Content uploaded by Ian Macdonald
Author content
All content in this area was uploaded by Ian Macdonald on Nov 25, 2014
Content may be subject to copyright.
British Journal
of
Nutrition
(1994),
71,
835-848
835
The effect
of
meal size on the cardiovascular responses
to
food
ingestion
BY
MICHAEL
B.
SIDERY AND IAN
A.
MACDONALD
Department
of
Physiology and Pharmacology, University
of
Nottingham Medical School,
Clfton Boulevard, Nottingham NG7 2UH
(Received 23 July 1993
~
Revised
1.5
September 1993
-
Accepted 23 September 1993)
Cardiac output (CO
;
indirect Fick), blood pressure (BP) and heart rate
(HR;
oscillometry), superior
mesenteric artery blood flow (SMABF; Duplex Doppler) and calf blood flow (CBF; venous occlusion
plethysmography) were recorded in the fasted state and for
120
min following the ingestion of
1, 2,
and
3
MJ, high-carbohydrate meals in eight healthy females. BP was unchanged following food.
HR
(P
<
0.0005)
and CO
(P
<
0.005)
rose significantly following all three meals. Integrated increments in
CO over the postprandial period were greater after
3
MJ compared with the
1
and
2
MJ meals
(P
<
0.05).
SMABF rose significantly following all three meals. The pattern of blood flow response was
significantly different between the
1
and
3
MJ meals (interaction effect
P
<
0.02,
ANOVA), with blood
flow
after the
3
MJ meal being significantly greater than flow after the
1
MJ meal at
15,60,
and
90
min.
Similarly, the pattern of response was significantly different after the
2
and
3 MJ
meals (interaction
effect
P
<
0.03,
ANOVA), with blood flow being significantly greater at
15
and
90
min after the
3
MJ
meal. CBF fell significantly in the first
15
min after the
3
MJ meal and then recovered towards baseline
values. No other significant changes in CBF were recorded. There are substantial peripheral and central
cardiovascular changes after food in man and there appears to be a relationship between meal size and
the extent of these changes.
Cardiac output: Blood flow: Meal size: High-carbohydrate meal
In the fasted, healthy subject, blood supply to the gastrointestinal tract accounts for a
substantial proportion of the cardiac output (Donald, 1983). In the postprandial state
blood flow to the mesenteric organs increases markedly from these fasted levels (Norryd
et
al.
1975; Moneta
et
al.
1988). Despite the mesenteric hyperaemia associated with food
ingestion, cardiovascular homeostasis is maintained, with blood pressure in the young
remaining unchanged (Kelbaek
et
al.
1989; Heseltine
et
al.
1990), whilst both heart rate and
cardiac output increase (Gladstone, 1935; Waaler
et
al.
1990).
The pattern of the cardiovascular responses to food is partly dependent upon the
composition
of
the meal (Qamar
&
Read, 1988; Sidery
et
al.
1990). There
is
also a
relationship between the postprandial increment in cardiac output and meal size, with a
larger and more prolonged cardiac response following a large meal when compared with
a small meal (Waaler
et
al.
1991). It was suggested that the greater increment in cardiac
output following the larger meal is a consequence of a greater demand in the mesenteric
bed.
The rate of absorption of nutrients in tissues following food ingestion is a function of the
blood flow to those vascular beds (Laakso
et al.
1990). The functional advantages of a rapid
nutrient delivery to the periphery following absorption from the gut are clear and an
increased cardiac output would facilitate this process. The aim of the present investigation
836 M.
B.
SIDERY
AND
I.
A. MACDONALD
was to study the effect of meals of differing energy content but constant fat:carbohydrate
ratio on superior mesenteric artery blood flow, cardiac output and blood flow in skeletal
muscle, using the calf as a muscle bed. Sampling of arterialized venous blood allows
simultaneous measurements of plasma insulin and whole-blood glucose. Samples were also
stored for noradrenaline analysis. Although there are limitations in the degree to which
plasma noradrenaline levels reflect sympathetic nervous system activity (Esler
et
al.
1988),
there is a surprisingly good correlation between such measurements and the activity of the
sympathetic nervous system assessed by recording sympathetic nerve firing in the leg
(Wallin
et
al.
1981).
METHODS
Eight healthy female subjects (body mass index (BMI) 21.3-25.2 kg/m2, age range 21-26
years) were recruited for the study. None were taking any medication other than the oral
contraceptive pill. All gave written informed consent to the study which was approved by
the University of Nottingham Medical School Ethics Committee.
On three occasions, following an overnight fast of 10 to
15
h duration, the subjects were
studied supine in a temperature-controlled room (dry bulb 26f2"). On arrival in the
laboratory, subjects rested supine for 30 min, during which time the monitoring equipment
was attached and a cannula for blood sampling inserted retrogradely under local
anaesthetic into a vein on the dorsum of the right hand. The cannula was kept patent with
a slow infusion of 154mM-NaC1 and the hand rested in a box circulated with warm
air (55-60") to obtain 'arterialized
'
venous blood samples. The total volume of saline
infused did not exceed 350ml on each occasion. The arterialization of venous blood
for the estimation of arterial blood glucose levels has been validated previously
(McGuire
et
ul.
1976). Heating of the hand leads to a reduction in the transit time and thus
minimizes the extraction of glucose by the hand tissue, providing a realistic alternative to
arterial sampling. This method of hand heating does not affect body temperature or
forearm blood flow on the other side (Gallen
&
Macdonald, 1990) but as yet it is not known
whether this method provides accurate estimation of arterial plasma catecholamine levels.
Following the rest period two sets of measurements
of
all variables were made at
15
rnin
intervals. Following this the subjects sat up and ate (no longer than 15 rnin was taken to
complete the meals in all cases), in a randomized order, a standardized high-carbohydrate
meal containing either 1, 2, or
3
MJ (approximately
84%
energy from carbohydrate; see
Table
1).
Subjects immediately returned to the supine position and measurements were
made at 15 and 30 rnin after completion of the meal and subsequently at
30
rnin intervals
for another 90 min (a total of 120 min postprandially). Blood samples were taken at 30 rnin
intervals. At least 1 week passed between each visit. Volume occupied by the meals did
differ, with the 1 MJ meal being approximately 350 ml, the
2
MJ meal 520 ml and the 3 MJ
meal
585
ml. The observers measuring blood pressure, cardiac output and limb blood flow
were blinded as to the energy content of the meals.
The subjects were also studied on
a
fourth occasion, in which the same protocol was
followed, except that
580
ml water at room temperature was ingested following the baseline
period. No blood samples were taken and post-water measurements were made every
15
min for
45
min only. In addition, superior mesenteric artery blood flow was measured
5
min after water ingestion.
Cardiac output was measured using the indirect Fick principle, monitoring respiratory
gases with
CO,
as an indicator. The subjects were attached to the breathing equipment
using a mouthpiece and wearing a nose-clip. CO, concentrations were measured using an
infrared
CO,
analyser (901 Mk 2;
P.
K. Morgan, Chatham, Kent).
CO,
production was
determined from measurements of ventilation rate (with a flowmeter) and mixed expired
CARDIOVASCULAR
RESPONSES
TO
MEAL
SIZE
837
Table
1.
Meal
composition
Percentage energy from
Meal size Ingredients Starch Simple
sugars
1
MJ
Cornflakes and skimmed milk
54
29
2
MJ
Cornflakes, skimmed milk, bread and honey
48
36
3 MJ Cornflakes, skimmed milk, sugar, bread and honey
43
44
air. End-tidal Pco, was used to estimate systemic arterial CO, tension (Pa,,,) and mixed
venous (pulmonary artery) CO, tension (Pv,.,,) was determined with a CO, rebreathing
technique. The CO, concentration in the rebreathing mixture was approximately 10
YO.
Rebreathing continues until there is no difference between expired and inspired
CO,
concentration (approximately 8-10
s)
measured at the mouthpiece. Cardiac output was
calculated from the measurements of
Piico,
and the estimated Paco2 and PVco,. This
technique correlates well
(Y
0.96, 95
Yo
CI of the difference -0.37 to
+
0.47 litres/min) with
cardiac output measurements made by thermodilution (Cowley
er al.
1986). The validity of
a
CO,
rebreathing method to determine cardiac output, particularly at rest, has been
questioned (Reybrouck
&
Fagard, 1990). However, in that case the method used involved
an exponential-based estimation
of
PV,.,?. The equilibrium method used in the present
study to measure PVCoiL appears to be more reliable, showing closer agreement than the
exponential method with cardiac output measured using both the direct Fick (Muiesan
et
al.
1968) and dye dilution (Hinderliter
et
al.
1987) techniques. The reproducibility of
PVco,
measurement by the equilibrium method was also shown to be excellent (Muiesan
er
al.
1968).
Time-integrated increases in cardiac output over the postprandial period were also
calculated. The cardiac output increments above baseline for the first 15 min period and the
subsequent 30,60,90 and 120 rnin measurements were calculated and the total extra blood
volume pumped by the heart over each time period and for each subject calculated from
these cardiac outputs and expressed in litres with the standard error of the mean.
Heart rate and systemic arterial blood pressure were measured by an automated
oscillometric device (Accutorr 1A; Datascope, Paramus,
NJ,
USA) with the cuff placed
around the right upper arm. Measurements were made once every
5
min and the mean of
three measurements was calculated for the 15 min time windows; the mean of six was
calculated for each of the
30
min time windows. The coefficient of variation
of
repeated
measurements in fasted subjects using this technique in this laboratory is
4
YO
for heart rate,
2%
for systolic blood pressure and 2% for diastolic blood pressure.
Superior mesenteric artery (SMA)
flow
was measured by transcutaneous Doppler
ultrasound (Diasonics Prisma
;
Diasonics International, Les Vlis, France) with a convex
linear array probe with variable receiver characteristics. The B-mode imaging system
operates with a centre frequency of 3.5 MHz. Doppler frequency of the probe is 3 MHz.
The anatomical position of the SMA makes visualization of the proximal part of the vessel
using ultrasound comparatively easy. Complete data on SMA flow for all three occasions
were obtained in seven
of
the eight subjects.
The vessel of interest was visualized with a sagittal scan of the abdomen. Care was taken
to ensure that the entire vessel lumen was insonated and the sample volume was placed in
the proximal part
of
the artery, several centimetres from the bifurcation
of
the SMA from
838
M.
B.
SIDERY AND
1.
A.
MACDONALD
the aorta. This avoids the introduction of error in flow measurements as a result of
turbulent flow which may be present near the junction of the two arteries.
The angle of insonation was recorded and used
to
convert the Doppler Shift values (kHz)
into blood flow velocity (cm/s). Care was taken to ensure that wherever possible the same
angle of insonation was used in each individual (mean angle of insonation was 39",
SD
3.5"). Error brought into the flow calculation due to an angle of insonation
of
39"
would be in the range of
7
YO
(Gill, 1985). Recordings were made with the subjects' breath
held in mid inspiration and mean values
of
time-averaged velocity (TAV) were taken from
at least eight Doppler waveform complexes. Using manually operated on-screen callipers
systolic vessel diameter was measured during the baseline period of each study session. It
was assumed that vessel diameter remained unchanged during the experimental protocol
and the mean of the vessel diameter measurements for each subject was used with the TAV
in the volume flow calculations. Blood flow was calculated from the equation:
blood flow
=
3.142
x
D2
x
TAV
x
60/4 ml/min,
where D is the vessel diameter.
A comparison of SMA flow measurement by Duplex ultrasound (calculated from the
time-averaged flow velocities) and electromagnetic flowmetry has been made in which a
strong correlation between the two techniques was found (Nakamura
er
al.
1989). The same
paper reported a coefficient of variation in the measurement of volume flow very similar to
our own.
Calf blood flow was measured by venous occlusion plethysmography (Greenfield
et
al.
1963) with mercury-in-silastic strain gauges (Whitney, 1953). An occlusion cuff placed
around the thigh was inflated to 40 mmHg to prevent venous return from the limb. Inflation
took place in a cyclical manner, with the change in calf circumference being measured using
the strain gauge. Flow both in and out of the foot was prevented during measurements
using an occlusion cuff placed around the ankle and inflated to
200
mmHg. During each
measurement period, a minimum of six measurements of
flow
were made and the mean
value used in the subsequent statistical analysis. The coefficient of variation for
measurement of calf blood flow by venous occlusion plethysmography on different days
is 11.5
O/O
(Roberts
et
af.
1986). In our laboratory under resting conditions a 10% alteration
in blood flow is the minimum change that can be detected using this technique.
Vascular resistances for the calf and the SMA were calculated from blood flow values
and mean arterial blood pressure (obtained using the Accutorr 1A). For these calculations
it was assumed that brachial artery pressure was a reliable index
of
calf and mesenteric
perfusion pressures, although disparity between central and peripheral blood pressure
measurements have been noted (Rowel1
et
a/.
1968). Any differences are likely to be
minimized in healthy volunteers, and be similar on the three occasions.
Arterialized blood samples were used to measure blood glucose
(YSI
23 AM; Yellow
Springs Industries, Yellow Springs,
OH,
USA) immediately. The remainder of the
arterialized blood samples were centrifuged and the plasma separated. Plasma (3 ml) was
mixed with
75
pl
EGTA-glutathione (antioxidant) and stored at
-80"
for later
determination of noradrenaline and adrenaline concentrations using HPLC with
electrochemical detection (Macdonald
&
Lake, 1985). All samples for any one subject were
run on the same day. The intra-assay coefficient
of
variation was 6
YO
for noradrenaline and
8
YO
for adrenaline, the inter-assay values being
8
YO
and
10
YO
at the levels of catecholamine
present in these samples. Plasma was also stored at -20" for subsequent determination of
insulin concentration by radioimmunoassay using a double-antibody technique developed
in-house. The intra-assay coefficient of variation is
8
Yo
and the inter-assay value is 12
YO
over a range of plasma insulin from 5 to
50
mU/1.
CARDIOVASCULAR RESPONSES
TO
MEAL SIZE
Statistical analysis
Statistical analysis of the results was performed by two-way analysis of variance with
repeated measures (ANOVA) using the package BMDP (BMDP Statistical Software;
Los
Angeles, CA, USA). Where the ANOVA indicated a significant treatment--time interaction
the exact level of significance at each time point was calculated using
a
paired
t
test, using
the variance term for the interaction from the ANOVA table, with a Bonferroni
correction applied for multiple testing.
For clarity, data are presented in the Figures as changes from baseline, each point being
the mean with its standard error. The data reported in the text on the responses to the meals
are the maximum changes from baseline values, and the
95
%
confidence intervals
(CI)
of
the changes unless stated otherwise.
839
RESULTS
Control study
There were no significant changes in any of the measured variables following the ingestion
of water.
Systolic and diastolic
blood
pressure
There was no difference in blood pressure in the fasted state on the three occasions. Neither
systolic nor diastolic blood pressure changed significantly following any of the meals in
these young subjects.
Heart ra
t
e
Heart rate rose significantly following all three meals
(P
=
O.OOOOl), with a maximum rise
of 5.9 beats/min after the
1
MJ meal (95
YO
CI
of the increase 2.4 to 9.4 beats/min), 10.5
beats/min after the
2
MJ meal
(95%
C1 of the increase 6.5 to 143 beats/min) and 13.3
beats/min after the 3 MJ meal (95
%
CI of the increase
8.8
to 17.8 beats/min). There was
a significant difference in the pattern of response in heart rate following the
1
and 3 MJ
meals (interaction effect
P
=
0.0005,
ANOVA; Fig.
1).
Heart rate was significantly higher
15 min after the 3 MJ compared with the 1 MJ meal
(t
test with Bonferroni correction).
There was no difference between the pattern of response after the 2 and 3 MJ and the
1
and
2 MJ meals. There was a significant difference in individual maximum increases in heart
rates between the 1 and 2 MJ meals only
(t
test
P
<
0.05;
Fig. 2).
Cardiac output
Cardiac output rose significantly after all the meals
(P
=
0.0001).
The maximum rise in
cardiac output following the
1
MJ meal was 2.19 l/min (95% CI of the increase 0.96
to
3.42 l/min), 1.78 l/min after the 2 MJ meal (95
YO
CI of the increase
0.76
to 2.79 I/min) and
3.1
5
1/min after the 3 MJ meal (95
%
CI
of the increase
1.58
to 4.72 I/min; Fig. 1). There
was a significant difference between individual maximum increases in cardiac output
between the 2 and 3 MJ meals
(t
test
P
=
0.03). The mean time-integrated postprandial
increments in cardiac output after the three meals were
80.1
(SE
24.5), 104.1
(SE
42.2) and
244.0
(SE
49.7)
1
respectively, with the time-integrated cardiac output after the
3
MJ meal
being significantly greater than after the 2 MJ meal
(t
test
P
=
0.04; Fig. 2).
Totul peripheral resistance (TPR)
TPR fell significantly following all three meals (95
%
CI of the change
-
3.5 to
-
10.7,
-
3.3
to
-
17.3, and -3.1 to
-20.1
respectively). However, the pattern of responses was
different, with TPR remaining low throughout the experimental period after the largest
840
M. B.
SIDERY AND
I.
A.
MACDONALD
v)
W
P
z
-22
"
-15
0
15
30
45
60
75
90
105 120
r
u
-0.5l
I
I
I,
IIt,
0
15
30
45
60
75
90
105 120
-16
I
I
I
I
I
I
I
I
0
15
30
45
60
75
90
105
120
Time
(min)
Fig.
1.
Heart rate (beats/min;
a),
cardiac output (I/min;
b)
and total peripheral resistance (TPR;
c)
changes from
baseline values (shown as a dotted line) following the ingestion of
1
(O),
2
(A)
and
3
MJ
(0)
meals. Values are
means with their standard
errors
represented
by
vertical lines. There was a significant rise in both heart rate and
cardiac output following all the meals. TPR fell initially after all meals. The pattern of response was different
between the largest and smallest meals for all three variables.
meal and recovering to values not significantly different from baseline values
30
min after
the small meal (interaction effect
P
=
0.02,
ANOVA;
Fig.
1).
SMA
bloodj2ow
SMA
flow rose significantly following all three meals, with a peak flow
of
836 ml/min being
reached at
15
min after the
1
MJ
meal (95
YO
CI
of
the change 66 to 474 ml/min), and a
peak flow
of
825 ml/min being reached at 60 min after the
2
MJ
meal (95
%
CI
of
the
CARDIOVASCULAR
RESPONSES
TO
MEAL
SIZE
84
1
820
2.g
720
C
.-
-
?<
620
2g
g
520
Eo
420
.E
Q
zz
320
=m
220
120
I
~
I
1
2
3
Meal
size
(MJ)
Fig.
2.
Graphs showing the relationship between meal size
(MJ)
and
(a)
the maximum increase in heart rate
(beats/min),
(b)
the time-integrated increase in cardiac output above baseline
(I)
and
(c)
the total
blood
flow above
baseline through the superior inesenteric artery
(SMA;
ml/min).
The
maximum increase in heart rate after the
2
and
3
MJ meals was significantly greater than after the
1
MJ
meal. The time-integrated increase in cardiac output
was significantly greater after the
3
compared with the
1
and
2
MJ
meals. The total
blood
flow above baseline
through the
SMA
was significantly greater after the
3
compared with the
1
MJ
meal
only.
change
240
to
673
ml/min).
A
peak flow of 1189 ml/min was reached
15
min after the
3
MJ
meal (95%
CI
of the change
514
to 956 ml/min). The pattern of blood flow response was
significantly different between the
1
and
3
MJ
meals (interaction effect
P
=
0-01,
ANOVA),
with blood flow after the
3
MJ
meal being significantly greater than flow after the
1
MJ
meal at
15,
60,
and 90 min
(t
test with Bonferroni correction at each time point). Similarly,
the pattern of response was significantly different after the
2
and
3
MJ
meals (interaction
effect
P
=
0.02,
ANOVA),
with blood flow being significantly greater
at
15
and 90 min after
the
3
MJ
meal
(t
test with Bonferroni correction; Fig.
3).
842
1000
F
900
800
E
700
&
600
2
500
v)
400
300
.-
-
v)
g
200
:
100
r
uo
M.
B.
SIDERY AND
I.
A. MACDONALD
-100'
I I
I
I
I
I
I
'
0
15
30
45
60
75
90
105 120
Time
(min)
Fig.
3.
Superior mesenteric artery blood flow (SMABF; ml/min;
a)
and calf blood flow (m1/100 ml per min;
b)
changes from baseline values (shown as a dotted line) following the ingestion of
1
(O),
2
(A)
and
3
MJ
(0)
meals.
Values are means with their standard errors represented by vertical lines. SMABF increased following all three
meals. The pattern of response was different between the largest and smallest meals. Calf blood flow fell
significantly following the
3
MJ meal only. The rise in flow from the
15
min value to the value at
90
rnin after the
I
MJ meal and
120
min after the
3
MJ meal was significant.
The total volumes
of
blood flow above baseline in the postprandial period after the three
meals were 12.1
(SE
5.9), 32.9
(SE
8.9) and 47.7
(SE
6.6) 1. Total flow volume after the 3 MJ
meal was significantly greater than that after the
1
MJ meal only
(t
test
P
=
0.02; Fig. 2).
Peak
flow
velocity
Peak flow velocity increased significantly following all three meals
(P
<
0.001).
The
maximum changes in peak flow velocity after the three meals were 42 cm/s (95
%
CI of the
increase 15 to 69 cm/s),
50
cm/s (95
YO
CI of the increase 29 to
7
1
cm/s) and
8
1
cm/s (95
Yo
CI
of
the increase 63 to 99 cm/s) respectively. There was no significant difference in the
pattern of flow velocity changes after the three meals.
Calf
blood
jlo
w
Although calf blood flow tended to fall initially following all three meals, there was a
significant fall in the first 15 rnin only after the 3 MJ meal (95
O/O
CI of the change -0.36
to
-
3.22 m1/100 ml per min). The trend was then for calf blood flow to rise, although flow
did not rise above baseline after any of the meals during the experimental period. The rise
from the
15
rnin calf blood flow value to the 120 min value following the 3 MJ meal was
CARDIOVASCULAR
RESPONSES
TO
MEAL
SIZE
843
TT
(a)
7
-30-15
0
15
30
45
60
75
90
105120
65
-
-
55
-
T
-30
-15
0
15
30
45 60 75
90
105 120
Time
(min)
Fig.
4.
Blood
glucose
(a)
and plasma insulin
(b)
before
(t
0)
and after the ingestion
of
1
(0).
2
(A)
and
3
MJ
(0)
meals. Values are means with their standard errors represented by vertical lines. There was a significant rise
after
all
three meals. The pattern
of
response in blood glucose was significantly different between the
3
MJ
meal
and the
two
smaller meals. The plasma insulin response was different between the
3
and
2
MJ
meals and the
smallest meal. The area under the glucose curve was greater after the
3
MJ
meal compared with the
2
MJ
meal
only.
significant (95
YO
CI
of the increase 0.27 to 3.97 m1/100 ml per min). The rise from the
15 min value
to
the value at
90
min following the
1
MJ meal was also significant (95
YO
CI
of the increase
0.5
to 3.78 m1/100 ml per min; Fig. 3).
Blood glucose
Blood glucose rose significantly after all the meals
(P
=
0.0001, ANOVA). There was no
significant difference between the peak blood glucose concentrations following any of the
three meals. The areas under the glucose curves following the
2
MJ and the
3
MJ meals
were significantly different
(t
test,
P
=
0.001).
There was no difference between the
1
and
3 MJ meals (Fig.
4).
Plasma insulin
Plasma insulin concentrations rose significantly following the three meals
(P
=
0.0001,
ANOVA). Again, there was no difference in the peak insulin response following any of the
three meals. There was a difference in the pattern of response following the
1
and 3 MJ
meals (interaction effect
P
=
0.002,
ANOVA; Fig.
4).
Plasma noradrenaline and adrenaline
Baseline plasma noradrenaline levels were 1.17
(SE
0.09),
1.30
(SE
0.22)
and 0.90
(SE
0.13) nmol/l before the 1, 2 and
3
MJ meals respectively. Plasma noradrenaline rose
only after the
3
MJ meal
(P
=
0.0001,
ANOVA), from
a
baseline value of 0.83
844
M.
B.
SIDERY AND I.
A.
MACDONALD
2.0
5
1.8
1
-
1.6
E,
I
=
1.4
1.2
2
1.0
it
m
0.8
0
c
-30-15
0
15 30 45
60
75
90
105
120
Time
(min)
Fig.
5.
Plasma noradrenaline changes from baseline values following the ingestion of
1
(O),
2
(A)
and
3
MJ
(0)
meals. Values are means with their standard errors represented by vertical lines. Plasma noradrenaline rose only
after the
3
MJ
meal
(P
<
0.009,
ANOVA). This was the only time point at which plasma noradrenaline was
significantly above baseline.
(SE
0.06) nmol/l to
1.38
(SE
0.18)
nmol/l
15
min after the meal (95 YO CI of the increase
0.07
to 1.03 nmol/l). This was the only time point at which plasma noradrenaline was
significantly above baseline (Fig.
5).
Baseline plasma adrenaline levels were
0.20
(SE
0.03),
0.25
(SE
0.09)
and 0.18
(SE
0.03)
nmol/l before the
1,2
and
3
MJ meals respectively. Plasma adrenaline did not change after
ingestion of any of the meals.
DISCUSSION
Cardiac output rose significantly following ingestion of all three high-carbohydrate meals.
The increase was most rapid following the
1
MJ meal, peaking at
15
min. Cardiac output
peaked between
30
and 60 min after the
2
and 3 MJ meals. There is some variance as to the
time at which the peak cardiac response is achieved following food ingestion (Fagan
et
al.
1986; Waaler
et
al.
1990, 1991), although meal composition appears to be important in the
time course of these changes (Dagenais
et
al.
1966). The cardiac response after both the
1
and
2
MJ meals is reflected in the mesenteric response, with both peaking at
15
min after
the
1
MJ and 60 min after the
2
MJ meal. The time courses of the initial changes in cardiac
output do not reflect the initial blood flow changes in the mesenteric bed following the 3 MJ
meal, when maximum hyperaemia in the mesentery was attained within the first
15
min and
the cardiac output peak at about 60 min. However, both cardiac output and mesenteric
blood flow were maintained at elevated levels throughout the postprandial period after the
3
MJ meal. Postprandial increases in stroke volume have been demonstrated previously
(Kelbaek
et
al.
1987,
1989; Waaler
et
al.
1991)
although not consistently (Fagan
et
al.
1986). Stroke volume and heart rate have been shown to contribute equally to the
postprandial rise in cardiac output (Waaler
et
al.
1991).
In the present study calculated
stroke volume did increase significantly after all three meals.
A
fall in TPR may account
for a fraction of the postprandial rise in stroke volume (by reducing afterload), but
increases in stroke volume of
41
YO
(Kelbaek
et
al.
1989) suggest a rise in pre-load and
possibly in contractility of the heart. Thus, the exact mechanism of the increase in stroke
volume is not clear either from the literature or from this study.
A
‘dose-response’ relationship exists between meal size and the total volume of blood
pumped by the heart above baseline over a
2
h period (Waaler
et
al.
1991). The present
study confirms and extends this relationship and also demonstrates that the same
CARDIOVASCULAR
RESPONSES
TO
MEAL
SIZE
845
relationship exists between the meal size and both the maximum change in SMA flow and
the total volume of blood flowing through the artery over a 120 min postprandial period.
Maximal hyperaemia was reached within
15
min after ingestion of two of the three
meals. This rapid response following meals high in carbohydrate has been observed after
both solid (Sidery
et
al.
1990) and liquid (Moneta
et
al.
1988; Qamar
&
Read, 1988)
carbohydrate meals. In the present study maximal flow following the intermediate meal was
not reached until 60 min after food, but blood glucose levels peaked 30 min after food. A
delayed hyperaemia has been observed after a high-fat meal which does not coincide with
the peak blood glucose levels (Sidery
et
al.
1990). This observation, that peak blood glucose
levels and peak mesenteric blood flow do not always coincide, is confirmation that the
mechanisms mediating mesenteric blood flow are complex and not fully understood.
In the present study TPR fell following all three meals, with no difference in the extent
of the fall. This is despite an increasingly greater initial hyperaemia with an increase in meal
size. There was an initial fall in calf blood flow following the 3 MJ meal only. Assuming
that blood pressure measured at the upper arm reflects peripheral blood pressure, the
recorded fall in blood flow in the calf is the result of increased vascular resistance: could
this increase in resistance in the calf be responsible for the fact that values of TPR are no
different after the meals, despite the significantly greater initial mesenteric response
following the 3 MJ meal? A postprandial increase in vascular resistance in vascular beds
other than those associated with digestion and absorption has been demonstrated
previously (Sidery
et
a/.
1990). Skeletal muscle blood flow increased significantly following
the initial fall after ingestion of the largest of the meals.
Blood glucose rose significantly following all three meals. There was a slight anomaly
between the blood glucose response following the
1
MJ and the intermediate size meal.
Different glycaemic responses to meals of identical sugar, starch and energy contents but
varying insoluble and soluble dietary fibre contents have been demonstrated (Torsdottir
et
al.
1989). The results of the present study may be a consequence of
a
difference in the
nature of the carbohydrate in the
2
and
1
MJ meals. Peak mesenteric blood flow was not
reached until 60min after the intermediate meal, in contrast to the other meals. This
different pattern of response might also be related to carbohydrate type, and clearly
warrants further study.
Plasma noradrenaline rose significantly following the largest of the three meals only.
Increases in plasma noradrenaline have been observed in man following glucose ingestion
(Rowe
et
al.
1979, 1981). Similarly, increases in sympathetic activity, detected using
microelectrode recordings, have been shown following ingestion of glucose and xylose in
man (Berne
et
al.
1989). A link between the increased plasma noradrenaline levels and the
cardiovascular changes associated with food ingestion is not clear. There is a differential
effect of meals of different composition on plasma catecholamines, with 120% (0.98 to
2.22 nmol/l) increases in plasma noradrenaline levels following a 2.4 MJ high-
carbohydrate meal and no changes after a high-fat meal of the same energy content
(Heseltine
et
al.
1990).
To summarize, a
'
dose-response
'
relationship between the cardiac and mesenteric
response and the energy content of meals has been demonstrated.
A
similar relationship
between cardiac output and meal size has been observed for meals ranging from
approximately
2
to
5
MJ (Waaler
et al.
1991). There was a decrease in vascular resistance
in the mesenteric bed after food and substantial increases in superior mesenteric blood flow
were observed. The mechanism responsible for the substantial postprandial increments in
cardiac output is not clear. TPR falls after food ingestion (Fagan
et
al.
1986; Kelbaek
et
al.
1989) and blood pressure remains either unaltered (Heseltine
et
al.
1990) or there is
a slight widening of pulse pressure (de Mey
et
al.
1987) in young, healthy subjects. The
846
M.
B.
SIDERY AND
I.
A. MACDONALD
cardiac changes seen in the postprandial state may in part be mediated by the autonomic
nervous system. Unloading of peripheral baroreceptors would lead to increased activity of
sympathetic cardiac fibres. Reduced atrial filling and distension would lead to increased
vasomotor tone.
Deleterious effects of food ingestion are well documented in angina patients, with a
significant reduction in exercise capacity and the time until depression
of
the ST segment
of the electrocardiogram in the postprandial state compared with the fasted state (Fagan
et
ul.
1982; Cowley
et
al.
1991). Similarly, there is a reduction in exercise tolerance in
chronic heart failure patients after food ingestion (Muller
et
al.
1992). The mechanisms of
these effects are unclear, but the evidence suggests that abnormal central and peripheral
cardiovascular responses may be responsible. In angina the degree of myocardial ischaemia
is the same during exercise
in
the pre- and postprandial states (Colles
et
ul.
1993), lending
credence to the hypothesis that increased postprandial myocardial oxygen consumption
and not reduced coronary blood flow results in the earlier onset
of
ischaemic pain. Reduced
exercise tolerance in heart failure is almost certainly due
to
abnormal peripheral blood flow
responses. In view
of
this, food ingestion may act to exacerbate the already abnormal state
in skeletal muscle during exercise.
Postprandial hypotension has been observed in both institutionalized elderly (Lipsitz
et
al.
1983) and also in healthy elderly subjects (Lipsitz
&
Fullerton, 1986; Heseltine
&
Potter,
1990).
High-carbohydrate meals result in significantly greater falls in blood pressure
compared with meals
of
any other composition (Potter
et
al.
1989). The mechanism for this
is
not clear. These subjects do not demonstrate significant cardiac responses to food
ingestion, despite similar vascular changes in the mesenteric bed to those seen in the young
(Sidery
et
al.
1993). Food ingestion represents a substantial cardiovascular challenge, with
the magnitude of the cardiovascular responses being directly related to the energy content
of the meal ingested. The implications of such findings for those patients with compromised
cardiovascular and autonomic function and the elderly are clear.
A
smaller meal elicits a
smaller central and mesenteric response. The impact of such meals in individuals with
pathology of the autonomic or cardiovascular systems will be less, hopefully with
cardiovascular homeostasis being maintained through more subtle vascular changes which
are more likely to be within the capabilities of these patients. Advice to patients with regard
to meal composition is more complex. High-carbohydrate meals elicit rapid central and
mesenteric responses, and, as mentioned, tend to lower blood pressure in the healthy elderly
more than any other meal type. However, high-fat meals, whilst delaying peak
cardiovascular responses, result in more prolonged mesenteric hyperaemia than other
meals (Moneta
et
al.
1988;
Sidery
et
al.
1994). Small, mixed meals containing less-readily
available, complex carbohydrate are likely to be the least challenging of all meal types to
the individuals discussed.
This work was supported by a Project Grant from the Wellcome Trust.
REFERENCES
Berne, C., Fagius,
J.
&
Niklasson,
F.
(1989). Sympathetic response
to
oral carbohydrate administration: evidence
from microelectrode nerve recordings.
Journtz/
of
Clinical lnwsrigaiion
84,
1403-1409.
Colles,
P.,
Juneau, M., Gregoir,
J.,
Larivee,
L..
Desideri,
A.
&
Waters,
D.
(1993). Effect
of
a standardized meal
on the threshold
of
exercise-induced myocardial ischemia in patients with stable angina.
Journa/oftl?e
Anrericcm
Cowley,
A.
J.,
Fullwood,
L.
J.,
Stainer,
K.,
Harrison,
E.,
Muller, F.
&
Hampton,
J.
R.
(1991). Post-prandial
Cowley,
A.
J..
Murphy, D.
T.
&
Stainer,
K.
(1986).
A
non-invasive method
for
measuring cardiac output: the
Cd/<’gtJ
of
Ct~di~l~g~’
21,
1052-1057.
worsening
of‘
angina: all
due
to changes in cardiac output.
L3riti.d~
Heart
Journal
66,
147-150.
effect
of
Christmas lunch.
Lancer
ii,
1422-1424.
CARDIOVASCULAR
RESPONSES
TO
MEAL
SIZE
a47
Dagenais, G. R., Oriol, A.
&
Gregor, M. (1966). Haemodynamic effects of carbohydrate and proteins in man: rest
and exercise.
Journal
(j'
Applied Physiologj, 21,
1157-1
162.
de Mey,
C.,
Enterling,
D.,
Brenel,
E.
&
Meineke,
I.
(1987). Postprandial changes in supine and erect heart rate,
systemic blood pressure and plasma noradrenaline and renin activity in normal subjects.
European Journal
qf
Clinicul
Phorrncicology 32,
47 1476.
Donald,
D.
E.
(1
983). Splanchnic circulation. In
Handbook
qf
Physiology,
section
2,
part
1.
Bethesda
:
American
Physiological Society.
Esler, M., Jennings,
G.,
Korner, P., Willett,
I.,
Dudley,
F.,
Hasking,
G.,
Anderson, W.
&
Lambert,
G.
(1988).
Assessment of human sympathetic nervous system activity from measureinents
of
norepinephrine turnover.
H.ipertension
I
I,
3-20.
Fagan,
T.
C., Gourley, L. A., Sawyer, P.
R.,
Lee, J.
T.,
Corneux, C.
B.,
Oexman, M.
J.
&
Gaffney,
T.
E. (1982).
Cardiovascular effects of a meal
:
clinical implications.
Clinical
Research 30,
7A.
Fagan,
T.
C., Sawyer, P.
R.,
Gourley, L. A,, Lee,
J.
T.
&
Gaffney, T. E. (1986). Postprandial alterations in
haemodynamics and blood pressure in normal subjects.
American Journal
qf
Curdiology 58,
636-641.
Gallen,
I.
W.
&
Macdonald,
I.
A.
(1990). Effect of
two
methods of hand heating on body temperature, forearm
blood flow and deep venous oxygen saturation.
American Journul
qf
Physiology 259,
E639-E643.
Gill,
R.
W. (1985). Measurement of blood flow by ultrasound: accuracy and sources
of
error.
Ultrasound iri
Medicine and Biologj, 11,
625-641.
Gladstone,
S.
A.
(1935). Cardiac output and related functions under basal and postprandial conditions.
Arc/iive.s
of
Internal Medicine 55,
533-546.
Greenfield, A.
D.
M., Whitney, R. J.
&
Mowbray, J.
F.
(1963). Methods for the investigation of peripheral blood
flow.
Briti.vh Medical
Bulletin
19,
101-109.
Heseltine,
D.
&
Potter,
J.
F.
(1990). Commentary: postprandial hypotension in elderly people.
Age andAg~ii7g 19,
Heseltine, D.. Potter.
J.
F.,
Hartley,
G.,
Macdonald,
1.
A. &James,
0.
F. W. (1990). Blood pressure, heart rate
and neuroeiidocrine responses to a high carbohydrate
and
a high fat ineal in healthy young subjects.
Clinical
Science
79.
5 17-522.
Hinderliter, A. L., Fitzpatrick, M.
A,,
Scliork, N.
&
Julius,
S.
(1987). Research utility of non-invasive methods
for
measurement
of
cardiac output.
Clinicul
Pharmacology and Therapeutics 41,
419425.
Kelbaek. H., Munck,
O.,
Christensen, N. J.
&
Godtfredsen, J. (1987). Autonomic nervous control of postprandial
haemodynainic changes at rest and upright exercise.
Journal
of
Applied Physiology 63,
1862--1865.
Kelbaek,
H.,
Munck,
0..
Christensen, N. J.
&
Godtfredsen,
J.
(1989). Central haemodynamic changes after a
meal.
Britivh Heart Journul61,
506-509.
Laasko, M., Edelman,
S.
V.,
Brechtel,
G.
&
Baron,
A.
D. (1990). Decreased effect of insulin to stimulate skeletal
muscle blood flow in obese man. A novel mechanism for insulin resistance.
Joiirnal
of'
Cliniccil
lnvestigation
85,
18461 852.
Lipsitz, L.
A.
&
Fullerton, K. J. (1986). Postprandial blood pressure reduction in healthy elderly.
Journal Ofrhl.
Ainuricun
Gerontologicul
Societj,
34,
267-270.
Lipsitz, L. A,, Nyquist, R., Wei,
J.
Y.
&
Rowe, J.
W.
(1983). Postprandial reduction in blood pressure in the
elderly.
Nivv
England Joirrnal
of
Medicine 309,
8 1.
Macdonald,
1.
A.
&
Lake, D. M. (1985). An improved technique for extracting catecholamines from body fluids.
Journal
of
Neuro.vcicv7ce Methods 13,
239-248.
McGuire, E. A. H., Helderman,
J.
H., Tobiii, J.
D.,
Andres, R.
&
Bergman, R. (1976). Effects
of
arterial versus
venous sampling on glucose kinetics in man.
Journal
of
Applied
Phjsiologj~ 41,
165-173.
Moneta. G.
L.,
Taylor, D. C., Helton, W.
S.,
Mulholland. M. W.
&
Strandness,
D.
E.
Jr (1988). Duplex
ultrasound measurement of postprandial intestinal blood flow: effect of ineal composition.
Gastrornterologj~
95.
12941
301.
Muiesan,
G.,
Sorbini,
C.
A.,
Solinas,
E.,
Grassi,
V.,
Casucci,
G.
&
Pedz. E. (1968). Comparison
of
C0,-
rebreathing methods and direct Fick methods
for
determining cardiac output.
Journal
qf
Applied Pliy.siolog~~ 24,
Muller. A.
F.,
Hawkins, M., Batin, P., Evans,
S.
&
Cowley, A. J. (1992). Food in chronic heart failure:
improvement in central haemodynamics but deleterious effects on exercise tolerance.
European Heurt Jorrrnul
13.
1460~-1467.
Nakamura,
T.,
Moriyasu,
F.,
Ban, N., Nishida,
O.,
Tamada, T., Kawasaki,
T.,
Sakai, M.
&
Uchino, H. (1989).
Quantitative measurement of abdominal arterial blood flow using image-directed Doppler ultrdSOnOgrdphy:
superior mesenteric. splenic, and common hepatic arterial blood flow in normal adults.
Jouriiul
o/
Clinicul
Ultrasouncl
17,
261-268.
Norryd, C., Dencker,
H.,
Lunderquist,
A.,
Olin,
T.
&
Tylen,
U.
(1975). Superior mesenteric blood flow during
digestion in inan.
Act0 Chirurgica Scundinuvica 141,
197-202.
Potter, J.
F.,
Heseltine, D., Hartley.
G.,
Matthews, J., Macdonald,
I.
A.
&
James,
0.
F.
W.
(1989). Effects
of
meal
consumption on the postprandial blood pressure, catecholamine and insulin changes in elderly subjects.
Clinical
Science
77,
265-272.
Qamar, M.
I.
&
Read, A.
E.
(1988). Effects
of
ingestion of carbohydrate, fat, protein and water on the mesenteric
blood flow in man.
Scandinuvinn
Journul
of
Gusrroen/erologj
23,
2630.
233-235.
424 -429.
848
M.
B.
SIDERY AND
I.
A. MACDONALD
Reybrouck,
T.
&
Fagard,
R.
(1990). Assessment of cardiac output at rest and during exercise by a carbon dioxide
rebreathing method.
European Heart Journal
11,
Suppl.
1
I.
21-25.
Roberts, L.
N.
&
Breckenridge, A. M. (1986). The reproducibility of limb blood flow measurements in human
volunteers at rest and after exercise by using mercury
in
silastic strain gauge plethysmography under
standardized conditions.
Clinical Science
70,
635-638.
Rowe, J. W., Young,
J.
B., Minkaer, K. L., Stevens,
A.
L., Pallotta, J.
&
Landsberg, L. (1981). Effect of insulin
and glucose infusions
on
sympathetic nervous system activity in normal man.
Diabetes
30,
219-225.
Rowe, J. W., Young, J.
B.,
Stevens, A., Kilgore, A,, Pallotta, J.
&
Landsberg, L. (1979). Insulin increases plasma
noradrenaline in man independent of changes in blood glucose.
Clinical Research
27,
594A.
Rowell, L. B., Brengelmann,
G.
L., Blackmon, J. R., Bruce, R.
A.
&
Murray,
J.
A.
(1968). Disparities between
aortic and peripheral pulse pressures induced by upright exercise and vasomotor changes in man.
Circulation
37,
954964.
Sidery, M. B., Blackshaw, P.
E.
&
Macdonald,
I.
A.
(1994). Superior mesenteric artery blood flow and gastric
emptying in humans and the differential effects of high fat and high carbohydrate meals.
Gut
35,
186190.
Sidery, M.
B.,
Cowley, A.
J.
&
Macdonald,
I.
A.
(1993). Cardiovascular responses
to
high fat and high
carbohydrate meals in healthy elderly subjects.
Clinical Science
84,
263-270.
Sidery, M. B., Macdonald,
I.
A., Cowley, A.
J.
&
Fullwood, L.
J.
(1990). Cardiovascular responses
to
high-fat and
high-carbohydrate meals in young subjects.
American Journal
of
Physiology
261,
H143gH1436.
Torsdottir,
I.,
Alpsten, M., Anderson,
H.,
Schweizer,
T.
F.,
Tolli.
H.
&
Wiirsch,
P.
(1989). Gastric emptying and
glycemic response after ingestion of mashed bean
or
potato flakes in composite meals.
American Journal
of
Clinicnl Nutrition
50,
1415-1419.
Waaler, B.
A,,
Eriksen, M.
&
Janbu,
T.
(1990). The effect of a meal
on
cardiac output in man at rest and during
moderate exercise.
Acta Physiologica Scandinavica
140,
167-1 73.
Waaler, B. A,, Eriksen,
M.
&
Toska,
K.
(1991). The effect of meal
size
on
postprandial increase in cardiac output.
Acta Physiologica Scandinavica
142,
33-37.
Wallin, B. G., Sundloff, G., Eriksson, B.-M., Dominiak,
P.,
Grobecker,
H.
&
Lindblad, L. E. (1981). Plasma
noradrenaline correlates to sympathetic muscle nerve activity in normotensive man.
Acta Physiologica
Scandinavica
11 1,
69-73.
Whitney, R. J. (1953). Measurement of volume changes in human limbs.
JournalofPhysiology (London)
121,
1-27.
Printed
in
Great Britain
