Prediction of the length of the femur from its fragments in a Sri Lankan population

Abstract

ABSTRACTIn many forensic circumstances, the femur is fragmented due to decomposition and mutilation, thus necessitating development and application of regression equations to estimate the length of the femur from its fragments. In this study, regression equations were derived for the prediction of the femoral length, utilizing 11 standard femoral measurements (maximum length of femur, epicondylar breadth, vertical head diameter, transverse head diameter, vertical neck diameter, transverse neck diameter, neck length, subtrochanteric anterior–posterior diameter, subtrochanteric transverse diameter, mid-shaft anterior–posterior diameter and mid-shaft transverse diameter) obtained from 86 femora belonging to a contemporary Sri Lankan population. Univariate regression equations with moderate to high correlations (0.226 to 0.714) were formulated for the estimation of the maximum femoral length from the measurements of fragments of the femur. Of the single variables, the transverse neck diameter emerged as the best predictor (r = 0.714, SEE = 18.6596) of the maximum length of the femur. The multiple regression equations using a combination of variables yielded higher correlations (0.774–0.857) with lower error estimates (14.18–16.96). The regression equations presented in this study can be used to estimate the length of the femur of unknown, mutilated and dismembered human skeletal remains with considerable accuracy.