“One size does not fit all.”
~ responding to William Broad and Michaelle Edwards
One of my first yoga teachers, David Swenson, used to say “there are fears that help us to live and fears that prevent us from living.” When we tell our children, “Don’t touch that stove, it is hot!” we are using fear to help them live their lives safely.
But often fear is used in an unskillful way, which prevents us from experiencing life in its fullness. Fear in teaching can be used skillfully, to draw our attention to potential problems, or it can be dogmatically, to prevent our own awareness from developing and defaulting to blind obedience to rules. When we are children, such rules are important because we do not yet possess the faculties to determine what is safe and what is dangerous, but as we grow older, the journey of life helps us to build our own awareness, a process yielding wisdom.
Unfortunately, fear is often employed in yoga training in a dogmatic, unskillful way.
Recently there has been a number of articles written by William Broad and Michaelle Edwards warning yoga students that certain yoga postures and movements, such as flexing the hips with legs straight, are dangerous, and could lead to significant hip injuries necessitating hip replacement surgery. Is this fear skillful or not?
We can assume both writers have their readers’ best interest at heart: William Broad concludes his recent Elephant Journal article with the wish “Onward to better yoga”. Who can dispute his intention? Michaelle Edwards hopes that “no one should be hurt doing yoga.” [Found in the comment section of this article.]
Again, we would all agree with this. But, while their intentions are honorable, the fear they are creating may prevent many people benefiting from the postures and movements decried.
The articles create the impression that yoga will cause hip damage due to excessive and unnecessary flexion of the hips. Is this a realistic fear that will help student prevent injuries, or is this a fear that will prevent many students from benefiting from these movements?
We have to undertake our own evaluation of the thesis and the evidence provided for it.
That sounds very bad, but it begs the question: did yoga cause these problems?
Mr. Broad implies that this is so, but he uses anecdotal evidence, which is the weakest form of scientific evidence. [For a fuller description of scientific evidence, see my article on How to Critically Analyze Yoga Articles.] Even doctors and scientists fall afoul of the anecdote because individual instances can be very compelling, but anecdotes are are not proof: at best they can point the way towards a line of study wherein proof may be found.
Michaelle Edwards also uses anecdotal evidence to support her thesis that yoga causes hip injuries. She states in the comments section of both my article and William Broad’s article, “Yoga teachers with hip replacements, Beryl Bender, Judith Lasater, Mary Dunn, Dharma Mittra, George Purvis, … the truth be told … Yoga teachers are getting hip replacements.” The clear implication raised by both writers is that yoga can cause hip damage.
There is a big logical leap taken when a writer moves from something can cause damage to something will cause damage, but in the minds of the student, hearing that something can cause damage is scary. We assume that if we do it, it will cause the damage we fear. Ms. Edwards does soften her stance somewhat when she allows, “You may use logic to state that there is no scientific proof that yoga caused [hip damage]. But yoga did not prevent it either.”
Is the fear created warranted? Should it prevent you from doing forward folds in your yoga practice, which Ms. Edwards claims is the cause of the stress in the hips which lead to degeneration?
Let’s look at these assertions more closely and see. As mentioned, anecdotal evidence is the weakest form of evidence in science because there is no breadth to the analysis. We may be able to find one incidence of something occurring, but this does not mean it will always occur. Anecdotal evidence is commonly used by advertisers who use a celebrity to make some claim, but what is not stated are all the other people who did the same thing but did not get the same result. For every yogi that Ms. Edwards’ claims had a hip replacement we can find hundreds of other yogis who did not have a hip replaced. A much stronger level of scientific evidence is correlation: can we correlate yoga practice with hip replacement?
A proper scientific investigation into this question would begin with determining how many women undergo hip replacement surgery in the general population, and at what age. Then the researchers would ask, how many people practicing yoga on a regular basis undergo this surgery and at what age. If the results of examining a large population showed that the yogis were getting hips replaced at a higher rate and/or at an earlier age than the general population, then there would be evidence that yoga practice was correlated with hip damage.
However, if the results showed that the yogis were having fewer surgeries and/or having the surgeries later in life, then the study would be showing yoga practice was inversely correlated with hip damage, and that yoga is good for the hips. Unfortunately, neither Mr. Broad nor Ms. Edwards offer such studies in evidence for their thesis, and perhaps there are no such studies to draw from.
Mr. Broad does utilize anecdotal correlation when he notes that surgeons he has contacted have noticed an increasing number of yoga students coming to them, but is this because yoga is causing hip problems, or is this simply because so many women are now doing yoga? Naturally the number of patients coming to surgeons who do yoga will rise if the number of women doing yoga in the general population is also rising. Mr. Broad is citing an unproven correlation.
In lieu of solid, preferred level of evidence, the writers have relied upon anecdotes. Let’s look at Ms. Edward’s anecdotal evidence. Remember, her thesis is that yoga can cause hip damage therefore one should avoid hip flexions. “Truth be told…” she said, “Yoga teachers with hip replacements, Beryl Bender, Judith Lasater, Mary Dunn, Dharma Mittra, George Purvis.“ She offers no source for this information, so I decided to verify her anecdotal claims directly.
I contacted these yogis.
I first discovered that Dharma Mittra has never had a hip replacement, even though he is 74. Here truth was not told by Ms. Edwards, but perhaps she was confusing a knee replacement Dharma Mittra had with a hip replacement. His knee surgery was necessitated by an accident that occurred 55 years ago while he was with the Brazilian Army. Yoga did not cause the underlying pathology that needed surgery.
George Purvis did have a hip replaced due to an underlying pathology called avascular necrosis. Again, yoga did not cause this condition, but George did report that his surgery was delayed many years beyond the date his orthopedist expected. Perhaps yoga helped delay his surgery and slowed down his degeneration. That is a speculation: we don’t know either way.
Mary Dunn died in 2008 so I could not check with her, but she was a close friend of George Purvis who told me that she had a congenital condition with her hips, which was the direct cause for the need for her surgeries. (This aligns with findings in the Swiss study often cited by Ms. Edwards and Mr. Broad: problems with the hip is most closely correlated with developmental issues of the hip as a youth and not with structural changes later in life, but we will come back to that study in a moment.)
Judith Lasater did have one hip replaced, but if yoga was the cause, she wonders why both hips didn’t need replacing? After all, her yoga practice was not only on one side. Beryl Bender Birch did have both hips replaced but also wonders whether yoga was a contributor to the need for surgery or whether it, indeed, delayed the necessity by years. Like Dharma Mittra, she too suffered an accident years ago that damaged her right hip. In both Beryl’s and Judith’s situations, they were very active in many endeavors such as running, dancing, and high impact aerobics which could also have been a cause of the degeneration of the hip. What is cause and what is effect is very unclear in these anecdotal incidences. No clear correlation is seen.
Ms. Edwards’ admits that she can’t prove yoga caused these problems but she does exclaim that yoga didn’t prevent the damage. Unfortunately, this point is not relevant to her thesis that yoga causes hip damage, but it may not be easy to understand why it is irrelevant. If we change the statement slightly, the illogic is more obvious:
- My mother ate oranges.
- My mother died of lung cancer.
- Eating oranges did not prevent my mother’s lung cancer.
- Therefore—oranges are dangerous and should be eaten with great care.
The first three premises are true, but the conclusion is faulty. To say that yoga did not prevent the degeneration of certain yogi’s hips is not evidence that yoga is dangerous for the hips.
In fact, George Purvis believes yoga was good for his hips and delayed his need for surgery. William Broad also states “dozens of books and articles hailed yoga as great hip therapy. The titles include: Heal Your Hips, Easy Yoga for Arthritis, and Therapeutic Yoga for the Shoulders and Hips.”
If we cannot use anecdotal evidence to show that yoga is dangerous for our hips, what about the Swiss study that both writers cite? This study does use both logical reasoning and correlational analysis, and I admit it is a well-done study. But does it actually support the thesis? In short—no. In fact, the study shows the opposite finding to what Mr. Broad and Ms. Edwards are claiming. I have already pointed this out to Ms. Edwards in the comment section of my earlier article, and I will extract some of these points again here.
The authors of the Swiss study several times mention that the cause of hip problems are pre-existing pathologies of the hip, (similar to Mary Dunn’s situation) not stresses caused by movements. They state this several times as their overarching theory and the whole point of their study. For example, in their abstract they note, “primary osteoarthritis [referred to as OA] is … secondary to subtle developmental abnormalities and the mechanism is femoracetabular impingement [FAI] rather than excessive contact stress.” That is a mouthful to understand but basically, their point is that, not only is the majority of hip OA caused by developmental abnormalities, especially in youth, but almost all problems are also caused by such developmental issues.
They restate this theory in their concluding discussion: “There is an increasing body of evidence that most hips that fail to OA without severe deformities of the joints do so with the mechanism of [impingement] based on subtle or moderate morphologic abnormalities of the joint components.” In other words, impingement is a result of hip abnormalities, not a cause of them.
Once more they reiterate this when they say their “theory proposes … OA of the hip … are indeed caused by minor developmental deformities [which] cause arthritis to develop from … femoracetabular impingement.” [Page 2] They cite several examples where it is the shape of the bones (specifically the acetabulum when it is in retroversion) and the femur (when it has a “pistol grip” shape) that cause the problem. They actually cite several statistics showing how common bone morphology is to the occurrence of OA: “… approximately 90% of the cases of adult hip OA were associated with some developmental abnormality.” [Page3]
The study states, “The pincer-type FAI produces a rather slow process of degeneration and occurs more often in women between 30 and 40 years of age engaging in activities with high demand like yoga or aerobics.” The context here is, however, that these women already had an underlying developmental abnormality and “high demand” exercises tended to bring out degeneration in some of these women.
Of course there are yoga styles that are “high demand,” but there are also many yoga styles that are not aerobic in nature, so it is not clear if they meant to implicate all yoga styles, nor do they imply that all women are in danger. They are talking about women who have hidden developmental abnormalities in their hips. Unfortunately, they do not define what “high demand” means. Interestingly, however, at the end of their paper they do refer to another study these same authors did, where they do make clearer what they mean by “high demand”. I
n that study they said, “High velocity movements, frequently occurring during athletic exercise may play a detrimental role.” Again, not all forms of yoga practice fit this description.
The authors state that impingement is a consequence of development abnormalities; they do not state that it arises from hip flexion, as Ms. Edwards believes. They do admit that for people who have these abnormalities, which give rise to the FAI, then “the most critical motion is internal rotation of the hip in 90° flexion.” [Abstract] They also cite another study in their discussion, at the end of their article, that found 90° hip flexion with internal rotation can increase acetabular rim pathology, but again this is for people who have the predisposition. It is very easy for someone to reach 90° of flexion: just sitting in a chair can give you that.
Should we all avoid sitting with our hips flexed 90°?
Clearly not, for only certain people in certain situations are likely to be in such danger, and again, only if they also add internal rotation to their movement.
Ms. Edwards’ cautions are against women doing deep hip flexions approaching 180° with straight legs, but the researchers do not talk at all about this amount of flexion. Flexion approaching 180° would actually move the place of impingement well away from the anterior, superior pincer positions toward the posterior, superior acetabulum rim, which is not the area that the researchers are worried about.
Also, it is commonly knowledge among yoga teachers that bending the knees in forward folds makes hip flexion easier, not harder. With knees bent, the hamstrings are more relaxed and the pelvis freer to flex. If Ms. Edwards wants to reduce the amount of hip flexion students do, she should be advising them to keep their legs straight! But, fortunately, larger degrees of flexion in the hip moves the possible point of impingement away from the place the researchers are concerned about.
William Broad also cites the culpability of these underlying pathologies in the hip or to the femur and offers links to sites that graphically show this. He doesn’t mention, however, that these pathologies are the cause of the problem, not yoga! In all cases these link show impingement happening to people with pincer or cam lesions. This is an underlying pathology caused, not by yoga, but (according to the Swiss study) by childhood developmental pathologies. And, these animations do not show that flexion is causing the impingement, but rather internal rotation, which is also stated in the Swiss study. (One video highlights adduction as well as internal rotation.) Also, all three videos show only a mild amount of flexion, on the order of ~90°. Flexion by itself does not cause impingement. Flexion of 90° and internal rotation, or internal rotation on its own, can cause FAI, but only for those individuals who have these underlying developmental problems. Ms. Edwards’ thesis that deep flexion of the hip is responsible for hip trauma is not supported by this study or by the videos offered by Mr. Broad.
Mr. Broad also cites Mel Robin: “Mel called bone conflicts ‘largely an unrecognized aspect of our practice’ and said the problem can be avoided ‘if we practice in a more mindful way.’ ” I would agree with Mel Robin – we can all benefit from practicing more mindfully and being aware of compression. But compression itself is not a problem; the problem is too much or too little compression. Too much compression leads to arthritis and degeneration. Too little however leads to osteopenia and a dangerous weakening of the bones. We need to find the Goldilocks’ position. The fear of going too far can result in the opposite problem, of not going far enough. This is why fear based teacher is a poor substitute for teaching a student how to sense her own body and come to know what is right for her. She needs to notice sensations: if a student experiences pain during or after her yoga practice, something is amiss and greater care and investigation is warranted.
The point raised by Mr. Broad and Ms. Edwards could be generalized to be—yoga can cause injuries. That is not debatable: anything can cause an injury. But this is not the same as saying—yoga will cause injuries. Like them, I don’t believe anyone should be hurt doing yoga and all teachers should take care to help their students avoid risky practices, but what is risky for one student is not necessarily risky for another student. We are all different.
This bears repeating—we are all different! This is known as human variation. Yes, there are some people who have significant arthritis in their hip socket, often due to over-stressing the acetabulum repetitively, and stress there could lead to a need for hip replacement.
There are other people at the opposite end of the spectrum that suffer osteoporosis from not stressing the hips enough, and these people do need to get up off their couches and stress that joint to help it regain strength. What is dangerous for one person is therapeutic for another. To say that no one should ever stress his or her hips out of fear of arthritis may lead to osteoporosis! To overstress a joint out of fear of osteoporosis can lead to arthritis.
The solution is not to avoid stress at all costs, but to find out what is the appropriate amount of stress for that joint for that person. This requires that the student be trained to notice the sensations that arise in her practice. The thesis that no one should ever do deep forward bends is not therapeutic for every body. Some people do need to do this in order to get the right amount of stress into her tissues. Others do not. We can’t apply a blanket prescription that will apply to everyone. I know many very flexible students who cannot get a stress at all to their hamstrings unless their legs are straight. For these students, keeping the legs straight is good for them. For many other students, as Ms. Edwards advises, they would be properly advised to bend their knees.
Bending over with legs straight or almost straight and with a good amount of hip flexion is a natural movement for the human body: check out pictures of workers in the third world planting rice. Often you will see them bending over with legs straight or nearly so, and their hips are in deep flexion.
These people have very healthy hips and backs and don’t suffer from pain that someone in the West might have doing the same movements. We sit in chairs. Our backs are weakened, muscularly and facially. We need to take care and bend our knees a lot, but, again, not everybody.
There are many people who have strong, healthy hips and spines and they can move in this natural way.
Beware dogma—the insistence that there is only one right way to do something—and beware teachers who claim, “I know the right way!” Every body is different. It would be nice and easier for teachers if there were only one way that worked for everyone, but due to the nature of human variations we are all unique. What is dangerous for you, with your unique biology and biography, may well be therapeutic for me.
The challenge for the teacher is finding what works for the student who is in front her in this moment. Teaching with fear is rarely helpful. Much better is to teach the student how to determine for herself what works for her. This begins with intention and attention.
We don’t need fear to motivate our students; we do need understanding and awareness.
Ideally, we need to teach students how to feel what is right for her body and develop her own wisdom.
She is unique, and so is what will work best for her.
Love elephant and want to go steady?
Editor: Bryonie Wise