Review of black cohosh-induced toxicity and adverse clinical effects

Abstract

Black cohosh (Actaea racemosa L) has been utilized for centuries by Native Americans as a traditional herbal medicine. The rhizome and root extract from black cohosh (BCE) is one of the most popular herbal dietary supplements worldwide. Due to its claimed estrogen-like effects, contemporary uses of black cohosh products are primarily for alleviating menopausal and perimenopausal symptoms. However, recent studies indicate that BCE is not only ineffective for menopausal therapy, but also induces genotoxicity through an aneugenic mode of action (MoA). Adverse effects induced by BCE have been reported in humans, with many case studies documenting outcomes ranging from mild reactions to acute liver damage and even death. Consequently, concerns about the safety of BCE have emerged. There are more than 100 chemical constituents in black cohosh products, including triterpene glycosides (>40 chemicals), polyphenols (>20 chemicals), and nitrogenous compounds (>70 chemicals). Therefore, commercially available BCE products can differ markedly in composition, leading to the potential for variable bioactivities among these complex commercial products. This review presents the latest information on the toxicological effects of BCE from both in vivo and in vitro experiments and summarizes the adverse effects of BCE in human clinical trials.

1. Introduction

Dietary supplements belong to a special category under the general umbrella of foods in the United States (U.S.), since the Dietary Supplement Health and Education Act of 1994 (DSHEA) stipulates that such products do not require the U.S. Food and Drug Administration (FDA) to approve their safety and effectiveness.¹ Many consumers of dietary supplements believe that these products are safer and more effective compared to synthetic drug compounds, due to their natural origins and a long history of human use.^2,3 While herbal dietary supplements have been used as primary sources for disease prevention and treatment in developing countries,² their use is also widespread in the U.S. Over half of adults reported using at least one dietary supplement within the past 30 days,⁴ and about 18% of them used herbal dietary supplements.⁵ In a recent survey between January 2017 and March 2020, an estimated 15.6 million U.S. adults consumed at least one of six botanical products associated with liver liability within the past 30 days.⁶ Retail sales of herbal supplements have steadily increased from 2000 to 2023. In 2023 alone, the U.S. sales reached 12.6 billion dollars, with an annual growth rate of 4.4%.⁷

Actaea racemosa L, formerly called Cimicifuga racemosa and commonly known as black cohosh, is a perennial plant with woody rootstock, native to Eastern North America, with a range extending from southern Ontario to Georgia and from Wisconsin to Arkansas.^8,9 The Actaea or Cimicifuga family was historically used as a vermifuge by Native Americans, as its generic name means “drive away bug” in Latin. The native name for black cohosh, “squaw root”, reflects its traditional use in women’s health for centuries, particularly for stimulating menstrual flow.¹⁰ Additionally, black cohosh has been used to relieve other symptoms, including inflammation, cough, diarrhea, and muscle pain.¹⁰ Black cohosh was first listed in the U.S. Pharmacopeia in 1820 and later included in the National Formulary in 1936.^11,12 In the late nineteenth century, it was introduced to Europe, where it has been used as an alternative to hormone replacement therapy for over 70 years.^10,13 This natural product is widely regarded as an effective treatment for menopausal symptoms, such as hot flashes and night sweats.¹⁴

Due to these potential benefits, black cohosh has become one of the most popular herbal dietary supplements, available over the counter in various forms, most commonly as a black cohosh extract (BCE). In 2023, it ranked 23^rd among the top-selling herbal supplements in the mainstream U.S. market, with total sales reaching 19.9 million dollars.⁷ According to the most current estimates, about 1.2 million of U.S. adults used black cohosh products within the last 30 days.⁶ However, some clinical studies have questioned the efficacy of BCE in treating menopausal symptoms due to its limited estrogenic activity.^15–19 Inconsistent results have been reported regarding BCE’s effectiveness in attenuating hot flashes.^18,20 Other potential benefits, such as its anti-inflammatory and anti-cancer properties, have been observed in animal models and in vitro studies, but have not yet been tested in humans.^15,21,22

A significant safety concern regarding BCE is its potential hepatotoxicity, prompting regulatory agencies in Australia, Canada, the European Union, the United Kingdom (U.K.), and the U.S. to issue cautionary labels for BCE-related products.^11,23 In addition, the complexity of active constituents identified in BCE raises concerns regarding their potential toxicity and interactions with endogenous enzymes, pharmaceuticals, or other natural products in humans.^24–27 Considering the potential public health risks associated with the concentration, composition, and contaminants of BCE products, as well as their interactions with drugs or other dietary supplements, it is important to characterize the toxic effects of BCE. Such research can also facilitate standardizing BCE quality and regulating its application in humans.

Black cohosh was nominated for toxicological evaluation by the U.S. National Toxicology Program (NTP) in 1999 based on the following reasons: (1) its long history of use for menstrual and menopausal symptoms in women; (2) its widespread application as a proprietary product; and (3) the lack of chronic safety studies in animals and humans at that time.^28,29 Since then, the NTP has conducted several studies on BCE. Initial sub-chronic toxicity tests (3 months) in female rats and female mice resulted in dose-dependent hematological changes and chromosomal damage, as evidenced by megaloblastic anemia (MA) and an increased micronucleus (MN) frequency in peripheral erythrocytes, respectively.^28,30 In the subsequent 2-year gavage study, BCE induced dose-dependent chromosomal damage, indicated by increased frequencies of micronuclei in the peripheral blood of female mice during the 3- and 12-month interim evaluations. However, BCE was not carcinogenic in female mice.²⁹ The genotoxicity of BCE has also been observed in human lymphoblastoid TK6 cells.^31,32

Recently, we conducted a literature search on PubMed using the keyword “black cohosh” alone and in combination with “toxicity”. The number of publications in this field has increased every year. Among a total of 1,018 publications on black cohosh (from 1851 to 2025; searched on May 19, 2025), more than 90% reported its beneficial activities, while only 89 papers, published between 2001 and 2024, focused on its toxicity and adverse effects. To better understand black cohosh-induced toxicity and adverse clinical effects, this review summarizes the general toxicity, hepatotoxicity, and potential genotoxicity and carcinogenicity of BCE based on both in vivo and in vitro experiments, as well as the reported adverse effects of BCE in humans.

2. Chemical components of black cohosh

Previous studies have conducted chemical component analyses and quality control assessments for various BCE products.^33–37 Both roots and rhizomes from black cohosh are economically important for BCE preparation.⁸ During sample preparation, different solvents (such as ethanol, methanol, iso-propanol, and butanol) can be used either alone or mixed with water to extract black cohosh. Specifically, water-mixed solvents yielded higher quantities of BCE samples than organic solvents alone.³⁸ For example, a 50–70% ethanol mixture is commonly used due to its high BCE yield, with a typical drug to extract ratio ranging from 4.5–8.5:1.^39–41 The mixture of active constituents in BCE can be analyzed using high-performance liquid chromatography-mass spectrometry (HPLC-MS), and more than 100 constituents have been identified, including triterpene glycosides (>40 chemicals),^25,42–44 polyphenols (>20 chemicals),^34,42,45,46 and nitrogenous compounds (>70 chemicals)^47,48 (Table 1). Regardless of the extraction solvent used, triterpene glycosides have been consistently reported as the main active BCE components, although their amounts may vary.³⁸

Table 1.

Chemical constituents in BCE.

Class	Major chemical constituents (CAS number)	References
Triterpene glycosides (triterpenoids, saponins) (>40 chemicals)	23-epi-26-deoxyactein (CAS 501938-01-8), 25-anhydrocimigenol xyloside (181765-11-7), 25-O-acetyl-cimigenol xyloside (27994-12-3), 26-deoxyactein (264624-38-6), 26-deoxycimicifugoside (214146-75-5), acetylshengmanol xyloside (62498-88-8), actein (18642-44-9), cimiaceroside A (210643-83-7), cimicifugoside (66176-93-0), cimicifugoside H1 (163046-73-9), cimicifugoside H2 (161097-77-4), cimigenol (3779-59-7), cimigenol xyloside (27994-11-2), cimiracemosides [A (264875-61-8), B (290821-38-4), C (256925-92-5), D (290821-39-5), E (290821-40-8), F (264875-61-8), G (289632-43-5), H (290821-41-9)]	[25,41–44]
Polyphenols (phenolic acids, flavonoids) (>20 chemicals)	caffeic acid (331-39-5), cimicifugic acids [A (205114-65-4), B (205114-66-5), D (219986-51-3), E (219986-67-1), F (220618-91-7), G (874351-11-8)], cimiracemate A (478294-16-5), cimiracemate B (478294-17-6), ferulic acid (1135-24-6), fukiic acid (35388-56-8), fukinolic acid (50982-40-6), isoferulic acid (537-73-5), piscidic acid (469-65-8), salicylic acid (69-72-7) formononetin (485-72-3), kaempferol (520-18-3)	[34,41,42,45,46]
Nitrogenous compounds (alkaloids) (>70 chemicals)	Allocryptopine (485-91-6), cimipronidine (865266-71-3), cimitrypazepine (1422514-69-9), cyclo-cimipronidine (1111268-98-4), cytisine (485-35-8), dopargine (1111269-01-2), N-methylcytisine (486-86-2), N-methylserotonin (1134-01-6), norsalsolinol (34827-33-3), protopine (130-86-9), reticuline (485-19-8), salsolinol (27740-96-1)	[41,47,48]

The chemical profiles and compositions of BCE products on the market can vary significantly, largely due to harvest conditions, collection season, and plant tissue used.⁴⁹ To standardize the quality of BCE, the two most abundant triterpene glycosides, 23-epi-26-deoxyactein and actein, are commonly used as markers, with their amounts in total triterpene glycosides being around 3% and 3.4%, respectively.^25,36,50,51 In addition to analytical chemistry methods, DNA barcode identification is also used to monitor the quality of BCE products. A previous study screened 36 dietary supplements labeled as BCE and found that 27 (75%) samples contained two matK nucleotides that exactly matched the BCE reference. The remaining nine (25%) samples showed a different sequence pattern, aligning with three Asian Actaea species: A. cimicifuga, A. dahurica, and A. simplex.³⁵ Usually, licensed or registered BCE products are considered authentic, while adulteration and/or mislabeling have been seen in 21% of tested BCE dietary supplements.⁵²

The first commercial product, Remifemin, is a standardized proprietary isopropanolic extract of black cohosh, containing 1 mg of 23-epi-26-deoxyactein per tablet or around 2.5% triterpene content.¹⁵ Developed by Schaper & Brümmer in Germany, Remifemin has been used for seven decades, and its pharmaceutical effects and recommended doses in both animals and humans have been documented.¹³ On the other hand, the NTP study applied a single lot of BCE product obtained from PlusPharma, Inc. (Vista, CA), which was extracted with 50% ethanol and standardized to 7.8% (w/w) total triterpene glycosides and manufactured by Frutarom Switzerland Ltd. The LCMS quantitative analysis showed it contained 0.47% actein and 2.09% 23-epi-26-deoxyactein.²⁹ Another study analyzed 11 different BCE products and found that the amount of seven selected triterpene glycosides ranged from 0.7% to 4.7%, with eight of these 11 products claiming to contain 2.5% triterpene glycoside content on their labels.⁵³

Furthermore, a recent study emphasized that the biological-response profile, rather than the chemical composition, should be the primary basis for investigating the pharmaceutical effects and toxicity of BCE products.³⁷ This study screened 14 BCE samples, including the NTP reference. Although only nine (64%) samples had chemical profiles closely matching the NTP BCE samples, all BCE samples exhibited similar gene expression profiles for toxicologically relevant hepatic receptor pathways in primary human hepatocytes. These included aryl hydrocarbon receptor (AhR), constitutive androstane receptor (CAR), pregnane X receptor (PXR), farnesoid X receptor (FXR), and peroxisome proliferator-activated receptor alpha (PPARα).³⁷

3. Adverse clinical effects of BCE in humans

3.1. General adverse effects

An early study conducted a systematic literature search from Australia, Canada, the U.K., and the World Health Organization (WHO), along with data from 16 manufacturers of BCE preparations, and reported that the adverse effects of BCE were observed as mild, rare, and even reversible in clinical trials.⁵⁰ Another study involving more than 2,800 patients indicated that 5.4% of patients experienced adverse events after BCE uptake, with 97% of these adverse effects being minor.⁵⁴ The most common adverse effects include gastrointestinal symptoms, musculoskeletal disorders, and connective tissue disorders.⁵⁵ Among the 2,016 Hungarian women treated with BCE for 12 wk, 1,876 patients were fully evaluated for the treatment of menopausal symptoms, and 1.9% of them had specified unexpected adverse effects, such as extremities, gastric pain, and allergic reactions.⁵⁶ In addition, unlike the rodent studies (see Section 4), 3 to 12 months of BCE uptake did not alter blood laboratory parameters in 13 clinical trials, including two studies with a two-armed, randomized, and open design,^57,58 one study with a prospective, open, uncontrolled drug safety study,⁵⁹ and ten randomized, double-blind studies.^60–70 Taken together, the overall adverse effects of BCE in humans are mild and transient, with most symptoms disappearing after discontinuing BCE intake. Only one death was reported, in which the patient experienced cardiopulmonary arrest after taking a herbal mixture containing BCE, and the death was most likely related to pennyroyal herb.⁷¹

3.2. Hepatotoxicity in humans

Hepatotoxicity is one of the most noticeable adverse effects of BCE in humans and has been widely reported in women receiving BCE treatments. About 100 cases of liver damage, such as acute hepatitis,^72–74 autoimmune hepatitis,⁷⁵ and acute liver failure^73,74,76,77 have been associated with BCE treatments. These cases have raised concerns among regulatory agencies across different countries regarding the potential liver toxicity of BCE. In February 2006, the Therapeutic Goods Administration (TGA) in Australia became the first to label BCE with a warning notice for its potential hepatotoxicity.⁷⁸ Later that year, the Medicines and Healthcare products Regulatory Agency (MHRA) in the U.K. and the European Medicines Agency (EMA)—Committee on Herbal Medicine Products (HMPC) also expressed similar concerns about BCE products in May and July.¹¹ In August of the same year, the Natural Health Products Directorate (NHPD) in Canada issued an advisory, noting that even though the possible link between BCE and liver injury remained unclear, consumers should stop using BCE if adverse effects, such as fatigue, weakness, and loss of appetite, developed.¹¹

The underlying mechanisms of BCE-induced hepatotoxicity are largely unknown. Previous studies involving young patients, in vivo study in hPXR/CYP3A4 double transgenic mice, and in vitro studies using liver microsomes have consistently showed that BCE does not inhibit CYP1A2, 2E1, or 3A4.^79,80 However, one human study showed that BCE inhibited CYP2D6, but the effect was minimal⁸⁰ and not observed in other human studies.⁸¹ Additionally, one in vitro study using CYP Isozyme Bioassays found that BCE treatments inhibited CYP1A2, 2C9, 2D6, and 3A4 in a concentration-dependent manner.²⁶ Meanwhile, for the hepatotoxicity observed in women of peri- to post-menopausal age (39–72 years old), it is important to note that most of these patients had multiple chronic diseases and were taking multiple medications.^11,82 This suggests that other drugs could potentially be contributing to the observed hepatotoxicity. Furthermore, although specific BCE constituents have shown potential CYP inhibition, the recommended BCE dose for humans is too low for those constituents to reach an effective concentration needed for CYP inhibition. For instance, the average BCE dose is 40–80 mg, and after distribution and metabolism in the humans, the remaining amount of total triterpene glycosides (approximately 8 mg) is far less than the amount required to exert CYP inhibition, according to their IC₅₀ values on different CYP enzyme in in vitro studies.²⁶

Current available literature suggests a weak link between BCE use and liver toxicity. For instance, 11 FDA MedWatch cases, involving 4 duplicate reports in published papers and 7 nonduplicate reports have suggested a possible connection between BCE and liver injury.^11,72,77,83 However, a meta-analysis on five randomized, double-blind, and controlled clinical trials found no connection between BCE and adverse effects on liver functions, including aspartate aminotransferase (AST), alanine aminotransferase (ALT), and γ-glutamyl transpeptidase (γ-GT).⁸⁴ This weak causality between BCE and liver toxicity is mainly due to the lack of detailed description of patient conditions, the specific BCE products used, and the absence of essential liver parameter measurements.²³ Except for one study showing BCE-induced liver injury in a 44-year-old woman with no other medications,⁸⁵ most studies reported that patients were taking multiple drugs or herbal mixtures, complicating the determination of causality between BCE and liver injury.⁷⁷ For example, one case study associated with BCE hepatotoxicity also involved a mixture of herbal medicines containing Greater Celandine and Buchus leaf with Pulegone, which are both associated with acute hepatitis and liver toxicity.⁷² Another case of BCE-induced fulminant hepatic failure with micronodular cirrhosis was likely influenced by modest alcohol consumption or possible preexisting liver diseases undetected by the core biopsy, rather than BCE treatment.⁷⁷ The weak causality is also due to a lack of information on the quality and identity of BCE products used in studies. As highlighted in FDA MedWatch reports, the BCE products were often not characterized for their identity or quality. The details on the sources, preparation, and purification of BCE products were limited, and the doses and treatment durations were frequently not specified.¹¹ Additionally, most spontaneous case studies did not measure ALT and ALP levels, which are essential standard biomarkers for assessing BCE-induced liver injury.⁸²

Several human studies have also noted that the limitations of analysis methods significantly impact the assessment of BCE’s causality in liver injury. For example, drug-induced liver injury (DILI) can be assessed using the Roussel Uclaf Causality Assessment Model (RUCAM) and its simplified version, the Maria and Victorino (M&V) scale. Both models have indicated a probable or possible link between BCE and DILI.^77,86 However, both models typically require rechallenging with the drugs, which is not possible for BCE due to safety concerns for patients.⁸⁶ Another common analysis method, the Naranjo scale, has been found to be not liver-specific.^82,84 When comparing the Naranjo scale with the liver specific International Organizations of Medical Science (CIOMS) analysis, the latter consistently showed lower causality scores in previous spontaneous reports.⁸² CIOMS analysis of 69 reported cases showed that the liver injury induced by BCE was unlikely or excluded in all reports, except for one study that used an unidentified BCE product.^82,87 In addition, weak causality was also observed when CIOMS was applied to 16 published case reports and 24 spontaneous studies.⁸⁷

Despite the findings of weak causality and the limitations of various assessment methods, the increasing number of case studies and accumulating evidence worldwide have led most regulatory agencies to continue issuing warning labels on the BCE products. These warnings specifically alert consumers to the potential for liver toxicity, especially for individuals with preexisting conditions, including potential liver disease, a history of liver transportation, and immune-mediated liver injury.^23,88 This cautionary approach reflects the ongoing concerns about the safety of BCE products, even in the absence of definitive evidence linking them directly to liver injury.

3.3. Potential genotoxicity in humans

The NTP’s cross-sectional study involving 23 women taking BCE and 28 women without BCE for at least 3 months, all of whom (18 ~ 70 years old) were healthy and not undergoing other medical treatments, found no significant difference in MN frequency and hematological parameters between the two groups.⁸⁹ The lack of difference between the groups may be attributed to several factors, including the higher doses of BCE used in rodents compared to humans. For instance, when the lowest rodent dose (250 mg/kg/day, induced MN induction in mice) was converted to the equivalent human dose using allometric scaling, it came out to 20 mg/kg/day, which is about 30 times higher than the recommended human dose of 0.67 mg/kg/day for a 60-kg woman or 40 mg/day.^28,90 Additionally, factors such as age selection and lifestyle could influence the results. For example, older women tend to have lower serum cobalamin levels than younger women, but the study observed the normal cobalamin levels, indicating that older women may be taking supplements that mask potential effects of BCE on cobalamin metabolism.⁸⁹ This study highlights the importance of optimizing experimental design and volunteer conditions in future studies to better capture any subtle but potentially significant biological effects of BCE.

4. In vivo and in vitro toxicology of BCE

4.1. General toxicity of BCE

BCE has generally been shown to reduce body weight in rodents. For example, female Sprague-Dawley rats treated with 6,700 mg/kg/day of CR BNO 1055 extract (a BCE product) in their diet for 1.5 months exhibited reduced body weight gain,⁹¹ though food consumption did not differ between the control and BCE-treated groups. In a sub-chronic study, female Wistar Han rats and female B6C3F1/N mice treated with repeated doses of BCE (up to 1,000 mg/kg/day) for 3 months showed no significant effects on survival or body weight.²⁸ Similarly, a 92-day BCE study found that 1,000 mg/kg/day of BCE did not decrease survival in female B6C3F1/N mice, but body weight and body weight gains were significantly lower in the treated group compared to the control.³⁰ In the NTP 2-year study, female Sprague-Dawley rats dosed with BCE during gestation and lactation had lower mean body weights (within 10%); and after a 2-year BCE treatment, the treated female rats and perinatally exposed male rats showed approximately 13% and 11% less mean body weights compared to controls, respectively.²⁹ The study also observed a 40% reduction in mean body weight in B6C3F1/N mice treated with the highest BCE dose (1,000 mg/kg/day) for two years. Despite these body weight reductions, survival rates were not affected in either rats or female mice. Other clinical observations, such as uterine dilation, hemorrhage, thrombus formation, ulcers, and ovarian atrophy, were only found in the BCE-treated rats, but not in female mice.

BCE has been shown to alter lipid profiles in previous animal studies. Female Sprague-Dawley rats treated with the maximum tolerated dose of 35.7 mg/kg/day BCE exhibited reduced serum free fatty acids at both 6 and 24 h, along with decreased triglyceride levels at 6 h.⁹² Another study in female Sprague-Dawley rats treated with 6,700 mg/kg BCE in chow for 1.5 months also reported a reduction in triglycerides, intra-abdominal fat accumulation, and an increase in plasma LDL-cholesterol.⁹¹ However, in a 90-day study, no significant changes in lipid metabolism biomarkers were observed at any tested doses (up to 1,000 mg/kg/day) in either female B6C3F1/N mice or female Wistar Han rats.²⁸

It is well-documented that BCE induced hematological effects in rod ents.^28,30,93 As summarized in Table 2, a sub-chronic study found that a 90-day BCE treatment in female B6C3F1/N mice and female Wistar Han rats resulted in decreased red blood cell (RBC) counts, along with increased mean cell volume (MCV) and mean corpuscular hemoglobin (MCH).²⁸ In addition, a decrease in hemoglobin (Hgb) was observed only in female B6C3F1/N mice. BCE also upregulated four genes regulating cobalamin metabolism, suggesting that the observed hematological changes were consistent with MA.^28,93 MA is often caused by deficiencies or perturbations in folate and cobalamin metabolism, which disrupt DNA synthesis and interfere with the re-methylation of homocysteine to methionine.^94,95 In a follow-up study, female B6C3F1/N mice treated with 1,000 mg/kg/day BCE for 92 days did not increase folate and cobalamin serum levels, but did exhibit significantly elevated biomarkers of folate and cobalamin deficiency, such as homocysteine and methylmalonic acid, supporting the hypothesis that BCE perturbates the metabolism of folate and cobalamin.^30,93 While BCE-induced anemia in rodents is usually not life-threatening, prolonged exposure may lead to more detrimental effects on target sites, such as the bone marrow.²⁸ A 2-year chronic NTP study in female Sprague-Dawley rats and female B6C3F1/N mice with BCE (up to 1,000 mg/kg/day) reported abnormal hematological parameters in both species at 12 months. Mice exhibited more severe hematological effects than rats, as evidenced by abnormal metarubricytes at 3 months and decreased erythrocyte counts and Hgb concentrations at 12 months.²⁹

Table 2.

Hematological effects and genotoxicity of BCE in the rodent studies.

	Study 128		Study 230,93		Study 329
Animal model	Female Wistar Han rats	Female B6C3F1/N mice	Female B6C3F1/N mice	Sprague Dawley rats	Female B6C3F1/N mice
BCE doses (mg/kg/day b.w.)	0, 15, 125, 500, and 1,000	0, 62.5, 125, 500, and 1,000	0 and 1,000	0, 75, 250, and 750 (for time-mated female rats and their offspring)	0, 30, 100, 300, and 1,000
Treatment period	Gavage for 90 days	Gavage for 92 days	Gavage for 2 years (Male: perinatal exposure only, female: perinatal exposure and postweaning dosing)	Gavage for 2 years (postweaning dosing only)
Hematological parameters:				NA	3 months	12 months
Red blood cells (106/μL)	↓	↓	↓		↓	↓
Hemoglobin (g/dL)	—	↓	—		—	↓
Hematocrit (%)	—	↓	—		—	↓
MCV (fL/cell)	↑	↑	↑		↑	↑
MCH (pg/cell)	↑	↑	↑		↑	↑
MCHC (g/dL)	—	—	—		—	—
Reticulocyte (103/μL)	↓	—	↓		—	—
Platelet (103/μL)	↑	↑	—		NA	NA
Chemistry parameters:	NA	NA		NA	NA	NA
Homocysteine (μmol/L)			↑
Methylmalonic acid (ng/mL)			↑
Genotoxicity (MN assay):			NA	NA
Micronucleated reticulocytes	↑	↑			NA	NA
Micronucleated erythrocytes	NA	↑			↑	↑

NA, not available; —, no change; MCV, mean cell volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; MN assay, micronucleus assay.

The mechanisms underlying BCE-induced disruption of folate and cobalamin metabolism have also been investigated. An in vitro study found that a concentration of 50 mg/mL BCE did not inhibit dihydrofolate reductase, a key enzyme in folate metabolism.³⁰ However, due to species differences, the folate and cobalamin requirements, storage, and metabolism differ between humans and mice, making it challenging to directly extrapolate findings from rodent models to humans.³⁰

A study investigating BCE-associated immunotoxicity reported minimal effects on the immune system. Female B6C3F1/N mice were administered BCE daily by oral gavage (up to 1,000 mg/kg/day) for 28 days, matching the dose levels used in previous 90-day toxicity studies.⁹⁶ The only observed effect was an increase in cytotoxic T cell activity, while no significant changes were detected in T-dependent antibody responses of the humoral immune system, innate immunity, or bone marrow cellularity and colony-forming units.⁹⁶

4.2. Hepatotoxicity of BCE

Mild liver toxicity associated with short-term BCE treatment has been observed in rodent studies. In female Sprague-Dawley rats, a 35.7 mg/kg/day BCE treatment induced lipid accumulation in the liver, as shown by Oil Red O histology staining, accompanied with signs of mild toxicity, including microvesicular lipid droplets, at 24 h post-treatment.⁹² Similarly, in female Wistar Han rats, a 90-day BCE treatment at 500 and 1,000 mg/kg/day resulted in increased liver weight and minimal to mild liver necrosis.²⁸ In a 2-year study, female B6C3F1/N mice treated with 500 and 1,000 mg/kg/day BCE also exhibited minimal liver necrosis, with a significant increase in incidence at the higher dose group.²⁹

The cytotoxicity potential of BCE in hepatocytes was further investigated using human hepatoma HepG2 cells. A previous study found that BCE at concentrations up to 50 μg/mL for a 72-h exposure did not inhibit HepG2 cell growth.²⁶ However, an extended 96-h treatment induced cell proliferation inhibition, with a median inhibitory concentration (IC₅₀) of 37 μg/mL.⁹² Since BCE did not induce severe hepatotoxicity in either rodent models or cultured HepG2 cells, previous reports concluded that BCE-related liver toxicity may occur through indirect mechanisms, such as immune-mediated liver injury or potential herb-drug interactions affecting drug metabolism.²⁶

Animal studies revealed that the hepatotoxicity of BCE is related to its potential cytotoxicity and herbal-drug interactions, as evidenced by changes in the expression of genes involved in cell apoptosis and hepatic metabolism in a microarray analysis. Female Sprague-Dawley rats treated with a single dose of 35.7 mg/kg BCE and sacrificed 24 h later showed downregulation of genes related to cell growth (Id1, lgfbp3), cell cycle (Ccnd1), xenobiotic metabolic response (Hamp), metabolic process (Aldh1a4), fatty acid synthesis (SCD1), and transport (Syt12) in liver tissues.⁹² Repression of Cyclin D and cell cycle arrest at G1/S boundary also contributed to the anti-proliferative effects of BCE in rat livers. It has been reported that BCE inhibited mitochondrial oxidative phosphorylation, directly inhibiting ATP generation,^92,97 and further showed mixed effects on apoptosis as evidenced by increased Caspase 9 and IAP5 expression, as well as suppressed cytochrome C and BAX.⁹² Longer-term BCE treatment also altered enzyme expression related to hepatic metabolism. In a 28-day study, in female C57BL/6J mice treated with BCE (up to 500 mg/kg/day) exhibited hepatic induction of Cyp3a11, a murine homolog of human CYP3A4, through activation of nuclear receptor PXR, which may lead to the hepatic toxicity.⁹⁸ However, BCE had no effect on CYP3A4 expression in hPXR/CYP3A4 double-transgenic mice. Similarly, a 92-day BCE study in female B6C3F1/N mice (up to 1,000 mg/kg/day) identified the most upregulated genes by BCE treatment were three drug-metabolizing enzymes, including Cyp2c18, Cyp7a1, and Akr1b7; while the top three enriched canonical pathways, based on QIAGEN Ingenuity Pathway Analysis, were hypoxia signaling in the cardiovascular system, Nrf2-mediated oxidative stress response, and protein ubiquitination pathway.⁹³ BCE treatments also inhibited CYP1A2, 2C9, 2D6, and 3A4 in a concentration-dependent manner through in vitro CYP Isozyme Bioassays.²⁶ The IC₅₀ of the three different BCE products (75% and 80% ethanol extracts and 40% isopropanol) ranged from 21.9 to 65.0 μg/mL. Another BCE study reported an IC₅₀ of 27 μg/mL for CYP3A4 inhibition.⁹⁹ Similar inhibitory effects were observed with isolated triterpene glycosides, including Cimiracemoside A, 23-epi-26-deoxyactein, and actein, which inhibited CYP2C9 and CYP3A4 activities, with IC₅₀ values ranging from 25 to 51.3 μM, while showing no effect on CYP1A2.²⁶

However, the overall hepatotoxicity of BCE in rodents was minimal compared to its hematological effects. Most importantly, there were no significant changes in clinical chemistry parameters which associated with liver toxicity, including sorbitrol dehydrogenase, bile acids, alkaline phosphatase, and alanine aminotransferase, in a 90-day female rat study.²⁸ The only observed change was a mild reduction in albumin levels in the rat serum following the treatment with 1,000 mg/kg/day, but this was insufficient to indicate hepatic injury.

4.3. Genotoxicity of BCE

It has been reported that BCE treatment for 15 days exerts toxicity through oxidative stress in the livers of ovariectomized female Wistar Han rats with renovascular hypertension.¹⁰⁰ The dose (0.6 mg/kg/day) used was close to the recommended dose (40 mg/kg/day) for a 60-kg woman. An in vitro study also showed that exposure to 100 μg/mL BCE for 1 h induced oxidative stress in vascular smooth muscle cells isolated from Sprague-Dawley rats.¹⁰¹ BCE therefore has the potential to induce genotoxicity considering that oxidative stress is a primary source of indirect DNA lesions in mammalian cells.¹⁰² Hierarchical clustering analysis further showed that BCE clustered with other genotoxic compounds, such as tubulin binding vinca alkaloids and DNA alkylators.⁹²

As summarized in Table 2, a sub-chronic BCE study involving 90-day BCE treatment in female B6C3F1/N mice and female Wistar Han rats showed a significant increase in chromosomal damage, as indicated by a higher micronucleated RBC frequency.²⁸ Mice showed a more pronounced responses to BCE at lower doses compared to rats. For instance, a significant increase in micronucleated RBC was found in mice treated with 125 mg/kg/day BCE, while similar effects were only observed in rats treated with 1,000 mg/kg/day BCE.²⁸ Results from the NTP 2-year rodent study with BCE treatment up to 1,000 mg/kg/day also showed elevated peripheral micronucleated RBCs in B6C3F1/N mice.²⁹

Several in vitro studies have been performed to provide relevant data on BCE’s effects in cultured human cells (Table 3). Under the physiological level of folic acid (300 nM), 125 μg/mL BCE for 4 h and 95 μg/mL BCE for 24 h increased MN frequency in human TK6 cells, while co-cultured with a higher folic acid concentration (3,000 nM) effectively attenuated this chromosomal damage.³¹ In the MultiFlow DNA damage assay, BCE was shown to increase phospho-histone H3 and p53 nuclear translocation after 24 h of treatment, but did not increase the expression of common DNA double strand damage marker γ-H2AX.^31,90 This pattern indicates an aneugenic mode of action (MoA). Similar to the well-known aneugen colchicine, which inhibits polymerization of microtubules and can disrupt cobalamin absorption, leading to MA and MN formation,¹⁰³ BCE may contain similar constituents. Its gene expression pattern also resembles that of tubulin-binding vinca alkaloids.⁹² Another study showed that BCE treatment for 24 h directly destabilized microtubules in vitro, which may be responsible for its aneugenic MoA.⁹⁰ Our study further showed that in human TK6 cells, BCE over 75 μg/mL increased MN frequency after a 24-h treatment, while BCE at 250 μg/mL caused acute DNA damage peaked within 0.5 h and gradually returned to the baseline after 24 h in the comet assay.³² Although BCE at 250 μg/mL for 4 h and 24 h induced γ-H2AX expression in TK6 cells, this induction was more likely associated to apoptosis rather than genotoxicity, as indicated by concurrent positive caspase 3 expression.³² Since aneugen predominantly generates apoptotic γ-H2AX expression,¹⁰⁴ these results speculated that BCE is a potential aneugen, which is in agreement with the NTP study.

Table 3.

Genotoxicity of BCE in the in vitro studies.

	Study 131			Study 290		Study 332				Study 429
In vitro model	TK6 cells		Bacteria	TK6 cells		TK6 cells		HepG2		Bacteria
BCE concentrations	Up to 1,000 μg/mL		Up to 6,000 μg/plate	Up to 500 μg/mL		Up to 250 μg/mL		Up to 500 μg/mL		Up to 10,000 μg/plate
Treatment time	4 h	24 h	20 min	4 h	24 h	4 h	24 h	4 h	24 h	20 min
Mutagenicity:
Ames test			Equivocal in TA98 with 10% S9							Equivocal in TA98 with 10% S9
Genotoxicity:
MN assay	↑	↑		NA	↑	NA	↑	NA	↑
Comet assay	NA	NA		NA	NA	↑	↑*	—	—
YH2AX	—	—		—	—	↑	↑	NA	NA
P53	—	↑		—	↑	—	↑	NA	NA
p-Histone 3	↑	↑		↑	↑	NA	NA	NA	NA

MN assay, micronucleus assay; NA, not available; *, not significant; —, no change.

Interestingly, when human TK6 cells were exposed to BCE at concentrations up to 1,000 μg/mL for 24 h with the co-treatment of rat liver S9 mixture for 4 h, no MN induction was observed.³¹ A 24-h BCE treatment in human HepG2 cells also showed negative results in both the MN assay (<200 μg/mL BCE) and the comet assay (up to 500 μg/mL BCE).³² Compared to TK6 cells, which lack metabolic components, HepG2 cells may metabolize BCE through hepatic CYP450 enzymes and phase II antioxidant or detoxification enzymes. Thus, the response to BCE-induced genotoxicity is largely dependent on the cell type, particularly its capacity for chemical metabolism and detoxification.

4.4. Carcinogenicity of BCE

In the NTP 2-year gavage study, no evidence of carcinogenic activity was observed in female B6C3F1/N mice treated with BCE at the doses of 30, 100, 300, and 1,000 mg/kg/day.²⁹ In contrast, female Sprague-Dawley rats treated with BCE at doses of 75–750 mg/kg/day for 2 years, with exposure beginning perinatally (during gestation and lactation) and postweaning (5–6 wk of age), showed slightly increases in the incidence of uterine squamous cell papillomas. However, these increases were not statistically significant and were considered equivocal evidence of carcinogenic activity in these animals.²⁹ Another study investigated if BCE influences mammary cancer development and progression using the MMTV-neu mouse model, which is widely used to study HER2-positive breast cancer.¹⁰⁵ Hemizygous female MMTV-neu mice were fed either a control diet or BCE-supplemented diet (40 mg/day) from 2 months of age until 16 months. There was no significant difference in the development of primary mammary tumors at 16 months between the two groups, with 93.9% tumor incidences (123 of 133 mice) in the control group and 92.9% (118 of 127 mice) in BCE treatment group. However, the total tumor volume at the time of death was significantly higher in the BCE-treated group compared to control mice, indicating that BCE contributed to increased cystic fluid accumulation in tumor tissue.¹⁰⁵ In addition, BCE-treatment was associated with a significantly higher incidence of lung metastases in tumor-bearing mice. These results indicate that further studies are needed to explore the potential effects of BCE in women, both with and without breast cancer, as well as its influence on metastatic breast cancer.¹⁰⁵

4.5. Toxicity of active constituents in black cohosh

Due to BCE’s complex components (Table 1), little is known about the active constituents responsible for its bioactivity and toxicity. Using five computational toxicology software, a previous study performed computational predictions for four botanicals, and 37 chemical constituents contained in black cohosh were predicted “positive” for various liver toxicity.¹⁰⁶ However, these predictive results have not been validated through follow-up studies using the in vivo or in vitro toxicity evaluations. Table 4 shows two triterpene glycosides, 23-epi-26-deoxyactein^26,102,107 and actein,^{102,107–112} and one nitrogenous compound, salsolinol,^{27,102,113,114} with reported cytotoxicity and genotoxicity. Based on their toxicity profiles and the total amount found in BCE products, actein appears to be the most likely contributor to BCE-induced toxicity and genotoxicity.¹⁰²

Table 4.

Summary of toxicity and genotoxicity of three BCE constituents.

Chemical	% in BCE	Cytotoxicity	Reference	Genotoxicity	Reference
23-epi-26-deoxyactein	abundant (3.4%)	Inhibited cell growth in human cells:		Negative in micronucleus assay	102
		MCF7 cells, IC50= 32 μM (96 h)	107
		TK6 cells, IC50=1 75 μM (24 h)	102
		In vitro CYP isozyme bioassays:
		Inhibition on CYP3A4 IC50= 28.7 μM, CYP2C9 (IC50=25.9 μM) and CYP2D6 (IC50=100 μM)	26
Actein	abundant (6.4%)	Inhibited cell growth in human cells through G1 cell cycle arrest and apoptosis:		Oxidative related DNA damage (γ-H2AX and positive in comet assay), Chromosome damage (MN formation) in TK6 cells,	102
		MCF7 cells, IC50=21–30 μM (96 h)	107,108
		Colon HT29 cells, IC50=35 μM (96 h)	112	Activated MAPK pathway
		MDA-MB-453 cells, IC50=12 μM (96 h)
		TK6 cells, IC50=50 μM (24 h)	102
		HMEC-1 cells, IC50= 65 μM (48 h)	109
		SNU-216 cells, IC50= 30 μM (24 h)	110
		Bladder cancer cells (BIU87, T24, T739, 5637), IC50=5–10 μM (24 and 48 h)	111
		In vitro CYP Isozyme Bioassays:
		Inhibition on CYP3A4 (IC50=35.2 μM), CYP2C9 (IC50= 25 μM) and CYP2D6 (IC50=51.1 μM)	26
Salsolinol	small	Inhibited cell growth:
		TK6 cells, IC50=150 μM (24 h)	102	Oxidative related DNA damage in MCF-10A cells,	27
		SH-SY5Y cells, IC5C1=200–250 μM (24 h)	113,114	Activated JNK and NF-κB pathway	114

5. Concluding remarks

Herbal dietary supplements have been considered to have health benefits in humans with a long history in many cultures, however, the concerns about their safety and toxicity such as hepatotoxicity have also been raised.^115–118 Black cohosh, one of the top-selling herbal dietary supplements, is primarily marketed to women for treating various disorders and relieving menstrual and perimenopausal symptoms. Its products available over the counter come in various forms, such as powdered whole plant, liquid extracts, and dried extracts in pill form. In the U.S., 40 to 50 million women suffer from menopausal symptoms, and around 51% of these women use complementary and alternative medicine, including herbal dietary supplements that often contain BCE as an active component.¹¹⁹ However, a systematic review has suggested that there is insufficient evidence to support the efficacy of black cohosh for treating menopausal symptoms.⁵⁵

The toxicity and adverse effects of black cohosh in humans have been increasingly reported. Its common side effects in various systems include weight gain, headache, heart arrhythmia, difficult breathing, nausea, vomiting, breast pain/enlargement, muscle pain, and skin rashes.⁸⁶ BCE-induced hepatotoxicity can range from mild, self-limited liver injury to acute liver damage, fulminant hepatic failure, and even death.^11,26 Consequently, health agencies in the U.K., U.S., Australia, Canada, and European Agency have issued warning labels on BCE products, advising caution due to potential hepatotoxicity.

Clinical investigations into BCE-induced toxicity face several challenges. First, high quality and contaminant-free raw black cohosh plant is of primary importance, because accidental misidentification or deliberate adulteration could harvest other related species. A recent report indicated that 42% of black cohosh rhizome samples (135 of 320 samples) were adulterated and/or mislabeled, especially the products sold as dietary supplements.⁵² Another report showed 46% of 33 samples of black cohosh food supplement products and 67% of black cohosh herbal drugs/preparations (out of 27 samples) had questionable quality.¹²⁰ The main adulterants of black cohosh are Asian Actaea species (such as A. cimicifuga, A. dahurica, and A. simplex) and American species (e.g., A. pachypoda, A. podocarpa, and A. rubra);^35,52,120 all these species also contain some 9,19-cycloartenol type triterpene glycosides and phenolic acids like North American black cohosh (Table 1). The contributions of these adulterants to the adverse effects of BCE cannot be ruled out without sufficient data. Second, BCE products on the market lack standardization, and the choice of solvents used during BCE preparation can influence the concentration of active constituents, potentially leading to various biological responses in humans.¹²¹ Also, BCE contains more than 100 different constituents, and interactions among these constituents, or between these constituents and other drugs, may further complicate toxicity assessments.¹⁰⁶ The complex composition of BCE makes comprehensive toxicity studies difficult. Third, some studies have involved limited sample sizes of participants, with either young people (28 ± 6 years) receiving standardized BCE therapy or older people (53 ± 2 years) using non-standardized BCE therapy.^79,89 Thus, human case studies are further complicated by the fact that people in their 50s and beyond often take multiple medicines and may have preexisting conditions, making it challenging to address the adverse effects attributable solely to BCE.

In summary, current available literature suggests both the beneficial and adverse effects of black cohosh products. However, little is known about black cohosh’s exact mechanisms of action and the active constituents responsible for its pharmacological and toxicological activities. The evidence from both in vivo models and in vitro cell studies suggests that BCE can exhibit toxic and genotoxic effects, with its potential to cause hepatotoxicity in humans ranging from mild reactions to acute liver damage. These concerns raised questions regarding the public health safety, especially for vulnerable populations, such as patients with autoimmune hepatitis or hormone-sensitive cancers. Further investigations, including animal studies, in vitro assessments, and more randomized controlled clinical trials in the target population, should be conducted to evaluate the safety and effectiveness of BCE, particularly its long-term effects.