Introduction to Human Nutrition

Body Composition 23

groups of subjects such as pregnant women or body-
builders. However, other predictive methods also
have their limitations in these subjects.
TBW can also be predicted from weight and height,
and numerous prediction formulae have been pub-
lished. The standard error of estimate of these equa-
tions is, however, considerable.

Skinfold thickness measurements

Body fat is located both internally and subcutane-
ously. If one assumes a constant relationship between
subcutaneous fat and total body fat, then total body
fat can be estimated by measuring the amount of the
subcutaneous adipose tissue. The amount of subcu-
taneous tissue can be estimated by measuring the
thickness of the subcutaneous fat layer at different
sites of the body using a skinfold caliper, infrared
interactance, or ultrasound measurements. In a given
age group, the relation between subcutaneous fat and
total fat is indeed relatively constant. However, the
relationship is different between males and females,
females having relatively more internal fat (Box 2.9).
Thus, it is possible by using age- and gender-specifi c
prediction equations to assess the total amount of
body fat by measuring skinfolds at different sites of
the body.
Skinfolds can be measured all over the body. The
most often measured skinfolds for the assessment of

total body fat are skinfolds on the upper arm biceps

(Figure 2.5) and triceps, under the scapula (subscapu-

lar) and above the iliac crest (suprailiac). The sum of

more skinfolds is normally used to reduce the error

in measurement and to correct for possible differ-

ences in subcutaneous body fat distribution between

subjects within the same age and gender group.

Various prediction formulae for body fat from skin-

fold thickness have been published. For children, in

whom the relationship between skinfold thickness

and body fat depends on biological age, separate for-

mulae must be used.

Measuring skinfolds adequately requires a trained

and experienced observer, otherwise excessively large

errors in the assessment of the body fat can occur. A

disadvantage of the method is that the subject has to

be partly undressed. This hampers the use of the

method in epidemiological studies. In bed-ridden or

seriously ill patients the measurement of the trunk

Box 2.9

From Table 2.6 it can be seen that for the same amount of subcu-

taneous fat (identical skinfold thickness) women have more body

fat than men. This is because of the higher internal (organ) fat

content in women. It can also be seen (in both females and males)

that at equal skinfold thickness older people have more body fat:

with age the amount of internal fat increases.

Box 2.8

Recent studies have shown that the relationship between body

mass index (BMI) and body fat percentage differs among ethnic

groups. For example, compared with Caucasian populations some

Asian populations have 3–5% more body fat for the same BMI,

age, and gender. These differences can be explained by differences

in body build or frame size, subjects with a smaller frame having

more body fat at the same BMI.

These differences can have important consequences for the

defi nition of obesity (based on BMI cut-off values) and the preva-

lence of obesity in a population. In Indonesia, obesity has recently

been redefi ned as BMI ≥ 27 kg/m^2. At this BMI, Indonesians have

a similar body fat to Caucasians with a BMI of 30 kg/m^2. The lower-

ing of the cut-off point for obesity from 30 to 27 kg/m^2 increased

the prevalence of obesity from less than 5% to over 10%.

Recently an Expert Consultation of the World Health Organiza-

tion (WHO) resulted in new guidelines to redefi ne “action points”

in non-Caucasian populations. For this not only was the different

relationship between BMI and body fat percentage important, but

also the high levels of cardiovascular risk factors at low BMI values

(WHO, 2004).

Figure 2.5 Measurement of the biceps skinfold.