Feverfew for migraine prevention: the half-attack result isn't the whole story

A feverfew trial can sound convincing if you hear only that attacks fell from 4.8 to 2.9 a month. But in that same trial, people taking a placebo went from 4.8 to 3.5. The difference between the groups was 0.6 attacks a month - roughly one fewer attack every couple of months on average, in one study of one extract. And that result didn’t settle the broader evidence: other trials disagreed, so the 2015 Cochrane review couldn’t calculate a dependable combined estimate.

That distinction matters if you’re looking for a way to prevent recurrent migraine. Feverfew has been studied as something taken over time to reduce future attacks, not as a rescue remedy when an attack has already started. If you’re in the middle of an attack, a claim about prevention doesn’t answer your immediate problem. A promising number from one preparation isn’t a reason to swap out prescribed care.

Where the half-attack figure came from

#

Feverfew (Tanacetum parthenium) has a long history of medicinal use and is sold for migraine. Its history is real; a dependable clinical effect is another question. Laboratory research into compounds including parthenolide offers possible mechanisms, but it doesn’t tell us whether a bottle on a shelf will prevent an individual’s attacks (NCCIH; Memorial Sloan Kettering).

The Cochrane reviewers found six double-blind, placebo-controlled trials involving 561 people, with searches through January 2015. Products, study designs and measures of success differed too much for the reviewers to pool the results. Three older positive trials were small; two other rigorous trials found no statistically significant difference. That’s not proof the plant can never help. But it does make a confident promise hard to defend.

In the largest trial, 218 people took either a stable feverfew extract or placebo. Monthly attacks fell by 1.9 in the feverfew group and 1.3 in the placebo group - hence the 0.6-attack between-group difference. Read the 1.9 figure on its own and the treatment looks much stronger than the comparison supports. The same trial didn’t establish a significant difference in attack intensity or duration, nausea or vomiting (Cochrane review). Nor does the result tell you what to expect from a different extract, much less fresh leaves or a combination supplement. What if the box says “standardized”? That label alone can’t turn one trial of one preparation into a result for every preparation.

And a later review hasn’t neatly resolved the dispute. NCCIH describes a 2020 review of seven studies and 634 participants as finding inconsistent results. Those studies likely overlap with Cochrane’s trials; the participant counts aren’t two independent piles of evidence. Cochrane called the combined efficacy evidence low quality, and a 2016 NICE surveillance assessment found it too inconsistent to support recommending feverfew. That assessment isn’t a ban, either.

The cautions come before the capsule

#

If you’re pregnant, don’t take feverfew: NCCIH warns it could affect uterine contractions. Breastfeeding safety isn’t established, and trial results in adults don’t establish safety for children. These aren’t footnotes to the efficacy question. They’re reasons to pause before asking whether a particular extract might help.

Feverfew may slow blood clotting. If you take anticoagulants or antiplatelet medicines, have a bleeding disorder, or have surgery coming up, tell the clinician managing your care about any feverfew use (and bring the product label). NCCIH advises stopping it at least two weeks before scheduled surgery, so raise it with your surgical team early rather than making last-minute changes on your own. Memorial Sloan Kettering also flags possible drug interactions, while noting that the clinical importance of some proposed interactions remains unknown. So this isn’t a claim that every pairing causes serious bleeding. But uncertainty isn’t a good reason to leave feverfew off your medication list. Don’t stop prescribed blood thinners or migraine medicines to make room for an herb. Include over-the-counter pain medicines and other supplements when you discuss the full list.

There are more ordinary downsides, too. Digestive upset and mouth irritation have been reported; chewing fresh leaves is particularly associated with mouth sores. People sensitive to ragweed or related plants should ask about allergy risk, and topical feverfew can irritate skin (NCCIH; Memorial Sloan Kettering). After prolonged use, headache, muscle stiffness and anxiety on stopping have also been reported. Yet there’s no reliable rate for those symptoms or established taper plan in the evidence reviewed here - another reason to speak with a clinician before changing regular use (Memorial Sloan Kettering; University of Rochester Medicine).

The trials’ mostly mild, short-term adverse-event reports can’t answer every longer-term safety question. And because the preparations varied, they don’t give us a dependable consumer dose or a brand to recommend (Cochrane). A tidy supplement label doesn’t fix that gap.

A better conversation about prevention

#

So what if your migraines are frequent enough that you want fewer of them? Bring the pattern - frequency, disability, medicines used and any change in symptoms - to a qualified health professional. A headache diary can help make that conversation specific rather than relying on memory. Ask what established preventive options fit your situation and how their risks compare with what you’re already taking. The AHS/AAN medication-prevention update is about medication choices and shared decisions; it isn’t a head-to-head feverfew comparison.

Don’t treat a sudden, extremely painful headache as another routine migraine. New weakness, trouble speaking or seeing, a seizure or confusion calls for urgent medical help (NHS). And if your familiar headache pattern has changed or attacks are becoming more frequent, make a clinical appointment rather than using an herb to put off an assessment.

Feverfew isn’t an absurd question to bring up - one good-sized trial did find a small frequency advantage. But the trials don’t agree, the extract matters, and the safety cautions matter even if you never buy a capsule. The honest answer is a possible signal, not a reliable prevention plan.