In a 48-person randomised trial with calories held equal, the vegan arm showed shifted immune and inflammatory methylation — and two of three clocks read it as slower aging.

Diet studies are usually confounded by weight. A diet that changes what people eat also changes how much they weigh, and the resulting biology could follow from either — which is why a trial that holds calories constant is worth more than one that does not.

A randomised trial in MedComm from the University of California San Diego and the University of Freiburg does hold them constant. Forty-eight healthy adults ate the same standardised diet for a week, then were randomly assigned to a vegan or a meat-rich diet for a month, with the two designed to supply similar calories so that diet composition could be separated from weight loss.

The outcome measured was not weight or cholesterol but the epigenome: more than 800,000 DNA methylation sites across the genome, which is where a cell records which genes it is inclined to read.

The check that makes the rest credible

Methylation studies have a standing problem, which is that blood is a mixture of cell types and a methylation signal can reflect a change in the proportions of cells rather than a change inside them. This study turned that into a validation.

From methylation alone, the vegan diet predicted lower neutrophil proportions and higher CD4 T cell proportions. Those predictions closely matched the blood cell counts measured directly in the same trial.

“That agreement told us we weren’t just seeing changes on a computer screen,” said Lukas Karbacher, the paper’s first author and a neuroscience graduate student at the UCSD School of Medicine.

It is the most persuasive element in the paper. An epigenetic inference that independently reproduces a measurement taken by a different method is evidence that the methylation signal is tracking something real, which is not something most methylation analyses can show.

The direction is also coherent with the inflammation framing. Neutrophils are the first-line inflammatory cells; CD4 T cells are the coordinating arm of adaptive immunity. A shift from the former toward the latter is the pattern you would predict if a diet were reducing low-grade inflammation.

The pathways

Beyond cell composition, the vegan diet was associated with increased promoter methylation in several cancer-related and cell growth pathways — and since methylation at a promoter generally reduces how much a gene is read, that points toward those pathways being turned down.

Elsewhere the team observed reduced activity in lipid metabolism and increased activity in insulin signalling, DNA repair and cellular stress responses.

“Our genes are not our destiny,” said Jerome Mertens, a co-senior author and associate professor of neurosciences at UCSD. “The remarkable finding isn’t simply that one diet outperformed another. It’s that the epigenome responded measurably in just four weeks, showing that our biology is far more dynamic.”

Max Storz of the University of Freiburg Medical Center, the other co-senior author, put it more narrowly: “Our study adds to growing evidence that nutrition can shape biology at the molecular level.”

Pathway results of this kind need a caveat the release does not supply. When methylation changes are modest and distributed across many of 800,000 sites, the step from site-level data to a named pathway runs through enrichment analysis, which involves choices about thresholds and gene sets. Different defensible choices can surface different pathways from the same data, and the release reports no statistics by which to judge how robust any particular pathway is.

Three clocks, two answers

The aging result is the one the headline rests on, and the interesting part is that it is not unanimous.

Two epigenetic clocks designed to predict health outcomes showed slower biological aging in the vegan group. A third clock, built to predict chronological age, showed a different pattern.

“These findings reinforce the idea that not all measures of biological aging capture the same biology,” Mertens said.

That is the honest reading and it should temper the finding rather than decorate it. Epigenetic clocks are statistical models trained to predict something from methylation, and what they predict depends on what they were trained on. A clock fitted to chronological age learns the marks that drift with time; a clock fitted to mortality or disease risk learns marks associated with being unwell. They are different instruments, and in this trial they disagreed.

There is also a basic interpretive difficulty with running any of them over four weeks. These clocks were built on cross-sectional data spanning decades of human life, and their behaviour over a one-month dietary change is not something they were validated for. A shift in a clock reading after four weeks is a shift in a biomarker, not a demonstrated change in the rate at which a body is aging.

What the authors say it does not show

The research team is explicit, and their caveats are worth reporting as they gave them. The study was small, lasted one month, and the methylation changes were modest and spread across many sites. The authors state that it does not show that a vegan diet prevents cancer or reverses aging.

Those limits have teeth. Forty-eight participants split between two arms is roughly 24 each, which is a small sample for genome-wide analysis and smaller still for subgroup patterns. The participants were healthy adults, so the findings describe what a diet does to a well population rather than what it does to disease. And no clinical outcome of any kind was measured — no disease, no event, no symptom. Every result is a molecular intermediate.

The isocaloric design is a genuine strength and it also narrows the question. By removing weight change, the trial isolates composition — which is scientifically cleaner and further from the real world, where changing to a plant-based diet usually does change weight, and where the weight change may be doing much of the work.

Nothing here is a reason to change what anyone eats, and the authors do not suggest otherwise. What the trial establishes is that a month is long enough for diet to leave a readable mark on the epigenome, which is a statement about the responsiveness of the measurement rather than about anybody’s health.

The other authors are Stefanie Kowarschik, Manching Ku and Roman Huber of University of Freiburg Hospital, and Ann-Kathrin Lederer of the University Medical Center of the Johannes Gutenberg University Mainz. Lederer’s position within the Academic Center for Complementary and Integrative Medicine was financed by the Ministry of Science, Research and Arts of Baden-Württemberg; the rest of the study was funded from institutional resources. The authors declare no conflicts of interest, and the paper was published on 3 August 2026.

---

Notes

1. Karbacher, L., Kowarschik, S., Ku, M., Huber, R., Lederer, A.-K., Storz, M., Mertens, J. “A vegan diet epigenetically modulates inflammatory pathways and biological aging: Genome-wide DNA methylation analysis of a one-month isocaloric vegan versus meat-rich dietary intervention.” MedComm, 2026. DOI: 10.1002/mco2.70899.
2. University of California – San Diego. “One month on a vegan diet shifts epigenetic patterns associated with aging and inflammation.” EurekAlert!, 4 August 2026. https://www.eurekalert.org/news-releases/1138813.

Photo by Daria Shevtsova on Pexels.com