Knee valgus with internal valgus moment

[Paulo Lucareli]

Dear all,

I have just completed a survey of patients with knee pain during triple jump and the result shows that during landing, people have peak knee abduction associated to the external adductor moment and power absorption. I've reviewed the data several times and confirmed the result. However , some of my colleagues are questioning the possibility of having abduction knee simultaneously with the medial GRF.

Is my impression that the function of the external varus moment of the knee is to halt the valgus movement, particularly hip adduction and internal rotation, together with the before mentioned reasons (trunk lean and weak abductor muscles). We firmly believe that this data differs from that which was expected. However, we must take into account the fact that the activity studied involved a high impact and high speed. The contact between the feet and the ground was minimal and this did not allow movements to occur with a range of motion that is typical of the traditional dynamic valgus pattern of movement.

I'd like to have more opinion on this topic.

Thanks.

[Ton van den Bogert]

Re: Knee valgus with internal valgus moment

Paulo,

I agree that this result is opposite to what is expected, so you are right to scrutinize the results.

What was the magnitude (in degrees) of the abduction you found? Abduction is usually a much smaller motion than flexion, and you can easily get "kinematic crosstalk" if the first axis of the joint coordinate system (JCS) is not well aligned with the functional flexion axis. This will cause some of the flexion to appear in the ab-abduction motion. A good reference is: Ramakrishnan & Kadaba, J Biomech 1991 (http://www.ncbi.nlm.nih.gov/pubmed/1744154). Figure 2b shows the problem very nicely. You could do a similar sensitivity analysis to see if a slightly different JCS axis would cause the abduction to disappear or even turn into adduction.

If you have double checked your inverse dynamic calculations, I would trust the joint moment more than the joint angle, in this case.

Ton van den Bogert

[Paulo Lucareli]

Re: Knee valgus with internal valgus moment

Dear Ton,

Thanks for your comments and paper.

The abduction is 7.5 +/- 2.5 degrees. Peak knee flexion is 55 +/- 4.9 degrees. The knee abductor moment is 2.1 +/- 0.4. I understood and agree with you comment, but I found a reference: Powers, CM JOSPT 2010 (http://www.ncbi.nlm.nih.gov/pubmed/20118526). Figure 2B show the possibility of knee pain patients move the center of mass away from the stance limbincreases the distance from the resultantground reaction force vector and the kneejoint center, thereby increasing the varusmoment at the knee.

I attached a video and a "skeleton" representing my data to you to see the "problem". https://drive.google.com/open?id=0Bz...bGc&authuser=0


All the best.

[Ton van den Bogert]

Re: Knee valgus with internal valgus moment

Paolo,

Thanks for the nice video!

Your GRF vector is medial to the knee, so this indeed suggests an external varus moment. The mass and acceleration of the shank and foot also play a role, but this would probably not change the direction of the moment.

It is hard to imagine a valgus motion happening at the same time as a varus load being applied. This would require a large active internal valgus moment (lateral hamstrings). I don't think that this is possible. We usually think of the internal varus-valgus moment as being passive and therefore opposite to the varus-valgus motion.

I still don't quite trust your abduction angle result. If the 7.5 degrees is real, this indicates that the medial compartment of the knee loses contact and the MCL elongates by about 50 x sin(7.5) = 6.5 mm or more.

As I mentioned earlier, there can be an error from crosstalk (55 degrees flexion is a lot, and can easily cause a false abduction reading). Also you use a marker cluster on the thigh, so there is a skin movement error. This error can be about 10 degrees (http://www.ncbi.nlm.nih.gov/pubmed/9239553) but that was for running and a different marker set. In that case the error would have increased the varus angle rather than the valgus angle. You have a different movement and a different marker set, so it is hard to guess what the effect of skin movement would be. I only know that it could easily be 10 degrees error.

Ton

[Chris Kirtley]

Re: Knee valgus with internal valgus moment

This is a great topic for discussion. I've been out of gait analysis for a few years, and I'd be interested to hear if there have been any new solutions to the problem.

As I experienced it, the accuracy of the old Kadaba (VCM/Vicon/Gage/Conventional) model is highly dependent on the marker on the lateral femoral condyle. It takes some degree of experience nd luck to align this with the sagittal (clinical) knee axis. I confess I never felt completely confident and it was only when the results came in that I was reassured that it was OK. As far as I know there's still better technique than skilled guessing. Initially Vicon encouraged the use of the Knee Alignment Device (KAD) but I don't think it really caught on. Richard Baker came up with a software version DynaKAD <http://clinicalgaitanalysis.com/bodybuilder> which is ingenious, but again as far as I know is little used in practice.

Other marker sets/models use different approches. The Cleveland model (originally associated with Motion Analysis Corp) used a wand to manually define the knee axis. I never used it myself but observed it in use and was impressed by the results. Then there are the 6DoF models by NIH (C-Motion). I never quite understood how the axis is defined in these.

I suspect this is one of the great unanswerable questions in biomechanics, and would encourage everybody to offer a solution!

Dear all,

I have just completed a survey of patients with knee pain during triple jump and the result shows that during landing, people have peak knee abduction associated to the external adductor moment and power absorption. I've reviewed the data several times and confirmed the result. However , some of my colleagues are questioning the possibility of having abduction knee simultaneously with the medial GRF.

Is my impression that the function of the external varus moment of the knee is to halt the valgus movement, particularly hip adduction and internal rotation, together with the before mentioned reasons (trunk lean and weak abductor muscles). We firmly believe that this data differs from that which was expected. However, we must take into account the fact that the activity studied involved a high impact and high speed. The contact between the feet and the ground was minimal and this did not allow movements to occur with a range of motion that is typical of the traditional dynamic valgus pattern of movement.

I'd like to have more opinion on this topic.

Thanks.

[Ross Miller]

Re: Knee valgus with internal valgus moment

The 6DOF models for the thigh/shank/knee as I recall are based on directly aligning the long axis of the segment to pass through the two joint centers:

- Long axis of segment (k) is defined from knee to hip (thigh) and ankle to knee (shank)
- Frontal axis (j) is the cross product of k and the axis between the femoral condyles
- Sagittal axis (i) is j x k

So the femoral markers will affect alignment of all three axes, but I think their axis is not necessarily parallel to either segment's sagittal axis.

Ross

[Paulo Lucareli]

Re: Knee valgus with internal valgus moment

Ton,

The values that a mentioned before is not the ROM during the landing. It is the peak knee abduction angle. The mean range knee abdution angle is 2 degrees.

Do you still believe that it is came from crosstalking? Despite the cluster in the thigh I processed the data using plug in gait model.

Thanks for all comments. This thread has helped my research group.

[Ryan Mahaffey]

Re: Knee valgus with internal valgus moment

No solution, only a comment...

I used Richard Baker's DynaKAD in my thesis on paediatric gait analysis. Repeatability of knee joint frontal plane and hip transverse plane motion was improved using DynaKAD against using no thigh marker correction. I did find larger transverse plane range of motion at the ankle joint using DynaKAD (crosstalk?) and wondered if lateral shank marker position would benefit from a similar correction procedure. However, whilst lateral thigh marker position is based on minimising knee varus/valgus angles I'm not sure what lateral shank marker position could be based on?

I should note that I don't use DynaKAD for gait analysis but it was a great tool to understand thigh marker placement issues, which (I think) was the aim from Richard Baker's article (http://www.sciencedirect.com/science...67945799000275)

Ryan

This is a great topic for discussion. I've been out of gait analysis for a few years, and I'd be interested to hear if there have been any new solutions to the problem.

As I experienced it, the accuracy of the old Kadaba (VCM/Vicon/Gage/Conventional) model is highly dependent on the marker on the lateral femoral condyle. It takes some degree of experience nd luck to align this with the sagittal (clinical) knee axis. I confess I never felt completely confident and it was only when the results came in that I was reassured that it was OK. As far as I know there's still better technique than skilled guessing. Initially Vicon encouraged the use of the Knee Alignment Device (KAD) but I don't think it really caught on. Richard Baker came up with a software version DynaKAD <http://clinicalgaitanalysis.com/bodybuilder> which is ingenious, but again as far as I know is little used in practice.

Other marker sets/models use different approches. The Cleveland model (originally associated with Motion Analysis Corp) used a wand to manually define the knee axis. I never used it myself but observed it in use and was impressed by the results. Then there are the 6DoF models by NIH (C-Motion). I never quite understood how the axis is defined in these.

I suspect this is one of the great unanswerable questions in biomechanics, and would encourage everybody to offer a solution!

[Paulo Lucareli]

Re: Knee valgus with internal valgus moment

Hi Rian,

I've used the same procedure (DynaKAD) with patella marker instead of thigh marker with good results also.

Paulo

[Ton van den Bogert]

Re: Knee valgus with internal valgus moment

Not sure if I understand you correctly, but I think you are saying that the angle only changes by 2 degrees during the landing. The 7.5 degrees is then mainly a DC offset.

2 degrees is small enough that I would not worry about it being in the wrong direction. An abduction motion of 2 degrees is too small to be measured reliably during impact. Crosstalk and skin movement error can easily be much larger than 2 degrees. There is no way to tell which of these two errors was responsible. Well, maybe one clue: if the abduction angle has the same pattern as flexion, I would suspect crosstalk.

Ton

[Zahra Safaeepour]

Re: Knee valgus with internal valgus moment

Hi Ryan

I tried to use the DynaKAd.mod on paediatric subjects but I almost can not run the model in Nexus 2.4. I was wondering if you could explain how you applied it.

Thanks,
Zahra

[Allan Carman]

Re: Knee valgus with internal valgus moment

I would agree with Ton in that the mean 7.5 + 2.5 degs of abduction is too large during stance and beyond expected mechanical limits of the knee joint. At first thought it could be the result of an internally rotated thigh medio-lateral axes from the true anatomical knee flex/ext axis (> 10 degs of internal rotation on average?) producing excessive knee abduction. However the magnitude of non-linear error (x-talk) produced will be dependent on knee flexion angle and would produce a large knee abd/add range of motion. If however you have a small abd/add ROM and it appears more like an offset then it is more likely a mis-alignment of the shank segment into abduction rather than cross talk from mis-alignment of the thigh segment. The rigid cluster on the quadriceps has the potential to exacerbate the effect of skin movement error on axes misalignment. As will the use of what appears to be a wand in the lower leg, particularly during impact? I prefer to have markers placed directly on the skin.
The skeletal motion depicting the lower leg alignment and knee position in the sagittal plane does not represent what I see in the video of the same subject or with the knee abduction you describe. The subject’s knee appears to track over the foot during weight bearing; this is in contrast to the laterally directed lower leg alignment and lateral knee joint position depicted in the skeleton. Which, if what I am seeing is correct, will produce large and erroneous adduction knee moments if used for kinematics or kinetics. I would therefore not trust the external adduction knee moments you describe based on this skeletal representation. The distorted leg segments in the graphical representation could be the result of the model used; constrained 3DoF joints or a 1DoF knee joint does not allow for inaccuracies in defining segment axes or skin movement error and could produce this distortion in segment alignment.
On weight bearing in normal gait it is typical to see 2-3 degrees knee adduction, internal rotation of the lower leg and pronation of the foot. I would presume these rotations are linked and mechanical in nature, but with these small knee adduction rotations it is still debatable to whether it is real or error (as Ton mentioned). Although my experience is based around gait I would still expect something similar in landing.

[Ryan Mahaffey]

Re: Knee valgus with internal valgus moment

Hi Zahra,

I applied DynaKAD in workstation. I have not tried in Nexus

Regards

Ryan

[Wadim Korneew]

Re: Knee valgus with internal valgus moment

I agree that this result is opposite to what is expected, so you are right to scrutinize the results.