Histological age assessment in a prehispanic Maya sample from Xcambó, Yucatan, Mexico: Benefits and limitations

The present study explores four different regression formulae for the histological assessment of age at death in ribs and their performance in a prehispanic Classic Maya population sample consisting of 57 individuals from Xcambó, Yucatan, Mexico. Regressions employed include the original age regression formula by Stout and Paine (1992), Cho et al.'s (2002) two formulae for samples of indeterminate ethnicity, as well as a formula published by Valencia et al. (2010) adjusted specifically for populations of Maya descent. In addition to applying these methods, we report histomorphometric variables (total cortical area (TA), cortical area (CA), relative cortical area (CA/TA%), osteon population density (OPD), and osteon cross-sectional area (On.Ar.)) from the 6th rib and compare these variables across groups within our sample defined by macroscopic age and sex, as well as with results reported for modern reference samples used by Cho et al. and Valencia and colleagues. Our study shows mean CA and CA/TA% are relatively high across all ages in the prehispanic Maya sample, especially in females, indicating a high degree of robusticity. OPD is high when compared with samples used by Cho et al., but similar to the modern Maya reference sample. Comparison of histological age at death estimates reveals interesting patterns of deviation; specifically Cho et al.'s formulae both deviate strongly from all other age estimates. Calculated mean net difference between Cho et al.'s and macroscopic age estimates, for example, is nearly 16 years. Both, Stout and Paine, and Valencia et al., result in age reconstructions more similar to macroscopic estimates (mean net difference around 8 years). Since Cho et al.'s formulae are unique in employing On.Ar., CA, and TA, in addition to OPD, OPD-based regression formulae may perform better in archaeological samples. However, some of the deviation observed could result from differences in histomorphometric variables between modern reference and archaeological samples, the outcome of complex biocultural processes. Continued analyses of histomorphological variation between differing reference and archaeological samples will be necessary to improve histological assessments of age at death in archaeological contexts.