Alcohol: What the evidence actually says

For decades, a glass of wine a day was considered heart-healthy wisdom. The science has moved on — and the picture is more complicated than either side of the debate admits.

What it is

Alcohol — ethanol — is a psychoactive substance produced by fermentation. The International Agency for Research on Cancer (IARC) classifies it as a Group 1 carcinogen — a substance for which there is sufficient evidence of causing cancer in humans. This is the same classification given to tobacco, asbestos and ionising radiation.¹ That classification has been in place for decades. What has changed recently is the scientific consensus on what constitutes a safe level of consumption — and the answer is increasingly uncomfortable for moderate drinkers.

Why it matters

The J-curve that wasn't

For much of the late 20th century, epidemiological research appeared to show a protective effect of light-to-moderate drinking, particularly for cardiovascular disease. The famous "J-curve" suggested that one or two drinks a day reduced heart disease risk compared to abstaining. This became the basis for a generation of public health messaging that moderate drinking — a glass of wine with dinner — was not just permissible but possibly beneficial.

That consensus has now largely collapsed. A landmark 2024 analysis re-examined the studies behind the J-curve and found a systematic flaw: many studies classified sick people who had quit drinking as "abstainers," inflating apparent mortality in the non-drinking group and making moderate drinkers look comparatively healthy.² When studies control for this "sick quitter" bias, the protective effect of moderate alcohol largely disappears.³

A 2024 study in JAMA Network Open followed more than 135,000 adults aged 60 and older. It found no improvement in longevity at any level of drinking. Starting at around two drinks per day, the risk of dying early began to climb — and the increase in risk was stronger for women than for men at equivalent drinking levels.⁴

Alcohol and cancer

The cancer picture is more established — and more concerning. The IARC links alcohol to at least seven cancer types — mouth, throat, oesophagus, liver, bowel, and female breast cancer — with sufficient evidence to classify the relationship as causal.¹ The biological mechanism involves ethanol breaking down in the body into acetaldehyde, a compound that damages DNA. This process occurs regardless of the type of drink, its price, or its quality.

Critically, there is no identified threshold below which this carcinogenic effect stops. The WHO European Region data shows that approximately half of all alcohol-attributable cancers occur in people drinking "lightly or moderately" — less than 1.5 litres of wine, or fewer than 3.5 litres of beer per week.⁵ The dose-response relationship is linear: less alcohol means less risk, but there is no floor.

In January 2023, the World Health Organization published a formal statement: "When it comes to alcohol consumption, there is no safe amount that does not affect health."¹

What about the Tromsø Study and longevity specifically?

A 2023 Norwegian cohort study following men aged 50–54 over 40 years found that those who drank rarely or never — even a few times a year — had better survival rates than those who drank more frequently. The relationship was consistent and dose-dependent.⁶ This aligns with Mendelian randomisation data (studies that use genetic variants to better establish causation) showing alcohol consumption is associated with shorter lifespan in both men and women.⁷

The takeaway for anyone optimising for a long healthy life is that alcohol, unlike most lifestyle factors, does not have an evidence-supported "beneficial" zone. The lowest measurable risk is at the lowest measurable consumption.

How units work

The unit system used by the UK's National Health Service (NHS) is the clearest way to compare drinks across type and strength. The formula is straightforward:

Units = ABV (%) × Volume (ml) ÷ 1000

So a 175ml glass of wine at 12% ABV contains: 12 × 175 ÷ 1000 = 2.1 units.

A pint of higher-strength lager at 5.2% ABV: 5.2 × 568 ÷ 1000 = 3.0 units.

The system makes visible what drink labels obscure: a "standard" glass of wine is often 2–3 units, not one. A craft beer at 6.5% ABV in a 568ml glass is nearly 4 units. People who believe they drink within guidelines frequently don't, because the mental model of "one drink = one unit" is simply wrong for most common drinks.

The 100 Great Years app uses NHS unit values for all nine standard drink types. These values are fixed and pre-calculated — tap to log, and the maths is done for you.

Our approach in 100 Great Years

The Alcohol Tracker uses the NHS unit calculation standard but applies a conservative weekly threshold: 14 units per week for men, 7 units per week for women. The male threshold matches both NHS guidance and the US Dietary Guidelines for men. The female threshold matches the US Dietary Guidelines — more conservative than the NHS's single 14-unit limit for all adults — reflecting the evidence that women face higher health risk at equivalent consumption levels.⁴

We are explicit that these are longevity-optimised thresholds, not average-population guidelines. The NHS figure of 14 units was designed to capture "low risk" across a general adult population. Our target user is actively working to extend their healthy years — a goal that warrants more conservative targets.

We don't think you need to be teetotal. The research shows a clear dose-response relationship: the less you drink, the lower the risk. Tracking what you actually consume — in units, not rough drink counts — is the first step to making an informed choice.

One important note on calories: alcohol contains 7 kcal per gram, making it almost as calorie-dense as fat. These calories are metabolically distinct from food calories — they displace fat oxidation, meaning the body preferentially burns alcohol rather than stored fat while it's present in the bloodstream. The app surfaces alcohol calories separately in the Nutrition widget, not as a warning, but because it's information worth having.

Practical guidance

• Track in units, not drinks. A large glass of wine is 3 units. A pint of strong lager is 3 units. "Just a couple" can easily mean 6–8 units.
• Alcohol-free days matter. Spreading consumption across the week is somewhat better than concentrating it — but total weekly units remain the primary metric.
• Context affects risk. Drinking with food, drinking wine (the modest finding from some studies), and avoiding episodic heavy drinking all appear to attenuate — but not eliminate — risk.
• If you have a personal or family history of any of the seven alcohol-related cancers, the evidence for zero or minimal consumption is stronger still.
• The app does not track mood in relation to alcohol — but if you notice patterns (drinking to manage stress, for example), that is worth discussing with your GP.

100 Great Years perspective

We've tried to be honest here rather than reassuring. Alcohol is deeply embedded in social life, and that reality matters — isolation and lack of social connection also affect longevity meaningfully. We're not advocating abstinence as a blanket recommendation. But we do think people deserve an accurate picture of what the current evidence says, free from the decades of industry-influenced research that overstated any benefits.

The unit system is a tool for awareness, not a morality framework. What you do with the information is your decision. Our job is to make sure the information is right.

Find out how alcohol is affecting your score

Get your free Health and Wealth scores in 5 minutes.

This article is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional before making decisions about your health.