# The impact of age-related changes in the skull on sex estimation using morphoscopic traits

## TL;DR

This study examines how aging affects sex estimation using skull traits in a South African population, finding that age has a limited impact compared to population-specific factors.

## Contribution

The study evaluates the influence of age on cranial sex estimation accuracy using a large, age-diverse sample from South Africa.

## Key findings

- Only the nuchal crest showed significant age-related differences in female skulls.
- Age-specific standards slightly improved accuracy for younger individuals but reduced it for older ones.
- Population-specific standards are more critical for accurate sex estimation than age pre-selection.

## Abstract

The five sexually dimorphic traits of the skull described by Walker are frequently employed in skeletal analyses for sex estimation. Previous research has highlighted various changes in the craniofacial complex associated with advancing age, as cranial remodelling persists into adulthood. Age has been recognised as one potential factor contributing to misclassification when using the Walker traits. This study aimed to evaluate the influence of age on the expression of sexually dimorphic cranial traits and the impact of that influence on sex estimation in a South African sample. The traits were scored on a sample of 453 skulls between 14 and 108 years of age with an equal sex distribution. Only the nuchal crest exhibited significant differences in score distributions among the females in the sample, indicating differences between individuals younger than 40 years and those older than 40 years. Classification models showed a slight increase in accuracy for younger individuals when age-specific standards were applied, while accuracy for older individuals decreased slightly. However, the difference in accuracy with age-specific standards was comparable to accuracies obtained with all ages pooled. The results suggest age did not have as substantial an impact on trait expression as population affinity, and prior knowledge of age did not significantly influence the accuracy of the method. Thus, the pre-selection of age before the estimation of sex using the Walker traits is not currently necessary when conducting skeletal analyses in a contemporary South African population. However, the use of population-specific standards is paramount to more accurate classification.

The online version contains supplementary material available at 10.1007/s00414-025-03568-1.