Alcohol withdrawal symptoms begin 6 to 24 hours after the last drink and follow a predictable three-stage arc: early tremulous symptoms peak at 24 to 48 hours, withdrawal seizures cluster between 12 and 48 hours, and delirium tremens strikes between 48 and 96 hours in roughly 5 percent of withdrawing patients. The timeline matters because each stage carries a different risk profile and a different treatment window. Roughly 50 percent of people with alcohol use disorder who stop abruptly develop withdrawal. Only 3 to 5 percent progress to the life-threatening DTs stage. Knowing exactly where you or someone else sits on that timeline determines whether home monitoring is safe or whether medical supervision is required.
The GABA Glutamate Rebound That Drives Every Symptom
Alcohol is a central nervous system depressant that works by binding to GABA-A receptors, the brain's main inhibitory brake, and simultaneously suppressing NMDA glutamate receptors, the main accelerator. Chronic heavy drinking forces the brain to adapt: it downregulates GABA receptors and upregulates glutamate receptors to maintain normal excitability against a constant sedative tide. This is tolerance at the receptor level.
When alcohol is withdrawn, the adaptation is suddenly unmasked. Inhibitory signaling collapses while excitatory signaling surges. The result is a hyperadrenergic state — elevated norepinephrine, increased sympathetic outflow, and a brain that is effectively running with the brake line cut. Tremor, sweating, tachycardia, and anxiety are the direct downstream products of that imbalance. Seizures and DTs reflect the same mechanism at higher intensity.
In plain terms: the brain spent months building extra accelerators to fight the alcohol brake. Removing the brake leaves those accelerators unopposed.
The key clinical consequence is that symptom severity tracks drinking duration and volume more closely than it tracks the patient's stated "last drink time." A person drinking 20 or more standard drinks daily for years has far more receptor remodeling than a binge drinker of two weeks.
Quantified Risk: Who Withdraws and How Badly
About 50 percent of individuals with alcohol use disorder experience withdrawal when they stop or sharply reduce intake, per the DSM-5 framework used since 2013. Within that group, the severity distribution is roughly:
- Mild withdrawal (tremor, anxiety, insomnia, nausea): approximately 80 percent of symptomatic cases.
- Withdrawal with seizures: 3 to 10 percent of untreated withdrawing patients; seizure risk is highest in those with prior withdrawal seizures.
- Delirium tremens: 3 to 5 percent of all withdrawing patients, but up to 35 percent mortality if untreated. With modern intensive care, DTs mortality falls to 1 to 4 percent.
Repeated withdrawal episodes sensitize the brain. A 2019 analysis in the Journal of Clinical Psychiatry found that each prior medically supervised detox raises the risk of DTs in later episodes, a phenomenon called kindling. A patient who had mild symptoms on their first detox may seizure on their fourth. This is the single most underappreciated variable in withdrawal risk stratification, and it means patient history outweighs current symptom score when predicting escalation.
Alcohol Withdrawal Stages and Timeline Table
The time course below reflects untreated or minimally treated withdrawal in a dependent drinker. Treating with benzodiazepines flattens and delays the curve but does not eliminate the seizure or DTs windows.
| Stage | Time after last drink | Typical symptoms | Peak and risk notes |
|---|---|---|---|
| Stage 1: Autonomic | 6 to 12 hours | Tremor, sweating, nausea, tachycardia, anxiety, insomnia | Tachycardia over 100 beats per minute and systolic BP rise of 10 to 20 mmHg; symptoms peak at 24 to 36 hours |
| Stage 2: Hallucinosis | 12 to 24 hours | Visual, auditory, or tactile hallucinations with intact sensorium | Affects up to 25 percent of withdrawing patients; resolves within 24 to 48 hours; patient remains oriented |
| Stage 3: Seizures | 12 to 48 hours | Generalized tonic-clonic seizures, usually 1 to 6 episodes | 90 percent of withdrawal seizures occur in this window; most are single or brief clusters |
| Stage 4: Delirium tremens | 48 to 96 hours | Disorientation, severe agitation, fever, profound autonomic instability, hallucinations | 3 to 5 percent of withdrawers; mortality up to 35 percent untreated; the true emergency |
| Resolution | 5 to 7 days | Symptoms wane; sleep and mood remain disrupted | Craving and sleep disturbance can persist for weeks to months |
The commonly cited "24 to 72 hour" figure for withdrawal is a simplification. The full clinically relevant arc is 6 hours to 7 days, and the dangerous tail sits at 48 to 96 hours — after many people assume the worst has passed.
Comparing Withdrawal Severity Scales and Treatment Settings
Two validated tools drive most treatment decisions. The following table contrasts them against the practical question of where a patient should be treated.
| Tool or setting | What it measures | Score interpretation | Best used for |
|---|---|---|---|
| CIWA-Ar | 10 symptom items, total 0 to 67 | Under 8 mild; 8 to 15 moderate; over 15 severe or DTs risk | Symptom-triggered benzodiazepine dosing in supervised care |
| AWS scale (DSM-5 based) | 10 criteria, 2 or more required | Count of criteria, not severity-weighted | Diagnosis, not dose titration |
| Home monitoring | Symptom tracking plus a sober adult present | Safe only if CIWA-Ar under 8 and no seizure or DTs history | Mild first-time withdrawal |
| Medically supervised detox | Continuous or scheduled assessment | Any score above 15, prior seizures, or prior DTs | Moderate to severe or high-risk patients |
| Inpatient or ICU | Continuous monitoring, IV therapy | Severe DTs, fever, instability | DTs, comorbid illness, failed outpatient care |
The numeric reconciliation matters here: a CIWA-Ar score of 8 to 15 is "moderate" but does not automatically require hospitalization. A prior seizure or DTs episode, by contrast, drops the admission threshold regardless of current score. In practice: risk history overrides the number on the scale.
Non-Obvious Mechanisms and Corrected Misconceptions
Misconception one: withdrawal is only dangerous if symptoms look severe early. In reality, the most lethal stage, DTs, often follows a deceptively calm period. A patient can score low at hour 24 and crash at hour 72. This biphasic pattern is why a single good day is not a discharge criterion.
Misconception two: alcohol withdrawal is the same as a hangover. They share tremor and nausea, but a hangover resolves in under 24 hours and does not involve receptor-level adaptation. Withdrawal carries measurable seizure and mortality risk that a hangover does not.
A counterintuitive detail: benzodiazepines, the first-line treatment, work by restoring GABA signaling, but in DTs the required doses can be 5 to 10 times the standard starting dose because of the same receptor downregulation that caused the withdrawal. Underdosing is a more common cause of treatment failure than overshooting, per the 2020 ASAM clinical practice guideline.
Another edge case: kindling means a lighter drinker with multiple past detoxes can outrank a heavier drinker on first detox in DTs risk. Clinicians who triage by current volume alone miss these patients.
Red Flags That Require Emergency Care
Seek emergency evaluation immediately for any of the following. Do not wait to see if symptoms improve.
- A seizure, especially a first-ever seizure or one lasting more than 5 minutes.
- Confusion or disorientation with hallucinations — this signals DTs, not hallucinosis.
- Fever above 100.4°F (38°C) with agitation or profuse sweating.
- Heart rate persistently above 120 beats per minute, or blood pressure above 180/110 mmHg.
- Repeated vomiting preventing oral intake, or signs of dehydration.
- Any chest pain, shortness of breath, or new focal neurological deficit.
What to Actually Do: A Practical Protocol
If withdrawal is mild and there is no history of seizures or DTs, home management under supervision is reasonable. Positive actions, in order:
Track and time. Log the last drink hour and check symptoms against the timeline table at 6, 12, 24, 48, and 72 hours. A symptom diary gives a supervising clinician actual data.
Hydrate with electrolytes. Replace fluids with oral rehydration solution, not plain water alone; thiamine deficiency is common in chronic drinkers and can cause Wernicke encephalopathy during withdrawal. Take 100 mg thiamine daily — a low-cost, low-risk step often started in emergency departments.
Do not self-medicate with alcohol. "Tapering" by drinking delays but does not prevent withdrawal, and it resets the receptor adaptation clock.
Arrange a sober adult for the first 72 hours. The highest-risk window for seizures and DTs is unattended at night.
Contact a clinician before hour 12 to discuss whether a benzodiazepine taper or supervised detox is appropriate. Do not start benzodiazepines without a prescription; they carry their own dependence risk.
Plan for days 5 to 30, not just day 5. Sleep disruption, irritability, and craving persist well past the acute phase and predict relapse. Link to ongoing care before the acute symptoms fade.
Limits of the Evidence and Where Uncertainty Remains
The 48-to-96 hour DTs window is well established, but the exact receptor changes that determine who progresses remain partly inferred from animal models. Genetic variation in GABA-A receptor subunits likely explains why two patients with identical drinking histories diverge, but no validated genetic test guides triage today. The kindling effect is supported by observational cohorts and is widely accepted clinically, yet its precise dose-response relationship is still characterized imperfectly. What is not uncertain: the timeline, the seizure window, and the 3 to 5 percent DTs rate are consistent across decades of cohort data, and they justify treating any seizure or confusion during withdrawal as an emergency rather than a wait-and-see event.