Chrysanthemum

What is Chrysanthemum?

Five cultivars of Chrysanthemum × morifolium (Ramat.) Hemsl. are listed in the Chinese Pharmacopoeia as Chrysanthemi Flos, and the species has been used medicinally for at least 2,200 years. The part used is the dried capitulum, the whole flower head, harvested at full bloom and dried before the petals brown. Hao and colleagues, writing in Frontiers in Plant Science in 2022 (13:973197), place the genus at around 40 species plus a large number of hybrids and cultivars, and date the tea use to roughly 2,000 years, beginning in the Han dynasty.

The genus is messier than the label suggests. Four Chrysanthemum species turn up in the food and medicine trade: Chrysanthemum × morifolium, Chrysanthemum indicum L., Chrysanthemum lavandulifolium (Fisch. ex Trautv.) Makino, and Chrysanthemum zawadskii Herbich. Several plants once filed under Chrysanthemum have been moved out. Pyrethrum daisy is now Tanacetum cinerariifolium (Trevir.) Sch.Bip., formerly Chrysanthemum cinerariifolium, and it is the commercial source of pyrethrin insecticides. The edible leafy green sold as shungiku is now Glebionis coronaria (L.) Cass. ex Spach, formerly Chrysanthemum coronarium. Costmary is Chrysanthemum balsamita L., and it is the only "Chrysanthemum" that appears anywhere in the United States food additive regulations, at 21 CFR 172.510, cleared for use in alcoholic beverages only. Chrysanthemum flower itself does not appear in 21 CFR 182.20, the generally recognized as safe list for essential oils and natural extractives, although both chamomiles do.

The five pharmacopoeial cultivars are Boju from Bozhou in Anhui, Chuju from Chuzhou in Anhui, Gongju from Huangshan in Anhui, Hangju or Hangbaiju from Zhejiang, and Huaiju from Jiaozuo and Wenxian in Henan. They are distinguished in the monograph largely by shape and size of the flower head. Hangju runs 2.5 to 4 cm across and is dish-shaped, often stuck together in clumps. Boju is inverted conical, 1.5 to 3 cm, and discrete. Chuju, Gongju and Huaiju all sit at 1.5 to 2.5 cm and differ in how the ligulate florets twist and whether glandular dots are visible.

Common Names

• Ju Hua (菊花), the standard Chinese name for C. morifolium flower

• Ye Ju Hua (野菊花), wild chrysanthemum, the separate drug from C. indicum

• Chrysanthemi Flos, the Latin pharmacopoeial name; Chrysanthemi Indici Flos for C. indicum

• Florist's chrysanthemum, mum, Hangbaiju, Gongju, Chuju, Boju, Huaiju, Taiju, Fubaiju

• Kikuka (Japan), Gukhwa (Korea)

Primary Active Compounds

• Caffeoylquinic acids: chlorogenic acid, cryptochlorogenic acid, 3,5-O-dicaffeoylquinic acid (isochlorogenic acid A), 4,5-O-dicaffeoylquinic acid, 1,5-O-dicaffeoylquinic acid

• Flavone glycosides: luteolin-7-O-glucoside (luteoloside, cynaroside), apigenin-7-O-glucoside (apigetrin), luteolin-7-O-glucuronide, acacetin-7-O-rutinoside (linarin, buddleoside), diosmetin-7-O-glucoside, tilianin

• Flavone aglycones: luteolin, apigenin, acacetin

• Volatile oil: 1,8-cineole, camphor, endo-borneol, bornyl acetate, sabinene, filifolone, beta-myrcene

• Sesquiterpene lactones of the guaianolide type, including arteglasin A

• Polysaccharides and free amino acids

Key Note

Chrysanthemum is a medicinal and culinary flower, not a nutrient. Nothing in it is required by human metabolism, there is no deficiency state, and no intake recommendation exists from any nutrition authority. Its interest is entirely pharmacological, and the compounds that carry that pharmacology vary more between two jars of chrysanthemum than most buyers would believe.

Primary Functions & Benefits

Chrysanthemum's traditional indications in the Chinese Pharmacopoeia are wind-heat colds, headache and dizziness, red and swollen eyes, diminished eyesight, sores, and abscess swelling. Modern work maps onto a narrower set of measurable effects, most of them still preclinical.

Antioxidant capacity in vitro and in serum

The caffeoylquinic acids do most of this work. Lu and colleagues (Frontiers in Pharmacology, 2022, 13:809482) fingerprinted 30 flower head samples across five cultivars and ran the radical scavenging assays on 15 of them. Chlorogenic acid, 3,5-O-dicaffeoylquinic acid, 4,5-O-dicaffeoylquinic acid, kaempferol-3-O-rutinoside and one unassigned peak were the main contributors to overall radical scavenging. Luteolin-7-O-glucoside was not significantly correlated with any of the three assays. DPPH clearance reached 85.35 percent for the best sample, a Taiju, against 70.77 percent for a Hangbaiju in the same assay set.

Urate lowering, in combination

The single strongest clinical signal for any chrysanthemum preparation comes from DKB114, a 1:2 weight blend of C. indicum flower extract and Cinnamomum cassia bark extract. In vitro it inhibits xanthine oxidase with an IC50 of 104.4 micrograms per milliliter.

Ocular surface and eye fatigue, in combination

The one large randomized trial that includes chrysanthemum is a five-ingredient eye formula, covered under Timeline of Effects below. Chrysanthemum was never isolated.

Anti-inflammatory signaling

Linarin, diosmetin-7-O-glucoside, tilianin, isochlorogenic acids A and C and 1,3-O-dicaffeoylquinic acid tracked with suppression of IL-1beta, IL-8 and MMP9 expression in a zebrafish colitis model, though the extract tested there was chrysanthemum stem and leaf rather than flower head (Li and colleagues, Molecules, 2022, 27(7):2114). No human inflammatory endpoint has been tested with chrysanthemum alone.

Blood pressure, animal and inhalation data only

Gao and colleagues (Pharmaceutical Biology, 2016, 54:2895 to 2900) gave polyphenol-rich C. morifolium extract at 75 to 150 mg/kg daily for four weeks to renovascular hypertensive Sprague-Dawley rats. Blood pressure fell 5.7 to 9.6 percent, cardiac weight index fell 10.9 to 20.1 percent, and cardiomyocyte cross-sectional area fell 8.3 to 30.4 percent.

Forms & Standardization

Whole dried flower heads (the tea)

This is the dominant form worldwide and the one the traditional evidence rests on. Sold by cultivar name if the seller is careful, or as undifferentiated "chrysanthemum" if not. Flower head diameter is the single most useful thing a buyer can check: Hangju is 2.5 to 4 cm, the other four pharmacopoeial cultivars are 1.5 to 3 cm.

Pharmacopoeial marker set for Chrysanthemi Flos

The Chinese Pharmacopoeia sets minimum content limits for exactly three compounds: chlorogenic acid, luteolin-7-O-glucoside, and 3,5-O-dicaffeoylquinic acid. The assay itself has a known defect. Yang and colleagues (Zhongguo Zhong Yao Za Zhi, 2022, 47:1286 to 1292) tested the official method on columns from seven brands and found 3,5-O-dicaffeoylquinic acid co-eluting with isochlorogenic acid D on four of them and incompletely separated on the other three. The two methods agreed on the summed area of isochlorogenic acids A and D to within 3.0 percent, which is what proved the pharmacopoeial number is a combined peak rather than isochlorogenic acid A alone.

Pharmacopoeial marker for Chrysanthemi Indici Flos

Different drug, different standard. C. indicum is controlled on a single marker, linarin, at not less than 0.8 percent. Li and colleagues (PLOS ONE, 2023, 18(4):e0283498), measuring nine constituents across 70 germplasms, cite the finding that more than 70 percent of C. indicum material from Chinese producing areas fails that threshold.

How much the cultivar moves the numbers

Across Lu's 30 samples: chlorogenic acid 0.19 to 0.69 percent, a 3.6-fold spread; 4,5-O-dicaffeoylquinic acid 0.19 to 1.34 percent; 3,5-O-dicaffeoylquinic acid 0.33 to 1.87 percent, a 5.7-fold spread; luteolin-7-O-glucoside 0.15 to 2.36 percent, a 15.7-fold spread. Buying "chrysanthemum" without a cultivar is buying somewhere inside a 16-fold range on luteolin-7-O-glucoside and a 5.7-fold range on 3,5-O-dicaffeoylquinic acid, the single strongest correlate of the antioxidant readout.

Powdered extracts

Sold at ratios from 4:1 to 20:1 with no consistent marker. Extraction solvent matters and is almost never disclosed. The Chinese Pharmacopoeia method for C. indicum uses 70 percent aqueous methanol under 3-hour reflux on a 0.250 g sample, which is an analytical extraction, not a manufacturing one. Commercial extracts are generally water or ethanol. Chen and colleagues (Molecules, 2023, 28(2):579) compared two Imperial Chrysanthemum extracts and found the flavonoid-richer one carried cynaroside at more than 30 percent of the extract, which tells you how far two extracts of the same flower can diverge.

Essential oil

Steam-distilled, dominated by 1,8-cineole and camphor. Kim and colleagues (Molecules, 2018, 23(8):2063) detected 41 volatile organic compounds in C. indicum oil: 32 hydrocarbons, four acids, three alcohols, two ketones and one aldehyde. This is an aromatherapy and topical material, not an oral one.

What the human trials actually used

Neither of the two controlled human trials with a real endpoint used a standardized single-herb chrysanthemum extract. One used a 1:2 C. indicum and cassia blend, the other used chrysanthemum as one of five botanical inputs. No commercial chrysanthemum capsule on the market matches either.

Food Sources

Chrysanthemum is a food, and the tea is the food form. A standard brewing portion is 3 to 5 g of dried flower heads. Applying Lu's measured ranges to a 3 g serving gives 5.7 to 20.7 mg of chlorogenic acid, 9.9 to 56.1 mg of 3,5-O-dicaffeoylquinic acid, and 4.5 to 70.8 mg of luteolin-7-O-glucoside in the dry flowers before any of it reaches the water.

What brewing changes

Han and colleagues (Molecules, 2019, 24(2):329) compared a yellow cultivar, Gamguk, against a purple gamma-irradiated mutant, ARTI-Dark Chocolate. The yellow tea's phenolic yield was essentially flat across water temperatures and steeping times. The purple tea was time-sensitive and peaked at 100 degrees Celsius for 4 minutes. Both teas held ABTS radical scavenging between 61 and 76 percent under every condition tested. The practical reading: with standard yellow or white chrysanthemum, boiling water and a few minutes is enough, and longer steeping buys nothing measurable.

What cultivar changes in the cup

The purple mutant contained anthocyanins and linarin that were undetectable in the yellow tea, and carried more chlorogenic acid, acacetin-7-O-beta-glucoside and free luteolin. The yellow tea carried more luteolin-7-O-beta-glucoside, 3,5-dicaffeoylquinic acid, apigenin-7-O-beta-glucoside and apigenin. Same drink by name, different compound set.

Culinary use beyond tea

Petals go into congee, wine, hot pot and desserts in China, and into sashimi garnish and vinegared dishes in Japan. These are flavor and color quantities, well below anything pharmacological.

The green that is not this plant

Shungiku, the chrysanthemum green sold in East Asian groceries, is Glebionis coronaria, a different genus. Eating shungiku does not deliver Ju Hua chemistry.

Is food intake sufficient?

For anything chrysanthemum plausibly does, yes. Two to three cups of properly identified flower tea daily delivers more total caffeoylquinic acid and flavone glycoside than most capsules disclose, costs less, and comes with an inspectable raw material. This is a case where the flower beats the pill.

Who Should Take Chrysanthemum

• People who want a caffeine-free hot drink with measurable polyphenol content. The tea is the honest use case.

• Adults with asymptomatic hyperuricemia in the 7.0 to 9.0 mg/dL range, who are the only population in which a chrysanthemum-containing product has moved a hard clinical number in a placebo-controlled trial, and only as part of a cassia blend.

• People using chrysanthemum in a traditional formula under a practitioner who has selected the cultivar, such as Sang Ju Yin for early wind-heat presentations.

• People with screen-related eye fatigue who understand the trial evidence is for a five-ingredient lutein-based formula, not for chrysanthemum on its own.

• Anyone who wants a polyphenol source that is inspectable before purchase. Whole flower heads can be looked at, smelled and measured. Powder cannot.

Who has no evidence-based reason to take it: anyone buying it as a standalone capsule for blood pressure, liver support, immunity, or vision. Those claims rest on rat and cell data.

Who Should AVOID or Use Caution

Contraindications

• Anyone with known Asteraceae (Compositae) contact allergy or sesquiterpene lactone sensitivity. Hausen and Schulz (Archives of Dermatological Research, 1976, 255:111 to 121) isolated arteglasin A, a guaianolide sesquiterpene lactone, from C. indicum and confirmed it as an allergen in sensitized guinea pigs. Bleumink and colleagues (Contact Dermatitis, 1976, 2:81 to 88) patch-tested six patients with florist's chrysanthemum allergy against 25 sesquiterpene lactones and got positives to 15 of them, with alantolactone positive in all six.

• Anyone taking ciclosporin, tacrolimus or sirolimus. See Interactions.

• Known allergy to ragweed, mugwort, feverfew, arnica, tansy or chamomile, all of which share the alpha-methylene-gamma-butyrolactone structural motif that drives the reaction.

Use Caution

• People on antihypertensive drugs. The rat and inhalation data both point downward on blood pressure, and additive hypotension has not been ruled out.

• People with chronic actinic dermatitis. Ross and colleagues (Contact Dermatitis, 1993, 29:84 to 87) found that of 48 sesquiterpene-lactone-positive patients investigated for photosensitivity, 29 had abnormal monochromatic irradiation tests.

• Anyone taking a narrow-therapeutic-index drug cleared by CYP3A4.

• Pregnant and breastfeeding women. There is no human safety data at medicinal doses. Occasional tea is a different exposure from a daily concentrated extract.

• Anyone taking retinol or high-dose vitamin A. Wang and colleagues (Journal of Food Science, 2012, 77:H121 to H127) found C. morifolium extract raised basal serum retinol dose-dependently in rats and increased plasma retinol after an oral 45 mg/kg retinol dose.

Critical Safety Point

The allergy risk is not to chrysanthemum the concept, it is to a class of molecules shared across the daisy family, and sensitization is often occupational. In Bleumink's series, pyrethrum extract produced a positive patch test in one of the six chrysanthemum-allergic patients while purified pyrethrins were negative in all six. The allergen travels with the plant material, not with the insecticidal esters. A gardener or florist who reacts to handling mums should treat chrysanthemum tea as a plausible oral exposure to the same lactones.

Recommended Dosages

Dried flower heads, tea

3 to 5 g per brewing, one to two times daily. The Chinese Pharmacopoeia gives 5 to 10 g per day for Chrysanthemi Flos, and that is the ceiling the brewing portions should add up to. It is a pharmacopoeial figure drawn from clinical tradition, not a trial-derived dose, and no dose-response study in humans exists for the tea.

Chrysanthemi Indici Flos (Ye Ju Hua)

Dosed higher than Ju Hua in the Chinese Pharmacopoeia, 9 to 15 g per day against 5 to 10 g, and indicated first for sores, boils and abscesses, though the monograph also lists red eyes, headache and dizziness. It is not interchangeable with Ju Hua, and a product that lists only "chrysanthemum" gives you no way to tell which one you have.

Powdered extract capsules

No evidence-based dose exists. Products range from 250 mg to 1,000 mg per capsule at undeclared ratios, and there is no human trial of a single-herb chrysanthemum extract at any dose.

Combination products with human data

DKB114, the 1:2 C. indicum to Cinnamomum cassia blend, was tested for 12 weeks at 2,000 mg per day of the DKB114 ingredient, delivered as four 1 g tablets totaling 4 g of product, two in the morning and two at night, after meals. It was randomized in 80 adults and analyzed in 58. The five-ingredient eye formula was tested for 90 days at doses defined by their lutein content, 6, 10 or 14 mg, not by chrysanthemum content.

Essential oil

Inhalation and topical only, diluted. Not for oral use.

Duration

The DKB114 trial showed no separation from placebo at six weeks and a significant difference at 12. The eye formula separated at 45 days. For anything urate-related, 12 weeks is the fair trial. For tea used as a daily drink, there is no defined course and no evidence of cumulative risk across the 26-week rat study described under Toxicity.

Timing & Administration

Chrysanthemum tea is traditionally taken after meals and in the afternoon, consistent with its traditional cooling classification. There is no pharmacokinetic reason to prefer a time of day.

Brewing

Use water at 90 to 100 degrees Celsius. Han's data showed no benefit from extended steeping in yellow chrysanthemum, and 4 minutes at 100 degrees was optimal for the pigment-rich purple cultivar. Flowers can be re-steeped two or three times, with declining yield.

With or without food

Jia and colleagues (BioMed Research International, 2020, 2020:8420409) dosed rats orally with Flos Chrysanthemi extract at 10 g/kg and tracked 13 compounds. Chlorogenic acid reached Tmax at 0.25 hours, luteolin and apigenin at 0.50 hours, and cynaroside at 6.38 hours. Nothing in that profile argues for fasting administration.

Capsules

Take with food. The extracts are polyphenol-dense and acidic on an empty stomach.

What not to do

Do not take chrysanthemum within two hours of a CYP3A4 substrate with a narrow therapeutic window. Spacing is not a fix, but it is better than co-administration.

Timeline of Effects

Immediate, within minutes

Inhaled C. indicum essential oil lowered systolic blood pressure and heart rate and shifted the electroencephalogram toward theta and alpha with reduced beta and gamma in Kim's 2018 study. This is an acute aroma response, it is not a treatment for hypertension, and the study reported no sustained follow-up.

Hours, single dose

In Chen's 2023 randomized parallel trial, 37 healthy adults with a mean age of 25.2 years and mean BMI of 20.9 were split between a high-fat meal alone and the same meal with Imperial Chrysanthemum extract, so roughly 18 or 19 received the extract. Postprandial glucose was significantly lower in the extract group at the 1-hour mark, and total antioxidant capacity, superoxide dismutase and glutathione rose while malondialdehyde fell. Notably, incremental area under the curve did not differ between groups. The effect was on the shape of the curve, not the total exposure.

45 days

The five-ingredient eye formula reduced total eye fatigue symptom score and increased visuognosis persistence time at the first follow-up visit.

90 days

Same trial: Schirmer test tear secretion improved in the 10 mg and 14 mg lutein arms, first and average tear break-up time and tear meniscus height rose, and macular pigment optical density increased. Retinal thickness and volume did not change.

12 weeks

DKB114 showed nothing at six weeks and a 0.58 mg/dL serum urate drop at 12.

When to conclude it is not working

If you are taking chrysanthemum for a symptom rather than as a beverage, 12 weeks is the honest limit. Nothing in the literature supports a longer wait.

Benefits of Taking Chrysanthemum

Serum urate reduction, moderate evidence, combination product only

Park and colleagues (Nutrients, 2020, 12(12):3794, trial registration KCT0002840) randomized 80 adults with asymptomatic hyperuricemia between 7.0 and 9.0 mg/dL to DKB114 or placebo for 12 weeks. Every efficacy figure below comes from the per-protocol set of 58, 32 on DKB114 and 26 on placebo, not from the 80 randomized. Serum uric acid fell 0.58 plus or minus 0.86 mg/dL in the treatment group to 7.37 plus or minus 0.92, against 0.02 plus or minus 0.93 in placebo to 7.67 plus or minus 0.89, p equals 0.0229. C-reactive protein rose significantly within the placebo group (p equals 0.0324) but not within the treatment group, between-group p equals 0.0443. Apolipoprotein B fell 4.75 plus or minus 16.69 mg/dL on treatment and rose 3.13 plus or minus 12.64 on placebo, p equals 0.0189. The responder secondary was null: 11 of 32 on DKB114 reached a serum urate below 7.0 mg/dL against 4 of 26 on placebo, not a significant difference. The authors note that the recruited groups may be too small to draw conclusions about efficacy, and that the placebo arm shrank most in per-protocol selection. Chrysanthemum was one third of the blend by weight.

Eye fatigue and dry eye, moderate evidence, five-ingredient formula

Kan and colleagues (American Journal of Clinical Nutrition, 2020, 112:334 to 342, ChiCTR1800018987) randomized 360 adults across four arms for 90 days; 360 is the number randomized, and the published report does not let you check how many were analyzed at each visit. The formula combined lutein ester, zeaxanthin, and extracts of blackcurrant, chrysanthemum and goji berry. All three doses beat placebo on total eye fatigue score and visuognosis persistence at day 45 and day 90.

Postprandial glucose and antioxidant status, preliminary

Chen 2023, described above, 37 adults across two parallel arms.

Antioxidant capacity in the cup, well established in vitro

61 to 76 percent ABTS radical scavenging across every brewing condition Han tested.

What the evidence does not show

No trial has shown chrysanthemum alone lowers blood pressure in humans, improves liver enzymes in humans, treats conjunctivitis, prevents migraine, or affects any cancer endpoint. No trial has isolated chrysanthemum from the products above. Every effect listed as moderate evidence is attributable to a mixture.

Potential Negatives & Side Effects

Allergic contact dermatitis

The best-documented adverse effect. Ross's four-year series patch-tested 7,420 eczema patients and found 135 positives (1.8 percent) to sesquiterpene lactone mix 0.1 percent, of which 114 (84 percent) were judged clinically relevant. Presentations split across hands alone (36 percent), hands and face (24 percent), generalized eczema (20 percent) and face alone (11 percent).

Oral and systemic allergic reactions

Rarer and less well characterized than the contact form. Anyone with confirmed Asteraceae contact allergy should treat oral chrysanthemum as an untested exposure rather than a safe one.

Photosensitivity

Associated with sesquiterpene lactone sensitization rather than with chrysanthemum specifically, but 29 of 48 patients investigated in Ross's series had abnormal cutaneous monochromatic irradiation tests.

Blood pressure lowering

Mild and probably only relevant if you are already on antihypertensive medication.

Sulfur dioxide sensitivity

A real and underappreciated route. Sulfite-sensitive asthmatics reacting to a "pure herbal tea" are reacting to the processing, not the plant. See Quality.

What has not been reported

No hepatotoxicity signal, no nephrotoxicity signal, no withdrawal effect, no dependence, and no documented interaction with alcohol.

Deficiency Symptoms

Chrysanthemum is a flower people drink, not a nutrient the body requires. No enzyme uses it, no transport protein carries it, no tissue stores it, and no intake level has ever been set by any nutrition authority. There is no such thing as chrysanthemum deficiency, and anyone framing low intake as a shortfall is selling something.

What chrysanthemum addresses

Symptoms, in a traditional framework, and biomarkers in a small number of trials. Eye fatigue from screen use, in a five-ingredient formula randomized in 360 adults over 90 days. Serum urate, in a two-herb blend randomized in 80 adults and analyzed in 58 over 12 weeks. Postprandial oxidative stress markers after a fatty meal, in 37 adults split between two arms over 8 hours. None of these are deficiency corrections, and the Chinese Pharmacopoeia frames all of its chrysanthemum indications as conditions to treat rather than nutrients to replace.

Bottom line

Skipping chrysanthemum costs you nothing physiological. It is a low-risk drink carrying real polyphenols, with two combination trials behind it. Treat it as a beverage with some upside, not as something your body is missing.

Toxicity Symptoms

At high intake

Li and colleagues (Journal of Food Science, 2010, 75:T105 to T109) gave Sprague-Dawley rats a single oral dose of 15 g/kg of ethanolic C. morifolium extract and observed them for 14 days. No treatment-related deaths, and the maximum tolerated dose was estimated above 15 g/kg. In the 26-week study, gavage at 320, 640 and 1,280 mg/kg daily followed by a four-week recovery period produced no changes in body weight, food or water intake, hematology, blood biochemistry, organ weight or histopathology at any dose. There is no established human toxic dose because none has been reached.

Signs to reduce or stop

• Rash, itching, swelling of the lips or eyelids, or eczema on the hands after handling or drinking. This is the allergic reaction and it does not become safer with continued exposure.

• Lightheadedness on standing, especially if you take an antihypertensive.

• Wheezing or chest tightness after drinking a particular brand of dried flowers. Suspect sulfur dioxide residue and switch to a sun-dried, cultivar-named product.

• Any new symptom while taking a transplant immunosuppressant. Drug levels are the relevant measurement, and the prescribing team is who orders them.

General note

The narrow safety concern with chrysanthemum is not the plant's own chemistry. The 26-week rat study is clean at a dose 1,280 mg/kg daily, which is far beyond anything a human tea drinker achieves. The realistic hazards are allergy in daisy-sensitive people, drug interaction through CYP3A4, and contaminants introduced during drying and cultivation.

How Chrysanthemum Works

Caffeoylquinic acids as the antioxidant backbone

Chlorogenic acid and the dicaffeoylquinic acids are the mass-dominant phenolics in the flower and the main correlates of radical scavenging in Lu's spectrum-effect analysis. They are also the compounds sulfur fumigation destroys, which is why processing and mechanism are the same conversation.

Flavone glycosides and aglycones on different clocks

Jia's rat pharmacokinetics make the two-phase pattern explicit. Chlorogenic acid peaked at 0.25 hours with a half-life of 0.20 hours, essentially a flash exposure. Luteolin peaked at 0.50 hours with a half-life of 8.89 hours and a Cmax of 2,079.55 plus or minus 307.09 ng/mL. Apigenin-7-glucoside had a half-life of 13.87 hours. Cynaroside peaked latest at 6.38 hours with the highest Cmax of the set, 2,547.84 plus or minus 1,121.18 ng/mL. Acacetin, hyperoside, isoquercitrin, tilianin and hesperidin appeared only at the earliest sampling points and could not be curve-fitted at all.

Xanthine oxidase inhibition

The urate mechanism. DKB114 inhibits xanthine oxidase in vitro with an IC50 of 104.4 micrograms per milliliter, and in hyperuricemic rats at 100, 150 and 200 mg/kg it both reduced urate synthesis and increased urinary excretion, with 200 mg/kg strongest.

Cardiac energy metabolism in hypertensive rats

Gao's mechanism work found polyphenol-rich extract suppressed myocardial HIF-1alpha and GLUT-4 and raised PPARalpha, CPT-1a and PDK-4, shifting the hypertrophic heart back toward fatty acid oxidation.

Volatile oil and the nervous system

1,8-cineole and camphor are the dominant volatiles and are the stated basis of the acute blood pressure and EEG shifts on inhalation. Kim and colleagues traced the volatile organic compounds they identified in C. indicum oil back to these two constituents.

What the mechanism work does not establish

None of this has been demonstrated in humans with chrysanthemum alone. The pharmacokinetics are rat data at 10 g/kg. The cardiac mechanism is rat data. The xanthine oxidase inhibition is in vitro and involves cassia as well.

Synergistic Supplements

Cinnamomum cassia bark extract

The only pairing with a positive randomized human trial behind it, at 1:2 chrysanthemum to cassia by weight. The combination was described as synergistic on urate lowering in the preclinical work that preceded DKB114.

Lutein and zeaxanthin

Chrysanthemum's role in the Kan formula is as a flavone source alongside 6 to 14 mg of lutein ester plus zeaxanthin, blackcurrant and goji berry. The macular pigment optical density increase is attributable to the carotenoids. If you want the trial's eye effect, the carotenoids are the part to buy.

Honeysuckle flower (Lonicera japonica)

Paired with chrysanthemum in classical formulas and in commercial herbal teas. Both are chlorogenic acid sources, so the pairing is additive rather than complementary. Worth knowing that honeysuckle is a far weaker CYP3A4 inhibitor, IC50 1,466.3 micrograms per milliliter against chrysanthemum's 95.7.

Goji berry (Lycium barbarum)

The traditional eye pairing, and one of the five ingredients in the Kan formula.

What to skip

Stacking chrysanthemum extract on top of chrysanthemum tea. You get more of the same phenolics with less control over identity and processing.

Interactions & What NOT to Take

Ciclosporin and sirolimus, documented clinical case

The strongest interaction signal in the chrysanthemum literature. Dufay and colleagues (Journal of Pharmacy and Pharmacology, 2014, 66:1478 to 1490) investigated a clinical case of toxic blood levels of both drugs after a patient consumed a Hong Kong herbal tea blended from 24 ingredients. Testing the six universal constituents against CYP3A4, chrysanthemum was the most potent inhibitor of all of them, with an IC50 of 95.7 micrograms per milliliter. For comparison, dandelion was 140.6, licorice 148.4, bishop's weed 185.5, field mint 1,153.3 and Japanese honeysuckle 1,466.3.

Tacrolimus

Not tested with chrysanthemum, but a CYP3A4 substrate with the same narrow therapeutic window as ciclosporin and sirolimus. Treat it the same way.

Other CYP3A4 substrates to name specifically

Simvastatin, atorvastatin, midazolam, apixaban, rivaroxaban, amiodarone, nifedipine, and ivabradine. Ashour and colleagues (Pharmacognosy Magazine, 2017, 13:300 to 308) found aqueous C. indicum extract inhibited CYP3A4 by more than 85 percent at 100 micrograms per milliliter, placing it among the eight strongest of 57 traditional Chinese medicine plants tested.

Retinol and high-dose vitamin A

The 2012 rat study showed dose-dependent increases in both basal and post-dose plasma retinol with chrysanthemum extract. CYP1A1, CYP1A2 and CYP2B1 were themselves induced by the extract, but inducing those enzymes would lower retinol rather than raise it, so the induction cannot explain the higher exposure and the mechanism is unexplained. Anyone on isotretinoin or a high-dose retinol supplement should treat this as unresolved.

Antihypertensives

Amlodipine, lisinopril, losartan and their relatives. The signal is small and the data are animal, but additive hypotension is the plausible direction.

Allopurinol and febuxostat

Not a dangerous interaction, but chrysanthemum-cassia products work on the same enzyme. Duplicating the mechanism without telling your prescriber makes urate targets harder to interpret.

What has not been studied at all

Warfarin. There is no chrysanthemum warfarin interaction study in either direction. The absence of data is not reassurance.

Quality, Testing & Adulteration

This is where chrysanthemum actually fails buyers, and the failures are documented.

Species and cultivar substitution

Gu, Scotti, Reich, Kirchhof, Booker and Heinrich (South African Journal of Botany, 2022, volume 148) bought 63 commercial chrysanthemum samples across four markets: 21 from China, 23 from the United States, 13 from Europe and 6 from Taiwan. With two botanical reference standards added, the 65 samples analyzed comprised 50 C. × morifolium and 15 C. indicum. They note that linarin, the Chinese Pharmacopoeia marker for C. indicum, also turned up in some C. × morifolium samples, which undermines its use as a species discriminator. HPTLC fingerprints did separate the two species cleanly.

Cadmium

The most serious contaminant in that survey. Against a 1 mg/kg limit, 7 of 60 samples exceeded it: four from the European market, two from China and one from the United States. Twenty percent of C. indicum samples exceeded, against 9 percent of C. × morifolium. The worst sample, coded TAI-2, measured 4.5 mg/kg, four and a half times the limit. Copper exceeded its 20 mg/kg limit in 2 of 60. Arsenic and lead were within limits in every sample. Two samples were found to contain live insects after purchase.

Sulfur fumigation

Kang and colleagues (Zhongguo Zhong Yao Za Zhi, 2018, 43:242 to 247) compiled sulfur dioxide residue data on 862 batches across 35 Chinese medicinal materials. Measured against the Chinese Pharmacopoeia limits of 150 and 400 mg/kg, the average over-standard rate was 52.43 percent. Judged against the looser 500 mg/kg vegetable additive limit it fell to 29.47 percent. Either way, more than a quarter of the material in the trade exceeds the standard it should meet. Kang's own reading is less alarmed than the numbers look: they class sulfur dioxide as a low-toxicity substance, far less dangerous than aflatoxin, and argue that a small residue does not make the material toxic. Their recommendation is a revised, workable limit rather than a ban.

Pesticide residue

Feng and colleagues (Food Chemistry, 2022, 383:132431) profiled 122 samples of chrysanthemum and Lusterleaf Holly tea from two Chinese producing regions. Almost every sample contained pesticides, at concentrations from 0.0005 to 10.305 mg/kg. Nearly 40 percent of carbendazim and imidacloprid findings in the chrysanthemum teas exceeded the levels permitted under EC Regulation No. 396/2005. Eight pesticide transformation products were identified alongside their parent compounds. Their acute, chronic and cumulative risk assessments came out acceptable.

Wang and colleagues (Journal of Hazardous Materials, 2025, 493:138363) analyzed 111 market chrysanthemum products from China, then traced 17 high-frequency pesticides through the whole production chain. Field dissipation followed first-order kinetics with half-lives of 0.8 to 8.8 days. The decisive step is drying, which concentrates residues, correlated with each compound's log Kow and water solubility. Brewing transferred between 0.5 and 112.3 percent of a residue from dry flower into the infusion, rising with water solubility. Their combined risk assessment still came out acceptable for chrysanthemum consumption. A residue figure on a dry-weight certificate tells you very little about what ends up in the cup.

Pyrrolizidine alkaloids

Chrysanthemum does not make them, but co-harvested weeds do, and chrysanthemum tea falls under the EU's herbal infusion category. Commission Regulation (EU) 2020/2040, applicable from 1 July 2022, set 200 micrograms per kilogram for dried herbal infusions generally, 400 for a named list including chamomile and rooibos, 150 for Camellia sinensis tea, and 75 for dried infusions intended for infants and young children. Food supplements containing herbal ingredients are capped at 400. Those limits now sit in Commission Regulation (EU) 2023/915, Annex Section 2.4, which replaced the instrument 2020/2040 amended without changing the numbers.

Third-party testing

There is no USP monograph for chrysanthemum flower, and neither USP Verified nor NSF Certified for Sport lists a chrysanthemum product. The Chinese Pharmacopoeia is the quantitative monograph the trade works to, and its assay has the co-elution defect described above. This is an ingredient with no Western quality infrastructure at all.

What a certificate of analysis should show

• The binomial with authority and the cultivar name, not "chrysanthemum flower"

• Chlorogenic acid, luteolin-7-O-glucoside and 3,5-O-dicaffeoylquinic acid by HPLC, with the chromatographic conditions stated so you can see whether isochlorogenic acid D was resolved

• Sulfur dioxide residue in mg/kg with the method named

• Cadmium, lead, arsenic and mercury by ICP-MS, with cadmium reported against 1 mg/kg

• A multi-residue pesticide panel that names carbendazim and imidacloprid explicitly

• Drying method: sun-dried, hot-air dried or sulfur-fumigated

What a buyer can verify without a lab

Buy whole flower heads, not powder. Measure the diameter: Hangju is 2.5 to 4 cm, the others 1.5 to 3 cm, so a bag of 2 cm heads sold as Hangbaiju is mislabeled. Ask the seller for the drying method and the county of origin in writing. Duan and colleagues (BMC Plant Biology, 2024) sequenced the chloroplast genomes of Boju and Huaiju and compared them against published genomes of Hangbaiju and Gongju, finding all four spanning just 151,057 to 151,109 base pairs with no significant rearrangements, so routine single-locus DNA barcoding confirms the species but cannot tell you the cultivar. Your eyes and a ruler are currently better than a barcode.

Special Considerations

Transplant recipients

The clearest do-not-use population in this entire issue. The CYP3A4 case that started Dufay's investigation involved exactly this scenario.

Occupational exposure

Florists, nursery workers and chrysanthemum growers are the classic sesquiterpene lactone sensitization population. Sensitization acquired through the hands is relevant to what you drink.

Sulfite-sensitive asthmatics

Given a 52.43 percent over-standard rate for sulfur dioxide in the Chinese medicinal material trade, sun-dried, origin-declared flowers are the lower-exposure option for this group.

Pregnancy and lactation

No human data at medicinal doses. The rat reproductive toxicology has not been published. Occasional culinary tea is a different proposition from daily extract capsules, and the honest answer is that nobody knows where the line is.

Children

The EU sets its strictest pyrrolizidine alkaloid limit, 75 micrograms per kilogram dry, for infusions intended for infants and young children, which is a reasonable signal about herbal infusions generally in this age group.

People buying for a specific cultivar's reputation

Hangbaiju commands a premium. Because cultivar cannot be confirmed by routine DNA testing and marker content overlaps heavily between cultivars, cultivar claims are the least verifiable thing on a chrysanthemum label and the most likely place for an honest premium to be charged for something else.

Geographical origin claims

Stable isotope ratio analysis across six geographical indication areas of white chrysanthemum, and 3D fluorescence spectroscopy with chemometrics for Hangbaiju specifically, have both been developed to authenticate origin, which tells you the fraud is common enough to be worth building instruments for.

Research Status & Evidence Quality

Tier 1, established

• Chemical composition. Hundreds of compounds characterized across the genus, including 183 monoterpenoids and 207 sesquiterpenoids reported genus-wide, with 17 monoterpenes and 27 sesquiterpenoids in C. indicum specifically.

• In vitro antioxidant activity, reproducible across DPPH, ABTS, FRAP and hydroxyl radical assays.

• Sesquiterpene lactone allergenicity, documented since 1976 and confirmed in large patch test series.

• CYP3A4 inhibition in vitro, two independent laboratories.

Tier 2, moderate

• Serum urate reduction, one 12-week randomized double-blind placebo-controlled trial, 80 randomized and 58 analyzed, combination product.

• Eye fatigue and dry eye, one 90-day randomized double-blind placebo-controlled trial, 360 randomized, five-ingredient formula.

• Cultivar-driven variation in marker content, multiple independent surveys.

• Sulfur fumigation damage to marker compounds, replicated within one research group across two publications.

Tier 3, preliminary

• Postprandial glucose and antioxidant status, n equals 37, single meal.

• Acute blood pressure and EEG effects from essential oil inhalation, small uncontrolled study.

• Antihypertensive and cardioprotective effects, rat only.

• Anti-inflammatory and hepatoprotective effects, cell and animal only.

Tier 4, no credible human evidence

Immune support, detoxification, conjunctivitis treatment, migraine prevention, weight loss, skin whitening, and any cancer claim.

Research Limitations

Three problems dominate. First, no human trial has ever tested chrysanthemum alone. Both trials with real endpoints used mixtures, and in both cases the co-ingredients have their own mechanisms, so attribution to chrysanthemum is not possible. Second, no trial has stated its material's cultivar, marker content or drying method, which means the inputs are unspecified across a range that varies up to 15.7-fold on luteolin-7-O-glucoside. Third, the quantitative work is almost entirely Chinese and Korean and almost entirely on material sourced domestically, while the product a Western buyer picks up passed through a supply chain nobody in that literature sampled. Gu's 2022 survey is the only study that looked at what is actually sold in Europe and North America, and it found 7 of 60 samples over the cadmium limit.

There is also no dose-response study of any kind. Every dose in the traditional literature comes from clinical tradition, and the two trial doses come from proprietary formulation decisions.

Summary & Key Takeaways

Bottom Line

Chrysanthemum is a well-tolerated flower tea, rich in caffeoylquinic acids, with a real pharmacology and almost no human trial evidence of its own. The two positive randomized trials both used it as one component of a mixture, and neither tells you what chrysanthemum alone does. Buy it as tea, in whole flower heads, from a seller who will name the cultivar and the drying method. Skip the standalone capsule: it costs more, discloses less, and no trial supports it. Where a price premium is charged for a cultivar, understand that cultivar is the one claim on the label that routine testing cannot confirm.

Key Safety Points

• Chrysanthemum was the most potent CYP3A4 inhibitor among six herbal tea constituents tested after a documented clinical toxicity case involving ciclosporin and sirolimus, IC50 95.7 micrograms per milliliter. Tacrolimus is the same kind of narrow-window CYP3A4 substrate but has not been tested with chrysanthemum. This is the interaction to raise with a prescriber before drinking it.

• Anyone with daisy family contact allergy should treat chrysanthemum as a likely oral allergen. Arteglasin A and related guaianolides are the responsible molecules.

• Sulfur dioxide from fumigation is a genuine exposure. A 52.43 percent over-standard rate across 862 batches of Chinese medicinal material is not a rounding error.

• Cadmium is the heavy metal to ask about, not lead. One of 60 surveyed samples measured 4.5 mg/kg against a 1 mg/kg limit.

• Chrysanthemum is safe in rats at 1,280 mg/kg daily for 26 weeks with no observed effect. The risk profile is about contaminants, allergy and drug interactions, not intrinsic toxicity.

Special Note

The single most useful thing on a chrysanthemum label would be the drying method, and it is almost never there. Sun-dried and sulfur-fumigated flowers from the same field are chemically different products, with different marker content, different antioxidant activity and different residues, and only one of them is what the traditional evidence was built on. Until sellers print it, ask them, and buy from the ones who answer.