TMAO: One Molecule, a Whole-Body Conversation

TMAO is the molecule the internet loves to hate: the red-meat compound, the one that clogs your arteries. Then a 2026 paper caught it driving a skin disease instead. The twist is not that TMAO is worse than we thought. It is that no molecule is a villain on its own.

Lit Review Friday · TMAO: One Molecule, a Whole-Body Conversation · Published July 2026 · 15 min read

📝 In short

Is TMAO bad for you? It depends entirely on context. High circulating TMAO is associated with cardiovascular disease (Witkowski et al., 2020, Circ Res), a new paper ties it to eczema through the Th2 immune pathway (Yu et al., 2026, Immunity), and yet in one stroke study TMAO was actually lower than in healthy controls (Yin et al., 2015). The dose and the setting decide the outcome.
Do eggs or red meat cause eczema or heart disease? The food is not the villain. Whether the choline in an egg becomes a flood of TMAO depends on which gut bacteria process it (Yu et al., 2026). Same food, different gut, different result.
Can you lower TMAO? In mice and in human high-TMAO producers, allicin from raw garlic lowered it by shifting the gut bacteria doing the conversion (Panyod et al., 2022, NPJ Biofilms Microbiomes). Designed molecules can block the bacterial enzyme directly (Wang et al., 2015, Cell). Neither is a prescription.
What does any of this have to do with the gut? TMAO is a microbial signal: gut bacteria make its precursor from your food, and a diverse, balanced gut keeps that signal in proportion. The harm shows up when the ecosystem is depleted and lopsided.

Learn Something Weekly Podcast Listen: Episode 28

Also on Spotify.

The Molecule the Internet Loves to Hate

Every so often a single molecule gets cast as a villain, and the casting sticks. Trimethylamine N-oxide, TMAO, is one of those. If you have read anything about heart health in the last decade, you have met it as the bad guy: the compound your gut makes from red meat and eggs, the one linked to clogged arteries. And that reputation is partly earned. But I have a bias I will admit up front. Biology almost never builds machinery for no reason. If your body has enzymes dedicated to producing a molecule, that molecule usually has a job. So whenever something gets flattened into pure villain, I get suspicious that we are missing half the story.

A 2026 paper by Yu and colleagues in Immunity is what pulled me back to TMAO. They were not studying the heart. They were studying eczema, and they traced an itchy patch of skin all the way back to this same molecule. Laying the heart story and the skin story side by side does something useful. It stops TMAO from being a cartoon, and it surfaces the real lesson, which is not about one metabolite at all. It is about context.

Where TMAO Actually Comes From

Start with the supply chain, because it is the whole point. You eat a food containing choline or L-carnitine. Choline is abundant in eggs, fish, poultry, and legumes. Carnitine is concentrated in red meat. Both are normal and, in the case of choline, genuinely essential: your cell membranes, your liver, and the neurotransmitter acetylcholine all depend on it. In a balanced gut, that choline is quietly absorbed in your small intestine and used exactly as intended.

The fork in the road is microbial. Certain gut bacteria carry an enzyme called CutC (choline trimethylamine-lyase) that intercepts choline and cleaves off a fragment, releasing a gas called trimethylamine, or TMA. The TMA is absorbed, travels to the liver, and liver enzymes oxidize it into the stable molecule TMAO, which then circulates throughout the body. So TMAO is not something your food contains. It is something a particular configuration of your microbiome manufactures from your food.

⚠️ The food is not the villain The reflex when you hear "a nutrient in eggs feeds bacteria that make an inflammatory molecule" is to throw out the eggs. Please do not. Choline is an essential nutrient, and restricting it carries real risks to liver and brain function. The point of this whole body of research is the opposite of food fear: the same egg is nourishing in a healthy gut and a problem in a dysbiotic one. The variable is the ecosystem, not the breakfast.

Does TMAO Cause Heart Disease?

This is the famous chapter, and it is the most solid. Over more than a decade, Stanley Hazen's group at the Cleveland Clinic built the case that elevated circulating TMAO tracks with atherosclerotic cardiovascular disease, with the buildup of arterial plaque, and with platelets that clot too readily (Witkowski et al., 2020). This is some of the best-characterized microbial-metabolite biology we have, moving from association to mechanism through animal models and human cohorts. TMAO earned its scary reputation honestly. The prospective human data is what gives that reputation teeth: in 4,007 patients followed for three years, the highest-TMAO group carried roughly 2.5 times the risk of heart attack, stroke, or death, and TMAO predicted that risk even after adjusting for traditional risk factors, and even in people who looked low-risk on paper (Tang et al., 2013).

It is worth being precise about how solid, though, because honest science holds its own tensions. Some large analyses note that adjusting for kidney function weakens the TMAO-to-disease association, since TMAO is cleared by the kidneys, and at least one genetic (Mendelian randomization) analysis did not find a clean causal signal for cardiovascular disease (Canyelles et al., 2023). So the most defensible statement is that TMAO is a strongly associated, mechanistically plausible contributor to cardiovascular risk, not a settled single cause. Hold that nuance. It matters for what comes next.

📊 How TMAO is thought to harm arteries The proposed mechanisms cluster in three places: it alters cholesterol handling in artery-wall macrophages (foam-cell formation), it promotes endothelial dysfunction and inflammation in the vessel lining, and it makes platelets hyper-reactive, raising clotting risk. Each is a distinct pathway, which is part of why the molecule draws so much attention.

The Plot Twist: TMAO and Your Skin

Now the Yu 2026 paper. The researchers started in mice engineered to lack a single immune sensor in the gut lining, a receptor called TLR4. Without that sensor, the gut community shifted. The mice lost a keystone bacterium, Akkermansia muciniphila, the species that tends the protective mucus layer, and the CutC-carrying, choline-cleaving bacteria moved into the vacancy. TMAO went up. So far this is the same supply chain, just tilted toward overproduction.

What happens next in the skin is the new part, and it reads like a lock-and-key sequence. TMAO binds a host protein called PPP5 (protein phosphatase 5). That binding enhances PPP5's activation of a much larger regulator, PPAR gamma. And activated PPAR gamma reprograms naive immune cells into Th2 cells, the branch of the immune system that runs the allergic program. Th2 cells pump out interleukin-4 (IL-4), which drives IgE antibody production, which ends in the mast-cell, histamine-soaked, red and itchy reality of atopic dermatitis.

📊 The chain, link by link Gut dysbiosis lowers Akkermansia and raises CutC bacteria → more choline becomes TMA → the liver makes more TMAO → TMAO binds PPP5 → PPP5 activates PPAR gamma → naive T cells become Th2 cells → IL-4 rises → IgE rises → skin inflammation. The researchers proved each link was load-bearing: delete the PPAR gamma switch in T cells, and even mice with high TMAO did not get inflamed skin. The molecule is a lever, not a bystander.

Then they looked for the human fingerprint. People with atopic dermatitis had higher plasma TMAO than healthy controls, and the higher it ran, the worse their disease severity (measured on the clinical SCORAD scale) and the higher their IgE. Reaching for population scale, they used the UK Biobank and found that higher dietary choline intake was associated with a modestly higher risk of developing atopic dermatitis over more than a decade of follow-up.

⚠️ Mouse mechanism, human fingerprint The clean, causal, link-by-link proof is mouse work, and a mouse is not a person. What exists in humans is correlation: higher TMAO alongside more severe eczema and higher IgE, and a population-level association between choline intake and eczema risk. That is a real, repeatable fingerprint. It is not proof that lowering your TMAO will clear your skin. Both halves are true at once.

Why This Proves TMAO Is Not a Blunt Instrument

Here is the experiment that changed how I read the whole molecule. The same team took mice prone to psoriasis, a different skin disease driven by a different immune branch (Th17, with IL-23 and IL-17 signaling), and flooded them with TMAO. Nothing happened. The psoriasis did not worsen at all. TMAO has the key only for the Th2 eczema lock. It does not fit the Th17 psoriasis engine.

That specificity is the tell. A blunt poison damages everything it touches. A molecule that selectively amplifies one immune pathway and ignores another is doing something precise, which means it has a real biological role, not just a capacity for harm. And the precision shows up again in the heart literature in the most surprising way: in a study of large-artery stroke and transient ischemic attack, patients had TMAO levels that were lower, not higher, than asymptomatic controls (Yin et al., 2015). The simple "more TMAO is always worse" story does not survive contact with the full data.

"TMAO is not a villain. It is a signal, only as good or bad as the ecosystem sending it." From the episode

This is the thesis the episode is built on. We have the enzymes to make TMAO because it does work in the body. The harm is not the molecule existing. It is the dose and the destination: a depleted, lopsided gut overproducing the signal, delivered into a body primed to misread it. The poison is the context, not the compound.

Why Would Biology Make TMAO at All?

If TMAO can do this much damage, it is fair to ask why we make it in the first place. The answer reframes the whole story, because almost nothing in biology persists without a job, and TMAO's job is far older than our arteries. It is what biologists call an osmolyte, a small molecule that stabilizes the shape of proteins when the cellular environment turns hostile. TMAO braces proteins against forces that would otherwise unfold them, and it does this across an astonishing range of life (Yancey, 2005).

The most vivid example lives in the deep sea. Fish and sharks accumulate more TMAO the deeper they live, in an almost linear relationship with depth, because it counteracts the crushing hydrostatic pressure that would otherwise deform their proteins. There is even a proposed TMAO ceiling that may set the maximum depth at which fish can survive (Yancey and Siebenaller, 2015). Closer to home, your own kidney runs the same trick, loading methylamine osmolytes into its saltiest, most stressed layer to protect the cells that live there. TMAO is a stress-protective molecule with a deep evolutionary resume.

📊 The osmolyte, not the antioxidant A quick precision note, because it is easy to blur. TMAO's protective role is chemical chaperoning, holding protein structure together under stress. It is not an antioxidant, and in human tissue the cardiovascular literature actually associates it with more oxidative stress, not less. So the ancient protective job is real, and specific: it stabilizes proteins, it does not mop up free radicals.

That reframes the harm entirely. TMAO is not a poison that snuck into human biology. It is a stress-response molecule, and the disease associations we have been tracing look less like a rogue toxin and more like an ancient protective signal turning up in the wrong amount, in the wrong place, in a body under chronic strain. Which is exactly the pattern this whole story keeps finding: the compound is not the problem, the context is.

Why Does the Th2 System Even Exist?

If the Th2 branch causes so much misery in allergy and eczema, it is worth asking what it was ever for. The answer is parasites. The Th2 program, with its IgE antibodies and its mast cells, evolved largely to expel large invaders like intestinal worms. For most of human history, that system had a constant sparring partner. In much of the modern world, it does not.

📊 The worm teaser There is a real, decades-long line of research in which people are deliberately given controlled doses of hookworm to calm allergic and autoimmune disease. Helminths survive by quieting the host immune system, inducing regulatory T cells and anti-inflammatory signals, and that quieting spills over onto unrelated allergy and autoimmunity (Finlay et al., 2014; Weinstock, 2015). This is not a recommendation. It is a clue: the Th2 system was built to be calibrated by exposures we have largely lost, which may be part of why it now misfires at dust, pollen, food, and our own skin.

Put the worm story next to the TMAO story and a pattern appears. An immune branch built for a world full of parasites, now running without them, meets a gut producing the wrong chemical signals because its keystone species are missing. Neither the immune program nor the molecule is broken in isolation. The context they evolved for has changed underneath them.

Does TMAO Drive Histamine and Mast Cell Activation?

This is a question worth putting on the table honestly, because the mechanism invites it. The Yu pathway ends in IgE, and IgE classically binds mast cells, which degranulate and release histamine. So there is a textbook chain from TMAO-driven Th2 activity toward histamine. There is also a separate, well-documented route by which TMAO amplifies inflammation through the NLRP3 inflammasome and IL-1 beta (Wu et al., 2020). Two distinct inflammatory bridges, both real.

⚠️ Frontier, not finding Here is where honesty has to lead. A direct, published line from TMAO to mast-cell activation syndrome (MCAS) does not currently exist in the literature. The IgE-to-histamine biology is textbook, and the TMAO-to-IgE link is the Yu paper, but stitching them into a TMAO-causes-MCAS claim is an extrapolation, not an established result. We are flagging the thread because it is mechanistically reasonable and worth watching, not because the data is in. Treat it as an open question.

Can You Actually Lower TMAO?

If the problem is the conversion, the gut turning choline and carnitine into too much TMAO, then the leverage sits on the conversion, not on the food. And the conversion is movable. In one elegant study, allicin, the compound in raw garlic, lowered serum TMAO in carnitine-fed mice and, importantly, in human high-TMAO producers, by shifting which gut bacteria were doing the work (Panyod et al., 2022). On the pharmaceutical side, a structural analog of choline called DMB (3,3-dimethyl-1-butanol) non-lethally inhibits the microbial TMA-lyase enzyme, lowering TMAO and reducing atherosclerosis in mice without killing the bacteria outright (Wang et al., 2015). A small family of related enzyme inhibitors now exists (Oktaviono et al., 2023).

There is a mirror image worth naming. The same gut that can overproduce TMAO also produces the calming signals: short-chain fatty acids like butyrate from fiber fermentation, and indole metabolites from tryptophan that act on the skin's aryl hydrocarbon receptor to support barrier function. A diverse microbiome is constantly sending both kinds of messages. Dysbiosis tips the balance toward the inflammatory ones and away from the soothing ones. The intervention question, then, is less "how do I block one molecule" and more "how do I support an ecosystem that keeps the whole signal set in proportion."

⚠️ None of this is a prescription Eating raw garlic is not a treatment for heart disease or eczema, and DMB is an experimental compound, not an approved drug. The road from a mechanism in mice and a one-week human biomarker study to a validated therapy is long. This section is about the direction the science points, which is toward supporting the gut ecosystem, not about a protocol to follow.

Where Postbiotics Fit the Picture

Step back to the shape of the problem. The harm in these stories is rarely a single rogue molecule. It is a gut that has lost its keystone members and the balanced chemistry they maintain: the Akkermansia that tends the barrier, the fiber-fermenters that make butyrate, the tryptophan-handlers that make soothing indoles. When that community thins out, the signal mix tips, and a metabolite like TMAO gets to dominate a conversation it should only be one voice in.

The conventional repair options each have real limits. Fecal transplant is powerful but blunt and tightly regulated. Diet shifts the inputs but works slowly and unevenly. Probiotics reseed a few strains, and the eczema trials for them are honestly all over the map, partly because many of the most important gut species are strict anaerobes that do not survive a capsule on a shelf, and partly because a handful of newcomers rarely overwrite an entrenched ecosystem.

This is the logic behind the postbiotic approach, and we hold it as a thesis, not a settled fact. A healthy microbial community speaks eloquently in chemistry, producing a vast spectrum of signaling molecules (10,000+ molecular signals by current estimates) that no single strain reproduces. If the microbes that make those balancing signals are depleted, one logical intervention is to deliver the full spectrum of postbiotic signals directly, rather than trying to re-grow the entire community first. A full-spectrum postbiotic derived from healthy human donor microbiomes is built on exactly that idea: capture the emergent chemistry of a functional community and deliver it. We believe this is a plausible mechanism. It is a thesis, not a proven clinical outcome. We have not run the trial that would validate it for any condition discussed here, and we would like to. ThaenaBiotic is one tool in your toolbox, not a cure, and certainly not a reason to ignore a dermatologist or cardiologist.

To be explicit on the science boundary: nothing in the Yu paper or the TMAO literature tests a postbiotic against eczema or heart disease. The connection we are drawing is conceptual, about supporting the ecosystem that sets the signal balance, not a claim that any product lowers TMAO or treats a disease.

The Honest Limitations

Several, and they matter. The skin mechanism is proven in mice; the human data is associative. The cardiovascular causal story is strong but complicated by kidney-function confounding and a mixed genetic-causation picture. The TMAO-to-histamine-to-MCAS thread is an extrapolation with no direct evidence yet. And the direction of causation in human gut-and-skin studies is genuinely tangled: severe eczema brings chronic stress, repeated antibiotic courses for skin infections, and anxious elimination diets, each of which can disturb the gut on its own. That is protopathic bias, the real possibility that some of the dysbiosis is a consequence of the disease rather than its cause. The Yu mouse experiments are valuable precisely because they isolate causality in a way human observation cannot.

Frequently Asked Questions

Should I stop eating eggs or red meat to lower my TMAO?

No. Choline is an essential nutrient and the population associations are modest. The research consistently points at the gut ecosystem that processes the food, not the food itself, as the variable that determines how much TMAO is made (Yu et al., 2026; Panyod et al., 2022). Restricting essential nutrients based on an association is not supported.

Is TMAO always harmful?

No. It is associated with cardiovascular risk and, in a new paper, with eczema severity, but it is a normal product of human metabolism, it acts with surprising specificity, and in at least one stroke study it was lower in patients than controls (Yin et al., 2015). Its effect appears to depend heavily on dose and context.

Does garlic really lower TMAO?

In one study, allicin from raw garlic lowered TMAO in mice and in human high-TMAO producers by altering the responsible gut bacteria (Panyod et al., 2022). That is a promising biomarker result, not a proven clinical therapy, and it is one study. It points to a direction, supporting the gut ecosystem, rather than a dose to take.

Can probiotics cure eczema?

No, and the clinical trials are genuinely mixed. Some early-life prevention data is encouraging, but trials treating established eczema in children and adults often show no benefit over placebo. Over-the-counter probiotics are not an approved treatment for atopic dermatitis, and severe skin disease needs a board-certified dermatologist.

What is the gut-skin axis?

It is the two-way chemical relationship between the gut microbiome and the skin. Gut bacteria produce metabolites (TMAO, short-chain fatty acids, tryptophan-derived indoles) that enter the bloodstream and reach the skin's immune cells, helping set whether skin inflames or stays calm (Yu et al., 2026). The Yu paper is one of the most precise molecular maps of that axis to date.

The Bottom Line

Follow one molecule across two organs and the lesson stops being about the molecule. TMAO is a heart villain in some contexts, an eczema driver in another, surgically specific to the Th2 pathway and inert against psoriasis, and even lower than normal in certain stroke patients. None of that fits a simple good-or-bad label. It fits a signal whose meaning is set by the ecosystem producing it and the body receiving it.

💡 The takeaway We are not single organisms. We are ecosystems in constant chemical conversation with ourselves, and metabolites like TMAO are the words. The most durable response to a molecule behaving badly is not to wage war on the molecule or fear the food that feeds it, but to tend the microbial community that decides what the molecule becomes. Support the ecosystem, and you change what the conversation says.

References

Yu, L., Peng, S., Chen, X., et al. (2026). Intestinal dysbiosis exacerbates skin inflammation via microbial metabolite-driven Th2 cell differentiation. Immunity, 59(6), 1545–1560. https://doi.org/10.1016/j.immuni.2026.03.019
Witkowski, M., Weeks, T. L., & Hazen, S. L. (2020). Gut microbiota and cardiovascular disease. Circulation Research, 127(4), 553–570. https://doi.org/10.1161/CIRCRESAHA.120.316242 · FREE FULL TEXT
Tang, W. H. W., Wang, Z., Levison, B. S., et al. (2013). Intestinal microbial metabolism of phosphatidylcholine and cardiovascular risk. New England Journal of Medicine, 368(17), 1575–1584. https://doi.org/10.1056/NEJMoa1109400
Panyod, S., Wu, W.-K., Chen, P.-C., et al. (2022). Atherosclerosis amelioration by allicin in raw garlic through gut microbiota and trimethylamine-N-oxide modulation. NPJ Biofilms and Microbiomes, 8, 4. https://doi.org/10.1038/s41522-022-00266-3 · FREE FULL TEXT
Wang, Z., Roberts, A. B., Buffa, J. A., et al. (2015). Non-lethal inhibition of gut microbial trimethylamine production for the treatment of atherosclerosis. Cell, 163(7), 1585–1595. https://doi.org/10.1016/j.cell.2015.11.055 · FREE FULL TEXT
Canyelles, M., Borràs, C., Rotllan, N., et al. (2023). Gut microbiota-derived TMAO: A causal factor promoting atherosclerotic cardiovascular disease? International Journal of Molecular Sciences, 24(3), 1940. https://doi.org/10.3390/ijms24031940 · FREE FULL TEXT
Oktaviono, Y. H., Dyah Lamara, A., Saputra, P. B. T., et al. (2023). The roles of trimethylamine-N-oxide in atherosclerosis and its potential therapeutic aspect: A literature review. Biomolecules & Biomedicine, 23(6), 936–948. https://doi.org/10.17305/bb.2023.8893 · FREE FULL TEXT
Wu, K., Yuan, Y., Yu, H., et al. (2020). The gut microbial metabolite trimethylamine N-oxide aggravates GVHD by inducing M1 macrophage polarization in mice. Blood, 136(4), 501–515. https://doi.org/10.1182/blood.2019003990 · FREE FULL TEXT
Yin, J., Liao, S.-X., He, Y., et al. (2015). Dysbiosis of gut microbiota with reduced trimethylamine-N-oxide level in patients with large-artery atherosclerotic stroke or transient ischemic attack. Journal of the American Heart Association, 4(11), e002699. https://doi.org/10.1161/JAHA.115.002699 · FREE FULL TEXT
Yancey, P. H. (2005). Organic osmolytes as compatible, metabolic and counteracting cytoprotectants in high osmolarity and other stresses. Journal of Experimental Biology, 208(15), 2819–2830. https://doi.org/10.1242/jeb.01730 · FREE FULL TEXT
Yancey, P. H., & Siebenaller, J. F. (2015). Co-evolution of proteins and solutions: protein adaptation versus cytoprotective micromolecules and their roles in marine organisms. Journal of Experimental Biology, 218(12), 1880–1896. https://doi.org/10.1242/jeb.114355 · FREE FULL TEXT
Finlay, C. M., Walsh, K. P., & Mills, K. H. G. (2014). Induction of regulatory cells by helminth parasites: exploitation for the treatment of inflammatory diseases. Immunological Reviews, 259(1), 206–230. https://doi.org/10.1111/imr.12164
Weinstock, J. V. (2015). Do we need worms to promote immune health? Clinical Reviews in Allergy & Immunology, 49(2), 227–231. https://doi.org/10.1007/s12016-014-8458-3

This post accompanies the Lit Review Friday episode of Learn Something with Thaena.

Back to blog