The relation between mechanical loading and bone biodynamics on the periosteal and endosteal envelopes of long-bone diaphyses has been used extensively as the basis for hypothesizing differences in activity levels among and within skeletal populations. The Haversian envelope, though also responsive to loading, has received little attention in biomechanical analyses of skeletal remains despite the fact that this envelope is frequently well-preserved. A major obstacle in mechanical interpretations of cortical bone microstructure stems from difficulties in differentiating the relative contribution of systemic (e.g., circulating hormones) influences and mechanical influences to an observed histomorphology.

If the systemic remodeling can be partitioned from the mechanically induced remodeling in a dynamically loaded bone, a mechanical interpretation of the microstructure would be possible. Further, a method for deducing activity levels from the Haversian envelope could be derived. I investigate the efficacy with which systemic (nonmechanically induced) remodeling in the midshaft femur can be controlled for by standardization with remodeling in the rib of the same individual.

Histomorphometric data were collected from undemineralized thin sections of the midshaft femur and middle third of the 6th rib in 49 adults (26 ♂, 23 ♀) from Paloma, a Preceramic archaeological site on the central coast of Peru. The skeletal remains derive from three contiguous stratigraphic levels at Paloma. Previous hypotheses, suggesting that physical activity levels decreased over time, are tested using histologic and gross geometric data.

Femoral osteon population densities (OPD_F) are regressed onto rib OPD (OPD_R), from which residual scores (^R) are calculated (^ROPD_F)- According to the model, ^ROPD_F represents local femoral turnover not accounted for by systemic influences. Second moments of area (I_max, I_min, J) are calculated from the same femur sections used in the histological analysis. Bending and torsional moments are standardized for body mass and beam length by calculating residual scores from the regression of second moments onto [body mass x femur length'].

When residual scores from the gross geometric analysis (^RI_max, ^RI_min, ^RJ) are plotted against residual scores from the histologic analysis in the femur (r OPDf), significant, positive correlations are found between ^ROPD_F and ^RI_max and between ^ROPD_F and ^RJ, but not between ^ROPD_F and ^RI_min. The results indicate that intracortical remodeling in the midshaft femur, after correction for systemic influences, is greater in cross sections exhibiting greater bending and torsional rigidity and is lower in sections exhibiting lower rigidity. Additionally, trends in geometric scores by stratigraphic level at Paloma are similar to those derived from the histologic method, suggesting that the two measures reflect the same phenomenon—mechanical loading history.

These data confirm clinical and experimental studies addressing the influence of mechanical usage on local Haversian turnover, and suggest that a significant amount of nonmechanically induced turnover in the femoral cortex can be accounted for by standardization with the rib. Further, the use of rib-standardized remodeling rates in the femur can provide a reliable estimate of mechanical loading history in extinct populations.