Effects of a quercetin-rich onion skin extract on 24 h ambulatory bloodpressure and endothelial function in overweight-to-obese patients with (pre-)hypertension: a randomised double-blinded placebo-controlled cross-over trial

Verena Brll1, Constanze Burak1, Birgit Stoffel-Wagner2, Siegfried Wolffram3, Georg Nickenig4,Cornelius Mller4, Peter Langguth5, Birgit Alteheld1, Rolf Fimmers6, Stefanie Naaf7, Benno F. Zimmermann7,8,Peter Stehle1 and Sarah Egert1*1Department of Nutrition and Food Sciences, Nutritional Physiology, University of Bonn, 53115 Bonn, Germany2Institute of Clinical Chemistry and Clinical Pharmacology, University Hospital Bonn, 53127 Bonn, Germany3Institute of Animal Nutrition and Physiology, Christian-Albrechts-University Kiel, 24118 Kiel, Germany4Department of Cardiology, Angiology and Pneumology, University Hospital Bonn, 53127 Bonn, Germany5Department of Biopharmaceutics and Pharmaceutical Technology, Institute of Pharmacy and Biochemistry, JohannesGutenberg University Mainz, 55099 Mainz, Germany6Institute of Medical Biometry, Informatics and Epidemiology, University Hospital Bonn, 53127 Bonn, Germany7Institut Prof. Dr. Georg Kurz GmbH, 50933 Kln, Germany8Department of Nutrition and Food Sciences, Food Technology and Biotechnology, University of Bonn, 53117 Bonn, Germany

(Submitted 11 March 2015 Final revision received 25 June 2015 Accepted 7 July 2015)

AbstractThe polyphenol quercetin may prevent CVD due to its antihypertensive and vasorelaxant properties. We investigated the effects of quercetinafter regular intake on blood pressure (BP) in overweight-to-obese patients with pre-hypertension and stage I hypertension. In addition, thepotential mechanisms responsible for the hypothesised effect of quercetin on BP were explored. Subjects (n 70) were randomised to receive162mg/d quercetin from onion skin extract powder or placebo in a double-blinded, placebo-controlled cross-over trial with 6-week treatmentperiods separated by a 6-week washout period. Before and after the intervention, ambulatory blood pressure (ABP) and office BP weremeasured; urine and blood samples were collected; and endothelial function was measured by EndoPAT technology. In the total group,quercetin did not significantly affect 24 h ABP parameters and office BP. In the subgroup of hypertensives, quercetin decreased 24 h systolicBP by 36mmHg (P= 0022) when compared with placebo (mean treatment difference, 39 mmHg; P= 0049). In addition, quercetinsignificantly decreased day-time and night-time systolic BP in hypertensives, but without a significant effect in inter-group comparison. In thetotal group and also in the subgroup of hypertensives, vasoactive biomarkers including endothelin-1, soluble endothelial-derived adhesionmolecules, asymmetric dimethylarginine, angiotensin-converting enzyme activity, endothelial function, parameters of oxidation, inflammation,lipid and glucose metabolism were not affected by quercetin. In conclusion, supplementation with 162mg/d quercetin from onion skin extractlowers ABP in patients with hypertension, suggesting a cardioprotective effect of quercetin. The mechanisms responsible for the BP-loweringeffect remain unclear.

Key words: Quercetin: Blood pressure: Endothelial function: Hypertension: Cardiovascular diseases

Quercetin (3,3,4,5,7-pentahydroxyflavone) is one of the pre-dominant flavonoids, ubiquitously distributed in (edible) plants,and one of the most potent antioxidants of plant origin(1). Richsources of dietary quercetin are onions, kale, unpeeled apples,berries, citrus fruits and tea (Camellia sinensis). In Westernpopulations, crude estimates of mean dietary intake appear to

be 1030 mg/d(2,3). As demonstrated in cohort studies, dietaryintake of flavonoids in general and of quercetin in particular isassociated with a decreased risk for CVD(2,4). Although thephysiological mechanisms accounting for this benefit remainincompletely defined, animal studies and human interventionstudies have identified many relevant effects for example,

* Corresponding author: S. Egert, fax +49 228 733217, email s.egert@uni-bonn.de

Abbreviations: ABP, ambulatory blood pressure; ACE, angiotensin-converting enzyme; ADMA, asymmetric dimethylarginine; BP, blood pressure; DBP, diastolicblood pressure; IsoP, isoprostanes; MAP, mean arterial pressure; oxLDL, oxidised LDL; PAT, peripheral arterial tonometry; RHI, reactive hyperaemia index;SBP, systolic blood pressure; sICAM-1, soluble intercellular adhesion molecule 1; sVCAM-1, soluble vascular cell adhesion molecule 1.

British Journal of Nutrition, page 1 of 15 doi:10.1017/S0007114515002950 The Authors 2015. This is an Open Access article, distributed under the terms of the CreativeCommons Attribution licence (http://creativecommons.org/licenses/by/3.0/), which permits unrestrictedre-use, distribution, and reproduction in any medium, provided the original work is properly cited.

http://crossmark.crossref.org/dialog/?doi=10.1017/S0007114515002950&domain=pdf

supplementation of quercetin may reduce platelet aggrega-tion(5,6) and plasma concentrations of atherogenic oxidised LDL(oxLDL)(7). In vitro studies suggest that high concentrations ofquercetin (>1 M) have anti-inflammatory effects(8,9). Studies inanimal models (e.g. obese Zucker rats) suggest beneficialeffects of quercetin on obesity-associated metabolic disordersincluding insulin resistance and dyslipidaemia(1012). Werecently showed that in patients with high CVD risk phenotype,chronic supplementation with a supra-nutritional dose of150mg/d quercetin significantly reduced systolic blood pres-sure (SBP)(7). Similar findings have been reported by Edwardset al.(13) in hypertensive patients and, recently, by Zahediet al.(14) in women with type 2 diabetes mellitus, although withpharmacological quercetin doses (500 and 730 mg/d). Althoughseveral pathways have been suggested, the mechanisms bywhich quercetin possibly affects blood pressure (BP) are notwell-understood. These pathways include (i) improvement ofvascular function in an endothelium-dependent orendothelium-independent manner, (ii) decrease in oxidativestress and/or (iii) interference with the reninangiotensinaldosterone system(15).One limitation in the interpretation of the BP-lowering effect

of chronic quercetin supplementation is that all human studiespublished to date, including our own trial(7), only measured theoffice (clinic) BP in the resting state (typically in the morningwhile fasting) and did not integrate an ambulatory bloodpressure (ABP) monitoring. ABP monitoring is considered thegold standard for BP measurement, compared with clinic BPmeasurements, as it is superior in terms of reliability andvalidity(16,17). The superior predictive power of ABP monitoringis probably not only due to its higher number of readings,which increases the reliability of the measurement, but alsodue to its ability to capture the impact of stressors and otherenvironmental factors that occur in daily life and are likely toaffect BP(16). To the best of our knowledge, no previous studyhas examined the effects of quercetin on 24 h ABP profiles thusfar. Therefore, the aim of the present double-blinded, placebo-controlled cross-over trial was to systematically investigate theeffects of quercetin on arterial BP (office BP and 24 h ABPprofiles) in adults with pre-hypertension and stage I hyperten-sion, and to explore mechanisms involved in the BP-loweringefficacy of quercetin. Furthermore, effects of quercetin on lipidand glucose metabolism were investigated.

Methods

Subjects

Overweight-to-obese subjects were recruited from the com-munity of the city of Bonn, Germany, via public postings, flyersand advertisements in the local newspaper. From a total of 500interested subjects, 154 individuals aged 2565 years with a BMIof 2535 kg/m2 attended a screening, which included physicalassessments (body height and weight, resting BP, heart rate,waist and hip circumference), clinical assessments (liver func-tion, serum lipids and lipoproteins, glucose and uric acid,haematology, high-sensitive C-reactive protein (hs-CRP)),medical history and a dietary questionnaire.

Participants were included if they had the following traits ofthe metabolic syndrome: (i) central obesity (waist circumference94 cm for men and 80 cm for women); (ii) pre-hypertension(120139mmHg SBP and/or 8089mmHg diastolic bloodpressure (DBP)) or stage I hypertension (140159mmHgSBP and/or 9099mmHg DBP); (iii) dyslipidaemia (fastingserum TAG concentrations 17mmol/l and/or serumconcentrations of HDL-cholesterol

the onion skin extract powder (Allium cepa L.; Rudolf WildGmbH & Company KG) was 4125 % (dry mass 9594 %). Eachquercetin capsule contained 54 mg quercetin. Hard gelatincapsules (Coni-Snap) size 0 were supplied by Capsugel.Quercetin and placebo capsules were identical in shape andtaste. Capsule filling was carried out using a Dott Bonapacesemi-automatic capsule filling machine. Mannitol was obtainedfrom Fagron. Quality was checked by determining the homo-geneity of weight distributions of a sample of twenty rando-mised capsules taken from each batch. Furthermore, themicrobiological burden of the capsules was determined fol-lowing manufacture and before release. Primary packaging wasinto blister packages. The primary investigators, all study per-sonnel, and all the participants were blinded to the treatments.A quercetin dosage of 162mg per d (three capsules) wasselected to represent the 10- to 15-fold of the estimated mean

daily quercetin intake in Germany(20) and other Europeanpopulations(21). Plasma kinetics of this quercetin dosage afterbolus intake was examined previously(22,23).

Subjects were assigned to quercetin or placebo treatmentaccording to a block-wise randomisation scheme. Separatecomputer-generated randomisation schedules for men andwomen were created to stratify subjects by sex with an attemptto achieve a distribution between men and women of 50:50 ineach treatment group. Randomisation, allocation to either oneof the capsules and capsule handling were carried out by anindependent researcher (B. A.). Capsules were handed out ondays 0 and 21 of each treatment period with a surplus of 20 %.Leftovers and empty blisters were collected on days 21 and 42.Compliance was monitored by determining quercetin plasmaconcentrations after the treatment periods (see below), bycounting capsules at the end of the study and by instructing

Assessed for eligibility bytelephone (n 500)

Complete screening withphysical and clinical

assessment, medical historyand dietary questionnaire

(n 154)

Excluded (n 346) Not meeting inclusion criteria (n 206) Declined to participate (n 31) Other reasons (n 109)

Randomly assigned tocrossover trial (n 70)

Excluded (n 84) Not meeting inclusion criteria (n 74) Withdrew (n 6) Other reasons (n 4)

Allocated to receive quercetin(n 35)

Allocated to receive placebo(n 35)

Washout (n 34) Washout (n 34)

Drop out (n 1) Drop out (n 1)

Allocated to receive placebo(n 34)

Allocated to receive quercetin(n 34)

Analysed for outcomes in the total study group (n 68)

Office BP and heart rate

24 h ABP monitoring

Anthropometric data

Blood/urine analyses

Dietary intake of energy, nutrients and quercetin

Analysed for outcomes in subgroups

Office BP, heart rate and 24 h ABP monitoring in hypertensive patients selected according to BP classification (17, 24):

Vascular testing (n 48)

Enrolment

Crossover

Allocation

Analyses

- office BP and heart rate (n 59 and n 55 for quercetin and placebo) - 24 h ABP monitoring (n 31 and n 29 for quercetin and placebo)

Fig. 1. Flow diagram of participants. ABP, ambulatory blood pressure; BP, blood pressure.

Effects of quercetin on blood pressure 3

subjects to document capsule consumption, potentialobserved side-effects, deviations from their normal physicalactivity and any other observations considered relevant in astudy diary.Subjects were instructed to keep 3-d food records at the

beginning and at the end of each treatment period. Each recordrepresented the food and beverage intake of 2 weekdays and1 weekend day. These dietary records were used to calculatethe habitual dietary intake of energy, nutrients and quercetin.Sample size was calculated based on office BP data from our

previous trial(7,24) and expected changes in SBP. The calculationrevealed that forty-nine subjects had to complete the study toreach a 90 % power with a significance level of 001 to detect a3 mmHg difference in BP between the quercetin and placebogroups (SD of the within-subject was different from the SBP,38 mmHg). With the assumption of a 30 % dropout rate, weaimed to randomly assign approximately seventy participants.The measurements of 24 h ABP, office BP, heart rate, vascular

testing and anthropometric measurements were conducted atthe beginning and at the end of the two intervention periods.

Measurements

Office blood pressure and heart rate. Measurements ofoffice BP and heart rate were obtained using an automatic BPmeasurement device (boso carat professional, Bosch+Sohn) understandardised conditions according to the recommendations of theAmerican Heart Association Council on High Blood PressureResearch(25). Each participant sat quietly for 510min, after whichtheir arm was placed at heart level and SBP and DBP weremeasured at least twice in 3- to 5-min intervals. If BP measurementsvaried by 10mmHg, an additional measurement was performed.The accumulated measurements were then averaged to determineoverall SBP and DBP. The mean arterial pressure (MAP) wascalculated as follows: (DBP+1/3 (SBPDBP)).

24 h ambulatory blood pressure monitoring. The 24 h ABPand heart rate recordings were taken every 15min between06.00 and 22.00 hours (day-time) and every 30 min between22.00 and 06.00 hours (night-time) using an ABP monitor(Spacelabs monitor type 90207) on the non-dominant arm. Onthe day of measurement, subjects were instructed to maintain

their habitual activity level and to refrain from strenuousexercise. We assessed the following BP parameters: 24 h,day-time and night-time SBP, DBP, MAP, heart rate and thenocturnal dip in SBP and DBP.

Vascular testing. For vascular testing, we used the noninvasiveperipheral arterial tonometry (PAT) technology to assess thereactive hyperaemia index (RHI) and the augmentation index (AI)using the EndoPAT plethysmographic device (Endo-PAT2000).Details have been described elsewhere(26). In brief, the PAT signalindicates changes of the peripheral arterial tone in peripheralarterial beds by recording the arterial pulsatile volume changesfrom the fingertip. For that purpose, we placed a pair ofplethysmographic biosensors on both index fingers and a BP cuffon the upper arm of the study arm (left arm), whereas the rightarm served as the control arm (without a cuff). The recording ofthe PAT signal started after a resting period of at least 15min witha 1min standby-test period. If the signal recording was free frominterferences, the test period started with a 5min baselinerecording of the pulse wave amplitude. Subsequently, the BP cuffwas inflated for 5min to supra-systolic values to induce ischaemiawhereupon the cuff was deflated. This resulted in reactivehyperaemia while further recording for 5min. PAT signals wereanalysed with automated software (Itamar Medical), and RHI wascalculated as the ratio of the average PAT signal amplitude over a1min period starting 1min after cuff-deflation divided by theaverage PAT signal amplitude over a 35min period at thebaseline recording followed by normalisation of the values fromthe study arm to the control arm. The RHI correlates with brachialartery flow-mediated dilatation(27) and is significantly influencedby nitric oxide (NO)(28). A lower RHI was found in subjects withcoronary endothelial dysfunction(29) and in the presence of CVDrisk factors(30).

Anthropometrics. Body height was determined on a stadiometerto the nearest 01 cm. Body weight was determined to the nearest100 g. Waist circumference was measured midway between thelowest rib and the ilial crest, while the subject was at minimalrespiration. Hip circumference was measured at the height oftrochanteres majores. Body composition was measured by bioe-lectrical impedance analysis (Nutrigard-M, Multi Frequency Phase-Sensitive Bioelectrical Impedance Analyzer, Data Input). Fat-freemass (FFM) was calculated in accordance with Sun et al.(31);fat mass was calculated by subtracting FFM from body weight.

Blood/urine sample processing and analysis

Fasting venous blood samples were collected on the first andthe last visit of each intervention period between 06.30 and09.00 hours under standardised conditions. The subjects wereinstructed to abstain for 24 h from alcoholic beverages and weretold not to engage in strenuous exercise on the day beforeblood sampling. The last capsule was taken in the eveningbefore blood sampling. Blood was drawn into tubes containingEDTA, lithium heparin, fluoride or a coagulation activator(Sarstedt). Plasma/serum was obtained by centrifugation at3000 g for 15 min at 8C. Plasma/serum aliquots were

Table 1. Flavonoid analysis in onion skin extract powder*(Mean values and standard deviations)

% (g/100 g)

Compound Mean SD

Quercetin 442 015Quercetin dihexoside 004523 000027Quercetin hexoside 1 01557 00072Quercetin hexoside 2 179 0017Methylquercetin hexoside 00481 000030Kaempferol 0122 00017Methylquercetin 00740 000033

* Analyses were conducted in duplicate.

4 V. Brll et al.

immediately frozen in cryovials and stored at 80C until ana-lysis. All the laboratory measurements were performed withoutknowledge of the treatment. All serum and plasma samples ofone subject were analysed in the same assay run.Serum concentration of total cholesterol was measured using

polychromatic endpoint-measurement, whereas serum con-centrations of LDL-cholesterol, HDL-cholesterol, TAG and plasmaconcentrations of glucose were measured using bichromaticendpoint-measurement with a Dimension Vista 1500 analyser(Siemens Healthcare Diagnostics). Serum concentrations of apoB, A1 and hs-CRP were determined using nephelometric methodswith a Dimension Vista 1500 analyser.Serum insulin concentration was measured using a chemilumi-

nescentimmunometric assay with the Immulite 2000 analyser(Siemens Healthcare Diagnostics). Insulin resistance was measuredusing the homoeostatic model assessment (HOMA) and calculatedas the product of the fasting plasma insulin concentration (in U/ml)and the fasting plasma glucose concentration (in mmol/l), dividedby 225(32). HbA1c was measured using an HPLC-method witha Variant II analyser (Bio-Rad Laboratories). Serum concentrationsof angiotensin-converting enzyme (ACE) were measured using aphotometric method (Buehlmann Laboratories) with a DimensionVista 1500 analyser.Plasma ADMA (Immundiagnostik), plasma oxLDL (Immundiag-

nostik), serum endothelin-1 and serum-soluble adhesion moleculesE-selectin, soluble intercellular adhesion molecule 1 (sICAM-1)and soluble vascular cell adhesion molecule 1 (sVCAM-1) (R&Dsystems) were determined in duplicate using commerciallyavailable enzyme-linked immunoassay kits according to themanufacturers instructions and quality controls. Plasmaconcentration of L-arginine was determined using reversed-phaseHPLC as described previously(33).Analyses of plasma concentrations of quercetin, its

monomethylated derivatives tamarixetin (4-O-methyl quercetin)and isorhamnetin (3-O-methyl quercetin) as well as of thedehydroxylated quercetin metabolite kaempferol were carriedout using HPLC with fluorescence detection as describedpreviously(34). All the samples were treated enzymatically with-glucuronidase/sulphatase before the extraction of the flavonols.Total plasma flavonols were calculated as follows: total flavonols(nmol/l)=quercetin (nmol/l) +kaempferol (nmol/l) + isorhamnetin(nmol/l) + tamarixetin (nmol/l).First morning urine samples were collected on the first and

the last visit of the intervention periods, 0002 % of butylatedhydroxytoluene was added and the urine was frozen at 80 Cuntil analysis. From these urine samples, 8-iso-PG F2 (8-iso-PGF2) and 2,3-dinor-15-F2t-IsoP were measured by UHPLC-MS/MS. Urinary creatinine levels were determined using abichromatic kinetic method (Jaff method).

Self-reported dietary intake of energy, nutrients andquercetin

The self-reported intakes of energy, macronutrients, dietaryfibre and antioxidant pro-vitamins/vitamins were calculatedusing the computer-based nutrient calculation programmeEBISpro (University of Hohenheim) based on the German

Nutrient Database Bundeslebensmittelschlssel (Max Rubner-Institute). The quercetin intake was estimated using the USDAflavonoid database(35).

Statistical analyses

All the statistical analyses were performed using IBM SPSSstatistical software package (version 20). Differences betweensexes at screening were tested using the unpaired Studentst test or the MannWhitney U test. Baseline values werecompared between groups using paired Students t tests orWilcoxon signed-rank tests. Intra-group (baseline v. endpoint)and inter-group comparisons (changes during quercetin v.changes during placebo treatment) of normally distributed datawere performed using paired Students t tests. Intra-group andinter-group comparisons of data that were not normally dis-tributed, which was mainly the case for serum TAG, plasmaglucose, serum insulin, HOMA IR index, plasma oxLDL, serumhs-CRP and urinary IsoP, were conducted by Wilcoxon signed-rank tests (for details, see footnotes in Tables).

Interaction effects between the stage of hypertension andtreatment assignment were tested using a univariate ANOVAwith the respective variables as fixed factors. In all cases, avalue for P 005 (two-sided) was taken to indicate significanteffects. Pearsons correlation coefficient was used to assessrelationships between BP variables (SBP and DBP) and bio-markers of inflammation/endothelial function and also betweenBP variables and plasma flavonols. A test for carry-over effectsaccording to Kenward and Jones(36) was used. No carry-overeffects between the two treatment periods could be observed.All the analyses are presented on a per-protocol basis. For allsixty-eight participants, complete data sets were available.

All data were analysed for the whole study group (n 68) andalso for the subgroup of patients with hypertension. Office BPdata were classified according to Chobanian et al.(18). Classifi-cation of ABP data was conducted as described by Pickeringet al.(25) (for details see footnotes in Table 4).

Results

Subject characteristics at screening, compliance and dietaryintake

Characteristics of the participants at screening are presented inTable 2. As planned, all subjects were overweight (43 %) orobese (57 %), had a visceral fat distribution and were pre-hypertensive or stage 1 hypertensive. We observed sex differ-ences with respect to body height, body weight, waist and hipcircumference, waist:hip ratio and fasting serum concentrationsof TAG, total cholesterol, HDL-cholesterol and plasma glucose(Table 2).

Count of returned capsules indicated an almost full com-pliance of 982 (SD 26) % and 980 (SD 41) % during quercetinand placebo consumption, respectively. Compliance to quer-cetin supplementation was objectively confirmed by a markedincrease in plasma concentrations of quercetin and total flavo-nols by 11496 % (P< 00001) and 8286 % (P< 00001),respectively (Fig. 2(a) and (b)). The increase was measurable in

Effects of quercetin on blood pressure 5

all patients receiving quercetin. In addition, plasma concentra-tions of kaempferol, isorhamnetin and tamarixetin significantlyincreased after quercetin but not after placebo supplementation(data not shown). There was a high inter-individual variation inplasma quercetin concentrations already at baseline (range forall study subjects: 023305 nmol/l) and after quercetin sup-plementation (99113131 nmol/l).Analyses of 3-d dietary records indicated no significant inter-

group and intra-group differences in mean daily intakes ofenergy, protein, carbohydrates, total fat, fatty acids, cholesterol,antioxidants (e.g. vitamin E, vitamin C), dietary fibre andquercetin (data not presented in Tables). Dietary quercetinintake was 113 (SD 119) mg/d and 90 (SD 81) mg/d at thebeginning and at the end of the quercetin treatment, respec-tively, and 117 (SD 112) mg/d and 92 (SD 82) mg/d at thebeginning and at the end of the placebo treatment, respectively.Main dietary quercetin sources were onions, apples and tea.

Body weight, waist and hip circumference, bodycomposition and potential side-effects

Quercetin supplementation did not significantly affect bodyweight, waist and hip circumference, relative fat mass or fat-freemass (data not shown). Participants did not report any side-effects, neither during quercetin nor during placebo treatment.

Ambulatory blood pressure monitoring and office bloodpressure

In the entire study group (n 68), quercetin and placebo sup-plements did not significantly affect mean 24 h, day-time andnight-time ABP parameters (SBP, DBP, MAP, nocturnal dip inSBP and DBP). In addition, office BP (SBP and DBP, MAP) was

not significantly changed by quercetin or placebo treatment asanalysed for the entire study group (Table 3). Resting SBP atbaseline and the change in SBP from baseline to after inter-vention were significantly correlated, with those individualswith higher baseline SBP demonstrating greater reductions inSBP in response to the quercetin treatment (r 0508,P< 00001; n 68). We also found a significant interaction effectbetween stage of hypertension and treatment assignment forthe decreases in mean 24 h SBP (P= 0008) and in day-time SBP(P= 0024) but not for the decrease in night-time SBP(P= 0568) (data not shown). There were also significantinteraction effects between stage of hypertension and treatmentassignment for decreases in mean 24 h MAP (P= 0014) and inday-time MAP (P= 0032) but not for decreases in night-timeMAP (P= 0176) (data not shown).

Effects of supplementation with quercetin or placebo on ABPprofiles and office BP in the subgroup of hypertensive patientsare presented in Table 4. Note that there are different n valuesfor quercetin and placebo treatment because the initial BP ofthe participants at the beginning of the quercetin and placeboperiod slightly differed. In some cases, this physiological dif-ference led to a different BP classification. When comparedwith placebo (treatment difference, 39 (SD 111) mmHg;P= 0049, Table 4), quercetin significantly decreased mean 24 hSBP by 36 (SD 82) mmHg (P= 0022) from baseline in thesubgroup of stage 1 hypertensive subjects (n 31). Changes inday-time SBP and night-time SBP during quercetin treatmentwere not significantly different from placebo treatment.However, quercetin significantly decreased day-time SBPby 46 (SD 90) mmHg (P= 0014) and night-time SBP by 66(SD 99) mmHg (P= 0007) (Table 4). In addition, changes in24 h MAP, day-time MAP and night-time MAP during quercetindid not differ significantly from placebo. However, quercetinsignificantly reduced mean 24 h MAP, day-time MAP and night-

Table 2. Subject characteristics and blood parameters at screening(Mean values and standard deviations)

Total (n 68) Women (n 34) Men (n 34)

Mean SD Mean SD Mean SD Women v. men (P)

Age (years) 474 105 482 104 466 106 0396Body height (cm) 1738 96 1682 90 1793 66

Table 3. Measurements of the 24 h ambulatory blood pressure and office blood pressure in the total study group during the 6-week dietary supplementation with quercetin or placebo*(Mean values and standard deviations)

Quercetin (n 68) Placebo (n 68)

Baseline Endpoint Mean changeP intra-group

Baseline Endpoint Mean changeP intra-group P inter-group

Treatment difference

Mean SD Mean SD Mean SD comparison Mean SD Mean SD Mean SD comparison comparison Mean SD

24 h ABP monitoringSystolic BP (mmHg) 1333 102 1324 90 09 73 0329 1325 84 1324 96 01 60 0848 0460 08 91Diastolic BP (mmHg) 820 85 814 76 07 53 0304 814 71 809 85 04 45 0425 0711 02 68MAP (mmHg) 995 87 986 75 09 61 0227 988 70 984 83 04 47 0520 0977 05 77Heart rate (beats per min) 743 82 763 82 +20 53 0003 745 98 751 87 +07 59 0372 0189 +13 81Nocturnal dip in systolic BP (%) 110 66 113 57 +03 61 0726 107 56 120 48 +13 65 0105 0368 11 95Nocturnal dip in diastolic BP (%) 159 71 169 68 +10 64 0217 160 67 175 62 +15 73 0101 0669 06 106Day-time (06.0022.00 hours)

Systolic BP (mmHg) 1366 106 1357 94 09 76 0347 1355 85 1361 97 +06 62 0440 0172 16 92Diastolic BP (mmHg) 850 87 845 79 05 56 0457 843 72 842 86 00 47 0932 0538 05 69MAP (mmHg) 1025 88 1018 79 08 64 0315 1014 71 1017 84 +03 49 0640 0257 11 78Heart rate (beats per min) 762 87 782 86 +19 56 0006 767 105 775 92 +07 64 0367 0202 +13 84Night-time (22.0006.00 hours)

Systolic BP (mmHg) 1214 117 1203 104 11 90 0322 1209 105 1196 98 13 91 0238 0921 +02 140Diastolic BP (mmHg) 714 92 701 80 13 64 0106 708 87 694 85 14 64 0085 0946 +08 99MAP (mmHg) 885 99 870 80 15 74 0093 881 91 867 85 14 75 0139 0911 02 115Heart rate (beats per min) 669 87 692 88 +24 66 0004 674 98 681 91 +07 73 0459 0203 +17 107Office BP

Systolic BP (mmHg) 1444 168 1469 172 +25 107 0063 1439 144 1438 174 01 109 0937 0171 +26 153Diastolic BP (mmHg) 982 113 972 97 10 85 0338 973 100 961 110 12 73 0166 0863 +02 115MAP (mmHg) 1136 125 1138 116 +02 85 0879 1128 107 1120 126 09 77 0361 0472 +10 116Heart rate (beats per min) 646 79 657 83 +11 55 0112 639 91 656 80 +16 64 0040 0598 05 82

ABP, ambulatory blood pressure; BP, blood pressure; MAP, mean arterial pressure.* The two groups did not differ significantly with regard to any of the variables at baseline (paired Students t tests or Wilcoxon signed-rank tests).

Effects

ofquercetin

onbloodpressu

re7

time MAP from baseline by 28 (SD 73) mmHg (P= 0043), by33 (SD 80) mmHg (P= 0042) and by 67 (SD 83) mmHg(P= 0001), respectively (Table 4).Treatment with quercetin did not significantly change the

nocturnal dip in SBP and DBP. In addition, systolic and diastolicoffice BP were not significantly changed by quercetin or pla-cebo treatment in hypertensive patients (Table 4).

Serum lipids, lipoproteins, apolipoproteins, glucose andinsulin

No significant inter-group differences were found for the effectsof quercetin or placebo on serum lipids, lipoproteins, apoli-poproteins, glucose and insulin. Furthermore, quercetin sup-plementation did not significantly affect fasting serum totalcholesterol, LDL-cholesterol, HDL-cholesterol, TAG, apo B andA1, insulin, plasma glucose and HbA1c neither in the total study

group nor in the subgroup of hypertensive patients (intra-groupcomparisons, Table 5).

Soluble adhesion molecules, asymmetric dimethylarginine,angiotensin-converting enzyme, C-reactive protein,endothelial function and oxidation

For the total study group, quercetin supplementation did notsignificantly affect fasting serum concentrations of endothelin-1,sE-selectin and sVCAM-1 (Table 6). When compared withplacebo, quercetin did not significantly affect serum sICAM-1(treatment difference, 53 (SD 307) ng/ml, P= 0219; Table 6).However, quercetin decreased fasting serum sICAM-1 frombaseline by 82 (SD 172) ng/ml (intra-group comparison,P< 0001). Neither quercetin nor placebo significantly alteredfasting plasma concentrations of the endogenous NO synthaseinhibitor ADMA, the corresponding ratio of L-arginine to ADMA

Week 0 Week 6

Week 0 Week 6

700

800

600

500

400

300

100

0

200

700

600

500

400

300

200

100

0

Time

Time

Pla

sma

flavo

nols

(nm

ol/l)

Pla

sma

quer

cetin

(nm

ol/l)

(b)

(a)

Fig. 2. Fasting plasma concentrations of quercetin (a) (n 68) and total flavonols (b) (n 68) before and after the 6-week supplementation with quercetin (162mg/d;) orplacebo (). Values are means and standard deviations represented by vertical bars. *** Mean value was significantly different from baseline (P< 00001; intra-groupcomparison). Change during quercetin treatment was significantly different from change during placebo treatment (P< 00001; inter-group comparison). Total plasmaflavonols were calculated as follows: total flavonols (nmol/l)= quercetin (nmol/l) + kaempferol (nmol/l) + isorhamnetin (nmol/l) + tamarixetin (nmol/l). The two groups didnot differ significantly with regard to plasma concentrations of quercetin and total flavonol at baseline (Wilcoxon signed-rank tests).

8 V. Brll et al.

Table 4. Measurements of the 24 h ambulatory blood pressure and office blood pressure in the subgroup of hypertensive patients during the 6-week dietary supplementation with quercetin or placebo*(Mean values and standard deviations)

Quercetin Placebo

Baseline Endpoint Mean changeP intra-group

Baseline Endpoint Mean changeP intra-group P inter-group

Treatment difference

Mean SD Mean SD Mean SD comparison Mean SD Mean SD Mean SD comparison comparison Mean SD

24 h ABP monitoring (n 31) (n 29)Systolic BP (mmHg) 1415 73 1379 73 36 82 0022 1397 53 1402 64 +05 64 0690 0049 39 111Diastolic BP (mmHg) 887 66 866 59 21 62 0074 866 61 869 73 +03 39 0680 0124 28 83MAP (mmHg) 1064 65 1037 59 28 73 0043 1044 54 1047 65 +03 44 0710 0151 30 97Heart rate (beats per min) 761 83 779 86 +18 55 0081 766 108 775 97 +09 59 0445 0435 +13 79Nocturnal dip in systolic BP (%) 121 65 12 66 +03 54 0758 96 58 125 59 +29 71 0035 0142 28 87Nocturnal dip in diastolic BP (%) 161 77 169 76 +08 58 0477 139 76 173 78 +33 87 0051 0213 25 95

Day-time (06.0022.00 hours) (n 27) (n 23)Systolic BP (mmHg) 1464 67 1418 77 46 90 0014 1445 44 1446 63 +01 59 0964 0109 49 121Diastolic BP (mmHg) 929 58 904 63 25 69 0070 903 65 909 74 +05 29 0399 0111 36 90MAP (mmHg) 1108 59 10754 65 33 80 0042 1084 52 1088 65 +04 36 0644 0110 42 106Heart rate (beats per min) 792 90 808 91 +16 63 0211 790 99 805 93 +14 67 0311 0793 06 91

Night-time (22.0006.00 hours) (n 21) (n 27)Systolic BP (mmHg) 1347 84 1281 96 66 99 0007 1311 67 1261 98 50 106 0021 0667 22 188Diastolic BP (mmHg) 806 94 758 81 49 72 0006 788 65 757 82 31 76 0045 0289 35 118MAP (mmHg) 997 85 930 79 67 83 0001 966 69 929 86 38 89 0037 0345 38 144Heart rate (beats per min) 687 85 716 93 +29 63 0049 695 110 675 85 20 87 0235 0127 +43 98

Office BP (n 59) (n 55)Systolic BP (mmHg) 1471 162 1493 170 +22 109 0131 1478 127 1472 169 06 116 0685 0311 +23 163Diastolic BP (mmHg) 1003 106 986 96 17 87 0138 1007 80 988 102 18 75 0072 0972 01 124MAP (mmHg) 1159 117 1155 114 04 87 0718 1164 86 1149 118 14 80 0186 0676 +07 125Heart rate (beats per min) 650 76 659 80 +08 55 0256 644 89 658 79 +13 61 0109 0575 06 82

ABP, Ambulatory blood pressure; BP, blood pressure; MAP, mean arterial pressure.* The two groups did not differ significantly with regard to any of the variables at baseline (paired Students t tests or Wilcoxon signed-rank tests). All the subjects participated in both treatments. There are different n values for quercetin and

placebo treatment because the initial blood pressure of the participants at the beginning of the quercetin and placebo treatment period slightly differed. In some cases, this physiological difference led to a different blood pressureclassification.

Systolic BP>135 and/or diastolic BP>85mmHg (according to reference 25). Systolic BP>140 and/or diastolic BP>90mmHg (according to reference 25). Systolic BP>125 and/or diastolic BP>75mmHg (according to reference 25). Stage 1 hypertensive: systolic BP140 and/or diastolic BP90mmHg (according to reference 18).

Effects

ofquercetin

onbloodpressu

re9

Table 5. Fasting lipids, lipoproteins, apolipoproteins, insulin, glucose, HbA1c and HOMA IR Index in the total study group (n 68) during the 6-week dietary supplementation with quercetin or placebo*(Mean values and standard deviations)

Quercetin (n 68) Placebo (n 68)

Baseline EndpointMeanchange

P value intra-groupBaseline Endpoint

Meanchange

P value intra-group P value inter-group

Treatmentdifference

Mean SD Mean SD Mean SD comparison Mean SD Mean SD Mean SD comparison comparison Mean SD

Serum TAG (mmol/l) 181 109 183 137 +003 098 0848 176 125 172 137 004 068 0241 0628 +007 118Serum total cholesterol (mmol/l) 544 112 538 112 006 058 0395 545 113 538 109 008 055 0257 0874 +002 089Serum HDL-cholesterol (mmol/l) 139 038 136 037 003 019 0238 138 034 138 035 000 014 0929 0406 003 025Serum LDL-cholesterol (mmol/l) 345 090 341 091 004 043 0459 352 095 338 088 014 041 0008 0193 +010 063Serum apo B (g/l) 105 026 104 025 002 012 0257 105 026 104 025 002 012 0233 0864 +000 018Serum apo A1 (g/l) 163 028 160 026 003 017 0128 162 026 160 027 002 013 0240 0824 001 023Plasma glucose (mmol/l) 508 068 511 058 +002 047 0526 518 063 513 071 003 051 0671 0400 +007 071Serum insulin (pmol/l) 4782 3362 4660 3586 122 2448 0649 5115 3875 4639 3885 475 2614 0079 0197 +353 3916HOMA IR index 157 134 151 132 005 092 0911 172 142 152 141 017 099 0071 0226 +013 141HbA1c (%) 566 043 568 045 +002 019 0366 570 042 566 038 003 018 0146 0105 +005 027

HOMA IR, homoeostasis model assessment of insulin resistance.* The two groups did not differ significantly with regard to any of the variables at baseline (paired Students t tests or Wilcoxon signed-rank tests).

10V.Brllet

al.

Table 6. Adhesion molecules, parameters of oxidation and inflammation and vascular testing in the total study group (n 68) during the 6-week supplementation with quercetin or placebo*(Mean values and standard deviations)

Quercetin (n 68) Placebo (n 68)

Baseline Endpoint Mean changeP intra-group

Baseline Endpoint Mean changeP intra-group P inter-group

Treatment difference

Mean SD Mean SD Mean SD comparison Mean SD Mean SD Mean SD comparison comparison Mean SD

Serum endothelin-1 (pg/ml) 222 074 225 074 +003 046 0639 209 071 225 078 +016 052 0014 0129 013 071Serum sE-selektin (ng/ml) 362 136 371 137 +09 51 0144 367 138 369 140 +02 48 0712 0441 +07 73Serum sVCAM-1 (ng/ml) 5810 1619 5716 1434 93 997 0444 5725 1523 5954 1558 +229 999 0063 0055 322 1358Serum sICAM-1 (ng/ml) 2057 542 1975 481 82 172

(L-arginine:ADMA), serum ACE and serum CRP. Quercetin andplacebo did not significantly affect endothelial function (RHIand AI) (Table 6). In addition, plasma oxLDL and urinaryexcretion of 8-iso-PGF2 and 2,3-dinor-15-F2t-IsoP were notsignificantly changed by quercetin or placebo treatment in thetotal study group (Table 6).In the subgroup of hypertensive patients, neither quercetin

nor placebo significantly affected serum endothelin-1, sE-selectin, sVCAM-1, plasma ADMA, serum ACE and CRP (datanot in Tables). In addition, plasma oxLDL and urinary F2-IsoPwere not significantly changed by quercetin or placebo treat-ment in the subgroup of hypertensives (data not in Tables). Inthe subgroup of hypertensive patients, quercetin and placebosignificantly decreased serum sICAM-1 concentrations frombaseline to a similar extent (quercetin, 93 (SD 148) ng/ml,P= 0002; placebo, 86 (SD 159) ng/ml, P= 0007; treatmentdifference, 35 (SD 175) ng/ml, P= 0584) (data not in Tables).

Correlations between blood pressure and biomarkers ofinflammation/endothelial function

In the total study group and also in the subgroup of hypertensivepatients, baseline SBP and DBP did not significantly correlatewith baseline biomarkers of inflammation and endothelial func-tion (CRP, endothelin-1, sE-selectin, sVCAM-1, sICAM-1, ADMA,RHI and AI). In addition, changes in BP variables did not sig-nificantly correlate with changes in biomarkers of inflammationand endothelial function, neither in the total study group nor inthe subgroup of hypertensives (data not shown).

Discussion

The aim of this double-blinded, placebo-controlled, cross-overintervention study was to systematically investigate the effectsof quercetin on arterial BP (primary variable), endothelialfunction and further CVD risk markers in overweight-to-obeseindividuals with hypertension. Our major finding was thatquercetin supplementation for 6 weeks in a supra-nutritionaldosage significantly reduced 24 h systolic ABP in hypertensiveparticipants. In contrast to our hypotheses, we observed noeffect of quercetin supplementation on secondary variablessuch as endothelial function including RHI, sE-selectin,sVCAM-1, ADMA and endothelin-1. In addition, no effects onparameters of inflammation, oxidative stress and lipid andglucose metabolism were observed. We used a moderate supra-nutritional but non-pharmacological dose of quercetin fromonion skin extract, as these data should provide a rational basisfor the development of functional foods.

Blood pressure

Quercetin significantly reduced the mean 24 h SBP by 36(SD 82)mmHg in hypertensive participants, but not inpre-hypertensive individuals, suggesting that a threshold ofelevated BP might be required to detect a BP-lowering effectof quercetin. The significant interaction effect between stage ofhypertension and treatment assignment confirms this. In contrast to

our previous trial(7), we found no significant effects of chronicquercetin supplementation on office SBP. At first glance, thisfinding is surprising. Our study design (e.g. quercetin dose, inter-vention period) was similar to our earlier study. In addition, ourparticipants (pre-/hypertensive patients with abdominal fat dis-tribution) and sample size were well chosen in consideration withthe primary objective of the trial. In accordance with our previousstudy(7), we observed a relatively high inter-individual variation inBP-lowering effects of quercetin. In both studies, only 44 to 51% ofthe participants showed a decrease in systolic office BP, despitethe fact that compliance with supplementation in both trials wasvery high (9498%). Compliance was confirmed by a markedincrease in plasma quercetin concentrations (final concentrationsof both trials, 03045 mol/l) and also in the monomethylatedderivatives isorhamnetin and tamarixetin. However, in both trials,changes in BP were not significantly correlated with plasma totalquercetin, isorhamnetin and/or tamarixetin. These findings indi-cate that, in humans, the BP-lowering effects of chronic quercetinsupplementation are difficult to predict. In addition, it is not pos-sible to deduce a minimum plasma concentration or quercetindose required for a specific biological activity. In this regard, it isvery likely that the plasma concentration of total quercetin doesnot necessarily reflect the concentration at the target sites. Similarhigh inter-individual variance in BP-lowering effects of quercetincan be observed in the study of Edwards et al.(13) in human sub-jects treated with pharmacological doses of 730mg/d quercetin for4 weeks. The authors illustrated individual subject responses dur-ing each supplementation phase (placebo, quercetin) in theirpublication; however, they did not discuss these results.

Overall, the quercetin-induced reductions in SBP observed in thepresent trial and previously (range 26 to 88mmHg)(7,13,14,24)

are similar to those experienced following current recommendedlifestyle modifications to reduce elevated BP (e.g. reducing sodiumintake and body weight, increasing physical activity)(37,38). A 36mmHg decrease in SBP as observed in the present trial would beclinically meaningful when considered at the population level,particularly in view of the large population of people with pre-hypertension and stage I hypertension(39). In addition, similareffects have been found for other dietary flavonoids, especially forflavonols from cocoa products(40,41).

Endothelial function, inflammation and oxidation

The first mechanism of action we explored was that quercetin-induced reductions in BP are secondary to an improvement invascular endothelial function. Rationale for this hypotheses wasbased on data from (i) normotensive humans wherein acuteadministration of 200 mg quercetin increased plasma quercetinand NO metabolite concentrations while decreasing endothelin-1(42), (ii) hypertensive rats in which quercetin attenuatedhypertension and/or vascular dysfunction in a NO-dependentmanner(4345) and (iii) in vitro studies wherein quercetindecreased cellular production of endothelin-1(46,47) andendothelial-derived adhesion molecules(48,49). In the presentstudy, biomarkers of endothelial function (plasma endothelin-1,sE-selectin, sVCAM-1, ADMA, RHI and AI) were unaffected bythe quercetin treatment. Therefore, these markers cannotexplain a possible relationship between quercetin and BP.

12 V. Brll et al.

However, in view of the only modest decrease in SBP, ourstudy was probably not suited to demonstrate small changes inthese markers. In addition, baseline PAT indices andendothelial-derived adhesion molecules did not correlate withBP, emphasising that factors regulating endothelial functiondiffer at least in part from those regulating BP.In contrast to previous in vitro(8,9) and in vivo results

obtained in animal models(4951) showing an anti-inflammatoryeffect of quercetin, quercetin supplementation did not reduceserum concentrations of CRP. This may, in part, be explained bythe duration of the intervention and/or the quercetin dosageand/or the low-grade inflammatory state of our participants,which were insufficient to elicit measurable effects. In agree-ment with this finding and consistent with earlier studies inhumans(13,23,52), no changes in biomarkers of in vivo lipidperoxidation (plasma oxLDL and urinary IsoP) were observedfollowing quercetin administration. This may be attributed to analready sufficient antioxidant defence status of our hypertensivesubjects. All participants had an adequate dietary intake ofessential antioxidants (vitamins E, C, -carotene). In addition,they did not report smoking or excessive physical exercise.

Angiotensin-converting enzyme

The second mechanism we hypothesised was that chronicquercetin supplementation decreased arterial BP in hypertensivepatients by decreasing ACE activity. Rationale for this hypotheseswas based on experimental studies demonstrating that quercetininhibits ACE in vitro(53) and on animal-based evidence showinga decrease in ACE activity after quercetin treatment in rats(54,55).For example, Mackraj et al.(54) compared the long-termantihypertensive effects of captopril (ACE inhibitor) with those ofquercetin in Dahl salt-sensitive rats that were given daily injectionsof captopril, quercetin or vehicle. Although BP increased in vehicle-treated Dahl rats, it was significantly decreased compared withbaseline in both quercetin- and captopril-treated groups. Thedecrease in BP occurred in parallel with the down-regulation of theangiotensin-I receptor in the kidney, increased urine volume andincreased urinary sodium excretion, thus providing a potentialmechanism for the long-term BP-lowering effects of quercetin(15).However, although chronic quercetin administration decreased24 h ABP in our hypertensive patients, we found no decreasein plasma ACE activity. Our results are in accordance withdata recently reported by Larson et al.(56) showing that acute,quercetin-induced reductions in BP in hypertensive individualswere not secondary to a reduced ACE activity.

Lipid and glucose metabolism

In the present study, chronic quercetin supplementation for6 weeks did not affect fasting serum concentrations of totalcholesterol, LDL-cholesterol and HDL-cholesterol, TAG andapolipoproteins, supporting previous data in metabolicallyhealthy participants(23,57), in overweight-to-obese patients withmetabolic syndrome traits(7), in men with different apoEisoforms(52) and in women with type 2 diabetes mellitus(14). Incontrast, Lee et al.(58) found that supplementation with quer-cetin (100mg/d) for 10 weeks significantly reduced fasting

serum concentrations of total cholesterol and LDL-cholesteroland significantly increased serum HDL-cholesterol in malesmokers. In addition, supplements of quercetin-rich red grapejuice(59) and grape powder(60) favourably influenced plasmalipid profiles in human subjects. These inconclusive results maybe attributed to the duration of quercetin administration, thebaseline characteristics of the study participants and the dosageof the ingested quercetin. Experimental studies in rat liver cellsdemonstrated that quercetin at high concentrations (e.g. 25mol/l) may be involved in lipid metabolism by reducinghepatic fatty acid, TAG and cholesterol synthesis(61,62).

In accordance with previous human studies(7,13,24,52,60), querce-tin did not affect fasting plasma glucose, serum insulin, HOMA IRor HbA1c in our hypertensive patients, although Lee et al.

(58)

demonstrated a slight decrease (39%) in fasting serum glucoseafter quercetin treatment. Our results are in contrast to experi-mental reports demonstrating the hypoglycaemic effects of highquercetin doses (10mg/kg of body weight) in animal models(e.g. obese Zucker Rats)(1012).

Strengths and limitations

The major strengths of our study are the double-blinded placebo-controlled cross-over design, the relatively large sample size, thehigh compliance to treatment, the low dropout rate and theexamination of numerous markers of vascular function. However,this study also has a few potential limitations. First, we administereda quercetin-rich onion skin extract instead of pure quercetin(quercetin dihydrate) as we did in our previous trials(7,23,24). Onionskins are rich in quercetin aglycone(63), and the bioavailability ofquercetin from capsules filled with onion skin extract powder wasshown to be significantly higher than that from capsules filled withquercetin dihydrate (C Burak, V Brll, P Langguth, BF Zimmer-mann, B Stoffel-Wagner, P Stehle, S Wolffram, S Egert, unpublishedresults). Thus, we decided to use a quercetin-rich onion skin extractfor the present trial. We characterised the polyphenol spectrum ofthe extract (Table 1), but we cannot exclude that other unknowncomponents in the onion skin extract may have influenced ourfindings. Thus, strictly, the conclusion of our study is only true forquercetin-rich onion skin extract but not for pure quercetin.Second, we conducted an explorative data analysis. All parameterswere analysed for the whole study group and also for the subgroupof patients with hypertension. Due to the multiple test situations,we cannot fully exclude that the significant effect on 24 h SBP inthe subgroup of hypertensives was a chance finding. On the otherhand, we do not think that corrections for multiple testing aremeaningful in the context of explorative data analysis(64).

Conclusion

Regular ingestion of a supra-nutritional dose of 162mg/d quercetinfrom onion skin extract did not affect BP and endothelial functionin the whole study population of pre-hypertensive and stage 1hypertensive overweight-to-obese subjects. In the subgroup ofhypertensives, quercetin was capable of decreasing 24 h SBP,suggesting a cardioprotective effect of quercetin, but without effectson mechanistic parameters. Thus, the mechanisms responsible forthe BP-lowering effect of quercetin remain elusive.

Effects of quercetin on blood pressure 13

Acknowledgements

This study was supported by the German Research Foundation(S. E., grant no EG292/3-1). The German Research Foundationhad no role in the design, analysis or writing of this article.V. B., P. S. and S. E. designed the study together with G. N.

and C. M. who were the medical advisors. V. B., C. B. and S. E.recruited the subjects and conducted the study. B. A. performedthe blinding procedure. P. L. prepared the capsules. V. B., C. B.,B. S.-W., S. W., S. N., B. F. Z. and S. E. participated in datacollection and analysis. V. B., C. B. and S. E. performed thestatistical calculations with support from R. F.; V. B., P. S. and S.E. wrote the manuscript. V.B. and S. E. had primary responsi-bility for the final content. All authors read and approved thefinal manuscript.There are no conflicts of interest.

References

1. Crozier A, Jaganath IB & Clifford MN (2009) Dietary phenolics:chemistry, bioavailability and effects on health. Nat Prod Rep26, 10011043.

2. Erdman JW Jr, Balentine D, Arab L, et al. (2007) Flavonoidsand Heart Health: Proceedings of the ILSI North AmericaFlavonoids Workshop, May 31-June 1, 2005, Washington, DC.J Nutr 137, 718S737S.

3. Scalbert A & Williamson G (2000) Dietary intake andbioavailability of polyphenols. J Nutr 130, 2073S2085S.

4. Wang X, Ouyang YY, Liu J, et al. (2013) Flavonoid intake andrisk of CVD: a systematic review and meta-analysis of pro-spective cohort studies. Br J Nutr 111, 111.

5. Hubbard GP, Wolffram S, Lovegrove JA, et al. (2004) Ingestionof quercetin inhibits platelet aggregation and essentialcomponents of the collagen-stimulated platelet activationpathway in humans. J Thromb Haemost 2, 21382145.

6. Hubbard GP, Wolffram S, de Vos R, et al. (2006) Ingestion ofonion soup high in quercetin inhibits platelet aggregationand essential components of the collagen-stimulated plateletactivation pathway in man: a pilot study. Br J Nutr 96, 482488.

7. Egert S, Bosy-Westphal A, Seiberl J, et al. (2009) Quercetinreduces systolic blood pressure and plasma oxidised low-density lipoprotein concentrations in overweight subjects witha high-cardiovascular disease risk phenotype: a double-blinded, placebo-controlled cross-over study. Br J Nutr 102,10651074.

8. Boots AW, Wilms LC, Swennen EL, et al. (2008) In vitro andex vivo anti-inflammatory activity of quercetin in healthyvolunteers. Nutrition 24, 703710.

9. Nair MP, Mahajan S, Reynolds JL, et al. (2006) The flavonoidquercetin inhibits proinflammatory cytokine (tumor necrosisfactor alpha) gene expression in normal peripheral bloodmononuclear cells via modulation of the NF-kappabeta system. Clin Vaccine Immunol 13, 319328.

10. Rivera L, Moron R, Sanchez M, et al. (2008) Quercetinameliorates metabolic syndrome and improves theinflammatory status in obese zucker rats. Obesity (SilverSpring) 16, 20812087.

11. Jeong SM, Kang MJ, Choi HN, et al. (2012) Quercetinameliorates hyperglycemia and dyslipidemia and improvesantioxidant status in type 2 diabetic db/db mice. Nutr ResPract 6, 201207.

12. Kobori M, Masumoto S, Akimoto Y, et al. (2011) Chronicdietary intake of quercetin alleviates hepatic fat accumulation

associated with consumption of a Western-style diet in C57/BL6J mice. Mol Nutr Food Res 55, 530540.

13. Edwards RL, Lyon T, Litwin SE, et al. (2007) Quercetinreduces blood pressure in hypertensive subjects. J Nutr 137,24052411.

14. Zahedi M, Ghiasvand R, Feizi A, et al. (2013) Does quercetinimprove cardiovascular risk factors and inflammatory bio-markers in women with type 2 diabetes: a double-blind ran-domized controlled clinical trial. Int J Prev Med 4, 777785.

15. Larson AJ, Symons JD & Jalili T (2012) Therapeutic potential ofquercetin to decrease blood pressure: review of efficacy andmechanisms. Adv Nutr 3, 3946.

16. Zawadzki MJ, Graham JW & Gerin W (2012) Increasing thevalidity and efficiency of blood pressure estimates usingambulatory and clinic measurements and modern missingdata methods. Am J Hypertens 25, 764769.

17. Myers MG & Godwin M (2007) Automated measurement ofblood pressure in routine clinical practice. J Clin Hypertens(Greenwich) 9, 267270.

18. Chobanian AV, Bakris GL, Black HR, et al. (2003) The SeventhReport of the Joint National Committee on Prevention,Detection, Evaluation, and Treatment of High Blood Pressure:the JNC 7 report. JAMA 289, 25602572.

19. Alberti KG, Zimmet P & Shaw J (2005) The metabolic syn-dromea new worldwide definition. Lancet 366, 10591062.

20. Linseisen J, Radtke J & Wolfram G (1997) Flavonoid intake ofadults in a Bavarian subgroup of the national foodconsumption survey. Z Ernahrungswiss 36, 403412.

21. Zamora-Ros R, Forouhi NG, Sharp SJ, et al. (2014) Dietaryintakes of individual flavanols and flavonols are inverselyassociated with incident type 2 diabetes in Europeanpopulations. J Nutr 144, 335343.

22. Egert S, Wolffram S, Schulze B, et al. (2012) Enrichedcereal bars are more effective in increasing plasma quercetincompared with quercetin from powder-filled hard capsules.Br J Nutr 107, 539546.

23. Egert S, Wolffram S, Bosy-Westphal A, et al. (2008) Dailyquercetin supplementation dose-dependently increasesplasma quercetin concentrations in healthy humans. J Nutr138, 16151621.

24. Egert S, Boesch-Saadatmandi C, Wolffram S, et al. (2010)Serum lipid and blood pressure responses to quercetin vary inoverweight patients by apolipoprotein E genotype. J Nutr140, 278284.

25. Pickering TG, Hall JE, Appel LJ, et al. (2005) Recommenda-tions for blood pressure measurement in humans andexperimental animals: Part 1: blood pressure measurement inhumans: a statement for professionals from the Subcommitteeof Professional and Public Education of the American HeartAssociation Council on High Blood Pressure Research.Hypertension 45, 142161.

26. Axtell AL, Gomari FA & Cooke JP (2010) Assessing endothelialvasodilator function with the Endo-PAT 2000. J Vis Exp, issue44, 2167.

27. Kuvin JT, Patel AR, Sliney KA, et al. (2003) Assessment ofperipheral vascular endothelial function with finger arterialpulse wave amplitude. Am Heart J 146, 168174.

28. Nohria A, Gerhard-Herman M, Creager MA, et al. (2006)Role of nitric oxide in the regulation of digital pulsevolume amplitude in humans. J Appl Physiol (1985) 101,545548.

29. Bonetti PO, Pumper GM, Higano ST, et al. (2004) Noninvasiveidentification of patients with early coronary atherosclerosisby assessment of digital reactive hyperemia. J Am Coll Cardiol44, 21372141.

14 V. Brll et al.

30. Hamburg NM, Keyes MJ, Larson MG, et al. (2008) Cross-sectionalrelations of digital vascular function to cardiovascular risk factorsin the Framingham Heart Study. Circulation 117, 24672474.

31. Sun SS, Chumlea WC, Heymsfield SB, et al. (2003) Developmentof bioelectrical impedance analysis prediction equations for bodycomposition with the use of a multicomponent model for use inepidemiologic surveys. Am J Clin Nutr 77, 331340.

32. Matthews DR, Hosker JP, Rudenski AS, et al. (1985) Home-ostasis model assessment: insulin resistance and beta-cellfunction from fasting plasma glucose and insulinconcentrations in man. Diabetologia 28, 412419.

33. Frst P, Pollack L, Graser TA, et al. (1990) Appraisal of fourpre-column derivatization methods for the high-performanceliquid chromatographic determination of free amino acids inbiological materials. J Chromatogr 499, 557569.

34. Bieger J, Cermak R, Blank R, et al. (2008) Tissue distributionof quercetin in pigs after long-term dietary supplementation.J Nutr 138, 14171420.

35. U.S.Department of Agriculture (2013) USDA database for theflavonoid content of selected foods, release 3.1 http://wwwars usda gov/Services/docs htm?docid=6231

36. Kenward MG & Jones B (1987) The analysis of data from22 cross-over trials with baseline measurements. Stat Med 6,911926.

37. Appel LJ, Brands MW, Daniels SR, et al. (2006) Dietaryapproaches to prevent and treat hypertension: a scientificstatement from the American Heart Association. Hypertension47, 296308.

38. Whelton PK, Appel LJ, Sacco RL, et al. (2012) Sodium, bloodpressure, and cardiovascular disease: further evidence sup-porting the American Heart Association sodium reductionrecommendations. Circulation 126, 28802889.

39. Lewington S, Clarke R, Qizilbash N, et al. (2002) Age-specificrelevance of usual blood pressure to vascular mortality: ameta-analysis of individual data for one million adults in 61prospective studies. Lancet 360, 19031913.

40. Ellinger S, Reusch A, Stehle P, et al. (2012) Epicatechiningested via cocoa products reduces blood pressure inhumans: a nonlinear regression model with a Bayesianapproach. Am J Clin Nutr 95, 13651377.

41. Hooper L, Kay C, Abdelhamid A, et al. (2012) Effects ofchocolate, cocoa, and flavan-3-ols on cardiovascular health: asystematic review and meta-analysis of randomized trials. AmJ Clin Nutr 95, 740751.

42. Loke WM, Hodgson JM, Proudfoot JM, et al. (2008) Puredietary flavonoids quercetin and ()-epicatechin augmentnitric oxide products and reduce endothelin-1 acutely inhealthy men. Am J Clin Nutr 88, 10181025.

43. Duarte J, Jimenez R, OValle F, et al. (2002) Protective effectsof the flavonoid quercetin in chronic nitric oxide deficient rats.J Hypertens 20, 18431854.

44. Galindo P, Rodriguez-Gomez I, Gonzalez-Manzano S, et al.(2012) Glucuronidated quercetin lowers blood pressure inspontaneously hypertensive rats via deconjugation. PLOS ONE7, e32673.

45. Sanchez M, Galisteo M, Vera R, et al. (2006) Quercetindownregulates NADPH oxidase, increases eNOS activity andprevents endothelial dysfunction in spontaneouslyhypertensive rats. J Hypertens 24, 7584.

46. Ruef J, Moser M, Kubler W, et al. (2001) Induction ofendothelin-1 expression by oxidative stress in vascularsmooth muscle cells. Cardiovasc Pathol 10, 311315.

47. Zhao X, Gu Z, Attele AS, et al. (1999) Effects of quercetin onthe release of endothelin, prostacyclin and tissue plasminogen

activator from human endothelial cells in culture. J Ethno-pharmacol 67, 279285.

48. Ying B, Yang T, Song X, et al. (2008) Quercetin inhibits IL-1beta-induced ICAM-1 expression in pulmonary epithelial cellline A549 through the MAPK pathways. Mol Biol Rep 36,18251832.

49. Kleemann R, Verschuren L, Morrison M, et al. (2011)Anti-inflammatory, anti-proliferative and anti-atheroscleroticeffects of quercetin in human in vitro and in vivo models.Atherosclerosis 218, 4452.

50. Boesch-Saadatmandi C, Wagner AE, Wolffram S, et al. (2012)Effect of quercetin on inflammatory gene expression in miceliver in vivo - role of redox factor 1, miRNA-122 andmiRNA-125b. Pharmacol Res 65, 523530.

51. Mahmoud MF, Hassan NA, El Bassossy HM, et al. (2013)Quercetin protects against diabetes-induced exaggeratedvasoconstriction in rats: effect on low grade inflammation.PLOS ONE 8, e63784.

52. Pfeuffer MA, Auinger A, Bley U, et al. (2013) Effect of quer-cetin on traits of the metabolic syndrome, endothelial functionand inflammatory parameters in men with different APOEisoforms. Nutr Metab Cardiovasc Dis 23, 403409.

53. Loizzo MR, Said A, Tundis R, et al. (2007) Inhibition ofangiotensin converting enzyme (ACE) by flavonoids isolatedfrom Ailanthus excelsa (Roxb) (Simaroubaceae). Phytother Res21, 3236.

54. Mackraj I, Govender T & Ramesar S (2008) The anti-hypertensive effects of quercetin in a salt-sensitive model ofhypertension. J Cardiovasc Pharmacol 51, 239245.

55. Hackl LP, Cuttle G, Dovichi SS, et al. (2002) Inhibition ofangiotesin-converting enzyme by quercetin alters the vascularresponse to brandykinin and angiotensin I. Pharmacology 65,182186.

56. Larson A, Witman MA, Guo Y, et al. (2012) Acute, quercetin-induced reductions in blood pressure in hypertensive indivi-duals are not secondary to lower plasma angiotensin-converting enzyme activity or endothelin-1: nitric oxide.Nutr Res 32, 557564.

57. Conquer JA, Maiani G, Azzini E, et al. (1998) Supplementationwith quercetin markedly increases plasma quercetin con-centration without effect on selected risk factors for heartdisease in healthy subjects. J Nutr 128, 593597.

58. Lee KH, Park E, Lee HJ, et al. (2011) Effects of daily quercetin-rich supplementation on cardiometabolic risks in male smo-kers. Nutr Res Pract 5, 2833.

59. Castilla P, Echarri R, Davalos A, et al. (2006) Concentrated redgrape juice exerts antioxidant, hypolipidemic, and antiin-flammatory effects in both hemodialysis patients and healthysubjects. Am J Clin Nutr 84, 252262.

60. Zern TL, Wood RJ, Greene C, et al. (2005) Grape polyphenolsexert a cardioprotective effect in pre- and postmenopausalwomen by lowering plasma lipids and reducingoxidative stress. J Nutr 135, 19111917.

61. Gnoni GV, Paglialonga G & Siculella L (2009) Quercetininhibits fatty acid and TAG synthesis in rat-liver cells. Eur JClin Invest 39, 761768.

62. Glasser G, Graefe EU, Struck F, et al. (2002) Comparison ofantioxidative capacities and inhibitory effects on cholesterolbiosynthesis of quercetin and potential metabolites. Phyto-medicine 9, 3340.

63. Wiczkowski W, Romaszko J, Bucinski A, et al. (2008) Quercetinfrom shallots (Allium cepa L. var. aggregatum) is morebioavailable than its glucosides. J Nutr 138, 885888.

64. Rothman KJ (1990) No adjustments are needed for multiplecomparisons. Epidemiology 1, 4346.

Effects of quercetin on blood pressure 15

http://www ars usda gov/Services/docs htm?docid=6231http://www ars usda gov/Services/docs htm?docid=6231

Effects of a quercetin-rich onion skin extract on 24h ambulatory blood pressure and endothelial function in overweight-to-obese patients with (pre-)hypertension: a randomised double-blinded placebo-controlled cross-overtrialMethodsSubjectsStudy design

Fig. 1Flow diagram of participants. ABP, ambulatory blood pressure; BP, blood pressureMeasurementsOffice blood pressure and heart rate24h ambulatory blood pressure monitoringVascular testingAnthropometrics

Blood/urine sample processing and analysis

Table 1Flavonoid analysis in onion skin extract powder*(Mean values and standard deviations)Self-reported dietary intake of energy, nutrients and quercetinStatistical analyses

ResultsSubject characteristics at screening, compliance and dietary intakeBody weight, waist and hip circumference, body composition and potential side-effectsAmbulatory blood pressure monitoring and office blood pressure

Table 2Subject characteristics and blood parameters at screening(Mean values and standard deviations)Table 3Measurements of the 24h ambulatory blood pressure and office blood pressure in the total study group during the 6-week dietary supplementation with quercetin or placebo*(Mean values and standard deviations)Serum lipids, lipoproteins, apolipoproteins, glucose and insulinSoluble adhesion molecules, asymmetric dimethylarginine, angiotensin-converting enzyme, C-reactive protein, endothelial function and oxidation

Fig. 2Fasting plasma concentrations of quercetin (a) (n 68) and total flavonols (b) (n 68) before and after the 6-week supplementation with quercetin (162mg/d; ) or placebo (). Values are means and standard deviations reprTable 4Measurements of the 24h ambulatory blood pressure and office blood pressure in the subgroup of hypertensive patients during the 6-week dietary supplementation with quercetin or placebo*(Mean values and standard deviations)Table 5Fasting lipids, lipoproteins, apolipoproteins, insulin, glucose, HbA1c and HOMA IR Index in the total study group (n 68) during the 6-week dietary supplementation with quercetin or placebo*(Mean values and standard deviations)Table 6Adhesion molecules, parameters of oxidation and inflammation and vascular testing in the total study group (n 68) during the 6-week supplementation with quercetin or placebo*(Mean values and standard deviations)Correlations between blood pressure and biomarkers of inflammation/endothelial function

DiscussionBlood pressureEndothelial function, inflammation and oxidationAngiotensin-converting enzymeLipid and glucose metabolismStrengths and limitationsConclusion

AcknowledgementsACKNOWLEDGEMENTSReferences