Population variation in diaphyseal growth and age estimation of juvenile skeletal remains

Abstract

The rationale for developing and using population-specific juvenile age estimation methods is that groups with different “racial” or geopolitical memberships will presumably have dissimilar biological ancestry and, therefore, have inherently different growth and development schedules. However, this assertion is based on an outdated perspective of framing human variation that overlooks the impact of environmental factors on skeletal development. This study tests whether the cross-applicability of juvenile age estimation methods of diaphyseal lengths depends on (1) presumed shared ancestry, or (2) similarity in ontogenetic environment between the reference sample and the untested target sample or case. For this purpose, two juvenile age estimation methods developed from two vastly different reference samples were tested on four target samples ( N = 81). Individuals of the reference and test samples were categorized into two groups, African and European, which have been often used previously as proxies for ancestry and/or “race,” and assigned to either higher or lower quality of the ontogenetic environment for growth. Three measures, including mean error, mean absolute error, and precision percentage, were calculated and subjected to the t -test to assess whether methods’ validity and reliability depend on presumed shared ancestry or similar ontogenetic environments. The results support that the optimal performance of juvenile age estimation methods is conditional on the similarity in the ontogenetic environment between the reference and test samples. The humerus consistently provided the least biased estimates of age across the various test samples. The least variation in diaphyseal length was observed in children under the age of two. This indicates that, from birth to 2 years, age estimation methods are more universally applicable across various groups with different ontogenetic environments. We conclude that the similarity in ontogenetic environments is a better basis for developing and choosing appropriate methods, instead of relying on presumed shared ancestry.