Instrumented mouthguards (iMGs) enable in-game measurement of head acceleration events (HAEs) in contact sports, forming the basis for brain injury criteria (BICs). However, most commercial systems assume a 50th-percentile male head model for center-of-mass (COMh) transformations, which may bias risk estimates for athletes of different sex and body size. To quantify how anthropometric scaling assumptions influence BICs derived from iMG data, we analyzed 15,237 video-verified HAEs from 572 community rugby players (ages 10–38 years; 34–142 kg). Three scaling approaches were compared: (i) a one-size-fits-all baseline reference (Hybrid III 50th-percentile male), (ii) a sex- and weight-adjusted regression-based model, and (iii) a multiple-fixed-reference model incorporating three discrete headform sizes based on sex and body mass. Outcome metrics included peak linear acceleration, head impact criterion (HIC & PRHIC), head impact power (HIP lin/rot ) and kinetic energy (KE lin/rot ) computed using rigid-body transformations and optimized algorithms. Generalized additive models evaluated scaling effects across weight strata. Scaling assumptions significantly altered BIC magnitudes, particularly at size extremes. Female players modeled with male parameters exhibited up to 54% higher PRHIC and 18–27% higher HIP and kinetic energy values ( p < 0.001). For lighter players (<55 kg), the multiple-fixed-reference model reduced predicted rotational power by more than 60% relative to the baseline reference. Misclassification near operational thresholds (e.g., PLA ≥ 65 g) occurred in both sexes, with bias direction dependent on head-size assumptions. Anthropometric scaling exerts a substantial influence on iMG-derived BICs and threshold-based classifications of head-impact severity. Incorporating sex- and size-appropriate scaling, or pragmatic multiple-fixed-reference models, can reduce systematic bias and improve interpretability, and equity of head-impact surveillance in contact sports.
Building similarity graph...
Analyzing shared references across papers
Loading...
Melanie D. Bussey
University of Otago
Joshua P. McGeown
Auckland Institute of Studies
Sergio Dempsey
University of Auckland
Journal of Biomechanics
University of Auckland
University of Otago
Auckland University of Technology
Building similarity graph...
Analyzing shared references across papers
Loading...
Bussey et al. (Fri,) studied this question.
synapsesocial.com/papers/6a0d4e9df03e14405aa99e1f — DOI: https://doi.org/10.1016/j.jbiomech.2026.113378