Body Composition1

Body Composition1

Scott Going

Melanie Hingle

Joshua Farr

A person’s composition reflects his or her net lifetime accumulation of nutrients and other substrates acquired from the environment and retained in the body. These components, ranging from elements to tissues and organs, are the building blocks that give mass and shape and confer function to all living things. Body composition assessment methods allow scientists to describe how these components function and change with age, growth, and metabolic state. Clinicians rely on body composition measurements for diagnosis, judging disease risk, and determining efficacy of therapies to improve clinical outcomes. Serial body composition measurements are a reliable indicator of nutritional recovery from uncomplicated malnutrition or illness. Simple anthropometric measurements such as height (HT), weight (WT), and body mass index (BMI), as well as percent fat or lean mass, can be used to assess an individual’s status against a standard, or relative to that person’s “usual” over a specified period of time. These simple measures allow early detection of nutrient deficiencies or inadequate nutrient intake so that nutritional status can be improved through an individualized nutrition plan before disease occurs.

There is considerable interest in defining normal changes in human body composition during growth, maturation, and senescence. Defining normal is vital to understanding abnormal, which is associated with disease. This proposition is challenging, given the large variation that occurs within and among healthy individuals and the difficulty separating age-related from disease-related changes in older persons. Typically, descriptions of normal age trajectories have been based on a composite built on data from multiple studies that are usually cross-sectional, employ different methods, and are not population based (1). Few large-scale, population-based studies have been conducted to describe normal because of the cost and complexity of accurate body composition methods. Some reference data have been developed using anthropometry, dual-energy x-ray absorptiometry (DXA), and bioelectric impedance measurements obtained in the National Health and Nutrition Examination Survey (NHANES) (2). The anthropometric data have been used to describe age trajectories in the measured variables (e.g., HT, WT, and skinfolds) and estimates of composition. Chumlea et al (3) published reference data for body composition predicted using bioelectric impedance analysis (BIA) and Janssen et al (4) used BIA to develop reference data for predicted skeletal muscle (SM) mass. Laurson et al (5) developed percent body fat (PBF) growth curves for children and adolescents based on the NHANES III and IV surveys. These age trajectories should prove useful for defining typical changes in fatness in US boys and girls, although they are based on indirect, not direct, estimates of composition. Standards for adults are not well established.


Five-Level Model

Approximately 50 elements in the body are organized into 100,000 chemical compounds, approximately 200 cell types, and 4 main tissues. The central model underlying body composition assessment is the five-level model (Table 48.1), in which body mass is considered as the sum of all components at atomic, molecular, cellular, tissue or organ, and whole body levels (6). Methods are available for measurement of components on each level, and the levels are interrelated so that components on one level can be used to estimate components on another level. Certain rules, reflecting these relationships, are inherent in the five-level model, and ultimately the accuracy of assessments depends on the validity of these rules.

Atomic Level

Body mass is composed of 11 major elements. Four of them—oxygen, carbon, hydrogen, and nitrogen—make up greater than 96% of body mass. The major elements are linked to higher level components. Other important elements are calcium, potassium, phosphorous, sulfur, sodium, chlorine, and magnesium. Most of these elements can be estimated in vivo by neutron activation analysis or whole body counting (7), research methods that are not widely used in clinical practice but are useful for establishing models underlying simpler methods. Total body carbon, total body nitrogen (TBN), and total body potassium (TBK) can be used with appropriate models to derive total body fat (8), protein (8), and body cell mass (BCM) (9), although other approaches for estimating these components are more practical and widespread.








WT = 0 + C + H + N + Ca + P + K + S + Na + Cl + Mg



WT = F + W + P + Ms + Mo + G


WT = F + W + P + M


F = 2.747/BD – 0.714 (W) + 1.146 (Mo) – 2.0503



F = 2.75/BD – 0.714 (W) + 1.148 (M) – 2.05


WT = F + W + solids


F = 2.118/BD – 0.78 (W) – 1.354

WT = F + Mo + residual


F = 6.386/BD + 3.961 (M) – 6.09


WT = F + FFM


F = 4.95/BD – 4.50



F = WT – TBK/2.66 (males);

F = WT – TBK/2.51 (females)


F = WT – TBW/0.73






BCM = 0.00833 × TBK

ECS = TBCa/0.177

ECF = (0.9 × TBC/plasma Cl)


WT = AT + SM + bone + blood + others


WT = Head + neck + trunk + lower + extremities + upper extremities

AT, adipose tissue; BCM, body cell mass; BD, body density; C, carbon; Ca, calcium; Cl, chloride; CM, cell mass; ECF, extracellular fluid; ECS, extracellular solids; F, fat; FC, fat cells; FFM, fat free mass; G, glycogen; H, hydrogen; K, potassium; M, mineral as a fraction of WT; Mg, magnesium; Mo, osseous mineral as a fraction of weight; Ms, cell mineral as a fraction of weight; N, nitrogen; Na, sodium; P, phosphorus; S, sulphur; SM, skeletal muscle; TBCa, total body calcium; TBK, total body potassium; TBW, total body water; W, water as a fraction of weight; WT, weight.

Molecular Level

The molecular level consists of six major components: water, lipid, protein, carbohydrates, bone minerals (BMs), and soft tissue minerals. Models having from two to six components can be created. The two-component, fat mass (FM) and fat-free mass (FFM) model, in which all nonlipid components are combined in FM, is most common. FFM is the actively metabolizing component and is often used as the reference for metabolic or functional indexes. Models with more than two compartments are called multicompartment models. These models divide the FFM into additional components that can be quantified in vivo. These models are used to minimize errors related to assumptions underlying the two-component model. In many situations, two-component models are not valid, such as in children, the elderly, and sick and infirm persons. Relying on fewer assumptions by measuring more components improves validity and accuracy, although they are more expensive, more burdensome, and the potential greater accuracy can be offset by greater measurement error if individual components are not measured accurately (10).

Cellular Level

Conceptually, the cellular level provides for multiple models based on different cell types. In practice, the most common model includes three components: extracellular
solids, extracellular fluid, and cells. The cellular mass can be divided further into two components, fat and BCM. BCM is the actively metabolizing component at the cellular level (11). The terms fat and lipid are often used interchangeably, although their meanings differ. In body composition assessment, lipid includes all of the biologic matter extracted with lipid solvents. These extracted lipids include triglycerides, phospholipids, and structural lipids that occur in small quantities in vivo (12). In contrast, fats refer to the specific family of lipids consisting of triglycerides (6). Based on reference man (13), approximately 90% of extractable lipids in healthy adults is triglyceride, although this proportion differs with dietary intake and illness (14). The remainder, approximately 10% of the total body lipid (nonfat lipid), are mainly composed of glycerophosphatides and sphingolipids.

Tissue-Organ Level

The major components on the tissue-organ level include adipose tissue (AT), SM, visceral organs, and bone. Some tissue-organ level components are single solid organs such as brain, heart, liver, and spleen. Others, such as SM and AT, are interspersed throughout the body. In common usage, fat and AT are often interchanged although they are distinct and on different levels, and the difference is important when measuring their mass and metabolic characteristics. Although fat is found primarily in AT, intracellular triglyceride pools are found in the liver, SM, and other organs, particularly in conditions such as hepatic steatosis and various forms of lipidosis. There are also small circulating extracellular pools of triglycerides, mainly as lipoproteins. AT consists of adipocytes, extracellular fluid, nerves, and blood vessels. AT compartments are distributed throughout the body, and their metabolic properties differ depending on location (15). AT compartments are closely linked with disease risk. Visceral AT (VAT) and its association with metabolic dysregulation and cardiovascular disease are perhaps the best studied, although ectopic fat in intramuscular and perivascular depots has also been linked to disease risk (15).

Whole Body Level

On the whole body level, composition is divided into regions such as appendages, trunk, and head. Rather than discrete components, trunk, and appendages are usually described by anthropometric measures such as circumferences, skeletal lengths, breadths, and skinfold thicknesses (16). Other whole body measures include body WT, volume, density, and electrical impedance. Anthropometric indexes have a long history of use as surrogates for body composition. Waist circumference, for example, has been used to predict obesity-related morbidity and mortality (17). Upper arm circumference, especially when corrected for subcutaneous AT, is a common index of nutritional status. Estimation of components on other levels (e.g., FM and FFM) is another common use of measures made on the whole body level.

The remainder of this chapter emphasizes description of the major components on the chemical, cellular, and tissue-organ levels, specifically body fat (or AT) and its anatomic distribution, and FFM, its main constituents, (BCM, water, SM, and bone), and the predominate methods used to measure them. These compartments have direct health and functional implications and some are used to index nutrient and energy needs. A comprehensive survey of other methods has been published in the literature (18). Anthropometric methods are discussed elsewhere in this volume.

Steady State

An important concept underlying body composition assessment is the notion that when body mass and energy stores are stable, the major components remain stable and thus maintain predictable interrelationships. Although components on the five levels are distinct, they are related and can be used to estimate components on the same and other levels. For example, assuming a constant ratio of total body protein (TBP) to TBN (TBP/TBN = 6.25), TBN (elemental level) can be used to estimate protein (chemical level). Similarly, BCM (cellular level) can be estimated from TBK (BCM = 0.00823 × TBK) and SM (tissue level) can be estimated from both TBK and TBN (SM = 0.0196 × TBK – 0.0261 × TBN). The premise of stable conversion factors used to estimate one component from another and the validity and accuracy of any method depends on the degree of departure from steady state.



Skeletal size is a determinant of HT (19), which is correlated with FFM, the actively metabolizing cellular component and an important factor in estimating energy requirements. In adults, HT has been used to estimate ideal body WT (IBW) (20), which can be used to provide an estimate of daily nutrient needs to maintain a healthy WT for HT. Although body composition methods are needed to provide a precise estimate of metabolically active tissue, estimates such as these may be used to quickly calculate a reasonably accurate estimate of IBW in the field.

Body Weight

Body WT is used as an indirect measure of nutritional status because it is representative of body energy stores. Because of the tight regulation of carbohydrate and protein oxidation rates, any long-term changes in WT are assumed to reflect proportional changes in body fat stores. IBW is useful in establishing nutrient intake guidelines and setting parameters for a healthy WT range; however, an individual’s usual body WT (UBW) (rather than IBW) may provide additional information useful for evaluating an individual’s nutritional status. The difference between
current and UBW or IBW may be compared against clinical parameters to determine risk of morbidity and mortality. Body WT typically varies less than ±0.1 kg/day in healthy adults. WT loss of more than 0.5 kg/day indicates negative energy or negative water balance or both. A clinically significant rate of WT loss is considered to be 1% to 2% over 1 week, 5% over 1 month, 7.5% over 3 months, or 10% or greater over 6 months (21). Severity of WT loss may also be evaluated by the absolute WT reduction, which also has prognostic value. An absolute WT of 85% to 95% of UBW (or 80% to 90% of IBW) indicates mild malnutrition, 75% to 84% of UBW (or 70% to 79% of IBW) indicates moderate malnutrition, and 75% or less of UBW (or ≤69% of IBW) indicates severe malnutrition (21). Absolute WT reduction to less than 55% to 60% of IBW places an individual at the limits of starvation (22). In infirm individuals, a WT loss between 10% and 20% of pre-illness WT over 6 months has been associated with functional abnormalities (23), whereas a loss of more than 20% of pre-illness WT suggests significant protein-energy malnutrition (23). The minimum survivable body WT in humans is between 48% and 55% of IBW or a BMI of approximately 13 kg/m2.

Overconsumption of energy relative to requirements results in a positive energy balance which, if sustained, leads to WT gain and excess adiposity. Excess adiposity is associated with increased risk of morbidity and early mortality, because AT not only functions as a storage depot for excess energy, but also significantly influences endocrine function and metabolic and immune regulation. The maximum survivable body WT is approximately 500 kg (a BMI of ˜150 kg/m2) (24).

When using WT as an estimate of energy and protein needs, clinicians must consider factors that affect WT fluctuations or otherwise confound the assumption that WT is a surrogate of energy stores, such as rapid fluid shifts (intracellular to extracellular or intravascular to extravascular spaces) and accumulation of fluid secondary to inflammation. Edema and ascites and the medications used to treat them may cause fluid to shift to extracellular spaces, masking body composition changes and artificially increasing WT. Tumor growth or abnormal organ enlargement in disease states may cause an increase in WT and mask tissue loss (i.e., loss of fat or FFM). Morbidly obese individuals experiencing rapid, intentional WT loss may be at nutritional (and health) risk as WT (including lean mass and FM) decreases as a result of protein-calorie malnutrition and semistarvation. Finally, physical activity- or diet-induced changes in energy intake and expenditure affect glycogen mass (and its bound water) and body sodium, which is associated with fluid readjustment and WT fluctuations.


BMI (kg/m2)a


MEN ≤102 cm (≤40 in) WOMEN ≤88 cm (≤35 in)

MEN >102 cm (>40 in) WOMEN >88 cm (>35 in)+


Grade III protein-energy malnutrition


Grade II protein-energy malnutrition


Underweight (grade I protein-energy malnutrition)








Class I obese


Very high


Class II obese

Very high

Very high


Class III severe obesity

Extremely high

Extremely high

BMI, body mass index.

a BMI cutoff points represent World Health Organization standard for international classification, although cutoff points of 23, 27.5, 32.5, and 37.5 kg/m2 have been suggested for Asian populations as points for public health action (26).

Adapted from the National Heart, Lung, and Blood Institute. Guidelines on Overweight and Obesity. Available at: obesity/e_txtbk/txgd/4142.htm, with permission.

Body Mass Index

Ratios of WT to HT (WT/HT ratios) have a long history in studies of body habitus. BMI (WT, kg/HT, m2) is the favored index because HT squared minimizes the relationship between HT and WT, at least in adults. Although not a direct measure of adiposity, BMI is a widely used surrogate for composition, based on the tenuous assumption that excess WT results from body fat. Although BMI and body fat are correlated, use of BMI as an “adiposity” index is confounded by differences in body proportions (e.g., trunk-to-leg length ratio), fat distribution, and composition relative to HT. Individuals with greater than average muscularity, for example, may be misclassified as overweight or obese, and elderly individuals may be considered normal WT obese (i.e., a normal WT despite muscle and bone loss because of added FM). In addition, composition and location of excess WT vary with gender, race, and age, information not captured by BMI (25). Despite these limitations, BMI predicts disease risk and standard definitions of overweight and obesity are in use (Table 48.2). Revised definitions have been proposed for Asians, who clearly have a different BMI-adiposity relationship (26).

Differential changes in fat and FFM in boys and girls confound interpretation of BMI. Consequently, genderspecific, BMI-for-age percentiles are used in children and youth. Revised BMI growth charts for US youth were constructed with data from the NHANES surveys conducted before the rapid rise in pediatric obesity (27). The charts provide practical tools for clinicians to compare the growth of a child against the reference population and make inferences about nutritional status and risk with regard to overweight and obesity (28). In boys and girls less than 18 years old, underweight, overweight, and obesity are defined as age- and gender-specific BMI less than 5th, greater than 85th to less than 95th, and greater than or equal to 95th percentiles, respectively (29).


FFM is a heterogeneous compartment on the chemical level of analysis. Its primary constituents of intracelluar and extracellular fluid, protein, and osseous and nonosseous minerals can be combined to form various models on which assessment methods are based (see Table 48.1). Historically, FFM has been estimated most commonly from body density (BD) estimated by underwater weighing (30), TBK estimated by whole body counting (7), and total body water (TBW) estimated by hydrometry (31). Each approach relies on a conversion factor based on an assumption of a constant relationship between the measured component and FFM. In healthy young adults, the assumption of chemical constancy introduces relatively little error. However, significant changes in FFM components with growth and maturation, aging, and illness are well described (1, 32, 33, 34) and introduce significant error unless appropriate adjustments are made. Gender and race or ethnicity differences are known (35) as well as effects of physical training (36). It is imperative that use of constants and equations be restricted to the groups for which they were developed unless their validity in another group has been shown. Alternatively, and especially when the condition of steady state is not met, application of multiple component models improves accuracy (37), although the requirement for more measures increases cost and patient burden and limits their use.




AGE (y)


MO/FFM (%)

DFFM (g/cc)


MO/FFM (%)

DFFM (g/cc)










































































DFFM, density of fat free mass; FFM, fat free mass; Mo, osseous mineral as a fraction of weight; TBW, total body water.

Data from Boileau et al (120), Fomon et al (55), Haschke et al (121, 122), Lohman et al (123, 124), with permission, with some modifications of the estimates of Fomon et al (55) to provide for linear changes in body water and bone mineral with age.


Historically, hydrodensitometry (underwater weighing) was used to estimate body volume (BV) and BD, which was converted to estimates of PBF and FFM (30). For young children, the elderly, and infirm, disabled, and other special populations, complete submersion in water is very difficult, if not impossible. An alternative approach, air displacement plethysmography (ADP), uses pressure-volume relationships to estimate BV and BD. The most recent form of ADP, the Bod Pod (COSMED USA, Inc. [formerly Life Measurements, Inc.], Concord, CA), provides a reliable means of determining BV (38, 39) and eliminates the need for submersion in water. The procedure can be performed by children and adults, although it does require a breathing maneuver to measure thoracic gas volume that may be difficult for young children and patients with pulmonary disease.

A major source of error in densitometry is the model used to convert BD to composition. In the classic twocomponent model, the densities of fat and FFM are assumed to be 0.9 and 1.1 g/mL, respectively. Using these densities, it is possible to derive an equation for estimating percent fat from BD (see Table 48.1). The density of FFM is derived from its primary constituents, water, protein, and mineral, as well as their respective fractions and densities (Table 48.3). The more closely the FFM components and their densities fit the individual being measured, the more accurate the result will be.

Many studies have demonstrated considerable variation in FFM composition and thus density attributed to growth and maturation (40), aging (41), and specialized
training (42). Sex and racial differences also exist, and even within a population, considerable interindividual variation (37) invalidates the assumption of FFM chemical constancy. Consequently, multicomponent models (three-component and four-component models; see Table 48.1), which require fewer assumptions because more components are measured, are more accurate than the two-component model. In children and in patients with edema, combining a measure of TBW along with BD significantly improves estimation of FFM; similarly, in elderly patients and in patients with significant bone loss, combining a measure of body mineral with BD gives a more accurate estimate of FFM. When a multicomponent model is not feasible, accuracy can be improved by using a population-specific equation that has been adjusted for the anticipated changes that occur with growth, maturation, and aging (Table 48.4).

Jul 27, 2016 | Posted by in PUBLIC HEALTH AND EPIDEMIOLOGY | Comments Off on Body Composition1

Full access? Get Clinical Tree

Get Clinical Tree app for offline access