# Is chronological age predictable on panoramic images using the Demirjian method on third molars? Insights from a Turkish population-based study

## TL;DR

This study shows that the modified Demirjian method can accurately estimate age in Turkish adolescents using third molar development stages from panoramic X-rays.

## Contribution

The study validates the modified Demirjian method for age estimation in a Turkish population and identifies interaction effects between gender, impaction, and tooth development.

## Key findings

- Third molar development stages strongly correlate with chronological age in Turkish adolescents.
- Generalized linear models confirmed age as a significant predictor of tooth development stages.
- Interactions between gender and impaction classifications influence developmental patterns.

## Abstract

Third molars are crucial in forensic age estimation during adolescence due to their late and variable development. The modified Demirjian method employs a nine-stage classification (0-H) for assessing dental maturity. However, population-specific validation is essential for forensic accuracy. This study aims to evaluate the relationship between chronological age and M3 development stages in a Turkish population using panoramic radiographs, and to investigate the influence of gender and anatomical location, including impaction classifications (Pell-Gregory and Winter).

A total of 1656 individuals (mean age: 13.9±3.78 years) underwent digital panoramic radiography. Developmental stages of 2770 mandibular and 2553 maxillary M3s were assessed using the modified Demirjian method. Statistical analyses included Spearman correlation, Kruskal-Wallis tests and generalized linear models.

Third molar development strongly correlated with chronological age. Most teeth were in stages B-E, with mandibular M3s most frequent in stages C and E, and maxillary in C and D. Age differed significantly across all stages (p<0.001). Generalized linear models confirmed age as a significant predictor (R²=0.330 mandible; R²=0.337 maxilla). Gender and impaction classifications had no main effects, but significant interactions were found (Pell-Gregory×Gender: p<0.001; Winter×Pell-Gregory: p=0.016).

The modified Demirjian method reliably estimates age in Turkish adolescents, with high age-stage correlation. While gender and anatomical location alone do not alter developmental progression, their interaction with impaction patterns reveals complex growth dynamics. These findings support context-sensitive forensic protocols.