brand logo

Patients who self-medicate with herbs for preventive and therapeutic purposes may assume that these products are safe because they are “natural,” but some products cause adverse effects or have the potential to interact with prescription medications. The United States lacks a regulatory system for herbal products. Although only limited research on herbs has been published, St John's wort shows promise as a treatment for depression. Ginkgo biloba extract is possibly effective for cerebrovascular insufficiency and dementia. Feverfew is used extensively in Canada for migraine prophylaxis but needs more rigorous study. Ephedrine has been regulated by many states because its misuse has been associated with several deaths. Echinacea is being tried as an agent for immune stimulation, and garlic is under study for cholesterol-lowering properties, but both require more study. Physicians should educate themselves and their patients about the efficacy and adverse interactions of herbal agents and the limitations of our present knowledge of them.

A significant minority of American adults self-medicate with herbal preparations for preventive or therapeutic purposes. Sales of herbal products in the United States have increased sharply in recent years, according to industry reports. An estimated $2 billion was spent in health food stores in 1996 for herbs in bulk, as well as capsules, tablets, extracts and teas.1 A 1995 telephone survey of 136 customers who purchased supplements at two health food stores in Milwaukee during a 15-day period found that the respondents were taking a total of 805 supplements, an average of 5.9 supplements per person. Of this group, 85 percent reportedly had a regular physician.2

Patients may believe that herbal products are inherently safe simply because they are “natural.” Yet herbs contain hundreds of components, some of which can cause ill effects directly, while others can interact adversely with pharmaceutical agents. Such issues are not addressed in a systematic way in this country because the United States lacks a regulatory system for herbal products that are marketed as “foods” or “dietary supplements.” They are not regulated under federal drug laws; safety and effectiveness do not have to be demonstrated before these products are marketed. No legal standards are applied to their harvesting, processing or packaging, so the possibilities of poor quality, adulteration, contamination and varying strengths must be kept in mind when evaluating them.

Herbs are available in a myriad of forms and dosages, but some standardized products are beginning to appear. An herbal extract is considered standardized when a guaranteed level of a certain constituent or group of constituents from the herb is present in the final product. This level is usually expressed as a percentage of the weight of the extract. For example, standardized extracts of St. John's wort use one of the plant's constituents, hypericin, as the reference, with the concentration ranging from 0.13 to 0.3 percent.

This article discusses some of the popular herbs that our patients may be taking and reviews what is in the literature about their effectiveness, dosages, side effects, toxicities and possible drug interactions (Table 1). While some research studies have been published, most of them are in languages other than English.

EchinaceaImmune stimulationNo well-controlled studies define dosingHuman and animal studies: increases phagocytosis, lymphocytic activity, cellular respiration and activity against tumor cells; no bactericidal and bacteriostatic properties; probably should not be taken for more than 8 successive weeks; no documented side effects, drug interactions or toxicity
Traditional dosing:
Maximum daily dosage in adults: 6 to 9 mL of expressed fresh juice, or 1.5 to 7.5 mL of tincture or 2 to 5 g of dried root
FeverfewMigraine headache prophylaxis125 mg daily of standardized extract (standardized to 0.2% Parthenolide)[ corrected] Inhibits prostaglandin production; serotonin antagonist; not recommended in pregnancy; induces menstruation; caution with anticoagulants; no documented problems
GarlicAtherosclerosisNo well-controlled studes define dosingInhibits platelet function; increases serum insulin levels and improves glycogen storage; concerns about increased postoperative bleeding
Traditional dosing:
4 to 12 mg of allicin, or 0.4 to 12 g of dried powder or 2 to 5 g of fresh bulb daily
Ginkgo bilobaMild to moderatecerebral insufficiency; favorable for PMS and vertigo120 to 160 mg daily of standardized extract (standardized to 24% flavone glycosides and 6% terpene lactones)Complaints of transient headache; no documented severe side effects (one case report of spontaneous subdural hematoma in a healthy adult); caution with concurrent use of aspirin and NSAIDs
Dementia120 mg daily of standardized extract, as above
St. John's wortMild to moderate depression300 mg three times daily or 450 mg twice daily of standardized extract (standardized to 0.13 to 0.30% hypericin)Inhibits uptake of neurotransmitters serotonin, norepinephrine and dopamine, and binds to GABA receptors in vitro; MAOI activity in vitro, not in vivo; no major side effects—minor complaints of dry mouth, dizziness, constipation and confusion; photosensitivity with high dosages (600 mg three times daily of 0.24 to 0.32% hypericin); uterotonic in animals in pregnancy; no known drug interactions, but because of neurotransmitter activity, exercise caution if used with other antidepressants
Maximum in studies: 1,000 mg daily

St. John's Wort

St. John's wort (Hypericum perforatum) has been used as a folk medicine for numerous indications, including depression, bladder problems and dermatologic conditions. It is licensed in Germany for use in the treatment of anxiety, sleep disorders and depression.3 Its mechanism of action in vivo is unknown, but in vitro, it inhibits the uptake of the neurotransmitters serotonin, norepinephrine and dopamine and binds to γ-aminobutyric acid (GABA) receptors. Whether it crosses the blood-brain barrier is unknown.4 Continued research is needed to identify the constituents most responsible for the agent's activity so that preparations can be optimally standardized.

Most of the research on St. John's wort has taken place in Germany, where reports of more than 37 randomized trials have been published. Two meta-analyses have been published, one in a botanical journal5 and one in a medical journal.3 Both analyses led to the conclusion that St. John's wort is significantly superior to placebo and comparably effective to standard antidepressants (maprotiline, imipramine and amitriptyline) in the treatment of mild to moderately severe depression and produces fewer side effects. However, researchers concluded that although the data are promising, there is not yet sufficient evidence for St. John's wort to be accepted as an effective antidepressant preparation. No studies published to date compare St. John's wort with selective serotonin reuptake inhibitors (SSRIs), but the U.S. National Institutes of Health has recently funded such a study. Dose standardization, studies in severely depressed patients and long-term studies to assess the risk of relapse and the emergence of late side effects are also needed. St. John's wort is in the early stages of U.S. clinical trials as an antiviral agent.

The daily dosage used in clinical trials for depression has varied from 300 to 1,000 mg of the standardized extract in tablet or capsule form. The standardized extract has ranged from 0.13 to 0.30 percent of hypericin. (St. John's wort is also available as a tea or a tincture.)

Study subjects have complained of dry mouth, dizziness, confusion, constipation and other gastrointestinal symptoms.4 Photosensitivity has been demonstrated in a controlled clinical trial with high dosages of hypericin.6 There have been some reports of monoamine oxidase-inhibiting activity in vitro but not in vivo.6 Studies in animals have shown slight uterotonic effects, raising concern about use in pregnancy.6 No drug interactions have been documented in humans, but St. John's wort probably should not be used in combination with other antidepressants because of its potential neurotransmitter activity.7,8

Ginkgo Biloba

Ginkgo biloba health products are extracted from the world's oldest living tree. The seeds and leaves are used in traditional Chinese medicines. A concentrated standardized extract of Ginkgo biloba (GBE), with 24 percent flavone glycosides and 6 percent terpene lactones, has been studied to assess its value in increasing peripheral blood flow and dilating capillary vessels in patients with intermittent claudication and cerebrovascular insufficiency. A review of 40 European controlled trials of GBE, eight of them judged to be of good quality, concluded that GBE is effective for mild to moderate symptoms of cerebral insufficiency and is promising for use in patients with intermittent claudication.9 GBE has compared favorably with placebo in the treatment of premenstrual syndrome and vertigo.10 Several studies have examined GBE for antioxidant and cholesterol-lowering properties.10 One study11 showed that GBE stabilized or modestly improved the cognitive performance and social functioning of demented patients for the study period of six to 12 months.

The dosage of GBE used in clinical trials for cerebrovascular insufficiency was 120 to 160 mg daily; treatment for four to six weeks was required before positive effects were noticed.9 The dosage given for treatment of dementia was 120 mg daily in one study,11 but dosages as high as 240 mg a day have been reported in German studies.12

Transient headache has been documented as a side effect.9 A case report13 describes spontaneous bilateral subdural hematoma in a 33-year-old healthy woman who presented with headache and no history of head trauma or concurrent use of aspirin or nonsteroidal anti-inflammatory drugs (NSAIDs). She had taken 120 mg of GBE per day for two years. The use of GBE has not been associated with other severe side effects to date. The whole plant, seeds and pulp are occasionally associated with severe allergic reactions.10

The only published report of drug interaction with GBE involved a 70-year-old man who had been taking aspirin for three years following coronary artery bypass surgery. He developed a hyphema one week after he began taking GBE.14 Patients taking aspirin should exercise caution when adding GBE therapy.14,15


Feverfew leaf has a long history as a folk medicine for fever, headache and arthritis. A standardized extract is available. Canada's Health Protection Branch allows the marketing of feverfew as a nonprescription drug for the prevention of migraine headache.16 Feverfew appears to be an inhibitor of prostaglandin synthesis and a serotonin antagonist.17 In a large randomized, double-blind, placebo-controlled crossover trial,18 patients treated with feverfew experienced a significant decrease in headache incidence. In another study,19 patients who were switched to placebo after taking feverfew developed headaches, sleep disturbances, myalgias and arthralgias, a condition referred to as “post-feverfew syndrome.” Feverfew has shown no apparent benefit when compared with placebo in the treatment of rheumatoid arthritis.20

The daily dosage in studies for migraine headache prevention was 125 mg of a standardized extract containing no less than 0.2 percent of parthenolide. Feverfew has a short shelf life.16 No studies of long-term use or chronic toxicity have been published. Interactions with anticoagulants have not been documented but may be clinically important because aqueous extracts of feverfew show in vitro inhibition of platelet aggregation.17 Feverfew should not be used by pregnant women, because menstruation may be induced.17


Ephedrine, an isolated active constituent of the herb ma huang (ephedra), is marketed as an herbal product. Ma huang has been used in China for thousands of years as a treatment for respiratory conditions. Ephedrine has been sold over the counter in the United States since the 1930s for the treatment of asthma, in products such as Primatene. In the past 20 years, it has been promoted in health food stores as a “natural” herbal stimulant and weight loss product. Side effects of ephedrine include insomnia, motor disturbances, high blood pressure, glaucoma, impaired cerebral circulation, urinary disturbances and hyperglycemia.7,8 Reports of widespread misuse and several deaths related to ephedrine-containing products have led some states to impose a ban on their sale, except by pharmacists.21 In June, 1997, the U.S. Food and Drug Administration (FDA) issued proposed regulations that would limit the amount of ephedrine in dietary supplements.22

A combination of ephedrine and St. John's wort has been marketed as “herbal fenphen,” a nonprescription approach to weight loss, but no published scientific studies support its use. Pseudoephedrine, another active constituent of ma huang, has little adverse cardiac effect and is sold as a nasal decongestant in Sudafed and other such products.


Echinacea is used extensively in Germany and elsewhere to promote wound healing and stimulate the immune system. Numerous studies of its effects have been reported, but few are of good quality.23 Both human and animal studies show increases in phagocytosis, lymphocyte activity, cellular respiratory activity and activity against tumor cells, but direct bactericidal or bacteriostatic properties have not been demonstrated.23 In a double-blind German study, flu-like symptoms were significantly reduced in persons who took echinacea compared with those who took placebo. The benefit was evident within three to four days.24 In another double-blind study,23 patients taking echinacea had 36 percent fewer respiratory infections, a shorter duration of illness and less severe symptoms compared with subjects taking placebo. Researchers at Bastyr University, a naturopathic institution in Bothwell, Washington, are studying the effect of echinacea on the frequency and severity of respiratory tract infections.

A German Commission E monograph, a consensus statement of expert opinions, recommends that use of echinacea should not exceed a period of eight successive weeks and that the agent should not be prescribed for progressive systemic diseases such as acquired immunodeficiency syndrome, tuberculosis, collagen vascular diseases and multiple sclerosis.25

Echinacea is administered intravenously, orally and topically in Germany. No well-controlled studies have evaluated the appropriate formulation or the dosages that are available on the market. Herbal publications typically state that the maximum adult daily dosage is 6 to 9 mL of expressed, fresh juice, 1.5 to 7.5 mL of tincture (preferred because not all the constituents are water soluble) or 2 to 5 g of dried root.7 No reports on side effects, drug interactions or toxicity have been published.


Garlic has been used in a variety of ways in folk medicine for centuries. It has recently attracted attention for its reputed ability to slow the process of atherosclerosis. The scientific literature is contradictory. Two sets of reviewers concluded that no evidence existed to recommend it as an effective lipid-lowering agent.26,27 Results of a meta-analysis28 led to the conclusion that although the quality of the controlled trials was limited, the equivalent of one-half clove of fresh garlic daily could significantly reduce total cholesterol levels. Garlic has also been studied for possible antineoplastic activity.29 Other studies suggest a mild, short-term antihypertensive effect.30 Garlic has also been shown to increase serum insulin levels and improve glycogen storage.29 Garlic oil inhibits platelet function.28 No published studies have evaluated the effect of garlic and its extracts in diabetic patients or in patients treated with anticoagulants.29

The manner of consumption is important because stomach acidity inactivates allinase, the enzyme-releasing allicin considered by most authorities to be the primary active agent and a precursor to others. Allicin is absorbed quickly in the mouth. Thus, for maximum effectiveness, garlic must be well masticated before swallowing or ingested in the form of enteric-coated tablets containing carefully dried garlic powder that will be absorbed rapidly in the small intestine. Other formulations, such as garlic oil, are less reliably absorbed.31,32

No well-controlled studies define proper dosing. Garlic is used extensively for culinary purposes with no known ill effects, but the safety of long-term use of concentrated extracts is unclear.29 Some case reports indicate that nondietary or excessive intake of garlic may increase the risk of postoperative bleeding.33

Herbal Product Safety

The FDA becomes involved in safety issues for herbal products only after they are on the market and complaints are filed. Prescription and over-the-counter drugs are different; they go through a premarket approval process for safety.16 The safety issues that are beginning to be addressed are good manufacturing practices, labeling and postmarketing surveillance. Some herbal remedies have been found to contain lead, arsenic and other heavy metals, as well as pharmaceutical agents such as steroids and benzodiazepines.33 In February 1997, the FDA published proposed regulations for good manufacturing practices for dietary supplements, a category that contains herbal products.34

Under current law, only substances that are legally considered to be drugs can have labeling on the product package that includes approved uses, dosages, possible side effects, toxicity and contraindications. Since herbs are not classified as drugs, the package labels of herbal products give little guidance to the consumer. The report of the federal Commission on Dietary Supplement Labels, published in 1997,35 recommends further research and consideration of other regulatory options, including a system of approval for herbal products when requested by the industry. The report gives no indication as to when appropriate labeling for self-medicating with herbal products can be expected.35

Postmarketing surveillance for adverse reactions to herbal supplements has been limited. The Commission on Dietary Supplement Labels has recommended that the herb industry and the FDA work together to improve the agency's tracking system for adverse effects.35 Physicians are encouraged to report individual cases of apparent drug and herb interactions or suspected herb side effects to FDA MED WATCH at this telephone number: 800-FDA-1088.

Concern has been expressed that the uncontrolled promotion of herbal products may present hazards to persons who consume them. Some readers may recall events relating to laetrile, a component of apricot pits that was promoted as an alternative cancer therapy. Laetrile was eventually proved to be toxic (it contains cyanide) and was shown to be ineffective in the treatment of cancer. Herbs can enhance or block the effects of conventional drugs. Examples include the ability of ginkgo to potentiate aspirin or that of licorice to counteract antihypertensive medications.36

Learning About Herbal Agents

Physicians are encouraged to become acquainted with the herbal products their patients are consuming so they can give guidance about how patients' choices may affect their health or current therapeutic regimens. Reviewing the current medical literature will provide limited information on herbs. Other useful resources are listed in Table 2.

Review of Natural Products*
 Facts and Comparisons, 111 West Port Plaza,
 Suite 400, St. Louis, MO 63146-3098;
 Telephone: 314-878-2515
Herbal medicine chart
 Document no. 1390901
Pharmacist's Letter and Prescriber's Letter
 Telephone: 209-472-2240; fax: 209-472-2249
German Commission E monographs
 American Botanical Council,§
 P.O. Box 144345, Austin, Texas 78714-4345;
 Telephone: 512-926-4900
 Web site:
Schulz V, Hansel R, Tyler VE. Rational phytotherapy: a physician's guide to herbal medicine. 3d ed. New York: Springer, 1998.
McGuffin M, Hobbs C, Upton R, Goldberg A, eds. American Herbal Products Association's botanical safety handbook. Boca Raton, Fla.: CRC Press, 1997.

The German monographs mentioned in Table 2 were published between 1978 and 1994 by a division of the German Federal Health Authority, which is equivalent to the FDA. The monographs contain therapeutic information on herbs, including pharmacologic properties, pharmacokinetics, toxicology, uses and contraindications. Their content is based on traditional information about herbs as well as laboratory data, clinical studies, testimonials and physician input. No plan to update these monographs has been announced, although they are to be published in book form (and in English) by the American Botanical Council.

How to Talk with Patients

An effective strategy for talking with patients about herbal products must take into consideration the fact that these products are widely marketed, and many patients are taking them. Their reasons for doing so vary and include belief in a product's efficacy based on advertising, advice from friends or personal experience. Other reasons include dissatisfaction with the conventional health care system and a desire to be in control of their own life and health.

While there is some objective evidence for the value of some herbal products in some conditions, this evidence is typically tentative and incomplete. Information about toxicity and adverse interactions is similarly fragmentary.

Because the subject of medicinal herbs is complex, achieving scientific clarity and sorting through the advertising hyperbole will be a slow and arduous process.

Physicians often talk with patients about controversial subjects, so discussing the merits and drawbacks of herbs does not require new skills. Inquiring about the use of herbs, vitamins and other remedies should be a part of normal history taking, along with questions about smoking, alcohol use, exercise and other health promotion topics. Good communication requires openness to patient choices relating to unconventional remedies as the physician works toward shared decision making and “relationship-centered” care.37 Withholding judgment on the value of an unfamiliar remedy may be a wise course. As experts in conventional medicine and pharmaceuticals, physicians should discuss with each patient the treatments that are known to be effective for the condition. Patients should receive information about the potential for drug and herbal interactions. Simply telling patients to stop using a product if no clear risk is known can be harmful to the physician/patient relationship. Continuing the relationship while monitoring the patient's progress with an herbal product provides a mutual learning opportunity.

Continue Reading

More in AFP

Copyright © 1998 by the American Academy of Family Physicians.

This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP.  See permissions for copyright questions and/or permission requests.