Impact of Cocoa Products Intake on Plasma and Urine Metabolites: A Review of Targeted and Non-Targeted Studies in Humans

Cocoa is continuously drawing attention due to growing scientific evidence suggesting its effects on health. Flavanols and methylxanthines are some of the most important bioactive compounds present in cocoa. Other important bioactives, such as phenolic acids and lactones, are derived from microbial metabolism. The identification of the metabolites produced after cocoa intake is a first step to understand the overall effect on human health. In general, after cocoa intake, methylxanthines show high absorption and elimination efficiencies. Catechins are transformed mainly into sulfate and glucuronide conjugates. Metabolism of procyanidins is highly influenced by the polymerization degree, which hinders their absorption. The polymerization degree over three units leads to biotransformation by the colonic microbiota, resulting in valerolactones and phenolic acids, with higher excretion times. Long term intervention studies, as well as untargeted metabolomic approaches, are scarce. Contradictory results have been reported concerning matrix effects and health impact, and there are still scientific gaps that have to be addresed to understand the influence of cocoa intake on health. This review addresses different cocoa clinical studies, summarizes the different methodologies employed as well as the metabolites that have been identified in plasma and urine after cocoa intake.


Theobroma cacao L. composition
Cocoa (Theobroma cacao L.) is a tree from the Malvaceae family. Its seeds are covered by a sweet and sour mucilage which contains approximately 11% of sugars, mainly sucrose, and an acidic environment with a pH of about 3.5-3.8. Citric acid is the main organic acid present in the pulp. Others such as oxalic, phosphoric, malic and tartaric acid are also present [1][2][3][4][5][6].
Fat is the main component of cocoa beans: it accounts to almost 50% of the cotyledon dry weight, where 98% corresponds to neutral lipids, and 2% to polar lipids, mainly phospholipids and glycolipids. It has been reported that the major cocoa fatty acids are palmitic, stearic and oleic [7,8].
Proteins represent 17-20% of the dried bean [9,10]. Vicilin-like globulins and albumins are the two most abundant proteins and are key factors in quality development during fermentation. Concerning the amino acid profile, glutamic acid and aspartic acid showed the highest contents while cysteine showed the lowest ones, in both fermented and non-fermented cocoa derived products [11]. Student's and Wilcoxon tests. 1999 [62] Evaluate changes in plasma epicatechin levels, antioxidant capacity and plasma lipid oxidation products after chocolate intake.
HPLC coupled with a coulometric detector.
ANOVA with control for multiple measurements, and Tukey-Kramer test.
2000 [63] Evaluate (-)-epicatechin and its metabolites in plasma and urine after cocoa and chocolate intervention.
ANOVA and Tukey test. 2003 [65] Determine theobromine and caffeine in saliva, plasma and urine after dark chocolate intake. 2010 [59] Identify and quantify (-)-epicatechin conjugates in plasma and urine after chocolate intake. Uncontrolled.

healthy volunteers.
Average age of 23 years and average BMI of 22 kg/m 2 .
One-way ANOVA with Tukey test. 2012 [66] Analyze the relation between chocolate consumption, glucose metabolism, and exercise performance.
2014 [49] Evaluate the effects of intake of milk chocolate, cocoa powder or dark chocolate consumption on uric acid crystallization.
Cross-over. 11 males and 9 females with ages from 22 to 65 years.
2 × 20 g of a cocoa derived product/day (milk chocolate, chocolate powder and dark chocolate). Urine.
Baseline and urine obtained after 12 h overnight at the end of the intervention. HPLC-UV/Vis.
2018 [67] Notes: The abbreviation BMI corresponds to body mass index, and BW to body weight. 2007 [72] Analyze the effect of milk in the bioavailability of (-)-epicatechin from a cocoa powder.
Randomized and cross-over.   HPLC-DAD, HPLC-Q-ToF in positive ionization mode.
Baseline and urine obtained after 12 h overnight at the end of the intervention. HPLC-UV/Vis.
Randomized, double-masked, cross-over.  Evaluate the effect of cocoa intake on lipid profiles and biomarkers of oxidative stress, and arachidonic acid/ eicosapentaenoic acid ratio.
Randomized, three-arm parallel group.
Plasma and urine.
0, 1, 2, and 4 (at the beginning of the study and after 4 weeks of intervention).

HPLC-FLD-UV
ANOVA and Tukey's test. 2018 [58] Notes: The abbreviation BMI corresponds to body mass index, and BW to body weight.  [84,85] Evaluate the changes in humane urine metabolome after cocoa powder intake.
40 g of cocoa powder with 250 mL of milk. Urine.
HPLC-Q-ToF-MS in positive mode.

[87]
Apply an untargeted metabolomics approach to propose a model that can discriminate the urinary metabolome of regular cocoa product consumers and non-consumers, revealing dietary biomarkers, in a free-living population.
No dietary interventions or recommendations. Urine. Baseline.
Positive and negative ionization modes.

[55]
Notes: (1) Samples taken 2 h after intake were analyzed only by NMR, and not by LC-MS.
(2) At least with type 2 diabetes mellitus or with at least three conventional cardiovascular risk factors.
(3) Baseline data and urine samples were obtained from a PREDIMED study of 275 volunteers.

Amino Acid Metabolites Others Reference
Baseline N-methylguanine ---- [87] 0-6 3-methyluric acid 3-methylxanthine 3,7-dimethyluric acid 7-methyluric acid 7-methylxanthineAMMU (4) Caffeine Theobromine 3-methylxanthine 7-methylxanthine Theobromine --Xanthurenic acid - [87] Notes: (4) AMMU means 6-amino-5-[N-methylformylamino]-1-methyluracil. Table 8. Metabolites analyzed in urine and plasma after acute, short-term, or regular cocoa intake interventions, studied with non-targeted and targeted methodologies. In particular, studies that mentioned acute intake of cocoa are summarized in the following tables: (a) Tables 4 and 6 include information of different metabolites as concentrations measured in the biological samples and the corresponding sampling periods. The type of matrix analyzed is also indicated. Both tables summarize information of targeted studies, the first focused on plasma (Table 4) and the other in urine (Table 6) analysis; (b) Table 5 shows the time after intake of different chocolates, where the highest signals of the different metabolites were detected or expected in plasma samples, (c) Table 7 focuses on the main metabolites found in urine that discriminated between different periods of time after intake, with an untargeted approach.
On the other hand, discriminating metabolites in between regular/chronic consumers and non-regular/chronic consumers observed in urine and plasma are listed on Table 8. This covers both targeted and untargeted studies.
Metabolites were organized by means of the chemical group, mainly methylxanthines, flavonoids and conjugates, phenolic acids, and lignans. As for the data collected from untargeted studies, data organization was done in a way that respected as much as possible the authors' original classification. Additionally, in each category metabolites were organized alphabetically.

Methylxanthines
Theobromine showed some of the higher concentrations in plasma and urine after cocoa intake in different human intervention trials (Tables 4 and 5). One study [81] showed that unmetabolized theobromine in urine was close to 1.5 times higher when consuming methylxanthine enriched powder compared to the non-enriched cocoa powder, which indicates that renal clearance could vary in function of exposure. Due to a possible saturation of the pathway, a lack of enzymes and/or cofactors to process the metabolite may have had affected [81,88].
There is evidence that suggests that excretion also depends on gender and age. After consumption of chocolate in a cross-over study, theobromine excretion showed by women almost doubled man's levels and was also higher for volunteers younger than 29 years, and lower compared to the ones older than 54 years [84].

Polyphenols
Polyphenols such as (epi) catechin, and the corresponding glucuronides, sulfates, with or without methylations were the most frequently analyzed metabolites.
One example is a study that evaluated plasma and urine samples after drinking a cocoa beverage.
In general terms these metabolites reached maximum concentrations in plasma in less than 4 hours after intake (Table 4). One exception was (-)-epicatechin-7'-β-D-glucuronide, that in one study required 12.8 ± 4.8 h after cocoa intake to reach maximal concentrations [66].
When these metabolites are present in plasma, they could exert particular effects. For example, one study analyzed the impact of chocolate intake on oxidative damage, and observed an increase in plasma antioxidant activity and plasma lipid oxidation products decrease as epicatechin plasma levels increased [63]. Additionally, elimination rates have an influence on the probabilities of accumulation in blood or tissues, and the subsequent effect on health [93].
In this matter, when studying urine samples, the required time to reach maximal concentration was found to be largely variable. While (epi) catechin-O-sulfate isomer required a time in between 0-2 h after cocoa powder in milk intake to reach the higher concentration [77], (-)-epicatechin-7-β-D-glucuronide required a time in between 10-24 h after dark chocolate consumption [66].
Influence of age on metabolism was analyzed in an intervention that evaluated two cocoa powder drinks. Non-significant differences or small differences were observed for the different pharmacokinetic parameters, in the renal clearance, total excreted metabolites, and apparent volume of distribution in two age groups: 18-35 years and 65-80 years. Some of the small differences observed were associated with changes in renal function due to normal aging processes [82].
Studies converge in that increasing polymerization degree (n > 2) of procyanidins hinders their absorption. Procyanidin B2 dimer has been detected in plasma after consuming a cocoa beverage, showing a maximum concentration in the period between 0.5-2 h ( Table 5). In this same study, B5 dimer was not detected [68]. Similar results were observed in another study where volunteers drank samples with cocoa extracts that contained procyanidins from 2-10 units. In this study, B2 dimer was detected in plasma in approximately 80% of the volunteers, indicating maximum concentrations at 2 h after intake, and with concentrations close to the limit of detection. Additionally, B5 dimer was not detected in any of the plasma samples [80].
Polyphenol absorption and excretion can depend on the food matrix, but contradictory information exists. For example, excretion patterns of microbial phenolic acids metabolites measured in urine differed depending on the liquid vehicle of cocoa powder consumed: water or milk [78]. One randomized study compared the levels of (-)-epicatechin glucuronide in plasma after the intake of cocoa powder with milk or water, and reported no statistical differences in the results from each group [73]. Also, milk protein did not affect the bioavailability of polyphenols present in a chocolate drink [74]. In another study the excretion of different flavan-3-ol metabolites within 24 h after cocoa intake was close to 20% of the ingested dose if cocoa was provided as a water-based beverage, but only 10% if ingested as a milk-based cocoa drink [77]. When consuming two cocoa beverages, either prepared with water or with milk, the total excretion was not affected, but kinetics was different. In the same period of time after intake (-)-epicatechin glucuronide was the most concentrated metabolite when water was the vehicle-doubling the concentration of total excreted (-)-epicatechin sulfatesbut when milk was used, glucuronides and total sulfates reached similar concentrations in urine [75]. No significant differences in excretion patterns were found when comparing a variety of metabolites in urine after consuming cocoa powder in milk or in water [86]. It is suggested that polyphenol absorption and metabolism might be affected because of the interactions between milk polypeptides and polyphenols, which can decrease the bioaccessibility [94,95]. In addition, polyphenol absorption could also be affected by carbohydrates, as it has been reported that flavanol uptake could increase when simultaneously consuming these macronutrients [50].
Flavanol absorption and metabolism has also been demonstrated to be dependent on the stereochemistry. Differences in absorption and metabolism were detected after the consumption of equal amounts of (-)-epicatechin, (-)-catechin, (+)-catechin, and (+)-epicatechin in a cocoa based milk drink [79].
Non-absorbed flavonoids and phenolic acids could follow metabolic pathways that are carried out by the colonic microbiota [66,[90][91][92]. Some epicatechin metabolites can return to small intestine including non-absorbed epicatechin pass to the colon, which can account a total close to 70 % of the total ingested [96].
In general terms, these types of metabolites require longer times to reach mean concentrations in urine and plasma than the metabolites following phase II metabolism (Tables 4 and 6, Tables 7 and 8).
To illustrate, ferulic and m-hydroxyphenyl acetic acids showed maximum concentrations in between 24 and 48 h after chocolate intake, and vanillic acid was the only metabolite that returned to basal level concentrations in the 48 h of observation [65].
Research suggests that consumption of cocoa and derivatives could modulate human metabolism. A randomized controlled cross-over intervention trial analyzed the acute consumption of three types of chocolates in 42 healthy volunteers and concluded that several endogenous metabolites were affected (Tables 5 and 7). In particular, several amino acids, organic acids, creatinine, lactate, and N 1 -methylnicotinamide decreased meanwhile pyruvate, tyrosine, and p-hydroxyphenylacetate increased after the intervention. Flavan-3-ols, methylxanthines and their metabolites were suggested to modulate endogenous and colonic microbial metabolism [84].
Studies with longer periods of intervention detected additional metabolome changes due to regular cocoa consumption (Table 8).
There is also evidence that polyphenol consumption from cocoa products might change the gut microbiota, exerting prebiotic effects, and which could be related to the activation of anti-inflammatory pathways with benefits in the host and alter the obtained profile of metabolites. One study observed an increase of Lactobacillus spp. and Bifidobacterium spp. in contrast with a decrease of Clostridium spp. populations after daily consumption during one month of a flavanol-rich cocoa beverage, compared to a low flavanol cocoa beverage [33,97,98].

Other Metabolites
Metabolites different than methylxanthines and polyphenols metabolism have also been detected after cocoa intake. For example, trigonelline (N-methylnicotinic acid) and hydroxynicotinic acid may be a result of nicotinic acid metabolism. They were detected in urine only within 6 h after cocoa powder consumption. Also, diketopiperazines such as cyclo(ser-tyr) and cyclo(pro-pro) were reported for the first time associated to cocoa powder intake and are considered flavor and taste metabolites [86].
N-phenylpropenoyl-L-amino acids have also been metabolites of interest as there is some preliminary evidence of their biological activity. Thirteen of them have been reported in urine after cocoa intake [20].
One study observed that daily dark chocolate intake for 2 weeks modulated energy and hormone metabolism of humans with low and high anxiety traits, and also modulated their microbial gut metabolism. Urine analysis showed a reduction of stress hormone metabolites, and other stress-related metabolites presented a trend to change towards the low anxiety profile. Some of these metabolites are glycine, citrate, trans-aconitate, proline, DOPA, β-alanine, hippurate, and p-cresol sulfate [99]. This can eventually help understand the impact of cocoa intake on health.

Conclusions
A diversity of metabolites has been identified in urine and plasma after consumption of cocoa or cocoa derived products. Theobromine and caffeine have been some of the main methylxanthines identified together with their metabolites. Polyphenolic phase I and phase II metabolism leads mainly to (epi)catechin sulphates, glucuronides, and sulfoglucuronides. Non absorbed polyphenols are transformed by colonic microorganisms yielding a diversity of metabolites that range from phenolic acids to different valerolactones, and more complex metabolites.
Evidence shows that absorption, metabolism, and excretion of cocoa metabolites depend on the food matrix, the dose, age, gender, overall health status and other factors such as the polymerization degree (e.g., procyanidins), and stereochemistry (e.g., flavanols).
The development of clinical studies is fundamental to understand the metabolic pathways of different metabolites present in cocoa and further on the effects on health. Special attention must be given to properly design the experiments, and instrumental methodologies for the extraction and quantification of metabolites in biological samples.