Thoughts on assessing and training proprioception for joint instability

There are a few tests that propose to measure the original definition of proprioception like the upper or lower limb Y balance test and the use of laser pens to test error or joint position recreation. I stress here, as I did at the end of part 1 that we are measuring movement output and looking at its stability to maintain task performance under the environmental and organismal affordances.

But as we now know, the proprioceptive-kinaesthetic coupling is embodied and embedded within the environmental context and entwined with affective dimensions of bodily resonance. One of the major problems I see is that none of these aspects are considered when performing error tests with laser pens or comparing scores with the Y balance of the upper or lower quadrants.

Could we make it more ecologically friendly by using the PK-coupling framework and manipulate the environment we do the Y balance or laser error test in e.g. surface, noise, visual field? We could change the bodies orientation by using external loads like weighted vests or holding dumbbells. For people who are going to be engaging with other people in their valued activities or sport, you could utilise the presence of other people and with this change their bodily resonance e.g. an aroused state (sporting-type arousal!! This is not turning into an erotic novel…) to try and elicit an affective dimension in the person of interest.

These are all possibilities, but they seldom replicate what the output might be under the plethora of scenarios in the real world – and there is no way we can know this! This is likely why multiple meta-analyses of the Y balance tool highlight it lacks the predictive ability of the composite or individual score for the upper and lower quadrant tests for future injury (Eckart et al., 2025; Plisky et al., 2021).

Remember we are dealing with behaviours of non-linear systems, so input is not directly proportional to the output!

The concept of scoring is another issue we must grapple with. The scores of the laser pen and Y balance exercises either decrease or increase respectively over time. The problem is these changes are based off a linear paradigm of movement – the mean is the goal and the error rate the standard deviation from that, but we know that human movement (by extension PK-coupling) is not linear. The application of a linear-based test to a non-linear concept (movement) is problematic at best. We are assuming that a reduced error rate using a laser pen means we have improved our proprioceptive capacity, yet as discussed in part 2 the evidence for motor abundance and the consequential movement variability is inherently good (so long as it falls within the upper and lower boundaries) for task performance. So, a reducing score relative to our starting point suggests we are encouraging rigidity not variability following training – this is the opposite of what we want to achieve.

Following this perspective, it affords the Y balance test some ecological validity as an increasing score suggests an increased ability to successfully complete a task moving through the three planes of movement. Therefore, if we are looking at proprioceptive training for shoulder instability then an increase from our baseline score suggests improvements in variability if variability is defined here as total movement excursion in different directions.

In further dismay, I know of no tests that consider the non-linear aspect of movement when it comes to proprioceptive assessment. My thoughts at current are to run with the idea that we can measure output of the system and that we can manipulate at least some variables. If we have a movement of interest, say the lateral or diagonal cut, could that be our test and our benchmark for success is the ability to complete the cut under different conditions?

These might be as follows: the standard cut, cut with eyes closed at point of outside foot contact, cut with external load e.g. weighted vest, cut with unilateral load, cut onto unstable surface, cut with opponent running toward you, cut with different types of music to elicit different affective responses and bodily resonance… It is just an idea but one that could be applied to largely any scenario and any part of the body!

Considering the revised view of proprioception, I struggle to see the utility of the laser pen test and have reservations around for the Y balance test as neither appear to measure our intended target in a way that even closely replicates the real-world interaction of the self-ecological-other coupling that is ever-present. In the early stages of the rehabilitation I will use the Y balance test as a proxy for tracking the increased capacity to move in the three planes successfully but once symmetry is within 5% I stop using it as it serves little further utility.   

Perhaps then, our assessment of the client or athlete needs to identify the key movements required for successful performance of their task, then track their ability to successfully complete the fundamentals of the task under varying conditions. Our barometer for a successful response to training would be completion of the task under increasing variable conditions without a decline in performance.

The three dimensions of the PK coupling contingencies offer a framework for how one might approach assessment and rehabilitation.

From an assessment perspective, understanding these contingencies interact at the level of the self, the environment and self-environment-other should encourage questioning around these levels of integration. How does the person (and the wobbly joint) feel moving in relation to their opposite limb, how does movement feel when presented with tasks in their environment e.g. approaching a step down in the case of knee instability, or lifting the arm overhead to reach an object in the case of shoulder instability.

Do they feel different under different environmental conditions; under circumstances where it feels better or worse what has changed in their interpretation of possibilities for action.

Throughout these, pay attention to the bodily resonance. How is the movement making them feel or how do they feel before making the movement? If there are resonances generated, questioning around their thoughts, feelings and emotions during these movements (and consequences of movement) is important. During different movement scenarios, how does this bodily resonance change.

Using the BMJ model for return to activity, practise and performance, we can consider the PK coupling contingencies at difference phases and how they will progress.

Within the return to activity, the focus is on re-establishing a sense of basic control so that the limb can move relative to the rest of the body with comfort. Such a phase lends nicely to the PK self and PK environment. The focus should be on spatial-temporal self-orientation of the limb and what It can do relative to the contralateral limb and wider self. Body resonance here is related to how we feel moving our body, and limb relative to each other. Although not a hard and fast rule, closing the chain with our exercises when establishing limb-body orientation may be appropriate, the increased feedback from our internal control systems make it easier to integrate the limb into our bodily actions. Movements involving the whole body here should be considered; integration of the limb into the body follows the notion we are seeking to re-embody the limb within (our) bodily movement. Using resonance to our advantage, experimenting with use of music to evoke positive feelings of bodily resonance is something to consider.

Graded motor imagery (GMI) has shown promise for improving muscle strength (Li et al., 2022), range of movement (Araya-Quintanilla et al., 2020), reduced pain catastrophising (Araya-Quintanilla et al., 2020) and pain intensity (Li et al., 2022; Araya-Quintanilla et al., 2020; Sırlan et al., 2025) in people following total knee arthroplasty (Li et al., 2022) and persistent rotator cuff related shoulder pain (Araya-Quintanilla et al., 2020; Sırlan et al., 2025). The RESOLVE randomised trial focused on graded motor imagery in people with chronic low back pain and found between-group differences in improvements in pain of 1 (0-11 NRS scale) in favour of the intervention, which is above the clinically importance difference (Bagg et al., 2022).

The quality of studies investigating GMI are limited with some being single-arm trials (Araya-Quintanilla et al., 2020) and others showing no differences in outcomes at final follow up time points (Sırlan et al., 2025) so suggestions around implementation should be taken with caution. Nonetheless, in the return to activity and return to sport phase, graded motor imagery of task performances could be used as an adjunct to the physical completion of movements and tasks.

In a similar fashion to the priming principle (post-activation potentiation) used when combing heavy strength sets with plyometric exercises, using GMI or visualisation before or after performing movements could be utilised to create experiences of movement that the shoulder or joint of interest is not yet able to perform. I have no research evidence to suggest this is effective, but it would seem the GMI body of research shows some promise… it certainly is not going to do any harm!

In the enactive-ecological approach to pain, the persons’ opportunities for action become stuck and consequently reduce. I would argue this extends to the painful, wobbly joint. Considering the embodied-embedded nature of action of the person in their environment, visualisation and GMI offer the opportunity to create environments in our head and possibly expand our affordances without moving the painful joint. This may then translate into improved affordance when interacting with the real world.

The components of a program for joint instability are largely like a normal program. Load it with heavy stuff, in all planes through end range joint positions with a sprinkling of hopping, throwing and reactive strength.

Isometric training

Unsurprisingly isometric training follows the principle of specificity, where the greatest changes in neurological parameters occurring at or closest to the joint angle training (Dustin Oranchuk et al., 2019).

Isomeric training has three general categories known as overcoming (or ramping), yielding and ballistic. Overcoming (also known as ramping) involves producing force against an immovable object like a rack mid-thigh pull or a squat against an inverted rack pin; yielding involves holding a position for an allotted time (often with external load) and often until failure (hence the term yielding) and ballistic which involves rapid force production in an isometric fashion with multiple repetitions often performed in succession of one another.

If we go with the notion that instability is often position specific, the use of isometrics allows us to explore movements that push the joint, say the shoulder, into those positions of instability in a controlled, safe and relatively pain-free manner. Or if we can’t directly hit that angle because there is too much instability, then we can train positions either side of it, allowing there to be a transfer of training to that desired joint angle.

The review by Dustin Oranchuk suggests the following principles that should be considered when implementing isometric training

-        Training at long muscle lengths with isometric training results in larger adaptions across the entire joint angle spectrum for changes in force parameters, tendon stiffness and muscle hypertrophy compared to shorter muscle lengths

-        Improvements in peak muscle force may be better attained via isometric training that builds to a maximal effort over 3 seconds (ramp isometric training)

-        Explosive strength (RFD in the first 0-150ms) is likely best attained by rapidly building to maximal effort in 1 second before then stopping aka ballistic isometric training

-        The greatest changes in tendon aponeurosis complex occurred following ballistic style isometric training compared to overcoming isometric training (although there were very few studies investigating this!)

Building isometrics into your training could be achieved several ways. This might include dedicating days to isometric-only training, or you could superset it in a French contrast style training approach with your larger multi-joint movements. Personally, super setting the ballistic style isometrics with your heavier dynamic strength sets seems to work well due to the short burst, max intent nature of ballistic isometrics and the ability to combine it with the rest of the training. This works well for time constrained people – which is most of us!

Throwing (and catching)

Remember we said that different sensory afferents serve different functions. Some detect length, velocity and force changes… some are more active at end of range joint positions. Throwing and catching variations can be graded easily enough to be incorporated into a program. You can modulate the velocity, the position the shoulder moves through (you could start with a small degree of freedom, or you could focus on throwing or catch & release from EOR positions), the extent of kinetic chain input, the weight of the object.

With any throwing exercises, we can incorporate cognitive influences such as dual tasking or reactive processing easily. It is clear that throwing allows you to really excite the proprioceptors however you want, and you could argue you can even be a little specific with it…

Want to really stress the Pacinian corpuscles and Ruffini endings, then train the joint at end range positions.

Want to light up the Golgi organs, throw it as fast as humanly possible with a weighted ball… this can all be done with minimal training equipment.

Heavy strength training through the three planes

There is no set formula for this. Just lift heavy stuff around the joint, exposing it to multiple planes at any one time and with movements you can easily incrementally load.

Remember we are all guilty of spending a great deal of time in the sagittal plane but rarely pushing into the transverse (rotational) or frontal (side to side) plane. Incorporate variation into your training.

I think it is worth considering supramaximal eccentrics somewhere in the program, for a short block of training (2-4 weeks). We know the Golgi tendon is reactive to magnitude of force and once that threshold is reached it has an inhibitory effect on the force-producing capacity of that muscle. Cal Dietz is a coach worth following and reading up on around this, his Triphasic Training model uses blocks of supra or submaximal eccentrics and I have found it to be a potent stimulus when used judiciously!

As a side note, if we really want to desensitise its gain threshold to force, then we need to expose it to very large amounts of force for a period. Supramaximal eccentrics are the only way to truly achieve this.

Reactive strength and Plyometrics

Reactive strength is the ability of an individual to produce force rapidly. Paul Jarvis and colleagues published a systematic review in 2022 showing that there are significant, moderate correlations between reactive strength, isometric and isotonic strength, endurance, top speed, acceleration and change of direction performance.

The reactive strength index is calculated by dividing flight time or jump height by ground contact time during a drop jump and is thought to be a proxy measure of the stretch shortening cycle (or plyometric capacity) as suggested by Flanagan et al., (2008).

Plyometric and non-plyometric exercises fall under this umbrella of reactive strength. Before jumping straight into plyometrics, it is important to consider the ability to develop force absorption before acceleration. Or as I like to call it, breaks before gas!

Landing drills that focus on developing appropriate mechanics are important as you need to be able to absorb force appropriately before you can release it during the propulsion phase. The judicious use of plyometrics throughout a training program is essential to maximising performance and stability of a joint under rapid force generating scenarios.

Ground contact times (GCT) in the region of 10ms (0.1 seconds!) have been observed for elite hurdlers and sprinters, emphasising the need to address and train the fast SSC movements (Flanagan & Comyns, 2008). Considering these observations, you can split RFD exercises by their ground contact time fast (<v250ms ground contact time) versus slow (> 250ms ground contact time).

For the lower limb, plyometrics are often split by their GCT as we discussed. A-drills, ankling drills, drop jumps, bounding and sprinting are cardinal examples of fast SSC movements whilst CMJ, squat jumps, depth jumps are examples of slow SSC (GCT > 250ms).

Depth jumps taken in the context of slow SSC are primarily performed as a means of shock training, developed by Verchoshansky and as quoted by Natalia Verchoshansky “depth jump (shock training) produces a sharp muscle tension which creates and stimulates a high-intensity neurological adaptation across the motor system resulting in faster switching of the muscles from eccentric to concentric work and an overall larger increase in muscle power” (Verkhoshansky, 2012).

For the upper limb, there exists no division between fast and slow SSC. Movements are performed with maximal intent with the aim of minimising the eccentric, amortization and concentric phase of the movement. Exercises like the drop & catch exercise, band-assisted push ups, drop push ups and medicine ball slam variations are all examples. Ballistic isometrics, given their rapid nature would also classify under an RFD-based exercise.

We need to add chaos to controlled environments to replicate the infinite possibilities the person will encounter in their chosen activity.

The latter stages of the return to activity phase and certainly all stages of the return to sport and penultimate performance should introduce increasing chaos for joint instability. The premise is to increase exposure to variability and build the richness of behavioural outputs we previously discussed in parts 1 and 2.

Meredith and colleagues published a paper on the visual chaos-control continuum, this is a useful starting point for beginning to consider creating chaos (Chaput et al., 2024). Take a movement we know is fundamental to the athlete’s sport, the tackle for example. No one tackle is the same with the force characteristics, the visual stimulus, the position of the arm and body, preparedness of the athlete and cognitive load. We need to build this into the training to prepare the individual for returning to their chaotic environment. We could take the push up as our base movement pattern. To build in chaos we can have the athlete move one arm relative to the other or one leg relative to the other in the direction of a coloured cone depending on which one has been called out. We could have them standing close to a wall, perform a falling start into the wall and react last minute to an auditory stimulus to place the arms in certain positions to absorb force.

This concept of shifting from control to chaos is key for the return to sport and performance phase for the wobbly joint. Cognitive and visual loading tasks as discussed above are important, but we can also utilise music to evoke bodily resonance whilst performing these tasks, again, considering contact or combat sports where there are high levels of arousal, we need to have exposed the athlete to movements under stressful / high arousal conditions to emulate exposures they may face on the field. Another way to achieve this is through partner-based competitive drills that involve trying to physically manipulate the other individual to win.

Putting it all together using the COMP checklist

Programming can be as complex or as simple as you want it to be. Understandably, in the busy clinic environment we often do not have stacks of time so having useful frameworks to help structure thinking is something I find useful. COMP serves as a quick checklist and stands for Characteristic output, Movements and Performance.

Characteristic Output refers to the key qualities that the body needs to be able to express. These are focused around

·        Force characteristics (peak, velocity, positive or negative, rate of force and time)

·        Energy systems used e.g. alactic, glycolytic, aerobic

·        Psychological and emotional qualities

Movements refers to the fundamental patterns force is expressed through in their sport or valued activity and are focused around

·        Single leg or double leg loading

·        Planes of movement (frontal, sagittal, transverse)

·        Joint positions commonly achieved and range of positions likely to see (think end range position possibilities)

·        Linking of movement positions together to form patterns of movement i.e. how one movement from one position to another. For example, what are the movement considerations for linear run into deceleration when we are linking it onto a diagonal cut versus a 180-degree turn.

Performance considers the context specific qualities we might need for an edge – what makes the absolute best athlete in their chosen sport.

·        Reactive elements focused on the PK self-environment and self-other aspects.

o   Self-environment would be things like moving objects with undetermined flight paths e.g. balls / apparatus.

o   Self-other consider game athletes where you have teammates and opponents moving an interacting with the environment and the agent.

·        Sensory integration and reweighting. Examples would be performing a cut whilst visually tracking a target in the opposite direction

·        Context specific psychological or emotional qualities

References

Hillier, S., Immink, M., & Thewlis, D. (2015). Assessing proprioception: a systematic review of possibilities. Neurorehabilitation and neural repair, 29(10), 933-949.

Sibley, K. M., Beauchamp, M. K., Van Ooteghem, K., Straus, S. E., & Jaglal, S. B. (2015). Using the systems framework for postural control to analyze the components of balance evaluated in standardized balance measures: a scoping review. Archives of physical medicine and rehabilitation, 96(1), 122-132.

Stergiou, N., & Decker, L. M. (2011). Human movement variability, nonlinear dynamics, and pathology: is there a connection?. Human movement science, 30(5), 869-888.

Oranchuk, D. J., Storey, A. G., Nelson, A. R., & Cronin, J. B. (2019). Isometric training and long‐term adaptations: Effects of muscle length, intensity, and intent: A systematic review. Scandinavian journal of medicine & science in sports, 29(4), 484-503.

Jarvis, P., Turner, A., Read, P., & Bishop, C. (2022). Reactive strength index and its associations with measures of physical and sports performance: A systematic review with meta-analysis. Sports medicine, 52(2), 301-330.

Flanagan, E. P., & Comyns, T. M. (2008). The use of contact time and the reactive strength index to optimize fast stretch-shortening cycle training. Strength & Conditioning Journal, 30(5), 32-38.

Verkhoshansky, N. (2012). Shock Methods And Plyometrics: Updates And In-Depthexamination. CVASPS, hh, 12-44.

Tuthill, J. C., & Azim, E. (2018). Proprioception. Current Biology, 28(5), R194-R203.

Fuchs, T., & Koch, S. C. (2014). Embodied affectivity: on moving and being moved. Frontiers in psychology, 5, 508.

González-Grandón, X., Falcón-Cortés, A., & Ramos-Fernández, G. (2021). Proprioception in action: a matter of ecological and social interaction. Frontiers in Psychology, 11, 569403.

Li, R., Du, J., Yang, K., Wang, X., & Wang, W. (2022). Effectiveness of motor imagery for improving functional performance after total knee arthroplasty: A systematic review with meta-analysis. Journal of Orthopaedic Surgery and Research, 17(1), 65.

Araya-Quintanilla, F., Gutiérrez-Espinoza, H., Jesús Muñoz-Yanez, M., Rubio-Oyarzún, D., Cavero-Redondo, I., Martínez-Vizcaino, V., & Álvarez-Bueno, C. (2020). The short-term effect of graded motor imagery on the affective components of pain in subjects with chronic shoulder pain syndrome: open-label single-arm prospective study. Pain Medicine, 21(10), 2496-2501.

Sırlan, S., Alaca, N., Yarar, H. A., & Başcı, O. (2025). Graded motor imagery as an adjunct to comprehensive physiotherapy in chronic rotator cuff-related pain: a single blind randomized controlled trial. BMC Musculoskeletal Disorders, 26, 588.

Heino, M. T., Proverbio, D., Marchand, G., Resnicow, K., & Hankonen, N. (2023). Attractor landscapes: A unifying conceptual model for understanding behaviour change across scales of observation. Health Psychology Review, 17(4), 655-672.

Hiver, P. (2015). Attractor states. Motivational dynamics in language learning, 20-28.

Negrello, M., & Pasemann, F. (2008). Attractor landscapes and active tracking: the neurodynamics of embodied action. Adaptive Behavior, 16(2-3), 196-216.

Newell, K. M., Van Emmerik, R. E. A., & McDonald, P. V. (1989). Biomechanical constraints and action theory. Human Movement Science, 8(4), 403-409.

Vaz, D. V., Stilwell, P., Coninx, S., Low, M., & Liebenson, C. (2023). Affordance-based practice: an ecological-enactive approach to chronic musculoskeletal pain management. Brazilian journal of physical therapy, 27(5), 100554.

Bagg, M. K., Wand, B. M., Cashin, A. G., Lee, H., Hübscher, M., Stanton, T. R., ... & McAuley, J. H. (2022). Effect of graded sensorimotor retraining on pain intensity in patients with chronic low back pain: a randomized clinical trial. Jama, 328(5), 430-439.

Bayrak, A. (2025). Exploring injury profiles in professional football: evidence from a five-year study and the role of the functional movement screen. BMC Sports Science, Medicine and Rehabilitation, 17(1), 217.

Dorrel, B. S., Long, T., Shaffer, S., & Myer, G. D. (2015). Evaluation of the functional movement screen as an injury prediction tool among active adult populations: a systematic review and meta-analysis. Sports health, 7(6), 532-537.

Tabor, A., Keogh, E., & Eccleston, C. (2017). Embodied pain—negotiating the boundaries of possible action. Pain, 158(6), 1007-1011.

Latash, M. L. (2012). The bliss (not the problem) of motor abundance (not redundancy). Experimental brain research, 217(1), 1-5.

Plisky, P., Schwartkopf-Phifer, K., Huebner, B., Garner, M. B., & Bullock, G. (2021). Systematic review and meta-analysis of the Y-balance test lower quarter: reliability, discriminant validity, and predictive validity. International journal of sports physical therapy, 16(5), 1190.

Eckart, A. C., Ghimire, P. S., Stavitz, J., & Barry, S. (2025). Predictive utility of the functional movement screen and Y-balance test: current evidence and future directions. Sports, 13(2), 46.

Chaput, M., Simon, J. E., Taberner, M., & Grooms, D. R. (2024). From control to chaos: Visual-cognitive progression during recovery from ACL reconstruction. journal of orthopaedic & sports physical therapy, 54(7), 431-439.