Association between acromion morphological parameters and rotator cuff tears in Somali population—a three-dimensional computed tomographic study Authors: Maslah Idiris Ali, Yuan-quan Chen, Deng Xing-Hao, Yi-Tao Yang, Daha Osman Heile, Ayoub Hussein Ahmed, Jing-yi Hou, Yi Long, Rui Yang Published: 2024 ABSTRACT AbstractBackgroundPrevious research has established connections between the acromion index (AI), critical shoulder angle (CSA), acromiohumeral distance (AHD), and lateral acromion angle (LAA) with the incidence of rotator cuff tears (RCTs). Despite numerous international studies exploring the variability of shoulder joint morphology and parameters, research incorporating these parameters has yet to be conducted in Somalia. This study aims to analyze and describe the radiological parameters of acromion morphology in the Somali populations. Furthermore, to investigate the relationship of these parameters with rotator cuff tears within the Somali population.MethodsThe data and physical examination of 188 patients who had a CT scan of the shoulder from 2018 to 2023 were retrospective analyses, including 107 patients (49 females, 58 males) with RCT and 81 patients (20 females, 61 males) with non-RCT. Using Three-dimensional computed tomography (3D-CT), parameters such as the AI, CSA, AHD, and LAA were compared between the RCT and non-RCT groups. Two independent assessors measured Each parameter from anterior views of the scapula. The inter- and intra-observer reliability was evaluated using the intraclass correlation coefficient (ICC). Additionally, the relationship between these parameters and the presence of rotator cuff tears was explored, and the predictive value of each parameter was assessed through receiver operating characteristic (ROC) analysis.ResultsThe results showed excellent intra- and inter-observer reliability, with all ICC values above 0.75. Among these, the CSA exhibited the greatest measurement stability, with an intra-observer ICC of 0.929 and an inter-observer ICC of 0.911. Significant positive correlations were identified between the AI, CSA, and the presence of RCTs, with CSA exhibiting the strongest correlation (r = 0.629; P < 0.001). Additionally, significant negative correlations were found between the presence of RCTs and both AHD (–0.247) and LAA (–0.338; P < 0.001 for both). The ROC analysis revealed that the CSA is the most effective parameter for identifying the presence of RCT (area under the ROC curve. 0.857).ConclusionThe outcome of this study shows that AI, CSA, AHD, and LAA are independent factors associated with rotator cuff tears in the East African, particularly the Somali population. Specifically, larger AI, CSA, and smaller AHD and LAA are correlated with an increased likelihood of RCTs in this population. In patients with shoulder disease suspected of RCTs, using a 3D-CT scan to measure the CSA may be helpful, as it is the best predictor measurement.Clinical trial numberNot Applicable.Keywords: Rotator cuff tear, Somali population, Acromion morphology, Acromion index, Critical shoulder angle, Acromiohumeral distance, Lateral acromion angle, Three-dimensional computed tomography Introduction Shoulder pain is a common problem in the population, and rotator cuff tears (RCTs) account for most of such pains [1, 2]. This type of pain is usually slow and persistent, which can negatively affect the normal functioning of the shoulder and reduce the overall quality of life [3]. The exact cause of RCTs has yet to be determined. Still, several factors are believed to contribute to its development, including degenerative changes in tendons around the shoulder joint (intrinsic causes) and mechanical compression from the acromion due to external force or trauma (extrinsic causes) [2, 4, 5]. In 1986, the three most common anatomical morphologies of the acromion were identified as flat (type I), curved (type II), and hooked (type III) by Bigliani et al., [6] their findings revealed that the hooked acromion has a higher incidence of RCTs. In recent years, the link between the anatomical morphology of the acromion and RCTs has been an area of interest for orthopedic surgeons [5, 7–10]. Researchers have demonstrated that the hooked acromion may reduce the subacromial space, consequently increasing the pressure on the rotator cuff tendons and making them more susceptible to degeneration [1, 8, 11]. Numerous studies have examined the link between variations in acromion morphology and RCTs [2, 12, 13]. Anthropometric techniques have been employed to identify critical parameters, including the acromion index (AI), critical shoulder angle (CSA), acromiohumeral distance (AHD), and lateral acromion angle (LAA). Nyffeler et al., [7] reported that a larger lateral extension of the acromion generates greater upward force during shoulder abduction, increasing the risk of rotator cuff tear. Furthermore, a high Acromion index has been strongly associated with an increased susceptibility to RCTs. The evidence suggests a clear link between acromion morphology and RCTs, making these factors important in diagnosis and treatment. Moor et al., [14] proposed a new radiological parameter called CSA, designed to evaluate an individual’s scapular anatomy. The CSA is determined by measuring the angle between the glenoid’s superior and inferior margins and the acromion’s most lateral point. Assessing the glenoid slope and the lateral extension of the acromion, CSA offers a complete understanding of the occurrence of RCTs [15]. The AHD is defined as the shortest distance between the acromion and the humeral head [16]. It is commonly used to assess the subacromial space, providing diagnostic and prognostic value; AHD is frequently used to indicate rotator cuff integrity [17]. According to Goutallier et al., a reduction in AHD to less than 6 mm is highly specific for a full-thickness rotator cuff tear. It is also associated with a poor prognosis following rotator cuff tear repair [18, 19]. AHD decreases when the humerus migrates superiorly due to a RCTs [20]. Banas et al., [21] provided substantial evidence indicating a relationship between the LAA and alterations in the rotator cuff, as observed through MRI. They reported a link between the lateral tilt of the acromion and the occurrence of subacromial tears, establishing this angle as an effective diagnostic parameter for assessing patients who are RCTs. Research has shown that the AI, CSA, AHD, and LAA play crucial roles in the effective functioning and biomechanical stability of the shoulder joint. In particular, an increase in CSA and AI and a decrease in AHD and LAA may lead to the development of RCTs [7, 22–24]. To better understand shoulder anatomy, it is important to establish standard minimum values for these parameters. By examining how these parameters vary among different populations, we can gain valuable insights into shoulder disorders and their implications for diverse groups. However, studies on these parameters have predominantly been conducted in North American, European, and Asian countries, highlighting the need for research in other populations [25–29]. Several studies worldwide have examined various populations, including those in Brazil [30], Japan [29], South Africa [31], China [23], and Chile [32]. However, no research has yet incorporated these parameters in Somalia. This study aimed to analyze and describe the radiological parameters of acromion morphology in Somali populations and investigate the shoulder joint morphology parameters-such as CSA, AI, AHD, and LAA and their relationship to rotator cuff tear within the Somali population. Methods Patient selection This study was approved by the institutional review board of Kamil Diagnostic center (KDC-EC-23-011). A written informed consent was obtained from every participant. A retrospective screening was conducted on all the patients who underwent computed tomography (CT) scans of the shoulder in the Radiology Department between January 2018 and June 2023. Only Somali patients were included in the analysis. Inclusion criteria of the study were as follows: Patients aged 18 years or older, with no prior history of shoulder trauma or surgery, and having complete scapular CT scans. Exclusion criteria included: CT scans that did not capture the entire scapula, as well as cases of scapular tumors, fractures, bone defects, or patients with osteoarthritis (OA). During the initial visit, patients provided their age, gender, height, weight, and smoking habits. The diagnosis for all patients was made through a combination of physical examination and imaging findings. Radiographic assessment The Digital Imaging and Communications in Medicine (DICOM) data of shoulder scans were obtained from the Radiology Department of Kamil Diagnostic Center and imported into Mimics 21.0 (Materialise, Leuven, Belgium) for 3D reconstructions. The alignment of Three-dimensional (3D) shoulder models was carried out using Blender 2.81 software (Amsterdam, the Netherlands), following the methods described in previous studies [5, 33]. To standardize the reference system, a Cartesian coordinate system was established to align each 3D shoulder model. The aligning process involved several primary steps. First, the scapular plane of each 3D model was identified using three anatomical landmarks, the inferior scapular angle (I), the intersection of the scapular spine and the medial border of the scapula (M), and the center of the best-fit circle of the glenoid (C), which corresponds to the center of the inferior glenoid [33, 34]. The x-axis was then defined by the line connecting points M and C, while the y-axis was established as perpendicular to the scapular plane (C-I-M plane), and the z-axis was calculated as the cross product of the x- and y-axes. This process created a Cartesian coordinate system with the X-, Y-, and Z-axes, the anterior view of the 3D model was positioned perpendicular to the XZ-plane (scapular plane). as shown in (Fig. 1A, B). Coordinate system establishment of the scapular plane of each 3D model (right shoulder) was identified using three anatomical landmarks, the inferior scapular angle (I), the intersection of the scapular spine and the medial border of the scapula (M), and the center of the best-fit circle of the glenoid (C), which corresponds to the center of the inferior glenoid. The x-axis was then defined by the line connecting points M and C, while the y-axis was established as perpendicular to the scapular plane (C-I-M plane), and the z-axis was calculated as the cross product of the x- and y-axes Measurement of acromial parameters on 3-D shoulder models In the study, CSA, AI, AHD, and LAA were evaluated through an anterior Three-dimensional computed tomography (3D-CT) view of the scapula using Blender 2.81 software. AI was determined as the ratio of the distance between the glenoid plane and the lateral border of the acromion (GA) to the distance between the glenoid plane and the most lateral edge of the humeral head (GH) [7] (Fig. 2A). AHD measured the distance between the parallel line drawn from the inferior end of the acromion and the parallel line tangent to the humeral head, as described by Werner et al., [35] (Fig. 2B). CSA was calculated as the angle formed by the glenoid plane and a line connecting the inferior borders of the glenoid and the most inferolateral point of the acromion [15] (Fig. 2C). Lastly, LAA was measured as the angle between the glenoid plane and a line that runs parallel to the undersurface of the acromion [24] (Fig. 2D). Measurements of (A) Acromial index, (B) Acromiohumeral distance, (C) Critical shoulder angle, and (D) Lateral acromial angle Statistical analysis The study presents continuous variables as mean and standard deviation (mean ± SD) and categorical variables are represented by frequency and percentages. Two independent assessors, one radiologist and one shoulder and elbow surgeon with more than five years of experience, evaluated the acromial parameters. Reliability analysis assessed the Intraclass Correlation Coefficient (ICC) between the two observers, with an ICC value of 0.75 or higher considered reliable [36]. An independent sample t-test was used for continuous variables to calculate the difference between the groups. In contrast, the chi-square test was employed for categorical variables to compare the differences between the two groups. Spearman correlation analysis was conducted to explore the correlation between AI, CSA, AHD, LAA, and RCT, and an area under the ROC curve (AUC) analysis was used to compare the diagnostic efficiency of these parameters for RCT. A P-value of less than 0.05 was considered statistically significant. SPSS Version 25.0 (IBM, Armonk, NY, USA) was used for statistical analysis. Results Initially, 260 patients met the inclusion criteria for our study. However, 35 were excluded due to fractures of the scapula, humerus, and glenoid or due to tumors. An additional 22 patients were excluded due to poor-quality shoulder CT scans, and 15 were excluded due to glenoid defects. In the final analysis,188 patients were included. Among these, 107 patients had RCTs and 81 had intact rotator cuffs but presented with conditions such as frozen shoulder, SLAP lesions, or other associated abnormalities (Fig. 3). The baseline characteristics were largely comparable between the two groups, with a few notable differences. Patients in the RCT group were significantly older than those in the non-RCT group (p < 0.001). Additionally, the RCT group had a significantly higher proportion of males compared to females (p < 0.003). The summary of the patient’s characteristics can be found in (Table 1). Flowchart of the study process Table 1Baseline characteristics of the study patientsVariablesTotalRCTnon-RCTStatistic valueP Value(N = 188)(N = 107)(N = 81)Age, Mean ± SD48.70 ± 16.4357.42 ± 13.9435.98 ± 10.36< 0.001* Sex χ2 = 8.8370.003* Male119(63.3)58(54.2)61(75.3) Female69(36.7)49(45.8)20(24.7)BMI, Mean ± SD26.55 ± 2.7126.54 ± 2.7926.55 ± 2.613t = -0.1320.895 Dominant χ2 = 1.1010.294 Right157(83.5)92(86.0)65(80.3 Left31(16.5)15(14.0)16(19.7) Effected Side Right102(54.3)61(57.0)41(50.6)χ2 = 0.7590.384 Left86(45.7)46(43.0)40(49.4) MW χ2 = 9.5510.002* Yes41(21.8)32(29.9)9(11.1) No147(78.2)75(70.1)72(88.9) Diabetes χ2 = 0.0720.788 Yes60(31.9)35(32.7)25(30.9) No128(68.1)72(67.3)56(69.1) Hypertension χ2 = 0.9030.342 Yes41(21.8)26(24.3)15(18.5) No147(78.2)81(75.7)66(81.5) Smoking χ2 = 1.3480.246 Yes39(20.7)19(17.8)20(24.7) No149(79.3)88(82.2)61(75.3)* Each difference between groups was statistically significant (P < 0.05;χ2 test or independent-samples t-test) Baseline characteristics of the study patients * Each difference between groups was statistically significant (P < 0.05;χ2 test or independent-samples t-test) The measured acromial parameters, including AI, CSA, AHD, and LAA, were independently assessed on 3D-CT scans of the shoulder joint by two independent assessors. The results demonstrated excellent intra- and inter-observer reliability, with all ICC values above 0.75. Among these, the CSA exhibited the greatest measurement stability, with an intra-observer ICC of 0.929 and an inter-observer ICC of 0.911. A summary of the data is presented in (Table 2). Table 2The inter- and intra-observer reliability for parameter in the studyParameterMean ± SDICC (95% CI)P ValueAIAssessor 1–10.7167 ± 0.0630Assessor 1–20.7152 ± 0.72820.877(0.804–0.907)< 0.001*Assessor 20.7179 ± 0.07410.862(0.820–0.985)< 0.001*CSAAssessor 134.52 ± 3.09Assessor 234.50 ± 3.550.929(0.906–0.946)< 0.001*Assessor 234.51 ± 3.640.911(0.884–0.933)< 0.001*AHDAssessor 17.23 ± 2.27Assessor 27.30 ± 2.560.867(0.826–0.898)< 0.001*Assessor 27.17 ± 2.710.843(0.796–0.880)LAAAssessor 172.03 ± 6.11Assessor 272.57 ± 6.860.896(0.864–0.921)< 0.001*Assessor 272.57 ± 6.860.859(0.816–0.893)< 0.001** Each difference between groups was statistically significant (P < 0.05; Reliability Analysis) The inter- and intra-observer reliability for parameter in the study * Each difference between groups was statistically significant (P < 0.05; Reliability Analysis) Significant positive correlations were identified between AI, CSA, and the presence of RCTs, with CSA exhibiting the strongest correlation (r = 0.629; P < 0.001). Additionally, significant negative correlations were found between the presence of RCTs and both AHD (–0.247) and LAA (–0.338; P < 0.001 for both) (Table 3). Table 3Evaluation of the relationship between shoulder parameters and the Presence of RCTPresence of RCTParameter r P ValueAI0.325**< 0.001*CSA0.629**< 0.001*AHD-0.247**< 0.001*LAA-0.338**< 0.001*** At the 0.01 level (one-tailed), there is a significant correlation. * Each correlation was statistically significant (P < 0.05; r, Spearman correlation analysis) Evaluation of the relationship between shoulder parameters and the Presence of RCT ** At the 0.01 level (one-tailed), there is a significant correlation. * Each correlation was statistically significant (P < 0.05; r, Spearman correlation analysis) When comparing shoulder parameter measurements between the RCT group and the non-RCT group, significant differences were observed for all parameters: AI (P < 0.001), CSA (P < < 0.001), AHD (P < 0.001), and LAA (P < 0.001) (Table 4). In examinations of CSA and AI, measurements were significantly lower in the non-RCT group compared to the RCT group. Conversely, measurements for LAA and AHD were significantly higher in the non-RCTs group than in the RCTs group. Table 4Comparison of shoulder parameters between the RCT and non-RCT groupsVariablesMean ± SDRCTnon-RCTP Value(N = 188)(N = 107)(N = 81)AI, Mean ± SD0.72 ± 0.060.73 ± 0.060.70 ± 0.57< 0.001*(t = 24.92)CSA, Mean ± SD34.80 ± 3.0736.22 ± 2.9132.73 ± 1.91< 0.001*(t = 58.33)AHD, Mean ± SD7.15 ± 2.226.70 ± 2.107.81 ± 2.23< 0.001*(t=-20.15)LAA, Mean ± SD71.89 ± 6.0970.32 ± 5.2974.18 ± 6.45< 0.001*(t=-25.40)*Each difference between groups was statistically significant (P < 0.05; independent-samples t-test) Comparison of shoulder parameters between the RCT and non-RCT groups *Each difference between groups was statistically significant (P < 0.05; independent-samples t-test) ROC curve analysis was conducted on all shoulder parameters to assess the presence of RCTs. The cutoff values established for distinguishing between the RCT group and the non-RCT group were 0.713 for the AI, 34.25 for the CSA, 7.925 for the AHD, and 70.2 for the LAA. The results of the ROC and post hoc analyses are illustrated in (Fig. 4). Notably, CSA proved to be the most effective parameter for identifying the presence of RCTs, achieving an area under the ROC curve of 0.857, compared to 0.687 for AI (Table 5). While other parameters also showed some discriminatory power, they were less effective in distinguishing between the groups. Receiver Operating Characteristic (ROC) Curve Analysis of the Entire Cohort. The ROC curve illustrates the effectiveness of various shoulder parameters in predicting the presence of a rotator cuff tear (RCT). A higher area under the ROC curve (AUC) reflects greater diagnostic accuracy for the presence of RCT. AI, acromial index; CSA, critical shoulder angle; AHD, acromiohumeral distance; and LAA lateral acromial angle Table 5Effectiveness of shoulder parameters to Determine the Presence of RCTParameterAUC95%CISEP ValueAI0.6870.621–0.7540.040< 0.001*CSA0.8570.804–0.9110.028< 0.001 *AHD0.6390.572–0.7070.041< 0.001 *LAA0.6910.628–0.7550.040< 0.001**Each AUC was statistically significant (P < 0.05) Effectiveness of shoulder parameters to Determine the Presence of RCT *Each AUC was statistically significant (P < 0.05) Discussion While numerous studies in the literature have evaluated the impact of acromion morphological parameters on RCTs across various populations, this is the first study to assess these factors in the Somali population specifically. Nyffeler et al., [7] first introduced AI in 2006. This parameter, which is believed to help detect early rotator cuff degeneration, is significantly higher in patients with RCTs. Numerous studies have demonstrated that a high AI is a predisposing factor for RCTs [2, 27, 28, 37]. Miyazaki et al., [38] reported contrasting results from two different populations, showing that AI may be a risk factor for RCTs in a Brazilian but not a Japanese population. However, Moor et al., [28] and Balke et al., [26] have described those distinct scapular morphologies, including AI, as reliable radiological predictors of degenerative RCTs. Similarly, Kum DH et al., [37] found that AI was an effective predictive factor for RCTs in a Korean population. These findings are consistent with our study, where AI demonstrated a statistically significant difference between groups (P < 0.001). According to a biomechanical study, a higher CSA results in increased shear force at the humeral head and reduced direct compressive force in the glenoid, generating a superiorly directed deltoid force vector which causes the supraspinatus to work harder to counterbalance the elevated shear force and superior deltoid force [39]. The increased workload on the supraspinatus ultimately contributes to developing rotator cuff tears (RCT) [15, 39]. Our results showed that CSA in the RCT group was higher than in the non-RCT group (36.22° vs. 32.73°). These results in Somalia are consistent with recent evidence from other countries, which confirms that an increased CSA is associated with RCTs [15, 23, 28, 32, 40]. All of these studies reported that patients with RCTs had a higher CSA compared to non-RCT patients. A systematic review established a strong correlation between larger CSA and the presence of RCTs [34, 41]. Additionally, it has been reported that a higher CSA increases the risk of retears in repaired RCTs [40]. The AHD is between the acromion and the humeral head. Goutallier et al., [18] reported that an AHD of less than 6 mm is associated with degenerative RCTs. A decreased AHD, particularly during shoulder abduction, increases impingement between the humeral head and acromion, accelerating rotator cuff degeneration. In our study, we observed that the mean AHD was 1.11 mm lower in the RCT group compared to the non-RCT group, consistent with findings in the literature. Similarly, Weiner and Macnab found that 90% of patients with an AHD of less than 7 mm had RCTs [42], and a retrospective study by Kaur et al., [13] showed that AHD was lower in patients with partial and complete RCTs than the control group. This reduction in AHD is caused by superior migration of the humeral head observed in cases of RCTs, suggesting an association with tendinopathy [43–45]. However, while our study demonstrates this association, it is important to note that the present study does not establish AHD as a definitive risk factor for RCTs. In future studies, we aim to further investigate whether AHD is a risk factor or a consequence of RCTs. Understanding the causal relationship will be critical to determining the role of AHD in shoulder pathology. The relationship between LAA and RCTs was first described by Banas et al. [21] who reported that a low LAA is a risk factor for RCT. Supporting this, Tetreault et al. found that smaller LAA values reduce the subacromial space, increasing pressure on the rotator cuff [46]. Yu et al., [47] also reported that geriatric patients with full-thickness RCTs had a significantly smaller mean LAA than healthy geriatric participants. According to a systematic review and meta-analysis by Zaid et al., [48] a low LAA is significant risk factor for RCTs. Additionally, an LAA below 70° was observed exclusively in patients with RCTs and not in those with healthy rotator cuff [2, 24]. In a study by Balke et al., [2] the mean LAA was 77° in patients with RCTs, compared to 84.1° in the control group. In the present study, the mean LAA was found to be significantly lower by 3.86° in the RCT group compared to the non-RCT group. In this study, we utilized 3D-CT to measure shoulder anatomical parameters, including AI, CSA, AHD, and LAA. Using 3D-CT provides significant advantages over traditional X-rays, as it allows for accurate and reliable measurement of these parameters by establishing a standardized anterior view through scapular rotation and utilizing bony markers. As reported in previous studies, 3D-CT is a suitable alternative for measuring the CSA and offers valuable insights into the assessment of RCTs [33, 49]. In summary, 3D-CT improves the accuracy of anatomical evaluations, enabling enhanced preoperative planning and more effective clinical decision-making for shoulder disorders. The results of this study carry important clinical implications for the treatment of shoulder disorders within the Somali population. By pinpointing morphological factors, such as AI, CSA, AHD, and LAA, as significant risk factors for rotator cuff tears, healthcare providers can improve diagnostic precision and enhance screening processes for individuals experiencing shoulder pain. Additionally, these findings can guide the development of personalized treatment strategies for rotator cuff tears, influencing surgical decisions and preventive measures aimed at enhancing overall clinical outcomes. Timely identification of patients at higher risk allows for prompt interventions, ensuring that care is customized to meet the distinct anatomical traits of Somali patients. Strengths and limitations To the best of our knowledge, this study was the first to investigate the acromion morphological parameters within the Somali population. This study has certain limitations. First, this was a retrospective study. Second, selection bias could exist since some patients with poor CT scan quality were excluded from the study. Third, a significant limitation of this study is the age discrepancy between the RCTs and non-RCTs groups (mean age 57.42 ± 13.94 vs. 35.98 ± 10.36). Research has shown that older age is strongly associated with the incidence of rotator cuff tears, often more so than CSA or other morphological features [29, 50]. This age discrepancy may suggest that the higher incidence of RCTs in the older group is primarily attributable to their age rather than solely to differences in morphological parameters. Given this potential bias, caution is warranted when interpreting our findings. Future studies employing age-matched cohorts are necessary to isolate the effects of morphological parameters on RCTs risk and provide clearer insights into their role in shoulder pathology. Fourth, while numerous morphological factors contribute to shoulder joint diseases, we only selected theAI, CSA, AHD and LAA for comprehensive analysis, other factors such as the Superior glenoid and inclination (SGI), Acromion Slope (AS), and Sagittal Acromion Tilt (sAT) were not included. This could be a bias into our results, especially for patients with chronic rotator cuff injuries. Fifth, our study was performed exclusively within a Somali population, and the findings may not be generalizable to populations in other regions. However, despite these limitations, the study demonstrated the significant role that AI, CSA, AHD, and LAA play in rotator cuff tears. Conclusion The outcome of this study shows that AI, CSA, AHD, and LAA are independent factors associated with rotator cuff tears in the East African, particularly the Somali population. Especially larger AI, CSA, and smaller AHD and LAA are correlated with an increased likelihood of RCTs in this population. In patients with shoulder disease suspected of RCTs, using a 3D-CT scan to measure the CSA may be helpful, as it is the best predictor measurement. Acknowledgements We would like to express our sincere gratitude to all co-authors for their invaluable contributions to this study. We also extend our thanks to the editors and reviewers for their valuable advice and guidance in revising this manuscript. Their constructive feedback has been instrumental in enhancing the quality of our work. Abbreviations Acromial index Critical shoulder angle Acromiohumeral distance Lateral acromial angle Rotator cuff tear Non-Rotator Cuff Tear Three-dimensional computed tomography Digital imaging and communication in medicine Computed Tomography Three-Dimensional Body mass index Manual worker Intraclass correlation coefficient Area under the curve Confidence Interval Standard error Spearman correlation analysis Author contributions Rui Yang conceived and designed the study. J.Y.H designed the study and Writing - review & editing. MI. Ali Writing—Original Draft and performed the data analysis. Y.Q.C, D.X.H and Y.T.Y performed the data analysis and manuscript preparation. Y.L created the graphs and analysis. D.O.H and A.H. A contributed to data collecting.* Maslah Idiris Ali, Yuan-quan Chen and Deng Xing-Hao contributed equally to this work and are the co-first authors. Funding This study was funded by the Fundamental Research Funds for the Central Universities, Sun Yat-sen University (2020004), and National Natural Science Foundation of China (no. 82002342). Data availability No datasets were generated or analysed during the current study. Declarations Ethics approval and consent to participate institutional review board of Kamil Diagnostic center approved this study (KDC-EC-23-011). A written informed consent was obtained from every participant, and all ethical rules were considered during the study. The relevant guidelines by the Declaration of Helsinki were followed within the study. Consent for publication Not Applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Maslah Idiris Ali, Yuan-quan Chen and Deng Xing-Hao contributed equally to this work and are the co-first authors. Contributor Information Jing-yi Hou, Email: houjy7@mail.sysu.edu.cn. Yi Long, Email: longy68@mail.sysu.edu.cn. Rui Yang, Email: yangr@mail.sysu.edu.cn. References Associated Data Data Availability Statement No datasets were generated or analysed during the current study.