TIMOTHY G. LOHMAN
Recent developments in the field of body composition assessment include health-related standards based on both body mass index (BMI) and percentage of body fat; increased prevalence of obesity in the U.S. population based on BMI; multicomponent models for more accurate assessment of body composition; validation studies of dual-energy X-ray absorptiometry (DXA) as a reference method for body fat and lean as well as bone density; increased use of body composition methods in national surveys, multicenter research projects, and epidemiological investigations; search for indirect estimates of abdominal and visceral fat components and their relationship to health; and assessment of body composition changes with short- and long-term interventions. In this chapter, I review these developments to describe the advantages and limitations of various body composition models and methods.
The development of multicomponent models has enabled more accurate assessment of body composition, particularly in populations where the two-component model does not apply, for example, children, elderly, and certain athletic groups. BMI as a prediction of body fat has a relatively large error (standard error of estimate, or SEE) for the individual (4–6%). Use of field methods such as bioelectric impedance, skinfolds, and circumferences have errors between 3 and 4.5% depending on the investigation and population. Laboratory methods such as underwater weighing, total body water, and DXA have errors between 2.5% and 3.5%. Multicomponent models can estimate percent fat with 1.0–3.0% error. The selection of a body composition method is dependent in part on the accuracy desired, the equation available, as well as the setting, sample size, and population under study (Figure 11.1). BMI works better with