle during which the 90% CIs did not overlap for a consecutive 3% of the gait cycle were considered to be significantly different between groups (Drewes et al., 2009b). As a follow-up to this analysis in order to better understand the magnitude of difference between groups, between group Cohen's d effect sizes and associated 90% CIs were calculated (Eq. (1)) (Hopkins et al., 2009). Effect sizes were interpreted as weak (b0.2), small (0.21–0.39), medium (0.4–0.69), large (0.7–1.0), and very large (N 1.0).

0

Cohen s d ¼

Meancontrol −MeanACLR SDpooled

ð3Þ

3. Results 3.1. Hip frontal–knee frontal plane coupling During walking, the ACLR group had significant increases in VCV during mid-stance and late stance with very large effect sizes (Table 3, Fig. 1). During jogging, magnitude was greater in the ACLR group during initial swing (ACLR = 1.19°(0.38°), healthy = 0.79°(0.30°)). Vector angles were significantly decreased during swing (ACLR = 8.73°(1.33°), healthy = 19.28°(0.77°)) during jogging, indicating more proximal segment contribution to the coupled motion. 3.2. Hip frontal–knee sagittal plane coupling During walking, the ACLR group exhibited decreased magnitude during mid-stance, late stance, and swing with medium to large effect sizes (Table 4, Fig. 2). Vector angles were increased during walking at mid-stance with very large effect sizes (Table 5, Fig. 3). During jogging, vector angles were increased at late stance (ACLR = 9.48°(4.90°), healthy = 4.61°(1.28°)). 3.3. Hip frontal–knee transverse plane coupling During walking, the ACLR group had a significant increase in VCV during late stance (Table 3, Fig. 1). Magnitude was decreased during late stance and swing (Table 4, Fig. 2). Vector angles were increased during loading response, late stance, and swing (Table 5, Fig. 3). Effect sizes were very large for all significant differences during walking gait (Tables 3–5, Figs. 1–3). During jogging, vector angles were increased during swing (ACLR = 32.95°(5.03°), healthy = 14.82°(0.63°)). 3.4. Hip sagittal–knee frontal plane coupling During walking, VCV was decreased in the ACLR group at mid-stance (Table 3, Fig. 1), while magnitude was decreased during loading response (Table 4, Fig. 2) with large effect sizes. There were no significant differences during jogging. 3.5. Hip sagittal–knee transverse plane coupling During walking, the ACLR group had decreased VCV at mid-stance, however demonstrated increased VCV during late stance and swing (Table 3, Fig. 1). The ACLR group exhibited decreased magnitude during loading response and late stance through swing phase (Table 4, Fig. 2). Vector angles were increased during swing in the ACLR group (Table 5, Fig. 3). Effect sizes were medium to large for differences during walking (Tables 3–5, Figs. 1–3). During jogging, VCV was increased during swing (ACLR = 0.0665°(0.0002°), healthy = 0.0436°(0.0007°)). There were no differences in magnitude or vector angle during jogging. 3.6. Hip transverse–knee frontal plane coupling During walking, VCV was increased in the ACLR group at mid and late stance with very large effect sizes (Table 3, Fig. 1). Magnitude was increased in the ACLR group during swing phase with very large effect sizes (Table 4, Fig. 2). During jogging, magnitude was increased during 40%–53% of swing (ACLR = 1.37°(0.44°), healthy = 0.92°(0.36)), and 91%–93% of swing (ACLR = 1.91°(0.03°), healthy = 1.40°(0.01°)). 4. Discussion The ACLR group exhibited decreased magnitude of joint excursion in four joint couples during all phases of walking gait except for the hip transverse–knee frontal couple during late swing, which exhibited increased magnitude. Magnitude of the hip transverse–knee frontal couple was also increased in the ACLR group during two segments of

T.C. Gribbin et al. / Clinical Biomechanics 32 (2016) 64–71

67

Table 3 Vector coding variability (VCV) mean and standard deviation for ACLR and healthy groups for phases of gait with significant differences. VCV ranges from 0 indicating no variability, to 1 indicating maximal variability. Joint couple

Phase of gait (%)

ACLR

Healthy

Effect size (90% CI)

Hip frontal–knee frontal

24%–32% 49%–53% 51%–58% 25%–31% 25%–30% 53%–55% 67%–69% 27%–29% 45%–47%

0.44 (0.014) 0.30 (0.02) 0.27 (0.01) 0.19 (0.04) 0.18 (0.03) 0.41 (0.03) 0.16 (0.001) 0.39 (0.01) 0.42 (0.02)

0.29 (0.012) 0.20 (0.021) 0.19 (0.01) 0.27 (0.03) 0.26 (0.03) 0.27 (0.01) 0.11 (0.002) 0.29 (0.01) 0.28 (0.01)

11.7 (9.6, 13.8) 4.5 (3.5, 5.4) 7.3 (5.9, 8.7) −2.3 (−3.0, −1.7) −2.69 (−3.4, −2.0) 7.69 (6.3, 9.1) 35.85 (29.6, 42.2) 14.9 (12.3, 17.6) 12.0 (9.8, 14.1)

Hip frontal–knee transverse Hip sagittal–knee frontal Hip sagittal–knee transverse

Hip transverse–knee frontal CI = confidence interval.

swing during jogging. Vector angles increased in the ACLR group during all phases of walking gait. The ACLR group also exhibited increased VCV during mid and late stance phase of walking gait with the exception of the hip sagittal–knee frontal and hip sagittal–knee transverse couples during mid-stance. Traditional kinematic assessment typically evaluates single plane joint angles, which ignores the fact that joint motion during gait occurs across multiple segments in three planes. For example, previous research has shown that ACLR individuals have increased hip flexion angles during jogging than healthy controls (Kuenze et al., 2014). However, this does not explain motion in the knee or how motion in the two segments is related. By using vector coding analysis methods, these results help provide a more thorough understanding of the complex multi-segmental interactions that occur throughout the gait cycle in an ACLR population.

During walking gait, the ACLR group exhibited decreased magnitude of joint excursion, suggesting constrained motion at the joint in an effort to enhance stability during stance phase. This is consistent with prior findings of joint stiffening at mid-stance during walking (Hurd and Snyder-Mackler, 2007). As the body is presented with new constraints, such as injury, the system adapts by changing how it performs tasks. During gait, altering the magnitude of joint excursion is a potential adaptation. An increase in magnitude indicates that the joint couple has greater relative motion, while a decrease in magnitude indicates that motion is being restricted. The decreases in magnitude during areas where the body is transitioning between various stages of weight bearing may represent an adaptation in response to a previous event of instability, as limiting motion provides more control during the transitioning phases of gait.

Fig. 1. Effect sizes for mean differences in vector coding variability (VCV) between ACL reconstructed and healthy limbs during walking gait. Fig. 1A shows the specific location in the gait cycle when confidence intervals did not overlap with healthy controls for three or more consecutive points. Fig. 1B displays the effect size for significant differences between groups with 90% confidence intervals within each general gait phase (HFKF = hip frontal–knee frontal, HSKF = hip sagittal–knee frontal, HSKT = hip sagittal–knee transverse, HTKF = hip transverse– knee frontal, HFKT = hip frontal–knee transverse).

68

T.C. Gribbin et al. / Clinical Biomechanics 32 (2016) 64–71

Table 4 Magnitude (degrees) mean and standard deviation for ACLR and healthy groups for phases of gait with significant differences. Joint couple

Hip frontal–knee sagittal

Hip frontal–knee transverse Hip sagittal–knee frontal Hip sagittal–knee transverse Hip transverse–knee frontal

Phase of gait (%)

ACLR

Healthy

Effect size (90% CI)

22%–32% 35%–37% 48%–65% 80%–85% 49%–66% 82%–86% 13%–19% 9%–18% 53%–69% 88%–101%

0.51 (0.08) 0.45 (0.04) 1.89 (0.60) 2.28 (0.35) 0.35 (0.05) 0.67 (0.05) 0.93 (0.07) 0.82 (0.12) 1.15 (0.58) 1.55 (0.51)

0.67 (0.10) 0.60 (0.05) 2.19 (0.66) 2.46 (0.36) 0.59 (0.13) 0.90 (0.03) 1.01 (0.07) 0.92 (0.11) 1.39 (0.58) 1.05 (0.36)

−1.9 (−2.5, −1.3) −3.8 (−4.6, −2.9) −0.5 (−1.0, 0.03) −0.5 (−1.0, −0.02) −2.4 (−3.1, −1.8) −5.7 (−6.8, −4.6) −1.1 (−1.6, −0.5) −0.9 (−1.4, −0.4) −0.4 (−0.9, 0.1) 1.1 (0.6, 1.7)

CI = confidence interval.

Altering the relative contribution of each segment of the joint couple is another way the body may attempt to conform to new constraints, determined by the vector angle. The ACLR group displayed increased vector angles compared to the healthy group, indicative of more knee contribution to joint motion during walking. The largest difference in vector angles in the ACLR group was during the early and mid-stance phases, which are areas of weight acceptance and a transition from an eccentric contraction to gait propulsion. Increased knee contribution to joint motion during gait can be achieved through either decreased hip motion, increased knee motion, or both. After ACL injury and reconstruction, the effort to maintain knee joint stability appears to shift proximally, as shown by increases in frontal and sagittal plane hip moments (Kuenze et al., 2014; Osternig et al., 2000). The increased hip moment may be a compensation for quadriceps weakness following injury (Ingersoll et al., 2008b). The ACLR group did not exhibit differences in magnitude compared to

the healthy group during early and mid-stance in the hip frontal–knee transverse joint couple, but did display increased vector angles. This indicates that the magnitude of joint excursion was not different between groups, but the ACLR group accomplished the movement with decreased proximal segment contribution. This is suggestive that the hip is acting as a primary stabilizer in order to maintain normal gait mechanics through early and mid-stance and may represent an attempt to open more degrees of freedom at the hip to compensate for constraints at the knee. VCV was generally increased in the ACLR group through mid and late stance phases of walking gait, indicating increased stride-to-stride variability. This is consistent with previous literature which showed increased variability during the first 40% of the stance phase of a side-step cutting task, notably in the hip sagittal–knee frontal and hip transverse– knee frontal couples (Pollard et al., 2015). In the current study, significant increases in VCV for the hip sagittal–knee frontal and hip

Fig. 2. Effect sizes for mean differences in magnitude between ACL reconstructed and healthy limbs during walking gait. Fig. 2A shows the specific location in the gait cycle when confidence intervals did not overlap with healthy controls for three or more consecutive points. Fig. 2B displays the effect size for significant differences between groups with 90% confidence intervals within each general gait phase (HSKF = hip sagittal–knee frontal, HSKT = hip sagittal–knee transverse, HFKS = hip frontal–knee sagittal, HFKT = hip frontal– knee transverse, HTKF = hip transverse–knee frontal).

T.C. Gribbin et al. / Clinical Biomechanics 32 (2016) 64–71

69

Table 5 Vector Angle (degrees) mean and standard deviation for ACLR and healthy groups for phases of gait with significant differences. Joint couple

Phase of gait (%)

ACLR

Healthy

Effect size (90% CI)

Hip frontal-knee sagittal Hip frontal-knee transverse

31%–35% 8%–10% 36%–39% 48%–63% 85%–88% 84%–88%

12.24 (0.48) 28.42 (1.25) 40.18 (1.48) 43.56 (7.25) 24.44 (6.90) 45.26 (1.04)

5.58 (0.19) 16.01 (0.15) 23.86 (2.50) 24.44 (6.90) 8.95 (2.37) 30.73 (4.30)

18.0 (14.8, 21.2) 13.9 (11.4, 16.4) 8.0 (6.5, 9.4) 2.7 (2.0, 3.4) 2.2 (1.8, 2.8) 4.6 (3.7, 5.6)

Hip sagittal-knee transverse CI = confidence interval.

transverse–knee frontal joint couples were found during mid-stance phase (20%–40% of the gait cycle). While different methods for determining significant differences were used between the two studies, the findings support an ACLR individual's inability to find an optimal movement strategy during the weight bearing transition phases of various tasks. An increase in the movement variability of a system implies a decrease in sensorimotor control (Davids et al., 2003) but does not necessarily indicate a lack of joint stability. High variability may represent the body's inability to find an optimal movement pattern within the constraints presented due to a lack of coordination and motor control. Additionally, it is possible to constrain motion while still having inconsistent movement patterns, as demonstrated by the hip frontal–knee transverse joint couple during terminal stance. During jogging, ACLR variability resembled that of the healthy group. As the number of constraints increases, the number of options for a dynamical system to complete a task decreases (Davids et al., 2003). Therefore, because task constraints are increased in jogging compared to walking, motion control could be more deterministic at jogging velocity (Herb et al., 2013), potentially causing the variability in the

ACLR group to be similar to that of the healthy group. As velocity increases, there is an overall increase in task constraints by increasing both muscle force production and ground reaction force. Additionally, the increase in velocity will decrease stance time, limiting the amount of time that the body has to explore various ranges of motion to accomplish a task. This potentially explains why almost all changes were seen during the longer swing phase. Patients have an increased risk of developing degenerative osteoarthritis after ACLR (Gillquist and Messner, 1999; Lohmander and Roos, 1994; Lohmander et al., 2004, 2007; Myklebust and Bahr, 2005; Øiestad et al., 2009). Cartilage growth and thickness is a response to consistent loading and movement patterns in healthy individuals (Andriacchi, 2013). A consistent movement pattern can be interpreted as there being low variability in motion. After injury, changes in movement patterns increase the risk for developing osteoarthritis and chronic loading of the cartilage in areas not designed to absorb forces may be a factor in mechanical damage to the joint (Andriacchi, 2013; Andriacchi et al., 2009). Decreasing the consistency of movement may shift force distribution away from areas where the cartilage is thicker

Fig. 3. Effect sizes for mean differences in vector angle between ACL reconstructed and healthy limbs during walking gait. Fig. 3A shows the specific location in the gait cycle when confidence intervals did not overlap with healthy controls for three or more consecutive points. Fig. 3B displays the effect size for significant differences between groups with 90% confidence intervals within each general gait phase (HFKS = hip frontal–knee sagittal, HFKT = hip frontal–knee transverse, HSKT = hip sagittal–knee transverse).

70

T.C. Gribbin et al. / Clinical Biomechanics 32 (2016) 64–71

and introduce higher stresses to unconditioned areas. The results in this study indicate that the ACLR limb experiences increased variability, which may imply a lack motor control during gait (Davids et al., 2003). We believe that this could alter loading patterns at the knee, potentially altering attenuation strategies in order to function within these new constraints, and could be a contributing factor to the development of osteoarthritis after ACLR. The only couples that exhibited decreased variability in the ACLR group were the two joint couples involving the hip sagittal motion during mid-stance. The decreases in magnitude during loading response were also in the two hip sagittal joint couples. The decreased variability at mid-stance could be in response to the constraint in the loading response. However, instead of shifting to a more proximal contribution to motion, these particular couples are constraining motion and making it more consistent to provide stability through the two stages of weight bearing in early and mid-stance. While the constrained hip and knee motions have been previously described in single planes (Chen et al., 2012; Ernst et al., 2000; Gardinier et al., 2012; Gokeler et al., 2013; Kuenze et al., 2014; Noehren et al., 2013), the addition of stride to stride variability and coupled motion measures offers a more detailed perspective of adaptations involving hip flexion in an attempt to improve knee stability during early stages of gait. These measures should be considered as a potential contributor to the normalization of gait mechanics. There are several key limitations in the current study that should be considered when reviewing the results of this investigation. As this was part of a larger exploratory investigation, we did not exclude participants based on the amount time between initial injury and ACLR or the time between ACLR and study participation. This may have resulted in increased within group variability for our outcome measures which may limit the generalizability of our findings to a specific time period following ACLR. In addition, we did not match participants to healthy controls based on factors such as sex or age. Previous researchers have shown that females display lower variability measures than males during a side-step cutting task (Pollard et al., 2005). Our healthy group had a higher number of females compared to males than the ACLR group, potentially biasing our healthy VCV measure to the low end. However, we found both increases and decreases in ACLR VCV compared to healthy controls with large effect sizes. Therefore, we believe that this limitation would overestimate the size of the difference between groups rather than the significance of the results. Future studies using these methods should match reconstructed and healthy groups based on limb and sex. We also did not control for individual differences in static alignment in the current analysis. While it is possible that frontal and transverse plane alignment of the lower extremity may have an impact on hip and knee motion, this was the first investigation to approach this coupling relationship in the ACLR population and it is unclear which factors may impact our results. Therefore, we chose to present our data without controlling for possible confounding factors. Lastly, while we excluded participants that were not physically active, there was a broad range of patient demographics including age and BMI that may have contributed to both our significant and non-significant findings. In general, future investigations should aim to better control for descriptive factors with the potential of developing sub-groups within the ACLR population in order to improve generalizability and clinical application of findings. 5. Conclusion In conclusion, ACLR individuals experienced decreased magnitudes of joint excursion, primarily in the stance phase of gait. Individuals with ACLR also demonstrated increased vector angles, indicating more distal segment contribution to motion relative to the healthy controls. These changes are representative of an aberrant movement pattern when compared to healthy individuals, which could be a coping mechanism that increases risk for future instability events and degenerative

osteoarthritis. The increased VCV implies the ACLR group was less consistent in motion from stride to stride, implying that the body has not found an optimal way to adapt to the constraints imposed by ACL injury and reconstruction. References Andriacchi, T.P., 2013. Valgus alignment and lateral compartment knee osteoarthritis: a biomechanical paradox or new insight into knee osteoarthritis? Arthritis Rheum. 65 (2), 310–313. Andriacchi, T.P., Dyrby, C.O., 2005. Interactions between kinematics and loading during walking for the normal and ACL deficient knee. Knee Mech. Updat. Theor. Exp. Anal. 38 (2), 293–298. Andriacchi, T.P., S. Koo, and S.F. Scanlan, Gait mechanics influence healthy cartilage morphology and osteoarthritis of the knee. J. Bone Joint Surg.(American volume), 2009. 91 Suppl. 1: p. 95–101. Ardern, C.L., et al., 2014. Fifty-five per cent return to competitive sport following anterior cruciate ligament reconstruction surgery: an updated systematic review and meta-analysis including aspects of physical functioning and contextual factors. Br. J. Sports Med. 48 (21), 1543–1552. Brophy, R.H., Wright, R.W., Matava, M.J., 2009. Cost analysis of converting from singlebundle to double-bundle anterior cruciate ligament reconstruction. Am. J. Sports Med. 37 (4), 683–687. Chen, C.H., et al., 2012. Anteroposterior stability of the knee during the stance phase of gait after anterior cruciate ligament deficiency. Gait Posture 35 (3), 467–471. Davids, K., et al., 2003. Movement systems as dynamical systems: the functional role of variability and its implications for sports medicine. Sports Med. 33 (4), 245–260. Di Stasi, S.L., et al., 2013. Gait patterns differ between ACL-Reconstructed athletes who pass return-to-sport criteria and those who fail. Am. J. Sports Med. 41 (6), 1310–1318. Drewes, L.K., et al., 2009a. Dorsiflexion deficit during jogging with chronic ankle instability. J. Sci. Med. Sport 12 (6), 685–687. Drewes, L.K., et al., 2009b. Altered ankle kinematics and shank-rear-foot coupling in those with chronic ankle instability. J. Sport Rehabil. 18 (3), 375–388. Ernst, G.P., et al., 2000. Lower extremity compensations following anterior cruciate ligament reconstruction. Phys. Ther. 80 (3), 251–260. Ferber, R., Davis, I.M., Williams Iii, D.S., 2005. Effect of foot orthotics on rearfoot and tibia joint coupling patterns and variability. J. Biomech. 38 (3), 477–483. Gardinier, E.S., et al., 2012. Gait and neuromuscular asymmetries after acute anterior cruciate ligament rupture. Med. Sci. Sports Exerc. 44 (8), 1490–1496. Georgoulis, A.D., et al., 2010. ACL injury and reconstruction: clinical related in vivo biomechanics. Orthop. Traumatol. Surg. Res. 96 (8 Suppl), S119–S128. Gianotti, S.M., et al., 2009. Incidence of anterior cruciate ligament injury and other knee ligament injuries: a national population-based study. J. Sci. Med. Sport Sports Med. Aust. 12 (6), 622–627. Gillquist, J., Messner, K., 1999. Anterior cruciate ligament reconstruction and the longterm incidence of gonarthrosis. Sports Med. 27 (3), 143–156. Gokeler, A., et al., 2013. Return of normal gait as an outcome measurement in ACL reconstructed patients. A systematic review. Int. J. Sports Phys. Ther. 8 (4), 441–451. Hall, M., Stevermer, C.A., Gillette, J.C., 2012. Gait analysis post anterior cruciate ligament reconstruction: Knee osteoarthritis perspective. Gait Posture 36 (1), 56–60. Hamill, J., Haddad, J., McDermott, W., 2000. Issues in quantifying variability from a dynamical systems perspective. J. Appl. Biomech. 16, 407–418. Heiderscheit, B.C., Hamill, J., van Emmerik, E.A., 2002. Variability of stride characteristics and joint coordination among individuals with unilateral patellofemoral pain. J. Appl. Biomech. 18 (2), 110–121. Herb, C.C., et al., 2013. Shank-rearfoot joint coupling with chronic ankle instability. J. Appl. Biomech. Hopkins, W.G., et al., 2009. Progressive statistics for studies in sports medicine and exercise science. Med. Sci. Sports Exerc. 41 (1), 3–13. Hurd, W.J., Snyder-Mackler, L., 2007. Knee instability after acute ACL rupture affects movement patterns during the mid-stance phase of gait. J. Orthop. Res. Off. Publ. Orthop. Res. Soc. 25 (10), 1369–1377. Ingersoll, C.D., et al., 2008a. Neuromuscular consequences of anterior cruciate ligament injury. Clin. Sports Med. 27 (3), 383–404 vii. Ingersoll, C.D., et al., 2008b. Neuromuscular consequences of anterior cruciate ligament injury. Clin. Sports Med. 27 (3), 383–404. Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., 1990. Measurement of lower extremity kinematics during level walking. J. Orthop. Res. Off. Publ. Orthop. Res. Soc. 8 (3), 383–392. Kiefer, A.W., et al., 2013. Inter-segmental postural coordination measures differentiate athletes with ACL reconstruction from uninjured athletes. Gait Posture 37 (2), 149–153. Knoll, Z., Kocsis, L., Kiss, R.M., 2004. Gait patterns before and after anterior cruciate ligament reconstruction. Knee Surg. Sports Traumatol. Arthrosc. 12 (1), 7–14. Kuenze, C., et al., 2014. Jogging biomechanics after exercise in individuals with ACL-reconstructed knees. Med. Sci. Sports Exerc. 46 (6), 1067–1076. Latash, M.L., Turvey, M.T., 1996. Dexterity and Its Development. L. Erlbaum Associates. Lohmander, L.S., Ostenberg, A., Englund, M., Roos, H., 2004. Hight prevalence of knee osteoarthritis, pain and functional limitations in female soccer players twelve years after anterior cruciate ligament injury. Arthritis Rheum. 50 (10), 3145–3152. Lohmander, L.S., Roos, H., 1994. Knee ligament injury, surgery, and osteoarthritis. Truth or consequences? Acta Orthop. Scand. 65 (6), 605–609.

T.C. Gribbin et al. / Clinical Biomechanics 32 (2016) 64–71 Lohmander, L.S., Englund, P.M., Dahl, L.L., Roos, E.M., 2007. The long-term consequence of anterior cruciate ligament and meniscus injuries: osteoarthritis. Am. J. Sports Med. 35 (10), 1756–1769. McKeon, P.O., Hertel, J., 2006. The dynamical-systems approach to studying athletic injury. Athl. Ther. Today 11 (1), 31–33. McKeon, P.O., et al., 2009. Effects of balance training on gait parameters in patients with chronic ankle instability: a randomized controlled trial. Clin. Rehabil. 23 (7), 609–621. Miller, R.H., et al., 2010. Variability in kinematic coupling assessed by vector coding and continuous relative phase. J. Biomech. 43 (13), 2554–2560. Mullineaux, D.R., Uhl, T.L., 2010. Coordination-variability and kinematics of misses versus swishes of basketball free throws. J. Sports Sci. 28 (9), 1017–1024. Myklebust, G., Bahr, R., 2005. Return to play guidelines after anterior cruciate ligament surgery. Br. J. Sports Med. 39 (3), 127–131. Noehren, B., et al., 2013. Long-term gait deviations in anterior cruciate ligamentreconstructed females. Med. Sci. Sports Exerc. 45 (7), 1340–1347. Novacheck, T.F., 1998. The biomechanics of running. Gait Posture 7 (1), 77–95. Øiestad, B.E., Engebretsen, L., Storheim, K., Risberg, M.A., 2009. Knee osteoarthritis after anterior cruciate injury: a systematic review. Am. J. Sports Med. 37 (7), 1434–1443. Osternig, L.R., et al., 2000. Human hip and knee torque accommodations to anterior cruciate ligament dysfunction. Eur. J. Appl. Physiol. 83 (1), 71–76. Ounpuu, S., 1994. The biomechanics of walking and running. Clin. Sports Med. 13 (4), 843–863. Pollard, C.D., et al., 2005. Gender differences in lower extremity coupling variability during an unanticipated cutting maneuver. J. Appl. Biomech. 21 (2), 143–152.

71

Pollard, C.D., et al., 2015. Altered lower extremity movement variability in female soccer players during side-step cutting after anterior cruciate ligament reconstruction. Am. J. Sports Med. 43 (2), 460–465. Riley, M.A., Turvey, M.T., 2002. Variability of determinism in motor behavior. J. Mot. Behav. 34 (2), 99–125. Sadeghi, H., et al., 2000. Symmetry and limb dominance in able-bodied gait: a review. Gait Posture 12 (1), 34–45. Scanlan, S.F., Chaudhari, A.M., Dyrby, C.O., Andriacchi, T.P., 2010. Differences in tibial rotation during walking in ACL reconstructed and healthy contralateral knees. J. Biomech. 43 (9), 1817–1822. Sparrow, W.A., et al., 1987. Using relative motion plots to measure changes in intra-limb and inter-limb coordination. J. Mot. Behav. 19 (1), 115–129. Stergiou, N., Harbourne, R., Cavanaugh, J., 2006. Optimal movement variability: a new theoretical perspective for neurologic physical therapy. J. Neurol. Phys. Ther. 30 (3), 120–129. Tagesson, S., Oberg, B., Kvist, J., 2014. Static and dynamic tibial translation before, 5 weeks after, and 5 years after anterior cruciate ligament reconstruction. Knee Surg. Sports Traumatol. Arthrosc. Off. J. ESSKA. Tepavac, D., Field-Fote, E., 2001. Vector coding: a technique for quantification of intersegmental coupling in multicyclic behaviors. J. Appl. Biomech. 17 (3), 259–270. van Emmerik, R.E., van Wegen, E.E., 2002. On the functional aspects of variability in postural control. Exerc. Sport Sci. Rev. 30 (4), 177–183.