Osteoporosis is a disease characterized by bone loss affecting 10% of the US population over 50 years old. The spine is one critical area affected by the disease. The effectiveness of experimental treatments can be tested on an ovariectomized rat osteoporosis model. As a result, lumbar vertebral bodies are often mechanically tested in uniaxial compression in order to determine whether or not the mechanical properties of the bone in ovariectomized rats improve with treatment. The irregular shape of rat vertebral bodies requires some specimen preparation to create two parallel loading surfaces for uniaxial compression testing.

Two specimen preparation methods are reported in the current body of literature. One cuts the cranial and caudal surfaces to make them parallel to each other. The other cuts the caudal surface and uses bone cement to create a flat loading surface at the cranial end. In this thesis a total of twenty rat vertebral bodies were tested. Ten were prepared with a cut specimen preparation method and ten with an embedding method. Each specimen was tested in uniaxial compression and was microCT scanned before and after testing. Eleven parameters were calculated from the mechanical testing data and compared between the two groups using Student’s t-tests. The specimens were also categorized into six failure modes and locations observed in the microCT images.

The embedded specimens showed a lower stiffness (p = 0.026), greater yield displacement (p = 0.007) and apparent strain at failure (p = 0.050). These differences were largely attributed to the embedded specimens being 1 mm taller than the cut specimens. The shorter size of the cut specimens affected the mechanical parameters. The cut specimens were easier to prepare and were less sensitive to end effect failures. The embedded specimens kept the endplate, which distributes the load from the intervertebral disk through the vertebral body, intact. In addition, the embedded specimens exhibited two failure modes, endplate failure and failure at the center of the vertebral body, observed in ex vivo human lumbar vertebral body testing, which suggests the interaction of the vertebral body with the endplate is an important factor in vertebral body failure in uniaxial compression testing. In conclusion, neither preparation method showed an overwhelming advantage over the other, and experimental parameters should be considered when choosing a loading surface preparation method.