Medicinal Plants by Infusion: What the Science Really Shows

Nature has a pharmacy. But we have to be honest.

A large share of modern medications originally goes back to plant substances. Aspirin comes from willow bark. Morphine from the poppy. Taxol from the yew tree. Artemisinin against malaria from Artemisia annua. These substances act, strongly, measurably, sometimes life-savingly.

That means: plant medicine is no esoterics. Biologically active substances from plants can act at the cellular level. Studies document this. For some applications, very well.

But it also means something I want to tell you directly: the active substance, the dosage form and the route of administration decide everything. A study on oral chamomile tea is no evidence for a chamomile infusion into the bloodstream. That sounds self-evident. But it is a difference that regularly blurs in the literature on plant infusions.

In this article I will show you what the study data actually support, with the plant substances we use in integrative medicine. Honestly. With correct studies. With clear limits.

The most important point upfront Some plant active substances have strong clinical evidence for intravenous use. Others have good oral evidence, but none for IV. Yet others have only weak or missing study data for both routes. I will distinguish this, because it matters for your decision.

The three evidence levels in plant infusion therapy

Before we discuss individual substances: it helps to distinguish three evidence levels. They are often mixed up in the popular literature.

Tier 1, strong evidence IV silibinin in death-cap mushroom poisoning, as a recognized emergency treatment in hospital. Oral: milk thistle (NAFLD), artichoke (liver enzymes), peppermint oil (IBS). Several controlled studies, reproducible results.

Tier 2, moderate evidence Oral: ginkgo (mild to moderate dementia, weak recommendation in the German S3 guideline), curcumin (blood pressure, about 2 mmHg), chamomile (sleep quality). Effects real, but clinically modest or methodologically limited.

Tier 3, no IV evidence IV chamomile, IV lemon balm, IV peppermint: not a single published clinical trial. IV curcumin: not approved, safety concerns. IV glutathione: one randomised pilot trial with no advantage over placebo. Oral studies do not allow IV conclusions.

1. Milk thistle (Silybum marianum): the substance with the clearest IV indication

🌿

Silibinin, the substance with a recognized IV emergency indication

IV: recognized emergency indication Liver, detoxification, cell protection

Milk thistle is the plant for which intravenous administration is most clearly justified. The active substance is called silibinin, the main component of silymarin. For the intravenous form, marketed as Legalon SIL, there is one clearly defined, clinically recognized emergency indication.

I name it deliberately. A text that hides active substances protects nobody, it only takes away your ability to check what you have read. What matters is the frame around the name, and it reads: Legalon SIL is prescription-only, it is a hospital antidote, it is given under monitoring, and it has no place in an outpatient practice. Further fields of application are discussed in the specialist literature, and I do not address those publicly here. Whether any of this is relevant for you at all belongs in a medical consultation.

One clarification, because it is a matter of life and death: if you suspect mushroom poisoning, call the emergency number 112 or the poison control centre immediately and go to a hospital. This is not a case for a practice and not a case for an appointment. Every hour counts. I mention silibinin here only because this emergency shows how powerful a plant substance can be.

IV silibinin in death-cap mushroom poisoning

In poisoning with the death cap mushroom (Amanita phalloides), intravenous silibinin (Legalon SIL) is used clinically. This treatment belongs exclusively in the emergency department. As a mechanism it is discussed that silibinin may block uptake transporters of the liver cell, among them OATP1B3, through which alpha-amanitin enters the cell. That could interrupt the circulation of the toxin between gut and liver.

How solid is this? Controlled trials do not exist for ethical reasons, since no placebo group can be formed in a life-threatening poisoning. An analysis of almost 1,500 documented cases reports a mortality below 10 percent under silibinin, compared with more than 20 percent under penicillin. That paper comes from authors linked to the manufacturer, which needs saying. An independent ten-year analysis of Californian poison centre data found no advantage, which the authors explain by the fact that it is precisely the most severely poisoned patients who receive the drug. So the use is recognized. In the strict sense it is not proven.

Mengs U, Pohl RT, Mitchell T. Legalon SIL: the antidote of choice in patients with acute hepatotoxicity from amatoxin poisoning. Curr Pharm Biotechnol. 2012;13(10):1964–1970. DOI: 10.2174/138920112802273353. Albertson TE et al. Clin Toxicol (Phila). 2023;61(11):974–981. DOI: 10.1080/15563650.2023.2276674. Metadata checked via PubMed.

What belongs alongside a substance name: Legalon SIL is prescription-only and in several countries not even available. As with any infusion, hypersensitivity reactions are possible, and milk thistle belongs to the daisy family, so an allergy to that plant family matters. I deliberately give no dosages here. In an emergency the treating hospital team sets them, based on weight, course and laboratory values.

What the popular literature often overstates

For intravenous silibinin there is a small, open dose-finding study from 2008 in 36 patients suggesting an antiviral effect in chronic hepatitis C. There was no control group, and the strongest effect comes from a subgroup of nine patients. That study predates today's direct-acting antivirals, with which hepatitis C can as a rule resolve completely. As a treatment option, silibinin no longer plays a role there. In the largest study so far on oral silymarin in hepatitis C (Fried MW et al. 2012, JAMA, n=154) there was likewise no advantage over placebo. A single study is not a definitive proof of ineffectiveness either.

For daily practice: a meta-analysis of 55 randomised studies with 3,545 patients found a reduction of AST and ALT for oral silymarin, most clearly in fatty liver and viral hepatitis. In drug-induced and alcohol-related liver damage there was no significant effect, nor at a BMI of 30 or above. A Cochrane review from 2025 with 17 randomised trials and 2,069 participants concludes more cautiously that benefits and harms remain unclear, at low to very low certainty of evidence. Both belong side by side.

2. Artichoke (Cynara scolymus): good for liver and lipid metabolism, but no IV candidate

🌱

Artichoke extract, oral evidence, no IV benefit documented

Oral: moderate evidence IV: no studies

Artichoke extract is one of the best-studied liver plants in phytotherapy. Its bitter compounds, above all cynarin and luteolin, may stimulate bile secretion, may favourably influence hepatic lipid metabolism and may protect liver cells from oxidative stress. Precisely for that reason caution is needed if gallstones are known or the bile ducts are narrowed. Increased bile secretion can then trigger colic. In that case artichoke does not belong in self-medication. With an allergy to composite plants it is likewise unsuitable.

Network meta-analysis on liver enzymes 2024

Liu and colleagues analyzed in 2024 37 RCTs with 2,509 patients on natural substances and liver enzymes in NAFLD. Artichoke extract took first place for AST reduction (SUCRA 99.1 percent) among seven compared natural products.

Liu H et al. Effects of different natural products in patients with non-alcoholic fatty liver disease. A network meta-analysis of randomized controlled trials. Phytother Res. 2024;38(7):3801–3824. DOI: 10.1002/ptr.8182

Cholesterol meta-analysis, 9 RCTs

Sahebkar et al. analyzed 9 RCTs with 702 patients. Total cholesterol fell by about 17.6 mg/dL (about 6 to 7 percent), LDL by about 14.9 mg/dL (about 8 to 11 percent). Moderate, statistically significant effects, but no dramatic reductions.

Sahebkar A et al. Lipid-lowering activity of artichoke extracts: A systematic review and meta-analysis. Crit Rev Food Sci Nutr. 2018;58(15):2549–2556. DOI: 10.1080/10408398.2017.1332572

A note on diligence: the 2002 paper by Saller and colleagues is often cited for artichoke. It deals, however, with Iberogast (STW 5), a nine-herb combination preparation without artichoke. Such mix-ups are common in the phytotherapy literature.

The order that comes before everything else: elevated liver values are not a diagnosis, they are a call for work-up. Viral hepatitis, autoimmune processes, iron storage disease, alcohol and medications have to be ruled out first. Only after that does the question make sense whether artichoke or milk thistle can contribute something alongside. A plant extract is not an answer to an uninvestigated laboratory finding.

For practice: oral artichoke extract is used traditionally in phytotherapy for bloating and digestive complaints. Studies suggest that it may favourably influence liver values. That does not replace a diagnosis and it is not a medicine against liver disease. Intravenous artichoke has not been investigated in any controlled study.

3. Ginkgo biloba (EGb 761): strong mechanisms, mixed study data

🧠

EGb 761, anchored in German guidelines, IV form barely available

Oral: moderate evidence IV: historical, no longer standard

Ginkgo biloba is the most-studied plant active substance in neurology. For the standardized extract EGb 761 it is described that it may improve cerebral microcirculation, may inhibit platelet activating factor (PAF) and may act neuroprotectively. These are predominantly findings from laboratory and animal models.

A word on the name, because paraphrases are no help to you here. EGb 761 is not a brand, it is the designation of a precisely defined dry extract with a fixed content of flavonoids and terpene lactones. In Germany it is marketed above all as Tebonin, alongside generics containing the same extract. I write this down because "ginkgo" on a shelf can mean almost anything, from an arbitrary food supplement to an authorised medicine, and because the studies discussed below refer exclusively to this one standardized extract. So do not transfer the results to just any ginkgo product.

On its status: these preparations are authorised medicines and pharmacy-only, but not prescription-only. Because they are available without a prescription, German statutory health insurance as a rule does not cover them for adults. Available without a prescription does not mean harmless, though, see the box below.

German S3 Dementia Guideline 2023, recommendation on EGb 761

The current German S3 Dementia Guideline of DGPPN and DGN suggests using EGb 761 at a dose of 240 mg daily for cognition and activities of daily living in mild to moderate Alzheimer's dementia or vascular dementia with non-psychotic behavioural symptoms. The strength of recommendation is "weak in favour", the certainty of evidence is rated moderate. A weak recommendation means: it is an option, not a standard, and the effect is limited.

S3 Dementia Guideline (S3-Leitlinie Demenzen). DGPPN and DGN. Version 4.0, 28 November 2023. AWMF register no. 038-013, recommendation 65.

Cochrane Review tinnitus 2022

A Cochrane review from 2022 included 12 studies with 1,915 participants on Ginkgo biloba, in various preparations, not specifically EGb 761. For the core outcome on tinnitus burden, only two studies with 85 participants could be pooled. No advantage over placebo was found there, but the authors rate the certainty of this evidence as very low.

Sereda M et al. Ginkgo biloba for tinnitus. Cochrane Database Syst Rev. 2022;11(11):CD013514. DOI: 10.1002/14651858.CD013514.pub2

Important before you take ginkgo on your own: ginkgo may affect blood clotting. If you take anticoagulants or aspirin, if surgery is planned or if you tend to bleed, this needs to be discussed beforehand. A meta-analysis of coagulation parameters cited in the guideline context found no sign of an increased bleeding risk, yet product information and individual case reports still urge caution. A review that set controlled studies and case reports side by side concluded that EGb 761 did not measurably affect haemostasis in the controlled work, while a rare individual event still cannot be ruled out. Also reported are headache, gastrointestinal complaints, dizziness and allergic skin reactions. In the pooled analysis of the licensing trials, adverse events and discontinuation rates did not differ noticeably from placebo, though the manufacturer was involved in that analysis. And with an uninvestigated memory problem the rule is: work-up first, treatment second. Whether a ginkgo preparation suits you belongs in a medical consultation, not on a shelf.

On the IV form: the former German injectable formulation is no longer in regular distribution. In China, ginkgo injections are used in stroke, but the study quality is overall weak.

4. Curcumin (turmeric): a real anti-inflammatory, with important limitations

🧡

Curcumin, modest but real oral effects, critical IV safety questions

Oral: moderate evidence IV: only investigational, safety concerns

Curcumin is biologically active as an anti-inflammatory. In-vitro data and preclinical studies show this consistently. In the laboratory it can inhibit NF-kappaB, modulate cytokines and act as an antioxidant. These are laboratory findings, not statements about what happens in your body.

Dose-response meta-analysis on blood pressure and vascular function 2024

Dehzad and colleagues analyzed 35 RCTs in 2024. Systolic blood pressure fell by −2.02 mmHg (95 percent CI −2.85 to −1.18), diastolic by −0.82 mmHg. Statistically significant, but clinically modest, a 5 mmHg reduction is considered the clinical threshold. The authors themselves phrase it in their title as curcumin could improve blood pressure and endothelial function.

Dehzad MJ et al. Curcumin/turmeric supplementation could improve blood pressure and endothelial function: A grade-assessed systematic review and dose-response meta-analysis of randomized controlled trials. Clin Nutr ESPEN. 2024;59:194–207. DOI: 10.1016/j.clnesp.2023.12.009

IV curcumin: a fatality and safety concerns. The FDA investigated in 2017 a fatality after IV curcumin infusion in a 30-year-old patient, presumably caused by the emulsifier PEG 40 castor oil. There is so far no approval for intravenous curcumin anywhere. I therefore do not use it. Anyone who judges this differently has to justify that for themselves. I am only describing my own standard here.

What works for curcumin? Oral, highly bioavailable formulations (liposomal, micellar, with piperine) achieve higher plasma levels. For inflammatory states and metabolic syndrome, oral curcumin can be a sensible adjunct.

A note that is often missing: the highly bioavailable curcumin preparations in particular are suspected in individual case reports of straining liver values. Piperine can also inhibit the breakdown of other medicines via CYP3A4 and P-glycoprotein, so their levels may rise. With known gallstones or an obstruction of the bile ducts, curcumin is not suitable. If you take medication regularly, this needs to be discussed beforehand.

5. Polyphenols, resveratrol, EGCG: interesting basic research, weak clinical data

🌸

Resveratrol & EGCG, biochemically fascinating, clinically not robust

Oral: weak evidence IV: no studies

Resveratrol can influence sirtuins in the laboratory, enzymes that play a role in cell metabolism. Whether that achieves anything in humans is open and is debated controversially. EGCG from green tea can dampen inflammatory pathways in cell experiments. Both are laboratory findings, not statements about your body. In humans the effects are considerably more modest.

Important finding: EGCG is possibly not the active vascular compound in tea

Lorenz and colleagues studied 50 healthy men in a randomised crossover trial: green tea significantly improved flow-mediated vasodilation, whereas isolated EGCG at the same dose of 200 mg did not. High-dose EGCG from extracts is moreover linked with liver damage.

Lorenz M et al. Tea-induced improvement of endothelial function in humans: No role for epigallocatechin gallate (EGCG). Sci Rep. 2017;7(1):2279. DOI: 10.1038/s41598-017-02384-x

No clinical study has investigated IV polyphenols for cardiovascular endpoints. Polyphenol-rich nutrition has consistent epidemiological associations with vascular health, but the mechanisms are probably more complex than the isolated single substance via infusion.

6. Chamomile, lemon balm, peppermint: effective plants, wrongly classified

🌿

Calming herbs, well-documented orally, but never investigated as IV

Oral / tea: moderate evidence IV: nonexistent

Here clarity is especially important: the German word "Infusion" means both "herbal tea" and "intravenous infusion". These are two completely different medical realities.

Central misunderstanding resolved

All clinical studies on chamomile, lemon balm and peppermint use oral application. There is not a single published clinical study on intravenous administration of these herbs in humans. For peppermint oil there is a case report of acute lung injury after intravenous injection. Oral evidence may not be extrapolated to IV therapy.

Oral evidence, what studies show

Peppermint oil (enteric-coated capsules, oral): the best data in this group. In a meta-analysis of 9 studies with 726 patients, the analysis of 5 studies with 392 patients showed global symptom improvement in IBS considerably more often than under placebo (RR 2.23; 95 percent CI 1.78 to 2.81). Adverse events occurred more often than under placebo, most commonly heartburn. For infants and small children, peppermint oil is unsuitable around the face and nose because of the risk of laryngospasm.
Chamomile (oral): in a meta-analysis of 12 RCTs, sleep quality showed a standardised mean difference of −0.73 (95 percent CI −1.23 to −0.23). On insomnia severity, the single study evaluated for it showed no significant effect.
Lemon balm (oral): for lemon balm the human study data are too thin for me to quote reliable numbers here. I therefore leave them out.
Spearmint and peppermint are not the same: anti-androgenic effects are described for spearmint (Mentha spicata), not for peppermint (Mentha piperita). The data on this come from small studies.

7. What this means for practice: my approach at ViveCura

My work in integrative medicine combines phytotherapy with orthomolecular medicine, biologically active plant substances combined with micronutrients like vitamin C, glutathione, magnesium or targeted amino acids.

A pattern I see often

Persistent exhaustion after a past infection, together with mildly elevated liver values. Before anything at all is given, the work-up comes first. Only when it is clear where the values come from does targeted support make any sense. I deliberately do not write here about the course of individual patients.

"Not every plant belongs in an infusion. But some do, and then with clear justification, not from tradition."

Intravenously I use very little, and I say openly how thin the data are in part. For IV vitamin C and IV glutathione there are so far no robust clinical studies documenting a benefit for the complaints this article is about. I therefore use them sparingly, case by case and after weighing it up, not as a standard and not as a course of treatment. These infusions are not free of risk either: for high-dose vitamin C, haemolysis in G6PD deficiency, oxalate burden on the kidney and falsified blood glucose readings are described, among others, and IV glutathione is a compounded preparation without marketing authorisation and without safety data for long-term use. The classical herbs, milk thistle, artichoke and ginkgo, I use predominantly orally, where the evidence is strongest.

8. Overview: what the evidence really shows

Active substance	Best evidence	IV status	Clinical relevance
Silibinin (milk thistle)	Oral (liver values in fatty liver), IV (death-cap mushroom poisoning)	IV: recognized emergency indication in hospital (Legalon SIL, prescription-only)	Oral moderate, IV only in emergencies
Artichoke extract	Oral (liver values, cholesterol, digestive complaints)	IV: no studies	Good orally, no IV route
Ginkgo EGb 761	Oral (mild to moderate dementia, weak guideline recommendation), extract EGb 761, e.g. Tebonin, pharmacy-only, not prescription-only	IV: historical, barely available	Moderate for cognition orally
Curcumin	Oral (blood pressure −2 mmHg, inflammation)	IV: investigational, fatality known	Low to moderate orally
Resveratrol	Human data thin, no reliable numbers	IV: no studies	Unclear
EGCG (green tea)	Oral (questionable whether EGCG is the active compound)	IV: no studies	Unclear, liver risk at high doses
Peppermint oil	Oral (IBS, meta-analysis of 9 RCTs)	IV: case report acute lung injury	High orally, never IV
Chamomile	Oral (sleep: moderate evidence)	IV: no studies	Low to moderate orally
Lemon balm	Human data thin	IV: no studies	Unclear

9. An honest closing word on plant medicine

Plants act. That is no question. Aspirin, morphine, digitalis, quinine, the history of medicine is the history of substances that were first found in plants and then understood.

But not every plant belongs in a vein. And when the word infusion is used for an herbal tea, it describes something other than an infusion into a vein. This distinction protects patients. In the long term it also strengthens trust in the possibilities of integrative medicine.

What I observe in my consultations: plant active substances, orthomolecular micronutrients and an individualized diagnostic approach can move something, when they are justified, combined and accompanied. That is an observation, not a success rate. That is the craft of integrative medicine.

Important safety notes All the plant substances discussed here are pharmacologically active. For pregnancy and breastfeeding, for children and adolescents and with impaired kidney or liver function, robust safety data are missing for most of these substances. Do not take them on your own in these situations. If you take medication regularly, interactions are possible and this needs to be discussed beforehand. If you suspect poisoning, or with signs of liver damage such as yellowing of the skin or eyes, dark urine or severe upper abdominal pain, call the emergency number 112 or go to an emergency department immediately.

Sources

Ferenci P et al. Silibinin is a potent antiviral agent in patients with chronic hepatitis C not responding to pegylated interferon/ribavirin therapy. Gastroenterology. 2008;135(5):1561–1567. Open dose-finding study without a control group. DOI: 10.1053/j.gastro.2008.07.072
Fried MW et al. Effect of silymarin (milk thistle) on liver disease in patients with chronic hepatitis C unsuccessfully treated with interferon therapy. JAMA. 2012;308(3):274–282. DOI: 10.1001/jama.2012.8265
Shahsavari K et al. Are alterations needed in Silybum marianum (Silymarin) administration practices? A novel outlook and meta-analysis on randomized trials targeting liver injury. BMC Complement Med Ther. 2025;25(1):134. DOI: 10.1186/s12906-025-04886-y
Wang C et al. Silymarin for adults with metabolic dysfunction-associated steatotic liver disease. Cochrane Database Syst Rev. 2025;6(6):CD015524. DOI: 10.1002/14651858.CD015524.pub2
Liu H et al. Effects of different natural products in patients with non-alcoholic fatty liver disease. A network meta-analysis of randomized controlled trials. Phytother Res. 2024;38(7):3801–3824. DOI: 10.1002/ptr.8182
Sahebkar A et al. Lipid-lowering activity of artichoke extracts: A systematic review and meta-analysis. Crit Rev Food Sci Nutr. 2018;58(15):2549–2556. DOI: 10.1080/10408398.2017.1332572
Sereda M et al. Ginkgo biloba for tinnitus. Cochrane Database Syst Rev. 2022;11(11):CD013514. DOI: 10.1002/14651858.CD013514.pub2
Mengs U, Pohl RT, Mitchell T. Legalon SIL: the antidote of choice in patients with acute hepatotoxicity from amatoxin poisoning. Curr Pharm Biotechnol. 2012;13(10):1964–1970. Authors linked to the manufacturer. DOI: 10.2174/138920112802273353
Albertson TE et al. A ten-year retrospective California Poison Control System experience with possible amatoxin mushroom calls. Clin Toxicol (Phila). 2023;61(11):974–981. DOI: 10.1080/15563650.2023.2276674
Gauthier S, Schlaefke S. Efficacy and tolerability of Ginkgo biloba extract EGb 761 in dementia: a systematic review and meta-analysis of randomized placebo-controlled trials. Clin Interv Aging. 2014;9:2065–2077. Co-author employed by the manufacturer. DOI: 10.2147/CIA.S72728
Bone KM. Potential interaction of Ginkgo biloba leaf with antiplatelet or anticoagulant drugs: what is the evidence? Mol Nutr Food Res. 2008;52(7):764–771. DOI: 10.1002/mnfr.200700098
Dehzad MJ et al. Curcumin/turmeric supplementation could improve blood pressure and endothelial function: A grade-assessed systematic review and dose-response meta-analysis of randomized controlled trials. Clin Nutr ESPEN. 2024;59:194–207. DOI: 10.1016/j.clnesp.2023.12.009
US FDA. Investigates two serious adverse events associated with ImprimisRx's compounded curcumin emulsion product for injection. 2017. fda.gov
Lorenz M et al. Tea-induced improvement of endothelial function in humans: No role for epigallocatechin gallate (EGCG). Sci Rep. 2017;7(1):2279. DOI: 10.1038/s41598-017-02384-x
Khanna R et al. Peppermint oil for the treatment of irritable bowel syndrome: a systematic review and meta-analysis. J Clin Gastroenterol. 2014;48(6):505–512. DOI: 10.1097/MCG.0b013e3182a88357
Hieu TH et al. Therapeutic efficacy and safety of chamomile for state anxiety, generalized anxiety disorder, insomnia, and sleep quality: A systematic review and meta-analysis of randomized trials and quasi-randomized trials. Phytother Res. 2019;33(6):1604–1615. DOI: 10.1002/ptr.6349
Hauser RA et al. Randomized, double-blind, pilot evaluation of intravenous glutathione in Parkinson's disease. Mov Disord. 2009;24(7):979–983. DOI: 10.1002/mds.22401
S3 Dementia Guideline (S3-Leitlinie Demenzen). DGPPN and DGN. Version 4.0, 28 November 2023. AWMF register no. 038-013.

Shukri Jarmoukli | Doctor and mentor | Berlin
All information here serves general information purposes and replaces neither a medical examination, nor a diagnosis, nor an individual treatment recommendation. Medical decisions about intravenous therapies always require an individual medical assessment. In an emergency, call 112.

Medicinal Plants by Infusion: what the science really shows, and what it does not show