Cartilage mechanical responses during gait as in silico biomarkers for medial knee OA progression

Cartilage mechanical responses during gait as in silico biomarkers for medial knee OA progression | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Cartilage mechanical responses during gait as in silico biomarkers for medial knee OA progression Yixuan Zhang, Bryce Adrian Killen, Ikram Mohout, Miel Willems, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6032747/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 09 Oct, 2025 Read the published version in Scientific Reports → Version 1 posted 11 You are reading this latest preprint version Abstract Osteoarthritis (OA) is a common degenerative joint disorder affecting the whole joint, particularly characterized by articular cartilage breakdown, mostly affecting the knee’s medial compartment. Its prevalence is high with aging as an important risk factor. With a global aging population, understanding, preventing, and managing OA becomes increasingly important. Progression of structural knee OA is multifactorial, including biomechanical stressors, inflammatory responses, and genetic predispositions. Traditional attempts to identify biomarkers predicting structural OA progression focus on wet biochemical markers from blood, synovia, or urine. This study assesses in silico loading-related parameters of the cartilage mechanical response as promising predictors of OA progression. A novel MSK-FE workflow relating knee movement to contact pressures and cartilage tissue response was developed. Subjects presenting OA progression over 2 years exhibited elevated medial compartment loading magnitude and posterolateral location shift at baseline. Unsupervised k-means clustering, using strain histograms, successfully differentiated progressors from non-progressors and controls when combining contact pressure and cartilage tissue mechanical responses. This study demonstrates the potential of computationally efficient, in silico mechanical biomarkers to identify personalized OA progression risk after 2 years. This approach offers promising clinical benefit by identifying patients at risk of OA progression, making them eligible for preventative treatment strategies. Health sciences/Diseases/Rheumatic diseases/Osteoarthritis Health sciences/Anatomy/Musculoskeletal system/Cartilage Physical sciences/Engineering/Biomedical engineering Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Osteoarthritis (OA) is the most common chronic joint disease, affecting approximately 242 million people worldwide 1 and imposing a significant socio-economic burden 2 . The risk of OA increases with aging 3 , and with a growing global elderly population, the prevalence of OA is expected to reach 1 billion by 2025 1 . OA is associated with joint pain, inflammation, and reduced functionality, which diminish quality of life and limit physical activity 2 . This inactivity further increases the risks of obesity 4 , cardiovascular disease 5 , and deep vein thrombosis 6 among elderly adults. Cartilage degeneration is a hallmark of OA 7 . Current OA treatments primarily focus on improving joint functionality through physiotherapy and exercise or reducing inflammation and pain through anti-inflammatory drugs 8 , 9 . However, the current treatment cannot fully restore cartilage to its original healthy state. The rate of OA progression varies widely among patients, with some patients exhibiting rapid progression while others show no progression for years 10 . Early interventions can improve treatment outcomes and slow disease progression in patients at risk of fast progression 11 . These interventions can ultimately delay joint replacement surgery, the end-stage treatment that more than 50% of patients eventually require 12 . Therefore, early identification of OA patients at risk of progression is essential to implement efficient early treatments to slow disease advancement and maintain healthy joint function, thereby minimizing the risk of mobility loss and associated comorbidities. Mechanical loading is well-accepted to contribute to articular cartilage degeneration 7 . Altered mechanical loading, such as increased magnitude and changes location, is known to induce cartilage microstructural degeneration and perturb chondrocyte homeostasis in vitro 13 – 15 . The in vitro insights 16 could be further extended to in vivo joint loading during gait, thereby identifying mechanical biomarkers for OA progression based on locomotion patterns. In silico approaches offer an integrative framework that combines 3D motion capture data during locomotion with joint geometry and cartilage mechanical properties to describe the mechanical loading 17 – 20 and subsequent cartilage tissue mechanical responses 21 – 28 . Patient-specific internal joint contact pressures estimated by musculoskeletal (MSK) modeling revealed cross-sectional differences in the location and magnitude of medial 18 and total 19 , 20 knee loading between patients with medial knee OA and controls during gait. Applying patient-specific joint loading to finite element (FE) models of the whole knee joint and incorporating complex fibril-reinforced biphasic material properties of cartilage 21 – 28 , the mechanical responses on fibril strain and maximum shear strain can be estimated. This is highly relevant as in vitro experiments combined with FE modeling studies have identified these loading-associated cartilage responses as critical factors for cartilage microstructural degeneration, driving collagen degradation and proteoglycans depletion 16 , 27 , 29 , 30 . Therefore, such an integrated in silico framework holds the potential for identifying mechanical biomarkers for OA cartilage degeneration and structural disease progression in vivo based on mechanical loading and consequent cartilage response during locomotion. Indeed, in silico studies 19 , 21 , including recent work by our group 27 , confirmed the framework’s ability to identify OA-specific mechanical biomarkers in vivo based on patient-specific gait. However, these studies were limited to cross-sectional comparisons between individual OA patients and controls rather than longitudinal follow-up data. In this study, we introduce the analysis of in silico biomarkers of OA progression in a cohort of subjects identified through a novel modeling workflow that relates joint movement and loading to cartilage tissue mechanical response in the medial compartment during gait. We investigated if in silico biomarkers could distinguish in vivo OA progressors from non-progressors and controls based on a unique longitudinal historical data set 31 , 32 and a novel modeling workflow. Upon confirmation, these in silico functional biomarkers related to cartilage tissue loading could contribute to the identification of elderly subjects at risk for accelerated OA progression and facilitate their selective enrolment in early intervention and preventative rehabilitation strategies. Methods Figure 1 illustrates the study's workflow, detailing the integration of gait analysis, MSK modeling, and FE simulations to identify mechanical cartilage tissue parameters that differentiate OA progressors from non-progressors and controls. Data Collection This study is a secondary analysis of a prospective clinical study using a longitudinal dataset of healthy control and knee OA female participants 31,32 . Ethics approval was obtained from the local ethics committee UZ Leuven (Clinical trial number: S50534) - in accordance with the Declaration of Helsinki. All patients provided written informed consents. Patients with medial knee OA were selected for this analysis based on their baseline Kellgren-Lawrence (KL) scores 33 and a 2-year follow-up. Specifically, patients were required to have a KL-score > 0 in the medial compartment and a higher KL-score in the medial compartment than in the lateral compartment at baseline. OA patients were classified as progressors if their KL score in the medial compartment increased by at least 1 point over the two-year follow up period. Patients whose KL score remained unchanged over the 2 years were categorized as non-progressors, and those with a KL score of 0 in both compartments at both time points were classified as controls. Patients with a KL-score increase of 0.5, a reduced KL score, or lateral compartment progression were excluded. Following group allocation, a total of ten controls (C), nine progressors (P), and eleven non-progressors (NP) were retained for further analysis (Table 1). Detailed information for each subject is tabulated in supplementary Table S1 . Non-parametric Kruskall-Wallis tests were performed using RStudio version 2024.09.1+394 (RStudio, PBC, Boston, MA) to evaluate between-group differences. Table 1 . Summary of the demographic characteristics of control, OA progressors, and non-progressors. KL scores for each group are detailed for the medial compartment at baseline. Detailed are average ± standard deviation for weight, height, body mass index (BMI), and age at baseline. group number of subjects number of subjects with different kl scores at baseline average weight (kg) average height (m) average BMI (kg/m2) average age (year) Gait speed (m/s) First peak Second peak KL1 KL2 KL3 progressor 9 6 2 1 71.7 ± 7.00 1.60 ± 0.06 28.05 ± 3.92 66.7 ± 4.1 1.24 ± 0.18 1.15 ± 0.19 non-progressor 11 7 4 0 67.1 ± 14.0 1.61 ± 0.03 25.90 ± 4.77 64.6 ± 5.6 1.19 ± 0.21 1.13 ± 0.16 control 10 - - - 64.2 ± 10.3 1.61 ± 0.06 24.71 ± 3.76 63.0 ± 10.2 1.20 ± 0.24 1.16 ± 0.21 Previously collected gait data were analyzed 31,32 . including marker positions tracked using a 10MX Vicon Motion capture system at 100Hz, was synchronized with ground reaction forces acquired via AMTI in-ground force plates collected at 1000Hz. Participants performed a static calibration trial 34 in an anatomical position for 5 seconds, followed by over-ground walking trials at a self-selected pace. Marker trajectories and force plate data were processed using custom Matlab (MATLAB R2020b, The Math Works, Inc., Natick, Massachusetts, USA) scripts for subsequent musculoskeletal modeling. Musculoskeletal Model to Estimate Joint Contact Mechanism A state-of-the-art musculoskeletal modeling framework, OpenSim Joint Articular Mechanics (JAM) (https://github.com/clnsmith/opensim-jam), 35 was used to estimate joint kinematics, joint moments, and contact loading parameters. First, a generic OpenSim 36 model 37 was scaled to match each participant’s anthropometry using static trial marker positions. OpenSim-JAM integrates a unique knee joint contact model with standard OpenSim tools to estimate force-dependent kinematics for knee joint secondary coordinates (i.e., tibiofemoral joint internal/external rotation, adduction/abduction and translations, and six-degrees-of-freedom patellofemoral joint) which are dynamically consistent with ligament, muscle, and articular contact forces. Dynamic contact pressure across the knee joint surface, including the center of pressure (COP), was estimated using an elastic foundation model formulation. Articular contact pressures and COP were separately calculated for the medial and lateral compartments. One trial was selected randomly for each participant for further modeling. Joint angles and moments in both hip and knee joints, mean contact pressure, and center of pressure were extracted at the first and second peak of total tibiofemoral joint loading identified using a semi-automated approach through a custom-written Matlab script and exported for subsequent statistical analysis. Parameters were compared between groups at the first and second peaks of tibiofemoral joint loading. Non-parametric Kruskall-Wallis tests followed by Mann-Whitney U tests were then performed to evaluate between-group differences. All statistical analyses were performed using RStudio version 2024.09.1+394 (RStudio, PBC, Boston, MA). Finite Element Model to Estimate Cartilage Mechanical Response Contact pressure estimated by MSK models was extracted for the stance phase of gait and applied to a finite element (FE) model of the medial tibial compartment, as all subjects involved in the project have medial compartment OA. Identical generic cartilage geometries were used in the FE and MSK models. The FE model incorporated hexahedral meshes generated using ANSA (v21.0.1, BETA CAE Systems International AG, Switzerland) with 21,076 elements (C3D8P). This model used a fibril-reinforced poroelastic material (FRPE) 38–40 formulation, which was implemented in Abaqus (Abaqus 2021, Dassault Systèmes Simulia Corp., Providence, RI, USA) via a user-defined material subroutine (UMAT). The material properties were derived from unconfined compression tests on non-OA human cartilage 41 . Detailed information on the material model and properties are documented in the supplementary material (Table S2) . Subject-specific dynamic contact pressures estimated by MSK models were applied to the cartilage's articular surface, while the cartilage's bottom surface, attached to the subchondral bone, was fixed in the model. Three critical mechanical responses—compressive, fibril, and maximum shear strain 27 —were analyzed at the first and second peaks of tibiofemoral joint loading and the maximum value for each element during the entire stance phase of gait. The compressive strain, representing the cartilage deformation under compressive loading, is approximated by the absolute minimum principal strain. Based on the literature and in line with our previous work, fibril strain was found indicative of fibril degradation and maximum shear strain was leading to proteoglycan depletion during OA progression 16,29,30,42,43 . These parameters, including minimum principal strain, fibril strain and maximum shear strain were plotted using a histogram as percentage of volume in Matlab and compared between groups. Pilot Clustering Test to Identify OA Progressors Histograms of contact pressures across the medial compartment estimated by MSK models and cartilage mechanical responses estimated by FE models were used in a pilot clustering analysis to identify OA progressors using a time series k-mean clustering algorithm (tslearn 44 ). Data from each subject were input into the algorithm at the first and second peaks of joint loading, and the maximum value was observed throughout the stance phase. Two clusters were generated, and each histogram was iteratively assigned to the nearest cluster centroid based on the Euclidean distance until convergency. The correctness of clustering was checked to assess the credibility using contact pressures and cartilage mechanical responses as in silico mechanical biomarkers in distinguishing progressors from non-progressors and controls. Results No significant differences in BMI (p = 0.2878), age (p = 0.6707), walking speeds at the first (p = 0.9567) and second peaks (p = 0.9913) of tibiofemoral joint loading were observed. Given the relevance of medial compartment OA progression, the current results section will specifically focus on contact mechanics and cartilage mechanical responses in medial tibial cartilage. However, for completeness, knee and hip joint angles and moments are available in supplementary material in Figures S1 and S2 . In short, the progressors group showed a significantly more externally rotated knee than the control group ( p = 0.013) at the first peak. However, no significant differences in knee adduction moment were found between groups. Furthermore, the contact mechanism in the lateral compartment is illustrated in supplementary Figure S3 . Joint Contact Mechanics Progressors present higher mean contact pressures than controls and non-progressors (Fig. 2 b, top). The increased contact pressures observed in progressors were associated with differences in peak loading location (Figs. 2 a and b, middle and bottom). In the medial compartment, the COP was more posterior in progressors than in controls and non-progressors, especially at the first peak. Furthermore, at the second peak, a lateral shift in the COP was observed in progressors compared to non-progressors and controls (although not statistically significant). Cartilage Mechanical Response Representative tibial strain maps for the medial compartment, estimated by FE models, at the first and second peaks of tibiofemoral contact force are shown in Fig. 3 for each subject group. Compressive strains were calculated using minimum principal strain in cartilage. Progressors exhibited posterolateral shifts in areas with high strain values for all studied strains at the first peak compared to non-progressors and controls. This shift mirrored the displacement of the COP in the medial compartment shown in the MSK model. Due to higher contact pressures, progressors showed slightly larger cartilage volume experiencing higher compressive strain, fibril strain, and maximum shear strain at the first peak compared to non-progressors and controls (Fig. 4 left). Interestingly, a larger volume of cartilage in progressors is experiencing lower strains during the first peak. Similar differences between progressors and other groups are observed at the second peak for compressive and maximum shear strain (Fig. 4 middle). However, the volume difference in cartilage under lower strains is notably less than at the first peak. For fibril strain at the second peak, no clear difference was observed between the three subject groups (Fig. 4 middle). When examining the maximum strain of each element during the stance phase (Fig. 4 right), a clear frequency shift towards higher strains can be observed in all three mechanical responses of progressors compared to the other groups. Clustering analysis to identify OA progressors Unsupervised clustering results for contact pressures estimated from MSK modeling at the first and second peaks of contact forces are presented in Fig. 5 . Detailed clustering results for each subject are listed in Supplementary Table S3 . No dominance of OA progressors (red) or non-progressors (blue) and controls (grey) was observed in either cluster. Contrary to contact pressure, cartilage mechanical responses from the FE analyses showed promising results for clustering. Unsupervised clustering results for the best-performing parameters—compressive strain and maximum shear strain both at the first peak of contact force and fibril strain at the first peak of contact force and maximum of stance phase—are shown in Fig. 6 . Clustering results for other studied strains at the two peaks of contact force and maximum during the stance phase are provided in the supplementary material ( Figure S4 ). Notably, the identified cluster 1 is dominated by OA progressors (in red), while cluster 2 is dominated by non-progressors (in blue) and control subjects (in grey). Compressive and maximum shear strains at the first peak of tibiofemoral joint contact forces achieved identical clustering performance, with Cluster 1 comprising 66.7% of progressors and 18.2% of non-progressors (false positive). Cluster 2 contained 33.3% of progressors (false negative), 81.1% of non-progressors, and all controls. In comparison, fibril strain at the first peak showed more false positives (27.3% of non-progressors in Cluster 1) but fewer false negatives (22.2% of progressors in Cluster 2). At the maximum value during the stance phase, fibril strain classified more false negatives (44.4% of progressors in Cluster 2), though some of these misclassifications did not overlap with those identified by the compressive strain and maximum shear strain at the first peak Discussion This study is the first to identify mechanical cartilage response during gait as biomechanical biomarkers that distinguish patients with progressing medial knee OA from those with non-progressing or controls based on patient-specific gait data. Elevated contact pressure with a posterolateral shifted COP was found in the medial compartment of OA progressors compared to non-progressors and controls. This resulted in an altered cartilage mechanical response pattern characterized by both underloading and overloading strains. These in silico biomechanical biomarkers were evaluated using a preliminary unsupervised clustering analysis, with the cartilage mechanical responses from FE analysis revealing distinct biomechanical characteristics in OA progressors. This longitudinal study confirms increased contact pressures and a posterolateral shift of COP (Fig. 2 ) in the medial compartment were observed in patients with accelerated OA progression This increase in loading magnitude and changes in loading areas among progressors supports the hypothesis that in OA, cartilage mechanical environment is altered and make therefore disrupted the homeostasis due to an imbalance between joint loading and load-bearing capacity 45 . Our findings align with previous literature, which reported a more posterolateral COP in patients with established OA than controls 18 . The altered loading patterns suggest that changes in the location of the loading, rather than just the magnitude, may drive OA progression in its early stages. These results confirm the utility of such advanced musculoskeletal modeling approaches, which provide joint-level estimates of cartilage loading. The finding that no significant differences were found in the knee adduction moments (KAM) of progressors compared to non progressors or controls further highlights their utility and the need to consider direct estimates of joint loading – rather than analyzing joint level kinematics and moments (supplementary Figure S1 ). This contrasts previous literature suggesting KAM may act as a mediator for medial knee OA 46 , 47 . Within this study and others, the KAM's ability as a direct surrogate for joint contact forces is becoming increasingly unclear. To document the disrupted cartilage mechanical environment in OA, contact mechanics estimated by MSK were further applied to FE cartilage models to estimate cartilage mechanical responses. The altered contact mechanics in OA progressors induced changes in the mechanical response of the cartilage structure. The strains (Fig. 3 ) shifted posterolaterally in line with the COP. At the maximum value during the stance phase, higher strains were found in progressors, and a larger volume of cartilage exceeded established degeneration thresholds defined in vitro for fibril degradation (fibril strain = 0.1 27,30 ) and proteoglycans depletion (maximum shear strain = 0.4 48 or 0.5 16 ), indicating an elevated risk for cartilage degeneration in these patients. This aligns with previous computational studies comparing OA patients and controls 49 , 50 . However, the earlier study did not distinguish OA progressors and non-progressors. At the first and second peaks of tibiofemoral (TF) contact forces, cartilage mechanical response does not always increase proportionally as contact loading increases due to cartilage's non-linear and time-dependent nature. Signature histograms of compressive and maximum shear strains revealed a dual trend in progressors, where a larger volume of elements experienced low-range and high-range strains at the first and second peaks. This dual pattern indicated OA progressors experiencing both under- and overloading at the first and second peaks, suggesting that an altered loading pattern, rather than just increased magnitude, could also discriminate OA progressors. Interestingly, neither the contact mechanics nor cartilage mechanical responses differed between non-progressors and controls. Although earlier studies reported elevated knee joint loading 18 , 19 , 51 and increased cartilage volumes exceeding degeneration thresholds 49 , 50 in patients with knee OA compared to controls, these did not distinguish between progressing and non-progressing OA patients. Therefore, previously reported differences observed between OA patients and controls align primarily with progressors within the OA group. Unsupervised k-means clustering was employed for histograms of contact pressures and cartilage mechanical responses to assess further their feasibility in distinguishing OA progressors from non-progressors and controls. Although more information is provided with a histogram than the mean value, contact pressure could not discriminate OA progressors (Fig. 5 ). For the selected cartilage mechanical responses (Fig. 6 ), the clustering analysis produced distinct divisions, with one cluster predominantly containing progressors and the other containing non-progressors and controls. This distinct separation supports the idea that histograms of cartilage mechanical responses could serve as in silico biomarkers of OA progression. Notably, the unsupervised clustering analysis effectively captured the dual trend signature in compressive strain and maximum shear strain at the first and second peaks, along with the shift toward higher strain values at the maximum stance phase. This suggests that cartilage mechanical responses could be sensitive and reliable biomarkers for identifying fast future OA progression. Despite the success of this work, which clustered OA progressors based on cartilage mechanical responses at baseline, we know that more patient-specific gait data are needed to train and validate the clustering model in the future, particularly for clinical applications. Therefore, future studies will collect larger longitudinal datasets to refine these mechanical biomarkers and improve the accuracy of classifying OA progressors. Indeed, investigating more in detail individual subject results, two progressors and two non-progressors exhibited deviant mechanical responses that resembled the opposite group’s characteristics (see supplementary Table S3 , P01, P07, NP07 and NP08). These four subjects accounted for most of the false clustering results. Moreover, no specific characteristics were found in joint kinematics or age of those 4 outliners. However, the progressor outliners (P01 and P07) showed higher BMI compared to the group average (Table 1 and Table 1 ), indicating high joint loading. Whereas, the non-progressor outliners (NP07 and NP08) had lower BMI compared to their group average (Table 1 and Table 1 ). Furthermore, it is essential to acknowledge that mechanical loading and joint kinematics during gait are not the only factors contributing to cartilage degeneration during OA progression. Incorporating other factors, such as systemic inflammation, genetic predisposition, and high-impact physical activities 52 , 53 , could provide a more comprehensive understanding of OA structural progression and thus improve prediction accuracy for future clinical applications. While this study offers valuable insights, certain aspects could be refined to enhance the prediction of OA progression in future investigations. For instance, the study relied on clinical standard KL-score to assess structural OA progression. However, the assessment is based on x-ray imaging, which lacks the ability to capture detailed cartilage degeneration in the joint Future studies could incorporate Magnetic Resonance Imaging (MRI) to examine cartilage degeneration more precisely and compare the degeneration location with high-strain regions estimated by the in silico model, further validating the model prediction. In terms of the modelling framework, although the cartilage geometry was scaled to the size of each patient in the models, patient-specific cartilage geometry and tibiofemoral alignment were not included in the current workflow. These anatomical variations could further alter the contact mechanism in the joint 54 , potentially changing the mechanical responses of the cartilage. To incorporate patient-specific anatomy, a recent MSK model that directly estimates joint contact mechanics, including patient-specific cartilage geometry and tibiofemoral alignment 55 , could be integrated within the current MSK-FE modeling workflow as a next step. Additionally, this study used identical cartilage material properties for both control and OA subjects. This choice seems valid as most OA subjects in this study had a KL score of 1 or 2 at baseline, indicating early or moderate OA with limited cartilage degeneration 56 . However, incorporating OA state-specific material properties and initial cartilage conditions in future studies could enhance the workflow’s ability to predict in vivo cartilage degeneration. Lastly, the study focused on the chronic mechanical alterations caused by gait patterns in OA progressors. Post-traumatic OA progression, which can result from ligament rupture, cartilage lesions, or meniscus tears, is not assessed. As post-traumatic OA progression is particularly relevant for younger patients, our relevance of this workflow adapted to include the effect of meniscal injuries, cartilage lesions, or changes in ligament properties needs to be analysed in future studies. Notably, while high personalization of joint geometry and cartilage mechanical properties could enhance the estimation of cartilage mechanical responses, achieving this would necessitate additional imaging techniques, such as MRI or ultrasound. This would, however, increase clinical costs and substantially raise computational demands. Although such highly personalized analysis of cartilage mechanical response may eventually benefit patients-tailored rehabilitation programs, it would limit the feasibility of large-scale joint health screening in older people. Joint health screening on a large scale would enable early identification of individuals at risk of fast OA progression, allowing for timely preventive interventions to slow disease progression and preserve joint function. To make such screenings viable at a population level, cost-effectiveness and prediction accuracy must be balanced. Based on patient-specific gait with generic geometries and mechanical properties without including high personalization, this proposed workflow demonstrates its sensitivity to joint loading during locomotion and ability to differentiate structural OA progressors using in silico biomarkers derived from patient-specific cartilage mechanical responses. Furthermore, compared to previous MSK-FE workflows 26 , 27 , 49 , 57 , 58 , this approach significantly reduces computational cost and complexity, enhancing its potential for clinical screening of OA progression. In conclusion, using an integrated MSK-FE modeling approach, cartilage mechanical responses during gait effectively distinguished medial knee OA progressors (over two years) from non-progressors and controls. These mechanical responses could serve as sensitive in silico biomarkers to identify patients at risk of accelerated OA progression, further supported by a pilot unsupervised clustering analysis. With future validation in larger cohorts, this workflow could be adapted for clinical application, enabling screening for joint health in the elderly and early identification of patients at risk of OA progression. This would allow timely access to preventive interventions such as weight management, neuromuscular exercise 59 and gait retraining 60 , helping to preserve joint function and mobility and reduce risks of mobility loss. Declarations Acknowledgments This work was supported by the FWO Happy Joints project (G045320N), the FWO fundament fellowship (11PK924N), and H2020 Project OACTIVE (SC1-PM17-2017). Author contributions BAK and YZ were responsible for data processing and simulations. IJ, SAE, BAK, IM, MW and YZ contributed to the study design and workflow. SV and FL were responsible for baseline collection and longitudinal follow-up of subjects' gait and OA levels. IJ, SAE, BAK, and YZ were responsible for the initial draft of the manuscript and interpretation of the results. All authors contributed to the final drafting of the submitted version. Data availability statement The datasets generated during and/or analysed during the current study are available from the corresponding author on request. Additional information The authors declare no competing interests. References Steinmetz, J. D. et al. Global, regional, and national burden of osteoarthritis, 1990–2020 and projections to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet Rheumatol. 5 , e508–e522 (2023). Hunter, D. J., Schofield, D. & Callander, E. The individual and socioeconomic impact of osteoarthritis. Nature Reviews Rheumatology vol. 10 437–441 at (2014). https://doi.org/10.1038/nrrheum.2014.44 Loeser, R. F. Aging and osteoarthritis. Curr. Opin. Rheumatol. 23 , 492–496 (2011). Erlichman, J., Kerbey, A. L. & James, W. P. T. Physical activity and its impact on health outcomes. Paper 1: The impact of physical activity on cardiovascular disease and all-cause mortality: An historical perspective. Obes. Rev. 3 , 257–271 (2002). Lacombe, J., Armstrong, M. E. G., Wright, F. L. & Foster, C. The impact of physical activity and an additional behavioural risk factor on cardiovascular disease, cancer and all-cause mortality: A systematic review. BMC Public. Health 19 , (2019). Truong, A. D., Fan, E., Brower, R. G. & Needham, D. M. Bench-to-bedside review: mobilizing patients in the intensive care unit–from pathophysiology to clinical trials. Crit. Care . 13 , 216 (2009). Lories, R. J. & Luyten, F. P. The bone-cartilage unit in osteoarthritis. Nat. Rev. Rheumatol. 7 , 43–49 (2011). Conley, B. et al. Core Recommendations for Osteoarthritis Care: A Systematic Review of Clinical Practice Guidelines. Arthritis Care Res. 75 , 1897–1907 (2023). Marriott, K. A. & Birmingham, T. B. Fundamentals of osteoarthritis. Rehabilitation: Exercise, diet, biomechanics, and physical therapist-delivered interventions. Osteoarthr. Cartil. 31 , 1312–1326 (2023). Emrani, P. S. et al. Joint space narrowing and Kellgren-Lawrence progression in knee osteoarthritis: an analytic literature synthesis. Osteoarthr. Cartil. 16 , 873–882 (2008). Hawker, G. A. & Lohmander, L. S. What an earlier recognition of osteoarthritis can do for OA prevention. Osteoarthr. Cartil. 29 , 1632–1634 (2021). Weinstein, A. M. et al. Estimating the burden of total knee replacement in the United States. J. Bone Jt. Surg. 95 , 385–392 (2013). Bader, D. L., Salter, D. M. & Chowdhury, T. T. Biomechanical Influence of Cartilage Homeostasis in Health and Disease. Arthritis 1–16 (2011). (2011). Sun, H. B. Mechanical loading, cartilage degradation, and arthritis. Ann. N Y Acad. Sci. 1211 , 37–50 (2010). Varady, N. H. & Grodzinsky, A. J. Osteoarthritis year in review 2015: Mechanics. Osteoarthr. Cartil. 24 , 27–35 (2016). Orozco, G. A., Tanska, P., Florea, C., Grodzinsky, A. J. & Korhonen, R. K. A novel mechanobiological model can predict how physiologically relevant dynamic loading causes proteoglycan loss in mechanically injured articular cartilage. Sci. Rep. 8 , 1–16 (2018). Smith, C. R., Choi, W., Negrut, K., Thelen, D. G. & D. & Efficient computation of cartilage contact pressures within dynamic simulations of movement. Comput. Methods Biomech. Biomed. Eng. Imaging Vis. 6 , 491–498 (2018). Meireles, S. et al. Medial knee loading is altered in subjects with early osteoarthritis during gait but not during step-up-and-over task. PLoS One 12 , (2017). Marouane, H., Shirazi-Adl, A. & Adouni, M. Alterations in knee contact forces and centers in stance phase of gait: A detailed lower extremity musculoskeletal model. J. Biomech. 49 , 185–192 (2016). Kumar, R. et al. Polarization second harmonic generation microscopy provides quantitative enhanced molecular specificity for tissue diagnostics. J. Biophotonics . 8 , 730–739 (2015). Adouni, M. & Shirazi-Adl, A. Evaluation of knee joint muscle forces and tissue stresses-strains during gait in severe OA versus normal subjects. J. Orthop. Res. 32 , 69–78 (2014). Halonen, K. S., Mononen, M. E., Jurvelin, J. S., Töyräs, J. & Korhonen, R. K. Importance of depth-wise distribution of collagen and proteoglycans in articular cartilage-A 3D finite element study of stresses and strains in human knee joint. J. Biomech. 46 , 1184–1192 (2013). Halonen, K. S. et al. Deformation of articular cartilage during static loading of a knee joint - Experimental and finite element analysis. J. Biomech. 47 , 2467–2474 (2014). Halonen, K. S. et al. Importance of patella, quadriceps forces, and depthwise cartilage structure on knee joint motion and cartilage response during gait. J. Biomech. Eng. 138 , 1–11 (2016). Erdemir, A. et al. Deciphering the ‘Art’ in Modeling and Simulation of the Knee Joint: Overall Strategy. J. Biomech. Eng. 141 , 1–10 (2019). Esrafilian, A. et al. EMG-Assisted Muscle Force Driven Finite Element Model of the Knee Joint with Fibril-Reinforced Poroelastic Cartilages and Menisci. Sci. Rep. 10 , 1–16 (2020). Mohout, I. et al. Signatures of disease progression in knee osteoarthritis: insights from an integrated multi-scale modeling approach, a proof of concept. Front. Bioeng. Biotechnol. 11 , 1–12 (2023). Orozco, G. A. et al. Shear strain and inflammation-induced fixed charge density loss in the knee joint cartilage following ACL injury and reconstruction: A computational study. J. Orthop. Res. 40 , 1505–1522 (2022). Eskelinen, A. S. A., Mononen, M. E., Venäläinen, M. S., Korhonen, R. K. & Tanska, P. Maximum shear strain-based algorithm can predict proteoglycan loss in damaged articular cartilage. Biomech. Model. Mechanobiol. 18 , 753–778 (2019). Elahi, S. A. et al. An in silico Framework of Cartilage Degeneration That Integrates Fibril Reorientation and Degradation Along With Altered Hydration and Fixed Charge Density Loss. Front. Bioeng. Biotechnol. 9 , (2021). Baert, I. A. C. et al. Gait characteristics and lower limb muscle strength in women with early and established knee osteoarthritis. Clin. Biomech. 28 , 40–47 (2013). Mahmoudian, A. et al. Varus thrust in women with early medial knee osteoarthritis and its relation with the external knee adduction moment. Clin. Biomech. 39 , 109–114 (2016). Kellgren, J. H. Radiological assessment. Ann. Rheum. Dis. 494–502. 10.2307/3578513 (1957). Cappozzo, Catani, F. & Della Croce, U. Leardini, a. Position and orietnation in space of bones during movement. Clin. Biomech. 10 , 171–178 (1995). Smith, C. R., Lenhart, R. L., Kaiser, J., Vignos, M. F. & Thelen, D. G. Influence of Ligament Properties on Tibiofemoral Mechanics in Walking. J. Knee Surg. 29 , 99–106 (2014). Delp, S. L. et al. OpenSim: Open-source software to create and analyze dynamic simulations of movement. IEEE Trans. Biomed. Eng. 54 , 1940–1950 (2007). Lenhart, R. L., Kaiser, J., Smith, C. R. & Thelen, D. G. Prediction and Validation of Load-Dependent Behavior of the Tibiofemoral and Patellofemoral Joints During Movement. Ann. Biomed. Eng. 43 , 2675–2685 (2015). Ebrahimi, M. et al. Elastic, Viscoelastic and Fibril-Reinforced Poroelastic Material Properties of Healthy and Osteoarthritic Human Tibial Cartilage. Ann. Biomed. Eng. 47 , 953–966 (2019). Ebrahimi, M. et al. Structure–Function Relationships of Healthy and Osteoarthritic Human Tibial Cartilage: Experimental and Numerical Investigation. Ann. Biomed. Eng. 48 , 2887–2900 (2020). Elahi, S. A. et al. Guide to mechanical characterization of articular cartilage and hydrogel constructs based on a systematic in silico parameter sensitivity analysis. J. Mech. Behav. Biomed. Mater. 124 , (2021). Castro-Vinuelas, S. A. E. R. & Govaerts, A. Unconfined Compression Experimental Protocol for Cartilage Explants and Hydrogel Constructs: From Sample Preparation to Mechanical Characterization. Cartil. Tissue Engineering: Introduction . 271–287. 10.1007/978-1-0716-2839-3_1 (2023). Elahi, S. A. et al. Contribution of collagen degradation and proteoglycan depletion to cartilage degeneration in primary and secondary osteoarthritis: an in silico study. Osteoarthr. Cartil. 31 , 741–752 (2023). Hosseini, S. M., Wilson, W., Ito, K. & Van Donkelaar, C. C. A numerical model to study mechanically induced initiation and progression of damage in articular cartilage. Osteoarthr. Cartil. 22 , 95–103 (2014). Tavenard, R. et al. A Machine Learning Toolkit for Time Series Data. J. Mach. Learn. Res. 21 , 1–6 (2020). Tslearn. Andriacchi, T. P., Smith, R. L., Medicine, S. & Koo, S. A Framework for the in Vivo Pathomechanics of Osteoarthritis at the Knee: 2nd Special Edition on Musculoskeletal Bioengineering. Guest Editor : Kyriacos A. A Framework for the in Vivo Pathomechanics of Osteoarthritis at the Knee. (2004). 10.1023/B Miyazaki, T. et al. Dynamic load at baseline can predict radiographic disease progression in medial compartment knee osteoarthritis. Ann. Rheum. Dis. 61 , 617–622 (2002). Bennell, K. L. et al. Higher dynamic medial knee load predicts greater cartilage loss over 12 months in medial knee osteoarthritis. Ann. Rheum. Dis. 70 , 1770–1774 (2011). Orozco, G. A. et al. Prediction of local fixed charge density loss in cartilage following ACL injury and reconstruction: A computational proof-of-concept study with MRI follow-up. J. Orthop. Res. 39 , 1064–1081 (2021). Liukkonen, M. K. et al. Simulation of subject-specific progression of knee osteoarthritis and comparison to experimental follow-up data: Data from the osteoarthritis initiative. Sci. Rep. 7 , 1–14 (2017). Orozco, G. A. et al. Effect of patient specificity on predicting knee cartilage degeneration in obese adults: Musculoskeletal finite-element modeling of data from the CAROT trial. J. Orthop. Res. 10.1002/jor.25912 (2024). Kumar, D., Manal, K. T. & Rudolph, K. S. Knee joint loading during gait in healthy controls and individuals with knee osteoarthritis. Osteoarthr. Cartil. 21 , 298–305 (2013). Greene, M. A. & Loeser, R. F. Aging-related inflammation in osteoarthritis. Osteoarthr. Cartil. 23 , 1966–1971 (2015). Martel-Pelletier, J. et al. Osteoarthritis. Nature Reviews Disease Primers vol. 2 at (2016). https://doi.org/10.1038/nrdp.2016.72 Willems, M. et al. Population-based in silico modeling of anatomical shape variation of the knee and its impact on joint loading in knee osteoarthritis. J. Orthop. Res. 1–12. 10.1002/jor.25934 (2024). Killen, B. A., Willems, M. & Jonkers, I. P. R. E. P. R. I. N. T. An Open-source framework for the generation of OpenSim models with personalised knee joint geometries for the estimation of articular contact mechanics. J. Biomech. 177 , 112387 (2024). Roemer, F. W. et al. Heterogeneity of cartilage damage in Kellgren and Lawrence grade 2 and 3 knees: the MOST study. Osteoarthr. Cartil. 30 , 714–723 (2022). Esrafilian, A. et al. Toward Tailored Rehabilitation by Implementation of a Novel Musculoskeletal Finite Element Analysis Pipeline. IEEE Trans. Neural Syst. Rehabil Eng. 30 , 789–802 (2022). Mononen, M. E., Liukkonen, M. K. & Korhonen, R. K. Utilizing Atlas-Based Modeling to Predict Knee Joint Cartilage Degeneration: Data from the Osteoarthritis Initiative. Ann. Biomed. Eng. 47 , 813–825 (2019). Roos, E. M. & Arden, N. K. Strategies for the prevention of knee osteoarthritis. Nat. Rev. Rheumatol. 12 , 92–101 (2016). Richards, R., van den Noort, J. C., Dekker, J. & Harlaar, J. Gait Retraining With Real-Time Biofeedback to Reduce Knee Adduction Moment: Systematic Review of Effects and Methods Used. Arch. Phys. Med. Rehabil . 98 , 137–150 (2017). Additional Declarations No competing interests reported. Supplementary Files Supplemenaryfinaltosubmit.docx Cite Share Download PDF Status: Published Journal Publication published 09 Oct, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 15 Jul, 2025 Reviews received at journal 15 Jul, 2025 Reviewers agreed at journal 08 Jul, 2025 Reviews received at journal 04 Apr, 2025 Reviewers agreed at journal 04 Mar, 2025 Reviewers agreed at journal 28 Feb, 2025 Reviewers agreed at journal 28 Feb, 2025 Reviewers invited by journal 28 Feb, 2025 Editor assigned by journal 28 Feb, 2025 Submission checks completed at journal 27 Feb, 2025 First submitted to journal 27 Feb, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6032747","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":448417736,"identity":"0550fd19-3e92-48d9-bca2-27263a78c3ef","order_by":0,"name":"Yixuan Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABD0lEQVRIie2OMUvDQBTHXziIy9OuF6KtH+GFg6ikkK9iKaTLIY5u6hKXuuejOB4cxCWSNUK3QBcrBFwsIniRCl0OOwreb3iPB+/H/w/gcPxF2GZTPzoFh5tb7aZ4hQL8XfmhVxjuotAjxnz9AOnJ3pN+GVcaB1xG7SUshlZF+1lwXwE7m19kiWw0BoUUooClsCus5Ps5+KRkLGSnkZosCxH05MaqeHnwmQNSvYrFqVHSJpt9GOXarrAyNCmcGilaMMWIT0tmlHNbsUD70+Qo50TNKvbm1Qx51eoQaRnZUg7qKnp+zccp1VJ072UyHNxNbt/wajGypRyr78X74fOtwjYBYLQVzzr7n8PhcPxnvgDhZFYBtfFoVQAAAABJRU5ErkJggg==","orcid":"","institution":"KU Leuven","correspondingAuthor":true,"prefix":"","firstName":"Yixuan","middleName":"","lastName":"Zhang","suffix":""},{"id":448417737,"identity":"b012b736-e4b2-43a7-813e-9fd65de1c714","order_by":1,"name":"Bryce Adrian Killen","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Bryce","middleName":"Adrian","lastName":"Killen","suffix":""},{"id":448417738,"identity":"e0b7e872-109f-449c-b6c5-bbc6d42ed0d5","order_by":2,"name":"Ikram Mohout","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Ikram","middleName":"","lastName":"Mohout","suffix":""},{"id":448417739,"identity":"be1e61af-93c5-4732-bb0e-da6d5381f062","order_by":3,"name":"Miel Willems","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Miel","middleName":"","lastName":"Willems","suffix":""},{"id":448417740,"identity":"5bfb2b6d-233f-4885-b303-f7e3d218ce54","order_by":4,"name":"Frank Luyten","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Frank","middleName":"","lastName":"Luyten","suffix":""},{"id":448417741,"identity":"dca26c62-6313-4ffc-8adc-fc6d91115800","order_by":5,"name":"Sabine Verschueren","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Sabine","middleName":"","lastName":"Verschueren","suffix":""},{"id":448417742,"identity":"c3d42167-276f-4ec7-9e69-8edd7af9b9b8","order_by":6,"name":"Seyed Ali Elahi","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Seyed","middleName":"Ali","lastName":"Elahi","suffix":""},{"id":448417743,"identity":"742ed5cf-9fb0-4b7a-bcdc-f6451c3be762","order_by":7,"name":"Ilse Jonkers","email":"","orcid":"","institution":"KU Leuven","correspondingAuthor":false,"prefix":"","firstName":"Ilse","middleName":"","lastName":"Jonkers","suffix":""}],"badges":[],"createdAt":"2025-02-14 18:38:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6032747/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6032747/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-19371-2","type":"published","date":"2025-10-09T15:57:22+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":82549258,"identity":"b3ca52ec-52dc-43fa-8c2c-e21058fbf73c","added_by":"auto","created_at":"2025-05-12 19:29:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":401671,"visible":true,"origin":"","legend":"\u003cp\u003eStudy workflow. \u003cstrong\u003ea)\u003c/strong\u003eHealthy controls (C) and patients with medial knee OA are recruited for 3D motion capture and x-ray to assess Kellgren-Lawrence (KL) scores of both knees. OA patients are divided into progressors (P) and non-progressors (NP) groups based on the \u003cstrong\u003echange \u003c/strong\u003eof KL score in the 2-year follow-up. \u003cstrong\u003eb-c)\u003c/strong\u003eJoint contact mechanics are estimated by MSK modeling using 3D motion capture. The dynamic contact pressure of the stance phase estimated by MSK models is then applied to FE cartilage models. Group-level analysis of contact mechanics and cartilage tissue mechanical responses are performed to identify in \u003cem\u003esilico\u003c/em\u003e mechanical biomarkers characteristic for OA progression. \u003cstrong\u003ed)\u003c/strong\u003e A pilot clustering test uses cartilage mechanical responses estimated for individual subjects to identify OA progressors.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/b041cb15ea79723d1c5d90ca.png"},{"id":82549474,"identity":"495f9cf9-257e-4654-b1f3-03a1554ece88","added_by":"auto","created_at":"2025-05-12 19:37:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":359070,"visible":true,"origin":"","legend":"\u003cp\u003eJoint contact mechanics were estimated for groups of OA progressors (red), non-progressors (blue), and control subjects (grey) at the first and second peaks of tibiofemoral (TF) contact forces.\u003cstrong\u003e a)\u003c/strong\u003e Group-average tibial contact pressure maps. Anatomical directions of medial (M), lateral (L), anterior (A), and posterior (P) are noted in the bottom left corner. \u003cstrong\u003eb)\u003c/strong\u003e Mean medial compartment cartilage contact pressure (top) and medial compartment center of pressure (COP) in the anterior/posterior (middle) and medial/lateral (bottom) direction. Individual subject values are shown by circles, with the thicker circles indicating a higher KL score at baseline. Dark squares indicate the group average and vertical lines indicate the entire range. Stars * indicate significant between-group differences (p\u0026lt;0.05 in black and 0.05\u0026lt;p\u0026lt;0.1 in grey) using the Mann Whitney U test with a p-value marked. \u003cstrong\u003ec)\u003c/strong\u003eHistograms of medial compartment cartilage contact pressure per area of articular contact surface for each subject group.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/4bd5cf0f8027bcd337beb86f.png"},{"id":82549860,"identity":"cb62b4eb-6f76-4645-bda6-e4ad374d2b5e","added_by":"auto","created_at":"2025-05-12 19:45:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":281957,"visible":true,"origin":"","legend":"\u003cp\u003eFE Simulated strain maps in the medial cartilage of compressive strain, fibril strain, and maximum shear strain of representative subjects in the subject groups of OA progressors (P09), non-progressors (NP09), and control (C08) at first and second peaks of TF contact forces. Both progressor and non-progressor score KL1 in the medial compartment at baseline. Anatomical directions of medial (M), lateral (L), anterior (A), and posterior (P) are noted in the top right corner.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/ddd07d933708a0443e79206f.png"},{"id":82549475,"identity":"dc594b00-1f52-47ac-9eef-792d9ca44e28","added_by":"auto","created_at":"2025-05-12 19:37:56","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":274036,"visible":true,"origin":"","legend":"\u003cp\u003eHistograms of simulated compressive strain (absolute minimum principal strain), fibril strain, and maximum shear strain per cartilage volume in cartilage medial compartment of OA progressors (red), non-progressors (blue), and controls (grey). Strains are obtained at the first (left column) and second (middle column) of TF contact force peaks. The histogram for maximum strains of each element during the stance phase is also presented in the right column.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/eb6ecdcdffb88aa78105a8d2.png"},{"id":82549265,"identity":"428d6530-00fd-493e-af68-b4aa00d75915","added_by":"auto","created_at":"2025-05-12 19:29:56","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":235106,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure 5.\u003c/strong\u003e Unsupervised clustering results for contact pressure at a) the first peak and b) the second peak of TF contact forces. Histograms of individual OA progressors, non-progressors, and control subjects are presented in red, blue, and grey, respectively, with black lines indicating the average values of the clustered groups.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/8f62000e544a9ec2afb797e3.png"},{"id":82549479,"identity":"c9eee88e-50d2-49d0-996a-003a20b479ed","added_by":"auto","created_at":"2025-05-12 19:37:56","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":415743,"visible":true,"origin":"","legend":"\u003cp\u003eUnsupervised clustering results for histograms of key mechanical parameters, including \u003cstrong\u003ea)\u003c/strong\u003e compressive strain (absolute minimum principal strain) at the first peak of TF contact forces, \u003cstrong\u003eb) \u003c/strong\u003emaximum shear strain at the first peak of TF contact forces, \u003cstrong\u003ec)\u003c/strong\u003efibril strain at the first peak of TF contact forces, and \u003cstrong\u003ed) \u003c/strong\u003efibril strain at the maximum value of the stance phase. Histograms of individual OA progressors, non-progressors, and control subjects are presented in red, blue, and grey, respectively, with black lines indicating the average values of the clustered groups.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/c8b3845e09b675580093c8a4.png"},{"id":93420708,"identity":"021c8b34-62eb-43c1-8150-efd68313e393","added_by":"auto","created_at":"2025-10-13 16:10:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2789654,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/13a76cff-2e6a-4935-a44b-f866e62238e7.pdf"},{"id":82549477,"identity":"a9d2686a-c749-4fc2-af95-29a861924eeb","added_by":"auto","created_at":"2025-05-12 19:37:56","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":763931,"visible":true,"origin":"","legend":"","description":"","filename":"Supplemenaryfinaltosubmit.docx","url":"https://assets-eu.researchsquare.com/files/rs-6032747/v1/6ffbe3328f99f2d99e5948e3.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Cartilage mechanical responses during gait as in silico biomarkers for medial knee OA progression","fulltext":[{"header":"Introduction","content":"\u003cp\u003eOsteoarthritis (OA) is the most common chronic joint disease, affecting approximately 242\u0026nbsp;million people worldwide\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eand imposing a significant socio-economic burden\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. The risk of OA increases with aging\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, and with a growing global elderly population, the prevalence of OA is expected to reach 1\u0026nbsp;billion by 2025\u003csup\u003e1\u003c/sup\u003e. OA is associated with joint pain, inflammation, and reduced functionality, which diminish quality of life and limit physical activity\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. This inactivity further increases the risks of obesity\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, cardiovascular disease\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, and deep vein thrombosis\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e among elderly adults. Cartilage degeneration is a hallmark of OA\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Current OA treatments primarily focus on improving joint functionality through physiotherapy and exercise or reducing inflammation and pain through anti-inflammatory drugs\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. However, the current treatment cannot fully restore cartilage to its original healthy state.\u003c/p\u003e \u003cp\u003eThe rate of OA progression varies widely among patients, with some patients exhibiting rapid progression while others show no progression for years\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Early interventions can improve treatment outcomes and slow disease progression in patients at risk of fast progression\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. These interventions can ultimately delay joint replacement surgery, the end-stage treatment that more than 50% of patients eventually require\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Therefore, early identification of OA patients at risk of progression is essential to implement efficient early treatments to slow disease advancement and maintain healthy joint function, thereby minimizing the risk of mobility loss and associated comorbidities.\u003c/p\u003e \u003cp\u003eMechanical loading is well-accepted to contribute to articular cartilage degeneration\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Altered mechanical loading, such as increased magnitude and changes location, is known to induce cartilage microstructural degeneration and perturb chondrocyte homeostasis \u003cem\u003ein vitro\u003c/em\u003e \u003csup\u003e\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. The \u003cem\u003ein vitro\u003c/em\u003e insights\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e could be further extended to \u003cem\u003ein vivo\u003c/em\u003e joint loading during gait, thereby identifying mechanical biomarkers for OA progression based on locomotion patterns.\u003c/p\u003e \u003cp\u003e \u003cem\u003eIn silico\u003c/em\u003e approaches offer an integrative framework that combines 3D motion capture data during locomotion with joint geometry and cartilage mechanical properties to describe the mechanical loading \u003csup\u003e\u003cspan additionalcitationids=\"CR18 CR19\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e and subsequent cartilage tissue mechanical responses\u003csup\u003e\u003cspan additionalcitationids=\"CR22 CR23 CR24 CR25 CR26 CR27\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Patient-specific internal joint contact pressures estimated by musculoskeletal (MSK) modeling revealed cross-sectional differences in the location and magnitude of medial\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e and total\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e knee loading between patients with medial knee OA and controls during gait. Applying patient-specific joint loading to finite element (FE) models of the whole knee joint and incorporating complex fibril-reinforced biphasic material properties of cartilage\u003csup\u003e\u003cspan additionalcitationids=\"CR22 CR23 CR24 CR25 CR26 CR27\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, the mechanical responses on fibril strain and maximum shear strain can be estimated. This is highly relevant as \u003cem\u003ein vitro\u003c/em\u003e experiments combined with FE modeling studies have identified these loading-associated cartilage responses as critical factors for cartilage microstructural degeneration, driving collagen degradation and proteoglycans depletion \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Therefore, such an integrated in silico framework holds the potential for identifying mechanical biomarkers for OA cartilage degeneration and structural disease progression \u003cem\u003ein vivo\u003c/em\u003e based on mechanical loading and consequent cartilage response during locomotion.\u003c/p\u003e \u003cp\u003eIndeed, \u003cem\u003ein silico\u003c/em\u003e studies\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, including recent work by our group\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e, confirmed the framework\u0026rsquo;s ability to identify OA-specific mechanical biomarkers \u003cem\u003ein vivo\u003c/em\u003e based on patient-specific gait. However, these studies were limited to cross-sectional comparisons between individual OA patients and controls rather than longitudinal follow-up data. In this study, we introduce the analysis of \u003cem\u003ein silico\u003c/em\u003e biomarkers of OA progression in a cohort of subjects identified through a novel modeling workflow that relates joint movement and loading to cartilage tissue mechanical response in the medial compartment during gait. We investigated if \u003cem\u003ein silico\u003c/em\u003e biomarkers could distinguish \u003cem\u003ein vivo\u003c/em\u003e OA progressors from non-progressors and controls based on a unique longitudinal historical data set\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e and a novel modeling workflow. Upon confirmation, these \u003cem\u003ein silico\u003c/em\u003e functional biomarkers related to cartilage tissue loading could contribute to the identification of elderly subjects at risk for accelerated OA progression and facilitate their selective enrolment in early intervention and preventative rehabilitation strategies.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eFigure\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e1\u003c/strong\u003e illustrates the study\u0026apos;s workflow, detailing the integration of gait analysis, MSK modeling, and FE simulations to identify mechanical cartilage tissue parameters that differentiate OA progressors from non-progressors and controls.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eData Collection\u003c/h2\u003e\n\u003cp\u003eThis study is a secondary analysis of a prospective clinical study using a longitudinal dataset of healthy control and knee OA female participants\u003csup\u003e31,32\u003c/sup\u003e. Ethics approval was obtained from the local ethics committee UZ Leuven (Clinical trial number: S50534) - in accordance with the Declaration of Helsinki. All patients provided written informed consents.\u003c/p\u003e\n\u003cp\u003ePatients with medial knee OA \u0026nbsp;were selected for this analysis based on their baseline Kellgren-Lawrence (KL) scores\u003csup\u003e33\u003c/sup\u003e and a 2-year follow-up. Specifically, patients were required to have a KL-score \u0026gt; 0 in the medial compartment and a higher KL-score in the medial compartment than in the lateral compartment at baseline. OA patients were classified as progressors if their KL score in the medial compartment increased by at least 1 point over the two-year follow up period. Patients whose KL score remained unchanged over the 2 years were categorized as non-progressors, and those with a KL score of 0 in both compartments at both time points were classified as controls. Patients with a KL-score increase of 0.5, a reduced KL score, or lateral compartment progression were excluded.\u003c/p\u003e\n\u003cp\u003eFollowing group allocation, a total of ten controls (C), nine progressors (P), and eleven non-progressors (NP) were retained for further analysis (Table 1). Detailed information for each subject is tabulated in supplementary \u003cstrong\u003eTable S1\u003c/strong\u003e. Non-parametric Kruskall-Wallis tests were performed using RStudio version 2024.09.1+394 (RStudio, PBC, Boston, MA) to evaluate between-group differences.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e1\u003c/strong\u003e\u003cstrong\u003e.\u003c/strong\u003e Summary of the demographic characteristics of control, OA progressors, and non-progressors. KL scores for each group are detailed for the medial compartment at baseline. Detailed are average \u0026plusmn; standard deviation for weight, height, body mass index (BMI), and age at baseline.\u0026nbsp;\u003c/p\u003e\n\u003ctable style=\"border: none;width:445.0pt;border-collapse:collapse;\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"3\" style=\"width:48.0pt;border:solid #7F7F7F 1.0pt;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003egroup\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"3\" style=\"width:40.0pt;border:solid #7F7F7F 1.0pt;border-left:none;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003enumber of subjects\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" rowspan=\"2\" style=\"width:102.0pt;border:solid #7F7F7F 1.0pt;border-left:none;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003enumber of subjects with\u0026nbsp;\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003edifferent kl scores at baseline\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"3\" style=\"width:43.0pt;border:solid #7F7F7F 1.0pt;border-left:none;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eaverage weight\u0026nbsp;\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e(kg)\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"3\" style=\"width:43.0pt;border:solid #7F7F7F 1.0pt;border-left:none;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eaverage height\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e(m)\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"3\" style=\"width:43.0pt;border:solid #7F7F7F 1.0pt;border-left:none;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eaverage BMI\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e(kg/m2)\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"3\" style=\"width:43.0pt;border:solid #7F7F7F 1.0pt;border-left:none;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eaverage age\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e(year)\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width:85.0pt;border:solid #7F7F7F 1.0pt;border-left:none;padding:.45pt .45pt 0cm .45pt;height:14.15pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eGait speed (m/s)\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height:14.15pt;border:none;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width:43.0pt;border-top:none;border-left:none;border-bottom:solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;padding:.45pt .45pt 0cm .45pt;height:21.65pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eFirst peak\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width:43.0pt;border-top:none;border-left:none;border-bottom:solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;padding:.45pt .45pt 0cm .45pt;height:21.65pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eSecond peak\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height:21.65pt;border:none;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;padding:.45pt .45pt 0cm .45pt;height:11.1pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eKL1\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;padding:.45pt .45pt 0cm .45pt;height:11.1pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eKL2\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;padding:.45pt .45pt 0cm .45pt;height:11.1pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cstrong\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eKL3\u003c/span\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height:11.1pt;border:none;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:48.0pt;border:solid #7F7F7F 1.0pt;border-top:none;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003eprogressor\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:40.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e9\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e6\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e2\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e71.7\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 7.00\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.60\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 0.06\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e28.05\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 3.92\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e66.7\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 4.1\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.75pt .75pt 0cm .75pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.24\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style='font-size:12px;font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026plusmn;\u003c/span\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026nbsp;0.18\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#F4E7E7;padding:.75pt .75pt 0cm .75pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.15\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style='font-size:12px;font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026plusmn;\u003c/span\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026nbsp;0.19\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height:25.5pt;border:none;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:48.0pt;border:solid #7F7F7F 1.0pt;border-top:none;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003enon-progressor\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:40.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e11\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e7\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e4\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e0\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e67.1\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 14.0\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.61\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 0.03\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e25.90\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 4.77\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e64.6\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 5.6\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.75pt .75pt 0cm .75pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.19\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style='font-size:12px;font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026plusmn;\u003c/span\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026nbsp;0.21\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#DEEBF7;padding:.75pt .75pt 0cm .75pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.13\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style='font-size:12px;font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026plusmn;\u003c/span\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026nbsp;0.16\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height:25.5pt;border:none;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width:48.0pt;border:solid #7F7F7F 1.0pt;border-top:none;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003econtrol\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:40.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e10\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e-\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e-\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:34.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e-\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e64.2\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 10.3\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.61\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 0.06\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e24.71\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 3.76\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.45pt .45pt 0cm .45pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e63.0\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026plusmn; 10.2\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.75pt .75pt 0cm .75pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.20\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style='font-size:12px;font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026plusmn;\u003c/span\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026nbsp;0.24\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width:43.0pt;border-top:none;border-left:none;border-bottom: solid #7F7F7F 1.0pt;border-right:solid #7F7F7F 1.0pt;background:#E7E6E6;padding:.75pt .75pt 0cm .75pt;height:25.5pt;\"\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e1.16\u0026nbsp;\u003c/span\u003e\u003c/p\u003e\n \u003cp style='margin-top:0cm;margin-right:0cm;margin-bottom:0cm;margin-left:0cm;text-align:center;font-size:11.0pt;font-family:\"Arial\",sans-serif;line-height:normal;vertical-align:middle;'\u003e\u003cspan style='font-size:12px;font-family:\"Times New Roman\",serif;color:black;'\u003e\u0026plusmn;\u003c/span\u003e\u003cspan style=\"font-size:12px;font-family:Arial;color:black;\"\u003e\u0026nbsp;0.21\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height:25.5pt;border:none;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003ePreviously collected gait data were analyzed\u003csup\u003e31,32\u003c/sup\u003e. including marker positions tracked using a 10MX Vicon Motion capture system at 100Hz, was synchronized with ground reaction forces acquired via AMTI in-ground force plates collected at 1000Hz. Participants performed a static calibration trial\u003csup\u003e34\u003c/sup\u003e in an anatomical position for 5 seconds, followed by over-ground walking trials at a self-selected pace. Marker trajectories and force plate data were processed using custom Matlab (MATLAB R2020b, The Math Works, Inc., Natick, Massachusetts, USA) scripts for subsequent musculoskeletal modeling.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eMusculoskeletal Model to Estimate Joint Contact Mechanism\u003c/h2\u003e\n\u003cp\u003eA state-of-the-art musculoskeletal modeling framework, OpenSim Joint Articular Mechanics (JAM) (https://github.com/clnsmith/opensim-jam),\u003csup\u003e35\u003c/sup\u003e was used to estimate joint kinematics, joint moments, and contact loading parameters. First, a generic OpenSim\u003csup\u003e36\u003c/sup\u003e model\u003csup\u003e37\u003c/sup\u003e was scaled to match each participant\u0026rsquo;s anthropometry using static trial marker positions. OpenSim-JAM integrates a unique knee joint contact model with standard OpenSim tools to estimate force-dependent kinematics for knee joint secondary coordinates (i.e., tibiofemoral joint internal/external rotation, adduction/abduction and translations, and six-degrees-of-freedom patellofemoral joint) which are dynamically consistent with ligament, muscle, and articular contact forces. Dynamic contact pressure across the knee joint surface, including the center of pressure (COP), was estimated using an elastic foundation model formulation. Articular contact pressures and COP were separately calculated for the medial and lateral compartments.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOne trial was selected randomly for each participant for further modeling. Joint angles and moments in both hip and knee joints, mean contact pressure, and center of pressure were extracted at the first and second peak of total tibiofemoral joint loading identified using a semi-automated approach through a custom-written Matlab script and exported for subsequent statistical analysis. Parameters were compared between groups at the first and second peaks of tibiofemoral joint loading. Non-parametric Kruskall-Wallis tests followed by Mann-Whitney U tests were then performed to evaluate between-group differences. All statistical analyses were performed using RStudio version 2024.09.1+394 (RStudio, PBC, Boston, MA).\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eFinite Element Model to Estimate Cartilage Mechanical Response\u003c/h2\u003e\n\u003cp\u003eContact pressure estimated by MSK models was extracted for the stance phase of gait and applied to a finite element (FE) model of the medial tibial compartment, as all subjects involved in the project have medial compartment OA.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIdentical generic cartilage geometries were used in the FE and MSK models. The FE model incorporated hexahedral meshes generated using ANSA (v21.0.1, BETA CAE Systems International AG, Switzerland) with 21,076 elements (C3D8P). This model used a fibril-reinforced poroelastic material (FRPE)\u003csup\u003e38\u0026ndash;40\u003c/sup\u003e formulation, which was implemented in Abaqus (Abaqus 2021, Dassault Syst\u0026egrave;mes Simulia Corp., Providence, RI, USA) via a user-defined material subroutine (UMAT). The material properties were derived from unconfined compression tests on non-OA human cartilage\u003csup\u003e41\u003c/sup\u003e. Detailed information on the material model and properties are documented in the supplementary material \u003cstrong\u003e(Table S2)\u003c/strong\u003e. Subject-specific dynamic contact pressures estimated by MSK models were applied to the cartilage\u0026apos;s articular surface, while the cartilage\u0026apos;s bottom surface, attached to the subchondral bone, was fixed in the model.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThree critical mechanical responses\u0026mdash;compressive, fibril, and maximum shear strain \u003csup\u003e27\u003c/sup\u003e\u0026mdash;were analyzed at the first and second peaks of tibiofemoral joint loading and the maximum value for each element during the entire stance phase of gait. The compressive strain, representing the \u0026nbsp;cartilage deformation under compressive loading, is approximated by the absolute minimum principal strain. Based on the literature and in line with our previous work, fibril strain was found indicative of fibril degradation and maximum shear strain was leading to proteoglycan depletion during OA progression\u003csup\u003e16,29,30,42,43\u003c/sup\u003e. These parameters, including minimum principal strain, fibril strain and maximum shear strain were plotted using a histogram as percentage of volume in Matlab and compared between groups.\u003c/p\u003e\n\u003ch2\u003ePilot Clustering Test to Identify OA Progressors\u003c/h2\u003e\n\u003cp\u003eHistograms of contact pressures across the medial compartment estimated by MSK models and cartilage mechanical responses estimated by FE models were used in a pilot clustering analysis to identify OA progressors using a time series k-mean clustering algorithm (tslearn\u003csup\u003e44\u003c/sup\u003e). Data from each subject were input into the algorithm at the first and second peaks of joint loading, and the maximum value was observed throughout the stance phase. Two clusters were generated, and each histogram was iteratively assigned to the nearest cluster centroid based on the Euclidean distance until convergency. The correctness of clustering was checked to assess the credibility using contact pressures and cartilage mechanical responses as \u003cem\u003ein silico\u003c/em\u003e mechanical biomarkers in distinguishing progressors from non-progressors and controls.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eNo significant differences in BMI (p\u0026thinsp;=\u0026thinsp;0.2878), age (p\u0026thinsp;=\u0026thinsp;0.6707), walking speeds at the first (p\u0026thinsp;=\u0026thinsp;0.9567) and second peaks (p\u0026thinsp;=\u0026thinsp;0.9913) of tibiofemoral joint loading were observed.\u003c/p\u003e \u003cp\u003eGiven the relevance of medial compartment OA progression, the current results section will specifically focus on contact mechanics and cartilage mechanical responses in medial tibial cartilage. However, for completeness, knee and hip joint angles and moments are available in supplementary material in \u003cb\u003eFigures \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e and S2\u003c/b\u003e. In short, the progressors group showed a significantly more externally rotated knee than the control group ( p\u0026thinsp;=\u0026thinsp;0.013) at the first peak. However, no significant differences in knee adduction moment were found between groups. Furthermore, the contact mechanism in the lateral compartment is illustrated in supplementary \u003cb\u003eFigure S3\u003c/b\u003e.\u003c/p\u003e \u003cp\u003eJoint Contact Mechanics\u003c/p\u003e \u003cp\u003eProgressors present higher mean contact pressures than controls and non-progressors (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, top). The increased contact pressures observed in progressors were associated with differences in peak loading location (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003ea and b, middle and bottom). In the medial compartment, the COP was more posterior in progressors than in controls and non-progressors, especially at the first peak. Furthermore, at the second peak, a lateral shift in the COP was observed in progressors compared to non-progressors and controls (although not statistically significant).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eCartilage Mechanical Response\u003c/p\u003e \u003cp\u003eRepresentative tibial strain maps for the medial compartment, estimated by FE models, at the first and second peaks of tibiofemoral contact force are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e for each subject group. Compressive strains were calculated using minimum principal strain in cartilage. Progressors exhibited posterolateral shifts in areas with high strain values for all studied strains at the first peak compared to non-progressors and controls. This shift mirrored the displacement of the COP in the medial compartment shown in the MSK model.\u003c/p\u003e \u003cp\u003eDue to higher contact pressures, progressors showed slightly larger cartilage volume experiencing higher compressive strain, fibril strain, and maximum shear strain at the first peak compared to non-progressors and controls (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e left). Interestingly, a larger volume of cartilage in progressors is experiencing lower strains during the first peak. Similar differences between progressors and other groups are observed at the second peak for compressive and maximum shear strain (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e middle). However, the volume difference in cartilage under lower strains is notably less than at the first peak. For fibril strain at the second peak, no clear difference was observed between the three subject groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e middle).\u003c/p\u003e \u003cp\u003eWhen examining the maximum strain of each element during the stance phase (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e right), a clear frequency shift towards higher strains can be observed in all three mechanical responses of progressors compared to the other groups.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eClustering analysis to identify OA progressors\u003c/p\u003e \u003cp\u003eUnsupervised clustering results for contact pressures estimated from MSK modeling at the first and second peaks of contact forces are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003e. Detailed clustering results for each subject are listed in Supplementary \u003cb\u003eTable S3\u003c/b\u003e.\u003c/p\u003e \u003cp\u003eNo dominance of OA progressors (red) or non-progressors (blue) and controls (grey) was observed in either cluster.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eContrary to contact pressure, cartilage mechanical responses from the FE analyses showed promising results for clustering. Unsupervised clustering results for the best-performing parameters\u0026mdash;compressive strain and maximum shear strain both at the first peak of contact force and fibril strain at the first peak of contact force and maximum of stance phase\u0026mdash;are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e6\u003c/span\u003e. Clustering results for other studied strains at the two peaks of contact force and maximum during the stance phase are provided in the supplementary material (\u003cb\u003eFigure S4\u003c/b\u003e). Notably, the identified cluster 1 is dominated by OA progressors (in red), while cluster 2 is dominated by non-progressors (in blue) and control subjects (in grey). Compressive and maximum shear strains at the first peak of tibiofemoral joint contact forces achieved identical clustering performance, with Cluster 1 comprising 66.7% of progressors and 18.2% of non-progressors (false positive). Cluster 2 contained 33.3% of progressors (false negative), 81.1% of non-progressors, and all controls.\u003c/p\u003e \u003cp\u003eIn comparison, fibril strain at the first peak showed more false positives (27.3% of non-progressors in Cluster 1) but fewer false negatives (22.2% of progressors in Cluster 2). At the maximum value during the stance phase, fibril strain classified more false negatives (44.4% of progressors in Cluster 2), though some of these misclassifications did not overlap with those identified by the compressive strain and maximum shear strain at the first peak\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study is the first to identify mechanical cartilage response during gait as biomechanical biomarkers that distinguish patients with progressing medial knee OA from those with non-progressing or controls based on patient-specific gait data. Elevated contact pressure with a posterolateral shifted COP was found in the medial compartment of OA progressors compared to non-progressors and controls. This resulted in an altered cartilage mechanical response pattern characterized by both underloading and overloading strains. These \u003cem\u003ein silico\u003c/em\u003e biomechanical biomarkers were evaluated using a preliminary unsupervised clustering analysis, with the cartilage mechanical responses from FE analysis revealing distinct biomechanical characteristics in OA progressors.\u003c/p\u003e \u003cp\u003eThis longitudinal study confirms increased contact pressures and a posterolateral shift of COP (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003e) in the medial compartment were observed in patients with accelerated OA progression This increase in loading magnitude and changes in loading areas among progressors supports the hypothesis that in OA, cartilage mechanical environment is altered and make therefore disrupted the homeostasis due to an imbalance between joint loading and load-bearing capacity\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Our findings align with previous literature, which reported a more posterolateral COP in patients with established OA than controls\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. The altered loading patterns suggest that changes in the location of the loading, rather than just the magnitude, may drive OA progression in its early stages.\u003c/p\u003e \u003cp\u003eThese results confirm the utility of such advanced musculoskeletal modeling approaches, which provide joint-level estimates of cartilage loading. The finding that no significant differences were found in the knee adduction moments (KAM) of progressors compared to non progressors or controls further highlights their utility and the need to consider direct estimates of joint loading \u0026ndash; rather than analyzing joint level kinematics and moments (supplementary \u003cb\u003eFigure S1\u003c/b\u003e). This contrasts previous literature suggesting KAM may act as a mediator for medial knee OA\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e,\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Within this study and others, the KAM's ability as a direct surrogate for joint contact forces is becoming increasingly unclear.\u003c/p\u003e \u003cp\u003eTo document the disrupted cartilage mechanical environment in OA, contact mechanics estimated by MSK were further applied to FE cartilage models to estimate cartilage mechanical responses. The altered contact mechanics in OA progressors induced changes in the mechanical response of the cartilage structure. The strains (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e) shifted posterolaterally in line with the COP. At the maximum value during the stance phase, higher strains were found in progressors, and a larger volume of cartilage exceeded established degeneration thresholds defined \u003cem\u003ein vitro\u003c/em\u003e for fibril degradation (fibril strain\u0026thinsp;=\u0026thinsp;0.1\u003csup\u003e27,30\u003c/sup\u003e) and proteoglycans depletion (maximum shear strain\u0026thinsp;=\u0026thinsp;0.4\u003csup\u003e48\u003c/sup\u003e or 0.5\u003csup\u003e16\u003c/sup\u003e), indicating an elevated risk for cartilage degeneration in these patients. This aligns with previous computational studies comparing OA patients and controls\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. However, the earlier study did not distinguish OA progressors and non-progressors.\u003c/p\u003e \u003cp\u003eAt the first and second peaks of tibiofemoral (TF) contact forces, cartilage mechanical response does not always increase proportionally as contact loading increases due to cartilage's non-linear and time-dependent nature. Signature histograms of compressive and maximum shear strains revealed a dual trend in progressors, where a larger volume of elements experienced low-range and high-range strains at the first and second peaks. This dual pattern indicated OA progressors experiencing both under- and overloading at the first and second peaks, suggesting that an altered loading pattern, rather than just increased magnitude, could also discriminate OA progressors.\u003c/p\u003e \u003cp\u003eInterestingly, neither the contact mechanics nor cartilage mechanical responses differed between non-progressors and controls. Although earlier studies reported elevated knee joint loading\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e and increased cartilage volumes exceeding degeneration thresholds\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e in patients with knee OA compared to controls, these did not distinguish between progressing and non-progressing OA patients. Therefore, previously reported differences observed between OA patients and controls align primarily with progressors within the OA group.\u003c/p\u003e \u003cp\u003eUnsupervised k-means clustering was employed for histograms of contact pressures and cartilage mechanical responses to assess further their feasibility in distinguishing OA progressors from non-progressors and controls. Although more information is provided with a histogram than the mean value, contact pressure could not discriminate OA progressors (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003e). For the selected cartilage mechanical responses (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e6\u003c/span\u003e), the clustering analysis produced distinct divisions, with one cluster predominantly containing progressors and the other containing non-progressors and controls. This distinct separation supports the idea that histograms of cartilage mechanical responses could serve as \u003cem\u003ein silico\u003c/em\u003e biomarkers of OA progression. Notably, the unsupervised clustering analysis effectively captured the dual trend signature in compressive strain and maximum shear strain at the first and second peaks, along with the shift toward higher strain values at the maximum stance phase. This suggests that cartilage mechanical responses could be sensitive and reliable biomarkers for identifying fast future OA progression.\u003c/p\u003e \u003cp\u003eDespite the success of this work, which clustered OA progressors based on cartilage mechanical responses at baseline, we know that more patient-specific gait data are needed to train and validate the clustering model in the future, particularly for clinical applications. Therefore, future studies will collect larger longitudinal datasets to refine these mechanical biomarkers and improve the accuracy of classifying OA progressors.\u003c/p\u003e \u003cp\u003eIndeed, investigating more in detail individual subject results, two progressors and two non-progressors exhibited deviant mechanical responses that resembled the opposite group\u0026rsquo;s characteristics (see supplementary \u003cb\u003eTable S3\u003c/b\u003e, P01, P07, NP07 and NP08). These four subjects accounted for most of the false clustering results. Moreover, no specific characteristics were found in joint kinematics or age of those 4 outliners. However, the progressor outliners (P01 and P07) showed higher BMI compared to the group average (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), indicating high joint loading. Whereas, the non-progressor outliners (NP07 and NP08) had lower BMI compared to their group average (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Furthermore, it is essential to acknowledge that mechanical loading and joint kinematics during gait are not the only factors contributing to cartilage degeneration during OA progression. Incorporating other factors, such as systemic inflammation, genetic predisposition, and high-impact physical activities\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e,\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e, could provide a more comprehensive understanding of OA structural progression and thus improve prediction accuracy for future clinical applications.\u003c/p\u003e \u003cp\u003eWhile this study offers valuable insights, certain aspects could be refined to enhance the prediction of OA progression in future investigations. For instance, the study relied on clinical standard KL-score to assess structural OA progression. However, the assessment is based on x-ray imaging, which lacks the ability to capture detailed cartilage degeneration in the joint Future studies could incorporate Magnetic Resonance Imaging (MRI) to examine cartilage degeneration more precisely and compare the degeneration location with high-strain regions estimated by the \u003cem\u003ein silico\u003c/em\u003e model, further validating the model prediction. In terms of the modelling framework, although the cartilage geometry was scaled to the size of each patient in the models, patient-specific cartilage geometry and tibiofemoral alignment were not included in the current workflow. These anatomical variations could further alter the contact mechanism in the joint\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e, potentially changing the mechanical responses of the cartilage. To incorporate patient-specific anatomy, a recent MSK model that directly estimates joint contact mechanics, including patient-specific cartilage geometry and tibiofemoral alignment\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e, could be integrated within the current MSK-FE modeling workflow as a next step. Additionally, this study used identical cartilage material properties for both control and OA subjects. This choice seems valid as most OA subjects in this study had a KL score of 1 or 2 at baseline, indicating early or moderate OA with limited cartilage degeneration\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. However, incorporating OA state-specific material properties and initial cartilage conditions in future studies could enhance the workflow\u0026rsquo;s ability to predict \u003cem\u003ein vivo\u003c/em\u003e cartilage degeneration. Lastly, the study focused on the chronic mechanical alterations caused by gait patterns in OA progressors. Post-traumatic OA progression, which can result from ligament rupture, cartilage lesions, or meniscus tears, is not assessed. As post-traumatic OA progression is particularly relevant for younger patients, our relevance of this workflow adapted to include the effect of meniscal injuries, cartilage lesions, or changes in ligament properties needs to be analysed in future studies.\u003c/p\u003e \u003cp\u003eNotably, while high personalization of joint geometry and cartilage mechanical properties could enhance the estimation of cartilage mechanical responses, achieving this would necessitate additional imaging techniques, such as MRI or ultrasound. This would, however, increase clinical costs and substantially raise computational demands. Although such highly personalized analysis of cartilage mechanical response may eventually benefit patients-tailored rehabilitation programs, it would limit the feasibility of large-scale joint health screening in older people.\u003c/p\u003e \u003cp\u003eJoint health screening on a large scale would enable early identification of individuals at risk of fast OA progression, allowing for timely preventive interventions to slow disease progression and preserve joint function. To make such screenings viable at a population level, cost-effectiveness and prediction accuracy must be balanced. Based on patient-specific gait with generic geometries and mechanical properties without including high personalization, this proposed workflow demonstrates its sensitivity to joint loading during locomotion and ability to differentiate structural OA progressors using \u003cem\u003ein silico\u003c/em\u003e biomarkers derived from patient-specific cartilage mechanical responses. Furthermore, compared to previous MSK-FE workflows\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e,\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e,\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e, this approach significantly reduces computational cost and complexity, enhancing its potential for clinical screening of OA progression.\u003c/p\u003e \u003cp\u003eIn conclusion, using an integrated MSK-FE modeling approach, cartilage mechanical responses during gait effectively distinguished medial knee OA progressors (over two years) from non-progressors and controls. These mechanical responses could serve as sensitive \u003cem\u003ein silico\u003c/em\u003e biomarkers to identify patients at risk of accelerated OA progression, further supported by a pilot unsupervised clustering analysis. With future validation in larger cohorts, this workflow could be adapted for clinical application, enabling screening for joint health in the elderly and early identification of patients at risk of OA progression. This would allow timely access to preventive interventions such as weight management, neuromuscular exercise\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e and gait retraining\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e, helping to preserve joint function and mobility and reduce risks of mobility loss.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAcknowledgments\u003c/p\u003e\n\u003cp\u003eThis work was supported by the FWO Happy Joints project (G045320N), the FWO fundament fellowship (11PK924N), and H2020 Project OACTIVE (SC1-PM17-2017).\u003c/p\u003e\n\u003cp\u003eAuthor contributions\u003c/p\u003e\n\u003cp\u003eBAK and YZ were responsible for data processing and simulations. IJ, SAE, BAK, IM, MW and YZ contributed to the study design and workflow. SV and FL were responsible for baseline collection and longitudinal follow-up of subjects\u0026apos; gait and OA levels. IJ, SAE, BAK, and YZ were responsible for the initial draft of the manuscript and interpretation of the results. All authors contributed to the final drafting of the submitted version.\u003c/p\u003e\n\u003cp\u003eData availability statement\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe datasets generated during and/or analysed during the current study are available from the corresponding author on request.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdditional information\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSteinmetz, J. D. et al. Global, regional, and national burden of osteoarthritis, 1990\u0026ndash;2020 and projections to 2050: a systematic analysis for the Global Burden of Disease Study 2021. \u003cem\u003eLancet Rheumatol.\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e, e508\u0026ndash;e522 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHunter, D. J., Schofield, D. \u0026amp; Callander, E. The individual and socioeconomic impact of osteoarthritis. \u003cem\u003eNature Reviews Rheumatology\u003c/em\u003e vol. 10 437\u0026ndash;441 at (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nrrheum.2014.44\u003c/span\u003e\u003cspan address=\"10.1038/nrrheum.2014.44\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoeser, R. F. Aging and osteoarthritis. \u003cem\u003eCurr. Opin. Rheumatol.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 492\u0026ndash;496 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eErlichman, J., Kerbey, A. L. \u0026amp; James, W. P. T. Physical activity and its impact on health outcomes. Paper 1: The impact of physical activity on cardiovascular disease and all-cause mortality: An historical perspective. \u003cem\u003eObes. Rev.\u003c/em\u003e \u003cb\u003e3\u003c/b\u003e, 257\u0026ndash;271 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLacombe, J., Armstrong, M. E. G., Wright, F. L. \u0026amp; Foster, C. The impact of physical activity and an additional behavioural risk factor on cardiovascular disease, cancer and all-cause mortality: A systematic review. \u003cem\u003eBMC Public. Health\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTruong, A. D., Fan, E., Brower, R. G. \u0026amp; Needham, D. M. Bench-to-bedside review: mobilizing patients in the intensive care unit\u0026ndash;from pathophysiology to clinical trials. \u003cem\u003eCrit. Care\u003c/em\u003e. \u003cb\u003e13\u003c/b\u003e, 216 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLories, R. J. \u0026amp; Luyten, F. P. The bone-cartilage unit in osteoarthritis. \u003cem\u003eNat. Rev. Rheumatol.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, 43\u0026ndash;49 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConley, B. et al. Core Recommendations for Osteoarthritis Care: A Systematic Review of Clinical Practice Guidelines. \u003cem\u003eArthritis Care Res.\u003c/em\u003e \u003cb\u003e75\u003c/b\u003e, 1897\u0026ndash;1907 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarriott, K. A. \u0026amp; Birmingham, T. B. Fundamentals of osteoarthritis. Rehabilitation: Exercise, diet, biomechanics, and physical therapist-delivered interventions. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 1312\u0026ndash;1326 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEmrani, P. S. et al. Joint space narrowing and Kellgren-Lawrence progression in knee osteoarthritis: an analytic literature synthesis. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 873\u0026ndash;882 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHawker, G. A. \u0026amp; Lohmander, L. S. What an earlier recognition of osteoarthritis can do for OA prevention. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e, 1632\u0026ndash;1634 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeinstein, A. M. et al. Estimating the burden of total knee replacement in the United States. \u003cem\u003eJ. Bone Jt. Surg.\u003c/em\u003e \u003cb\u003e95\u003c/b\u003e, 385\u0026ndash;392 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBader, D. L., Salter, D. M. \u0026amp; Chowdhury, T. T. Biomechanical Influence of Cartilage Homeostasis in Health and Disease. \u003cem\u003eArthritis\u003c/em\u003e 1\u0026ndash;16 (2011). (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun, H. B. Mechanical loading, cartilage degradation, and arthritis. \u003cem\u003eAnn. N Y Acad. Sci.\u003c/em\u003e \u003cb\u003e1211\u003c/b\u003e, 37\u0026ndash;50 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVarady, N. H. \u0026amp; Grodzinsky, A. J. Osteoarthritis year in review 2015: Mechanics. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, 27\u0026ndash;35 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrozco, G. A., Tanska, P., Florea, C., Grodzinsky, A. J. \u0026amp; Korhonen, R. K. A novel mechanobiological model can predict how physiologically relevant dynamic loading causes proteoglycan loss in mechanically injured articular cartilage. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, 1\u0026ndash;16 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith, C. R., Choi, W., Negrut, K., Thelen, D. G. \u0026amp; D. \u0026amp; Efficient computation of cartilage contact pressures within dynamic simulations of movement. \u003cem\u003eComput. Methods Biomech. Biomed. Eng. Imaging Vis.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, 491\u0026ndash;498 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeireles, S. et al. Medial knee loading is altered in subjects with early osteoarthritis during gait but not during step-up-and-over task. \u003cem\u003ePLoS One\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarouane, H., Shirazi-Adl, A. \u0026amp; Adouni, M. Alterations in knee contact forces and centers in stance phase of gait: A detailed lower extremity musculoskeletal model. \u003cem\u003eJ. Biomech.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e, 185\u0026ndash;192 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumar, R. et al. Polarization second harmonic generation microscopy provides quantitative enhanced molecular specificity for tissue diagnostics. \u003cem\u003eJ. Biophotonics\u003c/em\u003e. \u003cb\u003e8\u003c/b\u003e, 730\u0026ndash;739 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAdouni, M. \u0026amp; Shirazi-Adl, A. Evaluation of knee joint muscle forces and tissue stresses-strains during gait in severe OA versus normal subjects. \u003cem\u003eJ. Orthop. Res.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 69\u0026ndash;78 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHalonen, K. S., Mononen, M. E., Jurvelin, J. S., T\u0026ouml;yr\u0026auml;s, J. \u0026amp; Korhonen, R. K. Importance of depth-wise distribution of collagen and proteoglycans in articular cartilage-A 3D finite element study of stresses and strains in human knee joint. \u003cem\u003eJ. Biomech.\u003c/em\u003e \u003cb\u003e46\u003c/b\u003e, 1184\u0026ndash;1192 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHalonen, K. S. et al. Deformation of articular cartilage during static loading of a knee joint - Experimental and finite element analysis. \u003cem\u003eJ. Biomech.\u003c/em\u003e \u003cb\u003e47\u003c/b\u003e, 2467\u0026ndash;2474 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHalonen, K. S. et al. Importance of patella, quadriceps forces, and depthwise cartilage structure on knee joint motion and cartilage response during gait. \u003cem\u003eJ. Biomech. Eng.\u003c/em\u003e \u003cb\u003e138\u003c/b\u003e, 1\u0026ndash;11 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eErdemir, A. et al. Deciphering the \u0026lsquo;Art\u0026rsquo; in Modeling and Simulation of the Knee Joint: Overall Strategy. \u003cem\u003eJ. Biomech. Eng.\u003c/em\u003e \u003cb\u003e141\u003c/b\u003e, 1\u0026ndash;10 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEsrafilian, A. et al. EMG-Assisted Muscle Force Driven Finite Element Model of the Knee Joint with Fibril-Reinforced Poroelastic Cartilages and Menisci. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 1\u0026ndash;16 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMohout, I. et al. Signatures of disease progression in knee osteoarthritis: insights from an integrated multi-scale modeling approach, a proof of concept. \u003cem\u003eFront. Bioeng. Biotechnol.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 1\u0026ndash;12 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrozco, G. A. et al. Shear strain and inflammation-induced fixed charge density loss in the knee joint cartilage following ACL injury and reconstruction: A computational study. \u003cem\u003eJ. Orthop. Res.\u003c/em\u003e \u003cb\u003e40\u003c/b\u003e, 1505\u0026ndash;1522 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEskelinen, A. S. A., Mononen, M. E., Ven\u0026auml;l\u0026auml;inen, M. S., Korhonen, R. K. \u0026amp; Tanska, P. Maximum shear strain-based algorithm can predict proteoglycan loss in damaged articular cartilage. \u003cem\u003eBiomech. Model. Mechanobiol.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e, 753\u0026ndash;778 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElahi, S. A. et al. An in silico Framework of Cartilage Degeneration That Integrates Fibril Reorientation and Degradation Along With Altered Hydration and Fixed Charge Density Loss. \u003cem\u003eFront. Bioeng. Biotechnol.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaert, I. A. C. et al. Gait characteristics and lower limb muscle strength in women with early and established knee osteoarthritis. \u003cem\u003eClin. Biomech.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 40\u0026ndash;47 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMahmoudian, A. et al. Varus thrust in women with early medial knee osteoarthritis and its relation with the external knee adduction moment. \u003cem\u003eClin. Biomech.\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 109\u0026ndash;114 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKellgren, J. H. Radiological assessment. \u003cem\u003eAnn. Rheum. Dis.\u003c/em\u003e 494\u0026ndash;502. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2307/3578513\u003c/span\u003e\u003cspan address=\"10.2307/3578513\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1957).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCappozzo, Catani, F. \u0026amp; Della Croce, U. Leardini, a. Position and orietnation in space of bones during movement. \u003cem\u003eClin. Biomech.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 171\u0026ndash;178 (1995).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith, C. R., Lenhart, R. L., Kaiser, J., Vignos, M. F. \u0026amp; Thelen, D. G. Influence of Ligament Properties on Tibiofemoral Mechanics in Walking. \u003cem\u003eJ. Knee Surg.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e, 99\u0026ndash;106 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDelp, S. L. et al. OpenSim: Open-source software to create and analyze dynamic simulations of movement. \u003cem\u003eIEEE Trans. Biomed. Eng.\u003c/em\u003e \u003cb\u003e54\u003c/b\u003e, 1940\u0026ndash;1950 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLenhart, R. L., Kaiser, J., Smith, C. R. \u0026amp; Thelen, D. G. Prediction and Validation of Load-Dependent Behavior of the Tibiofemoral and Patellofemoral Joints During Movement. \u003cem\u003eAnn. Biomed. Eng.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e, 2675\u0026ndash;2685 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEbrahimi, M. et al. Elastic, Viscoelastic and Fibril-Reinforced Poroelastic Material Properties of Healthy and Osteoarthritic Human Tibial Cartilage. \u003cem\u003eAnn. Biomed. Eng.\u003c/em\u003e \u003cb\u003e47\u003c/b\u003e, 953\u0026ndash;966 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEbrahimi, M. et al. Structure\u0026ndash;Function Relationships of Healthy and Osteoarthritic Human Tibial Cartilage: Experimental and Numerical Investigation. \u003cem\u003eAnn. Biomed. Eng.\u003c/em\u003e \u003cb\u003e48\u003c/b\u003e, 2887\u0026ndash;2900 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElahi, S. A. et al. Guide to mechanical characterization of articular cartilage and hydrogel constructs based on a systematic in silico parameter sensitivity analysis. \u003cem\u003eJ. Mech. Behav. Biomed. Mater.\u003c/em\u003e \u003cb\u003e124\u003c/b\u003e, (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCastro-Vinuelas, S. A. E. R. \u0026amp; Govaerts, A. Unconfined Compression Experimental Protocol for Cartilage Explants and Hydrogel Constructs: From Sample Preparation to Mechanical Characterization. \u003cem\u003eCartil. Tissue Engineering: Introduction\u003c/em\u003e. 271\u0026ndash;287. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/978-1-0716-2839-3_1\u003c/span\u003e\u003cspan address=\"10.1007/978-1-0716-2839-3_1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eElahi, S. A. et al. Contribution of collagen degradation and proteoglycan depletion to cartilage degeneration in primary and secondary osteoarthritis: an in silico study. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 741\u0026ndash;752 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHosseini, S. M., Wilson, W., Ito, K. \u0026amp; Van Donkelaar, C. C. A numerical model to study mechanically induced initiation and progression of damage in articular cartilage. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 95\u0026ndash;103 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTavenard, R. et al. A Machine Learning Toolkit for Time Series Data. \u003cem\u003eJ. Mach. Learn. Res.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 1\u0026ndash;6 (2020). Tslearn.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndriacchi, T. P., Smith, R. L., Medicine, S. \u0026amp; Koo, S. A Framework for the in Vivo Pathomechanics of Osteoarthritis at the Knee: 2nd Special Edition on Musculoskeletal Bioengineering. Guest Editor : Kyriacos A. A Framework for the in Vivo Pathomechanics of Osteoarthritis at the Knee. (2004). \u003cdiv class=\"ExternalRefDOI\"\u003e10.1023/B\u003c/div\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiyazaki, T. et al. Dynamic load at baseline can predict radiographic disease progression in medial compartment knee osteoarthritis. \u003cem\u003eAnn. Rheum. Dis.\u003c/em\u003e \u003cb\u003e61\u003c/b\u003e, 617\u0026ndash;622 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBennell, K. L. et al. Higher dynamic medial knee load predicts greater cartilage loss over 12 months in medial knee osteoarthritis. \u003cem\u003eAnn. Rheum. Dis.\u003c/em\u003e \u003cb\u003e70\u003c/b\u003e, 1770\u0026ndash;1774 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrozco, G. A. et al. Prediction of local fixed charge density loss in cartilage following ACL injury and reconstruction: A computational proof-of-concept study with MRI follow-up. \u003cem\u003eJ. Orthop. Res.\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 1064\u0026ndash;1081 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiukkonen, M. K. et al. Simulation of subject-specific progression of knee osteoarthritis and comparison to experimental follow-up data: Data from the osteoarthritis initiative. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e, 1\u0026ndash;14 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrozco, G. A. et al. Effect of patient specificity on predicting knee cartilage degeneration in obese adults: Musculoskeletal finite-element modeling of data from the CAROT trial. \u003cem\u003eJ. Orthop. Res.\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jor.25912\u003c/span\u003e\u003cspan address=\"10.1002/jor.25912\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumar, D., Manal, K. T. \u0026amp; Rudolph, K. S. Knee joint loading during gait in healthy controls and individuals with knee osteoarthritis. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 298\u0026ndash;305 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreene, M. A. \u0026amp; Loeser, R. F. Aging-related inflammation in osteoarthritis. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 1966\u0026ndash;1971 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartel-Pelletier, J. et al. Osteoarthritis. \u003cem\u003eNature Reviews Disease Primers\u003c/em\u003e vol. 2 at (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nrdp.2016.72\u003c/span\u003e\u003cspan address=\"10.1038/nrdp.2016.72\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWillems, M. et al. Population-based in silico modeling of anatomical shape variation of the knee and its impact on joint loading in knee osteoarthritis. \u003cem\u003eJ. Orthop. Res.\u003c/em\u003e 1\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jor.25934\u003c/span\u003e\u003cspan address=\"10.1002/jor.25934\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKillen, B. A., Willems, M. \u0026amp; Jonkers, I. P. R. E. P. R. I. N. T. An Open-source framework for the generation of OpenSim models with personalised knee joint geometries for the estimation of articular contact mechanics. \u003cem\u003eJ. Biomech.\u003c/em\u003e \u003cb\u003e177\u003c/b\u003e, 112387 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoemer, F. W. et al. Heterogeneity of cartilage damage in Kellgren and Lawrence grade 2 and 3 knees: the MOST study. \u003cem\u003eOsteoarthr. Cartil.\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 714\u0026ndash;723 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEsrafilian, A. et al. Toward Tailored Rehabilitation by Implementation of a Novel Musculoskeletal Finite Element Analysis Pipeline. \u003cem\u003eIEEE Trans. Neural Syst. Rehabil Eng.\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 789\u0026ndash;802 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMononen, M. E., Liukkonen, M. K. \u0026amp; Korhonen, R. K. Utilizing Atlas-Based Modeling to Predict Knee Joint Cartilage Degeneration: Data from the Osteoarthritis Initiative. \u003cem\u003eAnn. Biomed. Eng.\u003c/em\u003e \u003cb\u003e47\u003c/b\u003e, 813\u0026ndash;825 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoos, E. M. \u0026amp; Arden, N. K. Strategies for the prevention of knee osteoarthritis. \u003cem\u003eNat. Rev. Rheumatol.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 92\u0026ndash;101 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichards, R., van den Noort, J. C., Dekker, J. \u0026amp; Harlaar, J. Gait Retraining With Real-Time Biofeedback to Reduce Knee Adduction Moment: Systematic Review of Effects and Methods Used. \u003cem\u003eArch. Phys. Med. Rehabil\u003c/em\u003e. \u003cb\u003e98\u003c/b\u003e, 137\u0026ndash;150 (2017).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6032747/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6032747/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eOsteoarthritis (OA) is a common degenerative joint disorder affecting the whole joint, particularly characterized by articular cartilage breakdown, mostly affecting the knee\u0026rsquo;s medial compartment. Its prevalence is high with aging as an important risk factor. With a global aging population, understanding, preventing, and managing OA becomes increasingly important. Progression of structural knee OA is multifactorial, including biomechanical stressors, inflammatory responses, and genetic predispositions. Traditional attempts to identify biomarkers predicting structural OA progression focus on wet biochemical markers from blood, synovia, or urine. This study assesses \u003cem\u003ein silico\u003c/em\u003e loading-related parameters of the cartilage mechanical response as promising predictors of OA progression. A novel MSK-FE workflow relating knee movement to contact pressures and cartilage tissue response was developed. Subjects presenting OA progression over 2 years exhibited elevated medial compartment loading magnitude and posterolateral location shift at baseline. Unsupervised k-means clustering, using strain histograms, successfully differentiated progressors from non-progressors and controls when combining contact pressure and cartilage tissue mechanical responses. This study demonstrates the potential of computationally efficient, \u003cem\u003ein silico\u003c/em\u003e mechanical biomarkers to identify personalized OA progression risk after 2 years. This approach offers promising clinical benefit by identifying patients at risk of OA progression, making them eligible for preventative treatment strategies.\u003c/p\u003e","manuscriptTitle":"Cartilage mechanical responses during gait as in silico biomarkers for medial knee OA progression","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-12 19:29:51","doi":"10.21203/rs.3.rs-6032747/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-15T09:23:13+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-15T05:07:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"210094004575134450832351335560235225900","date":"2025-07-08T08:48:16+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-04T22:32:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"10524456142097310127935914698254594128","date":"2025-03-05T03:00:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"309796235533745418236162435545537617347","date":"2025-02-28T17:12:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"230055913672203574830585317488536680285","date":"2025-02-28T14:08:46+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-02-28T11:01:04+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-02-28T07:26:44+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-02-27T07:31:17+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-02-27T07:30:00+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"94253c39-5b4f-46ca-856c-9e96e46f67c0","owner":[],"postedDate":"May 12th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47725338,"name":"Health sciences/Diseases/Rheumatic diseases/Osteoarthritis"},{"id":47725339,"name":"Health sciences/Anatomy/Musculoskeletal system/Cartilage"},{"id":47725340,"name":"Physical sciences/Engineering/Biomedical engineering"}],"tags":[],"updatedAt":"2025-10-13T16:08:45+00:00","versionOfRecord":{"articleIdentity":"rs-6032747","link":"https://doi.org/10.1038/s41598-025-19371-2","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-10-09 15:57:22","publishedOnDateReadable":"October 9th, 2025"},"versionCreatedAt":"2025-05-12 19:29:51","video":"","vorDoi":"10.1038/s41598-025-19371-2","vorDoiUrl":"https://doi.org/10.1038/s41598-025-19371-2","workflowStages":[]},"version":"v1","identity":"rs-6032747","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6032747","identity":"rs-6032747","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}