INDEPENDENT THINKING. WHOLE-FOOD LIVING.NO ANIMALS / NO OIL
THENANODIET
← Oncology / Health Desk

Garlic, onions and gastric cancer: a large association with important limits

A meta-analysis links higher allium intake with lower gastric-cancer odds, but its evidence is mostly case-control.

Whole-food ingredients prepared without added oil

Illustrative food image, not a photograph of the study intervention.

What the review actually compared

The question in this 2011 meta-analysis was specific: was consumption of allium vegetables associated with gastric cancer? The researchers combined 19 case-control studies and two cohort studies, covering 543,220 people. Their search extended from January 1966 through September 2010, and the pooled analysis used a random-effects model. This was not a trial assigning people to eat garlic or onions. Its evidence concerned differences in reported food consumption and gastric-cancer occurrence across observational studies. Keeping that design in view matters before interpreting either the large headline association or the smaller dose-response estimate.

Zhou Y, Zhuang W, Hu W, Liu GJ et al.. Gastroenterology, 2011. PMID 21473867 · DOI

A striking highest-versus-lowest comparison

Across the included studies, the highest allium-consumption groups had a pooled odds ratio of 0.54 compared with the lowest groups, with a 95% confidence interval of 0.43–0.65. That is a comparison of odds, not a statement that eating garlic prevents 46 cancers in every 100 people. The abstract does not supply an absolute risk reduction for an individual eater. Analyses of garlic, onion, leek, Chinese chive, scallion, garlic stalk and Welsh onion showed similar directions, while onion leaf was the exception. The pooled result concerns gastric cancer, not every cancer site.

Zhou Y, Zhuang W, Hu W, Liu GJ et al.. Gastroenterology, 2011. PMID 21473867 · DOI

The dose-response number is not a prescription

The review also estimated an odds ratio of 0.91 for each additional 20 grams per day of allium vegetables, with a 95% confidence interval of 0.88–0.94. Importantly, that dose-response calculation came from case-control studies. It should therefore be read as a modeled association within that evidence, not an experimentally established serving target. The broad allium category also matters: the estimate is not proof that the same amount of garlic alone produces a specific benefit. Nor does the abstract establish whether one preparation method or an allium supplement would reproduce the association.

Zhou Y, Zhuang W, Hu W, Liu GJ et al.. Gastroenterology, 2011. PMID 21473867 · DOI

Why the authors asked for more research

The authors explicitly identified potential confounding and exposure misclassification as reasons for further study. Those limitations belong beside the numerical findings, not in small print after a promise of prevention. A synthesis can pool a large number of people without turning observational food comparisons into a randomized experiment. Here, 19 of the 21 included studies were case-control studies, and the dose-response estimate also relied on that design. The defensible takeaway is a gastric-cancer association worth investigating. It is not a demonstrated treatment effect, a supplement recommendation, or a quantified guarantee from adding onions to dinner.

Zhou Y, Zhuang W, Hu W, Liu GJ et al.. Gastroenterology, 2011. PMID 21473867 · DOI