Magnetic resonance imaging-based cortical thickness as a predictor of osteoporosis in rotator cuff tears

Background Osteoporosis is frequently unrecognised in patients undergoing rotator cuff repair. Shoulder magnetic resonance imaging (MRI), already standard of care, offers an opportunity for opportunistic osteoporosis screening. Objective To evaluate whether proximal humeral cortical bone thickness on shoulder MRI can predict osteoporosis in patients with rotator cuff tears (RCTs). Methods A retrospective, cross-sectional study included 130 female patients (mean age 68.8 ± 8.9 years) who underwent shoulder MRI and dual-energy X-ray absorptiometry (DXA) within 6 months. Average cortical bone thickness (CBTavg) and cortical bone thickness index (CBTg) were measured on coronal T1-weighted images. Patients were classified per World Health Organization DXA criteria (T-score ≤ −2.5) at the femur, the lumbar spine and either site (all-site). Only female patients were eligible for inclusion; no male data were collected. Results CBTavg and CBTg were significantly lower in osteoporosis groups across all classification schemes. For femur DXA, CBTavg at 4.5 mm achieved 100% sensitivity and 86% specificity (area under the curve (AUC) = 0.955). For all-site osteoporosis, CBTg at 0.26 yielded 100% sensitivity and 87.8% specificity (AUC = 0.958). Conclusion MRI-derived proximal humeral cortical thickness is a promising osteoporosis screening tool in RCT patients. A CBTavg below 4.5 mm should prompt DXA referral, improving preoperative bone health assessment without added cost or radiation. These cut-offs require external validation before clinical implementation.