Article Text

Download PDFPDF
Multidirectional instability of the glenohumeral joint: an unstable classification resulting in uncertain evidence-based practice
  1. Kajsa Johansson
  1. Correspondence to Dr Kajsa Johansson, Head of the division of Physiotherapy, Department of Medical and Health Sciences, Linköping University, Sweden; kajsa.johansson{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The aim of this editorial is to discuss the pros and cons of some of the current classification systems of shoulder instability, with special emphasis on instability in more than one direction, and to summarise the basis for future clinical reasoning and research.

The term ‘multidirectional instability’ (MDI) is commonly used in adolescents and young adults, who can sometimes voluntarily subluxate their glenohumeral joints. They report pain, instability and shoulder disability, and present with symptomatic abnormal movement of the joint in two or more directions; the abnormal movement may predominate in one direction. The joint laxity may be congenital (in which case, it is often bilateral), or may due to the microtrauma of highly repetitive end-range movements, for example, in overhead athletes.


Classifying patients with shoulder instability is a challenge. Rockwood1 described four types of instability and patients with MDI fit into type 3, voluntary subluxation with no history of trauma, or type 4, atraumatic involuntary subluxation. Later, Kessel and Bayley2 grouped them as having ‘habitual instability’, but provided no guide for treatment. In the late 1980s, Thomas and Matsen3 coined the acronym ‘AMBRI’ (atraumatic, multidirectional, bilateral, rehabilitation and possible inferior capsular shift) for patients with MDI. They included a treatment aspect, and later a second I was added, for ‘closure of the rotator cuff interval’.

Gerber and Nyffeler4 defined three classes of shoulder instability: static (no classic symptoms, characterised by humeral head displacement, associated with rotator cuff or joint disease identified with imaging techniques), dynamic (trauma with instability symptoms, restorable loss of articular congruency) and voluntary (able to dislocate at will).

Lewis et al5 then proposed the Stanmore classification in 2004. Based on clinical presentation, patients are placed in relation to one of three polar types including both, structural and neuromuscular explanations as well as aetiology. The advantage with this non-static model is that patients can be more or less accordant with a defined pole and change over time. Patients with MDI can be placed along the axis of polar type II (atraumatic/structural) and polar type III (muscle patterning/non-structural), whereas polar type I is structural instability due to trauma.5

Kuhn6 questioned the term MDI and focused the primary direction of translation instead. He presented the FEDS classification involving Frequency, Etiology, Direction and Severity of shoulder instability. This classification was more detailed than earlier systems and has content validity and high reliability.7 If structural lesions are added, the complexity of the instability, with interacting symptom-causing components, seems embraced.

In summary, there is no consensus regarding which classification system to use and there are no standardised diagnostic criteria for MDI in the literature. This is highlighted in the systematic review by Warby and colleagues8 who concluded that ‘the effect of exercise-based management compared to surgery for MDI is difficult to determine due to participant heterogeneity and a high level of bias across included studies’. Consequently, physiotherapists (and orthopaedic surgeons) base clinical management of patients with glenohumeral MDI mainly on experience.

Clinical classification

The first issue that needs to be solved is how to classify patients clinically. Structural (contractile and non-contractile elements) and non-structural elements are both involved, and need to be identified in the light of how they influence symptoms and shoulder disability. Future research faces the challenge of identifying which elements are important to address during treatment and by whom. Patient heterogeneity needs to be addressed before randomising volunteers into controlled trials. Additionally, it is crucial to agree on appropriate outcome measures to evaluate the efficacy of interventions. This has been highlighted in the ISAKOS committee consensus statement on shoulder instability.9

Functional analysis of the individual patient—a key starting place

There is a clear need for a better and more pragmatic way of classifying patients as a basis for treatment. The individual patient's clinical presentation, supported by anatomical structural diagnosis (where possible), is a better starting point than trying to fit every patient into a simplified classification system. Deficiencies in passive structures will only lead to symptoms of instability and pain if the patient has insufficient neuromuscular control. Focus on the functional analysis: quality of arm movement, neuromuscular control affecting shoulder stability, kinematics of the glenohumeral and scapulothoracic joint, and core stability. These core components are well known,10 but the challenge is how to put them all together.

Quality aspects of each component need to be assessed as does how they respond to treatment. It is crucial within the analysis to find out if a change in scapular and rotator cuff muscle activation patterns can affect the symptoms and increase functional stability throughout the kinetic chain over time. Then, if rehabilitation fails, surgery should be considered, but continuously trying to make patients with shoulder instability more ‘squared’ to fit in ‘the boxes’ of a classification system will counteract building the body of evidence for best clinical practice.


View Abstract


  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles