A weekly read on the gut, the brain, and the wiring in between. One clinical pearl, one paper, one product, one person. No hype, no selling, just the stuff I actually think about.
1. Practice
A molecule you do not make, in four diseases you have heard of.
Histidine is an amino acid, one of the building blocks of protein. It is in eggs, meat, fish, dairy, beans, and whole grains, and you cannot live without it. Your own cells break it down one way. Some of the bacteria living in your gut break it down another way, and their version leaves behind a compound called imidazole propionate. ImP for short. You do not make it. Something living in you makes it, out of a nutrient you cannot stop eating.
For most of medical history that would have been trivia. Then people started measuring it.
In 2018, Ara Koh and Antonio Molinaro, working in Fredrik Bäckhed's lab in Sweden, found that people with type 2 diabetes carry more ImP in their blood (Cell, 2018). Then they showed why it matters. ImP jams the signal insulin uses to tell your cells to take up sugar. Fed to mice, it made their blood sugar control worse. A waste product from a gut microbe, interfering with a hormone.
In 2025, a team in Madrid moved the story to the artery wall. They gave ImP to mice prone to artery disease, kept them on ordinary food, and the mice built plaque anyway, with no change in their cholesterol (Nature, 2025). The damage ran through inflammation. ImP flips a switch on immune cells, and blocking that switch stopped the plaque from forming. Plaque, with a clean cholesterol panel, driven by a molecule no standard blood test will ever see.
Two weeks ago the human outcome data landed. Across roughly 7,000 patients with coronary artery disease in four European studies, people with high ImP died more, had more heart attacks, and had more strokes, even after accounting for every standard risk factor (European Heart Journal, 2026). In one group, the highest levels more than doubled the risk of dying.
And late last year, a lab at UCLA showed the molecule does not stop at the body. ImP gets into the brain and changes activity in the hypothalamus, the deep region that runs hunger and stress (Cell Host Microbe, 2025). Mice with more ImP showed more stress behavior. In people, higher levels went along with worse stress coping and more emotional eating.
That is diabetes, artery disease, cardiovascular death, and stress. One molecule, made by a tenant.
Now the part almost everyone gets wrong.
You cannot order this test. ImP is measured on research equipment after special sample handling. There is no version your doctor can order, and no home kit worth your money. Anyone selling you one is ahead of the science.
The answer is not to eat less protein. This is the first place everyone goes, and the data say no. In a study of nearly 2,000 Europeans, the people with high ImP were not eating more histidine than anyone else (Nature Communications, 2020). What set them apart was the rest of the plate: less fiber, fewer vegetables and nuts, more saturated fat, mostly from cheese. The signal is not how much histidine goes in. It is who gets to it first.
And the microbe does not need to be common. Federico Rey, one of the senior scientists on this week's paper, put it plainly: the bacteria that make ImP live in a large fraction of people, but in small numbers, and a microbe does not have to be abundant to matter. We tend to picture the microbiome as a census, as if the answer is in the percentages. Sometimes the answer is one rare organism running one chemical reaction nobody else in the room can run.
2. Publication
One paper worth reading, in three sentences.
Researchers at the University of Wisconsin measured ImP in the blood of 1,196 adults whose memory and thinking tested normal, average age 61, and the people with more of it scored lower on tests built to catch the earliest slipping, and carried more of two blood markers that rise when brain cells are being damaged (Vemuganti, Kang, and colleagues, Nature Communications, 2026). The paper added two more legs: a common gene variant that raises blood ImP has also been tied to Alzheimer's risk in large genetic studies, and in mice, ImP in the drinking water meant more amyloid plaque in one Alzheimer's model, more of the tangled tau protein in another, and a leakier barrier between the bloodstream and the brain in ordinary animals. Hold it honestly, because the authors do: this shows a link, not proof of cause, the volunteers were mostly white and mostly women, and when I asked the first author whether the kidney explanation behind that gene finding was backed by kidney data yet, he told me straight out that it is not, that it is a working hypothesis, and that better data is a year or more away.
That last answer is why this is the pick.
3. Product
This week's pick: a bag of mixed nuts. Unsponsored, no affiliate relationship, no brand, and obviously not a BellyMD product.
Vegetables and nuts were the two food groups most clearly missing from the high-ImP diet in that European study. Nuts are also the food group with a real randomized trial behind them. In Spain, the PREDIMED trial assigned 7,447 adults at high heart risk to a Mediterranean diet with olive oil, the same diet with mixed nuts, or a low-fat control diet. The nut group had about a 28 percent lower rate of heart attack, stroke, or death from heart disease (New England Journal of Medicine, 2018).
Two honest notes, and the first matters more than the pick. That trial was published in 2013, pulled back by its own authors in 2018 when they found enrollment problems at a few sites, reanalyzed the hard way, and republished. The result held. A trial that caught its own error and survived the correction is worth more than one nobody ever checked.
Second: nobody has shown that eating nuts lowers ImP. The pick stands on the diet pattern being the same one, the trial being real, and the downside being a bag of almonds.
4. Person
One person worth following in this space.
Vaibhav Vemuganti, a graduate student at the University of Wisconsin-Madison, and first author on the paper above. He works between two labs that do not usually share a hallway: Federico Rey's, which studies gut bacteria, and Barbara Bendlin's, at the Wisconsin Alzheimer's Disease Research Center, which has been collecting blood samples and memory tests from the same volunteers for years. Neither lab could have produced this paper alone.
The senior authors have the titles. Vemuganti gets this slot because of how he showed up. I made a short video about his paper, and he found it and wrote to me first, with a thank-you, his mentor copied in, and an offer to answer questions. He answered every question in under a day, sent me the exact page where the gene data lives so I could look for myself and then, unprompted, told me what people were getting wrong about his own work: readers were concluding they should stop eating protein, and the data do not say that.
One more thing, and it is the reason the slot stayed his. When he read a draft of this issue, he wrote back asking me to add other people's names. Jea Woo Kang, at the Wisconsin Alzheimer's Disease Research Center, led the work connecting ImP to memory and blood markers in living people. James Hilser and Hooman Allayee, at UCLA, brought the genetics behind the gene finding. And he wanted credit paid upstream, to Fredrik Bäckhed's group in Sweden, where Ara Koh and Antonio Molinaro did the original work that made anyone care about this molecule at all.
He had a blank slate to spread his message and he used it to name six other people.
He’s reachable, too. He is on LinkedIn at linkedin.com/in/vaibhav-vemuganti-a7b79a183, and he wants questions from anyone who reads this. Take him up on it. The first author of the paper you just read about is almost never someone you get to ask.
A scientist who tells you what his own paper does not show, and who hands the credit to other people, is worth ten who do neither.
Barbara Bendlin, the other senior scientist, described where she wants this to go, and it is the most useful sentence in all the coverage. If someone finds a drug that lowers ImP, this becomes cholesterol: a number, with a drug, and a decision.
We are not there. But notice what had to be true for anyone to even ask. ImP was named in 2018. The Wisconsin volunteers had been giving blood and taking memory tests for years before that, back when nobody knew this molecule existed, let alone that it was worth measuring. The discovery did not need a new study. It needed a record that already existed, sitting in a freezer, waiting for a question nobody had invented yet.
That is the entire argument behind MeNome, our revolutionary patient symptom tracking app. We cannot tell you what your daily symptoms will turn out to mean. Nobody can yet. That is exactly why the record has to start now…you cannot go back and measure a year you did not record.
That's the four. See you next week.
Rick
From BellyMD
MeNome is our free app for keeping your own record. A minute a day on your gut, sleep, mood, and energy, and over time it shows you what shows up together. It lives at menome.belly-md.com.
I also make MGB+, a small line of gut-brain supplement formulas: Clear, Cool, and Calm, at belly-md.com. Structure-function support, not a treatment for any disease.
If you want to talk through your own gut-brain picture with me directly, I offer a gut-brain consultation. Details at belly-md.com.
Not medical advice. This newsletter is for education and does not create a physician-patient relationship. Nothing here is a diagnosis or a prescription. Talk to your own physician before starting, stopping, or combining any medication or supplement, including over-the-counter ones.