October 26, 1998


The role of iliopsoas in neck and back pain & its treatment

Pointers, Journal of the Shiatsu Association of Australia, 1998.

An exploration of the global significance of this muscle group, both posturally and in terms of its role in neck and back pain, dynamically and statically.

Anyone who knows anatomy will be able to tell you that iliopsoas spans the anterior surfaces of the transverse processes and the femur, attaching just below the greater trochanter – but, really, what does this tell us in a functional sense? In respect of neck or back pain, why should iliopsoas be given any more attention that, say, gamellus or any of the other 600-odd muscles around the body?

Lumbar lordosis

Before answering that question, let us reflect for a moment on the shape of the spine: as seen from the side, it exhibits the characteristic three curves, the cervical lordosis, as the rearward-facing concavity is known, the thoracic kyphosis, as the rearward-facing convexity of the upper and middle back is known, and the lumbar lordosis, as the rearward-facing lumbar concavity is known. The usual explanation for an excessive lumbar lordosis is that the abdominal muscles are insufficiently strong, which (on this account) lets the brim of the pelvis tilt forward, increasing the lumbar curve. Additionally, any patient with prominent or well-developed gluteus maximus (like a dancer or weight lifter) will often have the shape of their back labelled ‘sway back’ or (more technically) lumbar lordosis (or ‘hyper-lordosis’) by various practitioners because of the illusion of greater than normal curve caused by the shape made by the buttocks joining the lower back. Let us put this myth to rest here: the lumbar lordosis is one of the normal, and necessary, adaptations of the spine, thought by many anatomists to be the primary shock-absorbing mechanism of the body (in respect of longitudinal loads).

Abdominal muscles

Let us return to the role of the abdominal muscles in the shape of the lumbar curve for a moment. It is usually argued that this shape results from weak abdominals, but try this quick test. Stand up, relax, place one hand over your navel, and lift one knee to the chest. Can you feel the abdominal muscles working under your hand? Now try walking around, with the body relaxed. Are these muscles doing any work at all? The answer is no. Of course, it is possible to walk, sit, and stand with the abdominal muscles clenched, but this is far from normal, and in fact is characteristic of a chronically stressed person, and one of the first diagnostic clues when doing hara diagnosis. The point is, even if you do feel tension in these muscles (perhaps your normal way of being in the world) repeat the tests trying to keep those muscles relaxed and you’ll see that you can. The fact is that in much of normal daily life the abdominal muscles do little work apart from providing the small amount of tension required to hold the abdominal organs in their normal position (and some people avoid even this minimal amount of work), and hence cannot play a major role in shaping the lumbar lordosis: no tension means no effect, after all.

Static vs. dynamic

What about dynamic movements, or during lifting events? In dynamic activities (like running or other fast sports) the abdominal muscles are working all the time: they brace the trunk, so other muscles can be brought into play to move either the body itself, or specific limbs, against the inertia of the body, as when you throw or kick a ball, for example. During lifting, the abdominal muscles play a major role: using an automatic, reflex response called the Valsalva manoeuvre (named after the Italian anatomist who first discovered it) we all take in a breath, close both the throat and the anus, and by tightening the abdominal muscles we turn the entire trunk into an ‘inflated splint.’ Some anatomists claim that this reflex decreases the compression experienced by (say) T12 by up to 50%, compared to the theoretical forces that would otherwise be experienced by this vertebra (see Kapandji, for example). So for these reasons we need strong abdominal muscles, but in the majority of daily life they play no significant role in the shaping of the lumbar curves. So which muscles do?

Effect of tight hip flexors

Please refer to the accompanying illustration (p. 39 in second edition, ONBP; p. 63 in the third edition, and p. 69 in the current, fourth edition). As it is presented on the page, the quickest and easiest test of the length of iliopsoas it to lie down face up, with your legs stretched out on the floor. Can you press your lower back to the floor with the back of the legs held onto the floor? If you cannot, suspect tight iliopsoas. The bottom illustration shows why people with tight iliopsoas (and for a moment I am only considering bilateral tightness in these muscles) feel discomfort in the lower back when lying with the legs outstretched: the lower back is literally pulled from the floor by these muscles if they are not loose enough to let the legs rest on the floor. Please turn the page you are reading 90 degrees, so the figures represent a standing person: now you can see the effect iliopsoas has on the shape of the lower back when you look at the leftward-most depiction. Tight iliopsoas is the primary determinant of the shape of the lumbar curve, and people with tight iliopsoas display both an exaggerated lumbar curve and a protruding abdomen. But the effects are not in fact confined to this area of the body, and what I will develop now is one of the reasons we identify iliopsoas as a ‘sufficient’ cause (a term used in philosophy to denote a cause which, by itself, is enough to make something happen) of not only low back pain, but middle back pain and neck pain too. Let me explain.

The spine adapts

If iliopsoas is tight, the entire spine must adapt. The typical way the body achieves this is, as already mentioned, to allow the lower part of the spine to be pulled forward, and some textbooks show this. But if this occurs, the thoracic spine usually increases its curvature too, as the body seeks to keep itself balanced over its centre of gravity, and as this occurs the cervical spine increases its curve in order that the head be held roughly level. We are still talking about each of the iliopsoas pair being tight (bilateral) as a cause. If only one of the iliopsoas pair is tight, these changes still occur, but are made even stronger by the addition of small rotations at all segmental levels. To understand why this occurs, we need to know something of the extraordinary complexity and subtlety of the engineering of the spine. If the pelvis is tilted to one side, as in the case of an actual leg-length difference, one ankle pronating significantly more than the other, or when one half of the pelvis is smaller than the other (‘small hemi-pelvis’), then all lumbar vertebrae will be slightly rotated away from the lower hip side (‘contrarotated’), all the thoracic vertebrae rotated towards the short leg side, and all the neck vertebrae rotated in the same direction as the lumbar spine. This is in fact how lateral flexion is achieved in the normal spine: lateral flexion is achieved in part by momentary wedging of the intervertebral discs, by vertical and lateral movements of the facet joints, and by contrarotation away from the direction of the flexion. Students interested in learning more about this fascinating and complex movement will find a brilliant exposition in Kapandji’s The physiology of the joints, volume three, pp. 53ff.

Other indications

So, following the movements of the spine, we find that patients demonstrating a tilted pelvis almost always carry the pelvis slightly rotated with respect to the plane of the feet , with the body in the anatomical position. This is something we look for when examining patients: stand to one side and (with the feet in the anatomical position, 300mm or so apart, and inside edges parallel) look at the plane represented by the heels, and look then towards the line of the hips. One hip will be carried further forwards then the other. Normally (although not always) the higher hip will be carried forward. When we test the tightness of iliopsoas with respect to each other, we usually find that iliopsoas of the lower hip’s leg tests tighter. Visualise this area of the body for as moment: if one hip is lower, the spine is laterally curved away from this hip (i.e., it bends back towards the centreline of the body) and the individual vertebrae are rotated towards this hip (contrarotated away from the induced curve) and accordingly the transverse processes similarly rotated. Thus the origin and attachments of iliopsoas are closer on the lower hip side and this is what the body considers ‘normal,’ and one of the reasons iliopsoas tests tighter on this side.


We have a number of innovative exercises to loosen tight iliopsoas, but space does not permit their description. We strongly recommend self-stretches done with appropriate abdominal bracing and strong contractions of the relevant gluteus maximus in the ‘tail-tucking’ action (to counteract the normal tendency of these muscles to hyperextend the spine).

On our workshops we pay close attention to the requirements of i) making sure that the hip whose iliopsoas the patient is trying to stretch is in front of the other hip during the movement (as might be seen from above) to focus the stretch on iliospoas, and ii) an arm is used on the front leg’s knee to brace against, and keep the trunk vertical, and iii) the abdominal brace together with the tail tuck already described. The complexity of these instructions is necessary, as the body will defeat one’s best attempts to stretch these (or any) tight muscles.

In this brief note, I have selected only one of the many causes of neck and back pain, the subject of our six-day intensive workshops. Fuller details of the rationale touched upon can be found in my book, Overcome Neck & Back Pain. Only so much can be covered in words and pictures, of course, and we hope to see more practitioners who deal with these common problems at the workshops in the months to come.

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}