Since the implementation of high resolution peripheral quantitative computed tomography (HR-pQCT) technology in the early 2000’s, it has provided a non-invasive, low radiation approach to quantifying compartmental bone mineral density and bone microarchitecture in vivo1. The growth of HR-pQCT use in clinical research studies2 has led to its continued development and new applications, in particular its use in fracture prediction3, 4, 5, 6, 7, 8 and assessment of bone remodeling9. A fundamental challenge of HR-pQCT is that skeletal health is assessed based on small (∼10mm) scan regions, which highlights the importance of selecting a consistent and appropriate scan position.

In recent years, a divergence of protocols to standardize HR-pQCT image acquisition position has occurred. All methods start by defining the region of interest using a primary 2D scout view10; however, two approaches to define this scan position have emerged: (1) the fixed offset positioning, and (2) the relative offset positioning. For the fixed offset method, the scan operator places a reference line at a known landmark on the endplate of the bone, the scan region then begins a fixed distance of 9 mm and 22 mm10 proximal to the reference line at the radius and tibia, respectively (we focus on second-generation HR-pQCT; first-generation fixed offsets differ slightly10). The concern is that when using the fixed offset positioning, a shorter individual may result in an image captured more proximally, and in taller individuals an image captured more distally11. The relative offset positioning attempts to take into account anatomical variance by determining the region of interest based on the individual's bone length. For the relative offset positioning method, the operator measures the individual's limb length and sets the scan region as 4 % and 7.3 % of the bone length10 from the articular surface at the distal radius and distal tibia, respectively. However, the concern raised about a relative offset positioning method is that it assumes that the underlying bone structure (especially the size and location of the metaphysis) scales linearly with bone length, which to the best of our knowledge has not been verified. Therefore, the consistency in the acquisition of a specific anatomical site across the population remains unknown.

Discrepancies between the use of these two methods in site-specific scan protocols may be problematic for comparisons between studies and participants12. Previous investigation has shown that the proportion of trabecular bone to cortical bone differs significantly in the region where HR-pQCT measurements are performed13,14, indicating a considerable outcome difference as a result of small changes in acquisition offset. Studies investigating the impact of the use of different scan protocols have highlighted the importance of standardization to allow comparison across cohorts12,15, but whether a fixed or relative offset resolves this issue is not known.

The objectives of this study were (1) to determine how anatomical landmarks scale with bone length, and (2) evaluate a fixed and relative offset positioning for providing a consistent anatomical location for scan acquisition. We hypothesize that neither the fixed nor relative offset positioning will provide an anatomically consistent location for scanning.