The estimation of body mass from human skeletal remains
St. George, Karen R. Bottenfield
MetadataShow full item record
The ability to estimate body mass from human skeletal remains with a high degree of accuracy would be significant for the identification of identifying unknown individuals in a forensic anthropology context, documenting secular change in modern populations, and evaluating any prevalence in prehistoric populations. Modern research investigating body mass incorporates one of two models: morphometric and biomechanical. The morphometric model views the body as a cylinder, where weight estimates are gathered from extreme points such as the breadth of the pelvis. In contrast, the biomechanical model incorporates engineering principles and biology to understand the effects of mass on the human skeleton. Only the biomechanical model can accommodate extremes in body mass, such as those exhibited by modern populations. This study examined the accuracy of estimating body mass (obesity in particular) from human skeletal remains using a suite of traits shown to be significant in previous studies, including documented biomechanical analysis of obese individuals involving gait and sit-to-stand (STS) movements. It was hypothesized that using a combination of methods, body mass could be estimated with a high degree of accuracy. Using a large skeletal sample (n = 191), composed of male and females with documented age, weight, and height, the following three variables were examined: (1) the spinal manifestation of diffuse idiopathic skeletal hyperostosis (DISH), (2) osteoarthritis (OA) of the tibiae, and (3) external femoral dimensions. These were then subject to statistical tests. Spearman's rank-order correlation and Mann-Whitney U tests showed significant relationships between DISH and obesity in females (p<.05), but not for males. The presence and severity of OA of the medial condyles were also significantly related to BMI in females (p<.05). In males, the relationship between BMI and OA was only significant on the condyles of the right tibiae (p<.05). Finally, ANOVA and Pearson's product-moment correlation tests were performed to evaluate the cross-sectional dimensions of the femur. The effect of age, stature, and BMI were also examined. ANOVA results showed a significant effect between BMI and M-L cross-sectional dimensions among both sexes (p<.05). Initial Pearson's tests performed separately on males and females showed no significant correlations; however, after the sexes were pooled, small to moderate negative correlations between the M-L/A-P ratio along the diaphysis of the femur and BMI were found. Finally, multiple regression analyses were performed. The models for both sexes with all ten variables was statistically significant for BMI. The final accuracy rate was 78.48% for females and 84.37% for males. The primary goal of this study was to evaluate Moore's (2008) body mass estimation study. In this investigation, however, all dimensions of the femur were performed using an osteometric board and sliding calipers following the guidelines used by Agostini and Ross (2011). The results of this study paralleled many of the observations seen in previous studies, particularly the M-L lateral widening of the femur. Future research should continue to examine the relationship of DISH and OA with body mass, particularly regarding the varying manifestations between the sexes and confounding factors such as age.