What Alcohol Actually Does Inside Your Liver

Ethanol is handled in two steps, and the middle one is the problem. ADH in the cytosol turns it into acetaldehyde — far more toxic than the alcohol itself, and an IARC Group 1 carcinogen. ALDH2 in the mitochondria has to clear it fast. Both steps burn NAD⁺ → NADH, and that one shift — watch the balance beam tip — is what stops fat burning, stops gluconeogenesis, dumps pyruvate into lactate, and raises uric acid. Then there is the part nobody wants: ADH is already saturated, so alcohol leaves at a flat rate of about one drink an hour. Mash the ☕ coffee button as much as you like and watch the BAC line refuse to bend.

Try this: run Binge (4 drinks fast) and hammer ☕ drink coffee — the alertness bar jumps every time, and the BAC line does not move at all. Then switch to ALDH2 deficiency and watch the red acetaldehyde pile up until the pool is full.

The Science in Plain Language

What a “standard drink” actually is — and why your pour is bigger than you think

Every number on this page is measured in standard drinks, and a standard drink is a precise thing. In the United States it is 14 grams — 0.6 fluid ounces — of pure ethanol. That works out to a 12 oz (355 mL) beer at 5 %, a 5 oz (148 mL) glass of wine at 12 %, or a 1.5 oz (44 mL) shot of 40 % spirits. Those three contain the same amount of alcohol. Nothing else about them matters to your liver.

Now look at what people actually pour. A 16 oz pint of a 7 % craft IPA is about 2.4 standard drinks. A generous 9 oz glass of a 14 % red is about 1.8. A home-poured gin and tonic is routinely a double. Someone who says “I had three drinks” has quite often had six. This is not a character flaw; it is a measurement problem, and it is the single most common reason people are shocked by a breathalyser.

Why it matters here: the liver clears roughly one standard drink per hour. If your pour is double, your clock is double. Everything downstream — the acetaldehyde, the NADH shift, the hours until you are safe to drive — scales with the real number of grams, not with the number of glasses.

Two steps, one villain

Ethanol is dismantled in two moves. First, alcohol dehydrogenase (ADH), sitting in the cytosol of the liver cell, strips two hydrogens off and turns ethanol into acetaldehyde. Second, aldehyde dehydrogenase 2 (ALDH2), inside the mitochondria, turns acetaldehyde into acetate. Acetate becomes acetyl-CoA, drops into the TCA cycle, and leaves as carbon dioxide and water. Two enzymes, one intermediate.

That intermediate is the whole story. Acetaldehyde is far more reactive and far more toxic than the ethanol it came from. It is an aldehyde, which means it grabs onto things: it forms adducts with proteins, disabling them, and it forms adducts with DNA, which is how a molecule stops being merely unpleasant and starts being a carcinogen. The International Agency for Research on Cancer (IARC) classifies acetaldehyde associated with the consumption of alcoholic beverages as a Group 1 carcinogen — the same tier as tobacco smoke and asbestos. Alcoholic beverages themselves are also Group 1.

In most people this is survivable because ALDH2 is astonishingly good at its job. Its affinity for acetaldehyde is very high (a Km in the low micromolar range), so it hoovers up acetaldehyde almost as fast as ADH can make it. That is why a normal drinker’s blood acetaldehyde stays at only about 2–5 µmol/L even in the middle of a night out. The bottleneck holds. In the animation, that is why the red pool stays thin on the default scenario — and why it fills up alarmingly the moment you break ALDH2.

The master switch: NADH:NAD⁺

Here is the part that most people have never been told, and it is the part that explains almost everything else.

Both steps — ADH and ALDH2 — work by handing a hydride to NAD⁺, converting it to NADH. NAD⁺ is the cell’s oxidising currency, and dozens of essential reactions are queuing for it at any moment. When you drink, the liver oxidises ethanol at close to its maximum rate for hours, and the NADH:NAD⁺ ratio collapses. Free NAD⁺ becomes scarce. That single change cascades:

1. Fat oxidation stops. β-oxidation needs NAD⁺ (at 3-hydroxyacyl-CoA dehydrogenase, among others). With NAD⁺ unavailable, fatty acids arriving at the mitochondrion simply cannot be burned. They are re-esterified into triglyceride and stored. This is why fatty liver appears within days of heavy drinking — not years. And it is why it is reversible: stop drinking, NAD⁺ is regenerated, β-oxidation restarts, and the fat comes back off over weeks.
2. Gluconeogenesis stops. Making new glucose also needs NAD⁺ (at glyceraldehyde-3-phosphate dehydrogenase), and the lactate-to-pyruvate step that feeds it is running backwards. If you have glycogen in the tank you are fine — glycogenolysis covers you. If you do not (you have not eaten, or you have been drinking for a day), there is nothing left to make glucose from and blood sugar falls. Alcohol-induced hypoglycaemia is real, is dangerous, and is easily mistaken for simple drunkenness.
3. Pyruvate is dumped into lactate. Lactate dehydrogenase runs in reverse to regenerate NAD⁺ — the cell is desperate. Blood lactate rises. And lactate competes with urate for excretion at the URAT1 transporter in the kidney, so urate is retained instead of being passed. Uric acid rises, and that is the actual mechanism behind the beer-and-gout link. (Purines in beer add to it, but the lactate effect is the big one, and it happens with any alcohol.)
4. De novo lipogenesis switches on. All that acetate arrives as acetyl-CoA, and abundant NADH provides the reducing power. The liver starts building brand-new fat. So it is not only that fat cannot be burned — more fat is being made.

You can actually measure the shift indirectly: the lactate:pyruvate ratio in blood tracks the cytosolic NADH:NAD⁺ ratio. It normally sits around 10:1. Ethanol pushes it substantially higher. The beam on the diagram is a normalised picture of the same thing.

Zero-order kinetics: the liver’s clock cannot be hurried

ADH’s affinity for ethanol is far higher than the blood alcohol levels you actually reach. In practice this means that from about 0.01 % BAC upward, ADH is already flat out. It is saturated. Doubling the ethanol in your blood does not double the rate at which it is removed.

So elimination is not a decaying curve like most drugs. It is essentially a straight line: about 0.015 % BAC per hour, with an individual range of roughly 0.010–0.020 %/h. In practical terms, about one standard drink per hour. Look at the BAC trace on the animation: after the peak, it comes down in a ruler-straight line. That is not an artistic choice. That is what zero-order kinetics looks like.

Which means: nothing speeds it up. Not coffee. Not a cold shower. Not going for a run. Not chugging water. Not making yourself sick. Not a fry-up. Not fresh air, not a nap, not standing up straight and concentrating. There is no known safe intervention that increases the rate at which a human liver oxidises ethanol.

Food is the one that deserves a careful answer, because it half-works and people over-read it. Eating slows absorption — it keeps alcohol in the stomach, flattens the peak, and means you reach a lower maximum BAC from the same drinks. That is genuinely useful and genuinely protective. But it does not speed elimination. The area under the curve is what it is. Once the alcohol is in you, the time to zero is set by the dose and by your ADH, not by what was on the plate.

And caffeine is the most dangerous of the myths, which is why there is a button for it. Caffeine blocks adenosine receptors in your brain. It changes how awake you feel. It does not touch a liver enzyme; it cannot. Press the coffee button as many times as you like and watch the alertness bar leap while the BAC line carries on falling at exactly the same slope. A wide-awake drunk is still drunk — and is arguably more dangerous than a sleepy one, because they now believe they are fine and get in the car.

The only variable you control is the dose. After that, it is the clock.

CYP2E1, tolerance, ROS — and the paracetamol trap

ADH is not the only route. There is a second pathway, the microsomal ethanol-oxidising system (MEOS), built around the enzyme CYP2E1 in the smooth endoplasmic reticulum. At normal drinking levels it handles only a small share. But chronic drinking induces it — the liver makes a lot more CYP2E1 — and it becomes a serious contributor.

Two things follow, one that feels good and one that does not.

The one that feels good: ethanol is cleared faster. A heavy drinker really does eliminate alcohol more quickly than a novice, on the order of 0.020–0.025 %/h instead of 0.015. This is a large part of what “tolerance” actually is at the metabolic level. (The rest is neuro-adaptation in the brain, which is a separate and equally important story.)

The ones that do not: CYP2E1 runs on NADPH and oxygen rather than NAD⁺, and it is a leaky enzyme. It sheds reactive oxygen species, which drive lipid peroxidation and inflammation, and it steadily draws down the cell’s glutathione. Watch the green tank on the “Chronic” scenario: it starts lower and it keeps falling.

And now the trap. CYP2E1 is also the enzyme that converts paracetamol (acetaminophen) into NAPQI, a viciously reactive quinone imine. In a normal liver, a therapeutic dose produces a trickle of NAPQI and glutathione neutralises it instantly — you never notice. In a heavy drinker, both defences fail at the same time: CYP2E1 is induced, so NAPQI is made faster and in greater quantity; and glutathione is already depleted, so there is nothing to mop it up. Free NAPQI binds to mitochondrial proteins and the cell dies.

That is why a paracetamol dose that is entirely safe for most people can cause fulminant liver failure in someone who drinks heavily. Press the paracetamol button on the Chronic scenario and watch the pink NAPQI overwhelm an already-empty tank. The antidote is NAC (N-acetylcysteine), which supplies cysteine and refills the glutathione tank — and it works far better the earlier it is given.

Practically: if you drink heavily, treat paracetamol with genuine respect. Stay well under the labelled maximum, never stack doses, and never combine it with a binge. And do not simply swap to ibuprofen or naproxen instead — NSAIDs carry their own gastrointestinal bleeding risk with alcohol, which is a different way to end up in hospital.

The flush is a warning, not a quirk

ALDH2*2 is a single amino-acid substitution (Glu504Lys) in the ALDH2 enzyme. It leaves the enzyme with a small fraction of its normal activity. Worse, ALDH2 works as a tetramer — four subunits together — so a single bad copy poisons the whole complex. That means people with one variant copy are already strongly affected; they do not get a comfortable half-measure.

Roughly 30–40 % of people of East Asian ancestry carry it. That is hundreds of millions of people, and it is far and away the most common clinically significant enzyme variant in human alcohol metabolism.

What happens is exactly what the animation shows. ADH keeps making acetaldehyde at full speed — that step is untouched — and almost nothing clears it. Blood acetaldehyde climbs to something like 30–80 µmol/L, many times a normal metaboliser’s 2–5. Acetaldehyde is a vasodilator and a mast-cell provoker: the face and chest go red, the heart races, the head aches, nausea follows. This is the “Asian flush”, and it is usually treated as an embarrassing party trick.

It is not. Here is the hard part, said plainly: people who flush and drink anyway carry a substantially elevated risk of oesophageal squamous-cell carcinoma. This is one of the best-established gene–environment interactions in all of cancer epidemiology — the flushing phenotype plus regular drinking multiplies risk in a way that neither does alone. The mechanism is not mysterious: a Group 1 carcinogen is sitting at high concentration in the tissues that the drink passes through.

The flush is your body telling you that acetaldehyde is accumulating. It is a warning light on the dashboard. And this is why the popular workaround is so bad: H2 blockers and antihistamines (famotidine and the like) can blunt the visible redness — and they do nothing to the acetaldehyde. You have painted over the warning light while the engine is still burning. If you flush, the honest advice is the unwelcome one: drink much less, or not at all.

Fatty liver, then hepatitis, then cirrhosis — and what comes back

Alcohol-related liver disease is usually described as three stages. It is worth knowing which of them you can undo.

1. Steatosis (fatty liver). Appears within days of heavy drinking, and happens in the large majority of heavy drinkers. Defined as fat in more than about 5 % of the liver. Almost always silent — no symptoms, sometimes a slightly enlarged liver, sometimes nothing but a raised GGT. It is fully reversible. Stop drinking and it typically resolves over a few weeks. This stage is a genuine second chance and most people never know they were given one.

2. Alcoholic hepatitis. Now the liver is inflamed and cells are dying. Jaundice, a tender liver, fever, malaise. It ranges from mild and biochemical to a severe episode with real, substantial mortality. It is largely reversible if drinking stops — but a severe attack is a medical emergency, not something to ride out.

3. Cirrhosis. Scar tissue has replaced working liver. This one is not reversible. Progression can be halted and function can partly recover with complete abstinence — that is genuinely worth doing and it changes survival — but the architecture does not grow back. Roughly 10–20 % of long-term heavy drinkers reach it. Whether you are in that fraction depends on dose, on years, on sex (women progress at lower intakes), on obesity, on genetics, and on whether you also carry hepatitis C.

Two lab clues worth recognising. In alcohol-related liver injury, AST is usually higher than ALT, classically a ratio above 2:1, and both are usually only moderately raised — typically under about 300–400 U/L. That is very different from paracetamol injury, where ALT can run into the thousands. GGT rises with sustained heavy drinking and is a useful, if non-specific, flag. None of these is a diagnosis on its own, and a normal set of liver enzymes does not prove a liver is healthy.

The point of this section is not to frighten anyone. It is the opposite. The first two stages come back. The liver is one of the few organs that genuinely forgives, and stopping is the entire treatment — there is no supplement, no cleanse, no milk-thistle protocol that substitutes for it.

Thiamine before glucose — and the other things alcohol strips out

Thiamine (vitamin B1) is the one that saves lives, and the reason is worth understanding properly. Alcohol attacks thiamine from three directions at once: it impairs its absorption in the intestine, it reduces the liver’s ability to store it, and it interferes with its conversion into the active coenzyme. Body stores only last a matter of weeks.

Thiamine is the coenzyme for pyruvate dehydrogenase and α-ketoglutarate dehydrogenase — the enzymes that let a cell actually use glucose. So here is the trap that every emergency department knows: if you give glucose to a thiamine-depleted brain, the glucose consumes what little thiamine is left, and you can precipitate Wernicke’s encephalopathy — confusion, eye-movement paralysis, ataxia — which, untreated, can become the permanent amnesia of Korsakoff’s syndrome.

That is why alcohol-related admissions get thiamine first, and glucose second. It is not a ritual. It is a real, standard, order-of-operations decision that prevents irreversible brain injury, and it is one of the most quietly heroic bits of routine medicine there is.

The rest of the depletion list is less dramatic but still worth knowing:

• Folate (B9) — absorption impaired and urinary losses increased. Low folate causes macrocytic anaemia (big, ineffective red cells), and it also sits right at the junction between alcohol and cancer risk, because folate feeds the one-carbon pathway that repairs and methylates DNA.
• Vitamin B6 — acetaldehyde physically displaces the active form (pyridoxal-5-phosphate) from its carrier protein and accelerates its breakdown. This is a direct, mechanical consequence of the villain molecule.
• Magnesium — alcohol causes renal magnesium wasting. Low magnesium worsens tremor, cardiac arrhythmia, and the seizure risk in withdrawal. It is one of the most commonly missed deficiencies in heavy drinkers.
• Zinc — low in most people with significant liver disease. Contributes to loss of taste, poor wound healing, weakened immunity, and possibly to the leaky gut that drives endotoxin into the portal blood and fuels the inflammation.

The hangover, red wine, and the one line worth keeping

The hangover is not one thing, and this is why nothing fixes all of it. It is at least five things happening together: residual acetaldehyde and the inflammatory cytokine response to it; dehydration, because ethanol suppresses vasopressin (antidiuretic hormone) so you pass out more water than you drank; genuinely wrecked sleep, because alcohol is sedating early and then rebounds — REM is suppressed in the first half of the night and sleep fragments in the second; low blood sugar, from the gluconeogenesis block described above; and gut and immune activation.

Congeners — the non-ethanol compounds that give dark drinks their colour and character — make it measurably worse. Bourbon, brandy, dark rum and red wine produce reliably worse hangovers than vodka or gin at the same ethanol dose. That is a real, replicated finding, not folklore.

Water and electrolytes genuinely help the dehydration component. Food helps the glucose component. Nothing touches the acetaldehyde component except time, and no commercial “hangover cure” has ever been shown to change the rate at which your liver works — because, as above, nothing can.

And now “red wine is good for your heart.” This deserves a straight answer rather than a wink.

The idea came from observational studies that repeatedly found moderate drinkers had less heart disease than abstainers. The problem is the comparison group. Abstainers are contaminated with “sick quitters” — people who stopped drinking because they became ill — which makes non-drinkers look unhealthy for reasons that have nothing to do with alcohol. Moderate drinkers also tend to be wealthier, better educated, more physically active and better fed. The J-shaped curve had a lot of confounding baked into it.

So researchers asked the question a different way, using Mendelian randomisation. Genetic variants in ADH1B and ALDH2 alter how much alcohol a person drinks over a lifetime, and they are assigned essentially at random at conception — before income, diet, exercise or illness can confound anything. When lifetime alcohol exposure is instrumented that way, the apparent cardiovascular protection largely disappears, while blood pressure and stroke risk rise steadily across the range of intake. The honest summary of the current evidence is that there is no established cardioprotective dose of alcohol.

This is not moralising, and it is not a lecture. Plenty of people will read that and pour a glass anyway, and that is entirely their business. But you are entitled to the actual state of the evidence before you decide, rather than a marketing line from the 1990s. The polyphenols in red wine are perfectly real — you can get them from grapes, berries, dark chocolate and olive oil without the ethanol attached.

The one line worth keeping is this: the dose and the pattern are what matter, and the liver’s clock cannot be hurried. One drink an hour, spaced out, with food, is a fundamentally different chemical exposure from six drinks in two hours — even when the weekly total is identical, because the peak acetaldehyde, the peak BAC and the depth of the NADH shift are all set by the rate. And whatever the number ends up being, the only thing on earth that clears it is time.

↑ Back to the animation

Connections

• All Interactive Visualizations
• How Your Liver Actually Detoxifies
• Blood Sugar & Insulin
• Inside the Mitochondrion
• Alcoholic Hepatitis
• Cirrhosis
• Non-Alcoholic Fatty Liver Disease
• Acetaminophen (Paracetamol) Overdose
• Gout
• Liver Function Tests (ALT, AST)
• GGT
• Uric Acid
• Thiamine (Vitamin B1)
• Vitamin B6
• Folate (Vitamin B9)
• Glutathione
• NAC (N-Acetylcysteine)
• Magnesium
• Zinc