Analysis of Age‐at‐Death Estimation Through the Use of Pubic Symphyseal Data*

Abstract

The question of whether age parameters derived from an American population will reliably estimate age-at-death for East European skeletal populations is important since the ability to accurately estimate an individual's age-at-death hinges on what standard is used. A reference sample of identified individuals with known ages-at-death from the regions of the Former Yugoslavia (n = 861) is used to determine the age structure of victims and serves as the prior in the Bayesian analysis. Pubic symphyseal data in the manners of Todd (Am J Phys Anthropol, 3 [1920], 285; Am J Phys Anthropol, 4 [1921], 1) and Suchey-Brooks (Am J Phys Anthropol, 80 [1986], 167) were collected for n = 296 Balkan males and females and for n = 2078 American males and females. An analysis of deviance is calculated using an improvement chi-square to test for population variation in the aging processes of American and East European populations using proportional odds probit regression. When males and females are treated separately, there is a significant association among females and the population (df = 1, chi-square likelihood ratio = 15.071, p = 0.001). New age estimates for Balkan populations are provided and are based on the calculated age distribution from the Gompertz-Makeham hazard analysis and the ages-of-transition. To estimate the age-at-death for an individual, the highest posterior density regions for each symphyseal phase are provided.

A test of Suchey–Brooks (pubic symphysis) and Buckberry–Chamberlain (auricular surface) methods on an identified Spanish sample: paleodemographic implications

“…However, despite its popularity, application of this method outside the USA on current samples from autopsied French individuals (Baccino et al, 1999), Canadian colonists (Saunders et al, 1992), Portuguese and Italians from current cemetery-based skeletal collections (Hens et al, 2008;Santos, 1996) and populations from the Balkans (Djuri c et al, 2007) demonstrated biased age estimates and difficulty in determining the age of individuals over 40 years. Furthermore, Sinha and Gupta (1995) observed differences in the timing of age-progressive pubic changes between USA and Indian samples; Hoppa (2000) and Kimmerle et al (2008) observed similar differences between females from USA and England and between females from USA and Balkans, respectively. In fact, these results are not surprising because Brooks and Suchey's original work (1990) pointed out a wide range of variability, especially in phases III to VI.…”

Human age estimation from lower-canine pulp volume ratio based on Bayes’ theorem with modern Japanese population as prior distribution

“…As mentioned above, inverse regression is not appropriate for age estimation unless the age distribution of the target sample is same as that of the reference sample. Since the critique by Bocquet-Appel and Masset (1982) of the conventional regression approach (inverse regression), the Bayesian approach has been a promising methodology for a new ageestimation scheme (Lucy et al, 1996;Buckberry and Chamberlain, 2002;Gowland and Chamberlain, 2002;Storey, 2007;Kimmerle et al, 2008;Konigsberg et al, 2008;Nagaoka et al, 2008;Coqueugniot et al, 2010;LangleyShirley and Jantz, 2010;Thevissen et al, 2010;Nagaoka et al, 2012a, b). The Bayesian approach takes the age distribution of the target population (referred to as prior distribution) into the probability calculation and, therefore, the answer (referred to as posterior distribution) circumvents the above-mentioned problem as it is customized to the target population.…”

A comparison of three established age estimation methods on an adult Spanish sample

“…The key to the successful application of a particular method is an understanding of whether the method is accurate (correct), precise (refined), and repeatable from an intra-and interobserver standpoint when applied to unknown individuals outside of the original reference sample [e.g. [4][5][6][7][8]. However, the reference samples on which many of the original methods were based are among very few known age-at-death collections of sufficient sample size for testing purposes [9][10][11].…”