Controlling Anterior Shear During lsokinetic Knee Extension Exercise

Transcription

1 /82/O4Ol-OO23$02.OO/O THE JOURNAL OF ORTHOPAEDIC AND SPORTS PHYSICAL THERAPY Copyright O 1982 by The Orthopaedic and Sports Physical Therapy Sections of the American Physical Therapy Association Controlling Anterior Shear During lsokinetic Knee Extension Exercise DAVE JOHNSON, PT* The purpose of this paper is to present a means of controlling anterior shear force at the knee during isokinetic exercise. A dual-pad attachment to existing exercise equipment allows selection of the degree of anterior shear placed on the knee during exercise, ranging from a value which apparently places minimal stress on ligamentous restraints, through intermediate shear levels, to a value nearly that of standard exercise equipment. The special problems associated with anterior cruciate ligament (ACL) involvement are reviewed in terms of ACL function, failure, and healing. Calculated loading configurations illustrate the effect on shear afforded by the device. A pilot roentgenographic study indicates restoration of normal tibiofemoral alignment and joint surface velocity vectors during isometric extension with the device. This preliminary finding correlates with clinical observation. Clinical application includes isokinetic exercise of the chronically lax knee and rehabilitation of the knee after ACL repair or reconstruction. Use of an exercise apparatus during the latter stages of knee rehabilitation is widespread. The typical apparatus applies resistance to the shank at a point near the malleoli. This location allows considerable shear force at the knee; therefore. use of standard exercise equipment carries the risk of overstressing healing or repaired ligaments. Too frequently, the athlete with anterior cruciate ligament (ACL) involvement emerges from rehabilitation with excellent strength, power, and endurance, but has paid for these gains with increased ligamentous laxity due to excessive shear (or "anterior tibia1 translocation forceu4) produced during vigorous quadriceps exercise. A similar problem is encountered by the athlete with chronic anterior laxity. During extension, the tibia may subluxate anteriorly, causing an abnormal relationship of glide and roll. This phenomenon may be manifested by grating and crepitus at the tibiofemoral joint as the knee approaches terminal extension, and may result in increased articular surface wear. It is not within the scope of this paper to delineate rehabilitation procedures specific to ACL injury or repair, so the reader is referred to two excel- Mr. Johnson was a Staff Physical Therapist at Mercy Health Center. Dubuque. IA when th~s paper was prepared. He IS now a graduate student in Manipulative Therapy at the Western Australian Institute of Technology. Shenton Park. Western Australia lent articles relating to the subject.ls *' Regardless of specific approach, a properly designed exercise program must consider the disruptive potential of shear force throughout the stages of rehabilitation. Due to the above risks, some athletic rehabilitation clinics prefer to use only mass extension exercise such as cycling,4. 20 step-ups, leg press, or stair climbing well into the final stages of rehabilitation. While this type of exercise may protect the knee from excessive shear, it probably does not as fully or efficiently develop strength and power, as compared with isokinetic exerci~e.~' In view of this dilemma, a modification of existing isokinetic equipment is presented, and is hereafter referred to as the dual-pad device.t It appears to allow vigorous low- and high-speed exercise while controlling anterior shear during knee extension. The device is composed of two freely pivoting resistance pads, which are secured to the shank at proximal and distal positions (Fig. 1). The pads are connected by a bar that also freely pivots at a fulcrum attached to the input arm of the exercise apparatus. The device may be fitted to athletes of all sizes by altering the distance between the resistance t U. S. Patent applied for.

2 24 JOHNSON JOSPT Vol. 4, No. 1 Fig. 1. Dual-pad device in use with an Orthotron isokinetic exercise apparatus. Shear force may be selected by altering the fulcrum position but keeping the resistance pads in the positions shown. This particular position is effective in preventing anterior subluxation during isokinetic extension. pads. By changing the position of the fulcrum, one alters the balance of distal force (near the malleoli) and proximal force (at the tibia1 tubercle). REVIEW OF LITERATURE Ligaments at the knee joint have overlapping functions, and are thus termed primary and secondary restraints. For example, the ACL provides about 85% of the total restraint to anterior drawer, while all other ligaments combined account for the remaining 15%.' A ligament which acts as the primary restraint in one plane acts as a secondary restraint in another; thus, each ligament has multiple functions and each accessory motion is restrained by several ligaments. It follows that the loss of a primary restraint to a given motion causes the secondary restraints to bear the external force experienced by the joint. As Paulos et alz8 explain, intact secondary restraints prevent clinical detection of laxity, due to the small testing force applied. However, functional instability probably exists, due to considerably higher forces experienced during activity. Because secondary restraints are not designed to resist large in vivo loading, they may eventually stretch out and allow laxity to be clinically detected. Several authors report the incidence of the above phenomena in reference to the ACL.8, 10, Statistics compiled by Noyes et alz4 indicate that the clinical anterior drawer test is initially positive in only 24% of those knees with a torn ACL. If the initial diag- nosis does not implicate the ACL, proper precautionary measures may not be taken, and subsequent anterior instability is likely to become apparent as secondary restraints elongate. This history has been described as "the beginning of the end" for the knee,33 and is characterized by recurrent giving way, rotatory instability, and meniscal tearsqg If ACL involvement is diagnosed and treated, the subsequent healing and maturation process takes much longer than originally thought. In studies of immobilization and reconditioning of primates, Noyes et a1.26 noted only partial recovery of tensile properties at 5 months after resumed activity, and complete recovery required up to 12 months. Therefore, several authors recommend some type of ligament protection extending up to 1 year following injury A goal of rehabilitation after ligament injury is to increase functional stability, implying stability under static and dynamic loading.25 This is accomplished by training the muscles crossing the joint, with the intent not to stretch out the noncontractile elements during rehabilitation. To achieve this latter aspect, protection of healing ligaments must limit tensile forces to values below the failure level for the particular stage in the maturation process. The periods of time the healing structures are vulnerable to given degrees of stress are yet to be determined." Throughout this paper, no differentiation is made between rehabilitation of ligament sprains, repairs, or reconstructions. Although there certainly are differences in the time rates of healing of these entities, the general rehabilitative principles to be followed and the application of the dual-pad device are quite similar. In contrast to the need for protection, the beneficial effects of exercise may extend into the realm of ligament healing, although the exact nature of these effects is controversial. At the least, exercise reduces ligament strength loss due to immobilization and disuse If properly controlled, it allows the healing ligaments to remodel along the lines of stress, without el~ngating.~~ Exercise producing ligament forces larger than the critical value for a given stage of healing has the potential to disrupt maturing collagenous tissue. Forces below the critical value may provide "desirable feedback exposure" for yet undefined neurophysiological mechanisms that initiate repair and regulate its course.'

3 JOSPT Summer 1982 CONTROLLING SHEAR DURING KNEE EXERCISE 2 5 The normal knee or that with ACL insufficiency may be analyzed by the method of instant center~.~. 31 Normally, the instant center makes a smooth progression within the sagittal condylar silhouette as the knee is flexed. An instant center located on the joint surface indicates predominant roll, while one located at a distance from the contact area indicates predominant glide. Instant center analysis also yields an instant joint surface velocity vector. Normally, the vector is tangent to the articular contact area, indicating a smooth combination of glide and roll that does not tend to separate or impinge the joint surface~.~ In knees with ACL insufficiency, during the subluxated phase of the pivot shift maneuver, Tamea and Henning3' observed centrodes displaced anteriorly and located too near the joint line, indicating abnormal roll while the tibia is subluxated anteriorly. Frankel et al7 observed increased articular surface wear at the site of abnormal surface velocity in a knee with ACL laxity. Increased wear may be the result of "out of balance" motion between joint partners allowed by ligamentous laxity.25 MECHANICAL ANALYSIS OF PAD POSITION Before proceeding with a discussion of the dual-pad device, an issue that may need clarification is the effect of positioning a single resistance pad close to the knee joint during exercise Consider an athlete lifting a weight which is mounted at a distance from the extended knee joint, with the resistance pad located near the malleoli (Fig. 2). The total torque about the knee joint equals the sum of the moments produced by the 10-kilogram weight (external torque) and the below the knee (BK) segment (internal torque). In order to hold the knee in extension, the force produced by the quadriceps multiplied by the lever arm created by the patellar mechanism must equal this total torque. Thus, holding a 10-kilogram weight with a single resistance pad positioned near the malleoli requires an extension torque of 5.25 kilogrammeters. Consider an athlete lifting the same weight with the resistance pad mounted close to the knee joint (Fig. 3). Note that neither the external torque nor the internal torque about the knee has changed. Therefore, the extensor mechanism must produce the same torque to hold the knee in extension, despite the more proximal position of the resistance pad. The W 10 kgf ET (10)(0.40) ET kgf-n ET = IGR IT - (Sl(0.25) Cp - kgf IT kgf-m Fig. 2. Top view of terminal extension. W: weight being lifted. ET: external torque created by 10-kilogram weight, IGR: isokinetic gauge reading for an equivalent quadriceps contraction. 77: total torque created by quadriceps mechanism. IT: internal torque created by weight of BK segment, CF: contact force between athlete and resistance pad. L: lever arm of resistance pad. TT-rl'+IT TT - (10)(0.40)+(5)(0.25) TT kgf-m ET (lol(0.40) CF-4.0kgf-m kgf-m 0.20 rn ET - IGR IT (5)(0.25) Cp, 20 kgf IT kgf-m Fig. 3. Top view of terminal extension. Same symbols as in Figure 2. For an equivalent quadriceps contraction, reducing the resistance pad lever arm increases the athlete-pad contact force. same analysis may be applied to knee extension at any joint angle, and the internal and external torques for any given angle remain unaffected by changing the resistance pad position.

4 26 JOHNSON JOSPT Vol. 4. No. 1 In respect to isokinetic equipment, which measures torque and not force, an athlete extending the knee with a total torque of 5.25 kilogrammeters (as before), would expend 1.25 kilogrammeters to lift the BK segment (internal torque), and would apply a 4.0-kilogram-meter torque to the input arm of the isokinetic apparatus. The gauge reading, therefore, would reflect only the external torque created by the athlete. As explained above, external torque for any given joint angle would not be affected by changing the resistance pad position. Therefore, no difference in isokinetic torque measurement may be attributed to changing the resistance pad position. At least three parameters, however, are affected: stability of the seated athlete, contact force between the resistance pad and the athlete's shank, and shear force at the knee joint. If the athlete is not properly secured to the exercise table, proximal placement of the single pad or dual-pad fulcrum increases the tendency of the athlete to slide forward in the seat and the hip to flex during knee flexion (Fig. 4). This problem may account for decreased flexion torque readings, as part of the hamstring torque is expended in shifting the athlete's position. Fig. 4. Side view of knee flexion against single pad mounted close to knee joint. A: End position of flexion, inadequate fixation of seated athlete. Athlete slides forward, and hip flexes as knee achieves 90' flexion. Apparatus input arm excursion is less than 90'. 8: End position of flexion, proper fixation at waist and distal thigh. Apparatus input arm excursion is 903 Similarly, extension torque may be dissipated by returning the athlete to the initial position. In addition, an athlete may unintentionally exert submaximal effort (especially during extension) due to the unfamiliar "feel" of higher pad-shank contact force. Contact force, then, is another parameter which is affected by changing the resistance pad position. Since external torque is applied by the input arm to the shank, it is this torque (not total torque) that is used to calculate contact force. Figures 2 and 3 illustrate that contact force is doubled as the input arm length is halved. In summary, changing resistance pad position does alter contact force, but not external torque. Therefore, it requires the same quadriceps torque to lift a free weight or to register a given isokinetic gauge reading, regardless of resistance pad position (assuming proper stabilization and consistent effort). In reference to the dual-pad device, the total contact force shared by the two pads is also a function of input arm length (measured from joint axis to fulcrum position). Building on the previous example, assume that the athlete is of a size requiring 0.3-meter separation between the two resistance pads (Fig. 5). As the fulcrum is repositioned on the exercise apparatus, it also is moved in respect to the two pads, whose positions remain fixed. Table 1 summarizes calculated proximal and distal contact forces created by moving the fulcrum, while maintaining a fixed extension torque. Note that eight of the 10 dualpad forces shown are less than the standardposition. single-pad contact force. Decreased contact force results in a more comfortable "feel" during exercise. SHEAR DURING EXERCISE Determination of shear during isometric extension involves summation of anterior-posterior components of extension force, BK segment weight, and contact force (Fig. 6). For a given joint angle and extension force, moving the fulcrum proximally causes pad-shank contact force to increase and shear force to decrease. Thus, fulcrum position is the variable which influences knee joint shear force. Calculated anterior shear forces are shown in Table 2, which illustrates the possibility of selecting a broad range of shear stress placed on the knee. The above estimates of shear apply to joint angles of 30" or less, which is the range of largest anterior shear during e~tension.~

5 JOSPT Summer 1982 CONTROLLING SHEAR DURING KNEE EXERCISE 2 7 EX - IGR FF - 20 kgf The calculated shear forces of Table 2 would be applied to the ACL, since it provides primary resistance to anterior glide during terminal exten~ion.~. '' Additional restraint is provided by the secondary ligamentous restraints and articular contours of the compression-loaded tibiofemoral joint. By resisting anterior shear, an in- FQA kgf FQ kgf FF 0 20 kgf CP kgf CD kqf Fig. 5. A: Side view of terminal extension using dual-pad device. Same symbols as in Figure 2. FF: fulcrum force. For a quadriceps contraction equivalent to that of Figure 3, the fulcrum force equals the single-pad contact force. B: Enlarged schematic of dual-pad device. It splits the fulcrum force into proximal (CP) and distal (CD) contact forces. TABLE 1 Calculated contact forces during isometric extension ' lsokinetic gauge reading (kgf-m) Lever arm (m) Contact force (kg0 Single pad. standard position Dual-pad device t 1.9 and 9.5$ and and and and 4.4 Calculations based on anthropometric data adapted from S~nidt.~' t Fulcrum lever arm. $ Proximal and distal contact forces, respectively. 30 Regardless of joint angle, the dual-pad device applies posterior force to the shank in a balanced manner. By permitting less anterior shear near terminal extension, the device relieves tension on the restraints to anterior glide. The amount of additional stress placed on posterior restraints during extension is probably insignificant in comparison with their primary role of resisting posterior shear during flexion.' A FQ - TT d FQA 0 FQ (sin 19O) FS BK - 5 kgf FOA - BK - CP - CD Fig. 6. A: Side view of resolution of extension torque into anterior force component. Same symbols as in Figures 2 and 5. FQ: quadriceps force. d: lever arm of quadriceps force. FQA: anterior component of quadriceps force. Anthropometric data adapted from Smidt.30 B: Side view of anteriorposterior forces acting on the tibia during isometric extension. FS: shear force. Anterior shear is calculated by summation of the other forces. TABLE 2 Calculated anterior shear during isometric extension in terminal range ' lsokinetic gauge reading (kaf-m) Lever arm Anterior shear % of single (m) (kaf, pad shear Single pad. standard position Dual-pad device t Calculations based on anthropometric data adapted from S~nidt.~' t Fulcrum lever arm.

6 28 JOHNSON JOSPT Vol. 4, No. 1 tact ACL holds the tibia and femur in proper alignment during extension.12 In the absence of a functional ACL, the secondary restraints and articular contours attempt to control anterior glide during extension. Under the repeated stress of vigorous quadriceps exercise, secondary restraints may elongate and allow the articular surfaces to come into abnormal relationship. As the tibia subluxates anteriorly under the pull of the quadriceps tendon, the tibiofemoral joint may manifest crepitus and grating, which lead to increased articular surface wear. Calculation of hypothetical posterior shear during flexion is accomplished by a method similar to that for anterior shear during extension. The various levels of calculated posterior shear for an isometric hamstring torque are as shown in Table 3. As the knee is flexed, the dual-pad device has a decreasing effect on posterior shear (calculations not shown). Therefore, it is less effective in the range of maximal posterior shear. However, the device is extremely effective in controlling anterior shear in the range of critical importance, as shown in Table 2. Throughout the range of flexion and extension, the device exerts a stabilizing effect on the knee, and allows the therapist to select the degree of shear stress placed on the knee. CLINICAL PILOT STUDY A pilot study was undertaken to determine if the dual-pad device altered the sagittal alignment of the tibia and femur in a subject with severe anterior laxity. The subject's knee demonstrated over 1 centimeter of anterior drawer, positive pivot shift, but no posterior laxity. The TABLE 3 Calculated posterior shear during isometric flexion at 75 knee flexion ' lsokinetic gauge Lever arm 'OSterior shear % of single reading (kgf-m) (m) (kgf) pad shear Single pad, standard position Dual-pad device t Calculations based on anthropometric data adapted from SmidL30 t Fulcrum lever arm. history included a skiing injury (valgus and external rotation of tibia), medial meniscectomy, and gradually increasing anterior laxity over the ensuing 6 years. During isokinetic exercise using a single resistance pad mounted near the malleoli, the subject sensed anterior subluxation of the tibia and noted tibiofemoral grating and crepitus near terminal extension. During exercise with the dualpad device, the subject reported a sensation of knee stability throughout the flexion-extension cycle, and noted no tibiofemoral grating or crepitus. The pilot study involved taking nine mediolatera1 roentgenograms of the subject's knee at three angles near terminal extension under three loading conditions: zero external load with the shank held in space by the extensor mechanism, near-maximal quadriceps contraction with a single resistance pad positioned near the malleoli, and using the dual-pad device while exerting quadriceps contractions equivalent to those with a single pad. The subject was secured to a padded seat, and the single or dual pad was mounted on an Orthotron (Lumex, Inc., Ronkonkoma, NY) isokinetic apparatus. The fulcrum position used in the study was the most distal position that still eliminated tibiofemoral grating and crepitus during isokinetic exercise: 16.5 centimeters from knee axis, 10 centimeters from proximal pad, and 12.5 centimeters from distal pad. This position is analogous to the dual-pad configurations of Tables 1, 2, and 3 that employ a fulcrum lever arm of 0.20 meter. Figure 7 is a tracing compiled from the superimposition of nine roentgenograms. The respective positions of a clearly visible point on the intercondylar eminence are represented by nine dots. Each of the three interconnected triples represents an approximate arc of motion under the specified loading configurations. These arcs permit rough comparison of anterior-posterior tibiofemoral alignment. One may assume that the dual-pad arc most closely represents the normal tibiofemoral position for several reasons: absence of posterior laxity, subjective sensation of stability, and objective absence of grating and crepitus during exercise. Therefore, this tibia subluxated approximately 1 millimeter while the shank was merely held in space. It subluxated a total of 4 millimeters during extension against a single pad. This "voluntary quadriceps shiftv4 may account for tibiofemoral grating and crepitus, and

7 JOSPT Summer 1982 CONTROLLING SHEAR DURING KNEE EXERCISE 29 satisfactory in every way, except that pad-shank contact forces were higher. INSTANT CENTER ANALYSIS B standard-pad arc standard-pad \\\? oual-pad arc / zero-load arc Fig. 7. A: Tracing from superimposed roentgenograms. Nine dots represent positions of a point on the intercondylar eminence under three loading configurations as explained in text. B: Enlarged view of tracing. Angles of knee flexion for each point (as measured on the roentgenograms) illustrate anterior subluxation through the same range of motion. for sensible anterior displacement during isokinetic extension. Applying the previously mentioned method of estimating anterior shear to the three loading configurations of the pilot study yielded the following shear levels: zero load, 1.8 kilograms; dual pad, 7.1 kilograms; and single pad, 37.9 kilograms. Joint compressive force may account for the lack of anterior subluxation in the dualpad configuration, even though it involved greater anterior shear than the zero-load configuration. That exercise with the dual-pad device did not manifest grating and crepitus suggested that the device reduced anterior shear to the point that the compression-loaded articular contours were able to adopt a normal relationship of roll and glide during extension. Exercise with the fulcrum advanced only a few centimeters distal to the position used in this study resulted in crepitus and grating during terminal extension. Fulcrum positions proximal to that used in the study were lnstant center analysis was applied to three pairs of roentgenograms, with one pair representing each of the load configurations (Fig. 8). The tangential orientation of the dual-pad velocity vector represented a normal relationship of glide and roll. The zero-load vector was very nearly tangent, but the single-pad vector was clearly abnormal. The single-pad centrode was displaced anteriorly and located too near the joint line. Such a position correlates with the findings of Tamea and Henning3' for knees with anterior laxity. The zero-load centrode was also located somewhat close to the joint line, indicating a relatively larger proportion of roll. That the centrode and velocity vector were displaced and abnormal with the single pad, and apparently restored to normal with the dual-pad device, correlated with the displacement and restoration of tibiofemoral alignment described earlier. These findings offered exciting implications, but a controlled study of many subjects is needed to establish a firm conclusion. CLINICAL APPLICATION In clinical practice, the dual-pad device has proven helpful to athletes with known anterior Fig. 8. Tracing from superimposed roentgenograms. The anterior cluster of six dots represents the positions of one point on the intercondylar eminence, while the posterior cluster represents the corresponding positions of a point on the edge of the plateau. Anterior and posterior pairs for each loading configuration determine the centrode (D: dual pad, - S: single pad. Z: zero load). lnstant velocity vectors illustrate normal, tangential (D and 2).and abnormal orientation (S).

8 30 JOHNSON JOSPT Vol. 4, No. 1 laxity. The principles hypothesized by the foregoing isometric analysis have been found to apply equally well to exercise at low and high velocities. For athletes who experience tibiofemoral grating or crepitus due to anterior laxity, a fulcrum position somewhat closer to the proximal pad eliminates these problems at all exercise speeds. Thus, it appears that the dual-pad device would enable an athlete with a chronically hypermobile knee to increase strength, power, endurance, and functional stability without further increasing clinical anterior laxity. The device also offers the option of selecting the degree of stress placed upon healing or repaired ligamentous restraints to anterior glide. by positioning the fulcrum at the appropriate position. Once critical force levels have been experimentally defined for the various stages of ligament healing, the therapist may use these data to base the selection of fulcrum position. In the meantime, one may choose to initially employ the fulcrum position that subjectively and objectively allows apparently normal arthrokinematics for a knee with severe anterior laxity. If the device performs this function for a knee without an ACL contributing any restraint to anterior shear, it seems likely that it would create minimal or zero stress on an injured or repaired ACL. Provided the joint retains clinical stability, and in keeping with the estimated time rate of tension tolerance, the fulcrum could gradually be advanced distally over the long-term course of rehabilitation. Allowing controlled levels of tensile stress to a ligament during the maturation process may enhance the ligament's response to tension, and would gradually prepare it for functional stress. Alternatively, the therapist may retain a proximal fulcrum position if excess anterior laxity were already present. In addition, the device would benefit the athlete in whom no anterior laxity is detectable. Due to the overlapping functions of primary and secondary restraints, clinical anterior laxity may not initially be present. Use of the dual-pad device as a precautionary measure would be advised if involvement of the ACL were remotely suspected, so as not to take the chance of creating anterior laxity during neuromuscular rehabilitation. Beyond that, an athlete with an injury to any knee ligament might do well to use the dual-pad device in a sequence of increasing shear stress as the knee is rehabilitated. Special thanks to the University of lowa Athletic Trainers. Mercy Health Center. and Gary Soderberg. REFERENCES Butler DL. Noyes FR. Grood ES: Ligamentous restraints to anterior-posterior drawer in the human knee. J Bone Joint Surg 62- A: Cage JB: Strength imbalance and knee injury. Phys Sportsmed 8: Crowley E. Crowe J: Department of Physical Education. University of lowa, lowa City. IA (personal communication) Daniel D. Biden E. O'Connor J. et at: The quadriceps-anterior cruciate interaction. Presented at the Annual Meeting of the American Orthopaedic Society for Sports Medicine. Lake Tahoe. NE. June 1981 Dehne E. Torp RP: Treatment of joint injuries by immediate mobilization. Clin Orthop 77: Feagin JA: The syndrome of the torn anterior cruciate ligament. Orthop Clin North Am t0: Frankel VH. Burstein AH. Brook DB: Biomechanics of internal derangement of the knee. J Bone Joint Surg 53-A: Helfet AJ: Disorders of the Knee. pp Philadelphia: WE Saunders Co Isolated Joint Testing and Exercise: A Handbook for Using Cybex II and UBXT, p 10. Bay Shore. NY: Lumex. Inc James SL: Biomechanics of knee ligament reconstruction. Clin Orthop 146: Johnson RJ. Pope MH: Knee joint stability without reference to ligamentous function. In: Evarts CM (ed). American Academy of Orthopaedic Surgeons Symposium on Reconstructive Surgery of the Knee. Chap 2. St Louis: CV Mosby Co Kapandji IA: The Physiology of the Joints. Ed 2. Vol 2. pp New York: Churchill Livingstone Kennedy JC. Weinberg HW. Wilson AS: The anatomy and function of the anterior cruciate ligament. J Bone Joint Surg 56- A: Laros GS. Tipton CM. Cooper RR: Influence of physical activity on ligament insertions in the knees of dogs. J Bone Joint Surg 53-A: Larson RL: Combined instabilities of the knee. Clin Orthop 147:68-75,1980 LeVeau B: Williams and Lissner: Biomechanics of Human Motion. Ed 2. Chap 5. App A. Philadelphia: WE Saunders Co Lindahl 0. Movin A: The mechanics of extension of the knee joint. Acta Orthop Scand 38: Malone T. Blackburn TA. Wallace LA: Knee rehabilitation. Phys Ther 60: , 1980 McDaniel WJ. Dameron TB: Untreated ruptures of the anterior cruciate ligament. J Bone Joint Surg 62-A: McLeod WD. Blackburn TA: Biomechanics of knee rehabilitation with cycling. Am J Sports Med 8: t 980 Metzler KG. Metzler S: The Orthotron and knee torque. Phys Sportsmed 8: Nicholas JA: Rehabilitation of the knee in athletes. In: Helfet AJ (ed). Disorders of the Knee. Chap 23. Philadelphia: JB Lippincott Co Noyes FR: Functional properties of knee ligaments and alterations induced by immobilization. Clin Orthop 123: Noyes FR. Bassett RW. Grood ES. Butler DL: Arthroscopy in acute traumatic hemarthrosis of the knee. J Bone Joint Surg 62- A: Noyes FR. Grood ES. Butler DL. Malek M: Clinical laxity tests and functional stability of the knee: biomechanical concepts. Clin Orthop t46: Noyes FR. Towik PJ. Hyde WE. DeLucas JL: Biomechanics of ligament failure. J Bone Joint Surg 56-A: O'Donoghue DH: Treatment of acute ligamentous injuries of the knee. Orthop Clin North Am 4: Paulos L. Noyes FR. Grood ES, et al: Knee rehabilitation after anterior cruciate ligament reconstruction and repair. Am J Sports Med 9: Pipes TV. Wilmore JH: lsokinetic vs isotonic strength training in adult men. Med Sci Sports Exercise 7:

9 JOSPT Summer 1982 CONTROLLING SHEAR DURING KNEE EXERCISE Smidt GL: Biomechanical analysis of knee flexion and extension. 33. Torg JS. Conrad W. Kalen V: Clinical diagnosis of anterior J Biomech 6: cmciate ligament instability in the athlete. Am J Sports Med 31. Tamea CD. Henning CE: Pathomechanics of the pivot shift ma- 4: newer. Am J Sports Med 9: Wolff R (ed): The rehabilitation gospel according to the Bills. 32. Tipton CM. Matthes RD. Maynard JA, et al: The influence of Cramer First Aider 50: physical activity on ligaments and tendons. Med Sci Sports 35. Yocum LA. Bachman DC. Noble HB. Hoover RL: The deranged Exercise 7: knee: restoration of function. Am J Sports Med 6: