Morphometric study of the facial skeleton in Jordanians: A computed tomography scan-based study

Several studies have shown variability in osteometric measurements of the skull between populations. Therefore, each population should have specific standards to optimize the accuracy of identification. The aim of this study was to evaluate the sexual dimorphism in metric relations between anatomical points of the skull using CT scans in a Jordanian population by means of discriminant function analysis, to determine which continuous variables discriminate between sexes, and to examine the craniofacial changes according to age. 500 CT scans (240 males and 260 females) were used and a total of 11 craniofacial parameters were studied. The data were analyzed using distriminant function analysis. Sexual dimorphism was found mainly in the young adult group. Significant age related changes were noticed in minimum frontal breadth, orbital height and orbital index. In all subjects, the analysis of multivariate (dimorphic variables) and stepwise functions gave an accuracy of 58.8% and 57.0% respectively. Using stepwise analysis, the most dimorphic variables to estimate sex were maximum frontal breadth, bimaxillary breadth and orbital index. The multivariate analysis of all variables gave an accuracy of 58.8%. The percentages of correct sexing in Jordanian population were high using the single variable analysis in females only. In females, classification accuracies of 70.4% using bimaxillary breadth, 70.0% using the orbital index, and 68.1% using maximum frontal breadth were obtained. Additionally, discriminant analysis was conducted separately for each age group. Higher percentages of correct sexing were obtained only in young adults. The results of the discriminant function analysis did not appear to be significant. The results of discriminant function were insignificant in middle aged and elderly groups. Being female and young increases the percentages of correct sexing. Generally, low levels of accuracy of sexual dimorphism were obtained in our study, suggesting that population estimates of dimorphism are highly variable and the equations derived from the discriminant function analysis are not reliable in sex estimation in Jordanians.