Viewing balance exercises with eyes closed

For a long time, I have questioned prescribing balance exercises with eyes closed to athletes in sport. Regular readers of the blog will know that I continuously explore the clinical reasoning behind treatments and interventions but have a particular interest in exercise prescription. I have to admit that single leg balance with eyes closed is an example of exercise prescription that just doesn’t make sense to me, how many athletes close their eyes to perform a sport related task? I’m regularly seeing discussions online about “what is functional?” and most of the debates are based around semantics without much weight behind them but provide a good opportunity for people to have a little disagreement about something. To avoid getting into a debate about “functional” I thought it best to better understand the concepts and demands behind “balance” to see if I can answer the “why” behind balance exercise progressions.

Now stay like that for 1 minute or until another player throws a ball at your face
One argument for closing eyes during balance exercises is to remove the visual stimulus and encourage the athlete to challenge vestibular and proprioceptive senses. Remove one thing and make others compensate for this deficit. In a study of track athletes, sway velocity (cm/s) increased two-fold when athletes closed their eyes during a static balance test (here) but the only significant finding in the study was the difference in centre of pressure displacement (cm) between non-dominant and dominant limb across the medial-lateral plane. So, no difference between male and female athletes and no difference between “eyes open” and “eyes closed”.

So how does this explain the increase in sway velocity? The sway velocity is the area covered in both the anterior-posterior and medial-lateral planes of the centre of pressure per second, indicating speed of correction. The fact that the displacement between “eyes open” and “eyes closed” was not meaningful suggests that the demand on the fine motor correction increases. A decent argument to include “eyes closed” in a balance program, if that is the aim.

Static balance in dynamic sports

Compared to dynamic balance tests, static tests do not allow re-positioning of the centre of mass within the base of support, so the athlete becomes more reliant on smaller corrections. Different sporting populations have demonstrated varying abilities in static and dynamic balance skills, with gymnasts outperforming in static balance but soccer players demonstrating better dynamic balance (here).

This may seem obvious given the control on the balance beam vs changing direction to avoid an opponent. But actually, perhaps where the argument becomes more broad and complex.

As with any exercise selection, it needs to be appropriate to the aims of the rehabilitation program and the demands of the sport, taking into consideration open and closed skills and linking these to fixed gaze drills vs dynamic gaze drills.

Have we gazed over “skill”?

In a given skill, experts can recognise which cues are relevant and avoid information overload (Martell & Vickers 2004). Below is a slide from my presentation “3 sets of when?” It explains the concept that following any injury, the athletes ability to perform a given skill returns (temporarily) to novice level.

skill level injury

Take a skill like walking. Immediately after an ankle sprain, your ability to perform that skill at an expert level is decreased. A skill that has taken years to perfect, to become automatic, now becomes a task which requires concentration. Thankfully, the return to expert level doesnt take years (hopefully!) and this is where our exercise selection becomes crucial to optimally load and sufficiently challenge. We can’t presume that the pre-injury skill level is the same post-injury. We should also consider experience of the balance task specifically. I can think of experiences where athletes are standing on one leg on a Bosu throwing a reaction ball at a 45 degree trampoline. “Oh you’re no good at that are you… we need to address your balance”

I’ve digressed slightly from single leg balance with eyes closed… and actually I still haven’t discussed “gaze control”.

off on a tangent

Gaze control links specifically to experience of a task. Comparing those skilled at orienteering to non-skilled (here) demonstrated an increased ability of the orienteering folk (what do you call people that go/do orienteering?!) to employ a wide focus of attention and to shift efficiently within a peripheral field. The test very cleverly measured gaze control to flashing images with varying degrees of relevant and irrelevant information. What is interesting from this study was that the control group where physically active and proficient in other sports, but the “skill” advantage lay with the orienteering-iers. [shrugs and thinks “sounds right”].

I did not know that about balance!…

Elite athletes have heightened spatial awareness and processing capabilities vs their non-elite counterparts, where gaze control is cool and calm, with long duration of fixation of specific locations. This results in better body positioning end efficient limb actions (here). What better example than ballet. When comparing professional dancers to controls walking along a thin taped line, it was observed that experienced dancers focus far into space, delivering effortless and accurate movements where as controls looked down and focused on the line, moving with greater speed and less control (here). Dancers shift their neural control from somatosensory inputs and to an increased use of visual feedback, via peripheral fields and focused gaze control. Interestingly, sub-maximal exercise has been shown to increase visual attentional performance (posh words for reaction time) and a decreased time need to zoom focus of attention (here). This is useful for prescription considerations.

This efficiency has been demonstrated in other studies also, where the addition of a 4-week balance training program to Physical Education classes in school resulted in increased CMJ, Squat Jump and Leg Extension Strength (here). A time period that can’t be associated with physiological adaptations to muscles (regardless of time, they did balance exercises!) and even when a balance training program has been compared to a plyometric strength program (here). It is thought that improved centre of pressure is linked to spinal and supraspinal adaptations, due to high inter-muscular activation and co-ordination.

My question for any budding researchers out there… if there is a spinal level involvement here, can we utilise the contralateral limb at the very early stages of injury to improve balance on the injured side?

Finally, I get to my argument… balance is the output. Balance and proprioception are different entities, as are gaze strategies and balance. But they may all be interlinked via “skill.”

In researching this blog, I’ve certainly become more accepting of “eyes closed” as an addition to balance programs. But also think I’ve gained more clarity on appropriate prescriptions and the suitable progressions for individuals.

Perhaps “eyes closed” is not a progression, but a starting point!

Immediately post injury, we are looking to internalise feedback (intrinsic) and focus on local, fine movements. There are plenty of regressions within “eyes closed” balance that we can make the athlete safe from secondary injury. Graded progressions from static to dynamic, trying to keep the demands appropriate to the skill required to return the athlete to “expert”.

From here, our progressions should not be the removal of a visual stimulus, but instead optimising and enhancing gaze control:

  • Focus on a stationary target –> moving target
  • Head still –> head moving (repeat stationary and moving target progressions within this)
  • Static balance –> dynamic balance (repeat progressions above)

Essentially, we progress through from intrinsic cues to extrinsic cues, where gradually the athlete is thinking less and less about the mechanics of balance and more about skill execution and performance. We know that gaze control components improve with sub-maximal exercise, so our ordering of our program can reflect this. It is commonplace for balance exercises to be at the beginning of the program, but if balance is our primary aim for rehabilitation, perhaps it should be later in the schedule.

I don’t think this is too dissimilar to how most people prescribe exercises, but for me at least it has given me a better thought process into the “why” which ultimately should make rehabilitation programming more effective and efficient and therefore more elite.

Yours in sport,


Exercise Progression & Rehab Programs

A year or so ago, I put on a CPD evening for our part time staff at the football club discussing exercises and the clinical reasoning behind developing a program (needless to say I got talking about the use of clams for a quite a while – clam blog). In this presentation, I started drawing my reasoning process onto powerpoint using some coloured blocks to help visualise the theory that I was trying to describe.

The theoretical model was recently published in Physical Therapy in Sport and I thought I would use this blog to try and discuss it in a less formal way than the writing style allowed in publication.


The model (here) is designed to be fluid and adapted to any individual by any level of clinician. Let me quickly introduce the components:

A theoretical model to describe progressions and regressions for exercise rehabilitation (Blanchard & Glasgow 2014)


  • The triangular blocks (1) represent the fundamental exercise, the core ingredient that will remain throughout the progression. The arrows running up the side of the triangles represent an ongoing progression throughout the rehab process such as speed, duration, repetition etc. So basically, something that can’t be affected by the stimuli that are added or removed. If you add an unstable surface to an exercise, you can still progress by increasing the duration.
  • The coloured blocks represent a stimulus that will help the exercise progress. This can be one of two things;
  1. Internal – something that the patient has to focus on intrinsically. A decreased base of support for example, where the patient must focus on the balance element of an exercise.
  2. External – the addition of something to the exercise that takes the patients focus away from the movement or action they are performing – adding a ball to a running drill, or a verbal command that initiates a change in direction.

The blocks are interchangeable and can be added / removed at the clinicians discretion.

  • Adding a new block, which will progress the exercise, is accompanied by a regression of the “gradient” on the blue triangle. Creating a step-like progression across the model. As you progress with an internal or external stimulus, its important to bring the difficulty levels back down, so reducing repetitions or speed or duration. This allows the pateints to adjust to the new stimuli without fear of re-injury or task failure. When teaching a child to ride a bike with stabilisers, you don’t take them off and ask them to cycle at the same speed you did with them on. For that reason, you wouldn’t get someone going from 30 reps of a hamstring bridge straight into 30 reps on a single leg bridge as a progression. You would decrease base support and reduce reps to allow adaptation.
  • Adding a “block” doesn’t mean you have to add something to the exercise. The block represents a step up in their progression. So progressing from two legs to single legs is technically “taking away base of support” but is an addition to the ongoing progression.


Lets use an example, recently I started designing a program for a teenage footballer with a proximal adductor strain. New to professional football with no history of conditioning.

In the sub-acute stage, once intial pain had settled, we began looking at his movement patterns and stability and noticed a huge imbalance with his left sided control through sagittal and transverse planes compared to his right. He is left footed, so his plant leg (right) is used to supporting his body weight.

His body awareness and “physical literacy” was so poor we had to regress him right back to basics. The following represents a small proportion of a larger exercise program. I’m not usually an advocate of planks in a multidirectional sport like football, but in this case, his single plane control was so poor that I swallowed my pride and began with basic planks.


When I say basic, we reverted to short lever planks with the knees on the floor – this was the only was we could get him to control the relationship between his trunk and pelvis. Looking at the model, this short lever plank would be the singular blue triangle at the start (1). We built up the duration of the hold from 30 seconds to 90 seconds over time. This would be the arrow running up the gradient of the triangle.


The addition of the first block (2) was to increase the length of the lever so that he now has to hold a traditional plank. In doing so, we dropped from 90s hold back down to 30 seconds and over time, built up to 90s. (These are just arbitrary times, based on no real evidence).


The next block we added was a rotational element (3), but to ensure the progression wasn’t too sharp, I removed the long lever and returned to a short lever position. I then asked the player to move a light 1.25kg weight from his left side, with his right hand and place it on his right side. Then with his left hand etc etc. The purpose of this was to introduce a transverse task to a sagittal plane activity – as the arm moves from the ground and across the body, the player has to control the rotation through his trunk and avoid rotation at the pelvis. Instead of duration, we built up repetitions over time.


Now that we were confident he could hold a plank, and control rotation in a short lever plank, we could combine the two blocks as the next progression. Now in a long lever plank with a rotational element.


The next progression was to add an unstable surface (4). To do this, the player performed a plank with his thighs on a gym ball. This in itself was quite easy so we instantly added a rotational component with an unstable surface, gym ball pelvic rotations (see video here). So now on the model, we have the basic “plank” triangle at the top, a block underneath to symbolise the long lever, another block to symbolise rotational control and a third block to symbolise an unstable surface.


“The length of time required by an individual to master a task has

been described as a linear function that begins quite rapidly with

the introduction of a new task and then plateaus or slows over time

as practice continues (Gentile, 1998).”



This is a very simplistic example of how the model works, but hopefully it demonstrates the fluidity that is intended with it and how the blocks are interchangeable and can work independently or as part of a more complex progression. Every program you write will be individual and the progressions will be different, therefor every model will look different. Some will continue longer than others, some may be shorter than the one I’ve described here. Some will end up with taller columns due to the number of progressions. The width of one column compared to its neighbour may be different size due to the length of time it takes for the patient to master. And so on and so on. If I continued, hopefully I could have ended up with the player doing this:

But whats the use of that defending a counter attack?


Like many conversations I begin or poor jokes I tell, this may be one of those things that only makes sense in my head, but I would love to hear if it makes sense to others – if you think it works and examples of doing so.


Yours in Sport