Late comment for this topic:
studies using computer tomography suggest that good bone strength values in impact-trained track athletes are related to improved geometrical structure (enlarger bone cortices) rather than bone density. The training effect seems to be quite similar in tibia and femoral neck (e.g. Nikander et al. Osteoporos. Int 2010, 21:1687-1694; Heinonen et al. Bone 2001, 29:279-285; Nikander et al. JBMR 2005,20, 520-8 ). The interindividual differences in bone properties (strength/stiffness) in athletes are small and may not have significant influence on performance ?