643
Clinical Science (1984) 66, 643-648
Acute effects of moderate alcohol consumption on blood
pressure and plasma catecholamines
M. A. IRELAND, R . VANDONGEN, L. DAVIDSON, L. J . BEILIN
A N D 1. L. ROUSE
University Department of Medicine, Royal Perth Hospital, Perth, Western Austrcllia
(Received 28 January/27 June 1983; accepted 11 November 1983)
summary
1. This study examines the response of blood
pressure, plasma catecholamines and cortisol to
acute alcohol intake in young men with light to
moderate drinking habits.
2. Ingestion of alcohol was associated with a
highly significant increase in systolic blood
pressure and heart rate which occurred before
blood alcohol reached its peak concentration of
16.9 k 1.1 mmol/l (80 mg/100 ml). After an initial
non-specific rise, diastolic pressure fell below
values observed after drinking water only. This
predominant effect of alcohol on systolic blood
pressure is also seen with chronic alcohol consumption.
3. Drinking water and non-alcoholic cold
liquids caused a marked fall in plasma adrenaline
and a transient rise in noradrenaline concentration.
In contrast, drinking alcohol resulted in a relative
rise in adrenaline and a delayed increase in noradrenaline concentration.
4. Blood glucose increased after alcohol,
supporting a physiological effect of adrenaline on
liver glycogenolysis. Plasma cortisol concentration
was also significantly higher after drinking alcohol.
5. It is proposed that the relative rise in
adrenaline together with higher cortisol levels,
repeated over a variable period in susceptible
individuals, are implicated in the elevation of
blood pressure associated with long term alcohol
consumption. It concurs with observations in man
and experimental animals o f a slow pressor
mechanism mediated by adrenaline.
Correspondence: Associate Professor R. Vandongen, University Department of Medicine,
Royal Perth Hospital, Perth, Western Australia,
6000.
6 . The study emphasized that an evaluation of
the acute effects of alcohol requires careful
measurement of blood pressure, precise assay of
catecholamines and recognition of the confounding effects of drinking cold liquids.
Key words: alcohol, adrenaline, blood pressure,
cortisol, noradrenaline, pressor mechanisms.
Introduction
The close relationship between alcohol consumption, blood pressure and the prevalence of hypertension has received renewed interest [l-31. The
mechanisms responsible for this pressor effect of
alcohol are unknown. We have previously studied
the effect of moderate habitual alcohol consumption on blood pressure and sympatho-adrenal
activity in drinking and non-drinking males
matched for age and weight [4]. Although drinkers
had significantly higher systolic blood pressure,
plasma catecholamines and cortisol were similar
at rest and after various physiological stresses. It
could be concluded that the higher blood pressure
in drinkers is not caused by increased sympathoadrenal activity. However, initiating mechanisms
may be transient with each episode of drinking, or
may become obscured by the development of
secondary adaptive changes once hlgher blood
pressure is established. We considered it important
to examine under carefully controlled conditions
the acute effects of moderate alcohol consumption
on blood pressure and sympatho-adrenal hormones
using precise and sensitive assays.
Methods
We studied the acute effects of alcohol consumption on heart rate, blood pressure and sympatho-
644
M.A. Ireland et al.
ment (type of drink) and time, and the Newmanadrenal hormones in 14 male university students
Keuls test for within-group (time) comparisons
(mean 20.3 It SEM 1.2 years and mean weight 72 f
and to evaluate the significance of between-group
SEM 2.3 kg) with light to moderate drinking habits
(treatment) differences. Values are means k SEM.
assessed by a 7 day retrospective diary [3]. The
students were paid a nominal sum and the
protocol was approved by the Hospital's Ethics
Results
Committee. They were asked to avoid alcohol,
The 14 subjects taking part in this study were
coffee, tea and tobacco on the day of the study
and reported to the laboratory after a light lunch.
estimated from the questionnaire to drink 170 f
Subjects were studied recumbent as far as possible,
55 ml of alcohol/week.
in a quiet room where an indwelling needle was
As seen in Fig. 1, there was a significant
inserted into a forearm vein. After the subjects
increase in systolic blood pressure with time after
had been lying down for 20min, heart rate and
consumption of the drink with added alcohol
blood pressure were measured with an automatic
(F8,1w
= 2.46, P = 0.02) (repeated measures AOV;
oscillometric recorder (Dinamap, Applied Medical
degrees of freedom 8,104 refer to analysis of treatment-time interaction or effect of time only, and
Research, U.S.A.). Blood samples were drawn for
1,13 refer to analysis of treatment effect only).
determination of plasma adrenaline, noradrenaline,
This rise was apparent after the first drink and
cortisol, alcohol and glucose. Each subject then
before there was a major increase in blood alcohol
drank two 370ml volumes of cold water (4OC),
concentration (P<O.OI at 10 and 20min
each over consecutive 10 min periods. Most were
unable to drink lying down and adopted a semisitting position. Blood pressure and heart rate
were recorded and blood samples were drawn at
the end of each 10 min period, at 10 min intervals
for 30 min and at 20 min intervals for a further
60 min. Subjects returned the following day for an
identical study except that they now drank two
370 ml volumes of cold non-alcoholic beer (Birell,
Cooper and Sons, South Australia), to each
volume of which was added 0.5 ml of ethanol/kg
of body weight. When necessary they were permitted to void urine after 60min but this was
timed to be completed at least 15 min before the
+
.*x
next measurements and blood samples.
Seven of these subjects then participated in a
further study on a separate occasion, drinking the
non-alcoholic beer without alcohol added. The
subjects were not told beforehand whether or not
alcohol had been added to their drinks.
Blood samples for adrenaline and noradrenaline
assay were collected in chilled tubes containing
EGTA (0.25 mol/l) and p t a t h i o n e (0.2 mol/l).
Flasma was separated at 4 C, stored at -8OoC and
assayed by a modification [5] of the radio0
0
10 20 30 40 50 70 90 110
enzymatic method described by Peuler & Johnson 8
Time (min)
[ 6 ] .All samples obtained from one subject during
each study were assayed together. The interassay
FIG. 1. Effect of drinking water (M)
or noncoefficient of variation was 12.7% for noralcoholic beer with added alcohol (u)
on
blood pressure, heart rate and blood ethanol
adrenaline and 9.4% for adrenaline with a sensiconcentration in 14 men. Basal measurements
tivity limit of 0.10 nmol/l. Plasma cortisol was
were taken at Omin. Measurements at 10 and
measured by radioimmunoassay, alcohol by
20 min were made after completing the first and
enzymatic assay (Calbiochem Alcohol Kit) and
second drinks respectively. Volume of each drink
glucose by autoanalyser.
was 370 ml; alcoholic drinks contained 0.5 ml of
Statistical analyses were performed using the
ethanol/kg. Significance of difference between
Statistical Package for the Social Sciences. All data
treatment groups (Newman-Keuls test): tP <
were analysed by repeated measures analysis of
0.01, and from values at 0 min *P<0.05 and
variance (AOV), examining the effects of treat**P< 0.01 (Newman-Keuls test).
^ ^
4
-
Alcohol and blood pressure
645
compared with basal value at Omin, NewmanKeuls test). Systolic pressure rapidly declined after
the second drink. There was no change in systolic
pressure when water was drunk. Diastolic pressure
initially increased with both drinks (Fig. 1)
(FbID1
= 4.53, P < 0.0001) but subsequently fell
in the alcohol group (Fl,13
= 10.43, P < 0.007),
the fall coinciding with blood alcohol reaching its
maximum concentration. There was a highly sig
nificant and immediate effect of alcohol on heart
rate (F1,13 = 22.91, P < 0.0004), which was clearly
evident in the higher values after alcohol throughout the study (P< 0.01, compared with corresponding values after water, Newman-Keuls test).
Both time and treatment markedly influenced
plasma adrenaline concentration (Fig. 2) (FBIo4
=
5.79, P<O.OOOOl). Twenty and 30min after
alcohol mean plasma adrenaline concentration was
considerably higher than after water (P< 0.01),
where levels sharply declined. Examination of
individual values showed an increase of plasma
adrenaline over basal levels at 20 min in seven of
the 14 subjects drinking alcohol but only in one
subject after drinking water. At 40 min, however,
adrenaline concentrations were again similar.
There was an obvious initial rise in noradrenaline
concentration after both drinks (Fig. 2), which
was short-lived after water but sustained after
alcohol, indicating a significant effect of treatment with time (F8,104
= 5.69, P < 0.00001) and
difference between the groups at 90min (P<
0.01).
Although there was an overall fall in plasma
cortisol concentration with time this was less
= 6.13, P < O.OOOOl),
marked after alcohol (FhIo4
where levels were significantly higher at 30,40 and
50 min (P< 0.01) than after water.
Blood glucose concentration was significantly
higher at 20, 30 and 40min when alcohol was
taken (Table 1).
To ensure that the changes observed were due
to alcohol and not to the non-alcoholic beer to
which it was added, the effects of drinking the
non-alcoholic beer alone were examined in seven
of the above subjects. As shown in Fig. 3, there
was an early increase in diastolic pressure with
0
10 20 30 40 50 70 90 110 time (Fqa= 2.34, P < 0.05), similar to when they
drank water. Contrary to the fmdings with
Time (min)
alcohol, subsequent diastolic pressures did not
FIG. 2. Effect of drinking water (Mor
)nondiffer. There were no significant changes in
alcoholic beer with added alcohol (M)
on
systolic pressure or heart rate. As before, a highly
plasma concentration of adrenaline, noradrenaline,
significant rise in plasma noradrenaline concentracortisol and ethanol in 14 men. For further
tion (F848= 14.17, P < O.OOOOl), was observed
explanation and significance of differences see
with time (Fig. 4) after completion of the first
Fig. 1.
-
TABLE 1. Effect o f drinking alcohol (0.5 ml of ethanolfkg at 10 and 20 min), com-
pared with water, on blood glucose
Values are means k SEM ( n = 14). **Significantly different from water at P < 0.001.
~~
Blood glucose (mmol/l)
Time (min) ....
0
10
Water
5.9t0.3
Alcohol
5.9tO.2
5.7t0.3
6.0i0.2
20
30
40
110
5.2t0.2
4.8i0.1
5.420.3
5.220.3
7.1 t0.3** 7.8i0.4** 1.7*0.5** 4.8i0.3
M. A. Ireland et al.
T
0.0004), confirming the specificity of the response
to alcohol. There was a progressive fall with time
in plasma cortisol concentration (F8,t)8= 7.79,
P < 0.00001) with no significant difference
between the groups. There was no difference in
blood glucose between drinking water and nonalcoholic beer.
Discussion
Little reliable information is available on the acute
effects of moderate alcohol consumption on blood
pressure and plasma catecholamines. Previously
either no change [7] or a small rise in heart rate
and diastolic blood pressure [8] were reported
after moderate amounts of alcohol. Indeed, it
appears to have become accepted that acute
l o 2o 30 40 50 'O
'lo ingestion of alcohol does not affect blood pressure
Time (min)
191. Although we have demonstrated a consistent
FIG. 3. Comparison of the effect of drinking water
haemodynamic effect, the changes were relatively
(-1
and non-alcoholic beer, without d d e d
small, indicating the need for careful measureon blood Pressure and heart
ments under controlled conditions. Use of an
alcohol (-),
rate in seven men.
automatic blood uressure recorder Death facilitated detection 0; these changes b; eGinating
observer variability and error. Interpretation of
blood pressure changes, particularly diastolic
pressure, is also complicated by a non-specific rise
in pressure due to drinking. Whether this is due to
the low temperature or volume effect of the drink,
or to the subjects partly sitting up to drink, is
lit
not known. The relevance of the acute rise in
xx
systolic blood pressure to the long term effect of
alcohol is strongly supported by epidemiological
**
evidence of a predominant effect on systolic
pressure [2-41.
Increased urinary catecholamines have been
reported during heavy alcohol consumption [ 10,
111, a situation quite different from our study
I
P where mean blood alcohol concentration only
reached 16.9 1.1 mmol/l (80 mg/100 ml). The
availability of sensitive and precise assays of
plasma adrenaline and noradrenaline has allowed
a careful reappraisal of the response to alcohol.
Firstly, it has uncovered small but nevertheless
important and highly significant responses to
0
l o 2o 30 40 5o 70
110 drinking cold liquids. Failure to recognize and
Time (min)
control for these changes could result in the
erroneous belief that they were due to alcohol.
FIG. 4. Changes in plasma adrenaline, noradrenaline
and cortisol concentration, after drinking water
Furthermore the effects of alcohol on plasma
(0)and non-alcoholic beer (a), in seven men. For
catecholamines need to be interpreted in the
significance of differences see Fig. 1.
light of the effects of drinking cold liquids alone.
The early response in adrenaline and the sustained
rise in noradrenaline after alcohol therefore
(water) or second (non-alcoholic beer) drink. Comrepresent a relative increase in these plasma
pared with the effect of alcohol, this rise was not
levels. Also, in seven of the 14 subjects studied, a
sustained. The marked decrease in adrenaline
concentration also occurred after drinking the
definite increase in plasma adrenaline concentration followed alcohol drinking. That this
non-alcoholic beer (Fig. 4)
= 4.49, P <
1
*
Alcohol and blood pressure
change in plasma adrenaline concentration is of
physiological significance is suggested by the rise
in blood glucose, which is presumably a secondary
effect of adrenaline on the liver. Adrenaline
measured in forearm venous blood also underestimates arterial levels because of tissue re-uptake
and conjugation [12], and therefore alcohol may
increase adrenaline concentration at active
receptor sites more than is indicated by the
measurements reported here.
The rise in systolic pressure and heart rate and
the lack of fall in plasma adrenaline concentration
occurred before there was any substantial increase
in blood alcohol concentration. This resembles
the ‘cephalic’ phase of metabolic regulation
raising plasma catecholamines in anticipation of a
metabolic need [13]. It is conceivable that these
responses are mediated by smell, taste or other
visceral afferents modifying sympatho-adrenal
outflow. Rising blood alcohol concentration may
also directly stimulate adrenal medullary secretion
or influence adrenaline disposal mechanisms.
The findings suggest a causal relationship
between the level of circulating adrenaline, the
early rise in systolic blood pressure and heart
rate, and the later fall in diastolic pressure. This is
most readily explained by a rise in cardiac output
and a decrease in peripheral resistance (due to
stimulation of vascular p2 receptors), effects which
are reproduced by slow intravenous infusion of
adrenaline [14]. Alcohol, as a peripheral vasodilator, may also contribute directly to the fall in
diastolic pressure.
The persistent elevation of plasma noradrenaline
could be a reflex sympathetic response to falling
diastolic pressure, or to enhanced neuronal release
due to adrenaline stimulating presynaptic &
adrenoceptors [ 151.
Alcohol given acutely to non-alcoholics usually
[16], but not invariably [17], increases plasma
cortisol. Although we found only small changes in
cortisol levels specifically associated with alcohol
ingestion, these may add to the pressor action of
alcohol by amplifying the cardiac and vascular
effects of catecholamines [18].
Although an actual increase in plasma
adrenaline concentration was observed in only
50% of subjects drinking alcohol, the overall
response has to be viewed against the effect of
cold fluids tending to decrease levels. Individual
susceptibility may well determine the magnitude
of this response and any ensuing blood pressure
changes.
Adrenaline has been proposed as the cause of
essential hypertension [ 191 and a sustained rise in
blood pressure has been observed in rats continuously infused with small amounts of adrena-
647
line [15]. Such a slow pressor mechanisn may
require only small increments in plasma adrenaline
or cortisol to produce a gradual increase in blood
pressure and it is tempting to speculate that this
mechanism is implicated in the pressor effect of
alcohol. It is likely that the effects of alcohol are
dosedependent and that larger amounts result in
quantitatively greater responses.
Acknowledgments
This investigation was supported by the Raine
Foundation of the University of Western Australia
and the National Health and Medical Research
Council.
References
1. Klatsky, A.L., Friedman, G.D., Siegelaub, A.B. &
Gerard, M. J. (1977) Alcohol consumption and blood
pressure. New England Journal of Medicine, 296,
1194-1200.
2. Cooke, LM., Frost, G.W.,Thornell, I.R. & Stokes,
G.S. (1982) Alcohol consumption and blood pressure.
Medical Journal of Australia, 1,65-69.
3. Arkwright, P.D., Beilin, L.J., Rouse, LA., Armstrong,
B.K. & Vandongen, R. (1982) Effects of alcohol use
and other aspects of life style on blood pressure levels
and prevalence of hypertension in a working population. Circulation, 66,60-66.
4. Arkwright, P.D., Beilin, L.J., Vandongen, R., Rouse,
LA. & Lalor, C. (1982) The pressor effect of
moderate alcohol consumption in man: a search for
mechanisms. circulation, 66,515-519.
5. Joyce, D.A., Beilin, L.J., Vandongen, R. & Davidson,
L. (1982) Plasma free and conjugated catecholamine
levels during acute physiological stimulation in man.
Life Sciences, 30,447-454.
6.Peuler, J.D. & Johnson, G.A. (1977) Simultaneous
single isotope radioenzymatic assay of plasma
norepinephrine, epinephrine and dopamine. Life
Sciences, 21,625-636.
7. Riff, D.P., Jain, A.C. & Doyle, J.T. (1969) Acute
hernodynamic effects of ethanol on normal human
volunteers. American Heart Journal, 78,592-597.
8. Zsoter, T.T. & Sellers, E.M. (1977) Effect of alcohol
on cardiovascular reflexes. Joumal of Studies on
Alcohol, 38, 1-10.
9. Saunder, J.B., Beevers, D.G. & Paton, A. (1981)
Alcohol induced hypertension. Lancet, ii, 653-656.
10. Anton, A.J. (1965) Ethanol and urinary catecholamines in man. Clinical Pharmacology and Therapeutic~,6,462-469.
ll.Ogato, M., Mendelson, J.H., Mello, N.K. &
Majchrowicz, E. (1971) Adrenal function and alcoholism. 11. Catecholamines. Psychosomatic Medicine,
33,159-180.
12. Joyce, D.A., Beilin, L.J., Vandongen, R. & Davidson,
L. (1982) Epinephrine sulfaction in the forearm:
arteriovenous differences in free and conjugated
catecholamines. Life Sciences, 31, 2513-2517.
13. Landsberg, L. & Young, J.B. (1978) Fasting, feeding
and regulation of the sympathetic nervous system.
New England Journal ofMedicine, 298,1295-1301.
648
M.A. Ireland et al
14. Weiner, N. (1980) Norepinephrine, epinephrine and
the sympathomimetic mines. In: The Pharmacdogacal Basis of Therapeutics, p. 138. Ed. Gilman, A.G.,
Goodman, L.S. & Gitman, A. Macmillan, New York.
15.Majewski, H., Tung, L.H. & Rand, M.J. (1981)
Adrenaline induced hypertension in rats. Journal of
Gzrdiovawlar Phmacology, 3,179-185.
16. Ylikahri, R.H., Huttunen, M.O., Harkonen, M., Leino,
T., Helenious, T., Liewendahl, K. & Karonen, S.L.
(1978) Acute effects of alcohol on anterior pituitary
secretion of the trophic hormones. Journal of Clinical
Endocrinology and Metabolism, 46,715-720.
17. F’rinz, P.N., Roehrs, T.A., Vitaliano, P.O., Linnoh,
M. & Weitzman, E.D. (1980) Effect of alcohol on
sleep and night time plasma growth hormone and
cortisol concentration. Journal of clinical End*
crinologv and Metabolism, 51,759-164.
18.Sambhi, M.P., Weil, M.J. & Udhoji, V.N. (1962)
Pressor response to norepinephrine in humans before
and after corticosteroids. American Journal of Physiology, 203,961-963.
19. Brown, M.J. & Macquin, I. (1981) Is adrenaline the
cause of essential hypertension? Lancet, ii, 10791082.