Sex determination using discriminant function analysis in children and adolescents: a lateral cephalometric study

Abstract

The objective of this study is to test the validity of sex determination in children and adolescents using lateral radiographic cephalometry and discriminant function analysis. Fifty male and 50 female cephalograms of Taiwanese children were used (males and females with mean age of 15.52 +/- 1.38 and 15.67 +/- 1.54 years, respectively). Twenty-two cephalometric measurements were performed using computerized cephalometry. Statistical analysis shows that all measurements were sexually dimorphic (p < 0.05). Nine measurements, statistically validated and clinically relevant, were used for discriminant function analysis. A stepwise discriminant procedure selected seven of the nine variables, producing 95% accuracy in sex determination. Resubstitution classification reveals the same discriminant rate. Cross-validation classification (the leave-one-out method) reveals that the correct sex determination rate is 91%. However, the combination of four variables using both the stepwise procedure and the resubstitution method achieves a 92% accuracy rate. A cross-validation classification procedure with the same four variables resulted in a 91% accuracy rate. Therefore, this study uses four cephalometric measurements as the minimum number of traits yielding the maximum discriminant effectiveness of sex determination in children and adolescents.