The shoulder is an inherently loose articulation permitting a wide range of motion. The glenohumeral capsule remains lax throughout mid-range motion where stability is conferred by the dynamic action of the rotator cuff and the conforming articulation of the humeral head and the glenolabral surface. The capsuloligamentous structures about the glenohumeral joint act as passive restraints when capsular tension develops at the extremes of motion and contributes to glenohumeral stability. Each portion of the joint capsule performs its primary “checkrein” responsibility at different glenohumeral positions. The rotator interval develops tension in external rotation with the arm adducted. The middle glenohumeral ligament restricts motion with the arm in external rotation in the midrange of abduction, while the anterior band of the inferior glenohumeral ligament stabilizes the glenohumeral joint in full abduction and external rotation. Tension is developed in the posterior-superior capsule with the arm adducted and internally rotated. This tension shifts posterior-inferior with abduction and internal rotation. Contracture of any given portion of the glenohumeral capsule will cause reduction in glenohumeral motion as though the end range of motion were reached prematurely. In addition, some capsular contractures cause an obligate and abnormal translation of the humeral head during motion (7). The abnormal joint mechanics associated with posterior capsular contracture have been observed to result in superior migration, which can be associated with subacromial impingent or superior labral tearing (8,9). Chronic anterior asymmetric tightness may cause posterior translation with increased posterior glenohumeral joint reaction forces resulting in posterior glenoid wear (10). Normal shoulder range of motion requires a smooth glenohumeral articulation with a surrounding capsule of normal volume and compliance, unrestricted gliding of the rotator cuff tendons under the coraco-acromial arch and under the deep surface of the deltoid muscle, and normal translation of the scapula over the chest wall. Adhesions or contracture in any of these locations may result in shoulder stiffness.

At this time there is no accepted complete understanding of the pathophysiological basis for the development of adhesive capsulitis. Harryman and Lazarus (11) credit Reidel’s report
in the early German literature as the first to propose that a pathologic process originating in the glenohumeral capsule was responsible for shoulder stiffness. Neviaser (5) subsequently described the histologic findings of perivascular infiltration, capsular fibrosis, and capsular thickening associated with the clinical development of adhesive capsulitis. These findings were later confirmed by Lundberg (12), who noted capsular changes including an increase in the density of collagen and a glycosaminoglycan pattern similar to that found in repair tissue. Omari and Bunker (13) studied biopsies of rotator interval tissue in patients with frozen shoulder and described a dense matrix of Type 3 collagen populated with fibroblasts and myoblasts. Some authors have hypothesized that autoimmune processes might be involved in the development of adhesive capsulitis. Studies showing reduced levels of IgA as well as an increased incidence of the immunohistocompatability antigen HLA-B27 initially suggested this relationship; however, subsequent reports have failed to establish an autoimmune etiology or identify an immunologic test to diagnose this condition. The possibility of a relationship between Dupuytrens contracture and adhesive capsulitis has intrigued investigators for many years. Many of the histologic characteristics of Dupuytrens disease have been observed in capsuloligamentous biopsies obtained from patients with adhesive capsulitis. Investigators have suggested these two conditions may share a common biochemical pathway that leads to contracture (14,15). More recently, investigations into the activity of polypeptide growth factors and matrix metalloproteinases have suggested a possible contribution to the development of adhesive capsulitis through their effects upon the activation and migration of fibroblasts and their alteration of normal collagen matrix remodeling (16).

Adhesive capsulitis is commonly observed in patients during their fifth and sixth decades of life with an incidence of 2% to 5% in this age group, and a greater frequency in women than men. There is an increased occurrence in patients having suffered closed head injuries (as high as 25%), Parkinson disease (13%), autonomic disorders of the upper extremity (such as post-stroke shoulder-hand syndrome with an incidence of approximately 25%), and diabetes mellitus (with an incidence varying from 10% to 35%) (11,17,18,19,20). Prior authors have suggested an association between adhesive capsulitis and other medical disorders including thyroid disease, cardio-pulmonary disease, cervical spine degenerative disease, neoplasia in general (and more specifically in the lung), and the presence of personality disorders such as depression and anxiety disorder; however, no causal relationship has been clearly established for these entities.

Adhesive capsulitis is a disease that typically progresses through three clinical phases. Patients with this condition will have varying clinical complaints depending upon the phase they are experiencing at the time of interview. The initial inflammatory “painful” phase begins with a spontaneous onset of aching discomfort at rest and the development of pain with use. Pain at night that interferes with sleep commonly occurs and early limitation of motion begins to develop. The arm may be held in an adducted and internally rotated position to reduce tension in the inflamed glenohumeral capsule. Occasionally the patient attributes the onset of their problem to a trivial trauma, but more often no inciting injury can be recalled. This phase may last as long as 9 months according to Reeves (21).

The proliferative or “freezing” phase occurs next and is characterized by progressive and global loss of motion. Range of motion often becomes quite restricted during this portion of the disease process, and it is common for shoulder function to be dramatically reduced by the contracture present. Night pain usually continues, especially when lying on the involved side. Pain with the arm at rest may begin to decrease, although pain with use persists when the patient reaches the end of his or her available motion and the capsule is placed under increased tension. This phase is often described as lasting between 3 to 12 months; however, we have observed it to last significantly longer in numerous patients.

The third and final “thawing” phase is characterized by resolution of the painful contracture. The process is often slow and may be punctuated by periods where recovery seems to plateau before resuming its course of progress. Improvement may extend over a period of time ranging from 3 months to 3 years, but it can be prolonged up to 6 or 7 years. Even after the discomfort resolves and acceptable function returns, limitation in final range of motion is commonly present.

The physical examination of patients with adhesive capsulitis reveals a global reduction in range of motion with a marked decrease in glenohumeral translation also present. Examination of the opposite shoulder (if normal) is performed to identify the patient’s expected normal range of motion for comparison. Evaluation for limitation of pure glenohumeral motion (best measured in the supine position with the scapula immobilized) is often more demonstrative of the extent of contracture present than the measurement of total shoulder girdle range of motion (glenohumeral plus scapulothoracic motion). The latter, however, is more closely linked to the patient’s clinical perception of their ability to function, so frequently both types of shoulder motion are measured and followed. Patients with adhesive capsulitis will demonstrate at least a 20% reduction in range of motion, and findings of 50% or greater loss of motion are not uncommon.

Some degree of weakness is often noted when examining for shoulder girdle strength in patients with adhesive capsulitis. The magnitude of this finding may be misleading
however if the patient is experiencing an inflammatory component to their disease. When this is present, strength testing can produce substantial pain and result in a limited resistive effort. To obtain the most representative assessment of the patient’s true shoulder strength, resistive strength testing should be performed within the patient’s comfortable arc of motion, often testing elevation strength at approximately 30 to 45 degrees of elevation, and rotational strength with the arm at the side.

Pain is often reported to be present diffusely throughout the shoulder girdle; however tenderness with palpation is often greatest over the anterior subacromial bursa, the proximal biceps tendon, the rotator interval area and the anterior capsule. The posterior capsule, the lateral subdeltoid recess, and rotator cuff area often have less tenderness with the acromioclavicular joint often spared. Depending upon the phase of the disease process tenderness can be quite severe, so palpation is often best performed at the end of the exam to avoid patient guarding while examining for range of motion and strength.

The evaluation of a patient with adhesive capsulitis is not complete without an appropriate series of plain radiographs. True glenohumeral anterior-posterior views, along with axillary, scapular outlet and acromioclavicular views are considered necessary to exclude other shoulder girdle conditions which result in pain and stiffness. These films may often reveal osteopenia, but should not show any other definitive pathology. Additional radiographs of the neck, chest, or arm should be obtained if clinically indicated to exclude such problems as cervical radiculopathy, lung cancer, or humeral bone tumor.

Other types of advanced imaging have been used in patients with adhesive capsulitis. Magnetic resonance imaging (MRI) may demonstrate thickening of the inferior capsule, and when performed with intravenous gadolinium may reveal enhancement in the capsule or synovium (22). MRI arthrography and standard arthrography can show decreased intra-articular volume and will commonly reveal a reduction in the size of the inferior capsular recess. Additional findings can include variable distention of the biceps sheath or the subscapularis recess. Dynamic ultrasound has been shown to display a reduction in supraspinatus excursion with attempted shoulder movement (23). Radionucleide scanning has demonstrated increased uptake of technetium on “posterior views” in frozen shoulder (24), versus increased anterior uptake in subacromial conditions and uptake involving the distal upper extremity in patients with reflex sympathetic dystrophy. None of these advanced tests, however, have been shown to be diagnostic of adhesive capsulitis and often are unnecessary unless needed to exclude other diagnoses.

A decision regarding whether and when to provide treatment for a medical condition is most often based on a complete understanding of the natural history of that disease. Unfortunately, the natural history of adhesive capsulitis is poorly understood. Although reports on this disease and its treatment have been numerous, clinical observations have varied widely with regards to the degree of disability and the duration of symptoms encountered. The response to treatment has also varied widely due to the employment of many different treatment regimens with confounding variables and conflicting results. Historically, idiopathic adhesive capsulitis has been referred to as a self-limited condition that usually resolves in 1 to 3 years regardless of treatment (19,21,25,26). Numerous studies, however, have described symptoms and disability that persist in 20% to 50% of patients for as long as 7 to 10 years (21,27). For this reason, the treatment approach of benign neglect often accepted in the past seems inappropriate, and intervention to reduce its impact upon the lives of its victims is clearly appropriate.

Nonoperative treatment commonly begins with measures to reduce shoulder pain and inflammation including topical treatments of heat, ice, transcutaneous electric stimulation, transcutaneous salves and balms, acupuncture, massage and systemic medications [nonsteriodal anti-inflammatory drug (NSAID) class medicines and oral corticosteroids]. The relative contribution to recovery that any of these treatment measures provide remains unclear and their use at this time is best individualized based upon the response each individual patient manifests during their use. Many other treatment regimens including trigger point injections, suprascapular nerve block, ultrasound, and the injection of hyaluronic acid have been studied and found to have no demonstrable long term benefit. Intra-articular injections of corticosteroid medications are often used to relieve the symptoms of painful inflammation and some believe they may alter the inflammatory process within the glenohumeral capsule. Hannafin and Chiaia (28) have observed that the use of intra-articular steroids coupled with range of motion therapy can result in a “rapid and striking improvement” if administered early in Stage 1 adhesive capsulitis and can even provide “significant improvement” during Stage 2 disease. Other investigators have found no long term improvement in outcome when these injections are used (29).The mainstay of nonoperative treatment for adhesive capsulitis is the administration of range of motion stretching therapy. This may be accomplished by the patient on their own at home; it can be administered by a physical therapist, or most commonly as a combination of both. Review of numerous articles in the orthopaedic and physical therapy literature reveal the reported success rate for formal physical therapy to average between 50% to 70% of patients. Miller et al. (19) have reported their results with a patient administered home therapy program directly supervised by the treating orthopaedist, which they call “Orthotherapy.” At an average of 14 months treatment, they observed that 100% of patients
were improved, having painless range of motion within 20% of normal or having no restriction during activities of daily living. Other researchers, however, have had less success with nonoperative treatment regimens. Griggs et al. (30) reported that with an average 22 months of follow-up, 50% of their patients had abnormal function, 37% continued to have pain, and 10% were not satisfied. Shaffer et al. (27) reported results after an average follow-up of 7 years and found 60% of patients had a persisting deficit in range of motion measuring between 10% to 30% of normal, 50% of their patients had continuing pain or a sense of stiffness and 11% felt they experienced continuing restriction in function.