1Department of Archaeology, University of Cambridge, 2Department of Anthropology, Pennsylvania State University, 3Department of Radiology, University of Southern California, 4Department of Geology and Paleontology, Georgian National Museum, 5Department of Integrative Anatomical Sciences, University of Southern California, 6Evolutionary Studies Institute, University of the Witwatersrand, 7Department of Anthropology, University at Albany, SUNY, 8Department of Anthropology, Western University
April 16, 2020 , Platinum Ballroom
Variation in trabecular and cortical bone structure is often used to infer habitual behavior in the past. However, signals from both types of bone are rarely considered together and may even contradict each other. Here we examine trabecular and cortical bone structure in living people to understand the mechanical signals hidden in both types of bone.
We compare trabecular and cortical bone mechanical properties in pQCT scans of the tibia between groups of 83 male athletes (running, hockey, swimming, cricket) and sedentary controls using a Bayesian multilevel model.
All groups show distinct combinations of mean tibia midshaft J0.73, Imax/Imin, and trabecular bone density in the distal tibia. There is a very minor but credible correlation between midshaft cross-sectional properties and trabecular bone density (R2=0.06), but no relationship within any sporting categories.
All sporting groups have a credibly higher mean trabecular bone density compared to controls, but do not differ amongst each other. Cortical bone J0.73 is greater in high impact sports (cricket, running, hockey) compared to controls, but not between low impact swimming and controls. These results suggest that diaphyseal cortical bone adapts to high impact bending loads while trabecular bone density increases under any kind of high activity.
Individuals from the different categories overlap substantially, but the posterior distributions of mean mechanical properties are clearly separated by unique combinations of cortical and trabecular variables. As such, detailed group-level behavioral inferences are possible by combining cortical and trabecular properties in our analyses of past human behavior.
Funding: RCUK/BBSRC grant BB/R01292X/1, NSF BCS-1719187, NSF BCS-1719140