Full text of "Ciba Foundation colloquia on ageing"

See other formats

CIBA FOUNDATION
COLLOQUIA ON AGEEVG

Vol. 4. Water and Electrolyte Metabolism in Relation
to Age and Sex

A leaflet giving details of available earlier volumes in this series,

and also of the Ciba Foundation General Symposia, and Colloquia

on Endocrinology, is available from the Publishers.

CIBA FOUNDATION
COLLOQUIA ON AGEING

VOLUME 4

Water and Electrolyte Metabolism in Relation
to Age and Sex

Editors for the Ciba Foundation
G. E. W. WOLSTENHOLME, O.B.E., M.A., M.B., B.Ch.

and
MAEVE O'CONNOR, B.A.

With 85 Illustrations

LITTLE, BROWN AND COMPANY

BOSTON

THE CIBA FOUNDATION

for the Promotion of International Co-operation in Medical and Chemical Research
41 Portland Place, London, W.l.

Trustees:

The Right Hon. Lord Adrian, O.M., F.R.S.

The Right Hon. Lord Beveridge, K.C.B., F.B.A.

Sir Russell Brain, Bt.

The Hon. Sir George Lloyd -Jacob

Sir Raymond Needham, Q.C, F.S.A.

Executive Council:
Sir Raymond Needham, Chairman Professor Dr. Dr. h.c. R. Meier

Lord Beveridge Mr. Philip Mair

Professor A. Haddow, F.R.S. Professor F. G. Young, F.R.S.

Director^ and Secretary to the Executive Council :
Dr. G. E. W. Wolstenholme, O.B.E.

Deputy Director:
Dr. H. N. H. Genese

Assistant Secretary : Editorial Assistants :

Miss N. Bland Miss Cecilia M. O'Connor, B.Sc.

Miss Maeve O'Connor, B.A.

Librarian :

Miss Joan Etherington

This book may not be reproduced by
any means, in whole or in part, ivith-
out the permission of the Publishers

Published in London by

J. dh A. Churchill Ltd.

104 Gloucester Place, W.l

First published 1958

Printed in Great Britain

PREFACE

This volume represents the fourth colloquium in the Ciba
Foundation's programme for the encouragement of basic
research relevant to processes of ageing which was initiated
by the Trustees early in 1954. In line with the series of
conferences begun earlier on Endocrinology, these meetings
are arbitrarily described as Colloquia to distinguish them from
the single conferences on isolated subjects which are known
as Symposia.

This colloquium on Water and Electrolyte MetaboHsm in
Relation to Age and Sex brought together a number of people
working on these problems from very different angles, with
what success the reader may judge for himself. Membership
had to be limited to a small group, as usual, but it is hoped
that the published proceedings will have a world-wide
readership, and will prove to be of value to those workers
in this field who could not be asked to participate on this
occasion, as well as to others not so closely associated with
such research.

Professor McCance, who directed the meeting with firm
but friendly skill and split-second time-keeping, also gave
much valuable help to the Deputy Director in its organization
and planning. He and Dr. Widdowson have continued their
assistance with some much appreciated advice on editorial
matters.

To those to whom this book serves as an introduction to
the activities of the Ciba Foundation it should be explained
that it is an international centre which owes its inception and
support to CIBA Ltd. of Switzerland. Under the laws of
England it is established as an educational and scientific
charity and is administered independently and exclusively by
its eminent British Trustees.

vi Preface

The aim of the Foundation is to improve co-operation in
medical and cjiemical research between workers in different
countries and different disciphnes. At its 200-year-old house
in the medical centre of London the Foundation provides
accommodation for scientists of all nationalities, organizes
conferences, conducts a medical postgraduate exchange scheme
between Great Britain and France, arranges a variety of
informal discussions, awards two annual lectureships, and is
building up a library service in special fields. In general, the
Foundation assists international congresses, scientific institu-
tions and individual research workers as much as lies within
its power.

CONTENTS

PAGE

Chairman's opening remarks

R. A. McCance 1

The development of physiological regulation of water
content

hy E. F. Adolph ....... 3

Discussion: Adolph, Black, Heller, Shock, Swyer, Talbot 11

Cellular aspects of the electrolytes and water in body
fluids

by H. Davson ........ 15

Discussion: Adolph, Davson, Fejfar, Hingerty, Talbot,

Wallace ......... 32

Hypematraemia and hyponatraemia with special reference
to cerebral disturbances

by P. FouRMAN and Patricia M. Leeson ... 36

Discussion: Adolph, Black, Borst, Davson, Desaulles,

FouRMAN, Wallace, Young ..... 58

Glandular secretion of electrolytes

by J. H. Thaysen 62

Discussion: Adolph, Black, Davson, Desaulles, Karvonen,

Talbot, Thaysen, Wallace . . . . . . 73

Hormonal aspects of water and electrolyte metabolism
in relation to age and sex

by G. I. M. Sw\t:r 78

Discussion: Adolph, Bull, Davson, Desaulles, Fourman,
Heller, McCance, Milne, Scribner, Swyer, Talbot,
Thaysen, Wallace, Widdowson, Young ... 93

General Discussion: Borst, Davson, Hingerty, McCance,

RiCHET, Scribner, Talbot, Thaysen .... 99

Body water compartments throughout the lifespan

by H. V. Parker, K. H. Olesen, J. McMurrey and B.

Friis-Hansen ....... 102

Discussion : Black, Borst, Bull, Davson, Fejfar, Fourman,
Heller, Hingerty, KSecek, McCance, Olesen, Scrib-
ner, Shock, Swyer, Widdowson . . . . .113

vii

76550

viii Contents

PAGE

The effect of variable protein and mineral intake upon
the body composition of the growing animal

hy W. M. Wallace, W. B. Weil and Anne Taylor 116

Discussion: Fourman, Heller, Kennedy, McCance, Milne,

Talbot, Wallace, Widdowson . . . . .136

The effect of age on the body's tolerance for fasting, thirst-
ing and for overloading with water and certain electrolytes

byN.B. Talbot and R. Richie 139

Discussion: Adolph, Black, Bull, Fourman, Kennedy,

McCance, Talbot, Wallace . . . . .150

Clinical consequences of the water and electrolyte meta-
bolism peculiar to infancy

by E. Kerpel-Fronius . . . . . .154

Discussion: Adolph, Black, Bull, Davson, Fejfar, Four-
man, Heller, Kennedy, McCance, Shock, Talbot,
Wallace, Widdowson, Young . . . . .162

The effect of hormones of the pituitary and adrenal glands
on the elimination of sodium, potassium and a water load
in infant rats during the weaning period

by J. K&ecek, Helena Dlouha, J. JelInek, Jarmila

KSeckova and Z. Vacek . . . . .165

Differences in the pattern of electrolyte and water excre-
tion in young and old rats of both sexes in response to
adrenal steroids

by P. A. Desaulles ....... 180

Discussion: Adolph, Borst, Desaulles, Fourman, Heller,

Kennedy, K6,ecek, McCance, Milne, Swyer . . 195

The effect of age on the electrolytes in the red blood cells
of different species

by M. J. Karvonen ....... 199

Discussion: Black, Bull, Davson, Desaulles, Fourman,

Hingerty, Karvonen, McCance, Milne, Shock . . 206

The development of acid -base control

by Elsie M. Widdowson and R. A. McCance . . 209

Discussion : Adolph, Fourman, Karvonen, McCance, Milne,

Scribner, Widdowson, Zweymuller .... 220

General Discussion: Adolph, Black, Bull, Desaulles,
Fourman, Hingerty, Kennedy, Milne, Richet, Shock,
SwYER, Talbot, Wallace ...... 224

Contents ix

PAGE

The role of the kidney in electrolyte and water regulation
in the aged

fti/ N. W. Shock 229

Discussion: Black, Bull, Borst, Fejfar, Heller,

HiNGERTY, Milne, Scribner, Shock, Zweymuller . . 246

Age and renal disease

by G. C. Kennedy ....... 250

Discussion: Borst, Desaulles, Fejfar, Fourman, Ken-
nedy, McCance, Milne, Richet, Swyer, Talbot . . 260

Renal function in respiratory failure

by D. A. K. Black 264

Discussion: Black, Borst, Bull, Davson, McCance, Milne,

Scribner ......... 268

Water and electrolyte metabolism in congestive failure

by Z. Fejfar 271

Discussion: Borst, Fejfar, McCance, Milne, Olesen . 298

A case of magnesium deficiency

by W. I. Card, and I. N. Marks ..... 301

Discussion: Black, Card, Davson, Fourman, Hingerty,

McCance 309

Concluding remarks : Adolph, Davson, Swyer . . . 311

Chairman's closing remarks

R. A. McCance 315

List of those participating in or attending the Colloquium on
"Water and Electrolyte Metabolism in Relation to Age and

Sex",
28th-30th January, 1958

E. F. Adolph . . . Dept. of Physiology, University of Rochester

School of Medicine, Rochester, N.Y.

D. A. K. Black . . . Dept. of Medicine, Royal Infirmary, Univer-

sity of Manchester

J. G. G. BoRST . . University Dept. of Internal Medicine,

Binnengasthuis, Amsterdam

J. P. Bull . . . M.R.C. Industrial Injuries and Burns Research

Unit, Birmingham Accident Hospital,
Birmingham

W. I. Card . . . Gastro-intestinal Unit, Western General

Hospital, Edinburgh

H. Davson . . . Medical Research Council, Dept. of Physio-

logy, University College, London

P. A. Desaulles . . Pharmaceutical Dept., CIBA Ltd., Basle

Z. Fejfar . . . Institute of Cardiovascular Research, Prague

P. FouRMAN . . . Medical Unit, The Royal Infirmary, Cardiff

H. Heller . . . Dept. of Pharmacology, University of Bristol

D. J. Hingerty . . . Dept. of Biochemistry and Pharmacology,

University College, Dublin

M. J. Karvonen . . Dept. of Physiology, Institute of Occupa-

tional Health, Helsinki

G. C. Kennedy . . Dept. of Experimental Medicine, University

of Cambridge

J. KXecek . . . Institute ofPhysiology, Czechoslovak Academy

of Sciences, Prague

R. A. McCance . . . Dept. of Experimental Medicine, University

of Cambridge

M. D. Milne . . . Dept. of Medicine, Postgraduate Medical

School, London

K. H. Olesen . . . Beringsvej 5, Copenhagen

G. RiCHET . . . Clinique des Maladies Metaboliques, Hdpital

Necker, Paris

B. H. ScRiBNER . . Dept. of INIedicine, University of Washington,

Seattle; and Dept. of Medicine, Post-
graduate Medical School, London

N. W. Shock . . . Gerontology Branch, Baltimore City Hospitals,

Baltimore

List of Participants

G. I. M. SWYER .

N. B. Talbot

J. H. Thaysen
W. M. Wallace

Elsie M. Widdowson

Winifred Young

E. Zweymuller

Obstetric Hospital, University College Hos-
pital, London

Dept. of Pediatries, Massachusetts General
Hospital, Boston

Medical Dept., Rigshospitalet, Copenhagen

Dept. of Pediatrics, Western Reserve Univer-
sity, Cleveland, Ohio

Dept. of Experimental Medicine, University
of Cambridge

Queen Elizabeth Hospital for Children,
Hackney, London

University Children's Clinic, Vienna; and
Dept. of Experimental Medicine, University
of Cambridge

CHAIRMAN'S OPENING REMARKS

R. A. McCance

When I first became interested in electrolytes some 25 or 30
years ago, there were not many other people interested in the
subject. Indeed, if they had been collected together in this
room for a symposium, they would have rattled about like
peas in a pod. But we did not meet. The world was no larger
then but there were no fairy godmothers like the Ciba Founda-
tion to transport us from distant parts of the world to London
in machines flying at hundreds of miles an hour in order that
we might see each other. Now there are so many people
interested in electrolytes that if all of them were to come to a
meeting, we should have to hold it in Trafalgar Square, or if it
were wet, in the Festival Hall.

We owe our fairy godmother a lot of thanks.

The subject of electrolyte metabolism has developed enor-
mously. We realize now that electrolytes enter into practically
every reaction that takes place in the body, but we still know
very little about a great many of them. The functions of
magnesium, for example, are still very much of a mystery, and
if anybody here can throw any light on this element it would
be very stimulating. We still know extremely little about how
and why the total amounts of the various electrolytes in the
body are maintained; why and how their relationships change
with age; what part each individual cell is playing and what
effect a change in the rest of the body may have on an indi-
vidual cell. That brings me to the object of this colloquium.
If you look at your programme you see that we have been
asked to try to put together our knowledge and information
about water and electrolyte metabolism in relation to age and
sex. You will see how the days have been divided up. The
first day will be devoted to "General principles". Then we

AGEING— IV— 1 1

2 R. A. McCance

have "The developing organism", and lastly "Senescence
and disease". I recognize the problems that arise when a col-
lection of "experts" get together: some people who are going
to speak today may not have any experience at all of the new-
born baby or of the effect of age on electrolyte metabolism —
except perhaps on their own, and I hope they have not had too
much of that ! Prof. Wallace can hardly be expected to be very
interested in old age ; he would prefer, I dare say, to listen to a
paper about congenital heart failure rather than the one about
congestive heart failure which Dr. Fejfar is going to give.
One of the objects of the symposium, however, is that he shall
do it. People speaking on Thursdaj^ moreover, may not have
thought about a newborn baby's renal function since they were
one themselves! At the same time it is very useful to have a
collection of experts brought together like this, if they — so to
speak — play to the title. We must always try to keep before
us the object for which we have been brought together, that
is to say to pool our knowledge so far as possible about the
metabolism of electrolytes in relation to age and sex.

As a corpus for dealing with electrolytes we may be a little
bit light on hormones. We could do with a few more specialists
in this field — there may be some unknown ones here who will
introduce themselves later — I hope there are! We shall re-
quire their assistance and I hope they will not be afraid of
saying what they think, when they think it. They will have
little chance of being contradicted!

It is a great pity that we shall have one absentee. I am very
sorry that our colleague Kerpel-Fronius could not come. He
is an old friend of mine and a very old friend of paediatrics
and electrolytes. I saw him not so long ago and he was much
looking forward to this international gathering. I personally
think he would appreciate it very much indeed if we were to
send him a letter as from the conference, saying how much we
are missing him. With your permission I shall write a letter
and send it off as from all of us.

THE DEVELOPMENT OF PHYSIOLOGICAL
REGULATION OF WATER CONTENT

E. F. Adolph

Department of Physiology, School of Medicine and Dentistry,
University of Rochester, New York

The plan of this study is to single out one way of measuring
the physiological regulation of body water content. This way
will concern water exchanges, that is, water intakes and
outputs. By use of it, the ontogeny of regulatory responses to

Fig. 1. Rat in restraint frame. Drinking water is available in

removable beaker; urine is shed into funnel. From Adolph,

Barker and Hoy (1954).

excesses and to deficits of water will be traced. We and others
found that at birth the responses whereby constancy of body
water is maintained are small compared to those of older
animals. The several relations involved in this regulation will
be described largely by means of data on laboratory rats.

Water exchanges vary chiefly in the excretion through the
urinary tract and in the drinking into the alimentary tract.
They are measured upon a rat confined to a frame (Fig. 1).
The urinary bladder is reflexly emptied when the rat and
frame are raised and lowered, whereupon the urine enters the

4 E. F. Adolph

funnel and a tube held beneath it. Drink is taken from the
beaker, which can be freed from the frame and weighed at
intervals. The weight of the body, ascertained while the rat
is in the frame, measures any net change of body water
content, including evaporative losses.

When an adult rat has been forcibly given an excess of body
water, it promptly excretes water more rapidly than usual.
The urine flow varies linearly with the water excess present
in the body, as is shown when one plots the first hour's output

^^S,^^

RAT

H 4

INGESTIVE

/ URINARY

§n

OXIDATIVE --^

^-T-xz:z

/ EVAPORATIVE + FECAL

? -10

-4 -2 0 +2 +4 +6

WATER LOAD, PERCENT OF BODY WEIGHT

•10

Fig. 2. Equilibration diagram for water exchanges of adult

rat. Constructed from data of Adolph (1956) and Adolph,

Barker and Hoy (1954).

of urine after water is forced into the stomach in relation to
the amount of water excess or load (Fig. 2). When the rat
has been dehydrated by being deprived of water for various
periods of time, water is drunk as soon as allowed, and the
amount drunk is roughly proportional to the water deficit or
negative load. Excretion and ingestion are symmetrical
activities that specifically and appropriately compensate for
the disturbances of water content (Adolph, 1943). Many
tests seem to show that the accuracies of compensation by
drinking and by excreting are about equal when the water
loads are of equal magnitudes.

Physiological Regulation of Water Content 5

The relations of exchange to content shown in Fig. 2, the
equiUbration diagram, form a useful basis for understanding
the regulation of body water, and of many other body con-
tents. They show the specificity of the responses required
for constancy, the sensitivities with which they occur, their
promptness and their accuracy. A fixed set of relations,
therefore, automatically keeps the rat in water balance.
Similar relations have been worked out for a number of
other species among mammals, other vertebrates, and some

Bladder.

B/adder Wall ■

VisceraC Peritoneum

Parietal PerUoneum

Red us Muscle

Fa I

SAin

Flanc

~V\lire

Fig. 3. Bladder cannula and its method of placement in
infant rat. From Hoy and Adolph (1956).

invertebrates (Adolph, 1943). Much effort has also been
expended by investigators to find through what messages and
effectors the adult's automatic responses are excited and
mediated; those features will be largely neglected here.

Are these relations also present in young animals, and when?
Are they the same as in adults? This question we tried to
answer particularly for water excretion, and first for newborn
dogs (Adolph, 1943, p. 267). For rats we needed an accurate
method for measuring urine flow at all ages, and eventually
found it through placement of a plastic cannula in the bladder
(Fig. 3). Urine is thereafter collected by exserting a capillary
glass tube on the cannula, and measuring the position of the

6 E. F. Adolph

meniscus from minute to minute as urine collects in it (Hoy
and Adolph, 1956). Quantitative collections can also be made
without the cannula, at the urinary papilla or by bladder
puncture; during rapid urine flows these collections give the
same results as with the cannula (Heller, 1947; McCance and
Wilkinson, 1947; Falk, 1955).

Water excess, administered by stomach tube, gives rise to
very little diuresis at birth (Fig. 4). In the course of several

Fig. 4. Water diuresis at various ages in infant rats. Points
show mean and standard error at end of each period of urine
collection. DA = days after conception. From Falk (1955).

days the rat's response increases, until at about ten days
after birth the response per unit of body weight is of adult
size. The ages indicated on the graphs shown here are
reckoned from conception instead of from birth, the average
gestation time for rats being 21*3 days. Actually in human
infants the maturation of the diuresis was found by Ames
(1953) to be triggered by birth rather than by scheduled age,
since prematures acquired the diuretic response about as soon
after birth as postmatures did.

A familiar notion about the way in which water diuresis is
excited is to suppose that the neurohypophysis withholds its

Physiological Regulation of Water Content 7

antidiuretic hormone until the water excess is removed. This
theory is widely accepted for mammals generally. In infant
rats above five days of postnatal age we found that water
diuresis was inhibited by injecting pitressin (Fig. 5). But at
two days of postnatal age the diuresis was unabated by this

HOURS

Fig. 5. Water diuresis at two different
ages in infant rats (dash lines), and the
effects of pitressin injections at P upon
it (sohd Unes). DA = days after con-
ception. From Adolph (1957).

substance. It is unlikely that the foreign pitressin is in-
activated at one age and not at another, and possible that
infant renal tissues are insensitive to it (Heller, 1952). But
the most important conclusion is that diuresis can be aroused
by some other means than the withholding of the hormone in
the neurohypophysis. At this particular age of two days a

8 E. F. Adolph

response is thus uncovered which is mediated through some
other channel ordinarily masked by the known hormonal one.
The intensity of diuresis is a function of the water excess
at all ages (Fig. 6), but the regression differs with age. Actually
these data supply part of an equilibration diagram for infant
rats, and by it one can watch the regulatory relations coming
to maturity during early postnatal life. The unexcreted water
has been located as excess in plasma and several other tissues.
A possible theory of maturation is that some slowly develop-
ing process or structure limits the rate of water excretion.

WATER LOAD. 7o OF WT.

Fig. 6. Water exchange in urine in relation to body water
load at each of three different ages. DA = days after con-
ception. From Adolph (1957).

This theory is doubtful, since at every age still greater water
excess arouses faster excretion. Rather, the response, ex-
pressed by the ratio between excretory rate and water load,
is small at birth and becomes greater as age increases.

However, in order to see whether diuresis is impossible at
birth, we tested the capacity of the infant rat to respond to
several other stimuli of diuresis. To concentrated salt solu-
tions the diuretic response is practically nil at birth, and it
matures even later than the water diuresis (Fig. 7). Hypoxia
arouses a primary diuresis that is small at birth and becomes
greater a few days later ; it also, however, arouses a secondary

Physiological Regulation of Water Content 9

diuresis that is large and sudden even a few hours after birth.
Likewise, adrenahne or noradrenaUne induces a full-blown
diuresis on the very day of birth. Evidently the capacity for
excreting water at a high rate is present, but its arousal
depends on the particular form of stimulation. Consequently,
any discussion of structural inadequacies or functional im-
maturities seems beside the present main point, which is
that the specific responding system of the newborn rat is not
tuned to water excesses.

Hence, we are privileged to see a physiological regulation
increase in intensity in the growing individual. The regulation

'--"" '^^^ — ^--

■ q <

2fe

y "7 //

/ / /

-§y

/ /

en 0=

/ y

"i / y

' \-

/ ^^''

,^-'' DAYS OF AGE

0 10 20

Fig. 7. Courses of development of four types of diuresis in
rats. B = birth. From Hoy and Adolph (1956).

duly materializes, whether the rat has ever experienced a
water excess or not; the elements necessary for it are there,
some of them long before this materialization. What guides
the regulation's intensity and determines its point of adult
fixation is unknown. The fixation is still subject to a small
degree of adaptation resulting from previous exposure to
water excesses (Adolph, 1956).

The control of water intake, on the other hand, is much less
understood than the control of water elimination. In early
infancy, rats, like dogs (iVdolph, 1943, p. 267), refuse to drink
water, even after dehydration. According to K?ecek, Kfec-
kova and Dlouha (1956), as late as 28 days after birth young

E. F. Adolph

rats drink more milk than water in recovering from dehydra-
tion. But in the same circumstance they drink more water
than sahne. Even newborn rats distinguish between milk and
other fluids; at 17 postnatal days they distinguish between
water and salt solutions. Such sensory discriminations are
necessary before rats can link their intakes to specific de-
ficiencies of bodily constituents. The actual tying of water
drinking to water deficiency does not certainly occur until

1000

0.01 0.1

100 1000

Fig. 8. Relation of log water content to
log body weight in rats from foetus to
adult. B = birth. Numbers represent
exponents in parabolic equation relating
the two quantities. Data of Hamilton and
Dewar (1938), from Adolph (1957).

28 days after birth (Krecek, Kfeckova and Dlouha, 1956).
Already then the water intake of rats equals the water deficit
imposed upon them (Adolph, Barker and Hoy, 1954, fig. 13);
just as in the adults, the one-hour intake closely matches the
water deficit so long as the water deficit does not exceed six
per cent of the body weight.

Once the immediate regulations of water content are fixed,
the adult method of maintaining water balance is persistently
at work. But it is well recognized that the water content,

Physiological Regulation of Water Content 11

both absolute (body size) and relative to body solids, varies
with the age of the rat (Fig. 8). What controls the absolute
content of water and of each solute? The answer to this ques-
tion is not available. Obviously all the items that enter the
determination of growth and its correlatives participate in
these controls. This is a problem that has barely been
visualized, and one whose analysis may occupy many physio-
logists in the future.

In general, the ready corrections of water excesses and
deficits result from specific response systems for diuresis and
for water drinking. The systems vary between infant and
adult, not only quantitatively but possibly also in the medi-
ators and effectors used. Over a long lifetime, the regulation
depends also upon detectors of body size and proportions
whose characteristics and locations have not been determined.

REFERENCES

Adolph, E. F. (1943). Physiological Regulations. Lancaster: Cattell.

Adolph, E. F. (1956). Amer. J. Physiol, 184, 18.

Adolph, E. F. (1957). Quart. Rev. Biol., 32, 89.

Adolph, E. F., Barker, J. P., and Hoy, P. A. (1954). Amer. J. Physiol.,

178, 538.
Ames, R. G. (1953). Pediatrics, Springfield, 12, 272.
Falk, G. (1955). Amer. J. Physiol, 181, 157.
Hamilton, B., and Dewar, M. M. (1938). Growth, 2, 13.
Heller, H. (1947). J. Physiol, 106, 245.
Heller, H. (1952). J. Endocrin., 8, 214.

Hoy, p. a., and Adolph, E. F. (1956). Amer. J. Physiol, 187, 32.
Krecek, J., Kreckova, J. and Dlouha, H. (1956). Physiol Bohemo-

slov., 5, suppl., p. 33.
McCance, R. a., and Wilkinson, E. (1947). J. Physiol, 106, 256.

DISCUSSION

Shock: We have obtained some data in our laboratory on the age
differences in the antidiuretic response to pitressin. Some of the results
of these experiments are in accord with the concept that in many in-
stances the senescent animal returns to a type of response and behaviour
that is seen during the course of development. In these experiments we
measured the concentrating ability of the kidney rather than total urine
flows. Total urine flows are not useful for age comparisons since in older
subjects the number of functioning units is reduced and hence there is a

12 Discussion

lower total urine output. Our results are expressed in terms of the amount
of water reabsorbed from the glomerular filtrate, that is the urine /plasma
(U/P) ratio of inulin. A maximum water diuresis was induced by the
oral administration of water plus an intravenous infusion of 5 per cent
glucose. There were three groups of subjects — young, middle-aged and
old. The young group represents nine individuals aged 26-45, the middle-
aged group ten subjects from 46-65 years old, and the old group was
from 66-90. Under conditions of maximum diuresis the U/P ratio was
about 10 for all three groups of subjects. We gave 0-5 m-u./kg. body
weight of pitressin, not enough to cause a rise in blood pressure, but there
was a marked inhibition of the diuresis. The U/P ratio in the young group
increased to 120 within 10 minutes as compared to 75 in the middle-aged
and about 40 for the old. After a period of roughly 50 minutes the diur-
esis was again re-established in all three groups (Miller, J. II., and Shock,
N. W. (1953). J. Geront, 8, 446) (see Shock, Fig. 10, p. 240).

Heller: In connexion with your results. Prof. Adolph, I should like to
clear up a point which has led to some misunderstanding. Some years
ago (Heller, H. (1952). J. Endocrin., 8, 214) we were also interested in
the response of newborn and infant rats to vasopressin. Our experiments
were not suitable for establishing at w hat time after birth the rats first
responded to vasopressin. But we could determine by means of inulin
U/P ratios, i.e. by the same technique as that used by Dr. Shock in
man, at what postnatal age the antidiuretic response to vasopressin
became quantitatively comparable to the response of adult animals.
We found that this occurred only in rats older than 22 days. I would like
to stress this because some workers have misinterpreted these results:
they assumed that we had tried to show that a significant inhibition
occurred /or the first time after 22 days. I think that one must expect that
this datum of around 20 days may change somewhat in the hands of
other workers. Clearly a comparison between the antidiuretic responses
of adult and infant rats depends on the choice and strictness of appli-
cation of the criteria of comparison. But I think that our data agree with
some work which Dr. Falk did later (1955. Amer. J. Physiol., 181, 157).
She injected nicotine into infant rats and tried to find out at what post-
natal age sufficient vasopressin was secreted by the pituitary gland to
produce an inhibition of diuresis which would be quantitatively com-
parable to that in adult animals. She found that this occurred at about
17-22 days after birth.

Adolph: I think Dr. Falk (1955) got a significant inhibition consider-
ably before 17 days. She also injected vasopressin itself, and by the
method of collecting the urine which is expelled in response to perineal
stimulation in the infant rat, she was able to get significant inhibition in
the first week of postnatal life. There is evidence that antidiuretic
hormone or something comparable which could inhibit water diuresis
was then being put out by the animal.

Heller: This is precisely the misunderstanding to which I have been
referring. Dr. Falk did get responses to nicotine in animals three days
after birth, so you are quite right in saying that responses were obtained
much earlier than after 20 days of postnatal life. But she also compared

Discussion 18

the response of older animals with that of adults: they became com-
parable in quantitative terms only when the rats were 17-22 days old.

There is another point on which I should like to have your view s, Prof.
Adolph. We find that these responses of infant rats to vasopressin are
influenced not only by the age of the animals, but also by the litter size.
In other words, if there are fewer animals in the litter, they will be larger,
and that may influence the development of renal functions.

Adolph : We have not tested for litter size. In general we have used the
larger animals.

Black: I must apologize for introducing another hormone, but Prof.
Adolph's interesting observation reminded me of some recent work on
hypertonic over-hydration by INIcCance and Widdowson (1957. Acta
Paediat., (Uppsala), 46, 337). W^e may be tacitly assuming that in these
poor responses we are dealing with either renal immaturity or with this
very interesting hypotension, and I wondered whether the adrenal gland
came into this at all, since its histology changes very considerably from
foetal to neonatal life. Could a better water diuresis be obtained in these
newborn animals by giving them cortisone with the water load?

Adolph: Dr. Falk did some work on the administration of the cortical
adrenal substances. At the early ages these seem to have very little
effect on water diuresis and water excretion.

Swyer : I cannot speak about the rat, but so far as the human is con-
cerned the evidence seems to be that the infant adrenal is quite effective
in secreting glucocorticoids and probably aldosterone, at least in amounts
relative to its own size, so that the apparently deficient response of the
kidney does not appear to be due to lack of adrenal steroids. You cannot
improve the renal response by giving steroids. It might be a lack of renal
responsiveness to the water load rather than any insufficiency of hor-
monal equipment.

Heller: We have found (Heller, H. (1958). Mschr. Kinderheilk., 106,
81) that injections of cortisone into newborn or infant rats produce a
significant decrease of total bodv water. Much the same effect is obtained
with ACTH.

Adolph : I should like to make a small protest against the use of the
term 'renal immaturity'. If you want the 100-day-old rat to be the
criterion of everything, then everything else is either premature or
postmature. But if you want to consider that every animal has an opti-
mum for its own age, then the use of the word immaturity seems to me
undesirable. The same thing applies to hypotension: what is hypo-
tension for an adult is not hypotension for an infant.

Talbot: I should like to register a mild objection to this thesis about
immaturity. For instance one might say that the parathyroid-renal
phosphorus homeostatic mechanism of the human infant is at least
functionally immature at birth, presumably because the mother's
mechanisms have performed this homeostatic task for the infant while
it was in utero. As a result, the infant has a very small tolerance for
dietary phosphorus at birth. However, he develops the capacity to
handle phosphorus satisfactorily within a few weeks.

Have you any further information about this adrenaline-induced

14 Discussion

diuresis? Did it increase the ratio of water to solutes in the urine, or did
it increase the solute output?

Adolph: Adrenaline diuresis in infant rats does involve more solute
output than the water diuresis, but adrenaline diuresis is a water
diuresis in that the urine is very dilute. I do not think you could blame
all the adrenaline diuresis on the solute output itself.

With regard to immaturity and whether it takes experience for an
animal to have a diuresis, we can point to the fact that adrenaline diure-
sis has no experience-factor. We have tried to see whether we could get
more water diuresis in the infant animal by subjecting it to water loads
on successive days. There is a considerable variation in the amount of
water excretion which is produced, and we are unable to say that there
is any significant change due to previous experience with water. Our
provisional conclusion is that there is no adaptation apparent in the
animal subjected to repeated water-loading.

CELLULAR ASPECTS OF THE ELECTROLYTES
AND WATER IN BODY FLUIDS

Hugh Davson

Medical Research Council, Department of Physiology,
University College, London

The water and electrolyte contents of a complex organism
are almost entirely determined by the activities of the kidneys,
which operate primarily on the blood plasma and, through
that, on the extracellular fluid of the organism. Casual
fluctuations in the water and electrolyte contents of the
organism are therefore usually the consequence of fluctuations
in the composition of these two compartments of the body
plasma and extracellular fluid. The electrolytes and water of
the cells of the body are affected secondarily to these primary
fluctuations in the composition of the extracellular fluid and
plasma, and, for practical purposes at any rate, the factors
that can influence them primarily are usually ignored. Never-
theless, since the cells occupy a considerable fraction of the
total volume of the organism, and since there must be some
reciprocity between the electrolyte and water content of cells
and extracellular fluid, it is of some importance that we
understand the physical and chemical factors that determine
the electrolyte concentrations and volumes of the cells of the
body.

The Gibbs-Donnan Equilibrium. The application of the
Gibbs-Donnan equilibrium to the problem of the water and
electrolyte distribution between the plasma and extracellular
fluid is familiar to all who have concerned themselves with
the water balance of the organism. It will be recalled that the
most important consequence of the Gibbs-Donnan distribu-
tion of ions between the two fluids separated by the capillary
membrane that is supposed to be impermeable to the protein
molecules of plasma, is that the osmolarity of the plasma is

16 Hugh Davson

significantly higher than that of the extracellular fluid. This
is illustrated by Fig. 1, and it follows that an equilibrium will
only be achieved when a counter-pressure is exerted on the
plasma equal to the colloid osmotic pressure due to the plasma
proteins. The amount of this difference of osmotic pressure is
determined by the concentration and degree of dissociation
of the proteins. Because of the high molecular weights of the
plasma proteins, their concentration, expressed as moles per
litre, is small and the difference of osmotic pressure that must
be resisted, if the system is to remain stable, is correspond-
ingly small, namely 25 mm. Hg. As a result, the organism is
able to maintain a statistical equilibrium between plasma and
extracellular fluid by virtue of the capillary pressure; at the

Plasma Membrane Extracellular

Fluid

Na+ P~
Na+ CI-

Na+ CI-

Fig. 1. The plasma-extracellular fluid

system.

(P=protein).

arterial end of the capillary the pressure is greater than this
difference of osmotic pressure so that fluid flows into the
extracellular compartment; at the venous end the reverse
holds, and fluid is absorbed.

It is worth noting that by the term "impermeability"
to a solute — here the plasma proteins — we do not necessarily
mean an absolute barrier; this is an ideal case on which cal-
culations are based, but practically it seems unlikely that a
natural membrane is completely impermeable to any of the
naturally occurring molecules in solution in the fluids, and it is
sufficient for our purposes if by "impermeability" is meant
that the rate of transport of this solute across the membrane
is negligibly small compared with that of the other molecules
that we are considering — in the particular case of plasma and
exti-acellular fluid, the salts and water.

The cell membrane is a more selective barrier than the

Cellular Aspects of Body Electrolytes and Water 17

capillary endothelium, and is capable of imposing restrictions
on the movements of ions that are very much smaller than the
protein ions; as a result, it is conceivable that much larger
differences of osmotic pressure could be established, since
these smaller ions may be present in vastly higher concen-
trations than those of proteins with their large molecular
weights. Let us consider the erythrocyte; for simplicity we
may choose the cat or dog erythrocyte which shows no
accumulation of potassium. The distribution of ions is
indicated roughly in Fig. 2; the cell contains the protein
haemoglobin which behaves as an anion, so that we may
expect to be able to apply the Gibbs-Donnan equilibrium to
the diffusible ions. If the Na+, Cl~ and HCOg" ions could
diffuse across the membrane, the position would be entirely

Cell Membrane Plasma

Na+ Hb-

Na+ CI-

Fig. 2. The cat erythrocyte.
(Hb=haemoglobin).

analogous with that already considered, and the contents of
the cell would have a higher osmolarity than the surrounding
plasma, so that unless the membrane could resist the expan-
sion caused by an influx of water, the cell would have to swell,
and swell indefinitely since this difference of osmolarity must
prevail so long as the cell contains a higher protein concentra-
tion than that in the outside medium. Cell membranes are
not strong and would certainly not be able to resist the dif-
ference of osmotic pressure that would be developed, which
in this case would be several times higher than in the case
considered earlier, owing to the very high concentration of
protein in the red cell. We know that the cat erythrocyte is
stable, and we must ask: how? Theoretically, stability could
be achieved by making the membrane impermeable to salts,
i.e. to all the ions of the system. Alternatively, stability could
be achieved by making the cell permeable to anions only and

18 Hugh Davson

impermeable to cations such as Na+ and K+. In this way the
cell would be able to fulfil its function in the maintenance of
the acid-base balance of the body, permitting the Cl~ — HCOg"
exchange that mediates the buffer action of haemoglobin in
the cell.

It might be thought that by making the cell impermeable to cations,
such as Na +, we should be establishing conditions for a Gibbs-Donnan
equilibrium leading to a large excess of osmotic pressure ; however, the
concentrations of impermeable cations will be equal on both sides of the
membrane, so that any Donnan effect due to impermeable cations on
one side of the membrane will be counterbalanced by an equal effect due
to impermeable ions on the other side.

It is easy to show that an osmotic equilibrium between the
inside and outside of the cell is possible, in spite of the high
concentration of indiffusible protein anions within the cell;
thus the impermeability of the cell membrane to cations such
as Na+ confers on it a stability that would be lacking in the
presence of a permeability to this ion; in other words, the
colloid osmotic pressure of the cellular proteins can only
operate in the presence of a permeability to both Na+ and
anions. It is now well known, however, that cell membranes
do not show an absolute impermeability to such ions as Na+
or K+; the use of isotopes has permitted the demonstration of
an unequivocal exchange of these ions across the erythrocyte
membrane. The exchanges are very slow compared with the
exchanges of Cl~ and HCOg", but they do occur, so that we
must expect a constant movement of NaCl and NaHCOg into
the cell, associated with the migration of water, unless some
process prevents this. As is well known, the process that does
prevent it is an active transport of Na+ ions out of the cell;
the membrane is permeable to Na+ so that there is a continual
drift of this ion into the cell because of the demands of the
Gibbs-Donnan distribution, but by some process not under-
stood, metabolic energy of the cell is employed in driving the
salt out. Practically, in consequence, the cell may be des-
cribed as a cell impermeable to Na+ and therefore in stable
equilibrium with its environment. The total amounts of
water and electrolytes within the cell will be determined by

Cellular Aspects of Body Electrolytes and Water 19

two main factors — the osmolarity of the plasma and the
activity of this Na+-extrusion mechanism. The passage of
water across the cell membrane is very rapid, so that the cell
responds to changes in osmolarity of the plasma by virtually
instantaneous changes in its water content; in this way it
may be said to respond passively to changes in the plasma,
and its changes of water content and salt concentration may
be said to be secondary to primary changes determined
principally by the kidney. The operation of the second factor
— the Na+-extrusion mechanism — will influence the amount
of material — salts and water — in the cell, and it would be by
virtue of this mechanism that this type of cell could exert a
primary influence on the water and electrolyte content of the
organism. Thus, if the Na+-extrusion mechanism operated
more rapidly than the influx under the electrochemical
gradient, there would be a net loss of Na+ and of anions,
namely Cl~ and HCOg" ; this would decrease the osmolarity
of the cell and water would be lost to the plasma. Such a
shrinkage of cells is easily demonstrable by allowing them to
recover from the effects of putting the Na+-extrusion mechan-
ism out of action. Thus, when the cells are cooled, the metabolic
processes supplying energy can no longer work; Na+ enters
the cells accompanied by anions and they swxll. When the cells
are warmed, the metabolic processes begin, and the extra Na+
is excreted until the cells return to their normal volume. The
effects of agents that increase the permeability of the cell
membrane are of some interest; substances like alcohol or
urethane, in the appropriate concentration, can increase the
permeability of the cell membrane to Na+ and K+ to such an
extent that the Na+-extrusion mechanism is unable to keep
pace with the influx of this ion; thus, in spite of a normally
functioning metabolism the cell may swell; on removing the
agent it may return to its normal size.

The erythrocytes of most species contain K+ as their
principal cation, so that the cell maintains large gradients
of Na+ and K+ (Fig. 3). The condition for an osmotically
stable system could be given by an impermeability of cations,

20 Hugh Davson

as before, but once again studies with isotopes have shown
that both Na+ and K+ can pass across the membrane and an
active transport of Na+ out of the cell and of K+ into the cell
must be postulated to account for the osmotic stability of the
system.

It was considered at one time that a mere extrusion of Na + would
account for the osmotic stabiHty and high concentration of K + in the
cell, i.e. that the extrusion of Na+ would demand a replacement by K+.
It was pointed out, however (Davson, 1951), that this would lead simply
to an excretion of NaCl and NaHCOg from the cell, with a resultant
shrinkage. Extrusion of Na + will only lead to accumulation of K + if
exchange of K+ for Na+ is obligatory on the system in order to pre-
serve electrical neutrality. If anions can accompany the excreted Na +
then exchange for K + is not obligatory. In nerve and muscle, where the
concentration of non-permeable anions in the cell is very high, such a
sodium-excreting mechanism would cause accumulation of K +.

Cell Membrane Plasma

K+ Hb- Na+ Cl-

K+Cl-

FiG. 3. The human erythrocyte.
( Hb = haemoglobin).

Once again, the water content of such a system will be
determined by the osmolarity of the plasma and the activity
of the metabolic ionic pumps; thus, over-activity of the Na+-
excreting mechanism would lead to a shrinkage; over-
activity of the K+-accumulating mechanism would lead to a
swelling. It is interesting that the two processes show some
degree of linkage, in that Harris (1954) has shown that
accumulation of the one ion is associated with a nearly
equivalent excretion of the other; the linkage is not complete,
however, since on cooling erythz'ocytes swell as a result of
gaining more Na+ than they lose K+; when they are re- warmed
the extra Na+ is excreted and they return to their original
volume. The fact that the cell maintains its characteristic
water content and proportions of Na+ to K+ within fairly
narrow limits indicates that there is some homeostatic
mechanism controlling the rates of accumulation of K+ and

Cellular Aspects of Body Electrolytes and Water 21

excretion of Na+. The mechanism is not known; presumably
the active transport processes are sensitive to the concen-
trations of Na+ and K+, or more probably to the relative
proportions of these ions, in the cell.

The erythrocyte is a highly specialized cell, and it would
not be correct to assume that all cells of the body, or even the
majority, are based on a similar physiological plan so far as the
maintenance of salt and water content is concerned. The
striated muscle fibre has been studied very thoroughly, and it
may well be that this is far nearer to being a "typical cell",
so that we may now consider its main features from the pre-
sent point of view. The main point of difference between the
muscle cell and the erythrocyte lies in the low contents of Gl-
and HCO3-, these anions being replaced by organic anions

Fibre Membrane Extracellular

Fluid

K+ A

Na+ CI

Fig. 4. The muscle fibre.
(A~^indiffusible organic anions).

that apparently cannot diffuse across the plasma membrane;
schematically the situation is as in Fig. 4 where A" represents
these indiffusible anions. The system would be osmotically
stable were the membrane impermeable to Na+, i.e. the rest
of the ions, K+, CI", HCO3-, would distribute themselves
across the membrane in such a way that equal osmotic
activities would exist on both sides. Actually the cell mem-
brane is permeable to Na+, and the reason why the Na+, K+
and CI- ions do not redistribute themselves is because an
active extrusion of Na+, as fast as it penetrates, maintains an
effective impermeability to Na+. There is no need to postulate
an active accumulation of K+ in this case since, owing to the
high concentration of impermeable anions in the cell, the
extrusion of a Na+ ion must be associated with the penetra-
tion of a K+ ion, in the interests of electrical neutrality. Once
again, then, the cell may maintain equilibrium with its

22 Hugh Davson

environment, provided that an ion-excreting mechanism is
active. Loss of this, by cooUng the tissue or by metabohc
poisons, causes a loss of K+ and a gain of Na+, CI" and HCO3-,
the net effect being an increase in osmolarity with a consequent
swelling of the cells. Re-warming of the tissue may cause a
reversal of these changes (see, for example, Steinbach, 1954).

Thus, in all of the cell types that we have considered, the
system can be treated, theoretically at least, as a system that
maintains an osmotic equilibrium between the interior and
external fluids by virtue of an "effective impermeability" to
one or more ionic types ; if the membrane were truly imperme-
able to the ions in question, the osmotic equilibrium would
be independent of metabolic processes and could be described
as a true equilibrium ; in practice, the effective impermeability
is the result of a continuous process of active transport.
For the purposes of mathematical description this is equivalent
to an impermeability, at any rate under normal conditions;
under abnormal conditions, on the other hand, the precarious-
ness or instability of the equilibrium is shown by the cellular
oedema that follows either the failure of the ion-excreting
mechanism or such a large increase in the permeability of the
membrane that the mechanism can no longer keep pace with
the influx of Na+.

If these considerations are correct, we may expect to find
that by adding up the total osmolarities inside and outside
the cell the two totals should be equal within the limits of
experimental error. Probably the muscle fibre has been
studied most carefully from this aspect, and it would seem
from Conway's (1957) figures (Table I), that osmotic equili-
brium does exist between the cell and its environment. The
same is probably true of the erythrocyte and the nerve fibre,
but it must be remembered that the analytical techniques for
all the constituents of the cell are not so accurate that a dif-
ference of one or two per cent would be ascertained. Within
this limit, then, it seems quite safe to affirm that these cells
are in osmotic equilibrium with their environment.

Within recent years the possibility that mammalian cells

Cellular Aspects of Body Electrolytes and Water 23

are not in osmotic equilibrium with their extracellular fluid
has been seriously maintained, and an "osmotic pump",
driving water continuously out of the cell, has been postu-
lated. The experimental basis for this claim rests on the
observation that mammalian tissue slices, in particular
those of liver and kidney, swell when placed in "isotonic"
solutions of sodium chloride, Tyrode or Krebs (Sperry and

nON OF FROG MUSCLE AND

able I

PLASMA

EXPRESSED

AS M-MOLE PEl

H2O (after Conway,

, 1957)

Fibre

Plasma

Concentration

124

2-25

10-4 (3-6)*

109

4-9

140

1-25

1-5

77-5

HCO3

12-4

26-6

Phosphate

7-3

3-3

Sulphate

0-4

Phosphocreatine

35-2

—

Carnosine

14-7

NHa-acids

8-8

7-2

Creatine

7-4

2-2

Lactate

3-9

3-5

Adenosine triphosphate

—

Hexose monophosphate

2-5

—

Ghicose

—

Protein

0-6

2-2

Urea

Total

248-2

245-3

♦ Figure in brackets for sodium represents, according to Conway, the true intracellular con-
centration.

Brand, 1939; Opie, 1949), either at room temperature or at
0°. Robinson (1952) observed that the swelling could be
prevented or reversed by maintaining the tissue at 37°; he
found also that the swelling occurred in the presence of cyanide
at this temperature. Since swelling was prevented by using
strongly hypertonic solutions — 0-55-0 -60 m — he concluded
that the cells were iso-osmotic with these. It will be quite
clear from what has been said earlier that these facts may be
explained just as easily on the assumption that the electrolyte-

24 Hugh Davson

excreting system fails at low temperature or in the presence
of cyanide. Thus, soaking a muscle at 0° certainly leads to
swelling, but this is completely accounted for by the gain of
Na+ and Cl~; warming the muscle causes an excretion of
these ions and it returns to its original volume. The same
argument will apply to other tissues, and conclusive proof
that this is the principal explanation for the changes taking
place on cooling was provided by the elegant experiments of
Deyrup (1953) who showed that if the tissues were bathed in
iso-osmotic sucrose (0 • 3 m) they failed to swell. If the swelling
in Ringer solution had been due to a failure of a water-excret-
ing mechanism, substitution of salt for sucrose should have
had no effect, whereas if the swelling had been due primarily
to a penetration of NaCl, substitution of a non-penetrating
substance like sucrose would have prevented it. It seems
safe to conclude, then, that very large differences of osmolarity
between cell contents and their environment, such as those
postulated by Opie (1949) and Robinson (1952), do not occur.
The detection of smaller differences, that would demand a
water pump continuously excreting water from the cell to
maintain an osmotic steady state between cells and their
environment, must rely on very precise measurements of
osmolarity.

The depression of freezing point has been employed by a
number of workers with a view mainly to testing the claim
that mammalian tissues were hypertonic to plasma (Conway
and McCormack, 1953; Opie, 1954; Brodsky et al, 1953, 1956;
Conway, Geoghegan and McCormack, 1955 ; Itoh and Schwartz,
1956); but, as Conway's studies indicate, the interpretation
of the results is not easy, since an excised tissue, when ground
up at 0°, undergoes autolytic changes — in particular the
breakdown of adenosine triphosphate to inosinic acid,
ammonia and phosphate — that lead to a considerable increase
in osmolarity. It would seem from Conway's studies that
within the limits of accuracy of the cryoscopic method —
probably a few per cent — the tissue cells examined — liver,
kidney and muscle — are iso-osmotic with their environment.

Cellular Aspects of Body Electrolytes and Water 25

This does not mean, however, that the maintenance of
differences of osmotic pressure between cells and their environ-
ment by the excretion of water does not occur; it is well
known that such fluids as urine and saliva have osmolarities
that are vastly different from that of the plasma; and the
elaboration of these fluids is best described by invoking an
active transport of w^ater — i.e. the functioning of a 'Svater

Table II

Concentrations of ions (M-MOLE/kg.

H3O)

IN PLASMA,

AQUEOUS HUMOUR

AND CEREBROSPINAL FLUID OF THE RABBIT

Plasma

Aqueous

Humour

Na 151-5

CI 108

143-5

109-5

K 5-5

HCO3 27-4

5-5

HCO3

33-6

Ca 2-6

Lactate 7 • 9

2-3

Lactate

6-00

Mg 10

Phosphate 1-8

IMg

0-85

Phosphate
Ascorbate

100
100

Total 160-6

Total 145 • 1

Total 152 1

Total

151-1

Cations and Anions 305-7 Cations and Anions 303-3

Cerebrospinal Fluid

151

CI 129

3-5

HCO3 31-4

1-3

Lactate 2-6

0-8

Phosphate 0 - 5

Total

156-6

Total 163-5

Cations and Anions 320 • 1

pump". The cerebrospinal fluid would appear to represent
another example of a non-iso-osmotic fluid, and since it is in
such close relationship with the nervous tissue of the brain and
spinal cord, this lack of iso-osmolarity is of special interest,
suggesting as it does that these tissues, too, are not in osmotic
equilibrium with the blood. The results of a detailed analysis
of the ionic concentrations in plasma and cerebrospinal
fluid are shown in Table II: included are values for a similar

26 Hugh Davson

type of fluid, the aqueous humour — similar because both are
speciahzed tissue fluids fifling cavities and being virtually
free from protein. By summing the cations and anions it
becomes clear that the cerebrospinal fluid has a higher con-
centration than the plasma or the aqueous humour ; allowance
must be made for the lower concentrations of glucose and urea
in the cerebrospinal fluid, a difference amounting to some
5 m-mole; thus the cerebrospinal fluid is hyperosmotic by
some 9 m-mole. The amount is small — some 3 per cent — never-
theless it represents a diff*erence of osmotic pressure of some
160 mm. Hg, and it is presumably because the fluid is able to
drain away easily from its cavities that this pressure does not
develop, i.e. the difference in osmolarity is reflected in a
continuous influx of water from the blood rather than in the
development of a pressure, such as would happen were the
system completely closed. However, the really significant
point to be made in this connexion is that the cerebrospinal
fluid lies in such close relationship with the brain and cord
that it seems most unlikely, having regard to the rapidity
with which water may exchange between the two, that a
diff'erence of osmolarity could be maintained. That is, if the
cerebrospinal fluid is, indeed, hypertonic to plasma, then so
must the tissue of the brain and cord be. If this is true,
then we may postulate one of two things: either a water
pump that drives water out of the nerve cells into the
extracellular fluid where it passes back into the blood;
or alternatively the elaboration, by the capillaries of the
nervous tissue, of a hyperosmotic extracellular fluid. The
capillaries in this region of the body are certainly different
from those in the rest of the body and are responsible, pre-
sumably, for the so-called "blood-brain barrier"; to attribute
secretory activity to their endothelium is by no means an
unreasonable proposition. The important point to be made
here is that the diff'erence of osmolarity is small and thus
requires highly accurate analysis for its demonstration. Why
the cerebrospinal fluid and nervous tissue should have this
higher osmolarity is not clear; according to Flexner (1938),

Cellular Aspects of Body Electrolytes and Water 27

the high concentration of chloride in the cerebrospinal fluid,
which may be taken as a measure of this hyperosmolarity,
appears at an early stage in development — at about 40 days
in fact. It may be that the positive pressure of the cerebro-
spinal fluid depends for its maintenance on a difference of
osmotic pressure between it and the blood.

The factors determining the water and electrolyte contents
of connective tissue are probably simple, although they have
not been studied in great detail. If a piece of collagen, or
collagen plus mucoid, is placed in a saline medium, equivalent
to extracellular fluid, we may expect a Gibbs-Donnan equili-
brium to be established between this and the medium by

Na+ Cl-

Na+ Coll

Na+ CI

Na+ CI-

Fig. 5. Illustrating Gibbs-Donnan equilibrium
between collagen and extracellular fluid. In
this case there is no membrane separating
the two, the collagenous gel being a separate
phase.

virtue of the acidic nature of the protein and mucoid. The
situation might therefore be as in Fig. 5, i.e. essentially
similar to that obtaining with plasma separated by a mem-
brane from extracellular fluid. There is no membrane separat-
ing the two, however, and separation is maintained because
of a phase difference, the collagen-mucoid system being a gel,
the extracellular fluid a liquid. Chemical analysis of con-
nective tissue shows that there is, indeed, a Gibbs-Donnan
distribution of ions between it and plasma and therefore,
presumably, between it and extracellular fluid, the concen-
tration of chloride being less, and that of sodium greater, in
the connective tissue. There is, in consequence, a tendency
for water to pass into the connective tissue phase, the salts

28 Hugh Davson

continuously redistributing themselves so that the osmotic
pressure of this phase is greater than that of extracellular
fluid and of blood. The extent to which the system will
take up water will depend on the counter-pressure that can
be exerted or, failing that, what is really equivalent, the
mechanical rigidity of the system that will oppose distention.
Presumably in such tissues as tendon and skin the structural
rigidity of the system prevents an indefinite uptake of water,
and the system is stabilized with a water content of about
75 per cent. In the cornea of the eye, however, the situation

Table III

Comparison of eyes maintained

AT NORMAL AND

LOW

CORNEAL

temperatures

(Davson,

1955)

Water Content

Expt.

Temp.

Time

no.

(hr.)

(^./lOO g.
tissue)

(g-lg. solid)

82-8

4-8

—

77-2

3-4

82-8

4-8

—

770

3-35

821

4-65

—

78-2

3-6

78-5

3-65

—

77-8

3-5

is different; it consists, essentially, of a number of laminae of
collagen-plus-mucoid, sandwiched between two cellular layers,
the epithelium and endothelium. If the eye is excised and
stored in the cold, say at 4°, the cornea increases in water
content, due to absorption of aqueous humour. If instead of
being kept at 4° the eye is maintained at about 31° — the
normal temperature of the cornea — the tissue retains its
normal water content (Table III). It would seem, then, that
metabolic activity is preventing the collagen plus mucoid
from absorbing water and salts from the aqueous humour, and
this may be proved by first allowing the cornea to swell at the

Cellular Aspects of Body Electrolytes and Water 29

low temperature and then transferring the eye to a chamber
maintained at the higher temperature. In this case the ab-
sorbed water and salts are excreted back and the cornea
reacquires its normal hydration (Table IV). The secretory
activity that usually maintains the cornea in its normal state
of hydration — about 75 per cent water — may be due to both
the endothelium and epithelium, but whether it is due to an
active excretion of salt, e.g. sodium, or of water, remains to
be proved. The extraordinary tendency of the cornea to
take up water, by contrast with tendon or sclera, is presumably

Table IV
The effect of subsequent warming on eyes maintained for 15-18

HOURS AT 7°

(Davson, 1955)

Column A gives the water content after the period at 7° ; column B the water

content after a further period of 6-8 hours at 31°.

Water content

Expt.

no .

(g'Ig-

solid)

Change

(%)
24

{A)
4-35

(B)
3-3

4-45

3-7

4-65

3-9

related to the large quantity of mucoid present as a coating
over the individual collagen fibrils (Schwarz, 1953), and it
seems likely that changes in hydration are really the conse-
quence of changes in hydration of this colloid, the collagen
fibrils being pushed apart by the swelling. The Gibbs-
Donnan sweHing of the collagen-mucoid system of skin and
subcutaneous tissues may well be a factor in determining the
water content and the turgescence of the tissues. Thus it
would seem from McMaster's (1946) studies that the extra-
cellular fluid may, in normal circumstances, be something of
an abstraction, the space between cells and collagen fibrils
being occupied by a mucoid gel ; only when excessive amounts
of fluid are filtered from the plasma, or under experimental

30 Hugh Davson

conditions of injection of fluid into the tissue, is it possible to
speak of free fluid in the extracellular spaces. The nature of
the collagen and mucoid in these tissues may therefore exert
some effect on the water content of the tissues. In general, it
would seem that acute changes in this tissue extracellular
water are the result of changed factors of capillary filtration
and reabsorption, but it may well be that the long-term
steady-state level is influenced by the amount of mucoid in the
tissue. This presumably exerts its Gibbs-Donnan difference
of osmotic pressure, drawing fluid to it; the tendency is
opposed by the structural rigidity of the tissue, so that a
steady state is established, in contrast to the cornea where the
rigidity of the system is inadequate to permit a steady state,
a continuous secretory activity being necessary, and made
possible by the presence of cellular membranes lining the
tissue.

The possible ways in which the water compartments of the
body may be altered with age become evident from this
general review; thus, the activity of the ion-transporting
mechanisms of the cells tends to oppose a normal tendency
to cell oedema, with the result that a steady state is main-
tained with the cells having a characteristic ionic make-up and
percentage of water. A decrease in the metabolic activity of
the cells may be expected to result in the penetration of salt
and water into the cells; hyperactivity, on the other hand,
may cause a shrinkage of the cells, but the extent of this will
be limited by the demands of electrical neutrality; excessive
excretion of the Na+ ion must be associated with excretion of
some anion or with accumulation of K+; in the latter event
there will be no change in osmolarity, whilst the former process
is limited by the availability of diffusible anions. It seems
unlikely that a cellular dehydration could result from hyper-
activity of this sort, and it seems more likely that dehydration
of cells might be due to a loss of the indiffusible anions,
collectively indicated as A" in Fig. 4, but actually consisting
of proteins, organic phosphates, etc. If these were replaced by
such diff'usible anions as CI" and HCO3-, then the process of

Cellular Aspects of Body Electrolytes and Water 31

extrusion of Na+ would lead to an elimination of these ions
and it could well be that a new steady state would be estab-
lished at a lower level of internal K+ and Na+ concentrations.
Unfortunately, practically nothing is known of the factors
that control the normal activity of the salt-excreting system
of the cell.

The large differences in the amount of extracellular water
that take place with age may be, to some extent, associated
with differences in the amount of water per cell of the organ-
ism; thus, other things being equal, a decrease in cellular
water is reflected in a rise in the extracellular water, expressed
as a percentage. To prove this, however, it would be necessary
to measure not so much the percentage water in the cells as
the amount of water per cell, and this might be attempted by
relating the water to the deoxyribonucleic acid content of the
tissue. It seems more likely, however, that long-term fluctu-
ations in the fractions of intra- and extracellular water,
especially those taking place during development, will be
determined by changes in the number of cells in unit weight
of tissue rather than in changes of their size, and this could be
achieved by (a) multiplication or reduction of the number of
cells; (b) expansion or contraction of the extracellular space,
by changes in the quantity of connective tissue and in the
ability of this to hold fluid.

REFERENCES

Brodsky, W. a., Appelboom, J. W., Dennis, W. H., Rehm, W. S.,

MiLEY, J. F., and Diamond, I. (1956). J. gen. Physiol., 40, 183.
Brodsky, W. A., Rehm, W. S., and McIntosh, B. J. (1953). J. din.

Invest., 32, 556.
Conway, E. J. (1957). Physiol. Rev., 37, 84.
Conway, E. J., Geoghegan, H., and McCormack, J. I. (1955). J.

Physiol., 130, 427.
Conway, E. J., and McCormack, J. I. (1953). J. Physiol., 120, 1.
Davson, H. (1951). Textbook of General Physiology, p. 276. London:

Churchill.
Davson, H. (1955). Biochem. J., 59, 24.
Deyrup, I. (1953). J. gen. Physiol., 36, 739.
Flexner, L. F. (1938). Amer. J. Physiol., 124, 131.
Harris, E. J. (1954). Symp. Soc. exp. Biol., 8, 228.

32 Hugh Davson

Itoh, S., and Schwartz, I. L. (1956). J. gen. Physiol, 40, 171.

McMaster, p. D. (1946). Ann. N.Y. Acad. Sci., 46, 743.

Opie, E. L. (1949). J. exp. Med., 89, 185.

Opie, E. L. (1954). J. exp. Med., 99, 29.

Robinson, J. R. (1952). Proc. roy. Soc, 140 B, 135.

ScHWARZ, W. (1953). Z. Zellforsch., 38, 26.

Sperry, W. M., and Brand, F. C. (1939). Proc. Soc. exp. Biol, N.Y., 42,

147.
Steinbach, H. B. (1954). Symp. Soc. exp. Biol., 8, 438.

DISCUSSION

Talbot : I was most interested, Dr. Davson, in your comments about the
cellular oedema that occurs in 'sick' cells. It has been shown that ani-
mals deprived of potassium, and thereby subjected to a combination of
cellular potassium insufficiency and cellular sodium intoxication, show a
tendency to cellular oedema. We therefore wondered whether loss of
potassium from the cell was a factor which might interfere with its
sodium and water pump mechanisms.

Davson : We still do not really know what makes a cell stop accumu-
lating. Accumulation may be a matter of the development of some
anions inside the cell at the same time as the development of a process of
excreting the sodium. But if you get rid of sodium, something has got to
come in and it may be potassium. That eventually leads to the develop-
ment of more of these ions and to a condition in which there is a high
potassium concentration inside, and low sodium and chloride. When we
allow the system to cool or give it poison, then we find that sodium comes
in and potassium goes out ; but when we warm it up again the whole thing
reverses and we get back to the original state of affairs. Whether it is
that the cell will stop with a given potassium concentration ratio, or a
given concentration of sodium, or at a given size, we do not really know
for certain. In potassium deficiency, according to the papers I read rather
a long time ago, one found that potassium was substituted for by sodium.

Talbot: That is if sodium is available.

Davson : So you propose a condition where there is a sodium as well as
a potassium deficiency?

Talbot: You could have simple deprivation with loss of cellular potas-
sium, but without entrance of sodium in any appreciable amount. There
you have a relatively benign situation. When you superimpose cellular
sodium intoxication, things really begin to get mixed up. How do you fit
that in with your very interesting observations?

Davson : It is really a matter of thinking these things out as separate
problems as they arise, and there has been no systematic investigation of
this. We still have no idea of the mechanism of sodium excretion, and
what makes it stop. If one did know more, one would be able to fit in
the results with the general physiology of the organism.

Fejfar: In clinical medicine we now accept that active sodium trans-
port and potassium deficiency are very important factors. We assume,
when we analyse a muscle biopsy specimen, that we will get a good

Discussion 33

representative sample of what is going on in the organism. Is tliis a fair
assumption? A second point is that most of the work has been done on
kidney shoes. Are kidney shoes representative of the whole organism, or
only of the kidney tissue?

Davson: I was thinking in terms of the tissues that I have worked with
— not the kidney, but musoles and red oells. As far as I oan see, the
results of the work on the kidney cortex are essentially similar to those
obtained on the muscle. However there might be a confusing situation if
you got a lump of kidney tissue with fairly intact tubules as well as not so
intact tubules ; they could be accumulating sodium and indulging in their
special secretory processes which are quite different from those in nmscle.
I have never looked with any approval on work done with slices of these
specialized tissues.

Fejfar : Quite a lot of work has been done with kidney in Prague by Cort
and Kleinzeller (1956. J. Physiol., 133, 287) and that is why I asked you.
They support an active mechanism for sodium and passive mechanisms
for potassium and chloride.

Davson: The active accumulation of potassium by most of the cells
that have been studied has not had to be specifically invoked. It is almost
an unnecessary hypothesis for muscle, but on the other hand one finds
that the active transport of sodium is linked with that of potassium. If
one is an active process, the other must be too. From the responses to
changes of environment, one must say that potassium is following its
gradients of electrochemical potential. On the other hand, when the
matter is studied with isotopes and it is found out just how much sodium
is going in, it is seen that there is a linkage between the amount of sod-
ium crossing the membrane and the amount of potassium. It is not a
rigid linkage, however.

Fejfar: Cort and Kleinzeller find that the amount of potassium crossing
the membrane is usually smaller than the amount of sodium.

Davson : Yes, there is a 2 : 1 ratio. In the inuscle it is a certain propor-
tion, and in the red cell it is a different proportion. Certainly in the red
cell an active accumulation of potassium as well as of sodium has to be
invoked.

Fejfar: Roguski in Poland claims that one can judge general cellular
metabolism from the red cells themselves. We do not agree because the
red cell is not a respiring cell. Neubauer (personal communication) has
made a comparison of the water and electrolyte changes between muscle
biopsy specimens and red cells, and he could not find any similarity be-
tween them. He came to the conclusion that you could not judge
electrolyte changes from the red cell.

Davson: That is quite true. The mammalian red cell metabolism is
different; it is largely anaerobic, whereas the muscle and all the other
cells are mainly aerobic.

Fejfar: I was surprised to hear you say that when cells are poisoned
there is not only an influx of sodium, but also of chloride. We were
taught that chloride does not usually enter cells in significant amounts
and that only sodium does this, so we judge the extracellular fluid by the
cliloride present.

AGEING — IV— 2 33

34 Discussion

Davson : That would be a most dangerous conclusion to draw. If your
chloride space altered under experimental conditions, it could very well
be due to penetration of chloride into the cells.

Wallace : We have been working with tissues for some time from the
standpoint of hydrogen ion gradients between cells and extracellular
fluid. I have often discussed this work with investigators interested in
single cells and the events that occur within the cell. One often finds that
such workers are unwilling to accept the interpretations derived from
analytical values for whole tissues. They point out that the interior of the
cell is not homogeneous. Potassium and sodium do not appear to be
evenly distributed and the hydrogen ion concentration seems to vary
from locus to locus. I am certainly not ready to give up the study of
ions and their distribution in tissues, but I think one must always bear in
mind that membrane equilibria can only tell a part of the story. The
concept of the cell, particularly the muscle cell, as an "empty bag"
cannot be completely accepted.

Davson : In general I am in favour of your iconoclastic approach, but
you are basing most of your argument on the findings of the electron
microscopists and they are by no means above criticism themselves. They
are working on fixed tissue and talk about their endoplasmic reticulum.
It certainly appears as a most complicated system of canals, but one
wonders how real it is. Is one to abandon all hope of applying rather
elementary physical chemistry to our problems just because of these
complexities? We think of the cell as being bounded by a limiting mem-
brane with certain permeability characteristics. The electron micro-
scopists show us the membrane which does exist, but then they find little
holes or vesicles just next door to it. They say that what is happening is
that the membrane is opening up, the vesicle is coming in and they have
caughtit just as it was coming in. It may well be that they are right. We
have obviously got to be suspicious of treating things too simply — there
you are absolutely right. On the other hand, I am not willing to stop
applying elementary physical chemistry to problems of salt transfer just
because of these complexities.

Adolph : I should like to add something to the point about swelling and
shrinking with the accompanying transfers of electrolytes. When tissue
slices, not only kidney slices but also liver slices, and two tissues which
we did not have to slice, i.e. diaphragm and auricle, are transferred from
low temperature to high, or from anoxic media to oxygen, they shrink.
This shrinking in high temperature and oxygen is fully reversible any
number of times; for instance, in ten-minute periods, in low temperature
or in high, in nitrogen and in oxygen, we can get complete reversibility of
the swelling and shrinking (Adolph, E. F., and Richmond, J. (1956).
Amer. J. Physiol., 187, 437). This indicates that there is no permanent
damage to these tissues from the swelling and shrinking, and it also indi-
cates that the transfers are very rapid. It looks as though, if there is
electrolyte transfer, it is as rapid as that of water. But I am not con-
vinced that the electrolyte transfers are necessary for this swelling and
shrinking. We have no method of measuring the speed of the electrolyte
transfers, but we have a method of measuring that of the water transfers.

Discussion 35

Maybe someone can furnish data which will be more convincing on
whether the electrolyte transfers are equally rapid and reproducible.

Davson: I think the electrolyte transfer is very likely to be much
slower and to hold up the whole process. The water transfer is very
rapid in every cell, so I would say that what happens first is the move-
ment of the electrolyte and the movement of water would not require
much time. The evidence I am citing is largely based on work from Prof.
Conway's laboratory.

Hingerty : One of the main experimental difficulties, of course, is in
maintaining the normal condition of the cells. When you remove tissues
from an animal there is a very rapid increase in molecular concentration
in the cells due to breakdown of molecules such as glycogen, hexose esters,
phosphocreatine and adenosine triphosphate (Conway, E. J., Geoghegan,
H., and McCormack, J. (1955). J. Physiol., 130, -427). If you remove the
tissue directly into liquid oxygen, grind to a frozen powder and then take
a series of freezing point depressions on this frozen tissue maintained at 0°,
extrapolation back to zero time gives a value equal to that obtained for
the plasma (Conway, E. J., and McCormack, J. (195S). J. Physiol., 120, 1).
This certainly held for liver, kidney and muscle tissue of the rat and it
would be interesting to see these techniques applied to other tissues.

The swelling of the cells in anoxic conditions cannot be due to a failure
to pump out water, since the freezing point depressions of respiring and
non-respiring kidney slices are the same, and the effect of anoxia may be
interpreted as being due rather to cessation of the sodium pump. Break-
down of molecules may be partly responsible for the swelling but the main
effect appears to be caused by sodium and chloride entering the cell (some
potassium leaving), and water then entering to preserve osmotic balance
(Conway, E. J., and Geoghegan, H. (1955). J. Physiol., 130, 438).

HYPERNATRAEMIA AND HYPONATRAEMIA

WITH SPECIAL REFERENCE TO

CEREBRAL DISTURBANCES

Paul Fourman and Patricia M. Leeson

Medical Unit, Royal Infirmary, Cardiff

Introduction

An abnormal concentration of sodium in the extracellular
fluid often presents a puzzling problem for the clinician. As
is well known, a change in the total amount of the sodium or
of the water in the body can explain many instances — water
deficiency or sodium excess producing hypernatraemia,
water excess or sodium deficiency producing hyponatraemia.
But many cases appear to require more than a simple account
of gains and losses to explain them. Is this because a simple
explanation, such as a change in the amount of water in the
body, has been overlooked, or must one in such cases invoke
some new mechanism, possibly under the control of the
nervous system?

There have been a number of reports of "cerebral" hyper-
natraemia and hyponatraemia (Knowles, 1956; Edelman,
1956). With regard to hypernatraemia it seems likely that
some of the contradictions in the present views (Welt et al.,
1952; Higgins et al., 1954) might have been avoided, for in
hardly any of the patients reported could a frank water
deficiency confidently be excluded from the information
supplied. This question is discussed in the first section. The
subject of hyponatraemia seems much more difficult, but if
sodium deficiency is excluded, many of the remaining cases
can be accounted for by an abnormal retention of water
diluting the body fluids. In the second section we present
some new data on the problem, derived from a study of two
patients.

Hypernatraemia and Hyponatraemia 37

Before discussing the subject in more detail it may be
helpful to recall some of the factors which regulate the water
content of the body.

Regulation of water

Two mechanisms, closely linked, normally guard against
water depletion. One regulates the intake of water through
the sensation of thirst, the other the output of water through
the secretion of antidiuretic hormone. There are at least two
ways in which each may be invoked: the first, a rise in the
tonicity, the second, less well known, a fall in the volume of
the body fluids (Smith, 1957; Strauss, 1957).

A rise in the sodium content of the extracellular fluid
(ECF) is well known to produce thirst and to stimulate the
release of antidiuretic hormone (ADH). The effective stimulus
is not simply the rise in ECF tonicity : if the ECF tonicity is
raised with a substance like urea, which diffuses freely across
the cell membrane and raises the tonicity of both extracel-
lular and cellular fluid equally, this does not stimulate thirst
and antidiuresis to the same extent (Gilman, 1937). When,
however, the extracellular tonicity is raised by a substance
which does not diffuse into the cells, water leaves the cells
until the tonicity of extracellular fluid and cells are again
equal. The cells shrink. It is assumed that certain cells in the
hypothalamus respond to shrinking and stimulate the sensa-
tion of thirst and the liberation of ADH.

For the release of ADH there is much evidence that there
are localized receptors of this kind (Jewell and Verney, 1957;
Verney, 1957). Recently Andersson (1957) has also provided
additional evidence for a thirst centre. He found that when
he stimulated a certain area of the hypothalamus in goats,
they drank water as long as the stimulus went on, even to the
point of haemolysing their own red cells. With destructive
lesions in the same region, the goats would not drink water
when they obviously needed it. The thirst centre and the
receptors of the ADH mechanism are very close together,
but probably distinct.

38 Paul Fourman and Patricia M. Leeson

The position of these centres in the nervous system suggests
that their control involves more than a response to changes
in tonicity, and some purely nervous stimuli such as pain and
emotion may initiate, or inhibit, thirst or antidiuresis.

A fall in the volume of the ECF can stimulate thirst and
antidiuresis, presumably through nervous pathways (see
Rosenbaum, 1957; Strauss, 1957). Smith (1957) has dis-
cussed at length where the receptors for the stimulus to anti-
diuresis might be : some of them may be in the left auricle of
the heart (Henry and Pearce, 1956).

Hypernatraemia
Water deficiency

Normally, thirst and antidiuresis are stimulated by a very
small increase in extracellular tonicity, less than two per cent
(Wolf, 1950; Verney, 1957). A concentration of sodium ([Na])
in the plasma exceeding 150 m-equiv./l. may certainly be
regarded as abnormal. In a study of water deficiency pro-
duced experimentally in dogs, values of 160, and in one
animal that died a value of 186 m-equiv./l., were found
(Elkinton and Taffel, 1942); in a man made water-deficient
by Black, McCance and Young (1944) the [Na] rose to 160
m-equiv./l. In a patient from Texas reported by Gordon and
Goldner (1957) a value as high as 192 m-equiv./l. was reported.
He recovered.

The "dehydration reaction". The hypernatraemia of
water deficiency is not simply the result of the blood becoming
more concentrated, for in spite of the high blood level of
sodium there may be very little sodium in the urine; it is
retained in the body.

Allott (1939), who first drew attention to the problem of
hypernatraemia, found the urinary [Na] ranged from 2-5 to
9 m-equiv./l. in four of his patients. It now seems most likely
these low concentrations of sodium were a result of the
"dehydration reaction" first described by Peters (1948, 1952).
The mechanism of this reaction is not clear, though it appears
to be a renal response to a fall in blood volume; in this con-

Hypernatraemia and Hyponatraemia 39

nexion it may be recalled that two of Allott's patients had
had an alimentary haemorrhage.

It is partly through neglect of this phenomenon that some
authors have been led to place cases of hypernatraemia with a
low urinary sodium in a separate group.

Symptoms of water deficiency. There are several reasons
why authors describing neurogenic or cerebral hypernatraemia
may have overlooked a water deficiency. Though they often
state that there is no cUnical evidence of dehydration in their
patients (e.g. Cooper and Crevier, 1952), this does not in fact
mean very much. The word dehydration is used for two
clinical states : one of water deficiency alone, and the other of
salt deficiency which generally also entails a loss of water.
This usage implies that these deficiencies produce a similar
clinical picture, though it was made clear long ago that this
is not so (Kerpel-Fronius, 1935 ; Nadal, Pedersen and Maddock,
1941). Water deficiency is not clinically obvious unless it is
extreme, because the deficit is distributed throughout the
body water. In salt deficiency, on the other hand, the extra-
cellular fluid, though but a third of the total in volume, bears
the whole of the deficit; it is patients with the latter who have
the haggard look, the sunken eyes, the small pulse and low
blood pressure of dehydration. Patients with simple water
deficiency are ill, but there are no specific signs of the defici-
ency, the tongue may even be moist, and it is not obvious it is
water they lack. If in addition, as a result of a craniotomy
their faces are oedematous, it may even be mistakenly as-
sumed that they have accumulated water in excess. The
diagnostic difficulties are increased because, particularly in
older patients, some of the most striking symptoms of water
deficiency are cerebral rather than vascular, for instance
drowsiness and confusion, and disturbances of behaviour,
which can mimic a lesion of the frontal lobes. These symptoms
make it more difficult to give water; but they can be com-
pletely reversed with water.

Losses of water. Abnormally large losses of water may
go unrecognized. Extrarenal losses may be larger than is

40 Paul Fourman and Patricia M. Leeson

generally assumed ; and a good urinary output does not neces-
sarily mean there is no deficit of water, for it may represent
failure of conservation. In the unconscious or helpless patient
the intake depends on the physician's instructions and the
nurses' care. If the intake is less than the combined losses from
the skin, the lungs and the bowels, there must be a deficit of
water in the body and the plasma [Na] will eventually rise.

Some cerebral lesions are associated with a high fever, or
with excessive sweating, or with an abnormally rapid respira-
tion. With any of these the insensible losses of water may
increase from the normal value of some 800 ml. They have
rarely been measured, but in one patient they were thought
to be as much as five litres a day (Gordon and Goldner, 1957).

One expects the volume of urine to be small in water de-
ficiency, and its concentration high. But there are three ways
in which untoward renal losses of water may contribute to
water deficiency: diabetes insipidus from a failure of the
pituitary-hypothalamic mechanism; defective renal func-
tion; and osmotic diuresis. Neither the first nor the second
has always been excluded in cases reported as cerebral hyper-
natraemia. Diabetes insipidus possibly explains cases 1 and 3
of Cooper and Crevier (1952) and one case of Natelson and
Alexander (1955). The force of this explanation is emphas-
ized by a patient reported by Peters (1948), a young woman
whose serum [Na] rose from 140 to 171 m-equiv./l. in 24 hours
following an operation for craniopharyngioma which was com-
plicated by diabetes insipidus. In an incontinent patient a low
concentration of the urine may be the only clue to diabetes
insipidus, and the effect of pitressin should be tried in all
patients with hypernatraemia in whom this possibility exists.

The excretion of a large amount of solutes produces an
osmotic diuresis (McCance, 1945; Hervey, McCance and Tayler
1946; Rapoport et al., 1949). This happens in spite of a water
deficiency (McCance, Young and Black, 1944) and may even
be the cause of it.

Urea, sodium and chloride are the main osmotically active
constituents of the urine. The excretion of urea may be

Hypernatraemia and Hyponatraemia 41

increased by an abnormal breakdown of body protein or by
excessive protein in the diet. One hundred grams of protein
contain 16 g. of nitrogen, excreted as 34 g. or 570 m-osm. of
urea. Ten grams of sodium chloride provide 340 m-osm. It is
not unusual for unconscious patients to receive these amounts
in their feeds ; and their endogenous production of urea may
already be very large (Cooper etal., 1951). The hypernatraemic
patient of Natelson and Alexander (1955) presumably had
an osmotic diuresis when he was made worse with "non-
saline fluids", because these consisted partly of protein
hydrolysate equivalent to 100 g. of protein. In certain
neurological disturbances (Astrup, Gotzche and Neukirch,
1954; Whedon and Shorr, 1957) and in water deficiency itself
(Black, McCance and Young, 1944) the breakdown of body
protein may be greatly accelerated.

To detect a water deficit, the minimum data required are
the estimated intake and output of water and solutes, and the
volume and concentration of the urine. A water deficit is
confirmed if, with the administration of water, the elevated
plasma [Na] falls.

In many of the reports of cerebral hypernatraemia it is
impossible to decide from the data given what the water
balance was. The patients with hypernatraemia of Higgins
and his co-workers (1954) seem to have begun with a deficit of
water of about one litre. Subsequently their intake of water
may have been as little as two litres daily. Their exogenous
osmolar load was about 610 m-osm. We do not know what
was the total excretion; urine volumes and specific gravities
are not stated. The blood urea was high, and fell as the plasma
[Na] fell, when their intake of fluid was increased. In other
reports the data actually show there was a cumulative deficit
of water although the fact may have been disregarded
(Anthonisen, Hilden and Thomsen, 1954; Allott, 1957).

Failure of thirst. Even when losses of water do go un-
recognized by the clinician, there is no danger of water
depletion as long as the patient responds normally with thirst
and is able to drink. For example, in uncomplicated diabetes

42 Paul Fourman and Patricia M. Leeson

insipidus the plasma [Na] is not usually very much raised ; in
a patient of ours, a man of 28 with sarcoidosis, the plasma
[Na] was at times as high as 149 m-equiv./L, but he was then
very thirsty, and he would not tolerate the [Na] rising any
higher. On the other hand, patients who are apathetic, weak,
disorientated or unconscious may be unaware of thirst, or
unable to respond to it. In these patients even normal losses
of water may lead to water deficiency with hypernatraemia.
It is not unusual to have elderly patients with cerebro-
vascular disease who tolerate a plasma [Na] of 150 m-equiv./l.
without any complaint of thirst. But when they are given
water they retain it, and their clinical and biochemical
responses show they had a need for it. We do not know the
possible sites of the lesions which may interfere with the
sensation of thirst in these people. There is, however, some
evidence that in man (Leaf and Mamby, 1952; Engstrom and
Liebman, 1953), as in the rat (Stevenson, Welt and Orloff,
1950) and the goat (Andersson, 1957), neurological lesions
may interfere with the normal sensation of thirst.

We have had the opportunity of studying a boy of ten who
had had a large suprasellar craniopharyngioma removed by
Mr. C. Langmaid. There was no evidence of diabetes insipidus
before the operation. After the operation, however, while he
was in a stuporous state, his plasma [Na] ranged between
152 and 163 m-equiv./l. It remained high even when he
recovered, and was up and about, and receiving pitressin.
The boy did not complain of thirst and we think the lack of
thirst led to water deficiency and hypernatraemia. These are
some of the values before and after he received pitressin : —

Date Plasma sodium Urine vol : ml. Specific

m-equiv.jl. per 24 hours. gravity

Before pitressin 7 Nov. 161 1370

After pitressin 21 Nov. 156 860

25 Nov. 156 1420 1-008

The urine volume and specific gravity while he was having
pitressin suggest the treatment was inadequate, but he did

Hypernatraemia and Hyponatraemia 43

not respond, as does the ordinary case of diabetes insipidus,
with thirst. (He recovered spontaneously from his diabetes
insipidus, and from his hypernatraemia, after three months.)
Although this type of hypernatraemia might be termed cere-
bral, it is in fact a water deficiency due to the breakdown of
one of the mechanisms that normally ensure water balance.

Renal effects of water deficiency. Before leaving the
question of hypernatraemia due to w^ater deficiency it may
be noted that in many of the reported cases the disturbance
apparently produced a disorder of tubular function, mani-
fested by oliguria with isosthenuria or by the excretion of
urine with a high pH in the presence of a systemic acidosis
(Cooper and Crevier, 1952 (Case 4); Gordon and Goldner, 1957;
Allott, 1957). This suggests that severe water deficiency may
be accompanied by tubular damage; Allott (1939) noted a
tubular degeneration in two of his cases post mortem.

A tubular damage would help to explain the acidosis in at
least one of the patients of Higgins and his co-workers (1951).
It is not possible to say with any certainty whether these
patients were water-deficient, but all of them had a high
blood urea and in relation to this the urine volumes were
certainly small. It is also possible that in some patients
(e.g. Allott, 1957) polyuria with hyposthenuria represented
the diuretic phase of a tubular necrosis, itself the result of
dehydration.

To sum up the question of "cerebral" hypernatraemia, a
failure of the thirst mechanism, with or without a diabetes
insipidus, accounts for some of the cases that have been
described; and, as Gordon and Goldner (1957) have ably
illustrated, unrecognized renal or extrarenal losses of fluid
must account for many more.

If, as we believe, cerebral hypernatraemia is the result of
water deficiency then water will correct it, but only if enough
is given. Unfortunately most authors have underestimated
the amount of water required to correct a severe deficit.
Higgins and co-workers (1954) gave up to four litres to the
patients they thought were water-deficient. We give nearly

44 Paul Fourman and Patricia M. Leeson

this amount routinely. Gordon and Goldner gave one of their
two patients 8-24 litres in 24 hours and even this was not
enough to bring down his plasma [Na] to normal. As long as
the plasma [Na] remains high there can be no risk of water
intoxication.

Other forms of hypernatraemia.

It is possible to produce hypernatraemia by giving an
excess of salt (McCance, 1956), though more usually this
produces an isotonic expansion of the extracellular fluid with
oedema.

The homeostatic mechanisms may be so adjusted as to
maintain the plasma [Na] at a high level. In experimental
potassium deficiency the plasma [Na] was over 150 m-equiv./L,
although the absorption of sodium was small and the intake
of water as much as eight litres a day in one subject (Fourman,
1954). Hypernatraemia is often a feature of aldosteronism
(Conn, 1956) but whether or not this is the result of the
associated potassium deficiency cannot be stated. Recently
Zilva and Harris- Jones (1957) have discussed the possibility
of excessive adrenocortical activity producing hypernatraemia
by a shift of sodium from cells to ECF.

Hyponatraemia

We may arbitrarily define hyponatraemia as a plasma [Na]
lower than 180 m-equiv./l. It is obvious the concentration of
sodium in the plasma may fall because of a reduction in the
total amount of sodium in the ECF or because of an increase
in the amount of water.

Salt deficiency

A reduction of the total amount of sodium in the ECF is
the result of sodium deficiency.

We have already emphasized that the clinical effects of
sodium deficiency are easily recognizable. Lack of salt is un-
likely to arise unless, through sweating, vomiting, diarrhoea
or fistulous discharge, sodium is lost from the body, because

Hypernatraemia and Hyponatraemia 45

the kidneys normally conserve sodium efficiently. For the
same reason, in sodium deficiency there is virtually no
sodium in the urine. To this there is one exception, namely,
when the sodium deficit is actually the result of continued
loss through the kidney. This happens, of course, in Addison's
disease, and in "salt-losing nephritis". Furthermore, in
certain patients with cerebral lesions persistent renal losses
have been observed, even when the intake of sodium is much
reduced (Welt et at., 1952). The renal defect has been ascribed
to a loss of neural impulses affecting proximal tubular func-
tion (Cort, 1954). But the patient of Merrill, Murray and
Harrison (1956) with malignant hypertension was able to
maintain a normal sodium balance when his own kidneys
were replaced by a kidney which was transplanted from his
twin brother and therefore deprived of its nerve supply. It
does not seem then that a loss of nervous impulses is alone
responsible for a failure of the kidneys to conserve salt,
though the renal nerves do play a part in the response to salt
deprivation (Bricker et al., 1956) and to anoxia (Foldi,
Kovach and Takacs, 1955a, h). The mechanism of the defect
in "cerebral salt-wasting" remains obscure. Water excess
(see below) may produce a renal loss of sodium, and some
instances of so-called salt wasting may therefore be examples
of water retention.

Hyponatraemia from salt deficiency can, of course, be
corrected with salt.

A deficiency of sodium, producing hyponatraemia, can arise
without a loss of sodium from the body. The sudden accumul-
ation of a transudate in some part of the body produces a
relative lack of salt and water. If only water is provided the
[Na] falls. This state of affairs is seen most clearly after a
paracentesis, w^hen water, carrying sodium with it, may rapidly
reaccumulate in the abdominal cavity. The fall in blood
volume presumably stimulates thirst and the liberation of
ADH; for the patient, while drinking copiously, produces
only a small amount of concentrated urine containing very
little sodium (Nelson, Rosenbaum and Strauss, 1951).

46 Paul Fourman and Patricia M. Leeson

Water excess

Water excess is a well recognized cause of hyponatraemia
when patients are given too much water while unable to
excrete it at the normal rate (Wynn, 1956). This may happen
in renal failure, in adrenal and pituitary insufficiency, and
postoperatively, particularly after mitral valvotomy (Bruce
et at., 1955). Hyponatraemia from this cause is usually obvious
from the circumstances. Such patients may have no symp-
toms; sometimes they have the syndrome of water intoxica-
tion, with fits and other profound neurological disturbances.
They may have hypertension; they certainly do not have
hypotension. The face looks bloated, not drawn.

Both sodium deficiency and simple water excess respond to
the administration of hypertonic saline with a rise in the
plasma [Na] to normal which is subsequently maintained.

There remains for consideration a large group of cases
where the hyponatraemia does not produce symptoms and its
mechanism is obscure. Elkinton (1956) and McCrory and
Macaulay (1957) have recently reviewed this problem. The
hyponatraemia appears to be associated with an expanded
volume of ECF; and the kidneys do not excrete water or
retain sodium to bring back the tonicity of the plasma to
normal (Leaf and Mamby, 1952).

There are at least two possible explanations. The first is
that there is an abnormal stimulus to antidiuresis, say from
the "volume receptors", operating through the secretion of
ADH or in some other way (Kleeman et al., 1955; Ginsburg
and Brown, 1957). Pitressin given experimentally to normal
people leads to a retention of water, a fall in the plasma [Na]
and eventually an increased renal loss of sodium in spite of
the low plasma [Na] (Leaf et al., 1953; Weston et al., 1953;
Wrong, 1956).

The second possibility is that an abnormal hypotonicity of
the cells determines the hypotonicity of the ECF (Sims et al.,
1950; Rapoport, West and Brodsky, 1951).

McCrory and Macaulay (1957) described an infant with
diffuse cerebral damage and hyponatraemia. Her ECF

Hypernatraemia and Hyponatraemia 47

volume was greater than normal. The infant did not excrete a
dose of water at the normal rate and the authors thought she
was secreting an excess of ADH. An excessive secretion of
ADH would, of course, be appropriate only to a restricted
fluid intake. When her fluid intake was restricted the plasma
[Na] rose to normal.

Schwartz and co-workers (1957) have recently suggested
that an inappropriate secretion of ADH might account for
the hyponatraemia in two patients with carcinoma of the
bronchus whom they studied. They imply that there was an
abnormal stimulation of the receptors for maintaining the
volume of the body fluids. Their patients had normal renal
and adrenal function; they excreted a normal amount of
aldosterone. In one of them the plasma [Na] fell as low as
103 m-equiv./l., but the extracellular volume, far from being
reduced as in sodium deficiency, was expanded and there
was no evidence of peripheral vascular failure. The urine
was generally hypertonic to the plasma, and this is the
principal argument adduced by Schwartz and co-workers
that these patients were producing too much ADH. The
kidneys of these patients did not conserve sodium when their
fluid intake was unrestricted, though they did so when
large amounts of salt-retaining steroids were given. Schwartz
and co-workers do not comment on the rate of excretion of a
dose of water. But there is no doubt the kidneys did respond
normally to water deprivation. Under this stimulus the
urinary sodium fell and the plasma [Na] rose. Others have
also described this response to water deprivation in hypo-
natraemia (see Edelman, 1956). It might be interpreted as
the usual "dehydration reaction".

Some observations we have made on two patients with
unexplained hyponatraemia are relevant.

Case reports

Albert, aged 62, was admitted on 25th May 1957 in status epilepticus
accompanied by hyperpjTcxia and heavy sweating. He had been up
and about until then, although he had had a right hemiparesis for two
years, which had become worse two months before admission. His

48 Paul Fourman and Patricia M. Leeson

blood pressure was 180/80. His fits were rapidly controlled, but he then
had a bilateral spastic paralysis with extensor plantar responses, and
never regained consciousness. On the second day he stopped breathing
and respiration had to be maintained with a Beaver respirator for 12
hours. Subsequently he had a purulent bronchopneumonia and on the
fourth day a tracheotomy was done to enable a clear airway to be
maintained by suction. The bladder was kept drained by a Foley
catheter but the urine was not infected until the last days of his iMness.
He died on 11th August of bronchopneumonia.

At post-mortem there was a large area of softening in the left temporal
lobe. The vessels of the circle of Willis were very atheromatous. There
was evidence of an earlier hypertension ; the left ventricle was hyper-
trophied to a thickness of 22 mm. compared to 8 mm. in the right
ventricle, and the kidneys showed hypertensive changes. There was
remarkably little evidence of infection in them although there was a
purulent cystitis.

Albert was certainly water- deficient in the early days of his illness.
His extrarenal losses of water were large, and for the first three days his
total intake was only two litres. On 29th May his plasma [Na] was 137
m-equiv./l. but at the same time the volume of the packed cells in his
blood was 55 per cent. He was then given six litres of water in two days ;
the packed cell volume fell to 41 per cent and the plasma [Na] fell to
128 m-equiv./l. Subsequently his plasma [Na] fluctuated between 130
and 110 m-equiv./l. The blood urea was 34 mg. per 100 ml. and the
creatinine clearance 70 ml./min.

Ivor, aged 54, was admitted on 11th June 1957 having been ill for 18
days with acute peripheral neuropathy affecting mainly the motor
nerves and accompanied by an enlargement of the liver. The plasma
albumin (2nd July) was 2-9, and the total protein 6 g. per 100 ml.
The cause of his illness was not discovered. In the next five days he
developed a partial respiratory paralysis with bronchopneumonia. His
blood pressure, which had been normal, fell to 90/60. Subsequently
he was fed by tube ; and his purulent bronchial secretion was aspirated
through a tracheostomy. At the end of June he began slowly to recover
and was taking some food by mouth on 4th July, but almost immedi-
ately had a severe relapse. Tube feeding continued until the end of
July, by which time he was able to move his limbs, though they were
still very weak. He subsequently had three relapses and died in Decem-
ber. We have not the details of the latter stages of his illness.

Before he was fed by tube his intake of water was inadequate to cover
his losses, which were augmented by copious sweating associated with
his chest infection, and he must have sustained a considerable deficit of
water and probably of salt. The water deficiency was corrected on 17th
and 18th June by the administration of a total of 8 -9 litres of water, of
which he excreted only 3-5 litres during those two days. Consistent
with a "dehydration reaction", on 17th June his urine contained only
2 m-equiv. sodium in 24 hours. With the correction of his water deficit
his plasma [Na] fell from 133 to 120 m-equiv./l. in 24 hours. In spite

Hypernatraemia and Hyponatraemia 49

of the low plasma [Na], on 19th June he excreted 210 m-equiv. of
sodium in three litres of urine. The plasma sodium remained low, rang-
ing from 109 to 123 m-equiv. /I. until August, when it gradually rose to
133 m-equiv. /I. Except on two occasions, both early in his illness, one
associated with salt deficiency and both with lung infections, he did not
have peripheral vascular failure. His blood urea was 26 mg. per 100 ml.
and the endogenous creatinine clearance 85 ml./min.

The daily feed in these patients consisted of protein, 90 g., fat, 120 g.,
carbohydrates, 120 g., in four litres of fluid. Until 6th July it contained
170 m-equiv. of sodium and thereafter 68 m-equiv., of sodium; the
urinary excretion of sodium fell correspondingly in both patients.

Muscle analysis

The question whether the total sodium content of the body
was low, or normal, but diluted by an excess of water in the
ECF could be settled by an analysis of muscle.

Table I

Analyses of muscle from the two patients, compared

WITH

"normal" values

m-equiv. I kg. fat-free tissue

Water

per cent

Albert

74-4

31-8

45-5

91-2

Ivor

79 1

27-2

47-9

89-3

Talso, Spafford and

Blaw

77-6±0-6

19-l±3-9

33-7±6-4

940±5

•9

(1953)

Wilson (1955)

77-5

25-6±51

40 -.6^6 0

92-3±7

•6

Barnes, Gordon and

Cope

80-3±l-6

231±6-5

43-6±ll

91-3±8

(1957)

Analyses of plasma taken from the two patients at the time

OF THE muscle BIOPSY

m-equiv. jl.

Albert

89-6

124

5-5

Ivor

87-6

124

3-7

The specimens were taken from paralysed muscles in both
patients. The electrolyte contents are shown in Table I,
with "normal" values for specimens taken fx'om anaesthetized

50 Paul Fourman and Patricia M. Leeson

patients. The potassium content was normal. The sodium
content, far from being lower than normal, was in fact at the
upper limits of the normal. The chloride content was simil-
arly high. For this to happen with a low concentration of
sodium in ECF, the amount of ECF in the muscle samples
must have been larger than normal.

Hypertonic saline

The infusion of hypertonic saline produced only a transient
increase in the plasma [Na].

The response was studied in detail in Albert. He had 500 ml.
of 5 per cent sodium chloride (436 m-equiv.) infused over
about three hours on 15th June when his plasma [Na] was
initially 127 m-equiv./l. (Fig. 1).

The immediate response to this infusion was an osmotic
diuresis with an output of 7-3 ml./min. of urine containing
330 m-osm. and 155 m-equiv. of sodium per litre. During
the infusion he excreted 80 m-equiv. of sodium. The plasma
[Na] increased to 143 m-equiv./l. during the infusion and was
138 m-equiv./l. at the end. In the following 21 hours he
responded quite differently. He excreted only 55 m-equiv. of
sodium and his urine flow fell to 0-2 ml./min. with a concen-
tration of 696 m-osm. /I. He was thus retaining water and
diluting the sodium he had retained. Three days later his
plasma [Na] was again only 130 m-equiv./l.

Ivor had infusions of 300 ml. of 5 per cent sodium chloride
on 22nd June and 540 ml. on 24th June. We did not make
very detailed studies of his response, but the plasma [Na]
before and after the second infusion was 113 and 115 m-
equiv./l. During the first three hours of this infusion when
he had received 190 m-equiv. he excreted only 30 m-equiv.
Both the infusions were followed by a retention of water.

These are not the responses one would expect from salt-
depleted patients (Black, Piatt and Stanbury, 1950). They
imply that the osmolality of the body water was being
maintained even at the expense of increasing the volume of
the extracellular fluid. This is the normal response to hyper-

Hypernatraemia and Hyponatraemia

tonic saline (Crawford and Ludemann, 1951; Birchard,
Rosenbaum and Strauss, 1953; Papper et al., 1956), and
depends, of course, on the liberation of ADH (Holland and
Stead, 1951).

500-

NqX«j/fnl
200

lOO-

NoCtm,)

sol

'TTTTT^/Z/Z/kr^-r.-^^.

TIME-

CAM 12 6PM 12

Fig. 1. The changes in total sodium excretion, urine flow,

sodium concentration, and osmotic concentration of the urine

after the infusion of 500 ml. of 5 per cent sodium chloride

(436 m-equiv.).

Water deprivation

When fluid was withheld for 19 hours both patients pro-
duced a urine of small volume and high osmolality (Table II).
The osmolality was not as high as might be expected in normal
people ; but the osmolality of the plasma of both patients was

Albert

Ivor

904

870

0-27

Oil

22-6

14-6

243

267

112-117

112-123

52 Paul Fourman and Patricia M. Leeson

low. The ratio of urine to plasma osmolalities, which can
normally rise to about 4 with water deprivation, was 3-7 in
Albert and 3-3 in Ivor. The deprivation of water was associ-

Table II

Effects of depriving the two patients of water for 19 hours

Changes in urine and plasma
Maximum urine concentration (m-osm/1.)
Flow at maximum concentration (ml./min.)
[Na] ((jL-equiv./ml. of urine)
Plasma (m-osm./l.)
Change in plasma ([Na] m-equiv./l.)

ated with a great reduction in the renal excretion of sodium,
and the increases in the plasma [Na] were unexpectedly
large. They were not maintained however, for the plasma
[Na] had returned to the original levels after 48 hours.

Effect of water

The effect of a water load was adequately tested only in
Albert, who on 15th July received one litre of water in 30
minutes, by stomach tube. He excreted all of this water in
less than three hours, achieving a diuresis of 7-3 ml./min.,
with an osmolal concentration of 54 m-osm./L, and a sodium
concentration of 4 m-equiv./l. These low concentrations are
similar to the minimum values obtained in normal persons
(Schoen, 1957). The values for the plasma sodium before and
after the test were 115 and 112 m-equiv./l. Remarkably low
osmolal concentrations were found twice in the 24-hour
collections of urine from Albert. The values, 153 and 168
m-osm./L, were lower than in the plasma, in spite of the fact
that at these times the plasma [Na] was exceptionally low,
104 m-equiv./l.; these values were obtained on the days
immediately following administration of pitressin (see below).

We did not find any very low urinary concentrations in
eight 24-hour collections from Ivor that were tested. In one
specimen an osinolal concentration of 245 m-osm./l. was the
same as that of the plasma taken at that time.

Hypernatraemia and Hyponatraemia 53

Effect of potassium chloride

In view of Laragh's (1954) findings of a rise in plasma [Na]
with the administration of potassium chloride in patients with
hyponatraemia, we gave 100 m-equiv. of potassium chloride
on two successive days to both the patients. There was no
increase in the plasma [Na] and only a slight rise in the plasma
[K].

The data so far reported show that the renal excretion of
sodium could be made to vary from very small to very large
amounts, and, in particular, although sodium continued to be
excreted while the plasma concentration was low, the kidneys
were able to conserve sodium during the dehydration reaction.
But a rise in the plasma [Na] produced by hypertonic saline
was followed by retention of water which restored the osmo-
lality of the plasma to its original level.

Effect of pitressin

All these results might be taken to show that these patients
had an intact antidiuretic mechanism which operated to
maintain their plasma osmolality at a lower level than normal.
Their response, however, to exogenous ADH given as pitressin
was quite unexpected.

After one litre of water by intragastric drip the patients
received 100 m-u. of pitressin intravenously and 5 i.u. of
pitressin in oil intramuscularly. Urine was collected in hourly
periods for the following five hours; the gastric drip was
running throughout, but the amounts given after the initial
load were unfortunately not recorded. The results are shown
in Table III. In Fig. 2 they are compared with the results of
water deprivation. Ivor began with a concentrated urine,
but after the first hour the maximum osmolality achieved
after pitressin was some 500 or 600 m-osm. less than after
dehydration. The effect of pitressin was tested a second time
in Albert, and he then passed urine with a concentration of
215 m-osm. /I., that is, lower even than his own hypotonic
plasma. The low concentration of urine in these tests de-
pended on the comparatively high urine flow, and not on a

Paul Fourman and Patricia M. Leeson

reduced excretion of solutes. The rate of excretion of sodium
and of solutes was actually higher than with dehydration,
though lower than immediately before the pitressin was

Table III

Effects of pitressin in the two patients while their
hydration was maintained

Changes in urine and plasma Albert Ivor

Maximum urine concentration (m-osm./l.) 280 387*

Flow at maximum concentration (ml./min.) 1-5 2-4

[Na] ([z-equiv./ml. of urine) 33-2 60-2

Plasma (m-osm./l.) 233 243

* The results on the first collection (see Fig. 2) have been neglected.

given. Glomerular filtration rates were not measured. The
same batch of pitressin was shown to have normal activity
in other subjects.

Min

DEHYD^

1////////////77ZZ

PITRESSIN

3^-

DEHYD^

PITRESSIN

"'"^'ff /'^TTTfc-»—

PLASMA

TIME 8 9 10 II 12

7JULY

12 13 14 15 16

ALBERT '"""

9 lO II
7 JULY

2 12 13 14 15 16 VI

IVOR '^""

Fig. 2. Comparison of the changes in the flow and concentration of urine
following deprivation of water and following pitressin and a water load.

The difference between the effects of water deprivation and
pitressin is far greater than anything observed in normal
people (Jones and de Wardener, 1956), and indeed indicates
an almost complete failure to respond to pitressin in the

Hypernatraemia and Hyponatraemia 55

presence of a water load, while the response to water depriva-
tion was nearly normal. Pitressin was not entirely without
effect on the urine flow since this diminished.

The failure of Albert and Ivor to respond to pitressin might
represent the human counterpart of the experiments of
Wesson and co- workers (1950). Their dogs with an isotonic
expansion of the ECF did not respond to pitressin.

We have mentioned that the failure of response was not a
complete one, and it therefore remains possible that the
original expansion of the ECF represented an effect of the
patients' own ADH, as Schwartz and co-workers (1957)
postulated for their two cases. Although Schwartz and co-
workers do not remark on it, there were occasions when their
patient W. A., like Albert, produced a hypotonic urine
following an additional expansion of the ECF. These observa-
tions would be consistent with the suggestion that when the
ECF is expanded beyond a certain point the kidneys become
refractory to the action of ADH.

If we assume that an overproduction of ADH wasreponsible
for the hypotonicity of the ECF in Albert and Ivor, the
alternatives previously suggested still remain, whether the
stimulus to ADH production represented a homeostatic
mechanism for maintaining a hypotonic ECF in two people
who might have had "hypotonic" cells; or whether it repre-
sented a response to an abnormal stimulation of some un-
identified receptor.

Summary

The problem of hypernatraemia seems in general to be one
of water deficiency. That of hyponatraemia is sometimes
one of salt deficiency, but often one of excessive dilution of
the ECF with water. The latter seems to have been the fault
in the two patients we studied. Muscle biopsies revealed
normal or high sodium contents. In their responses to hyper-
tonic saline, water deprivation, and water loading their
homeostatic mechanisms were adjusted to maintain an
abnormally large volume of ECF with low tonicity. Though

56 Paul Fourman and Patricia M. Leeson

they produced a hypertonic urine of low volume when deprived
of water, they did not always produce a hypertonic urine with
pitressin and water. Under certain circumstances, therefore,
the kidney can excrete a hypotonic urine in the presence of
pitressin while retaining its ability to respond normally to
dehydration.

Acknowledgements

We are indebted to Dr. H. E. F. Davies for his help, to Mr. Emlyn
Morgan, Mrs. M. Lewis and Miss M. O. Seabright for technical assistance,
and to Professor Harold Scarborough for his valuable advice.

REFERENCES

Allott, E. N. (1939). Lancet, 1, 1035.

Allott, E. N. (1957). Lancet, 1, 246.

Andersson, B. (1957). In The Neurohypophysis, ed. Heller, H.

London: Butterworth.
Anthonisen, p., Hilden, T., and Thomsen, A. C. (1954). Acta med.

scand., 150, 355.
AsTRUP, P., GOTZCHE, H., and Neukirch, F. (1954). Brit. med. J., 1,

780.
Barnes, B. A., Gordon, E. B., and Cope, O. (1957). J. din. Invest., 36,

1239.
BiRCHARD, W. H., Rosenbaum, J. D., and Strauss, M. B. (1953). J.

appl. Physiol., 6, 22.
Black, D. A. K., McCance, R. A., and Young, W. F. (1944). J. Physiol,

102, 406.
Black, D. A. K., Platt, R., and Stanbury, S. W. (1950). Clin. Sci.,

9, 205.
Bricker, N. S., Guild, W. R., Reardan, J. B., and Merrill, J. P.

(1956). J. din. Invest., 35, 1364.
Bruce, R. A., Merendino, K. A., Dunning, M. F., Scribner, B. H.,
DoNOHUE, D., Carlsen, E., and Cummins, J. (1955). Surg. Gynec.
Obstet., 100, 295.
Conn, J. W. (1956). Arch, intern. Med., 97, 135.
Cooper, I. S., and Crevier, P. H. (1952). J. din. Endocrin. Metah., 12,

821.
Cooper, I. S., Rynearson, E. H., MacCarty, C. S., and Power, M. H.

(1951). J. Neurosurg., 8, 295.
CORT, J. H. (1954). Lancet, 1, 752.

Crawford, B., and Ludemann, H. (1951). J. din. Invest., 30, 1456.
Edelman, I. S. (1956). Metabolism, 5, 500.
Elkinton, J. R. (1956). Circulation, 14, 1027.
Elkinton, J. R., and Taffel, M. (1942). J. din. Invest., 21, 787.
Engstrom, W. W., and Liebman, A. (1953). Amer. J. Med., 15, 180.

Hypernatraemia and Hyponatraemia 57

FoLDi, M., KovACH, A. G. B., and Takacs, L. (1955a). Nature, Lond.,

176, 120.
FoLDi, M., KovACH, A. G. B., and Takacs, L. (19556). Acta med.

Acad. Sci. hung., 8, 19.
FouRMAN, P. (1954). Clin, Sci., 13, 93.
GiLMAN, A. (1937). Amer. J. Physiol., 120, 323.
GiNSBURG, M,, and Brown, L. M. (1957). In The Neurohypophysis, ed.

Heller, H. London : Butterworth.
Gordon, G. L., and Goldner, F. (1957). Amer. J. Med., 23, 543.
Henry, J. P., and Pearce, J. W. (1956). J. Physiol, 131, 572.
Hervey, G. R., McCance, R. A., and Tayler, R. Q. C. (1946). Nature,

Lond., 157, 338.
Higgins, G., Lewin, W., O'Brien, J. R. P., and Taylor, W. H. (1951).

Lancet, 1, 1295.
Higgins, G., Lewin, W., O'Brien, J. R. P., and Taylor, W. H. (1954).

Lancet, 1, 61.
Holland, B. C, and Stead, E. A. (1951). Arch, intern. Med., 88, 571.
Jewell, P. A., and Verney, E. B. (1957). Phil. Trans., 240 B, 197.
Jones, R. V. H., and de Wardener, H. E. (1956). Brit. med. J., 1, 271.
Kerpel-Fronius, E. (1935). Z. Kinderheilk., 57, 489.
Kleeman, C. R., Rubini, M. E., Lambdin, E., and Epstein, F. H.

(1955). J. din. Invest., 34, 448.
Knowles, H. C. (1956). Metabolism, 5, 508.
Laragh, J. H. (1954). J. din. Invest., 33, 807.
Leaf, A., Bartter, F. C, Santos, R. F., and Wrong, O. (1953). J.

din. Invest., 32, 868.
Leaf, A., and Mamby, A. R. (1952). J. din. Invest., 31, 60.
McCance, R. A. (1945). J. Physiol, 104, 196.
McCance, R. A. (1956). Canad. med. Ass. J., 75, 791.
McCance, R. A., Young, W. F., and Black, D. A. K. (1944). J.

Physiol, 102, 415.
McCrory, W. W., and Macaulay, D. (1957). Pediatrics, Springfield,

20, 23.
Merrill, J. P., Murray, J. E., Harrison, J. H., and Guild, W. R.

(1956). J. Amer. med. Ass., 160, 277.
Nadal, J. W., Pedersen, S., and Maddock, W. G. (1941). J. din.

Invest., 20, 691.
Natelson, S., and Alexander, M. O. (1955). Arch, intern. Med., 96, 172.
Nelson, W. P., Rosenbaum, J. D., and Strauss, M. B. (1951). J. din.

Invest., 30, 738.
Papper, S., Saxon, L., Rosenbaum, J. D., and Cohen, H. W. (1956).

J. Lab. din. Med., 47, 776.
Peters, J. P. (1948). New Engl J. Med:, 239, 353.
Peters, J. P. (1952). In Diseases of Metabolism, ed. Duncan, G.G.

Philadelphia: W. B. Saunders.
Rapoport, S., Brodsky, W. A., West, C. D., and Mackler, B. (1949).

Amer. J. Physiol, 156, 433.
Rapoport, S., West, C. D., and Brodsky, W. A. (1951). J. Lab. din,

Med., 37, 550.

58 Paul Fourman and Patricia M. Leeson

RosENBAUM, J. D. (1957). In Essays in Metabolism, ed. Welt, L. G.

Boston: Little, Brown, and Co.
SCHOEN, E. J. (1957). J. appl. Physiol, 10, 267.
Schwartz, W. B., Bennett, W., Curelop, S., and Bartter, F. C.

(1957). Amer. J. Med., 23, 529.
Sims, E. A. H., Welt, L. G., Orloff, J., and Needham, J. W. (1950).

J. din. Invest., 29, 1545.
Smith, H. W. (1957). Amer. J. Med., 23, 623.
Stevenson, J. A. F., Welt, L. G., and Orloff, J. (1950). Amer. J.

Physiol, 161, 35.
Strauss, M. B. (1957). Body Water in Man. London: Churchill.
Talso, p. J., Spafford, N., and Blaw, W. (1953). J. Lab. din. Med.,

41, 281.
Verney, E. B. (1957). Lancet, 2, 1237, 1295.
Welt, L. G., Seldin, D. W., Nelson, W. P. Ill, German, W. J., and

Peters, J. P. (1952). Arch, intern. Med., 90, 355.
Wesson, L. G., Anslow, W. P., Raisz, L. G., Bolomey, A. A., and

Ladd, M. (1950). Amer. J. Physiol, 162, 677.
Weston, R. E., Hanenson, I. B., Grossman, J., Berdasco, G. A., and

WoLFMAN, M. (1953). J. din. Invest., 32, 611.
Whedon, G. D., and Shorr, E. (1957). J. din. Invest., 36, 941.
Wilson, A. O. (1955). BriL J. Surg., 43, 71.
Wolf, A. V. (1950). Amer. J. Physiol, 161, 75.
Wrong, O. (1956). Clin. Sci., 15, 401.
Wynn, V. (1956). Metabolism, 5, 490.
ZiLVA, J. F., and Harris-Jones, J. N. (1957). J. din. Path., 10, 156.

DISCUSSION

Wallace : Hypernatraemia is seen very frequently in young infants with
dehydration secondary to diarrhoea. I think that there are two points
worth noting here. The first is that infants can lose large amounts of
water in their stools without losing physiologically equivalent amounts of
sodium. The sodium content of stool water can be very low. It is almost
as though the gut contents had been passed over an exchange resin. The
second item is that, in infants at least, the hypernatraemia is accom-
panied by an ever greater degree of hyperchloraemia. Since the flame
photometer came into the laboratory chloride has been a neglected ion.
We have wondered whether or not chloride might not be an ion with
much more autonomy than it is generally given credit for. In the child-
ren we have studied, gain of water and loss of chloride have been the
primary measurable events occurring during clinical recovery.

Davson: Does the gut remove the sodium from the normal faeces?

Wallace: In normal faeces there is very little sodium.

Davson : It may be that the active accumulation mechanism is set to
take up any sodium that is in the gut.

Wallace : A few stools from infants with the salt-losing type of adreno-
genital syndrome and with concurrent diarrhoea have been examined.
The sodium in stool water from these infants has been found to be much

Discussion 59

higher than we have found in the child with hypernatraemia. The urine
of the infant with hypernatraemia is also low in sodium. One finds both
the gut and kidney strongly retaining sodium beyond what might seem
an optimal degree. I wonder what this means?

Young: This hanging on to sodium without any excess excretion in the
urine is just what happens in experimental dehydration. If you are not
putting sodium into the body either by mouth or intravenously, there is
never a high output of sodium in the urine, even if the serum sodium is
rising. There is nothing extraordinary about that in the baby. Why the
kidneys function that way, I do not know, but they did so under condi-
tions of experimental dehydration in the normal adults studied by Dr.
Black, Prof. McCance, and myself (1944. J. Physiol. 102, 406).

Desaulles : Is there any possibility of making chromatograms of blood
and urine steroids in the kind of case you have just described, Prof.
Wallace? The aldosterone content was very high, wasn't it?

Wallace : We can obtain such chromatograms but I am always told that
close to a litre of blood or urine is required, and these are tiny children.

Desaulles: For aldosterone determination 100 ml. is enough. The con-
dition fits so well with the picture of a very high aldosterone output that
I wonder if those cases cannot be explained by the very high aldosterone
levels. In these all the sodium is retained without changes in the water
content. In the recovery period you have water retention and a decrease
in aldosterone. After that you reach a steady state, i.e. a new form
of equilibrium, though it is perhaps not the true equilibrium. That is
only a hypothesis for the moment, until we have more precise values.

Davson : Does aldosterone influence the absorption of water by the
intestine?

Desaulles: I have no precise data.

Black : I want to express agreement with Dr. Fourman, because I think
that none of the alleged clinical tests for water depletion, such as the
'fingerprint' test, are any good. There is also another possible cause of
so-called cerebral salt-wasting. We had a patient in with hemiplegia and
a period of hypotension. Ten days later he was mopping up about six
litres of saline fluid a day and losing it through his urine. The only sug-
gestion I can make is that during the period of hypotension he sustained
tubular damage and that later he was in a renal salt-losing state, in
which the cerebral part was just an accident. I have seen this before and
I think it is particularly liable to happen in older people who have a
smaller renal reserve.

Fourman : I think that is a very interesting comment. The very severe
dehydrations probably do produce renal lesions and we have been
wondering whether that accounts for the systemic acidosis, which is so
often a prominent feature.

Wallace: Chloride acidosis always occurs.

Fourman: What is the plasma bicarbonate?

Wallace: In our experience it is always low. Chloride is making bicar-
bonate forfeit its place in serum.

Desaulles : Dr. Fourman, was it possible to make steroid determinations
in your case?

60 Discussion

We have made an observation on animals that is not identical but
may point in the same direction as the observation you have made. If
adrenalectomized rats are given a very high salt load, hypernatraemia is
produced in a relatively short time. Firstly, then, the sensitivity to ADH
and pitressin decreases considerably. We did not get any serum values
but in the urine there is a strong dilution due to the greater urinary out-
put. Secondly, treatment with aldosterone in relatively high doses for
four or five days causes sensitivity to pitressin to disappear completely.

Fourmcm : With high aldosterone dosage there is certainly an expansion
of the extracellular fluid, and it may be that this expansion diminishes
the sensitivity to pitressin.

As regards the steroid assays, I do know that Schwartz and Bartter's
cases, which were analogous in many ways, were not salt-deficient; they
had an expanded extracellular volume and the aldosterone output in the
urine was normal. A. Gowenlock in Manchester measured the aldo-
sterone output in one of our patients and it was normal. We also did
17-ketosteroid assays as a crude measure of their corticoid output, and
the results were normal. It is obvious that the hyponatraemia does not
lead to a stimulation of the aldosterone output of the adrenal.

Desaulles : Could this be given the same interpretation as the findings
of Prader, Spahr and Neher (1955. Schweiz. rued. Wschr., 85, 1085)?
There may be some form of sodium-losing syndrome.

Adolph : It seems to me. Dr. Fourman, that in order to show that there
is something more to one of these syndromes than a lack of drinking
behaviour or drinking response on the part of the individual, you have to
perform your tests in a certain order ; you have to be sure that the patient
has plenty of water when you do the salt test and plenty of salt when you
do the water test. Could you have switched the tests around and still
obtained the same results?

Fourman : The saline load was done three weeks before the dehydra-
tion. The dehydration preceded the pitressin by one day in one of the
patients, by a week in the other patient. The pitressin test was accom-
panied by a load of water at the time. I do agree that one test can
influence another but I do not think that they did in this instance.

Borst : When a high or a low sodium concentration in the blood plasma
is maintained we believe that this is almost always due to an insuf-
ficient circulation. This insufficiency often results from dehydration, but
it may have other causes such as cardiac failure or hypoproteinaemia.
We found a high blood sodium in anaemic patients who had had
recurrent haemorrhages from peptic ulcer. They had no free access to
water and had been treated with abundant saline infusions; they had
substantial oedema. During several days the urine contained less
sodium than tap water, After a large transfusion of blood the sodium
excretion started and the blood sodium fell to a normal level. Simul-
taneously, the output of water increased and the urea concentration of
the urine, which had been very high, decreased. The counterpart was
observed in cachectic patients with anaemia and hypalbuminaemia who
adhered to a salt-free diet and who had a liberal intake of water. They
maintained a low blood sodium concentration in the presence of oedema.

Discussion 61

A large blood transfusion elicited a considerable water diuresis and the
blood sodium rose to normal, while the oedema fluid was excreted.

With both the high and the low sodium concentrations the circulation
was inadequate. In the first instance the excess of sodium and a less
considerable excess of water was excreted as soon as the normal blood
volume was restored. In the second the rise in blood volume led to the
elimination of the excess of water and of a less considerable excess of salt.

The interesting observations of Dr. Schwartz and Dr. Fourman show
that variations in circulation may not always be the primary factor in
the excretion of sodium and water. It is, however, difficult to distinguish
renal responses to variations in the circulation from other reactions on
the part of the kidneys. jNIoreover any considerable loss or retention of
salt and water has an effect on the circulation. The problem is that an
excess or an inadequacy of the circulation in patients cannot be measured
in a satisfactory way. Since this factor cannot be disregarded we have to
estimate it on the basis of indirect evidence.

GLANDULAR SECRETION OF ELECTROLYTES
JoRN Hess Thaysen

Medical Department A, Rigshospitalet, Copenhagen

The ducts or tubules of glands with external secretion are
usually quite complex in structure and morphologically they
differ to a considerable extent from gland to gland. It is,
therefore, reasonable to assume that the ducts do not merely
serve as pathways for the secretion formed in the acini, but
that they contribute somehow to the elaboration of the final
secretory product. This possibility has already been considered
in the past century by Merkel (1883), mainly on morphological
grounds, and by Werther (1886), who made a comparative
investigation of the concentration of salt in various types of
saliva. The results of these experiments were, however,
inconclusive, and in 1950 Babkin restated the need for a
study of the physiology of the glandular ducts. Since then,
certain advances have been made through comparative work,
by the application of concepts from modern renal physiology
and with the use of electrophysiological methods, relating
changes in membrane potentials to ionic transport. It is the
purpose of the present paper to review this work and to
present a theory of the mechanism of glandular electrolyte
secretion based on the available data.

Fig. 1 shows a comparison between the concentrations of
the main electrolytes in sweat, parotid saliva, tears and
pancreatic juice in relation to secretory rate, calculated in
milligrams per gram gland per minute. The following simil-
arities and differences between the four secretory products
are apparent from Fig. 1 :

The Excretion of Sodium:

In sweat and in parotid saliva the concentration of sodium
is smaller than the concentration of sodium in plasma and

Glandular Secretion of Electrolytes

varies with the rate of secretion. With increasing secretory
rate the concentration of sodium rises to about 60 m-equiv./l.
in the sweat and to about 90 m-equiv./l. in the parotid sahva,
but no definite maximum is reached in either secretion. This
finding conforms with the old work of Heidenhain (1868),
Langley and Fletcher (1889), Kittsteiner (1911, 1913), and
Hancock, Whitehouse and Haldane (1929).

120

2 160

SWEAT (1)

PAROTID SALIVA (2)

PANCREATIC JUICE (6)

20 AO 60 80
SECRETORY RATE

100 20 AO 60 60 100

(MG PER GRAM GLANO PER MINUTE)

Fig. 1. The concentration of the main electrolytes in sweat,
parotid saliva, tears and pancreatic juice in relation to secretory
rate (in milligrams per gram gland per minute). From the data of
1 : Schwartz and Thaysen (1956) ; 2 : Thaysen, Thorn and Schwartz
(1954); 3: Thaysen and Thorn (1954); and 4: Bro-Rasmussen,
Killmann and Thaysen (1956).

In tears and in pancreatic juice the concentration of sodium
in secretion water is about equal to the concentration of
sodium in plasma water and is independent of the rate of
secretion.

The Excretion of Potassium:

The concentration of potassium in all four secretions
independent of wide ranges of variation in secretory rate.

64 J0RN Hess Thaysen

In parotid saliva, however, a definite rise in potassium
concentration is noted at rates smaller than 15 mg. per gram
gland per minute. This finding is in agreement with the results
of Langstroth, McRae and Stavraky (1938) and Burgen (1956).
A similar rise in potassium concentration possibly occurs at
low rates of sweat secretion (Kuno, 1956), but could not be
demonstrated with the experimental technique employed by
Schwartz and Thaysen (1956). In the two other secretions a
rise in potassium concentration at low secretory rates has
never been observed.

The Excretion of Anions:

The main anion of sweat and tears is chloride. This anion
accounts for about 80 per cent of the sum of the concentrations
of sodium and potassium in the tear fluid. Chloride concentra-
tion of sweat is not depicted in Fig. 1, but Locke and his co-
workers (1951) found the following relation: sodium=
1-12 chloride-j-3 m-equiv./l.

The chief anion of parotid saliva and pancreatic juice is
bicarbonate. With increasing secretory rate the concentra-
tion of bicarbonate rises in both secretions and reaches a
maximum of about 60 m-equiv./l. in parotid saliva and about
90-130 m-equiv./l. in pancreatic juice. When this maximum
concentration (which is subject to individual variation) has
been arrived at, the concentration of bicarbonate remains
independent of further increases in the rate of secretion. The
concentration of chloride varies inversely with that of bi-
carbonate. In both secretions and at all rates the sums of the
concentrations of the two anions equal about 80-90 per cent
of the sums of the concentrations of sodium and potassium.

The following hypothesis has been put forward to explain
the demonstrated differences in the excretion of the main
cations. In all four glands a precursor solution is formed in
which the concentration of sodium is independent of the rate
of precursor formation. In the sweat and parotid glands, but
not in the other two glands, sodium is consequently reab-
sorbed by a process of a limited maximal capacity (Thaysen,

Glandular Secretion of Electrolytes

Thorn and Schwartz, 1954; Thaysen, 1955; Schwartz and
Thaysen, 1956; Bulmer and Forwell, 1956; Bro-Rasmussen,
Killmann and Thaysen, 1956). Like sodium, potassium is
transferred into the precursor at a constant concentration,
but it is not reabsorbed in any of the glands. The rise in
potassium concentration at the low secretory rates in parotid
saliva (and in sweat?) may be secondary to reabsorption of
water from the precursor as indicated by Langstroth, McRae
and Stavraky (1938) and by Thaysen, Thorn and Schwartz
(1954), and/or to an exchange between sodium and potassium
ions during the process of sodium reabsorption.

Table I

Comparison between the calculated concentrations of sodium and

POTASSIUM in the PRECURSOR SECRETIONS OF FOUR SECRETORY PRODUCTS
AND THE CONCENTRATIONS OF THE SAME IONS IN PLASMA WATER

SWEAT

PAROTID

LACRYMAL

PANCREATIC

PLASMA WATER

Na
K

79
9

112
19

K6
15

161
5

160
5

SUM

131

161

166

165

Fig. 2 shows a linear regression of the rate of sodium
excretion in parotid saliva on the rate of secretion. According
to the above hypothesis the values for slope and intercept in
Fig. 2 can be interpreted to mean that sodium is transferred
into the precursor solution at the rate of 0-112 microequiva-
lents per mg. of saliva discharged and that 2 • 4 microequiva-
lents are subsequently reabsorbed per gram gland per minute.
The sodium concentration of the sweat precursor has been
calculated in a similar manner from the data of Schwartz and
Thaysen (1956) and the values are compared to those of the
other secretions and to plasma water in Table I. According
to Table I the sums of the concentrations of sodium and
potassium in the presecretions of saliva and sweat are lower
than the sums of the concentrations of the same ions in the
two other secretions and in plasma water. No other cations

AQEIXG — IV— 3

J0RN Hess Thaysen

are present in parotid saliva and in sweat in sufficiently large
concentration to make up for this difference. Judging from the
results of Table I, the production of sweat and parotid saliva
should therefore involve secretion of hypotonic precursor
solutions, a process which a priori does not appear very likely.

• y» 0.112(t0.005)X-2.4(tO.A)

100

120

X«SECRETORY RATE (mg/gram gland /min)

Fig. 2. The relation between the rate of sodium
excretion in parotid sahva (in microequivalents
per gram gland per minute) and secretory rate (in
milligrams per gram gland per minute). The linear
regression has been calculated for all data at or
above a secretory rate of 60 milligrams per gram
gland per minute.

It must be emphasized, however, that the calculated figures
for precursor sodium concentration (and sodium reabsorp-
tion) in the sweat and parotid glands underestimate actual
values, since the regressions for sodium excretion on secretory
rate have been fitted to points which approach, but do not
reach, a rectilinear relationship within the observed range
(cf. Fig. 2). One explanation for this considerable splay in

Glandular Secretion of Electrolytes 67

the observed values from the asymptote could be that there
is a certain back-diffusion of water in the sequence of active
sodium reabsorption. As demonstrated below there is reason-
able qualitative evidence to suggest that water is, in fact,
reabsorbed from the precursors of sweat and parotid saliva.'
Fig. 3 illustrates that the concentration of urea in sweat,
tears, and parotid saliva remains proportional to the con-

600

(1) SWEAT-- 1,82(10,02) P-6(i3)
(ALL RATES)

(3) PAROTID SALIVA=Q724(tQ002)P-3(i3)
(RATES>0,5ml/nin)

100 200 300 400 500

CONCENTRATION OF UREA IN PLASMA (mg/k)0ml)

Fig. 3. The relation between the concentration of
urea m the plasma (P) and the concentration of
urea in sweat, parotid sahva and tears. From
the data of 1 : Schw^artz, Thaysen and Dole (1953) •
2 : Albrectsen and Thaysen (1955) ; and 3 : Thaysen
and Thorn (1954).

centration of urea in the plasma within a wide range of varia-
tion in the latter. This finding indicates that urea is excreted
m these secretions by a process of simple diffusion and not via
a specific secretory mechanism which might become saturated
by increasing load. Potentially urea may, therefore, be used
as a tracer for the movement of water within the secreting
glands in a similar manner as in the glomerular nephron.

J0RN Hess Thaysen

Fig. 4 shows the relationship between the S/P (secretion/
plasma) concentration ratio for urea and the rate of secretion
of sweat, parotid saliva, tears and pancreatic juice.

In tears and in pancreatic juice there is apparently dif-
fusion equilibrium between the secretion and the plasma at all

2.0-

1 SWEAT (1)

\
\
\
\
\

PAROTID SALIVA (2)

^V^

TEARS (3)

PANCREATIC JUICE (4)

1.5

1.0

0.6

0.8

SECRETORY RATE
Fig. 4. The relation between the S/P ratio for urea
and secretory rate in sweat, parotid saliva, tears
and pancreatic juice. From the data of 1 : Araki
and Ando (1953) (the curve is shown as a
broken line because it represents the approximate
mean of two determinations and because secre-
tory rate cannot be directly compared to that of
the other glands); 2: Albrectsen and Thaysen
(1955); 3: Thaysen and Thorn (1954); 4: Bro-
Rasmussen, Killmann and Thaysen (1956).

rates of glandular activity. On the basis of these findings no
statement can be made about the existence or non-existence
of an internal circulation of water in these glands.

In sweat and in parotid saliva S/P urea varies with the
rate of secretion. In the sweat S/P urea decreases from
2 or 3 at the low secretory rates to about 1 when sweating

Glandular Secretion of Electrolytes 69

becomes profuse (Araki and Ando, 1953; Bulmer, 1957). In
parotid saliva the ratio decreases from about 1 • 6 at low rates
of secretion to about 0-6 when the flow of saliva is brisk
(Albrectsen and Thaysen, 1955). Since no specific secretory
mechanism for urea exists in either gland, it is reasonable to
conclude that urea, which is diffusing into the gland with
some precursor solution, is raised to a concentration greater
than that of the plasma by reabsorption of water from the
precursor in a region of the gland which is less permeable to
urea than the site of precursor formation. The rate of change
in S/P urea with secretory rate suggests that water reabsorp-
tion represents a relatively constant quantity at all rates of
precursor formation, and it is not unreasonable to assume
that the reabsorption of water occurs as a mere passive
sequence of active sodium reabsorption.

Quantitative information about precursor formation and
water reabsorption can, however, hardly be gained from these
results or from similar "clearance" studies with other solutes.
Morphological and physiological evidence strongly argues
against the possibility that the secretion precursor represents
an ultrafiltrate of the plasma like the urine precursor of the
glomerular nephron. A "glandular inulin" probably does not
exist, and it is quite possible that exact knowledge about
the composition of the precursor secretions and about the
manner in which they are modified as they flow down the
glandular ducts can only be obtained by micropuncture
techniqiies.

However, Lundberg (1955, 1957«,&,c), working on the
electrophysiology of the submaxillary and sublingual glands
of the cat, has obtained results which provide indirect support
in favour of the hypothesis that sodium is reabsorbed from a
precursor secretion in some of the duct-possessing glands.

In the submaxillary gland, which produces a secretion in
which sodium concentration varies with secretory rate in
about the same manner as in parotid saliva and sweat,
Lundberg (1955) demonstrated that the lumen of the (striated?)
ducts becomes negative as compared to the hilus, when the

70 JoRN Hess Thaysen

gland is activated by stimulation of the chorda. A similar
internal duct negativity could not be demonstrated in the
sublingual gland (Lundberg, 1957a), which (like the lachrymal
and pancreatic glands) produces a secretion that is isotonic
with the plasma and has a sodium concentration of about 150
m-equiv./l. Provided that the potential changes on stimu-
lation can be regarded as the electrical signal of ionic trans-
port, Lundberg (1957a) concludes that there is a net transport
of cation from the lumen to the blood side in the ducts of the
submaxillary gland, but not in the sublingual gland. Although
the composition of the submaxillary secretion was not
measured simultaneously with the duct potential, the latter
appears large enough for it be to accepted that the reabsorp-
tion of anion is merely a passive sequence of active cation
transport.

With one microelectrode inserted into acinous cells and the
other electrode on the gland surface, Lundberg (1955, 1957a)
detected a considerable increase in the negativity of the
acinous cells on stimulation of the submaxillary as well as of
the sublingual gland. The lumen of the acini, likewise,
becomes negative as compared to the morphological interior,
but this negativity decreases slightly with continued stimu-
lation of the gland. These potential changes may be due to a
net transport of anion from the blood side into the glandular
lumen. In another paper Lundberg (1957c) directly demons-
trated this anionic dependence of secretion and secretory
potentials in the perfused sublingual gland. Substitution of
sodium chloride with sodium nitrate or sodium thiocyanate
caused the secretion to stop almost entirely and decreased
the potential changes. The secretory response and the
potentials reverted to normal when sodium chloride was
again added to the perfusate.

On the basis of the experiments quoted in the present
report, it appears reasonable to suggest the following mechan-
ism for the secretion of electrolytes and water by the duct-
possessing glands. Active outward transport of anions is a
main factor in the formation of the secretory products of all

Glandular Secretion of Electrolytes 71

glands. In some glands the chief anion transported is chloride
(sweat, tears, sublingual saliva); in others bicarbonate ions
are added in varying proportion, possibly due to the presence
of carbonic anhydrase in the cells (pancreatic juice, parotid
saliva, submaxillary saliva). It is reasonable to assume that
water moves in a merely passive sequence of ionic transport
from the blood side into the glandular lumen, and that the
presecretions of all glands are isotonic or nearly isotonic.

In certain glands (sweat, parotid and submaxillary) sodium
is reabsorbed from the precursor secretion as it flows down the
glandular duct system, and it is likely that anions move from
duct lumen to the blood side in a passive sequence of the
active sodium reabsorption. The chief anion reabsorbed in
this manner appears to be chloride, independently of whether
the primary secretion contains primarily chloride or primarily
bicarbonate ions. It can be seen from a glance at Fig. 1 that
the parotid and the pancreatic glands apparently form pre-
secretions of qualitatively similar composition, and that the
main difference in the anionic pattern of the final secretory
products is that chloride ions have been removed from the
saliva precursor. As a consequence of active sodium re-
absorption a certain quantity of water is, moreover, diffusing
back into the blood stream, although it is obvious that
water reabsorption does not occur isotonically as in the
proximal renal tubule.

It is only possible to speculate on the morphological sites of
the different ionic transports in the duct-possessing glands.
According to Fig. 5 it is, however, not unreasonable to
suggest that sodium reabsorption is located in the striated
intralobular ducts. Striated epithelium is present in the
parotid and submaxillary glands, which apparently reabsorb
sodium, but it is absent in the sublingual, pancreatic and
lachrymal glands, which show no evidence of sodium re-
absorption. The precursor secretions are probably formed
by the acini as well as by the cuboidal epithelium of the
intercalary ducts, the former producing a viscous secretion
with a high concentration of organic material, the latter a

J0RN Hess Thaysen

watery secretion with a low concentration of organic material
(cf. Babkin, 1950). With respect to the sweat gland it is

SWEAT

PAROUS

SUBMAX.

S <EC.

Na Na

Z> varies with
Na

secretory rate

SUBLING. PANCREAS LACRYMAL

S = E.C.

Na No

_D», independent of
Na

secretory rate

Fig. 5. Comparison between the histological structure of the six
main duct-possessing glands and their secretion of sodium ions.
Sjja = concentration of sodium in tlie secretion. E.C.Na = con-
centration of sodium in the extracellular fluid. The coil of the
sweat gland and the acini of the other glands are cross-hatched.
The epithelia of the ducts are illustrated by different symbols,
which refer to the schematic cross- sections at the bottom of the
figure. The cross-sections are (from left to right) : double-layered
epithelium of sweat duct; striated epithelium of intralobular
ducts ; high cylindrical epithelium of excretory ducts ; low cuboidal
epithelium of intercalary ducts.

suggested that precursor formation is located in the coil,
whereas reabsorption of sodium takes place in the duct.

REFERENCES

Albrectsen, S. R., and Thaysen, J. H. (1955). Scand. J. din. Lab.

Invest., 7, 231.
Araki, Y., and Ando, S. (1953). Jap. J. Physiol., 3, 211.
Babkin, B. P. (1950). Secretory Mechanism of the Digestive Glands,

2nd ed. New York : Hoeber.

Glandular Secretion of Electrolytes 73

Bro-Rasmussen, F., Killmann, S.-A., and Thaysen, J. H. (1956).

Acta physiol. scancL, 37, 97.
BuLMER, M. G. (1957). J. Physiol, 137, 261.

BuLMER, M. G., and Forwell, G. D. (1956). J. Physiol., 132, 115.
BuRGEN, A. S. V. (1956). J. Physiol., 132, 20.
Hancock, W., Wiiitehouse, A. G. R., and Haldane, J. S. (1929).

Proc. roy. Soc, 105 B, 43.
Heidenhain, R. (1868). Stud, physiol. Inst. Breslau, 4, 1.
Kittsteiner, C. (1911). Arch. Hyg., Berl., 73, 275.
KiTTSTEiNER, C. (1913). Arch. Hyg., Bed., 78, 275.
KuNO, Y. (1956). Human Perspiration. Springfield: Thomas.
Langley, J. N., and Fletcher, H. M. (1889). Phil. Trans., 180 B,

109.
Langstroth, G. O., McRae, D. R., and Stavraky, G. W. (1938).

Proc. roy. Soc. 125 B, 335.
Locke, W., Talbot, N. B., Jones, H. S., and Worcester, J. (1951).

J. din. Invest., 30, 325.
Lundberg, a. (1955). Acta physiol. scand., 35, 1.
LuNDBERG, A. (1957«). Acta physiol. scand., 40, 21.
Lundberg, A. (19576). Acta physiol. scand., 40, 35.
Lundberg, A. (1957c). Acta physiol. scand., 40, 101.
Merkel, F. (1883). Die Speichelrohren. Rektoratsprogramm. Leipzig:

Vogel.
Schwartz, I. L., and Thaysen, J. H. (1956). J. din. Itwest., 35, 114.
Schwartz, I. L., Thaysen, J. H., and Dole, V. P. (1953). J. exp. Med.,

97, 429.
Thaysen, J. H. (1955). Sekretionsstudier. Copenhagen: Diss.
Thaysen, J. H., and Thorn, N. A. (1954). Amer. J. Physiol., 178, 160.
Thaysen, J. H., Thorn, N. A., and Schwartz, I. L. (1954). Amer. J.

Physiol., 178, 155.
Werther, M. (1886). Pfliig. Arch. ges. Physiol., 38, 293.

DISCUSSION

Davson: As far as I can make out, Dr. Thaysen, you postulate that
there is a region through which the urea can pass quite easily, and later
on in the ducts there is a relative impermeability to urea. This is rather
in conflict with what people have thought in the past, because, on the
assumption that it penetrates into all cells very rapidly, urea has been
used to determine cell water. Your view certainly does fit in with what
is found with the cerebrospinal fluid and the aqueous humour; urea does
not penetrate those barriers easily. If one confined oneself to these in-
stances, then, one would say that urea did not penetrate cells easily at all.

Thaysen : Yes, I believe that the cells in the region of water reabsorp-
tion are less permeable to urea than the cells at the site of precursor
formation. In all probability the difference in permeability is, however,
relative rather than absolute. In other words, I do not think that the cells
at the site of precursor formation are so freely permeable to urea that the
concentration of urea in the precursor is equal to that of the plasma at all

74 Discussion

rates of secretion. Certainly this is not tlie case in the parotid gland
(Fig. 4, p. 68). Conversely, I do not venture to claim that the cells in the
duct are impermeable to an extent that would completely prevent urea
from diffusing back into the blood stream along the concentration
gradient created by water reabsorption. But the amount of urea diffu-
sing back through the relatively impermeable duct epithelium is limited
by the short span of time during which the secretion remains in the duct.
Urea may equilibrate rapidly over some cellular membranes, more
slowly over others. This difference is not important when one measures
total body water as the volume of distribution of urea, because one waits
until complete equilibrium has been established before the measurement
is made. But the difference is important in the rate-dependent process of
secretion, where the time available for diffusion becomes limiting.

Karvonen: In prolonged sweating the potassium concentration is
higher to start with and then gradually decreases. There is no similar
change in sodium or chloride and that would agree quite well with the
reabsorption and consequent storing of potassium in the tubule, whereas
sodium and chloride are not stored (Ahlman et al. (1953) Acta endocr.,
Copenhagen, 12, 140).

Thaysen : Yes, the first sample of sweat obtained after stimulation may
have a higher potassium concentration than the following ones. One
reason for this may be that the first sample is contaminated with cellular
debris, sebum and sweat residues on the skin surface.

Karvonen: It is not just the rinsing factor, because we paid quite a lot
of attention to rinsing the skin and we still get this difference ; the potas-
sium is probably stored in the gland or at least in the tubule.

Thaysen : In that case it cannot be contamination. Your finding is very
interesting to me, because we found exactly the same thing with the
parotid secretion. The first sample of saliva obtained after stimulation
invariably had a higher potassium concentration (and a higher urea con-
centration) than the following ones. This phenomenon occurred inde-
pendently of the rate at which the first sample was produced. We
speculated that the vigorous flow of saliva, caused by stimulation,
" pushed out" first a small amount of secretion, which had been produced
at the low secretory rates prior to stimulation, and which consequently
had a high concentration of potassium and urea and a relatively low
sodium concentration (1954. Amer. J. Physiol., 178, 155; 1955, Scand. J.
clin. Lab. Invest., 7, 231). Burgen (1956) also observed a high potassium
concentration in the first samples of saliva obtained after stimulation.

Wallace: I have kept quiet here because a baby usually does not
sweat until the age of 3-4 months, nor does he shed tears — he only
learns to do that later.

Karvonen : In Finland babies have hot sauna baths quite young and
Dr. Eila Kassila of the Children's Clinic, Helsinki, has made an investi-
gation on the composition of the sweat they produce during the saunas.
I do not think that much difference was found between baby and adult.

Wallace : It is of interest that there is a very specific disease in children,
cystic fibrosis of the pancreas, in which the ability of the sweat glands to
reduce the sodium concentration in sweat seems to be lost. It is of

Discussion 75

theoretical interest that this is a disease which manifests itself primarily
in the lungs and pancreas with gross pathology, and yet has this very
subtle physiological pathology in the sweat glands. Have you done any-
thing with that type of patient?

Thaysen: Yes, but I never did much with them. We did find a very
high sodium concentration in their sweat.

Wallace : A high sodium concentration in sweat is found in nephrosis,
and Dr. Warming-Larsen of Copenhagen has studied this problem. The
nephrotic child gaining oedema has a high sodium concentration in the
sweat yet a very small sweat volume ; but overnight, as he diureses, he
puts out an increased volume of sweat yet at the same time the sweat
sodium concentration falls. The net amount of sodium lost from the
sweating skin is the same whether he is oedematous or not. I would like
to know about the relation of ADH to sweat volume ; does ADH control
the sweat glands as well as the kidney?

Thaysen: That is interesting. Off-hand one would have guessed that
the sodium concentration of the sweat would have been low during
the phase of oedema formation and high when the patient started to
diurese. That would agree with what we know about the action of aldo-
sterone on the glands and with the results of sweat and saliva analyses
in other oedamatous states. Since the quantity of sodium excreted per
unit area of the skin per unit time remained constant, whereas the volume
of sweat increased when the child diuresed, an ADH effect might be a
possibility worth considering. However, as far as I am aware, it has been
shown that ADH has no effect on the volume of sweat produced ( Ama-
truda, T. T., Jr., and Welt, L. G. (1953). J. appl. Physiol, 5, 759; Pearcy
et al. (1956). J. appl. Physiol, 8, 621).

Adolph : Can somebody clarify the reports that tears are very hyper-
tonic when they are formed?

Davson : I did some analyses a long time ago, and we discovered that
the chloride concentration was equal to that of the blood. It is a very
difficult problem obtaining tears, because you have got to make the
person cry very hard to get enough to do an analysis.

Thaysen: In 1889 Massart {Arch. Biol, Paris, 9, 537) applied sodium
chloride solutions of varying concentration to the conjunctival sac of a
few test subjects. He never analysed the tear fluid, but from the reac-
tions of the test subjects to the different solutions he concluded that a
1-3 per cent solution of sodium chloride was isotonic with the tears.
According to Krogh and co-workers (1945. Acta physiol scand., 10, 88)
this experiment forms the only basis for the rather widespread statement
in physiological and pharmacological textbooks that tears are hypertonic
as compared to the plasma. In 1945 Krogh measured the osmotic pres-
sure of tears and found them to be isotonic. The finding w as confirmed
by Giardini and Roberts in 1950 {Brit. J. Ophthal, 34, 737).

Black: If you inject ^^K intravenously and then collect serial samples
of saliva you find that the specific activity of potassium in the saliva is
several times that of the specific activity of the potassium in plasma at
the same time. This behaviour is analogous to that in urine and suggests
to us that the potassium in saliva is, like that in urine, secreted by cells.

76 Discussion

and not just filtered from the plasma. One then wonders whether epith-
elium does not similarly push out potassium in exchange for the sodium
which is being reabsorbed — the sort of mechanism that is possibly under
aldosterone control.

I believe that although Conn has concentrated mainly on sweat in his
tests for aldosterone activity, he has also used saliva in a similar way.
With a rice diet we did not get any falling off in the sodium concentration
in the saliva, as far as we could determine.

Thaysen : An exchange mechanism between sodium and potassium ions
at the site of sodium reabsorption is certainly a very likely possibility.
This may be one factor causing the potassium concentration of the final
secretory product to exceed that of the plasma. However, glands which
apparently possess no sodium-reabsorbing mechanism may also have a
potassium concentration in their secretions, exceeding the plasma potas-
sium concentration. This applies, for example, to the lachrymal (Fig.l,
p. 63) and the sublingual glands (Lundberg, 19576). Therefore I believe
that two factors may be at stake. First, the presecretion is frequently
formed with a potassium concentration which exceeds that of the plasma
(and a correspondingly lower sodium concentration). Second, in some
glands additional potassium ions are added to the presecretion in ex-
change for reabsorbed sodium ions. Similarly, the adrenal steroids may
have a dual site of action in the glands. In contradistinction to the situa-
tion in the glomerular nephron, adrenal steroids may act on the gland
cells forming the presecretion and thus alter the Na/K ratio of the pre-
cursor, and they may act on the cells in the ducts which reabsorb sodium
ions from the presecretion in exchange for potassium ions. There is some
evidence indicating such a dual site of action of aldosterone on the glands
(Thorn et al. (1954). Fed. Proc, 13, 310), but I do not know of any con-
clusive experiments. One way of approaching the problem may be to
compare the effect of aldosterone on glands with and without a sodium-
reabsorbing mechanism, e.g. on the sweat or parotid gland as contrasted
with the lachrymal or pancreatic.

As regards your comment about the rice diet, sodium depletion, in-
duced by a low sodium diet, causes the concentration of sodium to
decrease and the concentration of potassium to increase in sweat as well
as in saliva (McCance, R. A. (1938). J. Physiol, 92, 208). However, the
response of the glandular epithelium to sodium depletion is both delayed
and incomplete as compared to that of the kidney tubule (Robinson et
al. (1955). J. cqypl. Physiol., 8, 159; Thorn et al. (1956). J. appl. Physiol,
9, 477).

Karvonen: Can anyone comment on the statistical finding that men
have lower sodium and potassium than women in their sweat, and that
the Na/K ratio in women is significantly lower than in men (Ahlman et
al (1953). J. clin. Endocrin. Metab., 13, 773)?

Desaulles: That is a very interesting challenge. We have similar
findings in animals, not in sweat but in urine, but I have absolutely no
explanation for it. It is just an observed fact.

Talbot : I wonder if those who are commenting on the sodium, chloride
and potassium concentrations in sweat all have in mind the relationship

Discussion 77

between rate of sweating and the concentration, because it varies
enormously.

Thaysen: Yes, that is very important. Secretory rate must be con-
trolled in all work on electrolyte composition of secretions, and compara-
tive studies can only be made on the basis of standard rates. This is of
course equally important whether one states the result in absolute con-
centrations of sodium and potassium or as the Na/K ratio. When the
developing organism is under study, secretory rate must be expressed per
unit weight of gland or some other parameter allowing for the influence
of growth. I believe that negligence of these important factors is the
main reason why the literature on variations with sex and age in the
secretion of electrolytes is, largely speaking, inconclusive and frequently
mutually conflicting.

With respect to Dr. Karvonen's remark I should like to add this com-
ment. I take it that the sweat tests have been done in the usual way, i.e.
with collection of sweat from a smaller or larger area of the skin, not
from individual glands, and that the difference between the men and
women is stated on the basis of comparable sweating rates per unit area
of the skin. It does tell us, then, that there is a difference between sweat-
ing of men and women, but it does not tell us anything about the reason
for the difference. As is well known, the number of functioning sweat
glands per unit area of the skin varies between individuals, between the
sexes and with age, as well as between different skin regions in the same
person. Comparable rates per unit area of the skin are therefore not
necessarily the same as comparable rates per gland. Physiologically it is
of course the rate per gland that matters and not the rate per unit skin
area. Let us take it that women have half the number of glands per
unit skin area that men have. Since the rate per unit skin area was
comparable, the mean flow per gland in the women would then be twice
that in the men. A higher sodium concentration in the swxat of the
women might therefore merely be due to the fact that secretory rate per
gland was larger. Let us take it that the men and women had an equal
number of glands per unit skin area. In that case the difference between
the sexes could not be due to a difference in the rate per gland, but
might well be due to hormonal or other factors. What I mean is that in
comparative work it is a prerequisite to determine not only secretory
rate, but also the number of functioning glands, if you want to make
deductions from your findings. A method for determination of the
number of functioning glands within the area of sweat collection has
been published by Dole and Thaysen (1953. J. exp. Med.. 98, 129).

HORMONAL ASPECTS OF WATER AND

ELECTROLYTE METABOLISM IN RELATION

TO AGE AND SEX

G. I. M. SWYER
Obstetric Hospital, University College Hospital, London

Nearly all the hormones may have some influence on
water and electrolyte metabolism. However, for most of them,
this effect is indirect and occurs only under highly abnormal
circumstances. Thus, the dehydration which exists in un-
controlled diabetes mellitus or in hyperparathyroidism is the
result, respectively, of gross deficiency of insulin or excess of
parathormone, and certainly does not point to any physio-
logical role of these hormones in water metabolism. The same
is essentially true of thyroid hormone and, though perhaps
with reservations, of the sex hormones and gonadotrophins.
Only posterior pituitary antidiuretic hormone (ADH) and
certain of the adrenocortical steroids are directly concerned
with the day-to-day and minute-to-minute adjustments
needed to maintain fluid and electrolyte homeostasis in mam-
mals. The major details of this hormonal control are well
known and it is not necessary to relate them here. It is pro-
posed, on the other hand, to examine how the influence of
hormones on fluid and electrolyte balance differs at various
ages and in the two sexes. In general, it is fair to say that
little attention has been paid to considerations such as these,
and for the most part knowledge is meagre.

In Infancy

Fluid and electrolyte control is notoriously inefficient at
birth and during the first few weeks or so of life. The late
development of the loop of Henle is generally considered to
be responsible for this (Hubble, 1957), the infant kidney being
unable, in consequence, to vary tubular reabsorption of water

Hormones and Water and Electrolyte Metabolism 79

and salt. There does not appear to be any inability to secrete
ADH or adrenocortical hormones, though it is possible that
the infant does lack the power to adjust the amounts secreted
with any precision. The endocrinological situation, therefore,
is essentially one of target-organ insensitivity due to im-
maturity.

An interesting hypothesis relating to neonatal weight loss
has been put forward by Gans and Thompson (1957). These
workers measured the urine output and its content of oestro-
gens and 17-hydroxy corticosteroids in six normal male
neonates during the first few days of life. The findings were
similar in all the infants. Large amounts of oestriol (up to a
milligram or more) were excreted on the first post-partum
day, the quantity falling rapidly during the next two or
three days to the order of 1 or 2 [ig. by the sixth day. Oestrone
and oestradiol were found to the extent of 1-2 (xg. during the
first and second days and then disappeared. There was a
decreasing excretion of urine during the first three to five
days, and by the end of this time, postnatal weight loss had
ceased. The excretion of 17-hydroxy corticosteroids showed
only minor fluctuations throughout. The specific gravity of
the urine was low at first but became more concentrated as
the excess of water was excreted, in spite of the fact that
fluid intake was increasing during this time.

Gans and Thompson suggest that part at least of the
hydraemia of the newborn infant is due to water retention
caused by the high circulating oestrogen level — the oestrogens
being, of course, of maternal origin. As the oestrogens are
excreted, the fluid excess is eliminated.

Adrenal hyperplasia

Adrenal hyperplasia is a disorder with a definite predi-
lection for the female sex. In Wilkins' series (Wilkins, 1957)
the ratio was 62 females to 19 males. The clinical manifesta-
tions of this disorder and its pathogenesis need not concern us
here. It is, however, relevant to observe that about one-
fourth of these patients have a tendency to loss of sodium and

80 G. I. M. SwYER

to elevation of the plasma potassium, as a result of which
early death may occur from dehydration and circulatory
collapse, or from cardiac arrest due to hyperkalaemia. Once
again, the number of females affected is some three times that
of males.

The mechanism for this sodium loss is not understood. Very
likely there is a defect in aldosterone synthesis, but it is also
possible that some of the abnormal steroids produced by the
hyperplastic adrenals may actually cause sodium loss. It is
well known that surprisingly large amounts of sodium
chloride and cortexone acetate (DOCA) may be needed to
remedy the electrolyte defects in these infants, suggesting
that more than mere replacement of deficient hormone is
necessary. However, the response to 9a-fluorohydrocortisone,
together with cortisone, may be far more satisfactory. In a
Ij-year-old patient of the writer's, a female pseudoherma-
phrodite with the salt-losing disorder, 10 mg. daily of DOCA
intramuscularly, together with large sodium supplements,
was necessary to maintain electrolyte balance. With only
0-25 mg. of 9a-fluorohydrocortisone daily by mouth, it was
possible to maintain balance with no sodium supplement at
all.

A small proportion of patients with adrenal hyperplasia
(about 6 per cent) may show hypertension. It is possible that
in these there is actually sodium retention. Bongiovanni and
Eberlein (1955) have demonstrated in such a patient a defect
in the synthesis of Cortisol different from that usually found
in adrenal hyperplasia. This patient was producing increased
amounts of cortexone and 17-hydroxycortexone; it is thought
probable that these steroids were responsible for the hyper-
tension.

Changes in Relation to Adolescence

Knowledge of endocrine changes in relation to adolescence
is rather sketchy. It is ably summarized by Tanner (1955).
The impact of these changes on fluid and electrolyte metabol-
ism is somewhat obscure. Certain morphological changes of

Hormones and Water and Electrolyte Metabolism 81

possible significance occur. Thus, a considerable growth
spurt in the weight of the adrenal gland, more in boys than
in girls, has been observed. It is almost entirely due to growth
of the cortex. The weight of the thyroid also shows an adoles-
cent spurt, but without any sex difTerence. Scanty data on
hormone excretion indicate a slow increase in the excretion
of oestrogen in both boys and girls during childhood, with a
marked increase at puberty in the case of girls, while in boys
the rate of increase hitherto manifested is merely maintained.
Androgen excretion is similar in the two sexes before puberty ;
after puberty there is a marked rise in the case of boys, but a
not unimportant rise also occurs in girls, no doubt as a result
of increased adrenocortical activity. There is a gradual rise in
the rate of secretion of adrenal corticoids, without sex dif-
ference, from birth to maturity. The increase appears to be
proportionate to body size, without any adolescent spurt.
The blood level of 17-hydroxy corticosteroids is much the
same at all ages, and the responsiveness of the adrenals to
stimulation by adrenocorticotrophic hormone is also un-
affected by age, except, of course, in so far as the adrenal
glands are smaller in children than in adults. A steady fall
in the serum protein-bound iodine over the years six to 15
parallels the fall in basal metabolic rate, and the precise
significance of this is obscure.

The sum total of these changes does not seem to have any
striking impact on fluid and electrolyte metabolism.

Effects of the Menstrual Cycle

An important sex difference is introduced by the cyclic
variations in hypothalamic-pituitary-ovarian (and perhaps
adrenocortical) function which determine the menstrual
cycle in females. It might well be expected that these
would lead to important fluctuations in fluid and electrolyte
balance.

Variations in body fluid during the menstrual cycle have
been recognized for a long time, but the first full description
of "premenstrual oedema" was given by Thomas (1933) who

G. I. M. SWYER

reported weight gains of up to 14 lbs. at or during menstrua-
tion in two women. Several other writers (see Chesley and
Hellman, 1957) have concluded that approximately 30 per
cent of women have weight gains associated with menstrua-
tion. The suggestion that premenstrual weight gain is due to

•F

m.Ea/l

10 15 20 25 5

CYCLE DAYS

10 15 20 25

Fig. 1. Salivary sodium and potassium concentrations and Na/K

ratios in two cycles from a normal woman. In this and other figures

the upper curve is of the basal body temperature in °F. The black

shapes represent menstrual periods.

water and salt retention, mediated by oestrogens, is due to
Thorn, Nelson and Thorn (1938). Long and Zuckerman
(1937) postulated a role of adrenal salt-retaining hormones
in the electrolyte imbalance causing premenstrual fluid
retention.

In a recent investigation, Chesley and Hellman (1957)

Hormones and Water and Electrolyte Metabolism 83

studied 23 normal young women and found that in one-third
of them the weight was maximal during the premenstrual
eight days — in accordance with earlier writers. Closer
analysis, however, failed to substantiate the physiological
basis of such weight gains, since, when they did occur, they

m.E(i./l

AEL

10 15 20 25 30 35 40
CYCLE DAYS

Fig. 2. A long, but ovular, cycle in a normal woman.

were slight and were not repeated from one cycle to the next.
It was further shown that the incidence of premenstrual
weight gain was the same as would be expected on a purely
random distribution of weight gains throughout the menstrual
cycle. These workers also studied the salivary sodium and
Na/K ratios throughout the cycle; they were unable to find
any consistent pattern of variations such as would have been

G. I. M. SWYER

compatible with increased adrenal salt-retaining hormone
secretion during the premenstrual phase.

The present author's own limited studies on salivary and
urinary Na/K ratios in the menstrual cycle have been directed

Urinorjf
Nq/K
ratio

m.E<i./l.

(«)

28 10

CYCLE DAYS

(b)

36
AEL

Fig. 3. Urinary Na/K ratios in two normal women. In (a) there

appears to be a peak at about the time of ovulation. In (b) the

ratio appears to be higher during the second half of the cycle.

mainly towards an attempt at elucidating the basis for so-
called premenstrual tension which is widely supposed to
depend upon premenstrual salt and fluid retention (see, for
example, Greene and Dalton, 1953, who consider an increased
oestradiol/progesterone ratio to be largely responsible).
The findings are in agreement with those of Chesley and

Hormones and Water and Electrolyte Metabolism 85

Hellman (1957) in that no precise pattern of variation in
salivary or urinary sodium and potassium concentrations or
Na/K ratios, either in normal women or in those complaining
of premenstrual tension, has been discovered.

Fig. 1 shows two cycles from a normal woman: the Na/K

m.Ea/l-

Na/K urine

Na/K saliva

100

60
mEq./l.

40
20

5 10 15 20 25 27

}L CYCLE DAYS

Fig. 4. Urinary and salivary Na/K ratios compared in a woman
who experienced premenstrual tension.

ratio appears to be high at the start of both cycles and there
is a distinct fall (mainly due to increased potassium secretion)
at what may be judged from the basal temperature record to
be the time of ovulation in the second cycle.

Fig. 2 shows a long but ovular cycle in another normal
patient (A.E.L.) No convincing pattern is discernible.
Fig. 3b shows the urinary Na/K ratios in another cycle from

G. I. M. SWYER

patient A.E.L. If anything, the ratio is higher in the second
half of the cycle — i.e. sodium retention is less premenstrually.
In Fig. Sa, the urinary Na/K ratio appears to rise sharply just
at the time of ovulation — i.e. at the time of an oestrogen
peak, when, according to the usual view, the tendency should
be towards sodium retention.

m.Eq./l. -6

HB CYCLE DAYS

Fig. 5. Salivary sodium and potassium concentrations and ratios
in a woman who experienced premenstrual tension.

Figs. 4-6 relate to women who experienced definite pre-
menstrual tension. In Fig. 4 the salivary and urinary Na/K
ratios are compared. The latter (note that its scale is ten
times that of the salivary Na/K ratio) is much more variable
than the former, and neither shows any definite pattern.
Certainly there is no evidence of sodium retention premen-
strually. Fig. 5 shows the salivary Na/K ratios in another

Hormones and Water and Electrolyte Metabolism 87

patient; they fluctuate violently but show no evidence of
premenstrual sodium retention.

Fig. 6 shows three consecutive cycles in a patient who ex-
perienced quite severe premenstrual tension. In the first

17a-oxjfprogestcrone copronote ETHISTERONE '
-p-u IZSmjlM. 80 mq./doY. 50

rn.Eq./!.

LIS

10 20 26
CYCLE DAYS

20 26

Fig. 6. Three cycles in a woman who experienced premenstrual

tension. For explanation see text.

T = tension. D = dysmenorrhoea. 17a-Oxyprogesterone capronate

was injected at the point marked [ ; ethisterone was administered

orally in doses of 80 and 50 mg. per day where indicated.

cycle, the Na/K ratio in the saliva, was definitely lower, due
to a lower sodium concentration, in the second half of the
cycle. An injection of 125 mg. of 17a-oxyprogesterone capron-
ate intramuscularly failed to affect the symptoms, but when
ethisterone, 80 mg. daily by mouth, was started three days
later the tension disappeared, in spite of the Na/K ratio

88 G. I. M. SwYER

remaining low. In the next cycle, 50 mg. ethisterone daily
was given from the 14th day of the cycle. There was no
tension (though the succeeding period was painful). Yet
again the Na/K ratio appears to have been on the whole
lower in the second half of the cycle. In the third cycle, no
treatment was given ; the usual premenstrual tension appeared
but this time the premenstrual Na/K ratios were the highest
in the cycle.

It must be confessed that the writer does not know how to
interpret these findings, beyond concluding that they do not
provide evidence for theories currently held to account for
premenstrual tension and its relief (which, in the writer's
experience, is by no means invariable) with progesterone or
its analogues.

Pregnancy

In no physiological circumstances do such profound hor-
monal changes occur as in pregnancy. The output of oestrogens
rises some thousandfold, of progesterone ten to twentyfold,
and of adrenocortical and thyroid hormones to less impressive,
but still significantly increased levels. A new hormone,
chorionic gonadotrophin, found only in pregnancy, of foetal,
and therefore partly paternal origin — a "foreign protein", to
some extent — appears in the circulation immediately after
implantation, rises to striking levels by about the 60th day
of gestation and then as rapidly falls to about one-quarter
the maximum level during the remainder of pregnancy. The
sum total of these changes is to produce a substantial degree
of fluid and sodium retention in all pregnant patients. Oedema
is of course common; its association with hypertension, with
or without albuminuria to give pre-eclamptic toxaemia, is
also not uncommon. Toxaemia is, for the obstetrician, one
of the remaining major problems he has to face. Its patho-
genesis continues in obscurity, in spite of extensive research.

Only one or two aspects of this vast problem will be dealt
with here.

That water and sodium are retained in considerable

Hormones and Water and Electrolyte Metabolism 89

quantity during pregnancy has been shown by numerous
balance studies (see Rinsler and Rigby, 1957 for references).
Chesley and Boog (1943) found an increased thiocyanate
space in normal pregnancy, the increase being still greater
in pre-eclamptic toxaemia. From this it was concluded that
much of the sodium retention was due to expansion of the
extracellular fluid (ECF) compartment. However, Gray and
Plentl (1954), using a sodium isotope dilution technique,
found little change in the sodium space and total exchange-
able sodium in normal pregnancy. They observed a total gain
of some 500 m-equiv. of sodium during the last six months of
pregnancy, which they felt could be accounted for by the
products of gestation and the expanded maternal blood
volume. The maintenance of an essentially unchanged non-
pregnant sodium space during normal pregnancy, despite the
rise in plasma volume, suggests that there is little change in
ECF.

The gain of sodium and water, with maintenance of a
normal total-exchangeable sodium value and with an in-
creased thiocyanate space, provides indirect evidence that in
normal pregnancy there is an alteration of cell permeability
with an increased maternal storage of intracellular sodium and
water. The increased intracellular storage of sodium, together
with the foetal requirements, are a drain on the salt content
of the ECF, which, if uncorrected, would lead to diminution
of the ECF and plasma volumes. It has been demonstrated
by Bartter and co-workers (1956) that a fall in ECF volume
without change in tonicity leads to a rise in aldosterone
excretion. Such a rise in aldosterone excretion occurs in
pregnancy (Venning and Dyrenfurth, 1956; Venning et al.,
1957; Rinsler and Rigby, 1957) and may form part of a
homeostatic mechanism for maintaining the ECF volume and
meeting the loss of sodium from the ECF into the maternal
cells and foetal tissues by increased renal reabsorption.

In pre-eclamptic toxaemia, clinical examination alone is
sufficient to demonstrate the expanded ECF compartment.
Expansion of this compartment was shown by Bartter and

90 G. I. M. SwYER

co-workers (1956) to cause a fall in urinary aldosterone
excretion in normal persons. In the pre-eclamptic patients
studied by Rinsler and Rigby (1957), the aldosterone outputs
were considerably less than those at the same stage of normal
pregnancy and it was concluded that this was because of the
expanded ECF compartment. The output of aldosterone in
these toxaemic patients is less, for a given urinary Na/K
ratio, than in the normal group; yet despite the low aldo-
sterone output, sodium retention is maintained or increased.
This suggests that a mechanism other than that of aldo-
sterone secretion may be responsible for the sodium retention
of pre-eclamptic toxaemia.

Labour, especially if prolonged, is another aspect of
pregnancy in which electrolyte disturbance may assume im-
portance. Hawkins and Nixon (1957) have demonstrated a
consistent loss of plasma water and increase in plasma specific
gravity after only 20 hours of labour, indicating a state of
dehydration long before the appearance of clinical signs. In
addition, they found an increase in plasma sodium and a
decrease in chloride and potassium. This, they suggest, is due
to increased renal excretion of chlorides necessitated by the
disturbance of acid-base balance due to ketosis resulting from
shortage of available glycogen. After 48 hours of labour, a
striking fall in plasma potassium and in circulating eosinophils
was seen. This is consistent with increased adrenocortical
activity, such as is known to occur after surgical operations
(MacPhee, 1953). In labour, this fall in plasma potassium
may be particularly important because of its influence on
uterine contraction. It is very probable that potassium
depletion in long labours materially adds to the inefficiency
of an already inert uterus.

Changes in Steroid Metabolism in Ageing Men and
Women

The most extensive study of this subject has been made by
the Worcester group (Pincus et al., 1955). Certain of their

Hormones and Water and Electrolyte Metabolism 91

conclusions, of possible relevance to our main theme, are as
follows :

Oestrogens. In men, the output of oestrogens remains re-
latively constant with increasing age; in women, on the other
hand, the output declines between the ages of 40 and 60 years,
reaching a level somewhat below that of men and thereafter
remaining constant. Of the separate fractions, oestrone and
oestradiol decline slowly in men, accompanied by an increase
in oestriol which makes the total oestrogen output appear
constant; in women the most marked decline in earlier
decades is in oestriol output, the least marked in that of
oestrone, while in the later decades further small declines in
oestrone and oestradiol are accompanied by an apparent
increase in oestriol. Oestriol is a metabolite, not a secretory
product as the other two may be ; its increase with advancing
age may therefore be due to lesser destruction of secreted
oestrogen.

Neutral Steroids. The rate of decline of 17-ketosteroids is
similar in both sexes. The urinary ketonic androgens are
higher in men than in women and decline more steeply in the
former, particularly during the earlier decades. During these
decades, the decline of androgens is steeper than that of 17-
ketosteroids, so that with advancing age the ratio of 17-
ketosteroids to androgens increases, albeit somewhat irregul-
arly. Since the androgenic activity of the 17-ketosteroids is to
be attributed chiefly to androsterone, it follows that the rate
of production of androsterone (and presumably of its pre-
cursors) decKnes more rapidly than that of the less andro-
genically active 17-ketosteroids. Though this might have been
expected for men, as a result of declining testicular function, it
is perhaps more surprising in women and suggests a decrease
in output of either adrenal or ovarian androgens, or both.

The ratio of androgens to oestrogens is higher for men
than for women at all decades until the ninth.

The output of adrenal corticosteroids is rather higher in
men than in women at all ages and varies but little with age.
In contrast, the non-ketonic steroids, a mixture of substances

92 G. I. M. SwYER

of doubtful origin, part adrenocortical and part perhaps
gonadal, decline with age much as do the 17-ketosteroids.
Thus the outputs of the various classes of neutral steroids
change with age in dissimilar fashion. Close study of the data
suggests that the steroids of adrenal origin are less affected
by age than are those derived from the gonads, but that
adrenal steroids are not uniform in behaviour in this respect.

This differential behaviour is clearly shown by the various
a-ketosteroids. The 11-deoxy steroids, androsterone and
aetiocholanolone, decrease regularly and markedly with
advancing age, in both men and women. In contrast, the 11-
oxygenated 17-ketosteroids decrease much less markedly
with increasing age in both sexes. The 11-oxyaetiocholano-
lones decrease least of all ; these substances derive chiefly from
Cortisol and its metabolites.

To evaluate the significance for fluid and electrolyte control
of these hormonal changes in ageing men and women is
none too easy. The most important of the above-mentioned
hormones from this point of view are the adrenal cortico-
steroids, the output of which changes least. Beyond that
simple statement it is unsafe to venture.

Nothing has hitherto been said about the role of antidiuretic
hormone of the posterior pituitary in the control of electrolyte
and fluid metabolism under the various circumstances dis-
cussed above. Though it is true that numerous reports have
appeared in the literature implicating ADH in a variety of
pathological states characterized by oliguria and oedema, it is
the opinion of van Dyke, Adamsons and Engel (1955) that "the
assays used to support this belief are so grossly inaccurate as
to make valueless any conclusions that have been reached."
If we may accept that opinion, nothing further need be said.

REFERENCES

Bartter, F. C, Liddle, G. W., Duncan, L. E., Barber, J. K., and
Delea, C. (1956). J. din. Invest., 35, 1306.

BoNGiovANNi, A. M., and Eberlein, W. R. (1955). Pediatrics, Spring-
field, 16, 628.

Chesley, L. C, and Boog, J. M. (1943). Surg. Gynec. Obstet., 77, 261.

Hormones and Water and Electrolyte Metabolism 93

Chesley, L. C, and Hellman, L. M. (1957). Amer. J. Obstet. Gynec,

74, 582.
Dyke, H. B. van, Adamsons, K., and Engel, S. L. (1955). Recent

Progr. Hormone Res., 11, 1.
Gans, B., and Thompson, J. C. (1957). Proc. R. Soc. Med., 50, 929.
Gray, M. J., and Plentl, A. A. (1954). J. clin. Invest., 33, 347.
Greene, R., and Dalton, K. (1953). Brit. med. J., 1, 1007.
Hawkins, D. F., and Nixon, W. C. W. (1957). J. Obstet. Gynaec., Brit.

Emp., 64, 641.
Hubble, D. (1957). Lancet, 2, 301.

Long, C. N. H., and Zuckerman, S. (1937). Nature, Lond., 139, 1106.
MacPhee, I. W. (1953). Brit. med. J., 1, 1023.

PiNCUS, G., DORFMAN, R. I., ROMANOFF, L. P., RUBIN, B. L., BlOCH,

E., Carlo, J., and Freeman, H. (1955). Recent Progr. Hormone
Res., 11, 307.

RiNSLER, M. G., and Rigby, B. (1957). Brit. med. J., 2, 966.

Tanner, J. M. (1955). Growth at Adolescence. Oxford: Blackwell.

Thomas, W. A. (1933). J. Amer. med. Ass., 101, 1126.

Thorn, G. W., Nelson, K. R., and Thorn, D. W. (1938). Endocrinology,
22, 155.

Venning, E. H., and Dyrenfurth, I. (1956). J. clin. Endocrin. Metab.,
16, 426.

Venning, E. H., Primrose, T., Caligaris, L. C. S., and Dyrenfurth,
I. (1957). J. clin. Endocrin. Metab., 17, 473.

Wilkins, L. (1957). The Diagnosis and Treatment of Endocrine Dis-
orders in Childhood and Adolescence. Oxford. Blackwell:

DISCUSSION

Milne : Dr. Swyer rightly stressed the difficulties of showing the cyc-
lical changes in electrolytes in the menstrual cycle. But there is one
change which has been found by all those who investigated it, and that is
the cyclical changes in organic acid excretion in urine. There is both high
citrate and high a-ketoglutarate excretion at the time of o\ailation and an
abrupt fall immediately premenstrual. That is very constant indeed.
The easiest way to produce changes in these organic acids experimentally
is by variation in the systemic acid-base balance. Body alkalosis, not the
pH of the urine, tends to cause a rise in excretion and acidosis a fall. It
struck me that the previous investigations of acid-base balance in the
menstrual cycle had been rather contradictory. Some workers claim
there is a cyclical change in serum bicarbonate and others suggest a
change in pCOg, but this has been contradicted in other papers. Dr.
Swyer, have you any data in your metabolic studies which relate to
acid-base balance in normal menstrual periods?

Swyer: No, but I am very interested to hear of these changes.

Scribner: Dr. Swyer, were your studies made with constant intake?

Swyer: No, we did not attempt that because our subjects were ordi-
nary ambulant persons and it was rather difficult to restrict them much.
It was hoped that if the changes were going to be sufficiently distinctive.

94 Discussion

with people who were keeping themselves on an ordinary kind of regime
they would show up in spite of any day-to-day variations. That is
certainly a deficiency in our studies, but I do not think it entirely
invalidates them.

Adolph : I would like to go further and suggest that if the balances or
the outputs reflect variations of intake, they might be just as valuable as
variations of output would be on a constant intake.

Thaysen: How were the Na/K ratios in the saliva done?

Swyer: They were obtained by collecting saliva first thing in the morn-
ing, as nearly as possible at the same time each day, for a fixed length of
time (five minutes). In one series, the first five minutes was collected,
and in another series the first five minutes was discarded and the second
five minutes collected, as I understand there is something significant in
that. We were unable to see any difference at all when done in these two
ways. The saliva was collected just by spitting into a bottle, and the only
stimulation was that the patients were chewing paraffin.

Thaysen : Your Na/K ratios showed very great fluctuations and it was
difficult for me to assess any cyclical change. The Na/K ratio may vary
considerably just because of changes in the rate, and these variations
are quite independent of hormonal or other influences. I do not think
that the estimation of the Na/K ratio permits one to dispense with the
necessity for measuring secretory rate.

Swyer : Another thing we did was to measure the volumes which were
produced in this fixed time, and try to correct for the variations in
volume. It did not seem to make any difference at all, but I do agree
that some of the variables might have been inadequately controlled.

Thaysen: I believe that you might find the ratio very reproducible
when you use a standard secretory rate.

Talbot: Dr. Swyer, you mentioned something about a will-o'-the-wisp,
sodium diuretic hormone of adrenal origin. Do you believe in its exis-
tence, and if so, have you or any of those here a solid notion as to the
nature of the beast?

Swyer: I certainly have no solid notion. It is an idea that has been
mooted to account for the apparent inability of normal amounts of
sodium-retaining hormone to counteract the sodium loss. I know there
have been very active searches for it, and that large amounts are found
in the salt-losing type of adrenal hyperplasia.

Desaiilles : We have worked quite a lot on this problem and we have
got something which is derived from the adrenal, but what it is we do not
really know. Dr. Wettstein (1958. Iva., 29, in press) has just described
how he found it and how he is working on it, but that is as far as we
have got.

Adolph: I would like to raise a general problem which Dr. Swyer
brought up. How is one to judge whether in labour there is dehydra-
tion? All the criteria by which we can judge of the existence of dehy-
dration would be in a very fluctuating state at such a moment, and I
realise that Dr. Swyer was not making any positive statements about it.
Is there any way in which we can judge hydration, dehydration or super-
hydration, as transitory states of the organism?

Discussion 95

McCance: It is a question of definition. Do you mean by dehydration
a rise in the tonicity of the extracellular fluid due to an increase in the
quantity of sodium there, or do you mean by dehydration a decrease in
the total amount of water in the body?

Adolph : I think one type of dehydration would exist if we are satisfied
that there is no change in the concentration, but a decrease in the volume.

Fourman : In the patients with hypernatraemia who have an increased
volume of the extracellular fluid, is this increase appropriate or inappro-
priate to the requirements of the cells? Is the hypotonicity something
determined by the cells or something imposed upon them? Water in-
toxication with its characteristic symptoms (Weir, J. F., Larsen, E. E.,
and Rowntree, L. G. (1922). Arch, inter?!. Med., 29, 306) exemplifies an
inappropriate imposition; here the hypotonicity of the extracellular
fluid is accompanied by a swelling of the cells. An "appropriate" fall in
tonicity and increase in volume of the extracellular fluid is not associated
with these symptoms. The patients I mentioned earlier do not have
evidence of water intoxication.

After every stress, these patients with hyponatraemia returned to
their original low concentration of extracellular fluid. One feels that the
concentration is determined by the cells — a new steady state. We do
believe that this low concentration of the extracellular fluid must be the
result of a low osmotic pressure of the cells. There are obviously different
kinds of hypotonicity of the extracellular fluid, with and without symp-
toms of water intoxication. When there are symptoms, the cells are
swollen. Miss Leeson and I have been wondering whether a lack of
symptoms means the cells are not swollen, but merely hypotonic.

Swyer: I was referring to the opposite problem, namely that in labour
dehydration is accompanied by lack of potassium. Presumably this
increase in specific gravity of the plasma and the apparent loss of plasma
potassium would not be consistent with normal functioning of the cells.
It might therefore complicate still further the prolongation of labour.

Fourman: I do need convincing that any case of high serum sodium,
which these patients have, is not a case of dehydration.

You made another very fascinating statement which I would be very
glad to have amplified. Not being a paediatrician I do not see very much
of these adrenogenital syndromes. In Addison's disease, on the other
hand, I think it would be exceptional to find that the patient with a high
plasma potassium as a leading feature would die of a cardiac arrest as a
result. The general experience is to find that there is depletion of sodium
and, incidentally, but not clinically important, a raised serum potassium.
I wonder whether there is a different abnormality in a simple lack of
sodium-retaining hormone, which would account for the — to me rather
surprising — predominance of changes in the serum potassium. We have
one patient who is being maintained after adrenalectomy with cortisone
but because she has heart failure we are not giving her any sodium-
retaining hormone. To my astonishment our problems in her are those
of transient paralysis with very high plasma potassium (9 m-equiv./l.).

Swyer: High plasma potassium is one of the outstanding features, as I
think Dr. Talbot wfll also agree, in adrenal hyperplasia. In the salt-losing

96 Discussion

variety you get figures up to 16 m-equiv./l. or more with survival, though
not for long. I have not personally encountered that, but it has been
seen in the Hospital for Sick Children, Great Ormond Street.

Young: I think there is a much simpler explanation for the young
infant's rapid rise in serum potassium under conditions of stress. The
babies with the adrenogenital syndrome feel poorly and vomit ; therefore
they take in very little water and become dehydrated. Their blood urea
goes sky high at the same time as the serum potassium, and I think that
both are due to a rapid rate of cellular breakdown secondary to the dehy-
dration. I have no real proof of this, but all neonates becoming dehy-
drated very quickly show a high serum potassium level.

Milne: In these cases in babies with high serum potassium, is the
myocardium less sensitive to the hyperkalaemia? This could be inferred
from the work of Widdowson and McCance (1956. Clin. Sci., 15, 361) on
serum potassium in foetal pigs. Anyone with experience of hyper-
kalaemia in acute renal failure in adults would find very severe ECG
changes long before the serum potassium reached 10 m-equiv./l., and
death usually occurs very shortly after the potassium reaches 10
m-equiv./l. These high figures rather startle me; I would like to know
what is happening to the ECG during the period of hyperkalaemia.

Davson : The effect of potassium on the heart is linked with that of
calcium. It may be that over long periods the calcium might rise too
and tend to compensate for the raised potassium.

Scribner : We have made some studies on dogs and we could not greatly
increase the tolerance of the dog to hyperkalaemia by giving calcium.
Large doses of calcium increased tolerance no more than 1 m-equiv./l.

Adolph : This unanswered problem may leave us with an age difference
in the susceptibility of the heart to potassium.

Young: Once the potassium goes up towards 10 m-equiv./l. in babies
they become desperately ill and they sometimes die. I do not think they
have any better tolerance to these very high serum potassium levels than
adults. If you take infants with the adrenogenital syndrome off all their
treatment in order to confirm the diagnosis, which may be difficult in
young males, it is a very frightening experience to see the heart mis-
behaving with both the clinical effects and the ECG changes of hyper-
kalaemia.

McCance: You seem to have found something in which the infant
appears to react in the same way as the adult.

Young : Kerpel-Fronius has over-emphasized, perhaps, the differences
in the physiology of the infant, but his points are all intended to under-
line the differences in the effect of stress on the infant. In the treatment
of infants, sometimes the physiologist's point of view has made the
clinician oversensitive. He is frightened to give infants the treatment
that would be appropriate for adults because of the differences in the
physiology of the infant — whereas the clinical condition must and can be
treated effectively as long as the relatively minor differences in physio-
logy are borne in mind.

There is one point I should like to make which refers to the papers by
Prof. Adolph and Dr. Swyer. It seems to me, since the baby tends to

Discussion 97

retain water before birth and still excretes high amounts of oestrogens
after birth, that the persisting influence of the mother's oestrogens in the
early days of life might be the explanation of the infant's poor response
to a water load.

Adolph: This oestrogenic influence seems to me a very interesting
possibility. Has anyone any data on the influence of the maternal
hormones upon water balances or exchanges?

Fourman: Another question is whether oestrogens do inhibit the
diuretic response to an overdose of water in adults.

Heller: We have been injecting sex hormones of various kinds into
newborn rats to see whether we could influence the amount of total body
water or whether we could retard its decrease as the animals get older.
These are only preliminary experiments, but so far neither oestrogens
nor progesterone have produced any effect.

Swyer : Gans and Thompson (1957) produced evidence that the hydrae-
mic neonate might retain fluid as a result of maternal oestrogens, and
that as maternal oestrogens were excreted, the weight fell and then
remained reasonably constant. So it does look very much as though at
least some of the fluid retention in the newborn infant is due to maternal
oestrogen. I do not think that that can be held to account for the poor
handling of the water load since that extends for the best part of the
first year, or so I understand.

McCance: No, only about 14- days, I believe.

Adolph: I think I can clarify this contradiction of ages to some extent.
If you read the literature up to 1923 you learn that in the first year the
human infant excretes water very slowly. Such conclusions were re-
ported by Lasch (1922. Z. Kinderheilk., 36, 42) and others, with inade-
quate methods of collecting urine. The problem was clarified when
Ames (1953. Pediatrics, Springfield, 12, 272) did some well-controlled
studies on the excretion of a water load in infants of 1, 3, 7, and 14 days
of age. She showed that within 14 days the excretion of a water load
becomes 63 per cent as great, and within 90 days even as great as in
the adult human, if one bases water load and excretion on unit body
weight.

Swyer: Does this also apply to resistance to dehydration and handling
of electrolytes?

Adolph: I do not think we have any good data on the resistance to
dehydration. We know much less about hydropaenia than we do about
superhydration .

Widdozvson : Dr. Talbot, if a newborn baby and an adult were deprived
of all water, which would live longer, and why?

Talbot : The minimum daily water expenditure of the small infant rela-
tive to his body water stores is ordinarily about twice as great as it is in
the adult. For this reason, the infant usually tends to become dehydrated
when deprived of water about twice as fast as the adult. If only this
relationship is taken into account, one would expect the adult to outlive
the infant. However, the infant is born with a "surplus" of water,
equivalent to about one day's water requirements, which he is meant to
shed during the first few days of life. The shedding of this surplus fluid

AGEING — IV — 4

98 Discussion

tends to delay the development of serious dehydration during the neo-
natal period. This process coupled with other attributes might enable
some newborn infants to survive total thirsting as long as an adult.

Heller: I seem to remember that what Gans and Thompson showed
was that there was a decrease of body water in the infant which was corre-
lated with the excretion of maternal oestrogens, but this does not
establish a causal relationship.

Swyer : I think the point they were trying to make was that there was
this parallel fall in oestrogen and in body water with no change in adrenal
steroid output. They put two and two together and thought one was due
to the other.

Wallace: Dr. Swyer, what about the situation of a diabetic woman and
her baby? In a great number of instances there is a very intense water
retention.

Swyer : I can counter that by saying what about the baby of a pre-
diabetic mother? It shows just the same changes before the mother has
diabetes. I do not think we know why the prediabetic mother has a
large baby — there have been suggestions that it is due to excess growth
hormone secretion by the mother, but there is no very convincing
evidence.

Wallace : This kind of baby generally seems to have a great deal of
water in him — more water than in equivalent weight normal babies.

Swyer : That is very true. The baby is large but it is not postmature —
indeed, it behaves more like a premature.

Wallace: Is that an oestrogen effect?

Swyer : I do not think we know.

Wallace: Very often during these discussions the words "inefficient"
and "immature" have been used to describe the newborn infant. Mr.
Peter Rickham in his book, "The Metabolic Response to Neonatal
Surgery" (1957. Harvard University Press), develops the point of view
that the newborn infant is tolerant of adverse experiences such as
fasting, thirsting and surgical trauma. Despite the fact that the new-
born has an extra load of water in his body and a low metabolic rate he
does seem to have a certain toughness that at a later date is not so
evident. "Immaturity" and "inefficiency" may not be synonymous.

Bull: I should like to support that observation. We see enough burnt
children and adults to be able to assess their comparative mortality in
given degrees of burning. Although it is widely stated that children react
badly to burning — and burning largely involves the problems of fluid
and salt management that we are talking about today — we failed to
find any evidence that the small children react any worse than their
elder brothers and sisters (Bull, J. P. and Fisher, A. J. (1954). Ann. Surg.,
139, 269), The prognosis falls steadily from about 30 years to old age.
We do not frequently see babies during their first 14 days, but at least
in the first year there is no evidence that they react worse than older
children and adults.

GENERAL DISCUSSION

Richet: Dr. Thaysen, mercury poisoning is supposed to inhibit
some enzymic actions and possibly reabsorption by tubular cells. I
should like to know something about the secretion of sweat during
mercury poisoning and whether you found any differences due to that
substance?

Thaysen: I have not done any experiments of this kind myself,
but studies on mercurial diuretics have been performed, not on the
sweat glands but on the salivary glands, by White and co-workers
(1955. J. din. Invest., 34, 246). White showed that there was no
significant effect of mercurial diuretics on salivary sodium, potassium
or chloride excretion.

Richet: Dr. Desaulles has reminded me that during chronic
mercuric poisoning, acrodynia for instance, there is an increase in
sweating.

Thaysen: That might be due to a cerebral effect of chronic mer-
cury poisoning rather than to a local effect of the mercury directly
on the glands.

Davson: It is rather a fortunate accident that the mercurials are
diuretics and that they have that specific action on the kidney
tubules. If you were to try and raise the mercury concentration in
the blood so as to put some specific mechanism apart from the kidney
out of action, you would kill the person anyway, because mercury
would interfere with so many other metabolic reactions if you really
could get a reasonable blood level of it for any length of time. So I
think investigation of it is out of the question.

Hingerty: Is there any evidence that plasma magnesium goes up
at the same time as plasma potassium? In hypersecretion of aldo-
sterone, plasma magnesium has been reported as being decreased in
a few cases. We found some years ago (Conway, E. J., and Hingerty,
D. J. (1946). Biochem. J., 40, 561) that when plasma potassium went
up in adrenalectomized rats it was accompanied by an almost
parallel increase in the plasma magnesium ; cellular magnesium also
went up but rather less.

Richet: We have made determinations of plasma magnesium in
more than 200 patients during acute and chronic renal failure.
During acute renal failure there is always an increase in plasma
magnesium concentration. Our technique with yellow titanium gives
normal values of 1 • 5 -1 • 7 m-equiv./l. In acute renal failure we some-
times get 3 -0-3 -5 m-equiv./l. serum magnesium. In contrast, serum

100 General Discussion

potassium is increased in only 20 per cent of our patients. We have
noticed that the serum magnesium increases more rapidly and more
frequently than serum potassium. During chronic renal failure we
have found exactly the same thing. The serum magnesium begins to
increase when the urea clearance is below 15 ml./min., even if serum
potassium remains normal for a long time.

McCance: That agrees with observations Miss Watchorn and I
made in 1932 (Biochem. J., 26, 54). We generally found that the
serum magnesium was high in chronic renal failure and indeed
searched for such cases when we wanted high values for our ultra-
filtration experiments.

Scrihner: I want to bring to your attention the work done by Dr.
Konrad Buettner, professor in the Division of Climatology at the
University of Washington, Seattle (1953. J. appl. Physiol., 6, 229).
His observations bear on the sweating data that we have heard
and also on considerations of cellular tonicity. If you study water
transfer through skin and exclude sweating, the normal human skin
will absorb water into the skin against an osmotic gradient that is five
times isotonic. In other words if you expose it to increasing concen-
trations of sodium chloride solution, the skin will take up water until
a concentration which is five times isotonic is reached. The mecha-
nism of absorption is not known and there has been no work to
elucidate why this occurs. The rate of absorption in an adult human is
about 20 ml./hr. for the total skin, and is correspondingly less for
smaller areas of skin. Such factors as the storage phenomenon etc.
have been excluded by the methods of undertaking this study. The
practical implications of this are perhaps of interest. For example, at
low rates of sweating, data on electrolytes in sweat may be abnor-
mally high throughout due to this absorption, and there is some chance
that by the proper control of conditions you may be able to absorb
water in survival experiments at sea, since sea water is only three
times isotonic.

Davson: What happens to the water? Is it immediately carried
away by the capillaries?

Scrihner: Yes. Deuterium studies have shown that. Ten — twenty
ml./m.^hr. are the actual figures for the absorption.

Talbot: In the last war in survival ration studies we immersed
some very dehydrated volunteer subjects in the equivalent of sea
water for an hour or so, and were unable to detect any absorption of
water through the skin, using changes in total body weight as an
index; so this is very interesting.

Scrihner: The problem of controlling sweating during these studies
is a difficult one and this investigator has gone to great lengths to
control this variable.

General Discussion 101

Davson: Was there also a control on whether salts were being
absorbed, Avhen they say that five time isotonicity would have stopped
it? Just the fact that the skin absorbs water does not mean that
salts are not absorbed as well.

Scrihner: The concentration of salts goes up in the outside fluid.
Also water is absorbed from capsules containing crystallized salts
such as sodium and calcium chloride which are separated from the
skin by a layer of air. The type of salt used determines the amount
of water vapour in the air. The results by this technique agree with
the hypertonic solution studies.

Hingerty: What salts have been investigated?

Scrihner: Sucrose, potassium chloride, sodium chloride. The
phenomenon is believed to be purely an osmotic effect.

Borst: Before the war Viennese clinicians reported on considerable
absorption of water by the skin in heart failure. The prognosis could
even be determined by studying the rate of absorption. Dutch
workers repeated the experiments but could not demonstrate any
absorption at all. However an absorption of 20 ml./m.^/hr. is less
than was expected according to the Viennese papers and it is possible
that a more exact technique would have given positive results.

BODY WATER COMPARTMENTS
THROUGHOUT THE LIFESPAN*

H. Victor Parker, Knud H. Olesen, James McMurrey
and Bent Friis-Hansen

Surgical Service and Laboratories of the Peter Bent Brigham Hospital,

Harvard Medical School, Boston, and

Queen Louise's Childreri's Hospital, Copenhagen

Our first knowledge of the composition of the body was
acquired during the last decades of the nineteenth century.
The methods used were desiccation and chemical analysis
which allowed the determination of the contents of water and
electrolytes in carcasses or in single organs. With the recent
introduction of the dilution methods a new field of study has
grown up based on the in vivo measurements of the total
quantities of body water and its partitions. Direct dilution
methods are now available for the measurement of total body
water and of the extracellular water. The intracellular water
is calculated as the difference between total body water and
extracellular water and is thus a derived value (Moore et al.,
1956).

A few comments should be made about the methods and
the evaluation of the measurements. In the material presented
the total body water has been determined as the volume of
dilution of deuterium oxide. In the children the extracellular
water has been measured as the volume of dilution of thio-
sulphate and in the adult groups as the volume of distribution
of radioactive bromide corrected for red cell bromide, for the
relative water contents of plasma and interstitial water, and
for the Donnan effect. As the volume of dilution of thio-

* This work was supported by a grant from the United States Atomic
Energy Commission to the Peter Bent Brigham Hospital (AT-(30-l)-733),
and by the Surgeon General, Department of the Army, through a contract
(DA-49-007-472) with Harvard Medical School and sponsored by the Com-
mission on Liver Disease, Armed Forces Epidemiological Board.

102

Body Water Compartments throughout Lifespan 103

sulphate is smaller than the corrected volume of dilution of
radiobromide the values for extracellular volumes will not be
directly comparable for the children and the adults. The same
will apply to the calculated intracellular water. All the
methods used are reproducible within the 5 per cent range.
As the absolute quantities measured are difficult to compare
from one individual to another it has become customary to
express the results as relative values. The standard of refer-
ence used is the body weight as this standard in our experience
has been the most simple. In the interpretation it is important
to realize that a rather large biological variation appears
within groups of the same age and sex.

Although the study of the body water compartments
throughout the lifespan is still fragmentary, certain trends
in relation to age and sex have appeared. It will be the
purpose of this paper to outline these features in a description
of the body water compartments during the three main phases
of life : growth, maturity and ageing.

Growth

Growth implies a variety of fundamental processes: cell
multiplication, increase in cell size, accumulation of extracel-
lular material, increase in fat and minerals.

The alterations in the body water compartments during
growth have been studied by Friis-Hansen (1956). From a
series of 93 normal children studied with deuterium oxide,
with thiosulphate or with both, a series of 31 individuals with
simultaneous measurements of all three water compartments
will be presented.

It appears from Table I that the absolute amounts of total
body water, of extracellular water, and of intracellular
water demonstrate an increase throughout infancy and child-
hood. It is seen that the intracellular water rises more
markedly than the extracellular water.

In Table II the three measurements are given as percentages
of body weight. The total body water shows a relative de-
crease throughout infancy and childhood with a most marked

104 H. Victor Parker, et al.

decrease during the first two years of life. The relative
decrease in extracellular water is more marked than the
decrease in total body water. The intracellular water demons-
trates about the same relative value throughout childhood.
It should be mentioned that no sex difference appeared in this

Table I

Body water compartments in children, absolute values

Age

Water compartments in
TBW ECW

litres

ICW

Number of
subjects

0-11 days

2-65

1-45

1-20

Ll-180 „

310

1-42

1-68

^2 years

5-40

2-36

3-04

2-7 „

8-96

3-40

5-56

7-14 „ 27-62 7-52 2010 1

series. The tendencies found in this group are similar to the
findings in the larger group including cases with single measure-
ments of total body water or of extracellular volume. A
statistical analysis of the larger group has shown that most of
the differences between the age groups are significant.

Table II

Body water compartments in children, relative values

Age

Values in per cent <
TBW ECW

of body

weight
ICW

Number oj
subjects

0-11 days

76-4

41-6

34-8

.1-180 „

72-8

34-9

37-9

1-2 years

62-2

27-5

34-7

2-7 „

65-5

25-6

36-9

7-14 „

64-2

17-5

4G-7

The relative decrease in total body water with advancing
age indicates a relative increase in total body solids, i.e. cell
solids, mineral solids and body fat. The total body solids
thus represent the fraction of the body which demonstrates
the highest degree of absolute increase during growth. This
increase in total body solids represents one important facet
in the alterations in body composition with advancing age.

Body Water Compartments throughout Lifespan 105

Within the body water compartments the measurements
with thiosulphate demonstrated a relative decrease of extra-
cellular water during growth. A similar degree of decrement
in extracellular space with advancing age has been reported
by Ely and Sutow (1952) using the thiocyanate method, and
by Cheek (1954) using the corrected bromide space. The
relative values for intracellular water in the series presented
stayed about the same throughout infancy and childhood.
No similar investigations are available in the literature, but it
is interesting that Corsa and co-workers (1956) found that
the total exchangeable potassium as related to body weight
stayed the same throughout infancy and childhood. As about
95-98 per cent of the exchangeable potassium must be present
within the cells their results can be taken as corroborative
evidence for Friis-Hansen's (1956) findings of the relative
constancy of the intracellular water.

An alteration in the interrelationship between the extra-
cellular and intracellular water during growth thus appears.
When the extracellular water is expressed as a percentage of
total body water the extracellular compartment decreases
from 55 per cent in the youngest group to 38 per cent and 28
per cent in the two oldest groups, again reflecting the relative
decrease of the extracellular water. This altered relationship
between the extra- and intracellular water is another impor-
tant facet in the body compositional changes during growth.

The alterations during growth could be produced in two
ways : (1) They could be due to a proportional alteration in the
composition of all tissues, or (2) they could be caused by an
intracellular increase in some tissues whereas other areas would
develop in a different way.

Histochemical studies are helpful in the interpretation of
this problem. Kerpel-Fronius (1937) found in studies of
muscular tissue from human newborn babies and from adults
a relative increase in intracellular phase during growth,
whereas such a change did not appear in the skin or in the
central nervous tissue. Kerpel-Fronius also drew attention
to the fact that the total muscle water had increased from

106 H. Victor Parker, et al.

29 per cent of total body water in the newborn baby to 51 per
cent of total body water in the adult and he stressed that an
increase in total muscular tissue rich in intracellular phase was
a prominent feature in the alterations in body composition
during growth.

Yannet and Darrow (1938) found in their studies of cats a
relative increase in intracellular phase during growth in
muscles, whereas only very small alterations appeared in
liver tissue or in brain tissue. In studies of growing chickens
Barlow and Manery (1954) reported a similar relative increase
in the intracellular phase in muscular tissue.

It appears from these studies that the alterations measured
with the dilution methods must be results of a development
varying quantitatively and qualitatively from one tissue to
another.

In conclusion the alterations in body composition during
growth can be described as a disproportional increase in total
body solids, total body water, extracellular water, and intra-
cellular water. When the values are related to body weight
the following trends are seen during growth: a decrease in
total body water, an increase in total body solids, a decrease
in extracellular water, and a relative constancy in intracellular
water. When the water compartments are related to total
body water the trend is for a relative decrease in extracellular
water and a relative increase in intracellular water.

Maturity

The body water compartments in adults will be described
with particular reference to the sex difference.

The material presented comprises ten normal males and
ten normal females at ages from 23 to 54 years, average age
in the middle thirties. The series was studied by H. V.
Parker in Dr. Francis D. Moore's laboratory, Peter Bent
Brigham Hospital, Boston (McMurrey et al., 1958). The
methods applied were : total body water was determined with
deuterium oxide ; the extracellular water was measured as the
radiobromide space, which was corrected for red cell bromide.

Body Water Compartments throughout Lifespan 107

for the relative water contents of plasma and interstitial
water, and for the Donnan effect. As the extracellular water
according to the method applied here shows a higher normal
value than is obtained with the thiosulphate method the
results for extracellular and intracellular water in this series
will not be directly comparable to the findings in the group of
children.

Table III
Body water compartments in adults, absolute values

Sex

Males
Females

Age range

23-54
23-51

Body weight
kg-
72-5
59-8

Water compartments in litres
TBW ECW ICW

38-9

28-7

16-8
13-3

22-1
15-4

The absolute average values for total body water, extra-
cellular water and intracellular water appear in Table III. As
expected all values are higher in the males than in the females,
corresponding to the higher average weight in the male group.
Most of the difference in total body water is accounted for by
the difference in intracellular water.

Table IV

Body water compartments in adults, relative values

Sex
Males

Females

Age
23-54

23-51

Weight kg.
72-5

59-3

Water compartments in per cent of body
iveight with standard error of the mean

TBW
54-3

±1-39
48-6

±1-47

ECW

23-4
±0-64

22-7
±0-54

30-9
±0-89

25-9
±0-96

In Table IV the average values are given in per cent of
body weight. The males contain 54-3 per cent of total body
water whereas the females contain 48-6 per cent. This dif-
ference is statistically significant (P=0-01). The relative
values for the extracellular water are very close to one another.
The intracellular water amounts to 30 • 9 per cent in the males
and to 25 • 9 per cent in the females. This difference is statistic-
ally significant (0-01 > P>0-001).

108 H. Victor Parker, et al.

The similarity of the relative values for the extracellular
water and dissimilarity of the relative intracellular water
volumes in the two sexes gains further support from other
parts of the same study. As is seen in Table V, simultaneous
studies of total exchangeable sodium and potassium were
carried out in these patients according to the method des-
cribed by Moore and co-workers (1956). The total exchange-
able sodium which was determined through an independent
measurement demonstrates relative values very similar in the
two sexes. As about 85 per cent of the total exchangeable
sodium can be accounted for in the extracellular space the
findings can be taken as supportive evidence for the correct-
ness of the very close relative values for the extracellular

Table V

Body water compartments and total exchangeable
electrolytes in adults

Sex Values related to body weight with standard error of the mean

ECW Cle Nae ICW Ke

(%) {m-equiv.lkg.) {m-equiv.jkg.) (%) (m-equiv.lkg.)

Males 23-4 29-3 39-5 30-9 480

±0-64 ±0-71 ±1-06 ±0-89 ±1-38

Females 22-7 28-6 38-3 25-9 39-4

±0-54 ±0-92 ±109 ±0-96 ±1-40

water in the two sexes. The relative values for the total
exchangeable potassium which was determined independently
of the intracellular water demonstrate a pattern very similar
to the findings of the intracellular water. In both measure-
ments the females have a relative value about 20 per cent
below the males. As 97 per cent of the total exchangeable
potassium must be within the cells this finding can be taken
as evidence for the correctness of the measurements of the
intracellular water. It is worth mentioning that a calculation
of the average intracellular potassium concentration in the
two sexes results in very similar values: 152 m-equiv. per
litre intracellular water in the males and 149 m-equiv. per
litre intracellular water in the females, and thus indicates that
no difference in cellular composition exists in the two sexes.

Body Water Compartments throughout Lifespan 109

It appears from the series that males have a higher relative
content of body water than females, confirming the results
with the deuterium oxide method reported by Edelman and
co-workers (1952a) and Ljunggren, Ikkos and Luft (1957).

This sex difference in body composition does not appear to
be due to a difference in the relative amounts of extracellular
water in the series presented. The extracellular water repre-
sented 22 • 7 per cent of body weight in the females and 23 • 4
per cent in the males. This similarity in the relative values
for extracellular water is in agreement with the findings of
Cheek (1953), of Reid and co-workers (1956) and of Ljunggren,
Ikkos and Luft (1957) using the corrected bromide space, of
Ljunggren, Ikkos and Luft (1957) using the thiosulphate
method, and of Griffin and co-workers (1945) using the
thiocyanate method.

The lower relative content of total body water in females
as compared to males in the series presented is due to a
relatively lower content of intracellular water in the females.
A similar difference in the content of intracellular water
appears in the series studied by Ljunggren, Ikkos and Luft
(1957) in which the intracellular water was calculated on the
basis of an extracellular space measured with radiobromide as
well as with thiosulphate. Further evidence of the relatively
lower content of intracellular water in females compared to
males is present in the consistent findings of a lower relative
amount of total exchangeable potassium in females as reported
by Edelman and co-workers (19526), Arons, Vanderlinde and
Solomon (1954), Blainey and co-workers (1954), Sagild (1956),
and Ljunggren, Ikkos and Luft (1957).

The lower relative body water in females indicates a higher
relative content of total body solids in females than in males.
As the relative amount of intracellular solids, as judged by
the relative values for intracellular water and total exchange-
able potassium, must be assumed to be lower in females than
in males, it seems justified to conclude that females must have
a higher relative amount of fat (or other non-cellular solids)
than males.

llO M. Victor Parker, et ah

When the body water compartments are related to total
body water as a standard of reference another sex difference
appears. In males the extracellular water accounts for 43 per
cent of total body water and in females for 47 per cent,
whereas the intracellular water amounts to 57 per cent of
total body water in the males and 53 per cent in the females.
The difference between these ratios is statistically significant
(P< 0-001). This difference in the distribution of the total
body water between the extracellular and intracellular
compartments can be explained as the result of a higher
development of tissues rich in intracellular material and
relatively poor in extracellular phase, such as muscle tissue,
in the males.

In conclusion: the sex difPerence in body composition is
outlined as a higher relative content of total body water, a
higher relative content of intracellular water and a lower
relative amount of total body solids and especially of body
fat, in males than in females. The total body water is distri-
buted with a lower extracellular fraction and a higher intra-
cellular fraction in males than in females.

Ageing

Our experiences in the old age group are based upon the
investigations carried out in seven apparently normal males
with an average age of 75 years and seven apparently normal
females with an average age of 68 years. This group was
studied in Dr. Francis Moore's laboratory (Parker, Olesen and
Moore, 1958). The methods used were the same as those
applied to the younger adults.

The essential findings in the old age group are presented in
Table VI.

A comparison between younger and older adults reveals the
following findings : total body water decreases from 54 • 3 per
cent to 50 • 8 per cent in males and from 48 • 6 per cent to 43 • 4
per cent in females. The extracellular water rises slightly in
males and decreases slightly in females. The intracellular
water decreases from 30 • 9 per cent to 25-4 per cent in males

Body Water Compartments throughout Lifespan 111

and from 25-9 per cent to 22-4 per cent in females. The
differences mentioned are not statistically significant except
for the decrease in intracellular water in males (0-01> P>
0001).

The tendency to a decrease in the relative values for total
body water found in both sexes is mostly due to a decrease in
intracellular water. From an unpublished study of Dr. N. W.
Shock (1957), in which the antipyrine space and the thio-
cyanate space were measured in a larger group of males, the
following data are of interest. A comparison of 23 subjects

Body

Table VI

WATER COMPARTMENTS IN YOUNGER AND IN OLDER ADULTS.
RELATIVE VALUES

Water compartments in per cent of body
weight with standard error of the mean

Sex
(Number)

Age

Weight
kg.

TBW

ECW

ICW

Males

23-54

72-5

54-3

23-4

30-9

(10)
Males

71-84

68-1

±1-39
50-8

±0-64
25-4

±0-89
25-4

Females

23-51

59-3

±1-55
48-6

±1-36
22-7

±0-58
25-9

(10)
Females

61-74

63-9

±1-47
43-4

±0-54
21-4

±0-96
22-4

(7)

±1-32

±0-45

±0-97

aged 40-49 and 32 subjects aged 70-79 showed that the
values for total body water related to body weight decreased
from 54-8 per cent to 50-9 per cent, and those for the cal-
culated intracellular water decreased from 30-5 per cent to
25-1 per cent. The extracellular water changed from 24-3
per cent to 25-8 per cent only. The same pattern of a slight
decrease in total body water and in intracellular water
related to body weight was seen in a male series studied by
Olbrich and Woodford-Williams (1956). Sagild's findings of a
decrease in total exchangeable potassium in the old age
groups of both sexes can also be interpreted as evidence of a
decrease in the intracellular phase related to body weight
(Sagild, 1956).

112 H. Victor Parker, et al.

From the uniform tendencies in these materials it seems
reasonable to conclude that the slight decrease in total body
water and in intracellular water related to body weight
reflects real alterations in the body composition with advanc-
ing age. With the decrease in the relative value for total
body water there is a relative increase in total body solids.
As the intracellular phase shows a relative decrease the
increase in total body solids must be assumed to be caused by
a relative increase in non-cellular solids, most probably body
fat.

The alterations in the extracellular water related to body
weight are not quite uniform and the changes are small. It is
of interest that extracellular water expressed as per cent of
total body water in both sexes shows a rise from younger to
older subjects, in the males from 43 per cent to 50 per cent, in
the females from 47 per cent to 49 per cent. This tendency is
also seen in Shock's and in Olbrich and Woodford-Williams'
series and indicates an altered relationship between the
extracellular and intracellular water.

In conclusion: the alterations in body composition in the
old age group as compared to younger adults were rather
small. A tendency to a relative decrease in total body water
and in intracellular water and a relative increase in total body
solids, most probably body fat, was found. The extracellular
water stayed essentially the same in values related to body
weight, but demonstrated a tendency to increase in per cent
of total body water.

Acknowledgement

We express our gratitude to Dr. Francis D. Moore, Moseley Professor
of Surgery, Harvard Medical School, and Surgeon-in-Chief, Peter Bent
Brigham Hospital, Boston, for permission to present data from his
laboratory.

REFERENCES

Arons, W. L., Vanderlinde, R. J., and Solomon, A. K. (1954). J.

din. Invest., 33, 1001.
Barlow, J. S., and Manery, J. F. (1954). J. cell. comp. Physiol., 43, 165.
Blainey, J. D., Cooke, W. T., Quinton, A., and Scott, W. C. (1954).

Clin. Sci., 13, 165.

Body Water Compartments throughout Lifespan 113

Cheek, D. B. (1953). J. appl. Physiol., 5, 639.

Cheek, D. B. (1954). Pediatrics, Springfield, 14, 5.

CoRSA, L. Jr., Gribetz, D., Cook, C. D., and Talbot, N. B. (1956).

Pediatrics, Springfield, 17, 184.
Edelman, I. S., Haley, H. B., Schloerb, P. R., Sheldon, D. S.,

Friis-Hansen, B. J., Stoll, G., and Moore, F. D. (1952«). Surg.

Gynec. Obstet., 95, 1.
Edelman, I. S., Olney, J. M., James, A. H., Brooks, L., and Moore,

F. D. (19526). Science, 115, 447.
Ely, R. S., and Sutow, W. W. (1952). Pediatrics, Springfield, 10, 115.
Friis-Hansen, B. J. (1956). Changes in Body Water Compartments

during Growth. Copenhagen: Munksgaards.
Griffin, G. E., Abbot, W. E., Pride, M. P., Muntwyler, E., Mantz,

F. R., and Griffith, L. (1945). Ann. Surg., 121, 352.
Kerpel-Fronius, E. (1937). Z. Kinderheilk., 58, 276.
Ljunggren, H., Ikkos, D., and Luft, R. (1957). Acta endocr., Copen-
hagen, 25, 187.
McMurrey, J. D., Boling, E. A., Davis, J. M., Parker, H. V., Mag-
nus, I. C, and Moore, F. D. (1938). Metabolism, in press.
Moore, F. D., McMurrey, J. D., Parker, H. V., and Magnus, I. C.

(1956). Metabolism, 5, 447.
Olbrich, O., and Woodford-Williams, E. (1956). In Experimental

Research on Ageing, p. 236, ed. Verzar, F. Basle : Birkhauser.
Parker, H. V., Olesen, K. H., and Moore, F. D. (1958). Surgical

Forum, American College of Surgeons. Philadelphia: W. B.

Saunders, in press.
Reid, a. F., Forbes, G. B., Bondurant, J., and Etheridge, J. (1956).

J. Lab. clin. Med., 48, 63.
Shock, N. W. (1957). Personal communication.
Sagild, U. (1956). Scand. J. clin. Lab. Invest., 8, 44.
Yannet, H., and Darrow, D. C. (1938). J. biol. Chem., 123, 295.

DISCUSSION

Hingerty : Are these differences in the intracellular water related to the
proportion of functional muscular tissue? Have you any comparative
data for women athletes, for example?

Olesen : We have no measurements on muscle mass, but we assume that
there may be differences due to variations in muscle mass.

Black: Have you analysed your subjects in terms of their occupation?

Olesen: We have not investigated that, but it could probably be done.
I have the impression that muscular females have higher exchangeable
potassium relative to body weight than the fat ones.

Kfecek : Babies of six months have the highest total body water. Have
you seen any relationship to the weaning of these babies at this period?

Olesen : I have no data on this question.

Widdowson : Dr. Olesen, can you tell us approximately at what age the
fat-free body tissue of the baby becomes adult, or chemically mature, as
regards its intracellular-extracellular relationships?

114 Discussion

Olesen: It appears from Dr. Friis-Hansen's material that chemical
maturity occurs about the age of twelve months.

Widdowson : Have you made any calculations of the body fat at differ-
ent ages?

Olesen : I have tried to compare the different groups and it seems that
there is a relative increase in body fat throughout childhood. It is a
slight one but it does exist if we accept that all the non-cellular solid
changes are changes in body fat. This calculation is quite apart from
possible changes in body minerals and I do not know to what extent
these would interfere.

Borst : Is there any relationship between the creatinine output and the
intracellular fluid?

Olesen : In the original description of the method of determination of
total exchangeable potassium from Dr. Moore's laboratory (Corsa et al.
(1950). .7. clin. Invest., 29, 1289), a relationship was found between
creatinine excretion and the amount of total exchangeable potassium.
This has not been studied in this particular series.

Heller : How far is it justifiable to take mean figures from ten young
adult females without considering the role of the menstrual cycle? Have
you had enough cases to pay attention to this point?

Olesen: No, but it would appear from what Dr. Swyer mentioned
yesterday that it would not mean very much, as the latest view is that
these body weight changes are randomly distributed throughout the
menstrual cycle.

Shock : It seems to me that we have two possible interpretations of this
age reduction in intracellular water. The interpretation I favour is that
the reduction in total intracellular water is a reflection of the loss of
functional cells or the loss of protoplasm, rather than a change in the
water concentration of the remaining protoplasm. Have we any other
evidence that would make one interpretation more probable than the
other?

Davson : I think that is a very sound point, because a cell can change in
size without there being a change in the relative value of the water or
solid contents of the organism. Is there any change in the histological
appearance of old tissue that would indicate whether the cells had be-
come smaller or larger?

Shock : I cannot answer this question and must refer it to the patho-
logist or histologist. In our own work we have been looking for indices
of the total amount of man left functioning at a given age. Surface area
leaves much to be desired as a criterion, but one can account very nicely
for the age reduction in basal metabolism in terms of cellular loss if body
water is used as the index. In other words, although the basal metab-
olism per unit of surface area goes down with age, the basal oxygen con-
sumption per unit of intracellular water does not change at all with age.
When you try this with renal function data, renal plasma flow per unit
of body water goes down just as much as the renal plasma flow per unit of
surface area.

Scribner: Total exchangeable potassium might possibly be a good
parameter for this measurement of protoplasm.

Discussion 115

Dr. Maclntyre of Hammersmith has made an interesting study (to be
published), in which he finds a direct correlation between either body
weight or body fat and the extracellular space as measured by bromide.
The implication of this correlation is that fat tissue has an extracellular
space relationship to its weight which is the same as that of non-fat
tissue. This relationship is consistent with the data presented by Dr.
Olesen.

Bull: This is in contradistinction, for instance, to the blood volume,
which is a poor function of total body weight or of fat, and is closely
related to lean body mass. I would suggest that blood volume and meta-
bolic rate are related to intracellular water and possibly to exchangeable
potassium rather than to extracellular water.

McCance: Do those who see many old people professionally get the
impression that they are fatter than middle-aged people? There are
often indications that in old age man is rather wasted and has not much
fat ; but perhaps his shrinkage is more in protoplasm than in fat.

Swyer: One possible interpretation is that fat people do not live so
long; most of the really old people are pretty thin.

Fejfar: My experience is that older people usually eat more than they
did when they were middle-aged — they eat more than they need to.

Shock : I have no information on what they eat, shall I say, spontan-
eously. But I do know that on many metabolic balance studies that we
carried out on middle-aged and older people, one of our primary prob-
lems was to get our older people to consume the diets which were eaten
by the middle-aged control group without much difficulty. The varia-
tions were usually in the protein intake, particularly when we tried to
increase it by adding meat three times a day. A great deal of coaxing was
needed to get our older people to consume diets of this kind.

Fourman: Dr. Olesen, Dr. Shock and others suggest from their data
that, in adults, the percentage of total body water that is extracellular
water increases with age. I would like to try to visualize what this means.
One should not think of the extracellular fluid as a bag of water. Ob-
viously about a quarter of it is accounted for by the plasma volume, and
perhaps a fifth by the lymphatic fluid ; but what about the rest? The rest
is a film of fluid which surrounds the cells and the fluid of the collagenous
tissue of the body. If the cells, the muscle cells in particular, without
changing in number, shrink with age, then one would get a change in the
relation between the volume of the muscle cells and the amount of fluid
bathing them, since a single cell when it shrinks increases its ratio of
surface area to volume. I wonder whether this is the explanation of the
increase in ratio of extracellular to intracellular water with age : a shrink-
age in each cell without change in the total number of the cells, but each
cell still having to have its film of fluid surrounding it.

THE EFFECT OF VARIABLE PROTEIN AND

MINERAL INTAKE UPON THE BODY

COMPOSITION OF THE

GROWING ANIMAL *

William M. Wallace, William B. Weil and
Anne Taylor

Department of Pediatrics, Western Reserve University School of
Medicine and Babies' and Children'' s Hospital, Cleveland, Ohio

The quantities of various nutritive substances in the
growing body at any given point represent the metabohc
integration of the daily additions to the body from the diet
from the time of conception. Measurement of the rate or
quantity of addition may or may not measure the nutritional
requirement for a given substance. Whether it does or not
will depend upon the requirement for synthesis and metabolic
transformation and upon the possibility of the body being
able to store the substance. Thus, the day-by-day accretion
of fat or glycogen cannot measure a requirement but the
accretion of protein and mineral may do so, once any capacity
for storage is exceeded. Information concerning requirements
for growth is usually obtained by measurements of external
balance for variable periods of time. The information ac-
quired concerning the requirements for growth and the com-
position of growth by this method is often strangely contra-
dictory and always incomplete. Much of the data so ob-
tained indicate that extensive storage of dietary components
occurs, or that the composition of the body tissues is variable
and dependent upon quantity and quality of the intake.

* This work was supported by grants from the Baker Laboratories, Inc.,
Cleveland, Ohio and the National Institute of Arthritis and Metabolic Diseases
of the National Institutes of Health, United States Public Health Service,
Grants numbers G-3754 and A-1032.

Presented in part at the meeting of the American Pediatric Society, May
9-11, 1956, Buck Hill Falls, Pennsylvania.

116

Effect of Variable Intake on Body Composition 117

That body tissues can vary significantly in composition except
under extreme conditions is difficult to reconcile with present-
day knowledge of tissue composition.

The experiments to be described here were undertaken in
an attempt to characterize the effects of high and low mineral
and protein intakes, in various combinations, upon the body
composition of the growing albino rat as determined by
direct whole body analysis. Previous work using this method
of approach has been concerned with single constituents and
not with the interrelationships of all of the components. The
data indicate little variability in composition for the collective
soft tissues of the body. The only intake-dependent relation-
ship that seems of significance is in the relative proportions
of skeleton to soft tissues.

Experimental Methods
A. Animals and Diets

Male weanling Sprague-Dawley strain rats were used in all
feeding experiments. Two groups of animals were used to
measure food consumption on the high and low protein diets.
In these experiments spill-proof feeding tunnels were used,
and the animals caged singly. The remaining groups of
animals were housed in units of four in steel wire cages with
open-mesh bottoms. Continuous access to unlimited quanti-
ties of food in open containers was allowed. Distilled water
was similarly offered from dropping bottles. All groups of
animals were allowed to grow for a period of 20-25 days.
This period of time was chosen as it allowed approximate
doubling of weight for the most slowly growing groups.

The experimental diets were compounded using powdered
fat-free cow's milk (Starlac, The Borden Company), electro-
lyte and vitamin-free casein (Nutritional Biochemicals
Corporation, Cleveland), dextrose, a fat mixture composed of
equal parts of corn oil (Mazola Corn Oil, Corn Products Re-
fining Co., Argo, Illinois) and hydrogenated vegetable oil
(Crisco, Proctor and Gamble, Cincinnati, Ohio), and a salt

118 W. M. Wallace, W. B. Weil and A. Taylor

mixture (NaHCOg, 7-4 g.; KCl, 12-0 g.; CaCOg, 12-0 g.;
(NHJaHPO^, 14-9 g. ; MgSO^, 2-5 g.; KI, 0-001 g.) to pro-
duce the compositions shown in Table I. The salt mixture
was compounded to imitate the ion ratios found in fat-
free cow's milk. Ferrous sulphate, 2-0 g., copper sulphate,
0 • 22 g. and aureomycin, 0 • 25 g. per kg. of diet were incorpor-
ated in the mixtures. A vitamin mixture (Vitamin Diet
Fortification, Nutritional Biochemicals Corporation) in quanti-
ties calculated to make all diets equal in this respect was
added to the mixtures.

Table I

Analysis of diets

let Protein

g./lOO g. Diet

Fat Carbo- Ash

hydrate

Other"^

m-mole 1 100 g. Diet
K CI Ca

PHE 23-4

300 35-5 602

16-90

32-4

28-3

23 1

21-8

PLE 23-4

30-0 35-5 308

8-64

16-6

14-5

11-8

10-9

PHE 12 0

320 50-5 6-02

2-4

16-90

32-4

28-4

23-1

21-3

PLE 12 0

320 50-5 308

2-4

8-64

16-6

14-5

11-8

10-9

ciskies 26-8

6-5 51-4 11-60

3-7

15-10

15-3

13-9

840

56-8

* Moisture + Fibre (calculated by difference)

Prior to the beginning of the feeding experiments, all
animals had been weaned to a commercially produced small
animal feed (Friskies, The Carnation Milk Company) known
to produce excellent growth, general health and reproduction
in the albino rat. Preliminary feeding trials with the high
protein experimental diets in comparison with the Friskie
diet indicated equal effectiveness as measured by weight gain,
general appearance, activity, gentleness and lack of morbidity.

Eight groups of animals were studied, namely :

1. Weanling group (WEAN) 70-80 g. rats weaned to
Friskies.

2. High Protein-High Electrolyte (HPHE), see Table I

3. High Protein-Low Electrolyte (HPLE), see Table I.

4. Low Protein-High Electrolyte (LPHE), see Table I.

5. Low Protein-Low Electrolyte (LPLE), see Table I.

Effect of Variable Intake on Body Composition 119

6. Rats fed Friskies by way of control. See Table I for
composition of this ration.

7. A high protein, high electrolyte group fed to measure
food consumption.

8. A similar group to No. 7 but fed the low protein, low
electrolyte diet.

At the end of the allotted period of growth (20-25 days) the
animals were etherized and 2 ml. of blood removed for
analysis either by heart puncture or tail incision. Killing was
accomplished by further ether exposure. The dead weight
was obtained and the abdominal cavity, thorax and skull
opened with heavy shears.* The whole body was then dried
in an oven at 85°-95° C. until a constant weight was reached
(4-5 days). During the drying process, the carcass was
further broken up with heavy shears. The disintegrated
carcass was extracted repeatedly with a cold mixture of equal
parts ethyl and petroleum ether and re-dried to constant
weight. The dried extracted carcass was then homogenized in
a Waring Blendor with 5 volumes of anhydrous acetone and
the solvent evaporated off and the material re-dried. This
process produces a fine homogeneous powder suitable for
quantitative analysis. The powder was stored in a desiccator.

B. Chemical Methods

Water. Calculated from weight loss after desiccation.

Fat. During the course of the analytical work, the fat
extraction method used as applied to tissues by Hastings and
Eichelberger (1937) was examined for completeness of fat
extraction when applied to whole carcass. Powdered carcass
was exhaustively extracted in the Soxhlet apparatus serially
using ether, alcohol and chloroform. This process increased
the degree of fat extraction to the extent of 1-5-4 g. per
animal. Analysis of the material subjected to such extraction

* Intestinal contents were not removed. Analysis of the total gastro-
intestinal tract and contents of similarly fed animals for water and fat-free
solids indicated that their inclusion does not appreciably alter the interpreta-
tion of the data.

120 W. M. Wallace, W. B. Weil and A. Taylor

indicated that its nitrogen content multiplied by 6-25 plus
the weight of its ash very closely approximated 100 per cent
of the material. Consequently, fat has been calculated in all
of the data by the relation: Fat = dead weight— water
weight— (nitrogen X6-254- ash weight). All of the constitu-
ents shown in Table II have been calculated as g., m-mole or
m-equiv. per 100 g. of protein plus ash (i.e. fat-free dry solids).

Ash. A sample of carcass powder was weighed after in-
cineration at 600° in platinum.

Nitrogen. Determined by macro-Kjeldahl analysis using
selenium as a catalyst.

Chloride. A micro modification of the method of Lowry and
Hastings (1942) was used with cold nitric acid filtrates.
Samples of the homogenized powder were also analysed
polarographically for chloride, using sulphuric acid filtrates,
with excellent agreement between the two methods.

Sodium, Potassium and Calcium. These were determined on
the ash after separation of calcium using methods previously
described (Bergstrom and Wallace, 1954).

Magnesium. Determinations were done on the ash using the
method of Fister (1950).

Phosphorus. This was determined on the ash by the method
of Fiske and Subbarow (1925).

All electrolyte and nitrogen analyses were in duplicate.

Results

The analytical data obtained in the experiments are shown
in Table II. For comparative purposes the whole body
analyses on the albino rat of Light and co-workers (1934) and
of Cheek and West (1956) are included. Also shown are the
average data of Widdowson and Spray (195 IB) for six normal
human newborn babies and the data for single whole adult
human bodies of Widdowson, McCance and Spray (1951^),
Forbes, Cooper and Mitchell (1953) and Mitchell and co-
workers (1945). The data for water, protein and ash have been
calculated per kilogram of fat-free body weight. The water
and electrolytes are also shown using as a reference standard

Effect of Variable Intake on Body Composition 121

lis

I I

i-i ^

°p^

3- -*3 «<3c:-

S3 ss §2 is 2s

2i J?c5' OCq «5JJ COOJ .-iJl" 05C? rllJl' (ijg ^^ t-I =5" Jj ^
S""^ ^'~' 15'^ O^ '^^ O^ <M--' (Mw r-jw rH^ 1-H

o t^ o

icco Ti<eo Oco

^ ;*P O^ ^1^ -^Q O^ t^P ^Q c:iQ aoQ <^Q n^ in^ eoP ifl^ o^Q <©^
^5^ o.- co^ c,^ ^w ^^ ^_ ^,_ „^ _^ «- «- wow w

OGO rHOS ^O (MO <-i'p, OOO Tt<0 eOo Mt^ 3r-l t^r-l IflrH O'* «0 5(M OO

S^^ 2-H S-}^ S-H ""-H g^^ ^4i ^S ^-H -H ---H «4^ S-H -H ^44 ^

CXJrH OicO ?? '^"* ?'? <^"^ C^^ "^S *?°P OC^ C5<N O^OO

(MC^I

lOrH

eoic

O'*

OrH

eoi-H

eoo

(M05

(MO

00 rH

(MCO

^-H

^-1^

^4H

^-H

?I5

'^-t^

^-1^

^-H

0|>

ino

rHQO

ifti>

oeo

(NiO

TjHrH

•^rH

(>lrH

Tl<0

OOO

OI>

Sh-h

^-H

"-H

iC(N

""-H

^+1

^22

■ lO

tOTt*

O !>■

eo<^i

rHQO

WOO

05 00

(Nt;-

l>(35

050

tM>

Tl<-^

(NO

mrH

t^05

rHO

ooo

t^(N

Tt<r-t

s-+^

CO^

T*<rH

<^-H

CO^

SH^

^-H

^4^

00
OrH

00 c5

Neo

(Nop

9"f

OSIM

(N5<J

(>JO

C0 01

OrH

OiH

OOO

T^O

(NO

rH+l

oco

«^

<N-H

eo^

^-f)

^4^

CO^

OStH

ol^

OOO

(MO

o§

m(»

T-HCO

4<M

■^r-l

=0^

00 rH

15®

lOO

,fl

-g

•<

<
+

en
+

t-,

o
o

1
>

'3
<

^«

1«

|«

l«

S«5

>,«3

bCoJ

•Sai

dbcQ

rite

<?«J

i«^

0(M OOm COo OCC

fom oo 2co mo

com cDO 5;(M mo

^+1 -H S-H -t^

(35 rH (M TflC;^

t-TH com rH^ r-i'^

eo<M mo 5'(M o9.

;2;-H -H S-H -H

eoeo oo Jrc'^ ^^

3-H -H ""-H -H

I I I I

a a -g

|fi ao ao fifi

j»a3 Saj fesQ a^J

122 W. M. Wallace, W. B. Weil and A. Taylor

100 g. of protein plus ash. This is equivalent to the commonly
used reference standard of fat-free dry tissue (vide supra).

Fig. 1 graphically presents the currently obtained data in
terms of grams of ash, protein and water per kilogram of fat-
free body. The grams of fat per kilogram of fat-free body are
shown to the right of the columns. It is evident that the
compositions of the fat-free bodies are essentially similar. The
relative proportions of water, ash and protein have not been
greatly modified by variation of the diet producing the growth

GROUP PROT ASH
ASH PROT

WEAN 5.0 2.0

ASH PROT

WATER

FAT

HPHE 5.6 1.8

HPLE 6.3

1.6

LPHE 4.5 22

LPLE 4.9

2.0

FRISKIES 4.8 2.1

500
(g./kg. Fat Free Body Wecomt)

Fig. 1. Ash, protein and water content calculated per

kilogram of body weight for the six groups. Fat per

kilogram of fat-free body is shown at the right.

increment. Only if body fat were included would gross
variation occur. The young animals (WEAN) are relatively
low in ash and protein and high in water; with growth the
bodies acquired relatively more ash and protein than they
did water. The fat contents of the animals on the low
protein diets are significantly higher than they are on the high
protein.

In Fig. 2 the absolute values for total fat-free body weight,
water, protein and ash for the five groups are shown as con-
trasted against the Friskie group as an arbitrary reference

Effect of Variable Intake on Body Composition 123

standard of growth. The high protein groups are very closely-
equivalent in weight, protein and water content to the
standard. The two low protein groups reach two-thirds of the
high protein groups with regard to weight, water and protein.
The degree of mineral accretion in the high protein animals is
significantly different, the high electrolyte group accreting

; — Body Weight (fat free)

Wean

WATER

PROTEIN

ASH

HPHE

HPLE

LPHE

LPLE

Friskies

1-152. g

M7. g.

29. g.

6.2g.

0% 50%

(% Friskies Weight)

100%

Fig. 2. Absolute quantity of gain of water, protein and ash

calculated as per cent of the Friskie or control group. The

dashed, jVertical lines indicate the body weights as a percentage

of the Friskies.

much more than the low, but less than the Friskie group
which was on an equivalent protein but higher ash-containing
ration. In the low protein groups, whether on high or low
electrolyte intake, the gain of ash is not significantly different.
It is evident that protein intake is a limiting factor allowing
exploitation of a high ash intake only on a high protein diet.
The low protein, low electrolyte animals show a greater
relative and absolute accretion of protein than do the low

124 W. M. Wallace, W. B. Weil and A. Taylor

protein, high electrolyte group. This is significant at the 1 per
cent level.

The protein to ash ratios shown in Fig. 1 and evident in
Fig. 2 indicate the main significance for body composition
resulting from diets of variable protein and electrolyte content.
The high-protein-fed animals have more protein in relation to

PROT. ASH IjA K l^o Q<

300

200

g./litre HjO

m-equiv./ioog.Prot. + Ash.

Fig. 3. Diagrammatic representation on the left is of the ash
and protein content calculated on a kilogram of water basis.
On the right the individual elements composing the ash and
their relationship to the sum of protein plus ash (fat-free dry
weight) are shown.

ash than do the low-protein-fed animals. Since bone contri-
butes 90 per cent of the ash, the ratios represent the soft tissue
to bone proportions in a very general yet valid way. It seems
evident that only on a high protein intake can the growing
body lay down maximal bony tissue. In the Friskie group
where the ash of the intake is very high and composed
chiefly of calcium salts, an even greater accumulation of ash

Effect of Variable Intake on Body Composition 125

occurs at the relative expense of soft tissue. Where this
relationship stops is not answered by the present data.

While all animals are grossly similar in body composition,
as shown in Fig. 1, certain significant differences can be found
upon more detailed examination of the data. The concentra-
tion of ash and protein in the body water and the nature of
the composition of the ash are shown in Fig. 3. It is evident,
as has been noted, that only in the weanlings and in the low
protein, high electrolyte group does a significantly different
amount of protein per unit of water appear.

All of the experimental data for individual constituents of
the body have been calculated using four reference para-
meters : i.e. grams or m-mole per whole body, per kilogram of
fat-free whole body, per kilogram of water and per 100 g. of
protein plus ash (fat-free dry tissue). All of these calculated
individual values have been compared among the four groups.
The following statements can be made :

I. The Effects on the Protein Content of the Body.

A. By Protein Intake.

Only in those animals on the high electrolyte diets did
increased protein intake result in increased protein content of
the body on any of the enumerated bases.

B, By Electrolyte Intake.

In the animals on the high protein intakes, the electrolyte
effect was variable depending upon the reference base used
for calculation. In the low-protein-fed animals a high electro-
lyte intake reduced the protein content of the body calculated
on any basis.

II. The Effects on the Mineral Content of the Body.

A. By Protein Intake.

On any basis of calculation, other than absolute body size,
the bodies of the animals fed a low protein intake, whether
with high or low electrolyte, contained more ash, calcium.

126 W. M. Wallace, W. B. Weil and A. Taylor

magnesium, sodium, chloride and phosphorus than those fed
a high protein intake.

B. By Electrolyte Intake.

The high electrolyte diets led to increased calcium and
decreased chloride in all groups calculated on any basis.

In the high protein groups the high electrolyte intakes also
resulted in more ash and less potassium when calculated on
any basis.

In Table II the serum concentrations of sodium, potassium,
chloride and total protein are shown for the four experi-
mental groups. The only consistent significant difference is
for the concentration of total serum protein. Serum protein
concentrations are higher in the high-protein-fed groups.
The lower protein concentration may indicate protein de-
ficiency in the low protein group and other evidence for such
deficiency is given below. The validity of serum protein con-
centrations as a reliable index of protein malnutrition can be
questioned. In this connexion it is of interest that the serum
protein concentration of the breastfed infant is lower than
that of the infant fed cow's milk (Tudvad, Birch-Andersen
and Marmer, 1957).

Animals in experimental groups No. 7 and No. 8 were fed
in such a manner as to allow accurate measurement of food
intake. The high protein group consumed 8 • 2 g. of ration per
animal per day in contrast to 9-3 g. per day for the low
protein group. The mean weights for the two groups at the
end of 23 days were 174 and 155 g. respectively. Calculation
of the caloric values for the whole bodies of these animals
shows that the high protein group contained 292 calories per
average animal (1,710 calories per kg.) and the low protein
group 263 calories per average animal (2,085 calories per kg.).
Calculation of the calories utilized for physiological activity
indicates that the low protein group expended 175 calories
more per animal for the period of observation than did the
high protein group. Increased spontaneous activity was
clearly evident in the low protein groups during the period of

Effect of Variable Intake on Body Composition 127

observation. Increased spontaneous activity with nutritional
deficiency has been previously noted (Forbes et al., 1935;
Bevan et aL, 1950).

Ca-m-mole/
100 g. Protein

150

100

V HPHE
• HPLE
+ LPHE
0 LPLE

/•

CAo I.70P- 51.1

0 48 Ca + 43 6

O 50 100

P- m-mole /lOO g. Protein

Fig. 4. Relationships of calcium and phosphorus to

protein in the experimental groups. For description of

method of construction, see text.

The data in Fig. 4 represent the calcium/phosphorus re-
lationship in the four principal experimental groups. On
the assumption that the protein content is a basic unit of
structure, the values are compared in relation to protein.
One advantage of this formulation is that the intercept of the

128 W. M. Wallace, W. B. Weil and A. Taylor

regression line on the X axis defines the amount of phosphorus
present in 100 g. of calcium-free protein. This value should
reflect primarily the phosphorus content of muscle tissue.
From the statistical analysis of the calcium-phosphorus
relationship, a correlation coefficient of + 0-90 was derived.
Further, by the analysis of variance technique, it has been
determined that the regression curve is a straight line, des-
cribed by the equations calcium = 1'70 phosphorus— 51*1
and phosphorus = 0-48 calcium -f 43-6 when both are
expressed as m-mole/100 g. protein, and calcium = 2-19
phosphorus —2-04 and phosphorus = 0-37 calcium + 1*35
when both calcium and phosphorus are expressed as g./lOO g.
protein. The X intercept is between 30-3 and 43-6 m-mole
phosphorus/100 g. protein or between 0-93 and 1-35 g. phos-
phorus/100 g. protein. It is of interest that the calcium/
phosphorus ratios of the four groups of rats studied by Light
and co-workers (1934) and the infants analysed by Widdowson
and Spray (1951) also lie on this regression line when their
values are calculated in this manner. This indicates that the
changes in phosphorus content of the various groups are
related to the changes in calcium and to the total amount of
protein present. The phosphorus concentration is constant in
the "soft tissue" (calcium-free protein), and the phosphorus
has a constant ratio to the calcium in the " skeleton " (calcium-
containing tissue).

It is also apparent from the figure that the calcium to
protein ratio is highest in the low-protein, high-electrolyte-
fed animals and lowest in the high protein, low electrolyte
group.

Discussion

The present data, like the very similar data of Widdowson
and McCance (1957) and Stanier (1957), indicate no real
evidence for storage or depletion of protein with varying
intake. The basis for such a judgment is made by examination
of data calculated using either a kilogram of fat-free whole
body or 100 g. of fat-free dry solids as a standard of reference.

Effect of Variable Intake on Body Composition 129

The rationale for the use of the latter standard has been dis-
cussed in detail elsewhere (Cotlove et al., 1951). While such a
reference point is essential for evaluation of acute shifts of
water and electrolytes in tissues, it may not be equally
applicable where the growth of a complex of tissues is in-
volved. In this latter situation it is essential that the relative
gain or loss of a substance in question be examined in regard
to a number of reference standards, as has been done here (see
Results). When the change in any constituent is consistent in
direction, regardless of the reference basis, it is probably a
real one, as has been noted above. However, when the change
is in one direction on one basis and in the opposite on another,
the question of gain or loss is difficult to assess. An example of
this from the current data is found in the change in potassium
content with change in protein intake in the animals on the
low electrolyte diets. The high-protein-fed animals were
larger and contained more potassium on an absolute basis.
When calculated per kilogram of fat-free body the potassium
concentrations were equal, but on a litre of water basis the
potassium was greater in the low protein group. Again,
referring this ion to fat-free dry solids, the high-protein-fed
animals would seem to have the highest content. For the
purposes of nutritional evaluation, it is valid to calculate
constituents as per unit of whole body inclusive of fat. When
this is done, an even greater number of permutations and
combinations can be found with regard to relative contents
of all substances. Until more is known concerning the distri-
bution, function and relationships of protein and electrolytes
in tissues, it would seem advisable to emphasize only those
changes which are relatively consistent.

When the present data are considered on this basis, the
composition of the body with regard to water, protein and ash
is the same despite variation of the components of the intake.
The whole body may be smaller or larger as limited by the
availability of certain crucial nutriments but its relative
composition remains unchanged. Only the relative size of the
skeletal mass in relation to soft tissue seems to be significantly

AGEING — IV — 5

130 W. M. Wallace, W. B. Weil and A. Taylor

susceptible to some variation by variation of dietary intake.
Even in relation to skeletal tissue the possibility of variable
composition is limited by another parameter, i.e. protein.
Thus, the composition of the body achieves an independence
from the environment, an independence that would seem
essential in a living system where metabolic function is carried
on by protein with its critical requirement for constancy of
water and ionic concentration.

The concept that the whole body or the cells of the body
may be enriched or depleted of their various chemical con-
stituents by variation of the dietary intake is widely supported
in the nutritional literature. By examination of retentions
during balance observations on growing infants, it may be
concluded that the higher the intake of a substance, the
greater will be its final concentration in the body per unit of
weight (Rominger and Meyer, 1927; Swanson and lob, 1933;
Stearns, 1939).

Correlation of weight gains of premature infants with the
protein and ash content of the milk fed has shown high
positive correlation with the increasing ash content (Kagan
et al., 1955). Conversely, possible support for the concept of
variable body composition stems from nitrogen losses after
trauma. Both animals and men maintained on low protein
intakes lose less nitrogen after trauma than do those with
prior optimal intakes (Munro and Cuthbertson, 1943; Cuth-
bertson, 1948). Holmes, Jones and Stanier (1954) found evi-
dence indicating that men shifted from very low protein
intakes to optimal intakes retained nitrogen far in excess of
that calculated from weight gain and external losses. The
use of the terms "depletion" and "deficiency" bears tacit
evidence for the belief in the concept of cellular impoverish-
ment during nutritional deprivation. The majority of the
evidence for the concept of variable storage of protein and
minerals and loss during deprivation stems from the technic-
ally hazardous techniques involving measurement of external
balances. The possibility of low correlation between apparent
retentions or losses and changes in body weight has not been

Effect of Variable Intake on Body Composition 131

commonly realized. The shortcomings of the balance method
are functions of such items as the effects of variable caloric
intake, quality and quantity of protein intake and mineral
ratios on the fat content of the body, the distribution of body
water and the relative size of body components such as
skeleton and muscle. These problems have been most com-
pletely explored in relation to evaluation of the problem of
protein adequacy (Mitchell, 1944; Allison, 1954; Calloway and
Spector, 1953; Spector and Calloway, 1953). It is also little
appreciated that systematic errors occur in the calculation of
apparent retentions that are cumulative in a positive direc-
tion, the magnitude of the cumulative error being in direct
proportion to the magnitude of the intake. This makes
difficult the comparison of retentions at variable intakes. The
relevance of this criticism with regard to calcium retentions
has been discussed by Mitchell and Curzon (1939) and by
Mitchell and co-workers (1945).

Examination of the composition of growth increments by
direct body analysis has shown that, once chemical maturity
is reached, the composition of the fat-free body with regard
to protein and ash is nearly constant, regardless of any pro-
cedures taken to modify weight gain (Moulton, 1923; Moulton,
Trowbridge and Haigh, 1922; Pickens, Anderson and Smith,
1940). As determined by direct body analysis the body com-
position of rats growing on mineral-poor diets shows little
change except for a deficit of calcium (Light et at., 1934).

The concept of variable cellular composition of the body is
difficult to reconcile with the knowledge of the composition
of tissues. All of the individual tissues of the albino rat have
been analysed for their water, protein, fat and mineral con-
tent by many investigators. All of these data show a mono-
tonous constancy when calculated on a fat-free basis. This
occurs despite almost infinite variation in the rations fed to
the animals. Unless special experimental conditions are
imposed, individual tissues seem to hold fast to their chemical
composition. The principle variation occurs with age (Lowry
et al., 1942). At any given age composition is constant. Even

132 W. M. Wallace, W. B. Weil and A. Taylor

with age the maximum change of water content is no more
than 1 per cent and of potassium 5 per cent.

Examination of whole body data, with certain saHent ex-
ceptions, also shows rather remarkable constancy. Fat is
probably the only component of the total body that can vary
within rather wide limits and still allow reasonable well-being
to exist. Variation from 10 to 50 per cent can occur without
apparent evidence of malfunction. The water content of the
fat-free body is more closely guarded. Variation of much
more than i- 5 per cent from a rather rigid norm results in
rapid increments of physiological disability. Moreover,
allowable variation of body water is primarily extracellular;
cellular water content, within the limits of viability, must be
confined to much smaller variations. Since protein is the
critical parameter against which water content must be
judged, it follows that protein concentration must also be
highly critical and susceptible to only minute variation. The
consideration applying to water must also hold for the chief
extracellular electrolytes, sodium and chloride. Deficit of
potassium in the whole body to the extent of approximately
25 per cent does occur, and is replaced by variable gains of
total body sodium (Schwartz, Cohen and Wallace, 1953;
Cheek and West, 1956). The studies of Sherman and Booher
(1931) show that the calcium content of the whole body is
widely variable in response to variation in the dietary intake.
Definition of the optimal body content of this ion is elusive.

In the discussion so far the point of view has been taken that
in order to justify the terms stored protein or mineral, these
must exist as physically demonstrable entities comparable to
glycogen and fat in the body. It would appear that the
essential organic structure of the body cannot be affected in
quality by adjustment of the diet. The careful chemical
analyses by Luck (1936) of rat liver proteins from animals
maintained on varying levels of protein intake indicate that
all fractions of the liver proteins have participated equally
in any "storage" process. Madden and Whipple (1940) have
defined the reserve store of protein as "... all of the protein

Effect of Variable Intake on Body Composition 133

which may be given up by an organ or tissue under uniform
conditions without interfering with organ or body function-
ing." This definition indicates primary physiological signific-
ance, not anatomical. In this view the primary requirement
for furthering understanding would be methods for character-
izing and distinguishing physiological depletion. The response
to repletion has been used to assess the degree of depletion in
such a physiological sense. The work of Madden and Whipple
(1940) and Cannon (1954) illustrates the fruitfulness of the
method for studying the metabolism of protein under con-
ditions of deficit. Cooke and co-workers (1952) and Schwartz,
Cohen and Wallace (1955) have applied the technique to
experimental potassium deficiency and Hansen (1956) to the
potassium deficit in kwashiorkor. The ability to survive in
stressful situations provides a further avenue of approach.
Baur and Filer (1957), employing the weanling pig growing
on diets similar to those used in the present experiments,
have shown differing abilities of animals growing on different
diets to resist water and caloric deprivation. Their data
indicate that animals maintained on low protein intakes
survive caloric deprivation to a greater degree than do those
maintained on high protein intakes. Conversely, the high-
protein-fed animals withstand water deprivation to a greater
degree than do their low-protein-fed companions. Sherman
(1946) has correlated calcium intake with life span and
reproductive life. A newly opened approach to the problem
of characterizing and assessing deficits in a physiological
sense is that of distinguishing structural versus enzyme protein
in tissues. Potter and Klug (1947) have shown that liver
octonoate and succinate oxidases are depressed in animals fed
varying levels of protein. Miller (1948) Lightbody and Klein-
man (1939) and Williams and Elvehjem (1949) have extended
these observations to a number of other tissue enzymes.

Summary and Conclusions

The composition of growth of the albino rat on high protein-
high electrolyte, on high protein-low electrolyte, on low

134 W. M. Wallace, W. B. Weil and A. Taylor

protein-high electrolyte and on low protein-low electrolyte
diets has been examined. Analysis of the whole body for
protein, water, fat, ash, sodium, potassium, chloride, calcium,
phosphorus and magnesium was performed on animals allowed
to double their weaning weights on the enumerated diets.

The animals on the low protein intakes grew significantly
less and their bodies contained more fat. The composition of
the fat-free bodies on a unit basis were all essentially similar
despite the variation of the food intake. The principle dif-
ference resulting from variation in intake was in the quantity
of the skeletal constituents in the various groups. The
animals consuming the low protein rations contained more
calcium and phosphorus on a unit basis than did the high-
protein-fed animals.

On the high protein intakes accretion of skeletal minerals
was dependent upon the level of electrolyte intake, being
higher in the high-electrolyte-fed animals. In the low-
protein-fed animals accretion of skeletal minerals was less
affected by the level of electrolyte intake.

Only in the animals on the high electrolyte diets did in-
creased protein intake result in increased protein content of
the body.

The significance of the data for nutritional evaluation is
discussed.

REFERENCES

Allison, J. B. (1954). In Methods for Evaluation of Nutritional
Adequacy and Status, ed. Spector, H., Peterson, M. S., and Friede-
mann, T. S. Chicago: Quartermaster Depot, U.S. Army.

Baur, L. S., and Filer, L. J. (1957). Personal Communication.

Bergstrom, W. H., and Wallace, W. M. (1954). J. din. Invest., 33,
867.

Bevan, W., Jr., Lewis, G. T., Bloom, W. L., and Abess, A. T. (1950).
Amer. J. Physiol., 163, 104.

Calloway, D. H., and Spector, H. (1953). Fed. Proc, 12, 410.

Cannon, P. R. (1954). In Methods for Evaluation of Nutritional
Adequacy and Status, ed. Spector, H., Peterson, M.S., and Friede-
mann, T. S. Chicago: Quartermaster Depot, U.S. Army.

Cheek, D. B., and West, C. D. (1956). J. clin. Invest., 35, 763.

Cooke, R. E., Segar, W. E., Cheek, D. B., Coville, F. E., and Darrow,
D. C. (1952). J. clin. Invest., 31, 798.

Effect of Variable Intake on Body Composition 135

CoTLOVE, E., HoLLiDAY, M. A., ScHWARTz, R., and Wallace, W. M.

(1951). Amer. J. Physiol, 167, 665.
CuTHBERTSON, D. P. (1948). Amer. J. Med., 5, 879.
FiSKE, C. H., and Subbarow, Y. (1925). J. biol. Chem., 66, 315.
FiSTER, H. J. (1950). Standardized Procedures for Spectrophoto-
metry. New York: Standard Scientific Supply Corp.
Forbes, E. B., Swift, R. W., Black, A., and Kahlenberg, O. J.

(1935). J. Nutr., 10, 461.
Forbes, R. M., Cooper, A. R., and Mitchell, H. H. (1953). J. biol.

Chem., 203, 359.
Hansen, J. D. L. (1956). S. Afr. J. Lab. clin. Med., 2, 206.
Hastings, A. B., and Eichelberger, L. (1937). J. biol. Chem., 117,

73.
Holmes, E. G., Jones, E. R., and Stanier, M. W. (1954). Brit. J.

Nutr., 8, 173.
Kagan, B. M., Hess, J. H., Lundeen, E., Shaeffer, K., Parker, J. B.,

and Stigall, C. (1955). Pediatrics, Springfield, 15, 373.
Light, A. E., Smith, P. K., Smith, A. H., and Anderson, W. E. (1934).

J. biol. Chem., 107, 689.
Lightbody, H. D., and Kleinman, A. (1939). J. biol. Chem., 129, 71.
LowRY, O. H., and Hastings, A. B. (1942). J. biol. Chem., 143, 257.
LowRY, O. H., Hastings, A. B., Hull, T. Z., and Brown, A. N. (1942).

J. biol. Chem., 143, 271.
Luck, J. M. (1936). J. biol. Chem., 115, 491.

Madden, S. C, and Whipple, G. H. (1940). Physiol. Rev., 20, 194.
Miller, L. L. (1948). J. biol. Chem., 172, 113.
Mitchell, H. H. (1944). Industr. Engng. Chem. {Anal.)., 16, 696.
Mitchell, H. H., and Curzon, E. G. (1939). Actualites sci. industr.,

No. 771.
Mitchell, H. H., Hamilton, T. S., Steggerda, F. R., and Bean,

H. W. (1945). J. biol. Chem., 158, 625.
MouLTON, C. R. (1923). J. biol. Chem., 57, 79.
MouLTON, C. R., Trowbridge, P. F., and H.\igh, L. D. (1922). Res.

Bull. Mo. agric. Exp. Sta., 55, 21.
MuNRO, H. N., and Cuthbertson, D. P. (1943). Biochem. J., 37, 12.
Pickens, M., Anderson, W. E., and Smith, A. H. (1940). J. Nutr., 20,

351.
Potter, V. R., and Klug, H. L. (1947). Arch. Biochem., 12, 241.
RoMiNGER, E., and Meyer, H. (1927). Arch. Kinderheilk, 80, 195.
Schwartz, R., Cohen, J., and Wallace, W. M. (1953). Amer. J.

Physiol., 172, 1.
Schwartz, R., Cohen, J., and Wallace, W. M. (1955). Amer. J.

Physiol., 182, 39.
Sherman, H. C. (1946). Proc. nat. Acad. Sci., Wash., 52, 682.
Sherman, H. C, and Booher, L. E., (1931). J. biol. Chem., 93, 93.
Spector, H., and Calloway, D. H. (1953). Fed. Proc, 12, 430.
Stanier, M. W. (1957). Brit. J. Nutr., 11, 206.
Stearns, G. (1939). Physiol. Rev., 19, 415.
SwANSON, W. W., and Iob, L. V. (1933). Amer. J. Dis. Child., 45, 1036.

136 W. M. Wallace, W. B. Weil and A. Taylor

TuDVAD, F., Birch-Andersen, A., and Marmer, I. L. (1957). Acta

paediat., {Uppsala), 46, 329.
WiDDOWSON, E. M., and McCance, R. A. (1957). Brit. J. Nutr., 11, 198.
WiDDOwsoN, E. M., McCance, R. A., and Spray, C. M. (1951). Clin.

Sci., 10, 113.
WiDDOWSON, E. M., and Spray, C. M. (1951). Arch. Dis. Childh.,26, 205.
Williams, J. N., Jr., and Elvehjem, C. A. (1949). J. biol. Chem., 181,

559.

DISCUSSION

Widdowson: May I suggest, Prof. Wallace, that you started your
experiments far too late. If you had started at 21 "Adolph days"
instead of 21 "Wallace days", you might possibly have got different
results. We have the feeling that a great deal happens during these first
three weeks of suckling and the whole subsequent growth and develop-
ment of the rat depends upon the amount of milk it receives during that
time. Rats suckled in litters of three weigh two to three times as much
at weaning as others suckled in litters of 16-20. This difference in weight
persists even though all the animals receive unlimited food from weaning
onwards. The chemical maturation of the tissues of the body, particu-
larly the skeletal muscle, is more rapid in the fast-growing rats.

Wallace: How are they different? Are they dilute?

Widdowson: The proportion of extracellular fluid in the bodies and
muscles of all the rats decreases with development, and the proportion
of intracellular constituents, nitrogen and potassium, rises, but the
changes take place more quickly in the fast-growing animals, so that they
reach chemical maturity at an earlier age.

Kennedy : We can say, too, that the general developmental history is
altogether different. For example, puberty in the female rat, as meas-
ured by vaginal opening, is at 30-35 days in the big rat and it may be 60
days in the small rat. All subsequent growth is also quicker.

Wallace: What happens if the smaller young rats are specially fed?

Kennedy: This experiment was first done by Parkes (1926 and 1929.
Ann. appl. Biol., 13, 374, and 14, 171). He did fantastic things like
suckling mice with rat foster-mothers and getting them up within 21
days to something like 75 per cent of an adult mouse's weight. I went
over this again, breaking the changes down week by week (1957. J.
Endocrin., 16, 9). I found the acceleration in growth rate due to an un-
limited milk supply was achieved almost entirely in the first week of life.
The difference between birth weight and the weight at the end of one
week might be fourfold; after that there was roughly a 50-60 per cent
increase per week and this went on after weaning, when food was un-
limited. Something happened within the early part of the suckling period
which determined the shape of the subsequent exponential growth curve,
and I think that one of the things was probably the development of
appetite regulation. The amount the animal ate became fixed in relation
to body weight, so naturally the bigger rat ate more and continued to
grow faster.

Discussion 137

Wallace: Can you change them by feeding them different diets?
Kennedy: After weaning this has no effect. I have increased the con-
centration of protein in our stock diet, which is usually 13 per cent, to as
high as 30 per cent, which is about what rat milk contains, without signi-
ficantly changing the growth rates of the large or the small weanlings.
We have not tried to change the diet of sucklings.

Widdoivson : It would be most interesting to give some rats in a litter
electrolyte and protein supplements by stomach tube from the day of
birth omvards, and allow the mother to suckle the whole litter so that
some would get a higher protein and electrolyte intake than others.
Analysis of the bodies at three weeks of age might show much bigger
differences than those reported by Prof. Wallace for his older rats.

Talbot: When you give a high as contrasted to a low protein intake,
how much protein do you give the rats per day relative to their absolute
growth increment?

Wallace: I suppose that you are referring to the question of "feed
efficiency" — the relation of grams of food consumed to grams of weight
gained. This was 1-81 g. food per gram gain of weight for the high-
protein-fed animals and 2-51 g. consumed per g. of gain for the low
protein group. Thus the low protein group were less efficient in this
regard. If gain of weight per gram of protein consumed is calculated
the values are 0-41 g. per g. gain and 0 • 30 g. per g. gain for the high and
low protein groups respectively. The high protein animals, however,
have a greater gain of protein per unit of weight gain.

Kennedy : In the two curves you showed us with 100 per cent difference
in concentration of protein, there was nothing like 100 per cent differ-
ence in growth. Therefore it seems to me that the feed efficiency must
have been in favour of the low protein diet.

Wallace: One of our reasons for doing this type of experiment was to
find out whether or not we could rely on balance measurements to meas-
ure the composition of growth. I think that the answer is a negative one.
Except for change in body fat content, the composition of the body of
the growing individual remains relatively constant over the periods in
which it is feasible to carry out such measurements. There are probably
extreme experimental conditions which do change body composition but
I do not believe that one can change lean body composition significantly
by changing the plane of protein intake. One can probably determine
more accurately the composition of growth by dilution techniques than
by the balance method.

McCance : What would be the effect of change in diet on electrolytes in
the body? Our conclusion at the moment is that it has little effect on
the composition of the cell.

Wallace: We cannot change the electrolytes in the cell; we can only
change the amount in bone. Muscle can be made to grow faster or bigger,
but its composition in terms of electrolytes cannot be altered.

Heller : Our experience is that you have to decrease the protein content
of the diet very considerably to produce changes in body composition.
We have recently been feeding weanling rats on cassava flour and African
plantains, that is to say on diets that produce kwashiorkor in infants.

138 Discussion

After about four weeks there was an increase of 5-7 per cent in total body
water, but the interesting thing is that the plasma potassium and plasma
sodium concentrations remained unchanged.

Milne : Prof. Wallace, the main change in calcium with these diets was
in the skeletal calcium. Have you any information on changes in soft
tissue calcium, particularly kidney calcium? In my experience it varies
tremendously in rats on different calcium diets.

Wallace : The calcium in the body is almost entirely skeletal and with
this kind of data it is impossible to say just where this calcium is. You
have to study the individual tissues.

Fourman : Do you think that the increase in bone which you suggested
took place is an increase in trabecular bone — so-called freely available,
mobilizable, bone tissue?

Wallace : We are not certain but think it is probably both cortical and
trabecular. We would like to know if the large animals on the high
electrolj^e intakes have more easily mobilizable bone tissue under
conditions of stress.

McCance : You began by putting up charts of balances showing that if
the diet contained more sodium and potassium, the child absorbed and
retained more. Yet you find by experiment that you do not alter the
composition of the body. Can you reconcile those observations?

Wallace: This is a purely technical matter on which I have strong
feelings. In a balance experiment the quantity of food entering the
body and the excreta recovered are always slightly less than the measure-
ments indicate. The more refined the technique the smaller this error is.
Also, the greater the concentration of a nutriment in the intake the
greater will be the error when compared with intakes of lower concen-
tration but of equivalent caloric value. When subtraction is used to
calculate the balance these errors accumulate. The errors in doing a
balance are not randomly plus or minus as is generally believed, but
systematically positive. Much of the arithmetical difficulty arises because
one must subtract two quite large numbers to obtain the usually very
small balance value. At zero intake the balance method becomes more
accurate. Body composition estimates such as can be made from Bene-
dict's and Gamble's fasting data agree with direct analysis data quite
well. However, body composition estimates made from balance data
with infants fed with cow's milk and human milk are always widely diver-
gent, even when weight gains are equivalent. The higher the intake of a
constituent the greater the apparent retention. Eventually the retention
becomes patently absurd.

THE EFFECT OF AGE ON THE BODY'S

TOLERANCE FOR FASTING, THIRSTING AND

FOR OVERLOADING WITH WATER AND

CERTAIN ELECTROLYTES *

Nathan B. Talbot and Robert Richie

Department of Pediatrics, Harvard Medical School and the Children's Medical
Service, Massachusetts General Hospital, Boston

As is well known, the body is equipped with homeostatic
systems designed to maintain water and electrolyte content
and concentration values at physiologically optimal levels.
The systems accomplish this task largely by equating output
with input. While rates of input can be varied widely without
overreaching the capacities of the homeostatic systems con-
cerned, nonetheless there are limits beyond which one cannot
go without getting into difficulty (Talbot, Crawford and
Butler, 1953; Talbot et al., 1955). Thus for each substance
there is a physiological minimum requirement or floor, which is
the least intake of the substance in question needed to balance
output and hence to prevent deficits where conservation forces
are acting maximally. There is also for each substance a
physiological maximum tolerance or ceiling which is defined as
the largest amount of the substance that can be taken and
eliminated without seriously disturbing body composition.
Rates falling between these two parameters may be said to
fall within the physiological or safe working range. When the
rate of administration of a substance falls outside this range
for an appreciable length of time, body composition deviates
from normal and manifestations of disordered homeostasis
develop as outlined in Table I.

* This paper is based on work supported by grant A-808 of the National
Institute of Arthritis and Metabolic Disease, by grants H-1529 and HTS
5139 of the National Heart Institute, United States Public Health Service,
and by a grant from the Commonwealth Fund of New York.

139

140

Nathan B. Talbot and Robert Richie

The manner in which a hmit to homeostatic capacity can
be recognized and defined is illustrated in Fig. 1 (Talbot et al.,
1956). Here it can be seen that this patient maintained a
normal potassium status, as judged from electrocardiographic
T waves and from serum potassium concentration, and
remained in potassium balance at rates of intake up to
approximately 70 m-equiv. per m.^ per day. These rates of

Table I

Indications that intake is physiologically excessive or insufficient

(adult values)

Sub-
stance

Too Much

Too Little

H2O

Water intoxication

Serum water >3 • 8 ml./m-osm.*

Hypohydration

Serum water <3- 4 ml./m-osm.*

Extracellular oedema

NaEt> 20%

Extracellular dehydration

NaEi> 120/0

Weakness; ECG

T waves | ;

Serum K >6-5 m-equiv. /I.

Weakness; ECG
T waves | ;
Ki i >20%

Serum P >6 mg.%

Osteomalacia

Nas = extracellular sodium.

Ki = intracellular potassium.

Corrected for urea.

intake could therefore be considered to be within his safe
working range. By contrast, higher rates of intake led to a
sustained positive balance and to the appearance of elevated
T waves and hyperkalaemia, which are taken to be signs of
potassium intoxication. Accordingly, it may be said that this
individual's ceiling of tolerance for potassium was about
70 m-equiv. per m.^ per 24 hours, a subnormally low value in
comparison with a normal ceiling of at least 250 m-equiv. per
m.2 and in keeping with the fact that he was suffering from
marked impairment of renal function.

Effect of Age on the Body's Tolerance

141

The same principles have been used in estimating the upper
and lower limits of body tolerance for water and certain
electrolytes for normal individuals of various ages, depicted
in Fig. 2. The upper limits shown in this figure are of necessity
approximate, being based on the relatively few data available
in the literature and the files of our metabolic unit (Talbot
et ah, 1952; Talbot, Crawford and Butler, 1953; Talbot et ah.

HEIGHT OF
T WAVES 2

IN LEAD IE
mm.

INPUT

a 60

OUTPUT
mEq / m^/ day

Fig. 1. Demonstration of physiological maximum tolerance for

potassium in a patient with impaired kidneys. (From Talbot

et al., 1956).

1955, 1956; Talbot, Richie and Crawford, 1958). In all
instances, they are intended to represent levels which can be
attained by healthy individuals within a day or so and not the
uttermost levels which can be attained after extensive prior
conditioning. The lower limits include normal growth re-
quirements for infants and children, a factor of relatively
small size after the first few months of life. It can be seen that
with the exception of young infants, individuals normally

142 Nathan B. Talbot and Robert Richie

utilize but a small segment of their homeostatic capacities.

In early infancy, the margins of safety are relatively quite

narrow, a fact long recognized by those interested in paediatrics.

The clinical significance of these homeostatic parameters

300
SODIUM
mEq/m*/24* zoo

POTASSIUM
inEq/m^/24» ioo

llllllllli /f/t/vee OF PHYSIOLOGIC TOLERANCE
RANGE OF NORMAL DIETARY INTAKE

•

ill '^''

liiMflPIIHif

H::,ll.:|i"!i:i:,,li:'l..l.;!;l'l!'::.ll!

ISO

PHOSPHORUS
mMol/mV24* loo

-jlgflfl^,:

^^*^:-l^:''-''''^

/ 6 3 6 9 12 18 Z
BIRTH WKS. ^-MONTHS— '

6 e 10 12 16 20
YEARS '

Fig. 2. Estimates of the safe working ranges of intake

for individuals of various ages and of the portions of

these ranges used by persons taking ordinary diets

for age.

may be visuaUzed by considering the length of time needed
for individuals of various ages to lose a significant portion of
their body stores when totally deprived of water or certain
other substances (Fig. 3). In calculating these time values,
attention has been given to the changes in body composition

Effect of Age on the Body's Tolerance

143

which occur during the growth period; in each case average
normal values for body composition and content were used
(Shohl, 1939; Forbes and Perley, 1951; Corsa et ah, 1956;
Friis-Hansen, 1957). Each substance has been considered
separately. In dealing with water, sodium and potassium,

DAYS OF
DEPRIVATION
TO PRODUCE

SERIOUS
DEPLETION

15% DECREASt IN BODY PROTEIN
LOSS = 259m/m2/24"' ? 9

15% DECREASE IN BODY SOOiUM
LOSS = 10 TiEq/m^/gA* X X

75% DECREASE IN BODY FAT (CALORIC) STORES
LOSSnaoOCal /m2 /2A' f V

15% DECREASE IN BODY POTASSIUM
LOSS = lOmEq/nn2 /24' O— O

15% DECREASE IN BODY WATER
STORES •— •

LOSSES' IWL + OBLIGATORY URINE

/ 6 3 6 9 12 18 2
BIRTH WKS --MONTHS^

3 4 6 6 10 12 16 20
YEARS ■

Fig. 3. Days of deprivation (ordinate) needed to produce the
percentage decrease in body content indicated for each
substance in individuals of various ages (abscissa). The
rates of loss indicated for each substance approximate to
physiological minimum output rates, of which some are indi-
cated by the lower boundaries of the physiological tolerance
ranges shown in Fig. 2.

rate of loss was taken as the physiological minimum require-
ment value indicated in Fig. 2. In considering body fat
(calorie) stores, energy expenditures were assumed to be at
the rate of 1,800 calories per m.^ per day (Macy, 1942) and to
be derived entirely from body fat. Body protein losses were
calculated assuming a basal rate of loss amounting to 25 g.
per m.2 per day, the minimum value attained by individuals

144 Nathan B. Talbot and Robert Richie

receiving at least 75 grams of carbohydrate per m.^ per day
(Gamble, 1946-7). It was arbitrarily decided that a 15 per
cent decrease in body water, sodium, potassium or protein or
a 75 per cent depletion of body fat (calorie) stores constituted
a significant and potentially serious loss.

As indicated by the upward trend from left to right of the
curves of Fig. 3, infants and children up to three years of age,
when deprived of any one of the substances represented, are
apt to become depleted two to four times faster than adults.
For example, infants will develop as serious a degree of water
depletion within one and a half days as adults do in the course
of about five days of total thirsting. Likewise, infants de-
prived of electrolytes or protein or calories may lose an
appreciable portion of their body stores of these items after
nine to 17 days of deprivation.* By contrast, it takes 20 to
35 days for adults to become similarly depleted under condi-
tions where homeostatic conservation forces are operating
efficiently. These observations indicate that in infants who
must be maintained by parenteral fluid therapy for more than
a few days, special attention should be given to the provision
not only of water, carbohydrate and the main extracellular
and intracellular electrolytes, but also of maintenance allot-
ments of calories and either preformed protein or amino acids.
The same would apply to older children and adults who are
depleted or have to be sustained by parenteral fluid therapy
for more than a week or ten days.

Fig. 4 deals with the opposite phenomenon of overloading.
Here again it has been necessary to make arbitrary decisions
concerning the size of the overload and the degree of retention
to be considered significant. It was decided to postulate rates
of input that were ten per cent in excess of adult physiological
maximum tolerance or ceiling values. The end-point values
for the retentions of toxic degree resulting from these physio-
logically excessive rates are related to the respective average
normal body content values at each age as follows : total body

* The rate of loss would be considerably greater under conditions of zero
carbohydrate intake (Gamble, 1946-7).

Effect of Age on the Body's Tolerance 145

water, +7 per cent (Wynn, 1956); potassium, +5 per cent
(Drescher et al., 1958); total body sodium (euproteinaemic
subjects), +30 per cent (Leaf, personal communication). In
the case of phosphorus the end-point chosen was elevation
of extracellular inorganic phosphorus concentration to 12 mg.

160

140

HOURS OF
OVERLOAD
TO PRODUCE
TOXIC EFFECTS '20-

30% INCREASE IN BODY SODIUM
GAIN =25 mEq/m2/24« X X

5% INCREASE IN BODY POTASSIUM
GAIN«25 mEq/m2/24'' o— o

SERUM PHOSPHORUS CONC. ELEVATION
TO 4mmol/L (I2.4mg.%)
GAIN « ISmMol /m2/24» ? f

7% INCREASE IN BODY WATER
GAIN •l.5L/m2/24» • •

/ 6 3 6 9 12 18 2
BIRTH WKS. --MONTHS—

3 4 6 e 10 12 16 20
YEARS '

Fig. 4. Hours of overload (ordinate) needed to produce the
percentage increase in body content indicated for each sub-
stance in individuals of various ages (abscissa). The rate of
gain is that which obtains when rate of input exceeds the
physiological maximum tolerance levels for adults shown in
Fig. 2 by approximately ten per cent.

per cent.* Individuals who have surpluses of these degrees
are apt to show the signs of intoxication listed in Table I.

As might be expected, Fig. 4 indicates that infants are
relatively much more vulnerable to overloading than older
children and adults. This is true not only in the relative
terms depicted here, but also in absolute terms because the
quantity needed to produce intoxication in a small individual

* This assumes no bodily capacity for cellular or skeletal storage of surplus
inorganic phosphorus, a point on which we have no objective information.

146

Nathan B. Talbot and Robert Richie

is not very great. The curves indicate that one is apt to
become water and phosphorus intoxicated before one becomes
potassium or sodium intoxicated. It is interesting that these
relations are in keeping with chnical observations on patients
with marked hmitation of renal function (Talbot et al., 1956).
One of the areas where the foregoing considerations appear
to have practical implications is with respect to parenteral
fluid maintenance therapy. Review of hospital practices

INTAKE AND OUTPUT

SUBJECT H.W. n SUBJECT Y-S.C. SUBJECT PT

24 48

TIME IN HOURS
Fig. 5. Intake and output of water and electrolytes by normal adult subjects
receiving a standard maintenance allotment of multiple electrolyte plus
dextrose solution in 24, 12 or 6 hours each day. (From Neyzi, Bailey and

Talbot, 1958).

reveals that some physicians give the total daily fluid,
carbohydrate and electrolyte allotment in a slow continuous
manner while others administer the total daily dose in a few
hours, allowing the patient to fast and thirst for the remainder
of the 24-hour period. The data shown in the right-hand sec-
tions of Fig. 5 (Neyzi, Bailey and Talbot, 1958) indicate the
ranges of output rate observed on two sets of three normal
adults maintained for three days on an ordinary dose (1,200
ml. per m.^ per day) of a solution containing, per litre, 50 g. of

Effect of Age on the Body's Tolerance 147

dextrose, 40 m-equiv. of sodium, 35 m-equiv. of potassium,
40 m-equiv. of chloride, 20 m-equiv. of lactate and 15 m-equiv.
of phosphate (Talbot, Crawford and Butler, 1953; Talbot et
al., 1955). The first set of subjects received their allotment by
mouth in an essentially continuous (hourly dose) manner, the

H20

L/m2/24»

400

mEq/m2/24»

200

240 120 go

r— REGIMEN

.. Il il

{

\l {i

400

mEq /m2/24» 200

0
400
CI
fnEq/m2/24» 200

SI i

iixii

10
YEARS

Fig. 6. Relations between rates of output observed for
subjects on various regimens shown in Fig. 5 (right-
hand section) and physiological ranges of excretory capa-
city shown in Fig. 2 (left-hand section). The solid
black circles indicate the average and the vertical bars
traversing them the ranges in output rate noted for the
individual subjects per the scales along the left-hand
ordinate. (From Neyzi, Bailey and Talbot, 1958).

second set at twice the rate for 12 hours each day and the
third set at quadruple the rate for six hours out of every 24.
As indicated by the length of the vertical lines at the right of
Fig. 6, those on the 24-hour regimen utilized but a small
fraction of their physiological ranges of excretory capacity in
accomplishing metabolic homeostasis. By contrast, those on

148

Nathan B. Talbot and Robert Richie

the 12-hour and especially those on the six-hour regimens
used almost fully their normal adult ranges of renal excretory
adjustment in the course of each 24-hour period. When the
homeostatic adjustments in water and electrolyte excretion
exhibited by these adult subjects are viewed with relation to
the infant ranges of homeostatic adjustment indicated by the
shaded zones of the left-hand sections of Fig. 5, it can be seen

% INCREASE

IN
BODY CONTENT

10-

- HgO

r^-T

^ (5 wk.lnfont

IjAdult
^yMPotential

M, /Adult
Hr~lobserved

Fig. 7. Percentage increases in body water, sodium and potassium content
(a) which actually occurred (black sections) during the 6-hour infusion period
in the 6-hour regimen subjects of Figs. 5 and 6; (b) which would have occurred
in these adults (adult potential levels), and (c) which would have occurred in
a small infant (5-week infant potential), had no homeostatic increase in output
rates above basal levels occurred.

that they are considerably greater than those of which such
young individuals are capable.

Fig. 7 depicts the percentage increases in body water,
sodium and potassium content which would occur during the
course of the infusion period if a day's total maintenance
allotment of 1,500 ml. per m.^ per 24 hours * were adminis-
tered in six hours to a patient who was unable to increase
rates of urinary output above the physiologically low levels
characteristic of fasting and thirsting, a situation which one

* This is an ordinary allotment for infants and children on our Service.

Effect of Age on the Body's Tolerance 149

may encounter in young infants and in patients undergoing
the stress of anaesthesia and surgery. As the columns show,
the percentage gains to be expected for infants are approxi-
mately twice as great as those to be expected for adults.
While the gains indicated for adults are borderline as regards
toxicity, those shown for infants are large enough to produce
distressing manifestations.

In summary, an attempt has been made to indicate in
approximate terms the limits of capacity of the body to adjust
output of water and certain other substances in accordance
with homeostatic needs, and to illustrate the clinical implica-
tions of such knowledge.

These thoughts are presented in the hope that they may
elicit constructive suggestions concerning these highly signi-
ficant, yet rather elusive phenomena.

REFERENCES

CoRSA, L. Jr., Gribetz, D., Cook, C. D., and Talbot, N. B. (1956).
Pediatrics, Springfield, 17, 184.

Drescher, a. N., Talbot, N. B., Meara, P., Terry, M., and Craw-
ford, J. D. (1958). Submitted for Publication.

Forbes, G. B., and Perley, A. (1951). J. clin. Invest., 30, 566.

Friis-Hansen, B. (1957). Acta paediat., (Uppsala), 46, Suppl. 110.

Gamble, J. L. (1946-7). Harvey Led., 42, 247.

Macy, I. G. (1942). Nutrition and Chemical Growth in Childhood.
Vol. I. Evaluation. Springfield: Thomas.

Neyzi, O., Bailey, M., and Talbot, N. B. (1958). New Engl. J. Med.,
in press.

Shohl, a. T. (1939). Mineral Metabolism. American Chemical Society
Monograph Series. New York: Reinhold Publishing Co.

Talbot, N. B., Crawford, J. D., and Butler, A. M. (1953). New
Engl. J. Med., 248, 1100.

Talbot, N. B., Crawford, J. D., Kerrigan, G. A., Hillman, D.,
Bertucio, M., and Terry, M. (1956). New Engl. J. Med., 255, 655.

Talbot, N. B., Kerrigan, G. A., Crawford, J. D., Cochran, W., and
Terry, M. (1955). New Engl. J. Med., 252, 856, 898.

Talbot, N. B., Richie, R., and Crawford, J. D. (1958). Metabolic
Homeostasis: Basic Considerations and Clinical Applications. A
Syllabus. In preparation.

Talbot, N. B., Sobel, E. H., McArthur, J. W., and Crawford, J. D.
(1952). Functional Endocrinology from Birth Through Adoles-
cence. Harvard University Press.

Wynn, V. (1956). Metabolism, 5, 490.

150 Discussion

DISCUSSION

Black : There seems to be some conflict between Dr. Talbot, who says
that large intakes should produce retention, and Prof. Wallace, who tells
us that large intakes produce large arithmetical errors. In this matter I
am on Dr. Talbot's side, and that is not entirely the emotional reaction
of someone who has done a certain amount of balance experiments. I
think we have some supporting evidence in that if balance experiments
are done on an adult person who has just had an operation and is on a
milk intake (in which the errors of measurement should be much the
same as those of excreta), there is quite a definite correlation between
intake and retention (Davies, H. E. F., Jepson, R. P., and Black,
D. A. K. (1956). Clin. Sei., 15, 61).

Bull : What was the nature of the load imposed in the experiments on
the tolerance of loading?

Talbot : The rate of intake of the substances in question is increased in
a stepwise manner which allows time for compensatory homeostatic ad-
justment in rate of output to take place. At each step, measurements are
made to find out whether the body content and /or concentration of the
substance is being kept within physiological limits by appropriate adjust-
ments of the rate of output. As rate of input is increased, it eventually
reaches a point where the body is unable to keep its content and concen-
tration values within normal limits by suitable adjustment of rate of out-
put. This point is considered to be the upper limit of physiological
tolerance or physiological ceiling for the substance in question. Rates of
input in excess of this ceiling level produce a tendency to abnormal
retention. For example, in the case of potassium, when the rate of input
exceeds the physiological ceiling value, body potassium content increases
above normals levels and hyperkalaemia develops, together with signs of
potassium intoxication.

McCance : I would like a firm definition of what you mean by tolerance
and capacity to eliminate. De Wardener did some experiments in which
he took large amounts of water every day for 7 or 14 days and although
he did not succumb and appeared to tolerate them perfectly well, there
were finite changes in his responses, sensitivities, etc. (de Wardener,
H. E., and Herscheimer, A. (1957). J. Physiol, 139, 42 and 53).

Talbot : In the case of water, the body normally can tolerate up to
approximately 15 litres per square metre or about 25 litres per adult per
day. These large quantities are eliminated simply by increasing the ratio
of water to solutes in urine to levels of 20 to 30 ml. per m-osm. It is diffi-
cult to exceed this ceiling value in the normal individual. On the other
hand, it is easy to exceed the water tolerance ceiling value in pan-ne-
phritics and postoperative patients who are unable to increase the water/
solute ratio of their urine above a few ml. per m-osm. and whose rate of
solute output may be low. Such individuals may be unable to take more
than 2 or 3 litres of water per square metre per 24 hours without retaining
water and developing water intoxication.

Kennedy : Some of these substances were orally administered, and some

Discussion 151

parenterally. It is said that one of the safeguards in oral ingestion of
water is the fact that eHmination goes on about as fast as absorption.

Talbot : As far as water, sodium and potassium are concerned, it is six
one way and half-a-dozen the other whether they are taken by vein or by
mouth. With phosphorus, where calcium and other substances may carry
it out in the gut, there may be some large differences.

Bull: I believe there is a speed of infusion beyond which this theory is
not correct. If, for instance, very frequent samples of blood are taken
during transfusion, when a solution which is not isotonic is being given,
very high values may be found. I agree that if the balance studies are
taken for 24 hours, the result will be the same. But you can reach values
acutely which are well outside what you consider to be the normal range,
though fortunately without apparent ill effects. The picture of homeo-
stasis varies very markedly with the period over which you are consider-
ing it. My colleague Dr. Graber finds that if you go back to the finer detail
you may pick up oscillations in values which reveal the mechanism more
clearly than do the long-term studies.

Talbot : Rates of input which are expressed per square metre per day
mean are intended to represent the average rate of input throughout the
24-hour period. In other words, the fact that one may take as much as 15
litres of water per m.^ per day does not mean that one could tolerate this
volume if it were given in a fraction of the day. Indeed, were one to give
the 15 litres in 12 rather than 24 hours, one would be giving it at the rate
of 2 X 15 or 30 litres per m.^ per 24 hours. Such a very high rate of
input would produce signs of intoxication only if it were sustained for a
sufficient length of time. Thus, 30 litres per m.^ per 24 hours would be 30
divided by 24, or 1 -3 litres per m.^ per hour. One would have to infuse
water at this rate for at least 70 minutes to produce the 5 per cent gain in
body water necessary to induce overt signs of water intoxication.

Another factor which enters into such consideration is adaptation time.
Some of the body's homeostatic mechanisms, such as those concerned with
water, potassium and sugar, can adapt quite fully within two or three
hours, while others, such as those responsible for phosphorus and sodium
homeostasis, may require two or more days. In considering the ceiling
and floor levels reported here, an effort was made to take this variable
into account and to set forth ceiling and floor levels which the normal
individual should be able to attain without becoming seriously disturbed
metabolically either during the period of adaptation or later.

While it may be possible to set up experimental circumstances in which
there are differences in the body's tolerance for water and the various
electrolytes when given intravenously as compared to orally, for all ordi-
nary practical purposes the body's tolerance for these substances is
about the same whether they be given by mouth or by vein.

Fourman : Is it not true to say that with an excessive intake of water
the individual will vomit, and with excessive intake of potassium the
individual will, extraordinarily promptly, get diarrhoea?

Talbot : It is true that loss of thirst and nausea constitute accessory
mechanisms which serve to protect the organism against the development
of water intoxication by the oral route. On the other hand, we have

152 Discussion

observed that rats offered gradually increasing quantities of potassium in
their diet ate and absorbed the relatively very large quantities needed to
produce a lethal degree of potassium intoxication. They did not develop
diarrhoea, nor did they vomit; they just became weak and died. Like-
wise, we have seen a patient with marked limitation in tolerance for
potassium due to advanced pan-nephritis become fatally intoxicated
with potassium as a result of drinking fruit juices.

Adolph: The study of tolerances is a very important aspect of the
general physiology of regulatory processes. Dr. Talbot, you estimated
tolerances in terms of single constituents, but in some of the situations
you described, such as the intravenous administrations, you were con-
cerned with several constituents at a time. Now when there is depletion
or excess of more than one constituent at a time the picture is very differ-
ent with respect to tolerance. For instance, there is a great difference
between taking pure salt and taking an isotonic solution of salt. I recog-
nize that this work is exploratory and that you are making your esti-
mates in the simplest way possible when you consider one component at
a time, but eventually I hope we shall have some estimates of tolerance
to multiple components.

This consideration of components seems to me to extend also to your
studies of composition, Prof. Wallace. If you went to your statisticians
still more often, would you not get into the study of multiple correlations
which would get us further than comparisons made two at a time?

Wallace: We have made a number of statistical multiple correlations.
It is often difficult to know just what they mean, once certain correlations
become evident. Our biggest problem has been to have any assurance as
to the proper parameter to which to refer growth. Should the reference
basis be body weight, fat-free weight, protein, ash, or water?

Adolph : What I want to bring out is that an organism probably has
some way of measuring the bodily composition which is very much more
complicated than saying, for instance, that magnesium is the fixed con-
stituent around which all others revolve. I think that without a study of
multiple correlations we will never be able to find whether there is a key
fixity by which homeostasis is guided to a definite volume and concentra-
tion to which the organism always returns. I do not know whether
any of our methods of representing homeostasis will be so similar to that
of the organism that we can predict what it does to get back to its fixity.

I should also like to remark on Dr. Talbot's choice of a key variable.
No doubt he has great reservations about the use of this term. What he is
trying to do, I gather, is to out-guess the organism as to what it is using
as a measuring stick by which it will return to its original composition, or
by which it will estimate what has to be done in order to defend itself
against disturbances. When we think that an organism is restoring its
potassium concentration, have we any assurance that that one restora-
tion is a prime objective in the adjustments which are going on?

Talbot : We agree with you that most if not all of the variables under
consideration are related to each other. For instance it is known that
body tolerance for potassium is impaired under conditions of zero sodium
intake and that tolerance for sodium is abnormally limited under condi-

Discussion 153

tions of zero potassium intake. On the other hand, it was thought that a
thorough exposition of available information on these relations at this
time would serve only to confuse the picture without adding greatly to
its significance. Certainly one cannot take and eliminate large loads of
electrolyte without an ample supply of water etc. Accordingly, it was
decided to define physiological maximum tolerance and minimum require-
ment levels for each substance under circumstances where the influence
of these types of factors should be minimal, i.e. under conditions where
the rates of intake of substances other than the one under consideration
were well within normal limits. Should these preliminary definitions
prove to be of value, it may become worthwhile to undertake to extend
and refine them more by detailed definitions of certain of the most
important interrelations.

You are correct in your deductions concerning our aims in defining
physiological key variables. The present definitions are of necessity ap-
proximate and potentially subject to modification and refinement. At
the same time they are pro\ang to be of value as indices of patient status
and as a point of departure for investigation.

CLINICAL CONSEQUENCES OF THE WATER
AND ELECTROLYTE METABOLISM
PECULIAR TO INFANCY

E. Kerpel-Fronius*

Department of Paediatrics, University of Pecs, Hungary

Disturbances in the volume and composition of the body
fluids occur more frequently in infancy than at other ages.
Among the reasons for this are :

(1) The high incidence of diarrhoea, malnutrition, and
certain congenital defects.

Diarrhoea is still one of the paediatrician's major concerns,
one of its main causes being colon bacilli, pathogenic only for
this age group.

Owing to their high caloric and protein requirements
infants easily succumb to malnutrition, which progresses
rapidly. The resulting expansion of the volume of their extra-
cellular body fluids, sometimes accompanied by asympto-
matic hyponatraemia, is a common disturbance of homeostasis
in some countries.

Congenital defects of the oesophagus, the pylorus, the
renal tubules, the adrenals, and the central nervous system
may also cause serious disturbances in the body fluids ; their
discussion is beyond the scope of this paper.

(2) Circulation, metabolism and renal excretion are all
maintained at high levels relative to the volume of the body
fluids.

(3) When growth is arrested by disturbances which
diminish the utilization of food, a fraction of the intake
normally retained is rejected, thus raising the solute load on
the kidneys.

* In the absence of Prof. Kerpel-Fronius, his paper was read for him by
Dr. Winifred Young.

154

Effects of Metabolic Disturbances in Infants 155

(4) Partly due to the interrelationships (2) and (3) kidney
function is readily impaired by stress.

Thus the high incidence of body fluid disturbances is
partly due to the occurrence of disease and partly to relatively
inefficient homeostatic defence mechanisms. The latter is well

m-osm/
500

450

400

350

300

250

200

150

100

Fig. 1. Lability of osmotic regulation in

10-day-old puppies.
Left column : salt- and protein-

free diet
Central column : normal
Right column : concentrated milk

illustrated by the observation that diets such as milk evapor-
ated to one-quarter of its original volume, or salt- and protein-
free food, bring about great changes in the tonicity of the
body fluids (Csapo and Kerpel-Fronius, 1933; Kerpel-
Fronius, 1933). After the first, the osmolarity of the blood
plasma in puppies rose to 526 m-osm. /I., 457 m-osm. being
accounted for by " hyperelectrolytaemia " ; after the second,
the electrolytes decreased to 232 m-osm./l. (Fig. 1). There

156

E. Kerpel-Fronius

was a water loss of over 20 per cent of body weight in the
first case, while in the second an increase in the water content
of all organs was observed. Such gross disturbances of
homeostasis may partly be due to the fact that although the
extracellular body fluids occupy a relatively high percentage

WElfiHT E.C.FLUID E.C.
%OFWEiaMT ABS.

PL/Kg

PL.ABS. HAEMA-
TOCRIT

CIRCUL. PAH-
TIME ClEARAMtE

Fig. 2. Extracellular fluids, circulation and PAH clearance

in the dehydration of a malnourished infant.
Values are represented as percentages of those found in
normal infants of the same age. The horizontal line indi-
cates the normal values (100 per cent); the distance of the
top of each column from the normal line shows percentage
deviations.
White column : before diarrhoea
Black column : after diarrhoea
E.C. — extracellular; PL. — Plasma.

of the body weight, the water reserves in infants are low in
relation to the functions they may be called upon to perform.
In order to reconcile this apparent contradiction, it is
helpful to consider the relationship of body fluid reserves to
circulation and kidney function in malnourished infants.
Malnutrition does not affect all systems of the body equally,
fat and muscle sustaining greater losses than the extracellular

Effects of Metabolic Disturbances in Infants 157

fluid compartment. Hence the size of the latter appears to
increase with the progress of malnutrition (Kerpel-Fronius
and Kovach, 1948; McCance, 1951; Keys et al, 1950).
Haemodynamically, however, it is not the amount relative
to body weight but the absolute amount of extracellular
fluid which is of importance. Fig. 2 illustrates a striking
example of a case studied in comparison with well nourished
infants of the same length, first in a state of malnutrition and
later after dehydration due to diarrhoea had supervened.
In the malnourished infant the volume of the extracellular
fluid showed a percentage increase before and even after
diarrhoea. However, the "absolute amounts", i.e. the fluid
volumes calculated as percentages of those in normally
nourished infants of the same length, were decreased. Since
the haematocrit readings were high, the circulation time
prolonged, and the renal clearances low, high water reserves
calculated as a percentage of the body weight w^ere clearly
insufficient to maintain circulation and kidney function. The
absolute volume of the water reserves, and not just the amounts
proportional to the body weight, must be maintained in order
to conserve a normal circulation and good renal function.

Let us now consider the normal infant. When compared
with the adult, his extracellular water reserves — although high
in terms of percentage of body weight — are strikingly low in
relation to other physiological needs, namely oxygen con-
sumption, insensible perspiration and cardiac output (Fig. 3).

Thus when compared on the basis of body surface, the infant
appears to have the same oxygen consumption and cardiac
output as the adult, but his systolic output (stroke volume)
and plasma volumes are only half those of the adult; in
order to achieve the requisite cardiac output with a relatively
low plasma volume, the pulse rate is double that of the adult.
His inulin and ^-aminohippuric acid (PAH) clearance values
are low in comparison with those of the adult and also in
relation to his own cardiac output and metabolism. All his
fluid compartments are strikingly low in proportion to meta-
bolism, insensible perspiration and cardiac output.

158

E. Kerpel-Fronius

Alternatively, on the basis of body weight, the infant's
metabolism, dermal loss of water and cardiac output appear
to be very high in relation to his total body water and plasma
volume, which occupy approximately the same space as in the

100%
80
60
40
20
0

Surface/ kg.

adult

■■■III

per unit of
body surface

CO.

SV. In PAH PI.

tot.e.c.

Perspir.

Fig. 3. Haemodynamics, fluid spaces and renal function of the infant as
percentages of values for the adult.

The data represent mean values for five infants aged 4 months, with body
weights of 5-5 kg., lengths of 61 cm. and surface areas of 0-30 m.^. The
basis of comparison in the upper part of the figure is the unit of body weight,
in the lower one that of body surface. The horizontal line, 100 per cent, shows
the normal values for adults, the height of each column giving the percentage
differences between adults and infants.

CO. — cardiac output ; P. — pulse rate ; S.V. — systolic volume ;
In. — inulin; PI. — plasma; e.c. — extracellular.

adult. This relationship holds true also for the extracellular
fluid volume, although this is higher than in the adult. Renal
clearances are proportional to fluid volumes and therefore
low in relation to circulatory and metabolic rates.

Effects of Metabolic Disturbances in Infants 159

Despite the marked differences between adults and infants
in some of the physiological constants which have been men-
tioned, these functions are certainly nicely adjusted to each
other even in the infant, and his defence mechanisms are
fully capable of meeting the normal demands upon them.
When put under stress, however, the fragility of the whole
system which maintains body fluid homeostasis is exposed.

Waterless |

^ = s^'

/1. 73m

Litres
40

Total body Loss in %
water l./i.73m2 of total body
water

•

■

10
5

Fig. 4.

Vo
25

15
10

Infant Adult Infant Adult

Significance of "equal" losses when expressed
per unit of body surface.

Under pathological conditions the consequences of the
peculiar interrelationship of these functions are as follows :

(a) Water or salt loads calculated according to surface area
will, in relation to total body water content, be double the
values of the adult. The same holds true for loss of water,
equal losses per unit of surface area being twice as high in the
infant in proportion to the body water (Fig. 4).

160 E. Kerpel-Fronius

(b) Water deprivation quickly exhausts the water reserves
which are low in relation to metabolism and, consequently, to
obligatory urine volume and dermal loss of w^ater.

(c) Because of the high cardiac output required for meta-
bolic processes, and the low reserves of water to guarantee its
maintenance, circulation is endangered by even smaller water
deficits, the more so since water losses occur rapidly. It will
be remembered that the small plasma volume of the infant
relative to the cardiac output is compensated for by a high
pulse rate to ensure adequate circulation.

(d) The vulnerability of the circulation facilitates a rapid
decrease in renal clearances, which even in the healthy infant
are low in relation to his high metabolic rate. Obviously, the
infant's rather poor renal blood flow is adjusted to, and only
maintained by a relatively high cardiac output. The renal
fraction has been calculated to be 10 per cent of the total
output of the heart in infants whereas it is 20 per cent in
adults.

As pointed out by McCance and Widdowson (1957) stagna-
tion of growth plays a role in the easily disturbed equilibrium.
In a growing animal a certain amount of the food goes to the
building of its tissues. If growth is arrested, an additional
solute load formed by this fraction of the intake presents
itself for excretion by the kidneys. This will result either in a
higher urine volume, or, if the kidneys are incompetent, in
hyperelectrolytaemia and azotaemia. McCance and Widdow-
son (1957) have shown that these effects are striking in fast-
growing animals and may under certain circumstances be of
importance to the human infant. On the basis of some of the
data compiled by the American Academy of Pediatrics (1957)
an estimate has been made of the effect of arrested growth on
solute load and renal water expenditure. Solute load may be
expected to rise 13 per cent in the infant who is fed on cow's
milk, and 57 per cent in the breastfed child, causing a con-
siderable increase in urine volume. When at the same time
extrarenal water expenditure is increased by high environ-
mental temperature, or diarrhoeal losses, the water balance

Effects of Metabolic Disturbances in Infants 161

may be threatened either by high urine volumes or, in the
case of renal inadequacy, by uraemia.

In summary, the mechanisms defending body fluid equili-
brium in the infant are more easily broken down owing to the
water reserves being low in relation to the high metabolic
rate and "strained" circulation. In circumstances of shortage
this small water pool is quickly exhausted, and it is also
easily flooded by loads which, in terms of body surface, are
equal to those for adults. By decreasing the small plasma pool
rapidly, water losses lead to slowing down of circulation.
Owing to the rapidly decreasing renal clearances, as well as
the high metabolic rate producing solutes at great speed, the
relatively small water pool cannot then keep up its constancy.
Deterioration is accelerated by arrested growth.

In conclusion a particular type of dehydration in which the
infant seems to be in a somewhat less difficult position than
the adult may be mentioned. In infantile pyloric stenosis, a
condition in which starvation and dehydration develop
together, a sharp decrease of about 50 per cent in oxy-
gen consumption has been observed by Varga (1957). We
have found that this diminution in oxygen requirements
protects against stagnating anoxia brought about by the
slowing down of circulation due to dehydration (Kerpel-
Fronius e^ aZ., 1951). A low metabolic rate will most probably
also diminish obligatory water expenditures and thus delay
the progress of dehydration. Since the metabolic rate de-
creases less in the semi-starved adult (Keys et ah, 1950), the
infant may possibly be more resistant to dehydration when
he is already suffering from starvation than an adult under
similar circumstances.

REFERENCES

American Academy of Pediatrics. (1957). Report of Commission on

Nutrition. Pediatrics, Springfield, 19, 339.
CsAPd, J., and Kerpel-Fronius, E. (1933). Mschr. Kinderheilk., 58, 1.
Kerpel-Fronius, E. (1933). Z. ges. exp. Med., 90, 676.
Kerpel-Fronius, E., and KovAch, I. (1948). Pediatrics, Springfield, 2,

21.

AQKINQ — IV— 6

162 E. Kerpel-Fronius

Kerpel-Fronius, E., Varga, F., Vonoczky, J. and Kun, K. (1951).

Helv. paediat. Acta, 6, 377.
Keys, A., Brozek, J., Henschel, A., Mickelsen, O., and Taylor,

H. L. (1950). The Biology of Human Starvation. Minneapolis:

Minnesota Press.
McCance, R. a. (1951). Spec. Rep. Ser. med. Res. Court. (Lond.),

no. 275.
McCance, R. A., and Widdowson, E. M. (1957). Brit. med. Bull., 13, 3.
Varga, F. (1957). Personal communication.

DISCUSSION

Davson : Has the subject of size per se been considered as opposed to
immaturity? The pulse rate of the baby was mentioned as being faster
than that of the adult and the reasons for it were based on the im-
maturity of the organism, whereas one finds that small adult animals
have very fast pulse rates. The rabbit pulse, for instance, is well into the
hundreds and the mouse pulse is even faster.

Young : I do not think it has been suggested that the pulse rate is high
because of immaturity : it is high because of the high metabolic rate in
relation to the other constants, and in order to keep up the cardiac output.

Adolph : The effect of body size on functions such as pulse rate and
respiration rate varies considerably in any one species. Among various
species of adults it is very definite because you can get a wide range of
body sizes and can calculate what the average difference of function is.
In one species, the rat, the breathing rate is almost constant with age,
whereas the ventilation varies enormously with age, and even relative
to body size it varies somewhat with age. The pulse rate varies in accor-
dance with body size only after the age of weaning, and I should say
that none of the body size rules apply uncomplicatedly during infancy.
There are other factors, and perhaps the factor of metabolic peculiarities
is one of them.

McCance : Would anyone with paediatric experience like to comment
on the metabolic rate in pyloric stenosis?

Young : Prof. Kerpel-Fronius only quoted the example of the meta-
bolic rate in pyloric stenosis because dehydration is so likely to occur in
that condition, where the baby is also malnourished. Dr. Varga has
studied a series of malnourished cases in which he showed that the meta-
bolic rate and the oxygen uptake were low.

Talbot: Could the results shown in Fig. 2. (p. 156) be explained on
the basis of starvation with hypoproteinaemia? As in the nephrotic
patient, hypoproteinaemia tends to result in hypovolaemia. This in turn
leads to sodium and water retention and to a tendency to the formation
of extracellular oedema. It is thought that these reactions represent an
attempt on the part of the body to restore vascular volume to a satis-
factory level.

Young : When this infant became dehydrated he still had a relatively
high volume of extracellular fluid as a percentage of body weight, but

Discussion 163

the absolute volume was very low relative to that of normal infants. At
this time he showed an increase in all these handicaps of failing function.

Talbot: Did he have a low absolute plasma volume?

Young: Yes, but it was not very low per kg. /body weight.

Talbot: That might be the answer to the problem.

Bull: I should like to support that because we often find changed
plasma volumes in burns, where the situation is similar to that of ne-
phrosis. The extracellular fluid volume is not a good index of circulatory
competence ; the plasma volume can alter independently of it.

Fejfar: The longer circulation time showed in this case would mean
that the cardiac output was lower, and one can say that in all circum-
stances where the cardiac output is inadequate, there is a decrease in
renal blood flow. It is not necessary for it to be connected with a decrease
in blood volume.

Black: With a very high pulse rate and low cardiac output there
must be a fantastic decrease in stroke volume. That may be just a part
of the diminished blood volume, or the newborn infant may have a
diminished stroke volume. Perhaps the heart size is small in relation to
body size.

Young : The great value of this paper is in explaining why the baby is
more susceptible to stress than the adult, although he appears to have
plenty of water. This particular way of setting out these relationships is
very valuable from that point of view. To some people it has always been
rather a puzzle that although the extracellular fluid volume is relatively
high, it still is not high compared with the phj^siological demands made
on it.

Heller : We are always talking about the large body water content or
the high extracellular fluid volume in babies and young animals. Are
they accidental, as it were — due for instance to some prenatal endocrine
influences — or have they any functional significance? I have always
been struck by the similarity between the water metabolism of the new-
born animal and baby and animals with experimental nutritional oedema.

Davson : It depends whether the large water content is necessitated by
the geometry of the animal. If you had a sparse number of muscle fibres,
then you would have a bigger extracellular space to fill out the gap. The
animal's extracellular geometry changes gradually and the space really
has no functional significance except in so far as a muscle with more
muscle cells in it per unit of weight is a more efficient muscle.

Fourman : There is not a bag, to be filled either by muscle or by water.
Again, it all depends on the size of the cells.

Is the extra water of the baby in the muscle, the connective tissue or
the skin?

Davson : In the adult animal you can correlate the amount of collagen
with the amount of extracellular fluid.

Widdoivson : Most of the extracellular fluid is in the skeletal muscle.
This is one of the biggest tissues of the body and it is the one which
changes most in composition with development. Tissues like the heart
and the liver change very much less in their extra- and intracellular
relationships with development. The heart, for example, is very much

164 Discussion

nearer its adult composition in foetal life than the skeletal muscle. I
think a great deal of this change is in the skeletal muscle and not in
connective tissue.

Fourman : Then is there a difference in the mode of growth of skeletal
muscle on the one hand, and liver and heart on the other? Does skeletal
growth occur simply by hypertrophy without cell multiplication, and do
heart muscle and liver grow by cell multiplication? Are babies' muscle
cells smaller than those of adults and their liver cells the same size?

Kennedy: By and large what you have said is right. There is con-
siderable hyperplasia in liver during growth although there is an over-all
expansion in size of the cells with age. There is a much bigger change in
muscle cell size than in the liver cells.

Fourman : If the extracellular fluid is considered as a film over the cells,
that would account for the fact that the percentage of extracellular fluid
does not change with age so much in liver as it does in muscle.

Kennedy : Within any one tissue it should be quite easy to test that,
because cell size data based on nucleic acid determinations are available
for many different ages in a number of species, and equally, extracellular
fluid determinations are available in the same tissues.

Wallace : Muscle composition does not change much with age per unit
of muscle ; you are talking about more muscle, not per kilogram of muscle.

Widdowson : I am talking about per unit of muscle. As I have just
said, skeletal muscle changes very much in composition during develop-
ment.

Fourman : Dr. Shock, is the water content in the muscle larger in old
people than in the young ones, since muscles do atrophy in old age? We
have had that answered indirectly in Dr. Olesen's paper, but are there
any direct analyses?

Shock : I cannot answer for the human, but we have some data on the
electrolyte and water composition of rat muscle tissue. We found that
the total water content per kilogram of muscle tissue does not change
significantly with age. There was a definite shift in the water distribu-
tion in that the extracellular phase increased as the intracellular phase
decreased. The potassium, phosphorus, and nitrogen contents all went
down, but the chloride and sodium contents went up. The ratio of
potassium to nitrogen and of phosphorus to nitrogen remained constant.
Our interpretation of this was in the light of our beliefs about the reduc-
tion in active protoplasm in old age. It is as if a certain mass of proto-
plasm had disappeared and been replaced by extracellular compounds
with the appropriate amount of sodium and chloride to make up the
total water composition.

Fourman: As I said, it is not a replacement, but — to borrow Dr.
Davson's expression — a geometrical necessity to keep a film of water
around the cells.

Kennedy : But you would need to know whether the atrophy was due
to a loss of whole structural units or to a change in the size of each unit.

Shock : We do not really know this. We have not done the histology on
these muscle tissues, but we have sent some to Dr. Warren Andrew for
examination.

THE EFFECT OF HORMONES OF THE

PITUITARY AND ADRENAL GLANDS ON THE

ELIMINATION OF SODIUM, POTASSIUM AND

A WATER LOAD IN INFANT RATS DURING

THE WEANING PERIOD

JiRi Krecek, Helena Dlouha, JiM Jelinek,
Jarmila KreCkova and Zdenek Vacek

Department of Ontogenetic Physiology, Institute of Physiology, Czechoslovak

Academy of Sciences, Prague, and Institute of Embryology of the

Medical Faculty of Charles^ University, Prague

HoMEOSTATic mcchanisms in infant animals differ from
those in adults of the same species. Mechanisms regulating the
metabolism of water and electrolytes change immediately
after birth, during the period the eyes open, at the time of
weaning, in connexion with sexual maturation and perhaps
also at other stages of postnatal development. In the present
paper we should like to draw attention to the time of weaning,
which seems to us to be one of the important stages in the
development of the regulation of water and electrolyte
metabolism.

The preweaning period in rats is relatively long. Up to the
14th day of life infant rats cannot survive without the mother
rat. They are usually weaned at the end of the third week but
according to breeders natural weaning occurs only at the end
of the fourth week. This agrees with the development of
thermoregulation, for infant rats can survive very low
environmental temperatures without the mother only at the
end of the fourth week (Capek et ah, 1956).

Up to the 14th or 18th day infant rats live on breast milk
only. This is the only source of water and electrolytes, if we
disregard the urine of litter-mates that is sometimes sucked
by the infant animals. From that time onward infant rats in
addition to breast milk also actively feed on solid food and

165

166 Ji^i Kre(5ek, et al.

drink water. Gradually the mechanisms for compensation of
thirst and hunger separate. At the end of the fourth week
infant animals cease to feed on breast milk and take in food
that is normal for adult animals.

We studied the active intake of water, electrolyte solutions
and milk in infant rats using the method of free choice as
known especially from the work of Richter (1936), Young
(1949), and Young and Chaphn (1949). We observed that in
infant rats weaned at the beginning of the third week of
postnatal life there is a significant change in the regulation
of water, electrolyte and milk intake at the end of the fourth
week. The regulation of sodium intake in relation to water
intake, especially, changes. According to Richter (1936)
appetite for individual components of the diet is an important
homeostatic mechanism and is determined by the needs of the
organism.

In order to be able to offer a physiological explanation for
changes in the regulation of sodium intake it is necessary to
throw light on the relation between mechanisms of self-
selection and other components of water and electrolyte
metabolism that can be studied better and more objectively.

The adrenals and the posterior lobe of the pituitary are of
special significance for the regulation of water and electrolyte
metabolism. For this reason we have studied the effects of
hormones from these two glands. Up to the present nothing
is known of a change in function of the adrenals or in the
effect of their hormones at the end of the fourth week of life
in the rat. Indirectly one might expect such a change from
the fact that the regulation of sodium intake depends on the
function of the adrenals (Richter, 1936). There is also no dif-
ference in the size of the glands in males or in females during
the fourth week.

More is known about changes in the role played by the
posterior lobe of the pituitary during this period. Heller
(1952) showed that up to the end of the fourth week of life the
rat kidney does not react to vasopressin during a water load
in the same way as that of the adult. In addition the ability

Hormones and Homeostatic Mechanisms 167

of the kidneys to eliminate an administered water load
changes and its ability to concentrate increases. According
to Falk (1955), however, infant rats older than three days
already react to vasopressin by cessation of diuresis and an
increased excretion of chloride. As both authors use different
methods it seemed useful to study this problem first, using
several methods, and also to study the effect of vasopressin
on the elimination of sodium and potassium. Opinions on the
natriuretic effect of vasopressin also differ and we believe
that this is due to different methodological approaches.
Schaumann (1949) and Heller and Stephenson (1950) observed
that vasopressin decreases the excretion of sodium in adult
rats, while Sawyer (1952) observed an increased elimination
of this electrolyte. The former authors administered the
hormone at the same time as the water load. Sawyer first
slightly prehydrated his animals and then gave them the
hormone and the water load. According to Heller (1952) the
ability of the rat kidney to eliminate a water load changes at
the time of weaning. We therefore always used rats with a
water load.

Infant rats were weaned on the 15th- 16th day after birth
and the whole litter left in one cage. They received a standard
synthetic diet without sodium chloride. They were allowed
to choose between water and a 3 per cent sodium chloride
solution. As we expected changes in the mechanisms studied
to occur at the end of the fourth week, infant animals aged
23 and 33 days were used. Loads of warm distilled water
were administered via a stomach tube in amounts of 4-5
ml./lOO g. body weight. Subcutaneously the animals received
saline (0-5 ml./lOO g. body weight) in which the substances
studied were dissolved. The elimination of a water load was
studied for three hours after its administration or, in the
case of vasopressin, for three hours from the first micturition.
Urine was collected at hourly intervals. The amount of urine,
together with the concentration of sodium and potassium,
was determined by use of a flame photometer.

Adult rats rapidly excrete urine with a low content of

168

JiM Kre^ek, et al.

sodium and potassium after administration of a water load.
Males excrete a water load less well than females.

In our experiments the excretion of a water load was the
same in infant rats as in the experiments of Heller (1952).

Renal
water
loss

Renal

sodium

loss

Renal

potassium

loss

Without prehydratlon

With prehydratlon

: ru

/tEq/lOO g
■450

-300

■150

^Eq/100 g
450

300
150

h 0

|23 days

■■33 days

Hh^

Fig. 1. The renal loss of water, sodium and potassium during the first three

hours after administration of a water load (4-5 ml./lOO g. body wt.) to

young rats aged 23 and 33 days without prehydration or with prehydration

(2 • 5 ml./lOO g. body weight).

There are no sex differences. There are, however, consider-
able differences between infant animals aged 23 and 33 days.
These can be seen in Fig. 1. Twenty-three-day-old animals
do not ehminate the total water load within three hours.

Hormones and Homeostatic Mechanisms 169

Older animals, however, excrete nearly half the water ad-
ministered and thus excrete body water via the kidneys.
Differences in sodium excretion are also apparent. Thirty-
three-day-old animals excrete three times as much body
sodium as younger rats. The difference between both age
groups studied disappears completely, or becomes much
smaller, if 2 • 5 ml. water/100 g. body weight is put into their
stomachs two and a half hours before the actual water load.
In that case more urine is excreted by the younger animals
and losses are reduced in the older age group. Sodium losses
are also decreased in the older age group to the same level as
in 23-day-old animals. No significant changes in potassium
excretion were observed.

Differences between the two age groups are thus not con-
stant. For this reason we assume that the difference is not
due only to changes in renal function but that regulatory
mechanisms are also concerned.

The effect of vasopressin was studied in animals receiving
one water load and in prehydrated rats. The elimination of
the water load was studied according to the method of Falk
(1955). In addition the effect on total water loss three hours
after the first micturition was studied. This procedure was
similar to that of Heller (1952) who determined total renal
excretion of a water load 145 minutes after administration of
the hormone and the water load.

After 10 or 25 m-u. vasopressin/100 g. body weight, no
significant differences between the two age groups could be
observed during water diuresis. This is in agreement with
Falk (1955). Yet 23-day-old animals react differently to
vasopressin than 33-day-old rats. This difference can be
seen in Table I. After a single water load vasopressin (the
table shows the results with 25 m-u./lOO g. body weight)
increases renal water losses in the younger animals, while in
the older group total renal water losses are reduced. The
sodium loss in older animals treated with vasopressin becomes
greater after prehydration only. In younger animals the
elimination of potassium is significantly greater than in the

170

JiRi Kre^ek, et al.

Q
<
O

(^
<i3

in
O m

W ^
H H

:^ ^

- w
P w

^ &

2 f^
H o

|S
|«

H >^
O K
Ph H

5 ^

O Q
S«

in
<
>

»n

CO X

^ rH

(M rH

^ CO

CO r-^

(Tl N

CO ^

o 6

6 6

6 o

-H -H

^ -i\

-f] -H

-H -H

— ~- Sj

in CO

05 ^

o CO

f '7'

ll"

6 c^^

6 r^

7^ 6

rH 6

►o

v^-2

o* «

GO GO

J> i>

iO CO

CO CO

I— ( T-i

rH rH

^ i

j!.-

t* I-

lO PH

05 ^

CO o

05 CO

(M GO

CO r^

*•£

CO l-H

i-H rH

r-* r^

1^2=-

-H ?1

-H -H

1 o "^

C5 iO

CO CO

»0 05

CO CO

:Lo S

05 GO

(N r-l

CO o

1> (N

I— 1

r-1 r-i

r-H r- 1

rH rH

•^1

= 1

GO i>

lO 1>

CO ^

CO CO

•S

i-l r-l

T-t 1-i

T? IC

CO o

CO CO

(N ©

•g^-^

o >n

CO CO

-^ rH

CO 05

s Tjt

(N CO

(M (M

CO (M

"•£

1.^^

-H 4^

-H -H

4^ 4^

I§1

F— 1

l> o

1> ©

X ^

>r5 o

■^ ^^i

GO >0

CO lO

T? in

r^ -^

I— 1 CO

r-l i-l

(M W

■s-l

< 1

l> 00

1> 00

»n ^

CO CO

T-t 1-<

rH r-i

«*

?!§>

&l^

CO CO

(N CO

Ci M

(M CO

(N CO

•>;

^§.s

+i "5 O

+j ""7" C o

^*^:S

4i~~rd f}

^O in

^ ?-^ S

^^"-"^

its

^B-

. + CI.

§11

l|^-l

il?

^ol^^^

o g ^ W)^

^2 '^.2

rr-t "" -^ rH &H

-2V ti-g §

^ > bJD g Ci*

gq,o»r::S

S^«o|

gO,o»n:S

^XB'%

I^KSi'^

IsKS^

tsffiSi'^

Hormones and Homeostatic Mechanisms

171

33-day-old rats. Thus vasopressin has a different effect in
23-day-old than in 33-day-old animals. Evidently there is a

23 days

Renal
water
loss

Renal

sodium

loss

ml-/100 g
-2-0

-1-0

n»0

days

ju^Eq/lOO g
450

Renal

potassium

loss

^Eq/100 g
450
300
•150

. 0 C

I icontrols

cortisone 0.5

cortisone 0.125

cortisone 0.25

Fig. 2. The effect of cortisone administered for six days in different doses on
renal loss of water, sodium and potassium during the first three hours after
administration of a water load (4-5 ml./lOO g. body wt.) to young rats aged

23 and 33 days.

change in the reactivity of the kidneys to this hormone at
that period. This might be due to functional differences in

172 JiM Krecek, et al,

kidney parenchyma or to the fact that from the end of the
fourth week a regulatory factor is present which can be
influenced by loading the organism with water. It therefore
seemed all the more interesting to us to find out whether the
function of the adrenals changes at the time of weaning.

After adrenalectomy the ability to eliminate a water load is
strongly reduced in infant rats. It is difficult therefore to use
this method for solving the problem. A less direct way was
chosen — a study of the effect of substances that act in a
similar way to the main corticoids. Cortisone or cortexone
was administered for six days in various doses to 18-23 and
28-33-day-old animals. Then a water load was given. It
appeared that the effect of these substances also depends on
the age of the rats.

The effect of cortisone is shown in Fig. 2. The elimination
of a water load, sodium and potassium was determined in rats
that received 0-125, 0-25 or 0-5 mg. cortisone/100 g. body
weight. The hormone has opposite effects in the younger
and in the older age groups. In 23-day-old animals it increases
the excretion of water (as it does in the 3-day-old rats of
Falk, 1955) and sodium, while in the 33-day-old rats it
decreases both. After a dose of 0 • 25 mg./lOO g. body weight,
renal water and sodium losses in the younger animals reach
approximately the levels of the older control animals. It
appears as if the administration of cortisone compensates for
a factor missing in the younger animals but present in the
older rats. This, however, is not borne out by the way in
which a water load is eliminated by the younger rats after
cortisone. Fig. 3 shows changes in the concentration of
sodium in the urine during the course of water diuresis in
normal animals and after cortisone (0-25 mg./lOO g.). In the
control 33-day-old animals the concentration rises as the in-
tensity of water diuresis falls. In the younger group there is no
such relationship and the concentration is not lowest during
the highest diuresis. If cortisone were only a substituting
substance the course of the curves of sodium concentration
ought to be the same in 23-day-old rats receiving cortisone

Hormones and Homeostatic Mechanisms

173

and 33-day-old controls. As this is not the case and as
in the younger animals increased natriuresis is mainly due
to increased concentration at the time of maximum water

23 days

I H2O
100

— water controls

— water cortisone
-- sodium controls

— sodium cortisone

mE/1 Na
100

33 days

% H20

- 100

P- — — — — — — — — — J

IKfqyrNa"
100 -

50 -
hours

J"'

I ^^^^^_^^^^ 2

Fig. 3. The effect of cortisone (0-25 mg./lOO g. body wt./day) on the

course of the excretion of a water load and the concentration of sodium

in the excreted urine in infant rats aged 23 and 33 days.

diuresis, relations are evidently more complex. This is also
borne out by the fact that the effect of cortisone in the younger
group is variably dependent on the dose used.

174

Ji^f Krecek, et al.

This is even more evident in the case of cortexone. This
was administered by the same route as the former substance

23 days

33 days

Renal
water
loss

Renal

sodiiira

loss

Renal

potassium

loss

^q/100 g
-450

300
150

^Eo/lOG
-450

-300

-150

0 r

■IftBTT^v;?

dZI

controls

cortexone o.l

cortexone 1.0

Fig. 4. The effect of different doses of cortexone, administered for six days,
on renal loss of water, sodium and potassium during the first three hours after
administration of a water load (4-5 ml./lOO g. body wt.) to young rats aged

23 and 33 days.

but in doses of 0 • 1 and 1 mg./lOO g. body weight. Results are
shown in Fig. 4. Lower doses of cortexone had an effect
similar to cortisone, quantitatively different in younger and

23 days

33 days

Controls

Cortisone

ACTH

Cortisone
+ ACTK

Fig. 5. The effect of cortisone (0-25 mg./lCO g. body wt./day) and
ACTH (0-2 i.u. /animal/day) administered for six days on the size of
the adrenal cortex in young rats aged 23 and 33 days. Stained with

Sudan Black.

facing page 175

Hormones and Homeostatic Mechanisms 175

older animals. In younger animals it increased renal losses,
which thus nearly reached the levels of the older controls.
In 33-day-old animals water losses decreased after cortexone.
The higher dose, however, had no effect on renal losses of
water in 23-day-old animals, whereas in 33-day-old rats it
further decreased renal losses. These doses, however, are
probably toxic. Sodium losses were never significantly
altered by either dose of cortexone in the younger group. In
33-day-old animals they changed in direct proportion to the
dose used. In both age groups cortexone decreases renal
potassium losses significantly.

Thus corticoids have a different effect on the elimination
of water and electrolytes after a water load in infant rats that
have not yet reached the age at w^hich they are normally
weaned, than they have in older animals. The opposite effects
in 23-day-old animals, depending on the dose used, indicate
that these hormones cause changes that mutually interfere
with each other.

We attempted to determine whether in addition to the
pharmacodynamic effect of these hormones there is also an
effect on the regulation of adrenal activity.

The weight of the adrenals of animals receiving cortisone or
cortexone, as indicated above, dropped to about the same
extent in both 23- and 33-day-old animals. Simultaneous
administration of ACTH in amounts usually sufficient to
maintain adrenal weights of hypophysectomized animals
(0-2 i.u. per animal) prevents adrenal atrophy in both
groups. This reaction is less obvious on histological studies.
Fig. 5 shows microphotographs of the adrenal cortices of 23-
and 33-day-old animals (controls; after cortisone [0-25 mg./
lOOg./day] ; after ACTH [0 • 2 i.u. animal/day] ; and after simul-
taneous administration of cortisone and ACTH). Preparations
were stained with Sudan Black so that both the width of the
cortex and the sudanophil layers can be seen. After ACTH
there are no obvious changes in the width of the cortex and
the sudanophil layer. After cortisone and cortisone plus
ACTH differences are evident. This is even more apparent

176

Jiiii K^ECEK, et al.

in Fig. 6, which shows the results of micrometric measure-
ments of the width of the cortex and the sudanophil layer as
obtained from serial sections of the adrenals. Four adrenals
from each group were measured. One hundred sections from
each gland were used and measurements were taken from

The size of the cortex
of the adrenal section
The lower part of the
column corresponds to
the part of the cortex
stainable with Sudan
Black or Oil Red 0

I I 23 days

controls cortisone ACTH

33 days

The part of cortex
stainable with Sudan
Black or Oil Red 0,
expressed in % of to-
tal thickness of the
section of the cortex

-60

controls cortisone ACTH

cortisone

Fig. 6. See Fig. 5.

several sites of those sections. DifPerences are largest after
cortisone. In 23-day-old animals the sudanophil layer
decreases in size while the total width of the cortex remains
unchanged. In 33-day-old animals the width of the cortex
decreases and thus the relative width of the sudanophil
layer is increased. After ACTH and cortisone the proportion

Hormones and Homeostatic Mechanisms 177

of the sudanophil layer increases in both age groups but in
the younger group the size of the whole cortex is smaller.
It is difficult to interpret these changes. It is certain, however,
that according to morphological criteria the adrenals of the
23-day-old animal react differently from those of the animal
aged 33 days. This would indicate that changes in the reac-
tivity of infant rats to a water load at the end of the natural
period of weaning, and to corticoids, are also conditioned by a
different reactivity of the adrenals and the adrenopituitary
system.

This hypothesis is further supported by results from ex-
periments in which the effect of ACTH and a combination of
ACTH and cortisone (0-25 mg./lOO g.) was studied on the
elimination of water, sodium and potassium after a water
load. Results are shown in Fig. 7. As has already been shown,
cortisone prevents retention of a water load in 23-day-old
animals and considerably increases renal water losses. ACTH
is without effect. After simultaneous administration of ACTH
and cortisone, water losses decrease in comparison to losses
after cortisone only. In 33 -day-old rats results are less
evident because of the large scatter. ACTH itself causes an
increase in sodium excretion in 23-day-old animals but in
combination with cortisone it is without effect on sodium
elimination and thus removes the latter's natriuretic effect.
This effect is probably due to the lower renal water losses.
In 33-day-old animals ACTH decreases sodium losses just as
do cortisone and cortisone combined with ACTH. The same
holds good for ACTH when combined with cortexone. ACTH
prevents atrophy of the adrenals after cortisone in infant
rats aged 23 days and also prevents the effect of cortisone on
sodium and water elimination. This is not the case in older
animals. This is in agreement with the histological picture
and with the differences between 23 and 33-day-old animals.

We have thus been able to show that there is a time
correlation between changes in homeostatic mechanisms
regulating the intake of water and electrolytes appearing in
infant rats at the time of natural weaning, and adrenal

178

JiRi K^ECEK, et al.

pituitary mechanisms regulating the metaboHsm of water and
electrolytes. At the end of the fourth week of life the effect

Renal
water
loss

Renal

sodium

loss

23 days
ml/lOO g

33 days

Renal

potassium

loss

^q/100 g
-450

-300

[-150

r^^^M

czi

controls

cortisone

ACTH

ACTH •» cortisone

Fig. 7. The effect of cortisone and ACTH (for doses and duration of adminis-
tration see Fig. 5) on renal losses of water, sodium and potassium during the
first three hours after administration of a water load (4-5 ml./lOO g. body wt.)
to young rats aged 23 and 33 days.

of vasopressin on elimination of a water load changes. This
is in agreement with Heller (1952). In addition, at this time
vasopressin begins to have an effect on sodium elimination.

Hormones and Homeostatic Mechanisms 179

This is probably conditioned by the presence of a regulating
mechanism which after previous loading with water increases
the reabsorption of sodium. At the end of the preweaning
period there is a considerable change in the effect of cortisone
and cortexone on elimination of water and sodium after a
water load. Even 33-day-old animals, however, do not react
quantitatively in the same way as adult animals. This is
evidently due to the fact that only after the 33rd day does the
male adrenal begin to differ from that of the female. It may
be assumed from the results presented here that the reactivity
of the adrenals changes at the time of weaning. That change
can be in relation to the change in homeostatic mechanisms
regulating the intake of water and sodium which occurs at the
time of weaning.

REFERENCES

Capek, K., Hahn, p., Krecek, J., and Martinek, J. (1956). Studies

on the Physiology of Young Mammals. Czechoslovak Academy

Publication.
Falk, G. (1955). Amer. J. Physiol, 181, 157.
Heller, H. (1952). J. Endocrin., 8, 214.

Heller, H., and Stephenson, R. P. (1950). Nature, Lond., 165, 189.
Krecek, J., and KreCkova, J. (1957). Physiol. Bohemoslov., 6, 26.
Krecek, J., Kreckova, J., and Dlouha, H. (1956). Physiol. Bohemo-

slov., 5, suppl., p. 35.
RiCHTER, C. p. (1936). Amer. J. Physiol, 115, 155.
Sawyer, W. H. (1952). Amer. J. Physiol, 169, 583.
ScHAUMANN, O. (1949). Experientia, 5, 360.
Young, P. T. (1949). Comp. Psychol Monogr., 19, No. 5, 1.
Young, P. T., and Chaplin, J. P. (1949). Comp. Psychol Monogr., 19,

No. 5, 45.

[Discussion of this paper was postponed until after the paper by Dr.
Desaulles. — Eds.]

DIFFERENCES IN THE PATTERN OF

ELECTROLYTE AND WATER EXCRETION IN

YOUNG AND OLD RATS OF BOTH SEXES

IN RESPONSE TO ADRENAL STEROIDS

P. A. Desaulles
Research Laboratories, Pharmaceutical Department, CIBA Limited, Basle

It is a known fact that, with advancing age, the cell mass
and, correspondingly, the cell water content of the animal
decrease. This, together with a constant or increasing
extracellular water content, appears to be one of the true
signs of ageing (McCance and Widdowson, 1951; Olbrich and
Woodford-WilHams, 1956).

Although the adrenals, and more especially the adrenal
steroids, play an important part in the maintenance of the
water and electrolyte balance, only comparatively little is
known about the influence of age on the activity of the
adrenals or on the sensitivity of the organism to adrenal
steroids in animals. We were therefore prompted to study in
rats of different ages the pattern of urine and urinary elec-
trolyte excretion after treatment with two genuine adrenal
steroids, aldosterone and Cortisol, following a load of physio-
logical saline solution amounting to 20 ml. per kg.

In view of the very complex interrelationship existing
between pituitary, gonads, and adrenals during the develop-
ment of the animal from birth to maturity and old age, we
have also studied rats of both sexes. These animals were
chosen in three different groups, ranging in age from (a) five
weeks to (b) fifteen weeks to (c) one year and more.

Methods

All experiments were performed on adrenalectomized rats
of the same breed, in order to avoid interference between
the steroids injected and the steroid output of the animal's

180

Effect of Adrenal Steroids on Body Electrolytes 181

own adrenals, as well as to avoid strain-bound differences in
sensitivity.

To test the action of steroids on urinary and electrolyte
excretion, we have used the method described in detail by
Desaulles and Meier (1956), the only difference being that,
instead of collecting urine from the fifth to seventh hour after
treatment and loading, we collected it in different groups from
the 30th minute to the second hour and a half, from the
first to the third hour, from the second to the fourth hour,
and from the seventh to the ninth hour following treatment,
this procedure enabling us to follow closely the excretion of
urine and of electrolytes. Both male and female animals
were used, the age groups being:

(a) animals about five weeks old and about 50 g. in weight,

(b) animals about 15 weeks old and about 150-180 g. in
weight,

All animals used in these experiments were kept isolated in
metal cages at constant temperature (26°) and relative humid-
ity (75 per cent), the number of animals per group varying
from six to 12. The animals were given full standard rat cake
(Nafag A.-G., St. Gallen) and water ad libitum until the
beginning of the experiment.

The steroids chosen, aldosterone and Cortisol, are known to
be secreted by the rat adrenals (Bush, 1953; Singer, 1957).
Cortisol was used as free alcohol, aldosterone was used as
DL-aldosterone acetate, the activity of which is just one half of
D-aldosterone (Schmidlin et al., 1955, 1957). All substances
were dissolved in sesame oil and injected intramuscularly.

The doses used in these experiments were chosen from
previous experiments (Desaulles and Meier, 1954; Desaulles,
1958) and lay within a dose range corresponding to sub-
maximal effects. For aldosterone acetate 0-01 mg./kg. was
given, and for Cortisol 5 mg./kg.

As the excretion of urine and urinary electrolytes differs

182

P. A. Desaulles

in amount in animals of differing age and weight, the results
are expressed as percentages of the values of control animals
for urinary excretion in ml., and for sodium and potassium
excretion in m-mole. The differences between the sodium/
potassium ratios of treated and control animals are, on the
other hand, expressed in absolute values.

Results
Effect of aldosterone

In the male rat, aldosterone produces a marked inhibition
of urinary output that is most pronounced in young animals

160

120
100
60
60
AO

+
0

lZ74^6789f

Fig. 1. Urinary excretion of adrenalectomized male rats of dif-
ferent age groups treated with aldosterone (0-010 mg./kg.).
Abscissa: Duration of experiment (hours); collecting period

2 hours.
Ordinate: Urinary excretion as a percentage of the values of
control animals.
Continuous line : 5-week-old rats.
Interrupted line: 15-week-old rats.
Dotted line : one-year- and more-old rats.

OH CH2OH

- 0— CH CO

....'

'''•*— 1

5«fc^

-" ^.

^«.

[T-''

■^ ""M

,— —

' >

^^:^^

'-/

^c^

^O/'^

^ _

and tends to diminish — at first in duration and then in
intensity — with increasing age (Fig. 1).

Effect of Adrenal Steroids on Body Electrolytes 183

Aldosterone prevents sodium excretion in a very marked
manner in about the same intensity and for about the same
duration (five to seven hours) in all age groups (Fig. 2).

The only difference to be noted is that in old animals the
onset of the sodium-retaining effect of the steroid is retarded,

%
zzo

zoo

ISO

140
120
100
80
60
40
20

I ■

^ OH CH2OH

0 — CH CO

,•.-.

^,'

^ —■—

-•^

> 'A

tr.^

-«--«<

^^-*

b-^— ^
^

c-j

•^

^-J^"-"^

¥m^

^*^^

IZ'i456789t

Fig. 2. Urinary sodium and potassium excretion of adrenal-

e<*I;omized male rats of different age groups treated ^^^th aldosterone

(0 010 mg./kg.).

Thick line : sodium excretion.
Thin line : potassium excretion.
Other figures as for Fig. 1.

the maximal effect falUng in the collecting period of the third
hour, instead of in the preceding period.

The effects of aldosterone on potassium excretion depend
upon the age groups in question.

In young animals, aldosterone does not affect potassium
excretion until the fourth collecting hour. From the fifth

184

P. A. Desaulles

hour onward it induces a clear-cut reduction in potassium
excretion, which reverts to normal in the ninth hour. On
animals of the adult group aldosterone has practically no
effect at all. In old rats, however, aldosterone markedly
enhances potassium excretion.

f.d

1.6
lA
1,2
1,0
Q8
0,6
OA
0,2

' cf

1 1

O — CH CO

" rv

^"^

u-,.. -

— *»

..."

J -I

•

,..— <

\ Z ■} 4. ? 6 7 8 9 f

Fig. 3. Urinary sodium/potassium ratio of adrenalectomized male

rats of different age groups treated with aldosterone.
Ordinate : Difference between sodium /potassium ratio of experi-
mental animals and controls.
Other figures as for Fig. 1.

If we consider the sodium/potassium ratio, we observe that
aldosterone reduces it markedly during the first hours of the
experiment in all groups (Fig. 3), its maximum occurring in
the first collecting period for young and adult groups, and
showing a certain delay (three hours) and greater intensity
( — 0 • 95 against —0 • 75 to —0 • 80) in the old age group. From
the fourth hour onward there is an increase in the ratio for

Effect of Adrenal Steroids on Body Electrolytes 185

young animals (due to potassium retention), whereas adult
and old animals return to a range within control values, the
adult group reacting more readily than the old animals.

In the female rat, aldosterone also reduces the urinary out-
put, but to a somewhat smaller extent than in males ( —40 per
cent on the average, against about —60 per cent in males)
(Fig. 4). As in males, young animals tend to respond more

160
140
120
100

60
AO

20
40
60
80

... — — ,

T OH CHgOH

O— CH CO

^N,

*•

'%,

•.

t
•

^^.-^J

•*•

"^^

C^^

7^"^

IZJ4^6 78 9t

Fig. 4. Urinary excretion of adrenalectomized female rats of dif-
ferent age groups treated with aldosterone (0 010 mg./kg.).
Figures as for Fig. 1.

markedly although there is a certain delay in the onset of the
effect. In contrast to males, with increasing age a short
period of urinary retention is followed by a strong diuretic
response.

On sodium excretion aldosterone exerts a very pronounced
inhibiting effect of about the same relative intensity as in
males in all age groups (Fig. 5). In contrast to that in males,
this effect is followed by a period of sodium excretion, most
marked in old animals (+80 per cent), the values returning
towards the norm in the ninth hour.

186

P. A. Desaulles

On potassium excretion the enhancing effects of aldo-
sterone are more marked and begin at an earlier age than in
males, old animals showing the most pronounced effect.

On the sodium/potassium ratio the effects are much more
marked than in the case of males (Fig. 6). Young animals
respond with a reduction that is marked ( —0 • 90), but of slow

%
zco

ISO
14^
120
100
30
60
40

+
0

+ OH CHjOH

O— CH CO

,.A

-^ -''

V '*•

' — i

?^^ y

**..

,'-'

^S^

\ — ^'<,

^^>^ *

I *

3r^--=H

// "^

RVV"'

9 t

Fig. 5. Urinary sodium and potassium excretion of adrenalecto-
mized female rats of different age groups treated with aldosterone.
Figures as for Fig. 2.

onset (maximum in the fifth hour), the values returning to
within control limits at the end of the experiment.

In adult and old females, the reduction in the sodium/
potassium ratio is more intense (—1-29 and —1-40 respect-
tively) and rapid in onset (maximum in the first collecting
period). This effect lasts longest in old animals.

The rapid lowering of the sodium/potassium ratio is

Effect of Adrenal Steroids on Body Electrolytes 187

followed, in contrast to the situation for male animals, by a
very pronounced and rapid rise (more in adult than in old
animals) to high positive values (+0-95 and +1'28, re-
spectively), this effect tending to return within control values
in the ninth hour.

1.6

1,0

0,6
OA

o,z

0,2
0,1
Q6
0,8
1,0
1,2

U . ^ ^

/ 2 54? (J 7 5 9/

Fig. 6. Urinary sodium/potassium ratio of adrenalectomized female

rats of different age groups treated with aldosterone.

Figures as for Fig. 3.

TT OH CH9QH

o-,!„^'

^"~

" (^

^^s^

•♦♦..

"s^

'\.

s^^

'^N

"-^,

•

•
«

""^nC

«
•

^■"""^

y*^^

r^^

.•<

•

.""

Effect of Cortisol

On urinary output, Cortisol has, as is well known, a marked
enhancing effect (Marcus, Romanoff and Pincus, 1950;
Desaulles, Schuler and Meier, 1955). In male rats this effect is
well developed, and ageing does not seem to modify it
markedly (Fig. 7.)

On sodium, Cortisol exerts initially a sUght retaining effect
that has already been reported (Dorfman, 1949; Johnson,

188

P. A. Desaulles

1954; Desaulles, 1958) and which is followed by enhanced
sodium excretion (Fig. 8). In males these effects tend to dis-
appear with advancing age.

On potassium, one observes the characteristic excretory
response whose intensity is particularly high in young animals,
its onset being somewhat more rapid in adult and old animals.

r — • ■

- ^ CHgOH

"x^-]

.....P<'

.'-::.

v^^

^•*

■ —

*•

■■■■ilSli

160
140
120
100
00
60
JO

+ jO^

I Z ? 4 f ^ f S 9 t

Fig. 7. Urinary excretion of adrenalectomized male rats of dif-
ferent age groups treated with Cortisol (5 mg./kg.).
Abscissa : Duration of experiment (hours) ; collecting period 2 hours.
Ordinate : Urinary excretion as a percentage of the values of control

animals.

Continuous line : 5-week-old rats.
Interrupted line: 15-week-old rats.
Dotted line : one-year- and more-old rats.

The effect of Cortisol on the sodium/potassium ratio is first
to lower it moderately in males, and to raise it afterwards to
high positive values (Fig. 9). This effect, most marked in
young animals, declines with increasing age.

In female rats, Cortisol has a stronger enhancing effect on
urinary output than in males (Fig. 10). With age, this effect
increases and a certain latency of onset seems apparent.

Effect of Adrenal Steroids on Body Electrolytes 189

On sodium, Cortisol has similar retaining effects in females
as in males and these disappear in old animals (Fig. 11).

As regards the enhanced sodium excretion which appears
later, females react differently from males. Instead of dis-

8 9 f

Fig. 8. Urinary sodium and potassium excretion of adrenalecto-

mized male rats of different age groups treated with Cortisol

(5 mg./kg.).

Thick line : sodium excretion.

Thin line : potassium excretion.

Other figures as for Fig. 1.

appearing with increasing age, the response remains high and
its onset is more rapid in ageing females, although in this
experiment animals of the adult group do not respond
clearly.

The effect of Cortisol on potassium excretion in females is
similar to that observed in males, i.e. it is enhanced, the
effects tending to decrease in intensity with age.

1,6
1,6
1,4
1.2
1,0
0,8
0,6
OA
0,2

CHoOH

HO ^^ ^°

" r ■

-OH

r^^

■ 4

^^s^

—^

* i.

^^*

^./-y^-^'

*^*«

^^'

Fig. 9. Urinary sodium/potassium ratio of adrenalectomized male
rats of different age groups treated with Cortisol (5 mg./kg.)

Ordinate : Difference between sodium/potassium ratio of experi-
mental animals and controls.

Other figures as for Fig. 1.

%
160

140

120

100

jiO

+
0

.... .

+ CH20H ^

r— — ^

HO CO — j

vH — ^--OH _•

Ar^

1
t

^y\s

J*-^

*--.

•**^

/^.

•
t

i»0

Vs.

^cl

Fig. 10. Urinary excretion of adrenalectomized female rats of dif-
ferent age groups treated with Cortisol (5 mg./kg.).

Figures as for Fig. 1.

1 ■

HO CO — ^

,A^xkJ

r"i

/ J

r""

^->:^

/''

**v^

// y

,— — "

k \

\r^

•N

••.

J/,''

^ 1

.-^

,••'

—0

^"*x

^^^^

:;<

/*N

%
zzo

zoo

180
160
140
IZO
100
80
60
40

+
0

I Z 'J 4 f 6 7 8 9 t

Fig. 11. Urinary sodium and potassium excretion of adrenalecto-
mized female rats of different age groups treated witli Cortisol
(5 mg./kg.). Figures as for Fig. 2.

1,8
1.6
1,4
l,Z
1,0
QB
0,6
0,4
0,2

$ a

.20H_
3

-OH

f* 1

^w^

,.-^'^

['•♦.

j j>

%,"

■•,

"*^o

^,,

9 /

Fig. 12. Urinary sodium/potassium ratio of adrenalectomized

female rats of different age groups treated with Cortisol (5 mg./kg.)

Figures as for Fig. 3.

192 P. A. Desaulles

On the sodium/potassium ratio, the efPects are comparable
to those obtained in males but are of greater intensity (more
than twice the values observed in males) and of more rapid
onset, and also tend to diminish rapidly with increasing age
(Fig. 12).

Discussion

From the experimental results presented, it follows that age
modifies the sensitivity of adrenalectomized rats to the
influence of the adrenal steroids investigated. These modifica-
tions are qualitative as well as quantitative, the sex of the
animals also playing an important role.

Whereas in male rats increasing age tends to reduce to
control values the inhibiting effects of aldosterone on urinary
output, it tends in females to induce a marked secondary
diuretic response. The primary retention of sodium produced
by aldosterone is of about the same order of magnitude in all
animals, whether male or female, but ageing greatly increases
the concomitant loss of potassium, this effect being particul-
arly clear in male and female rats of the old age group.

In contrast, Cortisol has an enhancing effect on diuresis
which, especially in females, tends to increase with advancing
age, whereas in males it is more intense from the onset and
remains of about the same order. The effects of Cortisol on
sodium excretion are profoundly different with advancing age
in rats of different sexes. In males these effects tend to dis-
appear completely. In females, on the other hand, they
appear earlier and remain of the same order of magnitude.

After Cortisol treatment we can observe comparable dif-
ferences in potassium excretion. Whereas young males
respond with an intense potassium excretion which drops
rapidly as the animals grow older, these changes are only
moderate in females, potassium excretion remaining high
until old age and its onset merely retarded.

These age and sex-bound differences become particularly
clear if we study the variations in the sodium/potassium ratio.
The sensitivity of the animals to the effects of aldosterone

Effect of Adrenal Steroids on Body Electrolytes 193

increases with advancing age, females showing much greater
differences than males. By way of contrast, sensitivity of the
animals to Cortisol diminishes with advancing age, females
showing here too a greater sensitivity than males.

The similarity of the curves of the sodium/potassium ratio
for aldosterone and Cortisol is also striking and leads us to the
problem of (a) the primary and (b) the secondary effects of
these substances, and furthermore to the problem of the
classification of adrenal steroids on the basis of what has been
considered their most important physiological effects.

From previous experiments with aldosterone one is inclined
to consider as primary effects both sodium retention as a
consequence of increased tubular resorption of sodium ion,
and potassium excretion as a consequence of the exchange
between sodium ions in the tubule cells (Cole, 1957; Stanbury,
Gowenlock and Mahler, 1958). Sodium retention remains of
about the same order of intensity and duration from youth to
old age in both males and females. It is concomitant potas-
sium excretion that rises strikingly with advancing age both
in males and females.

On the other hand, the diuresis induced by aldosterone,
which is most apparent in old female animals in the later
phases of the experiment, is most probably of secondary
origin, its causes lying in the effect of aldosterone on the
sensitivity of adrenalectomized animals to endogenous anti-
diuretic hormone (Gaunt, Lloyd and Chart, 1956).

As for Cortisol, its essential effect seems to lie in the very
marked potassium excretion which is regarded as running in
parallel with its catabolic effects.

Its effect on potassium excretion tends to diminish with
advancing age, male animals being here more susceptible than
females. Conversely, the diuretic and sodium-excreting pro-
perties of Cortisol seem to be caused essentially by the potent
antagonistic effect of this steroid on the sensitivity of the
animal to antidiuretic hormone; these properties tend to dis-
appear with increasing age in males but not in females.
Aldosterone and Cortisol tend to induce a greater diuretic

AGEING — IV — 7

194 P. A. Desaulles

response and concomitantly higher sodium excretion with
advancing age, especially in females.

This, together with the similarity of the changes in the
sodium/potassium ratio induced by aldosterone and Cortisol
during these experiments, even if the factors of ageing and sex
act differently on them, underlines certain similarities of effect
in a number of known adrenal steroids which have already
been stressed (Meier and Desaulles, 1956; Gaunt and Chart,
1958). Relative dosage, time, age, experimental conditions
and different stages of homeostasis are among the factors
modifying these similar patterns of effect. The relation of
homeostasis to the development of the animal organism is too
complex to permit of any definite statement. We have
simply tried to show that the properties of certain hormones
may be profoundly affected by such factors as sex difference
and increasing age, and that these differences may act in the
same or in quite different ways and thus contribute towards
a better understanding of pathophysiological changes due to
age.

Summary

It has been shown that in rats of differing age and sex the
sensitivity to the influence of aldosterone and Cortisol on
urinary electrolyte excretion varies greatly.

Whereas age tends to increase sensitivity of the animals to
the effects of aldosterone, their sensitivity to Cortisol by way
of contrast tends to diminish.

On the other hand, female animals show a greater respon-
siveness to these changes than male animals.

These results are discussed.

Acknowledgement

I should like to express my thanks to Mr. H. D. Philps (MA. Cantab.)
for his kind assistance in the preparation of the English text of this paper.

REFERENCES

Bush, I. E. (1953). Ciba Found. Colloq. Endocrin., 7, 210. London:

Churchill.
Cole, D. F. (1957). Endocrinology, 60, 562.

Effect of Adrenal Steroids on Body Electrolytes 195

Desaulles, p. a. (1958). In Aldosterone, ed. Muller, A. F., and

O'Connor, C. M., p. 29. London: ChurchilL
Desaulles, P. A., and Meier, R. (1954). Unpublished data.
Desaulles, P. A., and Meier, R. (1956). Schweiz. med. Wschr., 86,

1060.
Desaulles, P., Schuler, W., and Meier, R. (1955). Schweiz. med.

Wschr., 85, 662.
DoRFMAN, R. I. (1949). Proc. Soc. exp. Biol., N.Y., 72, 395.
Gaunt, R., and Chart, J. J. (1958). Symposium on Homeostatic

Mechanism, Brookhaven National Laboratory (in press).
Gaunt, R., Lloyd, C. W., and Chart, J. J. (1956). Colston Pap., 8, 233.
Johnson, B. B. (1954). Endocrinology, 54, 196.
Marcus, S., Romanoff, L. P., and Pincus, G. (1950). Endocrinology,

50, 286.
McCance, R. a., and Widdowson, E. M. (1951). Proc. R. Soc, 138 B,

115.
Meier, R., and Desaulles, P. A. (1956). Rev. iber. Endocr., 3, 565.
Olbrich, O., and Woodford-Williams, E. (1956). In Experimental

Research on Ageing, ed. Verzar, F., p. 236. Basle: Birkhauser.
Schmidlin, J., Anner, G., Billeter, J.-R., and Wettstein, A. (1955).

Experientia, 11, 365.
Schmidlin, J., Anner, G., Billeter, J.-R., Heusler, K., Ueber-

WASSER, H., WiELAND, P., and Wettstein, A. (1957). Helv. chim.

Acta, 40, 2291.
Singer, B. (1957). Endocrinology, 60, 420.
Stanbury, S. W., Gowenlock, A. H., and Mahler, R. F. (1958). In

Aldosterone, ed. Muller, A. F., and O'Connor, C. M., p. 155.

London: Churchill.

DISCUSSION

Adolph: Dr. Krecek, how do you account for what I take to be an
absence of water diuresis in rats at 23 days of age? Is it because they are
weaned early? Unweaned rats have a large water diuresis at this age.

Kfecek : Water diuresis always occurs in rats of 23 days of age, but
during the first three hours after a water load there is a retention of one-
twentieth of the load. This figure was arrived at from balance tests,
being the difference between water load and water excretion.

Heller : I am very pleased about the agreement between your findings
and ours, Dr. Kfecek. You use much the same technique as we did to
estimate the response of your animals to vasopressin, and you say that
you collect the urine for three hours after the injection. Did you have a
special reason for choosing this time interval?

Kfecek: Yes, it was because the pattern of diuresis changes after
the administration of vasopressin, so that after three hours the excre-
tion of the water load is complete.

Heller : For how long did your dose of vasopressin inhibit the water
diuresis of the adult animals which you used for comparison?

Kfecek: When we give enough vasopressin for maximum diuresis we

196 Discussion

find that in young animals there is very httle difference in water diuresis
as compared to that in animals 33 days old. Between adult animals and
33-day-old ones there is no difference, but between 23 and 33 days there
are variable, but statistically significant differences in the excretion of
the water load.

Heller: That is almost exactly what we found; our age groups were
20-22 and 29-31 days after birth.

Borst : Has diurnal rhythm been taken into account by Dr. Desaulles
and Dr. Kfecek? Big differences can arise if the controls and experiments
are not done at the same time each day.

Kfecek : Our experiments and controls were always done at the same
time in the morning. They were done in summer and in winter, with the
same results.

Desaulles : Ours were done very early in the morning.

Adolph: Did you run controls without the hormones?

Desaulles: Every group was run with controls.

Borst: Light is not important. In blind people the diurnal rhythm
remains normal if they are in light during the night and in the dark
during the day.

Desaulles : We cannot cope with every activity, but we did think that
light might be one of the problems.

Fourman: Dr. Desaulles, you drew an analogy between the effects of
aldosterone and of Cortisol on the excretion of sodium and potassium.
If one considers the excretion ratio of these two ions in the urine, the
effects do appear to be analogous. Fred Bartter and I first became
interested in this question in 1949, when we began some studies which we
completed about a year ago (1957. J. clin. Invest., 37, 872). In the human
we were impressed with the fact that Cortisol produces a large increase
in the excretion of potassium which is transient even if the administra-
tion of Cortisol is continued. It is not necessarily accompanied by a
retention of sodium, but it is associated with an increase in the pH of the
urine. With aldosterone, on the other hand, the loss of potassium is not
transient ; it is accompanied by retention of sodium and the pH of the
urine does not change. On the basis of these experiments we felt that the
effects of these two steroids on the electrolytes were quite different. We
even suggested that the early effect of Cortisol on potassium was secon-
dary to a release of potassium from the tissues.

Desaulles: That is my opinion too.

Fourman: What strikes me is that pharmacologists are mistaken as
long as they equate these end-effects of excretion of sodium and potas-
sium, and as long as they speak about the alteration in Na/K ratio and
use this as a measure of aldosterone effect. I think the early effect of
Cortisol on potassium may be a tissue effect ; on the other hand the effect
of aldosterone on the secretion of sodium may well be a renal effect, and
I think you think so too.

Desaulles: Partly, yes. The Cortisol effect on potassium is surely
cellular.

Fourman : The early rise in potassium excretion with Cortisol is probably
a cellular effect. The results will not be very reliable if you assay a

Discussion 197

hormone by a change in Na/K ratios in the urine, when the change is
produced by two different mechanisms.

Desaulles: I just wanted to show in this experiment that ages bring
changes, and sex too.

McCance : We are deahng here with the reactions and responses of an
end organ, and it is a Httle difficult, apparently, to disentangle them.

Fourman: The effects that I am speaking about concern the imme-
diate loss of potassium within eight hours of giving Cortisol. This im-
mediate large loss is completely out of proportion to any nitrogen loss,
and in fact precedes a measurable nitrogen loss from the body. I was not
concerned with the later catabolic response, only with the early potas-
sium loss which is quite transient, and which is what people are con-
cerned with when they assay so-called aldosterone activity in urine by
Na/K ratios.

Milne: I am confused by your statement. Dr. Fourman. You make
a clear distinction between potassium excretion following (a) Cortisol,
and (b) aldosterone. You tell us that the potassium excretion following
Cortisol is out of proportion to the nitrogen loss, and therefore is a true
potassium excretion. You say that the difference is that potassium comes
from the cells, but where do you think the potassium comes from after
aldosterone excretion?

Fourman : It does appear that potassium excretion after aldosterone
may be attributed to a change in the sodium-potassium exchange in the
renal tubule, whereas the large and early transient potassium excretion
with Cortisol is not necessarily accompanied by any retention of sodium,
and is associated with a rise in pH of the urine. Ultimately the potas-
sium has got to come from the cells in both cases. But in the first case
we are concerned with a primary renal effect, and in the second case I
suggest — and it is only a suggestion— that there may be a liberation of
potassium — presumably organically bound (in view of the alkaline
urine) — from the cells, and that may be called a primary cellular effect.

Kennedy: Dr. Desaulles, when you spoke about the influence of sex
were you thinking in terms of the actions of androgens or oestrogens?
Your animals were not spayed, but is there a true sex difference?

Desaulles: The effect could be changed by ablation of one of the so-
called specific sex organs. If you castrate males, you modify the results
of the experiment quite considerably; if you spay the female, the
changes are much less impressive, but there still remains a great
difference between the two sexes. I want to stress here a point that is
always a little puzzling to me: if you spay a female you produce a
marked adrenal enlargement, but if you castrate a male the enlarge-
ment of the adrenals is not so obvious. We do find quite a lot of sex-
bound differences in different functions of the animal, so I think that a
very important part is played by the gonads.

Kennedy : If I understand you rightly, there is still a difference in the
absence of both the adrenals and the sex organs.

Swyer : Is this difference after castration in the two sexes one which is
independent of the time after castration, i.e. after a long time do the
differences between the sexes become less?

198 Discussion

Desaulles: That is a very important point, because it is very well
known that if you castrate an animal and the time-lag is too great, the
responsiveness of certain sexual adnexal organs disappears. We used
the following method in our work. We castrated the animals, in these
and other similar experiments, and at different periods after castration
we tested the sensitivity of their sexual adnexa. We observed that
what was found by Parkes and Deanesley about 20 years ago is still
absolutely valid. You must begin the experiments between two and
three weeks after castration ; after that the sensitivity diminishes very
rapidly. If you wait from one to three months some responses disappear
completely, and you need very high dosages of the substance to obtain
resensitization of certain organs.

Swyer : I was thinking not so much of that, but of whether the re-
sponse to the adrenal steroids shows a sex difference which is diminished
but not entirely removed by castration.

Desaulles : I have not enough experience of all the effects that may be
considered to say anything definite about this point, but it still seems to
me that castration in itself does not suffice to abolish certain existing
differences between the sexes in their response to adrenal steroids.

THE EFFECT OF AGE ON THE ELECTROLYTES

IN THE RED BLOOD CELLS OF DIFFERENT

SPECIES

M. J. Karvonen

Department of Physiology, Institute of Occupational Health,
Helsinki

Two kinds of age changes may occur in the red blood cells.
The erythrocytes themselves have a definite length of life
which may be determined in various ways, whereas the
longevity of "fixed" tissue cells generally cannot be as
exactly indicated. Thus, as cells erythrocytes may be
"young" or "old". On the other hand, like any other cells
of the body, the red cells may be a part of a young or of an
old organism.

Cellular age

In order to study age changes in the erythrocytes as cells,
two principal ways are open. One of them is to produce
anaemia, e.g. by bleeding, and thus to stimulate erythropoie-
sis, so that a large proportion of the circulating cells will have
been produced within a relatively short period. The writer is
not aware of any systematic study of the red cell electrolytes
throughout the regeneration after acute bleeding. In micro-
cytic anaemias of man — which is the type seen also in bleeding
anaemia — the concentration of potassium in erythrocytes is
lower than normal (Maizels, 1936). In other types of anaemia
a change in the opposite direction may occur (Maizels, 1936;
Selwyn and Dacie, 1954; McCance and Widdowson, 1956).
However, changes in the electrolytes observed in any type of
anaemia are not necessarily dependent on the age of the
erythrocytes, but may be caused by many other factors
associated with anaemia.

199

200 M. J. Karvonen

Recently it has been claimed that other methods for study-
ing young or old erythrocytes might be feasible. According
to Borun, Figueroa and Perry (1957), after centrifugation of
blood the bottom layer contains the oldest cells, and the
surface the youngest ones. An analysis of the different layers
has shown that — at least in human adult blood — the packing
is closest and the amount of intercellular plasma lowest in the
bottom layer, but when the effect of different packing is
corrected there is no difference between the sodium and
potassium concentrations of the bottom and the surface
erythrocytes (Leppanen, personal communication). Serial
osmotic haemolysis has also been suggested as a means for
differentiating young and old erythrocytes (Simon and
Topper, 1957). The value of these methods is not yet clear.
However, the nature of the methods used suggests that
changes in the electrolyte metabolism of the erythrocytes
may be involved in their ageing, though such changes may not
necessarily result in differences in the concentration of sodium
and potassium.

Age of the animal

As a mixed population of different cellular ages, erythro-
cytes are easily available. The availability and development
of flame photometric analysis have been a stimulus for several
investigations of the electrolyte content of the red cells. It
has been found that in general the sodium and potassium
content of the erythrocytes in vivo is fairly stable, typical of
the species, and resistant to many physiological and pharmaco-
logical agents. However, in disease, particularly in febrile
states, erythrocytes tend to lose potassium and gain sodium:
in other words, the electrolyte composition of the erythro-
cytes moves closer to that of plasma.

Sheep and other ruminants. It may be inferred from results
published by Green and Macaskill (1928), and by Wise and
co-workers (1947) that the intracellular potassium concen-
tration is higher in the blood of young calves than in that of
adult cattle. These two papers were the first to indicate that

Age Changes in Red Blood Cells

201

the red cell electrolytes may change with age. The subject
was taken up by Hallman and Karvonen (1949) in another
species, sheep. A distinct difference between foetal and
adult Finnish sheep was observed, in the sense that the con-
centration of potassium in erythrocytes was higher in foetal
than in adult sheep. Fig. 1 shows the differences in both

50 100

Sodium millieq. per litre

Fig. 1. The concentration of potassium and sodium in the
erythrocytes of sheep foetuses (F) and their mothers (M)
belonging to the Finnish breed (Hallman and Karvonen,
1949). The corresponding figures for the red blood cells of
adult sheep of other breeds fall along the two straight lines
(Evans, 1957).

sodium and potassium concentrations. The sum of the two
electrolytes tends to remain constant with age.

Widdas (1954) confirmed this observation and found a
gradual decrease of the potassium and an increase of the
sodium with advancing foetal age.

The study by Hallman and Karvonen (1949) brought out
another interesting finding. In 1898 Abderhalden published

202 M. J. Karvonen

the first values for the sodium and potassium concentration
of adult sheep erythrocytes, and found that they belong to the
"low potassium — high sodium" type. In 1937 Kerr observed
higher potassium concentrations, with a large variation
between individual sheep. In the determinations of Hallman
and Karvonen, the erythrocytes of the Finnish sheep turned
out to be — contrary to those of Abderhalden — of the "high
potassium — low sodium" type, containing still more potas-
sium than the red cells of Kerr's sheep. Sheep erythrocytes
thus show a large range of individual variations in the
electrolyte composition.

The electrolytes are not the only constituents of the red
cells in which individual sheep differ. The solubility character-
istics of sheep haemoglobin obtained from different countries,
from different breeds, or from different sheep may also differ
(Karvonen, 1949; Karvonen and Leppanen, 1952). It was
natural, as a working hypothesis, to connect with each other
these differences in the red cell electrolytes and in the type
of haemoglobin. In the first five samples representing dif-
ferent breeds of sheep, haemoglobin prepared from the low
potassium erythrocytes actually showed a crystal habit
different from that of the high potassium cells (Karvonen
and Leppanen, 1952).

Since these early attempts the red cell electrolytes of sheep
have become the subject of intense study. The individual
differences in the electrolyte composition have been shown to
be permanent characteristics (Evans, 1957). The occurrence
of different types of red cells in a number of breeds has been
studied, and the genetics of the inheritance have been worked
out (Evans, 1954, 1957; Evans and King, 1955; Evans et al.,
1956; Evans and Mounib, 1957).

The application of paper electrophoresis to sheep haemo-
globins has shown that though there is a definite association
between the electrolytes in the red cells and the haemoglobin,
this association is not absolute (Harris and Warren, 1955;
Evans et al., 1956; Evans, Harris and Warren). On the other
hand, the haemoglobin present in the red blood cells has an

Age Changes in Red Blood Cells 203

influence on the concentration of potassium in the whole
blood of both high potassium and low potassium sheep, and
thus presumably also on the concentration of potassium in the
cells themselves (Evans et ah, 1956).

The study of individual differences between adult sheep thus
shows that the type of haemoglobin is associated with the red
cell electrolytes, but that other factors also play a role.

Haemoglobin changes with age: the haemoglobin of a foetus
differs from that of an adult, but after the production of the
adult type is once established, no further changes with age are
known to occur. For instance, the haemoglobin of a sheep of
the age of 14 years showed solubility characteristics identical
with that of younger animals (Karvonen, unpublished.)

The transition from foetal to adult life involves a change of
haemoglobin and of the red cell electrolytes. In sheep, these
two changes appear to start before delivery and to be com-
pleted some time after birth (Karvonen, 1949; Hallman and
Karvonen, 1949; Widdas, 1954). Whether the changes are
exactly parallel would be a subject of considerable theoretical
interest.

Other species. The effect of age on the electrolyte concen-
tration of red cells has been studied in few other species.
Remarkably enough, a relationship just opposite to that in
ruminants has been found: the sodium concentration is
higher and the potassium concentration the same or lower in
foetal than in adult erythrocytes, at least in man (Hallman,
Osterlund and Vara, 1954; Osterlund, 1955; McCance and
Widdowson, 1956), pig (McCance and Widdowson, 1956), and in
guinea pig (Widdas, 1954, 1955; Karvonen and Leppanen,
unpublished). The concentration of chloride changes in the
same direction as that of sodium.

Underlying mechanisms

It has been pointed out by Conway (1957) that the smaller
a cell, the more work per unit cell volume a "sodium pump"
must do in the same environment of plasma or extracellular
fluid, in order to keep the intracellular sodium at constant

204 M. J. Karvonen

level. A similar conclusion applies to an eventual "potassium
pump". The erythrocytes of a foetus are larger than those of
an adult. With constant activity of the electrolyte pumps an
increase in the cell sodium and a decrease in potassium would
be expected from foetal to adult life. This is the direction of
development in the ruminants, but not in the other species
examined. It is rather questionable whether the decrease in
cell size even in the ruminants is an important cause of the
changes of the red cell electrolytes.

With the aid of in vitro studies much progress has been
made in elucidating the mechanism of cation transfer across
the red cell membrane. The application of these methods to
the erythrocytes of the foetus suffers from a serious limitation :
the cells of foetuses (at least human and sheep) haemolyse
spontaneously and rather fast in vitro. To some extent the
rate of haemolysis is dependent on oxygen tension, high
oxygen tension increasing the rate of haemolysis, but haemo-
lysis also occurs at an appreciable rate in blood exposed to
nitrogen. Haemolysis in human cord blood may also be
retarded by administering ascorbic acid to the mothers before
delivery, but even so the rate of spontaneous haemolysis
remains considerably higher than in adult blood. An addition
of ascorbic acid in vitro is without effect (Raiha, 1956, and
personal communication).

Summary

Information on changes in the electrolyte metabolism of
individual erythrocytes during their life cycle is meagre.
However, the claims that young and old cells may be separ-
ated with the aid of centrifugation or serial haemolysis suggest
that changes in the electrolyte metabolism may be involved
in the ageing of the red cells. Differences in the actual sodium
and potassium concentrations have not, however, been
demonstrated.

In sheep and cattle the erythrocytes of a foetus contain
more potassium and less sodium than those of an adult. In
man, pig and guinea pig, a difference in the opposite direction

Age Changes in Red Blood Cells 205

has been observed. The eventual association of the difference
in red cell electrolytes with a difference in haemoglobins and
with a difference in cell size is discussed.

In vitro studies of foetal erythrocytes and, particularly,
their interpretation, are handicapped by a fast rate of spon-
taneous haemolysis in foetal blood. In man this may be
retarded by exposing the blood to nitrogen and/or by admin-
istering ascorbic acid to the mother before delivery, but even
so the rate of spontaneous haemolysis remains far above that
observed in adult blood.

REFERENCES

Abderhalden, E. (1898). Hoppe-Seyl. Z., 25, 65.

BoRUN, E. R., FiGUEROA, W. G., and Perry, S. M. (1957). J. din.

Invest., 36, 676.
Conway, E. J. (1957). Nature, Lond., 180, 1017.
Evans, J. V. (1954). Nature, Lond., 174, 931.
Evans, J. V. (1957). J. Physiol., 136, 41.
Evans, J. V., Harris, H., and Warren, F. L. Unpublished, referred to

by Evans and Phillipson (1957).
Evans, J. V., and King, J. W. B. (1955). Nature, Lond., 176, 171.
Evans, J. V., King, J. W. B., Cohen, B. L., Harris, H., and Warren,

F. L. (1956). Nature, Lond., 178, 849.
Evans, J. V., and Mounib, M. S. (1957). J. agric. Sci., 48, 433.
Evans, J. V., and Phillipson, A. T. (1957). J. Physiol., 139, 87.
Green, H. H., and Macaskill, E. H. (1928). J. agric. Sci., 18, 384.
Hallman, N., and Karvonen, M. J. (1949). Ann. Med. exp. Fenn., 27,

221.
Hallman, N., Osterlund, K., and Vara, P. (1954). Ann. Chir. Gyn.

Fenn., 43, 211.
Harris, H., and Warren, F. L. (1955). Biochem. J., 60, xxix.
Karvonen, M. J. (1949). In Haemoglobin, ed. Houghton, F. J. W., and

Kendrew, J. C., p. 29. London: Butterworth.
Karvonen, M-Jm and Leppanen,V. (1952). Ann. Med. exp. Fenn. ,30,14.
Kerr, S. E. (1937). J. hiol. Chem., 117, 227.
Maizels, M. (1936). Biochem. J., 30, 821.

McCance, R. a., and Widdowson, E. M. (1956). Clin. Sci., 15, 409.
Osterlund, K. (1955). Ann. Paediat. Fenn., 1, Suppl. 4.
Raiha, N. (1956). Acta paediat., Uppsala, 45, 176.
Selwyn, J. G., and Dacie, J. V. (1954). Blood, 9, 414.
Simon, E. R., and Topper, Y. J. (1957). Nature, Lond., 180, 1211.
WiDDAS, W. F. (1954). J. Physiol., 125, 18.
WiDDAS, W. F. (1955). J. Physiol, 127, 318.
Wise, G. H., Caldwell, M. J., Parrish, D. B., Flipse, R. J., and

Hughes, J. S. (1947). J. Dairy Sci., 30, 983.

206 Discussion

DISCUSSION

Davson : The most interesting thing here is the finding that these red
cells have a very high sodium concentration and alow potassium one. One
thinks of it at first as a primitive feature. On the other hand, when one
looks through the animal species in which it happens, it is most promi-
nent in the cat and dog, whilst the guinea pig, which we think of as a
rather primitive animal, has a very high potassium just like man. So it
has nothing to do with that. Then you also think of it as a failure to
develop a potassium-accumulation mechanism. There again, it is prob-
ably not to be considered as a failure at all. The red cell is derived from
a very highly developed nucleated cell and most likely the erythroblast
has the ability to concentrate potassium. Then when the cell becomes a
reticulocyte or an erythrocyte it loses the power of accumulation of
potassium. This 'loss' could be a development in the interests of eco-
nomy, because much less energy is required to maintain a cell with a low
concentration of potassium than with a high one, and the erythrocyte
has only an anaerobic source of metabolism.

Your results with the foetal cells are interesting, Dr. Karvonen. With
sheep, you find that the foetal erythrocytes have the high potassium and
it looks as if as they develop they lose the power of accumulating potas-
sium. But then, with the other species, we get the reverse. I think a lot
more work on the spontaneous haemolysis is necessary. Haemolysis
usually has a very definite cause and is usually due to the fact that the
permeability of the membrane becomes abnormally high and you get
this Donnan difference of osmotic pressure being exerted between the
plasma and the contents of the cell. Therefore, the most profitable line
of research would be to try and get conditions in which you could prevent
this haemolysis from occurring.

Milne : Is there any data available on the foetal levels in the cat and
the dog, which have this very high sodium content in the erythrocytes?

Karvonen: No, we have none.

Davson : The sheep can have as high a sodium content as the cat, yet
there is very little difference between foetal sheep. Would it be possible
to get a nucleated stage in the erythrocyte of the mammal and study its
potassium content? It could almost be done histochemically, just to get
a qualitative idea of the content.

Karvonen: That would be a very interesting thing to do.

Fourman : Tosteson reported a low erythrocyte potassium in sickle-cell
anaemia (1953. J. din. Invest., 32, 608). That confirms your view that
the level of potassium in the blood may be related to abnormal haemo-
globins ; is there any information on that, outside man?

Karvonen: In sheep, the type of haemoglobin affects the absolute
level of electrolytes within the same group. If you have sheep with low
potassium-containing red cells, and one of the animals has a different
type of haemoglobin, the electrolyte level in its red cells is also slightly
different.

Desaulles: Has not the same effect been described for some kind of

Discussion 207

deer? Deer may have sickle cells, and this is correlated with a certain
type of different haemoglobin.

Davson: I know the camel has ellipsoidal cells.

Bull : Is anything known about the relative efficiency of the different
kinds of red cells with respect to their function of carrying oxygen, in
relation to pH changes, carbon dioxide changes, etc.

Karvonen : Quite a lot is known about species and foetal-adult differ-
ences, but nobody has studied these aspects within one species, and at
the same time paid attention to the intra-species variations in intra-
cellular electrolytes.

Hingerty : It seems from the last three papers that there may be some
sort of late development of function as regards sodium and potassium
control. It may be something, according to Dr. Desaulles's work, that
develops in the rat at about 5-6 weeks, or something that increases the
efficiency of sodium — potassium exchanges across the cell membranes,
or the reabsorption rates in the renal tubules. During our potassium-
depletion experiments we found that in young rats up to six weeks of age
we could replace about 25 per cent of the muscle potassium by sodium on
potassium-deficient diets (Conway, E. J., and Hingerty, D. J. (1948).
Biochem. J., 42, 372). When we repeated the experiment we happened to
use rats of about nine or ten weeks old, and we found that the exchange
rates were much lower. Probably a greater efficiency develops in the
interval ; either the cell holds on to the potassium more efficiently or the
sodium pump works more efficiently. Possibly these changes are gradu-
ally developing in the growing animals and their responses to hormones
may also develop gradually.

Shock : One of the problems that occurred to us was whether the eryth-
rocytes that are formed in the normal course of turnover in the very old
individual can act as effectively as those in the young individuals. We
have not yet done the obvious experiment of producing a stress which
causes haematopoiesis, but we have examined the osmotic fragility of
red cells from individuals between the ages of 20 and 90, with about
ten individuals in each decade. With careful control of the pH, which
influences the fragility rather markedly, we found no striking evidence
of differences in the osmotic resistance of red cells taken from individuals
as old as 90, as compared with the young individual.

I also wonder whether there are subtle differences between the chemi-
cal structure of haemoglobin formed in an old individual as compared
with that in the young or middle-aged person. If the haemoglobin from
80-90-year-old individuals had been subjected to as detailed and careful
an analysis as that which resulted in the identification of the different
types of haemoglobin in the foetus, perhaps we would have found that
differences appear after a lifetime of utilizing the mechanism for making
haemoglobin.

McCance : Dr. Davson, can you comment upon the genetic side of this?
You spoke about the sodium pump ; what about the difference in haemo-
globin?

Davson: I cannot relate this at all. I do not see why a given type of
haemoglobin should be associated with a given electrolyte content.

208 Discussion

Black : On the genetic side it is very odd that one gets this scatter right
along the Une. One would think that, according to Mendel, one would get
segregation at the two ends of the line.

I was not clear whether there was an excess of fluid in the red cells.
In other words, in the foetal sheep or man was there an excess of
potassium per litre of red cells? Was there any difference in phosphate
content? Differences in phosphate content have been described, I
think, by Prankerd (1955, Clin. Sci., 14, 381) and others in connexion
with the sickle cell problem, and I wondered whether that side had
been gone into with foetal versus grown-up sheep.

Karvonen: I am afraid I gave a wrong impression when I said the
scatter was all along that line. There is a very clear concentration at
each end of the line but there is also a group in between. Within each
group, however, there is quite a considerable scatter which is due to a
permanent, individual characteristic of each sheep. The statisticians say
that there is quite a high intra-individual correlation.

The foetal cells contain more water than the adult cells. In sheep I
do not think that any determinations of the phosphate have been done,
but in man and in pig it has been found (McCance and Widdowson,
1956) that the phosphate of the foetal cells is higher.

Davson : It must be realized that when red blood cells are analysed,
very large numbers are used; there may well be differences in concen-
trations of potassium and sodium amongst the individual ones, and they
could well fall into groups which would never be discovered. Variations
in the Na/K ratio could be reflections of variations in the proportions of
high potassium and low potassium cells, which would give a continuous
scatter right along the line.

THE DEVELOPMENT OF ACID-BASE CONTROL

E. M. WiDDOWSON and R. A. McCance

Medical Research Council, Department of Experimental Medicine,
University of Cambridge

General Principles (as they apply to adults)

When the body of a healthy person is provided with the
diet normally eaten in Europe and the United States, it
produces in its metabolism more non-volatile anions than
cations. These "surplus anions" are excreted by the kidney
partly in combination with titratable hydrogen ions (the
titratable acidity) and partly as ammonium salts. The
ammonium salts usually account for rather more than 50 per
cent of the total. If the excess of non-volatile anions increases,
the pH of the urine falls and the titratable acidity increases,
but the excretion of ammonia also increases because a fall in
the pH of the urine is one of the things which raises the output
of ammonia ; and consequently the percentage of the surplus
anions excreted as ammonium salts remains about the same.
The excretion of ammonium salts is also increased (a) if the
pH of the urine is maintained at its lower limits for some
time by the continuous administration of acid or acid-forming
drugs. This is thought to be due to an increase in the activity
of the enzymes in the kidney which catalyse the formation of
ammonia and particularly of glutaminase (Davies and
Yudkin, 1952). (b) By an increase in the acid "load" (Rector,
Seldin and Copenhaver, 1955). Both (a) and (b) increase the
percentage of the surplus anions excreted as ammonium
salts, and good examples of the effects which may be observed
after continuous high dosage are given by Ryberg (1948).
As the pH of the urine rises progressively above 6-5 the
percentage of the total output of surplus non-volatile anions
excreted as ammonium salts may also rise and ultimately

209

210 E. M. WiDDOwsoN AND R. A. McCance

reach 100, because above pH 6-5 the excretion of titratable
acid falls more rapidly than the ammonia and is extinguished
before the excretion of ammonia, which continues at a de-
creasing rate up to pH 8. This tendency of the percentage to
rise as the pH of the urine goes above 6 • 5 is therefore exagger-
ated if the urines are titrated, as they mostly are nowadays, to
pH 7-4 instead of, as at one time, pH 8.

Dihydrogen orthophosphates are the main buffer acids
which can be titrated in a normal adult's urine, but this may
not be so in disease if there is a great excess of abnormal
organic acids of the right buffer strength in the urine, such as
(3 -hydroxy butyric acid or amino acids. Apart from the
phosphates and weak organic acids which contribute by their
presence to the titratable acidity, the surplus of non-volatile
anions in the urine is very largely due to sulphates, derived
from the metabolism of protein (Hunt, 1956). Chlorides are
generally balanced by the equivalent amount of fixed base
unless calcium or ammonium chloride has been taken to
produce a chloride acidosis.

The ability of the kidney to excrete hydrogen ions into the
tubules, and so to excrete the surplus non-volatile anions in
the way described, depends upon the activity of carbonic
anhydrase. Since it has been shown experimentally that the
degree to which the pH of the urine can be lowered depends
upon the activity of the carbonic anhydrase at any given time,
it may be that the lower and well-known limit of urinary pH
attainable by a normal person is an expression of the activity
of his carbonic anhydrase, but this is merely a suggestion at
the moment.

The New-born Period and Later Infancy

Complete collections of urine from three healthy baby boys
have been made for the first 48 hours of their lives, and again
over the whole of the 7th-8th day. These babies all passed
urine at the moment of birth and this was also collected.
Urine passed by two other babies at birth has also been
included in the series, and a 24-hour collection has been made

Development of Acid-Base Control 211

on four additional babies on the 7th to 8th day. Of the seven
babies investigated one week after birth, six were breastfed
and the seventh was fed on Ostermilk. Samples of blood
have been taken from the cord at birth, and from the femoral
vein at 48 hours and seven days. Urine has also been collected
for 24 hours from one child aged eight months and from one
aged one year, while six normal men and women have pro-
vided 24-hour urine collections to serve as the adult com-
parisons. The urines were collected and stored under toluene.
Determinations of pH, titratable acid, ammonia, creatinine,
phosphate, citrate and sulphate have been made on the urine,
and the sera have been analysed for creatinine, CO 2, chloride,
sodium and potassium.

The excretion of surplus anions

Fig. 1 shows the millimoles of surplus anions not combined
with fixed base (i.e. titratable acid plus ammonium salts)
excreted by the infants on the first, second and seventh days
of life and by the older infants. A figure for the adults is
indicated also. All the values are expressed per kg. of body
weight per day. The average pH of the adult urine was 6 or
a little over, while that of the babies was between 5 • 5 and 5 • 8,
and this has to be taken into account in considering some of
the results. The urine passed in the first and second 24 hours of
life contained less surplus anions per kg. of body weight than
that of the adults although the pH of the urine was lower,
which would have led one to expect a higher rather than a
lower anion excretion. This low rate of excretion was quite
sufficient to maintain the acid-base balance of the body, for
the serum CO 2 and chloride did not change. It is to be
attributed to the fact that the urine contains very little
phosphate or sulphate at this period (McCance and von Finck,
1947, and see later), owing to the small breakdown of tissue
protein (McCance and Strangeways, 1954). By the seventh
day the babies were taking nearly 500 ml. of breast milk a
day, which contained 9-5 g. protein or about 3 g./kg., and
they were passing about three times as much urine per kg. of

212

E. M. WiDDOWsoN AND R. A. McCance

body weight as the adults. Their excretion of surplus anions,
sulphates among them, per kg., had reached the adult level
although they were still excreting Httle or no phosphate.
The pH of their urine was a little higher than it was on the
first two days, and the increased volume may have been one
reason for this (McCance and von Finck, 1947; Hungerland,
1957).

At eight months to one year of age the babies excreted more

m-mole/kg./24h.
1.6.

14.

I -2.

o a"

0-6-

04-

02'

■

- ' I Adult

O- 24-7- I
24h. 48h. 8 day year

Fig. 1. Surplus anions (not combined

with fixed base) excreted by babies

during the first week and at 8 months

to 1 year of hfe.

surplus anions per kg. of body weight than the adults. This is
explainable by the high intake and metabohsm of protein per
kg. of body weight at this time of life. A child of one year
consumes about 3-5 g. protein per kg., which is two to three
times as much as an adult per kg., and only 8 or 10 per cent of it
is used for growth in contrast to the 50 per cent or so retained
in the neonatal period. The phosphates and the cystine and
methionine in the milk and other protein foods were probably
the main sources of the surplus anions.

Development of Acid-Base Control

213

Fig. 2 shows the percentage of the surplus anions excreted
with ammonia. For this it is possible to give a figure for urine
which was formed in utero and passed at the moment of birth
and which had a pH of over 6. It will be observed that,
although the pH of the urine passed at birth was higher than
that of the urine passed afterwards, the percentage of the
surplus anions excreted with ammonia was also very high
before birth, and of the order to be expected in adults with

80
70
60
SO
40
30
20
lO

•Adult

Before O- 24- 7- I
birth 24 K 48h. 8 day year

Fig. 2. Percentage of the surplus
anions excreted as ammonium salts.

very acid urines after taking large doses of ammonium chloride
for some days. The percentage of the surplus anions excreted
with ammonia in the first 48 hours and on the seventh day of
life has also tended in our series to be higher than that in the
urine passed by adults. This is probably because the babies'
urine contained so little phosphate, and consequently the
titratable acidity was low in relation to the total amount of
surplus anions to be excreted. It was not because the ability
of the newborn kidney to produce ammonia was greater than
that of an adult, for all the evidence is against this. Work

214

E. M. WiDDOwsoN AND R. A. McCance

which has been done on kidney shoes in vitro (Robinson, 1954),
and on renal glutaminase and ammonia production (Hines
and McCance, 1954) goes to show that, weight for weight, the
kidney of the newborn of other species contains less glutamin-
ase and produces less ammonia than that of the adult. Fig. 3
shows that the total amount of ammonia excreted per kg. of
body weight was in fact small in the first two days, but that
by a week, when the baby was taking in three times as much
protein as the adult per kg. of body weight, it had risen above
the adult level. The ability to form ammonia in response to
an acid load in the first day or two of life has not yet been

m-inole/kg./24h.
0-8.

0-6

O 4?

Adult

O- 24- 7- I
24h 48h. 8day year

Fig. 3. The amount of ammonia
excreted.

studied in man, but Cort and McCance (1954) found it to be
smaller in puppies two days old than in adult dogs. The
matter requires further investigation.

Fig. 4 shows the excretion of ammonia in millimoles/24
hours divided by the glomerular filtration rate (as measured
by the endogenous creatinine clearance) in ml. /minute. It
was possible to calculate this ratio for the urine passed at
birth, even though the rate of urine flow before birth was not
known, because the two functions being compared are both
expressed in terms of rates of urine secretion. The excretion
of ammonia was high in utero and in the newborn period in
relation to glomerular filtration rate. The glomerular filtration
rate at this time of life is very low by adult standards, and if
the endogenous creatinine clearance is a true measure of it,

Development of Acid -Base Control

215

it is evidently lower even than the excretion of ammonia.
By one year of age the glomerular filtration rate/kg. had risen
above that of adults (McCance and Widdowson, 1952), and
more ammonia and surplus anions per kg. were being ex-
creted (see Figs. 1 and 3); the amount of ammonia excreted
per ml. of glomerular filtrate was near the adult level.

0-7

0-6

0-5

t O 3

0-2

Adult

Before O- 24- 7- I
birth 24h. 48h. Sday year

Fig. 4. The ratio of the ammonia excreted

(m-mole/24 h.) to the glomerular filtration

rate (ml./min.).

The nature of the titratable acidity

Fig. 5 shows the excretion of titratable acid per kg. of body
weight by the babies and the adults. The amount excreted
was low during the whole of the first week, but it was rising
even though the urine still contained no phosphates. The
high excretion at a year is again related to the high intake of
protein at that age.

Fig. 6 shows the percentage of the titratable acidity due to
phosphate in the urine of an adult and in the urine of a breast-
fed baby in the first week of life. In the adult the percentage
depends upon the pH and, since the pH of the urine passed

216

E. M. WiDDOwsoN AND R. A. McCance

by the present series of adults was higher than that of the
newborn infants, the value for adults shown in Fig. 6 (70-80
per cent) has been taken from Gamble (1942). Phosphates
accounted for a very small fraction of the titratable acidity of

m-inole/kg./24 h.
lO •

0-8
06
0-4
02
O

--Adult

O- 24-7- I
24hL 48h. 8 day year

Fig. 5. The excretion of titratable
acid.

the infant's urine, which is due to the fact, already mentioned,
that the urine of breastfed infants contains so little phosphate
at this time of life.

Investigations are being made on the organic acids in the

Adult

^^^^^^

Infant

Phosphate

Organic acids

j Phosphate

Organic acids

lO 20 30 40 50 60 70 80
Percentage of titratable acidity

90 ICO

Fig. 6. The proportion of titratable acid due to phosphates and
organic acids in the urine (pH 5 • 5-6 • 0) of adults and infants.

urine during the first week of life. Citric acid is one of the
major constituents, and on the seventh day the breastfed
babies were found to be excreting 33 mg. citrate/kg. body
weight/24 hours (Stanier, personal communication). This

Development of Acid-Base Control 217

is more than the amount excreted by the adults in this series.
In so far as citric acid may be regarded as a product of the
metaboHsm of the kidney it cannot be classed as a surplus
anion although it contributes to the titratable acidity.

It is well known that infants on cows' milk mixtures have a
higher concentration of inorganic phosphorus in their serum
than breastfed infants; they excrete phosphates by the
seventh day of life, and the phosphate-organic acid relation-
ship is of the adult pattern, as it is also in the urine of infants
eight months to one year of age.

Foetal Life

In the uterus the acid-base balance of the whole conceptus
is regulated ultimately by the mother's lungs and kidneys,
but the foetal kidneys, membranes and placenta act as inter-
mediaries.

Urine has been taken from the bladders of five human
foetuses aged 10-20 weeks. It has always been found to be
hypotonic, due mainly to very low concentrations of sodium
and chloride. It appears to resemble the urine formed in
utero and passed at term which has been better investigated
and described elsewhere (McCance and Widdowson, 1953;
Hanon, Coquoin-Carnot and Pignard, 1955, 1957).

The pig has a gestation period of about 120 days. Between
the 20th and 60th day there is a rapid expansion in the volume
of allantoic fluid. The sac containing the fluid has free con-
nexion with the kidney through the urachus and bladder.
Its membranes also participate in exchanges with the mother.
Table I shows the composition of the fluid at 20 days, 45 days
and 60 days. At 45 days both mesonephros and metanephros
are functional, but the former is becoming less so. The
volume of fluid in the sac is very variable (Wislocki, 1935), but
it far exceeds the weight of the foetus. The osmolar concen-
tration falls greatly so that from 45 days it is only one-half
or one-third that of foetal serum (McCance and Dickerson,
1957). This fall in osmolar concentration is due largely to
a fall in the concentration of sodium and chloride. The

218 E. M. WiDDOWsoN AND R. A. McCance

concentration of potassium does not fall in the same way, and
the concentration of calcium rises. This calcium appears to
be held in solution by citric acid (Economou-Mavrou and
McCance, 1958).

The fluid at 45 days has been found to have a pH between
5-5 and 6, and a titratable acidity of about 10 m-equiv./litre.
The fluid contains ammonia, and ammonia appears to ac-
count for about 25 per cent of the titratable acid plus ammonia
found in it. The concentration of phosphates is always small,
and the acidity is almost entirely due to carbonic acid. The

Table I

The weight of the foetal pig and the volume and

COMPOSITION OF

ITS

allantoic

FLUID

Foetal age

20 days

45 days

60 days

Weight of foetus

01 g.

20 g.

100 g.

Volume of allantoic fluid

5 ml.

110 ml.

350 ml.

Composition of allantoic fluid

Osmolar concentration m-osm./l.

256

120

Urea m-mole/1.

8-4

10-3

Chloride m-equiv./l.

Sodium „ „

114

Potassium ,, ,,

Calcium mg./lOO ml.

—

Inorganic phosphorus mg./lOO ml.

—

pH rises quickly if the fluid is shaken or even if it is left in a
tube exposed to the air, and it was found necessary to collect
and analyse the fluid out of contact with air. Lutwak-Mann
and Laser (1954) found no "bicarbonate" in pig's allantoic
fluid at 20 days' gestation, but there is no doubt about the
presence of carbonic acid at 45 days.

Further investigation has confirmed the fact, first noted by
Lutwak-Mann (1955), that the chorioallantoic membrane
contains carbonic anhydrase. At 45 days the allantoic fluid
itself also had some carbonic anhydrase activity. On the
basis of material from three pregnant pigs the activities of
carbonic anhydrase may be given as foetal kidney + + +,
chorioallantoic membrane ++, allantoic fluid +• It is

Development of Acid-Base Control 219

hoped to extend the study to later stages of gestation, to
other membranes and to glutaminase.

It is an open question at present whether the fluid found in
the sac at 45 days is a hypotonic urine elaborated by the foetal
kidney and similar to that formed by the human kidney
before birth, or whether the fluid has been made hypotonic
and acid by the action of the membranes themselves. Small
amounts of fluid have been removed from the bladders of
foetuses at 45 days and it is probably possible also to with-
draw fluid from the large mesonephric duct, so that this
problem may be soluble without recourse to large-scale
experimental veterinary obstetrics. Should it turn out that
the composition of the fluid is being altered by the membranes,
their activities may contribute materially to our ideas about
the function of the renal tubules.

REFERENCES

CoRT, J. H., and McCance, R. A. (1954). J. Physiol., 124, 358.

Davies, B. M. a., and Yudkin, J. (1952). Biochem. J., 52, 407.

Economou-Mavrou, C, and McCance, R. A. (1958). Biochem. J. ,68, 573.

Gamble, J. L. (1942). Chemical Anatomy, Physiology and Pathology of
Extracellular fluid. 4th ed. Boston: Harvard Medical School.

Hanon, F., Coquoin-Carnot, M., and Pignard, P. (1955). Bull.
Acad. nat. med., 139, 272.

Hanon, F., Coquoin-Carnot, M., and Pignard, P. (1957). Et. neo-
natal., 6, 97.

HiNES, B. E., and McCance, R. A. (1954). J. Physiol, 124, 8.

Hungerland, H. (1957). Ann. Paediat. Fenn., 3, 384.

Hunt, J. N. (1956). Clin. Sci., 15, 119.

Lutwak-Mann, C. (1955). J. Endocrin., 13, 26.

Lutwak-Mann, C, and Laser, H. (1954). Nature, Loud., 173, 268.

McCance, R. A., and Dickerson, J. W. T. (1957). J. Embryol. exp.
Morph., 5, 43.

McCance, R. A., and Finck, M. A. von (1947). Arch.Dis. Childh.,22, 200.

McCance, R. A., and Strangeways, W. M. B. (1954). Brit. J. Nutr., 8, 21.

McCance, R. A., and Widdowson, E. M. (1952). Lancet, 263, 860.

McCance, R. A., and Widdowson, E. M. (1953). Proc. roy. Soc, 141 B,
488.

Rector, F. C, Seldin, D. W., and Copenhaver, J. H. (1955). J. din.
Invest., 34, 20.

Robinson, J. R. (1954). J. Physiol., 124, 1.

Ryberg, C. (1948). Acta physiol. scand., 15, 114.

WiSLOCKi, G. B. (1935). Anat. Rec, 63, 183.

220

Discussion

DISCUSSION

Zweymiiller: The identification of organic acids in urine by paper
chromatography is elegant and of general application. The Rp values
of the different organic acids are distinctly different and therefore a clear
separation on the paper is possible. We used the technique developed by
Nordmann and co-workers (1954. C.R. Acad. Sci., Paris, 238, 2459),
and Fig. 1 demonstrates the position on a two-dimensional descending
chromatogram of some non- volatile, water-soluble organic acids which

05
Et OH-NH3-H2O

Fig. 1 (Zweymiiller). The position of some organic
acids in the urine of a normal adult on a two-dimen-
sional descending chromatogram.

Ci = Citric acid, Ta = Tartaric, Ma = Malic,

Gly = Glycolic, a-ce = a-ketoglutaric, Su = Succinic,

Ac ^Aconitic, Glu = Glutaric, p-hy = [B-hydroxy-

butyric, La = Lactic, Hi = Hippuric.

Nordmann has found in the urine of normal adults. There is clear separa-
tion of citric acid, tartaric, malic, a-ketoglutaric, succinic, aconitic, lactic,
glycolic, hippuric, glutaric and p-hydroxybutyric acids. One spot applies
to both sulphate and phosphate, if there is any phosphate in the urine.
Using this method the organic acids give yellow spots on a blue-greenish
background. These spots have the advantage that they do not fade but
get more intense with time. We have so far examined urines passed by
newborn babies on the first, second and seventh days of life, but we have
not done enough to give a complete answer yet. Citric acid appears to be

Discussion 221

the major organic acid constituent of the urine which is passed im-
mediately after birth. In addition, urine passed during the first 24
hours of hfe contains mahc acid, glycohc, lactic, ^-hydroxybutyric,
succinic, and a-ketoglutaric acids, but not aconitic acid. With this
method one can detect a minimum of 20 y.g. of each of these organic
acids.

Adolph : Is there any appreciable accumulation of organic acids in the
newborn during the first week of life? At this stage the individual is very
insensitive to the hydrogen-ion concentration changes as far as the
breathing is concerned, and I was wondering whether it is also insensitive
as far as excretion is concerned.

Zweymiiller : We are now working on the detection and identification
of the organic acids found in the urine of normal newborn babies, and the
next problem will be to identify those found in the urine of hypoxaemic
newborn babies.

Karvonen : Did you find any pyruvate or does it come out with this
method?

Zweymiiller : We have not found a pyruvic acid spot, but we have not
added pyruvic acid to the urine so we do not know exactly where the
spot should appear on the paper.

Karvonen : I understand that increased excretion of pyruvate has been
found during the first few days of life (Tallqvist, H. (1952). Thesis,
Hameenlinna).

Zweymiiller : There is an interesting paper about some work on the
output of organic acids in potassium depletion in which pyruvic acid,
lactic acid, a-ketoglutaric acid, and citric acid were estimated, but this
was done on normal adults (Evans et al. (1957). Clin. Sci., 16, 53).

Fourman : The hydrogen ion in the allantoic sac must come from some-
where, it cannot be manufactured. It must come in the end from the
mother and since she cannot manufacture the hydrogen ion it must
ultimately come from her diet. So what happens if you feed alkali to
the mother pig?

Widdowson : We have not tried that.

Milne: It is well shown in your paper. Dr. Widdowson, how the new-
born baby copes with its normal environment. I would agree that the
organic acid level, especially that of citrate, is proportionally much
higher than in the adult. Ob\4ously in assessing the efficiency of the
kidney, particularly in excreting an acid load, one must give it a maxi-
mum challenge and, though I see the difficulties of this in human experi-
mentation, it would be extremely interesting to do this in the newborn
animal. There seem to be two separate aspects of excretion of acid by the
kidney. One is the ability of the kidney to excrete a maximum amount of
hydrogen ion per day and clearly that can only be assessed by giving a
prolonged acid load. The other is the ability of the kidney to maintain
a hydrogen ion gradient between urine and plasma, in other words the
production of a minimum urinary pH. I would be very interested in
having data on whether the minimum pH of adult urine is similar to the
minimum pH of newborn urine, whether the ammonia excretion can
increase on prolonged acid ingestion proportionally to that of the adult,

222 Discussion

and finally whether this very large citrate output in the newborn shows
the same tremendous lability to acid-base effect as it does in the adult,
in whom it can be reduced by quite small doses of acid or increased by
alkalinization, say by sodium bicarbonate.

Widdowson: We have not yet given an acid load to newborn babies,
although we should like to do so, but the experiment has been done on
puppies. The question about citrate is one for the future. We have so far
only studied three babies and this investigation is by no means complete.

McCance: The puppies have only been studied with respect to acute
acidosis (Cort, J. H., and McCance, R. A. (1954). J. Physiol, 124, 358).
The difficulty in an animal which is developing very rapidly is to separate
the effects of several days' administration of an acid-forming drug and
the natural development of the animal at that age. In the acute experi-
ments the puppies were very defective in their ability to produce am-
monia and they did not make a good response at all. They remained much
more acid internally. We have unfortunately not yet tried the effect of
altering the pH of the urine upon the excretion of citrates in the newborn
baby.

Scribner : We carried out some experiments in rats which seem to indi-
cate that the amount of citrate in the urine depends on the kidney tissue
level of citrate rather than on the blood citrate level. After intra-
peritoneal injection of either sodium or potassium bicarbonate, urinary
citrate increases 10- to 20-fold in one to two hours. Kidney tissue citrate
increases two to threefold. Blood citrate rises 10 per cent at most. The
response to intraperitoneal injection of citrate is quite different. We
used ammonium citrate to get away from changes in acid-base balance,
due to, say, injection of citric acid on the one hand or sodium citrate on
the other. After the injection of 0-0035 m-mole/kg. ammonium citrate
the blood level rises nearly 100 per cent, but there is little or no increase
in either kidney tissue citrate or urinary citrate. We concluded from
these experiments that the level of citrate in the urine under these
conditions is determined by the citrate level in the renal tubular cell and
is independent of the amount of citrate filtered through the glomerulus.

McCance : Dr. Milne, what determines the lower limits of pH which
the human and other kidneys can achieve?

Milne: I think this can only be answered conditionally. First, one
must state the stimulus, and secondly one must state the conditions of
the kidney at that moment. Ammonium chloride has been used as the
usual stimulus and I think no-one has ever produced a pH of human
urine below 4 • 4 by that method, but other stimuli seem able to produce
a considerably lower pH. The experiments of Schwartz, Jenson and
Relman (1955. J. din. Invest., 34, 673) showed this, where they infused
sodium sulphate in a sodium-depleted individual. There, quite clearly,
they got down to a urinary pH of 4 0, so that is a more effective stimu-
lus, and indeed this agrees in the rat. It is very difficult to produce
a highly acid urine in rats by most experiments. When it is given
ammonium chloride the rat seems to be able to keep up with the am-
monium intake and puts out ammonium chloride in its urine almost as
quickly as it is either injected or taken in the drinking water. But an

Discussion

223

acid urine in the rat can be produced by the same technique of sodium
depletion and intraperitoneal sodium sulphate, which is clearly a more
efficient stimulus to maximum acidity. Finally, one would agree that
the condition of the kidney has been shown quite clearly to be dependent
partly on body potassium stores. Potassium depletion, possibly by
decreasing intracellular high-energy phosphate bonds — though that is
purely speculation — will decrease the maximum osmolar gradient
between urine and plasma, and similarly it will decrease the maximum
possible hydrogen ion gradient. This effect is produced by potassium
deficiency on the two stimuli of ammonium chloride or sodium sulphate
injections.

McCance : The lower limits might be due to the activitj^ of carbonic
anhydrase having a ceiling in the human kidney. We know that if the
carbonic anhydrase is defective the lowering of pH is correspondingly
limited.

GENERAL DISCUSSION

Wallace: I should like to present a problem that arises when one
attempts to interpret chemical analysis of tissues from deficient
animals in terms of histological appearance. Skeletal muscle taken
from potassium-deficient animals is low in potassium, high in sodium,
high in its content of basic amino acids and probably low in bicar-
bonate content. When the muscle is examined histologically one
sees apparently normal cells lying side by side with grossly abnormal
cells. Which cells account for the chemical abnormalities? I have
wondered if a cell can tolerate any deficit at all. Possibly, for the cell,
it is an all-or-none phenomenon. Does a tissue as a whole become
deficient in a sort of quantum fashion, cell by cell rather than by an
over-all shared process by all of the cells? Is it not necessary to get
down to a truly cellular level to further our understanding?

Fourman: May I add to Prof. Wallace's problem? The kidney and
the heart show the morphological changes of potassium deficiency
before the other tissues. These two tissues, when they are analysed in
animals that have been made deficient in potassium, do not as a rule
show chemical evidence of potassium deficiency. I suppose they do if
you carry the deficiency far enough but as a rule they do not. It has
always been a puzzle to me why two tissues that have a normal
potassium content are the first tissues to show a potassium abnor-
mality. These two tissues are also ones that never rest, in the way
muscles do, and one wonders whether the fact that their function
requires the maintenance of a normal potassium content, with the
demand on the metabolic energy of the cell that this entails, carries
the seeds of their own destruction.

Wallace: The analyses of Orent-Keiles and McCoUum do show
deficits of potassium in cardiac muscle taken from deficient rats (1941.
J. hiol. Chem., 140, 337). However, most workers have not shown
the same thing.

Black: Jean Oliver and co-workers (1957. J. exp. Med., 106, 563)
have done work on the localization of the morphological defect in
the nephron of potassium-depleted animals, and this seems to be
limited to the proximal and the collecting tubules. Dr. Fourman's
difficulty may not be so real if the lesion is as sharply localized as
that. With analysis of the whole kidney that may just be a failure
to detect a limited local deficiency of potassium.

Milne: Part of the difficulty may be this: is not the necrosis or
degeneration in the cell possibly due to the fall in intracellular pH,
not primarily to potassium deficiency? I agree that kidney analyses

224

General Discussion 225

have not shown a potassium deficiency as in muscle, but they have
shown a fall in intracellular bicarbonate, and therefore presumably
a fall in intracellular pH. These experiments have not, as far as I
know, been done with the heart muscle, but by analogy one would
predict that the same situation may occur: the fall in intracellular
potassium may be small, but the fall in intracellular bicarbonate
and intracellular pH may be comparable to that in the kidney, and
possibly greater.

Fourman: Yes, unless you think as I did, that the fall in intra-
cellular pH is a result of the fall in intracellular potassium.

Milne: Direct analysis of tissue does not appear to support that.

Shock: The histological structure in Prof. Wallace's potassium-
deficient animal, which I presume was a young one, is quite similar to
the sections of muscle tissue from the old animals that Dr. Andrew
has prepared from our material, which show a reduction in potas-
sium content of the total muscle mass. There were fewer nice-
looking muscle fibres in Prof. Wallace's animal than we see in the
sections from the older animals, but there is a striking similarity
in that there are good-looking areas, as described by the pathologist,
with a lot of other material around them. I recall that a few years
ago there was quite a flurry about the electron microscopic studies
of mitochondria. In such pictures the mitochondria from cells of
old animals were presumed to look frayed and woebegone. Sub-
sequent experiments showed that dietary deficiencies and alterations
could produce similar changes in the mitochondria taken from cells
of young animals. If the few cellular changes we can observe in
older animals can be produced by nutritional and dietary alterations
in the young ones, it is possible that these 'age changes' are the
result of chronic malnutrition of the cells. This brings us to the
basic questions of what is adequate nutrition of a cell, and how can
it be maintained.

Wallace: What is old and what is young? To me a 30-day-old rat is
quite young, while to Dr. Widdowson it is as old as Methuselah.

Shock: To me a 10-12-month-old rat is a husky young adult, and
when I talk of an old animal I mean one that is 24 months old or at
least is at an age when 50 per cent of his contemporaries are dead.

Wallace: Young rats made potassium-deficient do show morpho-
logical changes in skeletal muscle. These changes can be almost
completely reversed in as little as 36-48 hours after potassium
administration. The lesions in cardiac muscle do not show this
rapid type of healing. It would be interesting to see if your old rats
have a slower repair time. Dr. Hingerty has already mentioned
that older rats chemically repair potassium deficiency more slowly
than do the young ones.

AGEING — IV — 8

226 General Discussion

Kennedy: Morrison and Gordon (1957. Fed. Proc, 16, 366, and
personal communication) have reported that a 24-month-old rat
starved of food but not water for 24 hours loses far more urea,
creatinine and potassium than a young one of comparable weight.
So there is a state of incipient potassium deficiency. We have also
found that the adrenals are usually pretty large in these old rats.

Shock: We have a done good many metabolic balance studies on
the human (Duncan et al. (1951). J. din. Invest., 30, 908; Duncan et
al, (1952). J. Geront, 7, 351 ; Bogdonoff ^/ al, (1953; 1954). J. Geront.,
8, 272; 9, 262; Watkin et al, (1955). J. Geront, 10, 268). We consis-
tently found that the older individuals, when given good protein in-
takes that resulted in positive nitrogen balances, retained potassium
in excess of the theoretical amount required for the nitrogen retained.
A good deal of this, I am sure, may be due to cumulative analytical
errors, but it has always seemed to me that the older animal will
work himself into a potassium deficiency if given the opportunity.

Black: Is not some of our difficulty here due to the limitations of
morphology? If we take as our criteria of morphological change the
fact that the tissue ' looks bad ' or ' looks moth-eaten ', then we are not
going to get anywhere in deciding the cause of this change. You can
hardly expect a cell to have a signpost saying ' I am too old ', or 'I
am potassium-deficient', and if we see the same change I do not see
how we can expect morphology to decide its aetiology.

Talbot: When you use the term 'potassium-deficient', Prof.
Wallace, do you wish us to think simultaneously about the correlated
fact of the cellular sodium excess? Cellular sodium intoxication may
actually be the provocative factor under some circumstances.

Wallace: Sodium excess is usually a corollary but not always. Some
cation, it would seem, must replace the deficit. Basic amino acids
have been shown to increase in potassium-deficient tissues as well as
sodium.

Talbot: We have just done some experiments where the absolute
losses of potassium due to starvation were greater per rat than some
of the losses incurred when feeding a zero potassium-normal sodium
intake. The animals which had lost this large amount of potassium
by simple depletion were asymptomatic ; it was only those that also
had cellular sodium intoxication that showed all the symptoms com-
monly considered characteristic of marked potassium deficiency.

Hingerty: Prof. Wallace, when you restored the potassium, morpho-
logically the tissue appeared perfectly all right in 36 hours. Did you
do the chemical analysis?

Wallace: Yes, we did, stimulated by your work (Conway, E. J.,
and Hingerty, D. J. (1948). Biochem. J., 42, 372). Unlike you we
found that sodium was lost simultaneously with a gain of muscle

General Discussion 227

potassium to normal (Schwartz, R., Cohen, J., and Wallace, W. M.
(1955). Amer. J. Physiol, 182, 39).

Swyer: The sex difference in these responses to various hormones,
and other matters which must either themselves have a hormonal
basis or must be genetically determined, still puzzle me. What is
the true sex basis? Is it a question of androgens and oestrogens, or
the ratio of these two sex hormones, or is it in fact due to some
characteristic which depends upon the presence of one X or two
X chromosomes?

Kennedy: It is probably something to do with species, but the
differences in size and growth between the castrate cockerel and the
castrate hen, and the same sort of thing in male and female castrate
rats, are very well known, and there is obviously a genetic difference
in the subsequent behaviour of the neonatal castrate. Some of the
early theories of ageing depended on body size, and one wonders how
much actual size, or organ development and growth as such, rather
than sex alone, affects the matter. The kidney of the male castrate
rat, even though it is castrated very young, is a much bigger organ
and in some senses, therefore, is a more developed or older organ than
that of a female rat. Purely structural factors may determine some of
the differences in what I think you call end-organ responsiveness.

Swyer: Is castration even shortly after birth early enough? After
all, the foetal testis has a very important role to play and intra-
uterine castration might avoid this difficulty.

Desaulles: That might possibly be helpful in determining the role
of the X zone. It is hard to imagine how the interrelationship be-
tween pituitary, adrenals and gonads acts just at the beginning of
life in the animal.

Milne: Is the control to the castrate male a spayed female?

Desaulles: They are quite different — that is the annoying point.

Kennedy: When he discussed renal function Dr. Shock pointed out
that there was some similarity between the old and the young kid-
neys in their inability to sustain water diuresis and so on. It has been
shown (Smith, H. (1951). The Kidney; Structure and Function in
Health and Disease. New York : Oxford University Press) that if you
take an animal of intermediate age and remove one of its kidneys and
half the other, then the initial response, at least, is a great diminution
in water diuresis, which may take four weeks to be restored to about
two- thirds normal. This may suggest that the period during which
the major changes in the newborn develop is during the unfolding of
the anlage of the kidney ; senescence in most animals that have been
studied similarly involves a loss of structural units. So again, simply
the amount of end organ which is there may be the important thing,
apart altogether from what is called the endocrine climate.

228 General Discussion

Adolph : I wish the structural picture agreed so well with the physio-
logical response to water loading. First of all, when you take out
one kidney from, say, a middle-aged rat, you do not reduce the water
diuresis much — it is more often a reduction of 20 per cent than of 50
per cent. Even if you take out a kidney and a half you still have 70
per cent of the response, and hypertrophy does not seem to be parti-
cularly important in restoring the response to near 100 per cent
(Adolph, E. F., and Parmington, S. L. (1948) Amer. J. Physiol., 155,
317). Similarly in the kidney of the newborn the number of nephrons
available, as far as anatomical studies show, is about 50 per cent of
that in the adult, and yet the diuresis may only be 10 per cent of the
adult's. As the diuresis develops in intensity with age, it gets far
ahead of the development of the number of nephrons or of any other
structure that has been counted in the kidneys. Enzyme studies
have been made to try and find something that would be parallel
either to the water diuresis or to the clearance increase with age. The
clearances in the newborn kidney, as far as they have been measured,
also develop rather slowly, but all of them are in parallel, at least in
the rat. This is not necessarily true in all species, because there seems
to be an exception in the rabbit (Levine and Levine. (1958). Amer.
J. Physiol., 193, 123). However, phenol red and inulin clearances are
proportional to one another at every age in the rat, while there is no
clear parallelism between any two properties of excretion except the
clearances.

Desaulles: When a heminephrectomized animal is submitted, eight
or ten days after operation, to a physiological saline load of about
20 ml. /kg., the output of urine during eight hours is much higher
than in an animal with two kidneys. It is not at all clear to me why
the output is so much higher; the dilution is greater, and less sodium
is given out.

Richet: It may be dependent on the amount of solutes per nephron.

Bull: I think Dr. Richet's suggestion is a likely one. The kidney
lesions in severe burns are probably due to a period of low circulatory
volume which damages certain nephrons in several different ways.
The resulting morphology may be very various but the functional
lesion is usually rather similar in producing an oliguria, with a failure
of concentration. This agrees best with the idea of fully functioning
surviving nephrons ; any nephrons that are damaged at all are right
out of the picture. This explanation also agrees with our finding that
in old patients there is a poorer response to water load and a slower
excretion of sodium.

THE ROLE OF THE KIDNEY IN ELECTROLYTE
AND WATER REGULATION IN THE AGED

N. W. Shock

Gerontology Branch, National Heart Institute, National Institutes

of Health, PHS, D.H.E. & W., Bethesda, and the Baltimore City

Hospitals, Baltimore, Maryland

The kidney is the first line of defence in maintaining appro-
priate concentrations of water and electrolytes in the internal
environment of all the cells in the body. Although there are
other avenues through which salts and water may be lost
from the body, and other factors which may enter into the
regulation of concentrations in local areas, it is the kidney
which carries the major burden of electrolyte and water
regulation. The kidney responds to a multitude of stimuli and
is blessed with large reserve capacities. It is the purpose of
this report to describe briefly some of our findings with
regard to age changes in renal function, to discuss the possible
mechanisms of these changes, and to discuss their relation to
the maintenance of certain physiological constants in the
aged.

In order for the kidney to serve its functions of regulating
water and electrolyte concentrations, as well as the volume of
extracellular fluid, blood must be delivered to it in adequate
amounts, glomerular filtrate must be formed, and the tubular
cells must selectively reabsorb and excrete substances in
accordance with a variety of stimuli to which the kidney must
respond. The application of clearance techniques makes it
possible to assess the nature of age changes in discrete renal
functions. The studies to be reported are based on ambulatory
male subjects between the ages of 20 and 90 years who were
found to be free from clinical evidence of renal disease as
judged by clinical laboratory tests and medical history. All

229

230

N. W. Shock

subjects were selected only after a thorough history and physi-
cal examination which excluded recent or remote renal
diseases, cerebrovascular accidents, coronary artery disease,
syphilitic or rheumatic heart disease, hypertension, or any
recent alterations in body weight. All tests were carried out
under basal conditions and subjects were hydrated with 600-
800 ml. water, given orally 1-2 hours before the test, and
200 ml. water were given at half-hour intervals during the

40 60 60

AGE-YEARS

Fig. 1. Change in standard diodrast clearance or effective renal

plasma flow with age. O O average values ml. plasma/min./

1-73 sq. m. body surface area.
(From: Shock, 1952).

test. The constant infusion method was followed, and four
clearance and four Tm periods of 10-14 minutes each were
taken according to the method of Smith, Goldring and Chasis
(1938). Fig. 1 shows the age change in effective renal plasma
flow as estimated from diodrast clearance (Shock, 1952).
Between the ages of 20 and 90 years there was a decline in the
effective renal plasma flow amounting to approximately 53
per cent. The regression equation relating the diodrast
clearance to age is: Clj^ = 840 — 6-44 X age (in years).

Age Changes in Renal Function

231

Although there is a substantial variation between subjects at
any given age, the trend is highly significant.*

The age decrement in glomerular filtration rate, as measured
by standard inulin clearance, is shown in Fig. 2. The regres-
sion of inulin clearance with age is expressed by the equation :
C1t„ = 153*2 — 0-96 X age (in years). The average decline

^In

over the age span 20-90 years was 46 per cent in this instance.|

•

1 1

—

150

•

- •

•
•

•

•
*

•<

•

•!•

•

• •

ex:

•

• ■

•

K,^^!*

•
•

•

o ^100

•

• •

•
•

"•

=> 2

—

•

• •

•

. • •

• •Ni

. : -

r^ \

•

g ^

• •

• -

^ ^50

—

■z.

• "

-//-J

1 1

•
1

AGE-YEARS
Fig. 2. Change in standard inulin clearance or glomerular filtration

rate with age. O O average values, ml. filtrate/min./l • 73 sq. m.

body surface area.

(From; Shock, 1952).

The fall in glomerular filtration rate is closely associated with
the fall in plasma flow so that the filtration fraction, calculated
as ratio of inulin clearance to the diodrast clearance, shows
only a slight increase with age (Fig. 3).

* In a different sample of subjects in whom renal plasma flow was estimated
from PAH (/?-aminohippuric acid) clearance (Watkin and Shock, 1955), the
regression equation was: CIpah = 820 — 6-75 X age (in years).

f In other groups of subjects the regression of inulin clearance on age was :
Clin = 157-0 — 1-16 X age (in years) (Watkin and Shock, 1955), and
Clin = 150-9 — 0-904, X age (Miller, McDonald and Shock, 1952).

232

N. W. Shock

o
o 20

t^ 10

•

• ^^^^. • • • . . • .

- • • —

40 50 60

AGE-YEARS

Fig. 3. Change in filtration fraction with age. O —
values, per cent of plasma filtered.

(From: Shock, 1952).

O average

70
^ 60

•-OJ
CO 3

^^50
0^40

e 2

1^30
z o

i^^

^ I ^ I I I p

50
AGE-YEARS

Fig. 4. Change in standard diodrast Tm with age. O-

•O average

values mg. diodrast iodine/min./l -73 sq. m. body surface area.
(From: Shock, 1952).

Age Changes in Renal Function

233

The maximum capacity of the renal tubule to excrete
diodrast also diminishes with age. Fig. 4 illustrates the
results of this test in the subjects studied. The average
diodrast Tm fell from 54-6 to 30-8 mg. iodine/1-73 m.^/min.
between the ages of 20 and 90 years. This represents a reduc-
tion of 43 • 5 per cent. The regression equation relating diodrast

400

u" 300-

O
O ro

200

100 —

20-29 30-39 40-49 50-59 60-69 70-79 80-89

AGE IN YEARS

Fig. 5. Decrease in maximal tubular reabsorptive capacity with
age. The slope is drawn to connect the mean values for each
decade. The vertical lines represent ± one standard error of the
mean, while the open circles define the limits of ± one standard
deviation of the distribution.

(From: Miller, McDonald and Shock, 1952).

Tm to age is: Tm^ = 66-7 — 0-40 x age (in years).* The
reabsorptive capacity of the renal tubular epithelium for
glucose also shows a comparable diminution with age, as
shown in Fig. 5. The average glucose Tm fell from 328 to
223 mg. glucose/1- 73m. 7min. between the ages of 30 and

* The maximum excretory capacity for PAH shows the following regression
on age: Tuipah = 120-6 - 0-865 X age (Watkin and Shock, 1955).

234 N. W. Shock

90 years. The regression equation is : Twlq = 432 • 8 — 2 • 604 X
age (in years). The maximum capacity for both a reabsorptive
and excretory mechanism in the renal tubules showed ap-
proximately the same percentage decrement with age.

The average inulin clearance per unit of Tm remains con-
stant between the ages of 20 and 90 years (Fig. 6). This
finding lends support to the hypothesis that a nephron loses
its function as a unit. In contrast, the diodrast clearance per

o
z
<

-J
o

T \ 1 \ \ 1 \ 1 \ \ 1 \ ^ 1 T

'• • • - ••_

J__J \ \ \ I \ I I I i I \ \ L

- 10 20 30 40 50 60 70 80 90

AGE YEARS

Fig. 6. Change in rate of glomerular filtration per unit of diodrast Tm.
O O average values.

(From: Shock, 1952).

unit of Tm decreases from an average value of 12-6 at age
30-39 to 9-7 at age 80-89 (Fig. 7). This steady decline in the
effective renal plasma flow per unit of tubular excretory
capacity indicates that the average amount of blood delivered
to each tubule, and by implication each nephron, declines with
age. Since we have been able to demonstrate a significant
reduction in resting cardiac output with age (Brandfonbrener,

Age Changes in Renal Function 235

Landowne and Shock, 1955), as shown in Fig. 8, a portion of
the reduction in renal plasma flow must be attributed to a re-
duction in total blood flow. However, in experiments to be
reported later calculations show that the age reduction in
renal blood flow is proportionally greater than the reduction
in cardiac output.

—

•

,•

—

•

• •

•

•o«^"^

^.^.

_ a

•

• -

•

• ,

•

•-oL^

•

—

•

•
•

• ••••

•

>\-

./^

)

—

U
O

•• \<

^y^^

"r •

•

—

•

••

•

—

AGE

IN YEARS

Fig. 7. Change in effective renal plasma flow per unit of diodrast Tm.
O O average values.

(From: Shock,. 1952).

Other experiments have shown that the reduction of
effective renal plasma flow in the aged cannot be ascribed
to permanent structural changes in the renal vascular bed
(McDonald, Solomon and Shock, 1951). Previous studies have
shown that the administration of a pyrogen to young people

236

N. W. Shock

results in a marked increase in effective renal plasma flow. In
order to assess age changes in the ability of the renal vascular
bed to dilate, glomerular filtration rate and renal plasma flow

K. 60

w 50

.. »•

50 60

Age years

Fig. 8. Stroke output per sq. m. surface area versus age. Each point

represents the average of two measurements in 49 subjects, of three

measurements in four subjects, and a single measurement in 14 subjects.

The hne indicates the simple linear regression for the data.

(From: Brandfonbrener, Landowne and Shock, 1955).

were measured in young, middle-aged, and old subjects
following the intravenous administration of 50,000,000
killed typhoid organisms (0 • 5 ml. typhoid-paratyphoid A and
B vaccine). The results of these experiments, based on the

Age Changes in Renal Function

237

average of 20 subjects in each age group, are shown in Fig. 9.
From the three curves at the bottom of the chart it is clear that
although the usual age difference in glomerular filtration rate
was present, there was no significant effect of the pyrogen in

,26

■ 1 1

1 1

1 1 1

§g 22
i^ .14

-^^^^

''*-=^=.:^-

•is

5 .10

n eoo-

z-^-^^"

;; 600-

< 400-
^ 200-

.-■or''

^,JO—

^a^

'^ — a — ^_

o -o -0--

^ l2Cr

z ^ 90

-^r^mr:

— •^^

^L-

.—t, A ^— •— fi _

"P 60

..^--<^-

..■O-

— 0-,— ^....^....^ _

— #

1 1

1^1 1

111'

120
MINUTES

160

200

Fig. 9. Changes in glomerular filtration rate (Cin),
effective renal plasma flow (Cpah)j and filtration
fraction during the pyrogen reaction. Fifty million
killed typhoid organisms were injected intravenously

at 0 time. O O mean values for 14 subjects

aged 70-85 years (O group). A A niean

values for 20 subjects aged 50-69 years (M group).

0 % mean values for 20 subjects aged 20-49

years (Y group).

(From: McDonald, Solomon and Shock, 1951).

either the young, middle, or old subjects. The three curves in
the centre of the graph show clearly that, beginning about 80
minutes after the administration of the pyrogen, there was a
slow continuous rise in renal plasma flow in all groups of sub-
jects. Although the mean absolute increases were greater for
the young than for the old group, where increments were

238 N. W. Shock

expressed as percentages of the base line values, the rise in
renal blood flow for the young, middle, and old groups was
76, 86, and 91 per cent respectively. As shown by the upper
three curves, the filtration fraction diminished markedly in
all subjects, indicating a fall in effective filtration pressure,
which would result from a greater vasodilatation at the
efferent than at the afferent side of the glomerulus if there
were no change in blood pressure. Actually, the diastolic
blood pressure dropped slightly in the middle and old groups,
but remained constant throughout the reaction in the young
group. At the height of the reaction the differences in the
filtration fraction, observed under resting conditions, com-
pletely disappeared. The small absolute changes in renal
plasma flow in the older subjects, following pyrogen, are
consistent with the anatomical findings of a progressive
decrease in the number of glomeruli in the aged kidney
(Moore, 1931). On the other hand, the time of onset and the
percentage increase in renal plasma flow were similar in the
different age groups. Consequently, it must be concluded
that the responsiveness to pyrogen of the vascular elements
remaining in the aged kidney is not qualitatively different
from that in the young kidney. It is inferred from these
experiments that the renal arterioles in the aged kidney are
capable of dilating, and that in the resting state there is a
functional vasoconstriction of the afferent arterioles in the
aged which, under resting conditions, diverts blood from the
kidney to other parts of the circulation.

To function effectively the kidney must respond to a
variety of stimuli. One of the most important signals for
altering the reabsorption of water by the renal tubule is the
antidiuretic hormone. Age differences in the inhibition of
water diuresis, following the intravenous administration of
small amounts of pitressin, have been observed (Miller and
Shock, 1953). In these experiments a maximum water
diuresis was established by the oral administration of 500 ml.
water at 6.00 a.m., followed by 250 ml. water at half-hour in-
tervals until completion of the test. To ensure maximum urine

Age Changes in Renal Function 239

flows, oral fluid intake was supplemented by the intravenous
administration of 5 per cent dextrose in distifled water, in
which appropriate quantities of inulin and sodium amino-
hippurate had been added at the rate of 8 ml./min. by a
constant infusion pump. Twenty-nine adult males, ranging
in age from 26 to 86 years, served as subjects. The total
sample was arbitrarily divided into three age groups: young
(no. = 9, age range from 26-45), middle (no. = 10, age range
from 46-65), and old (no. = 10, age range from 66-86). After
three control collection periods, 0-05 milliunits pitressin/kg.
body weight was administered intravenously. Subsequently,
six consecutive urine collections, each of 12 minutes duration,
were made. During the control periods, the average urine
flow for the young subjects was approximately 14 ml./min.;
middle-aged, 11 ml./min. and old subjects, 10 ml./min. The
urine/plasma (U/P) inulin ratio was calculated as an index
of water reabsorption. The results of this experiment are
shown in Fig. 10, where the U/P inulin ratio was plotted
against the urine collection period. During the control
periods, the U/P inulin ratios were approximately 10 for all
three age groups. Following the administration of pitressin,
prompt antidiuresis was noted in all three groups. Peak
antidiuresis and peak concentration of inulin were observed in
all three age groups during this period which was 12-24
minutes after pitressin. As indicated in Fig. 10, there was a
marked age difTerence in the antidiuretic response to this
standard stimulus. The young subjects showed the maxi-
mum response and the old subjects showed the minimum.
In Fig. 11, the relationship between the maximum observed
tubular response to the standardized dose of pitressin and age
is shown. Correlation coefficient was — 0-73, and the regres-
sion of the concentration on age was described as U/P inulin =
162 —1-6 X age (in years). Although the administered
pitressin resulted in a rise of blood pressure, it averaged only
10 mm. at two minutes after injection, and fell to control
levels within five minutes. These experiments indicate that,
in the older individual, there is an impairment in the

240

N. W. Shock

functional capacities of the tubular cells to perform osmotic
work on the glomerular filtrate.

The results of these observations lead to the concept that,
with increasing age, there is a gradual loss of nephrons in the

1 — I r

URINE COLLECTION PERIOD

Fig. 10. Mean values of U/P inulin ratio for each of
three age groups before and after the intravenous ad-
ministration of pitressin. Urine collection periods 1-9
represent nine consecutive 12-minute periods. Pitressin
was administered immediately after the conclusion of
period 3.

(From: Miller and Shock, 1953).

are

kidney. In addition to these structural losses there
functional changes. One of these is a gradual increase in the
vasoconstriction of the vascular bed of the kidney which
further reduces the flow of blood through it, even in the face

Age Changes in Renal Function

241

of the falling cardiac output. This vasoconstriction is func-
tional in character and can be removed by an appropriate
physiological stimulus. Although the tubular epithelium
responds to the stimulus of the antidiuretic hormone as
quickly in the old as in the young, the functional capacity of
the tubular epithelium to perform osmotic work shows a
gradual reduction with age.

opr

201
180 —
I 60 —
140
120
100

40 —

20 —

AGE IN YEARS
Fig. 11. Relationship between maximum U/P inulin following pitressin,
and age. The ordinate is the mean U/P ratio for periods 5 and 6.
(From: Miller and Shock, 1953).

Although these experiments serve to define certain limita-
tions in renal function with increasing age, we must turn to
other observations to tell us how effective the aged kidney
is in maintaining volume and concentration characteristics of
the extracellular fluid. With regard to electrolyte concen-
tration of the plasma, there is no evidence of any system-
atic changes with age. Although Videbaek and Ackermann
(1953) reported a slight rise in plasma potassium concentra-
tions, 4 -0-4 -5 m-equiv./l., between the ages of 25 and 90, the

242

N. W. Shock

trend was not statistically significant. The other major
electrolytes, sodium and chloride, do not show any age trend
(de Billis, 1954; Herbeuval, Cuny and Manciaux, 1954; Lippi
and Malerba, 1955). In our own laboratory we have found no

45
43

7.40

7.39

7.38

7.37
f

tflOj

"o^

dii

55 65

MEAN AGE

Fig. 12. Trends in the acid-base equilibrium of the
blood of males with increasing age. Average curves
from top to bottom include percentage of red cells,
serum pH at 38°, carbon dioxide tension expressed
in millimetres of mercury, serum bicarbonate and
blood carbon dioxide content, both expressed in milli-
moles per litre. The vertical lines indicate ± one
standard error of the mean. Data for the 25-year de-
terminations taken from : Hamilton and Shock (1936).

(From : Shock and Yiengst, 1950).

systematic age changes in the total osmotic pressure of the
plasma or its water content. The bicarbonate content of the
plasma and the pH do not show significant age trends (Shock
and Yiengst, 1950). Thus, under basal conditions the kidney
is able to regulate the acid-base equilibrium of the body
adequately, even to advanced ages (Fig. 12). Lewis and

Age Changes in Renal Function

243

Alving (1938) found some evidence that with increasing age
there is an accumulation of urea nitrogen in the blood. Their
data show a sUght rise in the fifth decade, but no significant
change during the sixth and seventh decades, with a rather
sharp increase after the 70th year. Most of the total rise from
a mean of 12-9 mg. urea N/100 ml. blood in the 30-40 age

r\c\

1 1

1 1 1 1

(

o 0 ^„

- o»o "o

0°

0 0 -
00 0 0 0

a 0

>^°

0 0 0 8

" 0

°y

,■•: i.'V '

' 1

0
0

.•.°

■ ■

a
■

•

- •: .

■

'..••!* :.'- -

■ -.

■

■ ■■.>

- ,•-/.'

'.•;'•

•.

■ ■

• • • _

^. 1 1

1 II 1

AGE YEARS

Fig. 13, Total blood volumes, ml. per kg., and plasma volumes,
ml. per kg., in 105 males. □ total blood volume determinations
from Gibson and Evans (1937). | plasma volume (Gibson and
Evans), Q total blood volume, 0 plasma volume (Cohn and
Shock).

(From: Cohn and Shock, 1949).

group to a mean of 21-2 mg. per cent in the 85-89-year-olds
occurred after the age of 70. It therefore appears that there is
some impairment in the excretion of nitrogenous substances
in the aged kidney, although capacity for maintaining electro-
lyte concentrations under resting conditions is still adequate.
With increasing age there is a reduction in the concentrating
ability of the kidney. The maximum specific gravity attained

244

N. W. Shock

after 12 hours of water deprivation falls from an average of
1-032 at age 20 to 1-024 at age 80-90. Although the absolute
magnitude of the decrement is small, it is statistically signi-
ficant (Lewis and Alving, 1938) and indicates impairment of
the concentrating ability of the kidney, which is no doubt a
reflection of the reduction in Tm as reported from our studies.
With regard to volume regulation, our observations on a

£.£.

-J

LjJ

U-

-1

l(=)

Ixl

T r

.. • - •
• • •• •

. ••

s • • ,

• • •

^h^

J L

20 30 40 50 60 70 80 90 100

Age Years

Fig. 14. Relationship between extracellular fluid space (thio-

cyanate space) and age in males.
(From: Shock, 1956).

series of 152 males failed to demonstrate any systematic
changes in either plasma volume (Cohn and Shock, 1949) or
in total extracellular fluid volume (Shock, Watkin and Yiengst,
1954) as estimated by thiocyanate determinations (Figs. 13
and 14).

Although the aged kidney has a capacity for maintaining
acid-base equilibrium of the plasma under resting conditions,
when an extra load is imposed upon it age differences appear.
Thus, for example, we have found that a single dose of

Age Changes in Renal Function 245

ammonium chloride produces displacements of the acid-base
equilibrium in both old and young subjects. However, young
individuals are able to readjust equilibrium within a period of
eight hours, following a single dose of 10 g. of ammonium
chloride, whereas the older subjects require as much as 24-36
hours for the process (Shock and Yiengst, 1948). When
repeated daily doses of 1 • 5 m-equiv. ammonium chloride/kg.
body weight /day were administered to normal subjects for
4-14 days, readjustment of the acid-base equilibrium occurred
within 5-7 days in the young subjects, but the aged subjects
(65-73 years) were unable to attain equilibrium under this
load of ammonium chloride (Hilton, Goodbody and Kruesi,
1955). It was also found that the degree of metabolic acidosis
induced by a standard dose of ammonium chloride showed a
greater severity in the older subjects than in the young. We
have now initiated a study of age differences in the ability of
the individual to regulate plasma and extracellular fluid
volume following the imposition of an oncotic load.

Thus, the evidence now available indicates that in spite of
the reduction in discrete renal functions with age, the kidney
retains sufficient capacity to regulate both concentrations and
volumes fairly closely under conditions of rest. However,
when experimental displacements are produced, age differ-
ences in the speed of readjustment appear.

There are obviously many other questions, such as age dif-
ferences in glomerular permeability and the activity of specific
cellular enzymes in the kidney, which remain unanswered.
Studies on cellular enzymes are now in progress in our labora-
tory, using the rat as an experimental animal. Although we
have found a reduction in the total oxygen uptake for kidney
tissue between the ages of 12 and 24 months in the rat, these
differences disappear when an appropriate correction for cell
number is introduced. There are, however, some specific
enzymes, such as succinoxidase, which show an age reduction
which is apparently not dependent on the number of func-
tioning cells in the kidney preparation (Barrows et al., 1957).
It is our aim to extend these observations to include the

246 N. W. Shock

capacity for concentrating specific substances, such as PAH, in
tissue slices removed from the kidneys of animals of different
ages. It is thus apparent that a great deal of research re-
mains to be done before we can interpret age changes in renal
physiology.

REFERENCES

Barrows, C. H., Jr., Yiengst, M. J., Shock, N. W., and Chow, B. F.

(1957). Fed. Proc, 16, 7.
BiLLis, L. DE (1954). Boll. Soc. ital. Biol, sper., 30, 370.
Brandfonbrener, M., Landowne, M., and Shock, N. W. (1955).

Circulation, 12, 557.
CoHN, J. E., and Shock, N. W. (1949). Amer. J. med. Sci., 217, 388.
Gibson, J. G., II, and Evans, W. A., Jr. (1937). J. din. Invest., 16, 317.
Hamilton, J. A., and Shock, N. W. (1936). Amer. J. Psychol., 48, 467.
Herbeuval, R., Cuny, G., and Manciaux, M. (1954). Pr. med., 62,

1555.
Hilton, J. G., Goodbody, M. F., Jr., and Kruesi, O. R. (1955). J.

Amer. geriat. Soc, 3, 697.
Lewis, W. H., and Alving, A. S. (1938). Amer. J. Physiol, 123, 500.
Lippi, B., and Malerba, G. (1955). Arch. E. Maragliano, 11, 839.
McDonald, R. K., Solomon, D. H., and Shock, N. W. (1951). J. din.

Invest., 30, 457.
Miller, J. H., McDonald, R. K., and Shock, N. W. (1952). J. Geront.,

7, 196.
Miller, J. H., and Shock, N. W. (1953). J. Geront., 8, 446.
Moore, R. A. (1931). Anat. Rec, 48, 153.
Shock, N. W. (1952). In Cowdry's Problems of Ageing, p. 614, 3rd ed.,

ed. Lansing, A. I. Baltimore: Williams & Wilkins.
Shock, N. W. (1956). Bull. N.Y. Acad. Med., 32, 268.
Shock, N. W., Watkin, D. M., and Yiengst, M. J. (1954). Fed. Proc.y

13, 136.
Shock, N. W., and Yiengst, M. J. (1948). Fed. Proc, 7, 114.
Shock, N. W., and Yiengst, M. J. (1950). J. Geront., 5, 1.
Smith, H. W., Goldring, W., and Chasis, H. (1938). J. din. Invest., 17,

263.
Videbaek, a., and Ackermann, P. G. (1953). J. Geront., 8, 63.
Watkin, D. M., and Shock, N. W. (1955). J. din. Invest., 34, 969.

DISCUSSION

Zweymiiller : One of the interesting things in your paper, Dr. Shock,
was this tendency for the glomerular filtration rate, Tnip^n and Tmo to
fall, which leads to the conclusion that the total number of nephrons is
diminished. Are the nephrons which are left, and particularly the
tubules, still able to elevate their function? Under normal physio-
logical conditions we have a Tmp^H» which means that under normal

Discussion 247

conditions this function has an upper limit. If Vitamin A is fed this
action is elevated, and it is called trophic action. It would be interesting
to give old people Vitamin A and see if this normal Tmp^H for physio-
logical conditions could be elevated in this way.

Shock : We infused lactate in some of these older people who had low
Tms and we found that this did raise the Tm by providing additional
substrate; you can almost double the Tm for PAH in both old and
middle-aged subjects (McDonald, R. K., Shock, N. W., and Yiengst,
M. J. (1951). Proc. Soc. exp. Biol., N.Y., 77, 686). In other words the
tubules that are still present in the old kidney, as far as we have been
able to determine, are just as good as in the young. This is all very dis-
tressing to me because I am convinced that there must be progressive
changes. The tubule just cannot be working beautifully today and gone
tomorrow, but unfortunately this is the way the data come out so far.

Heller : We all know that there has been a lot of difficulty in the choice
of parameters when attempting to compare renal function in adults and
infants. I should therefore like to ask Dr. Shock whether he has tried to
express his data in terms of other parameters like, for example, total
body water. That would seem important because it might reveal cor-
relations which may have a functional significance.

Shock: Yes, we have done that, and if you refer metabolism to total
body water you wipe out the age change. However, the age decrement
in renal function remains, even when calculated on the basis of body
water.

Hingerty: When you selected your subjects, Dr. Shock, did you ex-
clude obese patients?

Shock : We did not use any index of body weight to exclude patients,
but I would say immediately that in our population we do not see obese
people over the age of 65. These patients were all males so we do not
know anything about the weight of females.

Hingerty : It seems to me that the decline in kidney function sets in at
about the 50-year mark, and that is about the age when you would expect
a higher incidence of obesity in the general population.

Shock: Actually the body surface area decreases with increasing age
in all groups of subjects we have studied. The major factor that contri-
butes to this reduction in surface area is the body height, which goes
down more than body weight. There is a wide scatter in height in our
population, but there is a statistically significant linear decrement
between the ages of 30 and 90. There is no significant regression of
weight on age in the population of male^ that we have studied.

Black: Is there any serial change in the blood urea with age? It seems
very odd that if you give some lactate these tubules can hypertrophy in
function to twice their previous extent, and yet when you study them
without any stimulus they are apparently in a fairly low state of func-
tion. If the blood urea does not go up then it looks as if the remaining
tubules are perfectly able to cope with the diminished urea formation
within the body.

Shock : The subjects used in our renal function studies had no elevation
in blood urea because this was one of the selection criteria. Every

248 Discussion

individual in the renal series was able to concentrate his urine at least
to a specific gravity of 1 • 020 on a Fishberg routine. However, Lewis and
Alving (1938. Amer. J. Physiol., 123, 500) have published blood urea
levels in 100 subjects aged 20 to 80. They found little increment in
blood urea up to the age of about 70, but from 70 on it does increase in
their data.

I must make it clear that the increment in Tm following lactate infu-
sion occurs only during the time that the blood lactate level is raised. We
have not been able to show that it induced any kind of renal hypertrophy.

Fejfar: I would not expect the blood urea level to increase, because in
a paper on chronic nephritis, Brod (1948. Cas. Lek. des., 87, 711) showed
that the blood urea did not rise markedly in patients with low protein
intake unless the glomerular filtration rate decreased to less than 25-30
ml./min.; in your work the glomerular filtration rate was far above this
figure.

In congestive failure or other situations where cardiac output is inade-
quate, there is usually a decrease in renal blood flow, and an increase in
tubular reabsorption of water. The normal concentration test might
point to a diversion of blood from the kidneys due to this insufficient
cardiac output.

Shock : I did not perhaps make it clear that unfortunately we only got
the cardiac output method in operation rather late in the series, so that
the cardiac output results that I showed you in the average curve were
not determined on the same subjects as the renal functions. We are now
measuring cardiac output and renal function in the same subjects simul-
taneously. The crucial point to me is whether there is a change in the
percentage of cardiac output that gets through the kidney, and I just
cannot answer that at the moment.

Milne: I have some difficulty about this fall in glomerular filtration
rate without a rise in blood urea with advancing age. It seems to me that
this could only be possible if the older people were not taking in so much
protein, or if the urea back-diffusion was diminishing and therefore the
clearance of urea was approaching the inulin clearance. I should have
thought that a fall in glomerular filtration rate of this magnitude would
necessitate a rise in blood urea, although it might not of course go above
some arbitrary upper limit of normal such as 40 mg./lOO ml.

Black: My question on blood urea really referred to blood urea in a
population and not in an individual. I think Van Slyke showed that in
terms of a population, even with 80 per cent of normal urea clearance
there is a detectable increase in the blood urea. All our clinical experi-
ence is that the glomerular filtration can be down to 30 per cent without
the blood urea being outside the so-called normal range in that individual
but if you do it in a population you then find that even with an 80 per
cent clearance the level is raised.

Borst: Dr. Shock, you eliminated all diseased people, but at what
blood pressure was a man eliminated as not having normal kidneys?

Shock : We excluded anyone who had a systolic pressure greater than
160 and a diastolic greater than 90 mm. Hg. Prof. Olbrich (Olbrich et
al. (1950). Edinh. med. J., 57, 117) was doing similar renal functional

Discussion 249

experiments at almost the same time. He did not exclude subjects with
elevated blood pressures and his results on British subjects are almost
identical with those we found by excluding the individuals with ele-
vated blood pressures.

Scribner: The question posed by these data is: is the change in the
kidney function, as described, a result of disease in the kidney or a
wearing out with age, or is it simply a response to a decrease in the size
of the living organism? This all comes back to the point raised by Prof.
Heller and Prof. Borst: might not creatinine excretion, or total ex-
changeable potassium, be reasonable reference points?

Shock: We have done a good many creatinine determinations in
balance studies under conditions of a closely regulated diet. However,
I have never been able to convince myself that creatinine excretion gives
a stable value that is characteristic of the individual, because we have
seen some rather wide fluctuations that we have not been able to explain
satisfactorily. I tried it first with adolescent children and then gave it
up as I did not feel it could be determined as a characteristic constant
for the individual. But I am intrigued by the potentiality of the total
exchangeable potassium, and would like to study its changes with age.

Bull : The lines you showed in illustrating the decline of renal function
with age are practically identical with the lines for our mortality findings
in burns. By Probit analysis we can fit LDjo's for the areas of burning
which will produce death at different ages. It may be coincidence that
you chose your ordinates on just the right scale, but the lines are almost
the same in that they take off at just the same age and go down in the
same way. Burning is a severe stress. We have been talking about the
elderly having a reduced tolerance to stress, and burning is largely a
stress affecting water and electrolytes. The burn is a convenient
measurable lesion, and death occurs with a progressively smaller size
of burn with advancing years, which I think probably represents an
important aspect of the ageing of a regulation of water and salt.

AGE AND RENAL DISEASE

G. C. Kennedy

Medical Research Council, Department of Experimental Medicine,
University of Cambridge

Introduction

Senescence has sometimes been described as a deteriora-
tion in homeostasis. Dr. Shock showed us that the deteriora-
tion may be due to faihng renal function, and this reopens an
old question of whether the kidney cells themselves are less
able to do their work in old people, or whether diseases of the
kidney become more frequent with advancing age. It seems
generally agreed that pathological lesions, particularly of the
renal vessels, are very commonly found post mortem in old
people in whom they were unsuspected during life. Oliver
(1942) reviewed the controversy as to whether these lesions
originate from a primary atrophy of the kidney, or are merely
one of the results of generalized arteriosclerosis. He decided
in favour of arteriosclerosis. The other view, that the kidney
dies piecemeal, will be re-examined here because it seems
possible to show that the death of some nephrons leads to
pathological changes in the survivors, and some indirect
ways in which this may happen will be suggested.

One can raise objections to any theory of ageing. The
major defect of the definition in terms of homeostasis, it seems
to the present author, is that the newborn animal finds it just
as difficult to maintain a stable internal environment under
stress as does the senile one. An older definition by Minot
(1908), in more structural terms, described senescence as the
gradual loss by differentiated cells, throughout life, of the
ability to grow and to regenerate. This idea applies especially
well to the kidney, as we shall see.

250

Age and Renal Disease 251

Renal growth and regeneration

Both the tentative and the definitive foetal kidneys
develop from mesoderm, in intimate relation with the gonads.
So it is not altogether surprising that the adult kidney
resembles the other transient tissues, and its life cycle is not
completely synchronous with that of the rest of the body
(Kennedy, 1957). It would be disastrous to a species, of
course, if kidney and body got too far out of step, but any
tendency for this to happen during reproductive life would
be prevented by natural selection. There is some evidence,
however, that the kidney atrophies after the climacteric, and
in some species such as the rat this may limit life.

Most mammals develop their full complement of nephrons
soon after birth, and postnatal growth of the kidney consists
chiefly of lengthening of its tubules, at first by the growth of
new cells and later by hypertrophy of existing ones. When a
rat is about six months old, or a man about 30 years, the
number of glomeruli in their kidneys begins to decrease, and
it may fall to half the young adult value, without pathological
change, by eighteen months old in the rat or seventy years in
the man (Arataki, 1926; Moore, 1931; Roessle and Roulet,
1932). Moore and Hellman (1930) showed that removing one
kidney from a rat did not slow down the loss of nephrons from
the other, so that involution of the kidney is an even more
relentless process than that of the ovary, where removal of one
gland does delay the loss of oocytes from the other (Mandl and
Zuckerman, 1951).

Nowadays chemical analysis can be used to supplement
histology in determining the number and size of the cells in a
tissue. This is because one of the two forms of nucleic acids in
cells, deoxyribonucleic acid or DNA, is confined to the nuclei,
as the name suggests it ought to be, while the other, ribo-
nucleic acid or RNA, is distributed with the bulk of the
ordinary protein throughout the cytoplasm. So if DNA,
RNA and protein are determined at different stages during
the growth of a tissue, it is possible to distinguish between

252

G. C. Kennedy

an increase in nuclei, or hyperplasia, and an increase of
cytoplasm, or hypertrophy. This method has shown that the
principal increase in the number of nuclei in the kidney of
the rat occurs during the first three months of life, pari passu
with the main growth of the skeleton, and this agrees well
with histological findings.

There is a conflict of evidence about regeneration, however.
Rollason (1949) showed histologically that mitosis began in
the surviving kidney within forty-eight hours of unilateral
nephrectomy, whereas Mandel, Mandel and Jacob (1950)

Table I

The effect of unilateral nephrectomy on the composition of

THE surviving KIDNEY IN RATS AT DIFFERENT AGES

Age at
Ojyeration

Interval

before

Killing

Group

Total
nitrogen
{mg. per
kidney)

RNA

phosphorus
(mg. per
kidney)

DNA

phosphorus
(mg. per
kidney)

One
Month

Two
Weeks

Control (not
operated)

Kidney
removed

11-3
180

0-273
0-424

0-165
0-233

Three
Months

Six
Weeks

(

Control
Kidney
removed

19-4
29-4

0-378

0-488

0-183
0-227

Six
Months

Six
Weeks

1
I

Control
Kidney
removed

28 1
41 0

0-587
0-717

0-253
0-244

were unable to show any increase in kidney DNA even three
wrecks after the same operation. The difference apparently
depends on the age of the animals. Table I illustrates a
comparison made by the present author of the effect of
unilateral nephrectomy on the composition of the surviving
kidney in one-month, three-month and six-month-old rats.
In the youngest group, which were about the same age as
Rollason used, there was a rapid increase in DNA phosphorus.
In the middle group the DNA increased less than the RNA
and the nitrogen, and more slowly, as Mandel, Mandel and
Jacob had found. No hyperplasia at all occurred in the

Age and Renal Disease 253

kidneys of the six-month-old rats. It may be emphasized
that these findings accord very well with Minot's definition
of ageing. As will be shown later, hyperplasia can and does
occur in the tubules of older rats, but it does not then repre-
sent the normal primary response to loss of moderate amounts
of renal tissue, and some additional stimulus, possibly endo-
crine in nature, is probably involved.

Renal Senescence

The compensatory changes that we have been considering
are self-limiting, and once they have been achieved, the kidney
undergoes no further changes for many months. A different
sort of tubular change will now be considered. In rats killed
after 18 months of age very active hyperplasia has been found
in occasional tubules, at first widely scattered, affecting
principally the proximal convolutions, and quite unlike the
regular, orderly growth of cells in young rats' kidneys. At
this age a lot of nephrons have already disappeared, but one
would expect the surviving tubules to compensate for their
loss by hypertrophy rather than hyperplasia. Further, this
hyperplasia in ageing kidneys appears to be destructive rather
than helpful, because the tubules are often blocked and
functionless and eventually become dilated by hyaline casts.
As age increases still further the kidneys become greatly
enlarged and granular in appearance, and microscopically
they show chronic interstitial fibrosis, generalized tubular
dilatation, and hyaline or fibrotic changes in the glomeruli
and smaller vessels. These histological changes have been
described and illustrated more fully elsewhere (Kennedy,
1951, 1957). The terminal appearance has been studied by
numerous pathologists, but since no two agree on a morbid
anatomical diagnosis, there is no need to add to the confusion
here. The terms chronic glomerulonephritis (Wilens and
Sproul, 1938), nephrosis (Saxton and Kimball, 1941), pyelo-
nephritis (Goldblatt, 1947) and senile nephrosclerosis (Oliver,
1942) have all been used.

254 G. C. Kennedy

The pathological renal changes in old rats are almost in-
variably accompanied by great enlargement of the adrenals,
frequently by parathyroid hyperplasia, and in the later
stages, at least, by cardiac hypertrophy and hypertension.
Before considering further which is cause and which is effect,
a description will be given of a number of ways in which
similar renal lesions can be produced in much younger rats in
association with the same endocrine and vascular changes.

"Senile" changes after renal overloading
in younger rats

The first condition in which these lesions were found in
fairly young rats was in experimental hypothalamic obesity.
When the ventromedial part of the hypothalamus is des-
troyed electrolytically, the appetite of a rat may be doubled
for several weeks and the animal becomes grotesquely fat. In
view of the association of clinical obesity with renal disease
and hypertension, it is interesting that most of these fat rats
developed typical senile kidney lesions about nine months
earlier than unoperated controls (Kennedy, 1951). If the
animals were operated on at three months old, they survived
nine to 12 months before pathological lesions appeared in the
kidneys, but the kidneys became enlarged during the period
of overfeeding soon after the hypothalamic puncture. Moise
and Smith (1927) and Addis and Oliver (Oliver, 1945) showed
that the renal enlargement produced by a high protein diet
in rats could eventually cause pathological changes, and it
seemed possible that this might be the way in which the kid-
neys were damaged in hypothalamic overfeeding. As a first
step an examination was made of the chemical changes in the
kidneys during the earlier stages of development of the
obesity, while the food intake was very high. In Table II
these are compared with the changes found previously in the
surviving kidneys after unilateral nephrectomy, and they
followed an almost identical pattern. This suggested a con-
venient way to isolate the effect of simple kidney overloading
during overfeeding from any possible effect of the subsequent

Age and Renal Disease

255

adiposity and abnormal fat metabolism. If the normally-fed
rats with one kidney were to develop the same pathological
renal changes as the overfed rats with two, then it would be
reasonable to attribute the lesions to some effect associated

Table II

Composition of the kidneys of obese

KIDNEY removed, AT THREE AND SIX

RATS, OR OF RATS WITH ONE
WEEKS AFTER OPERATION

Time from
operation

Group

Total
nitrogen
{mg. per
kidney)

RNA

phosphorus
{mg. per
kidney)

DNA

phosphorus
{mg. per
kidney)

Three weeks

Control (not
operated on)

Kidney removed
Obese

19-4

27-8
29-9

0-378

0-504
0-488

0-183

0-210
0-227

Six weeks

Kidney removed
Obese

29-4
28-3

0-516
0-532

0-289
0-308

with overloading. They did develop the lesions at the same
time as the obese rats, at an average age of 15 months.
Table III illustrates the changes in composition of the kidneys
12 months after carrying out each type of operation on three-

Table III

Composition of the kidneys of obese animals, or of animals

WITH ONE kidney REMOVED, TWELVE MONTHS AFTER OPERATION
(operated at three MONTHS OF AGE)

Group

Control (not operated on)
Kidney removed
Obese

Total
nitrogen
{mg. per
kidney)

34-9
73-5
75-9

RNA

phosphorus
{mg. per
kidney)

0-749
1-429
1-972

DNA

phosphorus
{mg. per
kidney)

0-239
0-695
0-785

month-old animals. Note that in each case the final renal
breakdown occurred quite quickly and that rats killed during
the period between four months and a year old had large but
otherwise normal kidneys.

The period of latency is interesting, because in subsequent
experiments it became shorter with increasing age of the

256 G. C. Kennedy

animal at operation, and in fact the age at which the final
breakdown occurred was almost constant. To take the
extreme case, rats over a year old frequently failed to estab-
lish any new renal equilibrium after either type of overloading,
but rapidly developed pathological lesions.

The age at which renal failure occurred was advanced still
further by increasing the renal loading, either by a more
extensive partial nephrectomy, or by combining unilateral
nephrectomy with overfeeding. It is sometimes said that
different species tolerate the removal of different proportions
of their renal tissue. It is difficult to see how a valid compari-
son can be made when the critical amount of kidney depends
so much on the age of the animal. We found that weanling
rats recovered and survived for many months after losing
five-sixths of their kidneys, while nine-month-old adults
often developed acute tubular necrosis after the same opera-
tion. A probable explanation for the latent period in the
younger animals is that it represents the time for the further
loss of nephrons due to ageing to reduce the available kidney
below the critical level. It remains to consider the part played
by the associated metabolic and endocrine disturbance in
destroying the kidney.

Endocrine stimuli to renal hyperplasia

A number of hormones are reno trophic. They include
growth hormone (White, Heinbecker and Rolff, 1949),
thyroid hormone (Korenchevsky and Hall, 1944) and testo-
sterone (Korenchevsky and Ross, 1940). The results of
treatment with growth hormone are particularly suggestive.
Acute overdosage can lead to rapid kidney destruction, but
treatment of a young rat for only a few days, apparently
causing no damage at the time, can lead to the appearance of
pathological lesions months later (Selye, 1951). From the
limited descriptions and photographs available no difference
can be seen between these and the spontaneous lesions of
older rats or those which develop after partial nephrectomy.
Interpretation is complicated because partial nephrectomy is

Age and Renal Disease 257

among the measures that Selye uses, as he says, to "sensitize"
the rat to the damaging effect of hormones. Nevertheless,
we have found that no overgrowth of the kidney occurs in
hypophysectomized rats with hypothalamic lesions, although
they still have increased appetites, and other tissues, such as
the liver and gastrointestinal tract, hyperti'ophy "(Kennedy
and Parrott, 1958). We also confirmed, as White, Heinbecker
and RolfP (1941) first showed, that compensatory growth after
partial nephrectomy required the presence of the pituitary.
However, the late renal changes in our rats were associated
with a catabolic rather than an anal^olic state of the body as a
whole, so it seems unlikely that growth hormone was being
secreted in excess.

There remains the possibility that adrenal overactivity
plays a part in the final renal breakdown. Adrenal enlarge-
ment and the nephrotic character of the renal defect (Saxton
and Kimball, 1941) have been mentioned. A number of
workers have shown that complete or extensive partial
nephrectomy is followed by increased urea production
(Bondy and Engel, 1947; Persike and Addis, 1949; Persike,
1950; McCance and Morrison, 1956). This has recently been
shown to be due to increased protein catabolism in the liver
(Sellers, Katz and Marmorston, 1957), so it may well be a
result of increased adrenal activity. Overdosage with adrenal
steroids can certainly cause renal breakdown associated with
extensive tubular hyperplasia, although the immediate cause
may be potassium deficiency (Follis, 1948) or sodium reten-
tion (Ingle, 1958) associated with such experiments. We have
learned little from the serum electrolytes of our rats, because
any changes that might implicate the adrenal are obscured
by the general electrolyte retention of incipient uraemia.
Morrison and Gordon (1957), however, have shown that
increased urea excretion during starvation occurs both in
partially nephrectomized and senile rats before obvious renal
damage and is accompanied by an increased potassium loss.

Another renoprival effect that may hasten the end of the
kidney is hypertension, although again the exact relation

AGEING— IV— 9

258 G. C. Kennedy

between cause and effect is uncertain. Wilson and Byrom
(1939, 1941) showed that the production of hypertension by
"cHpping" one kidney could lead after a prolonged latent
period to pathological lesions in the other kidney. They
attributed these lesions to hypertension, because their
development seemed to be arrested and the hypertension
cured by removing the ischaemic kidney. Goldblatt (1947)
pointed out that all the lesions Wilson and Byrom had de-
scribed could occur spontaneously in rats without hyperten-
sion. More recent work, reviewed by Floyer (1957), suggests
that removal of the clip, so restoring some of the lost excretory
function, is a much better protective measure than removing
the ischaemic kidney, which frequently increases the hyper-
tension. The importance of extrarenal or renoprival factors
in producing permanent hypertension now seems well estab-
lished and certainly fits with our experience, and apparently
with Goldblatt' s, that hypertension and vascular changes are
a late feature of the spontaneous renal disease of rats.

Much remains to be done, but it is hoped that some pro-
gress has been made towards establishing the thesis, stated at
the beginning of this paper, that the essential vicious cycle of
renal disease in old age, in one species at least, is the destruc-
tion of surviving nephrons by overloading, after the normal
renal atrophy of old age has reached a critical stage.

Summary

The kidney of the rat, and of most mammals including man,
begins to atrophy while the animal is still young. Pathological
changes in the kidney become more frequent during involution.
Irregular and apparently purposeless hyperplasia of tubular
cells is a prominent feature of such lesions. Hyperplasia
occurs in the tubules of growing rats both as part of normal
development and as a response to a moderate increase in the
excretory load, but it is not normally seen after the main
growth of the skeleton is completed. The stimulus to normal
renal growth probably arises in the pituitary gland. It is

Age and Renal Disease 259

suggested that the loss of renal tissue in excess of a critical
amount leads to additional renotrophic stimuli, probably
related to overactivity of the adrenal cortex and to hyperten-
sion, which hasten the end of the remaining nephrons.

REFERENCES

Arataki, M. (1926). Amer. J. Anat., 36, 399.

BoNDY, P. K., and Engel, F. L. (1947). Proc. Soc. exp. Biol., N.Y., 66,

104.
Floyer, M. a. (1957). Brit. med. Bull., 13, 29.

FoLLis, R. H. (1948). The Pathology of Nutritional Disease. Spring-
field: Thomas.
GOLDBLATT, H. (1947). Physiol. Rev., 27, 120.
Ingle, D. J. (1958). Personal communication.
Kennedy, G. C. (1951). Proc. R. Soc. Med., 44, 899.
Kennedy, G. C. (1957). Brit. med. Bull, 13, 67.

Kennedy, G. C, and Parrott, D. M. V. (1958). J. Endocrin., in press.
KoRENCHEVSKY, V., and Hall, K. (1944). J. Path. Bact., 56, 543.
KoRENCHEVSKY, V., and Ross, M. A. (1940). Brit. med. J., 1, 645.
McCance, R. a., and Morrison, A. B. (1956). Quart. J. exp. Physiol.,

41, 365.
Mandel, p., Mandel, L., and Jacob, M. (1950). C. R. Acad. Sci., Paris,

230, 786.
Mandl, a. M., and Zuckerman, S. (1951). J. Endocrin., 7, 190.
MiNOT, C. S. (1908). The Problem of Age, Growth and Death. New

York: Putnam Press.
MoiSE, T. S., and Smith, A. II. (1927). Arch. Path. (Lab. Med.), 4, 530.
Moore, R. A. (1931). Anat. Rec, 48, 153.

Moore, R. A., and Hellman, L. M. (1930). J. exp. Med., 51, 51.
Morrison, A. B., and Gordon, J. (1957). Fed. Proc, 16, 366, and

personal communication.
Oliver, J. (1942). In Problems of Ageing, ed. Cowdry, E. V., 2nd ed.,

p. 302. Baltimore: Williams & Wilkins.
Oliver, J. (1945). Harvey Lect., 40, 102.
Persike, E. C. (1950). Arch, intern. Med., 85, 1.
Persike, E. C, and Addis, T. (1949). Amer. J. Physiol, 158, 149.
RoESSLE, R., and Roulet, F. (1932). Mass und Zahl in der Pathologic.

Berlin: Springer.
RoLLASON, H. D. (1949). Anat. Rec, 104, 263.
Saxton, J. A., and Kimball, G. C. (1941). Arch. Path. (Lab. Med.),

32, 951.
Sellers, A. L., Katz, J., and Marmorston, J. (1957). Amer. J.

Physiol, 191, 345.
Selye, H. (1951). First Annual Report on Stress, p. 16, 356. Montreal:

Acta, Inc.
White, H. L., Heinbecker, P., and Rolf, D. (1941). Amer. J. Physiol,

149, 404.

260 G. C. Kennedy

White, H. L., Heinbecker, P., and Rolf, D. (1949). Amer. J. Physiol.^

157, 47.
WiLENS, S. L., and Sproul, E. E. (1938). Amer. J. Path., 14, 201.
Wilson, C, and Byrom, F. B. (1939). Lancet, 1, 136.
Wilson, C, and Byrom, F. B. (1941). Quart. J. Med., 10, 65.

DISCUSSION

Swyer: You said that this renal damage in the obese rat might be a
question of protein overloading. Did you try feeding these rats on an
isocaloric diet, but with half the protein content?

Kennedy: I have tried it as a short-term experiment but I did not
carry it to its logical conclusion. There was no renotrophic effect.

Sivyer : Over- feeding is itself a stressful activity in the Selyeian sense
and that alone might lead to adrenal over-activity. Certainly there is
clinical evidence that it may. Obese people who give evidence of in-
creased adrenal steroid production may cease to do so after they have
been put on a diet and have had their weight brought down to normal.

Kennedy : To answer that I must challenge the question of whether in
fact stress ever produces renal lesions in the rat. I can do that quickly
by quoting some recent work by Crane, Baker and Ingle (1958. Endo-
crinology, 62, 216; and Crane and Ingle, Endocrinology, 62, 474), who
have studied a large number of so-called stresses which sound quite
barbaric, and have found that the only one which produces what Selye
calls the stressed kidney is exposure to cold. Selye has always said
that this is the most effective, and these workers now say that it is the
only effective stress. Under those circumstances the rats eat twice as
much food. If they are then fed isocalorically, as you suggest, with a
high caloric diet made up with carbohydrate and fat, they do not
develop lesions. These workers attribute renal lesions to overloading
with salt; I choose protein.

Talbot: Will you take this as evidence in favour of restricting the
protein intake of patients with handicapped renal function?

Kennedy: I can see that it would be a dangerous thing to press a
trophic stimulus like a high protein intake too far in an attempt to get
recovery. Are you thinking of chronic renal disease, or a recovery from
acute damage?

Talbot: Both.

Kennedy : Purely from my own findings I would have said that I could
see no point in producing additional renal growth in trying to help
recovery of the kidney by giving a high protein diet; if the object was
simply to replace protein lost from the body then my results, of course,
are not relevant. I think the problem of a high protein intake has to be
studied from this point of view on the human, and we cannot answer
from the work on the rat. Moreover, there may be a totally different
limitation to the structural renal reserve in the rat, which has a kidney
of completely different anatomical character.

Borst: We treat all patients with a kidney function of less than 10 per
cent of normal with a diet adequate in calories but very poor in protein

Discussion 261

(less than 20 g. daily). I have no comparison with a group of patients
who continued eating normal amounts of protein. I have the impression
that our patients can continue longer with their ordinary work. Their
nausea usually disappears, they often gain weight, and in other respects
are also in better condition. We have the paradox that reducing the
protein intake often results in a rise in serum albumin and sometimes in
a slight rise in serum haemoglobin. The protein-poor diet does not
prevent a gradual reduction of the kidney function. However this decline
is usually slow and the patients may have several years of useful life.
A high diastolic blood pressure is a very bad prognostic factor. As long
as we have no control group we cannot produce convincing evidence
that an untreated patient will not live as long as our 'maltreated'
patients.

Kennedy : Have you done any liver function tests in a situation where
serum albumin is falling in spite of a high protein intake, Prof. Borst?
There may be a possible connexion with the increased liver protein
breakdown when one removes the kidney (Sellers, Katz and Marmorston,
(1957). Amer. J. Physiol., 191, 345).

Borst: No liver function tests were done, and we only have data on
the serum proteins. There is no increased y-globulin as is usually found
in chronic hepato-cellular disease. We had, however, some evidence of a
deleterious effect of the low protein diet. More cases of tuberculosis were
seen than would be expected in similar patients on a normal protein diet,
and two patients died from miliary tuberculosis. Probably the extremely
low protein diet reduces the resistance against the tubercle bacillus in
spite of the fact that the patients do not lose weight.

Talbot: How do you define a low protein diet?

Borst: It is less than 20 g./day. To control the diet and determine
whether or not the patient adheres to it, 24-hour urine portions are
regularly examined for nitrogen excretion. We also determine creatinine
excretion to be sure that urine collection is complete. The 24-hour
creatinine output is very constant. This output is determined for every
kidney patient during clinical observation, and we use the figures for
comparison with the nitrogen output when the patients are under control
in the out-patient department. Many adhere to the diet and go along
very well for several years.

Fejfar: We have had similar experiences in Czechoslovakia. This
treatment originated in the experiments of Thomas Addis (1948.
Glomerulonephritis : Diagnosis and Treatment. New York : Macmillan),
who showed that partially nephrectomized rats kept on a higher protein
intake could not survive as long as the animals with a low protein
diet. We therefore started to use a low protein diet in all patients with
chronic glomerulonephritis. Usually we give 0 • 5-0 • 7 g./kg. body weight
per day in the diet (but no less than 0-5 g./kg.), plus the amount lost in
the urine. Of course, children and those with the nephrotic syndrome
are given larger amounts of protein. It is very difficult to judge long-term
results as we have no control group for this treatment. Nevertheless we
do think we can prolong the life of patients with chronic nephritis on
this low-protein diet.

262 Discussion

Richet : In populations that are said to eat a lot of proteins, for in-
stance Eskimos, what is the state of the kidney? Do such people
often die from chronic nephritis? They are generally said to eat 5,000
cal./day, mostly fat and proteins.

McCance : I think in fact the Eskimos do not eat a very high protein
diet, although they may eat a great deal of fat. They certainly tend to
die rather young, but mostly from accidents, I believe ; an old Eskimo is
a man of about 40-45.

Richet: Some work has been done by Lieb (1929. J. Amer. med. Ass.,
93, 20), by Thomas (1927. J. Amer. med. Ass., 88, 1559), and by Bischoff
(1932. J. Nutr., 5, 431), which seemed to demonstrate that a high protein
diet was absolutely harmless.

Dr. Talbot, you mentioned the amount of protein given in cases of
chronic nephritis. In Paris we put some chronic nephritic patients on an
almost protein-free diet, about 10 g./day. Three or four patients whose
death was not expected died after six weeks (Hamburger, J., Serane, J.,
and Cournot, L. (1951). Sem. Hop. Paris, 27, 2289). We therefore never
gave that kind of diet again to any patients for more than ten or fifteen
days. Also, we never give under 0 • 5 g./kg. in chronic cases, because under
that amount we have a lot of trouble and the patients become so weak
they would never live anyway ; we prefer to have a patient with perhaps
a shorter life, but healthy, than the other way round.

Fourman: Tiv. Kennedy, why did you imply a relationship between
catabolic reactions, adrenal hyperplasia and Selye's results with cor-
texone acetate?

Kennedy : The catabolism would require over-secretion of Compound
F, of course. However, Hechter and Pincus (1954. Physiol. Rev., 34, 459)
showed that in the rat the adrenal secretes chiefly Compound B anyway,
and there is not in fact the contradiction there would seem to be. An
over-secreting adrenal could damage the rat kidney and one would, at
the same time, get a catabolic effect.

Fourman: But you would not necessarily want to relate that to the
results with cortexone acetate?

Kennedy: Yes, in that cortexone acetate was the particular steroid
which was used in most of Selye's experiments.

Desaulles : In our laboratories Compound B has been shown to help in
inducing hypertension in the rat.

Kennedy: Does it produce renal lesions?

Desaulles: Only in enormous doses.

Milne: The recovery lesions of potassium depletion are similar in
appearance to the ageing kidney, as mentioned earlier by Dr. Kennedy.
The histological studies reported by Dr. Desaulles seem to show the same
dilatation of the tubules that was seen in Dr. Kennedy's cases. We
repeated these experiments with dietary potassium depletion and cor-
texone acetate injections, but we used very young rats and were unable
to repeat the effects which were shown so conclusively at Cambridge.
Is the ageing kidney, then, more susceptible to permanent damage from
potassium depletion? This would tie up with Dr. Fourman's suggestions
regarding cortexone acetate as given by Selye.

Discussion 263

Kennedy: Dr. Fourman and I have looked at potassium-deficient
kidneys together many times. We agreed then that they were closely
similar to the kidneys we found in old rats of our own colony, and that
this showed that the chronic potassium-deficient kidney is simply an
ageing kidney. Now I am not at all sure that they are not the same thing
anyway: that the ageing kidney is, in a sense, a potassium-deficient
kidney and that there is an element of adrenal over-activity about it.
As you say, Dr. Milne, this may really mean that older age groups are
more liable to potassium-deficient states and the renal consequences of
that.

Fourman : That seems to provide an explanation of why the death of
some nephrons appears to lead to pathological changes in the remainder.
From your studies. Dr. Kennedy, it seems reasonable to argue that in a
potassium-deficient kidney some nephrons die, and as a result in the
remainder there are ultimately pathological changes which are likely to
be worse in older rats.

Kennedy: It is a vicious cycle and we are coming into it at different
points.

McCance : Dr. Kennedy has performed a valuable synthesis in bring-
ing together over-nutrition, age, and lesions in the kidney. No-one asked
and I wish we knew what happens if these kidneys are overloaded with
water and with various other test substances.

RENAL FUNCTION IN RESPIRATORY FAILURE

D. A. K. Black

Departmeyit of Medicine, Royal Infirmary, University of Manchester

With increasing age, the functional capacity of the lungs
and of the kidneys declines. Respiration is embarrassed by
increasing rigidity of the chest wall, and there is also an
increase in the respiratory dead space of the lung itself in
older subjects (Comroe et al., 1955). The kidneys lose efficiency
in consequence of a progressive loss of nephrons, which may
reduce the nephron population to 60 per cent of the original
number; the impairment of renal function is indicated by a
fall in the clearance of inulin and of ^^-aminohippurate, and in
the maximal reabsorptive capacity for glucose (Tm(j) (Shock,
1952). The blood pH in old people is a little lower, and their
plasma returns more slowly to its previous level after imposed
loads of either acid or alkali. These various encroachments
on functional reserve are probably of no great moment in
healthy old folk leading a normal life; but they are brought
into prominence when respiratory function is pathologically
impaired by the related changes of chronic bronchitis,
bronchospasm, and emphysema. In an urban population, the
incidence of chronic bronchitis in old people has been found
to be 40 per cent (Sheldon, 1948); this common illness leads
in time to gross respiratory failure, with the patient afflicted
by anoxia, hypercapnia, and increased pulmonary vascular
resistance in varying degrees. There are several ways in which
advanced respiratory failure can increase the demands on
the kidneys, and also diminish their functional capacity.
This communication outlines the effects on renal function of
chronic hypercapnia and of cardiac failure secondary to
emphysema (cor pulmonale).

264

Renal Function in Respiratory Failure 265

Hypercapnia. The effects of acute hypercapnia, usually
induced by inhalation of 5-10 per cent CO 2, have been
reviewed by Pitts (1953). There is a fall in plasma pH and a
rise in PCO2; the urine formed is acid, and the reabsorption
of filtered bicarbonate is virtually complete, although the
amount of filtered bicarbonate has been increased by the
experimental procedure. Enhancement of bicarbonate re-
absorption is the most striking change in renal performance
induced by acute hypercapnia ; and it persists when the fall in
plasma pH is prevented by infusion of bicarbonate, so that in
this context rise in pCOa seems to be the more relevant
stimulus to bicarbonate reabsorption. The reabsorption of
bicarbonate is also increased in subjects depleted of potassium,
in whom intracellular pH is probably decreased; so it seems
quite likely that the effect of raised pCOg on bicarbonate
reabsorption is mediated by a fall in the pH of the renal
tubule cells. Apart from this rather striking change in
bicarbonate excretion the output of electrolytes is not
significantly affected by short periods of hypercapnia,
although there is a transient water diuresis (Barbour et al.,
1953).

It is not clear how far the information obtained from
studies of acute hypercapnia can be applied to the situation
of chronic hypercapnia found in emphysematous patients.
Here, a steady state has been established at a new level of
plasma pH and bicarbonate concentration. The electrolyte
composition of plasma and red cells in emphysematous
patients is different in several respects from that of normal
people in whom a comparable hypercapnia has been induced
acutely by CO2 inhalation (Plattsand Greaves, 1957). For
example, the fall in pH is much smaller in the emphysematous
patients, and the chloride content of both cells and plasma is
lower than in acute respiratory acidosis.

There are few observations on the renal response to chronic
respiratory acidosis in man. As part of a study on the effect
ofDiamox,Nadell (1953) reports observations on 24-hour speci-
mens of urine from two patients with respiratory acidosis.

266 D. A. K. Black

The mean urinary pH in these two patients was 6 • 26 and 6 • 67,
no more acid than specimens from two 'controls' with mean
pH of 6-48 and 6*20. The mean excretions of bicarbonate
were 7-5 and 15-2 m-mole/day, compared with 10-1 and
4-8 m-mole/day in controls. Ammonium excretion was some-
what higher, and titratable acidity somewhat lower in the
patients with respiratory acidosis than in the controls, and it
has been reported that renal glutaminase is increased in
experimental respiratory acidosis. There were no striking
differences in 24-hour output of sodium, potassium, or
chloride. These findings are consistent with the view that
renal adaptation has included increased synthesis of ammonia,
allowing the excretion of hydrion at a higher urine pH than
in acute respiratory acidosis, without increase in urinary
buffer (the excretion of phosphate was lower than in the
controls).

In preliminary observations on four patients with respira-
tory acidosis, my colleague Dr. J. Timoner has found a pH
range in urine of 5 • 1 to 6-7, with ammonium excretion up to
65 [jL-equiv./min. and titratable acidity up to 60 [ji-equiv./min.
After a standard load of ammonium chloride (0-1 g./kg. body
weight), two patients excreted 76-5 and 81-3 [ji-equiv. of
ammonia, and 26 • 3 and 46 • 2 [x-equiv. of titratable acid per
minute. The ammonium excretion is just above the normal
range found by Davies and Wrong (1957). These two patients
were aged 57 and 60, and seem to have retained the capacity
of the renal tubule cells to form ammonia in response to an
acid stimulus.

Renal function in cor pulmonale. In the cardiac failure
associated with emphysema, the cardiac output is commonly
increased, and the patient has warm extremities. Terminally,
the limbs become cold, the blood pressure falls, and the
cardiac output at this stage is reduced. Davies and Kil-
patrick (1951) showed that even in the high-output phase of
cor pulmonale the circulation through the kidneys and the
glomerular filtration rate were substantially diminished.
These findings have been confirmed by Lewis and his co-

Renal Function in Respiratory Failure 267

workers (1952). A moderate degree of urea retention, pre-
sumably on the basis of relative renal ischaemia, is common
in cor pulmonale (Simpson, 1957), as in other forms of heart
failure. In patients dying from heart failure, the output of
urine may be reduced to below 500 ml. /day, but complete
suppression of urine does not seem to have been recorded,
even in the terminal stages. It is perhaps of some interest,
therefore, that over the past ten years we have seen two
patients, both with cor pulmonale, who became anuric
(Black and Stanbury, 1958). One of them, a girl of 20 with
widespread bronchiectasis and a terminal bronchopneumonia,
had an eight-day period of extreme oliguria, during which
her blood urea rose to 158 mg./lOO ml. She was treated
conservatively, urine was again formed, and the blood urea
fell to 76 mg./lOO ml. She continued to pass considerable
amounts of dilute urine until her death a week after the end
of the anuric period. The second patient, a man of 44, passed
no urine for over 24 hours, and had no urine in his bladder
after death. Both these patients had hypotension and cold
extremities, and were presumably in the low-output phase of
cor pulmonale; but cardiac output could not of course be
measured. Both of them had central cyanosis, but only the
second had a raised pCOg in the plasma. The main factor in
causing anuria was probably renal ischaemia, but this may
have been aggravated by arterial desaturation.

Both these patients had hyperkalaemia and low plasma
sodium. This association is fairly common in patients with
acute renal failure, but we have seen it also in the absence of
renal failure and it may possibly represent a loss of potassium
from cells, with partial replacement by sodium.

These observations in patients with terminal cor pulmonale
are possibly of little more than academic interest; but they
perhaps constitute yet another argument for the early treat-
ment of intercurrent infections in patients with emphysema ;
such intercurrent infections may be apyrexial, and attended
by little apparent reaction, but they can precipitate the
patient into terminal low-output failure.

268 D. A. K. Black

REFERENCES

Barbour, A., Bull, G. M., Evans, B. M., Hughes Jones, N. C, and

LoGOTHETOPOULOS, J. (1953). CHn. Sci., 12, 1.
Black, D. A. K., and Stanbury, S. W. (1958). Brit. med. J., 1, 872.
CoMROE, J. H., FoRSTER, R. E., DuBOis, A. B., Briscoe, W. A., and

Carlsen, E. (1955). The Lung. Chicago: Year Book Publishers.
Davies, C. E., and Kilpatrick, J. A. (1951). Clin. Sci., 10, 53.
Davies, H. E. F., and Wrong, O. (1957). Lancet, 2, 625.
Lewis, C. S., Samuels, A. J., Daines, M. C, and Hecht, H. H. (1952).

Circulation, 6, 874.
Nadell, J. (1953). J. din. Invest., 32, 622.
Pitts, R. F. (1953). Harvey Lect., 48, 172.

Platts, M. M., and Greaves, M. S. (1957). Clin. Sci., 16, 695.
Sheldon, J. H. (1948). The Social Medicine of Old Age. Oxford

University Press.
Shock, N. W. (1952). In Cowdry's Problems of Ageing, p. 614, 3rd

ed., ed. Lansing, A. I. Baltimore: Williams & Wilkins.
Simpson, T. (1957). Lancet, 2, 105.

DISCUSSION

Milne : I am not convinced. Dr. Black, that the anuria you mentioned
in your two cases is in any way related to the chronic respiratory disease.
During the last influenza epidemic in this country some cases of anuria
were associated with influenza. I know of one case in Dundee and we
ourselves have personally studied three cases. Two of those we saw
recovered and one died. The one that died showed typical acute tubular
necrosis; the other two showed a clinical course typical of tubular
necrosis. None of these patients gave any sign of chronic respiratory
disease. They were typical Asiatic influenza cases, as shown by the epi-
demiology and serum tests, developing in previously healthy individuals ;
one case was uncomplicated and two cases were complicated by a secon-
dary staphylococcal pneumonia. A severe respiratory infection of itself
in some cases seems to be able to precipitate anuria, and I myself prefer
to relate your experience to infection rather than to the biochemical
changes of chronic respiratory acidosis.

My other point is a personal protest : I have a tremendous respect for
the work of Dr. Pitts and his colleagues, but I do think we should avoid
adopting this term, 'bicarbonate-bound base'. To the chemist bicar-
bonate is a hydrogen ion acceptor and therefore is a base itself. Bicar-
bonate is the base; bicarbonate-bound base to me is meaningless.

Black: In quoting from Pitts, I used his terminology, but I do not
accept responsibility for it.

When you say infection, do you mean infection leading to a fall in
cardiac output and renal vasoconstriction, or do you mean an infection
of the kidney?

Milne : No, certainly not an infection of the kidney. All I am stress-
ing is that these cases occurred in young adults without any evidence

Discussion 269

whatsoever of chronic respiratory disease, and that a severe respiratory
infection, for some reason that I do not know, may cause acute tubular
necrosis, for which there is autopsy proof in one case.

Black : This would really bring it into the whole group of peripheral
circulatory changes.

McCance: This seems to me a matter which is wide open to experi-
mental attack, and it might be coupled with stress tests.

Davson : The trouble is that the energy required for these active trans-
port processes is a small fraction of the whole and when the energy sup-
plies are interfered with to such an extent that active transport is affected,
the cell will be dead long before you can obtain any useful information.

Borst: We have just had an autopsy on a very obese patient who died
with bilateral cortical necrosis. I am ashamed to say that she had been
under-examined. As in Dr. Black's cases she was admitted with a
respiratory infection which was treated with penicillin, and in a few days
the infection was under control. She was up and about until we dis-
covered that she was producing no urine. On autopsy no abnormality
in the lungs was found. The necrosis involved a great part of the renal
cortex; there was no evidence of other renal disease. We thought that
it was a case of Pickwick's syndrome.

It was reported about 20 years ago that giving oxygen to patients with
respiratory failure resulted in an increased sodium output. In our cases
there was no definite effect on sodium output in spite of the fact that the
general condition of some of the patients improved markedly.

Bull: I was hoping that Dr. Black was going to bring evidence of a
normal decline in respiratory function, because in our patients both
renal and respiratory deaths are common, and there are many cases of
the combination of the two. If someone could show that respiratory
function declined in roughly the same way that renal function does that
would help us to understand this situation. I believe that tissues other
than the kidney must undergo a similar decline in function at the same
sort of rate with age to account for this rather remarkable mortality
experienced. We have now confirmed our findings on over 3,000 cases,
and we get exactly the same effect as we did eight years ago.

Black : There are indeed plenty of references to the decline of respira-
tory function with age. A summary has been given by Stuart-Harris and
Hanley (1957. Chronic Bronchitis, Emphysema, and Cor Pulmonale.
Bristol: Wright & Son).

Scrihner: As regards renal compensation, we had one patient with a
remarkable ability to compensate for respiratory acidosis. We were
interested in finding out whether high pCOg or low pH caused the coma-
like condition that patients with respiratory acidosis may develop when
treated with oxygen. Our interest began when we tried treating acute
renal failure by putting a cellophan bag in the stomach, a technique first
suggested by Dr. Schloerb of Kansas City. With this technique of gastro-
dialysis it is possible to remove tremendous amounts of hydrogen ion,
in fact usually so much that you have to put hydrochloric acid in the
dialysis fluid to prevent alkalosis in the patient. We turned this around
and applied it therapeutically to the respiratory acidosis patients in an

270 Discussion

attempt to corapensate them artificially by getting their serum bicar-
bonate levels up. We treated a 50-year-old man with acute respiratory
acidosis whose initial bicarbonate figure was 40 m-equiv./l. and the
blood pH, breathing room air, about 7 • 28. When he went into oxygen
he became unconscious rather quickly, presumably due to the decrease
in ventilation from the relief of anoxia. He was removed from oxygen
and over the next 18 hours dialysed through a cellophan bag in his
stomach, using a fluid containing 50 m-equiv./l. sodium bicarbonate
in 5 per cent glucose. The dialysis elevated his serum bicarbonate to
64 m-equiv./l. despite a negative sodium balance of 200 m-equiv. The
sodium was lost mainly in the urine. The high serum bicarbonate ele-
vated his blood pH, breathing room air, to 7-55. When he again went
into oxygen his blood pH fell to 7 • 45 and he did not become unconscious.
His anoxia disappeared despite the fact that his ventilatory rate slowed
from 9 litres per minute to 3 litres per minute. During the next 72 hours
his kidneys sustained his serum bicarbonate level above 60 m-equiv./l.
by excreting a normal amount of ammonia and titratable acidity.

Experience in this patient suggests that so-called "CO 2 narcosis" is
actually due to the low pH rather than the high pCOg. The results also
suggest that despite the high serum bicarbonate renal compensation for
the respiratory acidosis may be incomplete in this acute situation.
Gastrodialysis makes it possible to treat the acidosis without resorting
to sodium administration, which is contraindicated because of the heart
failure from cor pulmonale.

WATER AND ELECTROLYTE METABOLISM
IN CONGESTIVE FAILURE

Z. Fejfar

Institute for Cardiovascular Research^
Prague — Krc

The role of the kidney in congestive failure

The genesis of abnormal water and electrolyte metabolism
in congestive failure is at present generally attributed to im-
paired renal function. It was previously thought that in-
creased systemic venous pressure (and hence the imbalance of
Starling forces in the capillaries) was the main factor initiating
these phenomena. Warren and Stead (1944) observed in some
cardiac patients an increase in body weight after the adminis-
tration of salt before any significant rise in central venous
pressure. This indicated that another mechanism might be
responsible for the retention of salt and water in chronic
congestive failure. Merrill (1946) confirmed the earlier findings
of Seymour and co-workers (1942) that patients with con-
gestive failure have a diminished renal blood flow; moreover
he found that the decrease in renal blood flow was far greater
than the diminution of cardiac output.

It was, however, not clear whether the retention of electro-
lytes and water in chronic congestive heart failure was due to
a primary decrease in renal function or to the decrease in
renal blood flow and function as a consequence of the increase
in central venous pressure.

It appeared to us in 1947 (see Brod and Fejfar, 1949, 1950)
that only observations of haemodynamic events at the time
when water balance was changing could elucidate this problem.
Patients with heart disease on the borderline of right heart
failure usually have a low urine output during the day, but
an increased urine flow at night. This spontaneous diuTcsis

271

272 Z. Fejfar

reflects a temporary improvement of the impaired water
balance. It runs its course within a few hours. It was there-
fore possible to follow the sequence of events and investigate
the relationship between central venous pressure, systemic
and renal haemodynamic changes, and renal function.

Cardiac output, right auricular pressure, water content of
plasma, and renal function (renal blood flow, glomerular
filtration rate and excretion of electrolytes) were studied
from the early hours of the afternoon until the following
morning in ten normal subjects and 25 patients with heart
disease of different origin, 19 of them having congestive failure
of varying degree (Brod and Fejfar, 1949, 1950; Fejfar and
Brod, 1950a,b,d).

Cardiac output was measured by a direct Fick method and
right auricular pressure by a water manometer attached to
the cardiac catheter ; changes of water content in plasma were
assessed from the percentage change in plasma proteins,
haematocrit and the disappearance curve of Evans blue.
Renal plasma flow was estimated by the clearance of PAH
(j9-aminohippuric acid), glomerular filtration rate by the clear-
ance of inulin, and chlorides by the Van Slyke and Hiller
(1947) modification of Sendroy's method.

A nocturnal diuresis was observed in 11 patients with con-
gestive failure. In none of them was it preceded by a decrease
in right auricular pressure. On the other hand the increase in
urine output at night started in all these patients with an
elevation in renal blood flow. The decrease in urine flow at
night occurred in seven decompensated cardiacs; in all of
them it was associated with a diminution in renal blood flow
(Fig. 1). The increase in renal blood flow was not related to a
similar change in cardiac output, which increased simultane-
ously in only half of the investigated subjects.

There is thus evidence in dynamic observations that the
increase in central venous pressure in congestive failure is not
the primary cause of cardiac oedema, the main factor being
impaired renal function.

A low renal blood flow with a diminished glomerular

Water and Electrolytes in Congestive Failure 273

filtration rate and increased tubular reabsorption of electro-
lytes was also found in patients with left-sided failure and with
mitral stenosis without any clinical evidence of right-sided
decompensation, the central venous pressure being normal
(Fejfar and Brod, 1949; Blegen and Aas, 1950; Werko et al.,
1952a; Himbert et al, 1954; Werko et al, 1955).

^%40

Decompensated cardiac

./A

cases
O Number of observations 11-

-%40

~ \

2 1

— hps.-

1 2

- +hrs.

Fig. 1. Composite diagram showing percentage changes (A%)
in renal blood flow (Clp^n) from the level at 0 hrs (time at which
the urine flow began to change) in decompensated cardiacs.
In patients with no change in urine flow, 0 hrs was fixed arbi-
trarily at 7 p.m. (1) are patients with a nocturnal increase in urine
flow, (2) are patients in which the urine flow decreased at night,
while in (3) it did not change. See text for details. (Brod, J.,
and Fejfar, Z. (1950). Quart. J. Med., 19, 187.)

Fig. 2 presents the individual values of renal blood flow
in normal subjects and in patients with heart diseases. All
patients are divided into five groups according to the clinical
degree of heart failure.

In the first group are clinically compensated patients. The
second group includes patients with a slight to moderate dys-
pnoea on effort ; in the third are those with marked dyspnoea
on effort, orthopnoea or attacks of nocturnal dyspnoea and
acute pulmonary oedema. The fourth group covers patients
with signs of right-sided decompensation who responded well

274

Z. Fejfar

to digitalis, and in the fifth group are patients refractory to
the usual methods of treatment.

It may be seen that patients without right-sided failure
have a decreased renal blood flow in comparison with the
values in normal control subjects.

On the other hand increase of pressure in the renal vein
brought about by a partial occlusion (Selkurt, Hall and

2000-

•

1800

/600-

1200

;..

,.*

dOO^

60O

400-

200

) 2 3 A 5

1 2 3 AXS

12 3 4 5

L2 34 5

n A»

Fig. 2. Renal blood flow in normal subjects and in patients with rheumatic

(Rm and Rao)j hypertensive (H), ischaemic (I) and pulmonary (P) heart

disease. All patients are divided into five groups according to the clinical

degree of heart failure. See text for details.

Spencer, 1949), or by an increased abdominal pressure (Brad-
ley and Bradley, 1947), is followed by only a small diminution
of the renal blood flow.

Maxwell, Breed and Schwartz (1950) measured pressure in
the inferior vena cava in 17 healthy subjects and ten patients
with congestive failure. The mean pressure in healthy subjects
was 15-2 cm. HgO, and in patients with congestive failure
27 cm. HgO. From the measured values of pressure they
calculated that the increase of renal resistance due to the
elevation of pressure in renal veins would reduce renal blood
flow by about 14 per cent. The actual decrease in renal blood
flow in congestive failure is far greater (see Fig. 2).

Water and Electrolytes in Congestive Failure 275

Farber and co-workers (1951, 1953) studied in man the
effect of an increase of pressure in the vena cava produced by
means of a balloon above and below the orifice of the renal
veins. In both procedures there was a diminution of renal
blood flow, glomerular filtration rate and excretion of water
and electrolytes.

The increased central venous pressure in congestive failure
may, of course, contribute to reduction in renal function
(Briggs et al., 1948; Bradley and Blake, 1949; Earle et ah,
1949). It determines the distribution of retained water and
electrolytes, which in left-sided failure is in the lungs and in
congestive failure mainly in the lower part of the body.

The nature of renal changes in congestive failure.

The nocturnal increase of diuresis and renal blood flow in
our investigated patients with congestive failure was also
associated with an elevation of glomerular filtration rate and
with a decrease in tubular reabsorption of water and electro-
lytes. This may be seen in Fig. 3, which covers 20 spon-
taneous changes in urine flow in 14 patients with congestive
failure. The lower urine output was always taken as the
initial value (100 per cent).

The mean increase in diuresis was 187 per cent (range from
44 to 672 percent). This increase was associated in all instances
(as seen in Fig. 1) with an elevation in renal blood flow. This
latter increased on the average by 55-5 per cent (from 6 to
146 per cent). Only three times was the increase in renal
blood flow smaller than 20 per cent. In 14 subjects in whom it
was measured cardiac output (CO) rose significantly in six
instances, fell in three and did not change in five. It is clear
that the increase in renal blood flow could not depend on the
primary increase in CO. This is confirmed by an increase in
the renal fraction of cardiac output in all instances except
one, in which the renal fraction did not change.

Glomerular filtration rate at high urine flow was elevated
15 times, and unchanged five times. The average increase
was 27 • 1 per cent, with the range — 4 • 5 to + 82 per cent.

276

Z. Fejfar

The elevation of renal blood flow was effected in the great
majority by a decrease in postglomerular resistance, the
filtration fraction diminishing 17 times and increasing in only
three instances.

The increase in chloride clearance was of the same order as

1000

900

eoo

TOO-
600
500
iOO

1/ rn rjr^.., or rt PT rt^.. —SS.

CO Ctp^ RF Ct,„, FF Ctct'

Ct.r

► 55" ♦^S *27 -16 *2B *!57

Fig. 3. Percentage nocturnal changes in urine flow
(V), cardiac output (CO), renal plasma flow (ClpAg),
renal fraction of cardiac output (RF), glomerular
filtration rate (Cljn), filtration fraction (FF),
chloride clearance (Clcr) and in the ratio of chloride

clearance to glomerular filtration rate ( -^^ — I • The

mean percentage change (^) and range (□) in 14
patients with heart failure are presented. The
lower urine output was taken as 100 %.

the elevation of urine volume. The mean increase was 215
per cent, range 2 to 910 per cent. The ratio of chloride clear-
ance to glomerular filtration rate rose on an average by 157
per cent (range —17-6 to +530 per cent).

Water and Electrolytes in Congestive Failure 277

According to Wesson, Anslow and Smith (1948) some 85
per cent of the filtered sodium and chloride is reabsorbed by
an active mechanism in the proximal tubule, irrespective of
the amount filtered. The reabsorption of the remaining 15 per
cent of sodium and chloride is limited by a fixed maximal
rate at which the distal tubular cells are able to reabsorb
these electrolytes. Whenever the tubular chloride load
decreases with a fall in glomerular filtration rate in the
presence of this maximal reabsorption capacity, almost all
of the filtered chloride is reabsorbed. Merrill (1949), Mokotoff,
Ross and Leiter (1948), Selkurt, Hall and Spencer (1949),
Stead (1951) and others are of the opinion that in congestive
failure this mechanism leads to the maximum reabsorption of
electrolytes and water; that is to say that the diminution of
glomerular filtration is such that with a normal unchanged
tubular reabsorption, water and electrolytes are retained.

Our results are not in accord with the hypothesis of Wesson,
Anslow and Smith. In patients with severe congestive failure,
glomerular filtration rate did not rise towards normal levels
at the time of nocturnal diuresis ; in spite of this, the amount
of excreted chloride was far greater than the quantity of
chloride excreted at night in healthy subjects with a normal
glomerular filtration rate. Fig. 4 demonstrates that the
tubular reabsorption of chloride can vary markedly with a
constant tubular chloride load. It is clear, of course, that at a
given chloride load less chloride is reabsorbed at a high than
at a low urine flow.

The concentration of chloride in urine exceeded its plasma
level in only seven out of 24 observations at high urine flow.
The increased urine flow, therefore, cannot be explained on
osmotic grounds by an increased excretion of chloride.

The lower elimination of electrolytes and water in conges-
tive failure is, according to these findings, not caused only by
decreased glomerular filtration rate. Tubular reabsorption of
water and electrolytes increases as well. The same conclusion
is stated by Briggs and co-workers (1948), Kattus and co-
workers (1948), Davis and Shock (1949), Newman (1949),,

278

Z. Fejfar

Himbert and co-workers (1954), Cort (19556), Cort and
Fencl (1957), and others.

Doyle and Merrill (1957) studied renal function in 18
patients with congestive failure in a supine position and tilted
in a passive erect posture. The changes were qualitatively
similar to those in normal subjects. There was a further
depression of renal plasma flow, glomerular filtration rate and
also a decreased urine flow and a fall in the excretion of the

100%
99

•

fT.

,«*a^^

*^-:

,^ ^^

i^=V

^^^

■>rt^

:^-0-

::::

-----

:---=©

1 1

■•■i2

\ •

.u^^_

r-?^

5?:

ri^

«t^

"\^

95%

«fe

xClinul?'°°

1 -^ 1 1

fj LOW URINE FLOW m

LOW URINE, FLOW %

1 > iPci.Clirn

5 |66 17

j|. mM

8 ,9

17 18

Fig. 4. Relationship of the amount of the chloride filtered (Pd X Clinul.) and

^ X 100 ) in individual subjects at high and low urine

Values in individual subjects are connected with dotted lines.
N — normal control subjects ; C — patients with heart disease. See text for

details.

reabsorbed ( ^ ^,

VPci X Clinul
flows

electrolytes. In accord with our previous findings, with
nocturia the decreased urine flow in the erect posture was
closely correlated with changes in renal plasma flow. There
was, on the other hand, a very poor correlation between
changes in glomerular filtration rate and sodium excretion.

In these observations there was an indirect relationship
between the tubular reabsorption of electrolytes and water
and renal blood flow. The tubular reabsorption increased
when renal blood flow fell and vice versa.

This finding does not characterize congestive failure alone.

Water and Electrolytes in Congestive Failure 279

Bucht and co-workers (1953) studied the haemodynamic
changes together with the excretion of sodium in eight healthy
human subjects during muscular exercise of varying degree.
As long as the effort was small (oxygen consumption not
above 500 ml./niin.), an increase of CO was found without
significant effect on renal blood flow, glomerular filtration
rate or excretion of sodium. A greater muscular effort
(oxygen consumption about 1,000 ml./min.) was character-
ized by a marked increase in CO (almost double) and a
simultaneous fall in renal blood flow and the renal fraction of
CO. The excretion of sodium and water fell. Glomerular
filtration rate and pressure in renal veins did not change
significantly. Similar results were observed in patients with
heart disease (Judson et al., 1955; Himbert, Scebat and
Theard, 1956). Increase of tubular reabsorption was there-
fore responsible for the diminished excretion of sodium and
water.

The close relationship between renal blood flow and excre-
tion of electrolytes in congestive failure is striking. We have
expressed the opinion (Brod and Fejfar, 1950) that decreased
renal blood flow directly impairs the excretion of water and
electrolytes. A smaller glomerular filtration rate diminishes
tubular electrolyte load and, owing to a slower flow of tubular
urine, a greater proportion of the filtered amount is reabsorbed.
We could not, of course, exclude another possibility: that
increased reabsorption of water and electrolytes in the renal
tubules could occur parallel with, but independently of the
diminished renal plasma flow; i.e. the stimulus for the renal
vasoconstriction could directly influence the function of renal
tubules, leading to an increased reabsorption of salt and water.

Humoral and neural regulatory mechanisms in
congestive failure.

Some known humoral and neural factors can alter the
function of renal tubules. In the urine of patients with con-
gestive failure renin (Merrifl, Morrison and Brannon, 1946),
VEM (vaso-excitor material) and VDM (vasodepressor

280 Z. Fejfar

material) have been found (Edelman et al., 1950). Extracts
of urine from patients with congestive failure contain anti-
diuretic materal (Bercu, Rokaw and Massie, 1949, 1950) with
a great sodium-retaining activity (Deming and Luetscher,
1950a,b), which disappears when the patients become com-
pensated (Luetscher, Deming and Johnson, 1950, 1951).
The substance responsible for this is aldosterone (Luetscher
and Johnson, 1954). An increased excretion of aldosterone is
not characteristic only of congestive failure, but accompanies
nephrotic and cirrhotic oedema as well. A permanent increase
of aldosterone under these conditions is called secondary
aldosteronism (Conn, 1955; Bartter, 1956; Milne and
Muehrcke, 1956; Thorn et a/., 1956; Liddle, Duncan and
Bartter, 1956; Wolff, Koczorek and Buchborn, 1957).

The increased secretion of aldosterone in congestive failure
may be important in some patients, as can be seen from the
favourable effect of bilateral adrenalectomy (Thorn et al.,
1956).

Buchborn (1956) estimated the activity of plasma anti-
diuretic hormone (ADH) by a sensitive biological method on
the toad, together with serum osmolarity. He found a close
indirect correlation between the plasma ADH and serum
osmolarity in 14 normal subjects, in patients with hepatic
cirrhosis, in compensated cardiac patients, and also in patients
with congestive failure. The increased plasma level of ADH
in congestive failure is not therefore primary, being an ex-
pression of the homeostatic function of ADH, regulating
osmotic pressure in the organism (Buchborn, 1956).

Neither ADH nor aldosterone significantly influences cir-
culation in the kidneys. Their main effect is on renal tubules,
where they increase the reabsorption of water (ADH), or
sodium (aldosterone). In addition we have already indicated
that the vasoconstriction in the kidneys, together with
diminished elimination of sodium, occurs during a short
muscular effort (10 minutes, Bucht et al, 1953). The effect of
aldosterone would be slower. According to Bartter (1956) the
excretion of sodium in a patient with Addison's disease did

Water and Electrolytes in Congestive Failure 281

not start to fall until more than an hour after intravenous
injection of 40 [ig. aldosterone.

It would appear to us, therefore, that neither of these
humoral substances is the primary cause of the retention of
salt and water in heart failure.

The results of haemodynamic changes in human subjects
following intravenous injection of Dibenamine called our
attention to the importance of reflex (neurohumoral) regula-
tion in the genesis of haemodynamic changes in congestive
failure.

Blockade of adrenergic impulses by Dibenamine in patients
with heart failure caused a diminution of a high peripheral
vascular resistance and central venous pressure. Cardiac
output increased. Renal blood flow rose in a great majority
of investigated subjects, suggesting that this was independent
of the increase in CO. These changes were not produced by
blocking the adrenergic impulses in the heart or by an in-
creased secretion of adrenaline (Fejfar and Brod, 1950c,
1951, 1954; Brod, Fejfar and Fejfarova, 1951, 1954) (Fig. 5).
The increase in renal blood flow in seven out of nine patients
in congestive failure was accompanied by a rise in urine flow
and an increased elimination of sodium or chloride.

We were able to conclude from our results that, with the
onset of congestive failure, reflex (neurohumoral) vasocon-
striction develops in both arterial and venous circulation.
The function of this selective vasoconstriction may be to
secure a sufficient supply of oxygenated blood to working
tissues such as the heart and other muscles.

A haemodynamic pattern resembling chronic heart failure
(i.e. unequal distribution of blood supply to various organs,
increased utilization of oxygen in tissues, and an insufficient
CO) may also be found in clinical circumstances with a
diminished return of venous blood to the heart (e.g. mitral
stenosis, constrictive pericarditis), or when the amount of
circulating blood and oxygen decreases, as well as in acute
heart failure or peripheral circulatory failure (see Fejfar,
1958). A similar haemodynamic picture can be seen in severe

282

Z. Fejfar

differs from that
an increase in CO and by vaso-

muscular effort in healthy subjects.

found in heart failure by

dilatation in the skin due to increased temperature.

Haemodynamic changes in heart failure therefore do not
represent a new and special adaptation of the organism to the

Fig. 5. Changes in cardiac output (CO), peripheral
vascular resistance (TPR), blood pressure (P), right
auricular pressure (RAP) and renal plasma flow
(PAH) after Dibenamine in a subject with heart
failure. See text for details. (Fejfar, Z. (1957).
Acta cardiol. (Brux.), 12, 13.)

diminishing performance of the heart. They are a typical
reaction which appears in every situation in which CO is
inadequate for oxygen requirement in the tissues. This
reaction becomes a chronic feature during the development of
congestive failure and leads to retention of water and sodium.

Water and Electrolytes in Congestive Failure 283

A high central venous pressure and a secondary excretion of
humoral substances like aldosterone complicate the response.

Werko and co-workers (1955), in a study of systemic and
renal haemodynamic changes in 146 subjects with different
cardiac disorders, came to a similar conclusion. Their results
suggest that "the adrenergic impulses could contribute to
the diminished renal blood flow in severe heart disease before
any signs of congestion are apparent". They think one of the
factors causing the release of adrenergic impulses may be a
decreased stroke volume.

The origin of the afferent impulses of this functional
haemodynamic reflex is not known. There are, of course,
several pieces of evidence on the influence of nervous impulses
on diuresis. Viar and co-workers (1951) demonstrated an
increase in urine flow and excretion of sodium as the result of
a rising venous pressure in the head (following the compres-
sion of neck by a manometer cuff). Cort (1953), in agreement
with these results, found an increased diuresis with higher
elimination of sodium in subjects with the head lowered
(Trendelenburg position of 15°). The changes in renal blood
flow were not reported. Cathcart and Williams (1955) did not
confirm this.

Gauer and co-workers (1954) described an increase in urine
flow in anaesthetized dogs during the negative pressure
breathing period. This was also found in healthy human
subjects (Sicker, Gauer and Henry, 1952, 1954). The rise in
diuresis was not accompanied by increased elimination of
electrolytes (Na+ or K+). This water diuresis was thought to
be caused by stimulation of volume or stretch receptors
localized in the cardiovascular system in the thorax (left
atrium or pulmonary veins). The values of renal plasma flow
were not measured in these experiments. We do not know,
therefore, if the changes reported were produced by a direct
influence on the renal tubules without any change in renal
haemodynamics .

It is also difficult to use these findings to explain the electro-
lyte and water imbalance in heart failure. We have produced

284 Z. Fejfar

evidence (see above) that renal blood flow and a decreased
excretion of electrolytes occurs in left ventricular failure and
mitral stenosis without right-sided decompensation, when
there is an increased pressure in the venous side of the pulmon-
ary circulation.

On occasion, however, a sudden increase of pressure in this
part of the pulmonary circulation may be associated with a
rise of urine flow in patients with a heart disease. We have
followed haemodynamic changes in nine patients with acute
pulmonary oedema (Fejfar et al., 1958a); in three of them we
also studied renal haemodynamics and the excretion of
electrolytes. At the onset of recovery from pulmonary
oedema there was a depressed renal blood flow and the renal
fraction of CO started to increase before any significant changes
in cardiac output occurred. In two of these three patients the
rise in renal blood flow was accompanied by an increased
excretion of chloride (Fig. 6). A rise of pressure in the left
auricle and pulmonary veins is typical for acute pulmonary
oedema in patients with mitral stenosis or left ventricular
failure. It is therefore possible that this elevation of pressure
could influence renal blood flow, diuresis, and the excretion
of electrolytes. The diuresis was not, however, a water
diuresis as described by Sicker, Gauer and Henry (1952,
1954).

Gomori and co-workers (1954) studied renal circulation in
dogs with crossed circulation under hypoxaemia. They
found a decrease in renal blood flow in a dog whose head was
perfused from the other body by hypoxic (venous) blood.
Following denervation of the kidneys, this vasoconstriction
either disappeared completely or was insignificant.

Foldi and co-workers (1955) found in hypoxaemic dogs a
decrease in renal blood flow, excretion of water and electro-
lytes. In healthy subjects breathing a mixture of 10 per cent
oxygen there was also a decreased renal blood flow and elimi-
nation of electrolytes. On the other hand a low renal blood
flow, glomerular filtration rate and excretion of sodium
significantly increased in patients with congestive heart

Water and Electrolytes in Congestive Failure 285

BPmmHg.
200

Util.0,%

'^ "" /5 -^^ 16 "•' 17 -^^ Idhrs:
Fig. 6. Haemodynamic changes and renal excretion of chloride in a
patient with acute pulmonary oedema. BP— blood pressure ; F— pulse
irequency ; Util. Og— oxygen utilization in tissues (in percentage of 0„
supply); O2 cons.— oxygen consumption/min. ; CO— cardiac output;
Jsat. O2— arterial (A) and mixed venous (V) oxygen saturation as a
percentage; RBF— renal blood flow; RF— renal fraction of cardiac
output; V— urine flow in ml./min. ; Clci— chloride clearance ; P atr— right
auricular pressure; D— dyspnoea; C— cough; R— rales.

286 Z. Fejfar

failure inhaling 50 per cent oxygen plus 4 per cent carbon
dioxide for 30 minutes (Foldi et ah, 1956). According to these
authors renal changes are brought about by hypoxia in the
brain.

It is improbable, however, that every case of heart failure
is accompanied by cerebral hypoxia. The renal changes are
manifested, as shown above, in left-sided failure. The results
of Scheinberg (1950) indicate a decreased blood flow through
the brain in heart failure together with a rise in cerebral
vascular resistance. If the cerebral supply of oxygen is really
insufficient, we might expect quite the reverse: a diminution
of cerebral vascular resistance and an increase in cerebral
blood flow. This was actually demonstrated in man during
experimental hypoxaemia by Kety and Schmidt (1948).

We are of the opinion that the heart itself may be the
starting point for the haemodynamic functional changes
in heart failure, and in all situations in which CO is inade-
quate for the requirement in tissues, i.e. where oxygen utiliza-
tion in tissues increases (Fejfar, 1956, 1957, 1958). The basis
for this hypothesis will be briefly summarized:

(a) Myocardial utilization of oxygen is, even with physical
inactivity in healthy subjects, greater than that by the other
important organs of the body. Every rise in oxygen con-
sumption or utilization in tissues (muscular effort, anaemia,
mitral stenosis, etc.) is associated with coronary vasodilation,
an increase in the coronary fraction of CO, and vasoconstric-
tion in the kidneys.

(b) We have demonstrated that during the inhalation of
oxygen a normal CO in a healthy subject, or in compensated
patients, either does not change or decreases, while a low
cardiac output in heart failure increases (Fejfar, 1957;
Fejfar et al., 1958a).

(c) Gomori and co-workers (1954), in experiments cited above,
did not find an elevation of CO during isolated hypoxia of the
brain. On the other hand, when the isolated head of a dog
was perfused by arterial blood and the trunk supplied with
hypoxaemic blood (the dogs inhaled a mixture with a low

Water and Electrolytes in Congestive Failure 287

concentration of oxygen), CO rose in a similar way to the rise
observed in hypoxaemic hypoxia in intact animals,
(d) Harrison and co-workers (1927) concluded from their
studies on experimental hypoxaemia in dogs that the oxygen
tension in the myocardium is the most important factor
determining the rise in CO.

A direct efferent nervous influence on the kidneys was
demonstrated by Kaplan and Rapoport (1951) and Blake
(1952) in dogs with unilateral renal denervation. Tubular
reabsorption of sodium was less in the denervated kidney.
Bykov and Alexejev-Berkmann (1930, 1931) (see Bykov,
1952) found that a conditioned "water" diuresis in dogs may
be partly inhibited by denervation of the kidneys.

Renal blood flow was measured only in the experiments of
Kaplan and Rapoport (1951), where the increased renal
excretion of water and electrolytes after splanchnicotomy was
independent of changes in renal blood flow. Our experimental
results in patients with heart failure (see above) demonstrated
a close relationship between changes in renal blood flow and
tubular reabsorption of water and electrolytes.

A partial answer to this question can be found in the experi-
ments of Cort and Kleinzeller (1956) on isolated kidney
tissues of rabbits. Changes in transport of cations and water
were studied during two hours' exposure of kidney slices to
unoxygenated physiological saline at 0°, and then after 10
and 30 minutes of incubation in Krebs' phosphate saline with
oxygen at 25°. One kidney was decapsulated and denervated
14 days before the actual experiment. It was shown that there
was a greater influx of sodium into the denervated slices
during leaching at 0°, and a slowcjr expulsion of sodium from
the denervated kidney slices during the incubation period.
The changes in water content of the slices were in the same
direction as the shifts of sodium. The difl'erence between
denervated and innervated kidney was, however, not marked.
Potassium loss during the two-hour leaching period was
greater, and its reaccumulation during subsequent incubation
slower, in the denervated kidney.

288 Z. Fejfar

In six rabbits with bilateral denervation the resting clear-
ances of inulin and PAH were practically the same as in the
rabbits without renal denervation (Brod and Sirota, 1949).
Cort and Kleinzeller (1956) therefore conclude that the dif-
ferences described are due to a direct nervous effect on
tubular cells rather than to a change in renal blood flow.

It is difficult to compare results obtained from experiments
with tissue slices or in anaesthetized animals, with results
from human subjects, in which every disturbance of homeo-
stasis is immediately compensated for in several ways.
Neural and humoral regulation act simultaneously and it is
practically impossible to differentiate them. It seems,
nevertheless, that even in subjects with chronic heart failure,
retention of electrolytes and water is the result of haemo-
dynamic changes parallel with increased tubular reabsorption
of sodium and water. These changes may be initiated by a
reflex mechanism acting through adrenergic nerves. Increased
secretion of aldosterone and ADH is a secondary manifesta-
tion. This secondary aldosteronism may, however, prevail in
the long run, dominate the whole picture of chronic congestive
failure, and close the vicious circle.

Further consequences of retention of salt and
vs^ater in heart failure.

The retained sodium and water in congestive failure does
not enlarge the volume of extracellular fluid only. In patients
recovering from heart failure the reduction of body weight
was greater than the reduction in the amount of extracellular
fluid (Seymour et al., 1942), chloride output (Schroeder, 1950)
or sodium loss (Miller, 1950, 1951). This surplus water must
come from cells. In the development of congestive failure,
the water accumulates in both extracellular and intracellular
compartments.

At the same time changes begin in the concentration of
extracellular and intracellular electrolytes. The loss of cellular
potassium in congestive failure was described in 1930 by
Harrison, Pilcher and Ewing. It has been ascertained by

Water and Electrolytes in Congestive Failure 289

balance studies, and by analyses of muscle biopsies, that in
addition to the cellular loss of potassium there is an incre-
ment of sodium in cells (Iseri, Boyle and Myers, 1950; Iseri
et aL, 1952; Squires, Crosley and Elkinton, 1951a; Warner
et aL, 1952; Cort and Matthews, 1954; see also Elkinton and
Danowski, 1955; Cort and Fencl, 1957). Particularly im-
portant is the fact that potassium depletion occurs in subjects
treated by repeated injections of mercurial diuretics (Squires
et al., 19516; Cort and Matthews, 1954). In some of these
severely ill cases hyponatraemia and hypochloraemia with an
elevated concentration of bicarbonate may be observed.

Clinical diagnosis of potassium depletion in chronic conges-
tive failure is difficult to prove. Decompensated cardiacs
excrete negligible amounts of sodium and the stronger acid
radicals are excreted neutralized by potassium. Therefore
the typical finding of a far higher concentration of potassium
than sodium in the urine in congestive failure is not alone
sufficient proof of cellular loss of potassium.

Plasma levels of Na+, K+, and Cl~ are usually within the
normal range in decompensated cardiac patients.

Table I presents the relationship between plasma levels
of Na+, K+ and HCOa" and concentration of Na+ and K+ in
muscle biopsy specimens in 13 patients with various degrees
of heart failure. Concentrations of total muscle Na+ and K+
are expressed in m-equiv. 100 g. of fat-free dry solids (FFDS).
Normal values given by Cort (1955b) are about 13 i 2 m-equiv.
of Na+ and 45 ± 3 m-equiv. of K+.

It will be seen that all the patients had a decreased amount
of potassium in skeletal muscle. This K+ depletion was very
marked, although not all were treated with mercurial diu-
retics. Patient M.E. was not yet in right-sided failure. In all
patients, with the exception of A.Z., the plasma concentra-
tions of Na+ and K+ were within the normal range. In
the majority the concentration of bicarbonate was slightly
elevated. None of them showxd ECG changes typical of
potassium depletion.

The lowest figure of muscle potassium (11 '2 m-equiv./lOOg.)

AGEING — IV— 10

290

Z. Fejfar

Table I

Relationship between plasma levels of Na+, K+ and HCOj"

AND CONCENTRATION OF Na+ and K+ IN THE SKELETAL MUSCLE

MS — mitral stenosis; MI — mitral incompetence; Tri S — tricuspid stenosis;

Tri ins. — tricuspid insufficiency; H. + I.H.D. — hypertensive and ischaemic

heart disease. See details in text.

Name

Sex

Diagnosis

Age
years

Degree
of heart
failure

Muscle

Plasma

Note

Na+ K +

total total

m-equiv./

100 g.

FFDS

Na+ K + ECO;>~
m-equiv. 11.

E.5.

MS>MI
TriS

10-23

18-8

137

4-47

28-3

A.S.

Atr. sept,
def.

3-4

19-75

27-6

141

5-15

29-4

0 mercurial
diuretic

M.B.

13-9

28-3

150

4-54

28-0

M.E.

MI>MS

15-04

28-64

137

4-5

26-4

0 merciu-ial
diuretic

I.D.

MS>MI
postcommis.

23-39

21-52

145

5-75

28-1

0 mercurial
diuretic

M.D.

MI, bacterial
endocarditis

12-51

29-92

143

3-9

30-2

0 mercurial
diuretic

E.K.

MS, Tri S.

14-3

37-2

131

5-34

28-5

P.U.

MS, Tri ins.

27-1

25-46

145-1

4-56

28-1

A.Z.

MS,
postcommis.

22-5

11-2

126-5

4-02

14-8

8th day post-
operative

A.V.

10-3

21-76

148-5

5-2

31-1

H.Ch.

MS,
postcommis.

19-57

33-61

143-3

4-97

28-5

V.B.

H.+I.H.D.

16-91

39-88

143-5

4-98

29-4*

* not at the
same time

M.V.

MS,
postcommis.

20-78

34 06

141-5

4-44

26-8

was found in patient A.Z., with suppuration in the thoracic
wound one week after mitral commissurotomy, 24 hours
before death. He was by this time in severe metabohc acidosis.
The loss of about three-quarters of the muscle potassium was

Water and Electrolytes in Congestive Failure 291

probably not just a consequence of postoperative suppura-
tion; it must already have been present before the operation.

Experiences with two other patients with mitral stenosis
and congestive failure, who died within a week after operation
with a picture of combined peripheral and cardiac failure, led
us to the conclusion that a greater operative risk with mitral
commissurotomy in patients with congestive failure (group IV
in the usual classification) is associated with potassium
depletion and intracellular acidosis with increased retention
of sodium (Fejfar et al., 1958a).

Negative nitrogen balance following surgical operations is
connected with potassium depletion (Moore and Ball, 1952),
and it is clear that in patients with potassium depletion in
chronic congestive failure a further loss of potassium after
operation brings about various complications (shock, acute
heart failure, infection, slow recovery, etc.).

It follows that the laboratory diagnosis of potassium
depletion in chronic congestive failure is not easy to make. A
low serum concentration of Na+, as an indirect indicator, is
present only in very advanced stages. One should suspect
potassium depletion if there is a decrease of serum chloride
and a rise in HCOg" accompanying the usual urinary pattern
in heart failure (negligible concentration of Na+ and a marked
excretion of K+).

Analysis of a muscle biopsy specimen or balance studies,
which, together with measurement of total exchangeable K+,
are at present the only methods for detecting early stages of a
metabolic imbalance of electrolytes, are both rather compli-
cated for practical use.

It is therefore more useful to assume potassium depletion
in every patient with chronic congestive failure. The treat-
ment of every patient should be supplemented by a diet rich
in potassium. In more severe cases potassium salts are useful,
being particularly important in all patients treated with
mercurial diuretics. Cort (1955c) demonstrated in 12 patients
with congestive failure that potassium chloride, given some
days before the injection of mercury, potentiated its diuretic

292 Z. Fejfar

effect more than ammoniuin chloride and simultaneously
compensated the potential loss of potassium. As the loss of
potassium from the cells is probably connected with a break-
down of cellular glycogen and protein, it is advantageous to
add N hormones (methylandrostendiol) to the treatment.

It is not easy to correct completely a severe potassium
deficiency in chronic congestive failure. Even with a high
potassium intake it may be several weeks before cells become
saturated (Cort and Matthews, 1954).

There remain many unanswered questions. It is customary
to treat patients with congestive failure with a low sodium
diet. It has been shown, however, that a low sodium diet in
healthy subjects increases aldosterone excretion in the urine
(Luetscher and Axelrad, 1954; Liddle, Duncan and Bartter,
1956; Wolff et at., 1956a, h), while a diet rich in sodium has
led to a decrease of aldosterone activity in the urine (Luet-
scher and Curtis, 1955a, h\ Gordon, 1955; Bartter et al., 1956;
Garrod, Simpson and Tait, 1956).

Potassium administration also increases the excretion of
aldosterone (Laragh and Stoerk, 1955; Luetscher and Curtis,
1955a, b; Falbriard et ah, 1955; Bartter et al, 1956).

Laragh and Stoerk (1957) recently demonstrated that no
sodium-retaining activity was found in the urinary extracts
from dogs on a diet low in both sodium and potassium. When
the amount of potassium was increased, hyperkalaemia
developed and sodium-retaining activity appeared in the urine.
Similar results were observed in one patient suffering from
rheumatic heart disease with congestive failure. As long as he
was kept on a diet low in sodium (about 12 m-equiv. daily)
and a rather high potassium intake (140 m-equiv.), the
excretion of aldosterone was high (about 300 (JLg./24 hr.).
After the marked reduction of serum potassium to 2*7
m-equiv. by an injection of 2 ml. of Mercuhydrine together
with a low potassium diet, the excretion of aldosterone fell
to 35 [jLg. Restoration of a normal serum potassium level by
administration of potassium was again followed by a very
marked excretion of aldosterone in the urine (630 (xg./24 hr.).

Water and Electrolytes in Congestive Failure 293

During the whole course, the serum sodium level did not
change significantly. Laragh and Stoerk (1957) concluded
from these results that the higher serum potassium level is
probably a stimulus for the secretion of aldosterone.

If patients with heart failure respond to a low sodium and
high potassium intake in the same way as normal subjects,
our customary therapeutic procedure would assist in the
creation of secondary aldosteronism.

Reduction of body water increases the excretion of aldo-
sterone in normal subjects (Luetscher, Deming and Johnson,
1951, 1952; Beck et al, 1955; Falbriard et al, 1955; Bartter
et al., 1956; Garrod, Simpson and Tait, 1956). When the
volume of extracellular fluid rises, the urinary elimination of
aldosterone diminishes (Beck et al., 1955; Liddle et ah, 1955;
Muller, Riondel and Mach, 1956).

In patients with congestive failure and other oedematous
states there is on the contrary an expanded extracellular fluid
volume associated with a rise in the urinary excretion of
aldosterone. The explanation of this reversed reaction is at
present difficult. Wolff, Koczorek and Buchborn (1957)
argue that in congestive failure there must be a disturbance of,
or anew regulatory mechanism for the secretion of aldosterone.

Increased elimination of aldosterone in the urine was
found in the first week following surgical intervention
(Llaurado, 1955; WolfP, Koczorek and Buchborn, 1957) or
acute myocardial infarction without signs of congestive
failure (Wolff, Koczorek and Buchborn, 1957). This may be
explained by a diminution of extracellular fluid volume.
But one must not neglect the fact that in all such stressful
situations there is a raised adrenergic activity; and the same
stimulus may perhaps also lead to an increased production of
aldosterone, irrespective of the level of extracellular fluid
volume, as seems to be the case in congestive failure.

Summary

Retention of salt and water in heart failure is caused by
disturbed renal function. The main factors are a decreased

294 Z. Fejfar

renal blood flow and an increased tubular reabsorption of salt
and water. High venous pressure in the systemic circulation
is not the primary cause of this disturbed water balance.
It may, however, contribute to it.

In congestive failure there is not merely a simple retention
of extracellular electrolytes and water. Serious metabolic
changes may also occur. Great clinical significance should be
attached to cellular potassium depletion. The laboratory
diagnosis of the latter is difficult, the best method at present
being chemical analysis of muscle biopsy specimens. One must
consider this disturbance in every patient with heart failure,
and consequently treat all such patients with sufficient
potassium in the diet, or by administering potassium salts,
particularly when mercurial diuretics are used.

Consideration was given to the significance of regulatory
mechanisms responsible for renal dysfunction in congestive
failure. The primary role of reflex changes was stressed and
the present knowledge of the role of aldosterone and ADH
was discussed.

Acknowledgements

I should like to thank Drs. J. H. Cort and A. Hlavova and Miss D.
Rosicka for carrying out the muscle biopsy analyses.

REFERENCES

AxELRAD, B. J., Johnson, B. B., and Luetscher, J. A., Jr. (1954). J. din.
Endocrin. Metab., 14, 783.

Bartter, F. C. (1956). Metabolism, 5, 369.

Bartter, F. C, Liddle, G. W., Duncan, L. E., Barber, J. K., and
Delea, C. (1956). J. din. Invest., 35, 1306.

Beck, J. C, Dyrenfurth, I., Giroud, C. J., and Venning, E. H. (1955).
Arch, intern. Med., 96, 463.

Bercu, B. a., Rokaw, S. N., and Massie, E. (1949). J. Lab. din. Med.,
74, 1585.

Bercu, B. A., Rokaw, S. N., and Massie, E. (1950). Circulation, 2, 409.

Blake, W. D. (1952). J. din. Invest., 31, 618.

Bland, J. H. (1956). Clinical Recognition and Management of Dis-
turbances of Body Fluids. 2nd ed. Philadelphia: W. B. Saunders.

Blegen, E., and Aas, K. (1950). Acta med. scand., 138, 391.

Bradley, S. E., and Blake, W. D. (1949). Amer. J. Med., 6, 470.

Bradley, S. E., and Bradley, G. P. (1947). J. din. Invest., 26, 1010.

Water and Electrolytes in Congestive Failure 295

Briggs, a. p., Fowell, D. M., Hamilton, W. F., Remington, J. W.,

Wheeler, N. C, and Winslow, J. A. (1948). J. din. Invest., 27,

810.
Brod, J., and Fejfar, Z. (1949). Cos. Lek. ces., 88, 991.
Brod, J., and Fejfar, Z. (1950). Quart. J. Med., 19, 187.
Brod, J., Fejfar, Z., and Fejfarova, M. H. (1951). Shorn, lek., 53,

128.
Brod., J., Fejfar, Z., and Fejfarova, M. H. (1954). Acta med. scand.j

148, 273.
Brod, J., and Sirota, J. H. (1949). Amer. J. Physiol., 157, 31.
Buchborn, E. (1956). Klin. Wschr., 34, 953.
BucHT, H., Ek, J,, Eliasch, H., Holmgren, A., Josephson, B., and

Werko, L. (1953). Acta physiol. scand., 28, 95.
Bykov, K. M. (1952). Mozkova Kura a Vnitfni Organy. Praha: SZN.
Bykov, K. M., and Alexejev-Berkmann, I. A. (1930). Pfliig. Arch.

ges. Physiol, 224, 710.
Bykov, K. M., and Alexejev-Berkmann, I. A. (1931). Pfliig. Arch,

ges. Physiol., 227, 301.
Cathcart, E. S., and Williams, I. T. D. (1955). Clin. Sci., 14, 121.
Conn, J. W. (1955). J. Lab. din. Med., 45, 3.
CORT, J. H. (1953). J. Physiol., 122, 22P.
CoRT, J. H. (1955a). Physiol. Bohemoslov., 4, 14.
CoRT, J. H. (19555). Acta med. Acad. Sci. hung., 8, 347.
CoRT, J. H. (1955c). Cas. Lek. ces., 94, 244.

CoRT, J. H., and Fencl, V. (1957). The Body Fluids. Praha: SZN.
CoRT, J. H., and Kleinzeller, A. (1956). J. Physiol., 133, 287.
CoRT, J. H., and Matthews, H. L. (1954). Lancet, 1, 1202.
Davis, J. O., and Shock, N. W. (1949). J. din. Invest., 28, 1459.
Deming, Q. B., and Luetscher, J. A., Jr. (1950a). J. din. Invest., 29,

808.
Deming, Q. B., and Luetscher, J. A., Jr. (19506). Proc. Soc. exp,

Biol., N.Y., 73, 171.
Doyle, A. E., and Merrill, J. M. (1957). Clin. Sci., 16, 155.
Earle, D. p., Farber, S. J., Alexander, J. D., and Eichna, L. W.

(1949). J. din. Invest., 28, 778.
Edelman, I. S., ZwEiFACH, B. W., Escher, D. J. W., Grossman, R.,

MoKOTOFF, R., Weston, R. E., Leiter, L., and Shorr, E. (1950).

J. din. Invest., 29, 925.
Elkinton, J. R., and Danowski, T. S. (1955). The Body Fluids. Balti-
more: Williams & Wilkins.
Falbriard, a., Muller, A. F., NeJier, R, and Mach, R. S. (1955).

Schweiz. med. Wschr., 85, 1218.
Farber, S. J., Alexander, J. D., and Eichna, L. W. (1951). J. din.

Invest, 30, 638.
Farber, S. J., Becker, W. H., and Eichna, L. W. (1953). J. din.

Invest., 32, 1145.
Farber, S. J., and Soberman, J. (1953). J. din. Invest., 32, 566.
Fejfar, Z. (1956). II Europ. Congr. Cardiol. Abstracts, p. 30.
Fejfar, Z. (1957). Acta cardiol. (Brux.), 12, 13.

296 Z. Fejfar

Fejfar, Z. (1958). Acta cardiol. (Brux.), in press.

Fejfar, Z., Bergmann, K., Dejdar, R., Fejfarova, M., Honsova, H.,

Keszler, H., Spacer, B., and Valach, A. (1958a). Klinicko-

fysiologicka Studie se Zamefenim k Chirurgicke L6cbe Mitralni

Stenosy. Praha: SZN.
Fejfar, Z., and Brod, J. (1949). Cas. Lek. i^es., 88, 1352.
Fejfar, Z., and Brod, J. (1950«). Quart. J. Med., 19, 221.
Fejfar, Z., and Brod, J. (19506). Cas. Lek. ces., 89, 151.
Fejfar, Z., and Brod, J. (1950c). I Int. Congr. Cardiol., No. 6, p. 47.
Fejfar, Z., and Brod, J. (1950rf). / Int. Congr. Cardiol., No. 67, p. 186.
Fejfar, Z., and Brod, J. (1951). Shorn, lek., 53, 99.
Fejfar, Z., and Brod, J. (1954). Acta med. scand., 148, 273.
Fejfar, Z., Fejfarova, M., Bergmann, K., and Brod, J. (19586). Ill

World Congr. Cardiol., to be published.
Foldi, M., Kovach, a. G. B., Takacs, L., and Koltay, E. (1955). Acta

med. Acad. Sci. hung., 8, 19.
Foldi, M., Solti, F., Koltay, E., Megyesi, K., Rev, J., and Szasz, J.

(1956). Klin. Wschr., 34, 857.
Garrod, O., Simpson, S. A., and Tait, J. F. (1956). Proc. R. Soc. Med.,

49, 888.
Gauer, O. H., Henry, J. P., Sieker, H. O., and Wendt, W. E. (1954).

J. din. Invest., 33, 287.
Gaunt, R., Renzi, A. A., and Chart, J. J. (1955). J. din. Endocrin.

Metab., 15, 621.
GoMORi, P., Kovach, A., Takacs, L., Foldi, M., Szab(5, G., Nagy, Z.,

and Wiltner, W. (1954). Orv. Hetil., 95, 225.
GoMORi, P., and Takacs, L. (1956). Z. arztl. Fortbild., 50, 286.
Gordon, E. S. (1955). J. Lab. din. Med., 46, 820.
Hamilton, W. F. (1954). Minn. Med., 37, 36.
Harrison, T. R., Blalock, A., Pilcher, C., and Wilson, C. P. (1927).

Amer. J. Physiol., 83, 284.
Harrison, T. R., Pilcher, C, and Ewing, G. (1930). J. din. Invest.,

8, 325.
HiMBERT, J., Theard, A., Gele, P., ScEBAT, L., and Lenegre, J. (1954).

Arch. Mai. Coeur, 47, 747.
HiMBERT, J., ScEBAT, L., and Theard, A. (1956). Acta cardiol. {Brux.)^

11, 209.
ISERi, L. T., Boyle, A. J., and Myers, G. B. (1950). Amer. Heart J., 40,

706.
IsERi, L. T., Alexander, L. C, McGaughey, R. S., Boyle, A. J., and

Myers, G. B. (1952). Amer. Heart J., 43, 215.
Judson, W. E., Hollander, W., Hatcher, J. D., and Halperin, M. H.

(1955). J. din. Invest., 34, 1546.
Kaplan, S. A., and Rapoport, S. (1951). Amer. J. Physiol., 164, 175.
Kattus, a., Sinclair-Smith, B., Genest, J., and Newman, E. V.

(1948). J. din. Invest., 27, 542.
K^ty, S. S., and Schmidt, C. F. (1948). J. din. Invest., 27, 484.
Laragh, J. H., and Stoerk, H. C. (1955). J. din. Invest., 34, 913.
Laragh, J. H., and Stoerk, H. C. (1957). J. din. Invest., 36, 383.

Water and Electrolytes in Congestive Failure 297

LiDDLE, G. W., Bartter, F. C, Duncan, L. E., Barber, J. K., and

Delea, a. C. (1955). J. din. Invest., 34, 949.
LiDDLE, G. W., Duncan, L. E., and Bartter, F. C. (1956). Amer. J.

Med., 21, 380.
Llaurado, J. G. (1955). Lancet, 1, 1295.
LuETSCHER, J. A., Jr., and Axelrad, B. J. (1954). Proa. Soc. exp.

Biol., N.Y., 87, 650.
LuETSCHER, J. A., and Curtis, R. H. (1955a). Ann. intern. Med., 43, 658.
LuETSCHER, J. A., and Curtis, R. H. (19556). J. clin. Invest., 34, 950.
LuETSCHER, J. A., Jr., Deming, Q. B., and Johnson, B. B. (1950). J.

clin. Invest., 29, 1576.
LuETSCHER, J. A., Jr., Deming, Q. B., and Johnson, B. B. (1951). J.

clin. Invest., 30, 1530.
LuETSCHER, J. A., Jr., Deming, Q. B., and Johnson, B. B. (1952).

Ciha Found. Colloq. Endocrin., 4, 530. London : Churchill.
LuETSCHER, J. A., Jr., and Johnson, B. B. (1954). J. clin. Invest., 23,

1441.
]V^Iaxwell, M. H., Breed, E. S., and Schwartz, I. (1950). J. clin.

Invest., 29, 342.
Merrill, A. J. (1946). J. clin. Invest., 25, 389.
Merrill, A. J. (1949). Amer. J. Med., 6, 357.
Merrill, A. J., Morrison, J. L., and Brannon, E. S. (1946). Amer. J.

Med., 1, 468.
Miller, G. E. (1950). J. clin. Invest., 29, 835.
Miller, G. E. (1951). Circulation, 4, 270.
Milne, M. D., and Muehrcke, R. C. (1956). Proc. R. Soc, Med., 49,

883.
MoKOTOFF, R., EscHER, D. J. W., Edelman, I. S., Grossman, J., and

Leiter, L. (1949). Fed. Proc, 8, 112.
MoKOTOFF, R., Ross, G., and Leiter, L. (1948). J. clin. Invest., 27, 1.
Moore, F. D., and Ball, M. R. (1952). Metabolic Response to Surgery.

Springfield: Thomas.
MuLLER, A. F., Riondel, a. M., and IVL^ch, R. S. (1956). Lancet, 1, 831.
Newman, E. V. (1949). Amer. J. Med., 7, 490.
Scheinberg, p. (1950). Amer. J. Med., 8, 148.
ScHROEDER, H. A. (1950). Circulation, 1, 481.
Selkurt, E. W., Hall, P. E., and Spencer, M. P. (1949). Amer. J.

Physiol., 40, 157.
Seymour, W. M. B., Pritchard, W. H., Longley, L. P., and Hayman,

J. M. (1942). J. clin. Invest., 21, 229.
SiEKER, H. O., Gauer, O. H., and Henry, J. P. (1952). J. clin. Invest. ,

31, 662.
SiEKER, H. O., Gauer, O. H., and Henry, J. P. (1954). J. clin. Invest.,

33, 572.
Singer, B., and Wener, J. (1953). Amer. Heart J., 45, 795.
Squires, R. D., Crosley, A. P., and Elkinton, J. R. (1951a). Circula-
tion, 4, 868.
Squires, R. D., Singer, R. B., Moffitt, G. R., and Elkinton, J. R.

(19516). Circulation, 4, 697.

AGEING IV 11

298 Z. Fejfar

Stead, E. A. (1951). Circulation, 3, 294.

Thorn, G. W., Renold, A. E., Froesch, E. R., Crabbe, J. (1956).

Helv. med. Acta, 23, 4.
Van Slyke, D. D., and Hiller, A. (1947). J. biol. Chem., 167, 107.
ViAR, W. N., Oliver, B. B., Eisenberg, S., Lombardo, T. A., Willis,

K., and Harrison, T. R. (1951). Circulation, 3, 105.
Warner, F. G., Dobson, E. L., Rodgers, C. E., Johnston, M. E., and

Pace, N. (1952). Circulation, 5, 915.
Warren, J. V., and Stead, E. A., Jr. (1944). Arch, intern. Med., 73, 138.
Werko, L., Bucht, H., Ek, J., and Eliasch, H. (1952«). Nord. Med.,

47, 79.
Werko, L., Ek, J., Bucht, H., and Eliasch, H. (19526). Scand. J. din.

Lab. Invest., 4, 15.
Werko, L., Ek, J., Varnauskas, E., Bucht, H., Thomasson, B., and

Eliasch, H. (1955). Amer. Heart J., 49, 823.
Werko, L., Varnauskas, E., Eliasch, H., Ek, J., Bucht, H., Thomas-
son, B., and Bergstrom, J. (1954). Circulation, 9, 687, 700.
Wesson, L. G., Jr., Anslow, W. P., Jr., and Smith, H. W. (1948). Bull.

N.Y. Acad. Med., 24, 586.
Wolff, H. P., Koczorek, K. R., and Buchborn, E. (1956a). Verh.

dtsch. Ges. inn. Med., 62, 480.
Wolff, H. P., Koczorek, K. R., Buchborn, E., and Kohler, M.

(19566). Klin. Wschr., 34, 1105.
Wolff, H. P., Koczorek, K. R., and Buchborn, E. (1957). Schweiz.

med. Wschr., 87, 163.

DISCUSSION

McCance : Prof. Borst, can you bring together these discoveries about
nocturnal diuresis, reflex activity and aldosterone excretion?

Borst : The role of aldosterone should not be exaggerated. Heart failure
and nocturia can be seen in patients with Addison's disease ; therefore in
the disturbance in water and electrolyte excretion of heart failure and
of nocturia the effect of aldosterone cannot be the only factor. We
believe that the evidence is in favour of the theory that salt retention in
the presence of normal kidneys is always largely effected through the
same pathways. The same mechanism is responsible for the retention
after haemorrhage, in nephrosis, in cirrhosis and in heart failure. On the
other hand we assume that salt diuresis is also always effected through
the same pathway. The characteristics of this mechanism can best be
studied in the excellent experimental conditions provided by patients
with paroxysmal tachycardia accompanied by polyuria. The attack of
tachycardia elicits the typical 'salt diuresis', though blood volume and
extracellular fluid volume remain constant. The diuretic stimulus must
therefore result from the change in heart action. The pulse rate acutely
rises from 80 to 160 and after a certain period falls suddenly to the original
rate. The consecutive portions of urine in patients who are on a standard-
ized diet show a brisk water diuresis followed by a gradual increase in
sodium output. The excretion pattern is very characteristic and is in

Discussion 299

every respect similar to that following the rapid intravenous injection of
saline. These facts point to a dependence of the sodium and water output
on blood pressure or on blood flow ; there is no direct relation to volume.
A fact worth remarking is that the diuresis may continue several hours
after the tachycardia stops. This suggests that the effect of the abnormal
circulation on the renal tubules is mediated by a slowly acting mecha-
nism, possibly a renal hormone. Experiments in animals in which the
functions of the two kidneys have been compared also prove that the
adrenal is not essential and that the receptor must be in the kidney.
When one renal artery is gradually narrowed the sodium and water
excretion of the corresponding kidney may fall sharply before a fall in
PAH and creatinine clearance can be demonstrated. Probably the kidney
responds even to the slightest reduction in intrarenal blood pressure by
an increased tubular reabsorption of sodium chloride and water.

Fejfar: I quite agree with you in all points. It is also my personal
view that this reaction might start in the heart itself. In all these types
of circulatory disturbances (mitral stenosis, pericarditis, acute heart
failure, hypoxaemia, anaemia), and in muscular effort, the only common
factor is a very low oxygen content in the central venous blood. When
we gave oxygen to patients with normal cardiac output, the cardiac
output did not change. When oxygen was given to patients with a
lowered cardiac output, the output increased ; there is, therefore, indirect
evidence that if more oxygen is given to the heart muscle in congestive
failure the performance of the heart improves.

Milne: Have you any observations on a similar correlation, or the
reverse, in other conditions besides congestive heart failure associated
with nocturnal diuresis? In starvation and cortisone overdosage, parti-
cularly, a similar reversal of the normal diurnal rhythm may be seen.

Fejfar: No, I have no comments to make.

Milne : I would agree with all the points you make regarding the diag-
nosis of potassium deficiency in heart failure, but I think that most
clinicians are now using a very useful clinical method of diagnosis — excess
sensitivity to digitalis. Of course, as you say, it can be checked by
balance or exchangeable potassium if necessary.

Fejfar : You are right about digitalis, but of course this usually occurs
in advanced stages of potassium deficiency. When patients are in potas-
sium deficiency it takes weeks and weeks to restore the balance. This is
not just an academic question because we had three deaths due to these
metabolic changes shortly after mitral valvulotomy. When a patient
already has a negative balance with loss of potassium, the added opera-
tive trauma and hypotension will easily lead to so-called metabolic death.

Olesen: I have had the opportunity of studying the problem of the
diagnosis of potassium depletion in congestive heart failure with the
dilution methods used in Boston (McMurrey et al. (1956). Metabolism,
5, 447). I would say first that the diagnosis is not very easy; in fact it is
probably impossible to make it by the dilution methods alone. We
found, however, that there were very marked changes in the body
composition of these patients with congestive failure. There was a rela-
tive decrease in the total intracellular mass, as expressed either by total

300 Discussion

intracellular water or total intracellular potassium. This is a change
which may also be seen in severe weight loss without congestive failure,
and the situation is very difficult to evaluate because patients with con-
gestive failure will often have lost weight in the late stages. An inter-
esting finding was that although there were almost equal degrees of
congestive failure the average intracellular potassium concentration ap-
peared normal in the males but was low in the females. We have no
explanation for this finding.

The question to us, however, is whether a low average intracellular
potassium concentration means a reduction in the relative amount of
potassium or too much water in the cells. We cannot answer this. In
tissue analysis results we are faced with the same question : when there
is a low intracellular potassium concentration related to the intracellular
water, is there too little potassium or too much water? The relationship
of potassium to nitrogen or phosphorus does not seem to change very
much. This might suggest that it is as much an increase in water as it is a
decrease of potassium in the cells.

There are conflicting opinions on the balance studies. Most American
studies demonstrate a positive potassium balance during recovery from
congestive failure. However, most of these studies have been carried out
on low sodium/high potassium intake, and the high potassium intake may
explain the positive potassium balance. In a study made in Switzerland
a medium-sized intake of potassium was used and no positive potassium
balance during recovery from congestive failure was seen.

Milne : There seem to me to be two sides to this question of assessing
the cause of secondary aldosteronism in relation to the expansion and
contraction of body fluids. There is the physiological stimulus in haemor-
rhage, shock, etc., where, as you say, there is contraction ; and there is the
pathological stimulus in the nephrotic syndrome, cardiac failure, and
hepatic cirrhosis, where there is expansion. All this is really tied up with
the philosophy of volume receptors. It always seems to me to be im-
possible for the body to have a true volume receptor. The only way we
know of measuring volume is to pour fluid into a graduated cylinder. I
feel the only possible explanation is that the body is relating tension to
volume, and that the receptors are tension receptors for either static or
pulsatile tension. I think the stimulus is the same in all forms of secon-
dary aldosteronism and that the receptors must be on the arterial side of
the circulation.

Fejfar: I agree with you about volume and stretch receptors. I would
like to add that if one gives sodium to patients with congestive failure, the
aldosterone excretion decreases (Gordon, 1955); these people therefore
react in the same way as normal persons, although their actual levels of
aldosterone may be higher.

A CASE OF MAGNESIUM DEFICIENCY

W. I. Card and I. N. Marks

Gastro-intestinal Unit, Western General Hospital, Edinburgh

Our knowledge of the effects of magnesium deficiency in
man is so meagre that we feel warranted in presenting the
data from a single case and, though these data are not as
complete as one would wish, we believe they are sufficient to
allow useful though tentative conclusions to be drawn.

The state of magnesium deficiency in animals whether
experimentally produced or occurring as a natural state has
been recognized for some time (Kruse, Orent and McCollum,
1932; Greenberg and Tufts, 1938). In animals such as cows
the syndrome goes under various names (Blaxter, Rook and
McDonald, 1954); it can be cured by the injection of mag-
nesium salts and prevented by using magnesite dressings on
the pasture. In man there seems to be no clearly recognized
picture. There have been reports of various states associated
with lowered blood magnesium which have responded to
magnesium sulphate injections, and it is recognized that
various excitable states such as delirium tremens may be
associated with a low serum magnesium and may improve
with magnesium therapy (Flink et al., 1954; Martin, Mehl
and Wertman, 1952). A case described as tetany and associ-
ated with low blood magnesium has been reported in a child
(Miller, 1944).

Such observations are not wholly satisfactory since the
fraction of magnesium which exists in the plasma is so minute
that it must necessarily be a very imperfect reflection of the
state of magnesium in the body. The only satisfactory evid-
ence for a magnesium deficiency is clearly some measure of
the actual body store of magnesium. Fitzgerald and Fourman
(1956) have shown how very difficult it is in man, owing to

301

302 W. I. Card and I. N. Marks

the conserving action of the kidney, to deplete the body of
magnesium to any serious extent by taking a diet low in
magnesium. The opportunity occurred to us some four years
ago of treating a patient with an ileal fistula from which
extensive fluid and electrolyte losses occurred, and in whom a
magnesium-deficient state ultimately appeared.

For the purposes of this paper the precise clinical details
are irrelevant; it is sufficient to say that the patient was a
woman aged 34, suffering from ulcerative colitis, who had had
performed a proctocolectomy with ileostomy. The immediate
postoperative course was satisfactory but it became necessary
to refashion the ileostomy a fortnight later, and this was
followed by intestinal obstruction for which a further opera-
tion was performed. An ileal fistula then developed. Such a
fistula results in large fluid and electrolyte losses.

It is not of course possible in clinical practice to measure
electrolyte balances on all patients postoperatively, but it is
clearly necessary to have sufficient knowledge of their losses in
order to replace them effectively. The routine ward procedure,
which was followed in this case, is as follows:

A fluid balance chart is kept on which the amounts of all
fluids given orally and by intravenous infusion are noted, as
well as all losses whether urinary, faecal, by aspiration or by
any other route. In patients such as this woman, where the
intake of food is important, the food taken is recorded on a
slip of paper, so that the dietitian may make some estimate
of caloric or protein intake. From the fluid balance chart,
with, if necessary, the estimation of electrolytes in any
aspirated fluid, the necessary amounts of fluid, water, sodium,
chloride, and potassium, are prescribed for the next 12 or 24
hours. Serum electrolyte concentrations are measured, daily
if necessary, as in this case.

This procedure was carried out with this patient so that she
was kept in water, sodium, potassium, and chloride balance.
The CO 2 combining power remained within normal limits.
There was no rise in her blood urea and judging by the urinary
specific gravity reached the kidneys functioned well. Calcium

A Case of Magnesium Deficiency

303

gluconate was given intravenously but in insufficient amounts,
and in retrospect it is clear that she was in negative calcium
balance. No thought was given at this time to the possibility
or the significance of any magnesium loss.

In such an ill patient adequate nutrition and the replace-
ment of protein is very difficult to achieve and her oral food
intake was augmented by intravenous feeding. The fluids

INTAKE
(litres)^

URINE
OUTPUT

ILEAL FISTULA — WATER BALANCE

A.M. 1954

INTRAVENOUS
ORAL

PlTl

^^^^^P^^^^^m^^^^^^X^^^TTl^d^kT^^^^

I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
5- DAY PEROOS

Fig. 1. Chart showing fluid intake and urinary output over a five-month
period, with the appearance of symptoms one month after the onset.

given were glucose solutions, sodium lactate, and alcohol,
while a casein hydrolysate supplied nitrogen. Loss of blood
was replaced by blood transfusions. Despite all these measures
she undoubtedly lost weight.

Fig. 1 shows the extent of the fluid replacement necessary
over nearly five months, plotted in five-day periods, and it will
be seen that the losses were very great. At their maximum,
calculation shows that the fistula losses were of the order of
five litres a day. Since the patient at this time weighed less

304 W. I. Card and I. N. Marks

than 35 kg. she was losing the equivalent of about 15 per cent
of her body weight daily through the fistula.

The patient during this time was, of course, extremely ill
with consistently rapid pulse and occasional fever. Towards
the end of a month, however, an entirely new symptomato-
logy appeared. It was noticed that the patient became
excitable, apprehensive, and required doses of sedatives some
three or four times what would ordinarily be adequate. It
was indeed difficult to procure sleep. This excitable mental
state was an entirely new clinical picture to us and we finally
wondered whether it might not be due to magnesium de-
ficiency. Signs of tetany, in the sense of peripheral neuro-
muscular irritability, were lacking. An electrocardiogram was
within normal limits. Her serum calcium was 8-1 mg. per
cent.

Arrangements were therefore made for serum magnesium
estimations and magnesium sulphate was given intravenously.
In 24-48 hours the state of the patient altered very consider-
ably, the excitement disappeared and the ordinary doses of
sedative were able to induce sleep. Magnesium therapy was
therefore continued to repair the deficit, and balance studies
were started and continued for some three weeks. All mag-
nesium therapy was given intravenously and the magnesium
ingested orally was not increased. This is important in the
light of subsequent calculations.

Table I shows how the deficit prior to the institution of
therapy was calculated. It should be made clear that the loss
of fluid by fistula could not be measured directly, since a
complete collection was quite impossible. It was calculated
as follows : —

Fistula fluid loss = (Oral + Intravenous) Intake +

Metabolic water — (Urinary output + Extrarenal loss).

Calculated in this way the total volume of fistula loss over
the period was 109-4 litres. The magnesium content of the
fistula fluid before therapy was started was never measured.
We have therefore made the assumption that intravenous

A Case of Magnesium Deficiency

305

magnesium therapy does not alter the output of faecal
magnesium (McCance and Widdowson, 1939) and that this is
also true of the magnesium content of ileal fluid. If this
assumption is true, then we can calculate the magnesium
content before therapy by measuring it in the fistulous fluid
after therapy had started. On 18 days a sample of ileal fluid
was measured and the mean magnesium concentration was

Table I
Magnesium deficiency — A.M.

18 AprU-19 May, 1954.
Volume of fistula loss = (Oral + Intravenous) Intake + Metabolic water
— (Urinary output + Extrarenal loss)
= 109-4 1.

Magnesium loss

Fistula = 109-4 x 4-1 = 447 m-equiv.
Urinary = 19-4 x ? 1 = 19 m-equiv.

Magnesium intake

Oral = 105 m-equiv.
Intravenous = 15 m-equiv.

Total = 466 m-equiv.

Total =120 m-equiv.
Balance = —346 m-equiv.

Body weight 17.4.54 = 34 kg.

less fat 7% = 31-6 kg.
Body Mg at onset = 31-6 X -45 = 14-2 g. = 1180 m-equiv.

Deficit = 29 %

4-1 m-equiv. /I. The total loss of magnesium through the
fistula can now be calculated and is 447 m-equiv.

The urinary loss of magnesium cannot be measured in this
way since the infusion of magnesium salts has been reported
to increase the amount put out by the kidney (McCance
and Widdowson, 1939) and this was certainly true in this
patient. Since the kidney was functioning well as judged by
its concentrating power, the urinary concentration in the
period before symptoms occurred probably never rose above
1 m-equiv./l. This gives a total urinary loss of 19 m-equiv.

306 W. I. Card and I. N. Marks

The food intake of the patient over this period was small
and at times negUgible. The magnesium content of the food
taken has been calculated from food tables and amounts to
105 m-equiv. She had no drugs containing magnesium and
no toothpaste was used. Of the intravenous fluids given
none appeared to contain magnesium. The makers (Bengers)
kindly sent us an analysis of the casein hydrolysate (Casydrol)
given which contained only negligible amounts of magnesium.
The only magnesium given intravenously was that given in
whole blood. The total negative balance over this period
therefore amounted to some 346 m-equiv.

The weight of the patient at the beginning of the period was
34 kg. and, if we assume that the body at this stage contained
7 per cent fat, the total magnesium content of the body
according to the data of Widdowson, McCance and Spray
(1951) was 14-2 g. or 1,180 m-equiv. The patient therefore
over this period lost something like 25-30 per cent of her total
body magnesium. This calculation makes the assumption
that she was normal at the onset, but it is quite possible that
she was already depleted since she had had an ileostomy for a
month with an episode of intestinal obstruction needing
suction and fluid replacement.

The balance studies which followed the institution of
therapy are shown in Fig. 2. The magnesium content of a
sample of the fistulous fluid and of the urine was estimated
daily and the output of magnesium calculated as described.
The serum magnesium was estimated every few days.

The results show that with the therapy, the patient passed
into positive balance over this period and that in all she
retained some 279 m-equiv. of magnesium before the observa-
tions were discontinued. The results are in general accord
with the previous conclusions.

The serum magnesium showed a low figure at the time
of symptoms and rose with therapy but the estimations
are perhaps chiefly of value in emphasizing how little use
can be made of them as an index of magnesium deficit in
the body.

A Case of Magnesium Deficiency

307

MAGNESIUM DEFICIENCY AM.
20-5-54-II-6-54

2 5

i.E<^/L.

i£5^M MAG

t
SYMPTOMS

mEq

INTAKE

LOSS

300
200

+ 39

+ 13

lOO

■ URINARY
♦65 = 279 m Eq

20MAY
- 24 MAY

25 MAY
-30MAY

3J MAY
5JUNE

6JUNE
IIJUNE

Fig. 2. Chart showing the effect of magnesium therapy in

producing a positive magnesium balance, and its effect on the

serum magnesium.

Discussion

When first seen the symptomatology of the patient in this
state was extremely puzzling. The clinical picture was quite
unusual and something we had not encountered before. The
patient was apprehensive, "on edge", and proved extremely
difficult to sedate. She was very ill at the time and there may
well have been earlier manifestations which passed unnoticed.
The animal behaviour as described by Greenberg and Tufts
(1938) in rats, and in particular the apprehensive state de-
scribed in induced magnesium deficiency in calves by Blaxter,
Rook and MacDonald (1954), strongly recall the clinical
picture we saw. Magnesium deficiency in man may ultimately
proceed to a condition of tetany and even convulsions as it

308 W. I. Card and I. N. Marks

does in animals, but the state we observed bore no resemblance
to low calcium tetany as seen clinically.

The other point worth discussing is the level of depletion at
which these symptoms appeared. It seems likely from this
one case, and we have failed to find a comparable example in
the literature, that symptoms of what might be called moder-
ate severity appeared when something like 25-30 per cent
depletion of the total body magnesium had occurred. If we
may adduce evidence from animal experimental work,
Blaxter, Rook and MacDonald (1954) calculated that in
calves on magnesium-deficient diets symptoms appeared when
a deficit of about 25-30 per cent magnesium had occurred,
while at death it was estimated that 35 per cent of the magnes-
ium in the body was lacking. If this general conclusion is
true, it follows that the small deficits of 50-100 m-equiv.,
which have been described by various authors (Nabarro,
Spencer and Stowers, 1952), are unlikely to produce clinical
manifestations and in themselves hardly call for treatment.
In man, the conditions necessary to produce magnesium
depletion sufficiently severe to result in a recognizable clinical
state are unusual and can hardly be expected to occur with
any frequency.

REFERENCES

Blaxter, K. L., Rook, J. A. F., and MacDonald, A. M. (1954). J.

comp. Path., 64, 157.
Fitzgerald, M. G., and Fourman, P. (1956). Clin. Sci., 15, 635.
Flink, E. B., Schutzman, F. L., Anderson, A. R., Koonig, T., and

Eraser, R. (1954). J. Lab. din. Med., 43, 169.
Greenberg, D. M., and Tufts, E. V. (1938). Amer. J. Physiol., 121,

416.
Kruse, H. D., Orent, E. R., and McCollum, E. B. (1932). J. hiol.

Chem., 96, 519.
McCance, R. a., and Widdowson, E. M. (1939). Biochem. J., 33, 523.
Martin, H. E., Mehl, J., and Wertman, M. (1952). Med. Clin. N. Amer.,

36, 1157.
Miller, J. F. (1944). Amer. J. Dis. Child., 67, 117.
Nabarro, J. D. N., Spencer, A. G., and Stowers, J. M. (1952). Quart.

J. Med., 21, 225.
Widdowson, E. M., McCance, R. A., and Spray, C. M. (1951). Clin.

Sci., 10, 113.

Discussion 309

DISCUSSION

Fourman: When Dr. Fitzgerald and I started to produce an experi-
mental depletion of magnesium we had in mind to do what I had done
with potassium (1956. Clin. Sci., 15, 635). But we got nowhere near a
significant depletion; only some 70 m-equiv. of magnesium were lost
from the body in the course of a month's efforts. Afterwards we realized
that this was partly because the urinary and faecal losses became very
small when the intake was low.

Duckworth, Godden and Warnock (1940. Biochem. J., 34, 87) found
that the magnesium of bone makes up one-half of the body magnesium.
This forms a mobilizable store, which is probably why it is so difficult to
produce symptoms of a deficiency of magnesium (Blaxter, K. L., Rook,
J.A.F.,andMcDonald, A.M. (1954). J.comp.Path.,64Ao7). A depletion
of magnesium seems to bear little relation to what is called a clinical
magnesium deficiency by some workers, who have attributed the condi-
tion of tremors in patients with alcoholism to a low serum magnesium
(Flink et al. (1957). Ann. intern. Med., 47, 956). The plasma magnesium
must depend on more than the stores of magnesium in the body.

Dr. Card, what were the urinary losses of magnesium when you gave
the intravenous injections of magnesium? In our experiments, even with
the small deficits we had, we found that the urinary losses after injection
were less than when the subjects had no deficit.

Card : I have not got the figures for the amount of magnesium in the
urine in the early days of treatment. When you give intravenous mag-
nesium some does come through the urine, but these amounts were
variable (McCance and Widdowson, 1939). The lowest magnesium we
have ever got, without magnesium therapy, was down to 1 m-equiv. /I.,
and we have taken that as the concentration of the urine prior to mag-
nesium therapy. Even that may be too high when a patient is in a
deficient state.

Fourman: It would be very convincing if the injection of magnesium
produced little rise in the urinary magnesium, while in normal people it is
known to produce a large and prompt rise in the urinary excretion of
magnesium.

Davson: McCance established that the concentration of magnesium
in the cerebrospinal fluid was considerably higher than that in the blood
plasma. It may be that it is necessary to have a high concentration
surrounding the nerve cells to maintain a low level of excitability, in
much the same way as there is a low ^concentration of potassium which
also decreases with excitability.

Card: In the experiments where the calves ultimately died, with a big
deficit, the tissue magnesium was normal. The whole deficiency appears
to occur in the bones, and I think that, as Dr. Fourman suggested, there
must be states in which the magnesium is not available. There is one
example of magnesium tetany in the literature which is obviously not a
case of deficiency, in a child with osteochondritis ; so there may be bone
diseases in which this interchange is impossible, and acute states in
which magnesium deficiency can occur, entirely different from this
chronic deficiency loss. Greenberg and Tufts (1938) went to a good deal

310 Discussion

of trouble to find out which part of the brain was particularly affected ;
they thought it was the mid-brain, and pointed out various differences
from low-calcium tetanus.

Black : Was there any tremor in your patient, and what was the state
of the reflexes?

Card: There was no obvious tremor, but of course she was extremely
ill. She had a very rapid pulse, up to 160, which may have been partly
due to magnesium deficiency as the animals showed that too. The deep
reflexes were probably gone, but they might have gone in any case.

McCance: What do you mean by 'gone in any case', when the mag-
nesium deficiency was raising the excitability?

Card: I simply mean that in a patient in this extremely wasted state,
with very little muscle tissue remaining, we may not be able to elicit
reflexes, quite apart from any electrolyte disturbance. We did an ECG
and it was normal.

Hingerty : Were there any noticeable symptoms of muscular dysfunc-
tion when the plasma magnesium was above normal. Dr. Card? In ani-
mal experiments we tried to reproduce some of the symptoms of adrenal
insufficiency by raising the plasma magnesium by injecting magnesium
sulphate. When we got the plasma magnesium and muscle magnesium
up to the level seen in adrenal insufficiency, we got very similar disturb-
ances in the levels of the hexose esters, phosphocreatine and adenosine
triphosphate (Hingerty, D. J. (1957). Biochem. J., 66, 429).

Card: Again, she was extremely ill, and I would say there was nothing
detectable. Only gross changes in the clinical state would have been
noticed. I would repeat, the clinical condition itself was most striking.

CONCLUDING REMARKS

Adolph: It is easier, I find, to mention some of the things we have
omitted in this colloquium than to dwell on some of the things that
we have gone into. We are all concerned with studies of regulation,
some of us as observers of normal individuals and some of us by
trying to cut in on the mediators by administering hormones. Per-
haps the most important element in metabolic events, particularly in
respect to water and electrolytes, may be the detection by the body
and the cells themselves of departures from the normal. In other
words, we must recognize that for each one of the constituents which
we have been talking about as having a constancy, there is some sort
of a detection machine. The fact that there are so many machines
all in one small body or cell is something to bear in mind. Since
regulation involves intrinsic detections both for the body as a whole
and for each constituent compartment, how is it that we had nothing
to say about the cell's own assessment of its state? I suppose it is
entirely because nobody so far has found a method of cutting in on
messages which are being transmitted from the surface of a cell to
the interior of a cell, or the kinds of excitation which occur to produce
the response within a cell. If we could find out whether these detec-
tors and transmitters, if there be such, differ at differing ages, then
we would have a more intimate picture of physiological changes with
age. So far we have mainly had to content ourselves with seeing
whether we could show some morphological or biochemical change
with age. As I see it we have not yet got down to what a physiologist
could be really proud of in the measurement of age changes. In my
estimation we do not need to wait until we know what the nature of
these detectors and transmitters may be before we can tackle these
problems of assessment of the state of the responding system. We
can study many a responding system without having any knowledge
of the kinds of gadgets which are in it. Our ignorance of cell excita-
tions is well founded, I suppose, and yet it is disappointing. I hope
the future physiology of cells will, develop a knowledge of these
detectors, and of the way they change with age.

Next I want to try and needle you into thinking of age changes not
as changes of immaturity and senescence but as states in the organism
which are perhaps optimal for each of the age groups. A man of 80
years of age need not necessarily be considered inadequate in any
particular respect. If he has not got as high a clearance at 80 as he
had at 30, can that mean that he has no use for it? This point of
view may lead to a slightly different kind of evaluation of what we

311

312 Concluding Remarks

find, and certainly to a revision of the kind of language in which we
express our results. I think that if we adopt a more descriptive
terminology, and do not imply that one type of organism is inferior
to another, the physiologist, at least, can feel a little satisfaction.

My third point is that we have not done much in this conference
with the description of the intake side of metabolism ; we have talked
about water and electrolytes almost entirely from the point of view
of output. I realize that we all think that we know a little more
about output than we do about intake, but perhaps we should have
made up our minds before we began the meeting that we knew
enough about outputs to feel semi-comfortable and that we knew
sufficiently little about intakes to feel distinctly uncomfortable, so
we might plan to see what we can find out about them. Lots of
people think that a regulation consists in an organism taking in
everything in sight and then getting rid of what is excessive. In my
experience this is a distinct misconception because where intakes have
been studied, we find that they are at least as accurately regulated
and controlled as outputs. If you give an animal a water deficit of
5 per cent of the body weight and see how much water it takes in the
first half-hour of recovery from that deficit, you will find that its
accuracy of intake is equal to its accuracy of output when it has an
excess of water from the body of 5 per cent. This accuracy, then, is
of a kind that must be assessed when we talk about intakes. The
intakes are, so far as we know, specific in a number of instances. We
have not been able to recognize specific ways in which the organism
responds to each of its deficiencies, but we know that there are
specific recognitions for sodium, and there may be more specific
recognitions for some of the other components. If we can see how
the organism relates its intake to its deficits, and how specific those
relations are, we shall have made the sort of quantitative progress
that we have already been able to recognize with respect to excretion.

Davson: Prof. Adolph has spoken as a physiologist, and there is
very little left for me to do, except to re-emphasize what he has said.
The organism is most dependent upon the reactions of certain critical
cells which respond to minute changes in their environment, such as
changes in magnesium concentration. It seems quite miraculous that
the cell could respond in these circumstances; we know that it can
respond to a large jump in its external potassium, and we think we
know the theory of that, but we are usually concerned with barely
measurable changes in the cell's environment. Consider, say, the
olfactory organ. There you have a concentration of gas which is quite
undetectable by any chemical means and yet one can detect the
presence of this gas ; that means that your cell is responding to some
infinitely small change in its environment and, as Prof. Adolph has

Concluding Remarks 313

emphasized, that is the way in which we regulate both output and
input.

The Chairman created a precedent by quoting from a minor poet
last night, and I would like to quote from a major poet. Shakespeare
was, I think, a very good physiologist, and he described age by saying
"when age hath drunk his blood and filled his brow with lines and
wrinkles". Now those are two aspects that we have ignored. We
have been told about the extracellular volume but not whether the
blood volume has changed in age ; the wrinkles of the brow I think
must be determined partly by extracellular water, and also by the
state of the collagen under the skin.

Swyer: As one of those who have something to do with hormones I
have been struck by one or two points more forcibly than by others
in this conference. When hormones are considered in relation to
electrolyte metabolism in ageing and with regard to sexual differences
it seems to me that we have two sets of data, both incomplete. One
of them relates to changes in hormone production with age and sex,
and the other to changes in water and electrolyte metabolism with
age and sex. For example, we have the data on body compartments
that Dr. Olesen gave us, which were very interesting indeed, and I
wish I had known more about that side of the problem before I set
about my own task. We have, too, the experimental evidence on the
development of hormonal responses with age and sex, and on this
point I feel there is something very fascinating which was touched
upon in the discussion but not sufficiently elaborated. I feel that we
need to determine more precisely the exact effect of sex, whether
it is indeed hormonal or genetic. I would like to suggest to Dr.
DesauUes that an interesting extension of his experiments might
be to carry them out on rats which had been castrated in utero by the
technique of Jost, and subsequently had their sex determined by the
cytological techniques which are now so readily available.

Another point which I thought was brought out very well by Dr.
Fourman was this question of the differential action of Cortisol and
aldosterone, the one liberating potassium in the cells as a result of
protein catabolism, and the other altering the renal exchange of
sodium and potassium. The importance of taking this into account
in attempting to use urinary Na/K ratios as a measure of these
salt-retaining hormones was emphasized.

I feel I should say a little about some of the things which were not
quite left out but almost so: calcium seems to have come in for
remarkably little attention during this colloquium, and I think the
only mention of the parathyroid glands was made by Dr. Kennedy
this morning. It is true that the parathyroids have no effect on
water metabolism except in highly abnormal states, but like some

314 Concluding Remarks

other hormones which receive Httle attention I think their hormone
deserves more thought than we have given it. Among these other
hormones I would hke to mention perhaps the thyroid. In myxoe-
dema there is a profound alteration in water metabolism, and that
might have exercised our thoughts too. Growth hormone is another
one which may be very important in the development of some of the
responses which vary with age, particularly in the younger organism.
Finally, the data which Dr. Shock described to us and on which
Dr. Kennedy's experiments also have a bearing, raise the question,
not completely solved, of whether the variations in renal function
which occur in senescence are entirely due to the age changes in the
kidneys themselves, or whether they might also be partly influenced
by the changes in hormone levels at that age. I have in mind particu-
larly the altered relationship between the adrenal anabolic and cata-
bolic steroids, which apparently moves in favour of the latter.

CHAIRMAN'S CLOSING REMARKS

McCance: On the opening day of this meeting Prof. Adolph dis-
cussed the capacity of the infant kidney to maintain the composition
and volume of the extracellular fluids, and he gave us a picture of its
responses to water, salt, and various other kinds of loading as it
developed. He was really discussing the ability of an " end organ " to
maintain the composition of the body. He said nothing about the
fact that the composition of the infant's body differed from that of
adults. We heard nothing about why such differences existed and
how they were maintained, yet they are the very, essence of electro-
lyte metabolism at that age. But the next day differences in the
composition of the body were considered when Dr. Olesen told us that
the extracellular fluids are comparatively very much larger at the
time of birth, and at the age of which Prof. Adolph was speaking,
than they are in the adult. Prof. Heller then brought up the question
of whether this large volume of extracellular fluid in the infant was of
any value or had any function. Nobody took up this challenge or
discussed how the volume was normally maintained.

Prof. Kerpel-Fronius's paper, which was read by Dr. Young, intro-
duced some rather novel ideas which were discussed to some extent
but we missed the originator of them, and I would prefer to leave you
to make your own interpretation of them. However, I was interes-
ted in the point he made that the infant's water reserves and fluid
volumes were small relative to its normal requirements even for
the circulation and metabolic rate, quite apart from losses through the
skin. Dr. Davson brought the matter to a head, I felt, in insisting
that size must be clearly separated from immaturity in their effects
on somatic function.

Dr. Shock showed that in advanced old age, even apart from
disease, the end organ begins to respond in the same kind of way that
it does in very early life. In both cases the end organ seems quite
capable of doing the work which nature intended it to do in a healthy
person of that age, but when one subjects it to the stresses which it is
capable of correcting in the young adult, one can pick out signs of
weakness. He did not discuss the composition and volume of the body
fluids in old people. Are there any steady states, normal or abnormal,
due to senility, either in the cell or in the body as a whole? Something
like this may be the basis of senility. The inability of senile kidneys
to maintain internal acid-base control as perfectly as those of young
adults was an interesting point to me.

315

316 Chairman's Closing Remarks

Dr. Fourman gave us a good account of an abnormal steady state
in the body, maintained and religiously guarded by the end organ
and the sensitive organs, but we did not have time to discuss the
effect of this on the function of the body as a whole, or how the
abnormality had been created.

Dr. Davson gave a clear exposition about the way in which the cells
maintain their electrolyte metabolism and their internal structure. In
other words he discussed the cellular steady state as distinct from
bodily steady states. He pointed out, which is very important of
course, that the cellular steady state is maintained by the metabolism
of the cell itself.

Dr. Kf ecek, Dr. DesauUes and Dr. Swyer put my fears to rest about
the hormone balance of the colloquium. They demonstrated both
well-known and hitherto unknown ways in which the hormones can
be shown to affect the end organ, and something about how this effect
varies with age and with sex.

Dr. Thaysen gave what was to me a most interesting paper about
the way in which various glands elaborate and deliver their secretions
and particularly the electrolytes in them, and the way in which
their mode of action can be interpreted in the light of their final
product. The glands as a group are certainly worth further study for
no two seem to do the same thing. If we could only isolate them and
compare their metabolism with their secretions in relation to the
level of sodium, potassium, oxygen, etc., in the serum and blood, how
interesting it would be !

Dr. Karvonen's paper about the genetic control of electrolyte
metabolism in the erythrocytes was the only major contribution on
this general subject, but of course there are plenty of ways in which
we know that genetics and inheritance can affect electrolyte meta-
bolism. There are abnormal steady states in the body well known to
be under genetic control, such as the " hyperelectrolytaemia " of
infants. We have recently had male infants (brothers) under observa-
tion, in whom there has been a breakdown in acid-base control and
an abnormal steady state in the body fluids, due among other things
to a failure of the kidney to make and excrete ammonia. Genetic
aspects of electrolyte metabolism are going to become more important
as time goes on, and indeed a discussion of the hereditary trans-
mission of abnormal steady states and electrolyte metabolism would
be a very interesting one.

Prof. Wallace discussed the ability of the organism to maintain its
normal cellular steady states under various nutritional conditions.
He came to the conclusion that wide variations in specific intakes did
not affect the composition of the cells but they may apparently
greatly affect the amount of calcium and phosphorus in the bone.

Chairman's Closing Remarks 317

Dr. Talbot gave us a practical paper on the tolerance of the body,
particularly the developing body, to stresses caused by the adminis-
tration of too large and too small amounts of the electrolytes nor-
mally present in the body. In dealing with the responses of the body
as a whole rather than with the end organ responsible for the
restoration of the steady state he was showing us the results of tests
which had been discussed before in relation to the kidney.

Dr. Kennedy summarized and synthesized the information about
the effect of over-nutrition, age, and so on, on the kidney, and the
points have been thoroughly discussed. Dr. Black gave us a good
illustration of the way in which the end organ, again, can break down
and thus allow an abnormal steady state to develop, but why and
how it breaks down he did not decide.

Dr. Fejfar gave us a glimpse of some of the interesting work going
on in the Institute for Cardiovascular Research in Prague. His
subject was congestive heart failure, and he discussed the renal and
extrarenal reasons for the retention of water and salt. This con-
sideration of the production of an abnormal steady state and the
potassium deficiencies which might follow from it gave rise to a
discussion which will be fresh in your minds.

Dr. Card kept the subject of his paper secret till the last moment,
but in the end he had to come out with it. He gave us a fascinating
description of a patient with severe magnesium deficiency, which as
far as I know has never been described before. The results of his
metabolic studies made us realize how difficult it would be to repro-
duce the state of this patient experimentally, and we certainly know
more about the functions of magnesium than we did when I made my
opening remarks.

We could have had more about the body as a whole. We have not
heard as much as I should have liked about what maintains the
electrolyte make-up of the body. Why is it different at birth,
maturity and in old age? What maintains these steady states,
which together make up the composition of the body? What causes
departures from them, and how are the abnormal ones maintained?

One could go on asking questions for ever. Let us be satisfied ; we
have had a good colloquium. Thank you all for coming to it, and let
us all thank the Ciba Foundation for entertaining us so hospitably.

AUTHOR INDEX TO PAPERS

PAGE

Adolph, E. F. . . . 3

McCance, R. A. . . .209

Black, D. A. K.

264

McMurrey, J.

102

Card, W. I.

301

Marks, I. N.

301

Davson, H.

Olesen, K. H.

102

Desaulles, P. A.

180

Parker, H. V.

102

Dlouha, Helena

165

Richie, R. .

139

Fejfar, Z. .

271

Shock, N. W.

229

Fourman, P.

Swyer, G. I. M.

Friis-Hansen, B.

102

Talbot, N. B.

139

Jelinek, J.

165

Taylor, Anne

116

Karvonen, M. J.

199

Thaysen, J. H.

Kennedy, G. C.

250

Vacek, Z. .

165

Kerpel-Fronius,

154

Wallace, W. M.

116

Kfe^ek, J.

165

Weil, W. B.

116

Kfeckova, Jarmila

. 165

Widdowson, E. M.

. 209

Leeson, Patricia

319

SUBJECT INDEX

Acid -base balance, changes in due Age

to age, 224-245
development of, 209-223
during menstrual cycle, 93
in foetal life, 217-219
in old age, 242-243
Acidosis, respiratory (see Respiratory

acidosis)
ACTH, effect on adrenals, 175

effect on potassium excretion, 17G,

177, 178
effect on sodium excretion, 17G,

177, 178
effect on water loss, 176, 177, 178
Adolescence, water and electrolyte

changes during, 80-81
Adrenal corticosteroids, excretion

of, changes due to age, 91
Adrenal glands, control of sodium
intake, 166
effects of ACTH and cortisone,

175-176
effect of castration, 197-198
effect on diuresis, 13
Adrenal hyperplasia, effect on
water and electrolytes, 79-80
potassium excess in, 95
Adrenal steroids, effects of age on
influence of, 192-194
effect on kidney, 257, 262
effect on water and electrolyte
excretion, 180-194, 196-198
Adrenaline, effect on water diuresis,

9, 14
Adults, water in body of, 106-110
Age, body water changes due to,
110-112, 114, 115
causing changes in acid-base bal-
ance, 224^245
causing changes in effect of aldo-
sterone on urine, 182-187
causing changes in effect of pitres-

sin, 239-240
causing changes in extracellular

water, 31, 110-112, 114^115
causing changes in glomerular fil-
tration rate, 231, 238, 246

causing changes in hormonal con-
trol of homeostasis, 168-179

causing changes in intracellular
water, 110-112, 114, 115

causing changes in nitrogen excre-
tion, 243

causing changes in oestrogen excre-
tion, 91

causing changes in steroid metab-
olism, 90-92

causing changes to homeostatic
capacity, 142-149

cellular changes due to, 199-205

changes in ketosteroid excretion
due to, 91

effect on homeostasis, 139-153

effect on influence of adrenal
steroids, 192-194

effect on renal disease, 250-263

effect on starvation, 226

effect on water diuresis, 238-240

electrolyte changes due to, 241,
311-312

erythrocyte changes due to, 199-
205, 207

haemoglobin changes due to, 203,
207

pulmonary effects of, 264

renal effects of, 11-12, 227-228,
229-249, 253-254
Aldosterone, 59, 60

effect on potassium excretion,
183-184, 186, 192-194, 196-197

effect on sodium excretion, 183,
185, 192-194, 196

effect on sodium/potassium ratio,
184r-185, 186-187, 192-194, 196

effect on urinary output, 182,
192-194, 196

excretion in congestive heart fail-
ure, 280, 292-293, 298, 300
in pregnancy, 89-90
Allantoic fluid, 217, 218
Ammonia, excretion of, 209-210,
213-215
in respiratory acidosis, 266

321

322

Subject Index

Ammonium salts in metabolism,

209-210
Anaemia, erythrocytes in, 199
Anions, excretion of, 209, 210

in infancy, 211-213
Antidiuretic hormone, 12, 37, 46,
47, 53, 55, 92, 238-240
in congestive heart failure, 280
Anuria, due to respiratory infection,

268
Aqueous humour, concentration of

ions in, 25-26, 28, 29
Ash, in rat body, 120-124

relation to body composition, 118,
122

Bicarbonate, excretion of, in respir-
atory acidosis, 265-266
in pancreatic juice, 64
in parotid saliva, 64
Blood-brain barrier, 26
Blood volume, effect of age, 243
Body, composition of, effect of
protein and mineral intake,
116-138
water in, 102-115
Bone, magnesium in, 309

Calcium, effect of diet on, 120, 121,
127-128, 132, 138
in body of rat, 120, 121, 127-128,

132, 138
in foetal urine, 218
Carbonic anhydrase, 218, 223
control of urinary pH, 210
Cardiac output, effect on kidneys,
234, 248, 267
in congestive failure, 272, 276
Castration, 227

effect on adrenal glands, 197-198
Cells, age changes in, 199-205
electrolytes and water in, 15-35
electrolyte transfer in, effect of

heat, 19
membrane of , permeability of , 1 6-3 1
osmotic equilibrium of, 18
Cerebral hypoxia, in congestive

heart failure, 286
Cerebrospinal fluid, concentration

of ions in, 25-26
Children, water in body of, 103-106
Chlorides, effect of diet on, 120, 121,
126, 132

Chlorides

excretion of, in congestive heart

failure, 276, 277, 278, 284-285
in babies' urine, 211
in erythrocytes, 203
in foetal urine, 217
in rat body, 120, 121, 126, 132
in sweat, 64, 74
in tears, 64, 71
loss of, during labour, 90
Chorioallantoic membrane, 218
Circulation, effects of deficiency of

water, 160, 163
Citric acid, excretion of, 217, 218,

221, 222
Cold, effect on osmolarity of cell,

24
Congestive heart failure, aldo-
sterone excretion in, 280,
292-293, 298, 300
cerebral hypoxia in, 286
humoral factors, 279-288
neural factors, 279-288
renal changes in, 275-279
renal function in, 271-275
salt and water retention in, 288-
293

water and electrolyte metabolism
in, 271-300
Connective tissue, water and elec-
trolytes in, 27
Cor pulmonale, renal function in,

266-267
Cortexone, effect on potassium ex-
cretion, 174, 175
effect on sodium excretion, 174,

175, 177
effect on water loss, 174, 175
Cortisol, effect on potassium excre-
tion, 188, 189, 192-194, 196
effect on sodium excretion, 187-

188, 189, 192-194, 196
effect on sodium/potassium ratio,

190-192, 193-194, 196
effect on urinary output, 187, 188,
192-194, 196
Cortisone, effect on adrenal glands,
175-176
effect on diuresis, 13
effect on potassium excretion, 171,

172, 176, 178
effect on sodium excretion, 171,

172, 173, 176, 178
effect on water loss, 171, 172
Creatinine excretion, 249

Subject Index

323

Dehydration, effect on water intake, 4

in labour, 94, 95
Dehydration reaction, 38-39, 47
Diabetes insipidus, causing loss of

water, 39, 42-43
Diarrhoea, causing hypernatraemia,

58
Dibenamine, effect on kidney, 281,

282
Diet, effect on body composition,
117-138

effect on electrolytes, 116-138

effect on homeostasis, 143-144
Diuresis, effect of adrenal glands, 13

effect of adrenaline, 9, 14

effect of age, 6-10, 238-240

effect of cortisone, 13

effect of hypoxia, 8

effect of pitressin, 7-8, 11

effect of vasopressin, 12, 13

in congestive heart failure, 272-273,
275

Electrolytes, cellular aspects of,

15-35
changes in due to age, 241, 311-312
deprivation of, 144
during pregnancy, 88-90
effect of diet, 116-138
effect of hormones on, 313-314
effect of hypercapnia, 265
effect on mineral content of body,

125
effect on protein body content, 125
excretion of, response to adrenal

steroids, 180-194, 196-198
glandular secretion of, 62-77
hormonal aspects of, 78-98
in congestive heart failure, 271-300
in muscle, 164

in parenteral fluid therapy, 146-148
metabolism of, in infancy, 154-164
regulation of, by kidney, 229-249
total exchangeable in body, 108
See also under Sodium, Potassium,

etc.
Erythrocyte, electrolytes and water

in, 17-21, 199-208
in foetus, 204, 205, 206
in sheep, 200-203, 204, 206
Extracellular fluid, equilibrium

with plasma, 15-16
volume of, changes due to age,

244
Eyes, water content of, 28, 29

Fat, in body, 113, 114, 115, 129, 132
in rat body, 119

Fluids, metabolic disturbances, rea-
sons for, 154r-155

Foetus, acid-base balance in, 217-219
haemoglobin in, 203
urine in, 217

Gibbs-Donnan equilibrium, 15-18,

27, 28, 30
Glomerular filtration rate, changes

with age, 231, 238, 246
effects of pyrogen, 237
in congestive heart failure, 275,

277
Growth, body water changes due to,

103-106
diet in, 116-138
effect on electrolytes, 160
in mice, 136
in rats, 136

Haemoglobin, changes in due to age,
203, 206, 207

foetal, 203, 206, 207

in sheep, 202-203
Heart, effect of potassium on, 95-96
Heart failure, congestive {see Con-
gestive heart failure)
Homeostasis, disturbances of, in
infants, 154-157

effect of hormones on, 165-179

of water and electrolytes, effect of
age, 139-153
Hormones, effect on electrolytes,
313-314

effect on homeostasis, 165-179
Hydrogen ion gradients, 34
17-Hydroxycorticosteroids, effect

on water and electrolytes, 79
Hypercapnia, renal effects of, 265-

266
Hypernatraemia, and cerebral dis-
turbances, 36-44

due to diarrhoea, 58

due to water deficiency, 38-44
Hypertension, renal aspects of, 258
Hypertonic saline, effect on hypo-

natraemia, 50
Hyponatraemia, 44-55, 95

and cerebral disturbances, 36-37

and steroid output, 60
Hypothalamus, effect on thirst, 37
Hypoxia, effect on diuresis, 8

324

Subject Index

Infants, electrolyte metabolism in,

78-79, 154^164
water metabolism in, 78-79, 154-

164
water retention in, 96-98

17-Ketosteroids, excretion of,

changes due to age, 91
Kidney, blood flow in, 248
age changes, 234-235
in congestive heart failure, 272,
273-274, 276, 282, 283, 284,
287
changes in due to age, 227-228,

229-249, 253-254
changes in due to congestive failure,

275
concentrating ability of, age varia-
tions, 11-12, 243-244
diseases of, effect of age, 250-263
effect of Dibenamine, 281, 282
effect of obesity on, 254, 255, 260
effects of potassium deficiency on,

262-263
effects of pyrogen, 235-238
effect of water deficiency on, 43
enzymes in, changes in due to age,

245
function of, 229

in respiratory failure, 264-270
glomerular filtration rate, changes
due to age, 231, 248
in congestive heart failure,
275, 277
growth of, 251-252
hormonal damage to, 256-258, 262
in cor pulmonale, 266-267
lesions of, causing loss of water, 39
overloading of, producing "senile"

changes, 254-^255, 260
plasma flow in, changes with age,
229-231, 235
effects of pyrogen, 237-238
regeneration of, 252-253
role of in water and electrolyte

regulation, 229-249
tubular excretion of, changes due to
age, 233, 247

Labour, dehydration during, 94, 95
Lungs, effects of age on, 264

Magnesium, deficiency of, 301-310
signs of, 304, 307, 309-310

Magnesium

in body of rat, 120, 121

in bone, 309

in plasma, 99-100
Malnutrition, effect on body fluids,

156-157
Menstrual cycle, acid-base balance
during, 93

effect on water and electrolytes,
81-88

electrolyte changes during, 93

sodium/potassium ratios during,
83-88
Mental excitement, due to mag-
nesium deficiency, 304, 307, 309-

310
Mercury poisoning, excretion of

sweat in, 99
Metabolic disturbances in infants,

154-164
Metabolism, comparison between

infant and adult, 157-159
Mineral, intake of, effect on body

composition, 116-138
Muscle, analysis of in sodium de-
ficiency, 49-50

composition of, 23

electrolytes in, 21-22, 164, 224-225

potassium in, 289-291

water in, 21-22, 113, 163-164

Nephrectomy, effects of, 252, 255,

257
Nitrogen, excretion, changes due to
age, 243

in rat body, 120, 121

Obesity, effects on kidney, 254, 255,

260
Oestrogens, effect on water reten-
tion, 79, 84, 86

excretion of, changes due to age, 91
Osmotic diuresis, 40

Pancreatic juice, bicarbonate in, 64
sodium excretion in, 63, 65, 71
urea in, 68-69
Parenteral fluid therapy, 144,

146-148, 151
Parotid saliva, bicarbonate in, 64
potassium excretion in, 63, 64,
65, 74, 75

Subject Index

325

Parotid saliva

sodium excretion from, 62-63,

65, 66, 69, 71
urea in, 67-69, 75
Phosphate, excretion, in respiratory
acidosis, 266
in babies' urine, 211, 213, 215,
216
Phosphorus, effect of protein intake,
121, 127-128
excess of, 144, 145-146
in body of rat, 120, 121, 127-128
intake of, 142
Pitressin, effect on hyponatraemia,
53-55
effect on water diuresis, 7-8, 11
variation of effects due to age,
239-240
Pituitary gland, effect on electro-
lytes, 166, 167
Plasma, concentrations of ions in,
25-26
equilibrium with extracellular fluid,

15-16
magnesium in, 99-100
potassium in, 65
sodium in, 65
urea in, 67
Potassium, accumulation of in cell,
32, 33
deficiency of, effects due to, 140
effects on kidney, 262-263
in congestive heart failure,
289-292, 299-300
deprivation of, causing cellular

oedema, 32
effect of protein intake, 121, 126,

129, 133
effect on heart, 95-96
excess of, effects due to, 140, 141,
144-146, 152
in adrenal hyperplasia, 80, 95
exchangeable amounts in body, 108,

109, 111, 114
exchange of, in cell, 18, 19, 20, 21,

22, 24, 30, 34
excretion of, after water loading,
167-169
effect of ACTH and corti^ne,

176, 177, 178
effect of aldosterone, 183-184,

186, 192-194, 196-197
effect of cortexone, 174, 175
effect of Cortisol, 188, 189,
192-194, 196

Potassium

excretion of, effect of cortisone, 171,
172, 176, 178
effect of vasopressin, 170
in parotid saliva, 63, 64, 65, 74,

75
in respiratory acidosis, 266
in sweat, 63, 64, 65, 74, 76, 77
in body of rat, 120, 121 , 126, 129, 133
in erythrocytes, 200-202, 204, 206,

207, 208
in foetal urine, 218
in muscle, 224-225, 289-292
in plasma, 65
in saliva, during menstrual cycle,

83-88
loss of, 226

during labour, 90
ranges of intake, 142
Potassium chloride, effect on

hyponatraemia, 53
Potassium pump, 204
Pregnancy, aldosterone excretion
during, 89-90
sodium retention during, 88-89
water and electrolyte changes

during, 88-90
water retention during, 88-89
Premenstrual oedema, 81-83
Protein, breakdown of, causing
osmotic diuresis, 41
in renal disease, 260-262
intake of, effect on body compo-
sition, 116-138
Pulmonary oedema, 284-285
Pyrogen, effects on kidney, 235-238
Pyruvic acid, in urine, 221

Respiratory acidosis, 265-266,

269-270
Respiratory failure, renal function

in, 264^270

Saliva, bicarbonate in, 64

potassium in, 63, 64, 65, 74, 75,
83-88

sodium in, 62-63, 65, 66, 69, 71,
83-88

sodium /potassium ratios in, 94

urea in, 67-69, 75
Sex, differences in body water,

107-110, 113
Sheep, erythrocytes in, 200-203, 204,

206
Skin, water absorption by, 100-101

326

Subject Index

Sodium, and adrenal function, 166
deficiency of, causing liyponatrae-
mia, 44-45
effects due to, 140
See also Hyponatraemia
effect of protein intake on, 121, 126,

132
effect on water intake, 37
excess of, 140, 144, 145-146, 226

See also Hypernatraemia
exchangeable amounts in body, 108
exchange of, in cell, 18, 19, 20, 21,

22, 24, 30, 34
excretion of, 62-63

after water loading, 167-169

during exercise, 279

effect of ACTH and cortisone,

176, 177, 178
effect of aldosterone, 183, 185,

192-194, 196
effect of cortexone, 174, 175,

177
effect of Cortisol, 187-188, 189,

192-194
effect of cortisone, 171, 172,

173, 178
effect of vasopressin, 170
in congestive heart failure,

277, 288, 289, 299
in pancreatic juice, 63, 65, 71
in parotid saliva, 62-63, 65,

66, 69, 71
in respiratory acidosis, 266
in sweat, 62-63, 65, 66, 69, 71,

74, 75, 76, 77
in tears, 63, 65, 71, 75, 76
in body of rat, 120, 121, 126, 132
in erythrocytes, 200, 201-202, 203,

206, 207, 208
in foetal urine, 217
in plasma, 65
in saliva, during menstrual cycle,

83-88
in submaxillary gland, 71-72
loss of, in adrenal hyperplasia,

79-80
ranges of intake, 142
retention of, in congestive heart
failure, 288
in pregnancy, 88-89
Sodium/potassium ratios, during
menstrual cycle, 83-88
effect of aldosterone on, 184-
185, 186-187, 192-194, 196-
197

Sodium/potassium ratios

effect of Cortisol on, 190-192,
193-194, 196

in saliva, 94
Sodium pump, 203, 207
Starvation, effect of age, 226
Steroid metabolism, changes due

to age, 90-92
Stress, effect on kidney, 260
Sublingual gland, electrolytes in,

69-70
Submaxillary gland, electrolytes

in, 69-70
Sweat, 100

chloride in, 64, 74

in mercury poisoning, 99

potassium excretion in, 63, 64, 65,
74, 76, 77

sodium excretion in, 62-63, 65, 66,
69, 71, 74, 75, 76, 77

urea in, 67-69

Tears, chloride in, 64, 71

sodium excretion in, 63, 65, 71, 75,
76

urea in, 67-69
Thirst, effect of, 143, 144

failure of, 41-43
Thirst centre, 37
Toxaemia of pregnancy, 88, 89-90

Urea, excretion of, 40-41, 73
in pancreatic juice, 68-69
in parotid saliva, 67-69, 75
in plasma, 67
in sweat, 67-69
in tears, 67-69
Urine, acids excreted in, 210, 215-
217, 221, 222
detection of, 220-221
ammonium salts in, 209-210
in babies, 210-211
in foetus, 217

magnesium excretion in, 305, 309
output, effect of aldosterone on,
182, 192-194, 196
effect of Cortisol, 187, 188, 192-
, 194, 196
pH of, 209-210, 211-212, 215, 221,

222
potassium in, during menstrual

cycle, 84-88
sodium in, during menstrual cycle,
84-88

Subject Index

327

Vasopressin, effect on electrolytes,
167, 170
effect on water diuresis, 12, 13, 169,
170, 195
Venous pressure, in congestive

failure, 272, 274, 275
Vitamin A, effect on kidney function,
247

Water, cellular aspects of in body,
15-35
content, control of, 10-11
deficiency of, causing hypernatrae-
mia, 38-44
effects of, 140, 160, 163
in children, 160
renal effects, 43
symptoms, 39
deprivation of, effect on hypo-

natraemia, 51-52
diuresis, at various ages, 6-10
effect of adrenaline, 9, 14
effects of age, 238-240
effect of pitressin, 7-8, 11
in congestive heart failure, 272-
273, 275
effect of load in rats, 167-168
effect of vasopressin on loss of, 169,

170
excess of, effect on hyponatraemia,
52
effect on diuresis, 6, 8
effect on urine output, 4
effects due to, 46-47, 140, 144,

145-146, 150, 151
in children, 159
exchange of in body, 3
excretion, during exercise, 279
response to adrenal steroids,
180-194, 196-198
extracellular, in adults, 106-110
in children, 103-106

Water

extracellular, variations with age,

31, 110-112, 114, 115
in body, 102-115

effect of age on, 110-112, 114,

115, 180
effect of growth, 103-106
measurements of, 102-103
in body of rat, 119, 122, 123
in muscles, 113, 163-164
in parenteral fluid therapy, 146-148
intake of, control of, 9-10
intracellular, effects of age, 110-
112, 114, 115
in adults, 106-110
in children, 103-106
loss of, 39-41, 195

during labour, 90

effect of ACTH and cortisone,

176, 177, 178
effect of cortexone, 174, 175
effect of cortisone, 171, 172,

176, 178
following adrenalectomy, 172
metabolism, hormonal aspects of,
78-98
in congestive heart failure, 271-

300
in infants, 96-98, 154-164
in malnutrition, 156-157
in pregnancy, 88-90
movement of, in cell, 19, 20, 22, 25,

27-29, 34
physiological regulation of, 3-14
ranges of intake, 142
regulation of, 37-38

by kidney, 229-249
retention of, effects of oestrogen, 79
in congestive heart failure, 288
in pregnancy, 88-89
in premenstrual period, 81-83
tolerance to excess, 150, 151
Water load, effects of, 170

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Ciba Foundation colloquia on ageing"

See other formats