Skip to main content
NCBI home page
Shoulder & Elbow logoLink to Shoulder & Elbow
. 2020 Mar 24;13(4):388–395. doi: 10.1177/1758573220910093

The effect of load and plane of elevation on acromial stress after reverse shoulder arthroplasty

Jason S Lockhart 1, Murray T Wong 1, G Daniel G Langohr 1, George S Athwal 1,✉, James A Johnson 1
PMCID: PMC8355643  PMID: 34394736

Abstract

Background

Acromial fractures are a substantial complication following reverse shoulder arthroplasty, reported to affect up to 7% of patients. Previous studies have shown that implant placement affects acromial stress during elevation of the arm in the scaption plane. The purpose of this study was to investigate the results of arm loading and variation in plane of elevation on acromial stresses.

Methods

Nine elevation angles (0°–120°), in three planes of elevation (abduction (0°), scaption (30°), and forward elevation (60°)), and three hand loads (0, 2.5, 5 kg) were investigated. Finite element models were generated using computed tomography data from 10 cadaveric shoulders (age 68 ± 19 yrs) to determine acromial stress distributions. Models were created for a lateralized glenosphere (0, 5, 10 mm), inferiorized glenosphere (0, 2.5, 5 mm), and humeral offset (−5, 0, 5 mm).

Results

For all planes of elevation (0°, 30°, 60°) and hand loads (0, 2.5, 5 kg) investigated, glenoid lateralization consistently increased acromial stress, glenoid inferiorization consistently decreased acromial stress, and humeral offset proved to be insignificant in altering acromial stress. Abduction resulted in significantly higher peak acromial stresses (p = 0.002) as compared to scaption and forward elevation.

Conclusions

In addition to implant position and design, patient activity, such as plane of elevation and hand loads, has substantial effects on acromial stresses.

Level of evidence

Basic science study

Keywords: reverse total shoulder arthroplasty, complication, acromial fracture, stress fracture, shoulder, arthroplasty

Introduction

Reverse total shoulder arthroplasty (RTSA) is a surgical procedure that reverses the native shoulder joint to restore stability and range of motion to the rotator cuff deficient patient. The center of rotation (COR) with RTSA is shifted medially compared to the native shoulder. The artificially created COR position increases deltoid participation and decreases the force required for elevation of the arm.1

Postoperative acromial fractures are a common complication after RTSA that are reported in up to 7% of patients.2 Acromial fractures have been found to lead to inferior clinical outcomes, resulting in decreased range of motion, weakness, and pain.2–4 The cause of postoperative fracture is thought to be fatigue failure due to increased stress levels above the operational microdamage (yield) threshold of bone (approximately 60 MPa), rather than a traumatic fracture caused by a single stress above the ultimate strength of bone (approximately 120 MPa).5,6 Note, the bone thresholds provided are approximated for healthy bone and can vary by case amongst individuals, and in particular with respect to the elderly and osteoporotic population.5 Significant factors that affect acromial fatigue fracture include the alteration of moment arms created by the deltoid due to RTSA biomechanics, and osteopenia, which is common in the elderly RTSA patient population.

The effect of implant placement on acromial stress during elevation in the 30° scaption plane has been previously studied.7 It was found that an inferior medial glenosphere and a medialized humeral implant decrease acromial stress during scaption (arm elevation in the scapular plane). However, knowledge related to the effect of implant placement on acromial stress is limited to the unloaded scenario in the scaption plane of elevation. As such, the purpose of this finite element analysis study was to determine the effect of implant position on acromial stress when factors such as the plane of elevation (0°, 30°, 60°) and loading (0, 2.5, 5 kg) were varied. We hypothesized that (1) due to the unique morphology of the scapula and acromion, varying the plane of elevation would affect stresses, and that (2) acromial stresses would increase to above the failure threshold of cortical bone with increased hand load.

Materials and methods

Three-dimensional models of the glenohumeral joint were created from CT data of 10 cadaveric specimens (Figure 1, average age: 68 ± 19 yrs). Wickham et al.8 identified seven independent deltoid segments and their insertions, which were used to model deltoid loading as seven force vectors applied to their anatomic areas of insertion on the scapula and humerus. The scapula and humerus were both oriented based on the International Society of Biomechanics standards.9 To match in vivo position the scapula was rotated 10° about the Zs axis to account for the anterior tilt.10 For arm elevation in the scaption plane (30°), the humerus was constrained to the glenoid using CT data for accurate positioning, and the humeral XY plane (XYh) was set parallel to the XY plane of the scapula (XYs) to initialize the humerus in the adduction position. The humerus was rotated about the X axis of the scapula’s coordinate system (Xs) to simulate elevation. To perform arm elevation in other planes of elevation, the humeral coordinate system was rotated about the Y axis (YH) by the corresponding angle (−30° or 30°).

Figure 1.

Figure 1.

Open in a new tab

Anatomical mapping used to set up the mathematic algorithm (described later). (I) The three planes of elevation studied. (II) The seven lines of action for the seven Wickham et al.8 segments of the deltoid in the 0° adduction position. (III) The scapula and humerus coordinate systems with origins.

A traditional reverse total shoulder implant (38 mm glenosphere, 155° neck-shaft angle, 20 mm humeral offset) was reconstructed using SolidWorks (Dassault Systèmes, France) to represent a shoulder post-RTSA. In the baseline implant configuration, the glenoid baseplate was placed at the inferior edge of the glenoid fossa with zero lateralization. RTSA implant position was altered by increasing glenosphere lateralization (GLAT) (0, 5, 10 mm) and inferiorization (0, 2.5, 5 mm), as well as varying the humeral component offset (15, 20, 25 mm, Figure 2).

Figure 2.

Figure 2.

Open in a new tab

The three implant placement parameters that were varied in the model: (I) GLAT (0, 5, 10 mm), (II) GINF (0, 2.5, 5 mm), and (III) HLAT (15, 20, 25 mm).

The force that each of the seven deltoid segments applies to the acromion was calculated by a mathematical algorithm (Matlab, MathWorks, Massachusetts) at static positions during arm elevation between 0° and 120° at 15° increments for RTSA models in all implant configurations and hand loads. The algorithm utilizes a moment balance based on the vectors applied by each deltoid segment, hand load, and arm weight as well as their respective distances from the joint’s COR, as previously reported.7

Loading was applied to model arm lifting during everyday tasks. Loads of 0, 2.5, and 5 kg were applied to the hand along with a patient mass of 75 kg11 for elevation in all planes (0°, 30°, 60°). The resultant moment was calculated as a cross product of the weight of the applied load and the distance from the COR of the shoulder joint to the center of mass of the hand. The distance between the COR of the shoulder joint and the center of mass of the hand was determined using anthropometry.12

The deltoid forces output from the moment balance was input into Abaqus (Dassault Systèmes, France), a finite element (FE) software to compute the stress on the acromion. The FE model consisted of the scapula segmented into 160,000 ± 20,000 quadratic tetrahedral elements, sized according to a previously performed mesh convergence study.13 The inferior portion of the scapula was rigidly constrained and the deltoid forces were tied to each segment’s physiologic insertion region on the acromion.8 The outcome variable was the maximum principal stress in cortical bone. The maximum principal stress is the variable most closely related to the ultimate failure behavior of cortical bone. Outputting peak stress levels in the acromion provided trends in acromial stress by investigating relative results between configurations; however, maximums represent a small area of the bone, meaning the stress levels may exceed the threshold of bone strength without the acromion fracturing.

This process was completed for each of the 0° (abduction), 30° (scaption), and 60° (forward elevation) planes of elevation. For each plane of elevation, the paths between the deltoid origins and COR vary. In the investigation of plane of elevation and loading, 2187 individual cases (nine angles, nine implant positions, three planes of elevation, three hand loads) were tested on 10 different cadaveric based models for a total of 21,870 simulations.

Acromial stress was also evaluated based on the location on the acromion. The acromion was split into three regions as defined by Levy et al.14 (Figure 3) based on the types of fractures that are clinically observed. Region I involves fractures caused by the anterior and middle deltoid segments, region II involves the entire middle deltoid segments, and region III involves both posterior and middle deltoid segments.

Figure 3.

Figure 3.

Open in a new tab

Visualization of the Levy regions on the acromion. Each region is defined by the deltoid segments that cause fractures localized to said region (region I: anterior and middle, region II: middle, region III: posterior, middle).8

Three-way repeated-measures ANOVAs (abduction angle, plane of elevation, hand load) were performed for each implant position across each specimen.

Results

All percentages herein are reported as percent difference to compare between experimental measures.

Plane of elevation

The peak acromial stress observed for neutral implant positioning during unloaded humeral elevation from 0° to 120° in the scaption plane was 24 ± 4 MPa occurring at 45° of elevation (Figure 4). In abduction, the peak acromial stress was 61 ± 6 MPa occurring at 75° of elevation, while in forward elevation, the peak stress was 24 ± 3 MPa occurring at 90° of elevation.

Figure 4.

Figure 4.

Open in a new tab

Maximum acromial stress during elevation (0–120°) for the baseline implant configuration in all three planes of elevation ((I) abduction = 0°, (II) scaption = 30°, (III) forward elevation = 60°) with loading (0, 2.5, 5 kg).

Humeral elevation in abduction resulted in an increase in peak acromial stress of 58 ± 3.0% (+20.0 ± 14 MPa, p = 0.002) over all elevation angles compared to the scaption plane, whereas elevation in the forward plane decreased peak acromial stress insignificantly by 10 ± 30% (−3 ± 8 MPa, p = 0.28), compared to the scaption plane.

Loading

Increasing hand load from 0 to 5 kg in the scaption plane increased peak acromial stress by 109 ± 4% (39 ± 9 MPa) over all elevation angles, compared to the unloaded state (p < 0.001, Figure 4). During elevation in the abduction plane, increasing hand load from 0 to 5 kg increased the peak acromial stress by 102 ± 6% (59 ± 9 MPa) over all elevation angles, compared to the unloaded state (p < 0.001, Figure 4). Increasing hand loading in the forward elevation plane increased the peak acromial stress by 98 ± 14% (38 ± 10 MPa) over all elevation angles, compared to the unloaded state (p < 0.001, Figure 4). Increasing hand weight generally affected all planes of elevation similarly and raised stress levels above the yield threshold (60 MPa) for cortical bone.

GLAT

For all planes of elevation (0°, 30°, 60°) and hand loads (0, 2.5, 5 kg) investigated, glenoid lateralization (GLAT) consistently increased acromial stress (Figure 5).

Figure 5.

Figure 5.

Open in a new tab

Change in maximum acromial stress as a function of implant placement and applied load in each plane of elevation ((I) 0°, (II) 30°, (III) 60°). Implant placement includes GLAT, GINF, HMED, and HLAT. GINF: glenosphere inferiorization; GLAT: glenosphere lateralization; HLAT: humeral component lateralization; HMED: humeral component medialization.

The increase in maximum stress on the acromion as a result of GLAT in abduction was 7 ± 1% (4 ± 1 MPa, p < 0.001) for 5 mm and 16 ± 2% (9 ± 1 MPa, p < 0.001) for 10 mm; compared to the scaption plane where the increases were 9 ± 1% (4 ± 0 MPa, p < 0.001) for 5 mm and 19 ± 2% (8 ± 1 MPa, p < 0.001) for 10 mm; and in the forward plane 11 ± 3% (4 ± 1 MPa, p = 0.009) for 5 mm and 21 ± 3% (8 ± 1 MPa, p < 0.001) for 10 mm.

Glenosphere inferiorization (GINF)

For all planes of elevation (0°, 30°, 60°) and hand loads (0, 2.5, 5 kg) investigated, glenoid inferiorization (GINF) consistently decreased acromial stress.

The decrease in peak stress in the acromion as a result of GINF in abduction over all elevation angles was 3 ± 1% (1.8 ± 0.3 MPa, p = 0.001) for 2.5 mm and 5 ± 1% (3 ± 1 MPa, p = 0.002) for 5 mm; compared to the scaption plane where the decrease was 3.6 ± 0.4% (1.4 ± 0.2 MPa, p < 0.001) for 2.5 mm and 6 ± 1% (2.5 ± 0.3 MPa, p < 0.001) for 5 mm. The decrease in acromial stress in the forward elevation plane was 2 ± 1% (0.8 ± 0.4 MPa, p = 0.3) for 2.5 mm and 5 ± 2% (2 ± 1 MPa, p = 0.041) for 5 mm.

Humeral component medialization (HMED) and lateralization (HLAT)

Humeral medial and lateral offset did not significantly affect acromial stresses (Figure 6).

Figure 6.

Figure 6.

Open in a new tab

Maximum acromial stress as a function of implant placement (I) GLAT, (II) GINF, (III) HMED and HLAT and abduction angle compared to standard positioning averaged over all loads, regions, and planes of elevation. GINF: glenosphere inferiorization; GLAT: glenosphere lateralization; HLAT: humeral component lateralization; HMED: humeral component medialization.

Stress by location on the acromion

The average stress acting on each Levy region14 of the acromion during arm elevation over all implant configurations, loads, and planes of elevation was found to consistently be highest in region II (p < 0.001), followed by region III (p = 0.05) with the lowest stress occurring in region I of the acromion (II>III>I, Figure 7).

Figure 7.

Figure 7.

Open in a new tab

Maximum acromial stress as a function of acromial regions and abduction angle averaged over all loads, implant placements, and planes of elevation.8

Discussion

Variations in the plane of elevation (abduction, scaption, and forward elevation) and the amount of weight the hand is lifting have significant effects on the stresses the acromion is experiencing. Acromial stresses were significantly higher for humeral elevation in the abduction plane as compared to more forward planes of elevation. This was likely a result of the unique shape of the scapular spine and acromion, as it is relatively unsupported laterally. As the middle aspect of the deltoid is largely responsible for abduction, it would follow that greater stresses would be experienced by the acromion in abduction as it overhangs the shoulder analogous to a diving board. Additionally, indirect muscles paths associated with the abduction plane may be introducing a more aggressive loading state. Understanding high stress arm positions may be beneficial in the early postoperative period to minimize exposure of the acromion to stresses that are in range of the failure thresholds of cortical bone (60 MPa yield threshold and 120 MPa ultimate threshold).

Inferior glenosphere positioning significantly reduced acromial stress in all planes and for all loads, agreeing with results previously reported for unloaded scaption.7 GINF decreases acromial stress likely because the shoulder joint’s COR is shifted causing the moment arm of the deltoid to lengthen. In a moment balance, an increased moment arm reduces the corresponding deltoid force required to combat gravity and hand loads. Decreased deltoid forces lead to decreased acromial stress because deltoid forces are applied directly to the surface of the acromion. The effect of GINF on stress increases in magnitude in a linear trend with respect to applied load (Figure 5) and exhibits similar behavior in all three planes of motion suggesting that this behavior is consistent over the entire range of motion.

GLAT resulted in a significant increase in acromial stress likely a result of the alteration of stress in the opposite manner of inferiorization; by shifting the joint’s COR laterally, the deltoid’s moment arm is shortened. Similar to GINF, the effect of lateralization scales in magnitude with respect to applied load. The effect of GLAT is magnified as applied load increases, making the effect of GLAT more relevant in high stress scenarios. Humeral medial and lateral offset produced an insignificant effect on stress over varying planes of elevation. The effects of humeral offset on stress do not scale with applied load making humeral offset less of a factor when hand loads are greater, which is when stress levels are at their highest. Therefore, humeral offset may have a diminishing effect on acromial stress as stress levels approach the yield threshold (60 MPa). The yield (operational microdamage) threshold of 60 MPa for cortical bone is the level at which the rate of damage is greater than the rate of repair. If bone is subject to stress above this threshold for an extended period of time or over multiple events, a stress fracture may occur.6 Elevation of the arm under the application of hand loading shows that the acromion of an RTSA patient experiences stress above the operational microdamage threshold as a result of hand loads consistent with everyday tasks. Furthermore, the ultimate strength of cortical bone (120 MPa) was exceeded in the coronal plane (0°) and approached in the scaption (30°) and forward flexion planes (60°) at larger hand loads (5 kg). Based on the results from this computational study, hand loads introduce stress to the acromion that is consistent with multiple mechanisms of cortical bone failure of cortical bone. Additionally, when incorporating required hand lifting with plane of elevation, patients may be counseled to lift in forward elevation, if required, rather than in abduction.

The limitations of this study are associated with implementing a mathematical model and the necessary assumptions. The shoulder joint is modeled using the assumption that it is a static one-dimensional moment balance. The only muscle considered for elevation in this model is the deltoid, and it is assumed that contraction of the deltoid follows the activation-squared-minimization principle. These assumptions have been found to mimic the physiologic case after RTSA with reasonable accuracy.15–18 The mass of the patient is assumed to be 75 kg with average anatomical segmentation in order to represent the 50th percentile male, and hence results may vary as patient mass varies.11 However, due to the parametric nature of this study, we are confident that relative results would be similar.

The strength of using a mathematical model is the capability to test an extensive number of implant scenarios with precise positioning. In this study, comparisons are made between configurations, which improve accuracy as relative results eliminate some systematic error associated with the assumptions at the foundation of the model. The model’s accuracy is also strengthened by the fact that average stress in acromial regions follows the expected magnitude order of regions II>III>I determined by Levy et al.,14 suggesting appropriate proportioning of force relative to the anterior, middle, and posterior deltoid.

Overall, the results of this study show that humeral elevation in the forward planes produces lower acromial stresses than elevation in the abduction plane. The addition of weight to the hand increased acromial stresses exceeding the yield threshold and approaching the ultimate strength of cortical bone for all planes of elevation investigated. The results of this study extend current knowledge regarding RTSA implant designs, in that inferior positioning of the glenoid has a positive effect on acromial stress, while lateralization has a negative effect. These effects scale as load is increased making it important in loaded scenarios that may lead to stress fractures. Additionally, humeral component offset, both lateral and medial, was observed to have minimal effect on acromial stresses for the planes of elevation and hand loads examined. These results confirm that implant parameters, plane of elevation, and the weight an arm lifts have significant effects on acromial stresses.

Contributorship: JL and MW researched literature. All authors were involved in study design, methods, and reviewing the manuscript. JL performed data analysis and wrote the first draft.

Guarantor: GA.

Declaration of Conflicting Interests: The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: George S. Athwal is a consultant for Wright Medical and Exactech. No company was involved with this study.

Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD: Jason S Lockhart https://orcid.org/0000-0002-8516-5546

Ethical Review and Patient Consent

Not applicable.

References

  • 1.Hsu SH, Greiwe RM, Saifi C, et al. Reverse total shoulder arthroplasty-biomechanics and rationale. Oper Tech Orthop 2011; 21: 52–59. [Google Scholar]
  • 2.Hamid N, Connor PM, Fleischli JF, et al. Acromial fracture after reverse shoulder arthroplasty. Am J Orthop 2011; 40: E125–E129. [PubMed] [Google Scholar]
  • 3.Zhou HS, Chung JS, Yi PH, et al. Management of complications after reverse shoulder arthroplasty. Curr Rev Musculoskelet Med 2015; 8: 92–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Teusink MJ, Otto RJ, Cottrell BJ, et al. What is the effect of postoperative scapular fracture on outcomes of reverse shoulder arthroplasty? J Shoulder Elbow Surg 2014; 23: 782–790. [DOI] [PubMed] [Google Scholar]
  • 5.Frost HM. Bone’s mechanostat: a 2003 update. Anat Rec A Discov Mol Cell Evol Biol 2003; 275: 1081–1101. [DOI] [PubMed] [Google Scholar]
  • 6.Pepper M, Akuthota V, McCarty EC. The pathophysiology of stress fractures. Clin Sports Med 2006; 25: 1–16. [DOI] [PubMed] [Google Scholar]
  • 7.Wong MT, Langohr GDG, Athwal GS, et al. Implant positioning in reverse shoulder arthroplasty has an impact on acromial stresses. J Shoulder Elbow Surg 2016; 25: 1889–1895. [DOI] [PubMed] [Google Scholar]
  • 8.Wickham JB, Brown JMM, McAndrew DJ. Muscles within muscles: anatomical and functional segmentation of selected shoulder joint musculature. J Musculoskelet Res 2004; 8: 57–73. [Google Scholar]
  • 9.Wu G, Van Der Helm FCT, Veeger HEJ, et al. ISB recommendation on definitions of joint coordinate systems of various joints for the reporting of human joint motion – part II: shoulder, elbow, wrist and hand. J Biomech 2005; 38: 981–992. [DOI] [PubMed] [Google Scholar]
  • 10.Henninger HB, Barg A, Anderson AE, et al. Effect of deltoid tension and humeral version in reverse total shoulder arthroplasty: a biomechanical study. J Shoulder Elbow Surg 2012; 21: 483–490. [DOI] [PubMed] [Google Scholar]
  • 11.Terrier A, Reist A, Merlini F, et al. Simulated joint and muscle forces in reversed and anatomic shoulder prostheses. J Bone Joint Surg Br 2008; 90: 751–756. [DOI] [PubMed] [Google Scholar]
  • 12.Winter DA. Biomechanics and motor control of human movement. Vol. 2, 1990. Hoboken, New Jersey, USA: John Wiley & sons.
  • 13.Reeves JM. An in-silico assessment of stemless shoulder arthroplasty: from CT to predicted bone response, https://ir.lib.uwo.ca/etd/5398 (2018, accessed 25 September 2018).
  • 14.Levy JC, Anderson C, Samson A. Classification of postoperative acromial fractures following reverse shoulder arthroplasty. J Bone Joint Surg Am 2013; 95: e104–e104. [DOI] [PubMed] [Google Scholar]
  • 15.Ackland DC, Roshan-Zamir S, Richardson M, et al. Muscle and joint-contact loading at the glenohumeral joint after reverse total shoulder arthroplasty. J Orthop Res 2011; 29: 1850–1858. [DOI] [PubMed] [Google Scholar]
  • 16.Kaufman KR, Au KN, Litchy WJ, et al. Physiological prediction of muscle forces – II. Application to isokinetic exercise. Neuroscience 1991; 40: 793–804. [DOI] [PubMed] [Google Scholar]
  • 17.Scepi M, Faure J-P, Ridoux N, et al. A three-dimensional model of the shoulder girdle. Forces developed in deltoid and supraspinatus muscles during abduction. Surg Radiol Anat 2004; 26: 290–296. [DOI] [PubMed] [Google Scholar]
  • 18.Anderson FC, Pandy MG. Static and dynamic optimization solutions for gait are practically equivalent. J Biomech 2001; 34: 153–161. [DOI] [PubMed] [Google Scholar]

Articles from Shoulder & Elbow are provided here courtesy of SAGE Publications

RESOURCES