Exchangeable Sodium and Blood Volume
in Normotensive and Hypertensive Humans
on High and Low Sodium Intake
By WALTE1R J. BROWN,
AND
M.D., FArrH K. BROWN, M.S.,
IQBAL KmsHAN, M.D.
JR.,
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
SUMMARY
Eight normotensive and eight essential hypertensive humans were studied when on
high and low sodium intake. Total 24-hr exchangeable body sodium, blood volume,
serum sodium concentration, hematocrit, body weight, and mean arterial pressures,
when the subjects were supine and upright, were measured. Plasma and nonplasma
sodium fractions, nonplasma weight, and plasma, cell, and whole blood volumes, as
referred to body weight and height, were calculated. Changes in these parameters
with diet were also calculated. Specific differences between normotensives and hypertensives were found, in absolute values and with changes with diet. The hypertensives
fell into two groups with regard to exchangeable sodium: in four the values varied
with diet and in the other four, the values remained low regardless of diet. The
"variable-sodium" patients had high upright arterial pressures when on high salt intake and normal upright pressures when on low salt intake. "Low-sodium" patients
had high arterial pressures on both diets. Changes in upright pressures of hypertensives
correlated significantly (r = 0.73) with changes in nonplasma exchangeable sodium.
Additional Indexing Words:
Hypertension (human)
Sodium intake in diet
RESTRICTION of the sodium intake of
patients with essential hypertension often helps lower their arterial pressures. This
effect has been attributed, at least in theory, to
secondary reduction of blood volume (plasma
fraction ),' alteration of sodium within and
Plasma volume
Blood cell volume
around smooth muscle cells,2 3 and to changes
in amount and composition of extracellular
fluid, among other possible causes.4
Because unsalted food does not taste very
good, most patients prefer to be treated with
thiazide derivatives, which tend to deplete the
body of sodium.10 Recent studies have investigated combined changes in body sodium and
blood volume after treatment with thiazides."1
It seems well to know what these changes are
when the body is deprived of sodium by
dietary restriction alone.
During a study of relationships between
alterations in exchangeable body sodium
(ENa) and alterations of venous reactivity to
postural changes, we have made measurements on a total of 22 patients on 50 different
days while they received diets designed to
produce marked changes in ENa. It has been
shown'2 that changes in dietary sodium intake
are accompanied by changes in ENa and that
From the Departments of Medicine and Physiology,
Medical College of Georgia, Augusta, Georgia, and the
Department of Physiology and Pharmacology, College
of Veterinary Medicine and the Institute of Comparative Medicine, University of Georgia, Athens, Georgia.
Supported by grants from the Georgia Heart
Association and the National Institutes of Health,
U. S. Public Health Services (HE-00240 and, General
Research Grant 1-S01-FR05601-03). The Clinical
Investigation Unit at Talmadge Memorial Hospital
was supported during this study by Grant FROO61-07
from the National Institutes of Health.
Published as Institute of Comparative Medicine
manuscript no. 796.
Received August 14, 1970; revision accepted for
publication December 16, 1970.
508
Circulation, Volume XLIII,1 April 1971
EXCHANGEABLE SODIUM AND BLOOD VOLUME
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
measurement by isotope dilution gives more
reproducible estimates of such changes in
body sodium content than do input-output
studies.13 As part of this study we also
measured blood volumes, since changes in
vascular filling could influence venous reactivity. In eight normotensive and eight essential
hypertensive subjects we have records of both
ENa and blood volumes when they were on a
salt-loaded diet and again when they were
receiving a salt-poor diet. We measured
weights, hematocrits, and morning arterial
pressures as well. These values and some
relationships among them are reported here.
The findings on venous reactivity are being
prepared for later publication.
Methods
Subjects of this study were sixteen human
volunteers, eight of whom were normotensive and
free of known cardiovascular disease, and eight of
whom were currently being seen for hypertension
of unknown etiology (i.e., "essential" hypertension) at Talmadge Memorial Hospital in Augusta,
Georgia. Both sets of subjects included males and
females over a broad range of ages, weights, and
heights. Normotensive subjects were six males
and two females, ranging in age from 27 to 58
years (mean, 40 ± 10), in weight from 57 to 80
kg (mean, 66.4 ± 8.5), and in height from 157 to
183 cm (mean, 171 -+- 8). Two were Negro; the
rest were white. Hypertensive subjects were five
males and three females, ranging in age from 19
to 56 years (mean, 41 ± 14), in weight from 52
to 93 kg (mean, 69.1 + 12.7), and in height from
153 to -180 cm (mean, 167 -+- 9). Three were
Negro; five were white. All hypertensive subjects
were without active medication for at least 2
weeks prior to participation in this study. None
showed evidence of hyperaldosteronism; serum
potassium concentrations were in the normal
range. Renal function was judged to be normal on
the basis of normal blood urea nitrogen and serum
creatinine concentrations and creatinine clearance
values. None of the subjects gave evidence of
cardiac failure before or during the study. No
edema formation was detected during high salt
intake. All subjects were hospitalized on the
Clinical Investigation Unit during their time on
the study, so that dietary sodium intake could be
controlled and urinary output measured. All were
ambulatory.
Diets
The low sodium diet contained less than 20
mEq sodium/day. The high sodium diet consisted
Circulation, Volume XLIII, April 1971
509
of regular hospital meals plus supplementary salt
to make the total intake at least 250 mEq
sodium/day (15 g NaCl). In all but one patient,
the supplement was in the form of tablets taken
at mealtime. That one patient had the supplement added to his food. All patients were allowed
to add salt to food from the shaker if they wished,
when on high salt intake. Their high salt intake
varied, therefore, but in all cases was well above
even a "high-normal" intake of 10 g NaCl/ day.
Immediately after change in diet, urinary
sodium outputs changed rapidly for about 3 days
until they reached a new level. We waited until
sodium outputs had passed this period of rapid
change so that measurement of 24-hr exchangeable sodium could be made during a stabilized
period. The total interval between change of diet
and study day was at least 6 days, most often 7
days, and occasionally slightly longer. Stabilized
daily sodium outputs on high and low intake,
respectively, were 337 + 65 and 14 + 14 mEq
for normotensives, and 297±95 and 26+ 17
mEq for hypertensives. Although these average
values for the two groups differed slightly, there
was considerable overlap, so that average normotensive and hypertensive outputs did not differ
significantly when compared by t-test.
Measurements
Each subject's height was measured on admission. Body weight was measured each morning
before breakfast, after the patient had voided.
Nonplasma weights were calculated as: total
weight - (plasma volume x 1.030). Systolic and
diastolic pressures were measured by sphygmomanometer, with the patient supine and again
after standing for 5 min. Mean pressures, calculated as: diastolic pressure + 0.40 (systolic minus
diastolic), were used in analysis of changes with
diet because they best reflect adequacy of tissue
perfusion.14 Urine was collected, volumes were
measured, and sodium concentration was determined by flame photometer.
Total Exchangeable Sodium
The 24-hr total exchangeable sodium (ENa)
was determined for each study day, using 22Na
(Abbott). On the morning of the day before the
study day, a blood sample was withdrawn from
the subject and centrifuged to yield serum for a 1
ml "blank" sample. Then an injection of saline
solution was given, containing approximately
5/.Ci 22Na. The following procedure was used to
determine the exact amount of radioactivity in the
injected solution. A stock solution of the isotope
wag prepared from the commercially supplied
22Na solution, so that it would contain approximately 1 ,Ci 22Na in each ml of physiological
saline. Six milliliters of this solution were drawn
510
BROWN ET AL.
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
into a syringe, and the syringe was weighed on a
measure unbound iodine, we diluted 1 ml of stock
solution with 4 ml 10% trichloroacetic acid to
balance to the nearest 0.01 mg. Then 1 ml of this
precipitate albumin. The supernatant fluid was
solution was ejected into a test tube for counting
removed after centrifugation, and 1 ml was
by the low-background, well-type scintillation
counter used throughout this study (Picker
counted for 1311 radioactivity. This count, corNuclear Autowell; sodium and iodide backrected for dilution by trichloroacetic acid repregrounds approximately 60 and 50 counts/min,
sented the unbound 131J of the stock solution. The
respectively). The syringe was again weighed.
amount was found to vary considerably from one
The remaining 5 ml were injected intravenously
batch to another (range, <1% to approximately
into the subject, and again the syringe was
7%). However, the amount of free 1311 in any one
weighed. In this manner we obtained highly
batch was constant for as long as 2 weeks after
accurate measurements of the amounts of stock
the shipment was received. Of course, many
solution used for counting and calculating the
different shipments were used during this study
actual radioactivity of the injected solution. We
because the half-life of 1311 is short, and fresh
regarded this degree of accuracy as necessary
isotope must continually be supplied. The average
with these samples because of their high
free 131I content of all batches used was 3.4%.
concentrations of isotope. A very small variation
Counts/min/ml of free 1311 were subtracted
in quantities could produce a large variation in
from counts/min/ml of the stock solution to yield
actual radioactivity. After counting to one million
effective counts/min/ml. The "zero time" radiocounts in a recorded time, the known number of
activity of the blood was determined by extrapocounts/min (from the stock solution sample of
lation of the slope of the line plotted on
known weight) was expressed as counts/min/mg
semilogarithmic paper from the radioactivity of 1and multiplied by the weight of the injected
ml blood samples drawn 20, 40, and 60 min after
solution to yield total counts/min injected into
injection of the isotope. Blood volume was
the patient. Then 24-hr urine output was
calculated by:
effective counts/min/ml of injected solution
"instantaneous" (zero time) counts/min/mli
Cell volumes and plasma volumes were calculated
collected for counting of radioactivity "lost" via
the urine. Twenty-four hours after the injection, a
by use of venous hematocrits. These were
blood sample was drawn for determination of
determined, in quadruplicate, by centrifugation of
serum 22Na activity by scintillation count. (Low
blood in capillary tubes. No correction was made
concentration samples such as blood and urine
for estimated difference between venous hematowere counted for 30 min.) Concentration of
crit and "whole body" hematocrit, since it was
serum sodium was determined from this sample
possible that "whole body" hematocrit could be
also. Total 24-hr ENa was then calculated as:
altered in unknown ways in hypertensives, whose
total counts/ min received - counts/ min lost in urine
ENa = serum Na ( mEq/cml x
counts! min/ml serum
and the result was then divided by body weight
plasma/interstitial fluid relationships have been
to yield mEq/kg of ENa.
shown to be abnormal.5 For uniformity of
procedure,
we used only venous hematocrit for all
Blood Volume
patients.
Blood volume was measured on the moming of
Serum Electrolytes
the study day, by dilution of 1311 (Risa-131,
Abbott). The usual procedures'5 16 were folSodium and potassium concentrations of serum
lowed, with certain technical modifications for
were determined by flame photometer for samples
improvement of accuracy and reliability. A
taken on the morning of the study day. These
description of these modifications follows. The
measurements were made in duplicate in each of
measurement of samples containing high concentwo separate laboratories. Agreement between
tration of isotope was done by weighing the
values from the two laboratories was good,
syringe, as described for the 22Na procedure. The
usually differing by not more than 2 mEq/liter.
When values were not identical, a calculated
measurement of "free" 1311 was judged necessary
average value was used.
because unbound ("free") iodine is removed very
rapidly from the circulating blood and should not
Results
be considered as part of the equilibrating
Normotensive
Controls
(Table 1)
quantity of injected 131I. To do so would give
These
subjects
had
lower
plasma volumes,
falsely high values for the diluting volume. To
Circulation, Volume XLIII, Api 1971
EXCHANGEABLE SODIUM AND BLOOD VOLUME
0
0
6
A
511
N
d
d
0
0
A
A
A
A
A
A A A
0
._
"0
A A
A A
~4
00
-
66dZ
0o0
PC0 m mm
oo
C> ",d c
0
A
o~~~~~~ oI'
z
m
o
r-
OOC
m)S
t_~"4i
o,:
U
6o
z
in
0 to 0
A Ao
o
oCt
cc
0CCo
o
t--c =
_
In = L
Cq
c* O CtD
Ci 00 _~ CS -~ Cn
A
~
o
.
L
o~~
c::
C
zz.
C4 CNSoOCm
-o
o
t- _ =
C OCc
dO
C;
cc Le
krt-L
cn CS
0
0
++
42
c3
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
r-
0
C.,
4)
X CO CM
N
Cq
°°R 't
MK t CS
mP
M O
t
C
p
-o
be
0
0
ce
0)
LM cq
m
B
0
CM CN
Co
rk.
1.1CO cq
-
C0
alO cl:
-4)
r.'40
CO
m1-H
0
O=
m
CO
M
COo
t9o
X
_ea
0
CO
-
In en
CD
= o CS
mOo~~~~C co
et
Co
0° Mo
N cs
--
4._
t- C.
P-
_1 N
,,
42
4)
c32
"° o. cl
4)
0
CQ
o
z
.5
CO
ce
z
p
Co
M
Mn
o
CO
M
CO
000
-H
M
m
u-
zg
m t-m
o o o
_- r_ c
CS.
M
,c
m
tz
ce
=s
g
be
CO
-Hm
cfl
0
c
kmH
-H-
00
b
0
0
=~
M
-c
pm
0
C-
to
-
0 23
4. C
U '"
c,
)o
._
Cd
p~~S
4)2O
2)C)
42~c
CiclainVo
eXLII,Ap) 97
C))2
9-,p
d
~.
01
C
"
BROWN ET AL.
512
N
-5 C~ CScs
CS
_
0
A A A
AA
o o 0
6
00
C C=
6C
C
m
-4
A\ A A
00 0
PO
AAA
C
A
A66
A 6C
AAAd
Co
t-
_..4 e_- M
c Ct cm
-"o C>
c,]
m QC
---1
:
?.-
Ce
o
..~~~~~~~iA
>C
cp
c
oLCL00 oi3oi
*0 zz
zzG
Zo
*
1~4 00
X
0n
_q
'
.
m m i 00
P-4
oo
O
o
0d
m
^;i -9 ~.
i5
C; CO
c
CS
.
C
zz
0o t-.
cd d
o
CS Ct
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
ce oc
0a)Lz
"-
to
1C
O" t-me
0
00
C0
eq
D
.
--
_-
*C-.C';iC: ns
c8
1=
C.S
cl)
0
cq
.~ .
LC =
o5
O " C0
0
(M t~-
-
in
_~
Ci
c
o
'Itt
-
c:0f
Cb
0
CO
C5>
M
in cc1
c4
-1
4
~4
ci
0c
0~
4)
*3
10
kC-
oo oq
C.
c
mS oo
C
Cd
0C
0
mO COrl
0
L
C cqUDtC)
X C "
0-
in
I-
cc
m~
0
LO
cc
m:
0
eci
c,
=
"
m
CIG?.)
c
Le
m < "_4
cs
CO
m
cq00 cqI
m
cq
-
',I,
m
Cq
-_
*.;M
O
11
"-
*
000
-
.
c,]
cq
00
CS~
coo000
o-
m
cc1
0
000
00
$
CS M
~
e'q
q
00
A A
C?I
6r_
c6
CCI
_
:-:5b
bE
C.
.; ,X
;
E
X
bE
w 0
z
g
:3
-
W
X
E°Sgga
8..o>4.
v
v
Circulation, Volume XLIII, April 1971
EXCHANGEABLE SODIUM AND BLOOD VOLUME
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
lower nonplasma weights, and smaller
amounts of sodium in the plasma at lower
concentrations when they were on low salt
intake than when they were on high salt
intake. The differences in plasma volume were
evident whether related to weight or height,
or treated as absolute values. They were
reflected in similar differences in absolute
blood volume, and blood volume as related to
height. They were obscured wnen blood
-olume was related to body weight, because
the nonplasma body weight also changed with
diet. The alterations in plasma volume were
also evidenced in hematocrits, which reflected
hemoconcentration on low salt intake. Cell
volumes remained constant regardless of diet.
Mean arterial pressures showed no significant
changes with diet, and individuals' changes in
arterial pressure showed no correlation with
changes in other parameters.
Hypertensive Patients (Table 2)
As a group, these individuals showed no
significant changes in blood volumes with diet.
Their plasma volumes, as absolute values or as
related to weight, showed no significant
changes. When related to height, their plasma
volumes were slightly smaller on low than on
high sodium intake. Sodium content and
concentration of plasma and nonplasma tissues
were significantly smaller on low than on high
salt intake. Total weight, and the nonplasma
fraction of it, were both significantly smaller
on low than on high sodium intake. Cell
volumes remained relatively constant. The
average hematocrits on both diets were
comparatively high, reflecting hemoconcentration in some individuals. However, the variation within the group was such that these
average values are not significantly greater
than those of the controls. All mean arterial
pressures for this group were significantly
higher than comparable pressures for the
control group. No significant pressure changes
with diet were evidenced for the group as a
whole.
The figures in tables 1 and 2 demonstrated
several specific differences between the hypertensive and control groups. There was also one
general difference between the two groups:
Circulation, Volume XLIII, Apfil 1971
513
the values from hypertensives usually covered
a wider range than did the values from the
normotensives. That is, there was more
individual variation among hypertensives than
among control subjects. In regard to plasma
sodium and plasma volumes, this was not the
case, but in nearly all other measurements it
was.
This observation suggested the possibility
that the hypertensives could be divided into
subgroups that would show greater uniformity
of values and responses. Examination of
changes in the 24-hr exchangeable sodium
values indicated that these patients showed
two different kinds of response to dietary
sodium changes. Four patients had higher
than average values on high sodium intake,
and dropped to lower than average values on
low intake. The other four had lower than
average values regardless of diet. Therefore,
we divided the hypertensive group on the
basis of their ENa values into two subgroups
and designated them "variable-sodium hypertensives" and 'low-sodium hypertensives."
This subdivision was made for the purpose of
closer analysis of the variations seen among
the eight hypertensive patients reported here.
It would not be appropriate to project such a
classification to include all patients with
essential hypertension, on the basis of this
small sample. The analysis is presented in
tables 3 and 4.
Variable-Sodium Hypertensives (Table 3)
These patients had expanded and contracted blood volume (plasma fraction only)
with salt loading and restriction. They also
showed changes in plasma sodium content
with diet, but the changes in concentration
(i.e., serum mEq/liter) were slightly more
variable and did not meet the test of
significance. Their changes in sodium content
and concentration in nonplasma tissues were
significant, as were the changes in weight of
their nonplasma tissues. Hematocrits- of two of
these individuals were high (>48) even on
high sodium intake when their plasma volumes were relatively expanded (as compared
to their own low-sodium plasma volumes).
When sodium intake was low, the average
OO4cb
BROWN ET AL.
514
~4
CQ
o c
A A A
C.. P4 Pi
A AA
le L-
d
c
d
C5
A A
Zs C..
A A
A
A
At A
1
c. o=. o
C.O
c
d d
o o
= c co
cq -cq t- soe
oo m me 1- = ds4 t 00
m C: =
1-0 10
dId --- .~ "
1.1
id
.L .~
-1 1.0 *l
g
A
CS
A
A
A
d
A
14
.l
en.C
c:4
_
m.~ m
.l
A
A A
G COlp
O- t~
o6 Ino
0
r-
d
~c
GC
-0 t-
=
---
Gd Gd Cd G.
N N 00 =
t_m
10 CD )L N
.
..4 O ~~~~~
.^ C?:> CS
X ~~~~~~~~~~~~
.
Gd Gd
9
.
.
N~~
cl
O
o
:o
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
O= m
cz
:
._
Oe
cs
cc
m
---
m
c:
C
_I
m to
oc:
C: cXs 00
00..
lc=
n.o
c9 00 C:>
I" e
q
_. _-
CO
m
c;
.>
CO
u:
4@
o
Ei x
o cq
O
c
-
ce =
= c
0
c$m
m
.gt
Z -H
0
PO
9
E..4
:k0
,Q
0
c$
4)
a
CS
cc
gN 0
o-m cc CCX
cq
*
Ld
Lt 1-1
c
CQ
4C5> °
m N cq
c
-H
m
m
"
OC
m0 =
._
m.
m
=n
N
N
d
CD
XTtl
_
Co =
z
m m
O= = N
CX
N
c: C=
c
cc
,f
m
C)
=d
et
F
X
Li
X
c)
4)
=
.>U2 z
N
-4
a:
_
,
1= c, h,.
cn
n:
3
OD c cc
m m mll m
IR1
c) =
<D *
_
zv
.
r cz
O r
iP
X
C.t
ce
cc
.
gv
=ua~ ~ ~ ~ ~ ~ ~ca~ ~ ,x~
a a = E~ ~ ~ ~ ~ ~ ~c
.;
I.1
g;
C9
_2
I.i
Bw i gwX i w> a =
a
E w <
;a<>o_~ ~ ~ ~ ~ ~ ~c
s i .B
S a S i A~~~~~~~O
ce~~~~~~~Cicltoz
omeXll
pl17
EXCHANGEABLE SODIUM AND BLOOD VOLUME
515
Table 4
Low-Sodium Hypertensive Patients*
Na intake in diet
High Na
Low Na
(mean 4i SD)
(mean 4 SD)
Parameter
ENa (mEq)
Total
Plasma
2514
312
2202
Nonplasma
Weights (kg)
Total
73.023
70.718
Nonplasma
ENa concentrations
Total (ENa/kg)
Serum (mEq Na/liter)
Nonplasma mEq/nonplasma kg
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
Blood volumes
Total ml
Plasma volumes
Total ml
ml/kg
ml/cm
Cell volumes
Total ml
ml/kg
ml/cm
Hematocrits
Mean arterial pressures (mm Hg)
Supine
Upright
-
17
17
71.238
68.968
-
-
3.4t
1.7
33.30
137.3
30.00
-
-
-
-
3.2t
-
960
-
-
-
-
t
P
640
69
570
134
9
124
2.051
0.847
2.015
N.S.
N.S.
N.S.
16
16
1.785
1.751
2.328
2.085
N.S.
N.S.
4.3
1.7
4.0
1.28
2.0
1.25
0.970
1.633
1.000
N.S.
N.S.
N.S.
-6
-1.3
0
-0.039
-0.437
0.025
N.S.
N.S.
N.S.
4.4
3922
54.8
23.2
1100
7.1
5.2
2237
31.2
13.3
360
4.0
1.3
2204
31.2
13.1
500
4.0
2.1
33
0
0.2
0.432
0
0.504
N.S.
N.S.
N.S.
1679
22.3
9.9
42
680
3.9
3.6
7
1718
23.6
10.1
43
630
4.3
3.2
5
-39
-1.3
-0.2
-1
-0.307
-0.717
-0.339
-0.870
N.S.
N.S.
N.S.
N.S.
21§
133
122
16§
17§
-3
-6
-0.351
-1.032
N.S.
N.S.
3916
53.5
23.2
ml/kg
ml/cm
*4 patients.
tO.05 > P >
JP = 0.05
§0.01 > P
34.58
139.3
31.25
2380
303
2078
560
51
510
Difference
130
116
2.3t
-
14t
-
0-02 }Significance levels of differences between hypertensives' 'values and normotensives' values
for same parameter and condition when compared by (Group Mean t-test.28
hematocrit of the group was signfficantly
greater than normal. Cell volumes showed no
significant changes or differences from comparable control values.
Mean supine arterial pressures in these
individuals were virtually identical on high
and low sodium intake. On low intake, the
mean pressure was significantly greater than
the comparable normotensive control value;
on high sodium intake, the variable-sodium
hypertensives' and normotensives' mean pressures were not statistically different, apparently because the normotensives' pressures were
at their highest average level in these
circumstances and not because the hypertensives' pressures were low.
Mean upright arterial pressures showed a
Cisrculation, Volume XLIII, April 1971
wider margin between high and low sodium
intake values, and each individual within this
group did have lower mean pressure on low
sodium intake. Even so, the average drop in
pressure of 20 mm Hg did not meet the test of
significance, probably because of the wide
range of pressure differences (-4 to -46) and
the small size of the group. The mean upright
pressure on high sodium intake was significantly higher than normal, and the mean
upright pressure on low sodium intake was not
significantly greater than normal.
Low-Sodium Hypertensives (Table 4)
These patients maintained relatively constant blood volumes (both plasma and cell
fractions), sodium content, concentration of
BROWN ET AL.
516
Table 5
Correlations Between Changes in Mean Upright Arterial Pressure and Changes in
Other Variables in All Subjects
A Pressure vs.:
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
A Total mEq ENa
A Plasma mEq ENa
A Nonplasma mEq ENa
A Total weight
A Nonplasma weight
A Total ENa/kg
A Serum mEq Na/liter
A Nonplasma mEq ENa/nonplasma kg
A Blood volume (ml)
A Blood volume (ml/kg)
A Blood volume (ml/cm)
A Plasma volume (ml)
A Plasma volume (ml/kg)
A Plasma volume (ml/em)
A Cell volume (ml)
A Cell volume (ml/kg)
A Cell volume (ml/cm)
A Hematocrit
Correlation coefficients
Hypertensives*
Normotensives*
0.72t
-0.11
0.19
-0.15
-0.38
-0.26
-0.04
-0.40
- 0.08
0.24
0.29
0.24
0.23
0.29
0.24
0.16
0.22
0.14
-0.25
0.15
0.73t
0.27
0.21
0.65
0.14
0.67
0.06
0.02
0.05
0.13
0.06
0.13
-0.04
-0.06
-0.02
-0.13
*8 subjects.
tSignificant at a/c level.
plasma and nonplasma tissues, and hematocrits, regardless of diet. All these values were
comparable to those of normotensive controls
on low sodium intake. They showed considerable variation in nonplasma weight, such that
for the group the average changes were not
significant.
These patients had higher than normal
mean arterial pressures in both supine and
upright positions, regardless of diet.
The observation that variable-sodium hypertensives had upright arterial pressures in
the normal range when their salt intake was
restricted, whereas the low-sodium patients
maintained their high pressures, suggested
that the drop seen in pressure could be related
to the degree of depletion of exchangeable
sodium. The variable-sodium hypertensives
also lost plasma volume, and that, too, could
conceivably be related to the pressure drop.
Therefore, we calculated the correlations
between the change in mean upright arterial
pressure with diet, and all the other changes
reported here. These correlations are presented in table 5. It can be seen that the only
statistically significant correlations are between changes in pressure in hypertensives
and their changes in exchangeable body
sodium content, specifically the nonplasma
fraction of that sodium content. Correlations
between sodium concentrations (A Total ENa
mEq/kg and A Nonplasmna mEq/nonplasma
kg) and A Pressure reflected this relationship
also, with r values of 0.65 and 0.67, respectively. These coefficients have a confidence level
for a sample of this size between 5 and 10%,
which is higher than customarily accepted as a
test of significance. No other correlations in
table 5 even approach a significant level.
Discussion
Sodium Changes
The most apparent conclusion that can be
drawn from the results of this study is that the
fall in mean arterial pressure seen in some
hypertensive patients when on a low-salt diet
is closely related to their loss of exchangeable
sodium from nonplasma tissues. Interestingly,
it was the change in pressure that was
correlated with the change in nonplasma
sodium. When we tested for correlations
Circulation, Volume XLIII, Apri 1971
EXCHANGEABLE SODIUM AND BLOOD VOLUME
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
between absolute pressures and absolute
nonplasma sodium values, we found none that
were significant.
Many efforts have been made to demonstrate abnormality of sodium content and
concentration of tissues and fluids of hypertensives. Early studies, especially, reported
high serum sodium concentrations. More
recent studies, including this one, emphasize
the frequency with which these concentrations
are in the low-normal range.17 Clear abnormality of total exchangeable body sodium has
been difficult to demonstrate except in cases
complicated by heart failure.8' 18 However,
high sodium content of certain vascular tissues
taken from hypertensive animals and humans
has been repeatedly demonstrated. Also,
increased water content ("water-logging") of
these tissues has been implicated.9
The Friedmans3 have pointed to the importance of dynamic changes in sodium equilibrium across cell membranes in determining
vascular muscle contraction. They have demonstrated, in rat tail vasculature, that changes
in sodium equilibrium are accompanied by
changes in smooth muscle contraction. The
results that we report here are compatible
with the thesis that changes in sodium in the
environment induce changes in vascular contraction. Specifically, in the hypertensive
patients we studied, when the amount of
nonplasma exchangeable sodium was reduced
in response to dietary restriction, the arterial
pressure dropped in proportion to that sodium
reduction.
Of course, vascular muscle contraction is
not the only determinant of arterial pressure.
To what extent changes in heart rate or
strength of contraction entered this response
we do not know. Changes in blood viscosity
should not be discounted. Some hematocrits
were definitely high. Because the cardiovascular system is highly integrated, it seems certain
that changes in these parameters, and in
others as well, must have their effect on
measured changes in mean pressure. Nevertheless, the most clearly demonstrable relationship in these patients was between the
Circulation, Volume XLIII, April 1971
517
changes in nonplasma sodium and the changes
in mean upright arterial pressure.
Plasma Volume Changes
The second conclusion is that changes in
plasma volume induced by alteration of
sodium intake bear little direct relationship to
arterial pressure changes in the patients
studied here. Reports frequently link these
two effects of sodium depletion, on the basis
that reduction in filling volume of a closed
system should result in reduction of filling
pressure. In a static system (such as a gasfilled reservoir of fixed capacity) this effect
would occur, but the circulatory system is far
from static. In the normal, active person, a
reduction in blood volume will be accompanied by a decrease in capacity due to
vasoconstriction, so that pressure is maintained. Cardioacceleration is part of this
sympathetic response, also.'9 Thus, it is not
surprising that the reductions in plasma
volume reported here were not correlated
with reductions in pressure. Of course, if
sympathetic vasoconstriction is prevented by
treatment with an autonomic blocking agent,
then reduction in plasma volume is accompanied by a drop in pressure.20
The observation that there was no significant correlation between changes in pressure
and changes in nonplasma weight seems to
indicate that changes in extracellular water do
not influence arterial pressure. This conclusion
is not warranted from these data. We have no
way of knowing what fraction of the nonplasma weight change was "free" extracellular
water, what fraction was intracellular water,
and what fraction may have been associated
with electrolyte-binding matrices surrounding
cells. These water fractions could be changing
in the same or different directions, and the
measurements made here could not distinguish such changes. Studies do indicate that
the relationship between plasma volume and
extracellular water is altered in some hypertensive individuals,5 but the exact nature of
this alteration has been difficult to define.
Blood Volume Regulation
Studies have shown that blood volume,
BROWN ET AL.
518
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
especially the plasma fraction, tends to be
lower than normal in hypertensive patients.5 21
This abnormality is especially striking if volumes are expressed in relation to body weight,
and this was true of the patients reported here.
The four "low-sodum" hypertensives had blood
volume/kg values, when on high salt intake,
which were significantly lower than comparable values for normotensives, and they showed
virtually no change with diet. Thus, these individuals appeared to be regulating their
blood volume so as to maintain it at a low
level, comparable to that of normotensives on
low salt intake, whether their actual salt intake
was high or low.
What determines the amount of blood in
the vasculature? Although much is known
about the regulation of blood volume, clearly,
much remains to be understood. Studies have
shown that the volume of blood tends to vary
closely with the longitudinal dimension of the
body.'9 22 Since the main, large distributing
and collecting vessels are oriented longitudinally, this observation suggests that the degree
of filling of these large vessels is a major
determinant of total blood volume. This idea
would apply especially to the veins, which
contain a larger proportion of the total blood
volume than the arteries, and it concurs with
the belief that volume receptors exist, at some
point in the venous capacitance system, which
function to maintain constancy of blood
volume.'9
Following this line of reasoning, one can see
that a short person who carries 20 lb. of
surplus tissue (i.e., fat), requiring perfusion,
would be at a greater circulatory disadvantage
than a tall person carrying an extra 20 lb. It is
interesting to note that shorter and heavier
people are statistically more subject to hyper-
tension.23' 24
Normal Values
Since the introduction of radioisotope dilution techniques for measurements on humans,
considerable information on the normal range
of values for blood volumes and body
electrolytes has been accumulating.25 Each
additional set of observations helps in understanding normal processes, as do those pre-
sented here pertaining to regulation of blood
volumes and electrolyte concentrations when
dietary sodium is increased or diminished.
Diagnosis and treatment of abnormal states
will have their greatest chance for success
when normal responses are well understood.
Acknowledgment
We gratefully acknowledge the advice and assistance of Mr. Al Irby, formerly of the Medical College
of Georgia, and Mr. Rollie Harp of the University of
Georgia Computer Center, with statistical work
related to this report. Also we want to express our
appreciation to Dr. Philip Dow of the Medical
College of Georgia for his valuable consultations and
support.
References
1. HOUSEL EL: Diet and anti-hypertensive therapy.
In Hypertension: Recent Advances, edited by
AN Brest, JH Moyer. Philadelphia, Lea &
Febiger, 1961, p 482
2. FRUIEDMAN SM, FRiEDMAN CL: Effects of ions on
vascular smooth muscle. In Handbook of
Physiology II: Circulation, Vol. II, edited by
WF Hamilton, P Dow. Washington, D.C.,
American Physiological Society, 1963, pp
1135-1166
3. FRIEDMAN SM, FRIEDMAN CL, NAKASHIMA M:
Sodium and the regulation of peripheral
vascular resistance. In Hypertension XIII,
edited by S. Rodbard, LA Saperstein, JE Wood.
Proceedings of the Council for High Blood
Pressure Research, American Heart Association. New York, American Heart Association,
1965, pp 178-186
4. PAGE IH, MCCUBBIN JW: The physiology of
arterial hypertension. In Handbook of Physiology II: Circulation, Vol. III, edited by WF
Hamilton, P Dow. Washington, D.C., American Physiological Society, 1965, pp 2163-2208
5. TARAZI RC, DUSTAN HP, FROHLICH ED: Relation
of plasma to interstitial fluid volume in
essential hypertension. Circulation 40: 357365, 1969
6. FRIEDMAN SM, FRIEDMAN CL: The ionic matrix
of vasoconstrictioii. Circ Res 20-21 (suppl. II):
II-147-155, 1966
7. DAVIDOV M, GAVIuLOVICH L, MROCZEK W, FINNERTY F: Relation of extracellular fluid volume
to arterial pressure during drug-induced saluresis. Circulation 40: 349-355, 1969
8. HOLLANDER W, CHOBANIAN AV, BuRmows BA:
Body fluid and electrolyte composition in
arterial hypertension. I. Studies in essential,
renal, and malignant hypertension. j Clin
Invest 40: 408-415, 1961
9. TOBIAN L JR: The relationship of sodiumn to
Circulation, Volume XLIII, April 1971
EXCHANGEABLE SODIUM AND BLOOD VOLUME
hypertension. (Clinical observations). In Hypertension: Recent Advances, edited by AN
Brest, JH Moyer. Philadelphia, Lea & Febiger,
1961, pp 90-94
10. FUCHS M: Human pharmacology of thiazide derivatives. In Hypertension: Recent Advances,
edited by AN Brest, JH Moyer. Philadelphia,
Lea & Febiger, 1961, pp 247-249
11. HANSEN J: Hydrochlorothiazide in the treatment
of hypertension. Acta Med Scand 183:
317-321, 1968
12. DoLE VP, DAHL LK, CORTZIAS CC, DzIEWIATKOWSKI DD, HAmuS C: Dietary treatment of
hypertension. II. Sodium depletion as related
to the therapeutic effect. J Clin Invest 30:
584, 1951
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
13. WILSON GM, OLNEY JM, BROOKS L, MYRDEN J,
BALL MR, MOORE FD: Body sodium and
potassium II. A comparison of metabolic
balance and isotope dilution methods of study.
Metabolism 3: 324, 1954
14. GUYTON AC: Regulation of mean arterial
pressure, and hypertension. In Textbook of
Medical Physiology, Ed 3, Philadelphia, W. B.
Saunders Co., 1968, pp 353-354
15. SILVER S: Radioactive Isotopes in Medicine and
Biology. Philadelphia, Lea & Febiger, 1962, pp
196, 231
16. BAKER CH: Cr51-labeled red cell, I131-fibrinogen, and T-1824 dilution spaces. Amer J
Physiol 204: 176-180, 1963
17. FROHLICH ED: Plasma sodium, potassium,
calcium, and magnesium concentrations in
essential hypertension. Amer J Med Sci 248:
419-423, 1964
18. Ross EJ: Total exchangeable sodium in hypertensive patients. Clin Sci 15: 81-91, 1956
Circulation, Volume XLIJI, April 1971
519
19. SJOSTRAND T: Blood volume. In Handbook of
Physiology II: Circulation, Vol. I, edited by
WF Hamilton, P Dow. Washington, D.C.,
American Physiological Society, 1962, pp
51-62
20. WELLER JM, HOOBLER S: Salt metabolism in
hypertension. Ann Intern Med 50: 106-114,
1959
21. TIBBLIN G, BERGENTZ SE, BJURE J, WILHELMSEN L: Hematocrit, plasma protein, plasma
volume, and viscosity in early hypertensive disease. Amer Heart J 72: 165-176, 1966
22. CHIEN S, UsAMi S, SIMMONS RL, McALLusTER
FF, GREGERSEN MI: Blood volume and age:
repeated measurements on normal men after 17
years. J Appl Physiol 21: 583-588, 1966
23. SPIEGELMAN M, THOMPSON KJ: Short range
blood pressure changes in a group of middleaged males. In The Epidemiology of Hypertension, edited by J Stamler, R Stamler, TN
Pullman. New York, Grune & Stratton, Inc.,
1967, pp 401-402
24. STRANG JM: Obesity. In Diseases of Metabolism,
edited by GG Duncan. Philadelphia, W. B.
Saunders Co., 1965, pp 707-709
25. EDELMAN IS, LEIBMAN J: Anatomy of body
water and electrolytes (review). Amer J Med
27: 256-277, 1959
26. ROESSLER EB: Significant figures in statistical
constants. Science 84: 289-290, 1936
27. BAILEY NTJ: Statistical Methods in Biology.
New York, John Wiley & Sons, 1959, pp
155-157
28. DIXON WJ, MASSEY FJ JR: Introduction to
Statistical Analysis. New York, McGraw-Hill
Book Co., 1957, pp 115-129
Exchangeable Sodium and Blood Volume in Normotensive and Hypertensive
Humans on High and Low Sodium Intake
WALTER J. BROWN, JR., FAITH K. BROWN and IQBAL KRISHAN
Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017
Circulation. 1971;43:508-519
doi: 10.1161/01.CIR.43.4.508
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX
75231
Copyright © 1971 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/43/4/508
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles
originally published in Circulation can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for
which permission is being requested is located, click Request Permissions in the middle column
of the Web page under Services. Further information about this process is available in the
Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/