Isotonic and hypertonic sodium loading in supine humans

Isotonic and hypertonic sodium loading in supine humans

L . J . A N D E R S E N , 1 T. U. S. J E N S E N , 2 M. H. B E S T L E 1,3 and P . B I E 1

1 Department of Medical Physiology, Panum Institute, University of Copenhagen, Copenhagen, Denmark

2 Danish Aerospace Medical Centre of Research, Rigshospitalet, Copenhagen, Denmark

3 Department of Clinical Physiology, Herlev Hospital, Herlev, Denmark

ABSTRACT

The hypothesis that hypertonic saline infusion induces a greater natriuresis than infusion of the same

amount of sodium as isotonic saline was tested in 8 supine subjects on fixed sodium intake of

150 mmol NaCl day±1. Sodium loads equivalent to the amount of sodium contained in 10% of

measured extracellular volume were administered intravenously over 90 min either as isotonic saline

or as hypertonic saline (850 mmol L±1). A third series without saline infusion served as time control.

Experiments lasted 8 h. Water balance and sodium loads were maintained by replacing the excreted

amounts every hour. Plasma sodium concentrations only increased following hypertonic saline

infusion (by 2.7 � 0.3 mmol L±1). Oncotic pressure decreased signi®cantly more with isotonic saline

(4.1 � 0.3 mmHg) than with hypertonic saline (3.2 � 0.2 mmHg), indicating that isotonic saline

induced a stronger volumetric stimulus. Renal sodium excretion increased more than a factor of four

with isotonic and hypertonic saline but also increased during time control (factor of three). Cumulated

sodium excretions following isotonic (131 � 13 mmol) and hypertonic saline (123 � 10 mmol) were

statistically identical exceeding that of time control (81 � 9 mmol). Plasma angiotensin II decreased

in all series but plasma ANP concentrations and urinary excretion rates of endothelin-1 remained

unchanged. In conclusion, hypertonic saline did not produce excess natriuresis. However, as the two

loading procedures induced similar natriureses during different volumetric stimuli, part of the

natriuresis elicited by hypertonic saline could be mediated by stimulation of osmoreceptors involved

in renal sodium excretion. The supine position does not provide stable time control conditions with

regard to renal excretory function.

Keywords angiotensin II, human subjects, osmoreceptors, sodium homeostasis.

Received 5 October 1998, accepted 12 January 1999

It is generally accepted that renal sodium excretion is

controlled primarily by mechanisms sensitive to chan-

ges in body ¯uid volume. However, in several animal

experiments, infusion of hypertonic saline induces a

natriuretic response that exceeds the response to iso-

tonic saline although the latter is associated with a

greater volumetric stimulus (Emmeluth et al. 1990,

Huang et al. 1995). This excess natriuresis may be in-

duced by stimulation of speci®c osmoreceptors in-

volved in the control of renal sodium excretion

(Thornborough et al. 1973, Zucker & Kaley 1976,

Emmeluth et al. 1992, McKinley et al. 1992). Recently,

we found that an increase in plasma sodium concen-

tration may also contribute to the natriuresis of intra-

venous salt loading in normal seated subjects (Andersen

et al. 1998). However, compared with other species, the

response to hypertonic saline infusion seemed to be

relatively small in humans and osmostimulated nat-

riuresis could only be demonstrated in subjects on high

sodium intake (Andersen et al. 1998). It could be hy-

pothesized that the natriuretic response to hypertonic

saline is dependent on the relative hypervolemia that is

associated with ingestion of a high sodium diet. Ac-

cordingly, sodium depletion inhibits osmostimulated

natriuresis in dogs (Gilmore & Nemeh 19841 ). The

purpose of the present investigation was to further

elucidate osmoregulatory control of sodium excretion

following isotonic and hypertonic salt loading. Subjects

were placed in the supine position. Compared with the

seated position this increases the load to central volume

receptors allowing us to investigate whether this mod-

erate hypervolemia could augment the response to

hypertonic vs. isotonic saline loading. In addition, the

duration of the experiment was extended by 3 h

Correspondence: Lars Juel Andersen, MD, Panum Institute, building 6.5.8, 3 Blegdamsvej, DK-2200 Copenhagen, Denmark.

Acta Physiol Scand 1999, 166, 23±30

Ó 1999 Scandinavian Physiological Society 23

compared with our previous protocol (Andersen et al.

1998).

METHODS

Experiments were performed in eight healthy male

volunteers. Subjects were 24±30 years old, weighing

74.4±85.1 kg. All gave informed consent and the study

was approved by the Ethics Committee of Copenhagen

(j.nr. KF 01±011/96).

Approximately 2 weeks prior to the experiments,

each subject's extracellular volume (ECV) was deter-

mined as the distribution volume of [51Cr] EDTA

according to the method described by Brùchner-Mor-

tensen (1982). Brie¯y, 3.5 MBq of [51Cr] EDTA was

injected intravenously and four venous samples were

drawn 180, 200, 220 and 240 min after injection. ECV

measurements and all experiments were performed af-

ter subjects had been on a controlled diet containing

150 mmol NaCl day±1 for 4 days. Sodium turnover

was veri®ed by measuring 24 h urinary sodium excre-

tion the day before the experiment.

The night before the experiment the subject slept at

the laboratory. At 07.00 hours he was awakened and

consumed a light standardized low-salt breakfast (3

slices of toast, 15 g marmalade, and 500 mL of tap

water). The subject was instrumented with two cathe-

ters (Ven¯on, 18 gauge) in super®cial cubital veins for

blood sampling and infusion of saline. After the in-

sertion of the catheters, the subject emptied his bladder

and was weighed. The subject remained supine

throughout the study and was allowed to stand up only

for micturition. The degree of hydration was main-

tained throughout the experiment by drinking of water

in amounts equal to those voided plus 1 mL min±1 as

replacement for the insensible2 water loss. Thirst

sensation was not monitored. Room temperature was

kept around 24 °C.

Each experiment lasted 8 h. After 1 h of baseline,

saline was infused over the following 90 min. After

infusion, sampling continued for another 5� h. Two

series of infusion plus time control were performed on

separate days. In one infusion series (ISOT), a salt load

was infused as 0.9% saline in amounts equal to 10% of

measured ECV, i.e. a load of 1.44 � 0.01 L of isotonic

saline. In the other infusion series (HYPR), an identical

amount of sodium was infused as a 5% solution

(0.26 � 0.01 L of 5% saline). Infusions were adminis-

tered by an automatic infusion pump (LifeCareÒPump, Model 4, Abbott, Ill, USA3 ). The time control

series (CONTR) was identical to infusion series, but

without saline infusion. At least 2 weeks of recovery

were allowed between each experiment.

Urine was collected by voluntary micturition during

the last minute of every hour. After each urine collec-

tion, urinary concentration of sodium was measured and

the excreted amount of sodium was replaced by i.v.

injection of isotonic saline at the beginning of the sub-

sequent period in order to maintain the sodium load.

The injected volume was subtracted from the volume of

water administered orally to maintain hydration status.

Blood (12 mL) was sampled in heparinized tubes at

the beginning and the end of the baseline period, and in

the middle of every remaining 1-h period for mea-

surements of sodium, osmolality, haematocrit, protein

and oncotic pressure. Baseline values are presented as

the mean of the ®rst two blood samples. Additional

blood samples for hormone analyses (20 mL) were

collected in prechilled polyethylene tubes containing

aprotinine (300 KIU mL±1) and EDTA (3 lmol mL±1)

during baseline and in the middle of period 2, 4, 6 and 8.

The withdrawn blood was replaced by twice the

amount of saline. Blood samples were centrifuged at

+4 °C and plasma was stored at )18 °C until analysed

for concentrations of angiotensin II (ANG II) and atrial

natriuretic peptide (ANP).

Haematocrit was determined by centrifugation4 .

Oncotic pressure was determined by a colloid osmo-

meter (4400 Colloid Osmometer, Wescor, Utah, USA5 ).

A SMAC3 instrument (Bayer) was used to measure

haemoglobin and plasma protein concentrations. Urine

and plasma sodium concentrations were measured by

an ion-selective electrode system (KNA-2, Radiometer)

and osmolality by freezing-point depression (model

3DII, Advanced Instruments, MA, USA6 ).

Arterial systolic and diastolic pressures were re-

corded semiautomatically four times every hour

(Propaq 102, Protocol Systems Inc., Or, USA7 ) and

the mean of these values were used to calculate

mean arterial pressure (MAP) from the formula:

MAP � diastolic pressure + 1/3 ´ pulse pressure.

Hormone concentrations in plasma and urine sam-

ples were quanti®ed by radioimmunoassays after ex-

traction. Thawed samples were acidi®ed with 4% acetic

acid and peptides were extracted by use of C18 Sep-Pak

cartridges (Waters, MA, USA8 ), as previously described

(Emmeluth et al. 1994). After elution, the samples were

dried and stored at )18 °C in tubes topped with N2

until analysis for hormone immunoreactivity. ANP was

determined according to the assay procedure described

earlier (SchuÈtten et al. 1987) by use of an antibody

(RAS8798) purchased from Peninsula Laboratories9 ,

CA, USA. Extraction recovery was 85%. Detection

limit was 1.5 pg mL±1 sample. Intra-assay coef®cient of

variation was 6%. Immunoreactivity of ANG II was

determined by use of an antibody (Ab-5±030682)

produced by P. Christensen10 . The assay procedure has

been described by Kappelgaard et al. 197611 ). Extraction

recovery was 85%. Detection limit was 1.4 pg mL±1

sample. Intra-assay coef®cient of variation was 5%. The

Osmostimulated sodium excretion á L J Andersen et al. Acta Physiol Scand 1999, 166, 23±30

24 Ó 1999 Scandinavian Physiological Society

procedure for measuring endothelin-1 concentrations

in extracted urine samples has previously been de-

scribed (Emmeluth et al. 1992). The detection limit was

0.16 pg mL±1 urine and extraction recovery was 91%.

Intra-assay and inter-assay coef®cients of variation were

5.2% and 6.5%, respectively. Hormone results have not

been corrected for incomplete recovery. Measurements

of plasma concentrations of arginine vasopressin failed

for technical reasons.

Statistics

Results are presented as mean � SE. Data were sub-

jected to one-way analysis of variance (ANOVA) for re-

peated measures (Winer 1971). In case of signi®cantly

large F-values, all possible differences were evaluated

by Newman±Keuls' test. Selected differences between

series were evaluated using paired Student's t-test. Level

of signi®cance was in all cases 0.05.

RESULTS

Subjects ECV ranged between 13.9 and 14.8 L with a

mean of 14.4 � 0.1 L (18.1 � 0.2% of body wt.).

Twenty-four hour sodium excretion was 147 �

12 mmol during the last day of controlled sodium intake.

Plasma sodium concentration and osmolality in-

creased during HYPR and peaked immediately after

end of infusion (from 139.2 � 0.3 to 141.9 � 0.3

mmol L±1 and from 289.9 � 1.3 to 293.3 � 1.1 mos-

mol L±1, respectively). Plasma sodium concentration

remained signi®cantly elevated throughout the obser-

vation period although values decreased after end of

infusion (Fig. 1). Plasma osmolality decreased during

the recovery period to values not signi®cantly different

from the preinfusion value (Table 1). However, plasma

osmolality also decreased during CONTR and ISOT

and the average plasma sodium concentration and

Figure 1 Plasma concentration of sodium. Values are mean � SE.

e: control; n: isotonic saline infusion; d: hypertonic saline infusion.

Values marked with * are signi®cantly different from the pre-infusion

value by ANOVA (P < 0.05) and Newman±Keuls' test.

Table 1 Plasma osmolality, plasma protein, oncotic pressure and haemotocrit

Sampling time, minutes after the start of urine sampling

Period, no 1 2 3 4 5 6 7 8

Time, min 30 90 150 210 270 330 390 450

Plasma osmolality,mosmol kg±1 H2O

CONTR 289.3 � 0.7 287.2 � 1.0* 286.0 � 1.0 285.8 � 0.6* 285.3 � 0.6* 284.8 � 1.1* 284.3 � 0.7* 283.3 � 0.7*

ISOT 287.5 � 0.6 287.4 � 0.8 285.1 � 0.9* 285.5 � 0.6* 285.6 � 0.9* 285.8 � 0.9* 285.5 � 1.1* 284.5 � 0.7*

HYPR 289.9 � 1.3 292.3 � 1.4 293.3 � 1.1* 291.8 � 0.8 290.4 � 1.3 290.4 � 1.1 290.3 � 1.0 290.6 � 1.2

Plasma protein concentration, g L±1

CONTR 66.3 � 1.0 65.3 � 1.0 65.7 � 0.9 65.8 � 0.8 64.8 � 0.8 64.8 � 0.8 64.2 � 0.7* 64.7 � 0.8

ISOT 66.3 � 1.1 62.5 � 1.1* 59.0 � 0.9* 61.1 � 1.1* 62.1 � 0.8* 60.3 � 1.1* 60.4 � 1.2* 61.0 � 0.9*

HYPR 66.4 � 0.9 63.5 � 0.9* 60.4 � 0.9* 61.8 � 0.9* 62.5 � 0.7* 60.8 � 0.9* 61.1 � 0.6* 61.0 � 0.6*

Oncotic pressure, mmHg

CONTR 24.7 � 0.3 24.3 � 0.3 24.3 � 0.2 24.5 � 0.3 23.8 � 0.3* 23.9 � 0.4* 23.5 � 0.2* 23.8 � 0.3*

ISOT 24.7 � 0.4 22.4 � 0.5* 20.5 � 0.4* 21.4 � 0.3* 21.9 � 0.4* 21.2 � 0.4* 21.0 � 0.3* 21.0v0.4*

HYPR 24.5 � 0.3 23.0 � 0.2* 21.4 � 0.2* 22.1 � 0.2* 22.3 � 0.3* 21.6 � 0.3* 21.6 � 0.2* 21.6 � 0.3*

Haematocrit, %

CONTR 43.2 � 0.8 42.4 � 0.9* 42.6 � 0.9* 42.8 � 1.0 42.3 � 0.8* 42.3 � 0.9* 41.8 � 0.8* 41.8 � 0.9*

ISOT 44.0 � 0.9 42.0 � 1.1* 40.7 � 1.0* 40.6 � 1.0* 41.3 � 1.0* 40.8 � 1.1* 40.8 � 1.1* 40.9 � 1.1*

HYPR 44.3 � 0.8 42.8 � 1.0* 41.3 � 0.9* 41.4 � 0.9* 41.8 � 0.9* 41.4 � 1.0* 41.2 � 1.0* 41.3 � 1.0*

Values are mean � SE. isotonic (ISOT) or hypertonic (HYPR) saline was infused intravenously between 60 and 150 min. Results marked with *

are signi®cantly different from period 1 (P < 0.05) by analysis of variance and Newman±Keuls' test.


Acta Physiol Scand 1999, 166, 23±30 L J Andersen et al. � Osmostimulated sodium excretion

osmolality from 120 to 480 min was signi®cantly higher

during HYPR than during the two other series.

Plasma oncotic pressure, plasma protein and ha-

ematocrit decreased immediately after start of saline in-

fusions (Table 1 and Fig. 2). These decreases were more

pronounced during ISOT than during HYPR. The mean

decrease in oncotic pressure during ISOT from period 3±

8 was to 85.9 � 0.8% of basal which was a signi®cantly

greater reduction than during HYPR (89.0 � 0.6%).

During CONTR, haematocrit also decreased signi®-

cantly but to a smaller extent than during ISOT and

HYPR. Oncotic pressure and plasma protein also de-

creased in the latter observation periods (Table 1).

Sodium excretion increased in all experimental series,

including CONTR during which sodium excretion in-

creased gradually throughout the experiment from a

baseline of 82 � 15 to 256 � 12 lmol min±1 (Fig. 3).

During the ®rst 4 h, hypertonic and isotonic saline in-

fusion induced very similar natriuretic responses

(84 � 10 to 306 � 28 lmol min±1 and 99 � 19 to

300 � 31 lmol min±1, respectively). By the end of the

observation, the natriuresis following isotonic saline

exceeded that following hypertonic saline (414 � 29 vs.

349 � 33 lmol min±1). However, the cumulative sodi-

um excretion during ISOT (131 � 13 mmol) was not

statistically different from that observed during HYPR

(123 � 10 mmol). The responses to both infusion series

exceeded that of the CONTR series (81 � 9 mmol).

Urine ¯ow increased gradually during CONTR from

2.7 � 0.4 to 6.2 � 0.6 mL min±1 observed in the last

experimental period. Urine ¯ow also increased during

ISOT from 2.4 � 0.5 to 10.6 � 0.4 mL min±1 but de-

creased from 2.4 � 0.5 to 1.8 � 0.3 mL min±1 during

HYPR (Fig. 4). Free water clearance followed the same

pattern as urine ¯ow with increases from )0.2 � 0.4 to

2.8 � 0.6 mL min±1 and from )0.5 � 0.3 to 6.0 � 0.4

mL min±1 during CONTR and ISOT, respectively, and

a decrease from 0.2 � 0.4 to )2.3 � 0.1 mL min±1

during HYPR.

A marginal increase in mean arterial pressure and

diastolic blood pressure was observed in the very last

period of observation during ISOT (Table 2). Diastolic

arterial blood pressure also increased slightly but sig-

ni®cantly during CONTR whereas no changes were

observed during HYPR. Systolic arterial blood pressure

remained unchanged in all series. Heart rate tended to

decrease but this was only signi®cant during CONTR

Figure 2 Relative change in plasma oncotic pressure. Values are

mean � SE. e: control; j: isotonic saline infusion; d: hypertonic

saline infusion. Values marked with * are signi®cantly different from

the preinfusion value by ANOVA (P < 0.05) and Newman±Keuls' test.

Figure 3 Renal sodium excretion. Values are mean � SE. e: control;

j: isotonic saline infusion; d: hypertonic saline infusion. Values

marked with * are signi®cantly different from the preinfusion value by

ANOVA (P < 0.05) and Newman±Keuls' test. Values marked with #

are signi®cantly different from corresponding value in the hypertonic

saline series by Student's t-test (P < 0.05).

Figure 4 Urine ¯ow. Values are mean � SE. e: control; j: isotonic

saline infusion; d: hypertonic saline infusion. Values marked with *

are signi®cantly different from the preinfusion value by ANOVA

(P < 0.05) and Newman±Keuls' test.



and HYPR. Creatinine clearance did not change mea-

surably in any of the series (Table 2).

Basal plasma concentrations of ANG II varied be-

tween series (Table 3). Samples were not measured in

the same assay, which may explain at least some of this

variation. However, the relative changes seem com-

patible (Fig. 5). From period 4 to 8, mean plasma ANG

II had decreased to 24 � 6, 27 � 5 and 61 � 6% of

preinfusion levels during ISOT, HYPR and CONTR,

respectively.

Basal plasma concentration of ANP varied between

30 � 3 and 36 � 3 pg mL±1 and remained unchanged

in all series (Table 3). Urinary excretion rates of en-

dothelin-1 varied between 18 � 3 and 23 � 3 pg min±1

and also remained statistically unchanged in all series

(Table 3).

Table 2 Haemodynamic variables and creatinine clearance

Sampling time, minutes after the start of urine sampling

Period, no 1 2 3 4 5 6 7 8

Time, min 30 90 150 210 270 330 390 450

Mean arterial blood pressure, mmHg

CONTR 80 � 2 81 � 2 82 � 2 82 � 2 81 � 2 80 � 2 82 � 2 83 � 2

ISOT 81 � 2 81 � 2 83 � 2 84 � 3 84 � 2 83 � 2 84 � 2 86 � 3*

HYPR 84 � 3 86 � 2 85 � 2 82 � 2 82 � 2 83 � 2 84 � 3 85 � 2

Systolic blood pressure, mmHg

CONTR 116 � 4 117 � 4 117 � 4 118 � 3 114 � 4 115 �4 117 � 4 118 � 4

ISOT 117 � 3 115 � 5 115 � 4 117 � 3 117 � 3 115 � 3 117 � 3 117 � 3

HYPR 120 � 4 120 � 4 120 � 3 116 � 3 117 � 3 118 � 3 120 � 3 120 � 2

Diastolic blood pressure, mmHg

CONTR 61 � 2 62 � 2 65 � 2 65 � 2* 65 � 1* 63 � 1 64 � 2 65 � 2*

ISOT 63 � 2 64 � 2 67 � 2 67 � 2 68 � 2 67 � 2 68 � 2 70 � 3

HYPR 65 � 3 68 � 2 68 � 2 64 � 2 65 � 2 65 � 2 66 � 3 67 � 2

Heart rate, beats min±1

CONTR 58 � 3 58 � 4 56 � 3 56 � 3 54 � 2* 53 � 3* 54 � 3* 52 � 2*

ISOT 57 � 3 56 � 3 55 � 3 57 � 4 53 � 2 53 � 2 52 � 3 52 � 2

HYPR 56 � 3 58 � 3 56 � 3 55 � 3 54 � 3 53 � 3 51 � 3* 51 � 3*

Creatinine clearance, mL min±1

CONTR 152 � 9 145 � 7 145 � 15 137 � 2 132 � 2 142 � 6 128 � 9 138 � 9

ISOT 158 � 8 144 � 7 154 � 13 152 � 9 150 � 9 152 � 10 147 � 8 140 � 11

HYPR 154 � 9 132 � 2 154 � 3 144 � 7 152 � 7 132 � 4 137 � 3 147 � 5

Values are mean � SE. isotonic (ISOT) or hypertonic (HYPR) saline was infused intravenously between 60 and 150 min. Results marked with *

are signi®cantly different from period 1 (P < 0.05) by analysis of variance and Newman±Keuls' test.

Table 3 Plasma hormone concentrations and urinary excretion rate of endothelin-1

Sampling time, min after the start of urine sampling

Period, no 1 2 4 6 8

Time, min 30 90 210 330 450

Plasma ANG II, pg ml)1

CONTR 14.2 � 3.0 11.2 � 2.6 9.3 � 2.7* 9.5 � 2.3* 9.9 � 2.7*

ISOT 6.4 � 1.3 3.8 � 1.1* 1.7 � 0.4* 1.2 � 0.3* 1.4 � 0.3*

HYPR 8.3 � 1.8 5.9 � 0.9* 3.0 � 1.0* 2.1 � 0.7* 2.5 � 0.8*

Plasma ANP, pg ml)1

CONTR 36 � 3 35 � 3 35 � 3 31 � 3 30 � 3

ISOT 30 � 6 28 � 2 35 � 3 32 � 3 31 � 3

HYPR 34 � 3 34 � 3 36 � 3 32 � 3 33 � 2

Urinary excretion rate of endothelin-1, pg min)1

CONTR 18 � 3 19 � 2 16 � 2 19 � 2 23 � 4

ISOT 18 � 3 14 � 2 13 � 2 15 � 3 51 � 25

HYPR 23 � 3 27 � 2 24 � 5 30 � 3 26 � 3

Values are means � SE. Isotonic (ISOT) or hypertonic (HYPR) saline was infused intravenously between 60 and 90 min. Results marked with *

are signi®cantly different from period 1 (P < 0.05) by analysis of variance and Newman Keuls test.



DISCUSSION

We compared the natriuretic responses to hypertonic

vs. isotonic saline infusion in normal supine subjects to

elucidate the effect of osmostimulation on the renal

handling of an acute sodium load. Experiments were

performed under well-de®ned and strictly controlled

conditions with regard to sodium status, individual

adjustment of dosage, and replacement of water and

salt losses throughout the study.

Stimulation of osmoreceptors

As expected, plasma sodium concentration and plasma

osmolality increased immediately after the start of in-

fusion of hypertonic saline and peaked by the end of

infusion (Fig. 1). The next 2 h plasma sodium con-

centration and osmolality decreased somewhat al-

though the excreted amounts of sodium were replaced

intravenously period by period throughout the experi-

ments. Probably, the load was not evenly distributed

throughout the ECV immediately after administration

but 2 h seems to be an unexpected long time for

reaching equilibrium. According to previous ®ndings in

normal subjects by Watenpaugh et al. (1992), a rapidly

administered intravenous salt load is evenly distributed

throughout ECV in less than 60 min. Slower trans-

portation of sodium to other compartments, e.g. the

intestines could be responsible for the fall in plasma

sodium after stop of infusion but this cannot be de-

termined from the present data. At any rate, plasma

sodium concentration remained higher than that ob-

served during CONTR and ISOT and only during

HYPR could osmoreceptors be stimulated as a conse-

quence of the loading procedure.

Stimulation of volume receptors

The decreases in oncotic pressure indicate that isotonic

saline induced a greater increase in plasma volume than

hypertonic saline. Accordingly, the stimulus to central

volume receptors should be greater during ISOT than

during HYPR. This notion is supported by our recent

®ndings in normal subjects, where similar isotonic saline

infusions indeed induced increases in central venous

pressure (CVP) and plasma volume that were signi®-

cantly higher than those induced by 5% saline (Ander-

sen et al. 1998). Changing body position to supine may

alter the CVP responses to salt loading. However, the

larger increase in CVP following isotonic saline was

independent of the initial sodium status of the subject

and therefore maintained at different levels of central

blood volume. Schrier et al. (1969) also found a greater

expansion of plasma volume and ECV when a given salt

load was infused as isotonic saline compared with 5%

saline. Thus, there are several lines of evidence in favour

of the notion that the salt loading induced a stronger

volumetric stimulus during ISOT than during HYPR.

Natriuretic responses to salt loading

Provided that volumetric stimuli dominate over os-

motic stimuli in the control of renal sodium excretion,

ISOT should be associated with a more pronounced

natriuresis than HYPR. However, the ®rst 4 h after

start of infusion, the natriuretic responses during

HYPR and ISOT were very similar. Therefore, we

suggest that part of the natriuresis during HYPR was

mediated by the increase in plasma sodium concentra-

tion and this increase was suf®cient to compensate for

the difference in the volumetric loads in the two series.

During the last 2 h of observation the natriuresis of

ISOT did exceed that of HYPR. But at this time plasma

sodium concentration had decreased somewhat to-

wards control values, which may explain the relative

decrease in sodium excretion by the end of HYPR. In

dogs, the excess natriuresis following hypertonic vs.

isotonic saline loading has been shown to decline after

end of infusion even in cases when plasma sodium

concentration remained elevated (Emmeluth et al.

1990). Unfortunately, most other animal studies on

osmostimulation and sodium handling have been ter-

minated shortly after end of infusion, making it im-

possible to evaluate the time frame of the natriuretic

responses to hypertonic vs. isotonic salt loading

(Emmeluth et al. 1992, 1994, 1996, Huang et al. 1995).

Very recently we demonstrated that isotonic and

hypertonic saline loading induced identical natriuretic

responses in seated subjects on high sodium intake. In

contrast the response to isotonic saline exceeded the

response to hypertonic saline in the same subjects when

they were on low sodium diet (Andersen et al. 1998).

Figure 5 Relative change in plasma angiotensin II concentration.

Values are mean � SE. e: control; j: isotonic saline infusion; d:

hypertonic saline infusion. Values marked with * are signi®cantly

different from the preinfusion value by ANOVA (P < 0.05) and

Newman±Keuls' test.



Thus, the responses to salt loading in these supine

subjects are qualitatively similar to those observed in

seated subjects on high sodium intake. The modest

central hypervolemia associated with the supine posi-

tion may have potentiated the response to hypertonic

saline infusion compared with the response in seated

subjects on low sodium intake. It is remarkable how a

change in posture from seated to supine has a profound

effect on renal excretory function during time control

conditions. Renal sodium excretion increased substan-

tially during CONT and the supine position does

therefore not provide a steady-state situation. In con-

trast, the seated position appears to provide very stable

conditions over several hours without regard to hae-

modynamic, humoral and renal excretory functions

(Andersen et al. 1998). These ®ndings emphasize the

importance of the inclusion of a time control series.

Alternatively, the supine position should be avoided in

acute studies of water and electrolyte homeostasis.

In several animal models, the natriuresis following

hypertonic infusion has been found to exceed that

following isotonic or hypotonic infusion by a factor of

2±5 (Childers & Schneider 1981, Emmeluth et al. 1990,

Huang et al. 1995) even when protocols are very similar

to the present one (e.g. Emmeluth et al. 1990). How-

ever, infusion of hypertonic saline has also failed to

induce excess natriuresis in one study on dogs (Schrier

et al. 1969). Papper et al. found that hypertonic and

isotonic saline infusion induced similar natriuretic re-

sponses in supine subjects even when the hypertonic

infusion increased plasma sodium concentration more

than 8 mmol L±1 (Papper et al. 195612 ). The absence of

excess natriuretic responses following hypertonic

loading in humans indicates that the operational range

of osmostimulated natriuresis in humans is smaller than

that observed in other species.

Natriuretic mechanisms

Osmostimulated natriuresis is probably mediated by a

humoral signal as renal denervation does not abolish

the response to central osmostimulation (Emmeluth

et al. 1996). However, the efferent signal to the kidney

is not yet identi®ed.

Plasma ANG II decreased by 82% and 74% during

ISOT and HYPR, respectively. In seated subjects re-

ceiving similar salt loads, plasma ANG II only de-

creased by » 40±50% (Andersen et al. 1998). The

somewhat greater suppression of ANG II observed in

the present study may be explained by the different

body positions of the subjects during the two sets of

investigations. Plasma ANG II decreased by some 33%

during CONTR whereas plasma levels remained un-

changed during time control in the seated subjects.

Corrected for these differences, the effect of a given

salt load on the suppression of ANG II seems sur-

prisingly constant. If ANG II is maintained at high

levels by exogenous administration during saline infu-

sion in humans (Singer et al. 1994) and dogs (Andersen

et al. 1998, abstract) the natriuresis is markedly inhib-

ited. Together, these ®ndings strongly indicate that a

decrease in plasma ANG II is a necessary condition for

the natriuresis of salt loading and ANG II may be the

main controller of sodium excretion in response to salt

loading as well as to changes in posture.

In contrast, plasma ANP remained unchanged in all

series. Apparently, the relative slow infusion rate did

not provide a volumetric stimulus that was powerful

enough to stimulate ANP secretion, not even during

ISOT. Similarly, plasma ANP was unaffected by

hypertonic as well as isotonic salt loading in seated

subjects (Andersen et al. 1998).

Endothelin-1 has been found to be natriuretic in

small doses in dogs (Sandgaard & Bie 1996) and the

urinary excretion rate of endothelin-1 is increased

with increasing sodium excretion following stimula-

tion of central osmoreceptors suggesting that en-

dothelin-1 may be involved in the natriuresis of

osmostimulation (Emmeluth et al. 1996). However,

urinary excretion of endothelin-1 did not change

measurably in the present study and it is unlikely that

the observed changes in sodium excretion were in-

duced by endothelin-1.

The changes in urine ¯ow and free water clearance

indicate that plasma vasopressin decreased during

CONTR and ISOT but increased during HYPR. This

notion is also supported by previous ®ndings in seated

subjects (Andersen et al. 1998). Although vasopressin is

involved primarily in water balance it may have some

antinatriuretic effect in humans (Andersen et al. 1990).

It is therefore possible that some of the natriuresis

during ISOT and CONTR could be contributed to a

fall in plasma AVP. Similarly, increased plasma vaso-

pressin may have prevented an excess natriuresis during

HYPR.

Haemodilution has been shown to be a potent

natriuretic stimulus, and this may contribute to the

natriuresis observed in all series (Johansen et al. 1997).

To what extent haemodilution has a direct renal effect

or acts through changes in for example ANG II is

unknown. At any rate, the haemodilution seemed more

pronounced during ISOT, which should augment an

excess natriuresis in response to isotonic saline. The

increases in blood pressure during CONT and by the

end of ISOT were less than 5 mmHg and we ®nd it

highly unlikely that natriuresis occurred as a conse-

quence of this minute increase in blood pressure. There

were no signi®cant changes in creatinine clearance, but

this parameter is not a very precise estimate of glo-

merular ®ltration rate in humans.



SUMMARY

The supine position did not provide a stable base line

situation with regard to renal salt and water excretion.

The natriureses seen in response to hypertonic and

isotonic saline infusion were very similar over the ®rst

5 h despite indications that isotonic saline provides a

stronger volumetric stimulus than hypertonic saline.

Consequently, it is suggested that part of the natriuresis

following hypertonic saline was mediated by the in-

crease in plasma concentration of sodium. The re-

sponses were very similar to those previously found in

seated subjects on high sodium intake. Unlike many

animal experiments, the natriuresis following hypertonic

infusion did not exceed that following isotonic saline

and the operational range of osmostimulated natriuresis

in humans appears smaller than seen in other species.

The natriuresis occurring in all series may be mediated

by decreases in plasma ANG II concentration but not

by plasma ANP or urinary levels of endothelin-1.

The authors thank Bodil Svenson, Birthe Lynderup Christensen,

Sigurd Kramer Hansen, Inge Pedersen, Barbara Sùrensen, and Trine

Eidsvold for expert technical assistance. We also wish to thank

Dr Steen Levin Nielsen for expert assistance in calculations of

subjects extracellular volume. The experiments were supported by

The Danish Medical Research Council, Danish Foundation for the

Advancement of Medical Science, Kong Christian den Tiendes

Foundation, Engineer August Frederik Wedell Erichsens Foundation,

Direktùr Jacob Madsens & Hustru Olga Madsens Foundation, and

The Danish Heart Foundation.

REFERENCES

Andersen, J.L., Andersen, L.J., Sandgaard, N.C.F. & Bie, P.

1998. Servo controlled elevation of blood pressure by

angiotensin II during volume expansion in conscious dogs.

FASEB J 12, A681 (Abstract).13

Andersen, L.J., Andersen, J.L., SchuÈtten, H.J., Warberg, J. &

Bie, P. 1990. Antidiuretic effect of subnormal levels of

arginine vasopressin in normal humans. Am J Physiol 259,

R53±R60.

Andersen, L.J., Norsk, P., Johansen, L.B., Christensen, P. &

Bie, P. 1998. Osmoregulatory control of sodium excretion

after sodium loading in humans. Am J Physiol 275,

R1833±R184214 .

Brùchner-Mortensen, J. 1982. The extracellular ¯uid volume

in normal man determined as the distribution volume of

[51Cr] EDTA. Scand J Clin Lab Invest 42, 261±264.

Childers, J.W. & Schneider, E.G. 1981. Effect of hypertonic

and hypotonic infusions on aldosterone in conscious

sodium-depleted dogs. Clin Sci 61, 191±199.

Emmeluth, C., Drummer, C., Gerzer, R. & Bie, P. 1992. Roles

of cephalic Na+ concentration and urodilatin in control of

renal Na+ excretion. Am J Physiol 262, F513±F516.

Emmeluth, C., Drummer, C., Gerzer, R. & Bie, P. 1994.

Natriuresis in conscious dogs caused by increased carotid

[Na+] during angiotensin II and aldosterone blockade. Acta

Physiol Scand 151, 403±411.

Emmeluth, C., Goetz, K.L., Drummer, C., Gerzer, R.,

Forssmann, W.G. & Bie, P. 1996. Natriuresis caused by

increased carotid Na+ concentration after renal

denervation. Am J Physiol 270, F510±F517.

Emmeluth, C., SchuÈtten, H.J., Knigge, U., Warberg, J. &

Bie, P. 1990. Increase in plasma sodium enhances

natriuresis in response to a sodium load unable to

change plasma atrial peptide concentration. Acta Physiol

Scand 140, 119±127.

Gilmore, J.P. & Nemeh, M.N. 1984. Salt depletion inhibits

cerebral-induced natriuresis in the dog. Am J Physiol 247,

F725±F728.

Huang, W., Lee, S.L. & SjoÈquist, M. 1995. Natriuretic role of

endogenous oxytocin in male rats infused with hypertonic

NaCl. Am J Physiol 268, R634±R640.

Johansen, L.B., Bie, P., Warberg, J., Christensen, N.J.,

Hammerum, M., Videbaek, R. & Norsk, P. 1997.

Hemodilution, central blood volume, and renal responses

after an isotonic saline infusion in humans. Am J Physiol 41,

R549±R556.

Kappelgaard, A.M., Nielsen, M.D. & Giese, J. 1976.

Measurement of angiotensin II in human plasma: technical

modi®cations and practical experience. Clin Chim Acta 67,

299±306.

McKinley, M.J., Lichardus, B., McDougall, J.G. & Weisinger,

R.S. 1992. Periventricular lesions block natriuresis to

hypertonic but not isotonic NaCl loads. Am J Physiol 262,

F98±F107.

Papper, S., Saxon, L., Rosenbaum, J.D. & Cohen, H.W. 1956.

The effect of isotonic and hypertonic salt solutions on the

renal excretion of sodium. J Lab Clin Med 47, 776±782.

Sandgaard, N.C.F. & Bie, P. 1996. Natriuretic effect of non-

pressor doses of endothelin-1 in conscious dogs. J Physiol

(Lond) 494, 809±818.

Schrier, R.W., Fein, R.L., McNeil, J.S. & Cirksena, W.J. 1969.

In¯uence of interstitial ¯uid volume expansion and plasma

sodium concentration on the natriuretic response to

volume expansion in dogs. Clin Sci 36, 371±385.

SchuÈtten, H.J., Johannessen, A.C., Torp, P.C., Sander, J.K.,

Bie, P. & Warberg, J. 1987. Central venous pressure ± a

physiological stimulus for secretion of atrial natriuretic

peptide in humans? Acta Physiol Scand 131, 265±272.

Singer, D.R., Markandu, N.D., Morton, J.J., Miller, M.A.,

Sagnella, G.A. & MacGregor, G.A. 1994. Angiotensin II

suppression is a major factor permitting excretion of an

acute sodium load in humans. Am J Physiol 266,

F89±F93.

Thornborough, J.R., Passo, S.S. & Rothballer, A.B. 1973.

Receptors in cerebral circulation affecting sodium excretion

in the cat. Am J Physiol 225, 138±141.

Watenpaugh, D.E., Yancy, C.W., Buckey, J.C., Lane, L.D.,

Hargens, A.R. & Blomqvist, C.G. 1992. Role of atrial

natriuretic peptide in systemic responses to acute isotonic

volume expansion. J Appl Physiol 73, 1218±1226.

Winer, B.J. 1971. Statistical Methods in Experimental Design.

McGraw-Hill, New York.

Zucker, I.H. & Kaley, G. 1976. Natriuresis induced by

intracarotid infusion of hypertonic NaCl. Am J Physiol 230,

427±433.



Isotonic and hypertonic sodium loading in supine humans

Documents

Transcript of Isotonic and hypertonic sodium loading in supine humans