Stomach cancer remains one of the deadliest malignancies globally, and understanding modifiable risk factors is critical — especially as alcohol consumption is already an established carcinogen for several other cancers. This large pooled analysis challenges the assumption that alcohol plays a uniform or negligible role in gastric carcinogenesis, revealing instead that its influence is highly context-dependent.
Drawing on harmonized individual-level data from 20 prospective cohorts encompassing over two million participants and nearly 8,400 incident gastric adenocarcinoma cases, researchers found only weak evidence that heavy alcohol consumption (≥30 g/day) raises overall stomach cancer risk compared to light drinkers (HR 1.06, 95% CI 0.96–1.16). However, stratified analyses told a more nuanced story. Among never-smokers, heavy drinking was associated with a 20% elevated risk (HR 1.20, 95% CI 1.02–1.42), a statistically significant interaction suggesting that tobacco may mask or confound alcohol's carcinogenic signal in the gastric mucosa. A comparable 21% increased risk was observed across Asian study populations, potentially reflecting distinct genetic profiles in alcohol metabolism — particularly aldehyde dehydrogenase (ALDH2) polymorphisms prevalent in East Asian populations — that allow acetaldehyde to accumulate at higher concentrations. Anatomically, modest risk elevations appeared for proximal non-cardia subsites such as the fundus and body, but not for distally located non-cardia cancers.
This analysis is methodologically strong: prospective design, large sample, individual-level covariate harmonization, and long follow-up (9–29 years). Yet it remains observational, and residual confounding — particularly from Helicobacter pylori infection, diet, and socioeconomic factors — cannot be fully excluded. The subgroup findings, while plausible mechanistically, should be considered hypothesis-generating rather than definitive. Overall, this is an important confirmatory and clarifying contribution: alcohol's role in stomach cancer is real but conditional, concentrated in populations with specific genetic or behavioral profiles rather than universally elevated across all drinkers.
