One of the most common problems at the knee is what has loosely been described as anterior knee pain. This term encompasses a huge range of syndromes and is most effectively treated conservatively.

As Insall (1981) and others have shown the muscular factors involved with patello-femoral joint dysfunction are of considerable importance. Quite why conditioning the quadriceps alleviates pain is not entirely known but its efficacy is unquestionable. Many authors have stressed the importance of an imbalance between the moments generated by the vastus medialis obliquus and vastus lateralis. Unfortunately there remains no reliable way of testing the balance between the VM and VL. EMG could be considered to be semi-quantitative and there are thousands of studies examining the EMG ratio between VL and VM.

Patello Femoral Joint (PFJ) & Isokinetics

When we test the quadriceps isokinetically in patients with PFJ dysfunction there is a definite dependence on the test velocity. This is evident in both the strength scores and in the shape of the MAP curve (enhanced in the lower testing velocities).

Testing for PFJ dysfunction was marred in the early years by researchers such as Elton et al (1985) and Macyntyre et al (1988) as they carried out tests which revealed no significant differences in either concentric or eccentric strength.

Unfortunately both studies involved the use of high angular velocities (180 and 200 degrees/second respectively). At high speed the joint is exposed for a much shorter time to the force which in turn creates a lower load on the joint and hence reduces potential inhibition. At high speeds it could also be argued that the reflex arc may be too slow to inhibit the quadriceps.

Tests performed at the lower velocities were different. Hoke et al (1983) used a test velocity of 30 degrees/second and showed that quadriceps strength and curves were significantly different to the unaffected side. Nordgren et al (1983) also performed tests at low velocities (6, 12 and 60 degrees/second). They found that both men and women demonstrated high reductions in quadriceps strength (22% in men and 34% in women). These findings were further validated by Dvir et al (1990) who found strength reductions of 27% in men and 35% in women concentrically and 44% in women and 41% in men eccentrically. If these findings had been made in 1985 by Elton et al it could be speculated that validation of isokinetic testing over arthroscopy for the diagnosis of chondro malacia patellae would have been made.

Testing at low velocities should be restricted to sets of no more than 6 repetitions performed maximally and reassessed every 2-3 weeks otherwise further symptoms may be invoked by testing.

Pain and Load

Many studies (including Dvir et al, 1991a) have shown that variations in pain score do not correspond with variations in strength.

Pain and load, however, showed strong correlations (load is the product of the magnitude of the moment and the period of time of exposure). Eccentric contractions demonstrate this more readily than concentric contractions. There are two reasons often cited for this.

Higher moment is generated during eccentric contraction.

Patella kinematics

(during concentric contraction the contact area within the patello-femoral notch diminishes yet with eccentric extension the knee flexes and hence the contact area becomes larger. This may be seen as a misnomer as with increased contact comes improved pressure distribution).

The Moment Curve

Basically a perfect moment curve from an unaffected joint muscle unit with good neuromuscular facilitation should have a curve which looks like an inverted 'U' seen here.

If there is a sudden shut off of the contractile activity this curve assumes an irregular shape. Hart et al. (1985) has shown a common oscillatory phenomenon. In our example we refer to a single (but this could even be a double, triple etc.) conspicuous peak in the curve.

 The Break Phenomenon

This is a widely disputed expression of PFJ dysfunction. It was first observed by Nordgren et al. (1983) who took 3 different moment curves on the same patients (preoperatively, under intra-articular anesthesia and during rehabilitation). The first real analysis of the phenomenon was done by Dvir et al. (1991b). They described the break phenomenon as a perturbation in the curve which exceeded a drop of 10% or more (an example of a break curve appears below).

In our example M = the maximum pre break force and L = the lowest recorded force during break. A i = the angle of break and P = the angle of lowest force.

Breaks occur exclusively during eccentric contractions and seem to be associated with a relief of pain at the same moment. This tends to be at about 45 degrees of flexion (reported by Dvir et al. 1991a and Hart et al. 1985). Interestingly, 75 degrees has been calculated to be the point of maximal PFJ reaction force.

Concentric loads at 30 degrees/second are actually greater than eccentric loads at 60 degrees/second but concentric loads do not demonstrate breaks. We could speculate that the load is not responsible of the break phenomenon which could only leave the velocity of movement responsible.

Most breaks can be seen at 30 degrees/second. Lower test velocities than this tend to have exceedingly long exposures and hence the velocity relationship is more constant masking the break.

PFJ Dysfunction Assessment Model

Use of the isokinetic machine to diagnose anterior knee pain as PFJ dysfunction has been the subject of a lot of research (most of which still needs validation). Dvir and Halpirin (1992) looked at assigning a positive and negative value to certain symptoms according to certain criteria.

Every subject was classified according to the following triad.

Strength +/-

Pain +/-

Break +/-

This would give a 3 point description e.g. '++-'. This would mean that the subject demonstrated significant strength reduction (over 20%), pain (modified Borg scale over 0) but no break phenomenon.

A relatively high proportion of the patients tested (27%) have been found to classify as negative on all factors and 24% of subjects were classified as positive on all factors. 40% classified as 2 positives and one negative the rest classified as 2 negatives and a positive (or unable to complete the test!).

The 40% value was found to reflect the accuracy of all other clinical tests. This would suggest that no other accepted test for diagnosing anterior knee pain as PFJ dysfunction is any more accurate than the isokinetic test (if in doubt test isokinetically).

Conclusion

Isokinetic tests can be used to diagnose PFJ dysfunction and monitor treatment progression. It also offers a quantitative outcome measure.