Abstract
<p>Bone remodelling is an important process with numerous clinical implications. It is generally
assumed that the bone will experience the same bone remodelling response to the stimulus under
tension and compression. However, laboratory studies suggest that the bone could exhibit
alternative bone remodelling under tension and compression as the result of different bone
remodelling mechanisms; where alternative bone remodelling is defined as two or more
different/alternative bone remodelling responses in relation to the same stimulus. Here, tension
and compression are assumed to be the same stimulus, where the absolute value of the strain
stimulus is used. Very little research has been conducted into this phenomenon and if it occurs
under physiologically applied loading conditions. Therefore, it is unclear how prevalent alternative
bone remodelling is under physiological loading conditions.</p>
<p>To investigate if the bone does exhibit alternative bone remodelling under tension and compression
under physiological loading conditions, this study examined the bone remodelling behaviour of
the proximal femur of 11 male subjects (mean age ± SD: 70.91 ± 2.78) taking part in a one-year
hopping clinical exercise trial. This was achieved by comparing the change in density of each
subject, measured by quantitative computerised tomography, against the mechanical stimulus,
determined by subject specific finite element simulations. Here the stimulus was determined to be
the principal strain experienced in hopping minus the principal strain experienced during everyday
activities which were either walking or stair climbing. The stimulus-remodelling relationships
were determined by comparing the change in density and stimulus using a previously published
bone remodelling algorithm.</p>
<p>This study managed to demonstrate consistent differences in the bone remodelling observed under
tension and compression. Where it was observed that at lower stimuli the compressive regions
experience a higher change in density in relation to the stimulus in comparison to the tensile
regions, until a cross-over point where tension experiences a higher change in density in
comparison to compression (p < 0.01). This cross over point occurs at a stimulus of approximately
670 µε which corresponds to a change in density of approximately 3.5%, which is higher than the
typical change in density experienced in the cortical bone during exercise. This cross over between
the two stimulus-remodelling relationships is thought to be the result of the Hueter-Volkmann law. </p>
<p>It was also observed that the bone remodelling under compression has a higher variance in the
change in density in relation to the stimulus, in comparison to tension (p < 0.01), which is thought
to the be result of linear microdamage formation and subsequent remodelling initiation under
compression. This high variance in the change in density under compression causes for the
stimulus-remodelling relationships under tension and compression to merge somewhat at lower
stimuli, which can give the impression of a single stimulus-remodelling relationship. Furthermore,
differences in the residual regression biases under tension and compression was observed, where
the bone remodelling under tension demonstrated significantly lower regression biases in
comparison to the bone remodelling under compression (p = 0.141).</p>
<p>Evidence from this study also suggests that the tensile and compressive stimulus-remodelling
relationships are independent of each other, but correlations of varying strength were made
between the observed characteristics and the subjects’ age and BMI, with the correlations being
typically stronger under tension. However further research is warranted into these correlations due
to the small sample size and population demographic used in this study. </p>