Herbal medicines

Definitions

There is no internationally recognized definition of a herbal medicine. The European Scientific Co-operative on Phytotherapy (ESCOP) has defined herbal medicinal products as “medicinal products containing as active ingredients only plants, parts of plants or plant materials, or combinations thereof, whether in the crude or processed state”. Definitions provided in a European Union (EU) directive on Traditional Herbal Medicinal Products [ ] are given in Table 4 , along with some other terms used in relation to herbal medicines [ ].

Usage

In the UK, retail sales of complementary medicines (licensed herbal medicines, homeopathic remedies, essential oils used in aromatherapy) were estimated to be £72 million in 1996, an increase of 36% in real terms since 1991 [ ]. This, however, was likely to be a gross underestimate, as popular products sold as food supplements, including Ginkgo biloba and garlic, were not included. According to a detailed analysis of the herbal medicines marketed in Germany and France, total sales of herbal products in those countries in 1997 were US$1.8 billion and US$1.1 billion respectively [ ]. In 1994, annual retail sales of botanical medicines in the USA were estimated to be around US$1.6 billion; in 1998, the figure was closer to US$4 billion [ ]. In the period from 2007 to 2009 the market in complementary medicines in the UK grew by 18% and was estimated to be worth about £213 million.

Two surveys of US adults carried out in 1991 and 1997/98 involved over 1500 and over 2000 individuals respectively [ , ]. The use of at least one form of complementary therapy in the 12 months preceding the survey increased significantly from 34% in 1990 to 42% in 1997. Herbal medicine was one of the therapies showing the most increase over this time: there was a statistically significant increase in self-medication with herbal medicine from 2.5% of the sample in 1990 to 13% in 1997 [ ]. Disclosure rates to physicians of complementary medicine use were below 40% in both surveys [ ]. Furthermore, 18% of prescription medicine users took prescription medicines concurrently with herbal remedies and/or high-dose vitamins. These aspects of user behavior clearly have implications for safety.

The reasons for the popularity of herbal medicine are many and diverse. It appears that complementary medicine is not usually used because of an outright rejection of conventional medicine, but more because users desire to control their own health and because they find complementary medicine to be more congruent with their own values, beliefs, and philosophical orientations toward health and life [ ]. Also, users may consult different practitioners for different reasons. An important reason for the continuing increase in use is that consumers (often motivated by the lay press) consider herbal medicines to be “natural” and assume that they are free from harms. However, this notion is dangerously misleading; adverse reactions have often been associated with the use of complementary therapies [ ]. Furthermore, complementary therapies may not only be directly harmful (for example adverse reactions to a herbal preparation), but like other medical treatments have the potential to be indirectly harmful (for example through being applied incompetently, by delaying appropriate effective treatment, or by causing needless expense) [ ].

In a study of how 13 patients defined adverse events associated with manual therapy techniques in Canada, they defined mild, moderate and major adverse events according to pain and other symptoms, taking into account their intensity, functional impact, and duration, and by ruling out alternative causes, in ways that differed from those of clinicians [ ].

In a questionnaire survey of 800 subjects living or working in suburban Kuala Lumpur, Malaysia, 610 replied; they had misconceptions regarding the safety of complementary and alternative medicines and over-the-counter medications, and usually underestimated the risks [ ]. Medication history and previous experience of adverse reactions had significant effects on their behavior.

General adverse effects and adverse reactions

In principle, herbal medicines have the potential to elicit the same types of adverse reactions as synthetic drugs; the body has no way of distinguishing between naturally occurring compounds and synthetic or semisynthetic compounds.

Herbal medicines consist of whole extracts of plant parts (for example roots, leaves) and contain numerous potentially active molecules. Synergy is normally assumed to play a part in the medicinal effects of plant extracts, and medical herbalists have always claimed that whole plant extracts have superior effects over single isolated constituents. Similarly, it is also claimed that combinations of herbs have synergistic effects. There is in vitro and/or in vivo evidence to support the occurrence of synergism between constituents in certain herbal extracts [ , ], but clinical evidence is lacking, and it is in any case uncertain how far the principle extends. Synergy is also taken to mean an attenuation of undesirable effects, another key tenet of herbalism being that the toxicity of plant extracts is less than that of a single isolated constituent. However, theoretically, plant constituents could also interact to render a herbal preparation more toxic than a single chemical constituent. Virtually no evidence is available to substantiate either hypothesis. It is also important to determine whether herbal treatments that have been shown to be as effective as conventional drugs have a better safety profile. Contrary to the belief of many herbalists, long-standing experience is by no means a reliable yardstick when it comes to judging the risk of adverse reactions [ ].

A hospital-based study from Oman suggested that 15% of all cases of self-poisoning seen in this setting are with traditional medicines [ ]. In a case series from Thailand in patients with oral squamous cell carcinoma, the use of herbal medicines before the first consultation with a health-care professional increased the risk of an advanced stage almost six-fold [ ] and survey data from the USA suggested that herb–drug interactions may be a significant problem in a sizeable proportion of patients [ ].

Several herbal medicines pose serious problems for surgical patients, for example through an increased bleeding tendency [ , ]. Vulnerable populations also include children [ ], and too few safety data are available to recommend herbal medicines during pregnancy or lactation [ ]. Several investigators have pointed out the potential of herbal medicines to harm certain organs, for example the liver [ ] or the skin [ ]. Laxatives are often based on herbal extracts, and the risks of herbal laxatives have been emphasized [ ].

Direct effects associated with herbal medicines can occur in several ways:

• hypersusceptibility reactions;

• collateral reactions;

• toxic reactions;

• drug interactions;

• impurities (contamination and adulteration);

• false authentication;

• lack of quality control.

Some of these relate to product quality. While there are some data on certain of these aspects, information on other aspects is almost entirely lacking. For example, there are case reports of interactions between conventional medicines and complementary (usually herbal) remedies [ , ], but further information is largely theoretical [ ].

Even a perfectly safe remedy (mainstream or unorthodox) can become unsafe when used incompetently. Medical competence can be defined as doing everything in the best interest of the patient according to the best available evidence. There are numerous circumstances, both in orthodox and complementary medicine, when competence is jeopardized:

• missed diagnosis;

• misdiagnosis;

• disregarding contraindications;

• preventing/delaying more effective treatments (for example misinformation about effective therapies; loss of herd immunity through a negative attitude toward immunization);

• clinical deterioration not diagnosed;

• adverse reaction not diagnosed;

• discontinuation of prescribed drugs;

• self-medication.

The attitude of consumers toward herbal medicines can also constitute a risk. When 515 users of herbal remedies were interviewed about their behavior vis a vis adverse effects of herbal versus synthetic over-the-counter drugs, a clear difference emerged. While 26% would consult their doctor for a serious adverse reaction to a synthetic medication, only 0.8% would do the same in relation to herbal remedies [ ].

The only way to minimize incompetence is by proper education and training, combined with responsible regulatory control. While training and control are self-evident features of mainstream medicine they are often not fully incorporated in complementary medicine. Thus the issue of indirect health risk is particularly pertinent to complementary medicine. Whenever complementary practitioners take full responsibility for a patient, this should be matched with full medical competence; if on the other hand, competence is not demonstrably complete, the practitioner in question should not assume full responsibility [ ].

Incidence of adverse effects and adverse reactions

Much of the information on adverse reactions associated with herbal medicines is anecdotal, and assessment and classification of causality is often not possible. Likewise, there have been few attempts to determine systematically the incidence of adverse reactions to non-orthodox therapies.

Of 1701 patients admitted to two general wards of a Hong Kong hospital, 3 (0.2%) had adverse reactions to Chinese herbal drugs; two of the three were serious [ ]. In a retrospective study of all 2695 patients admitted to a Taiwan department of medicine during 10 months 4% were admitted because of drug-related problems, and herbal remedies ranked third amongst the categories of medicines responsible [ ]. In an active surveillance adverse drug reaction reporting program conducted in a family medicine ward of the National Taiwan University Hospital, Chinese crude drugs were responsible for five hospital admissions (22% of the total) or 12% of all adverse reactions observed in the study [ ]. This is a part of the world where the herbal tradition is particularly strong; the figures may not apply elsewhere.

The incidence of contact sensitization associated with topical formulations containing plant extracts was significant when evaluated in 1032 consecutive or randomly selected patients visiting patch test clinics in The Netherlands [ ].

In a 5-year toxicological study of traditional remedies and food supplements carried out by the Medical Toxicology Unit at Guy’s and St Thomas’ Hospital, London, 1297 symptomatic enquiries by medical professionals were evaluated [ ]. Of these, an association was considered to have been confirmed, probable, or possible in 12, 35, and 738 cases respectively. Ten of the confirmed cases were related to Chinese or Indian herbal remedies. As a result of these findings, in October 1996 the UK Committee on Safety of Medicines extended its yellow card scheme for adverse drug reactions reporting to include unlicensed herbal remedies, which are marketed mostly as food supplements in the UK (the scheme had always applied to licensed herbal medicines) [ , ]. This was an important milestone in herbal pharmacovigilance.

A report from the Uppsala Monitoring Centre of the WHO has summarized all suspected adverse reactions to herbal medicaments reported from 55 countries worldwide over 20 years [ ]. A total of 8985 case reports were on record. Most originated from Germany (20%), followed by France (17%), the USA (17%), and the UK (12%). Allergic reactions were the most frequent serious adverse events and there were 21 deaths. The authors pointed out that adverse reactions to herbal medicaments constitute only about 0.5% of all adverse reactions on record.

In an audit of 100 patients admitted to a UK hospital for acute medical emergencies, 24 were taking herbal remedies (a total of 40 products); the most popular product was garlic [ ]. Twenty of these 24 patients were also taking prescription medications and 11 were taking herbal remedies that have either documented toxicity or known interactions with prescription drugs. However, the use of a herbal preparation was documented in the case notes in only one case.

In a 5-month survey on the use of phytomedicines by 1063 patients, based on a pre-structured questionnaire, in the outpatient department of an urban university general hospital in Italy, of the1044 women who completed the interview 491 (47%) reported taking at least one herbal compound [ ]. Of these 491 patients, 272 (55%) consumed only one phytomedicine, while 219 (45%) also took traditional drugs. Adverse reactions were reported by 47 (9.6%), including: gastrointestinal (due to dandelion, propolis, and fennel); cardiovascular (after liquorice, ginseng, and green tea); dermatological (after propolis, thyme, arnica, and passion-flower); and neurological (after guaraná and liquorice). Conventional medicines that were taken simultaneously and were potentially involved in adverse reactions were antihypertensive drugs, benzodiazepines, non-steroidal anti-inflammatory drugs, antibiotics, and oral contraceptives. In five cases adverse reactions were sufficiently serious to justify admission to hospital. In 29/47 cases the adverse reaction was not communicated to the doctor. The authors confirmed previous observations that consumers of herbal remedies act differently with regard to reporting an adverse reaction (serious or minor) to their physician, and that many adverse manifestations to herbal remedies are not monitored.

Perioperative events in 601 patients undergoing major elective surgery who had taken traditional Chinese herbal medicines presurgically have been studied in a teaching hospital in Hong Kong [ ]. Of these patients, 483 (80%) had taken self-prescribed herbal medicines and 47 (8%) had taken herbal medicines prescribed by practitioners of traditional Chinese medicine in the 2 weeks before surgery. The crude incidences of any combined end-points of preoperative, intraoperative, and postoperative events were 23% (19–26%), 74% (71–78%), and 63% (59–66%) respectively. Compared with non-users, patients who took traditional Chinese herbal medicines by prescription were more likely to have a preoperative event. The authors presented four case reports to highlight the effect of traditional Chinese herbal medicines by prescription on preoperative prolongation of the activated partial thromboplastin time and hypokalemia. In contrast, there was no significant association between the use of any type of traditional Chinese herbal medicines and the occurrence of either intraoperative or postoperative events. They emphasized that the use of traditional Chinese herbal medicines by prescription near the time of surgery should be discouraged, because of the increased risk of adverse events in the preoperative period.

Awareness of the need for surveillance of adverse reactions to natural health products has stimulated the implementation of a reporting system for suspected adverse reaction in Italy [ ]. From April 2002 to March 2007, 233 spontaneous reports of suspected adverse reactions to natural health products were collected. A large proportion of the suspected adverse reactions were serious: hospitalization was reported in 35% of cases; 6% reported life-threatening clinical events, and there were two fatal events. Most of the reported cases involved herbal products (66%); 21 reports were associated with 27 homeopathic preparations, most of which contained a mixture of substances; 14 reports attributed the suspected reactions to products containing propolis.

Pharmacovigilance of herbal products

Measures to improve the report standardization and detection of suspected adverse reactions to complementary and alternative medicines have been prompted in many regions. In a review of all suspected adverse reactions spontaneously reported to the Swedish Medical Products Agency between 1987 and 2006 (64 493 reports), 778 reports concerned with 967 suspected adverse reactions were related to 175 different products [ ]. The main suspected adverse reactions were urticaria (8.3%), rashes (7.4%), and contact dermatitis (5.7%). The substances that were most commonly implicated were purple coneflower ( Echinacea purpurea ) (8.1%), purple coneflower + Siberian ginseng ( Eleutherococcus senticosus ) + malabar nut ( Adhatoda vasica ) (7.3%) and ginkgo leaf ( Ginkgo biloba ) (6.7%). In 221 reports, at least one reaction was categorized as serious, the most frequent being pulmonary embolism (1.7%), mixed liver reactions (2.8%), and anaphylactic reactions (2.0%). Eleven of the serious suspected adverse reactions had a fatal outcome. Considering the limitations of spontaneous reporting of suspected adverse reactions by health-care professionals, it has been suggested that the signal detection power of spontaneous reporting systems should increase, that new monitoring systems are needed, and that consumers and CAM providers should have the ability to report adverse events during such therapy [ ].

A review in China showed that traditional Chinese medicine (TCM) had led to several adverse drug reactions that pharmacovigilance in TCM was problematic, although great efforts have been made to improve it [ ]. In order to report adverse reactions to interventions in randomized controlled trials transparently, it has been suggested that Consolidated Standards of Reporting Trials (CONSORT) for TCM should include background information on adverse reactions to each intervention, specific outcome assessments of adverse reactions, and interpretation of the occurrence of adverse reactions in a structural report [ ]. To improve reporting related to traditional Chinese medicine, recommendations have been given on constructing a title and basic principles of writing an abstract and the text of reports [ , ]. Frameworks for reporting adverse drug reactions anecdotally have also been suggested [ , ].

Indian Ayurveda and Unani medicines and Sri Lankan Siddha medicines have been monitored by a pharmacovigilance program in India since 2008 [ ]. In South India, a pharmacist-coordinated program has been initiated to improve the reporting of adverse reactions to complementary and alternative medicines [ ].

Adverse events related to medicinal plants and their derivatives (herbal medicines) have also been documented by the Brazilian Pharmacovigilance System since January 1999 [ ].

Asian herbalism

Most reports of adverse reactions to herbal remedies, of which there have been many reports [ ], relate to Chinese herbal medicines [ ]. This is a matter of growing concern, particularly because in many Western countries the popularity of Chinese herbalism is increasing. This is happening in the almost complete absence of governmental control [ ] or of systematic research into the potential hazards of Chinese herbal formulations [ ]. Chinese herbal decoctions are usually composed of a mixture of up to 20 different kinds of herbs. Although they are traditionally thought to cause fewer adverse reactions than conventional medicines, a review of the Chinese literature has shown that of 31 cases reported in 25 articles which dealt with adverse reactions, 10 were severe [ ]. The main adverse reactions were allergic reactions and gastrointestinal, nervous system, and hematological disorders. All the patients recovered after withdrawal of the decoction and symptomatic treatment.

Most serious adverse reactions to Chinese herbal remedies are associated with formulations containing aconitine, anticholinergic compounds, aristolochic acid, podophyllin, or contaminating substances [ ]. Problems with Chinese herbal formulations are intensified because of nomenclature, since common, botanical, and Chinese names exist side by side, making confusion likely.

In a German hospital specializing in Chinese herbalism, of 145 patients who had been treated within 1 year 53% reported having had at least one adverse event attributable to Chinese herbal medicines [ ]. Nausea, vomiting, and diarrhea were the most common complaints. In the same institution about 1% of 1507 consecutive patients treated with Chinese herbal mixtures had clinically relevant rises in liver enzymes [ , ]. Glycyrrhiza radix and Atractylodis macrocephalae rhizome were most consistently associated with such problems. In most of these cases there were no associated clinical signs and the abnormalities tended to normalize without specific therapy and in spite of continued treatment with the Chinese herbal mixtures.

When 1100 Australian practitioners of traditional Chinese medicine were asked to complete questionnaires about adverse reactions to Chinese herbal mixtures, they reported 860 adverse events, including 19 deaths [ ]. Each practitioner had encountered an average of 1.4 adverse events during each year of full-time practice.

A physician prospectively monitored all 1265 patients taking traditional Chinese medicines at his clinic during 33 months [ ]. Liver enzymes were measured before the start of therapy and 3 and 10 weeks later. Alanine aminotransferase activity was raised in 107 patients (8.5%) who initially had normal values; of these, about 25% reported symptoms such as abdominal discomfort, looseness of bowels, loss of appetite, or fatigue.

A retrospective analysis of all adverse events related to herbal medicines and dietary supplements reported to the California Poison Control System has given data on the risks of the adverse effects of herbal medicines [ ]. Between January 1997 and June 1998, 918 calls relating to such supplements were received. Exposures resulting in adverse reactions occurred most often at recommended doses. There were 233 adverse events, of which 29% occurred in children. The products most frequently implicated were zinc (38%), Echinacea (8%), witch hazel (6%), and chromium picolinate (6%). Most of the adverse events were not severe and required no treatment; hospitalization was required in only three cases.

Ayurvedic medicine

Ayurvedic medicine is a form of Hindu traditional medicine in the Vedic tradition. It relies greatly on the use of plant-derived medicines, to which heavy metals are often added in a supposedly detoxified state, because of their reputed therapeutic properties. However, if the detoxification process is not strictly followed during manufacture, it is possible for the resulting product to contain large amounts of heavy metals [ ] (see below).

“Legal highs”

Plants that contain hallucinogenic drugs may be marketed online as “legal highs”. In a study of 39 websites, 1308 products were found and evaluated [ ]. The products took the form of “pills” (47%), smoking materials (30%), and single plant materials/extracts (18%). Most of the products were claimed to be stimulants (42%), sedatives (32%), or hallucinogens (13%). Ingredients were not listed in 40% of products, contraindications in 82%, adverse reactions in 92%, and drug interactions in 86%. The five most common products were Salvia divinorum (salivinorin A), Kratom (mitragynine), Hawaiian baby woodrose seeds (lysergic acid amide), fly agaric (ibotenic acid, muscimol), and Genie (JWH018, CP47497).

In another study of 28 websites 119 products were evaluated. Some of the products (47%) were likened to illicit drugs, typically marijuana (48%) or ecstasy (23%) [ ]. The most common product ingredients were Ephedra alkaloids (27%), Salvia divinorum (17%), kava (10%), guaraná (10%), Acorus calamus (10%), and damiana (10%). Effect claims often involved use of the product as a hallucinogen (51%) or stimulant (39%). Adverse reactions were mentioned by 64% of the sites and drug interactions by 54%.

Reviews

Many authors have reviewed the risks of herbal medicines in general terms [ , ] and several reviews have covered specific topics, including:

• the toxicity of medicinal plants [ ];

• adverse reactions to herbal products in general [ ];

• adverse reactions in specific countries, for example the USA [ ] and Malaysia [ ];

• adverse effects on specific organs [ ], such as the cardiovascular system [ ], the liver [ , ], and the skin [ , ];

• adverse reactions to herbal medicines in vulnerable populations: elderly patients [ ], pregnant women [ ], and surgical patients [ , ];

• carcinogenicity [ ];

• adverse reactions to herbal antidepressants [ ];

• adverse reactions to Chinese herbal medicaments [ , ];

• adverse reactions to Ayurvedic medicines [ ];

• herb–drug interactions [ ];

• pharmacovigilance of herbal medicines [ ].

Cardiovascular

Respiratory

Nervous system

Sensory systems

Psychiatric

Gastrointestinal

Liver

Herbal medicines can be hepatotoxic, even those that are claimed to have hepatoprotective effects. The possible hepatotoxic effects of herbal medicines should be monitored, even though they are often regarded as “natural”, mild, and non-toxic.

Hepatotoxicity has been attributed to traditional medicines in patients with hepatitis B infection [ ]. In a pilot study to review the clinical course of drug-induced liver damage in a Singapore tertiary hospital 29 patients were identified, of whom 15 had consumed traditional Chinese medicines and four had taken antituberculosis drugs [ ].

In a 26-month prospective study at an Asian tertiary hospital to test whether drug-induced liver injury in 31 patients caused by traditional complementary and alternative medicines was related to adulterants, 23 had hepatocelluar damage, six had cholestasis, and two had a mixed pattern of injury [ ]. Chinese medicines were implicated in 17 cases, followed by Malay medicines in five. Adulterants were found in nine of the 31 traditional medicines that were available for chemical analysis. The authors concluded that drug-induced liver injury in Asia has a different pattern of causes from that in the West.