Changes in Body Chemical Composition With Age Measured by Total-Body Neutron Activation S. H. Cohn,
A. Vaswani,
I. Zanri,
J. F. Aloia,
Total-body levels of calcium and phosphorus (reflecting skeletal mass) and total-body levels of potassium (reflecting muscle mass) were measured by neutron activation analysis in 39 men and 40 women ages 30-90 yr. In order to intercompare the total body calcium (TBCa) values in a heterogeneous population, such as this, it was necessary to normalize the data for skeletal size. The normalization consisted of dividing the absolute calcium level by the predicted calcium level for each individual matched to a set of critical parameters. The parameter used in the computation of normal values were age, sex, muscle mass, i.e., total body potassium (TBK) and height. For the calcium data of the women, it was necessary to add an age correction factor after the age of 55 yr. The calcium ratio (mean mtio of the predicted to measured TBCa) in men was 1.000 f 7.8% and in women 0.996 f 7.1%. The TBCa of normal males and females can thus be predicted to =t 13% (at the 90% confidence level). An exception to this was found in males (70-90 yr) who exhibited a mean calcium ratio >1.13.
M. S. Roginsky,
and
K. J. Ellis
The derivative of TBCa with time was determined for this population of men and women by taking into account the dependency of calcium on three time dependent variables, height, TBK, and an explicit age correction factor in the case of the women. The mean mte of loss of TBCa in women was 0.37% and 1 .l % per year before and after menopause (50 yr). In the males, the avemge mte of loss of TBCa was 0.7% per year after 50 yr of age. The pattern of total body phosphorus (TBP) loss with age paralleled that of TBCa as the mtio of TBP/TBCa was mther constant with age. The constancy of the mtio suggests that the mineml composition of bone does not change significantly with age. The mte of loss of TBK with age was also related directly to that of TBCa. The mean mtio of TBK/TBCa was 9.9 in females and 8.0 in males and this ratio remained relatively constant from 30-70 yr. Thus, the mechanism responsible for the loss of bone with age, whether nutritional deficiency or decreased gonadal function and physical activity may also be responsible for the loss of muscle mass with age.
T
HE QUANTITATIVE ASSESSMENT of the chemical composition of the human body as a function of age is of basic interest as well as of clinical value in diagnosing metabolic disorders. Data on constituents such as total body calcium also serve as objective criteria in the evaluation of the efficacy of therapeutic programs for such metabolic disorders as osteoporosis and renal osteodystrophy. The changes in chemical composition of the body with age can be measured for the first time, in vivo, by the technique of whole-body neutron activation (TBNAA). In the past, various in vivo methods have been used for the study of These measurements were performed in accordance with the principles of the Declaration of Helsinki and the regulations of the U.S. Department of Health, Education and Welfare. An informed consent was obtained for each subject. From the Medical Research Center, Brookhaven National Laboratory, Upton, N. Y. Received for publication May 5. 1975. Supported by the U.S. Energy Research and Development Administration. Reprint requests should be addressed to Dr. S. H. Cohn, Medical Research Center. Brookhaven National Laboratory, Upton, N. Y. 11973. 0 1976 by Grune & Stratton, Inc. Metabolism,
Vol. 25, No. 1 (January),
1976
85
86
COHEN
ET Al.
body composition including x-ray densitometry, whole-body counting, underwater weighing, gross displacement, and chemical analysis of biopsy samples. For the elements of the body that exchange readily, radioactive tracers have, of course, been the basis of one of the most useful techniques for measuring body composition. Unfortunately, elements such as calcium exchange with tracers to only a very minor extent. It was for the analysis of total-body calcium that total body neutron activation (TBNAA) was first developed and applied successfully in medical research.14 Since the technique is noninvasive and affords a minimal radiation dose (0.28 rem) to the subject, it is a suitable tool for the study of body composition in normal individuals. In a previous study, the levels of total-body calcium were measured by neutron activation analysis in a group of normal women.s It was the object of the present study to extend these observations by measuring total-body calcium in a population of normal males. In addition, total-body phosphorus levels were measured in both populations to determine the composition of the skeletal system. The changes in levels of calcium and phosphorus with age in adult men and women are also compared with the corresponding changes in their cellular mass (i.e., total-body potassium). The interrelationship of the body composition of these elements, as well as their relationship to anthropometric parameters of growth (height and weight) were also investigated. In a cross-sectional study such as this, in which individuals exhibit a wide range of body size and habitus, it is necessary to normalize the data so that only the effects of aging are represented. The absolute levels of calcium, phosphorus, and potassium in the total body are thus expressed in ratio to the calculated “normal” value for an individual of matching sex, age, and body size. The measurement of total-body constituents such as calcium and phosphorus are of enhanced clinical value if the “normal” value for the particular subject studied is known for comparison. MATERIALS
AND
METHODS
Forty white females and 39 males ages 30-90 were analyzed by total-body neutron activation analysis (TBNAA). All the subjects were active and in good health; none were institutionalized. None of the subjects had a previous history or current symptoms of metabolic, renal, or cardiovascular disease. None of the postmenopausal women were receiving estrogen replacement therapy. No patients were receiving medications known to influence calcium, phosphorus, or potassium metabolism. Total-body neutron activation analyses were performed with neutron exposure provided by an array of fourteen 50-Ci encapsulated 23sPu,Be neutron sources.4 In this technique, the subjects are uniformly exposed to a beam of partially moderated fast neutrons which induce the reaction 48Ca(n,rf49Ca, and “P(n,cr) **Al. The absolute levels of the induced 49Ca and **AI (from P) are then counted in a whole-body counter. From these data, absolute levels of calcium and phosphorus are calculated. The in vivo activation technique provides an accuracy and precision of + 1.0% for total body calcium and f 3.9% for total body phosphorus as measured in an anthropomorphic phantom.4 The Brookhaven whole-body counter, with its on-line computer facility, was also used to quantitate the absolute levels of total-body potassium (TBK) by the measurement of 40K. The accuracy and precision for measuring TBK is f 3.3% in an anthropomorphic phantom.6 In order to intercompare the TBCa values in a heterogeneous population, such as this, it is necessary to normalize the data for skeletal size. The TBCa values were normalized by dividing the absolute calcium value by a calcium level computed for each individual, matched to a set of
BODY
CHEMICAL
Tablo
wt. Age
NO.
Females 30-39
6
(33.5) 40-49
11
W.1) 50-59
10
(54.1) 60-69
7
(63.4) 70-79 (72.5) Males 30-39 (34.7) 40-49 (45.7)
9 6
1.
Body
lit.
Kg.
cm
Composition TBK g 76.6
-
TBK 'b
55.6
156.6
zt10.4'
zt1.7
zt12.0
+9.a
63.2 ~1~12.9 69.8
161.6 13.6 159.1
88.1 *11.8 81.3
1.027 zt8.5 0.959
jz18.6 63.1
zt5.9 153.4
~1~12.9 75.3
h14.9 6
a7
COMPOSITION
0.987
in Adult TBCa g 785 +4.5
And Women -
TBCa COP 1.014
TBP (e) 402
-
TBP TBCa
0.51
TBCo TBK
10.3
klO.5
849 +8.7 804
0.991 zt6.5 0.996
396 +10.8 360
0.47 zt8.5 0.45
9.70 +7.1 9.92
zk9.3 1.029
l 13.2
zk5.7 0.985
h18.0 346
zt15.8 0.49
*a.5 9.30
+7.8
+15.7
*11.8
k7.9
701
h5.2
+7.4
zt9.6
h12.2
156.4 k2.7
67.8 zt7.8
0.957 zt6.2
634 *15.9
78.0
173.9
1.046
1110
0.999 ztl1.4
0.986
326 *17.3
510
zk13.1
-
zt5.7
57.9 iZ8.0
141.2
Men
0.49 h8.6
0.47
zk16.3
k4.9
~18.4
zt8.1
+15.3
h6.3
h16.9
86.0 zt12.7
177.7 a2.1
142.2 +11.9
0.994 zk5.4
1077 zt13.2
0.933 ~8.7
l 17.3
0.44 zt8.4
473
zt18.6
zk10.7
9.64 h13.4
7.92 zk7.7 7.60
l lO.l
50-59 (54.0)
6
75.6 h-7.5
168.1 a5.4
121.3 Lt15.9
1.031 zt7.2
996 zt15.4
0.986 zt7.5
420 +19.3
0.42 zk7.1
8.24 Lt9.7
60-69 (63.2) 70-79
6
76.1 h12.2 70.0
173.9 a2.7 169.5
128.3 +13.1 105.4
1.048 i9.0 0.975
1055 +11.5 1075
0.989 k 10.9 1.109
435 Lt17.5 479
0.41 Lt7.1 0.45
8.29 +12.6 10.34
h11.7 70.8 ztl6.5
a5.4 161.5 h4.7
zk19.8 93.7 zk16.5
+11.7 0.973 h11.3
l 10.8 1006 h19.7
+6.6 1.158 h10.2
+12.1 439 &lo.6
zk10.8 0.45 ~~16.4
zt8.7 10.75 jzl9.8
(74.5) 80-89 (84.5)
6 6
l CoetTtcient of variation (%SD). TBK-total body potossium,measured. KP-total body potassium, predicted. TBCo-totalbodycalcium,meosured. Co,-total body calcium, predicted. TBP-totolbody phosphorus,measured.
critical parameters. The calculated value is called the “normal” value for the individual. The parameters utilized in the computation of the normal values are sex, age, lean body mass (K), and skeletal size. The principle of the derivation of the following empirical relationship has been previously described.5*7 The equation for predicting normal calcium value (Cat,) based on the 79 normal individuals is presented in Appendix A. The measured total-body calcium (TBCa) divided by the predicted or normal calcium (Ca,) is referred to as the calcium ratio, denoted TBCa/Cae. In the female population after the age of - 55, there was a marked decrease in the calcium level. Thus it was necessary to apply an age correction factor to the coefficient (Y for females older than 55 yr.5 The age correction factor was determined empirically. The coefficients for the characteristic equations were determined by standard curve fitting to the data. The deficiency in terms of looo/, of the model is the age correction factor. In the derivation of the prediction equation for males, data on men over 80 yr were not included because of their large variability. Deviation of the calcium ratio from the value of I .OO in a normal individual reflects measurement errors and statistical variation. One can arbitrarily define an abnormal TBCa value for an individual as one which produced a ratio differing from 1.00 by > f 0.13 (90% confidence level, I .62 SD). As with the calcium data, the total body potassium was also normalized for body size. The mean potassium ratio (measured TBK divided by a predicted normal K) is presented in Table I. The em-
aa
COHEN
pirical relationship and sex” is presented
employed to in Appendix
predict A.
the
normal
potassium
(K,)
as a function
of
ET Al.
age,
size
RESULTS
The mean values for the total-body calcium (TBCa), phosphorus (TBP), and potassium (TBK) in male and female adults as a function of age are tabulated in Table 1. The variability in the values in each decade group for the individual elements is large (Fig. 1). The curves shown in Fig. 1 were developed from the mean of TBCa levels for males and females for each decade. The normalized calcium data, corrected for body size (i.e., the calcium ratio TBCa/Ca,) for each individual are presented in Fig. 2. The mean Ca ratio for each age group is approximately 1.00 (see Table l), but in females the variability about the mean increased from 5.7% to 11.4% markedly with increasing age. In the male population, the variability in the ratio ranged from 6.3% to 10.9x, and also tended to increase with age. In males above the age of 70 yr, the value of the calcium ratio increased significantly. The mean Ca ratio for the 8th decade was 1.11 f 6.6x, and for the 9th decade 1.16 f 10.2%. In the group of males 67-90 yr of age, six of the 13 subjects had ratios greater than 1.13 (90% confidence level). The mean total body potassium values (which reflect the muscle mass) for each age decade are also presented in Table 1. The mean TBK values show a decrease with age as previously reported.6 The mean K ratio for all the males was 1.011 + 8.7% (SD) and for the females 0.991 f 8.7% (SD). Only three males and one female had normalized TBK values > 2 SD (‘t 16%) from the mean. The variability of the mean values for each age group was rather constant with age except for a slight increase in males > 70 yr. In addition, the relationship between skeletal mass (TBCa) and muscle mass (TBK) is expressed in terms of the mean TBCa/TBK ratio for each decade in
1400 1300 I
5 1000 4 s 900
I
700 600 500
i I 30
1 40
I 50
. I 60 70 AGE (year)
’ I 00
1 90
100
Total-body calcium as a func1. of age in adult males and fomahs.
Fig.
tion
BODY
CHEMICAL
Rg. 2. The calcium sured calcium/predicted function of age.
4
a9
COMPOSITION
mtio calcium)
(meaas a
30
40
50
60
70
80
90
AGE ( year)
Table 1. The mean ratio for women was 9.8; for men (30-69 yr) it was 8.0. The change in the individual ratios with time is displayed in Fig. 3. There was a significant increase in this ratio for males past 70 years of age, to a mean of 10.5. The absolute values for total-body phosphorus (TBP) as a function of age and sex show a large variability (Table 1). The TBP values decrease with age in both males and females similar to the pattern observed with TBCa. The individual phosphorus values are also expressed in terms of their TBCa values. The change of the individual TBP/TBCa ratio for males of each decade are con-
30
40
50
60 AGE (yeor
70 I
80
90
Fig. 3. Total-body as a function of age.
calcium/potassium
90
COHEN
ET AL.
sistently lower than for females over the entire age span and both male and female ratios tend to show a slight decline at 65 and 55 yr, respectively. DISCUSSION
Total Body Calcium ( TBCa) The variation in the TBCa values in both males and females in any age decade is very large, ranging up to 19.7% (SD). The absolute amount of calcium reflects both the size of the individual and the degree of mineralization. Clearly, an “average” value based on sex and age alone is inadequate as a reference value because of the strong dependance of TBCa on the size of the individual. The mean value for each age decade is plotted in Fig. 1. It can be seen that these values do not provide a well defined reference standard against which an individual’s measurement can be assessed. For each age decade, the mean Ca values of females are 20%-40% lower than the males in absolute terms. However, when the calcium is expressed in terms of the calcium ratio (TBCa/Ca,) there is no difference between the sexes. That is, when the Ca is “normalized” for the size and habitus of the subject, the apparent difference based on sex disappears. The importance of the normalization of the data for the various body constituents lies in the fact that for the absolute measurement to be of clinical value, the normal value for each individual must be known to provide a reference against which to evaluate the level of calcium, phosphorus, or potassium in that individual. Body weight, body surface area, and muscle mass have been widely used as reference parameters for normalizing levels of various elements. The first two may be valid in the case of normal individuals over a wide range of body build, but neither alone has a satisfactory theoretical basis. Muscle mass, as measured by TBK, would seem to be a useful parameter of body size. Indeed, the muscle mass and height appear to be the simplest valid parameters for normalization of the data, even over the wide range of body build studied here. The equations for predicting TBCa, involving height and TBK, gave a very high correlation between the measured and the predicted calcium. A single equation satisfactorily predicted the TBCa for the male population. In the male population from 30-79 yr, the average calcium ratio was 1.000 f 7.8%. Male subjects 80-90 were excluded as there appeared to be a significant increase in TBCa in these elderly male subjects. For example, although the sample is small, 11 out of 17 of the men, 67 yr of age and older, had calcium ratios > 1.00. Six men had Ca ratios > 13% above their predicted value (ages 67, 70, 75, 82, 83, and 90). The extremely high value observed in the 90 yr old male (1.36) corresponded to an unusually high total body calcium of 1302 g. There is no obvious explanation for this very high total body calcium level in this individual or in the other elderly subjects. One possibility is that these high TBCa levels are the result of a selection process or they may represent increased extraosseous or periarticular deposits of calcium associated with old age. It was found that an explicit correction for age (in addition to the correction implicit in the height and potassium values) improved the correlation between predicted and measured calcium for females. The mean calcium ratio for the 40 women was 0.996 f 7.1% (SD). Of the 40 women, one had a Ca ratio below the
BODY
CHEMICAL
91
COMPOSITION
normal range and only one woman (age 72) had a calcium ratio greater than 1.13. Thus, with the present mathematical models, the TBCa for a normal man or woman can be predicted to within + 13% at the 90% confidence level. Total Body Phosphorus (TBP)
Levels of TBP were found to decrease with age. The variability in the data tended to be greater than was observed with TBCa data. It is more illuminating to express the phosphorus level in terms of calcium since they are stoichiometrically related in the skeleton. The ratio of P/Ca averaged 0.48 in the males over the entire life-span and 0.50 in the females (Fig. 4). For comparison, the P/Ca ratio in four cadavers analyzed chemically was reported to be 0.54 f .08.‘* Further, it is not clear why the male P/Ca ratio is slightly, but consistently, lower than that of the female ratio. It was observed that the mean P/Ca ratio fell slightly to a minimum value at 65 and 55 yr in males and females, respectively (Fig. 4); there is no obvious explanation for this phenomenon. The relative constancy of the mean P/Ca ratio may derive from the fact that with the loss of bone, both the TBCa and TBP fall proportionately. The variability of this ratio, however, is as large as 18.6%. This may be related to the fact that while almost all of the Ca (99%) is located in the skeletal system, only 80x-90% of the phosphorus is located in the skeleton. The nonskeletal phosphorus in the soft tissues and body fluids is involved in other metabolic processes, thus it is not unreasonable that there is a higher variability for TBP. Conversely, the finding that the ratio of P/Ca does not change significantly with age would suggest that the composition of bone which contains the major fraction of the TBCa and TBP in the apatite crystal, does not change significantly with age. Total-Body Potassium (TBK)
The mean TBK values decreased with age in males and females. The variability in each age decade remains relatively constant until the age of 80. Because K is the dominant intracellular cation found primarily in muscle and I
I
I
I
I
F 2
I -* ---0
0
I
FEMALE MALE
0.70
zz
i
I
0 0 .
0.30
’ 30
Fig. 4.
I 40
Total-body
I 50
I 60 AGE
phosphorus/calcium
I 70
I 80
I 90
(year) as a function
age.
100
92
COHEN
ET AL.
viscera, it has been used as an index of body cell mass. The variability of this intracellular component thus appears not to change with age in normal subjects. In absolute terms, females are 33%-46x lower than the males for each age decade. However, when the K is expressed as TBK/K, (i.e., normalized for body size) the mean potassium ratios are not different between the sexes. As previously established, the predicted potassium value (K,) correlates well with the measured TBK in normal subjects.6 If the ratio of Ca/K is considered, the average value for females is about 20% higher than that for males. That is females have - 20% less muscle mass per unit skeletal mass than males. Further, the value of the ratio does not change significantly with age except for males over 70 yr. This indicates that the loss of muscle mass is directly proportional to the loss of TBCa with age for most of the population. This relationship has previously been observed in a number of metabolic disorders.8 The significantly increased Ca/K ratio in 12 of the 13 elderly males closely approximates that of the normal females. A possible explanation for this increased ratio, is an increase soft tissue calcification, frequently observed in elderly persons. Rate of Loss of Calcium and Phosphorus From the Skeleton Almost all the studies on the loss of calcium from the skeleton have been on localized areas of the skeleton, and further they have been cross-sectional rather than longitudinal studies. An excellent review of the various techniques used for measuring bone loss was presented by Newton-John.9 The results of these studies, which concentrate on small areas of the skeleton (particularly measurements employing two-dimensional radiographs) cannot readily be extrapolated to whole bones, and are not always applicable to the whole skeleton. The present cross-sectional study has attempted to circumvent these difficulties in two ways: first, an examination of the loss of calcium and phosphorus from the entire skeleton, and secondly, by normalization of the data for the large variability in size, characteristic of a heterogeneous population of any age. It is clear that with this large variability in absolute levels of TBCa and TBP in each age group, the use of the average value for each decade can lead to quite erroneous conclusions. With a sufficiently large population, the error is, of course, minimized. However, once the individual data are normalized for the effects of skeletal size and sex, it is possible to estimate the rate of loss of skeletal mass with age in a cross-sectional study. Estimates of the time of onset of bone loss vary widely from 20 to 66 yr, depending on the criteria used and the bone selected?-” For example, studies of bone loss from radial and metacarpal cortex indicate no significant loss in women before 50 yr and before 60 yr in men. However, the present study suggests that the overall skeletal calcium loss commences within the fourth decade for women and the fifth decade for men. In the present study, the onset of the decline of skeletal bone mass was arbitrarily taken to be 35 yr for females and 45 yr for males. The change in skeletal mass after maturity (starting in the fourth and fifth decades, respectively) appears to be linear with time, at least through the ninth decade. The marked increase in rate of loss of Ca that occurs
BODY
CHEMICAL
93
COMPOSITION
after 55 yr in women, presumably associated with menopause, is superimposed on the slower rate of loss which starts at - 35 yr.5 In the male population, the decrease of TBCa starts at - 45 yr and continues over the life span. The rate of loss of TBCa is approximated with a linear function over the entire age span. The original equation for predicting calcium in males presented above is satisfactory without an additional age correction at least until 79 yr of age. The overall rate of change of calcium with time for the population of men and women in this study was determined in the following manner. Since height and body potassium decline with age, their effect on calcium loss, implicit in the equation for calcium prediction, must be considered along with the explicit age correction factor. Thus, the derivative of the calcium value with time takes into account the dependency of calcium on the three time-dependent variables (see Appendix B). The rate of decrease of muscle mass (as determined by measurement of @‘K) with time was previously measured in a large population: 250 females and 250 males, 30-79 yr of age.6 The mean rate of decrease of potassium with time for the females was found to be 0.45 g/yr and for the males 1.16 g/yr after 50 yr.6 The mean rate of loss of height with time was also determined for the same populations. The value for this loss was found to be 0.15 cm/yr for females and 0.28 cm/yr for males. Most observations on loss of height with age have been cross sectional on selected populations or abnormal populations. In Adams’ study (the only longitudinal study) the height loss was twice as great in women as men, but it is to be noted that the age group studied was between 55 and 64 yr.” Adams incidentally reported no change in height in 48% of the men and in 15% of the women, emphasizing great variability in the end response. In the present study, in order to analyze “normal” losses, any women who had any suggestion of osteoporosis were excluded. This exclusion may explain the small or average loss of height per year in women noted in the present study. The combination of the changes of height and potassium with age along with the age correction factor, (Y yields a range of values for the loss of calcium in females from 3 g/yr (for ages 35-54) to 7.01 g/yr (for ages 70-79) (see Table 2). For males the loss of calcium ranges from 6.2 g/yr at - 50 yr to 7.98 g/yr at 80-90 yr (Table 2). For the “average” woman, this amounts to a mean rate of loss of calcium of 0.38% per year for ages 35-54, and 1.08% per year over the Table
2.
Rate of loss of total-body
calcium
with
Females
age Males
AW
elv
%JY
30-39
3.00
0.39
-
-
40-49
2.90
0.37
-
-
SO-54 55-59
2.96 8.20
0.36 1.06
6.44
0.65
60-69 70-79
7.40 7.01
1.07 1.11
6.50 7.81
0.62 0.73
So-90
-
-
7.98
0.79
elv
WY
94
COHEN
ET Al.
age span 55-79. For the average man the mean loss of calcium was 0.70% per year after 50 yr. While no directly comparable data are available for comparison of total skeletal calcium loss, there are two useful longitudinal studies reported in the literature on the rate of loss of bone mass for the appendicular skeleton. Garn” measured the change in cortical thickness of the metacarpals of adult women over a 23-yr period, and reported a loss of thickness of 0.25 mm per decade. When this value is corrected for subperiosteal bone formation it yields a value of 0.75% per year. This value is somewhat lower than that observed in the present study for the female population. The corresponding rate of loss of cortical thickness in males was 0.45% per year, a value considerably lower than the results reported here. Adams” also carried out a longitudinal study on the effects of aging in white females over an 1 I-yr period, using the same end point, and reported a loss of 1.34% per year. The corresponding loss of cortical bone width for white males in the same study was 0.86% per year, slightly higher than observed here. From the present study it appears that for the average female, the average loss of bone calcium amounts to 3.8 g/yr, over the period 30-54 yr of age, and 7.6 g/yr after 55 yr of age. The average loss in males is 7 g/yr over the period 50-90 yr. Thus, for women, a total of 250 g of Ca are lost normally over the period of 30-80 yr; for men 210 g of calcium are lost over this time. This amounts to - 28% of the total body calcium in an average woman at 30 years of age, and - 20% of the total calcium in men. While the mechanism of bone loss with age is not as yet fully understood, it must ultimately result from an imbalance in the bone formation and bone resorption rates. The interesting point is that the mean loss of Ca in women of 0.37% and 1.1% per year, before and after menopause (50-60 yr), is equivalent to an average negative calcium balance of only 8 and 21 mg Ca per day, respectively. In males the negative calcium balance from 50-90 yr amounts to only - 19 mg per day. It is to be noted that these levels of Ca loss are within the experimental error of Ca balance techniques. The universal regression of calcium and phosphorus levels with age is similar to that of gonadal function. The accelerated rate of loss of total body calcium and phosphorus in women that commences at menopause may be associated with a progressively marked imbalance in anabolic and catabolic steroids that also occurs at 50-60 yr. It has been hypothesized that gonadal hormones normally retard bone resorption; thus their decline with time could result in an increased bone resorption rate. ACKNOWLEDGMENT The authors are indebted to Mr. M. Stravino and Mr. J. Rothmann for their ful technical assistance in performing the analyses in the previous study.
dedicated
and
skill-
REFERENCES 1. Chamberlain MJ, Fremlin JH, Peter DK, Philip H: Total body calcium by whole-body neutron activation: New technique for study of bone disease. Br Med J 2158 I, 1968 2. Palmer HE, Nelp WE, Murano R, et al:
The feasibility of in vivo neutron activation analysis of total body calcium and other elements of body composition. Phys Med Biol 13: 269, 1968 3. Cohn SH, Dombrowski CS, Fairchild RG:
BODY
CHEMICAL
In vivo activation Int J Appl Radiat
95
COMPOSITION
analysis of calcium Isot 21:127, 1970
in man.
4. Cohn SH, Shukla KK, Fairchild RG: Design and calibration of a broad-beam 238Pu,Be source for total body neutron activation analysis. J Nucl Med 13:487-492, 1972 5. Cohn SH, Vaswani A, Zanzi I, Ellis KJ: The effect of aging on bone mass in adult women. Am J Physiol (in press) 6. Ellis KJ, Shukla KK, Cohn SH: A predictor for total-body potassium based on height, weight, sex and age: Application in metabolic disorders. J Lab Clin Med 83:716727, 1974 7. Cohn SH, Ellis KJ, Wallach S, Zanzi I, Atkins HL, Aloia JF: Absolute and relative deficit in total skeletal calcium and radial bone
mineral content in osteoporosis. J Nucl Med 15:428-435, 1974 8. Ellis KJ, Cohn SH: The correlation between skeletal calcium mass and muscle mass in man. J Appl Physiol38:455460, 1975 9. Newton-John HR. Brian Morgan D: The loss of bone with age, osteoporosis and fractures. Clin Grtho 71:229-252, 1970 IO. Garn SM, Rohmann CG, Nolan P: Relations of development and aging, Ed. J.E. Birren. Springfield, Illinois: Thomas, 1964, p 43 1 I. Adams P, Davies GT, Sweetname P: Osteoporosis and the effects of aging on bone mass in elderly men and women. Q J Med 39: 60-615, 1970 12. Widdowson EM, McCance RA, Spray CM: The chemical composition of the human body. Clin Sci IO:1 13-124, 1950
96
COHEN
APPENDIX
A.
Equation
for
(Co,) Cap
Predicting
in Normal
the
Total
Body
ET AL.
Calcium
Subjects
112
= a HK
where H = height(m) K = body
potassium
(g)
a = 56.62 for female
subiects
5 55
a = 56.62 - 0.38 (A - 55) for females a = 54.5 for males Equation
for predicting
K, = a W”’
the total
body
> 55 yr
potassium
(K,)
in the normal
subjects.
Hz
where:
a = 5.52 - 0.014
A
males
a = 4.58
A
females
- 0.010
A = we (Y) W = weight (Kg) H = height(m)
APPENDIX
B.
Derivation
of Equation
Co,=
Rate
of Change
of Calcium
VlH
aK
!!!!$ = K’lz,.,g+yK ca aa
ac0
for
’
-‘/?a,,2+aK1/z?
at 1 co
aK
at
Ca aH
T’X-B;-+TKat+Kat For Women
For Men
a = 56.62 -
= a,
at
g
*Ref.
= -0.38 aK -= at
-0.15’
afi -= at
6. = predicted
K = total
body
total
body
potassium
calcium
(g).
(g).
H = height(m). a = coefficient
a = 54.5
= -0.45’
aH -= at
Co,
- 0.38 (A-55)
with age correction
factor.
-1.16’
- 0.28*
With
Time
A. Vaswani,
I. Zanri,
J. F. Aloia,
Total-body levels of calcium and phosphorus (reflecting skeletal mass) and total-body levels of potassium (reflecting muscle mass) were measured by neutron activation analysis in 39 men and 40 women ages 30-90 yr. In order to intercompare the total body calcium (TBCa) values in a heterogeneous population, such as this, it was necessary to normalize the data for skeletal size. The normalization consisted of dividing the absolute calcium level by the predicted calcium level for each individual matched to a set of critical parameters. The parameter used in the computation of normal values were age, sex, muscle mass, i.e., total body potassium (TBK) and height. For the calcium data of the women, it was necessary to add an age correction factor after the age of 55 yr. The calcium ratio (mean mtio of the predicted to measured TBCa) in men was 1.000 f 7.8% and in women 0.996 f 7.1%. The TBCa of normal males and females can thus be predicted to =t 13% (at the 90% confidence level). An exception to this was found in males (70-90 yr) who exhibited a mean calcium ratio >1.13.
M. S. Roginsky,
and
K. J. Ellis
The derivative of TBCa with time was determined for this population of men and women by taking into account the dependency of calcium on three time dependent variables, height, TBK, and an explicit age correction factor in the case of the women. The mean mte of loss of TBCa in women was 0.37% and 1 .l % per year before and after menopause (50 yr). In the males, the avemge mte of loss of TBCa was 0.7% per year after 50 yr of age. The pattern of total body phosphorus (TBP) loss with age paralleled that of TBCa as the mtio of TBP/TBCa was mther constant with age. The constancy of the mtio suggests that the mineml composition of bone does not change significantly with age. The mte of loss of TBK with age was also related directly to that of TBCa. The mean mtio of TBK/TBCa was 9.9 in females and 8.0 in males and this ratio remained relatively constant from 30-70 yr. Thus, the mechanism responsible for the loss of bone with age, whether nutritional deficiency or decreased gonadal function and physical activity may also be responsible for the loss of muscle mass with age.
T
HE QUANTITATIVE ASSESSMENT of the chemical composition of the human body as a function of age is of basic interest as well as of clinical value in diagnosing metabolic disorders. Data on constituents such as total body calcium also serve as objective criteria in the evaluation of the efficacy of therapeutic programs for such metabolic disorders as osteoporosis and renal osteodystrophy. The changes in chemical composition of the body with age can be measured for the first time, in vivo, by the technique of whole-body neutron activation (TBNAA). In the past, various in vivo methods have been used for the study of These measurements were performed in accordance with the principles of the Declaration of Helsinki and the regulations of the U.S. Department of Health, Education and Welfare. An informed consent was obtained for each subject. From the Medical Research Center, Brookhaven National Laboratory, Upton, N. Y. Received for publication May 5. 1975. Supported by the U.S. Energy Research and Development Administration. Reprint requests should be addressed to Dr. S. H. Cohn, Medical Research Center. Brookhaven National Laboratory, Upton, N. Y. 11973. 0 1976 by Grune & Stratton, Inc. Metabolism,
Vol. 25, No. 1 (January),
1976
85
86
COHEN
ET Al.
body composition including x-ray densitometry, whole-body counting, underwater weighing, gross displacement, and chemical analysis of biopsy samples. For the elements of the body that exchange readily, radioactive tracers have, of course, been the basis of one of the most useful techniques for measuring body composition. Unfortunately, elements such as calcium exchange with tracers to only a very minor extent. It was for the analysis of total-body calcium that total body neutron activation (TBNAA) was first developed and applied successfully in medical research.14 Since the technique is noninvasive and affords a minimal radiation dose (0.28 rem) to the subject, it is a suitable tool for the study of body composition in normal individuals. In a previous study, the levels of total-body calcium were measured by neutron activation analysis in a group of normal women.s It was the object of the present study to extend these observations by measuring total-body calcium in a population of normal males. In addition, total-body phosphorus levels were measured in both populations to determine the composition of the skeletal system. The changes in levels of calcium and phosphorus with age in adult men and women are also compared with the corresponding changes in their cellular mass (i.e., total-body potassium). The interrelationship of the body composition of these elements, as well as their relationship to anthropometric parameters of growth (height and weight) were also investigated. In a cross-sectional study such as this, in which individuals exhibit a wide range of body size and habitus, it is necessary to normalize the data so that only the effects of aging are represented. The absolute levels of calcium, phosphorus, and potassium in the total body are thus expressed in ratio to the calculated “normal” value for an individual of matching sex, age, and body size. The measurement of total-body constituents such as calcium and phosphorus are of enhanced clinical value if the “normal” value for the particular subject studied is known for comparison. MATERIALS
AND
METHODS
Forty white females and 39 males ages 30-90 were analyzed by total-body neutron activation analysis (TBNAA). All the subjects were active and in good health; none were institutionalized. None of the subjects had a previous history or current symptoms of metabolic, renal, or cardiovascular disease. None of the postmenopausal women were receiving estrogen replacement therapy. No patients were receiving medications known to influence calcium, phosphorus, or potassium metabolism. Total-body neutron activation analyses were performed with neutron exposure provided by an array of fourteen 50-Ci encapsulated 23sPu,Be neutron sources.4 In this technique, the subjects are uniformly exposed to a beam of partially moderated fast neutrons which induce the reaction 48Ca(n,rf49Ca, and “P(n,cr) **Al. The absolute levels of the induced 49Ca and **AI (from P) are then counted in a whole-body counter. From these data, absolute levels of calcium and phosphorus are calculated. The in vivo activation technique provides an accuracy and precision of + 1.0% for total body calcium and f 3.9% for total body phosphorus as measured in an anthropomorphic phantom.4 The Brookhaven whole-body counter, with its on-line computer facility, was also used to quantitate the absolute levels of total-body potassium (TBK) by the measurement of 40K. The accuracy and precision for measuring TBK is f 3.3% in an anthropomorphic phantom.6 In order to intercompare the TBCa values in a heterogeneous population, such as this, it is necessary to normalize the data for skeletal size. The TBCa values were normalized by dividing the absolute calcium value by a calcium level computed for each individual, matched to a set of
BODY
CHEMICAL
Tablo
wt. Age
NO.
Females 30-39
6
(33.5) 40-49
11
W.1) 50-59
10
(54.1) 60-69
7
(63.4) 70-79 (72.5) Males 30-39 (34.7) 40-49 (45.7)
9 6
1.
Body
lit.
Kg.
cm
Composition TBK g 76.6
-
TBK 'b
55.6
156.6
zt10.4'
zt1.7
zt12.0
+9.a
63.2 ~1~12.9 69.8
161.6 13.6 159.1
88.1 *11.8 81.3
1.027 zt8.5 0.959
jz18.6 63.1
zt5.9 153.4
~1~12.9 75.3
h14.9 6
a7
COMPOSITION
0.987
in Adult TBCa g 785 +4.5
And Women -
TBCa COP 1.014
TBP (e) 402
-
TBP TBCa
0.51
TBCo TBK
10.3
klO.5
849 +8.7 804
0.991 zt6.5 0.996
396 +10.8 360
0.47 zt8.5 0.45
9.70 +7.1 9.92
zk9.3 1.029
l 13.2
zk5.7 0.985
h18.0 346
zt15.8 0.49
*a.5 9.30
+7.8
+15.7
*11.8
k7.9
701
h5.2
+7.4
zt9.6
h12.2
156.4 k2.7
67.8 zt7.8
0.957 zt6.2
634 *15.9
78.0
173.9
1.046
1110
0.999 ztl1.4
0.986
326 *17.3
510
zk13.1
-
zt5.7
57.9 iZ8.0
141.2
Men
0.49 h8.6
0.47
zk16.3
k4.9
~18.4
zt8.1
+15.3
h6.3
h16.9
86.0 zt12.7
177.7 a2.1
142.2 +11.9
0.994 zk5.4
1077 zt13.2
0.933 ~8.7
l 17.3
0.44 zt8.4
473
zt18.6
zk10.7
9.64 h13.4
7.92 zk7.7 7.60
l lO.l
50-59 (54.0)
6
75.6 h-7.5
168.1 a5.4
121.3 Lt15.9
1.031 zt7.2
996 zt15.4
0.986 zt7.5
420 +19.3
0.42 zk7.1
8.24 Lt9.7
60-69 (63.2) 70-79
6
76.1 h12.2 70.0
173.9 a2.7 169.5
128.3 +13.1 105.4
1.048 i9.0 0.975
1055 +11.5 1075
0.989 k 10.9 1.109
435 Lt17.5 479
0.41 Lt7.1 0.45
8.29 +12.6 10.34
h11.7 70.8 ztl6.5
a5.4 161.5 h4.7
zk19.8 93.7 zk16.5
+11.7 0.973 h11.3
l 10.8 1006 h19.7
+6.6 1.158 h10.2
+12.1 439 &lo.6
zk10.8 0.45 ~~16.4
zt8.7 10.75 jzl9.8
(74.5) 80-89 (84.5)
6 6
l CoetTtcient of variation (%SD). TBK-total body potossium,measured. KP-total body potassium, predicted. TBCo-totalbodycalcium,meosured. Co,-total body calcium, predicted. TBP-totolbody phosphorus,measured.
critical parameters. The calculated value is called the “normal” value for the individual. The parameters utilized in the computation of the normal values are sex, age, lean body mass (K), and skeletal size. The principle of the derivation of the following empirical relationship has been previously described.5*7 The equation for predicting normal calcium value (Cat,) based on the 79 normal individuals is presented in Appendix A. The measured total-body calcium (TBCa) divided by the predicted or normal calcium (Ca,) is referred to as the calcium ratio, denoted TBCa/Cae. In the female population after the age of - 55, there was a marked decrease in the calcium level. Thus it was necessary to apply an age correction factor to the coefficient (Y for females older than 55 yr.5 The age correction factor was determined empirically. The coefficients for the characteristic equations were determined by standard curve fitting to the data. The deficiency in terms of looo/, of the model is the age correction factor. In the derivation of the prediction equation for males, data on men over 80 yr were not included because of their large variability. Deviation of the calcium ratio from the value of I .OO in a normal individual reflects measurement errors and statistical variation. One can arbitrarily define an abnormal TBCa value for an individual as one which produced a ratio differing from 1.00 by > f 0.13 (90% confidence level, I .62 SD). As with the calcium data, the total body potassium was also normalized for body size. The mean potassium ratio (measured TBK divided by a predicted normal K) is presented in Table I. The em-
aa
COHEN
pirical relationship and sex” is presented
employed to in Appendix
predict A.
the
normal
potassium
(K,)
as a function
of
ET Al.
age,
size
RESULTS
The mean values for the total-body calcium (TBCa), phosphorus (TBP), and potassium (TBK) in male and female adults as a function of age are tabulated in Table 1. The variability in the values in each decade group for the individual elements is large (Fig. 1). The curves shown in Fig. 1 were developed from the mean of TBCa levels for males and females for each decade. The normalized calcium data, corrected for body size (i.e., the calcium ratio TBCa/Ca,) for each individual are presented in Fig. 2. The mean Ca ratio for each age group is approximately 1.00 (see Table l), but in females the variability about the mean increased from 5.7% to 11.4% markedly with increasing age. In the male population, the variability in the ratio ranged from 6.3% to 10.9x, and also tended to increase with age. In males above the age of 70 yr, the value of the calcium ratio increased significantly. The mean Ca ratio for the 8th decade was 1.11 f 6.6x, and for the 9th decade 1.16 f 10.2%. In the group of males 67-90 yr of age, six of the 13 subjects had ratios greater than 1.13 (90% confidence level). The mean total body potassium values (which reflect the muscle mass) for each age decade are also presented in Table 1. The mean TBK values show a decrease with age as previously reported.6 The mean K ratio for all the males was 1.011 + 8.7% (SD) and for the females 0.991 f 8.7% (SD). Only three males and one female had normalized TBK values > 2 SD (‘t 16%) from the mean. The variability of the mean values for each age group was rather constant with age except for a slight increase in males > 70 yr. In addition, the relationship between skeletal mass (TBCa) and muscle mass (TBK) is expressed in terms of the mean TBCa/TBK ratio for each decade in
1400 1300 I
5 1000 4 s 900
I
700 600 500
i I 30
1 40
I 50
. I 60 70 AGE (year)
’ I 00
1 90
100
Total-body calcium as a func1. of age in adult males and fomahs.
Fig.
tion
BODY
CHEMICAL
Rg. 2. The calcium sured calcium/predicted function of age.
4
a9
COMPOSITION
mtio calcium)
(meaas a
30
40
50
60
70
80
90
AGE ( year)
Table 1. The mean ratio for women was 9.8; for men (30-69 yr) it was 8.0. The change in the individual ratios with time is displayed in Fig. 3. There was a significant increase in this ratio for males past 70 years of age, to a mean of 10.5. The absolute values for total-body phosphorus (TBP) as a function of age and sex show a large variability (Table 1). The TBP values decrease with age in both males and females similar to the pattern observed with TBCa. The individual phosphorus values are also expressed in terms of their TBCa values. The change of the individual TBP/TBCa ratio for males of each decade are con-
30
40
50
60 AGE (yeor
70 I
80
90
Fig. 3. Total-body as a function of age.
calcium/potassium
90
COHEN
ET AL.
sistently lower than for females over the entire age span and both male and female ratios tend to show a slight decline at 65 and 55 yr, respectively. DISCUSSION
Total Body Calcium ( TBCa) The variation in the TBCa values in both males and females in any age decade is very large, ranging up to 19.7% (SD). The absolute amount of calcium reflects both the size of the individual and the degree of mineralization. Clearly, an “average” value based on sex and age alone is inadequate as a reference value because of the strong dependance of TBCa on the size of the individual. The mean value for each age decade is plotted in Fig. 1. It can be seen that these values do not provide a well defined reference standard against which an individual’s measurement can be assessed. For each age decade, the mean Ca values of females are 20%-40% lower than the males in absolute terms. However, when the calcium is expressed in terms of the calcium ratio (TBCa/Ca,) there is no difference between the sexes. That is, when the Ca is “normalized” for the size and habitus of the subject, the apparent difference based on sex disappears. The importance of the normalization of the data for the various body constituents lies in the fact that for the absolute measurement to be of clinical value, the normal value for each individual must be known to provide a reference against which to evaluate the level of calcium, phosphorus, or potassium in that individual. Body weight, body surface area, and muscle mass have been widely used as reference parameters for normalizing levels of various elements. The first two may be valid in the case of normal individuals over a wide range of body build, but neither alone has a satisfactory theoretical basis. Muscle mass, as measured by TBK, would seem to be a useful parameter of body size. Indeed, the muscle mass and height appear to be the simplest valid parameters for normalization of the data, even over the wide range of body build studied here. The equations for predicting TBCa, involving height and TBK, gave a very high correlation between the measured and the predicted calcium. A single equation satisfactorily predicted the TBCa for the male population. In the male population from 30-79 yr, the average calcium ratio was 1.000 f 7.8%. Male subjects 80-90 were excluded as there appeared to be a significant increase in TBCa in these elderly male subjects. For example, although the sample is small, 11 out of 17 of the men, 67 yr of age and older, had calcium ratios > 1.00. Six men had Ca ratios > 13% above their predicted value (ages 67, 70, 75, 82, 83, and 90). The extremely high value observed in the 90 yr old male (1.36) corresponded to an unusually high total body calcium of 1302 g. There is no obvious explanation for this very high total body calcium level in this individual or in the other elderly subjects. One possibility is that these high TBCa levels are the result of a selection process or they may represent increased extraosseous or periarticular deposits of calcium associated with old age. It was found that an explicit correction for age (in addition to the correction implicit in the height and potassium values) improved the correlation between predicted and measured calcium for females. The mean calcium ratio for the 40 women was 0.996 f 7.1% (SD). Of the 40 women, one had a Ca ratio below the
BODY
CHEMICAL
91
COMPOSITION
normal range and only one woman (age 72) had a calcium ratio greater than 1.13. Thus, with the present mathematical models, the TBCa for a normal man or woman can be predicted to within + 13% at the 90% confidence level. Total Body Phosphorus (TBP)
Levels of TBP were found to decrease with age. The variability in the data tended to be greater than was observed with TBCa data. It is more illuminating to express the phosphorus level in terms of calcium since they are stoichiometrically related in the skeleton. The ratio of P/Ca averaged 0.48 in the males over the entire life-span and 0.50 in the females (Fig. 4). For comparison, the P/Ca ratio in four cadavers analyzed chemically was reported to be 0.54 f .08.‘* Further, it is not clear why the male P/Ca ratio is slightly, but consistently, lower than that of the female ratio. It was observed that the mean P/Ca ratio fell slightly to a minimum value at 65 and 55 yr in males and females, respectively (Fig. 4); there is no obvious explanation for this phenomenon. The relative constancy of the mean P/Ca ratio may derive from the fact that with the loss of bone, both the TBCa and TBP fall proportionately. The variability of this ratio, however, is as large as 18.6%. This may be related to the fact that while almost all of the Ca (99%) is located in the skeletal system, only 80x-90% of the phosphorus is located in the skeleton. The nonskeletal phosphorus in the soft tissues and body fluids is involved in other metabolic processes, thus it is not unreasonable that there is a higher variability for TBP. Conversely, the finding that the ratio of P/Ca does not change significantly with age would suggest that the composition of bone which contains the major fraction of the TBCa and TBP in the apatite crystal, does not change significantly with age. Total-Body Potassium (TBK)
The mean TBK values decreased with age in males and females. The variability in each age decade remains relatively constant until the age of 80. Because K is the dominant intracellular cation found primarily in muscle and I
I
I
I
I
F 2
I -* ---0
0
I
FEMALE MALE
0.70
zz
i
I
0 0 .
0.30
’ 30
Fig. 4.
I 40
Total-body
I 50
I 60 AGE
phosphorus/calcium
I 70
I 80
I 90
(year) as a function
age.
100
92
COHEN
ET AL.
viscera, it has been used as an index of body cell mass. The variability of this intracellular component thus appears not to change with age in normal subjects. In absolute terms, females are 33%-46x lower than the males for each age decade. However, when the K is expressed as TBK/K, (i.e., normalized for body size) the mean potassium ratios are not different between the sexes. As previously established, the predicted potassium value (K,) correlates well with the measured TBK in normal subjects.6 If the ratio of Ca/K is considered, the average value for females is about 20% higher than that for males. That is females have - 20% less muscle mass per unit skeletal mass than males. Further, the value of the ratio does not change significantly with age except for males over 70 yr. This indicates that the loss of muscle mass is directly proportional to the loss of TBCa with age for most of the population. This relationship has previously been observed in a number of metabolic disorders.8 The significantly increased Ca/K ratio in 12 of the 13 elderly males closely approximates that of the normal females. A possible explanation for this increased ratio, is an increase soft tissue calcification, frequently observed in elderly persons. Rate of Loss of Calcium and Phosphorus From the Skeleton Almost all the studies on the loss of calcium from the skeleton have been on localized areas of the skeleton, and further they have been cross-sectional rather than longitudinal studies. An excellent review of the various techniques used for measuring bone loss was presented by Newton-John.9 The results of these studies, which concentrate on small areas of the skeleton (particularly measurements employing two-dimensional radiographs) cannot readily be extrapolated to whole bones, and are not always applicable to the whole skeleton. The present cross-sectional study has attempted to circumvent these difficulties in two ways: first, an examination of the loss of calcium and phosphorus from the entire skeleton, and secondly, by normalization of the data for the large variability in size, characteristic of a heterogeneous population of any age. It is clear that with this large variability in absolute levels of TBCa and TBP in each age group, the use of the average value for each decade can lead to quite erroneous conclusions. With a sufficiently large population, the error is, of course, minimized. However, once the individual data are normalized for the effects of skeletal size and sex, it is possible to estimate the rate of loss of skeletal mass with age in a cross-sectional study. Estimates of the time of onset of bone loss vary widely from 20 to 66 yr, depending on the criteria used and the bone selected?-” For example, studies of bone loss from radial and metacarpal cortex indicate no significant loss in women before 50 yr and before 60 yr in men. However, the present study suggests that the overall skeletal calcium loss commences within the fourth decade for women and the fifth decade for men. In the present study, the onset of the decline of skeletal bone mass was arbitrarily taken to be 35 yr for females and 45 yr for males. The change in skeletal mass after maturity (starting in the fourth and fifth decades, respectively) appears to be linear with time, at least through the ninth decade. The marked increase in rate of loss of Ca that occurs
BODY
CHEMICAL
93
COMPOSITION
after 55 yr in women, presumably associated with menopause, is superimposed on the slower rate of loss which starts at - 35 yr.5 In the male population, the decrease of TBCa starts at - 45 yr and continues over the life span. The rate of loss of TBCa is approximated with a linear function over the entire age span. The original equation for predicting calcium in males presented above is satisfactory without an additional age correction at least until 79 yr of age. The overall rate of change of calcium with time for the population of men and women in this study was determined in the following manner. Since height and body potassium decline with age, their effect on calcium loss, implicit in the equation for calcium prediction, must be considered along with the explicit age correction factor. Thus, the derivative of the calcium value with time takes into account the dependency of calcium on the three time-dependent variables (see Appendix B). The rate of decrease of muscle mass (as determined by measurement of @‘K) with time was previously measured in a large population: 250 females and 250 males, 30-79 yr of age.6 The mean rate of decrease of potassium with time for the females was found to be 0.45 g/yr and for the males 1.16 g/yr after 50 yr.6 The mean rate of loss of height with time was also determined for the same populations. The value for this loss was found to be 0.15 cm/yr for females and 0.28 cm/yr for males. Most observations on loss of height with age have been cross sectional on selected populations or abnormal populations. In Adams’ study (the only longitudinal study) the height loss was twice as great in women as men, but it is to be noted that the age group studied was between 55 and 64 yr.” Adams incidentally reported no change in height in 48% of the men and in 15% of the women, emphasizing great variability in the end response. In the present study, in order to analyze “normal” losses, any women who had any suggestion of osteoporosis were excluded. This exclusion may explain the small or average loss of height per year in women noted in the present study. The combination of the changes of height and potassium with age along with the age correction factor, (Y yields a range of values for the loss of calcium in females from 3 g/yr (for ages 35-54) to 7.01 g/yr (for ages 70-79) (see Table 2). For males the loss of calcium ranges from 6.2 g/yr at - 50 yr to 7.98 g/yr at 80-90 yr (Table 2). For the “average” woman, this amounts to a mean rate of loss of calcium of 0.38% per year for ages 35-54, and 1.08% per year over the Table
2.
Rate of loss of total-body
calcium
with
Females
age Males
AW
elv
%JY
30-39
3.00
0.39
-
-
40-49
2.90
0.37
-
-
SO-54 55-59
2.96 8.20
0.36 1.06
6.44
0.65
60-69 70-79
7.40 7.01
1.07 1.11
6.50 7.81
0.62 0.73
So-90
-
-
7.98
0.79
elv
WY
94
COHEN
ET Al.
age span 55-79. For the average man the mean loss of calcium was 0.70% per year after 50 yr. While no directly comparable data are available for comparison of total skeletal calcium loss, there are two useful longitudinal studies reported in the literature on the rate of loss of bone mass for the appendicular skeleton. Garn” measured the change in cortical thickness of the metacarpals of adult women over a 23-yr period, and reported a loss of thickness of 0.25 mm per decade. When this value is corrected for subperiosteal bone formation it yields a value of 0.75% per year. This value is somewhat lower than that observed in the present study for the female population. The corresponding rate of loss of cortical thickness in males was 0.45% per year, a value considerably lower than the results reported here. Adams” also carried out a longitudinal study on the effects of aging in white females over an 1 I-yr period, using the same end point, and reported a loss of 1.34% per year. The corresponding loss of cortical bone width for white males in the same study was 0.86% per year, slightly higher than observed here. From the present study it appears that for the average female, the average loss of bone calcium amounts to 3.8 g/yr, over the period 30-54 yr of age, and 7.6 g/yr after 55 yr of age. The average loss in males is 7 g/yr over the period 50-90 yr. Thus, for women, a total of 250 g of Ca are lost normally over the period of 30-80 yr; for men 210 g of calcium are lost over this time. This amounts to - 28% of the total body calcium in an average woman at 30 years of age, and - 20% of the total calcium in men. While the mechanism of bone loss with age is not as yet fully understood, it must ultimately result from an imbalance in the bone formation and bone resorption rates. The interesting point is that the mean loss of Ca in women of 0.37% and 1.1% per year, before and after menopause (50-60 yr), is equivalent to an average negative calcium balance of only 8 and 21 mg Ca per day, respectively. In males the negative calcium balance from 50-90 yr amounts to only - 19 mg per day. It is to be noted that these levels of Ca loss are within the experimental error of Ca balance techniques. The universal regression of calcium and phosphorus levels with age is similar to that of gonadal function. The accelerated rate of loss of total body calcium and phosphorus in women that commences at menopause may be associated with a progressively marked imbalance in anabolic and catabolic steroids that also occurs at 50-60 yr. It has been hypothesized that gonadal hormones normally retard bone resorption; thus their decline with time could result in an increased bone resorption rate. ACKNOWLEDGMENT The authors are indebted to Mr. M. Stravino and Mr. J. Rothmann for their ful technical assistance in performing the analyses in the previous study.
dedicated
and
skill-
REFERENCES 1. Chamberlain MJ, Fremlin JH, Peter DK, Philip H: Total body calcium by whole-body neutron activation: New technique for study of bone disease. Br Med J 2158 I, 1968 2. Palmer HE, Nelp WE, Murano R, et al:
The feasibility of in vivo neutron activation analysis of total body calcium and other elements of body composition. Phys Med Biol 13: 269, 1968 3. Cohn SH, Dombrowski CS, Fairchild RG:
BODY
CHEMICAL
In vivo activation Int J Appl Radiat
95
COMPOSITION
analysis of calcium Isot 21:127, 1970
in man.
4. Cohn SH, Shukla KK, Fairchild RG: Design and calibration of a broad-beam 238Pu,Be source for total body neutron activation analysis. J Nucl Med 13:487-492, 1972 5. Cohn SH, Vaswani A, Zanzi I, Ellis KJ: The effect of aging on bone mass in adult women. Am J Physiol (in press) 6. Ellis KJ, Shukla KK, Cohn SH: A predictor for total-body potassium based on height, weight, sex and age: Application in metabolic disorders. J Lab Clin Med 83:716727, 1974 7. Cohn SH, Ellis KJ, Wallach S, Zanzi I, Atkins HL, Aloia JF: Absolute and relative deficit in total skeletal calcium and radial bone
mineral content in osteoporosis. J Nucl Med 15:428-435, 1974 8. Ellis KJ, Cohn SH: The correlation between skeletal calcium mass and muscle mass in man. J Appl Physiol38:455460, 1975 9. Newton-John HR. Brian Morgan D: The loss of bone with age, osteoporosis and fractures. Clin Grtho 71:229-252, 1970 IO. Garn SM, Rohmann CG, Nolan P: Relations of development and aging, Ed. J.E. Birren. Springfield, Illinois: Thomas, 1964, p 43 1 I. Adams P, Davies GT, Sweetname P: Osteoporosis and the effects of aging on bone mass in elderly men and women. Q J Med 39: 60-615, 1970 12. Widdowson EM, McCance RA, Spray CM: The chemical composition of the human body. Clin Sci IO:1 13-124, 1950
96
COHEN
APPENDIX
A.
Equation
for
(Co,) Cap
Predicting
in Normal
the
Total
Body
ET AL.
Calcium
Subjects
112
= a HK
where H = height(m) K = body
potassium
(g)
a = 56.62 for female
subiects
5 55
a = 56.62 - 0.38 (A - 55) for females a = 54.5 for males Equation
for predicting
K, = a W”’
the total
body
> 55 yr
potassium
(K,)
in the normal
subjects.
Hz
where:
a = 5.52 - 0.014
A
males
a = 4.58
A
females
- 0.010
A = we (Y) W = weight (Kg) H = height(m)
APPENDIX
B.
Derivation
of Equation
Co,=
Rate
of Change
of Calcium
VlH
aK
!!!!$ = K’lz,.,g+yK ca aa
ac0
for
’
-‘/?a,,2+aK1/z?
at 1 co
aK
at
Ca aH
T’X-B;-+TKat+Kat For Women
For Men
a = 56.62 -
= a,
at
g
*Ref.
= -0.38 aK -= at
-0.15’
afi -= at
6. = predicted
K = total
body
total
body
potassium
calcium
(g).
(g).
H = height(m). a = coefficient
a = 54.5
= -0.45’
aH -= at
Co,
- 0.38 (A-55)
with age correction
factor.
-1.16’
- 0.28*
With
Time
