Eight preservatives turned up associated with higher blood pressure in a French cohort of 112,395 adults followed for a median of eight years. One of them was ascorbic acid. Yes, the molecule you call vitamin C.

And of the eight, it was the only one that also independently associated with cardiovascular disease after the statisticians corrected for multiple testing.

I read that table twice to make sure I had it right.

The NutriNet-Santé study, published this month in the European Heart Journal, is what the lead author calls the first study of its kind to look across a wide range of preservatives against cardiovascular health at population scale. Mathilde Touvier’s INSERM team and PhD candidate Anaïs Hasenböhler tracked 112,395 French adults from 2009 onward, capturing diet in three-day food diaries every six months and watching for new hypertension and new cardiovascular events. Roughly 5,500 participants developed high blood pressure; about 2,450 developed cardiovascular disease.

Then they ran the list against the seventeen preservatives that showed up in at least one in ten people’s diets. Eight came back associated with hypertension. Read the list and notice the shape of it:

Three of those are the names a journalism textbook would have predicted: nitrite, sulphite, sorbate, the workhorses of the cured-meat and processed-snack aisles. Four are molecules wellness culture would have bet money on as the safe, natural ones. Vitamin C twice (ascorbic acid and its sodium salt). The synthetic vitamin C isomer used as a meat-color preservative (erythorbate). And the active acid in lemon juice. The eighth is rosemary extract, marketed for years as the “natural antioxidant” option for replacing synthetic preservatives on a clean label. Naming a molecule after a herb you can grow on a windowsill does not appear to have insulated it from the statistical signal.

Sit with the vitamin C entry for a second, because it is the load-bearing weirdness here. Ascorbic acid was supposed to be the inert one. It is on every “safe additives” list. Its sodium salt, sodium ascorbate, is what processed-meat manufacturers add precisely to keep nitrite chemistry from doing the worst of its known damage. And both of them, plus their synthetic cousin erythorbate, plus citric acid, landed in the same flagged column as nitrite itself.

Now, before anyone reorganizes their pantry: the individual hazard ratios live in the modest-effect range. The headline numbers come from the group-level analysis. Among the heaviest consumers, the European Society of Cardiology release on the paper reports a 29% higher hypertension risk from non-antioxidant preservatives as a class and 16% higher cardiovascular disease risk from the same class, with antioxidant preservatives clocking in at 22% higher hypertension risk. At the level of single additives the CVD signal mostly washed out after correction for multiple comparisons, and ascorbic acid is the lone individual molecule that survived as an independent CVD association. The JAMA news brief summarizes it cleanly: eight of seventeen preservatives associated with hypertension, one with CVD incidence.

Preservative class risk signals in heaviest consumers
29
% higher hypertension risk from non-antioxidant preservatives
22
% higher hypertension risk from antioxidant preservatives
16
% higher CVD risk from non-antioxidant preservatives
Among heaviest consumers in the French cohort, both preservative classes flagged for hypertension – and the non-antioxidant class also for cardiovascular disease. Source: European Heart Journal / NutriNet-Santé (Hasenböhler et al., 2026)

So what is the biology actually doing? I came into the paper wanting that answer and the honest reply is that no one knows yet. The authors flag the same gap and call for the bench work to catch up. The textbook story for nitrite has a name, nitrosamine formation in the gut, but it does not explain why ascorbic acid is associated with the same outcome instead of protecting against it the way the food-chemistry literature predicts. The simplest explanation that fits the data is uncomfortable: the relevant exposure may not be the molecule alone but the additive cocktail it ships with, the sodium load it travels alongside, the ultraprocessed package it stabilizes, and the cumulative shelf-life engineering that keeps the whole food category alive on the shelf. You can’t pull the molecule out of the matrix it was eaten in.

That distinction matters because vitamin C the nutrient is not the exposure in this study. Nobody followed 112,000 people taking 500 mg supplements or eating bell peppers. The signal attaches to ascorbic acid as it shows up in your diet labeled E300, embedded in industrially processed foods, sitting alongside seven other E-numbers from the flagged list. The food is the exposure. The additive is the marker for it.

Then Touvier said the quiet part. “We need a re-evaluation of the risks and benefits of these food additives by the authorities in charge, such as the EFSA in Europe and the FDA in the USA, for better consumer protection.” Translation, in academic-French politeness, to the regulators: you signed off on these additives without long-term cardiovascular outcome data, the long-term cardiovascular outcome data have arrived, and your sign-off is now obligated to defend itself in public.

Worth noticing who paid for the work. NutriNet-Santé runs on the French Ministry of Health, INSERM, INRAE, CNAM, Sorbonne Paris Nord, the European Research Council, and the French National Cancer Institute. The authors disclose no for-profit funding and no conflicts. A publicly funded cohort just produced a finding the regulated companies will not enjoy reading, with no industry consulting fees in the margin notes.

Now the limits, said honestly. This is observational. Food intake is self-reported. The cohort is French internet users who chose to join a nutrition study, which means it skews more health-conscious, more educated, and more female than the general population. Heavy preservative consumers also eat more ultraprocessed food in general, weigh more on average, exercise less, and sleep worse, and even the most careful statistical adjustment can’t fully separate the additive from the package it arrived in. The authors corrected for the usual suspects. They cannot correct away the fact that these molecules do not travel alone.

But the confounding is not a reason to shrug, because the additive approvals were granted on safety files that assumed exactly the kind of narrow exposure the matrix problem rules out. If a regulator’s “safe at approved exposures” file was built on tox studies dosing the isolated molecule into rodents, and the population-scale signal is about the molecule eaten as part of an ultraprocessed cocktail, the original safety file did not test the relevant exposure. That is the gap Touvier is pointing at. That is the file she is asking EFSA and the FDA to reopen.

What I’ll be watching: whether anyone runs the food-matrix experiment, dosing ascorbic acid at realistic preservative concentrations inside an industrial-food matrix rather than as an isolated mg/kg dose, and whether EFSA actually opens a docket or runs the clock until the press cycle moves on. A publicly funded cohort of 112,000 people just put eight additives on a list and named one of them an independent cardiovascular signal. The regulators are on the clock, and silence will be its own answer.

Sources

  1. European Heart Journal – Hasenböhler et al., NutriNet-Santé: preservative food additives and incident hypertension and cardiovascular disease (advance article, 2026)
  2. European Society of Cardiology – Common food preservatives linked to high blood pressure and heart disease (press release; includes Touvier quote on EFSA and FDA re-evaluation)
  3. JAMA – Anderer, “Some Common Food Additives Linked to Hypertension and CVD” (June 12, 2026)
  4. ScienceDaily – Researchers found 8 common food additives linked to high blood pressure and heart disease (June 17, 2026)