Estimation of sex based on pelvic girdle measurements in a Jordanian population: a computed tomographic study

Sex estimation from skeletal measurements that display sexual dimorphism is fundamental in forensic anthropology. Our aim was to establish population-specific discriminant equations for predicting sex from pelvic girdle measurements in a contemporary Jordanian population using 3D multidetector computed tomography (MDCT). A total of 302 pelvic MDCT scans (144 males and 158 females) were used and a total of 16 linear measurements and two angles were studied. Discriminant function analysis (DFA) was used to evaluate the possibility of sex categorization. All measurements were sexually dimorphic, except pubic length and maximum breadth of sacral alae (p >= 0.05). The multivariate analysis yielded an accuracy rate of 97.4% and a sex bias of -0.3%. Subpubic angle, angle of greater sciatic notch, innominate height, transverse pelvic inlet and ischial length were selected by stepwise analysis as best discriminators of sex, with an accuracy rate of 98.0% and a sex bias of -0.2%. Subpubic angle was the best single pelvic parameter in sex estimation with an accuracy of 88.1%. Lower classification accuracies and higher sex bias were obtained using univariate analysis. Considering multiple pelvic girdle measurements simultaneously, CT scans of living Jordanians provides a reliable source of osteometric data from which DFA equations can be derived.