Cistanche

What is Cistanche?

Cistanche deserticola Y.C. Ma entered CITES Appendix II in 2000. Cistanche tubulosa (Schenk) Wight, the species inside most Western capsules, was never listed at all. Both are sold as "cistanche," and that single fact governs almost everything confusing about this ingredient.

Cistanche is a holoparasitic plant in the Orobanchaceae. It has no chlorophyll and no functional root system of its own. It germinates only when a chemical signal from a compatible host root reaches the seed, then attaches by a haustorium and draws water, sugars and minerals from that host for three to five years before pushing a fleshy stem above the desert sand. The part harvested and sold is that succulent stem, cut in spring before or just after emergence, not the root, despite the wording used in most retail copy and in the CITES annotation itself.

Host specificity is not a botanical footnote here. C. deserticola parasitizes the saxaul shrubs Haloxylon ammodendron and Haloxylon persicum. C. tubulosa parasitizes Tamarix chinensis and other Tamarix species. C. salsa (C.A. Mey.) G. Beck grows on Kalidium, and C. sinensis G. Beck on Reaumuria, Ammopiptanthus and Potaninia. Each pairing produces a different chemical profile, and the host plant is the single biggest constraint on supply.

Common Names

Rou cong rong (肉苁蓉), Cistanches Herba, desert ginseng, desert hyacinth, suosuo dayun (the C. deserticola trade grade), guanhua roucongrong (the C. tubulosa grade), broomrape, and in Japan, nikuju you.

Primary Active Compounds

Phenylethanoid glycosides account for more than 80 percent of the characterized active fraction. The two marker compounds are echinacoside and acteoside, the latter also called verbascoside. Others isolated from the genus include isoacteoside, tubuloside A, tubuloside B, cistanoside A, cistanoside C, campneoside I, campneoside II, 2'-acetylacteoside, crenatoside, decaffeoylacteoside and the cistantubulosides B1, B2, C1 and C2. Beyond the phenylethanoids, 27 iridoids and a group of lignans have been described, plus polysaccharides, mannitol, betaine and fructose.

Key Note

Cistanoside A has been detected in C. tubulosa and not in C. deserticola, and 2'-acetylacteoside has been reported only in C. deserticola among the four species examined in one HPLC survey. The two species are chemically distinguishable in a laboratory. They are not distinguishable on a supplement label that says only "Cistanche."

Primary Functions & Benefits

Cistanche is classified in Chinese medicine as a gentle yang tonic for the kidney, with a secondary role as a moistening laxative. The modern research program has largely followed three of those threads and added a fourth.

• Bowel function. The mannitol in C. deserticola is an osmotic laxative in its own right, which is the most mechanistically straightforward thing cistanche does. The Chinese Pharmacopoeia indication for dry constipation in the elderly predates any of the pharmacology.

• Physical function in older adults. Two Japanese placebo-controlled trials from the Tohda group at the University of Toyama tested C. tubulosa ethanolic extract on walking ability and on postoperative sensory recovery, with modest positive results on specific endpoints and null results on others.

• Cognition. A C. tubulosa water extract has been a licensed prescription drug in China since 2005 for vascular dementia at 1,800 mg per day. Western supplement claims lean on this without mentioning that the indication is a disease indication in a different regulatory system.

• Androgen and performance claims. This is where the marketing sits and where the human evidence is thinnest. One randomized trial in 48 men, published in 2025 and sponsored by a cistanche manufacturer, is doing nearly all of the work.

The phenylethanoid glycosides themselves are broadly antioxidant and anti-inflammatory in cell and rodent models, acting on NF-kappaB, Nrf2/HO-1, PI3K/Akt and MAPK signaling. That breadth is a reason for caution rather than enthusiasm, because a compound reported to modulate everything has usually been tested at concentrations the gut will never deliver.

Forms & Standardization

Raw sliced herb (yin pian). The Chinese Pharmacopoeia article Cistanches Herba, and its daily dose of 6 to 10 g in decoction. This is the form all the traditional literature describes.

Wine-processed cistanche (jiu congrong). The standard pharmacopoeial processing step, in which sliced stem is steamed with rice wine. Processing changes the glycoside ratios, and animal work shows wine-processed cistanche polysaccharides raise acteoside absorption more than unprocessed.

Water extract, 70 percent phenylethanoid glycosides. The pharmaceutical-grade material. The specification EFSA assessed in 2021 requires at least 70 percent total phenylethanoid glycosides and 25 to 45 percent echinacoside; across seven applicant batches the measured values were 73.5 to 79.8 percent and 31.9 to 38.9 percent. Production is water extraction, concentration, ethanol precipitation and spray drying.

30 percent ethanolic extract. The form used in both Toyama trials. The locomotive syndrome material was 17.48 percent echinacoside and 5.41 percent acteoside; the cervical myelopathy material was 27.8 percent echinacoside and 6.2 percent acteoside.

60 percent ethanolic extract, 2:1. The form used in the 2025 muscle strength trial. The paper reports the solvent, the temperature (70 degrees C), three 2 hour extractions and a 2:1 drug to extract ratio, and does not report echinacoside or acteoside content at all.

Retail capsules. The common US offering is a 500 mg or 1,000 mg C. tubulosa capsule claiming 50 percent echinacoside and 10 percent acteoside. That echinacoside figure sits above the entire 25 to 45 percent range of the pharmaceutical specification EFSA reviewed, and above every one of the seven batches in that dossier. It is not impossible, but it is a claim that should come with an HPLC chromatogram rather than a number on a panel.

Two marker standards exist and they are not the same. The Chinese Pharmacopoeia requires echinacoside plus acteoside combined at not less than 3 mg/g, or 0.30 percent, for C. deserticola, and not less than 15 mg/g, or 1.5 percent, for C. tubulosa. A fivefold difference in the legal minimum for two plants sold under one name is the clearest signal available that they are not interchangeable.

Food Sources

Cistanche is a food in exactly one place. In Inner Mongolia, Gansu, Ningxia and Xinjiang it is sliced into soups and congee, steeped as tea, and steeped in liquor as cistanche wine. Kazakh communities in Altay, Xinjiang record it in ethnobotanical surveys as a medicinal food plant. China's National Health Commission and State Administration for Market Regulation formalized this in announcement 2023 No. 9, dated 9 November 2023, which added desert cistanche to the catalog of substances traditionally used as both food and Chinese medicinal material, bringing that list to 102 entries. The China National Center for Food Safety Risk Assessment had already reviewed Alxa desert cistanche against food safety requirements in 2016, and by June 2022 some 60 registered Chinese health food products contained C. deserticola or its extracts.

There is no Western food source. Cistanche appears in no grocery category, no national food composition database entry of consequence, and no traditional European or American preparation. The compounds are not unique to it: acteoside was first isolated from Verbascum and is present in olive leaf, lemon verbena and Rehmannia glutinosa, and echinacoside is named for Echinacea, where it occurs in Echinacea angustifolia root.

The honest arithmetic on food intake: a Chinese culinary preparation using the pharmacopoeial 6 to 10 g of sliced stem, at the C. deserticola minimum of 0.30 percent combined markers, delivers 18 to 30 mg of echinacoside plus acteoside. The cervical myelopathy trial delivered 816 mg of those two compounds daily. Food use and trial dosing are separated by a factor of roughly 30, so eating cistanche is a cultural practice, not a route to the doses in the literature.

Who Should Take Cistanche

Cistanche is not a nutrient and nobody needs it. The people with a defensible reason to try it are narrow.

• Adults over 60 with mild decline in walking speed and step width, which is the population where the 12 week locomotive syndrome trial found its one significant result, and only in the subgroup over 60 years old.

• People recovering from decompression surgery for degenerative cervical myelopathy with persistent sensory symptoms, which is the population in the 24 week trial that improved on the visual analog scale for pain and numbness.

• Adults with chronic dry constipation who have already tried and tolerated ordinary osmotic agents, since mannitol and the polysaccharide fraction give a plausible mechanism and Chinese use for this indication is 1,800 years old.

• Untrained men beginning a resistance program who understand that the strength and testosterone findings come from one 8 week trial of 48 men with 12 per arm, sponsored by a cistanche producer.

Anyone taking it specifically to raise testosterone should understand what the one relevant trial actually reported, which is covered below in the benefits section.

Who Should AVOID or Use Caution

Contraindications

• Pregnancy and breastfeeding. Both the EFSA novel food application and the Chinese food catalog exclude these groups. The applicant's own reason was "insufficient data," and nothing has changed since.

• Children. No pediatric data exist at any dose.

• Anyone with a history of intracranial hemorrhage. In the pooled safety analysis of 1,076 vascular dementia patients on 1,800 mg per day, one cerebral hemorrhage was classified by the study investigators as possibly related to treatment and graded severe.

• Anyone with a seizure disorder. In the same pooled analysis, one case of epilepsy was graded severe and classified as probably related.

Use Caution

• Active loose stool, chronic diarrhea or inflammatory bowel disease. Constipation appeared three times in the related adverse event list, but the laxative direction is the dominant effect and the mannitol fraction is dose dependent.

• Low blood pressure or aggressive antihypertensive therapy. In the 48 week open-label Alzheimer's study of 18 patients on 1,800 mg per day, diastolic blood pressure was 6 to 7 mmHg lower than baseline at weeks 24, 36 and 48.

• Existing liver enzyme elevation. Three of 18 patients in that same study showed raised alanine aminotransferase by week 48, and one participant in the 39 person cervical myelopathy trial had a transient enzyme elevation attributed to the extract that normalized within three months. In rats, C. tubulosa ethanol extract at 0.8 g/kg produced mild hepatic edema.

• Hormone sensitive conditions. Rodent work shows induction of the steroidogenic enzymes CYP11A1 and CYP17A1. No human endocrine safety data exist in men with prostate disease or in women.

Critical Safety Point

On 24 November 2020 the EFSA Panel on Nutrition, Novel Foods and Food Allergens adopted its opinion on Cistanche tubulosa stem water extract, published as EFSA Journal 2021;19(1):6346, and concluded in one sentence that "the safety of the NF has not been established." The panel's reasons were the 12 adverse events across 1,076 treated patients that investigators judged definitely, probably or possibly related, two of them severe, plus genotoxicity testing that did not follow EFSA's tiered strategy and repeated dose studies that were not GLP compliant. That is a regulator with a full dossier in hand declining to clear a purified extract for adult food supplement use at 2 g per day. No US supplement label mentions it.

Recommended Dosages

Sliced raw herb, decoction: 6 to 10 g per day, the Chinese Pharmacopoeia range, which is also the daily consumption figure carried in Qinghai provincial food standards.

30 percent ethanolic extract, locomotive function: 1,800 mg per day, the dose in the 12 week Toyama trial, supplying roughly 315 mg echinacoside and 97 mg acteoside at that material's stated 17.48 and 5.41 percent.

30 percent ethanolic extract, postoperative sensory symptoms: 2,400 mg per day, the dose in the 24 week cervical myelopathy trial, supplying 667 mg echinacoside and 149 mg acteoside.

70 percent phenylethanoid glycoside water extract: 1,800 mg per day is the licensed Chinese prescription dose for vascular dementia and the dose in all four human studies submitted to EFSA. Taiwan's food authority caps food use of this extract at 450 mg per day, a fourfold lower ceiling, which is the most conservative number any regulator has put in writing.

60 percent ethanolic 2:1 extract, strength: 5 g per day in two 2.5 g sachets, the protocol in the 2025 trial. Marker content unstated, so this dose cannot be converted to any other form.

Retail capsules: most labels suggest 500 to 1,000 mg per day. At a genuine 50 percent echinacoside that would be 250 to 500 mg of echinacoside, inside the range used in trials. At an overstated assay it could be a fraction of that, and nothing on the label lets a buyer tell which.

Duration

Published trial durations are 8 weeks (strength), 12 weeks (walking), 24 weeks (sensory symptoms), 30 and 90 days (memory, combination product) and 48 weeks (open-label Alzheimer's). Nothing longer than 48 weeks has been published in humans, and the 5 year periodic safety update filed for the Chinese prescription product covered 844 patients with no reported adverse events, which is a passive pharmacovigilance number rather than a trial result.

Timing & Administration

Echinacoside absorption depends partly on the sodium dependent glucose transporter SGLT1. Blocking that transporter with phloridzin cut apparent permeability of both echinacoside and acteoside to 20 percent of baseline, while the calcium channel blocker verapamil and the GLUT inhibitor phloretin had no effect. Whether a meal helps or hinders has not been tested directly in humans, so there is no evidence based instruction to give.

The matrix matters more than the timing. Whole C. tubulosa extract raised the apparent permeability of both echinacoside and acteoside threefold compared with the isolated compounds, which is an argument against buying purified echinacoside and an argument for a characterized whole extract.

In the trials, dosing was split. The 2025 strength trial gave one sachet after breakfast and one after lunch. The Chinese prescription product is two 300 mg capsules three times daily. The laxative effect is the reason timing is discussed at all, and the one trial that specified timing dosed after meals rather than at bedtime.

Timeline of Effects

Days 1 to 7. Stool softening, if it happens, happens here. This is osmotic and needs no accumulation.

Weeks 4 to 8. The 2025 strength trial ran 8 weeks and reported its between group differences at that endpoint only, with no interim measurement, so nothing can be said about when strength changes appeared. The 30 day memory trial in 117 healthy adults aged 30 to 65 found improvements at 30 days, but with a combination product containing Ginkgo biloba.

Week 12. The first timepoint at which the locomotive syndrome trial measured anything. Step width and 5 m gait speed differed from placebo at p = 0.046 in the over 60 subgroup, which is a finding at the edge of significance in the 21 people over 60, 13 on extract and 8 on placebo.

Weeks 12 to 24. The cervical myelopathy trial found visual analog scale improvement at both 12 and 24 weeks, with the 24 week values not clearly larger than the 12 week values.

Weeks 24 to 48. Diastolic blood pressure reduction in the open-label Alzheimer's study was detectable at week 24 and persisted at weeks 36 and 48.

The published trials measured everything other than bowel effects at 12 weeks or later. At doses supplying at least 300 mg of echinacoside daily, no trial has shown a benefit appearing after 12 weeks that was absent at 12 weeks.

Benefits of Taking Cistanche

Strength and testosterone, one trial. Tao and colleagues, Nutrients 2025;17:2965, randomized 48 men, 24 resistance trained and 24 untrained, to 5 g/day of a C. deserticola 2:1 ethanolic extract or placebo for 8 weeks, 12 per arm. In untrained men, endpoint one repetition maximum bench press was 64.16 plus or minus 8.55 kg versus 56.04 plus or minus 8.82 kg, squat 78.12 plus or minus 10.77 kg versus 67.70 plus or minus 10.08 kg, and serum testosterone 773.77 plus or minus 130.01 ng/dL versus 636.49 plus or minus 98.85 ng/dL, p less than 0.01. In trained men, testosterone was 846.15 plus or minus 141.26 versus 732.16 plus or minus 117.27 ng/dL, and squat and maximal voluntary isometric contraction differed, but one repetition maximum bench press did not, at 143.12 plus or minus 3.55 kg versus 141.04 plus or minus 4.05 kg, and neither did repetitions to failure, at 19.16 plus or minus 1.33 versus 18.91 plus or minus 1.24. Cortisol, C reactive protein and creatine kinase were all lower in the cistanche arms.

Now the qualifications, because they are not optional. Every one of those testosterone values, placebo included, sits inside the ordinary adult male reference interval. The trial reports no baseline testosterone values in the extract I could obtain, so the between group endpoint difference cannot be separated from baseline imbalance in groups of 12. The extract's echinacoside and acteoside content is never stated. No botanical authentication is described. The material came from a single supplier and the study was sponsored by Hohhot Shancheng Biotechnology Co., Ltd., a cistanche company. This is a real randomized trial and it is also the only one of its kind, at n = 12 per cell, with an interested sponsor and no marker characterization. It is a reason to run a bigger trial, not a reason to believe the category claim.

Walking ability in people over 60. Inada, Tohda and Yang, Nutrients 2021;13:264. Thirty two randomized, 26 completed, 15 on 1,800 mg/day C. tubulosa 30 percent ethanolic extract and 11 on placebo, 12 weeks, ages 40 to 80. Step width on the two step test and gait speed on the 5 m walk improved in the over 60 subgroup, p = 0.046. Skeletal muscle mass of trunk and limbs did not change, so whatever improved was not muscle bulk.

Sensory symptoms after cervical decompression. Tohda, Inada, Yasuda, Seki and colleagues, Next Research 2026;3:101092. Thirty nine patients, 19 on 2,400 mg/day and 20 on placebo, 24 weeks, double blind. Visual analog scale for pain and numbness improved versus placebo at 12 and 24 weeks, and Japanese Orthopaedic Association sensory score for the trunk improved. Grip strength, motor function and total JOA score did not differ.

Cognition, in combination only. Chen and colleagues, Phytotherapy Research 2024, randomized 100 middle aged and elderly Chinese adults to a C. tubulosa plus Ginkgo biloba tablet delivering 72 mg echinacoside and 27 mg flavonol glycosides daily for 90 days. Mini Mental State Examination rose from 26.5 to 27.1 and Montreal Cognitive Assessment from 23.4 to 25.3, both p less than 0.001, alongside falls in total tau and phosphorylated tau at T181, S199, T231 and S396. Gao and colleagues, Frontiers in Pharmacology 2026, ran the same combination in 117 healthy adults for 30 days, registered as ChiCTR2400084102. Neither trial can attribute anything to cistanche, because there was no cistanche-only arm.

Potential Negatives & Side Effects

The best characterized human safety dataset is the one EFSA rejected the conclusion of. Across a phase II trial (115 versus 112 completers), a phase III trial (330 versus 111 completers) and a phase IV cohort of 625 completers, all at 1,800 mg per day for three months, 46 subjects experienced 49 adverse events. Those completer counts total 1,070 and 223, while the safety analysis was run on the larger treated populations of 1,076 on cistanche extract and 229 on comparator. The rate was 3.72 percent on cistanche extract and 3.93 percent on the comparator, ergoloid mesylates at 6 mg per day. Twelve events, about 1.1 percent, were judged definitely, probably or possibly related to cistanche.

Named and graded, those 12 were: moderate gastritis (definitely related); moderate constipation, severe epilepsy and mild sleepiness (probably related); and severe cerebral hemorrhage, moderate gastric hemorrhagic ulcer, moderate lethargy, constipation twice, dizziness, fatigue and sleepiness (possibly related).

From the other trials:

• Grade 2 eczema in one cervical myelopathy participant, resolving after discontinuation.

• Transient liver enzyme elevation in one cervical myelopathy participant, normalizing within three months.

• Alanine aminotransferase rises in one of 18 Alzheimer's patients at week 24 and three of 18 at week 48, magnitudes not reported.

• Nausea in one Alzheimer's patient at week 8, and a hallucination recurrence at week 12 in a patient who had hallucinated at baseline.

• Diastolic blood pressure 6 to 7 mmHg below baseline from week 24 onward.

• No adverse event results at all from the 48 man, 8 week strength trial, which describes its monitoring procedure and its tolerability scale in the methods and then reports no tolerability findings anywhere.

One gastrointestinal bleeding event and one cerebral hemorrhage across the vascular dementia dataset, in a population already on other medication, do not establish causation. They do argue against dismissing the bleeding question, which nobody has studied properly.

Deficiency Symptoms

Cistanche is a plant, not a nutrient. Human beings have no requirement for echinacoside, acteoside or any other cistanche constituent, no biochemical pathway that fails without them, no recommended intake and no deficiency state. Nobody has ever been diagnosed with a cistanche deficiency because there is no such thing.

What cistanche addresses

The conditions in the literature are age related declines and disease states, not deficiencies: slowed gait and reduced step width in adults over 60, chronic dry constipation in the elderly, residual sensory symptoms after cervical spine decompression, and vascular dementia in the Chinese drug indication. Each of those has causes that have nothing to do with cistanche intake.

Bottom line

Framing cistanche as something the body is short of is marketing. The correct question is whether adding a plant extract with 0.83 percent oral bioavailability of its main marker changes a measurable outcome, and the answer so far is "sometimes, modestly, in older adults, on a few specific endpoints."

Toxicity Symptoms

At high intake

Animal toxicology on cistanche is reassuring and largely irrelevant to the human question, because the doses are enormous. A 90 day feeding study in 80 Sprague Dawley rats fed diets of 2, 4 and 8 percent powdered C. deserticola found no treatment related changes and set the no observed adverse effect level at 7.8 g/kg body weight in males and 8.0 g/kg in females. For C. tubulosa extract, single oral doses of 2, 4 and 8 g/kg in mice caused no mortality over 14 days, the Ames test was negative in five Salmonella typhimurium strains, the mouse micronucleus assay was negative at 0.5 to 2 g/kg, and six month studies set a rat NOAEL of 1.65 g/kg, the highest dose tested, with 1.5 g/kg in dogs described as 100 times the clinical dose. One rat study did find mild hepatic edema at 0.8 g/kg of C. tubulosa ethanol extract given intragastrically for 20 days.

EFSA looked at the same class of studies and noted that none of the repeated dose work complied with GLP or the relevant OECD guidance, and that the genotoxicity package skipped the required in vitro micronucleus test. High NOAELs from non compliant studies are weak evidence, not strong evidence.

Signs to reduce or stop

Loose or urgent stools, new abdominal pain or dyspepsia, unusual lightheadedness on standing, any dark stool or unexplained bruising, new right upper quadrant discomfort, and any rash. Liver enzyme elevation is silent, and the trials that ran past 12 weeks at 1,800 mg per day and above tracked it with blood chemistry at every study visit.

General note

There is no established human toxic dose and no published report of an overdose case. The ceiling anyone has set is regulatory rather than toxicological: 450 mg per day of the 70 percent phenylethanoid glycoside extract for food use in Taiwan, and no authorized food use at all in the European Union.

How Cistanche Works

The pharmacology has a bioavailability problem that most write ups skip. Oral bioavailability of echinacoside in rats is 0.83 percent. Whatever cistanche does systemically, it is not doing it through intact echinacoside in plasma at meaningful concentrations.

What reaches circulation is largely metabolites. Gut bacteria hydrolyze echinacoside to hydroxytyrosol and to 3-hydroxyphenylpropionic acid, the major microbial metabolite of caffeic acid. Hydroxytyrosol is the olive polyphenol with its own substantial antioxidant literature, which means part of cistanche's activity is plausibly a slow release hydroxytyrosol delivery system operating through the colon. This also explains why the polysaccharide fraction matters: rat work shows that both unprocessed and wine processed cistanche polysaccharides shift gut microbiota composition, raise short chain fatty acid output and increase acteoside absorption.

For the androgen claim, the proposed mechanism is direct testicular. In rats given 0.4 and 0.8 g/kg of C. tubulosa ethanol extract for 20 days, sperm count rose 2.3 and 2.7 fold, motility 1.3 and 1.4 fold, abnormal forms fell to 0.76 and 0.6 of control, and serum testosterone and progesterone rose, with increased expression of the cholesterol side chain cleavage enzyme CYP11A1 and of 17-alpha-hydroxylase/17,20-lyase CYP17A1. Those are the rate limiting steps of testosterone synthesis, so the mechanism is coherent. It is also a rodent mechanism at doses equivalent to many grams per day in a human, demonstrated by immunohistochemistry and western blot rather than by hormone kinetics.

A second mechanism has one suggestive paper behind it. Echinacoside above 1 micromolar stimulated growth hormone secretion from rat pituitary cells, and the effect was blocked by a ghrelin receptor inverse agonist, which points at the growth hormone secretagogue receptor. That is a cell culture finding with molecular docking support and no human confirmation.

Synergistic Supplements

• Ginkgo biloba. The only combination with human randomized data. Two trials, 100 participants over 90 days and 117 over 30 days, used tablets delivering 72 mg echinacoside plus 27 mg flavonol glycosides. The trials establish that the combination works better than placebo and say nothing about whether cistanche contributes.

• Cistanche polysaccharide fraction. Not a separate supplement so much as an argument for buying a whole extract rather than isolated echinacoside. Whole C. tubulosa extract tripled the apparent intestinal permeability of echinacoside and acteoside relative to the pure compounds, and polysaccharide pretreatment for 21 days increased acteoside absorption in rats.

• Cuscuta seed (tu si zi) and Eucommia ulmoides. Traditional pairings in Huan Shao Dan and in the 1644 formula Bushen Cishi Wan. Traditional co-use is a reason to expect tolerability, not a reason to expect additive effect.

Combining cistanche with other CYP3A4 inhibiting botanicals, or with additional osmotic laxatives, has no supporting data and two obvious ways to go wrong.

Interactions & What NOT to Take

Almost every interaction below is predicted from in vitro enzyme data or from a single observed physiological effect. None has been documented in any published human interaction study, and that absence is itself the finding.

Predicted from CYP inhibition. Echinacoside inhibits CYP1A2, CYP2E1, CYP2C19 and CYP3A4 in vitro. The drugs where that would matter most are the narrow therapeutic index CYP3A4 substrates tacrolimus, cyclosporine and everolimus; simvastatin and lovastatin, where inhibition raises myopathy risk; the direct oral anticoagulants apixaban and rivaroxaban; and midazolam. For CYP2C19, clopidogrel is the awkward one, because it is a prodrug and inhibition would reduce rather than increase its effect. For CYP1A2, theophylline, tizanidine and clozapine. For CYP2E1, chlorzoxazone.

A contradiction worth naming. The same rat study that reported raised testosterone also reported increased CYP3A4 expression, which is induction, the opposite direction to the in vitro inhibition. The net clinical effect of cistanche on CYP3A4 in humans is unknown, and "unknown in both directions" is a worse position than a clean inhibition signal.

Bleeding. Warfarin, apixaban, rivaroxaban, clopidogrel and aspirin. The signal is one severe cerebral hemorrhage and one gastric hemorrhagic ulcer among 1,076 patients, both judged possibly related. That is far short of a demonstrated interaction and far past the point where it should be ignored in anyone already anticoagulated.

Blood pressure. Amlodipine, lisinopril, losartan and any other agent being titrated to target. A 6 to 7 mmHg diastolic reduction sustained over 24 weeks is small in a healthy person and not small in someone already near their floor.

Thyroid medication. In the 90 day combination trial, thyroid stimulating hormone fell while triiodothyronine and free triiodothyronine rose. That shift was measured in a combination product over 90 days, and it has not been examined in anyone taking levothyroxine or methimazole.

Bowel medication. Cistanche opposes loperamide and adds to bisacodyl, senna, lactulose and polyethylene glycol. The mannitol content alone makes this predictable.

Seizure medication. One severe epilepsy event was judged probably related. That single case is the entire human record bearing on levetiracetam, lamotrigine, valproate or any other anticonvulsant, and nothing further has been published.

Quality, Testing & Adulteration

This is the section that should decide whether you buy.

Documented adulteration. Wang and colleagues, Frontiers in Plant Science 2018, sequenced the ITS2 region of 251 cistanche and adulterant samples, developed four nucleotide signatures of 30 to 37 base pairs and six species specific primers, then tested 66 commercial products including extracts and Chinese patent medicines. The adulteration rate was 36.4 percent: 19.7 percent adulterated with Cynomorium songaricum or Cistanche sinensis, and 16.7 percent substituted outright with Cynomorium songaricum, Cistanche sinensis or Boschniakia rossica. Cynomorium songaricum, sold in its own right as suo yang, was the most common adulterant. Separately, Cistanche salsa is routinely retailed by wholesalers as Cistanches Herba.

Why DNA testing is hard here. Cistanche is sold heavily processed, often wine steamed and dried, which degrades DNA and defeats standard barcoding. A 2025 platform combining multiplex PCR with MALDI-TOF mass spectrometry across four diagnostic SNPs in nuclear ITS and chloroplast rpl16 detects 0.07 percent genomic DNA, equivalent to 6.8 pg per microliter, and 1 percent C. deserticola powder in a dried mixture, distinguishing it from C. tubulosa and six further adulterant species, C. phelypaea, C. sinensis, C. salsa, C. ambigua, C. ridgewayana and C. rosea. Validated on 27 dried specimens from Egypt, Ethiopia, Kazakhstan, the United Arab Emirates and Uzbekistan, it found 11 to be C. tubulosa and 8 to give ambiguous signals.

Marker content is not a constant. Across 47 wild samples of four species, C. sinensis had no detectable echinacoside at all, while C. tubulosa measured 28.38 mg/g echinacoside with only 8.12 mg/g acteoside, a lopsided ratio that still clears the 15 mg/g combined minimum on echinacoside alone. Content varies by origin, with Inner Mongolia samples significantly higher than those from Minqin City in Gansu; by harvest season, with spring harvested C. deserticola reported at over 90 percent higher acteoside than autumn; and by position on the stem, with echinacoside declining from bottom to top.

What a certificate of analysis should show. The binomial with authority, either Cistanche deserticola Y.C. Ma or Cistanche tubulosa (Schenk) Wight. A DNA identity result, not a TLC fingerprint, since TLC will not separate the species reliably in processed material. Separate HPLC quantities for echinacoside and acteoside with the assay method named, not a single "total phenylethanoid glycosides" number. Heavy metals against real limits: the EFSA assessed specification used arsenic at or below 0.5 mg/kg, lead 0.3, cadmium 0.2 and mercury 0.1. Residual solvents, because ethanol precipitation is part of the standard process and the assessed specification also set a benzene limit of 2 mg/kg, a parameter EFSA specifically flagged as genotoxic and carcinogenic. Sulfur dioxide, because sulfur fumigation remains common in the Chinese crude herb trade and degrades glycosides. Harvest region and season.

What third party testing does not cover. There is no United States Pharmacopeia monograph for cistanche, so no US compendial identity test, assay or reference standard exists and no product can be USP verified for it. The operative standard is the Chinese Pharmacopoeia HPLC assay, written for the crude herb rather than for a 10:1 extract capsule. A buyer's practical options are a supplier who publishes a lot specific HPLC chromatogram with the species named, or nothing.

Special Considerations

The CITES question is real for one species and absent for the other. Annotation #4 exempts seeds, in vitro cultures, cut flowers of artificially propagated plants, Vanilla and Cactaceae fruits, Opuntia and Selenicereus stems and flowers, and finished products of Euphorbia antisyphilitica. Extracts and powders of C. deserticola are not on that list. Whether a given importer holds the permits is not something a consumer can check from a bottle, and the commercial incentive to relabel C. deserticola material as C. tubulosa, or to say nothing at all, is obvious.

The supply chain has shifted to cultivation. China reports over 200,000 hectares under cistanche cultivation and roughly 8,500 tons of annual production, against the roughly 400 to 500 tons of annual purchases recorded at the beginning of the 1980s and the 120 tons of world trade described in the 2000 CITES proposal, which put production at the time of writing at about 70 tons in Inner Mongolia and 50 tons in Xinjiang. Cultivation requires establishing the host first, Haloxylon ammodendron for C. deserticola or Tamarix chinensis for C. tubulosa, then inoculating and waiting three to five years. Atriplex canescens is under investigation as an alternative host. The 2000 proposal identified large scale destruction of the host plant alongside over exploitation of cistanche itself, without ranking one above the other.

"Desert ginseng" is a marketing coinage. Cistanche has no botanical relationship to Panax ginseng, which is in the Araliaceae, contains ginsenosides rather than phenylethanoid glycosides, and has a separate and much larger clinical literature. Nothing shown for ginseng transfers.

Regulatory status differs by continent, sharply. Approved Chinese prescription drug for vascular dementia since 2005. Food catalog substance in China since 9 November 2023. Food ingredient in Taiwan capped at 450 mg per day since 2017. Listed among drug and health products in Canada. Novel food application refused as unestablished for safety in the European Union in 2021. Sold freely as a dietary supplement in the United States with no premarket review of any kind.

Research Status & Evidence Quality

Tier 1, randomized placebo controlled human trials of cistanche alone. Three. Locomotive syndrome, n = 26 completers, 12 weeks, one endpoint significant at p = 0.046 in a subgroup. Degenerative cervical myelopathy, n = 39, 24 weeks, visual analog scale positive and motor endpoints null. Muscle strength and testosterone, n = 48, 8 weeks, positive on every endpoint in untrained men but null on bench press and repetitions to failure in trained men, manufacturer sponsored, no marker characterization. Total participants across all three: 113.

Tier 2, randomized trials of cistanche in combination. Two, both C. tubulosa with Ginkgo biloba, 100 and 117 participants, 90 and 30 days. Positive for cognition and memory. Uninterpretable for cistanche specifically.

Tier 3, controlled trials in a disease population with an active comparator. The Chinese vascular dementia program, phase II with 115 versus 112 and phase III with 330 versus 111, both against ergoloid mesylates rather than placebo. EFSA noted that using an approved drug as comparator limits what can be concluded, and that assessment concerned safety rather than efficacy.

Tier 4, uncontrolled human data. An 18 patient open-label Alzheimer's study over 48 weeks, and a 625 patient phase IV cohort.

Tier 5, animal and in vitro. Large and growing. The androgen mechanism, the ghrelin receptor and growth hormone finding, the constipation mechanisms and essentially all of the neuroprotection literature sit here.

Research Limitations

• No published trial of cistanche alone reports erectile function, libido or sexual satisfaction as an endpoint, primary or secondary. The category's dominant marketing claim has zero direct human evidence.

• The only human testosterone data is one 8 week trial with 12 participants per cell, sponsored by a cistanche manufacturer, reporting endpoint values that sit inside the normal reference range in both arms, with no botanical authentication and no marker assay of the test material.

• Extracts differ so much between trials that doses do not translate. A 2:1 ethanolic extract at 5 g, a 30 percent ethanolic extract at 1,800 or 2,400 mg, and a 70 percent phenylethanoid glycoside water extract at 1,800 mg are four different interventions.

• Species is often unstated. Several trials specify C. tubulosa; the strength trial specifies C. deserticola; most reviews pool them.

• Sample sizes are small. The largest placebo controlled trial of cistanche alone randomized 48 people.

• Publication and language bias are likely. Most of the pharmacology and much of the clinical work is Chinese and Japanese, and negative results in supplement research are underpublished everywhere.

• The single largest human safety dataset, 1,076 patients, was assembled as a drug dossier in a diseased population against an active comparator, which is why EFSA judged it unusable for clearing a food supplement.

Summary & Key Takeaways

Bottom Line

Cistanche is two different plants sold under one name, and the difference is legally and chemically material. Cistanche deserticola has been CITES Appendix II listed since 2000 under an annotation that covers extracts, and must meet a Chinese Pharmacopoeia minimum of 0.30 percent combined echinacoside and acteoside. Cistanche tubulosa has no CITES listing and must meet 1.5 percent. Western labels routinely name neither the species nor a marker assay, and 36.4 percent of 66 commercial products tested by DNA in 2018 were adulterated or substituted, usually with Cynomorium songaricum. On efficacy, three placebo controlled trials of cistanche alone, totaling 113 participants, support modest effects on gait speed in adults over 60, on sensory symptoms after cervical decompression, and on strength and serum testosterone in one manufacturer sponsored 8 week study. Nothing in the human record supports the sexual function claims the category is sold on, because no trial of cistanche alone has measured them.

Key Safety Points

• EFSA concluded on 24 November 2020 that the safety of C. tubulosa stem water extract at 2 g per day had not been established, citing 12 related adverse events among 1,076 patients including one severe cerebral hemorrhage and one severe epilepsy case.

• Avoid in pregnancy, breastfeeding and childhood, with a history of intracranial bleeding, and with a seizure disorder.

• Liver enzyme elevations appeared past 12 weeks at concentrated extract doses, in one of 18 patients at 24 weeks, three of 18 at 48 weeks, and one of 19 in a 24 week trial.

• Diastolic blood pressure fell by roughly 6 to 7 mmHg on long term high dose use, in the one study that measured it.

• Echinacoside inhibits CYP1A2, CYP2E1, CYP2C19 and CYP3A4 in vitro while a rat study shows CYP3A4 induction, so the direction of any human interaction with tacrolimus, simvastatin, apixaban, clopidogrel or theophylline is genuinely unknown.

Special Note

Buy on the certificate of analysis or do not buy. The document should name the species with its authority, report echinacoside and acteoside separately by HPLC rather than as one lumped phenylethanoid number, give a DNA identity result rather than a thin layer chromatography fingerprint, and state heavy metals, residual solvents and sulfur dioxide against numeric limits. A retail claim of 50 percent echinacoside sits above the 25 to 45 percent range of the pharmaceutical grade specification EFSA reviewed, and above all seven batches in that dossier, which means it is a number that should be proven rather than asserted. Where the label says only "Cistanche extract 10:1," the honest reading is that the seller either does not know which plant is in the capsule or has decided not to say.