Frozen shoulder: a sticky problem

In the first of a 2-part article, Andrew Hamilton looks at the etiology of the condition known as ‘frozen shoulder,’ and the recent thinking on conservative treatment options for restoring shoulder functionality

Frozen shoulder, more correctly known as adhesive capsulitis (AC), is a common cause of shoulder pain and loss of motion in the over 40s. In ‘primary’ cases of AC, there is no significant reason for the onset of pain or stiffness, whereas secondary AC can occur following an injury, trauma, surgery, or illness. In younger people below the age of 40, primary AC is rare, however, and is more likely to be secondary – for example following a rotator cuff injury. When it does occur, however, it can result in considerable disruption to any sports activity, especially given the typically long-timescale it requires to resolve.

Causes of AC

As its terminology implies, adhesive capsulitis describes a condition where the joint capsule becomes inflamed and ‘sticky’ (see figure 1), making the whole joint stiff and difficult/painful to move. The etiology of AC is poorly understood. We do know, however, that as the capsule becomes inflamed, scar tissue forms, which causes pain and leaves less room for the humerus to move through its normal range of motion.

Histological evidence from patients with AC suggests a fibro-proliferative mechanism for the condition(1), while on the macro-scale, there is thickening and contracture of the anterior capsule – especially the coracohumeral ligament and the middle glenohumeral ligament – leading to reduced glenohumeral joint volume, and shoulder ROM (particularly external rotation in neutral and mid-elevation(2)). Other studies have suggested a link between AC risk and some metabolic conditions such as diabetes and raised blood lipids(3,4).

From an athlete’s perspective, there’s some evidence that shortening of one or more of the ligaments of the shoulder (eg, through poor posture) may also increase the risk of AC; athletes whose training results in overdeveloped and/or shortened pectoral and anterior deltoid musculature may, therefore, be at increased risk(5). Also, prolonged immobility (for example after a rotator cuff injury or shoulder fracture) is known to be a risk factor for AC – a good reason for athletes to rehab a shoulder injury as rapidly as possible! Despite all of these risk factors however, many primary cases of AC remain completely idiopathic in nature, something that can be very frustrating for clinicians and sufferers alike.

Figure 1: Schematic representation of AC


The joint capsule shortens as inflammation progresses, restricting the range of motion of the humerus.

AC characteristics

AC typically displays a progression through three distinct phases: freezing, frozen, and thawing (see figure 2):

  1. In the initial phase (freezing), patients typically present with marked pain that comes on over a period of a few weeks with activity. At this point, there is no noticeable loss of range of movement.
  2. The frozen (adhesive) phase typically lasts for 3-9 months, with significant stiffness and pain at the extremes of movement. Patients presenting in this phase often hold the arm in adduction and internal rotation. There is usually a global (and painful) restriction of shoulder movement, with almost complete loss of external rotation.
  3. The thawing (resolution) phase typically lasts for 9-18 months and is characterized by steadily diminishing levels of pain and stiffness. However, the improvements are only slow, and it may significantly longer for complete thawing to occur, with some studies suggesting a timeframe of up to 3-10 years(6-9).

Figure 2: Schematic representation of typical time course and symptom severity in AC


AC can be challenging to diagnose in its early stages and to differentiate from other common shoulder disorders. In addition to the clinical features described above (in particular the almost complete loss of external rotation in passive movement), the typical pattern of AC onset is also helpful in making a diagnosis. Caution is needed, however; while the consensus view is that a diagnosis of AC can usually be confirmed in the clinic(10), some research has demonstrated that many of the clinical identifiers for early-stage AC previously proposed are not validated by evidence(11).

In cases where doubt remains, MRI imaging is desirable; a thickening in the joint capsule and the affected ligaments will provide strong evidence for AC(12,13), and very recent evidence demonstrates MRI can detect distinct morphological changes associated with the phases of AC(14). Routine X-rays for suspected frozen shoulder are not recommended. Among other studies, a recent comprehensive review study has concluded that X-ray imaging offers little over a diagnosis based on history and clinical examination alone(15).

Treatment options

The use of non-steroidal anti-inflammatory medication (NSAIDs) can help provide some symptomatic relief, but there’s little evidence it ameliorates the disease progression. Physiotherapy may be somewhat more successful; studies examining the role of physiotherapy in the early and mid-stage of AC have shown an improvement in pain scores, functionality, and range of motion(16-22). When conservative treatment options are unsuccessful, surgery or injections remain options, particularly in the second and third phases of AC. The most common approaches are manipulation under anesthesia (MUA), arthroscopic capsular release, and capsular distension injections(23).

Physio vs. manual therapy vs. injections

Because of the idiopathic nature of AC, treatment options remain equivocal. For example, some research suggests that patients receiving physiotherapy alone have better clinical outcomes than patients undergoing MUA(17), whereas other research suggests that ‘supervised neglect’ may provide better results at two years than an intensive physiotherapy regime(24). The efficacy of MUA is also far from clear; a study published earlier this year concluded:

“The efficacy of MUA in the treatment of frozen shoulder cannot be drawn from the current literature(25).”

A 2014 review study investigated the efficacy of manual therapy and exercise – in particular, how it compared to glucocorticoid injections(26). Analyzing data from 32 randomized controlled trials (RCTs) and quasi-randomized trials (a total of 1836 patients), it concluded that: “the best available data show that a combination of manual therapy and exercise may not be as effective as glucocorticoid injection in the short-term.” These findings fit with a previous meta-review, which evaluated(27):

• Steroid injection
• Sodium hyaluronate injection
• Supervised neglect
• Physical therapy (mainly physiotherapy)
• Acupuncture
• Distension and capsular release

The authors concluded that there was limited clinical evidence on the effectiveness of treatments for primary frozen shoulder – although, in terms of cost-effectiveness, steroid injections alone may be more cost-effective than steroid plus physiotherapy or physiotherapy alone.

Another injection therapy that may be of value in AC (in the short term at least) is the use of the non-steroidal anti-inflammatory ‘Ketorolac.’ In a recent randomized controlled trial, on 160 patients, Ketorolac and hyaluronic acid injection therapy were compared(28). Patients were split into two groups of 80, and the pain scores for both groups were recorded (using UCLA shoulder rating scale) before treatment and then again at follow up after four weeks. The results showed that mean pain decrease was significantly better in the NSAID group (26.7) compared to the hyaluronic acid group (21.7).

Stretching success

One conservative approach that may be of value is the use of stretching. A study by Indian researchers looked at the effectiveness of sustained stretching of the inferior capsule in the management of a frozen shoulder – specifically the efficacy of a shoulder counter traction apparatus on the range of movement, pain, and function in patients with a frozen shoulder(29). This study took 100 patients and compared 20 minutes a day for five days per week for two weeks of:

• Physiotherapy only
• Countertraction and physiotherapy

When counter traction was given in addition to physiotherapy, the scores for shoulder flexion improved from 94.1° at baseline to 161.9° after the intervention (see figure 3). Abduction range of movement increased from 90.4° to 154.8° after the treatment, while pain decreased from a score of 8.00 to 3.48. Overall, 60% of the participants were improved to the fourth stage of satisfactory joint function (according to the Oxford Shoulder Score) in the counter traction group compared with just 18% in the physio-only group. Whether or not these improvements remained over an extended period was not investigated, but the initial results were encouraging nevertheless.

Figure 3: Counter traction treatment and improvements in the range of motion(29)

Blue bars show before and after goniometer scores for shoulder flexion, while green bars show before/after scores for abduction.

Another (very recent) study reviewed the efficacy of Mulligan’s mobilization with movement (MWM) techniques on peripheral joint pathology, including AC(30). (see link at video 1).

Using only randomized, controlled high quality or medium quality trials, the researchers analyzed 18 studies with 753 participants in ten separate meta-analyses for the effects of this technique on the range of movement (ROM). They concluded that “MWM seems to produce better therapeutic results in comparison to sham, passive, other activities, or no therapeutic approach, regarding the improvement of joint ROM in specific peripheral joint pathologies, and consistently in all movement directions for shoulder adhesive capsulitis.” However, they also cautioned that due to the relatively small number of studies covering each pathology, more research, with longer follow-up periods is needed to validate this approach further.

Video 1: The concept and application of a Mulligan’s mobilization for AC


In summary

AC presents considerable challenges for the clinician. It is idiopathic in origin and takes a considerable amount of time to resolve (even in the best-case scenario). There remains much uncertainty about effective conservative treatment approaches. Although physiotherapy and manual therapy remain the first treatment strategy, the evidence that they speed healing in the longer term is weak. The proof that injection therapy is somewhat more robust, however, and more recent evidence suggests that stretching techniques involving counter traction and Mulligan’s mobilizations could be useful tools in the clinician’s armory. In part two of this article, we will look at the evidence for and against more technological treatment approaches in AC, including pulsed radiofrequency therapy and guided ultrasound.

1. J Bone Joint Surg Br. 1995;77(5 ):677
2. J Bone Joint Surg Am . 1989 ;1(10 ):1511
3. South Med J. 2008 Jun;101(6):591-5
4. Rheumatol Int. 2014 Jan;34(1):67-74
5. Open Orthopaedics Journal, 2013, 7, (Suppl 3: M10) 352-355
6. Br Med J 2005;331(7530): 1453-6
7. Curr Rev Musculoskelet Med 2008; 1(3): 180-9
8. Br J Gen Pract 2007; 57: 662-7
9. Scand J Rheum 1975; 4(4): 193-6
10. Open Orthopaedics J, 2013, 7, (Suppl 3: M10) 352-355
11. Phys Ther. 2014 Jul;94(7):968-76
12. J Bone Joint Surg Br 1995; 77(5): 677
13. J Bone Joint Surg Am 1989 ;71(10): 1511
14. Indian J Radiol Imaging. 2019 Jan-Mar;29(1):19-24
15. Musculoskeletal Care. 2019 Jun;17(2):288-292
16. J Bone Joint Surg Am 2000; 82(10): 1398.
17. Arch Orthop Traum Surg 1995; 114(2) : 87-91.
18. Ann Rheum Dis 1984; 43(3): 353-60.
19. Arthritis Rheum 2003; 48(3): 829-38.
20. Physiotherapy 2002; 88(8): 450-7.
21. Br Med J 1998; 317(7168): 1292-6.
22. Cochrane Database Syst Rev 2003;(2): CD004258
23. Shoulder Elbow 2010; 2(4): 294-300
24. J Shoulder Elbow Surg 2004; 13(5): 499-502
25. EFORT Open Rev. 2019 Mar 19;4(3):98-109
26. Cochrane Database Syst Rev. 2014 Aug 26;8:CD011275
27. Health Technol Assess. 2012;16(11):1-264
28. Pak J Pharm Sci. 2019 May;32(3):953-956
29. Clin Orthop Relat Res. 2014 Jul;472(7):2262-8.
30. J Manipulative Physiol Ther. 2019 Jul 16. pii: S0161-4754(18)30270-7

Share this
Follow us