Alcohol withdrawal is a hyperexcitable rebound state that begins when blood alcohol falls after a period of heavy, sustained drinking. The brain adapts to constant alcohol by dialing down its own inhibitory GABA signaling and ramping up excitatory glutamate activity. When alcohol is removed, that counter-regulation is unmasked, producing tremor, anxiety, and autonomic surge within 6 to 24 hours. The timeline runs in three phases: early (6–24 hours), seizure window (12–48 hours), and delirium tremens (48–96 hours). Roughly half of heavy drinkers develop withdrawal, 3 to 5 percent of those progress to delirium tremens, and untreated DTs carry a mortality of up to 5 percent.
The Mechanism: Why Removal, Not Drinking, Causes Symptoms
Alcohol is a central nervous system depressant that binds directly to GABA-A receptors and enhances inhibitory signaling. It also suppresses NMDA-glutamate receptors, which normally drive excitation. With repeated heavy exposure over days to weeks, the brain restores balance by down-regulating GABA-A sensitivity and up-regulating NMDA receptor density. This is neuroadaptation, and it is measurable.
The clinical consequence is tolerance while drinking and rebound when drinking stops. Remove the alcohol and the reinforced glutamate system fires essentially unopposed. The result is a sympathetic surge: heart rate climbing 20 to 30 beats per minute above baseline, systolic pressure rising 10 to 30 mm Hg, and tremor from increased neuromuscular excitability.
In plain terms: the brain spent weeks pressing its foot on the brake to offset alcohol, and when alcohol leaves, the accelerator is still pressed down.
The most dangerous downstream effect involves the NMDA receptor's calcium channel. Sudden glutamate excess floods neurons with calcium, and lowering of the seizure threshold follows. This explains why withdrawal seizures cluster in a narrow window rather than appearing randomly. It also explains the counterintuitive finding that repeated withdrawal episodes kindle each other: each detox lowers the threshold for the next, which is why a person's fifth withdrawal can be worse than their first despite identical drinking.
The Withdrawal Timeline, Hour by Hour
Symptoms follow a predictable sequence, and the timing of onset is one of the strongest predictors of severity. The table below reflects the standard clinical course in untreated adults, per the 2020 American Society of Addiction Medicine (ASAM) clinical practice guideline.
| Phase | Time after last drink | Typical features | Approximate risk level |
|---|---|---|---|
| Prodromal | 0–6 hours | Mild anxiety, insomnia, sweating; blood alcohol may still be falling | Low |
| Early withdrawal | 6–24 hours | Tremor, nausea, tachycardia (heart rate over 100), hypertension, irritability | Moderate |
| Seizure window | 12–48 hours | Generalized tonic-clonic seizures in 2–5 percent of untreated patients; often 1–3 seizures | High |
| Alcoholic hallucinosis | 12–48 hours | Visual, auditory, or tactile hallucinations with clear sensorium | High |
| Delirium tremens (DTs) | 48–96 hours | Disorientation, severe autonomic instability, fever, agitation | Critical |
| Protracted withdrawal | 2 weeks–12 months | Dysphoria, sleep disruption, craving, mild cognitive fog | Variable |
Two features of this timeline are frequently missed. First, seizures can precede any dramatic physical signs, so a patient who looks calm at hour 14 can seize at hour 20. Second, DTs almost never begin before 48 hours, which means the highest-risk hours fall after many patients have been sent home from an emergency department.
Quantifying Risk: Who Progresses to Severe Withdrawal
Not everyone who stops drinking after a binge develops withdrawal. The strongest predictors are the quantity and duration of prior drinking, a history of previous withdrawal, and prior withdrawal seizures or DTs. A score of 8 or higher on the Clinical Institute Withdrawal Assessment for Alcohol, revised (CIWA-Ar, a 10-item scale scored 0 to 67), generally marks moderate-to-severe withdrawal needing pharmacologic treatment.
The mortality figures deserve precision because they are often quoted loosely. Untreated delirium tremens historically carried mortality of 15 to 20 percent. Modern intensive management has cut that to roughly 1 to 5 percent. The shift is attributable to benzodiazepine protocols and supportive care, not to a milder disease.
A quantified comparison that rarely appears in consumer explainers: the risk of seizure in untreated withdrawal is about 2 to 5 percent, but among patients with a prior withdrawal seizure that figure rises to roughly 20 to 30 percent on subsequent detox. This kind of relative risk escalation is why a single prior seizure changes triage entirely.
Delirium Tremens: The High-Mortality Edge Case
DTs is the severe end of the spectrum, defined by withdrawal plus delirium (disorientation, fluctuating attention) and severe autonomic hyperactivity. It affects roughly 3 to 5 percent of patients hospitalized for withdrawal, but accounts for a disproportionate share of deaths. Core features include tachycardia above 120, fever, profuse sweating, and marked agitation.
The non-obvious detail here is the role of kindling and age. Older patients, those over 65, have substantially higher DTs mortality even with treatment, and patients with concurrent medical illness such as pneumonia, pancreatitis, or liver disease face higher risk. Hypoglycemia, hypomagnesemia, and thiamine deficiency commonly coexist and worsen confusion, so DTs management always includes correction of these deficits.
Per the 2020 ASAM guideline, benzodiazepines remain first-line for DTs, dosed to symptom control rather than a fixed schedule in severe cases. This symptom-triggered approach reduces total medication exposure by 50 to 60 percent compared with fixed dosing in appropriate patients.
Withdrawal vs. Hangover: A Comparison That Clarifies Risk
A common misconception conflates a hangover with withdrawal. They share symptoms but differ in mechanism and danger. The table below separates them.
| Feature | Hangover | Alcohol withdrawal |
|---|---|---|
| Cause | Acute toxicity, dehydration, sleep disruption after a single episode | Neuroadaptation unmasked after sustained heavy use |
| Typical onset | Within hours of drinking ending | 6–24 hours, peaking later |
| Duration | 24 hours or less | 2 days to a week (acute), months (protracted) |
| Autonomic severity | Mild | Marked tachycardia, hypertension, tremor |
| Seizure risk | Essentially none | 2–5 percent untreated; higher with prior seizures |
| Medical danger | Low | Can be fatal (DTs) |
The practical implication: a hangover resolves with time and fluids, whereas withdrawal can escalate and warrants clinical assessment when autonomic signs are prominent.
Misconceptions and Professional Caveats
Several beliefs are common and wrong. One is that withdrawal is purely psychological. The autonomic and seizure components are physiologic and can be measured on a monitor. Another is that a "safe" amount of drinking prevents withdrawal; in dependent drinkers, even a substantial reduction in daily intake can trigger it, because the brain has adapted to a higher set point.
A subtler caveat concerns the kindling phenomenon. Because each withdrawal episode can lower the seizure threshold, repeated cycles of drinking and abrupt cessation produce progressively worse symptoms. This argues against unsupervised repeated detox and for medically supervised withdrawal in anyone with prior severe episodes.
Finally, symptom-triggered dosing is not universal. Patients with a history of DTs or withdrawal seizures are often treated with around-the-clock benzodiazepines rather than as-needed dosing, because waiting for symptoms to escalate is risky in that subgroup. The threshold logic must be reconciled explicitly: a CIWA-Ar of 8 or higher triggers treatment in most protocols, but in patients with prior DTs or seizures, prophylaxis may begin at lower scores and after shorter observation, because their absolute risk is higher at any given score.
What to Actually Do: Practical Actions and Red Flags
For anyone at risk, the safest path is medical assessment rather than willpower. Positive actions include:
- Arrange supervised withdrawal if there is any history of seizures, DTs, or daily heavy drinking for more than two weeks.
- Keep a recent drink log with amount and timing, since onset timing drives triage decisions.
- Restore thiamine and fluids under guidance, as deficiency and dehydration worsen confusion.
- Avoid abrupt cessation without support in dependent drinkers, and taper under clinical direction.
- Re-evaluate at the 48-hour mark, the peak DTs window, even if early symptoms seemed mild.
Seek emergency care immediately for any of these red flags:
- A seizure, especially a first one in this context.
- Confusion, disorientation, or hallucinations.
- Heart rate persistently above 120 or fever.
- Severe agitation or inability to keep fluids down.
- Chest pain or shortness of breath.
Limitations and the State of the Evidence
Much of the withdrawal literature is observational, and severity-prediction tools like CIWA-Ar were validated largely in inpatient settings, which limits their precision in outpatient or home settings. The kindling hypothesis is well supported in animal models and consistent with clinical observation, but human prospective data remain limited. Individual thresholds for progression to DTs are not fully predictable, so clinical judgment supplements any score.
What is established is the mechanism, the timeline, and the mortality gradient across severity. What remains uncertain is how to individualize prophylaxis for borderline cases and how to best manage protracted withdrawal, for which evidence is sparse. The safest stance is early, supervised assessment for anyone with meaningful risk factors, because the dangerous window is predictable even when the individual outcome is not.