Age prediction formulae from radiographic assessment of skeletal maturation at the knee in an Irish population

• Published in: Forensic science international 2014-01, Vol.234 (C), p.188.e1-188.e8
• Main Authors: O’Connor, Jean E, Coyle, Joseph, Bogue, Conor, Spence, Liam D, Last, Jason
• Format: Article
• Language: English
• Subjects: Adolescent; Age; Age Determination by Skeleton - methods; Age estimation; Child; Epiphyses - anatomy & histology; Epiphyses - diagnostic imaging; Epiphysis; Female; Femur - anatomy & histology; Femur - diagnostic imaging; Fibula - anatomy & histology; Fibula - diagnostic imaging; Forensic Anthropology; Forensic sciences; Humans; Ireland; Jurisdiction; Knee; Knee Joint - anatomy & histology; Knee Joint - diagnostic imaging; Linear Models; Male; Males; Osteogenesis; Pathology; Population; Radiographs; Regression analysis; Sex Characteristics; Skeletal maturation; Tibia - anatomy & histology; Tibia - diagnostic imaging; Young Adult
• ISSN: 0379-0738; 1872-6283
• DOI: 10.1016/j.forsciint.2013.10.032

Abstract Age estimation in living subjects is primarily achieved through assessment of a hand–wrist radiograph and comparison with a standard reference atlas. Recently, maturation of other regions of the skeleton has also been assessed in an attempt to refine the age estimates. The current study presents a method to predict bone age directly from the knee in a modern Irish sample. Ten maturity indicators (A–J) at the knee were examined from radiographs of 221 subjects (137 males; 84 females). Each indicator was assigned a maturity score. Scores for indicators A–G, H–J and A–J, respectively, were totalled to provide a cumulative maturity score for change in morphology of the epiphyses (AG), epiphyseal union (HJ) and the combination of both (AJ). Linear regression equations to predict age from the maturity scores (AG, HJ, AJ) were constructed for males and females. For males, equation-AJ demonstrated the greatest predictive capability ( R2 = 0.775) while for females equation-HJ had the strongest capacity for prediction ( R2 = 0.815). When equation-AJ for males and equation-HJ for females were applied to the current sample, the predicted age of 90% of subjects was within ±1.5 years of actual age for male subjects and within +2.0 to −1.9 years of actual age for female subjects. The regression formulae and associated charts represent the most contemporary method of age prediction currently available for an Irish population, and provide a further technique which can contribute to a multifactorial approach to age estimation in non-adults.