Test of the auricular surface sex estimation method in fetuses and non-adults under 5 years old from the Lisbon and Granada Reference Collections.

Non-adult sex estimation is an active field of forensic inquiry as morphological variations between males and females are subtle, but observable, even from intrauterine development. The objectives of this study are threefold: to test the validity of the auricular surface method for sex estimation (Int J Osteoarchaeol 27:898-911, 2017) in fetuses and children under the age of 5 years old; to evaluate if health conditions, reported as the cause of death, influence its accuracy; and to detect possible secular trends in sexual dimorphism. One-hundred and ninety-seven skeletal individuals from the Lisbon and Granada Identified Collections were studied. Individuals were divided according to the hormonal peaks (< 0, 0-2, < 2, and 2.1-5 years old), cause, and year of death (before and after 1960). As in previous studies, two ratios (FI/CF and DE/AD) and two qualitative variables (OM and MRS) showed the highest frequencies of correct estimation (0.81-0.86). The correct sex allocations increased when the discriminant function (0.85) and logistic regression (0.86) were applied. Males of the age groups < 0 and 2.1-5 years were all correctly sexed by both formulae, and the same was observed for the female probabilities of adequate allocation. The cause and year of death were identified as variables without statistical significance. It is proposed that this method can be incorporated with confidence into the multifactorial laboratory protocols for non-adult sex estimation from skeletal remains.