Journal of Biomechanics

The relationship between external forces and bone strain in running is often complex and nonintuitive. We used Groucho running (i.e., running with exaggerated knee flexion) as a model to dramatically reduce the vertical ground reaction force (VGRF) and examined the relationship between peak VGRF and finite element (FE)-predicted tibia-fibula bone strain. Nine physically active males ran on an instrumented treadmill at 2.8 m/s with their preferred running technique, increased knee flexion (Groucho), and exaggerated knee flexion (Ex Groucho) in a randomized order. Strains at the tibia-fibula midshaft were calculated using computed-tomography-based FE modeling with loads and boundary conditions calculated from an inverse-dynamics based musculoskeletal model. Pressure-modified von Mises strain was used to quantify the peak strain (90th percentile strain) and strained volume (volume of bone experiencing strains above 3000 {micro}{varepsilon}). We further explored the relationship between peak VGRF, lower leg angle, and FE-predicted strain variables. The results showed that a 15.8% and 22.9% reduction in VGRF during Groucho and Ex Groucho, respectively, had no significant effect on FE-predicted peak strain (p > 0.304) and strained volume (p>0.053). Changes in peak VGRF did not correlate with FE-predicted strain variables (p>0.54) while changes in lower leg angle in the sagittal plane were moderately correlated (r>0.65; p<0.047). Our findings suggest that reductions in peak external forces do not always coincide with reductions in bone strain, especially in cases where running kinematics are dramatically altered. This work has important implications for designing gait retraining interventions based on reductions in external force measures.

Older adults face an increased risk of falls that may have severe consequences for their well-being. Routine, accessible clinical screening may help mitigate fall risk through early detection of balance impairments. Portable force plates offer a convenient and practical solution for balance assessment in clinical settings. A new force-plate-based balance measure, the intersection-point-height, has shown particularly promising results in its ability to distinguish between healthy and impaired balance behaviors. However, the intersection-point-height measure requires measurement of shear force during standing, which exhibits magnitudes of less than 0.2% of normal forces (body weight), taxing the dynamic range of most sensor technologies. The ability of existing force plates to measure such low-magnitude shear forces observed during quiet standing is currently unknown. This study presents a force-plate performance assessment method to evaluate shear-force measurement errors and quantify the uncertainty of the intersection-point-height measure. The method was applied to test a commonly used laboratory-grade portable force plate. While the device successfully captured sagittal-plane intersection-point-height at the lowest frequencies, low signal strength prevented precise readings in the frontal plane. Thus, the tested device only marginally met the precision required for quiet-standing analysis, underscoring the critical need for systematic performance validation of portable force plates prior to clinical use. Future efforts should focus on evaluating alternative portable force plates and exploring economical design improvements to enhance shear-force measurement precision.

ObjectiveThis study investigates knee joint biomechanics during lunges under varying tibial angles and external loads using musculoskeletal modeling and finite element analysis. The goal is to provide a biomechanical basis for understanding knee loading patterns and optimizing sports training and rehabilitation. MethodsTwenty-six healthy young men performed lunges under tibial inclination angles relative to the ground (60{degrees} and 90{degrees}) and two external load conditions (bodyweight and an additional 98 N external load). Kinematic and kinetic data were captured using motion capture and force plates. Musculoskeletal models were used to estimate joint moments, range of motion, and stiffness, with data analyzed using two-way repeated-measures ANOVA. Finite element analysis was performed at 90{degrees} tibial angle to evaluate tissue stress and displacement. ResultsThe joint moment at a 60{degrees} tibial angle was much higher than at a 90{degrees}. External load showed significant effects on knee stiffness, with lower rotational stiffness in the horizontal plane (P < 0.001) and lower coronal plane stiffness at 90{degrees} (P = 0.012) under the 98 N external-load condition, indicating reduced resistance to angular displacement in these planes. Under the 90{degrees} tibial-angle condition with external load, peak stress and displacement were concentrated in the posterior horn of the meniscus, with a maximum displacement of 3.12 mm. ConclusionThe anterior tilt of the tibia increased sagittal-plane knee loading, while external load mainly reduced joint stiffness in the coronal and horizontal planes. Under the 90{degrees} loaded condition, the concentration of stress and displacement in the posterior horn of the meniscus suggests a mechanically unfavorable loading pattern rather than direct evidence of injury risk. These findings may provide useful biomechanical information for load management during lunge-based training and rehabilitation.

Most widely used lower-limb musculoskeletal models are derived from male anatomy and adapted to female participants solely by linear scaling, which may not capture sex-specific differences in pelvic and hip geometry. We developed a population-averaged, female lower-limb musculoskeletal model, built from MRI-based models of a cohort of 25 adult women using thin-plate-spline muscle-path mapping, bilateral symmetrisation, and wrapping-surface optimisation. We hypothesised that this average model, adapted to a new individual by standard linear scaling alone, would reproduce that individual's MRI-based model's walking biomechanics more closely than a linearly scaled generic male-based model. We also expected that this advantage would be concentrated in pelvis- and hip-dependent outputs rather than distributed evenly across all joints. Using 5-fold cross-validation, the scaled average-female model and the scaled male model were each compared against the held-out individual's MRI-based model across gait kinematics, joint moments, muscle moment arms, muscle forces/activations, and joint reaction forces. The average-female model outperformed the male model in every output category (Holm-corrected p [&le;] *10-5), supporting our primary hypothesis. Consistent with our secondary hypothesis, differences were largest and most sustained for pelvis tilt, hip flexion, and gluteal/adductor moment arms and forces, and smaller for knee and ankle kinematics. Some divergence remained localised to early-stance knee kinematics and patellofemoral loading. The population-averaged female musculoskeletal model is freely available on SimTK https://simtk.org/projects/aver_fem and is recommended for studies involving female participants, particularly when pelvic and hip biomechanics are the primary outcomes.

Humans tend to walk at speeds that minimize energy expenditure and time duration. While most walking experiments examine individuals in relative isolation, everyday locomotion frequently occurs in social contexts. We investigated whether walking speed is modulated in response to a social interaction during a cooperative task. Participants completed 96 randomized walking trials where they approached and retrieved boxes varying in distance (2.5-10 m) and mass (0-6.8 kg). Boxes either rested on the ground or were handed off by an experimenter, signaling prosocial effort benefitting the participant. Approach speed was measured with inertial measurement units placed at the feet and fit to a saturating exponential function of walking distance using a nonlinear mixed-effects regression model. Based on the energy-time optimization framework, we hypothesized that participants would approach more quickly when larger boxes were held at farther distances, to reduce energy and time costs of the experimenter, despite exerting more effort themselves. Participants walked 8.5% faster (1.29 m s-1 versus 1.19 m s-1; p = 3.85 x 10-6) when the largest box was held out by the experimenter versus left on the ground. However, the manner in which the box was held had no influence on approach speeds (p = 0.31). Exploratory analyses identified modest trends between experiment responses and individual characteristics, but none reached statistical significance. The findings suggest that locomotor decisions reflect not only an individuals own energy and time costs but also the perceived costs borne by others. This study demonstrates that social context can meaningfully influence walking behavior. Summary StatementHumans walk more quickly in response to the perceived energy and time others spend helping them, even at the cost of exerting more energy themselves.

Purpose: Identifying head impacts linked to brain injury in sport remains challenging. Instrumented mouthguards quantify head-impact kinematics, and finite element (FE) modelling can transform these data into brain strain estimates, which may better reflect injury risk than kinematics alone. Here, we examined associations between mouthguard-measured kinematics, FE-derived strain, and plasma brain injury biomarker GFAP following head impacts. Methods: We analysed 41 video-verified impacts from male Australian football players, including 22 assessed for concussion (17 diagnosed) and 19 unassessed. Instrumented mouthguards recorded peak linear acceleration (PLA), peak rotational acceleration, and peak rotational velocity (PRV). Brain strain was estimated using the Imperial College FE brain model, and plasma GFAP was quantified using Simoa. Biomechanical-GFAP associations were examined using Spearman correlations and segmented regression. Results: For impacts overall, plasma GFAP was moderately correlated with PLA ({rho}=0.46, 95% CI: 0.20-0.66), PRV ({rho}=0.53, 95% CI: 0.20-0.78), and strain ({rho}=0.60, 95% CI: 0.32-0.80). Associations were stronger within concussion cases for strain ({rho}=0.86, 95% CI: 0.58-0.97) and PRV ({rho}=0.64, 95% CI: 0.15-0.93). Piecewise regression identified strain levels above which strain-GFAP relationships steepened across the whole-brain and brainstem. In concussion cases, supra-threshold brainstem strain was associated with greater symptoms. Conclusion: Finite element brain strain may better predict brain injury risk following a sport-related head impact than peak acceleration metrics. Stronger associations with plasma GFAP, particularly among concussion cases, and evidence of a biomechanical threshold, support the use of biomarker-informed strain measures in future risk modelling and the development of brain injury screening thresholds.

The cytoskeleton is a dynamic biopolymer network whose shear rheological properties are crucial for cellular physiology and pathology. However, its mechanical behavior spans multiple spatiotemporal scales, and the coupling of dynamic remodeling and viscoelastic dissipation mechanisms poses a challenge for traditional models to comprehensively capture complex cellular responses. This study aims to establish a multiscale cytoskeletal network model that integrates the bio-chemo-mechanical properties of local linked proteins, the viscoelasticity of actin filaments, and their deformation states. Developing a boundary-modified finite element method with an incremental iterative algorithm, we demonstrated the dynamic remodeling of network and the resultant rheological properties of cytoskeleton by extending the predictive time scale to one thousand seconds. The results not only reproduced the short- and intermediate-term power-law creep behavior and long-term strain plateau response of the cytoskeletal network observed in shear rheological experiments, but also indicate that the synergy among the chemo-mechanical coupling of cross-linked proteins and the bending-to-tension transition of actin filaments govern both the network remodeling and its power-law response evolutionary, whereas the steady-state properties of actin filaments determine the long-term network behavior. Simulations of cancerous and drug effects show that cancer-induced softening and reduced filament viscosity lead to accelerated cytoskeletal responses and decreased apparent shear modulus, respectively; and drug-enhanced filament prestress, along with promoting association or inhibiting dissociation of cross-linked proteins, can effectively increase the steady-state shear modulus. These findings advance the understanding of the spatiotemporal evolution and pathological mechanisms of cellular mechanical responses and provide insights for regulating polymer network performance.

The metabolic cost of walking is widely used to evaluate human performance and effectiveness of clinical interventions. Decades of laboratory research, largely based on treadmill experiments, have established a canonical relationship between walking speed and metabolic cost, and suggested that ageing shifts this relationship upward, implying reduced efficiency in older adults. However, this relationship has not been well tested during overground walking across matched speeds. We compared healthy younger (n=16; 26{+/-}2yr) and older (n=11; 74{+/-}3yr) adults across eight outdoor overground walking trials at different speeds: preferred walking speed (PWS), three fixed speeds (0.8, 1.2, 1.6 m{middle dot}s-{superscript 1}), and four speeds at {+/-}5% and {+/-}10% of PWS. Contrary to our hypothesis, older adults did not show higher gross or net metabolic cost of walking (GCOW and NCOW) than younger adults at any speed; rather, both trended consistently lower in older adults, reaching significance for GCOW at 0.8 m{middle dot}s-{superscript 1} only. Comparisons of resting metabolic rate and respiratory exchange ratio to prior reference groups did not indicate that our older cohort was unusually fit. Independent of age, GCOW was significantly higher at 0.8 m{middle dot}s-{superscript 1} than at the remaining speeds (1.2-1.6 m{middle dot}s-{superscript 1}), confirming that walking at slower speeds increases GCOW. These findings challenge the view that ageing intrinsically increases the energetic cost of walking, suggesting instead that previously reported upward shifts in cost may reflect treadmill-specific constraints or speed effects. Future work is needed to explore direct comparisons of outdoor, overground walking with treadmill walking at fixed speeds in both younger and older adults.

Ankle stiffness is decreased during movement compared to posture; however, the etiology of this decrease remains unknown. Determining what gives rise to this decrease is critical for understanding how humans successfully interact with their physical world and how that ability is compromised by functional impairments. While the triceps surae and Achilles tendon primarily dictate ankle stiffness, the relative contributions across posture and movement remain unknown. Therefore, our study sought to quantify the relative contributions of the muscle and tendon to ankle stiffness and how those contributions differ between posture and movement. We used our technique, which combines B-mode ultrasound imaging with joint-level perturbations, to quantify ankle, muscle, and tendon stiffness simultaneously. Since ankle, muscle, and tendon stiffness all scale with torque, participants matched torque between posture and movement tasks. During posture, the Achilles tendon is the dominant contributor to ankle stiffness. However, during movement, the triceps surae and Achilles tendon contribute more equally to ankle stiffness, which can be attributed to a significant decrease in muscle stiffness during movement. Here, we provide the first empirical data on how state-dependent properties of the triceps surae and Achilles tendon contribute to ankle stiffness in conditions relevant to locomotion.

Background: Newly available arm angle data offers a new dimension to understand rising rates of arm injury in MLB pitchers. Purpose: To evaluate the relationship between arm angle, pitch characteristics, and elbow and forearm injury in MLB pitchers.<br><br> Study Design: Retrospective cohort study; Level of evidence, 3 Methods: Statcast data from 2020 to 2025 and MLB injured list (IL) data were used to evaluate arm angle and pitch characteristics in relation to elbow and forearm injuries. Results are presented with and without requirements on prior season workload and for same-season and next-season injury incidence. A generalized additive model (GAM) was used to capture non-linear dependence and interactions between selected features and injury incidence to the elbow or forearm. Average marginal effect (AME) odds ratios are reported for main effect terms. Results: N = 3,812 pitcher-seasons were included. 29% pitchers who underwent UCLR did so in the same season as a forearm injury (tmean=44, tmedian=27 days to surgery). Arm angle, fastball usage, and their interaction were the three most predictive features. Arm angle was positively related to incidence of injury (ORmeanAME=1.014), fastball usage was inversely related to incidence of injury (ORmeanAME=0.243), and arm angle moderated the effect fastball usage at high arm angle, where increased usage was no longer protective. Slider velocity (ORmeanAME=1.072), spin rate (ORmeanAME=1.001), and usage (ORmeanAME=2.039) also significantly predicted injury risk. Fastball velocity was not significant in any fit, with ORmeanAME=0.999 across all fits. Fit-level Nagelkerke R2 values ranged from .019 to .052. Conclusion: Fastball usage and arm angle, not velocity, predicted elbow and forearm injury risk among MLB pitchers, and arm angle was the single most predictive feature. The heterogeneity of risk factors as a function of arm angle, and the novelty of MLB arm angle data, may explain why fastball usage has been previously underexplored as a risk factor. Keywords: baseball; arm angle; fastball velocity; fastball usage; spin rate; UCL; ulnar collateral ligament; elbow injury; forearm injury; Statcast

Adolescent idiopathic scoliosis (AIS) alters spinal geometry and may influence the biomechanical response of the spine during functional postures. However, posture-dependent changes in spinal loading and paraspinal muscle forces in AIS remain poorly understood. This study investigated the effects of trunk posture on intervertebral loading and paraspinal muscle forces using a subject-specific musculoskeletal model of an adolescent with AIS. The spinal deformity was reconstructed from biplanar radiographs and incorporated into a full-body musculoskeletal model. Flexion, extension, lateral bending, and axial rotation were simulated at three incremental magnitudes, with motion distributed across the thoracolumbar spine. Intervertebral compressive and lateral forces around the curve apex and forces in the erector spinae (ES) and multifidus (MF) muscles were evaluated. Trunk flexion produced the greatest compressive loading, reaching 337 N at the curve apex and 372 N two levels below the apex at 30{degrees} flexion. Lateral bending produced pronounced direction-dependent loading: concave-side bending increased lateral forces, whereas convex-side bending increased compressive forces. Axial rotation produced similar but smaller direction-dependent changes. Paraspinal muscle forces were consistently asymmetric, with concave-side dominance of the ES and convex-side dominance of the MF. Flexion and convex-sided movements generally produced greater muscle imbalance, while increasing posture magnitude amplified spinal loading and muscle forces. These findings demonstrate that trunk posture, movement direction, and magnitude substantially influence the biomechanical environment of the scoliotic spine and should be considered when evaluating spinal mechanics in AIS.

Acute subdural hematoma (ASDH) is a prevalent injury with high mortality and morbidity, often resulting from bridging vein (BV) disruption secondary to cortical relative motion. As a thin membrane enveloping the brain surface and anchoring BVs, the pia mater is hypothesized to play a critical mechanical role in cortical response and hence ASDH pathogenesis. Finite element (FE) head models are valuable tools to predict ASDH occurrence during impacts. However, the pia mater is often represented as an elastic material in existing FE head models, despite experimental evidence reporting its nonlinear mechanical behavior. In this study, both linear (Young's modulus of 11.5 MPa) and nonlinear (the stress-strain curve derived from pial tension tests) material models of the pia mater were implemented in one FE head model. The models were subjected to three experimental impact loadings, one of which was known to cause ASDH and two of which were not. Results demonstrated that, across all simulated impacts, the model with nonlinear pia mater properties predicted larger cortical displacements and BV responses than the linear model. For the impact with known ASDH occurrence, the predicted BV strain was 0.17 for the nonlinear model and 0.094 for the linear model, with only the former approaching the reported rupture strain range of the BV-superior sagittal sinus complex (0.29 {+/-} 0.13). These findings verified the mechanical importance of the pia mater in cortical responses and hence the prediction of ASDH, suggesting that conventional linear pia modeling might over-constrain cortical motion, leading to underestimation of BV strain and ASDH risk. The current study supported the adoption of experimentally derived nonlinear pia mater properties in FE head models to improve the reliability of ASDH prediction.

BackgroundThe use of specialized perturbation systems has become an increasingly popular approach for investigating walking stability. By accelerating or decelerating one belt, researchers can induce slip- and trip-like perturbations in a controlled laboratory setting. However, many existing studies rely on specialized perturbation systems that require expertise in device-specific software and handling of the equipment, limiting the accessibility of perturbation-based gait research to laboratories with access to such equipment. To address this limitation, we developed an open-source method capable of inducing slip- and trip-like perturbations using a standard split-belt treadmill. Here, we 1) describe the hardware and software components of the system, 2) validate the applications accuracy and precision, and 3) characterize the stability demands imparted by the perturbations with spatial stability measurements. Measured perturbation onset delay and duration were compared to the desired onset timing and programmed duration in addition to step length, step width, minimum mediolateral margin of stability, and sagittal-plane whole-body angular momentum range during the perturbed and recovery steps. ResultsFive participants with traumatic unilateral transtibial limb loss experienced perturbations consisting of brief, rapid increases or decreases in unilateral treadmill velocity, eliciting a "slip" or "trip". The mean (standard deviation) onset delay was 183.3 (9.7) ms, or 24.18% (1.91%) of stance duration. Mean perturbation duration was 239.90 (7.5) ms, 18.14% longer than the intended duration. The perturbations produced measurable changes in gait stability, such as increased step length during the perturbed step and step width during the subsequent recovery step in addition to increased minimum mediolateral margin of stability and sagittal whole body angular momentum. ConclusionThis open-source method successfully induced instability in individuals with impaired balance, demonstrating its feasibility as an accessible alternative to specialized perturbation systems. Future work will focus on refining both the software and hardware components to further improve timing, accuracy, and consistency.

Knee flexion contracture (KFC) may occur in the late stages of arthropathies, including osteoarthritis and haemophilic arthropathy. The impact of KFC on neuromuscular control remains unclear, particularly for the affected ankle plantar flexors. Surface electromyography (EMG) is widely used to assess muscle activation patterns, whereas intermuscular (EMG-EMG) coherence provides insight into common neural input. In this study we compared the neural control of ankle plantar flexors during gait between individuals with haemophilia and KFC (chronic; n = 8), and healthy individuals without (control; n = 15) and with an artificial constraint (artificial; n = 15). Bipolar EMG from plantar flexors was recorded during 30-m overground walking (1 m/s). Intermuscular coherence was estimated at 8-60 Hz during the stance phase and significance was determined using a permutation method. The chronic group showed greater knee flexion than controls (24-29 deg vs 2-20 deg), higher EMG amplitude at foot contact, and increased intermuscular coherence in the alpha (8-12 Hz) and beta (12-30 Hz) bands at mid-stance. Despite comparable sagittal knee kinematics between constrained conditions (chronic: 24-29 deg; artificial: 20-32 deg), early-stance EMG amplitude and mid-stance beta-band intermuscular coherence were higher in the chronic group across plantar-flexor pairs. Increased plantar-flexor activation in the chronic group suggests higher neural drive, while higher intermuscular coherence reflects greater common input to the plantar flexors. These findings indicate that limited ROM alone does not explain the altered neural control of plantar flexors, suggesting neural adaptations associated with non-neurological chronic KFC.

ObjectiveDescribe and compare the biomechanical severity of head acceleration events (HAEs) associated with diagnosed concussion in elite mens and womens rugby league using instrumented mouthguards (iMGs) and evaluate the diagnostic accuracy and screening performance of severity metrics within the current Head Injury Assessment (HIA) process. MethodsA prospective cohort of 398 men and 252 women from Super League teams wore iMGs across 515 matches. Following a matching and data quality screening procedure, 123 HIAs (106 men, 17 women) were matched to HAEs, 52 of which were diagnosed concussions (44 men, 8 women). High-magnitude asymptomatic control HAEs (22,676 men, 3,200 women) were sampled proportionally to the number of HIAs. A range of biomechanical severity metrics were calculated for HAEs. Statistical comparisons between outcomes were made. Receiver operating characteristic (ROC) analysis evaluated diagnostic accuracy (ability to predict diagnosed concussions within the HIA). Precision-recall analysis evaluated screening performance (ability to discriminate observable concussion signs from asymptomatic events). ResultsDiagnosed concussions had greater severity than asymptomatic HAEs across all metrics in both sexes. For diagnostic accuracy, area under the ROC curve ranged from 0.61 to 0.72 in men and 0.53 to 0.86 in women. For screening performance, optimal thresholds in several metrics provided theoretical improvements to precision over current thresholds used in rugby, but recall remained <0.20. ConclusionThese findings support integrating iMG-derived severity metrics into a multimodal, clinician-led HIA pathway as objective adjuncts for diagnosis and screening, while reinforcing that they cannot replace clinical judgement or other assessment modalities. What is already known on this topicInstrumented mouthguards are increasingly used in rugby to quantify head acceleration events and trigger Head Injury Assessment (HIA) alerts, but current screening thresholds based on simple peak kinematics have low sensitivity for identifying HAEs linked with visible concussion signs, and very few iMG-measured concussions, particularly in women, have been reported in current research. What this study addsThis study provides the largest dataset of iMG-measured concussions in any sport, shows that concussive HAEs are more severe than asymptomatic events across multiple biomechanical metrics in both sexes, and identifies several severity metrics with diagnostic accuracy comparable to existing HIA sub-tests and modest theoretical improvements over current screening thresholds. How this study might affect research, practice or policyThese findings support incorporating iMG-derived severity metrics as objective adjuncts within clinician-led HIA pathways, highlight the need for sex-inclusive iMG datasets and multimodal concussion identification, and may inform future refinement of iMG screening thresholds in elite rugby and other contact sports.

We present three quantitative methods: 1) estimation of exoskeleton mechanical power and energy ratio from published data, 2) a systematic review of the exoskeleton literature on reported energy ratios, and 3) timing correction analysis of the replication experiment, to address concerns raised by Collins et al. (2026) about Luo et al. (2024). Together, these analyses support the reported metabolic reductions and the validity of exoskeleton control via learning in simulation. The critique rests on an unsupported premise: that exoskeleton energy ratios above 4 are physiologically implausible. This premise of Collins et al. (2026) is not supported by the cited evidence, and the error originates in their own cited source. Sawicki and Ferris (2009), the paper they invoke as authority for the limit of 4, state explicitly that "reported values of the muscular efficiency range from 0.10 to 0.34, with many sources assuming an average of [~]0.25." The value of 4 corresponds to this average, it is not a physiological ceiling. Treating an average as a physiological upper limit is a fundamental error. The published exoskeleton literature further contradicts the claim, including work by the authors of the critique themselves (Collins et al., 2015: 4.3; Young et al., 2017: 5.0) and independent work (Malcolm et al., 2013: 4.8; Seo et al., 2017: 6.7). In contrast, our walking energy ratio is 2.4, calculated directly from Fig. 4 of our paper. Our device delivers higher peak torque (14.1 Nm vs. 10.9 Nm, Lim et al., 2019) and achieves a slightly larger metabolic reduction (24.3% vs. 21%). Independent groups have since demonstrated meaningful metabolic reductions using learning-in-simulation frameworks, including Barati et al. (2026, 15.2% mean and 22.5% maximum) and Zhou et al. (2025, [~]20% during running). The claim of Collins et al. (2026) that this problem "remains unsolved" is directly contradicted by these independent results. The experiment in the critique is not a valid replication of our method. Our controller is a neural network with [~]10,000 parameters learned through deep reinforcement learning in musculoskeletal simulation; the critique instead applies a pre-programmed fixed torque curve with no learnable parameters. Beyond this, the replication contains three methodological errors: 1) a heel-strike timing assumption producing offsets up to 30% of the gait cycle; 2) an averaged torque profile that discards subject-specific control; and 3) a device [~]50% heavier than ours (4.8 kg vs. 3.2 kg) without measuring the metabolic penalty of the added weight. The critique also misreports Samsung data, with reported values approximately double those in the original publication, errors that directly underpin their physiological limit argument.

This study examined how tempo, dynamics, and string influence upper-extremity physical exposure in professional violinists and how exposure variability is distributed among musical characteristics, between-subject differences, and residual variability. Twelve violinists performed seven standardized scales while bilateral upper-arm and wrist kinematics and shoulder and forearm muscle activity were recorded. Linear mixed-effects models showed that faster tempo increased right upper-arm velocity and bilateral forearm activity while reducing right upper-arm and wrist ranges of motion. Louder dynamics increased bilateral forearm and right trapezius activity and right-wrist ranges of motion. Higher-posture strings increased right upper-arm elevation and right shoulder muscle activity. Variance analysis identified exposures predominantly related to musical characteristics, jointly related to musical characteristics and between-subject differences, predominantly related to between-subject differences, or mainly unexplained. These findings support future exposure prediction from musical characteristics and targeted prevention through repertoire-based workload management, structured recovery, and individualized technique-focused strategies.

Sleep deprivation is associated with impaired endurance performance and an increased risk of running-related injury. Previous research has identified alterations in running biomechanics following a single night of sleep deprivation under laboratory conditions. However, whether these biomechanical changes can be detected using wearable technology remains unknown. Twenty-one recreationally active runners completed submaximal treadmill running under both normal sleep and total sleep deprivation conditions in a randomized crossover design. Biomechanical features were extracted simultaneously using a full-body motion capture system and a trunk-mounted wearable sensor. Five machine learning classifiers were evaluated in two classification tasks: a within-subject task using paired recordings from the same individual, and a between-subject task performed without individual baseline data. Within-subject classification consistently exceeded chance level for both measurement systems, with best accuracies of 85% for the wearable sensor (Logistic Regression) and 83% for the motion capture system (Random Forest). These findings indicate that sleep deprivation produces a systematic and individually consistent biomechanical signature during running. In contrast, between-subject classification failed across nearly all models and systems, with accuracies remaining close to chance level ([~]50%), demonstrating that inter-individual variability obscures the sleep-deprivation signal in the absence of personalized baseline data. Both systems converged on temporal organization, loading-related variables, and stride-to-stride variability as the most discriminative feature domains. Contrary to expectations, the laboratory motion capture system did not outperform the wearable sensor. Together, these findings demonstrate that individualized, baseline-referenced monitoring is essential for detecting sleep-deprivation-related changes in running gait, and suggest that a single trunk-mounted wearable sensor may provide a practical solution for real-world monitoring when paired recordings are available.

Abstract Objective HYROX is a rapidly growing hybrid fitness competition combining running with functional exercise stations. Our objective was to describe the 12-month prevalence, characteristics and severity of self-reported HYROX-related injuries. Methods We conducted an international cross-sectional online survey of 418 HYROX athletes. The primary outcome was the self-reported 12-month period prevalence of at least one HYROX-related injury; secondary outcomes included an exposure-adjusted lower-bound rate per 1000 hours of total training exposure and the profile and severity of the most significant injury. Associated factors were examined by multivariable logistic regression. Results Overall, 208 of 418 participants (49.8%, 95% CI 45.0 to 54.5) reported at least one HYROX-related injury. The exposure-adjusted lower-bound rate was 1.65 reported injuries per 1000 hours of total training exposure. Injuries mainly affected the lower extremity, most commonly the knee (20.8%); tendon-related complaints were the leading type (41.6%) and most were of gradual onset. Among participants with severity data, 20.3% reported more than 28 days of training interruption or no return to their previous performance level. Higher HYROX-specific training frequency was the only factor independently associated with injury reporting (adjusted OR 1.61, 95% CI 1.20 to 2.16; p = 0.001). Conclusion Approximately half of respondents reported at least one HYROX-related injury during the preceding 12 months, predominantly involving gradual-onset lower-extremity complaints. Higher HYROX-specific training frequency was associated with injury reporting, although the cross-sectional design precludes causal interpretation. Prospective, exposure-based surveillance is needed to quantify HYROX-specific injury incidence and burden and examine whether training frequency, load distribution and recovery contribute to injury risk.

Objective: To identify brain injury metrics suitable for supporting sports head injury assessment by evaluating their association with brain tissue strain and consistency across sports. Methods: Head kinematics from 3,139 impacts in boxing, mixed martial arts, and rugby matches were recorded using instrumented mouthguards and used to calculate nine brain injury metrics. Impacts were simulated using an anatomically detailed finite element brain model to estimate peak 95th-percentile maximum principal strain (MPS) in the brain and brainstem, a measure of tissue deformation associated with long-term pathology. Sport-specific ordinary least squares models estimated xE, the metric value equivalent to a reference MPS of 0.21. Metric-MPS correlations and xE uncertainties were quantified using 5000 bootstrap resamples. Cross-sport consistency was assessed using the coefficient of variation (CV) of sport-specific median xE values, and uncertainty using the normalised confidence interval size (NCIS). Results: XGB, an extreme gradient boosting strain-prediction model, showed the strongest and most consistent correlations with whole-brain (r=0.924-0.974) and brainstem MPS (r=0.887-0.954) across all sports. PRV, BrIC and UBrIC also correlated strongly with whole-brain (r=0.724-0.930) and brainstem MPS (r=0.739-0.900), whereas HIC15 and HARM showed weaker correlation with MPS, particularly in rugby. XGB showed the lowest cross-sport variability (CV=0.034) and uncertainty (median NCIS=0.056). HARM, DAMAGE and HIC15 showed the greatest sport dependence (CV=0.575-0.588) and uncertainty (median NCIS=0.331-0.791). Conclusions: XGB, BrIC, and UBrIC demonstrated the strongest associations with brain tissue strain and the greatest consistency across sports. This study provides a biomechanically informed framework for selecting suitable metrics for sports HIA protocols.