Introduction
Anatomists demonstrated many years ago that organs grow at different rates, and that these rates can differ from the growth rate of the body as a whole (Forbes, 1978). Further, auxologists have shown that children grow at a variety of rates: they can play lento or allegro (Tanner, 1986). The first case corresponds to a normal process where growth is organised in successive steps, while in the second case individual variation due to genetic and/or environmental factors influences the growth process. This variation makes it difficult to predict adult body composition from childhood measurements. However, growth is affected by hormonal status, so that childhood is a good time to study the relationship between anthropometry and hormonal status, and to analyse the influence of environmental factors such as nutrition. In general, patterns of growth give more useful information than absolute levels of anthropometric measurements. Better understanding of factors influencing body composition can improve prediction of adult status and help to propose strategies for reducing the risk factors of various diseases.
Use of anthropometric measurements
Anthropometric measurements can be used in several ways: directly (e.g. skinfolds), as indices (e.g. weight/height2, the Quetelet or body mass index (BMI)), areas (e.g. upper arm muscle area (UMA) based on arm skinfolds and arm circumference) or in regression equations relating body density to anthropometric measurements for a reference population. In addition, various ratios can be used to predict body shape and proportion.
Direct measurements and the BMI predict the level of fatness, while UMA and the regression equations predict body composition (i.e. fat mass (FM), fat-free mass (FFM) and % body fat (%BF)).