Boneset Safety: Pyrrolizidine Alkaloids and Liver Risk

This is the most important page in this set, and the reason is an asymmetry. Boneset's benefit claims sit on tradition and cell-culture work, with no adequate human trial of any kind. Its principal hazard sits on a compound class whose human liver toxicity is established, characterised at the molecular level, and documented in named poisoning outbreaks. Established harm mechanism, absent benefit evidence. That is the honest shape of the plant, and it is why this article exists rather than a paragraph of hedged cautions at the bottom of a benefits page.

To be clear about what is and is not being claimed: boneset is not white snakeroot and it is not comfrey root. There is no documented epidemic of boneset hepatotoxicity, and the alkaloid levels reported in the plant are described in the analytical literature as low. But “low” is not “absent,” the toxicity of this class is cumulative, and the absence of case reports for a minor, uncommonly used herb reflects absent surveillance at least as much as it reflects safety. Those two facts together are the whole argument of this page.

What Pyrrolizidine Alkaloids Are
The Liver Makes the Poison
Hepatic Sinusoidal Obstruction Syndrome
Cumulative Dose, Delayed Damage
What Is Actually Known About Boneset’s Alkaloids
The Numbers This Page Will Not Give You
Identity: Eupatorium, Eutrochium, Ageratina
Why the Adulteration Defence Backfires
Genotoxicity and the Cancer Question
What Preparation Does and Does Not Change
A Duration Limit, Not a Dose Limit
Who Must Not Use It At All
Interactions, and What “No Known Interactions” Means
The Non-Alkaloid Risks
Warning Signs and Which Tests Matter
Evidence Ledger: Harm Tier Against Benefit Tier
Key Research Papers
Connections

What Pyrrolizidine Alkaloids Are

Pyrrolizidine alkaloids are a large family of plant secondary metabolites — well over six hundred structures are known — produced by species scattered across several unrelated families. Three matter most: the Boraginaceae (comfrey, borage, alkanet, viper's bugloss), the Asteraceae tribes Senecioneae and Eupatorieae (ragwort, groundsel, coltsfoot, butterbur, and the Eupatorium group), and the Fabaceae genus Crotalaria. They exist as chemical defences against herbivores.

The critical structural distinction is between saturated and unsaturated pyrrolizidine alkaloids. A pyrrolizidine alkaloid with a saturated core is comparatively innocuous. It is the ones with a 1,2-unsaturated necine base — the group referred to in the literature as dehydropyrrolizidine alkaloids, or 1,2-unsaturated PAs — that carry the hepatotoxicity, and it is this subgroup that has been reported in analyses of Eupatorium perfoliatum and related species. The distinction is not a technicality: it is the difference between a compound that is a curiosity and one that is a liver toxin, and a source that says only “contains pyrrolizidine alkaloids” without specifying which type has left out the important part.

They also occur as free bases and as N-oxides, which are more water-soluble and therefore extract more readily into a hot aqueous infusion. N-oxides are reduced back to the parent alkaloid in the gut and body. A tea is not therefore protective; if anything the aqueous chemistry favours the more extractable form.

The Liver Makes the Poison

The most counter-intuitive and most important fact about this class is that the alkaloids as eaten are not the toxic agents. They are bioactivated — the body converts them into the reactive species that does the damage, and the organ that performs the conversion is the organ that is harmed.

The sequence, as characterised in the toxicology literature:

The alkaloid is absorbed from the gut and travels by the portal vein directly to the liver — the first-pass route, delivering the highest concentration to the most vulnerable tissue.
Hepatic cytochrome P450 enzymes, principally the CYP3A subfamily, oxidise the 1,2-unsaturated necine base to a dehydropyrrolizidine (pyrrolic) ester.
That metabolite is a highly reactive electrophile with a very short half-life. It alkylates whatever nucleophile is nearest — protein thiols, and DNA.
Because it is so short-lived, it acts essentially where it is made: in and immediately around the hepatocytes and, critically, the endothelial cells lining the hepatic sinusoids, the smallest vessels of the liver.

Two consequences follow, and both are important for understanding why the risk profile looks the way it does.

First, the toxicity is local and organ-specific by mechanism, not by dose distribution. This is not a poison that circulates and happens to reach the liver; the liver manufactures it on site. That is also why lung and, less often, other tissues can be involved — there is some extrahepatic activation capacity — but hepatic injury dominates.

Second, anything that changes CYP3A activity plausibly changes the risk. That is a mechanistic inference rather than a demonstrated interaction, and it has not been studied for boneset, but it is the reason a herb of this class sits uneasily alongside the very large number of drugs, foods and supplements that induce or inhibit CYP3A4.

Hepatic Sinusoidal Obstruction Syndrome

The characteristic injury is hepatic sinusoidal obstruction syndrome (SOS), historically called hepatic veno-occlusive disease. It is a distinctive lesion and it is not the same thing as ordinary drug-induced hepatitis.

The reactive metabolite damages sinusoidal endothelial cells. Those cells detach and embolise downstream, obstructing the small hepatic venules. Blood cannot leave the liver normally, so pressure rises upstream: the liver becomes engorged and tender, fluid accumulates in the abdomen as ascites, and over time the obstruction drives fibrosis, then cirrhosis, then liver failure. The clinical picture is therefore one of post-sinusoidal portal hypertension — painful hepatomegaly, ascites, weight gain from fluid, and jaundice — rather than the transaminase spike of a classic hepatitis.

The human evidence for this is not theoretical. It comes from repeated mass-poisoning episodes in which contaminated grain or medicinal plant material was consumed, from clusters associated with traditional herbal remedies, and from well-described case series — including a substantial Chinese literature on SOS following consumption of Gynura segetum, a PA-containing plant used medicinally. The lesion, its cause, and its mechanism are as well established as anything in herbal toxicology. Outcomes range from recovery through chronic liver disease to death.

Nothing in that paragraph is a boneset case report. It is a description of what the compound class does when enough of it is consumed, established in humans, in named outbreaks, with a characterised mechanism. That is the tier of evidence boneset's hazard sits on — class-level and firm — against a benefit tier of tradition and cell culture.

Cumulative Dose, Delayed Damage

The feature that makes this class genuinely awkward for a casual herbal user is that injury accumulates and appears late.

There is no acute warning signal at low exposure. A cup of tea containing a trace of these alkaloids produces no symptom, no taste cue, and no laboratory abnormality. The absence of an immediate reaction carries no information about whether damage occurred.
Endothelial loss is not fully repaired. Repeated small insults deplete a finite population of cells, and the fibrotic response to obstruction is itself progressive.
Presentation can be months or years after exposure begins, by which point the causal link is easy to miss — and easy for a clinician to miss too, since nobody volunteers a herbal tea habit as relevant history.
Susceptibility varies. Bioactivation depends on individual CYP3A activity, on competing detoxification capacity including glutathione status, and on liver health at baseline. Two people with the same intake do not carry the same risk, and there is no way to know in advance which one you are.

The practical translation is uncomfortable but simple: a person cannot tell from how they feel whether their boneset intake has been harmless. That is why the guidance for this class of herb is framed in terms of total lifetime exposure and duration rather than a per-cup safe amount.

What Is Actually Known About Boneset’s Alkaloids

Boneset was for a long time described in herbal literature as PA-free or of unknown PA status, and older texts sometimes still say so. That changed with targeted analytical work. Chemists including Betz, Colegate, Upton, Gardner and colleagues applied modern methods to boneset and related species and reported detection of dehydropyrrolizidine alkaloids in Eupatorium perfoliatum. The finding is that the toxic subclass is present, at levels the analytical literature describes as low.

Several honest qualifications belong with that:

The number of published analyses is small. This is not a well-surveyed plant. A handful of studies on a handful of samples is what exists.
Plant alkaloid content is intrinsically variable. Across the PA-bearing species generally, content varies with plant part, growth stage, season, geography, soil, and individual plant genetics, sometimes by more than an order of magnitude. There is no reason to expect boneset to be an exception, and no published survey establishing its range.
No public survey of commercial product exists. What is in a specific bag of dried boneset or a specific bottle of tincture on a shelf is unknown, and there is no routine testing regime that would answer the question for a retail buyer.
“Low” interacts badly with “cumulative.” For a toxin whose risk is a function of total intake over time, a low per-serving content shifts the problem from acute to chronic. It makes single use less concerning and habitual use more so — it does not remove the issue, it relocates it.

The Numbers This Page Will Not Give You

It would be easy to fill this section with confident figures — micrograms per gram of herb, micrograms per cup, a daily limit, a lethal dose, a regulatory threshold. This page does not, and stating the refusal is more useful than supplying a guess.

No micrograms-per-gram figure for boneset is asserted here. The published analyses are few, the sample base is narrow, and biological variability across PA plants is large. A single reported figure quoted as though it characterised the species would be misleading.
No micrograms-per-cup figure is asserted. That would require knowing the content of the herb used, the extraction efficiency for each alkaloid and its N-oxide under real brewing conditions, and the volume drunk. Two of those three are unpublished for boneset.
No specific regulatory threshold is quoted. Several authorities have set exposure limits or product limits for pyrrolizidine alkaloids, the values differ between jurisdictions and between food and medicinal categories, and they are revised. Quoting one figure without its jurisdiction, category, date and basis would produce false precision. The general position is what matters and it is consistent: exposure to 1,2-unsaturated PAs should be kept as low as reasonably achievable, and these compounds are not treated as having a safe intake for chronic consumption. Check a current primary source for figures.
No lethal or threshold dose is asserted. Human dose-response for this class comes from poisoning outbreaks with poorly quantified exposure. There is no reliable number.

A page that refuses these numbers is more trustworthy than one that invents them, and a reader who needs an actual figure needs a current regulatory document, not a health website.

Identity: Eupatorium, Eutrochium, Ageratina

Boneset sits inside a genus that has been taken apart, and the fallout is a live safety problem rather than a filing curiosity.

Eupatorium perfoliatum — boneset. The plant this page is about. Its diagnostic feature is the perfoliate leaf arrangement: opposite leaves fused at the base so the hairy stem appears to pass straight through a single continuous leaf, repeatedly, up the stem. Flat-topped clusters of small dull-white flower heads in late summer. Wet ground.
Eutrochium — the Joe-Pye weeds. Formerly Eupatorium, now a separate genus. Tall wetland plants with leaves in whorls of three to five (not fused pairs) and dusky mauve-pink domed flower clusters. A different herb with a different traditional use; boneset directions do not transfer to it, and PA content in this group should not be assumed to match boneset's in either direction.
Ageratina altissima — white snakeroot. Formerly Eupatorium rugosum. This is the dangerous one. It grows in the same region, flowers small and white at the same season, and killed frontier families through “milk sickness” — its toxin, tremetol, passing into the milk of cattle that grazed it. Nancy Hanks Lincoln is traditionally held to have died of it. Its leaves are ordinary, heart-shaped and stalked, clearly separate from the stem. Note that its hazard is a different toxin entirely, not a pyrrolizidine alkaloid.

The practical rule: the fused, stem-piercing leaves are the check, and they are unmistakable once you have seen them. If a specimen does not show them, it is not boneset. And if you are looking at dried, chopped, bagged herb, you cannot perform this check at all — which is the subject of the next section.

Why the Adulteration Defence Backfires

A familiar argument is made in defence of PA-containing herbs, and it is worth engaging with properly because it is partly correct. The argument runs: historical poisonings attributed to a given herb were really caused by a substituted species — germander sold as skullcap, a misidentified composite sold as coltsfoot, ragwort or groundsel contaminating a harvest — so the plant itself is exonerated. People who raise this point are doing real work; adulteration and misidentification genuinely have caused harm that was then misattributed, and accurate reattribution matters.

But notice what the argument establishes and what it does not. It does not establish that the herb in your jar is the herb on the label. In most historical incidents the implicated material was never analysed, so the substitution claim is itself often an inference. And the argument's own logic runs the wrong way for the consumer: if the hazard is that you cannot tell which plant you actually bought, then unverified loose herb is more dangerous, not less. The standard exculpation of the species turns into the strongest available argument against buying unauthenticated material.

For boneset this is sharpened by three specifics. Its genus has been split, so nomenclature in trade and in older literature is inconsistent. It has a genuinely poisonous look-alike sharing its habitat, season and flower colour. And the one reliable field character — the perfoliate leaves — is destroyed by drying and chopping, which is how the herb is sold. Wild-harvesting boneset without confident identification of the living plant is the highest-risk way to obtain it, and buying it dried from an unverified source is the second.

Genotoxicity and the Cancer Question

The reactive metabolite alkylates DNA as well as protein, forming characteristic DNA adducts. The 1,2-unsaturated pyrrolizidine alkaloids are accordingly treated as genotoxic carcinogens in the toxicology literature, with tumour induction demonstrated in rodents, and this is one of the main reasons regulators frame their advice as “as low as reasonably achievable” rather than setting a tolerable daily intake in the usual way.

Two honest points about how much weight this should carry for a reader.

It should be taken seriously in kind: a genotoxic mechanism has no threshold in principle, which is exactly why chronic low-level intake is the pattern of concern and why “the levels are low” is not a complete answer.

It should not be inflated in degree: rodent carcinogenicity was generally established at exposures far above what a person drinking an occasional herbal tea would encounter, and human cancer attribution for dietary PA exposure remains a matter of inference and modelling rather than direct epidemiology. Nobody has shown that boneset tea causes cancer in humans, and this page does not claim it. What the genotoxicity does is remove the option of arguing that a small chronic exposure is definitionally harmless.

What Preparation Does and Does Not Change

Several plausible-sounding manoeuvres are believed to reduce PA exposure. Most do not work.

Boiling or long steeping does not destroy them. These alkaloids are thermally stable at kitchen temperatures. Hot water is an extraction method, not a detoxification.
Water extraction is not selective in your favour. The N-oxides are the more water-soluble forms, so an aqueous infusion pulls them out efficiently — and they are reduced to the parent alkaloids in the body.
Tinctures concentrate. An alcoholic extract is a concentration step, and a dropper dose of tincture is not comparable to a cup of tea in constituent delivery.
Drying and storage do not reliably eliminate them. Content can shift, but not in a controlled or predictable direction you can depend on.
“PA-free” processing is a real technology, applied to other herbs. Certified PA-depleted products exist for butterbur and, in some markets, for comfrey and coltsfoot, produced by extraction processes designed to remove the alkaloids and verified analytically. This site is not aware of any comparable certified PA-depleted boneset product. If one existed, the certification and the analysis would be the thing to check — not the label claim.

A Duration Limit, Not a Dose Limit

This is the single most useful reframing on the page. For most herbal cautions, the relevant limit is a dose: do not exceed X per day. For a cumulative genotoxic hepatotoxin, the relevant limit is a duration, because the quantity that matters is the integral of intake over time.

That means the risk pattern for boneset is the opposite of the intuitive one:

A cup or two of correctly identified boneset tea during a single feverish illness is a small exposure, and this page does not claim it is dangerous.
A daily cup as a “winter immune tonic” through a whole season is a fundamentally different exposure, and it is the one the mechanism argues against.
Repeated seasonal courses year after year accumulate, and nothing resets between them.
Stacking PA-containing herbs is the worst case — boneset with comfrey tea, coltsfoot, borage, butterbur or a comfrey-containing blend. The exposures add, and a reader tracking each herb separately will underestimate the total.

This is also precisely why the misreading of the plant's name matters. A reader who thinks boneset is for bones is reaching for it for a chronic musculoskeletal complaint — taken daily, for months, exactly the pattern the alkaloids make unwise — for an indication the plant has no traditional or scientific claim to whatsoever. The etymology is not trivia; it points readers at the worst available use pattern. See Boneset for Fever and Influenza, and Where the Name Comes From.

Who Must Not Use It At All

For these groups the position is not “use cautiously.” It is do not use.

Pregnancy. Pyrrolizidine alkaloids cross the placenta, and the classical PA literature includes fetal and neonatal liver injury following maternal exposure. Fetal metabolic and detoxification capacity is not adult capacity. This is an absolute contraindication.
Breastfeeding. These alkaloids pass into milk — that transfer is the entire mechanism of “milk sickness” in the related toxic plant, and it is documented for PAs in general. A nursing infant has no way to refuse the dose.
Infants and children. Reported PA poisonings have consistently fallen hardest on children, whose smaller mass and immature hepatic handling both work against them.
Any existing liver disease. Hepatitis of any cause, fatty liver, cirrhosis, prior drug-induced liver injury, alcohol-related liver disease. Reduced reserve plus a hepatotoxin is a bad combination.
Anyone taking hepatotoxic or hepatically metabolised medication, or drugs known to induce CYP3A4 — on mechanistic grounds, since increased bioactivation is the plausible direction of the interaction.
Long-term or habitual use by anyone. There is no population for whom chronic boneset intake has an established benefit to weigh against this.

Additionally, and separately from the alkaloids: anyone with a known Asteraceae allergy should avoid it, and anyone using other PA-containing herbs should count the total rather than each item.

Interactions, and What “No Known Interactions” Means

Sources sometimes state that boneset has no known drug interactions. That phrase needs translating, because it means the opposite of what a reader assumes.

“No known interactions” here means no interaction study exists. Nobody has administered boneset to human volunteers alongside a probe drug and measured anything. That is absent data, not a clean bill of health, and it is a stated finding of this review rather than an omission from it.

What can be said on mechanistic grounds, clearly labelled as inference rather than demonstration:

CYP3A4 modulators are the theoretical concern, because CYP3A-mediated oxidation is the activating step. An inducer would plausibly increase bioactivation; an inhibitor would plausibly decrease it. Neither has been shown for boneset.
Additive hepatic burden with any hepatotoxic drug or supplement is the most straightforward concern and needs no interaction study to be plausible.
Sesquiterpene lactones are thiol-reactive, which raises a general question about glutathione-dependent detoxification capacity being drawn on from two directions at once.
Absent surveillance cuts a specific way here. Boneset is used by people with febrile illness who may also be taking antipyretics, including paracetamol/acetaminophen — a drug whose own toxicity is hepatic and glutathione-dependent. No study has looked at this combination. The asymmetry of consequences favours caution even with nothing documented.

The Non-Alkaloid Risks

Three further issues are unrelated to the alkaloids and are more likely to be what a user actually notices.

Emesis and purging. This is boneset's most reliable pharmacological effect and the traditional literature is unanimous about it: a large or strong dose causes nausea, vomiting and loose stools, and the herb was sometimes used deliberately as a purge. Practically, this means the margin between an unpleasantly strong tea and a gut-emptying one is narrow, and an over-brewed cup can make a person who is already febrile and dehydrated considerably worse. Vomiting on top of a fever is not a trivial adverse event.
Asteraceae allergy and cross-reactivity. As a member of the daisy family, boneset can provoke reactions in people sensitised to ragweed, chrysanthemum, marigold, chamomile, feverfew, arnica or echinacea. Reactions range from oral and respiratory symptoms to systemic allergy.
Contact dermatitis. Sesquiterpene lactones are among the best-characterised plant contact allergens, and Compositae dermatitis from handling or applying plants of this family is well described in the dermatology literature. This is a handling risk for anyone harvesting or processing the herb, independent of drinking it.

Warning Signs and Which Tests Matter

Not medical advice, and not a screening protocol. These are the patterns that should prompt someone to seek assessment and to mention any herbal use, including teas, without being asked.

Symptoms pointing at the liver:

Yellowing of the whites of the eyes or the skin (jaundice).
Dark, tea-coloured urine, or unusually pale stools.
Pain, fullness or tenderness under the right ribs.
Abdominal swelling or rapid weight gain from fluid (ascites) — particularly characteristic of sinusoidal obstruction rather than ordinary hepatitis.
Persistent nausea, loss of appetite, and unexplained fatigue.
Itching without a rash.
Easy bruising or bleeding, which suggests impaired clotting factor synthesis.
Confusion or unusual drowsiness — a late and serious sign requiring urgent care.

What a clinician would look at: a liver panel including ALT, AST, alkaline phosphatase, GGT and bilirubin; albumin and prothrombin time or INR as measures of synthetic function; a platelet count; and imaging with Doppler assessment of hepatic venous flow if sinusoidal obstruction is suspected, since the diagnosis is vascular rather than purely biochemical. The single most valuable thing a patient can contribute is an honest, complete list of every herb, tea, tincture and supplement taken and for how long — herbal exposure is routinely missed because it is not asked about and not volunteered.

Evidence Ledger: Harm Tier Against Benefit Tier

Placing the two side by side makes the asymmetry visible rather than argued.

1,2-unsaturated PAs cause hepatic sinusoidal obstruction syndrome in humans — tier: established, human, mechanism characterised, outbreaks documented.
1,2-unsaturated PAs are genotoxic and carcinogenic in animals — tier: established in animals, mechanism characterised, human extrapolation by inference.
Dehydro-PAs are present in Eupatorium perfoliatum — tier: analytically demonstrated, small number of studies, quantity not surveyed.
Boneset is emetic in large doses — tier: consistent traditional observation across two centuries of clinical literature.
Boneset relieves fever and shortens influenza — tier: tradition only; no controlled human trial exists.
Boneset extract interferes with influenza A attachment — tier: single in-vitro study, hydroalcoholic extract, no animal or human follow-up.
Boneset has anti-inflammatory activity — tier: in vitro, cell models, replicated within one research group.
Boneset stimulates appetite as a bitter — tier: tradition; the closest human bitter-receptor physiology points the other way.
Boneset sets or heals bones — tier: no traditional claim, no mechanism, no research. The name is a misreading.

Read top to bottom, the best-supported statements about boneset are all statements about harm and chemistry. That is not an argument that the plant is dangerous in a single feverish week; it is an argument that anyone contemplating regular use is trading a firmly established hazard against an entirely unestablished benefit.

Key Research Papers

Given as PubMed topic searches so the links keep working. No specific alkaloid content figure, threshold or dose is asserted anywhere on this page; readers needing numbers should consult a current primary regulatory source.

Fu PP, Xia Q, Lin G, Chou MW. Pyrrolizidine alkaloids — genotoxicity, metabolism enzymes, metabolic activation and mechanisms. Drug Metabolism Reviews, 2004. The standard account of bioactivation and DNA adduct formation. Find on PubMed.
Ruan J, Yang M, Fu P, Ye Y, Lin G. Metabolic activation of pyrrolizidine alkaloids: insights into the structural and enzymatic basis. Chemical Research in Toxicology, 2014. Why the 1,2-unsaturated structure is the one that matters. Find on PubMed.
Edgar JA, Molyneux RJ, Colegate SM. Pyrrolizidine alkaloids: potential role in the etiology of cancers and other chronic disease. Chemical Research in Toxicology, 2015. The argument for taking chronic low-level dietary exposure seriously. Find on PubMed.
Stegelmeier BL and colleagues. Pyrrolizidine alkaloid plants, metabolism and toxicity. Journal of Natural Toxins, 1999. Broad review covering the plant sources and species differences in susceptibility. Find on PubMed.
Chen Z, Huo JR. Hepatic veno-occlusive disease associated with toxicity of pyrrolizidine alkaloids in herbal preparations. Netherlands Journal of Medicine, 2010. Human cases arising specifically from herbal use. Find on PubMed.
Lin G and colleagues. Hepatic sinusoidal obstruction syndrome associated with consumption of Gynura segetum. Journal of Hepatology, 2011. A large, well-characterised human series from medicinal-plant PA exposure. Find on PubMed.
DeLeve LD, Shulman HM, McDonald GB. Toxic injury to hepatic sinusoids: sinusoidal obstruction syndrome. Seminars in Liver Disease, 2002. The pathophysiology of the lesion itself. Find on PubMed.
Prakash AS, Pereira TN, Reilly PE, Seawright AA. Pyrrolizidine alkaloids in human diet. Mutation Research, 1999. Dietary exposure routes, including herbal teas, honey and contaminated grain. Find on PubMed.
Pyrrolizidine alkaloids in herbal teas and herbal drugs: analytical surveys. The work showing that measured content in retail herbal products is variable and frequently unexpected — the reason unverified material is the core problem. Search PubMed on this topic.
Dehydropyrrolizidine alkaloids in boneset and related species. Analytical detection in E. perfoliatum by Betz, Colegate, Upton, Gardner and colleagues — the boneset-specific finding this page rests on. Search PubMed on this topic.
Roeder E. Medicinal plants in Europe containing pyrrolizidine alkaloids. Die Pharmazie, 1995. The classic inventory of which medicinal species carry them — useful for counting total exposure across a herbal regimen. Find on PubMed.
Transplacental and lactational transfer of pyrrolizidine alkaloids. The basis for the absolute contraindications in pregnancy and breastfeeding. Search PubMed on this topic.
White snakeroot, tremetol and milk sickness. The look-alike, whose hazard is a different toxin entirely. Search PubMed on this topic.
Compositae contact dermatitis and sesquiterpene lactone sensitisation. The handling risk shared across daisy-family herbs. Search PubMed on this topic.

Back to Table of Contents

Connections

Back to Table of Contents

A note on what this page is

This is educational information about a class of plant toxin and a herb that contains it. It is not medical advice, not a diagnosis, and not a substitute for assessment by a qualified healthcare professional. If you think you may have been harmed by a herbal product, stop taking it and seek medical advice, taking the product and its label with you. If you are pregnant, breastfeeding, giving anything to a child, or have any liver condition, do not use boneset.