Age-dependent change in the 3D structure of cortical porosity at the human femoral midshaft

DML Cooper, CDL Thomas, JG Clement, AL Turinsky… - Bone, 2007 - Elsevier
DML Cooper, CDL Thomas, JG Clement, AL Turinsky, CW Sensen, B Hallgrímsson
Bone, 2007Elsevier
Microstructural change associated with cortical bone remodeling has been extensively
explored with 2D techniques. However, relatively little is known regarding the 3D dynamic
microstructure of cortical bone. Therefore, we employed micro-CT imaging to investigate 3D
remodeling-related change in the structure of cortical bone porosity across the human
lifespan. Anterior femoral midshaft specimens (n= 51 male, 28 female) spanning 18 to 92
years of age were scanned with 7 μm nominal isotropic resolution. Canal volume fraction …
Microstructural change associated with cortical bone remodeling has been extensively explored with 2D techniques. However, relatively little is known regarding the 3D dynamic microstructure of cortical bone. Therefore, we employed micro-CT imaging to investigate 3D remodeling-related change in the structure of cortical bone porosity across the human lifespan. Anterior femoral midshaft specimens (n=51 male, 28 female) spanning 18 to 92 years of age were scanned with 7 μm nominal isotropic resolution. Canal volume fraction (Ca.V/TV), mean diameter (Ca.Dm), mean separation (Ca.Sp), degree of anisotropy (DA), connectivity density (Ca.ConnD), and number (Ca.N) were calculated for subperiosteal cylindrical regions of interest. Ca.N was calculated in 2D (Ca.N2D) and 3D (Ca.N3D). Regression was used to examine the relation between the structural parameters and age. Additionally, the impact of sex, height, and weight were investigated collectively (MANCOVA) and individually (ANCOVA). For all analyses, Ca.V/TV and Ca.Dm were inverted (Ca.V/TV−1, Ca.Dm−1) to establish normality and linear relations with age. Ca.N values (2D and 3D) were non-linearly (quadratic) related to age, increasing until the 6th decade then decreasing. This relation was only significant for the pooled sexes Ca.N3D values (p=0.012). Ca.ConnD was positively related to age (p<0.05), while all remaining 3D parameters, except DA for males (p=0.070), were negatively related (p<0.05). In all cases, the relation with age was strongest for females. MANCOVA revealed that age was the only significant (p<0.001) covariate overall. Univariate ANCOVA indicated significant differences between the sexes for Ca.V/TV−1 and Ca.Dm−1 (p=0.018 and 0.010, respectively). Relative to males, females had lower values for these parameters, translating into larger mean canal diameter and overall porosity. Body weight had a significant (p=0.043) positive relation with Ca.Dm−1, indicating lower weight was also associated with increased mean canal diameter. Therefore, while age was the most important factor, sex and body size were found to play a role in parameters related to canal size and the overall level of porosity. This study is unique in that changes in cortical bone microstructure were examined across the adult human lifespan in three rather than two dimensions.
Elsevier