Anterior Tibial Translation During Progressive Loading of the ACL-Deficient Knee During Weight-Bearing and Nonweight-Bearing Isometric Exercise

Anterior Tibial Translation During Progressive Loading of the ACL-Deficient Knee During Weight-Bearing and Nonweight-Bearing Isometric Exercise H. john Yack, PhD, PT, 01' lias M. Riley, BS, ATC, PAC2 Terry R. Whieldon, MS, ATC3 M any protocols for rehabilitating the quadriceps following anterior cruciate ligament (ACL) injury or repair use progressive loading of the quadriceps during both weight-bearing and nonweightbearing exercises (5,14-16.22.26). The rehabilitation concern that distinguishes weight-bearing exercises from nonweight-bearing exercises is thought to be the level of strain placed on the ACL and other passive restraining mechanisms. In vivo studies comparing closed kinetic chain exercises (weight-bearing) with open kinetic chain exercises (nonweightbearing) have demonstrated significantly less anterior tibial translation during closed kinetic chain exercises in subjects diagnosed with ACLdeficient knees (7.17,18,24,28). Less anterior tibial translation is associated with less strain to the ACL and other passive restraining structures (6). The decreased anterior tibial translation attributed to weight-bearing exercise is related to the ability of knee joint compressive forces. with possible concomitant hamstrings coactivation, to counter anterior shear forces (8,9,12,19,20,24). Cadaver studies have shown that Many protocols for rehabilitating the knee following anterior cruciate ligament (ACL) injury or repair call for the use of both weight-bearing and nonweight-bearing exercises. However, not much is known about the strain these exercises place on the passive restraining mechanisms. The purpose of this study was to examine the effect of progressive loading of the knee extensors on anterior tibial translation during weight-bearing and nonweight-bearing isometric exercise. Fourteen subjects diagnosed as ACL-deficient via arthroscopic surgery participated in the study. An arthrometer was used to measure anterior tibial translation during weight-bearing and nonweight-bearing exercises with the knee in 20" of flexion and with a quadriceps load equivalent to 25,50, 75, and?1)00/0 of body weight. Hamstring muscle activations during all testing were monitored to eliminate the possibility of substantial cocontractions. During the weight-bearing exercise, anterior tibial translation was found to be significantly less than that measured during the lachman's evaluation and the 50, 75, and 1000/0 nonweight-bearing exercises. For the nonweight-bearing exercises, anterior tibial translation was either equivalent to or greater than the lachman's evaluation. During the weight-bearing exercises, anterior tibial translation was not different under the different loading conditions. During the nonweight-bearing exercises, anterior tibial translation progressively increased with increased loading. The results from this study imply that weight-bearing exercises minimize strain to the passive restraining structures whereas nonweight-bearing exercises will strain these structures, and the magnitude of this strain appears to be proportional to the knee extensor moment. Key Words: anterior cruciate ligament, rehabilitation, biomechanics ' Associate Professor, Physical Therapy Graduate Program, College of Medicine, 2600 Steindler Building, The University of lowa, lowa City, IA 52242 ' Physicians Assistant, Orthopaedic Surgery, Albany, NY ' Physical Therapist; Fitness, Sports and Physical Therapy; Orchard apark, NY joint geometry in combination with axial compressive forces will limit anterior tibial translation even when the menisci, ACL, and medial collateral ligaments have been removed (8,9,19,2O). Compressive forces at the knee may be generated by weight-bearing forces and by the muscle force component acting along the long bones. In contrast, quadriceps contractions performed during open kinetic chain (nonweight-bearing) exercises at knee angles of less than 60" generate shear forces that increase anterior tibial translation (2,3,13,23,25). JOSPT Volume 20 Number 5 November 1994

During these nonweight-bearing exercises, anterior tibial translation may reach a maximal limit as the intensity of the quadriceps contraction increases (7). In in vivo studies, the effect of exercise on anterior tibial translation has generally been examined with the quadriceps exercised at a specific load. The ability to generalize these results to other loading conditions has not been established. The purpose of the study was to examine the effect of progressive loading of the knee extensors on anterior tibial translation during weight-bearing and nonweight-bearing isometric ex- Compressive forces at the knee may be generated by weightbearing forces and by the muscle force component acting along the long bones. ercise. By comparing anterior tibial translation during increased quadriceps loading under both conditions, the clinician will have a clearer understanding of how to progress patients following ACL injury. METHODOLOGY The study included 14 subjects, eight males and six females, with a mean age of 28 years (range 18-50 years). Subjects were all diagnosed as ACLdeficient within the past 24 months via arthroscopic surgery, which was performed at least 2 months prior to testing. Subjects were healthy except for their knee injury and had not undergone any type of ACL repair or reconstructive surgery. All subjects received information on the testing protocol and signed an institutional review board approved informed consent form before participating. A health history was completed on all subjects to screen out individuals for whom exercise would be contraindicated. Instrumentation A Knee Signature System arthrometer (Orthopedic Systems, Inc., Hayward, CA) was used to measure anterior tibial translation and to monitor knee joint angle. The extraskeletal metal tubular system of the Knee Signature System was aligned and secured in place using Velcrom straps. Anterior tibial translation was measured by a potentiometer with a fixed position over the tibial tuberosity and a movable arm positioned over the center of the patella (Figure 1). The electrogoniometer used to measure knee flexion was aligned with the axis of rotation of the knee joint estimated to be around the lateral femoral epicondyle. Alignment was adjusted so that flexing and extending the leg produced minimal anterior and lateral movement of the potentiometer throughout the range of motion. The reliability of the Knee Signature System for knee extension has been reported to be less than f 1.5 mm with the 90% confidence interval (2 1). The validity of the Knee Signature System has been suggested by concurrently measured strain on the ACL and anterior tibial translation using the Knee Signature System (6). The muscle activation level of the medial hamstrings was monitored with surface electromyography (EMG) and was assumed to be representative of all parts of the hamstrings. Skin overlying the muscle belly on the involved leg was shaved and abraded with an abrasive paste. Eight-millimeter diameter silver/silver-chloride electrodes were applied to the midportion of the muscle bellies. The EMG signal from the medial hamstrings was amplified, fullwave rectified, and low-pass filtered FIGURE 1. The arthrometer (Knee Signature System) fitted to the lower limb (lateral view). Potentiometer (6) is used to measure anterior tibial translation with the movable arm W positioned over the patella. Potentiometer (C) is used to measure the knee joint angle and is aligned with the knee joint center. at 3 Hz to obtain a linear envelope representation using a custom-made EMG telemetry system (P. Guy Circuit Design, Kitchner, Ontario, Canada) (27). The EMG signal was recorded and monitored using a chart recorder. A Kistler (Kistler Instrument Corporation, Amherst, NY) force platform was used to measure the vertical component of the ground reaction force data. Because measurements were made during quiet standing, the shear components of the ground reaction force were negligible and not included in the analysis. The output from the four vertical channels of the force platform were summed to provide a measure that represented the total vertical force. These same force measures were used to calculate the effective point where the vertical force was acting (center of pressure) using the method provided by the manufacturer. The amplification of the force platform signal was adjusted so that Volume 20 Number 5 November 1994 JOSPT

subjects standing with both feet on the force platform produced a full scale reading on a voltmeter. With the voltmeter within the subject's field of view, it was possible for the subject to monitor the relative amount of weight on the force platform. Procedures An 89-N Lachman's evaluation was performed to assess the amount of anterior tibial translation (1). During this assessment, the subject was seated upright on a Cybex I1 (Cybex, Division of Lumex, Ronkonkoma. NY) isokinetic dynamometer chair, with the involved knee set at 20" of flexion using thigh and ankle straps. The Knee Signature System was a p plied to the affected lower limb. Using the force applicator, also part of the Knee Signature System, an 89-N force was applied to the test leg just below the tibial tubercle in the posterior direction to set the knee in a neutral position. In the same manner, the applicator was placed behind the leg at the level of the tibial tubercle, and a 89-N force was applied in the anterior direction. The remaining portions of the testing protocol were divided into weight-bearing and nonweight-bearing conditions and were performed with the Knee Signature System attached to the ACLdeficient limb. For the weight-bearing conditions, subjects were instructed to stand with both feet on the force platform to determine a measure of total body weight. Subjects were then instructed to straddle the force platform so that the foot on the involved side remained in the center of the platform and parallel to the other foot, which was placed off the platform. Subjects were asked to maintain their knee in 20" flexion. This angle would allow near maximal anterior tibial translation as well as require a substantial extensor moment at the knee, and the subject could maintain the angle during data JOSPT * Volume 20 * Number 5 * November 1994 collection. Subjects were further instructed to alter the amount of weight on the test leg using feedback from the voltmeter. While the test leg remained at 20" of knee flexion, anterior tibial translation was measured with the subject bearing 100, 75, 50, and 25% of body weight. The four weight-bearing conditions were presented in a random order. The knee extensor moment during the nonweight-bearing exercise was made equivalent to the knee extensor moment in the weight-bearing exercise (28). For each of the weightbearing conditions, the subject stood on a force platform and squatted to a 20" knee angle, holding that position for a maximum of 20 seconds. The force platform output was used to determine the magnitude and point of application of the subject's ground reaction force. Simultaneously with the force platform measurements, a Polaroidm photograph was taken of the subject in the squatting position. Using these photographs and the force platform data. Differences between weight-bearing and non weight-bearing conditions exist above the 2% level but not at the 2% level. measurements were made of the distance of the knee joint center from the line of action of the ground reaction force (Figure 2). This enabled calculation of the knee extensor moment, assuming no cocontraction by the knee flexors. During nonweight-bearing testing, subjects were positioned on the Cybex chair. Waist and thigh straps were used to hold the torso and thigh in place. The knee on the test limb was positioned in 20" of flex- FIGURE 2. Diagrammatic representation of the tested lower limb of a subject standing on a force platform during the weight-bearing exercise. The line of action of the weight force (W) is shown with the magnitude and location of this weight force determined from the force platform outputs. This weight force will produce a flexor moment (Mf) equal to the product of the weight force W) multiplied by the distance from the!he of action of W to the knee joint center (moment arm X). In order for subjects to maintain this knee joint angle, their quadriceps must generate a net knee extension moment (M J equivalent to the flexor moment (Mf). ion. The axis of rotation of the movement arm of the Cybex was aligned with the electrogoniometer used to measure knee flexion. The distal leg, just proximal to the ankle joint, was attached to the movement arm. A 2.2-kN strain gauge, attached to the movement arm of the Cybex, was used to monitor the force generated by the tibia acting on the pad attached to the strain gauge. Knowing the distance from the knee joint center to the center of the strain gauge where the distal leg was attached, the moment generated during the nonweight-bearing exercise could be regulated and made equivalent to the moment developed during the weight-bearing exercise. With the test leg strapped to the movement arm, the subject per-

formed an isometric quadriceps contraction while anterior tibial translation was measured. Subjects were discouraged from pushing against the back support during the testing. Hamstring muscle activation levels were monitored during all testing to ensure that a substantial cocontraction of the hamstrings and quadriceps did not occur. Prior to testing, a maximum hamstring contraction was recorded. The recorder output was monitored during testing. No trial was accepted for analysis in which greater than 10% of the maximum activity level was detected. action between factors (p <.05). Further analysis of the results showed that while differences among loading conditions existed for the nonweight-bearing condition, no such differences were detected for the weight-bearing condition. It was also apparent that differences between weight-bearing and nonweight-bearing conditions existed above the 25% level but not at the 25% level. The mean data for the weightbearing and nonweight-bearing trials are shown in Figure 3. The Table summarizes the results of post hoc testing of the data indicating where significant differences were found. During all weight-bearing conditions, anterior tibial translation was significantly less than the translation that occurred during the Lachman's evaluation. During the nonweightbearing conditions, anterior tibial translation was not significantly different from the Lachman's evaluation, except for the 100% nonweight-bearing condition in which anterior tibial translation was significantly increased. Data Analysis A two-way (2 X 9) analysis of variance (ANOVA) with repeated measures was used to compare weight-bearing and nonweight-bearing conditions across the different loading conditions. When significant differences were detected (p <.05), a Student Newman Keuls test was used to make multiple comparisons. RESULTS The results from the ANOVA showed that there were significant differences between the weight-bearing and nonweight-bearing conditions ( p <.05) and between the different loading conditions (p <.05) and that there was significant inter- V 2596 5096 7596 10096 25% 50% 75% 100% WEIGHT BEARING NON-WEIGHT BEARING FIGURE 3. Comparison of anterior tibial translation for weight-bearing and nonweight-bearing isometric exercises with the knee in 20" of flexion. Mean values and standard errors of anterior tibial translation for 14 subjects are shown for the different loading conditions of both exercises. The heavy horizontal line indicates the anterior tibial translation magnitude during the 89-N lachman's evaluation. Lachman 100%WB 75%WB 50%WB 25%WB 100%NWB 75%NWB 50%NWB 25%NWB 8.97 6.88 6.95 6.54 6.38 11.47 10.44 10.07 8.77 Lachman 100% WB < 75% WB < NS 50% WB < NS NS 25% WB < NS NS NS 100% NWB > > > > > 75% NWB NS > > > > < 50% NWB NS > > > > < NS 25% NWB NS NS > NS NS < < < TABLE. Post hoc comparisons of anterior tibial translation measures for weight-bearing (WE) and nonweight-bearing (NWB) isometric conditions with the knee in 20" of flexion. The greater than (>) and less than (<) symbols refer to the direction of significant differences when the data identified in the first column are compared with the data in the first row (eg., the < symbol in column 2, row 2 means that anterior tibial translation for the 100"/0 W8 condition was significantly less than the anterior tibial translation during the lachman's evaluation). Nonsignificant (NS) differences are also labeled. Mean values (mm) for each condition are given beneath the column headings. 250 Volume 20 Number 5 November 1994 JOSPT

Between conditions, comparisons of anterior tibial translation during nonweight-bearing exercise at 50, 75, and 100% loads were significantly greater than anterior tibial translation measured during all weight-bearing conditions. There were no differences between the nonweight-bearing 25% load and any of the weight-bearing conditions. Within-condition comparisons demonstrated no differences in anterior tibial translation during the weight-bearing conditions. In contrast, anterior tibial translation measures during nonweight-bearing conditions were significantly different, with the exception of comparisons between the 50 and 75% loading conditions. The differences in anterior tibial translation between the 25 and 100% conditions were calculated for the weight-bearing and nonweightbearing conditions for each subject. For the weight-bearing condition, the mean difference was.81 mm (range 2.89 to -2.08 mm), with three of the subjects having differences of greater than 2 mm. For the nonweight-bearing condition, the mean difference was 2.75 mm (range 4.54 to -.36 mm), with three of the subjects having differences of less than 1 mm. DISCUSSION The results of this study support previous research that has demonstrated less anterior tibial translation under weight-bearing conditions compared with nonweight-bearing conditions. In addition, progressive loading of the lower limb under weight-bearing conditions did not increase anterior tibial translation. It has also been shown that anterior tibial translation during different nonweight-bearing loading conditions may result in anterior tibial translation equal to or greater than anterior tibial translation during an 89-N Lachman's evaluation. JOSPT Volume 20 Number 5 November 1994 In the present study, weightbearing exercises were effective in limiting anterior tibial translation below what is seen in a Lachman's evaluation and below what occurred in the nonweight-bearing exercises. It can be inferred from these results that strain on the mechanisms that limit anterior tibial translation is restricted during weight-bearing exercise. On the other hand, during nonweight-bearing exercise, the limits of anterior tibial translation established during the Lachman's test were either equaled or, in the case of the 100% loading condition, exceeded. The greater anterior tibial translation during nonweight-bearing exercise indicates additional strain to the passive knee joint structures beyond that experienced during weightbearing exercise. These results agree with other research showing that weight-bearing exercises lin-' 11t anterior tibial translation compared with nonweight-bearing exercises (9.28). In individuals with an intact ACL, the primary passive structure for limiting anterior tibial translation between O and 60" otmotion is the ACL (9,10,19). In the ACL-deficient knee, however, anterior tibial translation is limited by secondary restraints, ie., the posterior joint capsule and collateral ligaments (4). Thus, nonweight-bearing exercise places repeated strain on tissues that may already be injured or surgically repaired or whose secondary function is to control anterior tibial translation. The hamstrings are capable of limiting anterior tibial translation by providing a posterior shear force and bv contributing to the cornpressive forces acting at the knee. C' riven the angle of insertion of the hamstrings with the knee in 20" of flexion, it is questionable whether the muscle can generate a significant shear force relative to the anterior shear force that is generated by the quadriceps (20). At the knee angle tested, therefore, cocontraction of these muscles may function to add to the compressive forces. In this experiment, the role of the hamstrings in controlling anterior tibial translation was minimized by maintaining activation levels below 10% of maximum activation. The knee response, as measured by anterior tibial translation under various loading conditions, was different within the weight-bearing and nonweight-bearing exercises. Similar amounts of anterior tibial translation occurred during weight-bearing exercise for the various loading conditions. By contrast, in the nonweightbearing exercises, there was a direct relationship between the level of loading and anterior tibial translation. Hsieh and Walker defined an "uphill" theory to explain the relationship between axial loading and tibial translation (1 2). According to their theory, in order for tibial translation to occur, the tibia must overcome the geometrical conformity that exists between the tibia and the femur. With the knee joint approximated, anterior tibial translation occurs as the result of shear forces that move the femoral condyles both posterior and uphill. Thus, a relationship between shear force and axial force regulates the magnitude of tib ial transition. In the static conditions investigated in our study, the shear forces developed for different levels of knee extensor torque were apparently unable to overcome the associated increase in approximation forces due to the increased weightbearing force. While significant differences occurred between the weight-bearing and nonweight-bearing conditions. some individuals who did not conform with the larger group deserve mention. Examination of the differences in anterior tibial translation between the 25 and 100% conditions revealed a subgroup of three individuals who did not show substantial tibial translation (> 1 mm) during nonweight-bearing exercises. In these individuals, anterior tibial translation was less than what was

measured during the Lachman's evaluation. The results for this group are consistent with the findings of other research in which nonweight-bearing exercise did not necessarily result in excessive anterior tibial translation (1 1). One explanation of these contrary results is that additional secondary forces generated by the subjects as they pushed A subgroup of three individuals did not show substantial tibial translation during non weight-bearing exercises. against the chair back increased the axial compressive forces (29). For these individuals, nonweight-bearing exercise will not strain the passive restraining structures to the same extent as seen in most subjects. The other subgroups of three individuals demonstrated greater than expected translation during weightbearing exercise (> 2 mm). We speculate that these individuals may have a different joint geometry with shallow seating of the femoral condyles on the tibia that permitted greater joint translation as the shear forces increased. This translation, however, was still less than what occurred during the Lachman's evaluation. Thus, weight-bearing exercise for this group is still a viable method of limiting anterior tibial translation. For either subgroup, no other distinguishing features could be identified to further categorize these individuals. Thus, possible individual variations in performance or anatomy which may result in different responses must be considered. SUMMARY The results of this study have implications for current rehabilitation protocols used with either conservatively treated or surgically treated ACL-deficient patients. While standard nonweight-bearing hamstring exercises are typically performed early in the rehabilitation process, the sequencing and type of quadriceps exercises are less well defined. Ideally, the quadriceps exercise program should be designed to regulate the amount and timing of stress applied to the injured or repaired ligament. Many protocols advance patients through both weight-bearing and nonweight-bearing exercise regimes in which loading of the quadriceps is progressive. The implications of our results are that patients may be able to perform isometric quadriceps weight-bearing exercises with loads equivalent to 100% body weight without causing excessive strain to the passive restraints. When used during the initial part of the rehabilitation program, these exercises have the potential to meet the criteria for appropriate rehabilitation exercise. On the other hand, progressive loading during nonweight-bearing exercise generally produces progressive strain on the passive restraining mechanisms, something that may not be desired until the later stages of rehabilitation. JOSPT REFERENCES Anderson AF, Snyder RB, Federspiel CF, Lipscomb AB: Instrumented evaluation of knee laxity: A comparison of five arthrometers. Am I Sports Med 20: 135-140, 1992 Arms SW, Pope MH, lohnson RI, Fischer RA, Arvidsson I, Eriksson E: The biomechanics of anterior cruciate ligament rehabilitation and reconstruction. Am 1 Sports Med l2:8-18, 1984 Branch TP, Hunter R, Donath M: Dynamic EMC analysis of anterior cruciate deficient legs with and without bracing during cutting. Am Sports Med 17(1):35-4 1, 1989 4. Butler DL, Noyes FR, Crood E: Ligamentous restraints to anterior-posterior draw in the human knee. 1 Bone joint Surg 62A:259-270, 1980 5. DeMaio M, Noyes FR, Mangine RE: Principles for aggressive rehabilitation after reconstruction of the anterior cruciate ligament. Orthopedics l5:385-392, 1992 6. Fleming B, Beynnon B, Nichols C, lohnson R, Pope M: An in vivo comparison of anterior tibial translation and strain in the anterior medial band of the anterior cruciate ligament. I Biomech 26:5 1-59, 1993 7. Franklin ll, Rosenberg TD, Paulos LE, France EP: Radiographic assessment of instability of the knee due to rupture of the anterior cruciate ligament. / Bone loint Surg 73A:365-372, 1991 8. Crood ES, Suntay WI, Noyes FR: Biomechanics of the knee-extension exercise. I Bone loint Surg 66A:725-734, 1984 9. Henning CE, Lynch MA, Click KR: An in vivo strain gauge study ofelongation of the anterior cruciate ligament. Am I Sports Med 13(1):22-26, 1985 10. Hirokawa S, Solomonow M, Lu Y, Lou ZP, D'Ambrosia R: Anterior-posterior and rotational displacement of the tibia elicited by quadriceps contraction. Am I Sports Med 20:299-304, 1992 11. Howell SM: Anterior tibial translation during a maximum qmdriceps contraction: Is it clinically significant? Am I Sports Med 18:573-578, 1990 12. Hsieh HH, Walker PS: Stabilizing mechanisms of the loaded and unloaded knee joint. I Bone joint Surg 58A(1):87-93, 1976 13. lonsson H, Karrholm 1, Elmqvist LC: Kinematics of active knee extension after tear of the anterior cruciate ligament. Am I Sports Med 17:796-802, 1989 14. Malone TR, Garrett WE: Commentary and historical perspective of anterior cruciate ligament rehabilitation. I Orthop Sports Phys Ther 15:265-269, 1992 15. Mangine RE, Noyes FR: Rehabilitation of the allograft reconstruction. Orthop Sports Phys Ther 15:294-302, 1992 16. Mangine RE, Noyes FR, DeMaio M: Minimal protection program: Advanced weight bearing and range of motion after ACL reconstruction- Weeks I to 5. Orthopedics 15:504-515, 1992 17. Markolf KL, Bargar WL, Shoemaker SC, Amstutz HC: The role of joint load in knee stability. I Bone joint Surg 63A(4):570-585, 1981 18. Neuschwander D, Drez D, Paine R, Volume 20 Number 5 November 1994 JOSPT

Young /C: Comparison of anterior laxity measurements in anterior cruciate deficient knees with two instrumented testing devices. Orthopedics l3:299-302, 1990 19. Noyes FR, Crood ES, Butler DL, Malek M: Clinical laxity tests and functional stability of the knee: Biomechanical concepts. Clin Orthop 146.34-89, 1980 20. Renstrom P, Arms SW, Stamwyck TS, lohnson R/, Pope MH: Strain within the anterior cruciate ligament during hamstring and quadriceps activity. Am / Sports Med 14(1):83-87, 1986 2 1. Riederman R, Wroble RR, Crood ES, VanCinkel L, Shaffer BL: Reproducibility of the Knee Signature System. Am I Sports Med 19(6):660-664, 199 1 22. Shelbourne KD, Nitz P: Accelerated rehabilitation after anterior cruciate ligament reconstruction. Am / Sports Med l8:292-299, 1990 23. Shoemaker SC, Markolf KL: Effects of joint load on the stiffness and laxity of ligament deficient knees: An in vitro study of the anterior cruciate and medial collateral ligaments. 1 Bone joint Surg 67A: 136-146, 1985 24. Solomonow M, Baratta R, Zhou BH, Shoji H, Bose W, Beck C, D'Ambrosia R: The synergistic action of the anterior cruciate ligament and thigh muscles in maintaining joint stability. Am / Sports Med 15(3):207-2 13, 1987 25. Wang C-1, Walker PS: Rotary laxity of the human knee joint. / Bone joint Surg 56A(I): 16 1-1 70, 1974 26. Wilk KE, Andrews /R: Current concepts in the treatment of anterior cruciate ligament disruption. 1 Orthop Sports Phys Ther 15:279-293, 1992 27. Winter DA: Biomechanics and Motor Control of Human Movement, New York: john Wiley & Sons, Inc., 1990 28. Yack HI, Collins CE, Whieldon 1): Closed kinetic chain exercise for the ACL deficient knee. Am / Sports Med 2 1 :49-54, 1993 29. Yack Hl, Washco LA, Whieldon TI: Compressive forces as a limiting factor of anterior tibia1 translation in the ACL deficient knee. Clin / Sports Med, 1994 (in press) JOSPT Volume 20 Number 5 November 1994