Alcohol and Health: Is There a Safe Amount?
Why the 'glass of wine is good for you' finding turned out to be a study artefact, what WHO and Mendelian randomisation show, how alcohol affects the heart, brain, liver and cancer risk.

Contents
The idea that moderate drinking protects the heart persisted for decades in popular culture and even in some official guidance. In recent years the picture has changed: stricter methods showed that the "protective effect" was largely an artefact of study design. Here is what is known today — without moralising and without reassuring half-truths.
Measuring the dose
The confusion starts with units. What counts is pure ethanol, not the volume of the drink.
| Drink | Volume | Pure alcohol |
|---|---|---|
| Beer 5 % | 500 ml | 20 g |
| Wine 12 % | 150 ml (glass) | 14 g |
| Wine 12 % | 750 ml (bottle) | 71 g |
| Spirits 40 % | 50 ml | 16 g |
| Sparkling 11 % | 125 ml | 11 g |
The low-risk threshold in the combined analysis of 83 prospective studies (Wood et al., Lancet 2018) was about 100 g of pure alcohol per week. Above that, life expectancy falls in proportion to intake: roughly 100–200 g a week costs around six months of life, while over 350 g costs four to five years.
Why "a glass of wine is good for you" was a mistake
Observational studies showed a J-shaped curve for decades: moderate drinkers had lower mortality than non-drinkers. The problem lay in who counted as a non-drinker.
The second blow came from Mendelian randomisation — a method using genetic variants as a natural experiment. Parts of the East Asian population carry ALDH2 and ADH1B variants that determine alcohol tolerance: carriers physiologically drink less, and that distribution is random with respect to lifestyle. The large study by Millwood and colleagues in 500,000 participants showed that the more alcohol genetics predicted, the higher the risk of stroke and hypertension — linearly, with no protective segment of the curve.
Читайте также: Blood Pressure Explained: Numbers, Measurement and What to Do
What alcohol affects
| System | Effect | Threshold |
|---|---|---|
| Cancer | Group 1 carcinogen: mouth, pharynx, larynx, oesophagus, liver, bowel, breast | Risk rises from low doses |
| Heart and vessels | Raised blood pressure, cardiomyopathy, atrial fibrillation | Noticeable at moderate intakes |
| Liver | Steatosis → steatohepatitis → fibrosis → cirrhosis | Depends on dose and duration |
| Brain | Impaired memory, atrophy, increased dementia risk | Dose-dependent |
| Sleep | REM suppression, fragmentation in the second half of the night | From a single drink |
| Weight | 7 kcal per gram plus disinhibited appetite | Any amount |
| Mental health | Worsens anxiety and depression with regular use | Dose-dependent |
Atrial fibrillation deserves special mention. The link between alcohol and this arrhythmia is consistent enough to have earned its own name — "holiday heart syndrome". A randomised trial showed that abstinence in drinkers with atrial fibrillation substantially reduced episode frequency.
Breast cancer warrants a separate note because risk rises from the very lowest intakes: each 10 g of pure alcohol a day is associated with roughly a 7–10 % increase in risk. The mechanism is linked to raised oestrogen levels and the action of acetaldehyde.
Читайте также: Menopause Explained: What Changes and What Helps
Alcohol, sleep and recovery
The most noticeable everyday effect is on sleep. Alcohol speeds up falling asleep through sedation, which is why people use it as a sleep aid. But as the ethanol is metabolised, the second half of the night brings a rebound: sleep becomes shallow and fragmented, REM is suppressed, heart rate rises and heart rate variability falls. Wearable trackers make this strikingly visible — even two glasses of wine three to four hours before bed noticeably lower recovery scores.
For people who train there is a further effect: alcohol after exercise suppresses muscle protein synthesis and slows glycogen replenishment.
Читайте также: Sleep Hygiene: The Rules That Actually Work
Читайте также: How Often Should You Train: Frequency, Volume and Recovery
What to do in practice
A realistic position: total abstinence isn't the only option, and cutting back delivers benefit in proportion.
- Count grams of alcohol, not glasses. A bottle of wine is 70 g — most of the weekly low-risk threshold.
- Build in alcohol-free days. Three or four consecutive days give the liver time to recover and break the habit loop.
- Don't drink within 3–4 hours of bed. It is the cheapest sleep-quality improvement available.
- Separate alcohol from emotion. Drinking to relieve stress and anxiety is the mechanism that reliably escalates intake.
- Don't start "for your health". No current guideline recommends that a non-drinker begin drinking.
Читайте также: Triglycerides: Normal Range and How to Lower Them
The bottom line
The belief that moderate drinking is beneficial did not survive scrutiny by stricter methods: the protective effect proved to be a design artefact, and genetic studies show risk rising linearly. Alcohol is an established carcinogen that raises the risk of several cancers even at low intakes, and simultaneously a driver of hypertension, arrhythmia, liver damage and wrecked sleep. No safe level has been established, but risk is dose-dependent — meaning any reduction pays off. Count grams rather than glasses, build in alcohol-free days, and don't drink before bed.
FAQ
Is there a safe level of alcohol?+
Per the WHO position (2023), no safe level of alcohol consumption for health has been established. Risk of several cancers begins rising from the first drink. That does not mean every dose is equally harmful — risk is dose-dependent — but science no longer identifies a completely safe threshold.
What about the research showing red wine is good for you?+
The core problem is methodological. The 'non-drinker' group often included former drinkers who quit because of illness, which artificially worsened their outcomes. Mendelian randomisation studies, free of that bias, find no protective effect.
What counts as moderate drinking?+
Guidance varies by country. A widely used low-risk threshold is no more than 100 g of pure alcohol per week (about 10 standard drinks), with several alcohol-free days. That is a threshold of low risk, not zero risk.
Does alcohol really increase cancer risk?+
Yes. IARC classifies alcohol as a group 1 carcinogen — established as causing cancer in humans. The strongest links are with cancers of the mouth, pharynx, larynx, oesophagus, liver, bowel and breast. For breast cancer, risk rises appreciably even at low intakes.
How does alcohol affect sleep?+
It speeds up falling asleep but wrecks sleep architecture: it suppresses REM in the first half of the night, fragments sleep in the second, raises heart rate and lowers heart rate variability. Hence feeling unrested after an evening of drinking even with adequate sleep duration.
References
- 1.WHO. No level of alcohol consumption is safe for our health, 2023
- 2.GBD 2016 Alcohol Collaborators. Alcohol use and burden for 195 countries and territories. Lancet, 2018
- 3.Wood AM et al. Risk thresholds for alcohol consumption: combined analysis of 83 prospective studies. Lancet, 2018
- 4.Millwood IY et al. Conventional and genetic evidence on alcohol and vascular disease aetiology. Lancet, 2019
Read next

The Mediterranean Diet: What It Actually Is and How to Build It
The most studied eating pattern in the world: the PREDIMED evidence, what the diet contains, a weekly framework, and how to adapt it to an ordinary supermarket and budget.

Sauna and Longevity: What Finnish Studies Show
Finnish cohorts link frequent sauna use to lower heart and all-cause mortality. We cover the mechanisms, safety, and the limits of the evidence.

Health Check-Ups by Age: Which Tests Are Worth It
A decade-by-decade guide to screening that earns its place, based on USPSTF grades, plus the popular tests that do more harm than good in healthy adults.

Blue Zones: What Holds Up and What Does Not
A careful look at the Blue Zones: the record-keeping problem Saul Newman uncovered, the branding around it, and which famous habits have real evidence.

Stroke: Warning Signs, the FAST Test and Prevention
How to recognise a stroke in under a minute with FAST, why every minute counts, the difference between ischaemic and haemorrhagic stroke, and which risk factors you can actually control.

Triglycerides: Normal Range and How to Lower Them
What triglycerides measure, which values count as normal or high, why levels above 500 mg/dL are urgent, and which diet and lifestyle changes actually move the number.