---
canonical_name: Feverfew
alternate_names: Tanacetum parthenium, Chrysanthemum parthenium, Pyrethrum parthenium, Matricaria parthenium, Bachelor's Buttons, Featherfew, Midsummer Daisy, Wild Chamomile, Altamisa
canonical_topic: Feverfew for Health & Longevity
short_topic_lc: feverfew
creation_date: 2026-0825-1220
creator_ai_fullname: Opus 5
ep_keywords: Migraine Prophylaxis, Sesquiterpene Lactones
---

# Feverfew for Health & Longevity
<section id="top" markdown="1"></section>
Evidence Review created on 08/25/2026 using [AI4L](https://github.com/forever-healthy/AI4L) / Opus 5  

**Also known as:** Tanacetum parthenium, Chrysanthemum parthenium, Pyrethrum parthenium, Matricaria parthenium, Bachelor's Buttons, Featherfew, Midsummer Daisy, Wild Chamomile, Altamisa  

  
## Motivation

<!-- This motivation section was written last, after every other section of this review was complete, so that it reflects the full scope of the evidence rather than a preliminary impression of it. -->

Feverfew (*Tanacetum parthenium*) is a small, aromatic, daisy-like plant of the aster family that has grown in European gardens for centuries and served as a household remedy for far longer. Its dried leaves, and concentrated extracts made from them, are sold today mainly to make migraine attacks less frequent. Attention centres on a bitter compound in the leaf, parthenolide, which dampens the inflammatory signalling that runs inside cells.  

The plant's English name descends from a Latin word meaning to drive out fever, and older herbal practice used it for fevers, aches, and difficult childbirth. Its modern reputation began in Britain in the 1970s, when people with migraine who chewed the fresh leaves reported fewer attacks and hospital doctors decided to test the claim properly. The same leaf compound has since been pursued as a starting point for anti-inflammatory and anti-cancer drugs.  

This review examines what the human and laboratory record shows: how feverfew is thought to act, which effects hold up and which remain unsettled, what harms have been documented, how preparations differ, and how it is used in practice.  

**[Benefits](#expected-benefits) - [Risks](#potential-risks--side-effects) - [Protocol](#therapeutic-protocol) - [Conclusion](#conclusion)**  

  
## Recommended Reading

This section collects independent overviews and expert commentary that place feverfew within the wider field of migraine prevention and plant-derived anti-inflammatory pharmacology.  

<!-- Author's search statement: On 16 August 2026 I ran real-time searches for high-level, substantial treatments of feverfew and of herbal migraine prophylaxis. Two independent searches were performed for each priority expert platform: (1) a web search of "<expert name> feverfew"; and (2) an on-site search using the platform's own search function. On-site results: foundmyfitness.com returned a single hit, a listener Q&A episode in which feverfew is not a subject of discussion; hubermanlab.com returned "No results found."; lifespan.io returned "No Articles Found."; peterattiamd.com returned no feverfew content by web or site search; chriskresser.com returned a single passing mention of feverfew inside a podcast transcript on cerebral blood flow, but no article, episode, or lecture that examines the herb; lifeextension.com returned a substantial migraine protocol that discusses feverfew by name. Beyond the priority platforms I searched PubMed for narrative reviews and expert commentary, and the wider web for clinician-facing digests. Systematic reviews and meta-analyses were deliberately excluded here and routed to the Systematic Reviews section; encyclopedias, wikis, forums, mainstream media, commercial product pages, and database or monograph entries (for example Memorial Sloan Kettering's herb database and drugs.com) were excluded as ineligible types. -->

* [Nutraceuticals and Headache 2024: Riboflavin, Coenzyme Q10, Feverfew, Magnesium, Melatonin, and Butterbur](https://pubmed.ncbi.nlm.nih.gov/39853578/) - Tepper & Tepper, 2025  

  A headache specialist's narrative appraisal of six non-prescription migraine preventives, arguing feverfew may work while inconsistent leaf preparations and thin safety data justify restraint in recommending it.  

* [Exploring the Phytochemistry, Signaling Pathways, and Mechanisms of Action of Tanacetum parthenium (L.) Sch.Bip.: A Comprehensive Literature Review](https://pubmed.ncbi.nlm.nih.gov/39457613/) - Kashkooe et al., 2024  

  The most current single source on feverfew's constituents, the signalling pathways parthenolide acts on, and its documented toxicity, spanning migraine, epilepsy, neuropathic pain, and cancer research.  

* [Migraine Headache](https://www.lifeextension.com/protocols/neurological/migraine) - Sandhaus & Decker  

  A longevity-oriented protocol that places feverfew among nutrient and botanical migraine preventives, contrasting the dried-leaf and carbon-dioxide-extract trial results and naming the inflammatory mediators feverfew suppresses.  

* [Spotlight On: Nutraceuticals](https://americanmigrainefoundation.org/resource-library/nutraceuticals/) - Zubair & Kuruvilla  

  Patient-facing survey of the therapeutic category feverfew belongs to — non-prescription agents aimed at nerve inflammation and cellular energy supply in migraine — with an evidence rating for each.  

* [Headaches and Complementary Health Approaches: What the Science Says](https://www.nccih.nih.gov/health/providers/digest/headaches-and-complementary-health-approaches-science) - National Center for Complementary and Integrative Health  

  A clinician-directed digest of trial evidence across the non-drug headache category, useful for judging feverfew against butterbur, magnesium, and riboflavin on the same evidentiary yardstick.  

Note on the priority platforms: only Life Extension holds a substantial treatment of feverfew. Direct site searches of foundmyfitness.com, hubermanlab.com, lifespan.io, peterattiamd.com, and chriskresser.com returned no article, episode, or lecture that examines feverfew in any depth; the single FoundMyFitness hit is a wide-ranging listener Q&A in which the word appears without the herb being evaluated, and the only Chris Kresser hit names feverfew in one sentence of a podcast transcript on cerebral blood flow. Feverfew is a narrow, single-indication botanical that sits outside the metabolic, sleep, exercise, and longevity-pharmacology subjects those platforms concentrate on.  

  
## Grokipedia

<!-- Author's search statement: On 16 August 2026 I searched grokipedia.com directly with the browser tool for "Feverfew". The search returned 35 results; the dedicated species article is titled "Tanacetum parthenium" and was opened and read to confirm it is the primary page for the intervention. -->

* [Tanacetum parthenium](https://grokipedia.com/page/Tanacetum_parthenium)  

  The dedicated species article covers botany, naturalised range, taxonomy, parthenolide and flavonoid content, traditional use, and the mixed migraine trial record, including the negative rheumatoid arthritis result.  

  
## Examine

<!-- Author's search statement: On 16 August 2026 I searched examine.com directly for "Feverfew". The site search returned one intervention page, /supplements/feverfew/, which was then opened and read to confirm it is the primary, dedicated page for the intervention. -->

* [Feverfew](https://examine.com/supplements/feverfew/)  

  Examine's graded supplement page draws on 446 participants across five trials, gives standard dosing of 100–300 mg standardised to 0.2–0.4% parthenolide, and flags topical sensitisation.  

  
## ConsumerLab

<!-- Author's search statement: On 16 August 2026 I searched consumerlab.com directly for "Feverfew". The site search returned no product review, no clinical update and no answer page dedicated to feverfew; the herb appears only inside broader member-only answers on migraine supplements, supplements to stop before surgery, and herbs to avoid when breastfeeding. -->

ConsumerLab has published no product review, clinical update, or dedicated article for feverfew. A direct search of consumerlab.com returns only broader member-only answer pages — on migraine supplements, on supplements to stop before surgery, and on herbs to avoid when breastfeeding — in which feverfew is mentioned in passing but no feverfew product has been independently assayed for potency or purity. No ConsumerLab testing of feverfew products exists as of 16 August 2026.  

  
## Systematic Reviews

This section lists the systematic reviews and meta-analyses that bear on feverfew, covering both the claimed benefit (migraine prevention) and the principal countervailing consideration (safety and tolerability), drawing on randomised controlled trials (RCTs — studies in which participants are randomly assigned to the active treatment or to a dummy treatment).  

<!-- Author's search statement: On 16 August 2026 I ran a real-time PubMed search for (feverfew OR "Tanacetum parthenium") AND (systematic review OR meta-analysis), across title/abstract and publication-type fields, returning 13 records, and a second search for feverfew migraine randomised controlled trials returning 17 records. Selection was prioritised by whether the paper is genuinely a systematic review or meta-analysis of feverfew, then by recency, study size, citation standing, and relevance; narrative reviews and primary trials were excluded from this section. -->

* [Feverfew for preventing migraine](https://pubmed.ncbi.nlm.nih.gov/25892430/) - Wider et al., 2015  

  The Cochrane review of six placebo-controlled trials in 561 patients; finds a 0.6 attacks-per-month advantage but rates the certainty low.  

* [Systematic review and meta-analysis of Tanacetum parthenium: evaluating its efficacy in migraine relief](https://pubmed.ncbi.nlm.nih.gov/41422534/) - Nelaturi et al., 2025  

  The current pooled analysis, covering nine double-masked trials and 899 participants, with separate estimates for attack frequency, duration, severity, and accompanying symptoms.  

* [The efficacy and safety of feverfew (Tanacetum parthenium L.): an update of a systematic review](https://pubmed.ncbi.nlm.nih.gov/11276299/) - Ernst & Pittler, 2000  

  The dedicated safety-side review: systematically assembles the adverse-effect record alongside efficacy and concludes the harms are mild, transient, and without major safety signal.  

* [Feverfew for migraine prophylaxis: a systematic review](https://pubmed.ncbi.nlm.nih.gov/22435410/) - Saranitzky et al., 2009  

  Explains why the trials cannot be pooled, and raises the long-term cyclooxygenase-2 (an inflammation-driving enzyme) suppression question that later reviews largely dropped.  

* [Herbal treatments for migraine: A systematic review of randomised-controlled studies](https://pubmed.ncbi.nlm.nih.gov/32310327/) - Lopresti et al., 2020  

  Places feverfew beside butterbur, curcumin, and seven other botanicals in nineteen trials, applying one risk-of-bias standard across the whole category.  

  
## Mechanism of Action

Feverfew's activity is attributed chiefly to parthenolide, a germacranolide-type sesquiterpene lactone that accounts for most of the plant's sesquiterpene content (roughly 0.1–0.2% of dried leaf), alongside flavonoids such as luteolin and apigenin, a volatile oil, and coumarins.  

Parthenolide carries a reactive α-methylene-γ-lactone group that covalently binds cysteine 179 of IKKβ (IκB kinase beta, the enzyme that switches on the master inflammatory controller). That blocks NF-κB (nuclear factor kappa B, the protein complex that turns on inflammatory genes) and cuts output of TNF-α (tumour necrosis factor alpha, a principal inflammatory messenger), IL-1 (interleukin-1, another inflammatory messenger), COX-2 (cyclooxygenase-2, the enzyme that manufactures inflammatory prostaglandins), and MCP-1 (monocyte chemoattractant protein-1, which recruits immune cells), as [Kwok et al.](https://pubmed.ncbi.nlm.nih.gov/11514225/) established. An older line of work found whole-leaf extracts [inhibit serotonin release from platelets](https://pubmed.ncbi.nlm.nih.gov/2860288/) and block arachidonic acid liberation from white cells — the original migraine hypothesis.  

Two readings compete. One holds that migraine benefit follows from damping trigeminal inflammatory signalling. The other notes that [oral parthenolide is undetectable in plasma](https://pubmed.ncbi.nlm.nih.gov/15122077/) and concludes systemic NF-κB blockade cannot be operating, so any effect must come from other leaf constituents — an argument also used against feverfew working at all.  

Pharmacologically parthenolide is non-selective, alkylating cysteine thiols on many proteins rather than one receptor. At oral doses to 4 mg daily, plasma levels stayed below 0.5 ng/mL. Animal data indicate wide tissue distribution, rapid clearance, a plasma half-life under about an hour, and elimination dominated by glutathione conjugation plus hepatic oxidation.  

  
## Historical Context & Evolution

Feverfew's original use was as a febrifuge — the English name descends from Latin *febrifugia*, "fever-driver" — and Greek and later European herbals also assigned it to headache, joint pain, insect bites, and menstrual and obstetric complaints, the last reflecting a reputed uterine-stimulant action that still governs its contraindications.  

Its move into headache research was patient-led rather than laboratory-led. In 1970s Britain, migraine sufferers who chewed fresh leaves reported fewer attacks; the reports reached the City of London Migraine Clinic, where E. S. Johnson's group ran the first controlled test. That 1985 trial withdrew feverfew from seventeen habitual users: those switched to placebo showed a significant rise in headache frequency and severity and in nausea and vomiting, while those kept on the herb were unchanged — [evidence of prophylaxis, though obtained in an already-selected population](https://pubmed.ncbi.nlm.nih.gov/3929876/). A larger Nottingham crossover trial in 72 volunteers then reported [fewer and less severe attacks and less vomiting on dried leaf](https://pubmed.ncbi.nlm.nih.gov/2899663/), without change in attack duration.  

German work in the 1990s shifted the field from variable dried leaf to a reproducible supercritical carbon-dioxide extract, MIG-99, developed and funded by its manufacturer, whose employees co-authored both pivotal trials. The [dose-finding study missed significance overall](https://pubmed.ncbi.nlm.nih.gov/12230594/), succeeding only in a pre-specified high-frequency subgroup; the [confirmatory 170-patient trial](https://pubmed.ncbi.nlm.nih.gov/16232154/) then met its endpoint. Opinion has not settled: extraction chemistry, patient selection, trial size, and sponsorship each still separate the positive from the null results.  

  
## Expected Benefits

<!-- Before writing this section I ran dedicated searches for the complete benefit profile of feverfew across PubMed (migraine prophylaxis, acute migraine, rheumatoid arthritis, dermatology, oncology, phytochemistry), the two current meta-analyses, the Cochrane review, Examine's outcome grades, Grokipedia's species article, and clinician-facing nutraceutical reviews, to check that no claimed benefit domain was omitted. -->

### High 🟩 🟩 🟩

#### Reduction in Migraine Attack Frequency ⚠️ Conflicted

Taken daily, feverfew lowers how many migraine attacks occur per month. The proposed route is suppression of trigeminal inflammatory signalling together with reduced serotonin release from platelets. Nine double-masked placebo-controlled trials in 899 participants were pooled in [a 2025 meta-analysis](https://pubmed.ncbi.nlm.nih.gov/41422534/) showing about one fewer attack per month with consistent results, and the largest single trial of the standardised extract found [1.9 versus 1.3 fewer attacks](https://pubmed.ncbi.nlm.nih.gov/16232154/). The evidence base is large but internally inconsistent: [Cochrane](https://pubmed.ncbi.nlm.nih.gov/25892430/) rates its certainty low because the positive trials were small and two rigorous trials were null.  

**Magnitude:** 0.6 fewer attacks per 28 days than placebo in the Cochrane analysis; pooled mean difference −1.11 attacks (95% confidence interval −1.23 to −0.99 — the range within which the true value most likely falls).  

### Medium 🟩 🟩

#### Shorter Duration of Individual Attacks ⚠️ Conflicted

Attacks that do occur may end sooner. The mechanism is presumed to be the same anti-inflammatory damping applied to an attack already under way rather than to its initiation. The [2025 pooled analysis](https://pubmed.ncbi.nlm.nih.gov/41422534/) found a statistically significant shortening across the contributing trials, but heterogeneity was near-total, meaning the individual studies disagree sharply about the size of the effect; the earlier Nottingham crossover trial found [no change in the duration of individual attacks](https://pubmed.ncbi.nlm.nih.gov/2899663/) at all. The estimate should therefore be read as a direction rather than a dependable number.  

**Magnitude:** −4.43 hours per attack (95% confidence interval −7.63 to −1.23), with heterogeneity of 98% across the pooled trials.  

### Low 🟩

#### Reduced Attack Pain Intensity ⚠️ Conflicted

Whether feverfew makes attacks less painful, as opposed to less frequent, is unresolved. Early dried-leaf trials reported lower severity scores; the current pooled estimate points the same way but does not reach statistical significance, and [Cochrane](https://pubmed.ncbi.nlm.nih.gov/25892430/) reports no significant difference on intensity in the one large rigorous trial.  

**Magnitude:** Pooled mean difference −0.63 on the trials' pain scales (95% confidence interval −1.48 to 0.21), p = 0.14 (the p-value is the probability the result arose by chance; 0.14 is too high to exclude it) — [not statistically significant](https://pubmed.ncbi.nlm.nih.gov/41422534/).  

#### Faster Relief of an Attack Already Under Way

A sublingual feverfew-plus-ginger preparation, taken at the first mild phase of an attack, outperformed placebo on two-hour pain freedom in a 60-patient pilot. This is a combination product funded by its manufacturer, so the feverfew contribution cannot be isolated.  

**Magnitude:** 32% versus 16% pain-free at two hours, and 63% versus 39% achieving pain relief, in [the pilot trial](https://pubmed.ncbi.nlm.nih.gov/21631494/).  

#### Protection of Skin Against Ultraviolet-Induced Redness

A parthenolide-depleted topical extract scavenges reactive oxygen species, restores cellular thiols, and induces DNA-repair enzymes, reducing visible redness after ultraviolet exposure in a controlled clinical test. The work was performed by employees of the consumer-goods manufacturer that developed the extract.  

**Magnitude:** Direction is a significant reduction in erythema (skin redness) versus placebo 24 hours after ultraviolet exposure, holding for topical application of the parthenolide-depleted extract only; [the source study](https://pubmed.ncbi.nlm.nih.gov/18071724/) reports no outcome figure for the size of that reduction.  

### Speculative 🟨

#### Systemic Anti-Inflammatory and Immune Modulation

Parthenolide blocks NF-κB in cells, and feverfew was traditionally an arthritis remedy, but the one controlled human trial was [negative](https://pubmed.ncbi.nlm.nih.gov/2673080/) and swallowed parthenolide reaches no measurable blood level. The basis is mechanistic.  

#### Selective Activity Against Cancer Stem Cells

Parthenolide and its orally bioavailable analogue eradicate leukaemia stem cells while sparing normal blood-forming cells [in preclinical models](https://pubmed.ncbi.nlm.nih.gov/17804695/). No human study has tested either against a cancer endpoint; the basis is animal data alone.  

  
## Benefit-Modifying Factors

* **Baseline attack frequency:** The clearest predictor of response. The extract's [dose-finding trial](https://pubmed.ncbi.nlm.nih.gov/12230594/) reached significance only in the pre-specified subgroup with at least four attacks per 28 days at baseline; people with sparse attacks have little room to demonstrate benefit.  

* **Preparation and parthenolide content:** Dried whole leaf, alcoholic extract, and supercritical carbon-dioxide extract are not interchangeable. Null trials cluster among alcoholic extracts and unstandardised leaf; positive results cluster among fresh or freeze-dried leaf and the standardised carbon-dioxide extract.  

* **Glutathione-transferase variants:** Parthenolide is cleared largely by glutathione conjugation, so the common *GSTM1* and *GSTT1* null deletions (missing copies of genes for enzymes that neutralise reactive compounds) plausibly raise tissue exposure and, with it, both effect and sensitisation risk.  

* **Sex:** Migraine is roughly threefold more common in women, and every feverfew trial population has been female-predominant. No trial has reported a sex-stratified efficacy analysis, so sex differences in benefit remain untested rather than absent.  

* **Pre-existing conditions:** Concurrent medication-overuse headache blunts any prophylactic agent. Chronic migraine (15 or more headache days monthly) was an exclusion in every feverfew trial, so benefit in that group is unevidenced.  

* **Age:** Trials enrolled adults aged roughly 18–65. Benefit in adults past 65 is untested, and unstandardised dried leaf becomes a poorer choice with age as concurrent antiplatelet therapy grows more common.  

  
## Potential Risks & Side Effects

<!-- Before writing this section I ran dedicated searches of the drug- and herb-reference literature and the primary safety record: the Cochrane adverse-event tables, the Ernst & Pittler safety systematic review, the Kashkooe toxicity review, PubMed searches on feverfew contact dermatitis, feverfew safety, feverfew coagulation, and the Compositae/sesquiterpene-lactone dermatology series, plus drug-reference and consumer-safety summaries (drugs.com, Memorial Sloan Kettering, ConsumerLab's surgery and breastfeeding answers). -->

### High 🟥 🟥 🟥

#### Gastrointestinal Upset and Oral Mucosal Ulceration

The commonest complaints are nausea, abdominal discomfort, and aphthous ulcers (small painful mouth sores), the latter reported particularly with chewing fresh leaf, where direct contact with sesquiterpene lactones irritates the mucosa. These are the adverse events that most often drove withdrawal in trials. [Cochrane](https://pubmed.ncbi.nlm.nih.gov/25892430/) records them as mild and transient across all six included trials, with no serious events; the [safety-focused systematic review](https://pubmed.ncbi.nlm.nih.gov/11276299/) reaches the same conclusion. Encapsulated leaf and the carbon-dioxide extract largely avoid the oral lesions.  

**Magnitude:** In the largest trial, adverse events judged possibly treatment-related occurred in [8.4% on the extract versus 10.2% on placebo](https://pubmed.ncbi.nlm.nih.gov/16232154/); across Cochrane's trials, gastrointestinal complaints and mouth ulcers were the most frequent events reported.  

### Medium 🟥 🟥

#### Allergic Contact Dermatitis and Asteraceae Cross-Sensitisation

Parthenolide is a strong skin sensitiser and the reference allergen for the Compositae (daisy) family, [cross-reacting with chamomile, arnica, chrysanthemum, and ragweed](https://pubmed.ncbi.nlm.nih.gov/6195862/). Sensitised people react not only to handling the plant but to [airborne parthenolide](https://pubmed.ncbi.nlm.nih.gov/17300241/) and to cosmetic creams containing whole feverfew. Reactions are eczematous, can generalise, and are usually permanent once established; the parthenolide-depleted cosmetic extract was engineered specifically to remove this hazard. Risk falls on gardeners, growers, and topical users far more than on capsule users.  

**Magnitude:** Direction is a steep rise in reaction risk with cutaneous or airborne exposure and in anyone already patch-test positive to the sesquiterpene lactone mix; the dermatology literature reports patch-test series rather than an incidence figure for feverfew users.  

### Low 🟥

#### Post-Feverfew Withdrawal Syndrome

Abruptly stopping after months of daily use has been described as producing rebound headache, anxiety, poor sleep, and joint stiffness lasting days to weeks. The mechanism is unknown. The evidence is user-survey and case description assembled in [the safety review](https://pubmed.ncbi.nlm.nih.gov/11276299/), not controlled withdrawal data.  

**Magnitude:** Not quantified in available studies. No controlled discontinuation trial has ever been run, so the syndrome's frequency and severity rest entirely on retrospective self-report from long-term users.  

#### Prolonged Bleeding and Altered Coagulation

Feverfew extracts [suppress granule secretion in platelets and white cells](https://pubmed.ncbi.nlm.nih.gov/2860288/), giving a plausible antiplatelet action. A [documented case](https://pubmed.ncbi.nlm.nih.gov/34434419/) links high-dose use to deranged coagulation tests and prolonged vaginal bleeding, judged a probable reaction. Clinical relevance is greatest around surgery and with concurrent anticoagulants.  

**Magnitude:** Direction is toward longer bleeding times, holding at high or prolonged doses and when combined with antiplatelet or anticoagulant agents; the literature reports no incidence figure, only in vitro platelet data and isolated cases.  

#### Uterine Stimulation in Pregnancy

Feverfew's traditional role as an emmenagogue (an agent used to bring on menstruation) and its uterine-stimulant reputation underlie a universal contraindication in pregnancy, echoed by the [safety review](https://pubmed.ncbi.nlm.nih.gov/11276299/) and by breastfeeding guidance. [A rat reproductive screen](https://pubmed.ncbi.nlm.nih.gov/16781113/) found fetal growth effects and embryo toxicity; no human teratogenicity data exist.  

**Magnitude:** Not quantified in available studies. Because avoidance is universally advised, no exposure cohort or controlled study of pregnancy outcomes after feverfew use has ever been assembled.  

### Speculative 🟨

#### Sustained Cyclooxygenase-2 Suppression and Cardiovascular Risk

Sustained cyclooxygenase-2 inhibition drives the cardiovascular signal of some anti-inflammatory drugs, and [one systematic review](https://pubmed.ncbi.nlm.nih.gov/22435410/) flagged that years of feverfew use have never been studied. No event data exist; the basis is mechanistic reasoning.  

#### Depletion of Cellular Antioxidant Reserves

Parthenolide's reactive lactone ring consumes glutathione, the cell's principal thiol antioxidant. Whether chronic high-dose intake lowers antioxidant reserve in humans is untested; the basis is cell-culture chemistry, with no clinical or biomarker study addressing it.  

  
## Risk-Modifying Factors

* **Prior Compositae sensitisation:** The single strongest risk multiplier. Established patch-test positivity to the sesquiterpene lactone mix, arnica, or chamomile predicts reaction to feverfew, including to creams and to airborne exposure while gardening.  

* **Atopic background:** Atopic dermatitis and long-standing hand eczema raise Compositae sensitisation rates in dermatology series, because a compromised skin barrier increases allergen penetration during handling of the fresh plant.  

* **Glutathione-transferase variants:** *GSTM1* and *GSTT1* null genotypes remove enzymes that neutralise reactive compounds, so carriers plausibly retain parthenolide longer, which may increase both sensitisation and mucosal irritation. Not yet tested directly.  

* **Baseline coagulation and platelet markers:** A platelet count below 100 × 10⁹/L, a raised clotting time, or von Willebrand disease (an inherited clotting-factor deficiency) turns a theoretical antiplatelet effect into a real bleeding hazard.  

* **Sex:** The one [published bleeding case](https://pubmed.ncbi.nlm.nih.gov/34434419/) involved menorrhagia (abnormally heavy menstrual bleeding) in a premenopausal woman on high-dose feverfew, so menstruating women carry a documented, if rare, sex-specific manifestation of this risk.  

* **Pre-existing conditions:** Active peptic ulcer, inflammatory bowel disease, and recent gastrointestinal bleeding amplify the gastrointestinal side-effect burden; scheduled surgery within two weeks converts platelet inhibition into a perioperative problem.  

* **Age:** Adults past 65 more often take antiplatelet or anticoagulant drugs and tolerate gastrointestinal irritation less well, so the same dose carries more consequence than in a 40-year-old.  

  
## Key Interactions & Contraindications

* **Anticoagulants (warfarin, apixaban, rivaroxaban, dabigatran):** Caution; potential additive bleeding through platelet inhibition. Mitigation: avoid combination where possible, or check international normalised ratio (a standardised measure of clotting time) at two and six weeks after starting.  

* **Antiplatelet drugs (aspirin, clopidogrel, ticagrelor, dipyridamole):** Caution; additive suppression of platelet granule release, raising bruising and bleeding risk. Mitigation: reserve feverfew for patients not on dual antiplatelet therapy, and stop it 14 days before any procedure.  

* **Over-the-counter non-steroidal anti-inflammatory drugs (NSAIDs — the ibuprofen class: ibuprofen, naproxen, high-dose aspirin):** Caution; overlapping cyclooxygenase and antiplatelet effects plus additive gastric irritation. Mitigation: take feverfew with food and avoid daily concurrent use during acute-treatment cycles.  

* **Serotonergic migraine and psychiatric drugs (sumatriptan, rizatriptan, sertraline, venlafaxine):** Monitor; feverfew alters platelet serotonin release, a theoretical rather than documented interaction. Mitigation: no dose change indicated; watch for unusual bruising, since antidepressants of this type independently impair platelet function.  

* **Antiplatelet supplements (ginkgo, garlic extract, high-dose fish oil, vitamin E, ginger, curcumin, white willow bark):** Caution; cumulative bleeding risk. Mitigation: limit to one antiplatelet botanical at a time, and separate any surgical or dental procedure by 14 days.  

* **Additive migraine-prophylactic supplements (magnesium, riboflavin, coenzyme Q10, butterbur):** Beneficial additive effect on attack frequency without a known safety conflict. Mitigation: introduce one agent at a time over four-week blocks so that responders and non-responders can be told apart.  

* **Elective surgery, dental extraction, and neuraxial anaesthesia (spinal or epidural injection):** Relative contraindication around the procedure because of platelet effects. Mitigation: discontinue at least 14 days beforehand and disclose use at the pre-anaesthetic assessment.  

* **Other interventions — topical feverfew cosmetics:** Caution; whole-feverfew creams have provoked dermatitis in people already sensitised, whereas parthenolide-depleted cosmetic extracts have not. Mitigation: choose the parthenolide-depleted form, or patch-test on the forearm for 48 hours.  

**Populations who should avoid Feverfew:**  

* Pregnancy in any trimester, and breastfeeding, on grounds of uterine-stimulant activity and absent safety data  
* Known allergy or patch-test positivity to Asteraceae/Compositae plants (chamomile, arnica, ragweed, chrysanthemum, echinacea)  
* Children under 2 years, and children generally in the absence of paediatric dosing evidence  
* Bleeding disorders, or a platelet count below 100 × 10⁹/L  
* Therapeutic anticoagulation with an unstable or supratherapeutic international normalised ratio (above 3.0)  
* Elective surgery, dental extraction, or neuraxial anaesthesia scheduled within 14 days  

  
## Risk Mitigation Strategies

* **Choose a standardised extract over raw leaf:** A supercritical carbon-dioxide extract at 6.25 mg three times daily gives reproducible parthenolide content, avoiding both the under-dosing of degraded leaf and the mucosal ulceration caused by chewing fresh leaves.  

* **Never chew the fresh plant:** Swallowing encapsulated leaf or extract removes direct mucosal contact with sesquiterpene lactones, which is the specific exposure route behind aphthous mouth ulcers and lip and tongue swelling.  

* **Screen for daisy-family allergy before starting:** A history of reaction to chamomile, arnica, ragweed, or chrysanthemum, or a positive sesquiterpene lactone mix patch test, identifies the people in whom allergic contact dermatitis and systemic allergic dermatitis are likely.  

* **Wear gloves when handling the plant:** Growers and gardeners acquire sensitisation cutaneously and from airborne parthenolide, so gloves plus long sleeves and avoiding the plant in flower prevent occupational Compositae dermatitis.  

* **Taper over two to four weeks rather than stopping abruptly:** Reducing the daily dose by roughly a quarter each week is the practical countermeasure to post-feverfew withdrawal syndrome, with its rebound headache, anxiety, insomnia, and joint stiffness.  

* **Stop 14 days before any procedure:** A two-week washout exceeds the platelet lifespan turnover needed to restore normal granule secretion, addressing the perioperative bleeding risk documented in the coagulation case report.  

* **Take with food and cap the dose:** Keeping intake within 100–300 mg of standardised leaf daily and dosing after meals reduces the nausea and abdominal discomfort that account for most trial withdrawals.  

* **Set a 12-week decision point:** Recording monthly migraine days and stopping if attacks have not fallen by at least 30% prevents indefinite exposure to bleeding, sensitisation, and withdrawal risk for no benefit.  

  
## Therapeutic Protocol

* **Standardised extract approach:** The regimen with the strongest trial support is the supercritical carbon-dioxide leaf extract, MIG-99, at 6.25 mg three times daily, continued for at least 12–16 weeks before judging response.  

* **Dried-leaf approach:** The older and cheaper route is 50–150 mg of dried or freeze-dried leaf daily, standardised to at least 0.2% parthenolide; Examine's dosing range is 100–300 mg at 0.2–0.4% parthenolide.  

* **Competing approaches side by side:** Conventional prophylaxis (topiramate, propranolol, antibodies against CGRP — calcitonin gene-related peptide, a nerve signal that drives migraine) and botanical prophylaxis are alternative first moves, differing in evidence weight and cost rather than one being standard.  

* **Who popularised each:** Dried leaf was brought into testing by Johnson's group at the City of London Migraine Clinic; the carbon-dioxide extract by Weber & Weber with Diener's Essen department; the sublingual feverfew–ginger route by Cady's Headache Care Center.  

* **Time of day:** Prophylactic dosing is not time-critical; the extract's three-times-daily schedule spreads exposure across waking hours. Morning dosing with breakfast is customary and reduces the nausea some users report on an empty stomach.  

* **Half-life and its consequence:** Parthenolide's plasma half-life is under about an hour and oral levels are undetectable, which is the pharmacological reason divided daily dosing is used and why missed doses are not made up.  

* **Single versus split dosing:** The positive extract trials used three divided doses; dried-leaf trials used a single daily capsule. Splitting is the better-evidenced choice for extracts, while single dosing suits leaf preparations and aids adherence.  

* **Genetic considerations:** No pharmacogenetic test guides feverfew dosing. *GSTM1* and *GSTT1* null status is the only plausible modifier of parthenolide clearance, and *MTHFR* variants (a gene for the folate-processing enzyme) affect migraine susceptibility rather than feverfew response.  

* **Sex-based considerations:** No trial has reported sex-stratified dosing or efficacy. The one [documented case of menorrhagia](https://pubmed.ncbi.nlm.nih.gov/34434419/) arose in a premenopausal woman taking 2400 mg daily, well above the customary range.  

* **Age-related considerations:** No dose adjustment is established for older adults, but starting at the low end (50 mg leaf daily) is prudent past 65, where concurrent antiplatelet drugs and reduced gastric tolerance are common.  

* **Baseline biomarkers:** Baseline attack frequency is the practical response predictor: the [extract trials](https://pubmed.ncbi.nlm.nih.gov/12230594/) favoured people recording at least four attacks per 28 days. A baseline headache diary is therefore part of the protocol, not an optional extra.  

* **Pre-existing conditions:** Medication-overuse headache should be resolved before a prophylactic trial, since it masks response. Active peptic ulcer, bleeding disorders, and Compositae allergy redirect the choice to a different prophylactic entirely.  

  
## Discontinuation & Cycling

* **Intended duration:** Feverfew is used continuously rather than as a short course. Examine notes the effect building over the first 12 weeks and being maintainable indefinitely; the longest controlled exposure is only 16 weeks, so open-ended use is unstudied.  

* **Withdrawal effects:** Abrupt cessation after months of use has been described as producing rebound headache, anxiety, insomnia, and muscle and joint stiffness — the so-called post-feverfew syndrome, documented from user surveys rather than controlled withdrawal studies.  

* **Tapering protocol:** A stepwise reduction of about 25% of the daily dose each week over two to four weeks is the practical approach, prompted by the withdrawal reports; no trial has compared tapering with abrupt cessation.  

* **Cycling:** No evidence supports cycling for efficacy, and no tolerance has been demonstrated. Because benefit accrues over roughly 12 weeks, planned interruptions would more likely forfeit effect than preserve it.  

* **Re-assessment point:** A defined stop rule matters more than cycling. If monthly migraine days have not fallen meaningfully by week 12–16, continuing exposes the user to sensitisation and bleeding risk without return.  

  
## Sourcing and Quality

* **Parthenolide standardisation is the decisive specification:** A stated minimum of 0.2% parthenolide is the threshold Health Canada requires for licensed products. Unstandardised leaf products have historically been found to contain little or no detectable parthenolide.  

* **Preparation type determines the evidence that applies:** Supercritical carbon-dioxide extract and fresh or freeze-dried leaf carry the positive trial results; alcoholic extracts feature in the null trials. The label should name the extraction method, not merely "feverfew".  

* **Third-party testing:** Because feverfew is regulated as a dietary supplement rather than a drug, an independent certification mark — NSF, USP Verified, Informed Choice, or a published certificate of analysis — is the only routine assurance of identity and content.  

* **Botanical identity and adulteration:** [DNA-barcoding surveys of herbal capsules](https://pubmed.ncbi.nlm.nih.gov/31708772/) have found widespread substitution, contaminant species, and undeclared fillers across the category — 27% of nearly 6,000 products tested worldwide; species-verified sourcing matters more here than for synthetic compounds.  

* **Reputable suppliers:** Among widely available brands, NOW Foods (standardised to 0.5–0.7% parthenolide) and Nature's Way (standardised extract at 0.7% parthenolide) publish parthenolide content and testing documentation; European pharmacies stock registered traditional herbal products carrying a regulator-assessed specification.  

* **Storage and shelf life:** Parthenolide degrades with heat, light, and time. Recent lots, opaque containers, and storage below 25 °C away from light preserve content; a product without a use-by date is unreliable.  

  
## Practical Considerations

* **Time to effect:** Not an acute remedy. Trials measured outcomes over months, and Examine describes the effect strengthening across the first 12 weeks; a fair trial therefore runs 12–16 weeks with a headache diary throughout.  

* **Common pitfalls:** Judging response after two or three weeks; using an alcoholic extract or unstandardised leaf and inferring feverfew "does not work"; chewing fresh leaves and acquiring mouth ulcers; and stopping abruptly, then attributing withdrawal headaches to returning migraine.  

* **Regulatory status:** In the United States feverfew is a dietary supplement under the Dietary Supplement Health and Education Act, so it is not assessed for efficacy before sale. In the United Kingdom and European Union it is sold as a registered traditional herbal medicine.  

* **Cost and accessibility:** Neither expensive nor hard to obtain — typically well under a dollar a day, sold without prescription in pharmacies and online, which places it among the cheapest migraine prophylactics available.  

* **Payer and funding incentives:** Feverfew costs a fraction of anti-CGRP antibody therapy, so insurers and national health systems have a structural interest in favouring cheap options, while trial funding flows overwhelmingly from manufacturers of branded prophylactics — bias operating in opposite directions.  

* **Guideline positions and who issues them:** Neurology and headache societies have graded feverfew inconsistently, from a positive recommendation to no recommendation, [as compared across guidelines by Rajapakse & Pringsheim](https://pubmed.ncbi.nlm.nih.gov/26954394/). Those societies' members derive practice income largely from prescribing the branded prophylactics feverfew would displace.  

  
## Interaction with Foundational Habits

* **Sleep:** Indirect and generally favourable. Feverfew has no sedative or stimulant action and does not disturb sleep architecture; the benefit is second-order, through fewer nocturnal and early-morning attacks. The exception is withdrawal, where abrupt cessation after long use has been reported to cause insomnia and restlessness for days to weeks.  

* **Nutrition:** Direct and practically relevant. Taking feverfew with food blunts the nausea and abdominal discomfort that drove most trial withdrawals. It depletes no known nutrient. In people sensitised to Asteraceae, dietary cross-reactants — chamomile tea, echinacea, lettuce, artichoke, and sunflower seeds — can provoke systemic allergic dermatitis.  

* **Exercise:** Essentially none in either direction. Feverfew neither blunts training adaptation nor enhances performance, and no timing relative to workouts is indicated. The only practical caveat is contact sport or activity with a high bruising or bleeding risk, where the antiplatelet effect argues for caution alongside other antiplatelet agents.  

* **Stress management:** Indirect. Feverfew has no measured effect on cortisol or the stress response, and no trial has assessed it. Because psychological stress is among the most common migraine triggers, stress-reduction practice and feverfew act on the same outcome by separate routes and are reasonably paired rather than substituted.  

  
## Monitoring Protocol & Defining Success

Baseline assessment centres on a four-week headache diary recording migraine days, attack duration, pain intensity, and acute-medication doses; without it, response cannot be judged and the response-predicting threshold of four or more attacks per 28 days cannot be applied. Baseline bloods are modest: a full blood count with platelets, prothrombin time with international normalised ratio for anyone on or likely to start anticoagulants, and liver enzymes as a general tolerability reference. Patch testing to the sesquiterpene lactone mix precedes topical or occupational exposure in anyone with an atopic or plant-allergy history.  

Thereafter the diary is reviewed at 4, 8, and 12–16 weeks, and every 6 months once stable. Coagulation testing is repeated at 2 and 6 weeks only where an anticoagulant or antiplatelet drug is co-administered; otherwise bloods are repeated annually.  

| Biomarker | Optimal Functional Range | Why Measure It? | Context/Notes |
|---|---|---|---|
| Monthly migraine days (headache diary) | ≥50% reduction from a 4-week baseline; ≥30% is the minimum signal worth continuing for | The only outcome the trials actually measured; defines success | Recorded daily rather than retrospectively. Baseline of ≥4 attacks per 28 days predicted response in the extract trials |
| Acute medication days per month | <10 days/month for simple analgesics, <5 for triptans | Detects medication-overuse headache, which masks any prophylactic effect | Triptans are the prescription drugs taken to stop an attack once it starts (sumatriptan, rizatriptan). Rising use despite falling attack counts signals overuse, not treatment failure |
| Platelet count | 150–400 × 10⁹/L | Baseline reference before an agent with antiplatelet activity | Part of a full blood count (CBC). Below 100 × 10⁹/L, feverfew is contraindicated |
| Prothrombin time / international normalised ratio (INR) | 0.8–1.2 if not anticoagulated; within the prescribed target if anticoagulated | Detects the coagulation derangement documented with high-dose use | Only needed with concurrent anticoagulants or before surgery. Conventional labs report INR without a functional-medicine variant |
| Alanine aminotransferase (ALT) | 10–26 U/L (women), 10–33 U/L (men) | General tolerability reference for a long-term botanical | ALT is a liver enzyme that leaks into blood when liver cells are stressed. Fasting not required. Conventional upper limits run to 40–55 U/L, well above the functional range |
| High-sensitivity C-reactive protein (hs-CRP) | <1.0 mg/L | Tracks the inflammatory tone feverfew is proposed to lower | hs-CRP is a general blood marker of body-wide inflammation. Drawn when free of infection; a recent cold invalidates the result. Best paired with fasting glucose for context |
| Sesquiterpene lactone mix patch test | Negative at 48 and 96 hours | Identifies the Compositae sensitisation that drives the main non-trivial harm | No numeric range exists; result is read as positive or negative by a dermatologist. Performed before topical or occupational exposure |

Qualitative markers worth tracking alongside the labs:  

* Attack severity and functional disability on attack days, rated 0–10, rather than attack counts alone  
* Early warning symptoms — yawning, food craving, mood shift — and whether they still progress to full attacks  
* Nausea, abdominal discomfort, and any mouth ulceration, which are the earliest signals of intolerance  
* Sleep quality and daytime energy, both indirectly improved when nocturnal attacks fall  
* Skin condition on the hands and face, especially in anyone handling the plant or using topical products  
* Bruising, gum bleeding, or unusually heavy menstrual flow, the practical early markers of the platelet effect  

  
## Emerging Research

* **Parthenolide in Alzheimer's disease biology:** [NCT07236190](https://clinicaltrials.gov/study/NCT07236190), an open-label Phase 2 trial at Massachusetts General Hospital, is giving a parthenolide-plus-ipriflavone combination to 40 adults with early Alzheimer's changes, tracking plasma p-tau217 (a blood marker of tau protein damage) over six months. Recruiting since April 2026.  

* **Orally bioavailable parthenolide analogues:** The dimethylamino analogue [eradicates leukaemia stem cells while sparing normal progenitors](https://pubmed.ncbi.nlm.nih.gov/17804695/) (Guzman et al., 2007) and solves parthenolide's absorption problem. Whether that translates to humans is the single largest open question for feverfew's non-migraine claims.  

* **Standardisation as the rate-limiting variable:** The [2025 meta-analysis](https://pubmed.ncbi.nlm.nih.gov/41422534/) of Nelaturi et al. closes by calling for large trials with standardised preparations. Until extraction method and parthenolide content are fixed across trials, pooled estimates will keep inheriting the heterogeneity that currently undermines them.  

* **Evidence that could weaken the case:** The two rigorous null trials — the [dose-finding study's intention-to-treat analysis](https://pubmed.ncbi.nlm.nih.gov/12230594/) (Pfaffenrath et al., 2002; counting every randomised participant, including drop-outs) and the [riboflavin–magnesium–feverfew combination trial](https://pubmed.ncbi.nlm.nih.gov/15447697/) (Maizels et al., 2004) — remain unexplained. A properly powered replication returning null would collapse the current pooled estimate.  

* **The bioavailability problem:** [Undetectable plasma parthenolide at 4 mg daily](https://pubmed.ncbi.nlm.nih.gov/15122077/) (Curry et al., 2004) is the deepest challenge to every systemic mechanism proposed. Pharmacokinetic work using purified compound at higher doses would either rescue the mechanism or force a local-action explanation.  

* **Registry evidence beyond migraine:** Completed registered work remains combination-based, such as [NCT04759040](https://clinicaltrials.gov/study/NCT04759040), a 120-participant trial of a feverfew-containing migraine formulation. No registered trial currently tests feverfew monotherapy for any longevity-relevant endpoint.  

  
## Conclusion

Feverfew is a garden herb whose dried leaf and concentrated leaf extracts are taken to make migraine attacks less frequent. Its bitter main compound blocks a central inflammatory switch inside cells and curbs the release of a signalling chemical stored in blood platelets, which is the usual explanation for whatever benefit it delivers.  

The clearest signal is a modest drop in the number of attacks each month. It is real but small, and the record is uneven: several early positive studies were tiny, two careful ones found nothing, and summaries that combine studies reach differing conclusions depending on which ones they admit. Effects on how long attacks last and how much they hurt are weaker or absent. Claims of broad anti-inflammatory and anti-cancer action rest on cells and animals; human testing in joint disease showed no benefit, and the amount of active compound reaching the blood after a swallowed capsule appears vanishingly small, which undercuts any whole-body explanation.  

Harms are mostly minor — stomach upset and mouth sores — with one substantial exception: a genuine and usually permanent skin allergy in people sensitive to daisies and related plants. Stopping abruptly after long use has been described as bringing back headaches, restlessness, and stiffness. Pregnancy is the one absolute exclusion.  

Much of the strongest trial evidence was paid for by the companies selling the preparations tested, and the headache societies whose ratings shape practice draw member income largely from the prescription alternatives.  

**[Top](#top) - [Benefits](#expected-benefits) - [Risks](#potential-risks--side-effects) - [Protocol](#therapeutic-protocol)**