Effect of chronic coffee consumption on aortic stiffness and wave reflections in hypertensive patients

Transcription

1 (2007) 61, & 2007 Nature Publishing Group All rights reserved /07 $ ORIGINAL ARTICLE on aortic stiffness and wave reflections in hypertensive patients CV Vlachopoulos, GG Vyssoulis, NA Alexopoulos, AI Zervoudaki, PG Pietri, KA Aznaouridis and CI Stefanadis Hypertension Unit, 1st Cardiology Department, Athens Medical School, Hippokration Hospital, Athens, Greece Objective: Aortic stiffness and wave reflections are important markers and prognosticators of cardiovascular risk. Caffeine increases acutely aortic stiffness and wave reflections. Furthermore, chronic coffee consumption is associated with increased aortic stiffness and wave reflections in normotensive subjects. In the present study, we aimed to assess the association between chronic coffee consumption, and aortic stiffness and wave reflections in hypertensive patients. Design: Epidemiological survey. Setting: Hypertension Unit, University Hospital. Subjects methods: We examined 259 never-treated hypertensives (age years, 165 males) without diabetes mellitus, who were asked to describe in detail the type and amount of coffee they consumed. Carotid-femoral pulse wave velocity (PWV) and augmentation index (AIx) were measured non-invasively as indices of aortic stiffness and wave reflections, respectively. Results: When controlled for gender, age, height, smoking status, heart rate, mean pressure, HDL cholesterol and hscrp, AIx was found to be higher with increasing daily coffee consumption. Post hoc analysis revealed that all groups of coffee consumption had higher AIx compared to no-consumption. PWV did not differ among groups of daily coffee consumption. Each participant had 35% higher relative risk of having high AIx for each cup (150 ml) of coffee per day, and 40% higher relative risk for each 10 cup-years. Conclusions: Coffee consumption is associated with increased wave reflections, but not aortic stiffness in never-treated hypertensive patients. This finding may have important clinical implications for cardiovascular health in hypertensive subjects. Sponsorship: None. (2007) 61, ; doi: /sj.ejcn ; published online 13 December 2006 Keywords: caffeine; coffee; hypertension; pulse wave velocity; wave reflections Introduction Correspondence: Dr CV Vlachopoulos, Hypertension Unit, 1st Cardiology Department, Athens Medical School, Hippokration Hospital, 17, Kerassoundos Street, Athens 11528, Greece. Guarantors: CV Vlachopoulos and NA Alexopoulos. Contributors: CVV conceived the idea, designed the study, interpreted the results and drafted the manuscript. GVV interpreted the results and drafted the manuscript. NAA interpreted the results and drafted the manuscript. AIZ and PGP collected data and drafted the manuscript. KAA interpreted the results and drafted the manuscript. CIS interpreted the results and drafted the manuscript. Received 10 March 2006; revised 4 September 2006; accepted 11 October 2006; published online 13 December 2006 Coffee is a widely consumed beverage worldwide, and caffeine is probably the most frequently ingested pharmacologically active substance in the world; however, studies on the effect of coffee on cardiovascular risk are not conclusive. Although some of them have demonstrated a positive association between coffee consumption and cardiovascular risk, others have shown no association (LaCroix et al., 1986; Kleemola et al., 2000; Lopez-Garcia et al., 2006). Furthermore, we and others have previously shown that there is a J- or a U- shaped relationship between coffee consumption and the risk of developing acute coronary syndromes (Panagiotakos et al., 2003; Happonen et al., 2004). Caffeine and risk of hypertension is another unresolved issue; whereas coffee or caffeine acutely increases blood pressure (Robertson et al., 1978; Pincomb et al., 1996; Hartley et al., 2000; Corti et al., 2002), the chronic effect of coffee consumption on blood pressure is less well known. Despite the fact that tolerance may ensue in some subjects

2 (Robertson et al., 1981; Lovallo et al., 2004), there are data that coffee consumption is associated with chronic increases in blood pressure (Jee et al., 1999; Rakic et al., 1999; Klag et al., 2002; Noordzij et al., 2005; Savoca et al., 2005). Although not yet fully integrated in clinical practice, large artery stiffness and arterial wave reflections are important determinants of left ventricular function, coronary blood flow and mechanical integrity of arteries (Nichols and O Rourke, 1998; Vlachopoulos and O Rourke, 2000), and they have been identified as independent markers and prognosticators of cardiovascular risk (Stefanadis et al., 1997, 2000; Laurent et al., 2001; London et al., 2001; Weber et al., 2004; Sutton-Tyrrell et al., 2005). Hypertension is associated with increased aortic stiffness and enhanced wave reflections; furthermore, increased aortic stiffness and enhanced wave reflections are implicated in the pathogenesis of systolic hypertension (Stefanadis et al., 1997; Nichols and O Rourke, 1998; Vlachopoulos and O Rourke, 2000). On the other hand, it has been shown that caffeine exerts an acute detrimental effect on aortic stiffness and wave reflections both in normotensive and in hypertensive subjects (Vlachopoulos et al., 2001, 2003a, 2003b, 2004a; Mahmud and Feely, 2001; Karatzis et al., 2005). The chronic effect of coffee consumption on arterial elastic properties is less well studied. As we have previously shown, coffee consumption has a chronic detrimental effect on aortic stiffness and wave reflections in normotensive subjects (Vlachopoulos et al., 2004a, 2005a). However, the association between chronic coffee consumption, and arterial stiffness and wave reflections in hypertensives, that is, a population with already impaired elastic properties, has not been examined yet. Accordingly, the aim of the study was to investigate the chronic effect of coffee on aortic pulse wave velocity and wave reflections indices in hypertensive patients. Subjects and methods Study population The study population consisted of 259 consecutive patients, age years (mean7standard deviation (s.d.)) with uncomplicated, never-treated essential hypertension, as defined in the Joint National Committee VII guidelines, who were examined in the Hypertension Unit of our institution. Of them, 165 (64%) were males and 94 (36%) females. None of them had diabetes mellitus or used hypolipidaemic agents. All subjects were studied in the morning after an overnight fast in a quiet, air-conditioned room. Following a 15-min rest, arterial blood pressure was measured three times with a mercury sphygmomanometer applied on the right arm. Subsequently, arterial function measurements were obtained. The study was approved by the local ethics committee, and all the patients gave written informed consent. Coffee intake assessment All participants were asked to describe their usual frequency of coffee consumption during the last year, using a caffeine frequency questionnaire, validated with the same methodology as the food frequency questionnaire developed by the Unit of Nutrition of Athens Medical School (Katsouyanni et al., 1997). The participants were asked to describe in detail the type of coffee they consumed (instant, Greek type, filtered, espresso or cappuccino ) and the amount of coffee (number of teaspoons) they used for their specific coffee preparation. All reported types of coffee were then adjusted for one cup (150 ml) of coffee and caffeine concentration of 80 mg (McCusker et al., 2003; Lelo et al., 1986). According to the distribution of coffee consumption, we categorized usual coffee consumption as: (1) none, (2) low (o200 ml/day), (3) moderate ( ml/day) and (4) high (4450 ml/day). For better evaluation of the chronic effect of coffee, we also calculated the cup-years consumption of each participant, as the mean daily consumption of coffee (in cups) multiplied by the years of coffee consumption. The consumption of other caffeine-containing beverages, that is, cola drinks and tea were also recorded. According to self-reported data, none of the participants took medications (whether prescribed or over-the-counter) that contained caffeine. Evaluation of arterial elastic properties The pulse travels at a higher velocity in a stiff aorta and vice versa. Carotid-femoral pulse wave velocity (PWV), an established index of aortic stiffness (Nichols and O Rourke, 1998; O Rourke et al., 2002; Vlachopoulos et al., 2004b), was calculated from measurements of pulse transit time and the distance travelled between two recording sites (pulse wave velocity ¼ distance (meters)/transit time (seconds)) using a validated non-invasive device (Complior, Artech Medical, Pantin, France), which allows online pulse wave recording and automatic calculation of pulse wave velocity (Asmar et al., 1995). Two different pulse waves were obtained simultaneously at two sites (at the base of the neck for the common carotid and over the right femoral artery) with two transducers. The distance was defined as: (distance from the suprasternic notch to femoral artery) (distance from carotid artery to the suprasternic notch). Measurement of wave reflection indices Augmentation index (AIx) of the central (aortic) pressure waveform was measured as an index of wave reflections (Vlachopoulos and O Rourke, 2000; Vlachopoulos et al., 2001, 2004b; O Rourke et al., 2002). Augmented pressure is the pressure added to the incident wave by the returning reflected one and represents the pressure boost that is caused by wave reflections and with which the left ventricle must cope. AIx (defined as augmented pressure divided by pulse pressure and expressed as a percentage) is a composite measure of the magnitude of wave reflections and arterial 797

3 798 Aortic Pressure (mmhg) AP Pi PP Pulse Wave Analysis Augmentation Index = AP/PP Figure 1 Representation of pulse waveform analysis: This is the ascending aortic pressure waveform that is synthesized from the pressure waveform of the radial artery (not shown) that the tonometer records. An inflection point in the waveform identifies the merging point of the incident and the reflected wave. AIx, augmentation index; AP, augmented pressure; Pi, inflection point; PP, pulse pressure. stiffness, with the latter affecting timing of wave reflections (Figure 1). Larger values of AIx indicate increased wave reflections from the periphery and/or earlier return of the reflected waves as a result of increased pulse wave velocity (owing to increased arterial stiffness), and vice versa. Timing of the reflected waves (Tr), that is, the time the pulse wave needs to travel to the periphery and return to meet the incident wave, is an index of pulse wave velocity and was also calculated. All these indices were measured by using a validated, commercially available system (SphygmoCor, AtCor Medical, Sydney, Australia), which employs the principle of applanation tonometry and appropriate acquisition and analysis software for non-invasive recording and analysis of the arterial pulse. The technique has been described in detail previously (Nichols and O Rourke, 1998; Pauca et al., 2001). In brief, from radial artery recordings, the central (aortic) arterial pressure was derived with the use of a generalized transfer function, which has been shown to give an accurate estimate of the central arterial pressure waveform and its characteristics (Nichols and O Rourke, 1998; Pauca et al., 2001; Gallagher et al., 2004). Waveforms of radial pressure were calibrated according to sphygmomanometric systolic and diastolic pressure measured in the brachial artery as there is practically negligible pressure pulse amplification between the brachial and the radial artery (Nichols and O Rourke, 1998). Statistical analysis Sample size calculations were based on data from our unit, which showed that the standard deviations of PWV and AIx for hypertensive patients with characteristics similar to those of our study population were 1.2 m/s and 12%, respectively. Therefore, we estimated that 24 subjects per group would provide 80% power at the 5% level of significance to detect a difference of 1 m/s in PWV and of 10% in AIx among the study subgroups. Continuous variables are presented as mean values7s.d. (normally distributed variables), or median (interquartile range) (skewed variables), whereas qualitative variables are presented as absolute and relative frequencies. Normality was tested using the Kolmogorov Smirnov criterion. Logarithmic transformation was performed for skewed distributions, that is, high sensitivity C-reactive protein (hscrp), before any parametric analysis. Among groups, comparisons of continuous variables were performed using Student s t-test for unpaired samples or oneway analysis of variance (ANOVA). The association of the subgroups of coffee consumption (no consumption, low, moderate and high consumption) with indices of arterial function after adjusting for potential confounders was evaluated with analysis of covariance (ANCOVA). In ANCO- VA analyses, the clinical variables that correlated significantly in univariate analysis or have biological relevance with arterial function indices were introduced as confounders. Arterial function indices between each two consumption subgroups were compared after correcting the P-value for multiple comparisons using the Bonferroni post hoc test. Contingency tables and the w 2 test were applied to compare categorical parameters. Correlations between variables were evaluated by calculation of Pearson s correlation coefficient for the normally distributed variables (and hscrp after logarithmic transformation). In order to evaluate the correlation between arterial function indices and the cupyears of coffee consumption (highly skewed), the Spearman correlation coefficient was used. Multivariable logistic regression analysis was used to estimate the adjusted odds ratio (OR) for having high AIx (dependent variable: AIx X36% (highest quartile) ¼ 1, and AIx o36% (first, second and third quartiles) ¼ 0) according to the consumption of coffee (main independent variable, expressed in ml of daily consumption or in cup-years). The independent variables of each of these models comprised eight parameters that were chosen because of statistical correlation and/or biological relevance, plus ml of daily coffee consumption or cup-years. The overall significance of these models was evaluated with likelihood ratio tests. All tests were two tailed and exact P-values o0.05 were considered statistically significant. Data analysis was performed with SPSS software, version 10.1 (Chicago, IL, USA). Results Coffee intake and clinical characteristics Table 1 illustrates demographic and clinical characteristics of the participants. Of them, 64 (25%) reported no coffee consumption, whereas 42 (16%) reported low coffee consumption, 128 (49%) reported moderate coffee consumption and 25 (10%) reported high coffee consumption. There was

4 Table 1 Characteristics of the participants according to coffee consumption No consumption Low consumption Moderate consumption High consumption P-value 799 Number (%) 64 (25) 42 (16) 128 (49) 25 (10) Male sex (%) 37 (58) 26 (62) 81 (63) 21 (84) Age (year) Height (cm) Weight (kg) o0.02 Body mass index (kg/m 2 ) o0.05 Smoking (%) 18 (28.1) 19 (45.2) 57 (44.5) 17 (68) o0.01 Total cholesterol (mg/dl) HDL cholesterol (mg/dl) Triglycerides (mg/dl) Glucose (mg/dl) hscrp (mg/dl) 0.99 (1.65) 1.99 (2.67) 1.08 (1.58) 1.17 (1.58) Heart rate (bpm) Peripheral SP (mm Hg) Peripheral DP (mm Hg) Peripheral PP (mm Hg) Peripheral Mean P (mm Hg) Aortic SP (mm Hg) Aortic DP (mm Hg) Aortic PP (mm Hg) Aortic mean P (mm Hg) Abbreviations: ANOVA, analysis of variance; DP, diastolic pressure; HDL, high-density lipoprotein; hscrp, high sensitivity C-reactive protein; PP, pulse pressure; SP, systolic pressure. Categorical variables are presented as relative frequencies, while continuous variables with normal distribution as mean7s.d.; hscrp (skewed variable) is presented as median (interquartile range). P-values derived from w 2 test for categorical variables and from one-way ANOVA for normally distributed variables (and for hscrp after log transformation) between the different groups of coffee consumption. no difference between the groups regarding peripheral and central systolic, diastolic, mean or pulse pressures. Other caffeinated drinks Tea consumption was very low in our study population; only 18 (7%) subjects reported tea consumption, so this variable was omitted from further analyses. Consumption of cola drinks was reported by 105 (40.5%) subjects. Cola drinks consumption did not differ among different groups of coffee consumption (44, 43, 38 and 40% in no-, low-, moderate and high-consumption groups, respectively, P ¼ 0.887), and arterial function indices did not differ between cola drinkers and non-drinkers (AIx, vs % respectively, P ¼ 0.577; PWV, vs m/s respectively, P ¼ 0.674). Other caffeinated drinks are not popular in middle-aged Greeks. Coffee intake and indices of arterial function AIx was lower in males than in females ( vs %, Po0.001) and higher in smokers than in nonsmokers ( vs %, Po0.01). There was a significant correlation between AIx and age (r ¼ 0.474, Po0.001), height (r ¼ 0.555, Po0.001), weight (r ¼ 0.448, Po0.001), body mass index (r ¼ 0.132, Po0.05), heart rate (r ¼ 0.251, Po0.001), diastolic blood pressure (r ¼ 0.129, Po0.05), mean pressure (r ¼ 0.297, Po0.001), total cholesterol (r ¼ 0.140, Po0.05) and high density lipoprotein (HDL) cholesterol (r ¼ 0.297, Po0.001). PWV was not different between males and females ( vs m/s respectively, P ¼ 0.243) or between smokers and non-smokers ( vs m/s respectively, P ¼ 0.325). There was a significant correlation between PWV and age (r ¼ 0.383, Po0.001), height (r ¼ 0.156, Po0.02), heart rate (r ¼ 0.144, Po0.05), systolic pressure (r ¼ 0.364, Po0.001), pulse pressure (r ¼ 0.364, Po0.001), mean pressure (r ¼ 0.268, Po0.001) and hscrp (r ¼ 0.213, Po0.001). PWV was not associated with the cup-years of coffee consumption (r ¼ 0.028, P ¼ 0.651). On the contrary, AIx was strongly associated with the cup-years of coffee consumption (r ¼ 0.264, Po0.001). Indices of arterial function were compared among different groups of coffee consumption taking into account the effect of possible confounders using ANCOVA. When controlled for gender, age, height, smoking status, heart rate, mean pressure, HDL cholesterol and hscrp, AIx was found to be higher with increasing daily coffee consumption (Po0.002, Figure 2a), although Tr did not differ (P ¼ 0.487, Figure 2b). Post hoc analysis revealed that all groups of coffee consumption had higher AIx compared to no consumption group (Figure 2a). When controlled for gender, age, body mass index, smoking status, heart rate, mean pressure, HDL cholesterol and hscrp, PWV did not differ among groups of daily coffee consumption (P ¼ 0.156, Figure 2c). Finally, neither central, nor peripheral systolic, diastolic, mean and

5 800 Wave reflections. Despite the lack of association between coffee consumption and aortic stiffness, we observed that a positive association between coffee consumption and wave reflections exists. Indices of wave reflections are composite; they are affected not only by the timing of the reflected waves, that is, at what point during the cardiac cycle they return to the ascending aorta, but also by the magnitude of the reflected waves, that is, how much of the forward wave is reflected at the periphery. The timing of the reflected waves (indicated by Tr) depends greatly on large arteries stiffness (mainly aortic): neither PWV nor Tr was found to be associated with coffee consumption. However, the magnia % Augmentation Index Overall P< P<0.02 P<0.002 P< b sec Timing of Reflected Waves 139 P= no low mod high no low mod high no low mod high Daily Coffee Consumption Daily Coffee Consumption Daily Coffee Consumption 139 c m/sec Pulse Wave Velocity 7.54 P=0.156 Figure 2 Values of augmentation index (a), timing of reflected waves (b) and pulse wave velocity (c) in the different groups of daily coffee consumption after adjustment for gender, age, height (for AIx and Tr) or body mass index (for PWV), smoking status, heart rate, mean pressure, HDL cholesterol and hscrp Table 2 Results from two different logistic regression models that evaluated the relative risk of a participant to have high AIx (above 75th percentile, i.e. AIxX36%) for each ml of daily coffee consumption or for each cup-year Dependent variable a Main independent variable B(Exp) P-value AIx Daily coffee (ml) o0.05 AIx Cup-years of coffee (ml a years) o0.05 Abbreviations: AIx, augmentation index; HDL, high-density lipoprotein; hscrp, high sensitivity C-reactive protein. a In each model adjustments were performed for gender, age, smoking status, height, heart rate, mean pressure, HDL cholesterol and hscrp. increases acutely aortic stiffness and wave reflections both in normotensive and hypertensive patients (Vlachopoulos et al., 2001, 2003a, b, 2004a), and that chronic coffee consumption is associated with increased aortic stiffness and wave reflections in healthy normotensive subjects without cardiovascular risk factors, except from smoking (Vlachopoulos et al., 2004a, 2005a). Interestingly, the findings of the present study are not a direct extrapolation of the findings in normotensive subjects, because, contrary to those, only wave reflections are affected, indicating a different behaviour of large and small arteries in hypertension. pulse pressures were found to be different among various groups (data not shown). Logistic regression analysis showed that the adjusted OR of a participant to have high AIx (above 75th percentile, i.e., AIxX36%) was (1.002, Po0.05) for a 1 ml/day increase of coffee consumption, and (1.034, Po0.05) for a 1 cup-year increase of coffee consumption (Table 2). For a more comprehensive interpretation, we calculated that an increase of daily coffee consumption by one cup (150 ml) is associated with a 35% increase in the odds of having a high AIx (OR ¼ 1.35). Accordingly, in subjects with a 10 cup-years higher coffee consumption, there is a 40% increase in the odds of having a high AIx (OR ¼ 1.40). Discussion To the best of our knowledge, this is the first study to demonstrate that in hypertensive patients coffee consumption is associated with increased wave reflections, whereas it is not associated with aortic stiffness. This study extends the findings of our previous studies that showed that caffeine Mechanisms Aortic stiffness. Aortic stiffness is already increased in hypertensive patients, irrespective of whether this is due to the increased distending pressure or to the alterations in the intrinsic properties of the arterial wall (Asmar et al., 1995; Stefanadis et al., 1997; Nichols and O Rourke, 1998). According to our results, the stiffening effect of coffee in the aorta is probably less prominent in the already stiff aortas of hypertensive patients, than in normal aortas. In other words, it appears that coffee does not have a stiffening effect on the aorta additive to that of hypertension.

6 tude of the reflected waves depends mainly on the tone of the arterioles. Vasorelaxation results in decreased wave reflection indices, whereas vasoconstriction has the opposite effect. As it has been demonstrated, caffeine exerts its main cardiovascular effects through the antagonism of adenosine and the release of catecholamines, both resulting in vasoconstriction (Robertson et al., 1978; Smits et al., 1987; Varani et al., 1999; Corti et al., 2002; Lane et al., 2002; Brodmann et al., 2003; Saboury et al., 2003). In this context, it has been shown that norepinephrine infusion increases augmentation index (Wilkinson et al., 2001; Stewart et al., 2003). Thus, chronic coffee consumption may induce increased magnitude of wave reflections, which in turn increases augmentation index, despite the fact that pulse wave velocity and, hence, timing of reflected waves are not affected. Chronic coffee consumption is associated with increased inflammatory markers (Zampelas et al., 2004), denoting lowgrade inflammation. Despite the fact that acute inflammation decreases wave reflections (Vlachopoulos et al., 2005b), chronic low-grade inflammation has been linked to enhanced wave reflections in some studies (Kampus et al., 2004; Kullo et al., 2005; Mahmud and Feely, 2005). In our study, however, we found no association between coffee consumption and hscrp, a finding that clearly deserves further investigation. Clinical implications Our study has important clinical implications. Enhanced wave reflections augment aortic systolic pressure and, by this, left ventricular afterload. When they return late, as in normal subjects, they facilitate coronary perfusion. In contrast, early return of reflected waves, as in hypertensives, compromises coronary perfusion. Furthermore, by increasing pulse pressure, they increase pulsatile stretch of the arteries leading to mechanical fatigue of their elastic components. Accordingly, wave reflections are important markers and prognosticators of cardiovascular risk (London et al., 2001; Weber et al., 2004; Sutton-Tyrrell et al., 2005). Our study provides an explanatory mechanism for the results of previous investigations that showed an increase in cardiovascular risk with chronic coffee consumption (LaCroix et al., 1986; Panagiotakos et al., 2003; Happonen et al., 2004). Limitations The high consumption group consisted of 25 subjects, a number that was sufficient to elucidate differences in augmentation index in this group. However, this number was not adequate to perform subanalyses in various subgroups of our population, that is, in females only, or in non-smokers only. The consumption of other nutrients not included in our analyses may have confounded the results of our study; however, there is no apparent reason why these could be unequally distributed in the various groups. In conclusion, chronic coffee consumption is associated with increased wave reflections in hypertensive patients, although it does not affect the already impaired aortic elastic properties. Given the impact of wave reflections on cardiovascular performance and risk, and the widespread consumption of coffee worldwide, our results may have important implications for cardiovascular health in hypertensive subjects. References Asmar R, Benetos A, Topouchian J, Laurent P, Pannier B, Brisac A-M et al. (1995). Assessment of arterial distensibility by automatic pulse wave velocity measurement: validation and clinical application studies. Hypertension 26, Brodmann M, Lischnig U, Lueger A, Pilger E, Stark G (2003). The effect of caffeine on peripheral vascular resistance in isolated perfused guinea pig hind limbs. J Cardiovasc Pharmacol 42, Corti R, Binggeli C, Sudano I, Spieker L, Hanseler E, Ruschitzka F et al. (2002). Coffee acutely increases sympathetic nerve activity and blood pressure independently of caffeine content: role of habitual versus nonhabitual drinking. Circulation 106, Gallagher D, Adji A, O Rourke MF (2004). Validation of the transfer function technique for generating central from peripheral upper limb pressure waveform. Am J Hypertens 17, Happonen P, Voutilainen S, Salonen JT (2004). Coffee drinking is dose-dependently related to the risk of acute coronary events in middle-aged men. J Nutr 134, Hartley TR, Sung BH, Pincomb GA, Whitsett TL, Wilson MF, Lovallo WR (2000). Hypertension risk status and effect of caffeine on blood pressure. Hypertension 36, Jee SH, He J, Whelton PK, Suh I, Klag MJ (1999). The effect of chronic coffee drinking on blood pressure. A meta-analysis of controlled clinical trials. Hypertension 33, Kampus P, Kals J, Ristimae T, Fischer K, Zilmer M, Teesalu R (2004). High sensitivity C-reactive protein affects central haemodynamics and augmentation index in apparently healthy persons. J Hypertens 22, Karatzis E, Papaioannou TG, Aznaouridis K, Karatzi K, Stamatelopoulos K, Zampelas A et al. (2005). Acute effects of caffeine on blood pressure and wave reflections in healthy subjects: should we consider monitoring central blood pressure? Int J Cardiol 98, Katsouyanni K, Rimm EB, Gnardellis C, Trichopoulos D, Polychronopoulos E, Trichopoulou A (1997). Reproducibility and relative validity of an extensive semi-quantitative food frequency questionnaire using dietary records and biochemical markers among Greek schoolteachers. Int J Epidemiol 26, S118 S127. Klag MJ, Wang NY, Meoni LA, Brancati FL, Cooper LA, Liang KY et al. (2002). Coffee intake and risk of hypertension. The Johns Hopkins Precursors Study. Arch Intern Med 162, Kleemola P, Jousilahti P, Pietinen P, Vartiainen E, Tuomilehto J (2000). Coffee consumption and the risk of coronary heart disease and death. Arch Intern Med 160, Kullo I, Seward J, Bailey K, Bielak L, Grossardt B, Sheedy II P et al. (2005). C-reactive protein is related to arterial wave reflection and stiffness in asymptomatic subjects from the community. Am J Hypertens 18, LaCroix AZ, Mead LA, Liang KY, Thomas CB, Pearson TA (1986). Coffee consumption and the incidence of coronary heart disease. N Engl J Med 315,

7 802 Lane JD, Pieper CF, Phillips-Bute BG, Bryant JE, Kuhn CM (2002). Caffeine affects cardiovascular and neuroendocrine activation at work and home. Psychosom Med 64, Laurent S, Boutouyrie P, Asmar R, Gautier I, Laloux B, Guize L et al. (2001). Aortic stiffness is an independent predictor of all-cause and cardiovascular mortality in hypertensive patients. Hypertension 37, Lelo A, Miners JO, Robson R, Birkett DJ (1986). Assessment of caffeine exposure: caffeine content of beverages, caffeine intake, and plasma concentrations of methylxanthines. Clin Pharmacol Ther 39, London GM, Blacher J, Pannier B, Guerin A, Marchais S, Safar M (2001). Arterial wave reflections and survival in end-stage renal failure. Hypertension 38, Lopez-Garcia E, van Dam RM, Willett WC, Rimm EB, Manson JE, Stampfer MJ et al. (2006). Coffee consumption and coronary heart disease in men and women: a prospective cohort study. Circulation 113, Lovallo WR, Wilson MF, Vincent AS, Sung BH, McKey BS, Whitsett TL (2004). Blood pressure response to caffeine shows incomplete tolerance after short-term regular consumption. Hypertension 43, Mahmud A, Feely J (2001). Acute effect of caffeine on arterial stiffness and aortic pressure waveform. Hypertension 38, Mahmud A, Feely J (2005). Arterial stiffness is related to systemic inflammation in essential hypertension. Hypertension 46, McCusker RR, Goldberger BA, Cone EJ (2003). Caffeine content of specialty coffees. J Anal Toxicol 27, Nichols WW, O Rourke MF (eds) (1998). McDonald s Blood Flow in Arteries, 4th edn Edward Arnold: London, pp , , Noordzij M, Uiterwaal C, Arends L, Kok F, Grobbee D, Geleijnse J (2005). Blood pressure response to chronic intake of coffee and caffeine: a meta-analysis of randomized controlled trials. J Hypertens 23, O Rourke M, Staessen J, Vlachopoulos C, Duprez D, Plante G (2002). Clinical applications of arterial stiffness; Definitions and reference values. Am J Hypertens 15, Panagiotakos DB, Pitsavos C, Chrysohoou C, Kokkinos P, Toutouzas P, Stefanadis C (2003). The J-shaped effect of coffee consumption on the risk of developing acute coronary syndromes: the CARDIO2000 case-control study. J Nutr 133, Pauca A, O Rourke M, Kon N (2001). Prospective evaluation of a method for estimating ascending aortic pressure from the radial artery pressure waveform. Hypertension 38, Pincomb GA, Lovallo WR, McKey BS, Sung BH, Passey RB, Everson SA et al. (1996). Acute blood pressure elevations with caffeine in men with borderline systemic hypertension. Am J Cardiol 77, Rakic V, Burke V, Beilin LJ (1999). Effects of coffee on ambulatory blood pressure in older men and women: a randomized controlled trial. Hypertension 33, Robertson D, Frolich JC, Carr RK, Watson JT, Hollifield JW, Shand DG et al. (1978). Effects of caffeine on plasma renin activity, catecholamines and blood pressure. N Engl J Med 298, Robertson D, Wade D, Workman R, Woosley RL, Oates JA (1981). Tolerance to the humoral and hemodynamic effects of caffeine in man. J Clin Invest 67, Saboury AA, Divsalar A, Ataie G, Amanlou M, Moosavi-Movahedi AA, Hakimelahi GH (2003). Inhibition study of adenosine deaminase by caffeine using spectroscopy and isothermal titration calorimetry. Acta Biochim Pol 50, Savoca MR, MacKey L, Evans CD, Wilson M, Ludwig DA, Harshfield GA (2005). Association of ambulatory blood pressure and dietary caffeine in adolescents. Am J Hypertens 18, Smits P, Boekema P, De Abreu R, Thien T, van t Laar A (1987). Evidence for the antagonism between caffeine and adenosine in the human cardiovascular system. J Cardiovasc Pharmacol 10, Stefanadis C, Dernellis J, Tsiamis E, Stratos C, Diamantopoulos L, Michaelides A et al. (2000). Aortic stiffness is a risk factor for recurrent acute coronary events in patients with ischaemic heart disease. Eur Heart J 21, Stefanadis C, Dernellis J, Vlachopoulos C, Tsioufis C, Tsiamis E, Toutouzas K et al. (1997). Aortic function in arterial hypertension determined by pressure diameter relation: effects of diltiazem. Circulation 96, Stewart AD, Millasseau SC, Kearney MT, Ritter JM, Chowienczyk PJ (2003). Effects of inhibition of basal nitric oxide synthesis on carotid-femoral pulse wave velocity and augmentation index in humans. Hypertension 42, Sutton-Tyrrell K, Najjar S, Boudreau R, Venkitachalam L, Kupelian V, Simonsick EM et al. (2005). Elevated aortic pulse wave velocity, a marker of arterial stiffness, predicts cardiovascular events in wellfunctioning older adults. Circulation 111, Varani K, Portaluppi F, Merighi S, Ongini E, Belardinelli L, Borea PA (1999). Caffeine alters A2A adenosine receptors and their function in human platelets. Circulation 99, Vlachopoulos C, Alexopoulos N, Panagiotakos D, O Rourke M, Stefanadis C (2004b). Cigar smoking has an acute detrimental effect on arterial stiffness. Am J Hypertens 17, Vlachopoulos C, Dima I, Aznaouridis K, Vassiliadou K, Ioakeimidis N, Aggeli C et al. (2005b). Acute systemic inflammation increases stiffness of large arteries. Circulation 112, Vlachopoulos C, Hirata K, O Rourke M (2001). Pressure-altering agents affect central aortic pressures more than is apparent from upper limb measurements in hypertensive patients: the role of arterial wave reflections. Hypertension 38, Vlachopoulos C, Hirata K, O Rourke M (2003b). Effect of caffeine on aortic elastic properties and wave reflection. J Hypertens 21, Vlachopoulos C, Hirata K, Stefanadis C, Toutouzas P, O Rourke MF (2003a). Caffeine increases aortic stiffness in hypertensive patients. Am J Hypertens 16, Vlachopoulos C, Kosmopoulou F, Panagiotakos D, Ioakeimidis N, Alexopoulos N, Pitsavos C et al. (2004a). Smoking and caffeine have a synergistic detrimental effect on aortic stiffness and wave reflections. J Am Coll Cardiol 44, Vlachopoulos C, O Rourke M (2000). Genesis of the normal and abnormal arterial pulse. Curr Probl Cardiol 25, Vlachopoulos C, Panagiotakos D, Ioakeimidis N, Dima I, Stefanadis C (2005a). Chronic coffee consumption has a detrimental effect on aortic stiffness and wave reflections. Am J Clin Nutr 81, Weber T, Auer J, O Rourke MF, Kvas E, Lassnig E, Berent R et al. (2004). Arterial stiffness, wave reflections, and the risk of coronary artery disease. Circulation 109, Wilkinson IB, MacCallum H, Hupperetz PC, van Thoor CJ, Cockcroft JR, Webb DJ (2001). Changes in the derived central pressure waveform and pulse pressure in response to angiotensin II and noradrenaline in man. J Physiol 530, Zampelas A, Panagiotakos DB, Pitsavos C, Chrysohoou C, Stefanadis C (2004). Associations between coffee consumption and inflammatory markers in healthy persons: the ATTICA study. Am J Clin Nutr 80,