Two cups a day of a fungus-fermented Chinese brick tea for 12 weeks: LDL cholesterol fell and arteries widened better than with plain hot water. The artery result is about two and a half times what tea trials usually find

Most of the world's tea is green, black or oolong. Fuzhuan is a "dark" tea from China that goes through an extra step: the leaves are pressed into bricks and kept warm and moist until a fungus, Eurotium cristatum, grows through them as small yellow spots. Tea makers call the spots "golden flower" and count them as a sign of quality. The tea has been drunk in the border regions of northern and western China since the 16th century, and it began in Jingyang, in Shaanxi province. In mice and rats, Fuzhuan tea and its extracts have lowered blood fats and changed gut bacteria. Until now there had been very little evidence in people.

A trial published on 28 September 2026 in Frontiers in Nutrition, by researchers at Shaanxi Provincial People's Hospital and Xi'an Jiaotong University, is the first proper test. In September 2021 they recruited 120 healthy adults aged 20 to 65 from one rural community in northern China, people who did not normally drink tea and took no regular medicines. Half were randomly given two cups a day of Fuzhuan tea, each brewed from a 6 g capsule in 300 mL of water at 90 °C for five minutes in a machine built for the study. The other half drank the same amount of hot water. Everyone kept their usual diet and activity for 12 weeks. Ten people dropped out, five from each group, and the analysis covers the 110 who finished. Their average age was about 54.

The trial had two main measures. The first was blood fats. LDL cholesterol, the kind most closely tied to heart disease, fell by 0.36 mmol/L (about 14 mg/dL) in the tea group and did not change in the water group. In the authors' statistical model, the difference between the groups was 0.40 mmol/L, or about 15 mg/dL. That is a sizeable drop for a drink. A 2015 review of 10 trials of black tea, with 411 people, found an average LDL fall of 4.6 mg/dL. Triglycerides fell in the tea group, but not by significantly more than in the water group, and neither total cholesterol nor HDL differed between the groups.

The second measure was how well the main artery in the arm widens. Flow-mediated dilation (FMD) is checked with ultrasound after a blood pressure cuff has been tightened on the forearm for five minutes and then let go. The artery normally widens as blood rushes back, and that depends on its inner lining. FMD rose by 5.5 percentage points in the tea group and fell by 1.2 in the water group, a gap of about 6.7 points. A 2011 review of nine placebo-controlled tea studies found an average gain of 2.6 points for two to three cups a day. The tea group in this trial also started lower (11.0% against 14.4%). People who start low on a measure often score higher the next time just by chance. The authors adjusted for the starting values and the difference held, at about 6.3 points.

The result that is harder to explain is the second artery test. After FMD, each person was given a small dose of nitroglycerin under the tongue. Nitroglycerin relaxes the muscle in the artery wall directly, without needing the inner lining. This nitroglycerin-mediated dilation also rose by about 6 points more in the tea group. Tea studies usually report a better FMD but no change in the nitroglycerin response, and an earlier green tea study the authors cite found exactly that. A big change in how the artery muscle itself responds, in healthy people after 12 weeks of a drink, would be unusual. Nobody was blinded, and the comparison was plain water, so tea's caffeine and the simple fact of being given something went only to one group. The people who did the ultrasound did not know who was in which group. The blood fat figures also don't quite add up. In the tea group, LDL fell and HDL rose, yet total cholesterol went up slightly. The difference must be carried in other particles, mainly the triglyceride-rich ones, and triglycerides went down. That can happen when LDL is measured directly, as it was here, but the paper doesn't discuss it.

The gut bacteria part of the study is weaker than its title suggests. The researchers sequenced stool samples at the start, at 6 weeks and at 12 weeks. Overall diversity didn't change. The mix of species shifted slightly, and 77 species looked different between the groups at first. Once the analysis allowed for testing so many at once, none of them remained significant, as the authors say. The bacteria they highlight, which make short-chain fatty acids, are a hypothesis, not a finding. Short-chain fatty acids themselves were not measured.

Some context on who is behind it. The trial ran in late 2021 and was published almost five years later. It was paid for by Shaanxi provincial grants, including an industrial innovation project and programmes to develop traditional Chinese medicine. Two co-authors work at a government key laboratory for Fu tea processing in Xianyang, and the tea came from a Xianyang tea company. The authors declare no conflicts of interest. Our reading: this is a reasonable early trial that points the same way as other tea research, with an LDL drop that is large for a drink. The artery results are bigger than tea usually achieves, and one of them doesn't fit how tea is thought to work. The authors themselves call for a double-blind, placebo-controlled trial across several centres. Until then, if you like dark tea, there is no reason to stop. If your LDL is high, tea is not a substitute for what your doctor recommends.