Critique 273: Circulating metabolites may illustrate relationship of alcohol consumption with cardiovascular disease

Two papers were recently published which consider the complexity of liver and gut-generated circulating metabolites of alcoholic beverages, such as wine, and their specific roles in human health (referred to as metabolomics). These papers collectively suggest that the 60 or more alcohol-associated metabolites following consumption, may be related to the risk of cardiovascular and other diseases.

Title: Circulating metabolites may illustrate relationship of alcohol consumption with cardiovascular disease
Authors: Li Y; Wang M; Liu X; Rong J; Miller PE; Joehanes R; Huan T; Guo X; Rotter JI; Smith JA; Yu B; Nayor M; Levy D; Liu C; Ma J
Citation: BMC Medicine (2023)
Author’s Abstract
Background Metabolite signatures of long-term alcohol consumption are lacking. To better understand the molecular basis linking alcohol drinking and cardiovascular disease (CVD), we investigated circulating metabolites associated with long-term alcohol consumption and examined whether these metabolites were associated with incident CVD.
Methods Cumulative average alcohol consumption (g/day) was derived from the total consumption of beer, wine, and liquor on average of 19 years in 2428 Framingham Heart Study Offspring participants (mean age 56 years, 52% women). We used linear mixed models to investigate the associations of alcohol consumption with 211 log-transformed plasma metabolites, adjusting for age, sex, batch, smoking, diet, physical activity, BMI, and familial relationship. Cox models were used to test the association of alcohol-related metabolite scores with fatal and nonfatal incident CVD (myocardial infarction, coronary heart disease, stroke, and heart failure).
Results We identified 60 metabolites associated with cumulative average alcohol consumption (p < 0.05/211 ≈ 0.00024). For example, 1 g/day increase of alcohol consumption was associated with higher levels of cholesteryl esters (e.g., CE 16:1, beta = 0.023 ± 0.002, p = 6.3e − 45) and phosphatidylcholine (e.g., PC 32:1, beta = 0.021 ± 0.002, p = 3.1e − 38). Survival analysis identified that 10 alcohol-associated metabolites were also associated with a differential CVD risk after adjusting for age, sex, and batch. Further, we built two alcohol consumption weighted metabolite scores using these 10 metabolites and showed that, with adjustment age, sex, batch, and common CVD risk factors, the two scores had comparable but opposite associations with incident CVD, hazard ratio 1.11 (95% CI = [1.02, 1.21], p = 0.02) vs 0.88 (95% CI = [0.78, 0.98], p = 0.02).
Conclusions We identified 60 long-term alcohol consumption-associated metabolites. The association analysis with incident CVD suggests a complex metabolic basis between alcohol consumption and CVD.

In conjunction with:
Title: Exploring human metabolome after wine intake – A review
Authors: Lekka P; Fragopoulou E; Terpou A; Dasenaki M
Citation: Molecules. 2023 Nov 15;28(22):7616.
Author’s Abstract
Background 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.
Method 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 faeces samples, encompassing both short-term and long-term studies of the consumption of wine and wine derivatives.
Results 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.
Conclusions 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.

Forum comments

This preprint of an impressive study, which has not been peer reviewed yet, reports on 60 metabolites changing in long-term alcohol consumption using the Framingham Heart Study Offspring participants of whom the alcohol consumption has been well-monitored over a long period of time. Metabolite changes have been associated with cardiovascular disease (CVD) incidence to better understand the molecular basis between alcohol consumption and CVD. The study builds on some previous publications using metabolomics profiles predicting other health outcomes in the Framingham study (Wang et al., 2011).

Metabolomics profiling acts as a potent and high throughput tool offering new insights on disease pathogenesis and has potential in the early diagnosis. Early detection of pathogenesis through profiles used as a biomarker for a disease holds promise for early diagnosis and treatment.

However, metabolomics profiling is subject to numerous limitations, specifically in the case of an epidemiological setting. A factor like alcohol consumption may affect numerous pathways, many of these being unknown and therefore potentially missed. In this case, the targeted metabolite profiling included various negatively and positively charged metabolites as well as lipid metabolite species. In total 211 metabolites were included in the analysis, whereas about 40,000 metabolites are being identified in the Human Metabolome Database (HMDB)[1]. Of these 211 metabolites, 60 metabolites were changed in long-term alcohol consumption, which may suggest that some pathways may have been missed.

One of the main metabolites being identified and significantly associated with the cumulative average total alcohol consumption adjusted for potential confounders was cholesteryl palmitoleate, a cholesterol ester involved in cholesterol metabolism. Only one other cholesterol ester was positively associated with alcohol consumption namely cholesteryl eicosapentaenoic acid. We already know that cholesterol metabolism is important in CVD aetiology, but the role of this specific cholesterol bound fatty acid is unclear. Palmitoleic acid (C16:1) is one of the main monounsaturated fatty acids of the omega-7 fatty acid family. Palmitoleic acid may be a major product of endogenous lipogenesis in the liver (Shramko et al., 2020) and its influence on the cardiovascular system is inconsistent.

It would have been interesting to know if this cholesterol ester was located in HDL particles or in LDL particles. HDL particles are important in reverse cholesterol transport (Marques et al., 2018), the disposal route of cholesterol from the vascular wall, whereas LDL particles are essential in the opposite direction bringing cholesterol to the vascular wall giving rise to so-called atherosclerotic plaques. The authors have adjusted their association analysis for various confounding factors including total and high-density lipoprotein. Could this mean that HDL particles are more enriched in cholesteryl palmitoleate with alcohol consumption, indicative for an increased reverse cholesterol transport? Some nutrition intervention studies have shown that moderate alcohol consumption with a meal changes postprandial lipoprotein composition. These effects were transient and still observed up to 11 hours after the meal, resulting in a raised HDL-cholesterol/apo A-I ratio. The effects were considered anti-atherogenic and may contribute to the observed protection against coronary heart disease by moderate alcohol consumption (Van Tol et al., 1998). Other human intervention studies have shown that moderate alcohol consumption does stimulate reverse cholesterol transport (Rohatgi et al., 2014; Van Der Gaag et al., 2001).

Similar observations were made for phospholipids (Hendriks et al., 1998). In a diet-controlled randomized trial with alcohol either as wine, beer and spirits with a meal, HDL phospholipids were increased still at 13 hours after a meal. These data correspond well with what has been reported in this metabolomics study.

Metabolic profiles may be influenced by many different behavioural and environmental factors like diet, sleep pattern, drinking pattern, treatments, living conditions and many other variables. It is likely that not all or that some have only been partially corrected for. So, identifying metabolites that can be used as a biomarker for CVD risk may be challenging. In a more controlled setting, e.g. in a nutrition intervention where at least alcohol consumption is controlled for, in quantity, pattern and as many as possible other confounding factors may increase the chances of identifying metabolic pathways related to for instance established CVD biomarkers. In case of the moderate alcohol CVD association, some suggest that most of the mechanisms and therefore the metabolites involved have been identified (Mukamal et al., 2005).

Also, because of the observational nature of the findings without experimental validation, as the authors state in their discussion section, causality cannot be inferred. The very descriptive nature of this research and the presence of many potential and various well-substantiated pathways involved in the association between moderate alcohol consumption and CVD risk makes one wonder what this study may add to the well-established association. In searching PubMed for papers on human studies reporting on metabolites in cardiovascular diseases 1417 studies were identified. How these studies relate to the findings based on this targeted panel is unclear. So, the added value of this type of research may be limited since the combination of a large group of people with heterogenous lifestyles analysed with the generalized but limited methodology of targeted metabolomics may rather confuse the existing concepts than add to our understanding of the well-documented moderate alcohol – CVD relationship.

Specific Comments from Forum Members

Forum member Mattivi suggests that “the study by Li et al. (2023) addresses, in its targeted metabolomics analysis, a considerable number of 211 circulating metabolites, which still remains very limited compared to the huge complexity of the human metabolome. With the aim of investigating the impact of alcoholic beverage consumption on observed concentrations.

On the one hand, a strength of this approach is to identify a limited number of significant metabolites (7, corresponding to 3.3% of those studied) that are impacted in the same direction by alcoholic beverage consumption (i.e. significantly associated with the cumulative consumption of all three types of alcoholic beverages), and deserve to be further investigated with analyses aimed at expanding the number of analytes in the chemical classes of interest (CE16:1, LPC 20:5, PC 32:0, PC 32:1, PC 34:1, PC-B 36:4, and fumarate-malate).

On the other hand, this study confirms how the consumption of the different types of beverages (beer, spirits, and wines) also induces drink-specific changes to the concentrations of a large number of circulating metabolites (81, accounting for 38.4% of the investigated compounds). Among the metabolites observed, the strength of associations is in many cases different for different beverages; in some cases, the direction in association with consumption is significant and specific only to consumption of the specific beverage.

This aspect again highlights the importance of not underestimating the role of the matrix and consumption patterns, which can often be associated with the type of alcoholic beverage consumed, even from the perspective of impact on the metabolome.

Forum member Harding comments that he was “struck by the comprehensive nature of review of Lekka et al. (2023) and in particular the large number of clinical intervention studies already undertaken to inform causation, as shown in Table 1.  In the light of these, what is the point of conducting more epidemiology?  Should not the epidemiology come first, to indicate what human studies are sensible to do?  What possible use are epidemiological findings, which make their best (and often very approximate) estimate of wine consumption by a particular group in the population as in Li et al. (2003), when there are already so many studies on real subjects that control exactly how much wine is consumed, and then record the metabolical outcome?  Unless of course this epidemiological study identified new metabolites that merit further study, and I don’t know whether it has.  I realise that Lekka et al. (2023) is only about wine, and the epidemiology covers all alcohol consumption, but this does not invalidate the general point.”

Forum member Ellison suggests that “Lekka et al. (2023) is an extremely important contribution to the interpretation of existing and future studies by providing data to show the mechanisms of health effects related to the consumption of wine and other beverages containing alcohol.  The approach presented can be used to explain why the regular and moderate intake of wine with food, in particular, has been shown to markedly decrease the occurrence of cardiovascular diseases and to decrease the risk of total mortality.  Among cohort studies, such findings have been amazingly consistent over many decades.  And, despite some biased or even fraudulent attempts by a small group of scientists to deny such results, continue to be exhibited in essentially all well-done, scientifically balanced epidemiologic studies today.

Studies of alcohol intake and health are generally based only on the self-report of alcohol intake.  The use of the investigation of metabolic factors, as described in this paper, can add key information by which to substantiate (or question) the reported alcohol intake.  Further, and more importantly, it can provide evidence on the mechanisms of effects, for both observational studies and intervention trials, when seeking to determine the net health effects associated with the consumption of wine and other beverages containing alcohol.  While epidemiologic studies attempt to adjust for a multitude of potential confounders of effect (e.g., the pattern of drinking, the type of beverage, whether with or without food, socio-economic factors, diet, medical conditions such as obesity, etc.), the additional use of metabolomics can greatly advance our ability to determine the degree to which the effects seen are truly related to alcohol consumption or to confounding variables.  The relationships described in this paper suggest that considering metabolic data will greatly strengthen the demonstrated beneficial effects on health from moderate drinking.”

Forum member Mattivi considers that “the study by Lekka et al. (2023) shows how numerous the metabolic pathways are which, based on human intervention studies, are significantly impacted by wine consumption.

In the specific case of this beverage, which is a fermented fruit juice, not surprisingly all the metabolites and catabolites typical of fruit consumption are found. A very large part of the compounds observed are the result of the host-microbe co-metabolism (highlighting the fundamental role of the intestinal microbiome), and many of these also have prolonged persistence in biofluids.

In recent years, research has shown us how the intestinal microbiome can have an importance not unlike that of the liver for the metabolism of xenobiotics. And an individual diversity that can easily cause the circulating concentrations of some classes of metabolites (for example, gamma valerolactones) to vary by an order of magnitude with the same dietary intake.

The quantities of alcoholic beverages consumed and the methods of consumption (in association or not with food, with or without episodes of excessive consumption, etc.) are also expected to impact on the health of the microbiome, and therefore on how the numerous bioactive compounds of the wine, can be processed, possibly modulating the concentrations of dozens of microbial catabolites with potential bioactivity. This review highlights the metabolism deriving from the fact that man does not only consume “ethanol”, but specifically – in the case of wine – a vegetal matrix of extraordinary compositional complexity.”

Concluding comments

Forum member Ellison added that “a discussion of this paper with colleagues at Boston University led to questions related to the fact that in most observational studies of alcohol effects, there are no data on other dietary factors. It is well known that the consumption of alcohol is related to other components of the diet, in that wine consumers and moderate drinkers tend to report healthier diets than consumers of other beverages or heavy drinkers.  Undoubtedly, many of the metabolites reported in this paper are affected by diet.  This makes it difficult to determine if other dietary constituents may have contributed to the reported effects on metabolites attributed to alcohol.  It makes the use of controlled trials even more important.”


Hendriks, H. F. J., Veenstra, J., Van Tol, A., Groener, J. E. M., & Schaafsma, G. (1998). Moderate doses of alcoholic beverages with dinner and postprandial high density lipoprotein composition. Alcohol and Alcoholism, 33(4), 403–410.

Marques, L. R., Diniz, T. A., Antunes, B. M., Rossi, F. E., Caperuto, E. C., Lira, F. S., & Gonçalves, D. C. (2018). Reverse Cholesterol Transport: Molecular Mechanisms and the Non-medical Approach to Enhance HDL Cholesterol. Frontiers in Physiology, 9, 526.

Mukamal, K. J., Jensen, M. K., Gronbaek, M., Stampfer, M. J., Manson, J. E., Pischon, T., & Rimm, E. B. (2005). Drinking frequency, mediating biomarkers, and risk of myocardial infarction in women and men. Circulation, 112(10), 1406–1413.

Rohatgi, A., Khera, A., Berry, J. D., Givens, E. G., Ayers, C. R., Wedin, K. E., Neeland, I. J., Yuhanna, I. S., Rader, D. R., De Lemos, J. A., & Shaul, P. W. (2014). HDL Cholesterol Efflux Capacity and Incident Cardiovascular Events A BS TR AC T. NEJM.Org. N Engl J Med, 25(18), 2383–2393.

Shramko, V. S., Polonskaya, Y. V, Kashtanova, E. V, Stakhneva, E. M., & Ragino, Y. I. (2020). The Short Overview on the Relevance of Fatty Acids for Human Cardiovascular Disorders. Biomolecules, 10(8).

Van Der Gaag, M. S., Van Tol, A., Vermunt, S. H. F., Scheek, L. M., Schaafsma, G., & Hendriks, H. F. J. (2001). Alcohol consumption stimulates early steps in reverse cholesterol transport. Journal of Lipid Research, 42(12).

Van Tol, A., Van Der Gaag, M. S., Scheek, L. M., Van Gent, T., & Hendriks, H. F. J. (1998). Changes in postprandial lipoproteins of low and high density caused by moderate alcohol consumption with dinner. Atherosclerosis, 141(SUPPL. 1).

Wang, T. J., Larson, M. G., Vasan, R. S., Cheng, S., Rhee, E. P., McCabe, E., Lewis, G. D., Fox, C. S., Jacques, P. F., Fernandez, C., O’Donnell, C. J., Carr, S. A., Mootha, V. K., Florez, J. C., Souza, A., Melander, O., Clish, C. B., & Gerszten, R. E. (2011). Metabolite profiles and the risk of developing diabetes. Nature Medicine, 17(4), 448–453.

Comments on this critique by the International Scientific Forum on Alcohol Research were provided by the following members:

Henk Hendriks, PhD, Netherlands

Creina Stockley, PhD, MBA, Independent consultant and Adjunct Senior Lecturer in the School of Agriculture, Food and Wine at the University of Adelaide, Australia

R. Curtis Ellison, MD, Section of Preventive Medicine/Epidemiology, Boston University School of Medicine, Boston, MA, USA

Richard Harding, PhD, UK

Erik Skovenborg, MD, specialized in family medicine, member of the Scandinavian Medical Alcohol Board, Aarhus, Denmark

Fulvio Ursini, MD, Dept. of Biological Chemistry, University of Padova, Padova, Italy

Andrew Waterhouse, PhD, Department of Viticulture and Enology, University of California, Davis

Fulvio Mattivi, MSc, Head of the Department Food Quality and Nutrition, Research and Innovation Centre, Fondazione Edmund Mach, in San Michele all’Adige, Italy

Giovanni Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy