"Natural" or Alternative Medications for Migraine Prevention

Randolph W. Evans, MD; Frederick R. Taylor, MD


Headache. 2006;46(6):1012-1018. 

In This Article


Feverfew (Tanacetum parthenium) is species-specific dried chrysanthemum leaves. According to 1 peer reviewed report, preparations of feverfew have shown a >400% variation in dosage strength of the known active ingredient parthenolide.[9] Additionally, because feverfew also contains melatonin, some uncertainty exists as to whether parthenolide is even the major active ingredient in feverfew, although the MIG-99 trials of parthenolide might provide some insight into this issue.[10] Despite a very recent positive MIG-99 trial, the totality of evidence, including systematic reviews, does not support feverfew as representing a "definitely effective" therapy for migraine, nor has its safety with long-term use been established. In short, there is Grade B evidence of feverfew's utility as a migraine therapy. Should another successful trial of MIG-99 6.25 mg pass peer review and be reported, then Grade A evidence would exist (albeit with what might prove to be a rather small therapeutic gain [TG] over placebo).

In a systematic review, Vogler et al reported on randomized controlled trials (RCTs) involving feverfew for migraine prophylaxis conducted prior to 1998.[11,12,13,14,15,16] Five studies qualified by Jadad score as adequate; 1 has been published in abstract form only, and only 216 subjects in total have been studied.[17] Vogler et al concluded, "In view of the popularity of feverfew, perhaps the most striking finding was the paucity and low average quality of the existing RCTs on the subject." Three studies, including two-thirds of the total patients, showed greater benefit from feverfew than from placebo, but the study by De Weerdt et al, with the highest Jadad score and use of a standardized and constant concentration of parthenolides, showed no benefit.[15] Perhaps, this was due to the latter's use of an alcoholic extract, as 2 of the positive studies (totaling 67 patients) involved a powdered extract, while the active agent used in negative studies was an ethanolic extract.[12,13] With this "evidence" that the feverfew extraction technique might be critical to the agent's efficacy, work commenced on a supercritical CO2 extraction method which produced a highly stable, "highly enriched" parthenolide extract named MIG-99. In 2002, Pfaffenrath et al reported the results of a dose-finding prospective RCT, involving 147 subjects treated with this extraction.[18] The investigators studied three doses of MIG-99, and none was significant for the primary endpoint relative to placebo. A subset of subjects with episodic migraine of high frequency did seem to benefit, and the researchers consequently called for more study. In a 2004 Cochrane Database Systematic Review, Pittler and Ernst reviewed 5 double-blind trials with high Jadad scores involving a total of 343 patients, and noting the mixed results from the RCTs, the authors concluded that feverfew lacked convincing evidence of efficacy.[19] In follow-up to the 2002 study, Diener et al performed a second multi-center RCT, wherein patients randomized to the active arm received MIG-99 6.25 mg TID.[20] The investigators reported reduction in headache attacks per month during the second and third recording periods (from 4.8 attacks to 2.9 with MIG-99 vs. 3.5 attacks with placebo (ie, a reduction in attacks/month of 1.9 vs. 1.3) attacks per month (P = .0456, and OR 3.4). Greater than 50% reduction in headache attacks per 28 days occurred in 30.3% of the feverfew group and 17.3% of the placebo group, a TG of 13% (P = .047). Adverse events occurred in 8.4% of subjects receiving feverfew and 10.2% of placebo controls.

From the RCT date, feverfew adverse events include sore mouth and tongue (including ulcers), swollen lips, loss of taste, abdominal pain, and GI disturbances. A "post-feverfew syndrome" of joint stiffness and aches and increasing headaches was reported in 25% of placebo subjects in the trial which randomized patients who had been using feverfew for several years to continued feverfew therapy or placebo.[12] In the MIG-99 RCTs, adverse events were similar between all groups, with ≥1 adverse event occurring in 35%.[19] Adequate long-term studies, including extension safety trials, appear not to have been performed as of this writing.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.