sign in sign up
<<
Home , Cafestol, Caffeine, Enprofylline, Kahweol

European Journal of Clinical Nutrition (1999) 53, 831±839 ß 1999 Stockton Press. All rights reserved 0954±3007/99 $15.00 http://www.stockton-press.co.uk/ejcn

Coffee, and : a critical review

M-L Nurminen1*, L Niittynen2, R Korpela2 and H Vapaatalo1

1Institute of Biomedicine, Department of Pharmacology and Toxicology, University of Helsinki, Finland; and 2Valio Ltd, Research and Development Centre, Helsinki, Finland

Objective: We review the published data relating to intake of and caffeine on blood pressure in man. We also refer to studies on the possible mechanisms of actions of these effects of caffeine. Design: The MEDLINE and Current Contents databases were searched from 1966 to April 1999 using the text words `coffee or caffeine' and `blood pressure or hypertension'. Controlled clinical and epidemiologic studies on the blood pressure effects of coffee or caffeine are reviewed. We also refer to studies on the possible mechanisms of action of these effects of caffeine. Results: Acute intake of coffee and caffeine increases blood pressure. Caffeine is probably the main active component in coffee. The pressor response is strongest in hypertensive subjects. Some studies with repeated administration of caffeine showed a persistent pressor effect, whereas in others chronic caffeine ingestion did not increase blood pressure. Epidemiologic studies have produced contradictory ®ndings regarding the association between blood pressure and coffee consumption. During regular use tolerance to the cardiovascular responses develops in some people, and therefore no systematic elevation of blood pressure in long-term and in population studies can be shown. Conclusions: We conclude that regular coffee may be harmful to some hypertension-prone subjects. The hemodynamic effects of chronic coffee and caffeine consumption have not been suf®ciently studied. The possible mechanisms of the cardiovascular effects of caffeine include the blocking of receptors and the inhibition of . Descriptors: coffee; caffeine; blood pressure; hypertension; adenosine receptors; phosphodiesterases

Introduction Benowitz, 1994). Caffeine produces a variety of adverse effects, including gastrointestinal disturbances, , Coffee is one of the most widely used non-alcoholic and (Chou & Benowitz, 1994). It can beverages in industrialized countries. Caffeine is an impor- also induce and sometimes even cardiac tant component of this : a 150 ml cup of coffee (Mehta et al, 1997), and it has been suggested contains about 60 ± 120 mg of caffeine (Barone & Roberts, that caffeine may be potentially hypertensive (James 1997; 1996). Other common dietary sources of caffeine are Jee et al, 1999). and soft . The caffeine content in tea is aproxi- Caffeine is readily absorbed from the gastrointestinal mately 20 ± 40 mg per 150 ml cup, and that of cola soft tract and rapidly distributed in all the body ¯uids (Chou & drinks ranges from 15 to 24 mg per 180 ml serving. The Benowitz, 1994). There is, however, considerable inter- consumption of coffee in Nordic countries is among the individual variability in the rate of absorption. The peak highest in the world; for instance, in Denmark the mean plasma concentrations of caffeine are usually obtained daily caffeine intake is approximately 7 mg=kg (Barone & within 30 ± 120 min of oral intake (Robertson et al, 1978; Roberts, 1996). In addition, caffeine is used as an adjuvant Smits et al, 1985, 1986a). Caffeine is extensively metabo- in many prescription and over-the-counter , e.g. in lized in the via a complex set of reactions. The combination with nonsteroidal anti-in¯ammatory drugs in primary pathway is oxidative N- to para- formulations and with in drugs for , , and direct ring oxida- treating (Chou & Benowitz, 1994). Combination tion to 1,3,7-trimethyluric acid (Tang-Liu et al, 1983). The of caffeine and has also been reported to reduce elimination half-life of caffeine in man ranges from 2 ± weight in obese subjects, whereas both substances given 10 h, with a mean of about 4 ± 5 h (Robertson et al, 1981; alone had no effect (Astrup et al, 1992). Tang-Liu et al, 1983; Smits et al, 1985). A dose-dependent Caffeine exerts a variety of stimulatory effects upon the may occur, which means that , and it is probably the most widely involved in become saturated (Tang-Liu et al, used psychoactive substance. It also increases the respira- 1983). The metabolism is slowed during (Knutti tory rate and causes bronchodilatation, stimulates , et al, 1981) and in women taking oral contraceptives and increases diuresis (Robertson et al, 1978; Chou & (Callahan et al, 1983). On the other hand, the clearance rate of caffeine is greater in smokers than in non-smokers *Correspondence: M-L Nurminen, Institute of Biomedicine, Department (Parsons & Neims, 1978). of Pharmacology and Toxicology, PO Box 8, FIN-00014 University of Coffee and other caffeine-containing beverages are Helsinki, Finland. E-mail: marja-leena.nurminen@helsinki.®. widely consumed on a daily basis. It is therefore important Guarantor: M-L Nurminen. to de®ne the possible risks and bene®ts associated with Contributors: L Niittynen wrote the chapter dealing with acute intake of caffeine intake, in order to be able to better inform coffee and caffeine, R Korpela and H Vapaatalo wrote the chapter dealing with the mechanisms and M-L Nurminen is responsible for the other parts both health professionals and the public. The possible of the manuscript. association between coffee consumption and increased Coffee, caffeine and blood pressure M-L Nurminen et al 832 risk of coronary disease has been debated for decades intake on blood pressure (Table 1). However, the results without any clear agreement of causal relation (Wilson et have remained inconsistent. Some of these studies found al, 1989; Greenland, 1993; Chou & Benowitz, 1994; that habitual consumption of coffee or caffeine was asso- Kawachi et al, 1994; Thompson, 1994). Possible risk ciated with slightly elevated blood pressure (Lang et al, factors predisposing consumers to caffeine-induced coron- 1983a, b; Birkett & Logan, 1988; LoÈwik et al, 1991; Burke ary heart disease may include blood pressure and plasma et al, 1992; Narkiewicz et al, 1995). However, many of the levels elevated by coffee. epidemiologic studies showed no relation (Shirlow et al, The objective of this overview is to summarize current 1988; Wilson et al, 1989; HoÈfer & BaÈttig, 1993; Lewis et knowledge of the effects of coffee and caffeine intake on al, 1993), and some even showed a small inverse associa- blood pressure. The possible mechanisms of these tion between self-reported coffee consumption and blood responses will also be discussed. The relation between pressure (Periti et al, 1987; Stensvold et al, 1989; Salvag- coffee consumption and plasma lipids has been reviewed gio et al, 1990; Gyntelberg et al, 1995; Wakabayashi et al, elsewhere (Thelle et al, 1987; Pirich et al, 1993, Thomp- 1998). Consumption of coffee appears to be positively son, 1994) and is therefore not included in this article.SBP, associated with an increased risk of thromboembolic systolic blood pressure; DBP, diastolic blood pressure. stroke in middle-aged hypertensive men (Hakim et al, 1998). Methods Development of tolerance during habitual use may The MEDLINE and the Current Contents literature explain why coffee consumption in some epidemiologic searches for published articles in English from January studies did not appear to have a clear effect on blood 1966 to April 1999 were carried out using the keywords pressure. However, a large cross-sectional study with over `coffee or caffeine' and `blood pressure or hypertension'. 5000 participants found that caffeine consumption within Articles concerning human epidemiological and controlled the last 3 h was associated with signi®cantly higher blood clinical studies that have relevant data on the effects of pressure than when caffeine was not consumed in the last coffee or caffeine on blood pressure were included in the 9 h (Shirlow et al, 1988). A smaller study with 338 female present review. In addition, reference lists were reviewed to subjects also con®rmed that recent coffee consumption was obtain applicable articles. We also referred to animal and in related to increased blood pressure (HoÈfer & BaÈttig, 1993). vitro studies that dealt with the possible mechanisms of In some of the epidemiologic studies blood pressure mea- actions of the cardiovascular effects of coffee or caffeine. surements were taken under conditions when fasting was The inclusion criteria for clinical trials were a controlled required from the participants (Salvaggio et al, 1990; Lewis study design and random assignment. Exclusion criteria et al, 1993; Gyntelberg et al, 1995). This may have led to were an open study design, the numerical values of the an underestimation of the blood pressure levels in the blood pressure not given or statistical analysis not per- studies with negative results, because caffeine deprivation formed. has been associated with lower blood pressure in habitual consumers (James, 1994; Phillips-Bute & Lane, 1998). Epidemiologic studies on coffee and blood pressure Thus, the inverse association between coffee intake and Several cross-sectional and longitudinal epidemiologic blood pressure observed in some epidemiologic studies studies have evaluated the effects of coffee and caffeine (Salvaggio et al, 1990; Gyntelberg et al, 1995) is consistent

Table 1 Epidemiologic studies on coffee and blood pressure

Reference Country Population Age (y) Main results

Cross-sectional studies Birkett & Logan (1988) Canada 2 436 M ‡ F > 17 DBP increased by 1 mmHg at usual level of caffeine intake (181 mg=day) Burke et al (1992) Australia 843 M ‡ F 60 ± 87 BP was positively related to coffee drinking in subjects treated with antihypertensive drugs. No signi®cant association between coffee drinking and BP in untreated subjects was found HoÈfer & BaÈttig (1993) Switzerland 338 F 20 ± 40 Actual coffee consumption on the testing day was associated with elevated BP (0.7 ± 2 mmHg per cup of coffee). Habitual coffee consumption was unrelated to BP Lang et al (1983a) Algeria 1 491 M ‡ F 15 ± 70 DBP was 2 mmHg higher among coffee drinkers than among nondrinkers Lang et al (1983b) France 6 321 M ‡ F 18 ± 60 SBP was 2 ± 4 mmHg higher among coffee drinkers than among nondrinkers Lewis et al (1993) USA 5 115 M ‡ F 18 ± 30 Caffeine intake (up to 800 mg=day) was unrelated to BP LoÈwik et al (1991) Netherlands 255 M ‡ F 65 ± 79 Coffee consumption was positively correlated with BP among women Narkiewicz et al (1995) Italy 887 M ‡ F 18 ± 45 Daytime SBP was 2.5 mmHg higher in men who consumed coffee  4 cups=day than in men who did not drink coffee Periti et al (1987) Italy 500 M ‡ F 18 ± 62 SBP and DBP were 0.8 and 0.5 mmHg lower per daily consumed cup of coffee Salvaggio et al (1990) Italy 9 601 M ‡ F 18 ± 65 Increasingly lower BP with increases in coffee consumption; those who drank 4 ± 5 cups=day had SBP and DBP 2 ± 3 and 1 mmHg lower than nondrinkers Sharp & Benowitz (1990) USA 247 A positive correlation between plasma levels of caffeine and BP in infrequent caffeine users but not in habitual users Shirlow et al (1988) Australia 5 147 M ‡ F 20 ± 70 Caffeine intake within the last 3 h was associated with 2 ± 4 mmHg higher BP. Average daily caffeine consumption was not associated with BP Stensvold et al (1989) Norway 29 027 M ‡ F 40 ± 42 An inverse U-shaped relation between coffee consumption and BP, the mean values being lowest for 0 and  9 cups=day Wakabayashi et al (1998) Japan 3 336 M 48 ± 56 SBP and DBP were 0.6 and 0.4 mmHg lower per daily consumed cup of coffee Longitudinal studies Gyntelberg et al (1988) Denmark 2 975 M 53 ± 74 DBP was negatively associated with coffee consumption in coffee drinkers Wilson et al (1989) USA 5 209 M ‡ F Middle-aged Coffee consumption was unrelated to SBP

SBP, systolic blood pressure; DBP, diastolic blood pressure; BP, blood pressure. Coffee, caffeine and blood pressure M-L Nurminen et al 833 with the lower blood pressure in regular users of coffee in hypertensive or hypertension-prone subjects than in during caffeine withdrawal. normotensive subjects (Smits et al, 1986a; Greenstadt et al, 1988; Sung et al, 1994; Lovallo et al, 1996b). The effect has been shown to be additive or enhanced during Acute intake of coffee or caffeine and blood pressure mental and physical stress (Greenstadt et al, 1988; Pincomb et al, 1988, 1991; Lovallo et al, 1989, 1991, 1996b; Myers A rise in blood pressure after acute intake of coffee has et al, 1989; Lane et al, 1990; France & Ditto, 1992). been shown in many studies (Table 2). It seems that It has been suggested that the pressor effect of caffeine caffeine is the active component responsible for this is stronger in older subjects than that seen in the young ones effect, because in several studies regular coffee intake (Izzo et al, 1983; Massey, 1998). No racial or sex differ- increased blood pressure, whereas decaffeinated coffee ences have been found in blood pressure responses to had no effect (Smits et al, 1985; Ray et al, 1986; Myers caffeine (Myers et al, 1989; Strickland et al, 1989; Bak et al, 1989; Nussberger et al, 1990). In addition, no & Grobbee, 1990; James, 1994). differences between the pressor effects of regular coffee The pressor effect of acute caffeine intake is stronger in and caffeine have been found (Casiglia et al, 1992). persons who do not normally consume caffeine than in A single dose of caffeine (200 ± 250 mg, equivalent to habitual users of caffeine (Izzo et al, 1983; Casiglia et al, two to three cups of coffee) increases systolic blood 1992). The degree of blood pressure response associated pressure by 3 ± 14 mmHg and diastolic blood pressure by with a single dose of caffeine seems to be inversely related 4 ± 13 mmHg in normotensive subjects (Table 2). The to the plasma caffeine concentration at the time of admin- pressor effect coincides with the increase in plasma caf- istration, i.e. the greatest blood pressure response occurs in feine concentrations (Robertson et al, 1978; Haigh et al, those subjects with the lowest caffeine concentration 1993; Sung et al, 1994). Blood pressure usually elevates (Robertson et al, 1981). within 30 min, and the maximal increase occurs 60 ± 120 min after caffeine intake (Robertson et al, 1978; Free- stone & Ramsay 1982; Izzo et al, 1983; Nussberger et al, Repeated coffee or caffeine intake and blood pressure 1990; Sung et al, 1994). The increase in blood pressure may last over 2 ± 4 h (Freestone & Ramsay 1982; Casiglia Despite the well-described pressor response to acute intake et al, 1992; Sung et al, 1994; Bender et al, 1997). The of coffee or caffeine, the evidence for long-term effects on pressor response to caffeine seems to be more pronounced blood pressure is inconclusive. The results of the studies

Table 2 Controlled studies on the effects of coffee or caffeine on systolic blood pressure (SBP) and diastolic blood pressure (DBP) after acute administration

Subjects Change (mmHg)

Reference n Age (y) Dose SBP DBP

Normotensive subjects Astrup et al (1990) 6 20 ± 32 100 mg caffeine ‡ 2 ‡ 3 200 mg caffeine ‡ 20 400 mg caffeine ‡ 6 ‡ 6 Bender et al (1997) 12 21 ± 26 5 mg=kg caffeine ‡ 9 ‡ 4 Casiglia et al (1992) 15 24 ± 30 2 cups of coffee ‡ 3 ‡ 4 200 mg caffeine ‡ 6 ‡ 7 France & Ditto (1992) 48 16 ± 36 250 mg caffeine ‡ 3 ‡ 6 Haigh et al (1993) 8 67 ± 82 250 mg caffeine ‡ 12 ‡ 7 Lane & Williams (1987) 30 19 ± 28 250 mg caffeine ‡ 7 ‡ 4 Lane et al (1990) 25 18 ± 36 3.5 mg=kg caffeine ‡ 8 ‡ 8 Lovallo et al (1989) 34 21 ± 35 3.3 mg=kg caffeine ‡ 6±7 ‡ 4±8 Lovallo et al (1991) 39 27 Æ 1 3.3 mg=kg caffeine ‡ 7±8 ‡ 5±7 Lovallo et al (1996b) 24 20 ± 39 3.3 mg=kg caffeine ‡ 6 ‡ 4 Nussberger et al (1990) 8 24 ± 28 250 mg caffeine ‡ 12 ‡ 13 Passmore et al (1987) 8 21 ± 38 90 mg caffeine ‡ 5 ‡ 8 180 mg caffeine ‡ 7 ‡ 7 360 mg caffeine ‡ 11 ‡ 8 Pincomb et al (1988) 44 20 ± 36 3.3 mg=kg caffeine ‡ 4 ‡ 5 Pincomb et al (1991) 34 20 ± 35 3.3 mg=kg caffeine ‡ 6±10 ‡ 4±11 Ray et al (1986) 9 4 mg=kg caffeine ‡ 8 ‡ 11 Robertson et al (1978) 9 21 ± 30 250 mg caffeine ‡ 14 ‡ 10 Smits et al (1983) 12 17 ± 38 2 cups of coffee ‡ 5 ‡ 11 Smits et al (1985) 8 21 ± 25 300 ml coffee ‡ 4±5 ‡ 8±9 Smits et al (1986a) 10 19 ± 37 2 cups of coffee ‡ 5 ‡ 11 Sung et al (1994) 12 30 ± 45 3.3 mg=kg caffeine ‡ 9 ‡ 8 Hypertensive or borderline hypertensive subjects Freestone & Ramsay (1982) 16 500 ml coffee ‡ 10 ‡ 7 (  200 mg caffeine) Goldstein & Shapiro (1987) 18 37 ± 60 200 mg caffeine ‡ 8 ‡ 6 Lovallo et al (1996b) 24 20 ± 39 3.3 mg=kg caffeine ‡ 11 ‡ 8 Smits et al (1986a) 10 18 ± 56 2 cups of coffee ‡ 13 ‡ 11 Sung et al (1994) 18 30 ± 45 3.3 mg=kg caffeine ‡ 12 ‡ 11

SBP, systolic blood pressure; DBP, diastolic blood pressure. Coffee, caffeine and blood pressure M-L Nurminen et al 834 Table 3 Controlled clinical studies on the effects of coffee or caffeine on blood pressure (BP) in habitual coffee users after repeated administration

Subjects Daily Reference n Age (y) Treatments dose Duration Main results

Normotensive subjects Ammon et al (1983) 8 20 ± 30 8 cups 4 weeks Changing from decaffeinated coffee to Decaffeinated coffee 8 cups 4 weeks caffeinated coffee increased BP by 3 mmHg Bak & Grobbee (1990) 107 18 ± 33 Filtered coffee 4 ± 6 cups 9 weeks from coffee decreased SBP by Boiled coffee 4 ± 6 cups 9 weeks 3 mmHg. Filtered coffee and boiled coffee No coffee 9 weeks showed a similar BP response Bak & Grobbee (1991) 69 25 Æ 4 Caffeine 75 mg64 ± 6 9 weeks No differences in BP between groups Placebo 9 weeks receiving caffeine and placebo tablets Burr et al (1989) 54 18 ± 58 Instant coffee  5 cups 4 weeks SBP was 3 mmHg higher when subjects Decaffeinated coffee  5 cups 4 weeks received caffeinated coffee than whey they No coffee 4 weeks abstained from coffee van Dusseldorp et al 45 24 ± 45 Filtered coffee 5 cups 6 weeks Use of decaffeinated coffee decreased SBP (1989) Decaffeinated coffee 5 cups 6 weeks and DBP by 1.5 and 1.0 mmHg van Dusseldorp et al 54 39 Æ 6 Boiled coffee 6 cups 11 weeks Boiled coffee increased SBP by 3 mmHg (1991) 39 Æ 9 Boiled and ®ltered coffee 6 cups 11 weeks compared with boiled and ®ltered coffee. 39 Æ 8 No coffee 11 weeks Abstinence from coffee had no effect on BP HoÈfer & BaÈttig (1994) 120 20 ± 45 Instant coffee 5.3 cups 9 days No differences in BP between groups Decaffeinated coffee 5.3 cups 9 days receiving caffeinated and decaffeinated coffee Robertson et al (1981) 18 21 ± 52 Caffeine 250 mg63 7 days Tolerance developed to the pressor effect of Placebo 7 days caffeine within 4 days Rosmarin et al (1990) 21 35.5 Filtered coffee 3.6 cups 8 weeks Coffee had no signi®cant effect on BP No coffee 8 weeks Superko et al (1991) 181 46 Æ 11 Filtered coffee 3 ± 6 cups 2 months BP did not signi®cantly change following 48 Æ 9 Decaffeinated coffee 3 ± 6 cups 2 months change to decaffeinated coffee or 45 Æ 11 No coffee 2 months abstinence from coffee Hypertensive subjects Robertson et al (1984) 18 20 ± 44 Caffeine 250 mg63 7 days Tolerance developed to the pressor effect of Placebo 7 days caffeine after ®rst day

SBP, systolic blood pressure; DBP, diastolic blood pressure; BP, blood pressure.

Table 4 Studies on the blood pressure effects of coffee or caffeine using continuous ambulatory blood pressure (BP) monitoring

Subjects Daily Reference n Age (y) Treatments dose Duration Main results

Normotensive subjects Green & Suls (1996) 13 34.1 Caffeinated coffee ad 8 cups 1 day Caffeine increased daytime SBP and DBP Decaffeinated coffee 1 day maximally by 3.6 and 5.6 mmHg James (1994) 36 18 ± 52 Caffeine 1.75 mg=kg63 1 ± 7 days Caffeine increased SBP and DBP maximally Placebo 1 ± 7 days by 6.0 and 5.2 mmHg after 1 week's consumption Jeong & Dimsdale (1990) 12 29 ± 59 Instant coffee 3 cups 1 day Caffeinated coffee increased SBP and DBP Decaffeinated coffee 3 cups 1 day 4 mmHg during the workday Lane et al (1998) 19 37 Æ 7 Caffeine 50 mg6 2 9 h SBP and DBP were 4 and 3 mmHg higher 250 mg62 9 h when the larger dose of caffeine (250 mg62) was consumed during the workday Myers & Reeves (1991) 25 21 ± 43 Placebo ! Caffeine ! 5 days Caffeine increased daytime SBP and DBP by Placebo 200 mg62 5 days 3 mmHg during the ®rst day. BP returned to 3 days baseline by third day (data not shown) Rakic et al (1999) 22 54 ± 89 Instant coffee 5 cups 2 weeks No signi®cant differences in BP between Caffeine-free diet 2 weeks abstainers and coffee drinkers. Superko et al (1994) 150 Caffeinated coffee 4.5 Æ 1.1 cups 2 months Changing from caffeinated to decaffeinated Decaffeinated coffee 4.5 Æ 1.1 cups 2 months coffee increased daytime SBP by 3 ± 5 mmHg. No coffee 2 months Abstinence from coffee decreased SBP and DBP by 2 ± 4 and 2 ± 3 mmHg Hypertensive or borderline hypertensive subjects Eggertsen et al (1993) 23 28 ± 74 Instant coffee 3 ± 4 cups 2 weeks No signi®cant differences in mean 24 h BP Decaffeinated coffee 3 ± 4 cups 2 weeks between the treatments MacDonald et al (1991) 50 26 ± 63 Instant coffee  3 cups 2 weeks No signi®cant differences in mean 24 h BP Decaffeinated coffee  3 cups 2 weeks between the treatments No caffeine 2 weeks Normal diet 2 weeks Rakic et al (1999) 26 54 ± 89 Instant coffee 5 cups 2 weeks Mean 24 h SBP and DBP was 4.8 and 3.0 Caffeine-free diet 2 weeks mmHg higher in coffee drinkers than in abstainers

SBP, systolic blood pressure; DBP, diastolic blood pressure. Coffee, caffeine and blood pressure M-L Nurminen et al 835 with repeated administration of coffee or caffeine are blood pressure by comparing the coffee group with the no- summarized in Table 3. The effects of repeated caffeine coffee control group (Jee et al, 1999). Studies which intake have also been studied in controlled studies using compared caffeine tablets with decaffeinated coffee were continuous ambulatory monitoring to measure blood pres- excluded. According to the meta-analysis, systolic and sure (Table 4). diastolic blood pressure increased by 0.8 mmHg Some of the long-term studies have shown that caffeine (P < 0.001) and 0.5 mmHg (P < 0.01), respectively, for induced persistent pressor effects (by 3 ± 6 mmHg) in habi- every cup of coffee during long-term coffee consumption tual consumers (Ammon et al, 1983; Burr et al, 1989; James, (Jee et al, 1999). 1994; Rakic et al, 1999). Changing from caffeinated to In hypertensive subjects the combination of coffee and decaffeinated coffee (van Dusseldorp et al, 1989; Superko produced a stronger and more sustained pressor et al, 1994) or abstinence from coffee (Bak & Grobbee, 1990; response than each stimulus alone (Freestone & Ramsay, Superko et al, 1994) resulted in a slight fall (by 2 ± 5 mmHg) 1982). Moreover, in a cross-sectional observational study in blood pressure. However, other studies reported that using 24 h ambulatory blood pressure monitoring, moderate chronic caffeine consumption or abstinence from caffeine smokers and coffee drinkers with mild hypertension had were not accompanied by signi®cant changes in blood signi®cantly higher daytime blood pressure than non- pressure (Rosmarin et al, 1990; Bak & Grobbee, 1991; van smokers and those who did not drink coffee (Narkiewicz Dusseldorp et al, 1991; Superko et al, 1991; HoÈfer & BaÈttig, et al, 1995), which suggests that the effect might recur 1994). In one double blind trial in obese subjects, reduction in throughout the day, despite the increased caffeine catabo- blood pressure was observed during an energy-restricted diet lism in smokers (Parsons & Neims, 1978). combined with caffeine (200 mg63) for 6 months (Astrup et al, 1992). However, blood pressure also reduced in the Hemodynamic effects of caffeine placebo group, and there was a pronounced weight loss by more than 10 kg in both groups. Because reduction in body Acute intake of coffee or caffeine increased vascular weight lowers blood pressure in overweight subjects (Ste- resistance, indicating a vasoconstrictor effect (Pincomb vens et al, 1993), it is very dif®cult to differentiate the et al, 1985, 1988, 1991; Smits et al, 1986a; Lovallo et al, in¯uence of caffeine alone in this study. 1991; Casiglia et al, 1992; Sung et al, 1994). The pressor Daytime ambulatory blood pressure has generally been effect has often been accompanied by a small decrease in higher on caffeine days than on caffeine-free days (Jeong & the (Izzo et al, 1983; Smits et al, 1983, 1985, Dimsdale 1990; James 1994; Superko et al, 1994; Green & 1986a; Robertson et al, 1984; Sung et al, 1994), or heart Suls, 1996; Lane et al, 1998; Rakic et al, 1999), whereas some rate was not sign®cantly affected even after high doses of studies which reported ambulatory blood pressure record- caffeine (Robertson et al, 1981; Ammon et al, 1983; ings as 24 h means have failed to detect this pressor effect Casiglia et al, 1992; Bender et al, 1997). The caffeine- (MacDonald et al, 1991; Eggertsen et al, 1993) (Table 4). induced lowering of the heart rate may be a re¯ectory Most of the long-term studies were of normotensive bradycardic response to pressor action or a direct effect on subjects, and there is relatively little information on the the cardiac or sino-atrial node. The re¯ex activation of long-term effects of caffeine in hypertensive patients baroreceptors is supported by a recent study using power (Robertson et al, 1984; MacDonald et al, 1991; Eggertsen spectral analysis, in which an increase in parasympathetic et al, 1993; Rakic et al, 1999). nerve activity was observed in normotensive subjects The failure to detect an increase in blood pressure caused (Hibino et al, 1997). On the other hand, a single dose of by repeated adminstration of caffeine may be due to the caffeine reduced baroreceptor sensitivity simultaneously development of tolerance. The pressor response has been with an increase in blood pressure in normotensive subjects reported as diminishing within a few days of the start of (Mosqueda-Garcia et al, 1990). The effects were not seen regular intake of caffeine (Robertson et al, 1981, 1984; after multiple doses of caffeine over one week, possibly Ammon et al, 1983; Myers & Reeves, 1991). However, due to the development of tolerance to the hemodynamic the tolerance may be partial, because in some studies effects of caffeine. caffeine was still able to elevate blood pressure during Palpitations sometimes associated with caffeine intake habitual consumption (Burr et al, 1989; Jeong & Dimsdale may be explained by an increased level of altertness or by 1990; James 1994; Green & Suls, 1996). The pressor an increased force of myocardial contraction (Myers & response to caffeine may be regained after a relatively Reeves, 1991). A positive inotropic effect of caffeine has short period of abstinence. Caffeine taken after only an been reported (Bender et al, 1997), although in general, overnight abstinence of 10 ± 12 h increased blood pressure little or no effect on cardiac output at rest has been noted in regular coffee drinkers (Freestone & Ramsay, 1982; Isso (Pincomb et al, 1985, 1991; Casiglia et al, 1992; Bender et al, 1983; Ray et al, 1986; Goldstein & Shapiro, 1987; et al, 1997). However, during stress an enhancement of Pincomb et al, 1988, 1991; Myers et al, 1989; Lane et al, myocardial contractility has been found to be accompanied 1990; Lovallo et al, 1991, 1996b; Sung et al, 1994). A by an increase in cardiac output (Pincomb et al, 1988; pressor response to the second cup of coffee of the day Lovallo et al, 1991). The chronic effects of caffeine on within 1 ± 2 h of the ®rst morning cup has also been observed cardiac output have not been studied. (Lane & Manus, 1989; Goldstein et al, 1990). Thus, the cardiovascular effects of caffeine may persist, at least in part, Mechanisms of the cardiovascular effects of coffee during chronic consumption of caffeine. It is possible that and caffeine some people have a higher sensitivity to caffeine or a lesser susceptibility to the development of tolerance to the pressor The summary of the possible mechanisms of the cardio- effect of caffeine. vascular effects of coffee and caffeine has been presented A recent meta-analysis of 11 controlled clinical trial in Table 5. The best known pharmacologically active with 522 participants assessed the effect of coffee intake on substance in coffee is trimethylxanthine caffeine. The Coffee, caffeine and blood pressure M-L Nurminen et al 836 Table 5 Possible mechanisms of the cardiovascular effects of caffeine channels, as well as phospholipases C and D (Fredholm, 1995; Olah & Stiles, 1995). Stimulation of A receptors  Antagonistic effects on adenosine receptors 2a  Inhibition of phosphodiesterases (increase in cyclic ) activates adenylate cyclase and probably voltage-sensitive  Activation of the sympathetic nervous system (release of channels. The two subtypes of receptors may be catecholamines from adrenal medulla) co-expressed and activated simultaneously in the same cells  Stimulation of adrenal cortex (release of corticosteroids) (Fredholm, 1995). The actions of caffeine are therefore  Renal effects (diuresis, natriuresis, activation of the renin-angiotensin- complicated. aldosterone system) Adenosine also inhibits the release of many neural transmitters, e.g. noradrenaline, , , glutamate and GABA (Fredholm & Dunwiddie, 1988). The main metabolite of caffeine, , also increased excitatory transmitter release seems to be more inhibited blood pressure in man (Benowitz et al, 1995). Some of the than that of the inhibitory ones (Fredholm, 1995). This may cardiovascular effects of coffee may be mediated by a be related to activation of potassium channels via the A1 variety of substances other than caffeine, such as the lipid receptors. By blocking these receptors, caffeine may facil- compounds and , which have been impli- itate the release of . cated as the components responsible for the hyperlipidemic Caffeine not only acts at the level on the effect of un®ltered coffee (Urgert & Katan, 1997). Other adenosine system, but in rats it can increase plasma con- chemicals identi®ed in coffee include chlorogenic acid, centration of adenosine, probably via a receptor-mediated carbohydrates and peptides (Chou & Benowitz, 1994). mechanism at concentrations achieved by the daily human Both normal and decaffeinated coffee have been shown use of coffee (Conlay et al, 1997). to contain muscarinic compound(s) that constricted coron- ary arteries in vivo (Kalsner, 1977). An extract which Vascular effects and calcium produced an abrupt of blood pressure and Caffeine produced transient contractions of arterial smooth heart rate in rats and relaxed aortic rings in vitro has also muscle at millimolar concentrations in vitro (Kalsner, been isolated from regular and decaffeinated instant coffee 1971; Karaki et al, 1987; Sato et al, 1988; Moriyama et (Tse, 1992). al, 1989), which were more marked in preparations from spontaneously hypertensive rats than from normotensive Inhibition of phosphodiesterases rats (Moriyama et al, 1989). Caffeine-induced vascular Caffeine could in¯uence vascular tone by inhibiting phos- contractions were reduced or blocked by removal of extra- phodiesterases, with consequent increases in cyclic AMP cellular calcium or by the inhibition of calcium release and cyclic GMP levels (Kramer & Wells, 1979). However, from the (Sato et al, 1988; Mori- caffeine is a non-speci®c inhibitor of phosphodiesterases yama et al, 1989). It has been suggested that the mobiliza- and is only effective in vitro at much higher concentrations tion of intracellular calcium by the so-called calcium- (50% inhibitory concentration of 390 ± 500 mmol=l; Kramer induced calcium release mechanism is crucial in caffeine- & Wells, 1979) than are achieved in vivo in plasma after induced (Karaki et al, 1987). On the two to three cups of coffee (20 ± 40 mmol=l; Izzo et al, venous side in vitro, caffeine induced phasic contractions, 1983; Smits et al, 1983; Cassiglia et al, 1992). Caffeine which were dependent on both the in¯ux of extracellular induced vasoconstriction in humans, whereas , calcium and the release of intracellular calcium (Shima- a xanthine derivative which has prominent phosphodiester- mura et al, 1991). ase inhibiting effects, caused vasodilatation (Smits et al, Studies on the role of calcium in¯ux through calcium 1989). Thus, inhibition of phosphodiesterases appears not channels in the pressor response to caffeine in man have to be the major mechanism mediating the cardiovascular been contradictory, because pretreatment with the slow L- effects of caffeine. type calcium channel antagonist had no effect (Smits et al, 1986b), whereas reversed the Antagonism of adenosine receptors increase in blood pressure caused by caffeine in normoten- Caffeine exerts many of its actions by antagonizing adeno- sive subjects (van Nguyen & Myers, 1988). sine receptors, of which four subtypes have been cloned In addition to vasoconstriction, caffeine has also been and characterized (Fredholm, 1995; Olah & Stiles, 1995). shown to inhibit contractions in vascular Adenosine dilates several vascular beds (Ohah & Stiles, (Ahn et al, 1988; Sato et al, 1988), partly due to reduced 1995). Based on studies with knockout mice lacking A2a- cytoplasmic calcium and partly due to reduced sensitivity receptors and having elevated blood pressure (Ledent et al, of the contractile elements to calcium (Sato et al, 1988). 1997), adenosine seems to act as a physiological vaso- Caffeine-induced relaxations in rat aortic rings consist dilator. Thus, caffeine may cause an increase in blood partly of a nitric-oxide-dependent component and also of pressure by antagonizing this effect of adenosine. The a component involving the inhibition of phosphodiesterases major targets of caffeine are A1 and A2a receptors, which (Hatano et al, 1995). can be blocked by caffeine at low micromolar concentra- tions similar to those that occur in plasma after consump- Sympathetic nervous system tion of a few cups of coffee (Fredholm, 1995). In Acute administration of coffee or caffeine increases the normotensive men, intravenous caffeine has been shown plasma concentrations of catecholamines, particularly adre- to reduce the hemodynamic effects of adenosine naline, and the urinary of catecholamines (Smits et al, 1989). (Robertson et al, 1978; Izzo et al, 1983; Smits et al, Adenosine A1 and A2a receptors have partially opposite 1985; Pincomb et al, 1988; Lane et al, 1990; Nussberger cellular effects. Stimulation of A1 receptors causes the et al, 1990; Narkiewicz et al, 1995). The increase in plasma inhibition of adenylate cyclase and some types of calcium is similar in normotensive and hypertensive channels, and the activation of several types of potassium subjects (Smits et al, 1986a). Coffee, caffeine and blood pressure M-L Nurminen et al 837 In animal experiments caffeine activates hydroxy- chronic coffee consumption and plasma renin activity has lase (Snider & Waldeck, 1974; Datta et al, 1996), the rate- been observed in mildly hypertensive men (Palatini et al, limiting in catecholamine . In addition, 1996). Thus, the importance of the renin ± angiotensin ± caffeine releases adrenal catecholamines, both directly and aldosterone system in the cardiovascular effects of caffeine through neurogenic stimulation (Berkowitz & Spector, remains unclear. 1971; Snider & Waldeck, 1974). It has also been speculated that the caffeine-induced increase in plasma catechola- mines is due to inhibition of the extraneuronal metabolism Conclusion of catecholamines (Kalsner, 1971). The acute pressor effect of coffee and caffeine is well An increase in blood pressure by caffeine was seen in documented, but the reports on the chronic effects on blood some normotensive (Izzo et al, 1983) and hypertensive pressure are inconsistent. Large inter-individual variations subjects (Potter et al, 1993) as well as in adrenalectomized in caffeine metabolism may contribute to the variability in patients (Smits et al, 1986a) without an increase in plasma the hemodynamic responses to chronic caffeine consump- catecholamines. Caffeine also elevates blood pressure in tion. Differences in the extent of tolerance may also under- subjects with autonomic dysfunction, in whom sympathetic lie the variations in sensitivity to blood pressure elevation responsiveness is lacking (Onrot et al, 1985; van Soeren caused by caffeine. Although there is at present no clear et al, 1996). Therefore, the pressor response to coffee is not epidemiologic evidence that coffee consumption is causally purely the result of an increase in circulating catechol- related to hypertension, high intake of coffee may be an amines or the activation of the sympathetic nervous system. additional risk factor in the development of hypertension at the individual level, because some people seem to have a Effects on adrenal cortex higher sensitivity to the cardiovascular effects of caffeine, In normotensive and hypertensive subjects, caffeine e.g. due to long-lasting stress or to a generic susceptibility caused stress-like increases in plasma adrenocorticotropin to hypertension. We conclude that coffee and caffeine in (ACTH) and cortisol concentrations (Lovallo et al, 1996a) large doses may be harmful to some hypertensive or and potentiated stress-related increases in plasma cortisol hypertension-prone subjects. Before more de®nite conclu- (Pincomb et al, 1988; Lovallo et al, 1989; Lane et al, sions can be drawn, long-term clinical trials, as well as 1990). Corticosteroids can in¯uence vascular tone, e.g. by prospective epidemiologic studies with large numbers of enhancing reactivity to other vasoactive substances, such as participants, are needed. catecholamines (Walker & Williams, 1992). Acknowledgements ÐSupported by grants from the Academy of Finland Renal effects and the Biomed-2 (BMH4-CT96-0979). The authors would like to thank Acute ingestion of caffeine slightly increased urinary Mrs Mimi Ponsonby for her excellent language revision. volume and sodium excretion in humans (Robertson et al, 1978; Passmore et al, 1987; Nussberger et al, 1990). The References role of atrial natriuretic peptide (ANP) in the and natriuretic effects of caffeine is unclear, because the plasma Ahn HY, Karaki H & Urakawa N (1988): Inhibitory effects of caffeine on contractions and calcium movements in vascular and intestinal smooth concentrations of ANP were not affected by acute intake of muscle. Br. J. Pharmac. 93, 267 ± 274. caffeine (Nussberger et al, 1990), whereas a signi®cant Ammon HPT, Bieck PR, Mandalaz D & Verspohl EJ (1983): Adaptation increase in circulating ANP was observed during 2 weeks of blood pressure to continuous heavy coffee drinking in young of regular caffeine consumption (Eggertsen et al, 1993). volunteers. A double-blind crossover study. Br. J. Clin. Pharmac. 15, 701 ± 706. The plasma concentrations of antidiuretic hormone (vaso- Astrup A, Toubro S, Cannon S, Hein P, Breum L & Madsen J (1990): pressin) remained unchanged after an acute intake of Caffeine: a double-blind, placebo-controlled study of its thermogenic, caffeine (Izzo et al, 1983; Nussberger et al, 1990). On metabolic, and cardiovascular effects in healthy volunteers. Am. J. Clin. the other hand, in rats caffeine caused a marked dose- Nutr. 51, 759 ± 767. Astrup A, Breum L, Toubro S, Hei P & Quaade F (1992): The effect and dependent increase in renal prostaglandin excretion (Engel- safety of an ephedrine=caffeine compound compared to ephedrine, hardt, 1996), which could, at least partly, explain the caffeine and placebo in obese subjects on an energy restricted diet. A caffeine-induced diuresis. double blind trial. Int. J. Obes 16, 269 ± 277. Repeated administration of caffeine for one week caused Bak AAA & Grobbee DE (1990): A randomized study on coffee and blood renal vasoconstriction and lowered renal plasma ¯ow in pressure. J. Hum. Hypertens. 4, 259 ± 264. Bak AAA & Grobbee DE (1991): Caffeine, blood pressure, and serum normotensive volunteers ingesting a high-salt diet (Brown lipids. Am. J. Clin. Nutr. 53, 971 ± 975. et al, 1993). Plasma renin activity was increased, which Barone JJ & Roberts HR (1996): Caffeine consumption. Chem. could have caused the renal vasoconstriction via increased Toxicol. 34, 119 ± 129. angiotensin II concentrations. Acutely administered caf- Bender AM, Donnerstein RL, Samson RA, Zhu D & Goldberg SJ (1997): Hemodynamic effects of acute caffeine ingestion in young adults. Am. feine increased plasma renin activity in experimental ani- J. Cardiol. 79, 696 ± 699. mals (Ohnishi et al, 1987) and in man (Robertson et al, Benowitz NL, Jacob P, Mayan H & Denaro C (1995): Sympathomimetic 1978, 1981), probably through the antagonism of adenosine effects of paraxanthine in humans. Clin. Pharmac. Ther. 58, 684 ± 691. receptors on juxtaglomerular cells (Kuan et al, 1990). In Berkowitz BA & Spector S (1971): Effect of caffeine and theophylline on some studies, however, no signi®cant effect of acute peripheral catecholamines. Eur. J. Pharmac. 13, 193 ± 196. Birkett NJ & Logan AG (1988): Caffeine-containing beverages and the caffeine intake on plasma renin activity was observed prevalence of hypertension. J. Hypertens. 6(Suppl 4), S620 ± S622. (Passmore et al, 1987; Nussberger et al, 1990), or else Brown NJ, Ryder D, Nadeau J (1993): Caffeine attenuates the renal the increase was attenuated after chronic administration of vascular response to angiotensin II infusion. Hypertension 22, caffeine (Robertson et al, 1981, 1984; Eggertsen et al, 847 ± 852. Burke V, Beilin LJ, German R, Grosskopf S, Ritchie J, Puddey B & 1993). A decrease in plasma renin activity after an acute Rogers P (1992): Association of lifestyles and personality character- intake of coffee has also been reported (Smits et al, 1983, istics with blood pressure and hypertension: a cross-sectional study in 1986a,b). Moreover, an inverse association between the elderly. J. Clin. Epidemiol. 45, 1061 ± 1070. Coffee, caffeine and blood pressure M-L Nurminen et al 838 Burr ML, Gallacher JEJ, Butland BK, Bolton CH & Downs LG (1989): Izzo JL, Ghosal A, Kwong T, Freeman RB & Jaenike JR (1983): Age and Coffee, blood pressure and plasma lipids: a randomized controlled trial. prior caffeine use alter the cardiovascular and adrenomedullary Eur. J. Clin. Nutr. 43, 477 ± 483. responses to oral caffeine. Am. J. Cardiol. 52, 769 ± 773. Callahan MM, Robertson RS, Branfman AR, McComish MF & Yesair James JE (1994): Chronic effects of habitual caffeine consumption on DW (1983): Comparison of caffeine metabolism in three nonsmoking laboratory and ambulatory blood pressure levels. J. Cardiovasc. Risk 1, populations after of radiolabeled caffeine. 159 ± 164. Metab. Dispos. 11, 211 ± 217. James JE (1997): Is habitual caffeine use a preventable cardiovascular risk Casiglia E, Paleari CD, Petucco S, Bongiovi S, Colangeli G, Baccilieri factor? Lancet 349, 279 ± 281. MS, Pavan L, Pernice M & Pessina AC (1992): Haemodynamic effects Jee SH, HE J, Whelton PK, Suh I & Klag MJ (1999): The effect of chronic of coffee and puri®ed caffeine in normal volunteers: a placebo-con- coffee drinking on blood pressure Ð a meta-analysis of controlled trials. trolled clinical study. J. Hum. Hypertens. 6, 95 ± 99. Hypertension 33, 647 ± 652. Chou TM & Benowitz NL (1994): Caffeine and coffee: effects on health Jeong D-U & Dimsdale JE (1990): The effects of caffeine on and cardiovascular disease. Comp. Biochem. Physiol. 109C, 173 ± 189. blood pressure in the work environment. Am. J. Hypertens. 3, 749 ± Conlay LA, Conant JA, deBros F & Wurtman R (1997): Caffeine alters 753. plasma adenosine levels. Nature 389, 136. Kalsner S (1971): Mechanism of potentiation of contractor responses to Datta U, Noailles, P-AH, Rodriguez M, Kraft M, Zhang Y & Angulo JA catecholamines by methylxanthines in aortic strips. Br. J. Pharmac. 43, (1996): Accumulation of tyrosine hydroxylase messenger RNA mole- 379 ± 388. cules in the rat mesencephalon by chronic caffeine treatment. Neurosci. Kalsner S (1977): A coronary vasoconstrictor substance is present in Lett. 220, 77 ± 80. regular and `decaffeinated' forms of both percolated and instant van Dusseldorp M, Smits P, Thien T & Katan MB (1989): Effect of coffee. Life Sci. 20, 1689 ± 1696. decaffeinated versus regular coffee on blood pressure. A 12-week, Karaki H, Ahn HY & Urakawa N (1987): Caffeine-induced contraction double-blind trial. Hypertension 14, 563 ± 569. in vascular smooth muscle. Arch. Intern. Pharmacodyn. Ther. 285, van Dusseldorp M, Smits P, Lenders JWM, Thien T & Katan MB (1991): 60 ± 71. Boiled coffee and blood pressure. A 14-week controlled trial. Hyper- Kawachi I, Colditz GA & Stone CB (1994): Does coffee drinking increase tension 18, 607 ± 613. the risk of coronary heart disease? Results from a meta-analysis. Br. Eggertsen E, Andreasson AÊ , Hedner T, Karlberg BE & Hansson L (1993): Heart. J. 72, 269 ± 275. Effect of coffee on ambulatory blood pressure in patients with treated Knutti R, Rothweiler H & Schlatter C (1981): Effects of pregnancy hypertension. J. Intern. Med. 233, 351 ± 355. on the pharmacokinetics of caffeine. Eur. J. Clin. Pharmac. 21, 121 ± Engelhardt G (1996): Effect of acetylsalicylic acid, , caffeine 126. and a combination of these substances on the renal prostaglandin E2,6- Kramer GL & Wells JN (1979): Effects of inhibitors on keto-prostaglandin F1a, , creatinine and electrolyte excretion of the cyclic levels and relaxation of pig coronary arteries. Mol. rat. Arzneim.-Forsch.=Drug Res. 46, 509 ± 512. Pharmac. 16, 813 ± 822. France C & Ditto B (1992): Cardiovascular responses to the combination Kuan C-J, Wells JN & Jackson EK (1990): Endogeneous adenosine of caffeine and mental arithmetic, cold pressor, and static exercise restrains renin release in conscious rats. Circul. Res. 66, 637 ± 646. stressors. Psychophysiology 29, 272 ± 282. Lane JD & Manus DC (1989): Persistent cardiovascular effects with Fredholm BB (1995): Adenosine, adenosine receptors and the actions of repeated caffeine administration. Psychosom. Med. 51, 373 ± 380. caffeine. Pharmac. Toxicol. 76, 93 ± 101. Lane JD & Williams RB (1987): Cardiovascular effects of caffeine and Fredholm BB & Dunwiddie TV (1988): How does adenosine inhibit stress in regular coffee drinkers. Psychophysiology 24, 157 ± 164. transmitter release? Trends Pharmac. Sci. 9, 130 ± 134. Lane JD, Adcock RA, Williams RB & Kuhn CM (1990 ): Caffeine effects Freestone S & Ramsay LE (1982): Effect of coffee and cigarette smoking on cardiovascular and neuroendocrine responses to acute psychosocial on the blood pressure of untreated and diuretic-treated hypertensive stress and their relationship to level of habitual caffeine consumption. patients. Am. J. Med. 73, 348 ± 353. Psychosom. Med. 52, 320 ± 336. Goldstein IB & Shapiro D (1987): The effects of stress and caffeine on Lane JD, Phillips-Bute B & Pieper C (1998): Caffeine raises blood hypertensives. Psychosom. Med. 49, 226 ± 235. pressure at work. Psychosom. Med. 60, 327 ± 330. Goldstein IB, Shapiro D, Hui KK & Yu JL (1990): Blood pressure Lang T, Bureau JF, Degoulet P, Salah H & Benattar C (1983a): Blood response to the `second cup of coffee'. Psychosom. Med. 52, pressure, coffee, tea and consumption: an epidemiological 337 ± 345. study in Algiers. Eur. Heart J. 4, 602 ± 607. Green PJ & Suls J (1996): The effects of caffeine on ambulatory blood Lang T, Degoulet P, Aime F, Fouriaud C, Jacquinet-Salord M-C, Laprugne pressure, heart rate, and mood in coffee drinkers. J. Behav. Med. 19, J, Main J, Oeconomos J, Phalente J & Prades A (1983b): Relation 111 ± 128. between coffee drinking and blood pressure: analysis of 6,321 subjects Greenland S (1993): A meta-analysis of coffee, myocardial infarction, and in the Paris region. Am. J. Cardiol. 52, 1238 ± 1242. coronary death. Epidemiology 4, 366 ± 374. Ledent C, Vaugeois J-M, Schiffmann SN, Pedrazzini T, El Yacoubi M, Greenstadt L, Yang L & Shapiro D (1988): Caffeine, mental stress, and Vanderhagen JJ, Costentin J, Heath JK, Vassart G & Parmentier M risk for hypertension: a cross-cultural replication. Psychosom. Med. 50, (1997): Aggressiveness, hypoalgesia and high blood pressure in mice 15 ± 22. lacking the . Nature 388, 674 ± 678. Gyntelberg F, Hein HO, Suadicani P & Sùrensen H (1995): Coffee Lewis CA, Caan B, Funkhouser E, Hilner JE, Bragg C, Dyer A, Raczynski consumption and risk of ischaemic heart disease Ð a settled issue? JM, Savage PJ, Armstrong MA & Friedman GD (1993): Inconsistent J. Intern. Med. 237, 55 ± 61. associations of caffeine-containing beverages with blood pressure and Haigh RA, Harper GD, Fotherby M, Hurd J, Macdonald IA & Potter JF with lipoproteins. Am. J. Epidemiol. 138, 502 ± 507. (1993): Duration of caffeine abstention in¯uences the acute blood Lovallo WR, Pincomb GA, Sung BH, Passey RB, Sausen KP & Wilson pressure responses to caffeine in elderly normotensives. Eur. J. Clin. MF (1989): Caffeine may potentiate adrenocortical stress responses in Pharmac. 44, 549 ± 553. hypertension-prone men. Hypertension 14, 170 ± 176. Hakim AA, Ross GW, Gurb JD, Rodriguez BL, Burch®eld CM, Sharp DS, Lovallo WR, Pincomb GA, Sung BH, Everson SA, Passey RB & Wilson Yano K & Abbott RD (1998): Coffee consumption in hypertensive men MF (1991): Hypertension risk and caffeine's effect on cardiovascular in older middle-aged and the risk of stroke: the Honolulu heart program. activity during mental stress in young men. Health Psychol. 10, J. Clin. Epidemiol. 51, 487 ± 494. 236 ± 243. Hatano Y, Mizumoto K, Yoshiyama T, Yamamoto M & Iranami H Lovallo WR, al'Absi M, Blick K, Whitsett TL & Wilson MF (1996a): (1995): Endothelium-dependent and -independent vasodilation of Stress-like adrenocorticotropin responses to caffeine in young healthy isolated rat aorta induced by caffeine. Am. J. Physiol. 269, H1679 ± men. Pharmac. Biochem. Behav. 55, 365 ± 369. H1684. Lovallo WR, al'Absi M, Pincomb GA, Everson SA, Sung BH, Passey RB Hibino G, Moritani T, Kawada T & Fushiki T (1997): Caffeine enhances & Wilson MF (1996b): Caffeine and behavioral stress effects on blood modulation of parasympathetic nerve activity in humans: Quanti®cation pressure in borderline hypertensive caucasian men. Health Psychol. 15, using power spectral analysis. J. Nutr. 127, 1422 ± 1427. 11 ± 17. HoÈfer I & BaÈttig K (1993): Coffee consumption, blood pressure tonus and LoÈwik MRH, Hofman Z, Kok FJ, Wedel M (1991): Nutrition and blood reactivity to physical challenge in 338 women. Pharmac. Biochem. pressure among elderly men and women (Dutch Nutrition Surveillance Behav. 44, 573 ± 576. System). J. Am. Coll. Nutr. 10, 149 ± 155. HoÈfer I & BaÈttig K (1994): Cardiovascular, behavioral, and subjective MacDonald TM, Sharpe K, Fowler G, Lyons D, Freestone S, Lovell HG, effects of caffeine under ®eld conditions. Pharmac. Biochem. Behav. Webster J & Petrie JC (1991): Caffeine restriction: effect on mild 48, 899 ± 908. hypertension. Br. Med. J. 303, 1235 ± 1238. Coffee, caffeine and blood pressure M-L Nurminen et al 839 Massey LK (1998): Caffeine and the elderly. Drugs Aging 13, 43 ± 50. Rosmarin PC, Applegate WB & Somes GW (1990): Coffee consumption Mehta A, Jain AC, Mehta MC & Billie M (1997): Caffeine and cardiac and blood pressure: a randomized, crossover clinical trial. J. Gen. arrhythmias. An experimental study in dogs with review of literature. Intern. Med. 5, 211 ± 213. Acta Cardiol. 52, 273 ± 283. Salvaggio A, Periti M, Miano L & Zambelli C (1990): Association Moriyama K, Osugi S, Shimamura K & Sunano S (1989): Caffeine- between habitual coffee consumption and blood pressure levels. induced contraction in arteries from stroke-prone spontaneously hyper- J. Hypertens. 8, 585 ± 590. tensive rats. Blood Vessels 26, 280 ± 289. Sato K, Ozaki H & Karaki H (1988): Multiple effects of caffeine on Mosqueda-Garcia R, Tseng C-J, Biaggioni I, Robertson RM & Robertson contraction and cytosolic free Ca2‡ levels in vascular smooth muscle of D (1990): Effects of caffeine on barore¯ex activity in humans. Clin. rat aorta. Naunyn-Schmiedeberg's Arch. Pharmac. 338, 443 ± 448, Pharmac. Ther. 48, 568 ± 574. Sharp DS & Benowitz NL (1990): Pharmacoepidemiology of the effect of Myers HF, Shapiro D, McClure F & Daims F (1989): Impact of caffeine caffeine on blood pressure. Clin. Pharmac. Ther. 47, 57 ± 60. and psychological stress on blood pressure in black and white men. Shimamura K, Moriyama K & Sunano S (1991): Caffeine-induced con- Health Psychol. 8, 597 ± 612. traction of rat portal vein and effects of K-depolarization, Na-removal Myers MG & Reeves RA (1991): The effect of caffeine on daytime and low temperature. J. Smooth Muscle Res. 27, 75 ± 85. ambulatory blood pressure. Am. J. Hypertens 4, 427 ± 431. Shirlow MJ, Berry G & Stokes G (1988): Caffeine consumption and blood Narkiewicz K, Maraglino G, Biasion T, Rossi G, Sanzuol F & Palantini P pressure: an epidemiological study. Int. J. Epidemiol. 17, 90 ± 97. (1995): Interactive effect of cigarettes and coffee on daytime systolic Smits P, Hoffman H, Thien T, Houben H & van't Laar A (1983): blood pressure in patients with mild essential hypertension. J. Hyper- Hemodynamic and humoral effects of coffee after b1-selective and tens. 13, 965 ± 970. nonselective b-blockade. Clin. Pharmacol. Ther. 34, 153 ± 158. van Nguyen P & Myers MG (1988): Cardiovascular effects of caffeine and Smits P, Thien T & van't Laar A (1985): Circulatory effects of coffee nifedipine. Clin. Pharmac. Ther. 44, 315 ± 319. in relation to the pharmacokinetics of caffeine. Am. J. Cardiol. 56, Nussberger J, Mooser V, Maridor G, Juillerat L, Waeber B & Brunner HR 958 ± 963. (1990): Caffeine-induced diuresis and atrial natriuretic peptides. Smits P, Pieters G & Thien T (1986a): The role of epinephrine in the J. Cardiovasc. Pharmac. 15, 685 ± 691. circulatory effects of coffee. Clin. Pharmac. Ther. 40, 431 ± 437. Ohnishi A, Li P, Branch RA, Holycross B & Jackson EK (1987): Caffeine Smits P, Thien T & van't Laar A (1986b): In¯uence of slow calcium- enhances the slow-pressor response to angiotensin II in rats. Evidence channel blockade on the cardiovascular effects of coffee. Eur. J. Clin. for a caffeine ± angiotensin II interactin with the sympathetic nervous Pharmac. 30, 171 ± 175. system. J. Clin. Invest. 80, 13 ± 16. Smits P, Schouten J & Thien T (1989): Cardiovascular effects of two Olah ME, Stiles GL (1995): subtypes: characterization and the relation to adenosine antagonism. Clin. Pharmac. and therapeutic regulation. Ann. Rev. Pharmac. Toxicol. 35, 581 ± 606. Ther. 45, 593 ± 599. Onrot J, Goldberg MR, Biaggioni I, Hollister AS, Kincaid D & Robertson Snider SR & Waldeck B (1974): Increased synthesis of adrenomedullary D (1985): Hermodynamic and humoral effects of caffeine in autonomic catecholamines induced by caffeine and theophylline. Naunyn-Schmie- failure. Therapeutic implications for postprandial . New deberg's Arch. Pharmac. 281, 257 ± 260. Engl. J. Med. 313, 549 ± 554. van Soeren M, Mohr T, Kjaer M & Graham TE (1996): Acute effects of Palatini P, Canali C, Graniero GR, Rossi G, de Toni R, Santonastaso M, caffeine ingestion at rest in humans with impaired epinephrine dal Follo M, Zanata G, Ferrarese E, Mormino P & Pessina AC (1996): responses. J. Appl. Physiol. 80, 999 ± 1005. Relationship of plasma renin activity with caffeine intake Stensvold I, Tverdal A & Foss OP (1989): The effect of coffee on blood and physical training in mild hypertensive men. Eur. J. Epidemiol. lipids and blood pressure. Results from a Norwegian cross-sectional 12, 485 ± 491. study, men and women, 40 ± 42 years. J. Clin. Epidemiol. 42, 877 ± 884. Parsons WD & Neims AH (1978): Effect of smoking on caffeine Stevens VJ, Corrigan SA, Obarzanek E, Bernauer E, Cook NR, Hebert P, clearance. Clin. Pharmac. Ther. 24, 40 ± 45. Mattfeldt-Beman M, Oberman A, Sugars C & Dalcin AT (1993): Passmore AP, Kondowe GB & Johnston GD (1987): Renal and Weight loss intervention in 1 of the Trials of Hypertension cardiovascular effects of caffeine: a dose-response study. Clin. Sci. Prevention. The TOHP Collaborative Research Group. Arch. Intern. 72, 749 ± 756. Med. 153, 849 ± 858. Periti M, Salvaggio A, Quaglia G & DiMarzio L (1987): Coffee con- Strickland TL, Myers HF & Lahey BB (1989): Cardiovascular reactivity sumption and blood pressure: an Italian study. Clin. Sci. 72, 443 ± 447. with caffeine and stress in black and white normotensive females. Phillips-Bute BG & Lane JD (1998): Caffeine withdrawal symptoms Psychosom. Med. 51, 381 ± 389. following brief caffeine deprivation. Physiol. Behav. 63, 35 ± 39. Sung BH, Whitsett TL, Lovallo WR, al'Absi M, Pincomb GA & Wilson Pincomb GA, Lovallo WR, Passey RB, Whitsett TL, Silverstein SM & MF (1994): Prolonged increase in blood pressure by a single oral Wilson MF (1985): Effects of caffeine on vascular resistance, cardiac dose of caffeine in mildly hypertensive men. Am. J. Hypertens 7, output and myocardial contractility in young men. Am. J. Cardiol. 56, 755 ± 758. 119 ± 122. Superko HR, Botz W, Williams PT, Albers JJ & Wood PD (1991): Pincomb GA, Lovallo WR, Passey RB & Wilson MF (1988): Effect of Caffeinated and decaffeinated coffee effects on plasma lipoprotein behavior state on caffeine's ability to alter blood pressure. Am. J. cholesterol, apolipoproteins, and lipase activity: a controlled, rando- Cardiol. 61, 798 ± 802. mized trial. Am. J. Clin. Nutr. 54, 599 ± 605. Pincomb GA, Wilson MF, Sung BH, Passey RB & Lovallo WR (1991): Superko HR, Hyll J, DiRicco C, Williams PT, Bortz WM & Wood PD Effects of caffeine on pressor regulation during rest and exercise in men (1994): Effects of cessation of caffeinated-coffee consumption on at risk for hypertension. Am. Heart J. 122, 1107 ± 1115. ambulatory and resting blood pressure in men. Am. J. Cardiol. 73, Pirich Ch, O'Grady J & Sinzinger H (1993): Coffee, lipoproteins and 780 ± 784. cardiovascular disease. Wien Klin. Wochenschr. 105, 3±6. Tang-Liu DD-S, Williams RL & Riegelman S (1983): Disposition Potter JF, Haigh RA, Harper GD, Fortherby M, Hurd S & Macdonald of caffeine and its metabolites in man. J. Pharmac. Exp. Ther. 224, IA (1993): Blood pressure, plasma catecholamine and renin responses 180 ± 185. to caffeine in elderly hypertensives. J. Hum. Hypertens. 7, 273 ± Thelle DS, Heyden S & Fodor JG (1987): Coffee and choleserol in 278. epidemiological and exerimental studies. Atherosclerosis 67, 97 ± 103. Rakic V, Burke V & Beilin LJ (1999): Effects of coffee on ambulatory Thompson WG (1994): Coffee: brew or bane? Am. J. Med. Sci. 308, blood pressure in older men and women Ð a randomized controlled 49 ± 57. trial. Hypertension 33, 869 ± 873. Tse SYH (1992): Cholinomimetic compound distinct from caffeine con- Ray RL, Nellis MJ, Brady JV & Foltin RW (1986): and caffeine tained in coffee. II: muscarinic actions. J. Pharm. Sci. 81, 449 ± 452. effects on the task-elicited blood pressure response. Addict. Behav. 11, Urgert R & Katan MB (1997): The cholesterol-raising factor from coffee 31 ± 36. beans. Am. Rev. Nutr. 17, 305 ± 324. Robertson D, FroÈlich JC, Carr RK, Watson JT, Holli®eld JW, Shand DG & Wakabayashi K, Kono S, Shinchi K, Honjo S, Todoroki I, Sakurai Y, Oates JA (1978): Effects of caffeine on plasma renin activity, catechol- Umeda T, Imanishi K & Yoshizawa N (1998): Habitual coffee con- amines and blood pressure. New Engl. J. Med. 298, 181 ± 186. sumption and blood pressure: a study of self-defence of®cials in Japan. Robertson D, Wade D, Workman R, Woosley RL & Oates JA (1981): Eur. J. Epidemiol. 14, 669 ± 673. Tolerance to the humoral and hemodynamic effects of caffeine in man. Walker BR & Williams BC (1992): Corticosteroids and vascular tone: J. Clin. Invest. 67, 1111 ± 1117. mapping the messenger maze. Clin. Sci. 82, 597 ± 605. Robertson D, Hollister AS, Kincaid D, Workman R, Goldberg MR, Tung Wilson PWF, Garrison RJ, Kannel WB, McGee DL & Castelli WP (1989): C-S & Smith B (1984): Caffeine and hypertension. Am. J. Med. 77, Is coffee consumption a contributor to cardiovascular disease? Arch. 54 ± 60. Intern. Med. 149, 1169 ± 1172.