Herb–drug interactions

What it is

The effect a herbal preparation has on a prescribed medicine taken alongside it — either weakening it to the point of failure, or strengthening it to the point of harm.

This is not a fringe concern. Surveys consistently find that a large minority of people taking prescription medicines also take herbal or dietary supplements, and that most of them have not told their doctor. The commonest reason given is not secrecy but assumption: that something sold in a supermarket is not the kind of thing a doctor needs to know about.

Where it came from

As a recognised clinical problem, from the late 1990s.

St John's Wort is the case that created the field. It became enormously popular for low mood in the 1990s, and then reports began arriving that could not be ignored: transplant patients rejecting organs, people with HIV losing viral control, unplanned pregnancies. In 2000 the Lancet published the transplant findings, and regulators across Europe, the US and Australia issued warnings.

The mechanism turned out to be specific and well understood, and it is worth understanding because almost every interaction in this article works one of these two ways.

How it is used

Not a practice — a hazard. This article exists so a reader can recognise the shape of the problem, not so they can manage it themselves.

What we can and cannot say

We can say there are essentially two mechanisms, and they run in opposite directions.

Induction — the drug stops working. A plant compound makes the liver produce more of the enzymes that break a drug down, so the drug is cleared faster and its blood level falls. St John's Wort induces cytochrome CYP3A4 and the transporter P-glycoprotein, and CYP3A4 alone metabolises something like half of all prescribed drugs. The result is silent failure: nothing feels wrong, and the medicine simply stops doing its job.

Inhibition — the drug becomes too strong. The reverse: the enzyme is blocked, the drug accumulates, and a normal dose becomes an overdose. Grapefruit is the famous example and is not even a herb.

We can name the interactions that matter most, because these are the ones that recur in every clinical reference:

• St John's Wort with hormonal contraception, several antidepressants, warfarin, antiretrovirals, immunosuppressants, some chemotherapy agents and some anticonvulsants. It is also a serotonergic agent in its own right, so combined with an SSRI it carries a serotonin syndrome risk on top of everything else.
• Ginkgo, ginger, garlic, ginseng and feverfew with warfarin, clopidogrel and other anticoagulants — additive bleeding risk. The surgical literature is the reason many anaesthetists now ask about supplements before an operation.
• Liquorice with antihypertensives and diuretics. Glycyrrhizin causes potassium loss and sodium retention, raising blood pressure and, at sustained intake, producing a syndrome that looks like a hormonal disorder.
• Kava and valerian with sedatives, benzodiazepines and alcohol — additive central nervous system depression.
• Green tea extract with warfarin, through vitamin K content, quite apart from the separate hepatotoxicity problem the extract carries on its own.
• Berberine-containing plants with a wide range of drugs, through strong CYP inhibition.

We can say the highest-risk situations are identifiable. Interactions matter most where the therapeutic window is narrow — warfarin, digoxin, lithium, antiepileptics, immunosuppressants — because a small shift in blood level crosses from working to dangerous. They matter most for people on several medicines at once, for pregnancy and breastfeeding, for anyone with liver or kidney impairment, and in the fortnight before surgery.

We can say the evidence base is thinner than the risk deserves. Most known interactions come from case reports and pharmacokinetic studies rather than trials, because nobody is going to randomise transplant patients onto St John's Wort. Absence of a documented interaction for a given plant usually means nobody has looked, and reading it as reassurance is the error this article most wants to prevent.

And we can say what actually reduces the risk, which is the practical point of the whole thing:

Tell your pharmacist and your doctor everything you take. Herbal, over-the-counter, imported, recommended by a relative, all of it. Community pharmacists check interactions constantly, it takes minutes, and the reaction you are afraid of — being judged — is not what happens. Bring the actual packet, because the label is the only reliable statement of what is in it.

Do not stop a prescribed medicine to make room for a herbal one. If a practitioner suggests that, what they have told you is about them.

This article is not medical advice and cannot be. It describes a category of risk so that it can be recognised. Anything specific belongs with the people who can see your actual medicine list.

Further reading

• The Australian TGA and the UK MHRA both publish plain-language interaction warnings.
• Piscitelli et al. on indinavir and St John's Wort (The Lancet, 2000) — the paper that started it.
• Your pharmacist, genuinely, before anything on this list.