This article is adapted from the book Your Body, Your Yoga by Bernie Clark.
Does your standing leg look like that of the woman in the photo below doing king dancer pose (natarajasana)—slightly hyperextended when standing on the one foot? Have you ever been corrected and told to “micro-bend” your knees? Were you told this is to protect your joints and ligaments? Well, maybe it is—but maybe a little hyperextension of the knees is actually the healthiest thing for your knees!
Here is an interesting quotation regarding the knees: “A common error made by clinicians is to regard 0° of extension as full extension”.(1) In the yoga community most teachers try to remove hyperextension of the knees from all standing postures. It turns out that this is not always wise. The researchers in the study cited above concluded, “The results of this [study] confirm our belief that most people have some degree of hyperextension, and it is paramount that clinicians consider this when evaluating and treating a patient with a knee injury.” This means that hyperextension may be just as natural and okay for you as it is for the woman in the photo doing natarajasana.
“Hyper” can mean excessive, which has negative connotations. We don’t want to excessively extend any joint, and the knees are no exception. But “excessive” is a relative term: A hyperextended knee is excessively extended relative to what? A common assumption is that a healthy joint is one that opens to 180°… and any movement past 180° is defined as being “hyper” (and anything less than 180° as “hypo”).
Figure 1a (below) shows a knee that is neither hyperextended nor hypoextended: The hip joint is above the knee joint, which is above the ankle joint. But if this is the limit to how far you can move your knee, it is neither normal nor healthy! The second leg (b) is slightly hyperextended and the third even more so (hyperextension greater than 5° is sometimes termed genu recurvatum which is Latin for “backward-bending knee”). (2)
FIGURE 1. The ideal versus the reality of hyperextension. While anatomy books show (a) as the ideal alignment of the knees, the average person hyperextends their knees around 6° (b). Many people, however, can hyperextend the knees to 17° or more (c).
When we review all those who hyperextend their knees, it turns out that almost everyone extends past the magical 180° line. The average amount of hyperextension exhibited, however, varies greatly: 99% of women hyperextend their knees, with the average being 6.7° of hyperextension,(3) while 95% of men hyperextend, their average being 5.5°.(4)
Thus, most people have genu recurvatum—including the students who are constantly being corrected in yoga class.
Researchers are quite divided as to whether hyperextension of the knee predisposes us to injury.(5) While everyone with lax knee ligaments hyperextends, not everyone who hyperextends has lax ligaments.(6) Despite the view in the yoga community that the 180° alignment of the knee is ideal, allowing the knee to be slightly hyperextended is actually necessary for human bipeds.
Even though it sounds negative to “lock the knees” and “hang out” in the ligaments when we stand, we reduce muscular energy thereby—allowing the ligaments, rather than the muscles, to do the work of supporting our posture. Our species evolved to do this. Walking chimpanzees expend far more energy than walking humans, because the chimps have to keep their knees constantly bent. Humans can extend the knee, and this saves on calories.(7)
We hyperextend around 5° when we walk, which is, not surprisingly, right around the amount that average people can do.(8) If we did not slightly hyperextend the knees while standing or walking, the quadriceps would be constantly working to hold our posture, which would be tiring and could lead to chronic shortening of the quadriceps.
It may be surprising to discover that some hyperextension is actually very normal and useful. If, after a knee injury, we cannot hyperextend to our usual limit, the knee will feel constrained and stiff. We may then compensate for the loss of range of motion by creating unhealthy movement patterns.(9) Restoring full range of motion—including hyperextension—after injury or surgery prevents knee pain and abnormally thick scar tissue (called arthrofibrosis).(10) We need 5° to 7° of hyperextension for full functional activity.
(On the other hand, we can do too much of anything. People who sit all day long, which can cause their leg muscles to weaken, may tend to overdo hyperextension in the knees while standing. For these people, it may indeed be better to microbend the knees while in their yoga postures, and to work on strengthening their leg muscles.)
In figure 2 below, we see two yogis in natarajasana: Are they at risk? Perhaps, but not necessarily!
FIGURE 2.Both yogis are hyperextending at the knee, but the yogi on the left has the most hyperextension: 12°. This may or may not be a problem for her—it depends on the health of her ligaments.
There are two ways we can hyperextend the knee: dynamically or statically. In dynamic hyperextension, the knee is forced backwards due to a sudden trauma or movement. The most frequently damaged ligament in knee injuries is the anterior cruciate ligament (ACL), and a common cause of ACL injury is a sudden stress on the knee while it is hyperextended (such as a sports contact injury that forces the knee backwards).
Note that the ACL prevents the femur from sliding backward over the tibia, and the posterior cruciate ligament (PCL) prevents the tibia from sliding backward under the femur. Knee injury most often occurs when we make a sudden stop while running, especially when combined with a change of direction, or when we jump and land with the legs extended.(11),(12) Unfortunately, women are two to eight times more likely to damage their ACLs than men, and this happens most often in the landing that follows a jump.(13) (Women tend to activate their quadriceps initially upon landing, whereas men activate the hamstrings first.) (14)
Excessive hyperextension of a knee bearing a load increases the risk of ACL damage. If in addition to the hyperextension, the knee is twisted, the risk of damage increases.(15) Interestingly, the incidence rate of ACL injuries is far lower in gymnastics than in skiing, handball, soccer, basketball, or contact sports.(16) (Which of these sports seems to you most similar to yoga?)
In none of the reported cases surveyed was static hyperextension cited as a cause of knee damage. In the yoga environment, our stresses are mostly static: We rarely subject the knee to large, transient, dynamic stress. With the exception of Kundalini yoga and occasionally Ashtanga yoga, which may employ quickly jumping into and out of postures, most yoga styles are slow. A hyperextended knee may experience static stresses while in the postures, but the levels of these stresses are far below a healthy knee’s tolerance.
In several standing postures, such as king dancer, triangle pose (trikonasana), tree pose (vrksasana), and half moon pose (ardha chandrasana), one leg is fully extended and statically supporting most or all of the body’s weight. This is illustrated both in figure 2 and in the artist’s depiction in figure 3 of B.K.S. Iyengar performing king dancer and half moon. In many photographs, Iyengar displayed an average amount of hyperextension in his supporting knee: between 5–7°.(17) Almost everyone can hyperextend their knee while standing, so it is to be expected that most yoga practitioners will do so in their postures. The question is: Should they?
FIGURE 3. Hyperextension between 5° and 7° is normal. B.K.S. Iyengar hyperextended his knees in half moon pose and king dancer.
As shown in figure 4a, there is no bony stop at the back of the knee, as there is in the elbow joint. The ligaments, joint capsule, and tendons restrict posterior movement. One study found that 85% of the stress pulling the tibia forward is absorbed by the ACL.(18) Another study, however, found that the oblique popliteal ligament (found in the back of the knee) was the main ligament preventing hyperextension of the knee, contributing about 37% of the resistance.(19) This latter study began with the complaint that “a primary stabilizer against knee hyperextension has not been identified.”
FIGURE 4. (a) The ligaments of the knee: anterior/lateral oblique view (right knee). (b) The plateau at the top of the tibia is angled downward from front to back.
Time for some numbers. (But not to fear: If you happen to get lost, just jump to the last line of this paragraph.) How much stress is being placed on the knee in a statically held, hyperextended yoga pose? The greater the degree of hyperextension, the greater the amount of tension being placed on the ligaments. A healthy ACL can withstand over 2,000 newtons (N) (which is equivalent to a force of 450 pounds)(20) and the PCL twice that.(21) We can offer a rough approximation of how much stress a hyperextended ACL experiences while standing.(22) For a 60 kg (132 lbs.) person standing in mountain pose (tadasana) with 5° of hyperextension, trigonometry tells us that each knee experiences an average of about 26 N (~6 lbs.) of force pushing to the back of the knee, which is well within the ACL’s tolerance levels.(23) But what about the woman with the greater hyperextension in figure 2? If we assume that she weighs about 60 kg (~132 lbs.), the amount of posterior stress being placed on the standing knee with 12° of hyperextension is about 122 N (or ~27.5 lbs.).(24) Again, this is less than 10% of a healthy ACL’s tolerance level of 2,000 N! It is highly unlikely that 5° of passive hyperextension will be a concern for most people.
The causes of hyperextension beyond normal are many and varied: neurological deficits, muscle weakness, ligament laxity or injury. Interesting that strong quadriceps engagement can also cause hyperextension! (That will pull the tibia forward under the femur, which is hyperextension). But quadriceps that are too weak to stabilize the knee may also lead to hyperextension, as we then end up using the ligaments to provide the stability that the muscles cannot.(25) For this reason, people with congenital muscle weakness, for example, cerebral palsy or muscular dystrophy, or people who have suffered a stroke, often rely on hyperextension to support the knee. Over time, however, this can create too much laxity in the joint.
Human variation in the shape of the tibia can also contribute to the amount of hyperextension that’s natural for you. The top of the tibia, called the tibial plateau, slopes from the front to the back, as shown in figure 4b. How much slope people have is variable. One study found that the slope is about 7 to 11°, although the range of human variation is roughly 1° to 19°.(26)
How much do you hyperextend? The answer may depend upon laxity in your ligaments—which may or may not need fixing—or upon the shape of your bones, which will not change.
With a greater slope, there is greater stress on the ACL, which explains why injuries to the ACL are more likely for people who have a greater tibial slope.(27) However, here is an interesting observation: Those who possess more slope tend to exhibit less hyperextension than those who have less slope.(28) How much do you hyperextend? The answer may depend upon laxity in your ligaments—which may or may not need fixing—or upon the shape of your bones, which will not change.
It is worth noting that when the knees are hyperextended, some students compensate by tilting their pelvis anteriorly (forward); this increases the normal lumbar curve (and is called hyperlordosis), which can have negative effects along the spine and higher up the body. Again, every body is different, and while hyperextension may be quite acceptable for the knees, for you it may cause lower back problems or other issues. Indeed, some yoga teachers feel that hyperextension of the knee, while biomechanically safe, negatively affects the energetic qualities of the postures and the breath. Personally, I don’t believe it to be a valid concern for the majority of people because we have evolved to stand with hyperextended knees.
For those who do have knees at risk and/or have too much hyperextension, they can reduce hyperextension (see below), but this may prevent the stress needed to keep the knees healthy.(29) One way to meet both intentions of minimizing stress while maintaining the knee’s range of motion is to use a yin yoga practice to stress the joint safely and passively while it is hyperextended, and then use co-contraction during an active practice to prevent hyperextension. In this way, we put no one at risk, but we are also not overly protective of those who do need some stress in the knees.
One technique to reduce the level of stress on the ACL caused by hyperextension is co-contraction of the quadriceps and the hamstrings (see figure 5).
FIGURE 5. Hyperextending the knees in staff pose (dandasana): (a) While sitting with the legs straight in front of you, tighten the thighs and notice whether your heels rise up. If they do, you are hyperextending the knees; (b) Continue to push the thighs to the floor, but also push the heels down; this co-contraction engages the hamstrings and reduces hyperextension.
Hyperextension can be caused by over-activation of the quadriceps, or, more weakly, by the actions of the calf muscles (specifically the gastrocnemius). The common cue to lift the kneecap in standing postures, thus engaging the quads, does not stop hyperextension and may actually increase it. The quads pull the tibia forward, as shown in figure 6, creating more stress on the ACL. (The gastrocnemius pulls the femur backwards, again resulting in an increase in tension on the ACL.)
FIGURE 6. Co-contraction. Contraction of the quadriceps pulls the tibia anteriorly, creating stress on the ACL, but contracting the hamstrings pulls the tibia posteriorly, reducing ACL stress. Co-contracting both muscles at the same time helps to maintain extension but reduces ACL stress.
To reduce the stress on the ACL, we need to pull the tibia posteriorly, and we can do that by engaging the hamstrings, as shown in staff pose in figure 5b.
We can also apply this technique to reducing hyperextension in standing poses—for example, triangle pose as shown in figure 7.
FIGURE 7. Co-contraction in triangle pose. We can engage the quadriceps by attempting to stretch the mat beneath us. We can engage the hamstrings by simultaneously hugging the thighs together or by simulating sliding the feet together. These two actions may reduce or eliminate hyperextension of the knee.
It is okay and indeed healthy for most people to hyperextend their knees in static yoga postures, but for some people it may be dangerous. The level of static stress created by standing postures that hyperextend the knee(s) is quite moderate compared to the stress we experience while climbing stairs or running. However, students who have damaged or compromised ligaments, people who hyperextend beyond the normal range (i.e., more than 10°),(30) older people, who naturally have weaker ACLs,(31) or anyone recovering from knee surgery may well be advised to reduce the stress on their ACL.
Keep in mind, however, that while too much stress can be harmful, there is a danger in too little stress. As some researchers have observed, “The absence of strain on ligaments, even those that have not been injured, has been shown to have harmful effects.” (32)
As with all tissues, we need to stress the ACL! And that means, for the vast majority of yoga students, that not only is it okay to hyperextend the knee in yoga postures—it is actually a very good idea.
1. From M.S. De Carlo and K.E. Sell, “Normative Data for Range of Motion and Single Leg Hip in High School Athletes,” in Journal of Sports Rehabilitation 6.3 (1997): 246-55.
2. The literature varies on when to classify the hyperextended knee as genu recurvatum, with standards ranging from 0° to 10° of hyperextension.
3. In this study, one standard deviation was 2.7° which means that 68% of the women surveyed fell within the range of 4° to 9.4°, but the total range of the whole group was from zero to 17°. “Normal” is defined to be within 2 standard deviations which covers 95% of the population, and the normal range of knee extension for women was from 1.3° to 12.1°. See M.S. De Carlo and K.E. Sell, “Normative Data for Range of Motion and Single Leg Hip in High School Athletes,” in Journal of Sports Rehabilitation 6.3 (1997): 246-55.
4. The standard deviation for men in the above study was of 2.5°, with the overall range running from zero to 17.5°.
5. See R. Ramesh et al., “The Risk of Anterior Cruciate Ligament Rupture with Generalized Joint Laxity” in Journal of Bone and Joint Surgery 87-B.2 (2005): 800-3.
7. See Daniel Lieberman, The Story of the Human Body: Evolution, Health and Disease (New York: Pantheon, 2013) page 42.
8. The range of hyperextension required for walking on level ground varies between 2° – 7°; See N.B. Reese and W.D. Bandy, Joint Range of Motion and Muscle Length Testing, 2nd edition, Saunders (2010) page 333.
9. See De Carlo and Sell, “Normative Data.”
10. See K.D. Shelbourne, D.V. Patel and D.J. Martini, “Classification and Management of Arthrofibrosis of the Knee Following ACL Reconstruction,” American Journal of Sports Medicine 24 (1996): 857-62. And, see K.D. Shelbourne and R.V. Trumper, “Preventing Anterior Knee Pain Following ACL Reconstruction,” American Journal of Sports Medicine 25 (1997): 41-7.
11. See R. Ramesh et al., “The Risk of Anterior Cruciate Ligament Rupture with Generalized Joint Laxity” in Journal of Bone and Joint Surgery 87-B.2 (2005): 800-3.
12. Landing from a jump and changing directions while running are the most common causes of ACL injuries. See Yohei Shimokochi and Sandra J. Shultz, “Mechanism of Noncontact Anterior Cruciate Ligament Injury,” Journal of Athletic Training 43.4 (2008): 396-408.
13. See Hewett et al., “Anterior Cruciate Ligament Injuries in Female Athletes: Part 1, Mechanisms and Risk Factors,” American Journal of Sports Medicine 34 (2006):299-311. And, see Boden et al., “Noncontact Anterior Cruciate Ligament Injuries: Mechanisms and Risk Factors,” Journal of American Academy of Orthopaedic Surgeons 18.9 (2010): 520.
14. This is probably due to a difference in average size rather than gender per se. Women tend to be smaller than men on average and people with smaller bodies have shorter thighs and smaller tibial plateaus over which to distribute stress. See Hewett et al., “Anterior Cruciate Ligament Injuries in Female Athletes,” and Boden et al., “Noncontact Anterior Cruciate Ligament Injuries.”
15. See Shimokochi and Shultz, “Mechanisms of Noncontact Anterior Cruciate Ligament Injury.”
16. See T.E. Hewett, S.J. Shultz, and L.Y. Griffin (eds.), Understanding and Preventing Noncontact ACL Injuries (Champaign, IL: Human Kinetics, 2007), 25.
17. For example, see Iyengar, Light on Yoga, plate 591 on page 421 and plate 19 on page 75.
18. See R.F. Escamilla et al., “Effects of Technique Variations on Knee Biomechanics During the Squat and Leg Press,” Medicine and Science in Sports and Exercise 33.9 (2001): 1552–66.
19. See Morgan PM, et al., “The role of the oblique popliteal ligament and other structures in preventing knee hyperextension” Am J Sports Med. 2010 Mar; 38(3): 550-7.
20. See S.L. Woo, J.M. Hollis, D.J. Adams, R.M. Lyon, and S. Takai, “Tensile Properties of the Human Femur-Anterior Cruciate Ligament-Tibia Complex. The Effects of Specimen Age and Orientation,” American Journal of Sports Medicine 19.3 (1991): 217–25.
21. See Escamilla et al., “Effects of Technique Variations.”
22. Admittedly, this is a gross approximation. Researchers use a lot of computer modeling to make more accurate estimates (for one example, see Kevin B. Shelburne et al., “Effect of Posterior Tibial Slope on Knee Biomechanics during Functional Activity”). While my estimates are coarse, they are not far off what the detailed analyses yield. The whole point of this little exercise is simply to show that the reality of the stress experienced in a hyperextended, static standing posture is nowhere near the tolerance of these ligaments.
23. If you would like to follow some math: The sine of 5° is about .087, the force of gravity is 9.8, and weight on each leg is 30 kgs, which multiply out to ~26N. This is admittedly lower than the findings of K.L Markolf, J.F. Gorek, J.M Kabo and M.S. Shapiro, “Direct Measurement of Resultant Forces in the Anterior Cruciate Ligament. An In Vitro Study Performed with a New Experimental Technique,” Journal of Bone and Joint Surgery, American Volume 72.4 (1990): 557-67], who found passive hyperextension of cadaver’s knees of 5° created about 110N of stress (25 pounds). But, to hyperextend dead tissue may well take more force than what living tissue requires.
24. If you would like to follow more math: The sine of 12° is about .208, the force of gravity is 9.8, and her weight is 60 kgs, which multiply out to ~122N.
25. See J.K. Loudon, H.L. Goist and K.L Loudon, “Genu Recurvatum Syndrome.”
26. The variation in the average angle depends upon which side of the plateau one measures. The lateral side averages 7.2° while the medial side averages 10.7°. See Matsuda et al., “Posterior Tibial Slope in the Norman and Varus Knee,” American Journal of Knee Surgery 12.3 (1999): 165-8. Other studies report less slope but all studies found lots of variation
27. See Kevin B. Shelburne et al., “Effect of Posterior Tibial Slope on Knee Biomechanics during Functional Activity,” Published online 20 September 2010 in Wiley Online Library (wileyonlinelibrary.com).
28. See Terauchi et al., “Sagittal alignment of the knee and its relationship to noncontact anterior cruciate ligament injuries,” American Journal of Sports Medicine 2011 May; 39(5).
29. See “Developing Antifragility in Practice: The necessity of stress” on Yoga International.
30. Given the statistics cited earlier on the average amount of hyperextension for men (5.5°, with one standard deviation of 2.5°), the 95% “norm” is 10.5° or less for men (compared with 12.1° or less for women).
31. The strength of the ACL does decline dramatically with age. See S.L. Woo et al., “Tensile Properties of the Human Femur-Anterior Cruciate Ligament-Tibia Complex."