Knee joint movements in subjects without knee pathology and subjects with injured anterior cruciate ligaments. (Research Report).

Relative amounts of joint surface rolling and gliding may be inferred when there is movement at the knee through the application of the concept of the path of instantaneous centers of rotation (PICR). (1-4) For example, when the PICR is located close to the point of contact between joint surfaces, more rolling than gliding occurs. (1,2) Conversely, when the PICR is located far from the point of contact, less rolling than gliding occurs. Based on their visual observations, Frankel et al (1) and Gerber and Matter (3) reported that PICR patterns in human knees with injured anterior cruciate ligaments (ACLs) differed from the patterns in knees without pathology. Similarly, researchers investigating PICR in canine stifles with injured ACLs reported observable deviations from PICR patterns in stifles without pathology. (4) In these studies, however, the observed differences in PICR patterns were not quantified. Nevertheless, it is reasonable to hypothesize that changes in intrinsic joint surface rolling and gliding movements, as suggested by altered PICR patterns, may occur in knees with injured ACLs.

Frankel et al, (1) Gerber and Matter, (3) and Mitton et al (4) did not examine PICR or measure the amount of joint surface rolling and gliding under weight-bearing (WB) conditions. Tibiofemoral joint compression forces are greater during WB knee extension than during non-weight-bearing (NWB) knee extension. During WB knee extension, the knee joint is loaded through body weight as well as through muscle activity. This is more loading than occurs during NWB knee extension, when the leg is unobstructed as it moves through space and the knee is loaded primarily through muscle activity. (5,6) As joint compression increases, joint surface geometry regulates relative joint surface rolling and gliding. (2,6) As joint compression decreases, the articular ligaments become more involved in regulating joint surface rolling and gliding. (2,6) Consequently, several researchers have shown that the knee cruciate ligaments are loaded in the opposite manner from each other during movement and are loaded differently when in WB and NWB. (7-11) Anterior cruciate ligament strain is greater than posterior cruciate ligament (PCL) strain during NWB knee extension, whereas PCL strain is greater than ACL strain during WB knee extension. Presumably, this is because greater anterior displacement of the tibia with respect to the femur is believed to occur during NWB. (9,10)

Collectively, the results of these studies (7-11) suggest that knee joint surface rolling and gliding movements, and therefore PICR, may differ between WB and NWB movements. Hollman et al, (12) using PICR to estimate knee joint surface rolling and gliding in subjects without knee joint pathology, reported that greater joint surface gliding occurs during NWB knee extension. These data, however, have not been published in a peer-reviewed publication. Given that the ACL provides the primary restraint to anterior tibial displacement, (13) we hypothesized that joint surface gliding may be exaggerated in knees with injured ACLs during WB and NWB.

Electromyography (EMG) can provide some insight into variations in muscle activity. In EMG studies, increased muscle activity about a joint is associated with increased joint compression. (9) We believe, therefore, that it is reasonable to expect that altered muscle activity may also influence the relative amount of rolling and gliding and the PICR. Evidence suggests the hamstring muscles contribute to knee joint stability. (9,10) In the absence of the structural integrity of the ACL, hamstring muscle co-contraction is thought to provide stability during knee extension either by virtue of its posteriorly directed force, its contribution to increased joint compression, or a combination of both mechanisms. (5,9,14-16)

Data from several studies indicate that people with an injured ACL or reconstructed ACL demonstrate greater hamstring muscle EMG activity or altered hamstring muscle timing during functional activities than do subjects without knee pathology. (17-21) In our opinion, this suggests that a change in muscle coordination may be necessary to improve stability in knees with injured ACLs. We contend that an increase in hamstring muscle activity may be sufficient to counteract the increased gliding that might otherwise be present. Electromyographic analyses that accompany the measurement of joint surface rolling and gliding movements, in our opinion, can determine whether differences in muscle activity are associated with changes in joint surface rolling and gliding.

The primary purpose of our study was to compare joint surface rolling and gliding between subjects without knee pathology and subjects with injured ACLs during NWB and WB movements. A secondary purpose was to determine whether EMG activity of selected lower-extremity muscles, particularly the hamstring muscles, differed between these groups. We hypothesized that greater joint surface gliding would occur in knees with injured ACLs than in knees without pathology, assuming that an increase in hamstring muscle activity would not sufficiently counteract the hypothesized change in rolling and gliding movements. If a change in rolling or gliding did not occur, we hypothesized that greater hamstring muscle activity would be present in subjects with injured ACLs.

Methods

Subjects

Fifteen adult subjects with injured ACLs (9 men and 6 women; mean age=26 years, SD=7, range=18-36; mean height=173 cm, SD=8, range=160-185; mean body mass=81 kg, SD=20, range=54-113) and 15 age-and sex-matched subjects without knee pathology (9 men and 6 women; mean age=25 years, SD=6, range=18-36; mean height=172 cm, SD=11, range=155-193; mean body mass=69 kg, SD=9, range=52-91) participated in this study. The number of subjects satisfied a statistical power (1 - [beta]) greater than 0.80, based on an analysis of joint surface rolling among initial subgroups of 7 subjects with injured ACLs and 11 subjects without knee pathology. The effect size was calculated to be 0.29 using procedures described by Cohen and Cohen. (22) We determined the effect size in order to make our findings more likely to be statistically significant. We did not determine potential effects based on any claim of clinical or biomechanical meaningfulness. Subjects with injured ACLs were recruited from: (1) the Department of Orthopaedic Surgery, Washington University School of Medicine, St Louis, Mo, (2) the undergraduate and graduate student populations of Washington University, St Louis, Mo, and (3) the Athletics Department, Clarke College, Dubuque, Iowa (Tab. 1).

Data from subjects with injured ACLs were included only if an orthopedic surgeon diagnosed the ACL injury (confirmed by magnetic resonance imaging or by surgery following data collection for this study), either in isolation or with an associated injury to the medial collateral ligament (MCL), lateral collateral ligament (LCL), or meniscus. The MCL, LCL, and meniscus are secondary anterior stabilizers of the knee joint and provide stabilization in the absence of an intact ACL. (13) Combined, these structures contribute only up to 14% of the total resistance to anterior tibial displacement (13); therefore, we included subjects with injuries to these structures in the study. Additional requirements were that each subject have, at minimum, active range of motion of the knee of [greater than or equal to] 10 degrees to 90 degrees of knee flexion (relative to a 0[degrees]-180[degrees] notation system (23)) and an active straight leg raise of 60 degrees or greater in a supine position. This position was chosen in an effort to reduce potential effects of short hamstring muscles on knee joint movements. All subjects with injured ACLs had at least 120 degrees of knee flexion, although 3 subjects with acute injuries had mild knee joint effusion at the time of testing and had deficits in full extension of up to 5 degrees. We made a judgment that those with mild effusion, assuming they could move through the knee range of motion required of the study, would not have substantially adverse effects on the results and were therefore included in the study. Potential subjects with injured ACLs were excluded if they had any history of neurological or neuromuscular pathology.

Comparison subjects without knee pathology were recruited from a population of convenience--the undergraduate and graduate student populations at Washington University in St Louis. Potential subjects were excluded from the study if they had any history of knee injury or neurological or neuromuscular pathology. Comparison subjects were allowed to take part only if they satisfied the following criteria: (1) active range of motion of the knee of 0 to 130 degrees of flexion or greater, (2) an active straight leg raise in a supine position of 60 degrees or greater, (3) grade 5 force production capability of the right quadriceps femoris and hamstring muscles based on manual muscle testing techniques, (24) and (4) negative findings on the Lachman's test, posterior drawer test, valgus and varus ligament stability tests, and McMurray's meniscus test. (25) We did not examine the reliability of these measurements.

Instrumentation

Movements were recorded with 2 Panasonic AG-455P SVHS video cameras. * Kinematic data were processed with Ariel Performance Analysis System (APAS) software, APAS99 version 3.5. ([dagger]) The APAS software provides accurate measurements of linear (mean error less than 2 mm) and angular (mean error less than 0.3[degrees]) standards (26) and reliable measurements of angular velocity data, with intraclass correlation coefficients exceeding .85 at various speeds of movement. (27) Coordinate transformations and instantaneous center of rotation (ICR) calculations are described elsewhere (28) and were performed using custom-written programs in Microsoft Excel 97, version SR-1. ([double dagger])

Electromyographic signals were collected with bipolar surface electrodes having an on-site pre-amplification gain of 310. ([section]) Each active electrode pair consisted of two 8-mm-diameter silver-silver chloride electrodes with an interelectrode distance of 35 mm. The on-site preamplifiers had common mode rejection ratio ratings of approximately 105 dB at 60 Hz with high direct current input impedance of 100,000 M[OMEGA] and bandwidth frequencies of 8 Hz to 31 kHz. The EMG signals were processed with APAS ([dagger]) software, and normalization procedures were performed with custom-written programs in Microsoft Excel. ([double dagger])

A Bertec force platform ([parallel]) was used as a video synchronization and EMG triggering device. Subjects made contact with the force platform while initiating their movement, which simultaneously turned on an external light source that was visible in the video fields of both video cameras and initiated EMG sampling. This triggering mechanism enabled videographic images and EMG data to be synchronized.

Procedure

All subjects signed an informed consent form approved by the Washington University School of Medicine Institutional Review Board. The primary investigator (JHH) conducted a brief physical examination to assess knee joint range of motion and assessed muscle force and knee stability in subjects without knee pathology. Active range of motion of the knee was measured with a universal goniometer using techniques described by Norkin and White, (23) and muscle force was measured manually using techniques described by Kendall et al. (24) Only the involved side was tested in subjects with injured ACLs. In subjects without knee pathology, the side tested was selected in a random manner to equalize the distribution of right and left knees between the groups. For 11 subjects without knee pathology, we randomly selected whether the right or left knee would be tested. After we determined that 15 subjects would be tested per group, the side tested in the group without pathology was assigned so that there would be an equal number of right and left knees tested in the study.

Surface EMG electrodes were taped over the mid-muscle belly of each subject's vastus lateralis, medial hamstring (ie, the semitendinosus), medial gastrocnemius, and gluteus maximus muscles. Electrodes were placed in a manner that we believe was in parallel with the line of action of the muscle. To facilitate signal conduction, each electrode site was rubbed with an alcohol wipe prior to electrode attachment. Electrode locations were found by following the description of Ericson et al, (29) and specific locations were identified by palpating the respective muscle bellies. Once electrodes were applied, EMG signals were sampled during maximal voluntary isometric contractions (MVICs) for each muscle group.

Photoreflective markers, which were 2 cm in diameter, were placed on the lateral thigh and lateral leg to provide rigid-body representation of the thigh and leg segments. The markers were placed on the thigh 10 cm distal to the greater trochanter and 5 cm proximal to the lateral epicondyle and on the leg 1 to 2 cm distal to the fibular head and 1 to 2 cm proximal to the lateral malleolus, which minimized skin movement artifact. (30) We did not examine the reliability of placing these markers.

Subjects sat on either a 35.6-cm (14-in) or 40.6-cm (16-in) wooden chair, depending on the subject's height and lower-extremity limb length that allowed the knee to be in approximately 100 degrees of flexion in its resting position. Subjects performed 5 repetitions each of the NWB and WB movements. To do the NWB movement, subjects extended their leg to a position of maximal knee extension. Subjects did the WB movement by executing a 2-legged sit-to-stand movement. Subjects initiated both testing conditions by exerting pressure on the force platform, which allowed us to synchronize the kinematic and EMG data. Because joint rotation at slow angular speeds (<30[degrees]/s) decreases PICR measurement accuracy, (31) we attempted to control the speed of movement by instructing subjects to perform their movements over a period of approximately 1 second.

Data Processing

Reflective marker spatial locations and instantaneous center of rotation (ICR) calculations. M1 spatial location data were recorded at 60 Hz, the sampling frequency of the video cameras. Images were captured, compressed, and stored on the computer hard drive with APAS software. ([dagger]) Subsequently, the video files were transformed via direct linear transformation. (32) Three-dimensional spatial locations of reflective markers defining the thigh and leg were autodigitized and smoothed between 3 and 6 Hz using a second-order Butterworth digital filter. Cutoff frequencies for each marker were selected by using spectral power analyses in the "Filter" module of APAS software. ([dagger]) Spatial location data were transformed to a local reference system fixed in the thigh segment with its origin located at the lateral femoral epicondyle. The PICR data were calculated in a manner originally described by Winter (31) and detailed by Hollman and Deusinger (28) and Loudon. (33) The joint's estimated instantaneous angular velocity ([omega]) and the leg segment's estimated instantaneous tangential linear velocity (V) in the sagittal plane of motion were used to calculate locations of the ICR by identifying the distance (R) of the ICR from the lateral malleolus marker by

V = [omega] x R.

In Cartesian coordinates represented by a right-handed coordinate system, the location of the ICR is calculated as the distance ([R.sub.x], [R.sub.y]) from the lateral malleolus as follows:

- [R.sub.y] = [V.sub.x]/[omega]

and

[R.sub.x] = [V.sub.y]/[omega].

The accuracy and precision of ICR measurement that we used is based on data obtained with a model from which the ICR was a stationary and known entity. Calculations yielded a mean error magnitude of less than 1 mm and a standard deviation of 4.3 mm. (28) The reliability coefficient for the method when applied to human knee joint PICR measurement exceeds .80. (33) The ICR coordinate locations calculated for the first 3 digitizable trials from each movement condition were used for obtaining mean condition-dependent PICR patterns for each subject.

Intrinsic knee joint surface movements model. Coordinates of the ICR were obtained for each participant through the range of motion and extracted at 10-degree intervals from 90 degrees to 10 degrees of knee flexion. Knee joint surface movements were examined on 2 levels. First, a mean PICR for each movement condition was plotted and analyzed visually to examine observable differences between the subject groups and between the WB and NWB movements. Second, participants' ICR coordinate data were applied to a planar mathematical knee joint model (Fig. 1) that was used to calculate relative proportions of knee joint surface rolling and gliding.

[FIGURE 1 OMITTED]

Dimensions of the model were adapted from morphology studies of the distal femoral articular surface (34-36) to formulate a geometrical configuration of the femur in the sagittal plane. Three geometrical components constituted the model. One component was a circle having a 2-cm radius that represents the posterior condylar curvature. A second circle with a 4-cm radius represented the distal-anterior condylar surface curvature. The third geometrical component of the model was an ellipse fitted between the 2-cm and 4-cm curves that created a smooth transition between the 2 circles to represent the continuous condylar surface of the femur. The proximal tibia was modeled as a linear straight-line surface. Coordinates for the femoral surface contact points were calculated at 10-degree intervals along the elliptical surface of the model.

Based on the selected femoral contact point locations and the ICR data obtained experimentally, the relative proportion of rolling to gliding was calculated using the slip ratio described by O'Connor and Zavatsky. (37) The slip ratio is defined by

slip ratio = [s.sub.m]/[s.sub.f]

where [s.sub.m] is the displacement between successive contact points on the convex femoral surface and [s.sub.f] is the displacement between successive contact points on the flat tibial surface. The variables [s.sub.m] and [s.sub.f] are approximated by

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

and

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where R is the radius of curvature of the model, [r.sub.icr] is the radius from an ICR location to its respective femoral contact point, and [theta] is the fixed rotation angle (10[degrees]) between successive contact points. Ultimately, percent rolling (% rolling), defined by

% rolling = [[s.sub.m]/[s.sub.m] + ([s.sub.m] - [s.sub.f]) x 100

and percent gliding (% gliding), defined by

% gliding = 100 - % rolling

were the potential outcome measures of rolling and gliding movements. We used % gliding for statistical procedures in this study. There are no reliability data published in peer-reviewed journals for these measurements. With unpublished data obtained in our laboratory from samples consisting of 11 subjects without knee pathology and 7 subjects with injured ACLs, the within-session reliability coefficients (specifically, the relative generalizability coefficients (38)) for calculating % rolling in knees without pathology is .83 and in knees with injured ACLs is .92.

EMG. The EMG data were sampled at 1,000 Hz. We used this sampling rate to satisfy the Nyquist theorem, which states that the sampling rate must be greater than or equal to 2 times the highest frequency component in the analog signal. Therefore, we were able to sample muscle fiber firing frequencies up to 500 Hz. All raw signals were full-wave rectified and subsequently processed through a linear envelope with a 50-millisecond floating window. The processed EMG data obtained from the test conditions then were normalized by dividing the magnitude of a muscle's EMG activity by its peak MVIC magnitude and multiplying by 100, and the normalized EMG values were expressed as a percentage of MVIC (% MVIC). The normalized EMG values, which were synchronized with knee angles, were extracted at respective 10-degree knee angle positions and used for data analysis. Because we consider lower-extremity EMG data to be highly reliable within testing sessions for movements similar to those used in this study. (39) data from one repetition in each movement condition was processed for each subject.

Data Analysis

Movements. A 3-way mixed model analysis of variance (ANOVA) with [alpha]=.05 was conducted to test the null hypothesis that no difference occurred in rolling and gliding movements between subjects without knee pathology and subjects with injured ACLs. The dependent variable was % gliding. The between-subjects factor was group (subjects without knee pathology and subjects with injured ACLs). The within-subjects factors were movement (NWB and WB) and knee angle (90[degrees], 80[degrees], ..., 10[degrees]). We tested the main effects of group and movement and the group interactions with movement and knee angle to determine in which condition or conditions and at which knee angle or angles rolling and gliding movements differed between groups. If either a group x movement or group x movement x knee angle interaction existed, we then conducted a mixed-model ANOVA on each movement to determine whether differences occurred between groups within a movement. If a group x knee angle interaction existed, we conducted post hoc t tests with a Bonferroni-adjusted alpha to determine the knee angle or angles at which % gliding differed. The main effect of movement was tested to determine whether joint surface rolling and gliding differed between the WB and NWB movements across groups.

EMG. A 4-way mixed-model ANOVA with the same between-subject factor (group) and 3 within-subjects factors--including movement, knee angle, and muscle (vastus lateralis, medial hamstring, medial gastrocnemius, gluteus maximus)--was conducted to test the null hypothesis that' no difference existed in % MVIC between groups ([alpha]=.05). We chose this analysis to determine whether differences in EMG activity occurred between subjects with injured ACLs and subjects without knee pathology. If the group main effect or interactions involving the group factor were significant, then a mixed-model ANOVA was performed on each muscle to determine which muscle or muscles differed between groups.

Results

Movements

The PICR patterns and % gliding data for both groups are presented in Figures 2 and 3, respectively. Within each movement condition, the PICR for subjects with injured ACLs is located further from the joint surface than for subjects without knee pathology (Fig. 2), suggesting that greater joint surface gliding occurs in knees with injured ACLs. This is most evident in the last 20 degrees of knee extension in the NWB movement.

[FIGURES 2-3 OMITTED]

The statistical analysis of % gliding revealed that overall, there was no group main effect (F=2.44, df=1,28, P=.13, 1-[beta]=0.36), but there were interactions in each effect containing the group factor (group x movement: F=9.95, df=1,28, P=.004; group x knee angle: F=2.51, df=8,224, P=.013; and group x movement x knee angle: F=3.01, df=8,224, P=.003). In the NWB movement, there was no group main effect (% gliding: subjects without knee pathology=43.3% [+ or -] 1.8% [mean [+ or -] SEM], subjects with injured ACLs=45.0% [+ or -] 2.2%; F=0.27, df=1,8, P=.608). There was, however, a group x knee angle interaction (F=5.55, df=8,224, P<.001) in the NWB movement. Post hoc t tests revealed that, at 10 degrees of knee flexion, greater joint surface gliding occurred in knees with injured ACLs than in knees without pathology (% gliding=51.1% [+ or -] 1.3% and 43.0% [+ or -] 1.7%, respectively; [t.sub.28]=-3.44, P=.004). In the WB movement, there was a group main effect (% gliding: subjects without knee pathology=40.8% [+ or -] 2.3%, subjects with injured ACLs=44.8% [+ or -] 2.4%; F=5.51, df=1,8, P=.026), indicating that greater joint surface gliding occurred in knees with injured ACLs than in knees without pathology throughout the knee range of motion tested. In summary, joint surface gliding was greater in knees with injured ACLs than in knees without pathology at full knee extension (10[degrees] of knee flexion) in the NWB movement and throughout the range of motion (10[degrees]-90[degrees]) in the WB movement.

There was a main effect for movement (F=9.67, df=1,28, P=.004). More joint surface gliding occurred in the NWB movement than in the WB movement (% gliding: WB=42.8% [+ or -] 1.7%, NWB=44.2% [+ or -] 1.4%) across both groups. A movement x knee angle interaction (F=12.11, df=8,224, P<.001) revealed that the difference in % gliding between movements was particularly apparent at full knee extension (% gliding: WB=33.7% [+ or -] 1.3%, NWB=47.1% [+ or -] 2.6%; [t.sub.29]=5.44, P=.001). More joint surface gliding occurred in the NWB movement than in the WB movement among subjects in both groups, particularly at full knee extension.

EMG

The EMG data are illustrated in Figure 4. Overall, EMG activity was greater in the WB movement than in the NWB movement (F=81.56, df=1,28, P=.001). Electromyographic activity also differed among muscles (F=39.10, df=4,112, P=.001) and across knee angles (F=7.47, df=8,224, P=.001). There was no group main effect, however, nor were there any interactions involving the group factor (Tab. 2). Because EMG activity did not differ between subjects without knee pathology and subjects with injured ACLs (F=2.00, df=1,28, P=.169, 1-[beta]=0.36), no additional analysis of EMG data was conducted.

[FIGURE 4 OMITTED]

Discussion

Knee Joint Surface Movements

We found that PICR patterns during NWB knee extension (Fig. 2A) in subjects with injured ACLs moved farther from the joint surface as the knee approached full extension. We found that % gliding was greater than 50% (Fig. 3A) at the 10-degree knee flexion angle, indicating that, at full knee extension, more gliding than rolling occurs in knees with injured ACLs than in knees without pathology. These results disagree with those of Gerber and Matter, (3) who reported that the PICR shifts anteriorly and inferiorly (ie, toward the joint surface) within the femoral condyle during NWB, particularly in the range of 20 to 40 degrees of knee flexion. Our findings support the conclusion that the PICR in knees with injured ACLs moves away from the joint surface rather than toward the joint surface and that increased joint surface gliding occurs (Figs. 2A and 3A). Because the knee is extending, the increased gliding motion of the tibia in an anterior direction means, in our opinion, that greater anterior tibial displacement occurs in knees with injured ACLs than in knees without pathology during full knee extension in NWB. Our results agree with those of Vergis et al (40) and Lysholm and Messner, (41) who reported that more anterior tibial displacement occurs in knees with injured ACLs during both passive (40) and active (40,41) knee extension conditions.

A second finding of our study is that, although less joint surface gliding occurred in the WB movement than in the NWB movement for both groups throughout the WB motion, greater joint surface gliding did occur in knees with injured ACLs than in knees without pathology (Fig. 3B). Researchers (9,42,43) have found greater increases in anterior tibial displacement, anterior shearing forces, and ACL strain during knee extension in NWB than during knee extension in WB. For example, Jenkins et al (44) demonstrated with knee arthrometer measurements that isometric quadriceps femoris muscle contractions produced greater anterior tibial displacement during a NWB movement than during a WB movement. Jenkins et al reported intraclass correlation coefficients ranging from .74 in the WB condition to .96 in the NWB condition. We did not examine reliability as part of our study. Results of such studies have led some clinicians to use WB exercise for many lower extremity disorders, particularly for the rehabilitation of patients with injured or reconstructed ACLs. (45,46) Our findings, however, suggest that increased joint surface gliding may still be prevalent in knees with injured ACLs during WB movements.

EMG

Several researchers (17-21) have indicated that muscle activity during the stance phase of the gait cycle differs between individuals without knee pathology and individuals with injured ACLs. People with injured ACLs are thought to produce more or earlier hamstring muscle activity in the gait cycle than people without knee pathology. Therefore, we hypothesized that subjects with injured ACLs may have exhibited increased hamstring muscle activity, particularly during WB movement. We found no differences in muscle activity, however, between subjects without knee pathology and subjects with injured ACLs. The lack of increased hamstring muscle activity may partially account for the increase in joint surface gliding seen in the knees with injured ACLs, in particular throughout the WB movement. Renstrom et al (16) and Solomonow et al (47) reported that hamstring muscle activity reduces anterior tibial displacement. Our data suggest that, in the absence of increased hamstring muscle activity in individuals with injured ACLs, anterior tibial displacement may not be reduced to a level equivalent to the displacement that occurs in knees without pathology.

The power analysis for our study was based on rolling and gliding data only. Muscle activity effects may have been present, but they were not detected because of low statistical power of the EMG portion of our study (1-[beta]=0.36). The effect size of differences in EMG magnitude between groups in our study was equal to 0.27, which would have required 60 subjects to obtain a statistical power of 0.80.

Clinical Relevance

Measurement of PICR may be used to examine joint surface rolling and gliding movements and may provide a basis for quantifying knee joint dysfunction or instability. Additional research incorporating PICR measurements and joint surface rolling and gliding movements may be useful. For example, some people with injured ACLs are able to perform functional tasks, including athletic activities, without the knee instability or pain often associated with ACL injury. (48) Factors that differentiate people who are able to cope with their injured ACL from those who are not able to cope have not been fully elucidated. (48,49) An ability to quantify rolling and gliding movements may provide insight into why some people function better with damaged ACLs than others.

Limitations

The methods and model used in this study were based on several assumptions that may differ to varying degrees among people. We believe, therefore, that the application of PICR is probably not yet appropriate for clinical decision making. The method for measuring PICR and our model for calculating % gliding were based on the assumption that the knee joint is primarily a joint with a single degree of freedom and a joint that rotates about a flexion/extension axis (represented 2-dimensionally by the PICR) and translates in the sagittal plane. The knee is actually a 3-dimensional joint that incorporates secondary rotations in the frontal (abduction/adduction) and transverse (axial rotation) planes of motion. The assumption that knee movements can be represented by general plane motion discounts potential effects of axial rotation (the "screw home mechanism") on the calculation of rolling and gliding. Changes in the screw home mechanism may account for the differences in rolling and gliding movements we observed. The evidence of whether differences occur in the screw home mechanism during NWB and WB movements, however, is conflicting. (50,51)

Our mathematical knee model for estimating joint surface rolling and gliding may also be limited because the geometric components of the model represent average joint surface geometry obtained from a limited number of subjects in 3 morphological studies, (34-36) and we did not distinguish between the medial and lateral femoral condyles. The size and radii of curvature of the bony segments and joint surfaces comprising the knee joint vary among people. (34-36) Nevertheless, several researchers (52-54) have found that the radii of the posterior portions of the medial and lateral condyles that articulate with the tibia are not meaningfully different. Because movement and group interaction effects were present in our study, we believe our model was appropriate for addressing our goal of comparing average rolling and gliding movements between groups.

Some of our other measurements also contribute to limitations. Although skin markers are assumed to represent rigid body motion of underlying skeletal segments, it is well known that a violation of the rigid body assumption occurs with the use of skin markers in motion analysis. (30,55,56) Less error, however, is typically introduced for sagittal-plane measurements at the knee than either frontal- or transverse-plane measurements, and marker placement methods do exist that minimize skin or soft tissue movement error. (30) We attempted to minimize soft tissue movement error by modifying typical marker positions as recommended by Cappozzo et al. (30) Furthermore, previous research (33) has demonstrated high correlations (Pearson product moment coefficients [r] ranging from .78 to .88) in cadaver knees between ICR locations determined through use of skin markers with videographic motion analysis and through the use of bone markers with fluoroscopy. We therefore believe our methods were sound.

Conclusion

Knee joint surface rolling and gliding movements differ between subjects with injured ACLs and subjects without knee pathology. More joint surface gliding occurs in knees with injured ACLs at full extension during NWB knee extension and throughout the range of motion during WB knee extension. No differences in EMG activity occurred between the 2 groups in the vastus lateralis, medial hamstring, medial gastrocnemius, and gluteus maximus muscles. Because the ACL accounts for up to 86% of resistance to anterior tibial displacement in the knee (13) and was the primary variable of interest that differed between groups, we conclude that the ACL has a role in regulating joint surface rolling and gliding movements. We believe that the absence of a structurally intact ACL results in the excess joint surface gliding associated with increased anterior tibial displacement during both NWB and WB knee extension. Our results also suggest that anterior tibial displacement may not be completely reduced as a function of WB movement. We believe further investigation into the efficacy of WB exercise and potential effects of WB movement on knees with injured ACLs is warranted.

Table 1.
Characteristics of Subjects With Injured Anterior Cruciate Ligaments

                                                     Month
Subject   Age                                        After
  No.     (y)   Sex      Knee    Pathology (a)       Injury

 1        21    Female   Right   ACL/meniscus           1
 2        18    Male     Right   ACL/MCL/meniscus       1.5
 3        18    Female   Right   Failed ACL repair      1.5
 4        18    Female   Left    ACL                   24
 5        36    Female   Right   ACL                    6
 6        31    Male     Left    ACL                   96
 7        33    Male     Left    ACL                    2
 8        32    Male     Left    ACL                  120
 9        24    Female   Left    ACL                    1.5
10        32    Male     Left    ACL                    1.5
11        34    Male     Left    ACL/meniscus          84
12        27    Female   Left    ACL                    1
13        34    Male     Left    ACL/LCL/meniscus       3
14        20    Male     Left    ACL                   10
15        19    Male     Right   ACL                    7

(a) ACL=anterior cruciate ligament, meniscus=medial meniscus,
MCL=medial collateral ligament, LCL=lateral collateral ligament.

Table 2.
Summary of Mixed-Model Analysis of Variance for Electromyographic Data

Source of Variance        df       SS         MS        F      P

Between Subjects
  Group                     1    2313.54    2313.54    2.00   .169
  Error                    28   32419.80    1157.85

Within Subjects
  Movement                  1   25327.46   25327.46   81.56   .001 (a)
  Group x movement          1      28.83      28.83    0.09   .763
  Error                    28    8694.84     310.53

  Muscle                    4   60248.32   15062.08   39.10   .001 (a)
  Group x muscle            4    1594.44     398.61    1.03   .393
  Error                   112   43142.40     385.20

  Knee Angle                8    1635.20     204.40    7.47   .001 (a)
  Group x knee angle        8     369.36      46.17    1.69   .103
  Error                   224    6130.88      27.37

  Movement x muscle         4   10116.32    2529.08   16.97   .001 (a)
  Group x movement x        4     564.40     141.10    0.95   .440
    muscle
  Error                   112   16692.48     149.04

  Movement x knee angle     8    8401.20    1050.15   37.06   .001 (a)
  Group x movement x
    knee angle              8     181.76      22.72    0.80   .601
  Error                   224    6348.16      28.34

  Muscle x knee angle      32    3152.32      98.51    4.88   .001 (a)
  Group x muscle x
    knee angle             32     611.20      19.10    0.95   .555
  Error                   896   18099.20      20.20

  Movement x muscle x
    knee angle             32    9841.92     307.56   13.94   .001 (a)
  Group x movement x
    muscle x knee angle    32     636.80      19.90    0.90   .625
  Error                   896   19765.76      22.06

(a) Indicates significance, [alpha]=.05.

* Panasonic Broadcast & Television Systems Co, Franklin Park, IL 60131.

([dagger]) Ariel Dynamics Inc, 6 Aliconte St, Trabuco Canyon, CA 92679.

([double dagger]) Microsoft Corp, One Microsoft Way, Redmond, WA 98052-6399.

([section]) Motion Control, Division of Iomed Inc, 1290 West 2320 South, Ste A, Salt Lake City, UT 84119.

([parallel]) Bertec Corp, 819 Loch Lomond Ln, Worthington, OH 43085.

References

(1) Frankel VH, Burstein AH, Brooks DB. Biomechanics of internal derangement of the knee: pathomechanics as determined by analysis of the instant centers of rotation. J Bone Joint Surg Am. 1971;53:945-962.

(2) Zatsiorsky VM. Kinematics of Human Motion. Champaign, Ill: Human Kinetics; 1998.

(3) Gerber C, Matter P. Biomechanical analysis of the knee after rupture of the anterior cruciate ligament and it primary repair: an instant-centre analysis of function. J Bone Joint Surg Br. 1983;65:391-399.

(4) Mitton GR, Ireland WP, Runyon CL. Evaluation of the instantaneous centers of rotation of the stifle before and after repair of torn cruciate ligament by use of the over-the-top technique in dogs. Am J Vet Res. 1991;52:1731-1737.

(5) Hsieh HH, Walker PS. Stabilizing mechanisms of the loaded and unloaded knee joint. J Bone Joint Surg Am. 1976;58:87-93.

(6) Zheng N, Fleisig GS, Escamilla RF, Barrantine SW. An analytical model of the knee for estimation of internal forces during exercise. J Biomech. 1998;31:963-967.

(7) Escamilla RF, Fleisig GS, Zheng N, et al. Biomechanics of the knee during closed kinetic chain and open kinetic chain exercises. Med Sci Sports Exerc. 1998;30:556-569.

(8) Grood ES, Suntay WJ, Noyes FR, Butler DL. Biomechanics of the knee-extension exercise. Effect of cutting the anterior cruciate ligament. J Bone Joint Surg Am. 1984;66:725-734.

(9) Lutz GE, Palmitier RA, An KN, Chao EYS. Comparison of tibiofemoral joint forces during open-kinetic-chain and closed-kinetic-chain exercises. J Bone Joint Surg Am. 1993;75:732-739.

(10) Stuart MJ, Meglan DA, Lutz GE, et al. Comparison of intersegmental tibiofemoral joint forces and muscle activity during various closed kinetic chain exercises. Am J Sports Med. 1996;24:792-799.

(11) Wilk KE, Escamilla RF, Fleisig GS, et al. A comparison of tibiofemoral joint forces and electromyographic activity during open and closed kinetic chain exercises. Am J Sports Med. 1996;24:518-527.

(12) Hollman JH, Deusinger RH, Zou D, Matava MJ. Estimation of knee joint surface rolling/gliding kinematics via instant center of rotation measurement. In: Proceedings of the American Society of Biomechanics Annual Meeting. Chicago, Ill: American Society of Biomechanics; July 19-22, 2000.

(13) Buffer DL, Noyes FR, Grood ES. Ligamentous restraints to anterior-posterior drawer in the human knee. J Bone Joint Surg Am. 1980;62: 259-270.

(14) Imran A, O'Connor JJ. Control of knee stability after ACL injury or repair: interaction between hamstrings contraction and tibial translation. Clin Biomech (Bristol, Avon). 1998;13:153-162.

(15) Pandy MG, Shelburne KB. Dependence of cruciate-ligament loading on muscle forces and external load. J Biomech. 1997;30:1015-1024.

(16) Renstrom P, Arms SW, Stanwyck TS, et al. Strain within the anterior cruciate ligament during hamstring and quadriceps activity. Am J Sports Med. 1986;8:83-87.

(17) Ciccotti MG, Kerlan RK, Perry J, Pink M. An electromyographic analysis of the knee during functional activities, II. The anterior cruciate ligament-deficient and -reconstructed profiles. Am J Sports Med. 1994;22:651-658.

(18) Kaalund S, Sinkjaer T, Arendt-Nielsen L, Simonsen O. Altered timing of hamstring muscle action in anterior cruciate ligament deficient patients. Am J Sports Med. 1990; 18: 245-248.

(19) Lass P, Kaalund S, LeFevre S, et al. Muscle coordination following rupture of the anterior cruciate ligament: electromyographic studies of 14 patients. Acta Orthop Scand. 1991;62:9-14.

(20) Limbird TJ, Shiavi R, Frazer M, Borra H. EMG profiles of knee musculature during walking: changes induced by anterior cruciate ligament deficiency. J Orthop Res. 1988;6:630-638.

(21) Shiavi R, Zhang L, Limbird T, Edmonstone M. Pattern analysis of electyomyographic linear envelopes exhibited by subjects with uninjured and injured knees during free and fast speed walking. J Orthop Res. 1992;10:226-236.

(22) Cohen J, Cohen P. Applied Multiple Regression Correlation for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates; 1983.

(23) Norkin CC, White DJ. Measurement of Joint Motion: A Guide to Goniometry. 2nd ed. Philadelphia, Pa: FA Davis Co; 1995.

(24) Kendall FP, McCreary EK, Provance PG. Muscles: Testing and Function. 4th ed. Baltimore, Md: Williams & Wilkins; 1993.

(25) Magee DJ. Orthopedic Physical Assessment. 3rd ed. Philadelphia, Pa: WB Saunders Co; 1997.

(26) Klein PJ, DeHaven JJ. Accuracy of three-dimensional linear and angular estimates obtained with the Ariel Performance Analysis System. Arch Phys Med Rehabil. 1995;76:183-189.

(27) Wilson DJ, Smith BK, Gibson JK. Accuracy of reconstructed angular estimates obtained with the Ariel Performance Analysis System. Phys Ther. 1997;77:1741-1746.

(28) Hollman JH, Deusinger RH. Videographic determination of instantaneous center of rotation using a hinge joint model. J Orthop Sports Phys Ther. 1999;29:463-469.

(29) Ericson MO, Nisell R, Arborelius UP, Ekholm J. Muscular activity during ergometer cycling. Scand J Rehabil Med. 1985;17:53-61.

(80) Cappozzo A, Catani F, Leardini A, et al. Position and orientation in space of bones during movement: experimental artefacts. Clin Biomech (Bristol, Avon). 1996;11:990-100.

(31) Winter DA. Biomechanics and Motor Control of Human Movement. 2nd ed. New York, NY: John Wiley & Sons Inc; 1990.

(32) Abdel-Aziz YI, Karara HM. Direct linear transformation from comparator coordinates into object space coordinates in close-range photogrammetry. In: Papers From the 1971 ASP Symposium on Close-Range Photogrammetry. Falls Church, Va: American Society for Photogrammetry; January 26-29, 1971:1-18.

(33) Loudon JK. Reliability of two videographic methods to obtain knee PICR in healthy human adults. Clin Kinesiol. 2000;54:4-12.

(34) Elias SG, Freeman MAR, Gokcay EI. A correlative study of the geometry and anatomy of the distal femur. Clin Orthop. 1990;260: 98-103.

(35) Nuno N, Ahmed AM. Sagittal profile of the femoral condyles and its application to femorotibial contact analysis. J Biomech Eng. 2001;123: 18-26.

(36) Zoghi M, Hefzy MS, Fu KC, Jackson WT. A three-dimensional morphometrical study of the distal human femur. Proc Instn Mech Engnrs [H]. 1992;206:147-157.

(37) O'Connor JJ, Zavatsky A. Kinematics and mechanics of the cruciate ligaments of the knee. In: Mow VC, Ratcliffe A, Woo SLY, eds. Biomechanics of Diarthrodial Joints. Vol 2. New York, NY: Springer-Verlag; 1990:197-241.

(38) Shavelson RJ, Webb NM. Generalizability Theory: A Primer. Newbury Park, Calif: Sage Publications Inc; 1991.

(39) Isear JA, Erickson JC, Worrell TW. EMG analysis of lower extremity muscle recruitment patterns during an unloaded squat. Med Sci Sports Exerc. 1997;29:532-539.

(40) Vergis A, Hindriks M, Gillquist J. Sagittal plane translations of the knee in anterior cruciate deficient subjects and controls. Med Sci Sports Exerc. 1997;29:1561-1566.

(41) Lysholm M, Messner K. Sagittal plane translation of the tibia in anterior cruciate ligament-deficient knees during commonly used rehabilitation exercises. Scand J Med Sci Sports. 1995;5:49-56.

(42) Henning CE, Lynch MA, Glick KR. An in vivo strain gage study of elongation of the anterior cruciate ligament. Am J Sports Med. 1985;13: 22-26.

(43) Markolf KL, Gorek JF, Kabo M, Shapiro MS. Direct measurement of resultant forces in the anterior cruciate ligament. J Bone Joint Surg Am. 1990;72:557-567.

(44) Jenkins WL, Munns SW, Jayaraman G, et al. A measurement of anterior tibial displacement in the closed and open kinetic chain. J Orthop Sports Phys Ther. 1997;25:49-56.

(45) Fu FH, Woo SLY, Irrgang JJ. Current concepts for rehabilitation following anterior cruciate ligament reconstruction. J Orthop Sports Phys Ther. 1992;15:270-278.

(46) Shelbourne KD, Nitz P. Accelerated rehabilitation after anterior cruciate ligament reconstruction. Am J Sports Med. 1990;18:292-299.

(47) Solomonow M, Baratta R, Zhou BH, et al. The synergistic action of the anterior cruciate ligament and thigh muscles in maintaining joint stability. Am J Sports Med. 1987;15:207-213.

(48) Rudolph KS, Eastlack ME, Axe MJ, Snyder-Mackler L. Movement patterns after anterior cruciate ligament injury: a comparison of patients who compensate well for the injury and those who require operative stabilization. J Electromyogr Kinesiol. 1998;8:349-362.

(49) Snyder-Mackler L, Fitzgerald GK, Bartolozzi AR, Ciccotti MG. The relationship between passive joint laxity and functional outcome after anterior cruciate ligament injury. Am J Sports Med. 1997;25:191-195.

(50) Blankevoort L, Huiskes R, de Lange A. The envelope of passive knee joint motion. J Biomech. 1988;21:705-720.

(51) Lafortune MA, Cavanagh PR, Sommer HJ III, Kalenak A. Three-dimensional kinematics of the human knee during walking. J Biomech. 1992;25:347-357.

(52) Asano T, Akagi M, Tanaka K, et al. In vivo three-dimensional knee kinematics using a biplanar image-matching technique. Clin Orthop. 2001;388:157-166.

(53) Kurosawa H, Walker PS, Abe S, et al. Geometry and motion of the knee for implant and orthotic design. J Biomech. 1985;18:487-499.

(54) Siu D, Rudan J, Wevers HW, Griffiths P. Femoral articular shape and geometry: a three-dimensional computerized analysis of the knee. J Arthroplasty. 1996;11:166-173.

(55) Lucchetti L, Cappozzo A, Cappello A, DellaCroce U. Skin movement artefact assessment and compensation in the estimation of knee-joint kinematics. J Biomech. 1998;31:977-984.

(56) Reinschmidt C, van den Bogert AJ, Nigg BM, et al. Effect of skin movement on the analysis of skeletal knee joint motion during running. J Biomech. 1997;30:729-732.

JH Hollman, PT, PhD, is Assistant Professor, Department of Physical Therapy, Clarke College, 1550 Clarke Dr, Dubuque, IA 52001-3198 (USA) ([e-mail]). Address all correspondence to Dr Hollman.

RH Deusinger, PT, PhD, is Assistant Professor, Departments of Internal Medicine and Biomedical Engineering, Washington University School of Medicine, St Louis, Mo.

LR Van Dillen, PT, PhD, is Assistant Professor, Program in Physical Therapy, Washington University School of Medicine.

MJ Matava, MD, is Assistant Professor, Department of Orthopaedic Surgery, Washington University School of Medicine.

Dr Hollman, Dr Deusinger, and Dr Van Dillen provided concept/research design and writing. Dr Hollman provided data collection. Dr Hollman and Dr Van Dillen provided data analysis. Dr Hollman and Dr Deusinger provided project management. Dr Matava provided subjects. Dr Deusinger provided facilities/equipment and institutional liaisons. Dr Deusinger, Dr Van Dillen, and Dr Matava provided consultation (including review of manuscript before submission). The authors thank Shirley Sahrmann, PT, PhD, FAPTA, Scott Minor, PT, PhD,Jack Engsberg, PhD, Kevin Truman, PhD, and Dequan Zou, DSc, for reviewing portions of the manuscript and Janice Loudon, PT, PhD, ATC, for laying the groundwork for this research.

The Washington University School of Medicine Institutional Review Board approved the protocol for this study.

The results of this study were presented in a poster at the 47th Annual Meeting of the Orthopaedic Research Society, February 25-28, 2001, San Francisco, Calif, and in an abstract in the Transactions of the 47th Annual Meeting, Orthopaedic Research Society.

This project was funded in part by NIH-NCMMR training grant #5T32HDO743405.

This article was submitted September 21, 2001, and was accepted March 28, 2002.