dietary perturbation of calcium metabolism normal man:...

Dietary Perturbation of Calcium Metabolism

in Normal Man: Compartmental Analysis

JAMESM. PHANG, MoNEsBERMAN,GERALDA. FINERMAN, ROBERTM. NEER,LEONE. ROSENBERG,and THEoDomRJ. HAHNwith the technical assistanceof LILLIAN FIsHER and ALNORAGRANGER

From the Metabolism Branch, National Cancer Institute, and the MathematicalResearch Branch, National Institute of Arthritis and Metabolic Diseases,National Institutes of Health, Bethesda, Maryland 20014

A B S T R A C T The effect of dietary calcium intake oncalcium metabolism was studied in eight normal volun-teers by multicompartmental analysis of radiocalcium andbalance data. In paired studies of six normal subjectson normal and high or low calcium intakes, necessaryand sufficient criteria were used to determine changes incalcium metabolic parameters produced by alterations indietary calcium. These changes involved gastrointestinalcalcium absorption rate, renal and endogenous fecal rateconstants, and bone resorption rate. Bone accretion rateand compartment sizes need not change between thepaired studies. The changes of parameters involving kid-ney, gut, and bone were in a direction to support calciumhomeostasis and were compatible with the pattern ofchanges produced by parathyroid hormone. However,the source of the stimulus for hormone secretion was notapparent since plasma calcium concentrations showedno significant difference between paired studies. Theimplications of these findings relative to control of hor-mone secretion, calcium regulatory mechanisms, andmetabolic bone disease are discussed.

INTRODUCTION

The clinical significance of dietary calcium intake is wellrecognized. High or low calcium dietary intakes havebeen included in the therapeutic regimen for a number of

Dr. Finerman's present address is Johns Hopkins Hos-pital, Baltimore, Md. Dr. Neer's present address is Massa-chusetts General Hospital, Boston, Mass. Dr. Rosenberg'spresent address is Yale-New Haven Hospital, New Haven,Conn.

Address requests for reprints to Dr. James Phang, Metab-olism Branch, National Cancer Institute, National Institutesof Health, Bethesda, Md. 20014.

Received for publication 1 April 1968 and in revised form28 June 1968.

metabolic disorders in man (1-4). To date, the evidencesupporting these dietary approaches to therapy has beenaccumulated through animal studies and analysis of vari-ous disease states by balance and tracer techniques(5-10). However, no distinct pattern of metabolic re-sponse to dietary calcium perturbation has yet emerged.It is the purpose of this report to demonstrate the re-sponse in calcium metabolism to dietary calcium per-turbation in normal human subjects and to relate theseresponses to calcium homeostasis.

METHODS

Materials. Eight normal volunteers were studied on theMetabolism Branch of the National Cancer Institute. Eachhad normal screening studies which included physical exami-nation, complete blood count, serum calcium, phosphorus,creatinine, alkaline phosphatase, urinalysis, sterile urine cul-ture, and chest and bone X-rays. All studies were done withthe subjects kept indoors in air-conditioned wards and withcontrolled physical activity. Metabolic diets of constant com-position were ingested by the subjects for the durationof the study. Paired metabolic balance and kinetic studieswere done on six of the volunteers. In the paired studies,each subject was equilibrated for 4-6 wk on 800 mg ofcalcium and 1200 mg of phosphorus daily dietary intakebefore the initial study. Another 4-6 wk were allowed forreequilibration at a new level of dietary calcium intakebefore starting the second calcium kinetic and metabolicbalance study. Three of the six subj ects had their dietarycalcium intakes reduced to 200 mg/day. The other threehad their calcium intake increased to 2000 mg' of calciumper day. The increase in calcium intake was accomplishedwith milk in the case of K. B. (subject No. 6) with anaccompanying increase in phosphorus intake. In R. M. (sub-ject No. 4) and S. C. (subject No. 5), calcium lactate tabletswere given. Two other subjects (Nos. 7 and 8) were studiedon high and low calcium intakes without a preceding studyon 800 mg of calcium intake. In total, four subjects werestudied on low calcium intake and four subj ects on highcalcium intake. The summary of the studies on the eightsubjects is given in Table I.

The Journal of Clinical Investigation Volume 48 1969 67

TABLE ISummary of Balance and 47Ca Kinetics Studies Performed on Eight Subjects

Diet

Subject Sex Age Weight Height Study Ca P Remarks

kg cm g/day

1. S. G. F 21 79 173 I 0.81 1.23 Normal calcium dietary intakeII 0.22 0.97 Low calcium dietary intake

2. A. M. F 21 45 152 I 0.78 1.10 Normal calcium dietary intakeII 0.19 0.72 Low. calcium dietary intake

3. J. K. M 27 56 172 I 0.95 1.26 Normal calcium dietary intakeII 0.22 0.80 Low calcium dietary intake

4. R. M. F 24 60 155 I 0.83 1.13 Normal calcium dietary intakeII 2.06 1.16 High calcium intake; calcium lactate supple-

ments

5. S. C. M 26 85 181 I 0.93 1.71 Normal calcium dietary intakeII 2.10 1.71 High calcium intake; calcium lactate supple-

ments

6. K. B. M 23 78 179 I 0.79 1.67 Normal calcium dietary intakeII 2.14 2.62 High calcium intake; increased milk intake.

This volunteer had isolated renal glyco-suria.

7. V. M. M 21 72 183 I 0.24 1.07 Low calcium dietary intake

8. N. B. M 21 67 187 I 2.12 2.26 Highcalcium intake;increased milk intake

GI Tract

V.a = _VA

Feces UrineFIGURE 1 A schematic representation of calcium metabolism. Vs are transfer rates expressedin g/day of calcium. Vl = ingested, V. = absorbed, Vf = endogenous fecal, V1 = total fecal,Vu = urinary excretion, Vo+ = bone accretion, V0 = bone resorption. Xs represent the rate con-stant from a compartment expressed as day'. Xf = gastrointestinal rate constant from compart-ment 1 into feces. Xu = rate constant from compartment 1 into urine; W= rate constant fromcompartment 4 into nonexchanging calcium in bone. M1, M2, M8, M4, MT= pool sizes in gramsof calcium for compartments 1, 2, 3, 4, and total exchangeable calcium pool, respectively,a= fractional absorption of calcium and A = calcium balance in g/day.

68 Phang, Berman, Finerman, Neer, Rosenberg, and Hahn

TABLE I ICalcium Metabolic Parameters Derived from Multicompartmental Analysis

of 47Ca Kinetics and Metabolic Balance Studies

Vi VU VP a Va a v XV VO, Vo MT

Low calcium intake1. S. G. I 0.812 0.178 0.602 +0.032 0.333 0.410 0.153 4 0.009 0.099 --0.007 0.417 --0.005 0.385 6.7 ±t0.4

II 0.221 0.125 0.170 -0.074 0.152 0.688 0.080 -0.005 0.081 ±i0.006 0.491

2. A. M. 1 0.781 0.125 0.691 -0.035 0.210 0.267 0.122 ±-0.007 0.112 ±i0.008 0.298 ±i0.005 0.333 6.3 ±0.3II 0.192 0.101 0.169 -0.078 0.119 0.620 0.104 ±-0.006 0.097 -±0.006 0.376

3. J. K. I 0.947 0.278 0.601 +0.068 0.477 0.504 0.251 --0.019 0.119 --0.010 0.547 ±t0.009 0.479 8.0 ±0.4II 0.224 0.174 0.168 -0.118 0.165 0.737 0.176 ±t0.014 0.098 ±40.009 0.665

7. V. M. 0.236 0.210 0.190 -0.164 0.149 0.631 0.192 ±i0.013 0.087 ±i0.009 0.382 --0.021 0.543 16-±t11

High calcium intake4. R. M. I 0.833 0.143 0.627 +0.063 0.297 0.357 0.129 ±t0.006 0.087 4-0.005 0.272 ±t0.005 0.209 5.2 --0.2

II 2.055 0.184 1.712 +0.159 0.457 0.222 0.170 --0.007 0.109 ±i0.006 0.113

5. S. C. I 0.927 0.203 0.709 +0.013 0.377 0.407 0.158 ±-0.007 0.112 4-0.009 0.363 ±t0.024 0.350 11 ±-2II 2.099 0.205 1.914 -0.020 0.367 0.175 0.158 -0.007 0.126 -0.009 0.383

6. K. B. I 0.785 0.358 0.601 -0.174 0.369 0.461 0.288 ±i0.002 0.133 -±0.013 0.531 ±=0.011 0.705 8.7 ±-0.4II 2.136 0.356 1.670 +0.110 0.638 0.299 0.269 ±40.020 0.151 --0.014 0.421

8. N. B. 2.118 0.307 1.778 +0.033 0.495 0.234 0.230 ±t0.009 0.105 ±40.007 0.530 ±t0.014 0.497 11 ±1-

The parameter values on high or low calcium intakes are shown below the values on base line (800 mg/day) dietary calciumintake. For explanation of the symbols, see Fig. 1. The parameter values from 47Ca kinetics were obtained from the constrainedpaired studies (see text), in which Vo+ and MTwere held constant.

10 ,uc of "7CaCl, dissolved in 2.0 ml of sterile isotonicsaline was given in a rapid intravenous injection on themorning of the 1st day of the study. Serum calcium specificactivity, as well as urine and fecal cumulative radioactivity,were monitored for 18 days. During this same period, bal-ance data for calcium, phosphorus, and nitrogen were col-lected in three 6-day pools. The isotope counting techniquesand the methods used for balance collections and chemicaldeterminations have been previously described (11).

Data analysis. The kinetic behavior of the calcium iso-tope as seen in serum, urine, and stool can be described bya four compartment model. Although many four compart-ment models can satisfactorily fit the data, the series modeldiscussed by Neer, Berman, Fisher, and Rosenberg (11)was used as the basis of comparison. Steady-state compart-ment sizes, rate constants, and flow rates were calculatedfrom the tracer and balance data by the Simulation, Analy-sis, and Modeling (SAAM) computer program of Bermanand Weiss (12, 13).

A physiological interpretation of the four compartmentseries model is shown diagrammatically in Fig. 1. Flowrates and rate constants are designated by Vs and Xs, respec-tively. V, is the rate of dietary calcium entering the gastro-intestinal tract. Va is the rate of calcium absorption intothe labile calcium pool. The fractional gastrointestinal ab-sorption, a, is the ratio of the rate of absorption to therate of ingestion (a=V/V.I). VO is the rate of calciumresorption from bone. Calcium may leave the labile poolthrough accretion into nonexchangeable bone at a rate, Vo+,and by losses to urine and feces at excretion rates Vu andVr, respectively. The theory and assumptions for this multi-compartmental analysis have been discussed in detail in aprevious report (11).

Six of the subj ects underwent paired studies on normaland either high or low calcium intakes, and therefore eachserved as his own control for analysis of the dietary per-turbation. There is no reason to expect alterations in everyparameter of the compartmental system after a dietarycalcium change. The rationale in the analysis, therefore, wasto determine a necessary and sufficient set of parameterchanges as an hypothesis to explain the observations (14).To accomplish this computationally, both sets of data frompaired studies were used jointly in the computer leastsquares fitting procedure, with the constraint that one ormore of the kinetic parameters have identical values forboth studies.

RESULTS

Model testing. Kinetic data from paired studies werefirst analyzed individually. The least squares fittingprocedure demonstrated significant parameter changesonly in Au and Xf between each of the paired studies.Since other combinations of parameter changes may alsofit the data, formal model testing was used to rigorouslytest these initial results.

No significant difference in serum calcium concentra-tion was seen between the normal and high or lowcalcium intake periods (25 paired postabsorption bloodsamples from each subject; normal and low intake,10.46 ±0.05 mg/100 ml and 10.37 ±0.04 mg/100 ml,respectively, P < 0.2; normal and high intake, 10.27 +

0.04 mg/100 ml and 10.32 ±0.04 mg/100 ml, respectively,

Dietary Perturbation of Calcium Metabolism in Normal Man 69

TABLE IIIIntercompartnmental Rate Constants, Transfer Rates, and Steady-State Pool Sizes for the

Studies on Normal Subjects

Mean Mean Combinedlow high mean with

1. 2. 3. 7. Ca 4. 5. 6. 8. Ca standardS.G. A. M. J.K. V.M. diet R.MI. S.C. K.B. N.B. diet deviation

MI 1.24 ±-0.08 0.99 40.07 1.10 ±0.08 1.1 ±0.2 1.1 1.05 ±0.04 1.44 ±0.09 1.29 40.09 1.4 ±0.1 1.3 1.2 ±0.2M2 0.96 ±0.09 0.65 ±0.05 1.14 ±0.11 1.5 ±(0.4 1.1 0.75 ±0.08 2.1 ±0.4 1.4 ±t0.2 0.8 ±0.1 1.3 1.2 ±0.9M13 2.1 ±0.2 1.9 40.1 2.4 ±0.2 3.4 ±i1.2 2.4 1.4 ±0.2 2.7 ±0.3 2.8 ±t0.5 2.6 ±+0.2 2.4 2.4 ±-0.6M4 2.3 ±0.3 2.8 ±-0.2 3.4 ±0.2 9.6 ±8.8 4.5 2.0 ±:0.2 4.6 ±t1.6 3.1 40.4 5.9 40.7 3.9 4.2 ±2.5Mir 6.7 ±t0.4 6.3 ±-0.3 8.0 ±)0.4 16 ±11 9.2 5.2 ±0.2 11 ±-2 8.7 40.4 11 ±t1 8.9 9 ±4X21 24 ±6 21 46 28 ±7 22 ±4 24 26 4±5 15 ±t2 24 ±:7 25 44 23 24 44X2t 3(0 ±8 32 ±8 27 ±6 16 ±3 26 36 ±8 10 ±3 22 ±46 42 ±5 28 27 410X32 5.1 ±1. 0 6.4 ±0.8 5.3 ±1.1 2.3 ±0.5 4.8 5.5 ±1.4 1.5 40.7 4.2 ±1.1 8.7 41.5 5.0 4.9 ±2.3X23 2.1 ±-0.4 2.0 ±(0.2 2.3 ±0.5 0.9 ±0.2 1.8 2.7 ±0.7 1.0 ±-0.3 1.8 ±0.6 2.6 ±0.4 2.03 1.9 ±40.7X43 0.42 ±0.11 0.42 ±(0.05 0.55 4±0.12 0.26 ±()0.06 0.41 0.62 40.16 0.21 ±0.02 0.53 40.22 0.62 40.06 0.50 0.46 ±40.16X34 0.21 ±0.08 0.18 40.05 (1.22 ±0.115 0.06 ±40.03 0.17 0.30 ±0.07 0.042 ±0.004 0.31 ±0.11 0.18 ±0.03 0.21 0.19 ±0.10X5o 0.18 ±(0.02 ().11 ±0.01 0.16 ±0.01 0.04 ±10.0)4 0.12 0.14 ±0. 01 0.08 ±0.03 0.17 40.02 0.09 ±t0.01 0.12 0.12 ±0.05P21 3(0 46 21 ±5 31 ±t6 24 ±6 26 27 ±5 21 ±3 32 ±6 35 ±6 29 28 ±5P12 29 ±6 21 ±5 30 ±6 24 ±6 26 27 ±5 21 ±43 31 46 34 ±5 28 27 ±5P32 5.0 ±i1.6 4.2 ±0.3 6.0 ±0.8 3.4 ±0.8 4.7 4.1 ±0.7 3.1 41.0 5.8 41.0 7.0 ±2.0 5.0 4.6 ±1.4P23 4.5 ±0.6 3.9 ±0.3 5.5 ±0.8 3.1 40.8 4.3 3.8 40.7 2.8 4±1.01 5.3 ±1.0 6.8 ±-2.0 4.7 4.2 ±1.4P43 0.9 ±-0.1 0.8(1 ±0.06 1.3 ±-0.2 (1.9 ±0.2 1.0 0.9 ±-0.1 (1.55 ±0.06 1.5 O0.4 1.6 ±0.3 1.1 1.1 ±10.4P34 0.5 -4-0.1 0.50 4-0.06 (0.8 -4-0.2 0).5 -J(). 2 0.6 0.6 -40.1 0.19 -4-0.06 1.() -40.4 0.11 -40.4 0.48 0.54 -4-0.3

Mi, M2, MI, M4, MT = pool sizes in grams of calcium for compartments 1, 2, 3, 4, and total exchangeable calcium pool, respectively. The Xijs designatethe rate constants from compartment j into comnpartment i, and pijs are the corresponding transfer rate. These rate constants and pool sizes were ob-tained from the paired analysis and were constrained at identical values between the studies on base line and high or low calcium dietary intakes.

-0.2 rCo

g /day 0o

4-0.2

+0.4H+0.6

+0.8

p -0.5g/doy 0

+0.5

+1.0_+1.5_

2 1, i,Normal

U02 Urine

tSloo/

0 6 12 18 02 I

FIGURE 2 Typical metabolic balancedata on normal and low calcium diets.Data are expressed according to the con-vention of Reifenstein. The normal studyis shown oin the left and the low calcium

study on the right. There was an inter-vening equilibration period of 6 wk be-tween the studies. Negative calciumbalance was produced without significantchanges in phosphorus ornitro6en bal-

6 12 18 ances.2 3

1i =220 mg/day


Ng /day

DAYPERIOD

l= 810 mg/day

SERUMANDURINE SPEC/F/C ACTIVITY0 800 mg intake* 200mg intake

a

FIGURE 3 (a) Serum and urine specificactivity curves from the same subject onnormal and low calcium intake plottedsemilogarithmically vs. time in days. (b)Cumulative excretion of '7Ca in urineand feces with normal and low calciumintake in the same subject.

20

P < 0.4). However, there were distinct differences inrenal excretion rate Va (Table II). Since there wasno significant change in serum calcium concentration,the difference in renal calcium excretion rates was likelydue to a change in renal calcium clearance. To testwhether such a change would necessitate a change in Au,

the paired studies were fitted jointly with common Au

values, while changes in all other parameters and poolsizes were free to adjust. A good fit of the tracer datacould not be obtained for this hypothesis. A similar testwas performed for the gastrointestinal rate constant, At,

with comparable results. Thus, the testing led to theconclusion that both Af and Au must change with an

alteration in the level of calcium dietary intake.

Since Au and Af had to change, the next questionasked was whether all the differences in tracer kineticsproduced by changes in dietary calcium intake could beexplained by changes only in Au and At. This was testedby fitting both sets of data jointly with the constraint thatall corresponding compartment sizes and As, except Auand Af, have the same value between the paired studies.A good fit of both sets of data was obtained. These re-sults rigorously corroborated the initial unconstrainedanalysis and demonstrated that the changes in Au and Atwere not only necessary but also sufficient to accountfor the changes in '7Ca kinetics on the low and high cal-cium diets. In addition, the alterations in total fecalcalcium and calcium balance with dietary calcium per-


0.50

<D

2C-,

-J

4

a:

(O 0.10

ILJ

too

0)00

'.. 0.05

z

z0

4ca:

0.01

z0

IL.

4-i

5 10 15DAYS-

LOWCALCIUM INTAKELO

HIGH CALCIUM INTAKE A" ,,_ to ~~~~~~~~~~~~0.14+ 0.01F' 0.297 '0.209 0.272 +005 0 +

\0O.497 / .113-

0.087+0.005 0.129 + 0.0060.109 + 0.006 0.170 + 0.007 MT= S.2 + 0.2

Af AUFIGURE 4 Typical results of simultaneous multicompartmental analysis of paired studies in two volun-teers on low (subject S.G.) and high (subject R.M.) dietary calcium intakes. Parameter values wereobtained from the paired analysis in which pool sizes, Vo+, and intercompartmental rate constantswere constrained at identical values. Using necessary and sufficient criteria, changes were seen inVa, Vo-, Xf and X. (see text). The values for these four parameters on base line (800 mg/day) calciumintake are shown above the value on high or low calcium intake.

turbation required changes in bone resorption rate, Vo,and gastrointestinal calcium absorption rate, Va. Thus,changes in calcium metabolism secondary to dietary cal-cium perturbation may be characterized by changes inmodel parameters representing kidney (Au), gut (Af,a),and bone (V0). The values calculated by the computerfor normal and high or low calcium intakes subject tothese constraints are summarized in Tables II and III.

Low calcium intake. With a 200 mg/day calciumintake, the calcium balance became negative in all pa-tients (Fig. 2), but no significant changes were seenin phosphorus or nitrogen balances. Excretion of radio-activity into urine and stool was decreased (Fig. 3 b).Although the early points in the serum specific activitycurves for normal and low calcium intakes were coin-cident (Fig. 3 a), there was a definite decrease in the

slope of the curve on low calcium intake beginningat about day 3. Since bone accretion rates could be heldconstant in the face of negative calcium balance, boneresorption rates had to be increased (Vo = Vo+- A).In all three subjects studied on low calcium intake, simi-lar changes at kidney, gut, and bone were seen. The re-nal rate constant, Au, decreased by a mean of 31%, thegastrointestinal rate constant, Ar, fell by 19% and thebone resorption rate, Vo-, increased by 20%. Althoughthe rate of dietary calcium absorption, Va, fell from amean of 0.36 g/day on normal intake to a mean of 0.14g/day on low calcium intake, the fractional absorption,a, increased by a mean of 81%. A typical response pat-tern is shown in Fig. 4.

High calcium intake. The direction of the changewas, in general, opposite that for the low calcium stud-


,k 4-

FIGURE 5 Typical balance data on normaland high calcium intakes. Positive calciumbalance was produced but no changes wereseen in phosphorus or nitrogen balances.There wsas an intervening equilibration pe-

6 12 18 0 6 12 18 riod of 6 Adwk betw-een studies.2 3

= 2060 mg/day

dies. In subject R. M., Va rose from 0.30 to 0.46 g ofcalcium per day on calcium lactate supplementation. Thecalcium metabolic balance was positive (Fig. 5), andthe excretion of radiocalcium in urine and feces in-creased (Fig. 6b). There was an increase in the slopeof serum specific activity curve but, as in the low cal-cium intake studies, early points in the specific activitvcurves where coincident in the paired studies (Fig. 6 a).Necessary and sufficient changes to explain the response

to high calcium intake involved the same plarametersas those for lowv calcium intake, i.e., Au, Af, V., and VJ-(Fig. 4).

In subject S. C., there was no increase in V. despitesupplementation with calcium lactate to a daily intakelevel of 2000 mg of calcium (Table II). There were alsono other changes in model parameters except for Af.The gastrointestinal rate constant, Af, and fractional ab-sorption, a, both fell while total fecal calcium excretion

rate, VF, was greatly increased. Thus, it appears thatin this subject, the gastrointestinal tract alone was ade-quate to maintain homeostasis in the face of high dietarycalcium intake.

In subject K. B., Va weas increased on a 2000 mg/daycalcium intake. Accompanying this increase in absorbedcalcium was an increase of Af and a decrease in Vow.There was no change, how-ev-er, in V. or Au. The addi-tional calcium for this subject wncas given by increasingmilk intake, and therefore there was a commensurateincrease in phosphorus intake. This may, in part, ex-

plain the lack of change in renal calcium excretion (seeDiscussion).

Pattern of homeostatic r-esponise. The constancy ofthe size of compartment 1, M1, reflected the unchangedmean values for postabsorptive serum calcium concen-

trations between paired studies. Thus, calcium homneo-stasis was p)reserv-ed in the face of dietary calcium

Dietary Perturbation of Calcium Metabolism in Normal Man 7 3

23,T,Normol

0

+0.2

+0.4

+0.6 .

+0.8

+1.0 .

+1.2

+1.4

+1.6

+1.8

+2.0

Cag/day

pg/day

Ng/doy

Urine

Stool

+15 L

DAY

PERIOD

0

2 3

IVA = 830 mg/day

v

a SERUMANDURI/NE SPEC/F/C ACTIVITYCU ling fIn-uA,

n P * 2000mg intakeJ-)

0-

w0

0 0.10.C- 0

z .iz

o.0~~~~~~~~~~~~~~~

0.01 . ,I,.,I,

bCUMULATIVERADIOACTIVITY

Calcium IntokeURINE a 800mg

0.30 2000mgz FECAL 0 800mgo * 2000mg ,

>0.20-J

5 10 15 20DAYS

FIGURE 6 (a) A comparison of serum and urine specificactivity curves on normal and high calcium intake in thesame subject. (b) Cumulative excretion of 'Ca in urineand feces of normal and high calcium intake in the samesubj ect.

manipulation. This preservation was possible becausechanges in calcium absorption rate, V., were balancedby compensatory changes in Au, At, and Vo. The frac-tional changes in Au, Af, and V-o between paired studieswere plotted against fractional changes in Va (Fig. 7).Fitting these data by the method of least squares, it wasfound that SVo/Vo = - 0.33 8Va/Va; SAu/Au = 0.53 SVa/Va; SAt/Af = 0.27 SVa/Va.

DISCUSSION

Model testing. The set of parameter changes deter-mined by necessary and sufficient criteria is not mathe-matically unique but provides the simplest hypothesis toexplain the observations in tracer data. In particular,

no changes in intercompartmental As and Ao+ were neces-sary. This finding implies that the paired serum "7Cadisappearance curves were superimposable if correctedfor urinary and fecal excretion (Au, At). It may be thattwo or more parameters were changed in combinationto produce this result fortuitously, but this is unlikely.It is also possible that there may have been subtlechanges beyond the resolution in our data.

It is likely that the hypothesis for parameter changesin tracer data (As) is correct for the above considera-tions. However, the likelihood of the sufficient hypothesisfor steady-state parameters, i.e., compartment sizes,VO+, Vo, cannot be forcefully argued. These parametersare sensitive to variations in the model for the site ofsteady-state input. Although there is little evidence sup-porting the entry of gastrointestinal calcium, Va, or re-sorbed calcium from bone, Vo, into a pool other thanplasma, one must leave open these possibilities. Futurestudies may require extensions of the hypothesis pro-posed in this report.

Calcium regulation. Plasma calcium concentrationin man is very closely regulated by parathyroid hor-mone (PTH). Thyrocalcitonin (TCT) may also be in-volved in this regulation. PTH secretion is stimulatedby hypocalcemia, and the hormone acts on kidney, gut,and bone to elevate plasma calcium levels (15). TCTdecreases plasma calcium concentration by decreasingbone resorption and is secreted in response to hyper-calcemia (16). Acting alone, or perhaps in concertwith TCT, PTH maintains plasma ionized calcium innormal individuals with little measureable fluctuation.

The independent variable in our studies is the levelof ingested calcium. Kinetics and balance data providequantitative estimates of the homeostatic responses atkidney, gut, and bone after the subject has adjusted tothe change in calcium intake. This regulatory adjust-ment may be complex and involve a number of factors.

Low calcium intake. With a decrease in dietary cal-cium intake, Vl, there was a distinct pattern of metabolicresponse. Absorbed calcium, Va, fell from 0.34 ±0.013g/day on 0.8 g/day intake to 0.15 ±0.02 g/day on 0.2g/day intake. Accompanying this decrease in Va, therewas an increase in bone resorption rate (Vo-) and a de-crease in renal and gastrointestinal rate constants (Au,Af). The fractional changes at these three sites (8Vo-/Vo-, SAu/Au, SAt/At) were of a similar magnitude(- 25%), and suggest a common mechanism. SincePTH has effects on kidney, gut, and bone which paral-lel the observed changes, it is attractive to hypothesizethat increased PTH activity is the common factor medi-ating the target organ responses. Unfortunately, it hasnot yet been possible to test this directly. However,there is evidence from animal studies which supportsthis hypothesis (17-19).

In spite of a response pattern to conserve calcium, a


0203

01

Ixulxu

M-0.53+0.03-0.8-0.6-0.40.260.2 040.x0.8 0002

31

Sxfgalf

M= 0.27+0.0I-0.8-0.6-0.480.2 V 6 -L°0.2 0.40.60.8

0 Low Calcium* High Calcium

FIGuRE 7 Linear regression of the fractional change in Vo-, X., and xt on frac-tional change in V^. For each subject, fractional change in these parameters was

calculated as the ratio of the difference produced by high or low dietary calciumintake to the value on normal calcium intake in that subject. M is the slope of thestraight line passing through the origin and was calculated by the method of leastsquares Subject No. 6 is K.B. who had increased phosphorus intake accompanyingthe high calcium intake and also had isolated renal glycosuria.

negative calcium balance was produced in all four sub-jects on a 200 mg/day calcium intake. This is a well-known phenomenon. One may argue that negative bal-ance is only temporary and that additional time is re-

quired for complete metabolic adjustment. Even so,

metabolic balance can only be achieved after there hasbeen a definite decrease in total body calcium; i.e., plasmacalcium homeostasis occurs at the expense of calcium inbone.

Phosphorus intake. The inorganic phosphorus con-

tent of the diet was decreased (- 30%) during the lowcalcium intake studies (Table I). Phosphorus balance,however, remained unchanged (Fig. 2). Dietary phos-phorus is known to affect renal, intestinal, and osseous

calcium metabolism (20), but these effects were relatedto hypophosphatemia and marked negative phosphorusbalances, neither of which were present in our studies.Furthermore, a decrease in phosphorus intake would

0.5

,, 0.40

Em 0.3

0.2

0.1

0

DIETARY CALCIUM ABSORBED

0.2 0.8 2.0VY g/day

FIGURE 8 Rate of calcium absorption, Va, plotted againstrate of calcium intake, V,. Mean values for three intake ratesare shown. The curve was obtained by least squares fit usingthe expression; Va = kVs/(ks + V5).

alter urinary calcium in a direction opposite that ob-served in our studies. It is therefore unlikely that theresponse seen in the low calcium studies was significantlyaltered by the small change in phosphorus intake.

High calcium intake. A high dietary calcium intakeproduced changes in kidney and bone metabolism ofcalcium when there was an increased rate of calcium ab-sorption from the gastrointestinal tract. In subject R.M.,V. increased from 0.30 to 0.46 g/day with calcium lac-tate supplementation, but in subject S. C., Va remainedunchanged despite the same high level of calcium intakeas in R. M. This finding suggests that there may be in-dividual variation in the ability to utilize an increaseddietary calcium load. The previous dietary history hasbeen suggested as a factor in determining the level forV. with increased dietary calcium intake (9, 10). It hasalso been suggested that positive calcium balance froman increased calcium intake may be maintained only fora limited duration (5). It may be that the subjects withan increased V. will undergo further equilibration aftera period of positive calcium balance.

It is clear that an increase in Va produced a physio-logic response pattern involving changes in calciummetabolism at kidney, gut, and bone. Resorption of cal-cium from bone was decreased, whereas renal and gas-

trointestinal calcium clearances were increased. Thatthis pattern included the same organ sites as those seen

with calcium deprivation again suggests a change inhormone secretion may be the mediating factor.

In subject K. B., there were gastrointestinal and boneresponses consistent with those observed in the othersubjects, but no renal response was seen. Au remainedunchanged despite the increase in Va. However, sincethis subject was given milk to increase his calcium in-take, there was a concomitant increase in phosphorus


01030_7 2_

M:-0.33+ 0.02-0.8-0.6-0.4-0.2

0.2040.6 0.8- 0406

I

intake. It has been shown that high phosphorus intakewill decrease the urinary excretion of calcium in normalsubjects (21, 22). The lack of any increase in urine cal-cium excretion in K. B. may have been due to this effect.Also, K. B. had isolated renal glycosuria (11). Althoughno other renal tubular defects were found, the failureof renal adaptation to calcium loading might representa subtle form of nephropathy.

Gastrointestinal absorption of calcium. The rate ofcalcium absorption, V., is dependent on the rate of cal-cium ingestion, Vt. PTH control of this absorptionprocess has been proposed by several investigators (23,24). The efficiency at which the ingested calcium is ab-sorbed presumably is altered by PTH concentrationwhich, in turn, is sensitive to plasma calcium concentra-tion. A plot of V. vs. Vi in our studies suggests thatV. approaches saturation with increasing Vt (Fig. 8),and that their relationship may be expressed reasonablywell by a Michaelis-Menten type of equation,

Va = ki Vik2 + Vi

with values for kL and k2 of 0.66 and 0.77, respectively.This suggests the possibility of an intrinsic (nonhor-monal) control of calcium absorption. Such a local con-trol mechanism has been suggested by the studies ofKimberg, Schachter, and Schenker in rats (25). Hor-monal control of the gastrointestinal absorption process,however, is not inconsistent with our data, and it ispossible that both local and hormonal control mecha-nisms are present.

One might expect Xf to be affected in a manner paral-leling the changes in the efficiency of calcium absorption,since the gastrointestinal rate constant is a composite ofcalcium secretion from the plasma (x.) and reabsorp-tion from the gastrointestinal tract (x. * a) such thatXf = A. (i-a). The actual changes in Af observed, how-ever, are smaller than predicted from the above relation-ship, a fact which suggests that the reabsorption of cal-cium secreted from the plasma is less than absorption ofingested calcium. This hypothesis was first proposed byHeaney and Skillman (26). They suggested that se-creted calcium may be divided into absorbable and non-absorbable fractions such that only the absorbable por-tion is affected by changes in a. Recent studies byBirge' demonstrating variations in calcium absorptionalong the human gastrointestinal tract elaborates onHeaney and Skillman's hypothesis. As in the case ofV., changes in At reflecting changes in a may be ex-plained with or without invoking hormone regulation.

Kidney. The changes in Au occurred in the face ofconstant serum calcium concentration. It would be diffi-cult to exclude the possibility that small changes in ul-trafiltrable calcium or glomerular filtration rate wereresponsible for the changes in Au. More likely is a

change in renal tubular reabsorption of calcium. Anumber of studies show that renal tubular reabsorptionof calcium is increased by parathyroid hormone (27, 28).It is reasonable and consistent, therefore, to hypothe-size hormone regulation at the kidney in response tochanges in calcium dietary intake.

A small but significant difference in serum calciumconcentration with low calcium diets was reported byMacFadyen, Nordin, Smith, Wayne, and Rae (29).These authors then correlated changes in urinary cal-cium excretion with changes in renal calcium filteredload. In our study, no significant difference in serumcalcium concentrations was found between the normaland high or low calcium intake periods. This discrep-ancy between our data and that of MacFadyen et al. maybe due to the different durations employed for equilibra-tion on the low calcium diet. In MacFadyen's study,blood calcium determinations were obtained on 2 succes-sive days with an acute perturbation in dietary calciumintake on the 2nd day. It may be that they observed dif-ferences in serum calcium concentration during a transi-ent phase before completion of homeostatic responses inthe gastrointestinal tract and bone.

Bone. In view of constant plasma calcium concentra-tions, changes in bone resorption rate (Vo-) are mostlikely due to hormonal control. This has been suggestedby Jowsey and Gershon-Cohen (18) in studies of lowcalcium dietary effects in cats.

Stimulus for hormone secretion. The changes at kid-ney, gut, and bone produced by perturbations in dietarycalcium intake are consistent with the regulatory ef-fects of PTH. The stimulus for secretion of this hor-mone, however, is not apparent from our studies. It hasbeen clearly demonstrated in cows that plasma PTHconcentration is proportional to the degree of acutechange in plasma calcium levels (30). In our study, nosignificant difference in mean serum calcium concentra-tions could be detected at high and low calcium intakes.This would suggest either that hormone secretionchanges its sensitivity to plasma calcium with chronicstimulation, or that another triggering mechanism forPTH secretion exists. The former hypothesis has beensupported in part by studies in cows with milk fever(31). The latter is suggested by studies in other hor-mone controlled metabolic systems where intestinal ab-sorption may act as a triggering mechanism (32, 33),but there has been no investigation of this possibilityin calcium regulation in animals or in man. It must bepointed out, however, that we found small, hour to hour,fluctuations in plasma calcium concentration, and it isconceivable that these fluctuations may be factors inthe regulation of PTH or TCT secretion.

Pattern of response. The correlation of bone resorp-tion rate and renal and gastrointestinal rate constants togastrointestinal calcium absorption rate may be of diag-nostic and therapeutic significance. A number of rela-


tively simple techniques are available for determiningcalcium absorption1 (34, 35). Using these techniques,one may predict normal patterns of response in VC, Au,and Xf following a change in Va. Future kinetic studiesmay demonstrate abnormal patterns of response. Theseabnormal patterns may help to elucidate primary patho-genetic mechanisms in calcium disorders and metabolicbone disease.

ACKNOWLEDGMENTSThe success of these studies is due largely to the enthusiasticefforts of Mrs. Shirley Harris, Mr. George Washington,Mrs. Sylvia Downing, Mrs. Cordie Lee Montgomery, MissEileen Murnin, Mrs. Jeanne Kruhm, and the entire staff ofthe Metabolism Ward of the National Cancer Institute. MissDonna Wotring, Mrs. Geraldine Carr, Mrs. ChristineMathers, and Miss Sharon Morgart were most helpful inpreparation of the manuscript. To them we wish to expressour thanks and appreciation. Wewould also like to acknowl-edge the valuable advice of Mrs. Marjory Weiss and theadvice and encouragement of Dr. Nathaniel I. Berlin.

REFERENCES

1. Brooks, W. D. W. 1934. Calcinosis. Quart. J. Med. 3: 293.2. Whedon, G. D. 1959. Effects of high calcium intakes on

bones, blood and soft tissues; relationship of calciumintake to balance in osteoporosis. Fed. Proc. 18: 1112.

3. Forbes, A. P., and E. Dempsey. 1963. Nephrolithiasis.In Diseases of the Kidney. M. B. Strauss and L. G.Welt, editors. Little, Brown and Company, Boston.

4. Higgins, C. C., and R. A. Stratton. 1963. Nephrolithiasis.In Urology. M. F. Campbell, editor. W. B. SaundersCompany, Philadelphia. 1.

5. Schwartz, E., V. A. Panariello, and J. Saeli. 1965. Radio-active calcium kinetics during high calcium intake inosteoporosis. J. Clin. Invest. 44: 1547.

6. Menczel, J., R. Schraer, G. Pakis, A. S. Posner, andR. C. Likins. 1963. Effect of a low calcium diet on bonecrystallinity and skeletal uptake of 'Ca in rats. Proc. Soc.Exp. Biol. 112: 128.

7. Spencer, H., J. Menczel, I. Levin, and J. Samachson.1964. Absorption of calcium in osteoporosis. Amer. J.Med. 37: 223.

8. Carlsson, A. 1951. Metabolism of radiocalcium in relationto calcium intake in young rats. III. The effect of cal-cium deprivation on the absorption and on the excretionof radiocalcium. Acta Pharmacol. 7: 23.

9. Harrison, M., R. Fraser, and M. Bridget. 1961. Calciummetabolism in osteoporosis: acute and long-term re-sponses to increased calcium intake. Lancet. 1: 1015.

10. Nordin, B. E. C. 1961. The pathogenesis of osteoporosis.Lancet. 1: 1011.

11. Neer, R., M. Berman, L. Fisher, and L. E. Rosenberg.1967. Multicompartmental analysis of calcium kineticsin normal adult males. J. Clin. Invest. 46: 1364.

12. Berman, M. 1965. Compartmental analysis in kinetics.In Computers in Biomedical Research. R. Stacey and B.Waxman, editors. Academic Press Inc., NewYork. 2: 173.

13. Berman, M., and M. F. Weiss. 1967. SAAMManual.U. S. Public Health Service Publication No. 1703. U. S.Government Printing Office, Washington, D. C.

1Birge, S. J., W. A. Peck, M. Berman, and G. D. Whedon.Study of calcium absorption in man: a kinetic analysis andphysiologic model. Data to be published.

14. Berman, M. 1963. A postulate to aid in model building.J. Theor. Biol. 4: 229.

15. Rasmussen, H. 1961. Parathyroid hormone, nature andmechanism of action. Amer. J. Med. 30: 112.

16. Munson, P. L. 1966. Thyrocalcitonin. Ann. Intern. Med.64: 1353.

17. Engfeldt, B. 1950. Studies on parathyroidal function inrelation to hormonal influences and dietetic conditions.Acta Endocrinol. 5 (Suppl. 6): 1.

18. Jowsey, J., and J. Gershon-Cohen. 1964. Effect of dietarycalcium levels on production and reversal of experimentalosteoporosis in cats. Proc. Soc. Exp. Biol. Med. 116: 437.

19. Au, W. Y. W., and L. G. Raisz. 1965. Effect of vitaminD and dietary calcium on parathyroid activity. Amer. J.Physiol. 209: 637.

20. Lotz, M., T. Zisman, and F. C. Bartter. 1968. Phosphorus-depletion syndrome in man. N. Engl. J. Med. 278: 409.

21. Malm, 0. J. 1953. On phosphates and phosphoric acidas dietary factors in the calcium balance of man. Scand.J. Lab. Invest. 5: 75.

22. Spencer, H., J. Menczel, I. Lewin, and J. Samachson.1965. Effect of high phosphorus intake on calcium andphosphorus metabolism in man. J. Nutr. 86: 125.

23. Talmage, R. V., and J. R. Elliott. 1958. Influence ofparathyroids on intestinal absorption of radiocalciumand radiostrontium. Fed. Proc. 17: 160.

24. Rasmussen, H. 1959. The influence of parathyroid func-tion upon the transport of calcium in isolated sacs ofrat small intestine. Endocrinology. 65: 517.

25. Kimberg, D. V., D. Schachter, and H. Schenker. 1961.Active transport of calcium by intestine: effects ofdietary calcium. Amer. J. Physiol. 200: 1256.

26. Heaney, R. P., and T. G. Skillman. 1964. Secretion andexcretion of calcium by the human gastrointestinaltract. J. Lab. Clin. Med. 64: 29.

27. Kleeman, C. R., R. G. Rockney, and M. H. Maxwell.1958. The effect of parathyroid extract (PTE) on therenal clearance of diffusible calcium. J. Clin. Invest. 37:907.

28. Talmage, R. V. 1956. Studies on the maintenance ofserum calcium levels by parathyroid action on bone andkidney. Ann. N. Y. Acad. Sci. 64: 326.

29. MacFadyen, I. J., B. E. C. Nordin, D. A. Smith, D. J.Wayne, and S. L. Rae. 1956. Effect of variation indietary calcium on plasma concentration and urinaryexcretion of calcium. Brit. Med. J. 1: 161.

30. Ramberg, C. F., Jr., G. P. Mayer, D. S. Kronfeld, G. D.Aurbach, L. M. Sherwood, and J. T. Potts, Jr. 1967.Plasma calcium and parathyroid hormone responses toEDTA infusion in the cow. Amer. J. Physiol. 213: 878.

31. Buckle, R. M., L. M. Sherwood, G. D. Aurbach, andJ. T. Potts, Jr. 1967. Calcium homeostasis: relative im-portance of parathyroid hormone and thyrocalcitonin.In Symposium on Thyrocalcitonin and the C Cell,London.

32. Dupre, J. 1964. An intestinal hormone affecting glucosedisposal in man. Lancet. 2: 672.

33. McIntyre, N., C. D. Holdsworth, and D. S. Turner. 1965.Intestinal factors in the control of insulin secretion.J. Clin. Endocrinol. 25: 1317.

34. Avioli, L. V., J. E. McDonald, R. A. Singer, and P. H.Henneman. 1965. A new oral isotopic test for calciumabsorption. J. Clin. Invest. 44: 128.

35. Robinson, C. J., D. H. Gutteridge, G. F. Joplin, E. H.Belcher, and R. Fraser. 1964. Calcium absorption testsusing '7Ca. In Medical Uses of '4Ca Second Panel Re-port. International Atomic Energy Agency, Vienna.


dietary perturbation of calcium metabolism normal man:...

Documents