A Bayesian approach to the estimation of adult skeletal age: Assessing the facility of multifactorial and three-dimensional methods to improve accuracy of age estimation

In the estimation of age of the human skeleton, it is possible to form a reasonably accurate estimate for individuals younger than twenty years. The older the individual, the more difficult it becomes to determine age. Current methods of age estimation are biased toward data collected from considerable numbers of younger adults and small numbers of older adults. Additionally, adult age indicators are often limited in range and application. The result is consistent underestimation of adult age and uncomfortably large age ranges for adult materials.; This study consists of two interconnected analyses: (1) examination of three-dimensional patterns of age-related skeletal deterioration of the pubic symphysis to identify proportional shifts in physical topography corresponding to age-at-death estimates, and (2) application of a Bayesian approach to formulation of a multifactorial standard aimed at increasing accuracy of estimating adult age from the skeleton.; This study included skeletal remains of 135 white males and 70 white females of known age. The principal focus within this sample was on individuals between 45-70 years, subdivided into five-year intervals. The author scored individual remains based on fusion of the medial clavicle and sternum. The author also scored remains on key aspects of the physical appearance of the pubic symphysis and assigned each to an appropriate Suchey-Brooks phase. The first, fourth, and seventh ribs were examined and assigned to phases corresponding to comparison casts and methods developed by Iscan and colleagues.; The results of this study indicate that while general population trends in aging are found among individual skeletal characteristics, reducing the aging process to its principal components in five-year intervals does not generally result in more accurate estimation of age. Instead, the limited time intervals and key traits observed tend to reflect the enormous variability of the aging process at the level of the individual. Within each five-year age increment, individual variability in the aging process tends to overwhelm general trends. Despite inherent individuality in the aging process, multifactorial Bayesian analysis of skeletal traits may significantly increase confidence in estimation of the age-at-death for an individual, in some instances to 100% within a five-year age range.