Gut microbiome and plant-based diets

The human colon houses roughly 38 trillion microbes whose collective genome dwarfs our own. What we feed them — more than what we feed ourselves — determines which species flourish, which metabolites circulate in our blood, and, increasingly, which chronic diseases we are prone to. Plant-based diets change this ecosystem quickly, consistently, and in directions that most of the cardiometabolic literature considers favourable.

Fibre is the substrate

Human enzymes cannot break down most dietary fibre. Gut bacteria can. When Bacteroidetes, Faecalibacterium prausnitzii, Roseburia, Eubacterium rectale and their neighbours ferment resistant starch, inulin, pectin and beta-glucans, they produce short-chain fatty acids (SCFAs) — principally acetate, propionate and butyrate.

Butyrate is the preferred fuel of colonocytes and has been linked to lower intestinal inflammation, tighter epithelial junctions, and regulatory T-cell induction. Propionate reaches the liver via the portal vein and suppresses hepatic lipogenesis. Acetate enters peripheral circulation. Diets rich in legumes, whole grains, vegetables and fruit raise stool and plasma SCFAs in a dose-dependent fashion (De Filippis, 2016).

Diversity and the Mediterranean signal

In a landmark Italian cohort, De Filippis and colleagues showed that adherence to a Mediterranean pattern — largely plant-based, legume-heavy, olive oil forward — correlated with higher faecal SCFAs and a microbiota enriched in fibre-fermenting taxa. Omnivores who ate like vegetarians looked, in their microbiome, like vegetarians. The label on the diet mattered less than what was actually on the plate (De Filippis, 2016).

A larger multi-cohort analysis later linked Mediterranean adherence to lower cardiometabolic risk via specific microbial signatures, with the protective association attenuating in participants whose guts lacked the relevant fibre-degrading consortia (Wang, 2021). The microbiome was not a bystander — it was partly the mechanism.

TMAO: the animal-product conversion

The other side of the ledger is trimethylamine N-oxide (TMAO), a metabolite generated when gut bacteria convert L-carnitine (abundant in red meat) and phosphatidylcholine / choline (abundant in eggs, liver, and meat) into trimethylamine, which the liver then oxidises.

• Wang and colleagues (2011) first tied elevated plasma TMAO to incident major adverse cardiovascular events in a cohort of over 4,000 adults undergoing elective coronary angiography.
• Koeth and colleagues (2013) then demonstrated that omnivores, but not long-term vegans or vegetarians, produced a TMAO spike after a carnitine challenge — their microbiota had simply lost the machinery. Antibiotic suppression abolished the response in omnivores, confirming microbial mediation.

The practical implication: the same steak produces more TMAO in a habitual meat-eater than in someone whose gut has not been trained to make it. Diet shapes the factory that shapes the metabolite.

The Sonnenburgs and the vanishing microbiota

Justin and Erica Sonnenburg have argued that the low-fibre, high-fat Western diet is driving a slow-motion ecological extinction inside the human colon. In mouse work, successive generations on a fibre-poor diet progressively lost microbial taxa; re-introducing fibre did not restore them after a few generations — the lineages were gone (Sonnenburg, 2016).

Industrialised human guts harbour markedly lower diversity than those of hunter-gatherer and subsistence-farming populations, and that diversity tracks fibre intake more than any other single variable (Sonnenburg & Sonnenburg, 2019). A plant-based diet is not a cure for this, but it is arguably the most accessible way to arrest it.

How fast does it happen?

Fast. David and colleagues (2014) put volunteers on either an entirely animal-based or entirely plant-based diet for five days. Community structure shifted within 24 hours of food reaching the colon. The animal-based arm saw bloom of bile-tolerant organisms (Alistipes, Bilophila, Bacteroides) and a drop in fibre-fermenters; the plant-based arm showed the mirror image. Microbial gene expression shifted to match the available substrate before abundance did.

Responders and non-responders

Not everyone responds the same way. A Mediterranean or vegan shift produces large SCFA and diversity gains in some people and modest changes in others. Reasons include:

• Starting community composition — if the key fibre-degraders are already rare, there is less to expand.
• Transit time, pH, and bile acid pool — shape which niches exist.
• Genetics of the host — e.g. FUT2 secretor status.
• Fibre type and dose — soluble vs. insoluble, whole vs. isolated.
• Medications — particularly PPIs and recent antibiotic exposure.

Tomova and colleagues (2019) reviewed the vegan/vegetarian microbiome literature and concluded that, while the direction of effect is consistent (more Prevotella, more SCFA-producers, fewer Bilophila), the magnitude varies widely between individuals and cohorts.

Caveats worth naming

• Correlation is abundant; causation is thinner. Much of the human evidence is cross-sectional.
• “Plant-based” covers a wide range. An ultra-processed vegan diet (white bread, vegan cheese, sugary drinks) does not feed a diverse gut.
• TMAO is a biomarker with caveats — fish also raises it acutely, and its causal role, while supported by mechanistic work, is not settled.
• N-of-one variation is real. Population-level benefits do not guarantee individual ones.

The honest summary

Feeding a rainbow of whole plant foods to a human colon reliably expands fibre-fermenting taxa, raises SCFAs, lowers TMAO production capacity, and nudges the ecosystem toward patterns associated with lower cardiometabolic risk. The effect is fast, the direction is consistent, and the magnitude is personal. For most people, most of the time, eating more plants is one of the highest-leverage moves available for gut health.