Relationships between different levels of body composition

Relationships between different levels of body composition

The five levels of body composition are interrelated. This means that information at one level can be translated to another level. This is important as it forms the basis of many techniques used to determine body composition.

Box 2.2

Adipose tissue is made of adipocytes, which are cells that store triglycerides in the form of small fat droplets. Adipose tissue con-tains about 80% triglycerides and some 1–2% protein (enzymes), and the remaining part is water plus electrolytes. During weight loss adipose tissue decreases: the actual fat loss will be about 80% of the actual weight loss.

 After determining the amount of calcium in the body by, for example, IVNAA (atomic level), the amount of bone can be calculated assuming that a certain amount of total body calcium is in the skeletal tissue. Determination of total body potassium (by ⁴⁰K or IVNAA) enables the assessment of the body cell mass, as most of the body potassium is known to be intracellular. Skinfold thickness measurements (total body level) enable the assessment of body fat (molecular level). Formulae used for these calcula-tions are component based, property based, or some-times a combination. Component-based formulae are based on fixed relationships between components. An example is the calculation of total body water from measured hydrogen: the chemical formula of water determines the factor. Property-based formulae are based on established statistical relationships between variables. An example is the prediction of body fat percentage (body composition parameter) from skinfold thickness (property) (Box 2.2). Property-based formulae tend to be population spe-cific, which limits the widespread application.

Most body composition techniques that are in use today are based on assumptions, often derived from carcass analyses or experimentally derived from observational studies. Violation of these assumptions leads to biased results, and some methods are more prone to bias than others. In the following short description of different methodologies, the most important assumptions are highlighted.