Runners knee pain… A guest article by Tom Goom

So for the first time on ‘The Sports Physio’ I am handing the reins over to someone else and giving them the helm, but not to just any Tom, Dick or Harry we are very privileged to have our very own ‘special’ Tom, Tom Goom, a fellow physio with a wealth of experience both as a physio and as a runner. He writes regularly for his own very successful web site called the Running Physio (found here) where he gives heaps of advice and guidance to runners on injury prevention and rehab, today he writes for us on the subject of runners knee and sorts the myths from the facts surrounding the role a particular muscle called the VMO does or doesn’t have in it.
Patellofemoral Pain Syndrome, aka PFPS or Runner’s knee, is a common and sometimes debilitating problem that can be challenging to treat. A host of different treatment approaches have been tried with varying success. Exercises have been suggested for the glutes, quadriceps and core, stretches for ITB, hamstrings, calf and hip flexors. We’ve taped, mobilised, manipulated and massaged and yet a number of patients will find their pain remains. One area of discussion has been the role of Vastus Medialis Obliquus or VMO. This makes up part of the quadriceps – a group of four muscles that extend the knee and support the patella. This article will examine what role VMO has and whether it can be worked selectively with exercises.
Central to the theory behind recruiting VMO is the idea that PFPS may be caused my ‘mal-tracking’ of the patella. The patella rests in a groove on the femur, where it glides during movement of the knee. It acts almost like a pulley to direct the force of the quads into the tibia to extend the knee. A number of structures attach to the patella and can affect its movement. Some of these structures are thought to have a lateral pull on the patella (move it outwards) and others a medial pull (move it inwards). In PFPS it is thought that the structures pulling laterally are exerting more force than those pulling medially which results in the patella moving more laterally. This causes ‘mal-tracking’ which can lead to pain and inflammation. Sometimes evidence of this mal-tracking can be seen on x-ray or MRI where the patella appears to sit laterally within the groove.
VMO sits on the inside of the knee and, in theory, can help to pull the patella more medially and help with tracking. I say ‘in theory’ because, like many things, it’s a subject for debate in the Physio world.
Does VMO even exist?
The first spanner in the works for this theory is the rather big question of whether VMO as we think of it even exists! The quadriceps is made up of 4 muscles, Rectus Femoris, Vastus Lateralis, Vastus Intermedius and Vastus Medialis. Of these 4 it is thought that Vastus Medialis (VM) has 2 distinct sections to it VM Longus and VM Obliquus. However some dispute this. In 2009 Smith et al. completed a systematic review of the literature on this topic. They examined 26 papers including 699 healthy knees and 591 with PFPS. They found that in the majority of cases in both healthy knees and those with PFPS there was a substantial alteration in fibre alignment between the proximal and distal portions of VM. However they concluded that there was insufficient good quality evidence to state whether VM was composed of two separate components of VML proximally and VMO distally.
Putting that in plain English – they were unable to prove VMO exists as a separate section of the muscle.
Is VMO altered in patients with PFPS?
When we think of muscles we usually think of strength but this isn’t the only characteristic that needs to be examined here. Research has also explored VMO bulk and timing of VMO contraction as both are thought to play a part in PFPS.
Pattyn et al. (2011) examined VMO and quads bulk using MRI in what they claim is the first study of its kind in PFPS. They found VMO to be significantly smaller in those with PFPS than those without. They also reported that the total quadriceps size tended to be reduced in those with PFPS. They were unable to state whether PFPS resulted from decreased VMO size or caused it. In my clinical experience of PFPS, quads and VMO atrophy appears to be fairly common and specific to the painful side. However a number of factors are thought to affect quads bulk and swelling or pain are known to reduce quads activity so it is possible it happens as a result of PFPS.
There appears to be some agreement in the literature that reduced quads strength is a potential cause of PFPS although less evidence for specific weakness of VMO. Clinically we see many patients with PFPS have weaker quads on the affected side and respond well to a strengthening programme.
There are multiple studies on timing of muscle contraction looking at when VMO contracts relative to VL. The theory being that if VMO activates later than VL then the patella is moved laterally and this affects tracking. However, it is worth noting that the timing differences are usually miliseconds and there is some debate as to whether this amount of time is significant enough to cause mal-tracking.
Chester et al. (2008) performed a systematic review and meta-analysis of studies on timing of VMO and VL contraction in Anterior Knee Pain (of which PFPS is a potential diagnosis). They found a trend towards delayed activation of VMO relative to VL in those with Anterior Knee Pain compared to those without. However they also reported that the clinical and therapeutic significance was difficult to assess. Again establishing cause and effect is also complex – studies have shown the presence of pain or swelling to affect timing of muscle contraction – it’s feasible that pain from PFPS could cause the timing difference not result from it.
Can you selectively recruit VMO?
It would appear from the research that you can’t specifically exercise VMO on its own. Toby Smith (who also produced the research on the anatomy of VMO detailed above) completed a systematic review of electromyographic studies (research that used EMG to identify which muscles were active); Smith et al. (2009) assessed 20 studies including 387 participants and concluded that altering leg position or adding co-contraction did not enhance VMO activity over VL. They also acknowledged a number of methodological problems in the literature and that further research was required.
In this case the research seems to mirror what I see clinically – varying exercises doesn’t appear to target VMO in isolation. I also think muscles function together and are often anatomically more of a unit than 4 separate muscles working independently. Strengthening the quads as a unit appears to improve PFPS, it may be that this aids patella tracking or reduces patellofemoral load or that it helps control of movement. The literature appears to support this – Syme et al. (2009) compared general quads strengthening with selective VMO training for chronic PFPS. They found that both approaches resulted in significant improvements in pain, function and quality of life. So you might decide to target VMO specifically with exercises, but there is little evidence that a) this is possible or b) it is more effective than general quadriceps strengthening.
Baring all this in mind, how should we manage PFPS?
The key to managing PFPS is a graded exercise programme based on a comprehensive assessment. PFPS is a multifactorial problem. Weakness of quads or VMO may play a part but are just one of a number of potential causes. From the literature we can divide these causes into a broad categories – muscle weakness (especially quads and glutes), decreased control of movement (including poor control of single leg dip, altered proprioception and decreased trunk control), altered biomechanics (altered Q-angle, patella position, overpronation etc.), tissue tightness (ITB, quads and hips flexors, hamstring and calf muscles, altered patella mobility) and psychosocial factors (work, sport, mental health).
This is part of the reason why some research in this field is inconclusive – quads exercises will probably work for people who have weak quads as a cause of their PFPS but might not for those whose predominant cause is weak glutes or tight ITB. Strengthening exercises might not work at all if the underlying cause is continued overuse (e.g. high mileage runners).
A further important challenge with PFPS is undergoing rehab without aggravating symptoms. Many exercises that will help the condition can easily aggravate it. The way around this is to select exercises with low patellofemoral load and use a sensible graded approach. This, at its simplist, can be guided by pain – just work each exercise in a pain free range. Research McGinty et al. 2000 (p164) adds some precision to this approach,

“Both OKC (Open Kinetic Chain) and CKC (Closed Kinetic Chain) exercises can be utilised in the treatment of patients with patellofemoral pain if performed in a pain free range. CKC exercises may be better tolerated by the patellofemoral joint in the range of 0-45° of knee flexion. In this range, suggested exercises include step-ups, mini-squats, and leg presses.
OKC exercises may be better tolerated by the patellofemoral joint in the ranges 90-50° and 20-0° of knee flexion. In these ranges, suggested exercises include short arc isotonics, multiple angle isometrics, straight leg raises and quadriceps sets.
Performing CKC and OKC exercises in these specific ranges loads the quadriceps while minimising stress on the patella.”

For more details on open chain and closed chain exercises see my video. I won’t get into the specifics of exercise programmes for PFPS – that’s an article or two in its own right but more info is available in videos here and here. There is also excellent research from Scott Dye (2005) and is very helpful in this topic.
Final thoughts on VMO and PFPS – the evidence for the existence of VMO, our ability to exercise it selectively and the effectiveness of this approach over general quads work is very poor. A more evidence-based approach would be a general quads strengthening programme, including both open and closed chain exercises, working in a range with low patellafemoral load. This, of course, would form part of a comprehensive rehab approach based on detailed assessment



  1. Most informative article, as always. I have often questioned the methods often applied to exercise individual quad muscles that are diagnosed as being weak. As Tom says, it’s not as easy as simply pointing feet in or out. One thing that I have found can help is the use of Soft Tissue Mobilisation to separate the individual muscles of the quadriceps group (where possible) to promote decongestion. By doing this you can help each of the muscles to work optimally independently (as they are supposed to) and enhance overall movement performance. This done, normal quad strengthening exercises performed with optimum postural alignment will stimulate recruitment of all four muscles appropriately and allow strengthening where needed.

  2. I suffer from this pain, and found a way to manage it by exercising the VM muscle group (I’m not debating the existence of the VMO here). The exercise I found that worked most effectively, both in strengthening that particular muscle and also in allowing me to run again, is fairly simple. I use a high step exercise, consisting of two main exercises: start from behind the unit, leg to be exercised on the step–then pull yourself up by that leg slowly, and return. The other one is similar, except that you don’t stop in the middle, you pause briefly, then drop your leg in front of the stepper, like you’re stepping over the stepper. Then return to your original position behind the stepper.
    I’ll also note that I also suffer from dimorphism (right side is larger than left), and it’s the left quad that is undersized and that’s the knee sourcing the pain. Obviously, this may not work for everyone, but just wanted to share what is working for me.

    • Hi Greg
      Thanks for your comments, they sound like good all round quad strengthening exercises, thanks for sharing them
      With regards to your ‘diamorphism’ I wouldn’t worry too much, firstly every body is slightly off balanced we are not symmetrical, its deciding when asymmetry is significant or not that’s the challenge and also the size of a muscle doesn’t equate to its strength or power, some of the strongest people I know are the smallest
      All the best

  3. If you can’t selectively recruit VMO, what about foam rolling the VL before a strengthening exercise? It may inhibit the VL ,so the VMO has a higher input when performing the exercise/lift.

    • Hi, thanks for an interesting comment, foam rollering is a strange one, yes some evidence theorises that it works on descending inhibition but only for nocioception (pain) it has not been shown to reduce the neuromuscular connections so I don’t think it can reduce or inhibit muscle action this way, the neural system just doesn’t work like that, other wise everything we come into contact with would affect our muscle tone and strength, but an interesting thought thanks

Related news