March 1, 2001


Hidden causes of back problems

Hidden causes of back problems

Journal of the Royal Australian College of General Practitioners.

This is a technical article on the ‘hidden’ causes of back pain, written for doctors. Referenced.


In this note, I wish to share the results of workshops I have been running around Australia since February this year, 2000. These workshops offer the approach to treating neck and back pain outlined in the book [Overcome Neck & Back Pain (Simon & Schuster, revised 4th edition). The book advocates a structural analysis followed by a functional analysis, and treatment using an exercise-based approach. Approximately 1,350 people have attended the various workshops to date. All were long-term neck or back pain sufferers, all of whom had sought treatment previously from physiotherapists (around 30%), doctors (35%), chiropractors and osteopaths (over 60%), and a variety of other treatments. Most had sought treatment from more than one kind of practitioner. In the vast majority, any relief had been temporary (a day to a few days the most common response) and most participants expressed the desire to learn how to look after themselves, preferably freeing themselves from the necessity of seeking regular treatment.

Significant causes

In order of significance, the most important causes of back pain identified have been an actual leg-length difference, with or without tight iliopsoas. These muscles have been found to be tight in absolute terms (45%, using the standard test, following Kendall) or tighter on one side (55%; this includes a proportion of the previous figure). In order of frequency, the muscles in which the pain is experienced are quadratus lumborum and erector spinae when the pain is lower back and one sided, between the hip and the spine (around 60%), the fascia into which latissimus dorsi inserts when the pain is lower (10-15%; including sacro-iliac joint and ligament pain), and piriformis, about which more below. With respect to neck pain, the most significant cause and source of pain is levator scapulae, and the most important cause of referred pain in the arm and hand have been found to be the scalenus group.

Muscles, nerves, and pain

The assumption of the workshops is that the main sources of dysfunction (both in reduction in range of movement and pain) are the muscles (and sometimes nerves) of the body. The further assumption is that the pain is a combination of inaccurate body image (by this I mean that the stretch reflex is triggered inappropriately early in the range of movement, for a variety of reasons including protection patterns held over from an earlier trauma or other causes) and muscle tension, also deriving from a variety of causes, to be addressed below. Following the structural analysis described below, stretching exercises (or parts of standard exercises) are used diagnostically to compare key functions. In the vast majority, the location of the pain is in the tighter of paired muscles. Whether this is a result of pathology or is itself a pathology is a question to which I’ll return. Treatment consists of precise stretching exercises, usually parts of conventional exercises, using the Contract–Relax (C–R) approach to increase range of movement, reduce muscular pain, and to remake dysfunctional patterns and body image.

Pathology: a second look

One further assumption is that pathology identified in any reports brought along by the participants is not of the severity wherein surgical intervention is either recommended or likely to be effective. Most participants with sciatica, for example, have been told that the severity of any identified pathology did not indicate surgery. A small number (4 individuals; n=1,180) were seeking to avoid surgery. A significant number (110 individuals; n=1,180) had been informed that no significant pathology could be identified to explain their problems, and the most common recommendation in these cases was to seek a pain control clinic, or to use relaxation techniques or similar to learn to live with the pain. It is now well known that disc and joint pathology is normal, in the sense that two-thirds of a researched non-back pain suffering population displayed prolapsed or bulging discs, or serious joint pathology. Many had pathology at more than one vertebral junction. The researchers concluded that had any of these people been suffering back pain and had gone to their doctor and had the MRI analyses done, the conclusion would have been that the pathology identified was the cause (Jensen et al., 1994). A similar study a few months later for the neck showed virtually identical percentages. It is clear that pathology may exist benignly. It also seems possible that other causes may render pathology significant (in the causal sense), and one task is to identify such mechanisms. An actual leg-length difference is one such cause.

Leg-Length Inequality (LLI)

The workshops are providing evidence to support the claim that leg-length difference is significantly over-represented among back and neck pain sufferers, compared to the results of various studies done over the last 20 years. The standard figures are that 10–14% of the population have a leg-length difference of 9-10mm or more (see, for example, Giles & Taylor, 1985 or Rock, 1988). From the raw data provided in the radiological studies considered, calculation shows that differences in leg-length of 5mm or more are found in around 55% of the general population. Until now, the assumption has been that differences less than 10mm are unlikely to be significant. My work with athletes over the last ten years has suggested that smaller differences can be highly significant, this perspective being based on the displayed asymmetrical muscular development that cannot be attributed to any asymmetry of patterns of use. Such athletes demonstrate additional development in erector spinae on short leg side and additional development of the paravertebrals in thoracic spine opposite short leg most commonly, evident in athletes whose sport or training is primarily performed in the vertical load-bearing position. The inference here is that the body can be considered a ‘map’ of the forces that have acted upon it, and that a visual comparison of the shape of the muscles permits conclusions to be drawn about how the individual’s body uses itself. In one notable example, an experienced triathlete with a barely-discernible leg length difference (two or three millimetres) displayed clearly visible differences in the muscles mentioned. His weekly training schedule included 140–160km running. In a symmetrical activity like running, symmetrical morphology is expected. If asymmetrical morphology is found, one question to be answered is what forces might have produced the asymmetry.

Asymmetrical morphology

In the workshops we have seen similar asymmetrical development in the majority of attendees, few of whom have been athletes, and on questioning the vast majority did not regularly engage in asymmetrical activities. On testing (outlined below) around 75% displayed a tilting of the pelvis in the coronal plane, and many displayed a subtle recapitulation of the three normal curves of the spine in the coronal plane following the obliquity. This is Travell and Simon’s ‘S’-shaped adaptation. The ‘C’ shaped adaptation (see Travell & Simons, 1992, p.54) appears to be much less common, and is found more often in individuals whose whole-body flexibility is poor. The assumption is that tilting of the pelvis (with consequent alteration to the ideal relation of the spine to the forces of gravity) is the significant factor, which may be due to factors other than uneven length of the bones of the leg. This distinction is made because at least three causes of pelvic obliquity are not usually considered in measurements of leg length (a single pronating ankle, asymmetrical placement of hip joints in pelvis, and small hemi-pelvis). Additionally, hip width may be significant, as any particular leg length difference will tilt the pelvis further in someone with narrow hips. For these reasons, and the need to see how the body resolves the forces of gravity in the anatomical position, we favour a standing test of leg length. Many factors can vitiate lying leg length tests, a position elaborated in chapter eleven of my book.

The ‘Eyeballing’ approach to assessment

At the workshops, we begin with visual inspection of the whole body, looking for ankle pronation, symmetry of hips with respect to shoulders, level of shoulders, and placement of head. Comparison of left and right waist indentations are useful for women. We than ask the participant to stand with a small plastic block of 12 mm thickness under each heel in turn. When the block is under the longer leg, the distortion to the symmetry is obvious, both to observers and the participant. When placed under the shorter leg, the patient looks and feels more balanced. If the use of the block causes equal or similar distortion to the shape of the body on both sides, no significant difference is the conclusion. If a difference is found, a heel lift correction of just less than half the estimated difference is recommended. The reason for not recommending correcting the whole of the estimated difference is that we do not wish to render and adaptations made by the body possibly maladaptive. The thickness of the correction can be increased at later date if its use appears to help the problems; more likely if the attendee’s work is performed in the vertical load-bearing position.

One relationship between strength and flexibility

With respect to asymmetrical morphology of the muscles of the lumbar and thoracic spine, one additional observation needs to be fleshed out. Over the 18 years of running stretching exercise classes at the Australian National University (around 200 new students per semester), we have found an inverse relationship between strength and flexibility. For example, in the normal healthy pain-free adult, the standard tests of shoulder flexibility have revealed that the vast majority of right-handed people have reduced shoulder joint flexibility in all planes of movement. As far as the whole body is concerned, if the flexibility of any paired muscle group is measured, we have found that (in the individual who does not include stretching exercises in their normal routine) the stronger or visibly better developed muscle usually tests tighter. With respect to individuals who display asymmetrical muscular development following pelvic obliquity, we have found that the main loci of one-sided low back pain in the majority corresponds with the tighter half of certain pairs of muscles. Quadratus lumborum and erector spinae, and occasionally the obliques and the inferior-most fibres of latissimus dorsi are involved. In the majority of patients, the pain has found to be located in the tighter of a pair of muscles.

Tests of functional flexibility

Following structural analysis, four tests of functional flexibility are performed by the group, in pairs. Working in pairs facilitates good form in the tests, and teaches the participants how to correct the body’s natural tendency to avoid stretching the muscles that are tight. The tests are right/left rotation, right/left hip flexion (with both bent and straight legs), right left/lateral flexion, left/right hip extension, and left/right piriformis.

Improving the Straight leg Raising Test (SLRT)

If the attendee complains of sciatica, a supine straight leg-lifting test is performed and followed with a suitable piriformis exercise. The straight leg-lifting test is then performed a second time. What is most often found is that the leg can be moved further into the ‘normal’ range (usually reckoned to be in the 65–70 degree range in this test), and that the pain that was associated with the sciatica is now felt in the calf (less often, the hamstring). An additional stretching exercise is done for the indicated muscle, and the limb re-tested. In the majority of cases, the person who had been diagnosed with sciatica can allow the limb to be moved into the normal range after these exercises have been done. This approach can reduce false positives in this common test.

Common patterns

The most often found pattern of flexibility associated with a leg-length difference are hip and hamstring muscles test looser on the shorter leg, rotation of the lumbar spine looser when the shorter leg is taken to the floor in the ‘lumbar roll’ test, lateral flexion is tighter away from the short leg side, and the hip flexors are tighter on the short leg side. These patterns are not immutable, however, and we advise attendees to identify their own patterns and treat these, by beginning their stretching workouts with the tighter of the pairs, and finishing with the tighter pair and thus doing more work for these muscles over time.

Specific muscular causes

We have found tight hip flexors correlated with central low back pain, and a single tight hip flexor correlated with one-sided quadratus lumborum pain. As mentioned, the tighter hip flexor is usually found on the short leg side, but this is the least robust of the patterns. Tight hip flexors have been found to be highly correlated with an increased lumbar lordosis, often in association with bilateral tightness of quadratus lumborum and erector spinae. If the hip flexors (iliopsoas and rectus femoris) are tight, the leg will not extend past the line of the body (as is needed when walking, for example) without rotation of and/or anterior tilting of the pelvis. A suggested mechanism for pain in these patients is increased pressure on the facet joints together with the muscular causes mentioned. Lumbar flexion exercises usually provides immediate relief of the pain. Long-term tightness may contribute to increased kyphosis and additional compensating cervical lordosis, caused by the body’s need to carry its weight over the balance point. Effective stretching of the hip flexors is difficult: unless the trunk is braced, as stretching these muscles extends the lumbar spine, frequently eliciting the participant’s pain.


Another often-unsuspected cause of back and hip pain is one of the external hip rotators, piriformis. In about one fifth of the population, the peroneal branch of the sciatic nerve passes directly through this muscle rather than passing inferiorly to it (Travell & Simons, 1992, vol.II, p. 186 ff.). If this muscle is in spasm or simply very tight, enough clamping force can be produced on this branch of the sciatic nerve to cause pain in the muscles behind the hip joint and sciatic pain down the back of the leg. Typical histories include long periods of sitting.

This pain can feel the same as sciatica caused by disc impingement and difficult to distinguish from it from symptoms alone. If the patient has disc pathology a diagnosis of nerve-impingement induced sciatica may be made without there being any casual relationship between the pathology and the symptoms. I have developed an effective test and stretch for this area and, on average in past workshops, up to half the people present demonstrate this problem. The relief can be immediate and dramatic. See exercises 5, 6, and 7, and a practitioner-applied version, 16, in the book. In a future note, I shall consider the role of leg-length inequality and patterns of muscular tension in neck pain, and referred pain phenomena such as RSI and carpal tunnel syndrome.


Giles, L. G. F. and Taylor, J. R., 1985. Low-back pain associated with leg-length inequality. J. Aust. Chiropr. Assoc., 15: 135-145

Laughlin, K., 1989. Low back pain: review and prescription. In Is our future limited by our past? Freeman, L. (ed.). Proceedings of the third conference of the Australiasian Society for Human Biology. The Australasian Society for Human Biology, University of Western Australia

Jensen, M. C., Brant-Zawadski, M. N., Obuchowski, N., Modic, M.T., Malkasian, D., and Ross, J. S., 1994. Magnetic resonance imaging of the lumbar spine in people without back pain. New England Journal of Medicine, July 14. 331, No. 2: 69-73

Kendall, H.O., Kendall, F.P., and Wadsworth, G.P., 1971. Muscles, Testing and Function. 2nd edition. Williams and Wilkins, Baltimore. Rock, B. A., 1988. Short leg-a review and survey. J. Aust. Chiropr. Assoc., 18: 91-96

Travell, J.G. and Simons, D.G., Volume1, 1983; Volume II, 1992. Myofascial pain and Dysfunction: The Trigger Point Manual. Williams & Wilkins, Baltimore

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