It's been a while since we've thoroughly reviewed an injury, so today we'll be looking at another one of the "big five" most common running injuries. We've already seen how treatment for Achilles tendonitis has been revolutionized by specific eccentric exercises to remodel damaged tendon collagen; today's topic is iliotibial band syndrome, sometimes also referred to as (erroneously, it seems) iliotibial band friction syndrome. It is one of the most common running injuries and seems to be a problem both for recreational runners and for elites, accounting for somewhere between 8 and 10% of all injuries, depending on the study (Marti et al., Taunton et al.) Unfortunately, it's sometimes misunderstood, and there's a good bit of evidence indicating that current treatments centered around stretching, tissue manipulation, and anti-inflammatory drugs are incomplete. As usual, we'll go over some basic anatomy and terminology first, then delve into what the scientific literature has to say about this injury. Like before, I'll also include some common "tricks" runners use to overcome IT band problems, but I'll make it clear what's science and what's hocus-pocus magic.
The iliotibial band, commonly abbreviated as the "IT band," is a long, thick band of connective tissue (most properly referred to as a thickening of the leg muscle fascia) that serves to connect many of the major hip extensors and abductors (gluteal muscles and the tensor fasciae latae muscle) to the lower leg. More specifically, it connects to the tendons of the gluteus maximus, the main hip extensor, and the tensor fasciae latae (TFL—a short, straplike muscle that runs between the top of your pelvis and your femur), a hip stabilizer and abductor, to the top of the tibia, just below the knee. As such, it also helps stabilize and control the knee joint in addition to the hip. Most relevant for runners, it seems to stabilize the hip and knee at footstrike.
As you can see to the right, the IT band runs parallel to the quadriceps muscles and hamstrings. The black arrow points to the most common location of pain: the outside of the knee, just above the knee joint. However, this is not the only location of pain: sometimes ITBS can manifest itself higher up on the band, along the thigh or even near the greater trochantor of the femur. Regardless, the vast majority of ITBS cases involve significant pain on the lateral knee.
This location was widely assumed to be irritated by a small bony protrusion on the femur, called the lateral femoral epicondyle, illustrated to the left. The lateral epicondyle is fairly easy to feel by hand, and indeed the IT band appears to slide across the epicondyle during knee flexion. IT band pain is usually worse when the knee is at approximately 20-30 degrees of flexion, adding to the theory that the cause of IT band pain is friction between the IT band and the lateral femoral epicondyle—hence the name "iliotibial band friction syndrome." However, recent research, MRI imaging, and cadaver studies have called this assumption into question: in 2006 and again in 2007, Fairclouth et al. demonstrate rather convincingly that, as the IT band is really no more than a thickening of the fascia latae, which envelopes the entire musculature of the lateral leg and indeed is firmly attached to the femur near the epicondyle by thick, fiberous issue; it is not anatomically possible for the IT band to "slide" over the epicondyle as if it were a "free" structure like a tendon or ligament. But why is the IT band usually irritated over the lateral epicondyle, and why do patients sometimes respond to cortisone injections in the area? Fairclouth et al. propose that the tissue between the lateral epicondyle, which is comprised of fatty tissue rich in blood vessels and nerve endings, gets compressed by the IT band during running, particularly when the knee is at 20-30 degrees of flexion. While this is interesting in an academic sense, does it really matter to a runner who's got IT band problems?
Interestingly, it may: the distinction between compression instead of friction of the fatty tissue between the IT band and the bony protrusion on the femur may hold the key to its origins. If it was merely a friction issue, it seems that the solution would be fairly mundane: ice, rest, lower volume in training. But some interesting research in the last decade or so has elucidated an interesting possibility: the nerve endings in the fatty tissue between the IT band and the femur, called Pacinian corpuscles, function as proprioceptive feedback units, giving the brain information about what's going in in and around the body. This will become important when we return to the biomechanical origins of IT band problems, so don't forget about this fatty tissue and the nerve endings within!
As mentioned above, iliotibial band sydrome accounts for somewhere in the neighborhood of one in ten of all running injuries. Unlike some issues, IT band problems seem to affect runners at all levels of competition, from recreational runners to elites. Classically, IT band syndrome begins as a sharp or burning pain on the outside of the knee which occurs after a few miles of running. When aggravated, it may eventually become painful with daily activities like walking or descending and ascending stairs. Sitting for a long time also tends to aggravate the IT band, since (as mentioned above) the compression of the fatty tissue above the lateral epicondyle is strongest at 20-30 degrees of knee flexion. Runners find their pain is often greater while running downhill, since again, knee flexion increases during downhill running (not to mention impact forces).
Interestingly, a few sources claim that the IT band is actually less stressed by faster running, since the knee is less flexed at footstrike. While this makes intuitive sense, I suspect that part of this benefit is offset by the significantly greater impact forces created while running fast vs. jogging, so I am very hesitant to recommend (as some do) starting back running from IT band syndrome with short bursts of fast strides interspersed with walking, but if you are feeling particularly confident in theoretical biomechanics, you might consider giving it a shot.
What are the causes of IT band problems? A popular website lists the following as the "most common causes of ITBS" (with no sources cited I might add):
- Leg length differences
- Road camber - running on a slope for a long time
- Foot structure
- Excessive shoe breakdown - particularly it the outside of the heel
- Training intensity errors - increasing mileage or intensity too fast
- Muscle imbalances - particularly quads versus hamstrings
- Run/gait style factors - e.g. bow-leggedness, knock knees, etc.
Other ideas that are floated at some point or another include running on tight turns, excessive downhill running, or running on hard surfaces. The problem is that none of these have much (or in some cases, any) scientific evidence to back them up. There haven't been any rigorous studies that have connected any of these factors (except for 'training errors') with IT band injuries. And even the "training errors" theory isn't helpful—presumably, something went wrong biomechanically speaking, otherwise you would have injured another structure first. What made your IT band the weakest link in the chain?
To answer that, we have to look at the scientific literature. First, though, I should note that you shouldn't discredit factors like road camber, old shoes, excessive downhill or indoor track running, and so on. You may find for your particular case of ITBS, they may be a factor, but they aren't universally recognized. To that end, newer shoes, more varied running surfaces, and so on are never bad ideas. Some of the other factors, though, like trying to correct leg length discrepancies, may be more risky.
Turning to the science, here's the spoiler: the single most important factor in predicting and possibly treating IT band problems is hip abductor strength. Here's a review if you've forgotten what abduction is. To start, we'll look at some retrospective studies. These are the most simple kinds of investigations into an injury's root cause: you gather a group of runners with a particular injury, examine their gait, muscular strength, training habits, and so on to see if you can find anything in common. If so, you can then compare these results to a matched group of healthy runners to see if there is a difference between the groups. While it's easy to see why this alone doesn't prove a cause-effect relationship, it's a good first step in uncovering one.
Hip strength and ITBS
In 2000, Michael Fredericson and his colleagues at Stanford University published such a study. It examined 24 distance runners with ITBS, measured their hip abduction strength, and compared it to that of healthy runners. The injured runners were found to have significantly weaker hip abductors on their injured side compared to the healthy side, and were also found to have weaker hip abductors on both sides compared to healthy runners. The test for hip abductor strength was an isometric strength test, where the subjects were asked to abduct their hip as "hard" as possible against a dynamometer. While not identical to the kinds of stresses put on the hip during running, the abductors do work isometrically during the stance phase to hold the pelvis straight.
A classic sign of weak hip abductors is the trendelenburg gait, where the hips "drop" towards the unsupported side while running. As is illustrated on the left, a "drop" in the hips will necessarily require the stance phase leg to be adducted (moved towards the centerline of the body) moreso than if the hips were not "dropped." Since the IT band is essentially a thick strap of tissue that runs along the outside of the leg, it would not be a stretch (no pun intended) to propose that increased hip adduction increases strain on the IT band. And in fact, two studies confirm this (Ferber et al. and Noehrer et al.). The first, conducted by Irene Davis' lab (no relation) at the University of Delaware, measured hip, knee, and ankle biomechanics in two groups of healthy runners. One group had never had ITBS, while the other had previously been diagnosed with ITBS but had recovered. The subjects ran overground through an array of 3D cameras which tracked the motion of their legs. Using computer software, Davis and colleagues measured the motion of the ankle, knee, and hip joints during the gait cycle. The results were in line with what we'd expect based on our simple model above. Increased hip adduction and knee internal rotation (which would also logically increase strain on the IT band) were associated with a history of ITBS. In their own words:
However, aside from this variable [an increase in rearfoot inversion moment], these results begin to suggest that lower extremity gait mechanics [i.e. foot and ankle] do not change as a result of ITBS. Moreover, the similar results of the current study [...] suggest that the aetiology of ITBS is more related to atypical hip and knee mechanics as compared to foot mechanics. Therefore, the current retrospective study provides further evidence linking atypical lower extremity kinematics and ITBS. (Ferber et al.)
Interestingly, it seems that people who've suffered ITBS seem to pronate less than those who have not—probably ruling out "pronation correction" as a viable treatment option. More importantly, this study highlights that hip and knee mechanics are an important part of IT band issues. But did these changes in biomechanics happen because the runner became injured? Would we be able to predict who might get ITBS if we evaluated a group of completely healthy runners , then waited and observed who got hurt?
Irene Davis' group attempted to answer that question with a 2007 prospective study (Noehren et al.) which was designed as described above: a group of healthy female runners had their running mechanics evaluated using the same overground 3D camera system as above, and over the course of two years, they were followed via email. Some eighteen runners (out of 400 total evaluated at the outset of the study) developed ITBS during the study. When compared with a group of control subjects who remained healthy, the same tendencies were seen: the runners who would later develop ITBS exhibited differences in hip adduction and knee internal rotation. The knee, being a hinged joint between the tibia and femur, can be driven to internally rotate in one of two ways: either the tibia can internally rotate or the femur can externally rotate. Surprisingly, the tibial internal rotation in the injured runners was less than in the healthy group. The net knee internal rotation came entirely from femoral external rotation. Noehren et al. note that the main internal rotators of the femur (i.e. the muscles which should prevent femoral external rotation) are the tensor fasciae latae, the gluteus minimus, and the gluteus medius, which make up the hip abductor muscle group. It was this very same muscle group which was weakened in runners with ITBS in the Stanford study! A doctoral thesis by Alison Brown at Temple University also investigated muscle strength in runners with and without ITBS; interestingly, she found no difference in maximal strength, but a significant difference in endurance. Clearly, hip abductor strength plays a major biomechanical role in the development of ITBS.
Furthermore, John Fairclough's anatomical work on the iliotibial band (discussed above) suggests that the Pacinian corpuscles (a type of pressure-sensitive nerve ending) in the fatty tissue underneath the IT band as it crosses the knee are supposed to prevent excess strain on the IT band. In a healthy runner, the compression of the Pacinian corpuscles by the IT band triggers the hip abductor muscles to fire, reducing strain on the IT band by avoiding excessive hip adduction. But this protective mechanism fails when the main hip abductor muscles are dysfunctional or weak, as he explained to me via email:
The feed back [i.e. compression of the Pacinian corpuscles deep to the IT band near the epicondyle] should stimulate the Gluteus Medius, etc. to fire and stop the pelvis from adduction moment. If however there is an incompetence of the Abductors, mainly G. Medius, then the only major muscle to stop adduction is the Tensor Facia Lata of which the ITB is a part. It can resist abduction forces but is not designed to do so and hence it will over compress the distal tissues which is why on MRI often the fat pad [deep to the IT band] looks oedematous [read: swollen and damaged].
Using data from her lab's prospective study, Irene Davis created a computerized model of the stresses the IT band undergoes during running. Given a standard anatomic model of the skeleton, Davis and her colleges plugged in data about joint motion and impact forces (much like Dr. Casey Kerrigan's study I critiqued a few weeks ago) and computed how various biomechanical factors affected the strain (note I'm using "strain" loosely here; I really mean strain rate) on the IT band (Hamill et al.). As predicted, the computer model showed that the runners who went on to develop IT band syndrome displayed an increased strain rate on their IT band—moreover, the computer model was able to predict which side they would injure, as the model's data output showed a significant side-to-side difference in the injured runners (the injured side, of course, taking on more strain than the uninjured side). Runners who remained healthy showed no asymmetries in side-to-side IT band strain. While the model did not show a particularly strong connection between hip adduction, knee internal rotation, and strain rate, I suspect this was result of an unavoidably-imperfect and incomplete computer model.
At this point, the evidence overwhelmingly points to a biomechanical fault in the abductor muscles of the hip as the root cause for IT band syndrome. Weak or misfiring gluteus medius, gluteus minimus, or tensor fasciae latae muscles are unable to control the adduction of the hip and internal rotation of the knee, leading to abnormal stress and compression on the IT band. This muscular dysfunction manifests as excessive hip adduction and knee internal rotation, both of which increase strain on the iliotibial band and compress it against the fatty tissue between the lateral femoral epicondyle and the IT band proper, causing abnormal stress and damage. But although the pain is coming from the lateral knee, the root of the problem is coming from the hip muscles.
Biomechanical solutions for a biomechanical problem
|Fredericson et al. 2000. Healthy runners avg. 9.7-10.2|
This brings us back to Michael Fredericson's 2000 Stanford University study. Unlike many retrospective examinations of injuries, his study took the additional step of prescribing a six-week strengthening and stretching program. While the usefulness of stretching is questionable (see below), a strength program for rehabilitating the hip abductor muscles is, scientifically speaking, the most sound long-term solution to IT band syndrome. Unfortunately, there hasn't yet been a randomized, controlled trial of runners with ITBS—Fredericson's study didn't even have a control group! But his results were indeed impressive: following the six-week protocol consisting of two stretches and two strength exercises, 22 of the 24 athletes (92%) in the study returned to running. Additionally, their hip strength improved markedly: their strength values after rehabilitation were comparable with healthy runners who did not have ITBS (see figure to right). Is this the randomized clinical trial we need to "prove" the efficacy of hip abductor strengthening exercises? No. But, combining this with the reams of indirect evidence indicating hip abductor weakness as a major factor in the development of ITBS, am I comfortable recommending hip strength exercises as the prime choice for rehabilitation and prevention? Yes.
The following is the program that Fredericson et al. prescribed to their subjects in 2000. It's by no means perfect, and I have some suggested additions which I'll present separately, but currently it's the closest we have to an "approved" hip abductor strength program that's been evaluated in the scientific literature:
The Fredericson Protocol for ITBS:
The exercises prescribed by Fredericson et al. are illustrated below. The stretches, denoted (1) and (2), were performed three times per day, 15 seconds each on both sides. The strength exercises, denoted (3) and (4), started at one set of 15 repeats once per day and built by 5 repeats per day, assuming there was no soreness from the previous day, up to three sets of 30 repeats once per day. The program lasts six weeks. Additionally, nonsteroidal anti-inflammatories (presumably Advil, Aleve, or similar) were prescribed for the first week or so, until pain with daily activities disappeared.
|Click for larger image|
Note that you'll need a stretching rope to do stretch number one.
The athletes were instructed to avoid running for the six-week rehab program, though they could cross-train if it did not cause pain. As mentioned earlier, 92% of the athletes in the study recovered after this rehab period (though the lack of a control group makes it impossible to say why they got better: the 6 weeks' rest, the hip strength, the stretching, or the anti inflammatories!).
Recommended additions to the Fredericson protocol
I recommend doing three additional exercises for hip strength. Two of these I got from Dr. Rob Johnson, an orthopedist and practitioner in the Twin Cities (who incidentally authored one of the first major studies connecting hip muscle weakness to running injuries), and the third I got from a physical therapist in the Twin Cities area. I think these three are important additions because the Fredericson protocol lacks any external rotation component (hence the "clamshells") and also lacks any isometric exercises (hence the glute bridge and wall press). While the wall press doesn't carry the "stamp of approval" of any published researchers I know of, I doubt it'll be highly controversial—if you don't trust me, then axe these ones and just do the above protocol. For lack of anything better, you can start with 1x15 and build to 3x30 with these strength exercises as well. These can be done with a Thera-Band looped around your knees for increased difficulty.
|Adopted from mckinley.illinois.edu/handouts|
*Clamshell leg lifts
Very similar to side leg lifts (exercise 3 in the Fredericson protocol), but the knees are bent, meaning your legs "open" like a clam, externally rotating instead of simply abducting. As with the side leg lifts, go slowly and ensure the hips are straight above each other, not tilted forward or backwards.
|Adopted from uptownwomenscenter.com/topics/|
*Pelvic tilt into glute bridge, 5sec hold at top
This exercise is done lying on your back with your knees bent. "Tilt" your pelvis up by drawing your stomach in, then, keeping your pelvis "up," use your glute muscles to go into what's called a "glute bridge" and hold it for 5sec, then lower back down and "untilt" your pelvis.
|adopted from mikereinold.com|
*Isometric wall press 15x5sec
Stand perpendicular to a wall, with your shoulder, hips, and foot against it. Raise your inside leg up so your thigh is parallel to the ground and your knee is bent at 90 degrees (like you were stepping up onto a bench). Then use that raised inside leg to push against the wall. You should feel it in your glutes on the OUTSIDE leg as it resists (isometrically) the pressure from your inside leg/the wall.
Other treatments: stretching, rolling/manipulation, and more
Hip abductor strength is by far the most scientifically supported treatment for ITBS, but most doctors, trainers, and physical therapists will recommend stretching, rolling/manipulation, or both in addition (and unfortunately, sometimes instead of) hip abductor strength training. The evidence for these treatments is much more limited and their utility is much-debated among the research community.
While it's often reported that the IT band is "as strong as soft steel" (and is thus impossible to stretch), this is a bit of a mischaracterization. From what I can tell, that statement comes from a 1931 paper (Murray) which investigated the tensile strength of the fascia lata, the sheath that envelops the lateral muscles of the thigh (recall that the IT band is simply a thickening of this tissue). Indeed, the fascia lata shows a tensile strength of about 7800 psi (54 MPa in metric), which is comparable to a strong plastic or a soft metal. But this is the ultimate strength of the fascia lata—the point at which it breaks! Inside the body, it's reasonable to expect that it ought to remain in the elastic portion of its stress-strain diagram. The fascia lata stops behaving elastically (i.e. returning to its original dimensions after being stretched) at about 2200 psi, more in line with your average plastic or hard rubber. Either way, the fascia lata appears to be at least somewhat "stretchy." Whether stretching actually accomplishes anything is another issue.
In an actual experiment which involved performing IT band stretches on fresh cadavers with microstrain gauges implanted into the IT band (surely a riveting endeavor!), Falvey et al. report that the IT band itself is not particularly amenable to stretching, but mostly due to its numerous attachments to the femur, not its stiffness. Instead of focusing on the IT band itself, Falvey et al. recommend stretching or manipulating the main muscles which control the tension on the IT band—the tensor fasciae latae and the gluteus maximus:
The longitudinal and firm attachment (0.3mm average thickness) of the ITB to the full length of the femur means that the potential for physiological lengthening is limited [...] These findings highlight the tensioning role of gluteus maximus (working synergistically with TFL) in ITBS, concurring with other studies noting the substantial contribution of gluteus maximus to the ITB [...] The anatomical evidence for the current treatment regimens discussed previously [stretching the IT band and eliminating inflammation over the lateral femoral epicondyle] appears insufficient. While local treatment measures may have role in temporarily easing symptoms, they appear to treat symptoms rather than cause. Given that it appears that the muscular component of the complex plays an important role in tensioning of the ITB, treatment should be directed at TFL and gluteus maximus.
So presumably, any stretching should be directed at the tissue surrounding the IT band, not the IT band itself. I'm no expert on stretching, but it's easy to prove to yourself that stretch number (1) from the Fredericson protocol puts a good stretch on the gluteus maximus and that stretch number (2) is an effective stretch TFL stretch as well. I suspect the benefits of stretching have to do with loosening this peripheral muscular tissue, not the IT band proper.
|From Falvey et al.|
Likewise, any soft tissue manipulation (be it massage, Active Release Technique, Graston, or just plain old foam rolling) should also be directed at the tissue around the IT band. Knowing what we do about the irritation of the fat pad sandwiched between the lateral femoral epicondyle and the IT band (illustrated right), you should avoid any sort of "manipulation" of that area, since you'll only be making the problem worse. While there's very little scientific evidence to support soft tissue manipulation, many veteran distance runners (including yours truly) will swear by their foam rollers as a great weapon against IT band syndrome. Some even switch to a hard 3" PVC pipe if the foam roller feels too soft! While anecdotes aren't worth much scientifically speaking, it's awfully hard to argue against the experience of probably thousands of runners. My best guess is that foam rolling, massage, and the like break down scar tissue and/or muscle adhesions in the quadriceps, glute, TFL, and hamstring muscles that the fascia lata surrounds and connects to. But whether "scar tissue" and "muscle adhesions" even exist is a contentious issue, so I'll leave this as a guess!
Finally, I should mention anti-inflammatories and local corticosteroids (delivered either with an injection or via iontophoresis). While these can be good methods to control the swelling and reduce pain on the outside of the knee, you do have to address the root cause of the problem, which is likely to be hip abductor weakness and dysfunction. Don't completely write off the lateral knee pain, though—remember my first injury maxim is that pain represents real, physical damage: so if the outside of your knee hurts, it's probably been damaged! So avoid painful activities. You're only compressing and further damaging the sensitive tissue deep to the IT band. Treat the symptoms and the cause.
We've seen how IT band is an often-misunderstood running injury. The common conception is that the IT band is analagous to a thick cable which is irritated due to friction between the band and a bony protrusion of the femur just above the knee. From this point of view, the most logical treatment options are addressing the inflammation and irritation on the outside of the knee. But upon examining the scientific evidence, we've seen that it's anatomically impossible for the IT band to "slide" back and forth across the femur. Instead, the IT band and the highly innervated tissue beneath it is biomechanically linked to the hip abductor muscles. During running, the abductor muscles are supposed to control and limit hip adduction and knee internal rotation. When the abductor muscles, particularly the gluteus medius and minimus, are weak or dysfunctional, the body cannot control these factors—this is why runners prone to iliotibial band syndrome exhibit weak abductor muscles and abnormally high hip adduction and knee internal rotation while running. Although a randomized clinical trial showing the efficacy of hip abductor strengthening on treating and preventing iliotibial band syndrome has yet to be published, the preliminary and indirect evidence is very encouraging.
Until new research comes out, the best treatment plan for iliotibial band syndrome must include hip abductor strengthening. The six-week Fredericson protocol, detailed above, includes two such exercises. I've suggested three additional exercises to bolster the program with external rotator and isometric strength exercises. Additionally, some clinical studies (most notably Fredericson's) as well as anecdotal evidence support using stretching, rolling/manipulation, and anti-inflammatory drugs. In the case of stretching and rolling, these should be targeted at the tissue surrounding the IT band: for rolling and manipulation, the quadriceps, hamstrings, tensor fasciae latae (TFL), and glutes; for stretching, primarily the TFL and gluteus maximus.
One question I'm sure will come up is, "Do I have to take six weeks off from running for these strength exercises to work"? And as you might expect, I can't give a straight answer on that. If you've caught your case of ITBS early, probably not. You may even be able to get away with only a couple days off. But, at least in my experience, it takes about a month of doing hip strength exercises every day to start seeing significant results. This doesn't necessarily mean you have to take a month off from running, but it does mean that you should be patient if you aren't getting results quickly. In any case, any running injury that doesn't resolve in a week or so usually warrants a trip to the doctor, who may be able to help you identify some of the particular causes of your specific case of iliotibial band syndrome. Best of luck!
(in order of appearance)
1. Marti, B. et al. On the Epidemiology of Running Injuries. Am J Sports Med 1988; 16(3):285-94
2. Taunton, J.E. et al. A retrospective case-control analysis of 2002 running injuries. Br J Sports Med 2002;36:95–101.
3. Fairclough, J. et al. The functional anatomy of the iliotibial band during flexion and extension of the knee: implications for understanding iliotibial band syndrome. Journal of Anatomy 2006;208(3):309-316
4. Fairclough, J. et al. Is iliotibial band syndrome really a friction syndrome? J Sci Med Sport. 2007 Apr;10(2):74-6.
5. Fredericson, M. et al. Hip adductor weakness in distance runners with iliotibial band syndrome. Clin J Sport Med 2000;10:169–175.
6. Ferber, R. et al. Competitive female runners with a history of iliotibial band syndrome demonstrate atypical hip and knee kinematics. J Orthop Sports Phys Ther 2010;40(2):52-58
7. Noehren, B., Davis, I., and Hamill, J. Prospective study of the biomechanical factors associated with iliotibial band syndrome. Clinical Biomechanics 2007;22:951–956
8. Murray, G. Tensile strength and elasticity tests on human fascia lata. The Journal of Bone and Joint Surgery 1931; 13(2): 334.
9. Falvey, E.C. et al. Iliotibial band syndrome: an examination of the evidence behind a number of treatment options. Scand J Med Sci Sports 2010;20:580–587
10. Hamill, J. et al. A prospective study of iliotibial band strain in runners. Clinical Biomechanics;23(2008): 1018–1025