What is measured and what is estimated in body composition? A clarification framework for terminology and...

Body composition assessment has traditionally focused on estimating body mass components and the estimation processes required to obtain them in vivo. Within this context, the five-level model and the methodological taxonomy proposed by Wang and colleagues provided a rigorous conceptual framework that has strongly shaped the field. However, contemporary research and practice reveal persistent ambiguity in the use of terms such as direct, indirect, and doubly indirect. This ambiguity has become increasingly evident as methods such as anthropometry and bioelectrical impedance analysis are widely used not only to estimate body mass components, but also to obtain whole-body variables that are directly measured and interpreted as clinically, physiologically, or performance-relevant descriptors. Confusion often arises when directly measured outputs (e.g., skinfolds, girths, phase angle) are not clearly distinguished from model-derived estimates of body composition (e.g., fat mass, fat-free mass). Accordingly, we propose a clarification framework for in vivo body composition assessment that distinguishes: (1) directly measured whole-body variables, (2) model-derived estimates of body composition, and (3) hybrid assessment pathways combining measured and inferred outputs. In this framework, classification depends on the specific variable of interest and the estimation process linking measurement to output, rather than on assigning a fixed label to the instrument itself. The term direct refers only to an output that is measured rather than model-derived, and does not imply direct assessment of body components, greater validity, or superior methodological accuracy. The proposed framework provides a practical conceptual tool for distinguishing measured from estimated body composition outputs, thereby improving methodological transparency, interpretation, and reporting consistency across research and clinical practice.