In an elbow subluxation and dislocation investigation (O’Driscoll et al., 2000; 1992), O’Driscoll et al., designed an experiment to test the validity of the hypothesis that the elbow can be dislocated posteriorly with a functionally intact anterior medial collateral ligament (AMCL) and also determine the mechanism or kinematics of such dislocation and its clinical relevance. Thirteen upper extremities (seven right, six left) were used in this study. The humerus was dis-articulated from the shoulder joint as well as the radius and the ulna was dis-articulated from the wrist. All non-ligament soft tissues were removed, keeping the tissue around the elbow joint and the tendon insertion of the biceps, brachialis and triceps intact. The humerus was securely fixed in non-magnetic plastic frame and positioned so that the forearm moved in a horizontal plane with flexion and extension of the elbow for optimal testing of valgus instability. Loads of 1 kg for the biceps and brachialis and 2 kg for the triceps were used to simulate muscle tone. A 3 Space Isotrak electromagnetic tracking system was fixed to the plastic frame to which the humerus is firmly secured. The sensor was attached to the proximal of the ulna. The origin and insertion of the AMCL was digitized. The distance between the humerus and the ulna was calculated in real-time. External rotation and valgus moment with axial forces resulted in posterior elbow dislocation in twelve of the thirteen specimens with anterior medial collateral ligament intact (AMCL). O’Driscoll et al., assumed that the mechanism of dislocation during a fall on the outstretched hand would involve the body rotating internally on the elbow (O’Driscoll et al., 2000; 1992; O’Driscoll, 1999), which experiences an external rotation and valgus moment as it flexes. Experimental results also showed that dislocation occurs at 80 degrees of flexion with poster lateral rotation of 34–50° and 5–23 degrees of valgus moment. He also suggested that posterior dislocation can be reduced in supination. O’Driscoll et al., defined dislocation as the final of 3 instability stages resulting from posterolateral rotation, with a disruption of soft tissues progressing from lateral side to medial side.

In another experimental and 2D Finite Element (FE) investigation shown in (Wake et al., 2004), Wake studied the bio-mechanical analysis of the mechanism of elbow dislocation by a compressive force.

Fifty-three intact elbows where chosen for this study, where the humerus was cut transversely 90 mm proximally to the distal joint surface. The radius and the ulna were transected evenly at a point either 60 mm (short ulna model) or 90 mm extended ulna model) distally to the Coronoid tip. In this study, the humerus, radius, and ulna were fixed in dental resin to a depth of 30 mm in the specimen holders of the loading apparatus. Axial compressive loads were applied at 10 mm/min while the elbow joint being flexed at 15° of extension and 0°, 30°, 60° or 90° of flexion in custom made apparatus.

The loading experiments produced various dislocations and fractures of the humeral shaft (13%), supracondyle (30%) and radial or ulnar shaft (28%). Anterior fracture-dislocation [type II: with Olecranon fracture at 60° flexion position or 90° flexion position occurred when the load was applied at 60° and 90° of flexion. In the 60° flexion position and 60 mm forearm length, type II or combined types I and II occurred.

Posterior fracture-dislocation [type I: with Coronoid fracture, at 15° extension, at 0° flexion position, and 30° flexion position occurred from 15° of extension to 30° of flexion. All cases had a concurrent coronoid process and radial head or neck fractures. At 90° of flexion, the humerus catches on the Olecranon to develop a fracture of the Olecranon, which is presumed to have caused the failure of the posterior supporting systems.

Hyperextension at the elbow joint is another known mechanism of elbow dislocation. Tydral designed in (Tyrdal and Olsen, 1998) an experiment to produce a combined hyperextension and supination at the elbow joint and observed the lateral ligament lesions induced. In this study, ten elbow cadavers from five male donors with a mean age of 28.8 years were used. Relatively young donors were chosen since age-related changes are commonly expected in human tissue and specimens near the same age range as active athletes were wanted. All skin and fatty tissues were dissected away leaving the ligaments around the elbow joint and the forearm muscles intact. A three-dimensional loading apparatus was developed to study the kinematics of elbow dislocation. The humerus was mounted horizontally in the loading apparatus, and fixed with one screw and four hose clamps. The rotation was blocked by one lateral screw.

The forearm was connected to the mobile lever arm by two screws through the proximal ulna. The nylon line was fixed to an eyelet screw going through the distal radius from the volar side. Hyperextension force was applied in the form of bags filled with an increasing amount of water. The bags were also allowed to fall from 1.5 to 2 m. The loads were applied at the elbow being in maximal extension and in full supination to imitate the kind of trauma cased by a handball. The impact of the falling bags corresponded to the effects of a handball (450 g) hitting the distal forearm at different speeds (Tyrdal and Bahr, 1996). The experimental loads applied at the last trauma corresponded to the speed of a handball between 65 and 200 km/h. The hyperextension loads resulted in three different injuries to the ligaments: (1) anterior capsule rupture, (2) avulsion of the proximal insertion of the medial and (3) the lateral collateral ligaments. The lesion was only visible from the inside of the joint, but in some cases the lateral lesion could also be seen from the outside. Cartilage lesions of the Olecranon or the humerus were not observed.

Another possible mechanism of elbow dislocation is recurrent posterolateral rotatory instability (Tyrdal and Bahr, 1996). The combination of elbow dislocation with fractures of the radial head and the coronoid process has been termed the terrible triad by Hotchkiss in (Hotchkiss, 1996). Schneeberger designed in (Schneeberger et al., 2004) an experimental setup to evaluate the role of the radial head and coronoid process as posterolateral rotatory stabilizers of the elbow joint. Ten fresh-frozen upper extremities cadavers with no evidence of pathological changes at the elbow were used for this study; two were used for a pilot evaluation, and eight were used for measurements. The limbs were amputated through the proximal third of the humerus and dis-articulated at the wrist. All soft tissue around the elbow joint was left intact during the study.

The upper arm was fixed to a specifically designed frame using a large AO external fixator with two bicortical 4.5-mm Steinmann pins placed through the humeral shaft. A standardized surgical approach was designed to gain access to the coronoid process and radial head for this experiment. The approach consisted of two osteotomies-one of the lateral epicondyle and one of the ulnar insertion of the lateral ulnar collateral ligament-performed with an oscillating saw. The posterolateral rotatory displacement of the ulna was measured after application of a valgus and supinating torque (1) in seven intact elbows, (2) after radial head excision, (3) after sequential resection of the coronoid process, (4) after subsequent insertion of each of two different types of metal radial head prostheses (a rigid implant and a bipolar implant with a floating cup, and (5) after subsequent reconstruction of the coronoid with each of two different techniques in the same Cadaver elbow.

This vivo study showed that the posterolateral rotatory laxity averaged 5.4° in the intact elbow at 60° of flexion. Excision of the radial head in an elbow with intact collateral ligament caused a mean posterolateral rotatory laxity of 18.6° (p < 0.0001). The elbows with a defect of 50% or 70% of the coronoid, loss of the radial head, and intact ligaments could not be stabilized by radial head replacement alone, but additional coronoid reconstruction restored stability. Resection of 30%, 50%, and 70% of the coronoid process resulted in detachment of the annular ligament at the base of the coronoid of approximately 50%, 70%, and 90%, respectively. In clinical replacement, replacing the radial head with a rigid implant seems to restore stability better than replacing a floating prosthesis.

The three-column concept was proposed in (Watts et al., 2019) for elbow fracture with dislocations, into improving the understanding of injury patterns. The elements of the elbow joint were decomposed into three columns: medial, middle and lateral. This allowed for the proposal of a classification system for elbow fracture dislocations which can help in treating elbow injuries efficiently.

Magnetic resonance imaging and radiography datasets of 64 patients were used in (Schnetzke et al., 2021) to analyze the mechanism and injury patterns in elbow dislocations. Among the study’s findings, a “reversed Horii circle” was proposed, with a medial force of induction that originates and continues from medial to anterior. This term is used in the context of the “Horii circle”, the term used for the disruption of soft tissue from lateral to medial, described by O’Driscoll et al. (O’Driscoll et al., 1992; O’Driscoll, 1999)

This study experimentally investigated elbow dislocation using the arms of the Papio anubis baboon, as illustrated in Figure 3A (Alkork, 2011). Baboons were selected for their close anatomical similarity to the human elbow joint and their accessibility through established university research programs. In addition, baboon arms provide a cost-effective and logistically feasible alternative to human cadavers, which are limited in availability, highly regulated, and exhibit greater variability in donor health and tissue quality. The smaller scale and consistent anatomical features of baboon limbs make them well suited for controlled experimental setups and reproducible mechanical testing. Collectively, these factors support the use of the baboon as a valid and ethical intermediary model for studying elbow dislocation mechanisms prior to translation to human studies. Seventy female Papio anubis (2–5 years; mean weight, 10.8 ± 1.3 kg) were divided into two main groups. Group I included 62 arms used for the mechanical analysis of elbow dislocation, and Group II included 8 arms reserved for anatomical dissection. The humerus was dis-articulated from the shoulder joint in these elbows, and the radius and ulna were dis-articulated at the wrist. In Group I, all arms were thawed and dissected free of soft tissues except for the elbow capsule and ligaments that were left intact. Figure 3B shows the anterior medial collateral ligament (AMCL), posterior medial collateral ligament (PMCL), and lateral collateral ligament (LCL) in one of the used arms. The baboon cadaver’s upper extremities were harvested from non-related research studies at University of Illinois at Chicago (UIC) and approved by the UIC institutional Animal Care and Use Committee (ACC). All experiments conducted on the baboon cadaver arms were performed in accordance with the ethical guidelines and regulations set by the University of Illinois at Chicago Animal Care Committee (UIC-ACC). The use of baboon cadaver tissue was approved by the UIC-ACC under protocol number [2005-19970101]. In this study, no live animals were used. All authors complied with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines to ensure transparent and reproducible reporting of research involving animal-derived specimens.

The study involved 21 human cadaveric elbows (12 male, 9 female). Cadavers were stored at approximately −20 °C and thawed 24 h before testing. Bone mineral density (DXA) values ranged from −0.3 to −1.2 T-score. All specimens were donated for research purposes. Donor ages ranged from 45 to 82 years (mean = 63 ± 10 years), with no documented history of musculoskeletal or orthopedic disorders that could affect joint integrity. Each specimen was examined for visible deformities or degenerative changes prior to testing. When donor medical information was incomplete, this limitation was explicitly noted to maintain transparency in interpreting tissue behavior and mechanical response. Only one limb per donor was tested due to availability. When bilateral limbs were available, left/right selection was randomized. The use of human tissue in this study adhered to all relevant regulations and ethical guidelines, including the Declaration of Helsinki. Prior to use, written informed consent was obtained from the donors or their legal guardians. Studies involving human participants were reviewed and approved by the Institutional Review Board (IRB) of the University of Illinois at Chicago (Approval Number: 2005-19970102). All procedures were conducted in accordance with the ethical standards of the IRB. To ensure privacy and confidentiality, all donor identities were anonymized and de-identified before data analysis. The authors sincerely thank the donors and their families for their invaluable contributions to medical research. The cadaver arms were amputated through the proximal third of the humerus, and the radius and ulna were disarticulated at the wrist.

The target sample size for the cadaveric experiments ( n = 21 ) was determined based on both the feasibility and prior biomechanical investigations of upper-limb joints (Wake et al., 2004; O’Driscoll et al., 2000) typically employ 10–25 specimens to capture inter-individual variability. A preliminary pilot test with five cadaveric elbows (SD ≈ 580 N; expected difference ≈ 1000 N between flexion groups) indicated that a minimum of fourteen specimens would achieve 80% power ( α = 0.05 ) to detect a large effect size ( d ≈ 1.7 ) . Consequently, 21 samples were included to exceed this requirement and ensure adequate statistical strength while remaining within ethical and logistical constraints.

In all 21 cadaver arms the skin, subcutaneous tissues and muscles were excised. To preserve the integrity of the elbow joint, the lateral collateral ligament, medial collateral ligament and the radial collateral ligament were left intact along with the joint capsule. In each setup, and as in the use case of the baboon elbows, the humerus was fixed to a plate, and the ulna and radius were cemented to a metallic cup. Also, the elbows were configured with different flexion angles, and the ulna and radius were configured in pronation or supination. A jig apparatus was designed by Pro Engineer Software and manufactured for the experiments (see Figure 6). The apparatus is designed similar to the one used with the baboon arms but with modifications to allow it to hold more load and different elbow dimensions.

A total of 21 human cadaveric upper limbs were used in the experiments. Their demographic and positional characteristics are summarized in Table 7.

Each specimen was mounted and subjected to an increasing axial load at a constant rate of 10 mm/min until either (i) elbow dislocation occurred or (ii) fracture followed by dislocation was observed. The 21 arms were categorized into three experimental groups based on flexion angle and loading configuration. Table 8 summarizes the groups, their characteristics, and the average dislocation loads.

Figure 14 illustrates representative experimental setups, loading conditions, and X-ray images obtained during the cadaveric tests. The sequence shows progressive deformation and eventual failure at different flexion angles. Under pronation and increasing flexion, the ulna–humerus articulation maintained congruent contact surfaces for longer before posterior displacement occurred, requiring substantially higher axial loads to initiate dislocation. In contrast, hyperextension and neutral orientations displayed earlier separation and lower load thresholds.

The key experimental findings can be summarized as follows:

Overall, the human cadaver results align with the baboon model findings, demonstrating that pronation and lower flexion angles require higher axial loads to induce instability. This suggests that increased joint congruence and ligament tension in pronation contribute to greater resistance of the ulna–humerus articulation to posterior translation.

To provide a comparative visualization of human elbow behavior under similar loading regimes, an instability envelope was generated for the cadaveric specimens (Figure 15). This figure depicts how dislocation thresholds shift across flexion angles under axial compression, based on averaged data from Table 8. Greater flexion and forearm pronation correspond to increased joint stability, whereas lower flexion and neutral positions reduce resistance to posterior translation. The contour demarcations of Stages I–III illustrate the transition from soft-tissue yielding to combined bony–ligamentous failure, aligning with the fracture patterns observed experimentally at higher flexion angles.

The contour labels (I–III) in Figure 15 are intentionally illustrative bands, corresponding approximately to 90%, 100%, and 110% of the mean dislocation threshold curve. They serve as visual guides to indicate the progressive stages of instability defined mechanistically in this study, rather than strict quantitative boundaries.

Mixed-effects analyses across both the Papio anubis and human cadaveric models revealed a consistent mechanical pattern: elbow stability increased with flexion and was enhanced by pronation. Despite interspecies differences in absolute dislocation loads, the relative orientation effects were highly concordant, with load elevation observed in both models under flexed and pronated configurations. The baboon model demonstrated greater rotational sensitivity ( Δ ≈ 960 N between pronation and supination), whereas human elbows showed a smaller but statistically significant difference ( Δ ≈ 780 N). These parallel mechanical trends support the translational validity of the baboon model for studying elbow dislocation mechanisms and justify its use for parameterizing future computational and finite-element models of joint instability.

The combined experimental findings from the Papio anubis and human cadaveric models reveal a consistent biomechanical pattern: forearm pronation and increased elbow flexion substantially raise the axial load required for dislocation, reflecting enhanced joint congruence and tension within the collateral ligament complex. Across both species, bony failure often preceded complete soft-tissue rupture under axial compression, underscoring the primacy of osseous geometry and coronoid engagement in resisting posterior translation. These insights bridge comparative anatomy with applied biomechanics, highlighting that elbow stability depends on a finely balanced interaction between bone morphology and ligament restraint. Clinically, this integrative understanding can inform reconstructive and rehabilitation strategies by differentiating between rotational and compressive instability mechanisms—guiding surgeons toward load-specific approaches for restoring stability and preserving joint kinematics.

This study was performed under controlled, quasi-static loading without simulated muscle co-contraction; as such, stabilizing effects from dynamic neuromuscular control were not modeled and may elevate dislocation thresholds in vivo. Cadaveric and juvenile Papio anubis tissues differ from adult, living human tissue in viscoelasticity and failure tolerance, which may influence absolute load magnitudes while preserving relative angle–rotation trends. Forearm rotation was tested at end-range positions (pronation/supination), potentially exaggerating orientation effects relative to mid-range postures. At 90° flexion, fracture-dominant behavior limited pure soft-tissue dislocation analysis, and thresholds at that angle should be interpreted in the context of bony failure risk. Finally, sample sizes within some angle–rotation strata constrained precision of variance estimates; mixed-effects modeling mitigated inter-specimen variability, but replication with larger cohorts and explicit muscle actuation is warranted.

The findings from these experimental investigations on baboon and human cadaveric arms underscore the value of the baboon elbow as a biomechanical model for studying dislocation mechanisms. The similarity in dislocation patterns between the two species validates this approach, while the baboon model offers the additional advantage of lower cost and accessibility. The results also refine O’Driscoll’s earlier sequence of ligament failure, indicating that the medial collateral ligament tends to fail before the lateral collateral ligament.

Clinically, these results suggest that both ligament complexes should be evaluated and addressed during reconstruction, rather than focusing solely on the lateral side. In early, low-impact posterior dislocations, the joint often remains stable following closed reduction and early mobilization, preserving both bony and soft-tissue integrity. However, advanced dislocations are typically associated with extensive damage to both ligament complexes, requiring more comprehensive repair strategies.

Our experimental setup was designed to isolate the intrinsic mechanics of elbow stability under controlled, repeatable conditions. Group sizes were intentionally limited to allow precise statistical estimation. Muscle forces were excluded to observe the pure behavior of the bone–capsule–ligament complex, and fixtures were standardized to maintain alignment and consistent loading. While this approach clarified the sequential stages of dislocation and their load thresholds, future studies should incorporate dynamic loading with simulated muscle activation and integrate finite-element models to extend these findings toward clinical application.

The experimental findings from both the baboon and human cadaveric models provide an integrated biomechanical understanding of how ligament failure sequences govern elbow stability. Translating these mechanical insights into clinical practice is essential to guide diagnostic prioritization, surgical planning, and postoperative rehabilitation. The observation that the medial collateral complex fails before the lateral structures redefines the conventional approach that often prioritizes lateral repair alone. Clinically, this underscores the need for early recognition and targeted management of medial instability to prevent recurrent subluxation or chronic valgus laxity. To facilitate this translational step, a decision-making algorithm was developed (Figure 24) to bridge biomechanical patterns with surgical and rehabilitative protocols. The algorithm delineates a clear pathway from mechanism identification and imaging assessment to treatment selection and functional recovery, emphasizing the role of flexion–pronation positioning in preserving joint congruence and minimizing re-dislocation risk.

The study concludes with the creation of an illustrative animated video (click here to view video) that provides a visual representation of joint dislocation under axial and hyperextension loads. This video serves as an educational tool, illustrating the progression of dislocation mechanisms and enhancing understanding for students, clinicians, and researchers interested in elbow biomechanics, ultimately contributing to the field of orthopedics.

Future investigations should integrate dynamic simulation and finite-element analyses with instrumented joint rigs incorporating actuated musculature to better approximate physiologic motion. Expanding the comparative dataset to include additional primate species and a wider range of flexion–rotation combinations could further clarify evolutionary adaptations in elbow stability. Linking these biomechanical findings with patient-specific imaging and postoperative outcome data will enhance translational relevance and support the design of clinically validated reconstruction strategies.