The myths of foot orthoses… A guest article by Ian Griffiths

So I am absolutely chuffed to bits that I have managed to convince and lure another guest writer at 'The Sports Physio'. I'm even more chuffed that I have been able to convince one as well-respected and recognised as Ian Griffiths a highly specialised Sports Podiatist. Ian has a wealth of knowledge and experience in all things foot, ankle and lower limb injury wise, his full bio can be found here, he also has an excellent blog that's well worth reading here and he is an absolute must follow on Twitter where he regularly gives out free advice and tips. So Ian has kindly written a fantastic piece on dispelling the myths of what foot orthoses (not 'orthotics' as I've been incorrectly calling them) do and don't do… take it away Ian

Adam has very kindly asked me to write a piece for his blog and whilst I can’t necessarily promise the usual levels of controversy, I certainly feel I have picked a topic here in which some commonly held ‘myths’ can be dispelled and hopefully some healthy debate is generated off of the back of this. I also hope that we can give some coverage to answering some of the common things you may see or be asked regarding foot orthoses in your clinic. It is worth just noting two things before we get stuck in to the meaty stuff. Firstly, despite its length this is not intending to be an all encompassing piece which will answer all questions (the topic is just too large, and as you will see shortly not all questions have answers yet). Secondly, I need to make clear that in expressing my understanding of this topic here I am very much standing on the shoulders of giants – in particular my friend and mentor Dr Simon Spooner. I am going to try and keep this piece much closer to a blog than an academic article in format purely for ease of reading, so very few references present and instead just some suggested reading at the end. If anyone reading this is desperate for any formal references based on any comments I’ve made then I can make these available on request of course.

“Unlearning comes before learning”

Before taking on new ideas and concepts we often have to be willing to re-evaluate what we think we already know or were once taught. In the context of foot orthoses this is usually with regard to how they ‘work’, or what we are actually aiming to achieve by issuing them. Many of us were taught that orthoses are biomechanical devices (how can they be when they are not living tissue?) which will ‘correct’ foot posture by realigning the skeleton and/or holding the foot in its neutral position. Essentially, all the focus has historically been very firmly fixed on controlling the position and motion of the foot (kinematics). What I hope to give people a cursory introduction to in this blog is that this may not be as appropriate as was once thought.

So, let’s look at these points individually. Do they ‘correct’ foot posture? Well in order to do so it suggests we have a solid understanding of what is incorrect. This is far from the case, with no consensus on what a ‘normal’ foot is and no clear agreement on the foot level characteristics that will consistently predict or even correlate with lower extremity overuse injury. I don’t want to go off on a tangent on this one, and those that know me will already know my thoughts on this subject (with respect to “overpronation” etc) but it is an important question to ask. If we don’t know what position the human foot should be in then what is our goal when introducing orthoses? The idea that a foot should function in its neutral (STJ) position has hopefully long been put to bed, with not one single study ever performed (be it static or dynamic) showing this to be true of any sample of the population taken.

Even if we knew for sure that ‘STJ neutral’ was king, would foot orthoses hold the foot in this position? Well, no. They cannot physically hold the foot (they are inert materials) and again not a single shred of evidence exists showing that this is what orthoses do. The research performed on the kinematic responses to orthoses give us some very interesting data. We will talk about the limitations of foot orthoses research later, but for know all you need to know is that there are just as many studies available which show that rearfoot kinematics do not significantly change as there are that show they do. (NB Therefore easy to cherry pick references which support your beliefs on this topic should you be inclined to do so!). The best explanation of this based on our current understanding is that the kinematic response to foot orthoses (i.e. if someone shows any angular changes at all, and if so how much of a change) is subject specific/different for everyone. We do not really know why.

Image 1: Depicting the classic 'before and after' shot associated with orthoses introduction. this is a fake. i know as it is my wife's left foot and I took the pictures 30 seconds apart on a demo orthosis that wasn't even hers.

So do foot orthoses ‘work’? Well this will depend on how you define ‘work’. If you define it as eliciting a visual change at foot level then the answer is that they will not work in this way for everyone, and it will be difficult/impossible to predict those individuals that they will work for. If you define ‘work’ as improving pain levels and reducing musculoskeletal symptoms then this is also difficult to predict and will vary from person to person, so they could be deemed to sometimes not ‘work’ in this regard. However the bit to get your head round is that they do not need to change position or alignment to reduce pain levels. Why? We’ll come to that… And how else could we define ‘work’? We’ll come to that too…

Confused? I hope so. We all should be. Preferred the old ways we were all taught at undergraduate level with respect to foot orthoses because they were simpler/easier? Yeah, me too. (That’s why a lot of clinicians still stick with it!) A summary of the salient points so far:

• Orthoses are mechanical devices, not biomechanical devices
• We don’t know what the best/ideal/normal/correct foot posture is for everyone
• Even if we did, orthoses don’t hold the foot in this (or any) position
• Kinematic responses to orthoses are subject specific
• Orthoses do not need to change kinematics to reduce pain

So how do orthoses really ‘work’ then?

We know that they change alignment in some of the people some of the time. We know that they reduce pain and improve symptoms in some of the people some of the time. But what do they do in all of the people all of the time? They change kinetics. If you define ‘work’ as altering kinetic variables, then all devices are deemed to ‘work’ (Remember this won’t necessarily correlate with clinical success!) Kinetics are forces – all too easy to ignore (particularly clinically) by virtue of being invisible. Forces are what drive motion. They start it and they stop it [Newtons’ 1^st]. It is force which will be the key contributor to tissue damage. NB Although obvious it is worth briefly noting that they w
ill of course be a close relationship between kinetics and kinematics, but we will discuss them completely separately here to try and simplify things.

There are now many papers which have looked at the kinetic responses to foot orthoses. Most research will report on the kinetic and kinematic data separately. The first paper I read which really allowed the penny to drop for me regarding this entire concept was in 2003 [Williams et al]. To summarise, they measured both kinematics (angular changes) and kinetics (using inverse dynamics to calculate work loads in tissues) in three different environments (no orthoses, and two different orthoses environments of different levels of inversion) in runners. Interestingly there were no significant kinematic differences between all groups, but the demands placed on the tissues they measured significantly decreased as the level of inversion of the orthoses increased. Therefore these orthoses worked (i.e had a mechanical effect) without appearing to work (change alignment). Worth reiterating that whether or not this equates to clinical success is of course a different story and again brings us back to subject specificity.

Traditionally, if a patient returned and reported no improvements and on review it was deemed there was “not enough correction” then orthoses would often be adjusted and possibly have more wedging applied to them until such time as the patient reported improvement. We thought we were shooting for a situation where the foot was in a better position – could it be we may have been measuring and trying to change one thing, unknowingly altering another thing, and clinically succeeding by accident? Food for thought. Anecdotally, in my own clinical practice I often see patients getting great clinical outcomes from devices with seemingly little positional change observed. Anyone who has given out enough devices will have seen this too I’m sure. Just to further muddy the waters it is certainly the current belief that kinetic responses to an intervention will also be subject specific, and difficult to predict. (As well as being invisible). Nightmare.

So where does this leave the clinician?

Anyone issuing foot orthoses needs to have a sound clinical rationale for doing so, which should include an understanding of what they are trying to achieve and how best they feel it can be achieved. In light of the discussion so far you can see this is a far more complex set of considerations than simply looking at a foot and deciding it needs some ‘support’ merely based on its position either statically or dynamically. Anyone who tells you otherwise has unfortunately fallen prey of the Dunning-Kruger effect. Another summary before we go on:

• Foot orthoses will always change kinetic parameters
• Kinetic responses to orthoses will be subject specific
• The inability to measure/observe kinetics clinically is probably why disproportionate attention is given to kinematics

When we introduce foot orthoses we are essentially making an alteration of the foot level environment. It is key that we have some understanding of how the choices we make here may influence the kinetic parameters, and try to use our understanding of this to achieve our treatment goals. I’m sure most readers are familiar with the concept of equal and opposite reactions [Newtons 3^rd]. The fundamental mechanical function of foot orthoses is their influence of reaction forces at the foot-orthosis interface. Dr Spooner has postulated that they do this in one of three ways, and we can manipulate these to our advantage by virtue of three key design variables in our orthoses. This makes such mechanical/scientific sense to me that it has changed the way I practice and prescribe devices. It also has implications for picking off the shelf/prefabricated devices which we will talk a bit about later on. To try and give this some clinical context I will also need to take a slight detour away from orthoses briefly and talk about tissue loading and damage.

Kinetics at the ‘foot-orthosis interface’…

The meeting point of the foot and the orthosis shell is what we refer to as the foot-orthosis interface. When the foot applies a force onto an orthosis, the orthosis will apply a force back (the orthoses reaction force or ORF). It is our understanding of the reaction forces here and how decisions we make will affect them, which is important. Decisions we make will only modify these reaction forces in three possible ways (irrespective of what type of orthoses we are issuing):

1. Altering the magnitude of the reaction forces
2. Altering the vector of the reaction forces
3. Altering the temporal patterns of the reaction forces

Image 2: The foot-orthosis interface

So basically, all we can look to change at the foot-orthosis interface is how high the forces are, the direction they push in, and the time they spend there. That’s it. These factors may of course have subsequent effects (both desirable or undesirable) such as changing joint moments or changing tissue loads, and as we have already touched on they may or may not be accompanied by a kinematic response, but the only ways in which orthoses are thought to exert their mechanical (kinetic) effects at foot level are via one or more of points 1,2 and 3.

It is thought that we can only achieve alterations in one or more of the above three key kinetic variables by virtue of the following design options in our orthoses:

A. The surface geometry of the orthoses (shape/contour)

B. The load-deformation characteristics of the orthoses (stiffness)

C. The frictional characteristics of the orthoses (think top cover selection)

So when writing a prescription for a custom made device, or when picking from the hundreds of different prefabricated devices available on the market, these are the three key decisions the clinician has to make. How stiff does it need to be? What shape does it need to be? What top cover do I want (if any)? That’s it. There is nothing else in the clinicians’ control. Every single modification or design feature that exists and can be incorporated into an orthosis will change one or more of A, B, C and in turn will more than likely alter kinetics at the foot-orthosis interface via one (or more) of 1, 2, 3. It is making these decisions in the context of a real person with a real pathology that should be the goal of anyone implementing orthoses therapy as part of a rehabilitation strategy.

I have just realised that this probably needs a whole separate blog of its own, where we discuss joint axes, lever arms, external and internal moments etc but hopefully the main introductory message is clear – things are complex and not as simple as we once thought.

Still with me? It’s a bit intense I know, and further illustrates why some (lazy?) clinicians still just dish out ‘arch supports’ with the same prescription or from the same off the shelf company for all their patients. It’s easier. Lots.

So when do we give orthoses?

Let’s take a step back from the physics behind things for a while and get onto tissue damage. I make no excuses for grossly oversimplifying things here, but generally speaking you can summarise that with respect to musculoskeletal overuse injuries, ti
ssue dysfunction could be said to occur when a healthy tissue is loaded in an ‘abnormal’ way, when an unhealthy tissue is loaded in a ‘normal’ way or (and perhaps worst of all) an unhealthy tissue is loaded in an ‘abnormal’ way. It has been postulated that each of the body’s tissues has a zone of optimal stress and the tissue must be loaded within this zone to remain healthy. [This zone will of course will be different for each of the body’s tissues].

Thus, for a given individual with a given pathology, the aim of orthoses therapy should be that it helps facilitate the loading of the ‘target tissue’ to fall back within its zone of optimal stress (if it can).

The difficulty here of course (nothing in this game is ever easy) is that we do not necessarily know exactly what that zone is for a given tissue, and furthermore that it is likely to be a dynamic variable which itself is dependant on tissue health (e.g dysfunctional tissue = narrower zone). Worth noting here however, that for a given individual with a given pathology it is highly unlikely that there is one type or style of orthotic device which will result in a good outcome, meaning that one highly specific prescription for each patient and each dysfunctional tissue is unlikely. Dr Spooner refers to the concept of there being a range of ‘positive solution sets’ and I tend to describe it as an ‘envelope of success’. This is the reason that a lot of individuals will report at least some improvement in symptoms when given a device of any description for the first time (thus potentially perpetuating the aforementioned lazy clinicians’ beliefs that this is a simple game and they are on the right track) – if it puts the target tissue back within its zone of optimal stress then a good outcome will be achieved (even if the clinician did it by accident!)

It brings up a few more questions. If we are using this tissue stress approach to management of lower extremity overuse injury then how do we treat asymptomatic individuals? That’s a toughie. With the old adage ‘if it ain’t broke then don’t fix it’ ringing in your ears you still sometimes just see movement patterns that your clinical experience tells you may increase the risk of a tissue being loaded outside of its zone of optimal stress (even if the research doesn’t support you). You have to judge it on an athlete by athlete basis and make the call based on that individual, their level of conditioning, their training regime, their goals etc etc. Assessing/screening uninjured athletes is actually my least favourite thing to do. It is a very bold shout to claim with authority that you need to change something as it increases risk of injury (particularly in light of our current understanding of injury risk being poorer than we like to admit) in an asymptomatic individual.

Next question – do orthoses have to be worn for life? Another good one that gets asked a lot. Answer = it depends. For some individuals you may just want devices in situ to allow a tissue to be more appropriately loaded whilst rehabilitation and/or healing takes place. For others it is certainly the case that they may need those devices in place for the tissue to not be loaded outside of its physiological window and removing them may result in recurrence of injury. There is no one answer for all on this. Again you have to make the call based on the information you have in front of you at the time and your clinical experience. Remember despite what you may read from those with a financial motivation – there has never been a single piece of work published which has shown that orthoses ‘brace’ the foot or result in muscle weakness.

Last common question – custom made orthoses or off the shelf orthoses? The discussion about comparison of costs and durability aside, what is the main difference between these devices and when would you use one over the other? Well as already mentioned you have three main things to make decisions on (shell stiffness, shape and frictional characteristics) and you need to manipulate these to alter kinetic at the foot-orthosis interface in a way that will achieve your treatment goal. If there is a prefabricated device which ticks all the boxes then is it technically any worse than a custom made device? No. However it has to be pointed out that there are distinct advantages to designing the device yourself; mainly the far greater control you have over the prescription variables of interest. In my opinion the able clinician has both a very clear idea of the design features that may best achieve their goals and a good enough knowledge of the prefabricated orthoses market that they can identify one which is a reasonable off the shelf solution in delivering those design features (if there is one). Certainly beware the clinician who only recommends custom made devices exclusively, but also beware those that only recommend one type of prefabricated device.

Image 3: The main considerations with foot orthoses

Research Limitations…

I’ve alluded to the research several times, and in this day and age we are quite rightly trying to all implement more evidence based practice. Should we continue to do things just because we always have and they have always worked? Probably not. I’m not suggesting we stop them (for example the clinician giving out generic orthoses and getting good feedback) as we mustn’t forget the most important thing in all of this is the patient, but I do feel we should at least be constantly striving to understand the underlying mechanisms.

The problem with foot orthoses research is it is undeniably difficult to design. Individuals will often know if they are in the control group or not. It is not possible to have a ‘sham’ device in these studies as even this will alter kinetics at the foot-orthosis interface and therefore potentially have a mechanical effect. Quite often the orthoses given to pathological sample groups in a study do not appear to resemble the sort of orthoses that clinicians would give for that pathology in the ‘real world’. The list goes on. These difficulties may be why there is such disparity in the conclusions reached in orthoses research to date. All I urge people to do when they see a study incorporating foot orthoses is to look at the methodology rather than just the one line conclusion in the abstract.

Finally, and most crucially, you may remember me saying that kinetic responses to foot orthoses are variable and subject specific, just like kinematic responses. How many articles have you seen that have standardised for this? None. (can it even be done?) What they often do instead is give the subjects a device and monitor the kinematic responses or pain reductions and perhaps compare to a ‘control’ group. It is therefore highly likely that all individuals in these sort of trials are receiving a different kinetic ‘dose’ from the orthoses. This could be likened to a pharmaceutical company performing a trial on a drug and giving every subject a different (and unknown!) dose of the drug – how on earth would they learn anything about that drug?!

Summary…

For those still reading – if I haven’t managed to articulate things well enough that any of the above made sense I hope that at least I’ve managed to get across that the practice of issuing foot orthoses is complex, and our understanding of the subject (whilst not where we would like it to be) has certainly developed over recent years.
No doubt it will continue to develop, and I hope research looking at how orthoses prescription variables alter kinetic parameters, and in turn how these changes correlate with clinical success for a given pathology will be at the forefront. It seems to be the case that all the current research is looking at how good orthoses are for pathology X, or if they prevent pathology Y, or how do they compare to Physiotherapy in the treatment of pathology Z, when actually it’s pretty clear we need to take things back a few steps and actually investigate how they ‘work’ much more, and look to try and improve our study designs and methodologies.

So a last two bullet points for all:

• Significant consideration should be given to the subject specific responses to foot orthoses [will this change what you ‘promise’ your patients?]
• Orthoses goals may more appropriate to consider in reference to the zone of optimal stress rather than aiming for a kinematic ‘norm’

And to close out, a paragraph from that man Dr Spooner again which sums things up beautifully:

“The skilled clinician is capable of identifying the dysfunctional tissue (and the severity of injury to it); they understand the biomechanical function of the tissue during given activities; and they comprehend the manner in which each of the 3 orthoses design variables interact with one another, and ultimately their kinetic influence on the foot.”

Suggested Reading…

McPoil TG, Hunt GC: Evaluation and management of foot and ankle disorders: present problems and future directions. Journal of Orthopaedic and Sports Physical Therapy 21: 381, 1995.

Nester CJ: Lessons from dynamic cadaver and invasive bone pin studies: do we know how the foot really moves during gait? Journal of Foot and Ankle Research 2: 18, 2009.

Kirby KA, Spooner SK, Scherer PR, Schuberth JM: Foot Orthoses. Foot & Ankle Specialist 5: 334, 2012.

Williams DS, McClay I, Baitch SP: Effect of inverted orthoses on lower-extremity mechanics in runners. Medicine and Science in Sport and Exercise 35: 2060, 2003.