Dear Biomech-l users,
I am happy to share with you a summary of the responses on the definition of “joint stability”. I can certainly state that, from the number and the quality of the responses I have received, the topic of this discussion seems to be of great interest to the biomech community. Surely there seems to be the need to establish some standard definition and/or provide clinicians with indications and tools to objectively estimate the level of instability of joints.
I would like to point your attention to two definitions that have been put forward. The first one has been kindly sent by Dr A.Vleeming and it is a section from: Vleeming et al., European guidelines for the diagnosis and treatment
of pelvic girdle pain, Eur Spine J, 2007.
‘‘The effective accommodation of the joints to each specific load demand through an adequately tailored joint compression, as a function of gravity, coordinated muscle and ligament forces, to produce effective joint reaction forces under changing conditions’’

The second one has been forwarded by Dr Scott Tashman, and it is from: Franklin et al, Feedback Control of Dynamic Systems, 1987
"A stable system is one that always gives responses that are appropriate to the stimulus"

Other definitions (mathematical/physical) have been suggested and can be found in the responses.

After following and replying to all your interesting comments I would like to put forward this definition for you to consider:
"A joint can be defined as “stable” when, subjected to physiological loads, does not dislocate nor damage the articular surfaces and/or the supporting soft tissues and does not require extra work from the neuromuscular system to perform its tasks"

I understand this topic is still very much open for additional discussion, therefore do not hesitate to share your thoughts and suggestions. In the meanwhile thanks again to all of you who contributed to make this thread so interesting.

Best wishes,

Paolo Caravaggi

Dear Biomech-l subscribers, I was wondering if any of you is aware of an objective test to assess joint stability. According to my literature research neither a standard definition of joint stability nor standard evaluation tests to determine the degree of instability at joints have been established In most cases the level of instability is subjectively assessed by clinicians by applying dislocating forces to the joint. When more scientifically-objective approaches were taken, joint rotations to triplanar joint displacements (or rotations) are normally shown. However, although differences to the normal/stable joint are graphically presented, when/if the joint can objectively be considered unstable is not reported.

As far as our specific case, we are trying to quantify the level of instability at the proximal interphalangeal joint of the finger in-vitro through active flexion/extension of the joint following the release of supporting ligaments and the disruption of the joint by systematic resection of bone at its proximal aspect. We are indeed finding significant differences in the flex/ext rotation to joint displacement curves across different configurations (intact, disrupted..) but we are now facing the issue of establishing some kind of objective index for joint-stability. In other words, which variable is more relevant here and how far from the normal/non-pathological configuration a joint can still be considered to be stable? Any suggestions and/or further comments on this matter are welcome.


Paolo Caravaggi, PhD
Joint Biomechanics Lab, Orthopedics dep.
University of Medicine and Dentistry of New Jersey
185 South Orange Avenue, Newark, NJ 07103
Tel. +1 973 972 1426

You may find Dr. Manohar Panjabi's work on spinal instability of some relevance:

Journal of Electromyography and Kinesiology, Volume 13, Issue 4, August 2003, Pages 371-379



Dear Paolo
It might first be worth precisely defining what you mean by unstable. It maybe like trying to define the height of a 'low table' or its usefulness in terms of its height

Is the/a joint unstable when it exhibits an unusually large range of motion? Is it unstable when the joint has a large range in a plane that is not the usual primary plane of motion? Or is it unstable when the joint cannot maintain a fixed position under a certain load? or is it when the joint cannot maintain the expected supporting role during a dynamic or static action. Is it unstable just when the patient / subject says it is or is it when they experiences symptoms or show clinical signs that could be attributed to one or some of the above. If you have a fixed definition of an unstable joint, e.g. a certain joint is unstable when it exhibits 'x' range of motion under 'y' load during an activity of interest, is it then determined to be stable if it does not fulfill those criteria even tho the patient describes the joint as unstable because it gives way during the activity of interest. This is the problem with defining such a parameter, how do you apply it to your patient if they don’t fit in the box? Without those pre set criteria the clinician has the freedom to determine what the problem is with the joint (or any other part) based on his own skill, judgment and education. It might be useful to classify joint stability in an attempt to achieve some standardised comparative assessment data. However if the patient says it is better, then it is better. If the patient or the clinician thinks that the joint is problematic because of its instability, then make it more stable. Why would you want to precisely define how stable it is or unstable it was before, or after, the patient said it was better? Not critisizing, just some thoughts
Regards Dave Smith
Podiatrist, MSc App Biomech
FootHouse Podiatry
I've attached some mechanical stability papers for your reading pleasure. Enjoy.

Josh Cashaback, PhD candidate.
Biomechanics Lab, Kinesiology dep. McMaster University, ON., Canada
Paolo and Dave,
My experience with this question is from an engineering and not clinical perspective so please forgive any naïveté in that area. Classical control methods would consider an unstable system to be similar to one that is a ball sitting atop a sphere. Any slight perturbation will cause the ball to roll off in any direction with nothing to resist or control it. Likewise, human joints, though not as simple as this example, can exhibit similar characteristics. This, I assume, is the instability you are talking about. In the ball on the sphere example I could plot displacement (y-axis) vs. force (x-axis) and get a plot that shows any force can cause an infinite number of displacements (vertical line).Now suppose I changed to the top of the sphere to add a little dished out portion before the ball could roll down the side. Then, there would be some displacement/force curve that would become a vertical line once the force got too high. The slope of this curve would be considered compliance and eventually the compliance would be infinite (vertical line).Infinite compliance is instability. On a practical level with human joints the curves are not so simple to create or interpret, but for your specific joint you will need to come up with some way to plot compliance. Then you will need to determine some sort of threshold of where you consider instability to occur for that joint. You may consider plotting the 1st or 2nd derivatives of these curves to assist in this determination. I hope that helps.
Robb Colbrunn
Director, BioRobotics and Mechanical Testing Core
Cleveland Clinic
Cleveland OH USA
I agree that there is no standard definition of stability as applied to joint function. Traditional engineering-based definitions (e.g. bounded input/bounded output) don't really apply. One of my old engineering textbooks describes a stable system as one that always gives responses that are appropriate to the stimulus (Franklin et al, Feedback Control of Dynamic Systems, 1987). The system we are talking about (a joint) is complex, consisting of a combination of joint geometry, soft tissue constraints, muscles/tendons and a neuromuscular control system that is subjected to wide range of externally applied loads. So, for joint stability, what is the stimulus, what response do we measure and what distinguishes an appropriate response from an inappropriate one?

I propose that the stimulus should, as closely as possible, reflect the real-life loading that human joints are exposed to in daily life. For the knee (my area of greatest expertise), relevant activities would include walking or stair descent and perhaps running or jumping for athletes. It is during these activities that an individual with an injured knee is most likely to experience instability and damage other joint tissues, so the relevance of these movements to joint health is clear. I am no hand expert, but I assume that there are also some standard tasks that characterize functional use. By this definition, many measures traditionally used to assess stability are inappropriate. For example, simple laxity tests expose the knee to forces never experienced in daily living (e.g. uniaxial load, no compressive or muscle forces), and many studies that have shown little or no relationship between A/P knee laxity and functional outcomes in ACL-injured knees. Thus, though widely used, A/P laxity is poorly suited for assessing knee joint stability.

As to what response is measured, I suggest it should be relevant to the tolerance of the joint tissues. We have traditionally relied on joint translations and rotations, which are easy to describe but have little direct relevance to tissue function. Most joints have no fixed axes of rotation, and translations are described relative to floating or arbitrarily chosen points. It is not surprising that defining a threshold for significant instability has been difficult using these measures, since they cannot be easily related to tissue function. Some of the recent joint injury literature has begun to focus more on tissue-specific measures, such as the amount of elongation of a ligament graft or the magnitude and location of cartilage deformation. These are directly relevant to joint health, since they are likely to be related to conditions we care about (e.g. graft failure or osteoarthritis development).

If we choose tissue-relevant measures, then the question of what is an appropriate/stable vs. an inappropriate/unstable response becomes much easier to answer. An unstable joint is one that exposes joint tissues to damaging forces/deformations under functional loading conditions. Cadaver studies can be used to establish failure criteria, but the relevant thresholds may be much less than what would result in immediate tissue failure; chronic exposure to elevated loads can lead to gradual tissue destruction. We often don't know what the exact thresholds are for different tissues/measures, but we generally can get a good idea by examining the range of a particular measure during "high-demand" activities, or by relating the measure to degeneration or lost function in injured joints.

In the context of your hand/finger studies, a geometric model could be used to relate your flexion/displacement curves to tissue-specific measures. If arthritis is the concern, then the shear motions at the joint surfaces might be relevant. If ligament damage is most important, then some estimation of how the finger kinematics affect ligament elongation might be more appropriate.

I acknowledge that this definition for stability may be an academic ideal that is difficult to achieve in practice, especially in a clinical setting. But, I believe it also provides a framework by which simpler stability tests can be evaluated. For example, if we can show that a particular static laxity measure is a good predictor of dynamic stability (as described above), then we can establish it as a useful objective measure of joint stability. Right now, I don't think we really have a very good set of clinical tools for assessing stability for most joints.

Just my thoughts - comments appreciated!

Scott Tashman
__________________________________________________ _____
Scott Tashman, Ph.D.
Director, Biodynamics Laboratory
Associate Professor, Orthopaedic Surgery and Bioengineering
University of Pittsburgh
Orthopaedic Research Laboratories
Rivertech, 3820 South Water St.
Pittsburgh, PA 15203


Hi Paolo. Maybe we can be of some help. We wrote a new European guideline on pelvic girdle pain and worked on a definition on joint stability. This was necessary since we had to discuss in this guideline "pelvic instability" Terminology we did not like, because it does not fit the diagnostic picture of these patients. The new definition is based on studies we did in Spine and you can find them back in the reference list. Maybe it is also good to draw your attention to an upcoming world congress on low back pain. Many issues related to joint stability will be discussed (
Greetings and success Andry Vleeming
Prof Dr A.Vleeming
Ghent medical University
Belgium Chairman worldcongress lumbopelvic pain
I thought this might be a good topic to try and open to discussion, except for me, you and Robb it doesn't seem to be going that way so far, -- anyway: I understand what you want to do and in podiatry many have come up with posture indexes/indices to indicate normal joint kinematics during the static and dynamic situation. These are very nice for form filling where one can log comparative data to record change and 'prove' that the therapy 'worked'. However, for myself, I don't care how far outside the normal box my patient is, if at the same time they remain pain free and are able to successfully complete, to their own satisfaction, the daily and exceptional tasks that they choose to do. Robb Colbrun, who replied earlier, nicely explained, using engineering terms, how we would commonly define an unstable joint. The shoulder being the most unstable in those terms if we ignore the effect of the soft tissues that stabilise it and the way in which it is used most of the time. The knee is another very unstable joint and yet it works perfectly well most of the time despite having transverse, potentially dislocating, forces applied in many activities. Therefore I do not think that Robbs definition is complete enough to describe instability in the terms you are thinking of, I.E. it's usefulness (or its inadequacy) during an activity of interest. you wrote "it would be of interest to come up with some standard test that give a good estimation of the behaviour of the joint when in-vivo under physiological conditions. I am getting some good advices from other biomech-users and will definitely post a summary of the responses."There are many books that give standard / normal measures of RoM (range of motion) passive and active, these are useful as a reference point. Some also define RoMs outside these standards as hypermobile or unstable. This does not define a pathological condition itself and only perhaps the potential for pathology at best. Surely we can only judge the relevancy to pathology in terms of the patient of interest i.e. the individual and, in my opinion, one cannot describe a potential for pathology or reduction in performance in terms of an arbitrary or normal or standard RoM or measure of instability. On the contrary, often the joint that is unstable or hypermobile can enhance of facilitate the activity of choice and improve performance. The logic here is that just because (hypothetically) all Ford cars are red does not mean that all red cars are Ford's. And so just because (hypothetically again) all pathological joints are unstable doesn't mean all unstable joints are pathological, whatever measure you use to describe instability. I would be very interested in the other responses.
Regards Dave Smith
There exist precise definitions of structural instability in many books of mechanical Engineering, for example "Energy principles in Structural mechanics by T.R.Tauchert " ISBN 0-07-062925-0 pp201many examples of bars linked with springs that become unstable at certain angles are given. The principle apply for one or many degrees of freedom. if this seems of interest to you, write back and I may explain in more details.
yours very truly Paul Bourassa,
eng.U of Sherbrooke.
This is a very interesting discussion!

You may be interested in the work of Prof. Francisco Valero-Cuevas, who has done motor control and biomechanics research on the finger: has been pointed out, stability in a mathematical sense has an objective definition: a system is stable if at equilibrium a small perturbation results in the system returning to equilibrium. To use this definition then requires a statement of the system, the perturbation and the measured response. What is fascinating to me about "joint-stability" is that for a functional task it is dependent on the combined interaction of skeleton, muscles and nervous system. Clinically this poses an additional challenge. Not only does one need to identify changes instability of the joint, there is also a need to identify whether this is due to changes in soft-tissue, cartilage surfaces, muscle strength, neural activation, etc. I look forward to hearing about how you define stability for your problem.

Jeff Bingham
Bioengineering PhD Student
Neuroenginering Lab - Ting Group
Have a look at our paper (attached) in which I have used an algorithm to calculate the instantaneous axis of rotation of a movable platform. The variability of the axes of rotations was used to measure the repeatability of the axis location and this is analogous to what you might mean by joint stability. I have in fact used the same algorithm in a student project to calculate how variable the axis of a knee joint was comparing controls with ACL injured people performing an exercise which placed some stress on the knee normally constrained by the ACL and we have found greater movement of the axis after ACL injury.

It might be interesting to dig out my code and apply it to your problem whether or not you are at Uni Liverpool. Let me know if this sounds feasible to you.

Ciao Paolo, good point!
Alla caviglia anni fa leggevo il seguente: Garde, L. and Kofoed, H. Meniscal-bearing ankle arthroplasty is stable. In vivo analysis using stabilometry Foot and Ankle Surg 1996;2:137-143.

************************************************** *******************
Alberto Leardini
************************************************** *******************
Paolo and Biomech-l,

Let me start my comments with an example to illustrate the ideas of stability and control for the sake of subscribers who aren't sure. Many people believe the Wright Brothers invented the airplane in 1903.They did not. In 1901, Wilbur Wright noted there were plenty of flying machines and lots of engine designs to go with them. The ability to control these components, he explained, would usher in the age of heavier-than-air flight. The Wright Brothers own patents on devices to control the motion of their flying machine. In biomechanics, we know the skeletal system and the muscles that affect joint position as the Wrights understood their systems. Controlling the dynamic processes in gait to prevent falls, negotiate obstacles, carry loads, adjusting to terrain are analogous to maintaining altitude and course, maneuvering, take off and landing. Instability in any area of gait (or flight) may result in system failure or loss of function. A biological system has a functional envelope: homeostasis. In engineering terms, this is steady state In the uncontrolled condition, Newton's Laws are appreciated. A body in motion or at rest keeps doing that until acted on by a force changing position and speed setting a new steady state. A feedback control system responds to the outside effect to restore the steady state. In the airplane, the pilot is a crucial part of the feedback control although electronics and computerized systems are part of this system as well. When system failure occurs (crash), the structural and mechanical parts are examined but so are the command and control components. In joint motion, stability comes from the structural (skeletal) and mechanical (muscles, tendons, ligaments) parts and from the command and control components (central and peripheral nervous system). As an example of joint stability, consider the ankle. In the case where a clinically, excessive passive range of motion is noted and the subject is asymptomatic, a dynamic control system is at work, the ankle joint axis functions where it should and proximal and distal joints work within normal limits. Another ankle with normal range of motion subjected to a severe sprain which tore the lateral joint capsule has a disturbed proprioceptive feedback: functional instability. Adjacent joints are affected by the uncertainty in placing the ankle and the involved muscles function outside their normal phasic activities. I suspect you are correct in finding no stability criteria. The clinical assessment is too subjective and not dynamic. An investigation into the airworthiness of a bee or a stealth fighter based on form alone would reach a conclusion that neither could fly. I would recommend you continue to investigate dynamic challenges to your system. The Wrights performed extensive wind tunnel experiments (in vitro) before they were able to get their system operational. I think you'll find tracking the instantaneous axis of motion an important component of stability in general and that different criteria are appropriate based on joint size, type and function. At the PIPJ in particular, motion in the joint is affected by insertions distal to the joint and the extensor sling apparatus so it's no surprise you can find significant differences in flexion or extension indifferent configurations. What results would you expect (as compared to the in vivo system) if you kept each separate muscle connection intact and tested the range of motion it produces when other soft tissue is disconnected? Is the muscle still able to perform its function? Which conditions allow hyperextension or cause medial or lateral deviation? As a podiatrist, let me point out: the feet also have PIPJs and their functional requirements may be quite different from those in the hand. Hammer toes, claw toes, etc., may be the result of instabilities at these joints due to failure of neuromuscular components. When you know something about finger instability, maybe that can help understand toe instability.

James A. Furmato, DPM, PhD
Chief Engineer, Gait Study Center
Assistant Professor, Department of Orthopedics and Medicine
Temple University School of Podiatric Medicine
TUSPM Gait Study Center
148 N. 8th Street
Philadelphia, PA 19711
Phone 215-625-5370
Cellular Phone 609-933-2017