Rooibos and Health: What the Science Supports

This is the Ministry's standing record of what is known about rooibos (Aspalathus linearis) and health. Two things are settled and need no qualification: rooibos contains no caffeine, and it is rich in antioxidants. Almost everything else marketed as a benefit sits somewhere between "shown once, in a small human trial" and "shown only in mice" and "not shown at all." This page sorts the claims by the strength of the evidence behind them, so a reader can tell the settled facts from the promising ones from the invented ones. It documents the science; it prescribes nothing. Rooibos is a drink, not a treatment, and nothing here is medical advice.

What this record covers, in order: What is settled (the two facts that need no trial), what "antioxidant" does and does not mean, the best human evidence (a modest help for cholesterol), blood sugar (convincing in animals, unproven in people), where the whole body of human evidence stands, claims with no clinical support, safety, and a short closing rule for how to read a rooibos health claim. The sections are ordered by how strong the evidence is, settled first, unsupported last, so you can stop wherever your question is answered.

What is settled

Two facts about rooibos hold without qualification, and both are properties of the plant rather than findings that need a trial.

Rooibos is naturally caffeine-free. The caffeine was never there to remove, unlike a decaffeinated coffee or a decaffeinated black tea, where it is stripped out after the fact. This is why rooibos can be drunk in the evening, and why it is commonly offered to children and to expectant mothers (a cup is not the same as a medical clearance; anyone with a specific concern should ask their own doctor). The South African Rooibos Council and the published reviews are consistent on this point, and it is the firmest health fact about the drink.

Rooibos is low in tannins, the astringent plant compounds that turn black tea bitter on a long steep and that bind some of the iron in a meal. Rooibos carries far less of them than black tea does, which is why a cup does not turn bitter if you forget it steeping and why it sits more lightly alongside food.

Rooibos is rich in antioxidants. Its signature compound is aspalathin, a flavonoid (a C-glucosyl dihydrochalcone) that has so far been detected in almost no other plant, alongside a smaller relative called nothofagin. Green rooibos carries considerably more aspalathin than the familiar red type, because the oxidation step that reddens the leaf converts most of the aspalathin into other compounds. In healthy volunteers, some studies have measured a rise in the blood's antioxidant capacity for a few hours after a cup, though the effect has not shown up consistently across every lab assay used to test it.

The third fact carries a caution the first two do not, and it has its own section below.

What "antioxidant" does and does not mean

"Antioxidant" describes how a molecule behaves in a laboratory: it neutralises reactive oxygen species, the unstable molecules that can damage cells. Rooibos genuinely does this in a test tube, and aspalathin is a large part of why.

What the word does not do is promise a health outcome. A higher reading on a laboratory antioxidant assay, or a few hours of raised antioxidant capacity in the blood, is not the same as a longer life, a lower risk of disease, or any specific benefit a person can feel. The step from the assay to the body is exactly where marketing tends to get ahead of the science. McGill University's Office for Science and Society puts it plainly: the antioxidant content is a real laboratory finding that, on its own, "doesn't have much meaning for consumers." The compounds in rooibos are real and worth studying. What they do for health over a lifetime is still being worked out, which is the business of the sections that follow.

No food regulator has approved a specific health claim for rooibos. In the European Union, claims for botanical ingredients like rooibos sit on an "on hold" list that has never been formally evaluated or authorised. The absence of an approved claim is not proof that rooibos does nothing; it reflects that the human evidence has not been strong enough to clear the regulatory bar.

The best human evidence: a modest help for cholesterol

The strongest signal in actual people comes from the heart, from a single study, and it is real but small.

In a six-week trial, Marnewick and colleagues (2011) had forty adults, each carrying at least two risk factors for cardiovascular disease, drink six cups of fermented rooibos a day. Over the six weeks the blood-fat picture improved: LDL ("bad") cholesterol and triglycerides fell, HDL ("good") cholesterol rose, and markers of oxidative stress in the blood went down. The 2024 scoping review that pooled the human studies recorded the size of these moves as roughly a 15 percent fall in LDL, a 29 percent fall in triglycerides, and a 33 percent rise in HDL.

The honest reading of those numbers is set by the study's size, not by the percentages. It is one trial: forty people, all already at elevated risk, drinking six cups a day for six weeks, with no large confirmatory trial since. It is the closest rooibos comes to an evidenced benefit beyond simple refreshment, and it is still a long way from a settled fact. Anyone managing their cholesterol should treat rooibos as a pleasant addition to medical advice, not a replacement for it.

Blood sugar: convincing in animals, unproven in people

The diabetes research is the clearest case of a laboratory finding running well ahead of its human proof.

Aspalathin lowers blood glucose in diabetic mice. Kawano and colleagues (2009) reported this directly in a diabetic mouse model, and a 2018 systematic review and meta-analysis found that rooibos extracts and their polyphenols lower blood glucose overall in animal studies, though unevenly by compound: the pooled aspalathin studies alone did not reach significance, while two other rooibos compounds, PPAG and isoorientin, did. This mixed body of cell and animal work is why aspalathin is still taken seriously as a research compound, not yet proof that it works.

What is missing is the trial that matters. No clinical study has yet tested whether ordinary rooibos meaningfully controls blood sugar in people with diabetes. A few small human measurements have hinted at a dip in glucose after a single large serving, but they are preliminary and were not designed to answer the question. Until a proper trial is run, the diabetes story belongs to the mice. It is a reason for researchers to keep looking, not a reason for anyone to drink rooibos in place of treatment.

Where the whole body of human evidence stands

It helps to see the totality rather than one study at a time. The most complete survey is the 2024 scoping review, which gathered every human study it could find on rooibos and health: eighteen publications in all, covering 488 people between them, aged from six to eighty-three.

The review reported the same encouraging hints described above, in antioxidant status, in blood lipids, and in glucose, and then drew the cautious conclusion the numbers demand. In its own framing, the small number of human studies and the small total sample "make it very difficult to either confirm or refute" that rooibos has a significant effect on people, and it called for large, well-designed clinical trials before any firm health claim can be made. That is the durable state of the evidence: real laboratory promise, a handful of small and encouraging human studies, and no large trial yet to confirm them.

Claims with no clinical support

Some popular claims are not weak evidence; they are no evidence. The detox claims, the cure claims, and the weight-loss and anti-ageing claims have no clinical support behind them. Where a popular benefit is not named in the sections above, treat that absence as the answer. Rooibos has a genuine, modest research story, and dressing it up as a cure-all does the real findings no favours.

A note on vocabulary follows from this. "Natural" is a description of origin, not a guarantee of effect or of safety, and a plant compound that helps in a mouse does not automatically help, or even behave, the same way in a person.

Safety

For everyday drinking, rooibos has an excellent safety record, and the lack of caffeine makes it gentler than most teas.

The one caveat in the medical literature is rare and worth stating precisely. A small number of case reports describe liver inflammation in people who had been drinking rooibos. In the most detailed report, Engels and colleagues (2015) describe a man who developed acute liver injury after a month of daily rooibos taken together with buchu, another South African herb; his liver recovered after he stopped. Because the rooibos was taken alongside another herb, and because such reports are so rare against the volume of rooibos consumed worldwide, the likeliest explanations are the other herb in the blend, a contaminant, or the unusual quantity, rather than rooibos itself. The pattern is a useful reminder that "natural" and "harmless in any amount" are not the same thing. A few cups a day is not where any reported trouble lies.

How to read a rooibos health claim

A short rule covers most of what is on a label or in an advertisement. Caffeine-free and antioxidant-rich are settled, and may be stated plainly. A cholesterol benefit rests on one small trial and should be described as such, not as established fact. A blood-sugar benefit is shown in animals and not yet in people. Anything beyond these, a detox, a cure, weight loss, reversed ageing, has no clinical evidence behind it. When a claim outruns that ladder, the claim is ahead of the science, not the other way around.

So rooibos earns part of its reputation on the evidence and not yet the whole of it. It is a caffeine-free, low-tannin, antioxidant-rich drink, with one small trial suggesting a modest benefit for cholesterol and a good deal of laboratory promise that human studies have still to confirm. Drink it for what it reliably is, a calm, caffeine-free cup, and let the wonder-herb claims wait for the evidence that may or may not arrive.