Editorial NexoScience
Think about the last time a doctor or a survey asked you how much you drink. Chances are you rounded down, even without quite meaning to. It turns out that small habit has been quietly distorting decades of research into one of nutrition science’s most debated questions: does wine actually protect the heart?
PREDIMED, short for PREvención con DIeta MEDiterránea, is the largest randomized trial ever run to test the Mediterranean diet against cardiovascular disease, following participants at high cardiovascular risk for a mean of 4.8 years. Moderate wine drinking has long been treated as one of the diet’s cardioprotective ingredients, but nearly every study behind that idea leaned on people accurately reporting their own alcohol intake on a food frequency questionnaire, information that is notoriously skewed by memory, embarrassment and the simple wish to look healthier on paper.
A team of Spanish researchers decided to stop asking and start measuring. Tartaric acid is produced almost exclusively by grapes and rarely turns up anywhere else in the diet, which makes it a reliable short-term fingerprint of wine consumption, showing up in urine within days of drinking it. Using mass spectrometry, the researchers could measure it directly in urine samples, sidestepping self-report bias entirely.
The study drew on 1232 PREDIMED participants with a mean age of 68, a nested group made up of 685 people who went on to develop cardiovascular disease and a random subcohort of 625 others, with some overlap between the two. Over an extended follow-up with a median of nine years, the wider trial population racked up 222 cases of heart failure, 190 strokes, 138 non-fatal heart attacks and 286 cardiovascular deaths.
The biomarker matched what people actually drank, and then some. Urinary tartaric acid correlated with self-reported wine consumption, and once other factors were folded in, it predicted whether someone was a wine drinker with close to 80 percent accuracy. That gave the researchers confidence it was tracking real intake rather than something else entirely in the diet.
Moderate levels came with a real drop in cardiovascular disease. Participants were split into five groups by their baseline urinary tartaric acid. Those with 3 to 12 micrograms per milliliter, roughly equivalent to light wine consumption, had 38 percent lower cardiovascular disease risk than participants with almost none detected, though that result sat right at the edge of statistical significance. Those with 12 to 35 micrograms per milliliter, corresponding to moderate drinking of up to about a glass a day, had a clearer 50 percent lower risk. Both held up after adjusting for age, sex, smoking, physical activity, weight, blood pressure, cholesterol, diabetes and overall diet quality excluding wine itself. Looking at individual outcomes, the drop was strongest for heart attacks specifically.
But ask the same people what they drank, and the protective pattern disappears. When the researchers ran the identical analysis using the self-reported questionnaire data instead of the biomarker, people who said they drank the least wine showed no lower cardiovascular disease rate than heavier self-reported drinkers. The authors point to underreporting as the likely reason, with some heavier drinkers misclassifying themselves into the light or moderate categories and muddying the comparison.
Notably, the protection did not extend to the heaviest drinkers. Participants excreting more than 35 micrograms per milliliter of tartaric acid showed no significant reduction in cardiovascular disease risk compared to those with almost none, a pattern consistent with older research describing a J-shaped relationship between wine and cardiovascular mortality, where benefits cluster at light-to-moderate intake and vanish at higher doses.
None of this proves wine itself is protecting anyone’s heart. This was an observational study, and the authors are careful to note that a urine biomarker cannot fully separate wine’s own effect from the broader Mediterranean lifestyle that tends to come with it. Genetic studies designed to sidestep exactly this kind of confounding, known as Mendelian randomization studies, have generally found no protective link between alcohol and cardiovascular disease, a contradiction this paper does not resolve. Because PREDIMED enrolled older adults in Spain who already had elevated cardiovascular risk, the findings may not extend cleanly to other populations either.
The leading theory for why wine specifically, rather than alcohol in general, might matter involves polyphenols, the plant compounds responsible for much of wine’s color and flavor. Malvidin, procyanidin, catechin and tyrosol are the main ones found in wine, and earlier research within the same PREDIMED cohort has already linked higher polyphenol intake to fewer cardiovascular events on its own, independent of ethanol.
The study, “Urinary tartaric acid as a biomarker of wine consumption and cardiovascular risk: the PREDIMED trial,” was led by Inés Domínguez-López and Ramon Estruch and published in the European Heart Journal (DOI: 10.1093/eurheartj/ehae804).
