The 81st Annual Meeting of the American Association of Physical Anthropologists (2012)

Body mass prediction: the accuracy of estimating overweight individuals


Health Sciences Library and Informatics Center, University of New Mexico

Friday All day, Plaza Level Add to calendar

Body mass estimation is a vital part of many investigations within biological anthropology. Methods to estimate body mass are based on individuals that fall within a normal range of weight (BMI=18-24.9). Predictive equations are then utilized on all individuals, regardless of BMI range, in order to estimate body mass. This study seeks to ascertain how accurate four current methods of estimating body mass are for a sample of documented overweight individuals. For this research, 18 measurements were recorded on 174 skeletons (n=128 males and 46 females) from four documented collections (Hamann-Todd, Robert Terry, William Bass, and Maxwell Museum).

The predictive methods evaluated for this study are bi-iliac breadth and height (Ruff), subtrochanteric area (McHenry), femoral head dimension (Grine and colleagues), and the multiple element model (Daneshvari). A paired t-test was performed to evaluate the accuracy of estimating body mass on overweight individuals (BMI ranging 25-29.9), by sex. All predictive estimates are significantly different from the recorded weight (p-values<0.0001). All masses were underestimated and males had a greater difference between estimated and actual body mass than females. For estimating male body mass the method by Grine and colleagues produced the least average error (8.7 kg), while for females it was two methods, that by Daneshvari (10.18 kg) and Grine and colleagues(11.24 kg). Additionally, after a mass of 70 kg is attained, all methods cannot detect additional mass from skeletal measures suggesting that at a certain point the skeleton stops responding to the stressors from weight.

comments powered by Disqus