Frozen Shoulder Moves Beyond a Simple Stiffness Syndrome
Frozen shoulder is no longer easy to describe as a single-issue problem of painful restriction. A new systematic review has assembled a multidomain framework that treats frozen shoulder as a condition shaped by interacting clinical, metabolic, imaging, psychological, and inflammatory contributors rather than by capsular stiffness alone.
The review was built from 63 studies and used a structured evidence-weighted approach to organize the available literature into clinically interpretable spheres. The aim was not to force every study into one narrow explanatory model, but to map which domains repeatedly carried the strongest and most direct signal across the published record. In practice, that means the framework moves the discussion of frozen shoulder toward whole-person phenotyping, where symptom burden, endocrine status, structural tissue change, distress, sleep disruption, and inflammatory biology all sit in the same analytic field.
Clinical Stiffness and Symptom Duration Still Anchor the Phenotype
Within the clinical sphere, baseline stiffness and reduced range of motion emerged as dominant features, with symptom duration also carrying major weight. Those findings fit everyday practice: frozen shoulder announces itself through loss of movement, but the trajectory of that loss appears to matter just as much as the starting point.
A shoulder that has been stiff for months is not the same clinical problem as one in an early inflammatory phase, and the review treats that distinction as biologically meaningful rather than merely descriptive. Symptom duration likely captures more than chronology. It may also reflect how long the joint has been exposed to capsular fibrosis, pain-related guarding, and altered use patterns that reinforce immobility.
Metabolic and Endocrine Burden Emerges as a Major Driver
The strongest global contribution in the proposed framework came from the metabolic/endocrine sphere, which accounted for 24.0% of the provisional cross-sphere weighting. Diabetes-related burden and glycaemic control were the dominant elements in that domain, reinforcing the long-recognized link between disordered glucose metabolism and shoulder capsular disease.
That association is clinically important because it points to frozen shoulder as more than an isolated musculoskeletal complaint. Poor glycaemic control may intensify glycation of connective tissue, promote fibrosis, and alter the tissue remodeling response that should otherwise limit capsular thickening. The framework does not claim that diabetes explains every case, but it places metabolic status close to the center of phenotyping rather than at the margins.
In a busy clinic, this changes the frame of evaluation. A patient with persistent shoulder stiffness and poorly controlled diabetes may warrant a different expectation of duration, recovery pace, and therapeutic resistance than a patient without metabolic disease.
Imaging Contributes Structural Weight, Not Just Confirmation
The imaging sphere accounted for 22.9% of the provisional contribution and was led by capsular structural biomarkers. That matters because imaging in frozen shoulder has often been treated as a confirmatory tool. Here, it is positioned as a domain with genuine phenotyping value.
Capsular thickening, structural remodeling, and other tissue-level biomarkers likely reflect the anatomic expression of the disease process rather than a passive afterthought. When structural change is prominent, pain and motion loss may be tracking a more fixed fibroinflammatory state. The review’s weighting suggests that imaging can help separate patients whose dominant burden is structural restriction from those in whom other domains, such as psychological distress or metabolic burden, are more influential in daily function.
Psychological Distress and Sleep Burden Sit Close to the Core
The psychological sphere contributed 20.0% of the overall framework, with distress, mood disturbance, and sleep-related burden carrying the strongest signal. That is clinically believable. Chronic shoulder pain interrupts sleep, poor sleep amplifies pain sensitivity, and persistent distress can deepen disability even when the joint pathology itself is static.
This is where the review becomes especially useful. It refuses the old habit of treating psychological features as secondary reactions to the “real” orthopedic disease. In frozen shoulder, mood and sleep may influence how symptoms are experienced, how patients move, and how they respond to rehabilitation. The shoulder is not just stiff; the patient may also be exhausted, anxious, or demoralized, and those states can narrow functional tolerance in ways that matter at the bedside.
A one-track treatment plan is unlikely to capture that complexity.
Inflammatory Biology Remains a Smaller but Coherent Signal
The inflammatory sphere carried a 12.0% provisional contribution, the smallest of the five retained domains, but it remained biologically meaningful. The main retained signal involved IL-1β-related biomarker and susceptibility evidence.
That inflammatory component helps connect the clinical picture to a plausible pathophysiologic sequence. IL-1β sits within a network of cytokine activity that can promote pain, tissue activation, and fibrotic remodeling. Even though the inflammatory evidence base was limited to a single direct but lower-confidence study in the corpus, the framework preserves it because it fits the broader biology of fibroinflammatory disease. The signal is not strong enough to stand alone as a defining explanation, but it supports the idea that frozen shoulder may move through inflammatory and fibrotic phases rather than through one uniform mechanism.
Why the Weighted Framework Matters at the Bedside
The review’s central claim is not that every patient with frozen shoulder shares the same drivers. The claim is sharper than that: the condition is better represented as a weighted multidomain disorder, where the dominant contributors differ across patients and may overlap within the same patient.
That has practical consequences for assessment. A clinician evaluating frozen shoulder may need to consider motion loss, duration, diabetes burden, glycaemic control, psychological distress, sleep problems, and imaging phenotype together instead of in isolation. The framework also creates a rational basis for future treatment-allocation research. A patient with heavy metabolic burden may not follow the same course as one whose phenotype is dominated by mood, sleep disruption, and pain amplification. A patient with marked structural capsular change may need a different rehabilitation emphasis than a patient in whom symptoms still reflect a more active inflammatory state.
The review does not present this framework as a validated decision tool. It is an exploratory synthesis built from observational studies, with sphere weights derived from directness, consistency, prognostic or phenotyping relevance, methodological robustness, risk-of-bias profile, and domain specificity. Those weights should be read as evidence-derived estimates, not as clinical directives.
Validation Remains the Next Test for Personalised Care
The most important limitation is built into the design: this is a conceptual framework, not a prospectively tested model. The evidence base is uneven across domains, with far more studies in the clinical sphere than in the inflammatory sphere. Cross-sphere weighting therefore depends on how the available literature was distributed, how directly each study addressed frozen shoulder phenotypes, and how robustly each domain could be measured.
Future work will need to test whether these five spheres can actually separate patient subgroups, predict response to treatment, or improve outcomes when used in real clinical pathways. Until then, the framework’s value lies in forcing a more disciplined conversation about what frozen shoulder is and where its burden comes from: not one syndrome, but a cluster of interacting biological and behavioral processes that can present as the same stiff, painful shoulder.