Abstract
The decline in areal BMD (aBMD) in men may start as early as age 19, but it is debated whether proximal femoral strength follows the same pattern. We measured FN aBMD in a cross-sectional cohort of 1101 men aged 18-28 yr with two-dimensional DXA. Body height, weight, jump height, grip strength, and maximum knee extension torque muscle strength were measured with standard tests, after which BMI was calculated. We adopted a validated method to automatically obtain three-dimensional finite element (FE) models from 2D DXA images and predict proximal femoral strength. We compared FN aBMD and FE-predicted proximal femoral strength as a function of age. Linear mixed-effect models (LMEs) were used to evaluate the association between age (whole years), FN aBMD, FN area, jump height, grip strength, maximum knee extension torque, and FE-predicted proximal femoral strength. A negative correlation (r = -0.10, p = .001) was observed between FN aBMD and age between ages 18 and 28; no statistically significant correlation (r = -0.02, p = .55) between age and FE-predicted proximal femoral strength was found. There was a positive correlation between FE-predicted proximal femoral strength and the subjects' height (r = 0.22, p < .001) and weight (r = 0.35, p < .001). Higher jump height (r = 0.13, p < .001), grip strength (r = 0.23, p < .001), and maximum knee extension torque (r = 0.31, p < .001) were associated with higher FE-predicted proximal femoral strength. Femoral neck aBMD, FN area, and body height explained ~78% of the variations in proximal femoral strength in the LMEs. The inter-subject variations in proximal femoral strength were associated with FN aBMD, FN area, body height, and higher outcomes in jump height, grip strength, and maximum knee extension torque.