Author's personal copy Vertebral fractures in the elderly may not always be “osteoporotic” G. Jiang a ,J.Luo b , P. Pollintine c , P. Dolan b , M.A. Adams b, ⁎, R. Eastell a a Academic Unit of Bone Metabolism, Department of Human Metabolism, University of Sheffield, UK b Department of Anatomy, University of Bristol, UK c Department of Mechanical Engineering, University of Bath, UK abstract article info Article history: Received 1 February 2010 Revised 11 March 2010 Accepted 25 March 2010 Available online 31 March 2010 Edited by: Thomas Einhorn Keywords: Vertebral fracture Osteoporosis Trauma Radiographs ABQ method Introduction: Vertebral fractures in the elderly are often assumed to be “osteoporotic” and require anti- osteoporosis therapy. However, some of these fractures may represent traumatic injuries to vertebrae that havecomparativelynormalbonemineraldensity(BMD).Wehypothesizethatradiographicappearancescan be used to differentiate between “osteoporotic” fractures of vertebrae with low BMD and strength, and “traumatic” fractures of vertebrae with normal BMD and strength. Methods: 73 cadaveric specimens (each comprising two vertebrae with the intervening intervertebral disc andligaments)wereobtainedfromdonorsaged42to91(mean74)years.ArealBMDwasmeasuredinthe lateral projection for each vertebral body, using DXA. Each specimen was secured in metal cups containing dental plaster, and compressed to failure at 3 mm/s on a computer-controlled materials testing machine. Mechanical failure was detected by a reduction in the gradient of the load-deformation curve. Compressive deformation for each specimen was limited to 4 mm in order to prevent gross destruction of the vertebra. Radiographs,obtainedbeforeandaftermechanicalloading,wereassessedbyanexperiencedradiologist(GJ) who was blinded to BMD and mechanical data. The algorithm-based qualitative method (ABQ) was used to assign each specimen to two possible outcomes: no discernible fracture of either vertebra, or fracture. The latterwerefurtherclassifiedintospecimenswithosteoporoticfractureandthosewithtraumaticfracture,by applyingadditionalcriteriafordifferentialdiagnosis.TherelationshipoffailureloadtoBMDwastestedusing correlation. BMD and failure load for the three diagnostic outcomes were compared using one-way analysis of variance (ANOVA). Results: Failure load was proportional to BMD (R =0.63, p b 0.001). “Osteoporotic,”“traumatic” and “no discernible” fractureswerereportedin16,26and31specimensrespectively. “Traumatic” fracturespecimens had higher BMD and failed at higher loads than “osteoporotic” fracture specimens (p b 0.05). Conclusions: Somevertebralfracturesintheelderlymaybetraumaticratherthanosteoporoticinorigin.Our radiological criteria help to differentiate between them. © 2010 Elsevier Inc. All rights reserved. Introduction Vertebral body fracture is the most common type of osteoporotic fracture [1]. In vertebral osteoporosis, the endplate becomes weak- ened due to the loss of support from trabecular bone, and due to thinning of the endplate itself [2–4]. Therefore, in patients with osteoporosis, a moderate force generated from daily activity may be sufficient to induce vertebral fracture. However, a vertebral fracture identifiedinanelderlypersonmightbeduetoatraumaticeventthat occurred recently, or many years previously when the vertebra was underahighimpactforce,eventhoughbonedensitywasnormal. Traumaticvertebralfracturesarecommonlyseeninyoungadults, and may persist as permanent deformities. Accurate medical history can be difficult to ascertain due to poor recall in the elderly, and previousradiographsareoftenunavailableforcomparison.Traumatic vertebral fractures could possibly have produced minimal symptoms at the time of injury, or symptoms may have been ignored or interpretedinaccurately,perhapsduetothepresenceofotherinjuries [5,6]. For these reasons, some traumatic vertebral fractures in the elderly may be mistakenly treated as osteoporotic. The first author has developed clinical radiographic methods to distinguishbetweenthesedifferenttypesofvertebralfractureinvivo. However, the clinical criteria (Table 1) have not previously been validated against precise quantitative data concerning the force required to cause fracture. Mechanical loading tests on cadaveric spines can produce experimental vertebral fractures similar to those seen in vivo, using accurately controlled forces. Therefore, we investigatedwhetherthefailureloadandBMDofcadavericvertebral specimensthatwereinducedtofracturebymechanicaltestingvaried according to whether the fracture appearances on radiographs were classifiedas “osteoporotic” or “traumatic.” Bone 47 (2010) 111–116 ⁎ Corresponding author. Reader in Spine Biomechanics, Department of Anatomy, University of Bristol, Southwell St., Bristol BS2 8EJ, UK. Fax: +44 117 9254794. E-mail address: M.A.Adams@bristol.ac.uk (M.A. Adams). 8756-3282/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2010.03.019 Contents lists available at ScienceDirect Bone journal homepage: www.elsevier.com/locate/bone