Across 6,214,569 disease-free Korean adults followed from 2006-2009 through 2022, alcohol intake rose with risk for seven cancers in men and four in women, versus lifelong-abstainer-at-baseline reference groups. Men who drank faced a 9.96% lifetime incidence of alcohol-related cancers against 7.88% for abstainers (2.1-point gap); women showed 10.28% versus 9.89%. Eight of ten circulatory subtypes in men and two in women rose with drinking, while ischemic heart disease, heart failure and arrhythmias showed J- or U-shaped curves that vanished when light drinkers served as the reference. Positive associations also appeared in 14 further conditions in men and six in women.
The vanishing J-curve is the most instructive result. Classic claims of cardioprotection from light drinking have long been attributed to the 'sick quitter' problem, where abstainer groups include former drinkers who stopped for health reasons. Here the dip disappears once light drinkers anchor the comparison, which fits that critique and echoes Mendelian randomization work and the 2023 Zhao meta-analysis showing little mortality benefit from moderate intake. The scale and single-payer linkage make this unusually robust for an observational design, and the absolute figures are useful: a 2-point lifetime cancer increment in men is modest per person but large at population level. Limitations matter. Alcohol was self-reported at only two time points, with no beverage type or drinking pattern detail, and residual confounding from diet or socioeconomic factors can't be excluded. Drinking was far more common in men, so female estimates are less precise. Overall this is confirmatory rather than paradigm-shifting, but it strengthens the case that no intake level is clearly protective, and anyone drinking regularly has reason to discuss cutting back with their doctor.