Structural factors associated with femoral neck fractures and its prediction in Chinese males

Document Type

Journal Article

Publication Title

Journal of Clinical Densitometry

Volume

24

Issue

4

First Page

597

Last Page

602

PubMed ID

33500164

Publisher

Elsevier

School

School of Medical and Health Sciences

RAS ID

35526

Comments

Khoo, B. C. C., Wang, L., Lewis, J. R., Brown, K., Cheng, X., & Prince, R. L. (2021). Structural factors associated with femoral neck fractures and its prediction in Chinese males. Journal of Clinical Densitometry, 24(4), 597-602. https://doi.org/10.1016/j.jocd.2021.01.002

Abstract

We have recently proposed a new approach to evaluate 2D femoral neck (FN) structure, named the Minimal Model (MM), that comprised FN areal bone mineral density (FNaBMD) and FNWidth and 2 new internal structural measures; (1) the standard deviation of normalized mineral mass projection profile distribution (FNSigma), and (2) the displacement between center-of-mineral mass and geometric center of mineral mass projection profile (FNDelta). The contralateral hip of 67 FN fracture Chinese male patients had a QCT scan shortly after fracture and was compared to 156 community participants without hip fracture. The QCT scans were analyzed using Mindways software to enable DXA-equivalent 2D images to be obtained; MM variables were calculated from these images. In FN fracture and nonfracture participants, the 4 MM variables as well as age, weight and height were compared. Compared to nonfracture, fracture participants were older, weighed less and were taller. After adjustment for these differences FN fracture participants compared to nonfracture had mean ± SD lower FNaBMD 0.54 ± 0.11 vs 0.70 ± 0.11 g/cm2 (p < 0.001); larger FNSigma 1.05 ± 0.11 vs 0.98 ± 0.10 cm (p < 0.001); larger FNDelta 0.43 ± 0.09 vs 0.33 ± 0.09 cm (p < 0.001), however FNWidth did not differ 2.96 ± 0.35 vs 2.92 ± 0.34 cm. All variables except FNaBMD and FNWidth were correlated; however logistic regression identified increased age and height, reduced FNaBMD and increased FNSigma as independent contributors to differentiating participants with FN fracture from nonfracture. Area under ROC analysis identified significant improvement in discrimination with addition of FNSigma to the base model of Age and FNaBMD (C statistic 0.88 and 0.91, p = 0.019). These analyses identified important internal structural information available from 2D DXA imaging that contributes to discrimination of FN fracture in addition to low bone mass. This analytical approach may contribute to improved clinical FN fracture prediction, extending value of widely available DXA technology.

DOI

10.1016/j.jocd.2021.01.002

Access Rights

subscription content

Share

 
COinS