Selecting an appropriate reference sample for juvenile age estimation methods in a forensic context

The population on which forensic juvenile skeletal age estimation methods are applied has not been critically considered. Previous research suggests that child victims of homicide tend to be from socioeconomically disadvantaged contexts, and that these contexts impair growth. Thus, juvenile skeletal remains examined by forensic anthropologists may be short for age. Cadaver lengths were obtained from records of autopsies of 1256 individuals, aged birth to eighteen years at death, conducted between 2000 and 2015 in Australia, New Zealand, New Mexico, New York City, and Cuyahoga County. Growth status of the forensic population, represented by homicide victims, and general population, represented by accident victims, were compared using height for age Z-scores and independent sample t-tests. Cadaver lengths of the accident victims were evaluated against growth references using one sample t-tests to evaluate whether accident victims reflect the general population.Homicide victims are shorter for age than accident victims in samples from the United States, but not in Australia and New Zealand. Accident victims are more representative of the general population in Australia and New Zealand. Different results in Australia and New Zealand as opposed to the United States may be linked to higher socioeconomic inequality in the United States. These results suggest that physical anthropologists should critically select reference samples when devising forensic juvenile skeletal age estimation methods. Children examined in forensic investigations may be short for age, and thus methods developed on normal healthy children may yield inaccurate results.