Exploring Human Metabolome after Wine Intake—A Review

Wine has a rich history dating back to 2200 BC, originally recognized for its medicinal properties. Today, with the aid of advanced technologies like metabolomics and sophisticated analytical techniques, we have gained remarkable insights into the molecular-level changes induced by wine consumption in the human organism. This review embarks on a comprehensive exploration of the alterations in human metabolome associated with wine consumption. A great number of 51 studies from the last 25 years were reviewed; these studies systematically investigated shifts in metabolic profiles within blood, urine, and feces samples, encompassing both short-term and long-term studies of the consumption of wine and wine derivatives. Significant metabolic alterations were observed in a wide variety of metabolites belonging to different compound classes, such as phenolic compounds, lipids, organic acids, and amino acids, among others. Within these classes, both endogenous metabolites as well as diet-related metabolites that exhibited up-regulation or down-regulation following wine consumption were included. The up-regulation of short-chain fatty acids and the down-regulation of sphingomyelins after wine intake, as well as the up-regulation of gut microbial fermentation metabolites like vanillic and syringic acid are some of the most important findings reported in the reviewed literature. Our results confirm the intact passage of certain wine compounds, such as tartaric acid and other wine acids, to the human organism. In an era where the health effects of wine consumption are of growing interest, this review offers a holistic perspective on the metabolic underpinnings of this centuries-old tradition.


Introduction
According to the International Organization of Vine and Wine (OIV), global wine consumption has been steadily increasing over the years.In 2019, the estimated global wine consumption reached 244 million hectoliters, as shown by country in Figure 1, indicating its continued popularity worldwide [1].
Wine is produced and enjoyed in numerous countries around the world, including traditional wine-producing regions like France, Italy, Spain, and new-world producers such as the United States, Argentina, Australia, and Chile, as depicted in Figure 2. The widespread cultivation and consumption of wine illustrate its global popularity [2].
Nevertheless, wine's popularity and significance extend far beyond the present day.Its therapeutic use can be traced back to 2200 BC, making it the oldest known medicine [3].In the early 1990s, the French Paradox sparked an intense scientific investigation into the cause of lower death rates in France than in the USA, in spite of comparable plasma cholesterol levels and other similar risk factors.This inquiry revealed a noteworthy finding: moderate regular wine intake (1-2 glasses per day) has been linked to lower cardiovascular mortality and risk of heart disease [4].These advantages have been related to improved antioxidant capacity, lipid profile alterations, and anti-inflammatory actions.Following the initial link between the French Paradox and red wine consumption, groundbreaking research emerged [5][6][7].A comprehensive review study, with the aim to map the scientific research on wine and health, was conducted in 2013 by Aleixandre et al. and they came up with more than 400 papers published on this topic in the most productive journals between 2002-2011 [8].However, the pursuit of knowledge in this field continues unabated.In 2011, Contstanzo et al. conducted a comprehensive meta-analysis regarding fatal and non-fatal cardiovascular events, wherein they identified a J-shaped relationship between eleva ted consumption of wine and beer and vascular risk [9].Moreover, in the year 2023, Lucerón-Lucas-Torres et al. complete a meta-analysis in the association between wine consumption with cardiovascular disease and cardiovascular mortality.Their findings unveiled an inverse correlation between wine consumption and cardiovascular mortality, cardiovascular disease (CVD), and coronary heart disease (CHD) [10].Similar results were presented also by Lombardo et al. (2023) were the beneficial effect of wine consumption on antioxidant levels, markers related to thrombosis and inflammation, lipid profile, and improvements in the composition of gut microbiota, was reported [11].It is noteworthy to mention that, to the best of our knowledge, a review study regarding the metabolic alterations caused in the human organism due to wine consumption has not been published before.
Molecules 2023, 28, x FOR PEER REVIEW 2 of 48 Figure 1.Per capita wine consumption in 2019, quantified in liters of absolute alcohol per annum (approximately 0.12 L of pure alcohol are found in 1 L of wine) [1]. Figure 1.Per capita wine consumption in 2019, quantified in liters of absolute alcohol per annum (approximately 0.12 L of pure alcohol are found in 1 L of wine) [1].
To comprehensively explore the health benefits associated with wine consumption, it is crucial to first understand its complex chemical composition.Wine is a natural beverage derived from the direct fermentation of grapes and it consists of compounds originating from grape berries, yeast, bacteria and oak, all of which contribute to the unique chemical profile and health advantages of wine.Wine contains numerous components such as water, ethanol, organic acids glycerol, sugars, certain amino acids, volatile compounds, and polyphenolic compounds, among others [12,13].Hence, it is of utmost importance to comprehend the mechanisms by which these various classes of wine compounds can interact with the human body.The interpretation of wine compounds metabolism and metabolic pathways in the human body after wine consumption has been in the spotlight of research in the last years, with the application of "-omics" technologies playing an invaluable role in this upturn [12].Nevertheless, wine's popularity and significance extend far beyond the present day.Its therapeutic use can be traced back to 2200 BC, making it the oldest known medicine [3].In the early 1990s, the French Paradox sparked an intense scientific investigation into the cause of lower death rates in France than in the USA, in spite of comparable plasma cholesterol levels and other similar risk factors.This inquiry revealed a noteworthy finding: moderate regular wine intake (1-2 glasses per day) has been linked to lower cardiovascular mortality and risk of heart disease [4].These advantages have been related to improved antioxidant capacity, lipid profile alterations, and anti-inflammatory actions.Following the initial link between the French Paradox and red wine consumption, groundbreaking research emerged [5,6,7].A comprehensive review study, with the aim to map the scientific research on wine and health, was conducted in 2013 by Aleixandre et al. and they came up with more than 400 papers published on this topic in the most productive journals between 2002-2011 [8].However, the pursuit of knowledge in this field continues unabated.In 2011, Contstanzo et al. conducted a comprehensive meta- "Omics" refers to scientific branches that encompass the measurement and analysis of different disciplines, such as metabolomics (metabolites) including lipidomics (lipids), proteomics (proteins), and genomics (genes).The term "Omics" originates from the Latin suffix "ome", meaning many [14].In this current review, we chose to focus on metabolomics due to its increasing use in recent years for detecting food-based metabolites in human biological samples [15].
Metabolomics, which is the study of small molecules or metabolites, particularly those with a molecular weight below 1500 Da in biological samples [16,17], stands out as an advantageous technique compared to other "omics" approaches.Metabolites, including among other carbohydrates, amino acids, lipids, energy metabolites, vitamins, co-factors, nucleotides, and xenobiotics, could be influenced from the diet intake, environmental exposure, microbial metabolism, and pharmacotherapy [18].Analyzing how metabolites respond to stimuli provides a snapshot of an organism's metabolism, akin to a molecular fingerprint, effectively reflecting the organism's overall biological condition [19].This approach is being increasingly utilized in various areas of study, including the analysis of food components, the identification of diet-derived metabolites, assessment of their bioavailability and metabolism, examination of metabolic activities in the gut microbiota, and understanding the physiological response to specific dietary patterns, foods, or nutraceuticals [20].
This comprehensive review critically examines a wide range of metabolomic studies that have focused on investigating the changes in metabolic profiles following both shortterm and long-term wine consumption.All these 51 studies included in our review present the investigation of human metabolome of both healthy and unhealthy participants that consumed either wine, dealcoholized wine, or wine extract in capsule form.Biological samples were subjected to target and/or nontarget analysis using liquid chromatography and gas chromatography coupled with mass spectrometry, or H-NMR spectroscopy techniques.By meticulously analyzing and synthesizing the findings from these studies, we aim to provide a comprehensive and detailed understanding of the metabolic alterations that occur as a result of wine consumption.These alterations are observable across a range of compound classes, spanning from endogenous metabolic substances to wine compounds.For better understanding, we have compiled tables listing all the distinct compound classes that demonstrated significant differences compared to the samples taken prior to wine consumption or compared to control group.These tables could present an invaluable tool for the development of ready-to-use databases enabling the non-target/suspect screening of wine-related metabolites in biological samples.

Clinical Studies
Several clinical investigations have been conducted to assess the influence of wine consumption on the human metabolome.These studies can be classified into two main categories: interventional studies and observational studies [21].Findings from these investigations have been consolidated and are presented comprehensively in Tables 1 and 2. Furthermore, interventional studies were categorized into: parallel, crossover and randomized studies.Understanding the distinction between these types is essential in interpreting the findings and implications of the research.In parallel studies, different groups of subjects are assigned to distinct interventions, allowing for between-group comparisons [22][23][24][25][26][27][28][29][30][31].On the other hand, crossover studies enable each subject to serve as their own control by undergoing both interventions in a sequential manner [32,33].And lastly randomized controlled trials are randomly allocated to either the treatment or control arms [22,23,34].In Supplementary Table S1, a comprehensive compilation of information regarding interventional and observational clinical studies in the field of wine metabolomics research is presented.The significant majority (more than 70%) of the clinical studies concerning wine consumption were interventional studies.In these studies, before the intervention, the participants underwent a washout period during which they refrained from consuming polyphenol-rich foods, as well as wine and other alcoholic beverages (see Table 1).Following the washout period, the subjects entered an intervention period that lasted from a day or a specific number of 5-20 days or 3-8 weeks.During the intervention period, the participants were given explicit instructions to consume a predetermined volume of a particular wine daily.Alternatively, in some studies, participants were administered dealcoholized wine [31,32,35,44,45,47,51,54] or wine derivatives [36,[49][50][51][52]55,56].It is important to note that throughout the interventional studies, participants were instructed to maintain their regular diet and abstain from consuming other alcoholic beverages.Notably, among the selected studies, only the investigation conducted by Vitaglione et al. (2005) specifically explored the potential relationship between the type of meal consumed and the variety of red wine consumed by investigating resveratrol metabolites bioavailability [30].

Duration of the Study
The duration of interventional studies ranged from one day to 8 weeks.Specifically, as it can be seen in Table 1, fifteen studies had a duration of only one day, while three studies extended their interventions over 5 to 20 days.The majority of the studies encompassed a timeframe of 3 to 4 weeks.Notably, one study stood out with a substantial duration of 8 weeks [71].These variations in study duration highlight the diverse timeframes employed to examine the intended interventions.

Participants
The number of participants in a clinical study is crucial for ensuring statistical power and drawing meaningful conclusions.A large participant pool enhances the study's ability to detect effects or differences accurately.However, finding a diverse group of participants that meet the study's criteria is a time-consuming and resource-intensive task.As a result, most interventional clinical studies concerning wine consumption had a range of 5 to 61 participants (Table 1).Notably, more than half of these studies deliberately included both male and female participants, thereby achieving a balanced gender representation.In contrast, 14 out of 37 studies specifically recruited only male subjects.This decision was made to minimize potential confounding factors related to menstrual cycle phase variability in females, which could impact processes such as wine absorption, metabolism, and excretion [41].For instance, in the study conducted by Haas et al. (2022), the exclusive inclusion of male participants aimed to achieve sample homogeneity, considering potential differences in alcohol metabolism and trimethylamine-N-oxide (TMAO) metabolism between sexes [34].
Besides sex, also the health of participants plays a pivotal role in clinical studies due to its multifaceted significance.The majority of interventional studies in this review centered on healthy participants, affording valuable insights into the impact of wine consumption on individuals without pre-existing health conditions.Additionally, 5 studies focused on participants with type 2 diabetes (T2D) or those exhibiting at least three coronary heart disease (CHD) risk factors, or cardiovascular risk factors (CVRFs) shedding light on the potential benefits or risks of wine consumption for individuals with these conditions [44,45,47,48,54].It is worth noting that 3 studies within this review did not specify the health status of their participants [33,56,57], while an additional 3 studies specifically focused on participants with mild hypertension [36,49,50].Additionally, one study explored the impact of wine consumption on individuals diagnosed with coronary artery disease (CAD) [34].

Type and Amount of Wine/Wine Derivatives
The choice of wine variety consumed significantly influences the resulting impact on human metabolome [61].Within the interventional studies included in this review, it was customary for participants to consume a specific type of red, white, sparkling or dealcoholized wine as part of the study protocol.A small number of studies employed the consumption of wine extracts either as additives in wine or as individual supplements.The various types of interventions are presented in Table 1.
The interventional studies encompassed a range of red wine varieties, including Merlot, Pinot Noir, Cabernet Sauvignon, Cabernet Franc, Lemberger, Shiraz, as well as Spanish variety such as Tempranillo, and Italian varieties such as Lambrusco and Aglianico.These specific red grape varieties were purposefully selected based on their notable phenolic content [37,48] with some studies reporting also the phenolic profile of the utilized wines [46][47][48]54].In addition to regular red wines, a significant number of studies examined dealcoholized red wines.Moreover, although the majority of studies focused on regular red wine, three published papers explored white wine and only one study investigated sparkling wine [58].
The typical daily wine intake in the studies ranged from 200 to 272 mL.The selection of a 250 mL dosage was based on evidence indicating that this quantity of red wine is associated with beneficial effects [34].However, there were three studies that provided a lower amount of 120 mL [32,33,39], while two one-day interventional studies deviated from the norm by administering a higher volume of 500 and 600 mL of wine [30,43].

Observational Studies 2.2.1. Food Frequency Questionnaires (FFQs)
Fourteen metabolomic studies investigating the human metabolome after wine intake were observational studies.In the field of observational studies, traditional approaches to assessing regular eating habits have relied on self-reported tools to evaluate diets, such as food frequency questionnaires (FFQs).Consequently, many studies investigating wine consumption use these self-reported measures as a means of data collection.However, it is important to acknowledge that self-reported measures are inherently prone to both random and systematic measurement errors, which may compromise the accuracy and reliability of the findings [59].Recognizing these limitations, Regal et al. (2017) acknowledge the challenges associated with accurately recalling food/beverage consumption, especially concerning wine intake, due to social preconceptions surrounding alcoholic beverages.To address these challenges, they aimed to investigate the potential of resveratrol as a reliable biomarker for objectively measuring wine consumption within a specific population.The study concluded that resveratrol can serve as a valuable dietary biomarker only when used in conjunction to dietary tools since this stilbene is naturally present also in several plant species [61].

Participants
Observational studies, which rely on passive observation rather than active interventions, usually include a broader range of subjects.In the studies presented in this review the participant numbers range from 222 to 3559 individuals [63,65] with the exception of a single study featuring only 25 participants [61].More than half of the studies ensured representation from both sexes by incorporating both male and female subjects as participants.However, two specific clinical studies deviated from this approach and focused exclusively on female participants.One study examined female twins [63] while the other focused on women experiencing menopause [72].
Apart from gender, health status of the participants is of utmost importance.Among the observational studies reviewed seven chose to recruit participants from the general population [59,63,64,[66][67][68]70], while 3 studies specifically focused on healthy participants.Additionally, 3 studies focused on participants with T2D, CHD risk factors and CVRFs, shedding light on the potential benefits or risks of wine consumption for individuals with these conditions [12,60,71].Lastly, Playdon et al. ( 2016) included participants diagnosed with new or recurrent cases of colorectal adenoma [69].

Biological Samples, Analytical Techniques and Statistical Analysis
The process of sample preparation and analysis plays a pivotal role as it directly influences the quality and reliability of the study results.Tables 1 and 2 provide a comprehensive overview of the various protocols of sample preparation, analytical techniques and statistical analysis utilized in metabolomic studies related to wine consumption.Statistical analysis emerged as a prevalent approach in the majority of these studies, facilitating the elucidation of significant changes and trends in the metabolome.

Biological Samples Preparation
In metabolomic research, the choice of biological samples is pivotal to unraveling the intricate web of metabolic processes.Figure 3 illustrates the distribution of commonly collected biological samples, highlighting the prominence of urine, blood, and feces in metabolomic investigations.
Molecules 2023, 28, x FOR PEER REVIEW 14 of 48 CVRFs, shedding light on the potential benefits or risks of wine consumption for individuals with these conditions [12,60,71].Lastly, Playdon et al. (2016) included participants diagnosed with new or recurrent cases of colorectal adenoma [69].

Biological Samples, Analytical Techniques and Statistical Analysis
The process of sample preparation and analysis plays a pivotal role as it directly influences the quality and reliability of the study results.Tables 1 and 2 provide a comprehensive overview of the various protocols of sample preparation, analytical techniques and statistical analysis utilized in metabolomic studies related to wine consumption.Statistical analysis emerged as a prevalent approach in the majority of these studies, facilitating the elucidation of significant changes and trends in the metabolome.

Biological Samples Preparation
In metabolomic research, the choice of biological samples is pivotal to unraveling the intricate web of metabolic processes.Figure 3 illustrates the distribution of commonly collected biological samples, highlighting the prominence of urine, blood, and feces in metabolomic investigations.

Urine
Urine, one of the extensively examined biofluids in metabolomic research, holds great significance due to its ease of collection and its lower complexity.Besides its convenience, urine plays a crucial role as an excretory pathway for water-soluble metabolites and xenobiotics in the body.Consequently, the analysis of the urinary metabolome holds significant potential in providing valuable insights into diseases and dietary intake [16,73].As a result, as you can see in Tables 1 and 2, almost half of the metabolomic studies focused exclusively on investigating the urine metabolome after wine intake while an additional 10 studies explored both urine and blood samples for a more comprehensive study.
The two most popular sample preparation techniques for urine prior to chromatographic analysis were liquid-liquid extraction (LLE) with ethyl acetate or acetonitrile (ACN) and solid phase extraction (SPE).Additionally, in some studies, researchers performed enzymatic hydrolysis before extraction with the purpose of exclusively examining polyphenol aglycones [67,68].Lastly, some research groups preferred to carry out minimum sample preparation before analysis and chose the dilute-and-shoot [41,42,65,70] or direct injection method [24,25,69].In addition, a subset of studies that conducted gas chromatography (GC) analysis incorporated an additional step of derivatization.Among these studies, the preferred derivatization reagent was bis(trimethylsilyl)trifluoroacetamide

Urine
Urine, one of the extensively examined biofluids in metabolomic research, holds great significance due to its ease of collection and its lower complexity.Besides its convenience, urine plays a crucial role as an excretory pathway for water-soluble metabolites and xenobiotics in the body.Consequently, the analysis of the urinary metabolome holds significant potential in providing valuable insights into diseases and dietary intake [16,73].As a result, as you can see in Tables 1 and 2, almost half of the metabolomic studies focused exclusively on investigating the urine metabolome after wine intake while an additional 10 studies explored both urine and blood samples for a more comprehensive study.
The two most popular sample preparation techniques for urine prior to chromatographic analysis were liquid-liquid extraction (LLE) with ethyl acetate or acetonitrile (ACN) and solid phase extraction (SPE).Additionally, in some studies, researchers performed enzymatic hydrolysis before extraction with the purpose of exclusively examining polyphenol aglycones [67,68].Lastly, some research groups preferred to carry out minimum sample preparation before analysis and chose the dilute-and-shoot [41,42,65,70] or direct injection method [24,25,69].In addition, a subset of studies that conducted gas chromatography (GC) analysis incorporated an additional step of derivatization.Among these studies, the preferred derivatization reagent was bis(trimethylsilyl)trifluoroacetamide (BSTFA), while others utilized mixtures containing BSTFA or N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA), in combination with trimethylchlorosilane (TMCS).
Regarding sample preparation for proton nuclear magnetic resonance analysis (H-NMR), urine exhibits a notable pH range, typically ranging from 5 to 8, and this variance significantly influences the observed chemical shifts in NMR spectra.To effectively address this challenge, a phosphate buffer with a pH of 7.4 is routinely incorporated into the sample preparation process [74,75].This buffer ensures pH stability, minimizing variations across different urine samples and enhancing the overall consistency and reliability of the analysis.Additionally, sodium azide is added to maintain sample integrity by effectively controlling bacterial growth.
A significant obstacle encountered in the analysis of metabolites in intact serum/plasma is the substantial interference caused by the presence of a large quantity of serum/plasma proteins (6-8 g/dL) [75].Urpi-Sarda et al. (2005) [37] and Gonzalez et al. (2020) [40] conducted comprehensive studies investigating optimal sample preparation techniques for plasma according to the extraction recovery of resveratrol and other food metabolites, following moderate wine consumption [40].Both studies independently found that protein precipitation (PPT) was the most effective sample preparation when compared to liquidliquid extraction (LLE) and solid phase extraction (SPE) methods, which are the three most commonly used blood sample preparation techniques.Consistent with the approach taken for urine sample preparation, some researchers also apply enzymatic hydrolysis prior to implementing any extraction technique in plasma samples [29,33,38,39].
In studies employing gas chromatography (GC) for sample analysis, an additional derivatization step was implemented.The preferred derivatization reagent utilized by these studies was BSTFA.Furthermore, alternative derivatization mixtures comprising BSTFA in conjunction with TMCS [56], o-methylhydroxylamine, or pure MSTFA were employed by other researchers [52,55].In the current body of research, blood samples have not been employed for H-NMR analysis.

Feces
Feces are a valuable addition for assessing the combined pool of endogenous and microbial metabolites, often referred to as the hyperbolome [76].
In relation to the preparation of fecal samples, the methodologies implemented in the literature predominantly revolved around the dissolution of fecal matter to form a fecal solution.Subsequently, the fecal solution was subjected to extraction using acetonitrile (ACN) [23,26,28,35].Notably, only one study deviated from this approach and employed solid-phase microextraction (SPME) as an alternative technique for sample preparation [22].When it comes to the preparation of fecal water samples for H-NMR analysis, Jacobs et al. (2008) was the sole study that employed this analytical technique [36].In their approach, the fecal water samples were first alkalized using sodium hydroxide (NaOH), followed by acidification with formic acid.Furthermore, TSP was used as an internal standard for metabolite quantification.

Analytical Techniques
It is evident that numerous analytical methods can be applied to investigate the human metabolome following wine consumption, thus raising the question of which particular method from the 'metabolomics toolbox' should be selected.As illustrated in Figure 4, LC-MS, GC-MS, and H-NMR were frequently chosen.Additionally, many studies adopted a multi-faceted approach, combining multiple techniques within the same investigation.
standard for metabolite quantification.

Analytical Techniques
It is evident that numerous analytical methods can be applied to investigate the human metabolome following wine consumption, thus raising the question of which particular method from the metabolomics toolbox' should be selected.As illustrated in Figure 4, LC-MS, GC-MS, and H-NMR were frequently chosen.Additionally, many studies adopted a multi-faceted approach, combining multiple techniques within the same investigation.

LC-MS
Liquid chromatography (LC) is a widely employed separation technique, particularly for complex samples such as biofluids.It offers advantages like simplified sample preparation and shorter run times compared to gas chromatography (GC).The versatility of LC stems from the availability of various stationary phases and the ability to use different mobile phases, allowing for the analysis of a broad range of compounds, from moderately to highly polar substances, as well as low volatility and thermolabile compounds [17].Moreover, the coupling of LC with Mass Spectrometry (MS) offers unparalleled detection and identification capabilities, which allows not only the determination of wellknown metabolites with remarkable sensitivity and specificity, but also the untargeted study of unknown wine metabolites and related compounds [77,78].
Reverse phase liquid chromatography (RPLC) is most commonly used in metabolomics to separate polar metabolites and non-polar lipids.However, it may not retain highly polar or ionic compounds effectively.In such cases, hydrophilic interaction liquid chromatography (HILIC) complements RPLC by retaining polar or ionic metabolites, including compounds related to key metabolic pathways [16,79,80].The choice of a suitable stationary phase can be tailored to the specific compounds of interest, especially in targeted metabolomics.For untargeted metabolomics applications, using multiple column types is essential [81].In this review, most LC-based metabolomic analyses used RPLC, mainly C18 columns, with two studies using C8 columns [25,26].One study employed HILIC columns for overall analysis, while another used them specifically for measuring trimethylamine N-oxide (TMAO) [34,60].

LC-MS
Liquid chromatography (LC) is a widely employed separation technique, particularly for complex samples such as biofluids.It offers advantages like simplified sample preparation and shorter run times compared to gas chromatography (GC).The versatility of LC stems from the availability of various stationary phases and the ability to use different mobile phases, allowing for the analysis of a broad range of compounds, from moderately to highly polar substances, as well as low volatility and thermolabile compounds [17].Moreover, the coupling of LC with Mass Spectrometry (MS) offers unparalleled detection and identification capabilities, which allows not only the determination of well-known metabolites with remarkable sensitivity and specificity, but also the untargeted study of unknown wine metabolites and related compounds [77,78].
Reverse phase liquid chromatography (RPLC) is most commonly used in metabolomics to separate polar metabolites and non-polar lipids.However, it may not retain highly polar or ionic compounds effectively.In such cases, hydrophilic interaction liquid chromatography (HILIC) complements RPLC by retaining polar or ionic metabolites, including compounds related to key metabolic pathways [16,79,80].The choice of a suitable stationary phase can be tailored to the specific compounds of interest, especially in targeted metabolomics.For untargeted metabolomics applications, using multiple column types is essential [81].In this review, most LC-based metabolomic analyses used RPLC, mainly C18 columns, with two studies using C8 columns [25,26].One study employed HILIC columns for overall analysis, while another used them specifically for measuring trimethylamine N-oxide (TMAO) [34,60].
Mass spectrometry has been the primary platform for metabolomics applications because of its excellent sensitivity, selectivity, and remarkable compound annotation capabilities [16,17,82].Therefore, it is not surprising that the vast majority of the wine intake metabolomic studies used LC-MS analysis for the determination of both targeted and untargeted metabolites.
To perform untargeted metabolomics using mass spectrometry, different MS acquisition techniques have been proposed: full scan, Data Independent acquisition (DIA), or Data Dependent acquisition (DDA) [87].Both DDA and DIA have been used to address different questions: DDA relies on intensity-dependent precursor ion selection but potentially missing low-abundance metabolites in MS 2 analysis, while DIA theoretically captures MS 2 spectra for all precursor ions, enhancing low-concentration metabolite detection, but interpreting complex spectra poses informatics challenges [34,86].In a characteristic untargeted metabolomic study, Haas et al. (2022) used DDA to investigate the effects of red wine consumption on the plasma metabolome [34].

GC-MS
Gas chromatography-mass spectrometry (GC-MS) serves as a widely employed platform in metabolomics.However, one significant practical limitation is that many classes of metabolites, such as phenolic compounds, sugars, nucleosides, amino acids, among others, cannot be directly analyzed due to their polarity and lack of volatility, requiring the precedence of a derivatization step [73] This derivatization process is time-consuming, has limited throughput and carries a risk of errors, thus introducing variability and the possibility of artifacts [88].Additionally, GC-MS has a restricted mass range, and the molecular ion is frequently not detected due to fragmentation, which poses challenges for the identification of unknown compounds [79].Nonetheless, GC-MS, particularly GCGC-MS, offers notable benefits, including exceptional resolution, high sensitivity, and robustness [88].
In metabolomic studies after wine intake it is evident that gas chromatography (GC) emerges as the second most frequently employed analytical technique.Out of the 51 studies reviewed, a total of 13 studies opted for GC as the method of choice to identify the human metabolome following both long-term and short-term wine consumption (Tables 1 and 2).The majority of these studies used columns constituted of 5% phenyldimethylpolysiloxane in conjunction with polyethylene glycol (PEG), while some utilized columns consisting of 50% phenyl and 50% dimethylpolysiloxane. Additionally, one study employed a column based on polyvinylpolypyrrolidone (PVPP).Researchers primarily focused on the analysis of phenolics, lipids, and organic acids using GC-MS.Phenolic compounds, including 4-hydroxymandelic acid [49,56], pyrogallol [49,55], vanillic and its derivatives, as well as numerous flavonoids as elucidated by Donovan et al. (1999Donovan et al. ( , 2002) ) have been extensively investigated [32,33].Moreover, recent work by Belda et al. in 2021 has expanded the scope to include the analysis of short-chain fatty acids and medium-chain fatty acids using SPME-GCMS [22].

NMR
H-NMR spectroscopy offers several advantages in the field of metabolomics.Firstly, H-NMR is highly reproducible and provides quantitative measurements, ensuring reliable analysis of metabolites.H-NMR offers the advantage of absolute quantification, with signal intensities directly proportional to the concentration of nuclei in the sample.Additionally, it allows for the analysis of intact biological samples without the need for extensive sample preparation, reducing the potential for analytical variability.H-NMR also enables the identification of unknown metabolites, making it valuable for comprehensive profiling of complex metabolic mixtures.Another advantage is its non-destructive nature, which preserves the sample for further analysis or retesting.Furthermore, H-NMR facilitates the study of metabolic pathways and fluxes by utilizing stable isotope labeling techniques.With its ability to detect multiple atomic nuclei, including 1H, H-NMR offers flexibility in metabolite detection.Overall, these advantages make H-NMR spectroscopy a powerful and versatile tool for investigating metabolites and their role in biological systems [74,75,89].
In contrast to the aforementioned benefits, it is noteworthy that out of the 51 studies examined, a mere seven studies opted to utilize H-NMR as their analytical approach.Among these, two studies employed H-NMR specifically for target analysis of hippuric acid and SCFAs [36,55].This choice was motivated by the fact that hippuric acid stands out as the most abundant phenolic acid in urine, and its concentrations often tend to be underestimated when employing MS detection due to saturation effects.The remaining five studies employed nontarget analysis with the aim of assessing the effects of moderate wine, dealcoholized wine, or wine derivatives following four weeks of intake [44,45,47,49,50].In the metabolomic applications included in this review all H-NMR samples presented with internal standard TSP and most of them used sodium azide and phosphate buffer [44,45,47].
Studies using H-NMR tend to identify fewer statistically significant metabolite differences compared to those employing LC-MS or GC-MS techniques.For example, in H-NMR analysis, only the study by Van Dorsten (2010) detected 19 metabolites with significant changes before and after 4 weeks of a wine derivative consumption [49].In contrast, the remaining seven metabolomic studies utilizing H-NMR determined only 1 to 11 compounds.Additionally, H-NMR analysis is predominantly used for the determination of smaller molecular weight compounds, particularly ethanol, amino acids and their derivatives like threonine and alanine [44,47,50,55].Notably, only five phenolic compounds were detected to have statistical difference before and after wine consumption using H-NMR (3-hydroxyphenylpropionic acid, vanilmandelic acid, dihydroferulic acid, isovanillic acid, 3,4-dihydrophenylglycol) while 116 were detected with LC-MS.

Chemometrics
Statistical analysis plays a crucial and indispensable role in metabolomic studies, as it enables the identification of statistically significant metabolomic changes, the elucidation of potential biomarkers [25], and the exploration of intricate metabolic pathways influenced.
The statistical analysis employed in the reviewed studies revealed a diverse range of techniques used to assess the impact of wine consumption on metabolomic profiles as can be seen in Figure 5.The most frequently utilized parametric tests included Student's t-test and ANOVA, which examined mean differences and evaluated metabolite level variations among multiple groups or conditions (Tables 1 and 2).ANOVA was used in many different types of studies, for example Urpi-Sarda et al. ( 2015) used it to compare changes in phenolic me- The most frequently utilized parametric tests included Student's t-test and ANOVA, which examined mean differences and evaluated metabolite level variations among multiple groups or conditions (Tables 1 and 2).ANOVA was used in many different types of studies, for example Urpi-Sarda et al. (2015) used it to compare changes in phenolic metabolites in plasma and urine after wine intervention treatments, while Motilva et al. used ANOVA to exclude extreme outliers of metabolites after a 1-day of dealcoholized wine intake and wine derivatives intake.Non-parametric tests such as the Wilcoxon test [34,41,48,55] and Mann-Whitney U [41,45,55,62] test were employed to examine changes in paired and independent samples, respectively.Regueiro et al. (2014) used these tests to separate group 1 with 100 mL of wine intake, with group 2 with 200 mL of intake and group 3 with 300 mL of intake [41].Additional statistical techniques included the Shaprio-Wilk test (in 8 studies) to assess data normality assumptions, the Kolmogorov test (in 4 studies) to examine metabolite distribution characteristics and the Spearman test [34,66] to explore correlations between metabolites (Tables 1 and 2).
Overall, the primary objective of any metabolomic study is to generate comprehensive metabolic profiles from both test and control samples, facilitating the identification of potential biomarkers.Regarding sample preparation, achieving comprehensive extraction of the extensive sample metabolome is an essential yet challenging prerequisite.Exploring the potential benefits of automation in this process holds promise.It is apparent that several analytical methods can be employed for this purpose, thereby raising the question of which specific method from the "metabolomics toolbox" should be given preference.Nevertheless, it is currently not possible to provide a straightforward answer to this query.Metabolomics analysis necessitates the integration of pragmatism, skepticism, and the effective utilization of available technology.Consequently, none of the individual analytical techniques outlined can independently yield a comprehensive profile due to inherent limitations in sensitivity or potential biases towards specific analytes.To overcome these challenges, a prudent strategy for achieving global metabolic profiling involves employing a diverse range of complementary analytical methods, there by maximizing coverage within a reasonable time frame [83,84,90].

Metabolites-Biomarkers of Wine Intake
The present review paper offers a comprehensive documentation of the metabolites whose concentrations undergo significant changes after wine consumption, as monitored in biological samples after long-term, or short-term wine consumption.This extensive research yielded a list comprising approximately 600 compounds that demonstrate statistical differences before and after wine consumption.Rigorous processing and scrutiny of this data involved eliminating duplicate compounds that might appear with different names and excluding unknown entities that may have emerged from non-targeted analyses.The remaining 361 compounds were then classified into distinct categories based on their chemical structure: (i) phenolic compounds, (ii) lipids and lipid-like molecules, (iii) organic acids, (iv) amino acids and derivatives, (v) carbohydrates and carbohydrate conjugates, and (vi) organic compounds as it can be seen in Figure 6.This classification aligns with the preferred categorization used in The Human Metabolome Database, ensuring clarity and consistency in our analysis.
yses.The remaining 361 compounds were then classified into distinct categories based on their chemical structure: (i) phenolic compounds, (ii) lipids and lipid-like molecules, (iii) organic acids, (iv) amino acids and derivatives, (v) carbohydrates and carbohydrate conjugates, and (vi) organic compounds as it can be seen in Figure 6.This classification aligns with the preferred categorization used in The Human Metabolome Database, ensuring clarity and consistency in our analysis.
For the majority of these metabolites we were able to pinpoint their behavior, either up-regulating or down-regulating after wine intake; however for some metabolites this information was unknown, as it was only reported that they exhibited statistical significant differences.It is also important to note that all studies uniformly agree on the direction of change (increase or decrease) for the examined metabolites following wine consumption.For the majority of these metabolites we were able to pinpoint their behavior, either up-regulating or down-regulating after wine intake; however for some metabolites this information was unknown, as it was only reported that they exhibited statistical significant differences.It is also important to note that all studies uniformly agree on the direction of change (increase or decrease) for the examined metabolites following wine consumption.

Phenolic Compounds
Phenolic compounds emerged as the most abundant group of metabolites correlated to wine intake, comprising a remarkable total of 125 compounds.In Table 3 all phenolic metabolites that have been determined in biological samples after wine intake are presented.In addition to target screening applications, recent advancements in analytical techniques have facilitated the untargeted analysis of biological samples that can reveal the entire phenolic profile [22,24,34,59,64].These metabolites belong to different phenolic compound classes, encompassing flavonoids, hydroxybenzoic acids, catechols, gallic acid metabolites, hydroxycinnamic metabolites, methoxyphenols, coumarins, phenylpropanoic acids and lignans.
In our comprehensive analysis, out of the 125 different phenolic compounds that were observed to vary significantly in the human metabolome after wine consumption, most of them were increased.Notably, within the phenolic compounds group, it's important to emphasize that the categorization encompasses not only the parent compounds but also their glucosides and glucuronides forms.It is noteworthy that a substantial proportion of the phenolic compounds that exhibited an increase in concentration fall within the three categories of flavonoids, catechols, and gallic acid and metabolites.Conversely, the compounds displaying decreased concentrations predominantly belong to the hydrocinnamic acids, a category including caffeic acid sulfate, dihydrocaffeic acid 3-sulfate and piplartine.
Among the identified metabolites, resveratrol, vanillic acid, syringic acid, and (epi)catechin were the most characteristic compounds displaying statistical significantly differences after wine consumption.Resveratrol garnered remarkable attention as an essential component that brings to red wine's health benefits [5].Studies dating back to 2001 sought to unravel its pharmacokinetics and bioavailability [29,30,57,74], predominantly focusing on hydrolyzed samples of urine and blood.As analytical techniques advanced, researchers made significant strides in identifying various resveratrol metabolites, such as sulfates, glucuronides, and glucoside derivatives with a predominant presence in urine samples [40,48,66].(Epi)catechin has also drawn attention.It has been under investigation since 1999 for its remarkable ability to offer protection against cardiovascular diseases (CVDs) [33,91].In recent years, researchers have also identified its derivatives, particularly sulfates and methylated glucuronides [46,51].
Furthermore, a multitude of studies have documented the presence of gallic acid metabolites, encompassing gallic acid itself, methylgallic acids, methylgallic acid sulfates, and pyrogallol.Particularly intriguing observations have been made regarding 4-O-methylgallic acid, the primary metabolite of gallic acid [92], which has been detected in both urine and plasma of participants after wine intake [38,46,70].Conversely, 3-Omethylgallic acid was exclusively identified in feces samples [26,28,35], thus revealing distinct distribution patterns between isomers within biological specimens.
In addition to hydroxycinnamic acids, such as ferulic acid, which has been widely reported to possess diverse physiological functions [93,94], the detection of other significant metabolites after wine intake has been documented.Caffeic acid, known for its potential as a carcinogenic inhibitor [95], has been detected in both urine and plasma of participants after wine consumption.Furthermore, p-coumaric acid, displaying anti-inflammatory activities [96], has been identified in urine, plasma, and feces [35,46,54,76], in contrast to its isomer, m-coumaric acid, which has been infrequently detected in biological samples and has been found only in a few selected studies in urine and feces [22,46].Lastly, hydroxybenzoic acids, including the important syringic acid, with various therapeutic uses [97], have been found across all types of samples (urine, plasma, feces) of participants that have consumed wine, while vanillic acid, 3,5-dihydroxybenzoic acid and 3-(3-hydroxyphenyl) propionic acid have been detected in both urine and feces.As an overall remark it is important to note that, although phenolic metabolites have been in the spotlight and have been thoroughly investigated in wine metabolomic studies, the results of the most recent non-targeted analysis studies indicate that other classes of metabolites, such as the endogenous metabolites, might be even more affected by wine consumption [24,25,34,44,60]; however, so far, these metabolites haven't received the same level of attention.

Lipids and Lipid-like Molecules
A total of 84 lipid metabolites, related to wine consumption, have been identified in biological samples, mainly in plasma (Table 4).These lipid metabolites encompass a diverse array of categories, reflecting their wide-ranging chemical and biological functions.They were systematically categorized into thirteen (13) classes: short-chain fatty acids (SCFA), methyl-branched fatty acids, hydroxy fatty acids, medium-chain fatty acids (MCFA), long-chain fatty acids (LCFA), fatty acid esters, glycosylglycerols, fatty acyl glycosides, furanoid fatty acids, glycerophospholipids, glycerolipids, steroids and steroid derivatives, and monoterpenoids.
The majority of these lipids exhibited up-regulation, with a notable increase in certain microbial-derived metabolites, particularly short-chain fatty acids (SCFAs), as observed in the study by Belda in 2021 and Jiménez-Girón in 2015, 2014 [22,26,27].Additionally, there were lipid metabolites, such as capric acid and arachidonic acid, which displayed correlations with wine consumption solely in an observational study, as reported by Pallister et al. in 2016 [63].Moreover, 2,3-dihydroxyvaleric acid and sphingomyelins have been consistently detected in biological samples after wine intake, the first increasing and the second decreasing after wine-consumption as Haas et al. (2022) and Jacobs et al. (2012) reported [34,55].
While 2,3-dihydroxyvaleric acid's origin remains relatively unexplored, sphingomyelins, in contrast, have undergone extensive investigation and are recognized for their pivotal role in regulating lifespan [98].Additionally, it is worth noting that cholesterol and sphingomyelins' levels exhibit a coordinated regulatory relationship [99].

Organic Acids and Derivatives
The organic acids and derivatives class includes a total of 40 reported compounds, with 24 identified as aliphatic organic acids and 15 as benzene organic acids and substituted derivatives (Table 5).Intriguingly, the majority of these metabolites exhibited an increase in concentration following wine consumption, suggesting a potential metabolic response triggered by wine intake.In contrast, only a limited subset of five metabolites showed a decrease after wine consumption, including formic acid, 2R,3R-dihydroxybutyric acid, dimethylguanidino valeric acid (DMGV), tricarballylic acid, and 3-indoxylsulfuric acid [27,34,50,60].
Among organic acid metabolites, 3-hydroxyphenylacetic acid, tartaric acid, 4-hydroxyhippuric acid, 4-hydroxyphenylacetic acid, and hippuric acid were consistently detected in multiple studies within the metabolome of participants after wine or wine derivative consumption.Tartaric acid is one of the principals acids found in wine [100], and a confirmed biomarker for wine consumption [41].Equally noteworthy are the hydroxyphenylacetic group compounds, with 3-hydroxyphenylacetic acid being a metabolite of rutin [101] and 4-hydroxyphenylacetic acid a metabolite of polymeric proanthocyanidins [102], alongside 3,4-dihydroxyphenylacetic acid, deriving from microbiota-driven quercetin metabolism [103].Additionally, the hippuric group stands out, led by hippuric acid, a common urinary component that notably increases with the consumption of polyphenols, particularly (epi)catechin as present in tea [104,105].This elevation is attributed to the transformation of these phenols into benzoic acid and subsequently into hippuric acid, excreted in urine.Moreover, 4-hydroxyhippuric acid serves as a secondary metabolite of hesperidin [106] and has been detected post-milk consumption with cocoa powder [107], while its isomer, 3-hydroxyhippuric acid, arises from flavonoids and hydroxycinnamates metabolism [108,109].
Overall, although the number of organic acid metabolites that have been identified in wine metabolomic studies is relatively limited, their presence in the human metabolome following wine intake is documented a total of 90 times across 51 metabolomic investigations, establishing this category as one of the most frequently detected groups of metabolites in the context of wine consumption.

Amino Acids and Derivatives
Across the 51 metabolomic studies, 38 amino acids and derivatives were reported to be influenced by wine consumption (Table 6).Notably, this group exhibited the highest proportion of down-regulated compounds, with nearly 50% of amino acids and their derivatives displaying down-regulation.Jacobs et al. ( 2012) reported reduced levels of tyrosine, threonine, and lysine in fasted plasma, consistent with a prior study indicating that polyphenols can influence protein digestibility [55].Of particular interest, threonine was the only amino acid reported in two separate studies.Specifically, Jacobs et al. in 2012 observed a decrease in threonine levels after short-term (4 days) of wine derivative intake, while in Vázquez-Fresno et al. in 2016 noted a statistical difference before and after 28 days of wine consumption using ANOVA analysis [44,55].The remaining 37 compounds were each reported only once, and their fluctuations related to wine intake.

Carbohydrates and Carbohydrate Conjugates
In the investigation of all biological samples, specifically urine and plasma, only wine consumption (Table 7).Most of them showed an increase in their content while only N-acetylneuraminate, which is involved in aminosugar metabolism according to Haas et al. (2022), was decreased [34].
Within this category, a mere two compounds stood out: mannitol [44,45,47] and scylloinositol [60,62,63], consistently detected across three distinct research papers.Mannitol is a polyol that is naturally found in wine and is formed by heterofermentative lactic acid bacteria through fructose reduction.According to Vazquez-Fresno in 2012, the presence of this polyol in urine samples following red wine drinking could be due to its fast removal from the body before undergoing metabolism [47].Furthermore, scyllo-inositol has been linked to a variation in the SLC5A11 gene (rs4787294), which encodes a sodium-dependent glucose transporter that transports myo-and scyllo-inositol.SLC5A11 gene markers have been linked to susceptibility to systemic lupus erythematosus (SLE) [63].

Vitamins and Energy Compounds
Vitamins and energy-related compounds represent a smaller subset of metabolites that have been identified in recent metabolomic studies as being influenced by wine intake (Table 8).In these studies, only five compounds within this category were characterized, with analyses confined to plasma and urine samples [25,34,55].Notable findings included an increase in nicotinic acid (a microbial fermentation product of aromatic amino acids) as well as an increase in pantoic acid (a precursor of vitamin B5) and isocitric lactone.[34] Additionally, a decrease in phosphate and cytidine triphosphate was reported [25].

Other Organic Compounds
Lastly, the assortment of 45 compounds that did not find placement in the preceding categories were placed in the "other organic compounds" class.Among these compounds, 12 were classified as aliphatic compounds, and the remaining 33 were identified as cyclic.These 33 cyclic compounds were further segmented into various subgroups, including organoheterocyclic, benzene and supstituted derivatives, pyrimidines, purines, and alkaloids.
The fluctuation of these compounds in human metabolome after the consumption of wine is presented in Table 9.An up-regulation was noted in most of these organic compounds following wine intake.However, there were exceptions, such as the endogenous metabolite xanthine [26,27], and metabolites of different origins, like ajoene and cinnamyl cinnamate [25], which were reported to decrease.A particularly noteworthy finding was the down-regulation of stercobilin and urobilinogen, byproducts of bilirubin degradation, as elucidated by Jimenez in 2015 [27].Urobilinogen remaining in the intestine is oxidized to form brown stercobilin, which imparts the characteristic color to feces.This finding aligns with numerous studies suggesting that flavan-3-ol-rich sources like wine may influence the intestinal microbiota by enhancing beneficial bacteria while inhibiting other groups, such as Clostridium spp.[26,27].

Conclusions
The extensive review of 51 metabolomic studies led to some very interesting observations and conclusions.Noticeable alterations in the human metabolome were identified before and after wine consumption, following the metabolomic analysis of different biological samples and using various analytical techniques.This meticulous analysis of published metabolomic studies resulted to a comprehensive documentation and categorization of all metabolites that have been reported to appear in the human body after wine consumption or to be affected (increased or decreased) by it.This extensive list could be a valuable tool for researchers working in the field of wine metabolomics and could even constitute the keystone for the preparation of ready-to use MS databases, significantly facilitating non-target/suspect screening applications.
Furthermore, the distribution of metabolite categories in various bodily matrices reveals important insights.In general, urine stands out as the sample type with the highest number of detected metabolites and emerges as the primary matrix for identifying phenolic metabolites associated with wine consumption.When investigating lipids and amino acids and their derivatives, plasma is the matrix of choice, offering a suitable environment for lipid-related inquiries.Additionally, the presence of organic acids and their derivatives in urine, plasma, and feces samples suggests a dynamic equilibrium between metabolic pathways and excretion routes that can be modulated by wine consumption.In contrast, carbohydrates exhibit limited detection in these studies, likely due to their inherent resistance to the influence of wine consumption.This phenomenon arises due to the typically abundant presence of carbohydrates in the urine metabolome [111].
An intriguing observation also arises from studies conducted on individuals with good health versus those with underlying health conditions.In such a comparison, two compounds surfaced as being exclusive to studies involving individuals with health issues subsequent to wine consumption: mannitol and 3-methyl-2-oxovalerate showed a significant increase in urine and plasma samples after wine consumption from individuals with CAD, T2D or with ≥3 CHD risk factors [34,45,47].Regarding the duration of the wine intake, no clear differentiation in the human metabolome is perceived after short-term and long-term wine consumption.The studies by Motilva et al. (2016), investigating into acute dealcoholized wine consumption [51], and Urpi-Sarda et al. (2015) on 4-week dealcoholized wine consumption [54], indicate that the concentrations of phenols in both blood and urine exhibit no statistically significant differences between long-term and short-term consumption periods.
Overall, the findings presented in this review shed light on the intricate network of interconnected metabolites and pathways influenced by wine consumption, emphasizing the need for further research to comprehend their mechanistic relationships and overall health implications.It is imperative to underscore that the exploration of the interplay between wine and human health remains an ongoing endeavor, and this paper is expected to catalyze future investigations in the fields of metabolomics, biochemistry, and medicine within this domain.

Figure 1 .
Figure 1.Per capita wine consumption in 2019, quantified in liters of absolute alcohol per annum (approximately 0.12 L of pure alcohol are found in 1 L of wine) [1].

Figure 3 .
Figure 3. Distribution of collected biological samples in clinical studies of wine metabolism.

Figure 3 .
Figure 3. Distribution of collected biological samples in clinical studies of wine metabolism.

Figure 4 .
Figure 4. Analytical techniques employed in clinical studies of wine metabolism.

Figure 4 .
Figure 4. Analytical techniques employed in clinical studies of wine metabolism.

Molecules 2023 , 48 Figure 5 .
Figure 5. Statistical analysis employed in clinical studies of wine metabolism.

Figure 5 .
Figure 5. Statistical analysis employed in clinical studies of wine metabolism.

Figure 6 .
Figure 6.Categorization of metabolites affected by wine intake.Figure 6. Categorization of metabolites affected by wine intake.

Figure 6 .
Figure 6.Categorization of metabolites affected by wine intake.Figure 6. Categorization of metabolites affected by wine intake.

Table 1 .
Interventional Clinical Studies in wine metabolomics research.

Table 2 .
Observational Clinical Studies in wine metabolomics research.

Table 3 .
Phenolic metabolites affected by wine consumption.

Table 4 .
Lipids and lipid-like molecules affected by wine consumption.

Table 5 .
Organic acids and derivatives affected by wine consumption.

Table 6 .
Amino acids and derivatives affected by wine consumption.

Table 7 .
Carbohydrates and carbohydrate conjugates affected by wine consumption.

Table 8 .
Vitamins and energy-related molecules affected by wine consumption.

Table 9 .
Other Organic compounds metabolites affected by wine consumption.