点击显示 收起
1 From the Institute of Human Nutrition, College of Physicians and Surgeons, Columbia University (AH, SH, M-YS, and DG), the Obesity Research Center (SH, IJ, M-YS, and DG), the Children’s Hospital of New York (MH), and the Department of Cardiology, St Luke’s–Roosevelt Hospital (NK), New York.
2 Supported in part by National Institutes of Health grant R29-AG-14715 and an educational grant from Knoll Pharmaceuticals. 3 Reprints not available. Address correspondence to D Gallagher, Obesity Research Center, 1090 Amsterdam Avenue, New York, NY 10025. E-mail: dg108{at}columbia.edu.
ABSTRACT
Background: Children have a high resting energy expenditure (REE) relative to their body weight. The decline in REE during growth may be due to changes in body composition or to changes in the metabolic rate of individual organs and tissues.
Objectives: The goals were to quantify body-composition components in children at the organ-tissue level in vivo and to determine whether the observed masses 1) account for the elevated REE in children and 2) account, when combined with specific organ-tissue metabolic constants, for children’s REE.
Design: This was a cross-sectional evaluation of 15 children (aged 9.3 ± 1.7 y) and 13 young adults (aged 26.0 ± 1.8 y) with body mass indexes (in kg/m2) < 30. Magnetic resonance imaging–derived in vivo measures of brain, liver, kidney, heart, skeletal muscle, and adipose tissue were acquired. REE was measured by indirect calorimetry (REEm). Previously published organ-tissue metabolic rate constants were used to calculate whole-body REE (REEc).
Results: The proportion of adipose-tissue-free mass as liver (3.7 ± 0.5% compared with 3.1 ± 0.5%; P < 0.01) and brain (6.2 ± 1.2% compared with 3.3 ± 0.9%; P < 0.001) was significantly greater in children than in young adults. The addition of brain and liver mass significantly improved the model but did not eliminate the role of age. REEc with published metabolic coefficients underestimated REEm (REEc = 3869 ± 615 kJ/d; REEm = 5119 ± 769 kJ/d; P < 0.001) in children.
Conclusion: The decline in REE during growth is likely due to both a decrease in the proportion of some of the more metabolically active organs and tissues and changes in the metabolic rate of individual organs and tissues.
Key Words: Organ mass • fat-free mass • resting energy expenditure • magnetic resonance imaging • growth • children
INTRODUCTION
Resting energy expenditure (REE) is a basic biological parameter with implications for energy requirements, energy balance, and energy stores. Equations based on body weight (1) have been superceded by models based on the energy requirements of 2 distinct body-composition compartments: fat or adipose tissue and fat-free mass (FFM) or adipose-tissue-free mass (ATFM), which have markedly different specific energy requirements. FFM is the principal contributor to energy requirements, and total-body FFM is commonly used as a surrogate for metabolically active tissue. However, this practice still pools numerous organs and tissues. The brain, liver, heart, and kidneys account for 60–70% of REE in adults, whereas the combined weight of these organs is < 6% of total body weight (2–5). Skeletal muscle constitutes 40–50% of total body weight and accounts for only 20–30% of REE (2, 3, 5).
Compared with adults, children have a higher REE per kilogram body weight or per kilogram FFM (2, 5–8) that declines steadily during the growth years. Whether this decline in REE is due to changes in body composition or to changes in the metabolic rate of individual organs and tissues remains unknown. Holliday (5, 7) hypothesized that the decrease in REE during growth and development is secondary to changes in body composition. In the first year of life, organs grow in proportion to body weight; thereafter, organ growth rates decelerate (7). By the age of 5 y, total brain volume has reached 95% of adult size (9), and by 6 y, heart diameter is 80% of adult values (10). Skeletal muscle mass increases at a faster rate than does body weight after the first year of life (7). A reduction in organ growth coupled with an increase in skeletal muscle growth could account for a decrease in whole-body REE adjusted for FFM. This has been the basis for the hypothesis that the decline in REE during growth is a result of a decrease in the proportion of the more metabolically active FFM components (7).
Currently, there are no noninvasive methods for measuring the metabolic rates of organs and tissues in vivo. Magnetic resonance imaging (MRI) has been used to determine the mass of various organs and tissues in vivo (3); along with previously reported organ and tissue metabolic rate constants (2), the contribution of each organ and tissue to whole-body REE can be estimated. Using this approach, Gallagher et al (3) made calculations based on individual or combined organ mass; in that study, calculated REE was highly correlated with the measured values in young adults.
Thus far, there has been no investigation of the relation between REE and body composition in healthy, growing children for whom organ and tissue measures are available, thereby allowing for an evaluation of the relative importance of the 2 hypotheses in accounting for changes in REE during growth and development. That is, if REE in children can be accurately estimated from measured organ and tissue mass by using adult metabolic rate coefficients (Table 1), then we can conclude that the metabolic rate per unit organ or tissue mass remains relatively constant from childhood to early adulthood and that changes in body composition are responsible for the decline in REE. On the other hand, if REE in children is still underestimated after taking organ mass into account, then changes both in body composition and in metabolic rate coefficients are likely to play a role in the changing REE during the growth years.
View this table:
TABLE 1 . Organ and tissue coefficients used in developing models
The aims of the present study were to quantify the composition of FFM in children at the organ-tissue level and to determine whether the specific organ and tissue metabolic constants reported in the literature (2) are adequate to account for REE in childhood. We hypothesized that 1) the proportion of FFM as liver, brain, heart, and kidneys is greater in children than in young adults, and 2) the specific organ and tissue metabolic constants reported in the literature (2) applied to the observed organ and tissue masses will account for the observed REE in children.
SUBJECTS AND METHODS
Subjects
Children were recruited over a 3-y period through advertisements in local newspapers and on radio stations and through flyers posted in the local community. The adult subjects were previously described (3). Age (6–12 y for children; 23–29 y for young adults) and body mass index [< 30 (in kg/m2) for both groups] limits were set as requirements. Other inclusion criteria were that participants be ambulatory, not engage in vigorous exercise, and have no medical condition that could affect the variables under investigation. A medical evaluation that included a physical examination and screening blood tests was conducted on each potential subject. Only healthy individuals without any diagnosed medical condition and with normal thyroid hormone values were enrolled. The study was approved by the Institutional Review Board of St Luke’s–Roosevelt Hospital.
Body-composition measures
Body weight was measured to the nearest 0.1 kg (Weight Tronix, New York) and height to the nearest 0.5 cm by using a stadiometer (Holtain, Crosswell, United Kingdom). Total adipose tissue and skeletal muscle mass were measured by using whole-body multislice MRI. Subjects were positioned on the 1.5-T scanner (6X Horizon; General Electric, Milwaukee) platform with their arms extended above their heads. The adult adipose tissue and skeletal muscle protocol involved the acquisition of 40 axial images, 10-mm thick, at 40-mm intervals across the whole body (12). The pediatric protocol involved the acquisition of 35 axial images, 10-mm thick, at 35-mm intervals across the whole body. Note that MRI measures of adipose tissue are based on the analysis of gross cross-sectional images and include small amounts of nonfat materials (fluids, blood, etc). Also, small amounts of lipids distributed throughout the body (intramuscular, liver) are not identified in these cross-sectional images. Total adipose tissue and ATFM (ATFM = body weight - total adipose tissue by MRI) compartments therefore do not correspond exactly with measures of fat and FFM (FFM = body weight - fat by dual-energy X-ray absorptiometry) obtained by DXA.
Liver and kidney images were produced by using an axial spin-echo T1-weighted sequence with 5-mm slice thickness, no interslice gap, and a 40 x 40 cm2 (256 x 192/2 number of excitations) field of view. About 40 slices were acquired from the diaphragm to the base of the kidneys. Brain images (29) were produced by using a body coil with a fast-spin-echo T2-weighted sequence with 5-mm contiguous axial images and a 40 x 40 cm2 (256 x 256/1 number of excitations) field of view.
SLICEOMATIC 4.2 image analysis software (Tomovision, Montreal, CA) was used to analyze the images on a PC workstation (Gateway, Madison, WI). MRI volume estimates were converted to mass by using the assumed density for each tissue and organ (Table 1). In our laboratory, the technical error for repeated measurements of the same scan by the same operator for MRI-derived total adipose tissue and skeletal muscle volumes in adults is 1.1 ± 1.2% and 0.7 ± 0.1%, respectively (13). A sample of 8 MRI scans of the liver in healthy 26–70-y-old men and women was analyzed by 2 different operators to estimate reading error. The SD of the mass differences was found to be 0.14 kg with a mean weight of 1.58 kg.
Left ventricular mass was evaluated by using a two-dimensionally guided M-mode echocardiogram (Hewlett Packard 1500, Boise, Idaho) interfaced with strip chart recorder, two-dimensional video recorder, and either a 2.5- or 3.5-MHz probe. All subjects were studied while lying partially on the left side. Left ventricular dimensions were recorded from the parasternal long axis view at or below the tips of the mitral valve leaflets. The hard-copy strip chart recording was used for all measurements. End-diastolic and end-systolic dimensions were measured at the peak of the R wave and at the peak of the posterior wall motion, respectively, according to the American Society of Echocardiography convention (14). Wall thickness was measured by using the Penn convention (15), and left ventricular mass was calculated according to the formula of Devereux and Reichek (15). For all measurements, a minimum of 5 cardiac cycles was used. All echocardiographic recordings were read by a single cardiologist (NK), and the technical error for repeated measurements of the same scan by the same operator for left ventricular mass was 1.1%. Left ventricular mass was multiplied by a factor of 1.50 to obtain an approximate value for total heart mass (16).
A residual category consisted of body weight minus all other measured compartments.
Energy expenditure
Subjects reported to the study center in the morning after fasting overnight, and REE was measured by using the Columbia Respiratory Chamber-Indirect Calorimeter (17). After entering the thermoneutral chamber, the subjects rested comfortably on a bed with a plastic transparent ventilated hood placed over their heads for 40–60 min. Magnetopneumatic oxygen (Magnos 4G) and carbon dioxide (Magnos 3G) analyzers (Hartmann & Braun, Frankfurt, Germany) were used to analyze the rates of oxygen consumption and carbon dioxide production; the data displayed were then stored by the online computer system. Gas exchange results were evaluated during the stable measurement phase and were converted to REE (kJ/d) by using the formula of Weir (18). For a standard alcohol phantom, gas concentration measurements are reproducible to within 0.8%.
The REE (kJ/d) of each organ-tissue component (subscript i) was calculated by using the following equation (3):
RESULTS
Baseline characteristics
The baseline characteristics of the subjects are shown in Table 2. The race-ethnicity of the children included Hispanic (n = 2), African American (n = 3), and white (n = 10); that of the adults was Asian (n = 2), African American (n = 2), and white (n = 9). The young adults were significantly heavier, were taller, and had greater body mass indexes than did the children (P < 0.001 for all between-group differences). With respect to body composition, the young adults had significantly greater liver, heart, kidney, skeletal muscle, total adipose tissue, and residual mass than did the children (P < 0.001 for all between-group differences). Unlike the weight of all other organs, brain weight was not significantly different between the 2 study groups. This finding is consistent with reports that total brain volume reaches 95% of its adult size by the age of 5 y (9).
View this table:
TABLE 2 . Subject characteristics at baseline1
The proportional contribution of each organ or tissue to total body weight and ATFM are given in Figure 1A and B, respectively. In terms of individual organs, only the proportional contributions of the liver (P < 0.01) and brain (P < 0.001) to ATFM were significantly greater in children than in adults.
FIGURE 1. . A: Proportional contribution of each organ or tissue to total body weight. *,**Significantly different from adults: *P < 0.01, **P < 0.001. B: Proportional contribution of each organ or tissue to adipose-tissue-free mass (ATFM). *,**Significantly different from adults: *P < 0.01, **P < 0.001. Values given in boxes are means ± SDs.
REE modeling
Multiple regression analyses were carried out to examine the associations between REEm (in kJ/d) and several sets of independent variables (Table 3). REEm was the dependent variable in all models, and the 2 age groups were coded as children = 0 and adults = 1 in all analyses. Model 1, a basic model to predict REE from total adipose tissue and ATFM, showed that nonadipose tissue is a significant predictor of REEm (P < 0.001, r2 = 0.72, SEE = 773). The addition of age group in model 2 confirmed the significant role of age in predicting REEm (P < 0.001, r2 = 0.84, SEE = 608). In model 3, a new variable representing the sum of brain and liver mass, the only organs that were disproportionately larger in children, was added to model 2 to see whether information about the disproportionate size of these organs could account for the elevated REEm and render the age group variable redundant. The brain and liver mass variable made a statistically significant contribution to the model (P < 0.03), which then accounted for a total of 88% of the variance. However, age group continued to make a highly significant contribution (P < 0.001). Note, however, that the intercept ceased to be statistically important in this final model, suggesting that the model may have identified the important compartments that contribute to oxygen consumption so that an arbitrary constant in the formula is no longer necessary and that the regression line passes through zero.
View this table:
TABLE 3 . Regression models for the association of body-composition variables with resting energy expenditure (REE, kJ/d)1
Caution should be used in interpreting these multiple regression models because of the small number of subjects available. In particular, the coefficients should not be interpreted as reflecting actual energy costs because of the high collinearity among the variables.
As an alternative method of assessing the adequacy of disproportionate organ size in accounting for the differences in REE between children and adults, we compared calculated and measured REE for subjects in each group (3) with the use of the coefficients from Elia (2). As shown in Table 4, the values for REEc and REEm were significantly different in children (P < 0.001) but not in adults (P = 0.753).
View this table:
TABLE 4 . Resting energy expenditure (REE) in children and adults1
DISCUSSION
Using MRI and echocardiography, methods that have been previously used in adult body-composition studies (3, 13, 19), we quantified the mass of various organs and tissues in our subjects. The results show in vivo what has been previously shown in cadaver studies and hypothesized to be a factor in accounting for the elevated REE of children: that a difference exists in the proportion of some tissues and organs that make up the FFM in children and adults, specifically, the brain and liver (2, 5–8, 20–24). Our results, therefore, are consistent with the hypothesis that a decrease in the proportion of the more metabolically active organ mass may account for a decline in REE per kilogram body weight or per kilogram FFM during growth, as suggested by others (2, 5–8, 24).
Multiple regression analyses were conducted to explore the relations between REEm and various independent variables, although the use of this statistical tool was constrained by the small sample size. The addition of information about disproportionate brain and liver mass to the regression model that already included age group (model 3) contributed significantly to the model; however, age group continued to play a significant role in the model. The implication, therefore, is that other age-related factors, possibly hormonal (25), are additional significant determinants of REE.
Furthermore, estimating REE in children by using an approach previously validated in adults (3, 13) predicts only 76% (ie, REEc/REEm) of REEm. As shown in Figure 2, REEc consistently underestimated REEm in the pediatric group. Our results, therefore, do not support the hypothesis that differences in body composition are adequate to account for differences in REE between children and adults. Because REE in children cannot be accurately estimated from organ and tissue mass with the use of adult coefficients, we conclude that during growth and development, the metabolic rate per unit mass of individual organs and tissues may be higher.
FIGURE 2. . Difference between measured and calculated resting energy expenditure (REEm - REEc) versus age in children () and adults (). In children, REEm > REEc; in adults, there was no systematic difference.
Findings from brain studies (26, 27) suggest that the metabolic rate of the brain is indeed significantly higher in children than in young adults. In a small sample of children (n = 9), Kenney and Sokoloff (27) found that although cerebral oxygen consumption was relatively constant between the ages of 3 and 11 y, it was significantly higher than in young adults. These findings were later supported by Chugani et al (26), who, by using 2-deoxy-2[18F]fluoro-D-glucose and positron emission tomography, found that local cerebral metabolic rates for glucose in 29 children (aged 5 d to 15.1 y) were low at birth, increased to twice the adult rates by 3–4 y, and were maintained at high levels for the next 5–6 y. Local cerebral metabolic rates for glucose were then observed to decline until adult rates were reached by the end of the second decade. These data suggest that the metabolic rate of the brain is not a constant during growth and that its rate of change is nonlinear.
Findings from the above studies suggest that the decline in REE per kilogram body weight (or per kilogram FFM) could be due to changes in the metabolic rates of organs and tissues. An additional partial explanation for the underestimation in children is that REE in children includes the energy cost of laying down tissue. Assuming that children were to gain 31 kg weight (to match the weight of the adults in this study) between 9 and 17 y, this would represent 3.9 kg/y. Using the data of Webster (28), we can estimate that the amount of metabolizable energy needed to deposit 1 g of either protein or fat would be 53 kJ. Therefore, a 3.9-kg body weight gain would correspond to 206 700 kJ (3900 g x 53 kJ)/y or 556 kJ/d. The latter value could explain 50% of the underestimation of REE found in children.
Our study had some limitations. Along with the metabolic coefficients of Table 1, numerous constants were used, including assumed organ and tissue densities developed from reference man data (11); it is unclear whether these coefficients can be accurately applied in women and children. Moreover, organs and tissues were assumed to be homogeneous in composition. The MRI measurement protocol assumes that there are negligible amounts of, for example, infiltrated organ or tissue fat, edema, and cystic structures that would invalidate the assumed organ and tissue properties summarized in Table 1. Although the MRI measurement methods used in the present study can quantify smaller organs and tissues (eg, spleen, pancreas, thyroid gland, and skin), data are limited on the respective densities and oxygen consumptions of these organs and tissues. As a result, these components were grouped together as residual mass. Also, because of the small sample size, we were unable to further investigate the effects of sex, pubertal stage, and race on the REE–body composition relation in children (21, 24). Last, we cannot rule out the possibility that REEm in some children may be somewhat above true resting values because children tend to adhere less well to remaining in a nonfidgeting resting state during the REE measurement period. Some investigators have suggested that high-intensity exercise in adults may influence REE for up to 48 h after exercise (29). Because children spend more time in high-intensity play or training than do adults, it cannot be ruled out that the differences in REE between the adults and the children were partly influenced by differences in physical activity levels or a possible carryover effect of recent exercise.
In summary, these data confirm the hypothesis that the proportion of FFM as certain high-metabolic-rate organs, specifically, liver and brain, is greater in children than in young adults. However, after this disproportion was accounted for, the specific organ and tissue metabolic constants available in the literature (2) were not adequate to account for the REE in children. These results therefore imply that the decline in REE per kilogram body weight (or per kilogram FFM) during the growth years is likely due to both changes in body composition and changes in the metabolic rate of individual organs and tissues.
Further studies and techniques should aim at using noninvasive methods to quantify the metabolic rates of organs and tissues in children in vivo. Such information will help us to better understand the contribution of various organs and tissues to REE in healthy children. Moreover, better estimation of the energy expenditure of specific organs in this population could provide insights into individual variations in REE.
ACKNOWLEDGMENTS
AH was responsible for data analysis and manuscript writing; SH was responsible for data analysis and manuscript writing and provided advice and consultation; IJ was responsible for data analysis, specifically of the magnetic resonance imaging scans for organ and tissue volumes; M-YS was responsible for manuscript writing; MH was responsible for recruitment of children and data collection; NK was responsible for data collection, specifically all echocardiographic recordings; DG was responsible for study design, data collection, data analysis, and manuscript writing and provided administrative support, supervision, and advice. None of the authors had a conflict of financial or personal interest in any company or organization sponsoring this study.
REFERENCES