Degenerative rotator cuff tears in correlation with different anatomic shoulder parameters on MRI

Abstract Background Rotator cuff tears (RCTs) are a prevalent cause of shoulder pain, particularly in older adults. Since magnetic resonance imaging (MRI) provides better soft tissue assessment than other imaging modalities, it is used as the primary modality to diagnose and evaluate these tears, which frequently arise from degenerative changes. Various anatomical shoulder parameters, such as the Critical Shoulder Angle (CSA), Lateral Acromial Angle (LAA), Acromial Index (AI), and Acromiohumeral Distance (AHD), have been explored for their potential correlation with RCTs. However, there remains a lack of comprehensive data comparing these parameters simultaneously in relation to RCTs. Aim To assess the correlation between different anatomical shoulder parameters, on MRI, and degenerative rotator cuff tears (RCTs). Results This retrospective study included 67 participants, with data collected from the picture archiving and communications system (PACS) at Ain Shams University Hospitals. Participants underwent shoulder MRI examinations using a 1.5 T & 3 T machines, and anatomical measurements were assessed to determine their correlation with RCTs.The study revealed a strong positive correlation between the CSA and RCTs (r = 0.827, p < .0001), indicating that a higher CSA was associated with an increased likelihood of RCTs. Similarly, the AI demonstrated a significant positive correlation (r = 0.695, p < .0001). In contrast, the LAA and AHD were negatively correlated with RCT presence (r = − 0.542, p < .0001; r = − 0.413, p = .001), suggesting that lower values of these parameters were associated with higher RCT risk. The receiver operating characteristic (ROC) analysis indicated that CSA (Area Under the Curve AUC = 0.975) and AI (AUC = 0.900) were the most effective parameters for determining the presence of RCTs, while LAA and AHD exhibited limited discriminative power. Conclusion CSA and AI are valuable MRI parameters for diagnosing RCTs, showing strong correlation and high discriminative ability. In contrast, LAA and AHD, although being statistically significant, the low AUC suggests that they are not a strong predictor of RCT. These findings suggest that CSA and AI should be prioritized in the assessment of patients with suspected rotator cuff tears.