1 Introduction

Cardiovascular diseases are one of the principal causes of death and disability in the world [1]. Atherosclerosis is considered the main responsible factor of cardiovascular diseases [2]. In the atherosclerotic process, structural and functional changes of the endothelium are strictly involved in atherogenesis and related negative consequences [2]. It has been demonstrated that cardiovascular risk factors are able to increase the oxidative stress, therefore negatively affecting the equilibrium between nitric oxide (NO) and reactive oxygen species, comporting a fundamental decrease in NO availability [2, 3]. The enhancement of the oxidative stress may favor a serious alteration of the NO system, with a consumption of NO involved in a paradoxical increase of reactive oxygen species by NOS (H2O2, hydrogen peroxide; O2·−, superoxide; OH·, hydroxyl radical; ·ONOO–, peroxynitrite) [2,3,4].

Thus, in presence of pathogenic proinflammatory stimuli, the endothelium is “activated” and we observe the expression of adhesion molecules, and inflammatory cytokines on the endothelial surface [2, 3].

Because of the antiatherogenic, antithrombotic properties of NO and the proatherogenic prothrombotic actions of endogenous oxidants, a decreased NO availability with increased oxidative stress will result not only in impaired endothelium-dependent vasorelaxation and blood pressure regulation but also in the acceleration of atherogenesis and onset of acute atherotrombotic events [2, 3, 5]. According with this, endothelial dysfunction and activation have been described to promote the start of a “vicious circle” in which inflammatory factors promote monocyte and T-cell adhesion, foam cell formation, extracellular matrix digestion, as well as vascular smooth muscle migration and proliferation that lead to atherosclerotic plaque formation [3,4,5,6]. Witte et al. [7] reported that sICAM-1 and NO-mediated flow-mediated dilation are both related in healthy individuals, independently of cardiovascular risk factors and CRP, to the estimated risk of coronary heart disease, independently of each other.

Further, circulating levels of sCD40 ligand (sCD40L) are elevated in patients with stable and unstable angina and are associated with increased CV risk in healthy women [8].

With regard to evaluation of oxidative stress, 8-iso-prostaglandin (8-iso-PG) F2, a parameter of lipid peroxidation in vivo was associated with vasoconstrictive and platelet-activating properties, representing a surrogate for increased ROS production [8]. Indeed, increased urinary levels of 8-iso-PGF2 have been reported in association with several cardiovascular risk factors [8, 9].

Recent studies have demonstrated a potential and to some extent unanticipated and unexpected role of cocoa in “promoting health” [10, 11]. Indeed, a large body of evidence supports dietary intake of flavonoids and the specific class of flavanols from cocoa might exert some beneficial vascular effects, reduce the risk of cardiovascular morbidity and mortality, and contribute to the prevention of other chronic diseases [5, 10, 11]. Further, we showed that (–)-epicatechin, the flavanol mostly represented in cocoa, induced a dose-dependent NO-mediated vasorelaxation in isolated rat aortic rings precontracted by phenylephrine [12] and also recently reported that cocoa dose-dependently improved flow-mediated dilation and decreased arterial stiffness and circulating levels of ET-1 also by ameliorating office and monitored blood pressure in healthy subjects [12].

Conflicting results about effects of flavonoids on vascular function and endothelial markers of inflammation and activation have been reported [13,14,15]. Part of this inconsistency may be due to flaws in the study design, statistical power, controlled study design, dose of cocoa and amount of cocoa flavonoids tested varied over a wide range, and the dose–effect relation has never been evaluated [13,14,15].

Therefore, aim of the present study was to assess the effects of a range of doses of cocoa flavonoids on markers of endothelial and platelet activation and oxidative stress in healthy subjects.

2 Methods

20 subjects were recruited at the University of L’Aquila—Department of Internal Medicine and Public Health – among healthy volunteers between 18 and 70 years of age with a sitting systolic–diastolic BP (SBP/DBP) less than 140/90mmHg and BMI at least 19 and less than 30 kg/m2. Volunteers were excluded if they had a chronic–acute disease, including any kind of metabolic abnormality or major cardiovascular risk factor, or both. Habitual smokers and users of prescribed medication and dietary supplement, including those who used any type of medication or supplement, or both, diet restrictions due to any reasons within the last 2 weeks before entering and/or reported participation in another study 3 months before the study was started were also excluded.

To further avoid confounding factors, we also excluded individuals reporting daily intense sporting activities (> 10 h/week), and/or intolerance/allergy to cocoa-based products, and/or weight change > 10% body weight within the last 6 months before entering the study.

The study was approved by the responsible Ethics Committee (n° 27/2007) and all participants gave written informed consent. All clinical investigation have been conducted according to the principles expressed in the Declaration of Helsinki, and all participants gave written informed consent. The present study was part of a larger project and some of the study results have been previously reported with primary outcomes, while in this study we reported secondary outcomes established at the starting protocol definition [12].

2.1 Study Design

20 participants received five one-week periods of cocoa treatments according to a randomized, controlled, double-blind, cross-over design. Total flavonoid content of cocoa during the five one-week periods was randomly changed from 0 mg (control) to 80, 200, 500 and 800 mg/day on each week. Treatment phases were separated by a one-week washout period, during which volunteers consumed their habitual diets. In order to avoid any possible stress-dependent cardiovascular effect, all vascular evaluations were performed in complete relaxed behaviour (noninvasive devices, temperature-controlled room without noise and after at least 15 min at rest), with healthy individuals completely relaxed and without any condition of possible psychic or physical stress.

Individuals were instructed to maintain their usual diet and lifestyle, to avoid flavonoid-rich foods (tea, red wine and chocolate), dietary supplements or NSAIDs and maintain their usual daily intake of fruits and vegetables.

To ensure blinding of the study, all cocoa doses for each day were provided in coded, nontransparent, identical sachets. Before the start of the study and at the four subsequent measurement-visits the volunteers received the intervention products for the next intervention week. Volunteers received also 5 additional cocoa powder sachets to correct mistakes and spillage. The intervention products were ambient stable. Volunteers returned excess sachets and empty sachets to check compliance. Adherence to the study protocol was confirmed at each visit by a check list questionnaire of specified food items and by monitoring daily body weight and physical activity. The product was consumed dissolving the cocoa powder in 150 ml of hot water. Addition of milk and/or sugar was not allowed. Volunteers consumed one dose per day: fasting at least 3 h after breakfast.

Cocoa powder intake was of 10 g (including 0.2% sweetener, and 0.14% anti-caking agent) once a day and the amount of flavonoids in the resulting cocoa was measured with the Folin Ciocalteu assay using gallic acid as a standard. Composition of the test product is reported in a previous study [12]. Regardless of flavonoid content, each sachet presented with similar colour, taste and flavour. Caffeine, theobromine, sodium, potassium and magnesium content was similar in all products. Cocoa powder was provided by Barry Callebaut (Lebbeke-Wieze, Belgium).

2.2 Biomarkers of Endothelial Activation and Oxidative Stress

After each intervention phase, sICAM-1 and sCD40L concentrations were measured by an immunoenzymatic method (R&D Systems, Minneapolis, MN) and plasma total 8-iso-PGF2α levels were assessed by enzyme immunoassay (Assay Design Inc., Ann Arbor, MI) in fasting samples as previously described [16].

2.3 Statistical Analysis

Data are expressed as least square means ± standard error of the mean (SEM). Statistical analyses and power calculation were performed with SAS (version 9.1.3, 2004; SAS Institute Inc., Cary, North Carolina, USA).

Statistical power was based on results obtained in similar studies with healthy individuals who consumed a chocolate bar containing 500 mg polyphenols (11). In that study the mean ambulatory BP decreased by 6.5mmHg at a SD of 5.8mmHg. On the basis of these data a mean difference of 4 mmHg should be detected in a group of 20 individuals (two-sided, alpha = 0.05 and power = 0.8). In the same study FMD increased by 1.9% at a SD of 1.6% after cocoa polyphenol consumption. On the basis of these data a 1.5% improvement in FMD could be detected in 20 individuals with a very high power (0.99 at alpha = 0.05).

Results were evaluated using a linear mixed model with Proc Mixed Procedure with subjects treated as a random factor and treatment and sequence as fixed factors. Estimates for differences between mean responses at each dose [0 (control), 80, 200, 500 and 800 mg tea flavonoids per day] were computed using Dunnett-Hsu correction for multiple comparisons.

3 Results

3.1 Baseline Data

According to the inclusion/exclusion criteria participants presented with normal ranges for BMI, lipid profiles, and gluco-insulin metabolism. The baseline clinical characteristics of the 20 volunteers have been previously reported [12]. Twenty-three individuals were recruited; one smoker and two individuals reporting gastric disturbances were excluded. All the 20 individuals enrolled completed all the study phases. No dropps out were observed. Compliance was 100% in all the volunteers for all the study phases.

3.2 sICAM-1 Levels

Compared with control, cocoa decreased sICAM-1 levels [from 1190,2 (control) to 1123.0; 906.3; 741.7 and 625.6 pg/mL; after the different flavonoid doses, respectively (p for treatment = 0.0005)]. Compared with control, significant changes were observed for doses ≥ 500 mg of flavonoids (p = 0.019 and p = 0.001, for 500 and 800 mg, respectively) (Fig. 1).

Fig. 1
figure 1

Effects of cocoa on sICAM-1 levels in 20 healthy volunteers. Data are expressed as least square means with standard error of the mean. Data points with different superscripts are significantly different. Differences are considered significant when p < 0.05

Full size image

3.3 sCD40L Levels

Compared with control, cocoa decreased sCD40L levels [from 218.8 (control) to 210.2; 165.5; 134.5 and 128.4 pg/mL; after the different flavonoid doses, respectively (p for treatment = 0.006)]. Compared with control, significant changes were observed for doses ≥ 500 mg of flavonoids (p = 0.023 and p = 0.013, for 500 and 800 mg, respectively) (Fig. 2).

Fig. 2
figure 2

Effects of cocoa on sCD40L levels in 20 healthy volunteers. Data are expressed as least square means with standard error of the mean. Data points with different superscripts are significantly different. Differences are considered significant when p < 0.05

Full size image

3.4 8-iso-PGF2 Levels

Compared with control, cocoa decreased 8-iso-PGF2 levels [from 4703.9 (control) to 4670.7; 2000.1; 2098.4 and 2052.3 pg/mL; after the different flavonoid doses, respectively (p for treatment = 0.0008)]. Compared with control, significant changes were observed for doses ≥ 200 mg of flavonoids (p = 0.025; p = 0.034 and p = 0.029, for 200, 500 and 800 mg, respectively) (Fig. 3).

Fig. 3
figure 3

Effects of cocoa on 8-iso-PGF2 levels levels in 20 healthy volunteers. Data are expressed as least square means with standard error of the mean. Data points with different superscripts are significantly different. Differences are considered significant when p < 0.05

Full size image

4 Discussion

In this randomized, controlled, double-blind, crossover study we demonstrated for the first time that short-term cocoa consumption dose-dependently decreased the circulating levels of sICAM-1, sCD40L and 8-iso-PGF2 levels, therefore improving oxidative stress, proinflammatory mediators and lipid peroxidation, with a significant effect on circulating levels of these markers for higher dosages of flavonoids. Moreover, the present study is also the first reporting dose-dependent effects of cocoa flavonoids on markers of endothelial damage and oxidative stress. In line with this, and in contrast with the findings reported on endothelial function (only 10 g of cocoa with a very low caloric and flavanol intake per day were already significantly amelioring vascular function) [12] the current results suggest that effects we observed were effecctive only for higher dosages of flavonoids.

Althought indirectely, cardiovascular biomarkers of damage reflect early functional or morphological changes, well before overt disease manifests. The up-regulation of the expression of adhesion molecules in the vascular endothelium is the initiating event in this process and allows leucocyte and monocyte adhesion to the endothelial surface. Up-regulation of endothelial adhesins induces leucocyte and monocyte penetration into the sub-endothelial environment with consequent recruitment of additional circulating cells. The membrane-spanning protein CD40 is also up-regulated in activated endothelial cells and leucocytes and, after engagement with its natural ligand sCD40L, which is expressed in monocytes and T lymphocytes, amplifies atherogenesis by further promoting cytokine release and adhesion of circulating cells to the endothelium [2, 8, 16,17,18,19,20,21]. Moreover, oxidized, but not native low-density lipoprotein (LDL), up-regulates endothelial adhesins expression in vitro and promotes endothelial–leukocyte interactions both directly and by modulating endothelial response to cytokines. In fact, antioxidants prevent adhesion molecule expression in endothelial cells both in vitro and in vivo [8, 16,17,18]. Reactive oxygen species also reduce eNOS activity and increase NO breakdown and thereby reduce the bioactivity of NO, a powerful inhibitor of NF-KkB activation under physiologic conditions.

It has been hypothesised that this endothelial-dependent vascular imbalance is critical, not only in the initiation and progression of atherosclerosis, but also in the transition from a stable to an unstable disease state with the precipitation to acute vascular events [2, 4, 16,17,18]. According to this, it has been supposed that a dysfunctional endothelium may promote plaque activation, leading to a higher plaque vulnerability [2, 4].

Studies on endothelial dysfunction have focussed on two distinct mechanisms; dysfunction of the NO-mediated vasodilator effect of endothelium, and disruption of the equilibrium between pro- and antiinflammatory mechanisms. In keeping with this, circulating levels of sCD40L are elevated in patients with stable and unstable angina [20] and are associated with increased CV risk in healthy women [21]. With regard to evaluation of oxidative stress, 8-iso-PGF2, an index of lipid peroxidation in vivo endowed with vasoconstrictive and platelet-activating properties, represents a surrogate for increased ROS production [18, 19]. Accordingly, interesting feature of the present study is the use of this plasma redox biomarker to monitor oxidative stress levels. Indeed, it is of great importance to use such biomarkers in every day practice (either clinical or nutritional), since oxidative stress has been increasingly associated with many physiological and pathological conditions [22].

Indeed, 8-iso-PGF2, apart from being indicators of lipid peroxidation and being risk/progression factor for coronary heart disease, they are also involved in different additional biological conditions. In particular, it has been proposed that 8-iso-PGF2 exert cardiovascular effects by inhibiting angiogenesis through activation of the thromboxane A2 receptor Moreover, F2-isoprostanes have been implicated in the production of hepatic collagen, in the contraction of human bronchial smooth muscle, in platelet activation, in mitogenesis of vascular smooth muscle cells and in the vasoconstriction of the renal glomerular arterioles [22]. Moreover, associations have been found between sICAM-1 and cardiovascular mortality in both healthy individuals [7] and populations at high risk [23, 24]. In particular, Witte et al. [7] reported that in both sICAM-1 and FMD are related to the estimated risk of coronary heart disease in healthy subjects, independently of each other.

On the other hand, we previously reported that cocoa dose-dependently improved NO-dependent flow-mediated dilation and decreased arterial stiffness and ET-1 levels also by ameliorating office and monitored blood pressure [12]. According to this, our findings ulteriorly support the epidemiological evidence suggesting an inverse relationship between cocoa and cardiovascular disease [5, 15]. In line with our findings but in a different study design in obese patients, Stote et al. [25] reported that cocoa was able to decrease 8-isoprostanes, CRP and IL-6 but not sICAM-1 in response to an oral glucose load with a sort of dose-dependent effect and with medium dose having the maximal lowering effect for 8-isoprostanes and IL-6.

Of particular interest, a meta-analysis aiming to define the putative role of cocoa intake on biomarkers of inflammation and endothelial activation, reported that little evidence exists that consumption of cocoa-rich food may reduce inflammation, probably by lowering the activation of monocytes and neutrophils [26]. Specifically, authors observed that after consumption of cocoa-rich food, most studies suggested a decrease in inflammatory biomarkers in healthy subjects after 2–6 h. Nevertheless, Ellinger et al. [26] also report that different results are also linked to the different consumption of cocoa in studies (bolus and regular consumption studies) in healthy subjects, therefore suggesting about the possible difference in flavonoids bioavailability and effective after cocoa intake also considering the lack information on dose-dependent effects and also supporting the possibility that the different effects observed could depend on the extent of the basal inflammatory burden.

Exacly in line with our results, the recent COcoa Supplement and Multivitamin Outcomes Study (COSMOS) by Sesso et al. [27] strongly suggested that cocoa flavanols supplementation (500 mg), during a median follow-up of 3.6 y, significantly decreased CVD death by 27%. Our results support for the first time the possible fundamental role of the dose of flavonoids in the observed findings on inflammation, endothelial activation and lipid peroxidation. Indeed, althought a trend to decrease these markers is evident also for lower doses, we observed significant effects only after higher doses of flavonoid intake (≥ 200 mg of flavonoids 8-iso-PGF2 levels; ≥ 500 mg of flavonoids for sICAM-1 and sCD40L).

5 Conclusions

Atherosclerotic process is supported by early stages of endothelial dysfunction and activation, specifically supporting an inflammatory process enhancing a negative vicious cyrcle. According to this, it has been supposed that endothelial activation may promote plaque activation, leading to a higher plaque vulnerability [2, 4, 9] and thus, favouring cardiovascular events.

Our findings indicate that cocoa exerts significant effects in deacreasing vascular inflammation ednothelial activation and lipid peroxidation, also suggesting higher doses of flavonoids seem to be better to obtain this effect. As suggeted in the recent meta-analysis [26], the robust nature of our study design suggests that cocoa, with very-low calorie intake, presents with dose-dependent effects and satisfiy at least in part the lacking information on the argument. The observed responses suggest different and additional possible pathophysiological mechanisms by the way cocoa flavonoids might improve the cardiovascular risck possibly suggesting cocoa with high content in flavonoids could be considered in the complexity of a daily diet for cardiovascular prevention.

6 Study Limitations

The study provided short-term intervention phases, therefore suggesting evidence of nutritional effects only after a brief period. Our results support dose-dependent responses that should be confirmed in longer-term intervention studies.