Adhesive Capsulitis: Optimal Treatment of 'Frozen Shoulder'
Rick Sandor, MD
THE PHYSICIAN AND SPORTSMEDICINE - VOL 28 - NO. 9 - SEPTEMBER 2000
In Brief: Idiopathic adhesive capsulitis is an enigmatic shoulder disorder that causes pain and reduced range of motion from joint capsule fibrosis. In most cases, the fibrosis resolves and shoulder function returns to normal or near normal. Physicians' major tasks are to differentiate adhesive capsulitis from rotator cuff dysfunction and other causes of shoulder pain and to assist the patient in choosing appropriate treatment. Conservative therapies include rest, analgesia, and range-of-motion exercises. Other treatments include corticosteroid injections, capsular distention, manipulation under anesthesia, and surgical capsular release.
Adhesive capsulitis is a perplexing shoulder disorder that is both well recognized and controversial. The condition was first recognized as being distinct from glenohumeral arthritis at least as early as 1872, and it is defined as a "condition of unknown etiology distinguished by painful restriction of all shoulder movements, both active and passive, characterized by prominent reduction in the glenohumeral range of movement" (1).
Over the years, adhesive capsulitis has had many different names, including shoulder periarthritis, adherent subacromial bursitis, and frozen shoulder. Currently, adhesive capsulitis and frozen shoulder are the preferred terms and may be used interchangeably. The evolution in terminology reflects a somewhat improved understanding of the disorder's pathology, though overall the condition remains largely an enigma.
Adhesive capsulitis has typically been classified into two forms, primary and secondary (table 1) (1,2). In the primary or idiopathic form, no known precipitating event can be identified. The secondary form is associated with or attributable to other illnesses or events. Although in many cases the clinical disorder in the two types appears to be similar, there also can be important differences. For clarity, only the idiopathic form will be discussed.
TABLE 1. Classification of Adhesive Capsulitis (1,2)
The cause of adhesive capsulitis remains unknown. X-rays usually reveal normal joints or show minimal degenerative changes. Clinical signs and symptoms of a systemic inflammatory disorder are absent, and other joints remain unaffected.
Laboratory studies reveal no elevation of sedimentation rate or levels of C-reactive protein and show normal levels of circulating immunoglobulins and complement. Known immunologic components such as rheumatoid factor, antinuclear antibodies, and autoantibodies to smooth muscle, collagen, or cartilage are absent, and the frequency of HLA B27 is not increased among patients with adhesive capsulitis. No known infectious agent has been found. Recurrent hemarthrosis does not seem to be a factor since there is not an increased incidence in patients with coagulopathies (1).
Bone scans show diffuse isotope uptake on the affected side, but other features of reflex sympathetic dystrophy such as skin and vascular changes are absent (3). Some arthroscopic studies describe a proliferative synovitis that is especially evident in the early stages of adhesive capsulitis, but whether the final lesion is intra-articular or extra-articular remains controversial. Early arthroscopic studies described intra-articular adhesions particularly in the axillary fold, but most later studies found anterior capsular thickening, particularly at the coracohumeral ligament and superior glenohumeral ligament. Microscopic evaluation of these surgical specimens shows fibrosis and a decreased number of synovial cells (4). The later studies have for the most part found no intra-articular adhesions (1). Thus, the initiating event in adhesive capsulitis remains unknown, but the end result appears to be fibrotic thickening of the anterior capsule at the rotator interval (1,4).
The clinical profile of patients with idiopathic adhesive capsulitis is fairly well established. The disorder tends to affect women more than men, occurs in the fifth, sixth, and seventh decades of life, does not show a particular preference for handedness, and can on occasion become bilateral (up to 10%) (5,6).
Prevalence. The idiopathic form is uncommon. Good incidence and prevalence data are lacking, but one study estimated a 1-year incidence in patients seen in general practice ranges from 2/1,000 (7), to 2/100 (8). The prevalence in an elderly population has been shown to be less than 1/100 (9).
The secondary form may be more common. Studies show up to 20% of diabetic patients may develop capsulitis (10), as do a similar percentage of neurosurgical inpatients (11).
Disease stages. Three clinical stages of the disease have been described: freezing, frozen, and thawing. The freezing stage lasts from onset to between 10 and 36 weeks and is characterized by the most severe pain and a gradual diminution of articular volume (as assessed on arthrography) (6). The frozen stage lasts between 4 and 12 months. Pain decreases gradually but without appreciable improvement in motion. The thawing phase is marked by gradual return of motion and may be as short as 12 months but may last for years (6).
Natural history. Many patients will have residual signs and symptoms years after the onset of their disease. Mostly this is asymptomatic, mild loss of range of motion. In one study (6), more than half the patients followed for 5 years had mild loss of external rotation and abduction; however, in only 3 of 49 patients did the loss interfere with work or hobbies.
Another study (5) showed similar results in patients followed for up to 3.5 years: 40% of these patients did not regain full shoulder motion. Loss of range of motion, however, was less than 20°, mostly in abduction and external rotation. Many patients were unaware of their restriction and by inference had no functional disability. About 20% of the total had mild residual ache without measurable loss of motion. One patient of 40 had severe, persistent pain and significant loss of motion.
Such studies suggest a much more protracted clinical course than is usually stated for the disorder. Up to half of patients will have mild but detectable loss of external rotation and abduction that persists for years. Most of these patients will be unaware of their motion limits and have no functional disability, but up to 20% might have residual pain without loss of motion. Finally, 5% to 10% will be left with persistent significant pain, loss of motion, and disability.
Differential diagnosis. Adhesive capsulitis has a rather long differential diagnosis (table 2: not shown) (1), but a thorough history and physical examination should yield the correct diagnosis. Key features of the patient history include subacute onset of unilateral shoulder pain with little to no trauma or overuse, a distinct component of night pain, and marked limitation in shoulder movement. Cardiac, pulmonary, or gastrointestinal symptoms suggest referred pain. Inflammatory disorders tend to have bilateral and/or multiple other joint involvement and sometimes physical or systemic symptoms such as distinct morning stiffness, fever, or fatigue. High fever with severe pain suggests the possibility of infection. Major trauma or a distinct history of overuse indicates a possible fracture or rotator cuff pathology, respectively. A gradual onset of pain and dysfunction that worsens over the years and is sometimes associated with shoulder crepitation suggests glenohumeral arthritis. Finally, any history of cancer warrants consideration of metastatic disease.
Physical exam findings. A key physical exam finding in adhesive capsulitis is marked limitation in active and passive range of shoulder motion. All planes of motion seem to be affected. The limit should be equal for both active and passive movements in any one plane. In testing passive motion, the end point is firm but not quite as firm as with that of a bony block.
When assessing joint motion, it is important to measure the motion in the contralateral shoulder for comparison. Though the degree of motion loss can vary widely, typical values show 30% to 60% loss compared with the unaffected side. One way to test glenohumeral motion in both abduction and flexion is for the examiner to fix the scapula with downward pressure with his or her hand. Any subsequent movement will be from the glenohumeral joint.
Manual muscle testing of the rotator cuff muscles should reveal well-preserved muscle strength with little to no pain. One should also examine the neck to rule out any referred neck pathology. Finally, a system-specific exam is indicated if there is any suggestion of a cardiac, pulmonary, gastrointestinal, or inflammatory source of shoulder pain.
Imaging. Imaging evaluation should generally be limited to routine shoulder radiographs to assess the glenohumeral joint and subacromial space. The joints should be normal or show minimal age-related changes. Bone scintigraphy has been used and has revealed nonspecific abnormalities. Magnetic resonance imaging can show capsular thickening (12), but this finding has not been widely substantiated. Arthrograms can detect decreased joint volume. The advanced imaging studies should be reserved for cases in which the diagnosis is unclear and imaging is needed to rule out other sources of pain. Similarly, because adhesive capsulitis does not manifest specific abnormal laboratory findings, any lab workup should be done to assess only other sources of shoulder pain.
Many treatments have been devised to improve outcome in patients with adhesive capsulitis. Therapies include rest, analgesia, and simple range-of-motion exercises; more active home exercises, physical therapy with mobilization; oral corticosteroids; corticosteroid injection; capsular distention; manipulation under anesthesia; and arthroscopic capsular release (table 3) (13).
Rest, analgesia, and range-of-motion exercises. This regimen is the simplest of all treatments and is used when the patient presents earlier in the course of the disease when the shoulder is still quite painful. Rest is prescribed and perhaps even a sling is used. Pain medication is given as needed. Range-of-motion pendulum exercises and table crawls are begun as soon as pain allows (see the Patient Adviser, "Exercising the Frozen Shoulder").
Focused active home exercises. These exercises (figures 2 through 6 in the Patient Adviser) are used when the painful phase has resolved and patients need to regain their motion. (Since patients often first present at this phase, such treatment may be the only treatment needed.) Patients should be cautioned to avoid an overly aggressive exercise regimen, because some physicians believe that too rapid a return to movement and activity can lead to loss of function and increased pain (13).
Physical therapy. Formal physical therapy programs may help those who require a more structured rehabilitation program. Programs incorporate various modalities to decrease pain and employ passive mobilization to improve range of motion. Some believe that overly aggressive mobilization can result in worsening range of motion (13).
Oral prednisone. Oral prednisone has been prescribed by some clinicians on the presumption that adhesive capsulitis is an inflammatory disorder. The results demonstrate some relief of pain but no change in range of motion (14).
Corticosteroid injections. The rationale for corticosteroid injections is the same as for prednisone (15,16). Injection is generally intra-articular, but one study (16) showed equivalent results using subacromial injections. The dosing schedules vary from a single injection to one injection weekly for 6 weeks to one injection every 6 weeks for up to three injections. All corticosteroid injections generally bring moderate relief of pain but minimal change in range of motion.
Capsular distention. The procedure seeks to rupture the contracted capsule with an intra-articular injection of fluid or air (17,18). Capsular distention can provide substantial pain relief in some patients, but pain can be aggravated in less than 10°. Improvement in both abduction and external rotation can be in the 10° to 15° range.
Manipulation under anesthesia. In this procedure, capsular fibrosis is manually ruptured while patients are under general anesthesia or have had an interscalene brachial plexus block (19,20). Manipulation may require intensive pain control and physical therapy afterward. The technique has also been reported to cause humerus fracture and vascular and neurologic injury; another major concern is articular cartilage injury. The incidence of complications, however, is low; therefore, it is a reasonably common treatment in some centers. Reports show good improvement in pain and range of motion. One study (19) showed average improved abduction of about 80% and average improved external rotation of about 50% when the procedure was done within 7 to 8 months of onset.
Surgical capsular release. This procedure is done arthroscopically or via incision. Contracted tissues are directly visualized and released. Reports indicate that patients are 90% pain free and have normal to near normal motion within 3 months after surgery (4,21).
Choosing a treatment. It is difficult to draw reliable conclusions about the efficacy of one treatment versus another based on the current studies. This ambiguity stems from factors that include the design of the studies, the general lack of a control group, the timing of the intervention, and the variable natural history of the disorder.
When choosing treatment for a patient with adhesive capsulitis (figure 1), one always needs to keep in mind the favorable natural history of the disorder. More aggressive intervention should be reserved for patients whose pain and dysfunction are intolerable and who can't wait the required time for resolution (figure 2). The level of intervention should match the needs of the patient and skills of the physician.
After adequate counseling about what are reasonable expectations, most patients should do well with simple analgesia and simple range-of-motion exercises.
A subacromial injection with corticosteroids should be very strongly considered as an adjunct for two reasons. First, rotator cuff dysfunction seems to be a significant source of pain in adhesive capsulitis, and, second, rotator cuff dysfunction without adhesive capsulitis can exactly mimic the physical exam findings of adhesive capsulitis. Except in the nearly painless shoulder, I almost always proceed with a subacromial injection.
Finally, a rare situation is one in which a patient presents with an adhesive capsulitis-like syndrome within the first few days of onset. In these situations, a subacromial injection gives no relief, but an intra-articular injection with corticosteroids can provide dramatic and sustained relief of pain and improved motion. Therefore, these patients need to be managed much more aggressively.
Adhesive capsulitis is a disease of uncertain cause that can lead to prolonged pain and disability. The physician must differentiate a frozen shoulder from other possible causes. Because of the favorable natural history, patients can generally expect a good outcome, and in most cases conservative treatment is successful. Additional therapies with varying levels of risk and invasiveness are available for patients whose dysfunction does not respond to conservative modalities.
Dr Sandor is a medical orthopedist with Camino Medical Group in Sunnyvale, California. Address correspondence to Rick Sandor, MD, 401 Old San Francisco Rd, Sunnyvale, CA 94086.