Glyphosate, anxiety and mood: two independent studies, one oddly specific signal
A 2026 study found that glyphosate at legal exposure levels changed behaviour in mice, and proved the gut was the route. An independent human study points at the same narrow target. The reassuring review, meanwhile, was funded by Bayer.
Glyphosate, the active ingredient in Roundup and the most widely used herbicide on earth, turns up regularly in posts claiming it is quietly driving anxiety and depression. Usually with a real citation attached.
This one is worth walking through carefully, because the science underneath is better than the usual viral-claim fare, and the honest answer is genuinely unresolved rather than simply debunked.
The new study is not junk
In June 2026, a team at University College Cork published work in Molecular Psychiatry that is hard to wave away.
They exposed mice to glyphosate at regulatory acceptable daily intake levels: the amount regulators consider safe, not the enormous doses that make toxicology studies easy to dismiss. Male mice showed two things: impaired social novelty preference, and increased anxiety-like behaviour. Females showed reduced movement but few other robust changes.
The anxiety finding is not incidental. Gene expression in the amygdala, the region central to fear, threat appraisal and social processing, shifted in ways consistent with the behaviour. Other rodent work has reported similar anxiety-like effects under chronic exposure, so this is not a lone result.
Then they did the step that makes the study interesting. They transplanted gut microbiota from exposed mice into unexposed ones, and the social impairments travelled with the bacteria.
That is not a correlation. Within mice, it is a demonstration that the gut microbiome was the route by which glyphosate changed behaviour. It fits a growing framework its authors call ecological psychiatry: the idea that environmental chemicals may reach the brain through the microbiome rather than directly.
The human finding that matches
The mouse result would be easy to shelve as rodent work, except that an independent human study found something pointing the same way.
The ESPINA cohort, run by researchers at UC San Diego, NIH-funded and published in Environmental Health Perspectives, measured urinary glyphosate in adolescents living in Ecuadorian agricultural communities. It was detectable in 98.3% of them. And glyphosate showed a statistically significant negative association with exactly one domain: social perception, the ability to read emotion in other people.
Hold that next to the mice, who lost their preference for social novelty. Two independent teams, two species, converging on social processing. Not mood in general, not memory, but the specific business of engaging with other people.
That is a narrower and more interesting signal than “glyphosate causes depression,” and it is the kind of thing that disappears when a review pools every neurological outcome together and reports an average.
What about the studies showing nothing?
You will also find a prominent systematic review concluding that the highest-quality human studies show no association between glyphosate and depression. It is worth reading, and so is its funding statement.
It was sponsored by the Glyphosate Renewal Group, a consortium of glyphosate manufacturers including Bayer and Syngenta, contracted through Bayer AG. Its authors work for Exponent, a consultancy retained by those companies, and one co-author was employed by Monsanto for fifteen years.
That does not make its analysis wrong. But in a systematic review, the decisive choice is which studies count as high quality, and that choice sits with the authors. Industry-sponsored reviews reach favourable conclusions more often than independent ones across pharmaceuticals, tobacco and chemicals alike; it would be strange to assume this literature is the exception, particularly one where court disclosures have already shown manufacturer involvement in ghostwriting published reviews.
So: treat that null result as one input from an interested party, not as the settled answer.
Where the honest uncertainty actually sits
Other human data is genuinely mixed. NHANES analyses of US adults link urinary glyphosate to lower cognitive scores, and a follow-up in older adults found raised odds of recall and fluency impairment. But those are cross-sectional, and glyphosate clears the body in hours, so a single urine sample reflects roughly yesterday’s diet rather than years of exposure.
You will also see claims that farmers and farmworkers report more anxiety and mood disturbance than unexposed controls. They do. But agricultural work also comes with financial precarity, physical exhaustion, isolation, and exposure to many other chemicals. ESPINA is instructive here: 2,4-D was associated with more neurobehavioral domains than glyphosate was. If something in agricultural life raises anxiety, glyphosate has not been shown to be the agent.
What nobody has yet shown is the thing the viral posts assert: that glyphosate at dietary exposure levels causes depression or an anxiety disorder in humans. The study that would demonstrate it, following people for years, measuring exposure repeatedly and tracking mood, has not been done.
The mechanism people reach for, and why we are not leaning on it
The usual explanation runs: glyphosate blocks the shikimate pathway, gut bacteria use that pathway to make tryptophan, tryptophan makes serotonin, low serotonin causes anxiety and depression.
The first half is sound. Glyphosate does inhibit the shikimate pathway, many gut bacteria do rely on it, and disrupting the microbiome is a real effect. The microbiota-transfer experiment is the best evidence we have that the gut is the route.
The second half does not hold up. Tryptophan is an essential amino acid, which means humans get it from dietary protein, not from bacterial manufacture. How much reaches your brain depends mostly on what you eat and on competition with other amino acids at the blood-brain barrier, not on gut bacterial synthesis.
And the last link is weaker still. A 2022 umbrella review in Molecular Psychiatry examined serotonin concentrations, receptor binding, transporter levels and tryptophan-depletion studies together, and found no consistent evidence that depression involves lowered serotonin. That review is contested, and dozens of researchers published a rebuttal, but "low serotonin causes depression" is no longer a fact you can build an argument on.
Which matters here, because it is the plank most glyphosate-and-mood explanations rest on. The gut route looks real. The serotonin story attached to it is doing more work than the evidence allows.
How to hold both things at once
The mouse work is strong evidence of a mechanism. The human work is reasonable evidence that the mechanism is not producing a large, detectable effect on depression at real-world exposures.
A pathway can exist in a mouse and still be too small, too compensated-for, or too swamped by everything else to shape a human life. That is an ordinary situation in biology. What is unusual here is that the independent human and animal work agree on an oddly specific target, social behaviour, while the reassuring summary comes from the people who sell the product.
None of that proves harm. It does mean “no evidence of an effect” is a weaker statement than it first appears, and that the people best placed to fund the definitive study have little reason to.
If you are worried about it anyway
Reducing pesticide exposure is a reasonable personal choice, and washing produce and eating more whole foods are good ideas for reasons entirely unrelated to glyphosate.
What we would gently push back on is the leap to this is why I feel the way I feel. When you are depressed, a single clean external cause is an appealing story, and an expensive one, if it redirects you away from things with much better evidence behind them. Sleep, thyroid function, iron and B12 status, alcohol, and actual treatment all have far stronger links to mood than anything in the glyphosate literature.
If your gut symptoms are real, the gut-brain connection is worth understanding properly, and worth investigating with your doctor rather than by eliminating foods on the strength of a headline.
Mura is a screening and education tool, not a diagnosis. We flag patterns worth discussing with a clinician, and we try to be honest about the difference between a mechanism that exists and an effect that has been shown in people.
This article is for general information and is not a substitute for professional care. If you are in crisis or thinking about harming yourself, please contact your local emergency services or a crisis line. See crisis lines.