Quantifying sex, race, and age specific differences in bone microstructure requires measurement of anatomically equivalent regions

Individuals differ in forearm length. As microstructure differs along the radius, we hypothesized that errors may occur when sexual and racial dimorphisms are quantified at a fixed distance from the radio-carpal joint. Microstructure was quantified ex vivo in 18 cadaveric radii using high resolution peripheral quantitative computed tomography and in vivo in 158 Asian and Caucasian women and men at a fixed region of interest (ROI), a corrected ROI positioned at 4.5–6% of forearm length and using the fixed ROI adjusted for cross sectional area (CSA), forearm length or height. Secular effects of age were assessed by comparing 38 younger and 33 older women. Ex vivo, similar amounts of bone mass fashioned adjacent cross sections. Larger distal cross sections had thinner porous cortices of lower matrix mineral density (MMD), a larger medullary CSA and higher trabecular density. Smaller proximal cross-sections had thicker less porous cortices of higher MMD, a small medullary canal with little trabecular bone. Taller persons had more distally positioned fixed ROIs which moved proximally when corrected. Shorter persons had more proximally positioned fixed ROIs which moved distally when corrected, so dimorphisms lessened. In the corrected ROIs, in Caucasians, women had 0.6 SD higher porosity and 0.6 SD lower trabecular density than men (p < 0.01). In Asians, women had 0.25 SD higher porosity (NS) and 0.5 SD lower trabecular density than men (p < 0.05). In women, Asians had 0.8 SD lower porosity and 0.3 SD higher trabecular density than Caucasians (p < 0.01). In men, Asians and Caucasians had similar porosity and trabecular density. Results were similar using an adjusted fixed ROI. Adjusting for secular effects of age on forearm length resulted in the age-related increment in porosity increasing from 2.08 SD to 2.48 SD (p < 0.05). Assessment of sex, race and age related differences in microstructure requires measurement of anatomically equivalent regions.