Poster Presentation Annual Meetings of the Endocrine Society of Australia and Society for Reproductive Biology and Australia and New Zealand Bone and Mineral Society 2016

Is proximal femur geometry from DXA-derived 3D analysis predictive of pQCT-derived geometry of the tibia (#270)

Amy T Harding 1 2 , Conor Lambert 1 2 , Benjamin K Weeks 1 2 , Rossana Candiota Nogueira 1 2 , Steven L Watson 1 2 , Sally F Dzera 1 2 , Belinda R Beck 1 2 3
  1. Menzies Health Institute Queensland, Griffith University, Gold Coast, QLD, Australia
  2. School of Allied Health Sciences, Griffith University, Gold Coast, QLD, Australia
  3. The Bone Clinic, Brisbane, QLD, Australia


While DXA scans are the accepted osteoporosis diagnostic, their planar nature and inability to discriminate cortical from trabecular compartments are recognised limitations.  pQCT was developed so that bone morphology along with cortical and trabecular bone could be detected, but is limited by protracted scanning times and an inability to examine the most clinically relevant sites.  The recent development of software to determine proximal femur (PF) geometry from conventional DXA images addresses those limitations.



To determine whether DXA-derived 3D geometry at the PF is representative of pQCT-derived geometry at the tibia.



Apparently healthy men and women, screened for conditions and medications that influence bone, were recruited from the community.  DXA scans of the PF (Medix DR, Medilink) and pQCT scans of the 4% and 38% sites of the tibia (XCT-3000, Stratec) were conducted.  DXA scans were analysed using 3D Hip software (DMS Group, France) to derive femoral neck (FN) and total hip (TH) volume and cortical thickness.  Regression analyses were conducted.



Seventy-eight men (52.1±20.6yrs, 176.7±7.4cm, 81.6±13.8kg) and 156 women (54.4±18.3yrs, 163.2±6.3cm, 66.5±13.3kg) were recruited.  FN and TH total volume strongly predicted 53% and 57% of the variance in total area of the 4% tibia (P<0.001, respectively).  FN cortical and trabecular volume accounted for 44.5% and 49.0% of the variance in tibial 38% cortical and 4% trabecular areas, respectively (P<0.001).  TH cortical volume predicted 52.0% of the variance in cortical area at the 38% tibia (P<0.001).  Total FN cortical thickness explained 17.9% of the variance in cortical thickness at the 38% tibia (P<0.01).



Findings indicate DXA-derived 3D geometry of the proximal femur provides a representative index of lower extremity bone geometry, supporting the use of 3D Hip DXA analysis as an alternative to pQCT measures of bone geometry at less clinically relevant sites.