A garden herb taken daily to make migraine attacks less frequent. The drop in monthly attacks is real but small; studies disagree, and effects on attack length and pain are weaker or absent. Anti-inflammatory and anti-cancer claims rest on cells and animals. Harms are mostly minor, apart from a usually permanent daisy-family allergy. Pregnancy is the one absolute exclusion. (Full Review)
| Marker | Target | Why |
|---|---|---|
| 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 |
| Acute medication days per month | <10 days/month for simple analgesics, <5 for triptans | Detects medication-overuse headache, which masks any prophylactic effect |
| Platelet count | 150–400 × 10⁹/L | Baseline reference before an agent with antiplatelet activity |
| 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 |
| Alanine aminotransferase (ALT) | 10–26 U/L (women), 10–33 U/L (men) | General tolerability reference for a long-term botanical |
| High-sensitivity C-reactive protein (hs-CRP) | <1.0 mg/L | Tracks the inflammatory tone feverfew is proposed to lower |
| Sesquiterpene lactone mix patch test | Negative at 48 and 96 hours | Identifies the Compositae sensitisation that drives the main non-trivial harm |
Cadence: Headache diary reviewed at 4, 8, and 12–16 weeks, and every 6 months once stable. Coagulation testing repeated at 2 and 6 weeks only where an anticoagulant or antiplatelet drug is co-administered; otherwise bloods are repeated annually.