Cardiovascular effects of caffeine and caffeinated beverages

INTRODUCTION — Consumed in coffee, tea, and soft drinks, caffeine is the most widely used pharmacologically active substance in the world. Exposure to caffeine is generally of long duration, and the prevalence of exposure is very high (>80 percent) in most countries. The main sources of caffeine and the amount found in each are summarized in the table (table 1). Coffee and tea average 100 mg of caffeine per serving, colas approximately 50 mg per serving, and "energy drinks" as much as 250 mg per serving.

The effects of caffeine on the cardiovascular system are the subject of this review. Noncardiovascular effects of caffeine are discussed in detail elsewhere. (See "Benefits and risks of caffeine and caffeinated beverages".)

ACTIVE INGREDIENTS IN COFFEE — This topic focuses on the cardiovascular effects of caffeine. Most of the relevant studies of chronic caffeine ingestion have been performed in coffee drinkers. While coffee contains other compounds such as chlorogenic acid, potassium, magnesium, niacin, and lignans, we are not aware of evidence to suggest that these compounds influence cardiovascular outcomes in the amounts consumed by coffee drinkers. (See "Benefits and risks of caffeine and caffeinated beverages", section on 'Caffeine metabolism'.)

PHARMACOKINETICS — Caffeine is absorbed after oral, rectal, or parenteral administration, and maximal concentration is achieved in one hour [1]. Oral bioavailability is almost 100 percent; food slows the rate of absorption but does not limit the extent. Any effects on the cardiovascular system should return to baseline after approximately 10 and 60 hours as the half-life is 2 to 12 hours [2]. (See 'Cardiovascular effects' below.)

The apparent volume of distribution is between 0.4 and 0.6 L/kg, and less than 40 percent is bound to plasma proteins. Caffeine is distributed into all body compartments; it crosses the placenta and passes into breast milk.

Metabolism primarily occurs in the liver by demethylation and oxidation, with less than 5 percent excreted unchanged in the urine. The major metabolic pathway in humans is through the formation of paraxanthine (1,7-dimethylxanthine), leading to the principal urinary metabolites, 1-methylxanthine, 1-methyluric acid, and an acetylated uracil derivative. A minor pathway involves the formation of theophylline and theobromine, and their subsequent metabolism. In most patients, the drug obeys first-order kinetics; zero-order kinetics occur at higher concentrations because of saturation of metabolic enzymes.

The average half-life in plasma is five hours (range three to seven hours) but is somewhat shorter in children (three to four hours). The half-life may increase twofold in the last trimester of pregnancy. However, there is marked interindividual variation in the rate of elimination.

The main determinants of plasma caffeine concentration are the elimination half-life and time since last consumed. Since the half-life is approximately five hours, and it is usually consumed in separate doses throughout the day, the plasma caffeine concentration is normally highest in the late afternoon and lowest upon awakening in the morning [3,4]. Overnight abstinence of 8 to 12 hours leads to significant depletion of systemic caffeine by early morning, rendering the subject sensitive to its effects when re-exposure occurs [4].

There is an association between plasma levels of caffeine and toxicity and the activity of hepatic cytochrome CYP1A2. A study of 120 normal healthy volunteers found that CYP1A2 activity, biologic sex, and smoking behavior influenced the toxicity of caffeine; in particular, females and nonsmokers who experienced toxic effects had a lower caffeine N3-demethylation index (a measure of CYP1A2 activity) compared with females and nonsmokers who did not experience toxicity [5].

The rate of caffeine metabolism may be an important determinant of the effect of coffee intake on cardiovascular disease (CVD) risk (see 'Cardiovascular disease outcomes' below). This was suggested in a study that analyzed outcomes according to alleles of CYP1A2: Individuals homozygous for the CYP1A2*1A allele were rapid metabolizers of caffeine, while carriers of the CYP1A2*1F allele (55 percent of the study population) were slow metabolizers of caffeine [6]. Among rapid metabolizers, there was no association between coffee intake and nonfatal myocardial infarction. By contrast, slow metabolizers had a dose-dependent increase in risk (multivariate adjusted odds ratio 0.99, 1.36, and 1.64 for consumption of one, two to three, and four or more cups of coffee per day, respectively). These observations suggest that caffeine plays a role in the coronary risk associated with coffee intake.

CARDIOVASCULAR EFFECTS

Physiologic effects — Caffeine is a methylated xanthine derivative (1,3,7- trimethylxanthine) and is widely distributed in nature as a plant alkaloid. It affects cardiovascular hemodynamics as well as electrophysiologic properties of the heart. A number of molecular mechanisms have been detected in vitro for methylxanthines, but only some that take place at reasonably low concentrations are involved in the responses observed in vivo. These include:

●Inhibition of phosphodiesterase, which results in an elevation in myocardial cyclic adenosine monophosphate and a positive inotropic action on the myocardium.

●Inhibition of adenosine receptors, possibly preventing the negative inotropic effect elicited by adenosine [7].

●Facilitation of norepinephrine release from sympathetic nerve endings [8].

●Increase in intracellular calcium, mediated at high doses of caffeine by inhibition of calcium reuptake into the sarcoplasmic reticulum and at low doses by calcium release from the sarcoplasmic reticulum [9].

●Increase in sensitivity of the myofilaments to calcium [10,11].

Caffeine can directly increase the transmembrane calcium current that is responsible for the oscillatory afterpotential (ie, triggered activity) in in vitro and in vivo experimental models, in part via release of calcium from the sarcoplasmic reticulum [10,12-15].

The acute caffeine-induced pressor effect described below (see 'Blood pressure' below) is probably related to vasoconstriction due to:

●An increase in sympathetic activity and circulating catecholamine concentrations mediated by central nervous system stimulation [16,17]. However, the above observation that, in habitual drinkers, muscle sympathetic nerve activity is increased in the absence of a pressor response suggests that increased sympathetic activity alone is not sufficient [16].

●Antagonism of endogenous adenosine [18]. Subsequent upregulation of adenosine receptors probably explains the tolerance to chronic caffeine ingestion.

Atrial fibrillation — Caffeine does not appear to increase the risk of atrial fibrillation [19,20]. In a meta-analysis of seven observational studies with >100,000 individuals, caffeine exposure was not associated with an increased risk of atrial fibrillation (odds ratio 0.92; 95% CI 0.82-1.04) [21]. This lack of association was confirmed in a second meta-analysis [22] and a large population-based Danish cohort study [23]. By contrast, in adjusted analysis of data from the Multi-Ethnic Study of Atherosclerosis (MESA), participants who drank ≥1 cup of coffee/week had a higher incidence of atrial fibrillation compared with those who did not drink coffee, over an approximately eight-year follow-up (7 versus 4.5 percent, adjusted hazard ratio [HR] 1.40) [24]. Atrial fibrillation incidence was highest (9.8 percent) for the group consuming six or more cups per day, suggesting a potential dose-dependent relation. Strengths of the MESA design include a multiethnic cohort free from clinical cardiovascular disease (CVD). Limitations include self-recall of the exposure, which only evaluated at baseline, a high prevalence of smoking, and lack of consideration of confounders such as obesity, physical activity, and sleep.

In individuals with known paroxysmal atrial fibrillation, caffeine intake is a common trigger of subjectively reported arrhythmia symptoms [25,26]; however, subjective symptom reports in one study did not correlate with ECG-detected episodes of atrial fibrillation [27].

Other arrhythmias — Patients susceptible to cardiac arrhythmias should avoid consuming large quantities of caffeine, although modest amounts appear to be safe. Despite the theoretical relationship between caffeine and arrhythmogenesis, the available clinical evidence suggests that caffeine in doses typically consumed (ie, up to approximately 400 mg per day, which is roughly equivalent to up to five cups of coffee) (table 1) is not associated with incident atrial arrhythmias [21,28-32]. Nevertheless, there are patients who may be more sensitive to caffeine and note a relationship of palpitations to caffeine intake. In addition, there are case reports of arrhythmia in patients with underlying cardiac disease who have ingested excessive amounts of caffeine [33-37].

Three small randomized trials have evaluated the effect of caffeine on risk of arrhythmia [32,38,39]. The CRAVE trial, a randomized, case-crossover study assessed the effects of caffeinated coffee on ectopy and arrhythmias in 100 healthy United States adults [32]. Participants served as their own controls by consuming coffee or avoiding caffeine intake on alternate days. The frequency of atrial premature contractions, as measured by a continuously recording electrocardiogram (ECG) patch, did not differ with coffee consumption or caffeine avoidance. In contrast, the frequency of premature ventricular contractions was higher on days when coffee was consumed. In another trial of 51 patients with moderate to severe left ventricular systolic dysfunction randomized to caffeine (500 mg) or placebo, the frequency of ventricular and supraventricular premature beats, couplets, bigeminy, and nonsustained tachycardia was similar in both groups over approximately seven hours of continuous monitoring [39].

Data from observational studies also suggests that regular caffeine consumption does not correlate with an increased arrhythmia risk and may even have a protective effect [19,21,30,31,40-42]. As an example, in the prospective Cardiovascular Health Study, 1416 participants aged ≥65 years completed a dietary assessment (including frequency of consumption of caffeinated products) and underwent 24-hour ambulatory electrocardiography (Holter) monitoring [40]. There were no differences in the number of supraventricular or ventricular premature beats across different levels of caffeine intake, even after adjustment for potential confounders.

Caffeine does not appear to enhance the ability to induce an arrhythmia in the electrophysiology (EP) laboratory [28,43]. In one study, 80 patients with symptomatic supraventricular tachycardia (SVT) undergoing EP study prior to catheter ablation were randomized to caffeine (5 mg/kg) or placebo [43]. There was no difference between groups in SVT inducibility or the cycle length of induced tachycardias. In another prospective study, 22 patients with a history of SVT or fibrillation underwent EP testing before and one hour after coffee ingestion (275 mg of caffeine) [28]. The ability to induce an arrhythmia was unchanged in 10 patients, increased in six, and reduced in six; and rhythm severity was unchanged in 17 patients, more severe in two, and less severe in three.

Heart rate variability — Heart rate variability has been examined in a number of studies among habitual consumers of caffeine or coffee [44,45]. In a systematic review of 12 trials that evaluated the effect of 400 to 500 mg caffeine per day, there was not a consistent relationship with heart rate variability; however, the evidence as a whole supports the notion that 400 mg caffeine per day in healthy adults is not associated with an increases in heart rate variability [46].

Blood pressure — Caffeine can acutely raise the blood pressure (BP) by as much as 10 mmHg in individuals who are infrequently exposed [16,47-49], but there is little or no acute effect on BP in habitual coffee drinkers [16,49].

With regard to its impact in nonhabitual coffee drinkers, caffeine can potentiate (by approximately 5 to 10 mmHg) a rise in BP induced by stress, such as that occurring in the workplace [50,51]. The acute elevation BP caused by caffeine is more pronounced in hypertensive patients; compared with subjects with optimal blood pressure, those with hypertension had a >1.5-fold greater increase in blood pressure [2]. The hypertensive effect may be more prominent in older adult patients with hypertension [52].

The effects of chronic caffeine ingestion on BP are less clear, but the possibility of a slight increase in BP exists [46,53]. One systematic review and meta-analysis of 10 randomized controlled trials and five cohort studies found no effect of chronic coffee consumption on BP and the risk of development of hypertension [54]. However, two other meta-analyses found that coffee ingestion increased BP slightly. In the first of these two, systolic and diastolic BP increased by 1.2 and 0.49 mmHg, respectively [55]. In a meta-analysis of six prospective cohort studies with a mean follow-up of more than six years, the pooled relative risks [RR] for hypertension were 1.09 (95% CI 1.01-1.18), 1.07 (95% CI 0.96-1.20), and 1.08 (95% CI 0.96-1.21) comparing one to three, three to five, and more than five cups to less than one cup of coffee per day, respectively. Consistent with this observation is that similar reductions in BP may be seen when habitual coffee drinkers either abstain from coffee or switch to decaffeinated coffee [56-58].

The effect of coffee on BP may not be solely explained by caffeine. In a study of 15 volunteers (six habitual and nine nonhabitual coffee drinkers), intravenous caffeine increased sympathetic nerve activity in muscle and BP to a similar degree in both groups [16]. Caffeinated coffee raised the BP only in nonhabitual drinkers, despite similar plasma caffeine concentrations and a similar increase in muscle sympathetic nerve activity. Furthermore, nonhabitual drinkers had similar elevations in BP and muscle sympathetic nerve activity with caffeinated and decaffeinated coffee. By comparison, ingestion of tea does not appear to have a significant effect upon BP. A meta-analysis of five studies of tea ingestion among 343 subjects reported that BP was unchanged at a median duration of four weeks [59].

The likelihood of a pressor response is in part determined by the pharmacokinetics of caffeine in each individual. This was illustrated in a study of normotensive habitual coffee drinkers who ingested coffee after caffeine abstinence of 4.5 individual caffeine half-lives; the ensuing increase in BP was inversely related to the baseline plasma caffeine concentration [60]. Thus, rapid metabolizers of caffeine are more likely to have a pressor response after drinking coffee, despite the phenomenon of adaptation.

Diuretic effect — Caffeine and related methylxanthine compounds in higher doses may have a diuretic action. One literature review suggested that acute ingestion of caffeine in doses of at least 250 to 300 mg, equivalent to amounts found in two to three cups of coffee or five to eight cups of tea, results in short-term stimulation of urine output in individuals deprived of caffeine for a period of days or weeks [61]. A tolerance to the diuretic effect develops and diureses is diminished in those who regularly consume tea or coffee. Doses of caffeine equivalent to amounts found in standard servings of tea, coffee, and carbonated soft drinks appear to have little diuretic action.

Aortic stiffness — Aortic stiffness is an important determinant of the performance of the cardiovascular system and prognosis of cardiovascular risk. Caffeine acutely increases arterial stiffness, and the effect is synergistic with smoking [62]. The ingestion of 200 mg of caffeine (equivalent to two cups of coffee) increased pulse-wave velocity and augmentation index by 0.33 m/s and 4 percent, respectively, an effect that was augmented when combined with smoking (increased by 0.85 m/s and 17.4 percent, respectively). (See "Increased pulse pressure".)

Lipids — The evidence presented below suggests no major impact of filtered coffee consumption on the lipid profile. Thus, we do not recommend discontinuation of coffee consumption to improve the lipid profile. Most studies have shown no adverse effects of filtered coffee intake on plasma lipids [63-67]. In the prospective cohort study cited above of over 132,000 adults from the Health Professionals Follow-Up Study and the Nurses' Health Study, there was no association between the intake of filtered caffeinated coffee or filtered decaffeinated coffee and serum concentrations of total cholesterol, low-density lipoprotein (LDL) cholesterol, or high-density lipoprotein (HDL) cholesterol [63].

However, there may be a hyperlipidemic effect of nonfiltered coffee [64,68] which contains coffee lipid compounds (such as cafestol) that are removed by a paper filter [69]. In one study, the daily ingestion of four to six cups of boiled, but not filtered, coffee for a period of nine weeks increased serum total cholesterol by 18.5 mg/dL (0.48 mmol/L) and LDL cholesterol by 15 mg/dL (0.39 mmol/L), with no effect on HDL cholesterol [64]. A similar intake of filtered coffee had a similar effect on serum lipids as drinking no coffee.

CARDIOVASCULAR DISEASE OUTCOMES — Intake of small or moderate amounts of caffeine or caffeinated beverages (≤6 cups of coffee per day) is not associated with worse cardiovascular disease (CVD) outcomes including mortality, stroke, need for coronary revascularization, or sudden cardiac arrest. Thus, we do not recommend caffeine cessation to decrease CVD risk. It is reasonable to attempt a trial of caffeine cessation in those with unacceptable palpitations.

Cardiovascular disease incidence — There appears to be no adverse impact of caffeine or caffeinated beverage consumption on CVD risk based on many large observational studies [63,70,71].

A 2014 meta-analysis of 36 prospective cohort studies, which included nearly 1.3 million individuals, evaluated the relationship between coffee consumption and CVD risk (a composite of coronary heart disease events, stroke, heart failure, and CVD mortality) and found a significant nonlinear relationship (figure 1) [72]. Compared with individuals who consumed no coffee, the relative risk [RR] of CVD was:

●0.95 (95% CI 0.87-1.03) for those who consumed a median of 5 cups per day

●0.85 (95% CI 0.80-0.90) for those who consumed a median of 3.5 cups per day

●0.89 (95% CI 0.84-0.94) for those who consumed a median of 1.5 cups per day

Stroke prevention — Coffee and tea consumption have been associated with a lower incidence of stroke. The following are representative studies:

●The Swedish Mammography Cohort prospectively followed (mean follow-up of 10.4 years) 34,670 women without a history of CVD at baseline in 1997 [73]. After adjustment for other risk factors, coffee consumption was associated with a statistically significant lower risk of total stroke, cerebral infarction, and subarachnoid hemorrhage but not intracerebral hemorrhage. The multivariable RR of total stroke across categories of coffee consumption (less than one cup per day, one to two cups per day, three to four cups per day, and at least five cups per day) were 1.00, 0.78 (95% CI 0.66-0.91), 0.75 (95% CI 0.64-0.88), and 0.77 (95% CI 0.63-0.92), respectively.

●The Nurses' Health Study found that women who consumed at least four cups of coffee per day had a significant 20 percent lower risk of total stroke than those who seldom (<1 per month) drank coffee over a 24-year follow-up [74].

●In a prospective cohort study of male Finnish smokers, with a mean follow-up of 13.5 years, high consumption of coffee and tea was associated with a lower risk of cerebral infarction independent of known cardiovascular risk factors [75].

●A meta-analysis of 11 prospective studies (n = 479,689) with 10,003 cases of stroke reported that moderate coffee consumption was weakly inversely associated with risk of stroke [76]. Compared with no coffee consumption, the RR of stroke were 0.86 (95% CI 0.78-0.94) for two cups of coffee per day, 0.83 (95% CI 0.74-0.92) for three to four cups per day, 0.87 (95% CI 0.77-0.97) for six cups per day, and 0.93 (95% CI 0.79-1.08) for eight cups per day.

●In a large cohort study conducted between 2004 and 2015 among 487,377 persons in China, daily tea drinkers (mostly green tea) had a lower risk of incident ischemic and hemorrhagic stroke compared with non-tea drinkers (adjusted hazard ratio [HR] 0.92, 95% CI 0.89-0.95) [77]. The risk was even lower when analysis was limited to men (HR 0.89, 95% CI 0.85-0.93).

●In a prospective cohort study of 365,682 persons in the United Kingdom, those who drank two to three cups of coffee and two to three cups of tea per day had a 32 percent (HR 0.68, 95% CI 0.59-0.79) lower risk of stroke compared with those who did not drink tea and coffee, after a median follow-up of 11.4 years [78].

Mortality — Most [79-83], but not all [84], observational studies have found an inverse relationship between coffee consumption and all-cause and cardiovascular mortality. These findings have been noted in many different racial and ethnic groups [82,83]. Some data suggest a similar relationship between tea consumption and cardiovascular mortality [85]. The relationship between coffee and tea consumption and all-cause mortality is discussed separately. (See "Benefits and risks of caffeine and caffeinated beverages", section on 'All-cause mortality'.)

One of the largest of the studies finding an inverse relationship examined the association between coffee consumption and total and cause-specific mortality in 229,119 men and 173,141 women, all of whom were relatively healthy at baseline [81]. During more than five million person-years of follow-up, there were 33,731 and 18,784 deaths. After adjustment for tobacco-smoking status and other potential confounders, there was a significant inverse association between coffee consumption and all-cause and cardiovascular mortality. Adjusted hazard ratios for death due to heart disease among men who drank coffee, as compared with those who did not, were as follows: 0.93 (95% CI 0.85-1.02) for drinking less than one cup per day; 0.92 (95% CI 0.84-1.01) for one cup; 0.86 (95% CI 0.79-0.94) for two or three cups; 0.87 (95% CI 0.79-0.96) for four or five cups; and 0.88 (95% CI 0.78-1.00) for six or more cups of coffee per day (p<0.03 for trend). The respective HRs among women were 1.00 (95% CI 0.89-1.13), 0.91 (95% CI 0.81-1.03), 0.85 (95% CI 0.76-0.95), 0.78 (95% CI 0.68-0.90), and 0.72 (95% CI 0.59-0.88) (p<0.001 for trend).

However, in the Aerobics Center Longitudinal Study, which included 43,727 participants followed for an average of 17 years, convincing evidence for either a positive or a negative relationship was not found. With regard to cardiovascular mortality, the adjusted HRs for consumption averaging more than four cups of caffeinated coffee per day were not statistically significant (HR 1.15, 95% CI 0.89-1.49 and HR 0.73, CI 0.23-2.27, respectively) [84].

Sudden cardiac arrest — The effect of caffeine on the risk of sudden cardiac arrest has not been well studied. A case-control study of 117 patients with an out-of-hospital sudden cardiac arrest (all with a prior clinical history of coronary heart disease) and a group of unselected age- and gender-matched coronary heart disease control patients found that the consumption of more than 10 cups of coffee per day was a risk factor for sudden cardiac arrest after adjustment for accepted risk factors [86]. The consumption of large amounts of pure caffeine (in powder or tablets) can produce serious and fatal results [87].

SUMMARY

●Consumed in coffee, tea, soft drinks, and small amounts in chocolate, caffeine is the most widely used pharmacologically active substance in the world. The main sources of caffeine and the amount found in each is summarized in the table (table 1). (See 'Introduction' above.)

●The available evidence suggests that coffee consumption does not have a major impact on the lipid profile. However, it is possible that there is a slight increase in blood pressure with chronic coffee use. (See 'Lipids' above and 'Blood pressure' above.)

●Caffeine in doses typically consumed does not provoke arrhythmias, and its routine use in patients who have or are at risk for arrhythmia is acceptable. However, patients with underlying cardiac disease should be counseled to avoid ingesting "energy drinks" and other beverages with excessive amounts of caffeine. (See 'Other arrhythmias' above.)

●Concerns regarding possible deleterious cardiovascular effects of caffeine when used in small or moderate amounts (eg, ≤six cups of coffee per day) on cardiovascular outcomes such as death or stroke are not supported by the literature. There is insufficient evidence to draw conclusions about the cardiovascular effects of consuming more than this amount of coffee per day. (See 'Cardiovascular disease outcomes' above.)

●We do not recommend discontinuing caffeine intake for the purpose of improving cardiovascular outcomes. However, it is reasonable to attempt a trial of caffeine cessation in patients who are bothered by palpitations. (See 'Cardiovascular disease outcomes' above.)