Alcohol:
What the evidence
actually shows
in 2025

The J-shaped curve, showing lower cardiovascular mortality in light drinkers compared to abstainers, dominated the alcohol and health literature for 30 years. The problem is methodological. Most studies compared drinkers to a composite abstainer group that included former drinkers who had quit due to ill health, lifelong teetotallers with social isolation risk factors, and people on medications incompatible with alcohol. When studies control for this “sick quitter” confounding, the apparent cardiovascular benefit largely disappears. Mendelian randomisation studies, which use genetic variants associated with lower alcohol tolerance as a natural experiment, sidestepping the confounding, generally find no cardiovascular benefit from alcohol consumption. The Global Burden of Disease Study 2020 (published in The Lancet) found that accounting for both risks and benefits, the safest level of alcohol consumption is zero for people under 40, and very low for older adults even allowing for possible cardiovascular effects.

The carcinogenic mechanisms of alcohol are well-established and dose-dependent. Ethanol is metabolised to acetaldehyde, a directly mutagenic compound that damages DNA and impairs DNA repair mechanisms. Alcohol also generates reactive oxygen species, disrupts folate metabolism, acts as a solvent increasing absorption of other carcinogens in tobacco smoke, and raises oestrogen levels (increasing breast cancer risk). The seven cancers with established causal links include: mouth and throat, oesophagus, larynx, liver, bowel, and female breast cancer. The 2025 US Surgeon General advisory noted that despite alcohol being a more significant cancer driver than tobacco in terms of absolute case numbers attributable to it, public awareness of the alcohol-cancer link remains dramatically lower than awareness of the tobacco-cancer link.

Alcohol is commonly used as a sleep aid, based on the observation that it reduces sleep onset latency, you fall asleep faster. This is pharmacologically real but the overall effect on sleep quality is negative. Alcohol fragments sleep in the second half of the night, reduces REM sleep, increases slow-wave sleep early in the night followed by rebound wakefulness, raises resting heart rate, and activates the sympathetic nervous system during metabolism in the early morning hours. Even one or two drinks consumed within four hours of sleep produce measurable reductions in sleep quality and next-day cognitive performance. For people tracking HRV and resting heart rate, the acute effect of alcohol the night before is typically the largest single-night disturbance they will observe.

Alcohol consumption directly interferes with the anabolic signalling required for muscle repair and adaptation after training. Research by Parr et al. (2014) found that alcohol consumption post-exercise reduced muscle protein synthesis rates by approximately 24% compared to carbohydrate and protein alone, even when adequate protein was consumed. The mechanism involves suppression of the mTORC1 signalling pathway and direct inhibition of protein synthesis at the ribosomal level. Alcohol also impairs testosterone production acutely (within hours) and chronically with regular consumption, increases cortisol, disrupts IGF-1 signalling, and impairs glycogen resynthesis. For anyone training for strength, power, or body composition, regular alcohol consumption represents a consistent and quantifiable drag on adaptation.

Alcohol calories (7 kcal per gram) are calorie-dense and metabolically processed as a priority fuel, displacing fat oxidation. But beyond the caloric contribution, alcohol directly promotes visceral fat accumulation through multiple mechanisms: it stimulates appetite and reduces inhibitory control around food choices, disrupts the hormonal environment that regulates fat distribution (elevated cortisol, suppressed testosterone), and is associated with increased activity of lipoprotein lipase in visceral adipose tissue. Heavy drinking is consistently associated with elevated waist-to-height ratio and visceral adiposity independently of total calorie intake. Even moderate regular consumption is associated with greater visceral fat accumulation than equivalent caloric intake from other sources in several controlled studies.

The UK Chief Medical Officers recommend a maximum of 14 units of alcohol per week for both men and women, spread across three or more days. This guidance is designed to reduce the risk of the most serious harms from regular drinking at a population level. It is explicitly not a declaration that 14 units per week is safe, it is a harm reduction threshold below which the most severe chronic health consequences (liver disease, alcohol dependence, heavy-drinking-associated cancer risk) are substantially reduced. Cancer risk and sleep disruption apply at doses well below 14 units weekly. The UK unit system: one unit = 10ml or 8g of pure alcohol. A standard 175ml glass of 13% wine is approximately 2.3 units. A pint of 4% lager is approximately 2.3 units.

The evidence on alcohol does not require everyone to stop drinking. It does require honest accounting of what regular alcohol consumption costs in terms of sleep quality, recovery, body composition, muscle adaptation, and cancer risk. For someone who is seriously training, tracking recovery metrics, trying to reduce visceral adiposity, and investing in their long-term health, the question is whether the enjoyment or social value of regular drinking is worth the quantifiable physiological costs. That is a personal decision. The science provides the data for making it honestly rather than through the comfortable fiction of a glass of wine being good for the heart.