What causes knee valgus?

If you’re interested in cutting-edge biomechanics, please sign up to our free newsletter using the form below and receive a single free study review every month!

What causes knee valgus?

No strength and conditioning coach likes to see knee valgus in athletes when they are squatting.  But the exact causes of knee valgus are not certain. Many researchers and coaches have previously suggested explanations but few have been widely accepted. One of the most commonly held theories is that reduced gluteal activity levels allow for a greater degree of hip internal rotation and consequently also tibial rotation and therefore knee valgus.  Very recently, a group of researchers put together an interesting study that shows that this is probably not the case. In fact, it seems more likely that the lower leg muscles are the culprits. In this article, Chris Beardsley (@SandCResearch) reviews the study for you.

The study: Neuromuscular Characteristics of Individuals Displaying Excessive Medial Knee Displacement, by Padua, Bell and Clark, in Journal of Athletic Training, 2012

(Too much detail? Skip to the practical implications)


What’s the background?

As is well known in the fitness industry, researchers have identified that knee valgus is a risk factor for lower extremity injury.  It is most commonly implicated in injuries involving the anterior cruciate ligament (ACL) but it also features as a risk factor in respect of other leg injuries, including patellofemoral pain syndrome, knee osteoarthritis, medial collateral ligament sprains and more general knee cartilage and meniscus damage.

However, despite it being extensively studied, researchers have not been able to show conclusively what causes knee valgus.  Various explanations have been offered, including decreased gluteal activation or strength, increased hip-adductor activation and decreased ankle-dorsiflexion range of motion (ROM).  Indeed, Bell (2008) showed that subjects who displayed knee valgus displayed ankle-dorsiflexion ROM that was 20% lower than controls.


What did the researchers do?

What was the purpose of the study?

The researchers wanted to find out two things.  Firstly, they wanted to investigate hip and lower leg muscle-activation during a squat in subjects who displayed medial knee displacement (i.e. knee valgus) and in subjects who did not. Secondly, they wanted to examine the effects on muscle activation and knee valgus of using a 2-inch heel lift.  This was to explore whether reducing the required ankle-dorsiflexion ROM would have an impact.

Who were the subjects?

For the study, the researchers recruited 37 young and healthy subjects (30 women and 7 men).  None of the subjects were currently suffering from any lower extremity injury.  The researchers assigned the subjects to either a Valgus group or a Control group. The Control group included 19 subjects (15 women and 4 men) who exhibited a squat movement in which the knees remained over the toes in the frontal plane.  The Valgus group included 18 subjects (15 women and 3 men) who exhibited a squat movement during which the patella was observed to move medially and knee valgus was therefore observed.

What measurements did the researchers take?

During the various trials, the researchers used surface electromyography (EMG) to measure the activity of several muscles, including the gluteus maximus, gluteus medius, adductor magnus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior. The muscle activity was normalized for each of the muscles using maximum voluntary isometric contractions (MVICs). Importantly, the gluteus maximus was normalized while the subjects were prone in 90 degrees of knee flexion and 0 degrees of hip flexion.  This position is generally accepted as being the posture in which maximum gluteal activation can be achieved. The gluteus medius was normalized in the side-lying hip abduction position but no details were provided regarding the degree of abduction. The researchers also recorded the movements of the lower body using an electromagnetic motion-analysis system.

What tests did the subjects perform during the measurements?

The subjects carried out 5 double-legged overhead squat repetitions barefoot, with their feet shoulder-width apart, their toes pointed straight ahead and as if about to sit in a chair. The subjects squatted to 80 degrees of knee flexion, which was as deep as early pilot data suggested was reasonable.  The researchers gave them feedback once they reached this depth. The trials were repeated with a heel lift of 2 inches. The heel lift was introduced by way of a wooden block placed under the feet.


What happened?

Hip muscle activation

The researchers did not find any differences in gluteal activation between the Valgus or Control groups. However, they did note a difference in adductor activity in that the Valgus group displayed 34% greater adductor activity than the Control group (when collapsed across both ascending and descending phases and both with and without heel lift).  Similar differences in adductor magnus activity between the sub-groups can be seen in the chart below:


As can also be seen from the chart, the researchers found that there were no significant differences between the EMG activity of the hip muscles with and without the heel lift.

Lower leg muscle activation

The researchers found that the Valgus group displayed greater EMG activity of the tibialis anterior compared with the Control group during the descending phase of both the no–heel-lift and heel-lift conditions.  They also noted that the EMG activity when using heel lifts was less than when not using heel lifts for both Valgus and Control groups, again in the descending phase.  Both of these results can be seen in the chart below:


The researchers found that there was a trend for similar (but not identical) results in the ascending phase but they did not reach statistical significance. The researchers also observed that the gastrocnemius activity (both medial and lateral) in the descending phase without the heel lift was greater in the Valgus group than in the Control group. They also noted that the gastrocnemius activity (both medial and lateral) with heel lift was higher than without the heel lift in the Valgus group but not in the Control group.  They noted a trend for similar (but not identical) results in the ascending phase but they did not reach significance.

Frontal-plane knee displacement

Without a heel lift, the researchers reported that the Valgus group displayed greater medial knee displacement than the Control group. With a heel lift, the Valgus group displayed less medial knee displacement than without the heel lift. There was no effect on frontal knee displacement of the Control group of using the heel lift.


What did the researchers conclude?

What were the differences between Valgus and Control groups?

The researchers concluded that the EMG activity levels of the gastrocnemius, tibialis anterior and adductors were greater in subjects who displayed knee valgus than those who did not.  They also concluded that there were no differences in the EMG activity levels of the gluteal muscles in subjects with and without knee valgus.

What do the researchers think causes knee valgus?

The researchers suggest that the increase in gastrocnemius activity might have created greater plantar-flexion moments at the ankle – thereby increasing ankle-dorsiflexion stiffness.  They suggest that the increased anterior tibialis activity may have also increased the ankle-dorsiflexion stiffness through co-activation.

The researchers therefore propose that it is increased ankle-dorsiflexion stiffness that leads to restricted dorsiflexion ROM.  They suggest that this then leads to compensatory foot pronation and tibial internal rotation and therefore knee valgus.

What were the differences as a result of using the heel lift?

The researchers also concluded that the EMG activity of the tibialis anterior was reduced when using a heel-lift in all subjects, irrespective of whether they displayed knee valgus or not. They concluded that EMG activity of the gastrocnemius was also decreased when using a heel-lift but only in subjects who displayed knee valgus.  They noted that there were no differences in the EMG of the gluteals or hip adductors as a result of using the heel-lift.


What were the limitations?

The researchers observed the following limitations to their study:

  • The study was limited to an overhead squatting movement and different results might be observed in other movements, such as jump landings.
  • Additionally, this study was limited to taking EMG activity measurements in subjects whose knee valgus went away when they performed a squat with a heel-lift.  Different results might be observed in subjects whose knee valgus remains with a heel-lift.
  • The researchers did not measure all of the hip and leg muscles that may have a role in controlling frontal-plane movement of the knee joint, such as the quadriceps and hamstrings.
  • The study was not able to establish whether reducing the activity of the anterior tibialis and gastrocnemius would automatically lead to a reduction in knee valgus.


What are the practical implications?

Athletes who display knee valgus can temporarily use squat variants with small heel lifts or Olympic lifting shoes until they are able to squat normally without them.

Athletes who display knee valgus should work to develop ankle dorsiflexion ROM during squats using soft-tissue work, gastrocnemius stretching and ankle mobility drills.