Alcohol use disorder (AUD) is the most prevalent substance use disorder in nearly every country surveyed and a leading modifiable contributor to medical mortality, psychiatric morbidity, and trauma. DSM-5-TR collapsed the prior DSM-IV split of abuse and dependence into a single dimensional diagnosis of AUD graded mild, moderate, or severe by symptom count, while ICD-11 retains a distinct "alcohol dependence" construct under chapter 06. The clinician's job is fourfold: recognize the disorder despite frequent under-disclosure, manage withdrawal safely, initiate evidence-based pharmacotherapy and psychosocial treatment, and screen for the medical and psychiatric sequelae that drive long-term morbidity. Three FDA-approved medications (Naltrexone, Acamprosate, Disulfiram) are underused, and benzodiazepines remain the standard of care for medically significant withdrawal. The bottom line: AUD is a chronic relapsing condition treatable in primary care and psychiatric settings, and missed opportunities to intervene early are missed opportunities to prevent Wernicke encephalopathy, hepatic failure, suicide, and traffic deaths.

This overview draws on the APA Practice Guideline for the Pharmacological Treatment of Patients With Alcohol Use Disorder (2018), the NICE clinical guideline CG115 on alcohol-use disorders (with subsequent surveillance updates), the VA/DoD Clinical Practice Guideline for Substance Use Disorders (2021), the World Federation of Societies of Biological Psychiatry guidelines, DSM-5-TR, ICD-11, FDA prescribing information for naltrexone, acamprosate, and disulfiram, the most recent Cochrane reviews on pharmacotherapy for AUD and on management of alcohol withdrawal, and standard reference texts (Kaplan and Sadock's Synopsis of Psychiatry, Stahl's Essential Psychopharmacology). Citation pass completed 2026-05-15.

AUD is common, under-detected, and undertreated, and the gap between prevalence and treatment receipt is among the widest in psychiatry. Burden concentrates in young to middle adulthood, in men more than women, and in individuals with co-occurring psychiatric disorders.

Prevalence and treatment gap

• Twelve-month prevalence of DSM-5 AUD in U.S. adults is approximately 13.9% and lifetime prevalence approximately 29.1% in the NESARC-III survey.^[1]
• Globally, alcohol use disorders affect roughly 100 million people, with the highest age-standardized prevalence in Eastern Europe.^[2]
• Fewer than 1 in 10 individuals with past-year AUD receive any treatment, and fewer than 1 in 25 receive FDA-approved pharmacotherapy.^[1,3]

Demographics

• Male-to-female ratio is approximately 2:1, with the gap narrowing in younger cohorts.^[1]
• Median age of onset is the early 20s, with the highest 12-month prevalence in the 18-29 age band.^[1]
• Native American and white non-Hispanic populations show higher lifetime prevalence than Black, Hispanic, and Asian populations in U.S. epidemiologic data.^[1]

Comorbidity and consequences

• Approximately 40% of individuals with AUD meet criteria for another current substance use disorder; tobacco use disorder is the most common co-occurrence.^[1]
• Mood disorders, anxiety disorders, PTSD, and antisocial personality disorder are over-represented relative to the general population, with odds ratios typically 2-4.^[1,4]
• Alcohol is implicated in approximately 5% of global deaths and 5% of disability-adjusted life-years lost, with leading causes including liver disease, cancer, cardiovascular disease, suicide, and traffic injury.^[2]
• Lifetime suicide risk in individuals with AUD is roughly 10-fold the general-population rate.^[5]

Key risk factors

• First-degree family history of AUD (heritability estimated at approximately 0.5 in twin and adoption studies).^[6]
• Early age of first drink (before age 14 carries an approximately 4-fold increased lifetime risk vs. first drink at age 21).^[1]
• Male sex, single marital status, lower educational attainment, and adverse childhood experiences.^[1,4]
• Co-occurring psychiatric disorders, particularly PTSD and bipolar disorder.^[4]

AUD reflects converging neurobiological, genetic, and environmental vulnerabilities acting on reward, stress, and cognitive-control circuits. The current dominant framework is Koob and Volkow's three-stage cycle of binge/intoxication, withdrawal/negative affect, and preoccupation/anticipation, each mapped to distinct neural substrates.^[7]

Neurotransmitter pharmacology

• Acute ethanol potentiates GABA-A receptor chloride conductance and inhibits NMDA receptor glutamatergic transmission, producing sedation, anxiolysis, and disinhibition.^[8]
• Mesolimbic dopamine pathway release in the nucleus accumbens, driven indirectly via ventral tegmental area disinhibition, mediates positive reinforcement.^[7]
• Endogenous opioid release (beta-endorphin) at mu-opioid receptors contributes to euphoria and craving, the rationale for naltrexone.^[9]
• Chronic exposure produces compensatory NMDA up-regulation and GABA-A down-regulation; abrupt cessation unmasks glutamatergic hyperexcitability that drives withdrawal seizures and Delirium tremens.^[8]

Circuits and imaging

• Binge/intoxication maps to ventral striatum and ventral pallidum; withdrawal/negative affect engages the Extended amygdala (central nucleus of amygdala, bed nucleus of stria terminalis) with CRF and dynorphin recruitment; preoccupation/anticipation involves prefrontal cortex and insula.^[7]
• Structural MRI shows cortical and cerebellar atrophy and reduced hippocampal volume in chronic AUD, partially reversible with sustained abstinence.^[10]
• Functional MRI demonstrates blunted ventral striatal response to non-drug rewards and exaggerated response to alcohol cues.^[10]

Genetics

• Heritability is approximately 0.5-0.6.^[6]
• Protective variants in ADH1B (rs1229984) and ALDH2 (rs671, the East Asian "flushing" allele) accelerate acetaldehyde accumulation and substantially reduce AUD risk.^[6,11]
• GWAS implicate loci near ADH1B, KLB, GCKR, and CRHR1, with small individual effect sizes.^[11]

Environmental factors

• Early-life adversity, parental AUD, peer drinking norms, and alcohol availability and pricing all modulate risk.^[4]
• Trauma exposure and PTSD strongly predict transition from use to disorder, and AUD reciprocally increases trauma exposure.^[4]

DSM-5-TR diagnoses alcohol use disorder as a dimensional condition requiring at least 2 of 11 criteria within a 12-month period, with severity graded by symptom count. The diagnosis subsumes the older DSM-IV categories of abuse and dependence, and the 11 criteria map to four conceptual clusters.

DSM-5-TR criterion clusters

• Impaired control: drinking larger amounts or longer than intended; persistent unsuccessful efforts to cut down; excessive time spent obtaining, using, or recovering from alcohol; craving.^[12]
• Social impairment: failure to fulfill role obligations; continued use despite recurrent interpersonal problems; reduction or abandonment of important activities.^[12]
• Risky use: recurrent use in physically hazardous situations; continued use despite physical or psychological problems caused or worsened by alcohol.^[12]
• Pharmacological criteria: tolerance and withdrawal.^[12]

Severity, specifiers, and course

• Mild: 2-3 criteria; Moderate: 4-5 criteria; Severe: 6 or more criteria.^[12]
• Course specifiers: in early remission (3-12 months without meeting criteria except craving) or sustained remission (12+ months).^[12]
• Environmental specifier: in a controlled environment (e.g., locked inpatient unit, incarceration).^[12]
• On maintenance therapy may be specified for patients on agonist or antagonist medications, though this specifier is less commonly applied for AUD than for opioid use disorder.^[12]

Alcohol-induced and related diagnoses in DSM-5-TR

• Alcohol intoxication and alcohol withdrawal are coded separately.^[12]
• Alcohol-induced disorders include depressive, anxiety, bipolar, psychotic, sleep, sexual, and neurocognitive presentations, each with onset during intoxication or withdrawal specifiers.^[12]
• Alcohol-induced major or mild neurocognitive disorder may be amnestic-confabulatory (Korsakoff type) or non-amnestic.^[12]

ICD-11 differences worth knowing

• ICD-11 retains "alcohol dependence" (6C40.2) as a distinct diagnosis defined by impaired control, increasing priority of use, and physiological features; "harmful pattern of use" (6C40.1) and "hazardous use" sit on a use spectrum below dependence.^[13]
• ICD-11 alcohol dependence requires only two of three core features over 12 months (or daily/near-daily use over 1 month), a lower symptom-count threshold than DSM-5-TR severe AUD.^[13]

AUD presents along a spectrum from intermittent heavy episodic drinking to physiologically dependent daily use with multiorgan complications. Recognizing the syndrome requires attention to behavior, medical findings, and the specific time course of intoxication and withdrawal.

Alcohol intoxication

• Slurred speech, ataxia, nystagmus, impaired attention or memory, disinhibition, mood lability, and stupor at higher doses.^[12]
• Severe intoxication produces respiratory depression, hypothermia, hypoglycemia, and death; legal driving limit in most U.S. jurisdictions is a blood alcohol concentration of 0.08 g/dL.^[14]
• Tolerant chronic drinkers may appear clinically sober at blood alcohol concentrations that would render a naive drinker comatose.^[14]

Alcohol withdrawal — time course

• 6-24 hours: tremor, anxiety, insomnia, autonomic hyperactivity (tachycardia, hypertension, diaphoresis), nausea.^[15]
• 12-48 hours: alcohol withdrawal seizures, typically generalized tonic-clonic, often single or in brief clusters.^[15]
• 12-24 hours: Alcoholic hallucinosis — visual, auditory, or tactile hallucinations with intact orientation, distinct from delirium tremens.^[15]
• 48-96 hours: delirium tremens — global confusion, marked autonomic instability, hallucinations, and agitation; historical mortality 15-20% untreated, now under 5% with modern care.^[15-16]

Prodrome and course

• Initial heavy use typically begins in late adolescence to early 20s, with disorder onset clustered in the 20s and 30s.^[1]
• Course is typically relapsing-remitting, with periods of abstinence interspersed with relapse; sustained remission becomes more likely with each additional year of abstinence.^[17]
• Roughly one-third of individuals with AUD achieve sustained remission, often without formal treatment, though treatment substantially increases the likelihood and shortens time to remission.^[17]

Atypical presentations and red flags

• Older adults frequently present with falls, cognitive complaints, depression, or insomnia rather than overt heavy drinking.^[18]
• Women present at lower lifetime exposure with more rapid progression to medical complications (telescoping).^[1]
• Unexplained macrocytosis, elevated GGT, AST:ALT ratio greater than 2:1, or elevated carbohydrate-deficient transferrin should prompt screening.^[19]
• New-onset seizure in an adult, unexplained pancreatitis, atrial fibrillation in a young patient, or recurrent trauma warrant explicit alcohol-use history.^[19]

Heavy alcohol use can mimic, mask, or worsen virtually every major psychiatric and many medical syndromes. The clinical task is to separate primary psychiatric illness from alcohol-induced phenomena, and to recognize medical mimics whose missed diagnosis can be lethal.

Primary psychiatric mimics

• Major depressive disorder: depression that persists 4+ weeks after sustained abstinence is more likely primary; alcohol-induced depressive disorder typically resolves within 2-4 weeks of abstinence.^[20]
• Generalized anxiety disorder and panic disorder: anxiety symptoms during withdrawal often resolve over 1-3 weeks; persistence beyond that suggests a primary anxiety diagnosis.^[20]
• Bipolar disorder: high comorbidity with AUD; manic or hypomanic episodes preceding alcohol use, or persisting in abstinence, support a primary diagnosis.^[21]
• Primary psychotic disorders: alcoholic hallucinosis features intact orientation and resolves with abstinence, distinguishing it from schizophrenia.^[15]
• Borderline and antisocial personality disorders: pervasive interpersonal and behavioral patterns precede and persist beyond alcohol use.^[12]

Medical mimics of intoxication or withdrawal

• Hypoglycemia, hepatic encephalopathy, hyponatremia, and hypercapnia can mimic intoxication; check glucose, electrolytes, ammonia, and arterial blood gas when uncertain.^[19]
• Sepsis, meningitis or encephalitis, thyroid storm, pheochromocytoma, and sympathomimetic toxicity mimic withdrawal-like autonomic hyperactivity.^[15]
• Subdural hematoma in a heavy drinker who fell can present as evolving confusion misattributed to delirium tremens; threshold for head CT should be low.^[16]
• Wernicke encephalopathy presents with confusion, ophthalmoplegia, and ataxia; the classic triad is present in fewer than 20% of cases, so any chronic drinker with confusion warrants empirical thiamine.^[22]

Other substance-induced presentations

• Benzodiazepine and barbiturate withdrawal share the GABAergic withdrawal phenotype and can be clinically indistinguishable from alcohol withdrawal in the absence of a substance history.^[15]
• Stimulant intoxication and withdrawal, opioid withdrawal, and gamma-hydroxybutyrate withdrawal each have overlapping autonomic features and should be considered when the time course or response to benzodiazepines is atypical.^[15]

Assessment combines a non-judgmental clinical interview, validated screening, focused physical exam, and targeted laboratory work. The goal is to quantify use, identify withdrawal risk, capture comorbidities, and stage medical damage.

Interview approach

• Use open, normalizing questions ("Many people drink to manage stress — tell me about your drinking") and quantify in standard drinks per occasion, frequency per week, and pattern over the past 12 months.^[23]
• Ask about prior withdrawal complications (seizures, delirium tremens, hospitalization), longest period of abstinence, and prior treatment exposure.^[23]
• Screen explicitly for suicidality, intimate partner violence, driving under the influence, and access to firearms; AUD multiplies risk in each domain.^[5,24]
• Collateral history from a partner or family member, when consent allows, substantially improves accuracy.^[23]

Validated screening instruments

• AUDIT (Alcohol Use Disorders Identification Test, 10 items): a score of 8 or more is the standard threshold for hazardous drinking; the 3-item AUDIT-C is a rapid primary-care variant.^[25]
• CAGE: 4-item screen; 2 or more affirmative responses suggests problem drinking but is less sensitive in women and in less-severe use.^[26]
• CIWA-Ar (Clinical Institute Withdrawal Assessment for Alcohol, revised): a 10-item severity scale used to guide symptom-triggered withdrawal dosing.^[27]
• PAWSS (Prediction of Alcohol Withdrawal Severity Scale): identifies patients at risk for complicated withdrawal who need inpatient management.^[28]

Physical exam findings to seek

• Stigmata of chronic liver disease: spider angiomata, palmar erythema, jaundice, gynecomastia, caput medusae, ascites.^[29]
• Cerebellar findings: wide-based gait, dysmetria, intention tremor.^[10]
• Peripheral neuropathy and proximal myopathy in long-standing use.^[10]
• Cardiomyopathy: displaced apex, S3, signs of heart failure in established alcoholic cardiomyopathy.^[29]

Laboratories and imaging

• Mandatory in suspected significant use: CBC (macrocytosis), comprehensive metabolic panel (AST, ALT, GGT, bilirubin, albumin, glucose, electrolytes, magnesium, phosphate), INR, and lipase if abdominal pain.^[19,29]
• Carbohydrate-deficient transferrin and phosphatidylethanol are sensitive biomarkers of recent heavy use and useful for monitoring abstinence.^[19]
• Urine toxicology to detect co-use of benzodiazepines, opioids, stimulants, and cannabis.^[19]
• Pregnancy test in any patient of childbearing potential.^[12]
• Head CT for any chronic drinker with new confusion, focal neurologic findings, or unexplained fall.^[16]
• EKG to assess QTc and screen for atrial fibrillation, particularly in the holiday-heart pattern.^[29]
• ECG and echocardiography for suspected alcoholic cardiomyopathy; abdominal ultrasound or transient elastography for hepatic staging in established liver disease.^[29]

What NOT to order

• Routine head imaging in uncomplicated withdrawal without focal findings or trauma is low-yield.^[16]
• EEG in uncomplicated single withdrawal seizure with full neurologic recovery is generally not required; recurrent or focal seizures warrant neurology referral.^[15]

Treatment of AUD has two phases: safe management of withdrawal when physiologically dependent, then sustained pharmacotherapy and psychosocial treatment to prevent relapse. Strong evidence supports benzodiazepines for medically significant withdrawal and three FDA-approved relapse-prevention agents (naltrexone, acamprosate, disulfiram), yet fewer than 10% of patients with AUD receive any pharmacotherapy.^[1,3,30]

Pharmacotherapy

• Acute withdrawal — benzodiazepines are first-line and reduce seizures, delirium tremens, and mortality compared with placebo.^[31-32] Long-acting agents (chlordiazepoxide 25-100 mg PO q6h PRN or diazepam 5-20 mg PO q6h PRN) are preferred for self-tapering effect in patients without significant hepatic impairment.^[31-32]
• For patients with cirrhosis or advanced age, lorazepam 1-2 mg PO/IV q4-6h PRN or oxazepam is preferred due to glucuronidation-only metabolism and lack of active metabolites.^[31-32]
• Symptom-triggered dosing using the CIWA-Ar shortens total benzodiazepine exposure and treatment duration compared with fixed-schedule dosing in monitored settings.^[31]
• Phenobarbital is an evidence-supported alternative or adjunct in benzodiazepine-resistant withdrawal and in some emergency-department protocols, with comparable efficacy in selected populations.^[33]
• Adjuncts: alpha-2 agonists (clonidine, dexmedetomidine) and beta-blockers may treat residual autonomic symptoms but do not prevent seizures or delirium and should not be used as monotherapy.^[31-32]
• Anticonvulsants such as carbamazepine, gabapentin, and valproate have evidence in mild-to-moderate outpatient withdrawal but are not recommended as monotherapy in severe withdrawal.^[31,34]
• Relapse prevention — naltrexone 50 mg PO QD reduces heavy drinking days and relapse to heavy drinking compared with placebo; the extended-release IM formulation (380 mg monthly) addresses adherence and is similarly efficacious.^[30,35-36]
• acamprosate 666 mg PO TID supports maintenance of abstinence after detoxification, with greatest evidence in already-abstinent patients; renally cleared and contraindicated in severe renal impairment.^[30,37]
• disulfiram 250 mg PO QD produces an aversive acetaldehyde reaction with alcohol; benefits depend on adherence and are greatest in supervised administration settings.^[30,38]
• Off-label agents with supportive evidence include topiramate up to 300 mg PO QD (reduces heavy drinking; consider cognitive adverse effects) and gabapentin up to 1800 mg PO divided (modest benefit, especially in patients with prominent withdrawal-related craving).^[30,39-40]
• Baclofen has evidence in patients with hepatic impairment where naltrexone is contraindicated, though efficacy data are mixed.^[41]

Psychotherapy

• Motivational interviewing and motivational enhancement therapy improve engagement and reduce drinking in Brief interventions, particularly in primary care and emergency settings.^[42-43]
• Cognitive behavioral therapy targeting craving, high-risk situations, and coping skills shows moderate evidence for reducing relapse, often delivered in combination with pharmacotherapy.^[42,44]
• Twelve-step facilitation and Alcoholics Anonymous attendance are associated with rates of abstinence comparable to or better than CBT in head-to-head trials, including the Cochrane review of AA and 12-step facilitation.^[45]
• Contingency management produces during-treatment reductions in drinking, though durability after reinforcement ends is limited.^[46]
• Behavioral couples therapy reduces drinking and improves relationship outcomes in partnered patients.^[47]

Neuromodulation

• Repetitive transcranial magnetic stimulation targeting dorsolateral prefrontal cortex has shown modest reductions in craving in small trials, but evidence is insufficient to recommend routinely.^[48]
• Deep brain stimulation of the nucleus accumbens has been studied in severe treatment-refractory AUD in early-phase trials only and remains investigational.^[48]

Adjunctive

• Thiamine 100 mg orally daily should be continued for weeks to months after acute supplementation, particularly in patients resuming drinking or with persistent nutritional deficits.^[22]
• Treat comorbid depression, anxiety, and PTSD; SSRIs are reasonable first-line for comorbid mood and anxiety disorders, though they do not directly reduce drinking and may not be effective for depression that resolves with abstinence.^[49]
• Address tobacco use disorder concurrently; co-treatment improves outcomes for both conditions and does not worsen AUD outcomes.^[50]
• Linkage to mutual-help groups, peer recovery support, and recovery housing improves long-term outcomes when integrated with formal treatment.^[51]

Initiating naltrexone before discharge from inpatient detoxification, or within the first office visit after outpatient withdrawal, substantially improves uptake compared with deferred initiation.^[30]

Treatment-related harms in AUD cluster around three predictable areas: medication adverse effects, the medical and neuropsychiatric sequelae of ongoing drinking, and the limits of the trial evidence base itself. Most pharmacologic and psychosocial interventions are safe in absolute terms but require attention to renal, hepatic, and comorbid-medication context.

Common adverse effects of approved pharmacotherapy

• Naltrexone: nausea (most common), headache, fatigue, transient transaminitis at standard doses.^[35]
• Acamprosate: diarrhea, anxiety, insomnia, asthenia; large pill burden (two 333 mg tablets three times daily) limits adherence.^[37]
• Disulfiram: drowsiness, metallic taste, rash, peripheral neuropathy with prolonged use.^[38]
• Topiramate: paresthesia, anorexia, cognitive slowing ("dopamax"), word-finding difficulty, taste alteration.^[39]
• Benzodiazepines used for withdrawal: oversedation, paradoxical delirium in the elderly, respiratory compromise when combined with opioids.^[31]

Serious or rare adverse effects

• Disulfiram-ethanol reaction can produce hypotension, arrhythmia, myocardial infarction, and death; case fatality is rare but well documented.^[38]
• Disulfiram hepatotoxicity, including fulminant hepatic failure, mandates baseline and periodic LFT monitoring.^[38]
• Naltrexone-precipitated opioid withdrawal can be severe and prolonged with the long-acting XR-IM formulation; verify opioid abstinence for 7-10 days before initiation.^[36]
• Wernicke encephalopathy precipitated by glucose administration before thiamine is preventable and remains a sentinel teaching point.^[22]

Monitoring burden and discontinuation

• Disulfiram requires supervised administration and serial LFTs; programs without supervision show poor effectiveness.^[38]
• Naltrexone XR-IM requires monthly clinic visits and a clinician trained in gluteal injection; oral naltrexone requires daily adherence.^[36]
• Acamprosate dosing three times daily with dose adjustment for renal impairment limits practical use in many older patients.^[37]
• Abrupt discontinuation of any agent does not produce a withdrawal syndrome from the medication itself, but loss of pharmacologic relapse protection is common within weeks.^[30]

Limitations of the evidence base

• Most pharmacotherapy trials run 12-24 weeks; long-term efficacy data beyond one year are limited.^[30,35]
• Trial populations often exclude patients with significant medical or psychiatric comorbidity, the very patients clinicians most often see.^[30]
• Heterogeneity of outcome definitions (abstinence vs heavy drinking days vs no-heavy-drinking-days) complicates cross-trial comparison.^[30]
• Publication bias and industry sponsorship affect parts of the literature, particularly for newer or off-label agents.^[34,39]

Treatment principles for AUD are stable across populations, but dosing, screening cutoffs, and risk-benefit calculations shift meaningfully in pregnancy, older adults, comorbid medical and psychiatric illness, and culturally specific contexts.

Pregnancy and the perinatal period

• No safe threshold of prenatal alcohol exposure has been established; fetal alcohol spectrum disorders include facial dysmorphism, growth restriction, and neurodevelopmental impairment.^[46]
• Brief intervention and motivational counseling are first-line in pregnancy; pharmacotherapy is rarely used given limited safety data, with disulfiram and topiramate generally avoided.^[46-47]
• Severe alcohol withdrawal in pregnancy is treated with benzodiazepines because untreated severe withdrawal poses greater fetal risk than short-course benzodiazepine exposure.^[15,47]

Older adults

• Lower body water content and reduced first-pass metabolism produce higher blood alcohol concentrations at lower intake; NIAAA recommends no more than 1 standard drink per day for adults over 65.^[18]
• Presentations are often atypical: falls, cognitive complaints, depression, sleep disturbance, or medication noncompliance rather than overt heavy use.^[18]
• Short-acting benzodiazepines (lorazepam, oxazepam) are preferred for withdrawal in older adults and those with hepatic impairment because they lack active metabolites.^[15-16]

Comorbid medical illness

• Cirrhosis: avoid disulfiram; naltrexone is contraindicated in acute hepatitis or decompensated liver disease per FDA labeling; acamprosate is preferred as it is renally cleared.^[36-37]
• Renal impairment: acamprosate requires dose reduction at creatinine clearance 30-50 mL/min and is contraindicated below 30 mL/min.^[37]
• Chronic pain on opioid therapy: naltrexone precipitates opioid withdrawal and is contraindicated; consider acamprosate or topiramate instead.^[36]
• HIV, hepatitis C, and tuberculosis are overrepresented and warrant routine screening at AUD diagnosis.^[1,19]

Comorbid psychiatric illness

• Lifetime comorbidity of AUD with major depressive disorder, anxiety disorders, PTSD, bipolar disorder, and ADHD substantially exceeds chance; integrated treatment improves outcomes over sequential treatment.^[1,48]
• Substance-induced mood and anxiety symptoms commonly resolve within 2-4 weeks of abstinence; persistent symptoms after that window suggest a primary psychiatric disorder requiring independent treatment.^[12,48]
• Suicide risk is elevated approximately 10-fold over the general population and rises further with comorbid depression or recent loss.^[5]

Cultural and structural considerations

• The ALDH2 rs671 variant common in East Asian populations produces a flushing reaction and lower AUD prevalence but does not preclude AUD when drinking occurs.^[11]
• Indigenous populations in several countries carry disproportionate AUD-related mortality driven by historical, structural, and access factors rather than biological susceptibility.^[1]
• Stigma, immigration status, and confidentiality concerns are commonly cited barriers to disclosure; framing screening as a universal precaution rather than targeted inquiry improves detection.^[23]

AUD is a chronic relapsing disorder with substantial heterogeneity. Long-term outcomes are better than common clinical pessimism suggests, particularly when measured with realistic endpoints rather than lifetime abstinence alone.

Natural history and remission

• NESARC longitudinal data indicate approximately one-third of individuals with DSM-IV alcohol dependence achieve sustained recovery by 1-year follow-up without formal treatment, and rates rise with longer follow-up.^[17]
• Among treated individuals, the COMBINE and Project MATCH trials reported substantial reductions in percent heavy drinking days at 1 year compared with baseline, though outcomes were heterogeneous and most patients reduced rather than fully abstained.^[34,49]
• Each additional year of abstinence reduces the probability of relapse; risk remains elevated above general population baseline for at least 5 years.^[17]

Mortality and medical morbidity

• AUD shortens life expectancy by approximately 24-28 years in those with severe disease; alcohol contributes to roughly 5% of global deaths.^[2,50]
• Leading causes of alcohol-attributable death are cardiovascular disease, cancers (oropharyngeal, esophageal, hepatocellular, colorectal, breast), unintentional injury, and liver disease.^[2,50]
• Heavy drinking is causally linked to atrial fibrillation, hemorrhagic stroke, and hypertension, and the cardiovascular protective effects of light drinking reported in older studies have not survived recent Mendelian randomization analyses.^[51]

Functional outcomes

• Sustained remission is associated with measurable improvements in employment, housing stability, family functioning, and self-reported quality of life.^[17,49]
• Cognitive deficits in attention, executive function, and visuospatial processing partially recover with sustained abstinence, with most gains in the first year and continued slower recovery for years.^[10]

Several alcohol-related presentations require immediate stabilization. Recognizing them early prevents the most catastrophic outcomes: aspiration, seizure status, hepatic decompensation, and completed suicide.

When to admit for withdrawal

• History of complicated withdrawal (seizures or delirium tremens), CIWA-Ar above 15 despite outpatient management, significant comorbid medical illness, pregnancy, or absence of a reliable support person.^[15-16]
• PAWSS score of 4 or higher predicts complicated withdrawal and supports inpatient management with symptom-triggered or fixed-dose benzodiazepine protocols.^[28]

Suicide risk

• Acute intoxication disinhibits and impairs problem-solving, raising near-term suicide risk; firearm access during active drinking is a particularly lethal combination.^[5]
• Recent loss (relationship, employment, housing), comorbid depression, prior attempts, and male sex compound risk; means restriction and follow-up within 7 days reduce post-discharge attempts.^[5]

Acute agitation in withdrawal

• Symptom-triggered benzodiazepines (lorazepam 2-4 mg IV/IM or diazepam 5-10 mg IV) remain first-line; antipsychotics are second-line adjuncts for hallucinosis or agitation refractory to adequate benzodiazepines because they lower the seizure threshold.^[15-16]
• Phenobarbital protocols are increasingly used in emergency departments for severe or refractory withdrawal with comparable or superior outcomes in selected series.^[52]

In any patient with possible AUD presenting with altered mental status, hypoglycemia, or starvation, give thiamine before glucose — a glucose load can precipitate Wernicke encephalopathy in thiamine-deficient patients.^[22]

Several questions remain genuinely contested in AUD treatment, ranging from drinking targets to which pharmacotherapies should be considered "first-line" and how aggressively to deploy emerging agents.

Abstinence versus reduced-drinking goals

• Traditional U.S. addiction medicine prioritizes abstinence as the optimal goal, supported by NIAAA and most APA framing.^[30]
• European regulators have approved nalmefene for harm-reduction (reduced heavy drinking days) in some AUD populations, an indication never granted in the United States.^[53]
• Patient-centered care increasingly endorses reduced-drinking goals as legitimate treatment endpoints for patients unwilling or unable to pursue abstinence, with evidence that even non-abstinent reductions confer cardiovascular and hepatic benefit.^[30,53]

Naltrexone versus acamprosate as first choice

• The COMBINE trial found naltrexone superior on heavy drinking outcomes; some European trials and meta-analyses favor acamprosate on abstinence outcomes.^[34,37]
• A 2014 JAMA network meta-analysis found both have comparable effect sizes against placebo with overlapping confidence intervals.^[35]
• Practical choice is driven more by patient factors (renal versus hepatic function, opioid use, drinking goal) than by a clear head-to-head winner.^[30]

Topiramate and gabapentin off-label use

• Both are recommended by APA as alternatives for patients who decline or fail FDA-approved agents, despite lacking FDA AUD indications.^[30]
• Topiramate's cognitive adverse effects limit tolerability; gabapentin has misuse potential, particularly in patients with polysubstance use.^[39-40]

Benzodiazepine alternatives for withdrawal

• Phenobarbital monotherapy and gabapentin protocols are increasingly used for mild-to-moderate withdrawal, particularly in emergency-department fast-track pathways, though evidence quality remains lower than for benzodiazepines.^[16,52]
• Dexmedetomidine is adjunctive only; it controls autonomic hyperactivity but does not prevent seizures or delirium and should not replace benzodiazepines.^[16]

Psychedelic-assisted therapy

• A 2022 randomized trial of psilocybin-assisted therapy reported reduced heavy drinking days, but the trial was small and not yet replicated; psilocybin remains a Schedule I substance in the United States.^[54]
• Enthusiasm in trade press outpaces the evidence base; current guidelines do not endorse routine use outside research settings.^[30,54]

Cannabis as a "substitute"

• Some observational data suggest patients with AUD use cannabis to reduce drinking, but causal evidence is weak and the strategy is not endorsed by guidelines.^[30]

• DSM-5-TR AUD requires at least 2 of 11 criteria within 12 months; severity is mild (2-3), moderate (4-5), or severe (6+).^[12]
• DSM-5-TR removed the legal-problems criterion present in DSM-IV and added craving.^[12]
• Alcohol withdrawal seizures typically occur 12-48 hours after the last drink; delirium tremens peaks at 48-96 hours and historically carried 15-20% mortality untreated.^[15-16]
• First-line treatment for alcohol withdrawal is a benzodiazepine, dosed by symptom-triggered or fixed-schedule protocol guided by CIWA-Ar.^[15-16,31]
• Long-acting benzodiazepines (diazepam, chlordiazepoxide) are preferred in withdrawal except in elderly patients or those with significant hepatic disease, where lorazepam or oxazepam is safer.^[15-16]
• Give thiamine before glucose in suspected Wernicke encephalopathy to avoid precipitating or worsening it.^[22]
• Wernicke triad: ophthalmoplegia, ataxia, encephalopathy; the full triad is present in fewer than one in five cases, so treat empirically when any feature is present.^[22]
• Korsakoff syndrome is characterized by anterograde and retrograde amnesia with confabulation, and follows untreated Wernicke encephalopathy in many cases.^[22]
• FDA-approved pharmacotherapies for AUD are oral naltrexone, extended-release injectable naltrexone, acamprosate, and disulfiram.^[30,35]
• Naltrexone is contraindicated in patients on opioids or in acute opioid withdrawal.^[36]
• Disulfiram causes a reaction with alcohol mediated by inhibition of aldehyde dehydrogenase, producing flushing, nausea, tachycardia, hypotension, and rarely cardiovascular collapse.^[38]
• Acamprosate is renally cleared; dose-reduce at CrCl 30-50 mL/min and avoid below 30 mL/min.^[37]
• AST:ALT ratio greater than 2:1, elevated GGT, and macrocytic anemia are classic laboratory clues to heavy alcohol use.^[19]
• The ALDH2 rs671 polymorphism (common in East Asian populations) produces a flushing reaction and is associated with lower AUD risk but higher esophageal cancer risk in drinkers.^[11]
• Brief intervention by primary care clinicians reduces alcohol consumption and has the strongest evidence base of any AUD intervention.^[43]
• Approximately one in three individuals with AUD achieves sustained recovery; treatment substantially shortens time to remission.^[17]

No external funding. No conflicts of interest declared. Peer-review status: pending.