Academia and Clinic
The Risk–Beneﬁt Proﬁle of Commonly Used Herbal Therapies: Ginkgo, St. John’s Wort, Ginseng, Echinacea, Saw Palmetto, and Kava
Edzard Ernst, MD, PhD, FRCP(Edin)
Because use of herbal remedies is increasing, a risk– benefit profile of commonly used herbs isneeded. This article provides a clinically oriented overview of the efficacy and safety of ginkgo, St. John’s wort, ginseng, echinacea, saw palmetto, and kava. Wherever possible, assessments are based on systematic reviews of randomized clinical trials. Encouraging data support the efficacy of some of these popular herbal medicinal products, and the potential for doing good seems greater than thatfor doing harm. The published evidence suggests that ginkgo is of questionable use for memory loss and tinnitus but has some effect on dementia and intermittent claudication. St. John’s wort is efficacious for mild to moderate depression, but serious concerns exist about its interac-
tions with several conventional drugs. Well-conducted clinical trials do not support the efficacy of ginseng totreat any condition. Echinacea may be helpful in the treatment or prevention of upper respiratory tract infections, but trial data are not fully convincing. Saw palmetto has been shown in short-term trials to be efficacious in reducing the symptoms of benign prostatic hyperplasia. Kava is an efficacious short-term treatment for anxiety. None of these herbal medicines is free of adverse effects.Because the evidence is incomplete, risk– benefit assessments are not completely reliable, and much knowledge is still lacking.
Ann Intern Med. 2002;136:42-53. For the author affiliation and current address, see end of text. www.annals.org
hysicians’ need for reliable information on herbal medicinal products is considerable. In the United States, the popularity of complementary and alternativemedicine is growing at a remarkable and perhaps disquieting speed (1). Herbal medicine has grown faster than any other “alternative” treatment method in the United States (1, 2). The seven top-selling herbal medicinal products in the United States are Ginkgo biloba (total 1998 retail sales of $151 million), St. John’s wort ($140 million), ginseng ($96 million), garlic ($84 million), echinacea ($70million), saw palmetto ($32 million), and kava ($17 million) (3). (Garlic is not discussed in this article; the reader is referred to two recent meta-analyses [4, 5].) Physicians regularly see patients who self-prescribe herbal medicinal products but do not volunteer this information (1). Many herbal medicines have adverse effects, and most can interact with puriﬁed prescription drugs (6). Acomplete medical history should therefore include speciﬁc questions about the use of herbal medicinal products (7), and physicians must acquire sufﬁcient knowledge in this area to advise their patients responsibly. This article provides risk– beneﬁt proﬁles for the most commonly used herbal medicines. These proﬁles are based on systematic reviews where possible.
PubMed), EMBASE, CISCOM, and AMEDdatabases and the Cochrane Library were searched from their inception to October 2000. In addition, nine experts in the ﬁeld were asked for further details on systematic reviews, and the author’s personal ﬁles were searched for further articles. All systematic reviews of randomized clinical trials in humans were included. They were read in full, and key data were extracted according to criteriathat appear as column headers in Table 1. The quality of the reviews was estimated according to the guidelines of the Quality of Reporting of Meta-analyses statement (23). In essence, the following questions were asked: Are the objectives of the review clearly stated? Are the data sources stated? Are the search methods explained? Are inclusion and exclusion criteria stated? Was the validity of...