Acta Physiol Scand 1990, 140, 167-173
The effect of a meal on cardiac output in man a t rest and during moderate exercise B. A . W A A L E R , M. E R I K S E N and T. J A N B U Department of Physiology, University of Oslo, Norway
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 Physiol Scand 140, 167-173. Received 22 February 1990, accepted 31 May 1990. ISSN 0001-6772. Department of Physiology, University of Oslo, Norway. Cardiac output at rest increased by 11-63 yo in a group of healthy individuals after the consumption of a medium-sized, mixed meal. The maximum post-prandial levels of cardiac output were reached from 10 to 30 min after termination of the meal. Cardiac output values at rest fluctuate around a mean level, and this fluctuation was considerably more marked after a meal, when changes in cardiac output from one 15-s period to another could be of the order of 1-1.5 1 min-l. Recording of flow in the superior mesenteric artery before and also after a meal was successful in two subjects in whom anatomical conditions were favourable. Flow in the artery was approximately doubled from the fasting to the post-prandial situation, an augmentation that accounted for about 50% of the concomitant increase in cardiac output. The increases in cardiac output caused by 2-min bouts of standardized, moderate and rhythmic exercise were consistently larger in the post-prandial than in the fasting situation. It thus appears that any tendency for redistribution of blood flow, for example from the gastrointestinal tract to the working muscles, during moderately intense exercise is less marked after a meal than before. Key words : cardiac output, exercise, post-prandial circulation, splanchnic circulation.
During strenuous muscular exercise cardiac output (CO) is not only markedly increased but its distribution is also changed. Considerable restriction in the blood flow to resting tissues, and notably to gastrointestinal organs, has thus been demonstrated in exercising animals (Laughlin & Armstrong 1982, Armstrong et al. 1987). We have recently found evidence for a similar redistribution of flow during rhythmic exercise of moderate intensity in man (Eriksen et ul. 1990). W e thus consistently observed that flow to the working muscles increased more than did CO. T h i s would imply that even at this moderate level of physical activity some amount of flowing blood had been diverted to the exercising muscles Correspondence : Bjarne A. Waaler, Department of Physiology, University of Oslo, Karl Johansgt. 47, N0162 Oslo 1, Norway.
from other vascular beds, and possibly from the splanchnic circulation. I n this study we caused changes in the splanchnic circulation by giving a meal to fasting test persons. With the use of an improved method of Doppler ultrasonography (Eriksen & Wallee 1990) we recorded CO continuously both at rest and throughout brief periods of exercise both before and after the meal. I n addition, in two persons we were also able to record flow in the superior mesenteric artery before and after the meal. T h e aim of the study was twofold. First, we wanted to study the effect of a meal on CO and splanchnic circulation. Second, we wanted to see if the CO response to standardized bouts of exercise was changed in the post-prandial situation, in which splanchnic flow would presumably be increased.
167
168
B . A . Waaler et al.
M A T E R I A L S AND M E T H O D S Measiuements of CO were carried out as described by Eriksen & Wallee (1990), with Doppler ultrasonograph?, using a pulsed-wave Doppler velocimeter (SD-100, Vingmed Sound A/S, Horten, Norm!-), operated at 2.0 MHz. The transducer was positioned in the suprasternal notch, and the sampling volume was positioned about 20 mm upstream of the 1-alvular orifice. Instantaneous maximum velocit!- was calculated by the instrument’s built-in maximum velocity (CLJ estimator, and fed on-line to a microcomputer (-4pricot XI, ACT Computers, Birmingham, L K ) for further processing. T h e computer calculated systolic velocity integrals for each cardiac cycle during the measurements, gated by automatic ECG QRScomplex detections. Stroke volumes were calculated by multiplying velocity integrals by valvular cross-sectional area, calculated from internal aortic orifice diameters which were obtained by sector scan imaging (CFM-700, 1-ingmed Sound ,4/S). This calculation should give reliable estimates of stroke volumes according to the thorough methodological considerations of Eriksen & h’allee (l990), since the velocity profile at the valvular level is rectangular, and since this velocity is conserved as the maximum velocity in a jet some distance upwards in the aortic root. Beat-by-beat CO was calculated bj- dividing stroke volumes by the corresponding beat durations found from QRS-complex detections. Instantaneous heart rate was also calculated from the beat durations. All these calculated values were stored in the computer for later plotting and statistical analysis. .Measurement of superior mesenteric arteryyom. Flow in the superior mesenteric artery was measured by duplex ultrasound scanning (CFZl-700, Vingmed Sound), using an annular arraj- transducer, operated at 3.0 ZlHz for imaging and at 2.5 MHz for Doppler purposes. The superior mesenteric artery was imaged from the anterior abdominal wall, at a viewing angle giving the best image quality for diameter measurements. The diameter was measured in the fasting situation and just distal to the aortic offspring. Repeated measurements gave very reproducible values, but an average of three such individual caliper measurements was used. Because the intestinal contents caused a reduction in image quality after the meal, diameter measurement was not repeated in the post-prandial situation. Doppler recordings of maximal bloodstream velocity were made at the same location as the diameter measurements but with the transducer positioned somenhat lower on the abdominal aall, in order to reduce the angle between the direction of the bloodstream and that of the sound beam. Because of respiratory movements, long-lasting and continuous
registrations could not be achieved. Two measurements were made, separated by a few minutes, each measurement containing two or three consecutive cardiac cycles. The average of the results from the two measurements was used. In the fasting situation bloodstream velocity could be recorded in this artery in most people. It became much more difficult to achieve records in the period after a meal, as gastrointestinal gas and fluid content now tended to prevent a good insonication of the artery. In two lean test persons we were, however, successful in recording superior mesenteric artery blood velocity both before and in the period after a meal. Flow was calculated by multiplying time-averaged maximum 1-elocity by cross-sectional area. For this to be correct, the velocity profile must be rectangular. From hydrodynamic theory this was expected to be the case, and was confirmed by the actual narrowbanded velocity spectra obtained. Recordings of superior mesenteric artery flowduring leg exercise could unfortunately not be obtained as slight - but unavoidable - movements in the abdominal region made it impossible then to keep track of the artery. Sribjei.ts and experimental design. T h e experiments were carried out in seven subjects, two male and five female (age 22-50 years), who were not known to have an!- cardiovascular disease. They all gave informed consent. Each subject participated on 2 or 3 separate days in a session involving a meal and several C O measurements. The subjects arrived fasting (no food after 24.00 h on the preceding day) at 09.00 h. They then rested for at least 10 min, whereafter C O was recorded and followed for 5-10 min. These and all other measurements were made &-ith the subjects in the supine position. -4 standard meal was then served, and 15-20 min was allowed for the consumption of the food. The meal was a medium-sized and mixed one, consisting of 12&150 g of bread, with butter and spread (cheese, mackerel and liver paste) and 150-200 ml of skimmed milk. The size of the meal was somewhat modified according to the body weight of the test persons. Subjects in the 60-kg range received a meal of about 2800 kJ, whereas subjects in the 80-kg range received a meal of about 3.500 kJ. Within 10 min after the end of the meal, the recording of resting C O was started again as continuous measurements throughout 1&20 min periods, interrupted by short intervals without any recording. C O was followed for about 1 h after termination of the meal. In two of the test persons (both female), flow in the superior mesenteric arterl- was also recorded before and after the meal. The other five subjects (three female, two male) carried out 2-min tests of standardized, rhythmic
Post-prandial changes an cardiac output exercise both in the fasting situation and at certain intervals after termination of the meal. One or (usually) two such tests were made in the fasting situation, with 8 min of rest between the two. Twenty and 45 min after finishing the meal the subjects repeated the standardized 2-min exercise test. The exercise scheme applied has been tried out and used before (Wallee & Wesche 1988, Eriksen et al. 1990). The subjects lay on a bench with their legs from the knee level extending beyond the bench. Their heels were supported by a table slightly lower than the bench. Rhythmic, alternating, bilateral contractions of the quadriceps muscle groups were performed by extending the slightly flexed knee, while the thigh remained resting on the bench. By this manoeuvre the heels were lifted approximately 3-5 cm above the supporting table. The tension generated in the muscles was increased by attaching weights of 7-10 kg to straps fitted around the ankles. About 30% of maximal voluntary contraction force was developed by this procedure. The subjects exercised to the sound of a metronome. Each contraction and relaxation phase lasted for 2 s. The duration of each two-legged exercise period was 2 min. Statistics. Variance analysis was carried out on the exercise-induced increments in CO as recorded before and after a meal. The program used was the commercially available BMDP3V. A mixed model was used, in which the general effect of the meal could be isolated from interindividual variations and variations between test runs. The changes in resting CO from the fasting to the post-prandial situation were evaluated by one-sided Wilcoxon pair test.
Materials and Methods). In about half the meal tests, we observed, as is illustrated in the lower part of Fig. 1, that the full and maximum postprandial increase in CO had already developed 10-15 min after termination of the meal. I n the other half of the tests the increase developed more gradually, CO reaching its maximum value after 30 min (upper part of Fig. 1). I t has previously been demonstrated that CO in supine, resting persons shows considerable and spontaneous fluctuations around a mean level (Eriksen et al. 1990). One component in these fluctuations stems from the variations in heart rate and stroke volume occurring in synchrony with respiration. These latter, respiration-related variations can be removed through a suitable filtration procedure, as has been done in the tracings of Figs. 1 and 3. T h e marked, slow and more irregular fluctuations of CO with time are then revealed, as are clearly seen in the premeal parts of Figs. 1 and 3. A dominant finding in our test subjects was that these irregular CO fluctuations around mean levels were larger after a meal. This can clearly be seen in the tracings of Fig. 1, and most markedly in
Table 1. Maximum increments observed in resting CO after intake of a standardized meal. CO levels given are obtained by averaging measurements over 2-min periods. Values are given with an accuracy of 0.05 I min-'
RESULTS A definite post-prandial CO increase was observed after all 16 meal sessions in the seven persons tested (Table 1). T h e mean difference in CO values from the fasting to the post-prandial situation was highly significant (P< 0.01). However the magnitude of the increase varied considerably from person to person and also, to some extent, from one meal test to another in the siime person. T h e observed post-prandial CO increases ranged from some 0.6 to about 3.0 1 min-l, in relative terms from 11 to 63 yo. T h e time course of the CO development after a meal also varied considerably. T h i s is illustrated in Fig. 1, which shows post-prandial development of CO in two of the test persons, in whom CO was continuously recorded throughout succeeding 5-15 min intervals. T h e development of CO after a meal was not affected by the insertion of the two 2-min exercise tests (see
169
Subjects sex, weight
CO, fasting (1 min-')
Max post-prandial increase in CO (1 min-')
T.E. M, 67 kg L.W. M, 85 kg W.N. F, 71 kg
5.20 4.95 3.85 3.80 4.75 4.50 5.00 5.00 4.50 5.80 5.85 5.50 4.20 4.30 3.25 3.65
0.65 0.55 0.70 1.05 3.00 2.50 2.35 1.20 2.60 2.20 0.95 0.60 1.10 2.10 1.00 1.10
L.F.J. F, 62 kg T.F. M, 85 kg T.J. F, 55 kg T.F. F, 55 kg
170
B.A. Wader et al.
-4 5
-3 0
-15
15
0
30
45
Time (min) Fig. 1. Development of CO in t\$o supine subjects (L.W. above and L.F.J. below) before and after a standardized meal, the duration of which is indicated by the shaded bars. T h e gaps in the recordings correspond to periods in which short exercise bouts were carried out, or to periods during which the recordings were discontinued for technical reasons. To emphasize the general trends in CO development, respiratory variations in CO have been removed from the curves by lowpass filtration with a cut-off frequency of 0.1 Hz.
T.J.
T.F.
Fig. 2. CO and flow in the superior mesenteric artery before and after a meal in subjects T.J. (one test) and T.F. (two tests). The total height of the individual bars represents CO values obtained by averaging over 2-min measurement periods. Superior mesenteric artery flow is indicated by the shaded area. T h e left bar of each pair represents values obtained in the fasting condition, and the right bar represents values recorded about 30 min after completion of the meal.
Post-prandial changes in cardiac output
171
a-
7-
1
2
3
4
5
Time (min)
1
2
3
4
5
Time (min)
Fig. 3. Beat-by-beat development of CO before, during and after exercise periods of 2 min duration in one of the subjects. The left part depicts the fasting situation, the right the period 2 k 2 5 min after a standardized meal. The curves are drawn through points representing values from each individual cardiac cycle. Exercise started at the 2-min point.
4.0-
3.5.
-
7
-
i//
3.0-
.-c
E -
2.5-
0
a 2.0-
1.5-
1.0.
Fig. 4. Summarized results from all the exercise bouts carried out in our five subjects. Each cross represents the increase in CO caused by a standardized 2-min exercise period. The lines connect the value obtained in the fasting situation (left) with the one obtained in the subsequent postprandial period (right). The individual subjects are identified along the horizontal axis. the upper part. Before a meal CO could change from one 15-s period to another by some 0.5-0.75 1 min-l, whereas such fluctuations seen after a
meal were of the order of 1-1.5 1 min-'. When mean, individual CO values are given, as in Table 1 and in Figs. 2 and 4, these values are
172
B.A. Waaler et al.
therefore arrived at by averaging flow over periods as long as 2 min. In order to obtain some direct information about how the observed post-prandial increases in CO might be related to gastrointestinal blood flow, we attempted to measure flow in the superior mesenteric artery before and after intake of the standard meal, and these attempts met with success in two lean persons. The main results from three experiments with these two persons are given in Fig. 2. Flow in the superior mesenteric artery was approximately doubled from the fasting to the post-prandial situation in all of these three tests, and the increase equalled about 50% of the concomitant, total increase in CO. The effect of a meal on the CO response to our standardized scheme of rhythmic exercise is illustrated in Fig. 3, in which C O development in one premeal and one post-meal episode of exercise is depicted. The exercise-induced increase in CO after the meal is appreciably larger than that obtained before the meal. Both CO increases developed equally rapidly, but the post-meal one rose from a higher CO resting level. Figure 4 gives the differences between premeal and post-meal exercise-induced C O increments in all 11 experimental sessions in our five test persons. The figure shows how C O invariably increased more during exercise after a meal than before. The mean of all 11 exercise-induced C O increases before a meal was 2.21 1 min-'. The mean C O increment after a meal was 0.5 1 min-' (SD 0.06 1 min-') larger than this. Analysis of variance showed this difference to be highly significant ( P < 0.001). SD for the interindividual variation in CO increment values was found to be 0.47 1 min-'. Intraindividual variations of CO increment values (between test runs) had a SD of 0.21 1 min-'.
DISCUSSION In this study a significant increase in CO at rest was invariably recorded after consumption of a medium-sized, mixed meal. However there was a striking variability in the magnitude of this post-prandial increase, which in the different subjects ranged from 0.6 to 3.0 1 min-' or from 11 to 63% above the fasting CO value (Fig. 1 and Table 1). This variability in the postprandial C O increase is in good agreement with the results from previous studies. Grollman
(1929) found that the maximum increase in CO output varied between 0.5 and 0.9 I min-' after a light meal and between 1 and 2 1 min-' after a heavy meal. Similar ranges of CO increases were found by Gladstone (1935) after a mixed meal of moderate size and by Fagan et al. (1986) after a lunch meal. It is somewhat surprising that identical medium-sized meals cause such widely different individual increases in CO. One would expect the digestive processes involved after intake of the same amount and type of food to require a fairly similar increase in splanchnic blood flow. The question can be posed whether a variable degree of flow redistribution from other tissues to the splanchnic area might take place during digestion. The timing of the post-prandial increase in C O also varied, but the increase was always fully developed about 30 min after termination of the meal. A previous study has revealed considerable beat-by-beat fluctuations in C O at rest (Eriksen et al. 1990), largely in synchrony with the respiratory movements, as has also been demonstrated by others (Cummin et al. 1986, Guz et 01. 1987). However, a considerable temporal variation in resting CO also remains after removal by a suitable filtration procedure of the respirationrelated fluctuations in stroke volume (Figs. 1 and 3). An interesting finding in the present study was that these CO fluctuations were increased after a meal, when C O could vary from one 15s period to the next by as much as 1-1.5 1 min-I. We do not know why there are such large postprandial fluctuations in CO, but they might be related to some sort of intermittence in the digestive process, e.g. to the rhythmic gastric emptying and/or to the subsequent handling of chyme. Measurements in the superior mesenteric artery (Fig. 2) revealed a marked post-prandial increase in flow, equalling 50% of the increase in total CO. This would indicate that the augmentation in C O did directly reflect an increase in the blood supply to organs involved in digestion. In a previous study (Eriksen et al. 1990) we regularly observed a distinct pattern of circulatory response to bouts of standardized, moderate exercise. Thus, in the majority of tests, C O increased less than did peripheral flow through the femoral arteries to the working
Post-prandial changes in cardiac output muscles. T h e most likely explanation for such a discrepancy would be the occurrence, even at this moderate degree of exercise, of some redistribution of flow to the working muscles from other tissues. From experiments in animals (Laughlin & Armstrong 1982, Armstrong et al. 1987) it is known that the splanchnic circulation is markedly flow-restricted during strenuous exercise. We hypothesized that the splanchnic circulation might be similarly involved in our situation with moderate exercise. W e further hypothesized that a possible redistribution of blood flow from the splanchnic area might be more evident after a meal, when splanchnic flow would be increased. T h e results from the present experiments seem to refute such a hypothesis. T h e CO increase during bouts of standardized, moderate exercise was consistently larger in the post-prandial than in the fasting situation. Results from a previous study (Wallee & Wesche 1988) indicate that neither resting nor exerciseinduced flow in the femoral artery is influenced by a meal. Thus, the difference between the increases in femoral flow and in CO during exercise must have been less marked (or even absent) in the post-prandial situation, indicating that there was now less - if any - redistribution of blood flow to working muscles from other tissues. Direct and continuous measurements of flow in the superior mesenteric artery during bouts of exercise would clearly have helped answer the question of whether or not splanchnic flow is restricted during moderate exercise in man, and whether there is a difference in this respect between the fasting and the post-prandial situation. Unfortunately this exerimental approach was not possible for technical reasons (see Materials and Methods). O u r experiments indicate that splanchnic blood flow might have higher priority after the intake of a medium-sized meal than in the fasting situation. During moderate exercise in this post-prandial situation the increased demand for blood flow to the working muscles is thus apparently met more completely by augmentation in CO- and less through any redistribution of flow from other vascular areas, such as the splanchnic one.
173
This work was supported by grants from The Norwegian Research Council for Science and the Humanities and from Anders Jahre's Foundation for the Promotion of Science. We are grateful to professor Lars Wallse for fruitful discussions and for valuable help with the statistical analyses. REFERENCES ARMSTRONG,R.B., DELP, M.D., GOLJAN,E.F. & LAUGHLIN, M.H. 1987. Distribution of blood flow in muscles of miniature swine during exercise. j ' Appl Physiol62, 1285-1298. CUMMIN, A.R.C., IYAWE,V.I., MEHTA, N. & SAUNDERS, K.B. 1986. Ventilation and cardiac output during the onset of exercise, and during voluntary hyperventilation, in humans. 3 Physiol (Lond) 370, 567-583. ERIKSEN,M. & WALL0E, L. 1990. An improved method for cardiac output determination in man using Doppler technique. Med Biol Eng Comput (in press). ERIKSEN, M., WAALER, B.A., WALLBE, L. & WESCHE, J. 1990. Dynamics and dimensions of cardiac output changes in humans at the onset and at the end of moderate rhythmic exercise.3Physiol (Lond) 426, 423437. P.R., GOURLEY, L.A., LEE, FAGAN,T.C., SAWYER, J.T. & GAFFNEY, T.E. 1986. Postprandial alterations in hemodynamics and blood pressure in normal subjects. Am 3 Cardiol 58, 636-641. GLADSTONE, S.A. 1935. Cardiac output and related functions under basal and postprandial conditions. Arch Intern Med 55, 533-546. GROLLMAN, A. 1929. Physiological variations in the cardiac output of man. 111. The effect of the ingestion of food on the cardiac output, pulse rate, blood pressure, and oxygen consumption in man. Am 3 Physiol 89, 366-370. Guz, A. INNES, J.A. & MURPHY,K. 1987. Respiratory modulation of left ventricular stroke volume in man measured using pulsed Doppler ultrasound. 3 Physiol (Lond) 393, 499-5 12. M.H. & ARMSTRONG, R.B. 1982. Muscular LAUGHLIN, blood flow distribution patterns as a function of running speed in rats. Am 3 Physiol 243, H296H306. WALLBE,I. & WESCHE,J. 1988. Time course and magnitude of blood flow changes in the human quadriceps muscles during and following rhythmic exercise. 3 Physiol (Lond) 405, 257-273.
The effect of a meal on cardiac output in man a t rest and during moderate exercise B. A . W A A L E R , M. E R I K S E N and T. J A N B U Department of Physiology, University of Oslo, Norway
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 Physiol Scand 140, 167-173. Received 22 February 1990, accepted 31 May 1990. ISSN 0001-6772. Department of Physiology, University of Oslo, Norway. Cardiac output at rest increased by 11-63 yo in a group of healthy individuals after the consumption of a medium-sized, mixed meal. The maximum post-prandial levels of cardiac output were reached from 10 to 30 min after termination of the meal. Cardiac output values at rest fluctuate around a mean level, and this fluctuation was considerably more marked after a meal, when changes in cardiac output from one 15-s period to another could be of the order of 1-1.5 1 min-l. Recording of flow in the superior mesenteric artery before and also after a meal was successful in two subjects in whom anatomical conditions were favourable. Flow in the artery was approximately doubled from the fasting to the post-prandial situation, an augmentation that accounted for about 50% of the concomitant increase in cardiac output. The increases in cardiac output caused by 2-min bouts of standardized, moderate and rhythmic exercise were consistently larger in the post-prandial than in the fasting situation. It thus appears that any tendency for redistribution of blood flow, for example from the gastrointestinal tract to the working muscles, during moderately intense exercise is less marked after a meal than before. Key words : cardiac output, exercise, post-prandial circulation, splanchnic circulation.
During strenuous muscular exercise cardiac output (CO) is not only markedly increased but its distribution is also changed. Considerable restriction in the blood flow to resting tissues, and notably to gastrointestinal organs, has thus been demonstrated in exercising animals (Laughlin & Armstrong 1982, Armstrong et al. 1987). We have recently found evidence for a similar redistribution of flow during rhythmic exercise of moderate intensity in man (Eriksen et ul. 1990). W e thus consistently observed that flow to the working muscles increased more than did CO. T h i s would imply that even at this moderate level of physical activity some amount of flowing blood had been diverted to the exercising muscles Correspondence : Bjarne A. Waaler, Department of Physiology, University of Oslo, Karl Johansgt. 47, N0162 Oslo 1, Norway.
from other vascular beds, and possibly from the splanchnic circulation. I n this study we caused changes in the splanchnic circulation by giving a meal to fasting test persons. With the use of an improved method of Doppler ultrasonography (Eriksen & Wallee 1990) we recorded CO continuously both at rest and throughout brief periods of exercise both before and after the meal. I n addition, in two persons we were also able to record flow in the superior mesenteric artery before and after the meal. T h e aim of the study was twofold. First, we wanted to study the effect of a meal on CO and splanchnic circulation. Second, we wanted to see if the CO response to standardized bouts of exercise was changed in the post-prandial situation, in which splanchnic flow would presumably be increased.
167
168
B . A . Waaler et al.
M A T E R I A L S AND M E T H O D S Measiuements of CO were carried out as described by Eriksen & Wallee (1990), with Doppler ultrasonograph?, using a pulsed-wave Doppler velocimeter (SD-100, Vingmed Sound A/S, Horten, Norm!-), operated at 2.0 MHz. The transducer was positioned in the suprasternal notch, and the sampling volume was positioned about 20 mm upstream of the 1-alvular orifice. Instantaneous maximum velocit!- was calculated by the instrument’s built-in maximum velocity (CLJ estimator, and fed on-line to a microcomputer (-4pricot XI, ACT Computers, Birmingham, L K ) for further processing. T h e computer calculated systolic velocity integrals for each cardiac cycle during the measurements, gated by automatic ECG QRScomplex detections. Stroke volumes were calculated by multiplying velocity integrals by valvular cross-sectional area, calculated from internal aortic orifice diameters which were obtained by sector scan imaging (CFM-700, 1-ingmed Sound ,4/S). This calculation should give reliable estimates of stroke volumes according to the thorough methodological considerations of Eriksen & h’allee (l990), since the velocity profile at the valvular level is rectangular, and since this velocity is conserved as the maximum velocity in a jet some distance upwards in the aortic root. Beat-by-beat CO was calculated bj- dividing stroke volumes by the corresponding beat durations found from QRS-complex detections. Instantaneous heart rate was also calculated from the beat durations. All these calculated values were stored in the computer for later plotting and statistical analysis. .Measurement of superior mesenteric arteryyom. Flow in the superior mesenteric artery was measured by duplex ultrasound scanning (CFZl-700, Vingmed Sound), using an annular arraj- transducer, operated at 3.0 ZlHz for imaging and at 2.5 MHz for Doppler purposes. The superior mesenteric artery was imaged from the anterior abdominal wall, at a viewing angle giving the best image quality for diameter measurements. The diameter was measured in the fasting situation and just distal to the aortic offspring. Repeated measurements gave very reproducible values, but an average of three such individual caliper measurements was used. Because the intestinal contents caused a reduction in image quality after the meal, diameter measurement was not repeated in the post-prandial situation. Doppler recordings of maximal bloodstream velocity were made at the same location as the diameter measurements but with the transducer positioned somenhat lower on the abdominal aall, in order to reduce the angle between the direction of the bloodstream and that of the sound beam. Because of respiratory movements, long-lasting and continuous
registrations could not be achieved. Two measurements were made, separated by a few minutes, each measurement containing two or three consecutive cardiac cycles. The average of the results from the two measurements was used. In the fasting situation bloodstream velocity could be recorded in this artery in most people. It became much more difficult to achieve records in the period after a meal, as gastrointestinal gas and fluid content now tended to prevent a good insonication of the artery. In two lean test persons we were, however, successful in recording superior mesenteric artery blood velocity both before and in the period after a meal. Flow was calculated by multiplying time-averaged maximum 1-elocity by cross-sectional area. For this to be correct, the velocity profile must be rectangular. From hydrodynamic theory this was expected to be the case, and was confirmed by the actual narrowbanded velocity spectra obtained. Recordings of superior mesenteric artery flowduring leg exercise could unfortunately not be obtained as slight - but unavoidable - movements in the abdominal region made it impossible then to keep track of the artery. Sribjei.ts and experimental design. T h e experiments were carried out in seven subjects, two male and five female (age 22-50 years), who were not known to have an!- cardiovascular disease. They all gave informed consent. Each subject participated on 2 or 3 separate days in a session involving a meal and several C O measurements. The subjects arrived fasting (no food after 24.00 h on the preceding day) at 09.00 h. They then rested for at least 10 min, whereafter C O was recorded and followed for 5-10 min. These and all other measurements were made &-ith the subjects in the supine position. -4 standard meal was then served, and 15-20 min was allowed for the consumption of the food. The meal was a medium-sized and mixed one, consisting of 12&150 g of bread, with butter and spread (cheese, mackerel and liver paste) and 150-200 ml of skimmed milk. The size of the meal was somewhat modified according to the body weight of the test persons. Subjects in the 60-kg range received a meal of about 2800 kJ, whereas subjects in the 80-kg range received a meal of about 3.500 kJ. Within 10 min after the end of the meal, the recording of resting C O was started again as continuous measurements throughout 1&20 min periods, interrupted by short intervals without any recording. C O was followed for about 1 h after termination of the meal. In two of the test persons (both female), flow in the superior mesenteric arterl- was also recorded before and after the meal. The other five subjects (three female, two male) carried out 2-min tests of standardized, rhythmic
Post-prandial changes an cardiac output exercise both in the fasting situation and at certain intervals after termination of the meal. One or (usually) two such tests were made in the fasting situation, with 8 min of rest between the two. Twenty and 45 min after finishing the meal the subjects repeated the standardized 2-min exercise test. The exercise scheme applied has been tried out and used before (Wallee & Wesche 1988, Eriksen et al. 1990). The subjects lay on a bench with their legs from the knee level extending beyond the bench. Their heels were supported by a table slightly lower than the bench. Rhythmic, alternating, bilateral contractions of the quadriceps muscle groups were performed by extending the slightly flexed knee, while the thigh remained resting on the bench. By this manoeuvre the heels were lifted approximately 3-5 cm above the supporting table. The tension generated in the muscles was increased by attaching weights of 7-10 kg to straps fitted around the ankles. About 30% of maximal voluntary contraction force was developed by this procedure. The subjects exercised to the sound of a metronome. Each contraction and relaxation phase lasted for 2 s. The duration of each two-legged exercise period was 2 min. Statistics. Variance analysis was carried out on the exercise-induced increments in CO as recorded before and after a meal. The program used was the commercially available BMDP3V. A mixed model was used, in which the general effect of the meal could be isolated from interindividual variations and variations between test runs. The changes in resting CO from the fasting to the post-prandial situation were evaluated by one-sided Wilcoxon pair test.
Materials and Methods). In about half the meal tests, we observed, as is illustrated in the lower part of Fig. 1, that the full and maximum postprandial increase in CO had already developed 10-15 min after termination of the meal. I n the other half of the tests the increase developed more gradually, CO reaching its maximum value after 30 min (upper part of Fig. 1). I t has previously been demonstrated that CO in supine, resting persons shows considerable and spontaneous fluctuations around a mean level (Eriksen et al. 1990). One component in these fluctuations stems from the variations in heart rate and stroke volume occurring in synchrony with respiration. These latter, respiration-related variations can be removed through a suitable filtration procedure, as has been done in the tracings of Figs. 1 and 3. T h e marked, slow and more irregular fluctuations of CO with time are then revealed, as are clearly seen in the premeal parts of Figs. 1 and 3. A dominant finding in our test subjects was that these irregular CO fluctuations around mean levels were larger after a meal. This can clearly be seen in the tracings of Fig. 1, and most markedly in
Table 1. Maximum increments observed in resting CO after intake of a standardized meal. CO levels given are obtained by averaging measurements over 2-min periods. Values are given with an accuracy of 0.05 I min-'
RESULTS A definite post-prandial CO increase was observed after all 16 meal sessions in the seven persons tested (Table 1). T h e mean difference in CO values from the fasting to the post-prandial situation was highly significant (P< 0.01). However the magnitude of the increase varied considerably from person to person and also, to some extent, from one meal test to another in the siime person. T h e observed post-prandial CO increases ranged from some 0.6 to about 3.0 1 min-l, in relative terms from 11 to 63 yo. T h e time course of the CO development after a meal also varied considerably. T h i s is illustrated in Fig. 1, which shows post-prandial development of CO in two of the test persons, in whom CO was continuously recorded throughout succeeding 5-15 min intervals. T h e development of CO after a meal was not affected by the insertion of the two 2-min exercise tests (see
169
Subjects sex, weight
CO, fasting (1 min-')
Max post-prandial increase in CO (1 min-')
T.E. M, 67 kg L.W. M, 85 kg W.N. F, 71 kg
5.20 4.95 3.85 3.80 4.75 4.50 5.00 5.00 4.50 5.80 5.85 5.50 4.20 4.30 3.25 3.65
0.65 0.55 0.70 1.05 3.00 2.50 2.35 1.20 2.60 2.20 0.95 0.60 1.10 2.10 1.00 1.10
L.F.J. F, 62 kg T.F. M, 85 kg T.J. F, 55 kg T.F. F, 55 kg
170
B.A. Wader et al.
-4 5
-3 0
-15
15
0
30
45
Time (min) Fig. 1. Development of CO in t\$o supine subjects (L.W. above and L.F.J. below) before and after a standardized meal, the duration of which is indicated by the shaded bars. T h e gaps in the recordings correspond to periods in which short exercise bouts were carried out, or to periods during which the recordings were discontinued for technical reasons. To emphasize the general trends in CO development, respiratory variations in CO have been removed from the curves by lowpass filtration with a cut-off frequency of 0.1 Hz.
T.J.
T.F.
Fig. 2. CO and flow in the superior mesenteric artery before and after a meal in subjects T.J. (one test) and T.F. (two tests). The total height of the individual bars represents CO values obtained by averaging over 2-min measurement periods. Superior mesenteric artery flow is indicated by the shaded area. T h e left bar of each pair represents values obtained in the fasting condition, and the right bar represents values recorded about 30 min after completion of the meal.
Post-prandial changes in cardiac output
171
a-
7-
1
2
3
4
5
Time (min)
1
2
3
4
5
Time (min)
Fig. 3. Beat-by-beat development of CO before, during and after exercise periods of 2 min duration in one of the subjects. The left part depicts the fasting situation, the right the period 2 k 2 5 min after a standardized meal. The curves are drawn through points representing values from each individual cardiac cycle. Exercise started at the 2-min point.
4.0-
3.5.
-
7
-
i//
3.0-
.-c
E -
2.5-
0
a 2.0-
1.5-
1.0.
Fig. 4. Summarized results from all the exercise bouts carried out in our five subjects. Each cross represents the increase in CO caused by a standardized 2-min exercise period. The lines connect the value obtained in the fasting situation (left) with the one obtained in the subsequent postprandial period (right). The individual subjects are identified along the horizontal axis. the upper part. Before a meal CO could change from one 15-s period to another by some 0.5-0.75 1 min-l, whereas such fluctuations seen after a
meal were of the order of 1-1.5 1 min-'. When mean, individual CO values are given, as in Table 1 and in Figs. 2 and 4, these values are
172
B.A. Waaler et al.
therefore arrived at by averaging flow over periods as long as 2 min. In order to obtain some direct information about how the observed post-prandial increases in CO might be related to gastrointestinal blood flow, we attempted to measure flow in the superior mesenteric artery before and after intake of the standard meal, and these attempts met with success in two lean persons. The main results from three experiments with these two persons are given in Fig. 2. Flow in the superior mesenteric artery was approximately doubled from the fasting to the post-prandial situation in all of these three tests, and the increase equalled about 50% of the concomitant, total increase in CO. The effect of a meal on the CO response to our standardized scheme of rhythmic exercise is illustrated in Fig. 3, in which C O development in one premeal and one post-meal episode of exercise is depicted. The exercise-induced increase in CO after the meal is appreciably larger than that obtained before the meal. Both CO increases developed equally rapidly, but the post-meal one rose from a higher CO resting level. Figure 4 gives the differences between premeal and post-meal exercise-induced C O increments in all 11 experimental sessions in our five test persons. The figure shows how C O invariably increased more during exercise after a meal than before. The mean of all 11 exercise-induced C O increases before a meal was 2.21 1 min-'. The mean C O increment after a meal was 0.5 1 min-' (SD 0.06 1 min-') larger than this. Analysis of variance showed this difference to be highly significant ( P < 0.001). SD for the interindividual variation in CO increment values was found to be 0.47 1 min-'. Intraindividual variations of CO increment values (between test runs) had a SD of 0.21 1 min-'.
DISCUSSION In this study a significant increase in CO at rest was invariably recorded after consumption of a medium-sized, mixed meal. However there was a striking variability in the magnitude of this post-prandial increase, which in the different subjects ranged from 0.6 to 3.0 1 min-' or from 11 to 63% above the fasting CO value (Fig. 1 and Table 1). This variability in the postprandial C O increase is in good agreement with the results from previous studies. Grollman
(1929) found that the maximum increase in CO output varied between 0.5 and 0.9 I min-' after a light meal and between 1 and 2 1 min-' after a heavy meal. Similar ranges of CO increases were found by Gladstone (1935) after a mixed meal of moderate size and by Fagan et al. (1986) after a lunch meal. It is somewhat surprising that identical medium-sized meals cause such widely different individual increases in CO. One would expect the digestive processes involved after intake of the same amount and type of food to require a fairly similar increase in splanchnic blood flow. The question can be posed whether a variable degree of flow redistribution from other tissues to the splanchnic area might take place during digestion. The timing of the post-prandial increase in C O also varied, but the increase was always fully developed about 30 min after termination of the meal. A previous study has revealed considerable beat-by-beat fluctuations in C O at rest (Eriksen et al. 1990), largely in synchrony with the respiratory movements, as has also been demonstrated by others (Cummin et al. 1986, Guz et 01. 1987). However, a considerable temporal variation in resting CO also remains after removal by a suitable filtration procedure of the respirationrelated fluctuations in stroke volume (Figs. 1 and 3). An interesting finding in the present study was that these CO fluctuations were increased after a meal, when C O could vary from one 15s period to the next by as much as 1-1.5 1 min-I. We do not know why there are such large postprandial fluctuations in CO, but they might be related to some sort of intermittence in the digestive process, e.g. to the rhythmic gastric emptying and/or to the subsequent handling of chyme. Measurements in the superior mesenteric artery (Fig. 2) revealed a marked post-prandial increase in flow, equalling 50% of the increase in total CO. This would indicate that the augmentation in C O did directly reflect an increase in the blood supply to organs involved in digestion. In a previous study (Eriksen et al. 1990) we regularly observed a distinct pattern of circulatory response to bouts of standardized, moderate exercise. Thus, in the majority of tests, C O increased less than did peripheral flow through the femoral arteries to the working
Post-prandial changes in cardiac output muscles. T h e most likely explanation for such a discrepancy would be the occurrence, even at this moderate degree of exercise, of some redistribution of flow to the working muscles from other tissues. From experiments in animals (Laughlin & Armstrong 1982, Armstrong et al. 1987) it is known that the splanchnic circulation is markedly flow-restricted during strenuous exercise. We hypothesized that the splanchnic circulation might be similarly involved in our situation with moderate exercise. W e further hypothesized that a possible redistribution of blood flow from the splanchnic area might be more evident after a meal, when splanchnic flow would be increased. T h e results from the present experiments seem to refute such a hypothesis. T h e CO increase during bouts of standardized, moderate exercise was consistently larger in the post-prandial than in the fasting situation. Results from a previous study (Wallee & Wesche 1988) indicate that neither resting nor exerciseinduced flow in the femoral artery is influenced by a meal. Thus, the difference between the increases in femoral flow and in CO during exercise must have been less marked (or even absent) in the post-prandial situation, indicating that there was now less - if any - redistribution of blood flow to working muscles from other tissues. Direct and continuous measurements of flow in the superior mesenteric artery during bouts of exercise would clearly have helped answer the question of whether or not splanchnic flow is restricted during moderate exercise in man, and whether there is a difference in this respect between the fasting and the post-prandial situation. Unfortunately this exerimental approach was not possible for technical reasons (see Materials and Methods). O u r experiments indicate that splanchnic blood flow might have higher priority after the intake of a medium-sized meal than in the fasting situation. During moderate exercise in this post-prandial situation the increased demand for blood flow to the working muscles is thus apparently met more completely by augmentation in CO- and less through any redistribution of flow from other vascular areas, such as the splanchnic one.
173
This work was supported by grants from The Norwegian Research Council for Science and the Humanities and from Anders Jahre's Foundation for the Promotion of Science. We are grateful to professor Lars Wallse for fruitful discussions and for valuable help with the statistical analyses. REFERENCES ARMSTRONG,R.B., DELP, M.D., GOLJAN,E.F. & LAUGHLIN, M.H. 1987. Distribution of blood flow in muscles of miniature swine during exercise. j ' Appl Physiol62, 1285-1298. CUMMIN, A.R.C., IYAWE,V.I., MEHTA, N. & SAUNDERS, K.B. 1986. Ventilation and cardiac output during the onset of exercise, and during voluntary hyperventilation, in humans. 3 Physiol (Lond) 370, 567-583. ERIKSEN,M. & WALL0E, L. 1990. An improved method for cardiac output determination in man using Doppler technique. Med Biol Eng Comput (in press). ERIKSEN, M., WAALER, B.A., WALLBE, L. & WESCHE, J. 1990. Dynamics and dimensions of cardiac output changes in humans at the onset and at the end of moderate rhythmic exercise.3Physiol (Lond) 426, 423437. P.R., GOURLEY, L.A., LEE, FAGAN,T.C., SAWYER, J.T. & GAFFNEY, T.E. 1986. Postprandial alterations in hemodynamics and blood pressure in normal subjects. Am 3 Cardiol 58, 636-641. GLADSTONE, S.A. 1935. Cardiac output and related functions under basal and postprandial conditions. Arch Intern Med 55, 533-546. GROLLMAN, A. 1929. Physiological variations in the cardiac output of man. 111. The effect of the ingestion of food on the cardiac output, pulse rate, blood pressure, and oxygen consumption in man. Am 3 Physiol 89, 366-370. Guz, A. INNES, J.A. & MURPHY,K. 1987. Respiratory modulation of left ventricular stroke volume in man measured using pulsed Doppler ultrasound. 3 Physiol (Lond) 393, 499-5 12. M.H. & ARMSTRONG, R.B. 1982. Muscular LAUGHLIN, blood flow distribution patterns as a function of running speed in rats. Am 3 Physiol 243, H296H306. WALLBE,I. & WESCHE,J. 1988. Time course and magnitude of blood flow changes in the human quadriceps muscles during and following rhythmic exercise. 3 Physiol (Lond) 405, 257-273.
