Caffeine is the most widespread drug, at least in United States (1,2). It is consumed in coffee, tea, cola drinks, or chocolate in many countries as one of stimulators of higher nervous activity, although it influences many other body tissues. The effect of caffeine as well as its basic metabolite, paraxanthine, is believed to be associated with unselective inhibition of adenosine receptors (2,3).
Despite numerous studies of long-term caffeine use, information regarding changes in cardiac contractile function is limited. It has been shown that daily consumption of caffeine by humans is accompanied by a small increase in arterial pressure as well as by increased blood concentrations of catecholamines and fatty acids (4). However, prolonged caffeine consumption reduces the orthostatic pressor response and the increase in plasma catecholamine concentrations (5). The duration of the QRS complex is moderately increased (6), suggesting a delay of myocardial conductivity and may (with caffeine overdosage) cause cardiac arrhythmia.
In experimental cardiology, caffeine in high concentrations (10-20 mM) has been used for a long time in studies in isolated cardiomyocytes to cause rapid release of Ca2+ from sarcoplasmic reticulum. In lower concentrations, caffeine-induced increased Ca2+ leakage reduces the amplitude of Ca2+ transients in myocardial cells (7,8). This property of caffeine explains its negative inotropic effect, which is more pronounced in the isolated hearts of rats (8).
Cardiac performance with prolonged caffeine intake has been explored insufficiently. The only study we have found concerns long-term caffeine consumption during pregnancy and lactation (9). It therefore is of interest to investigate whether prolonged caffeine consumption may affect cardiac performance, especially in rats in which high caffeine dosage may be deleterious to cardiac contractile function. In the given study we have investigated the alterations of cardiac contractile function of adult rats both at initial and sustained periods of caffeine administration. The pump function of the isolated heart at functional loads and myocardial content of high-energy phosphates have been studied.
MATERIALS AND METHODS
In accordance with the rules of the Helsinki agreement, we studied male Wistar rats, 300-350 g, using both isolated hearts and in vivo methods. In the latter, a catheter (diameter 0.9 mm) was inserted into the left ventricle (LV) through the left carotid artery 2 days before the experiment, and its end was fixed at the neck. Experiments in conscious animals were started by connecting the catheter end to a Gould Statham P23Gb electromanometer (Gould Inc., Oxnard, CA, U.S.A.) and Gould 2600 recorder (Gould Inc., Cleveland, OH, U.S.A.). The LV pressure and its first derivative were recorded before and during caffeine infusion (10 mg/kg, i.v.).
In long-term experiments, caffeine (Sigma) was given by two procedures. The first group of rats was injected with caffeine, 20 mg/kg intraperitoneally, twice a day for 7 days, and the experiment was carried out the day after the last injection. The second group of rats consumed caffeine as 0.1% solution added to drinking water (10) for 8-9 weeks. At an average consumption of 50 ml water per day, each animal consumed approximately 150 mg/kg of caffeine daily.
Rats were injected with heparin (50 U/kg) and urethane (1.8 mg/kg) before heart isolation. The left atrium and aorta were cannulated, and isolated hearts were first perfused retrogradely through the aorta with standard Krebs solution containing 11 mM glucose and oxygenated with 95% O2 + 5% CO2 at 37°C (pH, 7.3-7.4). After 10 min, antegrade perfusion was started, with filling and resistance pressures set at 15 and 80 cm water, respectively.
A steel needle was inserted into the left ventricular (LV) cavity through its wall. The LV and aortic pressures as well as the first derivative of LV pressure were continuously monitored with Gould Statham P23Gb electromanometers and Gould 2600 recorder. Aortic outflow was measured with a Carolina Medical Electronics 501D flowmeter (King, NC, U.S.A.). The pump function of the heart was estimated by cardiac output that was calculated as a sum of the aortic output and the coronary flow, measured by collecting the fluid leaving the heart through the coronary sinus per minute. Cardiac volume-pressure work was calculated as a product of the cardiac output and mean pressure in the aortic chamber, representing the ejection pressure. The double product was calculated by multiplying LV developed pressure by heart rate. The index of diastolic stiffness was calculated as the ratio of the difference between LV end-diastolic and minimal diastolic pressure to LV filling volume.
After a 30-min stabilization period, three functional loads were consecutively applied. The volume loading was applied by a stepwise increase in filling pressure (5-25 cm water) at standard aortic pressure, 80 cm water. The resistance load was applied by a stepwise increase in aortic pressure from 60 to 110 cm water with further clamping of aortic outflow, the filling pressure being at an almost optimal level of 15 cm water. Frequency loading was accomplished by increasing the rate of electrical stimulation of the right atrium from 3.5 Hz to that rate at which a marked decrease in the cardiac output occurs, the filling and aortic pressures being at standard levels, 15 and 80 cm water, respectively. The stimulation was carried out using rectangular pulses 5-ms duration and amplitude equal to 1.5 thresholds (usually 1-2 V). Functional measurements were made at each step, which lasted 3-5 min. Intervals between loads were sufficient for the complete functional recovery. A cardiac distensibility curve was obtained at volume loading by determining the relationship between LV filling volume, which is equal to stroke volume, and LV end-diastolic pressure.
At the end of the experiment, after 10 min of steady work of the heart in its own rhythm and at standard load conditions, hearts were frozen in liquid nitrogen by Wollenberger clamps to determine myocardial content of adenine nucleotides, creatine phosphate, and creatine by conventional enzymatic methods. Data are presented as mean ± SEM. Student's t test (p < 0.05) was used for statistical significance.
Short-term effect of caffeine
In few preliminary experiments in conscious rats, the critical dose of caffeine that animals could tolerate at intravenous injection has been determined. In one of three experiments, 20 mg/kg caused death due to ventricular tachycardia. Hence 10 mg/kg was chosen as effective and introduced into three rats. In each case, this dose had immediate typical effects on cardiac activity. A reduction of LV systolic pressure (SP) and its first derivatives and an increase in heart rate were observed (Fig. 1). The relative alterations of inotropic and chronotropic parameters were roughly similar. In addition, the minimal LV diastolic pressure increased by 3-4 mg. These effects lasted for about 1 min, and later the parameters were returned to the initial level.
Long-term caffeine application
Prolonged caffeine application within 1 or 8-9 weeks had no essential influence on heart or body weight as compared with control groups. The ratio of heart weight/body weight after 1-week treatment was 4.08 ± 0.14 mg/kg, and in the control group, 4.34 ± 0.22 mg/kg; after 8-9 weeks of caffeine consumption, the values were 4.19 ± 0.17 and 4.56 ± 0.2 mg/kg, respectively (p > 0.05).
Volume loading of the isolated heart was accompanied in all experiments by increases in aortic and cardiac output. The latter in control experiments increased from 39 ± 3 ml/min at low filling pressure, 5 cm water, to 87 ± 9 ml/min at the highest filling pressure, 25 cm water. In both caffeine-treated groups, the curves relating these parameters to the filling pressure were very close to their respective controls. Functional parameters of the heart at maximal filling pressure also were very similar in both caffeine-treated groups and their respective controls (Table 1). However, cardiac and aortic outputs, which characterize cardiac pump function, were higher in the second control group as compared with the first one by 27-32% at moderate filling pressures. This was probably due to better LV distensibility in the second group (Fig. 2), allowing higher pump function to be maintained despite lower LVSP and its first derivative by 14 and 39%, respectively.
Resistance load in all experiments was combined with an increase in LVSP as well as in maximal values of +dP/dt and −dP/dt. Cardiac output was reduced insignificantly at a moderate increase of resistance, and decreased only on complete clamping of aortic outflow, when all inflow to the heart was ejected into the coronary vessels. Accordingly, the cardiac work index reached the maximum at medium values of resistance and was moderately reduced at the maximal resistance (Fig. 3). At the latter, hearts of rats treated with caffeine for 1 week developed lower maximal LVSP and aortic pressure by 14 and 12%, respectively, than the hearts from corresponding control groups (Table 2). In the group of rats consuming caffeine for 8-9 weeks, the parameters of pump function did not differ from the control, but the maximal double product was significantly higher (by 23%) than control (Table 2).
The increase in the double product at increasing resistance is almost linearly related to an increase in coronary flow (Fig. 4). A shift of this relation to the left was observed in the second caffeine-consuming group as compared with control.
The effect of electrical stimulation of the isolated heart has been studied in rats consuming caffeine for up to 8-9 weeks. An increased frequency of stimulation caused a decrease in the parameters of contractile function. Hearts of animals consuming caffeine demonstrated increased ability to carry out pump function at higher frequency. All hearts of controls (n = 8) successfully functioned at frequency 4-5 Hz, whereas at 6 Hz, pump function was maintained in six experiments. All hearts in the caffeine-consuming group (n = 8) successfully functioned at a frequency of 6 Hz, and the mean level of their pump function exceeded control by more than 60% at this rate (Fig. 5). Moreover, pump function was maintained in five experiments in this group at a rate of 7-7.5 Hz. In the majority of experiments, the decrease in aortic output was preceded by an increase in LV minimal diastolic pressure.
High-energy phosphate content
Measurements of high-energy phosphates were carried out at the end of the experiments, when some differences in functional parameters between groups were detected. The hearts in the caffeine-consuming group (8-9 weeks) differed from the control group in their lower heart rate (4.1 ± 0.21 vs. 4.8 ± 0.22 Hz; p < 0.05), higher stroke volume (0.33 ± 0.02 vs. 0.14 ± 0.04 ml; p < 0.01), and average aortic pressure (73 ± 1.5 vs. 65 ± 1.3 mm Hg; p <0.001).
Mean values of the myocardial content of adenine nucleotides and creatine phosphate were approximately identical in both groups (Table 3). The cross-correlation analysis between functional and metabolic parameters shows (Table 4) that the functional parameters are more closely correlated with the content of ATP than of CrP. The best correlation has been detected with the total content of adenine nucleotides (correlation coefficients >0.7), whereas ATP content and the sum of ATP + CrP demonstrated a slightly lesser correlation. Among functional parameters studied, the weakest correlation with energetic parameters was noted for coronary flow rate.
The results show that prolonged caffeine treatment of rats alters cardiac contractile function. A decrease in maximal LV developed pressure after 1-week administration of caffeine corresponds to its short-term action, and is believed to be related to its ability to deplete Ca2+ stores in sarcoplasmic reticulum and thus to reduce a Ca2+ fraction liberated during regular excitation. A reduced force of regular contractions is partially compensated by a lengthening of time of extracellular Ca2+ entry through slow channels (11,12).
After 8-9 weeks of caffeine consumption, opposite functional changes have been observed: a higher maximal level of double product and increased pump function at high rates of electrostimulation. These results, combined with preserved myocardial levels of high-energy phosphates, are new and were unexpected. Some other observations suggest that cardiac contractile function during prolonged exposure to cardioactive drugs may be altered in a direction opposite to the short-term effect. Like caffeine, doxorubicin (Adriamycin) also liberates Ca2+ ions from SR (13) and has an acute negative inotropic effect on the rat heart; however, after a 2-week period of drug washout, the maximal work of the isolated heart increases (14). Function of the isolated heart is reduced after prolonged application of high doses of catecholamines (15) and, in contrast, is increased after prolonged administration of verapamil or propranolol (16).
These results suggest that cardiac muscle successfully adapts to withstand the negative inotropic action of some cardioactive drugs. Although the precise mechanism in each case may be different, there is little doubt that it involves Ca2+ transport in cardiomyocytes. With prolonged caffeine application, force development may be facilitated because of concomitant lengthening of Ca2+ entry through slow channels (11,12) as well as of enhanced Na+/Ca2+ exchange. This suggestion is supported by the findings in transgenic mice, in which the expression of a protein ensuring Na+/Ca2+ exchange has been induced and results in accelerated Ca2+ transport to and from SR (17). Our data showing a greater cardiac relaxation at the maximal work and increased maximal intensity of the cardiac contractile function (Table 2, Fig. 5) are in accordance with this suggestion.
In addition, close connections between ryanodine receptors with which caffeine combines and adjacent mitochondria (18) suggest that some alterations in energy production may occur. Slight although insignificant increases in the myocardial ATP, total adenine nucleotides, CrP, creatine, as well as ATP/ADP ratio have been detected (Table 3). These alterations may be of importance at maximal functional loads in view of the close correlation between total adenine nucleotide content and LV systolic pressure, +dP/dt, double product, and minimal diastolic pressure (Table 4). In addition, an increase in catecholamine concentration in the heart caused by high doses of caffeine (19) may support both relaxation and contraction and thus contribute to increased cardiac output at high rate. Two factors, slightly increased myocardial energy stores and noradrenaline concentration, may explain the finding that hearts from caffeine-treated rats developed a higher intensity of function at a given coronary flow (20).
The results obtained in this and other experimental studies are not necessarily related to other species for at least two reasons. First, the release of Ca2+ through ryanodine receptors of SR is essentially important for myofibril activation in cardiomyocytes of rats but not other mammals (7). For example, isolated guinea pig hearts respond to 5-8 mM caffeine, whereas rat hearts respond to 0.2-0.4 mM(8). Second, the daily caffeine dose consumed by rats in drinking water (up to 150 mg/kg) exceeds by more than 10 times the level of caffeine consumption by humans, 10-12 mg/kg (4,21) and therapeutic caffeine plasma concentrations in humans, 0.05-0.1 mM,(21) are still lower than effective caffeine concentration in the isolated rat heart. It may be that caffeine affects cardiac performance in humans indirectly.
In humans, caffeine increases the rate of LV fiber shortening (22), heart rate, and systolic arterial pressure (4,23). These alterations may be related to an increased catecholamine level in plasma (4), heart, and brain, and acceleration of their exchange (19). Taking into account the ability of caffeine to interfere with the action of adenosine (2,3,21), a natural catecholamine antagonist, one may expect reenforcement of cate-cholamine action in these conditions. Obviously some acclimatization to short-term caffeine action develops, resulting in the disappearance of the usual short-term caffeine-induced responses of increased aortic pressure, heart rate, and blood glucose level (4,5). But the tolerance is not complete, as revealed by a persistent increase in plasma levels of norepinephrine (4), lactate, and triglycerides (24) with repeated caffeine consumption.
If extremely high doses of caffeine applied in this study to such sensitive species as the rat do not disturb cardiac contractile function and even result in increased maximal functional capacity of the heart, then one may suggest that regular consumption up to 10 cups of coffee would not substantially affect the contractile performance of the human heart. This suggestion is in accordance with the advice to moderate caffeine consumers to have little concern for the effect of caffeine intake on their health if their other lifestyle habits are also moderate (25). However, high caffeine intake may be deleterious in some diseases such as arrhythmias or hypertension, in which substantial activation of the sympathetic nervous system is undesirable. Long-term administration of caffeine to rats for up to 117 weeks also is associated with markedly reduced life span due to cardiovascular disease (79% of rats died, whereas in a control group, only 17% died) (20). Still, the problem of caffeine consumption by patients with diseased hearts requires further investigation.
Acknowledgment: We are grateful to Dr. S.F. Dugin for the help in management of in vivo catheterization of the rat heart.
The study was supported by Russian Fund for fundamental investigations (grant 96-04-49831).
1. Thompson WG. Coffee: brew or bane? Am J Med Sci
2. Chou TM, Benowitz NL. Caffeine
and coffee: effects on health and cardiovascular disease. Comp Biochem Physiol C Pharmacol Toxicol Endocrinol
3. Benowitz NL, Jacob P III, Mayan H, et al. Sympathomimetic effects of paraxanthine and caffeine
in humans. Clin Pharmacol Ther
4. Denaro CP, Brown CR, Jacob P III, et al. Effects of caffeine
with repeated dosing. Eur J Clin Pharmacol
5. Debrah K, Haigh R, Sherwin R, et al. Effect of acute and chronic caffeine
use on the cerebrovascular, cardiovascular and hormonal responses to orthostasis in healthy volunteers. Clin Sci (Colch)
6. Donnerstein RL, Zhu D, Samson R, et al. Acute effects of caffeine
ingestion on signal-averaged electrocardiograms. Am Heart J
7. Fabiato A. Calcium release in skinned cardiac cells: variations with species, tissues, and development. Fed Proc
8. Kapelko VI, Williams CP, Morgan JP. Intracellular calcium and mechanical function in isolated perfused hearts from rats and guinea pigs. Arch Int Physiol Biochem Biophys
9. Temples TE, Geoffray DJ, Nakamoto T, et al. Effect of chronic caffeine
intake on myocardial function during early growth. Pediatr Res
10. Kost CK Jr, Li P, Pfeifer CA, et al. Telemetric blood pressure monitoring in benign 2-kidney, 1-clip renovascular hypertension: effect of chronic caffeine
ingestion. J Pharmacol Exp Ther
11. Sham JS, Cleemann L, Morad M. Functional coupling of Ca2+
channels and ryanodine receptors in cardiac myocytes. Proc Natl Acad Sci U S A
12. Zahradnik I, Palade P. Multiple effects of caffeine
on calcium current in rat ventricular myocytes. Pflugers Arch
13. Kim DH, Landry AB III, Lee YS, et al. Doxorubicin-induced calcium release from sarcoplasmic reticulum vesicles. J Mol Cell Cardiol
14. Kapelko VI, Tsyplenkova VG, Khatkevich AN, et al. Morphological and functional estimation of acute and protracted cardiomyocyte alterations caused by Adriamycin in varied doses. Exp Clin Cardiol
15. Popovich MI, Kobets VA, Kostin SI, et al. Myocardial alterations induced by prolonged noradrenaline administration in various doses. Clin Cardiol
16. Veksler VI, Levitskaya EL, Khatkevich AN, et al. Myocardial adaptation to long-term action of substances associated with decreased intensity of cardiac function. Bioch Med Metab Biol
17. Terracciano CM, Souza AI, Philipson KD, et al. Na+
exchange and sarcoplasmic reticular Ca2+
regulation in ventricular myocytes from transgenic mice overexpressing the Na+
exchanger. J Physiol (Lond)
18. Duchen MR, Leyssens A, Crompton M. Transient mitochondrial depolarizations reflect focal sarcoplasmic reticular calcium release in single rat cardiomyocytes. J Cell Biol
19. Minana MD, Grisolia S. Caffeine
ingestion by rats increases nor-adrenaline turnover and results in self-biting. J Neurochem
20. Johansson S. Cardiovascular lesions in Sprague-Dawley rats induced by long-term treatment with caffeine
. Acta Pathol Microbiol Scand
21. Biaggioni I, Paul S, Puckett A, et al. Caffeine
and theophylline as adenosine receptor antagonists in humans. J Pharmacol Exp Ther
22. Bender AM, Donnerstein RL, Samson RA, et al. Hemodynamic effects of acute caffeine
ingestion in young adults. Am J Cardiol
23. Stubbs TA, Macdonald IA. Systemic and regional haemodynamic effects of caffeine
and alcohol in fasting subjects. Clin Auton Res
24. Astrup A, Toubro S, Cannon S, et al. Caffeine
: a double-blind, placebo-controlled study of its thermogenic, metabolic, and cardiovascular effects in healthy volunteers. Am J Clin Nutr
25. Leonard TK, Watson RR, Mohs ME. The effects of caffeine
on various body systems: a review. J Am Diet Assoc
Keywords:© 2000 Lippincott Williams & Wilkins, Inc.
Isolated heart; Caffeine; Pump function; Pressure development; High-energy phosphates