Tooth measurements determined adolescents' sex: 100% in girls, 97.8% in boys

The upper first molar is the strongest sex marker

In estimating sex in adolescents, skeletal indicators are often not enough. A group led by Hadwe tested on an Egyptian population how far one can get with measurements of the teeth and the dental arch alone.

A total of 150 plaster models of individuals aged 12 to 19 were digitised with a high-resolution three-dimensional scanner. In the Facad software, the mesiodistal and buccolingual dimensions of the permanent first molars, the mesiodistal widths of the upper canines, and the intercanine, interpremolar and intermolar arch widths were measured. On those data a model for sex estimation was built by discriminant analysis, and its predictive value was then tested.

All measured parameters were significantly larger in male subjects (p < 0.05). The strongest sexual dimorphism was shown by the upper first molar, in the buccolingual (26.95%) and the mesiodistal dimension (25.98%). The model built on those measurements correctly classified 97.8% of male and 100% of female subjects. The authors stress that the procedure is tied to the population on which it was developed and call for multicentre external validation before wider use. They present the model as a supplement to clinical assessment and forensic identification where skeletal indicators are unavailable or ambiguous.

Limitations

The model was developed and tested on a single population and on 150 models. High accuracy on the sample from which a model is derived regularly falls on independent validation, and transfer to other populations has not been tested.

For your practice

• The practical message is not forensic but documentary: three-dimensional scans of models retain the measurability that a plaster model loses through breakage or disposal.
• Standardised measurement of arch widths is a by-product of the orthodontic records you already keep, so it only needs to be stored in a form from which measurements can later be taken.
• Population specificity works the other way round as well, so normative values from foreign samples do not transfer automatically to Croatian patients.