Differences in Femoral Geometry and Structure Due to Immobilization

Reduction in bone mass of the lower extremity is well documented in individuals with paralysis resulting from spinal cord injury (SCI). The consequent osteopenia leads to elevated fracture risk with fractures occurring more commonly in the femoral shaft and supracondylar regions than the hip. A model has recently been described to estimate geometry and structure of the femoral midshaft from whole body scans by dual X-ray absorptiometry (DXA). Increases in femoral geometric and structural properties during growth were primarily related to mechanical loading as reflected by body mass. In this study, we investigate the relationship between body mass and femoral geometry and structure in adults with normal habitual mechanical loading patterns and those with severely reduced loading. The subjects were 78 ambulatory men (aged 20-72 yrs) and 113 men with complete paralysis from SCI of more than 4 years duration (aged 21 73 yrs). Subregional analysis was performed on DXA whole body scans to obtain bone mineral content (BMC, g), cortical thickness (cm), crosssectional moment of inertia (CSMI, cm4), and section modulus (cm3) of the femoral midshaft. All measured bone variables were significantly lower in SCI compared with ambulatory subjects: -29% (BMC), -33% (cortical thickness), -23% (CSMI), and -22% (section modulus) while body mass was not significantly different. However, the associations between body mass and bone properties were notably different; r2 values were higher for ambulatory than SCI subjects in regressions of body mass on BMC (0.48 vs 0.20), CSMI (0.59 vs 0.32), and section modulus (0.59 vs 0.31). No association was seen between body mass and cortical thickness for either group. The greatest difference between groups is in the femoral cortex, consistent with reduced bone mass via endosteal expansion. The relatively lesser difference in geometric and structural properties implies that there is less effect on mechanical integrity than would be expected from bone mass results alone. The reduced association in SCI subjects between body mass and bone properties is not unexpected. Although mean body mass differs little between ambulatory and SCI individuals, the association between body mass and in vivo skeletal loading is no longer present, as mechanical influences are removed except for transfer activities. The residual association is probably attributable to the strength of this influence during growth. These results highlight the importance of examining geometry and structure in conjunction with bone mass.