Ethanol: Acute Intoxication and the Withdrawal Spectrum
Ethanol is the oldest and commonest poison in the emergency department — and the most easily underestimated. On the way in, it slows the brain toward coma; if the drinker is chronic and the bottle is suddenly taken away, the very same brain rebounds into a storm that can kill. Acute intoxication and withdrawal are two ends of one adaptation, and the same molecule sits behind both. The danger is rarely the alcohol you can smell; it is the head injury, the low sugar, or the second drug you assumed away because the patient was "just drunk."
A 52-year-old man is brought in at 2 a.m., found slumped against a wall, reeking of alcohol, barely rousable. The easy diagnosis writes itself: drunk, let him sleep it off. But the resident who examines him is uneasy. His breathing is slow and shallow, his GCS is dropping, and there is a boggy swelling above one ear. A bedside glucose reads 2.4 mmol/L. This is not "just drunk": he has a subdural haematoma, he is profoundly hypoglycaemic, and his airway is at risk. He gets glucose — but only after thiamine — a definitive airway, and a CT head. Two beds down, a different man, sober now for thirty-six hours after his last drink, is drenched in sweat, pulse racing, seeing insects crawl the walls. One ward, two faces of the same drug: one sinking under alcohol, one convulsing from its absence.
Acute intoxication: two receptors, one sedated brain
Alcohol is a general CNS depressant, and it works by pushing the brain's two master switches in the same direction. Ethanol enhances the brain's main inhibitory system, GABA — it potentiates the GABA-A receptor, the same chloride channel that benzodiazepines and barbiturates act on, letting inhibition flood in. At the same time it blocks the main excitatory system, inhibiting the NMDA subtype of glutamate receptor. Turn the brakes up and the accelerator down at once, and the whole cortex slows. This is why alcohol shares a receptor family with the sedatives covered in the Central Nervous System chapter — and why combining it with benzodiazepines or opioids is so dangerous: they stack on the same GABA-A brake and drive respiration into the ground.
The clinical picture climbs a dose-dependent staircase downward: euphoria and disinhibition, then slurred speech and ataxia, then drowsiness and vomiting (with a real risk of aspiration in an obtunded patient), and finally respiratory depression, coma and death. There is no single "lethal level" — a chronic drinker may be walking and talking at a concentration that would render a naïve teenager comatose — but the trajectory is the same, and the airway is what kills first. The core of management is embarrassingly simple and often forgotten: protect the airway, support breathing, and watch. Ethanol has no antidote; the liver clears it while you keep the patient safe.
Think of ethanol elimination like a car park with a fixed number of exit gates. At low blood levels the gates are half-empty, so the more cars (alcohol molecules) arrive, the faster they leave — clearance rises with concentration (first-order). But drink enough and every gate is jammed with a car waiting to exit; adding more cars doesn't speed anything up. Now the exit rate is fixed — a steady number of cars per hour no matter how many are queuing (zero-order). This is why the last drinks "hang on" so long: the enzyme, alcohol dehydrogenase, is saturated, and blood alcohol falls by a roughly constant amount each hour rather than halving.
The metabolic traps: sugar and ketones
Ethanol metabolism doesn't just sedate — it hijacks the liver's redox balance, flooding it with NADH. That surplus shuts down gluconeogenesis (the liver's ability to manufacture new glucose) and diverts pyruvate away from making sugar. In someone with depleted glycogen stores — a malnourished chronic drinker, or above all a small child who has swallowed alcohol — the result is dangerous hypoglycaemia. Alcohol-induced hypoglycaemia in a child is a classic, lethal, and easily missed emergency: check a glucose in every intoxicated patient, and never attribute a reduced consciousness level to alcohol until the sugar is known.
A hungry drinker who keeps drinking can tip into a ketoacidosis that mimics diabetes. Alcoholic ketoacidosis is the second metabolic trap. A chronic drinker on a binge, eating little and vomiting, has no carbohydrate coming in and a liver forced by that NADH excess toward ketone production. They present with a raised-anion-gap metabolic acidosis, ketones, and vomiting — but, unlike diabetic ketoacidosis, the glucose is usually normal or low, not high. The treatment is not insulin; it is fluids and glucose (which switches off ketone production), alongside the thiamine every malnourished drinker needs. Recognising it hinges on not reflexively calling every ketoacidosis "diabetic."
The most dangerous phrase in the resus bay is "he's just drunk." Alcohol is a great imitator and a greater concealer. A depressed conscious level in an intoxicated patient may be the alcohol — or a head injury the drink caused by a fall, hypoglycaemia, an infection, a post-ictal state, or a co-ingested drug (opioids, benzodiazepines, a toxic alcohol). The blood alcohol level never explains everything on its own. Treat the intoxicated patient as a puzzle, not a punchline: glucose, a full neurological look, a low threshold for CT head, and a hunt for the second poison.
- Ethanol is a CNS depressant: it enhances GABA-A and inhibits NMDA glutamate receptors.
- At higher levels kinetics are zero-order (saturable) — blood alcohol falls a fixed amount per hour, not by halving.
- Progression is dose-dependent: disinhibition → ataxia → drowsiness/vomiting → respiratory depression and coma.
- Check glucose in every intoxicated patient — hypoglycaemia is common in children and the malnourished.
- Alcoholic ketoacidosis: raised-anion-gap acidosis with normal/low glucose — treat with fluids, glucose and thiamine, not insulin.
- There is no antidote for ethanol — care is supportive: airway, breathing, glucose, thiamine, and hunting the co-ingestant.
Withdrawal: the same adaptation, thrown into reverse
The brain that has drunk for years is not the brain it started with. Chronic alcohol keeps the GABA brake permanently pressed and the NMDA accelerator permanently blocked. The brain, seeking equilibrium, adapts: it down-regulates GABA-A receptors and up-regulates NMDA receptors to claw excitability back to normal in the constant presence of alcohol. Now take the alcohol away. The GABA inhibition that was propping the system up vanishes, while a surplus of hypersensitive NMDA receptors is suddenly unopposed by their usual blocker. The result is unbridled CNS excitation — the mirror image of intoxication. This is the same neuroadaptation of tolerance and dependence discussed in the Drugs of Abuse chapter, but here it is the removal, not the drug, that harms.
The syndrome unfolds on a predictable timeline. At roughly 6–12 hours after the last drink come the minor symptoms: tremor, anxiety, sweating, nausea, a fast pulse. Around 12–24 hours, some develop alcoholic hallucinosis — vivid hallucinations (often visual or tactile, the classic "insects") but with an otherwise clear sensorium and normal vital signs. Withdrawal seizures — generalised tonic-clonic, from that NMDA excess — cluster around 24–48 hours. And at the far, dangerous end, 48–72 hours out, comes delirium tremens: a confused, disoriented, hallucinating patient with an autonomic storm — fever, tachycardia, hypertension, profuse sweating. Delirium tremens is a genuine medical emergency with a mortality that is far from trivial if untreated.
Management: restoring the GABA tone
The logic of treatment falls straight out of the mechanism. If withdrawal is caused by a sudden collapse of GABA inhibition, the cornerstone is a drug that restores it: benzodiazepines, which act on the very same GABA-A receptor the alcohol was occupying. They calm the tremor, prevent and treat seizures, blunt the autonomic storm, and are the mainstay for delirium tremens. Modern practice favours symptom-triggered dosing — giving benzodiazepine only when a validated severity score (such as CIWA-Ar) says the patient needs it — which uses less drug than fixed schedules and avoids over-sedation. Long-acting agents like chlordiazepoxide or diazepam give a smooth, self-tapering course; but in significant liver disease, choose lorazepam, which is cleared by conjugation and doesn't accumulate active metabolites in a failing liver. This is the same benzodiazepine pharmacology taught in the Central Nervous System chapter, applied to a very different emergency.
One rule outranks almost everything else: thiamine before glucose. Chronic drinkers are typically thiamine (vitamin B1) depleted, and thiamine is the essential cofactor for the glucose-metabolising enzymes of the brain. Give a large glucose load to a thiamine-deficient brain and you drive those enzymes hard without their cofactor — you can precipitate Wernicke's encephalopathy, an acute, potentially reversible triad of confusion, eye-movement abnormalities (ophthalmoplegia/nystagmus) and ataxia. Miss or under-treat it and it can become Korsakoff's syndrome: a permanent, dense loss of the ability to form new memories, with confabulation. Hence the rule drilled into every trainee: give thiamine before (or with) glucose, never glucose alone, in anyone who might be alcohol-dependent or malnourished. This ties directly to the Nutrition and Endocrine sections, where thiamine's role and the deficiency states are covered in depth.
Benzodiazepine of choice: chlordiazepoxide or diazepam (long-acting, smooth taper) for most patients; lorazepam (or oxazepam) when there is liver disease, because it is glucuronidated and won't accumulate. Adjuncts: correct magnesium and other electrolytes, rehydrate, and treat coexisting alcoholic ketoacidosis or hypoglycaemia. Thiamine: high-dose parenteral thiamine in suspected or established Wernicke's, given before any glucose. A cautionary contrast lives one shelf over in the Toxic Alcohols chapter: for methanol or ethylene glycol poisoning the antidote of choice is fomepizole (an alcohol-dehydrogenase blocker), but ethanol itself can be used as a competitive substrate for the same enzyme when fomepizole is unavailable — the same molecule that poisons here becomes a treatment there.
- Chronic alcohol up-regulates NMDA and down-regulates GABA-A; stopping it unleashes CNS hyperexcitability.
- Timeline: tremor/anxiety at 6–12h → hallucinosis → seizures ~24–48h → delirium tremens ~48–72h.
- Delirium tremens = confusion + autonomic storm; a true emergency with meaningful mortality if untreated.
- Benzodiazepines are the cornerstone — symptom-triggered dosing; long-acting agents, or lorazepam in liver disease.
- Give thiamine before glucose — glucose without thiamine can precipitate Wernicke's encephalopathy.
- Wernicke's (confusion, ophthalmoplegia, ataxia) can progress to permanent Korsakoff amnesia if under-treated.
- Giving glucose before thiamine in a malnourished drinker — a large glucose load can tip a thiamine-deficient brain into Wernicke's encephalopathy.
- Labelling every obtunded, alcohol-smelling patient "just drunk" — and so missing the head injury, hypoglycaemia, sepsis or co-ingestant underneath.
- Under-treating alcohol withdrawal or relying on antipsychotics as first-line — benzodiazepines restore the missing GABA tone; neuroleptics lower the seizure threshold and don't treat the underlying deficit.
A malnourished man with a long alcohol history is brought in confused and drowsy with a capillary glucose of 2.1 mmol/L. What is the safest immediate management?
- Acute intoxication: ethanol is a CNS depressant (enhances GABA-A, blocks NMDA) with zero-order kinetics at higher levels, sinking dose-dependently toward respiratory depression and coma.
- Never accept "just drunk" — check glucose, look for head injury, ketoacidosis and co-ingestants; care is supportive, with glucose and thiamine.
- Withdrawal reverses the adaptation: up-regulated NMDA and down-regulated GABA leave the brain hyperexcitable — tremor (6–12h) → hallucinosis → seizures (24–48h) → delirium tremens (48–72h).
- Benzodiazepines are the cornerstone of withdrawal (lorazepam in liver disease); and thiamine before glucose prevents Wernicke's — which, missed, becomes permanent Korsakoff amnesia.
- Goldfrank's Toxicologic Emergencies — Ethanol, and Alcohol Withdrawal chapters.
- Katzung BG. Basic & Clinical Pharmacology — The Alcohols.
- Rang & Dale's Pharmacology — Drug dependence and CNS depressants (alcohol).
- British National Formulary (BNF) — Alcohol dependence; parenteral thiamine (Pabrinex) prescribing.
- UpToDate / TOXBASE — Management of moderate and severe alcohol withdrawal syndromes; ethanol intoxication in adults.
- Schuckit MA. Recognition and management of withdrawal delirium (delirium tremens). New England Journal of Medicine.
