Strong evidence for migraine prevention with pharmaceutical-grade extracts showing 1.9 fewer migraines per month. Traditional European use is extensive, but quality control is a crisis - commercial products vary 150-fold in active compounds.
Traditions: European herbalism, German phytotherapy
Multiple traditions agree on use.
Evaluated feverfew for migraine prevention (1984-1994). Part of 380 monographs for licensed medical prescribing in Germany.
Committee on Herbal Medicinal Products concluded feverfew herb preparations can be used for the prevention of migraine.
European Scientific Cooperative on Phytotherapy monograph recognizes feverfew for prevention of migraine headaches based on traditional use backed by modern clinical evidence.
Drug Identification Number granted for feverfew product. Recommended dosage: 125 mg dried feverfew leaf containing at least 0.2% parthenolide daily for migraine prevention.
Migraine prevention evidence is strong but complicated by quality issues. A 2025 meta-analysis of 899 participants found 1.11 fewer migraines per month and 4.43 hours shorter duration [1]. Phase III trial with 170 participants using MIG-99 extract found 1.9 fewer migraines per month, with odds ratio 3.4 favoring feverfew (p=0.0049) [2]. Critical issue: parthenolide not detectable in plasma at doses up to 4 mg daily despite clinical efficacy [3]. Massive quality control problem - commercial products vary 150-fold (0.02-3.0 mg parthenolide per dose), and NO commercial extracts tested matched label claims [4]. The research uses pharmaceutical-grade CO2 extract standardized for parthenolide, not typical commercial feverfew products.
899 participants across 9 double-masked placebo-controlled RCTs: 1.11 fewer migraines per month (95% CI -1.23 to -0.99, p<0.00001), 4.43 hours shorter duration (95% CI -7.63 to -1.23, p=0.007).
170 participants (89 feverfew, 81 placebo) for 16 weeks with MIG-99 CO2-extract 6.25 mg t.i.d.: 1.9 fewer migraines per month from baseline of 4.76 (p=0.0456), responder odds ratio 3.4 (p=0.0049). Adverse events 8.4% vs placebo 10.2% (p=0.654).
147 participants tested three doses. Subgroup with ≥4 baseline migraines: 6.25 mg t.i.d. optimal - 1.8 fewer migraines vs 0.3 placebo (p=0.02). No dose-related adverse effects up to 56.25 mg/day.
Parthenolide at doses up to 4 mg daily had no detectable plasma concentration (detection limit 0.5 ng/ml). Demonstrates severe bioavailability problem despite clinical efficacy for migraine.
Commercial products analysis: 150-fold variation (0.02-3.0 mg parthenolide per dose), 160-fold daily intake variation (0.06-9.7 mg/day). NO commercial extracts contained label-claimed parthenolide content. Different batches from same manufacturer varied significantly.
6 trials, 561 participants: 0.6 fewer migraines per month in most rigorous trial (n=218). Low quality evidence requiring confirmation. No serious adverse effects, feverfew generally well tolerated.
Gold standard with Phase III trial evidence. Best response in patients with ≥4 migraines per month. Minimum 12-16 weeks for full prophylactic effect. Pharmaceutical-grade with consistent batch parthenolide content. Adverse events similar to placebo (8.4% vs 10.2%).
Used in early clinical trials. CRITICAL LIMITATION: 150-fold commercial product variation makes therapeutic dose uncertain. To achieve ~18.75 mg parthenolide at 0.2% standardization requires ~9,375 mg dried leaf (not practical). Only use if third-party verified for parthenolide content.
Clinical practice recommendations. Must verify actual parthenolide content - label claims unreliable per Nelson 2002 study. Use only with Certificate of Analysis from independent laboratory.
Feverfew’s research profile shows strong evidence for migraine prevention, but a severe quality control crisis makes choosing the right product critical. This isn’t “may help headaches” - it’s specific, quantified migraine reduction in rigorous trials.
Migraine Prevention - Strong Evidence, Quality-Dependent:
The bioavailability paradox: Phase I cancer trial (Curry 2004) found parthenolide not detectable in plasma at doses up to 4 mg daily (detection limit 0.5 ng/ml) [4]. Despite this severe bioavailability problem, feverfew demonstrates clinical efficacy for migraine - may suggest local effects, gut-mediated mechanisms, or active metabolites not measured.
The quality control crisis (Nelson 2002):
This means most commercial feverfew products are therapeutically unpredictable or inadequate.
Mechanism: Parthenolide (germacranolide-type sesquiterpene lactone) is the primary bioactive compound. It inhibits the NF-κB pathway, reducing pro-inflammatory cytokines (especially TNF-α). This provides the anti-inflammatory and analgesic properties relevant to migraine prevention. Standardization to ≥0.2% parthenolide is the regulatory minimum, but clinical trials used pharmaceutical-grade products with verified consistent content.
Feverfew has extensive documented European traditional use - recognized across multiple regulatory bodies.
European Medicines Agency: Committee on Herbal Medicinal Products concluded feverfew herb preparations can be used for the prevention of migraine. Official European regulatory recognition [13]
German Commission E (1984-1994): Evaluated feverfew for migraine and headache as part of 380 monographs for licensed medical prescribing in Germany. Monographs still considered valid though not updated since 1994 [14]
ESCOP Monograph (2014): European Scientific Cooperative on Phytotherapy monograph titled “Tanaceti parthenii herba (Feverfew)” with therapeutic indication of prevention of migraine headaches. Recognition of traditional use backed by modern clinical evidence [12]
Canadian Health Protection Branch: Drug Identification Number granted for British feverfew product for migraine prevention. Recommended dosage: 125 mg of dried feverfew leaf preparation containing at least 0.2% parthenolide daily [13]
Traditional medicinal uses (European herbalism):
Plant parts used: Aerial parts, specifically leaves and flowering tops. High parthenolide content in leaves, flowering tops, and seeds. Low content in stalks and roots. Only leaves and flowering tops should be used medicinally.
Important gap: The clinical literature does NOT document traditional preparation methods (tincture ratios, decoction methods, tea preparation). Clinical trials focused on pharmaceutical preparations with standardized parthenolide content. For traditional preparation information, consult ethnobotanical and historical herbal texts.
The research makes one thing clear: you need pharmaceutical-grade standardized product, or you’re likely wasting your time. The 150-fold commercial product variation means most feverfew supplements are unpredictable.
For migraine prevention (strongest evidence):
MIG-99 CO2-extract: 6.25 mg three times daily (18.75 mg total daily). This is the gold standard from the Phase III trial with 170 participants where the responder odds ratio was 3.4 (p=0.0049) [2]. Take consistently at regular intervals (morning, midday, evening), preferably with meals to minimize GI effects.
Critical specifications: Pharmaceutical-grade CO2-extraction, standardized for parthenolide content, consistent batch-to-batch quality. This is NOT typical commercial feverfew.
Best candidates: Patients with ≥4 migraine attacks per month at baseline. The dose-response study found this subgroup had the clearest benefit [3].
Alternative if MIG-99 unavailable:
Standardized extract with verified parthenolide content ≥0.2%. You need third-party Certificate of Analysis from independent laboratory - do not trust label claims. The Nelson 2002 study found NO commercial extracts matched their label claims [5].
Reality check: Achieving MIG-99-equivalent doses (18.75 mg parthenolide daily) with dried leaf at 0.2% standardization would require ~9,375 mg dried leaf. Not practical.
What NOT to use:
Weeks 1-12 (initiation phase):
Weeks 12-16 (assessment window):
Beyond 16 weeks:
What you’re tracking: Migraine frequency per month (count them), duration of each migraine in hours, severity/intensity, rescue medication use. Compare to your pre-feverfew baseline.
Consider gradual tapering rather than abrupt discontinuation after prolonged use (>3 months). Anecdotal reports of rebound headache, anxiety, sleep disturbance with abrupt stopping, though not systematically studied in trials.
Baseline (4 weeks before starting):
During trial (track monthly):
Migraine frequency:
Migraine characteristics:
Rescue medication use:
Adverse effects to monitor:
RED FLAGS - Stop immediately:
What they might report (based on trials): “I went from 4-5 migraines per month to 2-3,” “My migraines are shorter - 4-6 hours instead of 8-10 hours,” “I need less rescue medication.”
This cannot be overstated: The quality control crisis with feverfew is severe. The Nelson 2002 study found 150-fold variation in commercial products and ZERO extracts matched label claims [5].
The standardization failure:
What to demand:
Why it matters:
The trials showing 1.9 fewer migraines per month [2], odds ratio 3.4 for responders [2], and 1.11 fewer attacks in meta-analysis [1] all used rigorously standardized pharmaceutical-grade products. An unstandardized supplement from an unknown manufacturer isn’t testing what was studied.
The bioavailability problem adds another layer: parthenolide isn’t detectable in plasma even at 4 mg doses [4], yet clinical efficacy exists. This paradox means we don’t fully understand the mechanism, making standardization and quality even more critical.
Stability requirements:
Red flags - Avoid:
Pharmacopoeia recognition:
This extensive pharmacopoeia recognition indicates feverfew’s medicinal legitimacy, but also highlights that quality standards exist because quality variation is a known problem.
Feverfew is one of the better-studied herbal migraine preventatives with solid evidence, but finding a quality product is the real challenge.
The case for trying it: Strong clinical evidence (1.9 fewer migraines per month in Phase III trial [2], 1.11 fewer in meta-analysis with 899 participants [1], responder odds ratio 3.4 [2]). Good safety profile (adverse events similar to placebo [2]). Multiple regulatory recognitions (EMA, German Commission E, Canadian Health Protection Branch, ESCOP). Effect size comparable to some conventional prophylactic medications. Best response in patients with ≥4 migraines per month [3].
The limitations: The quality control crisis is severe - 150-fold commercial product variation, NO commercial extracts matched label claims [5]. You need pharmaceutical-grade standardized product (MIG-99 or equivalent), which is expensive and may be hard to find. Effects take 8-12 weeks minimum to assess. About 60% of patients in some trials didn’t meet responder criteria. Bioavailability paradox (parthenolide not detectable in plasma [4]) means we don’t fully understand mechanism. Long-term safety (>6 months) not systematically studied. Absolute contraindication in pregnancy.
When it might work (if it works): You go from 5 migraines per month to 3. Your migraines last 4-6 hours instead of 8-10 hours. You need less rescue medication. The reduction is specific and measurable, not dramatic elimination.
When to skip it: You’re pregnant, nursing, or under 18. You have severe Asteraceae family allergies. You have infrequent migraines (<4/month) where the effect size won’t be meaningful. You’re unwilling to invest in pharmaceutical-grade products or wait 12-16 weeks for assessment. You need acute migraine treatment (this is prophylactic only). You’re on anticoagulants and can’t manage theoretical bleeding risk.
Safety profile: Cochrane review: “No major safety problems reported with feverfew use” [6]. Ernst & Pittler: “Mild and transient adverse effects only” [8]. Phase III trial: 8.4% adverse events (feverfew) vs 10.2% (placebo), not significant [2]. No dose-related toxicity up to 56.25 mg/day [3]. Most common effects: GI symptoms, mouth ulceration (11%), oral inflammation. No genotoxicity in 11-month study. Theoretical antiplatelet interaction not systematically studied - use caution with warfarin, aspirin, clopidogrel, NSAIDs. Consider discontinuing 2 weeks before surgery.
If trying: Use pharmaceutical-grade MIG-99 CO2-extract (6.25 mg three times daily, 18.75 mg total) or equivalent with Certificate of Analysis verifying parthenolide content. Take consistently with meals for 12-16 weeks before assessing. Track migraine frequency, duration, severity, and rescue medication use monthly. Best candidates: ≥4 migraines per month. Expected response: ~1.8-1.9 fewer migraines monthly, not elimination. If no benefit by 16 weeks, discontinue - you’re a non-responder. If responding, continue with awareness that long-term safety (>6 months) hasn’t been rigorously studied. Consider gradual tapering if discontinuing after prolonged use (>3 months).
The product selection problem is the biggest barrier. Most commercial feverfew won’t work because of quality issues [5]. You need to invest in pharmaceutical-grade standardized products with third-party verification, or you’re likely wasting your time and money on therapeutically inadequate supplements.
Duration: Minimum 12-16 weeks for full migraine prevention effect. Response may take 8-12 weeks to become evident. Dose-response study was 12 weeks, Phase III trial 16 weeks. Start tracking within first 4 weeks to establish baseline comparison.
What to notice:
You're tracking: How many migraines per month now vs before starting? Are individual migraines shorter? The effect size is specific - about 1.8-1.9 fewer migraines monthly, not elimination. 35% of participants in dose-response study reported at least one adverse event (similar across feverfew and placebo groups). Most common: GI symptoms, mouth ulceration (11% in larger series), oral inflammation. What you won't see: immediate effects - this is prophylactic, takes weeks. If no reduction in migraine frequency by 12 weeks, you're likely a non-responder.
Generally considered: safe
Contraindications:
Pregnancy/Nursing: Absolute contraindication in pregnancy due to potential emmenagogue activity (menstruation-promoting effects). Historically used for menstrual complications, suggesting reproductive effects. Not recommended for lactating mothers - unknown whether feverfew constituents are excreted in breast milk.
Good overall safety profile. Cochrane review: 'No major safety problems reported with feverfew use.' Phase III trial: adverse events 8.4% feverfew vs 10.2% placebo (p=0.654). Dose-response study: 35% reported adverse events (similar across all groups including placebo), no dose-related toxicity up to 56.25 mg/day. Most common adverse effects: GI symptoms (most frequent), mouth ulceration (11%), widespread oral inflammation with lip swelling and loss of taste, contact dermatitis (with plant exposure). Genotoxicity study: 30 migraine sufferers using feverfew for 11 months showed no chromosomal abnormalities. Chronic toxicity studies lacking - long-term safety (>6 months) not formally established. Theoretical antiplatelet interaction (not systematically studied) - use caution with warfarin, aspirin, clopidogrel, NSAIDs. Consider discontinuing 2 weeks before surgery. Anecdotal reports of rebound headache, anxiety, sleep disturbance with abrupt discontinuation - consider gradual tapering after prolonged use (>3 months). Comparative safety: 'Mild and transient adverse effects only' per Ernst & Pittler 2000 systematic review. QUALITY CONTROL CRITICAL: 150-fold product variation means many commercial products therapeutically inadequate or unpredictable.