Clinical Science (1972) 42,673-684.
EXCHANGEABLE POTASSIUM CONTENT A N D
DISTRIBUTION I N NORMAL SUBJECTS AND
URAEMIC PATIENTS O N CHRONIC HAEMODIALYSIS
V. RETTORI, T. GRAL, S. G. MASSRY A N D M. F. VILLAMIL
Cedars-Sinai Medical Research Institute, Los Angeles, California, U.S.A., and
Centro de Investigaciones Cardioldgicas, Buenos Aires, Argentina
(Received 19 October 1971)
SUMMARY
1. Exchangeable potassium, sulphate space and intracellular fluid have been
determined in eight normal males and in seven uraemic male patients on a regime of
twice weekly 8 h dialysis.
2. Intracellular water was more closely correlated with exchangeable potassium
than were body weight or lean body mass.
3. There was no correlation between exchangeable potassium and plasma or
erythrocyte potassium.
4. Uraemic patients had a 20% potassium depletion that may have been due to a
restricted oral intake (40 mmol/day) and too little potassium (0-1 mmol/l) in the
dialysis fluid.
Key words : haemodialysis, exchangeable potassium, uraemia.
Maintenance haemodialysis with low or potassium-free dialysis fluid has become an established
treatment of patients with chronic renal failure. Potassium depletion has been reported in
these patients by some authors (Johny, Lawrence, O’Halloran, Wellby & Worthley, 1970;
Seedat, 1969) but has not been observed by others (Johnson, Frohmert & Novak, 1969;
Morgan, Burkinshaw, Robinson & Rosen, 1970; Ram & Chisholm, 1969). The discrepancy
may be ascribed to differences of diet, time and frequency of dialysis and K+ concentration
in the dialysis bath, The dficulty in properly defining K+ depletion in these patients is another
important source of uncertainty. Total body K+ varies with sex and age. Moreover, the validity
of body weight which has been proposed as a reference standard for total body K+ has been
questioned (Muldowney, Crookes & Bluhm, 1957). Good correlations between exchangeable
K+and creatinine production estimated by daily urinary excretion, lean body mass or erythrocyte mass have been reported (Corsa, Olney, Steenburg, Ball & Moore, 1950; Muldowney
et al., 1957). However, these are not adequate reference standards for uraemic patients for the
Correspondence: Dr Mario F. Villamil, Centro de Investigaciones Cardiol6gicas, Marcel0 T. de Alvear 2270,
Buenos Aires, Argentina.
673
V. Rettori et al.
674
following reasons: (1) creatinine production is difficult to measure in anuric patients; (2) values
of lean body mass are derived from total body water, thus implying the absence of water
retention, which cannot be assumed in chronic renal patients; (3) the anaemia of uraemia
precludes the use of erythrocyte mass as a reference standard in this condition.
In this study these problems were approached by (a) examining several correlates of exchangeable potassium and @) comparing uraemic patients with normal subjects of the same
sex (males) and similar age distribution. In both groups exchangeable K + correlated better
with intracellular water calculated from total water and sulphate space than with any other
parameter. A 20% K + depletion was detected in uraemic patients dialysed against a bath
containing 0-1 mmol of K+/1.
MATERIALS A N D M E T H O D S
Eight normal male subjects and seven male patients with end-stage chronic renal failure that
were on maintenance haemodialysis were studied. The details of the investigation were fully
explained to the patients and normal subjects who volunteered to take part in this study.
Six of the seven patients had chronic glomerulonephritis. In four associated conditions
were (1) anephric transplant rejection, (2) Lawrence-Moon-Biedl syndrome, (3) dissecting
aneurysm and hypertension, (4) malignant hypertension. The seventh had chronic pyelonephritis and transplant rejection. Normal subjects were technicians and members of medical
staff. They were all healthy and free of renal disease as evidenced by normal blood urea,
creatinine clearance and urinary sediment. Patients were on twice weekly, 8 h coil dialysis.
TABLE
1. Plasma concentrations in uraemic patients of urea nitrogen, creatinine and electrolytes before and
after 8 h of dialysis
Urea nitrogen Creatinine
Caf
P
(mg/100 d) (mg/100 ml) (mg/100 ml) (mg/100 ml)
+
Before
dialysis
After
dialysis
85.6
f 45.3
(36-175)
16-8
8.5
2 6.9
k 3.1
(8.630)
(5.415)
8.9
f 1.0
(74-9.9)
42.1
k 36.0
(15-122)
9.1
k 5.0
(56-20)
4.8
zk 2.4
(2.5-10)
9.4
f0 7
(8.2-10)
Na+
(mmol/l)
K+
Haematocrit
(mmol/l)
(%)
136
2.3
(132-139)
4.3
f 0.8
(34-5.6)
137
f 6.4
(132-150)
k0.5
3.9
21.3
2.7
(17-24)
(3.3-4.8)
The values were measured in seven patients and are meansf SD and the range is given in parentheses.
Two patients were newly on the dialysis programme and had been treated with dialysis only
two and five times, respectively, before the estimation of their exchangeable K +, whereas the
other five patients had been maintained with dialysis for 1-17 months before the study. Five
patients were dialysed with a potassium-free dialysis fluid, including the two new patients. In
the remaining two patients the dialysis bath contained 1 mmol of K+/l. The patients had no
detectable oedema and were on a diet containing approx. 40 mmol K+/day. None of the
patients was on ion-exchange resin therapy at any time. Plasma concentrations of urea, creatinine and electrolytes, before and after dialysis, are summarized in Table 1.
675
Potassium in uraemia
Exchangeable K+ was measured with 42K+,
total body water with 3H20
and extracellular
water with 35S042-. The three radioactive compounds were given intravenously at the end of
dialysis in a single injection. The method for measurement of exchangeable K+ was essentially
as described by Veal1 & Vetter (1958). Urine (24 h sample) was collected when available and
42K+radioactivity lost in the urine was subtracted from the amount of radioactivity injected;
after 24 h, plasma, saliva and a freshly voided urine specimen (when available) were analysed
for 42Ky-radioactivity (well counter) and K+ concentration (flame photometry). At this time
specific radioactivities of different biological fluids were practically identical.
For total body water and extracellular space measurements, blood samples were drawn
30, 60, 90, 120 and 180 min after injection. 3H and 35S8-radioactivities were separated by
differential counting in a liquid scintillation-counter after 42Kradioactivity decreased to
background values (approx. 10 days). In normal subjects plasma 'S radioactivity decayed
exponentially from 60 to 180 min after injection. In uraemic patients with practically no renal
function plasma 35Sradioactivity remained constant from 90 to 180 min after injection (Fig. 1).
0.10 r
0.02
0
I
I
I
I
2
3
Time (h)
FIG.1. Disappearance curves of 35SO:- from plasma of eight normal subjects ( 0 ) and seven
uraemic patients (0) on chronic haemodialysis. Values are means f SD.
In both groups '
H radioactivity remained constant from 1 to 3 h after injection. The stable
values of plasma "S and '
H radioactivity were used for calculation of sulphate space and total
body water. When plasma sulphate radioactivity decayed exponentially, the log of sulphate
counts was plotted against time; the intercept at zero time of the straight part of the curve was
taken as the equilibrium value in plasma and used for calculation of sulphate space. The
sizes of the spaces (ml) were calculated from the equation : [injected radioactivity (c.p.m.)]/
[radioactivity/ml of plasma water (c.p.m.)] x Donnan factor. Plasma water was measured with
a refractometer. A Donnan factor of 1 was used for 3H20
and 0.90 for "S (Walser, Seldin &
Grollman, 1953).
V. Rettori et al.
676
Lean body mass was derived from the equation: total body water106732 (Pace & Rathbun,
1945). K + in erythrocytes was measured by the method of Villamil, Rettori, Simpson &
Kleeman (1970).
The significance of differences between normal and uraemic groups was tested by Student's t
test. Correlations between different values in each group were analysed in correlation matrix
on an IBM 360/50 256 Kb-computer.
RESULTS
Normal subjects (Table 2)
Exchangeable K+ in normal subjects did not correlate with body weight (Fig. 2), plasma K',
35SO:- space or lean body mass; correlation with the latter was slightly below significance
levels (r = 0.60) (Fig. 3). In this group exchangeable K + was negatively correlated with age
(P<0.025) but the best correlation was found between exchangeable K + and intracellular
water (P<O-Ol) (Fig. 4). As expected, an excellent correlation was found between lean body
mass and intracellular water (P<0401). These two parameters were also correlated with age
(P<0*05).
TABLE
2. Correlation matrix of normal group
Exchangeable K+
Age
Weight
Plasma
K+
Lean body
mass
- 0.75
-
-
-
(P<0.025)
Age
Weight
Plasma K +
-
-
-
-0.70
(P<0.05)
-
Intracellular 35SO: water
space
080
(P<001)
-0.70
(P<0-05)
092
Lean body mass
-
-
-
(P<0~001)
Intracellularwater
Only significant correlations are shown.
Uraemic patients (Table 3)
In these patients exchangeable K+ correlated with body weight (P<0.05)(Fig. 2) and lean
(Fig. 3) but the best correlation was with intracellular water (P<OOol)
body mass (P<0-05)
(Fig. 4). Exchangeable K+ did not correlate with age, plasma K" or '%O:- space.
Some other interesting correlations were found in this group. Plasma potassium was negatively correlated with 3'SO:- space (P<O.Ol) (Fig. 5) and with age (P<0*025) and 35SO$space was positively correlated with age (P<0.01).
Direrences between normal and uraemic groups (Table 4)
Mean values, standard deviation and range of age were very similar in both groups. The
mean weight was higher in the normal group but, because of the wide range of values, difference
between the groups was not statistically significant. In both groups the coefficient of variation
677
Potassium in uraemia
Normals :r = 0-27;
NS
Uraernics:r =0.71;P<0.05
y=358+30+13x
0
0
0
\
50
70
60
80
90
Body weight (kg)
FIO.2. Correlationbetween exchangeableK+and body weight in normalsubjects ( 0 )and uraemic
patients (0).The best fitted lie was derived from values of the uraemic group only.
Norrnak: r =0.60;
NS
Uraemics: r.0.74; PC0.05
y
s
-617 t 54.7f22.1~
0
0
0
40
50
60
70
80
Lean body mass (kg)
FIG.3. Correlation between exchangeable K+ and lean body mass in normal subjects (0)and
uraemic patients (0).The best fitted line was derived from values of the uraemic group only.
V. Rettori et al.
678
5-
Uraemics r = 0 96,P < 0 001
y = - 259+ 103k 14x
/6
I.
r
4-
I
/
/'
c
-E
c
Y
/
/
/
/
+
/
1
0)
E
ol
0
C
3-
L
0
x
W
2-
//
.
I-
I
I
I
r =0.07,P
I
< O 01
~ = 6 9 - 0 2 0 051r
+
3 L,
10
I
I
15
20
Sulphate space (I)
FIG. 5. Correlation between plasma K + concentration and sulphate space in uraemic patients.
Potassium in uraemia
679
for exchangeable K+ was lower when intracellular water was used as a reference standard
(11.0 and 7.4%) than when body weight (16-3and 17.0%) or lean body mass (18.2 and 16.7%)
were used.
Exchangeable K+ was significantly lower in the uraemic group, irrespective of the parameter
used as a reference standard. However, higher degrees of significance were obtained when
values were expressed either in absolute terms (P<0.005)or in mmol/l of intracellular water
(P<0.005)than when they were expressed per kg body wt. (P<O.O2) or per kg lean body mass
TABLE
3. Correlation matrix of uraemic group
Age
Exchangeable K+
--
Weight
Plasma
K+
071
-
(P<0.05)
Age
Weight
-
Lean body
mass
Intracellular
water
0.74
0.96
(P<O*Ool)
-
(P<0.05)
-
-0.78
space
0.82
(P< 0.01)
(P<0025)
-
35SO:-
0-89
0.77
(P<0.025)
(P<0.025)
Plasma K+
-
-0.87
( P i 0.01)
0.85
Lean body mass
(Pc 0.01)
Intracellular water
Only significant correlations are shown.
(P<0.02). The degree of depletion (approx. 20%) was independent of the duration of treatment. Two patients who had been treated by dialysis only two and five times were depleted to
the same extent as the rest of the group.
At 24 h after dialysis, plasma and erythrocyte K+ of uraemic patients were not significantly
different from normal subjects; Pco, and plasma HCO; were low in the uraemic group
(30*8+1.5 mmHg and 20.0f 1.8 mmol/l) but arterial pH was normal (7.42f0.01).
DISCUSSION
In the group of eight normal subjects age rather than weight was closely correlated with cell
mass as expressed by exchangeable K+, lean body mass (derived from total body water) or
intracellular water (Table 2). Previous reports (Edelman, Haley, Schloerb, Sheldon, FriisHanson, Stoll & Moore, 1952; Schwab, Dissman & Schubert, 1963; Steele, Berger, Dunning
& Brodie, 1950) have shown a negative correlation between age and total or intracellular
water. The fact that exchangeable K+ correlated with body weight much better in the uraemic
(Table 3) than in the normal group can be explained by the lower and less-variable fat content in the latter group. The following evidence supports the view that intracellular water
is the best reference point for exchangeable K+ in normal and uraemic groups: (1) in both
groups exchangeable K+ correlated better with intracellular water than with any other
500.0 > d
ZO.0 > d
20.0 > d
500.0 > d
Potassium in uraemia
68 1
parameter; (2) differencesin exchangeable K+ between normal and uraemic groups were more
significantwhen values were expressed as intracellular K+ concentration (Table 4).
The choice of 35SO:- rather than 82Br- for the estimation of extracellular space was based
on the fact that, being a divalent anion, sulphate is more thoroughly excluded from the intracellular space. Therefore, it needs no correction for erythrocyte penetration and is less apt
to penetrate cells when these are depolarized. This is convenient in the uraemic patients since a
decrease in membrane potential of striated muscle fibres may occur in uraemia (Bolte, Riecke
& Rohl, 1964) probably as a non-specific reaction to severe illness (Cunningham, Carter,
Rector & Seldin, 1971). Sulphate space is usually derived from a single blood sample taken at
30 min because complete equilibration in extracellular fluid and negligible renal excretion was
assumedtohavetakenplaceatthistime (Walser et ul., 1953). In normal subjectsvalues obtained
by this procedure are good approximations because, due to the slope of the slow part of the
curve, 30-min values agree fairly well with the intercepts at zero time. However, a decrease in
the slope will result in an under-estimation of the space. This possibility has been reported in
surgical patients (Cleland, Pluth, Tame & Kirklin, 1966; Roth, Lax & Maloney, 1969) and
can clearly be seen by inspecting the curve of uraemic patients in whom stable plasma values
were obtained. This curve also shows that: (a) contrary to previous reports (Walser et al., 1953)
disappearance of sulphate from extracellular fluid during the first 3 h can be accounted for
wholly by renal excretion; (b) at least 1 h is necessary for complete equilibration of sulphate in
the extracellular fluid. This finding agrees with more recent results (Kragelund & Dyrbye,
1967).
Lack of correlation between plasma and exchangeable K+ has been reported in uraemic
patients under maintenance haemodialysis (Hughes, Williams & Smith, 1967; Johny ei al.,
1970; Seedat, 1969) and in patients with other chronic diseases (Flear, Cooke & Quinton,
1967). It is possible that the characteristics of K+ depletion in humans vary according to the
duration of depletion (Flear et ul., 1967), mechanism of production and associated changes
(protein depletion, etc.). In preliminary studies (P. Henney, J. Maloney & M. F. Villamil,
unpublished work) we have found that K+ depletion of short duration in dogs is followed by a
pronounced decrease in plasma K+ with little change in muscle K+.
In uraemic patients, plasma K+ correlated well with extracellular volume (Table 3). Apparently, with increasing duration of the disease, there is a small tendency in these patients to
extracellular expansion. It is difficult to conceive that this can produce dilution hypokalaemia
since extracellular K + is in equilibrium with cellular stores. It is possible that in K+ depletion,
the volume of extracellular fluid is a critical factor in determining time of equilibrationbetween
extracellular and cellular K+ after each dialysis.
According to our results (Table 4), erythrocyte K+ does not reflect the state of cellular stores
in uraemic patients under chronic haemodialysis. This conclusion contrasts with previous
findings (Johny et al., 1970) which showed that a decrease in erythrocyte K+ correlated well
with a decrease in total body K+ during haemodialysis. It is very likely that serial determinations of erythrocyte K+ are useful for detecting changes in total body K+. However, a single
determination of erythrocyte K+ may be meaningless in K+ depletion of uraemic patients for
the following reasons: (1) the normal range of erythrocyte K+ is so wide [73-103 mmol/kg
fresh wt. in one study (Villamil et al., 197011that much K+ must be lost from the cells before
subnormal values are reached; (2) because of a shortened life span, erythrocyte population is
younger in uraemic patients (Joske, McAlister 8c Prankerd, 1956). Young erythrocytes have a
V. Rettori et al.
682
higher K+ content (Joyce, 1958; Keitel, Berman, Jones & Machchlan, 1955) which should
mask depletion.
The potassium depletion of the uraemic group was 17% when expressed as concentration in
intracellular water, but 32% in absolute amount. The possibility that it was partly due to K +
loss during previous dialysis should be considered. The 42Kwas given to the patients at the
end of the 8 h dialysis but the measurements were made 24 h later. Even if faecal K+ losses are
neglected and if it is assumed that the whole amount of K+ (80 mmol) ingested during the 2
subsequent days were retained, differences in exchangeable K+ before and after each dialysis
should be less than 4%.
The incidence of K + depletion found in other studies of uraemic patients under chronic
haemodialysis is variable. Table 5 summarizes our own results, and those of other authors. It is
difficult to draw a definite conclusion from this information because of its incompleteness.
TABLE
5. Summary of values from the literature concerning incidence of K + depletion in uraemic
patients on chronic haemodialysis
Reference
Present study
Seedat (1970)
Johny et al. (1970)
Ram & Chisholm
(1969)
Morgan et al. (1970)
Johnson et al. (1969)'
Time on
Duration of
dialysis dialysis regime
(h/week)
(month)
16
1442
16
-
30
24
36
0.5-17
4-24
1-16
20% of cases
just entered
the programme
3-14
0.75-26
17
6
Dietary K+
(mmol)
40
58
20
K + in
dialysis bath
(mmol/l)
K+
depletion
0-1
1
1
20
20-35
Variable in
half of the
cases
1
None
1.5
0
0
None
None
20% in half
of the cases
Probably
unrestricted
80
Unrestricted
Unrestricted
(%)
* Values correspond to two different programmes.
However, it seems beyond doubt that a K + concentration in the dialysis bath of 1-5mmol/l
and a liberal intake of K + will prevent K + depletion in these patients. Conversely, a restricted
oral intake coupled with a K+ concentration in the dialysis bath from 0 to 1 mmol/l seems to
produce K+ depletion in at least half of the cases. The situation is less clear when oral intake is
unrestricted and K + concentration in the dialysis fluid ranges between 0 and 1 mmol/l.
Apparently, K+ depletion may develop in 50% of the cases when K+ is omitted from the
dialysis bath and the duration of dialysis is increased to 36 h/week (Johnson et al., 1969).
Conversely, frequent dialysis with 1 mmol of K+/1 in the dialysis bath coupled with a liberal
protein intake apparently restores K + balance in uraemic patients (Ram & Chisholm, 1969).
Though there is no information in this latter study concerning the oral K + intake, it was probably unrestricted. It is possible that in some patients K+ depletion occurred before dialysis
Potassium in uraemia
683
and was not corrected. Potassium depletion has been reported in some or most uraemic nondialysed patients (Bittar, Watt, Pateras & Parish, 1962; Moore, Olesen, McMurrey, Parker,
Ball & Boyden, 1963; Villamil, Yeyati, Rubianes & Taquini, 1963; Ram & Chisholm, 1969;
Graham, Lawson & Linton, 1970). Two patients of our group who were dialysed only two and
five times had a K+ depletion of the same degree as the others.
In conclusion, it may be advisable not to restrict oral K+ intake in uraemic patients under
maintenance haemodialysis and not to keep K+ concentration in the dialysis bath below 1
mmol/l. Should hyperkalaemia become a problem between dialyses, it is probably better to
treat it by increasing the weekly hours of dialysis treatment rather than by restricting the K+
content of the diet and/or decreasing the K+ concentration in the dialysis bath.
ACKNOWLEDGMENTS
The authors are greatly indebted to Angel G6mez Macias, D.A., for his invaluable help in
computer processing of the data. Computing assistance was obtained from the Computing
Center of the Faculty of Medicine, University of Buenos Aires, Argentina. The work was
supported by USPHS Grant HE 11244.
REFERENCES
B I ~ A RE.E.,
, WAIT, M.F., PATERAS,
V.R. & PARISH, A.E. (1962) The pH of muscle in Laennec’s cirrhosis and
uraemia. Clinical Science, 23, 265-276.
BOLTE,H.D., RIECKE,G. & ROHL,D. (1964) Measurements of membrane potential of individual muscle cells
in normal men and patients with renal insUaciency. Proceedings of Second International Congress of
Nephrology (Prague, 1963), pp. 114-117. Karger, Basel.
CLELAND,
J., PLUTH,J.R., TANXE,W.N. & KIRKLIN,
J.W. (1966) Blood volume and body fluid compartment
changes soon after closed and open intra-cardiac surgery. Journal of Thoracic and Cardiovascular Surgery,
52,698-705.
CORSA,L., JR, OLNEY,J.M., JR, STEENBURG,
R.W., BALL,M.R. & MOORE,F.D. (1950) The measurement of
exchangeable potassium in man by isotope dilution. Journal of CIinicaI Investigation, 29, 1280-1295.
CUNNINGHAM,J.N., CARTER,
N.W., RECTOR,F.C., JR & SELDIN,
D.W. (1971) Resting transmembrane potential
difference of skeletal muscle in normal subjects and severely ill patients. Journal of Clinical Investigation, 50,
49-59.
EDELMAN,
I.S., H a w , H.B., SCHLOERB,
P.R., SHELDON,
D.B., FRIIS-HANSEN,
B.J., STOLL,G. & MOORE,
F.D. (1952) Further observations on total body water. 1, Normal values throughout the life span. Surgery,
GynecoIogy and Obstetrics, 95, 1-12.
FLEAR,
C.T.G., COOKE,W.T.& QUINTON,
A. (1967) Serum potassium levels as an index of body content.
Lancet, i, 458459.
GRAHAM,J.A., LAWSON,
D.H. & LINTON,A.L. (1970) Muscle biopsy water and electrolyte contents in chronic
renal failure. Clinical Science, 38, 583-591.
HUGHES,D., WILLIAMS,
R.E. & SMITH,A.H. (1967) Clinical studies on whole body potassium content measured
by gamma-ray spectrometry in health and disease. Clinical Science, 32,503-509.
JOHNSON,
W.J., FROHMERT,
P.P. & NOVAK,L.P. (1969) Potassium balance in patients maintained by long-term
hemodialysis against potassium-free dialysate. Proceedings of the Fourth ZnternationaI Congress of Nephrology (Stockholm, 1969), p. 303. Karger, Basel.
JOHNY,K.V., LAWRENCE,
J.R., O’HALLORAN,
M.W., WELLBY, M.L. & WORTHLEY,
B.W. (1970) Studies on total
body serum and erythrocyte potassium in patients on maintenance hemodialysis. The value of erythrocyte
potassium as a measure of body potassium. Nephron, 7,230-240.
JOSKE, R.A., MCALISTER,
J.M. & PRANKERD,
T.A.J. (1956) Isotope investigations of red cell production and
destruction in chronic renal disease. Clinical Science, 15, 511-522.
C
V. Rettori et al.
C.R.B. (1958) Uptake of potassium by parts of packed human blood cell column. Quarterly Journal of
Experimental Physiology, 43, 299-309.
KEITEL,H.G., BERMAN,
H., JONES,H. & MACLACHLAN,
E. (1955) The chemical composition of normal human
red blood cells, including variability among centrifuged cells. Blood, 10, 370-376.
KRAGELUND,
E. & DYRBYE,
M.O. (1967) Sulphate space in the human organism after intravenous a d m i t r a tion of radiosulphate. ScandinavianJournal of Clinical and Laboratory Investigation, 19, 319-324.
MOORE,F.D., OLSEN, K.H., MCMURREY,
J.D., PARKER,
V.H., BALL, M.R. & BOYDEN,C.M. (1963) The Body
Cell Mass and its Supporting Environment. Saunders, Philadelphia.
MORGAN,
A.G., BURKINSHAW,
L., ROBINSON,
P.J.A. & ROSEN,S.M. (1970) Potassium balance and acid-base
changes in patients undergoing regular haemodialysis therapy. British Medical Journal, i, 799-783.
MULDOWNEY,
F.R., CROOKES,
J. & BLUHM,M.M. (1957) The relationship of total exchangeable potassium and
chloride to lean body mass, red cell mass and creatinine excretion in man. Journal of Clinical Investigation,
JOYCE,
36, 1375-1381.
PACE,N. & RATHBUN,E.N. (1945) Studies on body composition. iii, The body water and chemically combined
nitrogen content in relation to fat content. Journal of Biological Chemistry, 158, 685-691.
RAM,M.D. & CHISHOLM,
G.D. (1969) Blood potassium in patients undergoing haemodialysis. Lancet, i, 260.
ROTH,E., LAX,L.C. & MALONEY,
J.V. (1969) Ringer’s lactate solution and extracellular fluid volume in the
surgical patient: a critical analysis. Annals of Surgery, 169, 149-164.
SCHWAB,
M.W., DISSMANN,
TH. & SCHUBERT,
W. (1963) Der Einfluss des Alters auf die Fliissigkeitsriiume des
Korpers. KIinische Wochenschrif, 31,1173-1181.
SEEDAT,Y.K. (1969) Exchangeable potassium study in patients undergoing chronic haemodialysis. British
Medical Journal, ii, 344-345.
STEELE,
J.M., BERGER,
E.Y., DUNNING,
M.F. & BRODIE,
B.B. (1950) Total body water in man. American Journal
OfPhysiology, 162, 313-317.
VEALL,N. & V E ~ RH., (1958) Radioisotope Techniques in clinical Research and Diagnosis. Buttenvorths,
London.
V., SIMPSON,E.F. & KLEEMAN,
C.R. (1970) A defect of red blood cell potassium
VILLAMIL,
M.F., RETTORI,
transport in acute myocardial infarction. Journal of Laboratory and Clinical Medicine, 76, 383-390.
VILLAMIL,M.F., YEYATI,
N., RUBIANES,
C. & TAQUINI,
A.C. (1963) Water and electrolyte of muscle in chronic
renal failure. Acta Physiologica Latino-Americana, 13, 184-192.
A. (1953) An evaluation of radiosulfate for the determination of the
WALSER, M., SELDIN,D.W. & GROLLMAN,
volume of extracellular fluid in man and dogs. Journal of Clinical Investigation, 32,299-311.