Sex estimation using anthropometry of feet and footprints in a Western Australian population

• Published in: Forensic science international 2013-09, Vol.231 (1), p.402.e1-402.e6
• Main Authors: Hemy, Naomi, Flavel, Ambika, Ishak, Nur-Intaniah, Franklin, Daniel
• Format: Article
• Language: English
• Subjects: Accuracy; Adolescent; Adult; African Americans; Aged; Analysis of Variance; Anthropometry; Anthropometry - instrumentation; Anthropometry - methods; Australia; Discriminant Analysis; Feet; Female; Females; Foot - anatomy & histology; Footprints; Forensic Anthropology; Forensic engineering; Forensic sciences; Genealogy; Humans; Male; Males; Methods; Middle Aged; Pathology; Population; Population standards; Sample variance; Sex; Sex Determination by Skeleton - methods; Sexual dimorphism; Statistics; Studies; Young Adult
• ISSN: 0379-0738; 1872-6283
• DOI: 10.1016/j.forsciint.2013.05.029

Abstract An important component of forensic investigation is the identification of deceased (and increasingly living) individuals, which is often the role of the forensic anthropologist. One of the most valuable steps towards identification is via a biological profile, developed through the application of population specific standards. In disaster victim identification scenarios, fleshed feet are often recovered in footwear; footprints are another potential source of trace evidence found at crime scenes. In medico-legal investigations, feet and footprints can be useful for extrapolating living height, it is thus expedient to determine whether sex can be estimated from the same anthropometric data. The aim of the present study is to develop accurate sex estimation standards for a contemporary Western Australian population from measurements of the feet and footprints. The sample comprises 200 adults (90 males, 110 females). Three bilateral linear measurements were taken from each foot and seven bilateral measurements were acquired from static footprints obtained using a Podograph. A precision test was first conducted to assess data accuracy and reliability. Measurement data are then analysed using a range of parametric statistical tests. Results show that males were significantly ( P < 0.001) larger than females for all foot and footprint measurements; cross-validated sex classification accuracies ranged from 71% to 91%. Although in many instances the sex bias was large (>±5%), this study provides viable alternatives for estimating sex in Western Australian individuals with accuracy equivalent to established standards developed from foot bones.