David S. Hungerford, M.D., Robert V. Kenna, and Darrell W. Haynes, M.S., Ph.D.


The purpose of this chapter is to review the characteristics of the forces which are transmitted across the knee during its normal use in performing the activities of daily living. It is not intended to be a comprehensive review of all that is known biomechanically about this complex joint, but rather to emphasize those points which impact directly on total knee arthroplasty. These points include the rotational stability of the normal knee, which is of paramount importance, not only for normal knee function, but also for the replaced knee.

Second, the forces crossing the knee will be analyzed in terms of magnitude and direction. Finally the mechanics of gait will be reviewed. The latter is important in designing and evaluating a rehabilitation program, which can spell the difference between success and failure in total knee arthroplasty. The combination of this chapter and the preceding one form the basis of our understanding of that reality which we are trying to simulate with our replacement surgery, i.e., the normal knee. This understanding of the normal knee is a prerequisite for a rational approach to knee replacement surgery and prosthetic design.


The preceding chapter has reviewed the kinematics of the normal knee, including the concept of two kinds of rotation; automatic and active. Both of these rotational movements have functional importance for the normal knee, and both have parameters which can be quantified. The kinematics as they relate to anatomical form have been considered, but it is important to note that the triaxial range of motion of the knee is orchestrated by muscular forces and load vectors as well as surface geometry. This is particularly true of active rotation which may or may not be occurring at any given degree of flexion. The limits to which active rotation may occur are defined not only by the anatomy of the tibia, femur, and menisci, but also by the tension which develops within the ligaments of the knee (3). Ligamentous tension which develops with motion is greatly influenced by the form of the femoral condyles and tibial plateaus, whether anatomical or prosthetic. For prosthetic components, ligament tension will also be influenced by the positioning of the components relative to the bony attachments of the ligaments. This will be more fully developed in Chapters 4 and 5.

In order to quantitate rotational stability under load, it is important to simulate normal loading conditions. Other studies assessing rotational stability of the knee under load have employed systems of jigs which constrain the knee and therefore impose a resistance to rotation which is largely a function of the jig and not solely a function of the knee itself (1). We have used a knee loading device which permits loading of the cadaver knee in such a way that abnormal constraints are not imposed by the jig.


Fresh cadaver knees with completely intact capsular structures and a minimum of 8 inches of femur and 6 inches of tibia were potted in fixation jigs usingWood's metal. The knees were then mounted in a knee testing device which allowed the

application of a vertical load. The degree of knee flexion was controlled by varying the quadriceps length (Fig. 3.1). The quadriceps tendon was mounted in a clamp and fixed to a threaded rod which could be lengthened or shortened in a holding clamp (Fig. 3.2). The rod was instrumented to measure quadriceps force required to resist the flexion of the knee under the applied load. The femoral jig was mounted on a transverse rod which could be positioned to align the femur so that the joint line would be horizontal with the knee in full extension. This transverse rod was attached to the loading frame through right and left sleeve bearings which allowed the rod to rotate, thus permitting flexion and extension of the knee. The tibial jig was fastened to the apparatus in such a way that flexion-extension, and varus-valgus shift were unconstrained. A foot plate which moved with the tibial mounting jig could be manually rotated to effect internal or external rotation of the mounted tibia in relation to the femur (Fig. 3.3). The foot plate was instrumented to show both torque and degree of rotation when a rotatory moment around the long axis of the tibia was applied to the knee.

Figure 3.1. Testing device for holding and loading the cadaver knee during rotational stability studies.

Figure 3.2. Lateral view of testing rig showing mechanism for flexing and extending the knee under load.

Figure 3.3. Tibia mounting attachment with instrumented rotating base plate.

Figure 3.4. A curve of automatic rotation plotted against knee flexion.

The knee was mounted in the test apparatus, the quadriceps length adjusted to control knee flexion, and a vertical load of 50 lb was applied. One hundred pounds of vertical load could also be applied with the knee in full extension and 20° of flexion. However, with greater degrees of flexion and the magnification of quadriceps force through the increasing flexor lever arm, it became difficult to hold the quadriceps tendon in the fixation clamp. Therefore, 50 lb of vertical load were applied for all degrees of flexion.

Once the knee was stabilized in the loading frame at the preselected degree of flexion, rotational movement was applied to the tibia. The potentiometer to measure degrees of rotation was connected to the x axis of an x-y plotter and torque, using a strain gauge instrumented torque arm, was measured on the y axis. A curve of torque versus degrees of rotation was generated with internal rotation showing to the right of the vertical axis and external rotation to the left (see Fig. 3.5).

Figure 3.5. A family of torque rotation curves for 0°, 20°, 40°, 60°, and 90° of flexion. The vertical mark in each curve signifies the neutral point. It can be seen that the neutral point is shifting to the right, which represents the automatic internal rotation of the knee with flexion. Torque in newton-meters is plotted on the x axis and degree of rotation on the y axis. Internal rotation is to the right of the neutral point, and external rotation to the left. (Reproduced with permission from D. S. Hungerford, B. V. Kenna, and K. A. Krackow: The porous coated anatomic total knee. Orthopaedic Clinics of North America, 13:103,1982.)

All knees were tested in 0°, 20°, 40°, 60° and 90° of flexion. The mechanism of loading, i.e. of applying a vertical load which works through the lever arms of the femur and tibia with flexion controlled by quadriceps tension, simulates normal loading of the knee in flexing and extending under load. The actual forces which are transmitted across the knee increased with increasing flexion so that at 90° of flexion 380 lb of quadriceps force were required to resist 50 lb of applied vertical load. These figures are very similar to the calculated quadriceps load in two-legged stance for the same degree of flexion.


Both with and without an applied load, flexion of the normal knee is associated with rotation of the tibia in relation to the femur. This so-called "screw-home" mechanism consists of internal rotation of the tibia, relative to the femur, with flexion and conversely external rotation with increasing extension. In the testing apparatus it was possible to determine the neutral point of rotation as the low energy point in the torque rotation curves. This allowed quantification of the screw-home mechanism under load. The results for the normal knee are shown in Figure 3.4. Most of this automatic rotation has taken place by 40° of flexion.

It is important to differentiate automatic from elective or active rotation. Automatic rotation occurs passively, without any externally applied torque, as the knee flexes and extends. At any given degree of flexion, torque applied to the tibia will produce rotation of the tibia around an axis which passes in the direction of the tibial shaft and at the approximate level of the tibial spine. The amount of force required to produce rotation, plotted against the degrees of rotation, produces a torque-rotation curve. Figure 3.5 represents a family of curves for a single knee at each of the given degrees of flexion. The neutral point for each curve is marked by a vertical line. Each division of the vertical axis represents 2 newton-meters of torque and each division on the horizontal axis represents 3' of rotation. Beginning with the upper left curve, it is apparent that the neutral point is shifting toward internal rotation of the tibia with increasing flexion. The relationships of the neutral points for each degree of flexion, using the neutral point in full extension as zero, can be seen by comparing the position of the neutral reference points for each curve to the scale at the bottom of the graph.


Figure 3.6. The same data in Figure 3.5 is not plotted against an absolute rotational reference point, starting with the knee in full extension. The solid line represents the automatic internal rotation with the isotonic lines of internal rotation above the solid line, and extrenal rotation below the solid line is shown. The x axis represents the degrees of knee flexion, and the y axis the degrees of rotation. The x· · · · x· · · · x line equals 2 newton-meters of rotatory torque and the 0· · · · 0· · · · 0 line equals 4 newton-meters of torque.

Figure 3.7. A schematic diagram of the distribution of force when standing on both feet.

Each degree of flexion has its own curve. It should also be noted that although the vertical loading remains the same, the joint reaction force is increasing and is, therefore, not the same for each degree of flexion. It is what would be experienced under constant loading of the knee under the influence of body weight, i.e., with increasing flexion, the knee will experience increasing load.

Figure 3.5 demonstrates that in full extension, under 50 lb of vertical loading, the knee is quite stiff. By 20° of flexion, a significant degree of tibial rotation can be accomplished in either direction by applying very low torque. This is also true for all levels of flexion beyond 20°. Also, for all degrees of flexion, external rotation of the tibia relative to the femur is accomplished with lower torque forces and to a higher degree than internal rotation. It is possible that the increasing slopes of the external rotation curves between 40°, 60° and 90° is a function of the increasing load of the knee due to the application of the vertical force through an increasing lever arm; i.e., as the knee flexes and the center of rotation moves away from the point of application of the load, the joint reaction force increases.

With internal rotation of the tibia on the femur, in any degree of flexion, the lateral femoral condyle moves posteriorly on the lateral tibial plateau, while the medial femoral condyle moves anteriorly on the mdeial tibial plateau. The opposite is occurring with external tibial rotation. Because of the 7-10° posterior slope of the plateaus, this also means that with anterior movement the condyle is also moving up the plateau while with posterior movement the condyle is moving down the plateau. It is this phenomenon which imparts a measurable and significant varus shift of the tibia with external tibial rotation and a valgus shift with internal tibial rotation. If this shift is prevented by the construction of the jig, a much stiffer curve will be seen.

Figure 3.6 shows another way to plot the same data. The x-axis represents the degree of knee flexion. The y-axis represents the absolute degrees of tibial rotation beginning with 0° at full extension. The solid line represents the zero energy level of the knee and is the same as Figure 3.4, representing the automatic or "screw-home rotation." External rotational force is required to move the tibial-femoral relationship off this line. Relative internal rotation is above this line and relative external rotation is below it. The · · · · · · x line represents the 2 newton-meter torque isotonic line and the 0· · · · 0 · · · · 0 line, the 4 newton-meter torque line. This means that the actual rotational relationship of the tibia to the femur in any given degree of flexion at any moment will be determined by the sum and direction of all rotatory forces. Within +2 newton-meters torque, there is a wide range of rotational potential, particularly in the critical areas of 20°-60° of knee flexion. Any changes of direction, which always takes place on a firm stable foot-floor contact base, imparts significant rotatory moments at the knee. If this rotatory moment can be dissipated through rotatory motion, then stresses on biological or prosthetic structures will be minimized. From the uniformly high failure rates of rotatively rigid hinge prostheses, it would seem that this rotational potential is also important for reducing the incidence of prosthetic loosening.

Figure 3.8. Radiograph of patient with significant arthritis in two-legged stance.

Figure 3.9. Same patient in single leg stance showing dramatic increase in the deformity.

Figure 3.10. Diagram shows the movement of nter of gravity, S6, proximally and away 'e supporting leg, S7, and its relationship leg as a whole. P = partial weight of the ;upported by the knee, A = mechanical lateral stabilizing forces. (Redrawn with sion from P. G. J. Macuet: Biomechanics Knee, Springer-Verlag, Heidelberg, 1976.

Figure 3.11. Free body diagram of the weight-bearing forces crossing the knee. 0 = the center of the radius of curvature of each condyle, X = the varus lever; y = the lateral lever arm through which the lateral structures of the knee are operating: and R = the resultant joint reaction force in single leg stance. (Redrawn with permission from P. G. J. Maquet: Biomechanics of the Knee, springer-Verlag, Heidelberg, lQBO (4).)

Figure 3.12. Single leg stance in patient with a significant varus deformity showing a lateral opening.


The second area of knee biomechanics which is of importance in total knee arthroplasty concerns an analysis of the static forces passing through the knee during double and single leg stance. When the body weight is borne equally on both feet at rest, or in the double stance phase of gait, the force which passes through the knee is only a fraction of body weight (Fig. . 3.7). Because of the double pilIar effect of both legs on the ground, there is no bending moment around either knee. It is important to realize that it is not under these circumstances that the knee is maximally stressed. In fact, if we look at an x-ray in double leg stance with the feet shoulder~width apart (Fig. 3.8), a normal stance position, we do not see the orientation of the lower extremity as it is in single leg stance (Fig. 3.9), and we may not appreciate the degree of deformity which is caused by such eccentric loading of the knee.


Figure 3.13. Whether the gait is narrow (A) or broadly based (B), the center of gravity under normal circumstances is brought over the foot contacting the ground. (Reproduced with permission from J. B. Saunders et al.: Journal of Bone and Joint Surgery, 35A:543, 1953(6).)

Analyzing the static forces passing through the knee, it is important to appreciate the orientation of the joint line (or the prosthesis) in single leg stance. As body weight passes onto the single leg, the center of gravity moves away from the supporting leg and up (Fig. 3.10). This shift occurs because the weight of the supporting leg is not included in the body mass to be supported by the knee while the suspended leg is included. This new functional center of gravity must be centered over the point of contact of the foot with the ground in order for equilibrium conditions to be satisfied.


Figure 3.14. Time dimensions of the walking cycle. (Reproduced with permission from V. T. Inman et ai.: Human Walking, Williams & Wilkins, Baltimore, 1981(2).)

In order to minimize movement of the body mass from side to side, the foot is brought toward the midline at heel strike as the center of gravity is displaced slightly towards the support side. In man, with upright single leg stance, this orientation is accomplished by the overall valgus orientation of the lower extremity which naturally brings the foot toward the midline. Actually the terms "varus" and "valgus" as applied to the lower extremity are somewhat confusing. As we look from the knee up the femoral shaft, we commonly refer to a valgus orientation of the shaft. As we look from the ankle up the entire leg, we speak of an overall valgus of the lower extremity. However, without changing the realities of orientation, when we look at the tibia from the knee looking toward the ankle, it is clear that normal alignment of the tibia is in mild but definite varus relative to the vertical. This must not be understood as a varus deformity, but the normal orientation of the lower leg. A varus deformity will only exist with a varus orientation of more than 3° to the vertical and a valgus deformity with less than 3°.

In single leg stance, therefore, the leg has a valgus orientation to the vertical and the plumb line from the center of gravity falls medial to the center of the knee (Fig. 3.11). This situation exerts a bending moment on the knee which would tend to open the knee into varus. In standing on one leg at rest with the knee fully extended, the ligaments and capsule are tight, in part because of the "screw-home" mechanism which has "close packed" the joint. These structures resist this bending moment. In the dynamic situation during gait, multiple muscles which cross the joint in the center or to the lateral side of center combine to provide a lateral resistance to opening of the lateral side of the joint due to this medial lever. These include the quadriceps-patellar tendon forces, the lateral gastrocnemius, popliteus, biceps and iliotibial tract tension. The sum of these can be represented as L in Figure 3.11 operating through lever arm Y. T his combination determines the magnitude and direction of the resultant vector of the tibial femoral joint load. With increasingknee varus the medial lever arm increases requiring an increased lateral reaction to prevent the joint from opening. Finally the ability to compensate is overcome so that lateral joint line opening does occur, and all of the joint reaction force is borne solely on the medial compartment (Fig. 3.12).

One of the consequences of this analysis for the diseases for which total joint replacement is done is that a single cane in the opposite hand does much to unload the knee and particularly to reduce the magnitude of the varus bending moment. Maquet (4) has calculated that 17.5 lb of force applied to a cane in the opposite hand will reduce knee loading by 46%. Of course the same will be accomplished for the postoperative total knee patient to provide decreased load and increased stability during the rehabilitation phase. A second consequence is to recognize the importance of fully correcting varus deformity when doing total knee replacement. Failure to do so results in large increases in joint load, an important part of the formula for normalizing the joint reaction force in the normally aligned leg. Therefore, in correcting severe valgus deformity we feel that it is important to lengthen these lateral stabilizers as needed rather than simply transect them (see Chapter 10).

Figure 3.15. Knee motion, and EMG of quadriceps muscle plotted on same time scale. Note that, after heel strike, the knee joint flexes (see knee angle and knee moment). To prevent excessive knee flexion the quadriceps is activated but undergoes elongation. With increasing speed of walking, the quadriceps may be used to prevent excessive knee flexion and to initiate knee extension. (Reproduced with permission from V.T. Inman et al.: Human Walking, Williams & Wilkins, Baltimore, 1981(2).)


The final area of biomechanics particularly relevant to total knee replacement is that of normal gait. This is the main area of function of the knee and the finer details are important not only for assessing the results of total knee replacement, but also for avoiding some of the pitfalls which may adversely affect the quality of the end result.

The first important aspect of normal gait is the overall orientation of the lower extremity. Whether the gait is wide or narrow based, the center of gravity is brought over the point of support during the single leg stance phase of gait (Fig. 3.13) (6). This, of course, would not be necessary if a walker, cane, or crutch were being used in the opposite hand. The narrow base gait is the norm, and the most energy efficient. The side to side deviation of the center of gravity is reduced to approximately 2 cm in each direction toward the support side or a total of 4 cm through the gait cycle involving both legs (2). With a waddling or broad based gait, this lateral displacement will be accentuated requiring greater energy for walking. However, the orientation of the lower extremity to the vertical and to the center of gravity will be the same during the single leg support phase of gait with both kinds of gait. It is this normal orientation which we are attempting to reconstruct during total knee arthroplasty.

Normal gait is divided into two phases: stance phase and swing phase. Walking is differentiated from running in that for walking, at some point in the gait cycle, both feet are on the ground providing a phase of double support. Each stance phase begins and ends with double support. Stance phase can be subdivided into heel strike (HS), foot flat (FF), heel off (HO), and toe off (TO). Figure 3.14 depicts a full cycle (2). The length of the period of double support depends upon the rate of the gait cycle. Stance phase will equal 50% of the cycle plus the length of the double support time. At a normal rate (3-4 mph), for a full cycle (one step with each leg) there will be two double support phases each occupying 10-12% of the gait cycle.

It is necessary that the orthopaedic surgeon and the physical therapist be aware of the normal phasic contraction of the muscles during the gait cycle. Many patients presenting for total knee arthropasty will not have used their muscles properly for years and will require intensive postoperative gait training. We have seen several patients who have complained of instability and giving way of their replaced knee only to find that they were not using the quadriceps power they had at the appropriate point in the gait cycle.

Figure 3.15 shows the relationships for quadriceps function to knee angle, gait cycle, and electromyographic (EMG) activity. Negative work refers to muscle activity as measured by EMG activity during elongation of the muscle. Positive work occurs during muscle shortening.


Figure 3.16. Knee motion and EMG of hamstring muscles plotted on the same time scale. Note that the principal action of the hamstrings is to decelerate the leg at the termination of swing phase. While electrically active, the muscles are being elongated. After the foot is flat on the walking surface, they continue to show electrical activity, presumably to assist in hip extension. (Reproduced with permission from V. T. Inman et al.: Human Walking, Williams & Wilkins, Baltimore, 1981 (2).}

Figure 3.17. Ankle motion and EMG of calf muscles plotted on the same time scale. Note that the calf muscles begin to show electrical activity after the foot is flat after heel strike. However, ankle motion reveals that the foot is undergoing dorsiflexion; thus the muscles are being elongated. At the peak of EMG, the heel begins to rise as the muscles start to plantar flex the ankle. Since the gastrocnemius also spans the knee, the negative work at the ankle can actually be viewed as positive work at the knee between foot fiat and heel rise (compare with (Fig. 3.15). (Reproduced with permission from V. T. Inman et al.: Human Walking, Williams & Wilkins, Baltimore, 1981(2).)

The figure begins with heel strike (HS) and ends with HS for the next step. Quadriceps contraction begins just before heel contact as positive work during the terminal degrees of the swing phase to stabilize the knee for heel contact. Between HS arid FF, the knee flexes 20°. This is a critical phase for attentuating the impact of heel strike. During this period of knee flexion, the quadriceps is doing negative work, i.e., it is lengthening during EMG activity. As the knee straightens again during mid stance, the quadriceps is again doing positive work. The small burst of activity just before and after toe off limits knee flexion in the swing phase. In rapid walking there may also be a small amount of positive work to initiate knee extension during swing phase. However, at normal walking speeds, extension during swing phase is a passive event.

Figure 3.18. Phasic actions of all the major muscle groups crossing the knee. Note that most of the muscles are active at the beginning and end of swing and stance phases. During midstance and midswing, there is minimal muscle activity although this is the period of maximal angular displacement. It seems that the principal action of the muscles is to accelerate and decelerate the angular motions of the leg. (Reproduced with permission from V.T. Inman et al.: Human Walking, Williams & Wilkins, Baltimore, 1981 (2).)

Patients who have walked with a stiff legged gait for a long time preoperatively sometimes have a difficult time relearning the controlled knee flexion between HS and FF. In such circumstances they may stabilize their knee at HS by extending the hip and hyperextending their knee. The quadriceps will not be activated, and therefore, the knee will not be dynamically statilized during the stance phase of gait. In addition, the force of heel strike will not be attentuated by knee flexion and the posterior capsule will be stressed leading to posterior knee pain. In addition, the patient will feel a great sense of functional instability even though the knee may be perfectly normal to ligament testing.

Figure 3.16 depicts the same parameters as Fig. 3.15, but referenced to the hamstring muscles. This group does positive work just after heel strike, working opposite the negative work of the quadriceps to stabilize the knee during the 20° of flexion which occurs between HS and FF. Just at and after toe off, the hamstrings contract to add additional flexion for clearance of the foot during the swing phase. Continuing contraction occurs, but now as negative work controlling the freely swinging knee. It is interesting to note that the hamstring muscles are electromyographically active throughout the swing phase of gait, whereas the quadriceps are silent for the most part.



Figure 3.19. Free body diagram of a patient ascending a step, m1 = quadriceps force; m2 = patellar tendon tension; KJR = knee joint reaction; PFJR = patellofemoral joint reaction; CG = center of gravity; x = flexor lever arm.


Figure 3.20. A free body diagram of an individual descending a step. Note the significant increase in knee flexion required and also the change in orientation of the tibial shaft to the vertical. CG = center of gravity; arrow indicates anterior subluxation thrust of the femur.

The triceps surae is usually thought of in reference to ankle motion. Not surprisingly they are active from foot flat to toe off (Fig. 3.17). However, since the gastrocnemius spans the knee joint, it produces both ankle plantar flexion, and also dynamically stabilizes the knee during the stance phase of gait. Compare Fig. 3.17 with the knee angle and gait aspects of Figs. 3.15 and 3.16.

The EMG activity of the muscles of the lower extremity throughout one full cycle is summarized in Fig. 3.18. Intensity of EMG activity is correlated with color intensity. It can be seen that most muscles are active at the beginning and end of stance phase during deceleration and weight transfer, and silent during mid stance and swing phase during maximum movement change.

Much of physical therapy following total knee replacement is oriented toward quadriceps rehabilitation and regaining range of movement. However, only 70° of flexion is necessary to walk with a normal gait. Substantial effort should also be directed toward hamstring and gastrocnemius rehabilitation as well as the proper phasic activities of these muscles in the gait cycle.


Certain aspects of going up and down stairs deserve special mention since they pose specific problems for patients with arthritis and with total knee replacements. Ascending stairs is always easier than descending for several reasons. The actual degree of knee flexion required to ascend stairs is determined not only by the height of the step, but also by the height of the patient (Fig. 3.19). For the standard 7" step approximately 65° of flexion will be required. This is no more flexion than required for normal walking. However, the 70° of flexion in walking occurs during the unloaded swing phase, whereas the leg is loaded in stair climbing. The patellofemoral joint reaction force in level walking approximates one-half body weight, whereas in stair climbing it rises to approximately 3.8 times body weight (5). This comes about because the body weight is acting through a long lever arm (Fig. 3.19). In stair climbing this lever arm can be reduced by leaning forward. Also, in stair climbing the tibia is maintained relatively vertical, which diminishes the anterior subluxation potential of the femur on the tibia.

Descending steps imposes a completely different set of circumstances. To descend the standard step which requires 65° of flexion to ascend, 85° of flexion is required (Fig. 3.20). In addition, the tibia is steeply inclined toward the horizontal, bringing the tibial plateaus into an oblique orientation. The force of body weight will now tend to sublux the femur anteriorly. This anterior subluxation potential will be resisted by the patellofemoral joint reaction force, and the tension which develops in the posterior cruciate ligament. In the absence of a posterior cruciate ligament, only the collateral ligaments are available to assist the patellofemoral joint reaction force in providing anterior-posterior stability. The collaterals are not well oriented to provide this assistance, and in total knee replacement would require perfect balancing to be of any usefulness. The worst set of circumstances arises in the patient with an unreplaced and/or painful patellofemoral joint and an absent posterior cruciate ligament. Also the post-patellectomy patient will suffer from diminished anterior posterior dynamic stability. In addition, the useful adjunctive technique of leaning forward to reduce the flexer lever arm is difficult to apply when ascending stairs.

Many patients with arthritis will report difficulty descending stairs normally. For some, this will also be true after total knee replacement. A simple remedy is to have them descend either sideways or backward, which is biomechanically the equivalent of ascending the stairs with its decreased mechanical and range of motion demands.


The various aspects of biomechanics of the knee which are of direct importance in total knee arthroplasty have been reviewed. Through new work, rotatory stability has been analyzed and quantified. The importance of rotational freedom and stability to normal knee function has been emphasized. Analysis of the static forces crossing the knee in single leg stance demonstrate the importance of proper alignment as well as the role of the lateral stabilizers. Finally, analysis of gait in level walking and stair climbing demonstrates the importance of the complicated phasic activity of the lower extremity muscles in providing stable, active knee function. Awareness of these factors will assist orthopaedic surgeons performing total knee arthroplasty in identifying and avoiding certain pitfalls in the rehabilitation of their patients.


  1. Hsieh HH, Walker PS: Stabilizing mechanisms of the loaded and unloaded knee joint. J Bone Joint Surg 58A:87, 1976.
  2. Inman VT, Rolston HJ, Todd F: Human Walking. Baltimore, Williams & Wilkins, 1981.
  3. Kapandji IA: The Physiology of the Joints, Vol.11. New York, Churchill Livingstone, 1970.
  4. Maquet PGJ: Biomechanics of the Knee. Heidelberg, Springer-Verlag, 1976.
  5. Reilly DT, Martens Mt Experimental analysis of quadriceps muscle force and patello-femoral joint reaction force for various activities. Acta Orthop Scanal, 43:126,1972.
  6. Saunders JB, Inman VT, Everhart HD: The major determinants in normal and pathological gait. J Bone Joint Surg 35A:543, 1953.