Petasites is a shrub with incredibly huge leaves, up to 3 feet in diameter, usually found in low-lying wet areas and marshes. The plant is sometimes called butterbur, perhaps because the large, soft leaves were used to wrap butter when trying to keep it cool. Other colorful names for the plant include butterfly dock, butter-dock, bog rhubarb, flapperdock, and umbrella leaves. By whatever name, the plant is quite toxic. Butterbur was used in the Middle Ages to try to treat the plague. More recently, there has been some evidence suggesting that it might help in the treatment or prevention of migraines, asthma, ulcers, and now hay fever.

I’ve seen several headlines promoting the use of butterbur tablets (also called petasites tablets) as a gentle, natural remedy for those who want to treat hay fever without antihistamines. There are two important questions: Do they work? And are they safe? I’m excited that people have been looking into butterbur. I’m always glad to see alternative remedies evaluated. These results are particularly important.

One study, conducted by the Petasites Study Group in Switzerland, took 125 adults with hay fever and randomly assigned then to take petasites (butterbur) 8 mg four times a day or cetrizine (Zyrtec) 10 mg once a day. The pills were disguised and dummy pills were included so that participants (and doctors) did not know at the time which treatment they were receiving. The results were published in the January 19th issue of the British Medical Journal. I was very hopeful as I began to read the study.

After 2 weeks, investigators found no difference in symptoms between the two groups - both treatments were reported to have improved hay fever allergies equally. However, I am quite skeptical of these industry-funded results. The purpose of ‘blinding’ a study is to eliminate bias. But they designed this study to show that these two treatments were the same - so interviewers would be tempted to score patients similarly, and the blinding does nothing to eliminate bias with this study design. Also - when scoring results, they did not look at patients’ reports of any allergy symptoms improved (sneezing, watery eyes, runny nose - anything!), just overall quality of life. Not a very sensitive way to find a difference. At best, this study had only 80% statistical power (a 20% chance that butterbur did not help at all), but the built-in bias and the insensitive tests used take away even that measure of confidence. I’m sad to say that this study did nothing to convince me that butterbur actually helps allergies (though it certainly may) - much less that it is as effective as cetrizine. But it is the side effect story that is of the gravest concern to me.

The total number of side effects observed in those two weeks was similar (16.4% of the petasites group and 17.2% of the cetrizine group). Those taking cetrizine most commonly reported drowsiness (even though it is supposed to be a non-sedating antihistamine). This side effect was considered significant by the investigators. Those taking petasites reported a broader range of side effects, including itching skin, itchy eyes, wheezing, diarrhea, GI upset, headache, and fatigue. The authors dismissed all of these because no single side effect occurred in more than two people. To me they suggest a pattern of potential allergic reactions, which makes sense in that butterbur is a relative of ragweed. One patient had elevated liver enzymes in the blood (enzymes that indicate the destruction of healthy liver cells). This was also dismissed. I wonder how many patients would have had elevated liver enzymes if the test went on for 4 weeks?

Natural petasites (butterbur) contains highly toxic alkaloids that are known to seriously damage the liver and to cause cancer. They have also been associated with potentially fatal blood clots and with chronic scarring of the lung. Even small traces of these alkaloids are considered quite dangerous. Some commercial preparations are available where these alkaloids have been removed. They may work as well as the natural product, but we do not know. Nevertheless, butterbur in any form should not be used by women who are pregnant or who are nursing - all the more so in countries like the United States where herbs and supplements are not well-regulated.

Of greater concern to me: does the very ingredient supposedly responsible for improving many of the symptoms cause long-term problems? The active ingredient S-petasin would be expected to have a possible impact on the reproductive system. The one study looking at this carefully, published in the September 30, 2000 Clinical Journal of Physiology, found that s-petasin caused a dramatic impairment of testicular function when given to rats at tiny trace doses. Baseline testosterone production fell.

I would certainly not recommend giving butterbur to children before puberty, unless more evidence becomes available. The evidence in the BMJ article was crafted to sound convincing, but did not prove that butterbur is either safe or effective for treating allergies (methodological flaws have long plagued butterbur research). By conducting the study in a backwards way, the investigators took the power out of their randomized, placebo-controlled, double blind study. Designing a study to find the truth yields valuable information. Designing a study to promote a product can leave you with nothing but questions. Still, the butterbur story teaches us how possible it is to manipulate statistical evidence and to convince the popular press. It also illustrates that “herbal” and “natural” do not necessarily mean “safe” or “gentle”.

• ‹ Yes, Asthma - But What Type?
• Big News: Reflux and Ear Infections? !!! ›