Elsevier

Bone

Volume 23, Issue 5, November 1998, Pages 417-424
Bone

Original Articles
Prediction of vertebral failure loads from spinal and femoral dual-energy x-ray absorptiometry, and calcaneal ultrasound: an in situ analysis with intact soft tissues

https://doi.org/10.1016/S8756-3282(98)00127-6Get rights and content

Abstract

The objective of the current study was to determine the correlation of spinal and femoral dual-energy X-ray absorptiometry (DXA) and calcaneal ultrasound, measured in situ with intact soft tissues, with the in vitro failure loads of lumbar vertebral bodies. Forty-nine cadavers with intact skin and soft tissues (32 men aged 82.1 ± 9.0 years, 17 women aged 83.1 ± 10.1 years) were examined. The bone mineral content (BMC), the projectional area, and the bone mineral density (BMD) of the lumbar spine and proximal femur were determined with DXA, and the ultrasonic properties of the calcaneus with quantitative calcaneal ultrasound. The fourth lumbar vertebra was then excised with adjacent intervertebral disks and its mechanical failure load determined, using a materials testing machine. Absolute fracture loads were significantly higher in men than in women, but they were similar after adjusting for body weight and height. Spinal DXA was significantly associated with vertebral failure load (r = 0.62 combined; r = 0.54 men; r = 0.58 women). Femoral DXA (r = 0.46) and calcaneal ultrasound (r = 0.48) showed somewhat lower correlation coefficients, with the speed of sound (SOS) being able to add predictive information in a stepwise regression model. Normalizing the vertebral failure loads to body weight and height reduced the correlations, with only spinal DXA yielding a significant relationship. Our data suggest that previous in vitro studies may have overestimated the association between spinal DXA and vertebral failure loads, presumably because measurements were performed on excised bones, but not in situ in the presence of soft tissue inhomogeneity. The results indicate that, even in a population of old age and under in situ conditions, spinal DXA may still be somewhat better than femoral DXA and calcaneal ultrasound in predicting vertebral failure loads.

Introduction

Vertebral fractures represent the classic hallmark of osteoporosis2, 10 with about 300,000–500,000 men and women being affected each year in the United States.10, 27 In cases of moderate or minimal trauma, the age-adjusted fracture rate of the vertebral body is generally higher in women, but above the age of 85 years the fracture rate is almost identical in both genders.10 In a recent study of clinically diagnosed vertebral fractures, 14% were found to be associated with severe trauma, 25% with a fall, and 9% with lifting; 33% had occurred spontaneously, and 16% were diagnosed incidentally.10 However, it has been estimated that only about one third of all patients with vertebral deformity seek medical attention; thus, only about 10%–15% of all vertebral fractures in the community are attributable to falls, whereas the vast majority appear to occur during normal everyday activity.10

Between 55% and 95% of all spinal fractures after the age of 45 are attributable to osteoporosis, depending on age, gender, and race.28 The direct health care expenditure involved in these fractures has been estimated to be $746 million in the United States in 1995.35 More importantly, spinal fractures lead to a substantial reduction in the quality of life due to back pain, impairment of well-being and mood, functional physical limitations, and disability.25 Because treatment of osteoporosis and for preventing osteoporotic fractures is available, noninvasive measurement techniques for the in vivo assessment of bone quantity and quality are required to allow for intervention at appropriate stages. However, it is difficult to determine the effectiveness of these techniques in predicting spinal fracture risk clinically, because many vertebral fractures pass unnoticed, because radiographs are required, and because the differentiation between vertebral fracture and deformity remains controversial. On the other hand, mechanical failure loads of skeletal regions can be determined in vitro under controlled loading conditions. Experiments of this nature can guide and focus clinical studies by selecting the most promising parameters and measurement techniques; however, the conditions under which the densitometric measurements are obtained should be similar to those in vivo.

Examining the relationship between the compressive strength of cadaveric lumbar vertebrae and quantitative computed tomography (qCT), McBroom et al.26 and Mosekilde et al.30 reported correlation coefficients between 0.68 and 0.76. When multiplying the bone density by the cross-sectional vertebral area, the coefficients were found to increase to between 0.80 and 0.86.4, 6 Cody et al.9 examined bone density in 18 regional subsets and found maximal correlations with failure load of 0.49 (L4–5) to 0.85 (T-11–L-1). Using multiple regression analyses, coefficients of >0.94 could be found. However, qCT has the disadvantage that it is relatively expensive and that the access is limited.

Dual-photon and dual-energy X-ray absorptiometry (DPA/DXA), on the other hand, are considerably less expensive and are more readily available than CT, and involve a relatively low radiation dose. Hansson et al.21 reported a correlation of 0.86 between bone mineral content (BMC), measured in excised lumbar vertebrae in the anteroposterior (AP) direction, and the load to fracture. Eriksson et al.14 compared qCT and DPA in the same specimens, and showed that the DPA-derived BMC was the better predictor (r = 0.80) than CT density multiplied by surface area (r = 0.74). The ultimate failure stress (load/cross-sectional area), however, was better predicted by qCT (r = 0.71 vs. r = 0.64). Ortoft et al.32 and Myers et al.31 indicated that lateral DXA measurements of excised lumbar vertebral bodies yielded a better association with ultimate failure loads than those taken in the AP direction. These findings were extended by Moro et al.29 who showed that the lateral BMD of L-2 correlated better with the failure load of lumbar (r = 0.89) and thoracic vertebrae (r = 0.94) than the AP BMD (r = 0.72 and 0.83).

In contrast to those studies, Bjarnason et al.5 reported much weaker correlations between the BMD and the in vitro failure loads of chemically dried lumbar vertebrae in 14 cadavers. They found that scans obtained in the lateral decubitus position gave a better correlation than those in the AP direction when measured in excised specimens (r = 0.71 vs. 0.51), but not when performed in situ in intact cadavers (r = 0.45 vs. 0.48). They observed, however, a correlation of r = 0.64 with in situ femoral DXA measurements and concluded that this parameter is more useful for assessing vertebral fracure risk than spinal measurements in any projection.

Ultrasound has more recently been advocated for assessment of fracture risk, its advantage being the low cost involved and the lack of ionizing radiation.17, 20 In a recent study, Cheng et al.7 related ultrasound measurements at the calcaneus to vertebral failure loads in 62 cadavers, and compared the correlation with that obtained for DXA and qCT measurements in excised specimens. They found that the fracture load could be best predicted with DXA-determined BMD (r = 0.80) and qCT (r = 0.78 for trabecular density × cross-sectional area), whereas calcaneal ultrasound (r = 0.37–0.41) was only a poor predictor. The investigators did, however, indicate that these findings contradicted emerging clinical data,3, 34, 36 which suggested that ultrasound has the same predictive ability as DXA. They reasoned that in vitro assessment of bone mineral density with DXA may be inherently more accurate than that performed in vivo, because of improved positioning and the absence of soft tissue variation. It is well known that the variable amount and inhomogeneous distribution of soft tissues and extraosseous calcifications may influence DXA measurements of bone mineral density,1, 15, 16, 18, 19, 37 and this has been a potential shortcoming of previous in vitro investigations.

The objective of the present analysis was therefore to determine and compare the correlation between spinal DXA, femoral DXA, and calcaneal ultrasound—when measured in situ with intact skin and soft tissues—with the in vitro failure loads of lumbar vertebral bodies in a relatively large study sample.

Section snippets

Materials and methods

Forty-nine formalin-fixed cadavers (from a macroscopic anatomy course) with intact skin and soft tissues were examined. Thirty-two cadavers were men, aged 82.1 ± 9.0 years (mean ± SD), and 17 were women, aged 83.1 ± 10.1 years. There was no significant difference in age between genders (p = 0.74), but the men were significantly taller (1.71 vs. 1.59 m; p < 0.001) and heavier (69.2 vs. 53.6 kg; p < 0.001) than the women (Table 1). The donors constituted a random sampling of the elderly

Results

The mechanical failure loads of L-4 varied between 1000 and 7870 N (Figure 2). In the men, the fracture load (3848 ± 1354 N) was significantly higher (p < 0.001) than in the women (2651 ± 1209 N) (Table 1). There was a moderate correlation of failure loads with body weight (r = 0.55, p < 0.001 for all subjects; r = 0.38, p < 0.05 for men; r = 0.57, p < 0.05 for women) and with the body height (r = 0.53, p < 0.001 for all subjects; r = 0.38, p < 0.05 for men, r = 0.43 for women). When

Discussion

In the current study we have determined the correlation of spinal and femoral DXA and calcaneal ultrasound, measured in situ with intact skin and soft tissues, with the in vitro failure load of the fourth lumbar vertebral body. We noted a significant, but only moderate, association of spinal DXA with the failure load of the fourth lumbar vertebral body, and a lower predictive ability of femoral DXA and calcaneal ultrasound. Nevertheless, calcaneal ultrasound was able to add independent

Acknowledgements

The authors thank Sabine Mühlsimer for photography help, and Markus Walz (Lunar, Germany) for his dedicated advice and support.

References (42)

  • D.C. Bauer et al.

    Quantitative ultrasound and vertebral fracture in postmenopausal womenFracture Intervention Trial and Research Group

    J Bone Miner Res

    (1995)
  • M. Biggemann et al.

    Prediction of the compressive strength of vertebral bodies of the lumbar spine by quantitative computed tomography

    Skel Radiol

    (1988)
  • K. Bjarnason et al.

    Anteroposterior and lateral spinal DXA for the assessment of vertebral body strengthComparison with hip and forearm measurement

    Osteopor Int

    (1996)
  • P. Brinckmann et al.

    Prediction of the compressive strength of human lumbar vertebrae

    Spine

    (1989)
  • X.G. Cheng et al.

    Prediction of vertebral strength in vitro by spinal bone densitometry and calcaneal ultrasound

    J Bone Miner Res

    (1997)
  • D.D. Cody et al.

    Correlations between vertebral regional bone mineral density (rBMD) and whole bone fracture load

    Spine

    (1991)
  • C. Cooper et al.

    Incidence of clinically diagnosed vertebral fracturesA population-based study in Rochester, Minnesota, 1985–1989

    J Bone Miner Res

    (1992)
  • S.R. Cummings et al.

    Bone mass measurements and risk of fracture in Caucasian womenA review of findings from prospective studies

    Am J Med

    (1995)
  • S.A. Eriksson et al.

    Prediction of vertebral strength by dual photon absorptiometry and quantitative computed tomography

    Calcif Tissue Int

    (1989)
  • M. Fischer et al.

    Phantom studies in osteoporosis

    Eur J Nucl Med

    (1993)
  • J. Frohn et al.

    Effect of aortic sclerosis on bone mineral measurements by dual-photon absorptiometry

    J Nucl Med

    (1991)
  • Cited by (56)

    • Neither altered incubation temperature during fetal development nor preferred rearing temperature improves leg bone characteristics of broilers

      2020, Journal of Thermal Biology
      Citation Excerpt :

      In this equation, the slope of the curve is given by a constant, which is obtained by multiplying the x variable, which expresses the bone stiffness (slope of the curve in the elastic phase). Bone strength, maximum force deformation, and rigidity were adjusted for body weight (according to Lockmûller et al., 1998). Femora and tibiotarsi of the right side of the body were tested for calcium, phosphorus, and ash.

    • Biomechanics of hip and vertebral fractures

      2020, Marcus and Feldman’s Osteoporosis
    • Biomechanics of Hip and Vertebral Fractures

      2013, Osteoporosis: Fourth Edition
    • Local irradiation alters bone morphology and increases bone fragility in a mouse model

      2010, Journal of Biomechanics
      Citation Excerpt :

      In fact, the cumulative incidence of insufficiency fracture (IF) following RTx can be upwards of 45%, as seen in women treated for cervical cancer (Kwon et al., 2008). Low bone mineral density (BMD) is often used to screen for fracture risk in older patients (Griffith and Genant, 2008; Lochmuller et al., 1998). However, in a study of IF’s post-RTx Konski and Sowers (1996) found that BMD at the fracture locations was not reduced.

    View all citing articles on Scopus
    View full text