Frozen Shoulder / Adhesive Capsulitis

In this brochure I wish to answer the following the following questions regarding a frozen shoulder.

  1. What is it?
  2. Why me?
  3. How do you know what I have?
  4. How does it affect me?
  5. What can be done to treat my shoulder pain?

What is it?

In order to proper understand this somewhat confusing condition I will first start by discussing a working definition of the condition of frozen shoulder based on the American shoulder and elbow surgeon’s society.   A frozen shoulder is condition of unknown cause which is characterised by significant stiffness both active and passive of the shoulder joint in the absence of any known underlying shoulder condition.   A frozen shoulder however is an often over used and applied to any person with a stiff and painful shoulder and the term frozen shoulder can often be a “waste can” diagnosis. A true frozen shoulder can be divided into a primary and secondary frozen shoulder.  The primary frozen shoulder or true frozen shoulder is as defined above with painful and stiff shoulder joint in the “absence of known intrinsic shoulder” in other words with no significant history of injury or investigations which can explain the pain and stiffness. A secondary frozen shoulder is shoulder pain and stiffness from any other causes such as shoulder injury for example a fracture.

The shoulder joint is a ball and socket joint not very dissimilar from the hip joint but having a striking difference in being far more mobile.  It has little bony constraint to movement and it is the surrounding cartilage, ligaments and capsule which provides shoulder stability but still allow an enormous range of motion.

Rotator cuff tendons surrounoding the shoulder joint

The capsule containing the ligaments is usually very loose.

Normal capsule

In a frozen shoulder the capsule initially becomes very inflamed which causes a deep ache in the shoulder joint radiating down the arm.


After the development of inflammation the surrounding capsule and ligaments shrink which restricts shoulder motion causing gradual and increasingly severe stiffness. The cause of the inflammation is unknown.  The body perceives the shoulder as being injured and therefore triggers an inflammatory response in order to try to heal it. As the inflammatory cascade occurs, healing tissues including white cells and platelets and subsequently fibroblasts migrate towards the shoulder joint. These migrating tissues attempt to heal the shoulder joint which is perceived by the body as being damaged, this creates a domino effect as more and more inflammatory factors are produced and ongoing inflammation and healing occurs. Like the falling dominos it does not matter what triggers the initial domino falling but subsequent dominoes continue to fall unrelated to the initial insult. Eventually the domino’s run out.The cells responsible for healing are called fibroblasts. typically tissue damage simulates migrations and growth of these fibroblasts and these cells are responsible for the production of scar tissue. Cytokines are the chemicals which trigger fibroblasts activity; they bond to the fibroblast and stimulate scar formation.  Special fibroblasts called myofibroblasts are responsible for shrinking scar tissue.  So when the fibroblasts accumulate in the shoulder joint, scar tissue is laid down and the myofibroblasts shrink that scar tissue. This inflammatory process occurs largely in the lining tissues of the shoulder joint and the capsule with their surrounding ligaments.

As a consequence of this the lying tissues of the joint (the synovium) and capsule with the shoulder ligaments which are normally extremely thin and elastic becomes thickened, hard and inelastic.   

Why Me?

Why me?

The cause of this condition remains poorly understood despite many attempts to explore the cause.  The clinical features of frozen shoulder warrant describing. Incidence The condition of frozen shoulder is very common. 2% of the general population develop a frozen shoulder, while diabetics have a risk of over 10%. Insulin dependent diabetics have a 40% chance of developing a frozen shoulder.   Women are more commonly affected than man, by a ratio of female to male 3 to 2. The most striking clinical feature of frozen shoulder is its predilection for people in their 50’s. Frozen shoulders can occur in people in mid to late 40’s to mid to early 60’s but far commonly are in their 50’s. Women are slightly younger than man on average when they develop this condition with the average age of women developing frozen shoulder is 52 years old and men 55 years old.

Also there appears to be a significant genetic increase risk. Once you have developed a frozen shoulder on one side you are at increased risk at developing it in the other side. 15% of people develop a frozen shoulder on both sides. rarely it can occur simultaneously(both at the same time) but more typically affects the other shoulder within the 5 years after the first shoulder. Surprisingly recurrence in the same shoulder is extremely unusual.

How do you know this is what I have?

As we said in the definition above, in the primary frozen shoulder there is no known underlying disorder.  As x-rays, ultrasounds and even MRI are good at picking up damaged tissues, these investigations are therefore often unfruitful.   There is an old expression “if it looks like a duck and quacks like a duck and walks like a duck, it must be a duck”. So we therefore rely on typical clinical features to suggest the diagnosis of a frozen shoulder.  If however features are not typical, such as these occurring in both shoulders or in an unusually young (less than 45 years old) or unusually old patient (greater than 65 years old), then further investigations are warranted. Blood tests to exclude thyroid disease, raised cholesterol or blood sugars may be of some benefit but will not help confirm the diagnosis. We therefore do not rely exclusively on investigations to confirm the diagnosis. Nevertheless they are of benefit in excluding other conditions such as an underlying arthritis.   The only investigation which can really assist in making the diagnosis is an Arthrogram in which dye is injected into the shoulder joint.  In a normal shoulder joint as we mentioned earlier the surrounding capsule is extremely thin and elastic and one can see on the MRI Arthrogram distension of the capsule.  In the presence of a frozen shoulder no distension occurs, confirming a tight capsule which is diagnostic of a frozen shoulder.

How does it affect me?

Typically the condition starts with often a very mild subtle onset of shoulder pain which becomes increasingly severe over the weeks. Usually after some time of increasing pain some stiffness in the shoulder joint is noted, typically by 6 months after the onset of symptoms the pain is at its worst and movement is also deteriorated to its lowest level. This worsening phase is often described as the freezing phase of a frozen shoulder, at approximately 6 months the condition stabilizes with little change in the pain or movement. This is often referred to as the frozen phase of a frozen shoulder. Over the subsequent 4-6 weeks typically pain gradually starts to settle and some weeks or months after that range of motion also starts to improve. We call this the thawing phase of a frozen shoulder. Gradually symptoms of pain and stiffness resolve. It is described as been like a locomotive, gradually building up speed, reaching its maximum speed and then gradually slowing down. This is very typical of a frozen shoulder where we do not see wild fluctuations in pain and stiffness as we do in other conditions.  In one study by Reeves he shows the typical symptoms occur anywhere between 1-3½ years with an average of 30 months of symptoms over the next 2½ years.  In another study by Hand, looking at the long term outcome of frozen shoulders at 4½ following the onset of symptoms they noticed that 60% of people had essentially regained a normal or near normal shoulder joint with minimal or no pain or stiffness.  Nevertheless even at 4½ 40% of patients still had some degree of ongoing symptoms.

On the whole these were usually quite mild but 5% of patients still had severe pain and stiffness, interestingly they noted that these 5% of patients are usually those with most severe symptoms at clinical onset and these patients had the worst long term outcome.

We described the natural history of the condition as what would occur if there was no intervention.

The natural history of frozen shoulder is extremely variable, some people have only short periods of pain and stiffness, some have a very long period, some have mild to moderate pain and stiffness, and some have extremely severe pain and stiffness.

What can be done to treat my shoulder pain?

Once we understand the natural history of this condition we can determine what can be done about it.  For many patients who symptoms are tolerable, a wait and see approach is quite reasonable. Typically if symptoms are tolerable at the 6 months following onset, no intervention will be required.

Physiotherapy may exacerbate the pain in the first few months following the onset of symptoms in what we call the freezing phase or the inflammatory phase of this condition. Nevertheless in the thawing phase when the inflammation has settled, physiotherapy and stretch exercises might accelerate recovery in range of motion.




Cortisone injections reduce inflammation. For it to be effective in frozen shoulders the injection has to be performed into the glenohumeral joint, not the bursa. They can provide 3-4 weeks of symptomatic relief. As this condition is only temporary and usually at its worse for a short period, this treatment is quite effective and usually only 1 or 2 injections are required to alleviate the worse of the symptoms.




  If the symptoms do not respond well to cortisone injections and are prolonged or intolerable, an arthroscopic capsular release is a key-hole day-surgery operation in which the inflamed and thickened capsule can be released. This allows often dramatic relief in the severe pain and accelerates recovery in the range of motion.



In conclusion once the diagnosis of a frozen shoulder has been made, a wait and see approach can sometimes be taken. If pain is disturbing a cortisone injection can be performed but to be effective in a frozen shoulder this has to be into the shoulder joint, not into the bursa which is the more common injection performed. If this relieves symptoms we can go back to a wait and see approach and repeat the injection if necessary. Nevertheless if this does not provide good relief and symptoms are severe or persistent (greater than 6 months), arthroscopic capsular release is an option which we will discuss with you.


Northern Beaches Physiotherapy and Sports Injury centre have posted  a blog based on my  article which may be of some interest: