Routine CT-based hounsfield unit measurements as predictors of osteoporosis : A comparative analysis with DXA and development of a nomogram for clinical application
- 1Department of Neurosurgery, Kanuni Sultan Süleyman Training and Research Hospital, Istanbul, Türkiye.
Abstract
Osteoporosis, a prevalent metabolic bone disease, is traditionally diagnosed using dual-energy X-ray absorptiometry (DXA). However, DXA accuracy can be affected by factors like degenerative changes and positioning errors. Recently, Hounsfield Unit (HU) measurements from routine computed tomography (CT) scans have emerged as a potential alternative for assessing bone mineral density. This study aims to evaluate the relationship between axial and sagittal CT HU measurements at lumbar levels (L1–L4) and DXA T-scores, and to develop a nomogram for individualized osteoporosis risk prediction. This retrospective study included 148 patients (120 females, 28 males; mean age 62.7 ± 11.2 years) who underwent DXA and non-contrast lumbar CT within 1 month. Axial and sagittal HU measurements were obtained at L1–L4. Analyses included correlation, between-group comparisons, multivariable linear and logistic regression, receiver operating characteristic analysis, interobserver Intraclass correlation coefficient, and bootstrap internal validation of a nomogram. L1–L4 sagittal mean HU correlated with DXA T-scores (Spearman ρ = 0.532, P < .001), compared with axial mean HU (ρ = 0.493, P < .001). The optimal sagittal HU cutoff for osteoporosis was ≤ 102.5 HU (sensitivity 81.6%, specificity 60.9%). The nomogram incorporating age, sex, body mass index, and sagittal HU demonstrated an apparent area under the curve of 0.868 and bootstrap optimism-corrected area under the curve of 0.853. The inverse adjusted association with age appeared only after adjustment for HU and should not be interpreted as a biological protective effect. Lumbar vertebral HU measurements, particularly sagittal HU values, may complement DXA for opportunistic osteoporosis risk assessment. The nomogram offers an individualized probability estimate, although external validation is required before routine clinical implementation.
Related
- Reliability, Readability, and Safety of Generative AI Chatbots for Osteoporosis Health Counseling: A Multidimensional Cross‐Sectional Comparative Study
- Epimedium koreanum Nakai attenuates glucocorticoid-induced osteoporosis by suppressing ferroptosis via the KEAP1/NRF2/SLC7A11/GPX4 axis: comparative pharmacodynamic evaluation of four epimedium species and proteomic and metabolomic analyses of epk
- Application of syringe-based device for pressure pain threshold measurements under autoalgometry settings
- Hounsfield Units–Anchored Trabecular Bone Attenuation on Low-Dose Chest CT Scanning: Opportunities and Gaps
- Application of a Quantitative POCT-FIT Based on Near-Infrared Fluorescence Immunochromatography in Population-Based Colorectal Cancer Screening: A Comparative Evaluation with a Centralized Testing System
BibTeX
@article{tekin2026routine,
title = {{Routine CT-based hounsfield unit measurements as predictors of osteoporosis : A comparative analysis with DXA and development of a nomogram for clinical application}},
author = {Abdurrahim Tekin and Engin Can and Evren Sönmez and Lokman Ayhan and Suna Dilbaz and Akın Öztürk and Harun Duran and Serdar Çevik and Nuri Serdar Baş},
journal = {Medicine},
year = {2026},
month = oct,
doi = {10.1097/md.0000000000051422},
url = {https://europepmc.org/article/PMC/PMC13727199},
note = {CC BY-NC},
}
RIS
TY - JOUR
TI - Routine CT-based hounsfield unit measurements as predictors of osteoporosis : A comparative analysis with DXA and development of a nomogram for clinical application
AU - Abdurrahim Tekin
AU - Engin Can
AU - Evren Sönmez
AU - Lokman Ayhan
AU - Suna Dilbaz
AU - Akın Öztürk
AU - Harun Duran
AU - Serdar Çevik
AU - Nuri Serdar Baş
PY - 2026
DA - 2026/10/30
JO - Medicine
DO - 10.1097/md.0000000000051422
UR - https://europepmc.org/article/PMC/PMC13727199
AB - Osteoporosis, a prevalent metabolic bone disease, is traditionally diagnosed using dual-energy X-ray absorptiometry (DXA). However, DXA accuracy can be affected by factors like degenerative changes and positioning errors. Recently, Hounsfield Unit (HU) measurements from routine computed tomography (CT) scans have emerged as a potential alternative for assessing bone mineral density. This study aims to evaluate the relationship between axial and sagittal CT HU measurements at lumbar levels (L1–L4) and DXA T-scores, and to develop a nomogram for individualized osteoporosis risk prediction. This retrospective study included 148 patients (120 females, 28 males; mean age 62.7 ± 11.2 years) who underwent DXA and non-contrast lumbar CT within 1 month. Axial and sagittal HU measurements were obtained at L1–L4. Analyses included correlation, between-group comparisons, multivariable linear and logistic regression, receiver operating characteristic analysis, interobserver Intraclass correlation coefficient, and bootstrap internal validation of a nomogram. L1–L4 sagittal mean HU correlated with DXA T-scores (Spearman ρ = 0.532, P < .001), compared with axial mean HU (ρ = 0.493, P < .001). The optimal sagittal HU cutoff for osteoporosis was ≤ 102.5 HU (sensitivity 81.6%, specificity 60.9%). The nomogram incorporating age, sex, body mass index, and sagittal HU demonstrated an apparent area under the curve of 0.868 and bootstrap optimism-corrected area under the curve of 0.853. The inverse adjusted association with age appeared only after adjustment for HU and should not be interpreted as a biological protective effect. Lumbar vertebral HU measurements, particularly sagittal HU values, may complement DXA for opportunistic osteoporosis risk assessment. The nomogram offers an individualized probability estimate, although external validation is required before routine clinical implementation.
SN - 0304-5412
SN - 1536-5964
SN - 0025-7974
ER -