Assessing Biological Mortality Bias From Deciduous Tooth Emergence.

OBJECTIVES: Traits found in a skeletal sample are frequently used to infer qualities of the living population from which the skeletons were drawn. However, traits observed in a mortality sample may exhibit biological mortality bias in that they may not accurately represent the same traits in the living sample. The purpose of this research was to assess biological mortality bias in deciduous tooth emergence, a trait that is used to estimate chronological age in skeletal samples. MATERIALS AND METHODS: Data on clinical tooth emergence were collected from longitudinal studies of Javanese, Guatemalan, and Bangladeshi children that included a living sample (those who survived through the study) and a mortality sample (those who died during the study). Parametric hazards analysis was used to test for differences in the timing of tooth emergence between the living and mortality samples. RESULTS: There were no significant differences between the living and mortality samples for Bangladesh and Java, although there was a trend toward delayed emergence in the Bangladesh mortality sample. The Guatemalan mortality sample exhibited advanced emergence of the posterior dentition for the mortality sample. No evidence of biological mortality bias was found in pooled analyses of the Bangladesh and Java samples or pooled analyses of all three samples. DISCUSSION: We found limited evidence that deciduous tooth emergence in a mortality sample (e.g., a skeletal series) would differ from the timing of emergence in the living population from which the mortality sample was drawn.