Flavan-3-ol-enriched dark chocolate and white chocolate improve acute measures of platelet function in a gender-specific way—a randomized-controlled human intervention trial

Authors


Correspondence: Dr. Baukje de Roos, Rowett Institute of Nutrition & Health, University of Aberdeen, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK

E-mail: b.deroos@abdn.ac.uk

Fax: +44-1224-438629

Abstract

Scope

We examined whether flavan-3-ol-enriched dark chocolate, compared with standard dark and white chocolate, beneficially affects platelet function in healthy subjects, and whether this relates to flavan-3-ol bioavailability.

Methods and results

A total of 42 healthy subjects received an acute dose of flavan-3-ol-enriched dark, standard dark or white chocolate, in random order. Blood and urine samples were obtained just before and 2 and 6 h after consumption for measurements of platelet function, and bioavailability and excretion of flavan-3-ols. Flavan-3-ol-enriched dark chocolate significantly decreased adenosine diphosphate-induced platelet aggregation and P-selectin expression in men (all p ≤ 0.020), decreased thrombin receptor-activating peptide-induced platelet aggregation and increased thrombin receptor-activating peptide-induced fibrinogen binding in women (both p ≤ 0.041), and increased collagen/epinephrine-induced ex vivo bleeding time in men and women (p ≤ 0.042). White chocolate significantly decreased adenosine diphosphate-induced platelet P-selectin expression (p = 0.002) and increased collagen/epinephrine-induced ex vivo bleeding time (p = 0.042) in men only. Differences in efficacy by which flavan-3-ols affect platelet function were only partially explained by concentrations of flavan-3-ols and their metabolites in plasma or urine.

Conclusion

Flavan-3-ols in dark chocolate, but also compounds in white chocolate, can improve platelet function, dependent on gender, and may thus beneficially affect atherogenesis.

Abbreviations
ADP

adenosine diphosphate

COL/EPI

collagen/epinephrine

PFA-100

Platelet Function Analyzer-100

TRAP

thrombin receptor-activating peptide

1 Introduction

Cardiovascular disease is widespread in industrialized countries [1] and accounts for more deaths each year than any other disease in Europe [2] and the United States [3]. Sound nutritional practices have been shown to reduce the risk of cardiovascular disease by up to 70% through modulation of risk factors [4]. Platelet function is related to the risk of deve-loping atherothrombosis, the principal cause of heart attack and stroke. Indeed, activated blood platelets play a central role in this chronic inflammatory condition as they are major components of thrombi that occlude arteries [5] and they also contribute to plaque formation within blood vessels in the early stages of atherogenesis [6].

Platelet function can be affected by various dietary components, including a wide range of plant-based products. Consumption of polyphenols from fruits, vegetables, herbs, spices, teas, and wines has been linked to improved platelet function in vitro and in vivo [7-11], however, consumption of cocoa beverages and chocolate rich in flavan-3-ols, produce the most consistent beneficial effects on platelet function [7]. Several human intervention studies have assessed the effects of cocoa or chocolate on platelets [7, 12-14], using only one or two methods to measure platelet function. This approach only provides limited insights into the complex physiological behavior of platelets, making it difficult to assess by which mechanisms flavan-3-ols could affect platelet function. Furthermore, many studies did not provide details on the exact amounts of flavan-3-ols that were ingested, and subsequently how much of the ingested amount was available to affect platelet function.

After ingestion, flavan-3-ols, like most polyphenols, undergo rapid metabolism by several enzymes in intestinal cells leading to the formation of sulfates, glucuronides, and/or methylated metabolites. These are partially absorbed into the circulatory system and undergo further metabolism in the liver followed either by reabsorption in the blood stream and excretion in urine or by recirculation into the small intestine within bile. The fraction of flavan-3-ols that is not absorbed from the small intestine proceeds to the colon and is metabolized by gut microbiota to mainly small phenolic compounds. Some of those are absorbed into the circulation [15, 16]. In general, the bioavailability of most polyphenols from food sources, including flavan-3-ols, is considered to be low [17].

In this study we assessed effects of acute consumption of flavan-3-ol-enriched dark chocolate, in comparison with standard dark and white chocolate, on platelet function by measuring platelet aggregation, ex vivo bleeding time and platelet activation in healthy subjects. Furthermore, we assessed the bioavailability of flavan-3-ols by measuring postprandial concentrations of these compounds and their metabolites in plasma and urine. This approach enabled us to assess whether availability of flavan-3-ols and their metabolites affects the efficacy of these compounds to beneficially affect platelet function.

2 Materials and methods

For detailed descriptions of subjects, study design, materials, and methods please see the Supporting Information.

2.1 Subjects and study design

We recruited 42 healthy subjects for this observer-blinded randomized-controlled acute intervention trial using inclusion and exclusion criteria as indicated in the Supporting Information. Figure 1A shows a flowchart of the study design, including randomization, treatments, and volunteer allocation and Fig. 1B shows the setup of an intervention day. The nutritional composition of the treatment chocolate bars, provided by Natraceutical Industrial S. L. U., Spain, is shown in Table 1, and a full list of ingredients for all chocolates is shown in the Supporting Information Table S1. Written informed consent was obtained from all subjects before participation. The study was approved by the Ethics Committee of the Rowett Institute of Nutrition and Health, University of Aberdeen (reference number 09–002) following review by the North of Scotland Research Ethics Committee. This trial was registered at clinicaltrials.gov as NCT01099150.

Table 1. Nutritional composition of chocolate bars
60 g chocolate contained:Flavan-3-ol- enriched DCStandard DCWhite chocolate
  1. Data were analyzed and provided by Natraceutical Industrial S.L.U., Valencia, Spain.

  2. a

    Matches 4% of the maximum recommended daily intake [18]; comparable to amount of cholesterol ingested when drinking 250 mL of low fat milk [19]. DC, dark chocolate.

Energy (kcal)314.48307.63306.62
Protein (g)7.297.620.10
Carbohydrates (g)17.7816.7024.55
Sugar (g)13.8912.7824.55
Starch (g)3.893.920
Lipids (g)23.2823.3723.11
Saturated fat (g)14.4114.4614.11
Polyunsaturated fat (g)0.860.860.93
Monounsaturated fat (g)8.018.066.65
Trans unsaturated fat (g)0.000.000.20
Cholesterol (mg)0.000.0012.24a
Fiber (g)7.998.530.00
Water (g)1.070.890.04
Caffeine (g)0.090.110.00
Theobromine (g)0.580.630.00
Sodium (mg)2.392.390.64
Figure 1.

(A) Study design. The study was designed as an observer-blinded randomized acute intervention trial with a latin-square format. (B) Intervention day setup. Volunteers came in after having fasted ≥ 10 h overnight. Arrows show time points when blood and urine samples were obtained. The star (*) marks the time point when 60 g chocolate were consumed together with 400 mL still table water. The circle (⁰) marks the time points when 200 mL still table water were consumed. DCR, flavan-3-ol-enriched dark chocolate; SDC, standard dark chocolate; WC, white chocolate.

2.2 Extraction and analysis of flavonoid content of the chocolate bars

Lipids were removed from chocolate bar samples as described by Cooper et al. [20]. Defatted samples were then extracted using a method similar to that reported by Cooper et al. [20]. This included the extraction of 40 mg defatted and dried chocolate powder using 950 μL 70% aqueous methanol. In addition, 50 μL 0.1 mg/mL galangin, dissolved in methanol, was added as internal standard and samples were heated to 70°C for 20 min. Supernatants were then filtered (pore size: 0.22 μm) and filtrates from each sample were analyzed by normal-phase HPLC as described by Prior et al. [21] and Gu et al. [22]. Samples were analyzed in duplicates and the resulting flavonoid contents of the chocolate bars are shown in Table 2.

Table 2. Flavonoid content of chocolate bars
60 g chocolate contained:Flavan-3-ol- enriched DCStandard DCWhite chocolate
  1. Data are presented as means (± SD). Samples were analyzed in duplicates. DC, dark chocolate; ND, not detectable.

Epicatechin (mg)257.0 (± 1.06)84.1 (± 0.67)ND
Catechin (mg)53.6 (± 0.27)25.8 (± 1.02)ND
Dimer B2 (mg)198.0 (± 1.22)74.4 (± 0.76)ND
Trimers (mg)168.0 (± 1.42)47.0 (± 2.23)ND
Tetramers (mg)105.5 (± 12.75)32.1 (± 4.40)ND
Pentamers (mg)125.4 (± 6.04)118.8 (± 39.50)ND
Total flavonoids (mg)907.4 (± 22.75)382.3 (± 45.20)ND

2.3 Assessment of platelet aggregation in platelet-rich plasma

Sample preparation for and measurement of light transmission aggregometry were carried out as described by us previously [23]. Briefly, the platelet number in platelet-rich plasma was adjusted to 300 × 109/L, and platelet aggregation was initiated by addition of 5 or 8 μmol/L adenosine diphosphate (ADP) or 20 μmol/L thrombin receptor-activating peptide (TRAP, a thrombin analog) and assessed using the light transmission method of Born [24] using a Helena PACKS-4 aggregometer over a period of 10 min. All measurements were carried out in duplicate. Aggregation was quantified based on area under the aggregation curve. All tests were carried out within 2 h after blood sampling.

2.4 Assessment of ex vivo bleeding time in whole blood

Within 90 min of blood sampling, Platelet Function Analyzer (PFA)-100 closure time was measured using a collagen/epinephrine-(COL/EPI-) coated cartridge.

2.5 Assessment of platelet activation in diluted whole blood

Sample preparation for and measurement of platelet activation were carried out as described by us previously [23]. Platelet activation was initiated by addition of 10 μmol/L ADP or 25 μmol/L TRAP and a combination of fluorochrome-labelled CD61 and CD62P, or CD61 and fibrinogen was used to quantify P-selectin expression or fibrinogen binding (and therefore integrin αIIbβ3 activation) on the platelet surface, respectively. Samples were analyzed for activation of platelets using a BD FACSArray Bioanalyzer. Results were analyzed as percentage of positive fluorescent platelets binding CD62P or fibrinogen in a total of 10 000 events as seen by flow cytometry.

2.6 Extraction of flavan-3-ols from plasma and urine

Epicatechin and conjugated epicatechin metabolites were extracted from plasma and urine as described by Saha et al. [25]. Samples for the standard curve were treated as described by Saha et al. [25], but were prepared by spiking human serum samples or human urine baseline samples with (+)-catechin, (-)-epicatechin, and procyanidin dimer B2.

2.7 Analysis of total epicatechin and distribution of flavan-3-ol conjugates in plasma and urine

Processed plasma and urine samples were analyzed using a LC-MS/MS approach as described by Saha et al. [25] with the only difference being the addition of procyanidin dimer B2 (m/z 577/289) as reference compound. For representative examples of all standard curves please see the Supporting Information Fig. S2. Values for “total catechins” from hydrolyzed samples are the sum of the individual values for (+)-catechin and (-)-epicatechin. Urinary analytes are given as amount of substance of the analyte per 1 mol creatinine to adjust for different fluid uptake and urine volumes.

2.8 Statistical analyses

Responses at 2 and 6 h after treatment were considered separately. Changes from baseline (t = 0 h) for all measurements, and total values of flavan-3-ol concentrations in plasma and urine, within the whole study population were analyzed by a mixed model, using the residual maximum likelihood approach. Random effect terms were subject and period. Fixed effect terms were period, treatment, gender, and the interaction between treatment and gender. Significance was tested by the Wald statistic, with estimated degrees of freedom in the denominator. Post-hoc multiple comparisons were carried out using two-tailed Student's t-tests. As some significant treatment × gender interactions were found, all outcomes were also analyzed within females and males separately by a mixed model using the residual maximum likelihood approach. These separate analyses allowed for any small differences in variance between males and females. Random effect terms were again subject and period. Fixed effect terms were period and treatment. The test for significance and post-hoc multiple comparisons were carried out as described above. For both, the total study population as well as females and males separately, data for catechin concentrations in plasma and urine, and dimer concentrations in urine showed non-normally distributed residuals and thus were log transformed before statistical analysis. We included age, hormonal status, and smoking behavior as potentially confounding factors into all analyses, however, none of these parameters had any significant effect on the results of any of the conducted measurements.

Differences between genders for all baseline measurements were analyzed using unpaired two-sample t-tests. Exploratory statistical analysis was used to assess associations between variables by the Pearson correlation coefficient and gender differences between correlations were assessed using the Fisher Z transformation of the correlation coefficients. All statistical analyses were done with GenStat version 13 (VSN International, UK) and differences were considered to be significant at the level of p < 0.05.

3 Results

3.1 Baseline characteristics of subjects

The baseline characteristics of the 42 healthy subjects are summarized in Table 3. COL/EPI-induced PFA-100 closure time at baseline was significantly longer in women compared with men (Supporting Information Table S2). There were no significant differences between genders in other measurements of platelet function at baseline.

Table 3. Baseline characteristics for study subjects
 TotalFemalesMalespa
  1. Data are presented as means ± SEM (range) or means (percentage of total).

  2. a

    p value for difference between genders (female versus male).

  3. *Significant difference between genders (p < 0.05). BP, blood pressure; RBC, red blood cells; WBC, white blood cells.

Number422616NA
Age (years)41 ± 2.0 (23–65)38 ± 2.4 (23–62)46 ± 3.4 (25–65)0.057
BMI (kg/m2)24.5 ± 0.41 (19.4–31.4)24.0 ± 0.54 (19.4–31.4)25.2 ± 0.60 (20.9–28.9)0.167
BP (mm Hg)
Systolic130 ± 2.7 (99–169)124 ± 3.2 (99–159)139 ± 3.8 (120–169)0.005*
Diastolic81 ± 1.5 (66–107)80 ± 1.8 (66–107)84 ± 2.4 (71–106)0.110
Pulse (bpm)72 ± 1.7 (48–95)75 ± 1.9 (61–95)67 ± 3.1 (48–86)0.029*
WBC (109/L)5.4 ± 0.17 (3.9–8.5)5.8 ± 0.23 (4.0–8.5)4.9 ± 0.19 (3.9–7.4)0.008*
RBC (1012/L)4.5 ± 0.06 (3.3–5.5)4.4 ± 0.06 (3.3–4.7)4.8 ± 0.09 (4.1–5.5)<0.001*
Platelets (109/L)238 ± 8.5 (171–449)258 ± 11.2 (174–449)205 ± 7.5 (171–259)<0.001*
Hemoglobin (g/dL)13.6 ± 0.17 (12.1–16.2)13.0 ± 0.11 (12.1–14.2)14.7 ± 0.24 (12.9–16.2)<0.001*
Hematocrit (%)40.2 ± 0.41 (36.2–46.0)38.7 ± 0.31 (36.2–42.2)42.7 ± 0.54 (39.0–46.0)<0.001*
Smoking3 (7.1%)2 (7.7%)1 (6.3%)NA
Contraceptives11 (26.2%)11 (42.3%)NANA

3.2 Effects on platelet aggregation in platelet-rich plasma

Platelet aggregation induced by 5 μmol/L ADP was significantly reduced 2 h after consumption of flavan-3-ol-enriched dark and standard dark chocolate, and 6 h after consumption of flavan-3-ol-enriched dark chocolate, compared with white chocolate, in men but not in women (p = 0.008 and p = 0.020 in men veresus p = 0.334 and p = 0.490 in women) (Fig. 2A). The interaction between treatment and gender was significantly different 2 h after consumption of flavan-3-ol-enriched dark chocolate (p = 0.014) and 6 h after consumption of white chocolate (p = 0.042) (Fig. 2A). Platelet aggregation induced by 8 μmol/L ADP was significantly decreased 2 h, but not 6 h, after consumption of flavan-3-ol-enriched dark and standard dark chocolate, compared with white chocolate, in men only (p = 0.010, p value for interaction between treatment and gender: p = 0.089) (Fig. 2B).

Figure 2.

(A) 5 μmol/L ADP-, (B) 8 μmol/L ADP-, and (C) 20 μmol/L TRAP-induced platelet aggregation in platelet-rich plasma. Boxes show median changes of AUCs, compared with baseline, as well as 25th and 75th percentiles. Error bars mark 10th and 90th percentiles. Filled circles show all remaining data including minimum and maximum values. Right-hand y-axes show percentage changes compared with baseline. *Significant difference between treatments (p < 0.05). Δ Significant interaction of treatment and gender (p < 0.05).

Platelet aggregation induced by 20 μmol/L TRAP was significantly decreased 2 h, but not 6 h, after consumption of flavan-3-ol-enriched dark chocolate compared with white chocolate in women only (p = 0.010, p value for interaction between treatment and gender: p = 0.213) (Fig. 2C).

3.3 Effects on ex vivo bleeding time in whole blood

COL/EPI-induced PFA-100 closure time was significantly increased 6 h, but not 2 h, after the consumption of flavan-3-ol-enriched dark chocolate compared with standard dark, but not white chocolate, in the total study population (p = 0.011) (Fig. 3). The interaction between treatment and gender was significantly different 6 h after consumption of white chocolate (p = 0.035) (Fig. 3). In women, COL/EPI-induced PFA-100 closure time was significantly increased 6 h, but not 2 h, after the consumption of the flavan-3-ol-enriched dark chocolate compared with white chocolate (p = 0.016) (Fig. 3). In men, COL/EPI-induced PFA-100 closure time was significantly increased 6 h, but not 2 h, after the consumption of both, flavan-3-ol-enriched dark and white chocolate, compared with standard dark chocolate (p = 0.042) (Fig. 3).

Figure 3.

Collagen/epinephrine-induced changes in ex vivo bleeding time. Boxes show median changes of closure times as well as 25th and 75th percentiles. Error bars mark 10th and 90th percentiles. Filled circles show all remaining data including minimum and maximum values. Right-hand y-axes show percentage changes compared with baseline. *Significant difference between treatments (p < 0.05). ΔSignificant interaction of treatment and gender (p < 0.05).

3.4 Effects on markers of platelet activation in diluted whole blood

Platelet P-selectin expression induced by 10 μmol/L ADP significantly decreased 2 h, but not 6 h, after consumption of flavan-3-ol-enriched dark and white chocolate, compared with standard dark chocolate, in men only (p = 0.002) (Fig. 4A). Additionally, the interaction between treatment and gender was significantly different 2 h after consumption of standard dark chocolate (p = 0.034) (Fig. 4A). Fibrinogen binding induced by 25 μmol/L TRAP was significantly increased 2 and 6 h after consumption of flavan-3-ol-enriched dark chocolate, compared with white chocolate, in the total study population (p = 0.014 and p = 0.021, respectively) (Fig. 4B) and 2 h, but not 6 h, after consumption of flavan-3-ol-enriched dark chocolate, compared with white chocolate, in women only (p = 0.041, p value for interaction between treatment and gender: p = 0.304) (Fig. 4B). Platelet P-selectin expression induced by 25 μmol/L TRAP, and fibrinogen binding induced by 10 μmol/L ADP, did not change significantly after consumption of any of the chocolates (Supporting Information Table S3).

Figure 4.

(A) Platelet P-selectin expression induced by 10 μmol/L ADP and (B) fibrinogen binding induced by 25 μmol/L TRAP. Boxes show median changes of activated platelets as well as 25th and 75th percentiles. Error bars mark 10th and 90th percentiles. Filled circles show all remaining data including minimum and maximum values. Right-hand y-axes show percentage changes compared with baseline. *Significant difference between treatments (p < 0.05). ΔSignificant interaction of treatment and gender (p < 0.05).

For the data presented in Figs. 2-4 please see Table S3 in the Supporting Information.

3.5 Levels of total epicatechin and distribution of flavan-3-ol conjugates in plasma and urine

Plasma concentrations of total (epi)catechin aglycones (posthydrolysis) significantly increased 2 h after consumption of enriched dark or standard dark chocolate compared with white chocolate (p < 0.001) (Fig. 5A), with mean concentrations peaking at around 1.20 μmol/L 2 h after consumption of flavan-3-ol-enriched dark chocolate. Levels of total plasma ca-techins decreased again 6 h postconsumption but the signifi-cant difference in plasma catechin concentrations between all chocolate interventions remained (p ≤ 0.001) (Fig. 5A). Urine concentrations of total catechins significantly increased 2 and 6 h after consumption of enriched dark or standard dark chocolate compared with white chocolate (p < 0.001) (Fig. 5B). For most individuals, 6 h after consumption of the enriched dark chocolate urinary catechin concentrations peaked at a mean value of 13.4 mmol/mol creatinine. Similar effects were observed for urinary procyanidin dimer B2 (p < 0.001), peaking at 57 μmol/mol creatinine at 6 h (Fig. 5C). For the data presented in Fig. 5 please see Table S4 in the Supporting Information.

Figure 5.

(A) Plasma catechin concentrations, (B) urinary catechin concentrations, and (C) urinary dimer concentrations. Urinary analytes (parts B and C) are given as amount of substance of the analyte per 1 mol creatinine to adjust for different urine volumes. Filled circles show medians and error bars mark 25th and 75th percentiles. Values in filled boxes on top of graphs show respective maximum values. *p ≤ 0.001, all treatments differ significantly from each other; +p < 0.001, flavan-3-ol-enriched dark chocolate is significantly different from standard dark and white chocolate; p ≤ 0.008, flavan-3-ol-enriched dark and standard dark chocolate are significantly different from white chocolate; Δ significant interaction of treatment and gender (p = 0.047); Δ1 significant interaction of treatment and gender (p = 0.017) for flavan-3-ol-enriched dark chocolate. Mean RF, mean response factor ± SEM for area ratio analyte/IS peak area and IS = 10 μg/mL taxifolin.

Peak flavan-3-ol concentrations appeared to be slightly higher in plasma and urine of women compared with men (p values for interaction between treatment and gender: 0.047 ≤ p ≤ 0.388, Fig. 5A and B, Supporting Information Table S4). This interaction of treatment and gender was statistically significant only for catechin concentrations 6 h after consumption of standard dark chocolate in urine (p = 0.047) (Fig. 5B, Supporting Information Table S4). However, this gender difference has most likely been introduced by correc-ting the urinary analyte concentrations for creatinine excretion to standardize for diuresis (Supporting Information Fig. S3). Indeed, it has been reported before that urinary creatinine is higher in men than in women [26-28] and we have also seen a significant difference in urinary creatinine between the two genders in our study (p ≤ 0.005, data not shown).

We assume that the plasma and urinary concentrations, as reported in Fig. 5A–C, are underestimated as the enzymatic hydrolysis of sulfates and methylated sulfates worked only partially. Using the MS in Multiple Reaction Monitoring mode (negative ion mode), we observed large ion currents consistent with the presence of (epi)catechin monosulfates (m/z = 369/289) and methyl-(epi)catechin monosulfates (m/z = 383/303) in the respective chromatograms in hydrolyzed urine samples, especially after consumption of flavan-3-ol-enriched dark chocolate (data not shown). Nevertheless, mean total catechin concentrations in hydrolyzed urine samples were still 5.8 times higher compared with catechin concentrations in nonhydrolyzed samples: 13.4 mmol/mol creatinine versus 2.32 mmol/mol creatinine, respectively. In addition, methyl-(epi)catechins could not be quantified due to a lack of the standard compounds and thus have not been accounted for in our estimations.

For the distribution of catechins and their human metabolites in nonhydrolyzed urine samples please see Suppor-ting Information Fig. S4. (Epi)catechin glucuronides were the most abundant metabolites, closely followed by (epi)catechin monosulfates and methyl-(epi)catechin monosulfates as seen previously [29]. However, because of the unavailability of standards for these compounds, results could not be quantified.

3.6 Correlation analysis

Changes in urinary catechins, urinary dimers, and plasma catechins significantly correlated with some of the platelet function measurements (Table 4).

Table 4. Significant correlations between measures of flavan-3-ol bioavailability and excretion and measures of platelet function as assessed by the Pearson correlation coefficient (r)
MeasurementsTotal population (n = 42)Females (n = 26)Males (n = 16)pa
 rprprp 
  1. a

    p value for difference in correlation between genders (female versus male).

  2. *Significant correlation between measurements (p < 0.05). FibB, fibrinogen binding; CT, closure time; Δ, change; P-sel, P-selectin

Δ catechins in plasma 2 h-0.310.006*0.170.0930.500.004*0.276
 Δ FibB 25 μmol/L TRAP 2 h       
Δ catechins in plasma 2 h-0.320.004*0.220.030*0.440.013*0.472
 Δ FibB 25 μmol/L TRAP 6 h       
Catechins in plasma 6 h-0.290.007*0.230.019*0.440.009*0.490
 Δ FibB 25 μmol/L TRAP 6 h       
Δ catechins in urine 2 h-0.300.001*0.360.000*0.170.2510.555
 Δ PFA-100 CT 6 h       
Δ dimers in urine 2 h-−0.280.003*−0.320.001*−0.250.1150.826
 Δ AUC 20 μmol/L TRAP 2 h       
Δ dimers in urine 2 h-−0.310.001*−0.270.003*−0.390.008*0.697
 Δ P-sel 10 μmol/L ADP 6 h       
Δ dimers in urine 6 h-−0.310.001*−0.390.000*−0.220.1710.589
 Δ AUC 20 μmol/L TRAP 2 h       

4 Discussion

In this study we saw that consumption of flavan-3-ol-enriched dark chocolate, compared with standard dark chocolate, bene-ficially affected measures of postprandial platelet function in healthy subjects in a gender-dependent way. As both dark chocolates only differed in the type of cocoa powder used to prepare the chocolates (Supporting Information Table S1), it is likely that flavan-3-ols are responsible for the observed anti-platelet effects, despite their relatively low bioavailabi-lity. White chocolate, compared with standard dark chocolate, beneficially affected platelet activation and ex vivo bleeding time, in men only. As the white chocolate did not contain any flavonoids (Table 2), other compounds would have to be responsible for the anti-platelet effects, for instance the small amount of whey protein (Supporting Information Table S1). Indeed, some peptides of whey protein have previously been shown to inhibit platelet function in vitro [30-33].

Acute consumption of flavan-3-ols from dark chocolate beneficially affected different measures of platelet function (Figs. 2-4) to a similar extent as seen in previous stu-dies [7, 13, 14] (please also see Supporting Information Fig. S5), and comparable with the anti-platelet effects of aspirin [7, 14]. Platelet aggregation and activation were mostly affected 2 h after consumption, when the maximum flavan-3-ol concentration in plasma was reached. Bleeding time, however, only significantly changed 6 h after consumption, and may therefore have been affected by different compounds, e.g., early colonic metabolites. However, effects on platelet function were gender-dependent. It is known that platelets from men generally respond stronger to activation of the α2 adre-nergic (receptor for epinephrine) and serotonin signaling pathways [34], and show stronger thromboxane A2 receptor-related aggregation responses [35]. We observed in our study that flavan-3-ol-enriched dark chocolate significantly inhi-bited ADP-induced platelet function in men only (Figs. 2A, B, and 4A, Supporting Information Table S3), implying that it may have inhibited the primary wave of platelet aggregation and P-selectin expression induced by exogenous ADP. The effect of flavan-3-ol-enriched dark chocolate may not have been strong enough though to affect the se-cond wave of aggregation, involving the thromboxane A2 pathway, therefore failing to lead to a decrease in integrin αIIbβ3 activation (and thus fibrinogen binding) in men. On the other hand, the lower amount of platelet thromboxane A2 receptors on women's platelets [36], may explain the inhibitory effect of flavan-3-ol-enriched dark chocolate on TRAP-induced platelet aggregation (Fig. 2C and Supporting Information Table S3). However, women's platelets generally show more severe responses to induction by ADP or thrombin receptor stimulation by binding more fibrinogen and forming larger platelet aggregates [34, 36, 37]. Indeed, in women we also observed a simultaneous increase in TRAP-induced fibrinogen binding after dark chocolate consumption (Fig. 4B and Supporting Information Table S3). It has been shown before that the integrin αIIbβ3 receptor is more prone to activation and less responsive to inhibitors in women [34, 37]. We currently cannot explain this appa-rent discrepancy between inhibition of aggregation and promotion of fibrinogen binding induced by the same agonist. COL/EPI-induced PFA-100 closure times were significantly larger in women at baseline (Supporting Information Table S2), and effects of flavan-3-ol-enriched dark chocolate on these closure times were also more pronounced in women compared with men (Fig. 3 and Supporting Information Table S3). Prolongation of COL/EPI-induced PFA-100 closure times after consumption of flavan-3-ols within cocoa products or grape seed extracts has been observed before [7, 38, 39], albeit mostly 2 h after consumption.

The use of oral contraceptives and/or menstrual phase is known to affect outcomes of platelet function, and could thus have affected our results for females. Indeed, integrin αIIbβ3 receptors on platelets from premenopausal women are more prone to activation than those on male platelets [37], and platelet function is significantly correlated with progesterone levels [40]. In our study, 11 premenopausal women used contraception (Supporting Information Table S5), but only five women used combined oral contraceptives, leaving us with insufficient statistical power to assess whether efficacy in platelet function was affected by the use of oral contraceptives.

The gender-dependent changes in platelet function could not be explained by gender-dependent differences in absorption or metabolism of the cocoa flavan-3-ols. Generally, women showed a trend for higher concentrations of flavan-3-ols and their metabolites in plasma and urine (Fig. 5 and Supporting Information Table S4), however we believe this trend was introduced by correcting the catechin and dimer contents of urine spot samples for creatinine to standardize for diuresis. Albeit a common approach, urinary creatinine is dependent on several factors, one of them being gender [26-28]. Indeed, no differences between genders were observed when comparing noncorrected urinary catechins and dimers or total amounts of excreted urinary catechins and dimers (Supporting Information Fig. S3). In our study, dose-normalized mean concentrations of the epicatechin aglycone in plasma agreed well with values reported in the literature, as did the maximum plasma concentration reached in a single subject, which was 6.6 μmol/L after 2 h [17, 41]. The high level of excretion of flavan-3-ols but relatively low concentrations found in plasma indicates a rapid turnover of these compounds in plasma. Indeed, the majority of flavan-3-ols undergo rapid metabolism by several enzymes in intestinal cells forming sulfates, glucuronides, and/or methylated metabolites before being absorbed into the circulatory system, followed by excretion of such compounds and metabolites within urine [15, 16]. Therefore, urinary concentrations of flavan-3-ols may provide a more realistic estimate of absorption in humans [42]. A lack of availability of pure (epi)catechin and methyl-(epi)catechin metabolites as standards for quantitative analysis and incomplete hydrolysis of sulfates may have led to an underestimation of catechin concentrations in plasma and urine. Additionally, we could only compare intensities of metabolites in postprandial nonhydrolyzed urine samples (Supporting Information Fig. S4). However, the distribution of (epi)catechin monoglucuronides, (epi)catechin monosulfates, and methyl-(epi)catechin monosulfates was similar to previously published data [43]. Since this distribution was also very si-milar in men and women, it is unlikely to have contributed to the observed differences in platelet function between genders.

Levels of flavan-3-ols in urine and plasma were significantly associated with several measures of platelet function (r ≤ 0.50) (Table 4). Interestingly, concentrations of plasma catechins positively correlated with TRAP-induced fibrinogen binding (r between 0.22 and 0.50 and p ≤ 0.030) (Table 4). This mirrors the observed increase in TRAP-induced fibrinogen binding after consumption of flavan-3-ol-enriched dark chocolate, which was statistically significant after 2 and 6 h for the total study population (p = 0.014 and p = 0.021, respectively) (Fig. 4B and Supporting Information Table S3). On the contrary, changes in urinary dimers 2 and 6 h after chocolate ingestion negatively correlated with TRAP-induced platelet aggregation (r between ‒0.28 and ‒0.39 and p ≤ 0.003) (Table 4). Again, this predicts what had been observed as TRAP-induced platelet aggregation was significantly decreased 2 h after consumption of flavan-3-ol-enriched dark chocolate, compared with white chocolate, in women (p = 0.012) (Fig. 2C, Supporting Information Table S3).

In conclusion, flavan-3-ol-enriched dark chocolate, but also white chocolate, improved several measures of postprandial platelet function in a gender-specific way. For the dark chocolate, the beneficial anti-platelet effects could only be partially explained by individual differences in absorption or metabolism of flavan-3-ols. Quantitative measurements of metabolites may, once standards become available, improve insights into the relation between efficacy and bioavailabi-lity. Beneficial effects on platelet activation and aggregation mostly diminished 6 h after consumption, and therefore re-gular consumption of small quantities of flavan-3-ol-rich dark chocolates may help to sustain these effects for longer. Nevertheless, increasing consumption of such products will also increase fat and sugar consumption, which may outweigh any beneficial anti-platelet effects. Future studies may want to focus on mechanistic measures, such as assessment of calcium mobilization, ADP secretion, and phosphorylation of different platelet receptors, to gain more insights into the anti-platelet effects of flavan-3-ols.

Acknowledgments

The authors thank the volunteers for participating in this study, and Niamh O'Kennedy, Eva-Maria Bachmair, Iris E Allijn, and Ja Yee Winnie Lo for assistance with the sample processing and analysis. This study was funded by the Scottish Government's Rural and Environment Science and Analytical Services Division (RESAS) and the Biotechnology & Biological Sciences Research Council (BBSRC), United Kingdom. All chocolates were specifically produced for this study and provided free of charge by Natraceutical Group, Valencia, Spain.

Potential conflict of interest statement: Elena Cienfuegos-Jovellanos was employed by the Natraceutical Group, Valencia, Spain during the implementation of this study. All other authors have declared no conflict of interest.

Ancillary