1280
R. C. STEINKAM?P, N, L. Conn, W. E. Stat, T. W. SarGeni and H. E, WALSH
«3% Bases of reference for body composition have had several approaches. Cadaver
esanalysis [14-17], body density [8, [8-21], total body water [22-24], combined density
and body water determinations [25, 26], absorption of non-toxic fat-soluble gases [27],
whole body counting of radioisotopic potassium (K4°) [28-30] and measurement of
--apultiple parameters by dilution techniques {3] have been used.
expe SIRI has critically evaluated the methods of determining body composition from
fluid spaces and from body density [31]. His conclusion that “combined measurements
‘of corporal density and of total body wateryield the only method for estimating body
composition that does not require a reference body nor an explicit description of the
composition of adipose tissue’ has been upheld by the work of ALLEN ef al. [26].
Theystate, “Srrt’s original simplifying deductions can be shown to hold throughout a
wide range in variation of bone mineral: fat: water: solids of healthy persons.”
The extensive body composition determinations by Mooreef a/. [3] using multipleisotope-dilution techniques provide an additional approach. Their skeletal weight
nomogram has as its basis the work of ALLEN ef a/. [26] plus a measure of body
potassium equivalents.
Additional approaches to body composition methodology have been measurements
of subcutaneous fat by roentgenography[32] and by caliper measurements[13, 33, 34].
These have been used to estimate body fat or body density. SHELDON’s approach has
been that of defining the body build by somatotype [35].
While muitiple techniques for measuring body composition have been developed,
few have been correlated in order to develop a practical clinical tool of determined
validity. The work of Brozek and Keys [19], BEHNKE[12], YOUNG and TENSUAN[11]
and Mooreef al. [3] are exceptions. However, for the most part, these studies have
been limited to one sex and to Caucasians. In addition there has been no study to
compare in a large numberof subjects lean body mass and body fat as determined by
thet otal body water—-body density technique with these as determined by whole body
potassium-40 for the same individuals.
There is considerable evidence that body composition is dependent upon multiple
factors including sex, genetic constitution, physical! activity and nutritional status [36].
Weare unaware, however, of previous attempts to correlate in man measured body
fat and lean body mass with diet, physical activity and with somatotype.
OBJECTIVES OF STUDY
The primary objectives of this study were:
to
1.
Todevelop for healthy adults valid regression equationsusing clinically applicable
anthropometric measurements which estimate total body fat.
To comparein the same subjects determinations of lean body mass and total body
fat as obtained by the total body water—body density technique and by the
measurement of whole-body potassium-40.
Additional objectives were:
3.
To relate estimated and measured total body fat to diet, physical activity and
somatotype.
4.
To relate whole-body cesium-1|37 to diet.
Measures of Body Fat and Related Factors in Normal Adults—I
1281
PURPOSE OF THIS REPORT
This report presents the selection of study subjects and the methods used. The
methods are considered under the following headings: anthropometry, laboratory
determinations (body density, total body water and whole-body counting of
potassium-40 and cesium-137), diet and physical activity assessment and somatotyping
technique. The anthropometric measurements are presentedin sufficient detail so that
the clinician may perform them exactly as recorded, a matter of import in employing
the regression equationsfor estimation of body fat with the accuracy determinedfor
the method.
Subsequentreports will present results of the various measurements, the regression
equations and interrelationships of the assessed factors.
SELECTION OF SUBJECTS
Altogether 2301 healthy* individuals aged 25-44 years were obtained as volunteer
subjects from many work or other groups in the San Francisco Bay area. Small as
well as large business concerns, companies, agencies and other groups(list appended)
were contacted through personnel offices or their associated medical facilities for
permission to request employee or group participation on a volunteer basis. Each
individual, qualifying by age, received a letter which described the purpose of the study
and the parameters to be measured and requested voluntary participation. Anthropometric measurements and diet and physical activity interviews were performed at
facilities of the company. Persons, randomly selected for the laboratory measurements,
were taken subsequently to the Donner Laboratory, University of California, Berkeley.
Only graduate students of this university were somatotyped at. the Cowell Memorial
Hospital Constitutional Research Laboratory.
Owing to limitations in the time schedule for laboratory work and the small numbers
of volunteers from several races, race-sex—age categories (vide infra) were arbitrarily
defined. From these categories, subjects were randomly selected for diet and physical
activity interviews and for laboratory measurements. If for any reason a subject so
selected was unable to continue in the study, another subject of the same race~sex—age
category was selected. The five arbitrary categories were:
I.
Caucasian males, 25-34 years old.
II.
Caucasian males, 35-44 years old.
IW.
Caucasian females, 25-34 years old.
IV.
Caucasian females, 35-44 years old.
V.
Negro males, 25-44 years old.
Table 1 presents the number by category, including oriental and other race groups,
who participated in each of the phases ofthe study.
*Healthy wasdefined as free of known disease, acute or chronic, upon statement ofthe
subject, and
having had no treatment by a physician for anyillness during the three months prior to
entry in the
study. Any person with obvious physical deformity or loss of limb was excluded. In addition
pregnancy or suspected pregnancy eliminated any woman from thestudy.