Feverfew: Traditional Use, Constituents and Evidence

My nan grew feverfew in a terracotta pot wedged between the lemon verbena and a sprawling rosemary bush. She called it by its old country name , "featherfoil" , and regarded it with a particular quiet reverence she didn't bother explaining to a ten-year-old. I only started looking into it properly years later, sitting in the State Library in Adelaide with a battered 1880s herbal on loan from the Rare Books collection. The plant has more history packed into its serrated little leaves than most people realise.

What is feverfew?

Tanacetum parthenium (formerly classified as Chrysanthemum parthenium) is a short-lived perennial in the Asteraceae family , the same sprawling family that includes chamomile, yarrow and echinacea. Native to the Balkans and the Caucasus, it naturalised across Europe centuries ago and has been cultivated since at least the medieval period. The common name derives, most likely, from the Latin febrifugia meaning "fever reducer", though the etymology is still debated by botanical historians.

In Australia it grows readily; perhaps too readily. It self-seeds along roadsides and in disturbed soils across temperate regions, including parts of South Australia. I've found it running wild on a gravel verge off Greenhill Road in the Adelaide Hills, identifiable by those strongly scented, yellow-green leaves and small white daisy flowers.

Traditional and historical use

The plant appears in John Gerard's Herball of 1597, where it is described as good "for them that are giddie in the head". Nicholas Culpeper, writing in 1653, was more expansive, he recommended it for everything from melancholy to difficult labour. Whether those attributions hold up to scrutiny is a different matter entirely, but they're a useful record of how European herbal practitioners conceptualised the plant across several centuries.

Traditional European use concentrated on three areas: feverish states (hence the name), gynaecological applications, and what historical texts variously called "the megrim", what we'd now recognise as the clinical description of migraine. That last association turned out to be the thread that pulled modern pharmacologists toward the plant in the late twentieth century.

It's worth noting that "traditional use" in phytomedicine is a complex evidentiary category; it tells us a community found a plant worth cultivating and using across generations, but it doesn't constitute proof of any mechanism or outcome. For more on how traditional use intersects with modern classification of botanical preparations, our phytomedicine glossary entry is a useful starting point.

Key phytochemical constituents

Feverfew's chemistry is reasonably well characterised. The constituent that drew the most early research attention is parthenolide, a sesquiterpene lactone concentrated in the leaves. Parthenolide has been studied in vitro for its interactions with nuclear factor kappa-B (NF-κB) signalling pathways, a finding that generated considerable interest, though in vitro results frequently don't translate cleanly to in vivo outcomes. The compound is thermolabile and volatile, which matters a lot for how dried or processed feverfew preparations compare to fresh leaf.

Beyond parthenolide, the plant contains:

• Flavonoids, including luteolin and apigenin, which have been examined in various phytochemical contexts.
• Volatile terpene compounds; the characteristic sharp scent of the fresh leaf owes partly to camphor and related monoterpenes. Terpene chemistry is something we cover separately in our terpene glossary entry, where you can see how these compounds behave across different plant families.
• Chrysanthenyl acetate and other esters, present in the essential oil fraction.
• Melatonin, identified in feverfew at measurable concentrations in at least one analytical study, though the significance of this finding for any human application remains entirely unclear.

Parthenolide content varies substantially between plant populations, growing conditions, and; especially, how the material is dried and stored. This variability is a genuine problem for research reproducibility and for the consistency of commercial preparations. A standardised extract, in the regulatory sense, requires defined minimum constituent levels; many older clinical trials used dried leaf of uncertain parthenolide content, which makes comparing their results genuinely difficult.

What the research actually shows, and its limits

The honest answer is: the evidence base is suggestive but thin, and the methodological problems are substantial.

The earliest controlled trial that attracted wide attention was published in The Lancet in 1988 by Johnson and colleagues at the City of London Migraine Clinic. It was a small (72-participant) randomised, double-blind, placebo-controlled crossover design. The results were cited widely in subsequent decades. But 72 participants across a crossover design, with the confounders around leaf standardisation I described above, is nowhere near sufficient to draw firm conclusions; and the authors were careful to say so at the time, even if the popular press was not.

A Cochrane systematic review (Pittler and Ernst, updated 2004) examined the available randomised controlled trials and found the evidence "encouraging but not conclusive". More recent reviews, including a 2011 analysis in Cephalalgia, noted that trials using standardised extracts showed more consistent results than those using raw dried leaf, but that trial sizes remained small and heterogeneity between study designs was high.

I'll be direct: the popular health media has, for decades, presented the evidence on feverfew as more settled than it is. That's not a knock on the plant, it's a knock on how phytomedicine evidence gets reported. The research is genuinely interesting. It is not definitive.

One area of newer biochemical interest involves parthenolide's interactions with inflammatory signalling cascades. Some researchers have noted structural parallels with how certain other bioactive plant compounds interact with receptor systems, there's even been preliminary in vitro work examining intersections with the endocannabinoid system, though this remains early-stage and should not be over-interpreted. The point is that feverfew's pharmacology is more layered than its "herbal headache remedy" reputation suggests.

For those interested in how plant compounds move through the body more broadly, the concepts of bioavailability and first-pass metabolism are directly relevant to understanding why the same dried herb can produce different physiological exposures across different people and preparation methods.

Safety considerations and known interactions

Feverfew is not without its cautions. Mouth ulcers and oral irritation are documented in users who chew fresh leaves; a traditional preparation method, and these effects are attributed to parthenolide's direct mucosal reactivity. Gastrointestinal effects have also been reported in clinical trials.

Cross-reactivity with other Asteraceae species is a meaningful concern for people with established plant allergies. The literature also flags potential interactions with anticoagulant medicines; this is pharmacologically plausible given some of the in vitro platelet aggregation data, though clinical evidence of a measurable interaction in humans is limited.

Abrupt discontinuation after prolonged use has been associated in some clinical reports with a cluster of symptoms, muscle stiffness, anxiety, disrupted sleep; sometimes called "post-feverfew syndrome". The mechanism is speculative. But it's been reported consistently enough that it appears across multiple independent clinical accounts and is worth knowing about.

Feverfew is contraindicated during pregnancy under standard herbal medicine guidelines, based partly on its traditional use as an emmenagogue and the theoretical risk of uterine stimulation. No adequate controlled human data exists on this, and the precautionary position reflects the general standard that safety in pregnancy cannot be assumed for any botanical preparation lacking specific safety data.

Regulation in Australia: how feverfew sits under the TGA

In Australia, feverfew preparations are regulated as complementary medicines under the Therapeutic Goods Administration (TGA) framework. A product can be listed on the Australian Register of Therapeutic Goods (ARTG), the ARTG, via the Listed medicine pathway if it contains feverfew at a level consistent with traditional use and carries only low-level indications that the TGA permits for listed products. "Listed" status means the TGA has assessed the manufacturer's evidence for quality and safety, but has not independently evaluated efficacy claims.

Permitted indications for listed products are drawn from the TGA's permitted indications database; manufacturers cannot make claims beyond these without seeking a higher level of regulatory evaluation (a "Registered" medicine). This matters because it means that what a product label can legally state in Australia is often considerably more circumspect than what you might read in overseas promotional material or wellness blogs.

Good manufacturing practice standards; GMP, apply to the manufacture of all listed medicines, which in principle should address some of the parthenolide variability concerns, though enforcement has historically been uneven across the industry.

Feverfew is not a controlled or restricted substance under the Poisons Standard. It is not subject to the kind of import controls that apply to, say, kava preparations (which have been subject to specific Australian import permit requirements under the Therapeutic Goods Act). But "unscheduled" does not mean "unregulated", all therapeutic goods claims remain subject to TGA oversight and the Australian Consumer Law.

If a practitioner wanted to use a feverfew-containing product that fell outside ARTG listing; for example, a novel standardised extract at a higher dose, the appropriate regulatory pathway would depend on how the product is classified. Our overview of the Special Access Scheme covers how unapproved therapeutic goods can be accessed in specific clinical contexts, though that pathway is primarily relevant to pharmaceuticals rather than listed complementary medicines.

Sources

• Listed medicines and registered complementary medicines, Therapeutic Goods Administration (TGA)
• Feverfew for preventing migraine (Cochrane Review); Pittler MH, Ernst E, NCBI/PubMed
• Randomised double-blind placebo-controlled trial of feverfew in migraine prevention, Johnson ES et al., The Lancet, via NCBI/PubMed
• Permissible indications for listed medicines, Therapeutic Goods Administration (TGA)

, Naomi Ellison, Botanical & phytomedicine writer