In healthy young subjects there is direct evidence for sympathetic vasoconstrictor activation after drinking water, but this is not accompanied by an increase in arterial blood pressure. A
marked pressor response to water ingestion has, however, been observed in elderly subjects and in patients with autonomic failure. We examined the effect of water ingestion on haemodynamic variables
and heart rate variability (HRV) markers of cardiac vagal control in ten healthy young subjects and four cardiac transplant recipients with confirmed persistent cardiac vagal denervation. In a random
order crossover protocol, changes in heart rate, blood pressure and measures of high frequency (HF) HRV were compared over time following the ingestion of 500ml and 20ml (control) of tap water. In
healthy subjects, after drinking 500ml of water the heart rate fell from 67.6±2.0 (mean±S.E.M.) to 60.7±2.4 beats/min (P<0.01), and the bradycardic response peaked between
20 and 25min. There were no significant changes in arterial blood pressure. Over the same time course, water ingestion caused increases in measurements of HF HRV: root-mean-square of successive RR
interval differences (RMSSD) increased by 13±2.7ms after 500ml versus 2±3.1ms after 20ml (P<0.05); HF power increased by 686±400 versus -63±322 (P<0.01).
In transplant recipients water ingestion was followed by a pressor response (range 13 to 29mmHg). These results provide evidence that water ingestion in normal subjects is followed by an increase in
cardiac vagal control that may counteract the pressor effects of sympathetic activation. We suggest that in the elderly, in transplant recipients and in autonomic failure, loss of this buffering
mechanism explains the pressor response to drinking water.
The essential daily activity of drinking water has recently been shown to have unexpected haemodynamic effects which may be of clinical importance [
1–3]. Jordan and colleagues noted that in patients with autonomic failure, the severity of their symptoms of orthostatic hypotension were substantially decreased shortly after drinking
water. They then demonstrated in these patients that drinking 0.5 litre of water caused an acute rise in blood pressure of between 30 and 100mmHg [
1,
3]. A more modest pressor effect of approx. 10mmHg was also demonstrated in healthy middle-aged and elderly subjects, but was not seen in young adults [
2]. There is now substantial evidence that this pressor response to water drinking, which is maximal at about 30min is due to an increase in sympathetic output. In patients with autonomic failure
the pressor response was accompanied by a rise in plasma noradrenaline [
2]. Studies in healthy young volunteers using direct measurement of muscle sympathetic nerve activity from the peroneal nerve also showed a rise in sympathetic nerve traffic after water drinking
[
4]. Despite this, there was no change in arterial blood pressure and notably no change, or even a slight reduction in heart rate. Thus it appears that sympathetic activation is a normal response
to water drinking but paradoxically, an associated pressor response is present only in those with autonomic dysfunction and in the elderly. This suggests that in young patients, an effective
'buffering' mechanism exists to counteract the pressor effect of sympathetic vasoconstrictor activation. It has been suggested that this may be due to a concomitant increase in cardiac vagal
activation, which prevents a rise in arterial pressure by reducing cardiac output [
2,
4]. The present investigation was designed to determine whether this was the case, by examining the effect of water drinking in healthy young volunteers on haemodynamic variables and heart rate
variability markers of cardiac vagal control. As a further test of the importance of autonomic control of the heart we also determined the haemodynamic response to water ingestion in cardiac
transplant recipients in whom there is a specific loss of cardiac innervation, whilst sympathetic vasomotor pathways remain intact.
Ten subjects (6 males, 4 females) aged between 24 and 34 years were studied. All subjects were normotensive, non-smokers, on no medication and were screened by history and physical examination to
exclude cardiovascular or neurological dysfunction. In addition four cardiac transplant recipients were recruited (2 males, 2 females) aged between 50 and 62 years and between 2 and 9 years post-
transplant. In each patient persistent cardiac vagal dennervation had been confirmed by power spectral analysis of ECG recordings and lack of measurable baroreflex sensitivity by the phenylephrine
bolus method [
5].
The study was approved by the South Birmingham Regional Ethics Committee and all subjects provided written informed consent prior to taking part. Subjects attended an initial habituation and
training visit to our dedicated clinical autonomic research laboratory. At this visit they were trained to breathe to an audio signal set close to the individual's resting respiratory frequency.
Subjects were studied in a random-order crossover design and all studies were commenced at 08.00 hours after a 12-h fast. Subjects were asked to avoid alcohol for at least 24h before each study
and to empty their bladder before starting the study. Transplant recipients withheld their regular medication on the morning of each study, with the exception of immunosuppressive therapy. The
laboratory temperature was kept constant at 23±1 °C and subjects rested in the semi-supine position throughout. A standard three-lead ECG signal was amplified, processed [high frequency
(HF) signal noise filter >500Hz], and digitized at 500Hz with the use of a National Instruments NB/MI0/16XH/18 analog-to-digital converter board (National Instruments Corporation, Newbury, Berks.,
U.K.). A continuous arterial pressure signal was obtained with the Portapres device (TNO Biomedical Instrumentation, Amsterdam, The Netherlands) and was similarly digitized. Respiratory excursion was
recorded from the amplified output of a standard strain gauge attached to an elastic strap around the subject's chest. All signals were displayed on the screen of a personal computer running
Laboratory View 5.0 software (National Instruments Corporation). In addition, blood pressure was recorded throughout the protocol at 5min intervals, taking the mean for three consecutive measurements
using automated arm cuff sphygmomanometry.
Subjects were rested for 30min, after which 5-min segments of the three digital signals were recorded during breathing at the predetermined frequency and stored to disk. Subjects were then asked
to drink tap water at 18°C as quickly as was comfortable. All subjects were randomly assigned to drinking 500ml or 20ml (control) of water during the first of two studies. The second volume was
given during a separate study visit 7 to 14days later. ECG and blood pressure monitoring continued for a further 45min whilst 5min recordings for analysis of heart rate variability were stored at 5,
20 and 35min after drinking during fixed frequency respiration.
The ECG series for analysis were coded so that the investigator performing the analysis was blinded to the nature of the study. All ECG series were reviewed and if necessary edited to exclude
ectopic and artifact signals. The RR intervals before and after any ectopic beats were replaced by interpolation from the previous and following sinus intervals. No signal containing >1% of
ectopic beats was used for analysis. R waves were detected by an individually adjusted threshold. Heart rate variability (HRV) was analysed off-line on data lengths of 256 RR intervals.
We used the standard time domain measures of root-mean-square of successive RR interval differences (RMSSD), and percentage of successive RR interval differences exceeding 50ms (pNN50). These
indices, based on successive differences in RR intervals, assess HF ('beat-to-beat') variation associated with respiratory sinus arrhythmia mediated principally by the vagus nerve [
6].
Power spectral analysis was performed with the use of the Burg algorithm (autoregressive modelling), with a model order between 8 and 20 selected to minimize the Akaike information criterion. The
power at each underlying frequency was quantified by decomposing the total variability signal according to previously published methods [
7,
8]. Powers at low frequency (LF; centred at 0.1Hz) and at HF (corresponding to the observed respiratory frequency) were thus determined.
Data for arterial blood pressure and RR interval in the water and control arms of the study were compared by a two-tailed paired Student's t test. Differences between groups for time and
frequency domain indices of HRV were determined with the Wilcoxon signed rank test for paired data. Statistical significance was defined by a value of P<0.05. Numerical values are expressed
as mean±S.E.M.
The frequency of metronomic breathing was within the range 0.18 to 0.22Hz. Baseline values for mean arterial pressure, RR interval, and indices of HF HRV were not significantly different in the
group before drinking 500ml or 20ml of water (
Table 1). Drinking 500ml of water did not result in any change in arterial blood pressure measured by Portapres or brachial sphygmomanometry (
Table 2). After ingesting 500ml of water there was, however, a fall in the heart rate in each of the ten subjects (
Figure 1). This decrease was maximal at 20min after ingestion. The heart rate returned to a level not statistically different from baseline values at 40min. In contrast, in the control arm of the
study, with the same subjects drinking 20ml of water, there was no significant change in the heart rate (
Figure 1).
Values are means±S.E.M.
Values are given as mean±S.E.M for the group comparing results after drinking 500ml of water with results after drinking 20ml of water as a control. ns, not significant (P>0.05);
vs, versus.
In the cardiac transplant recipients the resting heart rate was 89beats/min and the heart rate did not change significantly from baseline throughout the study period. Baseline systolic arterial
blood pressure was 137mmHg (range 124–144) and between 15 and 30min following ingestion of 500ml of water had increased in all four subjects, returning to baseline at 40–45min. The
changes in systolic pressure by automated brachial sphygmomanometry at 20min after drinking 500ml of water were +19mmHg, +13mmHg, +29mmHg and +14mmHg. Measures of HF HRV were extremely low or absent
in all transplant patients and remained so throughout the study.
DISCUSSION
The main finding of the present study was that water drinking in normal subjects caused a bradycardia, which our HRV analysis indicated was due to an increase in cardiac vagal activity. The vagus
nerve is inaccessible to direct recording in humans, but vagal modulation of heart rate can be assessed indirectly using HF ('beat-to-beat') variability. This is the only neural mechanism capable of
mediating variability at this frequency and HF indices such as RMSSD, pNN50 and the power of the HF peak (derived from power spectral analysis) are widely held to reflect the degree of vagal
modulation of heart rate [
6].
In addition, our results have confirmed that in normal young subjects, compared with the elderly or those patients with autonomic failure, ingestion of a moderate volume of water does not lead to
a rise in arterial blood pressure. This is in keeping with the results of previous investigators who, despite convincingly demonstrating a rise in sympathetic vasoconstrictor activity, found no rise
in arterial blood pressure using a Finapres device [
4,
9].
Our results therefore suggest that the physiological response to water drinking in healthy subjects may be an integrated response, consisting of an increase in sympathetic vasoconstrictor activity
coupled with a parallel increase in cardiac vagal activity. Thus the increase in total peripheral resistance is counteracted by a fall in cardiac output. Such an autonomic response is unusual, since
under most circumstances sympathetic and parasympathetic outflows are inversely related. The response is similar in some respects to the 'diving reflex' in which trigeminal nerve stimulation by cold
results in cardiac vagal activation and a profound bradycardia, as well as sympathetic activation and vasoconstriction [
10,
11]. The response to water drinking, however, differs in that it has a gradual onset, the consequent bradycardia is prolonged throughout 35–40min and it occurs in response to water at room
temperature. An integrated response of this nature would explain why, in these young volunteers and in those studied previously [
2,
4,
9], despite increases in sympathetic vasoconstrictor discharge and peripheral resistance, blood pressure remains stable, as a result of a vagally mediated fall in cardiac output. The fall in
cardiac output is likely to be mediated by reductions in both heart rate and stroke volume, since vagal stimulation has been demonstrated recently, in both humans and animal subjects, to have
negative inotropic effects [
12].
Whilst the present investigation was not designed to determine the nature of the afferent stimuli to such an integrated autonomic response, two mechanisms have been proposed to explain the
activation of sympathetic pathways in normal subjects. Gastric distension with water or a balloon in both animal [
13–16] and human [
17] studies has been repeatedly demonstrated to cause an increase in sympathetic activity. Decreases in portal osmolality have also been implicated in a reflex that results in sympathetic
vasoconstriction, a diuresis and sweating [
18,
19]. Similar afferent stimuli might also lead to an increase in cardiac vagal efferent activity; however, it must be noted that the characteristics of the observed response are not those that
would be expected from a conventional reflex. The bradycardic response we have observed following water ingestion has a slow increase in magnitude and a slow decay, a pattern more characteristic of a
response driven by the generation of humoral factors. Factors such as angiotensin II and atrial natriuretic peptide can affect brain centres with a poor blood–brain barrier such as the area
postrema and subcortical areas of the forebrain, either of which may be involved in the generation of the autonomic response to water ingestion.
Ageing is associated with marked reductions in cardiac vagal control even in healthy adults [
20,
21]. Similarly, baroreflex sensitivity is markedly reduced in elderly subjects [
22,
23]. We propose that a deficiency in the cardiac vagal component of this integrated response to water ingestion may explain the moderate rise in blood pressure that occurs in healthy elderly
subjects after drinking water.
In patients with chronic autonomic failure, residual sympathetic activity may be responsible for the pressor response, whilst loss of cardiac vagal control of heart rate means that this rise in
arterial blood pressure is not compensated for by a decrease in cardiac output. Some doubt remains regarding the mechanism of sympathetic activation, particularly in this group of patients who are
often unable to mount an adequate pressor response to overcome disabling postural drops in blood pressure. This has been highlighted recently by Mathias and colleagues, who postulated that in
patients with autonomic failure, the observed response may be due to restoration of intravascular volume [
3]. The increase in cardiac vagal activity after water drinking that we have shown would also be consistent with an increase in circulating volume and the consequent low pressure baroreceptor
response. This explanation is not, however, consistent with the previous reports of an increase in sympathetic nerve activity in response to water drinking in normal subjects. In addition, previous
studies of plasma volume after drinking water have found no significant change [
2,
4]. We therefore favour a neurally mediated pressor effect.
The aim of our studies with the transplant patients was to document the haemodynamic response to water drinking in a group whose autonomic pathology could be more clearly defined than in those
with the more diverse syndromes of chronic autonomic failure. This group of patients had been demonstrated to have complete persistent cardiac vagal denervation between 1–3 years after surgery.
The finding in these individuals of a pressor response to water ingestion (of a magnitude between that of elderly subjects and patients with autonomic failure), again suggests that cardiac vagal
innervation is required to prevent this pressor response. The rise in arterial pressure was variable but modest considering the intact peripheral sympathetic innervation and lack of cardiac vagal
innervation. This may be explained by the baseline hypertension of the transplant patients and their treatment with cyclosporin, which is well recognized to cause high basal levels of sympathetic
activity [
24].
The physiological response to water drinking in normal subjects appears to be complex, involving both limbs of the autonomic nervous system. Further investigations are required to define this
response in detail. Only when the normal response to water drinking is clearly understood will we be able to explain the powerful and potentially useful effects of water ingestion on blood pressure
in subjects with autonomic dysfunction.
REFERENCES
1 Jordan, J., Shannon, J.R., Grogan, E., Biaggioni, I. and Robertson, D. (1999) A potent pressor response elicited by drinking water. Lancet 353, 723
Medline 1st Citation 2nd
2 Jordan, J., Shannon, J.R., Black, B.K. et al. (2000) The pressor response to water drinking in humans: a sympathetic reflex? Circulation 101, 504–509
Medline 1st Citation 2nd 3rd 4th 5th 6th
3 Cariga, P. and Mathias, C.J. (2001) Haemodynamics of the pressor effect of oral water in human sympathetic denervation due to autonomic failure. Clin. Sci. 101,
313–319
Medline 1st Citation 2nd 3rd
4 Scott, E.M., Greenwood, J.P., Gilbey, S.G., Stoker, J.B. and Mary, D.A. (2001) Water ingestion increases sympathetic vasoconstrictor discharge in normal human subjects. Clin.
Sci. 100, 335–342
Medline 1st Citation 2nd 3rd 4th 5th
5 La Rovere, M.T., Mortara, A. and Schwartz, P.J. (1995) Baroreflex sensitivity. J. Cardiovasc. Electrophysiol. 6, 761–774
Medline 1st Citation
6 Task Force. (1996) Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the
North American Society of Pacing and Electrophysiology. Circulation 93, 1043–1065
Medline 1st Citation 2nd
7 Chowdhary, S., Vaile, J., Fletcher, J., Ross, H., Coote, J. and Townend, J. (2000) Nitric oxide and cardiac autonomic control in humans. Hypertension 36, 264–269
Medline 1st Citation
8 Pagani, M., Lombardi, F., Guzzetti, S. et al. (1986) Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man
and conscious dog. Circ. Res. 59, 178–193
Medline 1st Citation
9 Joannides, R., Moore, N., de la Gueronniere, V. and Thuillez, C. (1999) Effect of water on arteries. Lancet 354, 516
1st Citation 2nd
10 Anderson, H.T. (1966) Physiological adaptations in diving vertebrates. Physiol. Rev. 46, 212–243
Medline 1st Citation
11 Strauss, M.B. (1970) Physiological aspects of mammalian breath hold diving: a review. Aero. Med. 41, 1362–1381
1st Citation
12 Lewis, M.E., Al.-Khadidi, A.H., Bonser, R.S., Clutton-Brock, T., Morton, D., Paterson, D., Townend, J.N. and Coote, J.H. (2001) Vagus nerve stimulation decreases left
ventricular contractility in vivo in the human and pig heart. J. Physiol. 534, 547–552
Medline 1st Citation
13 Longhurst, J.C., Spilker, H.L. and Ordway, G.A. (1981) Cardiovascular reflexes elicited by passive gastric distension in anesthetized cats. Am. J. Physiol. 240,
H539–H545
Medline 1st Citation
14 Longhurst, J.C. and Ibarra, J. (1982) Sympathoadrenal mechanisms in hemodynamic responses to gastric distension in cats. Am. J. Physiol. 243, H748–H753
Medline 1st Citation
15 Vacca, G., Mary, D.A., Battaglia, A., Grossini, E. and Molinari, C. (1996) The effect of distension of the stomach on peripheral blood flow in anaesthetized pigs. Exp.
Physiol. 81, 385–396
Medline 1st Citation
16 Vacca, G. and Vono, P. (1993) The primary reflex effects of distension of the stomach on heart rate, arterial pressure and left ventricular contractility in the
anaesthetized pig. Pflugers Arch. 425, 248–255
Medline 1st Citation
17 Rossi, P., Andriesse, G.I., Oey, P.L., Wieneke, G.H., Roelofs, J.M. and Akkermans, L.M. (1998) Stomach distension increases efferent muscle sympathetic nerve activity and
blood pressure in healthy humans. J. Neurol. Sci. 161, 148–155
Medline 1st Citation
18 Haberich, F.J. (1968) Osmoreception in the portal circulation. Fed. Proc. 27, 1137–1141
Medline 1st Citation
19 Nakai, M. (1993) Cardiovascular responses to gastric hypo-osmolar stimulation in anesthetized dogs. Jpn. J. Physiol. 43, 335–346
Medline 1st Citation
20 Low, P.A. (1993) The effect of ageing on the autonomic nervous system. In Clinical Autonomic Disorders (Low, P.A., ed.), pp. 685–700, Little, Brown and Company, Boston
1st Citation
21 Agelink, M.W., Malessa, R., Baumann, B., Majewski, T., Akila, F., Zeit, T. and Ziegler, D. (2001) Standardized tests of heart rate variability: normal ranges obtained from
309 healthy humans, and effects of age, gender, and heart rate. Clin. Autonom. Res. 11, 99–108
1st Citation
22 Gribbin, B., Pickering, T.G., Sleight, P. and Peto, R. (1971) Effect of age and high blood pressure on baroreflex sensitivity in man. Circ. Res. 29, 424–431
Medline 1st Citation
23 Monahan, K.D., Tanaka, H., Dinenno, F.A. and Seals, D.R. (2001) Central arterial compliance is associated with age- and habitual exercise-related differences in cardiovagal
baroreflex sensitivity. Circulation 104, 1627–1632
Medline 1st Citation
24 Sander, M., Lyson, T., Thomas, G.D. and Victor, R.G. (1996) Sympathetic neural mechanisms of cyclosporine-induced hypertension. Am. J. Hyperten. 9, S121–S138
1st Citation
Received 30 October 2001/30 January 2002; accepted 22 April 2002
The Biochemical Society and the Medical Research Society ©
2002
Abstract
Baseline arterial blood pressure, RR interval and RR interval variability before the ingestion of 500ml and 20ml of water (paired data on
separate study days)
Changes from baseline values (D) in haemodynamic variables, time domain indices of HF HRV (RMSSD, pNN50) and
frequency domain indices (HF and LF absolute power) at various intervals of time following water ingestion
My CS toolbox
) or 20ml (
; control) of tap water