The immediate and short-term chemosensory impacts of coffee and caffeine on cardiovascular activity

Abstract

The immediate and short-term chemosensory impacts of coffee and caffeine on cardiovascular activity. Introduction: Caffeine is detected by 5 of the 25 gustatory bitter taste receptors (hTAS2Rs) as well as by intestinal STC-1 cell lines. Thus there is a possibility that caffeine may elicit reflex autonomic responses via chemosensory stimulation. Methods: The cardiovascular impacts of double-espresso coffee, regular (130 mg caffeine) and decaffeinated, and encapsulated caffeine (134 mg) were compared with a placebo-control capsule. Measures of four post-ingestion phases were extracted from a continuous recording of cardiovascular parameters and contrasted with pre-ingestion measures. Participants (12 women) were seated in all but the last phase when they were standing. Results: Both coffees increased heart rate immediately after ingestion by decreasing both the diastolic interval and ejection time. The increases in heart rate following the ingestion of regular coffee extended for 30 min. Encapsulated caffeine decreased arterial compliance and increased diastolic pressure when present in the gut and later in the standing posture. Discussion: These divergent findings indicate that during ingestion the caffeine in coffee can elicit autonomic arousal via the chemosensory stimulation of the gustatory receptors which extends for at least 30 min. In contrast, encapsulated caffeine can stimulate gastrointestinal receptors and elicit vascular responses involving digestion. Conclusion: Research findings on caffeine are not directly applicable to coffee and vice versa. The increase of heart rate resulting from coffee drinking is a plausible pharmacological explanation for the observation that coffee increases risk for coronary heart disease in the hour after ingestion.

---

Introduction

Caffeine is a widely-used behavioural stimulant consumed primarily in the beverages coffee, tea and caffeinated soft-drinks. It is also present in a wide range of pharmaceutical products.¹Caffeine has been reported to stimulate 5 of the 25 bitter taste receptors (TAS2Rs: 7, 10, 14, 43 and 46) located in the human oral cavity² as well as receptors found in the intestinal STC-1 cells lines.³ Therefore caffeinated beverages have the potential to elicit reflex autonomic (ANS) responses affecting the cardiovascular system (CVS) due to caffeine's chemosensory impact on gustatory and gastrointestinal tract (GIT) TAS2Rs. Such responses would be distinct from those resulting from antagonism of the adenosine receptors due to elevated caffeine plasma levels.⁴ This study investigated the initial 30 min post-ingestion impact of coffee and caffeine on the CVS.

Bitter taste

Bitter taste detection functions as a sensory warning system against the ingestion of potentially toxic or noxious substances.⁵ Bitter tasting substances occur in the tissues of both plants and animals and include alkaloids, terpenoids, saponins, amino acids and peptides.⁶ In mammals, approximately 30 broadly tuned divergent G-protein-coupled receptors serve as TAS2Rs.⁵ As well as being present in the oral cavity TAS2Rs have also been reported in the GIT,⁷ nasal⁸ and bronchial⁹ epithelium, brain¹⁰ and testes.¹¹ Stimulation of gustatory TAS2Rs activates neurons located in the nucleus tractus solitarius (NTS) and the parabrachial nucleus.¹² Similarly, stimulation of GIT TAS2Rs activates neurons in the NTS.¹³

Bitter agonists are not inherently toxic and some exhibit gastro-protective activity.¹⁴ Sensitivity to bitterness varies between animals according to their diet and digestive capability, e.g. carnivores are more sensitive to bitterness than herbivores.¹⁵ Human diets high in fruits and vegetables are associated with a lower incidence of cancer and CVS disease and many of the phytochemicals believed to be responsible for these health improvements taste bitter.¹⁶ Humans have incorporated many bitter tasting foods into their diet, e.g. the vegetables: bitter melon, Brussel sprouts, globe artichoke and lettuce; the fermented foods: cheese, soya sauce and miso; as well as the drinks: coffee, tea, beer, wine and aperitifs.⁶ The preference for bitter substances is referred as the Schweppes effect.¹⁷

Postprandial hyperaemia

Food ingestion triggers activation of intrinsic and extrinsic neuronal circuits and the release of regularity peptides, biogenic amines and lipid mediators.¹⁸ The digestive process includes the production and release of digestive enzymes, regulation of gastric emptying (GE), intestinal motility and the absorption and removal of food-derived substances. An increase of the splanchnic blood flow, referred to as postprandial hyperaemia (PPH), supports these activities. Understanding of the PPH mechanisms is limited and studies of both food digestion and GIT neural and circulatory dynamics have often been conducted outside normal physiological boundaries.¹⁹PPH involves sympathetic activation of the CVS commonly leading to increases of heart rate (HR), cardiac output (CO), blood pressure (BP) and splanchnic vascular resistance. PPH in the small intestine may be limited to those areas where chyme is present.²⁰PPH may also involve resetting of the baroreflex sensitivity (BRS) so that BP is kept stable while HR increases ensuring that an adequate blood supply is available to support PPH demands. PPH, as measured by increases in blood velocity in the celiac and superior mesenteric arteries, begins within a minute of the ingestion of food.²¹Chewing and taste increase blood velocity in the celiac but the superior mesenteric arteries.²²

It has been suggested that TAS2Rs may release regulatory peptides and activate vagal reflex pathways to initiate defensive measures which effectively prolong the exposure time of potential toxins to digestive activity. These defensive measures include a decreased GE, decreased food intake and increased secretary activity of the stomach, gallbladder, pancreas and intestinal wall.²³ Subsequently there has developed a research interest to determine whether bitter agonists have a role in regulating levels of food intake and satiety.²⁴ Bitter agonists have been shown to regulate the secretion of ghrelin, modify food intake and slow GE in mice.²⁵ Also in mice, bitter agonists have been shown to stimulate the release of cholecystokinin (CCK),²⁶ which is an important regulator of CVS function.²⁷ CCK has shown to be induced from hTAS2R14 cells following exposure to a bitter glycoside from Hoodia gordonii²⁸ and increased CCK plasma levels have been reported 15 min after of the drinking of caffeinated coffee but not decaffeinated coffee. Additionally, both coffees decreased gall bladder volume with the caffeinated coffee producing a greater change.²⁹ Bitter agonists from Gentiana lutea have also been shown to have gastro-protective effects on experimentally induce gastric lesions in rats.¹⁴ In contrast to the above findings, the bitter agonists quinine and naringin did not affect GE or hunger/fullness scores in humans consuming liquid food.³⁰

Coffee and caffeine

Despite its bitter taste, coffee is a popular drink that is consumed multiple times daily by large numbers of adults worldwide. Coffee's bitter taste is due to a group of quinides formed during the roasting process, rather than caffeine which is present in both the unroasted and roasted bean.³¹ Consequently it is possible to produce a decaffeinated coffee which tastes the same as regular caffeinated coffee but is virtually caffeine free.³² Like caffeine, the quinides are bitter and may also stimulate both gustatory and GIT TAS2Rs. Caffeine is absorbed rapidly and completely from the gut. After the intake of caffeine in solution, the plasma level peaks at 30 min, with 75% of the maximum level occurring after only 15 min.³³Caffeine in capsules or tablets can be expected to be absorbed more slowly than when in solution. Gelatine capsules will release their caffeine content approximately 10 min after intake, whereas the release-time from tablets depends on the type of excipients present.

The impact of coffee, tea and caffeine on the CVS has been extensively researched and reviewed in the acute period,^4,34,35 30 to 120 min following intake when caffeine plasma levels peak, and with long-term use,^4,34–38 but less is known about the more immediate effects of coffee drinking. At the caffeine levels encountered with normal usage, caffeine has a pressor effect which is attributed to antagonism of the adenosine receptors A₁ and A_2A.⁴ The intake of either coffee or caffeine by non-users and caffeine abstainers produces increases of BP in the acute period. Following several days of ingesting large amounts of caffeine “caffeine tolerance” develops and virtually no pressor effects follow large caffeine interventions of 250 mg or more.^4,34 It has been suggested that the extent of tolerance to caffeine challenges is dependent on residual caffeine plasma levels.^4,39 However, there is also evidence that the extent of caffeine tolerance may be a function of the size of the intervention dose. Habitual coffee and tea drinkers, following their normal consumption pattern, experienced increases in BP following the intervention doses of 67 and 133 mg but not for the larger dose of 200 mg.⁴⁰

The long term effects of habitual coffee ingestion on the CVS are also not completely clarified. A meta-analysis of prospective cohort studies reported that there was an increased risk for hypertension among consumers drinking 1 to 3 cups of coffee daily but not for higher daily intakes.³⁸ Meta-analyses of cohort studies report that habitual coffee consumption does not increase the risk of coronary heart disease (CHD),^41,42 and in fact reduces it for some subgroups.⁴² Conversely, a meta-analysis of case-controlled studies reported that habitual coffee consumption does increase the risk of CHD.⁴¹ Three hypotheses have been advanced to account for these varying outcomes. Firstly bias/confounding,^41,42 secondly genetic variation leading to abnormally high caffeine plasma levels⁴³ and thirdly that coffee ingestion produces a short-term adverse effect on the heart that can trigger a coronary event.⁴ The genetic hypothesis is based on research showing that slow metabolisers of caffeine, due to the allele CYP1A2*1F, consuming either moderate or large amounts of coffee have an increased risk of nonfatal myocardial infarction (MI).⁴⁴ The short-term adverse effect hypothesis is based on research indicating that in the hour after drinking coffee the risk of nonfatal MI is increased.⁴⁵ In contrast, a recent review on coffee and cardiac arrhythmia suggested that although “many physicians recommend avoidance of caffeine in patients with heart disease” there is no reason to restrict coffee intake.⁴⁶ This latter view is consistent with the longstanding view that drinking coffee affects vasculature tonus but not HR.⁴⁷

Hot drinks

The present study is concerned with the effect of coffee drinking on the CVS in the 30 min following ingestion. Only two previous studies^48,49 have investigated the impact of caffeinated beverages on the CVS during and directly after intake, i.e. before caffeine plasma levels are elevated in the acute period. These studies used hot drinks (water, and coffee and tea both containing various amounts of caffeine) with volumes of 400 mL and 300 mL respectively. Compared to a no-drink control, these studies reported transitory (<10 min) increases in HR and BP that were larger for the caffeinated drinks. A potential limitation in these studies is that the relatively large volume ingested may have elicited large somatosensory responses which masked chemosensory responses to caffeine. An additional limitation is that the CVS readings were obtained at three-minute intervals and therefore may not have been sufficiently accurate to track parameter changes.

Methods

Participants

This investigation conforms to the principles outlined by the Declaration of Helsinki and was approved by the University of Westminster Ethics Committee (03/04-08). Written informed consent was obtained from the participants, who were healthy volunteers recruited from the staff, students and associates of the University of Westminster, London. Participants were selected as habitual coffee and tea drinkers (two to six servings daily) who enjoy drinking unadulterated espresso coffee. Participants were required to continue with their habitual caffeine consumption for the duration of the study and compliance was verbally confirmed prior to each session. Hypertensives (systolic pressure (SP) >140 mmHg or diastolic pressure (DP) >90 mm Hg), smokers, pregnant women and those on prescribed medication were excluded from the study.

Test substances

Espresso coffee was chosen as the test form because it is straightforward to produce as a standard serving and is traditionally consumed in small volumes. Regular and decaffeinated “medium roasted Columbia” whole coffee beans were used for testing. A double espresso of approximately 67 mL made from 16.5 g of beans was produced for both types of beans using an espresso machine. The espresso caffeine content was analysed using HPLC and estimated at 130 mg for regular coffee and <10 mg for decaffeinated coffee.

Two coffees and two gelatine capsules were administered. The test substances were presented in a randomized double-blind design and all participants attended all four sessions. Test sessions were conducted at similar times (±1 h) of the day so as to minimize diurnal effects. The coffees were administered during the first two test sessions and the capsules were administered in subsequent sessions: (1) regular coffee containing 130 mg caffeine; (2) decaffeinated coffee; (3) a capsule containing 134 mg caffeine; and (4) a capsule containing cellulose (placebo-control). The caffeine, cellulose and gelatine capsule conformed to British Pharmacopeia standards. The disintegration time for caffeine capsule was estimated at 10 min.

Procedure

Prior to an experimental session, participants were required to abstain from food and drink (excluding water) for two hours. Participants were seated in a quiet room at a temperature of 21 to 23 °C with a haemodynamic monitoring system (Finometer PRO, Finapres Medical Systems, Amsterdam, The Netherlands) attached to their left hand. The initial five minutes of the session involved calibrating the Finometer (Return-to-Flow) and allowing the participants' CVS to stabilize.⁵⁰ A pre-ingestion recording of 130 s was obtained while sitting, after which the hot drink or capsule was ingested. Drinking was allowed to proceed at a pace determined by the participant so as to simulate normal usage and extended up to seven minutes. Capsules were administered with 67 mL of room-temperature water so that fluid intake was constant. Participants sat for 30 min post-ingestion before standing for three and a half minutes. The test sessions were conducted in silence, with the participants being permitted to read but not write while seated.

Cardiovascular measurements

The Finometer records the finger-pulse contour, providing continuous beat-to-beat measures of a number of CVS parameters. An infrared plethysmograph in a finger cuff records (200 Hz) the pulsation of the arterial diameter. Cuff pressure clamps the artery's unstretched diameter and is attuned so that finger arterial pressure is reflected in the cuff pressure. As well as providing enhanced sensitivity due to the availability of continuous BP readings,^51,52 the Finometer also provides CVS measures that are useful for assessing changes of the ANS. The measures include HR, heart period (HP), ejection time (ET), dp/dt (contraction force) and brachial SP and DP. Modelflow software, which is integrated into the Finometer system, automatically computes stroke volume (SV), cardiac output (CO), peripheral resistance (PR), arterial compliance (AC) and aortic (characteristic) impedance (AI), as well as the body-surface area adjusted values (Dubois and Dubois formula): “indexed stroke volume” (SVI), “indexed cardiac output” (COI) or the “cardiac index” and “indexed peripheral resistance” (PRI) values.^53–57 The indexed values are reported rather than the non-indexed values of SV, CO and PR values. Additionally, the diastolic interval (DI) (HP minus ET) was calculated in the statistical analysis program.

Baroreflex measures

Measures of spontaneous BRS were derived from the continuous recordings of HP and SP measures. The geometric mean was calculated in the time domain using a cross-correlation function (xBRS program v1.5.0.3 (BMEYE BV, Amsterdam, The Netherlands)) with standard settings (p = 0.01). The xBRS program has previously been validated against the EUROBAR database.⁵⁸

Analysis

Mean parameter values were extracted from the continuous recording for a pre-ingestion measure and for four functionally distinct post-ingestion measures. The post-ingestion phases were: Phase 1: 0 to 5 min; Phase 2: 10 to 15 min; Phase 3: 25 to 30 min; Phase 4: 31.5 to 33.5 min. Phase 1 commenced 30 s after swallowing. During pre-ingestion and Phases 1, 2 and 3 the participants were seated while in Phase 4 participants were standing. The period 30 to 31.5 min was allowed for body movement and the orthostatic reaction.

It was anticipated that Phase 1 would include any immediate responses resulting from stimulation of either somatosensory or bitter chemosensory receptors, located in either the oral cavity or GIT. Previous research indicates that CVS responses to the intake of hot drinks, including water, coffee and tea (regardless of their caffeine content) last for less than 10 min.⁴⁸ During Phase 2, unabsorbed coffee fractions would be present in the gut. Responses initiated in this phase would be chiefly the result of GIT TAS2R activation. As the disintegration time for the caffeine capsule was estimated at 10 min, it was anticipated that the GIT receptors would be maximally exposed to caffeine in this phase. During Phases 3 and 4 it was anticipated that most of the caffeine will have been absorbed from the gut³³ and responses initiated in this phase would be largely the result of elevated plasma caffeine levels. Participants were tested in two postures as previous work indicates that caffeine's effect on the vascular structures is modulated by posture.⁴⁰

Planned contrasts were used to compare the pre- and post-ingestion measures of the coffee conditions (Phases 1, 2, 3 and 4) and the caffeine capsule (Phases 2, 3 and 4), with the placebo-control capsule measures using within-participants Repeated Measures (2 × 2) ANOVA (SPSS v15, Chicago, USA). The significance level was set at p < 0.05 and the number of planned contrasts were equal to the number of degrees of freedom.

Results

Participant characteristics

The 12 female participants had a mean (standard deviation and range) age of 42 (±12.3, 21 to 61) years, a mean weight of 65 (±13.8, 45 to 98) kg, a mean height of 1.63 (±0.07, 1.52 to 1.74) m and a mean BMI of 25 (±5.6, 19.5 to 39.3) kg m⁻¹. Accordingly, the caffeine dose for both the regular coffee and the caffeine capsule was approximately 2 mg kg⁻¹. Two participants failed to complete one session each: one needed to visit the lavatory during the decaffeinated coffee session; the other participant experienced heart palpitations 15 min after ingesting the caffeine capsule. Consequently, data is omitted for these participants for Phases 2, 3 and 4. For the caffeine capsules the participant characteristics changed to: mean age 43 (±12.8, 21 to 61) years, a mean weight 62 (±9.6, 45 to 78) kg, mean height of 1.63 (±0.08, 1.52 to 1.74) m and mean BMI of 23 (±3.0, 19.5 to 28.8) kg m⁻¹.

The CVS parameter measures are presented in Table 1. Change values for significant parameter differences are given below. Change values were calculated as test substance change from baseline minus placebo change from baseline. Note: some of these values cannot be calculated directly from the table due to the missing data.

Table 1 Mean and standard deviation measures of the cardiovascular parameters for pre-ingestion and post-ingestion phases. Pre- and post-ingestion measures of the coffee conditions (Phases 1, 2, 3 and 4) and the caffeine capsule (Phases 2, 3 and 4) were compared with the placebo-control capsule measures using within-participants Repeated Measures ANOVA; * p < 0.05, ** p < 0.01, *** p < 0.001.^a

Parameter (unit)	Test	Pre-ingestion	Phase 1, 0 to 5 min	Phase 2, 10 to 15 min	Phase 3, 25 to 30 min	Phase 4, upright
a PC = placebo-control capsule, DC = decaffeinated coffee, RC = regular coffee, CC = caffeine capsule, HR = heart rate, DI = diastolic interval, ET = ejection time, SVI = indexed stroke volume, COI = indexed cardiac output, SP = systolic pressure, DP = diastolic pressure, AC = arterial compliance, AI = aortic impedance, PRI = indexed peripheral resistance, BRS = spontaneous baroreflex sensitivity, bpm = beats per minute, Lpm = liters per minute, MU = medical unit = mmHg s ml^−1.
HR (bpm)	PC	67.6 ± 5.3	66.5 ± 4.6	66.0 ± 4.3	66.0 ± 4.1	75.4 ± 6.6
	DC	69.2 ± 7.6	70.7 ± 7.5**	70.0 ± 8.0	70.1 ± 6.8	76.0 ± 9.2
	RC	68.3 ± 7.0	71.6 ± 7.2***	70.4 ± 6.7**	71.1 ± 6.4*	77.8 ± 7.2
	CC	65.7 ± 6.5	—	65.5 ± 6.0	65.5 ± 5.8	71.6 ± 4.3
DI (ms)	PC	585 ± 60	600 ± 54	604 ± 51	607 ± 46	526 ± 54
	DC	575 ± 76	555 ± 71**	568 ± 78	567 ± 68	529 ± 69
	RC	582 ± 82	548 ± 75***	560 ± 68**	553 ± 63*	507 ± 55
	CC	609 ± 64	—	612 ± 64	612 ± 62	557 ± 39
ET (ms)	PC	312 ± 15	315 ± 14	312 ± 15	310 ± 16	280 ± 19
	DC	307 ± 20	307 ± 21*	304 ± 23	301 ± 21	274 ± 25
	RC	310 ± 17	304 ± 18***	304 ± 19*	302 ± 18	274 ± 20
	CC	314 ± 21	—	315 ± 23	314 ± 22	288 ± 17
dp/dt (mmHg s^−1)	PC	904 ± 257	944 ± 228	935 ± 229	963 ± 218	1010 ± 191
	DC	904 ± 166	891 ± 145	914 ± 156	967 ± 162	1057 ± 200
	RC	827 ± 170	854 ± 166	899 ± 159	935 ± 141	1041 ± 168*
	CC	910 ± 185	—	969 ± 176	968 ± 231	1062 ± 192
SVI (ml m^−2)	PC	42.4 ± 10.2	43.2 ± 10.1	42.2 ± 8.9	41.7 ± 8.2	36.4 ± 8.2
	DC	43.2 ± 10.1	42.4 ± 8.8	41.6 ± 9.3	40.6 ± 8.4	38.5 ± 8.8
	RC	39.7 ± 8.5	39.3 ± 8.3	39.6 ± 7.4	38.7 ± 7.4	36.1 ± 7.9
	CC	42.1 ± 8.2	—	41.0 ± 7.3	41.0 ± 7.1	37.6 ± 7.7
COI (Lpm m^−2)	PC	2.85 ± 0.71	2.85 ± 0.67	2.77 ± 0.56	2.73 ± 0.50	2.71 ± 0.52
	DC	2.94 ± 0.67	2.95 ± 0.58	2.86 ± 0.60	2.81 ± 0.54	2.89 ± 0.62
	RC	2.69 ± 0.61	2.79 ± 0.61	2.76 ± 0.54	2.73 ± 0.53	2.79 ± 0.61
	CC	2.74 ± 0.49	—	2.66 ± 0.41	2.65 ± 0.37	2.67 ± 0.48
SP (mmHg)	PC	124.0 ± 11.7	125.8 ± 10.9	124.5 ± 10.0	125.8 ± 11.6	123.7 ± 12.1
	DC	122.7 ± 8.8	123.2 ± 8.4	124.5 ± 9.1	127.5 ± 11.0	127.6 ± 13.9
	RC	121.1 ± 8.9	122.5 ± 8.2	123.2 ± 8.0	124.5 ± 9.3	126.5 ± 12.6
	CC	124.5 ± 12.5	—	127.6 ± 10.4	128.0 ± 11.2	130.2 ± 13.9
DP (mmHg)	PC	74.4 ± 6.6	74.9 ± 7.4	74.6 ± 7.1	75.2 ± 6.9	76.3 ± 6.6
	DC	74.9 ± 6.3	75.9 ± 6.1	76.5 ± 5.4	78.3 ± 5.9	78.5 ± 6.8
	RC	75.7 ± 4.7	76.8 ± 4.5	76.5 ± 4,4	77.6 ± 4.9	79.3 ± 6.1
	CC	74.5 ± 7.2	—	76.4 ± 5.4*	76.8 ± 5.1	79.1 ± 7.5*
AC (MU)	PC	1.71 ± 0.36	1.69 ± 0.38	1.70 ± 0.35	1.69 ± 0.36	1.65 ± 0.37
	DC	1.68 ± 0.41	1.65 ± 0.39	1.62 ± 0.41	1.57 ± 0.43	1.58 ± 0.45
	RC	1.68 ± 0.41	1.66 ± 0.39	1.66 ± 0.40	1.63 ± 0.36	1.59 ± 0.44
	CC	1.67 ± 0.37	—	1.61 ± 0.35*	1.60 ± 0.34	1.55 ± 0.39*
AI (mMU)	PC	65.6 ± 7.1	65.8 ± 7.4	65.6 ± 6.9	65.9 ± 7.3	66.4 ± 7.6
	DC	66.1 ± 8.1	66.4 ± 8.2	66.7 ± 8.9	67.6 ± 9.5	67.8 ± 9.9
	RC	66.0 ± 8.3	66.3 ± 8.5	66.4 ± 8.4	66.7 ± 8.5	67.7 ± 9.2
	CC	66.9 ± 6.9	—	67.6 ± 6.3	67.6 ± 6.5	69.0 ± 7.3*
PRI (MU m^−2)	PC	0.75 ± 0.20	0.76 ± 0.20	0.76 ± 0.17	0.77 ± 0.17	0.78 ± 0.18
	DC	0.72 ± 0.19	0.72 ± 0.18	0.75 ± 0.20	0.78 ± 0.20	0.76 ± 0.22
	RC	0.79 ± 0.17	0.77 ± 0.19	0.77 ± 0.18	0.79 ± 0.19	0.79 ± 0.22
	CC	0.79 ± 0.21	—	0.83 ± 0.18	0.83 ± 0.20	0.85 ± 0.23
BRS (ms mmHg^−1)	PC	10.7 ± 5.2	10.9 ± 4.5	11.3 ± 5.0	10.8 ± 6.1	6.9 ± 2.8
	DC	11.7 ± 6.6	11.0 ± 5.9	10.8 ± 5.5	10.3 ± 5.7	7.3 ± 4.0
	RC	11.6 ± 5.7	11.0 ± 4.7	10.5 ± 4.4	10.3 ± 4.3	7.2 ± 3.2
	CC	11.1 ± 5.1	—	12.6 ± 6.5	10.9 ± 4.7	7.4 ± 3.2

---

Hot coffee

Regular coffee in Phases 1, 2, 3 and 4, and decaffeinated coffee in Phase 1 elicited cardiac responses while vascular parameters were unaffected. Increases in HR resulted from decreases of both DI and ET although the DI was most affected. For regular coffee the magnitude of the DI decreases and the HR increases were similar in Phases 1, 2 and 3.

Regular coffee. Phase 1: HR increased by 4.4 beats per minute (bpm) (F (1, 11) = 37.169, p < 0.001), DI decreased by 46 ms (F (1, 11) = 26.092, p < 0.001) and ET decreased by 8 ms (F (1, 11) = 46.251, p < 0.001). Phase 2: HR increased by 3.6 bpm (F (1, 11) = 13.992, p = 0.003) DI decreased by 42 ms (F (1, 11) = 12.264, p = 0.005) and ET by 5 ms (F (1, 11) = 4.965 p = 0.048). Phase 3: HR increased by 4.4 bpm (F (1, 11) = 7.628, p = 0.018) and DI decreased by 52 ms (F (1, 11) = 8.166, p = 0.016). Phase 4: dp/dt increased by 107 mmHg s⁻¹ (F (1, 11) = 6.246, p = 0.030).

Decaffeinated coffee. Phase 1: HR increased by 2.7 bpm (F (1, 11) = 12.877, p = 0.004), DI decreased by 31 ms (F (1, 11) = 11.877, p = 0.005) and ET decreased by 3ms (F (1, 11) = 5.714, p = 0.036).

Caffeine capsule

The caffeine capsule elicited vascular but not cardiac responses in Phases 2 and 4. Phase 2: DP increased by 2.0 mmHg (F (1, 10) = 5.148, p = 0.047) and AC decreased by 0.07 MU (F (1, 10) = 6.422, p = 0.030). Phase 4: DP increased by 2.9 mmHg (F (1, 10) = 6.584, p = 0.028), AC decreased by 0.09 MU (F (1, 10) = 5.475, p = 0.041) and AI increased by 1.3 mMU (F (1, 10) = 5.572, p = 0.040).

Correlation analysis

As there was a large range in age and BMI, correlations between these participant characteristics and the placebo-controlled change form baseline for the significant HR and AC changes were calculated. No correlations were significant.

Discussion

The present study has produced two novel findings: (1) the intake of caffeinated (≈130 mg) hot coffee elicits HR increases lasting 30 min; and (2) the intake of 134 mg encapsulated caffeine elicits short-lived vascular responses 10 to 15 min after ingestion. The difference between the present findings and the widely-held view, that caffeine affects the vascular system but not the heart,⁴⁷ is most likely due to the paucity of caffeine research in the pre-acute period (<30 min). Whereas the present study utilised continuous recording of CVS parameters, the two previous studies^48,49 investigating coffee and caffeine's effects in this period used intermittent (three minute) CVS recording. These intermittent recordings may not have been sufficiently sensitive to detect coffee's effects.

Coffee's impact on heart rate

The drinking of decaffeinated coffee produced an immediate and brief (Phase 1 only) increase in HR similar in length to that reported after the drinking of hot water.^48,49 These responses are likely to have been elicited by gustatory and/or GIT somatosensory receptors and do not appear to involve coffee's bitter quinides. In contrast to the decaffeinated coffee, regular coffee produced prolonged increases in HR (Phases 1, 2 and 3) which can be attributed to caffeine, as this is the major difference between the two coffees. Caffeine has the potential to stimulate both gustatory² and GIT³ TAS2Rs and elicit ANS responses. The finding, that the caffeine capsule did not elicit cardiac responses during Phases 2 and 3, indicates that it is caffeine's stimulation of the gustatory TAS2Rs, occurring during ingestion of regular coffee, rather than post-ingestion stimulation of the GIT TAS2Rs that elicited the ANS response. This ANS response includes, but is not necessarily limited to changes in cardiac activity. Although there is the possibility that the observed HR increases were the result of altered baroreceptor activity, acting to maintain a constant SP, this possibility can be excluded as there were no changes in the BRS values. The increase of dp/dt in Phase 4 only suggests that the ANS response elicited by regular coffee may vary with posture. However, this study has not fully investigated this possibility.

Previous studies on hot drinks have reported temporary HR increases of 10.5⁴⁸ and 6.0⁴⁹ bpm as well as BP increases. During Phase 1 the increases of HR were considerably smaller, 2.4 to 4.4 bpm and vascular parameters were unchanged. The differences in the present findings, using 67 mL, and the previous studies, which used 400⁴⁸ and 300⁴⁹ mL, may be due to the differing volumes of the test drink. Larger drink volumes can be expected to produce greater somatosensory stimulation and consequently elicit larger automatic responses. Thus the current report is likely to provide more accurate information about the effect of regular and decaffeinated coffee on the ANS and CVS than the previous studies. More research is required to define the relationship between the serving size, caffeine concentration and CVS responses.

The finding that coffee affects heart activity for at least 30 min after ingestion provides a plausible pharmacological explanation to account for the findings that CHD risk increases within an hour of ingesting coffee⁴⁵ and that long-term coffee consumption increases the risk of CHD.⁴¹ Elevated HR reduces myocardial oxygen consumption and energy utilisation as well as reducing diastolic coronary perfusion and may trigger ischaemic events.⁵⁹ Elevated resting HR is known to be an independent risk factor for CVS mortality and morbidity,⁵⁹ and there is evidence that CHD risk is reduced by the pharmaceutical treatment of patients with resting HR of 70 bpm or more.⁶⁰ What is unanswerable from the present findings is whether the observed increase in HR of 4 to 5 bpm is dose dependent or whether the magnitude of HR increases resulting from coffee drinking are dependent on pre-ingestion HR. Other questions regarding the impact of coffee ingestion include: what is the maximal HR increase; how long do HR increases last; whether HR increases vary with posture; whether similar HR increases occur during physical activities; and whether HR increases are additive to other HR increases resulting for physiological and psychological arousal. While the current finding that coffee did not affect HR but did affect dp/dt in Phase 4 suggest that coffee's impact may vary with posture, this finding cannot be generalised to the earlier post-ingestion periods. It well be that the coffee's effect on the ANS is declining by Phase 4. On the other hand, if caffeine's effect on HR is due to vagal withdrawal, as discussed below, then it may be that the vagal withdrawal resulting from two sources, the change to the upright posture and the effect of caffeine are not additive. Whether the finding that coffee ingestion elicits HR increases in the period when coffee consumers are most at risk for CHD is of clinical significance is yet to be determined.

Caffeine's impact on the vasculature

The caffeine capsule impacted on the vascular system during Phases 2 and 4. Phase 2 corresponds to the period when it was anticipated that the GIT caffeine receptors would be most exposed to caffeine. These parameter changes were short-lived (Phase 2 only) and occurred in the period when caffeine was present in the gut and before entering the plasma (Phase 3). It was not expected that a single caffeine capsule would produce a PPH CVS response similar to that of a large meal with increases of CO, BP and skeletal muscular resistance. However, the current finding that a CVS response, involving a decrease of AC and an increase of DP, was elicited during Phase 2 can be regarded as a likely PPH response²⁰ because it produces vascular changes that can enhance splanchnic circulation. These vascular changes are also in accordance with the hypotheses of how bitters may influence digestion.¹⁸ On the other hand, the decrease in AC and increases in DP and AI in Phase 4, which is the beginning of the acute period when little if any caffeine is in the GIT, are more likely due to caffeine's effect on the arterial adenosine receptors. These parameter changes in Phase 4 but not Phase 3, both of which are at the start of the acute-period, are consistent with our previous findings that the extent of the impact of caffeine varies with posture and is greater upright.⁴⁰

Regulatory peptides

CCK is released from entoeroendocrine cells in the gut wall in response to a meal and triggers a local PPM.²⁷ The magnitude of increases in plasma CCK, 15 min after drinking caffeinated coffee,²⁹ are similar to those increases experienced following a light meal.⁶¹ However, in the present study regular coffee did not produce vascular changes associated with PPM whereas the same amount of caffeine, when ingested in as the caffeine capsule, did produce vascular changes. It may be that it is the local concentration of caffeine, following the capsule's disintegration, which elicited decrease of AC rather than the total amount of caffeine ingested. More research is needed to define how caffeinated beverages influence the regulatory peptides.

Caffeine's effect on arousal

Caffeine is the most widely consumed drug in the world. It is the principal pharmacological human source of arousal and wakefulness⁶² and research involving the use of caffeine to affect mental performance has attracted a wide audience.^63,64 Mental performance tasks are commonly undertaken to assess an individual's level of psychological arousal while CVS measures serve as an indicator of physiological arousal. Mental arousal increases have been reported in the period 10 to 30 min after drinking caffeinated coffee, but only at 10 min after drinking decaffeinated coffee.⁶⁵ Mental arousal levels have also been reported to increase at 10, 30 and 60 min following the ingestion of caffeinated tea.⁴⁹ The pattern of HR changes, in this study, matches the pattern of changes in mental arousal elicited by the two types of coffee⁶⁵ and caffeinated tea.⁴⁹ As increases in both HR and arousal occurred before caffeine plasma levels are greatly elevated, we suggest that in the pre-acute period, it is the chemosensory stimulation of the gustatory TAS2Rs by the caffeine in beverages that elicits both the mental and physiological arousal. This suggestion is novel, as customarily caffeine arousal research focuses on the acute period where caffeine's plasma levels inhibit adenosine receptors.⁶³ Accordingly, arousal from caffeinated drinks may result from two distinct sources: in the pre-acute period from caffeine's stimulation of the gustatory TAS2Rs and in the acute period from caffeine's blockage of adenosine receptors.

While increases of HR may be attributed to either sympathetic activation or vagal withdrawal or a combination of both, the lack of change in the largely sympathetically controlled parameters dp/dt, AC and PR⁶⁶ following coffee ingestion suggests that the predominant elicited ANS response involves vagal withdrawal rather than sympathetic activation.

Coffee versuscaffeine

Although thousands of compounds have been reported present in roasted coffee beans, it is widely accepted that the major active ingredient of coffee is caffeine.⁶⁷ The present study supports this assertion with regard to drinking coffee. The 30 min long ANS changes, resulting from the drinking of regular coffee, were due to caffeine as decaffeinated coffee produced only brief ANS changes. What is less clear is why the caffeine in coffee did not produce the same vascular responses as a similar amount of encapsulated caffeine. As noted above, this may have to do with caffeine concentration in the gut but it may also result from other unrecognised factors. Consequently research findings on caffeine are not directly applicable to coffee and vice versa.

The findings infer that novel types of caffeinated products may provide novel patterns of arousal. When caffeine is taken to enhance mental performance, products where the caffeine is tasted may be more effective. Such products include “energy drinks”,⁶⁸ “food/energy bars” and chewing gum.⁶⁴

Limitations

The current participants were restricted to healthy, habitual coffee and tea drinkers who enjoy the taste of coffee without the addition of sugar or milk products. The number of participants was relatively small and research with larger groups is required to characterize the chemosensory impact of coffee and caffeine on the ANS and CVS. Furthermore, residual caffeine levels were not measured and so the observed changes can only be related to the differences in the drinks administered and not to plasma caffeine levels. Additionally, we did not test with preparations of hot water, with and without added caffeine, or decaffeinated coffee with added caffeine or cold coffee.

Caffeine's ANS impact via gustatory TAS2Rs is likely to be influenced by many factors that have not been addressed in this study. These factors include the type of coffee preparation, the caffeine level, the caffeine concentration or the addition of flavourings (sugar, dairy products, etc.). Care should be taken in extrapolating the current results to other groups, particularly those choosing not to use or who actively avoid caffeinated beverages. Notably, non-caffeine users have been reported to have a lower average detection threshold than habitual caffeine users (0.5 versus 1.2 mol L⁻¹).⁶⁹ Other less-healthy groups may also react differently, particularly groups with impaired CVS or ANS function. In particular, clarification is needed on the effect of drinking coffee by individuals at risk for CHD.

Conclusion

The impact of caffeinated beverages and pharmaceutical caffeine, at commonly consumed levels, on humans is not limited to antagonism of the adenosine receptors resulting from elevated caffeine plasma levels. It is also includes ANS and CVS responses elicited by caffeine's stimulation of the gustatory and GIT TAS2Rs. The current findings may be useful in establishing how coffee and tea produce arousal and help to explain the universal popularity of these bitter beverages. The results suggest that the increases in heart activity during the 30 min period following the intake of caffeinated beverages are primarily due to parasympathetic withdrawal rather than increased sympathetic activity. This finding provides a pharmacological mechanism that can be useful in investigations examining the possible link between coffee drinking and CHD. Furthermore, it is likely that substances, outside of the present study, may also modulate ANS and CVS activity through chemosensory stimulation.

References

1. K. L. Durrant, J. Am. Pharm. Assoc., 2002, 42, 625–637.
2. W. Meyerhof, C. Batram, C. Kuhn, A. Brockhoff, E. Chudoba, B. Bufe, G. Appendino and M. Behrens, Chem. Senses, 2010, 35, 157–170.
3. I. Masuho, M. Tateyama and O. Saitoh, Chem. Senses, 2005, 30, 281–290.
4. N. P. Riksen, G. A. Rongen and P. Smits, Pharmacol. Ther., 2009, 121, 185–191.
5. K. L. Mueller, M. A. Hoon, I. Erlenbach, J. Chandrashekar, C. S. Zuker and N. J. P. Ryba, Nature, 2005, 434, 225–229.
6. K. Izawa, Y. Amino, M. Kohmura, Y. Ueda and M. Kuroda, in Comprehensive Natural Products II, ed. M. Lew and L. Hung-Wen, Elsevier, Oxford, 2010, pp. 631–671.
7. S. V. Wu, N. Rozengurt, M. Yang, S. H. Young, J. Sinnett-Smith and E. Rozengurt, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 2392–2397.
8. M. Tizzano, B. D. Gulbransen, A. Vandenbeuch, T. R. Clapp, J. P. Herman, H. M. Sibhatu, M. E. A. Churchill, W. L. Silver, S. C. Kinnamon and T. E. Finger, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 3210–3215.
9. D. A. Deshpande, W. C. H. Wang, E. L. McIlmoyle, K. S. Robinett, R. M. Schillinger, S. S. An, J. S. K. Sham and S. B. Liggett, Nat. Med., 2010, 16, 1299–U1298.
10. N. Singh, M. Vrontakis, F. Parkinson and P. Chelikani, Biochem. Biophys. Res. Commun., 2011, 406, 146–151.
11. J. Fehr, D. Meyer, P. Widmayer, H. C. Borth, F. Ackermann, B. Wilhelm, T. Gudermann and I. Boekhoff, J. Comp. Physiol., A, 2007, 193, 21–34.
12. S. P. Travers and L. C. Geran, Physiol. Behav., 2009, 97, 592–603.
13. S. Z. Hao, C. Sternini and H. E. Raybould, Am. J. Physiol.: Regul. Integr. Comp. Physiol., 2008, 294, R33–R38.
14. Y. Niiho, T. Yamazaki, Y. Nakajima, T. Yamamoto, H. Ando, Y. Hirai, K. Toriizuka and Y. Ida, J. Nat. Med., 2006, 60, 82–88.
15. J. I. Glendinning, Physiol. Behav., 1994, 56, 1217–1227.
16. A. Drewnowski and C. Gomez-Carneros, Am. J. Clin. Nutr., 2000, 72, 1424–1435.
17. J. I. Glendinning, Physiol. Behav., 1993, 54, 141–150.
18. G. J. Dockray, Int. Dairy J., 2010, 20, 226–230.
19. D. M. Sartor and A. J. M. Verberne, Brain Res. Rev., 2008, 59, 140–154.
20. P. J. Matheson, M. A. Wilson and R. N. Garrison, J. Surg. Res., 2000, 93, 182–196.
21. N. Someya, M. Y. Endo, Y. Fukuba and N. Hayashi, Am. J. Physiol.: Regul. Integr. Comp. Physiol., 2008, 294, R1790–R1796.
22. N. Someya and N. Hayashi, Am. J. Physiol.: Regul. Integr. Comp. Physiol., 2008, 295, R1921–R1925.
23. E. Rozengurt and C. Sternini, Curr. Opin. Pharmacol., 2007, 7, 557–562.
24. M. C. P. Geraedts, F. J. Troost and W. H. M. Saris, Obes. Rev., 2011, 12, 470–477.
25. S. Janssen, J. Laermans, P. J. Verhulst, T. Thijs, J. Tack and I. Depoortere, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 2094–2099.
26. M. C. Chen, S. V. Wu, J. R. Reeve and E. Rozengurt, Am. J. Physiol.: Cell Physiol., 2006, 291, C726–C739.
27. T. A. Lovick, J. Chem. Neuroanat., 2009, 38, 176–184.
28. B. Le Neve, M. Foltz, H. Daniel and R. Gouka, Am. J. Physiol.: Gastrointest. Liver Physiol., 2010, 299, G1368–G1375.
29. B. R. Douglas, J. Jansen, R. Tham and C. Lamers, Am. J. Clin. Nutr., 1990, 52, 553–556.
30. T. J. Little, N. Gupta, R. M. Case, D. G. Thompson and J. T. McLaughlin, Am. J. Physiol.: Regul. Integr. Comp. Physiol., 2009, 297, R632–R639.
31. O. Frank, G. Zehentbauer and T. Hofmann, Eur. Food Res. Technol., 2006, 222, 492–508.
32. W. Heilmann, in Coffee: Recent Developments, ed. R. J. Clarke and O. G. Vitzthum, Blackwell Sciences Ltd, Oxford, 2001, ch. 5, pp. 108–124.
33. J. Blanchard and S. J. A. Sawers, Eur. J. Clin. Pharmacol., 1983, 24, 93–98.
34. M. L. Nurminen, L. Niittynen, R. Korpela and H. Vapaatalo, Eur. J. Clin. Nutr., 1999, 53, 831–839.
35. P. Nawrot, S. Jordan, J. Eastwood, J. Rotstein, A. Hugenholtz and M. Feeley, Food Addit. Contam., 2003, 20, 1–30.
36. M. Noordzij, C. Uiterwaal, L. R. Arends, F. J. Kok, D. E. Grobbee and J. M. Geleijnse, J. Hypertens., 2005, 23, 921–928.
37. J. M. Geleijnse, Vasc Health Risk Manag, 2008, 4, 963–970.
38. Z. Z. Zhang, G. Hu, B. Caballero, L. Appel and L. W. Chen, Am. J. Clin. Nutr., 2011, 93, 1212–1219.
39. P. Smits, T. Thien and A. Vantlaar, Am. J. Cardiol., 1985, 56, 958–963.
40. M. K. McMullen, J. M. Whitehouse, G. Shine and A. Towell, Food Funct., 2011, 2, 197–203.
41. F. Sofi, A. A. Conti, A. M. Gori, M. L. Eliana Luisi, A. Casini, R. Abbate and G. F. Gensini, Nutr., Metab. Cardiovasc. Dis., 2007, 17, 209–223.
42. J. N. Wu, S. C. Ho, C. Zhou, W. H. Ling, W. Q. Chen, C. L. Wang and Y. M. Chen, Int. J. Cardiol., 2009, 137, 216–225.
43. T. Celik, A. Iyisoy and B. Amasyali, Int. J. Cardiol., 2010, 144, 118–118.
44. M. C. Cornelis, A. El-Sohemy, E. K. Kabagambe and H. Campos, JAMA, J. Am. Med. Assoc., 2006, 295, 1135–1141.
45. A. Baylin, S. Hernandez-Diaz, E. K. Kabagambe, X. Siles and H. Campos, Epidemiology, 2006, 17, 506–511.
46. D. J. Pelchovitz and J. J. Goldberger, Am. J. Med., 2011, 124, 284–289.
47. T. G. Pickering, in Comprehensive Hypertension, ed. G. Y. H. Lip and J. E. Hall, Mosby, Philadephia, 2007, pp. 521–534.
48. P. T. Quinlan, J. Lane and L. Aspinall, Psychopharmacology, 1997, 134, 164–173.
49. P. T. Quinlan, J. Lane, K. L. Moore, J. Aspen, J. A. Rycroft and D. C. O'Brien, Pharmacol., Biochem. Behav., 2000, 66, 19–28.
50. T. G. Pickering, J. E. Hall, L. J. Appel, B. E. Falkner, J. Graves, M. N. Hill, D. W. Jones, T. Kurtz, S. G. Sheps and E. J. Roccella, Hypertension, 2005, 45, 142–161.
51. G. J. Langewouters, J. J. Settels, R. Roelandt and K. H. Wesseling, J. Med. Eng. Technol., 1998, 22, 37–43.
52. I. Guelen, B. E. Westerhof, G. L. van der Sar, G. A. van Montfrans, F. Kiemeneij, K. H. Wesseling and W. J. W. Bos, J. Hypertens., 2008, 26, 1321–1327.
53. B. P. M. Imholz, W. Wieling, G. A. van Montfrans and K. H. Wesseling, Cardiovasc. Res., 1998, 38, 605–616.
54. J. J. van Lieshout and K. H. Wesseling, Br. J. Anaesth., 2001, 86, 467–469.
55. J. R. C. Jansen, J. J. Schreuder, J. P. Mulier, N. T. Smith, J. J. Settels and K. H. Wesseling, Br. J. Anaesth., 2001, 87, 212–222.
56. I. Guelen, B. E. Westerhof, G. L. van der Sar, G. A. van Montfrans, F. Kiemeneij, K. H. Wesseling and W. J. W. Bos, Blood Pressure Monit., 2003, 8, 27–30.
57. J. J. van Lieshout, K. Toska, E. J. van Lieshout, M. Eriksen, L. Walloe and K. H. Wesseling, Eur. J. Appl. Physiol., 2003, 90, 131–137.
58. B. E. Westerhof, J. Gisolf, W. J. Stok, K. H. Wesseling and J. M. Karemaker, J. Hypertens., 2004, 22, 1371–1380.
59. A. Hjalmarson, Eur. Heart J. Suppl., 2007, 9, F3–F7.
60. K. Fox, I. Ford, P. G. Steg, M. Tendera and R. Ferrari, Lancet, 2008, 372, 807–816.
61. F. C. Leslie, D. G. Thompson, J. T. McLaughlin, A. Varro, G. J. Dockray and B. K. Mandal, QJM, 2003, 96, 870–871.
62. B. D. Smith, T. White and R. Shapiro, in Caffeine and Activation Theory, ed. B. D. Smith, U. Gupta and B. S. Gupya, CRC Press, Taylor and Francis Group, Boca Raton, USA, 2007, pp. 9–42.
63. R. J. Barry, A. R. Clarke, S. J. Johnstone and J. A. Rushby, Biol. Psychol., 2008, 77, 304–316.
64. J. E. Vanderveen, L. E. Arnmstrong, G. E. Butterfield, W. L. Chenoweth, J. T. Dwyer, J. D. Fernstrom, R. B. Kanarek, O. A. Levander and E. M. Sternberg, Caffeine for the Sustainment of Mental Task Performance, National Academy of Sciences, Washington, D.C., 2001.
65. A. Adan, G. Prat, M. Fabbri and M. Sànchez-Turet, Progress Neuro-Psychopharmacol. Biol. Psychiatry, 2008, 32, 1698–1703.
66. R. E. Klabunde, Cardiovascular Physiology Concepts, Lippincott Williams & Wilkins, Philadelphia, 2005.
67. I. Flament and Y. Bessiere-Thomas, Coffee Flavor Chemistry, John Wiley & Sons, Ltd, Cichester, 2002.
68. C. J. Reissig, E. C. Strain and R. R. Griffiths, Drug Alcohol Depend., 2009, 99, 1–10.
69. D. J. Mela, R. D. Mattes, S. Tanimura and M. R. Garciamedina, Pharmacol., Biochem. Behav., 1992, 43, 513–521.

---

Footnote

† Sources of support/funding: no external funding or sponsoring was involved in this study. Caffeine and placebo capsules were generously prepared and donated by Dr Brian Whitton of Whitehorse Nutriceuticals.

---