Herbal Remedies, Dietary Supplements, and Seizures


  • Alok Tyagi,

    1. Department of Neurology and Clinical Neurological Sciences, Beaumont Hospital and Royal College of Surgeons in Ireland, Dublin, Ireland
    Search for more papers by this author
  • Norman Delanty

    1. Department of Neurology and Clinical Neurological Sciences, Beaumont Hospital and Royal College of Surgeons in Ireland, Dublin, Ireland
    Search for more papers by this author

Address correspondence and reprint requests to Dr. A. Tyagi at Department of Clinical Neurophysiology, Royal Infirmary, North Staffordshire Hospital, Stoke-on-Trent, ST4 7LN, U.K. E-mail: alokrajani@yahoo.co.uk


Summary: The use of herbal remedies and dietary supplements is widespread throughout the world, and use may be increasing. These are taken for a wide range of perceived benefits, such as energy and memory enhancement and treatment of specific conditions. Individuals with and without epilepsy may use these substances and may not inform their treating physician unless specifically asked. Inquiry about herbal medicine and dietary supplement intake should now be part of routine clinical history taking. Anecdotal accounts suggest that some herbal substances may have anticonvulsant effect, but randomised double-blind controlled trails are lacking. Alternatively many herbals and dietary supplements may predispose to seizures in individuals without epilepsy and worsen seizure control in those with epilepsy. In this article, we review the potential anticonvulsant and proconvulsant effects of herbal remedies and dietary supplements and discuss the potential interaction between these herbal substances and conventional anticonvulsant medications.


The use of herbal medication and dietary supplements has increased dramatically in recent times. A significant number of these remedies are used for treating patients with neurologic or psychiatric complaints. There are potential benefits in the use of these medications for some conditions (e.g., St. John's wort for treating depression). However, most herbal and complementary medicines are used without specific evidence of benefit from controlled clinical trials. Some of these products may be anticonvulsant and thus of possible benefit in patients with epilepsy, whereas others may predispose to seizures in those with and without epilepsy. Conversely, side effects from herbs and dietary supplements may largely go unreported. Seizures have been known to occur in patients taking complementary therapies when no other cause has been found. To address issues pertaining to herbal medications and their proconvulsant and anticonvulsant properties, we conducted a literature search by using the following key words: herbs, herbal medications, herbal treatments, alternative therapies, traditional medicine, seizures, epilepsy, proconvulsant, anticonvulsant, fits, and convulsions.


Herbs are plants that are used for their medicinal properties. Plants are the basis of ≥30% of modern drugs (e.g., ergot from the fungus Claviceps purpurea, cardiac glycosides from foxglove (Digitalis purpurea), atropine from deadly nightshade (Atropa belladonna) (1). The earliest evidence of human use of plants for healing dates back to the Neanderthal period. During the Dark Ages, botanical gardens were created to grow medicinal plants for medical schools. Herbal medicine practice flourished until the 17th century, when more “scientific” pharmacologic remedies began to be favoured. During the 19th century and the first half of the last century, as scientific methods became widely accepted, the practice of botanical healing all but disappeared from the Western hemisphere. However, over the last 50 years, the use of herbal medications has increased dramatically again in Europe and North America. In recognition of the increasing use of herbal medicines and nontraditional remedies, the Office of Alternative Medicine was established at the National Institutes of Health in 1991 to evaluate promising alternative treatments for all disorders systematically. The World Health Organisation had, in 1974, encouraged developing countries to use traditional plant medicines to fulfil a need unmet by modern systems (2).

In 1975 Napralert was created at the University of Chicago, Illinois. This is currently acknowledged to be the most authoritative source of data pertaining to empirically validated studies on medicinal agents derived from natural products. This programme has continued to conduct systematic annual reviews of the world's scientific literature on pharmacologic activities and medicinal applications of naturally occurring substances. At present Napralert indexes ∼150,000 citations and abstracts corresponding to 19 major pharmacologic activities. A total of 5,680 individual citations falling under 114 specific central nervous system (CNS) activity subcodes in the data set on CNS effects have been identified in Napralert. Cumulative citations coded in Napralert as relating to antiseizure activity form the largest subgroup with 591 citations (3).

According to an estimate, 80% of the world's population relies primarily on traditional medicines for their health care needs (2). These medications may be taken in isolation or with allopathic medication. In developing countries, their use may be related to the unavailability of modern health care–delivery systems. Traditional medicine has deep roots in society in some countries: Ayurveda in India, Kampo medicines in Japan, and Chinese herbal medications (CHMs) in China. Estimates of the use of herbal medicine use in the United States differ, with studies concluding that between 3 and 93% of the population use herbs (2). An estimate of U.S. medicinal herb sales was $1.2 billion in 1996 (2). CHM has become increasingly popular, and in 1994, 600 clinics offered traditional Chinese medicine in the U.K. (1). Almost 70% of “Western” doctors in Japan prescribe kampo drugs regularly (2). Drug companies in India market herbal medications, which are widely used by allopathic practitioners.


Most studies report antiseizure activity of a nonallopathic preparation with animal models of epilepsy (Table 1)(4–34). Several studies in humans are discussed later. None of these studies has been properly randomised, blinded, or controlled.

Table 1.  Herbal remedies as anticonvulsants (animal studies)
Name of drug/herbStudiesSeizures-reductionAction
  1. PTZ, pentylenetetrazol; GABA, γ-aminobutyric acid; NMDA, N-methyl-d-aspartate; MES, maximal electroshock.

Elaecarpus sphaericus fruits (4)AnimalYes 
Aconitum alkaloid; 1-benzoylnapelline (5)Rat hippocampusYes 
Chotoko (Uncariae Uncis cam Ramlus) (6)MiceYes (glutamate induced) 
Volatile oil of Tetrapleura tetraptera fruits (7)MiceYes (leptazol induced) 
Arecaidine and guvacine from nut of Arecha
CatYes; (bicuculine induced)GABA
Nigerian herbs (9)MiceYes [PTZ and electroshock induced] 
Leaf essential oil of Psidium guyanesis (10)MiceYes (PTZ, picrotoxin and strychnine induced) 
Shitei-to (mixture of shitei, calyx of Diospyros kaki. L, Shokyo rhizome of Zingiber officinale and Choji, flower bud of Syzygium aromaticum (11)MiceYes (PTZ, bicuculine, and strychnine induced) 
Uncaria rhynchophylla and Gastrodia elata Bl. (12)RatYes (kainic acid induced) 
Beta-eudesmol present in Atractylode lancia and Mognolia Officinalis (13)Mice and ratYes (MES induced) compared with phenytoinNa chanel
Brahmighritham (14)RatYes (PTZ induced) compared to diazepam 
Saiko-ka-ryukotsu-borei-to (Kampo) (15)RatYes (reduced theophylline-induced seizures) 
Fruit essential oil of Pimpinella anisum (16)MiceYes (PTZ and MES induced) 
Essential oil of Eugenia caryophyllata (17)MiceYes (PTZ and MES induced) 
Chrysine (in Passiflora coerulea L. (18)MiceYes (PTZ induced) 
Qingyanshen (19)MiceYes (kainic acid induced) compared with phenytoin 
Methanol extracts of Viscum sapense L.f (20)MiceYes (bicuculine and PTZ induced) 
Brahmi Rasayan (21)MiceYes (chemoconvulsion and MES induced)GABA
Alkaloids of Aconitum sp. (22)RatYes (bicuculine induced) 
West African black pepper (Piper guineense) (23)MiceYes (audiogenic, NMDA and MES induced) 
BR-16A (Mentat) (24)MiceYes (PTZ induced)GABA-A
Shosaiko-to-go-keishika-shyakuyaku-to (Kampo medicine, mix of nine herbs) TJ-960 (25)MiceYes (PTZ induced) 
Root and stem extracts of Calliandra portoricensis (26)MiceYes (PTZ and MES induced) 
Essential oil of A. suaveolens (γ-decanolactone) (27)MiceYes 
Magnolia grandiflora L. seeds (28)RatYes (MES induced) 
Casimiroa edulis seed extracts (29)RatYes (MES and PTZ induced) compared with phenytoin and phenobarbital 
Inhaled lavender oil vapor (30)RatYes (MES, PTZ, and nicotitne induced)GABA
Siotone granules (Ashwagandha) (31)MiceYes (PTZ, MES, and strychnine induced)GABA
Rosmarinus officinalis (32)MiceYes (picrotoxin)GABA
Maprounea Africana (33)MiceYes (PTZ induced) 
Skullcap (Scutellaria lateriflora) (34) Anticonvulsant 

Chinese herbal medicines

The first known document on epilepsy in China appeared in The Yellow Emperor's classic of internal medicine, Huang Di Nei Ching, written by a group of physicians ∼770–221 B.C. The first classification of epilepsy, probably by Cao Yuan Fang in A.D. 610, listed five types of epilepsy (35). According to the old theory of Chinese medicine, the causes of epilepsy were thought to be “an abnormality of the liver.” The principle of treatment more commonly used is to subdue the “endogenous wind,” with agents such as “scorpion,”“centipede,”“antelope's horn,”“Bombyx batryticatus,”“rhizoma gastrodiar,” and Uncaria rhynchophylla. The tablet “Kang Jian Ling” is made from pepper, and “718” from Bombyx batryticatus. Their anticonvulsant effects range from 60 to 70% without side effects. In an open-label study in 32 patients with generalised major motor seizures, “Qingyangsen” (roots) was used as adjunctive treatment. Despite treatment with standard anticonvulsants (AEDs), each patient had at least four seizures a month before taking Qingyangsen. After a course of 2 to 9 months, seizures were controlled in nine patients, none of whom had side effects. The authors concluded that Chinese herbs can be used as adjuncts to improve the effects of AEDs (36).

In another open-label study, treatment with a herbal AED was compared with phenobarbitone (PB) in patients with epilepsy (37). One hundred patients with epilepsy were treated with a medication containing 13 different herbs. Forty patients acted as controls and were given PB in doses of 3 to 6 mg/kg/day. The duration of the study was 8 months. The herbal medication appeared to have an efficacy similar to that of PB, but appeared to have fewer side effects (37). Treatment with another herbal medication, Zhenxianling, was carried out in an open-label study in 239 patients with epilepsy treated for 6 months to 2 years (38). All patients had been treated previously with the standard AEDs, phenytoin (PHT), carbamazepine (CBZ), and valproate (VPA), although doses and duration of treatment were not reported. The treatment response was >75% seizure reduction in 66%, and a >50% reduction in a further 30%. The main drugs in the prescription were thought to be peach flower buds and human placenta (38). A number of other studies have shown similar results, and a summary of these can be found in a review article (39).

In an animal study, Qingyangsen in combination with PHT was found to reduce the peak level of hippocampal c-fos mRNA induction during kainic acid–induced acute seizures by 56%, whereas the inhibition of hippocampal c-fos expression during kainic acid–induced chronic seizures was disinhibited to control levels (40). It was postulated that Qingyangsen, when used in combination with PHT, may reduce the early production of fos protein and thereby inhibit the transcription of secondary genes that would contribute to the intracellular responses leading to acute seizures (40). Overall there is some evidence that a few Chinese herbal medications may have anticonvulsant activity, but randomised blinded studies are needed to substantiate this.

African traditional medicine

More than one third of 265 patients with epilepsy treated in an outpatient clinic of Lagos University Hospital had used alternative forms of treatment before attending the clinic (41). The main form of traditional medicine used was herbal preparations of various types. In addition, ∼10% of patients used herbs rubbed on scarification marks. Most of the herbal preparations were decoctions or infusions (41). Cow's urine concoction (CUC) is the traditional herbal preparation of choice for the treatment of childhood convulsions, regardless of cause, in the Yoruba-speaking people of Nigeria. Severe poisoning complicates its use, and this is the leading cause of drug poisoning in childhood in Western Nigeria (42). The main effects of CUC poisoning are respiratory depression, cardiovascular collapse, CNS depression, and hypoglycemia. The use of CUC as an AED is strongly discouraged (42).

Japanese “Kampo” medicine

Kampo medicine has been subjected to probably the most stringent and detailed analysis of all the possible complementary AED treatments. A number of medicines have been used to treat seizures, the most common and important being TJ-960. TJ-960 is mixture of nine herbal drugs, Paeoniae radix, Cinnamomi cortex, Bupleuri radix, Zingiberis rhizoma, Glycyrrhizae radix, Ginseng radix, Scutellariae radix, Pinelliae tuber, and Zizyphi Fructus (43). TJ-960 has been studied in 26 patients with epilepsy, comparing effects with 17 untreated patients (44). Sternberg's paradigm was used to examine change in cognitive function, especially short-term memory. After 8 weeks of treatment, one third of treated patients showed a 25% reduction in the number of seizures. In addition, reaction times for Sternberg's paradigm was significantly shortened. The authors concluded that besides improving seizure control, TJ-960 also improved cognitive function in patients with epilepsy (44). In another study, TJ-960 was shown to have an inhibitory effect on pentylenetetrazol-induced EEG power spectrum changes; one of the component herbal drugs, Paeoniae radix, is the important component drug (43).

Indian Ayurvedic medicine

Ashwagandha, Brahmirasayan, and Brahmigritham have been used for centuries to control seizures, in addition to their myriad effects on the CNS. BR-16A (Mentat) is widely used in conjunction with AEDs in current Indian medical practice. It has been reported to be of value in alcohol withdrawal seizures (45). A plethora of other plant-derived products has been reported to have some anticonvulsant activity based on weak anecdotal evidence. A list of some of these herbs with doubtful anticonvulsant activity is given in Table 2(46).

Table 2.  Plants with doubtful antiseizure activity
  1. Modified from Kulkarni and Verma

Acorus calamusKochia prostrata
Adonus vernalis 
Afraegle paniculataLeonurus cardiaca
 Licaria puchurymajor
Asparagus officinalis 
Astragalus centralpinusMarsilea rajasthanensis
 Marrubium vulgare
Cannabis sativa 
Canscora decussataNardostachys jatamansi
Capparis baduca 
Cinchona officinalisOleum chammonillae
Clausena anisata 
Cnestis glabraPanax ginseng
Convolvulus arvensisPatrina intermedia
Delphinium consolidaPicnomon acarna
Duboisia leichhardtiiPiper methysticum
 Pithecolobicum saman
Echinacea purpurea 
Echium vulgareRauwolfia serpentiana
Ergot alkaloidsRoylea elegans
Erythoxylum sp 
Euphorbia piluliferaSalvia nemorosa
 Solanum dasyphyllum
Galicia sp. 
Galium cruciataTaraxacum sp.
Gastrodia elataTechlea simplifolia
Haplophyllum perforatumThalictrum thumbergii
Heracleum sibiricumTrema orientalis
Herpestris monniera 

Seizures as a side effect of herbal remedies and dietary supplements

Side effects or toxicity from herbal remedies and dietary supplements can be due to a number of factors (34), including misidentification of plant species, unknown or ignored toxicity of a correctly identified plant, variability in chemical constituents of herbs, problems with nomenclature, adulteration with dangerous nonherbal ingredients, lack of dose-safety studies, variable potency of herbs depending on the climatic and soil conditions where they are grown, and mislabelling of herbs. A guide to the mechanism of possible side effects from herbal medications is outlined in Table 3(47). A review carried out by the National Poisons Unit, London, between 1983 and 1989, revealed that of 5,131 enquiries received, 968 related to herbal preparations, and of these, 245 (25%) were symptomatic (1). A number of herbal medications are associated with effects on the CNS.

Table 3.  Types of drug reactions to herbs
  1. Modified from deSmet (47).

Type A Pharmacologically predictable
 Dose dependent
 Preventable by dose reduction [e.g., Yohimbine (from yohimbine  bark) inducing hypertension]
Type B Idiosyncratic
 Pharmacologically unpredictable
 Toxicity does not correlate with dose
 Often immunologically mediated
 Often serious, potentially fatal (e.g., normal doses of yohimbine  causing allergic dermatitis and renal failure)
Type C Often over long-term therapy
 Well described, could be anticipated (e.g., muscular weakness in  users of herbal laxatives, due to hypokalemia)
Type D Delayed effects
 Not evaluated systematically; likely to be more apparent (e.g.,  hydroquinone carcinogenicity in mice indicates possible risks  from herbs containing arbutin)

Herbal remedies associated with seizures

We give a brief description of a number of important reports of seizures as a possible side effect of herbal medication.

  • 1A 45-year-old woman was seen after her third nocturnal seizure in 3 months. She had been taking Black cohosh root (Cimifuga racemosa), chaste tree berries and seeds (Vitex agnusrastus), and evening primrose oil. Investigations did not reveal any abnormalities. She discontinued the herbal medication and was prescribed CBZ. No follow-up was mentioned in the report. Black cohosh contains N-methyl cysteine and terpanoids (48).
  • 2A 6-year-old girl with cerebral palsy was reported to develop status epilepticus after a hot herb bath. Hyperthermia developed, followed by multiorgan failure, and she subsequently died. She had no history of seizures, and it is likely that seizures were a part of the hyperthermia syndrome, although status epilepticus causing rhabdomyolysis and death cannot be ruled out. No further details of the herb used were provided in the case report (49).
  • 3A young boy was admitted in status epilepticus with hyperthermia and hypotension. He had apparently eaten a small piece of root from Cicuta douglasii, which is more commonly known as water-hemlock, a weed found commonly in moist places in the United States. The authors subsequently tested the weed for its proconvulsive tendency in dogs given gallamine and artificially ventilated. EEG monitoring was done. In the four dogs tested, there were frank convulsive EEG patterns within 60 s of the intravenous administration of a diluted alcoholic extract (50).
  • 4A 19-year-old boy developed opsoclonus–myoclonus syndrome 3 days after the ingestion of a traditional African medicine for abdominal pain. This was the first time he had ingested this medication. His symptoms improved over 2 weeks [he was commenced on diazepam (DZP)]. Clinical and laboratory investigations did not reveal any of the known etiologic associations of the opsoclonus–myoclonus syndrome, although the anti-Li antibodies were not evaluated (51).

Other herbs reported to cause seizures in humans include (a) Bearberry (Arcostaphylos uva-ursi), used as an antirheumatic, which contains arbutin, and which may hydrolyse to hydroquinone, causing tinnitus, delirium, and seizures (1); (b) Ma Huang (Herba ephedrae), the major constituents of which are ephedrine, pseudoephedrine, and related compounds. It is sold in products for weight control and in those that are claimed to boost energy levels (52); (c) kava-kava, prepared from the crushed roots of the kava plant (Piper mysticum). This is either smoked or taken as tea and is thought to have antianxiety or calming effects (1); (d) Yohimbe (Pausinystalia yohimbe), a tree bark containing a variety of pharmacologically active substances. It is marketed in a number of products for body-building and is also used as an aphrodisiac. At high doses, it is a monoamine oxidase inhibitor (53); and (e) Monkshood (Aconitum sp.), used to treat rheumatism and arthritis, has been reported to cause convulsions (1).

Some commonly used herbs such as Ginkgo biloba (containing the ingredient Ginkgo), evening primrose oil, and starflower/borage (containing gamelonic acid, GLA) are thought to reduce the seizure threshold (54). It is probably not advisable to use them with AEDs, as they may diminish the anticonvulsant efficacy, or with other drugs known to lower seizure threshold (54), although this is controversial (55). Some other herbs also have been found to cause seizures in animal studies (Table 4)(56–60).

Table 4.  Herbal remedies as proconvulsants (animal studies)
Spilanthes acmella var. oleracea (56)Used in Brazil as analgesic
 hemostatic digestive larvicidal
Coriaria lactone (57) RatsYes
Choline supplements (58) RatFacilitation of kindled seizures
 in rats fed choline supplements
Marine wormwood oil (Artemisia caerulescens L.) (59) White miceYes
Stem bark of Solanum pseudoquina
 [(25S)-isosolafloridine] (60)
Used in Brazil as a tonic and
Rats and miceYes

In addition to causing seizures on their own, some herbal medications can interact with AEDs in patients known to have epilepsy (e.g., Shankpushpi, an Ayurvedic medication, interacts with PHT, lowering its serum level, thus increasing the probability of seizures (Table 5)(1,54,61–68). A large number of herbal medications are known to have an inhibitory effect on the cytochrome P-450–mediated metabolism [i.e., garlic (69), Echinacea, Licorice, chamomilla, Trifolium pratense (wild cherry) and dillapiol (70)]. This has important implications, as it is likely that there may be potential interactions with AEDs, if they are consumed concomitantly, leading to elevation of serum drug levels and causing adverse effects of AEDs.

Table 5.  Interactions between herbal remedies and anticonvulsants
Shankapushpi (61)EpilepsyPhenytoinLowered phenytoin levels, loss of seizure control
Septilin (62) CarbamazepineLowered carbamazepine levels
Sho-seiryu-to (63)Nasal allergy, colds, asthmaCarbamazepineDelayed oral absorption of carbamazepine with
simultaneous use; accelerated metabolism with
1 wk repeated pretreatment
Grapefruit juice (64) CarbamazepineIncrease bioavailability of carbamazepine
Eucalyptus oil (aromatherapy)
Asthma, diabetes, rheumatismPentobarbitoneReduced pentobarbitone-induced sleeping time,
probably due to induction of hepatic microsomal
enzyme complex
Pyrrolizidine (1)Herbal teaPhenobarbitonePhenobarbitone induces lethal transformation
of pyrrolizidines
Mentat (BR16A) (66) PhenytoinSuppresses phenytoin metabolism
Mentat (BR16A) (67) CarbamazepineIncrease bioavailability of carbamazepine
Paeoniae radix (68) PhenytoinDelayed absorption of phenytoin
Thujone (wormwood) and
(sage) (54)
Appetite stimulant, flatulencePhenobarbitoneReduce efficacy

Dietary supplements associated with seizures

Products marketed as dietary supplements include a diverse range of products, from traditional nutrients, such as vitamins and minerals, to such substances as high-potency free amino acids, botanicals, enzymes, animal extracts, and bioflavinoids that often have no recognised role in nutrition (71). Some cases of seizures, associated with these products, reported in the English literature include the following:

  • 1A 17-year-old girl who was 2 weeks postpartum was reported to have had two generalized seizures in quick succession 30 min after the ingestion of a “pick-me-up” pill. A computed tomography (CT) scan and a lumbar puncture were both normal. She was asked to discontinue the pill and started on PB. Two months later, her repeated investigations were again normal, and she was seizure free 6 months after this, despite discontinuing the PB. The tablet resembled “Histalet forte.” Analysis of the tablet revealed a large amount of caffeine, phenylpropanolamine, and pseudoephedrine (72). It is possible that this patient had cerebral venous thrombosis, which was not detected on the CT scan (CSF pressure measurement results were not mentioned), although the temporal relation of the seizures to the ingestion of the pills would make a causal relation more likely.
  • 2A 38-year-old woman with no history of seizures had two “petit mal” episodes. She had two further “petit mal” seizures before having a generalised tonic–clonic seizure. She had taken four tablets of an ephedrine-containing dietary supplement on the same day. During the next 4 days, she had five additional episodes of unresponsiveness, as well as a generalised seizure witnessed by the neurologist. Other possible causes of seizures were excluded. Since discontinuing use of the product, she has had no further seizures (73).
  • 3A male patient in his thirties was given a cumulative dose of 13.5 mg of molybdenum over an 18-day period as a dietary supplement. After this, acute psychosis developed, a series of “petit mal” attacks, and one “grand mal” seizure. He was given chelation therapy with calcium ethylenediaminetetraacetic acid (CaEDTA), after which his symptoms resolved. Single-photon emission CT (SPECT) studies and detailed neuropsychological testing revealed frontal lobe damage. One year later, the patient had residual problems with executive functioning (74).
  • 4A young woman who had an isolated seizure was assessed and found to have consumed a “Diet Fuel” supplement that contained guarana extract, whose pharmacologic action is primarily due to methylxanthine alkaloids (75).

Aspartamine (APM, saccharine) is another dietary supplement associated with seizures when used in doses >34 mg/kg body weight or when used in patients with phenylketonuria or in the population of people who are heterozygous for the PKU gene (76).


Ephedrine and related alkaloids (N-methyl ephedrine, pseudoephedrine, and phenylpropylamine) are the pharmacologically active moieties of the extract of Ephedra. Ephedra (Ma Huang) is commonly found in herbal weight-loss products referred to as herbal “fen-phen.” Ephedrine-containing products also are marketed as decongestants, bronchodilators, and stimulants. Other promoted uses include enhancement of athletic performance and body-building effects. Marketed uses of ephedrine-containing products such as “herbal ecstasy” include induction of a euphoric state and heightening of awareness and sexual sensations. These products are sold as “beyond smart drugs,” and one label indicates that the product acts on the same basis as MDMA (ecstasy). A number of combinations also include caffeine, another CNS stimulant. All of these drugs in isolation and in combination have the propensity to cause seizures. Recently the Food and Drug Administration (FDA) investigated >800 reports of adverse reactions associated with >125 different products that contained or were thought to contain ephedra alkaloids (71). Approximately 56% of the reported adverse effects occurred in persons younger than 40 years. The adverse events are broadly subdivided into those affecting the cardiovascular system and the CNS. Seizures have been widely reported after the consumption of ephedrine-containing products. Because many of these products are marketed as natural or promoted as foods, consumers may assume that these are safe and without side effects. A recent review of 140 reports of adverse events related to the use of supplements containing ephedra alkaloids that were submitted to the FDA over a 2-year period showed that 31% were definitely or probably related to supplement use, and a similar percentage were thought to be possibly related (77). Seven adverse events reported (5%) were seizures. Animal studies have shown that ephedrine intake can cause a decrease in the dose as well as the serum and CSF concentrations of theophylline needed to produce seizures in rats. This is in keeping with the observation that patients treated with therapeutic doses of ephedrine and theophylline exhibit a high incidence of CNS symptoms, including nervousness and insomnia, with an increase in EEG abnormalities. This is of obvious importance, as theophylline and ephedrine are often used together as bronchodilators, frequently as a fixed-dose combination, for the relief of asthma (78).

In response to these reports (especially cardiovascular side effects in the younger population), the FDA proposed a dosage limit of 8 mg every 6 h (24 mg/day) for ephedra alkaloids. The proposed rule also calls for a label advising consumers not to use an ephedrine-containing product for >7 days. The label also should include a warning that exceeding the recommended dosage may result in a heart attack, stroke, seizure, or death. However, life-threatening adverse reactions have been reported with doses of 1 to 5 mg (or 4 to 20 mg/day). There also are concerns that setting a dose limit may imply that a safe dose exists. Concomitant use of ephedrine-containing products and caffeine or other stimulants also should be discouraged.


Some herbal medicines may have an anticonvulsant effect; however, none has been scientifically tested in randomised blinded controlled studies. The rigorous evidence of safety and efficacy required by agencies such as the FDA for the approval of drugs is not necessary before a herbal medication or dietary supplement comes onto the market (79). Herbal medications may not be as harmless as perceived by most patients and the majority of doctors. As the use of herbal medications and dietary supplements increases, the industry should be brought under stricter regulations. Some of the herbal medications and dietary supplements may have interactions with AEDs, as shown in Table 5, or may be potentially neurotoxic and give rise to seizures, as summarised in Table 6. Therefore in assessing patients who have had seizures, it is important to specifically inquire about the use of alternative or complementary medicines. Most patients do not consider these to be “drugs” and therefore will not volunteer information about their use. Further clinical research, vigilance, and reporting of side effects is needed to define the risk associated with the use of various herbal medications and dietary supplements.

Table 6.  Herbal medications and dietary supplements known to be associated with seizures in humans
Black cohosh root (Cimifuga racemosa)
Cicuta douglasii (DC) or water-hemlock
Bearberry (Arcostaphylos uva-ursi)
Ma Huang (Herba ephedrae)
Kava Kava (Piper mysticum)
Yohimbe (Pausinystalia yohimbe)
Monkshood (Aconitum sp.)
Guarana extract
Ephedrine and related alkaloids (N-methyl ephedrine,
  pseudoephedrine, and phenylpropamine)