History

Until the late 1970s 75% of all isokinetic use and research was based on a single joint system - the knee. With more recent progress in rehabilitation and knee surgery this trend no longer exists. The basic design of isokinetic dynamometers (except for special purpose units) has not changed since the original instrumentation became available in the 1960s. The design is still better suited for knee testing and rehabilitation than any other joint (Dvir 1995). Although the knee has 2 major articulations the relevant one in this section is the tibio-femoral component.

The first major consideration when testing the knee is:

To use open or closed chain movement.

Once we have decided on a type of exercise there are four major considerations we need to address. These are:-

Alignment

Setting up the machine to get the subject in roughly the right position is only the first part of the task. Do not be afraid to move the seat or dynamometer to allow for better alignment of the biological axis of rotation. At the knee this changes throughout range (so we use a compromise position). Dvir (1995) tells us this extends through the lateral femoral epicondyle (although I believe alignment with the lateral joint line slightly anteriorly to the centre point generally offers better alignment throughout range see below). To check the alignment simply straighten and bend the knee and make sure that the attachment on the calf does not move up and down the shin (this can cause friction burns and does not allow the knee to rotate correctly).

TIPS:

If you find it hard to set the alignment correctly with the knee bent try doing it with it straight this often helps.

Subjects with limited extension will often lift their thigh from the seat as they reach terminal extension setting the axis of rotation slightly too far forwards (towards the patella) can help overcome this.

Small errors in alignment can be compensated for by the subject i.e. if you test the alignment and find the subject moves their body in the chair slightly continue to bend and straighten the knee until they stop adjusting and the alignment will be correct.

 

Here the lateral femoral epicondyle is shown as the red star. The actuator axis (the blue star on the machine and the centre of the lever arm's rotation) should be opposite the later femoral epicondyle or as I prefer the lateral joint line (the blue star on the subject)

Positioning and stabilization

Testing and exercise are generally performed in the sitting position although absolute testing is best performed in the prone lying position as this allows a greater range of motion and functional testing is best performed in the standing position.

Seated testing assumes that minimal femoral motion will occur as the chair and body act as distal stabilisers of the thigh. The subject usually sits with their back and thighs supported making approximately a right angle at the hip. The thigh support should extend to allow the appropriate amount of knee flexion. In most tests this would be somewhere towards the distal third of the thigh which would allow 75-90 degrees of flexion (the maximum knee flexion I have seen tested was 110 degrees whilst retaining reproducibility). This position then allows maximal extension (although debate rages over whether extension beyond -20 degrees should be permitted. Personally I would not test beyond 0 degrees extension as an absolute maximum, whilst subjects tend to find limitations beyond 5 degrees as irritating and tend to do large isometric contractions to try to complete the range). Although the angle of seat recline (from the semi-reclined to the upright position, i.e. 40-90 degrees), has little effect on quadriceps strength it has significant effects on hamstring strength. The optimal position is approximately 80 degrees (with a corresponding change in seat angle recline to give 90 degrees at the hips). This optimal position is suggested for both extensors and flexors as it allows the collection of good data over the least time.

Seated stabilization is normally accomplished using femoral and pelvic strapping, however, the optimal set-up is a bit more involved. The number of research papers available on the subject is incredible. Magnusson et al (1992) showed that stabilization with a thoracic strap and the hands was associated with the highest quadriceps strength whilst no stabilization produced the lowest score. Hart et al (1984) also showed the use of a thoracic strap to improve quadriceps strength, whilst Hanton and Ramberg (1988) found exactly the opposite. Use of a thoracic, pelvic and femoral strap decreased quadriceps strength when compared to minimal stabilization i.e. only gripping the sides of the testing table.

Interestingly, Currier (1977) whilst testing isometric strength found that gripping of the table increased strength whilst gripping of handles did not show such significant improvements. These results were elaborated upon by Bohannon (1986) when he tested various gripping devices compared to only gripping the table and found massive differences. Hence, most isokinetic dynamometers do not offer hand grips as an option.

Supine lying (laying on the back) should be used only if assessment or training are directed towards rectus femoris specifically. Stabilization should be accomplished by using the femoral and waist straps (authors recommendation). This position has become more popular as electromyography (EMG) is used to look at the role of vastus medialis and vatus lateralis in the positioning of the patella through range. Supine testing is more functional than seated testing but it is not as functional as performing the test in standing.

Prone lying allows for a much broader range of motion to be assessed. This position is generally used if the hamstring muscles are of particular interest (as stabilization of the knee flexion movement is easily achieved in this position). Stabilization is accomplished by allowing the subject to hold the seat edges and a femoral and waist strap should be applied. On the Cybex norm the seat does not lock into the down position this means that as the subject works the seat will raise up and crash back down. This can be prevented in the short term by wrapping the waist belt around both the patient and the chair. This appears to be an oversight on this machine.

Standing is a poor position for knee testing, however, it is a more functional position (stabilization is almost impossible and I would think undesirable as the whole point of using this position is because it is more functional).

TIPS:

Place two of your fingers behind the the subjects knee to make sure they are not too close to the chair with their popliteal vessels. Not can it be uncomfortable if the knee is squashed to the chair it also limits range of motion.

Position of the resistance pad

Whilst testing normal subjects (not usually patients!) the resistance pad is placed on a level with the inferior part of the pad immediately superior to the medial malleolus shown here (in other words the bottom of the pad touches the top of the medial malleolus seen as the red star).

This is because 70% of all subjects tested by Kramer et al (1989) found this the most comfortable with the other 30% preferring a position at two-thirds of the usable leg length (after you have spent an hour determining the usable leg length and then calculating a position two-thirds down this, your subject will be so fed up the results will be negated by poor subject motivation).

When using any selected location the subject should be free to maximally dorsiflex the ankle (as seen below). 

Close attention should be paid to not over tightening the strap around the shank as the resistance pad will, in all subjects, slide up and down the leg to some degree (this is because of the change in joint axis through range).

Siewert et al (1975) showed that the strength of both the extensors and flexors become successively smaller as the resistance pad is placed near the knee. This trend was established at all test velocities. Taylor and Casey (1986) have suggested that the reason for this phenomenon was increased intra muscular pressure which causes further divergence of the knee axis away from the actuators axis (or in other words the axis of knee rotation becomes greater which means that the axis of rotation you set at the machine must be further away from it). For every 1cm change an alteration of up to 5% in the values recorded can be expected.

These findings were supported by Kramer et al (1989) however, it is probably not that simple. Moving the resistance cell nearer the knee also shortens the dynamometer application arm and increases the angle between the arm and shank which when coupled with changes in neurophysiological inhibitory mechanisms, discomfort and pain all contribute to a general reduction in muscular strength. 

Consistency in the position of the resistance pads is, therefore, crucial.

Alignment

As open chain exercise is multi joint no specific recommendations can be made except during the leg press exercise care should be taken to ensure that the lever arm does not pass beyond the horizontal as this fundamentally alters the recruitment of the muscles around the hip.

Positioning and stabilization

The seat base should be angled to at least 10 degrees as this prevents the subject ‘submarining’ (sliding out of the chair from under the stabilizing belts) out of the chair, however, this is not appropriate for reciprocal motions (such as cycling) which do not present this problem. The seat back should be angled between 15 and 45 degrees to prevent the very high (often over 5 times body weight) forces encountered from moving the subject rather than the load cell.

The subject should be encouraged to hold the chair (and advised to avoid the Valsalva maneuver (MacDougall et al 1985)). Chest straps should be used but waist straps lead to increased intra abdominal pressure which may not be desirable (MacDougall et al 1985).

It has been argued that there is a need to reduce the anterior force of the pull of the quadriceps in patients with ACL deficiency. A dual pad including a distal and proximal part called the ‘anti-shear device’ was developed by Johnson (1982). A revision of this became available when Brown et al (1992) made adjustments. The use of this accessory pad has been validated by Timm (1985) and its incorporation in rehabilitation and testing has been strongly recommended by Dvir (1995).

As I have tested many ACL deficient and recently reconstructed patients with isokinetic open chain exercise without this expensive accessory, and had no problems, I do not see any good reason for recommending it's use. I must hasten to add that I test several ACL deficient/reconstructed patients daily!

The range of angular velocities used to test the hamstrings and the quadriceps is extensive. Borges (1989) chose an extremely low value of 12 degrees/second for one of the criterion velocities, whilst at the other end of the spectrum Ghena et al (1991) and Hall and Roofner (1991) tested subjects at velocities as high as 500 degrees/second. It is debatable whether the use of high velocities in knee testing gives significant data for interpretation. A high velocity at the knee is considered to be above 180 degrees/second. Some studies (Ghena et al (1991) being the most significant) have demonstrated only very small strength differences above 300 degrees/second at the knee. The greatest change in muscular strength tends to occur between 30 degrees/second and 120 degrees/second.

The findings of Hall and Roofner (1991) have revealed a moment angular velocity curve which may be easily extrapolated to give predictions of strength values at high values for most normal subjects. It would seem then that testing at very high velocities would provide no useful information to the clinician. However, there may be good reason to test and train at high speed for muscle performance for professional athletes. In fact muscle conditioning at velocities around 450 degrees/second may still constitute a genuine stimulus to the muscle, as has been recommended by Mangine and Noyes (1992).

Recommended range of test angular velocities

Any speed between 60 degrees/second and 180 degrees/second would generally meet most requirements for validity and the need for information about muscle performance. Between these ranges the subject tends to be comfortable and finds the movement reasonably easy to cope with. An added benefit is the very wide usage of these speeds in hundreds of studies. Very low and very high velocities are often contraindicated in most patients unless the purpose of the test is to provoke a specific reaction (testing at speeds outside the range of 60-180 degrees/second should be reserved only for professional athletes or very experienced clinicians).

Unfortunately the establishment of normal muscle strength values at the knee joint (and hence every other joint) is not yet complete. A data base this large would require enough subjects who share a number of similar traits (gender, age, activity level, fiber types, health status, anthropometric factors). The same protocols would have to be used for each subject (contraction type, velocity, testing procedures, measurement device, peak moment, average moment etc.). Given the amount of variables it would seem impossible to provide a dependable normative framework. Yet, here  we have one (from Neder et al. 1999).

PT = Peak torque W + Work TW + Total work AP + Average power TAE = Total Acceleration energy

Neder et al (1999) performed a prospective controlled randomized study to establish reference values for the prediction of concentric isokinetic knee strength and power in a sample of non athletic men and women aged 20-80 years. They tested at 60 degrees/second and 300 degrees/second.

Knee Treatment

Comparatively little research exists on treatment of knee problems compared to knee testing. The first thing to establish is whether the patient has normal muscle strength values in the quadriceps and hamstrings. A view of the values used in our department can be seen here.

Average peak torque (Nm) to body weight (Kg) at 60 degrees/second over a 75 degree ROM (0-75) seated with thigh stabilization.

If the patient does not have normal muscle strength values then these should be returned using the appropriate isotonic exercises. If this cannot be achieved then isokinetic treatment should be considered using the appropriate protocol.