Shoulder Impingement… some extra thoughts

Should we be using the term shoulder impingement or subacromial impingement? It’s certainly not a diagnosis, instead its rather a vague description that doesn’t tell us or the patient anything helpful about what is causing their shoulder pain, in fact it is completely misleading!
For starters shoulder or sub acromial impingement is actually a perfectly normal anatomical event!
Every time you raise your arm beyond a few degrees you are in fact compressing or ‘impinging’ the various soft tissue structures that reside in your sub acromial space between the humeral head and acromial arch, namely the sub deltoid bursa, Supraspinatus tendon and parts of the Infraspinatus and Subscapularis tendons.
So if compression of these structures is a normal action why do we use it to explain to a patient why their shoulder hurts when they lift it up? We don’t tell people with knee pain when they squat that they are getting meniscal impingement do we, but this normal anatomical event happens every time you bend the knee, its only when the mensicus is damaged do they get pain, so we use more accurate descriptive terms like meniscal lesion, tear, degeneration etc, so why don’t we do this for the shoulder? Surely its obvious that it’s only when the sub acromial structures become pathological do we get shoulder pain, it’s not the impingement that causes pain.
So why do we use this term?
Well if we are being honest, the main reason is that it is extremely difficult, if not impossible sometimes, to accurately distinguish the true source of pain to an individual structure in the sub acromial space, and so to ‘hedge our bets’ and perhaps to make it easier to explain to our patients, we use a non descript term like impingement instead of more accurate anatomical terms like sub deltoid bursitis, supraspinatus tendionpathy or tear etc.
However, I think this is a poor excuse and a mistake, I don’t think using the term impingement helps us or our patients at all.
First, in my opinion, impingement paints a rather gruesome picture for patients of things being adversely pinched and squashed and so can start some negative imagery and fear inducing psychosomatic processes, potentially increasing pain felt in the shoulder, as well as possibly making our patient more reluctant to move their arm, maybe adding to altered movement patterns or secondary shoulder stiffness, further compounding problems.
Impingement could be thought of a bit like those other terms we dislike as physios eg ‘slipped disc’ or ‘crumbling spine’… ok maybe not as bad, but it’s still a negative term that could affect our treatment even before its begun.
Secondly impingement doesn’t help us as clinicians as without knowing (or having an general idea) what structure is (most likely) at fault how can we effectively treat it, for instance a bursitis will be managed and treated completely differently from a tendinopathy, as would a cuff tear!
However being a realist I know that the term shoulder impingement is probably here to stay, it’s so well ingrained in our medical terminology to just disappear, so I have to resign myself to that, but I am forever the optimist, and I propose just like Dr Jeremy Lewis has said before me, that instead of using the term ‘impingement’ we should instead start using the term ‘sub acromial pain’ yes ok, it’s another undescriptive term but one that I think is more sound than ‘impingement’.
With this in mind we should also try harder to be more specific with our diagnosis by describing more accurately the most likely anatomical structure causing the shoulder pain for our patients, yes this is a challenge but there are methods that can give us better indication of suspected sources of pain in the sub acromial space. To start with we need to try to differentiate between the sub categories of sub acromial impingement, these are location, if it’s primary or secondary and if it’s internal or external.


There are two main areas that generate pain in the sub acromial space (actually three if we count internal impingement, and even possibly four if we include sub coracoid compression although it’s not part of the sub acromial space) these are the acromial arch and the coracoacromial arch, see the picture below
acromial arches
Sub Acromial Arch
This is compression occurring more posteriorly between the hard bony acromion arch and the humeral head, it affects mainly the supraspinatus tendon but can also affect a portion of the infraspinatus tendon as well as the sub acromial bursa that overlies these structures.
Sub Coracoacromial Arch
This is compression occurring more anteriorly between the ligamentous arch formed from the coracoacromial ligament and the humeral head, this affects again mainly the supraspinatus tendon but also part of the Subscapularis tendon as well as the sub acromial bursa again, and possibly the long head of biceps tendon, although this is debatable as it sit within its sulcus (groove) protecting it from compression.
Sub Coracoid
Finally a third area of potential compression not to forget although not part of the sub acromial space, is between the humeral head and the coracoid process, this is usually the Subscapularis tendon and the sub coracoid bursa in horizontal adduction, but it can also be parts of Supraspinatus in horizontal adduction when combined with internal rotation
Differentiation between Sub Acromial and Sub Coracoacromial Arch Compression
As you can see there is an overlap of structures that can cause pain in both of these areas, namely the supraspinatus tendon and the bursa, so true differentiation as I mentioned earlier is difficult if not impossible to achieve with just clinical testing alone, and as far as im aware no research has found any truly reliable and validated method of differentiation with physical testing, and there is also the question is there a need to?
However… there are, some positions that can give an ‘indication’ if one area is more a suspect over the other so maybe helping isolate a structure, for example placing the shoulder in abduction and internal rotation is thought to compress the supraspinatus tendon under the coracoacromial arch more, where as in more flexion and internal rotation it will compress it more under the bony acromial arch.
A cadaver study by Hughes et al also shows that some of these sub acromial structures are getting compressed in some differing shoulder positions than we originally thought, for example shoulder extension and external rotation seems to generate high compression between the supraspinatus and the bony acromial arch, but if this correlates with clinical findings and in ‘live’ shoulders is yet to be tested. (This would be a nice MSc or research study for someone to do, any takers?)

Primary or Secondary ‘impingement’

This is where it’s thought that the bony anatomy is to blame for adverse compression. One of the most common causes of this is thought to be the shape of the acromion of which there are three broad variations flat curved and hooked, also the acromial can developing and ‘grow’ bony spurs over time called osteophytes. If these are getting in the way within the sub acromial space they may need to be removed surgically so identification of these is important before you commence rehab and possible waste time and cause unnecessary pain to your patient, a simple x-ray is usually best to look at this, as well as checking for any calcific deposits in the tendons!
acromial types
This is where the shoulders biomechanics are thought to cause adverse ‘impingement’, so postural or sport related adaptations effect the shoulder joint positions, movements and imbalances that either allow the shoulder to ‘wobble’ or ‘move’ from is desired position or that make it doesn’t move enough or quickly enough when the arm is moved around.
There are numerous scenarios that are thought to cause this, rotator cuff weakness being one, for example when either or both the supraspinatus and subscapularis are overpowered by the deltiod muscle so allowing the humeral head to rise upwards and or forwards, losing humeral head centering on the glenoid so increasing the compression forces or time of compression on the structures in the sub acromial arch.
Another biomechanical cause that is thought to commonly occur can be adverse postural tightness of the anterior muscles around the shoulder and scapula with associated weakness of the muscles around the posterior of the shoulder and scapular, these allow postures that put the scapular in a forward and downward rotational position, this is though to place the aromial arch in a detrimental position even before the arm begins to move and so creates again adverse excessive compressive forces or increased times of compression of the structures, which eventually lead to overload, degeneration and pain.

External or Internal

Finally this tells us where the tendon has experienced insult on its top or underneath surface.
External impingement or bursal side tendon damage is done on the top of the tendon and can be caused by the factors we have described already, primary bony factors or secondary biomechanical issues and so you go and address these accordingly.

normal MRI

Internal impingement or articular surface tendon damage however is a completely different beast and a completely different mechanism of impingement that is usually seen in sports that involve throwing or repetitive over head actions such as tennis serves. It causes damage to the underside of the tendons usually the supraspinatus/infraspinatus junction from a loose or lax shoulder socket that allows a true mechanical pinch of the tendon in the shoulder socket, this normally causes shoulder pain down the back of the shoulder and arm, and so considered a different condition completely.
It is normally detected by placing the shoulder in full abduction and external rotation, which hurts and then relieved with downward pressure over the front of the shoulder, just like an apprehension test for a lax shoulder socket. They may also demonstrate instability issues around the shoulder such as a positive sulcus or load/shift tests.
So you can see there are some serious challenges in trying to identify any specific causes of shoulder pain from adverse sub acromial arch compression, and actually there is now research and thoughts that sub acromial compression may not be as significant or as common as we first thought with a lot of people with so called shoulder impingement symptoms, but that’s a different article.
In summary, I hope you can see that sub acromial compression issues cover a vast array of conditions and have multiple causes.
Better identification of factors that are unlikely to be solved with conservative treatment early on saves a lot of time and discomfort for your patient, and clinical testing needs to be done thoroughly and a full detailed history are paramount to help give indications and clues as to the most likely structure causing shoulder pain.
Finally I think shoulder or sub acromial ‘impingement’ is a rubbish, misleading, inaccurate and potentially pain inducing term, and I suggest we all start to use the term ‘sub acromial pain’ not impingement…
As always many thanks for reading!
Enjoy your sport and happy exercising



  1. This is a great blog!
    As a physiotherapist currently studying MSc manual therapy I have chosen my blogging assessment subject area as the big world of shoulder ‘impingement’! As you mention there is some great Jeremy Lewis work out there and I look forward to looking into this interesting subject area more deeply! Especially the debate about the diagnostic term of impingement and the evidence surrounding best clinical practice / treatments. Please follow me if you get the chance 🙂

  2. Hi Adam,
    I love reading your blogs and find them extremely interesting.
    You mention here about psychosomatic processes when it comes to shoulder impingement – sorry I know your not a fan of that term! I was wondering whether you find it a common occurrence to come across patients with a psychological element to their shoulder injuries and treatment/rehab? Also, at the risk of sounding a little cheeky, whether you had any tips on helping patients overcome this when the very basic of movements are too painful/fearful for the patient?
    Again, love the blogs, look forward to reading more!

    • Hi Nicole
      Thanks for your kind comments, yes is the simple answer to your question, firstly there are central nervous implications in shoulder tendinopathy pain as discussed in this paper by Chris Littlewood and also there is psychological considerations with all types of pain and more so with unstable shoulder, especially the type 3 group, which I talk about in my unstable shoulder series
      Hope that helps

  3. Great! Thanks for sharing this, after reading this I would’ve never thought that I ‘m already suffering to shoulder impingement. Now, I’ll be doing some research on how to relieve this pain I’m suffering. Thanks a lot.

Related news