Shu pu'er and the gut microbiome — small studies, big claims

Shu pu’er (熟普洱) is the only Chinese tea whose production was reverse-engineered from spoilage. The wet-piling process — wò duī (渥堆) — was formalised at the Kunming Tea Factory in 1973, deliberately accelerating the kind of microbial transformation that traders had observed accidentally on slow caravans north toward the Russian and Mongolian borders. Half a century later, that same microbial accident has become a research object. The pile is full of Aspergillus, Blastobotrys, Lichtheimia and a long tail of bacteria, and what they leave behind in the leaf includes a class of dark, polymerised pigments called theabrownins. In 2019 a team working through a Shanghai Jiao Tong University-affiliated hospital lab and Hong Kong Baptist University published a paper in Nature Communications arguing that theabrownins from ripe pu’er alter the gut microbiome of mice in ways that lower hepatic cholesterol — a paper tea-shop shorthand often mis-credits to “the Kunming labs”, which is not where the work was done. The paper was widely shared, often summarised as “shu pu’er fixes cholesterol via gut bacteria,” and almost as widely overinterpreted. I have spent more than a decade buying shu for the Russian and Mongolian markets — where the tea has been drunk daily, in volume, for generations — and I want to walk carefully through what this research actually demonstrates, what it does not, and how a drinker should hold the claim. The short version: the chemistry is genuine, the mechanism is plausible, the human data is a small real pilot rather than the mouse-only story usually told, and the marketing has still run several kilometres ahead of all of it.

What theabrownins actually are

Theabrownins are the high-molecular-weight pigments that give aged shu its near-opaque mahogany liquor. Chemically they are a heterogeneous polymer family — oxidation and polymerisation products of catechins, theaflavins and thearubigins, condensed with proteins, polysaccharides and caffeine residues. In fresh green tea they barely exist. In a well-fermented shu they can reach 8–14 percent of dry leaf mass, depending on pile temperature, turn frequency and the moisture curve over the 45–70 day fermentation. The Chinese national standard GB/T 22111-2008 for geographical-indication pu’er does not set a theabrownin minimum, but factory QC labs in Menghai and Xiaguan routinely measure them as a fermentation marker. Sensorily, a leaf with high theabrownin content brews dark and slick — the liquor coats a porcelain gaiwan with a faint ring after pouring out, and the texture in the mouth is what Mandarin tasters call huá (滑), slippery rather than astringent. They are also the reason a 15-year-old shu cake looks black where a five-year-old cake looks rust-brown. Importantly, theabrownins are not absorbed intact through the gut wall. They are too big. Whatever they do, they do in the lumen — on the microbiome, on bile acids, on sterol absorption — before being excreted. That fact is what makes the gut-microbiome framing biologically coherent, and it is also what makes any “detox” or systemic-antioxidant claim biologically incoherent.

The 2019 paper, read carefully

The paper in question is Huang, F., Zheng, X., Ma, X. et al., “Theabrownin from Pu-erh tea attenuates hypercholesterolemia via modulation of gut microbiota and bile acid metabolism,” Nature Communications 10:4971, published October 31 2019, out of the Shanghai Key Laboratory of Diabetes Mellitus and Hong Kong Baptist University’s School of Chinese Medicine. The authors isolated theabrownins from a commercial ripe pu’er extract, fed them to mice on a high-fat diet at graded doses up to roughly 225 mg/kg/day, and observed three things: reduced hepatic cholesterol, altered composition of the cecal microbiome with suppression of certain bile-salt-hydrolase-producing bacteria, and increased excretion of conjugated bile acids in faeces. What most popular summaries leave out entirely is that the same paper also ran a small open pilot in thirteen healthy human volunteers, who drank the tea infusion twice daily for four weeks — narrower measurements than a full lipid panel, but a real, if modest, human data point, not the mouse-only study it is usually described as. The proposed mechanism is elegant — theabrownins inhibit specific gut bacteria, which means more bile acids stay conjugated, which means the liver has to draw on cholesterol to make new ones, which means serum cholesterol drops. They replicated the microbiome shift by faecal transfer and showed the effect travelled with the bacteria. As mouse mechanism papers go, it is well-constructed. As a basis for telling humans to drink shu for their lipid panel, it has four large problems, which I will take in turn.

Problem one — dose translation

225 mg/kg/day in a mouse, scaled by the standard body-surface-area conversion (divide by about 12.3), comes out to roughly 18 mg/kg/day in a 70 kg human — about 1.3 grams of pure theabrownin per day. If a shu cake is 10 percent theabrownin by mass and you extract perhaps 60 percent of that into your cup, you would need to drink the soluble fraction of roughly 22 grams of dry leaf daily. That is not impossible — Mongolian households routinely brew 30–40 grams of brick tea a day boiled with milk and salt — but it is far above the 5–8 grams a Chinese drinker uses in a gōngfū session. The dose gap matters.

Problem two — the mouse is not the human

The murine microbiome shares perhaps 85 percent of its bacterial genera with the human one, but the relative abundances are wildly different, and the bile acid pool is differently composed — mice synthesise muricholic acids that humans do not. Effects on bile-salt hydrolase activity may not translate. The paper does not claim it does. The press releases did.

What human evidence exists

The honest answer is: very little, and almost none of it specific to isolated theabrownins. Beyond the small human pilot in the 2019 paper itself, a handful of Chinese clinical trials have tested whole pu’er extract on lipid or metabolic markers — Chu et al. (2011) in the Chinese Journal of Integrative Medicine, with ninety adults given pu’er extract capsules for three months, reported improvements across total cholesterol, triglycerides, LDL and HDL relative to placebo. None of these trials isolated theabrownins specifically, none measured the gut microbiome directly, and most used capsules of extract rather than brewed tea. A 2020 review in Frontiers in Nutrition counted fewer than ten human RCTs on pu’er of any kind, most with fewer than 100 participants and short follow-up. For shu specifically, dedicated human microbiome data beyond the 2019 paper’s own pilot remains sparse. This is not a body of evidence on which to build a health claim. It is a body of evidence on which to say “interesting, worth a larger trial.” The companion piece on tea.doctor covering the 2019 lipid data, aged sheng and lipids, walks through the same problem from the cholesterol-endpoint side.

The microbiome itself — what shu drinkers should know

Even setting aside theabrownins, shu pu’er arrives in the cup already shaped by microbes. The pile fermentation is dominated by Aspergillus niger and several yeasts; the finished cake carries dormant spores and a great deal of microbial metabolite residue. None of this is alive after a 95 °C brew, and modern factory shu is tested against GB 2762 for mycotoxin limits — aflatoxin B1 below 5 µg/kg, ochratoxin A below 5 µg/kg — so the safety questions raised in the 2010s by a few Guangzhou sampling studies have been substantially addressed by traceable producers. Where the microbiome story becomes interesting is in the prebiotic-like fibre and polysaccharide content of aged shu, which does survive brewing in some quantity. Pu’er polysaccharides are around 3–5 percent of the soluble matter in a strong brew, and in vitro they ferment readily with human stool inocula to produce short-chain fatty acids, particularly butyrate. This is a more modest claim than “theabrownins fix cholesterol,” and a better-supported one. A daily bowl of shu probably acts on the gut more as a gentle source of fermentable substrate than as a pharmacological agent.

A note from the northern markets

In Buryatia and northern Mongolia, süütei tsai — milk tea built from compressed brick, often shu-style or Hunan dark tea — is consumed at three to five bowls a day in winter. I have spent enough time at kitchen tables there to say that the population-level pattern is not one of pristine lipid panels and untroubled guts; it is one of people who tolerate a great deal of tea, salt and animal fat together. If shu were a powerful cholesterol drug, this would be a different demographic. The drink is wonderful. It is not medicine.

How to read a ‘shu pu’er and microbiome’ headline

When a wellness site claims that ripe pu’er “resets your gut microbiome” or “burns belly fat through theabrownins,” three checks will sort the serious from the silly. First — is the source a mouse study, an in vitro study, or a human RCT? Almost always the answer is the first two. Second — is the dose plausible from brewed tea, or is it a concentrated extract at multiples of normal intake? Almost always the latter. Third — does the article cite the original paper, or does it cite another article that cites the paper? You can trace the 2019 Huang et al. work through perhaps fifty popular pieces; the further from the source, the larger the claim becomes. None of this means shu is not worth drinking. It is one of the great Chinese teas — a 2008 Menghai 7572 brewed at 8 grams in a 150 ml gaiwan, four-second steeps after a 10-second rinse, gives a liquor that is sweet, earthy, slightly camphoraceous, and physically warming in a way that no green tea is. That experience is the reason to drink it. The cholesterol story is a footnote, possibly true, possibly not, certainly not settled. For the broader context on tea polyphenol pharmacokinetics, see our companion piece how much EGCG is actually in a real brew, which lays out the same dose-translation problem for green tea.

What I tell buyers and students

When I teach Russian and Mongolian buyers at sessions organised through tea.school, I frame shu pu’er like this. It is a fermented Chinese tea with a distinctive chemistry, including a pigment class — theabrownins — that has plausible activity on the gut lumen and bile acid pool. There is one well-cited mouse study and a handful of small human trials that hint at lipid effects. None of this rises to the standard of a clinical recommendation, and no responsible producer or vendor should pitch the tea as a treatment for anything. If a customer has hypercholesterolaemia, they need a cardiologist, not a bǐng (饼). What shu does offer, reliably, is warmth, digestibility after rich food, low caffeine relative to green tea, and the deep familiarity of a category that Mongolian and Tibetan drinkers have leaned on for centuries. Those are honest reasons to recommend it. The medical disclaimer that runs across the bottom of every tea.doctor page is not boilerplate — it is the literal scientific position. The chemistry is interesting. The claims are bigger than the evidence. Drink the tea because it is good tea.

References

1. Theabrownin from Pu-erh tea attenuates hypercholesterolemia via modulation of gut microbiota and bile acid metabolism — Huang F. et al., Nature Communications 10:4971, 2019
2. GB/T 22111-2008 — Geographical indication product: Pu'er tea — Standardization Administration of the People's Republic of China
3. GB 2762 — National food safety standard: Maximum levels of contaminants in foods — National Health Commission of the PRC
4. A randomized double-blind placebo-controlled study of Pu'er tea extract on the regulation of metabolic syndrome — Chu S.-L. et al., Chinese Journal of Integrative Medicine, 17(7), 492–498 (2011)
5. Polysaccharides from Pu-erh tea: structure, bioactivity and fermentability — a review — Frontiers in Nutrition, 2020
6. Microbial succession during wo dui fermentation of ripe pu-erh tea — Yunnan Agricultural University, Journal of Tea Science, 2017