Rehabilitation of Frozen Shoulder: A Systematic Review and Meta-Analysis of Multifactorial Interventions

Frozen shoulder (FS) is characterized by a multifactorial progression, often worsened by metabolic factors such as diabetes. Effective management may require a multidisciplinary approach that includes therapeutic exercise, physiotherapy, metabolic control, psychological support, and healthy lifestyle interventions. However, current evidence remains fragmented and predominantly focused on isolated domains, with limited integration of these factors within a biopsychosocial framework. The aim of this study was to comprehensively evaluate the influence of exercise-based interventions, as well as metabolic, psychological, and lifestyle-related factors, on clinical outcomes, including pain, function, range of motion, and quality of life, in patients with FS.

This systematic review with meta-analysis, reported following the PRISMA guidelines, included 31 studies published between 2010 and 2024, comprising randomized controlled trials, observational studies, Mendelian randomization studies, and qualitative studies. MEDLINE, Web of Science, CINAHL, SPORTDiscus, and Scopus were searched until April 2025. The risk of bias was assessed according to study design. The Cochrane Risk of Bias tool (RoB 1) was used for randomized controlled trials. Observational studies (including cross-sectional, case-control, and cohort studies) were assessed using the Newcastle-Ottawa scale (NOS). Mendelian randomization studies were evaluated using established methodological quality criteria for MR studies, including the assessment of instrument validity, pleiotropy, and heterogeneity. Qualitative studies were appraised using appropriate qualitative assessment criteria. The certainty of evidence was rated using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach.

Twelve studies evaluated physical interventions. Global effect summary of exercise versus control showed a small but statistically significant benefit favoring exercise (standardized mean difference [SMD], Hedges' g = 0.10; 95% CI 0.05-0.15) across clinical outcomes, particularly pain, function, and range of motion. Associations between FS and metabolic parameters (e.g., triglycerides, glucose, and HbA1c) were identified in 12 studies, although these results were synthesized narratively and no pooled effect estimates (e.g., odds ratios) were calculated due to heterogeneity. Seven studies showed that anxiety was consistently associated with FS severity, with potential impact on pain, functional outcomes, and quality of life, although this relationship may depend on sample characteristics. The risk of bias was low in exercise-related studies, unclear in those on quality of life, and high in metabolic studies. Overall, the certainty of evidence was rated as very low due to the risk of bias, heterogeneity, and imprecision.

Evidence supports a multifactorial, biopsychosocial model for FS. Exercise-based interventions were associated with small improvements in pain, function, and range of motion, while metabolic dysfunction and anxiety were identified as factors associated with FS severity. However, the certainty of evidence was very low and findings should be interpreted with caution due to high heterogeneity and methodological variability across studies. Future research should prioritize integrated, multidisciplinary strategies to improve FS treatment.