Effects of caffeine on arterial function and haemodynamics

In this issue of the journal, Vlachopoulos and O'Rourke [1] report the results of a study aimed at investigating the effects of ingestion of caffeine on peripheral (brachial) and transfer function-derived central (aortic root) blood pressure. Some important specific issues are raised by their study, in particular the place of central–peripheral blood pressure differences in assessing response and/or risk and, second, in view of the magnitude of the changes reported, the potential influence of the caffeine-induced circulatory effects on the overall cardiovascular risk. Caffeine, blood pressure and cardiovascular risk As the authors indicate, caffeine consumption is probably the most ubiquitous pharmacologically active compound used worldwide, being consumed in virtually all population groups, by men and women, by the young to very elderly and concomitantly with other recreational and therapeutic substances. In addition, the existing epidemiological literature is split between no effect [2–8] and a weak positive relationship of caffeine with cardiovascular risk [9–11]. In their study Vlachopoulos and O'Rourke [1] demonstrate a peak increase in systolic and diastolic blood pressure after ingestion of 250 mg of concentrated caffeine of approximately 11 and 7 mmHg, respectively, resulting in an increased pulse pressure amounting to 5 mmHg. With this dose, equivalent to 2–3 cups of coffee, a response still evident at 180 min post-ingestion was achieved. This is consistent with previous reports [12,13], one of which investigating the effects of a ‘more realistic’ dose (150 mg) of caffeine, that still showed effects on systolic blood pressure persisting after 90 min, although these latter investigators failed to show any significant change in diastolic blood pressure. The pressor response appears to be associated with, and probably triggered by, a marked increase in sympathetic cardiovascular drive (directly assessed by microneurographic technique), whose temporal response closely parallels the behaviour displayed by the haemodynamic variables [13]. Together, these results suggest that coffee drinkers could achieve a prolonged elevation above ‘usual’ blood pressure, with significant implications for increase in daytime total blood pressure load. Because elevated blood pressure, both combined systolic and diastolic and isolated systolic, is a clearly established cardiovascular risk [14], the demonstration of a prolonged pressor effect of caffeine in the brachial artery (upon which the longstanding epidemiology of hypertension is based) and, by inference, on central aortic blood pressure, requires consideration. What are the potential implications of caffeine-induced elevation in blood pressure? If coffee use translated to a persistent increase in ‘usual’ diastolic blood pressure of the magnitude reported by Vlachopoulos and O'Rourke [1] (approximately 2 mmHg at 2–3 h), this would be conventionally quoted as an increased risk of approximately 12% for stroke and 8% for coronary heart disease [15]. With the recent identification of pulse pressure and central pulse wave velocity, both dependent on elastic properties of the large arteries, as independent risk factors for cardiovascular mortality [16], a prolonged elevation in these parameters may also be a detrimental effect of caffeine ingestion. The increase in pulse pressure and pulse wave velocity observed by Vlachopoulos and O'Rourke [1] might equate to an increased risk of approximately 12 and 3%, respectively [16,17]. Although a persistent pressor effect would seem achievable in habitual coffee drinkers, the lack of a clearcut epidemiological association of coffee consumption, hypertension and increased cardiovascular risk possibly argues that an intermittent caffeine-induced rise in brachial or central blood pressure may not constitute the detrimental influence expected. Perhaps other factors (e.g. type of coffee, means of preparation or habituation) mitigate the effects demonstrated with concentrated caffeine. Regarding more acute greater increases in blood pressure, there is at least one report associating coffee ingestion with subarachnoid haemorrhage [18], and an increase in blood pressure of the magnitude seen by Vlachopoulos and O'Rourke at 30 min post-ingestion could conceivably precipitate an event in predisposed individuals. Methodological considerations Do we need estimated central as well as measured brachial blood pressures? At baseline, the study group recorded a difference between sphygmomanometrically measured brachial systolic and transfer function-derived central systolic blood pressure of the order of 10 mmHg, a difference consistent with previous published data using the same technique [19]. As expected, brachial and derived central diastolic blood pressure was very similar, and mean blood pressure is assigned to be equivalent as a requirement of the calculation process. It is interesting to observe that the effects of caffeine on blood pressure parameters were to produce a greater increase in derived central than measured peripheral systolic blood pressure, in effect decreasing the systolic difference associated with distal progression by approximatelly 2 mmHg and making brachial blood pressure a closer estimate of central blood pressure. A lack of distal increase in pulse pressure is normally equated with older, stiffer arteries consistent with an increased pulse wave velocity such as that shown in the present study. Of particular note is the prolonged elevation of central pulse wave velocity observed post-caffeine with no apparent decrease even 3 h post-ingestion. This is despite a declining blood pressure and augmentation index. As mentioned by Vlachopoulos and O'Rourke [1], this is at odds with the usual concept of augmentation pressure being a composite of pulse wave velocity and wave reflection. Pulse wave velocity is pressure dependent and increases with increasing systolic and diastolic blood pressure but remained elevated even when blood pressure (and augmentation pressure) was decreasing, This led the authors to suggest differential effects of caffeine on muscular and elastic arteries. This is an unexpected observation because an elevated aortic pulse wave velocity would be anticipated to increase central blood pressure by virtue of earlier arrival of reflected energy from distal aortic reflecting sites. In this instance, there appears to be some dissociation between measured pulse wave velocity and derived augmentation pressure and, in this regard, it is a pity that no results are given for other waveform parameters, such as the time to the systolic pressure inflection point, which may help to separate out mechanisms. Certainly, the discrepancy between brachial and derived central systolic blood pressure would result in an underestimation of the relative change in cardiac afterload using brachial blood pressure, but only of approximately 1%, and the clinical significance of this order of discrepancy is uncertain. Unfortunately, there are no published data concerning the use of a generalized transfer function in this type of study, and it should be remembered that the central blood pressure values reported are not directly measured. If, as described on a group basis, the central blood pressure increases substantially and the central–peripheral difference reduces, the implication is uncertain because the existing epidemiology is based on brachial blood pressure. There is no outcome data regarding level of central blood pressure; similarly, it is not known if differential change is indicative of healthy or maladaptive arterial function. Regarding assessment of effect, the small difference in change between central and peripheral pressures of approximately 2 mmHg over the first 60 min is a relatively small difference in total blood pressure load over that assessed solely from the brachial artery. Areas of further investigations In summary, Vlachopoulos and O'Rourke [1] report pronounced and prolonged effects on blood pressure following a large but by no means inconceivable dose of caffeine. Mechanistically, they demonstrate an increase in derived central augmentation pressure and a much more prolonged increased in central pulse wave velocity. These observations were associated with a small, but statistically significant, extra increase in derived central compared to measured brachial systolic blood pressure. These changes are consistent with previously demonstrated effects of coffee ingestion on arterial properties; however, the clinical significance remains uncertain. Longer follow-up (at least until return-to-resting values) and associated ambulatory blood pressure monitoring would help in our understanding of caffeine-induced effects on the cardiovascular system. Certainly, the results reported by Vlachopoulos and O'Rourke [1] support the long-standing practice in most laboratories of precluding ingestion of caffeine-containing beverages for at least 12 h before assessment of arterial function.