Herbal medicine and 'Silent Witness'

Could herbal remedies be deadly?

Michael Thomsen

A Review of the Potential Forensic Significance of Traditional Herbal Medicines by Professor Roger Byard

A response from Michael Thomsen


'Herbal remedies could be deadly, says forensic pathologist Professor Roger Byard' screams the headline in the Adelaide newspaper.

Why does the title of a scholarly article1 change from A Review of the Potential Forensic Significance of Traditional Herbal Medicines to Herbal remedies could be deadly, says forensic pathologist Professor Roger Byard2? Answer: When you want to dumb down medical writing to the masses and sensationalise the issues.

Prof. Byard's article basically states that use and abuse of herbal medicines should be considered in forensic pathology. I assume that no one will have any issue with this premise. The article then attempts to examine eight areas: toxic ingredients (mostly heavy metals), substituted ingredients (with the wrong herb), incomplete processing (of aconite for example, in TCM), accidental contamination (e.g. pesticides, aflotoxins), adulteration with pharmaceutical drugs, drug interactions, issues with surgery and the problem of identification and detection of organic toxins in the body.

These are all very valid issues for forensic pathology, and of course for the practice of herbal medicines. These are serious issues which all need regulation, policing and education.

The problem with Prof. Byard's article is that it contains too many inaccurate statements about commonly used herbal medicines. This response is an attempt to correct these inaccuracies.

The publication by Prof. Byard is presented as if it were a new study. He has, in fact, not performed any actual research. His paper is a review of the medical literature. The article is mostly based on other reviews and therefore on second-hand references. As result, it contains too many erroneous statements.

The concern about heavy metal contamination is not new. In recent times, the concern over heavy metal content first surfaced in December 2004 with the publication of an article in the Journal of the American Medical Association, which found that 14 Ayurvedic products sold in the Boston area of the United States contained potentially harmful levels of heavy metals.

Although there are published cases of terrible poisoning, especially in India and China - but also in Western countries where herbal medicines are either unregulated as food supplements or where the products have been purchased online or in local Asian shops selling unregulated products - we don't actually know how severe the problem is.

Prof. Byard himself mentions that only 0.2% of acute medical hospital admissions in Hong Kong were attributed to the use of herbal products in one study. He also quotes Ernst who summarised issues that occur in these cases as 'nonmedically qualified healers, lack of product standards, undeclared ingredients, nondisclosure of usage and long-term medication'.

According to the 2008 Annual Report of the American Association of Poison Control Centers' National Poison Data System (NPDS), there was not one death caused by a dietary supplement in 2008 in the US. The new 174-page annual report of the American Association of Poison Control Centers, published in the journal Clinical Toxicology, shows zero deaths from multiple vitamins; zero deaths from any of the B vitamins; zero deaths from vitamins A, C, D, or E; and zero deaths from any other vitamin. Additionally, there were no deaths whatsoever from any amino acid or herbal product. This means no deaths at all from blue cohosh, echinacea, ginkgo biloba, ginseng, kava kava, St John's wort, valerian, yohimbe, Asian medicines, Ayurvedic medicines, or any other botanical. There were zero deaths from creatine, blue-green algae, glucosamine, chondroitin, melatonin, or any homeopathic remedies. 61 poison centers provide coast-to-coast data for the U.S. National Poison Data System, which is then reviewed by 29 medical and clinical toxicologists. NPDS, the authors write, is 'one of the few real-time national surveillance systems in existence, providing a model public health surveillance system for all types of exposures, public health event identification, resilience response and situational awareness tracking'.2 The full report can be downloaded here.

Nonetheless, the issue of heavy metal contamination is being taken seriously by regulators and associations. In India the Department of AYUSH is working on mandatory testing for heavy metals and in the US, The American Herbal Products Association (AHPA) has published a heavy metals white paper it says will assist industry to comply with current Good Manufacturing Practice (cGMP).

Heavy metal contamination is a serious problem and it should be addressed by the regulators by imposing mandatory testing for all products, including ingredients (dried herbs and extracts) as well as finished products.

The addition of pharmaceuticals by unscrupulous manufacturers is of course also a serious issue which perhaps needs better policing by the TGA.

I must admit I am a bit of a fan of the British patho-drama, 'Silent Witness'. So I read with interest the section on Forensic Implications in Prof Byard's article:

A range of issues arise with the use of herbal medicines, including problems in determining what might constitute lethal levels of the active ingredients, how a particular herb may have contributed to death, and whether a herb may actually have caused death. The lack of diagnostic pathological findings at autopsy may also create difficulties.

The references given for this introduction have nothing to do with therapeutic herbal medicine. The first reference refers to a paper by Prof Byard himself(3) which describes two cases of poisoning with Datura spp. - presumably Datura stramonium, also known as thorn apple, from the deadly nightshade family. The second reference is about a person who self-injected with a toad extract, presumably to get high. The third reference refers to a case of injection of 5-methoxy-N,N-diethyltryptamine from an Auyrvedica formulation. I confess, I didn't know this drug. So I looked it up: 5-methoxy-N,N-diisopropyltryptamine (5-MeO-DIPT) is a tryptamine derivative and shares many similarities with schedule I tryptamine hallucinogens such as alpha-ethyltryptamine, N,N-dimethyltryptamine, N,N-diethyltryptamine, bufotenine, psilocybin and psilocin. Since 1999, there has been a growing popularity of 5-MeO-DIPT among drug abusers. This substance is abused for its hallucinogenic effects.4

Of course a forensic pathologist needs to know about such poisons, but they are not related to the use of herbal medicine for therapeutic treatments as it is practised in Australia and other countries.

Prof. Byard ends his article by stating:

The extent of the role of herbal medicines in the types of cases presenting to forensic facilities is yet, however, to be determined.

Exactly, couldn't agree more. Do the research so we can base our actions on facts, not speculation.

Prof. Byard also attempts to explain the possible interactions of herbal medicines with pharmaceutical drugs and possible dangers of using herbs during surgery. Unfortunately, he just quotes previous reviews instead of drilling down into actual case studies or trials. For the section of surgery he just quotes inaccurate statements about commonly used Western herbs from a review by Dasgupta published in 2006.5 The over-reliance on secondary references is a dangerous practice.

The review by Dasgupta is available here.


In the following section, I have tried to group the statements together and provide a comment.

Statement 1

St John's wort (Hypericum perforatum) induces several cytochrome P450 enzymes and the transporter protein P-glycoprotein and has significant effects on a range of drugs causing reduced anticoagulation by warfarin, decreased action of cyclosporine with acute organ transplant rejections, and intermenstrual bleeding in women taking the oral contraceptive pill.


While it is correct that St John's wort should be avoided or used with caution during therapy with warfarin and cyclosporine, the interaction with the oral contraceptive pill is likely to be clinically insignificant.

Three reports suggest a link between SJW and break-through bleedings in women taking contraceptives (twice within one week, once within 3 months after start of SJW treatment).6 The case of menstrual break-through bleeding three months after having started SJW therapy appears an unusually long delay for an interaction usually apparent after 8-10 days of continuous application, and might therefore be unrelated to SJW. In the two women where the bleeding was observed within one week of SJW intake, the reaction fits to the general picture observed in clinical trials with preparations rich in hyperforin. It has, in addition, been demonstrated that even if hyperforin-enriched St John's wort preparation are responsible for induction of breakthrough bleedings, the anti-ovulatory efficacy of the contraceptive is still not endangered.7, 8

In any case, St John's wort preparations with normal levels of hyperforin (typically around 1.5-2.0 %) have been demonstrated not to cause clinically relevant pharmacokinetic interactions, and thus possess a better safety profile compared to special extracts artificially enriched with hyperforin.9, 10

Statement 2

St John's wort: It has also significantly reduced the bioavailability of many drugs including serum digoxin, theophylline, amitriptyline, indinavir (an HIV-1 protease inhibitor), and methadone levels, the latter resulting in the re-emergence of withdrawal symptoms.


Only a few of the interactions hitherto detected can be considered clinically relevant. In all of these cases, such as treatment with cyclosporine, indinavir, warfarin, the patients are under a high level of surveillance, which should contribute to risk minimisation11-13, provided that the patients are able to discuss the use of St John's wort openly with their doctors - which we know is a problem as patients fear that their interests in herbal medicines may be ridiculed, dismissed or ignored by the doctor.

It can be quite safely assumed that the risk of clinically relevant interactions is low when preparations containing normal extracts, not enriched in hyperforin, are used. SJW preparations with normal levels of hyperforin (typically around 1.5-2.0 %) have been demonstrated not to cause clinically relevant pharmacokinetic interactions, and thus possess a better safety profile compared to special extracts enriched with hyperforin9, 10

Trials performed with a preparation void of hyperforin failed to show any pharmacokinetic influence in a study on plasma digoxin or contraceptive levels (14;15), Another study also failed to demonstrate any interaction between a low hyperforin extract and digoxin in 28 healthy volunteers (verum: n = 16; placebo: n = 12)16.

Statement 3

Garlic: increased risk of hemorrhage due to inhibition of platelet aggregation and increased fibrinolysis. Ginkgo: increased risk of hemorrhage due to inhibition of platelet activating factor.

The side effects of both of the above may be worsened if drugs that inhibit platelet aggregation are also being used. Gingko (Gingko biloba) and garlic (Allium sativum) have increased the risk of bleeding with anticoagulants, while garlic has increased the hepatotoxicity of paracetamol and enhanced the effect of oral hypoglycemic agents.

Other preparations may potentiate the action of warfarin thus increasing the risk of hemorrhage; these include Dan shen, Dong quai, Devils claw, ginseng, and Siberian ginseng.


It is an often repeated statement that ginkgo (Ginkgo biloba) and garlic (Allium sativum) may have antiplatelet or anticoagulant effects, which could potentially exacerbate the risk of gastrointestinal bleeding from non-steroidal anti-inflammatory drugs or corticosteroids, however, the reference is of just one case report, which discusses a rare and unusual event of excessive garlic ingestion causing a spontaneous spinal epidural haematoma, which would not be considered typical. If Prof. Bayard had bothered to look, he would have found a study from 2005 concluding: Ginkgo and ginger at recommended doses do not significantly affect clotting status, the pharmacokinetics or pharmacodynamics of warfarin in healthy subjects.17

I have not been able to find any evidence that devil's claw (Harpagophytum procumbens) interacts with warfarin (Medline search).

A study on a product containing Siberian ginseng (Eleutherococcus senticosus) found no significant interaction with warfarin.18

Prof Byard quotes the review of Dasgupta, 2006, for proof that ginseng interacts with warfarin. However, the study referenced in Dasgupta's article is an Australian study from 2004 which concludes:

Coadministration of warfarin with ginseng did not affect the pharmacokinetics or pharmacodynamics of either S-warfarin or R-warfarin.19

There has been just one documented case of interaction between Dong quai (Angelica sinensis) and warfarin. Dong quai increased the prothrombin time two-fold.(20) Dan shen (Salvia miltiorriza) may likewise also amplify the anticoagulant effect of warfarin and should be avoided or used with caution.21

Garlic has been shown to increase the toxicity of paracetamol? It is the other way round: Garlic may reduce the toxicity of paracetamol.

Acetaminophen, the most commonly sold over-the-counter antipyretic analgesic, is capable of causing severe and sometimes fatal hepatic damage in humans and experimental animals. The incidence of liver injury due to acetaminophen overdose, either with suicidal intent or by accident, is increasing. Garlic is among those medicinal plants famous for its different health protective effects. In this study, the protective effects of garlic extract on acute acetaminophen-induced liver injury were investigated using freshly isolated rat hepatocytes. The hepatocytes were isolated from Sprague-Dawley male rats by a two step collagenase model. Formation of Reactive Oxygen Species (ROS) and Glutathione (GSH) depletion were studied after addition of acetaminophen to cell suspensions. The effects of garlic extract on prevention of ROS formation as well as GSH depletion was investigated and compared with the effects of N-Acetyl Cysteine (NAC) as the standard treatment. Reactive oxygen species formation was assessed by a spectrofluorometry method and garlic extract was shown to be as effective as NAC in decreasing ROS formation induced by acetaminophen. Glutathione (GSH) levels of hepatocytes were determined using HPLC. Garlic extract was effective in preventing GSH depletion significantly (p < 0.05). It is concluded that garlic extract has an antioxidant effect and can protect hepatocytes from GSH depletion following NAPQI production.22

Garlic oil has been shown to improve the clearance of the toxic metabolites of the acetaminophen from the liver,23 and the garlic constituent ajoene, has been shown to exhibit a hepatoprotective effect against acetaminophen-induced liver injury in mice. A pretreatment with ajoene suppressed the rise in serum glutamic-pyruvic transaminase activity and the reduction in the hepatic reduced glutathione level. These effects of ajoene were observed dose-dependently (20-100 mg/kg). The pretreatment by ajoene also suppressed the decrease in hepatic protein thiol content resulting from acetaminophen administration.24

Garlic has anti-diabetic effects; however there are no reports of adverse interactions between oral hypoglycaemic agents and garlic. Most likely they can be used together.

Statement 4

Ephedra: increased risk of myocardial and cerebral ischemic events; potentially lethal interaction with monoamine oxidase inhibitors; intra-operative hemodynamic instability due to reduction of endogenous catecholamines; ventricular arrhythmias with halothane.


These effects are related to ephedrine. The judicious use of the herbal extract of Ephedra sinensis is effective in the treatment of asthma and carries very little risk of such side-effects. However, as a result of ephedrine misuse, the use of Ephedra is banned in many countries including Australia.

Statement 5

Asian ginseng (Panax ginseng) has potentiated the adverse effects of the monoamine oxidase inhibitors, enhanced the effect of oral hypoglycemic agents, decreased the effects of immunosuppressants, and had an additive effect with benzodiazepines.

Ginseng: hypoglycemia; reduction of anticoagulation properties of warfarin; increased risk of hemorrhage due to inhibition of platelet aggregation.


Ginseng has antidepressant and anxiolytic activity. The activity may be mediated partly through enhancing the monoamine neurotransmitter concentration and brain-derived neurotrophic factor expression in the hippocampus,25 but I have not been able to find a report on interaction with MAO inhibitors.

Ginseng may reduce hyperglycaemia in type 2 diabetic patients. Not a bad therapeutic activity, it may reduce the dosage or requirement of oral hypoglycaemic medication. There is no evidence is causes hypoglycaemia in healthy people. Ginseng has no effect on platelet function in vivo26 and as mentioned above, has no interaction with warfarin.

Statement 6

Certain herbal remedies such as Borage oil and Evening primrose oil contain gamolenic acid that lowers the seizure threshold in epileptics, thus counteracting drugs such as phenobarbitone and phenytoin.


Now, where does this crazy statement originate from? It seems to stem from an article by Miller, 'Herbal medicinals: selected clinical considerations focusing on known or potential drug-herb interactions', published in 1998.27 Miller quotes a book as the reference: Herbal Medicines: A Guide for Health-Care Professionals published in 1997.28 I haven't got this book so I can't check the references but in 2007 Puri29 published a review in regards to this issue:

The concern that evening primrose oil might cause epilepsy or seizures, or reduce the threshold for seizures, originated from two papers published in the early 1980s. These original reports are re-examined, and the association of evening primrose oil with seizures is shown to be spurious. Not only are linoleic acid and gamma-linolenic acid safe in epilepsy, with prolonged oral administration of linoleic acid and alpha-linolenic acid (in a 4:1 mixture) protecting rats from having seizures in four different epilepsy models, but the evening primrose oil-derived omega-6 fatty acid arachidonic acid inhibits sodium ion currents and synaptic transmission, while the evening primrose oil-derived eicosanoid prostaglandin E(1) appears to have anticonvulsant activity. In light of these findings, it is suggested that formularies should now remove seizures or epilepsy as a side-effect of evening primrose oil, and should remove a history of seizures or epilepsy as a contraindication to taking evening primrose oil.

Statement 7

Echinacea: allergic reactions, immunosuppression, reduced efficacy of immunosuppressants.


Allergic reaction to echinacea has been reported but it is extremely rare. Nevertheless, allergic reaction may occur.30 There is no clinical evidence that echinacea will reduce the efficacy of immunosuppressants. An in vivo study found no evidence of altered natural killer cell activity, T cell-mediated delayed-type hypersensitivity, or specific antibody formation in male rats given either a 225 mg/kg or 50 mg/kg of the commercial echinacea for 6 weeks. Antibody formation was significantly suppressed in female but not male rats given 250 mg/kg for 2 weeks of the commercial echinacea. The local products tested had no effect on antibody formation. We concluded that our study provided no supporting evidence for immunostimulatory activity by the echinacea preparations we examined and, in fact, may be immunosuppressive under some conditions.31

Statement 8

Kava and valerian: possible exacerbation of sedative effects of anesthetic agents.


This is poor paraphrasing from the review by Dasgupta who states that kava should be discontinued at least 24 hours before surgery because kava can increase the sedative effect of anesthetics. Valerian is not mentioned in the review by Dasgupta, so I assume that Prof. Byard has added valerian himself. Both kava and valerian are mild sedatives. I actually doubt it will have much of an impact compared to a pharmaceutical anaesthetic, but nevertheless, a herbalist would always recommend ceasing all herbal medication 24 hours prior to surgery.

Statement 9

Herbal products may also produce abnormal laboratory results by directly interfering with immunoassays, by increasing or decreasing concentrations of prescribed drugs, or by direct hepatotoxicity. The amount of interference due to cross-reactivity between herbs and drugs is also dependent on the assay method used withdrawal symptoms.


Yes, they may. Grapefruit is also known to influence drug metabolism. Let us have some more research because there is too much speculation, too much assuming and not enough facts.

Reference list

  1. Byard R. A Review of the Potential Forensic Significance of Traditional Herbal Medicines. www.adelaide.edu.au/news/binary5521/Journal.pdf. 2010.
  2. Saul A. No Deaths from Vitamins, Minerals, Amino Acids or Herbs Poison Control Statistics Prove Supplements' Safety . http://www.orthomolecular.org/resources/omns/v06n04.shtml. 2010.
  3. Byard RW, James RA, Felgate P. Detecting organic toxins in possible fatal poisonings--a diagnostic problem. J Clin Forensic Med 2002; 9(2):85-88.
  4. DrugBank. drug card for 5-Methoxy-N,N-diisopropyltryptamine (DB01441). http://www.drugbank.ca/drugs/DB01441. 2010.
  5. Dasgupta A, Bernard DW. Herbal remedies: effects on clinical laboratory tests. Arch Pathol Lab Med 2006; 130(4):521-528.
  6. Bon S, Hartmann K, Kuhn M. Johanniskraut: Ein Enzyminduktor? Schweiz Apoth Ztg 1999;(16):535-536.
  7. Hall SD, Wang Z, Huang SM, Hamman MA, Vasavada N, Adigun AQ et al. The interaction between St John's wort and an oral contraceptive. Clin Pharmacol Ther 2003; 74(6):525-535.
  8. Pfrunder A, Schiesser M, Gerber S, Haschke M, Bitzer J, Drewe J. Interaction of St John's wort with low-dose oral contraceptive therapy: a randomized controlled trial. Br J Clin Pharmacol 2003; 56(6):683-690.
  9. Müller SC, Uehleke B, Wöhling H, Petzsch M, Majcher-Peszynska J, Hehl EM et al. Effect of St John's wort dose and preparations on the pharmacokinetics of digoxin. Clin Pharmacol Ther 2004; 75(6):546-557.
  10. Müller SC, Majcher-Peszynska J, Uehleke B, Klammt S, Mundkowski RG, Miekisch W et al. The extent of induction of CYP3A by St. John's wort varies among products and is linked to hyperforin dose. Eur J Clin Pharmacol 2006; 62(1):29-36.
  11. Schulz V. Indicence and clinical relevance of the interactions and side effects of Hypericum preparations. Phytomed 2001; 8(2):152-160.
  12. Butterweck V, Derendorf H, Gaus W, Nahrstedt A, Schulz V, Unger M. Pharmacokinetic herb-drug interactions: are preventive screenings necessary and appropriate? Planta Med 2004; 70:784-791.
  13. Schulz V. Safety of St. John's Wort extract compared to synthetic antidepressants. Phytomedicine 2006; 13(3):199-204.
  14. Brattström A. Der Johanniskrautextrakt Ze 117. Wirksamkeit und Sicherheit. Dtsch Apoth Ztg 2002; 142(30):3695-3699.
  15. Will-Shahab L, Bauer S, Kunter U, Roots I, Brattstrom A. St John's wort extract (Ze 117) does not alter the pharmacokinetics of a low-dose oral contraceptive. Eur J Clin Pharmacol 2008.
  16. Arold G, Donath F, Maurer A, Diefenbach K, Bauer S, Henneike von Zeppelin HH et al. No relevant interaction with alprazolam, caffeine, tolbutamine, and digoxin by treatment with a low-hyperforin St. John's wort extract. Planta Med 2005; 71:331-337.
  17. Jiang X, Williams KM, Liauw WS, Ammit AJ, Roufogalis BD, Duke CC et al. Effect of ginkgo and ginger on the pharmacokinetics and pharmacodynamics of warfarin in healthy subjects. Br J Clin Pharmacol 2005; 59(4):425-432.
  18. Hovhannisyan AS, Abrahamyan H, Gabrielyan ES, Panossian AG. The effect of Kan Jang extract on the pharmacokinetics and pharmacodynamics of warfarin in rats. Phytomedicine 2006; 13(5):318-323.
  19. Jiang X, Williams KM, Liauw WS, Ammit AJ, Roufogalis BD, Duke CC et al. Effect of St John's wort and ginseng on the pharmacokinetics and pharmacodynamics of warfarin in healthy subjects. Br J Clin Pharmacol 2004; 57(5):592-599.
  20. Page RL, Lawrence JD. Potentiation of warfarin by dong quai. Pharmacotherapy 1999; 19(7):870-876.
  21. Chan TY. Interaction between warfarin and danshen (Salvia miltiorrhiza). Ann Pharmacother 2001; 35(4):501-504.
  22. Anoush M, Eghbal MA, Fathiazad F, Hamzeiy H, Kouzehkonani NS. The protective effects of garlic extract against acetaminophen-induced oxidative stress and glutathione depletion. Pak J Biol Sci 2009; 12(10):765-771.
  23. Kalantari H, Salehi M. The protective effect of garlic oil on hepatotoxicity induced by acetaminophen in mice and comparison with N-acetylcysteine. Saudi Med J 2001; 22(12):1080-1084.
  24. Hattori A, Yamada N, Nishikawa T, Fukuda H, Fujino T. Protective effect of ajoene on acetaminophen-induced hepatic injury in mice. Biosci Biotechnol Biochem 2001; 65(11):2555-2557.
  25. Dang H, Chen Y, Liu X, Wang Q, Wang L, Jia W et al. Antidepressant effects of ginseng total saponins in the forced swimming test and chronic mild stress models of depression. Prog Neuropsychopharmacol Biol Psychiatry 2009; 33(8):1417-1424.
  26. Beckert BW, Concannon MJ, Henry SL, Smith DS, Puckett CL. The effect of herbal medicines on platelet function: an in vivo experiment and review of the literature. Plast Reconstr Surg 2007; 120(7):2044-2050.
  27. Miller LG. Herbal medicinals: selected clinical considerations focusing on known or potential drug-herb interactions. Arch Intern Med 1998; 158(20):2200-2211.
  28. Newall CA, et al. Herbal Medicines: A Guide for Health-Care Professionals. London: Pharmaceutical Press, 1997.
  29. Puri BK. The safety of evening primrose oil in epilepsy. Prostaglandins Leukot Essent Fatty Acids 2007; 77(2):101-103.
  30. Mullins RJ, Heddle R. Adverse reactions associated with echinacea: the Australian experience. Ann Allergy Asthma Immunol 2002; 88(1):42-51.
  31. South EH, Exon JH. Multiple immune functions in rats fed Echinacea extracts. Immunopharmacol Immunotoxicol 2001; 23(3):411-421.