Contents - Previous - Next

Assessment and significance of body composition in infants and children

1. Techniques for estimating body composition
2. Changes in body composition during growth
3. The companionship of lean body mass and fat
4. Maintenance energy need is related to body size and composition
5. The energy cost of weight gain
Discussion (summarized by A. Ferro-Luzzi)


* Department of Pediatrics, University of Rochester School of Medicine and Dentistry, Rochester, New York, NY, U.S.A.


Following a brief presentation of the techniques used to estimate body composition, this paper deals with the changes in lean body mass (LBM) and body fat during normal growth, the effect of body composition changes on nutritional requirements and on the energy cost of growth, and the relationship between body composition and maintenance energy needs in adolescents and adults. In most situations involving significant changes in body weight, both body components participate in the change, and in the same direction, although not always in the same proportion.

The availability of modern techniques has made possible the study of certain aspects of body composition in living human beings with a minimum of discomfort to the subject. Such techniques permit the assessment of body composition at a given point in time as well as changes which may occur under various circumstances. The result has been the accumulation of much data on normal subjects, on those with various disease states, and on the perturbations induced by stimuli of various types (FORBES, 1987a). While current techniques were developed for use in adolescents and adults, several have been applied to the study of infants and young children. This paper will deal with several aspects of the field of body composition, with particular emphasis on the lean and fat components of the body.

1. Techniques for estimating body composition

Changes in body composition can be assessed by means of the metabolic balance technique, using nitrogen balance as an estimate of change in lean body mass. Many years ago BENJAMIN (1914) used this technique in infants, but was disappointed to find that the calculated increase in body protein during growth was far greater than expected. The problem is that retention of nitrogen, and also other elements, tends to be overestimated due in part to failure to take into account cutaneous losses. If, however, the balance is strongly positive, as in rapidly growing prematurely born infants or in infants and children recovering from malnutrition, such errors become less important though they never disappear completely. Examples of the use of the balance technique can be found in the publications of WATERLOW (1961), REICHMAN et al. (1981) and FJELD et al. (1989). Using energy balance as well as nitrogen balance, Reichman's group, for instance, was able to deduce the relative contributions of fat, carbohydrate, and protein to the weight gain of prematurely born infants.

However, the metabolic balance method can only provide information on the change in body content, not body content per se. Recent decades have witnessed the development of techniques for estimating body content of lean and fat, of body fluid volumes, and of skeletal mass in living humans. Estimates of lean and fat can be made by potassium-40 counting, by densitometry, and by assessment of total body water. While the last method, employing either deuterium or oxygen-18 dilution, can be applied to infants and children as well as adults, the first two are beset with technical difficulties in their use for the very young. A newborn infant, for instance, contains only about five grams of potassium, which means that the 40K signal is but a small fraction of the background noise, so that there is lack of precision. Satisfactory assays have been made, however, in 4- to 5-year-old children (FLYNN et al., 1970). The underwater weighing technique for determining body density, so commonly used in studies of adolescents and adults, is obviously unsuited to infants and young children. Attempts to determine body volume in young infants by the application of Boyle's law have not been successful thus far.

All of these methods carry the assumption that the composition of the lean body mass (potassium, water, and density) is known. The fact that all three moieties change during growth, and especially in young infants, raises problems in calculating the results.

A new technique is now under investigation. This is dual photon absorptiometry, which involves scanning the body with two gamma rays differing in intensity which emanate from a radioactive source. The claim is that computer analysis of the transmitted gamma rays can generate values for non-osseous lean tissue, skeletal mass, and total body fat. Adult subjects are required to lie still for 60 minutes, but PETERSEN et al. (1988,1989) have reported reasonable results using a 20-minute scan. The radiation dose is about 5 mR.

The ratio of extracellular fluid volume to intracellular fluid volume also changes during growth. As shown in Figure 1, this ratio is high in the fetus, and falls progressively during infancy and childhood to reach its lowest point in the early adult years, to be followed by a gradual increase during aging.

Figure 1. Ratio of ECF/ICF volume, plotted against the logarithm of postconception age, from the 6-month fetus to the 80-year-old adult. Combined sexes. B = birth. Data compiled from several sources. From FORBES (1987a) with permission.

COKINGTON et al., (1963) have studied the changes in body fluid volumes of infants during recovery from malnutrition. The ratio of ECF/total water is high at first, and falls to normal during recovery. The rate of fall in the ratio is more rapid on feedings providing 15% of energy from protein than on those providing 9%. On the other hand, MANNA (1963) in his studies of normal infants found no difference in body fluid volumes between infants given human milk (which has a relatively low protein content) and those fed artificial formulas. KAGAN et al. (1963) found that prematurely born infants fed a high-protein-high-ash formula gained more weight and had a higher ECF/ICF ratio than those fed human milk.

Two new techniques for estimating total body water are now being applied to the study of young infants. The total body electrical conductivity (TOBEC) method measures the changes in an electromagnetic field (5-10 MHz, 7 mW/cm2) when an object which conducts electricity, such as the human body, is placed in the field (COCHRAN et al., 1986). The bioimpedance method measures the resistance to the passage of a weak electrical current (100mA, 50 kHz) through the body (FJELD et al., 1990). Both measurements can be made very quickly, and neither carries an appreciable hazard. Estimates of the density of individual bones can be made by the technique of photon absorptiometry. The particular bone to be studied (usually the radius and/or ulna) is scanned by gamma rays emanating from a radioactive source, whence the attenuation of the rays is proportional to the amount of bone mineral present. Single photon absorptiometry, i.e., one monoenergetic gamma ray, has been used for single bones of very young infants (GREER et al., 1983), and the dual photon technique, i.e., two gamma rays of different energies, has been used to scan the entire skeleton of young infants (PETERSON et al., 1989). The former technique requires only a minute or so, while the latter requires the subject to lie still for 20 minutes. The radiation dose is 5-13 mR.

Using this technique it has been shown that prematurely born infants, and in some studies term infants, fed human milk with its relatively low Ca and P contents, have a lower bone density than those fed infant formula (CHAN et al., 1986). Thus, the early work of STEARNS (1939), who used the metabolic balance technique, has been confirmed.

Mention should also be made of the techniques of anthropometry, especially the measurements of subcutaneous fat-plus-skin thicknesses by means of special calipers, providing a gross and far from precise estimate of body fat. Urinary creatinine excretion provides an estimate of skeletal muscle mass (for references see FORBES, 1987a).

2. Changes in body composition during growth

While girls are only slightly fatter than boys at birth and throughout infancy and childhood, the sex difference in body composition becomes significant during adolescence. The adolescent spurt in weight and stature is accompanied by increases in lean body mass (LBM) and body fat; the boy acquires considerably more LBM than the girl, while she accumulates a larger percent body fat. Between the age of 10 and 20 years, the average boy puts on 33 kg LBM, the average girl only 16 kg. This sex difference remains when LBM is referred to stature: adolescent boys have a higher LBM:height ratio than girls. Not only is the adolescent spurt in LBM more intense in the male, but it also lasts longer; maximum LBM is achieved by the girl at about 18 years, but in the boy not until about age 20 (Figure 2).

Figure 2. Upper section: Mean body weight and lean body mass (LBM) during late childhood and adolescence. Lower section: Estimated LBM velocity; note the more intense and prolonged LBM spurt in boys. Author's data by potassium-40 counting.

This sex difference in LBM growth has obvious nutritional implications. Based on available data, the average boy will accumulate during the second decade of life 770 g Ca, 2.08 g Fe, 0.98 g Zn, and 7300 g protein. For the average girl these values are much less, namely 400 g Ca, 0.84 g Fe, 0.66 g Zn, and 3650 g protein. The sex difference is similarly striking for the estimated daily increments in body content at the peak of the adolescent growth spurt: 400 mg Ca for the boy versus 240 mg for the girl, 3.8 g protein for the boy, 2.25 g for the girl.

3. The companionship of lean body mass and fat

While the composition of the LBM is assumed to be fairly constant from mid-adolescence through most of the adult years, its size is not invariant. For example, tall individuals of all ages (including infants) have on average a larger LBM than those who are short; boys a larger LBM than girls (there is even a slight sex difference in infants), men a larger LBM than women.

It should be recognized that in many situations involving significant changes in body weight, both lean and fat participate in the weight change. With but few exceptions, a change in one component is accompanied by a change in the other, and in the same direction, though not always in the same proportion. In a sense, then, lean and fat behave as true companions.

The data in Table 1, compiled from the literature, show that the companionship rule holds for a number of situations: intentional underfeeding and overfeeding; spontaneous changes in body weight, i.e., free-living individuals on self-selected diets; and infants born to diabetic mothers. Whenever there is a significant change in body weight, LBM and fat both participate in the change; and of course normal growth involves a change in both LBM and fat.

Table 1. Composition of weight change


N, sex




Overfed adults

30 M, 18 F

+3 to 14 kg


density, K-40, N balance

Spontaneous weight gain (obese subjects)

1 M, 7 F

+6 to +22 kg


density, K-40

Spontaneous weight change (adults)

66 F, 60 M

-16 to +31 kg

0.38 *


Spontaneous weight change (lactating women)

37 F

-7 to +7 kg

0.29 *


Seasonal weight change (adults)

50 F

-5.0 to +2.1 kg


total body water

Induced weight change (adults)

14 M

-3.7 to +3.5 kg

0.38 *

N balance

Infants of diabetic mothers vs controls


+700 9

0.31 **

carcass analysis

* regression slope
** calculated for 38th week of gestation
References are given in FORBES, 1990.

However, not all individuals exhibit the same compositional response to weight changes induced by energy deficit or surfeit. Generally speaking, thin individuals lose proportionally more LBM and less fat during weight reduction than do obese individuals. This is why obese people have a greater tolerance for fasting. Likewise, thin people tend to gain a larger proportion of lean in response to high-energy diets with adequate protein content than do those who are obese. While the evidence for the companionship concept has been adduced from studies of adolescents and adults, it is entirely reasonable to assume that it pertains also to infants and children. Certainly, recovery from infant and childhood malnutrition involves an increase in both LBM and fat.

The companionship rule also applies to established states of under- and overnutrition. The vast majority of obese individuals, children and adults alike, have a supranormal LBM, which constitutes a quarter to a third of their excess weight. Individuals with cystic fibrosis and those with anorexia nervosa have a reduced LBM as well as less body fat (FORBES, 1987a, 1990).

However, there are a few pathological exceptions to the companionship rule. Obese children with the Prader-Willi syndrome have a reduced LBM in the face of an increased body fat content (FORBES, 1990). In this respect, this condition differs from the usual type of exogenous obesity. Such patients are known to have muscle hypotonia and hypogonadism, and it has been found that weight reduction can be achieved only with very low energy diets.

Rats who develop obesity as a result of experimentally produced lesions of the hypothalamus also exhibit a reduced LBM, reduced bone growth, and reduced body length in the face of massive accumulations of body fat; hence they do not provide a good model for the usual form of human obesity.

Another exception is the effect of anabolic steroids. These compounds promote nitrogen retention, and when given in large doses serve to increase LBM and to reduce body fat content (FORBES, 1987a). This is why athletes take steroids.

4. Maintenance energy need is related to body size and composition

Studies of adult women and adolescent girls of widely varying body size show that energy requirement bears a direct, and linear, relationship to body weight, to lean body mass, and to body fat content (FORBES, 1989). The calculated regression slopes for adult women average 18 kcal per kg weight, 51 kcal per kg LBM, and 23 kcal per kg fat. Thus, a woman weighing 61 kg needs 18 kcal more each day to maintain her weight than one who weighs 60 kg. Values for adolescent girls are slightly higher, due presumably to greater physical activity and the requirement for growth. Multiple regression analysis shows that LBM exerts a greater influence on energy need than does body fat. Other investigators have found that basal metabolic rate bears a closer relationship to LBM than to body weight.

5. The energy cost of weight gain

Normal individuals gain weight, and LBM, when given excess food, and it turns out that the increment in both weight and LBM are proportional to the total excess energy consumed during the overfeeding period. The slope of the regression line, which includes both sexes, indicates that the average energy cost is 8.05 kcal per gram of weight gain (FORBES et al., 1986). This observed value is very close to the value predicted from the estimated costs of fat (12 kcal/g) and protein (8.66 kcal/g) deposition based on the analysis by SPADY et al. (1976). 1

1 Observed fraction of weight gain due to LBM is 0.384, energy cost of LBM deposition is 8.66 kcal/g protein x 0.205 g protein/g LBM = 1.78 kcal/g. Hence, 1.78 x 0.384 +12 x 0.616 = 8.08 kcal/g weight.

This large difference between the cost of depositing LBM and fat means that the energy cost of weight gain will depend on the composition of that gain. As noted earlier, thin individuals tend to gain a large proportion of lean, so they will gain weight more efficiently than those with larger fat burdens whose tendency is to gain more fat (FORBES, 1987b).

Based on the available data, FOMON et al. (1982) have made estimates of the changes in the lean and fat components of the body in infants and children from birth to age 10 years. I have used these data to make a graph of weight velocity as a function of age, and a graph showing the contribution of body fat to the total weight change (Figure 3). Growth velocity is very high during the first 3 months of life, then rapidly falls to remain low throughout most of childhood. However, of importance to the present argument is the composition of this weight gain. During the first 4 months of life some 40-45% of the gain in weight consists of body fat, whereas by age 2 years this has dropped to about 7%. Hence the energy cost of weight gain is going to be higher in early infancy than at age 2 years. Using the factors set forth by SPADY et al. (1976), but with a slightly lower LBM protein content (17%), the estimated energy cost of weight gain during those early months is 5.0 kcal/g. BUTTE and her coworkers (1989) have measured the energy cost of weight gain in 1- and 4-month-old infants, and their observed value (average 4.5 kcal/g) is not too far from the theoretical estimate.

Figure 3. Lower section: Rate of change in body weight during infancy and childhood. Upper section: Fraction of body weight change due to fat. Estimates based on data provided by FOMON et al. (1982).

Because of the smaller increment in body fat (Figure 3) the calculated energy cost of weight gain in the 2-year-old child is only 2.2 kcal/g. Admittedly, it would be difficult to test the validity of this estimate since the normal growth rate at this age is only 6-7 g per day. However, it stands to reason that the energy cost of weight gain in children recovering from malnutrition will depend on the relative amounts of lean and fat which make up the gain.


This work has been supported by NIH grants HD 18454 and RR 00044.


BENJAMIN, E.: Der Eiweissnährschaden des Säuglings. Zeitschrift für Kinderheilkunde, 10,185-302 (1914).

BUTTE, N.F., WONG, W.W., GARZA, C.: Energy cost of growth during infancy. Proc. Nutr. Soc., 48, 303-312 (1989).

CHAN, G.M., MILEUR, L., HANSON, J.W.: Effects of increased calcium and phosphorous formulas and human milk on bone mineralization in preterm infants. J. Pediatr. Gastroenterol. Nutr., 5, 444-449 (1986).

COCHRAN, W.G., KLISH, W.J., WONG, W.W., KLEIN, P.D.: Total body electrical conductivity used to determine body composition in infants. Pediatr. Res., 20, 561-564 (1986).

COKINGTON, L., MANNA, F.M., JACKSON, R.L.: Changes in body composition of mal-nourished infants during repletion. Ann. N.Y. Acad. Sci., 110, 849-860 (1963).

FJELD, C.R., SCHOELLER, D.A., BROWN, K.H.: Body composition of children recovering from severe protein-energy malnutrition at two rates of catch-up growth. Am. J. Clin. Nutr., 50, 1266-1275 (1989).

FJELD, C.R., FREUNDT-THURNE, J., SCHOELLER, D.A.: Total body water measured by 18O dilution and bioelectrical impedance in well and malnourished children. Pediatr. Res., 27, 98-102 (1990).

FLYNN, M.A., MURTHY, Y., CLARK, J., COMFORT, G., CHASE, G., BENTLEY, A.E.T.: Body composition of Negro and white children. Arch. Environ. Health, 20, 604-607 (1970).

FOMON, S.J., HASCHKE, F., ZIEGLER, E.E., NELSON, S.E.: Body composition of reference children from birth to age 10 years. Am. J. Clin. Nutr., 35, 1169-1175 (1982).

FORBES, G.B.: Human Body Composition: Growth, Aging, Nutrition and Activity. Springer-Verlag, New York, 1987a.

FORBES, G.B.: Lean body mass-body fat interrelationships in man: dietary changes induce changes in both body components. Nutr. Rev., 45, 225-231 (1987b).

FORBES, G.B.: Maintenance energy needs for women as a function of body size and composition. Am. J. Clin. Nutr., 50, 404-405 (1989).

FORBES, G.B.: The companionship of lean and fat: Some lessons from body composition assays. In: New Techniques for Nutrition Research, R.G. WHITEHEAD (Ed.). Bristol-Myers Symposium, Cambridge, U.K. (1990).

FORBES, G.B., BROWN, M.R., WELLE, S.L., LIPINSKI, B.A.: Deliberate overfeeding in women and men: energy cost and composition of the weight gain. Br. J. Nutr., 56, 1-9 (1986).

GREER, F.R., LANE, J., WEINER, S., MAZESS, R.B.: An accurate and reproducible absorptiometric technique for determining bone mineral content in newborn infants. Pediatr. Res., 17, 259-262 (1983).

HANNA, F.M.: Changes in body composition of normal infants in relation to diet. Ann. N.Y. Acad. Sci., 110, 840-848 (1963).

KAGAN, B.M., FELIX, N., MOLANDER, C.W., BUSSER, R.J., KALMAN, D.: Body water changes in relation to nutrition of premature infants. Ann. N.Y. Acad. Sci., 110, 830-839 (1963).

PETERSEN, S., GOTFREDSEN, A., KNUDSEN, F.U.: Lean body mass in small for gestational age and appropriate for gestational age infants. J. Pediatr., 113, 886-889 (1988).

PETERSEN, S., GOTFREDSEN, A., KNUDSEN, F.U.: Total body bone mineral in light-for-gestational-age infants and appropriate-for-gestational-age infants. Acta Paediatr. Scand., 78, 347-350 (1989).

REICHMAN, B., CHESSEX, P., PUTET, G., VERELLEN, G., SMITH, J.M., HEIM, T., SWYER, P.R.: Diet, fat accretion, and growth in premature infants. N. Engl. J. Med., 305, 1495-1500 (1981).

SPADY, D.W., PAYNE, P.R., PICOU, D., WATERLOW, J.C.: Energy balance during recovery from malnutrition. Am. J. Clin. Nutr., 29, 1073-1088 (1976).

STEARNS, G.: Mineral metabolism of normal infants. Physiol. Rev., 19, 415-430 (1939).

WATERLOW, J.C.: The rate of recovery of malnourished infants in relation to the protein and calorie levels of the diet. J. Trop. Pediatr., 7,16-22 (1961).

Discussion (summarized by A. Ferro-Luzzi)

The first point raised in the discussion concerned the importance of making a clear distinction between lean body mass and fat-free mass. Forbes argued that the small amount of phospholipids in lean body mass was not much of a source of confusion and did not have an important effect on the accuracy of measurement.

Methodological problems of measuring fat content in infants and children were discussed next. The growth process is a major obstacle to assessing true body composition before tissues reach chemical maturity. Forbes mentioned as an example that cartilage accounted for up to one third of the weight of the skeleton at birth and was still around 20% at two years of age. Most measures of body composition in small children are therefore fraught with a certain degree of uncertainty. Moreover, measures of fat content at earlier ages appear to have been obtained mostly by indirect methods (body water and/or body potassium) rather than by more direct densitometry which unfortunately is not applicable at an early age. The helium dilution method looked promising, but it seems that it has failed to give consistent results till now.

The cost of depositing and losing body tissues (fat and lean) were discussed. Forbes thought that losing tissue should cost less than depositing it, but that there was not enough information on the exact cost of either. Since weight almost never remains completely stable but has a tendency to oscillate around a mean, this could imply that a small intake increment is needed to maintain that mean weight.

Contents - Previous - Next