Disabling stiffness. Severe, lightning bolt-like pain with attempted arm movements. Inability to sleep on the affected side. Absent findings on diagnostic imaging tests. The diagnosis? Adhesive capsulitis. Adhesive capsulitis, otherwise known as frozen shoulder syndrome (FSS), is frequently encountered by physical therapists who treat patients in orthopedic settings. This exquisitely painful shoulder disorder impairs active and passive range of motion, strength, and function to the degree that a patient may be completely unable to use the affected extremity. Sometimes, the disorder predictably comes on after a period of immobilization, but even more frequently, it can sneak up on its victim like a thief in the night.

Icy Hot

The healthy shoulder joint is a freely moving structure that is held in place by the glenohumeral joint capsule and supportive ligaments. It’s controlled by the surrounding rotator cuff and scapulothoracic musculature. The fibrous joint capsule encapsulates the ball and socket of the glenohumeral joint, attaching to the glenoid margin medially and the humeral neck and shaft laterally; the relatively lax fibers of the inferior capsule form the axillary recess and allow for freedom of movement with arm elevation. The joint capsule is supported by the glenohumeral ligaments and the rotator cuff tendons, which blend together with the capsular fibers. Internally, the joint capsule is lined with a synovial membrane, and it contains synovial fluid that helps lubricate the surfaces of the glenohumeral joint. With FSS, the joint capsule contracts, the amount of synovial fluid is diminished, and adhesions form in the synovial folds of the axillary recess, painfully restricting overhead movements of the shoulder.

Although the exact etiology remains unclear, the characteristic shoulder stiffness of FSS appears to begin with inflammation of the synovial lining of the glenohumeral joint capsule, which later undergoes fibrosis as it gradually thickens and contracts, causing restricted active and passive ROM. The initial event that triggers the inflammatory cascade is largely unknown. The contracture of the joint capsule is thought to result from adhesions forming on the surface of the capsule and from cytokine production that causes the proliferation of fibroblasts. Cells harvested from patients with refractory FSS are found to include inflammatory cells (mast cells, T and B cells, and macrophages) and proliferating fibroblasts, indicating a chronic inflammatory response that may be immunomodulated.¹ But capsular adhesions may only be in part responsible for the debilitating stiffness associated with FSS, as many experts contend that as the disorder progresses, soft tissue contractures of subscapularis and the subacromial rotator cuff muscles, damage to the coracohumeral ligament, and inflammation of the subacromial bursa may also be significant factors that contribute to pathologic stiffness.²

In the general population, the incidence of FSS is approximately 2%, of which 70% of affected individuals are female.³ FSS is most commonly diagnosed in individuals between the ages of 40 and 70, particularly in those who have undergone prolonged upper extremity immobilization or minor trauma, such as shoulder impingement syndrome.⁴ Other risk factors for FSS include thyroid disorders, autoimmune diseases, myocardial infarction, stroke, Parkinson’s disease, and cervical disc disease; patients who have type 2 diabetes mellitus are at five times the risk of developing FSS as compared to those without, and those with type 1 diabetes have a 40% lifetime risk of developing FSS.³ Patients who have experienced FSS in one shoulder rarely have a recurrence; however, up to 30% will develop FSS in the contralateral shoulder, perhaps secondary to underlying comorbidities.³ If FSS develops in the contralateral shoulder, it will usually occur within five years of the original event.

There are two main types of FSS: Primary (idiopathic) FSS, in which the condition arises seemingly “out of the blue” with patients unable to remember a specific incident that may have caused the pain and stiffness; and secondary FSS, in which patients can describe a condition or event that precipitated their shoulder dysfunction, such as an upper extremity trauma, surgery, or period of immobilization. Whether FSS is thought to be idiopathic or secondary to a precipitating event, the clinical manifestations are consistent in pattern of restriction, symptoms experienced, and phase. There are three characteristic phases of FSS (Table 1).

From start to finish, the pathology of FSS is generally thought to be self-limiting within approximately 18 to 30 months;⁶ however, many patients are left with significant ROM restrictions without the benefit of therapy. In most cases, therapists and patients agree that therapeutic intervention is desirable to hasten recovery and restore upper extremity function.

Snowed In

When speaking of frozen shoulder syndrome, Ernest Codman, MD, of Codman’s Exercises fame, purportedly commented, “This entity is difficult to define, difficult to treat, and difficult to explain from the point of view of pathology.” In this spirit, the majority of patients with FSS will often relate a frustrating history of progressive shoulder pain and stiffness over several weeks in which they have no recollection of actually injuring the shoulder. The pain is at first noticed primarily at night and with certain activities, and as the pain worsens, stiffness also worsens. Most patients complain of a global, deep aching sensation interspersed with sharp, shooting pains that may occur at rest, with quick or unexpected arm movements, or at end ranges of motion. Pain commonly refers into the deltoid insertion, and it often refers along the long head of the biceps tendon as well. Combined external rotation and abduction (e.g., combing the hair) is often the most painful functional movement, as this motion stretches the contracting anteroinferior glenohumeral joint capsule; conversely, combined extension and internal rotation (e.g., reaching into the back pocket or to fasten a bra) stretches the contracting posterior capsule, which also may be restricted and painful.

The diagnosis of FSS is normally made through history and physical examination, but because FSS can initially masquerade as other shoulder pathologies such as impingement syndrome, rotator cuff tear, inflammatory synovitis, or subacromial bursitis, a thorough workup is necessary. Diagnosis of undifferentiated Phase 1 FSS may require shoulder, neck, or chest X-rays to rule out calcific tendonitis, occult fractures, cervical pathology, and neoplasms (e.g., bone tumor, soft tissue tumor, Pancoast lung tumor); up to 26% of patients with such tumors may have been initially misdiagnosed with FSS.⁷ If idiopathic FSS is present, X-ray findings are typically normal, although long-standing FSS may result in radiographically observable disuse osteopenia.⁸ Blood tests may be ordered if there is a suspicion of infection, inflammatory arthritis, thyroid dysfunction, or diabetes; in one classic study, researchers found that out of 60 patients with idiopathic FSS, 28% had underlying occult diabetes.⁹ Bone scans and contrast MRI tests, though rarely performed for FSS, may reveal increased uptake and activity in the shoulder joint.

Magnetic resonance arthrography is typically not needed to confirm the diagnosis of FSS, but some experts maintain that with FSS, the amount of glenohumeral joint synovial fluid may shrink from a normal volume of 13 milliliters to as little as 5 milliliters.¹⁰ Others have noted that on MR arthrography, patients with FSS may show decreased joint volume but will more consistently demonstrate a significantly thickened coracohumeral ligament (on average 4.1 millimeters thick as compared to 2.7 millimeters in controls), and a thickened joint capsule in the rotator cuff interval (on average 7.1 millimeters thick, as compared to 4.5 millimeters in controls).¹¹

On Frozen Pond

The physical therapy evaluation must include a complete history and physical examination as well as a thorough review of past medical history, medications, functional limitations, and pain patterns. As always, the therapist must be prepared to rule out alternative diagnoses, but the PT must also be prepared to find nothing significant within the history or diagnostic imaging tests to date — this is the norm for most cases of primary, idiopathic FSS.

On physical exam, goniometric measurements of both active and passive ROM should be obtained to establish a baseline; such measurements are reproducible and acceptably reliable.¹² With FSS, both AROM and PROM will be restricted, and there is functional difficulty with reaching overhead. Cyriax’s classic “capsular pattern” of shoulder range of motion restriction (limited abduction, with more limited external rotation and less limited internal rotation) may or may not be present, depending on the patient’s individual characteristics and the anatomical position in which ROM is measured.^13,14 In the early stages of FSS when diagnosis may be difficult, the most consistent finding is decreased external rotation. Thoughtful evaluation of accessory glenohumeral joint mobility is important to assess the quality of end feel and the amount of joint excursion. In Phase 1 FSS, the end feel with PROM may be empty secondary to pain response, whereas in Phase 2 FSS, the end feel is more likely to be firm and rubbery (“capsular”) as joint inflammation subsides during the frozen phase of the disorder.¹⁵ Glenohumeral joint excursion is typically restricted in both anterior-posterior and inferior glide upon accessory mobilization, yet scapular movement may be excessive secondary to inhibited, weak scapular stabilizers and compensatory scapulothoracic hypermobility.

Compensatory shoulder movements will often result in palpable tightness and spasm in the upper trapezius muscle and significant periscapular strain and muscle guarding. There is frequently specific tenderness at the long head of the biceps tendon, as well as at the tendinous insertions of supraspinatus, infraspinatus, and teres minor. Disuse muscle atrophy may be present. Stretching into external rotation with the arm adducted at the patient’s side will usually reveal tightness of the subscapularis muscle as well as contracture of the anterosuperior portion of the glenohumeral joint capsule. Clinical strength testing of shoulder musculature is typically painful upon resistance, as are provocative clinical tests that evaluate impingement or tendinopathy of the soft tissue structures of the shoulder (i.e., Neer impingement test, Hawkins test, Speed maneuver, Empty Can test, and Yergason test, among others) perhaps because with FSS, pain may be generated not only from the joint capsule but also from intra-articular and extra-articular soft tissues.

Thawing the Frozen Shoulder

Treatment options for FSS range from benign neglect, in which the patient receives no intervention other than the passage of time; nonsteroidal anti-inflammatory drugs (NSAIDs), or oral/injected corticosteroids to address joint inflammation; physical therapy treatment to mobilize, stretch, and strengthen the shoulder; closed manipulation under anesthesia (MUA); distension arthrography, in which fluid is injected into the contracted joint to assist with joint expansion; arthroscopic capsular release; and open surgical release. Fortunately, up to 89.5% of patients with FSS will experience significant resolution of their pain and stiffness within an average of 3.8 months (+/- 3.6 months) using nonoperative treatment measures that consist of NSAIDs, physical therapy, and in some cases, one or more concurrent intra-articular corticosteroid injections.¹⁶

The use of NSAIDs is common in the treatment of FSS as a means of pain control in Phase 1 and Phase 2; however, this type of medication is not known to affect the underlying disease process or the duration of symptoms. In some cases, NSAIDs are used in conjunction with muscle relaxants or short-acting analgesic agents prior to a physical therapy session to enable improved compliance with treatment. At times, oral corticosteroids may be added, but intra-articular glenohumeral corticosteroid injections are more effective and offer fewer adverse effects, such as systemic hyperglycemia, than oral medication. In some cases, acupuncture may also be helpful to reduce painful shoulder symptoms while undergoing physical therapy.¹⁷ A recent study showed that a comprehensive approach that included physical therapy treatment consisting of joint mobilization, patient education, stretching exercises, and specific modalities was strongly linked to the fastest short-term improvement over four to six weeks; patients that responded more slowly derived benefits from adding intra-articular corticosteroid injections.¹⁸ When used rapidly within the first weeks of the physical therapy treatment period, the injection of intra-articular corticosteroids appears to hasten pain relief and improve patient response during therapy.¹⁹

With or without the addition of anti-inflammatory and pain-relieving medications, physical therapy is the cornerstone of treatment for FSS. Research is ongoing to identify the most effective combination of physical therapy treatment techniques that yield the greatest benefits for FSS, but it is clear that joint mobilization is an important component, especially when performed at end range positions (as compared to joint mobilization techniques administered within the pain free ranges of the shoulder).²⁰ Studies show that both passive abduction ROM and glenohumeral joint stiffness in the anterior-posterior direction reliably improves following intensive end-range joint mobilization in patients with FSS,²¹ and one biomechanical cadaveric study confirmed that when joint distraction is performed at end-range positions, there is the greatest improvement in glenohumeral joint mobility.²² In another study, Maitland accessory glenohumeral joint mobilization techniques, consisting of anterior-posterior mobilization during shoulder flexion and longitudinal distraction during shoulder abduction, yielded significant improvements in flexion, abduction, internal rotation, and external rotation ROM in patients with FSS, especially when combined with exercise.²³ Glenohumeral joint mobilization techniques performed with an anterior-to-posterior glide appear to be the most effective in increasing external rotation and reducing pain in patients with FSS than those performed in a posterior-to-anterior direction.²⁴ Active forms of shoulder manual therapy, including Mulligan Concept mobilization-with-movement techniques, also seem helpful.²⁵

Regardless of technique, joint mobilization of the painful frozen shoulder must always be applied tactfully and with consideration of the patient’s pain tolerance and phase of freezing. The critical feature is for therapists to get their hands on their patients in order to elicit change and not simply instruct them in independent exercises. In the initial painful “freezing” phase, patients may not be able to tolerate end range glenohumeral mobilization but instead may respond to mid-range oscillatory glides to reduce painful symptoms and enhance ROM. In this phase, rehabilitation must focus on patient education, low-grade joint mobilization to prevent loss of motion, and pain control. During the second, stiff, “frozen” phase of FSS, the PT may be able to progress toward applying end-range mobilization with distraction and frank capsular stretching. During this phase, more active exercises may be added as passive manual therapy techniques become more aggressive. Then, in the third “thawing” phase of FSS, when the patient is stiffness-dominant rather than pain-dominant, joint mobilization may be performed with more vigor in conjunction with the advancement of active exercises to continue stretching and then strengthening the shoulder. At this point, strengthening exercises and progressive activities tailored to the affected individual performed for as little as 4 weeks can result in significant functional gains with active shoulder elevation.²⁶ Throughout the entire rehabilitation process, the patient must perform a daily home exercise program to maximize gains made during therapy.²⁷

Whereas ongoing studies appear to confirm the value of joint mobilization and exercise as a means to improving patient function, flexibility, and pain levels, the use of passive modalities seems to fall short. Unless used briefly to reduce pain and facilitate stretching prior to treatment, passive modalities seem to have little value in improving outcomes for patients with FSS, as examined in the research to date. A study comparing therapeutic ultrasound with sham ultrasound treatment of patients with FSS could describe no benefits in pain or function attributable to the modality.28 In a recent large retrospective study of 2,370 patients with FSS, researchers found that the application of ultrasound, phonophoresis, iontophoresis, and massage actually reduced the likelihood that patients would experience improved physical function and decreased pain.²⁹

Deep Freeze

Despite the fact that the majority of patients with FSS respond very well to conservative, nonoperative treatment methods, in some cases the affected shoulder will remain extremely stiff and painful, necessitating further medical intervention. Options include closed manipulation under anesthesia (MUA), distension arthrography, and arthroscopic capsular release. Open surgical release of the joint capsule is rarely considered, since the natural evolution of FSS is that of progressive improvement, no matter how slow.

For quite some time the standard next step of treatment for frozen shoulders failing to rapidly advance in physical therapy has been to proceed to MUA. Especially when performed early (less than nine months after symptom onset), MUA can significantly improve ROM, pain, and function in up to 83% of patients with FSS; patients undergoing late MUA (9 to 40 months after symptom onset) do not fare quite as well.³⁰ Another subgroup of patients who exhibit a lesser degree of success following MUA are those with diabetes.³¹ Post MUA, physical therapy treatment is essential to continue skillfully mobilizing and moving the affected limb into the newly obtained ranges of motion to facilitate recovery.³² A home stretching program performed in addition to formal PT helps maintain progress, and in some cases, repeated positioning or splinting using a low load, long duration force can support the amount of time in which the patient receives end range stretching.³³ One novel study showed that positioning the freshly manipulated extremity in a sustained position of abduction and external rotation by merely tying the arm to the back of the bed was enormously helpful in restoring ROM.³⁴ Either way, the manipulated arm must be moved frequently after MUA to prevent adhesions from re-forming.

Despite the relative popularity of MUA for the recalcitrant shoulder, techniques to distend the contracted joint capsule via intra-articular injection are increasingly gaining ground. In most cases, arthrographic distension is performed in conjunction with the injection of intra-articular corticosteroids. Positive clinical outcomes for intra-articular corticosteroid injection with saline distension appear to be similar to MUA at two-year follow-up, without the concordant risks inherent to manipulation, such as humeral fracture.³⁵ One head-to-head study comparing patients with FSS undergoing MUA against those having arthrographic distension, also known as hydrodilation, conceded that although both groups enjoyed successful outcomes by six months post-procedure, 94% of patients with distension were satisfied with their results, as opposed to 81% of those with MUA.³⁶

Some studies show that adding capsular distension to the treatment of patients with FSS undergoing physical therapy results in significantly improved active shoulder movement that was sustained for six months,³⁷ whereas others recorded better pain scores, ROM (flexion, abduction, and external rotation), and function that began during the first week post-injection and persisted at one-year follow-up.³⁸ Even in patients with FSS who had experienced joint contracture for an average of 8.5 months derived significant benefits from arthrographic distension as an adjunct to physical therapy, and researchers feel that this approach may help shorten the recovery timeline of FSS.³⁹ However, some studies maintain that the benefits of arthrographic distension are derived from the addition of the corticosteroids, rather than the distension of the contracted joint capsule itself.⁴⁰ Overall, literature review articles examining the efficacy of arthrographic distension seem to agree that in most cases, the practice of intra-articular injection of the glenohumeral joint with corticosteroids combined with saline results in decreased pain, increased ROM, and improved function in patients with FSS by three weeks.41 Nonetheless, certain patients with severely restricted ROM from FSS may require arthroscopic release for pain relief and return of function.⁴² Regardless, physical therapy is needed to mobilize the painful joint, progress functional exercises for stretching and strengthening, and provide patient encouragement and guidance through the recovery process.

Warming Up

FSS is an extremely painful, debilitating condition that causes global shoulder pain, glenohumeral joint capsule contracture, muscle guarding and soft tissue tenderness, and functional weakness. Sleeping ability, workplace function, and even the most basic activities of daily living become compromised, leaving patients physically and emotionally vulnerable. The cornerstone of treatment for FSS involves physical therapy intervention with an emphasis on joint mobilization and exercise, regardless of whether other types of medical management, such as intra-articular corticosteroid injection, arthrographic capsular distension, or arthroscopic capsule release, are utilized. PT treatment is an essential component to recovery, hastening the return of quality of life and functional ability to the patient with frozen shoulder syndrome.