Morphometric sexual dimorphism of human sternum in a north Indian autopsy sample: sexing efficacy of different statistical techniques and a comparison with other sexing methods

Abstract

Forensic anthropologists have proposed different sex determination methods based on metric and non-metric skeletal features of various bones. The main objective of the present investigation was to evaluate sexual dimorphism in human sternum and make comparative analysis based on studies by several workers who used different methodologies and techniques. For this study, 343 sternums were collected and examined from 252 male and 91 female cadavers of a north Indian population. Significant differences were noticed between the mean sternal measurements measured directly on bone specimens in present study and those measured indirectly on radiographs or CT scan images reported by previous researchers. Comparative analysis of absolute sternal measurements showed concordance with some earlier Indian studies but discordance with most other population studies conducted worldwide. Limiting points gave much higher sex accuracy levels than the demarking points. The calculated multivariate DFA equation (y=-14.222+0.076MBL+0.137WS_1) and regression equation (y=36.675-0.198MBL-0.405WS_1) correctly estimated sex of 84.8% and 89.8% sternums, respectively. Multivariate analyses gave more striking results than the univariate ones. Though logistic regression analysis gave higher accuracy rates but discriminant function analysis was found to be a more reliable statistical tool for sex determination because of low sex biases. The sexing efficacies of the dry and wet sternums did not differ significantly, so bone specimens are equally reliable as the radiographs or CT scans for sex determination. The sternal area and the combined length were found the best parameters among all the criteria and methods considered. The changes in the body proportions, nutritional, environmental, climatic changes, etc., are attributable to secular changes in sterna measurements when compared to Jit et al. [13]. The osteometric techniques are better representative of biological identity of the deceased because of some inherent defects associated with radiographic methods. Like other sex-predictive bones of the human skeleton, the sternum can also be a valuable adjunct for sex determination of unknown subjects.