effect of coffee on blood pressure and cardiovascular disease in hypertensive individuals: a systematic review and meta-analysis

The American Journal of Clinical Nutrition, Oct 2011

Mesas, Arthur Eumann, Leon-Muñoz, Luz M, Rodriguez-Artalejo, Fernando, Lopez-Garcia, Esther

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:


effect of coffee on blood pressure and cardiovascular disease in hypertensive individuals: a systematic review and meta-analysis

The effect of coffee on blood pressure and cardiovascular disease in hypertensive individuals: a systematic review and meta-analysis1-3 Arthur Eumann Mesas Luz M Leon-Mun˜oz Fernando Rodriguez-Artalejo Esther Lopez-Garcia Background: The effect of coffee and caffeine on blood pressure (BP) and cardiovascular disease (CVD) in hypertensive persons is uncertain. Objective: The objective was to summarize the evidence on the acute and longer-term effects of caffeine and coffee intake on BP and on the association between habitual coffee consumption and risk of CVD in hypertensive individuals. Design: A systematic review and meta-analysis of publications identified in a PubMed and EMBASE search up to 30 April 2011 was undertaken. Data were extracted from controlled trials on the effect of caffeine or coffee intake on BP change and from cohort studies on the association between habitual coffee consumption and CVD. Results: In 5 trials, the administration of 200-300 mg caffeine produced a mean increase of 8.1 mm Hg (95% CI: 5.7, 10.6 mm Hg) in systolic BP and of 5.7 mm Hg (95% CI: 4.1, 7.4 mm Hg) in diastolic BP. The increase in BP was observed in the first hour after caffeine intake and lasted 3 h. In 3 studies of the longer-term effect (2 wk) of coffee, no increase in BP was observed after coffee was compared with a caffeine-free diet or was compared with decaffeinated coffee. Last, 7 cohort studies found no evidence of an association between habitual coffee consumption and a higher risk of CVD. Conclusions: In hypertensive individuals, caffeine intake produces an acute increase in BP for 3 h. However, current evidence does not support an association between longer-term coffee consumption and increased BP or between habitual coffee consumption and an increased risk of CVD in hypertensive subjects. Am J Clin Nutr 2011;94:1113-26. INTRODUCTION The association between coffee consumption and BP4 among normotensive individuals has been widely investigated. In a review of controlled trials, Nurminen et al ( 1 ) reported an acute increase in both SBP and DBP in the hours after the intake of caffeine; however, results on the effect of caffeine intake during 7 d were inconclusive. Also in a meta-analysis of 11 controlled trials that met many quality criteria, Jee et al ( 2 ) reported that coffee consumption for.1 d led to a slight increase in BP. Similarly, in a subsequent meta-analysis of randomized controlled trials of coffee or caffeine intake for 7 d, Noordzij et al ( 3 ) concluded that both interventions raised BP, although the effect of coffee was much smaller than that of caffeine. In contrast, cohort studies have found an association of habitual coffee consumption with either an increased risk of hypertension ( 4–7 ) or a decreased risk ( 8, 9 ). Last, although recent coffee consumption may transiently increase the risk of acute myocardial infarction ( 10 ) and stroke ( 11 ), particularly among infrequent drinkers, longitudinal studies suggest that habitual coffee consumption does not increase the long-term risk of CVD ( 12, 13 ). Unfortunately, the effect of coffee and caffeine on BP and CVD among individuals with hypertension is also uncertain. This is a relevant issue because, in these patients, even a slight increase in BP may raise it above levels considered safe. Moreover, coffee consumption might reduce the effectiveness of drug treatments for hypertension. Thus, this information would allow for an evidence-based clinical recommendation and for improving BP control in hypertensive patients. Of note, however, is that the most widely disseminated clinical guidelines on hypertension management do not comment on coffee consumption in the lifestyle recommendations ( 14, 15 ). Last, to our knowledge, no comprehensive and systematic overview of these topics has been published recently. Thus, this article has systematically reviewed the studies on the acute and longer-term effects of coffee and caffeine intake on BP and on the association between habitual coffee consumption and risk of CVD among hypertensive individuals. METHODS This review has followed the recommendations of the PRISMA statement ( 16 ) and the MOOSE statement ( 17 ), as appropriate. 1 From the Department of Preventive Medicine and Public Health, School of Medicine, Universidad Auto´noma de Madrid/IdiPAZ, and CIBER of Epidemiology and Public Health, Madrid, Spain. 2 Supported in part by “Fondo de Investigacio´n Sanitaria” research grant 09/00104. EL-G’s research was supported by a “Ramo´n y Cajal” contract. AEM was supported by a MAEC-AECID fellowship. 3 Address correspondence to E Lopez-Garcia, Department of Preventive Medicine and Public Health, School of Medicine, Universidad Auto´noma de Madrid, Avda Arzobispo Morcillo 4, 28029 Madrid, Spain. E-mail: esther. . 4 Abbreviations used: BP, blood pressure; CGA, chlorogenic acid; CVD, cardiovascular disease; DBP, diastolic blood pressure; HHQ, hydroxyhydroquinone; SBP, systolic blood pressure. Received March 23, 2011. Accepted for publication July 26, 2011. First published online August 31, 2011; doi: 10.3945/ajcn.111.016667. Am J Clin Nutr 2011;94:1113–26. Printed in USA. 2011 American Society for Nutrition Literature search A PubMed (http://www.ncbi.nlm.nih.gov/pubmed) and EMBASE (http://www.embase.com) search up to 30 April 2011 was conducted by using the key terms (words in the title or abstract of the manuscript) “coffee,” “caffeine,” “hypertension,” “hypertensive,” “hypertensives,” “high blood pressure,” “blood pressure,” and “mortality” and the CVD-related terms “ischemic heart disease,” “myocardial infarction,” “angina pectoris,” “arrhythmia,” “atrial fibrillation,” “stroke,” “heart failure,” “coronary heart disease,” “coronary bypass,” and “coronary angioplasty.” The complete PubMed and EMBASE searches are shown elsewhere (see Supplemental File 1 under “Supplemental data” in the online issue). Reference lists in articles retrieved from the electronic search were also scanned. Study selection and data extraction Two investigators (AEM and EL-G) independently selected the studies and extracted the data, and discrepancies were resolved by consensus. To examine the effect of coffee or caffeine consumption on change in BP, we selected controlled clinical trials, either randomized or nonrandomized, in which participants had an SBP 140 mm Hg and/or a DBP 90 mm Hg ( 18–25 ) or who were identified as having mild (26) or mild-to-moderate hypertension ( 27, 28 ). Trials with a co-intervention on other lifestyles factors, such as smoking or physical activity, were excluded. As was done in a previous review (3), the studies were classified into 2 groups, according to whether duration of coffee or caffeine intake was ,1 wk (indicating an acute effects) or 1 wk (indicating longer-term effects). To explore the association between habitual coffee consumption and risk of CVD in hypertensive individuals, we selected cohort studies in which the outcome variables were CVD incidence or mortality. Both cohort studies of hypertensive individuals and larger cohorts that presented results in a subgroup of hypertensive individuals were included. Quality assessment The clinical trials were assessed to determine whether they were randomized, whether the intervention was blinded, whether losses to follow-up were identified, and whether each BP reading was made at least twice. The cohort studies were assessed to determine whether the diagnosis of hypertension was based on self-reports or measurements, whether repeated measurements of coffee consumption were made during follow-up, whether losses to follow-up were identified, how morbidity and mortality of CVD were measured or validated, and the degree of adjustment for potential confounders, particularly other predictors of CVD risk and antihypertensive drugs. Synthesis of the data In each clinical trial, the net effect of coffee/caffeine on BP was calculated as the difference in mean SBP and DBP change between the intervention and control groups. In crossover designs, the effect of coffee/caffeine was calculated as the difference in BP between the intervention and control periods. In the meta-analysis of clinical trials on the acute effects of coffee/caffeine on BP, each period of time with BP measurements after the administration of coffee/caffeine was considered as an independent stratum (,60, 60 to ,120, and 120 to 180 min). When a study had more than one measurement in the same period of observation, we used the mean of those measurements. For example, in the study by Freestone and Ramsay ( 26 ), 3 strata were considered, because measures after caffeine intake were available for the first (mean BP change at 5, 15, and 30 min), second (BP change at 60 min), and third (BP change at 120 min) periods of observation in both the intervention and control groups. However, protocols a and b in the study by Potter et al ( 21 ) were considered as independent studies because they considered different times of caffeine abstinence before the intervention and included a different number of participants. Because none of the studies selected presented the SDi of the mean BP in each stratum, it was estimated by using the following formula: SDi ¼ SEMi 3 ðniÞ1=2 where SEMi is the SEM in each stratum and ni is the size of each stratum. In studies without information on SEM or other sources to calculate the SD ( 19, 26 ), the SEM was imputed in a standardized way by using the prognostic method (29) with the following formula: 2 3 R 3 SDbaseline 3 SDfinal where R is the correlation coefficient ( 30 ). Because none of the studies presented information on individuals to obtain R, the most conservative estimate was assumed, a minimum correlation of 0.50. This value was used in a previous meta-analysis of the effect of coffee on BP ( 3 ). Heterogeneity among studies was quantified by using the I2 statistic ( 31 ). To pool the results of trials of the acute effects of caffeine on BP, fixed-effects models were used when heterogeneity was low or absent (I2 , 20%); when heterogeneity was greater and it was considered appropriate to pool the results, random-effects models were used. A sensitivity analysis was performed by repeating the analyses after the studies that had the largest effect on the overall result were excluded. In addition, the analyses were stratified by caffeine dose (200, 250, or .250–300 mg), time of abstinence from caffeine before beginning the trial (9, 12, or 48 h), and antihypertensive treatment (yes or no) to examine the influence of these variables on the effect of caffeine on BP. Differences in results across strata were tested by using meta-regression models. Possible publication bias was explored with Begg’s funnel plots ( 32 ) and with Egger’s regression asymmetry test ( 33 ) in studies with a sufficient number of observations to allow for a meaningful analysis. Data were analyzed by using Stata software version 10.0 ( 34 ). Clinical trials of the longer-term effects of coffee consumption on BP were grouped according to the type of comparison examined: coffee compared with caffeine-free diet, coffee compared with decaffeinated coffee, and comparisons of coffees with different amounts of CGA and HHQ. Because of the small number of available studies and their heterogeneity, it was not considered advisable to pool their results by using meta-analytic techniques. RESULTS The flow of articles from the initial search to final inclusion in this systematic review is shown in Figure 1. The search of PubMed, EMBASE, and reference lists identified a total of 438 articles, 330 of which were excluded because their title or abstract was not related to the study associations. Of the 108 articles retrieved for critical review, 86 referred to the effect of coffee/caffeine on BP. Of these, 75 were excluded because they did not provide data among hypertensive individuals, were literature reviews, or had other specific criteria for exclusion detailed elsewhere (see Supplemental File 2 under “Supplemental data” in the online issue). Accordingly, 5 studies on the acute effects and 6 studies on the longer-term effects of coffee/caffeine on BP were finally included. In addition, 40 studies were identified on the effects of coffee on CVD incidence or mortality. After the review articles were excluded (ie, those that did not provide specific data on hypertensive individuals and those with insufficient data), the review was based on 7 cohort studies. Acute effects of caffeine on BP in hypertensive individuals The characteristics of the 5 studies selected are shown in Table 1. All were conducted in English-speaking countries and included persons between 20 and 82 y of age who were habitual coffee drinkers. All of these studies had a crossover design. Study quality was heterogeneous. Three studies were randomized ( 18, 21, 27 ), and all of the studies were double-blind except for the oldest study ( 26 ). The BP measurements were based on 2 readings in 3 studies, but Freestone and Ramsay ( 26 ) and Vlachopoulos et al ( 27 ) did not report this information. No losses to follow-up were reported, although in the study by Potter et al ( 21 ) one person did not participate in the second protocol because of illness. The BP cutoffs to define hypertension varied among studies and were not reported in 2 studies (26, ev 4 i s d n e e p tr p o ) e t e p s r y o h y f ti p e n ra b a ( e k e o th w s g H m m a e t 5 9 / 0 T 6 H 1 d l i s 2 r r d o e / k s e n p e i u ff rd c o F F / / 5 2 6 8 – – 0 5 3 6 6 1 ; 8 7 : b : a X , C , X D , , m m o o d d g g n n i i K K d d e e t t i i n n – – 0 0 9 g 9 / / g 0 H 0 6 6 H 1 m 1 tn m f o s c i t s i r e t c a r a h h t l C a e h s / e tu l a y t t s s e f i L e g y A x e S n n g i s e 4 – A 3 N , s u t i l g le ro in m , s e r s r u te lu n e i a n ab f i i t g d r a in o e v n h e i l l a ,l ic a n c i i l c m e r h o c , io ic b h p o a n r sm ehm leec ab o w g s h ,c io e it i i h ,a sa sw lgoo racd ilta rsek rse itw im ise re t o rm o k s th d k a tr o m o e y l o n s sm om rrh ea n on h a r C ( o o o Y N N N P P M B B B M D M B M M 3 7 2 2 X , C , P B , D C , , a i n l e a d tr e s w u S A , ) 8 2 ( , ) la 2 2 t ( e l a n e t s 3 e 9 t r 9 c 9 e 9 i 9 g 1 k 1 g a g H m m 5 0 1 – 0 9 P B d / s p u c 3 R A 0 0 1 , / 9 5 1 g – H 0 tic - ,r tr , ,s ;C s tirrsaacehC l/tltifseeaeyhhL ttssau rrsseeaaeokhvyovhym ;illrffssecaecceunononhoom tiirr-ffrsseeeecepnvoyho ;liillrrrssseeeeecenonovvpg ,,lirrrrsacaeeeceavouobhnyd iii;rrrssssceeeaeeodnndodon ,tiiitltrssaceaecaeebundbdbo ltitrreacaegnpnyopn llrrsseacaeokohoovyhm itrfrsssaehonoyodnub ,iitfrsecaeounodTH ,iitilirsaaeeeedbdpypm ,,,ttillrsseekouCVD tililrfrrrseceadvyunonon tlilfrrfaeeaacnd2ohuu tltillitfrrrssaceaekodgoonow ,:iititrsssaeeokngnoypnwm ,,tiliirrrrfseceececegydnum ,,llititlltrrseeecaopoponho lrsagudobdo ,.llirrsseeaeboodp3nbuPTB ,,.;irrrssezcaedoovondRXm o a o a k d m o o a m h m o s f h c to an nd l; a N V N M M B B R R R N N tn m C ( 1 E ex /F S M 5 e 6 g – , P , 2 P s ; e e r l u b ig l a F i a n v i a n A i N d ; e e t s r a o e p s e i r d re tr a a e d h n a ic y m ud e t h s c s a i n , i D s l H o I c ; o n t o o i r s p n e tn t r re e e p f y f h i d , , a ic 1 l 1 o2 t s a i d 27). Moreover, participants in these 2 studies ( 26, 27 ) continued to take their antihypertensive medication, whereas patients in the other studies did not receive treatment. Participants abstained from coffee and caffeine consumption before the study for periods ranging from 9 to 48 h. The meta-analysis of the acute effects of caffeine on SBP and DBP, respectively, grouped according to the time of observation (,60, 60 to ,120, and 120 to 180 min) is shown in Figures 2 and 3. Certain assumptions were made in the meta-analysis. First, in the 2 studies in which caffeine doses were adjusted for body weight ( 18, 19 ), caffeine intake (mg) was estimated as the product of the dose/kg times the mean reported weight (19). Because weight was not reported by Pincomb et al ( 18 ), it was assumed to be 82 kg (mean value in white men aged 20–39 y in the United States) ( 35 ). Second, in the study by Potter et al ( 21 ), we used BP measured in the standing rather than in the supine position, because this is more frequent in clinical practice. In all trials, 200–300 mg caffeine was administered (equivalent to ;1.5–2 cups filtered coffee). The overall change in SBP associated with caffeine over all studies and observation periods was 8.14 mm Hg (95% CI: 5.68, 10.61 mm Hg). SBP increased by 7.43 mm Hg in the first 60 min after coffee intake and remained elevated at the same level in the following periods. No heterogeneity among studies was detected in SBP elevation, either in each time period or over time (I2 = 0.0%) (Figure 2). Given the large contribution of the study by Pincomb et al ( 18 ) to the overall results (Figure 2), the analyses were repeated after this study was excluded; similar results were found. Moreover, in the stratified analyses, the effect of caffeine on SBP did not vary with caffeine dose in the range of 200 to 300 mg (Table 2). The increase in SBP due to caffeine was similar regardless of whether or not participants were taking antihypertensive medication (Table 2). For DBP, the overall change associated with caffeine intake was 5.75 mm Hg (95% CI: 4.09, 7.41 mm Hg). As for SBP, the increase in DBP was observed in the first 60 min and up to 3 h after caffeine intake (Figure 3). These results were obtained with random-effects models because some heterogeneity was observed in the results for the period 120–180 min (I2 = 36.1%). The heterogeneity is due to the larger increase in DBP observed in the study by Sung et al ( 19 ), which used higher doses of caffeine (300 mg) (Figure 3). The stratified and meta-regression analyses in Table 2 also suggest that DBP increased more in those who received .250 mg caffeine than in those who received smaller doses. The increase in DBP produced by caffeine did not vary with time of abstinence from caffeine before beginning the trial or with use of antihypertensive treatment (Table 2). The Begg’s funnel plots constructed with the 9 observations on the effect of 250 mg caffeine did not suggest publication bias for the results on SBP or DBP. Moreover, the Egger’s test P value was 0.411 for SBP and 0.583 for DBP. Thus, the results of the Egger’s test did not support the existence of publication bias for the effect of 250 mg caffeine. Unfortunately, we had only 3 observations on the effect of 200 mg caffeine and 4 observations on the effect of .250–300 mg caffeine. Accordingly, a formal assessment of publications bias was not considered advisable for these caffeine doses. Longer-term effects of coffee consumption on BP in hypertensive subjects The characteristics of the 6 studies selected are shown in Table 1. The studies were conducted in different countries ( 20, 22, 28 ), with the 3 most recent studies carried out in Japan ( 23–25 ). All included persons of both sexes were aged 20 y. Four studies had a parallel design ( 22–25 ) and 2 were crossover studies ( 20, 28 ). The quality of the studies was high. All were randomized and double-blind, except for the study by Rakic et al ( 22 ), which did not report whether the intervention was blinded. Loss to follow-up was ,10% in all studies, and in the study of Rakic et al ( 22 ) there were no losses. The BP values were based on the mean of 24-h ambulatory blood pressure monitoring in 3 studies ( 20, 22, 28 ), and on 2 BP readings in the remaining studies ( 23–25 ). The cutoff points to define hypertension were heterogeneous, and the oldest study ( 20 ) considered only the DBP value. The study subjects continued with antihypertensive medication in 2 studies ( 22, 28 ), and they abstained from coffee and caffeine during 1 or 2 wk beforehand in 4 studies ( 22, 24, 28 ). The longer-term effects of coffee intake on BP are shown in Table 3. Two of the studies ( 20, 28 ) presented BP data only at the end of the intervention and control periods. Thus, in the study by MacDonald et al (20), baseline BP was assumed to be the same as BP at the end of the period with the normal diet. In the study by Eggertsen et al ( 28 ), the effect of caffeine on BP was estimated by comparing only the final values of BP between the treatment branches. The duration of the intervention varied from 2 to 12 wk. No increase in BP was observed when coffee intake was compared with the caffeine-free diet and when coffee intake was compared with the decaffeinated coffee. Three studies examined the effect of drinking coffee containing different amounts of HHQ and/or CGA on BP. Chikama et al ( 23 ) compared the effect of coffee with reduced HHQ with that of normal coffee. In comparison with normal coffee, the reduced-HHQ coffee resulted in a decrease in BP after 4–12 wk of consumption. In the second study, Yamaguchi et al ( 24 ) studied the effect of coffee without HHQ and with different 1 The net change in blood pressure is the difference in blood pressure between the intervention and the control periods. Fixed-effects models were used for systolic blood pressure and random-effects models for diastolic blood pressure metaanalysis. 2 Significance of the difference between the categories of the stratification variables, obtained from meta-regression analyses. concentrations of CGA in comparison with normal coffee. The coffee without HHQ decreased BP, although the results did not show a clear dose-response with the concentration of CGA ( 24 ). Last, Ochiai et al ( 25 ) observed that, in comparison with a cup of coffee with a reduced concentration of HHQ and no CGA, a cup of coffee with a reduced concentration of HHQ but elevated concentration of CGA seemed to decrease SBP but did not modify DBP after 4, 6, and 8 wk of intervention. Habitual coffee consumption and risk of CVD in hypertensive subjects Five cohort studies ( 13, 36–39 ) were conducted in the United States, one in Finland ( 40 ), and the other in Sweden ( 41 ). Except for one study designed as a cohort of hypertensive individuals ( 36 ), the rest were large cohort studies in different populations with no clinical inclusion criterion, in which a subcohort developed hypertension during follow-up. Three cohorts included persons aged 30 y ( 13, 36, 38 ), and the others included those older than 48 y ( 37, 40, 41 ) and 64 y ( 39 ). The study quality was good. In 5 studies, hypertension was defined on the basis of measured BP values ( 36–40 ). Two studies were based on selfreported history of hypertension ( 41 ) or on a diagnosis of hypertension reported by nurses and were therefore very reliable ( 13 ). Habitual consumption of coffee was reported at the beginning of follow-up and in one study ( 13 ) was updated every 4 y. In all studies, information on CVD was confirmed by review of the medical record or death certificate. Length of follow-up varied from 4 to 25 y. Finally, in all cases, the results were adjusted for the main confounding factors. The studies included in this section examined different endpoints (Table 4). The 3 studies on CVD mortality found no excess risk of death associated with coffee or caffeinated drinks in hypertensive individuals ( 36, 38, 39 ). In contrast, the studies on coffee consumption and risk of stroke in hypertensive individuals yielded inconsistent results. Hakim et al ( 37 ) found that those who consumed 20 oz coffee/d (600 mL coffee/d), ;4 cups, had double the risk of thromboembolic stroke than did noncoffee drinkers; however, Lopez-Garc´ıa et al ( 13 ) and Larsson et al ( 41 ) found no excess risk, even in those with higher consumption ( 4 cups/d). Finally, Larsson et al ( 40 ) observed a lower risk of cerebral infarction in those who consumed 6 cups/d than in those with lower consumption (,2 cups/d). Only one study examined the effect of coffee on heart valve disease, and no association was found ( 39 ). DISCUSSION This review has shown that, in hypertensive patients, caffeine intake of 200 to 300 mg produced an important increase in BP, which was observed in the first 60 min after intake and persisted up to 180 min afterward. In contrast, drinking coffee for 2 wk did not appear to increase BP. These results are based on high-quality clinical trials. Finally, the cohort studies reviewed do not support an association between habitual coffee consumption and a higher risk of CVD in hypertensive individuals. These results are unique because they extend the evidence of the effect of coffee on BP to hypertensive individuals. The most comprehensive review to date on the acute effects of caffeine in normotensive individuals reported BP elevations from 2 to 12 mm Hg for SBP and from 3 to 11 mm Hg for DBP (the results of the studies were not pooled statistically) ( 1 ). In our study, hypertensive individuals who received caffeine showed an overall increase of 8 mm Hg in SBP and of nearly 6 mm Hg in DBP. Thus, caffeine has a comparable effect on normotensive and hypertensive individuals. In considering the longer-term effects of coffee consumption on normotensive individuals, 2 previous reviews showed elevated BP: 2.4 mm Hg in SBP and 1.2 mm Hg in DBP in the study by Jee et al ( 2 ) and 2.0 mm Hg in SBP and 0.7 mm Hg in DBP in the study by Noordzij et al ( 3 ). In our study, long-term coffee consumption did not show a clear increase in BP in hypertensive ) 1 . 2 0 2 2 2 2 ( ( ( ( 2 2 2 ( ( ( 2 2 2 2 4 8 0 2 1 1 ) 7 . 0 , 5 . 5 h t i ) 9 . 0 , 1 . 7 h t i n I , 1 . 6 h t i n I 7 7 n i h t i 2 2 2 T L R C C R 0 7 5 2 0 3 5 2 P tu B ee b D ff P fy co P a t o tu B m n n i e h t d n a s l a i r t l e l l a r a p r o f p u o r g l o r t n o c e h t d n a p u o r g d e t a e r t e h t w t e b s e g n a h c P B n i n o i th H t i H n e w ; v e r ls r e t n i , , e c P c , n b B e , D r e f f i d o u c s o s t e o rp r p d e o t ac lo i b d c n i i l o d t s d a n i a d a ; , d y i d c u a t s c i s n ’ l e a g t o e r o i l h h c c gu , a A m G a C Y ; In re . u s d s io e r r e p m re o e c f f t i n d e r h fe it d s h l t o i c o w t s o l r o p c o te t o a r c p i d te in a c d i if w n e e d d e in an c n b e c . r d , e e aan ,ab rssu .se iffd ,y ,s re c e h d e p un t u i t d d o is s l’s tsu loo r8 PB s b o tea l’a ilc L ien d t o m g l e t s a sy 240 achn .s an i o i cD ch ,P = t l a en ir a O SB cup eh tre InM and l; e 1 T v b 1 l’s la 2 3 so i s a a o t v r e a c s n e iso ffe no PB lu co it d c p e n ee m sa o r u e C -f s r n c Q o e H c d 3 e ) 2 g I , an C .9 h 2 c % C e b f ts o ce 3 . j 8 o b 1 N su ) C ( 3 E L P tu B b D P fy B i n Q A io S do H G t d m p e H C m sa to d [ u e n e h s r c t u i on ec id d w c d d e h t i o l r o g tr e n v o ti c c ( (a 9 5 9 7 , 3 . 5 2 ( h t i o l r o g tr e n v o it c c ( (a 9 5 9 7 h t i h t i 8 9 9 1 , ) im 7 ak (3 e itopn tih fok ilco raevg ton tih tr iltay m w is b e n w ea tr r C ( C : s tjlititltr-rssseeaaeaududvuM tlitlttrfaayoooRm itfrrfseeacoogo ,..:tisc00821npounom .,.,.,.()(026280301560 .,..,.;())a60312809d0n281 lrrrfscaaeecuvoboRR titlitrrfrseeacagoyoom .,.:itfsc01073nopunoom .,.,.,.()(012660641730 .,..,.;())a50630403nd1441 tlittrrfaeoyohoCRRDVm ititfrrfssceecaonupoogoonm .,.,..,.)(30108166039001 .,.,.)(a08d0n321350 .,.)(391640 itfrfscoon20puonoRm P,.,..t:/)-r(ez7nd32112do= .ttiifrfr(reaonoohon6000m liltit)rsaaeaeecavbonog itfrfsconopunooRm .ti/sseacgndpu51 jt:itrssseceeebuvnpyh1 .,..;)fr(o999030260RR ./tifsscc15upodonpunom ittirsseeeeanvph2ygn ,...:tj)(ssce740114108ub itfrrfseeacoogoR tititissecohhnonpuonwm .,:g10041PSBHmm .,,...)(090017170920 .,.,.,.()(096570701670 .,..,,.())a63009367d0n590 P.;titrr-caeeego1000dn=RR tiiittfssechnonopunohowm .,.:g1008098PBHDmm .,...,,.,()()087310806016270 ,..,..)(and070291060570 P.,.,.tr-)(edn0007807150, tiltfrrfayoooCRDHm /:itfscc1updoonpuonm P,.,..t-r)(edn274140780= .tl;rfaeaevvoh840RR tiirfsssceaepuonnoodm .,./:(fc027411pdu1o P.,.tr-)e040dn527= , D l o V ,I h C M lco tla fo ev ,B ,a ira ry is g ty o n n i m t te e i v , is r s ok ti n h e u sm lca tipo ,PB ithpy tion ,ex ica um ,s an ica s sy sn tu d ed eg ph co tsa an m , , ) 9 3 e = sa e n s ( i y d t i e l v a lav )20 t r o m tr a = D e n h ( H : n io /d t s p p m cu u s 1 n o c d e n e a f f 0 o m m ) 2 0 0 0 3 1 / = 0 16 n( g H ) y 45 5 3 6 1 = , /F n( y 1 . 0 1 , s e t a t S g 0 r 2 eb ,) n 9 e 3 re ( C H C C C C C g g P P B B S B y t i l a t r o m e s ) a 7 se 14 i d t = rea n( on ith i w t supm itaed irsk c n o ed ek feeco tssa rsea trso o no icn fo f C ) d e u n it n o C ( 4 E L B A T s n o i s u l c n o C s lt u s e r d e t s u j d a itre a a v itl u M s tr o c a f g n i d n u o f n o C ) n ( e m o c t u O 2 s e ir o g e t a c /r e u s o p x E n o i s lu e c iv in sen ro t f r ia ep r y e h t irc sa P B n o i ltaup taad o P n trao -up i ,trydu llfoow uno fo C e c n e r e f e R ) 0 0 9 = n ( e k o tr S : /opm 7 ,/sdp tion cu1 ,5k– c3u /sd pum ,o c cum /scpw ,/k2–w cp4u sn / o p u ffeo 1, to4 scup and e C ) mm ,2960 0 1 /1409 n(g= P H B ) 6 6y ,307 ,eanm :ea5g n(8= F y 4 2 , s e tt a S d e ti n U irca 09 a 0 -zG ,)2 e 3 op (1 L n o tir c ifan )82 la 4 r reeb n(= C :no adn i tp ,4 snum ,2–3 /sd o – p feec ,11 c5u fo , C fo reb itrhy nnm isev tno so (u trreepod tirseeno treehypn iilsadvu tr)eod -e hyp fo ind rep lf S ) 0 7 3y ,346 ,49–8 n(= F y 4 . 0 1 , n e d e w S individuals. Finally, longitudinal studies in normotensive individuals ( 42 ), and the studies reviewed in this work in hypertensive individuals, suggest that habitual coffee consumption is not associated with a higher risk of cardiovascular events. Our work makes some original methodologic contributions to the study of the acute effect of coffee and caffeine on BP. First, changes in BP after caffeine intake, from the beginning to the end of each study, were considered independently for each period of observation. This allowed evaluation of the temporal course of the effect of caffeine on BP in the first hours after its administration. Second, we found a suggestion of a dose-response relation between acute caffeine intake and DBP. Furthermore, the increase in BP was independent of the use of antihypertensive treatment. Given the small number and the heterogeneity of the studies, it was not possible to calculate an overall estimator of the longerterm effect of coffee on BP in hypertensive individuals. However, no study found a significant increase in BP; moreover, modifying the concentrations of some components of coffee—specifically decreasing HHQ and increasing CGA—produced reductions in BP. Coffee is a drink with a very complex chemical composition. Although the cardiometabolic effects of caffeine are well documented ( 1, 43–45 ), the effects of other substances in coffee are not yet well known. Recent studies suggest that some components of coffee, such as CGA and other phenolic compounds, magnesium, and trigonelline, improve glucose metabolism and reduce inflammation and endothelial dysfunction ( 46–48 ). In the long term, these effects would be seen in habitual coffee drinkers who have developed tolerance to caffeine. Thus, the harmful cardiovascular effects of caffeine would be offset by the beneficial effects of these other components. This may help explain the lack of association between coffee and long-term CVD risk in large cohort studies, in both normotensive ( 12, 40, 49–51 ) and hypertensive ( 13, 38, 39 ) persons. The results of this review have clinical implications for the control of hypertensive patients. On the one hand, the acute increase in BP may temporarily increase the risk of a cardiovascular event ( 10, 11 ). This is consistent with the increased risk of coronary disease and stroke in the hours after coffee consumption. Thus, hypertensive patients with uncontrolled BP should avoid consuming large doses of caffeine. Furthermore, the consumption of coffee or other caffeinated substances in the hours before measuring BP may elevate it and give the erroneous impression that BP is poorly controlled. However, we found no evidence to justify avoidance of habitual coffee consumption in well-controlled hypertensive patients; therefore, in these patients, medical recommendations should prioritize modification of other lifestyles, such as quitting smoking, controlling weight, and increasing physical activity. In conclusion, caffeine raises BP for 3 h after ingestion in hypertensive persons. However, the available evidence does not support the assertion that coffee consumption for 2 wk increases BP or that habitual coffee consumption raises the risk of CVD. Nonetheless, studies of the effect of prolonged coffee consumption (.2 wk) on BP are needed (Table 3). It is also necessary to investigate directly the influence of coffee or caffeine on the degree of BP control in hypertensive individuals and its possible variation with type of antihypertensive medication (eg, renin-angiotensin-aldosterone inhibitors, calcium antagonists, and b-blockers). Such evidence would make it possible to fine tune recommendations about coffee consumption in hypertensive individuals. Last, given that the studies on the association between coffee and risk of stroke found inconsistent results, further research is needed before we can rule out a deleterious effect of coffee consumption on each type of CVD. We thank Mercedes Corrales for designing the PubMed and EMBASE searches. The authors’ responsibilities were as follows—AEM, LML-M, FR-A, and EL-G: designed the study; AEM and EL-G: analyzed the data; AEM, LML-M, FR-A, and EL-G: wrote the manuscript; and FR-A and EL-G: had primary responsibility for the final content. All authors read and approved the final manuscript. The funders had no role in the design, implementation, analysis, or interpretation of the data. None of the authors had a conflict of interest. 1. Nurminen ML , Niittynen L , Korpela R , Vapaatalo H . Coffee, caffeine and blood pressure: a critical review . Eur J Clin Nutr 1999 ; 53 : 831 - 9 . 2. Jee SH , He J , Whelton PK , Suh I , Klag MJ . The effect of chronic coffee drinking on blood pressure: a meta-analysis of controlled clinical trials . Hypertension 1999 ; 33 : 647 - 52 . 3. Noordzij M , Uiterwaal CS , Arends LR , Kok FJ , Grobbee DE , Geleijnse JM . Blood pressure response to chronic intake of coffee and caffeine: a meta-analysis of randomized controlled trials . J Hypertens 2005 ; 23 : 921 - 8 . 4. Uiterwaal CS , Verschuren WM , Bueno-de-Mesquita HB , Ocke´ M, Geleijnse JM , Boshuizen HC , Peeters PH , Feskens EJ , Grobbee DE . Coffee intake and incidence of hypertension . Am J Clin Nutr 2007 ; 85 : 718 - 23 . 5. Hu G , Jousilahti P , Nissinen A , Bidel S , Antikainen R , Tuomilehto J . Coffee consumption and the incidence of antihypertensive drug treatment in Finnish men and women . Am J Clin Nutr 2007 ; 86 : 457 - 64 . 6. Palatini P , Dorigatti F , Santonastaso M , Cozzio S , Biasion T , Garavelli G , Pessina AC , Mos L . Association between coffee consumption and risk of hypertension . Ann Med 2007 ; 39 : 545 - 53 . 7. Zhang Z , Hu G , Caballero B , Appel L , Chen L . Habitual coffee consumption and risk of hypertension: a systematic review and metaanalysis of prospective observational studies . Am J Clin Nutr 2011 ; 93 : 1212 - 9 . 8. Klag MJ , Wang NY , Meoni LA , Brancati FL , Cooper LA , Liang KY , Young JH , Ford DE . Coffee intake and risk of hypertension: the Johns Hopkins precursors study . Arch Intern Med 2002 ; 162 : 657 - 62 . 9. Winkelmayer WC , Stampfer MJ , Willett WC , Curhan GC . Habitual caffeine intake and the risk of hypertension in women . JAMA 2005 ; 294 : 2330 - 5 . 10. Baylin A , Hernandez-Diaz S , Kabagambe EK , Siles X , Campos H . Transient exposure to coffee as a trigger of a first nonfatal myocardial infarction . Epidemiology 2006 ; 17 : 506 - 11 . 11. Mostofsky E , Schlaug G , Mukamal KJ , Rosamond WD , Mittleman MA . Coffee and acute ischemic stroke onset: the Stroke Onset Study . Neurology 2010 ; 75 : 1583 - 8 . 12. Lopez-Garcia E , van Dam RM , Willett WC , Rimm EB , Manson JE , Stampfer MJ , Rexrode KM , Hu FB . Coffee consumption and coronary heart disease in men and women: a prospective cohort study . Circulation 2006 ; 113 : 2045 - 53 . 13. Lopez-Garcia E , Rodriguez-Artalejo F , Rexrode KM , Logroscino G , Hu FB , van Dam RM. Coffee consumption and risk of stroke in women . Circulation 2009 ; 119 : 1116 - 23 . 14. The Seventh Report of the Joint National Committee on Prevention. Detection, evaluation, and treatment of high blood pressure-complete report . Available from: http://www.nhlbi.nih.gov/guidelines/hypertension/ jnc7full. htm (cited 13 October 2010 ). 15. Mancia G , De Backer G , Dominiczak A , Cifkova R , Fagard R , Germano G , Grassi G , Heagerty AM , Kjeldsen SE , Laurent S , et al. 2007 Guidelines for the management of arterial hypertension: the Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC) . Eur Heart J 2007 ; 28 : 1462 - 536 . 16. Liberati A , Altman DG , Tetzlaff J , Mulrow C , Gøtzsche PC , Ioannidis JP , Clarke M , Devereaux PJ , Kleijnen J , Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration . J Clin Epidemiol 2009 ; 62 : e1 - 34 . 17. Stroup DF , Berlin JA , Morton SC , Olkin I , Williamson GD , Rennie D , Moher D , Becker BJ , Sipe TA , Thacker SB . Meta-analysis of observational studies in epidemiology: a proposal for reporting. Metaanalysis Of Observational Studies in Epidemiology (MOOSE) group . JAMA 2000 ; 283 : 2008 - 12 . 18. Pincomb GA , Lovallo WR , McKey BS , et al. Acute blood pressure elevations with caffeine in men with borderline systemic hypertension . Am J Cardiol 1996 ; 77 : 270 - 4 . 19. Sung BH , Whitsett TL , Lovallo WR , al'Absi M , Pincomb GA , Wilson MF. Prolonged increase in blood pressure by a single oral dose of caffeine in mildly hypertensive men . Am J Hypertens 1994 ; 7 : 755 - 8 . 20. MacDonald TM , Sharpe K , Fowler G , Lyons D , Freestone S , Lovell HG , Webster J , Petrie JC . Caffeine restriction: effect on mild hypertension . BMJ 1991 ; 303 : 1235 - 8 . 21. Potter JF , Haigh RA , Harper GD , Fotherby M , Hurd S , Macdonald IA . Blood pressure, plasma catecholamine and renin responses to caffeine in elderly hypertensives . J Hum Hypertens 1993 ; 7 : 273 - 8 . 22. Rakic V , Burke V , Beilin LJ . Effects of coffee on ambulatory blood pressure in older men and women: a randomized controlled trial . Hypertension 1999 ; 33 : 869 - 73 . 23. Chikama A , Yamaguchi T , Ochiai R , Kataoka K , Tokimitsu I. Effects of hydroxyhydroquinone-reduced coffee on blood pressure in highnormotensives and mild hypertensives . J Health Sci 2008 ; 54 : 162 - 73 . 24. Yamaguchi T , Chikama A , Mori K , Watanabe T , Shioya Y , Katsuragi Y , Tokimitsu I . Hydroxyhydroquinone-free coffee: a double-blind, randomized controlled dose-response study of blood pressure . Nutr Metab Cardiovasc Dis 2008 ; 18 : 408 - 14 . 25. Ochiai R , Chikama A , Kataoka K , Tokimitsu I , Maekawa Y , Ohishi M , Rakugi H , Mikami H . Effects of hydroxyhydroquinone-reduced coffee on vasoreactivity and blood pressure . Hypertens Res 2009 ; 32 : 969 - 74 . 26. Freestone S , Ramsay LE . Effect of coffee and cigarette smoking on the blood pressure of untreated and diuretic-treated hypertensive patients . Am J Med 1982 ; 73 : 348 - 53 . 27. Vlachopoulos C , Hirata K , Stefanadis C , Toutouzas P , O'Rourke MF . Caffeine increases aortic stiffness in hypertensive patients . Am J Hypertens 2003 ; 16 : 63 - 6 . 28. Eggertsen R , Andreasson A , Hedner T , Karlberg BE , Hansson L . Effect of coffee on ambulatory blood pressure in patients with treated hypertension . J Intern Med 1993 ; 233 : 351 - 5 . 29. Ma J , Liu W , Hunter A , Zhang W. Performing meta -analysis with incomplete statistical information in clinical trials . BMC Med Res Methodol 2008 ; 8 : 56 . 30. Higgins JPT , Green S , eds. Cochrane handbook for systematic reviews of interventions, version 5.0.2 . In: The Cochrane Collaboration. Chippenham, United Kingdom: Wiley, 2009 . 31. Higgins JP , Thompson SG , Deeks JJ , Altman DG . Measuring inconsistency in meta-analyses . BMJ 2003 ; 327 : 557 - 60 . 32. Begg CB , Mazumdar M. Operating characteristics of a rank correlation test for publication bias . Biometrics 1994 ; 50 : 1088 - 101 . 33. Egger M , Davey Smith G , Schneider M , Minder C. Bias in metaanalysis detected by a simple, graphical test . BMJ 1997 ; 315 : 629 - 34 . 34. StataCorp. Stata statistical software: release 10 . College Station, TX: StataCorp LP , 2007 . 35. Ogden CL , Fryar CD , Carroll MD , Flegal KM . Mean body weight, height, and body mass index , United States 1960-2002. Adv Data 2004; Oct 27 : 1 - 17 . 36. Martin JB , Annegers JF , Curb JD , Heyden S , Howson C , Lee ES , Lee M. Mortality patterns among hypertensives by reported level of caffeine consumption . Prev Med 1988 ; 17 : 310 - 20 . 37. Hakim AA , Ross GW , Curb JD , Rodriguez BL , Burchfiel CM , Sharp DS , Yano K , Abbott RD . Coffee consumption in hypertensive men in older middle-age and the risk of stroke: the Honolulu Heart Program . J Clin Epidemiol 1998 ; 51 : 487 - 94 . 38. Greenberg JA , Dunbar CC , Schnoll R , Kokolis R , Kokolis S , Kassotis J . Caffeinated beverage intake and the risk of heart disease mortality in the elderly: a prospective analysis . Am J Clin Nutr 2007 ; 85 : 392 - 8 . 39. Greenberg JA , Chow G , Ziegelstein RC . Caffeinated coffee consumption, cardiovascular disease, and heart valve disease in the elderly (from the Framingham Study) . Am J Cardiol 2008 ; 102 : 1502 - 8 . 40. Larsson SC , Mannisto S , Virtanen MJ , Kontto J , Albanes D , Virtamo J . Coffee and tea consumption and risk of stroke subtypes in male smokers . Stroke 2008 ; 39 : 1681 - 7 . 41. Larsson SC , Virtamo J , Wolk A . Coffee consumption and risk of stroke in women . Stroke 2011 ; 42 : 908 - 12 . 42. Wu JN , Ho SC , Zhou C , Ling WH , Chen WQ , Wang CL , Chen YM . Coffee consumption and risk of coronary heart diseases: a metaanalysis of 21 prospective cohort studies . Int J Cardiol 2009 ; 137 : 216 - 25 . 43. Thong FS , Graham TE . Caffeine-induced impairment of glucose tolerance is abolished by beta-adrenergic receptor blockade in humans . J Appl Physiol 2002 ; 92 : 2347 - 52 . 44. Hartley TR , Lovallo WR , Whitsett TL . Cardiovascular effects of caffeine in men and women . Am J Cardiol 2004 ; 93 : 1022 - 6 . 45. Verhoef P , Pasman WJ , Van Vliet T , Urgert R , Katan MB . Contribution of caffeine to the homocysteine-raising effect of coffee: a randomized controlled trial in humans . Am J Clin Nutr 2002 ; 76 : 1244 - 8 . 46. Abel ED. Myocardial insulin resistance and cardiac complications of diabetes . Curr Drug Targets Immune Endocr Metabol Disord 2005 ; 5 : 219 - 26 . 47. Lopez-Garcia E , van Dam RM , Qi L , Hu FB . Coffee consumption and markers of inflammation and endothelial dysfunction in healthy and diabetic women . Am J Clin Nutr 2006 ; 84 : 888 - 93 . 48. Go´ mez-Ruiz JA , Leake DS , Ames JM . In vitro antioxidant activity of coffee compounds and their metabolites . J Agric Food Chem 2007 ; 55 : 6962 - 9 . 49. Andersen LF , Jacobs DR Jr, Carlsen MH , Blomhoff R . Consumption of coffee is associated with reduced risk of death attributed to inflammatory and cardiovascular diseases in the Iowa Women's Health Study . Am J Clin Nutr 2006 ; 83 : 1039 - 46 . 50. Sofi F , Conti AA , Gori AM , Eliana Luisi ML , Casini A , Abbate R , Gensini GF . Coffee consumption and risk of coronary heart disease: a meta-analysis . Nutr Metab Cardiovasc Dis 2007 ; 17 : 209 - 23 . 51. Bidel S , Hu G , Qiao Q , Jousilahti P , Antikainen R , Tuomilehto J . Coffee consumption and risk of total and cardiovascular mortality among patients with type 2 diabetes . Diabetologia 2006 ; 49 : 2618 - 26 .

This is a preview of a remote PDF: https://academic.oup.com/ajcn/article-pdf/94/4/1113/23872870/1113.pdf

Mesas, Arthur Eumann, Leon-Muñoz, Luz M, Rodriguez-Artalejo, Fernando, Lopez-Garcia, Esther. effect of coffee on blood pressure and cardiovascular disease in hypertensive individuals: a systematic review and meta-analysis, The American Journal of Clinical Nutrition, 2011, 1113-1126, DOI: 10.3945/ajcn.111.016667