Age at death estimation by cementochronology: Too precise to be true or too precise to be accurate?

Objectives Cementochronology based on annual deposition of acellular cementum is acknowledged for its superior performance for estimating age-at-death but is also disregarded because of its suspicious effectiveness. This article aims to provide a standardized framework for the validation of the technique and to define and test cementochronology's performance and limitations. Materials and methods To determine the boundaries of the cementum aging technique, we applied a certified protocol on a sample of 200 healthy canines from individuals of known age, sex, postmortem, and inhumation intervals from anthropological and anatomical collections. We scored readability and preservation of cementum and measured the agreement between estimates, i.e., the precision, and assessed the quality of the accordance between estimates and chronological age, i.e., the accuracy. To investigate the applicability on ancient material, 200 additional canines extracted from archeological assemblages were included. Accuracy and precision were analyzed for each age group in considering postmortem intervals and taphonomical conditions. Results A strong correlation was found between chronological age and estimates (r = .927; p = .000) but results reveal an association between readability of incremental lines and chronological age (p < .05) and a notable difference in both precision and accuracy between individuals under and over 50 years. Results also demonstrate that taphonomy can be a serious obstacle increasing imprecision by a factor of three. Discussion Improperly adopted, cementochronology can lead to precise but inaccurate estimations. If methodological, physiological, and taphonomical factors are taken into account, then, and only then, cementochronology will serve as a versatile and powerful tool for age-at-death estimation.