Predictive models for sex Estimation based on alveolar measurements in a contemporary Italian sample

Teeth play a significant role in reconstructive identification, aiding in the estimation of an individual's biological profile, such as population affinity, sex, and age at death. However, when teeth are lost postmortem-a common occurrence in forensic contexts-alternative methods are required. This study aimed to develop predictive models for sex estimation using alveolar measurements and to evaluate their reliability with a validation sample. Mesiodistal and buccolingual alveolar diameters from both arches were measured in 127 adults (64 males, 63 females) from the Certosa Cemetery osteological collection in Bologna. Binary logistic regression models were developed using a training subsample and tested on a hold-out subsample. The results indicated that the canine alveoli in both arches exhibited the highest sexual dimorphism, followed by those of the central incisor, first premolar, lateral incisor, and second premolar. In the training subsample, univariate predictive models achieved sex classification accuracy ranging from 57.3 to 83.9%, while multivariate models ranged from 68.0 to 91.3%. In the hold-out subsample of 19 individuals, sex was accurately estimated in 89.5% of cases, with a match rate of 63.2% compared with known biological sex. This study highlights the potential of alveolar measurements as a reliable alternative in forensic anthropology, offering a viable method for sex estimation when teeth are unavailable for measurement.