Is Comfrey Safe?

David Hoffmann B.Sc., FNIMH

(thanks to David for sharing this says jim)


Recent years have seen profound cultural changes that are transforming the sociopolitical environment within which the practice of herbalism is flowering. The dominant culture hs moved to embrace herbal ism, and it is no longer seen as fringe. Within this brave new world, the phytotherapy community faces both wonderful opportunities and profound challenges. One of the thorniest of these concerns the safety, toxicity, and contraindications of herbs.


Today, much is being written about herbs and herbalism that has little to do with plants or patients and everything to do with the belief system and dogma of the writer. The herbal “true believers” deny any problems that might exist with herbs because herbs are natural, other hand, some supposedly objective scientists can make even hawthorn sound as dangerous as crack cocaine. Some critical thinking is needed to traverse this maze of information and mis-information, evangelism and bigotry. In general, statements about the toxicity of a plant may be based on any of a number of factors.


Observation of actual cases. These may be published reports on individual responses to herbs or may come from published research on therapeutic trials in which research subjects were either healthy volunteers or undergoing treatment for a health condition.

Extrapolation from chemical effects to whole plant effects. Knowing the effects of a specific plant constituent provides little basis for conclusions about effects of the whole plant. An example here is a contraindication for salicylate-containing herbs in cases of peptic ulcer based on the fact that aspirin will aggravate peptic ulcer. While this may sound appropriate in theory; in reality the salicylate-rich plant meadowsweet (Filipendula ulmaria) works as an effective anti-inflammatory and vulnerary in such conditions—and causes none of the side effects of aspirin.

Extrapolation from animal evidence to humans. For example, the statement that St. John’s wort (Hypericum peforatum) causes photosensitivity is based on observance of phototoxicity reactions in light-colored range animals that graze on the plant.

Traditional knowledge, or belief based on experience with extensive use, is another potential source of information about risks and contraindications for herbs. However, this information must also be carefully considered. For example, feverfew is traditionally used for the prophylaxis of migraine in the form of a fresh leaf chewed daily, which may cause canker sores (aphthous ulcers). However, this often-repeated risk does not apply when the leaf is swallowed or when the herb is taken in tablet and tincture forms, which cause no buccal irritation.

Subjective opinion or belief system. Subjective statements from those considered experts in the field are often unquestioningly accepted as insightful, considered opinions. But who are the experts in these situations?

Consider the following quotation from The Honest Herbal, by Professor Varro Tyler. Charles C. Bennett, vice president of Public Education for the Arthritis Foundation, has suggested that inquiries concerning yucca be answered
”…by saying that there is no proper scientific evidence that yucca tablets are helpful in treating rheumatoid arthritis or osteoarthritis; that they are probably harmless; and that the real danger would be in taking yucca tablets instead of following proper and proven treatment procedures, which could lead to irreversible joint damage and possible disabilities.”
[1] Nothing need be added to this statement.

I beg to differ. This statement turns the opinion of Mr. Bennett about treatment advisability into a statement about a plant. The statement is about patient choice, and yucca has nothing to do with it.

Herb safety issues are complex. As an example, consider comfrey (Symphytum officinale), an herb with a very long history of traditional use. Comfrey has been used both internally (as a demulcent) and externally (as a vulnerary). The herb has become a topic of considerable debate, as it contains pyrrolizidine alkaloids (PAs), which have been shown to pose a real risk of hepatotoxicity.

There can be no question about the concerns posed by PAs to humans.[2] However, it must be remembered that we cannot demonize a naturally occurring substance simply because it is inappropriate for human consumption. PAs play vital roles in the ecology of many butterffies.[3] Several species feed on PA-containing plants. The butterflies incorporate the PAs into their tissues and thus become unpalatable to predators.[4] In the tropical danaid butterflies, PAs serve as essential precursors to pheromones that act as oviposition stimulants for females.[5]

However, comfrey is not simply a vehicle for the pharmacological effects of a specific constituent. The herb’s pharmacological impact is far more complex than that. The only time an herb can be viewed solely as an “organic drug delivery system” is when it contains pharmacologically active levels of potent constituents that override any potential whole plant effects. The fact that PAs pose risks cannot be simply used by extension to imply that comfrey is an unsafe herb. A recent, well-balanced review by Dorena Rode on toxicological findings concerning comfrey sheds considerable light on this subject.[6]

To address the question of whether or not comfrey is safe for therapeutic use, a range of issues must be considered, and not all of these involve toxicology. Webster’s Revised Unabridged Dictionary defines the word safe as meaning: Free from harm, injury, or risk; untouched or unthreatened by danger or injury; unharmed; unhurt; secure; whole; as, safe from disease; safe from storms; safe from foes. Unfortunately, there is a common assumption that might be paraphrased as “Herbs are natural, and therefore safe.” This is clearly not the case. Cliffs are natural and yet people fall off them, water is natural and yet peopie drown, arsenic is natural and yet it is lethally toxic.- To put this in the crudest terms, we are all going to die, and that too is natural!

Thus,- if “safe” is taken to mean completely risk-free, then the answer must be no, comfrey is not safe. For example, as with all herbs and drugs, there is a small chance that an allergic response will occur. As is discussed later in this chapter, it is more meaningful to consider the concepts of hazard and risk than “safety.” When these concepts are applied to comfrey, we see that the hazard it poses is its potential to cause an adverse effect, and the risk it poses is the statistical probability that the hazard will occur under specific exposure conditions.

All substances (including herbs) and all activities (including taking herbs internally) carry risks. The core therapeutic issue is whether the benefits outweigh the dangers. This therapeutic index is not the subject of this presentation, but it should be noted here that determining a risk-benefit relationship depends on the informed judgment of the clinician, not an absolute of pharmacology. Many very risky pharmaceuticals are prescribed because the clinician believes that the benefits outweigh the risks.[7]
To compound matters, questions of herb safety are posed in very divergent contexts and lead to similarly divergent answers. The assumptions, expectations, belief system, knowledge, and experience of both the person who asks the question and the person who responds will affect the outcome. A bureaucrat given the responsibility of regulating the herbal marketplace is not likely to consider the crucial risk-benefit complexities that a phytotherapist takes into account when reaching clinical decisions about a unique patient.

When the reports on comfrey toxicity are reviewed, some interesting points can be made. the following quotation comes from Dorena Rode, author of an insightful review on comfrey toxicity.
One might expect that new toxicity research or an unacceptable number of adverse reactions prompted these recent actions, but neither is the case.

The very few specific reports of human toxicity related to comfrey all come from the period between 1980 and 1990, when a number of cases of veno-occiusive disease were reported.[9] There is no question about the diagnoses. However, it is important to note that in these cases, the connection with comfrey was not considered in the context of other contributing factors. For example, concomitant illness, the use of prescription or over- the-counter hepatotoxic drugs (like acetaminophen, for example), and impaired nutritional status clearly increase the likelihood that PA-containing herbs will cause hepatotoxicity.

With minimal epidemiological data, what insights can be garnered from the laboratory research into toxicity? As with many statements about herbal toxicity, the evidence proffered comes primarily from rodent studies that utilized high levels of purified PAs. No systematic toxicity testing or clinical trials of comfrey have been performed. Although PA poisoning in humans does occur, it is most commonly a consequence of consuming plants other than comfrey.[10]

Such reliance on animal experimentation data and toxicity reports about other plants gives us little insight into the risks and therapeutic benefits of the human use of comfrey. Rode enumerates four limitations of the published research.


Not all PAs have similar toxicity. This group of alkaloids do not pose a uniform risk. Based on structure-toxicity studies, it can be concluded that the PAs in comfrey (such as symphytine, a retronecine monoester) are less toxic than those present in the plants Senecio, Crotolaria, and Heliotropium. These have actually caused human toxicity (for example, senecionine, a macrocyclic retronecine diester).[11]

Not all animals are susceptible to PA toxicity. As with most substances, responses to PAs among different animals vary greatly. Pigs, chickens, and rats are highly sensitive to poisoning by Senecio, whereas mice and sheep are resistant. However, and more significant, the response of one species to Senecio might not reflect its susceptibility to other PAs.[12] In addition, the route of administration can dramatically affect the toxic response. For example, rabbits are relatively resistant to chronic feeding of Senecio, but are killed by a single injection of the purified alkaloids.[13]  Although theoretically sensitivity to PAs, pigs readily accept comfrey as a food and show no adverse effects, even when comfrey represents 40% of their diet. Rats, however, appear to be very sensitive to the same PAs. When eating large amounts of comfrey or injected with comfrey PAs, rats develop the hepatic lesions indicative of PA poisoning.[14] This calls into question the validity of using rodent animal models as indicators of human response to PAs.

Comfrey species vary in PA content. Between 85% and 97% of the PAs in Symphytum officinale, the corn- frey commonly grown in American gardens, are built around the less toxic retronecine monoester. However, Russian comfrey (Symphytum x uplandicum), contains higher levels of the diester, which, is known to have a greater toxicity.[15] Bearing in mind the differences in toxicity among various PAs, and the variable distribution of PAs in different comfrey species and varieties, we might conclude that extrapolations of research results from one species to another may be unreliable.

Effects of isolated PAs might not be representative of whole plant use. As with many herbs, it is problematic to assume that the pharmacology of a specific constituent can be used to predict the pharmacology of the whole plant. Veterinary studies have shown that the formation of toxic PA metabolites is reduced by concurrent administration of the sulfur-containing amino acid methionine or cysteine.[16] Similarly, protein-deficient dietsenhance the toxicity of PAs.[17] Most toxicity studies used purified PAs, ignoring the potential protective effects of co-occurring nutrients present in the whole plant. This suggests that studies using purified PAs probably overstate the health risks associated with comfrey extracts or the whole plant.

So what can be concluded about the safe use of cornfrey? Again, the clinical concept of comfrey’s therapeutic index must be invoked. The therapeutic index compares the therapeutically effective dose to the toxic dose of a substance. This gives an indication of the relative safety of a drug or herb by providing a ratio of the dose that produces toxicity to the dose needed to bring about the desired therapeutic response. According to The Merck Manual, the term describes “informed clinical judgment,” but who are the relevant clinicians in the case of comfrey?

All of this might lead one to the conclusion that toxicologists’ statements about the safety of comfrey may be based on inadequate data. Unless benefits are compared with risks, we have only conjecture. Consider, for example, the case of nonsteroidal anti-inflammatory drugs (NSAIDs). Such medications can play an undeniably beneficial role in the control of arthritic inflammation, for example. But are they safe?

More than 30 billion NSAIDs are consumed annually in the United States. The major effect of all NSAIDs is to decrease the synthesis of prostaglandins by reversibly inhibiting cyclooxygenase, an enzyme that catalyzes the formation of prostaglandins from the precursor arachidonic acid. Prostaglandins enhance the inflammatory response, but also renal blood flow and cytoprotection of gastrointestinal mucosa. The 1998 report from the American Association of Poison Control Centers (AAPCC) documents 52,751 toxic exposures to ibuprofen alone, of which 13,519 were treated in health care facilities. Four deaths were reported. For the same year, the AAPCC also reports a high rate of complications arising from therapeutic use of NSAIDs, specifically, 100,000 toxic exposures resulting in hospitalization and 10,000 resulting in fatality. In the face of such toxicological data, it may seem strange that ibuprofen is considered safe enough for non-prescription use. However, the benefits are believed to outweigh the risks. Thus, the toxicity data are seen in the context of the therapeutic benefits.


The phytotherapist would say that the same process must be applied to comfrey before conclusive statements can be made. Thus, an assessment of comfrey’s safety must take into account more than the real and theoretical risks posed by PAs:
• How does the toxicity of a whole plant preparation compare with that of the actual PAs present?
• How does the toxicity of comfrey PAs compare to that of the well-studied PAs contained in toxic plants from the genus Senecio?
• How does the hazard posed by comfrey PAs relate to the risk of using comfrey in herbal teas and tinctures?
• What real risks are posed and what therapeutic benefits are offered by the use of comfrey?
• How do the benefits compare with the risks?

Unless the knowledge and experience of clinical phytotherapists are taken into account, the lack of mainstream clinical experience with comfrey might suggest it offers no therapeutic benefits.


Factors to Consider When Assessing Herb Safety
In general, a number of broad factors must be taken into account:
• Has traditional use of the plant revealed any knowledge of toxicity?
• Does the plant contain constituents with established toxicity?

• Are there inherent risks unique to the form in which the herb is administered? (For example, does topical use of comfrey pose a different level of risk than internal use?)

• Does objective data for both toxicity and efficacy exist? Are the research results valid enough to form some basis for approximate risk assessment?

• How does the science compare with the perceived risk expressed by the FDA, practitioners of orthodox medicine, or herbalists?

Based on the context in which the herb is being prescribed, selected, or purchased, what might be called iatrogenic risks arise—in other words, inadvertent risks introduced through mistakes, lack of knowledge, or other such factors. If use of comfrey was suggested by a clinician, did this clinician understand the parameters for safe use? At the other end of the spectrum, was the herb self- selected by a consumer under the influence of commercial hype?

The complex of issues concerning substance interactions must also be assessed. As discussed later in this chapter, the nature of interactions varies. One must consider the ways in which an herb might interact with prescription or over-the-counter drugs, other herbs, or even other modalities.
Sociopo1itical risks must also be evaluated. From certain cultural perspectives, the use of herbs by anyone outside of a professional elite poses some unique concerns. The informed use of therapeutic plants empowers people. However, in this context, any lessening of people’s dependence on the professional elite (or financial indebtedness to the industrial-medical complex) can be seen as a potential threat to the power structure of the dominant hierarchy. The more people develop their own relationship with nature, the less power the dominators will have. Just as attempts are made to discourage people about the green movement by labeling it “ecoterrorism,” the scare tactic used against herbs is often the issue of safety.
Of course, determining how to use herbs safely is a real issue, but it is one that can also be used for political ends. This does not necessarily imply a conscious political effort, but is more often simply the expression of the belief system in which the person is embedded. “Expert status” is a mutually supportive addiction that often inyokes a reflexive response. In mainstream culture, herbal expertise is rarely recognized as existing outside the walls of academia. However, if becoming an expert involves some form of culturally recognized academic achievement, how is this to be obtained in a culture that offers no Ph.D. programs in herbalism? By this I mean clinical phytotherapy programs, not ethnobotany. In a cultural milieu in which the knowledge, skills, and experience of the herbalist are devalued or ignored, how can our expertise be recognized?

How do we create a bridge between the knowledge and experience of a well-trained phyto therapist and the unique perspective and language of the scientific method? The plethora of issues concerning safety and toxicity is an area in which am trying to find meaningful corollaries,
and, as a simple-minded herbalist, make sense of this. In this time of rapid change, I want to sound a plea for traditional protocols. Here are some provisional guidelines.

Don’t jump to unwarranted conclusions about therapeutic indications and contraindica tion from research findings if:

• They are derived from animal experiments
This is partly a moral position, but there are also profound doubts about the validity of extrapolations from animal studies to human research applications. For example, does a
specific animal model actually test what the interpretation suggests it tests? Much of the
basic research on phytoestrogens uses a laborator test that assesses estrogen induced thickening of the uterine wall in rodents. This is all well and good but to use resuLts or such tests to arrive at human therapeutic protocols is absurd The mere fact that rodents do not menstruate calls into question the validity of such extrapolation This alone should inform us that the actual situation is much more complex than the one the test is designed to investigate
• The studies utilized isolated constituents not whole plants Even if we don’t invoke the biochemical synergy that may play a role in the actions of whole plant preparations, the stupendous number of potential interactions should make us wary of easy conclusions.
• They disagree with established phytothera peuti experience and protocols. As our modality has inherent value (a statement that is obviously a reflection of my own belief system and not statistical data) we should not reject protocols or plants because of Lack of research on their mechanisms of action.


[1] Tyler VE. The Honest Herbal: A Sensible Guide to the Use of Herbs and Related Remedies. Binghamton, NY The Haworth Press, 1993.

[2] Röder E. Medicinal plants in Europe containing pyrrolizidine alkaloids. Pharmazie 1995; 50:83—98.

[3] Mann J, Davidson RS, Hobbs JB, et al. Natural Products: Their Chemistiy and Biokgical Significance. Harlow, UK: Longman Scientific & Technical; New York: Wiley, 1994.

[4] Rothschild M, Edgar JA. Pyrrolizidine alkaloids from Senecio vulgaris sequestered and stored by Danaus plexippus. Journal of Zoo and Wildlife Medicine 1978; 186:347—9.

[5] Schneider D, Boppré M, Schneider H, et al. A pheromone precursor and its uptake in male Danaus butterflies. Journal of Comparative Physiology 1975; 97:245—56.

[6] Rode D. Comfrey toxicity revisited. Trends in Pharmalogical Sciences 2002; 23:497—9.

[7] Beers MH, Berkow R, eds. The Merck Manual of Diagnosis and Therapy. Whitehouse Station, NJ: Merck Research Laboratories, 1997.

[8] Rode D. Comfrey toxicity revisited. Trends in Pharmalogical Sciences 2002; 23:497—9.

[9] Stickel F, Seitz HK. The efficacy and safety of comfrey. Public Health Nutrition 2000; 3:501—8.

[10] Cheeke PR. Pyrrolizidine alkaloids. In: Natural Toxicants in Feeds, Forages, and Poisonous Plants, 2nd edition. Danville, IL: Interstate Publishers, 1998.

[11] Ibid.

[12] Rode D. Comfrey toxicity revisited. Trends in Pharmalogical Sciences 2002; 23:497—9.

[13] Pierson ML, et al. Resistance of the rabbit to dietary pyrrolizidine (Senecio) alkaloid. Research Communications in Chemical Pathology and Pharmacology 1977; 16:561—4

[14] Stickel F, Seitz HK. The efficacy and safety of comfrey. Public Health Nutrition 2000; 3:50 1—8.

[15] Muetterlein R, Aimold CG. Investigations concerning the content and the pattern of pyrrolizidine alkaloids in Symphytum officale L. (comfrey). Pharmazeutische Zeitung 1993; 138:119—25

[16] Cheeke PR. Pyrrolizidine alkaloids. In: Natural Toxicants in Feeds, Forages, and Poisonous Plants, 2nd edition. Danville, IL: Interstate Publishers 1998

[17] Schoental, R. Toxicology and carcinogenic action of pyrrolizidine alkaloids. Cancer Research 1968; 28:2237—46.


About Me Classes and Workshops Seeds and Stems Click here to e-mail