Two U.S. cohorts put daily sugary drinks at 2.45 times the gastric cancer risk

The comparison here runs between two ends of a habit. One group drank at least one serving a day; the reference group drank less than one serving a month ^[1]. Since roughly 65 percent of U.S. adults report at least one sugar-sweetened beverage a day ^[8], that reference group is a thin and unusual slice of the country. Contrasting extremes is a sensible way to find a signal in a cohort study, and it means the 2.45 figure describes the gap between a common habit and a rare one.

Andrew T. Chan, the senior author, said in a press release, "This is the first study to demonstrate an association between sugar-sweetened beverage intake and gastric cancer in a U.S. population. Gastric cancer is the fifth leading cause of cancer death worldwide, but we've known very little about how diet might contribute to this cancer" ^[9]. The work appeared in Gastro Hep Advances ^[11].

The case count is what limits the paper. Two hundred and seventy-eight gastric cancers among 112,284 people works out to about one case per 404 participants ^[1]. Those 278 then have to be divided across intake categories and two cohorts, and the account in Discover reports no per-category case counts and no confidence intervals ^[16].

The researchers counted regular carbonated drinks, punch, lemonade and sports drinks as sugar-sweetened ^[10], and they ran the same analysis on low-calorie carbonated drinks, where higher intake was not linked to greater gastric cancer incidence ^[4]. Higher overall fructose intake, by contrast, came with more cases ^[5]. If the association were driven by carbonation, or by the sort of person who buys a lot of packaged drinks, the artificially sweetened beverages should have carried some of it.

Helicobacter pylori is the major bacterial risk factor for this cancer, and the team had infection status for 940 participants, under one percent of the cohort ^[6][2]. Slightly more than a third of that subset showed evidence of infection, so somewhat over 313 people ^[6][3]. Sugary drink intake was not associated with whether someone carried the bacterium ^[6]. That argues against the simplest confounding story, but it rests on data from fewer than one participant in a hundred.

The thing this does not tell you is who these drinkers are otherwise. People who drink sugary beverages often may differ from those who rarely do in ways the researchers could not fully measure ^[15]. The participants were predominantly white, which leaves it unclear whether the association holds across racial and ethnic groups, and information on family history of gastric cancer was limited ^[12]. Earlier work has tied sugary drinks to colorectal, breast and liver cancers ^[13]. The researchers recommend further studies to confirm the connection and to uncover the biological mechanisms behind it ^[7].