The Paradox of Patellofemoral Pain Syndrome and Foot Biomechanics

Probably the most common injury seen in runners is “runners knee” – or patellofemoral pain syndrome (PFPS) or anterior knee pain or what it used to be called a long time ago, chondromalacia patallae. Typically the pain is around the patella and is aggravated by activity. The mechanism of the injury is overuse, but there is debate about what underpins that. It is probably some malalignment issues resulting in patella maltracking. Where the debate is, is if the cause of those malalignment problems come the foot or from the more proximal structures like the hip. There is evidence and arguments for both with no consensus as to which one it is. To blanket argue that one is more important that the other is nonsensical as there are good randomized controlled trials that support both.

Foot orthoses are widely used to treat patellofemoral pain syndrome and two RCT’s support their use (Collins et al & Eng et al), so there is no doubt they work. The model that the use of foot orthoses is based on is that the foot pronates excessively, so the tibia internally rotates more than it normally would, which internally rotates the knee causing the maltracking of the patella and the symptoms of┬ápatellofemoral pain syndrome. Foot orthoses are widely assumed to change that by preventing the pronation of the foot (and the RCT’s certainly show that when you try and do that, they do work clinically). The problem with this widely believed model is that the majority of the cross-sectional studies and all of the prospective studies show that there is no link between excessive pronation of the foot and patellofemoral pain (despite foot orthotics to treat excessive pronation actually working).

To further complicate it, the work of Bellchamber shows that it is the tibia that controls foot pronation and supination, not the other way around (however, the results showed a lot of intersubject variability and was not on people with patellofemoral pain syndrome), so this would suggest that the control comes from above and not below.

Adding to this, the proximal control group in the study by Collins et al did just as well as the foot orthotic group and the study by Coppack et al showed that patellofemoral pain can be prevented by proximal control exercises.

So, the paradox is that foot orthotics are widely used to treat PFPS and the RCT’s and our clinical experience tells it that they work; however, the model that the use of foot orthotics are based on is not right and not supported by the evidence. So why do foot orthotics work in PFPS and why do people with PFPS get better with proximal control exercises?

If you are a clinician and have a patient with PFPS, what should you do (other than the direct treatment of the injury)?; should you use foot orthotics (which the evidence supports; despite the flawed model they are based on) or should you use the proximal control exercises (which the evidence also supports). Unfortunately there is NO evidence to guide this, just the usual lot of rhetoric and propaganda with a lot of the usual argumentative fallacies, so what does one do in the absence of that evidence?

My proposed solution to that paradox is to consider a model (that I freely admits has no evidence to back it), but is totally consistent with the available evidence and is theoretically coherent and biologically plausible. Unfortunately we need these types of models as clinicians need to make decisions today. Clinical practice is based on ‘models’ or ‘frameworks’ and there is nothing wrong with that. However, those ‘models’ or ‘frameworks’ have to be consistent with all the available evidence.

Lets do a thought experiment: Lets assume for a moment, that yes it is patella tracking is the problem and lets also assume for a moment that proximal control is very important in PFPS, so exercises are done on the proximal structures to improve the alignment and control of the lower limb, but lets assume that the foot is rigid and does not move. If the foot does not move, then there is no way what you do above is going to work at changing the alignment or control of the lower limb.

Lets assume another situation in which the foot moves easily and the force needed to supinate the foot is really low. It will not take much of a change in the control from the more proximal structures to move the foot and change the alignment of the lower limb and help the PFPS.

So if the force needed from the proximal structures to supinate the foot varies from person to person along a continuum (ie from rigid and not moving to easily moved), then there is hypothetically a threshold for each individual in which the proximal control is not going to work as the force needed to supinate the foot (supination resistance) is too high. Many studies have demonstrated that one of the mechanical effects of foot orthoses is a reduction in the rearfoot inversion moment (ie its easier to supinate the foot). So, maybe the role of foot orthotics in PFPS is to lower that rearfoot moment, so that the proximal structures can control the motion of the lower limb.

If this was the case, then it would explain:

  • Why foot orthotics work in PFPS (ie to lower the rearfoot inversion moment and not necessarily change alignment)
  • Why the traditional model that the use of foot orthotics in PFPS was based on, while wrong still actually worked if you applied it (ie the rearfoot moment was reduced, and not necessarily changed alignment)
  • Why proximal control works on its own in many (ie maybe the force to supinate the foot was low in those people)
  • It is consistent with the work of Bellchamber mentioned above (and perhaps explains the wide variability in their data – ie the wide variability in supination resistance seen from person to person)
  • It is consistent with all the available evidence that I have seen (that may change in the future!)

While this model make sense, is coherent and biological plausible and can be used now to factor into clinical decision making, I will change it and my mind the minute there is some evidence that is inconsistent with it. As always: I go where the evidence takes me until convinced otherwise. This is where that evidence has taken me.

Last updated by .

, , ,

4 Responses to The Paradox of Patellofemoral Pain Syndrome and Foot Biomechanics

  1. Craig April 1, 2013 at 10:04 pm #

    I have had a couple of emails.To clarify:

    1. Foot orthotics work in patellofemoral pain syndrome (the evidence is clear)
    2. Proximal control (for lack of a better word) works in patellofemoral pain syndrome (the evidence is clear)
    3. There is NO evidence that minimalism or the 180-cadence works for patellofemoral pain syndrome (that does not mean that it doesn’t or that the evidence will or will not show that in the future)
    4. All I have proposed above is a model that attempts to reconcile inconsistencies in the literature

  2. Richard Bailey December 4, 2013 at 3:54 pm #

    Hi Craig – I enjoyed reading this article and have a couple of questions. Firstly – is there any biomechanical consequences of not being able to get to ‘toe-off- principally on the medial side of the foot. If my weight slides off towards the lateral column (?) will this ultimately affect the likelihood of vmo to fire ? In other words is there any evidence to show that hip and knee extension becomes less effective if I cant get through the big toe (windlass ?) side. Many thanks – Richard

  3. Gabe May 16, 2016 at 7:39 am #

    Hi Craig, another one for you… How would you differentiate the proximal (hip) from the distal (foot) as a contributor to valgus loading of the knee? I’m guessing you could have one without the other or both together? Cheers

    • Craig Payne May 16, 2016 at 8:05 am #

      My assumption if the foot is harder to move, then the distal moments are contributing more than the proximal.

      The real determinant of which one is greater will probably be the orientation of joint axes and the lever arms that tendons have to those axes –> that determines the moments.

Leave a Reply to Richard Bailey Click here to cancel reply.