Estimating body mass from skeletal material: new predictive equations and methodological insights from analyses of a known-mass sample of humans

Estimating body mass from skeletal material is a key task for many biological anthropologists. As a result, several sets of regression equations have been derived for cranial and postcranial material. The equations have been applied to a wide range of specimens, but several factors suggest they may not be as reliable as generally assumed. Specifically, since many of the equations were derived from small reference samples using proxies for key variables and/or mean data, the nature of the relationship between the skeletal variables and body mass has often not been adequately demonstrated. In addition, few of the equations have been validated on known samples, making their accuracy and precision uncertain. Lastly, because no study has used cranial and postcranial material from the same individuals, the two approaches have never been systematically compared. The present study responded to these issues by deriving new regression equations from cranial and postcranial material using a large sample of modern humans of known-mass and associated skeletal variables measured from CT data. The equations were then tested on an independent sample, also of known mass. The results show that the newly derived equations estimate mass more accurately than existing equations for most variables. However, improvements were modest and accuracy rates remained relatively low. In addition, variables that had previously been argued to be ideal predictors were not the most accurate, and the current criteria used to assess equations did not ensure reliability. Overall, the results suggest that body mass estimates must be used cautiously and that further research is required.