Alcohol Use and Cardiovascular Disease: A Scientific Statement From the American Heart Association

Authors, Journal, Affiliations, Type, DOI

• Mariann R. Piano (Vanderbilt University School of Nursing), Gregory M. Marcus (UCSF), Dawn M. Aycock, Jennifer Buckman, Chueh-Lung Hwang, Susanna C. Larsson (Karolinska Institute), Kenneth J. Mukamal, Michael Roerecke (Centre for Addiction and Mental Health, Canada)
• Circulation, 2025;152:e7–e21
• On behalf of AHA Council on Lifestyle and Cardiometabolic Health; Council on Cardiovascular and Stroke Nursing; Council on Clinical Cardiology; Stroke Council
• AHA Scientific Statement — comprehensive review of observational, RCT, and Mendelian randomization evidence
• DOI: 10.1161/CIR.0000000000001341

Overview

This AHA Scientific Statement provides a comprehensive evidence review of alcohol's relationship with cardiovascular disease across all major CV conditions. The central message is that heavy and binge drinking (≥3 drinks/day) is consistently harmful across every cardiovascular outcome studied, while low-to-moderate consumption (1–2 drinks/day) shows no clear risk to possible modest risk reduction for CAD/MI in observational studies — but Mendelian randomization studies do not confirm any causal protective effect. Alcohol has a clear dose-response relationship with AF and hypertension with no safe lower threshold identified. Alcoholic cardiomyopathy risk is modified by TTN truncating variants. The document emphasizes critical methodological limitations of the existing evidence base and calls for RCTs of light-to-moderate drinking.

Keywords

Alcohol consumption, atrial fibrillation, cardiovascular diseases, coronary artery disease, myocardial infarction, hypertension, risk reduction behavior, binge drinking, alcoholic cardiomyopathy, Mendelian randomization

Key Takeaways

Definitions and Drinking Patterns

• A US standard drink = 14 g (0.6 oz) of alcohol; global definitions range from 8–20 g, complicating cross-study comparisons.
• Low-risk alcohol use (NIAAA): ≤4 drinks/day and ≤14 drinks/week for men <65; ≤3 drinks/day and ≤7 drinks/week for women and men ≥66.
• Binge drinking: blood alcohol ≥0.08 g/dL (≈4 drinks for women, ≈5 for men within 2 hours); heavy drinking: binge on ≥5 days/month.
• 85% of US adults report lifetime alcohol use; heavy alcohol use increased ~20% during COVID-19 (2018–2022 sustained); >175,000 annual deaths in the US attributable to alcohol.

Methodological Considerations

• Self-reported alcohol intake underestimates true consumption, especially heavier drinkers; objective measures (phosphatidylethanol, transdermal sensors) increasingly used.
• Key biases: lifetime selection bias, reverse causality (sick abstainers), confounding by SES and lifestyle factors, reference group selection (non-drinkers vs abstainers vs occasional drinkers) critically affects results — Stockwell et al. showed reference group selection accounts for much discrepancy.
• Mendelian randomization (MR) studies use genetic variants as instrumental variables to mitigate confounding but remain subject to pleiotropy, weak instrument bias, and collider bias.
• Trial emulation (using observational data to mimic RCT design) is an emerging approach but not yet definitive.

Hypertension

• Consumption of <2 drinks/day does not significantly change BP vs. no alcohol use (meta-analysis of 36 RCTs, n=2865).
• ≥3 drinks/day raises SBP and DBP in a dose-dependent manner; biphasic effect — BP decreases at <12 hours post-consumption, then rises 12–24 hours post-consumption (SBP increase 3.7 mmHg).
• 1 drink/day associated with 1.25-mmHg higher SBP; 3 drinks/day with 4.9-mmHg higher SBP (meta-analysis >600,000 participants).
• Reducing intake by ~50% in heavy drinkers (≥6 drinks/day) produces mean SBP decrease of −5.5 mmHg and DBP decrease of −3.97 mmHg.
• Women show steeper hypertension risk slope at >1 drink/day; men show flattening above 3–4 drinks/day.

Coronary Artery Disease (CAD)

• Observational studies generally suggest lower MI risk with alcohol within US Dietary Guidelines limits vs. lifetime abstainers — but Mendelian randomization studies show NO significant association between genetically predicted alcohol intake and CAD risk.
• "Burden of proof" meta-analysis (Carr et al.): cohort studies show 5% average risk reduction across 0–3.5 drinks/day; MR studies show no statistically significant association.
• Binge/heavy episodic drinking eliminates any protective signal: moderate drinkers (<30 g/day) who also engage in heavy episodic drinking had pooled RR 1.12 for CAD (vs. 0.64 for moderate drinkers without binge episodes).
• Methodological limitations (confounding, sick-abstainer bias) limit causal interpretation.

Stroke

• Low-to-moderate drinking (≤2 drinks/day): modest 8–10% decreased ischemic stroke risk in meta-analysis of observational studies; however current-drinker-only analyses show a positive dose-response.
• Heavy drinking (>4 drinks/day): increased risk of all stroke types.
• MR studies consistently report positive association between genetically predicted alcohol and total stroke, ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage.
• 100 g/week increase in alcohol (≈1 drink/day) increases ischemic stroke risk by 13%, intracerebral hemorrhage by 17%, subarachnoid hemorrhage by 9% (combined analysis of 600,000 current drinkers, 19 high-income countries).

Atrial Fibrillation

• Heavier alcohol consumption predicts heightened AF risk; no clear threshold or differential effect by beverage type identified; relationship appears linear.
• Instrumental variable analyses support a causal relationship (Mendelian randomization).
• Data are conflicting whether 1 drink/day influences AF risk; alcohol abstainers are at lower AF risk than those who continue to drink.
• RCT evidence (DRINC trial): Abstinence in Australians consuming ≥10 drinks/week produced substantial reduction in AF burden.
• N-of-1 trials (I-STOP-AFib): Per-protocol analysis found alcohol was the only prespecified trigger causally associated with paroxysmal AF episodes.
• Mechanistic evidence: Alcohol acutely shortens pulmonary vein refractory periods (randomized, double-blind, placebo-controlled study); associated with left atrial enlargement and fibrosis; vagally mediated episodes more common in alcohol-related AF.
• Continuous transdermal alcohol sensors in paroxysmal AF patients showed heightened AF odds within hours of a drinking event.

Ventricular Arrhythmias and Sudden Death

• No consistent evidence that alcohol directly influences premature ventricular contractions, VT, or VF risk.
• ≈1 drink/day associated with lowest sudden death risk; heavy alcohol consumption associated with heightened sudden death risk.
• No relationship between alcohol and sinus node disease from longitudinal cohort studies; moderate consumption associated with lower sinus node disease risk.
• Alcohol has no effect or protective effect on AV conduction disease.

Alcoholic Cardiomyopathy (ACM)

• Long-term excessive alcohol use → dilated LV, normal/reduced LV wall thickness and mass, HFrEF in advanced stages.
• ACM prevalence: 68 per 100,000 US hospitalizations; mean hospitalization age ~56 years; higher prevalence in men; Hispanic men/women hospitalized at younger age (~53 years).
• Dose and duration: ≈7–15 standard drinks/day over 5–15 years associated with adverse systolic/diastolic ventricular function (case-control data). As few as 4 drinks/week was associated with increased odds of diastolic dysfunction in a cross-sectional study (Daka et al.).
• TTN interaction: Truncating TTN variants increase vulnerability to ACM — individuals with TTNtv reporting ~6 drinks/day over 5 years had significantly higher ACM risk. TTN is recognized as a genetic predisposition modifier.
• Sex difference: Women develop ACM at lower alcohol doses and shorter durations than men.
• Alcohol is classified as both an epigenetic trigger in the presence of underlying genetic variants and a non-genetic cause of DCM.
• Reduction in alcohol to <80 g/week (≈6 drinks/week) with guideline-concordant HF therapy improves ventricular function and prognosis.

Heart Failure

• Low-to-moderate (1–2 drinks/day): observational studies suggest inverse association with incident HF; MR study reports absence of association or inferred increased HF risk.

• 21 drinks/week: consensus across studies for increased HF risk (~50% increase at ≈21 drinks/week).

• In patients with structural/functional cardiac abnormalities: ≥5 drinks/week associated with increased progression of asymptomatic LV dysfunction to symptomatic HF (OR 5.0 [95% CI 1.7–15.5]) over 5.4 years (Wong et al.).

Special Populations

• Women: Higher risk of ACM at lower doses; steeper hypertension slope at >1 drink/day; underrepresented in alcohol-CVD RCTs.
• Older adults (50–80 years): Increased medication interaction risk due to polypharmacy and altered metabolism; 27% reported ≥6 drinks on at least 1 occasion in past year.
• Young adults (18–30 years): Binge drinking associated with endothelial dysfunction, coronary calcification, arterial stiffness, higher SBP; INTERSTROKE: binge drinking OR 5.44 (95% CI 1.81–16.4) for ischemic stroke age ≤45 years.
• Race/ethnicity: Genetic polymorphisms in alcohol-metabolizing enzymes (African/Asian ancestry) increase acetaldehyde production; lowest mortality at lower consumption levels among Black individuals vs. White.
• Diabetes: Small RCTs suggest 1 drink/week (wine) safe; conflicting data on BP and glycaemic effects; limited evidence overall.

Cardiovascular Biomarkers with Moderate Alcohol Use

• HDL cholesterol: ↑ 3 mg/dL; Lipoprotein(a): ↓ 0.04 g/L; Fibrinogen: ↓ 0.13 g/L; IL-6: ↓ 0.43 pg/mL; Adiponectin: ↑
• No significant change in: total cholesterol, LDL-C, triglycerides, CRP, TNF-α, leptin, fasting glucose/insulin/HbA1c
• Platelet aggregation reduced at 2 drinks/day but increased at heavier drinking; aspirin co-ingestion with 3–4 drinks potentiates bleeding time

Patient Education and Clinical Guidance

• Alcohol should be documented as a "seventh vital sign" with universal screening.
• Exposure to media reports of cardioprotective effects of moderate alcohol was associated with OR 1.67 (95% CI 1.02–2.74) for hazardous alcohol use among cardiology patients — reinforces need for careful clinical messaging.
• Clinicians should reinforce physical activity, tobacco avoidance, and healthy weight rather than implying alcohol has cardiovascular benefit.

Limitations of the Document

• Overwhelming majority of evidence is observational; prone to confounding, reverse causality, and selection bias.
• Women are underrepresented across alcohol-CVD research; pooled-effect estimates for women in BP trials remain uncertain.
• Mendelian randomization is subject to pleiotropy, collider bias, and weak instrument bias — not a definitive substitute for RCTs.
• No large-scale RCT of light-to-moderate drinking and cardiovascular outcomes yet completed; one Spanish trial (TANIT) includes MI as composite endpoint but recruitment/retention status uncertain.
• Standard drink definitions vary globally (8–20 g); cross-study comparisons limited.
• Potential confounding by SES, diet quality, meal-time drinking, and beverage type not consistently controlled.

Key Concepts Mentioned

• concepts/Heart-Healthy-Dietary-Patterns — alcohol avoidance as component of heart-healthy diet
• concepts/Alcoholic-Cardiomyopathy — dedicated concept for ACM mechanisms, dose-duration, genetics

Key Entities Mentioned

• entities/Atrial-Fibrillation — alcohol as modifiable causal risk factor; abstinence reduces AF burden
• entities/DCM — alcoholic etiology, TTN variant interaction, second-hit theory
• entities/Heart-Failure — dose-dependent HF risk; special risk in structural heart disease
• entities/TTN — truncating variants increase ACM vulnerability
• entities/HCM — not directly addressed but alcohol avoidance relevant to comorbidity management

Wiki Pages Updated

• wiki/sources/alcohol-cv-aha-2025.md (created)
• wiki/sourceindex.md (updated)
• wiki/wikiindex.md (updated)
• wiki/entities/Atrial-Fibrillation.md (alcohol section added)
• wiki/entities/DCM.md (ACM section expanded)
• wiki/concepts/Alcoholic-Cardiomyopathy.md (created)