Alcohol

The shifting consensus

For decades, observational studies suggested moderate alcohol intake (particularly red wine) was associated with lower cardiovascular mortality than abstinence (the “J-curve”). Subsequent work has:

• Identified that “abstainers” were often former drinkers or sick-quitters, biasing the reference group.
• Used Mendelian randomisation to test causality, with results suggesting the cardiovascular benefit largely disappears.
• Documented that even light drinking modestly raises risk of breast, oesophageal, oral, and several other cancers.

The 2022 Lancet GBD analysis concluded that for adults aged <40 in high-income settings, any amount of alcohol raises overall risk; for older adults the threshold is small (~1–2 standard drinks).

What current health bodies say

Guidance varies but has tightened. WHO: no safe level for cancer. Canadian guidance (2023): low risk = ≤2 drinks/week. UK: ~14 units/week spread across days. US: 1/day (women) or 2/day (men) historically; under review.

Practical framing

If you don’t drink, there is no health reason to start. If you drink, less is generally better than more, and any single drink does more than zero to cancer risk. Heavy drinking has well-known dose-dependent harms across nearly every organ system.

Related entries

Cancer (overview), Cardiovascular disease, Sleep optimization.