1284
R.C.Siiskamp, N. L. Conen, W. E. Sint, T. W. SARGENT and H. E. WALSH
16. Thigh diameter. At the mid-thigh, the calipers were placed in an anteroposterior position parallel to the floor with the subject standing and relaxed. The
caliper arms rested easily on the skin surface with no skin indentation made.
The following measurements (17-20) were made with Lange skinfold calipers
(Cambridge Scientific Intruments, Cambridge, Maryland) in accordance with the
techniques described by the Interdepartmental Committee on Nutrition for National
cn
Defense [37]. Spring tension of the calipers exert a pressure of 1Og/mm2 over a 40 mm?
surface area. The right side was used except for those subjects having a right lower
quadrant abdominal scar. In these subjects, the abdominal skinfold was measured on
the left side.
17. Arm. With the subject standing, right arm relaxed and flexed 90° at the elbow,
the calipers were applied adjacent to the fingers grasping a fold of skin and subcutaneous tissue over the triceps at mid-arm level.
18.
Scapula. With the subject standing, a fold of skin and subcutaneous tissue
over the inferior angle of the right scapula was grasped so as to parallel a line about
45° from the horizontal level going medially upward. The calipers were applied to the
skinfold adjacent to the grasping fingers.
19. Thorax. With the subject supine, breathing quietly and knees raised to relax
the abdominal muscles, the skin and subcutaneous tissue over the right 10th rib was
grasped parallel to the rib in the anterior axillary line and the calipers applied adjacent
to the grasping fingers.
20. Abdomen. With the subject in position as for thorax skinfold measurement,
a fold of skin and subcutaneoustissue of the abdomen1 in. to the right of the midline
halfway between the umbilicus and the symphysis pubis was prasped so as to parallel
the mid-abdominal line. The calipers were applied and the measurementread.
Laboratory determinations
Laboratory determinations were usually done within two weeks after the anthropometric measurements. Women were studied when non-menstruating. The body
weight and height were recorded at the time of the measurements for total body water
and body density. If the weight differed by +2 per cent of that obtained in the
anthropometric examination, the anthropometric measurements were retaken for
subsequent use in calculations.
|. Body density determinations by the method of Sir! [21] measure whole body
volume in a closed-circuit system which utilizes a gas-dilution principle. With the
subject in one chamber and helium in the other, the gases in the two systems are mixed.
The helium concentration, measured by thermal conductivity, is related to the volume
displacement by the subject. The standard deviation in volume determination is
estimated at +0.12 liter. Since the error is little affected by absolute volume, the
corresponding error in density ranges from +0.0024 g cm-3 for a 50 kg subject to
+0.0012 ¢ cm~3 for a person weighing 100 kg.
2. Total body water was measured after an oral tracer dose of tritiated water, given
on the same day as body density was determined. The methods employed involved
equipment and proceduresin general use. Four to five timed urine specimens,collected
within the first 24-hour period after the dose ingestion, were assayed for tritium with
a Tri-Carb liquid scintillation coincidence counter (Packard Instrument Co.,
LaGrange, Ill.}. The liquid scintillator formulated by WERBIN ef a/. [38] was used.
Measures of Body Fat and Related Factors
in Normal Adults— I
1285
It was found that 15 ml ofscintillator with 2.0 mi water (the
usual sample volume)
counted with anefficiency of 17 per cent. Samples
were always recounted with an
internal standard of tritiated water and appropr
iate corrections were made for
quenching. Vacuum distillation of urine specimens
reduced the volume to an appropriate amount. In general, counting, pipetting, weighing
, or other errors were maintained well below | per cent by suitable precautions.
3. Total body fat was calculated for each Subject from
the combined measurements
of body density and of total body water. The order
of magnitude of the error of this
calculation, based on the above values for the experimental
errors in density and water
determinations and for the variability in the combined
densities for protein and mineral,
has. been found to be +1.7 per cent of the total body
weight [25]. For a 70 kg
individual the standard deviation in the determination
of body fatis therefore +1.2kg
This erroris smaller than that expected when body fat
is estimated by either density or
water measurements since fewer variables are unaccou
nted for. In addition total bod
water—body density technique for measuring bodyfat is
completely independentofthe:
state of hydration of the subject.
4. Lean body mass was calculated fromthe body weight
less total body fat
S. Under the supervision of one of the authors, (T.W.S.)
, potassium-40 (K 4) and
cesium-137 (Cs!37) were determined using an Argonne
type whole-body counter [39]
Immediately before the subjects entered the countin
g chamber they showered (men
also shampooed their hair while women shampooed
their hair the night before
avoiding the use of hair sprays, rinses, etc.) and put
on cotton garments and disposable
slippers supplied by the laboratory. These precautions
were taken to removeall traces
of external activity, such as potassium salts on the
skin or scalp, or from fallout
occasionally found in dust on hair, skin and clothin g.
Whole body counts were
performed usually on the same day, and always within
one week, of the body density
and total body water measurements. The counting geomet
ry was that ofthe ‘tiltin
chair’ [39].
°
Cs187 activity was determined from the area underth
e 0.66 million electron volts
(MeV) gamma photopeak of this isotope, from 0.57
to 0.75 MeV. Cs137 counting
efficiency was determined in a group of eight volunteers
by injection ofcalibrated doses
of Cs13" (half-life—6.2 days); the mean counting
efficiency was 8.1 counts per
minute per nanocutie (nc) with a standard deviation of
0.96 counts/min/ne.
Calibration of the counter for K4° was performed
by injection of known doses of
K4? (half-life=12.5 hr) into nine volunteer subjects.
An average counting efficienc
of 1.09 +0. 10 counts/min/g with a standard deviation
of 9 .7 per cent was determined,
The total body potassium for each subject was obtaine
d by dividing K4° counts /min
by the efficiency factor. This calibration method is strictly
empirical, compar
ed
with the theoretical-empirical calibration obtained by
MENEELYet al. [40] The latter
method considers the subject’s height and weight
and yields values with a lower
standard deviation.
No such correlation between height and weight and countin
g
efficiency was foundin the present study, so the efficien
cy used here is simply the mean
of values for the subjects used for calibration.
Diet and physical activity assessment
Nutritionists (E.M.H. and M.H.R. during the first
part and E.M.H. and N.D.K
during the last part of the study) interviewed individually
each subject selected for