Association of Walch glenoid morphology and fracture patterns in proximal humeral fractures
PURPOSE: Proximal humeral fracture (PHF) assessment focuses on fracture pattern, bone quality, and patient factors, whereas glenoid morphology-central to biomechanics in glenohumeral arthritis and instability-is overlooked. We hypothesized that glenoid morphology consistent with Walch type B (posterior decentring) is independently associated with complex, tuberculum-involving PHFs. METHODS: This single-centre, retrospective cohort study analysed 919 patients with a PHF and shoulder computed tomography (CT). Two blinded shoulder-subspecialized orthopedic trauma surgeons independently classified all cases; agreement was quantified with Cohen's κ. Fractures were classified by the Neer and AO/OTA systems and glenoid morphology by the Walch classification. The primary outcome-tuberculum involvement (Neer IV/V)-was analysed by multivariable logistic regression adjusted for age at trauma, sex, and osteoarthritis. RESULTS: The cohort comprised 919 patients (72.6% female; median age at trauma 69 years). The Walch distribution in this CT-selected cohort was type A 42.7%, B 54.4%, C 2.7%, and D 0.2%; osteoarthritis was present in 59.5%. Walch agreement was moderate (κ = 0.53), with disagreement concentrated at the A1↔B1 decentring boundary (p < 0.001). Walch type B was independently associated with tuberculum involvement (adjusted OR 2.07, 95% CI 1.56-2.74; adjusted RR 1.30, 95% CI 1.17-1.45; both p < 0.001) and with greater AO/OTA fracture complexity (adjusted ordinal OR 1.60, p < 0.001). Female sex was independently associated with the outcome (adjusted OR 1.44, 95% CI 1.05-1.97, p = 0.023); age and osteoarthritis were not significant. CONCLUSION: In this CT-selected study population, glenoid morphology consistent with Walch type B was independently associated with complex, tuberculum-involving proximal humeral fractures. The association was statistically robust but clinically modest (AUC 0.62), and Walch type B morphology represented one of several correlates of fracture pattern rather than a standalone predictor. Glenoid morphology may therefore be worth recording during CT and fracture-pattern assessment; these hypothesis-generating findings are not sufficient to modify current treatment algorithms, but support the inclusion of glenoid parameters in future risk models and prospective studies. LEVEL OF EVIDENCE: Level III retrospective cohort study prognostic study.
Authors
- Jan‐Dierk Clausen (ORCID: https://orcid.org/0000-0001-6156-326X)
- Jorge Mayor (ORCID: https://orcid.org/0009-0006-2313-0846)
- Stephan Sehmisch (ORCID: https://orcid.org/0000-0002-5270-7249)
- Swantje Oberthür (ORCID: https://orcid.org/0000-0003-2742-7357)
- Axel Gänsslen (ORCID: https://orcid.org/0000-0003-2186-5049)
- Marcel Winkelmann (ORCID: https://orcid.org/0000-0003-4348-5725)
- Gökmen Aktas (ORCID: https://orcid.org/0009-0007-0646-8156)
- Hür Özbek (ORCID: https://orcid.org/0009-0001-0224-379X)
Institutions
- Medizinische Hochschule Hannover (DE)
Publication Details
- Journal
- European Journal of Trauma and Emergency Surgery
- Published
- 2026-11-17
- DOI
- https://doi.org/10.1007/s00068-026-03387-y
- Primary Topic
- Shoulder Injury and Treatment
- Type
- article
- Field-Weighted Citation Impact
- 0.00