evipedia.ai
← Back

Bulbine natalensis to Improve Testosterone

Evidence Review created on 09/14/2026 using AI4L / Opus 5

Also known as: Bulbine latifolia, Rooiwortel, Ibhucu, Ingcelwane, Broad-Leaved Bulbine, ProLensis

Motivation

Bulbine natalensis is a succulent herb from southern Africa whose stem and root have a long place in Zulu and Xhosa healing traditions. Over roughly the past fifteen years it has been sold internationally as an oral supplement aimed at men who want to raise their own testosterone rather than take a prescribed hormone.

Interest outside Africa grew out of a short series of rodent experiments run by one South African university group. Those experiments reported that low doses raised male hormone levels and sexual activity, that a higher dose did the opposite, and that liver and kidney measurements shifted at the very same doses. Only one study has ever been run in people, and it was built to check safety markers rather than to see whether the hormone moved at all.

This review examines what is known about Bulbine natalensis and male hormone levels: what the animal work shows and where it stops, what the single human study did and did not measure, what the plant and the products made from it actually contain, which harms have been observed or are suspected, and how the herb behaves alongside medicines and other supplements.

Benefits - Risks - Protocol - Conclusion

High-level sources that give an overview of Bulbine natalensis, of the plant’s chemistry, and of the supplement category it is sold in.

Of the six priority expert platforms, only hubermanlab.com has published anything naming this plant, and that item is listed above; both a web search and an on-site search were run for each of the other five, as recorded above.

Grokipedia

No Grokipedia article exists for Bulbine natalensis, nor for its accepted synonym Bulbine latifolia.

Examine

Bulbine natalensis

The most-read independent summary of the herb, and the source of the body-surface-area dose estimate sellers quote; it notes the organ-damage signal appears at the hormone-raising doses.

ConsumerLab

No ConsumerLab article, product review or quality test exists for Bulbine natalensis.

Systematic Reviews

Systematic reviews and meta-analyses bearing on the effect Bulbine natalensis is sold for and on its principal risk; none covers the plant itself.

Mechanism of Action

Bulbine natalensis stem and root carry anthraquinones (three-ringed plant pigments) — chrysophanol, aloe-emodin, knipholone and the bulbine-knipholones — alongside saponins (soap-like plant compounds), tannins, cardiac glycosides (heart-acting plant steroids) and phytosterols (plant versions of cholesterol).

The proposed route to testosterone is upstream, through the hypothalamic-pituitary-gonadal axis (the brain-to-testis hormone loop). In male rats, low doses raised luteinizing hormone (LH, the pituitary signal telling the testis to make testosterone) and follicle-stimulating hormone (FSH, the pituitary signal for sperm production), then testicular and blood testosterone, testicular cholesterol — the raw material the hormone is built from — and testicular alkaline phosphatase (ALP, an enzyme marking metabolically active tissue). Blood estradiol fell at every dose, which some read as blocking aromatase (the enzyme converting testosterone into oestrogen). The dose-response is biphasic: the highest dose tested lowered testosterone instead of raising it.

A competing reading holds that the hormonal shift is part of a general stress response rather than a specific androgenic action, because the same doses disturbed liver and kidney chemistry and tissue architecture, and because part of the observed erectile effect looks vascular and independent of any hormone change.

No human pharmacokinetic data exist: half-life, tissue distribution and receptor selectivity are uncharacterised. Cell work shows the extract activates two chemical-sensing switches, the pregnane X receptor and the aryl hydrocarbon receptor, inducing CYP3A4, CYP2C9, CYP1A2 and CYP2B6 (members of the cytochrome P450 enzyme family, which break down most medicines) along with P-glycoprotein (a pump that expels drugs from cells).

Historical Context & Evolution

The plant’s original use had nothing to do with hormones. In Zulu and Xhosa practice it is ibhucu, ingcelwane or rooiwortel (“red root”, for the orange sap the cut stem releases). Leaf gel was applied to wounds, burns, cracked lips, ringworm and insect bites; root infusions were drunk for diarrhoea, vomiting, convulsions, rheumatism, urinary complaints, diabetes and venereal disease. Alongside this, ethnobotanical surveys recorded it as an aphrodisiac used for male impotency — the thread that later carried it abroad.

Two lines of work moved it into the health-optimisation world. The first was chemical: analyses from the mid-1990s onward identified chrysophanol, knipholone and related anthraquinones, and a leaf-gel wound dressing reached the market in South Africa. The second, decisive, line was a series of rat experiments published between 2008 and 2010 by a University of Fort Hare group, reporting raised testosterone, raised pituitary hormones and improved copulatory performance at 25 and 50 mg/kg body weight.

Supplement manufacturers took up those findings almost immediately, and a standardised extract was trademarked and taken into a short human safety trial in 2012. Since 2019 a United States natural-products laboratory has been publishing the plant’s chemistry and its enzyme-interaction profile, work that has tightened what is known about composition without yet testing the hormonal claim in people. The species was also reclassified as Bulbine latifolia, which now splits the literature across two names.

Expected Benefits

High 🟩 🟩 🟩

No benefit reaches High: the qualifying class of evidence — a human clinical endpoint or a validated clinical surrogate shown in more than one trial — does not exist here, because no trial has measured testosterone, sexual function or body composition in people at all.

Medium 🟩 🟩

No benefit reaches Medium either: that would require the same class of human outcome in a single trial or in consistent observational data, and the one human study run on this herb measured safety chemistry only.

Low 🟩

Speculative 🟨

Increased Blood and Testicular Testosterone

Aqueous stem extract raised blood and testicular testosterone plus both pituitary hormones in male rats at 25–50 mg/kg (Yakubu & Afolayan, 2010). Basis is animal work only. No human trial has measured testosterone.

Enhanced Sexual Behaviour and Penile Reflexes

At 25–50 mg/kg, mount, intromission and ejaculation frequencies and penile reflexes rose in male rats while mount and intromission latencies fell (Yakubu & Afolayan, 2009). Rodent behaviour only; 100 mg/kg reversed every measure.

Accelerated Wound Healing ⭕️ Not Central to Improve Testosterone

Leaf gel sped wound contraction, collagen deposition and collagen maturation in pigs (Pather & Kramer, 2012). Animal work only. Bears on topical skin repair, not on androgen status.

Antimicrobial Activity ⭕️ Not Central to Improve Testosterone

Solvent fractions of the tuber inhibited several disease-causing bacteria and moulds in culture (Yakubu et al., 2012). Test-tube assays only. Bears on infection control, not on testosterone.

Glucose-Handling Activity ⭕️ Not Central to Improve Testosterone

Phenolic-rich extracts inhibited carbohydrate-digesting enzymes and altered insulin release in cultured cells (Oyedemi et al., 2023). Cell work only, never tested in people. Bears on blood-sugar control, not androgen levels.

Benefit-Modifying Factors

  • Drug-metabolising gene variants: Carriers of slow CYP2C9 or CYP3A4 variants (genes coding liver enzymes that break down both drugs and plant compounds) would clear the plant’s anthraquinones differently, plausibly changing exposure — though no pharmacogenetic study of this herb exists.

  • Baseline testosterone and SHBG: Any androgenic signal would be easiest to detect in men starting below roughly 350 ng/dL total testosterone. High sex hormone-binding globulin (SHBG, the protein that binds testosterone in blood) blunts a rise in the free, usable fraction.

  • Sex: Every efficacy finding is in male rats. In female rats the extract altered no reproductive parameter at 25–50 mg/kg. Nothing supports a benefit in women, and the hormonal direction would be unwanted for most.

  • Pre-existing conditions: A pituitary-driven route implies nothing to gain where the testis itself has failed (primary hypogonadism). Obesity, poor sleep and untreated sleep apnoea suppress testosterone through channels this herb does not address.

  • Age: Older men in the target range carry more illness, more medicines and slower liver clearance, so exposure runs higher for the same dose. Age-related testosterone decline has not been studied with this plant in any species.

Potential Risks & Side Effects

High 🟥 🟥 🟥

No risk reaches High: the qualifying class of evidence — a documented adverse event or clinical-endpoint change replicated across more than one human trial — does not exist, because only one 28-day human safety trial has been run and no published case report names this herb.

Medium 🟥 🟥

Liver and Kidney Marker Disturbance

Liver and kidney chemistry moves on this herb. In the only human trial — 28 days, 36 healthy men, part-funded by a supplement-industry sponsor — alkaline phosphatase rose and creatinine fell against placebo, both statistically significant but inside reference limits (Hofheins et al., 2012). In rats the same extract raised alanine and aspartate aminotransferase (ALT and AST, enzymes released when liver cells are stressed), gamma-glutamyl transferase (an enzyme flagging bile-duct stress) and bilirubin, and distorted liver and kidney tissue (Afolayan & Yakubu, 2009). Reversibility in people is unstudied.

Magnitude: Alkaline phosphatase rose 2.0 IU/L (+3%) against a 2.4 IU/L (−3.8%) fall on placebo (P < 0.04; P is the probability a difference this large arose by chance), and creatinine fell 0.08 mg/dL (−7.4%, P < 0.003), every value staying inside clinical reference limits; in rats at 25–100 mg/kg, 39% of liver indices and 21% of kidney indices were altered.

Low 🟥

Unreliable Dose Delivered by Commercial Products

What a labelled product contains varies enormously, so real intake is unknown. An analytical survey found the plant’s marker anthraquinones in only some labelled products (Bae et al., 2020). Testosterone-booster products also routinely carry nutrients above tolerable limits (Clemesha et al., 2020).

Magnitude: Anthraquinones were detectable in only 5 of 9 supplements labelled as containing the plant, and across those five the total ranged from 11.3 to 90.4 mg per labelled daily dose, an eight-fold spread; separately, 13 of 50 testosterone-booster products exceeded a regulatory upper intake limit for at least one nutrient.

Speculative 🟨

Herb-Drug Interaction via Drug-Metabolising Enzymes

Extracts and the plant’s knipholones activate two chemical-sensing receptors, inducing CYP3A4 and CYP2C9 (Husain et al., 2021) plus CYP1A2, CYP2B6 and P-glycoprotein (Husain et al., 2022). Cell culture only; no human study exists.

Unfavourable Blood Lipid Shift

In rats, 14 days of stem extract raised cholesterol and triglycerides and lowered HDL (high-density lipoprotein, the protective cholesterol carrier) (Yakubu & Afolayan, 2009). Rodent only; no human lipid data exist.

Loss of Effect and Reproductive Harm at Higher Doses

At 100 mg/kg the extract lowered rather than raised rat testosterone and progesterone, and raised embryo resorption in females (Yakubu & Afolayan, 2009). Rodent only; the human-equivalent threshold is unknown.

White Blood Cell and Platelet Count Shifts

Rat white cell counts rose over 14 days, with movement in neutrophils, lymphocytes, monocytes and platelets (Yakubu & Afolayan, 2009). Rodent only; the human trial found no blood-count change.

Genotoxicity Concern from Anthraquinone Content

The plant carries aloe-emodin and chrysophanol (Bae et al., 2019); European regulators judged this compound class genotoxic (DNA-damaging) and set no safe intake (European Food Safety Authority, 2018). Laboratory and rodent basis only.

Additive Cardiac Effect from Cardiac Glycoside Content

Screening of the stem detected cardiac glycosides alongside the anthraquinones (Yakubu & Afolayan, 2009); a drug reference warns these may add to digoxin’s effect. Qualitative screening only; no cardiac event has been reported.

Reduced Platelet Aggregation

Chloroform extract inhibited platelet clumping in the test tube, most strongly against adrenaline-triggered aggregation (reviewed by Musara & Aladejana, 2020). Isolated in-vitro reports only; no bleeding event has been described in people.

Rebound Testosterone Drop and Androgenic Side Effects

A first-hand account describes a sharp testosterone rise with back acne and aggression, then a crash below baseline after 7–10 days plus testicular pain (Huberman, 2023). Isolated report only; no controlled data exist.

Risk-Modifying Factors

  • Drug-metabolising gene variants: CYP2C9 poor metabolisers already clear warfarin and phenytoin slowly; layering an inhibitor of the same enzyme on top compounds that. UGT and SULT variants (enzymes tagging compounds for excretion) govern anthraquinone detoxification.

  • Baseline liver and kidney markers: Anyone starting with alanine aminotransferase above the upper reference limit, raised bilirubin, or an estimated filtration rate below 60 has less reserve against the liver-and-kidney shift seen in rodents.

  • Sex: All toxicology is in male rats except the reproductive study, where 100 mg/kg raised embryo loss in mated females. Pregnancy and breastfeeding are unstudied and the anthraquinone content argues against exposure.

  • Pre-existing conditions: Chronic hepatitis, fatty liver disease, chronic kidney disease, bleeding disorders and hormone-sensitive prostate disease each sit directly in the path of one of the observed or suspected signals.

  • Age: Older users carry more prescriptions, so the enzyme-induction signal matters more; hepatic blood flow and renal clearance also fall with age, raising exposure at any fixed dose.

Key Interactions & Contraindications

  • CYP3A4 substrates (tacrolimus, ciclosporin, simvastatin, apixaban, rivaroxaban): Caution to avoid. Enzyme induction can lower drug levels toward therapeutic failure — graft rejection or breakthrough clotting. Mitigation: avoidance of narrow-therapeutic-index agents, or drug-level monitoring where the pairing is unavoidable.

  • CYP2C9 substrates (warfarin, phenytoin, glimepiride, celecoxib): Caution. The plant both induces and inhibits this enzyme in cell work, so the direction is unpredictable — bleeding or loss of anticoagulation. Mitigation: weekly international normalised ratio checks for a month where the two are combined.

  • CYP1A2 and CYP2B6 substrates (theophylline, clozapine, olanzapine, bupropion, efavirenz): Caution. Induction can drop plasma levels and precipitate relapse or seizure-threshold change. Mitigation: separation of initiation from dose titration, with symptom and drug-level monitoring where assays exist.

  • P-glycoprotein substrates (digoxin, dabigatran): Caution. Induction of the efflux pump reduces absorption; with digoxin the plant’s own cardiac glycosides may separately add to its effect. Mitigation: a digoxin level at two weeks, or an anticoagulant not handled by this pump.

  • Over-the-counter paracetamol, ibuprofen and naproxen: Caution. Paracetamol adds hepatic load to an already uncertain liver signal; the anti-inflammatories add kidney load and, with the plant’s antiplatelet activity, bleeding risk. Mitigation: paracetamol held under 2 g daily.

  • Anthraquinone-bearing supplements (senna, cascara, aloe latex, rhubarb root): Caution. These stack the same hydroxyanthracene load that European regulators flagged, plus cumulative laxative effect. Mitigation: no concurrent use, with courses separated by at least four weeks.

  • Hepatically stressful supplements (green tea extract, kava, high-dose niacin, ashwagandha): Caution. Additive liver injury risk, the single most reported harm across the herbal supplement literature. Mitigation: one novel botanical at a time, with liver enzymes rechecked at four weeks.

  • Antiplatelet and anticoagulant supplements (fish oil, ginkgo, garlic extract, high-dose vitamin E): Caution. Additive inhibition of platelet aggregation raises bruising and surgical bleeding risk. Mitigation: cessation of all of them at least 14 days before any planned procedure.

  • Other androgenic or aromatase-directed supplements (Tribulus terrestris, Eurycoma longifolia, DHEA (dehydroepiandrosterone, a hormone the body converts into testosterone), boron): Caution. Additive intended effect, but also additive uncertainty and stacked liver load. Mitigation: one agent at a time, so any laboratory change is attributable.

  • Testosterone replacement therapy and hCG (human chorionic gonadotropin, a hormone drug that drives the body’s own testosterone production): Absolute contraindication in practice. A pituitary-stimulating agent is redundant against prescribed testosterone, which suppresses the same axis. Mitigation: none — the combination has no coherent rationale.

Populations who should avoid Bulbine natalensis:

  • Men with alanine or aspartate aminotransferase above 3× the upper limit of normal, or any active hepatitis
  • Cirrhosis of Child-Pugh Class B or C (moderate to severe liver scarring)
  • Estimated glomerular filtration rate below 60 mL/min/1.73 m²
  • Known or suspected hormone-sensitive prostate cancer, or prostate-specific antigen above 4 ng/mL
  • Men actively trying to conceive, given the reproductive findings at higher rodent doses
  • Anyone on warfarin, or with a diagnosed bleeding disorder, or within 14 days of scheduled surgery
  • Pregnancy and breastfeeding
  • Anyone under 18, and women, in whom no benefit has been proposed

Risk Mitigation Strategies

  • Baseline liver and kidney panel before first dose: A comprehensive metabolic panel plus gamma-glutamyl transferase establishes the reference against which the hepatorenal signal seen in rodents can be detected. Without it, any later abnormality is uninterpretable.

  • Four-week recheck of liver enzymes: Repeating alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase and bilirubin at 28 days matches the only human exposure window ever studied and catches drug-induced liver injury while it is still reversible.

  • Stop rule at 3× the upper reference limit: Discontinuing immediately if either liver enzyme exceeds three times the upper limit, or bilirubin rises, is the standard threshold for suspected herb-induced liver injury and prevents progression.

  • Single-agent introduction: Adding no other new supplement or medication for the first eight weeks keeps any lipid, enzyme or blood-count change attributable, mitigating the misattribution that makes herbal liver injury hard to catch.

  • Hard dose ceiling: Holding total intake at or below roughly 650 mg daily of a 10:1 extract stays inside the only dose ever given to humans and away from the rodent range where testosterone fell and embryo loss rose.

  • Medication reconciliation before starting: Reviewing every prescription against CYP3A4, CYP2C9, CYP1A2, CYP2B6 and P-glycoprotein handling mitigates the enzyme-induction risk, which is the harm most likely to matter clinically.

  • 14-day washout before surgery or dental extraction: Stopping two weeks ahead mitigates the antiplatelet signal and any residual enzyme induction affecting anaesthetic metabolism.

Therapeutic Protocol

  • The trial-anchored regimen: 325 mg of a standardised extract in the morning and 325 mg six hours later, 650 mg daily total, for 28 days. Popularised by The Center for Applied Health Sciences, which ran the only human exposure.

  • The allometric-scaling regimen: The alternative approach scales the optimal rat dose of 50 mg/kg by body-surface-area conversion to roughly 8 mg/kg — about 550 mg for a 150 lb adult. Popularised by Examine’s Kamal Patel.

  • The conventional alternative: Physician-supervised testosterone replacement, or correcting sleep debt, excess body fat and training load first. These have measured human endpoints that the botanical approach does not, and are presented here as equally standing options.

  • Best time of day: Split morning and early-afternoon dosing follows the human trial. Morning dosing also aligns with the natural testosterone peak, though no study of dose timing exists for this plant in any species.

  • Half-life: Uncharacterised. No human or animal pharmacokinetic study of the extract or its knipholones has been published, so no dosing interval can be derived from exposure data rather than convention.

  • Single versus split dosing: Split. The only human data used two doses six hours apart; the rodent work used once-daily oral dosing. Neither design compared schedules, so the split convention rests on precedent, not evidence.

  • Genetic polymorphisms: No pharmacogenetic data exist. CYP2C9 and CYP3A4 variants, and UGT or SULT conjugation variants, are the plausible candidates for altering exposure, since those pathways handle the plant’s anthraquinones.

  • Sex differences: All dosing precedent is male. Female rats showed no reproductive change at 25–50 mg/kg but embryo loss at 100 mg/kg, and no dose has been proposed for women in any source.

  • Age considerations: No age-stratified dosing exists. The human trial averaged 29 years, so the regimen carries no evidence at the older end of the target range, where hepatic and renal clearance are lower.

  • Baseline biomarkers: Total and free testosterone and sex hormone-binding globulin define whether any androgenic response is even detectable; liver enzymes define whether the protocol should begin at all.

  • Pre-existing conditions: Liver disease, chronic kidney disease and hormone-sensitive prostate disease each remove the dose range from consideration entirely rather than modifying it downward.

Discontinuation & Cycling

  • Intended duration: Short-term. The single human exposure lasted 28 days and the rodent studies 1–14 days, so nothing supports continuous lifelong use and no chronic-toxicity study exists in any species.

  • Withdrawal effects: None documented. No study has measured hormones, mood or sexual function after stopping, and no case report describes a withdrawal syndrome attributable to this plant.

  • Tapering: Not applicable. With no withdrawal signal and no receptor-downregulation data, abrupt cessation is the only protocol with any precedent; taper schedules circulating in user communities are invented rather than derived.

  • Cycling practice: Community protocols run roughly eight weeks on and four weeks off, reasoning from tolerance seen with other androgenic botanicals. No study has compared continuous with cycled administration in any species.

  • Rationale for a break: Scheduled interruption gives liver enzymes a window to normalise and makes a four-week recheck interpretable, which matters more here than any unproven tolerance argument.

Sourcing and Quality

  • Species verification: Labels may read Bulbine natalensis or Bulbine latifolia for the same plant, and Bulbine frutescens is a distinct species sold alongside it. A certificate of analysis naming the botanical species and plant part is the minimum.

  • Marker-compound standardisation: Products quantified for knipholone and total anthraquinones are the only ones whose dose means anything, since analytical survey work found several labelled products with none detectable.

  • Named material with human exposure: ProLensis, the trademarked standardised extract supplied for the 2012 safety trial, is the only Bulbine natalensis material ever given to people under controlled conditions; products naming it carry a dose precedent unnamed extracts do not.

  • Third-party testing: No Bulbine natalensis product currently carries NSF Certified for Sport, Informed Sport or USP Verified marks. In their absence, an independent laboratory certificate of analysis for identity, heavy metals and microbial load is the fallback.

  • Heavy-metal screening: Root material bought from open-air markets in South Africa has been reported to carry high aluminium and iron. Bulk powder of unverified origin therefore warrants an elemental-impurity certificate before use.

  • Extract ratio transparency: Dose precedent assumes a 10:1 extract. Raw, unextracted powder needs roughly ten times the mass for equivalence, so a label omitting the ratio makes the quoted dose uninterpretable.

  • Avoiding proprietary blends: The herb is frequently buried inside multi-ingredient testosterone formulas where its amount is undisclosed, which makes both dose control and attribution of any adverse change impossible.

Practical Considerations

  • Time to effect: Unknown in humans. Rodent hormone changes appeared within 1–7 days of dosing, and the only human trial ran 28 days without measuring any efficacy endpoint, so no human onset interval can be quoted.

  • Pitfall — assuming the rat dose transfers: The rodent optimum of 50 mg/kg is not a human dose. Body-surface-area conversion places it near 8 mg/kg, and taking the rat figure directly would land far into the range where the effect reversed.

  • Pitfall — stacking: The herb is usually bought inside a blend, so users often take it alongside several other liver-loading botanicals, making any enzyme rise unattributable and any benefit unassignable.

  • Pitfall — skipping laboratory testing: Because the observed organ signal and the claimed hormone signal appear at the same doses, going without baseline and follow-up panels removes the only way to tell which one is happening.

  • Regulatory status: Sold in the United States as a dietary supplement under DSHEA (the 1994 law letting supplements reach market without pre-approval), with no approved indication. The European Union has restricted the hydroxyanthracene compounds the plant contains.

  • Cost and accessibility: Neither exceptional. Bulk extract and capsules are widely available online at modest cost; the practical constraint is verified quality, not price or supply.

Interaction with Foundational Habits

  • Sleep: Direction indirect and possibly adverse. No sleep study exists for this plant; scattered user reports describe both improved sleep and palpitations. Sleep restriction itself suppresses testosterone more reliably than any botanical raises it, so sleep debt would mask any effect the herb had.

  • Nutrition: Direction strengthening for absorption, uncertain for outcome. Anthraquinones are fat-soluble, so a meal containing fat plausibly improves uptake, as the human trial’s capsule schedule implicitly allowed. Adequate dietary fat and zinc are prerequisites for testosterone synthesis; the herb does not substitute for either.

  • Exercise: Direction indirect. No trial has combined this plant with resistance training, and the rodent work involved no exercise. Resistance training raises testosterone acutely and improves body composition durably, which is the confounder that would make any self-experiment uninterpretable without controlled conditions.

  • Stress management: Direction potentially blunting. Chronic cortisol elevation suppresses the same brain-to-testis loop the herb is proposed to stimulate, so unmanaged stress works directly against the claimed mechanism. No cortisol measurement has been reported in any study of this plant.

Monitoring Protocol & Defining Success

Baseline testing precedes the first dose and serves two separate purposes: establishing whether an androgenic response is even detectable, and establishing the liver and kidney reference against which the rodent organ signal could be recognised. A morning draw between 8 and 10 a.m., fasted, covers the hormone panel, a comprehensive metabolic panel with gamma-glutamyl transferase, a lipid panel and a complete blood count.

Ongoing monitoring follows a cadence of 4 weeks, then 12 weeks, then every 6 months if use continues. The 4-week point matches the only human exposure ever studied and is where drug-induced liver injury usually declares itself; the 12-week point is the earliest at which a durable hormone change would be credible. Either liver enzyme above three times the upper reference limit ends the protocol rather than prompting a recheck.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Total testosterone 600–900 ng/dL The endpoint the herb is taken for Conventional range is far wider at 264–916 ng/dL; draw 8–10 a.m., two separate mornings
Free testosterone 15–25 ng/dL The biologically available fraction Calculated free testosterone from total and binding protein is more reliable than direct assay
SHBG 20–40 nmol/L Determines how much testosterone is usable SHBG is sex hormone-binding globulin, the protein binding testosterone in blood; rises with age and alcohol
LH 2–6 IU/L Distinguishes a pituitary-driven rise from a testicular one Pair with total testosterone in the same draw; high LH with low testosterone points to testicular failure
Estradiol (sensitive assay) 20–30 pg/mL Detects the oestrogen fall seen in rodents Request the liquid chromatography-mass spectrometry assay; standard immunoassay is unreliable in men
ALT Under 25 U/L Earliest marker of the hepatic signal Conventional upper limits of 40–55 U/L are too permissive to catch early injury; fasting not required
AST Under 25 U/L Confirms a hepatic rather than muscle source Interpret alongside creatine kinase (an enzyme leaking from strained muscle); hard training raises AST independently
ALP and total bilirubin ALP 45–90 U/L; bilirubin 0.3–1.0 mg/dL The pair that moved in the human trial and in rats ALP rose slightly on active supplement in the only human study; a bilirubin rise alongside enzymes is the serious pattern
eGFR and creatinine eGFR above 90 mL/min/1.73 m² Tracks the renal half of the rodent signal eGFR is estimated glomerular filtration rate, a calculated measure of kidney filtering; creatinine fell slightly on active supplement
Lipid panel HDL above 50 mg/dL; triglycerides under 100 mg/dL Tracks the adverse lipid shift seen in rats 12-hour fast; conventional HDL threshold of 40 mg/dL is less demanding
CBC with differential Within laboratory reference range Tracks the white cell and platelet movement seen in rats CBC is complete blood count; also catches the rise in red cell mass that androgen elevation can cause
PSA Under 1.0 ng/mL below age 50 Prostate safety under any androgenic agent PSA is prostate-specific antigen; avoid drawing within 48 hours of cycling or ejaculation

Qualitative markers worth tracking alongside the laboratory panel:

  • Morning erection frequency, recorded weekly
  • Libido, rated on a fixed 1–10 scale at the same time each week
  • Training performance: load moved on two fixed compound lifts
  • Energy and motivation through the afternoon
  • Sleep quality and any new palpitations
  • Skin changes, particularly new back or shoulder acne
  • Mood and irritability, ideally noted by someone else as well

Emerging Research

  • No registered trial of the plant: A ClinicalTrials.gov search for “Bulbine” returns zero studies of any status. No NCT identifier exists for this herb, so nothing is currently under way that would test the testosterone claim in people.

  • Microbiome-targeted nutrition for low testosterone: NCT07761546 is enrolling 112 men with functional secondary hypogonadism by invitation. It would strengthen the case for non-hormonal nutritional routes to the same endpoint this herb targets.

  • Adaptogen supplementation and strength: NCT07803068 is recruiting 40 physically active men for 12 weeks of botanical supplementation with resistance training, measuring strength and stress-response proteins. A null result would weaken the wider botanical-androgen case.

  • Adaptogens alongside protein in active males: NCT07716904 is recruiting 60 participants to test adaptogenic botanicals with whey protein on recovery and strength, an adjacent design that could inform whether this class produces measurable performance change.

  • Product-composition analysis: Work characterising marker compounds across commercial supplements (Avula et al., 2021) could weaken the case further by showing how little authentic plant material sold products contain.

  • Translating the interaction signal: Cell-level induction of CYP1A2, CYP2B6 and the efflux pump (Husain et al., 2022) needs human probe-drug studies to settle whether the enzyme effect is clinically real.

  • Non-hormonal pharmacology: A critical review of wound-healing mechanisms across Bulbine species (Voko et al., 2025) could strengthen the case for the plant’s topical uses while leaving the hormonal claim untouched.

Conclusion

Bulbine natalensis is a southern African succulent with a long traditional record in male sexual health, now sold internationally as an oral supplement said to raise testosterone. That claim rests almost entirely on rodent work from a single university group: low doses raised male hormone levels and sexual activity in rats, while a higher dose reversed the effect. Nothing comparable has been measured in people. The one human study was short, small and designed only to check safety chemistry; it was part-funded by a supplement-industry sponsor, and no independent human work has followed it.

Set against that thin benefit record are findings pointing the other way. In the same rodent experiments, liver and kidney chemistry and tissue structure shifted at the doses that raised hormone levels, and blood fats moved unfavourably. Laboratory work shows the plant strongly switches on the systems that clear many prescription medicines, so interactions are plausible even where none has been documented in people. Chemical analyses of products on sale found several containing none of the plant’s marker compounds at all, and a wide spread among those that had them.

What remains is a traditional remedy with an interesting animal signal, an entirely unmeasured human effect, an unresolved question about organ safety, and a supply chain that does not reliably deliver what the label states.

Top - Benefits - Risks - Protocol