- Email: [email protected]
A comparison of methods of assessment of body composition including neutron activation analysis of total body nitrogen
A Comparison of Methods of Assessment of Body Composition Including Neutron Activation Analysis of Total Body Nitrogen H. C. Lukaski,
J. Mendez,
E. R. Buskirk,
and
S. H. Cohn
Fourteen healthy men underwent determinations of total body nitrogen (TBN) by prompt gamma neutron activation analysis and total body potassium (TBK) by whole body counting to estimate the muscle and nonmuscle components of the fat-free body mess (FFBM) and their protein contents. Comparison of FFBM estimated from TBN and TBK (66.6 f 6.9 kg, meen + SD). densitometry (62.3 + 7.1 kg), TBK alone (62.2 + 8.0 kg), and TBW (63.9 + 7.8 kg) showed no differences among the techniques. Similarly, there were neither differences in fat mass nor percent body fat among the methods. Anal@s of the chemical composition of FFBM of this group showed TBK/FFBM = 62.6 + 2.3 mEq/kg. TBW/FFBM = 74.6 + 0.2%. TBN/FFBM = 32.74 f 1.09 g/kg, protein/FFBM = 20.5 + 0.7%. The calculated mineral content of the FFBM wes 6.4%. Theeevekres are strikingly similar to the values calculated by direct chemical analysis. It was concluded that the combinsd TBN-TBK method is a valid technique for estimating body composition in man.
T
HE INCORPORATION of body composition techniques into the practice of medicine, clinical nutrition, and applied physiology has provided normative information on the healthy population. These data are based upon the assumption that the body consists of two chemically distinct compartments, fat and fat-free.‘z If the body mass (BM) is known, determination of the fat-free body mass (FFBM] allows for the calculation of the fat mass (FM) by difference, and vice versa. The most reliable methods available for estimation of FM and FFBM require measurement of body density, total body potassium (TBK), or total body water (TBW). Estimation of body composition by each of these methods is based upon assumptions relating to the relatively constant composition of the FFBM or FM. Anderson3 proposed a three-compartment system of body composition analysis which utilized measurements of TBK and TBW to predict the body fat and the muscle and nonmuscle fractions of the FFBM. More recently, Burkinshaw et al.4 developed equations hased upon the different nitrogen/potassium ratio to estimate the muscle and nonmuscle portions of the FFBM and their respective protein contents. This approach requires determinations of TBK and total body nitrogen (TBN) which now can be measured accurately and safely by the prompt gamma neutron activation teehnique.5 The mass and the protein content of the muscle and nonmuscle FFB can be calculated and so provide a very useful tool to study objectively the effects of various physiological and pathological stresses on human body composition. However, before this method can be implemented in comparative studies, it is important to determine body composition in normal subjects. The purpose of the present study was to obtain baseline data on body composition calculated in several ways in a group of healthy young men and to compare the estimates derived from the TBN-TBK model with those determined by independent methods Me-,
Vd. 30. No. 8 (Aqustl, 198 1
(densitometry, TBK, and TBW). Also, the composition of the FFBM obtained from these various estimates is compared with the composition commonly assumed for the FFBM. MATERIALS
AND
METHODS
Fourteen healthy men, students at the Pennsylvania State University, volunteered to participate in this study. Subjects received a detailed description of the purpose and procedures, and gave their informed consent. This study was approved by the Biomedical Review Committee of the Pennsylvania State University and the Human Subjects Review Committee of the Medical Research Committee of the Brookhaven National Laboratory (BNL). Body composition was determined by densitometry at the Noll Laboratory using the underwater weighing system and the method of Akers and Buskirk. Residual volume was measured simultaneously with the underwater weighing by an open-circuit technique for nitrogen washout of the lungs.’ Percent body fat (% BF) was calculated from body density (D) according to Brozek et al.:* % BF = 100 [(4.570/D) FM and FFBM were calculated follows:
- 4.1421
(1)
from % BF and body mass (BM) as
FM = % BF (BM)/lOO FFBM = BM -FM
(2) (3)
where BM, FM, and FFBM are in kg. FM can be estimated by densitometry with a precision of + 0.5 kg.8 After the density determination, the subjects travelled to the Brookhaven National Laboratory for determinations of TBK, TBN, and TBW. The TBK content of each subject was measured exter-
From the Laboratory for Human Performance Research. The Pennsylvania State University, University Pork. Pennsylvonio. ond the Medical Research Center, Brookhoven Notional Loborotory, Upton. New York. Supported in part by PHS Grants CM0 1748 and AM-08311. NIH Grant RR0 7082, and NIH-NCI Contract YOI-CP-80207. Received for publication August 20. 1980. Address reprint requests to J. Mendez. 101 No11 Laboratory. Pennsylvania State University, University Pork. Pennrylvanio 16802. 0 1981 by Grune & Stratton, Inc. 00260495/81/3008-0007$01.00/0 777
778
LUKASKI ET AL.
nally by detection of 1.46 Mev gamma rays emitted from the naturally occurring radioactive isotope 40K which is a constant fraction (0.012%) of body potassium. The Brookhaven whole-body counter with its on-line computer facility was used to measure absolute levels of TBK. The counter system has a relatively invariant response to individual body size and radionuclide distribution.9 The accuracy and precision for measuring TBK is + 3.3%.r0 It has been determined there is a ratio of 62.7 Meq (2.45 g) potassium per kg FFBM (I I). Calculation of FFBM and FM is: FFBM
= TBK/2.45
(4)
FM = BM - FFBM
(5)
% BF = lOO(FM)/BM
(6)
Absolute levels of TBN were quantitated by using neutron capture prompt gamma ray analysis, IgN(n.y) “N, as described by Vartsky et al.’ Briefly, the capture of a thermal neutron by 14N is followed by the prompt emission (lo-*’ set) of a cascade of gamma rays. About 15% of the de-excitations takes place directly from the highest excited state to the ground state of 15N by emission of a 10.83 Mev gamma ray. Becaase no other body element has a neutron capture gamma ray at this energy, the 10.83 Mev gamma ray of “N is measurable among the more numerous captures by the other body elements. An internal standard based upon body hydrogen reduces the requirement for a highly uniform composite sensitivity (i.e., an equal number of counts per unit mass of nitrogen per unit dose) for absolute measurements.” While lying on a movable bed, subjects were irradiated from shoulder to knee in the prone and supine positions by a 60 x 13 cm rectangular beam of partially moderated neutrons from an 85 Ci 238Pu,Be source. Gamma rays were detected by two 15.24 x 15.24 cm NaI(TI) detectors positioned over the subject. The total measurement time was 20 min, and the radiation dose delivered to the body was 26 mrem. The precision of the TBN estimate is * 3%.5 TBW was determined by the dilution technique using tritiated water (HTO) as the tracer. After an overnight fast, subjects drank 50 &i of HTO (New England Nuclear, Boston, MA) mixed with 250 ml of orange juice. After a three hour equilibration period during which time no food or drink was consumed, a IO ml venous blood sample was drawn, and the plasma was extracted. Plasma water was collected by lyophilization according to Baling.” Plasma water tritium activity was assayed in duplicate with a liquid scintillation counter (Model 3385, Packard Instruments, Downers Grove, IL) using standard techniques. TBW was calculated from the HTO dose administered, corrected for any urinary isotope losses, and the observed tritium activity of the plasma water. The radiation dose incurred with the HTO ingestion was 3 mrem. FFBM was calculated from the TBW value assuming that water constitutes 73.2% of the FFBM:14 FFBM
= TBWl0.732
(7)
where FFBM is in kg and TBW is in L. FM and % BF were calculated using equations’ and.6 Compartmental analysis of the FFBM was derived from the TBK and TBN data using the equations of Burkinshaw et al.4 The theoretical basis of the model and the derivation of these equations are presented in detail in Appendices A and B. The muscle (Mm) and nonmuscle (Mn) components of the FFBM were calculated as follows: Mm = [ 19.25(TBK) Mn = [TBN
-TBN]/38.34
-8.45(TBK)]/20.2
where Mm and Mn are in kg and TBK and TBN are in g.
(8) (9)
Table 1. Physical Characteristics UV= 141 Mean + S.O.
Age. yr Height, cm Body Mass, kg
25.2
+ 4.1
177.1
_t 6.7
73.7
Body Density, g
. cc&
Total Body Potassium, g
RangeLimits 20-30 162.1-184.6
+ a.9
1.0647
58.0-86.4
* 0.0136
1.0383-
1.0897
152.5
+ 19.7
124.9-178.5
Total Body Nitrogen, kg
2.04
+ 0.23
1.68-2.40
Total Body Water, L
47.2
+ 6.0
37.0-51.4
The amount of nitrogen in the muscle (Nm) and the nonmuscle (Nn) compartments was calculated with the equations: Nm = [TBK -0.052
(TBN)]/0.066
(10)
Nn = [O.l I8 (TBN)
pTBK]/0.066
(11)
where Nm, Nn, TBK, and TBN are in kg. When Nm and Nn each are multiplied by 6.25 (assuming the mean nitrogen content of protein is 16%). an estimate of the protein content of muscle (Pm) and nonmuscle (Pn) fractions of the FFBM is obtained. One-factor ANOVA and standard regression analysis were performed on the body composition estimates determined by the various methods.” RESULTS
The physical characteristics of the subjects are given in Table I. Regression equations and correlation coefficients relating the measured values of TBK, TBN, and TBW are presented in Table 2. With the use of the TBK and TBN measurements and equations 8-l 1, the compartmental analysis of the FFB was conducted (Table 3). The mean calculated FFBM of 60.6 * 6.9 kg (mean _+ S.D.) was equal to the sum of 23.4 + 4.8 kg muscle and 37.2 f 4.8 kg nonmuscle FFBM. Thus, the muscle and nonmuscle portions represent 32 and 50% of BM, respectively. The protein content of the muscle portion was 4.39 f 0.91 kg and the protein content of the nonmuscle portion was 8.37 i 1.08 kg. The mean protein content of the FFB was 12.76 * 1.43 kg. Hence, the muscle component contained 34% of the total body protein, and the nonmuscle FFB included the remainder. No significant (P > 0.05) differences were observed among the estimates of FFBM, FM, and % BF determined by densitometry, TBK, TBW, and the combined TBN-TBK analysis (see Table 4). The combination of TBN-TBK predicted a slightly lower
Table 2. Prediction Equations and Correlation Coefficients (r) for Total Body Potassium (TBK), Total Body Water (TBW), and Total Body Nitrogen (TBN)
TEN, kg = 0.45 TBW, L = 3.75 TBN, kg = 0.37
+ 0.01 (TBK. g) + 0.28 (TBK. g) + 0.04 (TBW. L)
lN = 14. P < 0.001
S.D.
r*
0.08
0.94
2 08
0.94
0.06
0.96
779
BODY COMPOSITION AND TOTAL BODY NITROGEN
estimates is questionable. Similarly, failure to measure the residual volume during the hydrostatic weighing procedure can reduce significantly body density measurements and so overestimate body fat.*’ Whereas the experimental errors associated with the densitometry and whole body counting are related to procedural and technical considerations. errors of measurement of body water by isotope dilution are influenced by physiological phenomena. Assuming the subject is euhydrated, the error of the TBW measurement is dependent upon the degree of isotopic hydrogen exchange with nonaqueous exchangeable hydrogen in the body.22m’4 Based upon direct desiccation experiments in rats, Culebras et aLZ5 determined that HTO overestimated TBW by about 2%. This suggests the HTO method slightly underestimates body fat. Another error source inherent in human body composition analysis is the assumption of the validity of the “biological constants” used in the calculation of body composition from various measurements. Body composition determinations in man are complicated by uncertainty because they use indirect methods which are based upon information obtained from the direct chemical analysis of five cadavers of which only three can be considered normal,* plus data from carcass analysis. Thus, the accuracy of any in vivo estimate of body composition in man is unknown. The use of biological constants to derive either the FFBM or FM from primary measurements of body density, TBK, and TWB assumes the relatively constant chemical composition of the FFB and body fat. The FFB is assumed to have a density at 37°C of 1.1 g . cc-’ (1,26), a water content of 72%~-73%,‘7~‘9 and a potassium content of 60-70 mEq . kg- ’ in men and 50-60 mEq - kg-’ in women.-‘” The present study has demonstrated similar estimates of body composition obtained by independent methods. Therefore, using the components measured directly and the FFBM calculated from densitometry, it was possible to derive the chemical composition of the FFB (component/FFBM) of this group of healthy men. Table 6 summarizes the content of the various measured and derived components of the FFB, and shows the relatively constant composition of the FFB.
Table 3. Derived Mass and Protein Content of Fat-Free
Body (FFB) (N = 14)* Mean + SD.
Range Lhmlts
Mass
23.4
5 4.8
16.7-32.7
Nonmuscle, kg
37.2
i 4.8
29.1-45.2
FFB. kg
60.6
* 6.9
49.8-7
Muscle,
kg
1.6
Protein Content Muscle, kg
4.39
f 0.91
3.14-6.16
Nonmuscle, kg
8.37
f 1.08
6.54-10.15
12.76
+ 1.43
10.50-15.00
FFB. kg
*Values were obtamed from equations 8 to 1 1; proteln = N x 6.25; FFB = muscle + nonmuscle.
average value of FFBM, and higher mean values of FM and 5% BF than the other methods. Significant (P < 0.001) correlation coefficients (0.89-0.97) were calculated among the various estimates of body composition (Table 5). DISCUSSION
With the use of absolute measurements of TBK and TBN and the known differential ratio of the nitrogen/ potassium concentrations,4 the muscle and nonmuscle components of the FFB and their protein contents were estimated in a group of healthy men. Comparison of body composition estimates derived from TBNTBK model and those calculated by conventional methods showed no differences among the estimates. Results of studies comparing estimates of body composition in man have demonstrated different values among the various methods. Relative to estimates calculated from measurements of body density and TBW, significantly lower average FFBM and higher body fat values have been derived from ‘OK whole body counting.” ” There are two sources of error involved in any indirect estimate of body composition. Intrinsic errors are due to factors which affect the accuracy of the experimental measurement. For example, determinations of TBK are influenced by individual differences in body size and shape which can affect 40K gamma ray attenuation.’ Thus, absolute TBK levels are not measured and the validity of these body composition Table 4.
Estimates
Densltometry Fat-Free
62.3
kg
Range Lmvts Fat Mass,
Body Range
lFFBM
by Various Methods
(2 + S.D.)
Total Body
Total Body
Potassium
Water
TEN t
TBK
Model’
Body
Mass,
Range
of Body Composition
11.4
kg
15.2
%
+ 5.0
+ 5.6
5.2-26.3
Limits
calculated
7.1
5.0-21.3
Limits Fat,
f
50.7-73.6
as indicated
in Table
3.
62.2
+ 8.0
11.6
i
5.8
4.1-22.7 15.5
*
63.9
*
7.8
50.6-79.7
50.9-74.7
6.8
3.8-28.0
10.4
i
* 6.9
49.8-71.6 5.5
4.1-22.6 13.9
60.6
+ 6.2
5.4-26.5
12.3
+ 5.9
4.1-22.7 16.8
+ 7.0
5.4-28.0
LUKASKI ET AL.
780
Table 5. Correlation Coefficients for Comparison of Body Composition Estimates by Various Methods*
Densitometry
TotalBody Potassium ITBK)
Total Body Water (TBW)
Fat-Free Body Mass, kg Total Body Potassium
0.96
Total Body Water
0.97
0.95
TBN + TBK
0.96
0.96
0.97
Fat Mass, kg Total Body Potassium
0.92
Total Body Water
0.91
0.91
TBN + TBK
0.94
0.92
0.91
Body Fat, % Total Body Potassium
0.89
Total Body Water
0.90
0.89
TBN + TBK
0.91
0.90
0.90
lN = 14, P < 0.001.
On the average, the potassium content of the FFB was 62.6 mEq - kg-’ which is essentially the same as 62.7 mEq - kg-’ reported by Delwaide and Crenier” based upon determinations of FFBM calculated from TBW measurements. However, our value is substantially lower than the commonly used value of 68.1 mEq kg-’ obtained by direct chemical analysis of four cadavers by Forbes and Lewis.3’ Krzywicki et al.19 compared the techniques of densitometry, total body water, and whole-body 40K counting to estimate body composition in 233 male soldiers and reported that FM calculated using 68.1 mEq kg-’ was significantly greater than estimates derived from the other methods. Similarly, other investigations have demonstrated higher estimates of FM derived from 40K measurements using 68.1 mEq - kg-’ in military” and civilian male populations” as compared to independent techniques. Unfortunately, these authors did not provide the individual or mean TBK data necessary to calculate the TBK/FFBM. Our data indicate that the FFB contains 74.6% water. This agrees with the value of 73.5 + 0.04% reported by Loeppky et al.32 who measured TBW by HTO dilution and FFBM by densitometry in 35 men. Because the HTO technique may overestimate TBW by 2%,25 the corrected degree of FFB hydration was Table 6. Concentration of Total Body Potassium (TBK), Total Body Water (TBW), Total Body Nitrogen (TBN), Muscle, and Protein in the Fat-Free Body Mass (FFBM)* Mean+ SD. TBKIFFBM, mEq/kg TBWIFFBM.
L/kg*
62.6 0.746t
f 2.3 rt- 0.020
RangeLimits 59.7-68.7 0.720-0.800
32.74
? 1.09
3 1 .o-35.4
%
37.4
f 5.37
3 1 .O-48.9
Protein/FFBM. %
20.5
? 0.7
19.4-22.0
TBNJFFBM, g/kg Muscle/FFBM,
lFFBM calculated from densitometry according to Brozek et al.’ tlf corrected for non-aqueous hydrogen exchange, this value becomes 0.731.
73.1% which is similar to the direct desiccation value 73.2% observed in some mammals.‘4~23~33-3s Using average values derived from chemical analysis of three male cadavers, Brozek et al.2 calculated the FFB composition of their reference body. By weight, water represented 73.8%, protein 19.4%, and mineral 6.8% of the FFB. Widdowson and Dickerson36 have estimated the average protein content of the FFB to be 19.3%. Our mean value for this relationship was 20.5% which agrees with these estimates. Also, using our corrected value for hydration (73.1%) and protein concentration (20.5%), the mineral content of the FFB was calculated to be 6.4% which is similar to the 6.8% derived by Brozek et al.* Thus, the composition of the FFBM calculated in the present study supports the application of the different “biological constants” currently in use for the study of body composition in man. Furthermore, if TBN determined by neutron activation analysis or other techniques becomes popular, we propose the content of TBN and total body protein in the FFBM is 32.7 g - kg-’ and 204.6 g - kg-‘, respectively. Other investigators have employed a multimeasurement approach to estimate muscle mass. Using measurements of TBK and TBW, Anderson3 derived an average muscle mass of 26% BM and 32% FFBM in young adult men. Burkinshaw et al.4 used measurements of TBK and TBN to calculate a mean value of 28.5% BM and 36% FFBM in a group of 10 men. Similarly, in the present study the muscle mass calculated from data on 14 men aged 20-30 yr was 32% BM and 38% FFBM. Whether this discrepancy is due to differences in the experimental techniques utilized or to genuine differences in the subject groups is not known. The present study has reported the separate determinations of the muscle and nonmuscle components of the FFBM. Because no differences were observed among the estimates of body composition determined by the several independent techniques, it is concluded that the TBN-TBK method provides a valid measure of body composition in man. It is recognized, however, that the N/K ratio varies among the tissues of the nonmuscle component of FFB and that further research is needed to refine the working model. Also, our analysis of the chemical composition of the FFB demonstrated remarkable similarity to that proposed by the pioneers of human body composition analysis. APPENDIX
A
The values used to estimate the muscle and nonmuscle fractions of the fat-free body were taken from Burkinshaw et al.4 He derived the muscle potassium concentration as the weighted mean of 11 different values reported in the literature, with values ranging
BODY COMPOSITION
781
AND TOTAL BODY NITROGEN
from 3.00 to 4.17 g - kg-’ fresh fat-free skeletal muscle. His value for the concentration of nitrogen in muscle was calculated as the mean from the I.C.R.P. “Reference Man” composition,37 and the value determined by Widdowson and Dickerson,36 28.1 and 30.8 muscle, respectively. He g . kg-’ fresh fat-free compiled the potassium and nitrogen concentrations of the nonmuscle fat-free body similarly from the literature.36.37 A useful source of values for both the potassium and nitrogen concentrations in over 100 tissues in the standard man are summarized in Table 108 of the report of the task group on Reference Man3’ A summary of the values used by Burkinshaw et a1.4 is presented in the table below.
mass (Mn) from total body potassium total body nitrogen (TBN) as suggested shaw.4
de
Let: TBK = K; TBN = N; Km and Nm = muscle K and N; Kn and Nn = nonmuscle K and N; rm = Nm/Km; Nm = rmKm, and rn = Nn/Kn, Nn = rnKn. K = Km + Kn (b)
N=Nm+Nn(a) Substituting
in (a); N = rmKm
+ rnKn
from (b); Kn = K - Km then; N = rmKm
+ rn (K -Km)
Solving for Km; Km = (N -rnK)/(rm Concentration of Potassium and Nitrogen in the Muscle and Nonmuscle Portions of the Fat-Free Body
[Nl
fK1
g-kg ’
g . kg-’
Rearranging;
Km = (rnK -N)/(rn
Using the values from Appendix
-rn) -rm)
A:
N/K
Km = (19.25 K -N)/10.8
MUS&
3.55
30
8.45
Nonmuscle
1.87
36
19.25
is important to point out that the potassium concentrations used by Burkinshaw et a1.4 agree quite well with the values proposed by Anderson3 derived from data given by Forbes and Lewis,3’ 3.41 and I .90 components, g - kg-’ for muscle and nonmuscle respectively. The muscle nitrogen concentration is similar to values obtained for other mammals, 32.8 g . kg -‘,38 It
APPENDIX
(TBK) and by Burkin-
Mm = Km/[Km]
= (19.25 K -N)/38.34
Using the same type of derivation, Kn, Mn, Nm and Nn become: Kn = (N - rmK)/(rn
the equations
(8) for
- rm)
= (N -8.45K)/10.8 Mn = Kn/[Kn] Nm = (rmrnK
= (N - 8.45K)/20.2 - rmN)/(rn
- rm)
= (K -0.052N)/0.066
B
Nn = (rmrnK
Derivations of equations 8-l 1 used for the calculation of muscle mass (Mm) and the nonmus-
= (0.118N
- rnN)/(rn - K)/0.066
(9)
(10) - rm) (11)
REFERENCES 1. Keys A, Brozek J: Body fat in adult man. Physiol Rev 33:245-325, 1953 2. Brozek J, Grande F, Anderson JT, et al: Densitometric analysis of body composition: Revision of some quantitative assumptions. Ann NY Acad Sci 1 IO:1 13-140, 1963 3. Anderson EC: Three-component body composition analysis based on potassium and water determinations. Ann NY Acad Sci 110:189-212, 1963 4. Burkinshaw L, Hill CL, Morgan DB: Assessment of the distribution of protein in the human body by in vivo neutron activation analysis, in: International Symposium on Nuclear Activation Techniques in the Life Sciences. Vienna, Int. Atomic Energy Agency, 1918,t.A.E.A.~SM-221139, pp 787-798 5. Vartsky D, Ellis KJ, Cohn SH: In vivo quantification of body nitrogen by neutron capture prompt gamma-ray analysis. J Nucl Med 20:1158%1165, 1979 6. Akers R, Buskirk ER: An underwater weighing system utilizing “force cube” transducers. J Appl Physiol 26:649-652, 1969 7. Darling RC, Cournand A, Richards DW: Studies on the intrapulmonary mixture of gases. III. An open-circuit method for measuring residual air. J Clin Invest 19:609-618, 1940 8. Mendez J, Lukaski H: Variability of body density in ambula-
tory subjects measured in different days. Am J Clin Nutr 34:78-g I, 1981 9. Cohn SH, Dombrowski CS, Pate HR. et al: A whole-body counter with an invariant response to radionuclide distribution and body size. Phys Med Biol 14:6455658, 1969 10. Cohn SH, Dombrowski CS: Absolute measurement of whole body potassium by gamma-ray spectrometry. J Nucl Med 11:239246, 1970 11. Delwaide PA, Crenier EJ: Body potassium as related to lean body mass measured by total water determination and by anthropometric method. Human Biol 45:509-526, 1973 12. Vartsky D, Prestwich WV, Thomas BJ, et al: The use of body hydrogen as an internal standard in the measurement of nitrogen in vivo by prompt gamma neutron capture gamma-ray analysis. J Radioanal Chem 48:243-252, 1979 13. Boling EA: Determination of 4*K, “Na, **Br, and tritiated water concentration in man. Ann NY Acad Sci 110:246-254, 1963 14. Pace N, Rathburn EN: Studies on body composition. Ill. The body water and chemically combined nitrogen content in relation to fat content. J Biol Chem 158:685-691, 1945 15. Keppel G: Design and Analysis: A Researcher’s Handbook. Englewood Cliffs, Prentice-Hall, Inc., 1973 16. Steinkamp RC, Cohn NL, Gaffey WR. et al: Measures of
782
LUKASKI ET AL.
body fat and related factors in normal adults II. A simple clinical method to estimate body fat and lean body mass. J Chron Dis 18:1291-1307, 1965. 17. Myhre measurements 21:1251-1255,
LG. Kessler WV: Body density and potassium-40 of body composition as related to age. J Appl Physiol 1966
18. Murphy HF, Lohman TG, Oscai L, et al: Measurements of body potassium by whole-body counting and its application to physical education research, in Franks BD (ed): Proceedings of a Symposium on Exercise and Fitness. Chicago, Athletic Institute, 1969, pp 225-233 19. Krzywicki HJ, Ward GM, Rahman DP, et al: A comparison of methods for estimating human body composition. Am J Clin Nutr 27:138&1385, 1974 20. Clark DA, Kay TD, Tatsch RF, et al: Estimations of body composition by various methods. Aviat Space Environ Med 48:701704, 1977 21. Katch Fl, Katch VL: Measurement and prediction errors in body composition assessment and the search for the perfect prediction equation. Res Q Exer Sport 5 I :249-260, 1980 22. Culebras JM, Moore FD: Total body water and the exchangeable hydrogen. I. Theoretical calculation of nonaqueous exchangeable hydrogen in man. Am J Physiol: 232:R54-R59, 1977 23. Sheng H, Huggins with emphasis on total 32:63&647, 1979
RA: A review of body composition studies body water and fat. Am J Clin Nutr
24. Nagy KA, Costa DP: Water flux in animals: Analysis of potential errors in the tritiated water method. Am J Physiol: 238:R454pR465, 1980 25. Culebras JM, Fitzpatrick GF, Brennan MF, et al: Total body water and the exchangeable hydrogen. II. A review of comparative data from animals based on isotope dilution and desiccation, with a report of new data from the rat. Am J Physiol: 232:R60-R65, 1977 26. Behnke AR. Osserman
EF. Welham
WC: Lean body mass:
Its clinical significance and estimation from excess fat and total body water determinations. Arch Intern Med 91:685-601. 1953 27. Moore FD, Oleson KH, McMurrey JD, et al: The Body Cell Mass and its Supporting Environment. Philadelphia, Saunders, 1963 28. Novak LP: Total body water in man, in Bergner PE, Lushbaugh CC (eds): Compartments, Pool, and Spaces in Medical Physiology. Oak Ridge Tennessee, U.S. Atomic Energy Commission, 1967, pp 197-216 29. Siri WE: Body composition from fluid spaces and density: Analysis of methods, in Brozek J. Henschel A (eds): Techniques for Measuring Body Composition. Washington, D.C., National Academy of Sciences-National Research Council, 1961, pp 223-244 30. Boddy K, King PC, Womersley J, et al: Body potassium and fat-free mass. Clin Sci 44:6222625, 1973 3 1. Forbes GB, Lewis AM: Total sodium, potassium, and chloride in adult man. J Clin Invest 35:596-600, 1956 32. Leoppky JA, Myhre LG, Venters MD, et al: Total body water and lean body mass estimated by ethanol dilution. J Appl Physiol42:803-808, 1977 33. Pitts GC: Density and composition of the lean body compartment and its relationship to fatness. Am J Physiol 202:4455452, 1962 34. Kodama AM: In vivo and in vitro determinations of body fat and body water in the hamster. J Appl Physiol 31:218-222, 1971 35. Pitts CC, Bullard TR: Some interspecific aspects of body composition in mammals. in Body Composition in Animals and Man. Washington, D.C.; National Academy of Science, Publication #I 598, 1968. pp 45-70 36. Widdowson EM, Dickerson JWT: Chemical composition of the body, in Comar CL, Bronner F (eds): Mineral Metabolism, vol 2, Part A. New York, Academic Press, 1964, pp 22248 37. Report of the Task Group on Reference Man, Int Comm Radio1 Protection Report No. 23. Oxford, Pergamon Press, 1975 38. Mendez, J., Keys A: Density and composition of mammalian muscle. Metabolism 9:184-188, 1960
J. Mendez,
E. R. Buskirk,
and
S. H. Cohn
Fourteen healthy men underwent determinations of total body nitrogen (TBN) by prompt gamma neutron activation analysis and total body potassium (TBK) by whole body counting to estimate the muscle and nonmuscle components of the fat-free body mess (FFBM) and their protein contents. Comparison of FFBM estimated from TBN and TBK (66.6 f 6.9 kg, meen + SD). densitometry (62.3 + 7.1 kg), TBK alone (62.2 + 8.0 kg), and TBW (63.9 + 7.8 kg) showed no differences among the techniques. Similarly, there were neither differences in fat mass nor percent body fat among the methods. Anal@s of the chemical composition of FFBM of this group showed TBK/FFBM = 62.6 + 2.3 mEq/kg. TBW/FFBM = 74.6 + 0.2%. TBN/FFBM = 32.74 f 1.09 g/kg, protein/FFBM = 20.5 + 0.7%. The calculated mineral content of the FFBM wes 6.4%. Theeevekres are strikingly similar to the values calculated by direct chemical analysis. It was concluded that the combinsd TBN-TBK method is a valid technique for estimating body composition in man.
T
HE INCORPORATION of body composition techniques into the practice of medicine, clinical nutrition, and applied physiology has provided normative information on the healthy population. These data are based upon the assumption that the body consists of two chemically distinct compartments, fat and fat-free.‘z If the body mass (BM) is known, determination of the fat-free body mass (FFBM] allows for the calculation of the fat mass (FM) by difference, and vice versa. The most reliable methods available for estimation of FM and FFBM require measurement of body density, total body potassium (TBK), or total body water (TBW). Estimation of body composition by each of these methods is based upon assumptions relating to the relatively constant composition of the FFBM or FM. Anderson3 proposed a three-compartment system of body composition analysis which utilized measurements of TBK and TBW to predict the body fat and the muscle and nonmuscle fractions of the FFBM. More recently, Burkinshaw et al.4 developed equations hased upon the different nitrogen/potassium ratio to estimate the muscle and nonmuscle portions of the FFBM and their respective protein contents. This approach requires determinations of TBK and total body nitrogen (TBN) which now can be measured accurately and safely by the prompt gamma neutron activation teehnique.5 The mass and the protein content of the muscle and nonmuscle FFB can be calculated and so provide a very useful tool to study objectively the effects of various physiological and pathological stresses on human body composition. However, before this method can be implemented in comparative studies, it is important to determine body composition in normal subjects. The purpose of the present study was to obtain baseline data on body composition calculated in several ways in a group of healthy young men and to compare the estimates derived from the TBN-TBK model with those determined by independent methods Me-,
Vd. 30. No. 8 (Aqustl, 198 1
(densitometry, TBK, and TBW). Also, the composition of the FFBM obtained from these various estimates is compared with the composition commonly assumed for the FFBM. MATERIALS
AND
METHODS
Fourteen healthy men, students at the Pennsylvania State University, volunteered to participate in this study. Subjects received a detailed description of the purpose and procedures, and gave their informed consent. This study was approved by the Biomedical Review Committee of the Pennsylvania State University and the Human Subjects Review Committee of the Medical Research Committee of the Brookhaven National Laboratory (BNL). Body composition was determined by densitometry at the Noll Laboratory using the underwater weighing system and the method of Akers and Buskirk. Residual volume was measured simultaneously with the underwater weighing by an open-circuit technique for nitrogen washout of the lungs.’ Percent body fat (% BF) was calculated from body density (D) according to Brozek et al.:* % BF = 100 [(4.570/D) FM and FFBM were calculated follows:
- 4.1421
(1)
from % BF and body mass (BM) as
FM = % BF (BM)/lOO FFBM = BM -FM
(2) (3)
where BM, FM, and FFBM are in kg. FM can be estimated by densitometry with a precision of + 0.5 kg.8 After the density determination, the subjects travelled to the Brookhaven National Laboratory for determinations of TBK, TBN, and TBW. The TBK content of each subject was measured exter-
From the Laboratory for Human Performance Research. The Pennsylvania State University, University Pork. Pennsylvonio. ond the Medical Research Center, Brookhoven Notional Loborotory, Upton. New York. Supported in part by PHS Grants CM0 1748 and AM-08311. NIH Grant RR0 7082, and NIH-NCI Contract YOI-CP-80207. Received for publication August 20. 1980. Address reprint requests to J. Mendez. 101 No11 Laboratory. Pennsylvania State University, University Pork. Pennrylvanio 16802. 0 1981 by Grune & Stratton, Inc. 00260495/81/3008-0007$01.00/0 777
778
LUKASKI ET AL.
nally by detection of 1.46 Mev gamma rays emitted from the naturally occurring radioactive isotope 40K which is a constant fraction (0.012%) of body potassium. The Brookhaven whole-body counter with its on-line computer facility was used to measure absolute levels of TBK. The counter system has a relatively invariant response to individual body size and radionuclide distribution.9 The accuracy and precision for measuring TBK is + 3.3%.r0 It has been determined there is a ratio of 62.7 Meq (2.45 g) potassium per kg FFBM (I I). Calculation of FFBM and FM is: FFBM
= TBK/2.45
(4)
FM = BM - FFBM
(5)
% BF = lOO(FM)/BM
(6)
Absolute levels of TBN were quantitated by using neutron capture prompt gamma ray analysis, IgN(n.y) “N, as described by Vartsky et al.’ Briefly, the capture of a thermal neutron by 14N is followed by the prompt emission (lo-*’ set) of a cascade of gamma rays. About 15% of the de-excitations takes place directly from the highest excited state to the ground state of 15N by emission of a 10.83 Mev gamma ray. Becaase no other body element has a neutron capture gamma ray at this energy, the 10.83 Mev gamma ray of “N is measurable among the more numerous captures by the other body elements. An internal standard based upon body hydrogen reduces the requirement for a highly uniform composite sensitivity (i.e., an equal number of counts per unit mass of nitrogen per unit dose) for absolute measurements.” While lying on a movable bed, subjects were irradiated from shoulder to knee in the prone and supine positions by a 60 x 13 cm rectangular beam of partially moderated neutrons from an 85 Ci 238Pu,Be source. Gamma rays were detected by two 15.24 x 15.24 cm NaI(TI) detectors positioned over the subject. The total measurement time was 20 min, and the radiation dose delivered to the body was 26 mrem. The precision of the TBN estimate is * 3%.5 TBW was determined by the dilution technique using tritiated water (HTO) as the tracer. After an overnight fast, subjects drank 50 &i of HTO (New England Nuclear, Boston, MA) mixed with 250 ml of orange juice. After a three hour equilibration period during which time no food or drink was consumed, a IO ml venous blood sample was drawn, and the plasma was extracted. Plasma water was collected by lyophilization according to Baling.” Plasma water tritium activity was assayed in duplicate with a liquid scintillation counter (Model 3385, Packard Instruments, Downers Grove, IL) using standard techniques. TBW was calculated from the HTO dose administered, corrected for any urinary isotope losses, and the observed tritium activity of the plasma water. The radiation dose incurred with the HTO ingestion was 3 mrem. FFBM was calculated from the TBW value assuming that water constitutes 73.2% of the FFBM:14 FFBM
= TBWl0.732
(7)
where FFBM is in kg and TBW is in L. FM and % BF were calculated using equations’ and.6 Compartmental analysis of the FFBM was derived from the TBK and TBN data using the equations of Burkinshaw et al.4 The theoretical basis of the model and the derivation of these equations are presented in detail in Appendices A and B. The muscle (Mm) and nonmuscle (Mn) components of the FFBM were calculated as follows: Mm = [ 19.25(TBK) Mn = [TBN
-TBN]/38.34
-8.45(TBK)]/20.2
where Mm and Mn are in kg and TBK and TBN are in g.
(8) (9)
Table 1. Physical Characteristics UV= 141 Mean + S.O.
Age. yr Height, cm Body Mass, kg
25.2
+ 4.1
177.1
_t 6.7
73.7
Body Density, g
. cc&
Total Body Potassium, g
RangeLimits 20-30 162.1-184.6
+ a.9
1.0647
58.0-86.4
* 0.0136
1.0383-
1.0897
152.5
+ 19.7
124.9-178.5
Total Body Nitrogen, kg
2.04
+ 0.23
1.68-2.40
Total Body Water, L
47.2
+ 6.0
37.0-51.4
The amount of nitrogen in the muscle (Nm) and the nonmuscle (Nn) compartments was calculated with the equations: Nm = [TBK -0.052
(TBN)]/0.066
(10)
Nn = [O.l I8 (TBN)
pTBK]/0.066
(11)
where Nm, Nn, TBK, and TBN are in kg. When Nm and Nn each are multiplied by 6.25 (assuming the mean nitrogen content of protein is 16%). an estimate of the protein content of muscle (Pm) and nonmuscle (Pn) fractions of the FFBM is obtained. One-factor ANOVA and standard regression analysis were performed on the body composition estimates determined by the various methods.” RESULTS
The physical characteristics of the subjects are given in Table I. Regression equations and correlation coefficients relating the measured values of TBK, TBN, and TBW are presented in Table 2. With the use of the TBK and TBN measurements and equations 8-l 1, the compartmental analysis of the FFB was conducted (Table 3). The mean calculated FFBM of 60.6 * 6.9 kg (mean _+ S.D.) was equal to the sum of 23.4 + 4.8 kg muscle and 37.2 f 4.8 kg nonmuscle FFBM. Thus, the muscle and nonmuscle portions represent 32 and 50% of BM, respectively. The protein content of the muscle portion was 4.39 f 0.91 kg and the protein content of the nonmuscle portion was 8.37 i 1.08 kg. The mean protein content of the FFB was 12.76 * 1.43 kg. Hence, the muscle component contained 34% of the total body protein, and the nonmuscle FFB included the remainder. No significant (P > 0.05) differences were observed among the estimates of FFBM, FM, and % BF determined by densitometry, TBK, TBW, and the combined TBN-TBK analysis (see Table 4). The combination of TBN-TBK predicted a slightly lower
Table 2. Prediction Equations and Correlation Coefficients (r) for Total Body Potassium (TBK), Total Body Water (TBW), and Total Body Nitrogen (TBN)
TEN, kg = 0.45 TBW, L = 3.75 TBN, kg = 0.37
+ 0.01 (TBK. g) + 0.28 (TBK. g) + 0.04 (TBW. L)
lN = 14. P < 0.001
S.D.
r*
0.08
0.94
2 08
0.94
0.06
0.96
779
BODY COMPOSITION AND TOTAL BODY NITROGEN
estimates is questionable. Similarly, failure to measure the residual volume during the hydrostatic weighing procedure can reduce significantly body density measurements and so overestimate body fat.*’ Whereas the experimental errors associated with the densitometry and whole body counting are related to procedural and technical considerations. errors of measurement of body water by isotope dilution are influenced by physiological phenomena. Assuming the subject is euhydrated, the error of the TBW measurement is dependent upon the degree of isotopic hydrogen exchange with nonaqueous exchangeable hydrogen in the body.22m’4 Based upon direct desiccation experiments in rats, Culebras et aLZ5 determined that HTO overestimated TBW by about 2%. This suggests the HTO method slightly underestimates body fat. Another error source inherent in human body composition analysis is the assumption of the validity of the “biological constants” used in the calculation of body composition from various measurements. Body composition determinations in man are complicated by uncertainty because they use indirect methods which are based upon information obtained from the direct chemical analysis of five cadavers of which only three can be considered normal,* plus data from carcass analysis. Thus, the accuracy of any in vivo estimate of body composition in man is unknown. The use of biological constants to derive either the FFBM or FM from primary measurements of body density, TBK, and TWB assumes the relatively constant chemical composition of the FFB and body fat. The FFB is assumed to have a density at 37°C of 1.1 g . cc-’ (1,26), a water content of 72%~-73%,‘7~‘9 and a potassium content of 60-70 mEq . kg- ’ in men and 50-60 mEq - kg-’ in women.-‘” The present study has demonstrated similar estimates of body composition obtained by independent methods. Therefore, using the components measured directly and the FFBM calculated from densitometry, it was possible to derive the chemical composition of the FFB (component/FFBM) of this group of healthy men. Table 6 summarizes the content of the various measured and derived components of the FFB, and shows the relatively constant composition of the FFB.
Table 3. Derived Mass and Protein Content of Fat-Free
Body (FFB) (N = 14)* Mean + SD.
Range Lhmlts
Mass
23.4
5 4.8
16.7-32.7
Nonmuscle, kg
37.2
i 4.8
29.1-45.2
FFB. kg
60.6
* 6.9
49.8-7
Muscle,
kg
1.6
Protein Content Muscle, kg
4.39
f 0.91
3.14-6.16
Nonmuscle, kg
8.37
f 1.08
6.54-10.15
12.76
+ 1.43
10.50-15.00
FFB. kg
*Values were obtamed from equations 8 to 1 1; proteln = N x 6.25; FFB = muscle + nonmuscle.
average value of FFBM, and higher mean values of FM and 5% BF than the other methods. Significant (P < 0.001) correlation coefficients (0.89-0.97) were calculated among the various estimates of body composition (Table 5). DISCUSSION
With the use of absolute measurements of TBK and TBN and the known differential ratio of the nitrogen/ potassium concentrations,4 the muscle and nonmuscle components of the FFB and their protein contents were estimated in a group of healthy men. Comparison of body composition estimates derived from TBNTBK model and those calculated by conventional methods showed no differences among the estimates. Results of studies comparing estimates of body composition in man have demonstrated different values among the various methods. Relative to estimates calculated from measurements of body density and TBW, significantly lower average FFBM and higher body fat values have been derived from ‘OK whole body counting.” ” There are two sources of error involved in any indirect estimate of body composition. Intrinsic errors are due to factors which affect the accuracy of the experimental measurement. For example, determinations of TBK are influenced by individual differences in body size and shape which can affect 40K gamma ray attenuation.’ Thus, absolute TBK levels are not measured and the validity of these body composition Table 4.
Estimates
Densltometry Fat-Free
62.3
kg
Range Lmvts Fat Mass,
Body Range
lFFBM
by Various Methods
(2 + S.D.)
Total Body
Total Body
Potassium
Water
TEN t
TBK
Model’
Body
Mass,
Range
of Body Composition
11.4
kg
15.2
%
+ 5.0
+ 5.6
5.2-26.3
Limits
calculated
7.1
5.0-21.3
Limits Fat,
f
50.7-73.6
as indicated
in Table
3.
62.2
+ 8.0
11.6
i
5.8
4.1-22.7 15.5
*
63.9
*
7.8
50.6-79.7
50.9-74.7
6.8
3.8-28.0
10.4
i
* 6.9
49.8-71.6 5.5
4.1-22.6 13.9
60.6
+ 6.2
5.4-26.5
12.3
+ 5.9
4.1-22.7 16.8
+ 7.0
5.4-28.0
LUKASKI ET AL.
780
Table 5. Correlation Coefficients for Comparison of Body Composition Estimates by Various Methods*
Densitometry
TotalBody Potassium ITBK)
Total Body Water (TBW)
Fat-Free Body Mass, kg Total Body Potassium
0.96
Total Body Water
0.97
0.95
TBN + TBK
0.96
0.96
0.97
Fat Mass, kg Total Body Potassium
0.92
Total Body Water
0.91
0.91
TBN + TBK
0.94
0.92
0.91
Body Fat, % Total Body Potassium
0.89
Total Body Water
0.90
0.89
TBN + TBK
0.91
0.90
0.90
lN = 14, P < 0.001.
On the average, the potassium content of the FFB was 62.6 mEq - kg-’ which is essentially the same as 62.7 mEq - kg-’ reported by Delwaide and Crenier” based upon determinations of FFBM calculated from TBW measurements. However, our value is substantially lower than the commonly used value of 68.1 mEq kg-’ obtained by direct chemical analysis of four cadavers by Forbes and Lewis.3’ Krzywicki et al.19 compared the techniques of densitometry, total body water, and whole-body 40K counting to estimate body composition in 233 male soldiers and reported that FM calculated using 68.1 mEq kg-’ was significantly greater than estimates derived from the other methods. Similarly, other investigations have demonstrated higher estimates of FM derived from 40K measurements using 68.1 mEq - kg-’ in military” and civilian male populations” as compared to independent techniques. Unfortunately, these authors did not provide the individual or mean TBK data necessary to calculate the TBK/FFBM. Our data indicate that the FFB contains 74.6% water. This agrees with the value of 73.5 + 0.04% reported by Loeppky et al.32 who measured TBW by HTO dilution and FFBM by densitometry in 35 men. Because the HTO technique may overestimate TBW by 2%,25 the corrected degree of FFB hydration was Table 6. Concentration of Total Body Potassium (TBK), Total Body Water (TBW), Total Body Nitrogen (TBN), Muscle, and Protein in the Fat-Free Body Mass (FFBM)* Mean+ SD. TBKIFFBM, mEq/kg TBWIFFBM.
L/kg*
62.6 0.746t
f 2.3 rt- 0.020
RangeLimits 59.7-68.7 0.720-0.800
32.74
? 1.09
3 1 .o-35.4
%
37.4
f 5.37
3 1 .O-48.9
Protein/FFBM. %
20.5
? 0.7
19.4-22.0
TBNJFFBM, g/kg Muscle/FFBM,
lFFBM calculated from densitometry according to Brozek et al.’ tlf corrected for non-aqueous hydrogen exchange, this value becomes 0.731.
73.1% which is similar to the direct desiccation value 73.2% observed in some mammals.‘4~23~33-3s Using average values derived from chemical analysis of three male cadavers, Brozek et al.2 calculated the FFB composition of their reference body. By weight, water represented 73.8%, protein 19.4%, and mineral 6.8% of the FFB. Widdowson and Dickerson36 have estimated the average protein content of the FFB to be 19.3%. Our mean value for this relationship was 20.5% which agrees with these estimates. Also, using our corrected value for hydration (73.1%) and protein concentration (20.5%), the mineral content of the FFB was calculated to be 6.4% which is similar to the 6.8% derived by Brozek et al.* Thus, the composition of the FFBM calculated in the present study supports the application of the different “biological constants” currently in use for the study of body composition in man. Furthermore, if TBN determined by neutron activation analysis or other techniques becomes popular, we propose the content of TBN and total body protein in the FFBM is 32.7 g - kg-’ and 204.6 g - kg-‘, respectively. Other investigators have employed a multimeasurement approach to estimate muscle mass. Using measurements of TBK and TBW, Anderson3 derived an average muscle mass of 26% BM and 32% FFBM in young adult men. Burkinshaw et al.4 used measurements of TBK and TBN to calculate a mean value of 28.5% BM and 36% FFBM in a group of 10 men. Similarly, in the present study the muscle mass calculated from data on 14 men aged 20-30 yr was 32% BM and 38% FFBM. Whether this discrepancy is due to differences in the experimental techniques utilized or to genuine differences in the subject groups is not known. The present study has reported the separate determinations of the muscle and nonmuscle components of the FFBM. Because no differences were observed among the estimates of body composition determined by the several independent techniques, it is concluded that the TBN-TBK method provides a valid measure of body composition in man. It is recognized, however, that the N/K ratio varies among the tissues of the nonmuscle component of FFB and that further research is needed to refine the working model. Also, our analysis of the chemical composition of the FFB demonstrated remarkable similarity to that proposed by the pioneers of human body composition analysis. APPENDIX
A
The values used to estimate the muscle and nonmuscle fractions of the fat-free body were taken from Burkinshaw et al.4 He derived the muscle potassium concentration as the weighted mean of 11 different values reported in the literature, with values ranging
BODY COMPOSITION
781
AND TOTAL BODY NITROGEN
from 3.00 to 4.17 g - kg-’ fresh fat-free skeletal muscle. His value for the concentration of nitrogen in muscle was calculated as the mean from the I.C.R.P. “Reference Man” composition,37 and the value determined by Widdowson and Dickerson,36 28.1 and 30.8 muscle, respectively. He g . kg-’ fresh fat-free compiled the potassium and nitrogen concentrations of the nonmuscle fat-free body similarly from the literature.36.37 A useful source of values for both the potassium and nitrogen concentrations in over 100 tissues in the standard man are summarized in Table 108 of the report of the task group on Reference Man3’ A summary of the values used by Burkinshaw et a1.4 is presented in the table below.
mass (Mn) from total body potassium total body nitrogen (TBN) as suggested shaw.4
de
Let: TBK = K; TBN = N; Km and Nm = muscle K and N; Kn and Nn = nonmuscle K and N; rm = Nm/Km; Nm = rmKm, and rn = Nn/Kn, Nn = rnKn. K = Km + Kn (b)
N=Nm+Nn(a) Substituting
in (a); N = rmKm
+ rnKn
from (b); Kn = K - Km then; N = rmKm
+ rn (K -Km)
Solving for Km; Km = (N -rnK)/(rm Concentration of Potassium and Nitrogen in the Muscle and Nonmuscle Portions of the Fat-Free Body
[Nl
fK1
g-kg ’
g . kg-’
Rearranging;
Km = (rnK -N)/(rn
Using the values from Appendix
-rn) -rm)
A:
N/K
Km = (19.25 K -N)/10.8
MUS&
3.55
30
8.45
Nonmuscle
1.87
36
19.25
is important to point out that the potassium concentrations used by Burkinshaw et a1.4 agree quite well with the values proposed by Anderson3 derived from data given by Forbes and Lewis,3’ 3.41 and I .90 components, g - kg-’ for muscle and nonmuscle respectively. The muscle nitrogen concentration is similar to values obtained for other mammals, 32.8 g . kg -‘,38 It
APPENDIX
(TBK) and by Burkin-
Mm = Km/[Km]
= (19.25 K -N)/38.34
Using the same type of derivation, Kn, Mn, Nm and Nn become: Kn = (N - rmK)/(rn
the equations
(8) for
- rm)
= (N -8.45K)/10.8 Mn = Kn/[Kn] Nm = (rmrnK
= (N - 8.45K)/20.2 - rmN)/(rn
- rm)
= (K -0.052N)/0.066
B
Nn = (rmrnK
Derivations of equations 8-l 1 used for the calculation of muscle mass (Mm) and the nonmus-
= (0.118N
- rnN)/(rn - K)/0.066
(9)
(10) - rm) (11)
REFERENCES 1. Keys A, Brozek J: Body fat in adult man. Physiol Rev 33:245-325, 1953 2. Brozek J, Grande F, Anderson JT, et al: Densitometric analysis of body composition: Revision of some quantitative assumptions. Ann NY Acad Sci 1 IO:1 13-140, 1963 3. Anderson EC: Three-component body composition analysis based on potassium and water determinations. Ann NY Acad Sci 110:189-212, 1963 4. Burkinshaw L, Hill CL, Morgan DB: Assessment of the distribution of protein in the human body by in vivo neutron activation analysis, in: International Symposium on Nuclear Activation Techniques in the Life Sciences. Vienna, Int. Atomic Energy Agency, 1918,t.A.E.A.~SM-221139, pp 787-798 5. Vartsky D, Ellis KJ, Cohn SH: In vivo quantification of body nitrogen by neutron capture prompt gamma-ray analysis. J Nucl Med 20:1158%1165, 1979 6. Akers R, Buskirk ER: An underwater weighing system utilizing “force cube” transducers. J Appl Physiol 26:649-652, 1969 7. Darling RC, Cournand A, Richards DW: Studies on the intrapulmonary mixture of gases. III. An open-circuit method for measuring residual air. J Clin Invest 19:609-618, 1940 8. Mendez J, Lukaski H: Variability of body density in ambula-
tory subjects measured in different days. Am J Clin Nutr 34:78-g I, 1981 9. Cohn SH, Dombrowski CS, Pate HR. et al: A whole-body counter with an invariant response to radionuclide distribution and body size. Phys Med Biol 14:6455658, 1969 10. Cohn SH, Dombrowski CS: Absolute measurement of whole body potassium by gamma-ray spectrometry. J Nucl Med 11:239246, 1970 11. Delwaide PA, Crenier EJ: Body potassium as related to lean body mass measured by total water determination and by anthropometric method. Human Biol 45:509-526, 1973 12. Vartsky D, Prestwich WV, Thomas BJ, et al: The use of body hydrogen as an internal standard in the measurement of nitrogen in vivo by prompt gamma neutron capture gamma-ray analysis. J Radioanal Chem 48:243-252, 1979 13. Boling EA: Determination of 4*K, “Na, **Br, and tritiated water concentration in man. Ann NY Acad Sci 110:246-254, 1963 14. Pace N, Rathburn EN: Studies on body composition. Ill. The body water and chemically combined nitrogen content in relation to fat content. J Biol Chem 158:685-691, 1945 15. Keppel G: Design and Analysis: A Researcher’s Handbook. Englewood Cliffs, Prentice-Hall, Inc., 1973 16. Steinkamp RC, Cohn NL, Gaffey WR. et al: Measures of
782
LUKASKI ET AL.
body fat and related factors in normal adults II. A simple clinical method to estimate body fat and lean body mass. J Chron Dis 18:1291-1307, 1965. 17. Myhre measurements 21:1251-1255,
LG. Kessler WV: Body density and potassium-40 of body composition as related to age. J Appl Physiol 1966
18. Murphy HF, Lohman TG, Oscai L, et al: Measurements of body potassium by whole-body counting and its application to physical education research, in Franks BD (ed): Proceedings of a Symposium on Exercise and Fitness. Chicago, Athletic Institute, 1969, pp 225-233 19. Krzywicki HJ, Ward GM, Rahman DP, et al: A comparison of methods for estimating human body composition. Am J Clin Nutr 27:138&1385, 1974 20. Clark DA, Kay TD, Tatsch RF, et al: Estimations of body composition by various methods. Aviat Space Environ Med 48:701704, 1977 21. Katch Fl, Katch VL: Measurement and prediction errors in body composition assessment and the search for the perfect prediction equation. Res Q Exer Sport 5 I :249-260, 1980 22. Culebras JM, Moore FD: Total body water and the exchangeable hydrogen. I. Theoretical calculation of nonaqueous exchangeable hydrogen in man. Am J Physiol: 232:R54-R59, 1977 23. Sheng H, Huggins with emphasis on total 32:63&647, 1979
RA: A review of body composition studies body water and fat. Am J Clin Nutr
24. Nagy KA, Costa DP: Water flux in animals: Analysis of potential errors in the tritiated water method. Am J Physiol: 238:R454pR465, 1980 25. Culebras JM, Fitzpatrick GF, Brennan MF, et al: Total body water and the exchangeable hydrogen. II. A review of comparative data from animals based on isotope dilution and desiccation, with a report of new data from the rat. Am J Physiol: 232:R60-R65, 1977 26. Behnke AR. Osserman
EF. Welham
WC: Lean body mass:
Its clinical significance and estimation from excess fat and total body water determinations. Arch Intern Med 91:685-601. 1953 27. Moore FD, Oleson KH, McMurrey JD, et al: The Body Cell Mass and its Supporting Environment. Philadelphia, Saunders, 1963 28. Novak LP: Total body water in man, in Bergner PE, Lushbaugh CC (eds): Compartments, Pool, and Spaces in Medical Physiology. Oak Ridge Tennessee, U.S. Atomic Energy Commission, 1967, pp 197-216 29. Siri WE: Body composition from fluid spaces and density: Analysis of methods, in Brozek J. Henschel A (eds): Techniques for Measuring Body Composition. Washington, D.C., National Academy of Sciences-National Research Council, 1961, pp 223-244 30. Boddy K, King PC, Womersley J, et al: Body potassium and fat-free mass. Clin Sci 44:6222625, 1973 3 1. Forbes GB, Lewis AM: Total sodium, potassium, and chloride in adult man. J Clin Invest 35:596-600, 1956 32. Leoppky JA, Myhre LG, Venters MD, et al: Total body water and lean body mass estimated by ethanol dilution. J Appl Physiol42:803-808, 1977 33. Pitts GC: Density and composition of the lean body compartment and its relationship to fatness. Am J Physiol 202:4455452, 1962 34. Kodama AM: In vivo and in vitro determinations of body fat and body water in the hamster. J Appl Physiol 31:218-222, 1971 35. Pitts CC, Bullard TR: Some interspecific aspects of body composition in mammals. in Body Composition in Animals and Man. Washington, D.C.; National Academy of Science, Publication #I 598, 1968. pp 45-70 36. Widdowson EM, Dickerson JWT: Chemical composition of the body, in Comar CL, Bronner F (eds): Mineral Metabolism, vol 2, Part A. New York, Academic Press, 1964, pp 22248 37. Report of the Task Group on Reference Man, Int Comm Radio1 Protection Report No. 23. Oxford, Pergamon Press, 1975 38. Mendez, J., Keys A: Density and composition of mammalian muscle. Metabolism 9:184-188, 1960