The relation of femoral osteon geometry to age, sex, height and weight.

As computational modeling becomes an increasingly common tool for probing the regulation of bone remodeling, the need for experimental data to refine and validate such models also grows. For example, van Oers et al. (R.F. van Oers, R. Ruimerman, B. van Rietbergen, P.A. Hilbers, R. Huiskes, Relating osteon diameter to strain. Bone 2008;43: 476-482.) recently described a mechanism by which osteon size may be regulated (inversely) by strain. Empirical data supporting this relation, particularly in humans, are sparse. Therefore, we sought to determine if there is a link between body weight (the only measure related to loading available for a cadaveric population) and osteon geometry in human bone. We hypothesized that after controlling for age, sex and height, weight would be inversely related to femoral osteon size (area, On.Ar; diameter, On.Dm). Secondarily we sought to describe the relation between osteon circularity (On.Cr) and these parameters. Osteons (n=12,690) were mapped within microradiographs of femoral mid-diaphyseal specimens (n=88; 45 male, 43 female; 17-97 yrs). Univariate analysis of covariance was conducted (n=87; 1 outlier) with sex as a fixed factor and height, weight and log-transformed age as covariates. Weight was negatively related to On.Ar and On.Dm (p=0.006 and p=0.004, respectively). Age was significantly related to osteon and, it was also significantly related to circularity (all p<0.001). This relation was negative for On.Ar and On.Dm and positive for On.Cr (increasing circularity with age). On.Ar and On.Dm were found to be significantly different between the sexes (p=0.021 and p=0.019, respectively), with females having smaller osteons. No relation between sex and On.Cr was detected (p=0.449). Height was not significantly related to any of the geometric parameters. Partial eta-squared values revealed that age accounted for the largest proportion (On.Ar: 28%, On.Dm: 18%, On.Cr: 30%), weight accounted for the second largest (On.Ar: 9%, On.Dm: 10%) and sex accounted for the smallest proportion (On.Ar: 6%, On.Dm: 7%) of the variance in geometry. While previous studies have reported relations between osteon size and sex/age, we believe that our findings are the first to demonstrate a link with weight. We believe that this negative relation with weight is most probably mechanical in nature; however, alternative (endocrine) links between bone and adipose tissue cannot be ruled out by our design.