RESUMO
Mechanical loading of muscle action is concentrated at muscle attachment sites; thus there may be a potential for site-specific variation in cortical bone thickness. Humeri from an early 20th-century Finnish (Helsinki) and two medieval English (Newcastle, Blackgate and York, Barbican) populations were subjected to pQCT scanning to calculate site-specific cross-sectional cortical bone area (CA) for four locations and to measure cortical thickness at muscle attachment sites and non-attachment sites. We found that CA at 80% of humerus length was significantly reduced compared to more distal cross-sections, which can be due to reduced stresses at the proximal shaft. The principal direction of loading at 80% humerus length was towards mediolateral plane, likely due to fixing the humerus close to the torso. At 35% the main direction of loading was towards anteroposterior plane, reflecting elbow flexing forces. The principal direction of loading varied between populations, sides and sexes at 50% humerus length due to preference between elbow and shoulder joint; thus this location might be useful when trying to infer differences in activity. These changes are likely due to overall shaft adaptation to forces acting at the humerus. In addition, we found a potential for site-specific variation in cortical thickness; cortical bone at muscle attachment sites was significantly thicker compared to non-attachment sites. Lastly, CA at 35% of humerus length and cortical thickness at non-attachment sites decreased with age. These results underline the importance of muscle loading for bone mass preservation as well as indicate that a site-specific variation of bone mass is possible.