Recent advances in alcohol metabolism: from the gut to the brain

Abstract

Globally, alcohol is the most widely used psychoactive drug and a leading cause of premature death among individuals aged 15–49 years. Understanding the absorption, distribution, metabolism, and excretion of alcohol in the human body, otherwise known as alcohol pharmacokinetics, is essential for predicting its behavioral effects and toxic consequences. This review examines the evolutionary origins of alcohol consumption and metabolism, focusing on the activity of alcohol dehydrogenase enzymes across species, which serve as key catalysts in alcohol oxidation. It also highlights recent advances in understanding central alcohol metabolism and updates on the potential clinical significance of non-oxidative pathways of alcohol metabolism and endogenous alcohol production, particularly in the context of liver disease. In addition, the review inspects factors that modulate alcohol metabolism, including genetic polymorphisms, biological sex, food intake, women’s reproductive status, and clinical interventions such as medications and metabolic surgeries. Understanding these sources of variability in alcohol metabolism is crucial for identifying individual risk factors and tailoring strategies to reduce alcohol-related harm. This comprehensive review offers a current perspective on alcohol pharmacokinetics, valuable insights into its implications for health, behavior, and potential innovative therapeutic targets.

Graphical Abstract

Graphical abstract was prepared with a licensed version of BioRender.com.

1. Introduction

Alcohol, the most widely used psychoactive drug worldwide, is a major contributor to premature death among people aged 15 to 49 years, driven by patterns of excessive and harmful use (1). Establishing clear criteria for defining the dose of alcohol consumed —considering the complexity of alcoholic beverages—and understanding its metabolic fate after ingestion are critical steps for estimating exposure to the brain and other organs and predicting the behavioral and toxic effects of alcohol use. Given its importance to human health, several excellent reviews on alcohol metabolism have been published over the last two decades (2–6). In this review, we explore the origins of alcohol use by humans and its metabolism and provide an update on the state of the science of pharmacokinetics, that is, the absorption, distribution, metabolism, and excretion of alcohol. We highlight recent advances in the understanding of the systemic and central metabolism of alcohol, including the discovery of regional expression of key alcohol metabolism enzymes and their genetic variants. Special attention is given to the modulating effects of biological sex, fed state, reproductive status of women, and popular clinical interventions (e.g., gastric surgeries, medications) on alcohol metabolism, which in turn affect an individual’s risk for alcohol-related harm and alcohol use disorder (AUD).

1.1. Alcohol metabolism: An epic spanning millions of years

The origins of alcohol consumption and attraction by humans align with its production (herein, alcohol and ethanol will be used interchangeably). Archaeological evidence has revealed that the Natufian people, an ancient group of hunter-gatherers in Israel (circa 13,000 BCE), brewed beer from wheat and barley (7), and the inhabitants of the Neolithic village of Jiahu in China (circa 7000 BCE) made fermented beverages from rice, honey, and various fruits, demonstrating the complexity of the flavor and ingredients of alcoholic beverages, even in its nascent stages. Interestingly, alcohol dehydrogenase (ADH) genes that code for the primary enzymes responsible for alcohol metabolism originated much earlier—more than 400 million years ago.

Paleogenetics ─the study of ancient DNA─ has revealed that ADH genes are present in several branches of the tree of life, from plants to fungi and animals, underscoring the capability of many organisms to metabolize alcohol (8–12). The well-documented consumption of alcohol from food sources such as fermenting fruits by some non-primate animals (e.g., birds and fruit bats) (13–16) led to the hypothesis that exposure to ethanol-rich food environments favored the evolution of traits associated with ethanol metabolism (17), and the ability to detect its sensory properties. Locating the source of alcohol would be evolutionarily advantageous in leading the organism to an abundance of food high in sugar and carbohydrates (18). However, ADH genes predate natural ethanol availability, a finding that refutes this hypothesis (19). Notwithstanding, although ADH activity is widespread in the phylogenetic tree, there is a great diversity of ADHs within and across species, including encoding genes, their structure, and tissue expression (19).

There are three main superfamilies of ADH that arose independently throughout evolution. Although they are all dimeric metalloproteins, Type I ADH is a “medium-chain” family of dehydrogenases and is the most common in vertebrates; Type II is a “short-chain” family of dehydrogenases most common in insects (11), and Type III is more commonly known in microorganisms (20). In vertebrates, most alcohol is oxidized by Type I ADHs, and, as will be discussed in upcoming sections, 5 classes of Type I ADH isoenzymes are found in humans and non-human primates, each of which has different catalytic properties and patterns of tissue-specific expression (Table 1). Therefore, these different ADH families may have responded differently to environmental selective pressures.

Table 1.

Human alcohol dehydrogenase (ADH)

ADH Class	Official Gene Nomenclature	Former Gene Nomenclature	Subunit Nomenclature	Enzyme Nomenclature	Km*	Vmax*	Tissue expression
I	ADH1A	ADH1	α	ADH1A	4	30	Liver
	ADH1B*1	ADH2*1	β1	ADH1B	0.013	5.2	Liver, adipose, brain, breast
	ADH1B*2	ADH2*2	β2	ADH1B	1.8	190
	ADH1B*3	ADH2*3	β3	ADH1B	61	140
	ADH1C*1	ADH3*1	γ1	ADH1C	0.1	32	Liver, stomach, intestine
	ADH1C*2	ADH3*2	β2	ADH1C	0.14	20
II	ADH4	ADH4	π	ADH2	11	9	Liver
III	ADH5	ADH5	χ	ADH3	>1000	100	Most tissues
IV	ADH7	ADH7	σ (μ)	ADH4	30	1800	Stomach, esophagus, mucosae
V	ADH6	ADH6	Not identified	ADH5	?	?	Liver, stomach

The change of the names used for the ADH genes in the literature creates some challenges in interpretation; therefore, in this table, we included both the official and the former gene nomenclature.

^*Km values represent ethanol catalytic efficiency and are given in mM, and V_max values represent the maximal metabolic rate of the enzyme and are given in turnover numbers per minute. Data from kinetics were extracted from (113); data from tissue expression were extracted from (151, 152).

One ADH isoenzyme that has been the focus of research on how dietary factors shape metabolic pathways due to its high expression in the upper gastrointestinal tissue is the ADH class IV, also known as ADH4 (21, 22). A recent genetic study spanning more than 171 species and across 7 mammalian orders found little evidence that sugar-rich diets, such as floral nectar and ripe fruit, drove mutations in the ADH7 gene (coding for the ADH4 isozyme) that favored ethanol metabolism (12). The only notable trend was observed in frugivorous and nectarivorous bats (12). Even though ADH4 likely evolved to oxidize non-ethanol endogenous substrates, such as retinol (19), there was a significant evolutionary mutation in this gene in most hominids approximately 10 million years ago. This mutation resulted in a 40-fold increase in the capacity to metabolize ethanol (23) and coincided with when great apes shifted to terrestrial habitats, enabling them to better tolerate ethanol found in fermenting fruit on the forest floor. Carn et al. (24) hypothesized that while this metabolic adaptation may have led to the accumulation of calories and fat from fructose metabolism, it may now increase the risk of excessive consumption and alcohol-related harm.

1.2. History of alcohol practices and beliefs in humans

The dynamics of alcohol consumption are influenced by cultural norms and religious beliefs. For example, ancient Egyptians considered beer a staple of daily life and a vital element in religious ceremonies, while medieval European monastic orders became skilled vintners, harnessing the fermentation of wine for economic and spiritual purposes (25–28). Alcohol has been used for medicinal purposes, and the beliefs surrounding its efficacy have been ingrained in the traditional wisdom of many cultures. Before the discovery of successful inhalation anesthesia (29), alcohol was used as a sedative during surgeries and childbirth, as a stimulant for resuscitation (29), and as a medication to prevent premature labor (30). Tonics containing alcohol were believed to be healthy due to thermoregulatory and digestive benefits and common treatment for colds and fevers (31), and for several years, ethanol was used for its calories within parenteral nutrition (32). Up until the 1970s and the discovery of fetal alcohol syndrome, beliefs regarding drinking during pregnancy ranged from posing little or no risk (33) to benefiting mothers (34, 35). Likewise, the benefits of alcohol for lactating women and children went largely unchallenged by science until the 1990s. The folklore of many cultures suggests that drinking during lactation facilitates milk let-down, rectifies milk insufficiency, and soothes and nourishes breastfed babies (36, 37). A popular book for new mothers, first published in 1963, claimed that “... alcohol has special virtues for the nursing mother ... this is the one time in life when the therapeutic qualities of alcohol are a blessing!” (38). These beliefs persist even today despite the scientific evidence that refutes this lore (39, 40). Such historical and cultural perspectives highlight the evolving role of alcohol in society, reflecting its dual nature as both a valued resource and potential risk to health.

1.3. Pick your poison: Types of beverages containing alcohol

Beverages containing alcohol are as diverse as the cultures that produce and consume them. Many societies have their own alcoholic drinks, combining local sugar sources, microbes, and production methods with their unique cultural practices (26, 41). The most common beverages globally are beer, wine, and spirits. Among other things, they differ in the concentration of alcohol they contain, that is, the percentage (%) of alcohol by volume (ABV).

Because there is no international definition of a “standard drink,” it is important to quantify the amount of pure ethanol consumed in grams. The amount is calculated by multiplying the % ABV by the volume consumed (in mL) and the density of ethanol (0.789 g/mL at 20°C). In the United States, one standard drink contains 14 g of ethanol, which is equivalent to a 12-oz beer (5% ABV), 5 oz of wine (12% ABV), or 1.5 oz of spirit (40% ABV). For example, one standard drink of beer in the US is the product of ABV of the beer, volume of the drink, and alcohol density [(5ml of alcohol /100 ml) × 355 mL × 0.789 g/mL] corresponding to 14 g of ethanol. This formula can be used to define an ethanol dose (e.g., grams of ethanol per kilogram of body mass), which is useful in blood alcohol concentration (BAC) estimations (6).

Beer, wine, and spirits are the three classes of beverages that contain alcohol but vary in water content and non-ethanol compounds, such as flavors and congeners that are produced during fermentation and distillation. Congeners include other alcohols like methanol or non-alcohol aldehydes, esters, and flavonoids, that contribute to the overall complexity of the flavor, aroma, and color profile of the beverage (42) and, in turn, the positive aspects of the overall drinking experience. However, some individuals are sensitive to the negative effects of congeners.

In general, alcohol-containing beverages that are darker in color (e.g., bourbon, red wine) typically have higher congener levels than those lighter in color (e.g., vodka, white wine) and are linked to triggering more pronounced hangover symptoms with more severe headaches (43, 44). Although the mechanism linking congeners with headaches has been elusive, a recent in vitro study that focused on the congener quercetin offers some insight. In its metabolism, alcohol is first oxidized by ADH to acetaldehyde, a toxic byproduct (6), and then to acetate by aldehyde dehydrogenase (ALDH). Recent evidence suggests that quercetin, a flavonoid abundant in red grapes, is metabolized in the liver into quercetin-3-glucuronide, which strongly inhibits an isoform of ALDH (44). Reduced ALDH enzymatic activity causes acetaldehyde to accumulate in the bloodstream after alcohol consumption, triggering symptoms such as nausea, tachycardia, facial flushing, and headache (45). These findings suggest that it is the combined metabolism of quercetin and ethanol in red wine that leads to acetaldehyde buildup and may be a mechanism underlying the headaches associated with drinking 1 or 2 glasses of red wine. However, clinical studies that confirm these in vitro observations are warranted.

1.4. Endogenous alcohol: Auto-brewery syndrome

There is a medical condition, albeit rarely diagnosed, in which an individual experiences the effects of alcohol without actually ingesting alcohol. This condition, known as auto-brewery syndrome (ABS) (46), results from the endogenous production of ethanol through the fermentation of carbohydrates by microbes in the gastrointestinal system, oral cavity, and/or urinary tract (47–49). Individuals with ABS often have gastrointestinal disorders, such as Crohn’s disease, or a history of antibiotic overuse, both of which can alter the microbiome, allowing fungal or bacterial overgrowth and potentially contributing to the development of this condition (50–52). For example, ABS has been linked to the presence of yeast strains such as Saccharomyces cerevisiae, S. boulardii, Candida spp., and rare bacteria like Klebsiella pneumoniae, Enterococcus faecium, and Citrobacter freundii (53–55). While endogenous ethanol production is known to cause slight elevations in BAC, there is debate over whether ABS alone can cause BAC elevations to levels above 0.8 grams per liter of blood (g/L; equivalent to 0.08% BAC), which is the legal limit for driving in the U.S. and therefore could have clinical or forensic significance (6, 56).

Nevertheless, rare cases of ABS with medicolegal consequences have been documented. For example, one patient was acquitted of traffic-related legal charges after clinical and forensic evidence confirmed his ABS; he had BAC readings of 1.6 g/L upon admission to a highly controlled inpatient unit, and during the next 20 hours, his BAC readings remained high and even increased in the absence of exogenous alcohol consumption (57).

Similar to oral ethanol consumption, endogenous ethanol production has been implicated in the development of non-alcohol-associated fatty liver disease, NAFLD (now known as metabolic dysfunction-associated steatotic liver disease, MASLD), and non-alcohol-associatedsteatohepatitis NASH (now metabolic dysfunction-associated steatohepatitis, MASH) (54, 58–63). One of the strongest evidence that MASLD can result from endogenous alcohol production by the gut microbiota comes from a preclinical study (54). In this study, the transfer of a high-alcohol-producing strain, Klebsiella penumoniae —isolated from an individual with ABS and MASLD—into germ-free mice, either through oral gavage or fecal transplant, induced MASLD. Importantly, selectively eliminating this strain prior to fecal transplantation prevented the development of MASLD in recipient mice (54). Furthermore, more than 60% of patients with steatotic liver disease in a Chinese cohort were carriers of this bacterial strain (54), suggesting that endogenous alcohol production may contribute to liver and other diseases more frequently than currently recognized. A recent study of individuals with and without MASL/MASH confirms and extends these findings. Fasting BAC measured in the portal veins of individuals with MASL/MASH were significantly higher than in individuals without MASLD/MASH (medians; control 2.1 mM; MASLD 8.0 mM and MASH 21.0mM) (63). Additionally, the study shows that BAC, which was negligible at fasting, increased to measurable levels 2 hours after a meal tolerance test. Notably, individuals with more advanced liver disease exhibited the highest endogenous BAC, and BAC could be manipulated to increase or decrease by pre-treating patients with an infusion of an inhibitor of hepatic alcohol metabolism (4-methylpyrazole) or with a course of broad-spectrum antibiotics, respectively (63). These findings support a causal role of endogenous alcohol of microbial origin in liver disease and underscore the need for further research into this intriguing condition.

2. Down the hatch: From ingestion to circulation

2.1. Alcohol absorption

After consuming ethanol, the process of absorption begins. Ethanol is a small polar molecule absorbed by passive diffusion slowly from the stomach and more rapidly from the small intestine due to the large surface area created by the presence of villi. Therefore, the rate of gastric emptying plays a key role in determining alcohol’s absorption rate (5). While many factors influence gastric emptying, components of the beverage itself also affect the absorption rate. For example, carbon dioxide in alcohol-containing carbonated drinks (e.g., champagne) may speed up ethanol absorption by accelerating gastric emptying (64, 65), while beverages high in carbohydrates, like beer, may slow gastric emptying and delay alcohol uptake into the bloodstream (66, 67). The alcohol concentration in a beverage also significantly impacts its absorption rate. For beverages containing <30% ABV, the higher the concentration, the faster the absorption rate, leading to higher peak BAC levels (68). However, above 30% ABV, ethanol can irritate the mucosa and cause pylorospasm, which delays gastric emptying (69–71).

Factors beyond the beverage itself also influence alcohol absorption. Individual characteristics such as sex, reproductive status in women, medications, and whether alcohol is consumed with or without food all contribute to individual variability in alcohol absorption (see details in section 5). Food in the stomach slows gastric emptying, which delays alcohol absorption (69, 72). As a result, alcohol is absorbed more quickly in the fasted state than in the fed state, supporting the common advice to avoid drinking on an empty stomach (70, 71, 73, 74). Solid foods slow gastric emptying more than liquids, delaying alcohol absorption (74). Overall, the stomach empties in proportion to a meal’s caloric load rather than its macronutrient composition (75–77). Even small calorie differences matter, for example, peak breath alcohol concentrations (BrAC) is about 20% lower when the drink is sweetened with sugar than when the same drink is prepared with non-caloric sweeteners (78, 79). This reduction occurs partly because of a faster gastric emptying when ingesting drinks sweetened with non-caloric than caloric sweeteners (78).

The specific effects of individual characteristics and conditions/states on the pharmacokinetics of alcohol are discussed in more detail in the sections below.

2.2. First-pass metabolism of alcohol: stomach vs. liver

Ingested alcohol undergoes first-pass metabolism (FPM), meaning that a portion of the alcohol is metabolized by enzymes in the gastrointestinal tract and the liver before entering the bloodstream (Figure 1). Blood from the mesenteric veins that supply the gastrointestinal tract is collected by the portal venous system, which, together with the hepatic artery, supplies the liver, allowing alcohol metabolism to occur before the blood returns to the heart. However, the primary site of alcohol’s FPM —whether in the stomach or the liver— has been the subject of ongoing debate. Some investigators consider that the primary site for FPM of alcohol is in the liver and that gastric FPM is negligible (80, 81); others postulate that the stomach is responsible for most FPM of alcohol (82–84).

Figure 1. Modulators of first-pass metabolism (FPM) and their impact on ethanol bioavailability.

The FPM refers to the portion of ingested alcohol that is metabolized by enzymes (alcohol dehydrogenase (ADH) Class I-V, cytochrome P450, and catalase) in the gastrointestinal tract and the liver before entering the bloodstream. FPM increases (top yellow box), and as a result, bioavailability decreases when alcohol is consumed with food or drugs that delay gastric emptying or during breastfeeding. Conversely, FPM decreases (bottom yellow box), and bioavailability increases, when alcohol is consumed with non-competitive inhibitors of gastric ADH4, after gastric surgeries, in females compared to males, or in people with alcohol use disorder (AUD). Created with BioRender.com.

First, the strongest evidence supporting the stomach as the main site of FPM was the identification of several classes of ADH in the human stomach, particularly the gastric ADH class IV (21, 22). This enzyme is active at a range of alcohol concentrations that are typically achieved after alcohol ingestion, which are in the molar range (85, 86). Notably, the administration of non-competitive inhibitors of stomach ADH, such as Cimetidine, increases alcohol bioavailability when alcohol is ingested but has no effect when the same dose of alcohol is administered intravenously (87, 88).

Second, the consumption of alcohol with food delays the gastric emptying rate and increases the exposure time to gastric ADH, which logically would result in increased stomach FPM (89). Third, drugs that decrease or increase gastric emptying cause a respective increase or decrease in FPM (90). Fourth, FPM is eliminated after gastrectomy or when alcohol is administered intraduodenally (91, 92). While these lines of evidence support the role of the stomach as the site of FPM, it should be noted that the elimination of FPM by gastrectomy and its modulation by food or drugs that alter gastric emptying does not conclusively prove that the liver is not the main site for FPM. The hepatic metabolism of ethanol exhibits saturation kinetics, which makes it extremely sensitive to variations in the gastric emptying rate (81, 93). Therefore, delayed gastric emptying would result in a slow arrival of alcohol to the liver, allowing for a more efficient hepatic FPM (93). While gastrectomy eliminates alcohol FPM (91), it is unclear whether this is due to the removal of the gastric source of FPM or the saturation of the hepatic source of FPM by the delivery of alcohol as a bolus (in the absence of the gastric pylorus).

Recent data from a clinical study that used sleeve gastrectomy (SG), a gastric surgical procedure that removes most of the stomach (80–85%) but preserves the pylorus, provides a novel approach to disentangle gastric emptying rate from hepatic versus gastric FPM (94). In this study, BAC-time profiles obtained from women post-SG were compared to those of non-operated control women who were matched for age and body composition and who achieved comparable time to peak BAC (tmax) after drinking the same alcohol dose on an empty stomach. Because the two groups were matched in tmax, group differences in alcohol bioavailability were unlikely due to differences in gastric emptying rates. In addition, the same participants were administered alcohol intravenously using an alcohol clamp. The alcohol clamp method enables the assessment of alcohol elimination rates (AER in g/hr) independent of variations in alcohol absorption (95). This approach controls for the potential effects of SG factors on systemic AER, which might otherwise confound comparisons of alcohol bioavailability between groups (95). Compared to the control group, the bioavailability of ingested alcohol increased by 34% in women post-SG. The increased bioavailability among women post-SG was not explained by differences in systemic AER (94). Taken together, these data further support the hypothesis that the primary site for alcohol FPM takes place in the stomach among women. Whether this is also the case for men who have undergone SG surgery is an active research area.

2.3. Alcohol distribution

The ethanol that escapes FPM is distributed by the vascular system and diffuses to and from tissues across the capillary bed. As a small, highly polarized water-soluble molecule, ethanol diffuses into tissues in proportion to their water content (96). Ethanol does not bind to plasma proteins; in fact, ethanol can be used as a tracer to estimate total body water (TBW) with a precision that is on par with that obtained with the dilution method with heavy water (the gold standard method to determine TBW) (97, 98). The movement of ethanol is typically described as a passive diffusion process, where the intercompartmental concentration gradient is proportional to the rate of change. How fast alcohol equilibrates, even if briefly, between blood, extravascular fluids, and tissues depends on the cross-sectional area of the local capillary bed and the blood flow per gram of tissue (6). Therefore, highly perfused organs, with a higher blood flow rate per gram of tissue, such as the lungs, kidneys, brain, and liver, equilibrate more rapidly with arterial BAC than lower perfused ones, such as the skeletal muscles.

As a result, very early in the absorption phase and prior to reaching equilibrium, BAC in arterial blood is higher than in venous blood, with the greatest arteriovenous difference being ~ 0.2 g/L within 10 minutes after drinking on an empty stomach (99). Eventually, the circulation and tissue concentrations equilibrate. However, alcohol continues to be metabolized. Eventually, the direction of diffusion flips, and the tissue water space becomes the “source” of alcohol in the bloodstream (100–102).

Because the distribution of alcohol into fat is almost negligible (97), the fat-to-lean tissue ratio plays a key role in the volume of distribution of alcohol (Vd). Therefore, women, who generally have a higher percentage of body fat than men, reach a higher peak BAC when both sexes receive the same intravenous dose of alcohol per kilogram of body weight (103). Similarly, with aging, people tend to lose muscle mass and gain fat, which helps explain, at least in part, why men above 60 years of age reach higher peak BAC when ingesting the same alcohol dose per kilogram of body weight than younger men (104).

3. Alcohol elimination

Once in the bloodstream, 94–98% of alcohol is eliminated by metabolism, and the remaining fraction is eliminated by excretion, which is unchanged (Figure 2). While the vast majority of ethanol metabolism occurs in the liver and its coordinated action with the gastrointestinal tract (105), alcohol is also metabolized in the brain. In both sites, ethanol metabolism is prioritized over other carbon sources to provide energy (90, 106). Although brain metabolism does not significantly contribute to the inter-individual variation in BAC following alcohol ingestion, it is clinically significant due to its acute effects on behavior, reward, and metabolic dyshomeostasis. Additionally, less than 1% of the systemic alcohol undergoes non-oxidative metabolism (107). Despite its small contribution, this section includes a discussion on non-oxidative pathways because the resulting metabolites are of clinical importance (108, 109).

Figure 2. Pathways of ethanol elimination in the human body.

Once in the bloodstream, the majority (94–98%) of ethanol undergoes oxidative metabolism in the liver. Acetate can be oxidized in the liver or be shuttled out of the hepatocyte as a fuel source for other tissues (heart, muscle, brain). Non-oxidative metabolism accounts for less than 1%, producing metabolites such as fatty acid ethyl esters (FAEEs), phosphatidylethanol (PEth), and ethyl glucuronide/sulfate (EtG/EtS) in various organs. A small fraction (2–6%) of ethanol is excreted unchanged via breath (1–3%), urine (1–3%), and sweat (<0.2%), with implications for biomarker and diagnostic applications. Created with BioRender.com.

3.1. Systemic metabolism of alcohol

3.1.1. Oxidative metabolism

The liver is the main site of systemic alcohol metabolism, where most alcohol is first oxidized to acetaldehyde and then to acetate (2). These reactions are primarily catalyzed by the enzymes ADH and ALDH (Figure 3). This oxidation process involves the reduction of nicotinamide adenine dinucleotide (NAD+) to NADH, causing a significant shift in the liver’s redox state that can contribute to various metabolic disorders, including hypoglycemia, hyperlactacidemia, and fat accumulation (110). The microsomal ethanol-oxidizing system (MEOS), along with catalase, also contributes to alcohol metabolism and related cytotoxicity under certain conditions (111). In what follows, we describe each of the four enzyme systems: ADH, MEOS, catalase, and ALDH.

Figure 3. Oxidative pathways of hepatic ethanol metabolism.

There are three enzymatic pathways involved in oxidative ethanol metabolism. A) Alcohol dehydrogenase (ADH) pathway, the primary pathway for hepatic ethanol metabolism: Ethanol is oxidized to acetaldehyde mainly by Class I ADH in the cytosol. Acetaldehyde is subsequently converted to acetate by aldehyde dehydrogenase 1 (ALDH1) in the cytosol or ALDH2 in the mitochondria. In the mitochondria, acetate is converted to acetyl-CoA by acetyl-CoA synthetase and then enters the tricarboxylic acid (TCA) cycle as acetyl-CoA. B) Microsomal ethanol oxidizing system (MEOS), a secondary pathway that plays a more important role after binge drinking, and its activity can be upregulated after weeks of heavy drinking: The enzyme Cytochrome P-450 2E1 (CYP2E1) oxidizes ethanol to acetaldehyde in the endoplasmic reticulum, generating reactive oxygen species as byproducts. (C) Catalase pathway, minimal contributor to hepatic alcohol metabolism: Catalase, located within peroxisomes, metabolizes ethanol to acetaldehyde using hydrogen peroxide (H₂O₂) as a cofactor. When mitochondrial acetyl-CoA is abundant, acetate can be shuttled out of the hepatocyte as a fuel source for other tissues. Excess acetyl-CoA exits the TCA as citrate via the solute carrier family 25 member 1 (SLC25A1). ACLY: ATP citrate lyase; ACOT12: Acety-CoA thioesterase 12. Created with BioRender.com.

ADH is the primary enzyme that metabolizes alcohol. The human ADH gene family clusters on chromosome 4q21 in a region spanning ~370 kilobases (112, 113). Currently, seven human ADHs have been identified, of which six have been structurally and functionally characterized. However, the seventh (ADH6) has not been isolated as a protein in vivo, and its role in ethanol metabolism is still unknown (113) (Table 1). ADH accounts for about 3% of soluble protein in the liver. The Class I ADHs, ADH1A, ADH1B, and ADH1C, are responsible for most of the oxidation of ethanol in the liver and can form both homo- and heterodimeric forms of the three subunits, such as αα, αβ, ββ, βγ, γγ, among others. ADH4 is exclusively expressed in the liver, while ADH7 is the only ADH not expressed there (113). ADH5 is ubiquitously expressed in human tissues but has a low affinity for ethanol, so it does not significantly contribute to hepatic ethanol oxidation. As shown in Table 1, these isozymes have different kinetic properties, allowing for flexibility in the regulation of alcohol metabolism at different exposures.

MEOS, which involves cytochrome P450 enzymes (CYP2E1) in the endoplasmic reticulum, has a lower affinity for alcohol compared to ADH, making it less involved in the metabolism of moderate alcohol exposures. However, its key enzyme, CYP2E1, is upregulated by chronic alcohol consumption, increasing its role in metabolizing larger alcohol doses (90). For instance, animals fed a liquid diet with 36% of calories from ethanol showed a 7.5-fold increase in liver microsomal CYP2E compared to nonethanol pair-fed controls. Similarly, CYP2E1 content was found to be 4 times higher in liver biopsies from men with an AUD than in those from men who reported abstaining from heavy alcohol consumption (114). This upregulation of CYP2E1 explains, in part, the metabolic tolerance to alcohol and the cross-tolerance to other drugs that are also substrates of CYP2E1, such as pentobarbital and diazepam, in people with AUD (90).

Ketones and fatty acids also induce CYP2E1 (90). Findings from rodent studies show that diet-induced obesity leads to increased microsomal ethanol oxidation (115), and clinical research suggests that obesity is associated with a higher alcohol elimination rate (116) and increased CYP2E1-mediated clearance of acetaminophen (117). As shown in Figure 3, CYP2E1 not only metabolizes ethanol to acetaldehyde but also generates reactive oxygen species and hydroxyethyl radicals, which contribute to oxidative stress and potential liver damage (90). These effects can be particularly harmful as they may synergistically promote the development of steatohepatitis when heavy drinking is combined with a high-fat diet (118).

Catalase, a peroxisomal enzyme, is a heme-containing antioxidant that converts two hydrogen peroxide (H₂O₂) molecules to water and oxygen and plays an important role in the maintenance of cellular redox homeostasis (119). However, the contribution of this pathway to the hepatic metabolism of ethanol is minimal. This is likely because the oxidation of ethanol by catalase depends on the presence of H₂O₂, which is of limited availability under physiological conditions (90).

ALDH rapidly and efficiently converts the acetaldehyde produced by the three previously described pathways to acetate. In humans, there are 19 ALDH isozymes, with different tissue distributions and kinetic properties. However, only ALDH1 and ALDH2 isozymes are involved in acetaldehyde oxidation (120) (Table 2). ALDH1, a cytosolic enzyme with relatively low kinetic activity for acetaldehyde, is found in nearly all tissues, including the brain. In contrast, ALDH2 is a mitochondrial enzyme highly expressed in the liver and gastrointestinal tract, exhibiting high catalytic efficiency for acetaldehyde oxidation (121). As a result, except for individuals with non-functional ALDH2 (i.e., ALDH2*2, discussed in the section on genetic polymorphisms), circulating acetaldehyde levels in the blood remain extremely low —over 1,000 times lower than BAC ) – as exemplified in a study where individuals with a BAC of approximately 7,000 μM (equivalent to 0.3 g/L or 0.03%) had blood acetaldehyde levels below 2.4 μM (122).

Table 2.

Human aldehyde dehydrogenase (ALDH)

ALDH Family	Official Gene Nomenclature	Former Gene Nomenclature	Enzyme Nomenclature	Km*	Vmax*	Tissue expression
1	ALDH1A1	ALDH1	ALDH1A1	180	380	Liver, cornea
	ALDH1B1	ALDH5	ALDH1B1	55	655	Liver, smooth muscle
2	ALDH2*1	ALDH2	ALDH2	0.2	280	Ubiquostly, highest in liver, gastrointestinal
	ALDH2*2	ALDH2	ALDH2		τ

The change of the names used for the ALDH genes in the literature creates some challenges in interpretation; therefore, in this table, we included both the official and the former gene nomenclature.

^*Km values represent acetaldehyde catalytic efficiency and are given in μM, and V_max values represent the maximal metabolic rate of the enzyme and are given in turnover numbers per minute. τ inactive under physiological conditions. Data from kinetics were extracted from (113); data from tissue expression were extracted from (151, 152).

Recent preclinical studies have questioned the traditional view that the liver metabolizes most systemic acetaldehyde (105, 123). Compared to mice with a global alcohol dehydrogenase 2 gene knockout (Aldh2), which exhibit remarkably higher blood acetaldehyde concentrations following acute oral alcohol administration than wild-type controls, mice with hepatocyte-specific Aldh2 deletion showed only about a 50% increase in systemic acetaldehyde (123). Follow-up studies revealed that the combined deletion of Aldh2 in both the liver and gut raised acetaldehyde concentrations to those measured in global Aldh2 knockouts after alcohol exposure (105).

Moreover, Fu et al. (105) discovered that approximately 30% of the acetaldehyde produced in the liver was excreted into the gastrointestinal tract via the bile, where gut epithelial ALDH2 further converted it to acetate. Notably, manipulating bile flow dynamics within the liver-gut loop in the mice not only played a key role in acetaldehyde metabolism but also influenced drinking behavior. While increasing bile flow caused faster acetaldehyde clearance and higher alcohol intake, inhibiting bile acid transporters led to slower acetaldehyde clearance and reduced alcohol consumption (105). Future clinical studies should investigate the importance of enterohepatic circulation in the clearance of systemic acetaldehyde and the potential effects of drugs that affect bile flow on alcohol consumption and toxicity.

ALDH irreversibly oxidizes acetaldehyde to acetate, most of which enters the bloodstream (124) to supply other tissues (Figure 3). Besides serving as a key intermediate in energy metabolism and cellular regulation (125), acetate also produces central depressant effects (see Section 3.2). However, a full discussion of acetate metabolism is beyond the scope of this article; readers can consult recent comprehensive reviews for additional detail (125, 126).

3.1.2. Non-oxidative metabolism of alcohol

In the liver, ethanol is conjugated by uridine diphosphate-glucuronosyltransferase and sulfotransferase enzymes to produce ethyl glucuronide (EtG) and ethyl sulfate (EtS), respectively (Figure 4). Although these metabolites represent less than 0.5% of the ingested alcohol dose, they are relatively stable and persist in body fluids and tissues. After moderate drinking, EtG and EtS remain detectable for ~ 25 hours in blood, several days in urine, and EtG can be detected for weeks to months in hair, making them sensitive and specific biomarkers for recent alcohol consumption (127, 128). Incidental exposure to ethanol (e.g., mouthwash, hand sanitizer) can yield positive urinary EtG and EtS results, so an appropriate analytical cutt-off and combining results from different biomarkers is essential (128).

Figure 4. Non-oxidative metabolism of alcohol.

Although non-oxidative metabolism accounts for less than 1% of ethanol metabolism, it generates biomarkers of recent alcohol intake. Ethanol is conjugated by sulfotransferase to produce ethyl sulfate (EtS), detectable in urine. Ethyl glucuronide (EtG), produced by UDP-glucuronosyltransferase from UDP glucuronic acid (UDPGA) is detectable in urine and EtG in hair. Fatty acid ethyl esters (FAEEs) are synthesized through multiple enzyme systems. FAEEs contribute to the acute damaging effects of alcohol binge drinking and can be detected in blood and hair. Phosphatidylethanol (PEth), produced by phospholipase D is a commonly used biomarker of recent alcohol drinking that is stable in dried whole blood. Created with BioRender.com.

Fatty acid ethyl esters (FAEEs) and phosphatidylethanol (PEth) are additional non-oxidative metabolites of ethanol that can serve as alcohol consumption biomarkers (129). FAEEs also shed light on potential mechanisms underlying alcohol-related tissue injury, particularly in cases of heavy drinking. Although FAEE-synthesizing enzymes are expressed in many organs, clinical studies show that the highest FAEE synthase activity and FAEEs concentrations are in organs commonly damaged by heavy alcohol consumption, such as the liver and pancreas. Lower but detectable levels are also observed in the heart and brain (130).

Based on these findings, Laposata and Lange (130) proposed that FAEEs, independent of acetaldehyde, contribute to or exacerbate alcohol-induced organ damage. Data from several pre-clinical studies support this hypothesis. First, arterial administration of FAAEs in rats induces a phenotype characteristic of acute pancreatitis (131). Second, promoting FAEE synthesis (by inhibiting the oxidative pathway) worsens symptoms of alcohol-associated acute pancreatitis (132) while inhibiting FAEE synthesis (through pharmacological inhibition of an FAEE synthase) alleviates these symptoms in experimental models, both in vitro and in vivo (133).

Recent preclinical studies further suggest that FAEE plays a major role in the acute damaging effects of alcohol binge drinking and extends findings from the pancreas to the liver (108). Using multiple lines of genetically modified mice, Park and collaborators (108) showed that, compared to wild-type control animals, acute liver injury was worsened in mice lacking ADH1. As expected, these mice exhibited markedly increased BAC after receiving a high dose of alcohol by gavage, and this increase was associated with more hepatic endoplasmic reticulum (ER) stress, hepatocyte apoptosis, and lipolysis. Interestingly, in the same binge-drinking model, ALDH2-deficient mice experienced similar liver injury to wild-type animals, despite elevated serum acetaldehyde concentrations (108). Further studies indicated that administering ethanol, but not acetaldehyde, intraperitoneally reproduced the acute liver injury observed with alcohol binge drinking. Additionally, deleting ADH1 led to increased FAEE concentrations after alcohol gavage, while deleting the gene for Ces1d, an FAEE synthetase, reduced the alcohol-induced rise in FAEE concentrations. These findings strongly suggest that alcohol and FAEEs, rather than acetaldehyde, drive acute liver injury (108).

PEth, another important non-oxidative metabolite, is generated in the cell membranes of most organs when phospholipase D catalyzes the transphosphatidylation of ethanol with phosphatidylcholine. There are 48 known homologs of PEth in human blood, all derived from phosphatidylcholine as the precursor. Due to its 100% specificity and prolonged detection window, PEth is one of the most commonly used biomarkers of recent alcohol drinking. For instance, a single standard drink can produce detectable PEth levels for up to 12 days (134). Additionally, the stability and detectability of PEth in dried whole blood stored on filter paper cards make it highly practical for clinical, research, and forensic applications (109).

3.2. Central metabolism of alcohol

3.2.1. Oxidative metabolism

The same three hepatic oxidative enzymes discussed above are also expressed in the brain. However, unlike in the liver, the primary enzyme responsible for the first oxidative step from ethanol to acetaldehyde in the brain is catalase, accounting for approximately 60% of ethanol metabolism (at least in rodents). This is followed by Cytochrome P-450 (CYP2E1), while ADH plays only a minor role (135) (Figure 5).

Figure 5. Brain metabolism of ethanol.

The oxidation of alcohol in the brain involves the same three pathways as those described for the liver. However, the catalase pathway is the most important pathway in the brain, followed by CYP2E1, and ADH plays only a minimal role. In astrocytes: Ethanol is mostly metabolized to acetaldehyde by catalase and subsequently to acetate by ALDH2. Acetate is incorporated into the tricarboxylic acid (TCA) cycle as acetyl-CoA, contributing to glutamate synthesis. Glutamate is converted into glutamine by glutamine synthetase (GS) and transferred to neurons. In neurons: Ethanol is mostly oxidized to acetaldehyde by CYP2E1 in the endoplasmic reticulum, followed by conversion to acetate by ALDH1 in the cytosol. Acetate enters the TCA cycle as acetyl-CoA, contributing to glutamate synthesis. Glutamate is further converted to gamma-aminobutyric acid (GABA) via glutamic acid decarboxylase (GAD), modulating inhibitory neurotransmission. Importantly, glutamine supplied by astrocytes can be converted back to glutamate and GABA in neurons. Recent findings from preclinical models suggest that cerebellar astrocytic ALDH2-generated acetate mediates alcohol-induced elevation of GABA concentrations and its associated motor impairment (139). Created with BioRender.com.

CYP2E1 activity in the brain, similar to the liver, is induced following chronic ethanol treatment, increasing its contribution to ethanol metabolism in both rodents (136, 137) and humans (138). In rats chronically exposed to alcohol, CYP2E1 is predominantly localized in neuronal cell bodies within regions such as the cortex, hippocampus, basal ganglia, hypothalamic nuclei, and reticular nuclei in the brainstem (136). Interestingly, autopsy studies of individuals with AUD indicate that heavy alcohol consumption is associated with neuronal damage in many of these same regions, including the thalamus, hippocampus, and periaqueductal grey area of the midbrain (136).

The second oxidative step in brain alcohol metabolism, converting acetaldehyde to acetate, is primarily carried out by mitochondrial ALDH2. In addition to the effects of ethanol itself, its metabolites—acetaldehyde and acetate—have distinct neurological effects that may contribute to alcohol’s pharmacologic actions (140, 141). Variations in ALDH2 expression across different brain regions may remarkably influence local concentrations of acetaldehyde and acetate, thereby impacting behavior (106, 139).

While controversial, some researchers propose that ethanol functions as a prodrug, with brain-generated acetaldehyde potentially driving alcohol’s reinforcing effects (142). Preclinical studies using gene-specific modifications in the ventral tegmental area (VTA), a key midbrain region regulating reward, motivation, learning, and memory, support this hypothesis. Reducing brain-generated acetaldehyde by silencing catalase or increasing ALDH2 activity in the VTA decreases alcohol-induced-reinforcing effects and reduces voluntary ethanol intake (141). Conversely, increasing acetaldehyde production by expressing the hepatic enzyme ADH1B1 in the VTA enhances alcohol reinforcement and facilitates the initial acquisition of voluntary ethanol intake (141).

Interestingly, manipulating acetaldehyde concentrations in the brain does not affect alcohol intake in animal models that have consumed alcohol chronically for one to two months, suggesting that alternative mechanisms sustain long-term alcohol intake (143, 144). Some studies have also implicated salsinolol, a condensation product of acetaldehyde and dopamine, in the central effects of alcohol (142). Although salsolinol has been identified in the human brain (145), technical limitations have hindered efforts to determine the contributions of brain-generated acetaldehyde and salsolinol to human alcohol consumption. Therefore, the roles of these metabolites in human alcohol intake remain unclear.

Acetate is primarily produced in the liver during alcohol metabolism (Figure 3), but is also generated within the brain. Systemic concentrations of acetate are around 0.2 mM, but increase to 0.5 – 1.0 mM following alcohol intake. When acetate enters the tricarboxylic acid (TCA) cycle, it can be oxidized to generate adenosine, a neuromodulator that, like alcohol, has central depressant effects (140). After alcohol consumption, astrocytes preferentially use acetate as an alternative fuel to glucose (146). Clinical neuroimaging studies indicate that both acute and chronic heavy alcohol drinking increase cerebral acetate uptake and lower brain glucose metabolism (147, 148), with the cerebellum showing the greatest shift toward acetate use (reviewed in (149)).

Disentangling the contributions of brain-derived versus peripheral alcohol metabolism to its pharmacological effects has been challenging due to the lack of specific in vivo tools. However, recent technical advances in preclinical models are furthering our understanding of alcohol brain metabolism. Recently, investigators developed an innovative device that allows in vivo microdialysis to simultaneously sample brain alcohol, acetaldehyde, and acetate while animals are ingesting alcohol (150). Demonstrating the in vivo ethanol brain metabolism, they show that acetate concentrations rise even in animals with a low alcohol intake. This technique will allow the investigation of brain region-specific metabolism and its associated behavioral effects.

Additional recent technical advances that provide insights into the impact of brain-derived alcohol metabolism are the generation of pre-clinical models that allow the investigation of cell-type-specific ALDH2 distributions in a brain-region-specific manner (139). These studies suggest that acetate produced by astrocytic ALDH2 in the cerebellum plays a key role in ethanol-induced GABA synthesis and motor impairment (139). By generating mice with astrocytic or hepatocytic ALDH2 deficiency, Jin and collaborators showed that acetate derived from central—rather than peripheral—ethanol metabolism primarily drove the synthesis of gamma-aminobutyric acid (GABA), the most common inhibitory neurotransmitter in the brain after low-dose ethanol administration. Notably, they also found that the cerebellum exhibited the highest levels of ALDH2 mRNA expression in both mice and humans, suggesting potential translational relevance (139).

Other studies have also demonstrated the brain-specific distribution of ALDH2 in mice, particularly in neurons within the prefrontal cortex (106). Deleting the Aldh2 gene, specifically in the forebrain neuronal lineage, reduced alcohol intake, leading to lower BAC in knockout mice compared to wild-type mice. However, despite the reduced BAC, acetaldehyde accumulated in the prefrontal cortex of these mice, impairing behavioral performance in tasks related to spatial learning, working memory, and prefrontal cortical functions deteriorated (106). These findings underscore the importance of understanding human brain alcohol metabolism, particularly how region-specific and cell-type specific enzyme activity and individual factors influence its effects, including alcohol toxicity. For example, data from the Human Protein Atlas indicate that astrocytes, but not neurons, express catalase, whereas neurons, but not astrocytes, express CYP2E1 in humans (151, 152).

3.2.2. Non-oxidative metabolism

Ethanol can also undergo non-oxidative metabolism in the brain (153). Two forms of fatty acid ethyl ester synthetase were found in at least 10 different regions of the human brain, with higher activity observed in gray matter compared to white matter. Studies by Laposata et al. (153) confirmed that these enzymes are active at concentrations of ethanol that are typically reached after social drinking. Additionally, post-mortem brain tissue studies have shown that FAEEs are present in the brains of individuals who were acutely intoxicated at the time of death but not in those who were sober (153). These findings suggest that non-oxidative ethanol metabolism in the brain may contribute to the biochemical effects of alcohol and could have implications for understanding alcohol-related neurotoxicity (107). Further research is needed to explore these mechanisms.

3.3. Alcohol excretion

A small fraction of alcohol, typically less than 5%, is excreted intact through sweat, urine, and breath (4, 32); Figure 2). One important application for the excretion of alcohol in the breath is that given the high blood flow of the lungs, end-expiratory breath alcohol concentration closely follows arterial blood alcohol concentration (see (100) as an example), which allowed the development of the “breathalyzer” to non-invasively estimate brain and blood alcohol concentrations. A second potential application of this pathway is to accelerate ethanol excretion in cases of severe intoxication. Results from an innovative proof-of-concept pilot study support this concept (154). Since ethanol elimination via exhaled air is proportional to BAC and minute ventilation, increasing minute ventilation—while maintaining constant carbon dioxide levels—could enhance ethanol excretion in cases of extreme intoxication. To test the hypothesis, the investigators used a device to facilitate isocapnic hyperventilation, a technique previously shown to speed up the elimination of vapor anesthetics. In this pilot study, participants consumed a moderate alcohol dose during two separate visits. Using a counterbalanced design, their BrAC was measured under normal breathing (control) and during isocapnic hyperventilation. As hypothesized, isocapnic hyperventilation increased the ethanol clearance rate threefold (154). Although preliminary, these findings highlight the potential of this method for treating acute severe alcohol intoxication and underscore the need for further research.

The excretion of intact alcohol through sweat and transdermal diffusion provides another source to estimate BAC using transdermal alcohol concentration (TAC). However, this method faces significant challenges, as TAC is delayed relative to BAC and varies widely due to individual population differences in sweat gland density, sweat production, and other physiological factors (155). Nonetheless, the development of non-invasive, precise, and durable alcohol biosensors is a promising step forward in health monitoring and precision medicine (155). These tools align with ongoing efforts in pharmacokinetic modeling to better predict BAC curves ─ the focus of the next section.

4. Pharmacokinetic models: Predicting BAC

Pharmacokinetic modeling for ethanol began in the 1930s with Widmark (156) and continues today to either describe, predict, or control alcohol exposures resulting from an orally or intravenously administered dosage. These attempts have taken two forms; phenomenological and physiological, both attempting to describe the absorption, distribution, and elimination of alcohol from the body (157). Phenomenological models describe the ethanol concentration time course with generic compartments, the number defined by fit to experimental data (157). Physiologically based pharmacokinetic models attempt to describe the same ethanol time course with compartments based on functional anatomy and physiology to the extent possible (157). The desire to accurately describe the alcohol or alcohol-related concentration time course in various phases of the absorption, distribution, metabolism, and excretion processes or in specific phases is reflected in the number of compartments (158). These efforts are further complicated by attempts to describe the pharmacokinetics of the population vs an individual (158). Consequently, numerous models have been developed, each with its own strengths, limitations, and specific applications, and a detailed discussion of each model is beyond the scope of this review.

Generally, single-compartment models assume one volume for alcohol distribution and elimination, such as the Widmark and Wilkinson models (156, 159). Dual-compartment models commonly include central and peripheral distribution volumes with one or more elimination pathways; for example, the model used by Norberg compares the distribution space of alcohol and water (98). Multi-compartment models have also been used to describe distribution spaces of interest, such as specific organs or absorption kinetics. Examples include Wedel et al.’s exploration of factors including sex, alcohol dose, exercise, and food (160) and more recent efforts by Zekan et al. that expand the Norberg model (161). Other notable work includes Podéus’s gastric emptying models and the interaction between alcohol, food, and alcohol metabolites, along with studies by Sadighi, Leggio, and Akhlagi focusing on modeling multiple organ tissue concentrations under fasted and fed conditions (162). Pharmacokinetic modeling, in general, is governed by both the nature of the drug itself and its interactions with human anatomy and physiology. The key challenge lies in selecting a model appropriate for the task at hand while balancing complexity and the risk of overfitting the data (158).

4.1. Measurements: Breath vs. Blood

An important consideration in alcohol pharmacokinetic modeling arises from what, where, when, and in what context to measure alcohol or alcohol metabolites after administration. The brain alcohol exposure time course is one common outcome of interest, as that trajectory is a key contributor to intoxication. Unfortunately, it is generally difficult to directly measure human brain alcohol concentration. However, as a low-volume but high blood flow organ, brain alcohol concentration closely follows arterial concentration (shown in human participants with MRS, (163)). Noteworthy, as described above, there are significant time-varying arterio-venous differences in BAC (99), but the sampling of arterial, in contrast to venous blood, is challenging and poses an increased risk to study participants. Fortunately, end-expiratory breath alcohol concentration closely follows arterial blood alcohol concentration (see (100) as an example), leading to the use of the breathalyzer as a practical measure to predict brain alcohol concentration.

An alternative strategy to breath alcohol concentration measurement is the use of venus capillary arterialization to obtain an “arterialized” line for blood sampling. An “arterialized” line is a validated technique that uses a heated hand box to draw blood from a hand vein warmed for at least 15 minutes, eliminating the need for arterial catheterization (164). As the hand temperature rises, increased blood flow shortens the transit time through the tissue, making BAC in warmed hand veins a reliable indicator of arterial BAC (165). This technique has been proven to be critical for the assessment of peak BAC reached in populations whose maximum brain exposure takes place soon after oral alcohol ingestion (i.e., when gastric emptying is accelerated), such as in patients who underwent gastric surgeries (166) (see section below).

Depending on the modeling application goal, the selection of arterial blood vs. venous blood vs. breath alcohol, as well as the timing and frequency of those measurements, become critical yet generally poorly understood factors.

4.2. Zero-order kinetics

From a modeling perspective, the elimination of ethanol at lower concentrations follows first-order elimination kinetics, where the elimination rate increases linearly with rising alcohol concentration. However, this phase is limited because ADH enzymes become saturated at a BAC of around 0.15 g/L (81), roughly the level achieved after drinking less than one standard drink. Once saturation occurs, elimination shifts to zero-order kinetics, meaning alcohol is removed at a constant rate regardless of BAC (4). In this phase, hepatic metabolism is limited by the intrinsic blood flow to the liver. Since alcohol delivery via blood flow to the liver exceeds the metabolic rate, BAC can remain high and increase disproportionately with dose, potentially leading to fatal alcohol poisoning.

4.3. Saturation kinetics

The type of kinetics described above for alcohol is known as Michaelis-Menten kinetics (Figure 6) and is represented by the following equation:

$$\frac{dC\left(t\right)}{dt}=\frac{V_{max}^C\left(t\right)}{K_{m\mathrm{ }}+\mathrm{ }C\left(t\right)}$$

Figure 6. Enzymatic activity for common isoforms of alcohol dehydrogenase (ADH).

Each curve represents the relationship between alcohol concentration and enzymatic activity for a specific ADH isoform: ADH1A, ADH1B*1, ADH1C*1, and ADH1C*2. For illustrative purposes, the standard form of Michaelis-Menten Kinetics was used here with Km and Vmax from (113) (also see Table 1).

The instantaneous concentration of alcohol at a time t is represented by C(t). Change in concentration over time is represented by dC(t)/dt. Vmax represents the maximum metabolism rate. Km is known as the Michaelis-Menten constant and represents the concentration of the drug where metabolism is ½ of the maximal rate. Figure 5 depicts the enzymatic activity of several primary ADH isoforms as a function of alcohol concentration. It is important to note that the enzymatic activity and concentrations are measured in vitro, and therefore, estimates of Km and Vmax can vary as a function of the experimental conditions. While these studies provide valuable simulations for comparing the kinetics of different isozymes under the same conditions, they represent a simplified model of the complex physiological processes involved in alcohol metabolism in vivo.

5. Determinants of BAC

In the following sections, we will examine how genetic polymorphisms in alcohol-metabolizing enzymes, biological sex, reproductive status, and clinical interventions —such as gastric surgeries and medications that alter alcohol metabolism—impact an individual’s risk for AUD and alcohol-related harm (Figure 7).

Figure 7. Determinants of blood alcohol concentration (BAC).

Key factors influencing the excursion of BAC following alcohol consumption: Enzyme polymorphisms: Genetic variants in alcohol-metabolizing enzymes (e.g., ADH and ALDH) affect the rate of ethanol and acetaldehyde metabolism. Fed vs. fasted state: Food intake slows gastric emptying, delays alcohol absorption, and increases first-pass metabolism and systemic alcohol elimination rate. Biological sex and body composition: Variations in body composition, ADH activity, and hormonal factors contribute to sex-related differences in alcohol pharmacokinetics. Reproductive status: States like pregnancy and lactation can alter alcohol’s volume of distribution and bioavailability. Gastric surgery: Procedures such as sleeve gastrectomy or Roux-en-Y gastric bypass can accelerate alcohol absorption and decrease FPM, thereby increasing peak BAC. Medications: Drugs can affect the absorption and metabolism of alcohol, and alcohol can alter the pharmacokinetics or pharmacodynamics of concurrently administered medications. Created with BioRender.com.

5.1. Genetic polymorphisms in alcohol metabolizing enzymes

Genetic polymorphisms in the genes encoding ADH and ALDH enzymes result in isoforms with varying catalytic efficiencies for ethanol and acetaldehyde ((2, 121); Table 1). This, in turn, leads to substantial variability in alcohol metabolic rates that contribute to inter-individual differences in alcohol pharmacokinetics, pharmacodynamics, and susceptibility to AUD and alcohol-induced diseases (113, 167, 168).

5.1.1. ADH polymorphisms

Single nucleotide polymorphisms (SNPs) have been identified at the ADH1B and ADH1C loci with translation of the resulting isoforms having different catalytic efficiencies for ethanol (113, 120) (Figure 6). Further, the ADH1B alleles appear with different frequencies in different racial groups (summarized in (120, 167)). The ADH1B*1 form predominates in Caucasian ancestry and African ancestry populations, and ADH1B*2 predominates in East Asian populations (e.g., Chinese, Japanese, Korean). It is also found in about 25% of Caucasians with Jewish ancestry. The ADH1B*3 form is found in about 25% of individuals of African ancestry. For the ADH1C polymorphisms, ADH1C*1 and ADH1C*2 appear with about equal frequency in Caucasian ancestry groups, but ADH1C*1 predominates in African ancestry and East Asian populations.

The best-known functional ADH variant is the rs1229984 polymorphism (Arg48His), associated with the ADH1B*2 allele, which produces an enzyme with higher catalytic activity for converting ethanol to acetaldehyde. The ADH1B*3 variant (rs2066702) has been shown to exhibit lower activity compared to ADH1B*2. ADH1C shows two major functional variants, ADH1C*1 (rs1693482) and ADH1C*2 (rs698), that encode for isoforms that differ in in vitro activity, with the ADH1C*1 showing 1.5- to 2-fold greater activity than ADH1C*2 (113). There are additional genetic variations identified in non-coding regions of ADH genes, such as ADH4, that have been associated with altered gene expression as well as AUD risk, however, the effect of these variations on ADH activity and overall metabolism remains to be established (113, 169, 170).

Genetic variation in ADH7 has been characterized, and attempts have been made to associate these haplotype-based variations with alcohol pharmacokinetics (171). However, more work is needed to fully understand how these variations may impact gastrointestinal metabolism and exposure following alcohol consumption.

5.1.2. ALDH polymorphisms

The best-known ALDH genetic polymorphism is in ALDH2, with the variant form, ALDH2*2 (rs671) encoding for a low activity isoform that has a 100-fold higher Km for acetaldehyde, resulting in a nearly inactive enzyme under physiological conditions after drinking (113). This very prominent variant allele has been seen in about half of the East Asian populations studied (including the Han Chinese, Taiwanese, and Japanese) but has not been observed in populations of Caucasian ancestry (120, 172, 173).

5.1.3. Impact of ADH and ALDH polymorphisms on alcohol metabolism

Functional polymorphisms of genes for the alcohol metabolizing enzymes ADH and ALDH2, and differences in the prevalence of the polymorphic alleles in different ethnic populations, have resulted in several studies examining racial and ethnic differences in alcohol metabolism and the influence of ADH1B, ADH1C, and ALDH2 genotypes. In general, the ADH1B*2 polymorphism has been associated with an increase in alcohol metabolic rates. This has been most clearly seen in populations of Jewish ancestry, reflecting a more direct genetic effect (174). The effect is less clear in Asian ancestry populations that carry the ADH1B*2 variation, as a significant proportion also carry the ALDH2*2 polymorphism, which has a more profound effect on alcohol (and acetaldehyde) metabolic rates (discussed further below) (112, 175). Studies of the effect of ADH1B*3 polymorphism have been less consistent. An early study in individuals of Black/African American ancestry showed a higher alcohol disappearance rate (mg%/hr) for ADH1B*3 allele carriers (176), although a separate study, also conducted in an African American sample, did not replicate this effect (177).

Genetic polymorphisms in ALDH2 greatly influence the metabolism of acetaldehyde and, consequently, the individual’s response to alcohol and its toxic effects. The ALDH2*2 polymorphism results in a deficient ALDH2 isoform with a very low capacity for acetaldehyde metabolism. This leads to acetaldehyde accumulation and a condition known as the “alcohol flush reaction”—facial flushing, nausea, and discomfort—which discourages heavy drinking and is thought to have a protective effect against AUD. In fact, research has consistently demonstrated that individuals homozygous or heterozygous for ALDH2*2 have markedly higher acetaldehyde levels and reduced risk for AUD (122, 172, 178–180), although they also have increased cancer risk (181, 182). There have been no specific studies examining the effect of ALDH1A1 on alcohol metabolism in humans, likely due to its lower catalytic efficiency compared to ALDH2. However, it is possible that the combined effect of ALDH1A1 and ALDH2 polymorphisms may lead to considerable variability in alcohol metabolism, affecting tolerance, dependence, and alcohol-related diseases across individuals and populations.

5.1.4. Other genes involved in alcohol metabolism: CYP2E1 and rare variants

There are several CYP2E1 polymorphisms, and while their impact on alcohol metabolism has not been studied as much as ADH and ALDH variants, they are implicated in variability in alcohol metabolism and alcohol-induced liver injury (183).

Recent research has shed light on the role of rare genetic variants in alcohol metabolism (184). Through large-scale sequencing studies, Leger et al., identified a gene network associated with alcohol consumption, including both common variants, such as those in the ADH and ALDH family, and rare variants in genes, such as CYP, UDP-Glycosyltransferase (UGT), and two sulfotransferases (SULT). These later two genes are involved in non-oxidative ethanol metabolism and also contribute to variations in alcohol consumption behaviors. This study highlights the importance of rare genetic variants in understanding the genetic architecture of alcohol metabolism, expanding the scope of research in potential pharmacological targets beyond common variants in ADH and ALDH.

5.2. Fed vs. fasted

The food-induced lowering of BAC has been recognized for over a century (185, 186), yet the precise mechanisms underlying this effect remain only partially understood. As discussed earlier, food decreases alcohol bioavailability partly by slowing alcohol absorption, which increases FPM. Beyond these effects, food also significantly enhances the systemic elimination of alcohol (2). Studies have shown that compared to drinking in the fasted state, consuming alcohol after a meal (187–189) or even eating a few hours after drinking (189) increases alcohol elimination rate (AER) by 25–50%. Notably, food increases the AER even when alcohol is administered intravenously, demonstrating that this effect is independent of food’s impact on absorption and FPM (2, 103, 190).

Researchers have identified ingredients in food that could increase alcohol metabolism (191–194). For example, using a within-subject design, Lundquist and Wolthers (191) found that consuming isocaloric fructose, but not glucose, two hours after alcohol ingestion increased AER by 50%. Similarly, Soterakis and Iber (192) showed that the ingestion of fructose increased the elimination rate of intravenously administered alcohol by 32% compared to isocaloric glucose. However, attempts to use fructose in clinical settings for treating acute alcohol intoxication have had mixed results. While some replicated the fructose effect (195), others found no clinical benefit and noted potential harm, including elevated serum uric acid and lactate levels (196).

In line with a “fructose effect”, Rogers et al. used an intravenously administered alcohol clamp and found that a high-carb meal, but not an equicaloric high-fat or high-protein meal, increased AER (190). However, Ramchandani et al., also using an intravenous alcohol clamp, reported that eating isocaloric meals high in carbohydrates, fats, or proteins was equally effective at increasing AER, with all meals increasing AER by 45%, compared to a 12-hour fast (2). The discrepancy between the findings of Rogers et al. and Ramchandani et al. remains unclear but may be related to methodological differences, such as variance in fructose content in the carbohydrate portions of the different meals (i.e., a potential threshold for the fructose effect?).

Probable mechanisms underlying the food-induced AER increase include enhanced NADH reoxidation (197) and increased ADH activity (198), as suggested by preclinical models, as well as clinical evidence indicating that protein-and lipid-rich meals increase portal liver blood flow (199). Meanwhile, the search for “sobering” products continues, with current options including ADH and catalase enzymes, though their effectiveness has shown mixed results (e.g., (200, 201)). While the specific compounds and mechanisms behind the food-induced lowering of BAC remain unclear, one conclusion is certain: drinking on an empty stomach should be avoided.

5.3. Biological sex and body composition

Females are more susceptible to alcohol-related liver disease at lower levels of alcohol intake compared to males, which may partly result from sex-related differences in alcohol metabolism (202). Females generally exhibit higher peak BAC than males after consuming the same number of standard drinks or the same amount of alcohol per kilogram of body weight (6, 72, 203, 204). Two key factors likely explain this difference. First, females typically have a lower lean-mass-to-fat ratio than males, so alcohol is distributed in a smaller body-water volume (72, 203). Second, although results are mixed (see (92)), several studies report a reduced alcohol FPM in females, in part due to lower activity of ADH4 (6, 72, 203, 204). Some inconsistencies across studies may arise, in part, from the alcohol concentration used: the sex difference in FPM is evident with 10% (wine-strength) and 40% (spirit-strength) drinks, but not with 5% (beer-strength) beverages (72).

Additionally, several studies suggest that females eliminate systemic alcohol faster than males (205). Because most ethanol is metabolized in the liver, Kwo and collaborators (1998) investigated potential sex-related differences in liver volume and their relationship to AER. Using an intravenous alcohol clamp and radiological computerized tomography, they found no sex differences in liver volume or AER per unit liver volume, with both males and females eliminating, on average, 7.3 grams of ethanol per hour. However, because females typically have a larger liver volume relative to their lean body mass than males, they exhibited greater ethanol clearance per unit of lean body mass (206). A follow-up study using similar methods, which also included younger and older groups of males and females, found that regardless of the age group, males and females had similar AER per unit of liver volume (207). This study also replicated the finding that females, compared to males, have a larger liver volume relative to lean body mass. However, females in this sample had smaller absolute liver volumes and slower AER than males, highlighting the influence of liver volume as a source of variability in AER (207).

Expanding these findings, a recent study explored the relationship between body composition and AER in females, also using the intravenous alcohol clamp (116). The study found that females with obesity had a 52% faster AER than those with normal weight (116) (For an example of AER for individuals with different BMI, see Figure 8). The investigators controlled for these factors and indeed found that most of the association between obesity and AER was dependent on fat-free mass (FFM), consistent with earlier observations that individuals with obesity tend to have more FFM than those of the same sex, age, and height without obesity (208). Specifically, age and FFM together accounted for 72% of the individual variation in AER. Given that FFM strongly predicts lean liver volume— the functional portion of the liver involved in drug metabolism— (207) the dependence of AER on FFM likely reflects the robust relationship between FFM and liver volume, a major determinant of AER (206, 207, 209).

Figure 8. Impact of body composition on alcohol elimination rates (AER) using the breath alcohol clamping technique.

This illustration depicts the breath alcohol clamping technique, which uses intravenous alcohol infusions to achieve and maintain breath alcohol concentration (BrAC) at a target level for prolonged periods of time (2). The infusion rate is determined using a physiologically based pharmacokinetic model for alcohol that incorporates individualized estimates of model parameters based on each participant’s estimated total body water. BrAC measurements are obtained using a breath analyzer to provide feedback, ensuring that BrACs remain within 0.05g/L of the target. The feedback enables infusion rate adjustment to correct for parameter estimation errors and experimental variability. The top panels show the infusion rate versus time while the bottom panels depict BrAC versus time for two female participants: one with overweight (left) and one with obesity (right). The shaded regions represent the time periods used to calculate the AER, which provides an estimate of alcohol elimination independent of absorption variability. The dashed line indicates the target BrAC during the clamp (0.6 g/L). The data presented are from two participants in Seyesadjadi et al. (116). Vector elements for this figure were adapted from a licensed version of Adobe Stock (stock.adobe.com)

In addition to differences in body composition, sex-related differences in alcohol pharmacokinetics may also be influenced by reproductive hormones. Although research in this area remains limited, that sex hormones affect the activity of alcohol-metabolizing enzymes is suggested by the finding that castration in male rats reduced the in vivo rate of ADH degradation, leading to increases in ADH protein content and ADH activity to levels that are comparable to those of female rats (210). Similarly, patients treated for prostatic metastatic carcinoma with orchiectomy exhibit an increase in AER within one month after surgery (211). Additionally, dihydrotestosterone administration decreases ADH activity in hepatocytes in vitro and decreases AER in healthy males who received the drug over a two-week period (212).

Among women, investigation of the effects of the menstrual cycle on alcohol pharmacokinetics yielded mixed results (213). One of the best controlled published studies on this topic found no evidence that changes in sex steroids during the menstrual cycle influence alcohol pharmacokinetics (214). Mumenthaler and collaborators used a within-subject design in 24 females who consumed the same dose of alcohol during the menstrual and luteal phases of their cycle, and estradiol and progesterone were measured in blood for each study day. Despite remarkable differences in estradiol and progesterone concentrations in the menstrual and luteal phases, there were no pharmacokinetic differences. However, given the bidirectional effect between the acute effect of alcohol and reproductive hormones and its importance in health (e.g., breast cancer; (215, 216)), more research in this area is needed.

5.4. Reproductive states of pregnancy and lactation

The reproductive states of pregnancy and lactation constitute one of the greatest metabolic challenges experienced by women (217, 218). Overall, the increased plasma volume and lower ratio of lean muscle to adipose tissue during pregnancy and lactation can result in changes in the volume of drug distribution for drugs, including alcohol. Pregnancy and lactation are anabolic states that are orchestrated via hormonal and metabolic changes that redirect nutrients to specialized maternal tissues characteristic of reproduction, such as the placenta and mammary glands and, in turn, the developing fetus or infant, respectively (219). Lactation is the result of highly synchronized endocrine and neuroendocrine processes that begin during late pregnancy to prepare both the body and brain for motherhood (220).

Despite the pronounced metabolic changes (221), there is a paucity of experimental and clinical research on pregnant and lactating women in drug metabolism (222–224), hence their designation as therapeutic orphans (37). Indeed, most research has focused on the health risks for the nursing infant, not the mother (225). During pregnancy, alcohol ingested by the mother, along with its toxic metabolites, can cross the placental barrier into the fetal bloodstream and amniotic fluid, which the fetus swallows. The fetus can excrete alcohol via a) placental clearance back to the maternal blood flow and b) urinary excretion into the amniotic fluid or oxidate it (although at a much smaller fraction than the mother) by ADH and ALDH (226).

While we are not aware of research on the effects of the reproductive state of pregnancy on alcohol metabolism in women, per se, there is some evidence from a population-based, case-control study in Norway that variation in alcohol metabolism during pregnancy, as indicated by the ADH1C haplotypes of both the mother and child, is linked to an increased risk of oral cleft malformations in infants (227). Specifically, babies whose mothers engaged in heavy, intense drinking (consuming five or more drinks in one sitting) during the first trimester of pregnancy were more likely to develop oral clefts. The increased risk was only evident when either the mother or the baby carried the ADH1C haplotype associated with reduced alcohol-to-acetaldehyde metabolism, pointing to ethanol (more than acetaldehyde) as the teratogen for this specific birth defect.

Albeit limited, more research has focused on the effects of the lactational state on alcohol metabolism (225). The alcohol ingested by lactating women is transferred to human milk but is not stored in the breast tissue. Alcohol concentrations in breast milk parallel maternal bloodstream concentrations (40). In 1993, da-Silva (228) showed that BAC peaked later and that the mean area under the BAC-time curve (AUC), an indicator of systemic availability of the drug, was significantly smaller in lactating women when compared to a control group. However, the control group was comprised of some women who were tested during their postpartum period and others for whom the parity was not reported.

To isolate the effects of lactation on alcohol pharmacokinetics from other physiological changes associated with pregnancy and parturition (229–231), BAC of lactating women who were exclusively breastfeeding (Breastfeeding Group) were compared to a group of women whose babies were the same age but whose babies were exclusively formula fed (i.e., similar in parity but controls for recency of parturition in the absence of lactation), and a group of nulliparous women (i.e., controls for parity in the absence of lactation). All groups had equivalent age, BMI, and alcohol consumption patterns and were assessed after drinking a moderate dose of alcohol on two separate days: once after a meal and once on an empty stomach. As expected, food-delayed peak BAC, decreased alcohol bioavailability (i.e., smaller AUCs), and sped up the alcohol elimination rate in all groups, but overall BACs and AUCs were lower in the lactating group when compared to the two control groups, which did not differ from each other (Figure 9). That reproductive state of lactation did not affect the elimination of alcohol suggests that the act of lactating reduces alcohol bioavailability by increasing FPM of alcohol. To determine whether breast stimulation, which is necessary for sustaining lactation and impacts vagal activity and the release of hormones that affect gastric motility (e.g., oxytocin, cholecystokinin, gastrin) (232), we compared BAC in lactating women who were randomized to breast pump either before or after drinking a moderate dose of alcohol; both groups were tested both fasted and fed conditions (230). Women who breast pumped before drinking exhibited a reduced bioavailability of ethanol compared to those who pumped after drinking, particularly when assessed in the fed condition. Interestingly, women who pumped after drinking eliminated alcohol faster. Overall, these findings suggest that food intake and breast stimulation in lactating women have additive effects that enhance alcohol FPM.

Figure 9. Breath alcohol concentration (BrAC) and area under the BrAC-time curve (AUC) for lactating and non-lactating control women after drinking alcohol under fed and fasted conditions.

Left panel: BrAC time profiles of nulliparous women, formula feeding mothers, and breastfeeding mothers (averaged for both fed and fasted conditions). For the fed condition, a standardized breakfast was provided one hour before drinking. At time zero, participants consumed 0.4 g/kg of alcohol-containing beverage over 10 minutes. *P<0.05 Breastfeeding mothers vs. formula feeding mothers and nulliparous women at a given time point. Right panel: BrAC AUCs for each group, comparing fasted and fed states.^a different from ^b at P<0.05. For both graphs, data are mean ± SEM. Graph modified from (231).

5.5. Gastric Surgeries

The importance of including gastric surgeries in a review on alcohol pharmacokinetics in humans is twofold. First, due to its clinical impact: only certain gastric procedures are associated with increased AUD risk (233–236). Since the speed at which a psychotropic substance reaches the brain influences its addiction potential (237), comparing how different stomach surgeries affect alcohol pharmacokinetics can iluminate mechanism that underly risk of postoperative AUD.

As described earlier, gastrectomy was originaly used as a model to attempt to pinpoint the site of alcohol FPM. In 1989, Caballeria and collaborators demonstrated that total of partial gastrectomy, performed for stomach cancer or peptic ulcer disease, virtually eliminated FPM. After oral dosing, men reached BAC to those seen with intravenous administration (91). The fact that these patients were at higher risk of developing AUD after undergoing gastrectomy (238–241) suggests a biological mechanism linking stomach resection with alcohol misuse that is independent of patients having severe obesity, and challenges the common interpretation that AUD post-bariatric surgery is simply a case of “addiction transfer” (242).

Bariatric surgical techniques have since evolved, and each reshapes the gastrointestinal system, and potentially alcohol pharmacokinetics in distinct ways. Sleeve gastrectomy (SG) now accounts for 72% of US bariatric surgeries (243). It removes about 80–85% of the stomach but preserves the pylorus (Figure 10). Roux-en-Y gastric bypass surgery (RYGB), the second most common procedure at 21%, creates a small gastric pouch (~30 ml) connected to the distal jejunum. Laparoscopic gastric banding (LAGB), once common (44% in 2008), now represents less than 1% of surgeries due to poor long-term outcomes. In LAGB, a small adjustable silicone band is placed around the upper stomach. This creates a narrowed passage between the upper and lower parts of the stomach and limits how much food a person can eat at one time.

Figure 10. Impact of gastric surgeries on blood alcohol concentration (BAC).

Left panel: Illustrations of three gastric surgeries—Laparoscopic Gastric Banding (LAGB), Roux-en-Y Gastric Bypass (RYGB), and Sleeve Gastrectomy (SG)—depicting their anatomical modifications. Top right panel: BAC time profiles following alcohol consumption for individuals who underwent LAGB (white symbols), RYGB (black symbols), or SG (cyan symbols). Bottom right panel: BAC time profiles for individuals pre-surgery (white symbols) versus post-surgery for RYGB (black symbols) and SG (cyan symbols). Data are mean ± SEM. Graphs modified from (166, 244). Created with BioRender.com.

The first clear evidence that gastric surgery can accelerate alcohol absorption came in 2002, when Klockhoff and collaborators found that RYGB caused women to reach ~ 28% higher peak BACs—within 5 minutes, compared to 25 minutes in non-operated women matched for age and BMI (245). Later studies confirmed even larger increases (up to 100%) when arterialized blood (rather than venous blood) was sampled (246, 247). In contrast, LAGB leaves anatomy intact and does not alter alcohol pharmacokinetics (244, 248). Therefore, the higher AUD incidence after RYGB compared to LAGB, has a mechanistic basis (233, 235).

Evidence for SG emerged more slowly. Early studies that estimated BAC from breath were inconsistent. (248–250). The uncertainty that SG affected alcohol pharmacokineticswhich led some to consider SG safer than RYGB in terms of AUD (250).

One key issue with using a breathalyzer to estimate BAC is the protocol’s recommendation for a lag period of about 15 minutes from the end of drinking to the first breath sample to avoid contamination with alcohol in the oral mucosa. However, because peak BAC can occur within 10 minutes of drinking after gastric surgeries, BrAC cannot reliably estimate peak BAC in this clinical population (166). Studies that used arterialized venous blood samples and measured BAC using the gold-standard technique of headspace gas chromatography showed that SG was also associated with faster and higher peak BACs ((166, 244); Figure 10).

While it is well established that gastric surgeries speed up alcohol absorption and minimize alcohol FPM (244, 247, 251, 252), their effect on systemic alcohol elimination is less clear. Oral-dose studies hinted at a slower alcohol clearance post-surgery (166), but the differences disappeared when adjusted for body weight. An intravenous alcohol clamp to isolate alcohol elimination from absorption processes show that bariatric surgery had no independent effect on alcohol elimination rate (94).

Finally, it is interesting to note that RYGB and SG are effective partly because they dramatically raise endogenous glucagon-like peptide 1 (GLP-1), improving glycemic control and weight loss (253). Given the current enthusiasm for GLP-1 receptor agonists to treat alcohol misuse (reviewed in (254)), it seems paradoxical that surgeries that elevate GLP-1 also elevate AUD risk. The risk, however, follows a temporal pattern: 40–50% of patients with high-risk alcohol use before surgery reduced drinking within the first postoperative year (233, 255–257), but AUD incidence rises sharply two or more years later (233, 258, 259). This delay points to evolving neuroadaptations that, together with altered alcohol pharmacokinetics and pharmacodynamics, ultimately heightened vulnerability.

In summary, although RYGB and SG result in different anatomical changes to the gastrointestinal system, both reduce alcohol FPM and can double peak BACs without affecting systemic AER. The findings that SG and RYGB similarly affect alcohol pharmacokinetics align with the results from several (236, 259, 260) but not all (261, 262) studies, suggesting a comparable increase in AUD prevalence following both procedures. Beyond AUD, altered alcohol handling has wider clinical implications: patients with a history of metabolic surgery who are hospitalized for alcohol-associated liver disease (ALD) decompensate earlier and have higher cumulative mortality, despite being, on average, 12 years younger, than hospitalized patients with ALD who have no history of metabolic surgery (263).

5.6. Alcohol-drug interactions

Drugs can affect the absorption and metabolism of alcohol, potentially leading to higher BAC and an increased risk of adverse events (264). Conversely, alcohol can alter the pharmacokinetics or pharmacodynamics of concurrently administered medications. In the US, nearly 40% of adults take at least one medication that could interact negatively with alcohol, highlighting the importance of discussing alcohol use when prescribing drugs (265). In the following section, we include the mechanisms of some of the most common alcohol-drug interactions and highlight some medications that pose risks when combined with alcohol (for a more in-depth assessment of this topic, see a recent review in (266). We end this section with drugs that target the metabolism of alcohol with the goal of treating AUD.

5.6.1. Increased risk of adverse events

Benzodiazepines:

Benzodiazepines are commonly prescribed for anxiety, insomnia, seizures, muscle spasms, and alcohol withdrawal syndrome. Their metabolism primarily involves the CYP450 enzymes, mainly CYP3A4 and CYP2C19, and glucuronidation via UGT enzymes (or only glucuronidation) (267). Accordingly, due to competition for these metabolic pathways, it is likely that when consumed together, ethanol would raise blood concentrations of benzodiazepines. Several studies suggest that ethanol mostly affects the pharmacokinetics of benzodiazepines metabolized by CP450 but not those metabolized by glucuronidation alone. For example, the co-administration of alcohol with triazolam or diazepam increased AUC by 21–27% (268, 269). However, the pharmacokinetics of oxazepam, temazepam, and lorazepam remain largely unaffected (266). Nevertheless, even when pharmacokinetics is unaffected, the concomitant ingestion of alcohol and benzodiazepine can have synergistic sedative effects, suggesting that pharmacodynamic effects are at play (270, 271).

Barbiturates:

Barbiturates, such as phenobarbital, are a class of drugs known as central nervous system depressants. Before the discovery of benzodiazepines, they were used to treat anxiety and sleep disorders. Because of their high addiction potential and risk of overdose, they have been largely replaced by benzodiazepines; although they are still prescribed for surgeries or to treat seizure disorders. Barbiturates are metabolized by CYP2E1, so concurrent ethanol use competes for this enzyme, slowing barbiturate clearance (272), and raising plasma drug concentrations. However, clinical studies have also documented cross-tolerance between ethanol and barbituric such thatethanol clearance is faster after barbiturate administration (272). In addition, there are pharmacodynamically mediated additive sedative effects that heighten the risk of respiratory depression and overdose when these two drugs are combined (273).

Antidepressants:

Antidepressants include many different classes of drugs, such as selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors (SNRIs), tricyclic antidepressants (TCAs), and monoamine oxidase inhibitors (MAOIs). Several studies suggest that alcohol reduces antidepressant efficacy and exacerbates CNS side effects, though the pharmacokinetic interactions between alcohol and these drugs are not fully understood (266).

The dose of alcohol administered in combination with the drugs or the use of alcohol by the participants in the studies (e.g., moderate versus heavy users) can modify the findings in the results. For example, at a high BAC (~0.8–1.0 g/L), alcohol inhibited the hepatic first-pass metabolism of amitriptyline (TCA), which markedly increased amitriptyline plasma concentrations (274). However, a study that used a more moderate alcohol exposure (BAC~ 0.3 g/L) found that alcohol had no effect on amitriptyline’s pharmacokinetics (275) (but delayed the time of absorption of trazodone). The pharmacokinetics of bupropion, which is extensively metabolized by the CYP450 (CYP2B6), was not affected by moderate (~ 0.3–0.4 g/L) nor by high (~0.8 g/L) BACs (276). Most antidepressants are metabolized by CP2D6, but alcohol neither induces nor inhibits hepatic clearance via this specific enzyme (277). Interestingly, clinical studies show that CYP2D6 (and CYP2E1) is also expressed in the brain and is upregulated in people with AUD (137, 278). This suggests that some alcohol-drug interaction may arise from altered drug metabolism in the brain without affecting systemic drug concentrations.

Opioids:

Opioids are widely prescribed for pain management. Many opioids are metabolized primarily by CYP450 enzymes (e.g., codeine, fentanyl, and oxycodone) or UGT enzymes (e.g., hydromorphone and morphine) (279). One study showed that oxycodone, primarily metabolized by CYP3A4, combined with ethanol (0.3 or 0.6 g/kg), resulted in a significant reduction in peak BAC compared to ethanol alone (280). However, studies of morphine and hydromorphone, both metabolized primarily by UGT, did not affect alcohol metabolism (281, 282). In another study, ethanol (BAC of 0.1%) was associated with an additional 19% decrease in respiratory rate in subjects taking oxycodone (283). A recent meta-analysis found that those who consumed alcohol achieved higher maximal concentations (Cmax) and shorter time to Cmax (Tmax) with hydromorphone compared to placebo, although this was not observed with other opioids, including morphine (266). These findings show how variable the pharmacokinetic interactions can be between opioids and ethanol depending on the drug used. However, combining any opioid with alcohol can cause severe respiratory depression and death (284). Stringent clinical precautions must be taken when prescribing opioids to patients who consume alcohol, especially given the high potential for misuse associated with opioids.

Antibiotics:

Antibiotics include a wide range of medications with many metabolic pathways for the treatment of bacterial infections. A 2020 systematic review evaluated alcohol interactions with common antibiotics, highlighting risks such as reduced drug efficacy, liver toxicity, and disulfiram-like reactions (285). Studies have found that ethanol coadministration with erythromycin, a CYP3A4 substrate, and inhibitor, increases the drug’s lag time by 136% while decreasing Cmax by 15% and AUC by 27%, suggesting significant impacts on absorption (286, 287). Other antibiotics like isoniazid and rifampin carry a heightened risk of hepatotoxicity when combined with alcohol (285). Isoniazid is metabolized by N-acetyltransferase 2 (NAT2) and can inhibit CYP3A4 while inducing CYP2E1, although does not appear to have a pharmacokinetic interaction with ethanol (0.73 g/kg) (288, 289). Other common antibiotics, including ceftriaxone and trimethoprim-sulfamethoxazole, are associated with disulfiram-like reactions, producing symptoms such as flushing, nausea, and tachycardia (285). While consuming alcohol with antibiotics appears to carry great risks, evidence of alcohol’s impact on antibiotic pharmacokinetics remains inconsistent.

Non-steroidal anti-inflammatory drugs (NSAIDs):

NSAIDs are common prescription and over-the-counter (OTC) medications used for pain. Ibuprofen, metabolized by CYP2C9, does not show any significant pharmacokinetic interactions with ethanol (290). Aspirin, however, inhibits ADH activity, resulting in higher BACs, exacerbating mucosal damage (291). Despite the variability in PK interactions, NSAIDs and alcohol both impair gastric mucosal integrity. As little as one drink per day has been shown to raise the risk of GI bleeding by 37% among NSAID users (292).

Acetaminophen:

Acetaminophen is another common OTC medication used for pain. Chronic alcohol consumption induces CYP2E1, increasing the formation of the hepatotoxic acetaminophen metabolite NAPQI (293). Oneta et al. demonstrated that acute ethanol ingestion increased NAPQI formation by 20% in heavy drinkers, emphasizing the danger of acetaminophen use in this population (294). Though acute alcohol use temporarily inhibits CYP2E1, excessive alcohol consumption shortly before acetaminophen use still increases liver damage risk (295).

5.6.2. Therapeutic agents targeting ethanol metabolism

A few medications that act directly on ethanol metabolism and have been US FDA-approved or are under investigation for AUD or other alcohol are summarized here. Drugs that treat AUD through other pathways are not discussed; readers can find detailed reviews elsewhere (296–298).

Disulfiram:

Approved in 1951 as the first US FDA-authorized medication for AUD, disulfiram irreversibly inhibits ALDH1 and ALDH2. The resulting acetaldehyde buildup provokes an aversive reaction (flushing, nausea, tachycardia) whenever alcohol is consumed. Although effective with supervised dosing, it is no longer first-line therapy due to poor adherence, variable efficacy, and hepatotoxicity concerns (168, 296).

Cyanamide (calcium carbimide):

This drug is used in parts of Europe, Canada, and Japan, and it also inhibits ALDH via its active metabolite, nitroxyl (299). It has a shorter half-life than disulfiram but carries a higher risk of liver injury and is thus not frequently prescribed, and it is not US FDA-approved.

Metadoxine:

Approved in some European and Latin American countries, but not in the US (168), metadoxine is used to treat acute ethanol intoxication because it speeds ethanol clearance (300). The mechanism of action of this drug is incompletely understood, but it is likely through increased ALDH activity. Preliminary clinical trials also indicate that metadoxine may curb drinking and help maintain abstinence (301, 302). These early findings underscore the need for larger, double-randomized studies to establish its efficacy in AUD.

Fomepizole (4-methylpyrazole).

This is an FDA-approved drug for the treatment of acute ethylene glycol or methanol poisoning. Fomepizole works as an antidote by competitively inhibiting ADH (303). Interestingly, in ethanol-preferring rats, fomepizole reduces alcohol intake (304), hinting at possible therapeutic value for heavy drinking. However, ADH inhibition also diverts ethanol toward non-oxidative pathways, which are implicated in pancreatitis and hepatic damage (see above). This toxicity risk would need careful evaluation before fomepizole or similar ADH inhibitors could be considered as an AUD treatment.

Other ALDH2 approaches:

Many investigational drugs showed promise in pre-clinical studies, including selective and reversible ALDH2 inhibitors (e.g., ANS-6637) and genetic targeting of liver ALDH2 messenger RNA to locally reduce hepatic ALDH2 expression (305, 306). However, a Phase 2 clinical study evaluating ANS-6637 in treatment-seeking adults was terminated early due to concerns of liver toxicity (307), and efficacy results for a clinical trial using a hepatic ALDH2 mRNA inhibitor (DCR-AUD) are not yet publicly available (NCT05845398). Overall, targeting ethanol metabolism offers promise but requires caution due to risks from metabolic intermediates and non-target effects, and further research is needed.

6. Conclusions and remaining questions in the field of alcohol metabolism

This comprehensive review provided an update on the state of the science related to the evolutionary, biological, cultural, and clinical aspects of alcohol metabolism and its broader implications for human health. Despite advances in the field, critical gaps in understanding remain.

The influence of alcohol metabolism on consumption patterns and alcohol-related harm has long been recognized, particularly through studies on genetic polymorphisms, such as ALDH2 variants prevalent in individuals of Asian ancestry. As discussed earlier, these variants significantly increased blood acetaldehyde concentrations after alcohol consumption, leading to heightened aversive effects and reduced risk for AUD. In fact, the very first drug approved for the treatment of AUD was disulfiram, a drug that inhibits ALDH2 and, therefore, capitalizes on the resultant aversive effects of acetaldehyde accumulation after alcohol intake (308). However, with a few exceptions, research on alcohol metabolism has predominantly focused on peripheral processes, particularly hepatic metabolism. The recent observations from rodent studies by Fu and colleagues (105) highlight the need to better understand the coordinated action of the gut-liver axis on alcohol detoxification and ingestive behavior in humans. For example, clinical studies investigating how drugs that modulate bile secretion affect alcohol and acetaldehyde metabolism and alcohol subjective effects would help understand novel mechanisms impacting alcohol consumption in humans. A better understanding of human alcohol pharmacokinetics could help design innovative approaches that involve the modulation of ethanol metabolism for treating AUD (168).

While controversial, a growing number of reports suggest that endogenous ethanol production through fermentation in the microbiome contributes to liver disease and other signs and symptoms of alcohol intoxication. The auto-brewery syndrome is an understudied area, and it is unclear whether the disease is rare or whether it is rarely diagnosed. More studies on the impact of alcohol on the human microbiome, and how this may contribute to conditions like MASLD are needed.

Recent advancements in understanding brain alcohol metabolism opens new avenues for understanding individual differences in alcohol’s subjective effects and behavioral consequences. The discovery of regional and cell-specific distributions of alcohol-metabolizing enzymes —and the impact of their manipulation on motor and cognitive functions— presents groundbreaking opportunities for novel therapeutic targets. Further research could determine the behavioral impact of alcohol metabolism in other brain-specific regions, the effect of different patterns of alcohol intake on this local metabolism, and the timing of developmental alcohol exposure, such as drinking in youth or in utero exposure. Mechanistic studies in humans are challenging, but neuroimaging studies and in vitro human-banked tissue studies of specific brain regions could be instrumental in demonstrating translational value. In addition, ethanol alters the pharmacokinetics and pharmacodynamics of numerous medications, raising concerns about increased toxicity and impaired efficacy. Could some of the pharmacodynamic interactions between alcohol and other drugs be mediated by common pharmacokinetic processes at the central level?

The findings that Roux-en-Y gastric bypass and sleeve gastrectomy significantly alter alcohol pharmacokinetics, increasing peak BAC and potentially elevating the risk of AUD and alcohol-associated liver disease underscore the importance of tailored patient counseling and monitoring. Noteworthy, the American Academy of Pediatrics recently endorsed metabolic surgeries to treat severe obesity for adolescents aged 13 years and older (310). Around the same time, the Teen-Longitudinal Assessment of Bariatric Surgery (Teen-LABS) study reported that 47% of adolescents who underwent such surgeries screened positive for potential hazardous drinking or AUD symptoms eight years later (309). While metabolic surgeries are effective for severe obesity, it is critical to ensure that healthcare professionals, patients, and families understand the profound impact on alcohol metabolism post-surgery.

Addressing these gaps in knowledge by conducting translational research and targeted clinical interventions could lead to novel evidence-based strategies that mitigate alcohol-related harm, optimize therapeutic approaches for AUD, and enhance patient care.

Clinical Highlights.

Recent advances in alcohol metabolism research provide new insights into alcohol’s complex effects on behavior, health, and clinical outcomes. We review here how, beyond its primary metabolism in the liver and the gut, alcohol is also metabolized in the brain, where the locally generated metabolites, acetaldehyde and acetate, influence reward pathways, motor coordination, and sedation. Regional differences in enzyme activity, particularly in an isozyme of aldehyde dehydrogenase (ALDH2) in the cerebellum and prefrontal cortex, offer new explanations for individual variability in alcohol’s behavioral effects and long-term consumption patterns. Additionally, endogenous alcohol production by gut microbes is increasingly linked to metabolic dysfunction-associated steatotic liver disease, suggesting a broader impact on health than previously recognized. We also discuss how metabolic surgeries, such as Roux-en-Y gastric bypass and sleeve gastrectomy, can double peak blood alcohol concentrations due to faster absorption and reduced first-pass metabolism, significantly increasing the risk of alcohol use disorder and alcohol-related harm— an effect that warrants increased monitoring of alcohol use post-surgery, particularly in pediatric patients. Overall, this review highlights the importance of understanding alcohol’s metabolic pathways and integrating this knowledge into clinical practice to improve risk assessment, patient counseling, and tailored interventions.