European Journal of

Europ. J. Appl. Physiol. 36, 141-150 (1977)

Applied Physiology and Occupational Physiology 9 by Springer Verlag 1977

Estimation of Body Density and Lean Body Weight from Body Measurements at High Altitude Hari Bharadwaj, S. S. Verma, T. Zachariah, M. R. Bhatia, S. Kishnani, and M. S. Malhotra Defence Institute of Physiology and Allied Sciences, Delhi Cantt-ll0010, India

Abstract. Body density and other anthropometric data were obtained on 101 Indian soldiers who were continuously staying at high altitude (3920 m) for more than 10 months. Use was made of a human body volumeter, and body density was calculated from observed body weight and volume. Measurements were taken on the body using standard techniques. A stepwise linear regression analysis was performed to establish possible relationships of 36 body measurements with density and lean body weight. Thigh anterior, juxta-nipple skin folds and forearm and ankle circumferences were selected in the regression equation predicting body density. Multiple correlation coefficient (R) equal to 0.765 was obtained for this equation. For the predicted lean body weight, R equalled 0.930. The regression equations included body weight, thigh anterior and juxta-nipple skin fold thicknesses, and forearm circumference. Contribution of other body measurements in the regression of these parameters was not significant. The analysis also revealed that a new set of coefficients is required for the measurements included in the published regression equations. Key words: Body density - Anthropometry -- High altitude. Introduction Estimates of total body fat in the living individual are made from the knowledge of body density. Since the measurement of body density involves a tedious procedure, reasonably good estimates of body density are now made from the thickness of skin folds chosen at selected sites or even from some conventional anthropometric measurements. Large number of studies have been done on this aspect (Brozek, 1951; Pascale et al., 1956; Durnin and Rahaman, 1967; Sloan, 1967; Wilmore and Behnke, 1969) by Europeans and Americans, suggesting useful regression equations from which body density can be estimated. There is meagre evidence, however, that the published equations are applicable to subjects of varying racial groups. Use of the thickness of skinfolds at abdominal, triceps and chest sites for this purpose was examined by Sen and Banerjee (1958) in young Bengali College students. They

142

H. Bharadwaj et al.

concluded that the equation proposed by Brozek could be safely used for predicting body density from such measurements. All these studies, however, have been carried out at sea-level situations only, and the validity of regression equations have not been examined at high-altitude locations, particularly when men have spent considerable time there. The necessity for such a study was felt, when it was realised that body composition changes occur as a result of prolonged exposure to high altitudes (Bharadwaj, 19 7 2 - 1 9 74). Alter ations in body composition as a result of brief (8 days) but abrupt exposure to high altitude (4300 m) were studied by Surks et al. (1966) who concluded that the observed weight loss was due to loss of body fat. The authors also noticed significant shifts in protein muscle to the non-muscle fraction, although total body proteins remained unaltered. Consolazio et al. (1968) observed negative nitrogen and water balance after 4 weeks' exposure to the same altitude. Krzywicky et al. (1969) reported losses in body fat, protein, water and mineral contents in subjects exposed for 12 days at the same altitude. Although significant body composition changes have been observed among sealevel residents during short sojourns at high altitudes, paradoxically, the high-altitude natives do not appear to be much different with respect to body composition, as compared to the residents at sea-level. Siri et al. (1954) did not observe a significant difference in the amounts of body fat and water content of the high-altitude natives and sea-level residents. Picon et al. (1961) observed no difference in fat, bone mineral and cell solid contents between high-altitude (4300 m) natives and sea-level residents. The same authors, however, noticed a significantly increased extraeellular space among the high-altitude natives. Since the knowledge of body density provides the key to the secrets of body composition, in view of significant alterations in various compartments of the body at high altitude, its assessment by means of simple body measurements needs critical reappraisal.

Materials and Methods

A random sample of 101 young (18--30 years) and healthy soldiers of mixed ethnic origin was selected. The soldiers had been residing at altitudes ranging from 3960-4270 m continuously for 10 months. They were given a diet providing 4830 kcal per day. This contained 144 g protein, 747 g carbohydrate and 138 g fat, besides other essential nutrients. The study was carried out at a location 3920 m above sea level. Measurement of Body Density. The studies were conducted in the morning before the subjects had breakfast. Their body volume in full inspiration was measured by using the water displacement technique of Jones (1972). Residual lung volume and vital capacity (while submerged chin deep in water) were subtracted from total body volume. Body density was calculated by dividing the body weight in air by the volume so obtained. Residual lung volume was obtained by a modification of the three-breath nitrogen dilution method (Durnin and Rahaman, 1967; Rahn et al., 1949). Using experimentally determined body density, fat content in kg was estimated according to the formula of Brozek (1963), i.e., (4.57/D -- 4.142) x Body weight (kg). Lean body weight was obtained by subtracting total body fat from weight. Anthropometry. Stature was measured by Martins' anthropometer and nude body weight was taken on a sensitive Avery Beam Balance, capable of detecting a 25 g change in body weight. The thickness of

Estimation of Body Density and Lean Body Weight at High Altitude

143

skinfolds at various sites was measured with a Harpenden skinfold caliper exercising a pressure of 10 g/mm2. Body girths were measured by a flexiblesteel tape. Body widths were measured by a sliding caliper. In all, 36 body measurements were involvedin the study and are describedin Table 2. Measurements listed at serial Nos. 3, 5, 6, 8--10 and 12--23 were taken according to the procedure of Behnke (1961). The technique of Tanner et al. (1969) was adopted for those given at serial Nos. 4, 24-27, 29, 31 and 33-36 in the table. The remaining measurements were taken as follows: Head Circumference H, Circumference around vertex and point of chin. Thigh Circumference. Circumference at one-third subischial (stature-sitting height) distance above the knee joint. Abdomen H Skinfold. Thickness of the horizontal fold at the level of the navel and perpendicularly below the nipple. Thigh Posterior Skinfold. Thickness of vertical skinfoldon the posterior aspect of the thigh at one-third subischial distance above the knee joint. Calf Lateral Skinfold. The thickness of the skinfold in the lateral aspect at maximum calf girth level. The measurements involved in the published regression equations, which were tested at high altitude, were taken in the same way as described by the respective authors. Statistics. A stepwise linear regression analysis was performed on an IBM 360/44 computer. A sequence of multiple regression equations of the form: Y~-

a 0 + b I x 1 + b 2 x 2 + ...

bt xi...b p x,

were fitted in a stepwise manner to the data. The dependent variables were body density and lean body weight. At each step of the computation, one independent variable xi (i = 1, 2 . . . 36) was added to the regression equation. The variable added was the one which made the greatest reduction in the error sum of squares. It was the variable that had the highest partial correlation with the dependent variable, given that the variables preceding it had already been included in the regression equation.

Results The anthropometric characteristics of the combined sample are given in Table 1. In addition to the means and standard deviations of 36 body measurements, the table also includes the means and standard deviations for body density and lean body weight which were experimentally determined. The correlation coefficients of 36 body measurements with body density and lean body weight are shown in Table 2. The level of significance for each correlation coefficient is indicated separately. Intercorrelations of body measurements were also computed but these have not been shown for the sake of brevity. Table 3 gives the regression equations for predicting body density, the multiple correlation coefficient (R) and the standard error of the estimate for each equation. It also depicts the standard error of partial regression coefficients (bl, b z, b 3, b 4 and bs). Out of the 36 body measurements selected for the study, the stepwise linear regression technique has chosen thigh anterior, juxta-nipple, and biceps skinfold thicknesses and forearm and ankle circumferences which contribute significantly to estimating body density. The level of significance for each of the independent variables in every equation has been indicated.

144

H. Bharadwaj et al.

Table 1. Anthropometric characteristics of 101 troops at high altitude Subject

Variables

Units

Sample

no.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Body density Lean body weight Body weight Body height Shoulder circumference Sitting height Chest circumference Upper arm circumference Thigh circumference Knee circumference Calf circumference Ankle circumference Head circumference II Neck circumference Abdomen I circumference Abdomen II circumference Buttock circumference Forearm circumference Wrist circumference Biacromial width Bi-iliac width Elbow width Wrist width Knee width Ankle width Juxta-nipple skinfold Mid axillary skinfold Forearm skinfold Abdomen I skinfold Abdomen II skinfold Thigh anterior skinfold Thigh posterior skinfold Calf medial skinfold Calf lateral skinfold Biceps skinfold Triceps skinfold Subscapula skinfold Supra-iliac skinfold

g/ml kg kg cm cm cm cm cm cm cm cm cm cm cm cm cm em cm cm cm cm cm cm cm cm cm cm cm cm em cm cm cm cm cm cm cm cm

Mean

S.D.

1.0817 52.93 57.85 168.26 105.17 88.36 86.57 25.51 47.74 31.44 33.19 20.32 64.73 34.51 72.46 74.60 87.06 25.01 15.90 37.85 27.34 6.78 5.59 9.01 6.98 0.97 0.65 0.55 1.45 0.71 0.92 1.22 0.69 0.70 0.34 0.75 0.97 0.46

0.0119 5.14 6.11 5.96 4.18 2.59 3.84 1.74 2.70 1.49 1.90 1.02 2.01 1.53 4.97 5.73 3.88 1.22 0.63 1.70 1.28 0.36 0.29 0.39 0.35 0.47 0.22 0.15 0.94 0.39 0.41 0.46 0.28 0.24 0.06 0.29 0.31 0.15

T h e c o n t r i b u t i o n o f o t h e r b o d y m e a s u r e m e n t s in e s t i m a t i n g b o d y d e n s i t y is n o t s i g n i f i c a n t in f u r t h e r steps o f r e g r e s s i o n a n a l y s i s a n d h e n c e t h o s e s t e p s h a v e n o t b e e n s h o w n in t h e table. T h e i n c l u s i o n o f g r e a t e r n u m b e r o f i n d e p e n d e n t v a r i a b l e s in t h e r e g r e s s i o n e q u a t i o n o n l y m a k e s t h e e q u a t i o n unwieldly. T a b l e 4 gives t h e m u l t i p l e r e g r e s s i o n e q u a t i o n s for p r e d i c t i n g l e a n b o d y weight. O n l y f o u r r e g r e s s i o n e q u a t i o n s h a v e b e e n s h o w n in this t a b l e b e c a u s e o t h e r b o d y m e a s u r e m e n t s d o n o t c o n t r i b u t e s i g n i f i c a n t l y i n t h e e s t i m a t i o n o f l e a n b o d y weight.

Estimation of Body Density and Lean Body Weight at High Altitude

145

Table 2. Correlation of body measurements with body density and lean body weight Subject no.

Variables

Units

Coefficient of correlation (r) with body density

Coefficient of correlation (r) with L B W

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

Body weight Body height Shoulder circumference Sitting height Chest circumference Upper arm circumference Thigh circumference Knee circumference Calf circumference Ankle circumference Head circumference II Neck circumference A b d o m e n I circumference A b d o m e n II circumference Buttocks circumference Forearm circumference Wrist circumference Biacromial width Bi-ihac width Elbow width Wrist width Knee width Ankle width Juxta-nipple skinfold Mid-axillary skinfold Forearm skinfold A b d o m e n I skinfold Abdomen II skinfold Thigh anterior skinfold Thigh posterior skinfold Calf medial skinfold Calf lateral skinfold Biceps skinfold Triceps skinfold Subscapula skinfold Supra-iliac skinfold

kg cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm cm

--0.379 e -0.183 -0.189 -0.099 --0.264 b -0.254" -0.321 b --0.299 b --0.144 -0.241 a -0.159 0.005 -0.475 c -0.539 c --0.364 e 0.007 --0.105 0.040 -0.040 -0.035 0.038 -0.156 -0.027 -0.647 ~ -0.608 e --0.492 ~ -0.607 e -0.363 ~ -0.663 c -0.556 ~ -0.616 ~ -0.488 ~ -0.565 e -0.640 c -0.593 e -0.544 c

0.850 e 0.459 c 0.791 ~ 0.506 c 0.696 c 0.653 r 0.734 c 0.651 c 0.606 e 0.505 ~ 0.437 e 0.727 c 0.564 ~ 0.582 e 0.747 c 0.763 c 0.651 e 0.456 c 0.542 e 0.524 c 0.501 e 0.561 c 0.604 c 0.238 a 0.266 b 0.365 ~ 0.322 b 0.053" 0.235" 0.198" 0.295 b 0.240" 0.341 e 0.300 b 0.319 b 0.337 e

a

p < 0.05

b p < 0.01

The measurements

e p < 0.001

which enter these equations

a r e (1) b o d y w e i g h t , (2) t h i g h a n t e -

r i o r s k i n f o l d , (3) j u x t a - n i p p l e s k i n f o l d , a n d ( 4 ) f o r e a r m c i r c u m f e r e n c e . T h e t a b l e a l s o shows the multiple correlation coefficient R, the standard error of the estimate and the standard

e r r o r o f t h e p a r t i a l r e g r e s s i o n c o e f f i c i e n t s f o r all t h e e q u a t i o n s p r e d i c t -

ing lean body Correlation and lean body

weight. coefficients between observed and predicted values of body density weight in different situations have been presented

in T a b l e 5. T h e

oo

II

to e-* L~

-.-.I

I oo .o I .o

g

4~

0 9 4~ oo

0

(-D

~" 0,9,~ ,

g

0.725 ~

1 . 0 5 2 2 - - 0 . 0 1 2 0 X~ e -- 0 . 0 0 9 7 )(2 ~ + 0 . 0 0 2 0 X3 b

2.

3.

1.0741 -- 0 . 0 0 8 8 X1 b -- 0 . 0 0 8 6 X2 ~

+ 0 . 0 0 3 3 X3 ~ - - 0 . 0 0 2 3 X4 ~ -- 0 . 0 3 9 2 X5 a

0.766 ~

0.0078

0.0080

0.0083

0.0086

0.0089

Standard error of estimate

0.0031

0.0031

0.0031

0.0032

0.0022

bl

0.0028

0.0027

0.0028

0.0028

.

b2 .

.

0.0008

0.0008

0.0007

--

ba .

Standard error of partial regression coefficients

0.0010

0.0009

--

b4

0.0178

--

b5

P < 0.05 b P < 0.01 c p < 0.001 X 1 = t h i g h a n t e r i o r skinfold; X 2 = j u x t a - n i p p l e s k i n f o l d ; X 3 = f o r e a r m c i r c u m f e r e n c e ; X 4 = a n k l e c i r c u m f e r e n c e ; X 5 = b i c e p s skinfold; r - linear coefficient of correlation; R = multiple linear coefficient of correlation

"

5.

+ 0 . 0 0 3 2 )(3 ~ -- 0 . 0 0 2 8 X4 b

0.752 e

0.698 ~

1 . 1 0 0 8 -- 0 . 0 1 1 7 X~ e - - 0 . 0 0 8 5 X2 b

1 . 0 7 8 0 -- 0 . 0 1 0 2 X1 c -- 0 . 0 1 0 6 X2 ~

0.663 ~

1 . 0 9 9 4 - - 0 . 0 1 9 1 ](1 ~

I.

4.

r or R

Regression equation B o d y d e n s i t y (Y)

E q . no.

T a b l e 3. M u l t i p l e r e g r e s s i o n e q u a t i o n s f o r p r e d i c t i n g b o d y d e n s i t y f r o m b o d y m e a s u r e m e n t s

.m

taa t'O

+

t.ah bO

9

~

~ro

9

b-t

~'+

I

t'~

~

b

+

9

r~

0~"

g

bo

b

rgl

-

0.930 r

0.924 ~

0.913 ~

0.850 c

r or R

1.925

1.997

2.115

2.721

Standard error of estimate

0.0592

0.0420

0.0430

0.0446

bl

0.7757

0.7819

0.6375

--

b~

0.6577

0.6811

b~

Standard error of partial regression coefficients

X I = b o d y weight; X 2 = t h i g h a n t e r i o r skinfold; X 3 = j u x t a - n i p p l e skinfold; X 4 = f o r e a r m c i r c u m f e r e n c e ; r = linear coefficient o f c o r r e l a t i o n ; R = multiple linear coefficient o f c o r r e l a t i o n

" P < 0.05 b P < 0.01 c p < 0.001

e

- 7 . 9 6 4 3 + 0.8364XIC--2.8503Xf 2 . 3 3 6 0 X3 ~ + 0 . 6 9 6 1 3 2 4 b

2.8100 + 0.9615XI~--3.3802X2 2.4488X3 ~

3.

--

4 . 3 2 5 7 + 0 . 9 2 2 7 X I ~ -- 5 . 1 7 4 1 X f

11.500 + 0.7162X1 ~

2.

I.

Regression equation Lean body weight (I9

4. Multiple r e g r e s s i o n e q u a t i o n s for p r e d i c t i n g lean b o d y w e i g h t f r o m b o d y m e a s u r e m e n t s

Eq. no.

Table

0.2405

m

b,

e~

>

~zr

0~

o

o~

v

O e'~ ,.