Femoral Anteversion, Craig's test and Poor Advice

The intent of this article is to shed some light on lower extremity joint coupling from femoral versions. I hope that it is easy enough for non-healthcare folks to understand, yet detailed enough for healthcare providers to reference.

I am getting frustrated with patients coming into the clinic and telling me they received a gait analysis, or a bike fitting from someone with zero anatomy education but a self-perceived wealth of healthcare advice.

The impetus and motivation to write this article stems from a patient encounter a couple weeks ago. She is a triathlete with pain when she runs and bikes. Recently, she had a bike fitting and the bike fitter told her that her knee was tracking inward (toward the top tube) when she biked and she should make a conscious effort to force her knee outward at the top of the pedal stroke. There was also talk about putting a varus wedge between the cleat and the shoe to help force the knee straighten out. (Not only did he say that, but also that the knee was going inward because the ITB was too tight and she needed to stretch it…?)

Sorry if I’m stepping on toes here, but my opinion is that in a situation like that, the bike fitter should be sticking to properly fitting the bike to the person rather than dispensing treatment advice and modalities.  I’m all for bike fitters helping performance, but if someone has pain, I think there may be some boundaries that are crossed if they are prescribing stretches or installing orthotic wedges.  The same goes for coaches helping runners.

During my office examination, I did a standard procedure that I do on many patients: Craig’s test to check for femoral versions. This patient did have significant femoral anteversion (explained below in the video). This femoral anteversion explains some of the knee tracking inward and it reminded me of why people without any anatomical or biomechanical education should be cautious about dispensing therapeutic advice.

What is Craig’s test and femoral anteversion? Below, I have made a video explaining femoral versions/torsions and Craig’s test (how to easily screen for them). I have to used this video occasionally for clients of Running Reform. When clients submit their video to me and there are significant rotational distortions in their lower extremity, I ask them to perform this test at home and report back. Ruwe et al., (1992) proved this test to be more reliable than CT scan and I think it’s pretty easy to perform.

Please watch the 3 minute video here which explains a lot more than just Craig’s test:


Since femoral anteversion causes internal rotation of the femur and internal rotation of the femur is associated with greater knee valgus, one could assume that femoral anteversion is associated with knee valgus.

Figure 1: Running form with knee valgus

This idea makes sense and has been proposed by Hvid et al., (1982) and by Nguyen et al., (2007). However, Nguyen et al., (2009) used Craig’s test to evaluate femoral version and found no relationship to knee valgus which contradicts the previously listed studies. However, they tested knee valgus in a static double leg stance and this doesn’t mean that there would not be any dynamic valgus. In other words, standing still may not have increased knee valgus, but when running, jumping biking etc., there may be increased knee valgus in those with femoral anteversion. This is due to a couple reasons:

  1.  Merchant (1965) that the gluteus medius is at a mechanical disadvantage in those with femoral anteversion. Basically, the lever arm is reduced. Since the gluteus medius is important for maintaining frontal plane stability of the hip and knee (not allowing the hip to adduct or the knee to collapse inward).
  2.  Nyland et al. (2004), showed there is decreased activation of the gluteus medius, as measured by surface electromyography amplitude, in people with increased relative femoral anteversion.

When you put these together (reduced mechanical leverage plus reduced activation) it suggests that those with femoral anteversion would have a tendency to have increased dynamic valgus at the knee.

There are suggestions that femoral anteversion increases anterior knee pain most likely due to patellofemoral pain – Eckhoff et al., (1994) and Eckhoff et al., (1997). In addition, Takai et al., (1985) and Eckhoff et al., (1994) found increased arthritis in the patellofemoral joint was associated with femoral anteversion.

Moving on from femoral anteversion to femoral retroversion…When patients talk about getting a “knee replacement”, they are usually talking about arthritis in the tibiofemoral joint. This type of arthritis is generally on the medial (inside) compartment as it is usually prone to develop arthritis. Therefore, a patient at risk of this (family history, cartilage degeneration etc.) should be wary of anything that places extra pressure on the medial compartment. There are studies that show femoral retroversion increases the pressure on the medial compartment of the tibiofemoral joint. Bretin et al., (2011) not only showed that femoral anteversion predisposes the knee to a valgus deformity but also that femoral retroversion predisposes the knee to a varus deformity and also increases the pressure in the medial tibiofemoral joint. Kenawey et al., (2011) confirmed this notion by reporting that the medial tibiofemoral joint increased it’s pressure by nearly 30% when there was 20 degrees of femoral retroversion compared to anteversion. Again, Papaioannou et al., (2013) confirmed the increased compression medial compartment of the knee joint due to femoral retroversion.

So, if femoral retroversion is susceptible to knee joint pain and arthritis, and femoral anteversion is susceptible to patellofemoral pain and arthritis, everyone should be trying to force the knee straight, right? Not so fast. If a person with either retro or anteversion tries to rotate the leg so that the toes are pointing straight, the result may not be what you want. Essentially, the person is being asked to force the hip into a position where the head of the femur is no longer seated properly in the acetabulum which could lead to hip joint pathology. There are no studies that have evaluated this, but logic has to take over at some point. In his book Human Locomotion, Michaud states, “Unfortunately, once the femoral anteversion has been formed, attempts to modify an individual’s gait by having him/her walk with a straight gait pattern decreases the mechanical efficiency of the gluteus medius musculature and increases femoroacetabular contact pressures.”

Unfortunately, in many cases a person with femoral anteversion or retroversion will wrongly be instructed to keep their knee straight, or point their toes forward. They may be given a varus or valgus post in their running shoes or bike cleats to try and “straighten things out.” This can potentially be detrimental, since they essentially are being instructed to lose congruency in the hip socket which would increase contact pressures and risk potential injury.

So, as usual, the answer to this and many other conditions of the body is “it depends.” You can’t absolutely state with 100% confidence that a person with their knee going in toward the top tube should be forcing it to go straight via conscious efforts or with a cleat wedge. Then again, you can’t absolutely state with 100% that they shouldn’t either! Femoral versions need to be considered, tibial torsions need to be considered (perhaps in another blog post), patient genetic susceptibilities need to be considered, movement patterns need to be evaluated, patient symptoms need to be considered and the list goes on. Despite this, there will still be people with zero knowledge of biomechanics or anatomy dispensing advice with way more confidence than I have, even after I have considered the information above.

Will they guess correctly? Roll the dice and find out!