Millar, N. L. et al. (2022) Frozen shoulder.

Nature Reviews Disease

Primers, 8, 59. (doi: 10.1038/s41572-022-00386-2)

There may be differences between this version and the published version.
You are advised to consult the published version if you wish to cite from it.

http://eprints.gla.ac.uk/280247/

Deposited on 24 October 2022

Enlighten – Research publications by members of the University of Glasgow

http://eprints.gla.ac.uk
 1 Frozen shoulder
2

3 Neal L. Millar1†, Adam Meakins2, Filip Struyf3, Elaine Willmore4, Abigail L. Campbell 5, Paul D.
4 Kirwan6, Moeed Akbar1, Laura Moore7, Jonathan C Ronquillo 8, George A.C. Murrell 9, &
5 Scott A. Rodeo 7
6

8 1 Institute of Infection, Immunity and Inflammation, University of Glasgow, Glasgow, UK

9 2 Department of Trauma and Orthopaedics, West Hertfordshire Hospital Trust, England, UK
10 3 Department of Rehabilitation Sciences and Physiotherapy, University of Antwerp, Belgium
11 4 Physiotherapy Department, Gloucestershire Hospitals NHS Foundation Trust, England, UK
12 5 Department of Orthopaedic Surgery, Kaiser Permanente West Los Angeles, Los Angeles,
13 CA, USA
14 6 School of Medicine, Discipline of Physiotherapy, Trinity College Dublin, Dublin, Ireland
15 7 Department of Orthopaedic Surgery, Sports Medicine and Shoulder Service, Hospital for
16 Special Surgery, New York, NY, USA
17 8 De La Salle University Medical Center, Health Sciences Institute & Asian Hospital and
18 Medical Center, Philippines
19 9 Orthopaedic Research Institute, St George Hospital, University of New South Wales,
20 Sydney, Australia
21

22 email: [email protected]
†

23

1
24 Abstract
25 Frozen shoulder is a common debilitating disorder characterized by shoulder pain and
26 progressive loss of shoulder movement. Frozen shoulder is frequently associated with other
27 systemic conditions or occurs following periods of immobilization, and has a protracted
28 clinical course, which can be frustrating for patients as well as health care professionals. FS
29 is characterised by fibroproliferative tissue fibrosis, whereby fibroblasts, producing
30 predominantly type I and type III collagen, transform into myofibroblasts (a smooth muscle
31 phenotype), which are accompanied by inflammation, neoangiogenesis and neoinnervation,
32 resulting in shoulder capsular fibrotic contractures and the associated clinical stiffness.
33 Diagnosis is heavily based on physical examination and can be difficult depending on the
34 stage of disease or if concomitant shoulder pathology is present. Management consists of
35 physiotherapy, therapeutic modalities such as steroid injections, anti-inflammatory
36 medications, hydrodilation and surgical interventions; however, their effectiveness remains
37 unclear. Facilitating translational science should aid in development of novel therapies to
38 improve outcomes among individuals with this debilitating condition.

2
39 [H1] Introduction
40 Frozen shoulder1, also known as adhesive capsulitis, is a common shoulder disorder
41 manifesting as pain and progressive loss of shoulder movement. FS can be either primary or
42 secondary, which refers to whether the condition has come on spontaneously, with no known
43 cause or trauma (primary FS), or whether it is associated with trauma, surgery or other
44 pathology, such as subacromial pain (secondary FS). FS typically progresses through three
45 overlapping stages, with the predominate symptoms of pain and loss of motion (stage I:
46 inflammation/’freezing’), stiffness (stage II: ‘frozen’), and then resolution of symptoms (stage
47 III: ‘thawing’). However, this classification remains contentious, as many patients still
48 experience symptoms and functional restrictions long after this period.
49 FS is characterised by fibroproliferative tissue fibrosis (Figure 1) of the shoulder
50 capsule, which is thought to be modulated by mediators that include cytokines, growth
51 factors, and enzymes, in particular, matrix metalloproteinases (MMPs), with increasing
52 evidence of the involvement of inflammatory mediators and various immune cells. The
53 histological characteristic of FS is a matrix of type I and type III collagen containing
54 fibroblasts and myofibroblasts, resulting in an imbalance between tissue extracellular matrix
55 (ECM) degradation, remodelling and regeneration.Although knowledge of risk factors of FS,
56 pathophysiology, and enhanced treatments are still emerging, both basic and clinical
57 research (and consequently therapeutic advances) lag behind that in other musculoskeletal
58 conditions, such as inflammatory arthritis and osteoarthritis.
59 A true evidence-based model for the management of FS has yet to be defined, with a
60 wide spectrum of treatments available. Management varies according to the stage of the
61 disease and range from early pharmacotherapy and associated physiotherapy versus later
62 approaches such as surgery (manipulation under anaesthesia (MUA) and arthroscopic
63 capsular release (ACR)), extracorporeal shockwave therapy, hydrodilation, injections
64 (sodium hyaluronate injection, collagenase treatment, and experimental approaches that
65 require validation in clinical trials. FS therefore remains a challenge to treat, with a large
66 proportion of patients still failing to attain complete resolution of symptomology. Indeed, while
67 FS is often regarded as a self-limiting disease (1–2 year recovery), various studies have
68 shown that many of the symptoms associated with FS, such as stiffness and pain, persist in
69 20–50% of patients2-4. Thus, further work is required to identify more effective treatment
70 options for these patients. This Primer presents the current knowledge of the basic and
71 clinical science of FS and highlights its clinical presentation, natural history, risk factors,
72 pathoanatomy and pathogenesis. Furthermore, we provide evidence-based treatment
73 guidelines in the form of a proposed treatment algorithm. In addition, we aim to consolidate
74 and interpret the unmet needs in the field and discuss the barriers that need to be overcome
75 to attain better outcomes for all patients with FS.

3
 76

77 [H1] Epidemiology

78 [H2] Prevalence
79 The lifetime prevalence of FS is estimated to be 2–5% of the general population, and
80 FS affects ~8% of men and ~10% of women5,6. FS is most common in the fifth and sixth
81 decades of life, with the peak age in the mid-50s7. In up to 17% of patients with FS, the other
82 shoulder becomes affected within five years4,8.
83 It is debatable whether FS as a condition is truly unique to the shoulder. Indeed, there
84 are case reports of occurrences of adhesive capsulitis in the knee, hip and ankle9,10, although
85 they are exceptionally rare. Contractures and fibrosis do frequently occur in the knee and
86 elbow, although without the potential for the spontaneous resolution seen in the shoulder83,84.
87 [H2] Risk factors
88 FS has been linked to a range of comorbidities, including cardiovascular disease11,
89 Parkinson disease, stroke12, hyperthyroidism and, in particular, diabetes mellitus, where the
90 incidence of FS can reach close to 60%13-16. FS has also been linked to hypothyroidism17,
91 hyperlipidaemia18 and autoimmune diseases19. These comorbidities are found in more than
92 80% of individuals diagnosed with FS, with over 35% of affected individuals having more
93 than three associated conditions13. Other risk factors (Box 1) associated with FS are
94 smoking20, obesity7 and low levels of physical activity21. In addition, FS risk is increased in
95 individuals with Dupuytren’s disease, a fibrotic disorder of the palmar fascia that has a very
96 similar pathophysiology to FS22-24. In addition to an association with metabolic and hormonal
97 changes, FS has also been associated with abnormal shoulder mechanics and nerve
98 dysfunction. This link between primary nerve dysfunction and FS was first proposed in 1959
99 by Thompson and Kopell25, who proposed that reduced glenohumeral motion could result in
100 exacerbated scapulothoracic motion, thereby stretching the suprascapular nerve, leading to
101 a cycle of pain and shoulder dysfunction. Since then, FS has been identified in patients with
102 a variety of primary neurological conditions. FS is a cause of shoulder pain and dysfunction
103 in patients after radical neck dissection26, acute cerebrovascular aneurysm surger and
104 subarachnoid haemorrhage27 and in individuals with Parkinson disease28. Furthermore, FS,
105 as identified by shoulder capsule volume on arthrography, is the leading cause of hemiplegic
106 shoulder pain after stroke29.
107
108
109

110 [H1] Mechanisms/pathophysiology

111 [H2] From homeostasis to disease

4
112 The shoulder joint capsule is a lax fibrous sheath that encloses the joint. The healthy
113 capsule is collagenous in structure, composed primarily of dense type I collagen and elastic
114 fibre bundles with limited vessels and nerve fibres. The main cell type within this membrane
115 are fibroblasts, which maintain capsule health by producing ECM proteins that provide a
116 supportive yet flexible structure.
117 In FS, the typical collagen structure is disrupted by gradual fibrosis of this connective
118 tissue membrane and thickening of the adjacent synovial membrane30. These fibrotic changes
119 are accompanied by inflammation, neoangiogenesis and neoinnervation . The
31,32,33,34

120 consequence is a reduced joint volume and increased stiffness of the capsule, causing
121 restricted movement and pain. In the following sections we describe how the shoulder capsule
122 and associated structures progress from lax fibrous membrane to a fibrotic hypervascular
123 structure that drives the clinical course of FS.
124

125 [H2] Stages of FS

126 FS progresses through three characteristic stages,1 each with associated arthroscopic
127 and histological changes.6 Neviaser et al. initially described four stages of disease (stage I–
128 IV) in 19871, which was modified in 2010 to three clinically-based stages (stage I–III)35 (Figure
129 2). Stage I is characterized by pain without appreciable limitation in motion, and is associated
130 with an inflammatory synovial reaction on arthroscopy, and with hypervascular synovitis with
131 rare inflammatory cell infiltrates and normal capsular tissue on biopsy. Clinically, stage II
132 involves ongoing pain with progressive limitation in motion. Intra-articularly, there is ongoing
133 synovitis and progressive capsular contracture. On arthroscopy, there is hypervascular
134 synovitis and loss of axillary folding. Histology shows hypertrophic, hypervascular synovitis
135 now with perivascular and subsynovial scar formation. Stage III is marked by ongoing stiffness
136 clinically, and is associated with loss of the axillary recess, fibrosis, and minimal synovitis on
137 arthroscopy. Biopsy of patients with stage III FS reveals dense, hypercellular collagenous
138 tissue to mature fibrosis with a thin synovial layer, similar to other fibrosing conditions.
139

140 [H2] Inflammation

141 Recent years have seen the musculoskeletal scientific community direct its attention
142 to investigating the mechanisms underlying the inflammatory and fibrotic changes associated
143 with FS to elucidate the aetiological, cellular and molecular pathways. Although a single
144 unifying cause is yet to be identified, several key mechanisms have been implicated in the
145 pathogenesis of FS. One of these is chronic, unresolved inflammation. Histological analyses
146 of tissue biopsy samples from affected patients consistently reveal chronic inflammation,
147 which is associated with increased vascularity, fibroblast proliferation, synovial membrane
148 thickening and increased ECM deposition.7–10 Various immune cells have been identified in

5
149 capsular tissue from patients with FS, including B cells, macrophages, mast cells and T
150 cells36-38. There is growing evidence indicating a reciprocal homeostatic relationship between
151 immune cells and stromal cells within soft tissue, in both health and disease, and as we enter
152 the single-cell genomic age, there are emerging data of the presence of discrete subtypes of
153 immune cells in the capsule of patients with FS, including several subpopulations of dendritic
154 and T cells39. Immune cells and their mediators have been implicated in driving the
155 progression of many fibrotic disorders, and there are now the beginnings of a greater
156 appreciation for their role in soft tissue diseases. While it is simple to explain the presence of
157 immune cells in a purely pathological context, their homeostatic and inflammation-resolving
158 role in soft tissues is now evident. For example, a subtype of macrophage (those expressing
159 LYVE1 and MERTK) that has been identified in patients with rheumatoid arthritis (RA) who
160 are in remission40 are phenotypically similar to a population of macrophages that are present
161 in healthy shoulder capsule but are reduced in the capsule of patients with FS39. Loss of
162 these homeostatic or resolutory cells could indicate a function for these macrophages in
163 maintaining healthy tissue.
164

165 [H2] Pro-inflammatory cytokines

166 As FS has been historically described as a chronic fibrotic disease of the shoulder
167 capsule, the main emphasis of cytokine studies has been on the role of TGFß. Many studies
168 have unequivocally implicated TGFβ in fibrotic disease, and FS is no exception. TGFß is highly
169 expressed in FS tissue41 and can induce numerous cellular fibrotic responses, including ECM
170 protein production, fibroblast proliferation, increased myofibroblast differentiation and collagen
171 gel contractility42. The link to fibrosis will be discussed later in this section. Other inflammatory
172 mediators, including IL-1, IL-6, IL-10, GM-CSF, M-CSF, PGDF and TNF, are also dysregulated
173 in diseased capsule43,37 and may drive inflammatory and matrix responses. Fibroblasts
174 cultured from diseased capsule produced elevated levels of pro-inflammatory cytokines (such
175 as IL-6, IL-8 and CCL-20) in comparison to healthy capsular fibroblasts44.
176 Evidence suggests a prominent role for IL-17A in FS. FS tissue contains T cells (CD4+
177 and CD8+ T cells, among other subtypes), which produce IL-17A, whereas T cells are
178 predominantly absent from healthy shoulder capsule39. In this study, IL-17A induced greater
179 pro-fibrotic and inflammatory responses in FS fibroblasts compared with fibroblasts from
180 healthy tissue as a result of greater levels of the IL-17A signalling receptor (IL-17RA) on
181 fibroblasts from diseased shoulders. The potential pathological effects of IL-17A are notable
182 due to its similar effect observed in tendinopathy45, where anti-IL-17A treatment (secukinumab)
183 which is currently under clinical trial for this soft tissue disease46.
184 The levels of IL-33, which can also act as an alarmin (also known as a damage-
185 associated molecular pattern (DAMP)), are also elevated in FS tissue47. Alarmin release has

6
186 been described in other chronic musculoskeletal conditions, such as RA and osteoarthritis 48,49,
187 . A study examined H&E-stained capsular tissue from patients with FS and found fibroblastic
50

188 hypercellularity and increased vascularity as well as high levels of the alarmins IL-33, high-
189 mobility group protein B1 (HMGB1), S100A8 and S100A9; the levels of these alarmins were
190 correlated with the severity of patient-reported pain47. These alarmins can be released from
191 immune and stromal cells and may mediate crosstalk between the two compartments.
192 Advanced glycation end products (AGEs) have been associated with inflammation, and
193 these increased production and accumulation of which is seen in diabetes and routine ageing.
194 AGEs can act as immune modulators by attracting cells that release pro-inflammatory
195 cytokines to coordinate degradation and renewal of ECM. Capsular tissue of patients with FS
196 had higher immunoreactivity, blood vessel formation and perivascular adipocytes compared
197 with that in healthy capsule tissue51.
198

199 [H2] Neural and vascular changes

200 The hypervascularity that is associated with inflammation has also been proposed to
201 play a key role in the development of FS symptoms.20 Hypervascularity is prominent across
202 histological studies on FS, particularly in the rotator interval.7,8 This is the result of
203 neoangiogenesis, which is demonstrated by overexpression of the haematopoietic cell surface
204 marker CD349,21 and vascular endothelial growth factor (VEGF) in both diabetic22 and non-
205 diabetic23 patients with FS. Neoangiogenesis is accompanied by neurogenesis, which is likely
206 driven by increased expression of the nerve growth factor receptor p75.9 In patients with FS,
207 the degree of neo-innervation is correlated with the frequency of night pain and expression of
208 HMGB1.24 In addition to an increase in the density of nerves, there is also an increase in acid-
209 sensing ion channels (ASICs), calcitonin gene-related peptide (CGRP) and substance P13,25,
210 which are upregulated in hyperalgesia and chronic pain. CGRP in particular is a key connection
211 between the nervous and immune systems. CGRP is released by the synaptic terminals of
212 pain sensing neurons and acts on lymphocytes, macrophages and mast cells, among others,26
213 resulting in increased production of pro-inflammatory mediators and further immune cell
214 recruitment. In addition, expression of the melatonin receptors MTNR1A and MTNR1B is
215 upregulated in FS in response to the pro-inflammatory cytokines TNF and IL-1B27, which in
216 turn induces ASIC3 and IL-6 expression, leading to further pain and inflammation. Combined,
217 these features might explain why pain, particularly night pain, is such a prominent feature of
218 FS. Central sensitization in FS has not been comprehensively studied and so remains
219 speculative, but could explain why some patients are resistant to current interventions and
220 may benefit from a different approach.
221

222 [H2] Matrix changes

7
223 Fibrosis is the fundamental process manifesting in FS. Fibroblasts are the resident cell
224 within the joint capsule and are responsible for producing the ECM that forms the structure of
225 the tissue. In normal homeostatic conditions, type I collagen is the primary matrix protein
226 produced, whereas the more immature and disorganised type III collagen52 is deposited under
227 pathological conditions, owing to the requirement for accelerated ECM turnover. In addition,
228 the production of several other structural matrix proteins is increased in FS, including vimentin,
229 fibronectin and tenascin C53. Both matrix metalloproteinases (MMPs) and tissue inhibitors of
230 metalloproteinases (TIMPs), which regulate matrix remodelling, are dysregulated in FS.
231 MMP1–4, MMP7–9, MMP12–14 and TIMP1 and TIMP2 are implicated in FS53. These
232 proteinases have a vital role in ECM turnover, with the balance between MMPs and TIMPS
233 crucial in matrix remodelling and homeostasis, as highlighted by the development of FS in 50%
234 of recruited patients in an anti-cancer treatment trial using a TIMP analogue54.
235 Interestingly, many of the fibrotic facets of FS fibroblasts have been attributed to the
236 effects of increased TGFß production. TGFß has long been known to induce
237 transdifferentiation of fibroblasts to myofibroblasts, and myofibroblasts are a hallmark of FS
238 and other fibrotic conditions55 . In addition, there is now a greater appreciation of the
56 57

239 potential role of other cytokines, including IL-1, IL-4, IL-13, and IL-17A, in fibrosis. One such
240 aspect of fibrotic disorders that may be under cytokine regulation is the phenomenon of
241 fibroblast activation. Activated fibroblasts show higher expression of CD44, CD55, CD90
242 (THY1), CD106 (also known as VCAM1), CD248 (also known as endosialin), podoplanin,
243 uridine diphosphoglucose dehydrogenase, prolyl-4-hydroxylase and prolyl endopeptidase
244 FAP (also known as fibroblast activation protein) compared with control healthy fibroblasts,
245 which are associated with inflammatory cytokine and matrix dysregulation46. Elevated
246 expression of these proteins by fibroblasts is a phenotype of several musculoskeletal diseases
247 including frozen shoulder, and activated pathogenic fibroblasts produce more pro-
248 inflammatory proteins compared with healthy fibroblasts44. However, whether the increased
249 expression of these proteins is itself directly responsible for the pathological effects of activated
250 fibroblasts or whether it is just an epiphenomenon of fibroblast activation remains unclear58.
251

252 [H2] Metabolic factors

253 Multiple researchers have proposed that certain conditions, such as hyperlipidaemia
254 and hyperglycaemia, predispose patients with FS to propagation of pro-inflammatory and
255 pro-fibrotic signalling cascades. Multiple studies have found a strong association between
256 diabetes mellitus and FS,45–47 particularly in the setting of long-term hyperglycaemia.48–51 In
257 addition, FS in diabetic individuals tends to be prolonged and refractory to non-operative
258 treatment compared with that in non-diabetic individuals.52 This association is likely
259 multifactorial, resulting from chronic low-level inflammation in diabetic individuals as well as

8
260 the presence of AGEs. Pro-inflammatory cytokines that are consistently elevated in diabetic
261 patients, including TNF, IL-6 and IL-1B,53 are also present at high levels in the capsule and
262 synovium of patients with FS.8 Furthermore, AGEs show increased immunoreactivity in both
263 diabetic and non-diabetic patients with FS.54 AGEs contribute to fibrosis and inflammation
264 across other organ systems in diabetic individuals through multiple mechanisms.55 Firstly,
265 AGEs form cross-links between collagen molecules, leading to resistance to proteolysis and
266 reduced tissue compliance.56 Second, AGEs stimulate the production of pro-inflammatory
267 and pro-fibrotic cytokines and growth factors in stromal and immune cells through activation
268 of the receptor for AGEs.57 Finally, AGEs may also contribute to the imbalanced MMP and/or
269 TIMP activity that is found across diabetic organ systems.58
270 Elevation in serum lipids and cholesterol is also associated with the development of
271 FS, both in conjunction with diabetes and separate from it.47,59,60 Inflammatory lipoproteins,
272 which are associated with vascular inflammation and immune reaction, are independent risk
273 factors for the development of FS.61 Furthermore, the level of increase in serum lipids and
274 glucose is inversely correlated with the Constant score (a measure of patient-reported pain
275 and shoulder function) in patients with early FS,62 supporting the role of these blood markers
276 in disease progression. Transcriptional profiling of samples from patients with FS (using RNA
277 sequencing) revealed that the greatest differential gene expression was in the peroxisome
278 proliferator-activated receptor-γ (PPARγ) pathway,63 suggesting a central role for altered lipid
279 metabolism in the pathogenesis of FS. Interestingly, patients taking lipid-lowering
280 medications (such as statins) are not at an increased risk of developing FS, unlike those
281 taking anti-hyperglycaemic medications.47 This observation suggests that either a reduction
282 in serum lipids or lipid-lowering medications might be protective, which is consistent with the
283 known anti-inflammatory and anti-fibrotic effects of statins in other conditions59, 60.
284 In addition to hyperlipidaemia and hyperglycaemia, both hyperthyroidism and
285 hypothyroidism are associated with increased risk of developing FS.46,64,65 Calcitonin is likely
286 the connection between thyroid dysfunction and FS, as calcitonin deficiency is a feature of
287 both disorders.66,67 The connection between calcitonin and FS was first noted when
288 postmenopausal women being treated for osteoporosis with salmon calcitonin showed
289 improvements in their FS symptoms.68 Salmon calcitonin reduces TGFβ, type I collagen and
290 type III collagen synthesis as well as fibroblast adhesion in cultured cells,69 all of which are
291 key mediators of fibrosis in FS. These results were confirmed in a double-blind, randomized,
292 controlled trial in which intranasal calcitonin treatment improved shoulder pain and function
293 faster than placebo in patients with FS.70
294 In summary, the pathophysiology of FS is not yet clear but accumulating evidence is
295 starting to clarify the roles of inflammation, angiogenesis, neuromodulation, and fibrosis in this
296 disease (Figure 3).

9
297

298 [H1] Diagnosis, screening and prevention

299

300 [H2] Diagnosis

301 The diagnosis of FS is fraught with ambiguity, inconsistency, and confusion for
302 clinicians. Many patients can present with signs and symptoms of FS (pain, and global
303 restriction in movement) but not have pathological changes of the joint capsule61. Despite the
304 many diagnostic labels and familiar patterns of presentation with FS, there are currently no
305 formally recognised diagnostic criteria. Consensus studies indicate that pain, particularly at
306 night and with sudden or unexpected movements, along with a global loss of active and
307 passive movement of the shoulder, are reliable clinical identifiers62. While these are all
308 undoubtedly characteristic features of FS, they lack sufficient differential diagnostic capability
309 to distinguish FS from other shoulder pathologies.
310 Pain in FS is often reported in a wide and diffuse pattern around the shoulder,
311 scapula, chest and into the upper arm, usually above the elbow, which, in its early stages,
312 can make FS indistinguishable from other shoulder pathologies, such as rotator cuff
313 tendinopathy, joint arthrosis and pain from cervicogenic sources. Pain in FS is often
314 described as constant, deep, and severe. Loss of shoulder range of motion (ROM) is a key
315 feature of FS pathology but objective clinical markers that are deemed to constitute positive
316 findings are rather nebulous.
317

318 [H3] Clinical assessment. Loss of passive and active ROM is inherently associated with FS
319 but criteria are conflicting. Thresholds range from a reduction of 30% in two of three
320 unspecified directions63, to 50% loss of external rotation compared to the contralateral side64.
321 However, there is a lack of reliability in differentiating movement loss from capsule pathology
322 resulting from other potentially more serious pathologies or from self-limiting movement
323 owing to kinesiophobia and protective pain guarding61,65,66.
324 Reliably and accurately assessing shoulder movement in an individual with severe
325 pain is a clinical challenge. Often, what appears to be an abnormal loss of range can be a
326 patient self-limiting due to pain or fear. It is therefore recommended that movement is
327 assessed in a variety of positions with differing levels of support. For example, the key
328 movement of external rotation, if found to be reduced in standing position, should also be
329 assessed in the supine or lying position with the arm and trunk supported (Figure 5).
330 Similarly, assessing shoulder elevation by lifting the arm overhead could be compared with
331 lowering the head and trunk below a supported arm. A noticeable disparity in ROM is more

10
332 likely to represent kinesiophobia and movement inhibition as opposed to true capsular
333 restriction.
334 As capsular tissue is non-contractile, isometric muscle testing in the mid-range of
335 movements should elicit little pain provocation in patients with FS (Figure 5). This can be a
336 useful screening tool when considering other diagnoses such as rotator cuff tendinopathy.
337 Assessment of the cervical spine is also essential to eliminate possible cervicogenic
338 pathology such as nerve root irritation causing radicular pain.
339

340 [H3] Imaging. Plain radiographs of the glenohumeral joint are often suggested to be taken to
341 ensure that there is not substantial degenerative joint changes that could also present with
342 pain and motion loss and therefore confound the diagnosis of FS. However, in practice, a
343 working diagnosis can often be made on the basis of a good medical history and simple
344 clinical examination, with X-rays not necessarily required in primary care environments67. It
345 has been suggested that routine radiography may not confer superior diagnostic accuracy of
346 serious pathology to good clinical questioning and physical examination68.
347 The use of advanced imaging modalities such as ultrasonography and magnetic
348 resonance imaging (MRI) to diagnose FS has been proposed. Findings such as axillary
349 capsule thickening and/or obliteration of the axillary recess, coracohumeral ligament 66 and
350 rotator interval (RI) thickening, and/or hypervascularity are considered indicative of FS
351 pathology if the imaging results matches the clinical presentation69,70. Indeed, advanced
352 imaging in refractory FS cases can be extremely important in detecting undiagnosed soft
353 tissue tumours, although these undiagnosed tumours Are present in fewer than 1% of FS
354 cases66. However, imaging does not offer superior diagnostic information beyond a medical
355 history and physical examination and is therefore not recommended for routine workup71
356 however MRI may be useful if there is a clinical suspicion of another serious pathology with
357 similar symptomology to FS.
358

359 [H2] Screening and prevention

360 With new research and an associated understanding of the complex pathophysiology
361 of FS, it is increasingly apparent that the lack of clarity surrounding diagnosis of FS is, in
362 part, due to a historically oversimplified approach to this disease that does not consider the
363 heterogeneity of individuals with FS.
364 As discussed above, FS has been associated with myriad systemic diseases, such as
365 diabetes mellitus, cardiovascular disease and thyroid disorders. Although robust evidence of
366 a causal relationship between these conditions and the development of FS is lacking, there
367 are theories regarding the potential mechanisms that might underlie an increased risk of
368 developing FS. These conditions are associated with chronic low-grade inflammation21,

11
369 which has no mechanism of injury and is marked by elevated levels of active pro-
370 inflammatory cytokines but the absence of the increased neutrophil abundance associated
371 with acute inflammation72. The influence of hyperglycaemia on FS risk is mediated by pro-
372 inflammatory cytokines, which are elevated in the capsule and synovium of patients with FS
373 . Raised levels of serum cholesterol and pro-inflammatory lipoproteins have also been
73

374 detected in FS and are risk factors for cardiovascular disorders74. Thyroid-stimulating
375 hormone (TSH) levels also seem to be correlated with severity of FS17.
376 Routine haematological analyses and blood biochemical tests to assess for the
377 presence of inflammatory or metabolic markers are not routinely performed in patients with
378 FS. Although these markers have been shown to be risk factors, their prevalence across the
379 FS population and the impact they may have on disease trajectory and their relation to
380 causal mechanisms remains unknown.
381 A process that has received little attention to date is the role of chronic or persistent
382 pain. Chronic pain is now viewed as a long-term condition in its own right and has been
383 identified as a global health priority48,75. Central pain mechanisms are known to be present in
384 long-standing shoulder pain and could potentially play a greater role in FS than previously
385 considered76,77. Chronic pain could be compounded by low self-efficacy, pain perceptions
386 and health behaviours such as fear avoidance and kinesiophobia, which can be associated
387 with poor outcome78.
388 Individuals with traits of anxiety and depression might be at higher risk of longer
389 duration of symptoms and poorer prognosis79. The independent FS risk factors smoking and
390 obesity have the potential to further exacerbate levels of disability, as their presence, along
391 with sleep deprivation, lower pain thresholds80-84.
392 Individuals with recalcitrant symptoms for whom traditional mechanically-driven
393 treatments have been unsuccessful often require multi-specialty, multi-modal input to
394 address the complex physical, emotional and social dimensions that are the consequences
395 of chronic pain conditions. Like screening for existing comorbidities, validated screening
396 measures for dimensions such as fear avoidance beliefs, pain self-efficacy, sleep
397 disturbance and mood are not routinely used with this cohort of patients85-88. Further
398 research in this area is needed to determine whether such screening tools would be of value
399 when determining individual patient treatments and likely outcomes.
400

401 [H1] Management

402 The pathogenesis of FS remains incompletely understood. It is therefore unsurprising
403 that well-defined, evidence-based management guidelines are lacking. In general, a patient
404 with FS can seek non-surgical treatments, such as physiotherapy, medications, and

12
405 corticosteroid injections, or more invasive options, such as surgical interventions. Whether
406 disease duration can be influenced with treatment, and the efficacy of each intervention, is
407 unclear, as the evidence for most interventions is mixed35. Therefore, current treatment of FS
408 focuses primarily on symptom reduction, that is, pain relief and restoring mobility and function.
409

410 [H2] Non-operative management

411 There is consensus that non-operative management is the initial treatment of choice
412 for frozen shoulder89. Many non-operative management strategies have been suggested for
413 use in patients with FS. One of the reasons for this is that patients present with a wide array
414 of symptoms and varied levels of disability, which may relate to disease stage. Consequently,
415 it is suggested to adopt a treatment intervention suitable for the disease stage and pain level
416 of the patient and there is growing evidence for this approach90,91. In addition, as patients with
417 FS often have high pain levels and functional limitation in combination with a long duration of
418 symptoms, they are often motivated to try every possible intervention that might help them. As
419 symptoms may improve with time in a large proportion of those with FS, it is easy to consider
420 the intervention as the reason for improvement, when in fact this may not be the case.
421

422 [H3] Patient education. Informing the patient about FS and discussing the natural history is
423 one of the most important initial interventions. The mysterious and uncertain nature of FS can
424 be worrisome and perplexing. Good advice and education reduces patient anxiety and results
425 in subjective improvement of symptoms92; therefore, clearly explaining the evidence-based
426 knowledge of FS natural history, such as expected duration, can have substantial effects on
427 pain and function. It is important to inform patients of the options available to manage FS
428 themselves and to give them simple and clear strategies to modify their occupational or
429 recreational activities as required. It is therefore paramount that all healthcare providers
430 provide the same message to reduce confusion, contradiction, and negative stress factors.
431 Another important factor in patient education is noting the response to interventions or activity,
432 which differs for each stage; for example, in early FS, no increase in pain and inflammation
433 should be allowed, whereas in the middle and late stages, 24 hours of pain increase could be
434 allowed 93.
435

436 [H3] Physiotherapy.

437 Physiotherapy provides accelerated pain relief94 and/or improvement in ROM35,94-96 compared
438 with no treatment. However, these improvements are mostly short-term, without demonstrated
439 reduction in disease duration. It is suggested that the level of irritability of the patient be used
440 to define the appropriate intensity of the chosen management strategy93,97,98. Irritability levels

13
441 are mainly based on the intensity of pain. For example, in patients with high irritability (pain
442 level at least 7/10), the intervention should be at an intensity that is not inducing extra pain,
443 while in patients with moderate irritability (pain level 4–6/10) the intervention will increase in
444 duration and intensity, and patients with low irritability (pain <3/10) will be able to perform
445 increased duration stretches, with allowance for some pain or discomfort99.

446 Several mobilization and stretching techniques (for example, four-direction shoulder-
447 stretching100 and inferior capsular stretching101 ) are effective in early and late stages of FS for
448 pain relief 102,103
and can be recommended for increasing ROM and function . One of the
93,97

449 proposed mechanisms that might explain pain reduction in patients with FS involves the
450 sensory input that activates the endogenous pain inhibitory systems104. Further study is clearly
451 warranted to determine if endogenous pain inhibitory systems are indeed involved in manual
452 therapeutic interventions around the shoulder. However, for patients with FS who are in their
453 first high irritability stage, the use of passive mobilization or capsular stretching can be
454 counterproductive and can even increase the inflammatory response105. However, a study
455 comparing a combination of manual mobilisations and shoulder exercises to a glucocorticoid
456 injection found that the physiotherapeutic combination probably results in less improvement in
457 the short term but a similar number of adverse events106, although no clinically important
458 differences were noted at 6 or 12 months. Other mobilization techniques, such as Codman’s
459 pendulum exercises (passive mobilization of the shoulder while bent over), do not show a
460 substantial difference for pain or ROM107 compared with other techniques. Unfortunately, there
461 is insufficient evidence to quantify the ideal frequency of mobilization. The intensity of
462 stretching exercises should be determined by the patient’s irritability level, since stretching
463 beyond painful limits in a highly irritable patient results in poorer outcomes 93,97. In addition to
464 a patient’s irritability level, the Total End Range Time (TERT) can be used to report the dose
465 applied to the patient and evaluate progression . TERT is the total amount of time that the
108

466 joint is positioned at its end range and is proportional to the increase in passive ROM 109. The
467 importance of the right treatment intensity is highlighted again by a prospective study that
468 compared intensive passive stretching and manual mobilization to supportive therapy and
469 exercises within the pain limits, which demonstrated better shoulder function in the supportive
470 group at the end of the 2 years follow-up period110. However, currently there is little evidence
471 to support joint mobilizations over other non-operative interventions 97
. As such, the exact
472 effects of exercises, the extent to which they are effective, and the format of exercise therapy
473 that is the most effective is uncertain111. Preliminary evidence shows that supervised exercise
474 therapy is more effective than unsupervised exercise therapy at home112.

475 Resistance-based exercise may also have an important role in patients with FS,

14
476 although this approach has been poorly researched. The addition of strengthening exercises
477 to a multimodal programme with mobilization and electrostimulation seems to result in
478 improvements on pain, ROM, function, and muscle strength 113,114. These improvements were
479 not seen with the addition of scapulothoracic exercises, mobilization, and electromagnetic
480 therapy to a similar multimodal programme115-117.

481 The role of extracorporeal shock wave therapy (ESWT) has been investigated in the
482 treatment of frozen shoulder. In a randomized, double-blind, placebo-controlled trial comparing
483 radial ESWT to placebo shockwave therapy in 106 participants118, substantial improvement in
484 function, pain and ROM occurred in the group who received shockwave. In a trial of patients
485 with primary frozen shoulder119, focused ESWT produced superior pain outcomes compared
486 with oral prednisolone. A systematic review of 20 randomized controlled trials found some
487 evidence in favour of ESWT for reduction of pain in frozen shoulder, although the authors of
488 the review highlighted issues around the quality of evidence and were unable to perform a
489 meta-analysis. For now, definitive conclusions about the efficacy of ESWT as an adjunct to
490 treatment in frozen shoulder cannot be made120.

491 Other physiotherapy modalities, such as cold, heat, electrical modalities such as
492 transcutaneous nerve stimulation, pulsed electromagnetic field therapy or low-level laser
493 therapy, are proposed to have positive effects on pain in patients with FS. However, as these
494 modalities are typically applied as adjunctive interventions, the individual effect of each
495 technique on the natural course of FS is difficult to define. Consequently, there is only weak
496 evidence in favour of techniques such as shockwave therapy, shortwave diathermy, pulsed
497 electromagnetic field therapy, low-level laser therapy, therapeutic ultrasound, or electrical
498 stimulation to reduce pain and improve shoulder ROM in patients with FS97,121.

499 Mirror therapy is a promising mode of exercise therapy that seems to be effective in
500 the treatment of FS. This approach aims to restore the congruence between motor output and
501 sensory output 122
and has been beneficial for patients with FS for improving pain, function,
502 ROM in flexion and abduction and general health, although further research is needed 123.

503 Besides exercises that specifically target the shoulder, general physical activity is
504 recommended for general health, well-being93, improving mood and sleep, 124
and the
505 prevention of depression 124
. Physical activity can help to reduce or reverse the effects of a
506 sedentary lifestyle, which is often associated with an increase in chronic low-grade
507 inflammation and the development of insulin resistance 125.

15
508 [H3] Pharmacotherapy. Common medications for patients with FS include paracetamol or
509 acetaminophen, non-steroidal anti-inflammatory drugs (NSAIDs) and corticosteroids. The
510 evidence for the use of paracetamol in patients with FS is limited, but it may be useful when
511 there are contraindications to other medications 126-128. Paracetamol inhibits cyclo-oxygenase
512 and is active both peripherally and centrally. FS has been shown to be an inflammatory
513 process followed by fibrosis, and therefore theoretically NSAIDs should be more effective in
514 the early inflammatory stage than in the later fibrotic stage 14. However, this has not yet been
515 shown clinically. NSAIDs might be used for pain relief, but do not have an effect on ROM35.
516 In addition, NSAIDs influence the serotonergic system, which may provide some benefit in
517 modulation of perceived pain in addition to their direct anti-inflammatory effect126. Oral
518 corticosteroids provide quicker pain relief compared with placebo, but this effect has not
519 been seen in the long term, and in some cases this treatment exacerbate symptoms owing to
520 rebound pain after their discontinuation35,90,102,103.
521 Intra-articular corticosteroid injections (CSIs) are recommended in the inflammatory or
522 early stages of FS, prior to the emergence of capsular contraction, to provide pain relief and
523 reduce inflammation 129-132. Histologically, intra-articular CSIs have been associated with
524 decreased fibrosis proliferation 132. CSIs are more effective than placebo, but do not change
525 the long-term (6- and 12-month) outcome133. CSIs are more effective than physical therapy in
526 decreasing pain in the early stages of FS 35,90,102,103,134, but the difference is minimal in the
527 long term 94. CSI alone has no effect on ROM but a combination of CSI and physiotherapy
528 improves ROM 102. In general, CSI in early stage (stage I or II) FS results in greater
529 improvement in pain and function than in late (stage III or IV) FS135. Although risks are low,
530 there are potential complications with the use of intra-articular CSI, including avascular
531 necrosis, infection, muscle complaints, and pain increase 136-138. Intra-articular CSI can also
532 lead to a transient increase in serum glucose, which may be relevant in diabetic patients with
533 FS.
534

535 [H3] Alternative interventions. There is limited and mixed data for several other
536 interventions, including sodium hyaluronate injection139,140, suprascapular nerve block141,142,
537 collagenase treatment143, botulinum toxin144, and hydrodilatation145,146 for use in FS, with the
538 most supporting evidence for botulinum toxin and hydrodilatation.
539 Hydrodilatation therapy refers to intra-articular injection of a large volume of sterile
540 saline with or without corticosteroid to distend the capsule. Hydrodilatation therapy is a
541 promising intervention that is gaining in popularity over the past 10 years147-149. A meta-
542 analysis found that both CSI and hydrodilatation with corticosteroids provided superior short-
543 term pain relief, ROM improvement, and function compared with placebo, with ROM

16
544 improvements persisting to l beyond 24 months 149. Hydrodilatation with corticosteroids was
545 found to have a greater benefit than CSI150.
546 Following its successful use in Dupuytren’s disease, collagenase clostridium
547 histolyticum (CCH) has also been utilized to treat FS. . CCH is typically given in a series of 3
548 injections over 6 weeks. A randomized study showed improvement in subjective function with
549 CCH but no notable increase in ROM compared with placebo151. Another study found a
550 greater improvement in ROM at 3 months with CCH than with exercise therapy alone143.
551 Histological examination of capsular tissue in a rat model of FS revealed less fibrosis with
552 CCH injection than with CSI or saline152. These data support a potential role for CCH in the
553 management of FS.
554 In conclusion, while many interventions have been described, the most reliable
555 benefits are from steroid injection and NSAIDs in stage I, physiotherapy in stage II/III, and
556 advancement of physiotherapy to mirror or resistance exercises in Stage III (Table 1).
557

558 [H2] Operative management

559 After ruling out other causes of pain and stiffness of the shoulder, the patient should
560 be informed that the natural history of the condition is eventual resolution in most patients.
561 However, symptoms and disability persist in some cases, and surgical management may
562 provide a faster, more complete recovery. The aim of surgical approaches in FS is to release
563 the fibrous, thickened and tightened glenohumeral joint capsule and associated contracted
564 ligaments to improve ROM of the glenohumeral joint, and to decrease pain.
565 [H3] MUA. The aim of MUA is to tear the shoulder joint capsule and thereby improve ROM.
566 In an anaesthetized shoulder, the procedure involves applying a passive stretch to the
567 glenohumeral joint, in all shoulder ROM directions. There are conflicting opinions as to the
568 ideal time to perform MUA in patients with idiopathic FS, from as soon as FS is diagnosed
569
153,154
to up to 12 months of failed non-operative treatment 1,91,155. Several studies have cited
570 improved patient outcomes from MUA in more than 80% of subjects103,154,156-158.
571 [H3] ACR with or without MUA manipulation. ACR involves cutting and removing the
572 thickened, swollen, inflamed abnormal capsule under direct arthroscopic control. ACR is a
573 safe and an effective modality in treating FS155,156,159-163 and may offer distinct advantages
574 when compared to other methods of treatment. For example, direct visualization of the
575 affected joint allows for diagnostic confirmation and enables additional pathology to be ruled
576 out. The effectiveness of ACR has been demonstrated in multiple studies, with a dramatic
577 reduction in pain scores, increased ROM as well as overall increased shoulder
578 function163,164.161 165 160 162.
579 [H3] Comparison of ACR and MUA. UK FROST 166 is the largest multicentre randomized
580 trial in FS, which compared early structured physiotherapy 107, MUA and ACR. Early

17
581 structured physiotherapy and post-procedural physiotherapy programmes were standardized
582 at 12 sessions over 12 weeks. At the 12-month primary endpoint, most participants improved
583 to near full function, as determined by the Oxford Shoulder Score (43 out of 48), a quality of
584 life outcome score. Oxford Shoulder Scores were significantly better in the ACR group than
585 in the MUA and early structured physiotherapy groups (p <0.01), while Oxford Shoulder
586 Scores were higher with MUA than with early structured physiotherapy alone. Economic
587 analysis in the UK FROST study 166 showed that MUA is more costly (£276) than early
588 structured physiotherapy, while ACR was substantially more expensive (£1,734) than early
589 structured physiotherapy. All three treatments led to substantial patient recovery, with no
590 clear superiority for any approach. ACR resulted in patients requiring the least further
591 treatment, but is the most expensive, was associated with higher risks, suggesting it should
592 only be utilised when less costly and less invasive interventions fail. Early structured
593 physiotherapy was accessed faster, but more patients required further treatment.
594 In summary, the surgical options of MUA and ACR may provide an earlier, potentially
595 more complete resolution of pain and restoration of ROM and function, although these
596 interventions should be considered only after non-operative management approaches have
597 failed.
598

599 [H1] Quality of life

600 FS results in significant functional disability and reduction of quality of life, as shown
601 using various questionnaires and scores, such as the visual analogue scale , disabilities of
125

602 the arm, shoulder and hand (DASH) score, the 36-item Short-Form (SF-36) health survey, and
603 the Hamilton depression rating scale and anxiety scores78. It is not clear what factors predict
604 the severity of pain and disability as well as quality of life in patients with FS167. The prolonged
605 disease course in FS results in greatly impaired sleep and everyday activities and, therefore,
606 markedly affects the physical, psychological and social quality of life of patients168,169,167. FS
607 has been linked to anxiety and depression169. Comorbidities are associated with increased
608 disability and reduced quality of life in these patients but not with the severity of pain.
609 Psychiatric disorders can also affect pain, disability and quality of life as well as patients’
610 characteristic and objective symptoms170, but the effect of these parameters on FS requires
611 further study.
612 The patient’s perspective and experience has been largely overlooked in research of
613 FS. This is startling, particularly when considering the vast numbers affected by this condition
614 and the subsequent healthcare cost and implications of long-term symptoms and reduced
615 quality of life as a result of FS. A study72 exploring patients’ perception of FS treatment
616 highlighted the severe pain and loss of function that impacted the daily lives of patients with
617 FS, alongside sleep disturbance and inability to perform work duties. Delay to diagnosis was

18
618 a cause of frustration and worry for the patients interviewed, as the severity of pain often led
619 patients to suspect that a more sinister cause of pain might underlie the symptoms. Patients
620 also highlighted a lack of a definitive diagnosis alongside unclear pathways for management
621 of their condition and emphasized a mismatch between their perception of the impact of FS
622 and that of clinician’s. Although only involving a small number of patients, this study stresses
623 the need for a better understanding of FS, for both clinicians and patients. To improve a
624 patient’s experience with FS, a prompt diagnosis, a clear understanding of the treatment
625 options available and an explanation of the course of this painful condition are priorities.
626 Aligning treatment goals with those of patients suffering with FS should underpin clinicians’
627 interaction with patients who present with FS and future research into this condition. Clearly,
628 the patient’s voice has not been heard enough in studies thus far.
629
630 [H1] Outlook
631 Research in recent decades has provided improved understanding of known risk
632 factors and disease progression, and, importantly, insight into the basic mechanisms driving
633 the disease process, with the potential for new therapeutic targets. Despite affecting 5% of
634 the world’s population156, research into FS lags behind that of other musculoskeletal
635 conditions, and integration of new findings into a comprehensive treatment strategy that can
636 be applied across the spectrum of disease (from early- through to late-stage disease) and
637 medical practitioners (physiotherapists, primary care physicians, and surgeons) remains
638 elusive. Of note, emerging basic science research needs to be assimilated into clinical
639 practice to provide clinicians with a principal picture of the pathophysiological processes
640 involved in FS.
641

642 [H2] Physiotherapy advances

643 The component of physiotherapy that includes mobilization techniques beyond the
644 threshold of pain in early disease can be detrimental to patient engagement and is explained
645 by the unique mechanosensitive properties of fibroblasts and the fact that the inflammatory
646 response makes fibroblasts more sensitive to progressive mechanical stress110. However,
647 progressive stretching exercises up to tolerable pain levels results in an increase in the
648 MMP/TIMP ratio, thus favouring collagen remodelling, and importantly is superior to
649 supervised neglect171. Therefore, some mechanical stress is advantageous in promoting
650 remodelling of the ECM, especially in the later stage of the condition. Thus, further research
651 is now required on the role of precision ‘tailored’ physiotherapy guidelines for the treatment of
652 FS.
653

654 [H2] Translational advances

19
655 Novel bench to bedside treatment strategies have been suggested to intervene in the
656 inflammation–fibrosis cascade in different ways. Given the dominant role of the TGFβ
657 pathway in FS, gene silencing (with small interfering RNAs (siRNAs)) was utilized to knock
658 down SMAD4 (a central mediator of TGFβ signalling) in a rat model of FS induced by
659 immobilization172. Suppression of the TGFβ pathway impaired the inflammatory response
660 and myofibroblast differentiation and resulted in better shoulder ROM and an increased joint
661 volume in FS mice than in control rats. In a placebo-controlled, double-blind, randomized
662 trial, the thyroid hormone calcitonin (delivery by nasal spray) was moderately efficacious in
663 improving pain and functional outcomes in FS173. Another study found an association
664 between expression of alarmins (such as IL-33, IL-17A, and HMGB1) and pain levels47,
665 highlighting a potential role for anti-cytokine therapies in treating FS; however, additional
666 preclinical work is required before progressing towards potential first in human FS trials.
667

668 [H2] Clinical advances

669 Clinically, knowledge about pain management and the association between pain and
670 other psychological disorders, such as depression and anxiety, is expanding. In addition, the
671 central element of night pain in FS has a major impact on patients’ overall wellbeing owing to
672 interruption of sleep, and the emotional and societal implications of these factors are not well
673 understood. These issues may play an important role in the management of FS in the future
674 as these associations are clarified.
675 The lack of well-conducted prospective longitudinal studies to adequately investigate
676 prognostic risk factors for FS is a barrier to furthering our understanding of this pathology.
677 Although none of the risk factors discussed earlier are known to have a direct causative role,
678 their high prevalence in the FS population indicates that reducing or improving them would
679 be useful. Therefore, clear, simple, and actionable health promotion advice regarding
680 smoking cessation, physical activity, stress, and sleep levels, diet and weight management is
681 recommended. Even slight increases in physical activity and exercise can substantially
682 reduce the relative risks of both morbidity and mortality associated with many of the
683 proposed risk factors for FS174. Increasing walking time to just 2 hours a week significantly
684 reduces the risk of cardiovascular disease in individuals with diabetes175 and both aerobic
685 and resistance-based exercise demonstrably reduce morbidity risk176-178. In the near future,
686 ‘frozen shoulder’ could be considered an umbrella term for emerging subgroups of this
687 pathology, each with their own aetiology, disease progression, prognosis and management
688 strategies. While novel therapeutic modalities to treat FS are evolving, there should also
689 perhaps be a concomitant focus on preventing the important and avoidable risk factors
690 involved. Population health messages delivered consistently by health care professionals
691 across all sectors should be considered core tenets within musculoskeletal care. This should

20
692 include an honest, open yet compassionate discussion and acknowledgment that it may take
693 weeks, months or years for the benefits of lifestyle changes to make any demonstrable
694 impact on patient outcome179.
695 Clinical trials in FS remain challenging, with inconsistencies in outcome measures,
696 heterogeneity of patient stage at recruitment and the requirement for prolonged follow up.
697 Although advances have been made toward a consensus in terms of a core set of outcome
698 domains for use in shoulder trials180, several important issues need to be addressed,
699 including heterogeneity in study design, stage- or patient-specific treatment protocols, and
700 how we classify response to any treatment regimens – specifically those altering specific
701 physiotherapy regimes. Further research is now required to link the genetic, epigenetic,
702 environmental and therapeutic factors together so that curative or preventive therapies for FS
703 can be obtained. These findings should give impetus to the development of new diagnostic
704 techniques, evidence-based screening methods and more targeted personalised
705 interventions, which underscore the need for a multidisciplinary approach to the management
706 of FS.
707

708

709

710

711

712

713

714

715

716

717

718

719

720

721

722

723

724

725

726

727

728

21
729 References
730 1. Neviaser, R.J. & Neviaser, T.J. The frozen shoulder. Diagnosis and management. Clin
731 Orthop Relat Res, 59-64 (1987).

732 Classic orginal article describing frozen shoulder and treatment options.

733

734 2. Kim, D.H., et al. Is frozen shoulder completely resolved at 2 years after the onset of
735 disease? J Orthop Sci 25, 224-228 (2020).

736 3. Shaffer, B., Tibone, J.E. & Kerlan, R.K. Frozen shoulder. A long-term follow-up. The
737 Journal of bone and joint surgery. American volume 74, 738-746 (1992).

738 4. Hand, C., Clipsham, K., Rees, J.L. & Carr, A.J. Long-term outcome of frozen shoulder.
739 J Shoulder Elbow Surg 17, 231-236 (2008).

740 Key article looking at the longer term outcomes of frozen shoulder.

741

742 5. Walker-Bone, K., Palmer, K.T., Reading, I., Coggon, D. & Cooper, C. Prevalence and
743 impact of musculoskeletal disorders of the upper limb in the general population. Arthritis
744 Rheum 51, 642-651 (2004).

745 6. Sarasua, S.M., Floyd, S., Bridges, W.C. & Pill, S.G. The epidemiology and etiology of
746 adhesive capsulitis in the U.S. Medicare population. BMC Musculoskelet Disord 22,
747 828 (2021).

748 7. Kingston, K., Curry, E.J., Galvin, J.W. & Li, X. Shoulder adhesive capsulitis:
749 epidemiology and predictors of surgery. J Shoulder Elbow Surg 27, 1437-1443 (2018).

750 8. Rizk, T.E. & Pinals, R.S. Frozen shoulder. Semin Arthritis Rheum 11, 440-452 (1982).

751 9. Looney, C.G., Raynor, B. & Lowe, R. Adhesive capsulitis of the hip: a review. J Am
752 Acad Orthop Surg 21, 749-755 (2013).

753 10. Cui, Q., Milbrandt, T., Millington, S., Anderson, M. & Hurwitz, S. Treatment of
754 posttraumatic adhesive capsulitis of the ankle: a case series. Foot Ankle Int 26, 602-
755 606 (2005).

756 11. Kraal, T., Hekman, K. & van den Bekerom, M.P.J. What Is the Right Timing for
757 Arthroscopic Capsular Release of a Frozen Shoulder? Letter to the Editor. Orthop J
758 Sports Med 8, 2325967120903710 (2020).

759 12. Dias, R., Cutts, S. & Massoud, S. Frozen shoulder. BMJ 331, 1453-1456 (2005).

760 13. de la Serna, D., Navarro-Ledesma, S., Alayon, F., Lopez, E. & Pruimboom, L. A
761 Comprehensive View of Frozen Shoulder: A Mystery Syndrome. Front Med (Lausanne)
762 8, 663703 (2021).

763 14. Le, H.V., Lee, S.J., Nazarian, A. & Rodriguez, E.K. Adhesive capsulitis of the shoulder:
764 review of pathophysiology and current clinical treatments. Shoulder Elbow 9, 75-84
765 (2017).

22
766 15. Juel, N.G., Brox, J.I., Brunborg, C., Holte, K.B. & Berg, T.J. Very High Prevalence of
767 Frozen Shoulder in Patients With Type 1 Diabetes of >/=45 Years' Duration: The
768 Dialong Shoulder Study. Arch Phys Med Rehabil 98, 1551-1559 (2017).

769 16. Alhashimi, R.A.H. Analytical Observational Study of Frozen Shoulder among Patients
770 with Diabetes Mellitus. Joints 6, 141-144 (2018).

771 17. Schiefer, M., et al. Prevalence of hypothyroidism in patients with frozen shoulder. J
772 Shoulder Elbow Surg 26, 49-55 (2017).

773 18. Wang, J.Y., et al. Hyperlipidemia Is a Risk Factor of Adhesive Capsulitis: Real-World
774 Evidence Using the Taiwanese National Health Insurance Research Database. Orthop
775 J Sports Med 9, 2325967120986808 (2021).

776 19. Li, W., Lu, N., Xu, H., Wang, H. & Huang, J. Case control study of risk factors for frozen
777 shoulder in China. Int J Rheum Dis 18, 508-513 (2015).

778 20. Itoi, E., et al. Shoulder Stiffness: Current Concepts and Concerns. Arthroscopy 32,
779 1402-1414 (2016).

780 Classic article on general aspects of shoulder stiffness and management options.

781

782 21. Pietrzak, M. Adhesive capsulitis: An age related symptom of metabolic syndrome and
783 chronic low-grade inflammation? Med Hypotheses 88, 12-17 (2016).

784 22. Smith, S.P., Devaraj, V.S. & Bunker, T.D. The association between frozen shoulder
785 and Dupuytren's disease. Journal of shoulder and elbow surgery / American Shoulder
786 and Elbow Surgeons ... [et al.] 10, 149-151 (2001).

787 23. Green, H.D., et al. A genome-wide association study identifies 5 loci associated with
788 frozen shoulder and implicates diabetes as a causal risk factor. PLoS genetics 17,
789 e1009577 (2021).

790 24. Kilian, O., et al. Enhanced alpha 1(I) mRNA expression in frozen shoulder and
791 dupuytren tissue. Eur J Med Res 12, 585-590 (2007).

792 25. Thompson, W.A. & Kopell, H.P. Peripheral entrapment neuropathies of the upper
793 extremity. N Engl J Med 260, 1261-1265 (1959).

794 26. Giordano, L., Sarandria, D., Fabiano, B., Del Carro, U. & Bussi, M. Shoulder function
795 after selective and superselective neck dissections: clinical and functional outcomes.
796 Acta Otorhinolaryngol Ital 32, 376-379 (2012).

797 27. Tanishima, T. & Yoshimasu, N. Development and prevention of frozen shoulder after
798 acute aneurysm surgery. Surg Neurol 48, 19-22 (1997).

799 28. Papalia, R., et al. Frozen shoulder or shoulder stiffness from Parkinson disease?
800 Musculoskelet Surg 103, 115-119 (2019).

801 29. Wilson, R.D. & Chae, J. Hemiplegic Shoulder Pain. Phys Med Rehabil Clin N Am 26,
802 641-655 (2015).

803 30. Neviaser, A.S. & Neviaser, R.J. Adhesive capsulitis of the shoulder. J Am Acad Orthop
804 Surg 19, 536-542 (2011).

23
805 31. Ryu, J.D., et al. Expression of vascular endothelial growth factor and angiogenesis in
806 the diabetic frozen shoulder. J Shoulder Elbow Surg 15, 679-685 (2006).

807 32. Xu, Y., Bonar, F. & Murrell, G.A. Enhanced expression of neuronal proteins in idiopathic
808 frozen shoulder. J Shoulder Elbow Surg 21, 1391-1397 (2012).

809 33. Lho, Y.M., et al. Inflammatory cytokines are overexpressed in the subacromial bursa
810 of frozen shoulder. J Shoulder Elbow Surg 22, 666-672 (2013).

811 34. Ryan, V., Brown, H., Minns Lowe, C.J. & Lewis, J.S. The pathophysiology associated
812 with primary (idiopathic) frozen shoulder: A systematic review. BMC Musculoskelet
813 Disord 17, 340 (2016).

814 35. Neviaser, A.S. & Hannafin, J.A. Adhesive capsulitis: a review of current treatment. Am
815 J Sports Med 38, 2346-2356 (2010).

816 36. Hand, G.C.R., Athanasou, N.A., Matthews, T. & Carr, A.J. The pathology of frozen
817 shoulder. The Journal of Bone and Joint Surgery. British volume 89-B, 928-932 (2007).

818 Important basic science study looking at the pathophysiology of frozen shoulder

819

820 37. Kanbe, K., Inoue, K., Inoue, Y. & Chen, Q. Inducement of mitogen-activated protein
821 kinases in frozen shoulders. Journal of orthopaedic science : official journal of the
822 Japanese Orthopaedic Association 14, 56-61 (2009).

823 38. Dakin, S.G., et al. Tissue inflammation signatures point towards resolution in adhesive
824 capsulitis. Rheumatology 58, 1109-1111 (2019).

825 39. Akbar, M., et al. Translational targeting of inflammation and fibrosis in frozen shoulder:
826 Molecular dissection of the T cell/IL-17A axis. Proc Natl Acad Sci U S A 118(2021).

827 40. Alivernini, S., et al. Distinct synovial tissue macrophage subsets regulate inflammation
828 and remission in rheumatoid arthritis. Nat Med 26, 1295-1306 (2020).

829 41. Rodeo, S.A., Hannafin, J.A., Tom, J., Warren, R.F. & Wickiewicz, T.L.
830 Immunolocalization of cytokines and their receptors in adhesive capsulitis of the
831 shoulder. J. Orthop. Res. 15, 427-436 (1997).

832 First description of dysregulated cytokines and inflammation in frozen shoulder.

833

834 42. Kraal, T., et al. The puzzling pathophysiology of frozen shoulders - a scoping review. J
835 Exp Orthop 7, 91 (2020).

836 43. Bunker, T.D., J, R., S, B.K. & L, H.D. Expression of growth factors, cytokines and matrix
837 metalloproteinases in frozen shoulder. The Journal of Bone and Joint Surgery. British
838 volume 82-B, 768-773 (2000).

839 44. Akbar, M., et al. Fibroblast activation and inflammation in frozen shoulder. PLoS ONE
840 14, e0215301 (2019).

841 45. Millar, N.L., et al. IL-17A mediates inflammatory and tissue remodelling events in early
842 human tendinopathy. Sci Rep 6, 27149 (2016).

24
843 46. Millar, N.L., et al. Tendinopathy. Nat Rev Dis Primers 7, 82 (2021).

844 47. Cher, J.Z.B., et al. Alarmins in Frozen Shoulder: A Molecular Association Between
845 Inflammation and Pain. Am J Sports Med 46, 671-678 (2018).

846 Article discussing alarmins and clincal correlation of inflammation and pain.

847

848 48. Moayedi, M., et al. Contribution of chronic pain and neuroticism to abnormal forebrain
849 gray matter in patients with temporomandibular disorder. Neuroimage 55, 277-286
850 (2011).

851 49. Millar, N.L., Murrell, G.A. & McInnes, I.B. Alarmins in tendinopathy: unravelling new
852 mechanisms in a common disease. Rheumatology (Oxford) 52, 769-779 (2013).

853 50. Nefla, M., Holzinger, D., Berenbaum, F. & Jacques, C. The danger from within:
854 alarmins in arthritis. Nat Rev Rheumatol 12, 669-683 (2016).

855 51. Hwang, K.R., et al. Advanced glycation end products in idiopathic frozen shoulders. J
856 Shoulder Elbow Surg 25, 981-988 (2016).

857 52. Rodeo, S.A., Hannafin, J.A., Tom, J., Warren, R.F. & Wickiewicz, T.L.
858 Immunolocalization of cytokines and their receptors in adhesive capsulitis of the
859 shoulder. J Orthop Res 15, 427-436 (1997).

860 53. Cho, C.-H., Song, K.-S., Kim, B.-S., Kim, D.H. & Lho, Y.-M. Biological Aspect of
861 Pathophysiology for Frozen Shoulder. BioMed Research International 2018, 7274517
862 (2018).

863 54. Hutchinson, J.W., Tierney, G.M., Parsons, S.L. & Davis, T.R. Dupuytren's disease and
864 frozen shoulder induced by treatment with a matrix metalloproteinase inhibitor. J Bone
865 Joint Surg Br 80, 907-908 (1998).

866 55. Frangogiannis, N. Transforming growth factor-beta in tissue fibrosis. J Exp Med 217,
867 e20190103 (2020).

868 56. Pardali, E., Sanchez-Duffhues, G., Gomez-Puerto, M.C. & Ten Dijke, P. TGF-beta-
869 Induced Endothelial-Mesenchymal Transition in Fibrotic Diseases. Int J Mol Sci
870 18(2017).

871 57. Caja, L., et al. TGF-beta and the Tissue Microenvironment: Relevance in Fibrosis and
872 Cancer. Int J Mol Sci 19(2018).

873 58. Crowe, L.A.N., et al. Stromal “activation” markers do not confer pathogenic activity in
874 tendinopathy. Translational Sports Medicine 4, 268-279 (2021).

875 59. Taqueti, V.R. & Ridker, P.M. Lipid-Lowering and Anti-Inflammatory Benefits of Statin
876 Therapy: More Than Meets the Plaque. Circ Cardiovasc Imaging 10(2017).

877 60. Jeong, C., et al. Exploring the In Vivo Anti-Inflammatory Actions of Simvastatin-Loaded
878 Porous Microspheres on Inflamed Tenocytes in a Collagenase-Induced Animal Model
879 of Achilles Tendinitis. Int J Mol Sci 19(2018).

25
880 61. Hollmann, L., Halaki, M., Kamper, S.J., Haber, M. & Ginn, K.A. Does muscle guarding
881 play a role in range of motion loss in patients with frozen shoulder? Musculoskelet Sci
882 Pract 37, 64-68 (2018).

883 62. Walmsley, S., Rivett, D.A. & Osmotherly, P.G. Adhesive capsulitis: establishing
884 consensus on clinical identifiers for stage 1 using the DELPHI technique. Phys Ther
885 89, 906-917 (2009).

886 63. Sharma, S.P., Baerheim, A., Moe-Nilssen, R. & Kvale, A. Adhesive capsulitis of the
887 shoulder, treatment with corticosteroid, corticosteroid with distension or treatment-as-
888 usual; a randomised controlled trial in primary care. BMC Musculoskelet Disord 17, 232
889 (2016).

890 64. Lee, S.H., et al. Measurement of Shoulder Range of Motion in Patients with Adhesive
891 Capsulitis Using a Kinect. PLoS ONE 10, e0129398 (2015).

892 65. De Baets, L., Matheve, T., Traxler, J., Vlaeyen, J. & Timmermans, A. Pain-related
893 beliefs are associated with arm function in persons with frozen shoulder. Shoulder
894 Elbow 12, 432-440 (2020).

895 66. Quan, G.M., Carr, D., Schlicht, S., Powell, G. & Choong, P.F. Lessons learnt from the
896 painful shoulder; a case series of malignant shoulder girdle tumours misdiagnosed as
897 frozen shoulder. Int Semin Surg Oncol 2, 2 (2005).

898 67. Rangan, A., et al. Frozen Shoulder. Shoulder Elbow 7, 299-307 (2015).

899 68. Roberts, S., et al. Routine X-rays for suspected frozen shoulder offer little over
900 diagnosis based on history and clinical examination alone. Musculoskeletal Care 17,
901 288-292 (2019).

902 69. Choi, Y.H. & Kim, D.H. Correlations between clinical features and MRI findings in early
903 adhesive capsulitis of the shoulder: a retrospective observational study. BMC
904 Musculoskelet Disord 21, 542 (2020).

905 70. Do, J.G., Hwang, J.T., Yoon, K.J. & Lee, Y.T. Correlation of Ultrasound Findings With
906 Clinical Stages and Impairment in Adhesive Capsulitis of the Shoulder. Orthop J Sports
907 Med 9, 23259671211003675 (2021).

908 71. Lewis, J. Frozen shoulder contracture syndrome - Aetiology, diagnosis and
909 management. Man Ther 20, 2-9 (2015).

910 72. Minihane, A.M., et al. Low-grade inflammation, diet composition and health: current
911 research evidence and its translation. Br J Nutr 114, 999-1012 (2015).

912 73. Austin, D.C., Gans, I., Park, M.J., Carey, J.L. & Kelly, J.D.t. The association of
913 metabolic syndrome markers with adhesive capsulitis. J Shoulder Elbow Surg 23,
914 1043-1051 (2014).

915 74. Sung, C.M., Jung, T.S. & Park, H.B. Are serum lipids involved in primary frozen
916 shoulder? A case-control study. J Bone Joint Surg Am 96, 1828-1833 (2014).

917 75. Treede, R.D., et al. Chronic pain as a symptom or a disease: the IASP Classification
918 of Chronic Pain for the International Classification of Diseases (ICD-11). Pain 160, 19-
919 27 (2019).

26
920 76. Sanchis, M.N., Lluch, E., Nijs, J., Struyf, F. & Kangasperko, M. The role of central
921 sensitization in shoulder pain: A systematic literature review. Semin Arthritis Rheum
922 44, 710-716 (2015).

923 Important article looking at the role of central pain in the shoulder.

924

925 77. Noten, S., et al. Central Pain Processing in Patients with Shoulder Pain: A Review of
926 the Literature. Pain Pract 17, 267-280 (2017).

927 78. Bagheri, F., Ebrahimzadeh, M.H., Moradi, A. & Bidgoli, H.F. Factors Associated with
928 Pain, Disability and Quality of Life in Patients Suffering from Frozen Shoulder. Arch
929 Bone Jt Surg 4, 243-247 (2016).

930 79. Chiaramonte, R., Bonfiglio, M. & Chisari, S. A significant relationship between
931 personality traits and adhesive capsulitis. Rev Assoc Med Bras (1992) 66, 166-173
932 (2020).

933 80. John, U., et al. Tobacco smoking in relation to pain in a national general population
934 survey. Prev Med 43, 477-481 (2006).

935 81. Guneli, E., Gumustekin, M. & Ates, M. Possible involvement of ghrelin on pain
936 threshold in obesity. Med Hypotheses 74, 452-454 (2010).

937 82. Pisinger, C., et al. The association between active and passive smoking and frequent
938 pain in a general population. Eur J Pain 15, 77-83 (2011).

939 83. Kundermann, B., Krieg, J.C., Schreiber, W. & Lautenbacher, S. The effect of sleep
940 deprivation on pain. Pain Res Manag 9, 25-32 (2004).

941 84. Schuh-Hofer, S., et al. One night of total sleep deprivation promotes a state of
942 generalized hyperalgesia: a surrogate pain model to study the relationship of insomnia
943 and pain. Pain 154, 1613-1621 (2013).

944 85. van Noorden, M.S., et al. Pre-adult versus adult onset major depressive disorder in a
945 naturalistic patient sample: the Leiden Routine Outcome Monitoring Study. Psychol
946 Med 41, 1407-1417 (2011).

947 86. Nicholas, M.K. & Blyth, F.M. Are self-management strategies effective in chronic pain
948 treatment? Pain Manag 6, 75-88 (2016).

949 87. Bastien, C.H., Vallieres, A. & Morin, C.M. Validation of the Insomnia Severity Index as
950 an outcome measure for insomnia research. Sleep Med 2, 297-307 (2001).

951 88. Zigmond, A.S. & Snaith, R.P. The hospital anxiety and depression scale. Acta Psychiatr
952 Scand 67, 361-370 (1983).

953 89. Yip, M., et al. Treatment of Adhesive Capsulitis of the Shoulder: A Critical Analysis
954 Review. JBJS Rev 6, e5 (2018).

955 90. Cho, C.H., Bae, K.C. & Kim, D.H. Treatment Strategy for Frozen Shoulder. Clin Orthop
956 Surg 11, 249-257 (2019).

957 91. Redler, L.H. & Dennis, E.R. Treatment of Adhesive Capsulitis of the Shoulder. J Am
958 Acad Orthop Surg 27, e544-e554 (2019).

27
959 92. Jones, S., Hanchard, N., Hamilton, S. & Rangan, A. A qualitative study of patients'
960 perceptions and priorities when living with primary frozen shoulder. BMJ Open 3,
961 e003452 (2013).

962 93. Vermeulen, H.S., R; Hekman, KMC; Burg, DH van der; Struyf, F. De SNN
963 Praktijkrichtlijn Frozen Shoulder voor fysiotherapeuten 2017. (2017 ).

964 94. Brue, S., et al. Idiopathic adhesive capsulitis of the shoulder: a review. Knee Surg
965 Sports Traumatol Arthrosc 15, 1048-1054 (2007).

966 95. Sheridan, M.A. & Hannafin, J.A. Upper extremity: emphasis on frozen shoulder. Orthop
967 Clin North Am 37, 531-539 (2006).

968 96. Hanchard, N.C.A., et al. Evidence-based clinical guidelines for the diagnosis,
969 assessment and physiotherapy management of contracted (frozen) shoulder. (2011).

970 97. Kelley, M.J., et al. Shoulder pain and mobility deficits: adhesive capsulitis. J Orthop
971 Sports Phys Ther 43, A1-31 (2013).

972 98. McClure, P.W. & Michener, L.A. Staged Approach for Rehabilitation Classification:
973 Shoulder Disorders (STAR-Shoulder). Phys Ther 95, 791-800 (2015).

974 99. Duenas, L., et al. A Manual Therapy and Home Stretching Program in Patients With
975 Primary Frozen Shoulder Contracture Syndrome: A Case Series. J Orthop Sports Phys
976 Ther 49, 192-201 (2019).

977 Article giving key home exercises for frozen shoulder.

978

979 100. Griggs, S.M., Ahn, A. & Green, A. Idiopathic adhesive capsulitis. A prospective
980 functional outcome study of nonoperative treatment. J Bone Joint Surg Am 82, 1398-
981 1407 (2000).

982 101. Paul, A., Rajkumar, J.S., Peter, S. & Lambert, L. Effectiveness of sustained stretching
983 of the inferior capsule in the management of a frozen shoulder. Clin Orthop Relat Res
984 472, 2262-2268 (2014).

985 102. Favejee, M.M., Huisstede, B.M. & Koes, B.W. Frozen shoulder: the effectiveness of
986 conservative and surgical interventions--systematic review. Br J Sports Med 45, 49-56
987 (2011).

988 103. Georgiannos, D., Markopoulos, G., Devetzi, E. & Bisbinas, I. Adhesive Capsulitis of the
989 Shoulder. Is there Consensus Regarding the Treatment? A Comprehensive Review.
990 Open Orthop J 11, 65-76 (2017).

991 104. Teys, P., Bisset, L., Collins, N., Coombes, B. & Vicenzino, B. One-week time course of
992 the effects of Mulligan's Mobilisation with Movement and taping in painful shoulders.
993 Man Ther 18, 372-377 (2013).

994 105. Mertens, M.G., et al. An overview of effective and potential new conservative
995 interventions in patients with frozen shoulder. Rheumatol Int 42, 925-936 (2022).

996 106. Page, M.J., et al. Manual therapy and exercise for adhesive capsulitis (frozen
997 shoulder). Cochrane Database Syst Rev, CD011275 (2014).

28
 998 107. Zavala-Gonzalez, J., Pavez-Baeza, F., Gutierrez-Espinoza, H. & Olguin-Huerta, C. The
999 effectiveness of joint mobilization techniques for range of motion in adult patients with
1000 primary adhesive capsulitis of the shoulder: a systematic review and meta-analysis.
1001 Medwave 18, e7265 (2018).

1002 108. Kelley, M.J., McClure, P.W. & Leggin, B.G. Frozen shoulder: evidence and a proposed
1003 model guiding rehabilitation. J Orthop Sports Phys Ther 39, 135-148 (2009).

1004 109. Flowers, K.R. & LaStayo, P. Effect of total end range time on improving passive range
1005 of motion. J Hand Ther 7, 150-157 (1994).

1006 110. Diercks, R.L. & Stevens, M. Gentle thawing of the frozen shoulder: a prospective study
1007 of supervised neglect versus intensive physical therapy in seventy-seven patients with
1008 frozen shoulder syndrome followed up for two years. J Shoulder Elbow Surg 13, 499-
1009 502 (2004).

1010 111. Hanchard, N.C., et al. Evidence-based clinical guidelines for the diagnosis,
1011 assessment and physiotherapy management of contracted (frozen) shoulder: quick
1012 reference summary. Physiotherapy 98, 117-120 (2012).

1013 Excellent quick reference for management of frozen shoulder for the busy clinician.

1014

1015 112. Russell, S., et al. A blinded, randomized, controlled trial assessing conservative
1016 management strategies for frozen shoulder. J Shoulder Elbow Surg 23, 500-507
1017 (2014).

1018 113. Rawat, P., Eapen, C. & Seema, K.P. Effect of rotator cuff strengthening as an adjunct
1019 to standard care in subjects with adhesive capsulitis: A randomized controlled trial. J
1020 Hand Ther 30, 235-241.e238 (2017).

1021 114. Kumar, G., Sudhakar, S., Sudhan, S. & Jyothi, N. Subscapularis muscle spray and
1022 stretch technique with conventional physical therapy for the management of adhesive
1023 capsulitis. Biomedicine (india) 37, 511‐517 (2017).

1024 115. Celik, D. Comparison of the outcomes of two different exercise programs on frozen
1025 shoulder. Acta Orthop Traumatol Turc 44, 285-292 (2010).

1026 116. Junaid, M., et al. A COMPARATIVE STUDY TO DETERMINE THE EFFICACY OF
1027 ROUTINE PHYSICAL THERAPY TREATMENT WITH AND WITHOUT KALTENBORN
1028 MOBILIZATION ON PAIN AND SHOULDER MOBILITY IN FROZEN SHOULDER
1029 PATIENTS. International Journal of Physiotherapy 3, 316-319 (2016).

1030 117. Leclaire, R. & Bourgouin, J. Electromagnetic treatment of shoulder periarthritis: a

1031 randomized controlled trial of the efficiency and tolerance of magnetotherapy. Arch
1032 Phys Med Rehabil 72, 284-287 (1991).

1033 118. Hussein, A.Z. & Donatelli, R.A. The efficacy of radial extracorporeal shockwave therapy
1034 in shoulder adhesive capsulitis: a prospective, randomised, double-blind, placebo-
1035 controlled, clinical study, . European Journal of Physiotherapy 18, 63-76 (2016).

1036 119. Chen, C.Y., et al. Extracorporeal shockwave therapy improves short-term functional
1037 outcomes of shoulder adhesive capsulitis. J Shoulder Elbow Surg 23, 1843-1851
1038 (2014).

29
1039 120. Zhang, R., et al. Extracorporeal Shockwave Therapy as an Adjunctive Therapy for
1040 Frozen Shoulder: A Systematic Review and Meta-analysis. Orthopaedic Journal of
1041 Sports Medicine (2022).

1042 121. Page, M.J., et al. Electrotherapy modalities for adhesive capsulitis (frozen shoulder).
1043 Cochrane Database Syst Rev, CD011324 (2014).

1044 122. Ramachandran, V.S. & Altschuler, E.L. The use of visual feedback, in particular mirror
1045 visual feedback, in restoring brain function. Brain 132, 1693-1710 (2009).

1046 123. Baskaya, M.C., Ercalik, C., Karatas Kir, O., Ercalik, T. & Tuncer, T. The efficacy of
1047 mirror therapy in patients with adhesive capsulitis: A randomized, prospective,
1048 controlled study. J Back Musculoskelet Rehabil 31, 1177-1182 (2018).

1049 124. Hoffman, M.D. & Hoffman, D.R. Does aerobic exercise improve pain perception and
1050 mood? A review of the evidence related to healthy and chronic pain subjects. Curr Pain
1051 Headache Rep 11, 93-97 (2007).

1052 125. Leon-Latre, M., et al. Sedentary lifestyle and its relation to cardiovascular risk factors,
1053 insulin resistance and inflammatory profile. Rev Esp Cardiol (Engl Ed) 67, 449-455
1054 (2014).

1055 126. Nijs, J., Malfliet, A., Ickmans, K., Baert, I. & Meeus, M. Treatment of central
1056 sensitization in patients with 'unexplained' chronic pain: an update. Expert Opin
1057 Pharmacother 15, 1671-1683 (2014).

1058 127. van der Windt, D.A., van der Heijden, G.J., Scholten, R.J., Koes, B.W. & Bouter, L.M.
1059 The efficacy of non-steroidal anti-inflammatory drugs (NSAIDS) for shoulder
1060 complaints. A systematic review. J Clin Epidemiol 48, 691-704 (1995).

1061 128. D'Orsi, G.M., Via, A.G., Frizziero, A. & Oliva, F. Treatment of adhesive capsulitis: a
1062 review. Muscles Ligaments Tendons J 2, 70-78 (2012).

1063 129. Jain, T.K. & Sharma, N.K. The effectiveness of physiotherapeutic interventions in
1064 treatment of frozen shoulder/adhesive capsulitis: a systematic review. J Back
1065 Musculoskelet Rehabil 27, 247-273 (2014).

1066 130. Lorbach, O., et al. Nonoperative management of adhesive capsulitis of the shoulder:
1067 oral cortisone application versus intra-articular cortisone injections. J Shoulder Elbow
1068 Surg 19, 172-179 (2010).

1069 131. Widiastuti-Samekto, M. & Sianturi, G.P. Frozen shoulder syndrome: comparison of oral
1070 route corticosteroid and intra-articular corticosteroid injection. Med J Malaysia 59, 312-
1071 316 (2004).

1072 132. Hettrich, C.M., et al. The effect of myofibroblasts and corticosteroid injections in
1073 adhesive capsulitis. J Shoulder Elbow Surg 25, 1274-1279 (2016).

1074 133. Sun, Y., et al. Intra-articular Steroid Injection for Frozen Shoulder: A Systematic Review
1075 and Meta-analysis of Randomized Controlled Trials With Trial Sequential Analysis. The
1076 American journal of sports medicine 45, 2171-2179 (2017).

1077 134. Robinson, C.M., Seah, K.T., Chee, Y.H., Hindle, P. & Murray, I.R. Frozen shoulder. J
1078 Bone Joint Surg Br 94, 1-9 (2012).

30
1079 135. Ahn, J.H., et al. Early Intra-articular Corticosteroid Injection Improves Pain and
1080 Function in Adhesive Capsulitis of the Shoulder: 1-Year Retrospective Longitudinal
1081 Study. PM R 10, 19-27 (2018).

1082 136. Olafsen, N.P., Herring, S.A. & Orchard, J.W. Injectable Corticosteroids in Sport. Clin J
1083 Sport Med 28, 451-456 (2018).

1084 137. Walsh, L.J., et al. Adverse effects of oral corticosteroids in relation to dose in patients
1085 with lung disease. Thorax 56, 279-284 (2001).

1086 138. Gaujoux-Viala, C., Dougados, M. & Gossec, L. Efficacy and safety of steroid injections
1087 for shoulder and elbow tendonitis: a meta-analysis of randomised controlled trials. Ann
1088 Rheum Dis 68, 1843-1849 (2009).

1089 139. Lim, T.K., et al. Intra-articular injection of hyaluronate versus corticosteroid in adhesive
1090 capsulitis. Orthopedics 37, e860-865 (2014).

1091 140. Blaine, T., et al. Treatment of persistent shoulder pain with sodium hyaluronate: a
1092 randomized, controlled trial. A multicenter study. J Bone Joint Surg Am 90, 970-979
1093 (2008).

1094 141. Haque, R., Baruah, R.K., Bari, A. & Sawah, A. Is Suprascapular Nerve Block Better
1095 Than Intra-articular Corticosteroid Injection for the Treatment of Adhesive Capsulitis of
1096 the Shoulder? A Randomized Controlled Study. Ortop Traumatol Rehabil 23, 157-165
1097 (2021).

1098 142. Jung, T.W., Lee, S.Y., Min, S.K., Lee, S.M. & Yoo, J.C. Does Combining a
1099 Suprascapular Nerve Block With an Intra-articular Corticosteroid Injection Have an
1100 Additive Effect in the Treatment of Adhesive Capsulitis? A Comparison of Functional
1101 Outcomes After Short-term and Minimum 1-Year Follow-up. Orthop J Sports Med 7,
1102 2325967119859277 (2019).

1103 143. Badalamente, M.A. & Wang, E.D. CORR((R)) ORS Richard A. Brand Award: Clinical
1104 Trials of a New Treatment Method for Adhesive Capsulitis. Clin Orthop Relat Res 474,
1105 2327-2336 (2016).

1106 144. Joo, Y.J., et al. A comparison of the short-term effects of a botulinum toxin type a and
1107 triamcinolone acetate injection on adhesive capsulitis of the shoulder. Ann Rehabil Med
1108 37, 208-214 (2013).

1109 145. Nicholson, J.A., Slader, B., Martindale, A., McKie, S. & Robinson, C.M. Distension
1110 arthrogram in the treatment of adhesive capsulitis has a low rate of repeat intervention.
1111 Bone Joint J 102-B, 606-610 (2020).

1112 146. Wu, W.T., et al. Effectiveness of Glenohumeral Joint Dilatation for Treatment of Frozen
1113 Shoulder: A Systematic Review and Meta-analysis of Randomized Controlled Trials.
1114 Sci Rep 7, 10507 (2017).

1115 147. Catapano, M., Mittal, N., Adamich, J., Kumbhare, D. & Sangha, H. Hydrodilatation With
1116 Corticosteroid for the Treatment of Adhesive Capsulitis: A Systematic Review. PM R
1117 10, 623-635 (2018).

1118 148. Elnady, B., et al. In shoulder adhesive capsulitis, ultrasound-guided anterior
1119 hydrodilatation in rotator interval is more effective than posterior approach: a
1120 randomized controlled study. Clin Rheumatol 39, 3805-3814 (2020).

31
1121 149. Saltychev, M., Laimi, K., Virolainen, P. & Fredericson, M. Effectiveness of
1122 Hydrodilatation in Adhesive Capsulitis of Shoulder: A Systematic Review and Meta-
1123 Analysis. Scand J Surg 107, 285-293 (2018).

1124 150. Ladermann, A., et al. Hydrodilatation with corticosteroids is the most effective
1125 conservative management for frozen shoulder. Knee Surg Sports Traumatol Arthrosc
1126 29, 2553-2563 (2021).

1127 151. Fitzpatrick, J., Richardson, C., Klaber, I. & Richardson, M.D. Clostridium histolyticum
1128 (AA4500) for the Treatment of Adhesive Capsulitis of the Shoulder: A Randomised
1129 Double-Blind, Placebo-Controlled Study for the Safety and Efficacy of Collagenase -
1130 Single Site Report. Drug Des Devel Ther 14, 2707-2713 (2020).

1131 152. Karahan, N., et al. Can Collagenase Be Used in the Treatment of Adhesive Capsulitis?
1132 Med Princ Pract 29, 174-180 (2020).

1133 153. Minagawa, H. Silent manipulation for frozen shoulder. MB Orthop 25, 93-98 (2012).

1134 154. Itoi E, M.H. Manipulation under anesthesia for frozen shoulders. in Shoulder Stiffness:
1135 Current concepts and concerns (ed. Eiji Itoi, G.A., Greg Bain, Ronald Diercks, Dan
1136 Guttman, Andreas Imhoff, Augustus Mazzocca, Hiroyuki Suhaya, Yon-Sik Yoo) 201-
1137 208 (Springer, Berlin, 2015).

1138 155. van de Laar, S.M., van der Zwaal, P. Management of the frozen shoulder. Orthop Res
1139 Rev, 81-90 (2014).

1140 156. Hsu, J.E., Anakwenze, O.A., Warrender, W.J. & Abboud, J.A. Current review of
1141 adhesive capsulitis. J Shoulder Elbow Surg 20, 502-514 (2011).

1142 157. Dodenhoff, R.M., Levy, O., Wilson, A. & Copeland, S.A. Manipulation under anesthesia
1143 for primary frozen shoulder: effect on early recovery and return to activity. J Shoulder
1144 Elbow Surg 9, 23-26 (2000).

1145 158. Reichmister, J.P. & Friedman, S.L. Long-term functional results after manipulation of
1146 the frozen shoulder. Md Med J 48, 7-11 (1999).

1147 159. Arce, G.D.G., G.; Chen, J. The role of arthroscopic capsular release. in Shoulder
1148 Stiffness: Current concepts and concerns (ed. Eiji Itoi, G.A., Greg Bain, Ronald Diercks,
1149 Dan Guttman, Andreas Imhoff, Augustus Mazzocca, Hiroyuki Suhaya, Yon-Sik Yoo)
1150 212-219 (Springer, Berlin, 2015).

1151 160. Barnes, C.P., Lam, P.H. & Murrell, G.A. Short-term outcomes after arthroscopic
1152 capsular release for adhesive capsulitis. J Shoulder Elbow Surg 25, e256-264 (2016).

1153 161. Berghs, B.M., Sole-Molins, X. & Bunker, T.D. Arthroscopic release of adhesive
1154 capsulitis. J Shoulder Elbow Surg 13, 180-185 (2004).

1155 162. Le Lievre, H.M. & Murrell, G.A. Long-term outcomes after arthroscopic capsular release
1156 for idiopathic adhesive capsulitis. J Bone Joint Surg Am 94, 1208-1216 (2012).

1157 163. Segmuller, H.E., Taylor, D.E., Hogan, C.S., Saies, A.D. & Hayes, M.G. Arthroscopic
1158 treatment of adhesive capsulitis. J Shoulder Elbow Surg 4, 403-408 (1995).

1159 164. Warner, J.J., Allen, A., Marks, P.H. & Wong, P. Arthroscopic release for chronic,
1160 refractory adhesive capsulitis of the shoulder. J Bone Joint Surg Am 78, 1808-1816
1161 (1996).
32
1162 165. Diwan, D.B. & Murrell, G.A. An evaluation of the effects of the extent of capsular release
1163 and of postoperative therapy on the temporal outcomes of adhesive capsulitis.
1164 Arthroscopy 21, 1105-1113 (2005).

1165 166. Rangan, A., et al. Management of adults with primary frozen shoulder in secondary
1166 care (UK FROST): a multicentre, pragmatic, three-arm, superiority randomised clinical
1167 trial. Lancet 396, 977-989 (2020). Pivotal randomized controlled trial comparing
1168 MUA, physiotherapy and arthroscopic surgery in frozen shoulder.

1169 167. King, W.V. & Hebron, C. Frozen shoulder: living with uncertainty and being in "no-man's
1170 land". Physiother Theory Pract, 1-15 (2022).

1171 168. Mulligan, E.P., Brunette, M., Shirley, Z. & Khazzam, M. Sleep quality and nocturnal
1172 pain in patients with shoulder disorders. Journal of shoulder and elbow surgery /
1173 American Shoulder and Elbow Surgeons ... [et al.] 24, 1452-1457 (2015).

1174 169. Toprak, M. & Erden, M. Sleep quality, pain, anxiety, depression and quality of life in
1175 patients with frozen shoulder1. J Back Musculoskelet Rehabil 32, 287-291 (2019).

1176 170. Ebrahimzadeh, M.H., Moradi, A., Bidgoli, H.F. & Zarei, B. The Relationship between
1177 Depression or Anxiety Symptoms and Objective and Subjective Symptoms of Patients
1178 with Frozen Shoulder. Int J Prev Med 10, 38 (2019).

1179 171. Lubis, A.M. & Lubis, V.K. Matrix metalloproteinase, tissue inhibitor of metalloproteinase
1180 and transforming growth factor-beta 1 in frozen shoulder, and their changes as
1181 response to intensive stretching and supervised neglect exercise. J Orthop Sci 18, 519-
1182 527 (2013).

1183 172. Xue, M., Gong, S., Dai, J., Chen, G. & Hu, J. The Treatment of Fibrosis of Joint
1184 Synovium and Frozen Shoulder by Smad4 Gene Silencing in Rats. PLoS ONE 11,
1185 e0158093 (2016).

1186 173. Rouhani, A., et al. Calcitonin effects on shoulder adhesive capsulitis. Eur J Orthop Surg
1187 Traumatol 26, 575-580 (2016).

1188 174. Warburton, D.E., Nicol, C.W. & Bredin, S.S. Prescribing exercise as preventive therapy.
1189 CMAJ 174, 961-974 (2006).

1190 175. Gregg, E.W., Gerzoff, R.B., Caspersen, C.J., Williamson, D.F. & Narayan, K.M.
1191 Relationship of walking to mortality among US adults with diabetes. Arch Intern Med
1192 163, 1440-1447 (2003).

1193 176. Dunstan, D.W., et al. The Australian Diabetes, Obesity and Lifestyle Study (AusDiab)-
1194 -methods and response rates. Diabetes Res Clin Pract 57, 119-129 (2002).

1195 177. McLeod, J.C., Stokes, T. & Phillips, S.M. Resistance Exercise Training as a Primary
1196 Countermeasure to Age-Related Chronic Disease. Front Physiol 10, 645 (2019).

1197 178. Maestroni, L., et al. The Benefits of Strength Training on Musculoskeletal System
1198 Health: Practical Applications for Interdisciplinary Care. Sports Med 50, 1431-1450
1199 (2020).

1200 179. Dean, E. & Soderlund, A. What is the role of lifestyle behaviour change associated with
1201 non-communicable disease risk in managing musculoskeletal health conditions with
1202 special reference to chronic pain? BMC Musculoskelet Disord 16, 87 (2015).

33
1203 180. Ramiro, S., et al. The OMERACT Core Domain Set for Clinical Trials of Shoulder
1204 Disorders. J Rheumatol 46, 969-975 (2019).
1205

1206

34
1207 Author contributions
1208 Introduction (N.L.M., G.A.C.M., S.A.R.); Epidemiology (N.L.M., E.W., A.L.C., P.D.K);
1209 Mechanisms/pathophysiology (N.L.M., L.M., S.A.R., M.A., A.L.C. G.A.C.M); Diagnosis,
1210 screening and prevention (A.M, F.S, E.W.); Management (N.L.M., A.L.C., A.M., F.S., E.W.,
1211 P.D.K., G.A.C.M., J.R, S.A.R.); Quality of life (N.L.M., P.D.K.); Outlook (N.L.M., M.A, S.A.R.,
1212 A.L.C.); Overview of Primer (N.L.M.).
1213

1214 Competing interests

1215 The authors declare no competing interests.

1216

1217 Peer review information

1218 Nature Reviews Disease Primers thanks P.J. Millett, C.P. Charalambous, M.-M. Lefevre-
1219 Colau, who co-reviewed with A. Roren, and the other, anonymous reviewers for their
1220 contribution to the peer review of this work.

35
1221 Table 1. Treatment approaches for frozen shoulder based on stage of the condition.
Stage Characteristics Treatment approaches

Pharmacological Physical Other adjuncts

Stage I Inflammation NSAIDs Home exercises Patient education

CSI Hydrodilation
TENS
Stage I Freezing and frozen NSAIDs Physiotherapy: Patient education
Mobilization Hydrodilation
Shockwave
therapy
Stage III Thawing NA Physiotherapy: Surgical
Resistance- management if
based symptoms do not
dissipate
1222 NA, not applicable; NSAIDs, non-steroidal anti-inflammatory drugs; TENS, transcutaneous
1223 electrical nerve stimulation.
1224

36
1225 Figure legends
1226
1227 Figure 1. Structural changes during frozen shoulder. a | The healthy capsule is
1228 collagenous in structure, composed primarily of dense type I collagen and elastic fibre
1229 bundles with limited blood vessels and nerve fibres. The main cell type within this membrane
1230 is the fibroblast, which maintains capsule health by producing extracellular matrix (ECM)
1231 proteins that provide a supportive yet flexible structure. b | In FS, there is fibrosis and
1232 thickening of the connective tissue membrane as well as the adjacent synovial membrane17.
1233 c | Fibroproliferation results in an increased number of fibroblasts producing more ECM
1234 proteins, resulting in a dense and poorly organized fibrillar structure. These fibrotic changes
1235 are accompanied by inflammation, neoangiogenesis and neo-innervation 18,19,20,21. The
1236 consequence is a reduced joint volume and increased stiffness of the capsule, causing
1237 restricted movement and pain.
1238
1239 Figure 2. The stages of frozen shoulder. Frozen shoulder is classified into three clinical
1240 stages based on pain level and the severity of range of motion (ROM) limitation. Stage I is
1241 the inflammatory stage and includes gradually worsening pain but limited effect on ROM.
1242 Stage II involves plateauing of shoulder pain levels but is mostly associated with increasing
1243 stiffness that results in considerable loss of shoulder function that particularly affects a
1244 patients’ normal activities of daily living. Stage III is characterized by reduction of pain
1245 (particularly night pain), with pain usually at the end of ROM, and a very gradual
1246 improvement in stiffness over a number of months to years.
1247
1248 Figure 3. Molecular pathophysiology of frozen shoulder. A trigger, typically systemic (for
1249 example altered metabolic status), extrinsic (for example shoulder immobilization after
1250 trauma or surgery) or intrinsic (rotator cuff pathology) induces a pro-inflammatory, profibrotic
1251 environment in which various soluble factors influence cell behaviour. Substance P induces
1252 production and release of neuropeptides by mast cells, which affects fibroblast activation and
1253 matrix production. Pro-inflammatory cytokines, such as IL-1, IL-6, HMGB1 and IL-17A, and
1254 growth factors stimulate fibroblast activation, proliferation and positive feedback loops driving
1255 further cytokine and growth factor production. Cytokines also induce T cell activation and
1256 production of IL-17A, while the abundance of macrophage subsets, B cells and dendritic cells
1257 are all increased in FS human biopsy samples. All these factors, together with mechanical
1258 stress and matrix turnover imbalance, induce fibroblast transdifferentiation to myofibroblasts,
1259 which leads to tissue fibrosis and contracture.
1260
1261 Figure 4. Proposed algorithm for differential diagnosis of frozen shoulder. This
1262 algorithm may aid in differentiating frozen shoulder from other painful and/or motion-limiting
1263 conditions of the shoulder, such as glenohumeral joint (GHJ) osteoarthritis 48, subacromial
1264 pain and acromioclavicular joint (ACJ) pathology. Decisions are made based predominantly
1265 on physical examination (loss of passive external rotation) but other techniques and methods
1266 also provide useful diagnostic information. SLAP, superior labrum from anterior to posterior.
1267
1268 Figure 5. Key examination techniques for frozen shoulder. Various movements of the
1269 arm of the affected shoulder are used to assess pain and limitations of range of motion
1270 (ROM) in individuals with FS.
1271

37
1272 Box 1. Risk factors for frozen shoulder.
1273 Systemic risk factors
1274 [b1] Diabetes mellitus
1275 [b1] Hypothyroidism
1276 [b1] Hyperthyroidism
1277 [b1] Hypoadrenalism
1278 [b1] Hyperlipidaemia
1279 Extrinsic risk factors
1280 [b1] Cardiopulmonary disease
1281 [b1] Cervical degenerative disc disease
1282 [b1] Cerebrovascular disease
1283 [b1] Humeral fracture
1284 [b1] Parkinson’s disease
1285 [b1] Post axillary surgery (breast carcinoma)
1286 [b1] Radiotherapy
1287 Intrinsic risk factors
1288 [b1] Rotator cuff tendinopathy
1289 [b1] Rotator cuff tears
1290 [b1] Biceps tendinopathy
1291 [b1] Calcific tendinopathy
1292 [b1] Acromioclavicular arthritis
1293
1294

38
1295 Frozen shoulder is a fibroproliferative disorder of the shoulder characterized by pain and
1296 progress loss of shoulder mobility. In this Primer, Millar et al. provide an overview of the
1297 epidemiology, pathophysiology, diagnosis and treatment of FS, as well as how it affects
1298 patients’ quality of life.

39