Ombro congelado Folheto In-depth

O que você está sentindo

O ombro congelado geralmente começa como uma dor profunda e surda que se intensifica ao longo de semanas, em vez de surgir de repente. Para a maioria das pessoas, a dor é o aspecto mais incômodo no início, sendo frequentemente pior à noite. Deitar-se sobre o lado afetado torna-se impossível, e o sono interrompido é um dos aspectos mais exaustivos dessa condição.

Em seguida, o ombro começa a ficar rígido. Você tenta alcançar o cinto de segurança, um objeto numa prateleira alta ou o bolso, mas o braço simplesmente não se move. O que diferencia o ombro congelado da maioria dos outros problemas no ombro é que essa rigidez não se deve apenas à tentativa de evitar a dor — o ombro realmente não se move, mesmo quando outra pessoa tenta levantar o braço por você. A perda da capacidade de virar a mão para fora, como ao abrir uma porta ou enfiar o braço numa manga, é o sinal mais característico de tudo.

A maioria dos pacientes tem entre 40 e 60 anos. Alguns lembram de ter sofrido um pequeno trauma antes; muitos, porém, não têm nenhuma explicação. A condição é mais comum em pessoas com diabetes ou problemas na tireoide; além disso, se já ocorreu num ombro, há uma chance razoável de afetar o outro ombro posteriormente.

O que realmente está acontecendo

A articulação do ombro está inserida num invólucro de tecido resistente chamado cápsula. Normalmente, esse invólucro é frouxo e elástico, o que permite que o ombro se mova mais do que qualquer outra articulação do corpo. Na capsulite adesiva, a cápsula inflama-se, depois engrossa e contrai-se, fazendo com que o invólucro encolha. O espaço interno da articulação pode acabar ficando com menos da metade do seu tamanho normal.

Apesar do nome antigo “capsulite adesiva”, na verdade nada fica colado dentro da articulação. Não existem aderências semelhantes a cola. O tecido simplesmente se tornou mais tenso, o que é um problema completamente distinto; por isso alguns cirurgiões defendem que a condição deveria ser chamada apenas de contratura — ou seja, um estreitamento — do ombro.

Ainda não se compreende bem por que isso acontece, e é justo admitir isso. Existem evidências sólidas de que não se trata apenas de um problema localizado no ombro. Estudos sobre fatores hereditários demonstram que o diabetes tipo 1 realmente provoca a capsulite adesiva, em vez de apenas estar associado a ela; essa relação se dá por meio do nível de glicose no sangue, e não pelo peso ou pela carga mecânica. As alterações tissulares são muito semelhantes às da doença de Dupuytren na mão, razão pela qual ambas costumam aparecer na mesma pessoa.

Sejamos honestos sobre o tempo necessário para a recuperação

É muito provável que você leia que o ombro congelado é uma condição autolimitada, que se resolve completamente em cerca de 18 meses a dois anos. Queremos ser francos com você: essa afirmação é mais otimista do que as evidências científicas permitem.

A ideia de que a doença passa ordenadamente por três fases — “congelamento”, “ombro congelado” e “descongelamento” — vem de um pequeno estudo publicado em 1975, no qual os pacientes mantiveram os braços imobilizados em tipóias por até nove meses. Não é surpresa que os ombros tratados dessa forma tenham ficado rígidos e, posteriormente, recuperado a mobilidade quando o movimento foi permitido. Ao analisar todos os estudos disponíveis, os pesquisadores não encontraram evidências sólidas de que ombros não tratados realmente passem por essas fases e voltem ao estado normal. Na verdade, nenhum estudo demonstrou que as pessoas recuperem, sem tratamento, uma amplitude de movimento verdadeiramente normal, mensurável.

O que as evidências mostram, de forma mais útil, é o seguinte: a maior parte da melhora tende a ocorrer no início; depois, o progresso desacelera e estabiliza. Essa estabilização pode ser permanente. Conforme os diferentes estudos, entre um terço e metade das pessoas ainda apresentam alguma dor ou rigidez anos depois; porém, na maioria dos casos, esses sintomas são leves e não impedem que façam o que desejam.

Isso não significa que você não vá melhorar; a maioria das pessoas melhora bastante. Contudo, dizer “basta esperar, que tudo se resolve sozinho” não é um conselho neutro. Existem motivos válidos para tratar a dor de forma ativa, em vez de simplesmente suportá-la.

O que podemos fazer a respeito

No Mater Private Hospital Rockhampton, o Dr. Kieran Hirpara explicará as opções disponíveis, levando em conta o que mais o incomoda — a dor ou a rigidez — em vez de tentar enquadrá-lo em um estágio numérico.

A injeção de esteroides na articulação é o tratamento com maior respaldo científico para a fase dolorosa. Em diversos estudos, essa é a única intervenção que produz, de forma confiável, uma diferença perceptível na dor e na função do ombro, sendo eficaz já nas primeiras semanas. Sua limitação é que os benefícios tendem a desaparecer após cerca de três meses. O tratamento proporciona alívio da dor, melhora o sono e permite que o paciente volte ao trabalho; porém, não cura a rigidez.

Exercícios e fisioterapia também são válidos, desde que com expectativas realistas. Um programa simples de exercícios em casa traz benefícios mensuráveis. O que talvez surpreenda é que a intensidade dos exercícios faz pouca diferença: um ensaio que comparou alongamentos vigorosos até o limite da amplitude articular com movimentos suaves e indolores encontrou apenas pequenas diferenças, ligeiramente a favor da abordagem mais vigorosa para alguns movimentos, e outro estudo constatou que as pessoas que se exercitaram de forma suave, dentro dos seus limites de dor, tiveram melhores resultados — e, felizmente, não há evidências de que exercícios mais intensos causem danos. Portanto, faça o que seu corpo permitir. Não há comprovação de que suportar dor intensa acelera a recuperação.

A hidrodilatação, procedimento em que se injeta líquido para esticar a cápsula articular a partir do interior, é amplamente utilizada e apresenta baixo risco. Contudo, a literatura científica sobre ela é contraditória: uma revisão encontrou apenas efeitos pequenos tanto na dor quanto na mobilidade, e outras encontraram um ganho inicial em alguns movimentos, em comparação com a injeção de esteroides isolada, que não se manteve. É uma opção razoável, mas não a promovemos como solução definitiva.

A manipulação sob anestesia consiste em movimentar o ombro em toda a sua amplitude enquanto o paciente está dormindo, visando alongar a cápsula rígida. No maior estudo comparativo entre os principais tratamentos, essa técnica se mostrou a melhor opção geral.

A liberação capsular por artroscopia envolve o corte cirúrgico da cápsula rígida sob visão direta, em vez de esticá-la até que se rompa. Seu principal benefício é a durabilidade: nos estudos publicados, a recorrência da rigidez após a liberação é rara, enquanto após a manipulação um número significativo de pacientes volta a apresentar o problema, necessitando de um segundo procedimento. No grande estudo britânico, os pacientes submetidos à liberação foram os menos propensos a precisar de qualquer outro tratamento. Por ser uma cirurgia, apresenta riscos ligeiramente maiores e custos mais elevados; também é normal sentir mais dor nas primeiras semanas em comparação com quem fez manipulação — porém essa diferença tende a desaparecer por volta dos seis meses.

Vale ressaltar: nesse estudo, após um ano todos os três métodos apresentaram resultados muito semelhantes. Isso é uma excelente notícia, pois permite que a decisão seja tomada com base nas preferências do paciente, em suas condições de saúde e em suas circunstâncias pessoais, em vez de haver uma opção claramente superior.

Como o Dr. Hirpara costuma orientar o tratamento. Você não será orientado a suportar a dor por um número fixo de meses antes de receber qualquer intervenção. Na prática, ele geralmente sugere inicialmente a injeção de esteroides, pois para muitos pacientes isso reduz a dor o suficiente para que possam levar a vida normal enquanto o ombro se recupera. Caso a injeção não traga alívio adequado, ele propõe então a liberação capsular — sem exigir que o paciente aguente a dor por nove ou doze meses. Considerando que a ideia de “a condição sempre se resolve sozinha” não tem respaldo científico, não faz sentido passar um ano em dor para descobrir qual será o desfecho no seu caso.

O que esperar

Caso receba uma injeção, espere que a dor diminua durante a primeira ou segunda semana. Aproveite esse período para manter o movimento do ombro.

Se for submetido a uma manipulação ou liberação capsular, a maior parte da amplitude de movimento recuperada ocorre rapidamente — geralmente nas primeiras quatro a seis semanas — e, em geral, não há grandes mudanças após cerca de três meses. É importante começar a movimentar o ombro cedo; a fisioterapia normalmente é iniciada um ou dois dias após o procedimento. A maioria das pessoas retorna ao trabalho de escritório em uma a três semanas.

Se você tem diabetes, é razoável esperar uma recuperação mais lenta e um pouco menos completa do que a média; vale lembrar também que uma injeção de esteroides pode afetar os níveis de açúcar no sangue por alguns dias. Informe-nos caso o controle glicêmico seja difícil, pois isso influencia nossas recomendações.

Independentemente do procedimento escolhido, a recuperação leva meses, não semanas. A melhor coisa que você pode fazer é continuar usando o braço dentro dos limites do conforto e proteger o sono.

Quando procurar ajuda médica

Consulte o seu médico de família se a dor no ombro o acordar à noite ou se não melhorar após algumas semanas de repouso. Pergunte especificamente sobre a capsulite adesiva se o ombro estiver realmente difícil de mover, e não apenas dolorido.

Procure avaliação especializada mais rapidamente se não conseguir virar a mão para fora, se tiver diabetes, se a dor no ombro estiver afetando seu trabalho ou direção de veículos, ou se a dor for tão intensa que o impede de dormir. Não há vantagem em esperar muito tempo para ter a dor avaliada.

Procure atendimento médico imediato se o ombro ficar quente, vermelho e inchado; se você se sentir mal ou febril; se houver fraqueza ou dormência significativa no braço; ou se a dor tiver começado após uma queda ou lesão. Esses sinais indicam que não se trata de capsulite adesiva, mas sim de algo que requer avaliação mais urgente.

Em maior profundidade

Esta seção vai além do que você precisa saber para tomar suas próprias decisões de tratamento. Está aqui porque o ombro congelado é uma das condições em que o que os pacientes são informados e o que a pesquisa realmente demonstra acabaram se distanciando; além disso, algumas pessoas preferem conhecer o raciocínio por trás das recomendações a simplesmente aceitá-las sem questionar.

O que realmente está acontecendo na articulação

A cápsula articular é uma espécie de “manga” de colágeno que envolve a articulação. Na capsulite adesiva, ela fica inflamada e depois fibrosa: células chamadas fibroblastos se multiplicam e produzem novo colágeno de forma densa; algumas dessas células se transformam em miofibroblastos, capazes de contrair-se ativamente. Essa é a principal diferença em relação à cicatrização comum: o tecido não apenas engrossa, mas também se torna mais tenso. O volume da articulação pode cair de um valor normal de 10–15 mL para 3–4 mL.

O mesmo comportamento celular ocorre na doença de Dupuytren na mão; por isso essas duas condições aparecem juntas na mesma pessoa. Por esse motivo, um cirurgião de ombro defendeu que a doença deveria ser simplesmente chamada de contratura do ombro, em vez de “capsulite” [1]. Apesar do antigo nome “capsulite adesiva”, os cirurgiões que examinam o interior da articulação não encontram nenhuma aderência.

Por que a teoria das “três fases” tem menos fundamento do que parece

É comum ler que o ombro congelado passa por fases de “congelamento”, “congelado” e “descongelamento”. Esse modelo foi proposto em 1975. Ao investigarem sua origem, os pesquisadores descobriram que o autor confundiu as referências de dois artigos anteriores; mais importante ainda, seus pacientes mantiveram os braços imobilizados em tipóias por até nove meses. Um ombro imobilizado por tanto tempo torna-se rígido e, ao ser movimentado novamente, volta a ganhar mobilidade. As “fases” podem, portanto, descrever mais o tratamento do que a própria doença.

Uma revisão sistemática realizada para testar esse modelo não encontrou qualquer evidência a seu favor; ao contrário, os resultados foram opostos ao que ele prevê: a maior parte da melhora ocorre no início e depois desacelera, em vez de se intensificar até um “descongelamento” tardio [2]. Além disso, os profissionais da área não conseguem concordar de forma confiável em qual fase um determinado ombro se encontra; os próprios autores da definição padrão admitiram por escrito que os dados necessários para comprová-la “não estão disponíveis”.

A estatística de recuperação e por que existem duas respostas tão diferentes

O número frequentemente citado é que 94% das pessoas se recuperam sem tratamento. Esse dado provém de um estudo real, mas leva em conta apenas os pacientes que não receberam tratamento e retornaram para acompanhamento. A maioria do grupo original optou por tratamento ou desistiu do estudo. Considerando todos os participantes, os mesmos dados indicam cerca de 26% [2]. O termo “recuperado” significava que a amplitude de movimento ficava a até 10 graus da do outro ombro; em alguns desses pacientes, o outro ombro também estava com limitação de movimento.

O maior estudo de acompanhamento de longo prazo constatou que 41% ainda apresentavam sintomas após uma média de quatro anos. O ponto importante a se saber é que a melhora clínica praticamente para após cerca de três anos, em vez de continuar indefinidamente [4].

Trata-se de uma condição que afeta todo o corpo, não apenas o ombro

Um grande estudo genético identificou cinco regiões do DNA associadas ao ombro congelado e, por meio de uma técnica capaz de distinguir causa de coincidência, demonstrou que o diabetes tipo 1 realmente o provoca, em vez de apenas acompanhá-lo. Após levar em conta o diabetes, a relação entre excesso de peso e a condição desapareceu. Isso indica que o nível de açúcar no sangue, e não a carga mecânica, é o fator determinante [5].

O que as evidências da fisioterapia realmente dizem

Esta é a parte mais importante para quem está recebendo ou aplicando fisioterapia. Uma revisão Cochrane reuniu 32 ensaios clínicos envolvendo 1.836 pessoas [6]. Três conclusões se destacam.

Em comparação com uma injeção de esteroides, a fisioterapia manual apresenta piores resultados no início. No ensaio comparativo, seis semanas de mobilização passiva e exercícios supervisionados resultaram em uma melhora da dor 26 pontos menor e uma melhoria da função 25 pontos menor às sete semanas, em comparação com uma única injeção. Quanto à percepção de melhora ou recuperação pelos pacientes, os percentuais foram 46% para a fisioterapia e 77% para a injeção; ou seja, aproximadamente uma pessoa a mais se beneficiou a cada quatro submetidas à injeção. Entretanto, entre seis e doze meses, essa diferença desapareceu e deixou de ser relevante.

O grau de intensidade, a frequência e a técnica utilizadas parecem fazer pouca diferença. Um ensaio com 100 pacientes comparou mobilização vigorosa até o limite da amplitude articular com mobilização suave, sem dor: ambos os grupos melhoraram, e a técnica mais vigorosa obteve resultados estatisticamente melhores para alguns movimentos (elevar o braço para o lado aos três e doze meses, girá-lo para fora aos doze meses), mas os autores consideraram as diferenças pequenas [18]. Um estudo não randomizado com 77 pessoas encontrou o oposto: após dois anos, 89% das que se exercitaram de forma suave, dentro dos seus limites de dor, tinham uma pontuação do ombro normal ou quase normal, contra 63% das que receberam alongamentos intensivos [19]. Outro ensaio com 120 pacientes variou a frequência da mobilização até o limite articular (mais de duas vezes por semana, semanalmente ou menos de uma vez por semana); a diferença na amplitude de abdução foi inferior a oito graus, estatisticamente indistinguível de zero. Comparações diretas entre técnicas — deslizamento anterior versus posterior, exercícios PNF versus convencionais, uso de halteres versus movimentos sem equipamento — também não apresentaram diferenças significativas. Felizmente, não houve indícios de que a mobilização até o limite articular cause danos.

A maior lacuna metodológica: até hoje nenhum ensaio comparou fisioterapia manual associada a exercícios com a ausência total de tratamento. Assim, embora possamos afirmar que a fisioterapia não é claramente superior à injeção e que os detalhes de sua aplicação parecem não alterar os resultados, ninguém demonstrou efetivamente qual parte da melhora seria obtida mesmo sem qualquer intervenção.

Dois pontos importantes merecem atenção. O único ensaio de alta qualidade incluído na revisão analisou a fisioterapia após distensão articular; constatou-se que ela não influenciou a dor, a função ou a qualidade de vida, porém aumentou o percentual de pacientes que se consideraram recuperados e proporcionou um ganho de cerca de 13 graus na amplitude de abdução às seis semanas, efeito que desapareceu aos seis meses. No grande ensaio realizado no Reino Unido, o grupo de fisioterapia foi descrito como recebendo alongamentos manuais até a amplitude limitada; contudo, os registros mostram que a mobilização manual foi utilizada em apenas 11–27% paciente, e o alongamento supervisionado, em apenas um. O que, de fato, foi aplicado — e que produziu resultados comparáveis aos da cirurgia após um ano — foi orientação educativa, uma injeção, movimentos suaves supervisionados e um programa de exercícios domiciliares.

Quanto tempo leva para melhorar, e o tipo de liberação importa?

Tempo de recuperação. Após a liberação, a mediana dos pacientes ultrapassa o limiar de “houve uma melhora real” em cerca de um mês, e o limiar de “meu ombro está aceitável para mim” em aproximadamente quatro meses [14]. O tempo de afastamento do trabalho tem mediana de oito semanas, embora haja grande variação: 25% retornam ao trabalho em quatro semanas, e 75% em treze [15].

A manipulação restaura o movimento mais rapidamente. Esse é o trade-off em relação à maior durabilidade obtida com a liberação: a manipulação restaura mais cedo a amplitude de movimento; a liberação só alcança resultados comparáveis por volta do sexto mês, quando ambas as técnicas tornam-se indistinguíveis em todos os critérios avaliados. Um pequeno estudo comparativo constatou até que a manipulação isolada resultou na melhor rotação externa [16].

A quantidade de cápsula a ser liberada é realmente uma questão relevante; a resposta é “menos do que se imagina”. Uma revisão que analisou 18 estudos, 629 pacientes e 811 ombros comparou três abordagens: liberar apenas a parte frontal e inferior da cápsula, acrescentar a parte posterior ou realizar uma liberação completa em 360 graus. Concluiu-se que liberações menos extensas geram melhores resultados funcionais e menores índices de dor; a liberação posterior proporciona ganhos iniciais em rotação interna, porém temporários (embora aumente de forma duradoura a flexão); já a liberação total “pode não trazer nenhum benefício adicional”. As taxas de complicações não diferiram entre as três técnicas [17]. Dois ensaios randomizados sobre a extensão posterior apresentaram o mesmo padrão: recuperação inicial mais rápida, sem diferenças a partir do sexto mês. Portanto, uma cirurgia mais extensa não significa, necessariamente, melhores resultados.

Realizar ambas as intervenções simultaneamente não é a solução. No único estudo que comparou as três abordagens, a combinação de manipulação e liberação resultou na maior taxa de perda de rotação externa: 18%, contra 7% na manipulação isolada e 2% na liberação isolada [16].

O que o maior estudo clínico sobre tratamentos realmente revelou

Um estudo realizado no Reino Unido randomizou pouco mais de 500 pacientes para receber fisioterapia estruturada associada a injeção, manipulação sob anestesia ou cirurgia minimamente invasiva. Após um ano, os resultados dos três grupos ficaram dentro de poucos pontos uns dos outros, valor inferior à diferença que um paciente seria capaz de perceber [7]. Essa é, de fato, uma informação muito útil: significa que a escolha do tratamento pode ser feita com base no que melhor se adapta ao paciente e às suas circunstâncias, em vez de haver uma opção claramente superior às demais.

As diferenças entre os três métodos aparecem no número de vezes em que os pacientes precisam de tratamentos adicionais posteriormente. Nesse estudo, foi necessário tratamento complementar para 4% dos pacientes após a liberação ligamentar, 7% após a manipulação e 15% após a fisioterapia [7]. A manipulação sob anestesia apresentou o melhor custo-benefício; a liberação ligamentar foi a opção mais cara e gerou a maior parte das complicações graves observadas no estudo, embora essas fossem raras em termos absolutos.

Manipulação ou liberação — a questão da recorrência

É aqui que as duas intervenções realmente se diferenciam, e é esse o motivo pelo qual o cirurgião pode indicar a liberação mesmo quando os resultados gerais parecem semelhantes.

A manipulação estica a cápsula até provocar uma ruptura; porém, essa ruptura nem sempre ocorre onde se deseja. Na maior série de estudos, 792 ombros foram acompanhados por 17 anos — em 17,8% dos casos foi necessária uma segunda manipulação; entre pessoas com diabetes tipo 1, esse percentual subiu para 37,9% [9]. O acompanhamento mais longo, com média de 13 anos, revelou que 19,2% precisaram repetir a manipulação e 31,3% relataram recorrência em algum momento [10]. Dois fatores atenuam esse cenário: quase todas as recorrências ocorrem no primeiro ano (após cinco anos, a taxa cai para menos de 2%); além disso, uma nova manipulação tem eficácia semelhante à primeira [9][10].

A liberação divide a cápsula de forma intencional, sob visão direta. A necessidade de reoperação é bem menor: num estudo com acompanhamento médio de cinco anos, nenhum dos 32 pacientes precisou de nova cirurgia, havendo apenas um caso de recorrência em alguém que não completou a reabilitação [11]. A meta-análise que reuniu oito estudos e 768 pacientes considerou as duas técnicas comparáveis quanto à dor, função e amplitude de movimento (uma pequena vantagem da liberação na dor a longo prazo não atingiu uma magnitude clinicamente importante), com uma taxa maior de complicações graves após a liberação. Ela observou uma taxa maior, embora não estatisticamente significativa, de novas injeções de esteroides no grupo submetido à manipulação, e sugeriu que, em acompanhamentos mais longos, “o grupo submetido à manipulação pode apresentar mais recorrências” [8].

Dois pontos importantes devem ser considerados. Primeiro, os principais estudos sobre recorrência após manipulação vêm, em grande parte, da prática de um único cirurgião, portanto não constituem confirmações independentes entre si. Segundo, a liberação também não garante recuperação total: um estudo constatou que, dois a seis anos depois, a pontuação média do ombro ainda era de cerca de 77 de 100 [12]; além disso, num grupo que excluiu diabéticos, embora a maioria tenha apresentado melhora significativa, 11–22% ainda consideravam seu ombro insatisfatório após seis meses [13].

Em resumo: para tornar o ombro móvel e confortável, ambas as técnicas são praticamente equivalentes, sendo a manipulação a opção mais econômica. Porém, para manter esses resultados a longo prazo, as evidências favorecem a liberação. Essa diferença é especialmente relevante em pacientes diabéticos, naqueles com ombro muito rígido ou em quem já fez manipulação sem sucesso.

Referências

Advanced reading: the deeper science (optional)

[1] Bunker T. É hora de um novo nome para o ombro congelado: contratura do ombro. Shoulder Elbow. 2009;1(1):4-9. https://doi.org/10.1111/j.1758-5740.2009.00007.x [2] Wong CK, Levine WN, Deo K, Kesting RS, Mercer EA, Schram GA, et al. Histórico natural do ombro congelado: fato ou ficção? Uma revisão sistemática. Physiotherapy. 2017;103(1):40-7. https://doi.org/10.1016/j.physio.2016.05.009 [3] Vastamäki H, Kettunen J, Vastamäki M. Histórico natural do ombro congelado idiopático: estudo de acompanhamento de 2 a 27 anos. Clin Orthop Relat Res. 2012;470(4):1133-43. https://doi.org/10.1007/s11999-011-2176-4 [4] Hand C, Clipsham K, Rees JL, Carr AJ. Resultados a longo prazo do ombro congelado. J Shoulder Elbow Surg. 2008;17(2):231-6. https://doi.org/10.1016/j.jse.2007.05.009 [5] Green HD, Jones A, Evans JP, Wood AR, Beaumont RN, Tyrrell J, et al. Um estudo de associação genômica ampla identifica 5 loci associados ao ombro congelado e sugere o diabetes como fator de risco causal. PLoS Genet. 2021;17(6):e1009577. https://doi.org/10.1371/journal.pgen.1009577 [6] Page MJ, Green S, Kramer S, Johnston RV, McBain B, Chau M, et al. Terapia manual e exercícios para capsulite adesiva (ombro congelado). Cochrane Database Syst Rev. 2014;2014(8):CD011275. https://doi.org/10.1002/14651858.CD011275 [7] Brealey S, Northgraves M, Kottam L, Keding A, Corbacho B, Goodchild L, et al. Tratamentos cirúrgicos comparados à fisioterapia estruturada precoce no atendimento secundário para adultos com ombro congelado primário: o ensaio clínico randomizado de três braços UK FROST. Health Technol Assess. 2020;24(71):1-162. https://doi.org/10.3310/hta24710 [8] Zhao Y, Yang T, Feng C, Li L, Pang L, Zhao S. Liberação capsular artroscópica versus manipulação sob anestesia para ombro congelado refratário: uma revisão sistemática com meta-análise. Orthop Surg. 2024;16(7):1517-29. https://doi.org/10.1111/os.14077 [9] Woods DA, Loganathan K. Recorrência do ombro congelado após manipulação sob anestesia (MUA). Bone Joint J. 2017;99-B(6):812-7. https://doi.org/10.1302/0301-620X.99B6.BJJ-2016-1133.R1 [10] Fairclough A, Waters C, Davies T, Dacombe P, Woods D. Resultados a longo prazo após manipulação sob anestesia em pacientes com ombro congelado primário e secundário. Shoulder Elbow. 2023;15(2):173-80. https://doi.org/10.1177/17585732211070007 [11] Ranalletta M, Rossi LA, Zaidenberg EE, Bertona A, Tanoira I, Maignon GD, et al. Resultados a médio prazo após liberação capsular anteroinferior artroscópica no tratamento da capsulite adesiva idiopática. Arthroscopy. 2017;33(3):503-8. https://doi.org/10.1016/j.arthro.2016.08.024 [12] Mardani-Kivi M, Hashemi-Motlagh K, Darabipour Z. Liberação artroscópica na capsulite adesiva do ombro: estudo retrospectivo com 2 a 6 anos de acompanhamento. Clin Shoulder Elb. 2021;24(3):172-7. https://doi.org/10.5397/cise.2021.00311 [13] Pasqualini I, Tanoira I, Hurley ET, Ranalletta M, Rossi LA. A liberação capsular artroscópica proporciona resultados clinicamente significativos em pacientes com capsulite adesiva. Arthroscopy. 2024;40(4):1081-8. https://doi.org/10.1016/j.arthro.2023.08.083 [14] Pasqualini I, Rossi LA, Oyem PC, Tanoira I, Ranalletta M. Tempo necessário para alcançar resultados clinicamente significativos após liberação capsular artroscópica. Orthop J Sports Med. 2024;12(11):23259671241275653. https://doi.org/10.1177/23259671241275653 [15] Sedlinsch A, Berndt T, Rühmann O, Lerch S. Convalescença após liberação capsular artroscópica no ombro congelado. J Orthop. 2020;20:374-9. https://doi.org/10.1016/j.jor.2020.06.013 [16] Schoch B, Huttman D, Syed UA, et al. Tratamento cirúrgico da capsulite adesiva: comparação entre manipulação, liberação capsular e ambas as abordagens. Cureus. 2020;12(7):e9032. https://doi.org/10.7759/cureus.9032 [17] Sivasubramanian H, Chua CXK, Lim SY, Manohara R, Ng ZWD, Prem Kumar V, et al. Liberação capsular artroscópica para tratamento do ombro congelado idiopático: qual o grau necessário de liberação? Orthop Traumatol Surg Res. 2021;107(1):102766. https://doi.org/10.1016/j.otsr.2020.102766 [18] Vermeulen HM, Rozing PM, Obermann WR, le Cessie S, Vliet Vlieland TPM. Comparação entre técnicas de mobilização de alto grau e de baixo grau no tratamento da capsulite adesiva do ombro: ensaio clínico randomizado. Phys Ther. 2006;86(3):355-68. https://pubmed.ncbi.nlm.nih.gov/16506872/ [19] Diercks RL, Stevens M. Descongelamento suave do ombro congelado: estudo prospectivo de negligência supervisionada versus fisioterapia intensiva em setenta e sete pacientes com síndrome do ombro congelado acompanhados por dois anos. J Shoulder Elbow Surg. 2004;13:499-502. https://doi.org/10.1016/j.jse.2004.03.002

Evidence & references

This is the clinical evidence summary written for health professionals. It is technical, and it lists the research this page was built from. You do not need to read it to understand your treatment or to make a decision about it.

Overview

Frozen shoulder management guidelines provide evidence-based guidance and identify key areas for future research [2].
Magnetic resonance imaging findings in frozen shoulder should not replace clinical judgments regarding prognosis and treatment decisions [7].
Clinicians should monitor frozen shoulder patients with diabetes more closely and offer further treatment if pain or lack of function persist long-term [8].
Different treatment strategies for frozen shoulder may be appropriate depending on the location [9].
A strict definition of recalcitrant idiopathic frozen shoulder is needed to prevent unnecessary interventions, as current data should not be interpreted as a plea for surgery in patients with a mean symptom duration of 4 months [10].
The variety of participants, methods, interventions, and outcomes across trials provides limited new evidence to inform the non-surgical management and treatment of frozen shoulder [14].
Patients with a poor outcome or recurrent symptoms after manipulation under anaesthetic (MUA) should be offered a further MUA, with an expectation of a good outcome and a low complication rate [17].
Early surgical intervention might shorten the overall duration of symptoms in frozen shoulder and is not associated with inferior clinical outcomes compared with late surgical intervention [19].
There is insufficient evidence to reliably recommend a single treatment approach for frozen shoulder [32].
Further treatment was indicated in 41% of patients who could not tolerate more than 20 mL of injection during hydrodilatation [34].
Recurrence of frozen shoulder was more common in primary (33%) versus secondary (16%) cases [34].
Arthroscopic capsular release is a suitable option for refractory primary frozen shoulder syndrome, leading to faster and long-lasting recovery [35].
Manipulation under anaesthetic (MUA) and arthroscopic capsular release (ACR) are both good treatment options for primary frozen shoulder [70].
MUA is simpler but carries risks of serious complications, whereas ACR may be safer if performed by experienced surgeons and is convenient for patients with combined rotator cuff tears [70].

Anatomy & Pathophysiology

Primary (idiopathic) frozen shoulder consists of three phases: pain, stiffness, and thawing [3].
Secondary frozen shoulders may not exhibit all three phases and may not follow the exact chronology of primary frozen shoulder [3].
The pain phase of primary frozen shoulder involves a gradual onset of diffuse shoulder pain that is progressive over weeks to months [3].
Pain in the initial phase is usually worse at night and exacerbated by lying on the affected side [3].
Reduced arm use due to pain leads to stiffness [3].
The stiffness phase begins when patients restrict movement to seek pain relief and usually lasts 4 to 12 months [3].
During the stiffness phase, patients experience difficulty with activities of daily living, such as men accessing back pockets or women fastening brassieres [3].
A dull ache is present nearly all time during the stiffness phase, especially at night, often accompanied by sharp pain during range of motion at or near new endpoints [3].
The thawing phase lasts for weeks or months, during which motion increases and pain diminishes [3].
Without treatment, motion return in the thawing phase is gradual and may never objectively return to normal, although most patients subjectively feel near normal [3].
Frozen shoulder is characterized by progressive pain and stiffness that usually resolves spontaneously after about 18 months [4].
Histological features of frozen shoulder are reminiscent of Dupuytren’s disease, with active fibroblastic and myofibroblastic proliferation in the rotator interval, anterior capsule, and coracohumeral ligament [4].
Conditions particularly associated with frozen shoulder include diabetes, Dupuytren’s disease, hyperlipidaemia, hyperthyroidism, cardiac disease, and hemiplegia [4].
The etiology of frozen shoulder is still not known and understanding of the pathogenesis is limited [15].
Frozen shoulder is a poorly understood condition that typically involves substantial pain, movement restriction, and considerable morbidity [12].
The pathophysiology of frozen shoulder is complex, involving various pathophysiologic mechanisms [62].
A systematic review summarizes the tissue pathophysiology of primary frozen shoulder [25].
Primary frozen shoulder is defined by total elevation restricted to 135° or less [20].
In primary frozen shoulder, motion restriction is localized to the humero-scapular joint [20].
Primary frozen shoulder excludes cases with post-traumatic conditions, rheumatoid arthritis, osteoarthritis, hemiplegia, or other obvious changes explaining the range of motion decrease [20].
Secondary frozen shoulder involves decreased range of motion following a traumatic lesion, including soft tissue injury, intra- and juxtaarticular fractures, and other upper limb fractures [20].
Symptomatic subjects with frozen shoulder demonstrate substantial kinematic deficits during humeral range of motion [44].
Patients with frozen shoulder present with altered shoulder muscle activity and kinematics [78].
Pathomechanics of frozen shoulder are characterized by glenohumeral motion limitations, high tension in the anteroinferior glenohumeral capsule, and altered scapular motion [75].
The scapulohumeral rhythm (SHR) of the affected shoulder is inversely related to the severity of limitation of shoulder range of motion, suggesting a compensatory pattern [80].
The anatomical structure of passive shoulder restraints has no impact on the difference in passive joint position sense values between external and internal rotation in frozen shoulder [59].
The thickness of the inferior glenohumeral joint capsule in the 80° scapular plane elevated arm position is a highly reliable and valid method for assessment [79].
Imaging is an essential tool for evaluation of patients with shoulder pain to understand the extent of injury [37].
X-rays in frozen shoulder are normal, and their main role is to exclude other causes of pain and stiffness [4].
Post-traumatic stiffness may persist for some months after severe shoulder injury but is maximal at the start and gradually lessens, unlike the pattern of frozen shoulder [4].
Disuse stiffness can occur if the arm is nursed overcautiously, such as following a wrist fracture, but lacks the characteristic pain pattern of frozen shoulder [4].
Complex regional pain syndrome may follow acute trauma or be seen in patients with myocardial infarction or stroke, presenting features similar to frozen shoulder [4].
Rheumatoid arthritis and osteoarthritis can affect the shoulder and develop bilaterally, with diagnosis usually obvious on X-ray [4].
Shoulder stiffness is a typical outcome after bone or soft tissue injuries around the shoulder, including contusions, subluxations, dislocations, acromioclavicular joint injuries, clavicle and scapula fractures, and proximal humerus fractures in the elderly [61].
Repetitive, low-level trauma can cause localized contractures leading to motion loss in specific patterns [61].
Isolated posterior capsular contracture is the most commonly described localized contracture causing motion loss [61].
Surgical procedures such as anterior or posterior capsulorrhaphy, inferior capsular shift, and rotator cuff surgery can result in limitation of motion [61].

Classification

Frozen shoulder is a specific, painful, and debilitating condition affecting patients mainly in middle age [6].
Most patients clinically diagnosed with primary frozen shoulder had undiagnosed systemic abnormalities and/or intra-articular pathologies [5].
18F-FDG PET/CT is clinically relevant in diagnostically challenging cases, such as distinguishing the first phase of frozen shoulder from subacromial impingement [11].
Health professionals manage frozen shoulders differently for different phases of the condition [16].
Primary frozen shoulders are defined by total elevation restricted to 135° or less [20].
In primary frozen shoulder, the restriction of motion is localized to the humero-scapular joint [20].
Primary frozen shoulder is diagnosed when no findings in case history, clinical examination, or radiological examination explain the decrease in range of motion [20].
Cases with post-traumatic conditions, rheumatoid arthritis, osteoarthritis, hemiplegia, and other obvious changes are excluded from primary frozen shoulder classification [20].
Secondary frozen shoulder is characterized by decreased range of motion following a traumatic lesion [20].
Associated injuries in secondary frozen shoulder include soft tissue injury to the shoulder region, intra- and juxtaarticular fractures, and other fractures of the upper limb [20].
Frozen shoulder is classified into three consecutive stages according to Reeves [20].
Stage 1 of frozen shoulder is characterized by pain [20].
Stage 1 of frozen shoulder has a duration of 10 to 36 weeks [20].
In Stage 1 of frozen shoulder, there is no difference between men and women, no difference between affected dominant and nondominant shoulders, and no correlation with age [20].
In the early stages of frozen shoulder, there is a full range of movement under anesthesia [20].
Classification of frozen shoulder was found to be controversial [38].
There are lower rates of agreement among Japan Shoulder Society (JSS) members than American Shoulder and Elbow Surgeons (ASES) members for the definition of primary frozen shoulder [43].
There are lower rates of agreement among JSS members than ASES members for the classification of primary and secondary frozen shoulder [43].
There are lower rates of agreement among JSS members than ASES members for the divisions of secondary frozen shoulder [43].
There is disagreement among shoulder specialists regarding terminology for frozen shoulder [57].

Clinical Presentation

Frozen shoulder is a specific, painful, and debilitating condition that primarily affects patients in middle age [6].
Frozen shoulder is a common disease that causes significant morbidity [21].
Frozen shoulder is characterized by severe shoulder pain and functional restriction [24].
Frozen shoulder has considerable economic impact [28].
Frozen shoulder typically involves substantial pain, movement restriction, and considerable morbidity [12].
Frozen shoulder affects approximately 4% of the general population [24].
Frozen shoulder affects up to 59% of patients with diabetes mellitus [24].
The disease duration of frozen shoulder varies between 1 and 3 years [24].
The clinical course of primary (idiopathic) frozen shoulder consists of three phases: pain, stiffness, and thawing [3].
Primary frozen shoulder is characterized by progressive pain and stiffness which usually resolves spontaneously after about 18 months [4].
In Phase I (Pain), patients experience a gradual onset of diffuse shoulder pain that is progressive over weeks to months [3].
The pain in Phase I is usually worse at night and exacerbated by lying on the affected side [3].
In Phase II (Stiffness), patients restrict movement to seek pain relief, leading to stiffness that usually lasts 4 to 12 months [3].
During Phase II, patients have difficulty with activities of daily living, such as men reaching back pockets or women fastening brassieres [3].
In Phase II, a dull ache is present nearly all the time, especially at night, often accompanied by sharp pain during range of motion at or near new endpoints [3].
In Phase III (Thawing), motion increases and pain diminishes over weeks or months [3].
Without treatment, motion return in Phase III is gradual and may never objectively return to normal, though patients often subjectively feel near normal [3].
The natural history of frozen shoulder involves pain that increases in severity, prevents sleeping on the affected side, and subsides after several months [4].
As pain subsides in frozen shoulder, stiffness becomes more prominent and persists for another 6–12 months if untreated [4].
Movement is gradually regained in frozen shoulder but may not return to normal [4].
Patients with frozen shoulder are typically aged 40–60 years [4].
Patients with frozen shoulder may give a history of trauma, often trivial, followed by pain [4].
Physical examination of frozen shoulder usually reveals slight muscle wasting and tenderness [4].
Movements in frozen shoulder are always limited, and the shoulder may be extremely stiff in severe cases [4].
X-rays in frozen shoulder are normal; their main role is to exclude other causes of pain and stiffness [4].
Post-traumatic stiffness is characterized by stiffness without much pain that is maximal at the start and gradually lessens, unlike the pattern of frozen shoulder [4].
Disuse stiffness occurs if the arm is nursed overcautiously and lacks the characteristic pain pattern of frozen shoulder [4].
Complex regional pain syndrome may follow acute trauma or occur in patients with myocardial infarction or stroke, presenting features similar to frozen shoulder [4].
Rheumatoid arthritis and osteoarthritis can affect the shoulder bilaterally and are usually obvious on X-ray [4].
Primary frozen shoulder excludes cases with post-traumatic conditions, rheumatoid arthritis, osteoarthritis, hemiplegia, or other obvious changes in history or examination [20].
Secondary frozen shoulder involves decreased range of motion following a traumatic lesion, such as soft tissue injury or fractures of the upper limb [20].
Stage 1 (pain) of frozen shoulder lasts 10 to 36 weeks [20].
In Stage 1 of frozen shoulder, there is no difference in presentation between men and women, affected dominant and nondominant shoulders, or correlation with age [20].
Frozen shoulder following COVID-19 vaccination may present with clinical features similar to idiopathic frozen shoulder [39].
Frozen shoulder can occur after COVID-19 vaccination [41].
Frozen shoulder is particularly associated with diabetes, Dupuytren’s disease, hyperlipidaemia, hyperthyroidism, cardiac disease, and hemiplegia [4].
MR findings in frozen shoulder should not replace clinical judgments regarding further prognosis and treatment decisions [7].
Unusual stiffness and pain in the shoulder of a young female patient suggests a wide range of disease entities, including infection [23].
Misdiagnosing shoulder tumors as frozen shoulder syndrome is likely to cause a significant delay in making a correct diagnosis [40].
A detailed clinical history and examination is critical in the assessment of a painful, stiff shoulder to avoid misdiagnosis of conditions like malignant shoulder girdle tumours [1].

Investigations

A detailed clinical history and examination is critical in the assessment of a painful, stiff shoulder [1].
The term frozen shoulder should be reserved for a well-defined disorder characterized by progressive pain and stiffness which usually resolves spontaneously after about 18 months [4].
The histological features of frozen shoulder are reminiscent of Dupuytren’s disease, with active fibroblastic and myofibroblastic proliferation in the rotator interval, anterior capsule and coracohumeral ligament [4].
Patients may give a history of trauma, often trivial, followed by pain that gradually increases in severity and often prevents sleeping on the affected side [4].
Pain begins to subside after several months, but stiffness becomes more problematic; untreated stiffness persists for another 6–12 months before movement is gradually regained, which may not return to normal [4].
Clinical examination usually reveals slight muscle wasting and tenderness, with always limited movements and extreme stiffness in severe cases [4].
X-rays are normal in frozen shoulder, and their main role is to exclude other causes of pain and stiffness [4].
Disuse stiffness may occur if the arm is nursed overcautiously, such as following a wrist fracture, and lacks the characteristic pain pattern of frozen shoulder [4].
Complex regional pain syndrome may follow acute trauma or be seen in patients with myocardial infarction or stroke, and its features can be similar to those of frozen shoulder [4].
Both rheumatoid arthritis and osteoarthritis can affect the shoulder, with diagnosis usually obvious on X-ray, and rheumatoid arthritis may present with characteristic generalized symptoms and signs [4].
Unusual stiffness and pain in the shoulder of a young female patient suggests a wide range of disease entities, from simple frozen shoulder to infection [23].
Routine use of shoulder MRI scans in patients with frozen shoulder but without suspicion of additional pathology may not be indicated [31].
Imaging is an essential tool for evaluation of patients with shoulder pain, and understanding the extent of an injury with imaging is key to successful management [37].
Physicians should re-examine frozen shoulder patients with repeated plain radiographs and further imaging, especially MRI, if conservative therapy fails [56].
MR arthrography reveals characteristic findings in patients with frozen shoulder [58].
T2 signal hyperintensity and axillary capsule thickening are characteristic of the early stages of frozen shoulder, although MRI alone cannot completely define the disease stage [71].
The burning sign is an abnormal finding that appears in dynamic MRI of severe frozen shoulder [73].
Dynamic MRI semiquantitatively demonstrated a reduction in abnormal blood flow and improvement in clinical results after manipulation under cervical nerve root block (MUC) in patients with frozen shoulder [77].
Clinical improvement in patients with frozen shoulder was associated with a decrease in the coefficient of enhancement (CE) in the glenohumeral synovium following intraarticular injections of hyaluronate [81].
There may be a causal relationship between hypothyroidism and frozen shoulder [85].
Patients who underwent image-guided (ultrasound) injections had statistically significant greater improvement in shoulder pain and function at 6 weeks after injection compared to blind injections [86].
MR imaging of patients with severe frozen shoulder after MUC showed 29 capsule tears, 4 labrum tears, and 15 bone bruises of the humeral head [87].

Treatment

Non-Operative Management

Treatment for frozen shoulder aims to improve pain and function through a shared decision-making process, utilizing a step-up approach from conservative measures to invasive treatments if symptoms persist [28].
Health professionals manage frozen shoulders differently depending on the phase of the condition [16].
There is limited evidence of the effectiveness of different forms of treatment used for frozen shoulder, with many studies carrying a moderate to high risk of bias and omitting details of symptom duration or condition phase [74].
The variety of participants, methods, interventions, and outcomes across trials provides limited new evidence to inform non-surgical management [14].
Conservative treatment is effective for the treatment of frozen shoulder regardless of the severity of symptoms [45].
Intra-articular steroid injection is effective and safe for frozen shoulder, relieving pain, improving functional performance, and increasing range of motion [48].
Multisite corticosteroid injection therapy is more effective than single intra-articular injection in terms of pain relief, restoration of motion, and functional status for primary frozen shoulder [46].
Both multisite and single glenohumeral injections of corticosteroid are effective in patients with primary frozen shoulder [49].
Ultrasound-guided administration of corticosteroid injections into the glenohumeral joint increases the likelihood of successful injection but does not improve clinical outcomes at 12 weeks compared with blind administration [52].
Ultrasound-guided multisite injection is a nonsurgical treatment technique for frozen shoulder that emphasizes correct injection sites [72].
Hydrodilatation has emerged as a potential nonsurgical option in the management of frozen shoulder, but its role has yet to be fully clarified [60].
Management of frozen shoulder stage II to III using hydrodistension and a guided exercise programme by physiotherapists in primary care is an effective non-operative treatment strategy [66].
Recurrence of frozen shoulder after hydrodilatation was more common in primary (33%) versus secondary (16%) frozen shoulder [34].
A central nervous system-focused treatment approach is being evaluated in a randomized clinical trial to compare effectiveness against standard medical and physical therapy care [42].
The comparative effectiveness of low-level laser therapy versus muscle energy technique is being evaluated in a randomized controlled trial to determine the optimal treatment approach for frozen shoulder related to diabetes [47].

Operative and Procedural Management

Secondary frozen shoulder may be more recalcitrant to conventional conservative treatment [64].
The study on arthroscopic capsular release timing adds information for shared decision-making but should not be a plea for surgery for patients with a mean duration of symptoms of 4 months; a strict definition of recalcitrant idiopathic frozen shoulder is needed to prevent unnecessary interventions [10].
The arthroscopic 360° release is an effective and safe treatment modality for severe or recalcitrant frozen shoulder [53].
A combination of limited capsular release and manipulation under anaesthesia (MUA) for primary frozen shoulder is a safe and effective procedure resulting in marked improvement in pain, function, and range of motion [55].
The mid-term outcomes of transcatheter arterial micro embolization (TAME) for frozen shoulders resistant to conservative treatments are encouraging and warrant further evaluation [50].
Manipulation under anesthesia versus physiotherapy treatment in stage two of a frozen shoulder is being evaluated in a randomized controlled trial to provide evidence on the best treatment strategy [51].
Different treatment strategies for frozen shoulder may be appropriate depending on the location of pathology [9].

Complications

Frozen shoulder is a painful and debilitating condition [6].
Frozen shoulder causes significant morbidity [21].
The natural history of primary frozen shoulder consists of three phases: pain, stiffness, and thawing [3].
Primary frozen shoulder usually resolves spontaneously after about 18 months [4].
Without treatment, motion return in primary frozen shoulder is gradual and may never objectively return to normal [3].
In the long term, 41% of patients with frozen shoulder report some ongoing symptoms [30].
Postoperative frozen shoulder is a serious complication after shoulder surgery with an incidence of 11% [54].
Preoperative frozen shoulder negatively affects functional outcomes, including range of motion, at 6 months and 1 year postoperatively following arthroscopic rotator cuff repair [33].
Preoperative frozen shoulder positively affects rotator cuff healing [33].
Patients with diabetes should be monitored more closely for frozen shoulder, as they may experience persistent pain or lack of function long-term [8].
An age between 46 and 60 years is a statistically significant risk factor for developing frozen shoulder after simple arthroscopic shoulder procedures [29].
A previous history of contralateral frozen shoulder is a statistically significant risk factor for developing frozen shoulder after simple arthroscopic shoulder procedures [29].
Most patients clinically diagnosed with primary frozen shoulder have undiagnosed systemic abnormalities and/or intra-articular pathologies [5].
Recurrence of frozen shoulder after manipulation under anaesthetic (MUA) can occur, but late recurrence is uncommon [22].
Patients with poor outcomes or recurrent symptoms after MUA can be offered a further MUA with an expectation of good outcome and low complication rate [17].

Recovery

Phase I (Pain) involves a gradual onset of diffuse shoulder pain that is progressive over weeks to months [3].
Pain in Phase I is usually worse at night and exacerbated by lying on the affected side [3].
Phase II (Stiffness) usually lasts 4 to 12 months [3].
During Phase II, patients restrict movement to seek pain relief, leading to difficulty with activities of daily living such as reaching back pockets or fastening brassieres [3].
Phase II is accompanied by a dull ache present nearly all the time, especially at night, and sharp pain during range of motion at or near new endpoints [3].
Phase III (Thawing) lasts for weeks or months, during which motion increases and pain diminishes [3].
Without treatment, motion return in Phase III is gradual in most patients [3].
Objective motion may never return to normal, although most patients subjectively feel near normal due to compensation or adjustment in activities of daily living [3].
Frozen shoulder affects approximately 4% of the general population and up to 59% in patients with diabetes mellitus [24].
In the long term, 59% of patients with frozen shoulder have normal or near normal shoulders [30].
94% of patients with spontaneous frozen shoulder recover to normal levels of function and motion without treatment [69].
Idiopathic frozen shoulder is a self-limiting condition in which symptoms subside and full shoulder movement returns within a maximum of two years from the onset of symptoms [88].
Long-term outcome after manipulation under anaesthetic (MUA) for frozen shoulder is favourable with late recurrence being uncommon [22].
Patients with poor outcome or recurrent symptoms after MUA should be offered a further MUA with the expectation of a good outcome and a low complication rate [17].
Arthroscopic capsular release leads to a faster and long-lasting recovery in patients with refractory primary frozen shoulder syndrome [35].
The long-term results of arthroscopic capsular release in frozen shoulder were confirmed in 255 patients [36].
MUA in stage 2 frozen shoulder results in a faster recovery of range of motion and improved functional outcome compared to physiotherapy alone in the short term [68].
MUA in stage 2 frozen shoulder can be considered safe compared to physiotherapy alone in the short term [68].
Early surgical intervention might shorten the overall duration of symptoms in frozen shoulder [19].
Early surgical intervention is not associated with inferior clinical outcomes when compared with late surgical intervention [19].
Timing has a significant influence on the outcome of manipulation for frozen shoulders [90].
Diabetes is a poor prognostic factor for the conservative treatment of frozen shoulder [82].
Male sex is a poor prognostic factor for the conservative treatment of frozen shoulder [82].
Simultaneous bilateral involvement is a poor prognostic factor for the conservative treatment of frozen shoulder [82].
Subsequent bilateral involvement is a poor prognostic factor for the conservative treatment of frozen shoulder [82].
A longer duration of symptoms recorded at the first visit is a poor prognostic factor for the conservative treatment of frozen shoulder [82].
Preoperative frozen shoulder negatively affected most functional outcomes, including range of motion, at 6 months and 1 year postoperatively after arthroscopic rotator cuff repair [33].
Preoperative frozen shoulder positively affected rotator cuff healing after arthroscopic rotator cuff repair [33].

Key Evidence

[L4] A detailed clinical history and examination is critical in the assessment of a painful, stiff shoulder. [1] (10.1186/1477-7800-2-2)
[L1] This updated guideline provides evidence-based guidance for managing frozen shoulder and identifies key areas for future research. [2] (10.1177/17585732251335955)
[L4] Most patients clinically diagnosed with primary frozen shoulder had undiagnosed systemic abnormalities and/or intra-articular pathologies. [5] (10.5397/cise.2018.21.2.82)
[Paper] Frozen shoulder is a specific, painful and debilitating condition effecting patients mainly in middle age. [6] (10.1016/j.maturitas.2014.02.009)
[L4] MR findings in frozen shoulder should not replace clinical judgments regarding further prognosis and treatment decisions. [7] (10.1007/s00167-015-3887-y)
[L2] If high-quality studies can confirm the findings of this review, then clinicians should monitor frozen shoulder patients with diabetes more closely and offer further treatment if pain or lack of function persist long-term. [8] (10.1016/j.arrct.2021.100141)
[L5] Different treatment strategies for frozen shoulder may be appropriate, depending on the location. [9] (10.1016/j.jse.2018.03.010)
[L5] The study adds information for shared decision-making but should not be a plea for surgery for patients with a mean duration of symptoms of 4 months; a strict definition of recalcitrant idiopathic frozen shoulder is needed to prevent unnecessary interventions. [10] (10.1177/2325967120903710)
[L2] This is clinically relevant in diagnostically challenging cases, for instance in the first phase of frozen shoulder, which can be difficult to distinguish from subacromial impingement. [11] (10.1007/s00167-020-05937-2)
[L5] Frozen shoulder is a poorly understood condition that typically involves substantial pain, movement restriction, and considerable morbidity. [12] (10.1016/j.math.2014.07.006)
[L1] The variety of participants included/excluded in trials and the variety of methods, interventions and outcomes used across the trials provided limited new evidence to inform the non-surgical management and treatment of people with frozen shoulder. [14] (10.2340/16501977-2578)
[L4] The etiology of frozen shoulder is still not known and our understanding of the pathogenesis is limited. [15] (10.3109/03009749009096786)
[L4] Health professionals manage frozen shoulders differently for different phases of the condition. [16] (10.1111/j.1758-5740.2010.00073.x)
[L4] Patients with a poor outcome or recurrent symptoms of a frozen shoulder after a MUA should be offered a further MUA with the expectation of a good outcome and a low complication rate. [17] (10.1302/0301-620x.99b6.bjj-2016-1133.r1)
[L3] Early surgical intervention might shorten the overall duration of symptoms in frozen shoulder and is not associated with inferior clinical outcomes when compared with late surgical intervention. [19] (10.1016/j.jse.2020.07.023)
[L4] Frozen shoulder is a common disease which causes significant morbidity. [21] (10.5312/wjo.v6.i2.263)
[L3] Long-term outcome after MUA for frozen shoulder is favourable with late recurrence being uncommon. [22] (10.1177/17585732211070007)
[L5] Unusual stiffness and pain in the shoulder of a young female patients suggests a wide range of disease entities, from simple frozen shoulder to (albeit rarely) infection. [23] (10.1016/j.radcr.2020.08.006)
[L1] This systematic review presents a summary of what is currently known about the tissue pathophysiology of primary frozen shoulder. [25] (10.1186/s12891-016-1190-9)
[L4] Frozen shoulder is a painful and debilitating condition with considerable economic impact; treatment aims to improve pain and function through a shared decision-making process, with a step-up approach from conservative measures to invasive treatments if symptoms persist. [28] (10.1177/1758573215601779)
[L3] An age of between 46 and 60 years and a previous history of contralateral frozen shoulder were statistically significant risk factors. [29] (10.1302/0301-620x.97b7.35387)
[L3] In the long term, 59% of patients had normal or near normal shoulders and 41% reported some ongoing symptoms. [30] (10.1016/j.jse.2007.05.009)
[L4] Therefore, routine use of shoulder MRI scans in patients with FS but without suspicion of an additional pathology may not be indicated. [31] (10.1016/j.jseint.2022.05.009)
[L1] There is insufficient evidence to reliably recommend a treatment approach for frozen shoulder. [32] (10.1136/bmj.i4162)
[L3] Preoperative frozen shoulder positively affected rotator cuff healing but negatively affected most functional outcomes, including ROM, at 6 months and 1 year postoperatively. [33] (10.1177/2325967120934449)
[L4] Further treatment was indicated in 41% of patients who could not tolerate more than 20 mL of injection, and recurrence was more common in primary (33%) versus secondary (16%) frozen shoulder. [34] (10.1177/17585732221124914)
[Paper] In patients with refractory primary frozen shoulder syndrome, arthroscopic capsular release emerges as a suitable option that leads to a faster and long-lasting recovery. [35] (10.1016/j.eats.2015.06.004)
[L4] The long-term results of arthroscopic capsular release in frozen shoulder were confirmed in 255 patients. [36] (10.1186/s13018-018-0758-5)
[L4] Imaging is an essential tool for evaluation of patients with shoulder pain; understanding the extent of an injury with imaging is key to successful management. [37] (10.1016/j.csm.2013.03.009)
[L4] However, classification of frozen shoulder was found to be controversial. [38] (10.4055/cios.2020.12.1.60)
[L4] Frozen shoulder following COVID-19 vaccination may present with clinical features similar to those of idiopathic frozen shoulder. [39] (10.1016/j.xrrt.2023.09.013)
[L2] Misdiagnosing shoulder tumors as frozen shoulder syndrome is likely to cause a significant delay in making a correct diagnosis. [40] (10.1016/j.jse.2009.05.010)
[L4] Frozen shoulder can occur after COVID-19 vaccination, and musculoskeletal specialists should be aware of this diagnosis to identify and treat such patients early. [41] (10.1016/j.jseint.2022.02.013)
[Paper] The trial aims to compare the effectiveness of a CNS-directed treatment program versus standard medical and physical therapy care on outcomes in participants with frozen shoulder. [42] (10.1186/s13063-019-3585-z)
[L4] The survey shows lower rates of agreement among the JSS members than the ASES members for the definition of primary frozen shoulder, the classification of primary and secondary frozen shoulder, and the divisions of secondary frozen shoulder. [43] (10.1016/j.jos.2018.12.012)
[L4] Symptomatic subjects demonstrated substantial kinematic deficits during humeral range of motion. [44] (10.1016/s0003-9993(03)00359-9)
[L3] Conservative treatment is effective for the treatment of frozen shoulder. [45] (10.1016/j.asmr.2025.101149)
[Commentary] Multisite corticosteroid injection therapy is more effective in terms of pain relief, restoration of motion, and functional status than single intra-articular injection for the treatment of primary frozen shoulder. [46] (10.1016/j.arthro.2021.02.028)
[L2] The findings of the study may provide evidence on the efficacy of these interventions and most likely, the optimal treatment approach for frozen shoulder related to diabetes, which may guide clinical practice. [47] (10.1186/s13018-024-04735-7)
[L1] Intra-articular steroid injection is effective and safe for frozen shoulder, relieving pain, improving functional performance, and increasing range of motion. [48] (10.1177/0363546516669944)
[L1] Both treatments were effective in patients with primary frozen shoulder. [49] (10.1016/j.arthro.2021.01.069)
[Abstract] The mid-term outcomes of TAME for frozen shoulders that were resistant to conservative treatments are encouraging and warrant further evaluation. [50] (10.1016/j.jse.2016.11.031)
[L2] Successful completion of this trial will provide evidence on the best treatment strategy for patients with a stage two frozen shoulder. [51] (10.1186/s12891-017-1763-2)
[L1] In patients with frozen shoulder, US-guided administration of corticosteroid injections into the glenohumeral joint increased the likelihood of successful injection but did not improve clinical outcomes at 12 weeks compared with blind administration. [52] (10.2106/jbjs.21.01007)
[L4] The arthroscopic 360° release is an effective and safe treatment modality for severe or recalcitrant frozen shoulder. [53] (10.1016/j.jseint.2024.07.006)
[L2] Postoperative frozen shoulder is a serious complication after shoulder surgery, with an incidence of 11%. [54] (10.1007/s00402-016-2589-3)
[L4] A combination of limited capsular release and MUA for the treatment of primary frozen shoulder is a safe and effective procedure resulting in marked improvement in pain, function and range of motion. [55] (10.1177/1758573215578590)
[L4] Physicians should re-examine frozen shoulder patients with repeated plain radiographs and further imaging, especially MRI, if conservative therapy fails. [56] (10.1016/j.jse.2011.07.026)
[L4] This survey summarized the trend in prevalent practices regarding frozen shoulder among shoulder specialists and senior shoulder surgeons of SESI, revealing strong consensus on imaging and nonoperative management but disagreement on terminology and specific surgical complications. [57] (10.1177/23259671221118834)
[L4] MR arthrography reveals characteristic findings in patients with frozen shoulder. [58] (10.1148/radiol.2332031219)
[L3] The anatomical structure of passive shoulder restraints has no impact on the difference in passive joint position sense values between external and internal rotation. [59] (10.1186/s12891-016-0971-5)
[L5] Hydrodilatation has emerged as a potential nonsurgical option in the management of frozen shoulder, but its role has yet to be fully clarified. [60] (10.1302/2058-5241.2.160061)
[L5] This scoping review outlines the complexity of the pathophysiology of frozen shoulder and provides a comprehensive overview of pathophysiologic mechanisms. [62] (10.1186/s40634-020-00307-w)
[L4] Secondary frozen shoulder may be more recalcitrant to conventional conservative treatment. [64] (10.1016/j.jor.2015.01.030)
[L4] This service evaluation demonstrates that management of frozen shoulder stage II to III, as conducted by physiotherapists in a primary care setting utilizing hydrodistension and a guided exercise programme, represents an effective non-operative treatment strategy. [66] (10.1177/1758573217701063)
[L1] MUA in stage 2 frozen shoulder can be considered safe and results in a faster recovery of range of motion and improved functional outcome compared to physiotherapy alone in the short term. [68] (10.1016/j.jseint.2023.11.004)
[L4] We found 94% of patients with spontaneous frozen shoulder recovered to normal levels of function and motion without treatment. [69] (10.1007/s11999-011-2176-4)
[L5] MUA and ACR are good treatment options for primary frozen shoulder; MUA is simpler but carries risks of serious complications, while ACR may be safer if performed by experienced surgeons and is convenient for patients with combined rotator cuff tears. [70] (10.5397/cise.2020.00311)
[L5] T2 signal hyperintensity and axillary capsule thickening are characteristic of the early stages of frozen shoulder, although MRI alone cannot completely define the disease stage. [71] (10.1016/j.xrrt.2024.05.002)
[L5] This technique is presented for the nonsurgical treatment of frozen shoulder, emphasizing correct injection sites. [72] (10.1016/j.eats.2022.06.020)
[L4] The burning sign is an abnormal finding that appears in dynamic MRI of severe frozen shoulder. [73] (10.1016/j.jse.2016.06.003)
[L2] There is limited evidence of the effectiveness of different forms of treatment used for frozen shoulder, with many studies evaluating treatment effects carrying a moderate to high risk of bias and omitting details of the duration of symptoms or the phase of the condition. [74] (10.1111/j.1758-5740.2010.00067.x)
[L5] Pathomechanics of the frozen shoulder characterised by glenohumeral motion limitations should be considered complicated, as confirmed by high tension in the anteroinferior glenohumeral capsule and altered scapular motion. [75] (10.1016/j.jseint.2025.04.003)
[L4] Dynamic MRI semiquantitatively demonstrated a reduction in abnormal blood flow and improvement in clinical results after MUC in patients with frozen shoulder. [77] (10.1016/j.jseint.2021.12.007)
[L4] Patients with frozen shoulder presented with altered shoulder muscle activity and kinematics, and one-session of heat and manual muscle release showed beneficial effects on shoulder muscle performance, kinematics, mobility, and pain. [78] (10.1186/s12891-017-1867-8)
[L3] The thickness of the inferior glenohumeral joint capsule in the 80° scapular plane elevated arm position was shown to be a highly reliable and valid method. [79] (10.1016/j.jseint.2024.06.004)
[L4] SHR of the affected shoulder is inversely related to severity of limitation of shoulder range of motion, which suggests a compensatory pattern. [80] (10.1016/j.jbiomech.2007.09.004)
[L4] Clinical improvement in patients with frozen shoulder was associated with a decrease in the coefficient of enhancement (CE) in the glenohumeral synovium. [81] (10.1007/s00776-004-0766-7)
[L2] Diabetes, male sex, simultaneous bilateral involvement, subsequent bilateral involvement, and a longer duration of symptoms recorded at the first visit were identified as poor prognostic factors for the conservative treatment of frozen shoulder. [82] (10.1016/j.jos.2019.03.011)
[L2] Our MR analysis suggests that there may be a causal relationship between hypothyroidism and frozen shoulder. [85] (10.1186/s12891-024-07826-y)
[L1] Patients who underwent image-guided (ultrasound) injections had statistically significant greater improvement in shoulder pain and function at 6 weeks after injection compared to blind injections. [86] (10.1186/1471-2474-12-137)
[L4] MR imaging of patients with severe frozen shoulder after MUC showed 29 capsule tears, 4 labrum tears, and 15 bone bruises of the humeral head. [87] (10.1016/j.jse.2015.06.019)
[L4] In the great majority of patients idiopathic frozen shoulder is a self-limiting condition, in which symptoms subside and full shoulder movement returns within a maximum of two years from the onset of symptoms. [88] (10.2106/00004623-197860040-00030)
[L4] In the great majority of patients idiopathic frozen shoulder is a self-limiting condition, in which symptoms subside and full shoulder movement returns within a maximum of two years from the onset of symptoms. [89] (10.2106/00004623-197860040-00029)
[L4] Timing has a significant influence on the outcome of manipulation for frozen shoulders. [90] (10.1016/j.jos.2020.11.002)

References

[1] Lessons learnt from the painful shoulder; a case series of malignant shoulder girdle tumours misdiagnosed as frozen shoulder. International Seminars in Surgical Oncology. 2005. DOI: 10.1186/1477-7800-2-2

[2] British Elbow and Shoulder Society patient care pathway: Frozen shoulder. Shoulder & Elbow. 2025. DOI: 10.1177/17585732251335955

[3] Campbell S Operative Orthopaedics 4 Volume Set. ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTION WITH BONE-PATELLAR TENDON-BONE GRAFT > PRIMARY FROZEN SHOULDER.

[4] Apley And Solomon S Concise System Of Orthopaedics And Trauma. ADHESIVE CAPSULITIS (FROZEN SHOULDER).

[5] Is the Frozen Shoulder Classification a Reliable Assessment?. Clinics in Shoulder and Elbow. 2018. DOI: 10.5397/cise.2018.21.2.82

[6] Frozen shoulder – A stiff problem that requires a flexible approach. Maturitas. 2014. DOI: 10.1016/j.maturitas.2014.02.009

[7] Correlations of magnetic resonance imaging findings with clinical symptom severity and prognosis of frozen shoulder. Knee Surgery, Sports Traumatology, Arthroscopy. 2015. DOI: 10.1007/s00167-015-3887-y

[8] Diabetes as a Prognostic Factor in Frozen Shoulder: A Systematic Review. Archives of Rehabilitation Research and Clinical Translation. 2021. DOI: 10.1016/j.arrct.2021.100141

[9] Comparative proteome analysis of the capsule from patients with frozen shoulder. Journal of Shoulder and Elbow Surgery. 2018. DOI: 10.1016/j.jse.2018.03.010

[10] What Is the Right Timing for Arthroscopic Capsular Release of a Frozen Shoulder? Letter to the Editor. Orthopaedic Journal of Sports Medicine. 2020. DOI: 10.1177/2325967120903710

[11] The role of ¹⁸F‐FDG PET/CT in the diagnosis of frozen shoulder. Knee Surgery, Sports Traumatology, Arthroscopy. 2020. DOI: 10.1007/s00167-020-05937-2

[12] Frozen shoulder contracture syndrome – Aetiology, diagnosis and management. Manual Therapy. 2015. DOI: 10.1016/j.math.2014.07.006

[14] Clinical effectiveness of non-surgical interventions for primary frozen shoulder: A systematic review. Journal of Rehabilitation Medicine. 2019. DOI: 10.2340/16501977-2578

[15] Frozen Shoulder: Current Concepts. Scandinavian Journal of Rheumatology. 1990. DOI: 10.3109/03009749009096786

[16] Managing Idiopathic Frozen Shoulder: A Survey of Health Professionals' Current Practice and Research Priorities. Shoulder & Elbow. 2010. DOI: 10.1111/j.1758-5740.2010.00073.x

[17] Recurrence of frozen shoulder after manipulation under anaesthetic (MUA). The Bone & Joint Journal. 2017. DOI: 10.1302/0301-620x.99b6.bjj-2016-1133.r1

[19] Does the timing of surgical intervention impact the clinical outcomes and overall duration of symptoms in frozen shoulder?. Journal of Shoulder and Elbow Surgery. 2021. DOI: 10.1016/j.jse.2020.07.023

[20] Classifications And Scores Of The Shoulder. Classifications of frozen shoulder.

[21] Frozen shoulder: A systematic review of therapeutic options. World Journal of Orthopedics. 2015. DOI: 10.5312/wjo.v6.i2.263

[22] Long-Term Outcomes Following Manipulation Under Anaesthetic for Patients with Primary and Secondary Frozen Shoulder. Shoulder & Elbow. 2021. DOI: 10.1177/17585732211070007

[23] Subscapularis pyomyositis presenting as shoulder stiffness mistaken as frozen shoulder in young female: a case report. Radiology Case Reports. 2020. DOI: 10.1016/j.radcr.2020.08.006

[24] Frozen Shoulder. 2024.

[25] The pathophysiology associated with primary (idiopathic) frozen shoulder: A systematic review. BMC Musculoskeletal Disorders. 2016. DOI: 10.1186/s12891-016-1190-9

[28] Frozen Shoulder. Shoulder & Elbow. 2015. DOI: 10.1177/1758573215601779

[29] Frozen shoulder after simple arthroscopic shoulder procedures. The Bone & Joint Journal. 2015. DOI: 10.1302/0301-620x.97b7.35387

[30] Long-term outcome of frozen shoulder. Journal of Shoulder and Elbow Surgery. 2008. DOI: 10.1016/j.jse.2007.05.009

[31] Is routine magnetic resonance imaging necessary in patients with clinically diagnosed frozen shoulder? Utility of magnetic resonance imaging in frozen shoulder. JSES International. 2022. DOI: 10.1016/j.jseint.2022.05.009

[32] What is the most effective treatment for frozen shoulder?. BMJ. 2016. DOI: 10.1136/bmj.i4162

[33] Effect of Preoperative Frozen Shoulder on Clinical Outcomes After Arthroscopic Rotator Cuff Repair. Orthopaedic Journal of Sports Medicine. 2020. DOI: 10.1177/2325967120934449

[34] Recurrence of the frozen shoulder after hydrodilatation, what is the true incidence?. Shoulder & Elbow. 2022. DOI: 10.1177/17585732221124914

[35] Primary Frozen Shoulder Syndrome: Arthroscopic Capsular Release. Arthroscopy Techniques. 2015. DOI: 10.1016/j.eats.2015.06.004

[36] Clinical outcome of arthroscopic capsular release for frozen shoulder: essential technical points in 255 patients. Journal of Orthopaedic Surgery and Research. 2018. DOI: 10.1186/s13018-018-0758-5

[37] Imaging of the Shoulder with Arthroscopic Correlation. Clinics in Sports Medicine. 2013. DOI: 10.1016/j.csm.2013.03.009

[38] Definition, Diagnosis, Treatment, and Prognosis of Frozen Shoulder: A Consensus Survey of Shoulder Specialists. Clinics in Orthopedic Surgery. 2020. DOI: 10.4055/cios.2020.12.1.60

[39] Frozen shoulder after COVID-19 vaccination versus idiopathic frozen shoulder: similar clinical features and functional improvement at 1-year follow-up. JSES Reviews, Reports, and Techniques. 2024. DOI: 10.1016/j.xrrt.2023.09.013

[40] Tumors masked as frozen shoulders: A retrospective analysis. Journal of Shoulder and Elbow Surgery. 2010. DOI: 10.1016/j.jse.2009.05.010

[41] Frozen shoulder after COVID-19 vaccination. JSES International. 2022. DOI: 10.1016/j.jseint.2022.02.013

[42] A central nervous system-focused treatment approach for people with frozen shoulder: protocol for a randomized clinical trial. Trials. 2019. DOI: 10.1186/s13063-019-3585-z

[43] Representative survey of frozen shoulder questionnaire responses from the Japan Shoulder Society: What are the appropriate diagnostic terms for primary idiopathic frozen shoulder, stiff shoulder or frozen shoulder?. Journal of Orthopaedic Science. 2019. DOI: 10.1016/j.jos.2018.12.012

[44] Shoulder kinematics in subjects with frozen shoulder11No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated.. Archives of Physical Medicine and Rehabilitation. 2003. DOI: 10.1016/s0003-9993(03)00359-9

[45] Conservative Treatment for Frozen Shoulder Is Effective Regardless of the Severity of Symptoms. Arthroscopy, Sports Medicine, and Rehabilitation. 2025. DOI: 10.1016/j.asmr.2025.101149

[46] Editorial Commentary: Corticosteroid Injections and Physical Therapy Are Effective First-Line Treatments for Frozen Shoulder. Arthroscopy: The Journal of Arthroscopic & Related Surgery. 2021. DOI: 10.1016/j.arthro.2021.02.028

[47] Comparative effectiveness of low-level laser therapy versus muscle energy technique among diabetic patients with frozen shoulder: a study protocol for a parallel group randomised controlled trial. Journal of Orthopaedic Surgery and Research. 2024. DOI: 10.1186/s13018-024-04735-7

[48] Intra-articular Steroid Injection for Frozen Shoulder: A Systematic Review and Meta-analysis of Randomized Controlled Trials With Trial Sequential Analysis. The American Journal of Sports Medicine. 2016. DOI: 10.1177/0363546516669944

[49] A Multisite Injection Is More Effective Than a Single Glenohumeral Injection of Corticosteroid in the Treatment of Primary Frozen Shoulder: A Randomized Controlled Trial. Arthroscopy: The Journal of Arthroscopic & Related Surgery. 2021. DOI: 10.1016/j.arthro.2021.01.069

[50] Mid-term outcomes of prospective clinical trial of transcatheter arterial micro embolization (TAME) for resistant frozen shoulder. Journal of Shoulder and Elbow Surgery. 2017. DOI: 10.1016/j.jse.2016.11.031

[51] Manipulation under anesthesia versus physiotherapy treatment in stage two of a frozen shoulder: a study protocol for a randomized controlled trial. BMC Musculoskeletal Disorders. 2017. DOI: 10.1186/s12891-017-1763-2

[52] In Frozen Shoulder, US-Guided Versus Blind Administration of Intra-Articular Corticosteroid Injections Increased Accuracy of Injections but Did Not Improve Clinical Outcomes at 12 Weeks. Journal of Bone and Joint Surgery. 2021. DOI: 10.2106/jbjs.21.01007

[53] The clinical course and outcomes following arthroscopic frozen shoulder 360° release. JSES International. 2024. DOI: 10.1016/j.jseint.2024.07.006

[54] Incidence and prognostic factors for postoperative frozen shoulder after shoulder surgery: a prospective cohort study. Archives of Orthopaedic and Trauma Surgery. 2017. DOI: 10.1007/s00402-016-2589-3

[55] Limited capsular release and controlled manipulation under anaesthesia for the treatment of frozen shoulder. Shoulder & Elbow. 2015. DOI: 10.1177/1758573215578590

[56] Glenohumeral joint tuberculosis that mimics frozen shoulder: a retrospective analysis. Journal of Shoulder and Elbow Surgery. 2012. DOI: 10.1016/j.jse.2011.07.026

[57] Trends in Practice Among Shoulder Specialists in the Management of Frozen Shoulder: A Consensus Survey. Orthopaedic Journal of Sports Medicine. 2022. DOI: 10.1177/23259671221118834

[58] Frozen Shoulder: MR Arthrographic Findings. Radiology. 2004. DOI: 10.1148/radiol.2332031219

[59] Shoulder proprioception – lessons we learned from idiopathic frozen shoulder. BMC Musculoskeletal Disorders. 2016. DOI: 10.1186/s12891-016-0971-5

[60] Indications for hydrodilatation for frozen shoulder. EFORT Open Reviews. 2017. DOI: 10.1302/2058-5241.2.160061

[61] Rockwood And Matsen S The Shoulder. Arthroscopic Management of Prearthritic and Arthritic Conditions of the Shoulder and the Postarthroplasty Shoulder > Secondary Frozen Shoulder.

[62] The puzzling pathophysiology of frozen shoulders – a scoping review. Journal of Experimental Orthopaedics. 2020. DOI: 10.1186/s40634-020-00307-w

[64] The management of secondary frozen shoulder after anterior shoulder dislocation – The results of manipulation under anaesthesia and injection. Journal of Orthopaedics. 2016. DOI: 10.1016/j.jor.2015.01.030

[66] The effectiveness of ultrasound guided hydrodistension and physiotherapy in the treatment of frozen shoulder/adhesive capsulitis in primary care: a single centre service evaluation. Shoulder & Elbow. 2017. DOI: 10.1177/1758573217701063

[68] Improved range of motion after manipulation under anesthesia versus physiotherapy for stage two frozen shoulder: a randomized controlled trial. JSES International. 2024. DOI: 10.1016/j.jseint.2023.11.004

[69] The Natural History of Idiopathic Frozen Shoulder: A 2- to 27-year Followup Study. Clinical Orthopaedics & Related Research. 2012. DOI: 10.1007/s11999-011-2176-4

[70] Arthroscopic capsular release versus manipulation under anesthesia for primary frozen shoulder. Clinics in Shoulder and Elbow. 2020. DOI: 10.5397/cise.2020.00311

[71] Can magnetic resonance imaging distinguish clinical stages of frozen shoulder? A state-of-the-art review. JSES Reviews, Reports, and Techniques. 2024. DOI: 10.1016/j.xrrt.2024.05.002

[72] Ultrasound Guıded Multisıte Injectıon Technıque in the Treatment of Frozen Shoulder. Arthroscopy Techniques. 2022. DOI: 10.1016/j.eats.2022.06.020

[73] Characteristics of dynamic magnetic resonance imaging of idiopathic severe frozen shoulder. Journal of Shoulder and Elbow Surgery. 2017. DOI: 10.1016/j.jse.2016.06.003

[74] An Overview of Factors Relevant to Undertaking Research and Reviews on the Effectiveness of Treatment for Frozen Shoulder. Shoulder & Elbow. 2010. DOI: 10.1111/j.1758-5740.2010.00067.x

[75] Pathomechanics of glenohumeral capsule and scapula in idiopathic frozen shoulder: a study using a three-dimensional finite element model. JSES International. 2025. DOI: 10.1016/j.jseint.2025.04.003

[77] Reduction of abnormal blood flow in frozen shoulder after shoulder manipulation under ultrasound-guided cervical nerve root block: semiquantitative analysis using dynamic magnetic resonance imaging. JSES International. 2022. DOI: 10.1016/j.jseint.2021.12.007

[78] The immediate effect of muscle release intervention on muscle activity and shoulder kinematics in patients with frozen shoulder: a cross-sectional, exploratory study. BMC Musculoskeletal Disorders. 2017. DOI: 10.1186/s12891-017-1867-8

[79] Inferior glenohumeral joint capsule thickness in frozen shoulder via ultrasonography. JSES International. 2024. DOI: 10.1016/j.jseint.2024.06.004

[80] Three-dimensional scapular kinematics and scapulohumeral rhythm in patients with glenohumeral osteoarthritis or frozen shoulder. Journal of Biomechanics. 2008. DOI: 10.1016/j.jbiomech.2007.09.004

[81] Synovial response to intraarticular injections of hyaluronate in frozen shoulder: a quantitative assessment with dynamic magnetic resonance imaging. Journal of Orthopaedic Science. 2004. DOI: 10.1007/s00776-004-0766-7

[82] Is frozen shoulder completely resolved at 2 years after the onset of disease?. Journal of Orthopaedic Science. 2020. DOI: 10.1016/j.jos.2019.03.011

[85] Causal associations of hypothyroidism with frozen shoulder: a two-sample bidirectional Mendelian randomization study. BMC Musculoskeletal Disorders. 2024. DOI: 10.1186/s12891-024-07826-y

[86] Image-guided versus blind corticosteroid injections in adults with shoulder pain: A systematic review. BMC Musculoskeletal Disorders. 2011. DOI: 10.1186/1471-2474-12-137

[87] Magnetic resonance imaging and short-term clinical results of severe frozen shoulder treated with manipulation under ultrasound-guided cervical nerve root block. Journal of Shoulder and Elbow Surgery. 2016. DOI: 10.1016/j.jse.2015.06.019

[88] Brief Note The Natural History of 'Idiopathic' Frozen Shoulder. The Journal of Bone & Joint Surgery. 1978. DOI: 10.2106/00004623-197860040-00030

[89] The natural history of 'idiopathic' frozen shoulder.. The Journal of Bone & Joint Surgery. 1978. DOI: 10.2106/00004623-197860040-00029

[90] Does the timing of shoulder manipulation under ultrasound-guided cervical nerve root block for frozen shoulder affect the clinical outcome?. Journal of Orthopaedic Science. 2022. DOI: 10.1016/j.jos.2020.11.002