More fragility in core stability

The evidence for the role of exercise in chronic low back pain has been a bit of a theme on the blog here at BiM. We’ve discussed different perspectives regarding the evidence of its efficacy (see here, here and here), where it sits against other treatments in that regard and considered a recent review of its effects, or lack of, on parameters of spinal function.

A new paper by the same group, led by Anne Mannion, who performed that last review have had a closer look at the role of spinal stabilisation exercises in chronic low back pain, effectively asking the question – are the observed clinical improvements seen following a spinal stabilization exercise regime reflected in the sort of changes in abdominal muscle function that the core stability model would predict? Or simply – do patients get better after core stability exercises for the reasons we think they do?

Core stability is an interesting case. It divides opinion and yet stands as one of the preeminent models for treating back pain through rehabilitation. In the mid-nineties a rehabilitation movement was created, born from the experimental observation that deep abdominal and paraspinal muscles are altered in their activation patterns in patients with back pain. Those early experiments and subsequent ones gave the model plausibility but, for me, what was more fascinating than the phenomenon of altered trunk muscle function was the phenomenon of a clinical dogma thrusting its way relentlessly through the therapy world. From this fair but limited data, enthusiastic entrepreneurs and self-elected authorities duly sprinted with the ball, creatively developing detailed treatment approaches with strict and specific rules that far exceeded the actual data. “Contract this muscle but not that one, definitely not that one” or “move like this, not like that”, spreading the empirically unsubstantiated (but potentially harmful?) concept to therapists and patients alike of spinal segmental instability, wherein poorly controlled vertebral segments shear excessively resulting in pain. You too could have the answer to treating back pain as long as you attend these 5 sequential courses at $$$$ a pop. Colleagues were falling over each other to buy the equipment needed to apply this model that had been accepted as gospel, from not-too-pricey pressure biofeedback cushions to very-pricey-indeed real time ultrasound imaging devices and the rehabilitation philosophy of the long-late Joseph Pilates experienced a remarkable resurrection. As a case study into how new treatment approaches are adopted in our profession it is perhaps second to none. Maybe the model was correct – we couldn’t know at that time, but in my early clinical career core stability came to dominate thinking as an accepted truth way before we had a good answer about whether it worked.  Maybe a better approach to introducing new treatments might be that recently advocated in this great paper by and Kari Bø and Rob Herbert?

But I digress. This new study by Mannion and colleagues followed a single group of 37 chronic back pain patients through a 9 week programme of spinal stabilisation exercises. The measured a battery of clinical variables but also measured the aspects of trunk muscle function that the core stability approach would seek to change – anticipatory activation of the lateral abdominal muscles and voluntary activation of transversus abdominus during abdominal hollowing. They found that clinical improvements in disability were not related to changes in voluntary or anticipatory activation of the abdominal muscles. In fact the only variables that they measured that were associated with improvement were a reduction in catastrophising and the range of lumbar flexion.

So in this group it appears that where patients did improve it was unlikely to be as a result of improved core stability. By that interpretation it follows that those who improved did not uniquely respond because they were part of a subgroup whose back problem was the result of poor spinal stability. Since there is no comparison treatment group those improvements might simply reflect natural history, but they might also be due to the non-specific effects of care, the fear reducing effects of doing some (any) exercise or the belief that the spine was more stable.

This isn’t the biggest study and some might argue that the lack of attention to the function of paravertebral muscles such as multifidus is a problem. Nonetheless this study adds to the large body of evidence that has already suggested that, for chronic low back pain, exercise helps a little, but no specific exercise approach is clearly superior, and now in this group it appears that this very specific type of exercise does not seem to induce clinical improvements in the way that it is very specifically designed to. For the plausibility of core stability in the treatment of CLBP that is something of a blow.

About Neil

Neil OConnellAs well as writing for Body in Mind, Neil O’Connell is a researcher in the Centre for Research in Rehabilitation, Brunel University, West London, UK. He divides his time between research and training new physiotherapists and previously worked extensively as a musculoskeletal physiotherapist. He also tweets! @NeilOConnell

He is currently fighting his way through a PhD investigating chronic low back pain and cortically directed treatment approaches. He is particularly interested in low back pain, pain generally and the rigorous testing of treatments. Link to Neil’s published research here. Downloadable PDFs here.


Mannion, A., Caporaso, F., Pulkovski, N., & Sprott, H. (2012). Spine stabilisation exercises in the treatment of chronic low back pain: a good clinical outcome is not associated with improved abdominal muscle function European Spine Journal DOI: 10.1007/s00586-012-2155-9

Bø, K., & Herbert, R. (2009). When and how should new therapies become routine clinical practice? Physiotherapy, 95 (1), 51-57 DOI: 10.1016/


  1. There wasn’t a control group in the study? That’s surprising!

    Neil O'Connell Reply:


    This study wasn’t looking at whether core stability exs are effective or not. We could consider that question well answered by other trials. As such the design didn’t require a control group as a trial might. Rather the study wanted to see if any clinical improvements observed were related to the predicted improvement in trunk muscle function. They weren’t. And that’s why the study is a challenge to the theoretical basis of core stability in back pain.

  2. Steve Kamper says

    Hi Neil, I agree, I’ve heard several times that RCTs are an inappropriate method for testing complex interventions but no such objections to uncontrolled case series’ or experiments on healthy subjects. I’ve long suspected the real objection is to the findings rather than the method.
    Love to chat to you about this. Flick me an email, if you’ve time/are inclined.

  3. Steve Kamper says

    Nice post Neil, thought provoking as always. I am interested in the idea of the core stability thing and its widespread integration into (especially) physio culture. I have for a while had in the back of my mind the impression that the findings from experimental research are more readily incorporated into (or at least more readily accepted by) clinical practice than the findings from clinical research. This is obviously despite the fact that the latter studies are supposed to be designed for clinical application whereas the former are not to the same extent.
    At one stage I had an idea to try and see if this could be measured (i.e. the above is actually the case) but I ran out of steam thinking about the logistics. I still think it would be an nteresting study.

    Neil O'Connell Reply:

    Cheers Steve – really important points. This is a generalisation (but a fair one?) but in my experience physios tend to think a plausible mechanism is enough evidence to justify a treatment. That definitely leaves experimental evidence in the premium seat for guiding practice – observations from a lab study is often used to underpin or promote a whole treatment philosophy, by which point the finding has been stretched way beyond what it can actually tell us. I think it’s understandable but it has been at the heart of many failed treatments (consider most electrotherapy). The problem is once we therapists have a plausible mechanism and from there a treatment approach we all try it out, observe/ convince ourselves that it works, sell it to the willing audience who do the same, and then, when the RCTs come up short, conclude that we know that it works, that the trials must be wrong, didn’t identify the correct subgroups, or my favourite “trials just aren’t appropriate for physiotherapy”. Logical fallacies all.
    Plausibility is a start and should be a prerequisite, (but isn’t always a given: for instance acupuncture or fascial “release”, the list is endless). Changing the culture is difficult though, particularly when clinical trials earn a reputation for producing awkward results.
    The irony is that the culture of experimental research seems a bit behind that of clinical trials in terms of ensuring adequate power and controlling internal biases. Most of these studies that guide us are small and I reckon the risk of publication bias and false positives is higher in that field than it is for trials. Both types of research are vital but I agree – running with just an phenomena as evidence rather than evidence of efficacy is potentially problematic.
    Love your idea for a study – will chew it over, I had an idea of something different but on a similar theme a while back – we should natter.

    Andy Reply:

    A very succinct appraisal Neil, permission to quote (ad nauseum) please?



    Neil O'Connell Reply:

    be my guest on the proviso that you correct my grammatical errors!

  4. David Nolan says

    Its a very interesting journey indeed Jill, on how data has been misrepresented and re sold. I’m sure the western view on what is a body beautiful must take some blame too.

  5. Jill Wigmore-Welsh says

    Very interesting post. In the early 90’s when this swept in to the UK, I was working with elite rowers in the GB squads who frequently had lumbar disc issues. We were among the first wave of Physios trained in spinal and core stability work and peripheral joint segmental stabilisation exercises. At around the same time the basic “physio” core exercises were quite cleverly incorporated by Lynne Robinson’s (Body Control ) into her “take” on Pilates, but was never part of the original Joseph Pilates work. She gained a lot of support from Physios who trained with her and promoted Pilates because it included core work. The fitness and exercise industry is huge in the UK and fitness instructors and class teachers are constantly looking for innovative money making programmes. In my view its been a really interesting journey observing how Pilates has now gained almost folklore status as being ‘good for backs’ .

    Peter Faletto Reply:

    I couldn’t agree more. Pseudo specific training with words clients had not heard before turned Pilates into the cure all for every malady. The number of people writing articles for the US Physical Therapy Journals touting the benefits of Pilates seems to largely driven by their desire to certify/teach someone their method of Pilates. I don’t believe it is any more innovative, just repackaged with new buzzwords. My favorite question to ask them (or anyone) is how do they MEASURE core strength? I feel posture is the “functional” outcome of core strength/ening. Spinal symmetry may be the next “measure” of outcomes that may lead to some greater degree of accuracy in judging the “success” of a program.

  6. David Nolan says

    Thanks for the link Esther, Perhaps now we are better equipped to give an explanatory model which not only fits their expectations but actually makes a difference to people. Even though the model “you have a weak core, you need to strengthen you abdominals to give stability to your spine” may have helped some I also feel it has hindered many.

    Esther Reply:

    I guess different models will help different people or hinder people. I certainly found that not all patients liked the core stability model but other patients thought it made perfect sense.
    The thing I found interesting about the study by Toye and Barker was that even though patients who did well on the programme showed an understanding of the psychosocial aspects of pain, there was still a need to reconcille a physical explanation for symptoms with the psychosocial as they did not want to accept it was all in their head. This is where explaining pain from a neurophysiologal perspective has great potential to help patients by fulfilling their need to understand what is going on with their bodies and is a more evidence based explanatory model.

  7. Awesome post and study – given the growing understanding that there is not magic in any particular exercise routine, and that recovery from pain has as much to do with all aspects of our existence rather than just structure and pathokinematics, this is not a huge shocker. It might however be tough on some businesses.
    Probably a good idea to start to pay more attention to the aspects of movement/exercise routines that are associated with creating adaptive neuroplastic changes in body awareness, body image, fight/flight/freeze responses, breathing patterns, muscle tension, …, and that help restore the cognitive-emtional factors that ongoing pain negatively impacts – confidence, self-efficacy, locus of control, fear, anxiety, isolation, grief, …
    These things may sound a long way from PT. They are not. For most people the behaviour piece of CBT is much more real when attained through the body than through the mind and cognitions, and it is probably easier to change autonomic physiological processes when we use the body as well as the mind, than relying only on the mind.
    Just more theory I know.
    On the other hand, if there are some people with chronic non-specific LBP who get better when they go through core stability work or through Pilates, we should be looking for alternative reasons for the success (beyond more reductionist ones) and working to figure out the elements of those interventions that are beneficial.

    Neil O'Connell Reply:

    Thanks Neil and nice comments. Which factors trigger recovery is a tricky one and everyone seems to have a different view on how to approach the problem and which aspects of management hold most promise. What we really needs is that endangered species – a number of good RCTs of a specific approach that clearly show clinically meaningful superiority over existing treatments, and changes beyond the short term.

    Neil Pearson Reply:

    I get what you mean, and at the same time figure that since we are trying to change a human experience it is most likely that there will be multiple paths to recovery. As soon as we throw pain into the mix, we are trying to measure something that is illusive, and individual. Science would suggest that part of the motor control issue is related to pain – another protective response. Since these can be perpetuated by factors related to pretty much every aspect of our existence, our job will be to not latch onto the first treatment regime that shows some hope, but rather maintain our curiousity in finding more options to help alter the horrible nature of this human experience.

  8. How interesting. As one of the physios who were sucked into the core stability revolution in the 90’s I have seen patients benefit from these exercises despite the fact, as demonstrated here, there is little evidence that they actually improve core stability. There is an interesting qualitative paper by Toye and Barker ( who interveiwed patients taking part in a LBP rehab programme. This study found that even patients who had taken on board the psychosocial aspects of their back problem appeared to benefit from contructing an acceptible explanatory model that explained the physical aspects of their back problem (body out of balance). The concept of core stability is one that (possibly) makes sense to patients and fullfils this need for patients to understand what is going on with their bodies. Possibly another explanation from why patients benefit from these exercises?

    Neil O'Connell Reply:

    Really good points Esther, thanks. Maybe it is just a response to what appears to be a meaningful explanation of the symptoms. But as I suggest in the post, it is possible that this kind of explanation might cause avoidance and fear in some – the idea of an unstable segment, damaging forces etc.

    Regardless the rise of core stability has been fascinating. As someone who never really bought in (not for wholly rational reasons at the outset I must admit – more because I am fairly contrary, kicked against the wholesale adoption of this belief system a bit and also I found that the exercises were very difficult for the average person to comprehend) it was amazing to see it become “fact” overnight.