Design and validation of a biomechanics device for preclinical arthrofibrosis models Authors: Mason F Carstens, Nils Meißner, Matthew T Weintraub, Oksana Pichurin, Mark E Morrey, Joaquin Sanchez-Sotelo, Daniel J Berry, Roman Thaler, Matthew P Abdel Published: 2026 ABSTRACT AbstractAimsReliable assessment of joint stiffness is essential for studying arthrofibrosis in preclinical animal models. This study presents a new dedicated biomechanics device for measuring knee stiffness in mouse, rat, and rabbit models of arthrofibrosis, with the aim of validating its performance and comparing it to a previously validated device.MethodsThe new system integrates a torque load cell, stepper motor, and absolute encoder in a through-hole, unified configuration for precise continuous torque-angle measurement, with custom limb brackets ensuring species-specific alignment at the joint’s centre of rotation. To validate this device and compare it to a previously validated device, arthrofibrosis was induced using established models of extra-articular immobilization in mice and intra-articular violation with immobilization in rats and rabbits, followed by defined remobilization periods. Torque-angle curves were recorded ex vivo across species-specific torque ranges, and measurement variability between the new and a previous validated device was statistically compared using Brown-Forsythe’s test.ResultsThe onset of hyperextension in contralateral limbs occurred at mean angles of 152.4° (mice), 154.4° (rats), and 152.8° (rabbits), and at maximum torque, mean flexion angles were 79.2°, 87.3°, and 140.5°, respectively. No significant differences in measurement variability were observed between devices across all species.ConclusionThese findings demonstrate that our new device provides reproducible torque-angle data while maintaining the statistical robustness of the previous device and offering improved mechanical design and operational standardization. By enabling standardized, cross-species joint stiffness quantification, this platform improves cross-study comparability and strengthens the translational reliability of preclinical arthrofibrosis research.Cite this article: Bone Joint Res 2026;15(6):716–725.Keywords: Biomechanics, Device, Arthrofibrosis, arthrofibrosis, biomechanics, rabbit models, immobilization, rats, joint stiffness, Animal models, knee stiffness, flexion angles, knee Article focus To present and validate a new biomechanics device with a through-hole design for precise continuous torque-angle measurement in preclinical joint stiffness quantification. To present and validate a new biomechanics device with a through-hole design for precise continuous torque-angle measurement in preclinical joint stiffness quantification. To compare the measurement variability of the new device to a previously validated device across three distinct animal models of arthrofibrosis. To compare the measurement variability of the new device to a previously validated device across three distinct animal models of arthrofibrosis. To develop a reproducible and standardized platform that overcomes technical limitations to facilitate cross-study comparability in translational arthrofibrosis and joint stiffness research. To develop a reproducible and standardized platform that overcomes technical limitations to facilitate cross-study comparability in translational arthrofibrosis and joint stiffness research. Key messages The new biomechanics device successfully generates reproducible torque-angle data across mouse, rat, and rabbit models, with no significant difference in measurement variability compared to its predecessor. The new biomechanics device successfully generates reproducible torque-angle data across mouse, rat, and rabbit models, with no significant difference in measurement variability compared to its predecessor. The integrated design resolves major technical limitations of previous C-arm systems, such as encoder-motor slippage and difficult axial alignment for smaller joints. The integrated design resolves major technical limitations of previous C-arm systems, such as encoder-motor slippage and difficult axial alignment for smaller joints. This validated platform provides a foundation for standardized joint stiffness measurements in preclinical arthrofibrosis studies, addressing a critical barrier to cross-study comparison and therapeutic evaluation. This validated platform provides a foundation for standardized joint stiffness measurements in preclinical arthrofibrosis studies, addressing a critical barrier to cross-study comparison and therapeutic evaluation. Strengths and limitations The study rigorously validated a novel, integrated device across three distinct preclinical species using the Brown-Forsythe test, confirming high precision and statistical robustness. The study rigorously validated a novel, integrated device across three distinct preclinical species using the Brown-Forsythe test, confirming high precision and statistical robustness. The modular limb brackets and centre-of-rotation fixation system enable standardized joint stiffness quantification and are critical features for reducing technical confounders and improving cross-study comparability in translational arthrofibrosis research. The modular limb brackets and centre-of-rotation fixation system enable standardized joint stiffness quantification and are critical features for reducing technical confounders and improving cross-study comparability in translational arthrofibrosis research. The measurements were restricted to ex vivo passive flexion-extension of harvested limbs, which does not capture dynamic in vivo joint behaviour, the influence of surrounding musculature, or other planes of motion. The measurements were restricted to ex vivo passive flexion-extension of harvested limbs, which does not capture dynamic in vivo joint behaviour, the influence of surrounding musculature, or other planes of motion. Introduction Animal models are important for investigating the pathophysiology and treatment of musculoskeletal diseases. Small animal models, such as mice, rats, and rabbits, are widely employed in orthopaedic research due to their anatomical relevance, scalability, reproducibility, and cost-effectiveness.1,2 These models have facilitated the preclinical evaluation of surgical techniques, biomaterials, pharmacological therapies, and pathological mechanisms associated with joint injury and degeneration.3-5 A critical component of studying joint pathology, including conditions such as arthrofibrosis, is the ability to perform reliable biomechanical assessments. Arthrofibrosis, a debilitating complication following total knee arthroplasty and other joint injuries, is characterized by abnormal fibrous tissue deposition leading to restricted joint motion.6-12 Assessing joint stiffness is essential for quantifying disease severity and evaluating responses to therapeutic interventions. However, the small size of rodent and lagomorph joints presents unique challenges to mechanical testing.13,14 Existing methodologies often lack sensitivity, reproducibility, or species-specific adaptability, limiting their translational value.15,16 Importantly, the absence of a standardized, cross-species platform for quantifying joint stiffness remains a major barrier to reproducibility and meaningful comparison across preclinical arthrofibrosis studies. To address these limitations, we developed a new biomechanics device designed to provide precise, reproducible torque-angle measurements across mouse, rat, and rabbit models. By integrating mechanical improvements with species-specific alignment capabilities, the system was engineered to support standardized joint stiffness quantification. In this study, we compared the variability of measurements obtained with this device to those from a previously validated system to confirm preservation of statistical robustness while improving operational standardization. Methods New device configuration Our new biomechanics device (designated the MC-OrthoMech System), operates using three primary hardware components: a torque load cell, a stepper motor, and an absolute encoder (Figure 1). Compared to the previous version, this device incorporates a through-hole design that allows for precise axial alignment and coordinated function of these components. Configuration of new biomechanics device. Box labels indicate hardware components and arrows indicate hardware connectivity. The load cell measures the torque, and its amplifier enhances the electrical signal to be integrated in the USB Data Acquisition Box. The stepper motor is geared with the absolute encoder (angle measurement) and likewise integrated at the USB Hub. All integrated data is depicted on the PC user interface with real-time generation of the stiffness curve. Individual torque-angle data can be exported to an Excel (Microsoft, USA) document. The torque load cell (Transducer Techniques, USA) is surface-mounted and serves as the mechanical ground reference for all device motion. The stepper motor (Anaheim Automation, USA) drives a central shaft that passes through the centre of the load cell without direct contact — an intentional, non-intrusive configuration that ensures torque transmission is isolated to the mounted limb bracket rather than influenced by motor-shaft interaction. Positioned between the motor housing and the load cell, the absolute encoder (Anaheim Automation) engages the motor-driven shaft via geared contact and records angular displacement as the experimental joint is passively extended at a rate of 1°/s (Figure 2). At present, the device is intended for internal research use to ensure controlled implementation and methodological consistency across ongoing studies. Diagram of a new biomechanics device. The load cell measures the torque transmission of the experimental limb free of the confounding motor because of the through-hole design. The absolute encoder is geared with the rotating shaft of the stepper motor and ensures the two components are always inline. Ethical treatment of animals This study was approved by the Mayo Clinic’s Institutional Animal Care and Use Committee. We have adhered to the ARRIVE guidelines and have supplied the ARRIVE checklist in the Supplementary Material.17 All animals were assessed for weight before surgery; 16 mice (female, C57BL/6 J, The Jackson Laboratory, USA) weighed approximately 20 g, 16 rats (female, Sprague-Dawley, Inotiv, USA) 250 g, and 20 rabbits (female, New Zealand White, Charles Rivers Laboratories, USA) 2.5 kg. Accordingly, all animals were premedicated prior to surgery with a subcutaneous injection of buprenorphine (Fidelis Animal Health, USA) with an extended release at a dose of 1 mg/kg for analgesia and 30 mg/kg cefazolin (Hikma Pharmaceuticals, USA) for infection prophylaxis. All surgical procedures were done in a sterile field, and animals were anaesthetized with continuous flow isoflurane (Piramal Critical Care, USA). Following the surgery, all animals were returned to their native cages and allowed free range and food. Humane endpoints included observed loss of knee fixation (immobilization) and/or inability to thrive. Animals were observed frequently in the immediate postoperative period and throughout the experimental timeline by both research members (MFC, OP) and animal specialists. There were no surgical or postoperative complications, and no animals were removed from this study. Mouse model A previously published murine model of extra-articular knee immobilization was used to induce post-surgical joint stiffness for biomechanical testing.18 In the present study, this model was replicated using an eight week immobilization period followed by four weeks of remobilization. Briefly, a 21-gauge needle (Becton Dickinson, USA) was inserted lateral to the femur and advanced distally until exiting near the tibial tubercle. A 3-0 nonabsorbable suture (Ethicon, USA) was passed through the needle track. The procedure was then repeated medially, and the suture was looped around the femur and tibia and secured with a sliding knot in flexion to maintain immobilization. After eight weeks, the suture was removed, and mice were permitted unrestricted cage activity for four weeks prior to euthanasia and further biomechanical assessment. Rat model To model arthrofibrosis in rats, we employed a surgical technique that combines extra-articular immobilization with intra-articular joint violation, as described previously.19 The experimental timeline consisted of eight weeks of immobilization followed by four weeks of remobilization. A midline skin incision was made over the knee joint. A parapatellar approach was used to medially displace the patella, exposing the articular surfaces. Both cruciate ligaments were excised, and the posterior capsule was disrupted using a controlled hyperextension force. A singular 1 mm hole was drilled into each medial and lateral femoral condyle to simulate intra-articular injury. For immobilization, a 2-0 stainless steel wire (Ethicon) was looped around the femur and tibia while the knee was held in a flexed position and secured at the mid-shaft of the femur. After eight weeks, the wire was surgically removed, and animals were allowed unrestricted cage activity for four weeks prior to limb harvest. Rabbit model A validated rabbit model of post-traumatic knee arthrofibrosis was implemented, employing both intra-articular violation and external immobilization. The timeline included eight weeks of immobilization followed by 16 weeks of remobilization to replicate chronic joint stiffness development.20-29 Rabbits underwent surgery via a lateral parapatellar approach. The patella was subluxated medially, and the suprapatellar fat pad was excised to expose the cruciate ligaments, which were then transected. The posterior capsule was disrupted via hyperextension. Cortical defects were created in the medial and lateral femoral condyles using a 3 mm drill bit. A 2 mm transverse tunnel was drilled through the mid-shaft of the tibia, through which a half-threaded Kirschner wire (K-wire) (Stryker, USA) was passed. The distal end of the K-wire was bent to hook around the femur and secured with a stainless-steel nut at the tibial cortex to maintain the joint in flexion. After eight weeks, the K-wire was removed, and rabbits were returned to unrestricted activity for 16 weeks before limbs were collected for biomechanical testing. Biomechanical knee measurements Each animal model used a custom-designed bracket to secure the limb onto the biomechanics device. For mice and rats, the harvested limbs were prepared by embedding the femora in hollowed plastic cylinders using bone cement (Simplex HV, Stryker), while the rotating arm of the device engaged the tibia to measure torque across a passive range of motion. For rabbits, both the femoral and tibial canals of the harvested limbs were hollowed and reinforced with steel rods fixed in place with bone cement (Stryker), allowing stable fixation of the femur and precise engagement of the tibia by the rotating arm. These customized brackets ensured accurate positioning of the limb, placing the knee joint precisely at the device’s centre of rotation (Figure 3a). Rabbit knee positioning on new biomechanics device. a) Alignment of rabbit knee is in line with centre of rotation axis, with the centre of the joint aligned with the centre of the motor shaft. b) Custom limb bracket positioning rabbit limb at starting angle of 45° and rotated through a pre-determined total range of motion set by the user in the PC interface. Each limb was rotated through a total range of motion of 125° (from 45° to 170°) (Figure 3b). The device continuously recorded both torque and angle throughout the entire movement arc. The upper limit of 170° exceeded the physiological range for all species studied. To minimize confounding variables in torque analysis, the maximum angle used for data interpretation was limited based on measurements from contralateral (non-surgical) limbs. The torque ranges for data analysis (0.2 to 0.6 N·cm for mice, 1 to 10 N·cm for rats, and 25 to 150 N·cm for rabbits) were determined based on the maximum torque value achieved by all stiffened experimental limbs within each respective species. Unlike previous studies that reported joint stiffness at limited discrete torque values or single-point endpoints, our approach used these maximums to establish equal-interval torque increments across the measurement arc.18,19,28 This selection strategy was implemented to ensure maximum transparency in the propagation of torque-angle data, capturing the high-resolution biomechanical signature of the joint across its range of motion. Statistical analysis Angle measurements obtained with our new biomechanics device are presented as the mean and SD at each corresponding torque value. The sample sizes for this study were determined by the foundational studies that used biomechanical measurements from the previous device.18,19,24 The statistical criteria for all power analyses were 80% power and an α (Type I error rate) of 0.05. To assess significant differences in data variability during device comparisons, the Brown-Forsythe’s test was used. This test calculates the absolute deviations from the median at each measurement point and determines whether there is a significant difference in variability between two experimental groups. Complementary descriptive statistics, including the coefficient of variation, SD, standard error of the mean (SEM), and 95% CIs, were additionally calculated. Given that comparisons were performed at each torque within the torque range, p-values are reported as a range representing the minimum and maximum values observed across all intervals to demonstrate consistent statistical parity between devices (GraphPad Prism v.10.4.2, GraphPad Software, USA). Results Defining knee range of motion Contralateral limbs were used to determine the torque associated with knee hyperextension within the measured range of motion (45° to 170°). There were ten contralateral limbs each for mice, rats, and rabbits. Torque-angle curves were generated, and the onset of hyperextension was defined as the angle at which torque appeared to increase exponentially. In mice, the mean angle at the initiation of hyperextension was 152.4° (SD 4.7°), with a corresponding coefficient of variation of 3.1%. In rats, the mean angle was 154.4° (SD 8.2°) with a coefficient of variation of 5.3%, and in rabbits, the mean angle was 152.8° (SD 3.8°) with a coefficient of variation of 2.5%. A summary of this data is provided in Table I. Table I.Mean hyperextension angles to define knee range of motion.SpeciesTimeline (weeks)Hyperextension angle, °SD95% CICV, %SEMMouse8 Imm/4 Rem152.44.7149.1 to 155.83.11.5Rat8 Imm/4 Rem154.48.2148.5 to 160.25.32.6Rabbit8 Imm/16 Rem152.83.8150.1 to 155.52.51.2CV, coefficient of variation; Imm, immobilization; Rem, remobilization; SEM, standard error of the mean. Mean hyperextension angles to define knee range of motion. CV, coefficient of variation; Imm, immobilization; Rem, remobilization; SEM, standard error of the mean. New biomechanics device measurements in stiff joints There were ten stiff limbs each for mice, rats, and rabbits. Mean passive extension angles were calculated for all recorded torque values. At the upper limit of the established torque ranges, the following mean passive extension angles were observed: in mice, a torque of 0.6 N·cm produced a mean angle of 79.2° (SD 18.9°) (Figure 4a). For rats, the mean angle was 87.3° (SD 20.6°) at 10 N·cm (Figure 4b). For rabbits, a torque of 150 N·cm yielded a mean angle of 140.5° (SD 17.3°) (Figure 4c). Species-specific torque and angle values for all measurements can be found in Supplementary Tables 1 to 3. Passive extension angles (PEAs) measured by new biomechanics device. a) PEAs for mice (n = 10), b) rats (n = 10), and c) rabbits (n = 10). Shaded regions (grey) indicate SD. These measured values represent a significant reduction in knee mobility compared to the near-full extension (approximately 152° to 154°) observed in healthy contralateral limbs. Biologically, these restricted angles at maximum torque quantify the severity of the joint violation and immobilization induced in our surgical models. Clinically, this reduction in the passive range of motion reflects the hallmark of human arthrofibrosis, where pathological tissue deposition limits functional joint excursion. Comparison of variability between new and previous devices Brown-Forsythe’s test was used to assess whether the measurement variability between the previous and the new version of the biomechanics device differed significantly. This method utilizes absolute deviations from the median at each point measured, making it appropriate for our datasets, which did not meet the assumptions of normality. For mice, six stiff knees for the previous device and ten stiff knees for the new device were compared. There were no significant differences in variability observed between devices for mice after the remobilization period (p = 0.7 to 0.1) (Figure 5). For the rats, six stiff knees for the previous device and ten stiff knees for the new device were compared; again, there were no significant differences in data variability between the devices (p = 0.1 to 0.9) (Figure 6). For the rabbits, ten stiff knees from the previous device and ten stiff knees from the new device were compared. Likewise, there were no significant differences between devices (p = 0.3 to 1) (Figure 7). Comprehensive tables separated by species for all torque and angle values can be found in Supplementary Tables 4 to 6. Absolute deviations (Abs. Dev.) from median (Brown-Forsythe, box and whisker) for comparison of previous (blue) and new (red) devices of mouse passive extension angles. Torque range 0.2 to 0.6 N·cm. Individual values indicate individual mice (n = 6 for previous device, n = 10 for new device). ns, not significant. Absolute deviations (Abs. Dev.) from median (Brown-Forsythe, box and whisker) for comparison of previous (blue) and new (red) devices of rat passive extension angles. Torque range 1 to 10 N·cm. Individual values indicate individual rats (n = 6 for previous device, n = 10 for new device). ns, not significant. Absolute deviations (Abs. Dev.) from median (Brown-Forsythe, box and whisker) for comparison of previous (blue) and new (red) devices of rabbit passive extension angles (PEAs). Torque range 25 to 150 N·cm. Individual values indicate individual rabbits (n = 10 for previous device, n = 10 for new device). ns, not significant. Discussion This study demonstrated that our newly developed biomechanics device generates reliable joint stiffness measurements across three orthopaedic surgical animal models: mice, rats, and rabbits, with variability comparable to the previous system. These findings support the utility of the new device as a reproducible platform for preclinical studies of knee arthrofibrosis. Our new device is designed to improve a set of key aspects during data acquisition. The previous biomechanics device employed a C-arm design with torque and angular displacement measured by independent components. The load cell in both iterations was mounted to the base with the limb bracket fixed on top, which effectively isolated torque measurements from vibration artefacts generated by the motor mounted above. However, angular displacement was measured by a potentiometer connected to the motor through a rubber belt on aluminium gears. The inherent elasticity of the belt made it susceptible to slippage and wear, which risked misalignment between motor rotation and potentiometer readings and oftentimes required replacement. Because the torque and angle data streams were not integrated downstream of collection, accurate synchronization depended on aligning as many overlapping data points as possible. In the new device, these limitations were addressed by integrating the torque load cell, stepper motor, and absolute encoder into a through-hole configuration. The motor shaft passes directly through the load cell, and the limb brackets are mounted onto both, ensuring that torque is transmitted solely through the bracketed limb and not confounded by motor vibration. The absolute encoder, geared directly to the motor shaft, provides enhanced angular resolution compared with the earlier potentiometer and cannot drift out of register with shaft rotation. Importantly, the system allows predetermined torque or range of motion limits to be specified before testing, so all measurements are collected within the same defined interval. This is made possible by an intuitive user interface that automates calibration to a 45° starting angle, permits user-defined start and end positions, and displays torque-angle curves as they are generated. Since the encoder, motor shaft, and load cell are integrated through a downstream central data acquisition hub, torque and angle outputs are aligned in real time at the point of acquisition. Another limitation of the previous system was its origin as a downsized rabbit-knee device, which made axial alignment difficult when applied to smaller species such as rats and mice. The C-arm geometry restricted the placement of the tibial-engaging arm, particularly in mice, where consistent alignment of the joint’s centre of rotation with the device axis was challenging. The through-hole design of the new system resolves this problem by permitting the limb to be mounted and unobstructed along the axis of rotation, thereby ensuring precise alignment across mice, rats, and rabbits. Despite this improved mechanical design and robust integrated hardware updates, one of the primary objectives was to at least preserve the measurement integrity established by its predecessor. Consistent with that goal, our results show no significant differences in variability between devices across all tested species, as confirmed by Brown-Forsythe’s tests. These results support the device’s suitability as an replacement for the earlier device without introducing significantly increased experimental noise with the enhanced hardware. Consequently, the new system represents a major advancement; despite equivalent measurement variability, it is preferable for its superior mechanical reliability, ease of axial alignment, and the ability to generate continuous, synchronized torque-angle profiles that were previously difficult to achieve. The torque-angle profiles generated with the new device reflect consistent biomechanical behaviour and replicate prior findings. For example, in the murine model, the onset of hyperextension was detected at approximately 152°, which closely aligns with previously published biomechanical data using extra-articular fixation models to induce knee contracture.18 Similarly, the rat and rabbit models displayed expected stiffness patterns consistent with earlier reports involving intra-articular violation and capsular disruption.19,24 A critical barrier to cross-study comparison and repeatability has been the lack of standardized platforms for joint stiffness quantification, especially in preclinical surgical models. This device represents an advancement toward standardizing biomechanical protocols across different species. The modular limb brackets, species-specific torque calibration, and centre-of-rotation fixation system ensure that joint articulation is measured under controlled, repeatable conditions. These design features are critical for reducing technical limitations associated with limb positioning, which has been a notable confounder in previous preclinical work.30 A limited number of groups have published on their biomechanical devices for small animal joints, mostly relying on manual goniometry or torque application using external weights.31,32 These methods often result in poor reproducibility, inconsistent data collection parameters, and may overlook subtle but biologically meaningful differences in joint mechanics. In contrast, our new device’s motorized and encoder-driven system prevents operator bias and enables continuous torque-angle acquisition throughout a predetermined range of motion. In addition to harvested-limb systems, Reina et al33 previously validated a live passive extension device used to assess knee range of motion in sedated, muscle-relaxed rabbits prior to euthanasia. That platform was designed for longitudinal in vivo assessments and captured extension angle endpoints rather than continuous torque-angle data. While this live device is particularly valuable for tracking disease progression or treatment response in the same animal over time, it does not provide detailed biomechanical signatures or stiffness curves. Our new device addresses this limitation by enabling high-resolution, ex vivo measurements that reflect the full torque-angle profile of the joint. Together, these devices form a complementary suite of tools for multimodal characterization of joint mechanics in preclinical arthrofibrosis models. While our device demonstrated consistent performance, this study was limited to postmortem, passive measurements of harvested limbs. As such, it does not capture dynamic joint behaviour or the influence of surrounding musculature in vivo. Future iterations could incorporate methods to simulate active loading or approximate physiological weightbearing conditions, which would provide a more comprehensive understanding of joint mechanics. Additionally, while our findings confirm high precision and internal reproducibility, it is important to note that interoperator and inter-laboratory variability were not assessed in this study. Future studies with these species will be necessary to further contextualize the device’s performance and confirm its utility as a standardized platform. Furthermore, although the Brown-Forsythe test provided a robust assessment of variability, larger sample sizes may allow more detailed subgroup analyses, including factors such as sex, age, or surgical technique. Finally, long-term calibration monitoring and periodic validation studies will be essential to ensure accuracy and consistency as the device becomes a central platform for future preclinical studies. Standardization of biomechanical protocols remains a significant hurdle in preclinical research. By providing unified hardware architecture and automated data acquisition across multiple species, this system serves as a foundational platform that could potentially facilitate multicentre or cross-study comparisons. Such consistency is essential for future translational research, as it allows for the objective aggregation of data across different experimental cohorts, ultimately strengthening the reliability of therapeutic evaluations and the reproducibility of arthrofibrosis models. In summary, our new biomechanics device, the MC-OrthoMech System, offers precise, reproducible joint stiffness measurements in mouse, rat, and rabbit models of knee arthrofibrosis. It preserves the statistical robustness of its predecessor while delivering improved mechanical design and operational standardization. As a validated platform, this device will serve as the foundation for preclinical joint stiffness quantification and therapeutic efficacy. Author contributions M. F. Carstens: Conceptualization, Data curation, Formal analysis, Investigation, Writing – original draft, Writing – review & editing N. Meißner: Conceptualization, Data curation, Investigation, Writing – review & editing M. T. Weintraub: Conceptualization, Data curation, Investigation, Writing – review & editing O. Pichurin: Data curation, Investigation, Writing – review & editing M. E. Morrey: Formal analysis, Writing – review & editing J. Sanchez-Sotelo: Formal analysis, Writing – review & editing D. J. Berry: Formal analysis, Writing – review & editing R. Thaler: Formal analysis, Writing – review & editing M. P. Abdel: Conceptualization, Formal analysis, Funding acquisition, Writing – review & editing Funding statement The authors disclose receipt of the following financial or material support for the research, authorship, and/or publication of this article: this study was in part funded by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) of the National Institute of Health (NIH) under Award Number R01AR083873 (to MPA). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Additional internal funding was provided by the Mayo Clinic Research Space and Equipment Executive Subcommittee (to M. F. Carstens). ICMJE COI statement M. F. Carstens (no disclosures), N. Meissner (German Innovation Fund, Aesculap AG), M. T. Weintraub (no disclosures), O. Pichurin (no disclosures), M. E. Morrey (Bonebridge, Zimmer Biomet, Global Elbow Network), J. Sanchez-Sotelo (Parvizi, Precision OS, ACUMED, Journal of Shoulder and Elbow Surgery, American Shoulder and Elbow Surgeons, Stryker, Exactech, Orthobullets, Journal of Bone and Joint Surgery, Elsevier), D. J. Berry (DePuy, Elsevier, Wolters Kluwer, Johnson & Johnson, Orthopaedic Research and Education Foundation), Roman Thaler (no disclosures), M. P. Abdel (International Orthopedic Education Network, American Association of Hip and Knee Surgeons, Hip Society, Stryker). Data sharing The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request. Acknowledgements The authors would like to acknowledge the Andrew A. and Mary S. Sugg Professorship in Orthopedic Surgery for its philanthropic support that made such research possible (to MPA). IACUC Protocols: 4221-19-R25 (Mouse), 7731-24 (Rat), and 6201-21-R24 (Rabbit). Open access funding The authors report that they received open access funding for their manuscript from the National Institutes of Health (MPA, R01AR083873). Supplementary material The supplementary material includes comprehensive biomechanical measurements, including additional measures of variability. © 2026 Carstens et al. This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives (CC BY-NC-ND 4.0) licence, which permits the copying and redistribution of the work only, and provided the original author and source are credited. See https://creativecommons.org/licenses/by-nc-nd/4.0/ Data Availability The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request. References Associated Data Data Availability Statement The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request.