Original article 427

The combination of vitamin C and grape-seed polyphenolsincreases blood pressure: a randomized, double-blind,placebo-controlled trialNatalie C. Ward, Jonathan M. Hodgson, Kevin D. Croft, Valerie Burke,Lawrence J. Beilin and Ian B. Puddey

Background There is growing evidence that oxidative

stress contributes to the pathogenesis of hypertension and

endothelial dysfunction. Thus, dietary antioxidants may

beneficially influence blood pressure (BP) and endothelial

function by reducing oxidative stress.

Objective To determine if vitamin C and polyphenols, alone

or in combination, can lower BP, improve endothelial

function and reduce oxidative stress in hypertensive

individuals.

Design A total of 69 treated hypertensive individuals with a

mean 24-h ambulatory systolic blood pressure

>— 125 mmHg participated in a randomized, double-blind,

placebo-controlled, factorial trial. Following a 3-week

washout, participants received 500 mg/day vitamin C,

1000 mg/day grape-seed polyphenols, both vitamin C and

polyphenols, or neither for 6 weeks. At baseline and post-

intervention, 24-h ambulatory BP, ultrasound-assessed

endothelium-dependent and -independent vasodilation of

the brachial artery, and markers of oxidative damage,

(plasma and urinary F2-isoprostanes, oxidized

low-density lipoproteins and plasma tocopherols), were

measured.

Results A significant interaction between grape-seed and

vitamin C treatments for effects on BP was observed.

Vitamin C alone reduced systolic BP versus placebo

(S1.8 W 0.8 mmHg, P U 0.03), while polyphenols did not

(S1.3 W 0.8 mmHg, P U 0.12). However, treatment with the

combination of vitamin C and polyphenols increased

0263-6352 � 2005 Lippincott Williams & Wilkins

systolic BP (4.8 W 0.9 mmHg versus placebo;

6.6 W 0.8 mmHg versus vitamin C; 6.1 W 0.9 mmHg versus

polyphenols mmHg, each P < 0.0001) and diastolic BP

(2.7 W 0.6 mmHg, P < 0.0001 versus placebo;

1.5 W 0.6 mmHg, P U 0.016 versus vitamin C;

3.2 W 0.7 mmHg, P < 0.0001 versus polyphenols).

Endothelium-dependent and -independent vasodilation,

and markers of oxidative damage were not significantly

altered.

Conclusion Although the mechanism remains to be

elucidated, these results suggest caution for hypertensive

subjects taking supplements containing combinations of

vitamin C and polyphenols. J Hypertens 23:427–434 Q 2005

Lippincott Williams & Wilkins.

Journal of Hypertension 2005, 23:427–434

Keywords: blood pressure, endothelial function, oxidative stress,polyphenols, vitamin C

School of Medicine and Pharmacology, University of Western Australia and WestAustralian Institute for Medical Research (WAIMR), Perth, Australia.

Sponsorship: This study was funded by the National Health and MedicalResearch Council (NHMRC) of Australia (Project Grant 139067). All tablets wereprovided by Taractechnologies (Australia).

Correspondence and requests for reprints to Natalie Ward, School of Medicine &Pharmacology (RPH), Box X2213 GPO Perth, WA 6847, Australia.Tel: +61 8 92240381; fax: +61 8 9224 0246;e-mail: nward@cyllene.uwa.edu.au

Received 8 June 2004 Revised 2 September 2004Accepted 15 September 2004

IntroductionHypertension has been postulated to be a state of

increased oxidative stress [1]. This may be related to

increased production of free radicals in the arterial wall, a

reduction in nitric oxide bioavailability [1] and endothe-

lial dysfunction [2]. If oxidative stress plays a role in

hypertension and endothelial dysfunction, then dietary

antioxidants could be expected to beneficially influence

both blood pressure (BP) and endothelial function. Avail-

able evidence suggests water-soluble antioxidants may

play an important role and vitamin C and polyphenolic

compounds represent two major dietary water-soluble

antioxidants.

Cross-sectional population studies indicate a consistent

association of higher vitamin C intake or status with lower

BP [3] Controlled intervention trials have been incon-

clusive, with some suggesting that vitamin C supplemen-

tation may lower BP [4,5], while others have shown no

effect [6,7]. Studies examining the effect of polyphenolic

compounds on BP have provided conflicting results [8,9].

However, a recent report has shown a reduction in clinic

428 Journal of Hypertension 2005, Vol 23 No 2

BP following treatment with dark chocolate versus

polyphenol-free white chocolate, which may be due to

the cocoa polyphenols [10].

The effect of vitamin C on endothelial dysfunction also

remains unclear. A number of studies have demonstrated

vitamin C’s ability to improve vasodilation [11,12] in

hypertensive subjects, but not controls [13]. However,

Duffy et al. [14] observed no improvement in vasodilation

in hypertensive subjects, despite a reduction in BP [14].

Improvement in vasodilation by polyphenolic com-

pounds has been observed in subjects with coronary

artery disease [15], hyperlipidemia [16] and healthy sub-

jects [17]. However, studies have not been carried out in

hypertensive subjects.

The ability of vitamin C and polyphenols to reduce

markers of oxidative stress is also inconclusive. Vitamin

C has been shown to reduce urinary isoprostanes in non-

smokers [18], yet have no effect on urinary isoprostanes in

hypertensive subjects [4]. Polyphenols from tea [19], red

wine [20] and purple grape juice [21], have been shown to

reduce ex vivo LDL oxidation. However, effects of poly-

phenols on isoprostanes are variable [19,22,23].

Arachidonic acid can be metabolized by cytochrome

P450 enzymes to form 20-hydroxyeicosatetraenoic acid

(20-HETE), which has been postulated to play a role in

BP regulation and vascular tone [24,25]. Several in vitrostudies have shown inhibitory effects of various polyphe-

nols and phenolic acids on cytochrome P450 isoforms

[26,27]. The effect of vitamin C either alone or in combi-

nation with polyphenols on the activity of the cytochrome

P450 enzyme has not been previously investigated.

To date no studies have examined possible additive or

synergistic effects of vitamin C and polyphenols on BP,

endothelial dysfunction and markers of oxidative

damage. The aims of this study were to determine if

vitamin C and grape-seed polyphenols, alone and in

combination, lower ambulatory BP and improve endothe-

lial dysfunction in hypertensive subjects. We also sought

to determine if any changes in BP were related to changes

in oxidative stress, inflammation or 20-HETE.

Materials and MethodsSubjectsBetween February 2002 and May 2003, 74 men and

women (mean age 62 � 7 years) with a previous physician

diagnosis of hypertension were recruited from the Perth

general population to the School of Medicine and Phar-

macology at the University of Western Australia. All

individuals were taking one or more antihypertensive

drugs for � 3 months, had a mean 24-h ambulatory sys-

tolic BP of � 125 mmHg and at least one additional

cardiovascular disease (CVD) risk factor. Risk factors

included previous coronary or cerebrovascular event

> 6 months, hyperlipidaemia (total cholesterol > 6

mmol/l), use of lipid-lowering therapy, or smoking

> 5 cigarettes/day. All volunteers ceased any vitamin,

fish oil or antioxidant supplements for at least 3 weeks

prior to study entry. All usual medication was taken as

prescribed on the morning of each visit and maintained

for the duration of the trial. Exclusion criteria included:

previous coronary or cerebrovascular event � 6 months,

heart failure or unstable disease, premenopausal, use of

nitrate medication, use of oral contraceptive, body mass

index > 35 kg/m2, diabetes mellitus, fasting gluco-

se � 7 mmol/l, or elevated serum creatinine (men >110 mmol/l or women > 100 mmol/l). Individuals were

asked to limit tea and coffee intake to 3 cups/day, and

cease all red wine and commercial fruit juice for the

duration of the study. All individuals provided a written

informed consent. The study was approved by the Royal

Perth Hospital Human Ethics Committee and was car-

ried out in accordance with the Declaration of Helsinki

(1989) of the World Medical Association.

DesignFollowing a 3-week washout period, subjects were allo-

cated to a study treatment via block randomization using

computer-generated random numbers, devised by the

statistician. Volunteers were assigned to receive either

(1) 500 mg/day vitamin C and matched grape-seed poly-

phenol placebo, (2) 1000 mg/day grape-seed polyphenols

and matched vitamin C placebo, (3) 500 mg/day vitamin

C and 1000 mg/day grape-seed polyphenols, or (4)

matched placebo tablets for both grape-seed polyphenols

and vitamin C, for 6 weeks in a double-blind fashion

(Fig. 1). Tablets were taken twice daily at meal times as

250 mg and 500 mg of vitamin C and polyphenols,

respectively. Vitamin C doses chosen for the present

study as well as the treatment duration were in keeping

with previous studies [4,5]. Average daily intake of poly-

phenols was estimated to be approximately 1000 mg/day,

and the present dose was chosen to double polyphenol

intake. All tablets were identical in appearance, size and

colour and were prepared by Taractechnologies (Austra-

lia). All study personnel and participants were blinded to

treatment assignment for the duration of the study. The

chief investigator held the code for the tablets in a sealed

envelope, which was not broken until the end of the trial.

At the end of the washout and intervention periods, all

subjects completed a 24-h ambulatory BP monitoring,

provided a 24-h urine collection, underwent fasting bra-

chial artery ultrasound to assess endothelium-dependent

and -independent vasodilation and provided a fasting

blood sample.

Methods24-h ambulatory blood pressure

Twenty-four hour BP monitoring was performed using

an ambulatory BP monitoring device (Spacelabs 90207;

Antioxidant supplementation and blood pressure Ward et al. 429

Fig. 1

weeks-6

C nimatiV yad/gm 005

obecalp +

91 = n

yad/gm 0001

obecalp + slonehpyloP

81 = n

C nimatiV yad/gm 005

yad/gm 0001 +

slonehpylop

81 = n

obecalp + obecalP

91 = n

Randomizationweeks-3

letepmoc ton diD

0 = n

ton diD

2 = n

ton diD

2 = n

ton diD

1 = n

for Analysed forAnalysed

stniopdne

91 = n

stniopdne

61 = n

stniopdne

61 = n

stniopdne

81 = n

forAnalysed forAnalysed forAnalysed

letepmoc letepmoc letepmoc

Treated hypertensive patients

24-h systolic BP ≥ 125 mmHg +

≥ 1 CVD risk factors

Study design. BP, blood pressure; CVD, cardiovascular disease.

Australia), set to take oscillometric readings at 20-min

intervals while awake and 30-min intervals while asleep.

The monitor was fitted to the non-dominant arm 2.5 cm

above the antecubital fossa by a trained researcher. While

sitting and with the arm resting at heart level, BP was

calibrated with the monitor connected to a mercury

sphygmomanometer and at least three readings, recorded

by the monitor, were within � 7 mmHg of the readings

observed on the sphygmomanometer. Patients were

instructed to continue their normal routine and maintain

a diary while awake. A valid 24-h recording was accepted

as a minimum of 80% successful readings with all cali-

bration and error readings excluded from analysis. Read-

ings were aggregated for each hour, and mean BP

determined for the 24-h period and for awake and asleep

times based on the patient’s diary.

Brachial artery ultrasonography

Brachial artery ultrasonography was carried out as pre-

viously described [28]. Briefly, patients were studied fol-

lowing a 12-h fast and resting supine in a quiet,

temperature-controlled room (21–258C). Endothelium-

dependent vasodilation was assessed as response to

forearm ischaemic flow-mediated dilation (FMD) and

independent dilation as response to glyceryl trinitrate

(GTN)-mediated dilation. An experienced observer blind-

ed to the patient’s treatment status performed the analysis.

Responses were calculated as the percentage change in

brachial artery diameter from baseline at maximum peak

time. Reproducibility studies have previously demon-

strated an intrasubject coefficient of variation of 14.7

and 17.6% for FMD and GTN response, respectively [28].

Biochemical parameters

Oxidative stress was assessed via measurement of plasma

and 24-h urinary F2-isoprostanes by gas chromatography

mass spectrometry (GC/MS) [29]. Oxidized low-density

lipoproteins (oxLDL) were analysed using an enzyme-

linked immunosorbent assay (ELISA) kit (Mercodia,

Sweden) [30] and plasma a- and g-tocopherol concentra-

tions were analysed by high-performance liquid chroma-

tography (HPLC) with electrochemical detection using

Tocol as an internal standard [31]. Twenty-four hour

urinary 20-hydroxyeicosatetraenoic acid (20-HETE) was

analysed by GC/MS [32]. The Core Clinical Laboratory at

Royal Perth Hospital analysed high-sensitivity C-reactive

protein (HS-CRP) with particle-enhanced immunone-

phelometry using BN Systems (Dade-Behring, Germany)

and full blood picture was analysed using a Cell-Dyn 4000

analyser.

Compliance

Compliance was assessed via tablet count and analysis of

plasma vitamin C and urinary polyphenol metabolites.

430 Journal of Hypertension 2005, Vol 23 No 2

Table 1 Baseline patient characteristics

Vitamin C Polyphenols Vitamin C þ polyphenols Placebo

n 19 16 16 18Age (years) 59.5 � 5.9 61.3 � 6.3 62.3 � 7.1 63.6 � 8.2Gender (M/F) 12/7 12/4 10/6 14/4BMI (kg/m2) 28.7 � 3.6 27.7 � 3.4 28.6 � 2.6 29.3 � 4.3Smokers n (%) 1 (5) 2 (13) 1 (6) 0 (0)Mean 24-h SBP (mmHg) 133.6 � 9.8 133.6 � 11.6 138.8 � 11.0 134.1 � 8.2Mean 24-h DBP (mmHg) 80.8 � 7.9 79.8 � 9.8 79.6 � 11.0 78.2 � 7.1Plasma vitamin C (mmol/l) 40.8 � 14.6 41.8 � 17.2 35.2 � 17.4 40.1 � 12.3Urinary 3-HPP (mg/24-h)

�712 (346, 1467) 382 (219, 666) 518 (296, 906) 913 (517, 1613)

Mean result � SD or. �Geometric mean (95% confidence intervals). 3-HPP, 3-hydroxyphenylpropionic acid No significant differences. SBP, systolic blood pressure; DBP,diastolic blood pressure; BMI, body mass index; M, male; F, female.

Table 2 Differences in the use of each antihypertensive drug classand statin therapy between the treatment groups

Vitamin C PolyphenolsVitamin C þpolyphenols Placebo x2 P-value

ACE inhibitors 10 10 9 13 0.642At II receptor

blockers5 2 4 3 0.707

b-blockers 9 8 5 6 0.589Calcium

channel entryblockers

8 5 6 4 0.612

Diuretics 8 6 6 5 0.834Statins 14 12 12 16 0.653

ACE, angiotensin-converting enzyme; At II, angiotensin II.

Plasma vitamin C was measured by HPLC with electro-

chemical detection, using a modified version of a previous

method [33]. Twenty-four hour urinary excretion of 3-

hydroxyphenylpropionic acid (3-HPP), a phenolic acid

metabolite of proanthocyanadins [34], were analysed

using GC/MS [35].

Statistical analysis

Sample size was calculated using BP as the primary

endpoint; 16 subjects in each group provided us with

> 80% power to observe a 5 mmHg change in BP. Ambu-

latory BP was analysed using mixed models (PROC

Mixed) with the Statistical Analysis Systems (SAS Insti-

tute, Cary, North Carolina, USA). All other analysis was

carried out using the Statistical Package for the Social

Sciences (SPSS Version 11.5, Chicago, Illinois, USA).

Non-normally distributed data were log-transformed.

Results are presented as mean result � SEM or geometric

mean (95% confidence intervals) as appropriate. Univari-

ate and multivariate analysis was carried out with Bonfer-

roni post-hoc comparisons to assess the effect of treatment

on BP, endothelial function and markers of oxidative

damage, inflammation and 20-HETE production.

ResultsA total of 69 subjects completed the study. One subject

was excluded due to a myocardial infarction during the

study period; one ceased antihypertensive treatment; two

experienced gastrointestinal side-effects from the poly-

phenol tablets and one was unrelated to the study. The

primary analysis was according to the protocol and

involved all individuals who successfully completed

the study. Treatment groups were well matched at base-

line (Table 1), and there were no significant differences

between groups for age, body mass index, BP, urinary

phenolic acid excretion or plasma levels of vitamin C.

Number and type of antihypertensive medication did not

differ between the groups (Table 2).

Mean compliance assessed by tablet count, was reported

as 96%. Plasma vitamin C concentrations increased by

approximately 50% in those receiving vitamin C

(40.8 � 3.4 to 62.9 � 3.1 mmol/l, P < 0.001) and the com-

bination (35.2 � 4.4 to 62.9 � 2.9 mmol/l, P < 0.001).

Urinary excretion of the polyphenol metabolite increased

significantly following supplementation with the poly-

phenols (3-HPP, 382 (219, 666) to 1111 (628, 1966) mg/

24 h, P < 0.001). A similar increase was observed follow-

ing supplementation with the combination (3-HPP, 518

(296, 906) to 1033 (630, 1695) mg/24 h, P < 0.001).

There was a significant interaction between the vitamin

C and polyphenol treatments that affected BP. That is,

the effects of vitamin C and polyphenols were different

when provided in combination. Therefore results could

not be analysed for main effects and were analysed by

group, following adjustment for baseline levels. Treat-

ment with vitamin C alone significantly reduced systolic

BP versus placebo (�1.8 � 0.8 mmHg, P ¼ 0.02), while

treatment with polyphenols alone did not (�1.3 �0.8 mmHg, P ¼ 0.12). However, treatment with the com-

bination of vitamin C and polyphenols resulted in a

significant increase in mean 24-h systolic BP versus

treatment with placebo (4.8 � 0.9 mmHg, P < 0.0001),

vitamin C (6.6 � 0.8 mmHg, P < 0.0001) or polyphenols

(6.1 � 0.9 mmHg, P < 0.0001) (Fig. 2a). Treatment

with the combination also resulted in a significant

increase in diastolic BP versus placebo (2.7 � 0.6 mmHg,

P < 0.0001), vitamin C (1.5 � 0.6 mmHg, P ¼ 0.02) and

polyphenols (3.2 � 0.7 mmHg, P < 0.0001) (Fig. 2b).

Mean 24-h heart rate (HR) was significantly increased

following treatment with vitamin C compared with treat-

ment with placebo (1.6 � 0.6 beats/min, P ¼ 0.007) or

polyphenols (3.7 � 0.6 beats/min, P < 0.0001). Treat-

ment with the combination significantly increased HR

Antioxidant supplementation and blood pressure Ward et al. 431

Fig. 2

4−3−2−1−01234567

Mea

n s

ysto

lic B

P (

mm

Hg

)

a)(

*

‡†*

Poly + CV Poly + CV

Poly + CV

PolyPoly

Poly

CV 2−

1−

0

1

2

3

4

Mea

n d

iast

olic

BP

(m

mH

g)

b)(‡†*

CV

CV

3−

2−

1−

0

1

2

3B

eats

/min

c)( ‡* ‡

Mean 24-h ambulatory (a) systolic blood pressure (BP), (b) diastolic BP and (c) heart rate. Mixed model Statistical Analysis Systems (SAS) analysisfollowing adjustment for baseline, �versus placebo, yversus vitamin C and zversus polyphenols, P < 0.001. VC, vitamin C; Poly, polyphenols.

compared with polyphenols alone (3.2 � 0.6 beats/min,

P < 0.0001) (Fig. 2c). These changes were unaltered

after adjustment for age and gender. Mean systolic and

diastolic BP for awake and asleep periods were signifi-

cantly increased following treatment with the vitamin C

and polyphenol combination compared with placebo,

vitamin C alone and polyphenols alone (data not pre-

sented). Addition of dosage time to the model did not

alter the observed increases following treatment with the

combination and this variable was not statistically sig-

nificant (data not presented).

FMD and GTN responses were not significantly altered

by treatment with vitamin C alone, polyphenols alone or

the combination (Fig. 3). Plasma and urinary F2-isopros-

tanes, oxidized LDL or a- and g-tocopherol were not

significantly affected by any of the treatments (Table 3).

Markers of inflammation including HS-CRP, leukocytes

Fig. 3

2− Poly + CVPolyCV

1− 5.1−

0− 5.0.0 51.1 52

Mea

n c

han

ge

FM

D (%

)

(a)

Mean percentage change in (a) flow-mediated dilation (FMD) and (b) glyceryfollowing adjustment for age and gender with Bonferroni post-hoc compari

and monocytes were not altered by any of the study

treatments. However, neutrophils were significantly ele-

vated following treatment with the combination of vita-

min C and polyphenols (Table 3). Urinary 20-HETE

excretion (pmol/24 h) was not significantly different

between the groups and was not affected by any of the

study treatments (Table 3). Adjustment for age, gender,

smoking or alcohol intake did not alter the interpretation

of the findings.

DiscussionGiven the evidence linking hypertension with oxidative

stress, it is reasonable to speculate that antioxidants

would beneficially influence oxidative stress, BP and

endothelial function. However, the present study indi-

cates that a combination of vitamin C and grape-seed

polyphenols results in a clinically significant increase in

BP. In contrast, treatment with vitamin C alone resulted

(b)

CV4−3−2−1−0123456

Mea

n c

han

ge

GT

N (%

)

Poly + CVPoly

l trinitrate (GTN) response versus the placebo group. Univariate analysissons. VC, vitamin C; Poly, polyphenols.

432 Journal of Hypertension 2005, Vol 23 No 2

Table 3 Treatment effects on biomarkers of oxidative stress, plasma tocopherols, inflammation and urinary excretion of 20-HETE

Vitamin C Polyphenols Vitamin C þ polyphenols Placebo

Urinary 24-h F2-isoprostanes (pmol/24-h)§ �648 � 435 �1111 � 1355 �834 � 293 �348 � 260Plasma F2-isoprostanes (pmol/l)§ �232 � 93 �14 � 78 �80 � 87 �23 � 87Oxidized LDL (mg/ml)§ �0.01 � 0.43 �0.10 � 0.62 0.04 � 0.33 �0.12 � 0.57a-tocopherol �1.14 � 1.15 �0.58 � 1.04 0.38 � 1.00 0.50 � 0.50g-tocopherol 0.01 � 0.15 �0.11 � 0.12 0.01 � 0.12 0.13 � 0.06Leucocytes ( 109/l) �0.26 � 0.20 �0.01 � 0.18 0.33 � 0.25 �0.04 � 0.13Neutrophils ( 109/l) �0.26 � 0.13 �0.09 � 0.13 0.40 � 0.21�,y �0.08 � 0.09Monocytes ( 109/l) �0.05 � 0.04 0.02 � 0.02 0.04 � 0.02 0.01 � 0.03HS-CRP (mg/l)§ �4.88 � 4.89 �0.31 � 0.40 1.16 � 1.31 �0.24 � 0.4420-HETE (pmol/24-h)§ �18.7 � 138.3 7.5 � 75.1 �156.1 � 246.9 �68.1 � 160.4

Mean difference � SEM versus baseline. Between treatment group univariate analysis with Bonferroni post-hoc comparisons. �versus placebo. yversus vitamin C andzversus polyphenols, P < 0.05. §Data was log-transformed for analysis. HS-CRP, high-sensitivity C-reactive protein; 20-HETE, 20-hydroxyeicosatetraenoic acid; LDL, low-density lipoprotein.

in a small, but significant reduction in systolic BP. To our

knowledge this is the first study to demonstrate a pressor

effect of this combination supplement on BP. Although

this result is unexpected, it is in agreement with a recent

review that there is widespread misconception that the

use of antioxidants will always be beneficial. In fact,

depending on the circumstances, such compounds, which

are better described as redox agents, may be pro-oxidant

rather than antioxidant [36].

However, in our study, there was no effect of the combina-

tion supplement on any of the measured markers of

oxidative stress. This suggests that the increase in BP

may not have been a result of increased oxidative stress,

at least as measured by markers of lipid peroxidation

(urinary and plasma F2-isoprostanes), oxidized LDL or

plasma tocopherols. The effects of supplementation with

combination vitamin C and polyphenols on oxidative

stress have not been previously investigated. An in vitrostudy has shown a synergistic effect of phenolic acids and

ascorbate in preventing oxidation of LDL [37]. The

potential mechanism was thought to be the ability of

vitamin C to recycle the highly reactive phenoxyl radicals.

Alternatively, flavonoids have been shown to inhibit

uptake of ascorbate in cultured cells [38]. However, in

the present study, we have seen no apparent effects of

polyphenols on the bioavailability of ascorbate, with simi-

lar increases in plasma vitamin C with and without poly-

phenol treatment. Furthermore, there appears to be no

effect of vitamin C on the bioavailability of polyphenols.

An alternative mechanism investigated was a possible

effect of treatment on the production of 20-HETE.

20-HETE is a vasoactive cytochrome P450 metabolite

of arachidonic acid that has been implicated in the

pathogenesis of hypertension [24,25,39]. It is thought

to cause constriction via inhibition of potassium channels

[40], and contribute to the vascular actions of endothelin-1

and angiotensin [41,42]. Data on the regulation of

20-HETE in humans are limited. Studies have suggested

that polyphenolic compounds may be able to inhibit

various isoforms of cytochrome P450, which may alter

drug metabolism [26,27]. The present study however,

revealed that polyphenol treatment, either alone or in

combination with vitamin C, had no effect on 20-HETE

production. The possibility that the combination of vita-

min C and grape-seed polyphenols may be interfering

with the metabolism of antihypertensive drugs has not

been ruled out.

There is evidence to suggest that chronic or low-grade

inflammation is a risk factor for hypertension [43]. Alter-

natively, increased BP may itself be a stimulus for

inflammation, resulting from increased hydrostatic pres-

sure and strain, which in turn affects endothelial cell

function [43,44]. Inflammation may also contribute to

vascular problems associated with hypertension, via the

release of vasoconstrictors, reduction of nitric oxide and

promotion of platelet aggregation [43,45]. The increase in

circulating neutrophils following supplementation with

the combination is consistent with this explanation.

However, HS-CRP, total leucocyte count and monocytes

were not significantly altered, so although some of our

results provide support for an inflammatory mechanism,

the overall evidence is weak.

The non-significant reduction in FMD response follow-

ing treatment with the vitamin C, polyphenol combina-

tion suggests the increase in BP may be associated with a

small decrease in endothelium-dependent vasodilation.

The lack of significance for the change in FMD response

may be related to power. The study was originally pow-

ered for a main effects analysis. However, because of the

significant interaction for blood pressure, all analysis was

by group. This gave us 80% power to see a 3.8% change

in FMD response. We saw a < 1% change in FMD

response.

We also observed a small, but significant reduction in

systolic BP following treatment with vitamin C alone,

while the reduction seen with the polyphenols alone was

not significant. This is in contrast to the result observed

with the combination. Some previous studies have shown

a reduction in BP with vitamin C [4,5], while in others,

Antioxidant supplementation and blood pressure Ward et al. 433

vitamin C has had no effect [6,7]. Duffy et al. [4] provided

500 mg/day vitamin C, for 4 weeks, and observed

13 mmHg reduction in clinic systolic BP, but not diastolic

BP. Fotherby et al. [5] also examined 500 mg/day of

vitamin C for 3 months and observed a 3 mmHg reduc-

tion in ambulatory day systolic BP, but not diastolic or

clinic BP. There have been no other studies examining

the effects of grape-seed polyphenols on BP, but studies

examining other polyphenols [8,9] remain inconclusive.

In our study, although treatment with vitamin C alone

was able to reduce systolic BP, there was no correspond-

ing reduction in oxidative stress biomarkers or improve-

ment in endothelial function. This is in agreement with a

previous study that showed reductions in BP following

treatment with vitamin C, but no change in urinary

isoprostanes [4].

There is increasing evidence suggesting oxidative stress

is involved in the pathogenesis of hypertension. This has

resulted in the concept that dietary supplementation with

antioxidants may be able to reduce blood pressure by

reducing oxidative stress. However, the present study

provides evidence that combined supplementation with

vitamin C and grape-seed polyphenols actually increases

blood pressure in treated hypertensive individuals. While

the mechanism behind this increase in BP following

supplementation with the vitamin C and grape-seed

polyphenol combination remains to be elucidated, the

results indicate that the effect is unlikely to be due to

increased oxidative stress or 20-HETE production, but

still allow the possibility that it may be related to inflam-

mation or endothelial dysfunction. Although this result

was unexpected, it suggests that careful evaluation in

controlled clinical trials is required before the use of this

combination by treated hypertensive individuals can be

supported.

AcknowledgementsN.C.W. gratefully acknowledges the assistance of a

University of Western Australia Postgraduate Award.

The authors thank Michael Clarke, Jennifer Rivera

and Lisa Rich for expert technical assistance and the

volunteers who took part in the study.

References1 Kumar KV, Das UN. Are free radicals involved in the pathobiology of human

essential hypertension? Free Radic Res Commun 1993; 19:59–66.

2 Panza JA, Quyyumi AA, Brush JE Jr, Epstein SE. Abnormal endothelium-dependent vascular relaxation in patients with essential hypertension.N Engl J Med 1990; 323:22–27.

3 Ness AR, Chee D, Elliott P. Vitamin C and blood pressure – an overview.J Hum Hypertens 1997; 11:343–350.

4 Duffy SJ, Gokce N, Holbrook M, Huang A, Frei B, Keaney JF Jr, et al.Treatment of hypertension with ascorbic acid. Lancet 1999; 354:2048–2049.

5 Fotherby MD, Williams JC, Forster LA, Craner P, Ferns GA. Effect ofvitamin C on ambulatory blood pressure and plasma lipids in olderpersons. J Hypertens 2000; 18:411–415.

6 Ghosh SK, Ekpo EB, Shah IU, Girling AJ, Jenkins C, Sinclair AJ. A double-blind, placebo-controlled parallel trial of vitamin C treatment in elderlypatients with hypertension. Gerontology 1994; 40:268–272.

7 Kim MK, Sasaki S, Sasazuki S, Okubo S, Hayashi M, Tsugane S. Lack oflong-term effect of vitamin C supplementation on blood pressure.Hypertension 2002; 40:797–803.

8 Hodgson JM, Puddey IB, Beilin LJ, Mori TA, Burke V, Croft KD, et al.Effects of isoflavonoids on blood pressure in subjects with high-normalambulatory blood pressure levels: a randomized controlled trial. Am JHypertens 1999; 12:47–53.

9 Hosseini S, Lee J, Sepulveda RT, Rohdewald P, Watson RR. Arandomised, double-blind, placebo-controlled, prospective, 16 weekcross-over study to determine the role of Pycnogenol in modifying bloodpressure in mildly hypertensive patients. Nutr Res 2001; 21:1251–1260.

10 Taubert D, Berkels R, Roesen R, Klaus W. Chocolate and blood pressurein elderly individuals with isolated systolic hypertension. JAMA 2003;290:1029–1030.

11 Taddei S, Virdis A, Ghiadoni L, Magagna A, Salvetti A. Vitamin C improvesendothelium-dependent vasodilation by restoring nitric oxide activity inessential hypertension. Circulation 1998; 97:2222–2229.

12 Sherman DL, Keaney JF Jr, Biegelsen ES, Duffy SJ, Coffman JD, Vita JA.Pharmacological concentrations of ascorbic acid are required for thebeneficial effect on endothelial vasomotor function in hypertension.Hypertension 2000; 35:936–941.

13 Singh N, Graves J, Taylor PD, MacAllister RJ, Singer DR. Effects of a‘healthy’ diet and of acute and long-term vitamin C on vascular function inhealthy older subjects. Cardiovasc Res 2002; 56:118–125.

14 Duffy SJ, Gokce N, Holbrook M, Hunter LM, Biegelsen ES, Huang A, et al.Effect of ascorbic acid treatment on conduit vessel endothelial dysfunctionin patients with hypertension. Am J Physiol 2001; 280:H528–H534.

15 Stein JH, Keevil JG, Wiebe DA, Aeschlimann S, Folts JD. Purple grapejuice improves endothelial function and reduces the susceptibility of LDLcholesterol to oxidation in patients with coronary artery disease.Circulation 1999; 100:1050–1055.

16 Hodgson JM, Puddey IB, Burke V, Watts GF, Beilin LJ. Regular ingestionof black tea improves brachial artery vasodilator function. Clin Sci 2002;102:195–201.

17 Cuevas AM, Guasch V, Castillo O, Irribarra E, Mizon C, San Martin A, et al.A high-fat diet induces and red wine counteracts endothelial dysfunction inhuman volunteers. Lipids 2000; 35:143–148.

18 Huang HY, Appel LJ, Croft KD, Miller ER III, Mori TA, Puddey IB. Effects ofvitamin C and vitamin E on in vivo lipid peroxidation: results of arandomized controlled trial. Am J Clin Nutr 2002; 76:549–555.

19 Hodgson JM, Puddey IB, Croft KD, Burek V, Mori TA, Caccetta RA, et al.Acute effects of ingestion of black and green tea on lipoprotein oxidation.Am J Clin Nutr 2000; 71:1103–1107.

20 Abu-Amsha R, Croft KD, Puddey IB, Proudfoot JM, Beilin LJ. Phenoliccontent of various beverages determines the extent of inhibition of humanserum and low-density lipoprotein oxidation in vitro: identification andmechanism of action of some cinnamic acid derivatives from red wine. ClinSci 1996; 91:449–458.

21 O’Byrne DJ, Devaraj S, Grundy SM, Jialal I. Comparison of the antioxidanteffects of Concord grape juice flavonoids a-tocopherol on markers ofoxidative stress in healthy adults. Am J Clin Nutr 2002; 76:1367–1374.

22 Abu-Amsha CR, Burke V, Mori TA, Beilin LJ, Puddey IB, Croft KD. Redwine polyphenols, in the absence of alcohol, reduce lipid peroxidativestress in smoking subjects. Free Radic Biol Med 2001; 30:636–642.

23 Hodgson JM, Croft KD, Mori TA, Burke V, Beilin LJ, Puddey IB. Regularingestion of tea does not inhibit in vivo lipid peroxidation in humans. J Nutr2002; 132:55–58.

24 McGiff JC, Quilley J. 20-hydroxyeicosatetraenoic acid andepoxyeicosatrienoic acids and blood pressure. Curr Opin NephrolHypertens 2001; 10:231–237.

25 Roman RJ. P-450 metabolites of arachidonic acid in the control ofcardiovascular function. Physiol Rev 2002; 82:131–185.

26 Ow YY, Stupans I. Gallic acid and gallic acid derivatives: effects on drugmetabolizing enzymes. Curr Drug Metab 2003; 4:241–248.

27 Piver B, Berthou F, Dreano Y, Lucas D. Differential inhibition of humancytochrome P450 enzymes by epsilon-viniferin, the dimer of resveratrol:comparison with resveratrol and polyphenols from alcoholized beverages.Life Sci 2003; 73:1199–1213.

28 Woodman RJ, Playford DA, Watts GF, Cheetham C, Reed C, Taylor RR,et al. Improved analysis of brachial artery ultrasound using a novel edge-detection software system. J Appl Physiol 2001; 91:929–937.

29 Mori TA, Croft KD, Puddey IB, Beilin LJ. An improved method for themeasurement of urinary and plasma F2- isoprostanes using gaschromatography-mass spectrometry. Anal Biochem 1999; 268:117–125.

30 Holvoet P, Vanhaecke J, Janssens S, Van de WF, Collen D. Oxidized LDLand malondialdehyde-modified LDL in patients with acute coronarysyndromes and stable coronary artery disease. Circulation 1998;98:1487–1494.

434 Journal of Hypertension 2005, Vol 23 No 2

31 Su Q, Rowley KG, O’Dea K. Stability of individual carotenoids, retinol andtocopherols in human plasma during exposure to light and after extraction.J Chromatogr 1999; 729:191–198.

32 Rivera J, Ward N, Hodgson J, Puddey IB, Falck JR, Croft KD. Measurementof 20-hydroxyeicosatetraenoic acid in human urine by gaschromatography-mass spectrometry. Clin Chem 2004; 50:224–226.

33 McCulloch RK, Vandongen R. Measurement of ascorbic acid in plateletsand its relationship to polymorphonuclear leukocyte levels. Clin Chim Acta1992; 213:15–22.

34 Deprez S, Brezillon C, Rabot S, Philippe C, Mila I, Lapierre C, et al.Polymeric proanthocyanidins are catabolized by human colonic microflorainto low-molecular-weight phenolic acids. J Nutr 2000; 130:2733–2738.

35 Hodgson JM, Morton LW, Puddey IB, Beilin LJ, Croft KD. Gallic acidmetabolites are markers of black tea intake in humans. J Agric Food Chem2000; 48:2276–2280.

36 Azzi A, Davies KJ, Kelly F. Free radical biology - terminology and criticalthinking. FEBS Lett 2004; 558:3–6.

37 Vieira O, Laranjinha J, Madeira V, Almeida L. Cholesteryl esterhydroperoxide formation in myoglobin-catalyzed low density lipoproteinoxidation: concerted antioxidant activity of caffeic and p-coumaric acidswith ascorbate. Biochem Pharm 1998; 55:333–340.

38 Park JB, Levine M. Intracellular accumulation of ascorbic acid is inhibitedby flavonoids via blocking of dehydroascorbic acid and ascorbic aciduptakes in HL-60, U937 and Jurkat cells. J Nutr 2000; 130:1297–1302.

39 Imig JD, Zou AP, Stec DE, Harder DR, Falck JR, Roman RJ. Formation andactions of 20-hydroxyeicosatetraenoic acid in rat renal arterioles. Am JPhysiol 1996; 270:R217–R227.

40 Roman RJ. Does 20-HETE mediate the cGMP independent actions ofNO? High Blood Press Res Newsl 1999; Fall:20–22.

41 Croft KD, McGiff JC, Sanchez-Mendoza A, Carroll MA. Angiotensin IIreleases 20-HETE from rat renal microvessels. Am J Physiol 2000;279:F544–F551.

42 Hercule HC, Oyekan AO. Cytochrome P450 omega/omega-1hydroxylase-derived eicosanoids contribute to endothelin(A) andendothelin(B) receptor-mediated vasoconstriction to endothelin-1 in therat preglomerular arteriole. J Pharm Exp Ther 2000; 292:1153–1160.

43 Nakanishi N, Sato M, Shirai K, Suzuki K, Tatara K. White blood cell countas a risk factor for hypertension; a study of Japanese male office workers.J Hypertens 2002; 20:851–857.

44 Chae CU, Lee RT, Rifai N, Ridker PM. Blood pressure and inflammation inapparently healthy men. Hypertens 2001; 38:399–403.

45 Brown DW, Giles WH, Croft JB. White blood cell count: an independentpredictor of coronary heart disease mortality among a national cohort.J Clin Epidemiol 2001; 54:316–322.