Lumbo-Sacral Dysfunction and Treatment


Here are my thoughts on how I conceptualise lumbo-sacral segment dysfunction (LS) and approach the treatment of this area.

Of course, these ideas are hypothetical, not theoretical, being a set of empirical observations that I’ve tried to put into a framework.

Hopefully, you will find them interesting and already see these patterns in your practice. It might be worth a look at this article I wrote back in 2002 for some context on the terms and definitions I will use later.

I’ll gradually (and hopefully) add some videos over the next few months.


But first, some context about where I learned osteopathic technique. I studied at the UCO (formerly BSO – I use the two interchangeably) between 1994 and 1998 and did some technique tutoring for a couple of years after. I learnt more from teaching and practice life than I did during my undergraduate course, but this is only natural and I was building on the UCO’s comprehensive and well-taught technique syllabus.

Looking back, it seemed to me that the model of spinal functioning taught at the BSO at this time had two primary features;

  • Palpation and passive movement could reveal what one wanted to know about the ‘state’ of a spinal joint. Active movements were not so necessary and were more used to see global or regional ranges of movement and pain states.
  • HVT manipulation of the affected spinal joint did not need to be direction specific. Any thrust technique to the target joint would help as well as any other.

A feature of the BSO approach at this time was its lack of specificity. It is important to emphasise that the BSO tutors were aware of this and that this was seen as a positive thing in that it was more ‘pragmatic’ or ‘realistic’. So my comments in this respect should not be seen as critical or pejorative – on the contrary.

Special Areas

My first few years ‘on the job’ initiated some doubts about the above, particularly for the very top and bottom of the spine. I started to feel that there was something special about the upper cervical spine, the lumbo-sacral area and – possibly – about the 1st rib. In what ways?

  • The way that they lesioned.
  • The manner in which the lesion type affected active movements at that segment.
  • The manner in which the lesion type affected movements and function further afield.
  • And the nature of the technique required to correct a specific type of lesion.

This article is about the lumbo-sacral area. I will try at some point to cover these other ‘special’ areas. But for the rest of the spine (middle and upper lumbars, thoracics, lower and middle cervical areas, I felt that the BSO model broadly worked well, even for the junctional segments of C7-T1, T4-5 and T12-L1. These latter were special only in the technique challenges they posed, not because they behaved differently.

I appreciate that others have thought and written about these issues. I’m not trying to claim any originality here, but just to express a speculative world-view after 20 years or so of practice.


Before we start, a word about pain. All the following (logic, examples, and so on) is about patients with no pain. Pain will ‘confuse’ observation, analysis and choice of treatment for all of us, whatever we think about spinal function, including the way in which I do as described below.

For example, pain might cause conscious or unconscious antalgic movements. Or the peculiar disruption to movement caused by a problematic disc.  Even some swelling inside a facet joint will completely disrupt movement at that level.

So to keep things ‘pure’, the following assumes a pain and inflammation free patient.

You might ask ‘But surely most patients are in pain?’. Yes, but hopefully not after you have treated them and got them back in to;

  • Check on progress
  • See in what way they are now functioning differently (especially through active movements).
  • Make further functional improvements that might have a longer-lasting preventative effect.

A Brief Description of ‘Normal’ Function

What is normal? Let us define normal as symmetrical – or neutral – movements over a reasonable range, with each segment capable of playing its full part in that regional movement.

In young to middle-aged spines, non-discal disturbances of lumbar spine functioning invariably involve the development of asymmetry in lateral flexion. When this happens, we know that locomotion, standing and even sitting movements and postures are unbalanced, such that alterations in lower-extremity and thoracic/cervical function develop. The system will either cope with these changes, with or without adaptation, or it won’t.

Furthermore, extension of the lumbar spine – which for locomotion is more important than flexion – is directly affected by an asymmetry of lateral flexion. Or, put another way, a symmetrically moving lumbar spine in lateral flexion will always extend neutrally (again, in the absence of disc or disc/root problems).

By the way, definition wise I refer to the middle lumbar spine as L3-5, the upper as T12-L2 with the low lumbars being the lumbo-sacral joints. Not everyone will agree with this, of course.

All the following will justify why, inter alia, I place such emphasis on carefully observing the differences in sidebending in all three regions of the spine.

Early Progress

I’m sure you all appreciate the very steep learning curve of the first couple of years in practice. I think it is fair to say that a real appreciation of what palpation and passive movement testing can reveal begins then, and that – although it sounds a little odd – during the academic course itself, I was just going through the motions!

So one of my first discoveries in those early years was that during the side-lying exam, it was possible to reliably discern lumbar segment restrictions through passive movement testing. I remember how exciting this was, particularly the realisation that finding restrictions on the left side (for example), and treating them successfully, often correlated with improvements in bending to the left. At the time I knew that this sort of matched the insights and analysis of osteopaths such as Fryette, but I was not yet aware of the limitations of this way of thinking.

First Insights

I think my first dissonance was when I started to see the following patterns emerging.

A typical patient had the following;

  • Restricted sidebending LEFT in the lumbar area.
  • The usual checks of the mid and upper lumbar area (especially the left side) revealed no mechanical block impeding bending left in that area.

We’ll use the above as our base-line example for the rest of this article.

So I looked for problems in the left LS, because – as I had worked out from treating other parts of the lumbar spine – blocked bending on one side was likely due to locked joints on the same side. Why wouldn’t that apply to the LS?

So palpation and passive movement testing found stiffness / restricted movement in the low lumbar area on the LEFT side – with the right LS area seeming normal.

Decision? Treat the LEFT LS with a conventional rotation style lumbar roll HVT +/- some soft-tissue work preceding the HVT, or after.

Result? With the above pattern;

  • Treatment very often improved the restricted LEFT sidebending.
  • Indeed, because I always checked the patient’s balance on one leg, even in those days, I often found that balance on the right leg improved (see the following article for more on balance as a test for lumbar function).
  • Additionally, I found in the above examples that there were no differences before or after treatment in the supine leg length test (based on medial malleoli). This latter test I abbreviate to LLI.

Problems emerged when patients did not fall neatly into the above pattern;

  1. Sometimes the patient had a long right or left leg on the supine exam before treatment.
  2. Sometimes the patient’s active bending to the left was at fault (with bending right normal), but passive exam found a locked segment at the right low lumbar.
  3. Or the patient had a problem bending to the right, but I was sure the LS lesion was on the left.
  4. And combinations of the above.

I was stumped by the leg length differences noted above, but just put it to one side as an issue, being more interested in restoring natural sidebending and relieving pain.

For example, in 2) above, I would then treat the right low lumbar area using the conventional technique described above. This would often partially free up the segment, and relieve pain. Usually, though, it would make no difference or improvement to the poor left side-bending, or to the differences in leg length.

It was in this fashion that I carried on for a few more years until I attended a technique course at the ESO hosted by Barrie Savory. It was from Barrie that I first found a possible rational explanation for the exceptions noted above.

A New Technique

Barrie introduced the concept of leg-length discrepancy (LLI) allied to a particular way of treating it. Previously I was aware of the issue of LLI mainly in the context of femoral, tibial or scoliotic patterns, and had been puzzled by the persistence of LLI even when those conditions were not present. For the first time, I had to start thinking about the source of functional (correctable) LLI’s.

Barrie proposed that the origin of this type of leg-length discrepancy was a form of lumbo-sacral lesion (let us assume, as before, the left side), and the solution was a thrust technique that involved;

  • A side-lying setup, similar to a conventional ‘lumbar roll’.
  • The patient lying on his left side.
  • Similar amounts of rotation, but perhaps the legs in a more flexed hip position than the usual rotation style lumbar roll.
  • The thrust is with a left applicator hand/forearm on the left sacrum, in a cephalic direction.

Clearly, the thrust is a release – in a particular direction, and not involving any rotation in the thrust element – of the left lumbo-sacral joint. Barrie’s focus was on restoring the symmetry of leg-length in the supine position, which it seemed to do.

But I was more excited by observing that the successful technique seemed to involve (again, using the above example) an improvement in sidebending to the right. It looked as if I had a new way of dealing with the ‘exceptions’ to my traditional model of lumbar function and sidebending.

Take a particular example that I would previously have found puzzling;

  • The patient has a clear left sided LS block.
  • A short left leg supine.
  • An active sidebending deficit to the right side.

Before this insight, I would have corrected the left-sided lesion in the traditional rotation thrust way and worked on the left side soft tissues to permit right side-bending. This might have helped somewhat (and did seem to help reduce pain), but often would not fully improve right side-bending or correct the short left leg.

But this new technique of minimal rotation and cephalic sacral thrust seemed to achieve everything!

  • A normalisation of the passive movement and feel of the left LS.
  • Normal right side-bending – i.e. improved to the contralateral side of the lesion.
  • Correction of the LLI.
  • And improved balance on the left leg.

Barrie’s View

Barrie Savory argues that this type of lumbo-sacral lesion is far more common on the left side than the right. I’m not sure I agree with this, but I would agree with Barrie when he says that humans seem to be able to ‘drift’ into the left side lesion, whereas a trauma appears to be required for the right side. It is possible that this drift tendency on the left is something to do with handedness.

At the time, I wondered if the technique Barrie demonstrated was actually an SI technique. We had been taught these at the UCO based on a model of dysfunction of the posterior pelvic joints that postulated a kind of stable subluxation, where the altered ‘geometry’ of the pelvis could produce a short or long leg. But as time moved on after graduation I had formed the view that, at least in adults with no innate or acquired laxity in these joints (e.g. hypermobility syndrome, kick-boxers), SI joints were not mobile enough to produce these results. Anatomical research seemed to back up the idea that these joints just don’t manipulate. A controversial area, I know!

Of course, SI’s in most young and middle-aged adults do move, and they can get sore and irritable. They even have their own referral patterns. Accepting this, though, does not mean that they ‘release’ in the same fashion as the spinal joints above them.

The conclusion being, of course;

  • That when I found significant differences in the feel and passive movement of the LS –> SI complex, I should have not have attributed these to the SI.
  • And when I performed an SI technique and ‘got a pop’ along with some kind of passive movement improvement of the area, actually what I was treating was the LS.

Another Useful Test

Along with the careful observation of lateral flexion and one-legged balance, I found that (with the patient in the supine position) a useful additional test was getting the patient to lift one leg at a time. I started to record this as ‘leg weights‘.

Leg weights are always done with an extended knee. Due care is taken if the patient is in a lot of pain. Quite a lot of information can be gleaned from the feeling of weight, effort or heaviness that the patient experiences while holding the leg at about 20 degrees. It is not necessary to lift any further. I note the initial impression as they raise, and also what happens as the patient holds for 5 seconds. The critical thing is the comparison side to side, partly from what the patient reports and partly from the way the pelvis tilts.

If the spine is well working, balanced and pain-free spine, and there are no significant hip or knee issues, the leg weights ‘feel’ the same to the patient. Differences give really useful clues as to analysis and treatment.

Actually, in addition to the above caveats, there are some kinds of upper CSP problems and ‘core’ control asymmetries that seem to effect leg weights in the supine position.


But now I could analyse in the supine position and at the same time;

  • LLI.
  • Passive and palpatory exam of the L-S (this is a very convenient position to assess the L-S).
  • Leg weights.

Multiple L-S lesion types?!

So there appeared to be more than one type of L-S lesion. The first being the traditional ‘Fryette’ type, the second being Barrie’s. I now felt I could help troublesome sidebending patterns that I had previously found difficult. But were there other types of L-S lesions?

I found a third, I believe. To explain, take the standard base-line example given at the beginning of this article. Occasionally I observed the following;

  1. A pattern of sidebending restriction to the left.
  2. L-S dysfunction on the left.
  3. But with an LLI – always a long left leg.
  4. And a failure of 1-3 to respond to the conventional lumbar roll rotation thrust, which often would not release or cavitate.

In this situation, and referring back to the paragraph about SI joints a little earlier, I found that;

  • A conventional lumbar roll with the patient lying on the right side
  • With the applicator on the left innominate
  • With the thrust vector in a purely posterior and caudal direction

seemed to correct 1-4 above.

The innominate, of course, is only indirectly attached to the 5th lumbar – so the technique involves moving the innominate, dragging the sacrum with it, and thus releasing the 5th lumbar from the sacrum below.

Of course, in the example given above, there was no point in trying the Savory style technique, because this improves sidebending to the contralateral or opposite side, and in this case, there is no problem with the right L-S.

With this third type of lesion, I seemed to have squared the circle. Let us summarise all this.

Note that the logic in the table below carries on with the example of a left-sided L-S lesion. For the right side, simply reverse all the following.

Types of LS lesion

A note regarding the Type 2 technique. If it proves difficult to achieve an effective release of the L-S joint using this approach then another way is to place the patient on their right side and use a similar cephalic direction thrust but with the applicator on the left PSIS. This approach is less minimal than the sacral thrust in the table above, but can still work well.

What causes the LLI in this context?

My working assumption (once I have excluded all the other possible factors mentioned earlier in this article) is that certain types of lesion pattern at the L-S (type 2 and 3) cause a unilateral erector spinae tension pattern to occur that causes a short side of the lumbar spine. In the case of type 3, this seems to develop on the side opposite to the L-S lesion. This shortening (which ‘drags’ up the pelvis on that side, and is also responsible for the side-bending impairment to the opposite side) is thus present off or on weight-bearing and has nothing to do with joint considerations at the SI.

What Had I Learned So Far?

So I now had a working model of how to restore balanced lumbar sidebending (thus improving extension) in those situations where;

  • The patient had a normal mid and upper lumbar area.
  • There was no active disc or root issues.
  • There was no significant pain behaviour affecting movements.
  • There was no other disruption of lumbar function emanating from the thoracic, cervical or hip areas.

Where one or more of the above were present and causing their own movement effects, then it was a matter of analysing and treating in a staged way to ‘tease out’ the different elements of this combined pattern.

I was also learning that the three types of lumbo-sacral lesion seemed to have different remote effects from each other. For an explanation of what I mean by remote, please see this article. These effects are profound and explain why locally treating a problem in the thoracic or cervical areas sometimes leaves the patient with an incomplete improvement in function, even if symptoms have subsided. Further, these types of L-S lesions may themselves predispose to dysfunction secondarily occurring further up the spine. Being aware of these patterns is useful for both treatment and for understanding the chain of cause and effect.

This article is not the place to go into depth on these remote effects, but one example is worth considering now.

The ‘Type 2’ or Savory style L-S fault does seem to lead to adaptive secondary stiffness in the upper lumbar area of the contralateral side to the L-S lesion. Very often with this pattern, the patient presents with a failure of compensation and pain in the opposite side to the upper lumbar stiffness noted above. Treating the secondary stiffness will improve the pain state, but working back further in the chain by looking at the contralateral L-S helps even more.

Why Is The L-S special?

My working assumption – because of the extreme junctional nature of the segment. Through its location, its ligaments and muscular control and its varied planes of movements compared to middle and upper lumbar segments, it is far more unstable than the segments above. As we all know, maintaining 5th lumbar stability is quite a challenge, even for fit and active patients.

It may well be that the link between the local control system and the internal state of the L-S facet joint is so exquisitely complicated that the direction, speed and amplitude of an insult to the segment causes the local motor pool to react in different, but predictable, ways.

The LS joints might both become lesioned in different ways, either at the same time (for example, a fall, or a heavy lift while twisted) or at intervals (a left LS lesion might predispose to a different form of LS problem on the right). As you can imagine, this places an enormous strain on the muscular control system further up.

Perhaps any segment further up the spine is also capable (in principle) of reacting in different ways to shocks, but this is just less likely to happen to these more stable segments. This applies all the way up until we get to the upper CSP (C0-1 and C1-2).

At this other end of the spine, we really have to start thinking again!


I hope the above makes some kind of sense, and that osteopaths who are developing their own thought processes about spinal function and treatment might just find it of some use.

I haven’t carried out a historical review because this article has been about a personal journey, but I do appreciate that a lot of smarter people than me have thought and written about these issues over the last 100 years.

Leave a Reply