Body Composition 25From the mid-arm circumference, in combination
with the triceps skinfold thickness, information on
muscle mass and fat mass of the upper arm can be
obtained. Circumferences of the trunk at waist, hip,
and upper thigh level are used as indicators of body
fat distribution. The WHO suggests an upper limit
waist-to-hip circumference ratio above 0.85 for
females and 1.00 for males for abdominal fat distribu-
tion. Diameters can also be used to gain insights into
body fat distribution. A high sagittal diameter com-
pared with transverse diameter of the abdomen is
indicative of an enlarged amount of visceral fat.
However, it has to be kept in mind that the relation-
ship between these anthropometric parameters of
body fat distribution and the intra-abdominal fat
determined by CT or MRI scan is generally low.
Changes in internal fat detected by MRI are only
weakly associated with changes in these anthropo-
metric parameters of fat distribution.
Infrared interactance
The principle of infrared interactance is based on
differences in absorbance and refl ection of infrared
light in different tissues. When the absorbance
of near-infrared light (700–1100 nm) is measured
at well-defi ned sites of the body, information on
the thickness of the subcutaneous adipose tissue
layer can be obtained. In the prediction formulae
used, many other parameters are included, such as
weight, height, age, and gender, and it has been argued
that the prediction of body fat depends more on
these parameters than on the measured infrared
interaction.
Ultrasound measurements
Ultrasound measurements can also be used to study
several aspects of body composition. With ultrasound
measurements, the thickness of the subcutaneous fat
layer can be determined and total body fat can be
calculated. A good refl ection signal depends heavily
on the absence of connective tissue between adipose
tissue and muscle. The main diffi culty in ultrasound
measurements is the exact application of the ultra-
sound transducer perpendicular to the tissue without
any pressure. In the literature, several studies report
a good correlation between skinfold thicknesses mea-
sured by calipers and those measured by ultrasound.
The correlation of skinfolds with total body fat was
higher than when using ultrasound, suggesting that
skinfold thickness measurements have a better pre-
dictive value.
Internal abdominal body fat can also be assessed
with ultrasound. Studies have shown that ultrasound
measurements provide a better method than anthro-
pometry to assess internal abdominal fat depots.Bioelectrical impedance
In bioelectrical impedance, a small alternating current
is applied to the body. It is assumed that the body
consists of different components, of which water and
dissolved electrolytes are able to conduct the current.
Hence, body impedance is a measure of body water.
The electrical resistance or impedance of an electro-
lyte solution depends on several factors, of which the
most important are the amount of electrolytes (pure
water does not conduct the current), the kind of
electrolytes, and the temperature of the solution. If
currents of low frequency (<5 kHz) are used, body
impedance is a measure of ECW, as a low-frequency
current cannot penetrate the cell membrane, which
acts, with its layers of protein, lipids, and proteins, as
an electrical capacitor. With increasing frequencies
the capacitor features of the cell membrane diminish
and gradually ICW also participates in the conduc-
tance of the current, resulting in lower impedance
values at higher frequencies. Hence, at higher fre-
quencies, TBW is measured. TBW and ECW can be
predicted from impedance at high and low frequency,
respectively, using empirically derived prediction for-
mulae. Other parameters are often taken into consid-
eration, such as body weight, age, and gender.
Most prediction equations are based on statistical
relationships between empirically measured im-
pedance index values (height^2 /impedance) and body
water values obtained by dilution techniques such as
deuterium oxide dilution (for TBW) and bromide
dilution (for ECW). As body water in healthy subjects
is an assumed fi xed part (73%) of the FFM, bioelectri-
cal impedance measurements can also be used for the
prediction of the FFM and hence body fat percentage.
For those prediction equations, the impedance index
was related to measures of FFM, normally obtained
by densitometry or by DXA.
Body impedance depends on the frequency of the
current used and on body water distribution between
the extracellular and intracellular space and between
the different geometrical body compartments (legs,
trunk, and arms). This calls for extreme caution in the