Our quest for knowledge regarding body composition and how it affects our propensity for disease and overall health has intensified in recent years, driven in large part by the desire to better understand health concerns and risk of disability associated with obesity (Goodpaster 2002). Indeed, research has focused not only on absolute measures of fat and fat-free mass but also on how the distribution of these affects our risk of conditions such as type 2 diabetes, hypertension, cardiovascular disease, stroke and cancer, to name a few.
The primary goal of assessing body composition is to determine the proportion of fat mass relative to lean body mass. Fat mass comprises essential fat and storage fat, the former being the fat necessary to sustain normal physiological function and the latter consisting primarily of adipose tissue. Lean body mass, on the other hand, includes several components, including muscle, water, bone, connective tissue and internal organs.
The most common laboratory methods for assessing body composition include hydrodensitometry (hydrostatic, or underwater, weighing); air displacement plethysmography (ADP), using a BOD POD®; isotope dilution for measuring total body water; and dual-energy x-ray absorptiometry (DEXA) (Wagner & Heyward 1999). Field techniques, which are usually simpler and less expensive, include skinfolds, anthropometric measurements (e.g., body mass index [BMI], waist circumference and waist-to-hip ratio), near-infrared interactance (NIR) and bioelectrical impedance analysis (BIA). Ultrasound, computed tomography (CT) and magnetic resonance imaging (MRI) are also available but are less common and generally impractical due to lack of availability, cost and technical complexity (Wagner & Heyward 1999). A comparison of some of these techniques is included in Table 1.
Hydrodensitometry, ADP and isotope dilution are all based on a two-compartment model that assumes that the body is made up of fat and fat-free mass, without consideration for the distribution of fat mass or the different components of lean body mass. DEXA, on the other hand, is a method that uses a three-compartment model, which has the advantage of being able to differentiate bone mineral, fat tissue and lean soft tissue when determining percent body fat. Four-compartment models combine data from several different methods in an effort to further discriminate between the quantity and distribution of fat mass, and the individual components of fat-free mass, including bone mineral, protein and water (Wagner & Heyward 1999; Davies & Cole 1995).
Regardless of the method used, body composition assessment has widespread applications, from optimizing performance in athletes to assessing metabolic abnormalities in the elderly. More specifically, body composition assessment can be used to do the following (Heyward & Stolarczyk 1996):
- Identify a client’s health risk associated with total body fat or excessive accumulation of intra-abdominal fat.
- Promote a client’s understanding of health risks associated with too little or too much body fat.
- Monitor changes associated with specific diseases that alter body composition.
- Assess the effectiveness of nutrition programs and exercise interventions.
- Estimate ideal body weight and formulate dietary recommendations and exercise prescriptions.
- Investigate the relationship between body composition and increased morbidity and mortality, and between body composition and decreased function in the elderly.
- Monitor growth, development, maturation and age-related changes in body composition, especially in children.
- Formulate interventions to prevent chronic diseases later in life.
- Optimize athletic performance and evaluate the effectiveness of training regimens for athletes.
The choice of a method to measure body composition in healthy adults depends on several factors, including cost, availability, ease of use and the ultimate goals of the client. The good news is that most prediction equations have been validated for healthy adults and, as a result, generally have a higher degree of accuracy for this population. Nonetheless, research has shown that even among healthy adults, body composition can vary significantly based on ethnicity, age and gender. As a result, researchers have focused on developing prediction equations that account for these variations.
Hydrostatic weighing has been considered the gold standard for many years. But because it is time-consuming and clients may find it difficult and uncomfortable, this method may not be practical for many people (Wagner & Heyward 1999). Hydrostatic weighing does have excellent precision for estimating body density in adults (Wagner & Heyward 1999). Nonetheless, it is important to realize that the estimation of percent body fat from body density is only as accurate as the conversion equation used.
ADP is regarded as a viable alternative to hydrostatic weighing for measuring body composition in healthy adults because it is rapid, safe and noninvasive, and its reliability appears to be relatively high when used for this population. In addition, the majority of individuals seem to favor ADP over hydrostatic weighing. In a survey distributed to participants in a recent study, 92% either preferred ADP or had no preference (Demerath et al. 2002).
Skinfold measurements are extremely common in fitness settings because they are relatively inexpensive and simple to perform. The error in estimating percent body fat using skinfold measurements reportedly ranges from ±3% to ±11%, with the reliability of the measurements highly dependent on the skill of the examiner (Wang et al. 2000). Testers can improve the accuracy of results by training with an experienced and skilled technician, practicing on clients of different body types, carefully identifying and marking skinfold sites and following standardized procedures. Additionally, because the conversion from skinfolds to body fatness is influenced by ethnicity, gender and age, it is especially important to use an appropriate prediction equation for the individual being tested.
Waist circumference measurements are also useful, especially in clinical settings where they are particularly helpful for evaluating risk of specific diseases. Waist circumference measurements are highly reproducible and correlate with body fat mass in males and females (Wang et al. 2003). However, as with skinfolds, the reproducibility at any site depends on the skill of the technician.
BIA is another common method used to assess body composition in healthy adults. It offers the advantages of being noninvasive and relatively easy to perform. However, because BIA is based on several assumptions regarding the human body that are inaccurate, and because it requires adherence to strict pretest guidelines, the accuracy of the results has been questioned (Kushner 1992). To improve accuracy, Slinde and colleagues (2003) recommend that BIA measurements be taken in a fasting state after 10 minutes in a supine position.
For children, body composition assessment is important, not only for evaluating current nutrition and health status, but also for evaluating the propensity for obesity and chronic diseases later in life. Interpreting body composition in children, however, poses unique challenges due to the complex changes in body composition during childhood and adolescence (Sopher et al. 2004). As a result, interest is growing in the development of more accurate and reliable methods to assess body composition in children for both field/epidemiological and clinical use (Parker et al. 2003).
Siervogel and others (2003) conducted a comprehensive review of the literature regarding body composition assessment during puberty and concluded that techniques such as DEXA and hydrostatic weighing are better than BMI for assessing total body fat and fat-free mass in children. BMI is especially sensitive to body build and tends to overestimate body fat for young people with undersized legs for their height. As a result, it is considered to be only poor to fair in identifying overweight children.
In another study, Parker and colleagues (2003) compared the validity of six field and laboratory methods against a reference (a three-compartment model) in 42 healthy 10- to 14-year-old boys. The investigators evaluated ADP, skinfold measurement, body density, BIA (both hand-foot and leg-leg impedance) and isotope dilution. All methods resulted in statistically significant differences compared to the reference method, with the exception of isotope dilution and skinfolds. Even so, the authors ultimately concluded that “the validity, at the level of the individual child, is poor for all of the methods tested, with wide limits of agreement and large biases.”
In contrast to two-compartment and anthropometric methods, DEXA is one method that has shown promise for assessing body composition in young people. In a study of 411 children and adolescents, there was a predictable relationship between DEXA and a four-compartment reference method for measuring percent body fat; this relationship was not affected by gender, age, ethnicity, pubertal stage, height, weight or BMI. As a result, the investigators concluded that DEXA was a useful clinical tool for predicting metabolic abnormalities associated with excess body fat in this population (Sopher et al. 2004).
Aging is associated with progressive loss of muscle and bone mass, expanded extracellular fluid volumes, reduced body cell mass and increased body fat (Baumgartner 2000). Body composition assessment techniques used in the elderly should account for these variations.
Sarcopenic obesity (obesity combined with a reduction in the body’s muscle mass due to aging and inactivity), a condition that is prevalent in elderly individuals, is one example of abnormal body composition that is difficult to assess using standard anthropometric methods. Because individuals with sarcopenic obesity have a normal or low BMI but reduced lean body mass and increased body fat, simply measuring BMI is not enough. For identifying disordered body composition in the elderly, multicompartment methods that can distinguish between fat and fat-free mass are necessary (Baumgartner 2000).
Elderly individuals also tend to have loose connective tissue, and storage fat becomes less subcutaneous and more internalized with age, making it difficult to accurately assess body composition using skinfold measurements. As a result, BIA has been suggested as the preferred field method for estimating percent body fat in elderly individuals (Wagner & Heyward 1999). However, because of fluid volume changes and other variations in body composition associated with aging, an age- and gender-specific prediction equation should be used (Heyward & Stolarczyk 1996).
The utility of DEXA, hydrostatic weighing and a multicompartment model has also been assessed in postmenopausal women. Houtkooper and colleagues (2000) calculated changes in body composition in 76 postmenopausal women over the course of a 1-year exercise program. The results showed that compared to hydrostatic weighing and the multicompartment method, DEXA was the most sensitive method for assessing small changes in body composition in these subjects.
Accurately assessing body composition in obese individuals can be problematic. For instance, the amount of hydration, especially as a component of fat-free mass, is typically greater in obese individuals and is not proportionately distributed between intracellular and extracellular fluid compartments. Because standard two-compartment models are based on the assumption that water composes 73.8% of the fat-free mass and its distribution is constant and independent of body fatness, two-compartment models are likely to result in large errors in estimating body fat in obese men and women (Waki et al. 1991). Increasing levels of fatness also alter the ratio of subcutaneous fat to total body fat, which affects the relationship between skinfold measurements and total body density (Heyward & Stolarczyk 1996). In addition, skinfold compressibility—and the overall thickness of the skinfold exceeding what can be measured with the caliper—can affect the accuracy of this method. Thus experts generally agree that skinfold measurements should not be used to estimate body composition in obese individuals.
In contrast, BIA appears to have some promise for accurately assessing body composition in obese persons. It is important to realize, though, that the preprogrammed prediction equations included with BIA software are typically not validated for this population. The different body geometry and variations in fluid distribution among obese individuals affect the conductivity of the body to current, resulting in overestimation of fat-free mass (Cox-Reijven, van Kreel & Soeters 2002; Heyward et al. 1992). Therefore, if BIA is used in obese individuals, a prediction equation validated for this population should be employed (Heyward & Stolarczyk 1996). Examiners should be sure to record the resistance value given by the BIA because the majority of prediction equations to determine fat-free mass from BIA utilize this value.
In addition to BIA, anthropometric methods such as BMI and circumference measurements appear to be suitable alternatives for assessing body composition in obese persons. These methods are not only simple but can be used to predict risk of disease and disability, which is usually a primary concern for these individuals. Some prediction equations, most notably those developed by Tran and Weltman, use circumference measures to determine percent body fat. These equations are useful when evaluating body composition in obese individuals (Tran & Weltman 1989). Unlike BIA, anthropometric methods do not require adherence to pre-testing guidelines and can be performed quickly and reliably in most settings (Heyward & Stolarczyk 1996).
Assessing body composition in athletes is important for optimizing performance and evaluating the effectiveness of various training regimens (Vescovi et al. 2002). Additionally, regular body composition assessment can ensure that an athlete maintains overall health, which is essential in sports where achieving dangerously low levels of body fatness is viewed as advantageous but could actually hamper performance. Yet body composition assessment in athletes can pose unique challenges because of alterations in body composition as a result of specific training regimens and the physical requirements of a particular sport.
In general, athletes have greater bone mineral content, bone density and skeletal muscle mass than the general population, and consequently have a higher density than sedentary individuals (Heyward and Stolarczyk 1996). For some athletes with high bone mineral content (e.g., bodybuilders), standard methods based on two-compartment models such as hydrodensitometry have been shown to underestimate body fatness by as much as 3%. On the other hand, body density of active women with chronic amenorrhea (cessation of menses once it has begun) may actually be lower than that of sedentary women, which can result in overestimation of percent body fat by as much as 3% (Bunt et al. 1990). These findings emphasize the need to use prediction equations validated for athletes from specific sport backgrounds.
Skinfold measurements have been validated in athletes and appear to have relatively high accuracy across several sports. Based on a review of the literature, Heyward and Stolarczyk (1996) recommend using the sum of seven skinfolds (chest, midaxillary, triceps, subscapular, abdomen, anterior suprailiac and thigh) to estimate body density for athletic men, and the sum of four skinfolds (triceps, anterior suprailiac, abdomen and thigh) for athletic women.
In contrast to skinfold measurements, anthropometric methods appear to have less predictive accuracy in athletic men and women. Athletes typically do not have the characteristic fat distribution that is seen in obese individuals, and risk of disease is typically not a primary concern. Therefore, while circumference measurements and other anthropometric measures such as BMI and waist-to-hip ratio can be used, they may have limited applicability when considering the ultimate goals of the athlete.
ADP, which has the benefit of being noninvasive and relatively simple and rapid, has been evaluated in athletes. Unfortunately, however, the accuracy of this technique has been inconsistent with this population. In one study the accuracy of the BOD POD was evaluated in 80 female collegiate athletes and it consistently overestimated percent body fat compared with both hydrostatic weighing and skinfold measurements. Researchers concluded that while ADP may be highly reliable in certain populations, it cannot be recommended for use in lean female athletes (Vescovi et al. 2002). In contrast, one population for which ADP may be appropriate is male collegiate wrestlers. In a study of 66 Division I wrestlers, the BOD POD resulted in similar estimates of body density, percent body fat and fat-free mass as hydrodensitometry (Utter et al. 2003).
Because of the variability in the body composition of athletes in different sports, there does not appear to be one specific method with a high degree of accuracy that is applicable to all athletes. It is recommended that assessment methods not be used interchangeably; rather, one method should be used consistently over time. Regardless of the method used, however, the general consensus among experts is that body fatness should not go below 5% for males and 12% for females. It is also important to realize that optimal body weight and body composition to maximize performance will vary among individuals. Therefore, body composition goals for athletes should be determined on a case-by-case basis, rather than on a set of general standards developed for a particular sport.
Body Composition and Disease
Improved technology and recent research findings have improved our understanding of how fat distribution within specific regions of the body influences overall health and disease. This research has led to the data presented in Table 2, which shows the relative risk of disease as a function of BMI and waist circumference.
BMI has been studied extensively for its potential in predicting risk of premature death, disease and disability. Probably the strongest evidence comes from a prospective study of more than 1 million men and women that investigated the effects of age, race, sex, smoking status and history of disease on the relation between BMI and mortality (Calle et al. 1999). The results of this study showed that subjects with the highest BMI had significantly greater risk of death compared with those who had a BMI of 23.5 to 24.9. Furthermore, the authors concluded that the risk of death from cardiovascular disease, cancer or other disease increased with increasing weight, regardless of age or gender. The risk of type 2 diabetes has also been linked to BMI, with research demonstrating that the relative risk increases for every additional unit of BMI over 22 (Colditz et al. 1995). Waist circumference is another powerful predictor of type 2 diabetes, with individuals with a waist circumference in the highest quartile having an 11 times greater risk of type 2 diabetes than those with a waist circumference in the lowest quartile (Wei et al. 1997).
In addition to BMI and waist circumference measures, waist-to-hip ratio (WHR) has been correlated with certain diseases. As a general rule, a WHR of 1.0 or higher is considered “at risk” for undesirable health consequences such as heart disease and other conditions associated with being overweight, and a WHR of 0.90 or less for men and 0.80 for women is considered safe (NIH 2004). Likewise, waist circumferences greater than 40 inches for men and 35 inches for women are considered undesirable and associated with increased risk of disease (CDC 2004).
Technological advances in assessment techniques combined with greater focus on how fat distribution affects overall health have led to improved ability to predict future disability and risk of disease. While cost and availability are usually the primary considerations, the choice of an assessment technique should also be based on the following factors (Wagner & Heyward 1999):
- level of subject cooperation and comfort
- number of participants and amount of time available for assessment
- body composition variables to be quantified
- desired accuracy of the measurements
- frequency of measurements
- client’s ultimate goals
Regardless of the method used, the results are only as accurate as the measurement technique and prediction equation applied. It is important to follow the standard guidelines and protocols associated with the chosen method and use prediction equations specific to the individual being tested.
ACSM. 1998. ACSM’s Resource Manual for Guidelines for Exercise Testing and Prescription (3rd ed.). Philadelphia: Lippincott Williams & Wilkins.
Baumgartner, R.N. 2000. Body composition in healthy aging. Annals of New York Academy of Sciences, 904, 437–48.
Bunt, J.C., et al. 1990. Variation in bone mineral content and estimated body fat in young adult females. Medicine & Science in Sports & Exercise, 22, 564–9.
Calle, E.E., et al. 1999. Body-mass index and mortality in a prospective cohort of U.S. adults. New England Journal of Medicine, 341 (15), 1097–1105.
Centers for Disease Control (CDC) National Center for Chronic Disease Prevention and Health Promotion. 2004. Defining overweight and obesity. www.cdc.gov/nccdphp/dnpa/obesity/defining.htm; retrieved June 24.
Colditz, G.A., et al. 1995. Weight gain as a risk factor for clinical diabetes mellitus in women. Annals of Internal Medicine, 122, 481–6.
Cox-Reijven, P.L.M., van Kreel, B., & Soeters, P.B. 2002. Accuracy of bioelectrical impedance spectroscopy in measuring changes in body composition during severe weight loss. Journal of Parenteral and Enteral Nutrition, 26, 120–7.
Davies, P.S.W., & Cole, T.J. 1995. Body Composition Techniques in Health and Disease. Melbourne, Australia: Cambridge University Press.
Demerath, E.W., et al. 2002. Comparison of percent body fat estimates using air displacement plethysmogra-
phy and hydrodensitometry in adults and children. International Journal of Obesity, 26, 389–97.
Expert Panel. 1998. Executive summary of the clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. Archives of Internal Medicine, 158, 1855–67.
Goodpaster, B.H. 2002. Measuring body fat distribution and content in humans. Current Opinion in Clinical Nutrition and Metabolic Care, 5, 481-7.
Heyward, V.H., & Stolarczyk, L.M. 1996. Applied Body Composition Assessment. Champaign, IL: Human Kinetics.
Heyward, V.H., et al. 1992. Predictive accuracy of three field methods for estimating relative body fatness of nonobese and obese women. International Journal of Sport Nutrition, 2, 75–86.
Houtkooper, L.B., et al. 2000. Comparison of methods for assessing body-composition changes over 1 year in postmenopausal women. American Journal of Clinical Nutrition, 72, 401–6.
Kushner, R.F. 1992. Bioelectrical impedance analysis: A review of principles and applications. Journal of the American College of Nutrition, 11, 199-209.
National Institutes of Health (NIH). 1998. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. Bethesda, MD: Department of Health and Human Services; NIH; National Heart, Lung, and Blood Institute. Publication No. 98–4083.
Parker, L., et al. 2003. Validity of six field and laboratory methods for measurement of body composition in boys. Obesity Research,11 (7), 852–8.
Siervogel, R.M., et al. 2003. Puberty and body composition. Hormone Research, 60 (1, Suppl.), 36–45.
Slinde, F., et al. 2003. Bioelectrical impedance variation in healthy subjects during 12 hours in the supine position. Clinical Nutrition, 22 (2), 153–7.
Sopher, A.B., et al. 2004. Measurement of percentage of body fat in 411 children and adolescents: A comparison of dual-energy x-ray absorptiometry with a four-compartment model. Pediatrics, 113 (5), 1285–90.
Tran, Z.V., & Weltman, A. 1989. Generalized equation for predicting body density of women from girth measurements. Medicine & Science in Sports & Exercise, 60, 167–75.
Utter, A.C., et al. 2003. Evaluation of air displacement for assessing body composition of collegiate wrestlers. Medicine & Science in Sports & Exercise, 35 (3), 500–5.
Vescovi, J.D., et al. 2002. Evaluation of the BOD POD for estimating percent fat in female college athletes. Journal of Strength and Conditioning Research, 16 (4), 599–605.
Wagner, D.R., & Heyward, V.H. 1999. Techniques of body composition assessment: A review of laboratory and field methods. Research Quarterly for Exercise and Sport, 70 (2), 135–9.
Waki, M., et al. 1991. Relative expansion of extracellular fluid in obese vs. nonobese women. American Journal of Physiology, 261, E199–E203.
Wang, J., et al. 2000. Anthropometry in body composition. Annals of the New York Academy of Sciences, 904, 317–26.
Wang, J., et al. 2003. Comparisons of waist circumferences measured at 4 sites. American Journal of Clinical Nutrition, 77, 379–84.
Wei, M., et al. 1997. Waist circumference as the best predictor of noninsulin dependent diabetes mellitus (NIDDM) compared to body mass index, waist/hip ratio, and other anthropometric measurements in Mexican Americans—A 7-year prospective study. Obesity Research, 5, 16–23.