Affordable Access

Publisher Website

Heterogeneity and Spatial Distribution of Intravertebral Trabecular Bone Mineral Density in the Lumbar Spine is Associated with Prevalent Vertebral Fracture

  • Kaiser, Jarred1
  • Allaire, Brett2
  • Fein, Paul M1
  • Lu, Darlene3
  • Adams, Alexander1
  • Kiel, Douglas P.2, 4, 5
  • Jarraya, Mohamed6, 7
  • Guermazi, Ali6
  • Demissie, Serkalem3
  • Samelson, Elizabeth J.3, 4, 5
  • Bouxsein, Mary L.2, 8
  • Morgan, Elise F1
  • 1 Department of Mechanical Engineering, Boston University, USA
  • 2 Department of Medicine, Beth Israel Deaconess Medical Center, USA
  • 3 Department of Biostatistics, Boston University, USA
  • 4 Department of Medicine, Harvard Medical School, USA
  • 5 Hinda and Arthur Marcus Institute for Aging Research, Hebrew SeniorLife, USA
  • 6 Department of Radiology, Boston University School of Medicine, USA
  • 7 Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School
  • 8 Department of Orthopedic Surgery, Harvard Medical School, USA
Published Article
Journal of Bone and Mineral Research
Wiley (John Wiley & Sons)
Publication Date
Jan 16, 2020
DOI: 10.1002/jbmr.3946
PMID: 31886907
PMCID: PMC7145746
PubMed Central


The spatial heterogeneity in trabecular bone density within the vertebral centrum is associated with vertebral strength and could explain why volumetric bone mineral density (vBMD) exhibits low sensitivity in identifying fracture risk. This study evaluated whether the heterogeneity and spatial distribution of trabecular vBMD are associated with prevalent vertebral fracture. We examined the volumetric QCT scans of the L3 vertebra in 148 participants in the Framingham Heart Study Multidetector CT study. Of these individuals, 37 were identified as cases of prevalent fracture, and 111 were controls, matched on sex and age with three controls per case. vBMD was calculated within 5-mm contiguous cubic regions of the centrum. Two measures of heterogeneity were calculated: 1) interquartile range (IQR); and 2) quartile coefficient of variation (QCV). Spatial distributions of the trabecular vBMD were also calculated: anterior/posterior, central/outer, superior/mid-transverse, and inferior/mid-transverse. Heterogeneity and spatial distributions were compared between cases and controls using Wilcoxon Rank Sum tests and t -tests and tested for association with prevalent fractures with conditional logistic regressions independent of integral vBMD. Prevalent fracture cases had lower mean (SD) integral vBMD (134 (38) vs.165 (42) mg/cm3, p<0.001), higher QCV (0.22 (0.13) vs. 0.17 (0.09), p=0.003), and lower anterior/posterior rBMD (0.65 (0.13) vs 0.78 (0.16), p<0.001) than controls. QCV was positively associated with increased odds of prevalent fracture (OR=1.61; 95% CI: 1.04–2.49, p=0.034), but this association was not independent of integral vBMD (p=0.598). Increased anterior/posterior trabecular vBMD ratio was associated with decreased odds of prevalent fracture independent of integral vBMD (OR= 0.38; 95% CI: 0.20–0.71, p=0.003). In conclusion, increased trabecular vBMD in the anterior vs. posterior centrum, but not trabecular vBMD heterogeneity, was associated with decreased risk of prevalent fracture independent of integral vBMD. Regional measurements of trabecular vBMD could aid in determining the risk and underlying mechanisms of vertebral fracture.

Report this publication


Seen <100 times