3D-DXA-based Finite Element Models Improve Hip Fracture Prediction versus aBMD in the Japanese Population-Based Osteoporosis (JPOS) Cohort Study

Summary Hip fractures are among the most serious consequences of osteoporosis, yet many individuals who sustain hip fractures do not meet the diagnostic threshold based on areal bone mineral density (aBMD). Finite element (FE)-based mechanical descriptors may improve hip fracture prediction beyond aBMD. Purpose This study investigated whether FE-based mechanical descriptors derived from dual-energy X-ray absorptiometry (DXA) improve hip fracture discrimination beyond aBMD in the Japanese Population-Based Osteoporosis (JPOS) cohort. Methods The study included 1,636 women, of whom 85 sustained hip fractures during 20 years of follow-up. Subject-specific three-dimensional femoral models were reconstructed from DXA images using statistical shape and appearance modelling, and nonlinear FE analyses were performed to derive mechanical descriptors under simulated sideways-fall loading. Predictive performance was evaluated using discrimination, model fit, reclassification, and hazard ratio analyses. Results Over two-thirds of women who experienced hip fracture were not classified as osteoporotic based on aBMD. Several FE-based descriptors significantly improved fracture prediction relative to total hip and femoral neck aBMD across the evaluation frameworks. The best-performing descriptor, combining femoral strength with residual energy (F_U + W_R), achieved a significantly higher AUROC of 0.797 compared with 0.763 (p = 0.026) and 0.757 (p = 0.040) for total hip and femoral neck aBMD, respectively, and the highest hazard ratio (3.577 per SD). The same descriptors consistently performed best across all evaluation methods. Conclusion These findings demonstrate that FE-based descriptors provide meaningful information beyond aBMD and may improve identification of women at elevated hip fracture risk.

Authors

Institutions

Publication Details

Journal
medRxiv
Published
2026-11-18
DOI
https://doi.org/10.64898/2026.11.16.26368177
Primary Topic
Bone health and osteoporosis research
Type
preprint

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

3D-DXA-based Finite Element Models Improve Hip Fracture Prediction versus aBMD in the Japanese Population-Based Osteoporosis (JPOS) Cohort Study

Kuniyasu Kamiya, Jérôme Noailly, Masayuki Iki, 菜美 今井 et al.
2 links
medRxiv
Bone health and osteoporosis research
preprint

3D-DXA-based Finite Element Models Improve Hip Fracture Prediction versus aBMD in the Japanese Population-Based Osteoporosis (JPOS) Cohort Study

Kuniyasu Kamiya, Jérôme Noailly, Masayuki Iki, 菜美 今井, Akemi Morita, Sadanobu Kagamimori, Takahiro Tachiki, Jun Kitagawa, Katsuyasu Kouda, Abdollah Moslemi, Elham Alizadeh, Asako Kudo, Kazuhiro Uenishi, Junko Tamaki, Ludovic Humbert, Etsuko Kajita, Kouji Tsuda, Yuho Sato, Kumiko Ohara, Akemi Nitta, Masami Hamada, Naoyuki Takashima
preprint en

Abstract

Summary Hip fractures are among the most serious consequences of osteoporosis, yet many individuals who sustain hip fractures do not meet the diagnostic threshold based on areal bone mineral density (aBMD). Finite element (FE)-based mechanical descriptors may improve hip fracture prediction beyond aBMD. Purpose This study investigated whether FE-based mechanical descriptors derived from dual-energy X-ray absorptiometry (DXA) improve hip fracture discrimination beyond aBMD in the Japanese Population-Based Osteoporosis (JPOS) cohort. Methods The study included 1,636 women, of whom 85 sustained hip fractures during 20 years of follow-up. Subject-specific three-dimensional femoral models were reconstructed from DXA images using statistical shape and appearance modelling, and nonlinear FE analyses were performed to derive mechanical descriptors under simulated sideways-fall loading. Predictive performance was evaluated using discrimination, model fit, reclassification, and hazard ratio analyses. Results Over two-thirds of women who experienced hip fracture were not classified as osteoporotic based on aBMD. Several FE-based descriptors significantly improved fracture prediction relative to total hip and femoral neck aBMD across the evaluation frameworks. The best-performing descriptor, combining femoral strength with residual energy (F_U + W_R), achieved a significantly higher AUROC of 0.797 compared with 0.763 (p = 0.026) and 0.757 (p = 0.040) for total hip and femoral neck aBMD, respectively, and the highest hazard ratio (3.577 per SD). The same descriptors consistently performed best across all evaluation methods. Conclusion These findings demonstrate that FE-based descriptors provide meaningful information beyond aBMD and may improve identification of women at elevated hip fracture risk.

medRxiv
Kansai Medical University (JP), Kobe City College of Nursing (JP), Nihon Pharmaceutical University (JP), Universitat Pompeu Fabra (ES), Kagawa Nutrition University (JP), University of Kochi (JP), Osaka Medical and Pharmaceutical University (JP), Jin-ai University (JP), Kyoto Prefectural University of Medicine (JP), Nutrition 21 (United States) (US), University of Toyama (JP), Tottori University (JP), Nagoya University (JP), Josai International University (JP), Kitasato University (JP), Kōchi University (JP)
Japan Dairy Association, Univers Foundation, University of Tokushima, Agència de Gestió d'Ajuts Universitaris i de Recerca
Good health and well-being
Bone health and osteoporosis research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon