VMO Part I: the true key to patella stabilisation?

In the first part of this 2-part article, Chris Mallac looks at the function of vastus medialis oblique, which is widely believed to function as a patellar stabiliser. But as he explains, the research is far from conclusive.

Giro d’Italia, Rome, Italy – May 27, 2018 Bora-Hansgrohe’s Sam Bennett celebrates as he wins the final stage.

There are an overabundance of research articles pertaining to the function and dysfunction of the vastus medialis oblique (VMO), especially in relation to its role as a dynamic patella stabilizer and the role it plays in patella-femoral pain syndromes (PFPS). Indeed, it is one of the most studied muscles in sports medicine and rehabilitation. Disseminating the sheer volume of published content can become confusing; the importance of the VMO in preventing PFPS is still debated by experts in conflicting reports on the exact contribution that VMO plays in patella control and pain syndromes.

PFPS is a pathology related to the contact and tracking between the patella and the underlying femur and the trochlear groove that is housed in the distal end of the femur. PFPS is a common injury to the knee. It affects sports and movements that involve high patella forces whilst the knee is flexed such as stair climbing, running up hills, cycling and weight training1.

It has been estimated that 14-17%2 3or 25%4 of knee injuries presenting to the sports medicine practitioner involves the patella-femoral joint. It has been argued for a long time that VMO deficiency could lead to patella tracking issues and be the precursor to PFPS, and therefore exercises for the VMO are necessary to counteract PFPS5 6.

Part one of this of this article unpacks much of the presented research to discuss the relevant anatomy and biomechanics of the VMO as it is currently understood. In the next part, we look at how VMO dysfunction may relate to PFPS, how it may become dysfunctional in the presence of knee pain, and which exercises can best be utilised to rehabilitate the dysfunctional VMO.

Figure 1: Extensor apparatus of the knee

Anatomy and biomechanics

Vastus medialis (VM) is one of four quadriceps muscles (rectus femoris, vastus medialis, vastus lateralis and vastus intermedius) that forms the extensor apparatus of the knee (see figure 1). All four components function as extensors of the knee, with the rectus femoris having the additional function of flexion of the hip. The four heads of the quadriceps provide variable levels of force production towards knee extension with the rectus femoris/vastus intermedius providing 35% of the force, the vastus lateralis providing 40% and the vastus medialis providing 25%7.

The VM muscle arises from the medial border of the linea aspera, located on the posterior surface of the femur, and extends from approximately the lower end of the trochanteric line to the upper third of the medial supracondylar line. It has a unique anatomy in that it ‘wraps’ around the femur from back to front8. Furthermore, it has been suggested the VM is divided into two distinct components separated by a fascial plane)9 10

  1. The superior or longitudinal component (VML)
  2. The oblique or inferior component (VMO)

The VMO originates partially from the adductor longus and the adductor magnus, and has an orientation that is either oblique or transverse (50-55°to the long axis of the femur), with the insertion of the inferior or oblique section occupying half or more of the medial border of the patella11 12. These VMO insert as fleshy fibres into the medial patellar retinaculum and the upper half of the medial side of the patella13 14 15 16 17. The more proximal VML fibres run 15-20°medially and insert onto the medial margin and anterior surface of the aponeurosis, which merges with the aponeurosis of the vastus intermedius muscle18.

However, Carlson and Smith (2012) found evidence to refute the idea that the VML and VMO are two anatomically distinct muscles19. In their cadaveric studies, they did not find a delineation between the VML and VMO fibres and no distinct facial plane was found between the muscles, suggesting that these are not in fact two separate muscles. They also found that only a small number of fibres (22%) inserted directly onto the actual patella with the majority of fibres terminating onto the quadriceps tendon. They argue that the VMO and VML are the same muscle based on their anatomical studies.

Nevertheless, further evidence suggesting a differentiation between the VMO and VML has been highlighted by studies that have found that the VMO and VML have separate innervations. In a dissection study on 30 human vastus medialis muscles and their nerves, it was revealed that a consistent bipartite nerve supply from the posterior division of the femoral nerve existed20.

A short and slender nerve referred to as the ‘lateral branch’, supplies the upper lateral portion of the muscle. The other part, a ‘medial branch’, supplies the middle and lower portion of the muscle. There is an increase in the numbers of nerve fibres supplying the muscle as the muscle moves more distally, with the lowermost muscle fibres receiving the richest nerve supply. This is supported by other studies that show that VMO and VML are electrophysiologically distinct entities21 22.

VMO, VML and muscle fibre type
Studies have shown that VMO and VML have different functional demands as evidenced by different proportions of muscle fibre types in the two muscles(20).

VML and VMO significantly differ in the proportion of type 1 (59.6% vs. 44%) and type 2b (6.3% vs. 15%) fibres. The VML muscle is almost entirely composed of type 1 and type 2a fibres with almost no 2b fibres found. The proportion of slow-twitch type 1 fibres is nearly twice as high as the proportion of fast-twitch type 2a fibres. In VMO, the proportion of fasttwitch type 2 fibres is higher than the proportion of type 1 fibres.

These observations indicate that VML is a slower and more fatigue-resistant muscle than VMO. These characteristics correspond to the different functions of the VML, which is an extensor of the knee and to the VMO, which maintains the stable position of the patella in the femoral groove. The differences between the VMO and VML relative to their morphological anatomy, their dual innervation and fibre type composition suggests that the VML and VMO may have different functional roles at the knee in terms of knee extension and patella alignment.

Functional roles of VMO and VML

For many years, it has been believed that the primary role of the VMO is to not only contribute to knee extension torque but also to provide medial patella support in the trochlear groove of the femur during knee flexion and extension movements23 24 25 26 27 28 29 30 31 32. It has been suggested that this muscle is the main active medial stabiliser of the patella, which counteracts the lateral pulling forces acting on the patella from vastus lateralis (VL), the lateral retinaculum and the iliotibial tract33.

VMO morphology and patella pain
This exact debate — as to whether morphology of the VMO plays a role in patients with and without patella pain — was studied by Balcarek et al (2014), when they studied morphological parameters such as cross-sectional area fibre angulation and the caudiocranial extent of the VMO relative to the patella in patients with lateral patella dislocation and those without(34).

What they found was that VMO morphology does not differ in patients with and without lateral patella dislocation. They argue that there is no clear evidence to suggest that the VMO has a direct stabilising effect on the patella. They go on to suggest that structures such as the medial patellofemoral ligament, trochlear groove geometry and medial retinacular structures may play a more important role in patella stability.

This is further supported by other anatomical studies that show that the VMO is not the primary medial stabiliser of the patella, and that the medial patellofemoral ligament may provide 50-60% of the stabiliation force to the patella(35).

With the knee at full extension, the patella is situated above the trochlear groove. In this position, the patella has no inherent bone support from the natural ‘channel’ formed by the trochlear groove and is therefore free to move medially and laterally. As the knee flexes, the pull of the patella tendon ‘drags’ the patella downwards, towards the trochlear groove.

Once the patella enters the trochlear groove, it gains stability in the medial and lateral direction from the bony margins of the trochlear groove. Patellar maltracking is more pronounced between full extension and 20° flexion, when the patella is not in close articulation with the trochlear grove34. It is thought that the VMO contributes to orientating the patella into the groove by also contracting during knee flexion . This mechanism seems counterintuitive when it is understood that the quadriceps are knee extensors. During knee flexion the quadriceps should remain relaxed as they are direct antagonists to the knee flexors such as the hamstrings.

This phenomenon of VMO contraction during knee flexion was studied by researchers and it was found that in the healthy knee, the presence of obliquely oriented muscle fibres and a medial tendon insertion site explains the modulation of forces acting on the patella as the knee moves from extension to flexion.

Due to the oblique orientation of the VMO (50-55°) and the medial location of the tendon insertion on the patella, VMO contraction would produce medial patellar tilt and lateral rotation (inferior pole moving laterally in the frontal plane) during knee flexion35. As the knee flexes, a medially oriented (as opposed to superiorly orientated) pull from VMO would cause the patella to flex (or pull inferiorly, not superiorly) further and translate in the medial, distal, and posterior directions relative the femur and its trochlear groove36. This direction of pull would ‘drag’ the patella into the trochlea groove and create a stable patellofemoral joint.

This modulation of action is consistent with the hypothesis that in the healthy knee, the VMO stabilizes the patella at or near terminal extension during both knee flexion and knee extension37 38. These findings however pose the next obvious question: If these anatomical findings are found in the ‘healthy’ knee, does this infer that these anatomical findings are not present in knees with PFPS? That is, do patients develop PFPS due to altered morphology and anatomical structure as opposed due to genuine weakening of the VMO muscle?

Some studies have highlighted the role that VMO plays in reducing patellofemoral compression forces – an important consideration in patients with patella chondral or osteochondral defects. In another cadaveric study (and in computational modelling studies) it was determined that improving VMO function reduces the load carried by the lateral cartilage of the patellofemoral joint39.

Increasing the force applied by the VMO consistently decreased the percentage of the joint compression applied to the lateral cartilage and consistently increased the maximum medial pressure. Moreover, this lateral compression reduction is even more evident in the presence of a lateral cartilage lesion, highlighting the importance of VMO conditioning in reducing knee pain in those suffering from lateral compression issues with the patella.

This study also showed that the slight increase in the compression as the VMO force increased contributed to the medial pressure increase being more consistent than the lateral pressure decrease, although the decrease in the maximum lateral pressure was still significant. When the VMO force was increased from the weak to strong cases the other forces were decreased to maintain a consistent extension moment and minimize differences in compression40 41 However, many of these direct EMG, cadaveric and biomechanical computation studies have assessed the function of VMO in relation to VL in isolation, and in non-weight bearing conditions with movement of the femur constrained42.

Powers et al (2003)43 found that the patellofemoral kinematics change as the limb moves from non-weight bearing to weight-bearing positions. Also, isolation studies do not consider the role of other anatomical considerations such as tibial torsion, valgus knee collapse due to poor hip muscle control, overpronation of the rearfoot and midfoot, femoral anteversion abnormalities and inherent genu valgus44.

These all contribute to a medial displacement of the knee joint in relation to the hip and foot, thus increasing the Q angle of the patellofemoral joint45 46. Likewise, a decrease in hip abductor and external rotator strength results in greater variance in hip-to-knee position during functional tasks47 These can all contribute to tracking issues and lateral patella compression forces, even in the presence of a functional VMO.

Furthermore, many of these studies also utilise an experimental group that actually has anterior knee pain. It may be possible that the knee pain itself alters lower limb kinetics and kinematics and isolated VMO to VL function. Therefore the results may be misleading for exact cause and effect between knee pain and VMO function.

Finally, other more compelling factors at the hip and pelvis may result in abnormal patellofemoral loading, which in turn may lead to the onset of patellofemoral pain syndromes. Frontal plane hip control (using abductors to prevent a Trendelenburg gait or pelvic collapse) and sagittal plane control (using gluteals to control anterior pelvic tilt and subsequent hip flexion moments) are crucial to prevent aberrant movement, which may lead to increases in joint compression and patellofemoral loading.


Part 1 of this 2-part article has hopefully highlighted just how difficult it can be to fully understand the actual functional role that VMO plays in patellofemoral control. At best, we can only assume that the VMO has a medial stabilisation effect on the patella during knee flexion and extension movements. In part 2 however, we will discuss how VMO dysfunction does exist in the presence of knee pain – ie that VMO dysfunction may be the end result of knee pain, and not the actual cause. However, the presence of VMO dysfunction alone alerts the therapists that some form of VMO re-strengthening may be necessary to fully rehabilitate the athlete with knee pain.


  1. Manual Therapy; 1996. 1:60-6.
  2. Gend Med; 2008. 5:405–14.
  3. British Journal of Sports Med; 2002. 36:95– 101.
  4. Clinical Journal of Sport Med; 1997. 7:28-31.
  5. Sykes K and Wong YM (2001) Electrical activity of vastus medialis oblique muscle in straight leg raise exercise with different angle of hip rotation. Physiotherapy, 89: 423–430.
  6. American Journal Physical Medicine Rehabilitation, 2001, 80: 804–808.
  7. Journal of Orthopaedic Research; 1998. 16:136–143.
  8. Clinical Anat, 2005. 18(4), 281-289.
  9. Orthopedics; 1980. 3:880-883.
  10. Clinical Anat. 2009,22: 183–199.
  11. Clinical Anat, 2005. 18(4), 281-289.
  12. International Research Journal of Biological Sciences. 2012. 1(5), 70-73.
  13. Anat Rec. 1997, 249:135–142.
  14. Journal of Bone and Joint Surgery [Am] A 1968. ;50:1535.
  15. Aust Journal of Physiotherapy, 1997, 43: 277–281.
  16. Surgical and Radiologic Anatomy. 1989. 11, 17-22.
  17. Journal of Arthroplasty; 2008. 23:600– 604.
  18. Aust Journal of Physiotherapy, 1997, 43: 277–281.
  19. International Research Journal of Biological Sciences. 2012. 1(5), 70-73.
  20. Journal of Anatomy. 1990. 170, pp. 193-198
  21. Journal of Physical Therapy Science. 2014. 26: 101–104, 2014
  22. Clinical Biomechanics. 2012 July ; 27(6): 525–531.
  23. Orthopedics; 1980. 3:880-883.
  24. Physical Therapy; 1995. 75:672-683.
  25. Müller W: The knee: form, function, and ligament reconstruction. Springer-Verlag, 1983, pp 76–77.
  26. Medicine and Science in Sports Exercise, 2007, 39: 1153–1159.
  27. Journal of Bone and Joint Surgery. 53A, 1971. 749-758.
  28. Clinical Biomechanics, 2000. 15(5), pp. 335– 339.
  29. Journal of Bone Joint Surgery Am; 1990. 72:1424–1429.
  30. American Journal of Sports Med. 2004. 30: 447–456.
  31. British Journal of Sports Medicine. 2004. 38: 295–299.
  32. Orthop Clin North Am. 1992; 23:619–30.
  33. Clinical Sports Med; 2010. 29:379-98.
  34. Journal of Orthopaedic Research. 2009;27:561–570
  35. Journal of Orthopaedic Research; 1998. 16:136–143.
  36. Journal of Biomechanics. 2010 January 19; 43(2): 235
  37. Journal of Bone and Joint Surgery [Am] A 1968. ;50:1535.
  38. Journal of Bone and Joint Surgery [Br]; 1995. 77:225–231
  39. Journal of Orthopaedic Research; 2009. 27(5): 578–583.
  40. Journal of Orthopaedic Research; 2009. 27(5): 578–583.
  41. Journal of Biomechanics. 2010; 43: 612- 617.
  42. Journal of orthopaedic & sports physical therapy. 2010. 40(2); 42-51.
  43. Journal of Orthop Sports Phys Ther. 2003;33:677-685.
  44. Journal of Orthop Sports Phys Ther; 1998. 28:345–54
  45. Journal Orthop Research; 1988.6:499–508.
  46. Physical Ther. 1995;75:24-30.
  47. Physiother Res Int. 2009; http://dx.doi. org/10.1002/jor.2066310.1002/pri.449
Share this
Follow us