Effect of age on body composition
and resting metabolic
rate
NAOMI K. FUKAGAWA, LINDA G. BANDINI, AND JAMES B. YOUNG Charles A. Dana Research Institute and the Harvard Thorndike Laboratory, Department of Medicine, Beth Israel Hospital, and Department of Pediatrics and Division on Aging, Harvard Medical School, Boston, Massachusetts 02115; and The Clinical Research Center, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139
FUKAGAWA, NAOMI K., LINDA G. BANDINI, AND JAMES B. YOUNG. Effect of age on body composition and resting metabolic rate. Am. J. Physiol. 259 (Endocrinol. Metab. 22): E233-E238, 1990.-The relationship between fat-free mass (FFM) and resting metabolic rate (RMR) was compared in young men (n = 24; age 18-33 yr), old men (n = 24; 69-89 yr), and old women (72 = 20; 67-75 yr). Body composition was assessed using anthropometry, bioelectrical impedance analysis (BIA), and isotope dilution with “O-labeled water. RMR was measured at least twice using an open-circuit indirect calorimetry system with a ventilated hood. The results indicate that the different methods for assessing body composition vary substantially and should not be used interchangeably. Anthropometry was not adequate to assess group differences in body fatness, although skinfold measures may be appropriate for within-group comparisons. BIA correlated well with the isotope-dilution technique and may be a useful measure of FFM. Finally, RMR was lower in the old men than the young (1.04 t 0.02 vs. 1.24 t 0.03 kcal/min, P < 0.001) and remained lower even when adjusted for FFM estimated by isotope dilution (P < 0.001). RMR in the women was also lower (0.84 t 0.02 kcal/min), but in contrast to the difference between young and old men, RMR adjusted for FFM did not differ (P = 0.16) between old men and women. Therefore, it is clear that differences in FFM cannot fully account for the lower RMR in the old, suggesting that aging is associated with an alteration in tissue energy metabolism.
anthropometry, bioelectrical impedance analysis (BIA), and isotope dilution with 180-labeled water ( [lSO]HzO). Although there are numerous studies reporting the changes in subcutaneous skinfold thickness and body fat distribution in the elderly, there is very little information on which method best estimates body fatness in an old population. We, therefore, assessedhow well measures of body fatness derived from anthropometry and BIA correlated with percent body fat estimated by isotope dilution, which was chosen as our “gold standard.” One of the limitations of the isotope-dilution method, as with all indirect measures of body composition, is that it is based on the assumption that the components of FFM are of relatively constant composition. Our results indicate that the different methods for assessingbody composition in humans vary substantially and should not be used interchangeably; anthropometry measuresbody fatness, whereas BIA and [lSO]HzO measure hydrated lean tissue. Furthermore, it is clear from our data that differences in FFM cannot fully account for the lower resting metabolic rates in the old, suggesting that aging is associated with an alteration in tissue energy metabolism. MATERIALS
AND METHODS
Subjects. Twenty-four young (18-33 yr) and 24 elderly (69-89 yr) men and 20 elderly (67-75 yr) women in good health, as determined by history and physical examinations, were studied at the Clinical Research Centers AGING IS ASSOCIATED with a decrease in metabolic rate (CRC) of the Beth Israel Hospital (BIH) and Massachuthat has been attributed to an age-related decline in setts Institute of Technology (MIT). None were taking “active” cellular mass, i.e., fat-free mass (FFM) and a any medications at the time of investigation, and all had concomitant increase in body fat (l&18). Although many maintained stable body weights in the previous 6 mo. All studies in both children and adults have shown that subjects were carefully informed of the nature, purpose, metabolic rate is highly correlated with FFM (1, 3, 23, and possible risks of the studies before giving their 33), it is not clear whether the decline in metabolic rate written consent to participate. The protocol used was with aging is due entirely to a decrease in FFM or approved by the Clinical Investigation Committee at the whether it also reflects a reduction in the metabolic BIH and by the Committee on the Use of Humans as activity of lean tissue. A number of studies have exam- Experimental Subjects at MIT. ined the relationship between FFM and resting metabolic Protocol. All subjects were admitted to the CRC the rate (RMR) in obese and lean individuals and during evening before the study. After a 12-h overnight fast, hypocaloric feeding (10, 23); however, to our knowledge, they were awakened, asked to void, weighed in a hospital no studies have examined the relationship of FFM and gown, and returned to bed to rest until their metabolic RMR with advancing age. rate was measured for 45 min as described below. After In this study, the relationship between FFM and RMR completion of the measurement of metabolic rate, body was compared in young men, old men, and old women. composition was determined by isotope dilution using Because of concerns that the methods employed to esti- [ “O]H20 (n = 68), BIA (n = 66), and anthropometry (n mate FFM might affect the conclusions, we assessed = 60). Repeat measurements of metabolic rate were made at least one to two times during their stay in the CRC, body composition using several techniques, including fat-free mass; energy expenditure;
isotope dilution
0193-1849/90
$1.50 Copyright
0 1990 the American
Physiological
Society
E233
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
E234 TABLE
AGE
EFFECTS
ON
BODY
COMPOSITION
AND
METABOLIC
RATE
1. Subject characteristics P
Group Young
men
(n = 24)
Age, Y*
21tl
Height, cm Weight, kg Body mass index, kg/m2 Fat-free mass*, kg Total body fat*, kg %Body fat* Values
TABLE
are means
t
176.721.7 72.3k2.1 23.1t0.5 55.4t1.9 16.9tl.O 23tl
SE. * Estimated
2. Anthropometric
by isotope
Old men
(n = 24) 75tl 172.1t1.2 72.8t1.6 24.6t0.5 47.7t0.9 25.1tl.0 34t1
Old women
(n=
20)
72tl 158.8k1.2 59.4t2.2 23.5t0.8 35.8t0.8 23.6t1.5 3921
-
Young vs. old men
Old men vs. women
0.001 0.05 0.41 0.05 0.001 0.001 0.001
0.03 0.001 0.001 0.25 0.001 0.41 0.003
dilution.
measures in young men, old men, and old women P
Group Anthropometric Parameters
Young
men
(n = 20) Skinfold Tricep Biceps Subscapular Suprailiac Sum of 4 skinfolds Waist-to-hips ratio Values
are means
10.3kO.9 3.7rto.4 12.1t0.8 14.2t1.5 40.4t3.0 0.85t0.01
Old men
(n = 20) lO.lt0.6 4.2kO.4 16.6k1.6 14.7t1.1 45.622.7 0.95&0.01
Old women
(n = 20) 2O.lk1.4 12.4k1.4 20.8t1.6 18.9t1.9 72.2k5.5 0.86t0.02
Young vs. old men
Old men vs. old women
0.88 0.44 0.02 0.81 0.21 0.001
0.001 0.001 0.07 0.06 0.001 0.001
t SE in mm.
each after 15-30 min rest periods. Subjects were given practice sessions the evening before the studies to familiarize them with the ventilated-hood system of indirect calorimetry. The BIA method to assess body composition was used as previously described (28). Measurements were made with a portable impedance analyzer (BIA 103, RJL Systems, Detroit, MI) while the subjects lay comfortably in bed with their limbs abducted. Resistance (R) to the flow of a 50-kHz, 800-PA current was then measured on a Oto 1,000-Q scale. Anthropometric measurements were performed independently by two investigators. Skinfold thickness of the nondominant side of the body were measured in triplicate with a Lange skinfold caliper (Cambridge Instruments, Cambridge, MA) exerting a pressure of 10 g/mm2, and the mean of all measurements was used for the analyses. Regional fat distribution was estimated from the ratio of waist-to-hip circumferences. The waist was measured at the umbilicus in the men, but, because of differences in the position of the umbilicus in the women, the waist measurement used for the women was the smallest abdominal circumference. The hip measurement was made at the widest/broadest point over the buttocks in all groups. Skinfold measures were made at the triceps, biceps, subscapular, and suprailiac sites. Measurement of energy expenditure. An open-circuit indirect calorimetry system with a ventilated hood and interfaced to a desktop computer was used for all indirect calorimetric measurements at both BIH and MIT. The system used at the BIH was as follows. Air was withdrawn from the hood at 50 l/min, as measured by a rotameter (Fisher and Porter, Warminster, PA); concentrations of O2 and CO2 were measured using an electrochemical O2 sensor and infrared CO2 sensor (Ametek, Pittsburgh, PA). The output from the analyzers was
converted to digital signals by the computer for calculating the rates of 0, consumption, CO2 production, and metabolic rate, using Weir’s equation (34). Values were averaged and saved at I-min intervals after correction to standard temperature and pressure. The analyzers were calibrated before each measurement using standard gases, and the system was calibrated weekly using fixed concentrations of gas and found to be reproducible to within 3% of predicted values. The system used at MIT has been described elsewhere (2). The majority of the subjects (n = 47) were studied at BIH, and both young (n = 11) and old (n = 10) were studied at MIT, so that possible systematic errors introduced by differences between the indirect calorimetry systems were avoided. In addition, 10 men had their resting metabolic rates measured at both sites on separate occasions to determine the comparability of the two indirect calorimetry systems. No significant differences were found in measurements obtained at both MIT and BIH (P > 0.10, paired t test). The average difference between measurements at both sites was 0.02 t 0.01 kcal/min or 2.4%. Analytical methods. Body composition was determined by isotope dilution techniques using [180]H20 (27). Serum samples were obtained both before and 4 h after ingestion of [ “O]H20 [0.07 g [ “O]/kg estimated total body water (TBW)], and analyzed commercially for “0 enrichment using isotope ratio mass spectrometry (Geochron Laboratories, Cambridge, MA). The measurement of 180 enrichment varied Cl%. Oxygen dilution space (Do) was calculated from the increase in 180 enrichment of the 4-h serum relative to predose values as previously described (27). TBW was calculated from Do, assuming D = TBW x 1.01. FFM was calculated from TBW, asOsuming the relationship FFM = TBW/0.73 (21). Statistics. All data are presented as means t SE. The relationships between FFM and RMR in the three groups
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
AGE
EFFECTS
ON
BODY
COMPOSITION
were examined by analysis of covariance. Frequently, comparison of RMR data among subjects or between groups was done by expressing RMR per kilogram of FFM, i.e., by dividing RMR by FFM (17). However, the use of this ratio (RMR/FFM) is only appropriate when the correlation between RMR and FFM is perfect, i.e., r = 1.0, and when the mathematical equation relating the two parameters is one in which RMR is in constant proportion to FFM with an intercept equal to zero (17, 31). Analysis of covariance allows one to derive adjusted RMR values according to a specific relationship between RMR and FFM, thereby avoiding the assumption of a perfect correlation and zero intercept. The data were examined for parallelism, and results did not indicate that the slopes between the regression lines for each group were different. Age-gender was used as the class variable and FFM as the covariate. Pairwise comparisons of interest were determined a priori to be old men vs. young men and old men vs. old women. Comparisons between groups were made using the Student’s t test and Pearson product-moment correlations (BMDP Statistical Package, Los Angeles, CA). Intraclass correlations were calculated according to Zar (35). RESULTS
The characteristics of the volunteers are summarized in Table 1. The young men were taller than the elderly men (P < 0.05) but of comparable body weight (P = 0.41). The women were shorter and weighed less than the old men (P c 0.001). Body mass index, the ratio of body weight to height squared, was slightly, but significantly, higher in the old men compared with the young men (P < 0.05) and similar in both old men and women (P = 0.25). FFM estimated by isotope dilution was significantly lower in the old men (47.7 t 0.09 kg) than in the young men (55.4 t 1.9 kg; P < 0.001). FFM in the women (35.8 t 0.8 kg) was significantly lower than in the old men (P < 0.001). Total body fat, calculated as the difference between body weight and FFM, was greater in the old men compared with the young men (P < 0.001) but was similar in both groups of old (P = 0.41). The percentage of body weight as fat was higher in the old men than in the young men (P < O.OOl), reflecting the greater fat mass in the old men. Percent body fat was higher in the women than in the old men (P < O.OO3),reflecting the lower body weight of the women and hence greater relative fatness. Despite the differences in fat mass and in body fatness between young and old men, triceps (TSF), biceps (BSF), and suprailiac (SISF) skinfold thicknesses were similar in young and old men (Table 2); only the subscapular skinfold (SSF) was higher in the old men compared with the young men (P < 0.02). On the other hand, whereas elderly men and women differed in percent body fat but not in fat mass, TSF and BSF were substantially greater in the women than in the men (P < O.OOl), whereas SISF and SSF were only slightly greater (P < 0.06 and ~0.07, respectively). As suggested by the elevation in the waistto-hips ratio (WHR) in old men in comparison with the other two groups (P < O.OOl), differences in body fat distribution may contribute to the discordance between
AND
METABOLIC
E235
RATE
measurements of skinfold thickness and body fat among the three groups. Skinfold thicknesses within each study group, either individually or as the sum of the four measurements, generally correlated with measurements of body fat. TSF and BSF correlated highly with percent body fat in old men and women, but less well in young men (Table 3). SSF, however, correlated well with percent fat in young men and old women but not as well in old men. SISF, in contrast, did not correlate with percent body fat i.n either young or old men but did correlate highly with percent fat in the women. The sum of all four skinfolds correlated highly with percent body fat in all groups. Similarly, WHR was also strongly correlated with body fatness in all three groups, whereas BMI was significantly related to percent fat only in the two old groups and not in young men. These findings suggest that skinfold measures may serve as a valid index of fatness within specific age-gender groups, but may be considerably less useful as a means of comparing fatness among different agegender groups. Height squared divided by resistance obtained from BIA measurements correlated very strongly in all groups with the estimate of FFM based on isotope dilution (young men: r = 0.90, P c 0.001; old men: r = 0.85, P < 0.001; and old women: r = 0.87, P < 0.001, Fig. 1). Metabolic rate results are summarized in Table 4 and Fig. 2. FFM and RMR were significantly correlated within each group (young men: r = 0.82; old men: r = 3. Correlation coefficients for relationship between anthropometric measures and percent body fat TABLE
Young
Anthropometric Parameters
Skinfold Triceps Biceps Subscapular Suprailiac Sum of 4 skinfolds Waist-to-hips ratio Body mass index Values
Men
r
estimated
Old Men
P
r
Old Women P
r
P
0.45
0.06
0.55
0.39
0.10
0.51
0.50 0.29 0.47
0.04 0.23 0.04
0.39 0.37 0.58
0.11
0.007
0.68 0.75 0.80
0.59 0.16
0.008 0.47
0.71
0.001
0.47
0.63
0.001
0.74
by isotope
dilution.
0.01 0.02
0.09
0.70
0.001
0.69
0.001 0.001 0.001 0.001 0.038 0.001
0 0
80
1 0
0
0
0 0
l l
a0
0 0
0.
Young Men Old Men Old Women
+ 6b FFM FIG.
1. Relationship
fat-free mass (FFM) men, and old women.
between estimated
80
(kg)
height squared-to-resistance by isotope dilution in young
ratio and men, old
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
E236
AGE
EFFECTS
ON
BODY
COMPOSITION
4. Metabolic rate in young- men, old men, and old women TABLE
Resting Metabolic Rate
Young
Men
Old Men
Old Women
Mean Measured Adjusted for FFM WHO/FAO/UNU
1.24kO.03 1.04kO.O2* 0.84t0.02-f 1.13kO.02 1.03&0.02* 0.99+0.02$ estimates
1.23t0.02
1.02&0.01
0.85kO.02
Values are means & SE in kcal/min. FFM, fat-free mass. * P < 0.002 for comparison with young men; t P < 0.001 for comparison with old men; $ P = 0.16 for comparison with old men; § Ref. 11.
1 0.50
I 30
0
Young
l
Old
Men
Old
Women
A .
, 40
.
, 50
.
, 60
.
Men
, 70
1
{ 80
FFM (kg) FIG. 2. Relationship between resting metabolic mass (FFM) estimated by isotope dilution in young old women. Overall, r = 0.89, P < 0.001.
rate and fat-free men, old men, and
0.48; and old women: r = 0.74; as well as overall: r = 0.89, P < 0.001). The intraclass correlation coefficient of the multiple determinations of RMR within each individual was 0.968, suggesting that the measures were highly reproducible. Absolute RMR, representing the mean of two to eight measurements per individual, was lower in the old men than in the young (1.04 t 0.02 vs. 1.24 t 0.03 k cal/ min; P < 0.001); RMR remained lower in old men (P c 0.001) even when adjusted for FFM. RMR was also lower in the old women (0.84 t 0.02 kcal/ min) compared with the old men (P < 0.001) but, in contrast to the difference between young and old men, group means did not differ after adjustment for FFM (P = 0.16). The absolute values for RMR in all groups were similar to those estimated from the 1985 World Health Organization/Food and Agriculture Organization of the United Nations/United Nations University guidelines (11) (Table 4). DISCUSSION
As the elderly population increases in number, a growing health concern focuses on geriatric nutritional status and methods of nutritional assessment. In this study, measurements of skinfold thickness and bioelectrical impedance were compared with estimates of body composition obtained by isotope dilution in young and old individuals. In addition, the effect of age on the relationship between metabolic rate and FFM was examined. Although several references for anthropometric norms for the elderly have been established (5, 6, 15), the relationship between them and total body fat or percent
AND
METABOLIC
RATE
body weight as fat is not clearly established. Because the correlation between skinfold thickness and body fatness decreasesas body fat increases (l3), it might be expected that skinfold thickness would be a less robust index of body fatness in the old population because of the welldocumented increase in body fat with advancing age. Our data indicate that this is not the case, because skinfold thickness measurements in general correlated more highly with percent body fat in the old than in the young. Although others have found stronger relationships between TSF or SSF and total body fat or body fatness in children and young adults than reported here (24), these differences may be due, in part, to the small sample size and narrow range of body fatness observed within our young men (23 t 1%). Despite statistically significant correlations between several anthropometric measures and body fatness, a substantial portion of the variance in relative fatness (X50%) in all groups remained unexplained by anthropometry. On the other hand, the utility of skinfold thickness measurements in intergroup comparisons of body fat or body fatness is open to question. Despite substantial differences between young and old men in fat mass or percent body fat, neither TSF, BSF, nor SISF differed in the two groups. Similarly, although body fat was not different in old men and women and percent body fat was only slightly, albeit significantly, greater in the women, TSF and BSF were markedly increased in the women. The truncal skinfolds (SSF and SISF) were, however, slightly higher in the old women. Differences in the pattern of fat distribution in the groups studied may influence the interpretation of our findings, since the WHR differed significantly between the two groups of men and between the old men and women. These findings demonstrate that measures of anthropometry are not adequate to assess group differences in body fatness, although they may be appropriate for withingroup comparisons. Our results have potential implications for the estimation of FFM from anthropometric data. Such determinations rely on regression equations relating various skinfolds to some more direct measure of body fatness, e.g., body densitometry measurements (9). FFM is then calculated as the difference between body weight and the product of body weight and percent body fat. Any potential errors in skinfold thickness measurements or in their relation to body fatness are then magnified in the final calculation. Because the equations generally used to estimate body fat from skinfold measurements differ across age groups (9), a point that is consistent with the data presented here, age and gender-specific regression equations may be required to achieve valid information by this approach. Our results demonstrate that skinfold thickness measures differ between young and old men and between old men and women and hence should not be used as an index of body fat to assessthe relationship of FFM and RMR in young and old individuals without appropriate standardization. Recently, BIA has been adapted for use in the assessment of FFM in human subjects. BIA is rapid and noninvasive, and it has the advantage of being easy to use, even at the bedside. The results from BIA correlate
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
AGE
EFFECTS
ON
BODY
COMPOSITION
well with those obtained by densitometry and isotopedilution techniques (19,25,28). In the current study, the impact of age and gender on the relationship between BIA and FFM as estimated by isotope dilution of [l’O]Hz0 was examined. The ratio of height squared over resistance (H2/R), the immediate output of BIA, correlated significantly with FFM in all groups. Although FFM can be derived from H2/R, the derivation requires the use of a specific regression equation relating this ratio to an independent measure of FFM. The high correlations between H2/R and FFM observed in this investigation imply that both measures provide comparable estimates of TBW, from which FFM is derived and, furthermore , that in a study t such as this in which FFM is employed as a cova riate in the analysis of another variable (e.g., RMR), the use of the ratio H2/R might be equally suitable as a covariate in the analysis. In the present study, FFM was derived from measures of TBW. Because neutral fat does not bind water, measurement of TBW is a means of estimating the nonfat compartment of the body. The use of TBW in the estimation of body composition assumes a two-compartment model (fat and lean) in which FFM includes the extracellular fluid mass (EC W) ) intracellular components (ICW), and skeletal structures. This analysis also assumes that the ratio TBW/FFM is independent of age or gender. Because this latter assumption may not be entirely correct over the age ran .ges studied, the question arises whether our finding of diminished RMR in relation to FFM in young and old subjects might be attributable to inaccuracies inherent in this two-compartment model for body composition. Specifically, because the distribution of body water between extracellular and intracellular compartments is known to differ with age (ZZ), could the results depicted in Fig. 2 reflect a systematic overestimation of FFM in the two old groups and not a reduction in energy expenditure by lean tissue as we suggest? Because only TBW was measured in our study, no definitive answer to this question can be obtained from our data. It is, however, possible to estimate the changes in body water distribution between intracellular and extracellular spaces that would be required if group dif1.75 -I-
1
0 .
I 20
’ Total
I 30
= Body
A I 40
Young OldMen
Men
Old Women ’ I . 50
60
Water
(kg) 3. Relationship between resting metabolic rate (RMR) and body water (TBW) estimated by isotope dilution in young men, men, and old women. Dotted lines and respective midpoints repreaverage TBW of old groups if their RMR fell on same regression as for young groups.
FIG.
total old sent line
AND
METABOLIC
RATE
E237
ferences in RMR were due solely to age-related alterations in body water distribution. For the purposes of this discussion, ICW will be assumed to represent more closely than TBW the metabolically active compartment in which energy is expended. From the equations of Pierson et al. (ZZ), it can be estimated that in 21-yr-old men ICW is 66% of TBW. For the average RMR of the old groups to fall on the same regression line of RMR vs. TBW as for young men (Fig. 3), ICW would be only 55% of TBW in old men and only 46% in old women. The equations of Pierson et al. (ZZ), however, estimate ICW to be 64% of TBW in old men and 59% in old women. Thus the magnitude of the changes in distribution of TBW between ICW and ECW required to account for the lower rates of energy expenditure in old men and women appear far greater than the age-related alterations reported previously (22). Therefore, it is unlikely that the age-related reduction in RMR solely reflects alterations in body water compartments and not a difference in the rate of energy expended by lean tissue. The isotope-dilution method also assumes that the ratio TBW/FFM is relatively constant (-73%) and does not vary significantly in adults as age advances (26). Our calculations indicate that we must overestimate FFM by 19% in old men and by 44% in old women to account for their lower RMR, making it unlikely that the differences we observed between groups were due to intergroup differences in the ratio TBW/FFM. RMR is the major component of total daily energy expenditure (TEE), usually comprising 50-75% of TEE. Earlier studies in adults found that metabolic rate expressed in absolute terms (kcal/min) decreased with aging (4, 29), but it is unclear whether these changes are caused by alterations in body composition or in the metabolic activity of the active cellular mass. A decline in FFM with advancing age has been documented in both cross-sectional (7, 8) and longitudinal studies (14, 18, 30). Tzankoff and Norris (32) found that the differences in O2 consumption observed between young and old men were primarily due to differences in muscle mass as estimated by urinary creatinine excretion. Nonetheless, creatinine excretion is a measure of skeletal muscle and does not represent all active metabolic tissues of the body. Furthermore, although skeletal muscle is the major component of FFM, it is the least metabolically active component of FFM (16). In a longitudinal study, Keys et al. (18) concluded that the age-related decline in metabolic rate was caused by changes in body composition determined by body density measurement, but in that study subjects were only studied until age 50. To our knowledge, no studies of this kind have been reported in humans >50 yr of age. What factor(s) might underlie the effect of age on RMR was beyond the scope of the current study. In the present study, RMR was significantly lower in the old when expressed in absolute amounts (kcal/min) and remained lower when FFM was employed as a covariate in the analysis. Because RMR in younger subjects is influenced by antecedent diet and exercise, the possibility exists that age-related reductions in.food intake and in physical activity may contribute to our observations. Further studies will be required to examine the impact
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
E238
AGE
EFFECTS
ON
BODY
COMPOSITION
of these factors on RMR. On the other hand, differences in FFM appeared to explain the lower RMR in old women compared with the old men. These data are, therefore, in agreement with those of several investigators who found no difference between adult men and women aged 18-82 yr, in the relationship of RMR/FFM (20, 33), although their analysis did not consider age per se as a factor. These findings suggest that RMR falls with aging in men and women and that this decline in RMR is caused by a decrease both in FFM and also in the metabolic activity of FFM. In summary, these data emphasize the importance of the choice of the method used for assessing body composition in humans. Skinfold measures may be a useful indicator of regional fatness, but results are affected by age and gender. Bioelectrical impedance analysis correlates well with the isotope-dilution technique and appears to be a useful method for the measurement of FFM. Finally, it is clear that differences in FFM cannot fully account for the lower RMR in the old, suggesting that aging per se is associated with an alteration in tissue energy metabolism. We gratefully acknowledge the technical assistance of P. H. Lim, A. Thompson, M. A. Lee, F. Roingeard, and the staff of the Clinical Research Centers, the secretarial assistance of A. Keefe and J. Longstreet, and the volunteers whose participation made this work possible. This work was supported by National Institutes of Health Grants AG-00599, RR-00088, and RR-00954. Address for reprint requests: N. K. Fukagawa, Gerontology Division, Beth Israel Hospital, 330 Brookline Ave., Boston, MA 02215. Received
3 January
1990; accepted
in final
form
9 April
1990.
REFERENCES 1. BANDINI, L. G., D. A. SCHOELLER, AND W. H. DIETZ, JR. Energy expenditure in obese and nonobese adolescents. Pediatr. Res. 27: 198-283,199O. 2. BANDINI, L. G., D. A. SCHOELLER, J. EDWARDS, V. R. YOUNG, S. H. OH, AND W. H. DIETZ. Energy expenditure during carbohydrate overfeeding in obese and nonobese adolescents. Am. J. Physiol. 256 (Endocrinol. Metab. 19): E357-E367, 1989. 3. BERNSTEIN, R. S., J. C. THORNTON, M. U. YANG, Y. WANG, A. M. REDMOND, R. N. PIERSON, JR., F. X. PI-SUNYER, AND T. B. VAN ITALLIE. Prediction of the resting metabolic rate in obese patients. Am. J. Clin. Nutr. 37: 595-602, 1983. 4. BOOTHBY, W. M., J. BERKSON, AND H. L. DUNN. Studies of the energy of metabolism of normal individuals: a standard for basal metabolism with a nomogram for clinical application. Am. J. Physiol. 116: 468-484, 1936. 5. BURR, M. L., AND K. M. PHILLIPS. Anthropometric norms in the elderly. Br. J. Nutr. 51: 165-169, 1984. 6. CHUMLEA, W. C., A. F. ROCHE, AND D. MUKHERJEE. Some anthropometric indices of body composition for elderly adults. J. Gerontol. 41: 36-39, 1986. 7. COHN, S. H., D. VARTSKY, S. YASUMURA, A. SAWITSKY, I. ZANZI, A. VASWANI, AND K. J. ELLIS. Compartmental body composition based on total body nitrogen, potassium, and calcium. Am. J. Physiol. 239 (Endocrinol. Metab. 2): E524-E530, 1980. 8. COHN, S. H., A. VASWANI, I. ZANZI, AND K. J. ELLIS. Effect of aging on bone mass in adult women. Am. J. Physiol. 230: 143-148, 1976. 9. DURNIN, J. V. G. A., AND J. WOMERSLEY. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br. J. Nutr. 32: 77-97, 1974. 10. ELLIOT, D. L., L. GOLDBERG, K. S. KUEHL, AND W. M. BENNETT. Sustained denression of the resting metabolic rate after massive
AND
METABOLIC
RATE
weight loss. Am. J. Clin. Nutr. 49: 93-96, 1989. 11. FAO/WHO/UNU. Energy and protein requirements. Geneva: World Health Organization, 1985, p. 71-79. (Technical Report Ser. 724) 12. FLYNN, M. A., G. B. NOLPH, A. S. BAKER, W. M. MARTIN, AND G. KRAUSE. Total body potassium in aging humans: a longitudinal study. Am. J. Clin. Nutr. 50: 713-717, 1989. Pediatrics 13. FORBES, G. B. Lean body mass and fat in obese children. 34: 308-314, 1964. 14. FORBES, G. B. The adult decline in lean body mass. Hum. BioL. 48: 162-173, 1976. 15. FRISANCHO, A. R. New standards of weight and body composition by frame size and height for assessment of nutritional status of adults and the elderly. Am. J. Clin. Nutr. 40: 808-819, 1984. 16. HOLLIDAY, M. A., C. POTTER, A. JARRAH, AND S. BEARG. The relation of metabolic rate to body weight and organ size. Pediutr. Res. 1: 185-195, 1967. 17. KATCH, V. L. Use of the oxygen/body weight ratio in correlation analyses: spurious correlations and statistical considerations. Med. Sci. Sports 5: 253-257, 1973. 18. KEYS, A., H. L. TAYLOR, AND F. GRANDE. Basal metabolism and age of adult man. Metabolism 22: 579-587, 1973. 19. KUSHNER, R. F., AND D. A. SCHOELLER. Estimation of total body water by bioelectrical impedance analysis. Am. J. Clin. Nutr. 44: 417-424,1986. 20. OWEN, 0. E. Resting metabolic requirements of men and women. Muyo CLin. Proc. 63: 503-510, 1988. 21. PACE, N., AND E. N. RATHBUN. Studies on body composition. III. The body water and chemically combined nitrogen content in relation to fat content. J. BioZ. Chem. 158: 685-691, 1945. 22 PIERSON, R. N., JR., J. WANG, E. W. COLT, AND P. NEUMANN. Body composition measurements in normal man: the potassium, sodium, sulfate and tritium spaces in 58 adults. J. Chronic Dis. 35: 419-428,1982. 23 RAVUSSIN, E., B. BURNAND, Y. SCHUTZ, AND E. JEQUIER. Twentyfour-hour energy expenditure and resting metabolic rate in obese, moderately obese, and control subjects. Am. J. Clin. Nutr. 35: 566573, 1982. 24. ROCHE, A. F., R. M. SIERVOGEL, W. C. CHUMLEA, AND P. WEBB. Grading body fatness from limited anthropometric data. Am. J. Clin. Nutr. 34: 2831-2838, 1981. 25. Ross, R., L. LEGER, P. MARTIN, AND R. ROY. Sensitivity of bioelectrical impedance to detect changes in human body composition. J. Appl. Physiol. 67: 1643-1648, 1989. 26. SCHOELLER, D. A. Changes in total body water with age. Am. J. CLin. Nutr. 50: 1176-1181, 1989. 27. SCHOELLER, D. A., E. VAN SANTEN, D. W. PETERSON, W. DIETZ, J. JASPAN, AND P. D. KLEIN. Total body water measurement in humans with “O- and 2H-labelled water. Am. J. CZin. Nutr. 33: 2686-2693,198O. 28. SEGAL, K. R., M. VAN LOAN, P. I. FITZGERALD, J. A. HODGDON, AND T. B. VAN ITALLIE. Lean body mass estimation by bioelectrical impedance analysis: a four-site cross-validation study. Am. J. Clin. Nutr. 47: 7-14, 1988. M. J. YIENGST, A. H. NORRIS, G. 29. SHOCK, N. W., D. M. WATKIN, W. GAFFNEY, R. I. GREGERMAN, AND J. A. FALZONE. Age differences in the water content of the body as related to basal oxygen consumption in males. J. Gerontol. 18: l-8, 1963. 30. SHUKLA, K. K., K. J. ELLIS, C. S. DOMBROWSKI, AND S. H. COHN. Physiological variation of total body potassium in man. Am. J. Physiol. 224: 271-274, 1973. 31. TANNER, J. M. Fallacy of per-weight and per-surface area standards, and their relation to spurious correlation. J. Appl. Physiol. 2: l-15,1949. 32. TZANKOFF, S. P., AND A. H. NORRIS. Effect of muscle mass decrease on age-related BMR changes. J. AppZ. Physiol. 43: 10011006, 1977. 33. WEBB, P. Energy expenditure and fat-free mass in men and women. Am. J. Clin. Nutr. 34: 1816-1826, 1981. 34. WEIR, J. B. New methods for calculating metabolic rate with special reference to protein metabolism. J. Physiol. Lond. 109: I-9, 1949. AnuZysis. Englewood Cliffs, NJ: Prentice 35. ZAR, J. H. Biostutisticul Hall, 1984, p. 323-325.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
and resting metabolic
rate
NAOMI K. FUKAGAWA, LINDA G. BANDINI, AND JAMES B. YOUNG Charles A. Dana Research Institute and the Harvard Thorndike Laboratory, Department of Medicine, Beth Israel Hospital, and Department of Pediatrics and Division on Aging, Harvard Medical School, Boston, Massachusetts 02115; and The Clinical Research Center, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139
FUKAGAWA, NAOMI K., LINDA G. BANDINI, AND JAMES B. YOUNG. Effect of age on body composition and resting metabolic rate. Am. J. Physiol. 259 (Endocrinol. Metab. 22): E233-E238, 1990.-The relationship between fat-free mass (FFM) and resting metabolic rate (RMR) was compared in young men (n = 24; age 18-33 yr), old men (n = 24; 69-89 yr), and old women (72 = 20; 67-75 yr). Body composition was assessed using anthropometry, bioelectrical impedance analysis (BIA), and isotope dilution with “O-labeled water. RMR was measured at least twice using an open-circuit indirect calorimetry system with a ventilated hood. The results indicate that the different methods for assessing body composition vary substantially and should not be used interchangeably. Anthropometry was not adequate to assess group differences in body fatness, although skinfold measures may be appropriate for within-group comparisons. BIA correlated well with the isotope-dilution technique and may be a useful measure of FFM. Finally, RMR was lower in the old men than the young (1.04 t 0.02 vs. 1.24 t 0.03 kcal/min, P < 0.001) and remained lower even when adjusted for FFM estimated by isotope dilution (P < 0.001). RMR in the women was also lower (0.84 t 0.02 kcal/min), but in contrast to the difference between young and old men, RMR adjusted for FFM did not differ (P = 0.16) between old men and women. Therefore, it is clear that differences in FFM cannot fully account for the lower RMR in the old, suggesting that aging is associated with an alteration in tissue energy metabolism.
anthropometry, bioelectrical impedance analysis (BIA), and isotope dilution with 180-labeled water ( [lSO]HzO). Although there are numerous studies reporting the changes in subcutaneous skinfold thickness and body fat distribution in the elderly, there is very little information on which method best estimates body fatness in an old population. We, therefore, assessedhow well measures of body fatness derived from anthropometry and BIA correlated with percent body fat estimated by isotope dilution, which was chosen as our “gold standard.” One of the limitations of the isotope-dilution method, as with all indirect measures of body composition, is that it is based on the assumption that the components of FFM are of relatively constant composition. Our results indicate that the different methods for assessingbody composition in humans vary substantially and should not be used interchangeably; anthropometry measuresbody fatness, whereas BIA and [lSO]HzO measure hydrated lean tissue. Furthermore, it is clear from our data that differences in FFM cannot fully account for the lower resting metabolic rates in the old, suggesting that aging is associated with an alteration in tissue energy metabolism. MATERIALS
AND METHODS
Subjects. Twenty-four young (18-33 yr) and 24 elderly (69-89 yr) men and 20 elderly (67-75 yr) women in good health, as determined by history and physical examinations, were studied at the Clinical Research Centers AGING IS ASSOCIATED with a decrease in metabolic rate (CRC) of the Beth Israel Hospital (BIH) and Massachuthat has been attributed to an age-related decline in setts Institute of Technology (MIT). None were taking “active” cellular mass, i.e., fat-free mass (FFM) and a any medications at the time of investigation, and all had concomitant increase in body fat (l&18). Although many maintained stable body weights in the previous 6 mo. All studies in both children and adults have shown that subjects were carefully informed of the nature, purpose, metabolic rate is highly correlated with FFM (1, 3, 23, and possible risks of the studies before giving their 33), it is not clear whether the decline in metabolic rate written consent to participate. The protocol used was with aging is due entirely to a decrease in FFM or approved by the Clinical Investigation Committee at the whether it also reflects a reduction in the metabolic BIH and by the Committee on the Use of Humans as activity of lean tissue. A number of studies have exam- Experimental Subjects at MIT. ined the relationship between FFM and resting metabolic Protocol. All subjects were admitted to the CRC the rate (RMR) in obese and lean individuals and during evening before the study. After a 12-h overnight fast, hypocaloric feeding (10, 23); however, to our knowledge, they were awakened, asked to void, weighed in a hospital no studies have examined the relationship of FFM and gown, and returned to bed to rest until their metabolic RMR with advancing age. rate was measured for 45 min as described below. After In this study, the relationship between FFM and RMR completion of the measurement of metabolic rate, body was compared in young men, old men, and old women. composition was determined by isotope dilution using Because of concerns that the methods employed to esti- [ “O]H20 (n = 68), BIA (n = 66), and anthropometry (n mate FFM might affect the conclusions, we assessed = 60). Repeat measurements of metabolic rate were made at least one to two times during their stay in the CRC, body composition using several techniques, including fat-free mass; energy expenditure;
isotope dilution
0193-1849/90
$1.50 Copyright
0 1990 the American
Physiological
Society
E233
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
E234 TABLE
AGE
EFFECTS
ON
BODY
COMPOSITION
AND
METABOLIC
RATE
1. Subject characteristics P
Group Young
men
(n = 24)
Age, Y*
21tl
Height, cm Weight, kg Body mass index, kg/m2 Fat-free mass*, kg Total body fat*, kg %Body fat* Values
TABLE
are means
t
176.721.7 72.3k2.1 23.1t0.5 55.4t1.9 16.9tl.O 23tl
SE. * Estimated
2. Anthropometric
by isotope
Old men
(n = 24) 75tl 172.1t1.2 72.8t1.6 24.6t0.5 47.7t0.9 25.1tl.0 34t1
Old women
(n=
20)
72tl 158.8k1.2 59.4t2.2 23.5t0.8 35.8t0.8 23.6t1.5 3921
-
Young vs. old men
Old men vs. women
0.001 0.05 0.41 0.05 0.001 0.001 0.001
0.03 0.001 0.001 0.25 0.001 0.41 0.003
dilution.
measures in young men, old men, and old women P
Group Anthropometric Parameters
Young
men
(n = 20) Skinfold Tricep Biceps Subscapular Suprailiac Sum of 4 skinfolds Waist-to-hips ratio Values
are means
10.3kO.9 3.7rto.4 12.1t0.8 14.2t1.5 40.4t3.0 0.85t0.01
Old men
(n = 20) lO.lt0.6 4.2kO.4 16.6k1.6 14.7t1.1 45.622.7 0.95&0.01
Old women
(n = 20) 2O.lk1.4 12.4k1.4 20.8t1.6 18.9t1.9 72.2k5.5 0.86t0.02
Young vs. old men
Old men vs. old women
0.88 0.44 0.02 0.81 0.21 0.001
0.001 0.001 0.07 0.06 0.001 0.001
t SE in mm.
each after 15-30 min rest periods. Subjects were given practice sessions the evening before the studies to familiarize them with the ventilated-hood system of indirect calorimetry. The BIA method to assess body composition was used as previously described (28). Measurements were made with a portable impedance analyzer (BIA 103, RJL Systems, Detroit, MI) while the subjects lay comfortably in bed with their limbs abducted. Resistance (R) to the flow of a 50-kHz, 800-PA current was then measured on a Oto 1,000-Q scale. Anthropometric measurements were performed independently by two investigators. Skinfold thickness of the nondominant side of the body were measured in triplicate with a Lange skinfold caliper (Cambridge Instruments, Cambridge, MA) exerting a pressure of 10 g/mm2, and the mean of all measurements was used for the analyses. Regional fat distribution was estimated from the ratio of waist-to-hip circumferences. The waist was measured at the umbilicus in the men, but, because of differences in the position of the umbilicus in the women, the waist measurement used for the women was the smallest abdominal circumference. The hip measurement was made at the widest/broadest point over the buttocks in all groups. Skinfold measures were made at the triceps, biceps, subscapular, and suprailiac sites. Measurement of energy expenditure. An open-circuit indirect calorimetry system with a ventilated hood and interfaced to a desktop computer was used for all indirect calorimetric measurements at both BIH and MIT. The system used at the BIH was as follows. Air was withdrawn from the hood at 50 l/min, as measured by a rotameter (Fisher and Porter, Warminster, PA); concentrations of O2 and CO2 were measured using an electrochemical O2 sensor and infrared CO2 sensor (Ametek, Pittsburgh, PA). The output from the analyzers was
converted to digital signals by the computer for calculating the rates of 0, consumption, CO2 production, and metabolic rate, using Weir’s equation (34). Values were averaged and saved at I-min intervals after correction to standard temperature and pressure. The analyzers were calibrated before each measurement using standard gases, and the system was calibrated weekly using fixed concentrations of gas and found to be reproducible to within 3% of predicted values. The system used at MIT has been described elsewhere (2). The majority of the subjects (n = 47) were studied at BIH, and both young (n = 11) and old (n = 10) were studied at MIT, so that possible systematic errors introduced by differences between the indirect calorimetry systems were avoided. In addition, 10 men had their resting metabolic rates measured at both sites on separate occasions to determine the comparability of the two indirect calorimetry systems. No significant differences were found in measurements obtained at both MIT and BIH (P > 0.10, paired t test). The average difference between measurements at both sites was 0.02 t 0.01 kcal/min or 2.4%. Analytical methods. Body composition was determined by isotope dilution techniques using [180]H20 (27). Serum samples were obtained both before and 4 h after ingestion of [ “O]H20 [0.07 g [ “O]/kg estimated total body water (TBW)], and analyzed commercially for “0 enrichment using isotope ratio mass spectrometry (Geochron Laboratories, Cambridge, MA). The measurement of 180 enrichment varied Cl%. Oxygen dilution space (Do) was calculated from the increase in 180 enrichment of the 4-h serum relative to predose values as previously described (27). TBW was calculated from Do, assuming D = TBW x 1.01. FFM was calculated from TBW, asOsuming the relationship FFM = TBW/0.73 (21). Statistics. All data are presented as means t SE. The relationships between FFM and RMR in the three groups
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
AGE
EFFECTS
ON
BODY
COMPOSITION
were examined by analysis of covariance. Frequently, comparison of RMR data among subjects or between groups was done by expressing RMR per kilogram of FFM, i.e., by dividing RMR by FFM (17). However, the use of this ratio (RMR/FFM) is only appropriate when the correlation between RMR and FFM is perfect, i.e., r = 1.0, and when the mathematical equation relating the two parameters is one in which RMR is in constant proportion to FFM with an intercept equal to zero (17, 31). Analysis of covariance allows one to derive adjusted RMR values according to a specific relationship between RMR and FFM, thereby avoiding the assumption of a perfect correlation and zero intercept. The data were examined for parallelism, and results did not indicate that the slopes between the regression lines for each group were different. Age-gender was used as the class variable and FFM as the covariate. Pairwise comparisons of interest were determined a priori to be old men vs. young men and old men vs. old women. Comparisons between groups were made using the Student’s t test and Pearson product-moment correlations (BMDP Statistical Package, Los Angeles, CA). Intraclass correlations were calculated according to Zar (35). RESULTS
The characteristics of the volunteers are summarized in Table 1. The young men were taller than the elderly men (P < 0.05) but of comparable body weight (P = 0.41). The women were shorter and weighed less than the old men (P c 0.001). Body mass index, the ratio of body weight to height squared, was slightly, but significantly, higher in the old men compared with the young men (P < 0.05) and similar in both old men and women (P = 0.25). FFM estimated by isotope dilution was significantly lower in the old men (47.7 t 0.09 kg) than in the young men (55.4 t 1.9 kg; P < 0.001). FFM in the women (35.8 t 0.8 kg) was significantly lower than in the old men (P < 0.001). Total body fat, calculated as the difference between body weight and FFM, was greater in the old men compared with the young men (P < 0.001) but was similar in both groups of old (P = 0.41). The percentage of body weight as fat was higher in the old men than in the young men (P < O.OOl), reflecting the greater fat mass in the old men. Percent body fat was higher in the women than in the old men (P < O.OO3),reflecting the lower body weight of the women and hence greater relative fatness. Despite the differences in fat mass and in body fatness between young and old men, triceps (TSF), biceps (BSF), and suprailiac (SISF) skinfold thicknesses were similar in young and old men (Table 2); only the subscapular skinfold (SSF) was higher in the old men compared with the young men (P < 0.02). On the other hand, whereas elderly men and women differed in percent body fat but not in fat mass, TSF and BSF were substantially greater in the women than in the men (P < O.OOl), whereas SISF and SSF were only slightly greater (P < 0.06 and ~0.07, respectively). As suggested by the elevation in the waistto-hips ratio (WHR) in old men in comparison with the other two groups (P < O.OOl), differences in body fat distribution may contribute to the discordance between
AND
METABOLIC
E235
RATE
measurements of skinfold thickness and body fat among the three groups. Skinfold thicknesses within each study group, either individually or as the sum of the four measurements, generally correlated with measurements of body fat. TSF and BSF correlated highly with percent body fat in old men and women, but less well in young men (Table 3). SSF, however, correlated well with percent fat in young men and old women but not as well in old men. SISF, in contrast, did not correlate with percent body fat i.n either young or old men but did correlate highly with percent fat in the women. The sum of all four skinfolds correlated highly with percent body fat in all groups. Similarly, WHR was also strongly correlated with body fatness in all three groups, whereas BMI was significantly related to percent fat only in the two old groups and not in young men. These findings suggest that skinfold measures may serve as a valid index of fatness within specific age-gender groups, but may be considerably less useful as a means of comparing fatness among different agegender groups. Height squared divided by resistance obtained from BIA measurements correlated very strongly in all groups with the estimate of FFM based on isotope dilution (young men: r = 0.90, P c 0.001; old men: r = 0.85, P < 0.001; and old women: r = 0.87, P < 0.001, Fig. 1). Metabolic rate results are summarized in Table 4 and Fig. 2. FFM and RMR were significantly correlated within each group (young men: r = 0.82; old men: r = 3. Correlation coefficients for relationship between anthropometric measures and percent body fat TABLE
Young
Anthropometric Parameters
Skinfold Triceps Biceps Subscapular Suprailiac Sum of 4 skinfolds Waist-to-hips ratio Body mass index Values
Men
r
estimated
Old Men
P
r
Old Women P
r
P
0.45
0.06
0.55
0.39
0.10
0.51
0.50 0.29 0.47
0.04 0.23 0.04
0.39 0.37 0.58
0.11
0.007
0.68 0.75 0.80
0.59 0.16
0.008 0.47
0.71
0.001
0.47
0.63
0.001
0.74
by isotope
dilution.
0.01 0.02
0.09
0.70
0.001
0.69
0.001 0.001 0.001 0.001 0.038 0.001
0 0
80
1 0
0
0
0 0
l l
a0
0 0
0.
Young Men Old Men Old Women
+ 6b FFM FIG.
1. Relationship
fat-free mass (FFM) men, and old women.
between estimated
80
(kg)
height squared-to-resistance by isotope dilution in young
ratio and men, old
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
E236
AGE
EFFECTS
ON
BODY
COMPOSITION
4. Metabolic rate in young- men, old men, and old women TABLE
Resting Metabolic Rate
Young
Men
Old Men
Old Women
Mean Measured Adjusted for FFM WHO/FAO/UNU
1.24kO.03 1.04kO.O2* 0.84t0.02-f 1.13kO.02 1.03&0.02* 0.99+0.02$ estimates
1.23t0.02
1.02&0.01
0.85kO.02
Values are means & SE in kcal/min. FFM, fat-free mass. * P < 0.002 for comparison with young men; t P < 0.001 for comparison with old men; $ P = 0.16 for comparison with old men; § Ref. 11.
1 0.50
I 30
0
Young
l
Old
Men
Old
Women
A .
, 40
.
, 50
.
, 60
.
Men
, 70
1
{ 80
FFM (kg) FIG. 2. Relationship between resting metabolic mass (FFM) estimated by isotope dilution in young old women. Overall, r = 0.89, P < 0.001.
rate and fat-free men, old men, and
0.48; and old women: r = 0.74; as well as overall: r = 0.89, P < 0.001). The intraclass correlation coefficient of the multiple determinations of RMR within each individual was 0.968, suggesting that the measures were highly reproducible. Absolute RMR, representing the mean of two to eight measurements per individual, was lower in the old men than in the young (1.04 t 0.02 vs. 1.24 t 0.03 k cal/ min; P < 0.001); RMR remained lower in old men (P c 0.001) even when adjusted for FFM. RMR was also lower in the old women (0.84 t 0.02 kcal/ min) compared with the old men (P < 0.001) but, in contrast to the difference between young and old men, group means did not differ after adjustment for FFM (P = 0.16). The absolute values for RMR in all groups were similar to those estimated from the 1985 World Health Organization/Food and Agriculture Organization of the United Nations/United Nations University guidelines (11) (Table 4). DISCUSSION
As the elderly population increases in number, a growing health concern focuses on geriatric nutritional status and methods of nutritional assessment. In this study, measurements of skinfold thickness and bioelectrical impedance were compared with estimates of body composition obtained by isotope dilution in young and old individuals. In addition, the effect of age on the relationship between metabolic rate and FFM was examined. Although several references for anthropometric norms for the elderly have been established (5, 6, 15), the relationship between them and total body fat or percent
AND
METABOLIC
RATE
body weight as fat is not clearly established. Because the correlation between skinfold thickness and body fatness decreasesas body fat increases (l3), it might be expected that skinfold thickness would be a less robust index of body fatness in the old population because of the welldocumented increase in body fat with advancing age. Our data indicate that this is not the case, because skinfold thickness measurements in general correlated more highly with percent body fat in the old than in the young. Although others have found stronger relationships between TSF or SSF and total body fat or body fatness in children and young adults than reported here (24), these differences may be due, in part, to the small sample size and narrow range of body fatness observed within our young men (23 t 1%). Despite statistically significant correlations between several anthropometric measures and body fatness, a substantial portion of the variance in relative fatness (X50%) in all groups remained unexplained by anthropometry. On the other hand, the utility of skinfold thickness measurements in intergroup comparisons of body fat or body fatness is open to question. Despite substantial differences between young and old men in fat mass or percent body fat, neither TSF, BSF, nor SISF differed in the two groups. Similarly, although body fat was not different in old men and women and percent body fat was only slightly, albeit significantly, greater in the women, TSF and BSF were markedly increased in the women. The truncal skinfolds (SSF and SISF) were, however, slightly higher in the old women. Differences in the pattern of fat distribution in the groups studied may influence the interpretation of our findings, since the WHR differed significantly between the two groups of men and between the old men and women. These findings demonstrate that measures of anthropometry are not adequate to assess group differences in body fatness, although they may be appropriate for withingroup comparisons. Our results have potential implications for the estimation of FFM from anthropometric data. Such determinations rely on regression equations relating various skinfolds to some more direct measure of body fatness, e.g., body densitometry measurements (9). FFM is then calculated as the difference between body weight and the product of body weight and percent body fat. Any potential errors in skinfold thickness measurements or in their relation to body fatness are then magnified in the final calculation. Because the equations generally used to estimate body fat from skinfold measurements differ across age groups (9), a point that is consistent with the data presented here, age and gender-specific regression equations may be required to achieve valid information by this approach. Our results demonstrate that skinfold thickness measures differ between young and old men and between old men and women and hence should not be used as an index of body fat to assessthe relationship of FFM and RMR in young and old individuals without appropriate standardization. Recently, BIA has been adapted for use in the assessment of FFM in human subjects. BIA is rapid and noninvasive, and it has the advantage of being easy to use, even at the bedside. The results from BIA correlate
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
AGE
EFFECTS
ON
BODY
COMPOSITION
well with those obtained by densitometry and isotopedilution techniques (19,25,28). In the current study, the impact of age and gender on the relationship between BIA and FFM as estimated by isotope dilution of [l’O]Hz0 was examined. The ratio of height squared over resistance (H2/R), the immediate output of BIA, correlated significantly with FFM in all groups. Although FFM can be derived from H2/R, the derivation requires the use of a specific regression equation relating this ratio to an independent measure of FFM. The high correlations between H2/R and FFM observed in this investigation imply that both measures provide comparable estimates of TBW, from which FFM is derived and, furthermore , that in a study t such as this in which FFM is employed as a cova riate in the analysis of another variable (e.g., RMR), the use of the ratio H2/R might be equally suitable as a covariate in the analysis. In the present study, FFM was derived from measures of TBW. Because neutral fat does not bind water, measurement of TBW is a means of estimating the nonfat compartment of the body. The use of TBW in the estimation of body composition assumes a two-compartment model (fat and lean) in which FFM includes the extracellular fluid mass (EC W) ) intracellular components (ICW), and skeletal structures. This analysis also assumes that the ratio TBW/FFM is independent of age or gender. Because this latter assumption may not be entirely correct over the age ran .ges studied, the question arises whether our finding of diminished RMR in relation to FFM in young and old subjects might be attributable to inaccuracies inherent in this two-compartment model for body composition. Specifically, because the distribution of body water between extracellular and intracellular compartments is known to differ with age (ZZ), could the results depicted in Fig. 2 reflect a systematic overestimation of FFM in the two old groups and not a reduction in energy expenditure by lean tissue as we suggest? Because only TBW was measured in our study, no definitive answer to this question can be obtained from our data. It is, however, possible to estimate the changes in body water distribution between intracellular and extracellular spaces that would be required if group dif1.75 -I-
1
0 .
I 20
’ Total
I 30
= Body
A I 40
Young OldMen
Men
Old Women ’ I . 50
60
Water
(kg) 3. Relationship between resting metabolic rate (RMR) and body water (TBW) estimated by isotope dilution in young men, men, and old women. Dotted lines and respective midpoints repreaverage TBW of old groups if their RMR fell on same regression as for young groups.
FIG.
total old sent line
AND
METABOLIC
RATE
E237
ferences in RMR were due solely to age-related alterations in body water distribution. For the purposes of this discussion, ICW will be assumed to represent more closely than TBW the metabolically active compartment in which energy is expended. From the equations of Pierson et al. (ZZ), it can be estimated that in 21-yr-old men ICW is 66% of TBW. For the average RMR of the old groups to fall on the same regression line of RMR vs. TBW as for young men (Fig. 3), ICW would be only 55% of TBW in old men and only 46% in old women. The equations of Pierson et al. (ZZ), however, estimate ICW to be 64% of TBW in old men and 59% in old women. Thus the magnitude of the changes in distribution of TBW between ICW and ECW required to account for the lower rates of energy expenditure in old men and women appear far greater than the age-related alterations reported previously (22). Therefore, it is unlikely that the age-related reduction in RMR solely reflects alterations in body water compartments and not a difference in the rate of energy expended by lean tissue. The isotope-dilution method also assumes that the ratio TBW/FFM is relatively constant (-73%) and does not vary significantly in adults as age advances (26). Our calculations indicate that we must overestimate FFM by 19% in old men and by 44% in old women to account for their lower RMR, making it unlikely that the differences we observed between groups were due to intergroup differences in the ratio TBW/FFM. RMR is the major component of total daily energy expenditure (TEE), usually comprising 50-75% of TEE. Earlier studies in adults found that metabolic rate expressed in absolute terms (kcal/min) decreased with aging (4, 29), but it is unclear whether these changes are caused by alterations in body composition or in the metabolic activity of the active cellular mass. A decline in FFM with advancing age has been documented in both cross-sectional (7, 8) and longitudinal studies (14, 18, 30). Tzankoff and Norris (32) found that the differences in O2 consumption observed between young and old men were primarily due to differences in muscle mass as estimated by urinary creatinine excretion. Nonetheless, creatinine excretion is a measure of skeletal muscle and does not represent all active metabolic tissues of the body. Furthermore, although skeletal muscle is the major component of FFM, it is the least metabolically active component of FFM (16). In a longitudinal study, Keys et al. (18) concluded that the age-related decline in metabolic rate was caused by changes in body composition determined by body density measurement, but in that study subjects were only studied until age 50. To our knowledge, no studies of this kind have been reported in humans >50 yr of age. What factor(s) might underlie the effect of age on RMR was beyond the scope of the current study. In the present study, RMR was significantly lower in the old when expressed in absolute amounts (kcal/min) and remained lower when FFM was employed as a covariate in the analysis. Because RMR in younger subjects is influenced by antecedent diet and exercise, the possibility exists that age-related reductions in.food intake and in physical activity may contribute to our observations. Further studies will be required to examine the impact
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
E238
AGE
EFFECTS
ON
BODY
COMPOSITION
of these factors on RMR. On the other hand, differences in FFM appeared to explain the lower RMR in old women compared with the old men. These data are, therefore, in agreement with those of several investigators who found no difference between adult men and women aged 18-82 yr, in the relationship of RMR/FFM (20, 33), although their analysis did not consider age per se as a factor. These findings suggest that RMR falls with aging in men and women and that this decline in RMR is caused by a decrease both in FFM and also in the metabolic activity of FFM. In summary, these data emphasize the importance of the choice of the method used for assessing body composition in humans. Skinfold measures may be a useful indicator of regional fatness, but results are affected by age and gender. Bioelectrical impedance analysis correlates well with the isotope-dilution technique and appears to be a useful method for the measurement of FFM. Finally, it is clear that differences in FFM cannot fully account for the lower RMR in the old, suggesting that aging per se is associated with an alteration in tissue energy metabolism. We gratefully acknowledge the technical assistance of P. H. Lim, A. Thompson, M. A. Lee, F. Roingeard, and the staff of the Clinical Research Centers, the secretarial assistance of A. Keefe and J. Longstreet, and the volunteers whose participation made this work possible. This work was supported by National Institutes of Health Grants AG-00599, RR-00088, and RR-00954. Address for reprint requests: N. K. Fukagawa, Gerontology Division, Beth Israel Hospital, 330 Brookline Ave., Boston, MA 02215. Received
3 January
1990; accepted
in final
form
9 April
1990.
REFERENCES 1. BANDINI, L. G., D. A. SCHOELLER, AND W. H. DIETZ, JR. Energy expenditure in obese and nonobese adolescents. Pediatr. Res. 27: 198-283,199O. 2. BANDINI, L. G., D. A. SCHOELLER, J. EDWARDS, V. R. YOUNG, S. H. OH, AND W. H. DIETZ. Energy expenditure during carbohydrate overfeeding in obese and nonobese adolescents. Am. J. Physiol. 256 (Endocrinol. Metab. 19): E357-E367, 1989. 3. BERNSTEIN, R. S., J. C. THORNTON, M. U. YANG, Y. WANG, A. M. REDMOND, R. N. PIERSON, JR., F. X. PI-SUNYER, AND T. B. VAN ITALLIE. Prediction of the resting metabolic rate in obese patients. Am. J. Clin. Nutr. 37: 595-602, 1983. 4. BOOTHBY, W. M., J. BERKSON, AND H. L. DUNN. Studies of the energy of metabolism of normal individuals: a standard for basal metabolism with a nomogram for clinical application. Am. J. Physiol. 116: 468-484, 1936. 5. BURR, M. L., AND K. M. PHILLIPS. Anthropometric norms in the elderly. Br. J. Nutr. 51: 165-169, 1984. 6. CHUMLEA, W. C., A. F. ROCHE, AND D. MUKHERJEE. Some anthropometric indices of body composition for elderly adults. J. Gerontol. 41: 36-39, 1986. 7. COHN, S. H., D. VARTSKY, S. YASUMURA, A. SAWITSKY, I. ZANZI, A. VASWANI, AND K. J. ELLIS. Compartmental body composition based on total body nitrogen, potassium, and calcium. Am. J. Physiol. 239 (Endocrinol. Metab. 2): E524-E530, 1980. 8. COHN, S. H., A. VASWANI, I. ZANZI, AND K. J. ELLIS. Effect of aging on bone mass in adult women. Am. J. Physiol. 230: 143-148, 1976. 9. DURNIN, J. V. G. A., AND J. WOMERSLEY. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br. J. Nutr. 32: 77-97, 1974. 10. ELLIOT, D. L., L. GOLDBERG, K. S. KUEHL, AND W. M. BENNETT. Sustained denression of the resting metabolic rate after massive
AND
METABOLIC
RATE
weight loss. Am. J. Clin. Nutr. 49: 93-96, 1989. 11. FAO/WHO/UNU. Energy and protein requirements. Geneva: World Health Organization, 1985, p. 71-79. (Technical Report Ser. 724) 12. FLYNN, M. A., G. B. NOLPH, A. S. BAKER, W. M. MARTIN, AND G. KRAUSE. Total body potassium in aging humans: a longitudinal study. Am. J. Clin. Nutr. 50: 713-717, 1989. Pediatrics 13. FORBES, G. B. Lean body mass and fat in obese children. 34: 308-314, 1964. 14. FORBES, G. B. The adult decline in lean body mass. Hum. BioL. 48: 162-173, 1976. 15. FRISANCHO, A. R. New standards of weight and body composition by frame size and height for assessment of nutritional status of adults and the elderly. Am. J. Clin. Nutr. 40: 808-819, 1984. 16. HOLLIDAY, M. A., C. POTTER, A. JARRAH, AND S. BEARG. The relation of metabolic rate to body weight and organ size. Pediutr. Res. 1: 185-195, 1967. 17. KATCH, V. L. Use of the oxygen/body weight ratio in correlation analyses: spurious correlations and statistical considerations. Med. Sci. Sports 5: 253-257, 1973. 18. KEYS, A., H. L. TAYLOR, AND F. GRANDE. Basal metabolism and age of adult man. Metabolism 22: 579-587, 1973. 19. KUSHNER, R. F., AND D. A. SCHOELLER. Estimation of total body water by bioelectrical impedance analysis. Am. J. Clin. Nutr. 44: 417-424,1986. 20. OWEN, 0. E. Resting metabolic requirements of men and women. Muyo CLin. Proc. 63: 503-510, 1988. 21. PACE, N., AND E. N. RATHBUN. Studies on body composition. III. The body water and chemically combined nitrogen content in relation to fat content. J. BioZ. Chem. 158: 685-691, 1945. 22 PIERSON, R. N., JR., J. WANG, E. W. COLT, AND P. NEUMANN. Body composition measurements in normal man: the potassium, sodium, sulfate and tritium spaces in 58 adults. J. Chronic Dis. 35: 419-428,1982. 23 RAVUSSIN, E., B. BURNAND, Y. SCHUTZ, AND E. JEQUIER. Twentyfour-hour energy expenditure and resting metabolic rate in obese, moderately obese, and control subjects. Am. J. Clin. Nutr. 35: 566573, 1982. 24. ROCHE, A. F., R. M. SIERVOGEL, W. C. CHUMLEA, AND P. WEBB. Grading body fatness from limited anthropometric data. Am. J. Clin. Nutr. 34: 2831-2838, 1981. 25. Ross, R., L. LEGER, P. MARTIN, AND R. ROY. Sensitivity of bioelectrical impedance to detect changes in human body composition. J. Appl. Physiol. 67: 1643-1648, 1989. 26. SCHOELLER, D. A. Changes in total body water with age. Am. J. CLin. Nutr. 50: 1176-1181, 1989. 27. SCHOELLER, D. A., E. VAN SANTEN, D. W. PETERSON, W. DIETZ, J. JASPAN, AND P. D. KLEIN. Total body water measurement in humans with “O- and 2H-labelled water. Am. J. CZin. Nutr. 33: 2686-2693,198O. 28. SEGAL, K. R., M. VAN LOAN, P. I. FITZGERALD, J. A. HODGDON, AND T. B. VAN ITALLIE. Lean body mass estimation by bioelectrical impedance analysis: a four-site cross-validation study. Am. J. Clin. Nutr. 47: 7-14, 1988. M. J. YIENGST, A. H. NORRIS, G. 29. SHOCK, N. W., D. M. WATKIN, W. GAFFNEY, R. I. GREGERMAN, AND J. A. FALZONE. Age differences in the water content of the body as related to basal oxygen consumption in males. J. Gerontol. 18: l-8, 1963. 30. SHUKLA, K. K., K. J. ELLIS, C. S. DOMBROWSKI, AND S. H. COHN. Physiological variation of total body potassium in man. Am. J. Physiol. 224: 271-274, 1973. 31. TANNER, J. M. Fallacy of per-weight and per-surface area standards, and their relation to spurious correlation. J. Appl. Physiol. 2: l-15,1949. 32. TZANKOFF, S. P., AND A. H. NORRIS. Effect of muscle mass decrease on age-related BMR changes. J. AppZ. Physiol. 43: 10011006, 1977. 33. WEBB, P. Energy expenditure and fat-free mass in men and women. Am. J. Clin. Nutr. 34: 1816-1826, 1981. 34. WEIR, J. B. New methods for calculating metabolic rate with special reference to protein metabolism. J. Physiol. Lond. 109: I-9, 1949. AnuZysis. Englewood Cliffs, NJ: Prentice 35. ZAR, J. H. Biostutisticul Hall, 1984, p. 323-325.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (137.092.098.118) on August 7, 2018. Copyright © 1990 American Physiological Society. All rights reserved.
