Use and Interpretation of Anthropometry
Youfa Wang
Hyunjung Lim
Benjamin Caballero
1Abbreviations: %BF, percentage of body fat; BC, body composition; BMI, body mass index; CDC, Centers for Disease Control and Prevention; CVD, cardiovascular disease; DXA, dual-energy x-ray absorptiometry; IDF, International Diabetes Federation; IOTF, International Obesity Task Force; LMS, least means squared; NCHS, National Center for Health Statistics; SD, standard deviation; WC, waist circumference; WHO, World Health Organization; WHtR, waist-to-height ratio.
Anthropometry is defined as the measurement of humans for the purposes of understanding human physical variation. Anthropometric measures have been widely used for the assessment of nutritional and health conditions such as body composition (BC), malnutrition, and obesity. Changes in lifestyle, nutrition, and ethnic composition of populations lead to changes in body dimensions and BC. Anthropometric measures are more widely used in children than adults considering the routine needs of assessing growth (1, 2). The World Health Organization (WHO) has developed guidelines and growth references to provide guidance on the use and interpretation of anthropometric measures (1, 2). At present, weight and height are the most widely used anthropometric measurements, and their derivative, the body mass index (BMI), is the most commonly used indirect indicator of obesity and body adiposity (3, 4, 5).
This chapter describes the most commonly used anthropometric measurements, the indices derived from them, and the use and interpretation of these. Advances in developing growth standards and reference charts are discussed. Another chapter provides a detailed description of BC techniques.
COMMONLY USED ANTHROPOMETRIC MEASURE INDICES
The commonly used anthropometric measures in adults and children include weight, height, waist circumferences (WC), skinfold thickness (measured on different body sites), and a set of weight-for-height indices such as BMI. Often such measures or a combination of them are used as indicators of BC, such as percentage of body fat (%BF). These measures and their strengths and limitations are summarized in Table 49.1. Note that underwater weighing and dual-energy x-ray absorptiometry (DXA) are considered the gold standards for BC assessment.
Several studies have assessed the validity of anthropometric measures such as BMI, WC, and skinfold thickness for estimating body fat using DXA as the reference method. Results indicate modest to excellent agreement, with correlations ranging from 0.37 to 0.99 in adults (6, 7, 8) and children (9). The agreement is stronger in healthy subjects (R > .97) (8). One study found that accuracy of most of the skinfold thickness equations for assessment of body fat at the individual level was poor in 13- to 17-year-old adolescents compared with DXA (10). Others found that skinfold thickness measures are better predictors of %BF than other simple anthropometric methods such as BMI (11). In Asian adolescents, clinical validity of weight- and height-based classification for obesity screening is poor when compared with that defined based on a %BF greater than or equal to the 95th percentile. According to Youden’s index, a composite measure of accuracy indices indicating optimal sensitivity and
specificity rates, weight- and height-based classification only presented 48% in boys and 59% in girls (12).
specificity rates, weight- and height-based classification only presented 48% in boys and 59% in girls (12).
TABLE 49.1 COMMONLY USED ANTHROPOMETRIC MEASURES AND THEIR MAIN STRENGTHS AND LIMITATIONSa | ||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||
A newer anthropometric index, the waist-to-height ratio (WHtR), has been proposed as a useful indicator of central obesity and for screening cardiovascular disease (CVD) risk (13, 14). The WHtR was strongly correlated with %BF and fat distribution, which are associated with increased CVD risks (15, 16). Some research has indicated that WHtR is independent of age and eliminates the need for percentiles for children (17, 18). A WHtR cut-point of 0.5 has been recommended to classify central obesity in adults and children and for different ethnic groups (14). For example, the optimal value for WHtR was 0.5 for Japanese adults and its sensitivities of various proposed obesity indices for identifying clustering of defined and other risk factors (13). Among children, the WHtR showed high sensitivity and specificity (>0.90) compared with WC in Chinese children (8 to 18 years old) (17). In a study of British children (5 to 16 years old), WHtR decreased with age (18); it also increased greatly during the past 10 to 20 years, and was found to be more closely related to morbidity than was BMI (18).
WC has been recommended by the WHO and the International Diabetes Federation (IDF) as a measure of central obesity, which is a key component for defining metabolic syndrome (19). Studies indicate that WC is a good predictor of the risks for a number chronic diseases, such as CVD and type 2 diabetes, and is often a better predictor than BMI (20). A set of sex- and ethnic-specific WC cut-points have been recommended for adults (19, 21, 22, 23, 24); such as in men, 85 (Japan), 90 (by IDF for Asian and in countries such as China), 94 (Vietnam), 100 (France), and 102 (WHO international recommendation); and in women, 80 (by IDF), 85 (South Korea), 88 (WHO), and 90 (France and Japan). Previously, waist-hip ratio was used to measure central obesity, but later it was recommended that WC is adequate, whereas the ratio did not add much value.
DEVELOPMENT OF CUT-POINTS FOR ANTHROPOMETRIC MEASURES
One of the most common applications of anthropometric data is in the diagnosis or categorization of conditions such as underweight and overweight and the grading of their severity (3, 25, 26, 27). In addition, cutoff thresholds are used to elucidate variations in age, maturation, gender, ethnicity, and other “technical” factors that affect anthropometry “independently” or in conjunction with health or social causes or consequences as well as in applications such as policy formulation, social utility, and advocacy for particular problems and solutions (28). Different indicators and cut-points are needed for different application purposes. However, this notion may not be agreed by various user communities, because universal cut-points of simple
indicators often are considered easier to use and better for international comparisons.
indicators often are considered easier to use and better for international comparisons.
In growth references, certain Z-scores (e.g., +2 and −2) and percentiles (e.g., 5th, 85th, and 95th) often have been chosen for cut-points to classify problematic growth and nutritional status such as malnutrition or obesity. These criteria are based on statistical distribution rather than the associated health risks. Ideally, the criteria used should be established based on their associations with higher health risks. Cut-points for classifying “higher-risk” individuals and population groups should be based on the evidence of increased risk for morbidity, mortality, or/and impaired function (2). To assess the relationship between different anthropometric indicators and health outcomes is often more difficult in children than in adults. It is even harder to choose cut-points for “higher-risk” individuals. In children, the short- and intermediate-term health outcomes during childhood and adolescence and the longterm health outcomes in adulthood need be considered.
BODY MASS INDEX
BMI is a simple index of weight-for-height calculated as weight (kg)/height (m2) (kg/m2). BMI is commonly used to classify underweight, overweight, and obesity in adults and children worldwide. Many different BMI cut-points have been recommended and used over the past two decades. Some of them are becoming internationally used. However, it is a still a matter of debate what cut-points are more appropriate for specific populations considering racial and ethnic differences in BC.
Strengths and Limitations of Body Mass Index
An ideal measure of body fatness should meet several requirements.
It should be accurate in assessing the amount of body fat.
It should be precise with small measurement error.
The measure should predict risks of health consequences; that is, it should have a strong association with health outcomes.
It should be possible to develop some cut-points to separate individuals into different groups regarding their excess adiposity-related health risks.
For a measure to be useful in clinical setting or epidemiologic studies, it also should be accessible (in terms of simplicity, cost, and ease of use) and acceptable to the subjects (29).
BMI has most of these features, although none of the existing measures satisfies all of these criteria. BMI is identified as the best choice among available measures that can be easily assessed at low cost, and has a strong association with body fat and health risks.
However, as an indirect measure of adiposity, BMI has several limitations, especially when used for children (3, 30, 31, 32, 33). Some examples follow.
Children grow and gain lean body mass and adipose tissue at different rates, and there are large differences in between-population, and interindividual and intraindividual variations. Children’s maturation status and growth patterns affect their BC and BMI. Thus, the meaning of BMI may vary, being more complex in children than in adults (3).
BMI is positively associated with height in children and the association varies by age and gender (3, 33), although it is independent from height in adults.
There are biologic differences between ethnic groups and populations in BC, the relationships between BMI and %BF, and those between BC and morbidity.
Compared with classification of obesity based on %BF, BMI has a low sensitivity, although it has a high specificity (31, 32, 34).
Secular changes in height growth and in BC may complicate the interpretation of BMI.
Body Mass Index Cut-Points
BMI values are sex- and age-dependent. The same BMI may reflect different levels of body fat in different populations, in part because of differences in body build. The health risks associated with increased BMI are continuous, and the interpretation of BMI grading in relation to risk may differ across populations. Since the late 1990s, there has been debate on whether population- or ethnicity-specific BMI cut-points should be used for the classification of obesity (25, 26). Research has suggested some ethnic differences in the associations among BMI, %BF, fat distribution, and health risks (26, 30, 35, 36).
Adults
Table 49.2 shows the adult BMI cut-points recommended by the WHO for the classification of underweight, overweight, and obesity in adults. Different BMI cut-points have been recommended for some Asian and Pacific populations (24). To address the debate, a 2002 WHO Expert Consultation examined available evidence and made related recommendations (26). They concluded that the proportion of Asian people with a high risk of type 2 diabetes and CVD is substantial at BMIs lower than the existing WHO cut-point for overweight of 25 (26). However, the cut-point for observed risk varies from 22 to 25 in different Asian populations, and for high risk it varies from 26 to 31. They recommended that the current WHO BMI cut-points should be retained as the international classification; however, the cut-points of 23, 27.5, 32.5, and 37.5 were added as points for public health action. They recommended that countries should use all categories (i.e., 18.5, 23, 25, 27.5, 30, 32.5, and in some populations, 35, 37.5, and 40) for facilitating international comparisons. A review reported that 13 of the 18 identified cohort and cross-sectional studies indicated lower BMI cutoff points being more appropriate for Asian populations than BMI values of 25 and 30 (37).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
