Biostatistics Research Today is a free monthly online journal that collates and summarizes the latest research about Biostatistics, including details on statistics, uncertainty, probability, modeling.

Relationship between adiposity and body size reveals limitations of BMI.

Nevill AM, Stewart AD, Olds T, Holder R

Research Institute of Healthcare Sciences, University of Wolverhampton, WS1 3BD Walsall, UK. a.m.nevill@wlv.ac.uk

The aims of this study were to assess 1) whether the stature-adjusted body mass index (BMI) is a valid proxy for adiposity across both athletic and nonathletic populations, and 2) whether skinfold measurements increase in proportion to body size, thus obeying the principle of geometric similarity. The research design was cross-sectional, allowing the relationship between skinfold calliper readings (at eight sites and between specific athletic and nonathletic groups, n = 478) and body size (either mass, stature, or both) to be explored both collectively, using proportional allometric MANCOVA, and individually (for each site) with follow-up ANCOVAs. Skinfolds increase at a much greater rate relative to body mass than that assumed by geometric similarity, but taller subjects had less rather than more adiposity, calling into question the use of the traditional skinfold-stature adjustment, 170.18/stature. The best body-size index reflective of skinfold measurements was a stature-adjusted body mass index similar to the BMI. However, sporting differences in skinfold thickness persisted, having controlled for differences in body size (approximate BMI) and age, with male strength- and speed-trained athletes having significantly lower skinfolds (32% and 23%, respectively) compared with controls. Similarly, female strength athletes had 29% lower skinfold measurements compared to controls, having controlled for body size and age. These results cast serious doubts on the validity of BMI to represent adiposity accurately and its ability to differentiate between populations. These findings suggest a more valid (less biased) assessment of fatness will be obtained using surface anthropometry such as skinfolds taken by experienced practitioners following established procedures.

Published 12 December 2005 in Am J Phys Anthropol, 129(1): 151-6.
Full-text of this article is available online (may require subscription).

© 2005-2008 Biostatistics Research Today. All Rights Reserved.