BiomaLearnFoodsNutrition factsNutrientsGuidesAnswersEvidence

Butterbur: it worked in the trials, and the recommendation was withdrawn anyway

One name covers two different things here, and the science belongs to only one of them. Everything about migraine below comes from a proprietary root extract of Petasites hybridus sold as Petadolex, not from the vegetable in the produce aisle. A review of two randomized trials, one of 60 patients and one of 233, found that 150 mg a day reduced attack frequency, and produced more patients whose attacks halved, than 100 mg a day or placebo over three to four months. The two trials were too different to combine, so the reviewers left them unpooled, graded the evidence moderate only for a dose above the recommended one, and declined to recommend the extract until longer and more rigorous studies exist. The larger trial is where the familiar number comes from. In 245 patients aged 18 to 65 who had two to six attacks a month, 75 mg twice a day cut attack frequency by 48 percent over four months against 26 percent on placebo, in a per-protocol analysis rather than intention to treat. In that arm 68 percent of patients halved their attacks, against 49 percent on placebo. The dose below it landed at 36 percent and did not reach significance against placebo, at p = 0.127, so the amount on the label decides whether any of this applies. Then the recommendation reversed. In 2012 the American Academy of Neurology and the American Headache Society rated Petasites effective at their top level and told doctors to offer it. In 2015 the Academy stopped recommending it, on liver toxicity rather than on any new doubt about whether it prevents attacks. In the cases NIH LiverTox summarizes, liver injury began 2 to 12 weeks after people started a commercial product, and two of them needed urgent liver transplants.

Every statement here is tied to a source. The badge says what kind of evidence stands behind it: A is a meta-analysis or systematic review, E is the position of an expert body without its own analysis. How we verify →

The numbers, per 3.5 oz (100 g)

USDA entrykcalProteinFatCarbsFiber
Butterbur, (fuki), raw140.39 g0.04 g3.61 g— g
Butterbur, canned30.11 g0.13 g0.38 g— g
Butterbur, cooked, boiled, drained, with salt80.23 g0.02 g2.16 g— g
Butterbur, cooked, boiled, drained, without salt80.23 g0.02 g2.16 g— g

These are the USDA entries that are butterbur, not foods made from it; they differ in cut, part, pack or preparation. Linked names have a page with the full profile and per-portion figures.

How much is actually in it

What a real portion looks like

What cooking and storage change

What it does to your health

Safety, limits and interactions

What people believe that the data does not support

Who this works differently for

What is still unknown

The bottom line

The liver concern is not hypothetical. The NIH LiverTox monograph records cases of clinically apparent liver injury from several commercial butterbur preparations, suspected to come from residual pyrrolizidine alkaloid contamination. In the cases it summarizes, injury began 2 to 12 weeks after the product was started, with fatigue, nausea and jaundice, and two people needed urgent liver transplants while the others recovered. There is no count of users behind those cases, so there is no rate. Those alkaloids come out during manufacture and not in your kitchen. The NIH position is that only products processed to remove them, and labeled or certified PA-free, should be considered at all. The label is not proof. A 2012 analysis of 21 commercial butterbur products in the United States, which NIH reports rather than we found in the trial literature, judged 7 of the 21 to have petasin present with no detectable alkaloids. The ceiling that exists is on the alkaloids, not on the extract. The European Medicines Agency's herbal committee sets 1.0 microgram a day as the adult limit for these alkaloids from herbal medicines, counting the plant and any contamination together. For children it scales the limit to body weight, 0.5 microgram a day at 44 lb (20 kg). Germany's 1992 rules, which EMA cites and which name Petasites, allow 1 microgram a day for at most 6 weeks a year, or 0.1 microgram a day without a time limit. LiverTox gives 100 to 150 mg a day in 2 or 3 doses as the typical dose of the extract, which is a typical dose rather than a maximum. No source we found sets an upper dose of the extract or a maximum length of use. LiverTox also says interactions between butterbur and medicines have not been defined, which is missing data rather than clearance. Some people should leave it alone. Products containing the alkaloids should not be used in pregnancy or while breastfeeding, because of possible birth defects or liver damage, and for PA-free products the honest answer is that little is known, which is missing data rather than clearance. Butterbur belongs to the ragweed family, so it can set off allergic reactions in people who react to ragweed, chrysanthemums, marigolds or daisies. The Canadian Headache Society in 2012 listed butterbur among 11 preventives carrying a strong recommendation for episodic migraine, next to topiramate and propranolol, and that list was written before the alkaloid question closed in. We found nothing at all on how much of the extract reaches the bloodstream. The vegetable is a separate subject with separate numbers. USDA holds four butterbur records, and raw fuki carries 31.5 mg vitamin C, 655 mg potassium and 0.103 mg copper per 3.5 oz (100 g). Across the four records vitamin C runs from 11.9 to 31.5 mg and potassium from 12 to 655 mg, depending on whether the entry is raw, canned or boiled. None of those figures describe the extract. Boiling greens in a little water and draining them keeps about 60 percent of the vitamin C and 90 percent of the potassium, a retention factor USDA assigns to the whole vegetable group rather than to butterbur by name, dated 1975 and published in 2007.

More from the same food group (vegetables and vegetable products)

Sources

  1. btbr-daily-r1-13 · food composition dataset
    Butterbur, (fuki), raw (FDC 170385, SR Legacy): vitamin C 31.5 mg, potassium 655 mg, copper 0.103 mg per 100 g
    USDA FoodData Central, SR Legacy, FDC 170385 · checked 2026-09-25
  2. btbr-daily-r1-03 · randomized controlled trial
    This is a three-arm, parallel-group, randomized trial comparing Petasites extract 75 mg bid, Petasites extract 50 mg bid, or placebo bid in 245 patients with migraine. ... Petasites 50 mg PO bid was not significantly more effective than placebo on the primary study endpoints.
    Neurology, 2004 · checked 2026-09-25
  3. btbr-hf2-05 · agency monograph
    the typical oral dosage is 100 to 150 mg per day in 2 to 3 divided doses. ... Butterbur-drug interactions have not been defined.
    NIH NLM LiverTox, Butterbur (last update 2019) · checked 2026-10-08
  4. btbr-daily-r1-08 · agency position
    Some butterbur preparations have had PA toxins removed to ensure safety. Only butterbur products that have been processed to remove PAs and are labeled or certified as PA-free should be considered for use.
    NIH NCCIH, Headaches and Complementary Health Approaches: What the Science Says · checked 2026-09-25
  5. btbr-daily-r1-14 · retention factor table
    USDA-RETN-R6:3004, VEG,GREENS,BOILD,LITTLE WATER DRAIN, group Vegetables and Vegetable Products (11): vitamin_c_mg 60%, folate_total_ug 65%, thiamin_mg 85%, potassium_mg 90%, copper_mg 95%
    USDA Table of Nutrient Retention Factors, Release 6, 2007, Retn_Code 3004 · checked 2026-09-25
  6. btbr-daily-r1-01 · systematic review of RCTs
    Two trials totalling 293 patients (60 and 233 patients) were included in this review. Both trials investigated the proprietary Petasites root extract Petadolex. ... The extract at higher dose (150 mg) showed a greater decreased frequency of migraine attacks and a greater number of responders (improvement>50%) after treatment over 3-4 months than the extract at lower dose (100 mg) and placebo.
    Phytomedicine, 2006 · checked 2026-09-25
  7. btbr-daily-r1-02 · randomized controlled trial
    Over 4 months of treatment, in the per-protocol analysis, migraine attack frequency was reduced by 48% for Petasites extract 75 mg bid (p = 0.0012 vs placebo), 36% for Petasites extract 50 mg bid (p = 0.127 vs placebo), and 26% for the placebo group. The proportion of patients with a > or =50% reduction in attack frequency after 4 months was 68% for patients in the Petasites extract 75-mg arm and 49% for the placebo arm (p < 0.05).
    Neurology, 2004 · checked 2026-09-25
  8. btbr-daily-r1-12 · practice guideline
    Based on our review, 11 prophylactic drugs received a strong recommendation for use (topiramate, propranolol, nadolol, metoprolol, amitriptyline, gabapentin, candesartan, butterbur, riboflavin, coenzyme Q10, and magnesium citrate) and 6 received a weak recommendation (divalproex sodium, flunarizine, pizotifen, venlafaxine, verapamil, and lisinopril).
    Can J Neurol Sci, 2012 · checked 2026-09-25
  9. btbr-daily-r1-07 · agency monograph
    However, recent cases of clinically apparent liver injury have been reported with use of several commercial preparations of butterbur, suspected to be due to residual pyrrolizidine alkaloid contamination.
    NIH NLM LiverTox, Butterbur · checked 2026-09-25
  10. btbr-hf2-01 · agency position
    the HMPC agreed that an acceptable intake equivalent to 1.0 µg/day for an adult can be used as the limit for oral intake of PAs. ... For HMP with PA-containing herbal substances/preparations as active ingredient, the sum of PAs from the active ingredient and possible contaminations with PAs should not exceed the given limits.
    EMA HMPC, Public statement on the use of herbal medicinal products containing toxic, unsaturated pyrrolizidine alkaloids (PAs), EMA/HMPC/893108/2011 Rev. 1, 2021 · checked 2026-10-08
  11. btbr-hf2-03 · agency position
    Beside the long known PA content in some plants traditionally used as medicines such as from the genera Symphytum, Borago, Petasites and Tussilago (CPMP 1992) ... The maximum daily dose of such PAs for internal use is set at 1 µg for a duration of maximum 6 weeks per year and 0.1 µg without any limitation in the duration.
    EMA HMPC, Public statement on the use of herbal medicinal products containing toxic, unsaturated pyrrolizidine alkaloids (PAs), EMA/HMPC/893108/2011 Rev. 1, 2021 · checked 2026-10-08
  12. btbr-hf2-04 · agency monograph
    8 of the 10 cases were reasonably convincing and described a consistent clinical phenotype of fatigue, nausea and jaundice usually arising within 2 to 12 weeks of starting the product and marked by a hepatocellular pattern of injury and moderate-to-severe jaundice. ... Two cases resulted in urgent liver transplants, while the others apparently recovered, usually within 2 to 12 weeks.
    NIH NLM LiverTox, Butterbur (last update 2019) · checked 2026-10-08
  13. btbr-daily-r1-05 · practice guideline
    Petasites (butterbur) is effective for migraine prevention and should be offered to patients with migraine to reduce the frequency and severity of migraine attacks (Level A).
    Neurology, 2012 · checked 2026-09-25
  14. btbr-daily-r1-06 · agency position
    However, butterbur contains pyrrolizidine alkaloids (PA), which are hepatotoxic. The American Academy of Neurology stopped recommending butterbur in 2015 because of serious concerns about possible liver toxicity.
    NIH NCCIH, Headaches and Complementary Health Approaches: What the Science Says · checked 2026-09-25
  15. btbr-daily-r1-11 · agency position
    A 2012 study analyzed 21 commercial butterbur products in the United States for petasins (the active ingredient in butterbur) and PAs, and found only 7 products to be safe and effective, with petasin present and no detectable PAs.
    NIH NCCIH, Headaches and Complementary Health Approaches: What the Science Says · checked 2026-09-25
  16. btbr-daily-r1-09 · agency position
    Butterbur products with PAs should not be used during pregnancy or while breastfeeding because they may cause birth defects or liver damage. Little is known about whether it’s safe to use PA-free butterbur during pregnancy or while breastfeeding.
    NIH NCCIH, Headaches and Complementary Health Approaches: What the Science Says · checked 2026-09-25
  17. btbr-daily-r1-10 · agency position
    Butterbur may cause allergic reactions in people who are sensitive to plants such as ragweed, chrysanthemums, marigolds, and daisies.
    NIH NCCIH, Headaches and Complementary Health Approaches: What the Science Says · checked 2026-09-25
  18. btbr-hf2-02 · agency position
    If children are included in the usage of certain products the daily amount of toxic, unsaturated PAs has to be adjusted to the body weight of the age group: e.g. body weight of 20 kg would lead to an acceptable daily intake of 0.5 µg toxic, unsaturated PAs per day. Sensitive groups such as pregnant and breastfeeding women are also covered by the limit calculated above.
    EMA HMPC, Public statement on the use of herbal medicinal products containing toxic, unsaturated pyrrolizidine alkaloids (PAs), EMA/HMPC/893108/2011 Rev. 1, 2021 · checked 2026-10-08
  19. btbr-daily-r1-04 · systematic review of RCTs
    Moderate evidence of effectiveness is, thus, available for a higher than the recommended dose of the proprietary Petasites root extract Petadolex in the prophylaxis of migraine. Further rigorous studies are required to confirm effectiveness and safety in long-term use before treatment with Petasites root extract can be recommended as an alternative option in the treatment schedule for the prophylaxis of migraine.
    Phytomedicine, 2006 · checked 2026-09-25

Review. Reviewed on 2026-10-08 for evidence level, dose and population context, and claims a reader could misread. Bioma Learn has no human medical reviewer at this time, and we say so rather than invent one. Methodology. This is information, not medical advice.