Brief communication: cortical remodeling data are affected by sampling location.

It has been argued that techniques for estimating adult age-at-death from cortical histology are deleteriously affected by sampling location. This study uses nine complete femoral midshaft cross-sections to test the effect of sampling site on measurement of a standard histological variable, percent remodeled bone. Circumferential periosteal fields from four anatomically defined locations (anterior, posterior, medial, lateral) and four mechanically defined locations (maximum and minimum moments of area) were evaluated. Locations deviating from the periosteal surface toward the endosteal surface were also compared. Significant differences were found for both location and field placement. The anatomical axes exhibited greater variability than the mechanical axes, in particular the anterior location, a standard sampling site for age-at-death estimation techniques. More endosteal fields tended to show elevated levels of percent remodeled bone. This study demonstrates that circumferential and radial sampling locations are important considerations in deriving and applying predictive equations based on cortical remodeling.